Lege a Me Me tia Cae aL eee ee eT ey vatit ep i ha Sl PROS EST ae ate AS aN ill tae tee Fie teed © ON re nae ae a ae Ba nuk of Pa weak Pedals ete et cae ig : Sam tty eed Ae R OIA LE eta lesa mip im ES woe Cone wee ewteet, Bete “ ar aenlice BME Me On twine ey WPVM ee ho oS elegy nee eNews Ma de Sp Oe Haw Poa or Bae Mau drat are adens MOOS Gt at ale AUP TIS ets ‘ Date Palen, compa tues rode ater reese ee ed Pc letet Toten Rowe Pa au eeye Meese dete aia RE pa eS ae Nes Mee a Pal ees tea pa e ethearaeelaleistatg! ae wth a PD, wee RRA nd a athe ad alin, Sea ee Eo Le ae oe eens Eee ee vera Se Mars S eRe aad aa hata seers fet verge tanereyi Tat AE eas alles ay ete a eee 2r eae Tea Pheer ett Rlowtiana cep ie hy Mig ia sty BaP ges kid g Suet EI KE IED fy GaP asa a ness a a rr ee Media TERS a Rent tae sihdataced rele Pera Eee Re TS eas ee ee rae eo ee ee eee AVENE are ee are Mega dt ce BUNS ANGE fh eae a Dees str ets aa re ean ens heer as LeReNedery deren Poser tary: eNat Fd eS ERO age ok EY arioeee . wake VN MISES ANG Ved ye at Wet Tae ewer FePetate apneic ys Veta tls ay ie can Pees sigan a satan, Boaeseat . ! ‘ : . is oe aMas Bore te . Ree Caer ee Saye chants uy UNDE Spare eH PHS seed ey Ge Legh oy Sova eye egy Ss ene vate ene are pe esee sey pene es” ¥ ee rer sae daewer tetera tytn sy nee i Fp A iwey : Auge mpaerg yy ernG Tyeeryeres at - . er Oy areas wey PURE N tea wre ete Nea Tete tet Pare get nrg eied season dR ate fpedarytee Pee Ne te ener rae ros. arene woe ethene SVM Na tata a rte SON TRE Ud ewe ee nd te tet SW athe AAS A Wah gk sa Sa a EEA Legs a, SERIES AHO LOANS one et oe ae oe toe Powyeny ean aM NEN ee EUR UNS Ades SRE at eV ad eri ts VQe a aN aes, 7 Times ee uh dbo ARN Ss fe Sha BY ML LVRS RY St AQ eA, PNPERD EN Sy Oh eee ce Det EE TY POTN UU EN at eT ek ere re SC Pet aL eeneOraey Wr erved ds Poe ror err eee AA cet PRPS EN bette oe ONT eae Serves WERT Che Sa ad eth ga ti Pam Re Nae PRY TR TRE wey cys, eM TMD evan zg Ee EEUU Cee ba eet Poe Se Rens ee eee STEM ENE Nabe Se tee de ety Ve ed TN eS We en oe DENN CPTO NSB ee dete Pert ae tie ees at Ste Pether we emacs, AVES Geue NS Draven eee Babcee HYNES US Se URE Caer ae ba et eWartan ag Fosse A ete Seats ate aw areas Pare nes an Seer er ee Ses ok as Le VETOES bee UP le resets Om igen he se bueg BNL eral HOME Beattie tate SYN etre gee prone EOE?’ re ae 7 ‘ a 2 Prd 7 ' - v = : a yy es 7 ed . + = a ee a — = -_ ' rcs ' “— > MADRONO A WEST AMERICAN JOURNAL OF BOTANY VOLUME XLVI 2000 BOARD OF EDITORS Class of: 2001—ROBERT PATTERSON, San Francisco State University, San Francisco, CA PAULA M. SCHIFFMAN, California State University, Northridge, CA 2002—-NoRMAN ELLSTRAND, University of California, Riverside, CA CARLA M. D’ ANTONIO, University of California, Berkeley, CA 2003—FREDERICK ZECHMAN, California State University, Fresno, CA JON E. KEELEY, U.S. Geological Service, Biological Resources Division, Three Rivers, CA 2004—Davip Woop, California State University, Chico, CA INGRID PARKER, University of California, Santa Cruz, CA Editor—KRISTINA A. SCHIERENBECK California State University, Chico Department of Biology Chico, CA 95427-0515 kschierenbeck @csuchico.edu Published quarterly by the California Botanical Society, Inc. Life Sciences Building, University of California, Berkeley 94720 Printed by Allen Press, Inc., Lawrence, KS 66044 MApbRONO, Vol. 47, No. 4, pp. 11-111, 2000 TABLE OF CONTENTS Alverson, Edward R. (see Zika, Peter E, Edward R. Alverson and Loverna Wilson) Baldwin, Bruce G,,, President's report for Volume 47 =. _...2 eto oe ee ee Baldwin, Bruce G., Roles for modern plant systematics in discovery and conservation of fine-scale biodiversity Baldwin, Bruce G. (see also Mishler, Brent D., et al.) Barkley, Theodore M., Plomstc studies in contemporary botany 2.2022 ee ee Barron, Robin (see Corbin, Beth Lowe) Beidléman. Richard, Willis Linn Jepson—“The botamy man” 222.2122. 4b ee ee Beidleman, Richard (see also Jepson, Willis Linn) Bermejo- Velazquez, Basilio (see Ledig, E Thomas, et al.) Boyd, Robert S., Michael A. Wall, and James E. Watkins, Jr., Correspondence between Ni tolerance and hyperaccumulation in Strepianthus (Brassicaceae) ).__-- sa ee Boyd, Steve (see Provance, Mitchell C., et al.) Boyd, Steve (see also Soza, Valerie, et al.) Bramlet, David (see Provance, Mitchell C., et al.) Brooks, Matthew L., Review of A Natural History of the Sonoran Desert ed. by S. J. Phillips and P. W. Comus Bruederle, Leo P. (see Kuchel, Shannon D.) Buckmann, Allan (see Hrusa, G. Frederic, and Allan Buckmann) Ceska, Adolf (see Christy, John A.) Ceska, Oldriska (see Christy, John A.) Charlet, David A., Coupling species-level inventories with vegetation mapping — Christy, John A., Oldriska Ceska and Adolf Ceska, Noteworthy collections from Oregon and Washington Clark, Dina, and Tim Hogan, Notewothy collections from Colorado Capo-Arteaga, Miguel A. (see Ledig, E Thomas, et al.) Coleman, Konald A. Noteworthy collection from Anizona...°< 2 2.26 ee eee Constantine-Shull, Helen, and John O. Sawyer, Noteworthy collections from California Corbin, Beth Lowe, James L. Reveal and Robin Barron, Eriogonum spectabile (Polygonaceae): A new species MOM MOTUS aS ter CaO A tn IIa a te ee del Moral, Roger (see Wood, David M.) Dodd, Shana C., and Kaius Helenurm, Floral variation in Delphinium variegatum (Ranunculaceae) Ertter, Barbara, Our undiscovered heritage: Past and future prospects for species-level botanical inventory Ertter, Barbara (see also Jepson, Willis Linn) Ertter, Barbara (see also Mishler, Brent D., et al.) Ewing, Kern, Environmental gradients and vegetation structure on south Texas coastal clay dunes Felger, Richard, Noteworthy collections from Arizona and Sonora, Mexico 0 Felger, Richard Stephen, Noteworthy collections from Sonora, Mexico 2000.20.02 22222o- see eeen nnn e nnn cnn cecceneeee eee eeeceee ee =ee Flores-L6pez, Celestino (see Ledig, EF Thomas, et al.) Hart, Jefferey A. (see Hrusa, G. Frederic, and Jefferey A. Hart) Hartman, Ronald L., Review of Synthesis of the North American Flora. Version 1.0 by John T. Kartesz and CGINETS TO pe VS aia eer a Oe Helenurm, Kaius (see Dodd, Shana C.) Hipkins, Valerie D. (see Safiya, Samman) Hobbs, Richard J., Review of 2nd Interface Between Ecology and Land Development in California edited by J. EE aisccley. ivi. iBacr-Keeley. and ©.J. Fothernmomai 3. ee eee Hohday:susan.-A foristic study of Tsesi Canyons Arizomay 2.) 2s sae ae Hogan, Tim (see Clark, Dina) Hrusa, G. Prederic, Neteworthy collection from Califormmia ...... —.... 2 Hrusa, G. Frederic, and Allan Buckmann, Noteworthy collection from California oes Hrusa, G. Frederic, and Jeffrey A. Hart, Noteworthy collection from California... --_--_-e een Jepson, Willis Linn, Richard Beidleman and Barbara Ertter, Willis Linn Jepson’s “‘Mapping in forest botany” Knops, Johannes M. H., and Walter D. Koenig, Annual variation in xylem water potential in California oaks _. Koenig, Walter D. (see Knops, Johannes M. H.) Kuchel, Shannon D., and Leo P. Bruederle, Allozyme data support a Eurasian origin for Carex viridula subsp. VER ELTE LC GCL ACCC) ee ag ee cee Mie rg ee eee Ns ee Reece cee Kuykendall, Keli (see Zika, Peter EK, Keli Kuykendll and Barbara Wilson) Ledig, EF Thomas, et al., Locations of endangered spruce populations in México and the demography of Picea CN GATT CGN ATELG A (1 6 a a eo a ee nt a eee EMSS eee SRS A ee RIUM: PNT NEE PRN oy NT Mapula-Larreta, Manuel (see Ledig, EK Thomas, et al.) Markos, Staci (see Mishler, Brent D., et al.) 287 219 253 | 273m oTm 68 259 212 142 138 209 134 116 239, 10 211 2d 207 206 29 139 138 138 269 106 147 a 52) 2000] TABLE OF CONTENTS Mayer, Michael S., Laura M. Williams, and Jon P. Rebman, Molecular evidence for the hybrid origin of Opuntia PTO Crs (ACTACC AC) Meare oe Nee nee rp oe a ne Ne on eet eae Meyers-Rice, Barry A., Ramona Robison and John M. Randall, Noteworthy collections from California Mishler, Brent D., The need for integrated studies of the California flora Mishler, Brent D., et al., Introduction to The Jepson Herbarium 50th anniversary celebration and scientific symposium: Discovery, communication, and conservation of plant biodiversity in California 0... Moe, Richard, Electronic activities of the University and Jepson Herbaria Odion, Dennis, Seed banks of long-unburned stands of maritime chaparral: Composition, germination behavior, ATM LS ULL WA Veal VV ICED AAC eri ho SO wa ee oA Se a nee dee eee ee as SEG Oe ee Parnell, Dennis R. (see Saroyan, J. Phillip) Parsons, Lorraine S., and Adam W. Whelchel, The effect of climatic variability on growth, reproduction, and population viability of a sensitive salt marsh plant species, Lasthenia glabrata subsp. coulteri (Asteraceae) Preston, Robert E., Noteworthy collections from California Provance, Mitchell C., et al., Noteworthy collections from California Provance, Mitchell C. (see also Soza, Valerie, et al.) Randall, John M. (see Meyers-Rice, Barry A.) Rebman, Jon P. (see Mayer, Michael S.) Reveal, James L. (see Corbin, Beth Lowe) Reyes-Hernandez, Valentin (see Ledig, EK Thomas, et al.) Robison, Ramona (see Meyers-Rice, Barry A.) Safiya, Samman, Barbara L. Wilson and Valerie D. Hipkins, Genetic variation in Pinus ponderosa, Purshia tridentata, and Festuca idahoensis, commuity-dominant plants of California’s yellow pine forest Sanders, Andrew C. (see Provance, Mitchell C., et al.) Sanders, Andrew C. (see also Soza, Valerie, et al.) Saroyan, J. Phillip, Dennis R. Parnell, and John L. Strother, Revision of Corethrogyne (Compositae: Astereae) _ Sawyer, John O. (see Constantine-Shull, Helen) Schierenbeck, Kristina A., Editor’s report for Volume 47 Soza, Valerie, et al. Noteworthy collections from California 22222 Soza, Valerie (see also Provance, Mitchell C., et al.) Spickler, James C. (see Stillett, Stephen C.) Stephens, Scott L., Mixed conifer and red fir forest structure and uses in 1899 from the central and northern SilercayNen acre © alltG OU aces rere sere a ee eee ee Stephenson, Nathan L., Estimated ages of some large giant sequoias: General Sherman keeps getting younger Stillett, Stephen C., James C. Spickler and Robert Van Pelt, Crown structure of the world’s second largest tree _ Strother, John L., Hedosyne (Compositae, Ambrosiinae), a new genus for Iva ambrosiifolia Strother, John L. (see Saroyan, J. Phillip) Stuart, John D. (see Mahony, Thomas M.) Tiehm, Arnold, The taxonomic history, identity, and distribution of the Nevada endemic, Plagiobothrys PLOVNCTQIUES (CB OFAC TINAC CAG) © 0h SNe els Did sas Baer A ra a hea fac aclu gw eee Van Pelt, Robert (see Stillett, Stephen C.) Wall, Michael A. (see Boyd, Robert S.) Watkins, James E. (see Boyd, Robert S.) Wells, Philip V., Pleistocene macrofossil records of four-needled pinyon or juniper encinal in the northern Wiscaimo Deserts saya. GC alnrormic: el IN Otte cacao fa ee Whelchel, Adam W. (see Parsons, Lorraine S.) Williams, Laura M. (see Mayer, Michael S.) Wilson, Barbara (see Zika, Peter EK, Keli Kuykendll and Barbara Wilson) Wilson, Barbara L. (see Safiya, Samman) Wilson, Loverna (see Zika, Peter EF, Edward R. Alverson and Loverna Wilson) Windham, Michael D., Chromosome counts and taxonomic notes on Draba (Brassicaceae) of the Intermountain WeSta cls Witalean wa iniitiy oes oe cere a eee pa ae Ue ieee = wee CI Baa Ueyolhs sc. | sue ee ete Cue nee RO Wolf, Adrian L. (see Provance, Mitchell C., et al.) Wood, David M., and Roger del Moral, Seed rain during early primary succession on Mount St. Helens, RUNS) OVD ESS 10) Ui Repel a RP Oe SO ere oC eS RG Enterta er etre cae ncaa Pe Perey GRE SET oe TE OO peers =k Pe Zika, Peter F, Edward R. Alverson and Loverna Wilson, Noteworthy collections from Oregon and Washington __ Zika, Peter EK, Keli Kuykendll and Barbara Wilson, Noteworthy collections from Oregon DATES OF PUBLICATION OF MADRONO, VOLUME 47 Number 1, pages 1—70, published 8 December 2000 Number 2, pages 71—146, published 5 March 2001 Number 3, pages 147-216, published 25 June 2001 Number 4, pages 217-300, published 31 August 2001 ill 109 209 230 ZAG 265 195 174 138 hs) 164 89 hey 189 Statement of Ownership, Management, and Circulation (Required by 39 USC 3685) 2. Publication Number 3. Filing Date 5. Number of Issues Published Annually |6. Annual Subscription Price $27.00 Contact Person Roy Buck Telephone 510-848-4169 sy UNITED STATES B POSTAL SERVICE Madrono 4. Issue Frequency Quarterly 7. Complete Mailing Address of Known Office of Publication (Not printer) (Street, city, county, state, and Zip+4) California Botanical Society, Inc.; Herbaria, Life Sciences Building University of California; Berkeley, CA 94720 8. Complete Mailing Address of Headquarters or General Business Office of Publisher (Not printer) California Botanical Society, Inc.; Herbaria, Life Sciences Building University of California; Berkeley, CA 94720 9. Full names and Complete Mailing Addresses of Publisher, Editor, and Managing Editor (Do not leave blank) Publisher (Name and complete mailing address) California Botanical Society, Inc.; Herbaria, Life Sciences Building University of California; Berkeley, CA 94720 Editor (Name and complete mailing address) Kristina A. Schierenbeck; California State University, Chico; Dept. of Biology; Chico, CA 95429-0515 Managing Editor (Name and complete mailing address) 10. Owner (Do not leave blank. If the publication is owned by a corporation, give the name and address of the corporation immediately followed by the names and addresses of all stockholders owning or holding 1 percent or more of the total amount of stock. If not owned by a corporation, give the names and addresses of the individual owners. If owned by a partnership or other unincorporated firm, give its name and address as well as those of each individual owner. Hf the publicaiton is published by a nonprofit organization, give its name and address.) Full Name ‘Complete Mailing Address 5 A 5 es Herbana, Life Sciences Building California Botanical Society, Inc. University of California Berkeley, CA 94720 11, Known Bondholders, Mortgages, and Other Security Holders Owning or Holding 1 Percent or More of Total Amount of Bonds, Mortgages, or Other Secunties, If none, check box Full Name None ‘Complete Mailing Address 12. Tax Status (For completion by nonprofit organizations authorized to mail at special rates) (Check one) The purpose, function, and nonprofit status of this organization and the exempt status for federal income tax purposes: Has Not Changed During Preceding 12 Months oO Has Changed Dung Preceding 12 Months (Publisher must submit explanation of change with this statement) PS Form 3526, September 1998 (See Instructions on Reverse) Computerized Facsimile 13. Publication Title 14, Issue Date for Circulation Data Below Madrono April - June 2000 15, Actual No. Copies of Single Issue Published Nearest to Filing Date Average No. Copies Each Issue Extent and Nature of Circulation During Preceding 12 Months a. Total Number of Copies (Nef press run) 0 Paid/Requested Outside- County Mail Subscriptions stated on Form 3541. (Include advertiser's proof and 0 exchange copies, Paid In-County Subscriptions (Include advertiser's 0 b. Paid and/or proof and exchange copies) Requested Circulation Sales Through Dealers and Camers, Street Vendors, 0 Counter Sales, and Other Non-USPS Paid Distribution (4) ]Other Classes Mailed Through the USPS 0 ¢. Total Paid and/or Requested Circulation (Sum of 15b(1) , (2),(3), and (4)) > (0) d. Free Outside-County as Stated on Form 3541 Distribution by Mail (Samples,complim In-County as Stated on Form 3541 entary, and other free) Other Classes Mailed Through the USPS e. Free Distribution Outside the Mall (Carriers or other means) f. Total Free Distribution (Sum of 15d and 158) Viv g. Total Distribution (Sum of 15¢ and 151) h. Copies not Distributed |. Total (Sum of 15g. and h.) Vv Percent Paid and/or Requested Circulation 15c divided by 15q times 100) 16. Publication of Statement of Ownership 8 a ° o 2 Publication not required. | certify that all information furnished on this form is true and complete. | understand that anyone who furnishes false or misleading information on this or who omits material or information requested on the form may be subject to criminal sanctions (including fines and imprisonment) and/or civil sanctions (including civil penalties). Instructions to Publishers 1. Complete and file one copy of this form with your postmaster annually on or before October 1. Keep a copy of the completed form for your records. 2. In Cases where the stockholder or security holder is a trustee, include in items 10 and 11 the name of the person or corporation for whom the trustee is acting. Also include the names and addresses of individuals who are stockholders who own or hold 1 percent or more of the total amount of bonds, mortgages, or other securities of the publishing corporation. In item 11, if none, check the box. Use blank sheets if more space is required 3. Be sure to furnish all circulation information called for in item 15. Free circulation must be shown in items 15d, e, and f. 4. Item 15h., Copies not Distributed, must include (1) newsstand copies originally stated on Form 3541, and returned to the publisher, (2) estimated retums from news agents, and (3), copies for office use, leftovers, spoiled, and all other copies not distributed. 5. If the publication had Periodicals authorization as a general or requester publication, this Statement of Ownership, Management, and Circulation must be published; it must be printed in any issue in October or, if the publication is not published during October, the first issue printed after October. In item 16, indicate the date of the issue in which this Statement of Ownership will be published. Item 17 must be signed NQ Failure to file or publish a statement of ownership may lead to suspension of second-class authonzation. PS Form 3526, September 1998(Reverse) Computerized Facsimile OOK REVIEW NOUNCEMENT VOLUME 47, NUMBER | JANUARY-MARCH 2000 MADRONO A WEST AMERICAN JOURNAL OF BOTANY SEED RAIN DuRING EARLY PRIMARY SUCCESSION ON Mount St. HELENS, WASHINGTON David M. Wood and Roger del Moral ....cccccccccccccccccccccccccccccccceseesceseesesseees l ENVIRONMENTAL GRADIENTS AND VEGETATION STRUCTURE ON SOUTH TEXAS COAST- AL CLay DUNES ICT TUE WAIT ss cet retiectce sete narens ten ee eck hee eau te Ocenia ene tee ce eee 10 CHROMOSOME COUNTS AND TAXONOMIC NOTES ON DRABA (BRASSICACEAE) OF THE INTERMOUNTAIN WEST. 1: UTAH AND ARIZONA WV CHGCT AD | WAN GHGI seston ccsver ides ei ssgn eco Gee lane 21 A Fioristic Stupy OF TSEGI CANYON, ARIZONA SUS GIT OLIC CN aaa catctss sera aaleta tes shee os c2 eRe NRE tons Sane see Mee eR EEE 29 MIXED CONIFER AND RED FiR FOREST STRUCTURE AND USES IN 1899 FROM THE CENTRAL AND NORTHERN SIERRA NEVADA, CALIFORNIA SCO L. StCPNENS sasesecsascedes ccccgenensas Sh OE Nott) vison Ge PSA oss aed: 43 OLD-GROWTH ForREST ASSOCIATIONS IN THE NORTHERN RANGE OF COASTAL REDWOOD Thomas M. Mahony and-JOhan) Stuart visi icek hoy ccccssecensccccnecseceess 38) ESTIMATED AGES OF SOME LARGE GIANT SEQUOIAS: GENERAL SHERMAN KEEPS GETTING YOUNGER Nathan LESTCPHENS ON as Bol EG SAG es PS Sc eek a vos Sentinel ENTS 61 A NATURAL HISTORY OF THE SONORAN DESERT. EDITED BY S. J. PHILLIPS AND P. W. ComMuUS MGEECWAZBYOOR SAB oo BEERS SN ncn SA oiceccecacccteencees 68 ANNOUNCEMENT 5.2.27 i oca cc MER Sea iccccccve We cca ceeccadeneaenacess 70 PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL SOCIETY Ue Oe masts 9 feet Noe Bas MAproNno ( ISSN002229637) is published quarterly by the California Botanical Society, Inc., and is issued from the office of the Society, Herbaria, Life Sciences Building, University of California, Berkeley, CA 94720. Subscription information on inside back cover. Established 1916. Periodicals postage paid at Berkeley, CA, and additional mailing offices. Return requested. Postmaster: Send address changes to MADRONO, Roy Buck, % University Herbarium, University of California, Berkeley, CA 94720. Editor—KrisTINA A. SCHIERENBECK California State University, Chico Department of Biology Chico, CA 95429-0515 kschierenbeck @csuchico.edu Editorial Assistant—Davip T. PARKS Book Editor—Jon E. KEELEY Noteworthy Collections Editors—DIETER WILKEN, MARGRIET WETHERWAX Board of Editors Class of: 2000—Pame_a S. Sotis, Washington State University, Pullman, WA JOHN CALLAWAY, University of San Francisco, San Francisco, CA 2001—RobeErt PATTERSON, San Francisco State University, San Francisco, CA PauLa M. ScHIFFMAN, California State University, Northridge, CA 2002—-NorMAN ELLSTRAND, University of California, Riverside, CA Cara M. D’ Antonio, University of California, Berkeley, CA 2003—FREDERICK ZECHMAN, California State University, Fresno, CA Jon E. KEELEY, U.S. Geological Service, Biological Resources Division, Three Rivers, CA 2004—Davip M. Woop, California State University, Chico, CA INGRID PARKER, University of California, Santa Cruz, CA CALIFORNIA BOTANICAL SOCIETY, INC. OFFICERS FOR 2000—2001 President: BRUCE BALDwIn, Jepson Herbarium and Dept. of Integrative Biology, 1001 Valley Life Sciences Bldg. #2465, University of California, Berkeley, CA 94720. First Vice President: Rop Myatt, San José State University, Dept. of Biol. Sciences, One Washington Square, San José, CA 95192. rmyatt@email.sjsu.edu Second Vice President: Ros ScHLuIsING, California State University, Chico, Dept. of Biol. Sciences, Chico, CA 95424. rschlising @csuchico.edu Recording Secretary: DEAN KELcu, Jepson and University Herbarium, University of California, Berkeley, CA 94720. dkelch @sscl.berkeley.edu Corresponding Secretary: SUSAN BAINBRIDGE, Jepson Herbarium, University of California, Berkeley, CA 94720. suebain @ SSCL.berkeley.edu Treasurer: Roy Buck, % University Herbarium, University of California, Berkeley, CA 94720. The Council of the California Botanical Society comprises the officers listed above plus the immediate Past President, R. JoHNn Litre, Sycamore Environmental Consultants, 6355 Riverside Blvd., Suite C, Sacramento, CA 95831; the Editor of MApRONO; three elected Council Members: BiAN Tan, Strybing Arboretum, Golden Gate Park, San Francisco, CA 94122; James SHEvocK, USDI National Park Service, Pacific West Region, 600 Harrison Street, Suite 600, San Francisco, CA 94107; DIANE ELam, U.S. Fish and Wildlife Service, 3310 El Camino Avenue, Sacramento, CA 95825; Graduate Student Representative: KIRSTEN JoHANUS, Jepson Herbarium, University of California, Berkeley, CA 94720. This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). MaApbrONO, Vol. 47, No. 1, pp. 1-9, 2000 a, voice SEED RAIN DURING EARLY PRIMARY SUCCE HELENS, WASHINGTON Davip M. Woop Department of Biological Sciences 515, California State University, Chico, CA 95929 ROGER DEL MORAL Department of Botany Box 355325, University of Washington, Seattle, WA 98195 ABSTRACT Seed rain into sites undergoing primary succession on Mount St. Helens was measured from 1982 to 1986 and again from 1989 to 1990. Study sites were devastated in 1980 by pyroclastic flows of pumice, searing blasts, and lahars. Most sites were several km or more from seed sources. Seed rain density averaged 34 seeds 0.1 m~? yr™! in mid-elevation barren sites, 1083 seeds 0.1 m~? yr™! in mid-elevation vegetated sites and 2 seeds 0.1 m~ yr“! at subalpine barren sites. A total of 33 species was collected in traps. The relative abundance distributions of species were generally similar across years and sites. A few wind-dispersed species accounted for most of the seed rain: Anaphalis margaritacea (L.) Benth. & Hook., Epilobium angustifolium L., E. watsonii Barbey (E. ciliatum Raf.), Hieracium albiflorum Hook., and Hypochaeris radicata L. Seeds of trees and shrubs were virtually absent. The common species in the seed rain were also the most common species in the vegetation, although their absolute abundance is determined by environmental factors. Many uncommon species occurred in the vegetation that were not recorded in the seed rain. Two taxa common in the vegetation, Lupinus lepidus Douglas and Salix spp., were rare in the seed rain. For Salix, this is because seed dispersal occurred before traps were in place for the season. Lupinus lepidus is not wind dispersed and seeds are not likely to enter traps. We conclude that the seed rain on Mount St. Helens is apparently sufficient to initiate colonization but is depauperate in species. At present the vegetation generally reflects the incoming seed rain. One cause of succession is differential species availability at a site after a disturbance (Pickett et al. 1987). For vascular plants this differential avail- ability occurs mostly by vegetative regrowth, seed banks, or seed dispersal. In primary succession col- onization results mainly from seed dispersal. On large scale primary successional landscapes the in- put or “‘rain’”’ of seeds from long-distance dispersal is the main source for the establishment of most species, because seed banks and regrowth are ab- sent. A complete interpretation of primary succes- sion in particular, and community assembly in gen- eral, must therefore include measurements of the density and species composition of the seed rain along with an assessment of environmental and substrate conditions (Wood and del Moral 1987; del Moral 1993; Chapin et al. 1994; Booth and Larson 1998; Dlugosch and del Moral 1999). For example, the absence of a species from a particular site or seral stage could be due as much to its absence from the seed rain as to its inability to establish. Conversely, high abundance of a colonist may be explained as much by its abundant seed rain as by its environmental tolerance, growth rate, or com- petitive ability. Hypothesized mechanisms of suc- cession such as facilitation (Connell and Slatyer 1977; Morris and Wood 1989) must also consider differential species availability through the seed rain. Studies of seed dispersal generally are of two types: studies of individual species with the parent plant and its seed shadow as the focus (reviews in Harper 1977; Willson 1993), or studies of the long- distance seed rain into sites where specific seed sources cannot be identified precisely. Seed rain measurements are most appropriate for studies of community assembly in primary succession, but published studies are few (Ryvarden 1971; St6cklin and Baéumler 1996; Archibold 1980; Jefferson and Usher 1989; Chapin et al. 1994). This paper de- scribes the density and species composition of the seed rain in several contrasting regions and habitats undergoing primary succession on Mount St. Hel- ens, WA. The rate of vegetation recovery and plant species composition at various sites have been described for Mount St. Helens following the catastrophic eruption in 1980 (del Moral 1983; Wood and del Moral 1987; del Moral and Wood 1988; del Moral 1993; del Moral and Bliss 1993; del Moral and Wood 1993a, b; del Moral et al. 1995; del Moral 1998, del Moral 1999), but detailed species seed rain data have not been reported with the exception of Dale (1989) who sampled lower elevation lahar (mudflow) sites not included in this study. We pose these questions: For a given site, what is the density and species composition of the seed rain? How do islands of established, reproducing vegetation affect 2 MADRONO Seared * Zone North Fork Toutle River South Fork Toutle River N T on erietiomes Fic. 1. the local seed rain? Does the species composition of the seed rain reflect the species composition of the colonizing flora? Are there species present in the seed rain but absent as colonists? Are there spe- cies present as colonists but absent from the seed rain? STUDY AREA The Mount St. Helens volcano is in the Cascade Range of southwestern Washington at 46°12'N, 122°11'W. The catastrophic north-directed eruption of May 18, 1980, produced a variety of impacts including: a debris avalanche; pyroclastic flows, or incandescent flows of gas and pumice; and lahars, flows of water-saturated debris (‘“‘mudflows’’) trig- gered by rapidly melting snow and ice (Lipman and Mullineaux 1981; Decker and Decker 1981). Im- pacts on the vegetation are categorized into four regions: the blast zone, in which most life was de- stroyed; the blowdown zone, in which adult trees were knocked over but some saplings and under- story vegetation survived; the seared zone, in which trees remained standing but had their foliage singed by the hot gases of the blast, and the mud- flow (lahar) zone in which most vegetation was de- stroyed (Fig. 1). In addition to these four regions, a large region south of the crater received 5—20 cm [Vol. 47 WWQ ~ mith Creek Mudady\River Mudflows Location of study areas. PP = Pumice Plain sites, PA = Plains of Abraham sites, BC = Butte Camp sites. See text for details of seared, blowdown, and blast zones. of tephra (airfall ash and pumice) and most vege- tation survived. We studied three main areas separated by several km and at different elevations: Pumice Plain, Plains of Abraham (both in the blast zone), and Butte Camp (Fig. 1). The Pumice Plain is a 20 km?’ region between the crater and Spirit Lake that received the full force of the north-directed lateral eruption, re- ceiving a debris avalanche and pyroclastic flows. These new deposits now overlie what was formerly a montane forest of Tsuga heterophylla (Raf.) Sarg., Pseudotsuga menziesii (Mirbel) Franco, and Abies amabilis (Douglas) James Forbes (Krucke- berg 1987). Elevations range from 1000 to 1200 m. Although much of the Pumice Plain was initially flat to moderately hilly, numerous erosion gullies continue to form and deepen. The Plains of Abra- ham, also a pumice landscape, is located approxi- mately 3 km NE of the crater at 1350 m elevation. It received searing blasts and deposits of tephra, and is predominantly flat with numerous small gul- lies (del Moral and Wood 1993b). The pre-eruption vegetation of this high montane region was de- scribed by Kruckeberg (1987) as a region high in species richness of forbs and grasses but low in cover with scattered conifers. The Butte Camp re- gion is located on the southwest side of the volcano | TABLE 1. _ 2000] WOOD AND MORAL: SEED RAIN ON MOUNT ST. HELENS 3 MEAN + STANDARD ERROR OF SEED RAIN DENSITY (SEEDS 0.1 M-? YR~!) FOR 1983 TO 1986 ESTIMATED FROM _ FALLOUT TRAPS AT MID-ELEVATION BARREN SITES. Number in parentheses is number of traps. In 1983 and 1984, mean values for the Pumice Pond and Spirit Lake sites are significantly different. ~ Pumice Plain Sites 1983 1984 1985 1986 Pumice Pond Ior 93410) 38.4 + 9.6 (5) no data no data Spirit Lake 34.9 + 5.0 (10) 13.0 + .9 (9) 25.02 + 3.0 (10) 26.9 + 4.9 (10) at 1500 to 1600 m. This area of subalpine vegeta- tion was disturbed by several lahars and also re- ceived tephra deposits (del Moral 1983; del Moral and Wood 1986). However, seed rain was only measured at the primary succession lahar sites. METHODS Seed collections. Seeds of vascular plants were collected from both wet pitfall traps and dry fallout traps. Pitfall traps, whose primary purpose was to collect ground-dwelling insects (Edwards 1986), consisted of 10 cm diameter plastic cups filled with ethylene glycol, set flush with the ground surface, and covered by a plywood square elevated 1 cm above the cup with corner nails. Pitfall trap data were collected in 1982, 1983 and 1985. The fallout traps consisted of 33 < 33 cm (0.1 m’*) wooden frames 3 cm high with fine nylon mesh bottoms. Frames were filled with a single layer of used golf balls and set flush with the surface (Edwards 1986; Edwards and Sugg 1993). The golf balls (approxi- mately 3 cm in diameter) were used because they were an easily obtainable uniform sphere that mim- icked the size and surface texture of the pumice. We also wanted baseline density estimates of the seed rain for relatively flat, open ground. These baseline values may then be adjusted upwards if desired for sites of seed accumulation, e.g., against boulders or in gullies and other depressions (Dale 1989). Sticky traps (Werner 1975) or wet pitfall traps could yield over-estimates of density for flat ground in this open, windy environment due to seed accumulation (Johnson and West 1988). Thus, only fallout traps were used to estimate seed rain den- sity. Fallout traps were used in all years of the study except 1982, and were the only type of trap used in 1989 and 1990. Data from both pitfall traps and fallout traps were used to estimate relative abun- dance. Traps were set out in June of each year after snowmelt when the roads to the sites became ac- cessible and contents were collected approximately twice a month until November. Fallout traps were collected only once a year in late October, after the fall dispersal period and before sites became inac- cessible due to snow. Seeds were stored in alcohol or formalin-acetic acid-alcohol (FAA) and were identified using a reference collection obtained from field specimens and herbarium sheets. Al- though seed germination was not measured in this study, only those seeds with morphology and col- oration similar to viable seeds were counted. Ex- tensive seed germination experience with many species from Mount St. Helens suggests that the appearance of viability under a dissecting micro- scope is a good predictor of germination—most species had germination rates from 60 to 90% when abnormal-appearing seeds were excluded (Wood and del Moral 1987; Wood unpublished data). No- menclature follows Hitchcock and Cronquist (1973) with parenthetical updates from Hickman (1993) to correspond to Titus et al. (1998). Study sites. From 1982 to 1986 two sites on the Pumice Plain were sampled, Pumice Pond and Spir- it Lake. The Pumice Pond site was near the head- waters of the North Fork of the Toutle River on the northwest side of the Pumice Plain about 5 km NNW of the crater. Unfortunately, severe erosion at this site forced it to be abandoned in 1985 (Ed- wards and Sugg 1993). The Spirit Lake site was near the eastern edge of the Pumice Plain about 2 km south of Spirit Lake and 3 km north of the crater. Traps of both types (pitfall and fallout) were placed at 10 m intervals along 100 m transects, al- though resultant sample sizes vary because traps occasionally were filled with erosional material or lost (Tables 1 and 3). In 1986 mean percent cover of vegetation on the Pumice Plain in the vicinity of these sites was estimated at 0.09% (Wood and del Moral 1988) and had increased to only 1.4% by 1990 although some small patches exceeded 50% (Wood unpublished data). In 1989 and 1990, the number and placement of fallout traps was increased by including a greater variety of habitats within the Pumice Plain. Pumice Plain sites I and II were established in barren areas (defined as having <3% cover) close to the old Spirit Lake site at 1100 m elevation. Each site con- tained 16 fallout traps arranged in a 4 X 4 grid with traps separated by 10 m (hereafter referred to as a ‘**16-FT grid’’). Due to occasional trap disturbance (e.g., by ravens and elk) resultant sample size again varied (Table 2). The Pumice Ridge site was on an exposed, barren ridge 50 m above the Pumice Plain and contained 5 traps along a 30 m transect. The 16-FT grid Lupine Patch site was in a patch of dense flowering Lupinus lepidus Douglas, (>50% cover) a few hundred m from Pumice Plain I and II. The Willow Spring 16-FT grid was in a rela- tively open, moderately vegetated site (<20% cov- er) but was surrounded by a stand of dense, repro- ductively mature vegetation adjacent to a spring (Wood and del Moral 1988). This vegetation in- cluded Salix spp. (primarily S. sitchensis Bong. and 4 MADRONO [Vol. 47 TABLE 2. MEAN + STANDARD ERROR OF SEED RAIN DENSITY (SEEDS 0.1 M~? YR~!) IN 1989 AND 1990. Number in parentheses is number of traps. Means with the same letter within a year are not significantly different at P = 0.05 by Tukey’s HSD. MEB = mid-elevation barren; MEV = mid-elevation vegetated; HEB = high-elevation barren. Habitat 1989 1990 Butte Camp Sites Lahar I HEB 1.22 = 0.7 (15) no data Lahar II HEB 2.87 + 0.8 (16) no data Plains of Abraham Sites Abraham I MEB 11.8? = 1.7 (16) 5.37: =. 2.8 .(16) Abraham II MEB 9.0’ + 2.0 (5) 66.57 + 12.3 (4) Pumice Plain Sites Pumice Plain I MEB 24.17? © 4.1 (5) 50:67 = $2-(16) Pumice Plain II MEB 21.994 + 3.3 (15) 50.67 + 8.2 (16) Pumice Ridge MEB 68.2°4 + 19.4 (5) 9:68 e235) Lupine Patch MEV 94.1° 10.8 (16) 355.6? + 48.6 (16) Willow Spring MEV 1709.1 + 489.7 (16) 2174.4 + 645.1 (16) S. commutata Bebb), Anaphalis margaritacea (L.) Benth. & Hook., Epilobium angustifolium L., E. watsonii Barbey (E. ciliatum Raf.), Hypochaeris radicata L., and Lupinus lepidus. The Plains of Abraham area contained two sites in barren areas, one 16-FT grid (Abraham I) and one five-trap transect (Abraham II) similar to the Pumice Ridge site described above. Both sites were on nearly level ground and were spaced 200 m apart. In 1989 and 1990, mean percent cover on the Plains of Abraham was estimated at 0.12% and 0.23%, respectively (del Moral and Wood 1993b). The Butte Camp area contained two sites on la- hars (Lahar I and II), both 16-FT grids spaced 200 m apart. Percent cover on the Butte Camp lahars was estimated at 2—3% in 1989 (del Moral 1993). RESULTS Density. Seed rain density varied widely over both sites and years (Tables 1 and 2), from a low of 1.2 seeds 0.1 m~? yr“! at Lahar II in Butte Camp in 1989 to a high of 2174 seeds 0.1 m~ yr™' at Willow Spring on the Pumice Plain in 1990 (Table 2). In 1983 and 1984 the Pumice Pond site received more than twice as many seeds as did the Spirit Lake site (Table 1; t-test, log transformation, P < 0.01 in each year). In both 1989 and 1990, ANOVA revealed a significant difference among the Pumice Plain, Plains of Abraham, and Butte Camp areas as well as significant differences among sites within the Pumice Plain (Table 2; log transformation, P < 0.001 in each year). In both 1989 and 1990, Willow Spring had a significantly greater seed rain density than all other sites (Table 2; Tukey’s HSD multiple comparisons, P = 0.05). Lupine Patch had the sec- ond highest seed rain density in both years, al- though mean density at this site was not signifi- cantly different from Pumice Ridge in 1989 or Abraham II in 1990 (Table 2). Statistical tests were not performed on year-to-year differences within a site due to the lack of clear hypotheses, as variation could be due to unmeasured factors such as differ- ences in wind patterns or growing conditions and seed production in surrounding landscapes. When sites were classified by habitat, the varia- tion in density was reduced and a clearer pattern emerged. Mid-elevation barren sites (Pumice Pond, Spirit Lake, Pumice Plain I and II, Pumice Ridge, TABLE 3. RELATIVE ABUNDANCE OF COMMON SPECIES IN THE SEED RAIN FOR 1982 THROUGH 1986. The Pumice Pond and Spirit Lake sites are combined. See text for additional species. Distributions between years are not significantly different by a Wilcoxon Signed Ranks test. 1982 Anaphalis margaritacea 16 Epilobium angustifolium 26 Epilobium watsonii (E. ciliatum) <] Hypochaeris radicata 3 Hieracium albiflorum 3 Senecio sylvaticus 36 Lupinus lepidus 0) Number of pitfall traps 38 Number of fallout traps 0 Relative Abundance (%) 1983 1984 1985 1986 36 ZI 10 70 8 48 74 13 5 10 3] 5) 2 ) 2) 3 2 3 1 3 39 di 1 i 0 0 0 0 32 0 55) 0 20 14 10 10 2000] Abraham I and II) had an overall mean density of 33.6 seeds 0.1 m~ yr~', ranging from 5.3 at Abra- ham I in 1990 (Table 2) to 75.5 at Pumice Pond in 1983 (Table 1). High-elevation barren sites (Lahar I and II) had a much lower overall mean density of 1.9 seeds 0.1 m~ yr™!. The highest densities were recorded at mid-elevation vegetated sites (Lupine Patch and Willow Spring) where densities ranged from 94.1 at Lupine Patch in 1989 to 2174 at Wil- low Spring in 1990 (Table 2) with an overall mean density of 1083. Relative abundance. The relative abundance of the most common species in the seed rain is pre- sented in Tables 3 and 4. Relative abundance dis- tributions were generally consistent from year to year and from site to site. The most distinctive sites were Lahar I and II, where subalpine species char- acteristic of that habitat appear. However, no com- parison of abundance distributions is significantly different, either among years from 1982 to 1986 (Table 3), between years within a site, or among sites in 1989 and 1990 (Table 4; Wilcoxon Signed Rank Test, all P > 0.5). Unfortunately, separate es- timates of relative abundance for the Pumice Pond and Spirit Lake sites are not available because sam- ple collections from these sites were combined after counting the total number of seeds in a given trap. Six species accounted for 85% of the measured seed rain at the two Pumice Plain sites in 1982 and >90% from 1983 to 1986: Anaphalis margaritacea, Epilobium angustifolium, E. watsonii (E. ciliatum), Hypochaeris radicata, Hieracium albiflorum Hook., and Senecio sylvaticus L. (Table 3). Six species also accounted for >90% of the measured seed rain at all sites in 1989 and 1990 except for the subalpine sites Lahar I and II (Table 4). These were the same Six species listed above except that Lupinus lepidus replaced S. sylvaticus. The decline of S. sylvaticus and the increase of L. lepidus were the most note- worthy changes in relative abundance during this study. S. sylvaticus decreased from 39% relative abundance in 1983 to zero in 1989 and 1990 at all sites except Lahar I and II (Tables 3 and 4). Lupinus lepidus was not recorded from 1982 to 1986 but dur- ing the 1989-1990 sampling period it occurred at all sites except Abraham II at least once. A total of 33 species was collected, including two unidentified grasses (one seed each). Species not listed in Tables 3 or 4, all with three or fewer seeds trapped except as noted, are: Acer circinatum Pursh, Achillea millefolium L., Agoseris grandiflora (Nutt.) E. Greene, Agrostis sp., Antennaria sp., Carex mertensii Prescott, Carex rossii Boott, Carex sp., Cinna latifolid (Goeppert) griseb., Cirsium vul- gare (Savio) Ten., Epilobium luteum Pursh, Juncus parryi Engelm., Lactuca muralis (L.), Fresen. Pen- stemon cardwellii Howell, Salix spp. (12 seeds), Saxifraga ferruginea Graham, Senecio vulgaris L., Sitanion hystrix (Nutt.) J. G. Smith (Elymus ely- TABLE 4. RELATIVE ABUNDANCE OF COMMON SPECIES IN THE SEED RAIN FOR 1989 AND 1990. See text for additional species. All values are %. Distributions between years within a site or among sites are not significantly different by a Wilcoxon Signed Ranks test. Plains of Abraham II °89 11 20 Plains of Abraham I Lahar I Lahar II Willow Spring Lupine Patch °89 Pumice Ridge "89 Pumice Plains I Pumice Plains II oy) oe) 89 90 90 15 90 89 69 89 90 90 °90 °89 °90 "89 29 17 13 17 18 ii) 40 18 42 47 24 54 Anaphalis margaritacea 22 67 39 26 40 Epilobium angustifolium Epilobium watsonii WOOD AND MORAL: SEED RAIN ON MOUNT ST. HELENS oOonwnN 30 95 89 29 36 in 14 17 58 10 38 11 | 58 13 11 Hypochaeris radicata <1 se | 10 Hieracium albiflorum Lupinus lepidus N eo, 1% relative abundance at any site are present in the vegetation, and the most common Species in the seed rain were also the most common species in the vegetation during the study period (Wood and del Moral 1988; del Moral 1993; del Moral and Wood 1993a; see also St6cklin and Baumler 1996). No species with consistent, rela- tively abundant seed rain appeared to be excluded from establishing at least some individuals on Mount St. Helens due to a lack of ecological tol- erance. However, the absolute abundance in the vegetation on Mount St. Helens is determined by a host of other factors in addition to seed rain density including safe-sites for germination (Wood and Morris 1990; del Moral and Wood 1993b; Titus and del Moral 1998) and facilitation (Morris and Wood 1989; del Moral and Wood 1993a). Most of the common species in the seed rain have seeds adapt- ed for wind dispersal: a feathery coma in Epilobium angustifolium and E. watsonii, and pappuses in An- [Vol. 47 aphalis margaritacea, Hypochaeris radicata, Hier- acium albiflorum, and Senecio sylvaticus. The consistency in species composition of the seed rain among both years and sites suggests that the vegetation will also be similar from site to site, with the exception of the subalpine lahar sites. This prediction is upheld for sites of similar elevation except where patches of Lupinus lepidus have de- veloped (del Moral et al. 1995). The species com- position of the seed rain also gives some indication as to seed sources. Seeds of common montane spe- cies such as Anaphalis margaritacea, Epilobium angustifolium, E. watsonii (E. ciliatum), Hypo- chaeris radicata, Hieracium albiflorum, and Sene- cio sylvaticus probably originated in seared and blowdown forest 10—20 km to the west and north of the study areas (Fig. 1) where recovery of these Species occurred relatively rapidly (Halpern et al. 1990). Westerly prevailing winds likely transported these species up the Toutle River valleys to the Pumice Plains sites (Fig. 1). Willson (1993) reports a wide range in maximum dispersal distances of herbaceous species with wind dispersal adaptations, from a few m to >4000 m. The dispersal ability of Epilobium in particular is extraordinary—Solbreck and Andersson (1987) estimated the maximum dis- persal distance of E. angustifolium to be hundreds of km under windy conditions. Seeds of the ruderal species in the seed rain such as Cirsium arvense (L.) Scop., C. vulgare, Taraxacum officinale, Son- chus asper, Lactuca muralis, and Senecio vulgaris probably had their origin in low-elevation clearcuts or agricultural fields tens of km to the west. Dale (1989) captured several of these same ruderal spe- cies at lower elevation on the debris avalanche along the Toutle River to the west. The only shrub or tree species trapped besides Salix was one seed of Acer circinatum, in a sample of >75,000 seeds. Since the montane sites on Mount St. Helens will, in the absence of another eruption, eventually succeed to a coniferous forest, the low abundance of late-successional woody spe- cies suggests a strong seed dispersal limitation. Similarly, Chapin et al. (1994) detected no spruce seeds and negligible alder seeds in the pioneer stage of primary succession at Glacier Bay at dispersal distances comparable to those of this study (ap- proximately 10 km from seed sources for spruce and 3 km from alder sources). Although not de- tectable in the seed rain, conifer and shrub seed- lings such as Pseudotsuga menziesii, Abies amabi- lis, Tsuga heterophylla, Pinus contorta Loudon, Al- nus sinuata (Regel) Rydb. (A. viridis (chain) DC), Rubus spp., and Vaccinium spp. do occur in low numbers at most of the sites sampled here (see also del Moral et al. 1995). These individuals are either establishing from extremely low seed source inputs and/or our seed trap design did not adequately sam- ple their mode of dispersal (see below). Seed traps were designed to estimate the seed rain onto relatively flat, open ground. True densities 2000] may exceed our estimates in microsites where seeds accumulate, such as in depressions or wet sites, or about rocks (Dale 1989; Titus and del Moral 1998). Higher densities also may occur in vegetated sites where short-distance dispersal supplements the long-distance seed rain. For example, at Willow Spring and Lupine Patch, seeds produced on or near the site probably equaled or even exceeded the number of seeds arriving by long-distance dispers- al. Also, variation among traps was highest at Wil- low Spring, with standard errors of 29% and 30% of the density means for 1989 and 1990, respec- tively (Table 2). This suggests that established veg- etation islands augment the long-distance seed rain in a patchy manner, in contrast to barren sites which receive a more predictable, albeit low input. The sharp decline in the seed rain of Senecio sylvaticus (Tables 3 and 4) may be explained by its life history—a biennial, it exploits forest clearcuts for only one or two generations before being out- competed by more aggressive seral species (West and Chilcote 1968; Halpern et al. 1997). The 1980 eruption of Mount St. Helens apparently created brief but favorable growing conditions for S. syl- vaticus in surrounding forests that resulted in a pulse of seed rain in 1982 and 1983. The overall mean density of 33.6 seeds 0.1 m~? yr' for mid-elevation barren sites on Mount St. Helens is similar to that found by Ryvarden (1971; calculations from Rabinowitz and Rapp 1980), who reported 34.2 to 65.3 seeds 0.1 m ° yr! for primary succession at the base of a retreating glacier in Nor- way, and to that of St6cklin and Baéumler (1996) who found 12.5 seeds 0.1 m~’ yr“! for newly ex- posed terrain in glacial forelands in Switzerland. Archibold (1980) reported 240.0 to 380.0 seeds 0.1 m~’ yr! in stripmine wastes in Saskatchewan, but this higher figure may be due to the closer prox- imity of seed sources. The very low mean density of 2 seeds 0.1 m~? yr“! for the subalpine lahar sites was probably because seeds of well-dispersed spe- cies such as Anaphalis margaritacea and E. wat- sonii did not reach that elevation, and because seeds of species in the surrounding vegetation have poor adaptations for dispersal (Wood and del Moral 1987). Many species that occur in the vegetation on Mount St. Helens were not recorded in the seed rain. Most of these species are uncommon or rare. This suggests that either their seed rain is below our detection limits or that their mode or timing of dispersal is such that they eluded capture. Although we think that low species richness in the seed rain is more likely, our traps were designed to capture wind-dispersed seed and thus may have missed capturing seeds of species with other dispersal modes. One possible dispersal mode that may be important on Mount St. Helens is that of secondary wind dispersal across hard snow surfaces. Matlack (1989) showed that seeds of Betula lenta were dis- persed greater distances by secondary dispersal WOOD AND MORAL: SEED RAIN ON MOUNT ST. HELENS ig, than by primary dispersal to the ground. Because Mount St. Helens receives abundant winter snow and freeze-thaw cycles are common, hard surfaces conducive to secondary dispersal by wind probably occur. Water dispersal (hydrochory) is another un- measured variable. In addition to permanent streams, numerous small temporary streams com- monly develop during spring snowmelt and fall rains, and sheet flow occurs during particularly heavy rains. Seeds can be transported along these watercourses (St6cklin and Béumler 1996). Either secondary dispersal or water dispersal may be re- sponsible for the spread of non-wind dispersed spe- cies such as Lupinus lepidus and the occurrence of the late successional woody species listed above. Animal dispersal (zoochory) is another unmeasured vector. We consider animals to be less important than either wind or water, but we cannot rule out their effect. Plant taxa with fleshy fruits are rare on Mount St. Helens (e.g., Vaccinium, Rubus; del Moral 1993) suggesting that frugivory as a means of seed dispersal is also rare. However, birds and large mammals such as elk and coyotes travel long distances to the study sites and may disperse seeds by defecation or transportation in their feathers or hair. The potential importance of a rare colonization event that results in local seed production and pop- ulation spread should not be underestimated. Whereas relative abundance of a species in the seed rain is a good indicator of its relative abun- dance in the vegetation, the reverse is not neces- sarily true. A few species are common in the veg- etation but uncommon in the seed rain. These in- clude Lupinus lepidus, Salix spp., and Eriogonum pyrolifolium. Lupine is the species with the greatest disparity between its estimated seed rain density and its abundance. Lupine survived the eruption in a variety of high elevation sites around the volcano (del Moral 1983, 1993; del Moral and Wood 1986) and was present on the Pumice Plain as early as 1981 (C. Crisafulli personal communication), pos- sibly establishing from seeds or root fragments washed down from high-elevation survivors. Now lupine occurs across the Pumice Plain and other sites on Mount St. Helens (Morris and Wood 1989; Bishop and Schemske 1998; Titus et al. 1998). In spite of this early record of population growth, seeds of L. lepidus were not captured in the seed rain until 1989, presumably because of its limited seed shadow. Lupine seeds have no obvious dis- persal adaptations except for ballistic dispersal when legumes dehisce, but this type of dispersal probably only achieves a few m (Willson 1993). Thus the rapid increase of L. lepidus on the Pumice Plain was due to vigorous seedling recruitment in close proximity to early colonists, not to long dis- tance dispersal (Wood and del Moral 1988; Morris and Wood 1989). The low abundance of Salix spp. in the samples is probably due to a flaw in the sam- pling design. Salix began reproducing as early as 1985 at Willow Spring (Wood personal observa- 8 MADRONO tion) but each year due to impassible roads our traps were put out too late to catch dispersing wil- low seeds. The high abundance of Salix around Willow Spring would undoubtedly have contributed greatly to the seed rain at this site and would have resulted in lower relative abundances of other spe- cies such as E. watsonii. Eriogonum pyrolifolium, a dominant subalpine species, has relatively heavy, round seeds with no obvious dispersal adaptations (Wood and del Moral 1987) and thus its seed shad- Ow apparently did not extend to the seed traps. The vast majority of incoming seeds in the seed rain fail to establish. Vegetation remained generally sparse by 1990 in spite of a rain of hundreds of seeds m~’ yr! onto most sites. Previous studies demonstrated that limits to abundance on Mount St. Helens are set by environmental factors. Morris and Wood (1989) and del Moral and Wood (1993a) showed that Lupinus lepidus may facilitate the es- tablishment of several species including Anaphalis margaritacea, Epilobium angustifolium, and Hy- pochaeris radicata. Wood and Morris (1990) showed that manipulation of substrate moisture and microtopographic heterogeneity positively affected the rate of establishment of A. margaritacea and E. angustifolium. Del Moral and Wood (1993b) showed that most species on the Plains of Abraham established in favorable microsites more often than expected by chance. There is also a tradeoff be- tween seed mass and probability of establishment— heavier seeds have a greater likelihood of estab- lishing on Mount St. Helens due to increased seed- ling vigor but are less likely to disperse a long dis- tance (Wood and del Moral 1987; Wood and Morris 1990). Titus and del Moral (1998) further demon- strated the importance of microsites in seedling es- tablishment. Thus, the vegetation of early primary succession on Mount St. Helens is composed pri- marily of well-dispersed species in low abundance. Stochastic events such as chance colonization of species with low long-distance seed rain result in heterogeneous communities with little structure (del Moral et al. 1995). As succession proceeds, com- munity composition will become increasingly un- correlated with the long-distance seed rain. ACKNOWLEDGMENTS We thank L. Zemke and A. Ziegler for assistance in the herbarium in Seattle; R. Sugg and J. Edwards for design- ing the fallout traps and collecting early samples; V. Do- raiswamy and J. Marr for excellent work in seed identi- fication and counting; J. Hubbell for invaluable field as- sistance; and P. Frenzen and C. Crisafulli of the Mount St. Helens National Volcanic Monument for friendship and logistical support. Useful comments on the manuscript were provided by J. Hubbell, M. Potvin, J. Titus and S. James. This research was supported by a California State University, Chico, Research Award to David Wood, NSF grants BSR 82-07042 and BSR 84-07213 to Roger del Moral and BSR 89-06544 to Roger del Moral and David Wood. [Vol. 47 LITERATURE CITED ARCHIBOLD, O. W. 1980. Seed input as a factor in the regeneration of stripmine wastes in Saskatchewan. Canadian Journal of Botany 58:1490-1495. BisHop, J. G. AND D. W. SCHEMSKE. 1998. Variation in flowering phenology and its consequences for lupines colonizing Mount St. Helens. Ecology 79:534—546. BooTH, B. D. AND D. W. Larson. 1998. The role of seed rain in determining the assembly of a cliff commu- nity. Journal of Vegetation Science 9:657—669. CHAPIN, D. M. AND L. C. BLIss. 1989. Seedling growth, physiology and survivorship in a subalpine volcanic environment. Ecology 70:1325—1334. CHAPIN, FE S. II, L. R. WALKER, C. L. FASTIE, AND L. C. SHARMAN. 1994. Mechanisms of primary succession following deglaciation at Glacier Bay, Alaska. Eco- logical Monographs 64:149-—175. CONNELL, J. H. AND R. O. SLATYER. 1977. Mechanisms of succession in natural communities and their role in community stability and organization. American Nat- uralist 111:1119-1144. DALE, V. A. 1989. Wind dispersed seeds and plant recov- ery on Mount St. Helens debris avalanche. Canadian Journal of Botany 67:1434-1441. DECKER, R. AND B. DECKER. 1981. The eruptions of Mount St. Helens. Scientific American 244:68-80. DEL MorAL, R. 1983. Initial recovery of subalpine vege- tation on Mount St. Helens. American Midland Nat- uralist 109:72—80. DEL MoraAL, R. 1993. Mechanisms of primary succession on volcanoes: a view from Mount St. Helens. Pp. 79-— 100 in J. Miles and D. H. Walton (eds.), Primary Succession On Land. Blackwell Scientific Publica- tions, London. DEL MoRAL, R. 1998. Early succession on lahars spawned by Mount St. Helens. American Journal of Botany 85: 820-828. DEL MoraL, R. 1999. Plant succession on pumice at Mount St. Helens, Washington. American Midland Naturalist 141:101-114. DEL MoraL, R. AND L. C. BLiss. 1993. Mechanisms of primary succession: insights resulting from the erup- tion of Mount St. Helens. Advances in Ecological Re- search 24:1—66. DEL MorRAL, R. AND D. M. Woop. 1986. Subalpine veg- etation recovery five years after the Mount St. Helens eruptions. Pp. 215-221 in S.A.C. (ed.), Mount St. Helens: Five Years Later. Eastern Washington State University Press, Cheney, WA. DEL Mora, R. AND D. M. Woop. 1988. Dynamics of herbaceous vegetation recovery on Mount St. Helens, Washington, USA, after a volcanic eruption. Vegeta- tio 74:11-27. DEL MoraL, R. AND D. M. Woop. 1993a. Early primary succession on the volcano Mount St. Helens. Journal of Vegetation Science 4:223-—234. DEL MoRAL, R. AND D. M. Woop. 1993b. Early primary succession on a barren volcanic plain at Mount St. Helens, Washington. American Journal of Botany 80: 981-991. DEL MORAL, R., J. H. Tirus, AND A. M. Cook. 1995. Early primary succession on Mount St. Helens, Washing- ton, USA. Journal of Vegetation Science 6:107—120. DLUGOSCH, K. AND R. DEL MORAL. 1999. Vegetational het- erogeneity along elevational gradients. Northwest Science 73:12-18. Epwarps, J. S. 1986. Derelicts of dispersal: arthropod fall- 2000] out on Pacific Northwest volcanoes. Pp. 196—203 in W. Danthanaraya (ed.), Insect Flight: Dispersal and Migration. Springer Verlag, New York, NY. Epwarps, J. S. AND P. Succ. 1993. Arthropod fallout as a resource in the recolonization of Mount St. Helens. Ecology 74:954-958. HALPERN, C. B., P. M. FRENZEN, J. E. MEANS, AND J. E FRANKLIN. 1990. Plant succession in areas of scorched and blown-down forest after the 1980 eruption of Mount St. Helens, Washington. Journal of Vegetation Science 1:181—194. HALPERN, C. B., J. A. ANTOS, M. A. GEYER, AND A. M. OLSON. 1997. Species replacement during early sec- ondary succession: the abrupt decline of a winter an- nual. Ecology 78:621-—631. HARPER, J. L. 1977. Population Biology of Plants. Aca- demic Press, London. HICKMAN, J. C. (ED.). 1993. The Jepson Manual: higher plants of California. University of California Press, Berkeley. HitcHcock, C. L., AND A. CRONQUIST. 1973. Flora of the Pacific Northwest. University of Washington Press, Seattle. JEFFERSON, R. G. AND USHER, M. B. 1989. Seed rain dy- namics in disused chalk quarries in the Yorkshire Wolds, England, with special reference to nature con- servation. Biological Conservation 47:123—136. JOHNSON, C. K. AND WEST, N. E. 1988. Laboratory com- parisons of five seed-trap designs for dry, windy en- vironments. Canadian Journal of Botany 66:346—348. KRUCKEBERG, A. R. 1987. Plant life on Mount St. Helens before 1980. Pp. 3—23 in D. Bilderback (ed.), Mount St. Helens: One Year Later. University of California Press, Berkeley, CA. LIPMAN, P. W. AND D. R. MULLINEAUX. 1981. The 1980 eruptions of Mount St. Helens, Washington. U.S. Geological Survey Professional Paper 1250. MATLACK, G. R. 1989. Secondary dispersal of seed across snow in Betula lenta, a gap-colonizing tree species. Journal of Ecology 77:853-869. Morris, W. F AND D. M. Woop. 1989. The role of Lu- WOOD AND MORAL: SEED RAIN ON MOUNT ST. HELENS 2 pinus lepidus in succession on Mount St. Helens: fa- cilitation or inhibition? Ecology 70:697—703. PICKETT, S. T. A., S. L. COLLINS, AND J. J. ARMESTO. 1987. Models, mechanisms and pathways of succession. Botanical Review 53:335-—371. RABINOWITZ, D. AND J. K. Rapp. 1980. Seed rain in a North American tall grass prairie. Journal of Applied Ecology 17:793-—802. RYVARDEN, L. 1971. Studies in seed dispersal. I. Trapping of diaspores in the alpine zone at Finse, Norway. Nor- wegian Journal of Botany 18:215—226. SOLBRECK, C. AND D. ANDERSSON. 1987. Vertical distri- bution of fireweed, Epilobium angustifolium, seeds in the air. Canadian Journal of Botany 65:2177—2178. STOCKLIN, J. AND E. BAUMLER. 1996. Seed rain, seedling establishment and clonal growth strategies on a gla- cier foreland. Journal of Vegetation Science 7:45—S6. Titus, J. H., AND R. DEL MoRAL. 1998. Seedling estab- lishment in different microsites on Mount St. Helens, Washington, USA. Plant Ecology 134:13-26. Titus, J. H., S. Moore, M. ARNOT, AND P. J. Titus. 1998. Inventory of the vascular flora of the blast zone, Mount St. Helens, Washington. Madrono 45:146— 161. WERNER, P. A. 1975. A seed trap for determining patterns of seed deposition in terrestrial plants. Oecologia 59: 145-156. WEsT, N. E. AND W. W. CHILCOTE. 1968. Senecio sylva- ticus in relation to Douglas-fir clearcut succession in the Oregon Coast Range. Ecology 49:1101—1107. WILLson, M. FE 1993. Dispersal mode, seed shadows, and colonization patterns. Vegetatio 107/108:261—280. Woop, D. M. AND R. DEL MORAL. 1987. Mechanisms of early primary succession in subalpine habitats on Mount St. Helens. Ecology 68:780—790. Woop, D. M. AND R. DEL MoRAL. 1988. Colonizing plants on the Pumice Plain, Mount St. Helens, Washington. American Journal of Botany. 75:1228—1237. Woop, D. M. AND W. E Morris. 1990. Ecological con- straints to seedling establishment on the Pumice Plain, Mount St. Helens, Washington. American Jour- nal of Botany. 77:1411-1418. MADRONO, Vol. 47, No. 1, pp. 10—20, 2000 ENVIRONMENTAL GRADIENTS AND VEGETATION STRUCTURE ON SOUTH TEXAS COASTAL CLAY DUNES KERN EWING College of Forest Resources, Box 354115, University of Washington, Seattle, Washington 98195 ABSTRACT Clay dunes are unusual geological features that occur near playas, lagoons, or flats that are sometimes wet but dry out annually. If the sediment in these ephemeral bodies of water contains clay, and if there are strong prevailing winds, flakes or granules of clayey material are transported during the dry season and are caught by edge vegetation. The clay particles moisten in the dew or rain and stick together, eventually creating dunes that support vegetation. Known locally as Jomas, the clay dunes along the Gulf coast of Texas and Mexico reach their greatest stature near the mouth of the Rio Grande River, where this study was carried out. These dunes support ecologically unique vegetation assemblages. They sit, like islands, in hypersaline lagoons. Sharp environmental gradients separate halophytes from typical coast- al thornscrub vegetation. Endangered animal species such as ocelots live in the thornscrub. Development pressures along the border threaten their existence, and the construction of artificial Jomas has been proposed. In this paper I characterize four Joma plant communities. The first community is found in the adjacent hypersaline Flats, and is limited to halophytes. The second community is found in lower but still elevated salinities at the Edge of the Jomas. At low salinities atop the Jomas are the dense Thornscrub community, and a Mixed Halophyte and Thornscrub community that is hypothesized to be the result of disturbance. Analysis of elevation and salinity at plots along transects through the Jomas allows me to correlate individual plant species with salinity preferences and community membership. An interesting outcome is that while a number of species have fidelity to one community type, there are quite a few bridging species that are found in two community types. This information has important implications for the degree of precision required when attempting to restore or create the clay dune ecosystem. The clay dunes that occur along the Laguna Ma- dre on the southern part of the Texas Gulf coast are interesting both geologically and biologically. Clay dunes are found in association with playas in west Texas and New Mexico, and are reported in Aus- tralia and Africa. The coastal clay dunes described in this paper reach their greatest stature and number near the mouth of the Rio Grande River and de- crease to the north and south. Similarly, the vege- tation of the south Texas delta of the Rio Grande is unique and in some places luxuriant, but dimin- ishes as the distance from both the river and the Gulf increases. The clay dunes, or /omas as they are known lo- cally, are rendered more exotic by the fact that they exist as low hills with non-halophytic vegetation, sitting in the middle of extensive hypersaline wind tidal flats or lagoons. The lagoons are periodically inundated by wind tides or hurricanes, but are sub- jected to long periods of drying in the hot south Texas climate. Salts concentrate and the surface of the lagoons may become dry. In addition to the salinity and periodic droughts, there is a persistent brisk wind out of the southeast for much of the year. In this semiarid, semitropical climate, the warm months may be generally described as any but January and February, and even during these months, daytime temperatures above 30°C are com- mon. Lomas near Boca Chica, the beach north of and adjacent to the mouth of the Rio Grande River, are covered with thornscrub vegetation, some of which is commonly found over much of the Tamaulipan biotic province. In addition, species characteristic of the Jomas and other coastal areas are found, in- cluding Citharexylum berlandieri Robins., Mayten- us texana Lundell, Prosopis reptans Benth., Echeandia chandleri (Greenm. & Thomps.) M. C. Johnst., Monanthochloe littoralis Engelm., and Yucca treculeana Carr. Most of the lomas in the Boca Chica area are named, and the site for this work is called Loma Tio Alejos. It has a roughly north-south orientation, is 200 to 300 meters wide by 600 meters long, and rises about 7.5 m out of a lagoon, which is at an elevation of about 1.5 m. The south end of the Joma is 300 meters north of a bend in the Rio Grande River; the east side is about 12 km from the Gulf of Mexico. This site and much of the surrounding land is now part of the Lower Rio Grande Valley National Wildlife Refuge. It has long been known that non-halophytic veg- etation grows on lomas, and that hypersaline marshes and flats surround them. They are recog- nized as a unique biotic community by the U.S.EW.S., and are included in a proposed wildlife corridor running down the Rio Grande and up the coast. One impetus for this study is the extensive restoration program in place on the L.R.G.V. Na- tional Wildlife Refuge and the potential for con- 2000] structing or restoring /omas to create habitat. Off- refuge, there are significant pressures from en- croaching commercial and residential development and from the proposed construction of a new inter- national bridge and its infrastructure 8-km west of this site. Knowledge of typical vegetation composition on lomas and the relationships among vegetation, Joma elevation and soil salinity are critical to understand- ing what controls the vegetation structure. Loma vegetation is known to be dense, and often much shorter than it would be in other locations. In this study I measured vegetation composition, woody plant height and density, canopy cover, elevation and soil salinity in quadrats along transects across Loma Tio Alejos. Ordination and classification anal- yses were performed on cover data. The nature of the relationship between Joma elevation and salinity was determined, and species affinity to sites was related to salinity and elevation. Clay dunes. The clay dunes along the Gulf Coast in southern Texas have been remarked upon almost since the first accounts of the exploration of the area, probably because these explorations sought river mouths and disembarked from coastal areas. Coffey (1909), while on a soil survey of the region for U.S.D.A., saw the dunes and hypothesized that they were formed by granules of clay, which were blown off the surface of dried lagoons during hot, windy summers. The particles blew to the edge of the lagoon, were caught by vegetation or by wrack or debris, and began to accumulate. Rainfall or the humidity of the nights caused the particles to coa- lesce. Coffey further noted that they were found near the Rio Grande because the rains are seasonal and lagoons dry out; in more humid climates such conditions do not occur. Foscue (1932) noted that the dunes looked like small islands covered with brush. Huffman and Price (1949) and Price and Kornicker (1961) compared clay dunes all along the Texas and Mexican coast and determined that they existed along the mainland coast from Soto la Marina River in Tamaulipas, Mexico to St. Charles Bay (at the Aransas National Wildlife Refuge) in Texas. The dunes are highest at the Rio Grande (10 m), and become lower (1 m) in the more humid climates to the north and south. These authors es- sentially agreed with Coffey about the formation of the dunes, adding that they probably grow only dur- ing hot months (March to November), retain a loosely porous structure, and represent about 5000 years of growth since beginning of the current still- stand of sea level. During a seven year period of drought in the fifties, about a foot of loosely con- solidated pellets accumulated. Their height made the dunes attractive camp sites for the coastal In- dian tribes that fished in the area. Aboriginal arti- facts occur from about mid-dune to near the top foot, and European artifacts occur near the top. In EWING: VEGETATION OF CLAY DUNES 11 addition to their use by humans, the endangered ocelot (Tewes 1982) also uses clay dunes. Tamaulipan Thornscrub. Brown (1994) de- scribes the Tamaulipan biotic province as being one of several provinces that are semidesert scrublands. Such systems are dominated by thorny shrubs and small trees, and characterize much of the world’s tropic-subtropic zones. They are found in Australia (mulga), southern Africa (bush), South America (chaco-seco), Mexico (matorral) and Texas (chap- arral). They are drought-deciduous communities that occupy a position on a moisture gradient some- where between desert scrub and woodland or forest. They often have an irregularly layered overstory between 2 and 8 m in height, and are typically com- posed of spinose, microphyllous, and succulent life forms. Thornscrub is often in competition with grassland, and may increase under grazing pres- sures, with fire suppression, or on poorer soils. Muller (1947) observed that east central Coahui- la, southern Texas, northern Nuevo Leon and north- ern Tamaulipas all have a vegetation form that is similar. Shreve (1917) called it Texas semi-desert. Muller proposed that it be called Tamaulipan thorn shrub. The more luxuriant and _ tree-dominated forms found in south Texas and Tamaulipas were called Tamaulipan thorn forest. These environments differ from the adjacent Chihuahuan desert shrub in that they are found at lower elevations, have more rainfall, and are exposed to winds from the Gulf of Mexico. With these habitat differences are also found more thorny shrubs, an abundance of grasses, more luxuriant growth of shrubs, a richer flora, and more numerous characteristic species. With the increase in species there is also a greater number of variants of the vegetation formation. Blair (1950) included the area in Texas south of the Balcones fault line (which runs from Austin through San Antonio) in his Tamaulipan province. He described the biota of the province as neotropic, strongly diluted by Sonoran biota characteristic of the southwestern U.S. and parts of Mexico, and by biota characteristic of the forests blanketing the Gulf coastal plain. The climate is semiarid and megathermal. From the coast westward, the brush thins as available moisture declines. In Cameron County, at the southern tip of the state, average annual precipitation is just above 25 inches. Mean maximum temperature is 95° in July, mean mini- mum is 51° in January. Rainfall peaks during trop- ical storm season (centered on September). Long periods of drought, during which there is little or no rain for 4—6 months, are common; periods dur- ing which drought years occur for 3—5 years in suc- cession are also common. A strong, persistent hot wind blows out of the southeast for much of the year. Probably the earliest exhaustive description of the thornscrub vegetation of the Rio Grande delta was given by Clover (1937). Later Blair (1950), in 12 MADRONO his delineation of the biotic provinces of Texas, would call the area on the floodplain the Matamo- ran district of the Tamaulipan Biotic Province. Clo- ver justified the use of the term chaparral for the shorter vegetation of the area, saying that it referred to chaparro prieto (Acacia rigidula Benth.). Mes- quital is the term used for Prosopis glandulosa Tor- rey-dominated communities, and sacatal for grass- lands. Currently, two general types of brush habi- tats are recognized in the area. The first is referred to as riparian and scrub forests (associated with the Rio Grande, and producing taller vegetation); the second is upland thornscrub and thorn woodland (Jahrsdoerfer and Leslie 1988). Clover (1937) described the vegetation of the clay dunes near the coast as being similar to salt- affected thornscrub nearby, but being composed of shrubs twisted by the heavy winds. Dominants list- ed were Pithecellobium ebano (Berl.) C. H. Mull., Leucophyllum frutescens (Berl.) I. M. Johnst., Ziz- iphus obtusifolia (Torrey & A. Gray) A. Gray, Cas- tela texana (T. & G.) Rose, Randia rhagocarpa Standl., Forestiera angustifolia Torr., Prosopis glandulosa Torr. and Celtis pallida Torr. Between the clay dune “‘islands’”’ and the main chaparral- mesquital was a transition zone and sacahuistal (dominated by Spartina spartinae (Trin.) Merr.). USFWS (1997) added Citharexylem berlandieri, Erythrina herbacea, Dalea scandens (Mill.) R. T. Clausen, Echeandia chandleri and Sporobolus thar- piui Hithc. as being found exclusively in or near the loma-coastal brushland community. Johnson (1963) added that the windward sides of some of the dunes were covered with a thick growth of Sporobolus wrightii Scribn. (sacaton). METHODS In late October 1998, I began the establishment of two transects across Loma Tio Alejos. The first transect ran approximately east-west. It started in an unvegetated area of hypersaline lagoon to the west of the Joma, crossed 200-m of halophytic veg- etation, encountered the southern part of the loma and entered thornscrub vegetation. It climbed for the next 100 m to the ridgeline and then dropped, over the following 100-m, to the edge of halophytic vegetation on the east of the Joma. The transect then ended after it traversed 30 m of the halophytic vegetation. Rather than being symmetrical, the loma is kidney-bean shaped, so that transect two could be oriented in a north-south direction and cross the north end of the Joma almost perpendic- ular to its axis. The second transect was finished by the begin- ning of December. It started on the south side of the north end of the Joma and traveled for 30 m in halophytic vegetation, then entered the brush and traveled 80 m to the ridgeline. It then went down through a depression and up to another ridgeline, traversing extremely dense brush for about 70 m. [Vol. 47 The transect descended through thornscrub for 70 meters and was terminated about 50 m into the hal- ophytic vegetation to the north of the loma. Distance along transects was measured with sur- veying tapes, and station stakes were placed every ten meters (stations 0+00, 0+ 10, etc.). Differential leveling, employing a Keuffel and Esser optical level, was used to determine relative elevations along each transect. Elevations on transect one were tied to elevations on transect two by closing a leveling loop from one transect to the other along a trail which ran down the ridge of the Joma. The elevation of the transects was then fitted to a USGS contour map of the Joma and elevations from that map were used to register the high and low points surveyed. Vegetation was sampled in 10 < 20 m quadrats along most of both transects vegetation was so thick from station 0+20 to station 2+00 along tran- sect 2 that it was sampled using 5 < 20 m quadrats oriented with their long axis parallel to the ma- chete-cut line through the brush. By the conclusion of the investigation, 57 quadrats were measured and 76 plant species were found. In each quadrat the percent cover of herbaceous species was recorded, as well as the number of species, the number of individuals of each woody species, height and two crown-width dimensions for woody species. Cover for woody species was calculated by averaging the two dimensions and calculating the area of a circle with this average as the diameter. Soil samples for salinity measurements (Abbott 1967) were taken from each quadrat using a 3.8- cm diameter corer. Cores were generally 10-15 cm long, and were taken after removing organic matter and debris from the soil surface. Cores were ex- truded onto a tray and the length of each sample was measured, allowing the calculation of soil vol- ume. Wet weight of each core was measured, all were oven-dried at 80°C, and dry weights were measured. Dried cores were placed in flasks and a volume of water twice the original volume of the cores was added to each. The flasks were sealed with rubber stoppers and agitated for three days. They were then allowed to settle in a cold room for three days and the clear supernatant was removed with a syringe. Osmolality of the soil extract was measured with an Advanced Instruments Model 3300 Micro Os- mometer (Advanced Instruments, Inc, Two Tech- nology Way, Norwood, MA 02062). From the os- molality of the soil extract, values were calculated for the osmotically active solutes per unit volume of soil (‘‘a’’ below) and the apparent salinity of the soil solution at the time the sample was taken (“‘e”’ below) (Mahall and Park 1976). The following cal- culations were made to arrive at these two values. bX c bxXc = e = d f a 2000) where a = osmotically active solutes per unit volume of soil (m-osm. ml~'); b = osmolality (m-osm. ml~! water) of | soil extract (from freezing point | depression); c = volume of water added to dry soil core (ml); d = original volume of soil core (ml); e = apparent salinity of soil solution (m-osm. g™! water); f = weight of water in soil core (g). The soil solution salinity was converted to ppt for figures. Two data sets, one containing species cover in- formation for each quadrat and the second contain- ing elevation and soil salinity data at each quadrat were entered as the primary and secondary matrices in the multivariate software package PC-Ord (McCune and Mefford 1997). Classification was performed using TWINSPAN, and ordination was performed using DCA (detrended correspondence analysis), which uses the stand species matrix; the joint plot option was used to overlay environmental variable vectors over the DCA-generated ordination plot. TWINSPAN is a divisive classification technique that divides all initial stands in an analysis into two groups using an ordination, then iteratively refines the division. Each group formed is then divided into two new groups. (Jongman et al 1987). The mechanical division of existing groups can go on until some stopping rule is triggered (maximum number of divisions or minimum number of stands in a group may be specified). The actual selection of what divisions to accept may depend upon whether further divisions add to the explanatory power of the analysis (Gauch 1982). In this anal- ysis, four groups were used because each of the four groups represented a homogeneity of species composition and was characterized by similar dom- inant species. RESULTS Elevation. The elevation of the saline flats around the Jomas in this area is about 1.5 m, USGS datum. On this Joma, over a distance of 100-150 m, the measured transects rose to about 6.5—7 m at the height of land, then dropped again (Fig. 1). The elevation of the highest point on the Joma was es- timated at about 9 m. Vegetation. Seventy-six species were found with- in the sampling quadrats along the two transects across Loma Tio Alejos. Twenty-seven of them were erect, woody plants. There were also 8 grass- es, and a number of halophytes, weedy herbaceous Species, and herbaceous understory species. Echeandia chandleri, a lily limited to clay soils in south Texas and described as rare (USFWS 1997), was common here. Ophioglossum vulgatum L., a vascular cryptogam, was found growing under Pro- EWING: VEGETATION OF CLAY DUNES 13 sopis glandulosa trees. The vegetation gradient from the saline flats to the thornscrub of the lomas was short and steep, but there was surprising over- lap of thornscrub and halophytic species. Vegetation composition and cover data were an- alyzed using multivariate analysis. Classification of stands and species was accomplished using TWIN- SPAN. One of the products of a TWINSPAN anal- ysis iS a joint ordination of stands and species called a two-way table. The species list from a two- way table is ordered, i.e., the species that are on either end of the list are usually found in complete- ly different environments and species which are next to one another on the list are usually found together. On the list in Table 1, species at the top are from sites near the highest elevations of the loma, while species at the bottom occur mostly in the adjacent flats and at the Joma edges. I elected to stop the TWINSPAN procedure after four groups of stands had been generated because each of the four groups represented a relatively ho- mogeneous association of species and was charac- terized by a unique dominant species or group of dominants. I will describe the important species, the location of stands, and the general environment in which each group occurs. The first division segregated the high-diversity, non-halophytic thornscrub vegetation into one group. Woody thornscrub species and their associ- ated herbaceous understory or gap species domi- nate it. The other group from this division includes all of the species known to be halophytic, but hal- ophytes are not limited to the second group. The second division divided stands in the thorn- scrub vegetation group into two smaller groups, one of which is made up of species characteristic of more widely distributed coastal upland sites. All of the sites in this subgroup are found on the higher elevations of the north transect; the vegetation there is characterized by a dense canopy and shortened stature. On Figure 2, this association is called ‘**Thornscrub’’. The other subgroup in this division includes more salt-tolerant thornscrub which is found on the lower ends of the north transect and on the south transect. On Figure 2 this association is called ‘“‘Mixed Thornscrub and Halophytes’’. In- dicator species for the group are Prosopis reptans, Maytenus texana Lundell. and Ericamera austro- texana M. C. Johnst. Almost all of the Yucca tre- culeana Carr. and Prosopis glandulosa are also found in this grouping. This second subgroup could be further divided into closed canopy and open can- opy groupings, which would have slight species differences. The third division divides the stands containing mostly halophytic vegetation into two subgroups. The first subgroup is the association that makes up the low shrubby vegetation around the edges of the loma. On Figure 2 this is called ‘“‘Edge’’. Indicator species for this subgroup are Borrichia frutescens (L.) DC and Lycium carolinianum Walt. The sec- 14 MADRONO [Vol. 47 NORTH TRANSECT SALINITY 8.00 6.00 4.00 ELEVATION (m) 3.00 1.00 0.00 oe > o rm a oe FF SF SH Sf SH Nema VS eee eee eg es oe ee ge ee ee TF weer eT eT © 46 SOIL WATER SALINITY (ppt) S & © HD QO O AO xo eo a a) ou oe me ye Ri oS) & oS 9 S § QE S x STATIONS SOUTH TRANSECT SALINITY SALINITY ELEVATION (m) we x x x IN N x s « a ss We iy a S We ¥ s SOIL WATER SALINITY (ppt) © SD MP PO AP PD dP HO PP DO OD DO OD H GD PH OH WO H SP O_O HP HO P O | NE NIE RETRIEVE gk” ft oS ¥ SF FGF SF GF FF 7 9% 4 6 GF 9 9 9 G6 6 9 9 96 9 9 6 9 9 46 9 9 9 9 STATIONS Fic. 1. Elevation and salinity along the north transect (which runs north-south) and the south transect (which runs east-west) across Loma Tio Alejos, near the mouth of the Rio Grande River in south Texas. Salinity is represented by a dashed line, elevation by a solid line. The salinity is the apparent soil water salinity obtained by measuring the osmolality of a known volume of water into which a dried soil sample was mixed; salinity at the water content of the soil when the sample was taken was then calculated. ond subgroup is dominated by three species (Sali- cornia virginica L., Monanthochloe littoralis En- gelm. and Batis maritima L.) and little else; sites making up this subgroup occur in the extremely salty flats surrounding the /Jomas. On Figure 2 this association is referred to as ‘‘Flats’’. The same data set that was used to obtain a clas- sification of stands and species with TWINSPAN was subjected to an ordination analysis. Ordination allows the investigator to plot stands or species in a multi-dimensional space that can be interpreted in terms of environmental variation; Detrended Correspondence Analysis (DCA), an indirect ordi- nation technique, was used. Since an environmental matrix of stands by environmental data (salinity, elevation) was available, the joint plot option of PC-Ord was used to plot environmental vectors on the DCA ordination plots. The first DCA axis had and an eigenvalue of 0.859, indicating that a sub- stantial amount of the variation in the data set was accounted for by this axis. When plotted as a joint plot, the arrow representing salinity was almost 2000] identical to axis one, and the arrow representing elevation, while negatively correlated with salinity, was very close to axis one. Canonical Correspondence Analysis (CCA) al- lows the investigator to constrain the ordination axes to some combination of directly measured en- vironmental variables (Jongman et al. 1987). When this was done using elevation and salinity as the environmental variables, the eigenvalue for the first axis was 0.662, which is still high (Jongman et al. 1987). In the CCA analysis the arrows for elevation and salinity were highly correlated with axis 1. The results of these ordinations indicate that elevation and salinity are environmental factors that account for most of the variation in vegetation structure on the Jomas. They are inversely correlated; as eleva- tion increases, salinity decreases. Salinity. Salinity was high in the flats but dropped very quickly as the Joma elevation rose above that of the surrounding flats. (Fig. 1). Salinity per volume of soil and apparent salinity of soil wa- ter were highly correlated (r = 0.995), so apparent salinity in ppt will be used to discuss the relation- ship among elevation, vegetation and salinity. Av- erage soil water salinity in plots in the “‘Flats” veg- etation association was 64.1 + 11.8 ppt. This salin- ity 1s significantly (P = 0.05) greater than that in any of the other associations. Mean soil water sa- linity in the “‘Edge”’ association, 6.3 + 1.9 ppt, is higher than that in the ““Mixed”’ or “‘Thornscrub”’ groups, though not significantly so. Mean salinity in the “Mixed” group is 2.1 + 0.4, and in the “*Thornscrub”’ group is 2.0 + 0.3 ppt. DISCUSSION The ecotone between the hypersaline flats of La- guna Madre and the coastal thornscrub in south Texas has created interesting and unexpected veg- etation associations. The lomas or clay dunes are a microcosm of this contact, and they reach their maximum expression near the mouth of the Rio Grande River. Coastal thornscrub on the /omas and on the nearby mainland is valued because of its importance as wildlife habitat (ocelots, birds, and butterflies). It is part of a breathtaking diversity of vegetation; Jahrsdoerfer and Leslie (1988) indicat- ed that there were 265 native woody species in the thornscrub of southern Texas. It is the site of scarce and unusual plant species including the species of concern Echeandia chandleri, Citharexylum ber- landieri Robins. and the endemic Sporobolus thar- pti Hitchc. (Jahrsdoerfer and Leslie 1988). The La- guna Madre is unique in that it is a huge lagoon, which has little freshwater input except from trop- ical storms. The hot climate, persistent winds and a tendency to experience prolonged periods of drought have created a water body that becomes hypersaline. Many of the /Jomas near the site of this study rise out of wind flats at the edge of the La- EWING: VEGETATION OF CLAY DUNES 1S guna Madre; these flats are sometimes inundated but are very salty during most years. Many of the /omas near the mouth of the Rio Grande and behind Boca Chica beach are now pro- tected and part of the Lower Rio Grande Valley National Wildlife Refuge. This refuge and the La- guna Atascosa National Wildlife Refuge on the coast 20 km to the north are both sites of active vegetation restoration programs. Because of the sharp environmental gradients in areas adjacent to salt flats and lagoons, and because of unique tol- erances and preferences of plant species which would normally be selected for restoration plant- ings (in general, dominant and secondary woody species), information about the sorting of species along environmental gradients is important. Since the /omas rise like small islands out of low and level salt flats, their elevation was predicted to be an important environmental axis; this proved to be the case. Since /Jomas support species common to sites that are not salt-affected, and because they are surrounded by halophytic vegetation that is tolerant to high concentrations of salts, a salinity gradient was predicted. This also proved to be true. Other environmental factors may also shape veg- etation structure. Winds off the nearby gulf are per- sistent and may result in dwarfing of vegetation. Soils sampled were generally silty clays or clayey silts, with more clay in soils in the flats. Lomas have a history of use by people, and there are roads leading to them, around them and across them. There are excavation sites, dumps, and disturbed areas with weedy vegetation (primary weeds are the introduced pasture and lawn grasses Cenchrus cil- laris L., Cynodon dactylon (L.) Pers., Dichanthium annulatum Stapf. and Panicum maximum Jacq.); some of the vegetation plots occurred in these ar- eas. Multivariate analysis indicated that elevation and salinity change explains a very high amount of the variation in vegetation on Loma Tio Alejos. The analysis of data from 57 quadrats along two transects across the /oma was carried out by per- forming a classification procedure (TWINSPAN) and an ordination procedure (DCA). As a result of the TWINSPAN analysis, four vegetation group- ings or associations were identified. The groups were called ‘“Thornscrub”’, ‘‘Mixed Thornscrub and Halophytes’’, ‘‘Edge’’ and ‘‘Flats’’. The 16 plots in the Thornscrub group contained 36 plant species. For Mixed Thornscrub and Halophytes the numbers were 20 plots and 67 species, for Edge 13 plots and 41 species, and for Flats 8 plots and 11 species. All except one of the plots assigned to the Thorn- scrub group were on the north transect; that one occurred in very thick brush on the south transect. This association is made up of plants in very dense, short (3-4 m) vegetation. Vegetation on the north transect may have been subjected to greater wind intensity, because the south part of the Joma is par- tially protected from the wind by vegetation along 16 MADRONO [Vol. 47 TABLE 1. LIST OF SPECIES FOUND ALONG TRANSECTS AT LOMA TIO ALEJOS. Species have been ranked by TWINSPAN classification program so that those generally found at the higher elevation, lower salinity sites occur at the top of the list; those found at the lower elevations and in the saline flats are at the bottom. Frequency of occurrence in plots of each of the four TWINSPAN community types at the site (Thornscrub, Mixed Thornscrub and Halophytes, Edge and Flats) is shown for each species as a percentage. Numbers of plots of each of the community types are respectively n = 16, 20, 13 and 8. Thornscrub Mixed Edge Flats Castela texana 44 10 0 0) Pithecellobium pallens 6 0) 0 0) Malpighia glabra 25 5 0) 0) Bastardia viscosa 31 5 0 0) Rivina humilis 50 20 0) 0 Celtis pallida 63 30 8 0 Pithecellobium ebano 19 0) 0) 0) Phaulothamnus spinescens 81 45 8 0 Randia rhagocarpa 69 10 0) 0) Lycium berlandieri 13 5 0) 0 Aloysia gratissima 13 0 0) 0) Citharexylum berlandieri 94 85 8 0 Zanthoxylum fagara 100 85 0) 0) Karwinskia humboldtiana 63 35 8 0 Lantana horrida 38 60 0 0 Capsicum annum 6 20 0) 0) Schaefferia cuneifolia 22 40 0 0) Passiflora foetida 6 5 0 0 Cissus incisa 44 60 8 0 Forestiera angustifolia 19 30 8 0) Verbesina microptera 38 30 0 0) Zisiphus obtusifolia 13 20 8 0 Isocoma drummondii 19 80 15 0 Allowissadula lozani 19 aD 8 0 Yucca treculeana 13 60 15 0) Prosopis glandulosa 25 85 8 0 Eupatorium azureum 50 70 8 0 Leucophyllum frutescens 38 65 8 0 Condalia hookeri 6 10 0) 0) Echeandia chandleri 13 25 0) 0 Gymnosperma glutinosum 6 10 0) 0 Ericameria austrotexana 0) 80 8 0) Eupatorium incarnatum 0) 20 0) 0 Cenchrus incertus 0) 10 0 0 Atriplex acanthocarpa 0 5 0) 0) Wedelia hispida 0) 5 0) 0 Physalis cinerascens 0 20 0 0 Ophioglossum vulgatum 0) 20 0 0 Acacia farnesiana 0 5 0 0) Trixis inula 0 5 0 0 Dichanthium annulatum 0) 40 0 0 Croton cortesianus 0 5 0 0 Sida ciliaris 0 10 0 0 Malvastrum americanum 0 15 0 0 Ibervillea lindheimeri 0 10 0 0 Chenopodium ambrosoides 0 5 0) 0 Cenchrus ciliaris 0) 50 15 0 Sarcostema cynanchoides 0 30 15 0 Cynodon dactylon 0 10 8 0 Croton leucophyllus 0 10 8 0 Opuntia leptocaulis 6 30 31 0 Trandescantia micrantha 6 10 23 0 Acleisanthes obtusa 6 5 8 0 Borrichia frutescens 13 55 85 0) Oxalis drummondii 0) 5 8 0 Solanum eleagnifolium 0) 10 8 0) Sporobolus wrightii 0) 10 15 0 Panicum maximum 0 5 0 13 Oxalis dichondrifolia 6 15 15 0 2000] EWING: VEGETATION OF CLAY DUNES 17 TABLE 1. CONTINUED. Thornscrub Mixed Edge Flats Evolvulus alsinoides 0) 5 8 0 Maytenus texana 0) i Vd 13 Prosopis reptans 0) 70 85 25 Machaeranthera phyllocephala 0 45 46 0) Spartina spartinae 0 10 es 0) Suaeda linearis 0) 0) 23 25 Cressa nudicaulis 0) 10 ps: 25 Salicornia virginica 0) 0) 22 100 Talinum paniculatum 0) 5 8 22 Atriplex matamorensis 0) 0 0) 13 Lycium carolinianum O 20 85 0) Limonium nashii 0) 0) ai 0) Echinocactus setispinus var. setaceus 0) 0) 8 0) Distichlis spicata 0) O 8 0) Monanthochloe littoralis 0) 30 100 75 Opuntia engelmannii O 10 38 ig 0) 5 69 100 Batis maritima the river and by a road berm. Plots assigned to the Mixed group occurred in a more disturbed area on the north end of the north transect, and at the upper elevations of the south transect. There was erosion and open areas in both locations, so human distur- bance may have been partially involved in the cre- ation of such sites. Since salinity and elevation dif- ferences were inconsequential between plots in these two associations, the hypothesis that the Mixed association is generated by disturbance should be investigated more thoroughly. Any res- toration attempt would create a disturbed environ- ment, and so the Mixed association might be the expected mid-successional vegetation type on less- salty soils. Prosopis glandulosa, well-known as a self-seeder on open sites in South Texas (Archer et al. 1988), is a dominant species in the Mixed as- sociation but found in few plots in the more dense Thornscrub association. Plots in the Edge association occur in a band about 30—40 m wide around the Joma, and are vi- sually different from the adjacent brush because they are open, and woody vegetation in them is either short or widely spaced. At the lower and salt- ier sites in this association, plant diversity dimin- ishes to an average of 4 species per plot, and plants are generally less than a few decimeters tall. The saline Flats are characterized by a few species, and in places the salinity may become so great that no vascular plants occur. There was a substantial in- crease in soil salinity at the interface between the Edge association and the Flats (Fig. 1). At one un- vegetated quadrat in the flats, a soil water salinity of 187 ppt, roughly five times that of sea water, was measured. The mean soil water salinity of soil samples taken from plots in the Flats was 64 ppt, or almost twice the salinity of seawater. The Flats are truly the province of halophytes. Localized or general evaporation and concentration probably result in much higher localized salinity during drought pe- riods. Salinity in the Mixed and Thornscrub asso- ciations, on the other hand, fell under or around the 1.5—2 ppt salinity threshold that is generally con- sidered the point below which crop plants have no salinity problems (Hartman et al. 1990). Salinity in plots in the Edge association were generally in the range within which plants are likely to be affected, but not so high as to limit plant composition to halophytes. Each of the four vegetation associations identi- fied by classification analysis was characterized by a set of dominant species (based upon total cover). In analyzing the species composition of each as- sociation, it became evident that there were some species which preferred conditions found in sites limited to one vegetation association, but there were also many species that did quite well in two of the associations. For instance, Castela texana (T. & G.) Rose. and Randia rhagocarpa (Fig. 2a) were common in Thornscrub association and rare in the Mixed association. Citharexylum berlandieri (Fig. 2b) and Zanthoxylum fagara (L.) Sarg. were in both Thornscrub and Mixed associations. Yucca trecu- leana Carr. (Fig. 2c) preferred the Mixed associa- tion, while Borrichia frutescens (L.) DC (Fig. 2d), Maytenus texana Lundell. and Prosopis reptans were found in both Mixed and Edge associations. This combination of species with fidelity to an as- sociation and species which overlap associations continued with Lycium carolinianum Walt. (Fig. 2e) found in Edge, Monanthochloe littoralis and Batis maritima L. (Fig. 2f) found in Edge and Flats, and Salicornia virginiana L. (Fig. 2g) found in Flats. All of the species listed by Shindle and Tewes (1998) as recommended for the restoration of oce- lot habitat, as well as a broad variety of others, are found on the /Jomas. This paper has presented in- formation about the environmental preferences of loma species. The mixing of species which have 18 MADRONO A. RANDIA RHAGOCARPA NORTH TRANSECT o.co NORTH TRANSECT | IXED THORNSCRUB | NORTH TRANSECT ,_D-BORRICHIA FRUTESCENS THORNSCRUB MIXED THORNSCRUB ANO HALOPHYTES Fic. 2. ELEVATION (mm) ELEVATION (m) SOUTH TRANSECT MIXED THORNSCRUB aa | ee eG ie ee eh SOUTH TRANSECT Se ee Eee a ee MIXEO THORNSCRUB ANO HALOPHYTES NY ei FLATS SOUTH TRANSECT ee ee 2 es ee Presence of species in plots along transects. A solid vertical bar indicates that the species shown was present in a plot: a.) Randia rhagocarpa, occurring primarily in the Thornscrub association, b.) Citharexylum berlandieri, occurring in both the Thornscrub and Mixed associations, c.) Yucca treculeana, occurring primarily in the Mixed association, d.) Borrichia frutescens, occurring in both the Mixed and Edge associations, e.) Lycium carolinianum, occurring primarily in the Edge association, f.) Batis maritima, occurring in both the Edge and Flats associations, g.) Salicornia virginica, occurring primarily in the Flats association. narrow habitat ranges with species which have broader habitat ranges when planting a restoration project is wasteful of plant materials. For the plant installation phase of a restoration project, some spe- cies may be placed on the landscape with less pre- cision, but others require an exact understanding of the species preferences and the site conditions. Plant materials for restoration or creation of Jomas EWING: VEGETATION OF CLAY DUNES 19 NORTH TRANSECT E. LYCIUM CAROLINIANUM SOUTH TRANSECT MIXED THORNSCRUB ED THORNSCRUB pena ibantesiees AND HALOPHYTES NORTH TRANSECT F. BATIS MARITIMA fAND HALOPHYTES i NORTH TRANSECT G. SALICORNIA VIRGINICA } THORNSCRUB MIXED THORNSCRUB ‘AND HALOPHYTES H 1 i Fic. 2. Continued. will probably never be abundant, so placement of seedlings or seeds into the proper environmental zone will be critical to the success of a restoration project. In conclusion, this work has confirmed that coastal clay dunes or /Jomas are unique systems that are persistent over time. Unusual plant associations (thornscrub vegetation and halophytes in close proximity or mixed) and rare and threatened plants (Echeandia chandleri, Citharexylum_ berlandieri) are found on them; their individual vegetation structure can be complex. They are known to be valuable as wildlife habitat. Vegetation structure across lomas varies along environmental gradients, which can be predicted for the most part by mea- suring elevation and salinity. Reports in the litera- ture suggest that wind direction can also be an im- portant factor in vegetation composition and size (Clover 1937). For many centuries, these unique systems have been isolated and not greatly dam- Pit ia —-- es SOUTH TRANSECT aged. Population pressure and commercial devel- opment now pose a threat to the vegetation systems and the wildlife that they support. Restoration in other areas of south Texas Tamaulipan thornscrub has been undertaken successfully, and the core of an extensive wildlife corridor is being created along the coast and up the Rio Grande River. The resto- ration or creation of Joma vegetation to augment habitat and add to the wildlife corridor is an im- portant and achievable element of this restoration. The ability to restore unique ecosystems like the clay dunes, if indeed we have that ability, does not mean that there is no need for conservation of such unusual habitats. Conservation is an integral part of the U.S.EW.S. plan for development of a wildlife corridor in south Texas. Important parcels have been identified and a considerable acreage of land upon which dunes sit has been purchased. Resto- ration can augment the effectiveness of conserva- tion in a number of ways, including the creation of 20 MADRONO buffers, the increase in the effective size of a con- served parcel, the creation of corridors, and the ini- tiation of a successional trajectory that will even- tually result in an ecosystem that is not much dif- ferent from one at a conserved site. ACKNOWLEDGMENTS I would like to thank Chris Best, Monica Monk and Frank Gonzales for help in selecting this site, collecting data and reviewing results. Support from the U.S.EW.S was provided by Donna Howell, Ken Merritt, and Larry Ditto. Liz Van Volkenburgh kindly allowed me to use her new freezing point depression osmometer and Kari Stiles showed me how to use it. LITERATURE CITED ABBOT, W. 1967. Salinity determination by freezing-point depression osmometry. Lagunas Costeras, un Sim- posio. Mem. Simp. Intern. Lagunas Costeras. UNAM-UNESCO, Nov 28-30. Mexico, D.F:341- 348. ARCHER, S., C. SCIFRES AND C. R. BASSHAM. 1988. Au- togenic succession in a subtropical savanna: Conver- sion of grassland to thorn woodland. Ecological Monographs 58(2):111—127. BLAIR, W. FE 1950. The biotic provinces of Texas. The Texas Journal of Science 2(1):93-117. Brown, D. E. 1994. Biotic Communities: Southwestern United States and Northwestern Mexico. University of Utah Press, Salt Lake City. CLover, E. U. 1937. Vegetational survey of the Lower Rio Grande Valley of Texas. Madrofio 4(2):41—66 and 4(3):77—100. CoFFEY, G. N. 1909. Clay dunes. The Journal of Geology 17:754-755. CorRRELL, D. S. AND M. C. JOHNSTON. 1970. Manual of the Vascular Plants of Texas. Texas Research Foundation, Renner, Texas. EverITT, J. H. AND D. L. DRAwWE. 1993. Trees, Shrubs and Cacti of South Texas. Texas Tech University Press, Lubbock. Foscug, E. J. 1932. Physiography of the Lower Rio Gran- de Valley. Pan-American Geologist 57:263—267. GaucH, H. G. 1982. Multivariate Analysis in Community Ecology. Cambridge University Press, Cambridge. GOULD, E W. 1975. The Grasses of Texas. Texas Agri- cultural Experiment Station; Texas A&M University Press, College Station. HARTMANN, H. T., D. E. KESTER AND E T. DAvrgs. 1990. Plant Propagation Principles and Practices. Prentice [Vol. 47 Hall Career and Technology, Englewood Cliffs, New Jersey. HUFFMAN, G. G. AND W. A. PRICE. 1949. Clay dune for- mation near Corpus Christi, Texas. Journal of Sedi- mentary Petrology 19(3):118—127. JAHRSDOERFER, S. E. AND D. M. LESLIE. 1988. Tamaulipan brushland of the Lower Rio Grande Valley of Texas: Description, human impacts and management op- tions. U.S. Fish and Wildlife Service. Biological Re- port 88(36). U.S. Department of Interior. JOHNSTON, M. C. 1963. Past and present grasslands of southern Texas and northeastern Mexico. Ecology 44(3):456—466. JONGMAN, R. H. G., C. J. E TER BRAAK, AND O. FE R. VAN TONGEREN. 1987. Data analysis in community and landscape ecology. Pudoc, Wageningen, The Nether- lands. LONARD, R. L, J. H. EVERITT, AND FE W. Jupp. 1991. Woody Plants of the Lower Rio Grande Valley, Tex- as. Number 7 Miscellaneous Publications, Texas Me- morial Museum, The University of Texas at Austin. LONARD, R. I. 1993. Guide to Grasses of the Lower Rio Grande Valley, Texas. The University of Texas-Pan American Press, Edinburg, Texas. MAHALL, B. E. AND R. B. PARK. 1976. The ecotone be- tween Spartina foliosa Trin and Salicornia virginiana L., in salt marshes of northern San Francisco Bay. II. Soil water and salinity. Journal of Ecology 64:783-— 809. McCune, B. AND M. J. MEFFORD. 1997. Multivariate Anal- ysis of Ecological Data. Version 3.05. MjM Software, Gleneden Beach, OR. MULLER, C. H. 1947. Vegetation and climate of Coahuila, Mexico. Madrono 9:33-—57. PRICE, W. A. AND L. S. KORNICKER. 1961. Marine and lagoonal deposits in clay dunes, Gulf Coast, Texas. Journal of Sedimentary Petrology 31(2):245—255. RICHARDSON, A. 1995. Plants of the Rio Grande Delta. University of Texas Press, Austin. SHINDLE, D. B. AND M. E. TEwEs. 1998. Woody species composition of habitats used by ocelots (Leopardus pardalis) in the Tamaulipan Biotic Province. South- western Naturalist 43(2):273-279. SHREVE, F 1917. A map of the vegetation of the United States. Geogr. Rev. 3:119-125. TEWES, M. E. AND D. D. Everitt. 1982. Study of the en- dangered ocelot occurring in Texas. Year-end Report, U.S. Fish and Wildlife Service, Albuquerque, NM. U.S. FISH AND WILDLIFE SERVICE. 1997. Lower Rio Grande Valley and Santa Ana National Wildlife Refuges. Fi- nal interim comprehensive management plan and draft environmental assessment. U.S. Department of the Interior. VINES, R. A. 1974. Trees, shrubs and woody vines of the Southwest. University of Texas Press, Austin. MADRONO, Vol. 47, No. 1, pp. 21—28, 2000 CHROMOSOME COUNTS AND TAXONOMIC NOTES ON DRABA (BRASSICACEAE) OF THE INTERMOUNTAIN WEST. 1: UTAH AND VICINITY MICHAEL D. WINDHAM Utah Museum of Natural History, University of Utah, Salt Lake City, UT 84112-0050 ABSTRACT Of the 350+ species ascribed to Draba, nearly one quarter occur in the Intermountain Region of the western United States. Most of these Draba species have not been examined cytologically. This paper presents a total of 18 chromosome counts for 11 different taxa occurring in Utah, Wyoming, and Arizona. The chromosome numbers of D. juniperina, D. kassii, D. maguirei var. maguirei, D. rectifructa, D. sobolifera, D. spectabilis var. spectabilis, and D. subalpina are reported here for the first time. Counts differing from published reports are documented for D. asprella var. stelligera and D. cuneifolia var. cuneifolia. The taxonomic significance of the new chromosome counts is discussed for each species. Counts of n = 11 and n = 13 appear to be the first reports of those numbers in the genus, and they complete the continuous series of aneuploid base numbers extending from 8 to 16. It is suggested that the Intermountain West may be a center of diversity for aneuploid Draba, and that this assemblage of species provides a unique opportunity to study chromosomal evolution and speciation. Species assigned to Draba, considered to be the largest genus in the Brassicaceae (Rollins 1993), occupy a variety of habitats and occur on all con- tinents except Australia and Antarctica. The group achieves its greatest diversity in topographically complex, mountainous regions where the disjunct occurrence of suitable habitats seems to favor iso- lation and speciation (Payson 1917). A prime ex- ample of this is seen in the Intermountain Region of the western United States, broadly defined here as the territory extending from the continental di- vide to the Pacific Crest (Sierran-Cascade axis). Of the 350+ species attributed to Draba by Rollins (1993), nearly one quarter occur in this region and more than 50% of those are endemic to it. The Intermountain West is terra incognita as far as the cytology of Draba is concerned. Of the 57 taxa confined to this region, only 11 have been ex- amined chromosomally. Half of these are known from single counts, and none can be considered ad- equately sampled. By comparison, 37 of the 40 Draba species found in Canada and Alaska have been studied cytologically, thanks in large part to the diligent efforts of G. A. Mulligan (1966, 1970a, b, 1971a, b, 1972, 1974, 1975, 1976). Mulligan’s work on the high-latitude North American species of Draba (summarized in the 1976 paper) led to major advances in our taxonom- ic understanding of the genus. In addition to clari- fying species boundaries in several groups, his data provided the basis for the only modern infrageneric classification of North American Draba. Setting aside Draba (Erophila) verna L., a Eurasian intro- duction unrelated to the native species, Mulligan (1976) recognized three informal groups based on a combination of chromosome number, flower col- or, breeding system, and hybridization studies. All 17 of the white-flowered species studied by Mulligan exhibit euploid chromosome numbers based on x = 8. They clearly are related to Eurasian boreal species assigned by Schulz (1927) to the sec- tion Leucodraba DC. Another nine Canadian spe- cies were assigned to his yellow-flowered euploid alliance, which also frequents boreal habitats and has representatives in Eurasia. The remaining 13 Canadian taxa were placed in a yellow-flowered group characterized by aneuploid chromosome numbers of n = 9, 10, 12, 14, 15, and 37 (Mulligan 1976). Apparently restricted to North and South America, this assemblage of species appears more tolerant of the warm/dry conditions that prevail in much of the western United States. Mulligan’s (1976) informal classification of North American Draba is a vast improvement over the patently unnatural sections proposed by Schulz (1927). However, it can neither be used nor eval- uated phylogenetically until the chromosome num- bers of local Draba species have been determined. The goals of this study were: 1) to collect crucial chromosome data for Intermountain Draba species, 2) to critically assess current taxonomic treatments for the species sampled, and 3) to develop a set of chromosomally vouchered samples for a DNA analysis (Beilstein and Windham in prep.) designed to test the monophyly of Mulligan’s (1976) infor- mal species groups. MATERIALS AND METHODS Chromosome counts were made from flower buds of wild plants fixed in Farmer’s solution (3 Sp) MADRONO parts 95% ethanol: 1 part glacial acetic acid). Fixed materials were stored at —20°C for up to five years and transferred to 70% ethanol immediately before making slides. Buds (or dissected anthers in larger- flowered species) were macerated in a drop of 1% acetocarmine stain, which was mixed 1:1 with Hoy- er’s solution prior to setting the cover slip and squashing. Slides were examined with an Olympus BH-2 phase contrast microscope, and representative cells were photographed using Kodak Technical Pan 2415 film. A full set of voucher specimens was deposited at the Garrett Herbarium, Utah Museum of Natural History (UT). Duplicate vouchers were deposited at the herbaria listed in Table 1. To guide the discussion, I produced a compendium of pub- lished chromosome counts for all taxa studied and their putative relatives. This list was assembled by running all accepted names and synonyms from Rollins (1993) and Kartesz (1994) through Chro- mosome Numbers of Flowering Plants (Federov 1974) and a complete set of the Index to Plant Chromosome Numbers spanning the period 1966— 1995 (Omduff 1967, 1968; Moore 1973, 1974, 1977; Goldblatt 1981, 1984, 1985, 1988; Goldblatt and Johnson 1990, 1991, 1994, 1996 & 1998). The primary literature was consulted to verify critical taxonomic and geographic information for each North American count identified by this search. RESULTS My chromosome studies of Utah, Wyoming, and Arizona Draba species yielded a total of 18 counts for 11 different taxa (Table 1). Seven of these taxa have not been counted previously. These include D. Juniperina Dorn (n = 11), D. kassii Welsh (n = 11), D. maguirei C. L. Hitche. var. maguirei (n = 16), D. rectifructa C. L. Hitche. (n = 12), D. so- bolifera Rydb. (n = 13), D. spectabilis Greene var. spectabilis (n = 10), and D. subalpina Goodman & C. L. Hitche. (n = 13). Counts for two of the re- maining taxa, D. asprella Greene var. stelligera O. E. Schulz (n = 15), and D. cuneifolia Nutt. ex. T. & G. var. cuneifolia (n = 15), differ from numbers previously reported in the literature. Unexpected counts, especially those that disagree with the lit- erature, are documented photographically in Fig- ures 1—6. Determinations of n = 11 and n = 13 appear to be the first reports of those numbers in the genus, and they complete the continuous series of aneuploid base numbers extending from 8 to 16. In fact, this small sample of taxa includes every step in that aneuploid series except n = 9 andn = 14. DISCUSSION The plants herein referred to Draba albertina Greene originally were identified as D. stenoloba Ledeb. based on the treatment in A Utah Flora (Welsh 1993). Because D stenoloba has a chro- mosome number of n = 20 (Mulligan 1975), I was [Vol. 47 surprised when samples from two widely separated Utah populations yielded counts of n = 12 (Fig. 1) and 2n = 24. These determinations agree with pre- vious reports for D. albertina, including four counts from Alberta and one from the Northwest Territo- ries (Mulligan 1975). An additional count of n = 12 from Wyoming originally attributed to D. sten- oloba (Mulligan 1966) was reassigned to D. alber- tina in a subsequent paper by Mulligan (1975). Prior to detailed studies of the group (Mulligan 1975), Draba albertina was treated as a synonym or variety (nana) of D. stenoloba. After discovering that the two taxa had different chromosome num- bers, Mulligan recognized them as separate species based on correlated morphological and geographi- cal differences. The decision to classify these taxa as species also is supported by artificial hybridiza- tion experiments (Mulligan 1975), which indicate that any hybrids formed are completely sterile. According to Mulligan (1975) and Rollins (1993), D. stenoloba, with a chromosome number of n = 20 and mostly dendritic trichomes on the upper leaf surfaces, is rarely encountered south of the Canadian border. They assign most collections identified as D. stenoloba from the western United States to D. albertina, characterized by a chromo- some number of n = 12 and simple or once-forked adaxial leaf trichomes. My morphological studies of Utah specimens concur that typical D. stenoloba is not present in the state, and all collections iden- tified as such represent D. albertina. Both taxa be- long to Mulligan’s (1976) yellow-flowered aneu- ploid group. Draba asprella, a species endemic to Arizona and southern Utah, is represented by few herbarium collections and a confusing chromosome literature. A single count of n = +16 appears in the primary literature and the Indexes to Plant Chromosome Numbers. This count derives from a population in Coconino Co., AZ studied by Rollins and Riiden- berg (1971), which was not identified to variety in the original paper. Rollins (1993) attributes this count to var. asprella and reports an additional, ap- parently undocumented count of n = 16 for var. stelligera. The latter count is critical because it seems to place D. asprella in Mulligan’s (1976) yellow-flowered euploid assemblage, whereas my count of n = 15 (Fig. 2) would suggest an affilia- tion with his aneuploid group. I am confident of my determination, which is based on at least 40 cells from eight individuals. At this point, I am inclined to discount the undocumented euploid report and assign D. asprella to the yellow-flowered aneuploid group. In the upcoming field season, I hope to ob- tain accurate counts for all four varieties and de- termine whether var. stelligera is truly polymorphic with regard to chromosome number. The available literature provides two chromo- some counts for Draba cuneifolia. Rollins and Rut- denberg (1971) report a count of n = 16 from Pecos Co., TX. Although not identified to variety, this col- 2000] WINDHAM: CHROMOSOME COUNTS ON DRABA 23 TABLE 1. CHROMOSOME COUNTS ON DRABA FROM UTAH AND VICINITY. Counts differing from previously published reports are marked by an asterisk. Apparent first counts for a taxon are marked by a double asterisk following the relevant name. Letters before collection numbers identify the following collectors: ER = Eric Rickart; RS = R. Douglas Stone; JT = James Therrien; W = Michael Windham; TW = Theresa Windham; MEW = Maria Windham; MKW = Molly Windham. Herbaria housing voucher specimens are identified by upper case abbreviations (based on Holmgren et al. 1990) following the collection numbers. Draba albertina Greene 2n = 24 UT Emery Co. in South Hughes Canyon on the Wasatch Plateau (T14S, R7E, S30); W & ER 95-185 (UT) n= 12 UT Salt Lake Co. E of Guardsman Pass along State Route 152 in the Wasatch Mts. (T2S, R3E, S25); W 98-320 (MO, NY, UT) Draba asprella Greene var. stelligera O.E. Schulz n= 15, 2n = 30* AZ Coconino Co. along tributary of Bear Wallow Canyon E of Sedona (T17N, R6E, S10); W 95-250 (ASU, BRY, COLO, UT, UTC); W, TW & MKW 98-002 (MO, NY, UT) Draba cuneifolia Nutt. ex Torr. & A. Gray var. cuneifolia n = 15* AZ Yavapai Co. WNW of Sedona on the SW side of Fay Canyon (T18N, RSE, S30); W, JT & MEW 97-005 (MO, UT) n = 15* wT Washington Co. NE of Pinto on low hills overlooking road to Cedar City (T37S, RISW, S26); W 99-008 (MO, UT) Draba juniperina Dorn** n= 11 UT Daggett Co. along Browns Park-Clay Basin road in upper Jesse Ewing Can- yon (T2N, R24E, S1); W 96-152 (MO, NY, UT) n= 11 UT Daggett Co. along State Route 44 on N side of Spring Creek (T2N, R20E, S19); W 99-073 (COLO, MO, UT) n= 11 WY Sweetwater Co. just E of Richards Gap at S edge of Red Creek Basin (T12N, R105W, S22); W 00-012 (ASU, BRY, MO, UT) Draba kassii Welsh** n= 11 UT Tooele Co. in Goshute Canyon on E slope of the Deep Creek Range (T10S, R18W, S36); W 98-277 (ASU, COLO, MO, NY, UT) Draba maguirei C.L. Hitche.** var. maguirei n= 16 UT Cache Co. SE slope of Mt. Magog in the Bear River Range (T14N, R3E); W95-161 (ARIZ, ASU, BRY, COLO, CPH, DAO, ISTC, MO, NY, OGDE UC, US, UT, UTC) Draba nemorosa L. var. nemorosa n=8 UT Summit Co. N base of Windy Ridge on NE slope of the Uinta Mts. (T2N, RI19E, S24); W 99-072 (COLO, MO, NY, UT) Draba rectifructa C.L. Hitche.** n= 12 UT Juab Co. N of Mount Nebo near head of Gibson Creek (T11S, R2E, S19); W 96-204 (UT) Draba sobolifera Rydb.** n= 13 UT Piute Co. S side of Bullion Canyon in the Tushar Mts. (T28S, R5W, S11); W & RS 95-201 (ASU, BRY, COLO, MO, NY, OGDE UT) Draba spectabilis Greene var. spectabilis** n= 10 UT San Juan Co. SE of Gold Basin in the La Sal Mts. (T27S, R24E, S15); W95- 170 (ASU, BRY, COLO, CPH, MO, NY, OGDE UT, UTC) W & ER 97-188 (ISTC, UT) n= 10 UL San Juan Co. NW slope of South Peak in the Abajo Mts. (T34S, R22E); W95-182 (ASU, BRY, COLO, CPH, MO, NY, OGDE UT, UTC) Draba subalpina Goodman & C.L. Hitchc.** n= 13 UT Garfield Co. along tributary of Red Canyon on the Paunsaugunt Plateau (T36S, R42W, S1); W & MKW 92-037 (COLO, MO, UT); W 96-036 (DAO) n= 13 UT Garfield Co. near headwaters of Coyote Hollow on the Paunsaugunt Plateau (T36S, R42W, S1); W 98-129 (MO, NY, UT) n= 13 UT Iron Co. NW slope of Blowhard Mtn. on the Markagunt Plateau (T37S, ROW, S15); W 92-135 (BRY, MO, NY, UT, UTC) 74 MADRONO [Vol. 47 Fics. 1-6. Meiotic chromosome squashes for various Draba species. Solid spherical bodies in Figs. 1, 4, and 6 = nucleoli. Arrows identify overlapping pairs. 1. Diakinesis in D. albertina (n = 12). 2. Late diakinesis in D. asprella var. stelligera (n = 15). 3. Metaphase I in D. cuneifolia var. cuneifolia (n = 15). 4. Diakinesis in D. juniperina (n = 11). 5. Late prophase II in D. maguirei var. maguirei (n 16 at each pole). Faint spherical body near the center of each cluster = nucleolus. 6. Diakinesis in D. spectabilis var. spectabilis (n = 10). lection is presumed to represent var. cuneifolia based on geographic location. Hartman et al. (1975) also report n = 16 for a collection of the typical variety from Dallas Co., TX. Given this history, I was surprised to obtain clear preparations of n = 15 (Fig. 3) for two populations of D. cuneifolia var. cuneifolia from Arizona and Utah. These counts were confirmed in at least five cells from three dif- ferent plants in each population, so it seems likely that the apparent chromosomal polymorphism is real. It is interesting to note that my counts derive from the northwestern portion of the species distri- bution, whereas the two reports of n = 16 represent the southeastern portion of the native range. Further sampling is needed to determine whether chromo- some number truly is correlated with geography in D. cuneifolia. Such an investigation also should en- compass Draba reptans, (Lam.) Fern. which is con- sidered closely related (Hitchcock 1941) or inter- gradient (Welsh 1993) and apparently displays par- 2000] allel variation in chromosome number (Mulligan 1966; Love and Love 1982). Although D. reptans is placed in the aneuploid group by Mulligan (1976), the taxon is white-flowered and probably should be assigned to a separate group (Beilstein personal communication). Draba juniperina is endemic to pinyon-juniper woodlands at the northeastern edge of the Uinta Mountains near ““Three Corners’’, the point where Utah, Wyoming, and Colorado meet. The taxon, long thought to be related to D. oligosperma Hook., because of the shared occurrence of doubly pecti- nate trichomes, has a complex nomenclatural his- tory. It was first separated from the yellow-flowered D. oligosperma under the name D. pectinipila (Rol- lins 1953), a taxon typified on white-flowered spec- imens from alpine habitats in northwestern Wyo- ming. Dorn (1978) pointed out that the petals of D. pectinipila truly are white, but the flowers of pop- ulations from southwestern Wyoming and _ north- eastern Utah are yellow when fresh. Additional morphological features were found to correlate with flower color, geography, and habitat, which led Dorn (1978) to describe the Uinta populations as a new species, D. juniperina. Subsequent studies by Lichvar (1983) seemed to reinforce the distinctions among D. oligosperma, D. pectinipila, and D. juniperina but, in his most re- cent work, Rollins (1993) abandoned this taxono- my. Stating that designating “‘deviant types as in- dependent taxa... has done little to clarify the na- ture of the species as a whole”’ (Rollins 1993), he once again synonymized the segregate taxa under D. oligosperma. Kartesz (1994) followed suit, though Welsh (1986a) maintained juniperina as a variety of D. oligosperma without further comment. There has been little use of this combination, how- ever, because var. juniperina is described as having ““petals evidently white’’ (Welsh 1993), a character state not found in Utah specimens. The chromosome counts presented here for Dra- ba juniperina (Fig. 4) provide valuable insight into the taxonomy of this contentious species complex. Studies at two widely separated localities in Dag- gett Co., UT and one site in Sweetwater Co., WY revealed that D. juniperina is a sexually-reproduc- ing taxon with a chromosome number of n = 11. This is one of two numbers not previously docu- mented in Mulligan’s (1966, 1976) aneuploid se- ries, and clearly establishes this taxon as a member of the yellow-flowered aneuploid group. Draba oli- gosperma, on the other hand, is an apomictic taxon (Mulligan and Findlay 1970) with three reported chromosome numbers: 1) 2n = 32 from Alberta (Chinnappa and Chmielewski 1987), 2) 2n = +60 from Wyoming (Rollins 1966), and 3) 2n = 64 from seven populations in Alberta and one in Yu- kon Territory (Mulligan 1972). These numbers in- dicate that D. oligosperma belongs to Mulligan’s (1976) yellow-flowered euploid group. The difference in chromosome base numbers be- WINDHAM: CHROMOSOME COUNTS ON DRABA 25 tween D. oligosperma (x = 8) and D. juniperina (x = 11) is not trivial. The former is not a simple polyploid derivative of the latter and, if Mulligan (1976) is right in his assessment of relationships, they may belong to different major lineages. The two taxa are easily distinguished using the charac- ters listed by Dorn (1978) and Lichvar (1983), even where their ranges overlap. Even if they grew to- gether, which they apparently do not, there would be no opportunity for hybridization because D. oli- gosperma is apomictic and apparently does not pro- duce functional gametes (Mulligan and Findlay 1970). All of this provides a strong argument for maintaining Draba juniperina as a distinct species. Draba kassii is a very rare species endemic to a few canyons in the Deep Creek Mountains of west- ern Utah. Its relationships are obscure, with Rollins (1993) stating that it ‘“‘is not closely enough related to any known species of Draba to allow inferences as to its phylogeny.’’ Comparisons have been drawn to D. asprella (Welsh 1986b) and D. stan- dleyi J. EF Macbr. & Payson (Rollins 1993), though both authors suggest that the similarities may be superficial. Chromosome numbers have the poten- tial to play a crucial role in determining the rela- tionships of this species. At least 20 cells from five different plants clearly establish that the chromo- some number of D. kassii is n = 11 (Table 1). This number, which establishes the taxon as a member of Mulligan’s yellow-flowered aneuploid group, would seem to rule out a direct phylogenetic link to D. asprella (n = 15, 16?). The possibility of a relationship to D. juniperina, the only other species known to have n = I1, is intriguing. However, the two taxa do not appear closely related morpholog- ically, and any hypothesis of relationships will re- main speculative until additional Draba species (in- cluding D. standleyi) have been sampled chromo- somally. The phylogenetic affinities of Draba maguirei are as contentious as those of D. kassii, and it ap- pears that no recent author has ventured to discuss its possible relationships. On first describing the species, Hitchcock (1941) stated that it “‘is very striking and quite unlike any of the other Drabas from its immediate vicinity. Its closest relatives are probably those of the ventosa group...’’. Of the eight taxa comprising Hitchcock’s ventosa group, chromosome counts have been published for two (D. ventosa Gray and D. ruaxes Payson & St. John) and a third (D. sobolifera) is reported here. All three belong to Mulligan’s (1976) yellow-flowered aneuploid group with chromosome numbers based on x = 12 and 13. Thus, it is surprising to find that D. maguirei var. maguirei shows a euploid count of n = 16 (Fig. 5). Additional chromosome counts on D. maguirei and other members of Hitchcock’s ventosa group are needed to resolve this apparent conflict. Draba nemorosa L., a species of widespread oc- currence in both North America and Eurasia, was 26 MADRONO assigned by Mulligan (1976) to his yellow-flowered euploid group. All populations analyzed chromo- somally have shown n = 8, regardless of geograph- ic origin. In North America, there have been four counts from Alberta (Packer 1964; Mulligan 1966, 1975), one from Manitoba (Léve and Love 1982), two from Ontario (Mulligan 1975), and two from Saskatchewan (Mulligan 1966, 1975). It appears that my determination of n = 8 (Table 1) from Dag- gett Co., UT is the first report for the United States. None of the previous North American reports spec- ify variety, though most are surely var. nemorosa, the taxon to which my count is assigned following the taxonomy of Kartesz (1994). Although the gla- brous-fruited form (var. leiocarpa) is considered taxonomically insignificant by many authors, there does appear to be some geographic integrity to its occurrence. Therefore, it seems wise to maintain the distinction until North American populations are studied adequately. Although Hitchcock (1941) considers Draba rec- tifructa to be a close relative of euploid D. nemo- rosa, little evidence is cited to support such an as- sociation. Instead, it appears to be very closely re- lated to D. albertina, distinguished from that spe- cies mainly by its pubescent upper stems and pedicels. Ongoing studies of populations in north- ern Utah suggest that D. rectifructa and D. alber- tina hybridize when growing in close proximity. Thus, it is not surprising to find that D. rectifructa has a chromosome number of n = 12 (Table 1), identical to that of its putative aneuploid relative. Draba sobolifera, endemic to the Tushar Moun- tains of southern Utah, is considered a member of Hitchcock’s (1941) ventosa group with close affin- ities to D. cusickii Robinson & D. E. Schulz (Rol- lins 1993). The chromosome number of the latter species is unknown, but the two members of the ventosa complex previously reported, D. ventosa and D. ruaxes, show 2n = 36 and 2n = 72 (Mul- ligan 1971). They are considered to be triploid and hexaploid respectively, with a base number of x = 12. Cytological studies on a population of D. so- bolifera from Piute Co., UT reveal that it is a sex- ually-reproducing taxon with a chromosome num- ber of m = 13. This is the last number to be doc- umented in Mulligan’s (1966, 1976) aneuploid se- ries, and it firmly establishes this species as a member of the yellow-flowered aneuploid group. There are three previous chromosome counts for Draba spectabilis, all from Colorado and all as- signed to var. oxyloba (Greene) Gilg. & O. E. Schulz by Price (1980). The earliest report (Mulli- gan 1966) of n = 10 seemed to indicate that the species belonged in the aneuploid group. However, two subsequent counts of n = 16 andn = 16 + 2 by Price (1980) suggest an affinity to the yellow- flowered euploid assemblage. My determinations, apparently the first for var. spectabilis, are from two widely separated populations in San Juan Co., UT (Fig. 6). They agree with Mulligan’s (1966) report [Vol. 47 of n = 10 and point out the need for further sam- pling to determine the relationships and proper tax- onomy of D. spectabilis. Draba subalpina generally is restricted to a sin- gle geologic stratum, the Claron Formation of Bryce Canyon National Park and vicinity. Although recent authors have said little regarding its probable relationships, Hitchcock (1941) states that its clos- est relative is D. oreibata J. K Macbr. & Payson, a species under which it was subsumed prior to 1932. The latter taxon is endemic to central Idaho in its typical form, is similarly white-flowered, and shows a chromosome number of n = 16 (Hender- son et al. 1980). In light of its proposed relation- ships and the assumption that D. subalpina was a member of the white-flowered euploid group, the actual chromosome number was unexpected. Based on at least ten cells from five individuals in each of three populations, the chromosome count of Draba subalpina proves to be n = 13 (Table 1). Whether D. subalpina belongs to a relatively rare, white-flowered aneuploid group or is more closely related to some of its yellow-flowered congeners remains to be determined. The close proximity (ca. 60 km) of D. sobolifera, the only other species known to exhibit n = 13, raises intriguing possi- bilities regarding the relationships of white- and yellow-flowered aneuploids in Draba. Even with the small sample size of this nascent effort, it is clear that the taxonomic composition of the Intermountain Draba flora is quite different from the intensively studied assemblage of Canada and Alaska. In the latter, Mulligan (1976) assigned 17 species to his white-flowered euploid group, nine to the yellow-flowered euploid assemblage, and 13 to his yellow-flowered aneuploid group. In my sample from Utah, Wyoming, and Arizona, white-flowered euploids are not represented (unless D. cuneifolia belongs here) and yellow-flowered euploids are rare, comprising only D. nemorosa and possibly D. maguirei. Seven of the Intermountain taxa belong to the yellow-flowered aneuploid as- semblage and the remaining taxon (D. subalpina) is a white-flowered aneuploid of uncertain affinity. A growing number of chromosome counts for the region suggests that the Intermountain West may be a center of diversity for aneuploid Draba. With the discovery of both nm = 11 and n = 13 among local endemics, a complete series of base numbers extending from 8 to 16 has been docu- mented. Only n = 9 and n = 14 are missing from my sample, and those numbers have been con- firmed in other taxa from the region. This means that every major step in the process of aneuploid evolution is preserved among the Draba species of the Intermountain West. In this assemblage of Dra- ba species, we have an unprecedented opportunity to study the processes of chromosomal evolution and speciation in plants. With further cytological sampling and concurrent DNA studies of the group, we soon may be in a position to elucidate the evo- 2000] lutionary history of this interesting and diverse set of organisms. ACKNOWLEDGMENTS I wish to express my gratitude to the Utah Museum of Natural History (UMNH) and the staff of Wasatch-Cache National Forest (especially Wayne Padgett) for providing funding that made this work possible. The job of locating the rarer taxa and estimating the dates of meiosis was greatly facilitated by access to the extensive collections of the S. L. Welsh Herbarium (Brigham Young University). Special thanks are due R. Douglas Stone, who piqued my interest in Draba and whose help and dedication to the project (and ability to find buds undergoing meiosis) was a major factor in its success. I also thank Erin Kincaid (Exhibits Department, UMNH) for help with the figures, and Dr. Christopher Haufler (University of Kansas) for helpful comments on an earlier version of this manuscript. LITERATURE CITED CHINNAPPA, C. C. AND J. G. CHMIELEWSKI. 1987. Docu- mented chromosome numbers 1987: 1. Miscellaneous counts from western North America. Sida 12:409— 417. Dorn, R. D. 1978. A new species of Draba (Cruciferae) from Wyoming and Utah. Madrono 25:101—103. FEDEROV, A. (ed.). 1974. Chromosome Numbers of Flow- ering Plants. Otto Koeltz, Koenigstein. GOLDBLATT, P. (ed.). 1981. Index to plant chromosome numbers 1975—1978. Monographs in Systematic Bot- any from the Missouri Botanical Garden 5:1—553. . (ed.). 1984. Index to plant chromosome numbers 1979-1981. Monographs in Systematic Botany from the Missouri Botanical Garden 8:1—427. . (ed.). 1985. Index to plant chromosome numbers 1982-1983. Monographs in Systematic Botany from the Missouri Botanical Garden 13:1—224. . (ed.). 1988. Index to plant chromosome numbers 1984-1985. Monographs in Systematic Botany from the Missouri Botanical Garden 23:1—264. AND D. E. JOHNSON. (eds.). 1990. Index to plant chromosome numbers 1986-1987. Monographs in Systematic Botany from the Missouri Botanical Gar- den 30:1—243. AND . (eds.). 1991. Index to plant chro- mosome numbers 1988-1989. Monographs in Sys- tematic Botany from the Missouri Botanical Garden 40:1-238. AND . (eds.). 1994. Index to plant chro- mosome numbers 1990-1991. Monographs in Sys- tematic Botany from the Missouri Botanical Garden 51:1—267. AND . (eds.). 1996. Index to plant chro- mosome numbers 1992-1993. Monographs in Sys- tematic Botany from the Missouri Botanical Garden 58:1-276. AND . (eds.). 1998. Index to plant chro- mosome numbers 1994-1995. Monographs in Sys- tematic Botany from the Missouri Botanical Garden 69:1—208. HARTMAN, R. L., J. D. BACON, AND C. E BOHNSTEDT. 1975. Biosystematics of Draba cuneifolia and D. platycar- pa (Cruciferae) with emphasis on volatile and flavo- noid constituents. Brittonia 27:317—327. HENDERSON, D., A. CHOLEWA, AND N. REESE. 1980. in A. WINDHAM: CHROMOSOME COUNTS ON DRABA 2 Léve (ed.), IOPB Chromosome Number Reports LXVUI. Taxon 29:534. HOLMGREN, P. K., N. H. HOLMGREN AND L. C. BARNETT. (eds.). 1990. Index Herbariorum. Part I: The Herbaria of the World, 8th ed. New York Botanical Garden, Bronx, NY. KARTESZ, J. T. 1994. A Synonymized Checklist of the Vas- cular Flora of the United States, Canada, and Green- land, 2nd ed. Timber Press, Portland, OR. LICHVAR, R. W. 1983. Evaluation of Draba oligosperma, D. pectinipila, and D. juniperina complex (Crucifer- ae). Great Basin Naturalist 43:441—444. Love, A. AND D. Love. 1982. Pp. 120-126 in A. Léve (ed.). IOPB Chromosome Number Reports LXXIV. Taxon 31:119—128. Moore, R. J. (ed.). 1973. Index to plant chromosome numbers for 1967-1971. Regnum Vegetabile 90:1— 539. . (ed.). 1974. Index to plant chromosome numbers for 1972. Regnum Vegetabile 91:1—108. . (ed.). 1977. Index to plant chromosome numbers for 1973/74. Regnum Vegetabile 96:1—257. MULLIGAN, G. A. 1966. Chromosome numbers of the fam- ily Cruciferae. III. Canadian Journal of Botany 44: 309-319. . 1970a. Cytotaxonomic studies of Draba glabella and its close allies in Canada and Alaska. Canadian Journal of Botany 48:1431—1437. . 1970b. A new species of Draba in the Kanan- askis Range of southwestern Alberta. Canadian Jour- nal of Botany 48:1897—-1898. . 1971a. Cytotaxonomic studies of the closely al- lied Draba cana, D. cinerea, and D. groenlandica in Canada and Alaska. Canadian Journal of Botany 49: 89-93. . 1971b. Cytotaxonomic studies of Draba species in Canada and Alaska: D. ventosa, D. ruaxes, and D. paysonii. Canadian Journal of Botany 49:1455—1460. . 1972. Cytotaxonomic studies of Draba species in Canada and Alaska: D. oligosperma and D. incerta. Canadian Journal of Botany 50:1763—1767. . 1974. Cytotaxonomic studies of Draba nivalis and its close allies in Canada and Alaska. Canadian Journal of Botany 52:1793-1801. . 1975. Draba crassifolia, D. albertina, D. nemo- rosa, and D. stenoloba in Canada and Alaska. Ca- nadian Journal of Botany 53:745-—751. . 1976. The genus Draba in Canada and Alaska: key and summary. Canadian Journal of Botany 54: 1386-1393. AND J. N. FINDLAY. 1970. Sexual reproduction and agamospermy in the genus Draba. Canadian Journal of Botany 48:269—270. ORNDUwFF, R. (ed.). 1967. Index to plant chromosome num- bers for 1965. Regnum Vegetabile 50:1—128. . (ed.). 1968. Index to plant chromosome numbers for 1966. Regnum Vegetabile 55:1—126. PACKER, J. G. 1964. Chromosome numbers and taxonomic notes on western Canadian and Arctic plants. Cana- dian Journal of Botany 42:473—494. Payson, E. B. 1917. The perennial scapose Drabas of North America. American Journal of Botany 4:253—267. PricE, R. A. 1980. Draba streptobrachia (Brassicaceae), a new species from Colorado. Brittonia 32:160—169. ROLLINS, R. C. 1953. Draba on Clay Butte, Wyoming. Rhodora 55:229—235. . 1966. Chromosome numbers of Cruciferae. Con- tributions of the Gray Herbarium 197:43-—65. 28 MADRONO [Vol. 47 . 1993. The Cruciferae of Continental North Amer- WELSH, S. L. 1986a. New taxa and combinations in the ica: Systematics of the Mustard Family from the Arctic Utah Flora. Great Basin Naturalist 46:254—260. to Panama. Stanford University Press, Stanford, CA. . 1986b. New taxa in miscellaneous families from AND L. RUDENBERG. 1971. Chromosome numbers Utah. Great Basin Naturalist 46:261—264. of Cruciferae. II. Contributions of the Gray Herbari- . 1993. Cruciferae. Pp. 275-325 in S. L. Welsh, N. um 201:117-133. D. Atwood, S. Goodrich & L. C. Higgins (eds.), A SCHULZ, O. E. 1927. Cruciferae-Draba et Erophila. in: A. Utah Flora, 2nd ed. Jones Endowment Fund, Brigham Engler (ed.), das Pflanzenreich IV. 105(Heft 89):1—395. Young University, Provo, UT. MADRONO, Vol. 47, No. 1, pp. 29-42, 2000 A FLORISTIC STUDY OF TSEGI CANYON, ARIZONA SUSAN HOLIDAY Northern Arizona University, Flagstaff, AZ 86011 ABSTRACT The purpose of this study is to list the vascular flora present in Tsegi Canyon, Arizona, and to describe any change in flora that may have happened during the past hundred years. Plants were collected during the years 1994-1997. Three hundred and ten species representing seventy three families are reported to occur within the Tsegi drainage. Three percent of these species are endemic to the Colorado Plateau and twelve percent of the species found are non-native. A change in floristic composition is found to have occurred in the last century, correlated with a shift in habitat types in the canyon. The canyon of the late 1800’s had a slowly moving stream and marshes along a continuous alluvium. The present canyon has a faster moving stream that has eroded much of the alluvium to bedrock. One species, Cymopterus beckii is a new report for Arizona and also is listed as a candidate for rare and endangered species status. Floristics on the Colorado Plateau have not been widely studied, with perhaps the exception of the Grand Canyon (Phillips et al. 1987). The reason for this may be the remoteness of the area from civi- lization and difficulty in traveling to many parts of the plateau due to lack of roads and extremes in temperature. Although most of the plateau vegeta- tion consists of pinion and juniper in higher ele- vations and desert scrub in the lower, there are also canyons which harbor a very different flora, in- cluding some relict populations, as well as seeps and alcoves with their unique flora. The purpose of this paper is to describe one of the canyons on the plateau that includes year-round, running water as a small stream, as well as seeps and alcoves. Although this is the first compilation of flora done of Tsegi Canyon, the flora has been studied and described by various researchers previously. One of the first descriptions of the flora of the can- yon was done by Clute (1920). J. T. Hack (1945) also studied the canyon as part of his documenta- tion of the erosion/deposition cycles in northern Ar- izona. He, along with Dean (1969) and Weatherill (1953) describe the erosion of this area that oc- curred within the last hundred years, which could possibly be related to the environmental changes that occurred in the southwest during the first part of the century (Hastings & Turner 1965). An in- ventory was done for the park service in the 1970’s, with Brotherson et al. (1978) publishing a flora of Navajo National Monument. Historical collections were examined at the Walter B. McDougall Her- barium, the Deaver Herbarium, the U. of A. Her- barium, and at the herbarium at Navajo National Monument. Study area description. The Tsegi Canyon drain- age system is a complex of canyons that forms the headwaters of Laguna Creek, one water supply for the town of Kayenta, AZ. Tsegi Canyon, whose name in the Navajo language means ‘in the rocks’ or canyon, is in Navajo County in the northeast corner of Arizona, S40000E, 4054000N, 36°40'N, 110°30'W. Tsegi Canyon includes in its boundaries two of the three sections of Navajo National Mon- ument, Betatakin and Keet Seel. The head of Tsegi Canyon is located on the Or- gan Rock Monocline. This is an uplift that is fol- lowed by Highway 160, in Long Valley between the Shonto Plateau and Black Mesa. The canyon is a complex drainage cut into the Shonto Plateau on the east and Skeleton Mesa to the west. Six differ- ent geological formations are visible in the canyon. The top of the plateau is made of Navajo Sand- stone, the formation responsible for the magnificent sandstone cliffs at the top of the canyon. This layer is of Jurassic age and made of wind-blown sand and dunes. The layer just under the Navajo Sand- stone is the Kayenta formation consisting of gray- red sandstone and some clay shale. This layer is about 61 meters thick at the head of the canyon. Because the Navajo sandstone is porous, it allows percolation of water onto the top of the less porous Kayenta formation. The water moves laterally over the Kayenta formation to flow out and form seeps on the canyon walls. Exfoliation of the sandstone above the seeps causes the formation of the alcoves that were utilized by the Anasazi. The layers under the Kayenta formation are a part of the Glen Can- -yon group. In the canyon, it is represented by the Lukachukai member of the Wingate sandstone. This is a reddish-brown, cliff-forming sandstone of- ten responsible for the rockfall in the canyon. Un- der this is the Chinle formation. It is represented by two strata, the Churchrock member and the Owlrock member. The Churchrock member con- sists of brownish-red siltstone, mudstone, and fine- grained sandstone with small, white spots and streaks. Below is the Owlrock member which con- sists of reddish-brown siltstone and mudstone, and greenish-gray claystone laid down in the Triassic Age. There is also a limited exposure located north of the Tsegi Hotel, of Petrified Forest member with red, purple, and green/gray betonite claystone. On 30 MADRONO the bottom of the canyon is alluvial fill (Beaumont & Dixon 1964). There are three layers of alluvial deposition in Tsegi Canyon. The oldest layer is the Jeddito for- mation that was laid down before 3500 BC. This is overlaid by the Tsegi formation that was laid down between 3500 BC and AD 1300. The youngest lay- er of alluvium is the Naha formation that was laid down between 1450 and 1880 (Hack 1945). Al- though there is a small amount of post-1900 allu- vium, at the present there is more erosion happen- ing than deposition. The reasons for the deposition and erosion cycles is not definitely known. It has been suggested that rainfall, climate fluctuations, and land management practices may all be contrib- utors to this cycle, with perhaps, the climate being the most influential factor (Clay-Poole 1989). How- ever there is also evidence that human activity has affected arroyo cutting in the canyon. The cutting of the Tsegi-Naha arroyo in Keet Seel was preceded by the clearing of an aspen forest at the bottom of the canyon in the 1200’s. After the abandonment of Keep Seel, the area was redeposited with alluvium and recolonized by Quercus gambelii (Dean 1969). However, with the new arroyo cutting of the present century, the distribution of oaks have retreated to the upper side alluvium. The climate at Tsegi Canyon is arid with cold winters and hot summers. The daily average tem- perature at Tsegi is Celsius. Temperatures vary from highs of 340 to 380° C in July to lows of —230 to 130° C in the winter. The frost-free season averages about 155 days. Precipitation in the can- yon is variable from year to year. Over a 17-year span, the rainfall at the Betatakin Monument ranged from a low to 17.3 centimeters to a high of 47.7 centimeters (U.S.D.C. 1979-1996). The variability is caused by differences in winter precipitation and is also enhanced by the fact that monsoon rains are very spotty and usually do not equally wet all parts of the canyon (Dean 1969). METHODS Seventeen collecting trips to Tsegi Canyon were made between 1994 and 1997. The main focus of collecting was to include as many species as pos- sible for the floristic list. The collections were done between the months of April and October, as most of the vascular flora is dormant during the winter. Lower Tsegi Canyon was visited April 23, 1994; May 30, 1994; July 16, 1994; and August 12, 1994. Wildcat Canyon and Lone Cottonwood Canyon were visited June 19, 1995, July 26, 1995, and Au- gust 11, 1995. Upper Tsegi Canyon, Fir and Beta- takin Canyon were visited June 3, 1994; August 7, 1994; September 16, 1994; May 21, 1995; May 29, 1996; and August 8, 1996. Dowozhiebito and Keet Seel Canyons were visited June 18, 1994; May 5, 1996; September 28, 1996; and June 26, 1997. Four of the trips, May 29, 1996; August 8, 1996; Sep- [Vol. 47 tember 28, 1996; and June 26, 1997 were made at the request of the National Park Service as a survey for rare and endangered species. At this time, col- lections were made at Betatakin National Monu- ment and Keet Seel National Monument. However, the majority of the collecting was done on Navajo Tribal land. All specimens were pressed, dried, and stored at the Deaver Herbarium (ASC) at Northern Arizona University, with duplicates sent to the Navajo Trib- al Heritage Herbarium. Specimens were named fol- lowing Kartesz (1994). Previous collections located at the Deaver Herbarium (ASC), the museum of Northern Arizona Herbarium (MNA), and the Uni- versity of Arizona Herbarium (ARIZ) were used for comparison. A classification model, modified from Rowlands’ Colorado Plateau Vegetation Assessment and Clas- sification Manual was used to describe the various vegetation assemblages in the canyon. I determined dominant/co-dominants by using the largest sized plants that appeared to be the most abundant (Bon- ham 1989; Rowlands 1994). The other notable spe- cies are included to help describe the assemblages. Boundaries were determined using physical bound- aries, such as terrace levels and natural altitudinal separations. The assemblages were mapped out us- ing USGS topographic maps of the Betatakin, Keet Seel, and Marsh Pass quadrangles with additional information relating to location and size taken from aerial photos. The maps were drawn by hand and the areas of the vegetation assemblages determined using a Mackintosh scanner and NIH Image area analysis. RESULTS There were 310 species found in Tsegi Canyon during the study years. In comparison, 518 species were found in Canyon deChelly National Monu- ment (Halse 1973; Harlan 1976), 293 species were found at Navajo National Monument (Brotherson 1978), 376 species were found in Volunteer and Sycamore Canyon (Shilling 1980), and 326 species were found in the Walnut Canyon National Monu- ment (Arnberger 1947; Spangle 1953; Joyce 1976). There are eight vegetation assemblages described in this study for Tsegi Canyon which include the Pseudotsuga assemblage, Populus tremuliodes as- semblage, the Pinus edulis/Juniperus osteosperma assemblage, Quercus gambelii assemblage, Atri- plex/Artemisia assemblage, the Juncus marshland assemblage, Betula occidentalis assemblage, the Gutierrezia assemblage, and the Pucinnellia bad- lands assemblage. This classification is split into more detail than the USFS digitized classification system (Brown 1980) and some of the assemblages are combined compared to Rowlands (1994). Be- cause there is no standardized way to classify the complex systems of a riparian canyon on the Col- orado Plateau, this scheme was created based on both classification systems. 2000] Pseudotsuga occurred in shaded, mostly north and west facing areas in the side canyons of Tsegi. This assemblage consists of about 8% of the total cov- erage of vegetation sampled in the canyon. Most Pseudotsuga menziesii individuals encountered were older trees. A study of whether or not seedlings are present in high enough numbers to replace older trees would be of value for this area. The Pseudo- tsuga often graded down into populations of Populus tremuliodes in the moist side canyons, and some- times an individual fir could be found in the upper regions of a stream bed. Rarely, Pinus ponderosa could be found growing among the firs. In the upper Keet Seel canyon, few individuals of Abies concolor were found among the fir. Shrubs growing in this assemblage include Ssymphoricarpos oreophilius, Ribes cereum, Ribes inerme, and Amelanchier alni- folia. Herbaceous plants include Antennaria parvi- folia, Mahonia repens, Corydalis aurea, Galium aparine, and Valeriana acutiloba. This assemblage is included in the Cold Temperate Forest and Wood- lands, Rocky Mountain Montane Conifer Forest, Douglas fir-White Fir series, 122.311. Pseudotsuga menziesii Association in the Digitized Systematic Classification used by the Forest Service (Brown 1980). In Rowlands (1994), this is part of the Mon- tane Zone, Forest and Woodland Formation, and Pseudotsuga menziesii Series. Populus tremuloides creek bottoms include plants adapted to a shady, moist environment. Pop- ulus tremuloides covers less than 1% of the area studied, occurring in the upper areas of the Beta- takin-Fir Canyon side canyons. This assemblage is bounded above mostly by Pseudotsuga menziesii on the shady sides of the canyons and Quercus gambelii on the more sunny sides. Betula occiden- talis can be found further down the creek bed if there is running water, otherwise it usually ends abruptly in Pinus/Juniperus or sandy creek bottom vegetation. Other trees that can be found in this assemblage include Pseudotsuga menziesii and Prunus virginiana. Shrubs present include Rhus aromatica, Symphoricarpos oreophilus, Cornus sericea, Arctostaphylos pungens, Ribes leptanthum, Rosa wocodsii, Salix exigua, and Salix lasiolepis. Herbaceous plants include Equisetum arvense, Car- ex athostachya, Eleocharis palustris, Juncus arcti- cus, Fritillaria atropurpurea, Smilacina stellata, Poa praatensis, Erigeron speciosus, Silene menzie- sul, Lathyrus brachycalyz, Androsace septentrion- alis, Clematis ligusticfolia, Thalictrum fendleri, Heuchera parvifolia, and Mimulus rubellus. The Closest classification of this assemblage found in the USFS classification is the Great Basin Interior Strand, which is very non-specific (Brown 1980). Rowlands (1994) has a Populus tremuloides Series in his Forest and Woodlands Formation. However, I do not feel that this quite fits as it describes aspen in a pine forest where the pines will eventually suc- ceed the aspen. Here the aspen are the climax spe- cies, with aspen saplings replacing older trees. HOLIDAY: A FLORISTIC STUDY OF TSEGI CANYON, ARIZONA 31 The most common assemblage found on the Shonto Plateau is the Pinus edulis/Juniperus osteos- perma. This assemblage forms 50% of the plant communities mapped. In the canyon, this assem- blage is found in many of the upper ledges and south facing canyon sides. This assemblage surrounds the Monument headquarters and is common on the top of Skeleton Mesa. Shrubs that can be found in as- sociation with this assemblage include Ephedra vir- idis, Chrysothamnus nauseosus, Shepherdia rotun- difolia, Fendleera rupicola, Amelanchier utahensis, Cercocarpus intricatus, Cerccocarpus montanus, Holodiscus dumosus, and Yucca angustissima. Her- baceous plants include Allium macropetalum, Calo- chortus aureus, Bouteloua gracilis, Cymopterus acaulis, Artemisia dracunculus, Psilostrophe spar- siflora, Heterotheca villosa, Arabis perennans, Streptanthus cordatus, Echinocereus triglochidiatus, Opuntia polycantha, Astragalus ceramicus, Mirabi- lis multiflora, Ipomopsis aggregata, Castilleja lina- riifolia, Cordylanthus wrightii, and the parasitic Phoradendron juniperinum. The Pinus edulis/Junip- erus osteospermus assemblage is part of the Forest and Woodland Formation (Rowlands 1994) and clas- sified by the USFS as part of the Great Basin Conifer Woodland, Pinion-Juniper Series (Brown 1980). The area described as the Quercus gambelii as- semblage occurs on the upper terraces in side can- yons draining from west to east, though there are exceptions in the upper areas included in this study. This assemblage is estimated to cover about 2% of the study area. The main component of this assem- blage are thickets of Quercus gambelii with occa- sional larger trees of Quercus included. This area’s shading consists of thick leaf litter and generally is not as diverse as the more expansive Pinus/Junip- erus assemblage. Many plants found here grow in spaces in the thicket where more light can pene- trate. Other shrubs included in this assemblage are Ribes cereum and Prunus virginiana. The herba- ceous cover includes Juncus arcticus, Smilacinia stellata, Bouteloua curtipendula, Opuntia phaea- cantha, Erigeron utahensis, and Lathyrus brachy- calyx. Rowlands (1994) classifies this as the Tall Shrubland Formation, Quercus gambilii Series. The USFS Digitized Classification (Brown 1980) in- cludes this in the Cold Temperate Scrublands, Great Basin Montane Scrub, Oak-Scrub Series. The Atriplex/Artemisia assemblage dominate on the lower terraces above the main creek and at the mouths of the side canyons. This is the second larg- est vegetation assemblage covering about 22% of the total canyon. Although there may be an occa- sional Pinion or Juniper tree associated with this assemblage, the dominant larger plants are the shrubs Atriplex canescens and Artemisia tridentata. There are also occasional lone Elaegnus angusti- folia individuals of unknown origin, possibly plant- ed by members of the family that use the canyon (Melberg 1988), or distributed by birds. There is also a small stand of about four Ulmus pumila at 32 MADRONO one site, planted along the side of the dirt road. Shrubs associated with this area include chryso- thamnus viscidiflorus, Sarcobatus vermiculaatus, Poliomintha incana, and Cercocarpus intricatus. Herbaceous plants include Elymus smithii, Artemis- ia frigida, Helianthus peetiolaris, Senecio multilo- batus, Cryptantha crassisepala, Descurainia pin- nata, Salsola iberica, Astragalus amphioxys, Pha- celia ivesiana, Spaeralcea parvifolia, Mirabilis ox- ybaphoides, Gayophytum racemosum, Orobanche multiflora, Erigonum cernuum, Ranunculus testi- culatus, and Verbeena bracteata. Rowlands (1994) split these two shrub species into two separate se- ries in his classification scheme. However, in Tsegi Canyon, the two species were found in many places together, and thus would be hard to separate into separate assemblages. The closest classification found in the USFS classification manual (Brown 1980) is defined as the Cold Temperate Desertlands, Great Basin Desertscrub, Mixed Scrub Series. The Juncus marshland assemblages were usually found at the bottoms of most small side creek drainages and flattened areas below seeps. The Jun- cus assemblage is a minor component of the can- yon, consisting of less than 1% of the canyon sur- veyed. These areas were often heavily used by cat- tle and damaged by trampling. This assemblage had only occasional Elaeagnus angustifolia or Tamarix ramosissima. There was also one example of a Populus fremonti tree at the edge of one marshland. Herbs found in this assemblage include Juncus arc- ticus, Equisetum arvense, Polypogon monspelien- sis, Scirpus pungens, Aster frondosus, Conyza can- adensis, Taraxacum officinale, Crypthantha inae- quata, Lepidium virginicum, Epilobium ciliatum, Plantago major, and Ranunculus cymbalaria. Row- lands (1994) classifies this as a Marshland Forma- tion, Juncus arcticus Series. Using the USFS cias- sification this assemblage would be included in the Cold Temperate Marshlands, Great Basin Interior Marshland, Rush Series (Brown 1980). Betula occidentalis creek bottom assemblage was found only in the areas of Betatakin-Fir Canyon drainages and some north-draining side canyons of Keet Seel and upper Dowozhiebito Canyon. This assemblage accounted for about 1% of the area mapped. In the Betatakin area it was bounded above by Populus tremuoides and below by Elaeagnus angustifolia. In other side-canyons, Bet- ula was the uppermost tree species on the drainage floor. The Betula commonly grew along both sides of drainages that included year-round running wa- ter. Another tree species associated with this assem- blage was Acer negundo. Salix monticola and S. lasiolepis were also found growing among the birches. Herbaceous plants found in this assem- blage include Corallorhiza maculata, Toxicoden- dron rydbergii, and Chenopodium album. Like the Quercus assemblage, the density of the trees tend to shade the floor of the creek. This, along with occasional high water levels, tend to limit the num- [Vol. 47 ber of herbs present. Rowlands (1994) classifies this in the Montane Zone, Tall Shrubland Forma- tion, Betula occidentalis series. However, this would be more applicable if it was placed in a ri- parian formation, which is not included in this clas- sification scheme. In the USFS (Brown 1980) there is a Cold Temperate, Great Basin Interior Strand, which includes all riparian vegetation in the Great Basin Biome. The bottoms of the major drainages are classified as the Gutierrezia stream bottom assemblage. This assemblage covers about 18% of the canyon. The soils in this area are characterized by sandy depos- its that are typically scoured at least once a year by flooding. There is also quicksand after floods and other areas that are devoid of vegetation because of animal or human (automobile) use. This assem- blage also includes areas of bare sandstone and low, dry, sandy dunes. Shrubs that grow here include Gutierrezia sarothrae, Chrysothamnus depressus, and Artemisia frigida. Herb species that grow here include Equisetum hyemale, Chenopodium lepto- phyllum, Salsola iberica, Astragalus amphioxys, Nama retrorsum, Tripterocalyx micranthus, Oeno- thera pallida, and Verbascum thapsus. Rowlands (1994) has a Gutierrezia sarothrae series in his classification under the Low Shrubland Formation. The USFS (Brown 1980) includes this area under the Great Basin Interior Strand. Downstream the Betula individuals in the Beta- takin drainage is a stand of Elaeagnus angustifolia. Here, E. angustifolia is growing along both sides of the creek. Whether E. angustifolia is displacing the birch or growing in an area that is for some reason too low in the drainage for the birch is un- known. However, since E. angustifolia is an intro- duced species, and there is an area of interface be- tween the two species, and because the Elaeagnus seems to be spreading in the canyon (Melberg, 1994), the former possibility needs to be examined. Elsewhere in the canyon, E. angustifolia is mostly present as individual trees or young plants. Young plants were found growing in the Keet Seel drain- age, about seven miles from Betatakin, and other young plants were found in side canyons across and above the Betatakin drainage. This is one species that needs to be watched closely because it seems to be able to colonize some local areas and possibly out-compete native growth. The Puccinellia assemblage consists of few plant species growing on betonite clays. This assemblage covers less than 1% of the area mapped. The largest example of this assemblage is in upper Wildcat Canyon on the north-west facing side of the can- yon. Most of the plants are concentrated near a small seep. The area grades into the Pinus/Junip- erus assemblage above it on the canyon sides. Plants included here are Apocynum cannabinum and Puccinellia distans. The closest classification in Rowlands (1994) would probably be the Sub- montane Barren Formation 1408.03. In the USFS 2000] TABLE 1. HOLIDAY: A FLORISTIC STUDY OF TSEGI CANYON, ARIZONA 55 A COMPARISON OF THE TEN LARGEST FAMILIES OF VASCULAR FLORA OF TSEGI CANYON, PETRIFIED FOREST, N.P., CAPITOL REEF N.P., GRAND CANYON N.P., AND THE NAVAJO NATION. *Family not in top ten % flora, 'Kierstead, 1981, *Heil et al., 1993, *Phillips et. al., 1987, ‘Mayes and Rominger, 1994. Petrified! Capital? Grand? Family Tsegi Canyon Forest N.P. Resch N:P: Canyon N.P. Navajo* Nation Asteraceae 1920 19 20.4 16.6 17.6 Poaceae 12 19.6 14.0 13 11.7 Brassicaceae S55: 4.7 6.3 4.6 4.9 Fabaceae 2 5.6 8.4 4.8 8.1 Scrophulariacea a3 a 3 Jeik 6 Fe Chenopodiaceae 5) 6.8 a9 * SH Rosaceae Spe ‘i Zo 2 Boraginaceae Zo * oe 3.0 2 Cactaceae 2D * ss = a Salicaceae Dee a * 5 ** Total in Top Ten % of Flora 58.5% 72.4% 68.8% 53.6% 59.9% classification is it closest to the Cold Temperate Grasslands under the Great Basin Shrub-Grasslands (Brown, 1980). Alcoves, hanging gardens, and seeps are very specialized and variable components of the canyon. These plant communities vary according to direc- tional aspect, amount of sunlight received, depth of alcove, soil type, and amount and duration of water flow. Alcoves at Tsegi are created out of sandstone eroded by vertical movement of water across rock seams. Alcoves range from less than a meter to many meters in size, and may also include prehis- toric housing ruins. The species most common to alcove seeps include Mimulus eastwoodiae and Adiantum capillus-veneris. Other species found in cave seeps include Mentha arvensis, Selaginella mutica, Platanthera zothecina, Pragmites australis, Carex aurea, Carex lanuginosa, Carex specuicola, Oenotheera elata, Aquilegia micrantha, Epipactis gigantea, and Mimulus guttatus. Another type of seep found in the canyon comes out of a vertical canyon side, usually in loose, sandy soil. This type of seep receives a greater amount of sunlight in comparison to the hanging gardens in the alcoves. The plant composition in this type of seep includes grasses such as Avena fatua, Elymus canadensis, Elymus elymoides, Glyceria_ striata, Hordeum jubatum, Poa annua, Polypogon monspe- liensis, Schizachyrium scoparium, Secale cereale, and Sphenooopholis obtusata. Other plants at these sites include Cymopterus beckii, Apocynum can- nabinum, Artemisia ludoviciana, Aster frondosus, Solidago sparsiflora, Lithospermum incisum, and Glycyrrhiza lepidota. There is not a dominant spe- cies listed because of the differences of plant com- ponents present among various seeps. Seeps and alcoves in this area can harbor endemic, rare, and endangered plant species, such as Carex specuicola and Platanthera zothecina. DISCUSSION The floristic composition of representative plant families of the canyon is similar to the floristic composition of the Navajo Nation as last reported by Mayes and Rominger (1994; Table 1). The can- yon has a similar number of Asteraceae and Po- aceae when compared to other similar sites. This is surprising as Tsegi, and the Navajo Nation, are grazed while the national parks used for compari- son are not. Perhaps a more detailed study is need- ed to determine whether or not the grazing affects floristic composition. The differences in the fami- lies showing a small percentage of the total prob- ably can be accounted for by the concentration of specialized habitat types, such as alcoves and ele- vation at each site. Species considered endemic to the Colorado Pla- teau by Welsh (1993) include Calochortus aureus, Platanthera zothecina, Astragalus zionis, Astraga- lus cottamii, Astragalus sesquiflorus, and Cymop- terus beckii. The rare and endangered plants in- clude Platanthera zothecina, a candidate species, Carex specuicola, a listed threatened plant, and Nama retrosum, Penstemon pseudoputus, Astraga- lus cottami, and Cymopterus beckii, whose popu- lations are being watched. Aletes sessiliflorus, iden- tified by L. Constance, and Cymopterus beckii were new reports for the state of Arizona. Introduced species are listed in Table 2. These exotic plants species comprise about 10% of the total plant population of the canyon. Many of the weedy herbs may have been introduced by domes- tic grazing animals, whose feed is supplemented with commercial hay, and also by the disturbing of the land by off-road vehicles. Three introduced plants, Tamarix ramosissima, Elaeagnus angusti- folia, and Ulmus pumila were introduced purposely in the Southwest as shade trees and to aid in erosion control (Welsh 1993). Of these, FE. angustifolia is considered harmful in the canyon by the National Park Service. Attempts are currently being made to keep it out of Betatakin National Monument (Mel- berg 1996). There is evidence that suggests that the inner canyon has changed in the last 150 years. In 1916, L.C. Whitehead (MNA) collected Epipactis gigan- 34 TABLE 2. ExoOTIC PLANTS FOUND IN TSEGI CANYON. al., 1978. *State of Arizona Designated exotic plant Species Name Monocotyledoneae Poaceae Avena fatua Bromus tectorum Dactylis glomerata Erenopyrm triticeum Polypogon monospeliensis Polypogon semiverticillatus Secale cereale Dicotyledoneae Asteraceae Artemisia absinthium Cirsium vulgare Latuca serriola Sonchus asper Tagetes patula Taraxacum officinale Tragopogon dubius' Xanthium strumarium Brassicaceae Capsella bursa-pastoris Corisppora tenella’ Descurainia sophia Sisymbrium altissimum Chenopodiaceae Kochia scoparia Salsola iberica Elaegnaceae Elaeagnus angustifolia Fabaceae Trifolium repends' Medicago lupulina Medicago staiva Melilotus album Geraniaceae Erodium cicutarium Lamiaceae Draccocephalum tymiflorum Marrubium vulgare Plantaginaceae Plantago lanceolata Plantago major' Ranunculaceae Ranunculus testiculatus' Scrophulariaceae Verbascum thapsus Tamaricaceae Tamarix ramosissima Ulmaceae Ulmus pumila! Zy gophyllaceae Tribulus terrestris? MADRONO ‘Exotics found in Navajo Monument not listed in Brotherson et. species. Common Name Wild Oats Cheat Grass Orchard Grass Annual Wheat Grass Rabbitfoot Grass Eurasia/Africa Water Polypogon Eurasia Cultivated Rye Eurasia Absinthe Europe Bull Thistle Eurasia Prickly Lettuce Europe Spiny Sow Thistle Europe Marigold Mexico Dandelion Eurasia Eurasia Cocklebur Eastern U.S. Shepherds Purse Europe Musk mustard Asia Europe Tumble Mustard Europe Summer Cypress Eurasia Tumble Weed Asia Russian Olive Europe White Clover Europe Hop Clover Europe Alfalfa Europe White Sweet Clover Europe Storkbill Europe Horehound Eurasia Eurasia Eurasia Broadleaf Plantain Europe Bur Buttercup Eurasia Wooly Mullein Eurasia Tamarack Eurasia Siberian Elm Asia Puncture Vine Eurasia Origin Eurasia Eurasia Eurasia/Africa Central Asia tea, Pinus edulis, Abies concolor, Equisetum hie- male, Salix exigua, Populus tremuliodes, and Quer- cus gambelii from Tsegi Canyon. All of these spe- cies are now present in the canyon, though the Abies is now only found in one side canyon north- east of the Keet Seel ruin. Of these, Epipactis, Eq- uisetum, Salix, and Populus are usually found in more mesic soils. Clute (1920) described the can- yon as containing desert plants and some meso- phytes. He also stated, that at the time, there was a layer of ‘‘peat two feet thick”’ that contained snail shells. He felt that this was evidence of the previ- ously reported lakes and swamps. The most exten- sive historical collection was done by the Wetherill family in the 1930’s. Most of the plants in that col- lection are also found at the present and are indi- cated as such in the floristic appendix of this paper. Plants included in the Wetherill collection (Wyman 2000] 1951) that were not found in the present study or listed by Brotherson et al. (1978) in their survey of Navajo National Monument include Abronia fra- grans, Cologonia angustifolia, Amaranthus retro- flexus, Juniperus communis, Bromus anomalus, Bromus vulgaris, Pachystema myrsinites, Oeno- thera lavandulifolius, Eriophyllum lanosum, Achil- lea lanulosa, Setaria viridis, Pedicularis centranth- era, Bouteloua eriopoda, Sporobolus pulvinatus, and Nicotiana trigonophylla. These particular plant species could possibly survive in the microhabitats of the present Tsegi drainage. Why they were not found during my survey periods is unknown. AIl- though One Side Canyon was in the past called ‘‘Water lily Canyon’’, I could find no evidence of water lilies collected in the early 1900’s. CONCLUSION Tsegi Canyon is a complex canyon of numerous drainages. It has experienced a large amount of ero- sion in the last hundred years. This has changed the nature of the bottom of the canyon from marshy pools to a faster moving creek. Erosion has lowered the bottom of the creek and that lowered the water table and affected small seeps along the sides of the canyon. Some marshy areas still exist, but only in limited areas in protected side-canyons. There is a good possibility that changes in the fauna and flora have occurred along with local extinctions. Hopefully, more studies will be done here to doc- ument the flora and fauna in this remote area of Arizona. Information needs to be gathered to create an effective means of preserving this riparian area while still allowing usage by local inhabitants. ACKNOWLEDGMENTS I would like to thank Dr. Tina Ayers and Dr. Randy Scott for their guidance in preparing this paper, Dr. H. David Hammond for his help with identifying plants I found so frustrating, Bruce Melberg and the staff at Navajo National Monument for their help and support, Dr. Joseph E. Laf- erriere for his review and helpful comments, and my hus- band, Gordon, my sons, Lyle and Curtis, and fellow bota- nist Abril Perez, who accompanied me on several trips. LITERATURE CITED BONHAM, C. D. 1989. Measurements for Terrestrial Veg- etation. John Wiley and Sons. BROTHERSON, J. D., G. NEBEKEN, M. SKOUGARD, AND J. HOLIDAY: A FLORISTIC STUDY OF TSEGI CANYON, ARIZONA 55 FAIRCHILD. 1987. Plants of Navajo National Monu- ment. Great Basin Naturalist 38(1):19—30. BROTHERSON, J. D., RUSHFORTH, S. R., AND JOHANSEN, J. R. 1983. Effects of Long-term Grazing on Cryptogam Crust Cover in Navajo National Monument, Ariz. Journal of Range Management 36(5):579—581. Brown, D. E., Lowg, C. H., AND PAsg, C. P. 1980. A Digitized Systematic Classification for Ecosystems with an Illustrated Summary of the National Vege- tation of North America. USDA Forest Service Gen- eral Technical Report RM-73. CLAY-POOLE, S. T. 1980. Pollen Flora and Paleoecology of Holocene Alluvial Tsegi and Naha Formations, Northeast Arizona. Northern Arizona University. Flagstaff, AZ. CLUTE, W. N. 1920. Notes on the Navajo Region. The American Botanist. 26(2):38—47. DEAN, J. S. 1969. Chronological Analysis of Tsegi Phase Sites in Northeastern Arizona. University of Arizona Press. Tucson, AZ. Hack, J. T. 1945. The Changing Physical Environment of the Hopi Indians of Arizona. Papers of the Peabody Museum of American Archaeology and Ethnology, Havard University. 35(1). HASTINGS, R. H. AND TURNER R. M. 1965. The Changing Mile. University of Arizona Press. HEIL, K. D., PORTER, J. M., FLEMING, R., AND ROMME, W. H. 1993. Vascular Flora and Vegetation of Capitol Reef National Park, Utah. USDI, National Park Service Technical Report NPS/NAUCARE/NRTR-93/01. KARTESZ, J. T., AND KARTESZ, R. 1890. A Synonymized Checklist of the Vascular Flora of the United States, Canada, and Greenland, Vol. II, The Biota of North America. The University of North Carolina Press. Chapell Hill, NC. KIERSTEAD, J. R. 1981. Flora of Petrified National Park, Arizona. Thesis. Northern Arizona University. MAYES, V. O. AND ROMINGER, J. M. 1994. Navajoland Plant Catalog. National Woodlands Publishing Company. PHILLIPS, B. G., PHILLIPS, A. M. AND BERNZOTT, M. S. 1987. Annotated Checklist of Vascular Plants of Grand Canyon National Park. Grand Canyon Natural History Association, Monograph 7. ROWLANDS, P. G. 1994. Colorado Plateau Vegetation Ass- esment and Classification Manual. USDI, National Park Service, Technical Report NPS/NAUCPRS/ NRTR-94/06. U.S. DEPARTMENT OF COMMERCE, 1996 Climatological Data. WELSH, S. L., ATwoop, N. D., HIGGINS, L. C., AND GOoopD- RICH, S. 1993. A Utah Flora, Great Basin Naturalist Memooirs No. 9. WyMan, L. C. 1951. The Ethnobotany of the Kayenta Navajo. University of New Mexico Press. Albuquer- que NM APPENDIX 1 FLORA OF TSEGI CANYON The nomenclature in this flora follows Kartesz (1980). The letter designation for endemic species is ‘A’, exotic species is ‘B’, federal category C2 is ‘C2’, fegeral category C3 is ‘C3’, federally listed threatened and endangered is “T’, those included in the Wetherill collection are ‘W’. Selaginellaceae Selaginella mutica D. C. Eaton. Perennial; moist areas under cave seeps, Apr.—June. Equisetaceae Equisetum arvense L. Perennial; sandy, moist areas by springs and seeps, June—Aug. 36 MADRONO Equisetum hyemale L. Perennial; sandy, moist area near upper Laguna Creek. June—Aug., W. Adiantaceae Adiantum capillus-veneris L. Perennial; cave seeps and hanging gardens, May—Aug. W. Aspleniaceae Woodsia oregana D.C. Eaton. Perennial; seep run- ning over crack in sandstone cliff, Sept. Cupressaceae Juniperus osteosperma (Torr.) Little. Evergreen Tree; widespread on sandy flats, W. Ephedraceae Ephedra viridis Cov. Evergreen shrub; widespread on sandy flats, W. Pinaceae Abies concolor (Gord. & Glend.) Lindl. Evergreen tree; shady, upper areas of Keet Seel Canyon. Pinus edulis Engelm. Evergreen tree; widespread on sandy flats, W. Pinus ponderosa Doug}. Evergreen tree; sometimes found in side canyons. W. Pseudotsuga menziesii (Mirbel) Franco. Evergreen tree; found in shady side canyons, W. Agavaceae Yucca angustissima Engelm. ex Trel. Perennial; found on sandy flats, flowers in June, W. Yucca baccata var. baccata Torr. Perennial; north slope of Betatakin canyon, W. Commelinaceae Tradescantia occidentalis (Britt.) Smyth. Perennial herb; trail to Betatakin ruins and sandy areas in side canyons, May—June, W. Cyperaceae Carex aurea Nutt. Perennial herb; seeps and springs, June—Aug. Carex lanuginosa Michx. Perennial herb; seep near Keet Seel Ruins, June—Aug. Carex rossii EF Boott. Perennial herb; Betatakin canyon by trail bench, June—Aug. Carex specuicola J. T. Howell. Perennial herb; hanging gardens and seeps, June—Aug., A, T. Eleocharis palustris (L.) R. & S. Perennial herb; Betatakin creek, June—Aug. Scirpus pungens Vahl. Perennial herb; moist creek bottoms, June—Aug. Juncaceae Juncus arcticus Willd. Perennial herb; moist creek bottoms, May—Aug., W. [Vol. 47 Juncus bufonius L. Perennial herb; moist sand, low- er sidecanyons, June—Aug. Juncus saximontanus A. Nels. Perennial herb; by small creek, May—June, W. Liliaceae Allium macropetalum Rybd. Perennial herb; sandy flats, Apr.—May, W. Androstephium breviflorum Wats. Perennial herb; trail to Betatakin, Apr—May. W. Calochortus aureus Wats. Perennial herb; sandy flats, June—July. W. Fritillaria atropurpurea Nutt. Perennial herb; creekside, Fir Canyon, June—July W. Smilacina stellata (L.) Desf. Perennial herb; moist canyon bottoms, June—Aug. W. Orchidaceae Corallorhiza maculata Raf. Perennial herb; lower Betatakin canyon, under trees, June—July, W. Epipacis gigantea Doug]. ex. Hook. Perennial herb; hanging gardens and seeps, June—Aug., W. Platanthera zothecina Higgins & Welsh. Perennial herb; hanging gardens and cave seeps, including Betatakin ruin, July—Aug. A, C2, W. Poaceae Agrostis exarata Trin. Perennial herb; beside small stream in Keet Seel Canyon, June. Arista purpurea Nutt. Perennial herb, dry sandy soil, May—June, W. Avena fatua L. Annual herb; found in moist sand and flat seeps, July—Sept., B. Bouteloua curtipendula (Michx.) Torr. Perennial herb; tree shade in side canyons, Aug.—Sept., W. Bouteloua gracilis (H.B.K.) Lag. ex Steudel. Pe- rennial herb; found on sandy flats with pinion trees, June—Sept., W. Bromus carinatus H. & A. Perennial herb; sandy soil in side canyon, May—June, W. Bromus tectorum L. Annual herb; found on all sandy flats and near streams, April—Sept., B, W. Dactylis glomerata L. Perennial herb; sandy soil beside creek, July—Aug., B. Elymus canadensis L. Perennial herb; moist sand near seep, Aug.—Sept. Elymus cinereus Scribn. & Merr. Perennial herb; sandy soil beside creek, Aug.—Sept. Elymus elymoides (Rat.) Swezey. Perennial herb; wet sand of seep, April—May, W. Elymus smithii (Rybd.) Gould. Perennial herb; sandy areas of lower canyon, April—June. Elymus trachycaulus (Link) Gould ex Shinners. Pe- rennial herb; sandy soil, May—June. Eremopyrum triticeum (Gaetrn.) Nevski. Annual herb; dry sandy soil beside creek, July—Aug., B. Glyceria striata (Lam.) A. S. Hitch. Perennial herb; wet sandy soil by seep, June—July. Hilaria jamesii (Torr.) Benth. Perennial herb; dry sandy soil, June—July. 2000] Hordeum jubatum L. Perennial herb; wet sandy soil by seep, June—July. Hordeum pusilium Nutt. Annual herb; dry sandy soil, June—July. Muhlenbergia andina (Nutt.) A. S. Hitch. Perennial herb; found in sandy soil of dry creek bed, Aug.— Sept., W. Muhlenbergia pungens Thurben in Gray. Perennial herb; dry sandy soil, Aug.—Sept., W. Monroa squarrosa (Nutt.) Torr. Annual herb; found on sandy soil, July—Aug., W. Pragmites australis (Cav.) Trin. ex Steudel. Tall perennial herb; found at Betatakin ruin, June— Sept., W. Poa pratensis L. Perennial herb, moist soil in shade, June—Sept. Polypogon semiverticillatus (Forsskal) Hylander. Perennial herb, moist, sandy soil, June—July, B. Puccinellia nuttalliana (Schultes) A. S. Hitch. Pe- rennial herb, sand by seep, June—July. Schizachyrium scoparium (Michx.) Nash in Small. Perennial herb, sand by seep, July—Aug. Secale cereale L. Annual herb, moist sand by seeps, July—Aug., B. Sphenopholis obtusata (Mitchx.) Scribn. Annual herb, wet sandy soil, July—Aug. Sporobolus cryptandrus (Torr.) gray. Perennial herb, dry sand, Aug.—Sept., W. Stipa comata Trin. & Rupr. Perennial herb, sandy soil, May—June, W. Stipa hymenoides R. & S. Perennial herb, dry sand, May-June, W. Vulpia octoflora Walter. Annual herb, sandy soil creekside, May—June, W. Typhaceae Typha domingensis Pers. Perennial, below Keet Seel ruin and side canyons, June—Sept. Aceraceae Acer negundo L. Tree, near streams and seeps, May—Oct., W. Amaranthaceae Amaranthus blitoides Wats. Annual herb, sandy soil near streams, July—Sept. Anacardiaceae Rhus aromatica var. trilobata (Nutt.) Gray. Shrub, upper side canyon, June—Aug., W. Toxicodendron rydbergii (Small) Greene. Small shrub, upper side canyons, June—Sept., W. Apiaceae Aletes sessiliflorus Theobald & Tseng. Perennial herb, sand by streams, June—July, new report. Cymopterus acaulis (Pursh) Raf. Perennial herb, sand under trees, June—July, W. HOLIDAY: A FLORISTIC STUDY OF TSEGI CANYON, ARIZONA Si) Cymopterus beckii Welsh & Goodrich. Perennial herb near seeps, June—July, new report, A, C2. Cymopterus purpureus Wats. Perennial herb, clay soil near trees, June—July. Apocynaceae Apocynum cannabium L. Perennial herb, moist sand by seeps, June—Aug. Apocynum x medium Greene. Perennial herb, moist sand by stream, June—July. Asclepiadaceae Asclepias asperula (Decne.) Woodson. Perennial herb, sand, May—June, W. Asclepias latifolia (Torr.) Raf. Perennial herb, sun- ny sand, June—July. Asclepias speciosa Torr. Perennial herb, beside stream, July—Aug., W. Asclepias subverticillata (Gray) Vail. Perennial herb, sunny canyon sides, June—Aug. Asteraceae Ambrosia acanthicarpa Hook. Annual herb, sandy soil, Aug.—Sept., W. Antennaria neglecta Greene. Perennial herb, shade, sand, May—June, W. Antennaria parvifolia Nutt. Perennial herb, shade, June, W. Artemisia absinthium L. Perennial herb, dry sand, July—Aug. Artemisia campestris L. Perennial herb, dry sandy soil, shade, July—Aug., W. Artemisia carruthii Wood ex Carruth. Perennial herb, sandy soil, July—Aug. Artemisia dracunculus L. Perennial herb, shade, June—Aug., W. Artemisia frigida Willds. Perennial herb, sandy soil, June—Sept., W. Artemisia ludoviciana Nutt. Perennial herb, moist sand, July—Aug., W. Artemisia tridentata var. tridentata Nutt. Shrub, common in canyon, June—Aug., W. Aster frondosus (Nutt)T. & G. Annual herb, moist sand, Aug.—Sept., W. Aster glaucodes Blake. Perennial herb, creekbed, July—Sept. Brickellia californica Gray. Perennial subshrub, moist sand, July—Aug., W. Brickellia microphylla (Nutt.) Gray. Small shrub, sandy soil, Aug.—Oct., W. Brickellia oblongifolia Nutt. Perennial subshrub, dry clays, May—June. Chaetopappa ericoides (Torr.) Nesom. Perennial herb, dry sand, June—July, W. Chrysothamnus depressus Nutt. Low shrub, dry creek bottom, July—Sept. Chrysothamnus nauseosus (Pallas)Britt. Shrub, common in canyon, July—Sept. Chrysothamnus viscidiflorus (Hook.)Nutt. Shrub, lower canyon, July—Sept., W. 38 MADRONO Cirsium calcareum var. pulchellum (Greene)Welsh. Perennial herb, creek bottoms, July—Sept. Cirsium vulgare (Savi)Ten. Biennial herb, creek bottom, July—Sept. Conyza canadensis (L.)Crong. Annual herb, creek bottom, June—Sept. Erigeron bellidiastrum Nutt. Annual herb, sandy soil, July—Aug. Erigeron compactus Blake. Perennial herb, dry sand, May—June. Erigeron eatonii Gray. Perennial herb, sand, shade, May-June, W. Erigeron flagellaris Gray. Perennial herb, near seep, April—May. Erigeron lonchophyllus Hook. Perennial herb, moist sand, July—Aug. Erigeron pumilis Nutt. Perennial herb, sandy soil, June—July. Erigeron speciosus (Lindl.) D.C. Perennial herb, sandy soil July—Aug. Erigeron utahensis Gray. Perennial herb, sandy soil, sun, May—June. Gnaphalium chilense Sprengel. Annual herb, moist sand, July—Aug. Gutierrezia sarothrae (Prush)Britt. & Rusby. Small shrub, sand, sun, July—Sept., W. Haplopappus ameriodes (Nutt.)Gray. Perennial herb, dry clays, May—June. Haplopappus spinulosus (Prush)D.C. Perennial herb, sand, shade, July—Aug. Helianthus petiolaris Nutt. Annual herbs, sand, sun, July—Aug., W. Heterotheca villosa (Pursh)Shinn. Perennial herb, sandy soil, common, July—Sept., W. Hymenopappus filifolius Hook. Perennial herb, west facing walls, June—July, W. Hymenoxys acaulis (Pursh)Parker. Perennial herb, sandy soil, May—June, W. Lactuca serriola L. Biennial herb, sandy soil, Aug.—Sept., B, W. Lygodesmia grandiflora (Nutt.)T.&G. Perennial herb, clay soil, June—July. Machaeranthera canescens (Pursh)Gray. Biennial herb, sandy soil, July—Aug., W. Machaeranthera grindeliodes (Nutt.)Shinn. Peren- nial herb, dry sand, June—July, W. Petradoria pumila (Nutt.)Greene. Perennial herb, dry sand, June—July. Psilostrophe sparsiflora (Gray)W.Nels. Perennial herb, sandy soil, June—Aug., W. Senecio douglasii DC. Perennial herb, sandy soil below seep, July—Aug. Senecio multilobatus T.&G. Perennial herb, sandy soil, shade, May—June, W. Senecio spartioides T.&G. Perennial herb, sandy soil, Aug.—Sept., W. Solidago canadensis L. Perennial herb, dry creek bottom, July—Sept., W. Sonchus asper (L.)Hill. Annual herb, moist sand, July—Sept., B. [Vol. 47 Stephanomeria exigua Nutt. Annual herb, sandy soil, July—Aug. Stephanomeria tenuifolia (Torr.)Hall. Perennial herb, sandy soil, July—Aug. Tagetes patula L. Annual herb, moist sandy soil, July—Aug., B. Taraxacum officinale Weber ex Wiggers. Perennial herb, moist soil, May—Sept., B. Thelesperma subnudum Gray. Perennial herb, sandy soil, sun, June—July, W. Tragopogon dubius Scop. Biennial herb, dry sand, June—Aug., B. Verbesina enceliodes (Cav.)Benth.&Hook. Annual herb, sandy soil, July—Aug. Wyethia scabra Hook. Perennial herb, creekbed sand, June—July. Xanthium strumarium L. Annual herb, moist sand, June—Aug., B. Berberidaceae Mahonia repens (Lindl.)G.Don. Small evergreen shrub, shade, W. Betulaceae Betula occidentalis Hook. Small trees, along creek sides, May—Aug. Boraginaceae Cryptantha bakeri (Greene)Payson. Biennial herb, sandy soil, May—June. Cryptantha crassisepala (T.&G.)Greene. Annual herb, sand, sun, May—June, W. Cryptantha flava (A.Nels.)Payson. Perennial herb, sandy soil, May—June. Cryptantha fulvocanescens (Wats.)Payson. Peren- nial herb, dry sand, June—July. Cryptantha inaequata Johnston. Perennial herb, moist sand below seep, June—July. Cryptantha cinerea (Torr.)Crong. Perennial herb, sandy soil, sun, May—June. Cryptantha circumscissa (H.&A.)Johnston. Annual herb, sandy soil, May—June. Lappula occidentalis (Wats.)Greene. Annual herb, dry sand, April—May, W. Lithospermum insicum Lehm. Perennial herb, moist sand, June—Aug. Brassicaceae Arabis perennans Wats. Perennial herb, sand, shade, June—Aug., W. Arabis pulchra var. pallens Jones. Perennial herb, sandy soil, May—June. Capsella bursa-pastoris (L.)Medicus. Annual herb, near trails, May—June, B. Chorispora tenella (Pallas)DC. Annual herb, along trails, June—July, B. Descurainia pinnata (Walter)Britt. Annual herb, sandy soil, May—June, W. 2000] Descurainia sophia (L.) Webb ex Prantl. Annual herb, trail side, May—June, W. Dithyrea wislizenni Engelm. in Wisliz. Annual herb, sand, May—June, W. Lepidium montanum var. spathulatum (Rob- ins.)C.L.Hitch. Perennial herb, sandy soil, July— Aug., W. Lepidium virginicum L. Annual herb, moist sand by seep, July—Aug. Lesquerella intermedia (Wats.)Heller. Perennial herb, sandy soil, April-June, W. Pysaria newberryi Gray. Perennial herb, sandy soil, June—July. Sisymbrium altissimum L. Annual herb, sand dunes, July—Sept., B. Stanleya pinnata (Pursh)Britt. Perennial herb, clays, June—Aug. Streptanthella longirostris (Wats.)Rydb. Annual herb, sandy soil, shade, May—June. Strepthanthus cordatus Nutt. ex T:&G. Perennial herb, sand, shade, June—Aug., W. Thelypodium integrifolium (Nutt.)Britt.&Rose. Bi- ennial herb, sandy soil, June—Aug. Thlaspi montanum L. Annual herb, moist sand be- low seep, April—May. Cactaceae Coryphantha vivipara (Nutt.)Britt.&Rose. Perenni- al, sandy soil, shade, June. Echinocereus triglochidiatus Englem. Perennial, sandy soil, shade, June. Opuntia erinacea Englem. Perennial, sandy soil, sun, June. Opuntia fragilis (Nutt.) Haw. Perennial, sandy soil, shade, June, W. Opuntia phaeacantha var. discata (Griffiths)Benson & Walkington. Perennial, below Betatakin ruin, June. Opuntia polycantha Haw. Perennial, sandy soil, sun, June. Opuntia whipplei Engelm. Perennial, sandy soil, sun, June. Sclerocactus whipplei (Engelm.)Britt.&Rose. Pe- rennial, sandy soil, shade, June. Cannabaceae Humulus americanus Nutt. Perennial herb, dry creek bottom, June—July, W. Capparaceae Cleome serrulata Pursh. Annual herb, sandy soil, sun, May-July, W. Caprifoliaceae Symphoricarpos oreophilius Gray. Shrub, sandy soil, shade, June—Aug., W. Caryophyllaceae Arenaria fendleri Gray. Perennial herb, sandy soil, sun, May—June, W. HOLIDAY: A FLORISTIC STUDY OF TSEGI CANYON, ARIZONA 59 Silene menziesii Hook. Perennial herb, sandy soil, shade, May—June, W. Chenopodiaceae Atriplex canescens (Pursh)Nutt. Shrub, sandy soil, sun, July—Aug. Atriplex confertifolia (Torr.&Frem)Wats. Shrub, sandy soil, July—Aug. Atriplex powellii Wats. Annual herb, sandy soil be- low seep, July—Aug. Chenopodium album L. Annual herb, sandy soil, shade, June—Aug., B. Chenopodium leptophyllum (Mogq.)Wats. Annual herb, sandy soil, July—Aug. Chenopodium rubrum L. Annual herb, sandy soil, July—Aug. Corispermum villosum Rybd. Annual herb, sandy soil, July—Sept. Kochia scoparia (L.)Schrader. Annual herb, sandy soil, July—Aug., B. Salsola iberica Sennen & Pau. Annual herbs, sandy soil, sun, July—Sept., B, W. Sarcobatus vermiculatus (Hook.)Torr in Emory. Shrub, sandy soil, July—Sept. Suaeda torreyana Wats. Perennial subshrub, sandy soil, July—Aug. Cornaceae Cornus sericea L. Shrub, streamside, shade, May— Sept., W. Elaegnaceae Elaeagnus angustifolia L. Small tree, near creeks, May-Sept., B. Shepherdia rotundifolia Parry. Evergreen shrub, cliffsides, W. Ericaceae Arctostaphylos patula Greene. Shrub, sand, near creek, May—Sept., W. Euphorbiaceae Euphorbia lurida Engelm. Perennial herb, sand, sun, June—Aug. Euphorbia micromera Boiss. Annual herb, sand, sun, July—Sept. Fabaceae Astragalus amphioxys Gray. Perennial herb, dry sand, sun, June—July, W. Astragalus ceramicus Sheldon. Perennial herb, sandy soil, June—July. Astragalus cottamii Welsh. Perennial herb, sandy soil, shade, June—July, A, C3. Astragalus flavus Nutt. ex T.&G. Perennial herb, sandy soil, May—June. Astragalus lentiginosus Doug]. ex Hook. Perennial herb, sandy soil, sun, May—June. 40 MADRONO Astragalus mollissimus Torr. Perennial herb, sandy soil, June—July, W. Astragalus sesquiflorus Wats. Perennial herb, sand on canyon sides, June—Sept., A, W. Astragalus zionis Jones. Perennial herb, along trail, May-June, A. Glycyrrhiza lepidota Pursh. Perennial herb, moist sand, shade, July—Aug. Lathyrus brachycalyx Rybd. Perennial herb, sand, shade, July—Aug., W. Lupinus argenteus Pursh. Perennial herb, sand, sun, June—Aug. Medicago lupulina L. Annual herb, sand near creek, June—July, B. Medicago sativa L. Annual herb, sandy soil, sun, July—Sept., B. Melilotus alba Medic. Annual herb, sand, shade, June—July, B. Psoralidium lanceolatum (Pursh)Rydb. Perennial herb, sandy soil, June—July. Trifolium repens L. Perennial herb, sand near creek, June—July, B. Fagaceae Quercus gambelii Nutt. Small trees, sandy soil, canyon sides, May—Sept., W. Fumariaceae Corydalis aurea Willd. Annual herb, sandy soil, shade, June—July, W. Gentianaceae Swertia radiata (Kellogg) Kuntze. Perennial herb, sandy soil; June—Sept. Geraniaceae Erodium cicutarium (L.)L Her. Annual herb, sand, sun, May-—Sept., W. Geranium caespitosum James. Perennial herb, sandy soil, shade, June—Aug., W. Grossulariaceae Ribes cereum Dougl. Shrub, sand, shade, May— Sept., W. Ribes inerme Rydb. Shrub, side canyons, shade, June—Sept. Ribes leptanthus Gray. Shrub, side canyons, shade, June—July, W. Hydrangeaceae Fendlera rupicola Gray. Shrub, sandy soil, sun, May-July, W. Hydrophyllaceae Nama retrosum J.T. Howell. Annual herb, sand dunes, sun, June—July, A, C3, W. Phacelia ivesiana Torr. in Ives. Annual herb, sandy soil, sun, May—June. [Vol. 47 Lamiaceae Dracocephalum thymiflorum L. Annual herb, sandy soil, near creek, June—July, B. Hedeoma drummondii Benth. Annual or perennial herb, sand, June—July. Marrubium vulgare L. Perennial herb, sandy soil, Aug.—Sept., B. Mentha arvensis L. Perennial herb, moist sand, shade, Aug.—Sept., W. Poliomintha incana (Torr.)Gray. Shrub, sandy soil, May-June, W. Linaceae Linum aristatum Engelm. Annual herb, sandy soil, Aug.—Sept. Linum perenne L. Perennial herb, sandy soil, shade, July—Aug. Loasaceae Mentzelia albicaulis Dougl. ex Hook. Annual herb, sandy soil, June—July. Malvaceae Sphaeralcea parvifolia A. Nels. Perennial herb, sand, sun, May-July. Nytaginaceae Mirabilis linearis (Pursh)Heimerl. Perennial herb, sandy soil, July—Aug., W. Mirabilis multiflora (Torr.)Gray in Torr. Perennial herb, sand, shade, July—Aug. Mirabilis oxybaphoides (Gray)Gray in Torr. Peren- nial herb, sandy soil, Aug.—Sept. Tripterocalyx carneus (Greene)Galloway. Annual herb, sand, sun, July—Aug., W. Onagraceae Epilobium ciliatum Raf. Perennial herb, moist sand, shade, June—Aug., W. Gayophytum racemosum T.&G. Annual herb, dry sand, sun, May—June. Oenothera caespitosa Nutt. Perennial herb, sand, beside trail, June—July. Oenothera elata H.B.K. Biennial herb, shade, July—Sept. Oenothera pallida Lindl. Annual herb, sand, sun, June—July, W. sandy, Orobanchaceae Orobanche multiflora Nutt. Perennial herb, near A. tridentata Nutt., June—July, W. Plantaginaceae Plantago lanceolata L. Perennial herb, sandy soil, beside trails, May—June, B. Plantago major L. Perennial herb, moist sand, June—Aug., B. 2000] Plantago patagonica Jacq. Annual herb, dry sand, near trail, May-June, W. Polemoniaceae Gilia aggregata (Pursh)Sprengel. Perennial herb, sandy soil, sun, June—Aug., W. Gilia leptomeria Gray. Annual herb, sandy soil, sun, June—July. Gilia longiflora (Torr.)D.Don. Annual herb, sandy soil, shade, June—July, W. Leptodactylon pugens (Torr.)Nutt. Subshrub, sand, sun, May-July, W. Polygonaceae Erigonum alatum Torr. in Sitg. Perennial herb, sandy soil, sun, May—July, W. Erigonum cernuum Nutt. Annual herb, sandy soil, Aug.—Sept. Erigonum microthecum Nutt. Small shrub, sandy soil, Aug.—Sept., W. Polygonum aviculare L. Annual herb, sandy soil, Aug.—Sept. Polygonum douglassi Greene. Annual herb, sandy soil, June—July. Portulaceae Portulaca oleraceae L. Annual herb, moist sand, June—July. Portulaca retusa Engelm. Annual herb, sand be- hind trail, June—July. Talium parviflorum Nutt. Perennial herb, sandy de- pressions, May—June. Primulaceae Androsace septentrionalis L. Annual herb, sandy soil, shade, June—July, W. Ranunculaceae Aquilegia micrantha Eastw. Perennial herb, hang- ing gardens, June—July, W. Clematis lingusticifolia Nutt. Woody vine, canyon sides, shade, June—July, W. Delphinium andersonii Gray. Perennial herb, sand along trail, June—July, W. Ranunculus cymbalaria Pursh. Perennial herb in marshy areas, May—Sept., W. Ranunculus testiculatus Crantz. Annual herb, sandy soil, sun, May—June, B. Thalictrum fendleri Engelm. Perennial herbs, sand, shade, June—Aug., W. Rhamnaceae Rhamnus betulifolia Greene. Shrub, above pool, shade, June. Rosaceae Amelanchier alnifolia (Nutt.)Nutt. Shrub, sandy soil, shade, May—June. HOLIDAY: A FLORISTIC STUDY OF TSEGI CANYON, ARIZONA 41 Amelanchier utahensis Koehne. Shrub, sandy soil, sun, May—June, W. Cercocarpus intricatus Wats. Shrub, sand, sun, May-June, W. Cercocarpus montanus Raf. Shrub, sandy soil, par- tial shade, June—July. Holodiscus dumosus (Nutt.)Heller. Shrub, sandy soil, June—July, W. Prunus angustifolia Marsh. Small trees, Keet Seel ruin, May—July. Prunus virginiana L. Small tree, beside streams, June—July, W. Pursia mexicana (D.Don)Welsh. Shrub, sandy soil, sun, May—June, W. Pursia tridentata (Pursh)DC. Shrub, sandy soil, sun, May—June, W. Rubiaceae Galium aparine L. Annual herb, moist sand, shade, June—Aug. Salicaceae Populus angustifolia James. Tree, sandy soil, by streams, May. Populus fremontii Wats. Tree, sand, along streams, May. Salix exigua Nutt. Shrub, sand, sun, along streams, May-June. Salix laevigata Bebb. Small tree, sand, along stream, May—June, W. Salix lasiolepis Benth. Shrub, sand, along stream, May-June. Salix monticola Bebb. ex Coult. Shrub, moist sand, May-June. Santalaceae Comandra umbellata (L.)Nuatt. Perennial herb, sandy soil, sun, June—July, W. Saxifragaceae Heuchera parvifclia Nutt. in T.&G. Perennial herb, sandy soil, shade, June—July, W. Scrophulariaceae Castilleja chromosa A. Nels. Perennial herb, sandy soil, June—July. Castilleja linariifolia Benth. Perennial herb, sandy soil, shade, May—June, W. Cordylanthus wrightii Gray. Annual herb, sandy soil, shade, July—Aug., W. Mimulus eastwoodiae Rydb. Perennial herb, hang- ing gardens, July—Aug. Mimulus guttatus DC. Perennial herb, moist sand, shade, June—July. Mimulus rubellus Gray. Annual herb, sandy soil, shade, July—Aug. Penstemon barbatus (Cav.)Roth. Perennial herb, sand, July—Aug., W. 42 MADRONO Penstemon comarrhenus Gray. Perennial herb, sand, sun, June—July, W. Penstemon eatonii var. undosus Jones. Perennial herb, sand, shade, May—June, W. Penstemon pseudoputus (Crosswhite) N.Holmgren. Perennial, sand, June—July, A, C3. Penstemon rostriflorus Kellogg. Perennial, sand, shade, June—July. Verbascum thapsus L. Biennial, sand, sun, July— Aug., B. Veronica pergrina L. Annual herb, moist sand, June—July. Solanaceae Chamaesaracha coronopus (Dunal)Gray. Perennial herb, sandy soil, Aug.—Sept., W. Datura wrightii Regel. Annual herb, sandy soil, partial shade, July—Aug. Lycium pallidum Miers. Shrub, sand, sun, May— June, W. Pysalis hederifolia Gray. Perennial herb, sand, sun, July—Aug. Solanum jamesii Torr. Perennial herb, sand, by trail, July—Aug., W. [Vol. 47 Tamaricaceae Tamarix ramosissima Ledeb. Shrub, sand, along washes, May-June, B. Ulmaceae Ulmus pumila L. Tree, along sides of creek, May— June, B. Valerianaceae Valeriana acutiloba Rybd. Annual herb, sand, shade, July—Aug., W. Verbenaceae Verbena bracteata Lag.&Rodr. Perennial herb, sand, sun, Aug.—Sept. Viscaceae Phorodendron juniperinum Gray. Parasitic peren- nial, found on juniper trees. Zy gophyllaceae Tribulus terrestris L. Annual herb, sand, sun, July— Aug., B. Mapbrono, Vol. 47, No. 1, pp. 43-52, 2000 MIXED CONIFER AND RED FIR FOREST STRUCTURE AND USES IN 1899 FROM THE CENTRAL AND NORTHERN SIERRA NEVADA, CALIFORNIA Scott L. STEPHENS!” 'Natural Resources Management Department, California Polytechnic State University, San Luis Obispo, CA 93407 ?Current Address: Division of Forest Sciences, Department of Environmental Science, Policy, and Management; University of California, Berkeley, CA 94720-3114 stephens @ nature.berkeley.edu ABSTRACT Historical data collected from five ‘‘average’’ mixed conifer stands, four large mixed conifer stands, and four red fir stands from the central and northern Sierra Nevada by George Sudworth in 1899 were analyzed to determine historic forest structure including diameter distributions, basal areas, and snag and live tree densities. The effects of early logging operations on stand composition and structure is quantified by comparing characteristics of the trees that were harvested versus those unharvested in four mixed conifer stands. Average diameter at breast height (DBH) was 86 cm (34 inches) in the “‘average”’ mixed conifer stands, 110 cm (43 inches) in the large mixed conifer stands (this was equal to the average DBH of 8 mixed conifer stands sampled by Sudworth in the southern Sierra Nevada), and 77 cm (30 inches) in the red fir stands for trees greater than 30.5 cm DBH. Shade intolerant tree species dominated the ‘“‘average’’ mixed conifer stands, shade intolerant, intermediate, and shade tolerant species were abundant in the large mixed conifer stands, and Abies magnifica Andr. Murray dominated the red fir stands. Mean tree density for the “‘average’’ mixed conifer, large mixed conifer, and red fir stands was 229 trees/ha, 235 trees/ha, and 433 trees/ha, respectively. Average tree density was higher in Sudworths southern Sierra Nevada mixed conifer stands when compared to the central and northern Sierra Nevada. Snag density averaged 5/ha in the large mixed conifer stands and 17.5/ha in the red fir stands. Early logging operations removed the majority of the Pinus spp. and Pseudotsuga menziesii (Mirbel) France leaving large amounts of Calocedrus decurrens (Torrey) Florin and Abies concolor (Gordon & Glend.) Lindley. Information from this study can assist in the characterization of historic stand structure in these forest types. The absence of fire in the 20th century and past harvesting operations have modified the structure and ecosystem processes in the coniferous forests of the Sierra Nevada. An increase in the density of small shade tolerant trees has been produced in many forest types (Leopold et al. 1963; Hartesveldt and Harvey 1967; Vankat and Major 1978; Parsons and DeBendeetti 1979; Bonnicksen and Stone 1982) and this increase has resulted in a decrease in forest sustainability (Weatherspoon and Skinner 1996; van Wagtendonk 1996; Stephens 1998). Changes in climate over the last centaury may have also contributed to the changes in forest structure (Millar and Woolfenden 1999). Historical and prehistoric information on the Structure (density, size distribution, and species composition) of mixed conifer forests are relatively rare and they have been reviewed elsewhere (Ste- phens and Elliott-Fisk 1998). One of the methods that can be used to determine prehistoric forest Structure is the analysis of data from early forest inventories. These data provide quantitative infor- mation on historic forest structure, however, the re- sults from the analyses can be biased because the methods used to select the stands were frequently not recorded (Stephens and Elliott-Fisk 1998). Analysis of historical data have been done for the Stanislaus and Lake Tahoe Forest Reserves (Sudworth 1900), portions of the northern Sierra Nevada and the Transverse Ranges of southern Cal- ifornia (McKelvey and Johnston 1992), and por- tions of the southern Sierra Nevada (Stephens and Elliott-Fisk 1998). All of these studies discuss early logging operations but no work has been done that quantifies the effects of early logging at the stand level, quantifies the amount of hardwoods present historically in mixed conifer forests, determines historic snag densities and sizes, or differentiates between average and mature mixed conifer stands. Early logging operations affected the composi- tion and structure of Sierra Nevada forests, es- pecially between 1860 and 1950 (Laudenslayer and Darr 1990). In 1899, approximately 45 percent of the trees harvested in California were either Pinus ponderosa Laws (ponderosa pine) or Pinus lam- bertiana Douglas (sugar pine). Most early logging operations in the Sierra Nevada harvested all trees that were considered to be merchantable at the time of the harvest (Laudenslayer and Darr 1990). The viability of the California spotted owl (Strix occidentalis occidentalis) is receiving major atten- 44 MADRONO tion in California. The owl prefers to nest in mixed conifer forests with 80 percent of the nesting sites occurring in this forest type followed by 10 percent in Abies magnifica Andr. Murray (red fir) and 7 percent in Pinus ponderosa hardwood forests (Ver- ner et al. 1992). The remaining 3 percent of nests occur in eastside pine forests and foothill riparian- hardwood habitats in the western Sierra Nevada foothills (Verner et al. 1992). The habitat requirements of the California spot- ted owl have been investigated and it nests in old- growth forests with high canopy cover (Gutierrez et al. 1992). A relatively high number of snags and down logs are also correlated to the current nesting sites of the California spotted owl (Gutierrez et al. 1992) but no prehistorical data exist on the abun- dance of snags or fuel loads in this forest type mak- ing it difficult to describe the composition of the prehistorical habitat. The objective of this paper is to analyze mixed conifer and red fir forest inventory data acquired by George Sudworth in 1899 from the central and northern Sierra Nevada to further our understanding of forest conditions and their management in the late 19th century. Analysis includes snag and live tree densities, basal areas, diameter distributions, and quantification of the effects of early logging operations on stand composition and structure. STUDY SITE AND METHODS The historic data analyzed in this paper were ob- tained from the area of the central and northern Sierra Nevada that now includes the southern por- tion of the Tahoe National Forest, the El] Dorado National Forest, and northern portion of the Stan- islaus National Forest. Mixed conifer and red fir forests were surveyed in 1899 by George B. Sudworth while employed by the United States Geological Survey. The purpose of this survey was to inventory the forest reserves of the Sierra Nevada. The original unpublished field notebooks (Sudworth 1899) were the source of the inventory data analyzed in this paper. Sierra Nevada mixed conifer forests sampled by Sudworth were composed of white fir Abies con- color (Gordon & Glend.) Lindley (white fir), Abies magnifica, Pinus ponderosa, Pinis lambertiana, Pinis jeffreyi Grev. and Balf (Jeffrey pine), Calo- cedrus decurrens (Torrey) Florin (incense cedar), Pseudotsuga menziesii (Mirbel) Franco (Douglas- fir), and Quercus kelloggii Newb. (California black oak). The red fir forests were composed of Abies magnifica, Pinus jeffreyi, Pinus monticola Douglas (western white pine), Pinus contorta spp. murray- ana (Grev. & Balf.) Critchf. (lodgepole pine), and Tsuga mertensiana (Bong.) Carriére (mountain hemlock). Red fir forests are widely distributed and they can be found on both the west and east sides of the Sierra Nevada (Rundel et al. 1977). All stand data recorded by Sudworth in 1899 are analyzed in this paper with the exception of one | stand located in a pure Pinus jeffreyi forest because | of no replication in this forest type. Exact stand | locations are not given in the field notebooks but | references to rivers, mountains, and landmarks are | included (Sudworth 1899). Five “‘average”’ mixed conifer stands, four large mixed conifer stands, and four red fir stands were | recorded in the 1899 field notebooks (Sudworth 1899). Mixed conifer stand data were stratified into | two classes (average and large) whereas this was © not done in the southern Sierra Nevada analysis | (Stephens and Elliott-Fisk 1998) because no stands | were identified by Sudworth in his notebooks as _ having “‘average’’ characteristics. Sudworth recorded the species, diameter at breast | height (DBH), and number of 4.9 m (16 ft) logs for each tree greater than 30.5 cm (12 inches) DBH (one 28 cm DBH tree was recorded in a red fir — stand). Each stand was sampled with one 0.1 ha | (0.25 acres) plot. He recorded notes on regeneration © (estimate of density by species, not a complete | seedling inventory), forest floor depth, and other © information such as the revenue generated by early | [Vol. 47 | logging operations. He also frequently commented on the effects of early grazing and burning on the Sierra Nevada and his comments are summarized © below. The following values were calculated by aver- aging all stand data for each of the 3 forest types (“‘average’’ mixed conifer, large mixed conifer, red fir): number of snags per hectare, snag basal area, diameter and species of trees removed by early log- — ging operations, basal area per hectare by species, © number of trees per hectare by species (density), quadratic mean diameter by species, percent total basal area by species, and percent total stocking by species. Stand data are summarized and discussed, but a statistical analysis was not performed. Selection of an appropriate analysis method requires informa- tion on sampling procedures which are unknown for this early forest inventory (Stephens and Elliott- Fisk 1998). RESULTS “Average’’ mixed conifer stands. 'The five mixed conifer stands denoted as “‘average’’ by George Sudworth were dominated by moderate sized trees of several species. The average quadratic mean di- ameter for all trees over 30.5 cm DBH was 86 cm (34 inches). Average tree density was 229 trees/ha (92 trees/acre) (range 150-300 trees/ha). Average basal area was 130 m*/ha (558 ft?/acre) (range 94— 186 m2/ha). Table 1 summarizes all stand calcula- tions for the ‘“‘average’’ mixed conifer stands. The largest trees in the ‘“‘average”’ mixed conifer stands were Pinus lambertiana with an average DBH of 108 cm (42 inches). The largest Pinus lam- bertiana recorded in the inventory had a DBH of 2000) TABLE 1. STEPHENS: MIXED CONIFER AND RED FIR FOREST STRUCTURE AND USES IN 1899 45 AVERAGE CALCULATIONS OF GEORGE SUDWORTH’S 5 “‘“AVERAGE”” MIXED CONIFER STANDS IN THE CENTRAL AND NORTHERN SIERRA NEVADA IN 1899 (STANDARD ERROR). * Average value for all stands. Basal area DBH Percent (m2/ha) Trees/ha (cm) of total Percent of Tree [130]* [229]* [86]* basal area trees/ha Abies concolor 5.3 6.0 105.7 4 3 6.2) (6.0) (0) Calocedrus decurrens 26.0 54.0 80.6 20 24 (6.6) (11.2) (10.5) Pinus lambertiana 8.8 12 107.8 7 3 (4.2) (7.4) (22.8) Pinus ponderosa 56.6 106.0 83.9 43 46 (23.1) (37.0) (12:1) Pseudotsuga menziesii 30.9 38.0 101.2 24 16 (20.2) (24.6) (1.8) Quercus kelloggii 2.4 2 58.0 2 6 (1.4) (3.7) (7.2) 152 cm (60 inches). Pinus ponderosa was the most common species comprising 46 percent of total stocking and 44 percent of total basal area (Table 1). Abies concolor was rare in the stands accounting for only 3 percent of total stocking and 4 percent of total basal area. Calocedrus decurrens and Pseu- dotsuga menziesii were the next most common spe- cies, after Pinus ponderosa, respectively. The av- erage DBH of the Quercus kelloggii was the small- est of the species found in the mixed conifer stands, the conifer with the smallest average DBH was Calocedrus decurrens (Table 1). Pinus lambertiana, Pseudotsuga menziesii, and Abies concolor all had similar average DBH’s of approximately 105 cm whereas Pinus ponderosa and Calocedrus decurrens had average DBH’s of approximately 82 cm. Quercus kelloggii accounted for an average of 6 percent of stand stocking. No snags were recorded in the average mixed conifer stands (Table 2). Four of the “‘average’’ mixed conifer stands in- ventoried by Sudworth were in the process of being harvested in 1899. Sudworth’s notebooks recorded the diameter and species of all trees harvested and also recorded the same information on all trees that remained after the harvesting operation. All of the Pseudotsuga menziesii trees in these TABLE 2. CHARACTERISTICS OF SNAGS FOUND IN MIXED CONIFER AND RED FIR STANDS IN THE CENTRAL AND NORTH- ERN SIERRA NEVADA IN 1899. Density Average Average range Average basal density (snags/ DBH area Stand type (snags/ha) ha) (cm) (m/?/ha) Average mixed conifer 0 0 0 0 Large mixed conifer 5 0-10 108.7 4.6 Red fir 17.5 0-60 a 4.5 four stands were harvested along with 88 percent of the Pinus lambertiana trees (Table 3). The ma- jority of the wood harvested from these stands was from Pinus ponderosa trees with an average of 64 m*/ha (275 ft?/acre) removed and this was 2.4 times greater than Pseudotsuga menziesii which was the next most common species harvested. The amount of Calocedrus decurrens and Abies concolor trees harvested was low, averaging 13 percent and 33 percent, respectively (Table 3). The following comments were written by George Sudworth in the original field notebooks and in- clude information about regeneration and impacts from early European settlers (Sudworth 1899). September 27, 1899. Near Beech Sawmill (above Placerville) on Big Iowa Canyon. No reproduction (manzanita brush) but abundant a few yards distant. Grazed, no humus, all trees fire marked. September 28, 1899. South of Blair Sawmill (near Sly Park) on summit of ridge. All touched with fire, humus 1—2 inches in spots. Ample repro- duction of all species in patches. September 30, 1899. Sample on big hill south west of Grizzly Flat 0.5 mile. Humus all burned off. October 5, 1899. 2 miles east of Whitmore’s Mill TABLE 3. AVERAGE AMOUNT HARVESTED IN 4 “‘SAVERAGE”’ MIXED CONIFER STANDS LOCATED IN THE CENTRAL AND NORTHERN SIERRA NEVADA IN 1899. Basal _ Basal Trees/. area area DBH of ha cut cut cut trees (per- (per- (m’%/ _— cut Tree cent) cent) ha) (cm) Abies concolor 33:3 48.2 3.9 127 Calocedrus decurrens 125 12.5 1.6 88.9 Pinus lambertiana $7.5 92.8 17.1 104.3 Pinus ponderosa 57.9 64.2 64.3 103.3 Pseudotsuga menziesii 100.0 100.0 27.2 101.7 Quercus kelloggii Za0", “381 2.9 86.4 46 MADRONO Fic. 1. El Dorado county, 1899. Opposite Snyder and Sherman’s Ranch. Yellow pine (mixed conifer) forest on south slope of Silver Fork. Ponderosa pine 91—193 cm (36—76 inches) in diameter, 46—50 m high (150—165 feet), clear 8— 11 m (25-35 feet), ten in 0.1 ha (0.25 acre), 3—5 white fir (Abies concolor) same size. Cattle grazed. (Mill Creek, near Volcano), representing no cut stumpage, rolling flat 1000 feet above creek bot- tom. No humus, frequent burning destroyed all. Abundant reproduction of pines and cedar 5-8 years old, mostly under 4. October 5, 1899. Near Whitmore’s Mill but in shallow ravine. Taken as a whole mill operator es- timates output 10—20 thousand per acre. Abundant reproduction of all species. Taxus brevifolia Nutt., dogwood, and Acer macrophyllum Pursh abundant. Humus in part 3—6 inches. Large mixed conifer stands. The four large mixed conifer stands were dominated by large trees of several species and the average quadratic mean diameter at breast height was 110 cm (43 inches) for all trees above 30.5 cm DBH. Average tree den- sity was 235 trees/ha (94 trees/acre) (range 160— 300 trees/ha). Average basal area was large 215 m?/ ha (923 ft*/acre) (range 188—232 m*/ha). The stands inventoried by Sudworth were relatively open and dominated by large trees (Fig. 1). Table 4 summa- rizes all stand calculations for the large mixed co- nifer stands. Abies concolor was the most common species comprising 46 percent of total stocking, but only accounting for 34 percent of total basal area be- cause of their relatively small diameters. The larg- est tree inventoried in these stands was a Pseudo- tsuga menziesii and it had a DBH of 188 cm (74 inches). Pseudotsuga menziesii made up only 16 percent of the trees/ha but contributed 24 percent of the basal area of the stands because of their large size. Abies concolor trees were much more com- mon in the large mixed conifer stands when com- pared to the “‘average’’ mixed conifer stands (46 percent stocking versus 3 percent stocking, respec- tively). Abies concolor, Abies magnifica, and Calocedrus decurrens were the smallest trees with average qua- dratic mean diameters of approximately 93 cm. Pi- nus ponderosa and Pinus lambertiana were larger with average diameters of approximately 112 cm, 2000] STEPHENS: MIXED CONIFER AND RED FIR FOREST STRUCTURE AND USES IN 1899 47 TABLE 4. AVERAGE CALCULATIONS OF GEORGE SUDWORTH’S 4 LARGE MIXED CONIFER STANDS IN THE CENTRAL AND NORTHERN SIERRA NEVADA IN 1899 (STANDARD ERROR). * Average value for all stands. Basal area Percent of (m?/ha) Trees/ha DBH (cm) total basal Percent of Tree [215]|* [235]* [110]* area trees/ha Abies concolor IZ 107.5 97:3 34 46 (20.0) (38.2) (6.4) Calocedrus decurrens 26.7 32.5 92.0 We 14 (9:5) (6.3) (17.4) Pinus lambertiana 33.4 Pie i) 109.1 16 12 (18.3) (4.8) (20.5) Pinus ponderosa 10.5 10.0 LiS:3 5 4 (10.5) (10.0) (O) Pinus jeffreyi 13.8 10.0 120.4 6 4 (11.4) (LAY (21.4) Pseudotsuga menziesii 51.8 5755 £2327 24 16 (45.4) Gi) (11.8) Abies magnifica 6.2 10.0 88.9 3 4 (6.2) (10.0) (O) and the largest trees were Pinus jeffreyi and Pseu- dotsuga menziesii with average diameters of ap- proximately 122 cm. Quercus kelloggii was not re- corded in any of the large mixed conifer stands. Snag density averaged 5/ha with a range O-10/ ha (Table 2). Average snag quadratic mean diam- eter was 109 cm and snag average snag basal area was 4.6 m’/ha. The following comments were written by George Sudworth in the original field notebooks and in- clude information about regeneration and impacts from early European settlers (Sudworth 1899). September 3, 1899. 12—15 miles west of Bloods, north slope of Mokelumne River. 30 concolor 2—8 inches diameter, 100 under 6 inches. 5 sugar pine under 2 feet high. Thickets of Acer oblusifobium (glabrum?). September 8, 1899. South slope of Bear River, one half way up slope. Seedlings of all in spots near blue ceonothus when protected from tramping of cattle. No sheep here, but no humus. Abundant blue ceonothus chaparral. September 9, 1899. South lower slope of Silver Fork (American River) in rich bottom bench (at point where a little stream enters Silver Fork). Dense fir and cedar on outskirts, no seedlings with- in. Humus 2-3 inches, cattle grazing and sheep. September 21, 1899. 1.5 miles south of Merzns, across (west) of Dark and Multon Canyons (where Georgetown road crosses). South slope of South Fork of the Consumnes River (?) (Sudworth in- cluded the? mark and was probably referring to the American River). Abundant reproduction of all spe- cies 1-12 years old, all fire marked 15 years back. Humus 1.5 inches deep, soil sandy loam with rock. Red fir stands. Sudworth sampled four red fir stands during this inventory and all stands were dominated by Abies magnifica. The average qua- dratic mean diameter at breast height was 77 cm (30 inches) for all trees inventoried. Average tree density was 433 trees/ha (173 trees/acre) (range 180—610 trees/ha) for trees greater that 28 cm DBH. Average basal area was 202 m*/ha (867 ft?/ acre) (range 98—286 m*/ha). Table 5 summarizes all stand calculations for the red fir stands. TABLE 5. AVERAGE CALCULATIONS OF GEORGE SUDWORTH’S 4 RED FIR STANDS IN THE CENTRAL AND NORTHERN SIERRA NEVADA IN 1899 (STANDARD ERROR). * Average value for all stands. Basal area DBH Percent of (m?/ha) Trees/ha (cm) total basal Percent of Tree [202]* [433]* aa area trees/ha Pinus jeffreyi a2 2 25.0 128.1 16 6 (32.2) (25.0) (15.1) Abies magnifica 136.2 2120 80.1 67 63 (55-3) (93.5) (10.9) Pinus monticola 11.3 30.0 66.4 6 d (8.0) (19.2) (19.4) Pinus contorta 12.6 70.0 47.8 6 16 (12.6) (70.0) (12.0) Tsuga mertensiana 9.3 35.0 58.0 5 8 (9.3) (35.0) C45) 48 MADRONO [Vol. 47 Fic. 2. Amador county, 1899. Near sawmill 5—6 km (3—4 miles) below dam on Bear River. Forest fire in fir and pine, killed all seedlings, just started. The largest trees in the red fir stands were Abies magnifica and Pinus jeffreyi with DBH’s of 160 cm (63 inches). Abies magnifica was the most common tree in the stands accounting for 63 percent of all trees inventoried and 68 percent of average stand basal area (Table 5). The next most common tree found was Pinus contorta which accounted for 16 percent of all trees but only contributed to 6 percent of basal area because of the smallest DBH of any species in this forest type. Snag density averaged 17.5 per ha with a range O-—60 per ha (Table 2). Average snag quadratic mean diameter was 57 cm (22 inches) and average snag basal area was 4.5 m/?/ha. The following comments were written by George Sudworth in the original field notebooks and in- clude information about regeneration and impacts from early European settlers (Sudworth 1899). September 2, 1899. On foothill (above) Bear Meadow, north fork of Stanislaus River. No graz- ing, 40 young trees under 10 inches diameter. Hu- mus 4—6 inches deep, no herbaceous growth. 75— 100 seedlings 2—10 inches. September 7, 1899. On south slope 4—5 miles down on Silver Fork (near Silver Lake and Kirk- wood). Sheep grazing, no reproduction. Scattered bunches of blue ceonothus. Earth bare, rock and gravel. September 7, 1899. North side of Thimble Peak (west of Kirkwood Meadow). On volcanic and granite. No humus, grazed by sheep. Dense shade in part, no reproduction. 2 Abies magnifica down. September 13, 1899. On Rocky flat between Ly- ons and Blakley (south fork Silver Creek, west side of Pyramid Peak). Abundant 1 year fir seedlings, 50 fir under 10 feet, 20 Murr (Pinus contorta) pines 2-10 feet, 5 Pimo (Pinus monticola) 1—3 feet. Hu- mus | inch. Cattle grazed, no sheep within 5 years. DISCUSSION Sudworth’s noted recent evidence of fire in many stands and believed fires were ignited by sheep herders to increase forage production and by log- gers to consume slash fuels (Fig. 2). This burning apparently did not spread extensively because fire scar analysis in the Sierra Nevada have documented the almost complete removal of surface fires in 2000] Fic. 3. sheep, gullying. many mixed conifer forests in the 1860—1870’s (Kilgore and Taylor 1979; Swetnam et al. 1990; Swetnam et al. 1992; Caprio and Swetnam 1995) at the same time burning was reportedly being used by loggers and sheep herders (Sudworth 1900; McKelvey and Johnston 1992; Stephens and E]- lhiott-Fisk 1998). Regeneration was noted in the majority of “‘av- erage’? mixed conifer stands. More site resources (light, water, nutrients) were probably available for regeneration in the “‘average’’ mixed conifer stands because of their lower stocking and basal areas. Re- generation in mixed conifer forests probably oc- curred prehistorically when small gaps were created by the interaction of fire and locally high fuel loads (Stephens et al. 1999). Regeneration was noted in half of the red fir stands and livestock grazing was noted in all stands. Sudworth noted that in some high elevation sites sheep were actually grazing on conifer seed- lings (Sudworth 1899). Many photos in the collec- tion show complete bare mineral soils (Fig. 3) and STEPHENS: MIXED CONIFER AND RED FIR FOREST STRUCTURE AND USES IN 1899 49 2) co eiak (spe Pa E, 5 ~~ Pa: S Hoe ght) ne Ps Peta 4A Dae pa seedlings were reportedly also trampled by live- stock. Early logging operations had a dramatic effect on the species composition and diameter distribu- tions of mixed conifer stands sampled by Sudworth (Table 3). The majority of the Pinus spp. and Pseu- dotsuga menziesii were harvested in the stands leaving large amounts of Calocedrus decurrens and Abies concolor (Fig. 4). This type of logging op- eration has been described as “‘high-grading”’ be- cause of the preference for large trees of particular genera. In this period it was common for all mer- chantable trees to be removed during logging op- erations (Laudenslayer and Darr 1990). Abies con- color and Calocedrus decurrens were therefore left because they were of relatively low economic value late in the 19" centaury. Early logging operations coupled with a national fire suppression policy that began in the early 20" century favored shade tolerant species such as Cal- ocedrus decurrens and Abies concolor. Climate changes over this period (wetter than average) may 50 MADRONO eas Spt Saat PS Fic. 4. El Dorado county, 1899. Forest logged out 5—6 years ago. Sugar pine (Pinus lambertiana), large ponderosa pine (Pinus ponderosa), Douglas-fir (Pseudotsuga menziesii) taken out, Kellogg oak (Quercus kelloggii) and incense cedar (Calocedrus decurrens), 12—25 per ha., remain (5—10 per acre). Reproduction of incense-cedar, ponderosa pine, white fir, and Douglas-fir abundant 2—9 m high (6-30 feet), 2-8 years old. Ground grazed. Half mile south of Blairs Mill at Sly Park. Humas 4—10 cm deep (2-4 inches). Soil deep brown, sandy loam. have also led to an increase in tree densities in Si- erra Nevada forests. The management of Quercus kelloggii is receiv- ing increased attention in the Sierra Nevada Frame- work Project (SNFP) Environmental Impact State- ment because several rare species such as the Cal- ifornia spotted owl and Pacific fisher use this spe- cies for foraging and denning habitat (USDA 2000). Quercus kelloggii is shade intolerant, and therefore, would have difficulty living in areas dominated by large mixed conifers because it can be over-topped and killed which is one explanation of why it was not recorded in any large mixed co- nifer stands. Quercus kelloggii did contribute to 6 percent of average stand stocking on the less stocked “‘average’’ mixed conifer stands because these stands were composed by smaller trees, and therefore, more site resources were probably avail- able for the oaks. There was a great deal of variability in snag den- sities in the stands, particularly in the red fir forest type. Tree density was also much higher in the red fir forest type when compared to the mixed conifer forests. Snag basal area was almost identical in the large mixed conifer stands and red fir stands (4.6 m’/ha and 4.5 m/*/ha, respectively). Snag densities found in the large mixed conifer stands are on the lower end of the current requirements for California spotted owls (Verner et al. 1992). Very little snag information exists for red fir forests making it dif- ficult to compare this historic data to contemporary data. The average basal area recorded in the mixed conifer stands is high, even for those labeled as ‘average.’ The SNFP Environmental Impact Re- port is defining desired conditions in mixed conifer forests as having basal areas below 70 m’/ha (300 ft?/acre). The large mixed conifer stands Sudworth inventoried had over three times this basal area and the “‘average’’ mixed conifer stands had double the 2000] basal area. Some areas of mixed conifer forest in the Sierra Nevada have the ability to produce much larger trees. The average quadratic mean diameter of the large mixed conifer stands from this study (110 cm) is equal to those recorded in the 8 mixed conifer stands in the southern Sierra Nevada (110 cm) for all trees greater than 30.5 cm DBH. (Stephens and Elliott-Fisk 1998). Omitting Sequoiadendron gi- ganteum (Lindley) Buchholz (giant sequoia) data from the four Sequoiadendron giganteum —mixed conifer stands in the southern Sierra Nevada pro- duced an average DBH of the remaining trees of 111 cm which is also very similar to those recorded above. Average tree density was higher in the southern Sierra Nevada when compared to the central and northern Sierra (278 trees/ha compared to 235 trees/ha, respectively). Average stand basal area was also higher in the mixed conifer stands from the southern Sierra Nevada when compared to the large mixed conifer stands from this study (271 m/?/ ha versus 215 m/7’/ha, respectively). Since the av- erage DBH was equal in the mixed conifer stands the increase in basal area is a result of increased stocking in the southern Sierra Nevada (18 percent higher). Abies concolor was very rare in the “‘average”’ mixed conifer stands but was the most common tree in the large mixed conifer stands. Low amounts of Abies concolor in the ‘“‘average’’ mixed conifer stands may have occurred because these stands were probably less developed (younger) or they may have been in drier locations which would have favored pines over true fir species. In the southern Sierra Nevada Abies concolor contributed to 40 percent of average stand stocking and 28 percent of average stand basal area (Stephens and Elliott- Fisk 1998). In the large mixed conifer stands in this study, Abies concolor contributed to 46 percent of average stand stocking and 34 percent of average stand basal area indicating that Abies concolor was slightly more common in the central and northern Sierra Nevada stands sampled by George Sudworth. Pinus lambertiana was much more common in the southern Sierra Nevada when compared to the central and northern Sierra Nevada (19 percent of stocking, 36 percent of basal area versus 12 percent of stocking, 16 percent of basal area, respectively). This difference can be partially explained by the presence of Pseudotsuga menziesii in relatively large amounts (16 percent of stocking, 24 percent of basal area) in the northern Sierra Nevada where- as Pseudotsuga menziesii is not native to the south- ern Sierra Nevada. Both Pseudotsuga menziesii and Pinus lambertiana are classified as shade interme- diate (in between shade tolerant and shade intoler- ant) and therefore, Pseudotsuga menziesii may have occupied areas that Pinus lambertiana could have also dominated. STEPHENS: MIXED CONIFER AND RED FIR FOREST STRUCTURE AND USES IN 1899 51 CONCLUSION The mixed conifer stands sampled by George Sudworth in 1899 were dominated by large trees at relatively low densities. Shade intolerant species, particularly Pinus ponderosa, dominated the ‘“‘av- erage’” mixed conifer stands whereas the large mixed conifer stands were composed of shade tol- erant, intermediate, and shade intolerant species. Early harvesting operations removed the major- ity of the economically viable species (Pinus spp. and Pseudotsuga menziesii) and left a large amount of Calocedrus decurrens and Abies concolor. This practice coupled with fire suppression policies ini- tiated at the beginning of the 20" century promoted the establishment and growth of shade tolerant spe- cies. There was a large amount of variability in snag densities, particularly in the red fir stands. The red fir stands had the highest tree densities and Abies magnifica dominated in these stands. ACKNOWLEDGMENTS I posthumously thank George Sudworth and his inven- tory crew for collecting the original data analyzed in this paper. I am grateful to Craig Olsen for introducing me to George Sudworth’s field notebooks. I wish to give special recognition to Norma Kobzina, Librarian at the University of California Biosciences and Natural Resources Library, for her assistance on this project. LITERATURE CITED BONNICKSEN, T. M. AND E. C. STONE. 1982. Reconstruction of a presettlement giant sequoia-mixed conifer forest community using the aggregation approach. Ecology 63(4):1134-1148. Caprio, A. C. AND T. W. SWETNAM. 1995. Historic fire regimes along an elevational gradient on the west slope of the Sierra Nevada, California. in J. K. Brown, (ed.), Proceedings: Symposium on fire in wil- derness and park management. USDA Forest Service Gen. Tech. Rep. INT-320. Intermoutain Research Sta- tion, Ogden, UT. GUTIERREZ, R.J., J. WERNER, K. S. MCKELVEY, B. R. NOON, G. N. STEGER, D. R. CALL, W. S. LAHAYE, B. B. BINGHAM, AND J. S. SENSER. 1992. Habitat relations of the California spotted owl. USDA Forest Service Gen. Tech. Rep. PSW-133. Pacific Southwest Re- search Station, Albany, CA. HARTESVELDT, R. J. AND H. T. HARVEY. 1967. The fire ecology of sequoia regeneration. Proceedings of the Tall Timbers Fire Ecology Conference, Tallehassee, FL. 7:65—77 KILGoRE, B. M. AND D. TAyYLor. 1979. Fire history of a sequoia mixed conifer forest. Ecology 60(1): 129- 142. LAUDENSLAYER, W. E AND H. H. Darr. 1990. Historical effects of logging on forests of the Cascade and Sierra Nevada Ranges of California. Transactions of The Western Section of the Wildlife Society 26:12—23. LEOPOLD, S. A., S. A. CAIN, C. A. CoTTaM, I. N. GABRIEL- SON, AND T. L. KIMBALL. 1963. Wildlife management in the national parks. American Forestry 69:32—35, 61-63. =) MADRONO McKeLvey, K. S. AND J. D. JOHNSTON. 1992. Historical perspectives on forests of the Sierra Nevada and the Transverse Ranges of southern California: Forest con- ditions at the turn of the century. in J. Verner et al. (eds.), USDA Forest Service Gen. Tech. Rep. PSW- 133. Pacific Southwest Research Station, Albany, CA. MILLAR, C. I. AND W. B. WOOLFENDEN. 1999. The role of climate change in interpreting historical variability. Ecological Applications, vol. 9:1207—1216. Parsons, D. J. AND S. H. DEBENDEETTI. 1979. Impact of fire suppression on a mixed-conifer forest. Forest Ecology and Management 2(1):21-33. RUNDEL, P. W., D. J. PARSONS, AND D. T. GORDON. 1977. Montane and subalpine vegetation of the Sierra Ne- vada and Cascade Ranges. in M. FE Barbour and J. Maor (eds.), Terrestrial vegetation of California, John Wiley and Sons. STEPHENS, S. L. 1998. Effects of fuels and silvicultural treatments on potential fire behavior in mixed conifer forests of the Sierra Nevada, CA. Forest Ecology and Management 105:21-—34. , D. DULITZ, AND R. E. MARTIN. 1999. Giant se- quoia regeneration in group selection openings of the southern Sierra Nevada. Forest Ecology and Manage- ment 120:89-95. AND D. E. ELLiotT-Fisk. 1998. Sequoiadendron giganteum-mixed conifer forest structure in 1900— 1901 from the southern Sierra Nevada, CA. Madrofo, vol. 45(3):221—230. SuDWoRTH, J. B. 1899. Unpublished field note books of Sierra Nevada forest reserve inventory. (University of California, Berkeley, Bioscience and Natural Re- sources Library) . 1900. Stanislaus and Lake Tahoe forest reserves, California, and adjacent territory. Jn Annual reports of the Department of the Interior, 21st report of the U.S. Geological Survey, 56th Congress, 2nd session, [Vol. 47 | senate document #3. Washington D.C. Government | printing office. | SWETNAM, T. W., C. H. BAISAN, P. M. BRown, A. C. Ca- | PRIO, AND R. TOUCHAN. 1990. Late Holocene fire and © climate variability in giant sequoia groves. Bulletin of the Ecological Society of America 71(2): 342. , BAISAN, C. H., CApRio, A. C., TOUCHAN, R., AND P. M. Brown. 1992. Tree ring reconstruction of giant sequoia fire regimes. Final report to Sequoia-Kings | Canyon and Yosemite National Parks. Cooperative | agreement DOI 8018-1-002, Laboratory of tree ring research, University of Arizona, Tucson, AZ. UNITED STATES DEPARTMENT OF AGRICULTURE. 2000. In | Press. United States Forest Service Sierra Nevada | Framework Project Draft Environmental Impact © Statement. Pacific Southwest Region, Mare Island, CA. VAN WAGTENDONK, J. W. 1996. Use of a deterministic fire growth model to test fuel treatments. Sierra Nevada Ecosystem Project, final report to congress, vol. II, Assessments and Scientific Basis for Management Options. Davis: University of California, Centers for Water and Wildland Resources. VANKAT, J. L. AND J. Major. 1978. Vegetation changes in Sequoia National Park, California. Journal of Bioge- ography 5:377—402. VERNER, J., K. S. MCKELVEY, B. R. Noon, R. J. GUTIER- REZ, G. I. GOULD, AND T. W. Beck. 1992. The Cali- fornia Spotted Owl: A technical assessment of its cur- rent status. USDA Forest Service Gen. Tech. Rep. PSW-133. Pacific Southwest Research Station, Alba- ny, CA. WEATHERSPOON, C. P. AND C. N. SKINNER. 1996. Fire sil- viculture relationships in Sierra forests. Sierra Nevada Ecosystem Project, final report to congress, vol. II, Assessments and Scientific Basis for Management Options. Davis: University of California, Centers for Water and Wildland Resource. Maprono, Vol. 47, No. 1, pp. 53-60, 2000 OLD-GROWTH FOREST ASSOCIATIONS IN THE NORTHERN RANGE OF COASTAL REDWOOD THOMAS M. MAHONY College of Natural Resources and Sciences, Humboldt State University, Arcata, CA 95521 JOHN D. STUART! Department of Forestry, Humboldt State University, Arcata, CA 95521, (707) 826-3823 ABSTRACT Old-growth Sequoia sempervirens (D. Don) Endl. (redwood) forests occurring in northwestern Cali- fornia and southwestern Oregon were classified and described using data from 206 systematically placed plots. Data were collected from Jedediah Smith Redwoods State Park, Del Norte Coast Redwoods State Park, northern Redwood National Park, and the southwestern portion of the Siskiyou National Forest. Plot data were analyzed using TWINSPAN and polar ordination. Six associations within the redwood series were classified: Sequoia sempervirens/Polystichum munitum (Kaulf) C. Pres] (SESE/POMU), Se- quoia sempervirens-Pseudotsuga menziesii (Mirbel) Franco/Rhododendron macrophyllum D. Don (SESE- PSME/RHMA), Sequoia sempervirens-Tsuga heterophylla (Raf.) Sarg./Vaccinium ovatum Pursh (SESE- TSHE/VAOV), Sequoia sempervirens-Tsuga heterophylla/Polystichum munitum (SESE-TSHE/POMU), Sequoia sempervirens-Tsuga heterophylla/Rubus spectabilis Pursh (SESE-TSHE/RUSP), and Sequoia sempervirens-Alnus rubra Bong./Rubus spectabilis (SESE-ALRU/RUSP). Discriminant analysis was used to assess the relationships between abiotic site variables and classified floristic associations. Elevation and coastal proximity explained 81.1 percent of the variation among associations. Aspect and topographic position explained 14.2 percent of the remaining variation. Moisture was the primary environmental variable controlling the distribution of classified forest associations. Sequoia sempervirens (D. Don) Endl. (redwood) forests are endemic to coastal margins and mesic inland sites from central California to southern Or- egon. Along this broad latitudinal gradient, S. sem- pervirens is limited to a narrow belt 10 to 50 ki- lometers wide (Roy 1966; Fox 1989). The extreme northern range of S. sempervirens has not been ad- equately classified and described. Vast tracts of old- growth forest in Jedediah Smith Redwoods State Park, Del Norte Coast Redwoods State Park, and northern sections of Redwood National Park have been virtually ignored in the S. sempervirens liter- ature. The difficult access, steep terrain, and huge volume of coarse woody debris characterizing in- terior portions of these parks may explain the dearth of botanical information in the region. As a result of this relative isolation, these parks contain some of the most primeval and undisturbed old- growth (Helms 1998) redwood vegetation in exis- tence. Southwestern Siskiyou National Forest con- tains a patchy network of old-growth representing the northernmost natural S. sempervirens stands. Since they exist at the terminus of the redwood range, these stands are ecologically significant. They may give insight into processes affecting oth- er parts of the range, including gradients in soil moisture and temperature that affect species com- position and stand dynamics. 'To whom correspondence should be addressed. METHODS Study Area. The northern range of redwood, as defined in this study, includes Jedediah Smith Red- woods State Park, Del Norte Coast Redwoods State Park, and northern Redwood National Park, all lo- cated in northern California, and portions of Sis- kiyou National Forest located in southwestern Or- egon. It extends from 41°47’N to 42°10'N, and 124°4'W to 124°12'W. The study area is topograph- ically diverse—elevations range from sea level to over 490 m. Rocks of the Franciscan Formation, a subduction complex consisting of accreted frag- ments of oceanic crust and forearc sediments, un- derlay most of the region (Aalto and Harper 1989). Soils were mapped as predominantly Melbourne and Empire series by the California State Cooper- ative Soil-Vegetation Survey (Smith et al. 1977; Delapp et al. 1978). Crescent City, CA is the closest weather station to the study area. Precipitation data (1948—2000) indicated that maximum precipitation fell during December and January, averaging 27.7 cm and 29.6 cm, respectively. The least amount fell during July and August, averaging 1.0 cm and 2.0 cm, respec- tively. Annual average precipitation was 168.1 cm. The highest mean temperatures occurred in Cres- cent City during August and September, at 14.9°C and 14.6°C, respectively. The lowest mean temper- atures occurred in December and January, at 8.8°C 54 MADRONO TABLE 1. COVER ABUNDANCE SCALE AND MIDPOINTS USED IN OCULAR ESTIMATES. Cover class Range of cover (%) Class midpoints (%) 8 75—100 87.5 | 50-75 62.5 6 25-50 SES 5 5-25 15 4 1-5 3 3 0.1-1 0.6 Z 0.01—0.1 0.06 1 0.001—0.01 0.006 and 8.7°C, respectively (Western U.S. Climate His- torical Summaries 2000). Study area vegetation conforms to the Society of American Foresters redwood forest cover type (Eyre 1980). East and north of the study area, the Douglas-fir forest cover type dominates. Sampling. For Jedediah Smith Redwoods State Park, Del Norte Coast Redwoods State Park, and Redwood National Park, old-growth forest was identified on 1:12,000 color infrared aerial photo- graphs. Two hundred plots were stratified based on three elevation classes (O—105 m, 106—215 m, and >215 m), and placed onto USGS topographic quad- rangles in a systematic grid 485 meters apart. For Siskiyou National Forest, four old-growth polygons were identified on maps obtained from the USDA Forest Service GIS database, and trans- ferred to 1:24,000 USGS topographic quadrangles. Five plots were placed in each polygon via a sys- tematic sampling grid, with plot spacing propor- tional to polygon area. Riparian zones within Jedediah Smith Redwoods State Park were sampled separately to best char- acterize this unique and diverse vegetation. Fifteen sample plots were systematically placed approxi- mately 1500 m apart (total stream length/15) along Cedar Creek, Mill Creek, Clarks Creek, and several other unnamed perennial drainages within park boundaries. Of 235 plots slated for sampling, 206 were even- tually field checked. The remaining plots were not sampled because of difficult or dangerous access, or the plot was not in an old-growth forest. The 206 circular 0.05 ha (S00 m7) plots were thoroughly searched and all vascular plant species identified and recorded with an ocular cover estimate using a modified Braun-Blanquet cover abundance scale (Table 1; Mueller-Dombois and Ellenberg 1974). Tree species were tallied based on stem density in three height classes: O—3 m, 3-10 m, and >10 m. Basal area, taken from plot center, was estimated using a “‘cruise angle”’ sighting device for canopy (dominant, co-dominant, and intermediate crown classes) species. Elevation was determined with a pocket altimeter and topographic map. Slope angle was recorded in percent using a clinometer. Aspect [Vol. 47 was assessed with a hand compass. Distance from the ocean was estimated using a topographic map. Topographic position was recorded for each plot. Data Analysis. Two-Way Indicator Species Anal- ysis (TWINSPAN) (Hill 1979) was used to simul- taneously classify species and samples. Only spe- cies occurring in greater than 5 percent of plots were used in the analysis (Gauch 1982). Plots were analyzed in TWINSPAN with species cover cut levels of 0.6, 3.0, 15.0, 37.5, 62.5, and 87.5 percent. The 15.0 and 37.5 cut levels were weighted to em- phasize dominance (Stuart et al. 1996). TWIN- SPAN groupings were analyzed using a _ polar (Bray-Curtis) ordination, to further analyze and re- fine the TWINSPAN output. Species richness was determined by randomly selecting 5 plots from each association and calculating the mean number of species per plot (Stuart et al. 1996). In addition, stem density per hectare in three height classes and canopy species basal area were averaged for each association. Discriminant analysis was performed using NCSS 2000 (Hintze 1998) to relate floristic asso- ciations with abiotic site characteristics. Elevation, slope angle, coastal proximity, and a Moisture Equivalency Index (MEI) were used as abiotic vari- ables in the discriminant analysis. The MEI was adapted from Sawyer and Thornburgh (1974) and Matthews (1986). It incorporates topographic po- sition and aspect, two variables important to soil moisture. A lower index number (1—15) assumes greater soil moisture available to plants. RESULTS AND DISCUSSION TWINSPAN and polar ordination analysis pro- duced six groups that were interpreted as associa- tions (Fig. 1). Groups were consistent with vege- tation units observed in the field. All associations were in the Sequoia sempervirens series, with Pseu- dotsuga menziesii (Mirbel) Franco, Tsuga hetero- phylla (Raf.) Sarg., and Alnus rubra Bong. sub-se- ries. The first TWINSPAN division separated groups based on understories dominated by either Vaccinium ovatum Pursh or Polystichum munitum (Kaulf) C. Presl. Within these broad groupings, subsequent TWINSPAN division levels reflected groupings based on other indicator understory spe- cies such as Lithocarpus densiflorus (Hook. & Arn.) Rehder, Rhododendron macrophyllum_ D. Don, and Rubus spectabilis Pursh. The following association descriptions are presented from rela- tively dry types to wet types. A more detailed treat- ment of the associations can be found in Mahony (1999); The Sequoia sempervirens-Pseudotsuga menzie- sit/Rhododendron macrophyllum Association. Total vegetation cover averaged 85 percent, and total overstory cover averaged 68 percent. Over- stories were dominated by Sequoia sempervirens and Pseudotsuga menziesii, with mean cover values 2000] MAHONY AND STUART: OLD-GROWTH FOREST ASSOCIATIONS 55 Level Understories dominated by Understories dominated by 0 Vaccinium ovatum Polystichum munitum 206 Tsuga heterophylla Tsuga heterophylla 1 sparse dense Inland/non-riparian Coastal/riparian 77 129 Lithocarpus | Lithocarpus densiflorus densiflorus Alnus rubra Alnus rubra 2 common uncommon uncommon common 112 17 3 27 50 64 48 7 10 SESE-PSME/RHMA SESE-TSHE/VAOV SESE/POMU SESE-TSHE/POMU SESE-TSHE/RUSP SESE-ALRU/RUSP Fic. 1. Dendrogram of TWINSPAN classification. Numbers beneath lines represent the number of plots prior to division. Numbers above association acronyms are the number of plots in each classified association. Association acronyms are: SESE-PSME/RHMA = Sequoia sempervirens-Pseudotsuga menziesii/Rhododendron macrophyllum, SESE-TSHE/VAOV = Sequoia sempervirens-Tsuga heterophylla/Vaccinium ovatum, SESE/POMU = Sequoia sem- pervirens/Polystichum munitum, SESE-TSHE/POMU = Sequoia sempervirens-Tsuga heterophylla/Polystichum muni- tum, SESE-TSHE/RUSP = Sequoia sempervirens-Tsuga heterophylla/Rubus spectabilis, SESE-ALRU/RUSP = Se- quoia sempervirens-Alnus rubra/Rubus spectabilis. of 43 and 31 percent, respectively, and mean con- stancies of 96 and 100 percent, respectively (Table 2). Tsuga heterophylla was occasionally present but contributed minimal cover. Lithocarpus densiflorus dominated the sub-canopy. Basal area averaged 123 m’/ha (Table 3). The shrub layer was extremely dense. Vaccinium ovatum and Rhododendron macrophyllum dominat- ed, with mean cover values of 47 and 35 percent, respectively, and mean constancies of 100 percent each. Berberis nervosa Pursh, Gaultheria shallon Pursh, Rhamnus purshiana DC., and Vaccinium parvifolium Smith each had greater than 30 percent constancy but less than 3 percent cover. The herb layer was virtually absent. Polystichum munitum was the most dominant species in this lay- er with 7 percent cover and 93 percent constancy. Disporum hooker (Torrey) Nicholson, Galium tri- florum Michaux, Oxalis oregana Nutt., Trillium ovatum Pursh, and Viola sempervirens E. Greene were common but contributed negligible cover. The Sequoia sempervirens-Pseudotsuga menzie- sii/Rhododendron macrophyllum association was generally found on upper slopes and ridges in Sis- kiyou National Forest and Del Norte Coast Red- woods State Park. Elevations ranged from 58—470 m, averaging 312 m. Distance from the ocean av- eraged 8.5 km. Slopes averaged 43 percent, and Moisture Equivalency Index (MEI) scores averaged 9.8. Species richness averaged 13.6 species (Table 4). Vegetation dynamics. Lithocarpus densiflorus and Sequoia sempervirens dominated reproduction. Veirs (1979) suggested S. sempervirens and L. den- siflorus were components of the “‘climax’’ vegeta- tion and would remain in the stand regardless of disturbance such as fire. The presence of S. sem- pervirens in all height classes represents an uneven age structure for redwood. Pseudotsuga menziesii is a seral species that will disappear from stands without major disturbance (Daubenmire 1975; Veirs 1979; Eyre 1980). Low P. menziesii stem densities for the O—3 m and 3-10 m height classes, and high density of trees >10 m, suggested that a cohort resulted from disturbance, and additional disturbance will be necessary for continued pres- ence of P. menziesii in this association. Relationships to previous classifications. The Se- quoia sempervirens-Pseudotsuga menziesii/Rhodo- dendron macrophyllum association closely resem- bled the midslope stands encountered by Dyrness et al. (1972) in Wheeler Creek Research Natural Area, and Tanoak-coast redwood association stands described by Atzet and Wheeler (1984) for south- western Oregon. Other similar types include the Se- quoia sempervirens-Pseudotsuga menziesti/Vaccin- ium ovatum association described by Matthews 56 MADRONO [Vol. 47 TABLE 2. AVERAGE COVER AND CONSTANCY FOR SPECIES USED IN TWINSPAN ANALYSIS. Species reported are those with >50 percent constancy. See Figure | for plant association acronyms. Cov = average cover (%). Con = constancy (%). SESE- PSME/ SESE-TSHE/ SESE-TSHE/ SESE-TSHE/ SESE- RHMA VAOV SESE/POMU POMU RUSP ALRU/RUSP Cov Con Cov Con Cov Con Cov Con Cov Con Cov Con Maianthemum dilatatum <1 a2 Menziesia ferruginea 3 69 Pseudotsuga menziesii 31 100 17 78 Rhododendron macrophyllum 35-100 13 76 Viola sempervirens <1 74 <1 90 <1 61 Disporum hookeri <1 54 <1 66 Lithocarpus densiflorus 45 100 21 94 16 5 8 57 Vaccinium ovatum 47 100 J! 98 16 97 14 98 =) a7 2 70 Tsuga heterophylla 41 100 24 a 39 88 pal 86 Trillium ovatum <1 93 <1 98 <1] 98 | 69 =] 86 Vaccinium parvifolium 1 70 2 70 3 73 S, 94 3 86 Blechnum spicant 3 72 4 WS it 96 10. +100 <1 50 Polystichum munitum 7 93 12 100 55 100 67 100 30 ~=100 24 100 Sequoia sempervirens 43 96 37 ~=100 60 100 a6 98 22 86 36 80 Oxalis oregana =<] D2 <1 78 13 98 9 100 16 100 Disporum smithii =|. 67 <1 100 Rhamnus purshiana <1 60 1 a7 Gaultheria shallon 2 78 2 74 5 80 1 fis) 6 aM 10 80 Galium triflorum <1 54 = 1 61 <1 57 a | 60 Vancouveria hexandra =| 55 <1 i <1 10 Dryopteris expansa 2 1 1 86 2 50 Rubus spectabilis 5 63 25 100 22 100 Acer circinatum 19 VW Acer macrophyllum 10 a7 Adiantum aleuticum i | vial Corylus cornuta 14 57 Ribes bracteosum 1 57 Asarum caudatum =] 71 Al 70 Athyrium filix-femina <1 58 2 100 2 60 Tolmiea menziesii <1 Th 2 60 Alnus rubra 38 ~=100 Claytonia sibirica =<] 80 Marah oreganus > 80 Polypodium scouleri <1 50 Rubus parviflorus 5 90 Sambucus racemosa 3 70 Stachys ajugoides << 90 (1986) and the Sequoia sempervirens/Arbutus men- The Sequoia sempervirens-Tsuga heterophylla/ ziesti Pursh association described by Lenihan Vaccinium ovatum Association. (1986). This association might be considered an ex- Total vegetation cover averaged 88 percent, and tension of the Pseudotsuga-hardwood forests de- total overstory cover averaged 74 percent. Sequoia scribed by Sawyer et al. (1977). sempervirens, Tsuga heterophylla, and Pseudotsu- TABLE 3. MEAN BASAL AREA (M?/HA) FOR CANOPY SPECIES BY ASSOCIATION. SESE- SESE- SESE- SESE- SESE- PSME/RHMA TSHE/VAOV- SESE/POMU TSHE/POMU TSHE/RUSP ALRU/RUSP Sequoia sempervirens 86.0 114.0 165.0 170.0 73.0 87.0 Pseudotsuga menziesii 37.0 21.0 10.0 2.0 2.0 7.0 Tsuga heterophylla 0.4 23.0 15.0 23.0 11.0 0.0 Picea sitchensis 0.0 0.0 0.0 4.0 6.0 10.0 Abies grandis 0.0 3.0 1.0 0.2 0.0 3.0 Total basal area 123.4 161.0 191.0 199.2 92.0 107.0 | 2000] MAHONY AND STUART: OLD-GROWTH FOREST ASSOCIATIONS a7 - TABLE 4. ENVIRONMENTAL CHARACTERISTICS, TREE DENSITY IN THREE HEIGHT CLASSES, AND SPECIES RICHNESS FOR EACH _ ForREST ASSOCIATION. SESE- SESE- SESE- SESE- SESE- PSME/RHMA TSHE/VAOV- SESE/POMU TSHE/POMU TSHE/RUSP ALRU/RUSP Elevation (m) 312.0 161.0 143.0 114.0 67.0 136.0 ' Distance (km) 8.5 TA 6.4 55 6.8 3.6 Slope (%) 42.9 36.2 36.2 34.9 38.0 49.7 MEI (1-15) 9.8 9.0 HT ais 7.0 3 6.7 _ Stems/ha: 0-3 m 71.8 87.6 114.0 54.0 68.6 74.0 3-10 m 127.6 76.8 86.6 66.2 51.6 168.0 >10m 180.8 206.0 172.4 165.8 72.0 170.0 Sp. Richness 13.6 16.6 19.0 156 26.8 19.4 ga menziesii dominated the canopy, with mean cov- ers of 37, 41, and 17 percent, respectively, and mean constancies of 100, 100, and 78 percent, re- spectively (Table 2). Abies grandis (Douglas) Lind- ley appeared occasionally in the canopy. Lithocar- pus densiflorus was common in the subcanopy. Ba- sal area averaged 161.0 m’/ha (Table 3). The shrub layer was dense. Vaccinium ovatum dominated, averaging 51 percent cover and 98 per- cent constancy. Rhododendron macrophyllum had 13 percent cover and 76 percent constancy. Ber- beris nervosa, Gaultheria shallon and Vaccinium parvifolium each had greater than 40 percent con- stancy but less than 2 percent cover. Corylus cor- nuta Marsh. and Rhamnus purshiana occurred spo- radically. The sparse herb layer was dominated by Polys- tichum munitum, averaging 12 percent cover and 100 percent constancy. The Sequoia sempervirens-Tsuga heterophylla/ Vaccinium ovatum association was usually found on inland upper slopes and ridges in Jedediah Smith Redwoods State Park. Elevations ranged from 40— 460 m, averaging 161 m. Distance inland averaged 7.4 km. Slopes averaged 36 percent, and MEI scores averaged 9. Species richness averaged 16.6 species (Table 4). Vegetation dynamics. Tsuga heterophylla and Lithocarpus densiflorus dominated reproduction. Tsuga heterophylla seedlings were particularly abundant on downed logs. Combs (1984) noted a similar pattern of T. heterophylla regeneration in the Little Lost Man Creek Research Natural Area in Redwood National Park. He suggested that few seedlings would reach maturity because of vulner- ability to fire and disease. Daubenmire (1975) noted extensive T. heterophylla in all size classes in Je- dediah Smith Redwoods State Park, but believed the species would decline without disturbance. Veirs (1979) suggested that light ground fires, un- affecting the canopy, will favor T. heterophylla re- generation. The high density of 7. heterophylla and the complete absence of P. menziesii seedlings sug- gested a light fire regime, sufficient for S. semper- virens and T. heterophylla regeneration, but not for regeneration of P. menziesii. Relationships to previous classifications. The Se- quoia sempervirens-Tsuga heterophylla/Vaccinium ovatum association was unique compared to other redwood types described in the literature due to the importance of Tsuga heterophylla. While other red- wood classifications have noted the presence of T. heterophylla (Dyrness et al. 1972; Atzet and Whee- ler 1984; Lenihan 1986), none have shown such dominance by this mesic conifer. The Sequoia sem- pervirens/Berberis nervosa association described by Lenihan (1986), and the Tsuga phase of the Pseudotsuga-hardwood forests described by Saw- yer et al. (1977) were similar in composition to this association. The Sequoia sempervirens/Polystichum munitum Association. Total vegetation cover averaged 90 percent. Total overstory cover averaged 76 percent. Sequoia sem- pervirens dominated the canopy with 60 percent cover and 100 percent constancy (Table 2). Tsuga heterophylla was common, and Pseudotsuga men- ziesii appeared occasionally in the canopy. Abies grandis, Cupressus lawsoniana A. Murray and Um- bellularia californica (Hook. & Arn.) Nutt. oc- curred sporadically, contributing minimal cover. Lithocarpus densiflorus was ubiquitous in the sub- canopy. Basal area averaged 191.0 m*/ha (Table 3). Vaccinium ovatum dominated the relatively sparse shrub layer, averaging 16 percent cover and 97 percent constancy. Gaultheria shallon, Rhodo- dendron macrophyllum, and Vaccinium parvifolium each had greater than 40 percent constancy but less than 5 percent cover. Acer circinatum Pursh, Ber- beris nervosa, Corylus cornuta, and Rubus spec- tabilis occurred sporadically, contributing minimal cover. Herbaceous cover and species diversity was moderately high. Polystichum munitum dominated, averaging 55 percent cover and 100 percent con- stancy. Oxalis oregana was extremely common. The Sequoia sempervirens/Polystichum munitum 58 MADRONO association was found throughout the study area, generally on lower and middle slopes at moderate distances from the ocean. Elevations ranged from 21-369 m, averaging 143 m. Distance from the ocean averaged 6.4 km. Slopes averaged 36 per- cent, and MEI scores averaged 7.5. Species rich- ness averaged 19 species (Table 4). Vegetation Dynamics. Lithocarpus densiflorus and Sequoia sempervirens dominated reproduction. The moderate levels of Abies grandis, Tsuga het- erophylla, and L. densiflorus reproduction may be indicative of the light fire regime in intermediate to mesic sites referred to by Veirs (1979). However, he noted that these species exhibited an all aged pattern and can reproduce regardless of fire. Relationships to previous classifications. The Se- quoia sempervirens/Polystichum munitum associa- tion contained elements of the Sequoia sempervi- rens/Blechnum spicant (L.) Smith association de- scribed by Lenihan (1986), though Lenihan’s as- sociation appeared wetter. The dominance of Sequoia sempervirens, the sparse shrub layer, and the well-developed herb layer related this associa- tion to Becking’s (1967) Redwood-oxalis alliance. The Sequoia sempervirens-Tsuga heterophylla/ Polystichum munitum Association. Total vegetation cover averaged 92 percent, and total overstory cover averaged 75 percent. Sequoia sempervirens and Tsuga heterophylla dominated the canopy, with mean covers of 53 and 39 percent, respectively, and mean constancies of 98 and 88 percent, respectively (Table 2). Abies grandis, Lith- ocarpus densiflorus, Picea sitchensis (Borg.) Car- riere and Pseudotsuga menziesii occurred sporadi- cally, contributing minimal cover. Thuja plicata D. Don appeared occasionally in mesic sites. Basal area averaged 199.2 m’/ha (Table 3). The shrub layer was generally not well devel- oped. Vaccinium ovatum was the most abundant shrub, averaging 14 percent cover and 98 percent constancy. Menziesia ferruginea Smith, Rubus spec- tabilis, Vaccinium parvifolium, Gaultheria shallon, and Rhamnus purshiana each had greater than 60 percent constancy but less than 6 percent cover. The herbaceous layer was dense. Polystichum munitum dominated, averaging 67 percent cover and 100 percent constancy. Blechnum spicant and Oxalis oregana were common. The Sequoia sempervirens-Tsuga heterophyllal Polystichum munitum association was _ generally found at lower slopes and elevations, especially in southwestern areas of Jedediah Smith Redwoods State Park exposed to maritime influence. Eleva- tions ranged from 40-274 m, averaging 114 m. Distance inland averaged 5.5 km. Slopes averaged 35 percent, and MEI scores averaged 7. Species richness averaged 15.6 species (Table 4). Vegetation Dynamics. Tsuga heterophylla and Sequoia sempervirens dominated reproduction. Se- [Vol. 47 quoia sempervirens had fewer stems in the lower height classes relative to T. heterophylla, but the longevity and resilience of S. sempervirens makes abundant individuals in the reproduction layers un- necessary to ensure continued dominance. Relationships to previous classifications. The Se- quoia sempervirens-Tsuga heterophyllal/Polysti- chum munitum association, like Sequoia sempervi- rens-Tsuga heterophylla/Vaccinium ovatum, ap- peared unlike any previously described redwood types. It was similar in many respects to the mesic Tsuga/Polystichum association described by Frank- lin and Dyrness (1973) for Oregon Coast Range forests in the Tsuga heterophylla Zone. Addition- ally, it contained elements of the Tsuga-picea/Oplo- panax horridum/Athyrium filix-femina association of Picea sitchensis Zone forests described by Franklin and Dyrness (1973). It related tangentially to Lenihan’s (1986) Sequoia sempervirens/Blech- num spicant association. The Sequoia sempervirens-Tsuga heterophyllal Rubus spectabilis Association. Total vegetation cover averaged 94 percent, and total overstory cover averaged 55 percent. Sequoia sempervirens and Tsuga heterophylla were canopy dominants, averaging 22 and 21 percent cover, re- spectively. Both species had 86 percent constancy (Table 2). Picea sitchensis and Thuja plicata were occasional to common in mesic sites. Pseudotsuga menziesii occurred sporadically. Acer macrophyl- lum was common, especially near stream channels. Lithocarpus densiflorus was common in the sub- canopy. Alnus rubra and Sambucus racemosa L. appeared occasionally. Basal area averaged 92.0 m*/ha (Table 3). Rubus spectabilis dominated the dense shrub lay- er, averaging 25 percent cover and 100 percent con- stancy. Acer circinatum and Corylus cornuta were abundant in this layer having 71 percent and 57 percent constancy and 19 percent and 14 percent cover, respectively. Other common shrubs having greater than 40 percent constancy but less than 6 percent cover included Gaultheria shallon, Menzie- sia ferruginea, Rhamnus purshiana, Ribes bracteo- sum Douglas, Rubus parviflorus, Vaccinium ova- tum, and V. parvifolium. The herbaceous layer was dense and floristically diverse. Polystichum munitum dominated with 31 percent cover and 100 percent constancy. Oxalis oregana and Blechnum spicant were abundant. The Sequoia sempervirens-Tsuga heterophylla/ Rubus spectabilis association was restricted to in- terior perennial drainages in Jedediah Smith Red- woods State Park. Elevations ranged from 37-122 m, averaging 67 m. Distance from the ocean aver- aged 6.8 km. Slopes averaged 38 percent, and MEI scores averaged 1.3. Species richness averaged 26.8 species (Table 4). Vegetation Dynamics. Tsuga heterophylla and Lithocarpus densiflorus dominated reproduction. 2000] MAHONY AND STUART: OLD-GROWTH FOREST ASSOCIATIONS Sy) TABLE 5. STANDARD CANONICAL COEFFICIENTS USED IN DISCRIMINANT ANALYSIS. Variable Variate | Variate 2 Variate 3 Variate 4 Elevation —0.874009 0.717388 0.045912 —0.37663 1 Slope 0.189664 0.275203 0.477507 0.858742 Distance —0.828487 —0.300679 —0.599034 0.465420 MEI —0.376680 — 1.100974 0.339726 0.143424 Riparian conditions produced the wettest and most floristically diverse association encountered in the study area. Conifer basal area was greatly reduced compared to other associations. Sequoia sempervi- rens attained its lowest basal area, but still domi- nated conifer basal area. The streamside environ- ment allowed mesic woody species such as Acer macrophyllum, A. circinatum, Corylus cornuta, and Rubus spectabilis to thrive. Relationships to previous classifications. The Se- quoia_ sempervirens-Tsuga _ heterophylla/Rubus spectabilis association appeared much wetter than any redwood association previously described. It shared many of the same riparian components, such as high cover of herbaceous and hardwood species, described by Dyrness et al. (1972) for lower slopes in the Wheeler Creek Research Natural Area in southwestern Oregon, and appeared similar in many respects to the Tsuga heterophyllal/Acer cir- cinatum/Polystichum munitum-Oxalis oregana as- sociation described by Franklin and Dyrness (1973) for alluvial terrace vegetation in the Tsuga hetero- phylla Zone of Oregon. The absence of Sequoia sempervirens in Tsuga heterophylla Zone forests makes comparison difficult, however. The Sequoia sempervirens-Alnus rubra/Rubus spectabilis Association. Total vegetation cover averaged 93 percent, and total overstory cover averaged 71 percent. Sequoia sempervirens dominated the canopy, averaging 36 percent cover and 80 percent constancy (Table 2). Picea sitchensis was common in coastal sites. Pseu- dotsuga menziesii and Abies grandis occurred spo- radically. Alnus rubra dominated the subcanopy. Acer macrophyllum, Lithocarpus_ densiflorus, Rhamnus purshiana, Sambucus racemosa, and Um- bellularia californica appeared occasionally in the subcanopy. Basal area averaged 107.0 m’/ha (Table 3). Rubus spectabilis dominated the moderately dense shrub layer, averaging 22 percent cover and 100 percent constancy. Gaultheria shallon had 10 percent cover and 80 percent constancy. Other common shrubs included Acer circinatum, Corylus cornuta, Rubus parviflorus, R. ursinus Cham & Schldl., and Vaccinium ovatum with constancies greater than 20 percent but with less than 8 percent cover. The herbaceous layer was diverse. Polystichum munitum dominated, averaging 24 percent cover and 100 percent constancy. Oxalis oregana oc- curred sporadically, but was generally abundant when it did occur. The Sequoia sempervirens-Alnus rubra/Rubus spectabilis association was generally found along the Smith River, or on coastal bluffs in Del Norte Coast Redwoods State Park. Elevations ranged from 18-299 m, averaging 136 m. Distance from the ocean averaged 3.6 km. Slopes averaged 50 per- cent, and MEI scores averaged 6.7. Species rich- ness averaged 19.4 species (Table 4). Vegetation dynamics. Sequoia sempervirens dominated reproduction. Picea sitchensis was very common on coastal bluffs, and Pseudotsuga men- ziesii was common along the Smith River. Alnus rubra achieved high cover in the subcanopy. The mesic, high light environments of the Smith River floodplain and exposed coastal bluffs provided fa- vorable conditions for this shade intolerant hard- wood (Hibbs et al. 1994; Harlow et al. 1996). Ad- ditionally, natural disturbance from Smith River flooding likely enhanced the competitive ability of A. rubra, which is more tolerant of flooding and poor drainage than its associates (Hibbs et al. 1994). Tolerance of salt spray and resistance to windthrow allowed A. rubra to thrive along the coastal bluffs of Del Norte Coast Redwoods State Park. Periodic disturbances likely benefited the ser- al Pseudotsuga menziesii. Alnus rubra showed high stem densities in the 3-10 and >10 m height class- es, but minimal density in the 0-3 m class, indi- cating many stands may be recovering from distur- bance. Relationships to previous classifications. The Se- quoia sempervirens-Alnus rubra/Rubus spectabilis association was similar to coastal sections of the Wildcat Hills transect described by Zinke (1977), as well as the red alder series described in Sawyer and Keeler-Wolf (1995). It should be noted that a pure Picea sitchensis forest type may exist imme- diately adjacent to the coast in Del Norte Coast Redwoods, but was not sampled. Discriminant Analysis. Discriminant analysis re- vealed that elevation, coastal proximity, and topo- graphic position/aspect (MEI) were statistically sig- nificant (P < 0.01) in discriminating among floristic associations. Elevation and coastal proximity had the greatest influence on the first discriminant func- tion (Table 5). This function explained 81.1 percent of the variation between groups. MEI had the great- est influence on the second discriminant function, 60 MADRONO which explained 14.2 percent of group variation. Together, the first two discriminant functions, influ- enced by elevation, distance to the ocean, and MEI, explained 95.3 percent of group variation. The physiographic factors influencing floristic associa- tions, in decreasing order of importance, were ele- vation, coastal proximity, and aspect/topographic position (MEI). ACKNOWLEDGMENTS We thank John Sawyer and Larry Fox for their help in project design and manuscript review. Thanks to Jim Belsher for field assistance, and Stephen Underwood at Redwood National and State Parks for logistical support. Partial funding came from a MclIntire-Stennis grant. LITERATURE CITED AALTO, K. R. AND G. D. HARPER. 1989. Geologic evolu- tion of the northernmost coast ranges and western Klamath mountains, California. American Geophysi- cal Union, Washington, D.C. ATZET, T. AND D. L. WHEELER. 1984. Preliminary plant associations of the Siskiyou Mountain Province. U.S.D.A. Forest Service, Pacific Northwest Region. BECKING, R. W. 1967. The ecology of the coastal redwood forest and the impact of the 1964 floods upon red- wood vegetation. Final Report. NSF Grant GB No. 3468.91 p. (on file at Humboldt State University li- brary, Arcata, CA). Comes, W. E. 1984. Stand structure and composition of the Little Lost Man Creek Research Natural Area, Redwood National Park. M.S. thesis, Humboldt State University, Arcata, CA. DAUBENMIRE, R. W. 1975. The community status of the coastal redwood, Sequoia sempervirens: A report pre- pared for the National Park Service, Arcata, CA. DELappP, J. A., B. E SMITH, AND W. R. POWELL. 1978. Soil- vegetation map and legend-N% of Klamath quadran- gle (9D-1,2). California Cooperative Soil-Vegetation Survey. U.S.D.A. Forest Service, Pacific Southwest Forest and Range Experiment Station, Berkeley, CA. DyrRngess, C. T., J. E FRANKLIN, AND C. MASER. 1972. Wheeler Creek Research Natural Area, federal re- search natural areas in Oregon and Washington: A guidebook for scientists and educators. U.S.D.A. For- est Service, Pacific Northwest Forest and Range Ex- periment Station, Portland, OR. Supplement No. 1. Eyre, FE H. (ed). 1980. Forest cover types of the United States and Canada. Society of American Foresters, Washington D.C. Fox, LAWRENCE III. 1989. A classification, map and vol- ume estimate for the coast redwood forest in Califor- nia. Final report. Interagency agreement number 8CA52849. The Forest and Rangeland Resources As- sessment Program. California Department of Forestry and Fire Protection, Sacramento, CA. FRANKLIN, J. E AND C. T. DyRNEss. 1973. Natural vege- tation of Oregon and Washington. U.S.D.A. Forest Service General Technical Report PNW-8. Pacific Northwest Forest and Range Experiment Station, Portland, OR. Gaucu, H. G. 1982. Multivariate analysis in community ecology. Cambridge University Press, New York. Har.Low, W. M., E. S. HARRAR, J. W. HARDIN, AND FE M. [Vol. 47 WHITE. 1996. Textbook of dendrology. McGraw-Hill, Inc., New York. HELMs, J. A. (ed). 1998. The dictionary of forestry. So- ciety of American Foresters, Washington D.C. Hipss, D. E., D. S. DEBELL, AND R. F TARRANT. 1994. The biology and management of red alder. Oregon State University Press, Corvallis, OR. HILL, M. O. 1979. TWINSPAN: A FORTRAN program for arranging multivariate data in an ordered two-way table by classification of the individuals and attri- butes. Section of Ecology and Systematics, Cornell University. Ithaca, NY. HINTZE, J. L. 1998. NCSS 2000. NCSS, Kaysville, UT. LENIHAN, J. M. 1986. The forest associations of the Little Lost Man Creek Research Natural Area, Redwood National Park, CA. M.S. thesis, Humboldt State Uni- versity, Arcata, CA. Manuony, T. M. 1999. Old-growth forest associations in the northern range of redwood. M.S. thesis, Humboldt State University, Arcata, CA. MATTHEWS, S. C. 1986. Old-growth forest associations of the Bull Creek watershed Humboldt Redwoods State Park, California. M.S. thesis, Humboldt State Uni- versity, Arcata, CA. MUELLER-DoMBOIS, D. AND H. ELLENBERG. 1974. Aims and methods of vegetation ecology. John Wiley and Sons, New York. Roy, D. FE 1966. Silvicultural characteristics of redwood. U.S.D.A. Forest Service research paper PSW-28. Pa- cific Southwest Forest and Range Experiment Station. Berkeley, CA. SAWYER, J. O. AND T. KEELER-WOLF. 1995. A manual of California vegetation. California Native Plant Soci- ety, Sacramento, CA. SAWYER, J. O. AND D. A. THORNBURGH. 1974. Subalpine and montane forests on granodiorite in the central Klamath mountains of California. Report to Pacific Southwest Forest and Range Experiment Station and Humboldt State University, Arcata, CA. (on file at Humboldt State University library, Arcata, CA). SAWYER, J. O., D. A. THORNBURGH, AND J. R. GRIFFIN. 1977. Mixed evergreen forest. Pp. 359-381 in M. G. Barbour and J. Major (eds.), Terrestrial vegetation of California. John Wiley and Sons, New York. SMITH, B. FE, J. A. DELAPP, AND W. R. POWELL. 1977. Soil- vegetation map and legend-S% Crescent City quad- rangle (9A-3,4). California Cooperative Soil-Vegeta- tion Survey. U.S.D.A. Forest Service, Pacific South- west Forest and Range Experiment Station, Berkeley, CA. STUART, J. D., T. WORLEY, AND A. C. BUELL. 1996. Plant associations of Castle Crags State Park, Shasta Coun- ty, CA. Madrofio 43:273-291. VEIRS, S. D., JR. 1979. The role of fire in northern coast redwood forest dynamics. Proceedings from the Sec- ond Conference on Scientific Research in the Nation- al Parks, San Francisco, CA. WESTERN U.S. CLIMATE HISTORICAL SUMMARIES. 2000. Crescent City 1 N, California (042147), July 1948— April 2000. in Program for climate, ecosystem, and fire application. Desert Research Institute and USDI Bureau of Land Management, Reno, Nevada. Ac- cessed July 17, 2000. Available: www.wrcc.dri.edu. ZINKE, P. J. 1977. The redwood forest and associated north coast forests. Pp. 679-698 in M.G. Barbour and J. Major (eds.), Terrestrial vegetation of California. John Wiley and Sons, New York. MADRONO, Vol. 47, No. 1, pp. 61—67, 2000 ESTIMATED AGES OF SOME LARGE GIANT SEQUOIAS: GENERAL SHERMAN KEEPS GETTING YOUNGER NATHAN L. STEPHENSON U.S. Geological Survey, Western Ecological Research Center, Sequoia and Kings Canyon Field Station, Three Rivers, CA 93271-9651 ABSTRACT Using a method that combines information on tree size with growth rates determined from relatively short increment cores, I estimated the ages of several of the largest living Sequoiadendron giganteum (Lindley) Buchholz. Compared to the longest-lived S. giganteum known, which was at least 3266 years old, most of the large sequoias analyzed here were relatively young, with estimated ages of only 1650 to 2150 years. Thus, contrary to common supposition, the largest S. giganteum generally owe their great size to rapid growth, not to exceptional age. However, two of the largest S. giganteum were substantially older, with estimated ages of 2850 and 2890 years. There is a high probability that some S. giganteum living today are older than the oldest S. giganteum yet discovered. People have long been fascinated by the great size and longevity of Sequoiadendron giganteum (Lindley) Buchholz (giant sequoias), which grow naturally only in isolated groves on the western slope of California’s Sierra Nevada. Sequoiaden- dron giganteum are the world’s largest trees, reach- ing a maximum known bole volume of nearly 1500 m? (Hartesveldt et al. 1975; Flint 1987 and in press). Precise cross-dating of tree rings on cut stumps has shown that sequoias can reach at least 3266 years in age (R. Touchan personal commu- nication), making S. giganteum the third longest- lived, non-clonal tree species known, exceeded only by Pinus longaeva Bailey (bristlecone pine, 4844 years) of western North America’s Great Ba- sin (Currey 1965) and Fitzroya cupressoides (Mo- lina) Johnston, (alerce, 3613 years) of Chile and Argentina (Lara and Villalba 1993). Here I present age estimates for some large, well- known S. giganteum, thereby addressing one of the most frequently-asked questions about famous S. giganteum—namely, “how old is this tree?’ I ad- ditionally address two questions regarding S. gi- ganteum sizes and ages. First, are the largest S. gi- ganteum so massive because they are exceptionally old, as is often presumed, or because they have grown particularly rapidly? Second, are there likely to be any S. giganteum alive today that are older than the longest-lived S. giganteum yet known, which is known only from a cut stump? These questions are difficult to answer because the only way to precisely determine the age of liv- ing S. giganteum is to crossdate tree rings on in- crement cores that intersect the tree’s pith (Stokes and Smiley 1968). However, the tremendous girth of large S. giganteum usually makes it impossible to reach their piths with hand-driven increment bor- ers. Power increment borers with very long bits can sometimes be used to obtain cores that reach the pith (Echols 1969; Johansen 1987), but have sev- eral disadvantages, which include unacceptably large holes left in the trees, poor quality of many of the cores extracted, and unacceptable use of noisy power tools on and around popular and fre- quently-visited S. giganteum. I therefore estimated the ages of several large S. giganteum using a method that takes advantage of information from partial increment cores (cores that fall well short of a tree’s pith). The derivation and testing of the method is described in detail else- where (Stephenson and Demetry 1995). Unlike pre- vious attempts to estimate the ages of large S. gi- ganteum (e.g., Douglass 1946; Hartesveldt et al. 1975), this method has been tested on hundreds of S. giganteum stumps, does not systematically over- or underestimate tree ages, and offers confidence intervals on the final age estimates. METHODS Choice of individual Sequoiadendron giganteum for analysis. The primary criteria for choosing in- dividual S. giganteum for analysis were (1) the S. giganteum were among the largest known, and (2) the cores and other data needed for age estimation were already available (that is, no S. giganteum was to be cored solely for the purpose of this study). Specifically, for a given S. giganteum to be includ- ed, original increment cores or the necessary mea- surements from those cores had to be available, along with measurements of the tree’s bark thick- ness and diameter at the height at which the cores were taken. These data requirements limited the pool of S. giganteum available for analysis. While many large S. giganteum have been cored for stud- ies of human impacts (Hartesveldt 1962, 1965), ring-width chronology development (Brown et al. 1992; Hughes et al. 1996), climatic reconstructions (Hughes and Brown 1992), forest dynamics studies 62 MADRONO TABLE 1. PRESS AND PERSONAL COMMUNICATION). Bole Size rank volume Tree name (by volume) (m7?) General Sherman 1 1487 Washington 2 1355 General Grant 3 1320 Boole 7 1202 Grizzly Giant 27 963 Cleveland 36 887 Sentinel Not ranked 790 [Vol. 47 SEQUOIODENDRON GIGANTEUM SELECTED FOR ANALYSIS (SIZE RANKS AND BOLE VOLUMES ARE FROM FLINT IN Location Giant Forest, Sequoia National Park Giant Forest, Sequoia National Park General Grant Grove, Kings Canyon National Park Converse Basin Grove, Giant Sequoia National Monument Mariposa Grove, Yosemite National Park Giant Forest, Sequoia National Park Giant Forest, Sequoia National Park NOTE: Future discoveries of previously unrecognized large sequoias will probably change the ranking of sequoias smaller than the Boole tree. For example, the fourteenth largest sequoia known (the Ishi Giant of Kennedy Grove) was identified only in 1993 (Willard 1994; Flint personal communication). (Stephenson 1994), and fire history reconstruction (Swetnam 1993), only a limited subset of those S. giganteum have associated records of diameter at core height. Diameter at core height is essential for age estimation (Stephenson and Demetry 1995), and cannot be estimated readily from published di- ameters at breast height of individual S. giganteum. Cores are rarely taken exactly at breast height, and sequoia bole diameter usually changes rapidly with increasing distance from breast height. The following seven large S. giganteum were se- lected for analysis (Table 1). The General Sherman, Washington, and General Grant trees are the world’s three largest trees, with the General Sher- man and General Grant trees being among the most heavily visited of all S. giganteum. The Boole tree is the seventh largest, and is well-known as being the largest sequoia on lands managed by the U’S. Forest Service. The Grizzly Giant is heavily visited because of its craggy appearance and status as one of the two largest S. giganteum in Yosemite Na- tional Park, whereas the Cleveland tree is a lesser- known and seldom-visited tree in Sequoia National Park. Finally, the Sentinel tree is a well-known se- quoia beside the road at the southern entrance to Giant Forest in Sequoia National Park. The General Sherman, Washington, General Grant, Grizzly Giant, and Cleveland trees all were cored by R. J. Hartesveldt and his colleagues for various studies during the late 1950’s and early 1960’s. All cores and data sheets for these trees are archived at Sequoia National Park, except I was unable to locate the original core for the Washing- ton tree, and therefore relied exclusively on Har- tesveldt’s ring measurements for that tree. The Boole tree was cored by researchers from the Uni- versity of Arizona in 1992; those data were kindly supplied by L. S. Mutch. Finally, the Sentinel tree was cored by V. G. Pile and me in 1998 at the request of National Park Service staff, who wished to have an age estimate for displays near the tree. Estimating tree ages. I estimated ages of these seven S. giganteum following Stephenson and De- metry’s (1995) approach, which combines knowl- edge of tree size with information gained from par- tial increment cores. The derivation and biological basis of this approach are too lengthy to repeat here; interested readers are therefore referred to Stephenson and Demetry (1995). When tested on 231 sequoia stumps up to 3200 years old and 6.5 m in diameter, this approach gave age estimates that were within 10% of actual age 62% of the time, and within 25% of actual age 98% of the time, assuming that two 60-cm increment cores are avail- able for analysis; fewer or shorter cores gave less precise estimates. This level of precision is a sub- stantial improvement over that of previously pub- lished methods, which estimated tree age from di- ameter alone, by assuming that basal area incre- ment is constant through time, or by linear extrap- olation of growth rates from the innermost portion of an increment core (Stephenson and Demetry 1995). Sequoia age in years, a, was estimated according to the following equation, (2100) ee [1] — Cc ss —_—_—_—_—e_ i elas Aaa) where c is the full ring count of a partial increment core; g is the length of the innermost 100 rings of the increment core; r is the length g plus the length of the section of bole radius (extending to the tree’s pith) that was not sampled by the increment core; and d is given by the following equation: d = 0.230 + 0.759(100/g,4m) + 1.27r — 0.8487? + 0.1597° [2] Units for g and r are meters, whereas g,,,, iS the length of the innermost 100 rings of the increment core in mm. For reasons discussed in Stephenson and Demetry (1995), if r exceeded 3 m, r = 3 m was substituted into eq. 2 for calculating d. A sequoia’s pith usually is not at the geometric center of its bole. However, we typically have no way of determining the location of a living tree’s pith, and therefore cannot directly measure the val- | 2000] STEPHENSON: SEQUOIA AGES 63 TABLE 2. CONFIDENCE INTERVALS FOR §. GIGANTEUM AGE ESTIMATES BASED ON DIFFERENT NUMBERS AND LENGTHS OF INCREMENT CORES (FROM STEPHENSON AND DEMETRY 1995). Two 60-cm One 60-cm Two 30-cm One 30-cm cores core cores core 50% confidence interval —6.9 to 9.0 —8.4 to 9.4 —14.1 to 11.1 —13.0 to 11.8 95% confidence interval —23.7 to 19.5 —36.7 to 19.7 —45.8 to 26.4 —48:2 to 27.5 NOTE: The intervals are expressed as percentage of estimated sequoia age. For example, the —23.7% listed as one endpoint of the 95% confidence interval for two 60-cm cores means that 2.5% of the time, actual tree age will be more than 1.237 times estimated tree age. (Rephrased, 2.5% of the time estimated sequoia age will be at least 23.7% less, expressed in terms of estimated sequoia age, than actual sequoia age.) The 19.5% listed as the other endpoint of the interval means that 2.5% of the time, actual tree age will be less than 0.805 times estimated tree age. ue of r associated with a particular increment core. Therefore r was estimated as described by Ste- phenson and Demetry (1995). First, tree radius was calculated as half of tree diameter (determined by diameter tape) at the height at which the increment core was taken. Average bark thickness, determined by probes at several location around the bole, was then subtracted to determine tree radius inside the bark. From this, the length of the increment core, excluding the core’s innermost 100 rings, was sub- tracted, yielding an estimate of 7. Because increment cores shrink as they dry, the wet length of a core must be known for the most accurate application of eqs 1 and 2. However, for most of the S. giganteum analyzed here (the Sen- tinel tree being the one exception), wet lengths of cores were not recorded. My colleagues and I (un- published data) have found that the average shrink- age of hundreds of sequoia cores was about 2%. Thus, when the wet length of a core was not re- corded, it was estimated by multiplying the core’s dry length by 1.02. To improve accuracy, when several cores were available from a sequoia, a given core’s location on the bole had to be separated from that of the other cores by at least 90° of circumference to be includ- ed in the age estimation (Stephenson and Demetry 1995). Tree age at height cored was estimated by averaging the age estimates based on the individual cores (Stephenson and Demetry 1995). Some of the data used to estimate sequoia ages came from S. giganteum cored several decades ago. It was therefore necessary to account for the num- ber of years that have passed since a sequoia was cored. Because, for convenience, I wished to esti- mate all sequoia ages relative to the year 2000, I subtracted the year in which a core was taken from 2000, then added the result to estimated tree age. The method outlined above only estimates se- quoia age at the height at which the cores were taken. However, accounting for the time it took a tree to grow to the height cored potentially can add decades to the tree’s estimated age. To account for height growth, I multiplied the height of the core above ground level (in m) by 178x~°.’, where x is the (estimated) cumulative width, in mm, of the 10 rings that abut the tree’s pith. This empirical factor scales height growth to radial growth, and was de- rived from ring measurements of 41 smaller S. gi- ganteum which my colleagues and I cored to the pith both near ground level and near breast height (see Agee et al. 1986 for a similar approach). How- ever, because there is no way of knowing the actual cumulative width of the 10 rings that abut the pith of the large S. giganteum analyzed here, I assumed that the width was 27.5 mm, based on the average from measurements of more than 450 sequoia stumps (Table A in Huntington 1914). Thus, I as- sumed that large S. giganteum took 178 X (27.5) °°? = 7.5 years to grow each meter taller until core height was reached. However, with the exception of the Sentinel and Grizzly Giant trees, core heights were not recorded. I therefore esti- mated core heights for the other trees based on con- versations and correspondence with individuals in- volved in the corings (H. S. Shellhammer for the General Sherman, Washington, General Grant, and Cleveland trees, and R. Adams and L. Mutch for the Boole tree). Confidence intervals. Stephenson and Demetry (1995) showed that as both the number and length of increment cores increase, confidence in sequoia age estimates also increases (Table 2). However, the numbers and lengths of cores used did not always fall neatly into the categories in Table 2. To deter- mine confidence intervals, core lengths were there- fore rounded to the nearest category shown in Table 2 (either 30 or 60 cm). In two cases (the General Sherman and General Grant trees), three cores rath- er than two were used. However, since confidence is improved relatively little by increasing core num- ber (it is improved more by increasing core length; Table 2), confidence intervals for only two cores were used. The number of years elapsed between the year in which a tree was cored and the year 2000 was then added to the endpoints of the tree’s confidence intervals, as was the estimated number of years it took each sequoia to grow to the height at which it was cored. Admittedly, the latter step does not change a sequoia’s age confidence intervals to re- flect the uncertainty associated with estimating the number of years it took a sequoia to grow to the height cored. However, uncertainty added at this 64 MADRONO [Vol. 47 TABLE 3. DATA USED TO ESTIMATE THE AGES OF THE SELECTED S. GIGANTEUM. * Side of tree from which core was taken. > Confidence intervals (see Table 2) are: 1 X 30, one 30-cm increment core; 2 X 30, two 30-cm cores; 1 X 60, one 60-cm core; 2 X 60, two 60-cm cores. ‘ Estimated from length of innermost 154 rings. 4 Estimated from length of innermost 280 rings. Wet Wet length of Height Diameter Bark length innermost Ring of core at core thick- of full 100 rings count above Confidence height ness core of core of full ground Year interval Tree Core? (m) (m) (m) (m) core (m) cored used? General Sherman South G32) 0.127 0.387 0.148 S17 1.6 1964 2, X730 Northwest T3295 0,127 0.365 0.156 249 1.6 1964 East T3253 0.127 0.3922"".0 20 315 1.6 1964 Washington — 7.858 0.152 0.291 0.091 525 1.4 1963 1 xX 30 General Grant Southeast 6.705 0.203 O375 0.259 146 2.0 1964. 2 xX 30 West 6.705 0.203 0.376 0.180 233 2.0 1964 North 6.705 0.203 0.378 0.139 293 2.0 1964 Boole B (Northwest?) 7.45 0.090 0.418 0.124¢° 209 1.4 1992 1 xX 60 C (Northeast?) 7.45 0.090 0.639 0.1594 386 1.4 1992 Grizzly Giant Southwest 6.621 O.127, 0.289 0.175 206 3.05 1958 2, %230 East 6.621 O:r27 0.266 0.148 175 3.05 1958 Cleveland — a 5.613 0.127 0.347 0.045 598 1.6 1964 1 x 30 Sentinel Northeast 6.399 0.073 0.515 0.099 366 2.56 1998 2 X 60 Southwest 6.399 0.073 0.556 0.128 333 2.04 1998 stage is small compared to the uncertainty of esti- mating the tree’s age at core height. Statistics on the longest-lived sequoia known. As a yardstick for interpreting results, I used the age and size of the longest-lived sequoia known—a cut stump in Converse Basin, Giant Sequoia National Monument, designated CBR26 by its discoverers (R. Touchan and E. Wright of the University of Ar- izona’s Laboratory of Tree-Ring Research). To- General Sherman Washington General Grant Boole Grizzly Giant Cleveland Sentinel 1000 1500 2000 2500 3000 3500 4000 4500 Age (years) Fic. 1. Estimated ages of selected S. giganteum in the year 2000, with associated confidence intervals. The ver- tical line within each horizontal box indicates that tree’s estimated age. The ends of each box delimit the 50% con- fidence interval for that tree’s age, whereas the ‘‘whisk- ers’’ extending from each box delimit the 95% confidence interval. The dotted vertical line at 3266 years indicates the age of the oldest sequoia yet discovered (see the text). Because the innermost ring of a long core taken within a fire scar cavity at the base of the Boole tree has been crossdated to A.D. 143 by E. Wright of the University of Arizona (L. Mutch personal communication), the Boole tree is at least 1858 years old, as indicated by the asterisk. uchan has precisely crossdated 3207 rings on the stump. It is missing much of its sapwood, so the outermost ring dates to 1834. However, the exten- sive logging of Converse Basin Grove occurred be- tween 1893 and 1908 (Johnston 1983; Willard 1994). Thus, at least 59 years of sapwood are miss- ing, and the tree therefore was at least 3266 years old when it was cut. (It is unlikely that the tree exceeded 3290 years old, including the time it took the tree to grow to the height sampled by Touchan and Wright.) The stump is relatively small: 5.8 m in diameter near ground level and 4.3 m in diameter at the cut surface 2.2 m above ground level (R. Touchan personal communication). Even with sap- wood and bark intact, the tree’s diameter at 2.2 m above ground level was probably less than 5 m when it was cut, much smaller than any of the trees analyzed here (Table 3). While we will never know the volume of the living CBR26, it is clear that many hundreds of S. giganteum alive today (prob- ably well over one thousand) are larger than CBR26 was before it was cut (e.g., see Appendix 1 in Stohlgren 1991). RESULTS Table 3 presents the data used to estimate the ages of the seven large S. giganteum. Estimated ages ranged from 1650 years for the General Grant tree to 2890 years for the Cleveland tree (Fig. 1), averaging 2230 years. Though all of these S. gi- ganteum were much larger than CBR26, the lon- gest-lived sequoia known, five had estimated ages at least 1000 years younger than CBR26 (Fig. 1). In fact, the third-largest living sequoia (the General Grant tree) is estimated to be little more than half 2000] -as old as CBR26. Additionally, CBR26’s age lies well outside of the high end of the 95% confidence intervals of the five S. giganteum (Fig. 1). _ While there are exceptions (namely, the Wash- ington and Cleveland trees), the largest living S. _giganteum generally owe their great bulk to rapid growth, not to extraordinary age. For example, av- erage ring width from the cores of the (estimated) youngest sequoia (the General Grant tree, 1.82 mm) was more than three times that of the (estimated) oldest sequoia (the Cleveland tree, 0.58 mm). This notion is further supported by Huntington’s (1914) age data from more than 450 sequoia stumps (the accuracy of which is discussed in Stephenson and Demetry 1995). Huntington’s ten largest stumps av- eraged 6.0 m in diameter inside the bark, but only 1842 years old by direct ring count (the largest was 6.5 m in diameter but only 1347 years old). In sharp contrast, his ten oldest stumps averaged only 4.9 m in diameter inside the bark, but 2822 years old—1 m less in diameter but nearly 1000 years older. Membership in the two groups of stumps was al- most mutually exclusive; only one stump was both one of the ten largest and one of the ten oldest (see Fig. 1 in Stephenson and Demetry 1995). Thus, for whatever reason, S. giganteum that reach great age tend to have grown relatively slowly. Figure 1 indicates that there is a 25% probability that the Cleveland tree is older than CBR26, and a similar probability that the Washington tree is older. The probability that at least one of these two living trees (Cleveland or Washington) is older than CBR26 therefore is roughly 1 — (0.75)*, or 44%— nearly even odds. Given that the seven S. gigan- teum examined here are only a small sample of all potentially old, living S. giganteum (likely candi- dates would number well over one thousand), it seems highly likely that some S. giganteum living today exceed the age of CBR26. DISCUSSION There has been a long-standing belief that the largest S. giganteum are the oldest. This is well illustrated by tracing the history of age estimates for the General Sherman tree, the world’s largest tree. By the early 20" century, careful ring counts and crossdating had identified a handful of sequoia stumps more than 3000 years old, the oldest being about 3200 years old (Huntington 1914; Douglass 1919, 1945). (John Muir’s reported count of 4000 rings on the ‘‘Muir Snag” in 1875 has not been repeated and was almost certainly in error [Flint 1987], and other early claims of up to 11,000 rings counted on stump tops [Jordan 1907] cannot be taken seriously.) Since none of these old stumps approached the great size of the General Sherman tree, most natural historians concluded that the General Sherman tree must be more than 3500 years old (e.g., Fry and White 1930). Stewart (1930) believed that the General Sherman tree was STEPHENSON: SEQUOIA AGES 65 about 4000 years old, though he reported that an estimate based on “‘average number of rings count- ed ...in charred fragments from parts of the [Gen- eral Sherman tree’s] burned trunk, in connection with the actual counts of rings of felled trees . which have grown under conditions and situation similar to those of the Sherman tree’’ yielded an age of 5200 years. Popular publications, such as a 1931 program for a play performed among the se- quoias not far from the General Sherman tree, tend- ed to be more extravagant, proclaiming the tree to be 6000 years old (see also Hartesveldt et al. 1975). Ironically, the aforementioned play took place less than two months before the first quantitative esti- mate of the General Sherman tree’s age based on increment cores, by A. E. Douglass. Douglass, the founder of the modern science of dendrochronology, obtained six short cores from the General Sherman tree in 1931 (the year is mis- takenly given as 1935 in Douglass [1946]). He deemed two of the cores to be good enough to use for age estimation, finding that average ring width at 4.6 m above ground level was 0.81 mm. This ring width is less than that of Hartesveldt’s cores (Table 3) because it comes from a height where the General Sherman tree’s bole is narrower. Douglass stated that “‘[t]hese are ring sizes which, in relation to the total size of the tree and the probable rate at which rings increase in size toward the center, sup- plied an estimate of the age of the tree of 3500 years plus or minus 500 years”’ (Douglass 1946). I have found no quantitative description of how Douglass accounted for “‘the probable rate at which rings increase in size toward the [tree’s] center.” To shed light on Douglass’ age estimate, I ap- plied the approach outlined in this paper to his data. Douglass’ data yield an age of only 2380 years for the General Sherman tree in 1931, or 2450 years in 2000 (rounded to the nearest decade). This latter estimate is only 300 years older than the estimate based on Hartesveldt’s cores (Fig. 1), and is well within that estimate’s 95% confidence interval. However, I judge the estimate based on Hartes- veldt’s cores to be much more reliable than that based on Douglass’ cores. Specifically, the estimate based on Hartesveldt’s cores required that fewer key parameters be estimated (such as the diameter of the General Sherman tree at 4.6 m above ground level in 1931, needed for using Douglass’ data), and was based on three cores widely spaced around the tree’s bole, each of which was nearly twice as long as the longest of Douglass’ two adjacent cores. In contrast, an age estimate based on linear ex- trapolation of Douglass’ ring-width data, assuming no change in ring width toward the General Sher- man tree’s center (an unrealistic assumption), would yield an age of 3790 years in 1931. Thus, Douglass’ estimate of 3500 (+500) years apparent- ly was little different from an estimate based on a simple linear extrapolation, and did not adequately consider the increase in ring widths toward the pith. 66 MADRONO Douglass’ age estimate was widely quoted (and sometimes exaggerated) from 1931 until the 1960’s, when Hartesveldt et al. (1975) radically revised the estimate downward. Unlike Douglass, Hartesveldt and his colleagues explicitly stated their assumption as to how ring widths change within a tree: they assumed that basal area increment is constant (that is, trees add a constant amount of basal area each year). This is equivalent to substituting d = 2 into eq. 1 (Stephenson and Demetry 1995). Hartes- veldt’s notes (archived at Sequoia National Park) show that when he strictly adhered to this assump- tion, he estimated that in 1964 the General Sherman tree was only about 1600 years old. However, Har- tesveldt’s examination of growth patterns on se- quoia stumps measured by Huntington (1914) in- dicated that strict adherence to this assumption sometimes underestimated the ages of S. giganteum (Hartesveldt et al. 1975). Thus, apparently based on a combination of assumed constant basal area in- crement and judicious comparisons with Hunting- ton’s data, Hartesveldt and his colleagues (1975) cautiously stated that the General Sherman tree **... 18 less than 2500 years old.’”’ According to my calculations using their original cores and data, their statement has a more than 75% probability of being true (Fig. 1). As careful as Hartesveldt et al. (1975) may have been in stating that the General Sherman tree was less than 2500 years old, the National Park Service, perhaps unable to bear such a precipitous decline in the tree’s age, instead adopted 2500 years as the midpoint for a range encompassing the tree’s esti- mated age. At the time of this writing, Park litera- ture and the plaque at the General Sherman tree stated that the tree’s estimated age was “2300-— 2700 years.”’ Additionally, a popular book authored by Hartesveldt’s colleagues (Harvey et al. 1981) dropped the qualifier ‘“‘less than,”’ stating instead that the tree “*‘.. . is about 2,500 years old’’ (though a table on the same page gives the General Sher- man tree’s age as “*2,500—3,000”’ years!). The most recent estimate of the General Sherman tree’s age— 2150 years (Fig. 1)—is most closely aligned with Hartesveldt et al.’s (1975) original statement that the tree is less than 2500 years old. The relative youth of other famous S. giganteum may come as a disappointment to some. For ex- ample, the decline in the estimated age of the Griz- zly Giant tree has been even more precipitous than that of the General Sherman tree. Clark (1910) re- ported that the Grizzly Giant had been growing so slowly over the last few centuries that its rings (pre- sumably observed inside of a fire scar cavity) were ‘“‘as thin as wrapping paper, too fine to be counted with the unaided eye.”’ (On the contrary, measured ring widths [Table 3] and measured tree volume changes [W. Flint personal communication] both in- dicate that the tree has been growing quite rapidly.) Comparing these purported ring widths with those of some fallen S. giganteum, Clark concluded that [Vol. 47 *‘the Grizzly Giant must be not less than six thou- sand years old,”’ and that the tree was probably the oldest living thing on earth. Other early age esti- mates placed the Grizzly Giant at a more modest 3800 years old, while Hartesveldt et al. (1975) later suggested that the tree “‘... is perhaps only 2500 years old.’ At the time of this writing, the National Park Service reported the age of the Grizzly Giant as 2700 years. However, I estimate the tree to be only about 1790 years old (Fig. 1), and that the probability of it being at least 2700 years old is less than 2%. Hartesveldt and his colleagues (1975) of- fered solace to those disappointed by the suggestion that certain large S. giganteum might be younger than expected: * this [discovery] effects a change only in superlatives; the world’s largest trees are the world’s fastest-growing trees.”’ Some readers may be disappointed by the broad confidence intervals associated with age estimates in Figure 1. There is a great deal of uncertainty in estimating the ages of individual large S. gigan- teum, largely due to relatively abrupt and sustained changes in ring widths in the part of the bole not sampled by increment cores, and therefore invisible to us (Stephenson and Demetry 1995). Such changes in growth rates are due to unpredictable, site-specific events in the past, such as occasional, localized high-intensity fires (e.g., Mutch and Swet- nam 1995). Thus, though Figure 1 suggests that the General Sherman and Sentinel trees are the same age (2150 years), the broad confidence intervals ad- ditionally suggest that this correspondence is most likely a meaningless coincidence. However, most of the confidence intervals in Figure 1 are based on relatively short cores. Confidence intervals could be tightened somewhat in the future by taking longer cores and, in the case of the Washington and Cleve- land trees, more cores. ACKNOWLEDGMENTS I thank R. Touchan of the University of Arizona’s Lab- oratory of Tree-Ring Research for graciously supplying data on CBR26, and L. S. Mutch for supplying data on the Boole tree. Additional thanks go to V. G. Pile (USGS), who helped me core the Sentinel tree, and D. J. McGraw (University of San Diego), who sent copies of Douglass’ original notes and correspondence on the General Sher- man tree, and who inspired me to complete this work. A. Caprio, M. Crapsey, W. Flint, J. Keeley, D. McGraw, H. Shellhammer, D. Shenk, R. Touchan, W. Tweed, and two reviewers provided helpful comments on the manuscript. LITERATURE CITED AGEE, J. K., M. FINNEY, AND R. DE GOUVENAIN. 1986. The fire history of Desolation Peak. Cooperative Parks Study Unit, University of Washington, Seattle. Final contract report to the National Park Service, Coop- erative Agreement CA-9000-3-0004. Brown, P. M., M. K. HUGHES, C. H. BAISAN, T. W. SWET- NAM, AND A. C. Caprio. 1992. Giant sequoia ring- width chronologies from the central Sierra Nevada, California. Tree-Ring Bulletin 52:1—14. 2000) | CLARK, G. 1910. The big trees of California. Yosemite Valley, CA. | CurrEY, D. R. 1965. An ancient bristlecone pine stand in eastern Nevada. Ecology 46:564—566. - Douctass, A. E. 1919. Climatic cycles and tree growth, | Vol. I. Carnegie Institute of Washington Publication No. 289, Washington D.C. . 1945. Survey of sequoia studies. Tree-Ring Bul- letin 11:26—32. . 1946. Sequoia survey—III: miscellaneous notes. Tree-Ring Bulletin 13:5-8. EcHOLs, R. M. 1969. Powered drive for large increment borers. Journal of Forestry 67:123—125. FLINT, W. D. 1987. To find the biggest tree. Sequoia Nat- ural History Association, Three Rivers, CA. . In press. To find the biggest tree, 2nd ed. Sequoia Natural History Association, Three Rivers, CA. Fry, W. AND J. R. WHITE. 1930. Big trees. Stanford Uni- versity Press, CA. HARTESVELDT, R. J. 1962. The effects of human impact upon Sequoia gigantea and its environment in the Mariposa Grove, Yosemite National Park, California. Ph.D. dissertation. University of Michigan, Ann Ar- bor. . 1965. An investigation of the effect of direct hu- man impact and of advanced plant succession on Se- quoia gigantea in Sequoia and Kings Canyon Na- tional Parks, California. Report on contract number 14-10-0434-1421, USDI National Park Service, San Francisco, CA. , H. T. Harvey, S. H. SHELLHAMMER, AND R. E. STECKER. 1975. The giant sequoia of the Sierra Ne- vada. U.S. Department of the Interior, National Park Service, Washington D.C. Harvey, H. T., H. S. SHELLHAMMER, R. E. STECKER, AND R. J. HARTESVELDT. 1981. Giant sequoias. Sequoia Natural History Association, Three Rivers, CA. HUGHES, M. K., R. TOUCHAN, AND P. M. Brown. 1996. A multimillennial network of giant sequoia chronolo- gies for dendroclimatology. Pp. 225-234 in J. S. Dean, D. M. Meko, and T. W. Swetnam (eds.), Tree rings, environment, and humanity. Proceedings of the International Conference. Radiocarbon, Department of Geosciences, The University of Arizona, Tucson, AZ. AND P. M. Brown. 1992. Drought frequency in STEPHENSON: SEQUOIA AGES 67 central California since 101 B.C. recorded in giant sequoia tree rings. Climate Dynamics 6:161—167. HUNTINGTON, E. 1914. The climatic factor as illustrated in arid America. Carnegie Institute of Washington Pub- lication No. 192, Washington D.C. JOHANSEN, R. W. 1987. Taking increment cores with pow- er tools. Southern Journal of Applied Forestry 11: 151-153. JOHNSTON, H. 1983. They felled the redwoods. Trans-An- glo Books, Glendale, CA. JORDAN, D. S. 1907. The alps of the King-Kern divide. A. M. Robertson, San Francisco, CA. LARA, A. AND R. VILLALBA. 1993. A 3620-year tempera- ture record from Fitzroya cupressoides tree rings in southern South America. Science 260:1104—1106. Mutcu, L. S. AND T. W. SWETNAM. 1995. Effects of fire severity and climate on ring-width growth of giant sequoia after burning. Pp. 241—246 in J. K. Brown, R. W. Mutch, C. W. Spoon, and R. H. Wakimoto, (technical coordinators), Proceedings: symposium on fire in wilderness and park management, 30 March— 1 April 1993, Missoula, Montana. USDA Forest Ser- vice General Technical Report INT-GTR-320. STEPHENSON, N. L. 1994. Long-term dynamics of giant sequoia populations: implications for managing a pi- oneer species. Pages 56—63 in P. S. Aune, technical coordinator. Proceedings of the Symposium on Giant Sequoias: their place in the ecosystem and society, 23-25 June 1992, Visalia, California. USDA Forest Service Gen. Tech. Rep. PSW-151. AND A. DEMETRY. 1995. Estimating ages of giant sequoias. Canadian Journal of Forest Research 25: 223-233. STEWART, G. W. 1930. Big trees of the Giant Forest. A. M. Robertson, San Francisco, CA. STOHLGREN, T. J. 1991. Size distributions and spatial pat- terns of giant sequoia (Sequoiadendron giganteum) in Sequoia and Kings Canyon National Parks, Califor- nia. Technical Report No. 43, Cooperative National Park Resources Study Unit, University of California, Davis, CA. USDI National Park Service. STOKES, M. A. AND T. L. SMILEY. 1968. An introduction to tree-ring dating. University of Chicago Press, Chi- cago. SWETNAM, T. W. 1993. Fire history and climate change in giant sequoia groves. Science, 262:885-889. WILLARD, D. 1994. Giant sequoia groves of the Sierra Ne- vada: a reference guide. Privately published by D. Willard, P. O. Box 7304, Berkeley, CA, 94707. MADRONO, Vol. 47, No. 1, pp. 68—69, 2000 REVIEW A natural history of the Sonoran Desert. Edited by S. J. Phillips and P- W. Comus. 2000. Arizona-So- nora Desert Museum Press, Tucson AZ, and Uni- versity of California Press, Berkeley CA. 628 pp. Cloth $55.00 ISBN 0-520-22029-3 Paper $24.00 ISBN 0-520-21980-5. The staff and associates of the Arizona-Sonora Desert Museum wrote this book as a compilation of their training courses, research, and personal ex- periences in the Sonoran Desert of Arizona, Cali- fornia, and Mexico. It evolved from the Museum’s docent handbook, developed over 30 years as a training document for volunteer interpreters. It pro- vides a summary of numerous biotic and abiotic patterns and processes, emphasizing adaptations of desert organisms and the interrelationships between nature and humans, both past and present. It an- swers in nontechnical prose typical questions of visitors to the Sonoran Desert. As such, this book covers a wide range of topics within its 628 pages. The book begins by briefly describing what a desert is, and how the Sonoran Desert differs from other deserts of North America. The regional sub- divisions and biomes are nicely summarized based on the original work of Forrest Shreve. The book then launches into two chapters describing a cal- endar of natural events and ten nature watching hot- spots. Although these two chapters are informative and will undoubtedly be of great use to those plan- ning trips to this region, they would make more sense if read after the other chapters and should have been placed at the end of the book, possibly as appendices. A chapter on desert storms gets the book gets back on track. This is a key chapter appropriately placed near the beginning of the text. It introduces rainfall as a significant factor influencing the evo- lution of desert organisms and the limitations to human settlement in the Sonoran Desert. Most oth- er chapters that follow presuppose some informa- tion contained in this chapter. The next chapter on desert air and light breaks with the main theme of the book by presenting ex- planations couched in basic physics. Phenomena such as mirages, atmospheric shimmer, and dust devils are described using simple descriptions of light refraction and the influence of temperature on the behavior of air. Although it is not directly re- lated to the other chapters, it presents an entertain- ing, effective, and succinct summary and does not significantly detract from the flow of the book. Consecutive chapters on deep history, geologic origins, soils, human ecology, and _ biodiversity frame the current Sonoran Desert in the perspective of evolutionary and recent time scales, and set the stage for the meat of the book which is the ecology of plants and animals. The deep history chapter de- scribes changes in flora and fauna over geologic time and discusses relationships between their past and current distributions. The geologic origins chapter describes the geomorphological develop- ment of major geologic features. The soils chapter describes the development and physical and biolog- ical properties of soils, and their implications for plants and animals. The human ecology chapter de- scribes the influence of humans over the past 12,000 years, ranging from native American hunt- ing, gathering, and farming, to Anglo-American ag- riculture, mining, and dam building. The chapter on biodiversity defines different indices and scales of biological diversity, describes natural centers of di- versity, and explains how humans have reduced di- versity by introducing invasive species and con- verting native plant communities to monoculture farms. Plants are discussed in separate treatments of plant ecology, flowering plants and grasses. The chapter on plant ecology covers a wide range of topics in its brief 23 pages. Topics include rudi- mentary descriptions of flower anatomy and pho- tosynthesis, and more detailed descriptions of drought adaptations, pollination, seed dispersal, and flowering seasons. The chapter on flowering plants includes a sampling of the most common and in- teresting angiosperms (other than grasses) in the Sonoran Desert. It is organized by taxonomic fam- ily, within which a few representative species are presented, covering their description, range, and comments on ethnobotany and natural history. The chapter on grasses includes a relatively detailed ac- count of different grassland types and dominant species, including a few dominant aliens, and some original natural history accounts. Animals are presented in separate sections cov- ering invertebrates, birds, mammals, fishes, reptiles, and amphibians. Adaptations to life in the desert by these groups are summarized. Species accounts of the common or otherwise interesting taxa include descriptions of distinguishing characteristics, habi- tat, range, life history, and in some cases feeding behavior. The glossary is very brief, but only includes terms which are not referenced in the extensive in- dex which lists a wide variety of items including common and scientific names, geographic places, and descriptive terms. This book provides a comprehensive introduc- tion to the natural history of the Sonoran and Mo- jave deserts, because many examples presented and - 2000) _ species discussed are common to these two deserts. _ It would be an ideal text for a community college or undergraduate course on desert ecology. Upper _ division and graduate students would not find much new information in this book. The strength of the _ book lies in the natural history descriptions for in- _ dividual species. This information is typically given REVIEW 69 very short treatment in the floras and field guides of the region, and all readers should find this in- formation interesting and useful. Matthew L. Brooks. United States Geological Survey, Western Ecological Research Center, Box Springs Field Station, 6221 Box Springs Blvd. Riv- erside CA 92507 California Botanical Society—Meeting Program 2000-2001 Academic Year All Meetings are held at 7:30 pm in room 2063 in the Valley Life Sciences Building on the UC Berkeley campus. September 21, 2000 Predicting the future of Sierran conifer forests: no lessons from the past. John Battles, Professor, University of California, Berkeley October 19, 2000 Diversity in California’s serpentine plants: the roles of patchiness, grazing, and burning. Susan Harrison, Professor, University of California, Davis November 16, 2000 Restoration of oak woodlands and grasslands in California: an evolutionary perspective. Kevin Rice, Professor, University of California, Davis January 18, 2001 Explosive beauty: rare plant research and management at Lawrence Livermore National Lab’s high explosive test facility, site 300. Tina Carlsen, Project Leader and Ecologist, Lawrence Livermore National Lab February 21, 2001 ANNUAL BANQUET and SEMI-ANNUAL GRADUATE STUDENT MEETINGS **ENOTE CHANGE OF LOCATION: California State University, Chico The role of geology in molding the California flora Arthur Kruckeberg, Professor Emeritus, University of Washington March 15, 2001 Molecular phylogenetic studies in Rosaceae. Dan Potter, Professor, University of California, Davis April 19, 2001 Defenders or pretenders? Interactions between an African acacia tree and four symbiotic ants. Maureen Stanton, Professor, University of California, Davis May 17, 2001 Using DNA fingerprinting to study Sequoia sempervirens populations in Big Basin Redwoods State Park. Chris Brinegar, Professor, San Jose State University Volume 47, Number 1, pages 1—70, published 8 December 2000. SUBSCRIPTIONS—-MEMBERSHIP Membership in the California Botanical Society is open to individuals ($27 per year; family $30 per year; emeritus $17 per year; students $17 per year for a maximum of 7 years). Late fees may be assessed. Members of the Society receive Maprono free. Institutional subscriptions to MADRONO are available ($60). Membership is based on a calendar year only. Life memberships are $540. Applications for membership (including dues), orders for sub- scriptions, and renewal payments should be sent to the Treasurer. Requests and rates for back issues, changes of address, and undelivered copies of MApDRONO should be sent to the Corresponding Secretary. INFORMATION FOR CONTRIBUTORS Manuscripts submitted for publication in Maprono should be sent to the editor. It is preferred that all authors be members of the California Botanical Society. Manuscripts by authors having outstanding page charges will not be sent for review. Manuscripts may be submitted in English or Spanish. English-language manuscripts dealing with taxa or topics of Latin America and Spanish-language manuscripts must have a Spanish RESUMEN and an English ABSTRACT. Manuscripts and review copies of illustrations must be submitted in triplicate for all articles and short items (NOTES, NOTEWORTHY COLLECTIONS, POINTS OF VIEW, etc.). Follow the format used in recent issues for the type of item submitted. Allow ample margins all around. Manuscripts MUST BE DOUBLE-SPACED THROUGHOUT. For articles this includes title (all caps, centered), author names (all caps, centered), addresses (caps and lower case, centered), abstract and resumen, 5 key words or phrases, text, acknowledgments, literature cited, tables (caption on same page), and figure captions (grouped as consecutive paragraphs on one page). Order parts in the sequence listed, ending with figures. Each page should have a running header that includes the name(s) of the author(s), a shortened title, and the page number. Do not use a separate cover page or ‘erasable’ paper. Avoid footnotes except to indicate address changes. Abbreviations should be used sparingly and only standard abbrevia- tions will be accepted. Table and figure captions should contain all information relevant to information presented. All measurements and elevations should be in metric units, except specimen citations, which may include English or metric measurements. Authors of accepted papers will be asked to submit an electronic version of the manuscript. Microsoft Word 6.0 or WordPerfect 6.0 for Windows is the preferred software. Line copy illustrations should be clean and legible, proportioned to the MADRONO page. Scales should be in- cluded in figures, as should explanation of symbols, including graph coordinates. Symbols smaller than 1 mm after reduction are not acceptable. Maps must include a scale and latitude and longitude or UTM references. In no case should original illustrations be sent prior to the acceptance of a manuscript. Illustrations should be sent flat. No illustrations larger than 27 X 43 cm will be accepted. Presentation of nomenclatural matter (accepted names, synonyms, typification) should follow the format used by Sivinski, Robert C., in Maprono 41(4), 1994. Institutional abbreviations in specimen citations should follow Holmgren, Keuken, and Schofield, Index Herbariorum, 8th ed. Names of authors of scientific names should be abbreviated according to Brummitt and Powell, Authors of Plant Names (1992) and, if not included in this index, spelled out in full. Titles of all periodicals, serials, and books should be given in full. Books should include the place and date of publication, publisher, and edition, if other than the first. All members of the California Botanical Society are allotted 5 free pages per volume in MAprowno. Joint authors may split the full page number. Beyond that number of pages a required editorial fee of $40 per page will be assessed. The purpose of this fee is not to pay directly for the costs of publishing any particular paper, but rather to allow the Society to continue publishing MaproNo on a reasonable schedule, with equity among all members for access to its pages. Printer’s fees for illustrations and typographically difficult material @ $35 per page (if their sum exceeds 30 percent of the paper) and for author’s changes after typesetting @ $4.50 per line will be charged to authors. At the time of submission, authors must provide information describing the extent to which data in the manu- script have been used in other papers that are published, in press, submitted, or soon to be submitted elsewhere. ———eEE—E>EEeEeEeEEEEEEE——E——EEEEE————~ C ‘Ss oe IEW SPECIES JTEWORTHY LLECTIONS 133 30T. VOLUME 47, NUMBER 2 APRIL-JUNE 2000 ‘MADRONO A WEST AMERICAN JOURNAL OF BOTANY LOCATIONS OF ENDANGERED SPRUCE POPULATIONS IN MEXICO AND THE DEMOGRAPHY OF PICEA CHIHUAHUANA F. Thomas Ledig, Manuel Mdpula-Larreta, Basilio Bermejo- Velazquez, Valentin Reyes-Herndndez, Celestino Flores-Lopez, and Miguel A. OCI O= PTL CO Caractere ee tee ate resareeste ee eee rceee ene oat rete ate eeee tne seen ereeete eee 7 REVISION OF CORETHROGYNE (COMPOSITAE: ASTEREAE) J. Phillip Saroyan, Dennis R. Parnell, and John L. Strother .............000+. 89 CORRESPONDENCE BETWEEN Ni TOLERANCE AND HyPERACCUMULATION IN STREPTANTHUS (BRASSICACEAE) Robert S. Boyd, Michael A. Wall, and James E. Watkins, Jr. ..........0000000+- wi ANNUAL VARIATION IN XYLEM WATER POTENTIAL IN CALIFORNIA OAKS Johannes M. H. Knops and Walter D. KOenig .......ccccccccccccseessssessesnsenseeees 106 MOLECULAR EVIDENCE FOR THE HYBRID ORIGIN OF OPUNTIA PROLIFERA (CACTACEAE) Michael S. Mayer, Laura M. Williams, and Jon P. Rebman ...........0000000+ 109 FLORAL VARIATION IN DELPHINIUM VARIEGATUM (RANUNCULACEAE) Shana C. Dodd and KGitis THETCNUIIA SE EEG TG tite Bia vsvesessecccvevescess 116 CROWN STRUCTURE OF THE WORLD’S SECOND LARGEST TREE Stephen C. Sillett, James C. Spickler, and Robert Van Pelt .............c00000+ 7) ERIOGONUM SPECTABILE (POLYGONACEAE): A NEW SPECIES FROM NORTHEASTERN CALIFORNIA Beth Lowe Corbin, James L. Reveal, and Robin Barron ..........0000.000000000 134 ARIZONA. ooscccccs e000 heG GA 00s ER SG es scan cn Med duce Sc obs vassvaesvawen saves 138 CALTBORNYTA 0.000.500 BZA E vccoscec MELODIES. Shs ctncn dons MAUR ce cvsvbecenertoaessoeees 138 COLORADO siisees deca sites. ee ER ee ee 142 OREGON? ssssuccersusie So issoccat eee a cos cag teens A ccc ees 144 PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL SOCIETY Maprono (ISSN 0024-9637) is published quarterly by the California Botanical Society, Inc., and is issued from the office of the Society, Herbaria, Life Sciences Building, University of California, Berkeley, CA 94720. Subscription information on inside back cover. Established 1916. Periodicals postage paid at Berkeley, CA, and additional mailing offices. Return requested. Postmaster: Send address changes to MApRONO, Roy Buck, % University Herbarium, University of California, Berkeley, CA 94720. Editor—KristTINA A. SCHIERENBECK California State University, Chico Department of Biology Chico, CA 95429-0515 kschierenbeck @csuchico.edu Editorial Assistant—Davw T. PARKS Book Editor—Jon E. KEELEY Noteworthy Collections Editors—DIETER WILKEN, MARGRIET WETHERWAX Board of Editors Class of: 2000—Pame-a S. Sottis, Washington State University, Pullman, WA JOHN CALLAWAY, University of San Francisco, San Francisco, CA 200 1—RoBert PATTERSON, San Francisco State University, San Francisco, CA PAuLA M. ScuHIFFMAN, California State University, Northridge, CA 2002—NorMAN ELLSTRAND, University of California, Riverside, CA Cara M. D’ Antonio, University of California, Berkeley, CA 2003—-FREDERICK ZECHMAN, California State University, Fresno, CA Jon E. KEELEY, U.S. Geological Service, Biological Resources Division, Three Rivers, CA 2004—Davip M. Woon, California State University, Chico, CA INGRID Parker, University of California, Santa Cruz, CA CALIFORNIA BOTANICAL SOCIETY, INC. OFFICERS FOR 2000—2001 President: BRucE BALDwin, Jepson Herbarium and Dept. of Integrative Biology, 1001 Valley Life Sciences Bldg. #2465, University of California, Berkeley, CA 94720. First Vice President: Rov Myatt, San José State University, Dept. of Biol. Sciences, One Washington Square, San José, CA 95192. rmyatt@email.sjsu.edu Second Vice President: Ros ScuuisinG, California State University, Chico, Dept. of Biol. Sciences, Chico, CA 95424. rschlising @csuchico.edu Recording Secretary: DEAN KELCcu, Jepson and University Herbarium, University of California, Berkeley, CA 94720. dkelch @sscl.berkeley.edu Corresponding Secretary: SUSAN BAINBRIDGE, Jepson Herbarium, University of California, Berkeley, CA 94720. suebain @SSCL.berkeley.edu Treasurer: Roy Buck, % University Herbarium, University of California, Berkeley, CA 94720. The Council of the California Botanical Society comprises the officers listed above plus the immediate Past President, R. JouN LittLe, Sycamore Environmental Consultants, 6355 Riverside Blvd., Suite C, Sacramento, CA 95831; the Editor of MApRONO; three elected Council Members: Bian Tan, Strybing Arboretum, Golden Gate Park, San Francisco, CA 94122; James SHEvock, USDI National Park Service, Pacific West Region, 600 Harrison Street, Suite 600, San Francisco, CA 94107; DIANE ELAM, U.S. Fish and Wildlife Service, 3310 El Camino Avenue, Sacramento, CA 95825; Graduate Student Representative: KirsTEN JOHANUS, Jepson Herbarium, University of California, Berkeley, CA 94720. This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). -MADRONO, Vol. 47, No. 2, pp. 71-88, 2000 LOCATIONS OF ENDANGERED SPRUCE POPULATIONS IN MEXICO AND THE DEMOGRAPHY OF PICEA CHIHUAHUANA FE THOMAS LEDIG* Institute of Forest Genetics, Pacific Southwest Research Station, USDA Forest Service, 2480 Carson Road, Placerville, CA 95667 USA MANUEL MAPULA-LARRETA, BASILIO BERMEJO-VELAZQUEZ AND VALENTIN REYES-HERNANDEZ Centro de Genética Forestal, Universidad Aut6noma Chapingo, Apartado Postal No. 37, Chapingo, México, C.P. 56230, México CELESTINO FLORES-LOPEZ AND MIGUEL A. CAPO-ARTEAGA Departamento Forestal, Universidad Aut6noma Agraria Antonio Narro, Buenavista, Saltillo, Coahuila, México ABSTRACT Picea A. Dietr. (spruce) is an essentially boreal genus, but three endemic taxa occur in México. Co- ordinates were determined for all known stands to accurately map their range and stimulate their protec- tion, conservation, and study. Thirty-nine stands of Picea chihuahuana Martinez (Chihuahua spruce) in the Sierra Madre Occidental were found in three clusters, each separated by 2 to 2.5° of latitude. The southernmost stands occur just south of the Tropic of Cancer. The entire north-south range is 687 km. Mean elevation of southern and central clusters was 2675 m, but stands in the northern cluster averaged over 350 m lower in elevation. Picea chihuahuana was associated with steep-sided arroyos or at the base of barrancas (cliffs or gorges). Picea martinezii T. FE Patterson (Martinez spruce) was found in six stands in the Sierra Madre Oriental, at an elevation of about 2250 to 2650 m and all within 147 km of each other. Picea mexicana Martinez (Mexican spruce) occurred on two of the highest ridges in the Sierra Madre Oriental, about 5 km apart and at an elevation of about 3500 m, and on the highest point (3185 m) in Chihuahua in the Sierra Madre Occidental, 676 km to the west. It is probable that P. mexicana will be found on one or two other high ridges in the Sierra Madre Oriental. . Every P. chihuahuana over 0.3 m in height was counted, measured, and scored for mistletoe infection, fire damage, and crown dieback from unknown cause(s) in 21 stands. Similar observations were made for another 18 stands by Narvaéz et al. (1983) about 15 years earlier. The combined count was 42,610 P. chihuahuana, which includes 24,221 trees and saplings over 2 m tall and 18,389 seedlings under 2 m but over 0:3 m in height. The distribution of diameter classes in our sample of 21 stands was a reverse- J, suggesting that the species is reproducing. However, the ratio of seedlings to saplings and trees was less than 1.0 in all except four of the 39 stands, indicating that the species may actually be in jeopardy. Based on ring counts from increment cores and stumps, P. chihuahuana can reach 272 years of age. This is a relatively short life span compared to other North American spruces. The largest trees were 51 m tall and 125 to 150 cm in diameter-breast-high, and size was about average compared to its congeners in the United States and Canada. Many trees were in poor condition or damaged from cutting, mistletoe, top dieback, and fire. Contrary to expectations, the southern stands were in no poorer condition than the northern, and in fact, the incidence of mistletoe was highest in the north; this seems to be the first report of mistletoe on P. chihuahuana. Trees in southern stands were larger and older, and the ratio of seedlings to saplings and trees was highest in the southern and central stands and lowest in the north. RESUMEN Los abetos (Picea A. Dietr.) son un género esencialmente boreal. A pesar de lo anterior, tres taxa endémicos ocurren en México. Las coordenadas geograficas fueron determinadas para todos los rodales conocidos con el proposito de ubicar de una manera precisa su area de distribuci6n en un mapa y promover su protecciOn, conservacion y estudio. Treinta y nueve rodales de Picea chihuahuana localizados en la Sierra Madre Occidental estuvieron agrupados en tres clusters, cada uno separado por 2 a 2.5° de latitud. Los rodales mas surefios ocurrieron justamente al sur del Trépico de Cancer. El area entera de distribuci6n de norte a sur fue de 687 km. El promedio de elevaci6én de los clusters del sur al centro fue de 2675 m, sin embargo, en los rodales pertenecientes a los clusters del norte, estos promediaron 350 m mas bajos en elevaciOn. Picea chihuahuana estuvo asociada con arroyos en pendientes pronunciadas y barrancas. * To whom correspondence should be addressed. MADRONO Picea martinezii fue encontrada en seis rodales en la Sierra Madre Oriental, a una elevaci6n de cerca de 2250 m a 2650 m y localizados todos a una separacion de 147 km uno del otro. Picea mexicana ocurrid en dos de los puntos mas altos en la Sierra Madre Oriental separados 5 km uno del otro a una elevacion cerca de 3500 m, y sobre uno de los puntos mas altos (3185 m) en Chihuahua en la Sierra Madre Occidental, 676 km al oeste. Es posible que Picea mexicana pudiera ser encontrada sobre uno u otros dos de los puntos mas elevados en la Sierra Madre Oriental. Cada una de las plantulas de Picea chihuahuana arriba de 0.3 m de altura fue contada, medida, y registrada por presencia de infestaci6n de muérdago, dano por incendio, y muerte por despunte de la copa debido a causas desconocidas en 21 rodales. Observaciones similares fueron hechas para otros 18 rodales por Narvaéz et al. (1983) con 15 anos de anterioridad. La contabilizaci6n combinada fue de 42 610 plantas de Picea chihuahuana, la cual incluy6 24 221 darboles adultos y jévenes fustales arriba de 2 m de alto y 18 389 plantulas abajo de 2 m. La distribuci6n de clases de didmetro en nuestra muestra de 21 rodales tuvo una distribuci6n de una jota invertida, sugiriendo que la especie se esta reproduciendo. Sin embargo, la proporci6n de plantulas a jOvenes fustales y arboles adultos fue menos de 1.0 en tados, excepto en cuatro de los 39 rodales, indicando que la especie podria realmente estar en peligro de extincion. Sobre la base del conted de los anillos de crecimiento en virutas de madera y tocones, Picea chihu- ahuana puede alcanzar hasta 272 anos de edad. Esto es un ciclo de vida corto comparado con otras piceas norteamericanas. Los arboles mas grandes tuvieron 51 m de alto y de 125 cm a 150 cm de didmetro a la altura del pecho y su tamafno estuvo cerca del promedio comparado con sus congéneres en los Estados Unidos y Canada. Muchos 4rboles presentaron una condici6n pobre o estuvieron dafados por corta, muérdago, despunte de copa por causas indeterminadas, y fuego. Los datos pueden ser utilizados para monitorar la especie para cambios futuros. La evidencia de fosiles indica que el arca de distribuci6dn de Picea chihuahuana estuvo confinada hacia el norte durante el Holoceno, y nosotros especulamos que los rodales localizados mas hacia el sur podrian estar en una condici6n mas pobre que aquellos localizados hacia el norte. Contrario a las expectaciones, los rodales localizados hacia el sur no estuvieron en una condiciOn muy pobre comparados con aquellos localizados hacia el norte, y de hecho, la incidencia de infestaci6n por muérdago fue mas alta en el norte. Los arboles en los rodales del sur fueron mas grandes y viejos, y la proporci6n de plantulas a jo6venes fustales y [Vol. 47 arboles adultos fue mas alta en los rodales del sur y del centro, y mas baja en el norte. The occurrence of Picea A. Dietr. (spruce) in the subtropical latitudes of México is surprising. Spruce is a largely boreal genus which, depending on the taxonomist, includes 31 to 50 species (Dal- limore and Jackson 1923; Wright 1955; Bobrov 1970; Everett 1981). Three species of spruce occur in cool, temperate, montane forests of México. In the Sierra Madre Occidental, spruce forest occurs less than 38 km from the Rio Urique at the bottom of the Barranca del Cobre (the Copper Canyon), where bananas and citrus are grown. Palynological evidence suggests that the spruces of México are relicts stranded by a warming climate during the current interglacial. They may serve the same func- tion as canaries in a coal mine; i.e., by monitoring Mexican spruce populations, we may have an early signal of climate change projected for the next cen- tury (Mahlman 1997). The disappearance of cool temperate conifer forest in México is projected un- der three climate change scenarios (Villers and Tre- jo 1997). The most common spruce in México is Picea chihuahuana Martinez (Chihuahua or prickly spruce), which occurs in the Sierra Madre Occiden- tal, México’s western cordillera (Fig. 1). Picea chi- huahuana was first reported in 1942 from a site called Talayotes in the State of Chihuahua (Marti- nez 1953). Locally, it is called cahuite, cahuite es- pinoso, cahuite bravo, and pinabete espinoso, or, by the native Tarahumara Indians of the Barranca del Cobre, matego or mategoco. Vegetation, soils, and climate at sites where it occurs have been described by Gordon (1968), Narvaéz, Sanchez, and Olivas (1983), and Narvaéz (1984). Picea mexicana Martinez (Mexican spruce) was discovered in 1961 (Martinez 1961) on Sierra La Marta at the border between the States of Nuevo Leon and Coahuila (Fig. 1). Cerro El Morro on Si- erra La Marta is the highest point (ca. 3700 m) in the Sierra Madre Oriental, the range that parallels the east coast of México. A year earlier, spruce had been reported on Cerro Mohinora (Correll 1960), the highest point in Chihuahua (ca. 3300 m), and these were referred to as P. “‘hybrida” or P. “‘in- determinada’’. This spruce was very different from P. chihuahuana but similar to P. engelmannii En- gelm. (Engelmann spruce) that reaches its southern limit in the Chiricahua Mountains of southern Ar- izona about 700 km distant (Little 1971). Using morphological data and multivariate techniques, Taylor and Patterson (1980) found that the trees from Cerro Mohinora clustered with P. mexicana. Subsequently, P. mexicana and P. “‘indeterminada”’ were treated as a variety of P. engelmannii by Tay- lor, Patterson, and Harrod (1994); i.e., P. engel- mannii Parry var. mexicana (Martinez) Silba. In 1983 two stands of spruce, which Miiller-Us- ing and Alanis (1984) identified as P. chihuahuana, were discovered in the Sierra Madre Oriental in the State of Nuevo Le6n (Fig. 1). However, the nearest populations of P. chihuahuana were at least 510 km distant in the Sierra Madre Occidental, separat- ed from the Sierra Madre Oriental by the arid Mes- eta Central of México. Subsequently, Patterson | 2000] New Mexico w Sinaloa VA “Durango . e @ s Coahuila © ZN \ % iG al. and some geographic features mentioned in the text. (1988) decided that these trees were sufficiently dif- ferent from P. chihuahuana to merit species status, and named them P. martinezii T. E Patterson (Mar- tinez spruce). Needles of P. martinezii are flat and flexible and its cone scales have denticulate mar- gins, but the needles of P. chihuahuana are four- sided and stiff and its cone scales are rounded. The cones and seeds of P. martinezii are larger than those of P. chihuahuana. The ecology of P. mar- tinezii and P. mexicana in the State of Nuevo Leén was described by Cap6é et al. (1997). The endemic spruces are a minor element in the flora of México, yet potentially important from the standpoint of science, their unique contribution to the biodiversity of México, and their value as ge- netic resources. Picea chihuahuana and P. mexi- cana were included on a list of endangered arboreal taxa prepared for the Instituto Nacional de Inves- tigaciOnes Forestales y Agropecuarias (INIFAP) by Vera (1990), and all three species qualify as threat- ened under the guidelines of the International LEDIG ET AL.: SPRUCE IN MEXICO 73 30° N Oa aN [Vn %,)) * a 25° N mi uevo Leon - %& Monterrey iu a O c} 2 ™ 20°N ISTHMUS OF .. TEHUANTEPEC 15° N 95° W Map of México showing the locations of Chihuahua spruce (@), Martinez spruce (MI), Mexican spruce (A), Union for the Conservation of Nature and Natural Resources (Sanchez 1984; Sanchez and Narvaéz 1990). Picea chihuahuana occupies sites with some of the richest arboreal species diversities in the Sierra Madre Occidental (Gordon 1968), or in all of tem- perate North America, and for that reason its habitat should be a crucial focus for protection. The Sierra Madre Occidental was nominated by the IUCN as a global center of plant diversity (Anonymous 1991). Spruce in México may have retreated northward several times, most recently during Holocene warming. Spruce occurred at least as far south as the Isthmus of Tehuantepec (18°09'N) in the mid- Pliocene, 5 million years ago (Graham 1993). Pol- len in the ancient bed of Lake Texcoco, under Méx- ico City, and in Lake Chalco in the basin of Méx- ico, show that spruce occurred in the surrounding uplands at the end of the Pleistocene (Clisby and Sears 1955) and at least as recently as 7000 to 8000 74 MADRONO yr before present (B.P.; Lozano-Garcia et al. 1993; M.d.S. Lozano-Garcia, personal communication 1997). The nearest P. chihuahuana are now about 700 km northwest of México City in the Sierra Ma- dre Occidental, and P. martinezii occurs about 500 km north in the Sierra Madre Oriental (Patterson 1988). All P. mexicana may have had ranges as far south as México City, but P. chihuahuana is most likely to have occurred there. The topography of México is more conducive to migration of high el- evation taxa between México City and the Sierra Madre Occidental than between México City and the Sierra Madre Oriental, where P. martinezii oc- curs. In addition, the high endemism of the subal- pine habitats in the Sierra Madre Oriental suggests that they were not linked during the Pleistocene with the Transverse Volcanic Belt in which México City lies (McDonald 1993). In any case, palynolog- ical and genetic studies indicate that the range of spruce retreated northward since the Pleistocene and all Mexican spruces are now characterized by small, fragmented populations (Ledig et al. 1997). The southernmost stands of P. chihuahuana, Ar- royo de la Pista and Arroyo del Chino, lie a few kilometers south of the Tropic of Cancer (23°30’N). Only P. morrisonicola Hayata (Morrison spruce) of Taiwan grows at such southerly latitudes (Wright 1955). No comprehensive description exists to docu- ment the location of the Mexican spruces. Previ- ously published information is incomplete, incon- sistent, and, in some cases, incorrect. Confusion ex- ists in the spelling of place names, which appear in several variants. Coordinates are sometimes wrong. Many important publications on spruce in México are agency reports or proceedings and not easily obtainable. A census provides data important to conservation and a baseline against which to monitor the effects of climate change. The increase in temperature at the end of the last glacial was about equal to that projected after a doubling of atmospheric carbon dioxide, which may occur in less than half a cen- tury (Mahlman 1997). Spruce was so sensitive to the post-glacial warming that it was reduced from a widespread taxon or group of taxa that occurred as far south as Lake Texcoco to isolated relicts in the highest mountain ranges (Ledig et al. 1997). An excellent start was made on censusing P. chi- huahuana by Narvaéz et al. (1983), and on P. mar- tinezii by Miiller-Using and Alanis (1984). In Chi- huahua, Narvaéz et al. (1983) counted every P. chi- huahuana then known, but a census had never been undertaken for Durango, where the southernmost stands occur, and where the continued existence of P. chihuahuana may be most in jeopardy because of climatic warming. Our objective was to accurately map the location of all stands of spruce in México, to make them more accessible for scientific study, and to identify sites worthy of conservation. We anticipate that [Vol. 47 more stands will be found in time, because bota- nists have not yet completely explored the rugged mountains of México. Nevertheless, the present in- ventory is a beginning, and will probably stimulate additional exploration. Our second goal was to complete the census of P. chihuahuana begun by Narvaéz et al. (1983) to provide a baseline against which to measure future change, and to characterize the health of the stands as an aid in making man- agement decisions. METHODS Each known location of spruce in México was visited in 1997 or 1998 and its coordinates (lati- tude, longitude, and elevation) were determined with a geographic positioning system (GPS, Trim- ble’s GeoExplorer). No base stations were available in México, so the data were not corrected for sat- ellite signal degrade introduced by the U.S. De- partment of Defense through selective availability. However, even with selective availability, the co- ordinates should not be off more than 100 m and are probably within 50 m of the actual center of the stand. Elevation may vary substantially within stands, and in most cases we took readings at the center and the extremes. Observations were also made of exposure, slope, and land use. Distance between stands was calculated with an online program (Kindred 1997) and used to group them into geographically coherent clusters using Statistica’s clustering module (StatSoft 1995a). Municipio and land ownership (usually an ejido) were determined in the field. Municipios are divi- sions of states, roughly equivalent in size and po- sition in the political hierarchy to counties in the United States. Ejidos occur within municipios and are lands given to groups of peasants after the 1910 revolution (Vargas 1996). Ejido lands are often held and used communally. However, parcels may be owned individually if members of the ejido decide that this is appropriate. By reference to GPS coordinates and topographic features, stand locations were plotted on 1:50,000 maps variously published by: the Comision de Es- tudios del Territorio Nacional (CETENAL), Secre- tarfa de la Presidencia; the Departamento Carto- grafico, Secretaria de la Defensa Nacional; the In- stituto Nacional de Estadistica, Geografia e Infor- matica (INEGI); or the Coordinacié6n General de los Servicios Nacionales de Estadistica, Geografia e In- formatica. Duplicate maps are in the files of the Institute of Forest Genetics, Placerville, CA, and the Centro de Genética Forestal, Chapingo, México. We counted every P. chihuahuana in the State of Durango greater than or equal to 0.3 m tall. Dur- ing the census, the spruces were marked with paint to ensure that none was counted twice and none overlooked. The stands in Durango were censused during April and May, 1997, except for Arroyo del Chino and Arroyo del Agua, which were only dis- | 2000] covered in 1997, and were censused in August and September 1998. In 1998, we also counted spruces -at La Luisiana, Arroyo de las Ranas, La “‘Y’’, Las _Lajas, and Llano Grande in the State of Chihuahua. In June 1999, we counted spruce at Arroyo de Que- ‘brada near El Vergel, Chihuahua. Narvaéz (1984; _Narvaéz et al. 1983) censused every stand in Chi- huahua known in 1983, but La Luisiana, Arroyo de las Ranas, La “*Y”’, and Las Lajas were not known until 1997, and he omitted Arroyo de Quebrada be- cause he believed that it was in Durango. Our cen- sus of Llano Grande unintentionally provided a comparison with Narvaéz’ (1984) counts. In each of the 21 stands censused, we measured height of all spruces <3 m but 20.3 m tall with a measuring pole, and estimated height of the rest. Diameter-breast-high (dbh at 1.4 m) of spruce over 2 m tall was measured with a graduated scale, called a Biltmore stick. In most stands, more ac- curate measurements were made on 1000 m? plots in the stand interior. The number of plots in a stand varied from three to 13 to provide a sample equal to approximately 10% of the number of trees =10 cm dbh. On the plots, we measured dbh to the near- est millimeter with a diameter tape and height to the nearest meter with a Haga altimeter. In a few stands, plotless samples were measured, and in stands with 15 or fewer trees, we measured every tree with diameter tape and altimeter. We aged one or two of the large trees in a stand using increment cores taken from standing trees. The very largest trees inevitably had a rotten heart, and it was not possible to age them. We supple- mented the increment cores by opportunistically counting rings on stumps. Although P. chihuahu- ana are nominally protected, stumps were observed in many stands. In all, 29 trees were cored, and ring counts were made on 40 stumps. Finally, we scored condition of the P. chihuahu- ana. We recorded the number with mistletoe, and visually scored the severity of infection on a scale from one to three: 1) one-third of the crown in- fected, 2) two-thirds of the crown infected, 3) all of the crown infected. Many spruce had dead tops of unknown cause, and we classified these into three categories: 1) one-third of the crown dead, 2) two-thirds of the crown dead, 3) nearly 100% of the crown dead. To summarize these observations, we constructed weighted scores for mistletoe and for crown dieback: the score was the number of Spruce in each class multiplied by the severity rat- ing (1 to 3), divided by the total number of spruce. We also presented the data as the percentage of Spruce infected with mistletoe or with crown die- back, regardless of severity. If the tree was fire- scarred, that was also recorded. In reporting the census data, we excluded dead Spruce. For regressions of log(height) on log(dbh) and in calculating the means of height and dbh for the interior sample, we excluded dead trees, burned LEDIG ET AL.: SPRUCE IN MEXICO 75 trees, trees with 100% of their crown dead, trees with missing tops, and trees of sprout origin. Statistical analyses, including regressions, anal- yses of variance, and post-hoc comparisons of means, were made with Statistica modules for cor- relation matrices, breakdown and one-way ANO- VA, and tables and banners (StatSoft 1995b). How- ever, most of the data presented here for P. chihu- ahuana represent the entire metapopulation, not samples, so significance testing is not actually nec- essary. All differences are real. Where we present the results of significance tests, it is only to suggest the level of confidence had these stands been a ran- dom sample. RESULTS Thirty-nine stands of P. chihuahuana were lo- cated, six of P. martinezii, and three of P. mexi- cana. We could not locate the stand of 15 P. chi- huahuana trees listed in Narvaéz (1984) and San- chez and Narvaéz (1990) as Rio Verde, Ejido Ca- tedral. Martinez (1953) noted P. chihuahuana at Arroyos de Urichique, Cuervo, and Meguachic, but we are not certain where he meant. The location of all stands of spruce in México that were known and confirmed as of June 1999 are presented in Tables 1—4 and Fig. 1. Synonyms for place names previ- ously published are listed in the footnotes to Tables 1-4. Picea chihuahuana occurs between 23°20'N (south of the Tropic of Cancer) and 28°39’N. The distance between the northernmost and the south- ernmost stand is 687 km. The stands are grouped in three major areas with large gaps between, al- though Arroyo de Chachamori and Faldeo de Ce- bollitas are sufficiently isolated that they might be considered their own unique “‘clusters’’ (Fig. 1). The largest, northern cluster consists of almost all the stands in Chihuahua immediately north of the Barranca del Cobre, from Arroyo de Chachamori south to Rio Vinihueachi. The southern cluster con- sists of four stands in Durango: Arroyo de la Pista, Arroyo del Chino, Arroyo de las Lagunas, and Ar- royo del Infierno. The central cluster is largely in Durango with the exception of one stand, Arroyo de Quebrada, which is in Chihuahua near its border with Durango. For the 21 stands in Durango and Chihuahua at which we recorded site conditions, all were rocky, 16 abundantly so. In every case, the slope aspect was north to northeast, in only two cases varying as much as 20° west of north. The slope averaged 65%, and was less than 50% only at La Estancia Agua-Amarilla and Arroyo del Infierno, where some of the trees grew on level ground. The spruce generally occurred near the bottoms of canyons or barrancas, sometimes extending nearly to the ridge. Grazing animals were present in all stands except Arroyo de la Pista, Piedra Rayada, Arroyo del Chi- no, and La ‘Y’. Stumps were observed in some [Vol. 47 MADRONO 76 S69TC OLSC 0097 068¢ SCLC OOrC OSLC SPIT O8Sc 066¢ OSVC SCLC SLLC 009 S89C (ur) , AVA ut L,8C.901 uL 1, VC9OI w&S,CCIOT uv, 1C.901 uce,V1901 wl S,€1.901 wl T,81.901 u00,C 1.901 wt 1,LCE.901 iV 181.901 uSt,9T90I u80,9C.SOI wT, LEVOT uSO,€V.VOI wWO00,SV.VOT NM 2 m). We tabulated our data into various height categories, and a division at 3 m in height resulted in excellent agreement between Narvaéz’ (1984) count and ours; his 545 saplings and trees versus our 528, and his 370 seedlings versus our 378. Actually, our census also counted 15 dead or burned trees, and if these are added to our tree class, the total is 543, almost the same as Narvaéz’ (1984) number of 545. The difference between counts suggests that there has been a slight reduc- [Vol. 47 MADRONO oe) i OSET O8ETc OCCC SLVC SITET S8CTC SCVC OLCC CSET OVET OLCC SOT SOT O9TC O&Le (uw) y AST uSO,9€.L01 u80,9V.L01 uC 1, SVLOT u81,9VLOT uc V,ICLOT wVC,LELOI w0V,LELOl u9e , LELOI u10,67.L01 u80,LE.LOl wWOT,LELOT wl 1,SVLOT uSC,1VLOI u8S, 1TVLOI wl, 17901 NM = Ss eS al ee ee ee cance by Scheffe’s test (P = 0.052). | [ee] ie.e) 00 (oe) (oe) 00 oO oo [ee] . CUTE GUS CN CNET: UN RN The tallest tree of P. chihuahuana we measured was 51 m in height at Arroyo de la Pista and the largest diameter was 150 cm at Arroyo de Quebra- da. Considering stems =10 cm dbh, mean height of sampled trees (i.e., trees in the interior sample plots) varied among stands from 9.8 m at La Lu- ro Fy ins Gh, Oo A, isiana to 25.9 m at Arroyo del Infierno and mean SlExn knew Sn kn ew fxn 20650] diameter varied from 22.5 cm at Piedra Rayada to 6 = 6 = 6 = 6 = 6 c 6 S 6 = 6 ee = 55.6 cm at Arroyo del Infierno (Table 5). In most ON ON ON oN OO ON oN ND Ss Srererer er er 4742742! cases these values were larger than the mean co i i heights and diameters for the entire stand, but are A emphasized here because they were taken with a greater precision (Haga altimeter and diameter-tape S fe versus ocular estimate and Biltmore stick). Mean 6 Bi/E& € EF & € € & E §& height and diameter tended to be test in th sO O 0) 0) 5) 5) 5) ) ) greatest in the v € F EF F&F &F § & § & & southern cluster (20.5 m and 42.3 cm), less in the ror] > 5 5 5 5 5 5 5 5 5 central cluster (16.2 m and 31.7 cm), and least in 5 the four stands that we measured in the northern z cluster (12.2 m and 31.7 cm). If these were sam- ples, the differences, between the southern and northern clusters would be statistically significant according to Scheffe’s test. Maximum height and z diameter (using data for the entire stand) also de- ce nes creased south to north: mean maximum heights > & & § &§ were 45.0, 37.3, and 24.5 m, and mean maximum a 8 2 A HD D diameters were 105.9, 98.8, and 59.8 cm, respec- Bln 4H £ = § tively, for the southern, central, and northern clus- es aor 4 53 3 4 ters. Again, if these were samples, the southern and = & & Pes = . = central clusters would differ significantly from the & 6 6 @5€ € € 8s 8 northern cluster by Scheffe’s test. csc - EZ ca sc € § 5 5 The logarithms of height and diameter were A «£ << Eee ee Ve x 0 closely related in all 21 stands in which we took 6> 6: GO “s 2o. -c “6 to: 6 ; 3 USD FRB ses 8 data (Fig. 4). Because the number of observations fara ee ee was so large, tests for homogeneity of regression coefficients (p. 319 in Steel and Torrie 1960) for height on diameter indicated that the slopes differed beth between and within clusters. Analysis of vari- zs ance also showed that the intercepts differed among y) < clusters, and Scheffe’s test indicated that the inter- Ss &§ cepts for the northern cluster were significantly dif- ate “x 8 ferent from those for the central and southern clus- a|%3 3 Peltier os ae O ters. Although the differences were small, the pat- a) & 4 s € &€ € 8 8 tern was consistent; the smallest intercepts (most a, S| SebOrne. So oS 9 negative) and largest slopes occurred in the north oa oc SOuc. Vac and the greatest intercepts and smallest slopes in Ga a ee ee ae Va ee the south (Table 7). For a given diameter, trees were () = TABLE 3. LOCATION OF MARTINEZ SPRUCE STANDS IN THE STATE OF NUEVO LEON, Mexico. ' A municipio is a political division of a state. 7 Number and key of the 1:50,000 national topographic maps of the Estados Unidos Mexicanos, published by the Comision de Estudios del Territorio Nacional (CETENAL), Secretaria de la Presidencia. * In degrees, minutes, and seconds. * Elevation recorded by GPS at the center of the stand and rounded to the nearest 5 meters. ° Called La Encantada by Patterson (1988). © Ejidos are lands given to groups of peasants after the 1910 revolution and usually held communally. ’ Called El Butano by Paterson (1988). Longitude? Elev.* Latitude? (m) 215 W 99°47'30" Map? Municipio! Property Stand 23°53'24" Zaragoza Ejido® La Encantada Zaragoza La Tinaja> F14A17 Aramberri 1820 99°42'39" 24°02'17" Aramberri Private Property: Agua Fria Agua Fria G14C87 Aramberri 2120 99°44'04" 24°02'34" Aramberri Private property: Agua de Alardin Canada el Puerto I G14C87 Aramberri 2200 99°43'54” 24°02'26” Arambetri Private property: Agua de Alardin Canada el Puerto II G14C87 Aramberri 2220 99°43'55” 24°02'14" Arambetrri Private property: Agua de Alardin Canada el Puerto III G14C87 Rayones 100°07'37" 2180 Montemorelos 2 LOA Ejido La Trinidad Canon el Butano’ G13C46 MADRONO tallest in the southern cluster and shortest in the northern (e.g., 16.0 m vs. 12.4 m for a tree of 30 cm dbh). The oldest tree aged from increment cores or | from ring counts on stumps was at Arroyo de la Pista and was 272-years-old. All five trees that were | aged at Arroyo de la Pista, from stumps or from increment cores, were over 200-years-old. The condition of many stands was poor due to a | combination of factors, including mistletoe, fire, and top dieback of unknown cause(s) (Table 5). De- pending on stand, from 0% to 33% of the trees in the 21 stands that we observed were infected with mistletoe; from less than 1% to 46% had fire scars; from 1% to 44% had some degree of top dieback. The percentage of trees infected with mistletoe in- creased with latitude (r = 0.53), but there seemed to be no geographic pattern with relation to fire or | dieback. DISCUSSION Thirty-nine stands of P. chihuahuana were con- firmed. Only 4 sites in Durango and 20 in Chihua- hua were known when Narvaéz et al. (1983) pub- lished their account. The stands occur in three rath- er distinct geographic clusters, which we will call the southern, central, and northern clusters. The clusters are about 2 to 2.5° of latitude apart and © consist of 4, 12, and 23 stands, respectively. Within clusters, stands are usually within a 20 km radius. However, Arroyo de Chachamori and Faldeo de Cebollitas are somewhat more isolated and could be considered outliers. Although we have been searching for 15 years, we expect that more stands will be found in the future. Stands are often identified by locals, foresters, and botanists with a variety of names. They may be called by a topographic feature, such as a river or a peak, or by the ejido or property on which they occur. Spelling may vary among sources, because different maps use different variants of the same indigenous word. Situriachic and Situriachi are ex- amples. Furthermore, many names are in common use and confuse the situation. For example, Arroyo de Quebrada is frequently seen on maps, and three stands in Table 2 have been known as Quebrada locally. Llano Grande and Arroyo Ancho, which take their names from ranchos, were reversed by Narvaéz (1984) and Narvaéz et al. (1983). Thus, our Llano Grande corresponds to their Arroyo An- cho. Use of the GPS coordinates presented here will certainly help in eliminating confusion. Our coor- dinates can be compared to previously published data for 19 stands (Narvaéz 1984; Narvaéz et al. 1983). Their coordinates seem to be in error for Mategoina I, II, and III, where longitude is about 1° off, a mistake easily introduced when reading coordinates from topographic maps, as Narvaéz et al. (1983) must have done. Previously published [Vol. 47 | 4 | | i | 1 2000] TABLE 4. LOCATION OF MEXICAN SPRUCE STANDS IN THE STATES OF NUEVO LEON, COAHUILA, AND CHIHUAHUA, MExico. ' A municipio is a political division of a state. 5 z , rla de la Defensa Number and key of the 1:50,000 national topographic maps of the Estados Unidos Mexicanos, published by the Departamento Cartografico, Secreta Nacional or the Instituto Nacional de Estadistica Geografia e Informatica (INEGI). *? In degrees, minutes, and seconds. * Recorded by GPS at the center of the stand and rounded to the nearest 5 meters. > Synonyms: Sierra de la Martha (Taylor and Patterson 1980; Taylor et al. 1994; but in Nuevo Leon, not in Coahuila as reported there); El Carmen (Martinez 1961); La Carmen (Rushforth 1986). ° Ejidos are lands given to groups of peasants after the 1910 revolution and usually held communally. (m) Elev.* 3500 Latitude? Longitude? N W Map? San Rafael Municipio! Property Stand 100°21'48" 252 POW: Rayones, N.L. Private Property: Familia Sierra la Marta> G14C45 San Rafael Sanchez de la Pena Ejido® Nuncio LEDIG ET AL.: SPRUCE IN MEXICO 81 100°21'12” 3470 25°14'49" Arteaga, Coah. Sierra el Coahuilén G14C45 Mohinora 107°02'21” 3185 25°57'42” Guadalupe y Calvo, Ch. Private Property: El Venadito El Mohinora G13C13 3200 — ~ . — 3000 ° ° 2800 ° ° — (eo) fe) fo} E ° ° z ° {e) ~ 2600} ° 52 Damaged by fire. * The number of affected spruce in each severity class weighted by the class value (1—3) and divided by the total number of trees plus regeneration. ° The number of spruce affected expressed as a percentage of the total number of spruce =0.3 m in height. Height Stand Trees Regen (m) Arroyo de la Pista 599 809 24.7 Arroyo del Chino 40 10 [5.3 Arroyo de las Lagunas 185 1930 16.1 Arroyo del Infierno 107 174 259 Faldeo de Cebollitas 83 172 13,3 Arroyo de los Angeles 1570 4435 Lou Estancia-Agua Amarilla 834 855 17.9 La Medalla 694 1174 16.7 Arroyo del Agua 373 420 1:2 La Medallita 264 218 16.0 El Saltito 472 Toe 199 Arroyo de Rosales 14 22 19.4 Arroyo del Indio Ignacio 1563 2501 13D Piedra Rayada 2342 3204 12:5 Arroyo de Enmedio 316 832 15.4 Arroyo de Quebrada 765 475 15.4 Ea Ye 11 10 11.7 Llano Grande 480 407 13.5 Las Lajas [5 1a 10.8 La Luisiana o2 90 9.8 122 120 15:3 DBH_ Density DHSS ee Sa Scar? (cm) (ha!) Score+ %° Score* Joe (%) 46.7 63.0 0.001 0.0 0.351 19:5 5:0 36.4 36.6 0.340 26.0 0.500 44.0 46.0 30.6 97.5 0.00 0.0 0.010 0.9 0.9 55.6 21-7 0.00 0.0 OrI25 6.8 14 32.0 22.0 0.00 0.0 0.122 6.3 3.5 26.0 161.1 0.190 13.3 0.100 7.9 4.5 S27 VES ey 0139 11.2 0.377 19.3 11.0 31.8 112.9 0.054 3.4 0.065 4.8 BS 33.4 78.3 0.486 26.5 0.211 13.8 5.0 52.2 62.5 0.224 16.4 0.145 8.5 6.2 38.3 84.0 0.151 10.6 0.114 me 4.2 44.0 — 0.361 30.6 0.306 25.0 13.9 25. 130.0 0.114 7.9 0.118 8.0 3.2 22:0)? Lei OAT 8.2 0.097 Fine 4.3 2702 126.47 “0.072 cee) 0.083 3.3 0.6 34.0 122.9 0.348 251 0.379 26.0 17.3 Qt — 0.476 33.3 0.286 28.6 4.8 30.0 — 0.306 21.5 0.268 20.4 14.5 33.5 — 0.462 26.9 0.346 19.2 Pls 28.4 — 0.253 18.7 0.275 22.0 9.9 38.8 — 0.182 14.9 0.161 12.4 22:3 Arroyo de las Ranas : and ours, our count of 107 stems =10 cm dbh, or even 56 stems =30 cm, dbh is in excess of the 36 trees and 9 seedlings recorded at Arroyo del Infier- no by Gordon (1968). Combining our data with that of Narvaéz (1984), the total number of P. chihuahuana =2 m in height is a sizeable figure, 24,221, and the number of seed- lings <2 m and 20.3 m is about 18,389. Density of P. chihuahuana 210 cm dbh, derived from the 1000 m? plots located in the centers of the stands, averaged 94.8 ha™' with a range for 15 stands from 21.7 to 187.7 ha! (Table 5). Although the data are not completely comparable, density was slightly lower in the 18 stands reported by Nar- vaéz et al. (1983), which had a mean of 72.8 and a range of 25.4 to 117.0 trees >2 m tall ha~'. Picea chihuahuana stands are similar to but, perhaps, slightly less dense than stands of P. breweriana S. Watson, (Brewer spruce) which may range from about 125 to 320 trees =10 cm dbh ha”! (Thorn- burgh 1990), or stands of P. sitchensis (Bong.) Car- riére, (Sitka spruce) which have about 188 trees ha™' (Harris 1990). However, density of boreal and other north temperate spruces is much higher: 815 to 1324 trees ha"! for well stocked P. glauca (Moench) Voss (white spruce) stands in Alaska and Saskatchewan (Nienstaedt and Zasada 1990); 1110 to 1780 trees ha"! of P. mariana (Mill.) B.S.P. (black spruce) in Ontario (Viereck and Johnston 1990); 140 to 780 P. engelmannii =4 inches (ap- proximately 10 cm) dbh in mixed spruce-fir forests in Wyoming (Oosting and Billings 1951); and 121 to 480 trees ha~! P. rubens Sarg. (red spruce) in the Maritime Provinces and Maine (Blum 1990; Oosting and Reed 1942). Though we have not counted the trees of P. mex- icana, they are numerous, perhaps in the thousands, at all three confirmed sites. Rushforth (1986) mis- takenly reported that the stand at Sierra La Marta was reduced to six trees at the top of the range. A fire in 1975 destroyed the type locality on the lower slopes of Sierra la Marta, and most of the spruces, but many hundreds escaped damage in a cafiada (a precipitous cleft smaller than a canyon) high on the mountain. Miiller-Using and Alanis (1984) counted 68 P. martinezii 210 cm dbh at La Tinaja and 350 at Cafion el Butano. The number at Agua Fria, re- cently discovered, may exceed the number at Ca- fion el Butano. Nevertheless, P. martinezii is ex- tremely rare. Our data from increment-cored trees was insuf- ficient to construct precise height-age and dbh-age relationships because of the limited range in age of the trees we bored. However, Gordon (1968) pre- sented relationships based on stem analyses and Narvaéz (1984) presented scatterplots of height and diameter on age for 37 trees spanning a range from about 20 to 120 yr. Based on Narvaéz’ (1984) scat- terplots, spruce on the ten sites he sampled in Chi- 2000] i} 1 LEDIG ET AL.: SPRUCE IN MEXICO 83 | TABLE 6. NUMBERS OF TREES AND SAPLINGS (T = 2 M TALL), SEEDLINGS (S < 2 M TALL), AND THE RATIO SEEDLINGS/ | TREES + SAPLINGS FOR ALL 39 CONFIRMED STANDS OF CHIHUAHUA SPRUCE. ' We did not include seedlings <0.3 m tall; | _ data for Narvaez (1984) apparently does. ? Counts made in April and May 1997. > Counts made in August and Sep- tember 1998. * Count made in June 1999. ° Counts reported in Narvaez (1984), which differ from counts in Narvaez et al. (1983). © During seed collections, we found only 17 trees, most with dead tops, and stumps, indicating that the _ population has declined in the last 15 years. ’ Called Llano Grande by Narvaez (1984) and Narvaez et al. (1983). Stand T | Arroyo de la Pista?’ 919 Arroyo del Chino? 46 Arroyo de las Lagunas? 505 Arroyo del Infierno? 148 Faldeo de Cebollitas” 172 Arroyo de los Angeles? 2507 La Estancia-Agua Amarilla?’ 1195 La Medalla?’ LOL? Arroyo del Agua?’ 510 La Medallita’ 356 El Saltito? 656 Arroyo de Rosales? 21 Arroyo del Indio Ignacio? 2628 Piedra Rayada? 3564 Arroyo de Enmedio? 465 Arroyo de Quebrada* 877 Rio Vinihueachi° 1785 E] Pinabetal? 455 Las Trojas° 874 Napahuichi IP 1064 Napahuichi IP 209 Talayotes? 291 a “y’’? 13 Situriachi® 389 Las Aguilas® 548 El Realito° 587 E] Cuervo? 140 El] Ranchito® 217 La Tinaja° 99 Cerro de la Cruz>° 20 Llano Grande? 545 Llano Grande? 614 Arroyo Ancho>’ 127 Las Lajas? 19 La Luisiana?* 127 Mategoina I° 124 Mategoina II> 448 Mategoina III 207 Arroyo de las Ranas?* 137 Arroyo de Chachamori° 146 Total 24,221 huahua would be about 29 m tall at 100 yr of age. Gordon’s (1968) stem analyses suggest a height of 28 m at 100 yr on a moist site at Arroyo del In- fierno in Durango, but only 12 m for a poorer site. Based on a regression of height on age for the 29 trees we bored (r = 0.49), we would expect a height of about 24 m at age 100. Thus, P. chihuahuana probably grows slowly, about one-quarter to one- third of a meter per year, on average, over its first 100 years. Compared to other North American spruces, P. chihuahuana does not seem particularly long-lived. Most of the large P. chihuahuana that we cored had between 100 and 200 rings at breast height. How- S! Total S/T 489 1408 0.532 4 50 0.087 1610 ZANS 3.188 133 281 0.899 83 259 0.483 3498 6005 1.395 494 1689 0.413 856 1868 0.846 28) 795 0.559 126 482 0.354 548 1204 0.835 15 36 0.714 Ps22 4150 0.579 1982 5546 0.556 683 1148 1.469 363 1240 0.414 1579 3364 0.885 267 T22 0.587 780 1654 0.892 921 1985 0.866 150 359 0.718 299 590 1.027 8 2A 0.615 286 675 0.735 168 716 0.307 210 797 0.358 96 236 0.686 162 379 0.747 37 136 0.374 5 25 0.250 370 915 0.679 273 887 0.445 8 135 0.063 7 26 0.368 D5 182 0.433 29 153 0.234 164 612 0.366 65 272, 0.314 105 242 0.766 24 170 0.164 18,389 42,610 0.759 ever, the largest trees had heart rot, which made it impossible to age them. Sanchez and Narvaéz (1990) reported finding wood rot in 40% of adult trees, possibly caused by fungi belonging to one or more of the genera Alternaria, Helminthosporum, Nigrospora, Sporotrichum, or Trichoderma. The oldest trees that we bored were 272- and 244-yr- old at breast height, and several stumps had over 200 rings. By contrast to P. chihuahuana, the oldest P. breweriana might be 900 yr (Waring, Emming- ham, and Running 1975); P. sitchensis may live to 700 or 800 yr (Harris 1990); trees of P. engelmannii between 500 and 600 yr of age are “‘not uncom- mon’”’ (Alexander and Shepperd 1990) and can sur- 84 MADRONO Frequency (%) [Vol. 47 ononn non wo WwW WH oOnononnwnwnwno WO WO onononnwown wo WH NRNRANRAR NRARARAR NRENRARAR —- OoOwronon~ & © -~ OwrOR OD O - OwroOorkrR Oo O —_ b — DBH Classes Fic. 3. Frequency of observations in 5 cm-dbh classes plotted over the class mid-point for all Chihuahua spruce =0.3 m tall in the southern, central, and northern clusters of stands. vive to 680 yr (Brown et al. 1995); P. pungens Engelm. (blue spruce) may survive to 600 yr or more (Fechner 1990); and the maximum age in P. rubens is about 400 yr (Blum 1990). The oldest P. glauca on good sites may be 250 to 300 yr, similar to P. chihuahuana, but above the Arctic Circle, slow-growing P. glauca may reach 1000 yr (Nien- staedt and Zasada 1990). Picea mariana, alone of the North American spruces, may be slightly short- er-lived than P. chihuahuana, but P. mariana of 280 yr have been reported (Viereck and Johnson 1990). Miiller-Using and Lassig (1986) found a P. martinezii with a ring count of 279 at breast height. With regard to ultimate size, P. chihuahuana is average among the North American spruces. The largest trees were in the southern cluster and were about 50 m in height and 125 cm dbh. Size seemed to decrease from the southern to the northern clus- ter. The largest P. martinezii were only 32 m tall and 62 cm dbh (Miiller-Using and Lassig 1986). The largest P. breweriana, P. engelmannii, and P. pungens are very close in height and diameter to P. chihuahuana (Thornburgh 1990; Waring et al. 1975; Alexander and Shepperd 1990; Fechner 1990). Picea glauca may be slightly smaller, the largest being about 55 m by 120 cm (Nienstaedt and Zasada 1990). Picea mariana and P. rubens are substantially smaller, reaching 27 m height by 46 cm dbh and 35 m height by 61 cm dbh, respectively (Viereck and Johnson 1990; Blum 1990). On the other extreme, P. sitchensis is a giant, the largest being about 66 m by 510 cm (Harris 1990). Many P. chihuahuana were in poor condition or damaged. Within the 21 stands that we studied, from 1% to 44% of the trees had dead tops from unknown causes. The severity varied from a third to almost the complete crown, and incidence tended to be higher in smaller stands (Table 5). Dead tops were also observed by Narvaéz (1984) and Narvaéz et al. (1983), but only in 1.3% of the spruces in the State of Chihuahua. We found that 19% of the spruces had some degree of dieback in the five stands we observed in the northern cluster, includ- ing 181 of 887 spruce (20%) at Llano Grande, which was also scored by Narvaéz (1984). Narvaéz (1984) recorded only 4 of 915 spruce with top die- back, “‘puntisecos’’. Perhaps, the condition has be- come substantially worse in the 15 years between observations. Or, perhaps, Narvaéz (1984) scored only the most severely affected trees. In our sample of five stands in Chihuahua, only 2.2% of the trees =10 cm dbh had tops with complete dieback, but 33.2% had dieback of less severity. In any case, damage from top dieback and other causes is extensive, and further threatens this rare species. Fire scars were noted in every stand, and reached high proportions, up to 46% in some small- er stands. In the survey of Sanchez and Narvaéz (1990), only 3.4% of the spruce were damaged by fire. Even though P. chihuahuana is nominally pro- Arroyo del Infierno = -0.268 + 0.958x Log(HEIGHT) Arroyo de las Lagunas y = -0.065 + 0.852x Ps LEDIG ET AL.: SPRUCE IN MEXICO 2.4 0.0 85 Arroyo de la Pista y = -0.040 + 0.864x Arroyo del Chino y = -0.386 + 0.999x 0.6 2 1.8 2.4 Log(DBH) Fic. 4. Regression of log(height in m) on log(dbh in cm) for a sample of 190 Chihuahua spruces from the southern cluster of stands (Arroyo del Infierno, Arroyo de la Pista, Arroyo de las Lagunas, and Arroyo del Chino). tected, clandestine cutting occurs, as evidenced by the 40 stumps on which we made ring counts. Many other trees were topped for Christmas deco- rations. Our observation of mistletoe seems to be the first report on P. chihuahuana (Hawksworth and Wiens 1996), and we have not identified the species. We observed mistletoe on spruce in some stands, but not in all. Mistletoe was least common and the fre- quency of infected trees was least in the southern cluster and greatest in the northern cluster, where it exceeded 19%. Sanchez and Narvaéz (1990) did not mention mistletoe. Dwarf mistletoes are dam- aging to all other North American spruces north of México. Arceuthobium microcarpum causes heavy mortality to P. engelmannii in Arizona and New Mexico, but is not found farther north (Alexander and Shepperd 1990). It also colonizes P. pungens (Fechner 1990). In some stands, 36% of P. brew- eriana are parasitized by A. campylodum (Thorn- burgh 1990), which may destroy infected trees. A. pusillum is common and destructive on black spruce in eastern North America, but absent in the West (Viereck and Johnston 1990), and it also col- onizes P. glauca (Nienstaedt and Zasada 1990), and, occasionally, P. rubens (Blum 1990). Some stands seemed to be in poorer health than others, and these included some of the smallest stands. For example, the number of spruce at Ar- royo del Chino was only 50, and 44% of these had tops affected to various degrees by dieback, 46% were scarred by fire, and 26% had mistletoe. In addition, the ratio of seedlings to trees and saplings was only 0.087. It may be very difficult to conserve such stands, and we suggest that seeds or cuttings be collected as soon as possible to conserve their genetic resources. The most threatened stands can be identified in Tables 5 and 6. The most critical question with regard to the con- servation of P. chihuahuana is whether the stands are regenerating adequately. The distribution of spruce into 5 cm diameter-size classes forms a re- verse J-shape (Fig. 3) in most stands, particularly the large ones. This suggests that the older trees will be replaced. However, this may be misleading. The ratio of seedlings <2 m tall to trees and sap- lings =2 m is less than 1.0 in all except four of 39 stands for which data exist. One of these, Arroyo TABLE 7. REGRESSION COEFFICIENTS (B) AND INTERCEPTS (A) OF LOG(HEIGHT IN M) ON LOG(DBH in cm), with Standard Errors (in Parentheses), and Correlation Coefficients (R) in Chihuahua Spruce from Northern, Central, and Southern Clusters. ' Five stands are represented in the northern cluster, 12 in the central, and 4 in the southern. 7 Number of trees, from the interior plots in each stand. Cluster! n b a r northern 341 0.9524 (0.0221) —0.3145 (0.0306) 0.920 central 1235 0.8704 (0.0097) —0.0944 (0.0125) 0.938 southern 193 0.8458 (0.0188) —0.0461 (0.0247) 0.956 86 MADRONO de las Lagunas, was heavily logged somewhat over a decade ago, removing many of the large trees, which may explain the high ratio of 3.19 seedlings to saplings and trees. In many stands, the number of trees =10 cm dbh exceeds the number of seed- lings <2 m tall, suggesting that the stands are in trouble. We believe that the ratio of seedlings to trees is significant because P. chihuahuana probably regen- erates in the shade. Although ecological studies of P. chihuahuana are lacking on this point, other North American spruces, such as P. breweriana, P. engelmannii, P. glauca, P. rubens, P. mariana, and P. sitchensis are tolerant of shade. In fact, P. brew- eriana and P. engelmannii seedlings cannot survive strong sunlight (Thornburgh 1990; Alexander and Shepperd 1990; Ronco 1975), and regeneration of P. pungens, P. rubens, and P. sitchensis benefits from shade (Fechner 1990; Blum 1990; Harris 1990). All spruces north of México are thin-barked and highly susceptible to fire, and do not normally require disturbance for regeneration, with the ex- ception of the partially serotinous P. mariana. The ratio of seedlings to saplings and trees in some other North American spruces seems at least twice that observed in P. chihuahuana. However, comparison with other studies is difficult because of differences in methodology and presentation of data. In P. rubens, the ratio of stems =10 ft tall to stems >10 ft was 1.87 for a 60-yr-old stand in Maine (Oosting and Reed 1942). In eight stands of P. engelmannii in Wyoming, the ratio of stems <8 ft tall to stems >8 ft averaged 2.34, ranging from 0.24 to 7.33 (Oosting and Reed 1952). For seven stands in the Smoky Mountains of North Carolina and Tennessee, the ratio of P. rubens stems <2 inches dbh to stems =2 inches was 8.45, and in four stands in the White Mountains of New Hamp- shire, the ratio was 23.81, with a range of 22.26 to 26.85 (Oosting and Billings 1951). A 2-inch di- ameter limit is roughly equivalent to a 3 m tall P. chihuahuana (from Table 7). If Oosting and Bill- ings’ (1951) 1l-inch diameter limit is used (equiv- alent to a 1.5 m tall P. chihuahuana), the ratios of P. rubens regeneration to trees was 5.24 in the Smoky Mountains and 13.34 in the White Moun- tains. Global warming is a threat to the cool temperate conifer forest of México (Villers and Trejo 1997). Therefore, we might expect the stands in the south- ern cluster to be in greater decline than those in the central cluster, and those in the central cluster to be in greater decline than those in the northern cluster. This does not yet seem to be the case. The ratio of seedlings to saplings and trees was 1.18, 0.72, and 0.53 for the southern, central, and northern clusters, respectively. However, trees were larger on average in the south and smallest in the north, suggesting that northern stands are younger and will survive longer without replacement. [Vol. 47 ACKNOWLEDGMENTS This study was an undertaking of the Forest Genetic Resources Study Group/North American Forestry Com- mission/Food and Agricultural Organization of the United Nations. It was completed with the help of National Re- search Initiatives Competitive Grant Program award no. 95-37101-1916 and PSW Global Change Research Grant 95-02. We are grateful to Jesis SAnchez-Cérdova for shar- ing his great knowledge of Chihuahua spruce with us, and thank James A. Baldwin, Michael G. Barbour, G. E. Reh- feldt, and two anonymous reviewers for helpful comments that greatly improved the manuscript. LITERATURE CITED ALEXANDER, R. R. AND W. D. SHEPPERD. 1990. Picea en- gelmannii Parry ex Engelm. Engelmann spruce. Pp. 187—203 in R. M. Burns and B. H. Honkala, tech. coords., Silvics of North America. Vol. 1, Conifers. Agriculture Handbook 654. USDA, Forest Service, Washington, DC. ANONYMOuS. 1991. NS/S nominates Sierra Madre as a global center of plant diversity. The Seedhead News No. 34:4. Native Seeds/SEARCH, Tucson, AZ. BARBOUR, M. G. AND R. A. MINNICH. 1999. Californian upland forests and woodlands. Pp. 159-200 in M. G. Barbour and W. D. Billings (eds.), North American terrestrial vegetation, 2nd ed. Cambridge University Press, NY. BLum, B. B. 1990. Picea rubens Sarg. red spruce. Pp. 250-259 in R. M. Burns and B. H. Honkala (tech. coords), Silvics of North America. Vol. 1, Conifers. Agriculture Handbook 654. USDA, Forest Service, Washington, DC. Bosrov, E. 1970. History and systematics of the genus Picea A. Dietr. [in Russian]. Novosti Sistematiki Vys- shikh Rastenii 7:5—40. Brown, P. M., W. D. SHEPPERD, C. C. BRown, S. A. MATA, AND D. L. McCLain. 1995. Oldest known Engelmann spruce. U.S. Forest Service Research Note RM-RN- 534. Burns, R. M. AND B. H. HONKALA (tech. coords.). 1990. Silvics of North America. Vol. 1, Conifers. Agricul- ture Handbook 654. USDA, Forest Service, Washing- ton, DC. CAPO-ARTEAGA, M., S. WALENCIA-MANZO, C. FLORES- LOPEZ, AND S. BRAHAM-S. 1997. Informe final de ac- tividades del proyecto: autoecologia del género Picea en Nuevo Leon. Presentado a: Consejo Consultivo Estatal para la Preservaci6n y Fomento de la Flora y Fauna Silvestre de Nuevo Leon. Universidad Aut6n- oma Agraria Antonio Narro, Buenavista, Saltillo, Coahuila. CLIsBy, K. H. AND P. B. SEARS. 1955. Palynology in south- ern North America. part III: microfossil profiles under Mexico City correlated with the sedimentary profiles. Bulletin of the Geological Society Amer. 66:511—520. CorRRELL, D. S. 1960. A mule-train trip to Sierra Mohi- nora, Chihuahua. American Fern Journal 50:66—78. DALLIMORE, W. AND A. B. JACKSON. 1923. A handbook of Coniferae including Ginkgoaceae. Edward Arnold and Company, London. EveretTT, T. H. 1981. The New York Botanical Garden illustrated encyclopedia of horticulture, Vol. 9 Par-Py. Garland Publishing, NY-London. FECHNER, G. H. 1990. Picea pungens Engelm. blue spruce. Pp. 238-249 in R. M. Burns and B. H. Honkala (tech. 2000) coords.), Silvics of North America, Vol. 1, Conifers. Agriculture Handbook 654. USDA, Forest Service, Washington, DC. Gorbon, A. G. 1968. Ecology of Picea chihuahuana Mar- tinez. Ecology 49:880—896. GRAHAM, A. 1993. Historical factors and biodiversity in Mexico. Pp. 109-127 in T. P. Ramamoorthy, R. Bye, A. Lot, and J. Fa (eds.), Biological diversity of Mex- ico: origins and distribution. Oxford University Press, NY-Oxford. Harris, A. S. 1990. Picea sitchensis (Bong.) Carr. Sitka spruce. Pp. 260—267 in R. M. Burns and B. H. Hon- kala (tech. coords.), Silvics of North America. Vol. 1, Conifers, Agriculture Handbook 654. USDA, Forest Service, Washington, DC. HAwkKSworTH, F G. AND D. WIENS. 1996. Dwarf mistle- toes: biology, pathology, and systematics. Agricultur- al Handbook 709. USDA, Forest Service, Washing- ton, DC. HopkINs, A. D. 1938. Bioclimatics: a science of life and climate relations. Miscellaneous Publication No. 280. USDA, Washington, DC. KINDRED, D. ‘‘How far is it?’’ Distance calculator. (http:/ /www.indo.com/distance/, accessed 1997). Lepbic, F T., V. JACOB-CERVANTES, P. D. HODGSKISS, AND T. EGUILUZ-PIEDRA. 1997. Recent evolution and di- vergence among populations of a rare Mexican en- demic, Chihuahua spruce, following Holocene cli- matic warming. Evolution 51:1815—1827. LITTLE, E. L., JR. 1971. Atlas of United States trees. Vol. 1. Conifers and important hardwoods. Miscellaneous Publication No. 1146. Forest Service, USDA, Wash- ington, DC. LOZANO-GaRCIA, M. S., B. ORTEGA-GUERRERO, M. CAB- ALLERO-MIRANDA, AND J. URRUTIA-FUCUGAUCHI. 1993. Late Pleistocene and Holocene paleoenvironments of Chalco Lake, central Mexico. Quaternary Res. 40: 332-342. MAHLMAN, J. D. 1997. Uncertainties in projections of hu- man-caused climate warming. Science 278:1416— 1417. MARTINEZ, M. 1953. Las Pinaceas Mexicanas. Subsecre- tarfa de Recursos Forestales y de Caza, Secretaria de Agricultura y Ganaderia, México. MARTINEZ, M. 1961. Una nueva especie de Picea in Méx- ico. Anales del Instituto de Biologia de la Universi- dad Nacional de México 32:137-142. McDonaLp, J. A. 1993. Phytogeography and history of the alpine-subalpine flora of northeastern Mexico. Pp. 681-703 in T. PR. Ramamoorthy, R. Bye, A. Lot, and J. Fa (eds.), Biological diversity of Mexico: origins and distribution. Oxford University Press, NY-Ox- ford. MULLER-USING, B. AND G. ALAN{S-FLORES. 1984. Nuevos registros del pinabete de Chihuahua (Picea chihuahu- ana Martinez) en Nuevo Le6n propuesta para la pro- teccion legas de dos areas de especial interés ecol6- gico. Pp. 130—132 in Reunién Regional de Ecologia Norte, 25, 26 y 27 de abril 1984, Monterrey, Nuevo Le6n, México. Secretaria de Desarrollo Urbano y Ecologia, Subecretaria de Ecologia, México. MULLER-USING, B. AND R. LAssic. 1986. Zur Verbreitung der Chihuahua-Fichte (Picea chihuahuana Martinez) in Mexiko. Mitt. Deutsch. Dendrol. Ges. 76:157—169. NARVAEZ-FLORES, R. 1984. Contribucion al conocimiento de la ecologia de Picea chihuahuana Martinez. Un- published thesis. Universidad Aut6noma de Nuevo Leén, Monterrey, Nuevo Leon. LEDIG ET AL.: SPRUCE IN MEXICO 87 NARVAEZ-FLORES, R., J. SANCHEZ-CORDOVA, AND A. OLI- VAS-MEZA. 1983. Distribucion y poblacion de Picea chihuahuana [in Spanish]. Nota Técnica No. 6, Cen- tro de Investigaciones Forestales del Norte, Instituto Nacional de Investigaciones Forestales, Secretaria de Agricultura y Recursos Hidraulicos. NIENSTAEDT, H. AND J. C. ZASADA. 1990. Picea glauca (Moench) Voss white spruce. Pp. 204—226 in R. M. Burns and B. H. Honkala, tech. coords., Silvics of North America, Vol. 1, Conifers. Agriculture Hand- book 654. USDA, Forest Service, Washington, DC. OosTING, H. J. AND W. D. BILLINGS. 1951. A comparison of virgin spruce-fir forest in the northern and southern Appalachian system. Ecology 32:84—103. OosTING, H. J. AND J. KF REED. 1942. The establishment of a white birch community on cutover pulpwood land in northwestern Maine. Bull. Torrey Bot. Club 69:647—660. OosTING, H. J. AND J. EF REED. 1952. Virgin spruce-fir of the Medicine Bow Mountains, Wyoming. Ecological Monographs 22:69-91. PATTERSON, T. EK 1988. A new species of Picea (Pinaceae) from Nuevo Leén, México. SIDA 13:131—135. Ronco, E 1975. Diagnosis: “‘sunburned”’ trees. Journal of Forestry 73:31-35. RUSHFORTH, K. 1986. Mexico’s spruces—rare members of an important genus. Kew Magazine 3:119-—124. SANCHEZ-CORDOVA, J. 1984. Picea chihuahuana Mrtz., una conifera en peligro de extincion. Ciencia Forestal, Revista del Instituto Nacional de Investigaciones Fo- restales, Secretaria de Agricultura y Recursos Hidrau- licos 9(no. 51):51—63. SANCHEZ-CORDOVA, J. AND R. NARVAEZ-FLORES. 1990. Plan integral para la proteccién y fomento de Picea chihuahuana Martinez. Ecosistemas Forestales, Insti- tuto Nacional de Investigaciones Forestales y Agro- pecuarias, Secretaria de Agricultura y Recursos Hid- raulicos 1:37-—55. STATSOFT. 1995a. Statistica, Vol. III: Statistics I. StatSoft, Inc., Tulsa, OK. STATSOFT. 1995b. Statistica, Vol. I: General Conventions and Statistics I. StatSoft, Inc., Tulsa, OK. STEEL, R. G. D. AND J. H. Torrie. 1960. Principles and procedures of statistics with special reference to the biological sciences. McGraw-Hill, NY. TAYLOR, R. J. AND T. EF PATTERSON. 1980. Biosystematics of Mexican spruce species and populations. Taxon 29: 421-469. TAYLOR, R. J., T. F PATTERSON, AND R. J. HARROD. 1994. Systematics of Mexican spruce—revisited. Syst. Bot. 19:47-59. THORNBURGH, D. 1990. Picea breweriana Wats. Brewer spruce. Pp. 181—186 in R. M. Burns and B. H. Hon- kala (tech. coords.), Silvics of North America, Vol. 1, Conifers. Agriculture Handbook 654. USDA, Forest Service, Washington, DC. VARGAS-HERNANDEZ, J. J. 1996. Forest land ownership and management in México. P. 21 in D. L. Rogers and E T. Ledig (eds.), The status of temperate North American forest genetic resources. Report no. 16, University of California Genetic Resources Conser- vation Program, Davis, CA. VERA, G. 1990. Algunas especies arboreas raras 0 amen- azadas en México. Secretaria de Agricultura y Re- cursos Hidraulicos (SARH), Instituto Nacional de In- vestigaciones Forestales y Agropecuarias (INIFAP), México. VIERECK, L. A. AND W. E JOHNSTON. 1990. Picea mariana [Vol. 47 88 MADRONO (Mill.) B.S.P. black spruce. Pp. 227—237 in R. M. WARING, R. H., W. H. EMMINGHAM, AND S. W. RUNNING. Burns and B. H. Honkala (tech. coords.), Silvics of 1975. Environmental limits of an endemic spruce, North America. Vol. 1, Conifers. Agriculture Hand- Picea breweriana. Canadian Journal of Botany 53: 1599-1613. book 654. USDA, Forest Service, Washington, DC. VILLERS-RuIz, L. AND I. TREJO-VAZQUEZ. 1997. Assess- WRIGHT, J. W. 1955. Species crossability in spruce in re- ment of the vulnerability of forest ecosystems to cli- lation to distribution and taxonomy. Forest Science 1: mate change in Mexico. Climate Research 9:87—93. 319-349. - MaproNo, Vol. 47, No. 2, pp. 89-96, 2000 REVISION OF CORETHROGYNE (COMPOSITAE: ASTEREAE) J. PHILLIP SAROYAN 1838 Indiana, Vallejo, California 94590 DENNIS R. PARNELL Department of Biology, Santa Clara University, Santa Clara, California 95053-0001 JOHN L. STROTHER University Herbarium, 1001 Valley Life Sciences Building #2465, University of California, Berkeley, California 94720-2465 ABSTRACT Review of variation in morphological characters within and among natural populations and comparisons of greenhouse-grown plants, coupled with high pollen stainabilities in progeny from crosses of plants from different populations, led to the conclusion that Corethrogyne should be treated taxonomically as comprising a single species with two varieties. The revised taxonomy results in the new combination Corethrogyne filaginifolia var. californica. Corethrogynes occur, as scattered, local popula- tions, from Coos Co., Oregon, south through much of California into northern Baja California, from sea level along the immediate coast to ca. 2500 m in the Sierra Nevada. Keck (1959) recognized three species of Corethrogyne DC.; Ferris (1960) recog- nized seven species. Lane (1992, 1993) treated the type species of Corethrogyne and Lessingia Cham- iSSO aS congeneric and, based in part on the work summarized here, treated the corethrogynes of Keck and Ferris as two varieties in one species of Lessingia. Morphological variation within and among pop- ulations of these plants is considerable and is re- flected in the 33 basionyms that have been linked to the generic name Corethrogyne. The 30 ‘taxa’ referable here to Corethrogyne have been distin- guished mainly by traditional morphological char- acteristics such as habit, size and shape of leaves, number of heads per flowering stem, size of heads, and shape of involucres. Aspects of indument such as relative amounts of tomentum and/or stipitate- glandular hairs have also been used in drawing cir- cumscriptions. Field observations by Saroyan indicated that plants with characteristics of reputedly allopatric ‘taxa’ sometimes grow together in local popula- tions. Plants with heads (including rays) 15 vs. 30 mm wide (assignable to two distinct ‘species’), plants with leaves 5 vs. 19 mm wide (different ‘spe- cies’), or plants with involucres 7 vs. 12 mm long (different ‘species’ or different ‘varieties’) are found in local populations, represent extremes of continua, and exemplify the kinds of problems that are found in past taxonomies. Such problems of reconciling taxonomy and plants did not go unnot- iced by Keck or Ferris. ‘This imperfectly known sp. [C. leucophylla] recombines the characters of the other two [C. californica and C. filaginifolia sensu Keck] and needs further study.’’ —Keck (1959). ‘Relatively few individual collections in any given range completely conform to the type and the original description. Intermediate forms abound ....’’ —Ferris (1960). On transfer of Corethrogyne to Lessingia, Lane (1992) emphasized morphological and chemical (chloroplast DNA) similarities. We emphasize the differences, as did Jones (1977), who compared corethrogynes and lessingias and concluded, “‘It would serve no useful purpose to argue for a con- generic status.” The genera are readily distinguished: Plants perennial; heads radiate ..... Corethrogyne Plants annual; heads discoid or + radiant (corollas of peripheral florets often strongly zygomorphic) ee eee ee ee eee ae eee Lessingia MATERIALS AND METHODS This paper is based on an unpublished thesis (Sa- royan 1974). In addition to traditional field observations made up and down California and traditional review of specimens in herbaria (ca. 1500 sheets from 17 her- baria - see Acknowledgments), Saroyan (1974) studied morphological variation in samples from 20 populations of Corethrogyne from northern to southern California (A-T in Appendix 1). Each population was associated with one of three vege- tation types: type I, grasslands; type I, coastal scrub and chaparral; and type III, forest. For each of the study populations, one to five plants grown from seed in greenhouses in Hayward, CA were 90 MADRONO [Vol. 47 Head Width Number of Ray > " y, Flowers per Head Number of \ UE umoer o Wat Series of —d wk | Phylleries Involucre Length Leat Width aa, Os : | s a Involucre \\" | \ vi Y Width \\; i \/ / d Depth of Leaf \\Y Serrations Leaf Length Number af Serrations per Leaf Cypsela Width Win KP wy We pa of Ss =" a \ : se Ae per eens 4 ANE LP Ses eV Floral — Length Fic. 1. Characters of a generalized Corethrogyne. used in tests of self-compatibility and in inter-pop- ulation crosses and served as tests for phenotypic plasticity in the study populations. Saroyan (1974) chose characters that have been used in taxonomy of Compositae in general and/or in Corethrogyne in particular and that could be eas- ily assessed statistically (cf. Fig. 1). He determined mean, standard error of the mean, and range of val- ues for: 1) length of stem, 2) length of leaf, 3) width of leaf, 4) length of floral stem, 5) number of heads per flowering stem, 6) width of head, 7) number of ray florets per head, 8) length of involucre, 9) di- ameter of involucre, 10) number of series of phyl- laries, 11) number of serrations per leaf, 12) depth of leaf serrations, 13) length of cypsela, 14) width of cypsela, 15) length of pappus, 16) ratio length/ width of leaf, and 17) ratio length/diameter of in- volucre. To determine minimum adequate sample size for > ys Qe SP uth ea SIAN NE Stem Length representative statistics for morphological charac- ters in wild populations, Saroyan selected samples of 25, 50, and 100 plants from one such population; data from those three samples were pooled to give a fourth sample of 175 plants. Comparisons of F- test statistics for the four samples indicated that re- liable results could be achieved from a sample size of 50 individuals. Fifty plants were randomly se- lected in the field from each of the 20 study pop- ulations by use of a grid and a table of random numbers, yielding a grand total of 1000 plants sam- pled. Sampling of individual structures was also ran- domized. For some characters, all like structures (e.g., heads at mid anthesis) were removed from a plant and tossed into a paper bag, then one was extracted blindly for measurement. For other char- acters, the structure to be measured was chosen by association with a table of random numbers. Mean, 2000] standard error of the mean, and range were deter- mined for each set of 50 measurements. Chromosome counts were made from microspo- rocytes fixed in acetic ethanol (3:1, v:v) and stained in aceto-carmine for 2—13 plants from each study population. Pollen was stained in lactophenol-cot- ton blue for 1—13 plants (300 grains counted for each plant) from each study population. Vouchers for each study population and repre- sentative greenhouse-grown plants (including prog- eny from crosses made in the greenhouse) were de- posited in UC. RESULTS AND DISCUSSION Character by character comparisons across the 20 study populations confirmed our impression of gen- eral morphological continuity within Corethrogyne. Not only ranges of absolute values but even mea- sures of two standard errors of the mean for any one population overlapped with the same values or measures for some or all of the other 19 popula- tions (e.g., length of stem and width of head; cf. Fig. 2). Nevertheless, the five northern populations (A-E in Appendix 1) showed coherence and were somewhat distinct from the southern ones (F—T in Appendix 1) for some characters (e.g., number of heads per flowering stem and the ratio length/di- ameter of involucre; cf. Fig. 3). Variation in single characters seldom showed clear correlations with environmental or habitat pa- rameters. Plants from coastal populations usually had shorter, more prostrate stems; plants from in- land populations were usually more erect. The four populations with longest and narrowest leaves were the four associated with chaparral or coastal scrub vegetation; they were not distinctive for other char- acters. Morphologies of greenhouse-grown progeny from each of the study populations were similar to those of parental plants. Evidently, phenotypic plas- ticity is not strong for any of the characters exam- ined. All published reports of chromosome number (see standard indices; e.g., Goldblatt and Johnson 1992) for Corethrogyne have given 2n = 10 (some as n = 5; some as 2n = 5 II). Similarly, we found meiosis to be regular with 2m = 5 II in all of our samples, including the interpopulational hybrids. Pollen stainabilities in our study populations ranged from 80 to 97 percent (Appendix 1). One or more plants grown from seed collected from each study population were self-pollinated; none set viable seeds. Inter-population crosses of one to five pairs of plants in the combinations E X H, N X H, J X D, and M X C (cf. Appendix 1) yielded hybrids with no obvious irregularities at meiosis and with pollen stainabilities of 82 to 96 percent. In general aspect, some local populations of Cor- ethrogyne are strikingly different from others. Nev- ertheless, overlap and continuity in expression of SAROYAN ET AL.: CORETHROGYNE v1 most morphological characters, coupled with cross- compatibility between plants from quite disjunct and dissimilar populations, have led us to treat Cor- ethrogyne aS comprising a single species. Plants from the northwestern part of the range of the spe- cies usually have fewer heads on each flowering stem and have larger heads than do plants from other areas; we have recognized them as a variety. TAXONOMY CORETHROGYNE DC., Prodr. 5:215. 1836. —Type: Corethrogyne californica DC. Herbaceous or suffrutescent perennials, primary stems decumbent to ascending or erect, mostly 1— 10 dm long, usually densely white-tomentose, sometimes becoming glabrate and/or glandular dis- tally. Leaves alternate, often crowded at bases of stems, sessile or with bases of blades + decurrent on petioles, blades ovate to spatulate, oblanceolate, or linear, 1-7+ cm long, 3—19+ mm wide, becom- ing smaller, sessile, and bractlike distally, margins entire or variously toothed. Heads pedunculate or sessile, 1—-20+ per floral stem. Involucres hemi- spheric to campanulate, turbinate, or cylindric, 6— 14 mm long, 3—10 mm in diam. Phyllaries 30—90+, strongly graduated in 3—9 series, narrowly lanceo- late to linear, cartilaginous to scarious with herba- ceous, often spreading to squarrose tips, becoming deflexed as cypselae are shed. Ray florets 10—43 in 1 series, neutral, corollas purplish through violet and pink to white, laminae + linear. Disc florets 12—120+, bisexual, corollas yellow, actinomorphic, 4-8 mm long, tubes 0.6—1.4 mm long, glabrous, throats very narrowly cylindric, 2.8—5.5 mm long, often sparsely puberulent, lobes equal, narrowly lanceolate, 0.7—1.2 mm long, sparsely to densely glandular-puberulent abaxially, papillate-ciliolate on margins and/or adaxially; style branches linear with blunt to subulate appendages, + hispid with rigid yellow hairs, the appendages to half as long as the stigmatic lines. Cypselae cuneiform to linear, mostly 2—5 mm long, 5—7-ribbed, puberulent to pi- lose. Pappi of 35—65 coarse, unequal, brownish to reddish bristles 3-8 mm long. Chromosomes: 2” = 10. As treated here the genus is monotypic. CORETHROGYNE FILAGINIFOLIA (Hook. & Arn.) Nutt., Trans. Amer. Phil. Soc., ser. 2. 7:290. 1840 [1841]. = Aster? filaginifolius Hook. & Arn., Bot. Beechey voy. 146. 1833. = Corethrogyne californica DC. [var.] filaginifolia (Hook. & Arn.) Kuntze, Rev. gen. pl. 1:330. 1891 [illegit., oldest sp. epithet not used]. = Corethrogyne fi- laginifolia (Hook. & Arn.) Nutt. var. typica M. L. Canby, Bull. S. Calif. Acad. Sci. 26:10. 1927. = Lessingia filaginifolia (Hook. & Arn.) M. A. Lane, Novon 2:213. 1992. —Type: California, F. W. Beechey et al. s.n. (holotype: E!). 92 MADRONO [Vol. 47 390 — 80 eg ©). 70 Ee e 60 ) - 50 ° £ 40 oe e e 30 a) 20 10 A B C 0 E F GH t| J K LM N O P QR S T Population E Ee e e 2 40 ° uo nn» e e 2° 2 | x ) cw 20 = e =x 10 Width of A BC DB E F GH tJjy K LE MN O P QR S§ T Population Fic. 2. Dice diagrams for variation (mean, + two standard errors of the mean, and absolute range) in stem length and in width of head (cf. Fig. 1) in 20 populations (cf. appendix 1) of Corethrogyne (A-E = C. filaginifolia var. californica; F-T = C. f. var. filaginifolia). As indicated in discussion above, we recognize 10 mm, about equal to diameters.......... two + allopatric varieties within C. filaginifolia. 4 +++ sere eee C. filaginifolia var. californica KEY TO VARIETIES OF CORETHROGYNE FILAGINIFOLIA 1. CORETHROGYNE FILAGINIFOLIA (Hook. & Arn.) Nutt. var. FILAGINIFOLIA. 1. Floral stems usually branched, each with 3—6(1— Aster? tomentellus Hook. & Arn., Bot. Beechey 20+) heads; lengths of involucres (fresh, at anthe- sis) 6-14 mm, mostly twice the diameters ... voy. 146. 1833. = Corethrogyne tomentella ee ee ee C. filaginifolia var. filaginifolia (Hook. & Arn.) Torrey & A. Gray, Fl. N. Amer. 1’ Floral stems usually unbranched, each with 1(—5) 2:99. 1841. = Corethrogyne californica DC. heads; lengths of involucres (fresh, at anthesis) 5— [var.] tomentella (Hook. & Arn.) Kuntze, Rev. 2000] SAROYAN ET AL.: CORETHROGYNE 93 20 Number of Heads per Flowering Stem arpeeue A B C 0D E F G H It J K LE M N O P QR S T Population Retlo of Length/Diemeter of Involucre A BC 0 E F G H | J K LE MN O P QAR S T Populetion Fic. 3. Dice diagrams for variation (mean, + two standard errors of the mean, and absolute range) in numbers of heads per flowering stem and in ratio of length/diameter of involucre in 20 populations (cf. Appendix 1) of Coreth- rogyne (A-E = C. filaginifolia var. californica; F-T = C. f. var. filaginifolia). gen. pl. 1:330. 1891 [illegit., oldest sp. epithet Sci. 26:14. 1927. —Type: California, D. Douglas not used]. —Type: California, F. W. Beechey et s.n. (no specimen located; the plate confirms ap- al. s.n. (holotype: K!). plication of the name). Diplopappus incanus Lindl., Edward’s Bot. Reg. Aplopappus? haenkei DC., Prodr. 5:349. 1836. — 20:1693. 1835. = Corethrogyne incana (Lindl.) Type: “‘inter plantas Regiomontanas herb. Haen- Nutt., Trans. Amer. Phil. Soc., ser 2. 7:290. 1840 keani ab ill. de Sternberg ad studium missi con- [1841]. = Corethrogyne californica DC. [var.] servatur’’; specimens came from near Monterey, incana (Lindl.) Kuntze, Rev. gen. pl. 1:330. 1891 California, fide A. Gray (1876). Holotype may [illegit., oldest sp. epithet not used]. = Corethro- be in P; specimen in G-DC (microfiche!) con- gyne filaginifolia (Hook. & Arn) Nutt. var. in- firms application of the name. cana (Nutt.) M. L. Canby, Bull. S. Calif. Acad. Diplopappus leucophyllus Lindl. ex DC., Prodr. 5: 94 MADRONO 278. 1836. = Corethrogyne californica DC. [var.] leucophylla (Lindl. ex DC.) Kuntze, Rev. gen. pl. 1:330. 1891 [illegit., oldest sp. epithet not used]. = Corethrogyne leucophylla (Lindl. ex DC.) Jeps. [attributed to Menzies], Fl. w. Calif. 564. 1901. —Type: California, Monterey Co., near Monterey, A. Menzies s.n. (holotype: K!). Corethrogyne virgata Benth., Bot. voy. Sulphur 23. 1844. = Corethrogyne filaginifolia (Hook. & Arn.) Nutt. var. virgata (Benth.) A. Gray in S. Watson, Bot. Calif. 1:321. 1876. = Corethrogyne californica DC. [var.] virgata (Benth.) Kuntze, Rev. gen. pl. 1:330. 1891 [illegit., oldest sp. ep- ithet not used]. —Type: California, San Pedro, R. Hinds s.n. (holotype: K!). Corethrogyne incana (Lindl.) Nutt. var.? rigida Benth., Pl. hartweg. 316. 1849. = Corethrogyne filaginifolia (Hook. & Arn.) Nutt. var. rigida (Benth.) A. Gray, Synop. fl. N. Amer. 1(2):170. 1884. = Corethrogyne californica DC. [var.] rig- ida (Benth.) Kuntze, Rev. gen. pl. 1:330. 1891 [illegit., oldest sp. epithet not used]. = Corethro- gyne rigida (Benth.) A. Heller, Muhlenbergia 2: 256. 1906. —Type: California, Monterey Co., ‘In collibus arenosis juxta Monterey,” K. T. Hartweg “*1771(130)” (holotype: K!). Corethrogyne filaginifolia (Hook. & Arn.) Nutt. var. robusta Greene, Pittonia 1:89. 1887. —Lec- totype (here designated): California, Santa Bar- bara Co., San Miguel Island, Sep 1886, E. L. Greene s.n. (lectotype: CAS!). Corethrogyne viscidula Greene, Fl. francisc. 378. 1897. = Corethrogyne filaginifolia (Hook. & Arn.) Nutt. var. viscidula (Greene) D. D. Keck, Aliso 4:105. 1958. —Type: California, Monterey Co., ““Monterey,”’ 22 Jul 1888, C. C. Parry s.n. (holotype: ND-G!). Corethrogyne viscidula Greene var. greenei Jeps., Fl. w. Calif. 564. 1901. —Lectotype (here des- ignated): California, Alameda Co., ‘‘Niles,’’ 25 Jun 1896, W. L. Jepson ‘*14614” (ectotype: JEPS!). Corethrogyne virgata Benth. var. bernardina Abrams, Fl. Los Angeles 401. 1904. = Corethro- gyne filaginifolia (Hook. & Arn.) Nutt. var. ber- nardina (Abrams) H. M. Hall, Univ. Calif. Publ. Bot. 3:71. 1907. —Type: California, San Ber- nardino Co., Mentone, 10 Aug 1903, L. Abrams 2931 (holotype: DS!). Corethrogyne filaginifolia (Hook. & Arn.) Nutt. var. glomerata H. M. Hall, Univ. Calif. Publ. Bot. 3:72. 1907. —Type: California, San Ber- nardino Co., ““Oak Glen, Yucaipe Ranch, near Redlands,’’ Nov 1903, G. Robertson 117 (holo- type: UC!). Corethrogyne filaginifolia (Hook. & Arn.) Nutt. var. latifolia H. M. Hall, Univ. Calif. Publ. Bot. 3:70. 1907. —Type: California, Ventura Co., Ox- nard, 1901, J. B. Davy 7815 (holotype: UC!). Corethrogyne filaginifolia (Hook. & Arn.) Nutt. [Vol. 47 var. linifolia H. M. Hall, Univ. Calif. Publ. Bot. 3:71. 1907. = Corethrogyne linifolia (H. M. Hall) Ferris, Contr. Dudley Herb. 5:100. 1958. —Type: California, San Diego Co., ca. 1 km | south of Del Mar, 5 Aug 1906, K. Brandegee s.n. (holotype: UC!). Corethrogyne filaginifolia (Hook. & Arn.) Nutt. var. pacifica H. M. Hall, Univ. Calif. Publ. Bot. 3:73. 1907. —Type: California, San Diego Co., ‘Pacific Beach, near San Diego,” May—Oct | 1899, C. A. Purpus s.n. (holotype: UC!; isotype: US!). Corethrogyne brevicula Greene, Leafl. Bot. Ob- | serv. Crit. 2:26. 1910. = Corethrogyne filagini- — brevicula | folia (Hook. & Arn.) Nutt. var. (Greene) M. L. Canby, Bull. S. Calif. Acad. Sci. 26:12. 1927. tologue), C. R. Orcutt s.n. (holotype: US!). Corethrogyne flagellaris Greene, Leafl. Bot. Ob- serv. Crit. 2:26. 1910. —Type: California, Los Angeles Co., ““Along the seaboard at Redondo,”’ 25 May 1902, E. Braunton 280 (holotype: US!; isotype: DS!). Corethrogyne floccosa Greene, Leafl. Bot. Observ. Crit. 2:25. 1910. —Type: California, Santa Bar- | bara Co., “Elwood, near Santa Barbara,” 1908, A. Eastwood s.n. (holotype: US!). Corethrogyne lavandulacea Greene, Leafl. Bot. Observ. Crit. 2:27. 1910. —Type: California, Santa Catalina Island, Sep 1898, B. Trask s.n. (holotype: US!). Corethrogyne racemosa Greene, Leafl. Bot. Ob- sery.. Crit. 2:26. *‘Mountains of San Diego Co.,”’ Oct 1889, C. R. Orcutt s.n. (holotype: US!). Corethrogyne scabra Greene, Leafl. Bot. Observ. Crit. 2:25. 1910. —Type: California, Los Ange- les Co., Sep 1890, H. E. Hasse s.n. (holotype: US!). Sep Type: California, ‘“‘Mountains of | San Diego Co.,’’ Oct 1889 (label) or 1899 (pro- | 1910. —Type: California, | Corethrogyne sessilis Greene, Leafl. Bot. Observ. | Crit. 2:25. 1910. = (Hook. & Arn.) Nutt. var. sessilis (Greene) M. L. Canby, Bull. S. Calif. Acad. Sci. 26:15. 1927. Type: California, ““San Bernardino Mountains,”’ 23 Oct 1891, S. B. Parish 2233 (holotype: US!; isotype: UC!). Corethrogyne filaginifolia (Hook. & Arn.) Nutt. Corethrogyne filaginifolia | var. pinetorum I. M. Johnst., Bull. S. Calif. Acad. | Sci. 18:21. 1919. —Type: California, Los An- | geles Co., San Antonio Mts., Brown’s Flat, 1 Sep 1918, I. M. Johnston 2137 (holotype: POM!; iso- | type: DS!). Corethrogyne filaginifolia (Hook. & Arn.) Nutt. | var. peirsonii M. L. Canby, Bull. S. Calif. Acad. » Sci. 26:14. 1927. —Type: California, Los An- geles Co., Newhall, 7 Oct 1923, F. W. Peirson | 4159 (holotype: POM!; isotype: DS!). Corethrogyne filaginifolia (Hook. & Arn.) Nutt. | var. hamiltonensis D. D. Keck, Aliso 4:105. | 1958. —Type: California, Santa Clara Co., Mt. || \ | 2000] Hamilton, Aug 1914, H. M. Hall 9865 (holotype: lee NY '!). _ Primary stems predominately erect, mostly 1—10 dm long, usually branched distally. Leaf blades nar- -rowly spatulate to linear, entire or toothed. Heads (1-)5—20+ per floral stem. Involucres cylindric to turbinate, lengths (6-14 mm) mostly twice diame- ters (3-8 mm) at anthesis in living plants. Phyllaries mostly 4—9-seriate. Disc florets mostly 12—40, co- rollas mostly 4-6 mm long. Widespread through much of cismontane Cali- fornia from Placer Co. south through Sierra Nevada and in interior mountains of Contra Costa and Al- ameda cos. south through coast ranges to transverse ranges and Channel Islands and into northern Baja California. 2. Corethrogyne filaginifolia (Hook. & Arn.) Nutt. var. californica (DC.) Saroyan, comb. nov. = Corethrogyne californica DC., Prodr. 5:215. 1836. = Lessingia filaginifolia (Hook. & Arn.) M. A. Lane var. californica (DC.) M. A. Lane, Novon 2:213. 1992. —Type: “‘Nova California,” 1833, D. Douglas s.n. (holotype: G-DC, micro- fiche!; isotypes: BM!, K!). Corethrogyne obovata Benth., Bot. voy. Sulphur 22. 1844. = Corethrogyne californica DC. [var.] obovata (Benth.) Kuntze, Rev. gen. pl. 1:330. 1891 [illegit., oldest sp. epithet not used]. — Type: California, ““‘Bodegas,” 1841, R. Hinds s.n. (holotype: K!). Corethrogyne spathulata A. Gray, Proc. Amer. Acad. Arts 7:351. 1868. = Corethrogyne cali- fornica DC. [var.] spathulata (A. Gray) Kuntze, Rev. gen. pl. 1:330. 1891 [illegit., oldest sp. ep- ithet not used]. —Type: California, Humboldt Co., Shelter Cove, 1867, H. N. Bolander 6505 (holotype: GH!; isotypes: BM!, K!, US!). Corethrogyne caespitosa Greene, Fl. francisc. 378. 1897. —Type: California, San Mateo Co., Crys- tal Springs, 22 Jun 1886, E. L. Greene s.n. (ho- lotype: ND-G!; isotypes: PH!, US!). Corethrogyne californica DC. var. lyonii S. E Blake, J. Wash. Acad. Sci. 33:267. 1943. — Type: California, Merced Co., Cathedral Peak, 4 Jun 1941, G. S. Lyon 1572 (holotype: US!, for- merly in NA; isotypes: DS!, UC!). Primary stems predominately decumbent to as- cending, mostly less than 6 dm long, mostly un- branched. Leaf blades mostly obovate to spatulate, toothed. Heads mostly 1(—5) per floral stem. Invo- lucres hemispheric to campanulate, lengths (5—10 mm) mostly equalling diameters at anthesis in liv- ing plants. Phyllaries mostly 3—5-seriate. Disc flo- rets 30—120+, corollas mostly 6-8 mm long. Common as discrete, very local, usually dense populations from northern Monterey, San Benito, and western Merced cos. north through the North Coast Ranges of California into western Klamath Range in Coos Co., Oregon. SAROYAN ET AL.: CORETHROGYNE 95 QUESTIONABLE NAME AND EXCLUDED TAXA Corethrogyne californica DC. [var.] recurva Kun- tze, Rev. gen. pl. 1:330. 1891 [illegit., oldest sp. epithet not used]. —Type: We have not seen type material; the description, “‘Involucri bracteae ap- ice recurvae,’’ is insufficient to allow confident application of the name. Corethrogyne cana (A. Gray) Greene, Bull. Calif. Acad. Sci. 1(no. 4):223. 1885. = Diplostephium canum A. Gray, Proc. Amer. Acad. Arts 11:75. 1876. = HAZARDIA CANA (A. Gray) Greene, Pit- tonia 1:29. 1887. = Haplopappus canus (A. Gray) S. FE Blake, Contr. U.S. Natl. Herb. 24:86. 1922. —Type: Mexico, Baja California, Guada- lupe Island, 28 Mar 1875, E. Palmer s.n. (holo- type: GH). Corethrogyne detonsa Greene, Bull. Torrey Bot. Club 10:41. 1883. = HAZARDIA DETONSA (Greene) Greene, Pittonia 1:29. 1887. = Haplo- Pappus detonsus (Greene) P. H. Raven, Aliso 5: 343. 1963. —Type: Origin and collector un- known (holotype: CAS!). ACKNOWLEDGMENTS We thank B. Baldwin, T. Duncan, M. Lane, S. Markos, G. Nesom, R. Ornduff, R. Pimentel, R. Price, J. Semple, A. Smith, and S. Sundberg for sharing information and/or for helpful comments on earlier versions of this paper. We thank staff at BM, CAS, DS, E, G, GH, HAY, JEPS, K, ND, ND-G, NY, PH, POM, RSA, UC, and US for loans and/or for making specimens available for study. JPS ex- tends special thanks to Mary Lou Wilcox for guidance, support, and encouragement throughout the course of his thesis research. The original for Figure 1 was drawn by Randall E. Tribo. LITERATURE CITED FerRIS, R. S. 1960. Corethrogyne. Pp. 337-342 in R. S. Ferris, Illustrated flora of the Pacific States, vol. 4. Stanford University Press, Stanford. GOLDBLATT, P. AND D. E. JOHNSON. 1992. Index to chro- mosome numbers in plants. Monographs in System- atic Botany from the Missouri Botanical Garden. 40: Vi-vill,, 1-238; GRAY, A. 1876. Corethrogyne. Pp. 320-321 in S. Watson et al., Geological survey of California... Botany .. .. [Bot. Calif.], vol. 1. Welch, Bigelow, & Co., Univer- sity Press, Cambridge, Mass. JONES, D. T. 1977. Relationships of Lessingia Cham. and Corethrogyne DC. (Asteraceae). M.A. thesis, Califor- nia State University, Hayward. Keck, D. D. 1959. Corethrogyne. Pp. 1204—1207 in P. A. Munz, A California flora. University California Press, Berkeley. LANE, M. 1992. New combinations in Californian Lessin- gia (Compositae: Astereae). Novon 2:213. . 1993. Lessingia. Pp. 304-306 in J. D. Hickman (ed.), The Jepson manual. University California Press, Berkeley. SAROYAN, J. P. 1974. Variation in the genus Corethrogyne DC. (Astereae - Asteraceae). M.A. thesis, California State University, Hayward. 96 MADRONO [Vol. 47 APPENDIX 1. STUDY POPULATIONS OF CORETHROGYNE. Order of entries is north to south. Each entry follows the form: alphabetic identifier; latitude; longitude; elevation (m); distance from ocean (km); vegetation (grasslands, coastal scrub, chaparral, forest); geographic location (all in California); and average pollen stainability (number of plants sampled for | pollen stainability). Chromosome counts for one or more plants from each of the populations all yielded 2n = 5 II. | Voucher collections are in UC. A. 41°47’; 124°08’'; 75 m; 0.4 km; grasslands; Humboldt Co., 2.5 miles north of Patricks Point; 90%(4). B. 38°57’: 123°43'; 75 m; 0.04 km; grasslands; Mendocino Co., 500 yards east of Point Arena; 88%(7). C. 38°08’; 122°53’; 75 m; 4.8 km; grasslands; Marin Co., along Pierce Point Road, 1 mile south of Tomales Bay State Park; 87%(5). D. 38°05’; 122°45’; 200 m; 9.5 km; grasslands; Marin Co., on Inverness Ridge, 3 miles west of Inverness; 93%(7). E. 37°30’; 122°20'; 100 m; 11.25 km; grasslands; San Mateo Co., east bank of Upper Crystal Springs Reservoir 0.25 mile south of highway 92; 97%(11). F. 36°42’; 121°48’; 25 m; 0.4 km; grasslands; Monterey Co., 0.25 miles east of ocean, 1 mile south of animal shelter on dunes of Marina Beach; 93%(7). G. 36°38’; 121°46’; 125 m; 4.8 km; chaparral; Monterey Co., Fort Ord, break area of M-79 grenade range 83%(1). H. 36°37’; 121°56'; 10 m; 0.02 km; grasslands; Monterey Co., 1 mile north of limit of Asilomar Beach; 93%(13). I. 36°35’; 121°58’; 10 m; 0.03 km; grasslands; Monterey Co., opposite Seal Rock on 17-Mile Drive; 91%(2). J. 35°34’; 121°59'; 25 m; 0.03 km; forest; Monterey Co., opposite Cypress Point on 17-Mile Drive; 94%(9). K. 36°30’; 121°55'; 10 m; 0.02 km; grasslands; Monterey Co., vacant lot at Yankee Point, 1 mile south of Point Lobos; 80%(4). L. 36°23’; 121°30'; 40 m; 35.5 km; chaparral; Monterey Co., Hastings Reservation, along trail between bunk houses; 84%(5). M. 36°17’; 121°51'; 25 m; 0.8 km; grasslands; Monterey Co., along highway 1, 10 miles south of Point Lobos; 93%(5). N. 36°16’; 121°50’; 250 m; 0.8 km; coastal scrub; Monterey Co., coastal bluff, 0.25 mile west of highway 1, 15 miles south of Carmel; 92%(12). O. 35°52’; 121°27'; 300 m; 0.5 km; grasslands; Monterey Co., 1.2 miles north of Gorda; 94%(2). P. 35°39’; 121°14’; 10 m; 0.02 km; grasslands; San Luis Obispo Co., 3 miles north of San Simeon; 91%(3). Q. 34°56’; 120°38’; 150 m; 3 km; grasslands; Santa Barbara Co., 3 miles east of Point Sal; 85%(1). R. 34°35’; 120°25’; 100 m; 14.5 km; grasslands; Santa Barbara Co., 1 mile south of Lompoc, 0.5 mile west of highway 1; 98%(2). S. 34°13’; 117°12’; 2000 m; 100 km; forest; San Bernardino Co., 3 miles south of Lake Arrowhead, 0.5 miles west of highway 18; 86%(4). T. 33°00’; 117°15'; 200 m; 0.8 km; coastal scrub; San Diego Co., coastal bluff near Torrey Pines State Reserve; 90%(3). } Maprono, Vol. 47, No. 2, pp. 97-105, 2000 CORRESPONDENCE BETWEEN NI TOLERANCE AND HYPERACCUMULATION IN STREPTANTHUS (BRASSICACEAE) ROBERT S. BOYD, MICHAEL A. WALL, AND JAMES E. WATKINS, JR. Department of Biological Sciences and Alabama Agricultural Experiment Station, Auburn University, AL 36849-5407 ABSTRACT Nickel hyperaccumulation may be associated with increased Ni tolerance for some plant species that grow on serpentine soils. We contrasted the Ni tolerance of three species: a Ni hyperaccumulator (Strep- tanthus polygaloides A. Gray) endemic to serpentine soil, a congeneric non-hyperaccumulator also en- demic to serpentine soil (S. breweri A. Gray), and a species from the same family but not adapted to serpentine soil (Brassica oleracea L.). We assessed Ni tolerance by measuring germination and radicle elongation in test solutions varying in Nit? content. By both approaches, Ni tolerance was greatest for the hyperaccumulator, intermediate for the non-hyperaccumulator, and least for the unadapted species. A soil-based test of root elongation, using S. polygaloides and B. oleracea with two serpentine soils and one non-serpentine soil, showed a significant species-by-soil interaction. Root elongation of B. oleracea was inhibited in serpentine soil, whereas S. polygaloides showed reduced root elongation in non-serpentine soil. We concluded that these results are consistent with the hypothesis that Ni hyperaccumulation is a metal tolerance mechanism adopted by some species native to serpentine soils. These results also are consistent with other ecological functions of Ni hyperaccumulation, such as the elemental allelopathy or microsite tolerance hypotheses. Plant tissues vary widely in heavy metal concen- trations, although most plant species contain very low levels. Pais and Jones (1997) reported that spe- cies not adapted to high-Ni soils typically contain 0.3—-3.5 pg Ni/g dry wt. For these species, tissue Ni in the range 8—50 wg/g dry wt usually denotes a toxic Ni concentration (MacNicol and Beckett 1985). Serpentine soils often contain elevated levels of Ni (Kruckeberg 1984; Brooks 1987). Many plants native to these soils also contain elevated levels of Ni (Reeves 1992), often ranging from 10—100 pg Ni/g (Brooks 1987). Baker and Walker (1990) called these species “‘accumulators’’. A small pro- portion of plant species native to serpentine soils accumulate Ni to an extraordinary degree over a wide range of soil Ni concentrations (Morrison et al. 1980). Brooks et al. (1977) termed these plants “hyperaccumulators’’, defining them as containing at least 1000 pg Ni/g. Many workers have suggested that metal hyper- accumulation has an adaptive function. Metal hy- peraccumulators belong to a number of evolution- ary lines of dicotyledonous plants (Brooks 1987), Suggesting multiple independent evolution of this trait and therefore that metal hyperaccumulation has positive selective value. Boyd and Martens (1992) suggested four functions of metal hyperac- cumulation: metal tolerance, drought tolerance/ avoidance, defense against herbivores/pathogens, and interference with neighboring plants. To date most research has focused on defensive explana- tions (Boyd 1998). This work has shown that hy- peraccumulated metals can defend plants against herbivores (Boyd and Martens 1994; Martens and Boyd 1994; Pollard and Baker 1997; Sagner et al. 1998; Boyd and Moar 1999; Davis 1999; Jhee et al. 1999) and pathogens (Boyd et al. 1994; Ghad- erian et al. 2000). The remaining hypothesized functions of metal hyperaccumulation are relatively unexplored. Metal hyperaccumulation has been suggested to function as a mechanism for tolerating elevated soil metal contents (Boyd and Martens 1992). Metals could be removed from metabolically sensitive ar- eas of a plant’s cells or tissues by concentrating them in less sensitive locations (e.g., vacuoles, cell walls, epidermal cells, and trichomes). In some cases (e.g., Ernst 1972; Wild 1978), it has been pro- posed that abscission or loss of high-metal plant parts serves to dispose metals from the plant body. The difficulty with this “tolerance hypothesis” (sensu Boyd and Martens 1992) is that, although metal hyperaccumulators must surely be able to tol- erate very high tissue metal levels, there is no ev- idence that tolerance is an adaptive function of met- al hyperaccumulation. Thus, it is important to com- pare the metal tolerances of hyperaccumulator and non-hyperaccumulator species native to serpentine soils. If both have equivalent metal tolerance abil- ities, then this would be evidence contrary to the tolerance hypothesis. Increased metal tolerance of hyperaccumulator species relative to other species able to grow on serpentine soil is also crucial to the “interference hypothesis” (sensu Boyd and Martens 1992). Some authors (e.g., Gabbrielli et al. 1991; Baker et al. 1992) have suggested that the elevated metal con- tent of leaf litter produced by hyperaccumulators can lead to increased metal content of the surface 98 MADRONO soil beneath individual plant canopies. Less metal- tolerant species may be prevented from growing in these metal-enriched areas, resulting in lessened competition for the hyperaccumulator. Wilson and Agnew (1992) further suggested that metal hyper- accumulators might suppress the growth of less metal-tolerant species through surface soil metal enrichment and thus create areas dominated by rel- atively pure stands of the hyperaccumulator spe- cies. This interaction was termed ‘“‘elemental alle- lopathy”’ by Boyd and Martens (1998), due to the similarity of this phenomenon with allelopathy (Rice 1984). Boyd and Martens (1998) pointed out that experimental confirmation of elemental allelop- athy must demonstrate two facts. First, soil Ni lev- els must be significantly elevated in the vicinity of hyperaccumulator plants, relative to other micro- sites in serpentine habitats. Second, it must be dem- onstrated that hyperaccumulator species are more metal tolerant than potentially competing species. Thus, demonstration of differential metal tolerance between hyperaccumulator and non-hyperaccumu- lator serpentine species is one of the two principles essential to uphold the elemental allelopathy hy- pothesis. As explained above, both the tolerance hypoth- esis and the elemental allelopathy (interference) hy- pothesis require that metal hyperaccumulators be more metal-tolerant than other serpentine soil spe- cies. Some studies of hyperaccumulators (e.g., Kra- mer et al. 1997) have contrasted them with a con- gener that is not native to metalliferous soils and several studies have compared the metal tolerance of hyperaccumulator species with other species. The early work of Morrison et al. (1980) examined the Ni tolerances of seven Ni-hyperaccumulating species and two non-hyperaccumulating species of Alyssum. Root elongation tests showed the hyper- accumulators to be more Ni tolerant than non-hy- peraccumulators. Gabbrielli et al. (1990) used root elongation tests to contrast the Ni tolerance of the hyperaccumulator A. bertolonii Desv. with that of the non-hyperaccumulator serpentine species, Sile- ne italica L., and showed that the hyperaccumulator was much more Ni tolerant than the non-hyperac- cumulator. Homer et al. (1991) contrasted a Ni hy- peraccumulator (Alyssum troodii Boiss) and a non- hyperaccumulator (Alyssum saxatilis = Aurinia saxatilis (L.) Desv.), finding greater Ni tolerance for the hyperaccumulator by examining both bio- mass production and germination experiments. Kra- mer et al. (1996) obtained similar results, using a non-hyperaccumulating Alyssum species (A. mon- tanum) and contrasting its Ni tolerance with the Ni hyperaccumulator A. lesbiacum. Biomass produc- tion of the Ni hyperaccumulator was much greater than the non-hyperaccumulator when plants were grown in a Series of Ni-containing solutions. Shen et al. (1997) studied Zn tolerance of two serpentine soil species of Thlaspi by measuring biomass ac- cumulation as a function of Zn concentration in the [Vol. 47 growth medium. They found that the Zn hyperac- cumulator 7. caerulescens was more Zn tolerant than the non-hyperaccumulator T. ochroleucum. The research reported here was conducted to | compare the Ni tolerance of two annual species of | Streptanthus, both of which are endemic to serpen- | tine soils, but only one is a Ni hyperaccumulator. This New World genus in the Brassicaceae contains a single Ni hyperaccumulator species, S. polyga- loides A. Gray, and a number of non-hyperaccu- mulating taxa endemic to serpentine soils (Kruck- eberg 1984). Streptanthus polygaloides also is unique among hyperaccumulators due to its being an obligate annual. Other Ni hyperaccumulators, in- cluding those in Alyssum and Thlaspi, are peren- nials and thus the present work extends our under- standing of the relationship between hyperaccu- mulation and tolerance to annual species in another genus. METHODS Study species. Streptanthus polygaloides is an annual Ni hyperaccumulator endemic to serpentine soils in the western foothills of the Sierra Nevada in California (Munz and Keck 1968). Studies by Reeves et al. (1981) and Kruckeberg and Reeves (1995) have documented >1000 wg Ni/g dry wt in all parts of this species. We collected seeds from the Red Hills of Tuolumne County, California (Fa- vre 1987), approximating sample #6737 of Kruck- eberg and Reeves (1995). We selected S. breweri A. Gray to represent a non-hyperaccumulating species of Streptanthus. This species also is an annual and is also endemic to serpentine soils (Kruckeberg 1984). The ranges of these two species do not overlap because S. breweri is native to the Coast Ranges of California (Munz and Keck 1968). Analysis of leaves of this species by Kruckeberg and Reeves (1995) showed a maximum of 13 pg Ni/g. Specimens analyzed by Reeves et al. (1981) contained <9 wg Ni/g. We col- lected seeds from a population in Napa County, California, near sample #6742 of Kruckeberg and Reeves (1995). The third species used was broccoli, Brassica oleracea L. This species is also in the Brassicaceae, but is not adapted to serpentine soil. Seeds were obtained from a commercial source. Germination tests. Seeds were placed in 5-cm di- ameter petri dishes under 24-hr illumination at room temperature. Germination solutions were add- ed in sufficient quantity to completely cover seeds, and were replenished as needed during the experi- ments. Nickel concentrations used for seeds of S. polygaloides and S. breweri were 0, 4.3, 8.5, 13, 17, 26, 34, 43, 51, and 60 mmol/L. The experiment using B. oleracea seeds used fewer Ni*? concentra- tions, 0, 8.5, 13, 17, 34, and 51 mmol/L, because we assumed the Ni*? tolerance of this unadapted species would be easier to characterize. Nickel was i} || {| i | } } | i) } i}) i) } 1} | } 2000] supplied as NiCl, (Fisher Scientific). The use of NiCl, presented the possibility that inhibition could - occur due to increased Cl~! ion concentration rather | than Ni*t?. To test for this effect (as well as osmotic _ effects) on seed germination, solutions containing gl 2, 19, 25, 50, 75, 100, 125, 150, 175, 200, 225, and 250 mmol Ca*?/L (as CaCl, Fisher Scientific) _were used. Seeds of two species, S. polygaloides | _and B. oleracea, were used for Ca*? experiments. - Due to insufficient seed availability, these experi- ments were not conducted using S. breweri seeds. Seeds (ranging in number between 10 and 15) of each species were placed in each petri plate and monitored for germination (defined as emergence of the radicle from the seed coat) until germination declined to zero during a several-day period. Ten petri plates (replicates) were used for each combi- nation of species, Ni*?, and Ca*’ concentration. Germinating seeds were removed from petri plates to minimize their influence on the chemistry of the germination solutions. Percent germination was cal- culated for the seeds in each petri plate and these data were analyzed by one-way ANOVA for each species and each ion (Ni*? or Ca*’) used, after transformation (arcsine square root) so data would better fit ANOVA assumptions (Zar 1996). Fisher’s Protected Least Significant Difference (PLSD) test (Abacus Concepts 1992) was used for post-hoc means separation (a = 0.05). Root elongation tests in solution. Seeds of all three species were placed on moistened filter paper in petri dishes and allowed to germinate. Once rad- icles reached lengths of 1-3 mm, seedlings were removed and transferred to root elongation test so- lutions. At least three seedlings (up to five if more were available) were used in each petri plate. Three petri plates (replicates) of each species were used for each concentration of Ca‘? or Ni*?. Radicles of transplanted seedlings were allowed to elongate for 7 d at room temperature, after which they were re- moved from test solutions and the length of the primary root measured. All test solutions contained a background level of ions important for normal seedling development (Baker 1987). The background solution contained 0.77 mmol Ca*?/L as Ca(NO;),, 0.82 mmol Mg*?/ L as MgSO,, and 0.28 mmol K*!/L as KNO, (all obtained from Fisher Scientific). NiCl, was then added to the background solution to create concen- trations of 0, 0.085, 0.17, 0.34, 0.51, 0.68, 0.85, 1.4, and 1.7 mmol Ni*?/L. As in the germination experiment, the addition of Cl-' along with Ni*? presented the possibility that inhibition could occur due to increased Cl@! concentration rather than Ni*’. To test for this effect (and general osmotic effects), three Ca*? solutions were created with CaCl,, using the same _ back- ground solution used for the Nit? experiments, to create solutions containing 2.5, 5.0, or 10 mmol Ca*?/L. BOYD ET AL.: NI TOLERANCE AND HYPERACCUMULATION 99 Root lengths from seedlings in each petri plate were averaged, and the data analyzed by one-way ANOVA for each species and solution type (Ni*? or Ca*’) after log-transformation so data would bet- ter fit ANOVA assumptions (Zar 1996). Fisher’s PLSD test was used for post-hoc means separation (a <= 0.05). Root elongation tests in soil. This experiment de- termined if the tolerance results from Ni*? solutions had relevance to root performance in serpentine soils. Three soils were used for testing root elon- gation: two from serpentine sites in California and one from a non-serpentine site in Alabama. The first serpentine soil was collected from the Red Hills in Tuolumne County. The second was col- lected from a serpentine site along U.S. Highway 101 near the southern city limit of San Luis Obispo, San Luis Obispo County, California. The non-ser- pentine soil was collected from Auburn, Lee Coun- ty, Alabama. Approximately 1 L of soil was col- lected to 10 cm depth at each location and sieved to remove stones >2 mm dia. A subsample of each soil was used for-elemental analysis. Soil samples were double-acid extracted using 20 mL of extractant (0.05 M HCI1/0.025 M H,SO,) shaken with 5 g of dry soil for 5 min. The extract was analyzed for Ni using an atomic ab- sorption spectrophotometer (IL 251, Instrumenta- tion Laboratory, Franklin, MA), and for Ca, K, Mg, P, Cu, Fe, Mn, Cr, Pb, Co and Zn using an induc- tively-coupled argon plasma spectrometer (ICAP 9000, Jarrell-Ash, Franklin, MA). Soil from each location was put into small (5 cm diam.) petri plates to approx. 5 mm depth and moistened with deionized water. Seeds of two spe- cies, the unadapted B. oleracea and the hyperac- cumulator S. polygaloides, were used (not enough seeds of S. breweri were available). Seeds were germinated on moist filter paper. Germinating seeds (radicles 1-3 mm long) were transferred to the sur- face of the soil plates and allowed to grow for 4 d. Three seedlings were used for each replicate, and radicle lengths were averaged for each plate. Seed- lings of these two species differed in size, so that direct comparisons of root length were confounded by this factor. We minimized the influence of this innate size difference by relativizing mean root lengths for each species within a replicate. Means were expressed as a decimal fraction of the mean for that treatment which produced the largest mean root length. Mean root lengths for each of the three treatments were each divided by the largest value of those three data, resulting in a value of 1 for the largest mean and lesser values for the other two means. These relativized data were analyzed by two-way ANOVA, using transformed (arcsine square root) data to better meet the assumptions of ANOVA (Zar 1996). Soil collection site and spe- cies were main effect factors and the interaction term was included in the ANOVA model. Fisher’s 100 MADRONO Protected Least Significant Difference (PLSD) test (Abacus Concepts 1992) was used for post-hoc means separation (a = 0.05). RESULTS Germination tests. Germination of all three spe- cies was significantly affected by Ni*? concentra- tion. For S. polygaloides, ANOVA indicated a sig- nificant Ni*? effect (Foo. = 21, P < 0.0001). Ger- mination was relatively high (>50%) for even the most concentrated (60 mmol/L) solution (Fig. 1), and Fisher’s PLSD test indicated that germination was equivalent to that of the control for all solu- tions containing <17 mmol/L. Streptanthus breweri germination also was significantly affected by Ni*? (ANOVA: Fo. = 22, P < 0.0001). Germination was less than 50% for the higher Ni*? concentra- tions (>34 mmol/L), and a significant decline in germination relative to the control (Fisher’s PLSD test) was first observed at a lesser concentration than for S. polygaloides (13 mmol/L vs. 17 mmol/ L: Fig. 1). Brassica oleracea germination also de- clined significantly as Ni‘? concentration increased (ANOVA: F; 54 = 38, P < 0.0001). Germination declined to <50% at concentrations >8.5 mmol Ni*’/L, with Fisher’s PLSD test showing that ger- mination first declined significantly relative to the control at 8.5 mmol Ni*?/L. Experiments with Ca*? solutions also showed in- hibition of germination for all species (data not shown), but at greater concentrations than with Ni*? solutions. Germination of S. polygaloides was sig- nificantly affected by Ca*? concentration (ANOVA: F,.,;, = 110, P < 0.0001). Mean germination was >75% for those solutions containing up to 50 mmol Ca*t’/L, but declined significantly compared to the control (Fisher’s PLSD test), reaching 37% at 75 mmol/L. Germination of B. oleracea also was sig- nificantly affected by Ca*’ concentration (ANOVA: F,.;;, = 67, P < 0.0001). Mean germination was >72% for concentrations up to 75 mmol Ca‘*’/L, and then declined significantly relative to the con- trol (Fisher’s PLSD test) to 64% at 100 mmol/L. These results show that Cl~' did not produce the decreased germination observed with the Ni*? so- lutions, and that Ni*? played a significant role in decreasing seed germination in this experiment. Root elongation tests in solution. Solution Ni*? concentration significantly affected root elongation for all species tested. For S. polygaloides, ANOVA showed a significant Ni** effect (F,,; = 29, P < 0.0001). Mean root length declined significantly relative to the control at 0.085 mmol Ni*t?/L (Fish- er’s PLSD test, Fig. 2). Maximum inhibition (small- est mean root elongation) was observed at the high- est Ni*? concentration used (1.7 mmol Ni*?/L). Comparing means to that value, the lowest concen- tration of Ni*’ that resulted in maximum inhibition was 0.85 mmol/L (Fisher’s PLSD test). This value begins the maximum inhibition zone noted on Fig- [Vol. 47 | Streptanthus polygaloides 6 100 e 80 = = 60 ro) ~ ‘40 c ro) O 20 o oO 6 100 e 80 = = 760 o>) - 40 eS ® = 20 Co) sees 0 4.3 8.51317 26 34 43 51 60 100 c Brassica oleracea = 80 O = E 60 oO o>) _ 40 Cc ® 2 20 1) oO 0 0 8.5.13) dia r Sate, 54 Solution Ni*+? content (mmol/L) Fic. 1. Percent germination of the three experimental : species as influenced by Nit? content of the germination | test solution. Note that the x-axes are not linear. The | hatched bars in each graph indicate treatments for which | germination was lowered significantly from that of the | control (O mmol Ni*?/L) treatment. Error bars denote the | upper 95% confidence limit of each mean. ure 2, and represents the lowest Nit‘? level that — causes maximum root growth inhibition. Root growth of S. breweri also declined with increasing Ni‘? concentration (ANOVA: F,,; = 84, P < 0.0001). Again, means declined significantly rela- _ tive to the control at 0.085 mmol/L (Fisher’s PLSD _ test, Fig. 2). In this case, however, the maximum © inhibition zone began at a lesser Ni concentration than for S. polygaloides. For S. breweri, the maxi- mum inhibition zone started at 0.51 mmol Ni*?/L S. polygaloides Critical concentration 0.85 0 0.17 0.51 ie 0.085 0.34 0.68 1.4 e S. breweri oO <= 2 Critical cab) : = concentration re) Q (= (40) ® = : 0 0.17 0.51 0.85 a7 0.085 0.34 0.68 1.4 a B. oleracea co ies Critical e concentration io) © = © ® 7 0 0 0.17 0.51 0.85 1.7 0.085 0.34 0.68 1.4 Solution Nit? content (mmol/L) Fic. 2. Mean root lengths (cm) of seedlings of the three experimental species as affected by solution Ni‘? concen- tration. The hatched bars denote treatments that resulted in maximum inhibition of root elongation (all hatched bars are not statistically different from the 1.7 mmol Nit?/L solution, using Fisher’s PLSD test). Error bars denote the upper 95% confidence limit of each mean. (Fisher’s PLSD test comparison with mean value at 1.7 mmol/L, Fig. 2). Root growth of B. oleracea also was significantly depressed by Ni*? solutions (ANOVA: F,,, = 4.3, P = 0.0021). Mean root elongation again declined significantly at 0.085 mmol Ni*?/L (Fisher’s PLSD test, Fig. 2), but the maximum inhibition zone began at the lowest Ni concentration for all three species tested. For B. oleracea, the maximum inhibition zone extended BOYD ET AL.: NI TOLERANCE AND HYPERACCUMULATION 101 S. polygaloides Mean root length (cm) Mean root length (cm) B. oleracea Mean root length (cm) 0 ' 5.0 10 Solution Cat? content (mmol/L) Fic. 3. Mean root lengths (cm) of seedlings of the three experimental species as affected by solution Ca*? concen- tration. Error bars denote the upper 95% confidence limit of each mean. from 0.17—1.7 mmol/L (Fisher’s PLSD test com- parison with mean at 1.7 mmol/L, Fig. 2). Effects of the Ni*? solutions on root elongation did not result from Cl~' or overall osmotic concen- trations of the test solutions. For all species tested, root elongation in CaCl, solutions was not signifi- cantly depressed relative to control solutions at the highest concentration tested (10 mmol/L, Fig. 3). The Cl~' concentration of the 10 mmol Ca*’/L so- lution was much higher than that of the highest (1.7 mmol/L) Nit? solution. Thus, a Cl-! effect cannot explain the significant declines of root elongation observed for all species tested with the 0.085 mmol Ni*t’/L solution. However, a significantly positive 102 [Serpentine 1 Serpentine 2 1.0 Nonserpentine Cc £ 0.8 O od) 6 Th nS 8 - 0.4 ® = 5 0.2 cc 0 Brassica Streptanthus oleracea polygaloides Fic. 4. Mean relative root elongation of B. oleracea and S. polygaloides on three soil treatments (two serpentine and one non-serpentine soil). Error bars denote the upper 95% confidence limit of each mean. effect of Ca*’ was detected for two of the three species. Streptanthus breweri was the one species that lacked a significant Ca*’ effect (ANOVA: F; 5. ps Fame 0.0803). Streptanthus polygaloides showed a significant Ca*? effect (ANOVA: F,>5. = 8.2, P = 0.0004). Fisher’s PLSD test revealed that root elongation for the 2.5 mmol Ca‘*’/L solution was greater than for any other test solution. Bras- sica oleracea also showed a significant effect (ANOVA: F;,, = 12, P = 0.006). In this case, Fish- er’s PLSD test showed that root elongation for the control treatment was significantly lower than for the solutions with elevated Ca*?. In all cases, a sig- nificant decline in root elongation (either due to Cl-' or osmotic effects) was not observed with the Ca*? solutions used. Thus, we conclude that the in- hibitory effects of the Ni*? solutions were attrib- utable to Ni*?’. Root elongation tests in soil. Relativized root elongation values were not significantly affected by either of the two main effect factors (soil and spe- cies) in the ANOVA (for soil: F,7, = 0.40, P = 0.67, and for species: F, 7, = 0.86, P = 0.36). How- ever, the interaction term was highly significant (F,7, = 21, P < 0.0001). Inspection of mean rela- tive root elongation values revealed that the two species reacted in opposite ways to the soils (Fig. 4). Brassica oleracea root elongation was greatest in the non-serpentine soil (mean = 0.78), and less in both of the serpentine soils (Serpentine Soil 1 mean = 0.32; Serpentine Soil 2 mean 0.55). Fisher’s PLSD test showed that the mean for roots from the non-serpentine soil was significantly greater than for roots from Serpentine Soil 1, but the other pairwise comparisons were only margin- ally significant (P = 0.072 for each). On the other MADRONO [Vol. 47 TABLE 1. ELEMENTAL ANALYSIS OF THE THREE SOILS USED IN THE SOIL ROOT ELONGATION TESTS. Serpentine Soil 1 was collected from the Red Hills, Tuolumne County, Cal- ifornia, Serpentine Soil 2 was collected from San Luis Obispo County, California, and the Non-serpentine Soil | came from Auburn, Lee County, Alabama. Parameter Serpentine Serpentine Non-serpentine (ug g') Soil 1 Soil 2 Soil Ca 200 1960 484 K S32 15.6 16.8 Mg 493 742 59.0 P 5.10 17.8 60.2 Cu 0.23 Oot 1.41 Mn 27.6 56.1 20.4 Zn 0.92 4.82 53.0 Co | ge 4.47 0.14 Cr O:12 WA 0.085 Pb 0.443 11.4 12.8 Ni 80 128 <4 Ca/Mg ratio 0.41 2:6 82 hand, S. polygaloides showed depressed root elon- gation in the non-serpentine soil (mean = 0.29) and higher root elongation in both serpentine soils (Ser- | pentine Soil 1 mean = 0.83; Serpentine Soil 2 mean | = (0.73). Fisher’s PLSD test showed that the mean | for roots from non-serpentine soil was significantly less than for both serpentine soils (P < 0.0001 for | both comparisons) and that the means for roots from the serpentine soils did not differ from each other (P = 0.31). Elemental analysis of the soils used in the above | experiment showed several differences between the serpentine soils and the non-serpentine soil (Table 1). Notable were the elevated Ni levels in the two | serpentine soils, along with the lower Ca/Mg ratios | for those soils relative to the non-serpentine soil. Also, Serpentine Soil 2 had an unusually high Ca | content, giving it a higher Ca/Mg ratio than is usual | for serpentine soils (Brooks 1987). DISCUSSION Our experiments showed that the Ni hyperaccu- | mulator species, S. polygaloides, was more Ni tol- | erant than either the congeneric serpentine soil spe- — cies or the unadapted species. This contrast was | consistent for both the germination and the root | elongation experiments. This finding for Streptan- — thus is consistent with earlier work using species | of Alyssum and Thlaspi. Both Morrison et al. | (1980) and Kramer et al. (1996) reported greater Ni | tolerance, as measured by root elongation or bio- | mass comparisons, for Ni-hyperaccumulating Alys- sum species relative to non-hyperaccumulators that grew on Serpentine soil. Shen et al. (1997) reported | greater Ni tolerance in T. caerulescens relative to the non-hyperaccumulating serpentine species 7. ochroleucum. Thus, we can extend the generality of the correlation between Ni hyperaccumulation — ability and enhanced Ni tolerance to yet another i | || 2000] genus within the Brassicaceae, in this case to in- ‘clude an annual hyperaccumulating species. We ‘hope that additional tests of this hypothesis may be ‘undertaken using congeneric taxa from another family. The recent discovery by Reeves et al. (1996; 1999) of a large number of Ni hyperaccu- mulators from Cuba, many within the genera Phyl- _lanthus and Leucocroton (Euphorbiaceae) and Bux- us (Buxaceae), provides an excellent opportunity for such research. These genera would also allow extension of these questions to include woody (shrub, tree) growth forms. As pointed out earlier, equivalent metal tolerance between hyperaccumulator and non-hyperaccumu- lator species from serpentine soils would constitute evidence contrary to both the tolerance and ele- mental allelopathy hypotheses. However, our re- sults, plus those reported earlier (Morrison et al. 1980; Gabbrielli et al. 1990; Homer et al. 1991; Kramer et al. 1997; Shen et al. 1997), showed greater metal tolerance by the hyperaccumulator species studied. These results are consistent with both the tolerance and the elemental allelopathy hy- potheses. More information is needed to decide whether either of these hypotheses provides an evolutionary rationale for metal hyperaccumulation. For exam- ple, the elemental allelopathy hypothesis requires that soil Ni levels under hyperaccumulating plants are significantly higher than in other microsites. If this occurs, then the lesser Ni tolerance of co-oc- curring plant species might put them at a compet- itive disadvantage relative to hyperaccumulating species. Our (and other) Ni tolerance tests indicate that elemental allelopathy may indeed provide an adaptive rationale for metal hyperaccumulation. Unfortunately, little information on the microdis- tribution of soil Ni around hyperaccumulator plants is available. Two preliminary reports (Baker et al. 1992; Schlegel et al. 1992) indicated that localized Ni enrichment might occur. A third study (Boyd and Jaffré in review) also documented significantly greater surface soil Ni concentrations under cano- pies of the Ni-hyperaccumulating tree, Sebertia ac- uminata Pierre ex Baillon, compared to surface soil under the canopies of nearby non-hyperaccumula- tor tree species. In contrast, a study by Boyd et al. (1999), using the New Caledonian Ni hyperaccu- mulating shrub Psychotria douarrei (Beauvis.) Daniker, provided information contrary to this hy- pothesis. They analyzed soil Ni content under shrubs ranging in size from saplings to full-sized adults. No correlation of soil Ni with shrub size was detected, indicating that Ni enrichment was not oc- curring in that case. We also point out that enhanced tolerance of soil metals by hyperaccumulators may result in another ecological advantage apart from the elemental al- lelopathy hypothesis discussed above. Greater met- al tolerance would allow hyperaccumulators to ex- ploit relatively high-metal soil microsites that might BOYD ET AL.: NI TOLERANCE AND HYPERACCUMULATION 103 exist on serpentine sites. If other serpentine soil species are unable to grow (or are unable to grow well) in these microsites, hyperaccumulators might avoid competition for soil water/nutrients and thus gain a survival advantage. We call this the “‘mi- crosite tolerance’? hypothesis, to separate it from the elemental allelopathy hypothesis discussed pre- viously. Even for an annual species like S. polygaloides, enhanced Ni tolerance could allow colonization of relatively high-Ni microsites that are not exploited by other species, resulting in a survival advantage. The variability of metal levels within and between serpentine soil sites has been noted before (Kruck- eberg 1984), including recent studies on serpentine soils from California (Nicks and Chambers 1998). Field investigations of the relationship between S. polygaloides density and soil Ni levels would pro- vide evidence pertinent to this hypothesis, but to our knowledge such studies have not been con- ducted. We should note that the microsite tolerance function of hyperaccumulation might intergrade with elemental allelopathy. For example, seedlings of a metal hyperaccumulator that become estab- lished on a microsite containing elevated soil met- als might, over time, further elevate surface soil metal content and thus extend the boundaries of the high-metal microsite. The concentration of metals into surface soil, along with the expanded area of the microsite, could then provide a survival advan- tage via elemental allelopathy. The tolerance hypothesis is more difficult to completely falsify. Hyperaccumulators must, by def- inition, possess the ability to tolerate high tissue metal levels. Therefore, demonstration of greater Ni tolerance by hyperaccumulators does not contradict this hypothesis but does not provide definitive ev- idence that tolerance is an adaptive function of hy- peraccumulation. Our experimental result, demon- strating greater Ni tolerance of S. polygaloides rel- ative to S. breweri, is consistent with the tolerance hypothesis. However, several authors have suggest- ed that tolerance and hyperaccumulation are not strongly linked traits. Ingrouille and Smirnoff (1986) first suggested this for Thlaspi caerulescens, stating that Zn tolerance and Zn hyperaccumulation in that species may be independently inherited. Meerts and Van Isacker (1997) compared Zn hy- peraccumulation and tolerance among populations of Thlaspi caerulescens collected from high- and low-metal soil sites. They found that the low-metal populations were able to hyperaccumulate Zn to a greater extent, but the high-metal populations were more Zn tolerant. We suggest that other approaches can more clear- ly address this question. Perhaps the creation of non-hyperaccumulating mutants of a metal hyper- accumulator species, that can then be used to com- pare metal tolerances, can provide another way to test this hypothesis. Until that point, the tolerance hypothesis must be regarded as a possibly viable 104 explanation for the adaptive value of metal hyper- accumulation. Our soil-based experiment showed that solution- based root elongation tests may mirror the effects of Ni in soils, but suggested that factors other than Ni content may affect root elongation in serpentine soils. That Brassica root elongation was less in ser- pentine soils was not surprising, as this species does not possess adaptations that allow it to tolerate ser- pentine soil conditions. The result of the soil-based experiment with S. polygaloides (greater elongation in serpentine soils) was unexpected. One explana- tion for this result might be that S. polygaloides requires elevated soil Ni for optimum root growth. A higher metal requirement for hyperaccumulators has been suggested by some experiments (e.g., Shen et al. 1997) but not others (e.g., Morrison et al. 1980, Kramer et al. 1996). Results of our root elongation experiment do not show a Ni*? require- ment for S. polygaloides, as all species we tested showed a decline in root elongation at the lowest Ni*? level tested (0.085 mmol/L) relative to the control solution. In contrast, Nicks and Chambers (1995) reported lower biomass for S. polygaloides individuals grown in very low-Ni*? nutrient solu- tions, and suggested that some level of Ni*? in the growth medium was needed for optimum growth. We have noted mixed results in our own studies with Streptanthus polygaloides. In one experiment (Martens and Boyd 1994), plants grown on high- Ni greenhouse soil had less biomass than plants grown on low-Ni greenhouse soil. A second exper- iment (Boyd et al. 1994) showed the opposite re- sult. A second explanation may involve interactions between a hyperaccumulator and the soil microflo- ra. For S. polygaloides, this could be a positive (in the serpentine soil) or a negative (in the non-ser- pentine soil) interaction that could produce the dif- ference in root growth that we observed. Soil pathogens have been reported to limit plant growth when serpentine soil species are grown in low-Ni soil. For example, Tadros (1957) reported that soil- borne pathogens caused damping-off of seedlings of a serpentine soil non-hyperaccumulator Emmen- anthe species when seedlings were grown on non- serpentine soil. Also, Brooks (1987) reported that Ni hyperaccumulators in the genus Alyssum could be difficult to grow on low-Ni soil due to apparent pathogen attack. Certainly, it seems likely that the performance of serpentine soil species on their na- tive soils will be affected by similar organismal in- teractions, and that these interactions might be af- fected by the elevated metal contents of hyperac- cumulators. The possible consequences of these hy- peraccumulator/soil microflora interactions are only now being articulated (e.g., Boyd and Martens 1998). ACKNOWLEDGMENTS We thank M. Davis, R. Dute, D. Folkerts, and R. Locy for constructive comments on an earlier version of this MADRONO manuscript. This paper is Alabama Agricultural Experi- ment Station Journal No. 6-996053. LITERATURE CITED ABACUS CONCEPTS. 1992. StatView. Abacus Concepts, | Inc., Berkeley, CA. BAKER, A. J. M. 1987. Metal tolerance. Pp. 93-111 in I. H. Rorison, J. P. Grime, R. Hunt, G. A. E Hendry, | and D. H. Lewis (eds.), Frontiers of comparative plant | ecology. Academic Press, London. BAKER, A. J. M., J. PRocror, M. M. J. VAN BALGOOY, AND © R. D. REEVES. 1992. Hyperaccumulation of nickel by | the flora of the ultramafics of Palawan, Republic of | the Philippines. Pp. 291-304 in A. J. M. Baker, J. | Proctor, and R. D. Reeves (eds.), The vegetation of ultramafic (serpentine) soils. Intercept, Andover. BAKER, A. J. M. AND P. L. WALKER. 1990. Ecophysiology | of metal uptake by tolerant plants. Pp. 155-177 in A. | J. Shaw (ed.), Heavy metal tolerance in plants: Evo- lutionary aspects. CRC Press, Boca Raton. Boyp, R. S. 1998. Hyperaccumulation as a plant defensive | strategy. Pp. 181—201 in R. R. Brooks (ed.), Plants that hyperaccumulate heavy metals. CAB Internation- | al, Oxford. Boyp, R. S., T. JAFFRE, AND J. ODOM. 1999. Variation in | nickel content of the nickel-hyperaccumulating shrub | Psychotria douarrei (Rubiaceae) from New Caledo- . nia. Biotropica 31:403—410. Boyp, R. S. AND T. JAFFRE. In review. Phytoenrichment of soil Ni content by Sebertia acuminata in New Cale- donia and the concept of elemental allelopathy. South | African Journal of Science. Boyb, R. S. AND S. N. MARTENS. 1992. The raison d’etre for metal hyperaccumulation by plants. Pp. 279-289 | in A. J. M. Baker, J. Proctor, and R. D. Reeves (eds.), The vegetation of ultramafic (serpentine) soils. Inter- | cept, Andover. Boyp, R. S. AND S. N. MARTENS. 1994. Nickel hyperac- | cumulated by Thlaspi montanum var. montanum is | acutely toxic to an insect herbivore. Oikos 70:21—25. Boyp, R. S. AND S. N. MARTENS. 1998. The significance of metal hyperaccumulation for biotic interactions. Chemoecology 8:1-7. [Vol. 47 | Boyp, R. S. AND W. J. Moar. 1999. The defensive func- | tion of Ni in plants: Response of the polyphagous | herbivore Spodoptera exigua (Lepidoptera: Noctui- | dae) to hyperaccumulator and accumulator species of Streptanthus (Brassicaceae). Oecologia 118:218—224. Boyp, R. S., J. SHAW, AND S. N. MARTENS. 1994. Nickel | hyperaccumulation defends Streptanthus polygalo- ides (Brassicaceae) against pathogens. American | Journal of Botany 81:294—300. Brooks, R. R. 1987. Serpentine and its vegetation: a mul- tidisciplinary approach. Dioscorides Press, Portland, OK. Brooks, R. R., J. LEE, R. D. REEVES, AND T. JAFFRE. 1977. Detection of nickeliferous rocks by analysis of her- barium specimens of indicator plants. Journal of Geo- chemical Exploration 7:49—57. Davis, M. A. 1999. Characterization of the metal-based defense of Ni-hyperaccumulating plants. Ph.D. dis- sertation. Auburn University, Auburn, Alabama. ERNST, W. H. O. 1972. Ecophysiological studies on heavy metal plants in South Central Africa. Kirkia 8:125— 145. Favre, R. M. 1987. A management plan for rare plants in the Red Hills of Tuolumne County, California. Pp. } t | 2000] | | | } 425—427 in T. S. Elias (ed.), Conservation and man- agement of rare and endangered plants. California Native Plant Society, Sacramento. | GABBRIELLI, R., T. PANDOLFINI, O. VERGNANO, AND M. R. PALANDRI. 1990. Comparison of two serpentine spe- cies with different nickel tolerance strategies. Plant and Soil 122:271-277. GABBRIELLI, R., C. MATTIONI, AND O. VERGNANO. 1991. Accumulation mechanisms and heavy metal tolerance of a nickel hyperaccumulator. Journal of Plant Nutri- tion 14:1067—1080. GHADERIAN, S. M., A. J. E. LYON, AND A. J. M. BAKER. 2000. Seedling mortality of metal hyperaccumulator plants resulting from damping-off by Pythium spp. New Phytologist 146:219—224. Homer, E A., R. S. Morrison, R. R. BROOKS, J. CLEMENS, AND R. D. REEVES. 1991. Comparative studies of nickel, cobalt and copper uptake by some nickel hy- peraccumulators of the genus Alyssum. Plant and Soil 138:195—205. INGROUILLE, M. J. AND N. SMIRNOFF. 1986. Thlaspi cae- rulescens J & C Presl (T. alpestre L.) in Britain. New Phytologist 102:219—233. JHEE, E. M., K. L. DANDRIDGE, A. M. CHRISTY JR., AND A. J. POLLARD. 1999. Selective herbivory on low-zinc phenotypes of the hyperaccumulator Thlaspi caeru- lescens (Brassicaceae). Chemoecology 9:93—95. KRAMER, U., J. D. COTTER-HOWELLS, J. M. CHARNOCK, A. J. M. BAKER, AND J. A. C. SMITH. 1996. Free histidine as a metal chelator in plants that accumulate nickel. Nature 379:635—638. KRAMER, U., R. D. SMITH, W. W. WENZEL, I. RASKIN, AND D. E. SALT. 1997. The role of metal transport and tolerance in nickel hyperaccumulation by Thlaspi goesingense Halacsy. Plant Physiology 115:1641— 1650. KRUCKEBERG, A. R. 1984. California serpentines: flora, vegetation, geology, soils, and management prob- lems. University of California Press, Berkeley. KRUCKEBERG, A. R. AND R. D. REEVES. 1995. Nickel ac- cumulation by serpentine species of Streptanthus (Brassicaceae): field and greenhouse studies. Madro- no 42:458—469. MACNICcOL, R. D. AND P. H. T. BECKETT. 1985. Critical tissue concentrations of potentially toxic elements. Plant and Soil 85:107—129. MARTENS, S. N. AND R. S. Boypb. 1994. The ecological significance of nickel hyperaccumulation: a plant chemical defense. Oecologia 98:379—384. MEERTS, P. AND N. VAN ISACKER. 1997. Heavy metal tol- erance and accumulation in metallicolous and non- metallicolous populations of Thlaspi caerulescens from continental Europe. Plant Ecology 133:221— 23): Morrison, R. S., R. R. BROOKS, AND R. D. REEVES. 1980. Nickel uptake by Alyssum species. Plant Science Let- ters 17:451—457. BOYD ET AL.: NI TOLERANCE AND HYPERACCUMULATION 105 Munz, P. A. AND D. D. Keck. 1968. A California flora. University of California Press, Berkeley. Nicks, L. J. AND M. F CHAMBERS. 1995. Farming for met- als? Mining Environmental Management 3:15-—18. Nicks, L. J. AND M. EF CHAMBERS. 1998. A pioneering study of the potential of phytomining for nickel. Pp. 313-325 in R. R. Brooks (ed.), Plants that hyperac- cumulate heavy metals. CAB International, Oxford. Pais, I. AND J. B. JONES JR. 1997. The handbook of trace elements. St Lucie Press, Boca Raton, FL. POLLARD, A. J. AND A. J. M. BAKER. 1997. Deterrence of herbivory by zinc hyperaccumulation in Thlaspi cae- rulescens (Brassicaceae). New Phytologist 132:113— 118. REEVES, R. D. 1992. The hyperaccumulation of nickel by serpentine plants. Pp. 253-277 in A. J. M. Baker, J. Proctor, and R. D. Reeves (eds.), The vegetation of ultramafic (serpentine) soils. Intercept, Andover, MA. REEVES, R. D., A. J. M. BAKER, A. BORHIDI, AND R. BER- EZAIN. 1996. Nickel-accumulating plants from the an- cient soils of Cuba. New Phytologist 133:217—224. REEVES, R. D., A. J. M. BAKER, A. BORHIDI, AND R. BER- EZAIN. 1999, Nickel hyperaccumulation in the serpen- tine flora of Cuba. Annals of Botany 83:29—38. REEVES, R. D., R. R. BROOKS, AND R. M. MACFARLANE. 1981. Nickel uptake by Californian Streptanthus and Caulanthus with particular reference to the hyperac- cumulator S. polygaloides Gray (Brassicaceae). American Journal of Botany 68:708-—712. Rick, E. L. 1984. Allelopathy. Academic Press, Orlando, FL. SAGNER, S., R. KNEER, G. WANNER, J.-P. COSSON, B. DEUus- NEUMANN, AND M. H. ZENk. 1998. Hyperaccumula- tion, complexation and distribution of nickel in Se- bertia acuminata. Phytochemistry 47:339—347. SCHLEGEL, H. G., M. MEYER, T: SCHMIDT, R. D. STOPPEL, AND M. PICKHARDT. 1992. A community of nickel- resistant bacteria under nickel-hyperaccumulating plants. Pp. 305-317 in A. J. M. Baker, J. Proctor, and R. D. Reeves (eds.), The vegetation of ultramafic (ser- pentine) soils. Intercept, Andover, MA. SHEN, A. G., E J. ZHAO, AND S. P- MCGRATH. 1997. Uptake and transport of zinc in the hyperaccumulator Thlaspi caerulescens and the non-hyperaccumulator Thlaspi ochroleucum. Plant, Cell and Environment 20:898— 906. TaApDROS, T. M. 1957. Evidence of the presence of an eda- pho-biotic factor in the problem of serpentine toler- ance. Ecology 38:14—23. WILD, H. 1978. The vegetation of heavy metal and other toxic soils. Pp. 1301—1332 in M. J. A. Werger (ed.), Biogeography and ecology of southern Africa. Junk, The Hague. WILSON, J. B. AND A. D. Q. AGNEW. 1992. Positive-feed- back switches in plant communities. Advances in Ecological Research 23:263—336. ZAR, J. H. 1996. Biostatistical analysis. Prentice-Hall, En- glewood Cliffs, NJ. MApRONO, Vol. 47, No. 2, pp. 106-108, 2000 ANNUAL VARIATION IN XYLEM WATER POTENTIAL IN CALIFORNIA OAKS JOHANNES M. H. KNops School of Biological Sciences, University of Nebraska, 348 Manter Hall, Lincoln, NE 68588 WALTER D. KOENIG Hastings Natural History Reservation, University of California, Berkeley, 38601 E. Carmel Valley Road, Carmel Valley, CA 93924 ABSTRACT We measured late-summer predawn, daytime and the overnight recovery of xylem water potential for six years in Quercus lobata Nee, Q. douglasii Hook. & Arn. and Q. agrifolia Nee. Predawn xylem water potential was positively correlated with the rainfall in the previous year, indicating that low rainfall years are experienced as dry years by these oaks. Quercus douglasii had consistently lower xylem water po- tential than the other two species. Predawn values were consistently different among individuals trees and species, but the daytime and recovery values converged in the wet years. These results indicate that one- time measurements of predawn xylem water potential are a good indicator of species and individual tree differences in access to soil moisture. The pressure bomb technique is an easy, reliable method of measuring xylem water potential of trees in the field (Koide et al. 1991; Ritchie and Hinckley 1975; Turner 1981) and is widely used as an indi- cation of water stress of individuals (Callaway et al. 1991; DeLucia et al. 1988; Donovan and Ehler- inger 1994; Knops and Koenig 1994; Kolb and Da- vis 1994; Stringer et al. 1989). Implicit assumptions of these studies are that inter-annual variation in environmental conditions have a minimal effect on either the ranking of individuals or species-specific water stress such that differences in water stress are consistent from year to year. We tested this as- sumption in three species or California oaks by measuring xylem water potential in the same trees in six different years. Based on the large number of trees measured during the first year of the study, we previously reported differences among the species: Quercus douglasii Hook. & Arn. (blue oak), a winter decid- uous species, exhibited low xylem water potential values, indicative of little access to ground water; Q. lobata Nee (valley oak), a second winter decid- uous species, exhibited significantly higher values indicating good access to ground water; and Q. agrifolia Nee (coast live oak), an evergreen species with high daytime and predawn xylem water po- tential values, indicating limited transpiration dur- ing the dry part of the year (Knops and Koenig 1994). METHODS This study was conducted at the Hastings Natural History Reservation in central coastal California. The landscape consists of a mosaic of Mediterra- nean grasslands, oak woodlands, chaparral and ri- parian areas (Griffin 1988; Knops et al. 1995). Rainfall was measured daily and we used the an- nual total from September 1 of the previous year through August 30 of the current year. We measured xylem water potential of 14 Q. agrifolia, 13 Q. douglasii and 13 Q. lobata trees. These trees were selected from 250 trees of five species that are part of a long-term study examining acorn productivity (Knops and Koenig 1994, 1997; Koenig et al. 1994, 1996). Trees were selected to represent a gradient in acorn productivity within each species and are located throughout Hastings over a distance of approximately 3.5 km. Xylem water potential was measured using a pressure bomb (PMS Instruments Co.) in Septem- ber of each year, the end of the dry season when temperatures are hottest and water stress the great- est (Knops and Koenig 1994). Daytime measure- ment were made between 1300 and 1700, whereas predawn measurements on the same trees were made on the subsequent night between 0200 and 0600. Recovery was calculated as daytime minus predawn. Shoots with a minimum of three leaves and approximately 5 cm long were cut and imme- diately measured in the field. Shoots were not bagged, because we found no differences with sam- ples pre-bagged in plastic. All daytime shoots were cut in direct sunlight. We measured 2 shoots per tree, except if the values were more than 10% dif- ferent, in which case we measured a third shoot. Measurements were taken in 1991 and 1994 through 1998. In each year trees were measured at the end of the dry season in September. RESULTS There was a significant positive relationship be- tween the amount of rainfall and the predawn xy- | | 2000] | | a bon Staph amen cies ea sl a Pre-Dawn ay a 051 ‘a —_— = -1.0 — 5 OQ Oo -1.5 — s | 3 Q. agrifolia 5 -2.0 OQ. douglasii = Q. lobata * 2 ° Xylem water potential (MPa) Recovery Xylem water potential (MPa) 40 50 60 70 80 90 100 110 Rainfall (cm) Fic. 1. Predawn, daytime and recovery (calculated as daytime - predawn) xylem water potential of Q. lobata (n = 14), Q. douglasii (n = 13) and Q. agrifolia (n = 13). Given are the means +1 S.E. for the years (from left to right) 1994, 1996, 1991, 1997, 1995, 1998. Predawn xy- lem water potential, Q. agrifolia F = 184, R? = 0.98, P < 0.001; Q. douglasii F = 45, R? = 0.92, P < 0.003; Q. lobata F = 47, R? = 0.92, P < 0.003; midday xylem water potential Q. douglasii F = 11, R? = 0.73, P < 0.03; all other regressions are P > 0.05. lem water potential of all three species, but there was a significant relationship between daytime xy- lem water potential and rainfall only in Q. douglasii (Fig. 1). None of the species exhibited a significant relationship between the overnight recovery and rainfall (Fig. 1). Although all three xylem water potential mea- surements were significantly concordant among the individual trees over the six-years of the study, pre- dawn values were consistantly more similar from KNOPS AND KOENIG: OAK WATER POTENTIALS 107 year to year than either of the other measures (Ken- dall’s coefficient of concordance, predawn 0.653, day 0.298, recovery 0.184, all Chi-Square >39, all P < 0.000). Quercus douglasii and Q. agrifolia had consis- tently lower predawn xylem water potential values than Q. lobata (one way ANOVA with Scheffe’s post hoc comparison; Q. douglasii P < 0.05 in all 6 years; Q. agrifolia P < 0.05 in all years, except 1995). Q. douglasii had consistently the lowest daytime values (significant from Q. lobata in 4 out of 6 years, not in 1995 and 1998), with Q. lobata having intermediate values (significantly different from Q. agrifolia 2 out of 6 years, e.g., 1996 and 1997). Recovery was greater for Q. douglasii and Q. lobata than for Q. agrifolia (significant 4 out of 6 years, not in 1995 and 1998). DISCUSSION Does xylem water potential reflect rainfall? Pre- dawn xylem water potential values were signifi- cantly correlated with rainfall in all three species. This supports the assumption that predawn xylem water potential measured in the driest period of the year reflects relative water availability and that these oak species experience lower water status in dry years. Predawn values were consistently sig- nificantly different among the species, and the dif- ferences were largest in drier years. In contrast, daytime and recovery xylem water potential values converge in wet years (Fig. 1) and have only a lim- ited value in characterizing differences among spe- cies. Are individuals and species different over time? Measurements for individual trees were significant- ly concordant from year to year, more so for pre- dawn than daytime or recovery values. Thus, pre- dawn xylem water potential apparently reflects real and consistent differences among individuals in e1- ther their access to water, in their genetic ability to acquire water or in their ability to conserve water. Our data also support the hypothesis that pre- dawn xylem water potential values are consistently significantly different among the species, with Q. douglasii being more water stress tolerant, because of its low predawn xylem water potential and the significant relationship between rainfall and day- time xylem water potential. However, this scenario does not fit the other two species. These differences are consistent with the limited data on root distri- bution of these three species. Quercus lobata is re- ported as having a deep root system connected to the ground water (Griffin 1973), which might make it less sensitive to the previous 12 months rainfall and more sensitive to long term changes in ground water levels. Quercus agrifolia has an extensive shallow system (Canon 1914a, 1914b) and presum- ably lacks access to the previous winter precipita- tion, which is likely stored in deeper soil levels and Q. douglasii, which does not have consistent access 108 to groundwater (Griffin 1973). This also raises the alternative hypothesis that these oak species might differ in critical water potential for cavitation (e.g., the formation of unreversible air bubbles within the xylem vessels). Cavitation can be a significant cause of hydraulic conductivity loss within oaks due to water stress (Tognetti et al. 1996, 1998) and species specific differences in vulnerability of cav- itation have been reported for oak species (Cocard et al. 1996; Tognetti et al. 1996). Lastly, differences in water conservation caused by differences in phe- nology and physiology among the species may also have contributed to these patterns. The lack of a relationship for Q. agrifolia and Q. lobata correlation between midday xylem water po- tential and the rainfall and the lack between recov- ery and rainfall for all three species indicates that the degree to which xylem water potential recovers overnight is not dependent on rainfall. Instead, in- dividual trees may be able to lower their predawn xylem water potential, and in the case of Q. doug- lasii daytime xylem water potential, thereby in- creasing water uptake in the driest years. This does not imply that the activity of the trees is the same in each year, as the time that the stomates are open in the daytime might be correlated with the amount of water available for transpiration. Alternatively, this might also indicate a strict regulation for water loss for Q. agrifolia and Q. lobata, via stomatal conductance or adjustment in hydraulic conduc- tance to avoid cavitation and Q. douglasii might have a lower critical threshold for cavitation. Test- ing this would require measuring daily patterns of xylem water potential, hydraulic conductivity and cavitation, which we did not do as part of this study. ACKNOWLEDGMENTS Thanks to Louise Johnson for statistical advice, Sarah Hobbie, Bryan Foster, Lars Pierce, Bill Schlesinger for comments, Mark Stromberg and Mark Johnson for logistic help. LITERATURE CITED CALLAWAY, R. M., N. M. NADKARNI AND B. E. MAHALL. 1991. Facilitation and interfering of Quercus doug- lasii on understory productivity in central California. Ecology 72:1484—-1499. CANON, W. A. 1914a. Specialization in vegetation and in environment in California. Plant World 17:223—237. CANON, W. A. 1914b. Tree distribution in central Califor- nia. Popular Science Monthly 85:417—424. COCHARD, H., N. BREDA AND A. GARNIER. 1996. Whole tree hydraulic conductance and water loss regulation in Quercus during drought: Evidence for stomatal control of embolism? Annales des Sciences Foresti- eres 53:197—206. MADRONO [Vol. 47 | DeLucia, E. H., W. H. SCHLESSINGER AND W. D. BILLINGS. | 1988. Water relations and the maintenance of Sierran | conifers on hydrothermically altered rock. Ecology | 69:303-311. | DonovaN, L. A. AND J. R. EHLERINGER. 1994. Water stress | and use of summer precipitation in a Great Basin | shrub community. Functional Ecology 8:289—297. GRIFFIN, J. R. 1973. Xylem sap tension in three woodland oaks in Central California. Ecology 54:152-159. GRIFFIN, J. R. 1988. A natural history of Hastings Reser- vation, Carmel Valley, California. Unpublished manuscript on file at the Hastings Reservation. Uni- © versity of California, Berkeley. Kwnops, J. M. H., J. R. GRIFFIN AND A. C. ROYALTY. 1995. | Introduced and native plants of the Hastings reser- vation, central coastal California, A comparison. Bi- ological Conservation 71:115—123. Knops, J. M. H. AND W. D. KoeENIG. 1994. Water use strat- | egies of five sympatric Quercus species in central | coastal California. Madrono 41:290—301. Knops, J. M. H. AND W. D. KogENic. 1997. Site fertility © and leaf nutrients of sympatric evergreen and decid- uous species of Quercus in central coastal California. Plant Ecology 130:121-131. KOENIG, W. D., J. M. H. KNops, W. J. CARMEN, M. T. STANBACK AND R. L. MuMME. 1996. Acorn produc- | tion by oaks in central coastal California: influence | of weather at three levels. Canadian Journal of Forest | Research 26:1677—1683. KOENIG, W. D., R. L. MuMMgE, W. J. CARMEN AND M. T. STANBACK. 1994. Acorn production by oaks in cen- | tral, coastal California: variation within and among | years. Ecology 75:99-109. Kowe, R. T., R. H. RosBicHaux, S. R. Morse AND C. M. | SMITH. 1991. Plant water status, hydraulic resistance ; and capacitance. Pages 161-184 in R. W. Pearcy, J. | Ehleringer, H. A. Mooney and P. W. Rundell (eds.), | Plant Physiological Ecology. Chapman and Hall, | London. Ko_s, K. J. AND S. D. DAvis. 1994. Drought tolerance and xylem embolism in co-occurring species of coastal | sage and chaparral. Ecology 75:648—659. RITCHIE, G. A. AND T. M. HINCKLEY. 1975. The pressure | chamber as an instrument for ecological research. Ad- vances in Ecological Research 9:165—254. | STRINGER, J. W., P. J. KALISZ AND J. A. VOLPE. 1989. Deep | tritiated water uptake and predawn xylem water po- tentials as indicators of vertical rooting extent in a Quercus-Carya forest. Canadian Journal of Forest | Research 19:627—631. ToGneTTI, R., A. LONGOBUCCO AND A. RASHI. 1998. Vul- | nerability of xylem to embolism in relation to plant | hydraulic resistance in Quercus pubescens and Quer- | cus ilex. New Phytologist 139:437—447. TOGNETTI, R., A. RASHI, C. BERES, A. FENYVESI AND H. W. RIDDER. 1996. Comparison of sap flow, cavitation and water status of Quercus patraea and Quercus | cerris trees with special reference to computer to- . mography. Plant, Cell & Environment 19:928—-938. Turner, N. C. 1981. Techniques and experimental ap- | proaches for the measurement of plant water status. Pant and Soil 58:339-—366. MapRONO, Vol. 47, No. 2, pp. 109-115, 2000 MOLECULAR EVIDENCE FOR THE HYBRID ORIGIN OF OPUNTIA PROLIFERA (CACTACEAE) MICHAEL S. MAYER* AND LAURA M. WILLIAMS Department of Biology, University of San Diego, 5998 Alcala Park, San Diego, CA 92110-2492 JON P. REBMAN Department of Botany, San Diego Natural History Museum, Balboa Park, PO. Box 121390, San Diego, CA 92112-1390 ABSTRACT Opuntia prolifera Engelm., (Coastal Cholla) is common to the coastal sage scrub community extending from Ventura County, California to El Rosario, Baja California. On the basis of morphological interme- diacy, O. prolifera is suspected to have originated through hybridization between O. alcahes and O. cholla, both species of coastal and inland deserts of Baja California and Baja California Sur. For an independent test of this hypothesis, we generated RAPD banding patterns from exemplars of different populations of O. prolifera and the putative parents. In order to exclude other potential parents and to distinguish species-specific RAPD bands we included O. bigelovii Engelm., O. ganderi, O. tesajo, and O. wolfii (L. Benson) M. Baker in the screening. The results provide support for the hybridization hy- pothesis as well as some insight into the speciation process. Twenty-nine primers revealed 44 bands shared only between O. prolifera and one or the other putative parent. No other species included in the screening proved to be comparable alternatives to O. alcahes or O. cholla as the parents of O. prolifera. Unique bands are rare (=2) in O. prolifera compared with O. alcahes (=19) and O. cholla (=23). Trends in the degree of band sharing between O. prolifera and representatives of O. alcahes and O. cholla suggest a central Baja California origin of the species. The dynamic geologic and climatic history of Baja California has fostered a diverse and highly endemic flora on the peninsula, and one of the most Speciose genera is Opuntia (Cactaceae). The spe- ciation routes taken by Opuntia have also been di- verse: Many species are proven hybrids and many more exhibit multiple ploidal levels (D. Pinkava pers. comm.). One suspected hybrid, Opuntia pro- lifera Engelm., was until recently considered a spe- cies derived through cladogenesis. Opuntia prolifera (Coastal Cholla) occurs in the coastal sage scrub community adjacent to the Pa- cific Ocean between Cedros Island, Baja California and Ventura County, California. This taxon is trip- loid (Pinkava et al. 1992) and reproduces almost exclusively asexually, usually through dispersal of detached stem segments. Morphological interme- diacy of O. prolifera between O. alcahes F A. C. Weber and O. cholla FE. A. C. Weber in several char- acteristics (Table 1) has prompted speculation that O. prolifera may have arisen through hybridization of these species (Rebman 1995). Opuntia alcahes and O. cholla are desert taxa of Baja California and typically diploid (Pinkava et al. 1992; Rebman 1995). The two species commonly grow sympatri- cally without hybridizing (Fig. 1), however a hybrid Swarm involving the two species exists near El Ro- sario (Rebman 1995)—which is in the southern part of the range of O. prolifera, an area of overlap be- * Author for correspondence. tween the Sonoran Desert and the California Flo- ristic Province (Fig. 1). Despite the general inter- mediacy of O. prolifera, phenotypic plasticity of the putative hybrid and parent species prevents a strong case for hybridization to rest on morpholog- ical data alone. To subject the hybridization hypothesis to further scrutiny, we surveyed patterns of Random Ampli- fied Polymorphic DNA (RAPD) markers obtained from O. prolifera and its putative parents. The RAPD technique allows relatively quick assessment of a large number of highly polymorphic loci, largely of the nuclear genome (Welsh and Mc- Clelland 1990; Williams et al. 1990). Recent stud- ies have successfully applied the RAPD approach to questions of interspecific hybridization (Pham and Smith 1995; Barker et al. 1996; Daehler and Strong 1997) and hybrid speciation (Smith et al. 1995; Lifante and Aguinagalde 1996; Allan et al. 1997; Padgett et al. 1998). We used RAPD data to test if Opuntia prolifera exhibits the classic genetic expectations of hybrid taxa. If the putative parent species, O. alcahes and O. cholla, were sufficiently divergent genetically prior to a hybridization event, then the hybrid, 1.e., O. prolifera, should exhibit additivity of genetic markers specific to the parent species as well as a lack of unique markers (Gallez and Gottlieb 1982). Additionally, the sterile triploid nature of O. pro- lifera suggests the possibility that “‘fixed’’ hetero- zygosity (sensu Roose and Gottlieb 1976) in O. 110 MADRONO TABLE 1. SELECTED MORPHOLOGICAL CHARACTERISTICS OF OPUNTIA PROLIFERA AND ITS HYPOTHESIZED PARENTS, OQ. AL- CAHES AND O. CHOLLA, IN REGIONS OF SYMPATRY (FROM REBMAN 1995), Characteristic O. alcahes Inner tepal color yellow, green, or red- magenta long and narrow (3.5— 13 X 1.5—4.5 cm) Stem segment shape Tubercule length 4—22 mm Tubercule height 2-9 mm Spine shape short and thin (4—20 xX 0.3—0.5 mm) Areole size 3-5 X 2-4 mm Proliferating fruit rare prolifera could endow it with a higher overall num- ber of RAPD markers relative to its putative par- ents. Finally, patterns in the degree of band sharing between hybrid and parents can also be used to make preliminary inferences regarding the geo- graphic region in which the hybrid taxon arose, as well as the possibility that this event occurred mul- tiple times. METHODS Field collection and DNA extraction. Stem seg- ments were gathered from a single plant (exemplar) O. cholla light to dark pink O. prolifera magenta to deep red intermediate (7.5—12.6 short and wide (6—11.5 x 3.0—4.1 cm) x 3-—5.5 cm) 12—24 mm 20-35 mm 5—9 mm 10—20 mm intermediate (14-18 xX long and thick (20-35 0.7—0.9 mm) x 0.8—-1.3 mm) 5-8 X 3-5 mm 6-11 xk 3-5 mm yes yes at each location (Table 2). DNA was extracted from fresh or frozen stem tissue following a modification (Doyle and Doyle 1987) of the hot CTAB method of Saghai-Maroof et al. (1984). Initial RAPD screening. DNA extracts from [Vol. 47 | Opuntia alcahes, O. prolifera, and O. cholla were | subjected to DNA amplification via the polymerase | chain reaction (PCR) using the 100 10-mer primers | of RAPD Oligo Set 3 (Nucleic Acid-Protein Ser- — vice Unit of the University of British Columbia). | Each 25 wL reaction contained 1 unit of Promega | (Madison, WI) Taq polymerase, 1 X reaction buffer, TABLE 2. COLLECTIONS OF OPUNTIA FROM CALIFORNIA AND MEXICO ANALYZED IN THE PRESENT STUDY; PRECISE LAT./ | LONG. DATA ARE AVAILABLE UPON REQUEST. Exemplars are given abbreviated names for reference in text, tables, and | figures; B.C. = Baja California, B.C.S. = Baja California Sur; asterisk denotes collections used in initial screening only. Species Collection Location O. alcahes FE A. C. Weber alc 1 alc 2 alc:3 alc 4 O. bigelovii Engelmann var. Bigelovii O. cholla EF A. C. Weber cho 1 cho 2 cho 3 cho 4 O. ganderi (C. B. Wolf) J. Rebman & D. J. Pinkava O. prolifera Engelmann pro | pro 2 pro 3 pro 4 O. tesajo Engelmann O. wolfii (L. D. Benson) M. A. Baker Rebman s.n.* Voss 1174 Rebman 4157 Rebman 4835 Rebman 5183 Rebman 4956 Rebman 4158 Rebman 4501 Rebman 4827 Rebman 5184 Rebman 4973 Mayer 591 Rebman 3951 Rebman 3977 Rebman 5119 Rebman 4972 Rebman 3820 CA., San Diego Co., Quail Botanical Gar- dens B.C.S., Cape Region B.C., southwest of Catavifia B.C.S. near Rt. 1 and rd. to Punta Abreo- jos B.C. Sur, Sierra Guadalupe CA., San Diego Co., Hwy S-2 at Cane- brake B.C., southwest of Catavifia B.C.S., Sierra San Francisco B.C.S., Isla Margarita B.C.S., Sierra Guadalupe B.C., San Felipe Desert, n. of Laguna Diablo CA., San Diego Co., U.S.D. campus, West Point B.C., between La Bocana and Puerto San- | to Tomas B.C., s. of Punto Canoas B.C., near La Mision B.C., San Felipe Desert, n. of Laguna Diablo CA., Imperial Co., along I-8 at Mountain Springs Grade | 2000] California : ; wil | ail om * a Baja. ‘| California EI! Rosario alc 2 cho 1 \ cho 2 cho 4 Baja California Sur alc 4 @) == Opuntia prolifera \\ Opuntia alcahes [ Opuntia cholla Fic. 1. Ranges of Opuntia prolifera, O. alcahes, and O. cholla; locations of collections used for population-level comparisons are noted by abbreviated names listed in Ta- ble 2. 50km KA MAYER ET AL.: ORIGIN OF OPUNTIA PROLIFERA eT 1.5 mM MgCl, 0.1 mM of each dNTP, 0.2 pM of one primer, and 1| pl dilute DNA extract. After 2 min at 94°C, the following cycle was repeated 40 times: denaturing at 94°C for 15 s, annealing at 36°C for | min, and elongation at 72°C for 1 min. A final elongation segment was held at 72°C for an additional 6 min. The PCR products were separated electrophoretically in 2% agarose gels and banding patterns were visualized by staining with ethidium bromide and inspection under ultraviolet light. Of the 100 primers, twenty-one showed banding poly- morphism and the sharing of bands between ex- emplars of O. prolifera and either O. alcahes or O. cholla; therefore these primers were used in sub- sequent screening experiments. To replicate the patterns observed in the first round of screening and to identify bands shared be- tween O. prolifera and only O. alcahes or O. chol- la, we included other related species (O. bigelovii, O. ganderi, O. tesajo) in new screening experi- ments using the primers identified in the first round. We assumed that any marker that was also present in one of these additional species was a symple- siomorphic characteristic and not helpful in a rig- orous test of the hybridization hypothesis. Opuntia prolifera exhibited a total of five bands that it shared only with both the putative parents, six bands that it shared only with O. alcahes, and eight bands that it shared only with O. cholla. When O. prolifera was compared in the same way with O. bigelovii, O. ganderi, and O. tesajo, the numbers of exclusively-shared markers were zero, two, and one, respectively. Primary RAPD screening. The results of the ini- tial rounds of screening increased our confidence that Opuntia prolifera was a hybrid derivative of O. alcahes and O. cholla. We then examined the distribution of RAPD markers among populations within these species. We employed the same prim- ers that had proven useful in previous rounds of screening and increased our sample sizes of O. pro- lifera, O. alcahes, and O. cholla to include an ex- emplar from four populations of each taxon (Table 2). In addition, one exemplar each was included from O. bigelovii, O. ganderi, O. tesajo, and O. wolfii. This allowed us to (1) replicate previously observed patterns and identify additional bands shared only between the putative hybrid and its par- ents, (2) get a cursory look at intraspecific RAPD polymorphism, and (3) assess the degree of band sharing on a pairwise population level and, subse- quently, compare these data to the geographic dis- tribution of the populations represented. RESULTS Banding patterns derived from screening 16 ex- emplars using 29 RAPD primers revealed a greater number of markers in support of the hybridization hypothesis than did the initial comparisons (Table 3), presumably because more of the total variation 112 MADRONO TABLE 3. PRIMERS THAT RESOLVE MARKERS SHARED EX- CLUSIVELY BETWEEN OPUNTIA PROLIFERA AND ITS PUTATIVE PARENTS. A = O. alcahes, C = O. cholla, A+C = both species. Markers shared with O. prolifera Primer Sequence A C A+C UBC 202 GAGCACTTAC 2 0) 0 UBC 204 TTCGGGCCGT l 0) 0 UBC 218 CTCAGCCCAG 1 0) 0) UBC 219 GTGACCTCAG 1 ] 2 UBC 220 GTCGATGTCG 3 ] 0) UBC 225 CGACTCACAG 1 2 1 UBC 226 CGCGECTCTAT ] 2 0) UBC 227 CTAGAGGTCC 0 ] 0 UBC 228 GCTGGGCCGA 1 0 UBC 238 CTGTCCAGCA 0 I 0 UBC 245 CGCGTGCCAG 1 0) 6) UBC 246 TATGGTCCGG l 1 0) UBC 247 TACCGACGGA 0) 2 0) UBC 250 CGACAGTCCC 0) ] ] UBC 253 CCGTGCAGTA ] 2 0) UBC 259 GGTACGTACT 2 ] 0) UBC 260 TCTCAGCTAC l 0) 0) UBC 269 CCACTTCGEE ] 2 0) UBC 270 TGCGCGCGGG ] 2 0 UBC 275 CCGGGCAAGC 0 ] 0 UBC 281 GAGAGTGGAA 3 0 ] UBC 283 CGGCCACCGT l 0 0) within each species was assessed and more primers were successful. Pairwise comparisons between ex- emplars of O. prolifera and O. alcahes, or O. pro- lifera and O. cholla revealed 23 and 21 bands, re- spectively, present in at least one population of the two species compared, and found in no other spe- cies (Tables 3, 4). Of these 44 marker loci, the group of O. prolifera exemplars is polymorphic for at least 31 (>70%). A comparison between O. al- cahes and O. cholla detected just one shared band, which was unique to just one exemplar of each spe- cies. A comparison of O. prolifera exemplars against representatives of O. bigelovii, O. ganderi, O. tesajo, and O. wolfii revealed one, one, zero, and zero bands, respectively, that were exclusively shared. Of the aforementioned markers of hybrid- ization, only a small number are fixed in all ex- emplars of O. prolifera and O. alcahes (=3) or O. prolifera and O. cholla (=5) (Table 4). Five addi- tional bands were shared exclusively among O. prolifera and both putative parents. Opuntia prolifera did not possess a significantly greater number of bands (P = 0.97) compared with its putative parents: 167 bands total versus 164 in both O. alcahes and O. cholla (Table 4). Compar- ison of O. prolifera with its putative parents also revealed significantly (P < 0.01) fewer unique bands in O. prolifera (n = 2) than in either O. alcahes (n = 19) or O. cholla (n = 23) (Table 4). A factor analysis (Statview 5.0, SAS Institute, Inc. 1998) of TABLE 4. SUMMARY DATA FROM PRIMARY SCREENING OF RAPD PATTERNS IN OPUNTIA PROLIFERA (P) AND ITS PU- | TATIVE PARENTS QO. ALCAHES (A) AND O. CHOLLA (C). Characteristic P A C Total bands examined 167 164 164 Unique bands Z 19 23 Bands shared only with P — 28 21 Bands shared only with P, fixed for both taxa — 3 SS) Bands shared only between A and C — ] Bands shared only among A, C, and P 5 the RAPD data provided the means to assess over- all similarity among the exemplars included in the study. The first two factors account for 41.5% and 16.8% of the variance in the data set. Plotting the exemplars by their scores along factors one and two places O. prolifera clearly intermediate between O. alcahes and O. cholla (Fig. 2). Estimates of banding pattern similarity between pairs of populations of O. prolifera and O. alcahes or O. cholla were made in two ways: using the Simple Matching Coefficient (Sokal and Michener 1958) and the Coefficient of Jaccard (Sneath 1957). We tallied presence or absence of marker bands for all pairwise comparisons of exemplars of O. pro- lifera vs. O. alcahes or O. cholla. We ignored bands that were fixed for all exemplars of the two taxa being compared in an effort to minimize the effect of symplesiomorphies on the coefficient. The Sim- ple Matching Coefficient (SMC) was calculated by adding the matches (shared absences plus shared presences of markers) and dividing by the total number of matches and mismatches. The Coeffi- cient of Jaccard (CJ) omits shared absences from the numerator and denominator. We were con- cerned that the SMC would be biased by artefacts Factor 2 =e etn, O S25 oO dee Or. Pil -1 -.75 -.5 Factor 1 Fic. 2. Unrotated factor plot showing position of ex- emplars along factors one and two; refer to Table 2 for key to abbreviations. [Vol. 47 | | 2000] MAYER ET AL.: ORIGIN OF OPUNTIA PROLIFERA 113 TABLE 5. PAIRWISE SIMILARITY COEFFICIENTS BETWEEN EXEMPLARS MEASURING OPUNTIA PROLIFERA (PRO) X O. ALCAHES | (ALC) AND O. PROLIFERA X O. CHOLLA (CHO). Simple Matching Coefficients before slash, Coefficient of Jaccard after; see Table 2 for key to abbreviations. pro | pro 2 pro 3 pro 4 alc 1 0.539/0.143 0.524/0.231 0.359/0.242 0.225/0.184 alc 2 0.583/0.000 0.568/0.111 0.250/0.069 0.189/0.063 pale 3 0.475/0.160 OD12/0.259 0.525/0.424 0.342/0.308 alc 4 0.436/0.154 0.475/0.250 0.539/0.455 0.400/0.368 cho 1 0.528/0.150 0.421/0.120 O29 7/0212 0.447/0.364 cho 2 0.650/0.263 0.600/0.200 0.263/0.125 0.250/0.167 cho 3 0.528/0.105 0.526/0.182 0.243/0.152 0.368/0.273 cho 4 0.514/0.182 0.462/0.192 0.324/0.242 0.487/0.412 arising from poor amplification, thereby inflating cation: RAPDs are dominant, diallelic markers and estimates of similarity between two populations. As — thus may not show the same patterns of additivity expected, the SMC values were uniformly greater as codominant markers. Another surprising out- than the CJ values (Table 5), but in some cases the come was the RAPD polymorphism evident among two approaches yield different patterns of relation- exemplars of O. prolifera, indicating interpopula- ships among the populations. For example, the two tional genetic diversity. Because O. prolifera is sets of coefficients comparing pro 2 with the four only known to reproduce asexually, this variation populations of O. alcahes display almost opposite may signify one or more of the following: (1) mul- rankings by magnitude (Table 5), perhaps indicat- tiple independent hybrid origins of O. prolifera, (2) ing that the inclusion of shared absences does in- undetected sexual reproduction, or (3) genetic di- deed bias the SMC in this application. vergence via somatic mutations. We introduce these Considering, therefore, only the CJ values we see alternative processes briefly below, but leave a crit- that among all the exemplars of O. alcahes and O. ical analysis to future studies specifically targeted cholla, two exemplars of O. prolifera (pro 1 and 2) to discriminating among these phenomena. exhibited greater similarity, albeit by narrow mar- First, recurrent origin of a triploid O. prolifera gins, to alc 3 and cho 2 (Table 5). In contrast, pro would require either that multiple diploid-level hy- 3 and pro 4 exhibited greater similarity to alc 4. and _ bridizations must each have been followed by the cho 4. These relationships also had geographic sig- _ production of triploid offspring, or that a pairing of nificance: exemplars alc 4 and cho 4 were collected —_q diploid parent with a tetraploid parent must have from the same vicinity in northern Baja California occurred multiple times. Because both Opuntia al- Sur, and alc 3 and cho 2 were collected just 40 km = canes and O. cholla are diploid with rare exception apart, also in northern Baja California Sur (Fig. 1). (Rebman 1995), the latter scenario seems unlikely. If the former scenario occurred, diploid hybrids DISCUSSION should be common and widespread in the zone of Molecular evidence supports the proposition that Sympatry. However, only one diploid count has hybridization between Opuntia alcahes and O,. been documented for O. prolifera (Pinkava and cholla gave rise to O. prolifera. Forty-four RAPD _ Parfitt 1982). Despite the apparent obstacles to re- markers are shared only between O. prolifera and Cutting origins of O. prolifera, RAPD-based rela- one or the other parent species; no other candidates tionships among exemplars employed in the present emerge as comparable alternatives to O. alcahes or Study provide some evidence in its support. Two O. cholla as the parents of O. prolifera. As ex- exemplars of O. prolifera (pro 1 and 2) are more pected for a hybrid, O. prolifera exhibits signifi- closely related to alc 3 and cho 2 than to the other cantly fewer unique RAPD markers than its parent representatives of O. alcahes and O. cholla. In con- species. Moreover, multivariate analysis of marker trast, the other two exemplars of O. prolifera (pro distribution places exemplars of O. prolifera inter- 3 and 4) are more closely related to alc 4 and cho mediate between those of O. alcahes and O. cholla. 4. Furthermore, specimens alc 4 and cho 4 were Some results of this study, however, were con- collected from the same vicinity, and the locations trary to early expectations. First, O. prolifera band- of alc 3 and cho 2 were separated by just 40 km. ing patterns did not exhibit the greater numbers of Next, for sexual reproduction to be the source of loci predicted for a sterile hybrid or allopolyploid interpopulational variation, triploid O. prolifera (Table 4). This observation may indicate a relative- plants must give rise to triploid offspring. If meiosis ly low degree of divergence between O. alcahes could occasionally generate viable gametes of vary- and O. cholla, or a low amount of variation derived ing ploidy in O. prolifera, we should expect more from the actual hybridization event, or it may ex- _ ploidal levels than just triploid in these populations. pose a limitation of RAPD markers in this appli- Currently, only a hybrid swarm of the El Rosario 114 MADRONO area (Fig. 1) has yielded counts in O. prolifera that exceed triploidy, including a hexaploid—presum- ably an autopolyploid that formed through the fu- sion of two unreduced gametes (Rebman 1995). Lastly, reproduction in O. prolifera relies per- haps exclusively on establishment of detached stem segments (Rebman 1995). Long-term clonal growth allows for the possibility that somatic mutations in branch primordia could generate RAPD variation among populations of O. prolifera. The importance of somatic mutations in clonal species has long been suspected and is gaining more experimental support (Ellstrand and Roose 1987, reviewed in de Kroon and van Groenendael 1997). Origin of Opuntia prolifera. Although it is al- most uniformly triploid across its range, O. prolif- era could have originated as a diploid, through hy- bridization of diploid O. alcahes and O. cholla. Meiotic irregularities in this diploid hybrid allowed the production and subsequent fusion of a reduced and unreduced gamete, generating a triploid off- spring. This route from diploidy to triploidy has been seen repeatedly among cactus species (D. Pin- kava pers. comm.). A notable example is O. bige- lovil, a close relative of O. prolifera, which appar- ently arose as a diploid but is now predominantly triploid (D. Pinkava pers. comm.). If indeed O. pro- lifera originated in this way, some set of factors then allowed the triploid to surpass its diploid pro- genitor and thrive in the coastal sage scrub of the Californias, a habitat to which few other chollas are well-adapted. All exemplars of O. prolifera showed the great- est similarity to representatives of O. alcahes (alc 3 and 4) and O. cholla (cho 2 and 4) collected from the northern end of Baja California Sur, indicating a possible region of origin of O. prolifera. Surpris- ingly, this region is greatly disjunct from the pres- ent range of O. prolifera (Fig. 1). However, the repeated shifts in climate and vegetation in the his- tory of Baja California cautions us from excluding this proposition prior to further investigation. Establishment of Opuntia alcahes and O. cholla as the parents of O. prolifera now sets the stage for further population genetic studies, which should be aimed towards testing for recurrent origins of O. prolifera and the route by which it attained triplo- idy. ACKNOWLEDGMENTS We thank Douglas Gilbert and Michelle Darnell for as- sistance in the lab, Rick Gonzalez for help with statistical tests, and two anonymous reviewers for comments on an earlier version of this paper. This research was supported in part by a Faculty Research Grant from the College of Arts and Sciences of the University of San Diego to M. S. M., and by the Undergraduate Research Fund of the Associated Students of the University of San Diego. [Vol. 47 LITERATURE CITED ALLAN, G. J., C. CLARK, AND L. H. REISEBERG. 1997. Dis- tribution of parental DNA markers in Encelia virgi- nensis (Asteraceae: Heliantheae), a diploid species of | putative hybrid origin. Plant Systematics and Evolu- tion 205:205—221. BARKER, N. P., L. Y. KRUGER, J. H. VLOK, AND E. H. HArR- LEY. 1996. RAPD data suggest the Gamkapoort Aloe to be of hybrid origin. South African Journal of Bot- any 62:292—295. DAEHLER, C. W. AND D. R. STRONG. 1997. Hybridization between introduced smooth cordgrass (Spartina al- terniflora; Poaceae) and native California cordgrass (S. foliosa) in San Francisco Bay, California, USA. American Journal of Botany 84:607-611. DoyLe, J. J. AND J. L. DOYLE. 1987. A rapid DNA isola- tion procedure for small amounts of fresh leaf tissue. Phytochemical Bulletin 19:11—15. DE Kroon, H. AND J. VAN GROENENDAEL (EDS.). 1997. The Ecology and Evolution of Clonal Plants. Backhuys, Leiden, The Netherlands. ELLSTRAND, N. C. AND K. L. ROOSE. 1987. Patterns of | genotype diversity in clonal plant species. American Journal of Botany 74:123-131. GALLEZ, G. P. AND L. D. GOTTLIEB. 1982. Genetic evidence for the hybrid origin of the diploid plant Stephano- meria diegensis. Evolution 36:1158—1167. LIFANTE, Z. D. AND I. AGUINAGALDE. 1996. The use of Random Amplified Polymorphic DNA (RAPD) markers for the study of taxonomical relationships among species of Asphodelus sect. Verinea (Aspho- delaceae). American Journal of Botany 83:949—953. PapcETT, D. J., D. H. LEs, AND G. C. Crow. Evidence for the hybrid origin of Nuphar X rubrodisca (Nympha- ceae). American Journal of Botany 85:1468—1476. PHAM, T. V. AND J. FE SmituH. 1995. Genetic variation and hybridization in the narrow endemic Allium aaseae based on RAPD data. American Journal of Botany 82 (supplement: abstracts): 155—156. AND B. D. PARFITT. 1982. Chromosome numbers in some cacti of western North America-IV. Bulletin of the Torrey Botanical Club 109:121-128. , M. A. BAKER, AND R. D. WORTHINGTON. 1992. Chromosome numbers in some cacti of western North America-VI, with nomenclatural changes. Ma- drono 39:98-113. REBMAN, J. P. 1995. Biosystematics of Opuntia subgenus Cylindropuntia (Cactaceae), the chollas of Lower California, Mexico. Ph.D. dissertation. Arizona State University. Roose, M. L. AND L. D. GOTTLIEB. 1976. Genetic and biochemical consequences of polyploidy in Trago- pogon. Evolution 30:818—830. SAGHAI-MARoor, M. A., K. M. SOLIMAN, R. A. JORGENSEN, AND R. W. ALLARD. 1984. Ribosomal spacer lengths in barley: Mendelian inheritance, chromosomal lo- cation and population dynamics. Proceedings of the National Academy of Sciences, U.S.A. 81:8014— 8018. SmiTH, J. E, C. C. BURKE, AND W. L. WAGNER. 1995. In- terspecific hybridization in natural populations of Ha- waiian Cyrtandra: evidence from RAPD markers. American Journal of Botany 82 (supplement: ab- stracts):162. 2000] MAYER ET AL.: ORIGIN OF OPUNTIA PROLIFERA > SNEATH, P. H. A. 1957. Some thoughts on bacterial clas- nomes using PCR with arbitrary primers. Nucleic Ac- sification. Journal of General Microbiology 17:184— ids Research 18:7213-7218. 200. WILLIAMS, J. G. K., A. R. KUBELIK, K. J. LIVAK, J. A. SOKAL, R. R. AND C. D. MICHENER. 1958. A statistical RAFALSKI, AND S. V. TINGEY. 1990. DNA polymor- method for evaluating systematic relationships. Uni- phisms amplified by arbitrary primers are useful as versity of Kansas Science Bulletin 38:1409-1438. genetic markers. Nucleic Acids Research 18:6531-— WELSH, J. AND M. MCCLELLAND. 1990. Fingerprinting ge- 6535. MADRONO, Vol. 47, No. 2, pp. 116—126, 2000 FLORAL VARIATION IN DELPHINIUM VARIEGATUM (RANUNCULACEAE) SHANA C. Dopp! Department of Biology, San Diego State University, San Diego, California 92182 KAIUS HELENURM? Department of Biology, University of South Dakota, Vermillion, South Dakota 57069 ABSTRACT Delphinium variegatum is subdivided into three subspecies distinguished by three floral characters. Delphinium v. variegatum is found in central and northern California, while D. v. kinkiense (an endangered taxon) and D. v. thornei are endemic to San Clemente Island off the coast of southern California. Broad variation is documented in most natural populations for all three floral characters. Our results indicate that the two metric characters, lateral sepal length and lower petal blade length, provide no clear distinction between the taxa. Sepal color is the least ambiguous for differentiating the subspecies, but is problematic in distinguishing between D. v. kinkiense and D. v. thornei. Sepal color exhibits a complex pattern of variation on San Clemente Island in which northern populations generally contain primarily light-flowered individuals, southern populations generally contain primarily dark-flowered individuals, and central pop- ulations may contain substantial numbers of both light- and dark-flowered individuals as well as inter- mediates. However, one southern population contains primarily light-flowered individuals, and almost half of the populations contain individuals having sepal colors considered to represent the two different sub- species. Further taxonomic study including additional characters is recommended to determine whether D. v. kinkiense and D. v. thornei should be considered distinct taxa. Delphinium variegatum Torrey & A. Gray (Ran- unculaceae) is a perennial larkspur that is found in grassland and open woodlands of mainland Cali- fornia and San Clemente Island, the southernmost of the Channel Islands off the coast of southern California (Warnock 1990b). One subspecies, D. v. ssp. variegatum (Royal larkspur), is found exclu- sively on the mainland and ranges approximately from northern to central California, from the coast to the Sierra Nevada °?? foothills (Fig. 1). The other two subspecies, D. v. ssp. kinkiense (Munz) M. J. Warnock (San Clemente Island larkspur; Warnock 1990a) and D. v. ssp. thornei Munz (Thorne’s lark- spur; Munz 1969) are insular endemics found only on San Clemente Island. The Channel Islands are thought to provide refuge for a number of species with northern affinities, including D. variegatum (Raven and Axelrod 1978), that once extended far- ther south on the mainland during Pleistocene plu- vial cycles (Raven 1963). The island endemic subspecies of D. variegatum are vulnerable to extinction because of their limited distribution (Skinner and Pavlik 1994). Delphinium v. kinkiense is listed as endangered by the U.S. Fish & Wildlife Service (USFWS) and by the California Department of Fish and Game. However, the rarest of the subspecies, D. v. thornei, has no special legal 'Current address: S. C. Dodd Biological Consulting, 3786 Dana Place, San Diego, California 92103. ?Author for correspondence: Department of Biology, University of South Dakota, Vermillion, South Dakota 57069. E-mail: helenurm @usd.edu. status, although the USFWS considers it to be a species of concern. Both of these taxa are on the | California Native Plant Society List 1B (plants rare, | threatened, or endangered in California and else- where; Skinner and Pavlik 1994). The subspecies of D. variegatum are distin-_ guished primarily by three floral characters: sepal | color, lateral sepal length and lower petal blade | length (Warnock 1990b, 1993, 1997; summarized in Table 1; Fig. 2). However, there is overlap. among the subspecies. The mainland subspecies, D. v. variegatum, is differentiated from the two island subspecies by its deep royal blue flowers, as the. ranges for the two metric characters encompasses the variation observed in the entire species. The. two island subspecies are differentiated from each other by all three characters, in spite of consider- able overlap, with D. v. kinkiense having mainly white, smaller flowers and D. v. thornei having mainly bright blue, larger flowers. Munz (1974), interestingly, had described D. v. thornei as having | smaller flowers (sepals ca. 12 mm long) than D. v. kinkiense (which he recognized as a separate spe- cies, D. kinkiense; sepals 16-18 mm long). Current keys use sepal color to identify taxa (Warnock | 1990b, 1993, 1997), although the most recent also uses density of hairs on the base of the stem to. distinguish between D. v. variegatum and the island | subspecies (Warnock 1997). ! Casual observation of natural populations of D. | variegatum on San Clemente Island suggests that | hybridization may be occurring between D. v. kin-_ kiense and D. v. thornei in some populations. At 2000] DODD AND HELENURM: FLORAL VARIATION IN DELPHINIUM VARIEGATUM Lay San Clemente Island 10 : 6 9 : , 2) 11 SR , CaN 20 14 y | iy a D4 17 PNR HE e 23 16 OE . 18 | | 3 we N ee a: 20KmM Fic. 1. Distribution and sampled populations of Delphinium variegatum from mainland California (ssp. variegatum) and from San Clemente Island (ssp. kinkiense and ssp. thornei). the time this study was initiated, fewer than 15 pop- southern half. However, some populations in the ulations of D. variegatum were known on San Cle- central part of San Clemente Island include indi- mente [sland (along with scattered individuals), viduals exhibiting white, bright blue or intermediate with D. v. kinkiense occurring in the northern half flower colors. Natural hybridization has been doc- of San Clemente Island and D. v. thornei in the umented to regularly occur among other taxa in the 118 (a) (b) lateral sepal lower petal blade Fic. 2. Diagram of (a) side view and (b) front view of a D. variegatum flower. genus Delphinium (Warnock 1990b, 1997); natural hybrids are known between D. v. variegatum and D. hansenii, D. hesperium, D. parryi, and D. re- curvatum (Lewis and Epling 1954). The goals of this study were (1) to document variation in floral morphology in natural popula- tions of the subspecies of D. variegatum, (2) to evaluate the utility of the three floral characters for distinguishing D. v. kinkiense and D. v. thornei, and (3) to identify populations of D. variegatum on San Clemente Island as D. v. kinkiense, D. v. thornei or mixed populations. MATERIALS AND METHODS Study sites. Twenty-four populations of D. v. kin- kiense and D. v. thornei were located and sampled from San Clemente Island in 1996 (Fig. 1; Table 2). This represents all known populations and prob- ably all of the populations on the island; subsequent surveys have failed to reveal additional locations (Junak and Wilken 1998; S. Burckhalter, University of South Dakota, pers. comm.; K. Helenurm, per- sonal observation). Seven populations of D. v. var- legatum were sampled across its range, from Marin County in the north to southern Monterey County and east to Tuolumne and Mariposa Counties. All populations of the three subspecies occurred in open grassland habitat. Island populations were found only on west or northerly aspects, probably due to moister, cooler conditions in these areas. Flower collection and measurements. Thirty to forty-four flowering individuals were haphazardly sampled from large populations (Table 2). In small- er populations, all flowering individuals were sam- pled. Two flowers from each sample individual TABLE 1. FLORAL CHARACTERS USED TO DISTINGUISH THE FROM WARNOCK 1990B, 1993, 1997). Delphinium Floral character ssp. kinkiense Sepal color white to light blue (or lavender) Lateral sepal length 11-18 mm Lower petal blade length 4—9 mm MADRONO were measured for sepal color, lateral sepal lengths, | and lower petal blade lengths (Fig. 2). Sepal color | was measured by matching lateral sepals to a color chart (Royal Horticultural Society 1986). Colors were quantified by matching the color chart patches | to colors in Adobe Photoshop (1995) computer | software using a calibrated monitor. We recorded their hue, saturation, and brightness values using | the same computer system for all measurements. Values for brightness were used for analysis be- cause brightness (quantifying the degree of light- ness or darkness, ranging from 0 representing black to 100 representing white) best reflects Warnock’s | (1990b) descriptions of the subspecies and the range of variation in flower color we observed. Al- though differences in hue (the attribute of colors that permits them to be classed as blue versus lav- | ender or purple, for example) and saturation (the | degree of difference from a gray having the same lightness) occur, the quantifiable difference between [Vol. 47 | white, light blue, bright blue and deep royal blue © sepals is reflected in brightness values rather than hue or saturation. In all, 775 individuals were measured for floral — characters in all 24 San Clemente Island popula- tions, and 242 individuals were measured in 7. mainland populations, for a total of 1017 individ- uals. Analysis. Measurements of brightness and metric | characters were averaged for different flowers of the same individual. Associations among the dif- | ferent floral characters in D. v. kinkiense and D. v. thornei were addressed in three ways. First, t-tests | were used to test differences in lateral sepal and | lower petal blade lengths in individuals with light | versus dark sepal color. Second, correlations among | the floral characters were tested using Pearson’s . correlation analysis. Third, grouping of floral char- | acters was investigated using principal components analysis (PCA). All analyses were performed using © SYSTAT (1992). RESULTS Variation in floral morphology. Box plots of flo- | ral variation in D. variegatum illustrate broad vari- | ation in most natural populations on San Clemente | Island (Fig. 3). Sepal color is invariant, or nearly | so, in some populations (populations 1—7, 10). The | metric characters, lateral sepal length and lower | variegatum ssp. thornei ssp. variegatum light blue to bright deep royal blue, rarely blue white or lavender 17—21 mm 10—25 mm 6—11 mm 4—11 mm THREE SUBSPECIES OF DELPHINIUM VARIEGATUM (SUMMARIZED | | | | 2000) DODD AND HELENURM: FLORAL VARIATION IN DELPHINIUM VARIEGATUM 19 TABLE 2. POPULATION NUMBER, SUBSPECIES DESIGNATION, COLLECTION LOCATIONS, APPROXIMATE POPULATION SIZES (1996), AND SAMPLE SIZES OF DELPHINIUM VARIEGATUM. San Clemente Island populations are listed north to south. Population Population Sample number Subspecies Location size size Island 1 kinkiense Flasher Canyon 200 28 kinkiense Nots Drive 200 4] 3 kinkiense Pelican Canyon 2500 44 4 kinkiense Larkspur Canyon 150 40 5 kinkiense Stone Canyon 500 37 6 kinkiense Burns-Horton Canyon >1000 40 7 kinkiense Lower Twin Dams Canyon 16 ° 8 mix Boulder 200 40 ) mix Upper Twin Dams Canyon 1000 ee) 10 kinkiense Warren Canyon 200 S7 1] thornei Upper Middle Ranch Canyon I 32 12 mix Lower Middle Ranch Canyon 350 32 13 kinkiense Waynuk Canyon 1000 41 14 thornei North Norton Canyon 60 17 15 thornei South Norton Canyon 500 36 16 thornei Horse Canyon 16 6 17 thornei Box Canyon 150 30 18 thornei Cave Canyon 400 37 19 thornei Eagle Canyon 150 38 20 thornei Eagle-Bryce Canyon 7 6 Pai thornei Bryce Canyon 200 39 pe thornei Malo 300 31 23 thornei Canchalagua Canyon 3000 40 24 kinkiense Guds iis) 35 Mainland 1 variegatum Edgewood County Park 200 33 2 variegatum China camp State Park 100 38 3 variegatum Green Springs Road 40 22 + variegatum Chinese Station 200 40 5) variegatum Route 49 150 39 6 variegatum Nacimiento-Ferguson Road 200 31 a variegatum G14 250 39 petal blade length, are highly variable in most pop- ulations, with largely overlapping ranges. Mainland populations show a similar pattern (Fig. 3). Sepal color is relatively invariant in main- land populations with the exception of China Camp State Park, in which many white-flowered individ- uals occurred (18 of the 38 sampled). Mainland populations also show variation in metric charac- ters, but they generally have narrower distributions with fewer outside values. Edgewood County Park has longer lateral sepals and lower petal blades than the other mainland populations. Histograms of the three floral characters indicate lighter-colored and larger flowers for the island populations (treated together because of the broadly overlapping distributions noted above) than for mainland populations (Fig. 4). Sepal color is dis- tributed bimodally on the mainland only because of white-flowered individuals in China Camp State Park. The distribution of sepal color on San Cle- mente Island is clearly bimodal, with 375 of the 775 individuals sampled (48.4%) having white or very light blue flowers (henceforward ‘“‘light-flow- ered’’; brightness values from 88—100), 72 (9.3%) being intermediate (brightness values 56—87), and 328 (42.3%) having bright blue flowers (hencefor- ward ‘‘dark-flowered’’; brightness values 28-55). In contrast, the metric characters have unimodal distributions. The overall bimodal distribution of sepal color on San Clemente Island shows a geographic pat- tern. Northern populations generally contain pri- marily light-flowered individuals and southern pop- ulations generally contain primarily dark-flowered individuals (Fig. 5). Central populations may con- tain substantial numbers of both flower types as well as intermediates. Association among floral characters. The lateral sepal lengths of San Clemente Island individuals with light (brightness values 88-100) and dark (brightness values 28-55) sepal colors are signifi- cantly different (t = —5.78, df = 698, P < 0.0001), but their means (16.31 mm and 17.07 mm, respec- tively) and ranges (12.0-21.75 mm and 10.75-— 24.25 mm, respectively) are very similar. Likewise, lower petal blade lengths are significantly different for the two brightness classes (t = —6.11, df = 689, 120 MADRONO San Clemente Island 100 90 80 Sepal 70 brightness 50 40 30 Lateral sepal length (mm) Lower petal blade length (mm) 123 4 5 67 8 Fic. 3. [Vol. 47 | 9 10 1112 13 14 15 16 17 18 19 20 21 22 23 24 123 4 5 6 7 Mainland California Box plots of (a) brightness, (b) lateral sepal length, and (c) lower petal blade length in populations of D. variegatum. Median values (central line, defining the 50th percentile), upper and lower hinges (edges of the central box, defining the 25th and 75th percentiles), whiskers (extending to the farthest observation from the hinges not farther than 1.5 times the distance between the hinges), outside values (asterisks, observations farther from the hinges than | 1.5 times the distance between the hinges), and far outside values (open circles, observation farther from the hinges | than 3.0 times the distance between the hinges) are illustrated. P < 0.0001), but their means (6.86 mm and 7.22 mm, respectively) and ranges (5.0-9.5 mm and 5.0-10.12 mm, respectively) are almost identical. Three of the four means fall within the overlapping portion of the ranges described for the two subspe- cies (Warnock 1990b, 1993, 1997). A strong correlation exists between lateral sepal length and lower petal blade length (Pearson’s r = 0.619, P < 0.0001; Fig. 6). Weaker correlations ex- ist between brightness and lateral sepal length (Pearson’s r = —0.213, P < 0.0001) and between brightness and lower petal blade length (Pearson’s r = —0.216, P < 0.0001). PCA groups individuals primarily by flower col- or with a broad range of lateral sepal and lower petal blade lengths for each color class (Fig. 7). The | first two axes account for 58.11% and 29.20% of | the total variation, for a total of 87.31%. Plots of | the first and third and of the second and third axes | (not illustrated) are dense clouds of points showing , no structure. The deep royal blue sepal color of D. v. varie- gatum is significantly different from both light-— flowered and dark-flowered island plants (mean = 38.19, range = 30.0—100.0; F = 1778.9, df = 2, r° = 0.798, P < 0.0001; Tukey HSD multiple com- parison P < 0.0001 for both comparisons). Lateral | sepals in D. v. variegatum are significantly shorter | than in light-flowered and dark-flowered island © plants (mean = 15.66, range = 11.25—20.5; F = 2000] 400 Mainland 300 ’ ie] Number of individuals 200 100 RS SS ye 200 || Mainland SCI Number of individuals 100 10 11 12 13 14 15 Lateral 300 Wi Mainland ges 200 Number of individuals 100 28 32 36 40 44 48 5 Sepal DODD AND HELENURM: FLORAL VARIATION IN DELPHINIUM VARIEGATUM AA ' E3 = Ea a = 2 Ea Ea ‘ste Ea a 2 56 60 64 68 72 76 80 84 88 92 96100 brightness (%) 16 17 18 19 20 21 22 23 24 25 sepal length (mm) Lower petal blade length (mm) Fic. 4. Histograms of brightness, lateral sepal length, and lower petal blade length for populations of D. variegatum on San Clemente Island. X-axis values represent the minimum of a class. 37.2, df = 2, r? = 0.076, P < 0.0001; Tukey HSD multiple comparison P < 0.0001 for both compar- isons). Lower petal blades are also significantly shorter in D. v. variegatum (mean = 6.34, range = 4.00-12.15; F = 47.7, df = 2, r° = 0.097, P < 0.0001; Tukey HSD multiple comparison P < 0.0001 for both comparisons). DISCUSSION Floral variation. All three floral characters ex- hibit substantial variation within populations. The metric characters, lateral sepal length and lower petal blade length, exhibit unimodal distributions both on the mainland and on San Clemente Island. Mainland populations have smaller flowers than is- land populations, with the exception of Edgewood County Park in which lower petal blades lengths even exceed those of San Clemente Island plants. Sepal color is relatively invariant on the main- land, although it shows a bimodal distribution due to the high proportion of white-flowered individuals in China Camp State Park. In contrast, sepal color is clearly bimodally distributed on San Clemente Island. Most island populations of D. variegatum are highly variable in sepal color, although some consist primarily of white-flowered individuals. Floral characters and taxonomy. Warnock (1990b) divided D. variegatum into three subspe- cies primarily on the basis of sepal color, lateral sepal length, and lower petal blade length. The 122 Fic. 5. MADRONO gy 2 km Pie charts of flower color in San Clemente Island populations of D. variegatum. White areas represent pro- portion of individuals with white or light blue flowers (brightness values from 88 to 100), black areas represent individuals with bright blue flowers (brightness <55), and stippled areas represent individuals with intermediate colors (brightness values from 56 to 87). mainland populations we sampled generally fit Warnock’s (1990b) descriptions of D. v. variegatum (Table 1), although some individuals in Edgewood County Park have lower petal blade lengths ex- ceeding the described taxonomic range. Delphinium v. variegatum differs from the two island subspe- cies in generally having darker (deep versus bright blue) flowers (except for many individuals in China Camp State Park) and shorter lower petal blades (except in Edgewood County Park). Considerable population differentiation appears | to exist within D. v. variegatum. Of the seven pop- ulations we sampled, two are morphologically dis- tinct: China Camp State Park has a high proportion of white-flowered individuals (absent in the other | [Vol. 47 populations we sampled), and Edgewood County 2000] DODD AND HELENURM: FLORAL VARIATION IN DELPHINIUM VARIEGATUM 123 | 14 12 ssp. thornei 10 ssp. kinkiense Lower petal blade length 8 mm (mm) 5 4 2 10 12 14 16 18 20 22 24 26 Lateral sepal length (mm) t a Lateral sepal ' length (mm) a i 30 40 50 60 70 80 90 100 Sepal brightness (%) Owe Lower petal a blade length ae r Z (mm) I: ‘ f Ry . 30 40 50 60 70 80 90 100 Sepal brightness (%) Fic. 6. Scatterplots of sepal brightness, lateral sepal length, and lower petal blade length for individuals of all pop- ulations of D. variegatum on San Clemente Island. Taxonomic designations shown are from Warnock (1990b, 1993, 1997). Park has larger flowers than other populations. Sub- sequent to sampling, we discovered that Edgewood Park is the only population we sampled that occurs on serpentine soils. Warnock (1990b) considers ser- pentine soil populations of D. v. variegatum to be not well marked morphologically and did not rec- Ognize them as distinct taxa. Instead, Warnock (1997) comments that plants with large flowers are common in the San Francisco Bay area, either as scattered individuals or as populations made up largely of such individuals. In other species, plants growing on serpentine soils have often been docu- mented to be morphologically and genetically dis- tinct from plants growing on non-serpentine soils (Kruckeberg 1954; Mayer et al. 1994). Intensive sampling of additional natural populations of D. v. 124 MADRONO Factor 2 Fic. 7. (brightness values 28-55) flower color. variegatum may clarify whether serpentine soil populations are differentiated morphologically from non-serpentine soil populations or whether varia- tion is geographic in pattern. Delphinium vy. kinkiense and D. v. thornei also generally conform to Warnock’s (1990b, 1993, 1997) descriptions, in spite of some individuals having lateral sepals exceeding the described tax- onomic range (Fig. 6). However, our data do not support the separation of D. v. kinkiense and D. v. thornei on the basis of all three floral characters. Specifically, our results indicate that the two metric characters, lateral sepal length and lower petal blade length, provide no clear distinction between these taxa. Both exhibit a unimodal distribution on San Clemente Island so that any attempt to use them for delineating taxa is necessarily arbitrary. In addition, the majority of individuals fall within or near the area of overlap for these characters. Both lateral sepal length and lower petal blade length show a Statistically significant association with sepal color, indicating that light-colored flow- ers tend to be smaller than dark-colored flowers. These results are in agreement with Warnock’s (1990b) descriptions. However, their significance should be considered an artifact of large sample size. The almost identical means and almost com- pletely overlapping ranges of the metric characters for light- versus dark-colored flowers indicate that Scatterplot of the first and second PCA axes for floral characters in D. variegatum on San Clemente Island. Individuals are coded as having light (brightness values 88—100), intermediate (brightness values 56—87) or dark ’ [Vol. 47| Light Intermediate Dark Factor 1 the statistical differences have little taxonomic sig- | nificance. The remaining floral character, sepal color, is the’ least ambiguous for differentiating between D. v. | kinkiense and D. v. thornei but it is also problem- ' atic, exhibiting a more complex pattern of variation | on San Clemente Island than previously suspected. Although northern populations generally contain | primarily light-flowered individuals and southern | populations generally contain primarily dark-flow- | ered individuals, central populations may contain substantial numbers of both flower types as well as | intermediates. Moreover, this is only a general trend | as the southernmost population (Guds) is predom- inantly, although not exclusively, light-flowered. In addition, nearly half of the populations (11 of 24). contain both light- and dark-flowered individuals, | thus having individuals with sepal colors consid- ered to represent different subspecies. The complex pattern of variation observed for sepal color on San’ Clemente Island may be due to hybridization and subsequent introgression between the taxa. Genetic ' data may provide evidence regarding this possibil- ity. The discrepancy between our results and the tax- | onomic separation of D. v. kinkiense and D. v. thor- nei is probably due to our intensive sampling of all natural populations of D. variegatum on San Cle- | mente Island. The taxonomic descriptions of these | 2000] } taxa are based on examination of herbarium spec- ‘imens (M. Warnock, University of Missouri, Co- ‘Jumbia, pers. comm.) that may have represented ‘only a fraction of the variation found in natural populations. This is clearly a potential problem with any taxon endemic to remote locations, es- pecially in cases where access is highly restricted ‘(permission of the U.S. Navy is required to visit San Clemente Island). Classification of San Clemente Island popula- ‘tions. Because D. v. kinkiense is listed as endan- gered and D. v. thornei is merely considered a spe- ‘cies of concern by the USFWS, it is necessary for practical reasons to identify populations on San 'Clemente Island. Since our analyses show poor sep- aration between subspecies for lateral sepal and lower petal blade lengths, sepal color was used to classify populations as D. v. kinkiense, D. v. thor- nei, or mixed. Populations having at least 80% light-flowered individuals (brightness values be- tween 88 and 100) were classified as D. v. kin- _kiense, and populations having at least 80% dark- flowered individuals (brightness values below 56) were classified as D. v. thornei. Populations with less than 80% of its individuals in either brightness range were classified as mixed. This criterion is based on the observed bimodal distribution of sepal color. Using this criterion, there are 10 populations of D. v. kinkiense, 11 populations of D. v. thornei, and 3 mixed populations (Table 2, Fig. 5). Although an 80% criterion seems to be a weak basis for distinguishing taxa, it may be preferable to a stricter classification. If Warnock’s (1990b, 1993, 1997) descriptions are interpreted in con- junction with the bimodal distribution we have doc- umented, then individuals with brightness values from 88 to 100 may be classified as D. v. kinkiense (white to light blue flowers) and individuals with brightness values below 88 may be classified as D. v. thornei (light blue to bright blue flowers). Using this criterion, there are 7 populations of D. v. kin- kiense, 5 populations of D. v. thornei, and 11 mixed populations (Fig. 5). The results of this study indicate that D. v. kin- kiense and D. v. thornei are, at best, currently sep- arable only on the basis of sepal color. They may be more appropriately classified as varieties rather than subspecies or classified together as one sub- species (as defined by Stuessy 1990). However, it is not uncommon for plant taxa to be separated on the basis of morphological characters controlled by only one or two loci, such as flower color (Bach- mann 1983; Gottlieb 1984; Hilu 1983). Moreover, other characters such as flowering time may clearly distinguish D. v. kinkiense and D. v. thornei. The northern populations of predominantly light-flow- ered individuals flower earlier than the southern, dark-flowered populations (S. Junak, Santa Barbara Botanic Garden, pers. obs.), although this may have an environmental rather than a genetic basis. Fur- DODD AND HELENURM: FLORAL VARIATION IN DELPHINIUM VARIEGATUM 125 ther taxonomic study using additional characters should be conducted to decide whether the island taxa have been appropriately designated as separate subspecies. Correct taxonomic designation has practical implications for the survival of these taxa because only D. v. kinkiense has legal protection. ACKNOWLEDGMENTS The authors thank Steve Burckhalter, Steve Junak, Liz Kellogg and Jane Rombouts for providing many popula- tion locations on San Clemente Island, Toni Corelli and Mike Warnock for providing population locations on the mainland, and Jan Larson and Jennifer Stone for logistic support. This research was funded by the Natural Re- sources Office, Staff Civil Engineer, Naval Air Station, North Island, San Diego, California. LITERATURE CITED ADOBE PHOTOSHOP. 1995. Adobe Photoshop computer software, version 3.0. Adobe Systems Incorporated, San Jose, CA, USA. BACHMANN, K. 1983. Evolutionary genetics and the ge- netic control of morphogenesis in flowering plants. Evolutionary Biology 16:157—208. GoTTLigeB, L. D. 1984. Genetics and morphological evo- lution in plants. American Naturalist 123:681—709. HiLu, K. W. 1983. The role of single-gene mutations in the evolution of flowering plants. Evolutionary Bi- ology 16:97-128. JUNAK, S. A. AND D. H. WILKEN. 1998. Sensitive plant status survey, Naval Auxiliary Landing Field, San Clemente Island, California. Santa Barbara Botanic Garden Technical Report No. 1, Santa Barbara, CA, USA. KRUCKEBERG, A. R. 1954. The ecology of serpentine soils. III. Plant species in relation to serpentine soils. Ecol- ogy 35:267—274. Lewis, H. AND C. EPLING. 1954. A taxonomic study of Californian Delphiniums. Brittonia 8:1—22. Mayer, M. S., P. S. SOLTIS, AND D. E. SottTis. 1994. The evolution of the Streptanthus glandulosus complex (Cruciferae): genetic divergence and gene flow in ser- pentine endemics. American Journal of Botany 81: 1288-1299. Munz, P. A. 1969. California miscellany VII. Aliso 7:65— 71. . 1974. A flora of southern California. University of California Press, Berkeley, California, USA. RAVEN, P. H. 1963. A flora of San Clemente Island, Cal- ifornia. Aliso 5:289—347. AND D. I. AXELROD. 1978. Origin and relationships of the California flora. University of California Press, Berkeley, CA, USA. ROYAL HORTICULTURAL SOCIETY. 1986. Royal Horticultur- al Society colour chart (editions 1,2). Royal Horti- cultural Society, London. SKINNER, M. W. AND B. M. PAVLIK. 1994. California Na- tive Plant Society Inventory of rare and endangered vascular plants of California. California Native Plant Society, Sacramento, CA, USA. STUEsSsyY, T. E 1990. Plant taxonomy: the systematic eval- uation of comparative data. Columbia University Press, New York. SYSTAT. 1992. SYSTAT: Statistics, Version 5.2 Edition. SYSTAT, Inc., Evanston, Illinois, USA. 126 MADRONO [Vol. 47 }) WARNOCK, M. J. 1990a. New taxa and combinations in ed. The Jepson manual: higher plants of California, North American Delphinium (Ranunculaceae). Phy- 916-922. University of California Press, Berkeley, | tologia 68:1-6. CA. . 1990b. Taxonomic and ecological review of Cal- . 1997. Delphinium. Pp. 196—240 in Flora of North | ifornia Delphinium. Collectanea Botanica 19:45—74. America Editorial Committee ed. Flora of North | . 1993. Delphinium. Pp. 916—922 in J. C. Hickman America, Volume 3. Oxford University Press, NY. | | \ |'MADRONO, Vol. 47, No. 2, pp. 127-133, 2000 CROWN STRUCTURE OF THE WORLD’S SECOND LARGEST TREE STEPHEN C. SILLETT AND JAMES C. SPICKLER Department of Biological Sciences, Humboldt State University, Arcata, CA 95521 ROBERT VAN PELT College of Forest Resources, Box 352100, University of Washington, Seattle, WA 98195 ABSTRACT We studied the crown structure of the Washington Tree (Sequoiadendron giganteum (Lindley) Buch- holz) in Sequoia National Park, California. The tree was 77.3 m tall and 9.1 m basal diameter. Its total volume was 1403.2 m?, including the main trunk (1357.3 m*) and 46 reiterated trunks (45.8 m?). The main trunk was hollow, and 133.2 m?* of wood volume was missing. A 35-m deep, 2—3-m wide pit extended into the heart of the main trunk below 58 m. The microclimate at the bottom of the pit was dark, cool, and humid. Fire and fungal decay apparently contributed to the formation of the pit. Some charred wood was evident throughout the pit, but most of this had fallen away and been replaced by decayed wood. The walls of the pit in the lower 17 m were spongy, wet, and covered by fungal mycelia. Sequoiadendron giganteum (Lindley) Buchholz (giant sequoia) is an awe-inspiring species restrict- ed to 66 groves in California’s Sierra Nevada (Wil- lard 1995). These are the world’s largest living trees (Flint 1987), and some individuals are over 3000 years old (Hartesveldt et al. 1975). As such, they have attracted a great deal of scientific interest. They were among the earliest trees in North Amer- ica to be studied from a canopy perspective; rope techniques and an elevator were used to sample cones and arboreal arthropods in the 1970’s (Har- vey et al. 1980). However, the crown structure of ancient S. giganteum trees has never been studied. The crown structure of ancient Sequoia semper- virens (D. Don) Endl. trees, the closest living rel- atives of S. giganteum, has recently been the focus of research (Sillett 1999). Like many other conifers, including S. giganteum, S. sempervirens grows via a simple architectural model consisting of a vertical trunk that supports numerous horizontal branches. Ancient trees, which have endured centuries of wind and fire, develop highly individualized crowns consisting of multiple, resprouted trunks (1.e., vertically oriented stems) arising from other trunks and branches. One very complex S. semper- virens tree, for example, has a crown with 148 res- prouted trunks accounting for over 14 percent of its total wood volume (Sillett and Van Pelt 2000). Such extra trunks are reiterations of the tree’s ar- chitectural model (Hallé et al. 1978). They are in- distinguishable from free-standing trees except for their locations within the crown of a larger tree. Each reiterated trunk supports its own system of horizontal branches. This is the first rope-based study of crown struc- ture in an ancient S. giganteum tree. We used meth- ods developed in S. sempervirens to map the crown of the Washington Tree, the second largest S. gi- ganteum tree (Flint 1987). Our objective was to cal- culate the tree’s total volume, including reiterations. During our exploration of the tree, we discovered a deep pit extending into the heart of the tree’s main trunk. We compared microclimatic conditions in- side the pit with those on top of the crown. STUDY AREA The Washington Tree is located at 2085 m ele- vation near the center of Giant Forest in the south- ern Sierra Nevada of California (36°33’N, 118°45’'W). At 1212 hectares, Giant Forest is the second largest of the unlogged S. giganteum groves (Willard 1995). The average temperature ranges from 0.2°C in January to 17.9°C in July with an annual average of 8.1°C. Much of the area’s 110 cm precipitation falls as snow (482 cm average ac- cumulation), and a snowpack persists into spring. The summer is very dry; only 2 cm of precipitation falls between July and September (National Park Service). The Washington Tree grows near a large granite outcrop in a forest dominated by S. giganteum, Abies concolor (Gordon & Glend.) Lindley, and Pi- nus lambertiana Douglas. The area has been burned within the last 15 years as part of a prescribed burn- ing program carried out by the National Park Ser- vice. Abundant regeneration of S. giganteum and P. lambertiana is visible in the vicinity of the Washington Tree and the nearby Franklin Tree. METHODS Tree access. We accessed the Washington Tree by shooting a rubber-tipped fiberglass arrow trailing 10 kg test strength Fireline® filament over sturdy branches with a compound bow mounted to a spin- ning reel. A 3 mm nylon cord, followed by a 10 mm static kernmantle climbing rope, was then hauled over the branches. We anchored one end of 128 MADRONO [Vol. 47 the rope at ground level and climbed the other us- ing mechanical ascenders. We used a 20 m long arborist’s rope lanyard to access progressively high- er branches and to move laterally through the crown. We worked in the tree crown during a two- week period from May to June 1999. Crown mapping. We mapped crown structure of the Washington Tree by measuring dimensions of the main trunk and all reiterated trunks (both living and dead) over 5 cm basal diameter. All trunk di- ameters were measured directly using a graduated tape. The main trunk’s diameter was measured at the highest ground surface, which was 0.65 m above true ground level (i.e., the average of the © highest and lowest ground surfaces), at 2.5 m in- | tervals from true ground level to 15 m, and at5m_ | intervals above 15 m. Branches and reiterated | trunks prevented us from obtaining diameter mea- | surements at every height interval. In these cases, we measured trunk diameter as close to the regular interval as possible. Since the main trunk was hol- low (see RESULTS), we also measured the maxi- mum and minimum diameters of the hollow cavity | at 5 m intervals. A series of reiterated trunks extended well above the broken top of the main trunk. The tree’s total height was measured by lowering a tape from the topmost foliage to true ground level. We recorded the following data for each reiterated trunk: top | height, height of origin, basal diameter, and diam- eter at 5 m intervals along the length of the trunk. | For reiterated trunks arising from branches, we also recorded horizontal distance to main trunk, branch height, branch basal diameter, and branch diameter at reiteration. We only surveyed branches giving rise to reiterated trunks; no other branches were measured. Reiterated trunks were referenced to the main trunk by recording azimuths and distances be- tween them at 5 m height intervals. We used an Impulse® laser range finder (Laser Technology, | Inc.) to measure horizontal distances between _ trunks. All other measurements were made with the | aid of a compass, clinometer, and graduated tape. Since large trunks often gave rise to smaller trunks, we sketched crown structure and noted physical connections between trunks and branches. We also noted whether trunks were monopodial (1.e., con- sisting of a single axis), sympodial (1.e., consisting of successive axes), or otherwise broken. All in- formation was used to generate a tree crown dia- gram. Crown illustrations. We made an illustration of the entire Washington Tree from the ground. Major branches, clumps of foliage, burls, and kinks were Sa Fic. 1. Illustration of the Washington Tree prepared by Robert Van Pelt on the basis of crown structure data and photographs taken from the ground. 2000] noted on a sketch, and heights to these landmarks were measured with the Impulse® laser. We then took photographs of all portions of one side of the crown from as far away from the tree as possible. The illustration itself started with a ‘“‘skeleton,”’ which was based on the height and diameter mea- surements taken both from within the crown and from the ground. We used the photographs to pro- vide details of foliage, branches, and bark texture. The illustration was drawn at Y,,. scale, and human figures were added for additional scale. We also made a detailed illustration of the upper crown, in- cluding the entrance to the pit. Photographs and sketches made from a nearby hill supplemented sketches made from within the tree crown. Microclimate sampling. We measured light, air temperature, and relative humidity at three posi- tions in the tree’s crown. Sensors were lashed to the tree’s uppermost branches with nylon cord to obtain measurements from the top of the crown (77 m). Sensors were suspended on a rope to obtain measurements from the top of the pit (57 m) and the bottom of the pit (23 m). Microclimate mea- surements were made at 4-minute intervals over a 23-hour period (11 a.m. June 11 to 10 a.m. June 12). We used HOBO® RH/Temp Loggers to mea- sure air temperature and relative humidity and StowAway® Light Intensity Loggers to measure light intensity (Onset Computer Corporation). Calculations. We used Wendell Flint’s ground- level survey data for Washington Tree (Flint un- published) to determine the basal diameter of the main trunk. Direct tape measurements overestimate basal area by failing to account for missing wood in fire cavities and spaces between buttresses. We calculated the surface area of the tree’s ‘‘footprint”’ and converted this to a true basal diameter (diam- eter = 2[footprint area/t]”). Trunk volume was cal- culated by applying two different equations to the trunk diameter data. We used the equation for a parabolic frustum (i.e., volume = length/2*[Al + A2], where Al and A2 are the upper and lower trunk cross sectional areas) for sections of trunks that tapered slowly. We used the equation for a reg- ular conic frustum (i.e., volume = length*q/ 3*[lower diameter? + (lower diameter)*(upper di- ameter) + upper diameter’]) for sections of trunks that tapered more rapidly, such as within the crown. Using the latter equation, we also calculated the volume of branches supporting reiterations based on the limited data we collected on these branches (see above). RESULTS Volume. The Washington Tree is 77.3 m tall and 9.1 m basal diameter (Fig. 1). A large fire cave has been burned away from the main trunk near the ground. It consists of an inner portion that is 3.7 m high, 2.4 m wide, and 2.6 m deep as well as an outer portion that is 5.8 m high, 4.2 m wide, and SILLETT ET AL.: SECOND LARGEST TREE’S CROWN STRUCTURE 129 0.5 m deep. Above 58 m, the main trunk is broken (Fig. 2). A shield of wood extends 10 m above this break and terminates two structures: a partially dead reiteration and the splintered remains of the main trunk. According to our calculations, the Washington Tree’s main trunk has a volume of 1357.3 m*, including the volume occupied by the pit (see below). Reiterated trunks (see below) add an additional 45.8 m? of volume, the three largest reiterations accounting for 79.6 percent of this total. Thus, the Washington Tree’s total trunk volume is 1403.2 m*. Branches bearing reiterations on the Washington Tree (n = 22) add an additional 25.6 m?* volume, but we emphasize that our measure- ments of branch volume are incomplete. Reiterations. The Washington Tree has 46 reit- erated trunks arising from the main trunk (n = 3), other trunks (n = 10), and branches (n = 33) (Fig- ure 2). The largest reiteration, which is 1.7 m basal diameter and 16 m long, sprouts from the main trunk at the mouth of the pit and terminates in a dead broken stump that is 0.8 m diameter. It sup- ports 19 other trunks, including two dead and seven sympodial trunks. Like the main trunk, the largest reiterated trunk is hollow with a lesser pit that is 5.1 m deep. Another large reiteration sprouts from the end of a 1.1 m diameter branch emanating from the main trunk at the mouth of the pit. Its top is dead, and it supports four other trunks. Eight trunks emerge from the backside of the shield above the pit. Two of these trunks are completely dead, and two have dead tops. All of the rest of the Washing- ton Tree’s reiterations arise from branches emanat- ing from the main trunk. The largest of these trunks is 1 m basal diameter and 25.2 m long. It sprouts from the end of a 4 m long, 1.2 m diameter branch at 35.7 m above the ground. The Washington Tree’s largest branch, which supports three trunks, is 1.4 m basal diameter, 10 m long, and located 45.4 m above the ground. The pit. A 35-m-deep pit extends down into the heart of the Washington Tree from the break in its main trunk below the shield (Figs. 2 and 3). The pit is over 2 m diameter at the mouth, and it en- larges to over 3 m diameter farther down. At the bottom (i.e., 22.8 m above ground level), the pit is 3.05 by 0.62 m wide. Charred wood is evident along the walls of the pit throughout most of its length, and fungal decay becomes pronounced with increasing depth. Below 40 m, fungal mycelia are evident along the walls, and the wood is soft and wet. Humus has accumulated on ledges of rotting wood. Massive protrusions of dead wood extend from the walls of the main trunk into the pit, evi- dence of ancient branch bases. At the bottom of the pit, these branch bases are small, and the humus is deep, rich, and filled with rotting seed cones. An- nual rings are visible in the spongy wood, some of them perhaps 3000 years old (Stephenson in press). The pit occupies 133.2 m*, or about 10 percent, of 130 MADRONO [Vol. 47 | 80 70- wn i) | = —_—_ mo) = Oo KEY jot 8) o 40- ui > oO 2 fae) oc cong 2 30- : 20+ : 10- ; 0 ] le, a ae Sle pol 10 9 8 7 6 5 4 3 2 1 O Pood 93. aD GO, 17, 8D eO, OM 2 distance from central axis (m) Fic. 2. Crown diagram of the Washington Tree. All trunk diameters are drawn to the scale of the x-axis, which is expanded relative to the scale of the y-axis. Circles correspond to the basal diameters of reiterated trunks. Serrated edges indicate broken trunks. Branches bearing reiterated trunks are depicted with single straight lines. No other branches are shown. Gray lines indicate dead branches and trunks. Gray lines also indicate a basal fire cavity as well as maximum and minimum diameters of the pit. the main trunk’s volume. Thus, the Washington Tree’s total wood volume is 1270.0 m‘°. There are two narrow fissures in the main trunk far above the bottom of the pit (i.e., 59.5—-51.7 m and 36.3—33.4 m). The upper fissure is located in a large area of dead wood on the outside of the main trunk that extends from the shield to 49.4 m above the ground. Some light and wind pass into the pit through the opening, which is up to 10 cm wide. The lower fissure is much narrower. It affords no breeze, but there is enough light to support an epix- ylic green alga on the inner surface of the opening. Microclimate. At the bottom of the pit, the mi- croclimate was relatively dark, cool, and humid during the 23-hour sampling period. Light intensity during the day was less than 0.1 percent as high as on top of the crown and less than 1| percent as high as on top of the pit. Temperatures at the bottom of the pit remained a constant 4°C for the duration of the sampling period. The other locations were 4.5 to 11°C warmer than the bottom of the pit. The top of the pit was 0.5 to 3°C cooler during the day and up to 2°C warmer during the night than the top of the crown. Relative humidity at the bottom of the SILLETT ET AL.: SECOND LARGEST TREE’S CROWN STRUCTURE 131 2000] from a location 120° clockwise of the including the mouth of the pit, prepared by Robert > s. This view 1s S upper crown I Van Pelt on the basis of crown structure data and photograph Fic. 3. Detailed illustration of the Washington Tree view in Figure 1. 132 MADRONO pit remained a constant 99 percent for the duration of the sampling period. The other locations were 23 to 60 percent less humid than the bottom of the pit. The top of the pit was 5 to 25 percent more humid during the day and up to 10 percent less humid during the night than the top of the crown. DISCUSSION Tree size. The Washington Tree has been known as the world’s second largest tree for over a decade (Flint 1987), and our direct crown-level measure- ments confirm this fact. Among living trees, only the General Sherman Tree, whose main trunk has a volume of 1489 m/’, is larger (Van Pelt in press). The main trunk of the General Grant Tree, whose volume is 1357.3 m+, is identical in volume to the Washington Tree’s main trunk, but the General Grant Tree has a smaller volume of reiterated trunks (Van Pelt in press). However, the Washing- ton Tree’s total wood volume, including reitera- tions, is actually less than that of the General Grant Tree, the President Tree (1318.0 m’*, Van Pelt in press), and the Lincoln Tree (1275.2 m?, Van Pelt in press) because its main trunk is hollow. Among living S. giganteum, many trees are taller than the Washington Tree, and several trees (e.g., Ishi, Grant, and Boole) have larger bases (Van Pelt in press). The Washington Tree, however, is the largest living tree from 3.2 to 4.3 m (main trunk diameters 7.6 to 7.0 m) and from 50.0 to 58.5 m (main trunk diameters 4.6 to 3.8 m) above the ground. Prior to losing the upper part of its main trunk, the Washington Tree may have been the only living tree larger than the General Sherman Tree (Flint 1987). Crown structure. Like their close relative, Se- quoia sempervirens, ancient Sequoiadendron gi- ganteum trees can have complex crowns consisting of multiple reiterated trunks. Ancient trees of both redwood species frequently possess branches that gradually curve upwards and become increasingly trunk-like over time. Perfectly vertical reiterated trunks, however, appear to be less common on S. giganteum than on S. sempervirens. And unlike S. sempervirens, Old S. giganteum branches tend not to be heavily buttressed; they appear almost circular in transverse section. Furthermore, flagelliform branches and fusions between trunks and branches, which are common in ancient S. sempervirens crowns (Sillett 1999; Sillett and Van Pelt 2000), are rarely encountered on S. giganteum. It is difficult to compare the Washington Tree’s crown structure with that of other ancient S. gigan- teum trees because no others have been thoroughly mapped using rope-based methods of access. Ground-level surveys permit a few preliminary comparisons. Unlike the Washington Tree, several of the largest S. giganteum trees (e.g., Grant, Lin- coln, Stagg, Boole, Genesis) have very few (if any) reiterated trunks (Van Pelt in press). The General [Vol. 47 Sherman Tree’s crown, however, is highly reiterat- ed, and a few of these reiterations are larger than any on the Washington Tree (Van Pelt in press). The Franklin Tree (1222.7 m* volume, Van Pelt in press), which grows within 500 m of the Washing- ton Tree, also has a highly reiterated crown, but nearly all of its large reiterations are dead. Fire and fungi. Tops of the main trunks on many ancient S. giganteum trees are dead (Rundel 1973), and most trees have fire scars throughout their crowns. The Washington Tree is no exception. Trunks that have been hollowed out by fire are also commonly observed, but the Washington Tree’s pit may be unique. In reference to hollow trunks, John Muir wrote, “‘All of these famous hollows are burned out of solid wood, for no Sequoia is ever hollowed by decay”’ (1909). He clearly never ob- served the Washington Tree’s pit! There is no doubt that fungal decay has played a major role in creating the pit; the heartwood is rotting and spongy, and fungal mycelia are abun- dant. Indeed, microclimatic conditions in the lower portion of the pit promote fungal decay. After the top of the main trunk broke away, fire probably initiated formation of the pit, perhaps via a mech- anism similar to the one Muir observed on fallen logs during a fire (1909): After the great glowing ends fronting each other have burned so far apart that their rims cease to burn, the fire continues to work on in the centres, and the ends become deeply concave. Then heat being radiated from side to side, the burning goes on in each section of the trunk independent of the other, until the diameter of the bore is so great that the heat radiated across from side to side is not sufficient to keep them burning. It ap- pears, therefore, that only very large trees can receive the fire-auger and have any shell rim left. But precipitation accumulated in the newly formed pit, and the moist wood was ultimately colonized by decay fungi that increased its size over many years. Subsequent fires probably consumed much of the decaying wood (Piirto et al. 1984). We ob- served some charred wood on the walls of the pit to within 0.7 m of the bottom, so fire has clearly contributed to the hollowing of the pit to a great depth. However, most of the charred wood has fall- en away and been replaced by decaying wood. Thus, both fire and fungal decay were directly in- volved in the formation of the pit. ACKNOWLEDGMENTS Funding from Global Forest supported Sillett during this research (GF-18-2000-48). MacGillivray-Freeman Films and the Adventures in Wild California film crew provided logistical support. Dana Laughlin, Brett Love- lace, and Shanti Revotskie assisted with the field work. Erik Jules, Nathan Stephenson, and two anonymous re- | 2000] | viewers provided constructive comments on the manu- script. Finally, Wendell Flint graciously provided us with _ his ground-level survey data for the Washington Tree. LITERATURE CITED _ FLINT, W. D. 1987. To find the biggest tree. Sequoia Nat- ural History Association. Three Rivers, CA. HALLE, F, R. A. A. OLDEMAN, AND P. B. TOMLINSON. 1978. Tropical trees and forests: an architectural analysis. Springer-Verlag, New York. HARTESVELDT, R. J., H. T. HARVEY, H. S. SHELLHAMMER, AND R. E. STECKER. 1975. The giant sequoia of the Sierra Nevada. U.S. Department of the Interior, Na- tional Park Service, Washington, DC. Harvey, H. T., H. S. SHELLHAMMER, AND R. E. STECKER. 1980. Giant sequoia ecology. U.S. Department of the Interior, National Park Service, Scientific Monograph Series 12. Washington, DC. Muir, J. 1909. Our National Parks. The Riverside Press, Cambridge. SILLETT ET AL.: SECOND LARGEST TREE’S CROWN STRUCTURE 133 PurRTO, D. D., W. W. WILcox, J. R. PARMETER JR., AND D. L. Woop. 1984. Causes of uprooting and breakage of specimen giant sequoia trees. Bulletin 1909. Division of Agriculture and Natural Resources, University of California, Berkeley. RUNDEL, P. W. 1973. The relationship between basal fire scars and crown damage in giant sequoia. Ecology 54:210-213. SILLETT, S. C. 1999. Tree crown structure and vascular epiphyte distribution in Sequoia sempervirens rain forest canopies. Selbyana 20:76—97. SILLETT, S. C. AND R. VAN PELT. 2000. A redwood tree whose crown is a forest canopy. Northwest Science 74:34-43. STEPHENSON, N. L. 2000. Estimated ages of some large giant sequoia: General Sherman keeps getting youn- ger. Madrono (in press). VAN PELT, R. In press. Forest giants of the Pacific coast. University of Washington Press, Seattle. WILLARD, D. 1995. Giant sequoia groves of the Sierra Ne- vada: a reference guide. Dwight Willard, Berkeley, CA. MADRONO, Vol. 47, No. 2, pp. 134-137, 2000 ERIOGONUM SPECTABILE (POLYGONACEAE): A NEW SPECIES FROM NORTHEASTERN CALIFORNIA BETH LOWE CORBIN Lassen National Forest, 2550 Riverside Dr., Susanville, CA 96130 JAMES L. REVEAL Norton-Brown Herbarium, University of Maryland, College Park, MD 20742-5815 ROBIN BARRON 1731 Country Lane, Placerville, CA 95667 ABSTRACT Eriogonum spectabile, a new species of the subgenus Eucycla, is described from northeastern Plumas County in northeastern California, USA. It differs from the related E. pendulum of northwestern California and adjacent southwestern Oregon in being a shorter, more compact plant with more numerous branches at the base, narrower leaves with the pubescence equally distributed on both surfaces, longer petioles, reduced umbellate inflorescences, broadly campanulate involucres, and densely pubescent flowers with gland-tipped hairs among the silky-white ones. The new species is currently known only in an extremely limited area of glaciated andesite southeast of Lassen Peak. A new species of Eriogonum Michx. was en- countered during project field surveys on the Las- sen National Forest in northeastern California on 30 July 1997. This plant was immediately recog- nizable as distinctly different from other known Er- iogonum species in this part of the state, and indeed from all other known Eriogonum by the combina- tion of its low shrubby habit and densely pubescent flowers and fruits. Subsequent surveys have result- ed in a total of only three occurrences with about 250 plants total, all within about 1 km of each oth- Cr Eriogonum spectabile B. L. Corbin, Reveal, & R. Barron, sp. nov. (Fig. 1).—TYPE: USA, Califor- nia, Plumas Co., ca. 13 km north of Chester, ca. 1.9 km west-southwest of Hay Meadows trail- head to the Caribou Wilderness, Lassen National Forest, T30N, R7E, sect. 28 NE % of SW %, MDM, ca. 40°25’N, 121°12’W, 18 Aug 1998, Corbin and Earll 910 (Holotype: US; Isotypes DAV, JEPS, K, MARY, NY, RSA, Lassen Na- tional Forest herbarium.) Planta perennis, suffrutex, usque 2.5 cm alta; caules patentes dense ramosi fragiles; petioli breves 0.6—0.9 cm longi; lamina foliaris anguste elliptica, 0.5—1.7 cm longa, 0.4—0.7 cm lata, in superficiebus ambabus sericea cinereo-tomentosa, marginibus in- tegris planis vel revolutis praedita. Caules florentes Scaposi, primo albi-tomentosi deinde glabrati; inflo- rescentia umbellata; bracteae semifoliaceae, angus- ti-oblongae vel-ellipticae, albi-tomentosae; involu- cra solitaria, extus sericeo-tometosa atque dense gladularia intus glabra; dentes involucrales 5-7, acute triangulares, 1.0—1.2 mm longi; bracteolae numerosae lineares. Flores sub anthesi albi, sub gemmascentia frutescentiaque rosei vel subrubri, nervo medio fuscato ornati; tepala extus dense hir- suta intus subglabra, per anthesin 4.0—4.5 mm lon- ga; stamina inclusa vel exserta; antherae rubrae vel purpureae. Achenia 3—4 mm longa, tomentosa, ad basem subglobosam in rostrum crassum 3-angulum contracta. Plants low shrubs 1—1.5 dm high vegetatively and 1.7—2.5 dm high in flower, mostly 3—5.5 dm across; stems spreading, densely branched, brittle, arising from a stout, woody taproot (up to 5 cm across at the top of the taproot), the older branches with reddish-brown bark exfoliating in wide strips. Leaves arranged in open rosettes mostly at the base of the flowering stem or at the tips of exposed caudex branches, others sheathing shortly up the herbaceous stems; leaf blades narrowly elliptic, (0.7—)1.2—1.7(—2.2) cm long, (0.2—)0.4—0.7(—0.9) cm wide, equally densely gray and somewhat silky tomentose on both surfaces; leaf margins entire and plane to revolute; leaf apex broadly acute, the base cuneate. Petiole short, (0.2—)0.6—0.9(—1.3) cm long, silky tomentose. Petiole base elongate triangular, 3— 5 mm long, 2—4 mm wide, densely white tomentose without, glabrous within. Flowering stems scapose, erect (1.5—)6.1—13.3(-17) cm long, white tomen- tose, becoming glabrate at maturity. Inflorescences umbellate. Bracts mostly semifoliaceous, 2—4(—6), narrowly oblong to narrowly elliptic, 2-5 mm long, 1—2 mm wide, white tomentose. Peduncles slender, slightly spreading, (1.5—)2—8.5(—9.5) cm long, silky tomentose to glabrate at maturity, occasionally some peduncles have an extra whorl of bracts mid- length. Involucres solitary, broadly campanulate, the tube 2—3 mm long, (2—)3—4(—5) mm wide, silky tomentose and densely glandular without, glabrous 2000] CORBIN ET AL.: ERIOGONUM SPECTABILE 135 1cm 1cm Fic. 1. Eriogonum spectabile. Left-habit. Right-involucre. Drawn from Corbin 906. 136 MADRONO [Vol. 47 | within; involucral teeth 5—7, acutely triangular, 1— 1.2 mm long; bractlets numerous, linear, 2—3 mm long, fringed with long, silky hairs and minute gland-tipped cells. Pedicels 2.5—-4 mm long, gla- brous below, glandular and slightly hairy above. Flowers white (in anthesis) to pink or reddish Gn fruit, also in bud). Stipe essentially lacking, 0.1—0.2 mm long. Tepals 6, with slightly darker greenish to reddish bases and midribs (midribs greenish within and dark pink without), 4-6 mm long (mostly 4— 4.5 mm in flower, longer in fruit), densely hairy without with long, slender, silky-white hairs and short, capitate glands, essentially glabrous within except for minute glands and some hairs mainly along the midrib; tepals essentially similar, obovate, those of the inner whorl slightly longer than the outer whorl, united for less than % of their length. Stamens ? included to exerted, 2.5—3(—6.5) mm long, the filaments sparsely pilose at the base; an- thers red when fresh, purplish-red to purple when dried, 0.6—0.7 mm long, broadly ovate. Gynoccium with a style 1—1.3(-1.5) mm long. Achenes light brown, 3—4 mm long, tomentose, the subglobose base tapering to a stout, 3-angled beak. Paratypes. Topotypes—30 Jul 1997, Barron s.n. (Lassen National Forest herbarium), 18 Aug 1997, Barron s.n. (CHSC, JEPS, MARY, Lassen National Forest herbarium), 8 Sep 1997, Corbin et al. 861 (MARY), 8 Aug 1998 Corbin et al. 906 (Lassen National Forest herbarium). USA, California: Plu- mas Co., ca. 2.2 km SW of Hay Meadows trailhead to the Caribou Wilderness, T30N, R7E, sect. 28 SW % of SW %, MDM, 23 Sep 1997 Corbin 882 (MARY, Lassen National Forest herbarium); ca. 2.2 km WSW of Hay Meadows trailhead, T30N, R7E, sect. 28 NW % of SW %, MDM, 23 Sept 1997 Barron s.n. (Lassen National Forest herbarium). Eriogonum spectabile is most closely related to E. pendulum S. Watson; both are members of an as yet undescribed section of the subg. Eucycla (Nutt.) Kuntze. The new species is a shorter, more compact plant than E. pendulum with more numerous branches at the base, narrower leaves with the pu- bescence equally distributed on both surfaces, lon- ger petioles, reduced umbellate inflorescences, broadly campanulate involucres, and densely pu- bescent flowers with gland-tipped hairs among the silky-white ones. Eriogonum pendulum is found in extreme northwestern California (Del Norte Co.) and adjacent southwestern Oregon (Josephine Co.) in dry sandy soil in mixed evergreen forests not unlike that of E. spectabile (see below). The spe- cific epithet refers to the spectacular appearance of this small shrub, which is quite attractive. We sug- gest the common name ‘“Barron’s buckwheat’’ to acknowledge the first collector. Eriogonum spectabile appears to be limited to three small occurrences, all within one quarter sec- tion. The first discovered occurrence (and the type locality) is the largest; 194 plants were counted on 8 Sep 1997. The second occurrence had 54 plants | on 23 Sep 1997, and the third only three (also on | 23 Sep 1997). Numbers were similar in 1998 and | 1999 visits. All known locations are on the Lassen | National Forest in northeastern California within | about 5 km of Lassen Volcanic National Park, and about 1.2 km from the Forest’s Caribou Wilderness Area. This region is considered part of the southern limit of the Cascades Range. Extensive searches in | adjacent Lassen National Forest areas were made in 1997, 1998, and 1999. Lassen Volcanic National _ Park has been fairly well botanized in the past (Os- | wald et al. 1995, Gillett et al. 1961), but no collec- | tions from Lassen Volcanic National Park are | known. The Caribou Wilderness Area contains ex- | tensive apparently suitable habitat, much of which ! has not been surveyed, so Eriogonum spectabile — may occur there as well. | The new species grows in open areas on minor | ridges within a Pinus contorta Loudon. subsp. mur- | rayana (Grev. & Balf.) Critchf., Abies magnifica | Andr. Murray var. magnifica, and A. concolor (Gor- don & Glend.) Lindley forest, at 2010 to 2025 m | elevation. The general area is Quaternary glacial | deposits (Lydon et al. 1976), with moraines form- | ing low ridges interspersed by several small kettle | lakes. Glaciation is particularly evident at the type | locality, as shown by glaciated andesite bedrock or | large boulders with smooth and striated surfaces, and chatter marks on the larger surfaces. Arctostaphylos nevadensis A. Gray, which is abundant in this general area, is the species most closely associated with Eriogonum spectabile; how- ever, E. spectabile occurs only in the less common — openings between individuals of Arctostaphylos nevadensis. Other less abundant associates include Achnatherum occidentale (Thurbes) Barkworth subsp. californicum (Merr. & Burtt Davy) Bark- worth, Arctostaphylos patula E. Greene, Ceanothus prostratus Benth., Cymopterus terebinthinus (Hook.) M. E. Jones var. californicus (J. Coulter & Rose) Jepson, and Helianthella californica DC. var. nevadensis (Greene) Jeps. The area receives about 60 inches (150 cm) of precipitation per year, mostly as snow (Ranz 1969). Plants were in late flower on 30 Jul 1997 and 18 Aug 1997 visits, and in fruit in early Sep 1997. The spring of 1998 was much colder. The site was snow-covered on 17 Jun 1998; the plants were mostly vegetative on 22 Jul 1998, in bud on 8 Aug 1998, and in full bloom on 18 Aug 1998. Plants were in full flower on 26 Jul 1999, and still flow- ering and in early fruit on 31 Aug 1999. At all three occurrences, most individuals ap- peared to be mature shrubs, and many had dead wood about the base. Few seedlings (only four at the largest occurrence in 1998, and none at the smallest) or apparently young plants were ob- served. Not all plants at an occurrence flowered: only 18 percent of the second occurrence flowered in 1997, but 60 percent of the adult plants at the ~ 2000] _ type locality flowered in 1998. The apparent low _ rate of recruitment may indicate an uncertain future _ for this species. Although Eriogonum spectabile occurs relatively close to a wilderness area and national park, it is on national forest land without special designation. No human disturbance was observed at the three sites where the plant was found, but adjacent areas have been logged extensively. Firewood cutting is also common and numerous skid trails and wood- cutter roads criss-cross the general area. Besides potential human impacts, the low numbers and lim- ited distribution of E. spectabile suggest it is at risk of extinction from natural habitat changes. One change may be an increase in competing vegetation (particularly Arctostaphylos nevadensis, but also other shrubs and the coniferous overstory), perhaps due to climate change and/or a change in the fire regime. Wildfire effects are not certain; given the plant’s shrubby, presumably non-sprouting, long- lived growth form, it is likely that fire would kill existing plants. The species would then depend on seedling recruitment from the soil seed bank (of which we have no information) or from seeding in from adjacent areas, which is highly unlikely given its rarity and lack of obvious seed dispersal mech- anism(s). Another potential effect is browsing of flowering CORBIN ET AL.: ERIOGONUM SPECTABILE 137 stems. On a 30 Sep 1998 visit to the second oc- currence, nearly all flowering stems and some of the tips of the leafy shoots had been browsed, pre- sumably by deer. The result is a virtual lack of in- tact seeds produced from this occurrence in 1998. ACKNOWLEDGMENTS Special thanks to Vernon H. Oswald and Pete Figura for their help in recognizing the distinctiveness of this plant. Thanks to P. M. Eckel for the Latin translation. Thanks to Shannon Workman for the illustration. LITERATURE CITED GILLETT, G. W., J. T. HOWELL, AND H. LESCHKE. 1961. A flora of Lassen Volcanic Park. Wasmann Journal of Biology 19(1):1-185. Lypon, P. A., T. E. GAy, JR., AND S. W. JENNINGS. 1960. Geologic Map of California, Olaf P. Jenkins Edition, Westwood Sheet. Third printing, 1976. [Map at a scale of 1:250,000.] OswaLp, V. H., D. W. SHowErRs, & M. A. SHOwerRsS. 1995. A Flora of Lassen Volcanic National Park, California. California Native Plant Society, Sacramento, Califor- nia. [Revision of the original flora by Gillett et al 1961.] RANZ, S. E. 1969. Mean annual precipitation in the Cali- fornia region. U.S. Geologic Survey, Water Resources Division, Menlo Park, California. Reprinted in 1972. [Set of two maps at approximate scale 1:990,000. ] MADRONO, Vol. 47, No. 2, pp. 138-145, 2000 NOTEWORTHY COLLECTIONS ARIZONA HEXALECTRIS REVOLUTA Correll (ORCHIDACEAE).— Pima County, Baboquivari Canyon, and McCleary Can- yon; Santa Cruz County, Sawmill Canyon. Between 1371 and 1524 meters elevation in canyon bottoms and sides of canyons, under oaks and mesquite, in association with Arizona walnut. Previous knowledge. Previously known range was lim- ited to portions of northern Mexico, and the Big Bend area of Texas. Voucher specimens of H. revoluta deposited at the University of Arizona Herbarium (ARIZ), Tucson, AZ collected in Baboquivari Canyon by Toolin in 1981 and McCleary Canyon by McLaughlin in 1986 were orig- inally identified as H. spicata. Studies of fresh material in the field by the author indicated the plants are correctly H. revoluta. Significance. First record of this species in Arizona, and represents a western range extension of approximately 290 miles (483 km) and a northern range extension of approx- imately 210 miles (350 km) from Big Bend National Park. Not known from New Mexico. Hexalectris revoluta is not currently a candidate for Federal Endangered Species sta- tus, but should be considered for listing due to rareness across its range. The McCleary Canyon location was re- cently included within the boundaries of land being con- sidered for trade from the Forest Service to a mining de- veloper. That trade is not currently under consideration. —RONALD A. COLEMAN, University of Arizona, 11520 E. Calle Del Valle, Tucson, AZ 85749. CALIFORNIA ESCHSCHOLZIA RHOMBIPETALA E. Greene (PAPAVERA- CEAE).—Alameda County: Lawrence Livermore Nation- al Laboratory, Site 300, T3S R4E, SW % Sec. 29, elev. 850 ft, on N-facing crumbling clay bank, with Poa secun- da J. S. Presl, Bromus madritensis L. subsp. rubens (L.) Husnot, Avena barbata Link, Stylomecon heterophylla (Benth.) G. C. Taylor, Microseris douglasii (DC.) Schultz- Bip., Blepharizonia plumosa (Kellogg) E. Greene, 06 May 1997, R. E. Preston 1028 (DAV). Previous knowledge. Historically known from the in- terior foothills of the Hamilton and Diablo Ranges, with disjunct occurrences on the Carrizo Plains (W. Ernst, Ma- drono 17:281—294, 1964). Believed extinct (M. Skinner and B. Pavlik, Inventory of Rare and Endangered Vascular Plants of California, 1994) until rediscovered at Carrizo Plains in 1993 by David Keil and in 1995 by Curtis Clark (C. Clark, The genus Eschscholzia: California poppies and their relatives, http://www.intranet.csupomona.edu/ ~jcclark/poppy/, 2000). Significance. First east Bay Area record since 1949. Site 300 is near Corral Hollow, where the species was last collected by Peter Raven. Subsequent attempts by Raven and Clark to relocate the Corral Hollow occurrences were unsuccessful (California Natural Diversity Database, Rar- efind 2, Version 2.1.2, March 24, 2000 update; C. Clark, personal communication). Because the plants are small, they may be easy to overlook, and the plants may only appear in favorable years (C. Clark, personal communi- cation). TRICHOCORONIS WRIGHTI (A. Gray) A. Gray (ASTERA- CEAE).—Merced County: Merced National Wildlife Ref- uge, S of Mariposa Bypass, T9S, R12E, SW % Sec. 3, elev. 100 ft, 21 May 1997, R. E. Preston 1031 (DAV. CAS). Previous knowledge. Native to Mexico, Texas. In Cal- ifornia, known from four occurrences in Riverside County and four scattered locations in the Central Valley. Cali- fornia populations are presumed to be introductions (A. M. Powell in J. C. Hickman [ed.], The Jepson Manual: Higher Plants of California, 1993), although Skinner & Pavlik (1994) suggest that the species may be native to California. Previously thought to be extirpated in the Cen- tral Valley (Skinner and Pavlik 1994). Significance. First Central Valley record since 1953. Found growing in the bypass floodplain, with Eleocharis macrostachya Britton, Xanthium strumarium L., Malvella leprosa (Ortega) Krapov., Phyla nodiflora (L.) E. Greene, Polygonum arenastrum Boreau, and Frankenia salina (Molina) I. M. Johnston. SENECIO APHANACTIS E. Greene (ASTERACEAE ).—AI- ameda County: Corral Hollow, 0.5 mi NW of Tesla town site, T3S R3E, SE %4 of NE % S26, elev. 1500 ft, scattered on barrens, with Plantago erecta E. Morris, Bromus mad- ritensis L. subsp. rubens (L.) Husnot, Erodium cicutarium (L.) LHés, Hypochaeris glabra L., Erodium botrys (Cav.) Bertol., Medicago polymorpha, Avena fatua L., 21 April 1998, Robert E. Preston 1097 (DAV, CAS). Previous knowledge. Widely but infrequently collected in the California Coast Ranges south of San Francisco Bay; the Transverse Ranges; southwest California, includ- ing Santa Cruz Island; and Baja California. This species is included in list 1B of the CNPS Inventory (Skinner and Pavlik 1994). Significance. First record for Alameda County and first San Francisco Bay Area collection since 1933. In the Jep- son Manual, this species is reported to occur on drying alkali flats (T. M. Barkley in Hickman, 1993). However, information obtained from herbarium specimens indicates that it occurs on various substrates: clay; coarse sand; rock outcrops, including serpentinite; and soils with high gyp- sum content or high alkalinity. Common to all occurrences iS a conspicuous absence of vegetative cover. —ROBERT E. PRESTON, Jones & Stokes Associates, 2600 V Street, Suite 100, Sacramento, CA 95616. CALIFORNIA BOEHMERIA CYLINDRICA (L.) Sw. (URTICACEAE).— Sacramento Co., widely scattered colonies in riparian zone along both sides of Georgiana Slough separating Andrus and Tyler Islands, from approx. 2 mi SW of Georgiana slough divergence from Sacramento River to approx. 11 mi along the slough to near The Oxbow; less common in Snodgrass Slough immediately upstream off the Sacra- mento River. Elev. ~ sea level. G. F. Hrusa 14879, J. A. Hart, 11 Oct. 1998, 2 mi SW divergence of Georgiana ) if _ 2000] Slough from Sacramento River on W side. Rhizomatous colony in opening at waters edge at or slightly below high tide level, also in shade beneath adjacent Alnus rhombi- | folia Nutt. 38°07'50.4"N; 121°34'55.6"W (CDA and to be distributed.); G. F. Hrusa 15277, J. A. Hart, M. J. Hooper, 24 Nov. 1999. Snodgrass Slough at Delta Meadows State Park, on exposed and partially submerged logs. Elev. ~ sea level. (CDA). Previous knowledge. Native throughout the region east | of the Rocky Mountains occurring in bogs, marshes and other wet places. Collected in Arizona near the turn of the ' century but only recently redocumented there (J. Bouf- _ ford, Ariz-Nevada Acad. Sci. 26:42—43, 1992). Significance. First records for California. In addition to the collected sites listed above, the species has been ob- served near the town of Rio Vista on the Sacramento Riv- er. It is assumed here that Boehmeria cylindrica is intro- duced to California, based primarily on its general occu- pation of disturbed and rip-rapped riverbanks. Moreover, the heavy boat traffic throughout the Delta region would appear to provide ample opportunity for introduction and spread of this species; however, it has also been found growing in less disturbed conditions and it may be a pre- viously overlooked native opportunistically occupying disturbed situations. —G. FREDERIC HrusA, Herbarium CDA, California Dept. of Food & Agriculture, Plant Pest Diagnostics Cen- ter, 3294 Meadowview Rd., Sacramento, CA 95832; JEF- FREY A. Hart, H. A. R. T. Inc. (Habitat Assessment & Restoration Team) 13737 Grand Island Rd., Walnut Grove, CA 95690. CALIFORNIA ERYNGIUM CONSTANCE! Y. Sheikh (APIACEAE).—Son- oma Co., dense populations in two seasonal pools, one draining into the other, on summit of Diamond Mtn. 4 km SSW of the town of Calistoga. Elev. 685 meters, 38°32'30"N; 122°35'00"W. G. F. Hrusa 13582a (lower pool center), /3582b (lower pool periphery), and /3582c (upper pool), A. Buckmann. Oct. 05, 1996. Verified by L. Constance, November 1997. (CDA, UC/JEPS and to be distributed). Previous knowledge. Previously known only from two sites, one the type locality, at and near Loch Lomond in Lake Co., approx. 35 km NNE. Described in 1983 (L. Sheikh, Madrono 30:93—101, 1983). Significance. First record for Sonoma Co. Two sites are known near Loch Lomond in Lake County; the type lo- cality immediately N of the Loch Lomond townsite and a second approximately 3 km to the east. The habitat on Diamond Mtn. appears similar to that at the type locality, but is dominated by Quercus garryana Hook., Q. lobata Nee and Pseudotsuga menziesii (Mirbel) Franco rather than the Pinus ponderosa Laws., Quercus kelloggii Newb. mix at Loch Lomond. The Diamond Mtn. pools have been variously disturbed, and apparently a permanent spring which in past times fed both pools was closed some de- cades before the current owners took up residence. How this affected the local hydrology and flora is unknown. The plants at Diamond Mtn. do not match exactly the form at the type locality, the divergence most noticeable in the larger number of flowers per capitulum and the variable habit, ranging from slender and upright on the pool margins, to prostrate and stout in the deepest center NOTEWORTHY COLLECTIONS 52 of the lower pool. These and other similar populations in the Sonoma-Lake County region are currently under study to assess their relationship to both E. constancei and E. aristulatum Jepson. The type locality is currently listed in Title 14 of the Fish & Game Code as the Loch Lomond Vernal Pool Eco- logical Reserve. At present such safety cannot be claimed for the Diamond Mtn. locality as the pools are in an area under active viticulture development and forest harvest. However, the current landowners are aware of the pools’ botanical importance in addition to DFG and California Department of Forestry regulations concerning their pro- tection. DFG plans to pursue some form of permanent protection. —G. FREDERIC HRuSA, Herbarium CDA, California Dept. of Food & Agriculture, Plant Pest Diagnostics Cen- ter, 3294 Meadowview Rd., Sacramento, CA 95832. AL- LAN BUCKMANN, California Dept. of Fish & Game, Region 3. P.O. Box 47, Yountville, CA 94599. CALIFORNIA ONONIS ALOPECUROIDES L. (FABACEAE).—San Luis Obispo Co., Temettate Rd. approx. | km by road NW of intersection with Suey Creek Rd., in NW corner of S6, TIIN; R33W, SB meridian. 35°03 58:2 N: 120°23'59.9"W. Elev. 380 m. Occupying approximately % acre in open grazed woodland-savanna among Quercus agrifolia Nee, Pinus sabiniana Douglas, extending east- ward down a dry arroyo to edge of riparian zone. Perez & Parks s.n., 9 July 1998 (CDA), Hrusa 14732 a—g, 21 July 1998 (CDA and to be distributed). Hrusa 14732c, 14732d, and 14732h confirmed by R. B. Ivimey Cook (EXR) & S. Jury (RNG). Previous knowledge. Native to southwest Europe, North Africa. Adventive in central Europe. Significance. First record for North America. The prop- agule source(s) is unknown. First noticed by the landown- er 2 years previous, the population apparently expanded rapidly into a dense but currently still more or less local- ized colony. In 1998 it was found spreading downslope along a dry drainage leading to Suey Creek and so may be expected more widely, at least locally, in the near fu- ture. The plants are unpalatable to the horses and burros that graze through the local area and because the species may form a dense stand capable of excluding more pal- atable vegetation it is currently the target of active erad- ication efforts by the San Luis Obispo County Agricultural Commissioners office. As of July 2000, an intensive out- reach program to local residents by the Ag. Commission- ers office has not revealed additional populations; how- ever, the current one remains active. —G,. FREDERIC HrusA, Herbarium CDA, California Dept. of Food & Agriculture, Plant Pest Diagnostics Cen- ter, 3294 Meadowview Rd., Sacramento, CA 95832. CALIFORNIA AMBROSIA PUMILA (Nutt.) A. Gray (ASTERACEAE).— Riverside Co., Riverside, moist area along Arlington Av- enue, La Sierra Heights, 26 Aug 1940, Ruth Cooper s.n. (Riverside Community College Herbarium); Nichols Road wetlands area, northwest of Nichols Road and west of 140 Alberhill Creek, southwest of the I-15 Freeway, Lake El- sinore 7.5’ Quad., T5S RSW NW/4 S825, alt. 384 m, ca. 1300 plants in Arbuckle loam soil, annual grassland with Vulpia myuros (L.) C. Gmelin, [socoma menziesii (Hook. & Arn.) G. Nesom, Nassella pulchra, Bromus rubens, Er- odium botrys (Cav.) Bertol., Hirschfeldia incana (L.) Lagr.-Fossat, Hemizonia paniculata A. Gray, and Avena fatua L., 18 Jul 1997, D. Bramlet 2575 (UCR); Warm Springs Valley, Nichols road, 0.3 km by road W junction I-15 near head of Walker Canyon, ca. 2.8 km NW Lake Elsinore, alt. 383 m, 29 Jun 1997, Fred M. Roberts Jr. 5043 (RSA); Warm Springs Valley, NW of Lake Elsinore, about Alberhill Creek at the head of Walker Canyon near Durant Siding, SE base of Alberhill Mountain, along Nichols Road, 0.2 miles SW of junction with Collier Av- enue, near 33°42'N, 117°21'W, TSN RSW NW 4% S825, alt. 381 m, 22 Jul 1997, Steve Boyd 10017 (RSA). Previous knowledge. Reported only from San Diego County by The Jepson Manual (J. C. Hickman, ed., 1993, U.C. Press) and from “‘sw San Diego Co.”’ by Munz (A California Flora, 1959, U.C. Press) but a locality at Skunk Hollow, Riverside County is also known (Madrono, 1992, 39(2):157). This species is considered rare and endangered and is reported to be declining by the CNPS Inventory of Rare and Endangered Vascular Plants of California (Skin- ner and Pavlik, 1994, CNPS Inventory of Rare and En- dangered Vascular Plants of California, 5th Ed.). Significance. These specimens provide second and third localities for Riverside County and range extensions of 27 km WNW and 53 km NW from the previously reported site at Skunk Hollow. The La Sierra Heights plants, only now being reported 60 years after their collection, were from the wet alkaline areas that formerly existed from near the intersection of Arlington Ave. and Van Buren Blvd., and the Riverside Airport, to the vicinity of California Ave. at Jackson Street. Unfortunately, there is little chance the plants persist there as the area is now largely paved and urbanized, though a few small pockets of marginal habitat do remain. The record at RCC was represented by three replicate specimens! It is unfortunate that two of these were not distributed to larger herbaria where they would have more rapidly come to the attention of the bo- tanical community. Perhaps something could have been done to salvage part of this population if the location had been generally known. The Nichols Road population is under considerable threat due to its location in a rapidly urbanizing area. These discoveries highlight both the need to look for this species in other moist alkaline places in western Riverside County and the importance of checking small and inactive herbaria for significant records of other rare plants. DICORIA CANESCENS A. Gray (ASTERACEAE).—San Bernardino Co., Rialto [actually Colton, near the city boundary], degraded sand dunes on Slover Ave. just E of the tank farms and E of Riverside Ave., S of I-10 Freeway, very common, 14 Oct. 1993, Chet McGaugh s.n. (UCR); same location, 34°04’N, 117°22'W, alt. 320 m, ca. 200 plants seen in loose blowsand at summit of dune, in and near a disturbed OHV area, very local at this site and at another ca. 700 m further west, 14 Sept. 1999, A. C. Sand- ers, S. Boyd and M. Provance 23073 (RSA, UCR). Previous knowledge. A native plant of desert dunes, previously unrecorded on the coastal slope of California, except as a waif along the railroad near Elysian Park in Los Angeles (A. Davidson s.n. in 1892 and 1893, POM; H. M. Hall, Compositae of Southern California, Univ. of California Publ. Botany, Vol. 3, 1907). MADRONO Significance. First records of the species in natural hab- itat west of the deserts and a range extension of 55 km WNW from the Cabazon area. The species is well estab- lished at this site, which is a remnant of the formerly | extensive Colton Dunes. We cannot be certain that the species was not introduced, but it appears native and is | present in an arid interior valley in the best available nat- ural habitat on the coastal slope. It is surprising that this species has so long escaped detection if it is native at this site, but other typically desert species are also found near- by, including Camissonia campestris (E. Greene) Raven, Encelia farinosa Torrey & A. Gray, Eriophyllum wallacei [Vol. 47 | (A. Gray) A. Gray, Malacothrix glabrata A. Gray, and | Prosopis glandulosa Torrey. While this area has long had | an active botanical community, the total number of col- lectors has never been great, and it’s probable that the dunes have just not been thoroughly surveyed in the fall, | when few other plants are flowering. ERODIUM MALACOIDES (L.) Willd. (GERANIACEAE).— | San Bernardino Co., Blue Mtn., Grand _ Terrace, 34°01'14"N, 117°18'14”"W, alt. 396 m, abundant on the | pediment near a water tank on the W side of the mountain, associated with Heterotheca grandiflora Nutt., 4 Apr 1998, M. C. Provance 350 (UCR). Previous knowledge. Weed introduced from Europe. Previously reported in California only from the northern San Joaquin Valley and San Francisco Bay areas. Significance. First record for San Bernardino County | and southern California and a range extension of over 600 km from the San Francisco Bay region. This weed should be sought in other areas in southern California and its status and spread carefully monitored. KOELERIA PHLEOIDES (Villars) Pers. (POACEAE).—Riv- erside Co., Jurupa Mtns., Sunnyslope near Rattlesnake Mtn., Armstrong St., 0.1 mi. south of San Bernardino Co. line, 34°01'59"N, 117°24'53”"W, T2S RSW NW/4 S4, alt. 1050 ft, in hard dry soil in a disturbed field, 7 Mar 1998, M. C. Provance 174 (ARIZ, UCR). Det. by J. & C. Reed- er, 1999; Previous knowledge. Uncommon introduction, reported from scattered locations from Santa Barbara and Kern Counties through northern California. Significance. First record for Riverside County and a range extension of 250 km southeast from Santa Barbara County. MONARDELLA PRINGLEI A. Gray (LAMIACEAE).—San Bernardino Co., sand hills west of Colton, 17 May 1941, J. C. Roos 2472 (La Sierra College Herbarium). Previous knowledge. This species, a very local endemic of the Colton Dunes, has generally been thought extinct since 1921 (Skinner and Pavlik, 1994). Significance. This collection extends the known chro- nological range of this species by 20 years, but unfortu- nately we still have no evidence that the species has per- sisted until today. Like the Ambrosia record above, this record emphasizes the need to examine ail herbaria for informative collections. This population was reported on the label to have been ‘mutilated by grasshoppers’’, though the specimen preserved was not too badly dam- aged. The recent discovery of other noteworthy species on the remnants of the Colton Dunes offers hope that this species may yet be rediscovered. NAMA STENOCARPUM Gray. (HYDROPHYLLACEAE).— Orange Co., San Joaquin Hills, Emerald Canyon, 3 km up canyon from Pacific Coast Highway, 2.75 km SW of in- tersection of Laguna Canyon Rd. and El Toro Rd., Laguna 2000] Beach, 33°34'25"N, 117°47'10"W, a few plants on sandbar ‘in steep incised channel with Typha sp., Chenopodium ‘ambrosioides L., Juncus xiphoides E. Meyer and Bac- ‘charis salicifolia (Ruiz Lopez & Pavon) Pers., alt. 122 m, 21 July 1998, A. L. Wolf 402 (UCR); Laguna Lakes ‘(northernmost lake), W of Laguna Canyon Rd., Laguna Beach, 33°36'50"N, 117°45'30’W, alt. 118 m, 30+ plants 'on south edge of lake on drying margin with Petunia par- viflora A. L. Juss., Rorippa curvisiliqua (Hook.) Britten and Lythrum californicum Torrey & A. Gray, 26 June (1998, A. L. Wolf 358 (UCR); Lambert Reservoir, N of El ‘Toro Marine Corps Air Station, 33°41'32"N, 117°42'40"W; alt. 134 m, few plants on mudflat on S edge of reservoir with Ammania robusta Heer & Regel, Lythrum hyssopi- folium L., and Juncus bufonius L., 15 May 1998, A. L. Wolf 276 (UCR); Peters Canyon Channel, E side of chan- nel between Alton Pkwy and Barranca Pkwy, Tustin, 33°41'30"N, 117°49'17"; alt. 15 m, 2 plants in sediment basin adjacent to channel in the southwest portion of re- cently bladed field, 1 July 1998, A. L. Wolf 414 (UCR); Riverside Co., Mystic (San Jacinto) Lake, 1.4 miles SE of Jackrabbit Trail on Gilman Springs Rd., 1.4 mi S of Glen Eden Hot Springs, 33°52'31'N, 117°03'W, alt. 433 m, common on receding lakeshore with numerous herbs, 26 Sep 1999, A. C. Sanders & M. Provance 23120 (UCR, and to be distributed); same location, 15 Oct 1999, A. C. Sanders, D. Bramlet, M. Costea & T. Salvato 23173 (UCR, and to be distributed). Previous knowledge. A rare plant of seasonally moist areas from scattered counties, mainly coastal in southern California and extending south into Mexico. This species is considered rare in California but more or less common elsewhere (Skinner and Pavlik 1994). Not recorded on the mainland of southern California since 1939, nor anywhere in California since then, except for two collections on San Clemente Island (California Natural Diversity Database), the last in 1991. Significance. First records for Riverside County and first records for California, except for San Clemente Is- land, in over 60 years. This rare plant is still a member of the mainland California flora. QUERCUS PALMERI Engelm. (FAGACEAE).—Riverside Co., Jurupa Mountains, Rattlesnake Mt. near Crestmore Heights, ca. 4.2 km NW of downtown Riverside, 34°01.59'N, 117°23.81'W, T2S R5W S83, alt. 400 m, ca. 50 individual shrubs forming a dense and nearly homog- enous vegetation over an area of approximately 25 m X 8 m, growing in a nook on a rocky north-facing slope, associated with Prunus illicifolia (Nutt.) Walp., Ribes in- decorum Eastw. and Phacelia ramosissima Lehm., 14 Apr 1998, Mitchell C. Provance 441 (UCR); Jurupa Moun- tains, Rattlesnake Mt. (hill 1452) above Crestmore, NW of Riverside, Fontana 7.5’ Quad., 34°02’N, 117°23.5'W, T2S RSW center of W/2 S3, alt. 365-400 m, dense colony in a notch on the ridge and in adjacent rocky gully be- tween outcrops on the N-facing granitic slope, coastal Sage scrub with chaparral elements, Prunus illicifolia, Ri- bes indecorum, Eriogonum fasciculatum (Benth.) Torrey & A. Gray, Rhamnus crocea Nutt., Mimulus aurantiacus Curtis, Rhus trilobata Torrey & A. Gray, Salvia mellifera E. Greene, Toxicodendron diversilobum (Torrey & A. Gray) E. Greene, etc., 14 May 1998, A. C. Sanders and Mitch Provance 21848 (CAS, DAV, RSA, SD, UC, UCR). Previous knowledge. Occurrences of this species are patchy from Colusa Co., California south to Baja Califor- nia, Mexico and to the east in Arizona. In southern Cali- fornia, the species is most common in the Peninsular NOTEWORTHY COLLECTIONS 141 Range, San Jacinto Mountains and south, but also occurs on the desert slopes of the San Gabriel and San Bernar- dino Mountains. There is a record of a single plant in the Little San Bernardino Mountains. This species has been considered for inclusion in the CNPS Inventory of Rare and Endangered Vascular Plants of California. Significance. First known record of this species for the South Coast subregion (Hickman 1993) and a range ex- tension of approximately 38 km SSW from the nearest known populations on the north side of the San Bernar- dino Mtns., near Mojave River Forks. This population is well separated from the other known populations, and oc- curs far below the lowest elevation previously known for the species in the region. The rate of sexual reproduction in this population appears to be extremely low, although some of the plants do appear to be making a few healthy acorns. It is amazing that this conspicuous species has escaped detection on the outskirts of a large city that has had an active botanical community for over 100 years. For example, some early botanists and plant collectors who were Riverside residents, prior to the establishment of UC Riverside, include: Charlotte M. Wilder (whose house was in the Jurupa Mtns.), Fred M. Reed, Harvey Monroe Hall, David D. Keck, and Edmund C. Jaeger. In addition, Samuel B. Parish’s residence in San Bernardino was less than 15 km NE of the Quercus palmeri site. This discovery points up the need to continue searching for new species and important range extensions even in ‘‘well known” areas: an unexplored ridge can hide something new or interesting, and there are many unexplored ridges. —MITCHELL C. PROVANCE and ANDREW C. SANDERS, Herbarium, Dept. of Botany & Plant Sciences, University of California, Riverside, CA 92521-0124, VALERIE SOZA and STEVE BoypbD, Herbarium, Rancho Santa Ana Botanic Garden, 1500 N. College Avenue, Claremont, CA 91711, DAVID BRAMLET, 1691 Mesa Dr. A-2, Santa Ana, CA 92707 and ADRIAN L. WoLF, Harmsworth Associates, 36 Bluebird Lane, Aliso Viejo, CA 92656. CALIFORNIA All the following collections are from the San Bernar- dino Mountains. BRICKELLIA KNAPPIANA Drew (ASTERACEAE ).—San Bernardino Co., northeast side of Blackhawk Mountain, east of Blackhawk Canyon, 34°21'06"N_ 116°47'24"W, elev. 1372 m, shrub 1.8 m tall in alluvium of minor can- yon, northeast-facing wash, 10 Aug 1998, Valerie Soza et al. 409 (RSA, and to be distributed). Previous knowledge. This probable hybrid (treated as a species in all available manuals) between B. desertorum Cov. and B. multiflora Kellogg is known from only a few locations in the northern and eastern mountain ranges of the Mojave Desert, almost always at sites where both par- ents are present. Previously collected from the Argus, Fu- neral, Panamint, and Kingston Mountains, but with the type locality at an undefined site along the Mojave River. The type locality is given simply as ‘‘in the neighborhood of the Mohave River” (Pittonia, 1888, 1:260). The vague- ness of the collection data inspires little confidence that the site was particularly near the Mojave River, especially since the plant has not, in the 110 years since, been re- corded there. Significance. First record for the San Bernardino Moun- 142 tains, which extends the range of this taxon 90 km south- ward from the closest known locality in the Argus Range. BRICKELLIA MULTIFLORA Kellogg (ASTERACEAE).— San Bernardino Co., canyon running from west to east along the northern edge of Horsethief Flat, 34°19.487'N 116°45.841'W, elev. 1372 m, rocky drainage area near mouth of canyon, seasonally moist, subjected to carbonate scree Slides from along canyon walls, 29 Jun 1998, Mitch- ell C. Provance & Valerie Soza 792 (UCR). Previous knowledge. Uncommon shrub occurring from the northern and eastern Mojave Desert of California to Nevada, particularly in the mountains of Inyo County, the White Mountains of Mono County, at Little Lake in Kern County, and the Kingston, Clark and Granite Mountains of San Bernardino County. Significance. First record for the San Bernardino Moun- tains, which extends the range of this species about 100 km southwestward from the nearest known occurrence in the Granite Mtns. near Kelso. It is noteworthy that this species turned up at about the same time that its hybrid progeny, B. knappiana Drew, was also discovered in the range. It is interesting that several species traditionally known from the northern and eastern Mojave Desert (e.g., Baileya multiradiata A. Gray, Madrofio 43 (4):524, and the plants reported here) have recently been found on the northern slopes of the San Bernardino Mountains, as that area has begun to be explored away from the major routes of travel. CAMISSONIA PTEROSPERMA (S. Watson) Raven (ONA- GRACEAE).—San Bernardino Co., west of Horsethief Flat, above road 3N03A, 34°19'06"N 116°47'03’W, elev. 1768 m, scarce annual in open WNW-facing slope, 16 Jul 1998, Valerie Soza & Tasha LaDoux 390 (RSA); San Ber- nardino Mountains, northwest of Tip Top Mountain, east of Arrastre Creek, 34°15'59"N 116°43'45”W, elev. 1920 m, rare annual on open rocky north-facing lower slope, 30 Jul 1998, Valerie Soza & Tasha LaDoux 404 (RSA). Previous knowledge. Rare annual in northern mountain ranges of the Mojave Desert, e.g., Panamint and Clark mountains, and Inyo and White mountains east to Last Chance Range and Fish Lake Valley, to Utah and Oregon. Significance. First record for the San Bernardino Moun- tains, range extension of about 100 km southwestward from Clark Mountain. CORNUS GLABRATA Benth. (CORNACEAE).—Riverside Co., Morongo Indian Reservation, very locally common at a seep along a gully at the west end of Burro Flat, 33°59'N, 116°52’W, alt. 1160 m, 14 Nov 1997, A. C. Sanders & T. Tennant 21596 (DAV, RSA, SD, TEX, UC, UCR, UTC). Previous knowledge. Scarce in southern California and known from Riverside County only from a single collec- tion from the San Jacinto Mountains made in 1922 (P. A. Munz 5806, alt. 1500 m, Hemet Valley, frequent along banks of Pipe Creek) based on specimens at UCR and RSA. Significance. First record for the San Bernardino Moun- tains, second record from Riverside County and the first collection of this species from that county in 77 years. CYNANCHUM UTAHENSE (Engelm.) Woodson (ASCLE- PIADACEAE).—San Bernardino Co., E. of Horsethief Flat, 0.5 km N of the Arrastre Creek Dam, 34°19.44'N 116°45.77'W, elev. 1433 m, on a steep, barren, sandy, SE- facing slope, 23 Jun 1998, Mitchell C. Provance & Val- erie Soza 744 (UCR). Previous knowledge. Uncommon perennial occurring MADRONO [Vol. 47 | on the Mojave Desert of California and to Utah and Ari- | zona. | Significance. First record for the San Bernardino Moun- | tains and the Transverse Ranges and extends the range of | this species slightly (10 km southwest) from the nearest | known occurrence of Old Woman Spring on the southern Mojave Desert. GLYCERIA OCCIDENTALIS (Piper) J. C. Nelson. (PO- ACEAE).—Riverside Co., Morongo Indian Reservation, t very locally common in mud around the sag pond at the | southeast end of Burro Flat, 33°59.5'N, 116°51'’W, T2S RIE SE/4 S14, alt. 1150 m, 24 Apr 1996, A. C. Sanders | & S. Hawkins 18088 (RSA, UCR, and to be distributed). | Det. by Travis Columbus. Previous knowledge. Northwestern California and north to Idaho and British Columbia, the furthest south previ- ously known populations are apparently in San Mateo > County (P. A. Munz, 1968, Supplement to A California > Flora, University of California Press). | Significance. First record for southern California and a> range extension of 650 km from the San Francisco Bay area. The site where this species was collected is a shallow but permanently wet sag pond on the San Andreas Fault from which, reportedly, peat was formerly harvested. This bizarre disjunction in the distribution of a native plant | suggests that the species should be sought in other wet | areas in central and southern California. Travis Columbus | notes that this species is possibly not distinct from the | Eurasian species Glyceria declinata Brébiss. That species | has also been reported from northern California (e.g., A. | S. Hitchcock and A. Chase, 1935, Manual of the Grasses | of the United States; P A. Munz, 1959, A California Flora, | University of California Press), but is not mentioned in > the Jepson Manual. Glyceria declinata was reported in the | literature for California in 1957 by Beecher Crampton from Stanislaus County, south of Oakdale, further ESE from the San Francisco Bay area (Leaflets of Western Bot- any 8 (6):160). NICOTIANA ACUMINATA Hook var. MULTIFLORA (Philippi) Reiche (SOLANACEAE).—Riverside Co., Morongo In- | dian Reservation, lower Hathaway Creek Cyn., 33°58'N, | 116°52'W, elev. 838 m, disturbed roadside especially in moist areas, margins of riparian forest, 20 Aug 1998, A. | C. Sanders 22232 (UCR). Previous knowledge. An introduced weed from South | America, long known from northern California, but only I first reported in southern California in 1996 from San Ber- nardino County (Madrofio 43(2):334-—336). Significance. First record for Riverside County; extends | its southern California range 18 km southeast from Mill Creek Cyn., and further documents the establishment and | spread of this introduced weed in southern California. —VALERIE SOZA, Herbarium, Rancho Santa Ana Botan- ic Garden, 1500 N. College Avenue, Claremont, CA | 91711; MITCHELL C. PROVANCE and ANDREW C. SANDERS, Herbarium, Dept. of Botany & Plant Sciences, University | of California, Riverside, CA 92521-0124; and STEVE Boyp, Herbarium, Rancho Santa Ana Botanic Garden, | 1500 N. College Avenue, Claremont, CA 91711. COLORADO ALICIELLA SEDIFOLIA (Brandegee) J. M. Porter [Gilia sed- ifolia Brandegee| (RPOLEMONIACEAE).—Hinsdale Co., y i | | ] 2000] | San Juan Mts., Half Peak; 19 km SW of Lake City, T42N -R6W SEC25 SE%; 4110 m. South facing slope, gravelly ‘patches with no other vegetation, common locally. 5 Au- gust 1995, Susan Komarek 478 (COLO 457147). Previous knowledge. Evidently a rare endemic; thought extinct since original collection by Purpus in 1893 [Gun- nison County, Uncompahgre Range, Sheep Mt., 11,800’, July 1893. Purpus 697 (GH)]. Significance. First collection in 103 years. _ ASCLEPIAS INVOLUCRATA Engelmann ex Torrey (ASCLE- PIADACEAE).—Bent Co., 18 km NE of Las Animas, -T21S R51W SEC23 NW%; 1260 m. Transitional zone be- tween shortgrass and sandsage prairie, with Buchloé dac- tyloides, Psoralidium tenuiflorum, and Oligosporus filifol- ius. 7 June 1998, Dina Clark 686 (COLO 471324). Previous knowledge. Southern Great Plains, New Mex- ico, Arizona, and Mexico; one collection from Las Ani- mas County, Colorado in 1948 [Rogers 5834, May 31, 1948. (COLO 55248)] (Great Plains Flora Association, Flora of the Great Plains 1986). Significance. First Colorado collection in 50 years. The documented distribution of this species may become better known with increased field work in the eastern part of the state. ASTROLEPIS INTEGERRIMA (Hooker) Benham & Windham (PTERIDACEAE).—Las Animas Co., Mesa de Maya, Je- sus Mesa, near Colorado - New Mexico State Line, ca. 85 km ESE of Trinidad; T35S R54W SEC16 NE%; 1540 m. Dry, southwest facing slope in crevice of Dakota sand- stone outcrop. 19 September 1994, Dina Clark 582 & Car- olyn Crawford (COLO 455480). Previous knowledge. Arizona, Nevada, New Mexico, Oklahoma, Texas, and Mexico (Great Plains Flora Asso- ciation, Flora of the Great Plains 1986). Significance. First Colorado record. Apparently the northernmost record of this New World genus (Flora of North America Association, Flora of North America Vol. 221993). BOTHRIOCHLOA SPRINGFIELDIT (Gould) Parodi [Andropo- gon springfieldii Gould] (POACEAE).—Las Animas Co., Mesa de Maya, vicinity of upper Gotera Canyon, ca. 70 km ESE of Trinidad; T34S R55W SEC27 SE'%; 1700 m. South facing slope in grassy breaks. 5 July 1993, Dina Clark 204 & C. Deihl (COLO 455362); 4 August 1993, Dina Clark 216 & P. Deihl (COLO 455213). Previous knowledge. West Texas to Arizona (Great Plains Flora Association, Flora of the Great Plains 1986). Significance. First Colorado record. A range extension of ca. 100 km and the northernmost record for this species (Great Plains Flora Association, Atlas of the Flora of the Great Plains 1977). CHENOPODIUM CYCLOIDES Nelson (CHENOPODIACE- AE).—Weld Co., ca. 25 km ENE of Greeley; T6N R63W SEC34 SE%; 1430 m. Andropogon hallii-Calamovilfa lon- gifolia grassland on eolian deposited sand. 3 September 1997, Dina Clark 634 (COLO 469818). Previous knowledge. Southwest Kansas south through west Texas, west to southern New Mexico; known in Col- orado from Las Animas and Pueblo counties (Spackman et al., Colorado Rare Plant Field Guide 1997). Significance. Northward range extension of ca. 250 km for this rare, or perhaps overlooked, species. DIPLACHNE DUBIA (Kunth) Scribner [Leptochloa dubia Humboldt, Bonpland, & Kunth] (POACEAE).—Las An- imas Co., Mesa de Maya, ca. 115 km E of Trinidad; T33S NOTEWORTHY COLLECTIONS 143 RS51W SEC25 SE%; 1450 m. Bottom of dry, rocky, S facing slope. 4 September 1993, Dina Clark 283, T. Ho- gan, & R. Brune (COLO 455445). Previous knowledge. Texas, western Oklahoma, Arizo- na, and Mexico; also south Florida and Argentina (Correll and Johnston, Manual of the Vascular Plants of Texas 1970). Significance. First Colorado record. A range extension of ca. 500 km from previously known location (Neil Snow personal communication). ELEOCHARIS XYRIDIFORMIS (Fernald) Brackett (CYPER- ACEAE).—Logan Co., Vicinity of Sterling, Peetz Table, 40°56'36"N, 103°10'19"W; 1300 m. Low swale, on clay- sand substrate. 31 August 1997, W. A. Weber 19273 & Ron Wittmann (COLO 467546). Cheyenne Co., Three km E of Kit Carson on Hwy 40; 1250 m. Low swale in sand hills of high plains. 7 September 1997, W. A. Weber 19382, Ron Wittmann, & Dina Clark (COLO 467358). Previous knowledge. North Dakota and Montana, south to Texas and Mexico (Great Plains Flora Association, Flo- ra of the Great Plains 1986). Significance. First Colorado record. Probably over- looked previously, and to be expected more frequently with increasing field work in eastern Colorado. FESTUCA SUBULATA Trin. (POACEAE).—Rio Blanco Co., North Fork of White River near North Fork Camp- ground, White River N.E, ca. 20 km NE of Buford; TIN R9OOW SEC13; 2400 m. Riverbank with Alnus incand (L.) Moench and Salix drummondiana Hook. 9 August 1992, Nan Lederer 3661 (COLO 451683); 3 September 1992, Gwen Kittel (COLO 451607). Previous knowledge. Alaska to Alberta, south to Cali- fornia, east to Idaho, Montana, Utah, and Wyoming. Re- ported from northern tiers of Utah (Welsh, A Utah Flora 1993) and Wyoming (Dorn, Vascular Plants of Wyoming 19972), Significance. First Colorado record. Approaching the southern extent of its range in North America (known from the central Sierra Nevada of California). HELENIUM MICROCEPHALUM De Candolle (ASTERA- CEAE).—Las Animas Co., Mesa de Maya, ca. 105 km E of Trinidad; T33S R51W SEC22 NW%; 1500 m. Muddy edge of stock pond in shortgrass prairie. 16 September 1994, Dina Clark 554 (COLO 455583). Previous knowledge. Southwest Oklahoma, Texas, southern New Mexico, to northern Mexico (Great Plains Flora Association, Flora of the Great Plains 1986). Significance. First Colorado record. A species of the southern Great Plains, this record represents a range ex- tension of nearly 300 km. HETEROSPERMA PINNATUM Cavanilles (ASTERA- CEAE).—Las Animas Co., Mesa de Maya, ca. 75 km E of Trinidad; T34S R56W SEC20 NE%; 1700 m. Among sandstone slabs near drainage; shortgrass prairie. 17 Sep- tember 1994, Dina Clark 566 (COLO 455475). Previous knowledge. Mexican highlands, north to Ari- zona and Texas (Correll and Johnston, Manual of the Vas- cular Plants of Texas 1970). Significance. First Colorado record. A Chihuahuan spe- cies at its northernmost point in North America; an exten- sion of its previously known range from Santa Fe County, New Mexico and the Davis Mountains of Texas. REVERCHONIA ARENARIA Gray (EUPHORBIACEAE).— Bent Co., S shore of John Martin Reservoir; T23S R49W SEC17 SW¥% of NW%; 1150 m. Sandsage prairie on dune. 144 6 June 1998, Dina Clark 683 (COLO 471323); 30 July 1998, Dina Clark 740 (COLO 471325). Previous knowledge. Oklahoma, Kansas, Texas, New Mexico, Utah, Arizona, and northern Mexico (Great Plains Flora Association, Flora of the Great Plains 1986). Significance. First Colorado record. Northernmost rec- ord for this species; a range extension of ca. 200 km. TRITELEIA GRANDIFLORA Lindl. [Brodiaea douglasii Wats.] (ALLIACEAE).—Montezuma Co., Boggy Draw near Peel Reservoir; T38N RI4W SEC7 SW; 2400 m. Pine-oak vegetation; population of about 2000 plants in 8 ha area. 22 June 1998, Leslie Stewart 4 (COLO 470260, 470261). Previous knowledge. A showy species of the northwest; nearest reports from the Wasatch Range of northeast Utah (Welsh, A Utah Flora 1993) and northwestern Wyoming (Dorn, Vascular Plants of Wyoming 1992). Significance. First Colorado record. There is some spec- ulation this may have been transported by indigenous peo- ple as a food source. Specimen was collected by Charlotte Thompson, a seasonal employee of the U.S. Forest Ser- vice, in the process of clearing a timber sale. Population was reportedly protected from lumbering operation (Leslie Stewart personal communication). —DINA CLARK and TIM HOGAN, University of Colorado Herbarium (COLO), Campus Box 350, Boulder, CO 80309. OREGON CAREX CHORDORRHIZA Ehrh. ex L. f. (CYPERA- CEAE).—Clatsop Co., weedy cranberry fields, with C. arcta Boott, W of Cullaby Lake, elev. 4 m, T7N RIOW $22, 2 Oct 1999, P. F. Zika 14455 WTU; Coos Co., well- established weed in cranberry fields, with Vaccinium ma- crocarpon Aiton, Aster subspicatus, Lysimachia terrestris, between Spruce Hollow and Coquille River, elev. <50 m, T28S R14W S20 E'%, 20 June 1997, P. F. Zika 13217 & B. Wilson (OSC, MICH); Curry Co., weed in cranberry fields, with Agrostis exarata, 2.5 km W of Route 101, SW of Floras Lake, elev. ca. 15 m, T31S RISW S20, 21 Au- gust 1997, P. F. Zika 13365 & B. Wilson (OSC). Previous knowledge. Creeping sedge is a circumboreal sedge native in eastern and northern North America. The nearest native population is 380 km northeast in Okanog- an Co., Washington (Wooten 1334 WS). Significance. First report for Oregon. All sites are as- sociated with wetland cranberry agriculture, which began with the introduction of Vaccinium macrocarpon to Ore- gon in 1885. The cranberry vines are cut and hauled be- tween farms to establish new cranberry fields. The process also transports weed seeds from one farm to the next. Carex chordorrhiza is known as a weed in cranberry fields in Wisconsin (Eck, The American Cranberry, 1990), where it is a native and invades from adjacent wetlands (A. A. Reznicek, personal communication). Creeping sedge is not reported as a weed from cranberry farms else- where in North America (Eck 1990). Thus we suspect C. chordorrhiza arrived in Oregon during the transport of cranberry vines from Wisconsin to Oregon. There are sev- eral other eastern native wetland plants that are reported below as new weeds in our area. All probably originated as propagules attached to cranberry vines transported from the eastern United States. MADRONO [Vol. 47 ESCALLONIA RUBRA (Ruiz & Pav6n) Pers. (GROSSU- LARIACEAE).—Coos Co., naturalized weed on muddy | shore and on disturbed ground in adjacent fields, with | Gaultheria shallon Pursh, Myrica californica Cham., SE end of Lost Lake, NW of McTimmonds Road, elev. 7 m, | T29S RISW S36 NW%, 20 June 1997, P. F. Zika 13207) & B. Wilson (OSC). Previous knowledge. Native to Chile. Commonly cul- tivated west of the Cascade Mts. in Oregon and Washing- ton. | Significance. First report for Oregon as a garden escape. ‘ Naturalized in a wild setting in the New River Area of | Critical Environmental Concern, Coos Bay District of the. Bureau of Land Management. FUCHSIA MAGELLANICA Lam. (ONAGRACEAE).—Coos | Co., steep weedy roadside bank, with Rubus armeniacus, Lathyrus latifolius L., Route 101, Coos Bay, near North’ Bend town line, elev. ca. 10 m, T25S R13W S23, 18 June | 1997, P. F. Zika 13139 & B. Wilson (OSC); Curry Co., weed in hedge rows, with Rubus spectabilis Pursh, R. ar- | meniacus, R. laciniatus Willd., Pseudotsuga menziesii (Mirbel) Franco, near waste treatment plant, Wharf Street, . Brookings, elev. ca. 20 m, T41S R13W S6, 19 May 1997, - P. F. Zika 13083 (WTU); roadside weed, Route 635 at 0.3 km NW of Yorke Creek, elev. 24 m, T37S R14W S18 | NW% of SE%; 19 June 1997, P. F. Zika 13163, B. Wilson, | & V. Stansell (OSC). Previous knowledge. Native to Chile and Argentina. Common ornamental in Oregon and Washington, occa- sionally fruiting west of the Cascade Mtns. | Significance. First report in Oregon as a garden escape. Most likely dispersed by fruit-eating birds. HYPERICUM BOREALE (Britton) E. Bickn. (CLUSI- ACEAE).—Clatsop Co., weedy cranberry fields, W of Cullaby Lake, elev. 4m, T7N RIOW S22, 2 Oct 1999, P. F. Zika 14449 OSC, WTU; Coos Co., common weed in- moist sand, in and near cultivated cranberry fields, with Vaccinium macrocarpon Aiton, Lysimachia_ terrestris, Juncus effusus, Lower Fourmile Road, 1.5 km W of Route 101, elev. ca. 10 m, T30S RISW SI NW% of NW%, 20 August 1997, P. F. Zika 13338, B. Rittenhouse, B. New- house, & B. Wilson (OSC); sunny depression between cranberry fields, and nearby ditches, with Sisyrinchium californicum (Ker Gawler) Dryander, Lotus corniculatus L., Rubus ursinus Cham. & Schldl., Panicum occidentale, Croft Lake Road, 1.5 km W of Route 101, elev. ca. 15 m; T30S RI5W S11; 20 August 1997, P. F. Zika 13340, 13346 (OSC); Curry Co., cranberry fields and ditch mar- gins, well established weed, with Juncus bufonius L., 2.5 km W of Route 101, SW of Floras Lake, elev. ca. 15 m, T31S R1ISW S20, 21 August 1997, P. F. Zika 13367 & B. Wilson (OSC). Previous knowledge. Native to eastern North America, west to Illinois. Significance. First record for Oregon. At present it ap- pears to be restricted to cultivated cranberry fields, and adjacent moist, disturbed ground. JUNCUS BREVICAUDATUS (Engelmann) Fernald (JUNCA- CEAE).—Clatsop Co., weedy cranberry fields, W of Cul- laby Lake, elev. 4 m, T7N RIOW S22, 2 Oct 1999, P. F. Zika 14454 OSC, WTU; Curry Co., sandy banks near | cranberry fields, with J. bufonius L., J. canadensis, J. planifolius, 2.5 km W of Route 101, SW of Floras Lake, elev. ca. 15 m, T31S R1I5W S20, 21 August 1997, P. F. Zika 13366 & B. Wilson (OSC). | 2000] _ Previous knowledge. Native to eastern North America, | west to Minnesota. | Significance. First Oregon record. Another weed asso- ' ciated with cranberry agriculture. | JUNCUS CANADENSIS J. Gay ex Laharpe (JUNCA- ' CEAE).—Clatsop Co., weedy cranberry fields, W of Cul- 'laby Lake, elev. 4 m, T7N R1OW S22, 2 Oct 1999, P. F. | Zika 14452 OSC, WTU; Coos Co., mucky shoreline, with Carex aquatilis Wahlenb. var. aquatilis, Dulichium, Lysi- _machia terrestris, Juncus supiniformis Engelm., Scirpus _subterminalis Torrey, Muddy Lake, elev. 5 m, T30S, R1I5W S11 SW%, 20 August 1997, P. F. Zika 13353 & |B. Rittenhouse (OSC); shallow water emergent, with Po- tentilla palustris (L.) Scop., Gentiana sceptrum Griseb., Hypericum anagalloides Cham. & Schldl., Vaccinium ma- crocarpon Aiton, E shore of unnamed pond, W of Lost Lake, elev. 7 m, T29S RISW S36 NW% of NW¥%, 21 August 1997, B. Newhouse 97055 (OSC); Curry Co., well established weed in ditches and cranberry fields, with J. effusus L., J. falcatus E. Meyer, Scirpus setaceus L., 2.5 km W of Route 101, SW of Floras Lake, elev. ca. 15 m, T31S RISW S20, 21 August 1997, P. F. Zika 13360 & B. Wilson (OSC). Previous knowledge. Native to eastern North America, west to Minnesota. Reported as a weed in cranberry fields in New England (Sears et al., An Illustrated Guide to the Weeds of Cranberry Bogs in Southeastern New England, 1996). Significance. First report for Oregon. Another weed as- sociated with the cranberry industry. Our collections from NOTEWORTHY COLLECTIONS 145 undisturbed wetlands indicate this species is successfully invading native plant communities in southwestern Ore- gon. JUNCUS PELOCARPUS E. Meyer (JUNCACEAE).—Coos Co., sandy soil in swamps and cranberry bogs, Bandon; 2 Sept 1958; G. Scott s.n. (OSC); sandy shoreline of ir- rigation pond, with Lysimachia terrestris, Salix sitchensis Bong., Viola lanceolata, near cranberry fields, Croft Lake Road, 1.5 km W of Route 101, elev. ca. 15 m; T30S R1I5W S11; 20 August 1997, P. F. Zika 13339 (OSC). Previous knowledge. Native to eastern North America, west to Minnesota. Noted as a weed in cranberry fields in New England (Sears et al. 1996). Significance. First report for Oregon. Another weed as- sociated with the water systems for cranberry farming. SPIRAEA TOMENTOSA L. (ROSACEAE).—Coos Co., weed in cranberry fields, with Vaccinium macrocarpon Aiton, Lysimachia terrestris, Aster subspicatus, between Spruce Hollow and Coquille River, elev. <50 m, T28S R14W S20 E%, 20 June 1997, P. F. Zika 13213 & B. Wilson (OSC). Previous knowledge. Native to eastern North America, west to Minnesota. A weed in cranberry fields in New England (Sears et al. 1996). Significance. First record for Oregon. Another weed as- sociated with the cranberry industry. —PETER F ZIKA, KELI KUYKENDALL and BARBARA WIL- SON, Herbarium, Department of Botany and Plant Pathol- ogy, Oregon State University, Corvallis, OR 97331. Volume 47, Number 2, pages 71—146, published 15 March 2001 SUBSCRIPTIONS—MEMBERSHIP Membership in the California Botanical Society is open to individuals ($27 per year; family $30 per year; emeritus $17 per year; students $17 per year for a maximum of 7 years). Late fees may be assessed. Members of the Society receive Maprono free. Institutional subscriptions to MADRONO are available ($60). Membership is based on a calendar year only. Life memberships are $540. Applications for membership (including dues), orders for sub- scriptions, and renewal payments should be sent to the Treasurer. Requests and rates for back issues, changes of address, and undelivered copies of MApRONO should be sent to the Corresponding Secretary. INFORMATION FOR CONTRIBUTORS Manuscripts submitted for publication in MApDRONO should be sent to the editor. It is preferred that all authors be members of the California Botanical Society. Manuscripts by authors having outstanding page charges will not be sent for review. Manuscripts may be submitted in English or Spanish. English-language manuscripts dealing with taxa or topics of Latin America and Spanish-language manuscripts must have a Spanish RESUMEN and an English ABSTRACT. Manuscripts and review copies of illustrations must be submitted in triplicate for all articles and short items (NOTES, NOTEWORTHY COLLECTIONS, POINTS OF VIEW, etc.). Follow the format used in recent issues for the type of item submitted. Allow ample margins all around. Manuscripts MUST BE DOUBLE-SPACED THROUGHOUT. For articles this includes title (all caps, centered), author names (all caps, centered), addresses (caps and lower case, centered), abstract and resumen, 5 key words or phrases, text, acknowledgments, literature cited, tables (caption on same page), and figure captions (grouped as consecutive paragraphs on one page). Order parts in the sequence listed, ending with figures. Each page should have a running header that includes the name(s) of the author(s), a shortened title, and the page number. Do not use a separate cover page or ‘erasable’ paper. Avoid footnotes except to indicate address changes. Abbreviations should be used sparingly and only standard abbrevia- tions will be accepted. Table and figure captions should contain all information relevant to information presented. All measurements and elevations should be in metric units, except specimen citations, which may include English or metric measurements. Authors of accepted papers will be asked to submit an electronic version of the manuscript. Microsoft Word 6.0 or WordPerfect 6.0 for Windows is the preferred software. Line copy illustrations should be clean and legible, proportioned to the MADRONO page. Scales should be in- cluded in figures, as should explanation of symbols, including graph coordinates. Symbols smaller than 1 mm after reduction are not acceptable. Maps must include a scale and latitude and longitude or UTM references. In no case should original illustrations be sent prior to the acceptance of a manuscript. Illustrations should be sent flat. No illustrations larger than 27 X 43 cm will be accepted. Presentation of nomenclatural matter (accepted names, synonyms, typification) should follow the format used by Sivinski, Robert C., in MADRONO 41(4), 1994. Institutional abbreviations in specimen citations should follow Holmgren, Keuken, and Schofield, Index Herbariorum, 8th ed. Names of authors of scientific names should be abbreviated according to Brummitt and Powell, Authors of Plant Names (1992) and, if not included in this index, spelled out in full. Titles of all periodicals, serials, and books should be given in full. Books should include the place and date of publication, publisher, and edition, if other than the first. All members of the California Botanical Society are allotted 5 free pages per volume in MAprono. Joint authors may split the full page number. Beyond that number of pages a required editorial fee of $40 per page will be assessed. The purpose of this fee is not to pay directly for the costs of publishing any particular paper, but rather to allow the Society to continue publishing MApRONo on a reasonable schedule, with equity among all members for access to its pages. Printer’s fees for illustrations and typographically difficult material @ $35 per page (if their sum exceeds 30 percent of the paper) and for author’s changes after typesetting @ $4.50 per line will be charged to authors. At the time of submission, authors must provide information describing the extent to which data in the manu- script have been used in other papers that are published, in press, submitted, or soon to be submitted elsewhere. Ca iS M \s 3 Rial DO a ae als VOLUME 47, NUMBER 3 JULY-SEPTEMBER 2000 MADRONO A WEST AMERICAN JOURNAL OF BOTANY \) k ONTENTS ALLOZYME Data Support A EURASIAN ORIGIN FOR CAREX VIRIDULA SUBSP. VIRIDULA VAR. VIRIDULA (CYPERACEAE) Shannon D. Kuchel and Leo P. Brueder le ...cccccccccccccccccccccseccccsecccusecceueccens 147 THE TAXONOMIC History, IDENTITY, AND DISTRIBUTION OF THE NEVADA ENDEMIC, PLAGIOBOTHRYS GLOMERATUS (BORAGINACEAE) AT HOI LICHT eo tates ene tte ree otc r ee RE mene ere teers 159 GENETIC VARIATION IN PINUS PONDEROSA, PURSHIA TRIDENTATA, AND FESTUCA IDAHOENSIS, COMMUNITY DOMINANT PLANTS OF CALIFORNIA’S YELLOW PINE FOREST Safiya Samman, Barbara L. Wilson, and Valerie D. Hipkins ..............00++- 164 THE EFFECT OF CLIMATIC VARIABILITY ON GROWTH, REPRODUCTION, AND POPULATION VIABILITY OF A SENSITIVE SALT MARSH PLANT SPECIES, LASTHENIA GLABRATA SUBSP. COULTERI (ASTERACEAE) Lorraine S. Parsons and Adam W. Whel che .....0.....ccccccccceccccceeeeccceeeccceees 174 PLEISTOCENE MACROFOSSIL RECORDS OF FOUR-NEEDLED PINYON OR JUNIPER ENCINAL IN THE NORTHERN VIZCAINO DESERT, BAJA CALIFORNIA DEL NORTE Philip V. Wells72 cs eee LON gd WEI a ca eve ccesccsees 189 SEED BANKS OF LONG-UNBURNED STANDS OF MARITIME CHAPARRAL: COMPOSITION, GERMINATION BEHAVIOR, AND SURVIVAL WITH FIRE Dennis-GeOdion £4. \ oR itn vnc cnenncacnanccceness 195 YEW GENUS HEDOSYNE (COMPOSITAE, AMBROSIINAE), A NEw GENUS FOR [VA AMBROSIIFOLIA SOWA VOTER. ZZ ho EES 0c ISU iv ccceccassoeeceess 204 OK REVIEWS 2” INTERFACE BETWEEN ECOLOGY AND LAND DEVELOPMENT IN CALIFORNIA EbiTep By J. E. KEELEY, M. BAER-KEELEY, AND C. J. FOTHERINGHAM. U.S. GEOLOGICAL SURVEY OPEN-FILE REPORT 00-62. SACRAMENTO, CA ITCH ATO. TODDS Fee icsz ti NS Uae aa eee eee se 206 SYNTHESIS OF THE NORTH AMERICAN FLORA. VERSION 1.0 By JOHN T. KARTESZ AND CHRISTOPHER A. MEACHAM RONG eS EL QVM peice. 2c ere 207 ITEWORTHY CO ATETE ORIN come aerate oe Cte eA iene ean octet he Tafa tia al nin oN Aen a ae 209 JLLECTIONS LRIZONACAND AMIE ICOM sit cet cache tur tar, Mere mete Mince see ae eee couse ee 211 COREGONE eon reer Fe ee ee eee 213 WV EAS EIN GION caret eect tee ee es cca eee eNO ace tre calner eee es need erate ee ee er 214 (OUNCEMENTS PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL SOCIETY Maprono (ISSN 0024-9637) is published quarterly by the California Botanical Society, Inc., and is issued from the office of the Society, Herbaria, Life Sciences Building, University of California, Berkeley, CA 94720. Subscription information on inside back cover. Established 1916. Periodicals postage paid at Berkeley, CA, and additional mailing offices. Return requested. Postmaster: Send address changes to Maprono, Roy Buck, % University Herbarium, University of California, Berkeley, CA 94720. Editor—KrisTINA A. SCHIERENBECK California State University, Chico Department of Biology Chico, CA 95929-0515 kschierenbeck @csuchico.edu Editorial Assistant—JEANNIE TRIZZINO Book Editor—Jon E. KEELEY Noteworthy Collections Editors—DIETER WILKEN, MARGRIET WETHERWAX Board of Editors Class of: 2001—Robert PATTERSON, San Francisco State University, San Francisco, CA PAuLA M. ScuIFFMAN, California State University, Northridge, CA 2002—NorMAN ELLSTRAND, University of California, Riverside, CA Cara M. D’ Antonio, University of California, Berkeley, CA 2003—FREDERICK ZECHMAN, California State University, Fresno, CA Jon E. KEELEY, U.S. Geological Service, Biological Resources Division, Three Rivers, CA 2004—Davww M. Woon, California State University, Chico, CA INGRID Parker, University of California, Santa Cruz, CA CALIFORNIA BOTANICAL SOCIETY, INC. OFFICERS FOR 2000—2001 President: BRucE BALDwin, Jepson Herbarium and Dept. of Integrative Biology, 1001 Valley Life Sciences Bldg. #2465, University of California, Berkeley, CA 94720. First Vice President: Rop Myatt, San José State University, Dept. of Biol. Sciences, One Washington Square, San José, CA 95192. rmyatt @email.sjsu.edu Second Vice President: Ros SCHLISING, California State University, Chico, Dept. of Biol. Sciences, Chico, CA 95424. rschlising @csuchico.edu Recording Secretary: DEAN KELCcu, Jepson and University Herbarium, University of California, Berkeley, CA 94720. | dkelch @sscl.berkeley.edu ' Corresponding Secretary: | SUSAN BAINBRIDGE, Jepson Herbarium, University of California, Berkeley, CA 94720. | suebain @SSCL.berkeley.edu Treasurer: Roy Buck, % University Herbarium, University of California, Berkeley, CA 94720. The Council of the California Botanical Society comprises the officers listed above plus the immediate Past President, R. Joun Litre, Sycamore Environmental Consultants, 6355 Riverside Blvd., Suite C, Sacramento, CA 95831; the Editor of MADRONO; three elected Council Members: BIAN Tan, Strybing Arboretum, Golden Gate Park, San Francisco, CA 94122; James SHEvock, USDI National Park Service, Pacific West Region, 600 Harrison Street, Suite 600, San Francisco, CA 94107; DIANE ELam, U.S. Fish and Wildlife Service, 3310 El Camino Avenue, Sacramento, CA 95825; Graduate Student Representative: KirsTEN JoHANUS, Jepson Herbarium, University of California, Berkeley, CA 94720. This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). } Maprono, Vol. 47, No. 3, pp. 147-158, 2000 i j = : (_ ut 09 2008} LIGRARIES ALLOZYME DATA SUPPORT A EURASIAN ORIGIN FOR CAREX VIRIDULA SUBSP. VIRIDULA VAR. VIRIDULA (CYPERACEAE) SHANNON D. KUCHEL! AND LEO P. BRUEDERLE Department of Biology, Campus Box 171, University of Colorado at Denver, P.O. Box 173364, Denver, CO 80127-3364 ABSTRACT Carex viridula Michaux subsp. viridula var. viridula (Cyperaceae), the green sedge, occurs in wetland habitats distributed throughout northern and central North America. Its distribution also extends to the southern Rocky Mountain region in several disjunct sites, including alpine wetlands in Colorado, where it is rare. Populations of C. viridula from Colorado were investigated using starch gel electrophoresis of soluble enzymatic proteins coupled with substrate specific staining in order to describe genetic diversity and structure. The objective was to determine if Colorado populations exhibited the reduced genetic diversity expected of marginal populations when compared to other populations from North America and Europe. Genotypic data were collected for 15 enzyme systems encoded by 21 putative loci in 350 indi- viduals from seven populations in Colorado and in 179 individuals from eight populations from elsewhere throughout the range in North America. Data from all North American populations were compared with data previously reported from European populations of this species by Bruederle and Jensen (1991). No variation, either within or among North American populations, was detected at any of the loci. However, North American populations were genetically differentiated from European populations, with significantly more diversity maintained by European populations. The surprising lack of genetic diversity in North American populations is probably the combined result of high levels of selfing and inbreeding, restricted ecological amplitude, and genetic drift. Genetic bottlenecks are presumed to have occurred as a result of climate changes associated with Pleistocene glaciation or founding events associated with colonization of North America by proposed ancestral European populations. Since 1986, starch gel electrophoresis and allo- zyme analysis have been used to study genetic di- versity in no fewer than 43 species representing nine sections of the genus Carex. These studies have been useful not only in elucidating systematic relationships (e.g., Whitkus 1992; Ford et al. 1998), but also in providing indirect estimates of mating systems (e.g., Waterway 1990), identifying hybrid origins (e.g., Standley 1990), and revealing corre- lations between genetic diversity and structure and certain life-history traits (e.g., Jonsson et al. 1996). However, few studies have examined genetic di- versity in disjunct populations of a broadly distrib- uted species of Carex. Carex section Ceratocystis Dumort. (Cypera- ceae) comprises seven species worldwide, which collectively occur throughout much of the northern hemisphere, particularly in boreal latitudes and the subalpine (Crins and Ball 1988). Carex viridula Michaux subsp. viridula var. viridula, the green sedge, is the only representative of section Cera- tocystis in Colorado. Carex viridula is putatively one of the most re- cently derived members of section Ceratocystis, al- though it is not entirely clear when it diverged from its closest relative, C. viridula subsp. oedocarpa (N. ' Author to whom correspondence should be addressed. Current address: Environmental, Population, and Organ- ismic Biology, Campus Box 334, University of Colorado, Boulder, CO 80309-0334. J. Anderson) B. Schmid (Schmid 1984b; Crins and Ball 1989; Bruederle and Jensen 1991). It has been hypothesized that their common ancestor differen- tiated in West Europe. Thereafter, C. viridula is pre- sumed to have colonized the remainder of the tem- perate and boreal northern hemisphere, perhaps be- fore Pleistocene glaciation, by way of the Bering land bridge (Crins and Ball 1989). Carex viridula is a short-lived perennial with a densely caespitose habit. It is monoecious, charac- terized by a single terminal staminate spike and several sessile pistillate spikes. It has been sug- gested that C. viridula is a dispersal generalist, with possible transport by biotic, e.g., birds and mam- mals, and abiotic agents, e.g., wind and water (Schmid 1984a; Crins and Ball 1989). While there are no apparent impediments to outcrossing, the breeding system is predominantly selfing. In a study by Schmid (1984a) examining the life history of C. viridula, tests for self-compatibility in the field and in experimental gardens using fine mesh bags to control pollination were positive. Addition- ally, inflorescences from which the staminate spike had been removed had maximum seed sets of only 10% when growing in the immediate vicinity of other fertile plants. Similarly, Bruederle and Jensen (1991) attributed low genetic diversity (e.g., pro- portion of polymorphic loci and observed hetero- zygosity) and deviations from Hardy-Weinberg equilibrium in West European populations of C. viridula to selfing. Genetic diversity was appor- 148 tioned among populations with relatively little vari- ation found within populations. Ecologically, C. viridula is characterized by rap- id growth and development, small stature, short life-span, early reproduction, large reproductive ef- fort, and small population size (Schmid 1984a, b). As such, C. viridula is an early successional, r- selected species, or in Grime’s (1979) classification, a ruderal species. Although C. viridula typically oc- cupies moist, early successional sites characterized by fluctuating and unpredictable water levels, these can vary from calcareous, acidic, sandy, or organic shorelines; runnels in limestone barriers; wet mead- ows; marshes; on borders of streams, ponds, and lakes; and fens. This species’ success in colonizing is likely due to its tolerance of diverse and fluctu- ating environments, high phenotypic plasticity, and ability to reproduce quickly and profusely (Schmid 1984a, b; Crins and Ball 1989). Geographically, C. viridula is the most wide- spread taxon in section Ceratocystis, with a near circumboreal distribution. It is common throughout northern Europe, and much of northern and central North America; it 1s also scattered across the cen- tral and eastern parts of temperate Asia to the Pa- cific Ocean (Crins and Ball 1989). In North Amer- ica, its range extends south in the Rocky Mountain region to several disjunct sites in Colorado, Wyo- ming, Utah, and Nevada. In the Southern Rocky Mountains, C. viridula occupies an uncommon hab- itat, alpine wetlands, with habitat specificity con- tributing to rarity in this region (Rabinowitz 1981). The most significant threat to these rare populations may be habitat alteration and loss, as a result of peat mining and the draining of wetlands for irri- gation of surrounding ranchlands and diversion to municipal drinking water supplies. Although C. vir- idula has been assigned a state ranking of S1, in- dicating that it is critically imperiled in Colorado, it has received a global ranking of G5, indicating that it is demonstrably secure globally (Spackman Ct alin 1997), Extant Colorado populations of C. viridula are geographically marginal, occurring at the edge of the species distribution in North America. Further- more, North American populations, in general, are peripheral relative to West Europe, the putative center of diversity and origin for C. viridula (Crins and Ball 1989). Genetic theory predicts differenti- ation of marginal populations with respect to cen- tral populations, with reduced levels of genetic variation and greater population differentiation (Bruederle 1999). While marginal populations are expected to maintain a subset of the genetic varia- tion observed in central populations as a result of reduced gene flow (Yeh and Layton 1979), both random genetic drift and selection may cause the fixation of alleles that are rare in central popula- tions (Blows and Hoffman 1993). However, other factors in addition to distribution may influence levels and apportionment of genetic MADRONO [Vol. 47 diversity in species. Geographical range, succes- sional status, population size, life form, breeding system, and seed dispersal mechanism have all been demonstrated to have significant effects on ge- netic diversity and structure (Brown 1979; Hamrick et al. 1979; Loveless and Hamrick 1984; Karron 1987; Hamrick and Godt 1989; Hamrick et al. 1991; Barrett and Kohn 1991). Furthermore, levels and apportionment of genetic variation could be the consequence not only of life-history characteristics, but also of historical and evolutionary events such as genetic bottlenecks resulting from founder effect, glaciation, migration, and speciation (Lewis and Crawford 1995). The purpose of this research is to describe ge- netic diversity and structure in populations of C. viridula from Colorado relative to other North American and West European populations. Collec- tively, the aforementioned influential factors lead to three specific predictions regarding population ge- netic diversity and structure in C. viridula. First, life-history characteristics, such as self-compatibil- ity, restricted habitat, short-lived perennial growth, herbaceous habit, ruderal strategy, and small pop- ulation sizes, are expected to confer low levels of within-population genetic variation (Schmid 1984a, b; Crins and Ball 1988, 1989; Bruederle and Jensen 1991). Second, deviations from Hardy-Weinberg equilibrium and heterozygote deficiency are ex- pected as a result of the caespitose habit, which has been correlated with high levels of inbreeding and genetic substructuring. Finally, pronounced differ- entiation among populations is expected as a result of reduced gene flow, isolation, inbreeding, and ge- netic drift. It is expected that Colorado populations will be genetically differentiated from other North American populations, and that North American populations, in general, will be genetically differ- entiated from West European populations. MATERIALS AND METHODS Fifteen populations of C. viridula were sampled during the summers of 1998 and 1999 from the Pacific Northwest, Rocky Mountain, and Great Lakes regions of the United States and Canada (Fig. 1, Table 1). Sites were typically peatlands or other wetlands, with C. viridula occupying early succes- sional microsites along the shores of streams, springs, ponds, creeks, and swamps; or along road- sides, ditches, and ruts (Table 1). Population sam- ples ranged in size from six to 50 individuals. Be- cause C. viridula is caespitose, samples obtained from discrete, well-spaced clumps were assumed to represent different individuals. At each site, whole vegetative culms were harvested, placed in separate plastic bags with moist paper towels, and kept re- frigerated until extraction of soluble enzymatic pro- teins. Extraction procedures followed those previ- ously reported in Bruederle and Fairbrothers (1986). Voucher specimens for each population 2000] KUCHEL AND BRUEDERLE: EURASIAN ORIGIN FOR CAREX VIRIDULA 149 Fic. 1. Map of North America showing the locations of 15 Carex viridula subsp. viridula var. viridula populations sampled for allozyme analysis. Detailed locations are provided for seven Colorado populations. Population numbers correspond to those in Table |. have been deposited at the University of Colorado at Denver Herbarium and the University of Colo- rado Herbarium (COLO). Electrophoresis and staining followed Bruederle and Fairbrothers (1986) and Bruederle and Jensen (1991). Three gel-buffer systems and 15 enzyme stains were used to resolve 21 putative loci using 10.5% starch gels (Sigma-Aldrich, Inc.). Isocitrate dehydrogenase (IDH), malate dehydrogenase (MDH), 6-phosphogluconate dehydrogenase (6PGD), phosphoglucose isomerase (PGI), phos- phoglucomutase (PGM), and shikimic acid dehy- drogenase (SkDH) were stained on a discontinuous histidine-HCl system (Gottlieb 1981). Aspartate aminotransferase (AAT), acid phoshatase (ACP), and glyceradehyde-3-phosphate dehydrogenase (G3PDH) were stained on a tris-citrate system (Sol- tis et al. 1983). Alcohol dehydrogenase (ADH), dia- phorase (DIA), malic enzyme (ME), menadione re- ductase (MNR), superoxide dismutase (SOD), and triose phosphate isomerase (TPI) were stained on a discontinuous lithium-borate system (Soltis et al. 1983). Staining followed Soltis et al. (1983) with minor modifications for IDH, MDH, PGI, PGM, SkDH, ACP, G3PDH, ME, and TPI; Gottlieb (1973) for PGD and ADH; and Cardy et al. (1981) for AAT. Electrophoretic allozyme phenotypes were inter- preted genetically on the basis of segregation pat- terns, know substructure and intracellular compart- mentalization of enzymes, and previously observed electrophoretic patterns (e.g., Bruederle and Fair- brothers 1986). Data were collected as individual genotypes; these data have been deposited at the University of Colorado at Denver Herbarium and are available upon request (LPB). Standards rep- resenting most of the common electrophoretic var- iants for the section were incorporated into each of the gels to facilitate allele identification. Data were analyzed using BIOSYS-1 (Swofford and Selander 1981) to obtain common measures of genetic diversity including proportion of polymor- phic loci (P), mean number of alleles per locus (A) and per polymorphic locus (A,), observed hetero- zygosity (H,), and expected heterozygosity (H,). The distribution of genetic variation within and among populations was calculated using Nei’s (1973) gene diversity statistics and GENESTAT-PC (Lewis and Whitkus 1989). In order to assess dif- ferences in genetic diversity, data from North American populations were compared with data previously reported for European populations of this species using a r-test (Table 1; Bruederle and Jensen 1991). RESULTS The fifteen enzymes assayed for C. viridula are encoded by 21 putative loci: AAT-/, AAT-2, ACP- 1, ADH, DIA-1, DIA-3, G3PDH, IDH-1, IDH-2, MDH-1, MDH-2, ME, MNR, 6PGD, PGI-2, PGM- 1, PGM-2, SkDH, SOD, TPI-1, and TPI-2. Four ad- ditional loci (WVDH-3, PGI-1, ACP-2, and ACP-3) were not included in the analysis, because they did not exhibit consistent activity or clearly interpreta- ble banding patterns. North American C. viridu/a maintains no genetic diversity based upon this sample of 15 populations and 21 loci. None of the loci examined were poly- morphic. At every locus assayed, each of the 529 individuals was homozygous for the same allele; no heterozygosity or allozyme variation was observed 150 TABLE l. MADRONO [Vol. 47 | LOCATIONS AND SITE INFORMATION FOR 15 NORTH AMERICAN CAREX VIRIDULA MICHX. SUBSP. VIRIDULA VAR. VIRIDULA POPULATIONS SAMPLED FOR ALLOZYME ANALYSIS, AS WELL AS THREE WEST EUROPEAN POPULATIONS (BRUEDERLE | AND JENSEN 1991). Country, Pop. state, and no. province 1 CO, USA 2 CO, USA 3 CO, USA 4 CO, USA ) CO, USA 6 CO, USA ih CO, USA 8 OH, USA 9 MI, USA 10 MI, USA 1] WI, USA Sample size 50 50 50 50 50 50 50 28 2] 25 21 Location Park Co., Colorado, High Creek Fen, 13 km (8 m) S of Fairplay on U.S. Rte. 285 Park Co., Colorado, Sweet Water Ranch, 18 km (11 m) S of Fairplay on U.S. Rte. 285 Park Co., Colorado, Warm Springs Ranch, 5 km (3 m) S of Fairplay on U.S. Rte. 285 Jackson Co., Colora- do, Lone Pine, 23 km (14 m) W of Walden on Co. Rd. 16 Jackson Co., Colora- do, Bear Creek, 24 km (15 m) W of Walden on Co. Rd. 16 Grand Co., Colora- do, Haystack Mountain, 16 km (10 m) N of Sil- verthorne on St. Rte. 9 San Juan Co., Colo- rado, Andrew’s Lake, 10 km (6 m) S of Silverton on U.S. Rte. 550 Ottawa Co., Ohio, Quarry Rd., 1 km (0.6 m) SW of Lakeside on St. Rte. 163 Iosco Co., Michigan, 6 km (4 m) W of U.S. Rte. 23, N of Alabaster Rd. Mackinac Co., Mich- igan, 13 km (8 m) N of U.S. Rte. 2, W of Borgstrom Rd. Waushara Co., Wis- consin, Hills Lake, 6 km (4 m) E of St. Rte. 22, S of Co. Rd. H Latitude 39°05'N 39°03'N 39°09'N 40°44'N 40°45'N 39°55'N 37°43'N 41°31'N 44°12'N 46°12'N 44°09'N Longitude 105°58'W 105°58'W 106°03'W 106°34'W 106°35'’W 106°19'W 107°42'W 82°45'W 83°37'W 85°21'W 89°09'W Estimated popula- tion size 1000 1200 100 50 50 The 1000 1000 500 50 100 Habitat and microhabitat description scattered throughout well-developed peatland; coloniz- ing alongside the banks of a few streams scattered throughout well-developed peatland; coloniz- ing alongside large, deep ditch dug through peat- land growing on shore of spring and on moist areas adja- cent to spring, alongside shore of pond shore of pond, colo- nizing edge of one bank and on top of a few hum- mocks shore of creek, colo- nizing along fallen log and within ad- jacent rut near outlet of small alkaline spring, some peat accu- mulation, growing on top of small hummocks scattered throughout well-developed peatland; coloniz- ing shores of a few ponds scattered through moist areas in floor of old lime- stone quarry adjacent to swamp, colonizing burrow pit well-drained edge of sandy road shore of lake; grow- ing in rows of re- cent, successive colonizations 2000] KUCHEL AND BRUEDERLE: EURASIAN ORIGIN FOR CAREX VIRIDULA 151 TABLE 1. CONTINUED. Country, Estimated Habitat and Pop. state, and Sample popula- microhabitat no. province size Location Latitude Longitude tion size description 12 WA, USA 28 King Co., Washing- 47°40'N 121°37'W 1000 scattered throughout ton, Snoqualmie large bog/fen sys- Bog, 21 km (13 tem underlain by m) E of St. Rte. peat soil 203, N of N. Fork Co. Rd. 13 ONT, CAN 25 Norfolk Co., Ontar- 42°34'N 80°25'W 100 colonizing middle io, Long Point on and edges of Lake Erie, 48 km raised vehicular (30 m) S of Hwy. track 401 14 ONT, CAN 6 Peterborough Co., 44°40'N 78°80'W 100 shore of lake Ontario, Belmont Lake, 24 km (15 m) N of Hwy. 7 15 ONT, CAN 13 Nipissing Dist., On- —45°60'N 78°30'W 100 shore of lake tario, Radiant Lake, Algonquin Prov. Park, 16 km (10 m) N of Rte. 60 16 AUS 3 Trunnahutte, Austria, 47°30'N 11°70'E na growing in seepage 3.8 km (2.4 m) at base of slope , SSW of Trin 7. SWE 26 Asa, Sweden, 0.9 58°65'N 11°80'E na colonizing along km (0.56 m) hummocks and in WSW depressions in sat- urated soils of meadow 18 SWE 28 Skanor, Sweden, 1 56°10'N 12°90'E na colonizing around km (0.62 m) E (Table 2). As such, for all populations, the propor- tion of loci polymorphic (P), observed heterozy- gosity (H,), and expected heterozygosity (H,) were all zero. Similarly, the number of alleles per locus (A) was one, the minimum value for this statistic (Table 3). While low, West European populations exhibited higher levels of genetic diversity than North Amer- ican populations (Bruederle and Jensen 1991). On average, two of the 20 loci examined for West Eu- ropean populations (10.0%) were polymorphic (Ta- ble 2). Mean number of alleles per locus was 1.15, while mean number of alleles per polymorphic lo- cus was 2.0. Observed heterozygosity was 0.014 and expected heterozygosity was 0.039 (Table 3). Genetic diversity in West European populations was significantly higher than that in North Ameri- can populations when compared using a two sam- ple t-test assuming unequal variances for proportion of polymorphic loci (P < 0.10), mean number of alleles per locus (P < 0.05), observed heterozygos- w (P < 0.05), and expected heterozygosity (P < .10). rocks, hummocks, and in depressions of low, pastured meadow Levels of genetic diversity for the species across its sample range in North America and West Europe were extremely low. The mean proportion of poly- morphic loci was 1.7%. Mean number of alleles per locus was 1.03, while mean number of alleles per polymorphic locus was 2.0. Observed heterozygos- ity was 0.002 and expected heterozygosity was 0.007 (Table 3). Despite the differences in levels of genetic di- versity, West European and North American pop- ulations are, in fact, very similar. Mean genetic identity obtained from pairwise comparisons of populations from North America and West Europe was 0.987. Mean genetic identity among North American populations was 1.000, and among Eu- ropean populations was 0.974, ranging from 0.960 to 1.000 for the latter. Of the small amount of di- versity maintained by populations of C. viridula, the majority (G,, = 0.650) was due to differences among populations, both among West European and between North American and West European populations. ALLELE FREQUENCIES AT THREE POLYMORPHIC LOCI FOR 18 POPULATIONS OF CAREX VIRIDULA MICHX. SUBSP. VIRIDULA VAR. VIRIDULA. Population numbers correspond to those in Table 1. See text for allozyme nomenclature. TABLE 2. European Populations North American Populations 18 0.04 0.96 0.07 0.93 0.96 0.00 0.04 ig) 0.00 1.00 0.06 0.94 1.00 0.00 0.00 16 0.00 1.00 0.00 1.00 0.00 15 0.00 1.00 0.00 1.00 1.00 0.00 0.00 14 0.00 1.00 0.00 1.00 1.00 0.00 0.00 13 0.00 1.00 0.00 1.00 1.00 0.00 0.00 12 0.00 1.00 0.00 1.00 1.00 0.00 0.00 10 0.00 om N | 0.00 Allele Locus 0.00 1.00 0.00 1.00 1.00 0.00 0.00 0.00 1.00 0.00 1.00 1.00 0.00 0.00 0.00 1.00 0.00 1.00 1.00 0.00 0.00 0.00 1.00 0.00 1.00 1.00 0.00 0.00 0.00 1.00 0.00 1.00 1.00 0.00 0.00 0.00 0.00 1.00 0.00 1.00 1.00 0.00 0.00 0.00 1.00 0.00 1.00 1.00 0.00 0.00 0.00 1.00 0.00 1.00 1.00 0.00 0.00 b PGI-2 1.00 0.00 1.00 0.00 ADH 1.00 1.00 0.00 0.00 1.00 1.00 0.00 0.00 SOD 0.667 0.333 MADRONO [Vol. 47 DISCUSSION As expected, North American populations of C. viridula do maintain extremely low levels of ge- netic diversity within populations. However, due to the low levels of genetic diversity within popula- tions and subsequent lack of allozyme markers, no genetic differentiation was observed among popu- lations in Colorado or North America. Neverthe- less, North American populations were genetically differentiated from West European populations, with significantly more diversity maintained by West European populations. A number of factors could have contributed to the paucity of genetic di- versity observed in populations of this species. First, the plant allozyme literature reveals strong associations between genetic diversity and breeding system, with those species characterized by out- crossed breeding systems having significantly high- er levels of genetic diversity apportioned among in- dividuals within populations. On average, species characterized by selfing breeding systems maintain significantly lower levels of genetic diversity, in- cluding proportion of polymorphic loci, number of alleles per locus and per polymorphic locus, and observed heterozygosity (Brown 1979; Hamrick et al. 1979; Loveless and Hamrick 1984; Hamrick and Godt 1989; Hamrick et al. 1991). Carex viridula has been shown to exhibit popu- lation genetic structure and seed set suggestive of selfing (Schmid 1984a; Bruederle and Jensen 1991). The extremely low levels of genetic varia- tion found in North American populations of C. viridula in this study may be the result, in part, of such selfing. High levels of inbreeding attributable to selfing are expected to result in homozygosity and decreased genetic variability. Of the large num- ber of species of vascular plants that have been ex- amined similarly, at least 14 other taxa have been reported to maintain no detectable allozyme diver- sity (Table 4). Although an exact comparison be- tween these taxa and C. viridula is not possible, it is noteworthy that almost all of these taxa also show substantial levels of selfing. In graminoids, high levels of inbreeding have also been correlated with the caespitose growth form. Stebbins (1950) proposed a relationship be- tween growth form and breeding system among grasses, suggesting that species with a rhizomatous growth form are predominantly outcrossing due to the intermingling of genets. In contrast, the caes- pitose habit results in a growth form in which the nearest neighbor of a flowering culm is another culm from the same plant (e.g., ramet), thus pro- moting inbreeding, and specifically, selfing. Genetic evidence substantiating this phenomenon in the gra- minoid genus Carex was first reported by Bruederle and Fairbrothers (1986) and Bruederle (1987). Ad- ditional evidence supporting the relationship be- tween growth form and genetic variability in Carex 2000] TABLE 3. KUCHEL AND BRUEDERLE: EURASIAN ORIGIN FOR CAREX VIRIDULA 153 SUMMARY OF GENETIC DIVERSITY FOR 18 POPULATIONS OF CAREX VIRIDULA MICHX. SUBSP. V/RIDULA VAR. VIRI- DULA: SAMPLE SIZE (N), MEAN NUMBER OF ALLELES PER LOCUS (A) AND PER POLYMORPHIC LOCUS (A,,), PROPORTION OF POLYMORPHIC Loci (P), OBSERVED HETEROZYGOSITY (H,), AND EXPECTED HETEROZYGOSITY (#7.). Population numbers correspond to those in Table 1. A locus was considered polymorphic if the frequency of the most common allele did not exceed 0.95. Pop. No. N A 1 50 1.0 2 50 1.0 3 50 1.0 4 50 1.0 5) 50 1.0 6 50 1.0 7 50 1.0 8 28 1.0 ) ay 1.0 10 25 1.0 11 2] 1.0 12 28 1.0 13 25 1.0 14 6 1.0 15 13 1.0 Mean-N. America 529 1.0 16 3 1.15 17 26 1.05 18 28 125 Mean-W. Europe a7. 1.15 Mean-Species 586 1.03 was subsequently provided by Ford et al. (1991, 1998). A survey of the population genetic literature for the genus Carex revealed data for 29 taxa including six rhizomatous and 23 caespitose carices (Kuchel, unpubl.). On average, populations of rhizomatous species harbor high levels of genetic diversity, e.g., A — 2.26 = 0.12, P = 44.5 > 4.08%, and. i, = 0.171 + 0.038, while caespitose species have sig- nificantly less, e.g., A, = 2.03 + 0.09 (P < 0.05), P= 13.4 + 12.0% (P < 0.001), and H, = 0.042 + 0.04 (P < 0.001). Furthermore, whereas popu- lations of rhizomatous species are poorly differen- tiated (G,, = 0.159 + 0.053), caespitose species are well-differentiated with nearly half of all genetic diversity attributable to differences among popula- tions (G,, = 0.462 + 0.272). Although exceptions exist (Ford et al. 1998), it would appear that rhi- zomatous species maintain more variation within and less differentiation among populations, presum- ably due to outcrossing. Conversely, caespitose species have less variation within and more differ- entiation among populations, presumably due to in- breeding. As such, the extremely low levels of ge- netic variation found in North American popula- tions of C. viridula in this study may be the result, in part, of the caespitose growth form. Second, narrowly distributed plant species tend to maintain lower levels of genetic variation than more widespread species (Karron 1987; Karron et al. 1988; Hamrick and Godt 1989). The lower lev- els of genetic variation observed in narrowly dis- tributed species may be due to changes in allele A, P H, H. =. 0.0 (0.000 0.000 = 0.0 0.000 0.000 = 0.0 0.000 0.000 = 0.0 0.000 0.000 = 0.0 0.000 0.000 = 0.0 0.000 0.000 a 0.0 0.000 0.000 = 0.0 0.000 0.000 == 0.0 0.000 0.000 = 0.0 0.000 0.000 — 0.0 0.000 0.000 = 0.0 0.000 0.000 = 0.0 0.000 0.000 = 0.0 0.000 0.000 eu 0.0 0.000 0.000 = 0.0 0.000 0.000 2.0 15.0 0.017 0.070 2.0 5.0 0.006 0.006 2.0 10.0 0.018 0.042 2.0 10.0 0.014 0.039 2.0 1.7 0.002 0.007 frequencies due to chance (genetic drift and found- er effect) or strong, directional selection toward ge- netic uniformity in a limited habitat type (Karron 1987). Almost all of the aforementioned genetically invariable taxa are narrowly distributed (Table 4). Wolff and Jefferies (1987) hypothesized that the lack of diversity in one of these taxa, Salicornia europaea L., could be due to its restricted ecolog- ical distribution, despite the fact that it is geograph- ically widespread. Ecologically, this species is con- fined to chronically disturbed, early successional open habitats in coastal and inland salt marshes where individuals and populations are subject to considerable turnover and population re-establish- ment. The narrow habitat requirements, founding events, small population sizes, and possible selec- tion pressures experienced in such an environment could have contributed to the observed paucity of genetic diversity in S. europaea. Even though Carex viridula is_ distributed throughout boreal North America, its ecological distribution also appears to be narrow. Carex viri- dula is a habitat specialist, occurring only in highly disjunct wetland habitats. Additionally, C. viridula is confined to early successional microsites, which tend to be small, ephemeral, highly variable, and subject to repeated local extinction and coloniza- tion. As in Salicornia europaea, it is possible that the extremely low levels of genetic variation found in populations of Carex viridula in this study may be, in part, the result of this narrow ecological dis- tribution. It is interesting to note that a number of species 154 MADRONO [Vol. 47 TABLE 4. SUMMARY OF ALLOZYME LITERATURE FOR THOSE SPECIES HAVING No DETECTABLE ALLOZYME VARIATION. Geographic range/ Species Breeding system Ecological amplitude Inferred historical mechanisms Reference Carex viridula selfing narrow distribution Genetic drift: inbreeding ecologically and genetic bottleneck associated with founding events and/or climate changes during glaciation Rapid colonizer Disturbed habitats Bensoniella ore- _ selfing narrow distribution Genetic drift: inbreeding Soltis et al. 1992 gona (Abrams geographically and genetic bottleneck & Bacig) associated with range re- Morton strictions during glacia- tion Clonal growth Small population sizes Chrysosplenium _ selfing narrow distribution Genetic drift: inbreeding Schwartz 1985 liowense geographically and and genetic bottleneck Rydb. ecologically associated with climate changes during glaciation Clonal growth Small population sizes Howellia aqua- approaches obli- narrow distribution Genetic drift: inbreeding Lesica et al. 1988 tilis Gray gate selfing geographically and and genetic bottleneck ecologically associated with range re- strictions during glacia- tion Age of populations, not enough time to accumu- late variability and het- erozygosity Lespedeza lep- selfing narrow distribution Genetic drift: inbreeding Cole and Biesboer tostachya En- geographically and genetic bottleneck 1992 gelm. associated with range re- strictions during glacia- tion Oenothera hook- _ selfing narrow distribution Genetic drift: inbreeding Levy and Levin 1975 eri Torr. and geographically Age of populations, not Gray enough time to accumu- late variability and het- erozygosity Rapid colonizer Permanent translocation heterozygosity Pedicularis fur- pollinator re- narrow distribution Genetic drift: inbreeding Waller et al. 1987 bishiae S. quired for pol- geographically and and genetic bottleneck Wats lination, but ecologically associated with range re- possibly self- strictions during glacia- compatible tion and/or founding events Local population extinctions Disturbed habitats Pinus resinosa highly self-com- widespread geographi- Genetic drift: inbreeding Fowler and Morris Ait. patible cally and genetic bottleneck 1977; Allendorf et associated with range re- al. 1982; Simon et strictions during glacia- al. 1986; Mosseler tion et al. 1991 Age of populations, not enough time to accumu- late variability and het- erozygosity 2000] KUCHEL AND BRUEDERLE: EURASIAN ORIGIN FOR CAREX VIRIDULA TABLE 4. CONTINUED. Species Breeding system Geographic range/ Ecological amplitude Inferred historical mechanisms Reference Salicornia euro- paea L. (s.1.) Senecio mohav- ensis Gray Sullvantia ore- gana S. Wats. Taraxacum obli- quum (Fr.) Dahlst. selfing obligate selfing selfing agamospermous narrow distribution ecologically narrow distribution geographically narrow distribution geographically and ecologically narrow distribution geographically and ecologically Genetic drift: inbreeding and genetic bottleneck associated with range re- strictions during glacia- tion and founding events Rapid colonizer Genetic drift: genetic bottle- neck associated with founding events (recent colonization of North America) and/or climate changes during glaciation Rapid colonizer Genetic drift: inbreeding and genetic bottleneck associated with range re- strictions during glacia- tion Post-glacial range expan- sion through hybridiza- tion to form polyploid Jefferies and Gottlieb 1982; Wolff and Jefferies 1987 Liston et al. 1989 Solits 1982 Van Oostrum et al. 1985 Thuja plicata self-compatible Donn ex D. geographically Don Tragopogon selfing widespread geographi- pratensis cally Ownbey Typha domin- selfing widespread geographi- gensis Pers. cally reported to maintain no detectable allozyme diver- sity are rapid colonizers of disturbed habitats (Table 4). Carex viridula has also been described as a rap- id colonizer and ruderal species with rapid growth and development, small size, short life-span, early reproduction, large reproductive effort, and small population size (Schmid 1984a, b). In Switzerland, C. viridula often occupies newly disturbed sites, with many small and isolated populations. It is a pioneer in open, wet habitats, but is quickly ex- cluded successionally (Schmid 1986). Not surpris- ingly, most populations sampled for this study ap- pear to occupy early successional microsites, grow- ing along pond shores, stream banks, roadsides, ditches, and ruts, comprising a part of larger, later successional communities (Table 1). Schmid (1984b) hypothesized that such ruderal species with early successional populations would have high ge- narrow distribution cally and ecologi- agamospermous popula- tions Selection in more severe and less diverse environ- ments No explanation given Genetic bottleneck associat- ed with range restrictions unlikely; no evidence of physical barriers and as- sociated species show abundant genetic varia- tion Genetic drift: inbreeding Rapid colonizer Copes 1981 Roose and Gottlieb 1976 Mashburn et al. 1978: Sharitz et al. 1980 Genetic drift: inbreeding Clonal growth Rapid colonizer Disturbed habitats netic variability between, but low genetic variabil- ity and high plasticity within populations, as a re- sult of small population size, genetic drift, and di- rectional selection. This hypothesis is supported by the ecology of C. viridula (Schmid 1984b), as well as the present study of genetic diversity. Thus, low levels of genetic diversity found with- in populations of C. viridula may be attributed, in part, to effective breeding system and restricted ecological distribution. However, populations of this species in North America show substantially lower levels of genetic variability, even when com- pared to those means reported for selfing or nar- rowly distributed species—it should be reiterated that all putative loci examined were monomorphic. The most likely explanation for such low levels of polymorphism and heterozygosity is genetic drift. It is possible that a genetic bottleneck occurred at 156 some point in the history of these populations that eliminated all or most of the allozyme polymor- phism. One possibility is that a genetic bottleneck re- sulted from the founding of North American pop- ulations. During migration and dispersal, new pop- ulations may be formed by a small number of initial colonists. The genetic material of such populations is limited to those alleles introduced by these few founders and may not be representative of the spe- cies as a whole (Schwaegerle and Schaal 1979). Since European populations are the proposed pro- genitors of North American populations (Crins and Ball 1989), it would be expected that the allozymes present in North American populations would largely comprise a subset of those alleles present in European populations (Crawford 1983; Cole and Biesboer 1992). Indeed, North American popula- tions are genetically differentiated from West Eu- ropean populations, with West European popula- tions of C. viridula harboring higher levels of ge- netic diversity, and North American populations harboring a subset of that genetic diversity. These data suggest that a small number of individuals from the putatively ancestral European populations founded North American populations. Limited gene flow between populations would likely maintain the lower levels of genetic diversity observed in North American populations. Although populations of Bromus tectorum L. have only recently been intro- duced to North America from Eurasia, a number of similarities can be seen between this species and C. viridula. Both species exhibit low genetic variabil- ity within and high differentiation among popula- tions possibly as a result of founding events and a selfing breeding system. Ecologically, both are characterized by routinely disturbed habitats and high phenotypic plasticity (Novak et al. 1991). Another possibility is that a genetic bottleneck occurred at some point after the founding of North American populations. This bottleneck could have resulted in genetic uniformity in an original popu- lation or populations with a reduced geographic distribution, followed by spread of this species to the range now occupied (Lesica et al. 1988; Cole and Biesboer 1992). Climatic changes during the Pleistocene, particularly the Xerothermic period about 8500 to 3000 y B.P., have been suggested as a possible cause for genetic bottlenecks in a number of other North American species exhibiting ex- tremely low levels of genetic diversity (Table 4), as well as in other species of Carex (Waterway 1990). Additionally, it has been suggested that these pop- ulations have not had enough time to accumulate variation and differentiate since these events (Levy and Levin 1975; Ledig and Conkle 1983; Lesica et al. 1988; Liston et al. 1989). It has been suggested that allozyme variation de- tectable by electrophoresis may not provide a com- plete measure of genetic diversity in the genome (Mosseler et al. 1991; Mosseler et al. 1992). A MADRONO [Vol. 47 more direct analysis of variation in DNA, e.g., RAPDs, may provide the genetic markers necessary to infer genetic diversity and structure in C. viri- dula. Data from the present study do indicate that C. viridula is genetically depauperate over a large portion of its range. However, additional analyses of populations lying at the extreme northwestern, northeastern, and eastern edges of the distribution in North America, as well as in putative glacial re- fugia, should be carried out to confirm this. Anal- yses of additional Eurasian populations would also contribute to a reconstruction of the biogeography of C. viridula. ACKNOWLEDGMENTS We are grateful to the following people and organiza- tions for their assistance with collections: P. W. Ball, Uni- versity of Toronto; W. J. Crins, Ontario Ministry of Nat- ural Resources; A. Cusick, Ohio State Department of Nat- ural Resources; J. Sanderson, Colorado Natural Heritage Program; F Weinmann; and S. Komarek. We thank P. W. Ball, University of Toronto; B. A. Ford, University of Manitoba; K. A. Schierenbeck, California State Univer- sity, Chico; and D. EF Tomback and T. M. Roane, Univer- sity of Colorado at Denver, for review of the manuscript. This work was supported, in part, by grants from the Sig- ma Xi Grants-in-Aid of Research Program and the Colo- rado Native Plant Society. LITERATURE CITED ALLENDORF, E W., K. L. KNUDSEN AND G. M. BLAKE. 1982. Frequencies of null alleles at enzyme loci in natural populations of ponderosa and red pine. Ge- netics 100:497—504. BARRET, S. C. H. AND J. KOHN. 1991. Genetic and evo- lutionary consequences of small population size in plants: implications for conservation, p. 3—30. in D. A. Falk and K. E. Holsinger (eds.). Genetics and con- servation of rare plants. Oxford University Press, New York, N.Y. BLows, M. W. AND A. A. HOFFMAN. 1993. The genetics of central and marginal populations of Drosophila serrata. 1. Genetic variation for stress resistance and species borders. Evolution 47:1255—1270. Brown, A. H. D. 1979. Enzyme polymorphism in plant populations. Theoretical Population Biology 15:1—42. BRUEDERLE, L. P. 1999. Genetic differentiation of geo- graphically marginal populations in Carex mitchelli- ana (Cyperaceae): implications for conservation. Journal of the Torrey Botanical Society 126:1-8. . 1987. Allozyme variation in the genus Carex (Cyperaceae). XIV International Botanical Congress Abstracts 14:304. , AND D. E. FAIRBROTHERS. 1986. Allozyme vari- ation in populations of the Carex crinita complex (Cyperaceae). Systematic Botany 11:583-—594. , AND U. JENSEN. 1991. Genetic Differentiation of Carex flava and Carex viridula in West Europe (Cy- peraceae). Systematic Botany 16:41—49. Carpy, B. J., C. W. STUBER AND M. M. GoopMan. 1981. Techniques for starch gel electrophoresis of enzymes from maize (Zea mays L.). Institute of Statistics Mim- eograph Series No. 1317. North Carolina State Uni- versity, Raleigh, NC. Cote, C. T. AND D. D. BresBoER. 1992. Monomorphism, | § 2000] | | } reduced gene flow, and cleistogamy in rare and com- mon species of Lespedeza (Fabaceae). American Journal of Botany 79:567—575. _ Cores, D. L. 1981. Isoenzyme uniformity in western red cedar seedlings from Oregon and Washington. Ca- nadian Journal of Forest Research 11:451—453. ~ CRAWFORD, D. J. 1983. Phylogenetic and systematic in- ferences from electrophoretic studies, p. 257-87. in S. D. Tanksley and T. J. Orton (eds.). Isozymes in plant genetics and breeding, part A. Elsevier Science Publishers B. V., Elsevier, Amsterdam. CRINS, W. J., AND P. W. BALL, 1988. Sectional limits and phylogenetic considerations in Carex Section Cera- tocystis (Cyperaceae). Brittonia 40:38—47. . 1989. Taxonomy of the Carex flava complex (Cy- peraceae) in North America and northern Eurasia. II. Taxonomic treatment. Canadian Journal of Botany 67: 1048-1065. Fow er, D. P. AND R. W. Morris. 1977. Genetic diversity in red pine: evidence for low genic heterozygosity. Canadian Journal of Forest Research 7:343—347. Forb, B. A., P- W. BALL AND K. RITLAND. 1991. Allozyme diversity and genetic relationships among North American members of the short-beaked taxa of Carex sect. Vesicariae (Cyperaceae). Systematic Botany 16: 116-131. , D. A. R. MCQUEEN, R. E C. NAcZI AND A. A. REZNICEK. 1998. Allozyme variation and genetic re- lationships among species in the Carex willdenowii complex (Cyperaceae). American Journal of Botany 85:546—-552. GoTTLigB, L. D. 1973. Genetic differentiation, sympatric speciation, and the origin of a diploid species of Stephanomeria. American Journal of Botany 69:545— Do 3: . 1981. Gene number in species of Astereae that have different chromosome numbers. Proceedings of the National Academy of Sciences, USA 78:3726- 3729. GRIME, J. P. 1979. Plant Strategies and Vegetation Pro- cesses. John Wiley and Sons, New York, N.Y. HAMRICK, J. L., Y. B. LINHART AND J. B. MitTTon. 1979. Relationships between life history characteristics and electrophoretically detectable genetic variation in plants. Annual Review of Ecology and Systematics 10:173—200. , AND M. J. W. Gopt. 1989. Allozyme diversity in plant species, p. 43-63. in A. H. D. Brown, M. T. Clegg, A. L. Kahler and B. S. Weir (eds.). Plant pop- ulation genetics, breeding and genetic resources. Sin- auer, Sunderland, MA. , D. A. MURAWSKI AND M. D. LOVELESS. 1991. Correlations between species traits and allo- zyme diversity: implications for conservation biolo- gy, p. 75-86. in D. A. Falk and K. E. Holsinger (eds.). Genetics and conservation of rare plants. Oxford Uni- versity Press, New York, N.Y. JEFFERIES, R. L. AND L. D. GOTTLIEB. 1982. Genetic dif- ferentiation of the microspecies Salicornia europaea L. (sensu stricto) and S. ramosissima J. Woods. New Phytology 92:123-129. JONSSON, B. O., I. S. JONSDOTTIR AND N. CRONBERG. 1996. Clonal diversity and allozyme variation in popula- tions of the arctic sedge Carex bigelowii (Cypera- ceae). Journal of Ecology 84:449—459. KARRON, J. D. 1987. A comparison of levels of genetic polymorphism and self-compatibility in geographi- KUCHEL AND BRUEDERLE: EURASIAN ORIGIN FOR CAREX VIRIDULA log cally restricted and widespread plant congeners. Evo- lutionary Ecology 1:47—S58. , Y. B. LINHART, C. A. CHAULK, AND C. A. ROB- ERTSON. 1988. Genetic structure of populations of geographically restricted and widespread species of Astragalus (Fabaceae). American Journal of Botany 75:1114-1119. Lepic, EF T. AND M. T. CoNKLE. 1983. Gene diversity and genetic structure in a narrow endemic, Torrey pine (Pinus torreyana Parry ex. Carr). Evolution 37:79— 85. Lesica, P., R. KR LEARY, R. W. ALLENDORF AND D. E. BILD- ERBACK. 1988. Lack of genetic diversity within and among populations of an endangered plant, Howelia aquatilis. Conservation Biology 2:275—282. Levy, M. AND D. A. LEvIN. 1975. Genic heterozygosity and variation in permanent translocation heterozy- gotes of the Oenothera biennis complex. Genetics 79: 493-512. Lewis, P. O. AND D. J. CRAWFORD. 1995. Pleistocene re- fugium endemics exhibit greater allozymic diversity than widespread congeners in the genus Polygonella (Polygonaceae). American Journal of Botany 82:141— 149. , AND R. WHITKUS. 1989. GENESTAT-PC. Amer- ican Society of Plant Taxonomists Newsletter 2:15— 16. Liston, A., L. H. RIESEBERG AND T. S. ELIAs. 1989. Ge- netic similarity is high between intercontinental dis- junct species of Senecio (Asteraceae). American Jour- nal of Botany 76:383—388. Lovetess, M. D. AND J. L. HAMRICK. 1984. Ecological determinants of genetic structure in plant populations. Annual Review of Ecology and Systematics 15:65— oF: MASHBuURN, S. J., R. R. SHARITZ AND M. H. SmituH. 1978. Genetic variation among Typha populations of the southeastern United States. Evolution 32:68 1—685. MOsSsELeR, A., D. J. INNES AND B. A. RoBerts. 1991. Lack of allozymic variation in disjunct Newfoundland pop- ulations of red pine (Pinus resinosa). Canadian Jour- nal of Forest Research 21:525—528. , K. N. EGGER AND G. A. HUGHES. 1992. Low lev- els of genetic diversity in red pine confirmed by ran- dom amplified polymorphic DNA markers. Canadian Journal of Forest Research 22:1332—1337. Ner, M. 1973. Analysis of gene diversity in subdivided populations. Proceedings of the National Academy of Sciences, USA 70:3321-—3323. Novak, S. J., R. N. MACK AND D. E. SoLtis. 1991. Genetic variation in Bromus tectorum (Poaceae): population differentiation in its North American Range. Ameri- can Journal of Botany 78:1150—1161. RABINOWITZ, D. 1981. Seven forms of rarity, p. 205-217. in H. Synge (ed.). The biological aspects of rare plant conservation. John Wiley and Sons, Chichester, En- gland. Roose, M. L. AND L. D Gort ics. 1976. Genetic and bio- chemical consequences of polyploidy in Tragopogon. Evolution 30:818—830. ScHMID, B. 1984a. Life histories in clonal plants of the Carex flava group. Journal of Ecology 72:93-114. . 1984b. Niche width and variation within and be- tween populations in colonizing species (Carex flava group). Oecologia 63:1—5. . 1986. Colonizing plants with persistent seeds and persistent seedlings (Carex flava group). Botanica Helvetica 96:19-—26. 158 SCHWAEGERLE, K. E. AND B. A. SCHAAL. 1979. Genetic variability and founder effect in the pitcher plant Sar- racenia purpurea L. Evolution 33:1210—1218. SCHWARTZ, A. O. 1985. Lack of protein polymorphism in the endemic relict Chrysosplenium iowense (Saxifra- gaceae). Canadian Journal of Botany 63:2031—2034. SHARITZ, R. R., S. A. WINERITER, M. H. SMITH AND E. H. Liu. 1980. Comparison of isozymes among Typha species in the eastern United States. American Jour- nal of Botany 67:1297—1303. SIMON, J.-P., Y. BERGERON AND D. GAGNON. 1986. Isozyme uniformity in populations of red pine (Pinus resinosa) in the Abitibi region, Quebec. Canadian Journal of Forest Research 16:1133—1135. SoLTis, D. E. 1982. Allozymic variability in Sullvantia (Saxifragaceae). Systematic Botany 7:26—34. , C. H. HAUFLER, D. C. DARROW AND G. J. GAs- TONY. 1983. Starch gel electrophoresis of ferns: a compilation of grinding buffers, gel and electrode buffers, and staining schedules. American Fern Jour- nal 73:9—27. SoLTis, P. S., D. E. Sottis, T. L. TUCKER AND F A. LANG. 1992. Allozyme variability is absent in the narrow endemic Bensoniella oregona (Saxifragaceae). Con- servation Biology 6:131—134. SPACKMAN, S., B. JENNINGS, J. COLES, C. DAwson, M. MINTON, A. KRATZ AND C. SPURRIER. 1997. Colorado Rare Plant Field Guide. Prepared for the Bureau of Land Management, U. S. Forest Service and U. S. Fish and Wildlife Service by the Colorado Natural Heritage Program, Ft. Collins, CO. STANDLEY, L. A. 1990. Allozyme evidence for the hybrid MADRONO [Vol. 47. origin of the maritime species Carex salina and Car- | ex recta (Cyperaceae) in Eastern North America. Sys- | tematic Botany 15:182-191. | STEBBINS, G. L. 1950. Variation and evolution in plants. , Columbia University Press, New York, N.Y. | SWOFFORD, D. L. AND R. B. SELANDER. 1981. BIOS YS-1. | A Fortran program for the comprehensive analysis of | electrophoretic data in population genetics and sys- 7 tematics. Journal of Heredity 72:281-283. VAN OostTrRuM, H., A. A. STERK AND H. J. W. WISMAN. | 1985. Genetic variation in agamospermous microspe- \ cies of Taraxacum sect. Erythrosperma and sect. Ob- liqua. Heredity 55:223—228. WALLER, D. M., D. M. O’MALLEY AND S. C. GAWLER. 1987. Genetic variation in the extreme endemic Ped- | icularis furbishiae (Scrophulariaceae). Conservation Biology 1:335-—340. WATERWAY, M. J. 1990. Genetic differentiation and hy- | bridization between Carex gynodynama and C. men- | docinensis (Cyperaceae) in California. American Journal of Botany 77:826—838. WHITKUS, R. 1992. Allozyme variation within the Carex pachystachya complex (Cyperaceae). Systematic Bot- | any 13:146—153. Wo rr, S. L. AND R. L. JEFFERIES. 1987. Morphological | and isozyme variation in Salicornia europaea (s.1.) | (Chenopodiaceae) in northeastern North America. | Canadian Journal of Botany 65:1410—-1419. YEH, EF C. AND C. LAYTON. 1979. The organization of | genetic variability in central and marginal populations of lodgepole pine (Pinus contorta spp. latifolia). Ca-— nadian Journal of Genetics and Cytology 21:487—503. | - Maprono, Vol. 47, No. 3, pp. 159-163, 2000 THE TAXONOMIC HISTORY, IDENTITY, AND DISTRIBUTION OF THE NEVADA ENDEMIC, PLAGIOBOTHRYS GLOMERATUS (BORAGINACEAE) ARNOLD TIEHM 1550 Foster Dr., Reno, NV 89509 ABSTRACT Plagiobothrys glomeratus A. Gray is a western Nevada endemic restricted to areas of altered andesite. It is morphologically close to P. hispidus A. Gray. Plagiobothrys hispidus is more widespread and shows much variation as to number and size of nutlets. Reports of P. glomeratus from California are based on misidentifications of P. hispidus. Illustrations of the nutlets and a distribution map of both species are included. Current floristic studies in the Pine Nut Moun- tains of western Nevada uncovered a problem in defining Plagiobothrys glomeratus A. Gray. This study was undertaken to clarify the identity and dis- tribution of P. glomeratus. Gray (1885) described Plagiobothrys glomeratus from two collections by Katharine Curran (later Brandegee) taken between Virginia City and Car- son City, Nevada. In the same article Gray de- scribed P. hispidus A. Gray based on a collection, again taken by Curran, from the streets of Truckee in nearby California. These two sites are approxi- mately 27 air miles apart. Both taxa are members of section Plagiobothrys characterized by alternate leaves, lateral nutlet scars placed near or above the center of the nutlets, and not growing in seasonally saturated soils. They share the characteristic of hav- ing rather broad upper cauline leaves with P. jonesii A. Gray and P. kingii (S. Watson) A. Gray. Both P. kingii and P. jonesii have elongated nutlet scars along the ventral keel and an earlier spring flow- ering time in contrast with nutlet scars about as wide as long, placed at the end of the ventral keel, and a later spring to summer flowering time in P. hispidus and P. glomeratus. As such P. hispidus and P. glomeratus are more similar to each other than to any other species. The first treatment of P. glomeratus is that of Greene (1887) who described the genus Sonnea to accomodate P. glomeratus, hispidus, jonesii, and kingii. He later described Sonnea foliacea from the geographic area between the type localities of P. glomeratus and P. hispidus (Greene 1888). Johnston (1923) published a synopsis of Pla- giobothrys placing glomeratus and hispidus in his Sonnea group and kingii and jonesii in his Amsinck- lopsis group. He also reduced Greene’s Sonnea fol- lacea to a variety of P. hispidus and stated: ‘‘It is possible that the plant is a hybrid between P. his- pidus and P. glomeratus.”’ Tidestrom (1925) in his Flora of Utah and Ne- vada recognized the distinctiveness of P. hispidus and P. glomeratus but followed Greene in placing them in the genus Sonnea. He also maintained S. foliacea as a good species. Cronquist (1984) recognized P. glomeratus as an acceptable species with the comment “‘Reno south nearly to Carson City, rarely collected.’’ The only other published references for the distribution of P. glomeratus in Nevada are from the south side of Peavine Mountain where it is reported as occurring in an open pine stand (Billings 1992; Williams et rel Th 2) Plagiobothrys glomeratus is not included or mentioned in any flora covering California (Abrams 1951; Jepson 1925, 1943; Messick 1993; Munz 1968; Munz, and Keck 1959). Since none of the above references mention P. glomeratus or place it in synonymy I assume they did not have any evi- dence to believe it occurred in California. DeDecker (1990) reported P. glomeratus as new to California. Her records are Sweetwater Moun- tains, above Star City, DeDecker 5677 (RSA!) and Sierra Nevada, “‘The Bluffs,’ 0.6 miles NNE of Mammoth Rock, Bagley 300] (personal herbarium of Mark Bagley, Bishop, California!). A check with Roxanne Bittman of the California Natural Diver- sity Data Base, in Sacramento, CA, revealed no other known specimens from California. The DeDecker and Bagley records are the basis for in- cluding P. glomeratus in the California Native Plant Society inventory of rare and endangered vas- cular plants of California (Skinner and Pavlik 1994). I find both of these specimens to be P. his- pidus. The misidentifications likely come from the lack of understanding of P. hispidus not from the true nature of P. glomeratus. Few California ref- erences provide the nutlet size for P. hispidus. Munz and Keck (1959) and Abrams (1951) list the size aS 1 mm while Messick (1993) lists the size as 1-1.5 mm. Cronquist (1959, 1984) twice has dealt with P. hispidus and his descriptions are es- sentially the same. The one slight difference is nut- let length, 1-2 mm in 1959 and 1-2 (2.5) in 1984. I have found that the nutlets of Plagiobothrys hispidus vary in the number that mature. At the north end of its range many plants have four ma- 160 turing nutlets while at the south end one or two is the norm. The number of maturing nutlets greatly influences their orientation, shape, and size. If four nutlets mature they are vertically oriented, less than 2 mm long, have a definite dorsal keel, and are unevenly tuberculate or rugose-tuberculate (see Fig. 1, illustration A). This is the nutlet type illustrated in Cronquist (1959, 1984) and represented by the type collection of P. hispidus. When one or two nutlets mature they are horizontally oriented, up to 2.4 mm long, flat-backed with a more obscure keel, the end farthest from the scar is greatly expanded, and the roughness is more evenly appressed and not as evident. It is this nutlet type that is represented by DeDecker 5677 and Bagley 3001, the basis of the reports of P. glomeratus from California, and by the type of Sonnea foliacea (see Fig. 1, ilustra- tion C). On the other hand P. glomeratus is ex- tremely uniform with larger, mottled, shiny nutlets, and is edaphically restricted (see Fig. 1, illustration D). I can see how one could be misled in trying to identify the California specimens. The nutlet size does not fit the descriptions in Munz and Keck (1959) or Messick (1993). In checking Intermoun- tain Flora (Cronquist 1984) the illustration of P. hispidus is that of the smaller four nutlet type. The broad fat-ended illustration of P. glomeratus then becomes the logical choice. I agree with Cronquist (1984) in placing Sonnea foliacea in synonymy with P. hispidus. The ex- tremes seem distinctive but all stages of interme- diacy occur. For instance, many collections from the Truckee, CA area contain plants with nutlets of both the Aispidus and foliacea type. Gray (1885) in describing P. glomeratus de- scribes its distributions as: “‘Western part of Ne- vada, between Carson and Virginia City, 1883 and 1884, Mrs. Layne-Curran.”’ There are two sheets in the Gray Herbarium that fit Gray’s protologue. One is labeled “‘Geiger Grade, Aug. 1883, Curran s.n.”’ and the other “‘between Carson and Virginia, [un- dated], Curran s.n.”’ Selection of the ‘“‘Carson to Virginia” sheet as a lectotype was effectively done by Cronquist (1984). MADRONO [Vol. 47 | The results of this study indicate that Plagiob- othrys glomeratus is a western Nevada endemic re- | stricted to areas of altered andesite between 4860 and 6650 ft in elevation. These altered andesite ar- eas have shallow azonal soils nearly totally lacking in nutrients and with an acidic pH (3.7—4.0) (Bill- | ings 1992). Soils are so nutrient poor that they are not able to support the ubiquitous sagebrush, Ar- temisia tridentata Nutt., or other shrubs in any number (Billings 1950, 1992). This lack of com- | petition from shrubs has allowed relic stands of Si- erran conifers to persist in isolated pockets. The altered andesite areas are orangish light-brown in surface color and are dotted with dark green coni- fers. As such they are a conspicuous feature on the | hills around Reno (Billings 1950, 1992). Although concentrated in the Reno area there are outliers of altered andesite as far northeast as the Pah Rah Range, east to Ramsey in the Virginia Range, and | south to the Sweetwater and White Mountains of | California and P. glomeratus may eventually be found at some of these sites (Billings 1992). Plagiobothrys glomeratus is known from the Virginia Range in Storey and Washoe Counties, Carson Range of the Sierra Nevada, foothills north | of Reno, and from nearby Peavine Mountain, all in— Washoe County (see Fig. 2). Its distribution nearly i | ij matches that of the only other known altered an- | desite endemic, Eriogonum robustum E. L. Greene (type also collected by Curran in 1884). Both occur less than six miles from California and eventually | may be found there. Searches in the Truckee River | canyon west of Reno, and near Markleeville south- | southwest of Gardnerville, have so far proved fruit- | less. Plagiobothrys hispidus occurs from south-central — Oregon south and east through the eastern Sierra | Nevada of California and Nevada to the Mammoth | area in Mono and adjacent Madera Counties (see Fig. 2). There are outliers on Steens Mountain, Har- ney County, Oregon, Skeedaddle Mountain, Lassen — County, California, Granite Range, Washoe County, | Nevada, Pine Nut Mountains, Douglas County, Ne- vada, and the Masonic Hills and Sweetwater Moun- tains in Mono County, California. KEY TO PLAGIOBOTHRYS GLOMERATUS AND HISPIDUS Nutlets smooth and shiny, mottled, 2.4—3.0 mm long, horizontally oriented, P. glomeratus Nutlets unevenly tuberculate to pavemented with the roughness always readily discernable, up to 2.4 mm long, horizontally or vertically Ofiented......2¢05442 0% eee 4 Specimens of Plagiobothrys glomeratus examined, all from Nevada STOREY CoO., Virginia Range, Six Mile Can- yon, 4.2 road miles E of highway 341, Tiehm 12544 (BRY, CAS, NY, OSC, RENO, RSA, UC, UTC); Virginia Range, 1.1 road miles SE of N junction of highways 341 and 342 on highway 341, Tiehm 12542 (ARIZ, BRY, CAS, GH, MONT, NY, OSC, RENO, RM, RSA, UC, UNLV, UT, UTC, WS): P. hispidus | WASHOE CoO., Dandini Blvd. N of Reno, Nachlin- | ger 1375 (NY), Tiehm & Kelley 12522 (CAS, NY, © OSC, RENO, UC, UTC); west slopes of Peavine | Mountain., Nachlinger & Billings 1374 (NY); hill east of Black Panther Mine, 3 miles N of Reno, | Billings 1296 (RENO); Geiger Grade, Jul 1884, — Curran s.n. (DS); Geiger Grade to Virginia City, | Eastwood 14809 (CAS); Geiger Grade, Aug 1883, 2000] TIEHM: PLAGIOBOTHRYS GLOMERATUS 161 Fic. 1. A-C are nutlets of Plagiobothrys hispidus, D is nutlets of Plagiobothrys glomeratus. A is drawn from Steward 6798, Deschutes Co., OR (NY); B from Sonne s.n., Truckee, Nevada Co., CA (NY); C from Tiehm 12244, Pine Nut Mountains, Douglas Co., NV (RENO); and D from Eastwood 14809, Geiger Grade to Virginia City, Storey Co., NV (CAS). 162 Fic. 2. MADRONO 100 Kilometers is designated by solid circles and the distribution of P. hispidus is designated by open circles. Curran s.n. (GH); Virginia Range, Geiger Grade, 2.8 road miles E of highway 395 on highway 341, Tiehm 12540 (ARIZ, ASU, B, BRY, CAS, COLO, CS, DAO, GH, ID, K, KSC, LE, MICH, MO, MONT, MONTU, NY, OKL, OS, OSC, RENO, RM, RSA, SI, TEX, UC, UNLV, UTC, WIS, WS, WTU); Virginia Range, foothills E of the S end of Hidden Valley County Park, Tiehm 12547 (CAS, NY, OSC, RENO, RM, RSA, UC, UTC, WTU); Sierra Nevada, Carson Range, ridge on N side of N fork of Evans Creek, Tiehm 12548 (BRY, CAS, MICH, MO, MONT, NY, OSC, RENO, RM, RSA, UC, UNLV, UTC); Sierra Nevada, Carson Range, ridge divide between Hunter and Alum Creeks, Tiehm 12593 (CAS, NY, OSC, RENO, UC): COUNTY UNKNOWN, Nevada between Carson & Virginia, [undated], Curran s.n. (GH lectotype). ACKNOWLEDGMENTS I am grateful to the following people for supplying me with advice and information: Mark Bagley, Janet Bair, Roxanne L. Bittman, Kenton L. Chambers, Ron Kelley, Sally Manning, James D. Morefield, Janet L. Nachlinger, Kathleen Nelson, and Margriet Wetherwax. The map was prepared by Brian McMenamy. Comments from two | anonymous reviewers added substantially to this work. I am also grateful to the curators of the following herbaria [Vol. 47. Map showing parts of Oregon, Idaho, Nevada, and California. The distribution of Plagiobothrys glomeratus for access to their collections, for loans, or both: CAS, | DS, GH, JEPS, NDG, NY, ORE, OSC, RENO, RSA, UC. Kathryn (Kay) Corbett aptly illustrated nutlets of P. glom- eratus and P. hispidus. LITERATURE CITED ABRAMS, L. 1951. Illustrated flora of the Pacific States. 3: 1-866. Stanford University Press, Stanford, CA. BILLINGS, W. D. 1950. Vegetation and plant growth as af- fected by chemically altered rocks in the western — Great Basin. Ecology 31:62-—74. SLOPES OF PEAVINE MOUNTAIN. MENTZELIA 6, PART 1:32-—39. CRONQUIST, A. 1959. Boraginaceae, pp. 175-244. in C. L. Hitchock, A. Cronquist, M. Ownbey, and J. W. . 1992. ISLANDS OF SIERRAN PLANTS ON THE ARID Thompson, Vascular Plants of the Pacific Northwest, | Part 4. Univ. Wash. Publ. Biol. 17(4):1—510. A. H. Holmgren, N. H. Holmgren, J. L. Reveal, and _ 1984. Boraginaceae. Pp 207-293 in A. Cronquist, — P. K. Holmgren, Intermountain Flora 4:1—573. New | York Botanical Garden, Bronx, N.Y. _ 2000] DEDECKER, M. 1990. Additions to our flora. California Native Plant Society Bristlecone Chapter Newsletter 9(6):3. Gray, A. 1885. Contributions to the botany of North America. Proc. Amer. Acad. Arts. 20:257—310. GREENE, E. L. 1887. Some west American Asperifoliae. Pittonia 1:8—23. . 1888. New or noteworthy species III. Pittonia 1: 215-225. JEPSON, W. L. 1925. A manual of the flowering plants of California. University of California Press, Berkeley. . 1943. A flora of California. 3, part H:129—464. Associated Students Store, University of California, Berkeley. JOHNSTON, I. M. 1923. Studies in the Boraginaceae. 4. A synopsis and redefinition of Plagiobothrys. Contr. Gray Herb. 68:57—80. TIEHM: PLAGIOBOTHRYS GLOMERATUS 163 Messick, T. C. 1993. Plagiobothrys, Pp. 386-390. in J. C. Hickman (ed.). The Jepson Manual. Higher plants of California. University of California Press, Berkeley. Munz, P. A. 1968. Supplement to A California Flora. Uni- versity of California Press, Berkeley. , and D. D. Keck. 1959. A California Flora. Uni- versity of California Press, Berkeley. SKINNER, M. W. AND B. M. PAVLIk. 1994. Inventory of rare and endangered vascular plants of California. California Native Plant Society Special Publication No. 1, 5th ed. CNPS, Sacramento. TIDESTROM, I. 1925. Flora of Utah and Nevada. Contr. U.S. Natl. Herb. 25:1—665. WILLIAMS, M. J., J. T. HOWELL, G. H. TRUE, JR., AND A. TIEHM. 1992. A catalogue of vascular plants on Peav- ine Mountain. Mentzelia 6, part 2:3—83. MApRONO, Vol. 47, No. 3, pp. 164-173, 2000 GENETIC VARIATION IN PINUS PONDEROSA, PURSHIA TRIDENTATA, AND FESTUCA IDAHOENSIS, COMMUNITY-DOMINANT PLANTS OF CALIFORNIA’S YELLOW PINE FOREST SAFIYA SAMMAN,!, BARBARA L. WILSON? AND VALERIE D. HIPKINS? 'USDA Forest Service, FHP, AB-2S, PO. Box 96090, Washington, D.C. 9009-6090 ?Department of Botany and Plant Pathology, 2064 Cordley Hall, Oregon State University, Corvallis, OR 97331 3USDA Forest Service-NFGEL, 2375 Fruitridge Road, Camino, CA 95709 ABSTRACT Genetic diversity of Pinus ponderosa Laws., Purshia tridentata (Pursh) DC., and Festuca idahoensis Elmer at Black’s Mountain Experimental Forest in northwest California was evaluated using isozymes. This tree, shrub, and grass are all common, outcrossing, long-lived perennials that dominate their respec- tive layers of the same plant community. Genetic analyses were provided for diploid Pinus ponderosa and Purshia tridentata. A phenotypic analysis of isozyme band patterns was provided for tetraploid F. idahoensis and, for comparison, previous reports of fescue isozyme variation were reanalyzed using this method. Pinus ponderosa, Purshia tridentata, and Festuca idahoensis were highly genetically variable, with 75% to 92% polymorphic loci. For all three species, more than 90% of the genetic variation occurred within, rather than among, populations. This study compares genetic diversity in plant species of three life forms, while holding constant habitat, breeding system, and community domi- nance. The three plant species chosen for study are Pinus ponderosa Laws, Purshia tridentata (Pursh) DC, and Festuca idahoensis Elmer. They represent three life forms, tree, shrub, and grass, respectively. All three are common, widespread, outcrossing, long-lived perennials. All dominate their respective layers in the plant community at the study site. They do have life history differences; F. idahoensis is insect pollinated while the other two species are wind pollinated, and F. idahoensis is tetraploid while the others are diploid. The three species af- fect one another in a complex web of competitive and commensal relationships (e.g., Baron et al. 1966; Busse et al. 1996; Hall et al. 1995; vander Wall and vander Wall 1992). The study site, Black’s Mountain Experimental Forest, was established in 1934 in the Lassen Na- tional Forest, Lassen County, California. More than 60 y of experimentation and careful record keeping make the 4050-ha forest a uniquely valuable re- source for investigating the effects of different tim- ber management practices on eastside pine type for- ests. In 1993 the Black’s Mountain Interdisciplinary Research Program was established to study the ef- fects of forest management on various ecosystem components including vertebrates, insects, soil or- ganisms, and vegetation. This study provides base- Corresponding and proofs author: Valerie D. Hipkins, USDA Forest Service—NFGEL, 2375 Fruitridge Road, Camino, CA 95709. Phone: 530 642-5067. Fax: 530 642- 5093. E-mail: vhipkins @fs.fed.us line data for a long-term study of effects of silvi- | cultural treatments on genetic biodiversity. Genetic | and species biodiversity are elements of a healthy | ecosystem. Little is known about the effects of for- est management on the genetics of forest plants, although some forestry practices can profoundly in- » fluence tree genetics (Adams et al. 1998). | Four plots similar in topography and vegetation | were chosen for this study (Table 1). Genetic vari- ation in the three selected species was sampled in 1994 and 1995. Subsequently, three silvicultural | treatments (a timber cutting regime, fire, and graz- ing) were applied to the plots (Table 1). Genetic © diversity will be resampled in five to twenty years, to detect any effects from the silvicultural practices | initiated in 1995. METHODS In 1993, plots were chosen for an intensive, mul- tidisciplinary study of the effects of management © practices on the entire ecosystem. Midpoints of the four plots were 2.2 to 4.0 km apart. These plots have been treated similarly in the past, and all were grazed lightly until 1996, when some were fenced to exclude cattle (Table 1). All four plots sampled in the genetics study consisted of dry forest domi- nated by P. ponderosa, and all plots were similar (Table 1). Half of each plot was subsequently burned. The presence of small, unburned, long-term control plots in some burned split plots is ignored in this analysis. Sample collection. Permanent markers were es- tablished on a 100-m grid within each plot. In each plot, fifty grid points were randomly selected as | 2000) SAMMAN ET AL.: GENETICS OF PONDEROSA PINE, BITTERBRUSH, AND IDAHO FESCUE TABLE 1. 165 CHARACTERISTICS AND MANAGEMENT PLANS OF THE FOUR BLACK’S MOUNTAIN PLOTS SAMPLED IN THIS STUDY. Pine trees are mostly Ponderosa Pine with some Jeffrey Pine; fir trees are White Fir, Abies concolor (Gordon and Glend.) Lindley. Perennial grass cover was predominantly Idaho Fescue cover. Structural vertical diversity is a man- agement practice imposed by thinning two of the plots to reduce structural diversity. Half of each plot was subsequently burned; burned and unburned halves are labeled “‘B” and ““N,” respectively, in other tables. Information on plot vegetation from W. W. Oliver (unpublished data). Perennial ee cee grasses Structural Elevation Area pie eee Idaho Fescue (foliar Vertical Plot (meters) (hectares) Pine (frequency) — cover) Diversity Grazing 38 523-540 136 440 y 59% 4% high no 39 526-543 120 535 0 717% 6% low no 4] 575-580 108 440 16 22% 2% high yes 43 570-576 109 364 170 40% 2% low yes collection sites. All three species were collected at each grid point, if all three were present. If one or more was missing, a replacement sample was col- lected at a different, ranodmly selected grid point. Pinus ponderosa: Cones were collected from the tree nearest each selected grid point (207 total in- dividuals). Samples were collected in September 1994. Purshia tridentata: At each selected grid point, two individuals were flagged and sampled (404 to- tal individuals). The shrub closest to the grid point was the first sampled individual, and the second was the closest shrub on the opposite side of the gridpoint, on a line drawn from the first shrub through the grid point. A leafy shoot was cut from each individual, wrapped in wet paper towels, placed in a plastic bag, and stored on ice in the field. Samples were collected in June and July OD. Festuca idahoensis: Near each flagged P. triden- tata shrub, a fescue individual was sampled (385 total individuals). (If fescue was sparse, the fescue might be as much as 30 meters away from the P. tridentata.) If no fescue grew within 30 meters of a sampling point, another point was randomly se- lected for fescue sampling. From each fescue, a small rooted plug with about 20 leaves was col- lected, wrapped in wet paper towels, placed in a plastic bag, and stored on ice in the field. Samples were collected in July 1995. Sample preparation. Samples were prepared us- ing NFGEL standard operating procedures (Anon- ymous 1995). For P. ponderosa, seeds were ger- minated and the megagametophytes were ground in a 0.2 M phosphate buffer, pH 7.5. The slurry was absorbed onto 3 mm wide wicks prepared from Whatman 3MM chromatography paper, and stored at —70°C. Purshia tridentata and F. idahoensis leaf tissue was ground in a Tris buffer pH 7.5 (Gottlieb 1981); liquid nitrogen was used to freeze P. triden- tata tissue before grinding. One hundred fifty mi- croliters of slurry per sample was transferred into each of two microtiter plate wells, and plates were frozen at —70°C. For electrophoresis, the slurry was thawed and absorbed onto wicks. Electrophoresis. Methods of electrophoresis are outlined in Anon. (1995), and follow the general methodology of Conkle et al. (1982) except that most enzyme stains are modified. The following en- zymes were examined: aconitase (ACO), catalase (CAT), diaphorase (DIA), florescent esterase (FEST), fructose-1,6-diphosphate dehydrogenase (FDP), glutamate-oxaloacetate transaminase (GOT), glucose-6-phosphate dehydrogenase (G6PDH), glycerate-2-dehydrogenase (GLYDH), isocitrate dehydrogenase (IDH), leucine aminopep- tidase (LAP), malate dehydrogenase (MDH), malic enzyme (ME), phosphoglucomutase (PGM), phos- phogluconate dehydrogenase (6PGD), phosphog- lucose isomerase (PGI), shikimic acid dehydroge- nase (SKD), triosephosphate isomerase (TPI), and uridine diphosphoglucose pyrophosphorylase (UGPP). The enzymes examined, and the buffer systems used to resolve them, varied according to species (Table 2). All enzymes were resolved on 11% starch gels. Enzyme stain recipes for enzymes follow Anonymous (1995) except that GOT was stained using the recipe from Wendel and Weeden (1989). Two people independently scored each gel. When they disagreed, a third person resolved the conflict. For quality control, 10% of the individuals were run and scored twice. Both P. ponderosa and the morphologically sim- ilar Jeffrey Pine P. jeffreyi Grev. and Balf (Jeffrey Pine) occur in the research plots (Oliver, MS). For a few of the pine samples, alleles of multiple en- zymes differed from those seen before in P. pon- derosa (NFGEL, unpublished data). Such samples were omitted from analysis on the assumption that they were P. jeffreyi. Pinus ponderosa and P. tridentata are diploid. Most studied isozymes are known to show Men- delian inheritance in P. ponderosa (Linhart et al. 1989; O’ Malley et al. 1979), but no such informa- tion is available for P. tridentata. Genetic interpre- tations were inferred directly from isozyme _ phe- 166 MADRONO [Vol. 47 | TABLE 2. ISOZYMES EXAMINED FOR PINUS PONDEROSA, PURSHIA TRIDENTATA, AND FESTUCA IDAHOENSIS. LB = a lithium borate electrode buffer (pH 8.3) used with a Tris citrate gel buffer (pH 8.3) (Conkle et al. 1982). SB = a sodium borate | electrode buffer (pH 8.0) used with a Tris citrate gel buffer (pH 8.8) (Conkle et al. 1982). MC6 = a morpholine citrate | electrode and gel buffer (pH 6.1) (Conkle et al. 1982). MC8 = a morpholine citrate electrode and gel buffer (pH 8.0) (Anon. 1985). na = number of alleles per locus. ne = effective number of alleles per locus (Kimura and Crow 1964). | np = number of patterns per locus. For enzymes abbreviations, see text. Numbers in isozyme abbreviations refer to — different regions on gels, interpreted as different loci. Buffer: LB na ne SB na PINUS PONDEROSA ACO 4 2.8 CAT | 16 enzymes ADH 2 1.9 GOT1 4 26 loci LAPI 3 1.1 GOT2 5 LAP2 4 | GOT3 6 ME 4 Ie G6PDH 4 PGI1 2 1.0 UGPPI1 3 PGI2 3 Let UGPP2 6 PGM1 4 1.9 PGM2 i) 2.3 PURSHIA TRIDENTATA na ne na FEST l 1.0 CAT 2, 13 enzymes LAP 3 2.6 GOT 4 16 loci PGI l 1.0 6PGD 4 PGI2 3 1.0 TPH 2 PGM 3 2.0 TP12 l UGPP R) FESTUCA IDAHOENSIS np np ME l CAT l 11 enzymes PGI 17 GLYDH 3 treated as 12 loci PGM 3 GOT 1 3 GOT2 4 6PGD Z TPI 4 UGPP 16 notypes, based on knowledge of the generally con- served enzyme substructure, compartmentalization, and isozyme number in higher plants (Gottlieb 1981, 1982; Weeden and Wendel 1990). Genotypes of P. ponderosa were inferred from the segregation patterns of 10 megagametophytes per tree. Festuca idahoensis has a chromosome number 2n = 28 and is thought to be autotetraploid (Dar- lington and Wylie 1955). Because of the compli- cated banding patterns observed, and because of lack of crossing studies to determine the inheritance of bands in this species, we were unable to identify specific alleles and loci for some enzymes, includ- ing highly variable MDH, PGI and UGPP. There- fore, a phenotypic instead of genotypic analysis was performed. Data analysis. In order to allow future compar- ison with genetic diversity after all silvicultural techniques (logging, grazing, and burning) are im- plemented, the data were analyzed in terms of eight plots, which represent the half plots subsequently burned (Table 3). For P. ponderosa and P. tridentata, results were analyzed using Popgene version 1.21 (Yeh et al. 1997). A locus was considered polymorphic if an ne MC6 na ne MC8 na ne 1.0 FEST fi 1.5 DIA 4 2.9 1.1 MDHI1 3 1.0 FDP 1 1.0 1.1 MDH2 3 1.9 IDH1 3 1.7 L.3 MDH3 4 1.1 IDH2 5 1.2 1.1 6PGD 1 > 220) 1.9 6PGD2 4 ie 3.0 ne na ne 1.0 DIA 2 1.0 2.0 FDP 3 1.0 bal IDH l 1.0 1.0 MDH1 2, 1.1 1.0 MDH2 3 1.0 1.9 np DIA 6 SKD 4 alternate allele occurred at least once. Statistics cal- | culated included unbiased genetic distances (Nei 1978), expected heterozygosity (Nei 1973), inferred © gene flow (Nm = 0.25(1/F,,)/F,,; Slatkin and Barton © 1989), and F statistics (F = (H,—H,)/H,; Hartl and | Clark 1989). The gene diversity statistics H,, (ex- | pected heterozygosity at the species level) and H,, ) (expected heteozygosity at the population level) were calculated (Hamrick and Godt 1990). | For F. idahoensis, phenotypic diversity measures — were calculated from both band presence/absence | and multi-band patterns. For presence/absence data, — phenotypic diversity was measured by a polymor- | phic index (PI), based on the frequency of occur- rence of each band. PI = the sum of f(1—f), where f = the frequency of a band in a population (Chung et al. 1991). For multi-band patterns, phenotypic diversity measures include: (1) the number of bands found in each plot, (2) percent of stains that yield more than one band pattern, (3) the average number of band patterns per stain in each plot, and (4) Shannon-Weaver Diversity Index values (Shannon and Weaver 1949). The Shannon-Weaver Diversity Index uses the frequency of each band pattern in each plot. The larger the Shannon-Weaver Index, 2000] SAMMAN ET AL.: GENETICS OF PONDEROSA PINE, BITTERBRUSH, AND IDAHO FESCUE 167 TABLE 3. SUMMARY OF GENETIC DIVERSITY MEASURES IN PINUS PONDEROSA AND PURSHIA TRIDENTATA BY SPECIES AND BY PLot. All = overall statistics for the study. Mean = average over the 8 plots. SD = standard deviation. N = mean number of individuals sampled per locus, per population. %P = percent of all loci that are polymorphic. A = average number of alleles per locus. A, = the average number of alleles per polymorphic locus. H, = observed frequency of heterozygotes. H. = frequency of heterozygotes expected under Hardy-Weinberg equilibrium conditions. F = the fixation index, = (H,-H,)/H. (* = statistically significant difference (p < 0.05)). S-W = Shannon-Weaver diversity index. B and N following plot numbers indicated the half plots that were subsequently burned or not burned. N JP A ae H, H, F S-W Pinus ponderosa (26 1oc1) All 207 92% BETS 3.96 0.264 0.272 0.030 0.510 Mean 26 85% 299 3.09 0.265 0.271 0.021 0.483 SD 2.4 4.0 0.11 0.10 0.021 0.010 0.082 0.019 38B 24 88% Deis 2.96 0.263 0.269 0.023 0.477 38N 26 81% DAS 3.14 0.251 0.264 0.046 0.470 39B ea 92% 3.04 ere 0.283 0.287 O012 OSL) 39N 29 85% 2.81 3.14 0.240 O277 Oois3 0.495 41B 26 85% 253 3.04 0.280 0.280 0.002 0.495 41N 23 81% 2AT 3.19 0.257 0.255 —0.010 0.455 43B 23 88% 23 2.96 0.301 0.264 —0.140 0.467 43N 29 81% 24 3.09 0.244 O272 0.104 0.489 Purshia tridentata (16 loci) All 404 75% 2.38 2.89 0.138 0.157 Oa it7* 0.263 Mean 510) 55% 1.81 2.47 1.138 0.156 0.113 0.254 SD 8.0 9.6 0.12 0.26 0.006 0.005 0.049 0.011 38B 50 50% 1.69 2.38 0.135 0.148 0.087 0.235 38N 50 56% 1.69 Die 0.134 0.152 0.120* 0247 39B 62 38% 1275 3.00 0.140 0.156 0.101 0.254 39N 38 50% Lei 2.50 0.138 8 alin 0.091 0.246 41B ys 69% 1.94 2.36 0.130 0.155 0.165 0.260 41N 50 62% 2.00 2.60 0.147 O.157 0.059 0.262 43B 42 52% 1.75 220) 0.132 0.166 0.204* 0.269 43N 60 62% 1.94 2.50 0.147 0.159 0.076 0.261 the more diverse the plot. The distribution of the total variation within and among plots was deter- mined by partitioning the total Shannon-Weaver Diversity Index. The phenotypic relationships among plots were determined by calculating Hed- rick’s phenotypic identities (Hedrick 1971) for mul- ti-band pattern data, and by cluster and principle coordinate analyses of Jaccard’s Similarity Index for band presence/absence data (Chung et al. 1991; Rolf 1987). The hypothesis that all the plots had equal ge- netic diversity by species was tested by ANOVA using Excel (Microsoft 1997). In one analysis, plots were treated as blocks and the burned and unburned half plots as samples. In a second analysis, burned and unburned half plots were treated as blocks and the plots were treated as samples. RESULTS Pinus ponderosa, P. tridentata, and F. idahoen- sis were all genetically variable (Tables 3 and 4). Both percent polymorphic loci/enzyme and _ the Shannon-Weaver diversity index indicate that P. tridentata is the least variable of the three species. In the two diploid species for which they could be calculated, observed heterozygosity nearly equaled that expected under Hardy-Weinberg con- ditions. Therefore, the fixation index (F) within each plot, calculated for P. ponderosa and P. tri- denta, was low (F < 0.113) (Table 3). Although heterozygosity could not be calculated for F. ida- hoensis, we observed the unequal band staining and complex band patterns characteristic of heterozy- gous tetraploids (Soltis and Riesberg 1986). F,, values indicated that over 98% of the isozyme variation in P. ponderosa and P. tridentata was within, rather than between populations, and in- ferred gene flow was high (Table 5). G,, a measure of interpopulation diversity analogous to F,, but cal- culated from the Shannon-Weaver diversity index, was somewhat higher than the corresponding val- ues of F,, but indicated that in all three species more than 92% of the variation was within, rather than between, populations (Table 5). Genetic similarities among plots were corre- spondingly high. Unbiased genetic identities (Nei 1978) between plots were greater than 0.99 for P. ponderosa and P. tridentata. Hedrick’s distances (Hedrick 1971), calculated using band patterns, re- vealed a similarily greater than 0.98 for F. ida- hoensis. For F. idahoensis, band presence/absence data did not reveal differences between plots. No bands were unique to any plot. Two clusters ap- peared in a graph based on Jaccard’s similarity in- dex (not shown), but all eight plots were repre- sented in both clusters. 168 TABLE 4. quently burned or not burned. N #Bands %P* All 385 53 83% Mean 48.1 45.1 66% SD 8.6 2D 8.3% 38B 61 46 715% 38N 47 45 67% 39B 59 45 67% 39N 36 42 58% 41B 49 43 50% 41N 47 49 75% 43B 39 48 67% 43N 47 43 67% With one exception, measures of genetic vari- ability in plots 38, 39, 41, and 43 did not differ significantly for any of the three species. The only exception was the Shannon-Weaver diversity index for P. tridentata. Its Shannon-Weaver diversity in- dex values differed significantly among plots (Table 6), and values for plot 43 are higher than those of plot 38. Before fire treatments were applied, mea- sures of genetic variability were the same in all the half plots, except that the percent polymorphic loci and observed heterozygosity for P. ponderosa were consistently higher on the plots that would later be burned (Table 6). The mean heterozygosity for the species (H..,) and the sampled populations (H,,) (Hamrick and Godt 1990) were 0.271 and 0.266, respectively, for P. ponderosa, and 0.156 and 0.120 for P. triden- tata. DISCUSSION Pinus ponderosa. Genetic variability found in this study is higher than previously reported in comparable studies of this species (that is, in stud- ies that involved at least twelve loci and including both polymorphic and monomorphic loci) (Allen- TABLE 5. INTER-POPULATION DIVERSITY STATISTICS IN PI- NUS PONDEROSA, PURSHIA TRIDENTATA AND FESTUCA IDAHOEN- SIS AT BLACK’S MOUNTAIN. G, = a measure of inter-pop- ulational genetic differentiation derived from the Shannon-Weaver diversity index and analogous to F,. F, = Wright’s fixation index (Weir 1990). S-W = mean Shannon-Weaver diversity index. Nm = inferred gene flow, = 0.25(1-F,,)/F,. Species G, FP. Nm Pinus ponderosa 0.0406 0.0188 13.0301 Purshia tridentata 0.0708 0.0148 16.6675 Festuca idahoensis 0.0637 MADRONO GENETIC VARIATION IN FESTUCA IDAHOENSIS AT BLACK’S MOUNTAIN, FOR 12 ENZYMES. All = overall statistics | for the study. Mean = average over the 8 plots. SD = standard deviation. N = mean sample size/stain. #Bands = total | number of bands (in all stains), in the population. %P* = percent of all presumed loci (regions on the gel that each probably represent a locus or set of homoeologos loci) that have more than one band pattern. A* = mean number of | band patterns/stain. PI = polymorphic index based on band presence/absence data (see text). S-W = Shannon-Weaver | diversity index is based on band patterns. B and N following plot numbers indicated the half plots that were subse- | [Vol. 47 | A* Pi S-W oy We 3.804 0.563 3.41 3.666 0.521 0.24 0.211 0.015 3.83 3.400 O35 i eo) 3,639 0.498 3.58 3.851 0.543 3.08 3.840 Oot re IY, 32936 O27 3.58 3.530 O25 eee: 4.013 0.516 3.42 3.520 0.510 | dorf et al. 1982; Niebling and Conkle 1989; | O’ Malley et al. 1979; Woods et al. 1983; Yow et , al. 1992). Expected heterozygosity (H,,) in this © study equals 0.272 while the average H,, of the oth- | er studies equals 0.171. However, this level of ge- netic variability is consistent with previous NFGEL | research on P. ponderosa (in previous NFGEL | studies the average H,, = 0.231; NFGEL, unpubl.). | The higher genetic variability reported by NFGEL | for this species probably results from quality con- trol measures and highly standardized procedures » that allow repeatable detection of small differences | in enzyme migration distances. Including rare al- | leles (those with frequencies lower than 0.05) in | analyses may also contribute to the high genetic variability reported. Both NFGEL studies and pre- | viously published work indicate that P. ponderosa | subsp. ponderosa is more genetically variable than | P. ponderosa subsp. scopulorum (average from | NFGEL studies: H,, (subsp. ponderosa) = 0.247, — H., (subsp. scopulorum) = 0.235; average from oth- | er studies: H,, (subsp. ponderosa) = 0.161, H,, | (subsp. scopulorum) = 0.151). | Genetic variation in P. ponderosa was distributed | within, rather than among, the plots. More than | 90% of the isozyme variation often occurs within, — rather than among, P. ponderosa populations | (Hamrick et al. 1989, Linhart et al. 1981). Because P. ponderosa pollen can travel long distances (Latta | et al. 1998), and calculated gene flow among plots in this study is high (Table 5), the short distances © (2 to 4 km) between plots may limit genetic differ- | entiation. However, genetic differentiation has been | detected previously over small distances in P. pon- | derosa (Beckman and Mitton 1984; Mitton et al. | 1977; Mitton et al. 1980). In general, the genetic | similarily among plots provides a uniform back- | ground against which the genetic effects, if any, of | timber management practices will be detectable. | However, the higher initial percent polymorphic | | t -2000] SAMMAN ET AL.: GENETICS OF PONDEROSA PINE, BITTERBRUSH, AND IDAHO FESCUE 169 TABLE 6. ANALYSIS OF VARIANCE OF MEASURES OF GENETIC VARIABILITY IN PINUS PONDEROSA, PURSHIA TRIDENTATA AND | FESTUCA IDAHOENSIS. The fixation index F = (H,-H,)/H,. The variance ratio F = s,’/s,’, where s? = the variance of the | sample. * = statistically significant difference (p < 0.05). Half plots (half later burned) Plots 38, 39, 41 and 43 Measure of genetic variance F (variance ratio) probability — F (variance ratio) — probability Pinus ponderosa N = number of individuals/plot 1.0576 0.3434 O23 0.5886 P = % polymorphic loci IZ752 O-0118* 0.5460 0.6767 A = alleles/locus 0.2242 0.6526 223750 0.2107 A, = alleles/polymorphic locus 2.4440 0.1690 0.9505 0.4964 H, = observed heterozygosity 14.627 0.0087* 0.1305 0.9370 H. = expected heterozygosity 1.2605 0.3045 1.0604 0.4588 F = fixation index 3.5968 0.1067 0.3551 0.7893 SW = Shannon Weaver diversity index 0.6778 0.4418 1.4773 0.3478 Purshia tridentata N = number of individuals/plot 0.1076 0.7540 0.0059 0.9993 P = % polymorphic loci 0.5627 0.4815 3.8774 0.1117 A = alleles/locus 0.4615 0.3223 6.0143 0.0579 A, = alleles/polymorphic locus 0.0238 0.8824 1.5350 0.3355 H, = observed heterozygosity Be2005 0.1198 0.1758 0.9076 H, = expected heterozygosity 0.1197 0.7412 4.9700 O.0777. F = fixation index 229523 0.1366 0.2069 0.8868 SW = Shannon Weaver diversity index 0.0031 0.9574 6.6854 0.0489* Festuca idahoensis N = number of individuals/plot 1.8075 0.2274 0.4115 0.7542 Number of bands/enzyme 0.1617 0.7015 0.2770 0.8401 %P* = % polymorphic enzyme 0.1005 0.7619 0.3481 0.7938 A* = number of patterns/enzyme 0.4906 0.5099 0.2970 0.8269 PI = polymorphic index 0.1811 0.6853 1.4484 0.3542 SW = Shannon Weaver diversity index 5.6936 0.0543 O:3223 0.8104 loci and observed heterozygosity in burned than in unburned half-plots (Table 6) would have been con- sidered a treatment effect if this baseline study had not been done. Purshia tridentata. Bitterbrush has been the sub- ject of intense scrutiny, focused on interspecific re- lationships and management practices (e.g., Basile 1967) rather than genetic diversity. Secondary com- pounds have interferred with isozyme resolution in previous studies (S. Brunsfeld, pers. comm.). Ge- netic variation was less evenly distributed among plots in P. tridentata than in the other two species in this study (Table 3), possibly because P. triden- tata is pollinated by insects, rather than wind. Al- though plots were homogeneous for most measures, there were significant differences in the fixation in- dex (Table 3) and Shannon-Weaver diversity index (Table 6). Festuca idahoensis. Because polyploidy compli- cates gel interpretation, isozymes have been unde- rutilized for describing genetic variation in the fine- leaved fescues, Festuca subgenus Festuca, to which F. idahoensis belongs, and summary statistics are rarely reported. One should keep certain trends in mind when comparing a phenotypic analysis, like that performed for fescues, with genotypic analyses of isozymes. The proportion of polymorphic en- zymes (%PE in Tables 7, 8) is higher than the per- cent polymorphic loci (%P*) because stains may reveal both polymorphic and monomorphic loci for the same enzyme. %PE is reported because the sta- tistic is unambiguous. Percent polymorphic putative loci (%P*) is theoretically equal to %P of a genetic analysis, although for polyploid plants different re- searchers may parse band patterns into loci in dif- ferent ways. The number of patterns reported per putative locus (A*) should be somewhat greater than the number of alleles per locus, because any two alleles can produce three patterns. For exam- ple, AA homozygotes, BB homozygotes, and AB heterozygotes are counted as three different pat- terns. The polymorphic index (PI) is a rough mea- sure of heterozygosity, valid only for comparing populations within a study. Measures of similarity and diversity based on patterns (Hedrick’s distance and the Shannon-Weaver diversity index, respec- tively) may overestimate differences, because AA homozygotes, BB homozygotes, and AB heterozy- gotes are considered three equally different pat- terns. On the other hand, measures of similarity and diversity based on bands (Jaccard’s Similarity Index and the polymorphic index, respectively) produce uneven results among monomeric enzymes (which have two bands when heterozygous), dimers (which have three), and tetramers (which have five) (Got- tlieb 1977). The Gs statistic derived from partition- ing the Shannon-Weaver diversity index, like hier- archical F-statistics (Wright 1978), indicates wheth- er a greater proportion of variation resides within 170 MADRONO [Vol. 47 | TABLE 7. OVERALL ISOZYME DIVERSITY STATISTICS FOR TAXA FESTUCA SUBGENUS FESTUCA, FROM STUDIES THAT WERE | NOT LIMITED TO POLYMORPHIC ALLELES. Overall statistics are total values for the entire study. Chr. = Chromosome number (2X = 14, etc.). + = chromosome numbers from Markgraf-Dannenberg (1980); other chromosome numbers | were provided in the source article. Pops. = number of populations. N = mean sample size per population. Enz = | number of enzymes stained. Loci = number of regions on the gel that each probably represent a locus or set of | homoelogous loci. Summary statistics calculated by authors: %PE = percent of polymorphic enzymes. %P = percent | of polymorphic putative loci. AE = number of patterns per enzyme. A* = number of patterns per putative locus. Taxon Chr. Pops. N Enz. Loci JPE Tor AE A* — Source (F. ovina complex) F. auriculata 2X 4 20 10 15 90% 67% l i F. baffensis 4X , 26 10 15 90% 67% l : F. brachyphylla 6X 5 23 10 15 80% 67% l J F. brevissima 2X Z, 23 10 15 20% 13% ] i F. idahoensis 4X 8 48 11 12 82% 83% Sal S) F. idahoensis 4x 8 41 11 18 91% 67% 2.607 4 F. minutilfora 2X ] 5) 9 14 67% 21% ] F. roemeri v. klamathensis 4x 4 3] 11 18 91% 67% 2.78 4 F. roemeri v. romeri 4x 8 27 11 18 91% 712% 4.00 4 F. valesiaca 2X 3 3] 8 20 75% 50% 2a dD 1.65 2 (F. rubra complex) F. amythestina 2XT 2 58 8 20 62% 35% 22D 1.40 2 F. asperifolia l 40 8 20) 75% 45% PHO) 1.65 2 F. diffusa 6X, 8XT l 40 8 20 50% 20% 1.65 1.35 2 F. heterophylla 4Xt 8 efi 8 20 65% 40% De 1.60 2 F. nigrescens 4X, 6Xt 6 36 8 20 50% 35% 2.50 1.65 2 F. nigrescens 6X 3 27 11 18 82% 67% D3 4 F. peristerea ] 40 8 20 715% 30% LF es) P35 2 F. picturata 3 34 8 20 715% 35% 22 1.50 2 F. rubra 2X-10Xf 8 20 8 20 75% 45% 2.88 1.75 2 F. rubra 6X 6 38 10 100% 3 Sources: 1 = Aiken et al. 1993 = Angelov and Edreva 1987 (omitting anodal esterase), Angelov et al. 1988, Angelov 1992, 1993 Livesey and Norrington-Davies 199] Wilson 1999 = this study ARWN II or among populations. Although a phenotypic anal- polyploid taxa for which genetic interpretations are ' ysis 1S imprecise compared to genetic analyses of not possible. isozyme data, it does quantify isozyme variation Available studies indicate that these fescues are. and therefore allows comparisons among those highly polymorphic (Tables 7, 8), with the excep- TABLE 8. MEAN ISOZYME DIVERSITY STATISTICS PER POPULATION FOR FESTUCA SUBGENUS FESTUCA, FROM STUDIES THAT | WERE NOT LIMITED TO POLYMORPHIC ALLELES. Chr. = Chromosomes (2X = 14, etc.). Chromosome numbers were | provided in the source article. Pops. = number of populations. N = mean sample size per population. Enz = number | of enzymes stained. Loci = number of regions on the gel that each probably represent a locus or set of homoeologos , loci. Summary statistics calculated by authors: %PE = percent of polymorphic enzymes. %P* = percent of polymorphic | putative loci. A* = number of patterns per putative locus. S-W = Shannon-Weaver diversity index. Taxon Chr. Pops. N Enz. Loci %PE Tors AE A* Source — (fF. ovina complex) F. auriculata 2X 4 20 10 Is) 60% 43% l F. brachyphylla 6X =) ZS 10 15 67% 40% | F. brevissima 2X Z 23 10 15 68% 7% ] F. idahoensis 4x 8 48 1] 12 68% 66% 3.41 3 F.. idahoensis 4X 2 56 11 18 68% 50% 2.03 0.2032 2 F. minutilfora 2 l ) 9 14 67% 21% ] F. roemeri v. klamathensis 4x 4 31 1] 18 80% 53% 1.89 0.2376 2 F. roemeri v. romeri 4x 6 38 a) 18 67% 50% 1.94 0.2710 2 = Aiken et al. 1993 = Wilson 1999 = this study WON = 2000] SAMMAN ET AL.: GENETICS OF PONDEROSA PINE, BITTERBRUSH, AND IDAHO FESCUE 171] ‘tion of the uncommon F. brevissima Jurtzev. Vari- | ability in Black’s Mountain F. idahoensis was high -but consistent with previously reported fescue ge- netic variation. Our reported number of band pat- terns per putative locus was particularly high but not anomalous. For example, we detected 17 PGI band patterns (Shannon-Weaver diversity index = 1.91), and the same number were found for PGI in -a survey of 6 European Red Fescue (F. rubra L.) populations (S-W = 1.75) (Livesey and Norring- ton-Davies 1991). That study reported an average of 10 band patterns per enzyme in the three highly variable enzymes investigated. Genetic distances and identities among fescue populations are rarely reported. Genetic identities between populations of diploid arctic fescues are 0.934 (F. brevissima) and an average of 0.857 (F. auriculata Drobov aggregate) (Aiken et al. 1993). In both intra- and interspecific comparisons, Hed- rick’s identities exceed 0.95 between populations in the tetraploid F. idahoensis and Roemer’s Fescue (F. roemeri (Pavlick) E. B. Alexeev) species pair in northern California (Wilson 1999). The high (greater than 0.98) Hedrick’s identities among Black’s Mountain fescue populations are therefore expected. Because Black’s Mountain populations of F. ida- hoensis were similar, any effects of burning, graz- ing, and logging regimes on genetic variability will be detectable. Such fine-scale adaptation to local habitat variables has been seen in grasses, particu- larly in self-pollinating species (Bradshaw 1959; Clary 1975; Clegg and Allard 1972; Hamrick and Allard 1972; Kahler et al. 1980; Lonn 1993; Nevo et al. 1983). However, coarse adaptation has also been observed in both self- and cross-pollinated in- troduced grasses (Rice and Mack 1991; Rapson and Wilson 1988). Summary. The gymnosperm tree P. ponderosa, the dicot shrub P. tridentata, and the monocot grass F. idahoensis are not phylogenetically close. How- ever, they have strikingly similar patterns of elec- trophoretically detected genetic variation. All are genetically variable, with well over 90% of the variation within, rather than among, populations in the area studied. All three are common, widespread, long-lived, perennial, outcrossing species that dom- inate late successional stages in their plant com- munity. Plants with this series of characteristics tend to be more genetically variable than average, and to have their genetic variation within, rather than among, populations (Hamrick and Godt 1990). The high level of genetic variability detected in the three plants is consistent with observed trends in genetic variability (Hamrick and Godt 1990). The markedly lower genetic variation in P. tri- dentata as compared to the other two species is consistent with the tendency for insect pollinated plants and dicots to have much less isozyme vari- ation than wind pollinated plants and gymnosperms or monocots (Hamrick and Godt 1990). The value of the genetic diversity statistics H.,, and H.,, for P. ponderosa are more than one stan- dard deviation higher than comparable mean statis- tics reported (Hamrick and Godt 1990; Hamrick et al. 1992). Those for P. tridentata vary from that far above to somewhat below average, depending upon the comparison; they are low for woody angio- sperms (Hamrick et al. 1992), but most woody an- giosperms studied are wind-pollinated trees, and P. tridentata is an insect-pollinated shrub. For all three species, the percentage of genetic variability among populations is lower than the mean reported in comparable plants (Hamrick and Godt 1990), but few other studies compared populations growing in such close proximity in the same habitat. ACKNOWLEDGMENTS We thank Chris Bailey and Dawn Cambell for collect- ing P. ponderosa cones. We thank Laura French and Des- sa Welty for collecting P. tridentata and F. idahoensis, and for providing insights into collection methods. We thank Suellen Carroll, Patricia Guge, and Randy Meyer for electrophoresis and gel scoring. We thank Dr. Steven J. Brunsfeld and Manuela Huso for their input, Dr. Georgi Angelov for providing unpublished information about fes- cue isozymes, Maria Ivantchenko for translation from the Bulgarian, Tanya Bray for text editing, and anonymous reviewers for critical comments on the manuscript. This article is contribution no. 18 of the Blacks Mountain Eco- logical Research Project, Pacific Southwest Research Sta- tion, USDA Forest Service. LITERATURE CITED ADAMS, W. T., J. Zuo, J. Y. SHIMIZU AND J. C. TAPPEINER. 1998. Impact of alternative regeneration methods on genetic diversity in coastal Douglas-Fir. Forest Sci- ence 44:390—396. AIKEN, S. G., L. L. CONSAUL, J. I. DAVIS AND P. S. MANOs. 1993. Systematic inferences from variation in 1so- zyme profiles of arctic and alpine cespitose Festuca (Poaceae). American Journal of Botany 80:76—-82. ALLENDORF, EF W., K. L. KNUDSEN AND G. M. BLAKE. 1982. Frequencies of null alleles at enzyme loci in natural populations of ponderosa and red pine. Ge- netics 100:497—504. ANGELOV, G. B. 1992. Biosystematic study of species from the genus Festuca L. IV. Electrophoretic anal- ysis of the enzymes peroxidase, acid phosphatase, and catalase. Fitologia 43:58—68. ANGELOV, G. B. 1993. A biosystematic study of species of genus Festuca L. V. An electrophoretic analysis of the enzymes glucose-6-phosphate dehydrogenase, 6- phosphogluconate dehydrogenase, isocitrate dehydro- genase, and superoxide dismutase. Fitologia 44:47— 3 ANGELOV, G. B. AND A. EDREVA. 1987. Electrophoretic study of species of genus Festuca L. I. Electropho- retic patterns of esterase. Trudove IV Natsionalna Konferentsiy Botanika, Bulgarska Akademiya na Naukite [Proceedings of the Fourth National Confer- ence of Botany; Bulgarian Academy of Science], So- fia, Bulgaria. 172 MADRONO ANGELOV, G. B., S. I. KOJUHAROV AND A. EDREVA. 1988. Biosystematic study of the species of genus Festuca L. II. Electrophoretic analysis of the seed proteins and the enzymes glutamate-oxaloacetate-transaminase and glutamate dehydrogenase. Fitologia 35:13—23. ANONYMOUS. 1995. National Forest Genetic Electropho- resis Laboratory Standard Operating Procedures. NFGEL, USDA Forest Service, Camino, CA. BARON, E J., E. C. Norp, A. B. EVANKO AND W. J. MAKEL. 1966. Seeding conifers and buffer crops to reduce deer depredation. U.S. Forest Service Res. Note PSW- 100. BASILE, J. V. 1967. An annotated bibliography of P. tri- denta (Purshia tridentata (Pursh) DC.). Intermoun- tain Forest and Range Experiment Station. USDA Forest Service Research Paper INT-44. Ogden, UT. BECKMAN, J. S. AND J. B. Mitron. 1984. Peroxidase allo- zyme differentiation among successional stands of P. ponderosa. American Midland Naturalist 112:43—49. BRADSHAW, A. D. 1959. Population differentiation in Agrostis tenuis Sibth: morphological differentiation. New Phytologist 58:208—227. Busse, M. D., P. H. COCHRAN AND J. W. BARRETT. 1996. Changes in P. ponderosa site productivity following removal of understory vegetation. Soil Science So- ciety of America Journal 60:1614—1621. CHUNG, M. G., J. L. HAmrick, S. B. JONES AND G. S. DERDA. 1991. Isozyme variation within and among populations of Hosta (Liliaceae) in Korea. Systematic Botany 16:667—84. CLARY, W. P. 1975. Ecotypic adaptation in Sitanion hys- trix. Ecology 56:1407-1415. CLEGG, M. T. AND R. W. ALLARD. 1972. Patterns of genetic differentiation in the slender wild oat species Avena barbata. Proceedings of the National Academy of Sciences USA 69:1820-1824. COoNKLE, M. T., P. D. HopGskiss, L. B. NUNNALLY AND S. C. HUNTER. 1982. Starch Gel Electrophoresis of Co- nifer Seeds: A Laboratory Manual. Gen. Tech. Rep. PSW-64. Berkeley, CA: Pacific Southwest Forest and Range Experiment Station, Forest Service, U.S. De- partment of Agriculture. DARLINGTON, C. D. AND A. P. WyLIE. 1955. Chromosome Atlas of Flowering Plants. University Press, Aber- deen, Great Britain. Gott Lies, L. D. 1977. Electrophoretic evidence and plant systematics. Annals of the Missouri Botanic Garden 64:161—180. Gott Lies, L. D. 1981. Gene number in species of Aster- aceae that have different chromosome numbers. Pro- ceedings of the National Academy of Sciences USA 78:3726—3729. GoTTLieB, L. D. 1982. Conservation and duplication of isozymes in plants. Science 216:373—380. HALL, D. B., S. B. MONSEN AND V. J. ANDERSON. 1995. Influence of herbaceous understory on native shrub recruitment. Rangelands 17:12—14. HAmRrICk, J. L. AND R. W. ALLARD. 1972. Microgeograph- ical variation in allozyme frequencies in Avena bar- bata. Proceedings of the National Academy Sciences USA 69:2100—2104. HAMRICK, J. L., H. M. BLANTON AND K. J. HAMRICK. 1989. Genetic structure of geographically marginal popu- lations of P. ponderosa. American Journal of Botany 76:1559-1568. HAmrIck, J. L. AND M. J. W. Gopr. 1990. Allozyme di- versity in plant species, Pp. 43-63. in Brown, A. H. D., M. T. Clegg, A. L. Kahler and B. S. Weir (eds.), [Vol. 47. Plant population genetics, breeding, and genetic re- | sources. Sinauer Associates, Inc., Sunderland, MA. HAMRICK, J. L., M. J. W. Gopt AND S. L. SHERMAN-BROy- | LES. 1992. Factors influencing levels of genetic di- |, versity in woody plant species, Pp. 94-124. in W. T. | Adams, S. H. Strauss, D. L. Copes and A. R. Griffin | (eds.), Population genetics of forest trees. Kluwer Ac- | ademic Publishers, Boston. HARTL, D. L. AND A. G. CLARK. 1989. Principles of Pop- ulation Genetics, 2nd ed. Sinauer Associates, Sun- | derland, MA. | HEDRICK, P. W. 1971. A new approach to measuring ge- | netic similarity. Evolution 25:276—280. KAHLER, A. L., R. W. ALLARD, M. KRZAKOWA, C. FE WEHRHAHN AND E. Nevo. 1980. Associations between | isozyme phenotypes and environment in the slender © wild oat (Avena barbata) in Israel. Theoretical Ap- | plied Genetics 56:31—47. Kimura, M. AND J. E Crow. 1964. The number of alleles | that can be maintained in a finite population. Genetics — 49:725-—738. LaTTA, R. G., Y. B. LINHART, D. FLECK AND M. ELLIOT. | 1998. Direct and indirect estimates of seed versus © pollen movement within a population of P. ponder- | osa. Evolution 52:61—67. LINHART, Y. B., M. C. GRANT AND P. MONTAZER. 1989. | Experimental studies in P. ponderosa. I. Relationship | between variation in proteins and morphology. Amer- | ican Journal of Botany 76:1024—1032. LINHART, Y. B., J. B. MiITToN, K. B. STURGEON AND M. L. | DAvis. 1981. Genetic variation in space and time in | a population of P. ponderosa. Heredity 46:407—426. LIVESEY, V. AND J. NORRINGTON-DAVIES. 1991. Isozyme polymorphism in Festuca rubra L. Euphytica 55:72— ek LONN, M. 1993. Genetic structure and allozyme-micro- habitat associations in Bromus hordeaceus. Oikos 68: | 99-106. MARKGRAF-DANNENBERG, I. 1980. Festuca, Pp. 125-153. in Tugin, T. G., V. H. Heywood, N. A. Burges, D. M. Moore, D. H. Valentine, S. M. Walters and D. A. Webb (eds.), Flora Europea, Vol. 5. Monocotyledons. Cambridge University Press, Cambridge. MICROSOFT CORPORATION. 1997. Microsoft Excel SR-1. Redmond, Microsoft Corporation, Washington. Mitton, J. B., Y. B. LINHART, J. L. HAMRICK, AND J. S. BECKMAN. 1977. Observation on the genetic structure and mating system of P. ponderosa in the Colorado Front Range. Theoretical Applied Genetics 51:5—13. Mitton, J. B., K. B. STURGEON AND M. L. Davis. 1980. Genetic differentiation in P. ponderosa along a steep elevational transect. Silvae Genetica 29:100—103. Nel, M. 1973. Analysis of gene diversity in subdivided Populations. Proceedings of the National Academy of Sciences USA 70:3321-3323. Ne!, M. 1978. Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics 89:583-—590. Nevo, E., A. BEILES AND N. STORCH. 1983. Microgeo- graphic edaphic differentiation in hordein polymor- phisms of wild barley. Theoretical Applied Genetics 64:123-132. NIEBLING, C. R. AND M. T. CONKLE. 1989. Diversity of Washoe pine and comparisons with allozymes of P. ponderosa races. Canadian Journal of Forest Re- sources 20:298—308. OLIVER, W. W. MS. Black’s Mountain Interdisciplinary Research: Vegetation. 2000] O'MALLEY, D. M., E W. ALLENDORF AND G. M. BLAKE. 1979. Inheritance of isozyme variation and hetero- zygosity in Pinus ponderosa. Biochemical Genetics 17:233-250. _ Rapson, G. L. AND J. B. WILSON. 1988. Non-adaptation in Agrostis capillaris L. (Poaceae). Functional Ecology 2:479—490. | ROHLF, E J. 1987. NTSYS-pce: numerical taxonomy and multivariate analysis system for the IBM PC micro- computer (and compatibles). Applied Biostatistics, Setauket, New York. Rice, K. J. AND R. N. Mack. 1991. Ecological genetics of Bromus tectorum Ul: The demography of reciprocally sown populations. Oecologia 88:91—101. SHANNON, C. E. AND W. WEAVER. 1949. The mathematical theory of communication. University of Illinois Press, Urbana, IL. SLATKIN, M. AND N. H. BARTON. 1989. A comparison of three indirect methods for estimating average levels of gene flow. Evolution 43:1349-1368. SoLTis, D. E. AND L. H. RIESEBERG. 1986. Autopolyploidy in Tolmiea menziesii (Saxifragaceae): genetic insights from enzyme electrophoresis. American Journal of Botany 73:310-318. VANDER WALL, S. B. AND S. B. VANDER WALL. 1992. Es- tablishment of Jeffrey pine seedlings from animal caches. Western Journal of Applied Forestry 7:14— 20. WEEDEN, N. F AND J. E WENDEL. 1990. Genetics of plant SAMMAN ET AL.: GENETICS OF PONDEROSA PINE, BITTERBRUSH, AND IDAHO FESCUE E73 isozymes, Pp. 46—72. in D. E. Soltis and P. S. Soltis. (eds.), Isozymes in plant biology. Dioscorides Press, Portland, OR. WENDEL, J. EK AND N. EF WEEDEN. 1989. Visualization and interpretation of plant isozymes, Pp. 5—45 in D. E. Soltis and P. S. Soltis (eds.), Isozymes in Plant Bi- ology. Dioscorides Press, Portland, OR. Weir, B. S. 1990. Genetic Data Analysis. Sunderland, MA, Sinauer. WILSON, B. L. 1999. Fescue Taxonomy in the Pacific Coast States. Unpubl. Ph.D. Thesis, Department of Botany and Plant Pathology, Oregon State University, Corvallis, OR. Woops, J. H., G. M. BLAKE AND F W. ALLENDORF. 1983. Amount and distribution of isozyme variation in P. ponderosa from eastern Montana. Silvae Genetica 32: 151-156. WRIGHT, S. 1978. Evolution and the Genetics of Popula- tions, Vol. 4. Variability within and among natural populations. University of Chicago Press, Chicago. YEH, FE C., R.-C. YANG, T. BOYLE, Z’H. YE AND J. X. Mao. 1997. Popgene, the userfiendly shareware for population genetic analysis. Version 1.20. Molecular Biology and Biotechnology Centre, University of Al- berta, Edmonton, Alberta. Yow, T. H., M. R. WAGNER D. E. WOMMACK AND G. A. TUSKAN. 1992. Influence of selection for volume growth on the genetic variability of southwestern P. ponderosa. Silvae Genetica 41:326—333. MADRONO, Vol. 47, No. 3, pp. 174-188, 2000 THE EFFECT OF CLIMATIC VARIABILITY ON GROWTH, REPRODUCTION, AND POPULATION VIABILITY OF A SENSITIVE SALT MARSH PLANT SPECIES, LASTHENIA GLABRATA SUBSP. COULTERIT (ASTERACEAE) LORRAINE S. PARSONS! AND ADAM W. WHELCHEL2? 'Point Reyes National Seashore, Point Reyes Station, CA 94956 URS Corp., Box 290, 201 Willowbrook Blvd., Wayne, NJ 07474 ABSTRACT As with many other sensitive species in California, the range of Lasthenia glabrata Lindley subsp. coulteri (A. Gray) Ornd. (Asteraceae; Coulter’s goldfields) has been dramatically reduced in recent de- cades by urbanization. Many populations are small, isolated, and seemingly unstable. In this study, we conducted an autecological assessment of a small L. glabrata subsp. coulteri population at San Diego County’s San Dieguito Lagoon, using a large population at San Elijo Lagoon for comparison. The large population was not only more stable based on trends in seed production, but generally produced larger plants and more flowers and capitulescences than the small population. However, this relationship appears to be temporally variable and influenced significantly by climatic conditions, particularly rainfall totals and distribution. In a year with above-average rainfall, vegetative and reproductive yield of plants in the small population (San Dieguito Lagoon) matched or even exceeded that of plants in the large one (San Elijo Lagoon), which was subjected to prolonged inundation following heavy rains and a back-up of run- off and creek flows behind a dike system. Rainfall is linked not only to soil moisture, but to nutrient influx and cycling, variables that were strongly associated with group (marsh/monitoring year) separation and prediction in statistical analyses. When resources are sufficient, reproductive yield appears to be driven by other factors, the most probable of which is pollen supply. The relationship between rainfall and plant yield could prove integral to predicting long-term viability of L. glabrata subsp. coulteri pop- ulations, as above-average rainfall years are often sporadic and interspersed between lengthy periods of average or below-average rainfall in southern California. Conservation and enhancement of the remaining coastal salt marsh L. glabrata subsp. coulteri populations could perhaps be furthered by factoring this relationship into conservation and restoration projects and hydrologic regimes designed for managed wetland systems. Small populations of rare plant species face many genetic, demographic, and ecological chal- lenges. Small populations can suffer from reduced “fitness,” often undergoing one or more genetic bottleneck events that reduce genetic variation (Nei et al. 1975; Hamrick et al. 1979; Hedrick 1983; Ledig 1986; Barrett and Kohn 1991 and oth- ers). Fewer numbers also create a greater chance for normally outbreeding species to inbreed and become less fit through concentration of deleteri- ous alleles (Charlesworth and Charlesworth 1987 and others). Opportunities for gene flow between populations—and the potential for infusion of new alleles—may be minimal due to the dwindling number of populations and the distance between them. Small plant populations are intrinsically less appealing to pollinators (Powell and Powell 1987; Morgan 1999), which can further reduce fecundity and the potential for even limited outbreeding be- tween more distant individuals. Reduced genetic variation can also increase small populations’ sus- ceptibility to herbivory, pathogens, and stochastic factors such as floods and environmental and de- mographic variability (Shaffer 1981). In addition, population viability can be continually jeopardized by human-related disturbances or changes in wa- tershed or ecosystem conditions—sometimes the very changes believed to have made the species rare in the place. Even efforts to better manage, enhance, or restore systems in which rare plants occur can pose a threat if these activities do not balance their ecological requirements with those of other target plant and wildlife species and the ecosystem as a whole. Determining whether small populations are suc- cumbing to these challenges is not an easy task. Annual censuses are not only difficult, but often misleading unless conducted over several decades due to cryptic life history stages (1.e., seed banks) and normal fluctuations in population size that may have little impact on population stability or viability (Davy and Jefferies 1981; Schemske et al. 1994; Pavlik 1994 and others). A life table or population viability analysis (PVA) is often considered the op- timal approach for assessing population stability (Schemske et al. 1994; Pavlik 1994; Menges 1986). However, the probability of a long-lived seed bank immeasurably complicates performance of a life ta- ble or PVA for plant species (Pavlik 1994), despite arguments that there are ways to circumvent cal- culation of this unknown (Menges 1986). Some al- ternative approaches to assessing population stabil- ity involve performance of non-integrated demo- | graphic trend assessment, which focuses on overall ‘trajectories in survivorship, seed production, den- ‘sity of viable seed, and frequency of establishment \(Pavlik 1994). Morphological attributes associated with productivity or yield such as plant size/bio- /mass and flower number may be incorporated, as well (Menges 1986; Menges and Gordon 1996). These analyses are often improved through using either weedy congeners or large, more stable pop- ‘ulations of the same species for comparison (Pavlik | 1994). Consistent with the major role that extrinsic dis- ‘turbances or changes can play in population viabil- | ity, many studies on rare species include an eco- logical, as well as demographic, component '(Schemske et al. 1994). Some have criticized re- searchers for emphasizing autecology over demog- raphy in sensitive plant research, characterizing ecological research as premature in the absence of demographic information relevant to population vi- tal rates (Schemske et al. 1994). However, manag- ers of reserves and enhancement/restoration pro- jects often seek ecological information that might help them better manage reserves or design projects (Pavlik 1994). Moreover, ecological data can great- ly complement demographic assessments, particu- larly when the information is integrated to allow for identification of ecological constraints on key life history stages and variables associated with productivity (e.g., plant size, flower number) (Schemske et al. 1994; Pavlik 1994; Menges and Gordon 1986). San Dieguito Lagoon in San Diego County sup- ports a small population of a rare plant, Lasthenia glabrata Lindley subsp. coulteri (A. Gray) Ornd. (Asteraceae: Coulter’s goldfields). For a sensitive species, L. glabrata subsp. coulteri has a remark- ably diverse distribution. This annual is found in alkali playas in southern California’s arid inland ar- eas and salt marshes and vernal pools in the re- gion’s more moderate coastal areas (NDDB 1998; Skinner and Pavlik 1994; Hickman 1993). This diverse distribution has not spared the spe- cies from the threat of extirpation, however. All of these habitats have been negatively impacted to some extent by California’s extensive urbanization over the past 50 y (Skinner and Pavlik 1994). More than 90 percent of California’s wetland habitats, in- cluding marshes, vernal pools, and alkali playas, have been destroyed by commercial and residential development, and despite regulatory efforts at ef- fecting a ‘“‘no net loss”’ policy, this downward trend in wetland habitat acreage appears to be continuing. The wetland habitats that remain are often frag- mented, highly disturbed, and heavily impacted by outside influences such as nutrient and contaminant influx associated with watershed development. The toll these habitat losses and impacts has taken is apparent from the constriction of the species’ his- PARSONS AND WHELCHEL: CLIMATIC VARIABILITY AND RARE PLANT K75 toric range. In recent decades, its once extensive distribution throughout southern California has been reduced to a few marshes and vernal pools in San Diego, Ventura, and Santa Barbara counties and alkali playas in Riverside County (NDDB 1998). This precipitous decline in distribution prompted listing of L. glabrata subsp. coulteri as a species of concern (formerly C2) by the U.S. Fish and Wild- life Service and a species of limited distribution (List 1B) by the California Native Plant Society (CNPS). While its cousin, Lasthenia glabrata Lind- ley subsp. glabrata, is relatively common and has even a larger range than subsp. coulteri, other Las- thenia species that occur in vernal pool habitats such as Lasthenia burkei (E. Greene) E. Greene (Burke’s goldfields) and Lasthenia conjugens E. Greene (Contra Costa goldfields) are faced with similar threats in terms of potential extirpation (Skinner and Pavlik 1994). Within its historic coast- al range, L. glabrata subsp. coulteri often grows in high elevation areas of salt marshes—or the “‘high marsh’’—alongside another sensitive species, Cor- dylanthus maritimus Benth. subsp. maritimus (salt marsh bird’s beak), a state- and federally listed en- dangered species. San Dieguito Lagoon is one of six San Diego County coastal marsh systems that supports histor- ical and/or possibly reintroduced populations of L. glabrata subsp. coulteri. The species was once present at 10 San Diego County marshes (NDDB 1998), but probably in low abundance, as an early ecological study characterized it as only an “‘infre- quent”’ inhabitant (Purer 1942). Of the six remain- ing occurrences, three are believed to be small and relatively unstable populations, including the one at San Dieguito Lagoon. Over the past few decades, L. glabrata subsp. coulteri numbers at San Dieguito Lagoon have ranged from as low as zero in 1980 (Sea Science Services and Pacific Southwest Bio- logical Services, Inc. 1980) and six in 2000 (An- drea Thorpe personal communication) to as high as 1000 individuals during the mid- and late-1990s (MEC Analytical, Inc. 1993; L. Parsons and A. Whelchel, personal observation). Neighboring marshes such as San Elijo Lagoon, as well as re- portedly Los Penasquistos Lagoon, support annual populations consistently numbering as many as 5000 to 10,000 individuals (L. Parsons and A. Whelchel personal observation). San Dieguito Lagoon is also one of seven coastal marshes in San Diego County, CA, for which res- toration and/or enhancement activities have been or are being conducted or are proposed. As with other San Diego County marshes, this coastal lagoon has been subject to a number of historic watershed changes and disturbances, including damming of its river, agricultural and residential development, dik- ing, and intermittent mouth closures that impound water and create hypoxic conditions. With restora- 176 MADRONO tion and enhancement plans for San Dieguito La- goon currently being developed, there appeared to be a strong and immediate need for information on demographic and ecological aspects of this small and possibly unstable population and its potential for conservation and even future enhancement. Few studies have actually assessed demography or aut- ecological relationships of this or other Lasthenia species. The research that exists deals primarily with upland species (Lasthenia californica Lindley; Rajakaruna and Bohm 1999; Vivrette 1999) or fo- cuses on salinity tolerance (L. glabrata subsp. coul- teri; Kingsbury et al. 1976; Callaway et al. 1990; Callaway and Sabraw 1994). In 1996, a study was implemented to assess de- mographic and ecological characteristics of the L. glabrata subsp. coulteri population at San Dieguito Lagoon. For purposes of performing a comparative assessment, we broadened the scope of our study to include another L. glabrata subsp. coulteri pop- ulation that appeared to be larger and more stable in terms of plant numbers, the population at San Elijo Lagoon, located directly north of San Diegui- to Lagoon. Through this study, we hoped to gain insight into differences in survivorship, reproduc- tive potential (plant size, flower number) and suc- cess (seed set and seed production) between pop- ulations. When possible, we also attempted to track trends in demographic results for purposes of as- sessing population stability. By comparing autecol- ogy of a small and unstable population with that of a large and stable one, we hoped to increase our understanding of the biotic and abiotic factors that might be influencing variations in plant yield and population viability. We believe that this informa- tion could prove invaluable to resource managers charged with planning or implementing complex restoration or conservation projects, particularly projects with multiple species and ecosystem ob- jectives. Study area. San Dieguito Lagoon is located in Del Mar, CA, approximately 25 km north of San Diego (Fig. 1). The lagoon’s principal source of freshwater is the San Dieguito River, which has been dammed to create Lake Hodges, a reservoir in the inland area of the San Diego County. Water- shed of the lagoon totals 897 km’, 785.2 km? of which 1s behind dams (California Wetlands Infor- mation System 1996). During the summer and fall, the lagoon mouth sometimes closes for several weeks to a month until it re-opens either naturally or manually (1.e., using bulldozers). Vegetation communities occurring within the lagoon include salt and brackish marshes and, at its eastern end, freshwater marsh and limited riparian habitats. Ap- proximately 104 of the 240 hectares of wetlands once present at San Dieguito Lagoon still remain (California Wetlands Information System 1996). In addition to damming, the watershed of the lagoon has been altered considerably by development of [Vol. 47] a % \ Caraiff-% k | \py-the- 4 To Los Angeles | \ Sea 156 Km San Elljo ‘ \ ‘ d, | Lagoon | 4 \ | 0 Vp ] Kilometers eo : San Dieguito | et Lagoon : f To San Diego © DelMar Ve ssa : a | Fic. 1. Location of Lasthenia glabrata subsp. coulteri. (Coulter’s goldfields) study areas in San Diego County, CA. the adjacent floodplain and uplands for agriculture, | commercial and residential structures, and a race- | track/fairgrounds. The population of L. glabrata subsp. coulteri at this lagoon is located in a muted tidal basin in the southern portion of the lagoon, approximately 2 km from the mouth. San Elijo Lagoon is located in the Cardiff, CA, | approximately 27.5 km north of San Diego and di- | rectly north of San Dieguito Lagoon (Fig. 1). Sev- | eral creeks flow into the lagoon, one of which is dammed by Lake Wohlford (Escondido Creek). The | watershed for San Elijo Lagoon is 200.2 km? (Cal- | ifornia Wetlands Information System 1996). The : mouth of San Elijo Lagoon also closes during the | summer and fall, typically with more frequency and | longer duration than San Dieguito Lagoon. Current | wetland acreage at San Elijo Lagoon totals 230.4. ha and is comprised of salt marsh, brackish marsh, | freshwater marsh, and riparian habitat (California | Wetlands Information System 1996). The water- | shed of this lagoon has been altered considerably | by development of the adjacent floodplain and up- : lands for farming and commercial and residential | structures. The lagoon has also been separated into | a series of hydrologic ‘‘cells’? by construction of | a series of dikes and levees between 1880 and. 1940. By closing flood gates at one of the eastern | dikes, the area east of Interstate 5 is partially flooded from November through March for water- | fowl enhancement (Susan Welker personal com- | munication). The population of L. glabrata subsp. | 2000] PARSONS AND WHELCHEL: CLIMATIC VARIABILITY AND RARE PLANT 177 TABLE 1. MEAN ANNUAL TEMPERATURE AND RAINFALL AND NUMBER OF DAYS THAT LAGOON MOUTH WAS OPEN DURING |THE STUDY PERIOD (1996-1999). San Diego NWS-Lindbergh (NOAA; CDWR) San Dieguito Lagoon San Elijo Lagoon Temperature Departure No. of days (Jan—Dec) Departure Rainfall from mean No. of days mouth Year mean from mean (Oct—April) (% of mean mouth open* open** end (deg.C) (% of mean) _ total (cm) total) (Nov—Oct) % of year (Nov—Oct) % of year 1996 17.8 99.4 12.85 53.4 S51 96 80 pi, 1997 18 100.5 17.48 ded 358 98 153 42 1998 18.5 103.4 42.75 177.8 338 92 238 65 1999 15.6 Si 16.2 67.4 219%? 60 240 66 NOAA = National Oceanic Atmospheric Administration Center, National Climatic Data Center CDWR = California Department of Water Resources, California Data Exchange Center, CIMIS * = H. Elwany, unpublished data; San Diego County Department of Environmental Health ** = San Diego County Parks Department, unpublished data *** = Al closure events occurred after April 1999 and plant senescence coulteri at this lagoon occurs at the eastern end of the lagoon in an area that receives little to no di- rect tidal flow, although some subsurface tidal in- flow may occur. Mean annual temperature and rainfall data and data for the number of days the lagoon mouths were open for the study period (1996-1999) are provided in Table 1. Rainfall was 53 to 73 percent of average during October—April in 1996, 1997, and 1999 and 178 percent of average during those months in 1998. Annual mean daily temperature showed less variation (87 to 103 percent of average between 1996-1999), METHODS Annual Monitoring Demography. In general, demographic informa- tion was collected in 1996, 1997, 1998, and 1999 at San Dieguito Lagoon and in 1997, 1998, and 1999 at San Elijo Lagoon. To assess demography, 14 plants within each of 10 sampling locations (0.5 x 0.5 m plots) were haphazardly chosen and marked in late January or early February of each monitoring year, when the plants were | to 2 cm tall seedlings. The sampling locations were chosen as representative of the microhabitat diversity and environmental heterogeneity present with the pop- ulations’ existing range at each marsh. Mortality and phenology were assessed on a monthly or twice monthly basis for three months: February, March, and April. Mortality was assessed as the number of marked plants that died between marking and mid- April. Phenology was broken into three basic Stages: vegetative, in bud, and flowering. In addi- tion, plants were examined for signs of potential herbivory. In mid-April, when most plants had al- ready set seed, a minimum of 10 plants and a max- imum of 14 plants were harvested from each sam- pling location to determine aboveground plant height (cm), number of capitulescences (inflores- cences), capitulescence diameter (the diameter of the receptacle in mm excluding ray flowers), num- ber of flowers (total of disc and ray flowers), num- ber of seeds, and seed set (number of seeds/number of flowers). Seeds were also examined for signs of granivory or seed predation. There were some exceptions to the described de- mographic monitoring. Mortality was not assessed during the following marsh/years—San Dieguito Lagoon 1996 and 1999 and San Elijo Lagoon 1999—but plants were harvested in mid-April for measurement of plant height, capitulescence num- ber and diameter, flower and seed number, and seed set. Efforts were made to assess survivorship at San Dieguito Lagoon and San Elijo Lagoon in 1998, but a sedimentation event associated with higher-than- average water levels in the study area at San Elijo Lagoon caused marking materials (colored rubber- bands) to be obscured. The 1998 data for San Die- guito Lagoon was incomplete, as information on mortality was not collected in April. Biotic variables. Density of L. glabrata subsp. coulteri was assessed biweekly in 1997 using a 0.5 x 0.5 m quadrat subdivided into 25 1-dm?’ subquad- rats. One of the subquadrats was randomly chosen, and the number of L. glabrata subsp. coulteri in- dividuals present within the subquadrat was count- ed. Also, vegetative cover within the sampling plot was assessed One time per monitoring year using the 36 cross points of the subdivided 0. 5 X 0.5 m quadrat and recording the species or bareground oc- curring below the cross point. For analysis purpos- es, percent cover was calculated for L. glabrata subsp. coulteri, total vegetation cover, and cover of non-native species. Total vegetation cover included both native and non-native species. Native species were primarily coastal salt marsh inhabitants such as Salicornia virginica L. (pickleweed), Salicornia subterminalis Parish (glasswort), Frankenia salina (Molina) I. M. Johnston (alkali heath), Cressa trux- illensis Kunth (alkali weed), and Spergularia ma- rina (L.) Grisels. (sand-spurrey). Non-native spe- 178 MADRONO cies included Cotula coronopifolia L. (brass-but- tons), Mesembryanthemum crystallinum L. (crys- talline iceplant), Lolium multiflorum Lam. (Italian ryegrass), Parapholis incurva (L.) C. E. Hubb. (sickle grass), Polypogon monspeliensis (L.) Desf. (annual beard grass), and Poa annua L. (annual bluegrass). Abiotic variables. A total of 11 abiotic variables was assessed during each monitoring year, except 1999. The abiotic variables were soil pH, soil sa- linity, soil moisture, organic matter, ammonium, ni- trates + nitrites, phosphorous, cation exchange ca- pacity (CEC), calcium, magnesium, and potassium. Soil texture was assessed at both marshes in 1997. As this species grows in high marsh areas, which are only infrequently inundated, reduction-oxida- tion potential was not measured. Soil pH, soil sa- linity, and soil moisture was measured twice a month (1997) to monthly (1998) from 15-cm-deep soil core samples at all 10 sampling locations. Soil pH was measured by creating soil pastes in the field and measuring with an Oakton phTestr 3 (+0.01 pH resolution) field pH probe. Soil salinity was measured by expressing soil water from a syringe fitted with filter paper onto a refractometer, which reports salinities in grams per kilogram. Soil mois- ture was measured by removing 10- to 15-cm soil cores and assessing loss of mass on drying (Gard- ner 1986). For nutrient analysis, five of the 10 sam- pling locations at each marsh were randomly se- lected for subsampling twice each monitoring year (mid-February and mid-March). At these subsam- pling locations, approximately 100 g of soil was removed, air dried, and sent to A&L Western Ag- ricultural Laboratories (Modesto, California) for measurement of organic matter, ammonium, ni- trates + nitrites, phosphorous, cation exchange ca- pacity, calcium, magnesium, and potassium. In 1997, the laboratory also analyzed soil texture. The procedures described above were performed for both marshes during the years monitoring was per- formed, with the exception of San Dieguito Lagoon in 1996 and 1999 and San Elijo Lagoon in 1999. Soil salinity was not measured at San Dieguito La- goon in 1996, and pH, soil moisture, organic mat- ter, and other nutrients were only analyzed once during the 1996 monitoring year. Data for variables sampled more than once per season were averaged for analysis. Data analysis. Differences in plant population dynamics between marshes and sampling years were assessed by treating each “‘marsh/year”’ sam- pled independently and conducting a One-Way Analysis of Variance using the Systat computing package (SPSS, Chicago, IL). For the density com- parison, a t-test was conducted to test for differ- ences in plant density between marshes. When as- sumptions for parametric tests were not met, data were either transformed, or an equivalent, non- parametric procedure (e.g., Kruskal-Wallis) was | [Vol. 47 conducted. If a significant difference was found,| differences between particular means were ana-'! lyzed further by using either Tukey, T’-method (So-) kal and Rohlf 1981), or non-parametric Tukey-type| (Zar 1984) multiple comparison procedures. The) dependent variable plant height was log-trans-. formed for analysis. Discriminant function analysis was used to ex- plore the association between plant population dy-' namics and 13 biotic and abiotic factors within! marshes. Quadratic discriminant function analyses were performed, because they are less sensitive to’ dissimilarities in covariance matrices between, groups. Groupings used in analyses were based on results from the Analysis of Variance tests, with generally low yield plots separated from high yield plots. The analyses incorporated data from 1996—) 1998: no biotic and abiotic data were collected in 1999. As salinity data were not available for San) Dieguito in 1996, a preliminary analysis was per- formed using models that incorporated the salinity | variable, but not the San Dieguito 1996 data. If sa-' linity did not have a strong loading on any canon-, ical variable, a second or final analysis was per- formed using models that incorporate San Dieguito. 1996 data, but not the salinity variable. For the re-. productive success analyses in which groups were smaller, fewer than 13 variables were incorporated, using F-to-enter from the preliminary analyses as. the criterion. Discriminant function analyses were conducted using the Systat computing package. The following variables were log transformed for anal-. ysis: ammonium, nitrate + nitrite, calcium, and cat-. ion exchange capacity. | RESULTS Comparison of plant yield between marshes. One| of the intents of our study was to compare attributes | of a small, and perhaps unstable, population with those of a large and stable one. A component of this study involved assessment of demographic variables associated with survivorship and yield to determine the degree to which these populations ac- tually differ other than in estimated population size. As shown in Figure 2, there were statistically sig- nificant differences between marsh/years for mean survivorship (per 0.25 m? plot; ANOVA, F = 8.67, n = 30, P = 0.001), mean plant height (ANOVA, F = 56.8, n = 69, P < 0.001), mean capitulescence number (Kruskal-Wallis, Test Stat. = 46.3, n = 69, P < 0.001), mean capitulescence diameter (ANO-. VA, F = 19.6, n = 69, P < 0.001), mean flower number (Kruskal-Wallis, Test Stat. = 28.5, n = 49, P < 0.001), mean number of seeds produced (seed number) (Kruskal-Wallis, Test Stat. = 54.3, n = 69, P = 0.001), and mean seed set (mean number of seeds/mean number of flowers) (ANOVA, F = 25.1, n = 49, P < 0.001). The mean percent cover of L. glabrata subsp. coulteri within 0.25m? sam- pling plots also differed between marsh/years 179 PARSONS AND WHELCHEL: CLIMATIC VARIABILITY AND RARE PLANT Plant Height Survivorship avoul) djqssoajamsg Marsh/Y ear Marsh/Y ear Capitulescence Diameter (juejd/ageul) requinn 929s: enyjdeD Marsh/Y ear Seed Number Flower Number Quejd/agewm) Joquinn JaMOLy Marsh/Y ear Lasthenia Cover Cc Cc Seed Set Marsh/Y ear Marsh/Y ear = or < 0.001). Differences between particular means are indicated 2) Wt I Size: Means for dependent plant variables in marsh/years studied. Differences between marsh/years were significant for all analyses (P for all ANOVA/Kruskal-Wallis by small case letters. Bars represent plus or minus | SE (standard error): means seemingly without bars are ones with very small standard errors. Plot refers to individual sampling plots, which were O Fic. 2. 180 MADRONO (Kruskal-Wallis, Test Stat. = 26.12, n = 49, P < 0.001). Measurements of these variables were not assessed for all marsh/years. Specifically, survivor- ship data were only available for three marsh/years, and survivorship for San Dieguito Lagoon 1998 was assessed only through bud stage. Data for mean flower number, mean seed set, and cover of L. glabrata subsp. coulteri were only available for five marsh/years. Closer examination of the results and multiple comparison tests reveals some interesting trends. Survivorship through the vegetative stage was low- er at San Dieguito Lagoon 1997 than at San Elijo Lagoon 1997 and San Dieguito Lagoon 1998 (Fig. 2). However, despite these statistical differences, seedling survivorship at both marshes remained generally high (>70 percent). Based on biweekly censuses for marked individuals, most mortality oc- curred just prior to or after reproduction. In 1997, total survivorship at reproductive maturity (flow- ering stage) was 72 percent for San Dieguito La- goon and 95 percent for San Elijo Lagoon. As the San Dieguito Lagoon population matured more quickly than San Elijo Lagoon, it is probable that some of the mortality occurred after reproduction and therefore actually constituted senescence. The results suggest that, at least in 1997, the populations were following the Deevey Type I survivorship curve characteristic of stable populations (Pavlik 1994), in that the mortality inflection point fol- lowed onset of seed production. Furthermore, as the data for San Dieguito 1997 was recorded during a below-average rainfall year, low mortality cannot necessarily be ascribed to above-average environ- mental conditions. Means for plant height, capitulescence diameter and number, and flower number also showed some interesting relationships. Means were not only gen- erally lower at San Dieguito Lagoon than at San Elio Lagoon, but remarkably similar between years within the respective marshes. There was one ex- ception. In 1998, yield of the San Dieguito Lagoon plants was actually closer to that of the 1997, 1998, and 1999 San Elijo Lagoon plants. Means for plant height and capitulescence diameter suggested that San Dieguito Lagoon 1999 might be intermediate between low and high yield marsh/years, but those for capitulescence and flower number were equiv- alent to means in low yield years. Results for seed set and seed number were some- what more complex. Plants at San Dieguito Lagoon Set less séed in 1997 than in 1996; and seed set (number of seeds/number of flowers) was lower in both of these marsh/years than in San Dieguito La- goon 1999 and San Elijo Lagoon 1997 and 1999. For the total number of seeds produced, however, the 1996 San Dieguito Lagoon marsh/year was the least productive. The distinction between the re- maining marsh/years was less clearcut, but in terms of seed productivity, the ranking appeared to be, from lowest to highest, as follows: San Dieguito [Vol. 47 | Lagoon 1997 and San Elijo Lagoon 1998; San Die- | guito Lagoon 1999; San Elijo Lagoon 1999; San | Elijo Lagoon 1997; and San Dieguito Lagoon 1998. | Unlike the 1997 survivorship data, these results | suggest that the San Dieguito Lagoon population | might be less stable than that of the larger San Elijo | Lagoon one. Based on non-integrated demographic | trend analysis, seed production per individual of | stable populations should consistently equal or ex- | ceed that of a common congener or more stable | population (Pavlik 1994). With the exception of | 1998, seed production of the San Dieguito Lagoon | plants was typically lower than those at San Elijo Lagoon. The large differences observed in annual : population size may have only exacerbated this dis- parity in seed production between populations. Plant yield and influence of biotic and abiotic factors. The dissimilar patterns in sample means observed for mean seed number and seed set and | the other plant variables may relate to an underly- ing difference in how biotic and abiotic factors af- fect various stages or aspects of plant development. Based on these patterns, we decided to analyze our results by dividing our results into two grouping | structures—reproductive potential and reproductive — success. Reproductive potential measures the po- tential of the plant to be more reproductively suc- cessful through survivorship to reproduction (mor- tality), being larger (mean plant height), and pro-— ducing more capitulescences (mean capitulescence > number) and more flowers (mean capitulescence di- ameter and mean flower number). All of these vari- ables relate to an individual’s ability to outcompete another in terms of attracting pollinators or utilizing limited resources (e.g., water, nitrogen, etc.). Re- productive success measures the actual success of an individual in reproducing, as determined by seed number and seed set (percentage of flowers pro- ducing seed). For dependent variables such as plant height, capitulescence diameter and number, and flower number, marsh/years generally split into two groups based on yield, with San Dieguito Lagoon 1996, 1997, and 1999 in a low yield group (Reproductive Potential 1/RP1) and San Dieguito Lagoon 1998 and San Elijo Lagoon 1997, 1998, and 1999 in a high yield group (Reproductive Potential 2). For the dependent variables seed number and seed set, groupings were less distinct, but marsh/years were separated into three groups, with San Dieguito La- goon 1996 in a low yield group (Reproductive Suc- cess 1/RS1), San Dieguito Lagoon 1997 and San Elijo Lagoon 1998 in an intermediate yield group (Reproductive Success 2/RS2), and San Elijo La- goon 1997 and 1999 and San Dieguito Lagoon 1998 and 1999 in a high yield group (Reproductive Success 3/RS3). | Reproductive potential of a germinated seedling is typically affected by herbivory, environmental factors, and intra- and inter-specific competition. 2000] Herbivory can negatively affect individuals through consumption of either vegetative tissue or flowers, _which may weaken or kill the plant. Through the three years of study, no herbivory of vegetative tis- sue or flowers was observed at either marsh. The effect of intra- and inter-specific competition is not 'as directly observable and can be more complicat- ed. At high intra- or inter-specific densities, seed- | lings can compete for resources or light or become _ more attractive to herbivores. At later stages, how- ever, high densities of synchronously flowering in- _ dividuals, including non-native neighbors such as - Cotula coronopifolia, may also serve to attract pol- linators and thereby enhance reproductive success. Some estimates of L. glabrata subsp. coulteri density were collected in 1997, and densities ranged from three individuals (San Elijo Lagoon) to 140 individuals (San Dieguito Lagoon) per dm?. Overall, the 1997 sampling plots at San Dieguito Lagoon had higher densities per dm? (71.6 + SE 13.5) than those at San Elijo Lagoon (28.4 + SE 7.2) (t-test, t = 2.83, n = 20, P = 0.011). Densities of other species were not estimated, but cover of other native species ranged from O to 75 percent, and cover of non-native species ranged from O to 50 percent. Extremely low total mortality rates for vegetative and flowering individuals at both marsh- es in 1997 (~72 to 95 percent) suggests that either abundance of L. glabrata subsp. coulteri or other species was not high enough, or resources not lim- ited enough, to have induced either intra- and inter- specific competition at the seedling or vegetative stage during this year. While above-average rainfall may have increased seedling densities at San Die- guito Lagoon in 1998, the fact that 89 percent of the plants reached at least bud stage suggests that densities were not high enough to incur density- dependent mortality. Reproductive success is affected by all the same factors as reproductive potential, but other factors can limit reproduction, as well, specifically grani- vory (herbivory of unfertilized ovules or seed) and, for non-vegetative species such as L. glabrata subsp. coulteri, pollination success. Based on sta- tistical analyses, reproductive success was highest for the RS3 group (San Elijo Lagoon 1997 and 1999 and San Dieguito Lagoon 1998 and 1999) and lowest in the RSI group (San Dieguito Lagoon 1996). The fact that marsh/years with technically equivalent reproductive potential (San Dieguito La- goon 1996 and 1998) should have differing rates of reproductive success suggests that a different factor or suite of factors may be affecting seed number and seed set. As noted earlier, no consumption of entire flowers was observed during the three years of study, and low mortality rates indicate that most individuals survived to flowering and seed set. It is possible that competition among individuals for resources in- creased during the flowering stage, as ambient tem- peratures and rates of evaporation and evapotrans- PARSONS AND WHELCHEL: CLIMATIC VARIABILITY AND RARE PLANT 181 piration typically climb during the warm spring months. Flowering often coincides with a neap tide series, a period of extremely low tides that often decrease soil moisture and increase evaporation rates and soil salinity in higher marsh elevations. Some granivory was actually observed in seeds of San Dieguito Lagoon individuals in 1998. The extent of granivory was not quantified, but in gen- eral, the number of individuals and/or number of seeds per individual that appeared to have been af- fected was relatively low. While viability of the damaged seeds was not tested, the damage ap- peared extensive enough to render the seeds invi- able. Granivory, or pre-dispersal predation, was not observed in the other study years at this marsh, nor was it observed in seeds of plants from San Elijo Lagoon. The presence of organisms that would re- move seeds after dispersal, including ground-dwell- ing insects such as ants, was sporadic, and even when present, abundance was low. Ground-dwell- ing organisms observed within L. glabrata subsp. coulteri patches included ants (Formicidae), thrips (Thysanoptera), and rove (Staphylinidae) and other beetles (Bembidion sp. and Dermestidae) (Wesley Maffei personal communication). With the excep- tion of ants, these invertebrates are considered un- likely post-dispersal seed predators. None of these organisms, including the ants, were observed re- moving fallen seeds, nor were birds observed for- aging in these areas. Pollination was not included within the scope of our study, but many insects were observed visiting flowers during our sampling efforts. The primary insect visitors appeared to be solitary bees (An- drena pallidofovea and A. cercocarpi), beetles (Dermestidae, Geocoris sp.), beeflies (Bombyli- idae), flies (Bufolucilia sp. and Nemotelus sp.), but- terflies (Coenonympha californica), and halictine or ‘““sweat”’ bees (Lasioglossum sp.) (W. Maffei per- sonal communication; Robbin Thorp personal com- munication). Several of these visitors have the po- tential to effect pollination either through collecting pollen (e.g., Andrenidae or halictine bees) or for- aging on pollen or other flower parts (e.g., Der- mestidae). Overall, visitor numbers and species di- versity appeared to be lower at San Dieguito La- goon than at San Elijo Lagoon, although no formal pollinator observations were conducted. Based on these observations, we hypothesize that resources such as nutrients and perhaps even pollen may be the primary determinants or reproductive potential and success. Discriminant function analysis. To explore fur- ther the association between biotic and abiotic fac- tors and the groupings of marsh/years suggested by results of multiple comparison testing, discriminant function analyses were performed. The question posed by these analyses was two-fold. Using the groups suggested by multiple comparison testing, was there some combination of biotic and abiotic 182 MADRONO [Vol. 47 | ! i i TABLE 2. RESULTS OF THE DISCRIMINANT FUNCTION ANALYSES FOR THE REPRODUCTIVE POTENTIAL AND SUCCESS MODELS. | Reproductive Potential Model Canonical Variable: 0.002*pH — 0.193*organicmatter — 1.330*soilmoisture + 1.092*phosphorous — 0.005*potassium — 0.189*magnesium — 0.443*totalplantcover — 0.406*ammonium_log — 0.113*nitrates + ni- trites_log + 0.326*non-nativeplantcover — 0.487*calcium_log — 0.699*CEC_log Classification results: Actual groups RP1 RP1 16(100) RP2 0 Total % Correct Reproductive Success Model Canonical Variable 1: 0.386*pH — 1.235*soilmoisture — Results—Cases (%) Jackknifed Results—Cases (%) RP2 RP1 RP2 0) 15(94) 1(6) 15(100) O 15(100) 100 OF 0.65 1*organicmatter + 0.937*phosphorous — 0.438*potas- | sium + 0.106*magnesium + 0.422*totalplantcover + 0.756*ammonium-_log + 1.406*nitrates + nitrites_log Canonical Variable 2: 0.482*pH + 0.771*soilmoisture + 0.258*organicmatter — 1.265*phosphorous + 0.421*potas- sium + 0.786*magnesium + 0.253*totalplantcover + 0.754*ammonium_log + 0.497*nitrates + nitrites_log : ; Results—Cases Classification results: Actual groups RS1 (%) RS2 (%) RS1 10(100) 0 RS2 6) 11(100) RS3 6) @) Total % Correct Reproductive Potential Variables measured RP1 pH 759 % Soil moisture 29.6 % Organic matter 2.62 Phosphorous (ppm) 46.1 Potassium (ppm) 414.6 Magnesium (ppm) 1036.6 ] % Total plant cover 16:3 Log-Ammonium (ppm) Leds Log-Nitrates + Nitrites (ppm) 1.38 % Non-native plant cover Pa Log-Calcium (ppm) 3.42 Log-CEC (meq/100g) Lie Soil salinity (ppt) 58.7NI (%) Jackknifed Results—Cases (%) RS3 (%) RS1 (%) RS2 (%) RS3 (%) 0) 10( 100) 0) 0) 0 0) 11(100) 0) 10(100) 0 0 10(100) 100 100 Group Means for Models Reproductive Success RP2 RS1 RS2 RS3 7.83 7.43 118 7.90 40.3 25.9 3732 4] 3,09 1.84 3.39 4.11 S2ar 39.90 53.41 54.05 346.6 425.60 347.86 375.10 104.2 1024.60 1030.82 1156.35 64 89.8 65.9 Sch et 1.09 1.09 P22 1.06 1.68 0.96 1.06 8.5 25.0NI 19.2NI 9.9NI 3.76 3.35NI 3.65NI SIN 2.26 1.83NI 2.12NI 2.04NI 33.1NI NA 38.6NI 42.4NI NA-Not available; NI-Not included in models (see Methods and/or Results for explanation.) variables that would allow us to discriminate be- tween these groups? And, if so, what combination of variables would allow us to best predict the group to which the sampling location belonged? Two models were used. One model separated sam- pling locations into two groups (RPI, RP2) based on differences in sample means for reproductive potential variables, and another separated sampling locations into three groups (RSI, RS2, and RS3) based on differences in sample means for repro- ductive success variables. The groupings were es- sentially the same as described previously, except for the absence of San Dieguito Lagoon and San Elijo Lagoon 1999: no biotic and abiotic data were collected in 1999. The 13 biotic and abiotic vari- ables used in the models were: pH, salinity, soil moisture, organic matter, ammonium, nitrates + ni- trites, phosphorous, potassium, magnesium, Calci- um, cation exchange capacity, total vegetation cov- er, and non-native plant species cover. While soil | salinity appeared to be higher for RPI than RP2 (Table 2), it did not have a strong loading in pre- | liminary analyses for either model or correlation | with other variables and was therefore not incor- | porated into final analyses. According to the reproductive potential analysis, | the biotic and abiotic variables used discriminated well between the groups suggested by multiple | comparison results (F = 8.34, n = 31, P < 0.0001). | Canonical scores of group means were 2.21 for. RP! and —2.36 for RP2. The canonical discrimi- | nant function accounted for approximately 100 per- | cent of the total dispersion in the data. Based on the standardized functions, most of the group sep- | aration came from soil moisture, phosphorous, cal- cium, and cation exchange capacity. The canonical variable and a list of group means is provided in > Table 2. Using the canonical variable, the model | | | | } ( | 2000] PARSONS AND WHELCHEL: CLIMATIC VARIABILITY AND RARE PLANT 183 10.0 iy 6.6 el mM < eC 3.2 S =I < On-02 = MARSH/YEAR GROUPS O = oO RS1 O -3.6 |- RS2 t RS3 -7, Qh - /0 -36 -0.2 3.2 6.6 10.0 CANONICAL VARIABLE 1 Fic. 3. Canonical scores plot for the reproductive success model of the discriminant function analysis. Soil moisture and nitrate + nitrite concentrations appeared to provide most of the separation between RS1 and RS2/RS3. Abiotic factors were also able to separate RS2 and RS3, although the separation appeared weaker. was able to correctly predict or classify sampling locations 100 percent of the time Gackknifed clas- sification matrix = 97 percent). The reproductive success analysis was also suc- cessful at using nine of the biotic and abiotic vari- ables to discriminate between the three groups sug- gested by multiple comparison results (F = 18.04, n = 31, P < 0.0001; Fig. 3). Canonical scores of group means were (8.09, 0.03) for RSI, (—3.72, —1.29) for RS2, and (—4.00, 1.38) for RS3. The canonical discriminant function, which has two variables, accounted for 100 percent of the total dispersion in the data. The first canonical variable accounted for approximately 96 percent of the data dispersion. The canonical variables and a list of group means are provided in Table 2. As Figure 3 illustrates, most of the group separation comes from the first canonical variable, which appears to split RSI (San Dieguito 1996) from the other marsh/ years. The first canonical variable had strong load- ings for nitrates + nitrites, soil moisture, phospho- rous, ammonium, and organic matter. The second canonical variable involved separation of RS2 from RS3, although the degree of separation relative to RS1 appeared weaker. At least three of the sam- pling locations from RS2 appeared to associate more strongly with RS3, while one of the RS3 lo- cations was grouped with RS2 (Fig. 3). The second canonical variable had strong loadings for phos- phorous, magnesium, soil moisture, ammonium, and nitrates + nitrites. Using the canonical vari- ables, the model was able to correctly classify sam- pling locations 100 percent of the time (jackknifed Classification matrix = 100 percent). The strong separation of marsh/year groups ef- fected by soil moisture, phosphorous, cation ex- change capacity, and calcium supports the premise that resources are the primary limiting factors of reproductive potential. Standardized canonical co- efficients and group means for soil moisture, cation exchange capacity, and calcium concentrations point to a positive, perhaps even linear, relationship between resource factor and plant variabie (Table 2—canonical variable). The relationship between phosphorous concen- trations and L. glabrata subsp. coulteri yield ap- pears somewhat more complicated than that for soil moisture, cation exchange capacity, and calcium concentrations. Standardized coefficients suggest that elevated phosphorous concentrations may ac- tually drive the canonical score toward RP1, which had a low yield. In contrast, group means show highest phosphorous concentrations in marsh/years with the highest reproductive potential. This dis- parity between group means and standardized co- efficients for phosphorous also occurred in the re- productive success model. As shown in Figure 3, resources also appear to play a role, if perhaps a more limited one, in re- productive success. As hypothesized earlier, the suite of factors influencing vegetative yield ap- peared to be slightly different from that affecting seed production, which may account for the differ- ence in reproductive success observed between marsh/years with similarly sized individuals (San Dieguito Lagoon 1996 and San Dieguito Lagoon 1997). While soil moisture and phosphorous con- centrations featured prominently in both analyses, 184 MADRONO nitrogen concentrations (ammonium and nitrates + nitrites) appeared to have a larger effect on repro- ductive success than on reproductive potential. As was the case with phosphorous, standardized coef- ficients for nitrates + nitrites and ammonium sug- gest that elevated concentrations of inorganic nitro- gen may actually drive canonical scores toward RS1 and a reduced reproductive yield, although group means for ammonium were marginally high- er in RS3 than in either RSI or RS2. Based on standardized coefficients and group means, only or- ganic nitrogen sources such as organic matter ap- peared to contribute directly to enhanced seed pro- duction. The evidence for resource limitation of seed pro- duction is somewhat weaker for the RS2 and RS3 groups. The reproductive success analysis did pro- vide at least enough separation between RS2 and RS3 using canonical variable 2 to enable successful group differentiation and prediction. As with ca- nonical variable | (Table 2), standardized coeffi- cients for phosphorous in canonical variable 2 again appear to drive canonical scoring toward re- duced yield, despite the fact that phosphorous con- centrations were slightly higher in the group pro- ducing the most seeds (RS3). However, in contrast to canonical variable 1, inorganic nitrogen, along with soil moisture and magnesium, appeared to play a positive role in influencing reproductive yield. While the biotic and abiotic factors included in the analysis do enable successful separation be- tween RS2 and RS3, the slight to moderate overlap between groups displayed graphically in Figure 3 suggest that, at some resource level, the number of seeds produced may be driven by other factors not included in this analysis, which is pollen supply. DISCUSSION Soil moisture would seem an unlikely constraint in a salt marsh, but the high marsh represents a distinct ecotone in an aquatic environment. In gen- eral, high marsh species must contend with a com- plex series of hydrologic cycles: days or even weeks of flooding in the winter may be followed by months where the high marsh or marsh periph- ery is only inundated or saturated from subsurface flow during the highest high tides. The hydrologic complexity is compounded in managed lagoons, where the lagoon may be flooded deliberately to attract waterfowl or the tidal inlet may remain closed for most of the year even after winter storms elevate internal water levels. In several instances, water or moisture stress has been singled out as a primary factor limiting growth of species in the up- per marsh zones (Boorman 1971; De Leeuw et al. 1990). Conversely, too much water or waterlogging can negatively affect species adapted to the typi- cally well-drained soils of the high marsh or marsh periphery (Phleger 1971; Nestler 1977; Parrondo et the most probable of [Vol. 47) | al. 1978; Cooper 1982; Seliskar 1985; Adams and Bate 1994). Waterlogging may account for the anomalous eal sults recorded in 1998, when survival (L. Parsons | personal observation) and reproductive yield of the San Elijo Lagoon population plummeted and was/ significantly less than that of San Dieguito Lagoon. | Nineteen ninety-eight was the one year during our) study when rainfall was above average (178 percent of average during the months October—April; Table | 1). During that year, back-up of run-off and creek flows kept water levels within the eastern area of | the lagoon substantially elevated for weeks. In gen-. eral, reproductive yield of this population was ace | tually highest in the two years where rainfall was. slightly below average—1997 (73 percent of aver- | age) and 1999 (67 percent of average). In below- | average years, the current hydrologic management, regime, in which the sluice gates are closed for wa- terfowl enhancement and outflow is provided. through dips in a dike system, may actually en- hance the population by artificially maintaining sat- urated soil conditions within the eastern portion of | the lagoon. Conversely, the response of the San. Dieguito Lagoon population to rainfall is more con-. sistent with plants being limited by lack of water., Above-average rainfall during 1998 was directly. associated with dramatic increases in vegetative and reproductive yield. The positive association be- | tween rainfall and yield, combined with the strong, evidence of resource limitation in discriminant function analyses, suggests that, at San Dieguito) Lagoon, rainfall both directly and indirectly boosts | input and cycling of resources such as water and nutrients. . The importance of nutrient limitation in coastal | salt marsh plant communities has been well docu- | mented (Tyler 1967; Pomeroy et al. 1969; Valiela. and Teal 1974; DeLaune et al. 1979; Smart 1982;, Long and Mason 1983; Mitsch and Gosselink 1986; | Covin and Zedler 1988; Langis et al. 1991; Parsons. and Zedler 1997; Boyer and Zedler 1988 and. 1999). Our results generally show that higher yields | are linked to higher nutrient concentrations. The. seemingly negative relationship between phospho- rous and inorganic nitrogen concentrations and. plant yield observed in analyses could have resulted | from some indirect effect of nutrient influx, such as | greater competition with more abundant species for light, moisture, or nutrients (Bollens et al. 1998). However, neither total plant cover or cover of non- | native species factored strongly into the discrimi-. nant function analyses. Based on group means showing elevated levels of phosphorous and, to. some extent, ammonium in high yield plots (Table 2), it is more probable that these nutrients must in- | teract with other resource variables in such a way that yield is maximized in areas with moderate con- | centrations of phosphorous and inorganic nitrogen. If such an interaction exists, our analyses were not sensitive enough to detect it, as no strong correla- 2000] | tion was evident between biotic and abiotic depen- dent or predictor variables (correlation <71 per- | cent). In general, plasticity in growth or reproduction in relation to rainfall and changes in soil moisture and nutrient input should be expected in annual plant species within Mediterranean climates, even in aquatic systems such as salt marshes. These op- portunistic life forms must rely almost entirely on nature’s largesse to propagate, survive, and succeed as they have none of the mechanisms (e.g., deep taproots, strongly developed mycorrhizal associa- tions, waxy cuticle layer on leaves, etc.) that enable perennial plants to cope with drought and other cli- matic challenges. Several studies on salt marsh an- nuals, including an occurrence of L. glabrata subsp. coulteri at Carpinteria Marsh near Santa Barbara, California, have linked above average rainfall to increases in relative abundance (Allison 1992; Par- sons and Zedler 1997) and density, distribution, and biomass (Callaway and Sabraw 1994). Terrestrial species are also strongly influenced by soil moisture (Reynolds et al. 1997; Center for Conservation Bi- ology 1994), with yield for grassland members of Lasthenia such as L. californica optimized both during wet years and when growing in wet micro- sites (Hobbs and Mooney 1991, 1995). For peren- nial species, the effect of below average rainfall may be more subtle than for annuals, though no less significant, resulting in substantial reductions in seed set (Morgan 1999) and ultimately recruitment and population growth rates (Maschinski et al. 1997). Plasticity in reproduction can be exacerbated by inter-annual variability in other types of “‘re- sources”’ such as pollen. While no information ex- ists on the mating system of southern California coastal populations, in general, L. glabrata subsp. coulteri has been categorized as one of the 14 of 17 Lasthenia species that is self-incompatible (Orn- duff 1966). Several species of insects such as bees, beeflies, flies, and beetles were observed visiting flowers, although what role these species have in effecting pollination of L. glabrata subsp. coulteri is unknown. Based on the species’ presumed status as an entomophilous outcrosser, reproductive yield must depend to some degree on pollination success. As with their host species, pollinators, some of which are believed to nest in marshes or adjacent upland areas, can be affected by climatic variations and watershed disturbances, including flooding (Stephen et al. 1969). Given the myriad of ecological interactions in- volved, it is not surprising that the factors govern- ing reproductive potential and success of L. glabra- ta subsp. coulteri may prove complex both in terms of time and scale. A number of recent studies have Supported the potential for spatial or temporal het- erogeneity in resource and pollen limitations (McCall and Primack 1987; Zimmerman and Aide 1989; Campbell and Halama 1993; Lawrence 1993; PARSONS AND WHELCHEL: CLIMATIC VARIABILITY AND RARE PLANT 185 Parsons 1994; Parsons and Zedler 1997). Our study supports not only inter-annual heterogeneity in re- source limitations, but possibly intra-annual hetero- geneity, as well. For example, while reproductive potential and rainfall totals between October— March were similar for the 1996 and 1999 San Die- guito Lagoon populations, seed set was higher in 1999 than in 1996. A series of storms in early April 1999 may have eased resource constraints during the seed set period, allowing the sparse population of small plants to produce comparatively larger numbers of seed. In general, however, the complex hydrology of urbanized watersheds with dams, year-round urban run-off, and mouth closures would seemingly argue against a tight linkage be- tween rainfall patterns and resource inputs and cy- cling. CONCLUSIONS As we originally surmised, the L. glabrata subsp. coulteri population at San Dieguito Lagoon is not only smaller than the one at San Elijo Lagoon, but, based on trends in seed production, less stable, as well. For the most part, plants at San Dieguito La- goon were smaller and produced less flowers and capitulescences and seed than those at San Elhljo Lagoon. However, the nature of this relationship appears to be temporally variable and highly de- pendent on climatic conditions such as rainfall to- tals and distribution. In a year with above-average rainfall, yield of the San Dieguito Lagoon popula- tion was similar to and, in some ways, greater than that of the more stable one at San Elijo Lagoon. As rainfall is often linked directly and indirectly to in- puts and cycling of resources such as water and nutrients, the strong association found between re- sources and reproductive potential and, to some ex- tent, reproductive success is certainly not surpris- ing, although the relationship was not always either simple or linear. Too much water actually appeared to decrease survival and reproductive yield of the 1998 San Elijo Lagoon population by inducing ‘““waterlogging.”’ In addition, some nutrients such as inorganic nitrogen and phosphorous may require higher levels of other resources such as soil mois- ture before exerting a positive effect on growth or reproduction of L. glabrata subsp. coulteri. When resources are sufficient, seed production appears to be limited by other “‘resources,’’ the most probable of which is pollen supply. The importance of the relationship between cli- matic conditions and population productivity as- sumes a deeper significance when considering the long-term viability of the small San Dieguito La- goon population. Obviously, less seed will be pro- duced in years when few plants are present or plant vigor is reduced. Still, even when the San Dieguito Lagoon population was relatively large and pro- duced more seed per plant than the San Elio La- goon one, productivity of the San Dieguito Lagoon 186 MADRONO population as a whole was still comparatively low- er, because of the difference between marshes in population size. To some extent, the impact of con- sistently producing small numbers of seed could be offset if seed banks are long-lived and/or seed vi- ability and germination rates are high. No research has been specifically conducted on seed bank lon- gevity of L. glabrata subsp. coulteri, but studies on various Lasthenia species have documented long- lived seed banks (10 y; Vivrette 1999) and high germination rates in the field (25 to 69 percent; Thorp 1976) and laboratory (34 to 90 percent; Kingsbury et al. 1976; Callaway et al. 1990; Ra- jakaruna and Bohm 1999; Michael Wall personal communication, March 1999; Doug Gibson unpub- lished data). However, there are indications that germination or emergence from the seed bank for some Lasthenia species may be tightly regulated by the same climatic conditions (Vivrette 1999) that appear to negatively affect yield of L. glabrata subsp. coulteri, at least at San Dieguito Lagoon. Some evidence for this could be seen in the low number of plants present at San Dieguito Lagoon in 2000 (six plants; A. Thorpe personal communi- cation), when rainfall during the primary germina- tion period (October—January) totaled only 10.4 percent of average (San Diego NWS-Lindbergh; California Department of Water Resources, Cali- fornia Data Exchange Center). In drought years, then, both recruitment and individual yield could be reduced, thereby further diminishing productiv- ity of the population as a whole. Poor recruitment and yield in all but above-av- erage rainfall years is of concern for populations in a region such as southern California, where above- average rainfall years are sporadic and often inter- spersed between lengthy periods of drought or be- low-average rainfall. In San Diego County, below- average rainfall occurs 60 percent of the time, while above-average rainfall occurs about 40 percent of the time (Elwany et al. 1998). There are sugges- tions that variability of this already extremely vari- able climate may be increasing due to global warm- ing. Chronically low numbers of plants in average to below-average rainfall years can increase popu- lations’ susceptibility to genetic bottlenecks or ex- tinction due to stochastic or disturbance-related events. Long-term viability of small populations such as San Dieguito Lagoon will probably depend on whether the species can germinate and repro- duce successfully under average, as well as above- average, rainfall and climatic conditions. Future monitoring efforts should focus on assessing repro- ductive potential and success of this population un- der a variety of climatic and hydrologic conditions, as well as better defining pollinator relationships, breeding system, survivorship, seed bank dynam- ics, and field germination rates of L. glabrata subsp. coulteri. [Vol. 47 | IMPLICATIONS FOR MANAGEMENT AND RESTORATION | The information from this study will provide | both preserve and restoration managers with some | guidelines for future efforts to enhance or even re- | introduce L. glabrata subsp. coulteri into salt. marshes. Based on our results, L. glabrata subsp. | coulteri grows best in marshes with moist, but not | waterlogged, soils with low to moderate salinity, | high cation exchange capacity, high percentage of | organic matter, and moderate concentrations of. phosphorous, calcium, and possibly ammonium. To_ ensure a high potential for project success, man- | agers interested in conducting enhancement or re- | introduction projects should carefully evaluate site conditions and hydrologic management regimes. While the goal of restoration and enhancement con-_ tinues to revolve around creation of self-sustaining ecosystems, the reality is that many of our wetland ecosystems are now highly managed through tide. or sluice gates, dikes, culverts, mechanical mouth | breaching, and even deliberate floodings to attract. waterfowl. If management cannot be avoided, it can | perhaps be manipulated to provide benefits to spe- | cies other than waterfowl. Indeed, the high yield recorded at San Elijo Lagoon in years with below- | average rainfall may result in part from artificially | elevated soil moisture conditions created by back- | up of run-off and creek flows when sluice gates are closed during the winter. While, as a science, restoration ecology has moved away from a single-species management ap- proach, there is still a strong need for single-spe- | cies-focused research. Without carefully under- | standing the biotic and abiotic relationships that drive individual species within an ecosystem, we might be tempted to make gross generalizations. about the habitat linkages without ever really grasp- ing the framework of those linkages. For example, what functions of the high marsh are particularly | important for L. glabrata subsp. coulteri, and how do these needs complement or detract from those of other species inhabiting this fragile ecotone, such as Cordylanthus maritimus Benth. subsp. maritimus or Panoquina errans (wandering skipper butterfly)? Directed research on each of these species provides — the pieces for the larger ecosystem puzzle. It is up to restoration and preserve managers to put the puz- zle together in a manner that will maximize benefits © for as many species as possible, as well as the eco-. system as a whole. i ACKNOWLEDGMENTS The authors would like to thank Doug Gibson, John Boland, Troy Kelly, and Roger Briggs for their assistance _ with this project. We also thank Bonnie Peterson for her | contribution to data collection. The comments of Kathy Boyer, Jessica Martini-Lamb, and two reviewers improved this manuscript. LITERATURE CITED ADAMS, J. B. AND G. C. BATE. 1994. The effect of salinity and inundation on the estuarine macrophyte Sarco- cornia perennis (Mill.) A. J. Scott. Aquatic Botany 47:341-348. ALLISON, S. K. 1992. The influence of rainfall variability on the species composition of a northern California salt marsh plant assemblage. Vegetatio 101:145—160. BarrReETT, S. C. H. AND J. R. KOHN. 1991. Genetic and evolutionary consequences of small population size in plants: implications for conservation, Pp. 3—30. in D. A. Falk, and K. E. Holsinger (eds.). Genetics and conservation of rare plants. Oxford University Press, New York, N.Y. ~ BoLLeNns, U., S. GUSEWELL AND F KLorz_i. 1998. On the relative importance of nutrient availability and water level for the biodiversity of wet meadows. Bulletin of the Geobotanical Institute ETH 64(0):91—101. BoorMAN, L. A. 1971. Studies in salt marsh ecology with special reference to the genus Limonium. Journal of Ecology 59:103-—120. Boyer, K. E. AND J. B. ZEDLER. 1998. Effects of nitrogen additions on the vertical structure of a constructed cordgrass marsh. Ecological Applications 8(3):692— VOD: Boyer, K. E. AND J. B. ZEDLER. 1999. Nitrogen addition could shift plant community composition in a restored California salt marsh. Restoration Ecology 7(1):74— 85. CALIFORNIA WETLANDS INFORMATION SYSTEM. 1996. Cali- fornia Environmental Research Exchange System (CERES). California Resources Agency, Sacramento, CA. Electronic communication (web site). CALLAWAY, R. M., S. JONES, W. R. FERREN, JR. AND A. PARIKH. 1990. Ecology of a mediterranean-climate es- tuarine wetland at Carpinteria, California: plant dis- tributions and soil salinity in the upper marsh. Ca- nadian Journal of Botany 68:1139-—1146. CALLAWAY, R. M. AND C. S. SABRAW. 1994. Effects of variable precipitation on the structure and diversity of a California salt marsh community. Journal of Vege- tation Science 5:433—438. CAMPBELL, D. R. AND K. J. HALAMA. 1993. Resource and pollen limitations in seed production in a natural plant population. Ecology 74(4):1043-1051. CENTER FOR CONSERVATION BIOLoGy. 1994. Conservation of the palmate-bracted bird’s beak, Cordylanthus pal- matus. Stanford University, Stanford, CA. CHARLESWORTH, D. AND B. CHARLESWORTH. 1987. In- breeding depression and its evolutionary conse- quences. Annual Review of Ecology and Systematics 18:237-268. Cooper, A. 1982. The effects of salinity and waterlogging on the growth and cation uptake of salt marsh plants. New Phytologist 90:263-—275. Covin, B. AND J. B. ZEDLER. 1988. Nitrogen effects on Spartina foliosa and Salicornia virginica in the salt marsh at Tijuana Estuary, Calif. Wetlands 8:51—65. Davy, A. J. AND R. L. JEFFERIES. 1981. Approaches to the monitoring of rare plant populations, Pp. 219-232. in H. Synge, (ed.). The biological aspects of rare plant conservation. John Wiley & Sons, London, United Kingdom. DeELAung, R. D., R. J. BURESH AND W. H. PATRICK, JR. 1979. Relationship of soil properties to standing crop biomass of Spartina alterniflora in a Louisiana marsh. Estuarine Coastal Marine Science 8:477—487. PARSONS AND WHELCHEL: CLIMATIC VARIABILITY AND RARE PLANT 187 De LeEEvuw, J., H. OLFF AND J. P. BAKKER. 1990. Year-to- year variation in peak above-ground biomass of six salt marsh angiosperm communities as related to rain- fall deficit and inundation frequency. Aquatic Botany 36:139-151. ELWANY, M. H. S., R. E. FLICK AND S. AUJAZ. 1998. Open- ing and closure of a marginal southern California la- goon inlet. Estuaries 21(2):246—254. ELWANy, M. H. S. 1999. Professor. Center for Coastal Studies, Scripps Institution of Oceanography, La Jol- la, CA. Unpublished data. GARDNER, W. H. 1986. Water content in Methods of Soil Analysis, part 1: physical and mineralogical methods, 2nd ed. American Society of Agronomy—Soil Sci- ence Society of America. Madison, WI. Hamrick, J. L., Y. B. LINHART AND J. B. Mitron. 1979. Relationships between life history characteristics and electrophoretically detectable genetic variation in plants. Annual Review of Ecology and Systematics 10:173—200. HARPER, J. L. 1977. Population biology of plants. Aca- demic Press. London, United Kingdom. Heprick, P. W. 1983. Genetics of populations. Science Books International. Boston, MA. HICKMAN, J. C. 1993. The Jepson manual: higher plants of California. University of California Press, Berke- ley. KINGSBURY, R. W., A. RADLOw, P. J. MubIg, J. RUTHER- FORD AND R. RADLOW. 1976. Salt stress responses in Lasthenia glabrata, a winter annual composite en- demic to saline soils. Canadian Journal of Botany 54: 1377-1385. LANGIS, R., M. ZALEJKO AND J. B. ZEDLER. 1991. Nitrogen assessments in a constructed and natural salt marsh of San Diego Bay. Ecological Applications 1:40—51. LAWRENCE, W. S. 1993. Resource and pollen limitation: plant size-dependent reproductive patterns in Physalis longifolia. American Naturalist 141(2):296—313. Lepic, E T. 1986. Heterozygosity, heterosis, and fitness in outbreeding plants, p. 77-104. in M. E. Soule, (ed.). Conservation biology: the science of scarcity and di- versity. Sinauer Press, Sunderland, MA. Lona, S. P. AND C. EF MAson. 1983. Saltmarsh ecology. Cambridge University Press, Cambridge, United Kingdom. MASCHINSKI, J., R. FRYE AND S. RUTMAN. 1997. Demog- raphy and population viability of an endangered plant species before and after protection from trampling. Conservation Biology 11(4):990—999. McCALL, C. AND R. B. PRIMACK. 1987. Resources limit the fecundity of three woodland herbs. Oecologia 71: 431-435. MEC ANALYTICAL SYSTEMS, INC. 1993. San Dieguito La- goon restoration project biological baseline study, March 1992—May 1993, Vol. I, 2nd ed. Draft tech- nical memorandum. Prepared for Southern California Edison. Rosemead, CA. MENGES, E. S. 1986. Predicting the future of rare plant populations: demographic monitoring and modeling. Natural Areas Journal 6(3):13—25. MENGES, E. S. AND D. R. GORDON. 1996. Three levels of monitoring intensity for rare plant species. Natural Areas Journal 16(3):227—237. Mitscu, W. J. AND J. G. GOSSELINK. 1986. Wetlands. Van Norstrand Reinhold Company, New York, NY. MoraGan, J. W. 1999. Effects of population size on seed production and germinability in an endangered, frag- 188 mented grassland plant. Conservation Biology 13(2): 266-273. NATURAL DIVERSITY DATA BASE (NDDB). 1998. Califor- nia Department of Fish and Game. Sacramento, CA. Nel, M. Molecular population genetics and evolution. E]- sevier, New York, N.Y. NESTLER, J. 1977. Interstitial salinity as a cause of eco- phenic variation in Spartina alterniflora. Estuarine and Coastal Marine Science 5:707—714. ORNDUFF, R. 1966. A biosystematic survey of the gold- field genus Lasthenia (Compositae:Helenieae). Uni- versity of California Publications in Botany 40:1—92. PARRONDO, R. T., J. G. GOSSELINK AND C. S. HOPKINSON. 1978. Effects of salinity and drainage on the growth of three salt marsh grasses. Botanical Gazette 139(1): 102-107. PARSONS, L. S. 1994. Re-establishment of salt marsh bird’s beak at Sweetwater Marsh: factors affecting repro- ductive success. Master’s Thesis, San Diego State University, San Diego, CA. Parsons, L. S. AND J. B. ZEDLER. 1997. Factors affecting reestablishment of an endangered annual plant at a California salt marsh. Ecological Applications 7(1): 253-267. PAVLIk, B. M. 1994. Demographic monitoring and the re- covery of endangered plants. Pp. 322-350. in M. L. Bowles, and C. J. Whelan (eds.). Restoration of en- dangered species: conceptual issues, planning and im- plementation. Cambridge University Press, Cam- bridge, United Kingdom. PHLEGER, C. F 1971. Effect of salinity on growth of salt marsh grass. Ecology 52:908-911. PoMEROY, L. R., R. E. JOHANNES, E. P. ODUM AND B. ROFF- MAN. 1969. The phosphorous and zinc cycles and pro- ductivity of a salt marsh, Pp. 412—419. in D. J. Nel- son, and E C. Evans (eds.). Proc. 2nd Natural Sym- posium on Radioecology. POWELL, A. H. AND G. V. N. POWELL. 1987. Population dynamics of male euglossine bees in Amazonian for- est fragments. Biotropica 19:176—179. PuRER, E. A. 1942. Plant ecology of the coastal salt marshlands of San Diego County, California. Ecolog- ical Monographs 12(1):83—111. RAJAKARUNA, N. AND B. A. BOHM. 1999. The edaphic fac- tor and patterns of variation in Lasthenia californica (Asteraceae). American Journal of Botany 86(11): 1576-1596. REYNOLDS, H. L., B. A. HUNGATE, F S. CHAPIN, III AND C. M. D’ANTOoNIo. 1997. Soil heterogeneity and plant competition in an annual grassland. Ecology 78(7): 2076-2090. SAN DIEGO COUNTY DEPARTMENT OF PARKS, AND RECRE- ATION. UNPUBLISHED DATA. Recording of date of mouth opening, duration of opening, and type of opening (artificial vs. natural). San Diego, CA. MADRONO SCHEMSKE, D. W., B. C. HusBAND, M. H. RUCKELSHAUS, | C. GOoDwILLig, I. M. PARKER, AND J. G. BISHOP. | 1994. Evaluating approaches to the conservation of | rare and endangered plants. Ecology 75(3):584—606. | SEA SCIENCE SERVICES, AND PACIFIC SOUTHWEST BIOLOGI- CAL SERVICES, INC. 1980. Report on environmental | baseline studies for the San Dieguito Lagoon En- | hancement Plan, Del Mar, California. Prepared for the Department of Planning and Community Develop- ment, City of Del Mar and the State of California Coastal Conservancy. SELISKAR, D. M. 1985. Morphometric variations in five | tidal marsh halophytes along environmental gradi- | ents. American Journal of Botany 72(9):1340—1352. SHAFFER, M. L. 1981. Minimum population sizes for spe- | cies conservation. BioScience 31:131—134. SKINNER, M. W. AND B. M. PAVLIK. 1994. Inventory of | rare and endangered vascular plants of California. — California Native Plant Society, Sacramento, CA. SMART, R. M. 1982. Distribution and environmental con- | trol of productivity and growth form of Spartina al- | terniflora. in D. N. Sen, and K. S. Rajpurokit (eds.). | Contributions to the ecology of halophytes. Dr. Junk, | The Hague, The Netherlands. SoKAL, R. R. AND E J. RouLF. 1981. Biometry. W. H. Freeman and Company, New York, N.Y. STEPHEN, W. P., G. E. BOHART AND P. E Torcuio. 1969. : The biology and external morphology of bees. Agri- | cultural Experiment Station, Oregon State University, Corvallis, OR. THORP, R. W. 1976. Insect pollination of vernal pool flow- ers, Pp. 36—40. in S. Jain (ed.), Vernal pools, their | ecology, and conservation. Institute of Ecology Pub- lication No. 9. University of California-Davis, Davis, CA. TYLER, G. 1967. On the effect of phosphorous and nitro- | gen, supplied to Baltic shore-meadow vegetation. Bo- taniska Notiser 120:433—447. VALIELA, I. AND J. M. TEAL. 1974. Nutrient limitation in | salt marsh vegetation, Pp. 547-563. in R. J. Reinhold | and W. H. Queen (eds.). Ecology of halophytes. Ac- ademic Press, New York, NY. VIVRETTE, N. 1999. Coastal bluff vegetation change over 25 years on Santa Cruz Island in Fifth California Is- lands Symposium: schedule of symposium proceed- ings and events and presentation and poster abstracts. March 29—April 1, 1999. WALL, M. 1999. Seed curator. Rancho Santa Ana Botanic Garden at Claremont, Claremont, CA. Personal com- munication. ZAR, J. H. 1984. Biostatistical analysis. Prentice-Hall, En- glewood Cliffs, NJ. [Vol. 47. ZIMMERMAN, J. K. AND T. M. Ae. 1989. Patterns of fruit | production in a neotropical orchid: pollinators vs. re- source limitation. American Journal of Botany 76(1): 67-73. | “Maprono, Vol. 47, No. 3, pp. 189-194, 2000 PLEISTOCENE MACROFOSSIL RECORDS OF FOUR-NEEDLED PINYON OR JUNIPER ENCINAL IN THE NORTHERN VIZCAINO DESERT, BAJA CALIFORNIA DEL NORTE PHILIP V. WELLS Department of Ecology and Evolutionary Biology, University of Kansas, Lawrence, KS 66045 ABSTRACT Two late-Pleistocene Neotoma (wood rat) middens have been dated by four radiocarbon analyses at 10,000—10,200 and 17,470 radiocarbon years. Both deposits document by numerous macrofossils the abundance of Juniperus californica Carriere, but in the older deposit dominance is shared with Pinus quadrifolia Parl. Both deposits contain lesser quantities of the principal dominant shrub of the California chaparral, the chamise (Adenostoma fasciculatum Hook. & Arn.) together with the shrubby oak, Quercus turbinella E. Greene, and other chaparral genera at about 30°N in the northern part of the Vizcaino desert. The existing desert vegetation at both sites is dominated by giant columnar xerophytes and several species of low, desert shrubs, no trace of which has been detected in either of the dated middens. Although abundance of macrofossils of woodland trees with lesser amounts of chaparral shrubs, in conjunction with absence of any species of desert shrub, document a modest displacement of desert vegetation at moderate elevations (550-594 m) in the northernmost Vizcaino desert, this evidence cannot be extrapolated to include the entire peninsula of Baja California. Very substantial biogeographic and ecologically pertinent physiographic evidence suggest a major desert barrier in the central part of the peninsula that also may explain the high degree of endemism in the desert flora. RESUMEN En Desierto Vizcaino, Baja California del Norte, Mexico, en 550—594 m, dos depositos de epoca tardo- Pleistocene con datos 10,000—10,200 y 17,470 anos, documentan con macrofosiles numerosos la presencia de Pinus quadrifolia y/o Juniperus californica, con arbustos de encinal (Quercus turbinella, Adenostoma fasciculatum). Sin embargo, arbustos de desierto moderno ausente en los dos depositos. Some of the most spectacular desert vegetation in North America occupies a relatively restricted sector of northern Baja California, north of the Viz- caino Plain in the central part of the peninsula. Shreve (1951), who was very familiar with all the North American deserts, referred to it as a desert wonderland (on p. 108). This unique assemblage of bizarre xerophytes of gigantic stature extends from just south of the lofty Sierra San Pedro Martir, east of El Rosario on the Pacific Coast, south to about San Borja, a distance of about 250 km. The most spectacular scenery is more localized, reaching a peak in areas of weathered granitic rocks, such as the extensive tract north of Santa Ines or farther south near Portezuelo. The unique plant is the cirio tree (dria colum- naris Kell.), a remote relative of the common oco- tillo (Fouquieria splendens Engelm. also present) of the same family, but the cirio attains a height of more than 16 m and has a single main trunk with spongy wood that stores water. It has been likened to an inverted, spiny carrot with innumerable short branches arranged in dense spiral phyllotaxy. Al- though it resembles a giant columnar cactus, it has ordinary C3 photosynthesis, unlike most cacti which have crassulacean acid metabolism (CAM). The cirio is practically endemic to this sector of Baja California, reaching its southern limit on the high Tres Virgenes volcanoes at over 1600 m, east of San Ignacio. The equally impressive cardon cac- tus (Pachycereus pringlei, (S. Wats.) Britt. & Rose) has about the same northern limit on the Pacific slope, but extends south to the Cape in Baja Cali- fornia del Sur, where it is the ubiquitous dominant giant of desert elevations. Another pachycaul with a swollen trunk, the el- ephant tree (Pachycormus discolor (Benth.) Covi- lle) is prominent in the cirio area but extends south with the cardon, though not nearly to the Cape. Closer to the cardon in distribution is another dis- tinctive quasi-endemic tree, Viscainoa geniculata (Kell.) Greene, which associates with cirio in its more limited range. Sprawling beneath the vertical cardons is another large cactus, but of a horizontal mode of growth, the endemic Machaerocereus gummosus, (Engelm.) Britt. & Rose, with a wind- swept appearance and tart fruits (“‘pitahaya agria’’). The lesser cacti include various species of Opuntia, especially chollas (O. cholla Weber, O. molesta Brandegee). A lesser succulent is the euphorbia- ceous Pedilanthus macrocarpus Benth. The domi- nant understory shrubs are mainly dull gray like Ambrosia chenopodiifolia (Benth.) Payne, enliv- ened by weedy patches of Encelia californica Nutt. with its masses of sunflowers, and punctuated by the scarlet flowers of Beloperone shrubs (seen after the winter rains). 190 MADRONO TABLE 1. SITES: AGES: MACROFOSSILS CONIFERS Pinus quadrifolia (intact leaf fascicles) Juniperus californica (leafy twigs, seeds) CHAPARRAL Adenostoma fasciculatum var. obtusifolium (leaves) Quercus turbinella (acorns, leaves) Prunus lyonii (leaves, endocarps) Arctostaphylos glandulosa (nutlets of drupe) Eriodictyon angustifolium (leaf) It was in this most spectacular part of Baja Cal- ifornia that I first sought for evidence of Pleistocene vegetation early in 1968. The Neotoma method (Wells 1976) had already proven fruitful in the Mo- have, western Sonoran (“‘Colorado”’ desert, Cali- fornia), and Chihuahuan deserts (Wells 1966; Wells and Berger 1967). Equipped with a % ton 4 xX 4 pickup and camper, my lifetime friend, Jack Yri- zarry, and I crawled down the thousand-mile desert track to the Cape. ‘““The Baja Run” was then little more than a very rough, rocky trail with incredibly steep grades in places and was strictly one lane in the best stretches, where a speed of 15 mph might be briefly reached! The average speed with a big, rocking camper was about 5 mph. That there was any “‘road’’ at all was due to a sparse procession of big-wheeled trucks driven by native truckers, who provided “‘servicio particular’’ to isolated local rancheros. METHODS AND MATERIALS The Neotoma macrofossil method (first mono- graphed in Wells 1976) is deceptively simple, but requires thorough training in taxonomic and mor- phological botany and proficiency with existing flo- ras over wide areas of North America. Furthermore, it is essential that proficiency extends to minute de- tails exhibited by numerous tiny macrofossils. In- terpretation of the results requires training in ecol- ogy of the vegetation of North America and de- tailed knowledge of its physiography, geology, and climatology. Skilled sampling of the deposits is even more essential. The key to critical stratigraphic analysis is a focus on friable, macrofossil-rich layers. The friable stratigraphic layers are split off as separate units that disintegrate in the dry state into their con- stituent macrofossils. Thus, bulk-processing of mid- dens in homogenizing water baths is avoided. In- stead, the method of dry-processing provides a co- pious yield of macrofossils from discrete strati- graphic units that are carefully sorted on a PLANT COMPOSITION OF NEOTOMA MIDDENS FROM THE CIRIO-CARDON DESERT OF BAJA CALIFORNIA DEL NORTE. Locations shown on Fig. |: site 1 = northwest of Mision San Fernando at ca. 30°N; site 2 = northwest of Rancho! Santa Ines at 29°46’N. Relative abundances: +++ = major constituent; ++ = lesser; + = trace. [Vol. 47. San Fernando, at 594 m 10,000—10,200 + 135 (UCLA: 1365, 1366, Santa Ines, at 550 m 17,470 + 200 (Beta 9372) 1367) absent Satis +++ +++ ++ ++ ++ ++ + + + + multiple-mesh seive set. The macrofossils are sort- | ed as to species, weighed, and reported as percent biomass, or with simpler assemblages assigned rel- | ative abundances. Most species are present as mere | traces. There is often an overwhelming dominance ; of woody plants that fully justifies the vernacular | name “‘wood rat,’”’ coined by mammalogists for the | genus Neotoma. Aliquots of the same friable strati- | graphic layers burned for carbon dating are curated — separately as vouchers for documentation and fur- ther study. RESULTS Less than a week of crawling south of Ensenada | got us well within the northern part of the great cirio desert, northwest of Mision San Fernando. Ex- ploring all of the rock shelters we could spot with. 10X binoculars en route, we finally found an old wood rat (Neotoma) midden in a cavelet located in volcanic rocks. The midden was about one meter | thick and well within a secure shelter, large and dark enough to harbor a few bats. I knew it was | old because it contained none of the desert plants | mentioned above, but rather was composed of. woodland or chaparral trees and shrubs, predomi-_ nantly Juniperus californica, accompanied by leaves and acorns of a shrubby live-oak (Quercus turbinella) and tiny but numerous leaves of cha-— mise (Adenostoma fasciculatum), a principal dom- inant of the California chaparral (Table 1). Also present were a few nutlets of Arctostaphylos glan- dulosa Eastw. More remarkable were the large leaves and endocarps (cherry pits) of Prunus lyonii Eastw., the Catalina cherry. A thorough reconnaisance of the area showed a desert vegetation characterized by dominance of all of the giant columnar xerophytes and a profusion of © lesser cacti and Agavaceae, notably Agave shawii En- gelm., A. deserti Engelm., and Yucca whipplei Torrey. There were no chaparral shrubs except for one indi- vidual of the xerophytic monotype, Xylococcus. A fly in the ointment was the presence of a few, live Ju- WELLS: “niperus californica Carriére about 200 m from the _midden site at an elevation of 610 m (2000 ft). This -somewhat vitiated the significance of the juniper re- _cord, but the chaparral sclerophylls justified the three ‘radiocarbon dates kindly provided by Rainer Berger: 10,000 to 10,200 radiocarbon years before present ‘from top to bottom of the midden. This was cetainly not a pleniglacial date, but rather late-glacial to Ho- | locene transition (Table 1). _ Proceeding slowly south for another few days, we entered the extensive granitic area north of San- ta Ines, where we explored many Neotoma middens in cavities within the small exfoliation domes of the granitoid rocks. Unfortunately, all we saw were re- cords of desert plants. This set a pattern of more desert records all the way to La Paz. Needless to say, I was discouraged with the prospects, but later switched to mainland Mexico, where I have more unpublished records. Still later, I decided to con- centrate on the Great Basin (Wells 1983). Shortly after the 1983 paper appeared, William H. Clark of Albertson College, Caldwell, Idaho, kindly sent me samples of an obviously ancient wood rat midden from the Santa Ines granitic area. It contained Pinus quadrifolia, a four-needled pinyon pine species previously unrecorded in the Neotoma fossil record anywhere, so I had the stratum dated: 17,470 + 200 radiocarbon years before present (Beta 9322). In 1988, I responded to Bill’s generous invitation to visit the site to collect more material. Instead of a week’s drive below Tijuana, I made it in the afternoon of the same day, blacktop all the way! Additional material from the dated sector of the midden provided much more Pinus quadrifolia and Juniperus californica and a similar assemblage of chaparral sclerophylls previ- ously recorded in the 10,000 yr old midden, plus Er- lodictyon angustifolium Nutt. Consistently absent from both the 10,000 and 17,400 y-old deposits were any xerophytes of the existing desert vegetation (Ta- ble 1). The combination of chaparral shrubs, junipers, and pinyons was called encinal by Forrest Shreve, using a Spanish word for an oak community. The west slope of the Sierra San Pedro Martir at middle elevations has chaparral, partly dominated by Quer- cus turbinella and Adenostoma, associated with Pi- nus quadrifolia and Juniperus californica but mi- nus any desert xerophytes, aside from the chaparral Yucca, Y. whipplei. Thus, the encinal recorded in late-glacial and pleniglacial middens in what is now cirio desert has a nearly exact analog in the San Pedro Martir mountains to the north. DISCUSSION The contrast between the spectacular modern de- sert vegetation dominated by giant xerophytes like cirios and cardons at both Neotoma sites and the midden macrofossil evidence for a Pleistocene en- cinal, lacking desert xerophytes, staggers the imag- ination (Table 1). The recorded displacement was complete at this latitude, ca. 30°N. PLEISTOCENE ENCINAL IN CIRIO DESERT 19] Pinus quadrifolia, the four-needled pinyon, is one of the more mesophytic pinyon pines. At pre- sent it is restricted to moderately high elevations (ca. 1100 to 2100 m, or 3500 to 7000 ft), mainly on the western or Pacific slopes of the Peninsular Range as far south as the San Pedro Martir. In con- trast, the one-needled P. californiarum D.K. Bailey (1987) occurs on the rain-shadowed eastern slope of the Peninsular Range (Bailey, personal commu- nication 1975-1990; Wells 1995). Where the two distinct species occur on the same mountain, as on Mt. San Jacinto, California, P. guadrifolia forms a zone well above P. californiarum. The latter, one- needled, pinyon alone has a far southern disjunction on Cerro San Luis (to 1550 m) in the Sierra de Calamajue at about 29°N, an isolated peak not far north of the high San Borja Mountains (to 1700 m +), where no pinyon pines have ever been recorded (Fig. 1). The nominate subspecies of Pinus califor- niarum also extends disjunctly far to the north of P. quadrifolia on low, isolated mountains in arid parts of the Mohave Desert, e.g., the Coxcombs, Eagle, Old Woman, and northeast in the Provi- dence; it dominates the pinyon-juniper zone in all of these ranges (Wells 1995). From the distribution maps of Critchfield and Little (1966), one would have hypothesized P. ca- liforniarum (then under the Great Basin species P. monophylla Torrey & Frémont) to have extended farther south on the peninsula than P. quadrifolia during the Ice Ages, because they show the south- erly outlier of the former on Cerro San Luis. This may well have been true on the east or Gulf of California slope, where we have no Neotoma re- cords. The macrofossil record of P. guadrifolia we do have is from the Pacific side of the peninsular divide in the cirio zone at the substantial elevation of 550 m (1800 ft), less than 500 m below its mod- ern lower limit (Wells 1986). The assemblages of evergreen sclerophylls (Ade- nostoma fasciculatum, or chamise, a principal dom- inant of California chaparral, Prunus lyonii, Arc- tostaphylos glandulosa or Eastwood manzanita, etc.) are consonant with the presence of Juniperus californica or Pinus quadrifolia in the same strata of the Neotoma middens. Today, the juniper and four-needled pinyon both associate with a broad zone of chaparral below montane forest of Pinus Jeffreyi Grev. & Balf. that is mainly above 2500 m; the chaparral belt (largely dominated above by Arc- tostaphylos peninsularis Wells) extends down to <1000 m. Chaparral extends southward from California in the Peninsular Range to Cerro Matomi (to 1370 m), the southern extremity of the San Pedro Martir. Aside from small populations of manzanitas (A. pe- ninsularis) On isolated peaks: Cerro San Juan de Dios (to 1300 m at 30°N) and Cerro San Luis (1550 m: the southern limit of Pinus californiarum at 29°19’'N) there is a major disjunction of chaparral species to the high San Borja Mountains (to 1700 192 MADRONO [Vol. 47), 52°}- a, an oh 3le ; ' 4 | s } SQuinh® i H 30° | | J i 19 r 2X | i. fa™ Ds i} ; y * ne H 8Agua D Rots 6 1 SANS > \ Tecomaty aes | hat ek; tea meet: 1 Ex - Tate {| . i? a hintt 7 ~: | om q ieerie zZ : { RE ‘S. Bor { MUWteede seo PS Cabrigs bi Pedre 28 i SENG A Sr a® , oes LS a : is Cre rancisco Mtns, «sya 27° 26° es A s< ~~ ’ ad vane 6 : A, BUSS OAS TAN Re ” mn + cana ZEAF AG . — gh — culup — canyon land—meses — bedlands NSA ety Ef e us Sen % SEE Ry ies 4 aa whe atl UAT J Or caRmen | : Paws) NS he Ly 2 ' ooo LaTinet dy $F FESS VC Daizante \ Na as u QAZAE cm 4 4: ( va ny SSE ae B\Ni ALS Catalina , oO a9 25 sc tambon i CK al AEG ‘ Bib eyes fon * SR R41 Muertos 3 pa aPaat eS We Mois Ry aN t u - 2 Sure CTE DASE Bl Pol =~ +2 ay PS RR a. abate frst a | ~ a Tedos Sante rb ; ory ny hese \ +2 ‘\eaLaguna Mtns. : it + ne ae He Waar d gate del aba re cauyar J Fic. 1. Map of Baja California, showing principal mountains mentioned in text and numbered wood rat (Neotoma) midden sites of Pleistocene age in white circles: 1 = northwest of Mision San Fernando, 30°N, at 594 m (1950 ft). 2. = northwest of Rancho Santa Ines, 29°46'N, at 550 m (1800 ft). m) rising above the Gulf of California at 28°47'N south at 27°30'N there is a southern outpost of. (Fig. 1). The San Borjas are the southern limit of | chaparral on the lofty Tres Virgenes volcanoes (to — Juniperus californica and Arctostaphylos peninsu- 2000 m) above Santa Rosalia on the Gulf; the Tres laris, but lack pinyon pines (Reid Moran, personal Virgenes are also a southern outpost for cirio trees. | communication 1969; Wells 1972). Still farther Possibly the most unexpected macrofossil occur- | 2000] rence in the northern cirio desert at both Neotoma sites (Table 1) is the Catalina cherry tree (Prunus lyonii). This large-leaved cherry occurs in Califor- nia on the larger Channel Islands, but not on the mainland. In Baja California, however, this ever- _green cherry has widely disjunct stations in deep canyons of the isolated and inaccessible Sierra de San Francisco (or Francisquito) of northern Baja California del Sur at ca. 27°30’N (Fig. 1). The two ' Pleistocene records of Prunus lyonii in the northern part of the cirio zone suggest (but do not prove) a former continuity of range. A major biogeographic anomaly that might shed light on the Pleistocene location of the Sonoran De- sert with its rich array of endemic plants (about 30% of the 2500+ species are endemic: cf. Wells 1970), is posed by the distribution of pinyon pines (cf. Wells 1986). As discussed above, Pinus qua- drifolia is presently restricted to the Sierra San Pe- dro Martir south to about 30°20'N, and the Neoto- ma record at Santa Ines extends that to 29°46’N, a scant 80 km farther south. The one-needled Pinus californiarum has an isolated southern outpost on Cerro San Luis (to 1550 m) at 29°19'N. Neither of these pinyon pines is known from the higher Sierra San Borja (to 1700 m +), where Juniperus califor- nica reaches its southern limit. None of these three conifers occurs in Baja California del Sur, the southern half of the peninsula, which has suitably high mountains such as Tres Virgenes (to 2000 m), Sierra de Santa Lucia (to 2000 m), Sierra de la Gi- ganta (to 1770 m), and Sierra de la Victoria (to 2070 m) = Laguna Mountains (Fig. 1). Instead, the Laguna Mountains support an extensive zone of the three-needled Mexican pinyon, Pinus cembroides Zucc. (subsp. Lagunae Bailey), which has very dis- tinctive pink “‘endosperm”’ (gametophytic tissue). In all other species of pinyon pines the food re- serves of the seed are white. Pinus cembroides has a very wide distribution on the mainland of Mexi- co. There are also some live-oaks in the Lagunas identical to mainland species (Quercus reticulata H. & B., QO. tuberculata Liebmann). Other broad sclerophylls in the isolated Lagunas include the toyon, Heteromeles arbutifolia (Lind- ley) Roemer, Arbutus peninsularis Rose & Gold- man, and Garrya salicifolia Eastw. Most remark- ably, no taxa of Arctostaphylos, Ceanothus, or Ade- nostoma are known from any mountains of Baja California del Sur (S. de la Giganta, S. de Laguna). Absence of these three most characteristic genera of the California chaparral, including the two most Speciose genera, Arctostaphylos and Ceanothus, is strong evidence of a major isolating barrier in the central sector of the peninsula. The southernmost known occurrence of Arctostaphylos peninsularis is in the high Sierra de San Borja at 28°45'N (Wells 2000). The isolated Sierra de la Laguna pinyon-oak woodland is mainly above tropical deciduous for- est, which occupies the lower slopes and surround- WELLS: PLEISTOCENE ENCINAL IN CIRIO DESERT 195 ing foothills of the Lagunas that are on the Tropic of Cancer (23 %°N). The absence of Juniperus ca- lifornica or either of the two northern pinyons may mean that these conifers never migrated this far south. Almost certainly, had Juniperus californica colonized any of the high mountains of Baja Cali- fornia del Sur, possibly even the summer-rainy La- gunas, it may have survived to the present, inas- much as its niche is vacant, there being no other species of Juniperus in Baja California. Although the Neotoma macrofossil evidence from the northern cirio zone documents a modest displacement of desert vegetation by a mesophytic pinyon pine and evergreen chaparral, there is as yet no evidence as to how far south this Pleistocene climatic effect extended. Even if all the elevated areas north of the Vizcaino plain in the central sec- tor of Baja California were affected, there would be ample room for desert vegetation farther south (Fig. 1). The peninsula is immensely long, extend- ing far into subtropical latitudes. The apparent fail- ure of even the relatively xerophytic Juniperus ca- lifornica to colonize any of the mountains of the southern half of Baja California (it stopped far short in the San Borjas at 28°45’'N) suggests a major de- sert barrier in the central sector of the peninsula, where temperature-sensitive giant cacti, and quasi- endemics like /dria (cirio), Pachycormus, Viscai- noa and others may have survived the long Pleis- tocene periods of climatic displacement unscathed. ACKNOWLEDGMENTS Research supported by grants from the National Science Foundation: GB-5002, DEB 78-11187. Special thanks are extended to William H. Clark and his long-term research team at Santa Ines, chiefly from Caldwell, Idaho, who generously shared their discovery of the ancient pinyon midden near their annual camp at Santa Ines. LITERATURE CITED BAILEY, D. K. 1975. Personal communication 1975—1990. . 1987. A study of Pinus, subsection Cembroides: the single-needed pinyons of the Californias and the Great Basin. Notes Royal Botanic Garden, Edinburg 44:275-—310. CRITCHFIELD, W. B. AND E. L. LITTLE. 1966. Geographic distribution of the pines of the world. U.S.D.A. Forest Service Misc. Publ. 991. Moran, R. 1969. In letter. SHREVE, F 1951. Vegetation of the Sonoran Desert. Carnegie Institute of Washington Publication 591, Vol. 1. WELLS, P. V. 1966. Late Pleistocene vegetation and degree of pluvial climatic change in the Chihuahuan Desert. Science 153:970—975. 1970. Historical factors controlling vegetation patterns and floristic distributions in the Central Plains region of North America. Department of Ge- ology, University of Kansas, Special Publ. 3. . 1972. The manzanitas of Baja California, includ- ing a new species of Arctostaphylos. Madrono 21: 268-273. . 1976. Macrofossil analysis of wood rat (Neoto- ma) middens as a key to the Quaternary vegetational 194 MADRONO [Vol. 47 | history of arid America. Quaternary Research 6:223— . 1995. Recognizing the new single-leaf pinyon | 248. pine (Pinus californiarum Bailey) of southern Cali- . 1983. Paleobiogeography of montane islands in fornia. The Four Seasons 10:53-—58. the Great Basin since the last glaciopluvial. Ecol. Mo- . 2000. The Manzanitas of California, Mexico and | nogr. 53:341-382. the World. Published by the author; 151 p., 150 figs. | . 1986. Systematics and distribution of pinyons in , AND R. BERGER. 1967. Late Pleistocene history | the late Quaternary, Pp. 104-108. in R. L. Everett of coniferous woodland in the Mohave Desert. Sci- compiler. Proceedings: Pinyon-Juniper Conference. ence 155:1640—1647. i } “Maprono, Vol. 47, No. 3, pp. 195-203, 2000 SEED BANKS OF LONG-UNBURNED STANDS OF MARITIME CHAPARRAL: COMPOSITION, GERMINATION BEHAVIOR, AND SURVIVAL WITH FIRE DENNIS C. ODION' Department of Geography, University of California, Santa Barbara, CA 93106 ABSTRACT Seed germination requirements in the California chaparral have been described mainly from freshly collected seed. However, uncertainties remain because the behavior of seeds in the soil can differ. I studied germination of the seed bank in long-unburned stands of maritime chaparral in central coastal California. I quantified seedlings emerging from soil samples provided with appropriate temperature and moisture conditions following 1) no other treatment, 2) a heat treatment to optimize germination of heat-stimulated species, 3) the same heat with the addition of charred wood, and 4) the burning of chaparral stands prior to collection of samples. I compared germination in these treatments with seedling emergence in the field following fire. I also collected and divided samples into O0—2.5 and 2.5—7.5 cm depth fractions to evaluate abundance of seed at the surface and depth before and after fire. Seed of one annual had reduced germination following the heat treatment. Seeds of all other species common enough to evaluate statistically were heat tolerant. However, because seeds were found to be mostly near the surface, there was considerable mortality with fire. Moreover, seedling populations in the field only accounted for a fraction of the seed bank that survived fire, and seventeen species that ger- minated in samples did not germinate and/or emerge in the field. Most species’ germination and emergence was influenced in some way by heat and/or charate. Germination of two Ceanothus was dependent on heat. Adenostoma fasciculatum Hook. & Arn., Arctostaphylos purissima P. Wells, and two annuals had germination that was enhanced by heat and enhanced further when charate was added. Despite the 1m- portance of fire effects, there were no short-lived species having entirely fire-dependent germination. Germination and/or emergence of 3 species was negatively affected by charate. These germinated spar- ingly or not at all after fire. One of the most prominent evolutionary special- izations to fire exists in the germination ecology of seeds from plants found in Mediterranean shrub- lands, particularly those of Australia, South Africa, and California (Bond and Van Wilgen 1996). This subject has received considerable attention (Review by Keeley 1991), revealing a fascinating complex- ity of features that insure germination will coincide with the anomalously favorable conditions for seed- ling establishment that exist after fire. There are physical features such as bradyspory (or serotiny) where seeds stored in fruits and cones are released when heated by fire (Whelan 1995), and impervi- ous seed coats that open with the heat of fire (Swee- ney 1956; Quick and Quick 1961; Auld and O’Connell 1991). Physiologically dormant seed may be induced to grow following fire by chemi- cals washed from charred wood (Wicklow 1977; Keeley 1984, 1987; Keeley et al. 1985; Keeley and Pizzorno 1986), water soluble nitrogenous com- pounds (Thanos and Rundel 1996) and smoke (Keith 1997; Keeley and Fotheringham 1997, 1998). For each fire-related germination cue, there are multiple dormancy-releasing mechanisms that have evolved convergently among disparate floras (Baskin and Baskin 1998). In chaparral, germina- ' Present address: Marine Science Institute, University of California, Santa Barbara, CA 93106. tion without fire may also be inhibited by allelo- pathic chemicals washed from foliage or litter (Muller et al. 1968; McPherson and Muller 1969), and/or phytotoxins produced by soil microbes (Ka- minsky 1981). Fire eliminates these compounds. Seeds of chaparral plants range from readily ger- minable at the time of dispersal (non-refractory) to deeply dormant (refractory) as a result of multiple barriers to germination (Keeley and Fotheringham 1998). Some species produce a portion of seed that is refractory and a portion that is not (Emery 1988; Parker and Kelly 1989). Generalizations about the type(s) of dormancy species exhibit derive mainly from tests on freshly collected and stored seed. Germination of seeds residing in the soil may differ significantly, as has been documented for Adenos- toma fasciculatum Hook. & Arn. and Arctostaphy- los canescens Eastw. among others (Stone and Juhren 1951; Parker 1987; Keeley and Fotheringh- am 1998). Seeds exposed to allelopathic chemicals and phytotoxins found in chaparral soils may ex- hibit enforced dormancy (Muller et al. 1968; Mc- Pherson and Muller 1969; Keeley 1991). Therefore, it is imperative to study the soil seed bank to un- derstand how chaparral germination is controlled in nature. The potential for germination in the chaparral seed bank without fire is thought to be low for most species because seedlings are rarely apparent under 196 the shrub canopy. However, Christensen and Muller (1975), Tyler (1995), and Swank and Oechel (1991) reported considerable seedling growth under Ad- enostoma in plots protected from herbivory, sug- gesting more germination can occur without fire than is evident. In addition, Zammit and Zedler (1988, 1994) and Holl et al. (2000) found that many species germinated readily from chaparral soil seed bank samples. To test how much germination can occur in species’ seed banks without fire and how much requires heat and/or chemicals produced by fire, I compared emergence from controls and uni- form fire treatments that are known to maximize germination of refractory seed without inducing mortality. I then analyzed how much of the in situ seed bank was eliminated above and below 2.5 cm in the soil by fire in the chaparral stands. Finally, I enumerated seedling emergence in the field to com- pare germination in nature vs. in collected samples. The chaparral I studied is geographically-isolated and its environment differs in many respects from that found in the Transverse and Peninsular ranges inland. With Santa Ana winds absent, and a lower frequency of ignitions, coastal environments have likely supported less frequent and dynamic fire, at least prior to human dominance of the fire regime (Odion et al. 1992; Odion et al. 1993). The average lifespans of the Ceanothus spp. in maritime chap- arral are particularly short (Davis et al. 1988) com- pared to those inland (Keeley 1975, 1992). Death of the non-sprouters opens space for recruitment by numerous herbs and subshrubs (Odion and Davis 2000). These and other factors such as soil and cli- mate may help explain differences in post-fire re- generation in maritime vs. nearby inland chaparral reported by Tyler (1995); they may also have con- tributed to evolutionary divergence in maritime chaparral taxa that has produced endemic Arctos- taphylos and Ceanothus (Griffin 1978). I have eval- uated my germination data for any evidence that the environment and insular nature of the study area manifested variation in germination ecology. STUDY AREA Samples were collected from Adenostoma fascic- ulatum (hereafter Adenostoma) chaparral located near sea-level, within Vandenberg Air Force Base in central, coastal California as described in D’ Antonio et al. (1993), and Odion and Davis (2000). Substratum here is Pleistocene eolian sand (Dibblee 1950). Climate in the area is strongly in- fluenced by the prevailing onshore winds and cool ocean, and the temperature regime is mild, es- pecially for a chaparral environment. Maritime chaparral of the area has been described in detail by Davis et al. (1988). The average annual precip- itation 1s 35.3 cm. I counted annual rings from the obligate seeders, Arctostaphylos purissima P. Wells, and Ceanothus cuneatus (Hook.) Nutt. var. fasci- cularis (McMinn) Hoover (Keeley 1993) to estab- MADRONO [Vol. 47m lish that the chaparral had not burned for 75-80 y | at the site where more intensive sampling was un- | dertaken (site 1). Samples were also collected from | a second, nearby site which had not previously | burned for about 50 y. Both sites were dominated by Adenostoma. METHODS Transects consisting of 47 contiguous 1 m? plots were established in dense chaparral dominated by Adenostoma. Nine 5 cm diam, 7.5 cm deep cores of soil were obtained per plot at site 1 in the fall of 1988. Five cores were collected in fall, 1989 at site 2. Chaparral at both sites was burned soon thereafter, with low fuel moisture contributing to relatively intense fires (Odion and Davis 2000). I collected 5 cores per plot the day after each fire. Seed bank cores for each plot were composited, and 350 cc subsamples were removed from each ho- mogenized sample. Pre-burn samples from site 1 were given three treatments: 1) heat, 2) heat and charate (charred, pencil-sized Adenostoma stems collected after the fire and ground up) and 3) control (no heat or char- ate). Only the second of these treatments was used on pre-burn samples from site 2. Samples to be heated were spread to a depth of 2—3 mm on alu- minum cooking trays. Based on studies by Wright (1931), Sampson (1944), Sweeney (1956), Keeley and co-workers (several publications, see Keeley (1991), heating at 100°C for ~5 min typically pro- duces the greatest germination response among fire- recruiters, and is well within their heat tolerance. Given the slight insulation the soil would provide, I decided to use a 7 min duration. Heat-induced seed mortality is controlled predominantly by max- imum temperature, as opposed to duration (Borch- ert and Odion 1995), so it is unlikely that this change effected mortality. Heating was done in a forced-air oven. The subsamples were spread on sterile sand in 20 cm plastic pots. The amount of charate added was 2 rounded tablespoons (21.3 + 0.94 g, n = It): The pots were covered with clear plastic, pro- tected from herbivory, and kept moist out-of-doors under 50 percent shade cloth at Cal-Orchid Nursery in Santa Barbara, where temperature fluctuations were analogous to the field. Potting was complete in late November, at which time all samples were given their first watering. All the samples were ex- posed to outdoor temperatures from the time of col- lection through the subsequent growing season to provide natural temperature stratification. Seed bank sampling was also undertaken at ran- domly located ~1.5 m diam canopy gap areas ad- jacent to the site 1 transect. I took samples from the center of the gap as well as the edge and un- derstory of the adjacent Adenostoma canopy. These cores were separated into 0O—2.5 cm and 2.5—7.5 cm - 2000] _ depth fractions and given the heat and charate treat- | ment. Due to smaller amounts available, 175 cc | subsamples were spread over sand in 16 cm diam _ plastic pots placed with the others. _ All germinants were identified and removed from pots through the growing season. Nomenclature | follows Hickman (1993). Specimens whose identity - was uncertain were grown until it was determined. Five of the pre-burn samples treated to heat and _charate were removed from pots after emergence _ stopped. I repotted these the following autumn. No further germination occurred in these. RESULTS General patterns. Seed from 72 species germi- nated and emerged from samples collected along transects (Table 1). More than half (48) were an- nuals. Site 2 had greater diversity (60 vs. 48 spe- cies). Many of the same species emerged abun- dantly from samples from both sites (e.g., Aden- ostoma, Helianthemum scoparium Nutt., Crassula connata (Ruiz Lopez & Pavon) A. Berger, Cen- taurium davyi (Jepson) Abrams, and Navarretia atractyloides (Benth) Hook & Arn.). Despite this, there were only 33 species in common. In addition, two subshrubs, Mimulus aurantiacus Curtis and Lo- tus scoparius (Nutt.) Ottley, were abundant in site 2 samples and absent in those from site |. Several annuals were common in samples from one site but not the other. Nine species were non-native. All 9 are widespread weeds. Post-burn samples contained 44 species (1 non- native) and substantially reduced numbers of ger- minants. The reduction varied with species depend- ing on the proportion of seed present at depth, as described below. Reduced germination was also strongly correlated with the amount of soil heating that occurred where samples were located (Odion and Davis 2000). Thus, the number of post-burn germinants was much greater in samples from gaps vs. under the shrub canopy. Horizontal patterns of seed abundance are analyzed in Odion and Davis (2000). There were 20 and 31 species that germinated in the field respectively at the two sites (Table 1). Twelve were at both. Four species germinated in the field plots but not seed bank samples. There was only one individual of each. Seventeen species ger- minated in samples but did not germinate and/or emerge in the field, including species whose seed was among the most abundant (e.g., Centaurium davyi, Mimulus floribundus Lindley). Another, Crassula connata, was virtually absent in the field in the burn areas, although it was common in ad- jacent unburned chaparral. Field populations for most species were much smaller than post-burn seed bank populations—between ~5 and 14 times smaller for shrubs, and generally even smaller for other species. Germination treatments. Only two species, both ODION: GERMINATION OF MARITIME CHAPARRAL SEED BANK {97 perennial Gnaphaliums, had germination that was not affected by the heat treatment (Table 1). One of these, G. microcephalum Nutt. was significantly negatively affected by charate. Among heat-affect- ed species, the two obligate-seeding species of Ce- anothus, the subshrub Helianthemum scoparium, and the annual Trifolium microcephalum Pursh had significantly greater germination with heat alone, while the opposite occurred for the annual Calan- drinia ciliata (Ruiz Lopez & Pav6én) DC. (Table 1). Other important species that had a positive response to heat were also affected by charate. Adenostoma, Arctostaphylos purissima, and Lotus strigosus had significantly greater germination with heat and charate compared to heat alone. Germination of Centaurium davyiti and Crassula connata with heat and charate was not only signif- icantly lower than with heat alone, but also lower than with no treatment. Crassula was rare in the burn areas, and Centaurium did not occur there un- til the third year after fire. Both species were fairly common in the surrounding unburned chaparral. Mimulus floribundus had much lower germination with heat and charate than with heat alone, but heat and charate germinants outnumbered those in con- trol samples ( P > 0.05, NS). This species, though abundant in several samples, was absent from most. It was never observed in the field, including in un- burned chaparral. It is typically found in seasonal wetlands like two other species that were found in samples, but not in the field, Crassula aquatica (L.) Schonl. and Centunculus minimus L. With a relatively high proportion of seed at depth (76 percent below 2.5 cm in gap, edge, and under- story samples combined, Table 2), Arctostaphylos purissima had better survival (post-burn/pre-burn = 17 percent) than Adenostoma (site 1 = 2 percent, site 3 = 3 percent) which only had 22 percent of its seed below 2.5 cm. These survival percentages are from transect samples. The 2.5—7.5 cm depth samples had relatively little emergence of species with charate-enhanced germination. Seeds of other shrubs were not abundant enough to evaluate depth distribution. The high survival of Ceanothus cu- neatus at site 2 (33 percent) as well as results from a fuel translocation experiment (Odion and Davis 2000) suggest this obligate-seeder had a high pro- portion of seed at depth. Helianthemum scoparium was particularly abun- dant and not affected by charate, so the effect of the depth distribution of its seed is relatively clear. The subshrub had 53 and 75 percent of its seed bank below 2.5 cm in gap and understory samples respectively. Despite the greater proportion of seed at depth at understory plots, survival was similar there (22 percent) compared to gaps (24 percent). Survival along the site | transect was 9 percent (post-burn/pre-burn heat). After fire, 94 percent of Helianthemum scoparium seeds were below 2.5 cm in the soil in understory plots, only half were at depth in gap plots. 0 — vr 0 = DINONbY VINSSVAD ‘~ 07 0 — 0 0 Si 0) 0) 0) 0) 8°8 8°8 VOL 6) SISUAPDUDI DZKUOD S cI vv 0°T7 ZOO 8°8 8°0€ 9°19 8'°0€ pyofusnsuvjosnffip ayjuvz0yd a 0 0 0 0 0 OL +O LOI Co smuunu snjnouniuay 0 VVrV Cc 6VOE 0) 8°CS -C S9E +C SSET 9°L9OS IAAbDp WNn1LAnDjUay vO’ O°CC 9°SOI 0) 0) 0) 0) 0) DYJUDAINU DIUOSSIUDI cO LT OLCL 10 0 0 8°8 VL wnApunuoul wniplsJdavy 0 0 CLS 0 0 v9 -9'6£ OLT DIDI DIUMpUuvIv,) ve 0 9°6€ el 0 TLS +T6L 8°8 unyofiusnsup wunaspidy 0) 0) Col 0) O 0) 0) 6) sypjuap1zz0 saupydy 0 0) 0 0) 0) VY 0) 6) DINIOD siuayJUuy cl VV 9°€8 0) 0 O 0) 0 SISUIAID SIIDBDUY STVOANNY 0 0 0 0 v'Y 0 Ce v'Y ppida] D]]assoN 0 VY 0 0) 0) 8 OST OCVL 8 8IT snpunqiuoy snjnunw 0) 0) vv 0) 0) 0) @) 0 DyJafiadu vIaW CO’ 8'cS V8scl 0) (O6 9°SOT V9ECl OLTI WUNUISSIsOUDs wun1jDYdDUL 0) Vv SVL 0) 0) -CSE 0'86l 8 sil unjoydaso049u wunijoydouy 0) 0 0) 0) 0) 0) TET vr suagsauvos uniypydvuy SPL 0 VV 10 VV 88 V9C vv DSOGO]8 XaIDD SdudH TVINNaddg O cO 0 vv 0 0 0 0 0 1Uudx WNUYIOS iz cO 0 88 0 0 0 0 0 Nspjsnop wnuvjog © 0 SO 9°SL8 O O O 6) 6) SNIDIJUDAND SNNU1 a ie CLS 8° C8E 0) O 0 0) 0) sn14Ddoos SnloT < V9C Gel 88 0 0) 0) VV 0) DIDIUNI DI]AYAOH = OSV y9osrl 9°6S89 09 8 s8Il Vorel +9 6891 Vrii wunlipdoos wnwayjuni)aH 90 0 0°88 0 0 0 0 0 wnsoyusafuord winjjcydouy 0) 0 0 10 0) 88 CSE OLI synpa snjorqodivy SsanuHsans 0°99 VV SL 0 0 0) 0 0) psafiyJau D1IA]DS 0) O°CC Cel O 0) 0) 0 0) sida]o1sv] x1]DS CO" O 0) @) @) @) ) 0 WNIIDAIDUL SAG1lY 0 0 0 0) 0) ) rr CC SAPlOINIAA DIAMWUDIIA CO’ vv 0 80 TET Cie 19 6E 0 snssaidu snyjouvay TOl DSL CStrI Ol 6) Q°0¢E co CC 0 snjpauns snyjouvay 0) 0 O'VL 10 0) 88 88 C&l supynyid SIUDYIIVE 0 vv 0) sae V'8r ++7'067 +V'081 SOE puussiind sojkydvjsojoay cO0 2 a OeCo V?¢ OLI +4+9°8EL +P 9VT 8 Orl WNIDINIIISD{ DUOJSoUuapy vO" vv 9°6STC 0 0) 0 0) 0 DIWUALOJ]DI VISNUIIAY SaNudHs Play uiIng-jsOg ojyereyd pue yeoH PIP uiInq-jsOq oyeleyo pure yeapxy 1eoH JonU0Z Cc SUS [ OUs ‘¢ pue p SUUIN[OS UT UMOYS Ie dIY Jaye IedA JsIY OY) Ploy oy} ul pue sopdwies usng-jsod wo ddUaSIOWIY *(1S9} ao | MNS YURI SIEM -TRASNIY “GO" > d) Wao DANVSOU SOIVOIPUL — ‘JOdFJO JJeIBYO DATIISOd JUROYIUSIS & SOJOIPUL + +4 ‘JOaJo JeaYy IANISOd JUROYIUSIS eB SaJBOIPUT + “(SNINNTOD DQ aAaYHL LSalf) SLNANLVINL NaNd-dad ATMHL TY ONIGNTNYI ‘SaIdNVS ANVG Gas Lp WOT ONIONAWNY ‘IN add ALISNAG SV GdASSaadXY “SLNVNINAAD JO WAaWAN “[T ATAV tc | 199 ODION: GERMINATION OF MARITIME CHAPARRAL SEED BANK | 2000] "Soo Soo Coe oS 90° AN = oo eo o7o:S D ) i) i) CO’ Bere ANOno =) — N b en | oe) Ooo oom ooo eS ™ 00 oODN i DECI VV 0 9 LVe uiIng-]sog c SOUS VV VV 0 CLS VV 0 CLS CLS ICL 0 0 CLIT 0 88 8 OSC 0 Vel cel VV C6L 0 V9 9°19 Oy 9LI vv OCLS VV vril Cc 186 ajereyo pure yeapy = WO Semoeoeceooeeocooeomo ones OoeoneecoeeoeSeSeseeSeoeoose ee cO Pel ioe) wz Socom OoocoCcecoom eee aS (oe) — as Ne) os) Col . | I must confess, I have had a review copy of this | work on my computer for nearly a year. I have used | it numerous times almost daily while working on a checklist of the vascular plants of Colorado with | synonymy and on a list of exotic plants for Wyo- ming. Now, finally, a review. | Future plants for the Synthesis include county- | level occurrence data for the U.S., illustrations and | colored photos of all taxa available, and a random | access key to all North American species. With | Kartesz’s productivity and expertise and Mea- | cham’s “‘magic,”’ I anxiously await the appearance | of future editions, likely on DVD. —RONALD L. HARTMAN, Rocky Mountain Herbarium, | Department of Botany, University of Wyoming, Laramie © 82071-3165. “Maprono, Vol. 47, No. 3, pp. 209-216, 2000 NOTEWORTHY COLLECTIONS CALIFORNIA ARABIS PINZLIAE Rollins (BRASSICACEAE).—Mono Co.: Inyo Nat. For., Sierra Nevada, ridge ca. 1 km ENE of Two Teats, 37°43'03”N, 119°05'35”W, elev. 3200-3280 m, kru- mmholz woodland of Pinus albicaulis with Lupinus lep- idus var. lobbii, Leptodactylon pungens, Eriophyllum lan- atum var. lanatum, Eriogonum ovalifolium var. nivale, Pri- mula suffructescens, etc., one colony in fine, gravelly, vol- canic scree on SE-facing N slope just below ridgetop, 22 Aug. 1996, H. Constantine-Shull 1158 (HSC). Verified by J. D. Morefield, Nevada Natural Heritage Program. Previous knowledge. Known from 9 sites in Esmeralda Co., Nevada and 1 site in Mono Co., California at the north end of the White Mountains in granitic scree in steeply-sloped dry drainages on north to east aspects, elev. 3000-3350 m (J. D. Morefield and D. W. Taylor, Note- worthy Collections (California), Madrono 37:64—65, 1990; R. Rollins, in Hickman (ed.), The Jepson Manual, Higher Plants of California, 1993; J. D. Morefield, Status Report: Arabis pinzlae Rollins (1982, 1994), but not re- ported from the Sierra Nevada in these or any other avail- able flora (e.g., A. Howald, Vegetation and Flora of Mam- moth Mountain, Report for the Mammoth Mountain Ski Area, Mammoth Lakes, CA, 1983; M. Bagley, Sensitive Plant Survey for June Mountain and Rodeo Meadows, Re- port for the Mammoth/June Ski Resort, Inyo National For- est, Mono County, CA, 1988; N. EF Weeden, A Sierra Ne- vada Flora, 1996, etc.). Significance. First record of the species from the Sierra Nevada. A range extension of ca. 78 km WSW from the Boundary Peak region in the White Mountains. This rec- ord may suggest a possibly very recent Sierran origin for this newly developed taxon (H. Constantine-Shull, Flo- ristic Affinities of the San Joaquin Roadless Area, Inyo National Forest, Mono County, CA, 2000). Plants in the Sierran population fit A. pinzliae morphological charac- teristics; however, 2 out of the 15 plants measured had mature fruits 0.5 mm longer than expected for the taxon. Further analyses on this Sierran population may help to clarify the taxon’s distinction from A. platysperma A. Gray var. platysperma. Arabis pinzliae should be sought at additional locations in the central-eastern Sierra, es- pecially on adjacent ridges, and in the Glass Mountains ca. 38 km ENE of the Sierran population where there ap- pears to be suitable habitat. ARTEMISIA MICHAUXIANA Besser (ASTERACEAE).— Mono Co., Sierra Nevada, streamside ca. 1.75 km E of Inyo Crater Lakes, 37°42'N 119°02'W, elev. 2500 m, Abies magnifica var. magnifica forest with Pinus contorta subsp. murrayana Critchfield, in gravelly pumice soil in a wet area along a stream with Salix lemmonii, Carex nervina, Cicuta maculata var. angustifolia, Epilobium cil- tatum subsp. ciliatum, etc., 28 Jul. 1996, H. Constantine- Shull 987 (HSC); Moist pumice soil beside Deadman Creek ca. 2.25 km E of Two Teats, 37°42’N 119°04’W, elev. 2680 m, open avalanche zone with Artemisia triden- tata subsp. vaseyana, Populus tremuloides, Salix lemmon- il, Delphinium glaucum, 1 Sept. 1996 H. Constantine- Shull 1180 (HSC). Previous knowledge. In subalpine to alpine scree, talus, and drainages in the White and Inyo Mountains and in the Marble Mountains of the Klamath Region, north to B.C., Montana, Colorado (L. Abrams and R:S. Ferris, I]- lustrated Flora of the Pacific States, WA, OR, and CA, Vol. IV, 1960; PA. Munz, Supplement to A California Flora, 1968; L.M. Shultz in Hickman (ed.) The Jepson Manual, Higher Plants of California, 1993) but not re- ported from the Sierra Nevada in these floras or by any specimen recorded in the Cal Flora database, 1999. One undocumented sighting with no specimen was noted in Glacier Canyon in Yosemite (J. T. Howell, A list of the vascular plants of Tuolumne Meadows and vicinity, Sierra Club Nature Notes #13, 1944). This may be the reference to a Sierran range for this plant in A. Cronquist et al. Intermountain Flora Vol. 5—Asterales, 1994. Significance. First documented report of the species from the Sierra Nevada. A range extension of ca. 75 km from the Montgomery Peak region of the White Moun- tains. These eastern Sierran populations occur at lower elevations (2500—2680 m) than the 3000 m minimum el- evation recorded by the Jepson Manual and specimens cited in the Cal Flora database. ARTEMISIA LUDOVICIANA Nutt subsp. CANDICANS (Rydb.) Keck (ASTERACEAE).—Mono Co., Sierra Nevada, Min- aret Meadow ca. | km E of Minaret Summit, 37°39'27’N, 119°02'52”W, elev. 2690 m, in gravelly pumice soil along streamside in and below the meadow with Pinus contorta ssp. murrayana, Salix lemmonii, Lonicera involucrata vat. involucrata, Arabis holboellii var. pinetorum, 20 Aug. 1996 AH. Constantine-Shull 1224 (HSC). Previous knowledge. In dry woodland, shrubland from the northern Sierra Nevada and Modoc Plateau to Wash- ington, Montana, and Utah (L. Abrams and R. S. Ferris, Illustrated Flora of the Pacific States, WA, OR, and CA, Vol. IV, 1960; L. M. Shultz in Hickman (ed.) The Jepson Manual, Higher Plants of California, 1993; A. Cronquist et al. Intermountain Flora Vol. 5—Asterales, 1994; Cal Flora database, 1999). Significance. First report of the subspecies for Mono County. A range extension of ca. 222.6 km SSE from Donner Pass Ridge, Nevada Co., CA (Cal Flora database). —HELEN CONSTANTINE-SHULL, 825 UNION STREET, AR- CATA, CA 95521 AND JOHN O. SAWYER, JR., Dept. of Bi- ology, Humboldt State University, Arcata, CA 95521. These findings are presented in Constantine-Shull, H. 2000. Floristic Affinities of the San Joaquin Roadless Area, Inyo National Forest, Mono County, California. M.A. thesis. Humboldt State University, Arcata, CA. The thesis is also published with University Microfilms, Incorporated. A California Native Species Field Survey Form for Ar- abis pinzliae has been submitted to the Natural Diversity Data Base, California Department of Fish and Game. CALIFORNIA LIQUIDAMBAR STYRACIFLUA L.. (HAMAMELIDA- CEAE)—Sacramento County, on the north bank of the American River in the American River Parkway, a few hundred meters downriver of the Estates Drive entrance, 210 N38°33', W121°22’, 16 June 1998, J. M. Randall s.n. (DAV). A fruiting tree was found growing along the river in this semi-wild preserve bordered by residential areas. The tree appears to be a naturalized specimen, but it is possible that it was planted and then neglected. The nar- row strip of forested land along the river where the tree occurs is dominated by Quercus lobata, Populus fremon- tii, Fraxinus latifolia, and Acer negundo var. californicum and the exotic Sapium sebiferum. Prominent understory plants include Toxicodendron diversilobum and numerous exotics including Arundo donax, Conium maculatum, Raphanus raphanistrum, Rubus discolor, Saponaria offi- cinalis, and Vinca major. Previous knowledge. Native to the eastern North Amer- ica, from Connecticut to southern Illinois and south to Florida, west to Oklahoma and south again to eastern Mexico and Guatemala (H.A. Gleason and A. Cronquist 1991, Manual of Vascular Plants of Northeastern United States and Adjacent Canada. New York Botanical Garden, NY). This species, commonly called sweet gum, is com- monly cultivated as an ornamental tree, and widely rec- ognized for its unusual spiky-spherical fruits, beautiful 3 to 7-pointed leaves, and fall color. Many cultivars and hybrids have been developed with the other two species of Liquidambar that are native to Asia (A. L. Jacobson 1996. North American Landscape Trees. 10 Speed Press, Berkeley, CA). Its wood is valued for furniture, flooring and veneer. Reported as growing wild outside of its orig- inal range in a large population at one site in northeastern Illinois (E Swink and G. Wilhelm. 1994. Plants of the Chicago Region. 4" edition. Indiana Academy of Science, Indianapolis). Significance. First record in California. We know of no other reports of establishment of L. styraciflua west of its native range despite widespread cultivation of the species in low elevation areas of the western U.S. We have also been unable to find reports of L. styraciflua escaping cul- tivation on other continents. SAPIUM SEBIFERUM (L.) Roxb. (EUPHORBIACEAE).— Sacramento County, on the north bank of the American River in the American River Parkway, a few hundred me- ters downriver of the Estates Drive entrance, N38°33’, W121°22', 28 June 1998, B. A. Meyers-Rice 4#MR980603 (DAV). Approximately two dozen semi-mature trees were found growing along the river in a wildland preserve bor- dered by residential areas. Numerous seedlings have also been found on sandbars along the river. The narrow strip of forested land along the river is dominated by Quercus lobata, Populus fremontii, Fraxinus latifolia, and Acer ne- gundo var. californicum. Prominent understory plants in- clude Toxicodendron diversilobum and numerous exotics such as Arundo donax, Conium maculatum, Raphanus ra- phanistrum, Rubus discolor, Saponaria officinalis, and Vinca major. Previous knowledge. Native to China. Sapium sebifer- um (Chinese Tallow Tree) was introduced to the east coast of the USA in the late 1700s. It now occurs in every coastal state from North Carolina to south Texas, inland to Arkansas, and in Florida as far south as Tampa; over- seas it has escaped cultivation in Japan, Formosa, India, Pakistan, central and southern Europe, Martinique, and the Sudan (K. A. Bruce, et al. Natural Areas Journal, 17:255— 260, 1997). Sapium sebiferum is used in China as a source of soap and other products, and was introduced to the USA as a potential oilseed crop. In California it is used as a landscape tree and is valued for its attractive habit, MADRONO [Vol. 47 glossy green foliage that turns red in the fall, and showy} white seeds. Significance. First record in a California wildland. The} parent stock for these plants may be landscape trees from, the surrounding suburbs. Sapium sebiferum has been seen) in two other wildland locations in Placer County, at An-' telope Creek and at Strap Ravine (D. Bishop, pers.; comm.). The closest documented occurrence of Sapium sebiferum outside of cultivation is in coastal Texas, ap-| proximately 2600 km distant. This species has great po-| tential to become a serious weed in riparian forests of California’s Central Valley. Birds, especially finches and warblers, feed upon the seed and may help to spread the’ fruit (personal observation). Seed can also be transported, by water. SESBANIA PUNICEA (Cav.) Benth. (FABACEAE).—Sac- ramento County, on the margins of William B. Ponds wet-. land in the American River Parkway on the north side of | the River at the Arden Way entrance, N38°33’, Wi121°22) 28 June 1998, B. A. Meyers-Rice #MR980604 (DAV). Large numbers of plants, ranging in age from seedlings to’ mature, fruiting specimens, were found growing on the margins of the pond. This pond is a heavily developed! fishing pond within the American River Parkway’s wild- land area. Various sized specimens were also established | at the low flow edges of the American River, especially. on islands in the middle of the river. Total distribution in) the parkway is from Ancil Hoffman Park downstream to: the California Exposition floodplain. Previous knowledge. Native to South America (Argen-. tina, Brazil, Paraguay, and Uruguay). As an exotic species, in the USA, it occurs from northern Florida and southern) Georgia to eastern Texas. Previously known in California’ at only a few sites in Butte County (V. Oswald and L.) Ahart, 1994. Manual of the Vascular Plants of Butte Coun- | ty, California. California Native Plant Society. Sacramen- | to). It has been seen the area of Suisun Marsh in the Cal- i ifornia Delta, but has apparently been eradicated from this‘ location (A. Shapiro, pers. comm.). Another location in, northern Sacramento County is along Dry Creek within the Cherry Island Golf Course (R. Robison, personal com- | munication). In southern Africa it is a serious weed in) South Africa (Natal, Transvaal, and Cape Provinces), Le- | sotho, and Zimbabwe. Sesbania punicea is widely used as. an ornamental plant because of its attractive compound | leaves, bright sprays of red flowers and persistent winged | fruit. The species Sesbania tripetii is closely related, and | indeed may also be a synonymous name. The name “‘Dau- | bentonia punicea”’ is also a synonym. Significance. The first collection for Sacramento county. | Other than the Butte County locations, the closest reported | occurrence of Sesbania punicea is in far-eastern Texas. It. is unclear how this plant was transported to California. It is likely to become a serious weed in the riparian areas of | California’s Central Valley. It forms dense thickets, es- pecially in moist areas, in the southeastern USA and_ southern Africa. Its seeds are effectively transported by water. —Barry A. MEYERS-RICE, RAMONA ROBISON, JOHN M. | RANDALL, The Nature Conservancy, Wildland Invasive | Species Program, Department of Vegetable Crops and Weed Sciences, University of California, Davis, CA 95616. SONORA, MEXICO Salvia similis Brandegee (LAMIACEAE).—Municipio ‘de Guaymas, peninsula at S side of Algodones Bay, 2 km 'W of Cerro Tetas de Cabra summit, 27.9°N, 111.0°'W, ‘north-facing granite slope, elev. 5-20 m, desertscrub with ‘Stenocereus thurberi, Jatropha cuneata, Bursera micro- ‘phylla, Fouquieria diguetii; uncommon shrubs, about | m \tall, 17 March 1983, Burgess 6361 et al. (ARIZ). | Significance. First record for mainland Mexico. | Previous knowledge. Otherwise known from Baja Cal- ‘ifornia Sur (I. L. Wiggins, Jn, E Shreve & I. L. Wiggins, ‘part 2, 1964 Stanford Univ. Press) where it is rather wide- ‘spread and Isla San Pedro Nolasco (R. S. Felger and C. -H. Lowe, Natural History Museum of Los Angeles Coun- ty, Contributions in Science 285, 1976). | Notes. Apparently rare on the mainland although com- ‘mon on Isla San Pedro Nolasco on east-facing slopes near ‘the summit. Both of these populations are on north- and ‘east-facing granite slopes above the sea; there are very ‘few mainland habitats where these conditions are dupli- ‘cated. ERAGROSTIS SPICATA Vasey (POACEAE).—Municipio de Guaymas: Bahia San Carlos, Creston area, ca. 3 m elev., disturbed habitat, 8-10 m N of highway, among Prosopis glandulosa, perennial, forming dense clumps ca. 1—1.5 m tall, colony of dozen or so plants, 18 November 1984, Felger 84—544 & Valdez Zamudio (ARIZ, MEXU, TEX); 20 March 1986, Felger 86-67 & Sanders (ARIZ). Mex Hwy 15, 1.7 mile SE of Pitahaya (Belem, Rio Yaqui) junc- tion, elev. 10 m, 27°45'N, 110°24’W, coastal thornscrub, pond near roadside, densely shaded by Prosopis glandu- losa; robust grass to 1.8 m tall, 11 October 1985, Felger 55-1248 & Reichenbacher (ARIZ, MEXU). Mex Hwy 15, 3.6 mile S of Pitahaya (Belem, Rio Yaqui) junction, ca. 10 m, 27°42%'N, 110°22'W, in shade of mesquites; com- mon robust perennial to | m tall, 11 October 1985, Felger 85-1248 & Reichenbacher (ARIZ, MEXU). Municipio de Hermosillo: Hwy 24, 5.0 mile N of Sahuaral (4.7 mile N of Bahia San Agustin Rd junction), elev. 5 m, 28°23'N, 111°21'W; low lying, temporarily swampy area; fine-tex- tured silty-muddy soil; shallow standing water in lowest areas; Shade of mesquite; common, 12 October 1985, Fel- ger 85-1586 & Reichenbacher (ARIZ, MEXU). Significance. First record for Sonora. Previous knowledge. Texas, northeastern Mexico, Baja California Sur, and Argentina and Paraguay (F W. Gould and R. Moran, San Diego Society of Natural History Memoir 12, 1981; E O. Zuloaga, et al. Monographs in Systematic Botany from Missouri Botanical Garden 47, 1994). Notes. This large, perennial grass is a common and con- sistent element in the grassy, savanna-like swampy habi- tats of coastal west-central Sonora between the Rio Yaqui and Empalme (southeast of Guaymas), with outlier pop- ulations at San Carlos north of Guaymas and near Sa- huaral (east of Tastiota). The plants are reproductive at least in March and November. In west-central Sonora it often grows beneath mesquite (Prosopis glandulosa var. torreyana) with Kosteletzkya hispidula, Luffa operculata var. intermedia, Phyllanthus evanescens, Sesbania her- bacea, and grasses including Echinochloa crusgalli, Lep- tochloa fusca ssp. uninervia, L. panicea ssp. brachiata, L. viscida, Panicum hirticaule and Sporobolus airoides. PORTULACA JOHNSTONI J. Henrickson (PORTULACA- CEAE).—Municipio de Guaymas, 0.5 km W of Estero Soldado at ca. 1 km inland from shore (ca. 6 km E of NOTEWORTHY COLLECTIONS 211 Bahia San Carlos), ca. 2 m elev., coastal desertscrub, sandy soil, locally common, 18 November 1984, Felger 84-421 & Valdez Zamudio (ARIZ, MEXU, TEX). Significance. New record for Sonora and the Sonoran Desert. Previous knowledge. Known only from the type collec- tion in Coahuila in the Chihuahuan Desert (J. Henrickson, Madrono 28:78—79, 1981). Notes. No differences were noted between the Sonoran and Chihuahuan plants. The Sonora collection was made in an area of natural vegetation but near disturbed habitats in an area rapidly being urbanized. Immature seeds are reddish (rust-colored) throughout, the body becomes iri- descent blackish with maturity. Due to the radiating fim- briae on the seeds, unique in the genus, this plant is hereby given the common name ‘Punk Portulaca’ (Fig. 1). It would be interesting to study the origin and development of these fimbriae; the plants are otherwise similar to P. oleracea. XIMENIA PARVIFLORA Benth. var. GLAUCA DeFilipps (OLA- CACEAE).—Municipio de Guaymas: Canon La Pintada, large riparian canyon ca. 4 mile E of La Pintada (ca. 33 mile S of Hermosillo on Mex Hwy 15), riparian deserts- crub, on slopes of canyon side, not common, flowers dull yellow, Felger 3267 (ARIZ). Broad spiny shrubs ca. 1.5 m tall, apparently evergreen, herbage with a reddish cast; flowers dull yellow, the calyx red, the petals densely pu- bescent inside with many whitish and flattened hairs; flow- ering late May. Significance. First record for this genus in Sonora. Previous knowledge. Known only from Baja California Sur. Notes. A unique feature of this genus is the corolla, comprised of 4 or 5 free petals densely covered inside with hairs. These hairs are said to be brownish and barbed; in the Sonoran specimen the hairs are whitish when fresh but become brownish with age, and I do not find barbed hairs on these specimens, or those from Baja California Sur, or the half dozen specimens of X. americana at ARIZ. A previous collection from Baja California Sur (John- ston 3718; 1. M. Johnston, Proceedings, California Acad- emy of Sciences (4) 12:951—1218, 1924), misidentified as X. pubescens seems to be the source of the erroneous ref- erence to X. pubescens in Baja California (1. L. Wiggins, Flora of the Sonoran Desert, Joc. cit.; 1. L. Wiggins, Flora of Baja California, 1980 Stanford Univ Press). —RICHARD STEPHEN FELGER, Drylands Institute, PMB 405, 2509 North Campbell Avenue, Tucson, AZ 85719. ARIZONA AND SONORA VAUQUELINIA CALIFORNICA Subsp. SONORENSIS Hess & Henrickson (Rosaceae).—SONORA: Municipio de Gen. Plutarco Elias, Sierra Cubabi, middle-elevation on north- facing drainage; on granite with Dodonaea viscosa, Er- iogonum sp., Galium stellatum, Solanum hindsianum, vi- cinity 31°43'25"’N, 112°50’50”W, elev. ca. 750 m, shrub ca. 8 ft tall, 11 November 1998, Pate s.n. (ARIZ). [Sierra Cubabi] N 35 11 000 E, 3 27 500, ca. 700 m SE of highest point in area, 1130 m elev., granite with Stenocereus thur- beri, Justicia californica, Ferocactus covillei, Eriogonum wrightii, Encelia farinosa, scattered locally in protected areas, 19 March 1991, Baker 8130 & Johnson (ASU).— ARIZONA: Maricopa County, Barry M. Goldwater Air- force Range, Sand Tank Mountains, 32°39'43.3"N, 2A 112°19'51.2”W, 2800 ft, NE-facing steep rhyolite slope; uppermost elevational limit of Sonoran Desert including Acacia greggii, Anisacanthus thurberi, Calliandra erio- phylla, Carnegiea gigantea, Cercidium microphyllum, Coursetia glandulosa, Encelia farinosa, Ephedra aspera, Eriogonum fasciculatum, Fouquieria splendens, Lycium parishii, Olneya tesota, Opuntia chlorotica, Prosopis vel- utina, Viguiera parishii, Yucca arizonica, 10 October 1995, Felger 95-337, Wilson, Smith, & Speich (ARIZ). Sand Tank Mountains, 2 mile SW of Squaw Tit Peak, along rocky drainage with Juniperus, 3200 ft, 1 January 1995, Malusa s.n. (ARIZ). Previous knowledge. Documented only from the Ajo Mountains in Organ Pipe Cactus National Monument, Pima County, Arizona, and disjunct in local areas on the eastern slopes of the Sierra Juarez in northern Baja Cali- fornia (W. J. Hess and J. Henrickson, Sida 12:101-163, 1987). Significance and natural history notes on the species in Sonora. First record for this distinctive subspecies from mainland México and the second documentation of an Ar- izona population. Sierra Cubabi in Sonora is the nearest range to the Ajo Mountains, and ca. 30 km directly south- southwest from it. The Sand Tank Mountains in Arizona are ca. 54 km northeast of the Ajo Mountains. The Sonora and Ajo Mountain populations are within the Sonoran Desert, the Sand Tank Mountain population is mostly at the upper elevational limit of the desert. All three moun- tains support Sonoran Desert species to their summits. A locality mapped by Turner, et al. (Sonoran Desert Plants: an ecological atlas, 1995, Univ. of Arizona Press) as a “sighting”? west of the Ajo Mountains, in Yuma County, is presumed to be erroneous. Of the four subspecies of V. californica, only subsp. pauciflora was known for certain from Sonora (Hess and Henrickson, 1987). In Sonora it is documented from can- yons and slopes at the north end of the Sierra el Tigre, at 1140 m, where it is apparently quite rare. The several collections at ARIZ are probably taken from the same one or two shrubs. The substrate is rhyolite although limestone intrusions occur nearby. Here Vauquelinia occurs at the upper margin of Sonoran desertscrub merging to thorn- scrub and just below the oak zone (Quercus oblongifolia). Associated plants include Coursetia glandulosa, Dodon- dea viscosa, Fouquieria splendens, Fraxinus gooddingii, Hechtia montana, Juglans major, Prosopis velutina, Rhus microphyllum, and Yucca arizonica. Elsewhere this sub- species is known from limestone substrate in ecotone of Chihuahuan desertscrub and oak woodland, and lower el- evations in oak woodland. Subspecies californica occurs in both states of Baja California and southern Arizona including the Baboqui- vari Mountains near the Sonora border. It should be sought in nearby north-central Sonora. An observation of rose- wood in the Sierra del Viejo near Caborca (Turner et al. 1995, loc. cit.) may be this subspecies or subsp. sonoren- SUS. While admiring V. californica subsp. sonorensis on a field trip of the Arizona Native Plant Society to Alamo Canyon in the Ajo Mountains, Felger suggested that it should be sought in mountains in northwestern Sonora east of Sonoyta such as the Sierra Cubabi, mountains that have scarcely been botanically explored. Pate replied that she had indeed seen it there, and soon thereafter verified it with the record cited here. Subsequently Felger located the Baker specimen at ASU. The paucity of records for V. californica in northern Sonora seems unusual given its widespread occurrence MADRONO | ; [Vol. 47. and diversity (three subspecies) in adjacent southern Ari-. zona. In view of the discovery of the Sierra Cubabi pop- ulation, it seems that absence of records in much of north- ’ ern Sonora for many other species likewise may be due} to a lack of botanical exploration in remote and now often dangerous borderland territory. Distant areas are often far | better known (e.g., Martin, Yetman, Fishbein, Jenkins, | Van Devender, and Wilson, 1998, Gentry’s Rio Mayo | Plants, University of Arizona Press). Hess and Henrickson (1987) give a maximum size of 10 m for any Vauquelinia, 8 m for any of the four sub- | species of V. californica, and 7 m for subsp. sonorensis. A tree in Alamo canyon, in the Ajo Mountains, carefully measured by Robert Zahner and associates (National Reg- | ister of Big Trees, 1996, American Forests, Washington, | D.C.) was 14.3 m (47 ft) in height with an average crown | spread of 12.2 m (40 ft), and 2.0 m (78 inch) in girth at 1.4 m (4.5 ft) above ground level (original measurements | in English units). Thus the most xeromorphic taxon (Hess | and Henrickson 1987) in the genus contains the largest- sized individual. | | —RICHARD FELGER, Drylands Institute, PMB 405, 2509 © North Campbell Avenue, Tucson, AZ 85719, and Ami | Pate, Organ Pipe Cactus National Monument, Rt 1 Box 100, Ajo, AZ 85321. OREGON MYRIOPHYLLUM USSURIENSE (Regel) Maxim. (HALORA- GACEAE).—Clatsop Co.: Columbia River, Cathlamet Bay, Lewis and Clark National Wildlife Refuge, Russian - Island, 4 miles WNW of Knappa, T8N, R8W, Sec. 11. On mud along tidal channels, subject to daily freshwater tidal | inundation. 14 August 1992. J.A. Christy 8205 (MO, NY, — OSC, V). Previous knowledge. Taiwan, China, Japan, Russian Far East, British Columbia. First reported from North America by Ceska et al. (Brittonia 38:73-81, 1986), it is known from over a dozen sites in southern British Columbia, in- cluding Vancouver Island, with the earliest collection dat- ing from 1916. This is one of two collections from the estuary of the Columbia River of Oregon and Washington. Both sexes of M. ussuriense are present in British Colum- bia, but no flowers were found in either of the U.S. pop- ulations. We consider it to be a rare native species with an amphi-Beringian distribution. Significance. New to the United States; new to Oregon. —JOHN A. CHRISTY, Oregon Natural Heritage Program, 821 SE 14th Ave., Portland, OR 97214; OLDRISKA CESKA and ADOLF CESKA, PO. Box 8546, Victoria, B.C. V8W 3S2, Canada. W ASHINGTON MYRIOPHYLLUM USSURIENSE (Regel) Maxim. (HALORA- GACEAE).—Wahkiakum Co.: Columbia River, Julia But- ler Hansen National Wildlife Refuge, small wetland across Steamboat Slough from Price Island, TIN, R6W, Sec. 16. On mud along Steamboat Slough, subject to daily fresh- water tidal inundation. 29 July 1992. J. A. Christy 8164 (WSU). Previous knowledge. See report above for Oregon. _ 2000] Significance. New to the United States; new to Wash- ington. —JOHN A. CurRISTy, Oregon Natural Heritage Program, 821 SE 14th Ave., Portland, OR 97214; OLDRISKA CESKA and ADOLF CESKA, P.O. Box 8546, Victoria, B.C. V8W 382, Canada. OREGON CAREX LONGI Mack. (Cyperaceae).—Clatsop Co., wet, peaty cranberry field, 1.6 air km SSW of Cullaby Lake, Delmoor Loop Rd., elev. 4 m, T7N RIOW S27, 2 Oct. 1999, Zika 14470 (MICH, OSC, WTU). Previous knowledge. Native to eastern North America, west to Texas. A member of the Carex straminea complex in C. section Ovales. For an identification key to the spe- cies, see Rothrock et al. (Canadian Journal of Botany 75: 2177-2195, 1997). Long’s sedge is a common adventive in cultivated cranberries in New England. The seeds of this and other cranberry weeds were apparently introduced on the Pacific coast by the transport of Vaccinium macro- carpon vines between agricultural areas. Significance. First report for Oregon. CHAENOMELES SPECIOSA (Sweet) Nakai (ROSACEAE).— Jackson Co., spreading from cultivation to roadside ditch- es, ca. 25 seen, Applegate River valley ca. 12 air km NNE of dam at Applegate Reservoir, elev. 546 m, T39S R3W $28, 23 May 1991, Zika 11122 (OSC). Previous knowledge. Native to eastern Asia, and an oc- casional weed in eastern North America, W to Wisconsin. Flowering quince is a common ornamental in western OR. Significance. First record as an escape from cultivation in Oregon. COTONEASTER HORIZONTALIS Decne. (Rosaceae).—Lane Co., naturalized in Willow Creek Natural Area, with Frax- inus, Populus, Crataegus, West Eugene, elev. 122 m, T18S R4W S3, 9 July 1998, Voss 1936 (OSC); Tillamook Co., bird-sown weed in grassland, S aspect, Cascade Head, elev. 275 m, T6S R11W S14, 14 Aug. 1986, Zika 9989 (WTU). Previous knowledge. Native to China and planted for ornament W of the Cascades. Rock cotoneaster fruits heavily in autumn, and birds such as American robins (Turdus migratorius) are commonly observed eating the colorful fruit and dispersing the seed of this and other Cotoneaster species. Significance. First report as a wild plant in Oregon. COTONEASTER LACTEUS W. W. Sm. (ROSACEAE ).—Cur- ry Co., bird-sown weed in roadside thickets, Route 101, Winchuck, T41S RI3W S23, 24 Oct. 1990, Zika 11024 OSC; roadside thickets, Route 101, Brookings, T41S R13W S5, 24 Oct. 1990, Zika 11033 (OSC); Lane Co., weed in Quercus garryana woods, Morse Ranch, Eugene, elev. 120 m, 3 Dec. 1999, R. Love 9962 (WTU). Previous knowledge. Late cotoneaster is native to Chi- na, commonly planted W of the Cascades, and known to be invasive. Significance. First report as a garden escape for Oregon. COTONEASTER SIMONSII Baker (Rosaceae).—Coos Co., edge of coniferous woods, Route 101 near Saunders Lake, elev. 24 m, T23S R13W S35, 8 Sept. 1999, Zika 14322 (WTU); Curry Co., roadside thicket, Route 101, Win- chuck, T41S R13W S23, 24 Oct. 1990, Zika 11025 OSC; NOTEWORTHY COLLECTIONS 213 Tillamook Co., steep bank, Route 101, Nehalem, elev. 15 m, T3N R1OW S27, 17 Sept. 1999, Zika 14365 (WTU). Previous knowledge. Native to the Himalayas, Hima- layan cotoneaster is commonly planted as an ornamental in western Oregon, Washington, and British Columbia. Significance. First report for Oregon as a garden escape. ELEOCHARIS QUADRANGULATA (Michx.) Roem. & Schult. (CYPERACEAE).—Lane Co., large clone, long-estab- lished but local weed, wetland clay soil, full sun, with Ludwigia palustris, Eleocharis palustris, Juncus margin- atus, Ventenata dubia, degraded wet prairie by abandoned airport runway and old race track, West Eugene, elev. 117 m, T17S R4W S833, 8 July 1997, Alverson & Zika 13225 (OSC). Previous knowledge. Native to eastern North America, west to Texas. Square-stemmed spike-rush is considered native in California. Significance. First report for Oregon, but surely adven- tive, not native, on this site with a history of decades of disturbance in the industrial zone of Eugene. HYPERICUM MAJUS (A. Gray) Britton (CLUSIACEAE).— Clatsop Co., cranberry field, with Juncus canadensis, Del- mar Loop Rd., elev. 4 m, T7N RIOW S22, 2 Oct. 1999, Zika 14450 (OSC, WTU); Coos Co., cranberry field, with Juncus pelocarpus, Randolph Rd., 6 km N of Bandon, elev. 52 m, T28S RI4W S4, 7 Sept. 1999, Zika 14249 (WTU). Previous knowledge. Greater Canadian St. Johnswort is native to wetlands across northern North America, includ- ing British Columbia & Washington. However, a number of Pacific coast populations are adventive, in gravel pits, railroad yards, and cranberry farms, where they were un- doubtedly introduced from eastern cranberry states, along with Hypericum boreale, H. canadense, H. ellipticum, and Triadenum fraseri. Significance. First report for Oregon. All known popu- lations are weeds in agricultural settings, and are adven- tive, not native. LUZULA ARCUATA (Wahlenb.) Swallen subsp. UNALAS- CHCENSIS (Buchenau) Hultén (Juncaceae).—Hood River Co., wet sunny mossy banks of Lost Lake, elev. ca. 1000 m, 29 June 1924, Henderson 778 (ORE). Previous knowledge. Circumboreal and native, curved woodrush is known from collections in Washington on Mt. Rainier and Mt. Adams, 85 km N. Significance. First report for Oregon. LUZULA FORSTERI (Sm.) DC. (SJUNCACEAE).—Marion Co., lawn, Salem, Apr. 1910, Peck 5135 (WILLU). Previous knowledge. Southern woodrush is native to Europe. Significance. First report for Oregon. Perhaps only a waif, but should be sought in the Salem area again. POLYGONUM SAGITTATUM L. (POLYGONACEAE).— Clatsop Co., common, ditches and marshy borders of cul- tivated cranberry field, with Carex chordorrhiza, Juncus brevicaudatus, Delmar Loop Rd., elev. 4 m, T7N RI1OW S22, 2 Oct. 1999, Zika 14459 (OSC, WTU). Previous knowledge. Native to eastern North America, west to Colorado. Arrow-leaf tearthumb is weedy in cul- tivated cranberry fields in Massachusetts (Sears et al. An Illustrated Guide to the Weeds of Cranberry Bogs in Southeastern New England, 1996). Significance. First report for Oregon. ZANTEDESCHIA AETHIOPICA (L.) Spreng. (ARACEAE).— Curry Co., steep W aspect, base of eroding sea bluffs, 214 with Holcus lanatus, Equisetum telmateia, Nesika Beach, elev. 5—30 m, T35S R1IS5W S36, 20 May 1997, Zika 13085 (OSC). Previous knowledge. Native to Natal, and a weed in California. Altar lily was originally planted on a blufftop as an ornamental, and subsequently dropped downslope as erosion undermined garden areas. Persisting for many years and spreading across a sandy precipice, despite con- trol efforts. Significance. First report as an escape from cultivation in Oregon. W ASHINGTON BERBERIS DARWINII Hook. (BERBERIDACEAE).— Grays Harbor Co., steep shrubby slope, with dense Rubus armeniacus, Lonicera involucrata, Rubus spectabilis, N aspect, above Route 105 near Bigelow Rd., S of Aberdeen city limits, elev. 15 m, T17N ROW S20, 25 April 1998; Zika 13422 (WTU). Previous knowledge. Native to Chile and commonly cultivated at low elevations in western Oregon and Wash- ington. Darwin’s barberry is naturalized on the coast of Coos Co., Oregon, 400 km to the south. Significance. First record as a wild plant in Washington. CALLUNA VULGARIS (L.) Hull (ERICACEAE).—Pacific Co., sandy cranberry field, with Lythrum portula, Bidens tripartita, Pioneer Rd., 2.5 km NE of Long Beach, elev. 5m, TION R11W S10, 31 Aug. 1999, Zika 14200 (WTU). Previous knowledge. A common ornamental west of the Cascades in the Pacific Northwest. Heather is a weed on peaty soils and in cranberry farms in S British Columbia, 300 km N. Significance. First report as an escape from cultivation in Washington. CAREX LONGII Mack. (CYPERACEAE).—Grays Harbor Co., cranberry fields and drainage ditches, with Juncus effusus, | km N of County Line Rd., elev. 5 m, TISN R11W S7, 29 Sept. 1998, Zika 13641 (WTU); Pacific Co., ditch in cranberry field, with Lysimachia terrestris, Pio- neer Rd. 7 km N of Ilwaco, elev. 5 m, TION RIIW S10, 12 Sept. 1998, Zika 13592 (WTU); peaty disturbed ground, with Ledum groenlandicum, near Jim Street, 2 km NE of Seaview, elev. 5 m, TION RI1W S22, 31 Aug. 1999, Zika 14193 (WS). Significance. First report for Washington. COTONEASTER DIELSIANUS E. Pritz. ex Diels (ROSA- CEAE).—King Co., Thuja hedge, campus of Univ. of Washington, Seattle, elev. 25 m, T25N R4E S16, 6 Nov. 1999, Zika 14707 (WTU); Kitsap Co., woods near New Brooklyn Rd., Bainbridge Is., Puget Sound, elev. <100 m, T25N R2E S21, 28 Sept. 1999, Zika 14426 (US, WTU); Pacific Co., sandy edge of coniferous woods, NE of Black Lake, elev. 5 m, TION RIIW S28, 30 Sept. 1999, Zika 14436 (WTU). Previous knowledge. Diels’ cotoneaster is native to Chi- na, and planted for its ornamental fruit west of the Cas- cades. Significance. First report as a naturalized plant in Wash- ington. COTONEASTER FRANCHETH BOIS (ROSACEAE).—Grays Harbor Co., thickets, with Picea sitchensis, Rubus armen- iacus, Route 105, E of Ocosta, elev. 10 m, TI6N R11W $22, 19 Oct. 1998, Zika 13650 (WTU); King Co., thickets, MADRONO [Vol. 47 | Burbank Park, Mercer Is., Lake Washington, elev. 6 m, T24N RSE S6, 6 Oct. 1999, Zika 14537 (WTU); San Juan Co., edge of forest, near pond on Turn Point, with Sym- phoricarpos, San Juan Is., Puget Sound, elev. 8 m; T35N | R2W S18, 24 Oct. 1999, Zika 14643 (WTU). | Previous knowledge. Native to China, commonly cul- | tivated west of the Cascade Mtns. in the Pacific States. | Franchet’s cotoneaster is naturalized in western Oregon | and coastal California. | Significance. First record as a garden escape in Wash- | ington. | COTONEASTER LACTEUS W. W. Sm. (ROSACEAE).— — Grays Harbor Co., gravel roadbank, Route 109, NW of | Chenois Cr., elev. 20 m, TI8N R1I1W S15, 4 Oct. 1999, Zika 14517 (WTU); King Co., cracks in asphalt parking | | lot, Mercer Middle School, S Oregon St., Seattle, elev. 30 | m, T24N R4E S16, 25 Sept. 1999, Zika 14413 (WTU). | Significance. First report as a garden escape for Wash- | ington. COTONEASTER REHDERI Pojark. (ROSACEAE).—King | Co., shade of Pseudotsuga, Alder Crest School, 195th St | NE, elev. < 50 m, T26N R4E S4, 25 Sept 1999, Zika | 14414 (WTU); Kitsap Co., woods near Gazzam Lake, | Bainbridge Is., Puget Sound, elev. 90 m, T25N R2E S829, | 28 Sept 1999, Zika 14424 (OSC). | Previous knowledge. Bullate cotoneaster is native to China, and introduced as an ornamental west of the Cas- © cades. | Significance. First report as a wild plant in Washington. COTONEASTER SIMONSI] Baker (ROSACEAE).—Grays | Harbor Co., roadsides, with Alnus rubra, Tsuga hetero- phylla, Route 105 E of Ocosta, elev. 15 m, TI6N R11W S11, 16 Sept. 1999, Zika 14341 (WTU); King Co., open forest, Lincoln Park, 0.5 km N of Point Williams, Seattle, elev. 45 m, T24N R3E S26, 14 Sept. 1999, Zika 14325 (WTU); Kitsap Co., woods, near New Brooklyn Rd., Bainbridge Is., Puget Sound, elev. <100 m, T25N R2E S21, 28 Sept. 1999, Zika 14425 (OSC, WTU); Pacific Co., roadside, Jacobson Rd, Heather, elev. 5 m, TISN R11W S30, 1 Oct. 1999, Zika 14443 (WTU). Significance. First report for Washington as a garden escape. GLYCERIA CANADENSIS (Michx.) Trin. (POACEAE).— Grays Harbor Co., cranberry fields and ditches, with Jun- cus effusus, Evergreen Park Rd., elev. 5 m, TISN R11W S7, 29 Sept. 1998, Zika 13646 (WTU); Pacific Co., ditches with Leersia oryzoides, Jim Street, 4 km N of Ilwaco, elev. 5 m, TION RIIW S22, 12 Sept. 1998, Zika 13591 (WTU); ditches, N of Black Lake, elev. 5 m, TION R11W S28, 29 Sept. 1998, Zika 13636 (WTU). Previous knowledge. Native to eastern North America, west to Minnesota. Collected in ‘‘a cranberry marsh” in adjacent Clatsop Co., Oregon in 1929 (Henderson 11841 ORE), 30 km SE, where in 1999 it was a well established weed. Rattlesnake grass is also recorded as a rare weed in southern British Columbia. Significance. First collection for Washington. HELLEBORUS FOETIDUS L. (RANUNCULACEAE).—San Juan Co., common in meadow with Festuca arundinacea, Pteridium, Rubus armeniacus, near False Bay, San Juan Is., Puget Sound, elev. 20 m, T34N R3W S4, 30 May 1999, Zika 13766 (WTU). Previous knowledge. Stinking hellebore is an ornamen- tal native to Europe, planted west of the Cascade Mtns. in the Pacific Northwest. 2000] Significance. First record as an garden escape in Wash- ington. HYPERICUM BOREALE (Britton) E. Bickn. (Clusiaceae).— Grays Harbor Co., cranberry fields and drainage ditches, Evergreen Park Rd., elev. 5 m, TISN RI1IW S7, 29 Sept. 1998, Zika 13640 (WTU); bulldozed field, near Hogan Rd., North Bay, elev. 3 m, TI8N R1I1W S17, 4 Oct. 1999, Zika 14484 (WTU); Pacific Co., sandy banks near cran- berry fields, Pioneer Rd., elev. 5 m, TION R11W S10, 12 Sept. 1998, Zika 13594 (WTU); swale between sand dunes, marine beach near 10th St., Long Beach, elev. 2 m, TION RIIW S17, 1 Oct. 1999, Zika 14446 (WTU). Previous knowledge. Native to eastern North America, as far west as Minnesota. Northern St. Johnswort is a weed associated with cranberry agriculture on the Oregon coast in Coos and Curry Cos., 300 km to the south. Significance. First record for Washington. HYPERICUM CANADENSE L. (CLUSIACEAE).—Grays Harbor Co., cranberry fields and drainage ditches, with Juncus effusus, Evergreen Park Rd. elev. 5 m, TIS5N R11W S7, 16 Sept. 1999, Zika 14348 & Weinmann (WTU); ditch, Burrow Rd., N Bay, elev. 3 m, TI8N, R11W S17, 4 Oct. 1999, Zika 14483 (WTU); Pacific Co., sandy cranberry field, near Pioneer Rd., elev. 5 m, TION R11W S9, 30 Sept. 1999, Zika 14433 (WTU); ditches between cranberry fields, with Hypericum anagalloides, | km E of Long Lake, elev. 6 m, TISN R1I1IW S17, 19 Oct. 1998, Zika 13667 (WTU). Previous knowledge. Canada St. Johnswort is native to eastern North America as far west as Manitoba, and ad- ventive on cranberry farms on the Oregon coast in Coos and Curry Cos., 300 km to the south. Significance. First record for Washington. HYPERICUM ELLIPTICUM Hook. (Clusiaceae).—Grays Harbor Co., cranberry field, with Equisetum arvense, Cranberry Rd., Grayland, elev. 5 m, TISN R11W S6, 16 Sept. 1999, Zika 14342 (UC, WS, WTU); Pacific Co., moist sandy ground, cranberry field, with Oenanthe, Heather Rd., Heather, elev. 5 m, TISN R1I1W S20, 16 Sept. 1999, Zika 14353 (OSC, US, WTU). Previous knowledge. Pale St. Johnswort is native to eastern North America, west to North Dakota. Significance. First report for Washington. HYPERICUM MUTILUM L. (CLUSIACEAE).—King Co., shores of Phantom Lake, with Scutellaria lateriflora, Ly- simachia thyrsiflora, Juncus balticus, Bellevue, elev. 75 m, T24N RSE S2, 15 Oct. 1999, Zika 14605 & Weinmann (WTU); Skagit Co., wet bank, with Typha latifolia, My- osotis laxa, Potentilla palustris, Gandy Lake outlet creek, 5 km NW of Concrete, elev. ca. 245 m, T36N R8E S32, 1 Aug. 1989, Naas 5536 (WTU); Gandy Lake, opening in cattail marsh, elev. ca. 250 m, 24 Aug. 1973, Naas & Cheney 2751 (WTU). Previous knowledge. Native to eastern North America as far west as Oklahoma. Dwarf St. Johnswort is recorded as a weed at low elevations 900 km to the south, in Butte and Glenn Cos., California. Skagit Co. records were pre- viously identified as H. majus. Significance. First report for Washington. JUNCUS CANADENSIS J. Gay ex Laharpe (Juncaceae).— Clallam Co., Ericsons Bay, Lake Ozette, elev. 10 m, T30N RISW S8, 3 Aug. 1986, Buckingham et al. 3787, & Ceska 20607 (ONP) [herbarium of Olympic National Park]; Grays Harbor Co., cranberry fields, with Vaccinium ma- crocarpon, 1 km SE of Horseshoe Lake, elev. 5 m, TI5N NOTEWORTHY COLLECTIONS 215 R11W S6, 19 Oct. 1998, Zika 13658 (WTU); damp bull- dozed ground, with Juncus supiniformis, near Burrow Rd., North Bay, elev. 3 m, TI8N RIIW S17, 4 Oct 1999, Zika 14486 (WTU); Pacific Co., ditch near cranberry fields, with Glyceria canadensis, Jim Street, elev. 5 m, TION RIIW S22 W1/2, 12 Sept. 1998, Zika 13588 (WTU); cranberry fields, with Potentilla pacifica, 0.8 km E of Long Lake, elev. 5 m, TISN RIIW S17, 19 Oct. 1998, Zika 13670 (WTU); Skagit Co., Sphagnum mat, shore of Summer Lake, with Sarracenia purpurea, Eriophorum virginicum, Vaccinium oxycoccus, elev. 200 m, T33N RSE S21, 27 Sept. 1999, Zika 14419, Weinmann & Weinmann (MICH, WTU); small pond ca. 0.2 km N of Summer Lake, elev. 200 m, 27 Sept. 1999, Zika 14423, Weinmann & Weinmann (WTU). Previous knowledge. Native to eastern North America, west to Minnesota. Canada rush is known as a weed in wetlands and in cranberry fields in British Columbia and in Coos and Curry Cos., Oregon. Recent reports of J. brevicaudatus from Washington (Buckingham ef al., Flora of the Olympic Peninsula, 1995) are based on collections of J. canadensis. Significance. First documented report for Washington. JUNCUS DIFFUSISSIMUS Buckley (JUNCACEAE).—Cow- litz Co., moist sand flats, Cowlitz R., with Phalaris arun- dinacea, Salix sitchensis, Longview, elev. 2 m, T7N R2W S11, 26 Sept. 1998, Zika 13624 (WTU); sandy shoreline, Cowlitz R., N end of Castle Rock, 18 July 1994, Kollock & Wilson s.n. (OSC, WTU); moist gray sand, Toutle R., with Juncus bolanderi, 5 km N of Castle Rock, elev. 24 m, TION R2W S827, 15 Sept. 1998, Zika 13614 (WTU); cobble shore, S Fork Toutle R., 1.8 air km E of Toutle, elev. 137 m, TION RIE S29, 20 Oct. 1998, Zika 13671 (WTU). Previous knowledge. Native to the eastern United States, west to Kansas. Known as a weed in Sacramento Valley of California, 800 km to the south. Discovered in the Castle Rock area by Loverna Wilson and Kathleen Kollock in 1994. Significance. First report for Washington. Juncus dif- fusissimus successfully colonized riverine sand and ash deposits from the 1980 eruption of Mt. St. Helens. Slim- pod rush is now frequent on volcanic debris from the mouth of the Cowlitz R. upriver to the shores of S Fork Toutle R. However, the linear population continues up- stream of the ash deposits. This distribution suggests the population was originally introduced on private logging lands on the upper tributaries of S Fork Toutle R. At pres- ent J. diffusissimus is absent from suitable habitat on other tributaries in the Cowlitz R. basin. JUNCUS PELOCARPUS E. Meyer (Juncaceae).—Grays Har- bor Co., cranberry fields, Blake Rd., Grayland, elev. 5 m, TIS5N RIIW S6, 19 Oct. 1998, Zika 13659 (WTU), dis- turbed damp sandy ground, near Hogan Rd., N shore of N Bay, elev. 3 m, TI8N R1IIW S17, 4 Oct. 1999, Zika 14508 (WTU); Pacific Co., moist sandy ground, near ir- rigation pond, Cranberry Rd., elev. 5 m, TIIN R11 W S34, 12 Sept. 1998, Zika 13597 (WTU); ditches and cranberry fields, 0.8 km E of Long Lake, elev. 6 m, TISN RI1W S17, 19 Oct. 1998, Zika 13669 (WTU). Previous knowledge. Native to eastern North America, as far west as Minnesota. Brown-fruited rush was first recorded in 1958 as a cranberry weed in Coos Co., OR, 300 km to the south. Significance. The first report for Washington. LONICERA PILEATA Oliv. (CAPRIFOLIACEAE ).—Clal- 216 lam Co., E end of Lake Crescent, bird-sown shrubs scat- tered in forest and at edge of clearings, with Tsuga het- erophylla, Alnus rubra, elev. 195 m, T30N ROW 828, 22 Nov. 1997, Zika 13408 (WTU). Previous knowledge. Box-leaved honeysuckle is native to China and commonly planted west of the Cascades. Significance. First report as a garden escape in Wash- ington. SALIX PURPUREA L. (SALICACEAE).—Wahkiakum Co., mouth of Elochoman R., N shore, scattered on sand spit, with S. sessilifolia, elev. 3 m, TON R6E S828, 4 June 1999, Zika 13779 (CAN, WTU). Previous knowledge. Basket willow is native to Eurasia, and occasionally planted by weavers. It is weedy in east- ern North America, W to Colorado. Significance. First report as a wild plant in Washington. TRIADENUM FRASERI- (Spach) Gleason (CLUSI- ACEAE).—Pacific Co., cultivated cranberry field, 1 km N of Black Lake, elev. 5 m, TION R11W S28, 31 Aug. 1999, Zika 14176 (US, WTU); sandy cranberry field, S of Gile Lake, elev. 5 m, TION R11W S3, 30 Sept. 1999, Zika 14438 (WTU). Previous knowledge. Native to eastern North America, W to Saskatchewan. Marsh St. Johnswort was first de- tected as a weed in cranberry farms of British Columbia in 1913 (EK Lomer, pers. comm.), where it has been re- ported as 7. virginicum (Hueppelsheuser & Emery, A Field Guide to Common Weeds of Cranberries in British Columbia, 1996). Significance. First report for Washington. VACCINIUM CORYMBOSUM L. (Ericaceae).—Grays Harbor Co., cranberry fields near Hogan Rd., N shore of North Bay, elev. 3 m, TI8N R11W S17, 4 Oct. 1999, Zika 14502 (WTU); King Co., boggy N shore of Panther Lake, 6 km MADRONO [Vol. 47) S of Renton, elev. 75 m, T22N RSE S5, 15 Oct. 1999, | Zika 14610 & Weinmann (WTU); undisturbed boggy | shore of Tub Lake, 140th St., Burien, elev. 100 m, T23N\ R4E S16, 14 Oct. 1999, Zika 14596 & Jacobson (WTU):. common escape near large cultivated blueberry fields, Mercer Slough, Bellevue, elev. 5 m, T24N RSE SS, 6 Oct. t 1999, Zika 14553 & Weinmann (WTU); marshy NW. shoreline of Union Bay, Seattle, elev. 4 m, T25N R4E S16, 27 Aug. 1999, Zika 14143 & Jacobson (WTUV). Previous knowledge. Highbush blueberry is commonly | cultivated for fruit west of the Cascades. It is native to eastern North America, west to Texas. Birds disperse the seed. Significance. First report for Washington as an escape. from cultivation. | VACCINIUM MACROCARPON Ait. (Ericaceae).—Grays Har- | bor Co., roadside ditch, with Anthoxanthum, Rubus spec- tabilis, Route 105, 2.5 km E of Ocosta, elev. 15 m, TI16N . R11W S11, 16 Sept. 1999, Zika 14340 (WTU); Pacific | Co., peaty clearing, near Jim St., 2 km NE of Seaview, - elev. 5 m, TION R1IW S22, 31 Aug. 1999, Zika 14195} (WTU). Previous knowledge. Cranberry is native to eastern North America, west to Minnesota. First introduced as a crop plant in Oregon in 1885, and known as a local weed © in California and British Columbia. Reported as question- | ably escaped in Washington (Buckingham et al. Flora of | the Olympic Peninsula, 1995). Significance. First documentation as a naturalized spe- cies in Washington. —PETER FE ZIKA, Herbarium, Dept. of Botany, Box 355325, Univ. of Washington, Seattle, 98195-5325; Ep- WARD R. ALVERSON, Herbarium, Dept. of Botany & Plant Pathology, Oregon State Univ., Corvallis, OR 97331; and | LOVERNA WILSON, P.O. BOX 2284, Corvallis, OR 97330. Volume 47, Number 3, pages 147-212, published 25 June 2001 Mh SUBSCRIPTIONS—-MEMBERSHIP Membership in the California Botanical Society is open to individuals ($27 per year; family $30 per year; emeritus $17 per year; students $17 per year for a maximum of 7 years). Late fees may be assessed. Members of the Society receive MApRONO free. Institutional subscriptions to Maprono are available ($60). Membership is based on a calendar year only. Life memberships are $540. Applications for membership (including dues), orders for sub- scriptions, and renewal payments should be sent to the Treasurer. Requests and rates for back issues, changes of address, and undelivered copies of MApRONO should be sent to the Corresponding Secretary. INFORMATION FOR CONTRIBUTORS Manuscripts submitted for publication in MaprRoNo should be sent to the editor. It is preferred that all authors be members of the California Botanical Society. Manuscripts by authors having outstanding page charges will not be sent for review. Manuscripts may be submitted in English or Spanish. English-language manuscripts dealing with taxa or topics of Latin America and Spanish-language manuscripts must have a Spanish RESUMEN and an English ABSTRACT. Manuscripts and review copies of illustrations must be submitted in triplicate for all articles and short items (NOTES, NOTEWORTHY COLLECTIONS, POINTS OF VIEW, etc.). Follow the format used in recent issues for the type of item submitted. Allow ample margins all around. Manuscripts MUST BE DOUBLE-SPACED THROUGHOUT. For articles this includes title (all caps, centered), author names (all caps, centered), addresses (caps and lower case, centered), abstract and resumen, 5 key words or phrases, text, acknowledgments, literature cited, tables (caption on same page), and figure captions (grouped as consecutive paragraphs on one page). Order parts in the sequence listed, ending with figures. Each page should have a running header that includes the name(s) of the author(s), a shortened title, and the page number. Do not use a separate cover page or ‘erasable’ paper. Avoid footnotes except to indicate address changes. Abbreviations should be used sparingly and only standard abbrevia- tions will be accepted. Table and figure captions should contain all information relevant to information presented. All measurements and elevations should be in metric units, except specimen citations, which may include English or metric measurements. Authors of accepted papers will be asked to submit an electronic version of the manuscript. Microsoft Word 6.0 or WordPerfect 6.0 for Windows is the preferred software. Line copy illustrations should be clean and legible, proportioned to the MApRONOo page. Scales should be in- cluded in figures, as should explanation of symbols, including graph coordinates. Symbols smaller than | mm after reduction are not acceptable. Maps must include a scale and latitude and longitude or UTM references. In no case should original illustrations be sent prior to the acceptance of a manuscript. Illustrations should be sent flat. No illustrations larger than 27 X 43 cm will be accepted. Presentation of nomenclatural matter (accepted names, synonyms, typification) should follow the format used by Sivinski, Robert C., in MaprRoNo 41(4), 1994. Institutional abbreviations in specimen citations should follow Holmgren, Keuken, and Schofield, Index Herbariorum, 8th ed. Names of authors of scientific names should be abbreviated according to Brummitt and Powell, Authors of Plant Names (1992) and, if not included in this index, spelled out in full. Titles of all periodicals, serials, and books should be given in full. Books should include the place and date of publication, publisher, and edition, if other than the first. All members of the California Botanical Society are allotted 5 free pages per volume in MAprono. Joint authors may split the full page number. Beyond that number of pages a required editorial fee of $40 per page will be assessed. The purpose of this fee is not to pay directly for the costs of publishing any particular paper, but rather to allow the Society to continue publishing MaproNo on a reasonable schedule, with equity among all members for access to its pages. Printer’s fees for illustrations and typographically difficult material @ $35 per page (if their sum exceeds 30 percent of the paper) and for author’s changes after typesetting @ $4.50 per line will be charged to authors. At the time of submission, authors must provide information describing the extent to which data in the manu- script have been used in other papers that are published, in press, submitted, or soon to be submitted elsewhere. ‘VOLUME 47, NUMBER 4 OCTOBER—DECEMBER 2000 MADRONO A WEST AMERICAN JOURNAL OF BOTANY | Shay ae a Vi if es ‘ : { SPECIAL ISSUE THE JEPSON HERBARIUM SOTH ANNIVERSARY CELEBRATION AND SCIENTIFIC SYMPOSIUM: DISCOVERY, COMMUNICATION, AND CONSERVATION OF PLANT BIODIVERSITY IN CALIFORNIA—JUNE 16—18, 2000 | "ONTENTS INTRODUCTION | Brent D. Mishler, Barbara Ertter, Bruce G. Baldwin, and Staci Markos. ...... PANG) ROLES FOR MODERN PLANT SYTEMATICS IN DISCOVERY AND CONSERVATION OF FINE- SCALE BIODIVERSITY BYUCE GB GI GWU rctesea cecal ones hs pees pos sees ghee 219 THE NEED FOR INTEGRATED STUDIES OF THE CALIFORNIA FLORA Brent D. MUSIC coccc ccc cccocccccsvcccunnas COR Mee AeA CSO onc once oP vac cciscncenesevaes 230 Our UNDISCOVERED HERITAGE: PAST AND FUTURE PROSPECTS FOR SPECIES-LEVEL BOTANICAL INVENTORY Barbara Evtter ........h.Zarsc RET OO wo OS ore ovccccssccccsaccaseceteose Zot FLorISTIC STUDIES IN CONTEMPORARY BOTANY Theodore MeBGVidley na She SN fil beri he OU BGP GD. casneees 255 COUPLING SPECIES-LEVEL INVENTORIES WITH VEGETATION MAPPING DavidA& Chdrlet .. £6.82 BS SOA MED sv ccccceevcccenevescesons 259 ELECTRONIC ACTIVITIES OF THE UNIVERSITY AND JEPSON HERBARIA Richtivad MOC EAT AL GRS SS Hv co ne MDE ES NPE SS cs vesccaeeveccsenens 265 WILLIS LINN JEPSON’S “MAPPING IN ForEST BOTANY” Willis Linn Jepson, Richard Beidleman, and Barbara Ertter ...........0000+. 269 WILLIS LINN JEPSON—‘“THE Botany MAN” Richard G. BetQlennan 2.0.0 8EKE GINS oe neve cen EINES Shas sce cevevcconvecceenvecs 213 PRESIDENT’ S REPORT FOR. VOLWIMEM Ty oncccccccccccn00es Seis ccndeevececnssvcasesveces 2a) EDITOR’S REPORT FOR VOLUME 47 23252... cccccccccssscsscccscecescessceasecsvesceasnes 288 REVIEWERS OF MANUSCRIPTS icc oso ate ea ed oeens eaweudaceeeneree cesses 288 PND XG TON OI IV IAG Gates cviscs id eset et ties erence iv Suns eee ree raat 289 DEDICATION AND BIOGRAPH Y—ROBERT ORNDUEFF ........0...cc00ecccseeesceeeeeeeees 292 TABEE OF CONTENTS FOR: VOLUME 47ier eae eis ul DATES OF PUBEICATION geet artis Geeta eraer eee ences min ill Sie PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL SOCIETY Maprono (ISSN 0024-9637) is published quarterly by the California Botanical Society, Inc., and is issued from the office of the Society, Herbaria, Life Sciences Building, University of California, Berkeley, CA 94720. Subscription information on inside back cover. Established 1916. Periodicals postage paid at Berkeley, CA, and additional mailing offices. Return requested. PostmMAsTER: Send address changes to MaproNo, Roy Buck, % University Herbarium, University of California, Berkeley, CA 94720. Editor—KrisTINA A. SCHIERENBECK | California State University, Chico Department of Biology Chico, CA 95929-0515 kschierenbeck @csuchico.edu Editorial Assistant—JEANNIE TRIZZINO Book Editor—Jon E. KEELEY Noteworthy Collections Editors—DIETER WILKEN, MARGRIET WETHERWAX Board of Editors | Class of: 2001—RobBert PATTERSON, San Francisco State University, San Francisco, CA PAuLa M. ScuIFFMAN, California State University, Northridge, CA 2002—NorMAN ELLSTRAND, University of California, Riverside, CA | Cara M. D’ Antonio, University of California, Berkeley, CA 2003—FREDERICK ZECHMAN, California State University, Fresno, CA Jon E. Keetey, U.S. Geological Service, Biological Resources Division, Three Rivers, CA 2004—Davw M. Woon, California State University, Chico, CA INGRID PARKER, University of California, Santa Cruz, CA CALIFORNIA BOTANICAL SOCIETY, INC. OFFICERS FOR 2000—2001 President: BRUCE BALDwin, Jepson Herbarium and Dept. of Integrative Biology, 1001 Valley Life Sciences Bldg. #2465, University of California, Berkeley, CA 94720. First Vice President: Rop Myatt, San José State University, Dept. of Biol. Sciences, One Washington Square, San José, CA 95192. rmyatt @email.sjsu.edu Second Vice President: Ros SCHLIsING, California State University, Chico, Dept. of Biol. Sciences, Chico, CA 95424. rschlising @csuchico.edu Recording Secretary: DEAN KELCcH, Jepson and University Herbarium, University of California, Berkeley, CA 94720. dkelch @sscl.berkeley.edu Corresponding Secretary: SUSAN BAINBRIDGE, Jepson Herbarium, University of California, Berkeley, CA 94720. suebain @SSCL.berkeley.edu Treasurer: Roy Buck, % University Herbarium, University of California, Berkeley, CA 94720. The Council of the California Botanical Society comprises the officers listed above plus the immediate Past President, R. JOHN LitrLe, Sycamore Environmental Consultants, 6355 Riverside Blvd., Suite C, Sacramento, CA 95831; the Editor of Maprono; three elected Council Members: BiAN Tan, Strybing Arboretum, Golden Gate Park, San Francisco, CA 94122; James SHEvocK, USDI National Park Service, Pacific West Region, 600 Harrison Street, Suite 600, San Francisco, CA 94107; Diane Exam, U.S. Fish and Wildlife Service, 3310 El Camino Avenue, Sacramento, CA 95825; Graduate Student Representative: KirsTEN JoHANUS, Jepson Herbarium, University of California, Berkeley, CA 94720. This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). MAbDRONO, Vol. 47, No. 4, pp. 217-218, 2000 THE JEPSON HERBARIUM 50TH ANNIVERSARY CELEBRATION AND SCIENTIFIC SYMPOSIUM: DISCOVERY, COMMUNICATION, AND CONSERVATION OF PLANT BIODIVERSITY IN CALIFORNIA JUNE 16-18, 2000 This warm weekend in June witnessed an in- credible outpouring of energy, information, science, love of plants, and just plain fun, revolving around _ the golden anniversary of the founding of the Jep- son Herbarium. More than 210 registrants con- verged on the Valley Life Sciences Building, UC Berkeley, for a diverse array of activities. On Fri- day afternoon, many visitors from out of town toured the facilities, worked in the herbarium col- lection, and enjoyed a formal reception that lasted well into the evening. In addition, participants from around and outside of California met that afternoon to discuss innovative ways to pursue floristic proj- ects in western North America. A primary focus of this special networking and brainstorming session was the potential development of electronic ap- proaches to floristics within a collaborative frame- work, taking advantage of shared resources and joint funding opportunities. Saturday morning was devoted to a plenary ses- sion that set the stage for the main issues to be examined the rest of the day. Beginning this ses- sion, Barbara Ertter spoke on the need for basic surveys of biodiversity, to the extent that even in a relatively well studied state like California a dozen new species are discovered and described each year. Bruce Baldwin followed with several test cases from the California flora showing how knowledge of phylogenetic relationships of plants is essential to determining their proper classifica- tion, even at the species level; evolutionarily sig- nificant, yet often cryptic, biodiversity is every- where. Theodore Barkley (Botanical Research In- stitute of Texas) summarized the need for infor- matics tools to make the huge amount of data on biodiversity available to the broad array of ‘‘con- sumers”’ of biodiversity information, and Ken Berg (U.S. Fish & Wildlife Service) presented a con- vincing case for us to get out in the trenches im- mediately and do what we can to protect plant bio- diversity in this rapidly developing part of the world. Finally, Brent Mishler discussed the need for integrated studies of the California flora linking practical floristics and conservation biology with academic systematics, a subject at the heart of the scientific goals of the weekend. On Saturday afternoon two sets of concurrent sessions (three sessions per set), allowed the partic- ipants to break out into smaller groups, thereby fos- tering discussion and interchange on the basic sub- jects of the morning talks. We encouraged folks to go to sessions that were on subjects somewhat out- side their normal interests. Judging by the debates, interesting questions, and sharing of information in the halls afterward, it was a great success! Following a break for further tours of the Her- baria, we re-convened at the Radisson Hotel at the Berkeley Marina for a reception and the gala 50th Anniversary Celebration Banquet, with Jepson Trustee Chair Roderic Park as the Master of Cere- monies. After a truly sumptuous meal, Richard Bei- dleman gave an entertaining tribute to Willis Linn Jepson, with many wonderful photographs and an- ecdotes from his extensive biographical research. Bob Ornduff followed with equally entertaining views of several other historical figures in Califor- nia botany, noting how their careers and interests interacted with Jepson’s. One of the many treats of the evening was the presence of two of Jepson’s great nieces—Mrs. Margaret Van Eck and Mrs. Louise Condeff, as well as two of his former grad- uate students, Dr. Lincoln Constance and Dr. Mary Bowerman, and relatives of a third, Mrs. Virginia Bailey. Sunday was devoted to a fine set of field trips led by expert botanists to some of the places Jepson and his students loved. There were packed trips to Point Reyes, Mount Diablo, Santa Cruz sandhills, Redwood Regional Park, and Solano County, with topics ranging from bryophytes to restoration to en- dangered species. Despite all the important indoor scientific activities, deep down it is being out in the field with the plants that excites us most of all. We all know who Willis Linn Jepson was as a botanist, in part because of the fine articles by Rich- ard Beidleman in the Jepson Globe and this issue of Madrono, and the paper by Lincoln Constance in an earlier issue of Madronio (1995, volume 42, number 2). But, what we were celebrating this weekend was Jepson as an institution-builder. De- spite his reputation as a somewhat cantankerous in- dividualist, he had the foresight to envision the need for a permanent botanical institution devoted to the flora of California at one of the premier uni- versities in the state. He followed up this vision with a gift in his will, which in 1950 established the Jepson Herbarium. We owe the central position and guaranteed continuance that the University and Jepson Herbaria now hold at UC Berkeley largely to Jepson’s foresight. The association of the Jepson Herbarium with a world-class university is unique in the state, and nearly unique across the US. This 218 gives us a tremendous opportunity to interface tra- ditional approaches to plant systematics and con- servation with the most up-to-date scientific ad- vances, to the benefit of general research, formal university education, and public outreach. Since the herbarium was established in 1950, other creative, hard working people have devoted their lives and fortunes to the Jepson Herbarium. Rimo Bacigalupi, the first Jepson Curator, left his published work and collections, as well as an en- dowment fund which is used to enhance research and publication in the Herbarium. The second Jep- son Curator, Larry Heckard, likewise left a distin- guished body of work and created an endowment fund that provides grants to UC Berkeley botanists for a spectrum of studies. From the Curators on down, many generous people have enhanced the Herbarium over its history through their continued support, and it is the totality of these efforts that we recognized with this weekend-long celebration. We would in particular like to thank the many staff members, students, and associates who _ helped make possible the weekend celebration and this special issue of Madrorio—especially Betsy Ring- rose who was involved in all aspects of the orga- nization. MADRONO [Vol. 47 | Inclusion of the proceedings of the Jepson Her- | barium’s 50th Anniversary Celebration in the vol- ume of Madrono dedicated to the late Robert Orn- | duff is particularly fitting given Bob’s long history | of involvement with the Jepson Herbarium. In ad- | dition to a full biography of Dr. Ornduff, who died | only months after giving his post-banquet talk, the | proceedings consist of the morning plenary session talks, several afternoon presentations that comple- | mented the other papers particularly well, Richard | Beidleman’s post-banquet biography of Willis Linn | Jepson, and a previously unpublished manuscript | by Jepson himself that seemed exceptionally appro- — priate. Ken Berg was unfortunately unable to pro- | vide a written version of his excellent talk (‘‘Con- — serving California’s Flora: Who Will Answer the | Call?’’) due to career developments. Dr. Ornduff’s | post-banquet talk, “Piss and Vinegar: Skeletons in | our Botanical Closet,’’ was geared more to the lay | audience and accordingly was published in Fre- | montia (volume 28, pages 18-20). | —BRENT D. MISHLER, BARBARA ERTTER, BRUCE G. | BALDWIN, and Staci MARKOS, Jepson Herbarium, 1001 | VLSB #2465, University of California, Berkeley, CA | 94720. MADRONO, Vol. 47, No. 4, pp. 219-229, 2000 ROLES FOR MODERN PLANT SYSTEMATICS IN DISCOVERY AND CONSERVATION OF FINE-SCALE BIODIVERSITY BRUCE G. BALDWIN Jepson Herbarium and Department of Integrative Biology, 1001 Valley Life Sciences Building #2465, University of California, Berkeley, CA 94720-2465 ABSTRACT Systematic methods involving the use of DNA data and genealogical analysis have been widely applied to higher-level phylogenetic questions in plants but much less commonly to discovering plant lineages corresponding to minimal-rank taxa (i.e., species, subspecies, and varieties) or to refining plant classifi- cation at the finest levels. Recent research in the Jepson Herbarium integrating extensive field sampling, biosystematic data, and molecular phylogenetics provides examples from the California flora for assessing the value of modern systematic approaches as a means of discovering fine-scale plant diversity. Results have sometimes led to taxonomic changes at the levels most important for biodiversity assessment and have allowed resolution of systematic questions important to establishing conservation strategies. Angio- sperm groups newly resolved with molecular data include both morphologically distinctive and morpho- logically cryptic lineages that have been previously treated within more broadly circumscribed species, subspecies, or varieties. Taxonomic recognition of such newly resolved lineages is often necessary if taxonomy is to reflect monophyletic groups and fine-scale units of biodiversity. To promote discovery, recognition, and conservation of plant lineages, systematists are advised to sample widely within minimal- rank taxa (including rare taxa) in the field and in herbaria, to consider previous taxonomies, to voucher all collections, to examine multiple lines of systematic evidence, and to publish taxonomic changes, including nomenclatural changes. Scientists involved in biodiversity management and conservation are advised to regard circumscriptions of all taxa as hypotheses of natural groups, to recognize that those hypotheses are subject to change, and to welcome taxonomic and nomenclatural changes that reflect an improved understanding of natural groups. Conservation biologists are urged to bear in mind that species or infraspecific taxa are not necessarily the minimal units of biodiversity. To conserve evolutionary lineages (and potential for future evolution), plant managers must seek to conserve representative popu- lations of taxa from throughout their geographical and ecological distributions, must resist indiscriminate use of non-local germplasm in restoration efforts, and must consider cryptic biodiversity in regional conservation planning. Systematics is fundamental to understanding of biodiversity. The most widely recognized organis- mal units of biodiversity, 1.e., species, may not rep- resent natural evolutionary groups or may not re- flect the finest-scale natural groups that can be re- solved and described by systematists. Modern sys- tematic approaches that allow a_ genealogical perspective on biodiversity hold great promise as a means of achieving a refined taxonomy that better reflects evolutionary lineages of organisms throughout the tree of life (e.g., Angiosperm Phy- logeny Group 1998). To date, modern systematic approaches to resolving evolutionary relationships have been applied by plant systematists mostly to questions concerning groups of recognized species or higher-level taxa (see Soltis et al. 1998). Recent studies of higher-level plant phylogeny have yield- ed insights into broad-scale evolutionary and bio- geographic patterns that are directly relevant to bio- diversity assessment and prioritization of conser- vation efforts (e.g., Vane-Wright et al. 1991; Mish- ler 1995; Faith 1996; reviewed by Soltis and Gitzendanner 1998). Modern systematic methods have been applied less commonly to testing the naturalness of mini- mal-rank taxa (i.e., species, subspecies, and varie- ties), which are of most immediate concern to con- servation biologists, ecologists, and floristicians (e.g., Rieseberg et al. 1988; see Soltis et al. 1992). Studies of phylogeographic diversity—fine-scale, geographically structured evolutionary lineages corresponding to ‘‘Evolutionarily Significant Units”? (sensu Moritz 1994)—also have been ex- tremely limited for plants (e.g., Fujii 1997; Soltis et al. 1997; Olsen and Schaal 1999; Tremblay and Schoen 1999; Shaw 2000; also see Schaal et al. 1998; Schaal and Olsen 2000), especially by com- parison with the rich literature on animal phylo- geography (reviewed by Avise 2000). As noted by Moritz (1995, 1999) and Coates (2000), conserving independently evolving sets of populations not only preserves biodiversity but also may be the best strategy for conserving the potential for future evo- lution. On a regional scale, refined understanding of phylogeographic patterns across organismal groups may allow for improved resolution of bio- diversity hot-spots and identification of critical ar- eas for conservation attention (see Moritz and Faith 1998). In this paper, I present examples of previously 220 undetected diversity resolved from studies of the California flora conducted in the Jepson Herbarium. These studies illustrate both the potential and the proven value of applying modern systematic meth- ods to discovery of fine-scale plant diversity and to refining classifications of minimal-rank plant taxa. Finally, I make general recommendations for how systematists and other biodiversity scientists and planners may promote discovery and conservation of plant diversity. EXPLORATION IN THE FIELD AND LABORATORY Well-focused field exploration (see Ertter 2000a) and detailed systematic analysis are complementary components of an effective strategy for discovering plant diversity. As reviewed by Ertter (2000b), bo- tanical field exploration in western North America, often by non-academic professionals and amateurs, has been a continuing source of floristic novelties. Modern systematic approaches can contribute greatly to the process of discovery by offering a rigorous means of resolving the systematic status of recently discovered plant populations. For ex- ample, DNA sequence variation may clarify wheth- er phenotypically unusual populations or sets of populations belong within previously described, minimal-rank taxa or represent undescribed evolu- tionary groups (e.g., Baldwin 1999a). Data from DNA also may allow confident taxonomic place- ment of newly discovered plants that are evidently distinct from any described minimal-rank taxon but of uncertain position (e.g., Boyd and Ross 1997). Conversely, fine-scale systematic studies depend on extensive field sampling across the geographical and ecological distribution of taxa for assessing naturalness of groups and detecting any unrecog- nized diversity within a group. Phylogeographic studies have demonstrated the potential for discov- ery of geographically distinct, and often morpho- logically cryptic, evolutionary lineages within both widespread and narrowly distributed species (see Soltis et al. 1997; Moritz 1999; Avise 2000; Riddle et al. 2000). Systematic studies and floristic surveys that involve extensive geographic sampling of widespread taxa as well as locally restricted taxa are therefore advisable to maximize the potential for discovering unrecognized plant diversity. The prospect for floristic discoveries to result from more detailed systematic analyses of western North American plant groups appears great. Young lineages, which account for much of the endemic plant diversity in western North America, e.g., Cal- ifornia (Raven and Axelrod 1978), can be expected to exhibit mosaic or cryptic phenotypic variation from minimal divergence, differential sorting of an- cestral polymorphism through descendant lineages (see Maddison 1995), or hybridization (see Arnold 1997). Climatic and geologic upheaval and exten- Sive species turnover seen in the plant fossil record during the mid to late Cenozoic in western North MADRONO [Vol. 47m America has been associated with the rise of di- | verse lineages of annuals and perennials that are largely or entirely restricted to the region (Axelrod | 1992; Graham 1999). The high degree of endem- © ism, ca. 50% of species, in the California Floristic | Province, i.e., the Mediterranean-climatic region of | western North America (Raven and Axelrod 1978), | largely reflects high diversity in neoendemic groups wherein often only minimal morphological diver- gence has occurred between evolutionary lineages. | Systematists in California and elsewhere in west- | ern North America have long appreciated the com- | plexity of the regional flora and the need for in- depth evolutionary investigations to reveal natural | units of biodiversity. The San Francisco Bay Area | botanists Harvey Monroe Hall, Ernest Babcock, G. | Ledyard Stebbins, Jens Clausen, David Keck, Wil- | liam Hiesey, and others pioneered the incorporation | of genetic principles and experimental approaches | into systematics in a highly successful effort to re- solve evolutionary lineages and understand com- plex patterns of variation in the California flora (e.g., Babcock and Hall 1924; Stebbins 1950; Clau- sen 1951; also see Smocovitis 1997). Subsequent | developments in systematics now allow even more progress in discriminating natural plant groups and refining the taxonomy of western North American plants. Advances in phylogenetic theory and methodol- ogy, together with the development of accessible high-speed computers, now permit simultaneous analysis of large numbers of variable characters to produce rigorous hypotheses of evolutionary rela- tionships within plant groups, as well as estimates of support for resolved lineages (see Swofford et al. 1996). Character changes (resulting from mu- tations) allow diagnosis of monophyletic groups (=evolutionary lineages or clades), the most natural groups recognized by systematists (Hennig 1966; see Mishler 1995, 2000a, b)—plants belonging to monophyletic groups are more closely related to one another than to plants in other groups. Access to an ever-increasing number of macromolecular characters from DNA sequences has enhanced the prospects for systematists to attain fine-grained, ro- bust resolution of evolutionary lineages (see Hillis et al. 1996; Soltis et al. 1998). EXAMPLES OF RECENT PLANT DISCOVERIES FROM SYSTEMATIC STUDIES To illustrate the utility of modern systematic methods for discovery of plant groups, I present below some examples from research conducted in my lab at the Jepson Herbarium, principally on Cal- ifornian angiosperm lineages. Although categori- zation of the examples is somewhat artificial, three general types of problems are addressed: confusing variation within taxa, resolution of cryptic biodi- versity, and questionably distinct rare taxa. 2000] I. Variation within taxa reexamined. The first category of examples comprises groups wherein morphological variation within a taxon was of un- certain systematic significance until phylogenetic studies were undertaken. Deinandra bacigalupii: A narrow endemic mis- placed in a widespread species.—Deinandra is a species-rich genus of tarweeds (Madiinae, Com- positae) reinstated for members of Hemizonia sensu Keck (1959) that are most closely related to Hol- ocarpha (Baldwin 1999b). Until 1999, an ca. 8-rayed Deinandra from alkaline meadows in the Livermore Valley, California, was treated within D. [Hemizonia] increscens subsp. increscens, a mostly coastal taxon known otherwise from Santa Barbara County to Monterey County, California (Tanowitz 1982), more than 75 km south of Livermore Valley. Morphologically, the Livermore Valley tarweed is highly similar to D. increscens except in anther col- or and pappus characteristics. Robert F Hoover col- lected and left unidentified to species the Livermore Valley tarweed as early as 1966. Rimo Bacigalupi annotated the UC accession of the first known col- lection (by Hoover) as not matching any published species of Hemizonia. Dean Kelch and Robert Preston independently collected the Livermore Valley tarweed in the 1990’s and brought specimens to me with concerns that the plant was not identifiable with The Jepson Manual: Higher Plants of California (Hickman 1993). The characteristics of yellow to brownish, rather than “‘black”’ (1.e., dark purple), anthers in the Livermore Valley plants was in conflict with placement in Deinandra [Hemizonia] increscens and ray laminae of the plants were much too short for D. [Hemizonia] pallida. Further examination of the plants in comparison with other deinandras re- vealed that the pappus was much shorter and more irregular than in other populations then assigned to D. increscens (Baldwin 1999a). Chromosome counts of the Livermore Valley tarweed of 2n = 12 II, the modal chromosome number in Deinandra (as in D. increscens), were inconclusive about re- lationships of the plants (Baldwin 1999a). Results of phylogenetic analysis of nuclear ri- bosomal DNA (rDNA) sequence data, in concert with the morphological evidence, led me to con- clude that the Livermore Valley tarweed is not a member of Deinandra increscens or any other pre- viously recognized species of Deinandra (Fig. 1; Baldwin 1999a). Representatives of the two most divergent groups of D. increscens, i.e., D. i. subsp. increscens and D. i. subsp. villosa, were resolved as a well-supported monophyletic group to the ex- clusion of representatives of the other six recog- nized species of the ‘‘northern lineage’? of Dein- andra and the Livermore Valley tarweed. The Liv- ermore Valley tarweed does not appear to be of recent hybrid origin based on 10 unambiguous rDNA mutations not shared with any other sampled BALDWIN: DISCOVERY AND CONSERVATION OF PLANT LINEAGES 221 D. increscens subsp. increscens D. increscens subsp. villosa D. bacigalupii D. corymbosa subsp. corymbosa D. corymbosa subsp. macrocephala D. kelloggii D. pallida D. pentactis D. lobbii D. halliana 5 changes Fic. 1. The most parsimonious tree from phylogenetic analysis of 18S-26S nuclear ribosomal DNA sequences of the external and internal transcribed spacers for the north- ern lineage of Deinandra (Compositae—Madiinae; Bald- win unpublished data). The outgroup (D. minthornii) used for rooting the tree is not shown, nor are tree statistics and support values (to be published elsewhere). Note the extensive divergence of the Livermore Valley tarweed (D. bacigalupii) from other representatives of Deinandra and the remote phylogenetic position of D. bacigalupii from D. increscens, the species in which the Livermore Valley tarweed was earlier treated. plants from the “‘northern lineage”’ of Deinandra (Baldwin, unpublished data). Evidence from DNA substantially augmented morphological evidence for distinctiveness of the Livermore Valley tarweed from D. increscens. Rec- ognition of D. bacigalupii as distinct from D. in- crescens improves our understanding of diversity in Deinandra and rare plants in general in the Spring- town wetlands area near Livermore, California. Deinandra bacigalupii has been regarded as an ex- ample of a plant species that was discovered ‘‘in front of the bulldozer’? (Ertter 2000b), 1.e., that came close to being driven to extinction prior to being recognized as distinct. Layia gaillardioides: Interpopulational variation of phylogenetic significance.—Layia_ gaillardioi- des, the woodland layia, has been regarded as an example of a species displaying substantial mor- phological variation among populations (Clausen 1951). Ray laminae may be uniformly deep yellow or have white, greenish, or pale yellow apices de- 9 it Tomales Muir Beach { } { Saratoga Summit Fic. 2. MADRONO | } Knoxville 3 Isabel Creek Lewis Creek i t Basal leaf variation in the woodland layia, Layia gaillardioides (modified from Clausen 1951). Leaves sepa- rated by the vertical line correspond to outer (left) and inner (right) Coast Range populations in central and northern California. The populational differences shown here correspond to variation across three divergent molecular lineages that do not appear to constitute a natural group (Baldwin, unpublished data). pending on the population examined. Clausen (1951) noted that inner and outer Coast Range pop- ulations differ in stem thickness and degree of lob- ing of the basal leaves, characteristics that he con- cluded were heritable and ecologically significant (Fig. 2). Phylogenetic analysis of Layia has revealed ev- idence that L. gaillardioides as circumscribed at present represents an unnatural group (Baldwin, un- published data). Populations shown earlier by Clau- sen to be morphologically distinct constitute three distinct lineages that apparently are not most close- ly related to one another. Two lineages of L. gail- lardiodes are more closely related to L. carnosa, L. hieracioides, and L. septentrionalis than to a third lineage of L. gaillardioides. Each of the groups re- solved within L. gaillardioides is well-supported by unique rDNA mutations, and relationships among the lineages and related species are likewise robust based on rDNA data. Evolutionary lineages within the woodland layia conform to a typical phylogeo- graphic pattern except that the natural groups with- in L. gaillardioides do not constitute a clade and instead represent a paraphyletic or (conceivably) polyphyletic group (see Riddle et al. 2000 for sim- ilar examples). Recognition that L. gaillardioides has been circumscribed too broadly and comprises multiple natural groups, each probably warranting taxonomic distinction, may be an important con- servation concern given the evidently scattered dis- tribution and small size of woodland layia popula- tions. Sampling of additional populations and DNA regions is now underway to resolve the precise de- limitation of each evolutionary lineage within L. gaillardioides s. lat. prior to describing segregate taxa. Lessingia: Problems in species and varietal cir- cumscriptions in the “‘yellow group.’’—Systematic investigations by Markos (2000; also see Markos and Baldwin 2001) revealed an outstanding exam- ple of misinterpreted morphological variation in an- other lineage of annuals in the California Compos- itae, namely, in the “yellow group” of Lessingia (Astereae). Markos (2000) found that annuals in Lessingia constitute two major lineages that differ in disc corolla coloration—a_ pink-to-white-flow- ered lineage and a yellow-flowered lineage. Within the “‘yellow group,’ Markos (2000) used morpho- logical and molecular data to resolve three major natural groups that span the boundaries of widely recognized species and varieties. Markos (2000) found that different representa- tives of each of three taxa (Lessingia gladulifera, L. glandulifera var. glandulifera, and L. lemmonii) belong to different major lineages within the “‘yel- low group.”’ Morphologically, differences in shape of the style-branch appendages and presence or ab- sence of a maroon band in the corolla throat diag- nose the three primary lineages of yellow-flowered lessingias. Markos (2000) concluded from her phy- logenetic data that the accepted taxonomy of Les- singia underrepresents the actual biodiversity of the group and warrants substantial revision (S. Markos, in prep.) Il. Cryptic biodiversity. Modern systematic methods have promising potential for allowing dis- covery of natural plant groups that are morpholog- ically indistinguishable (or nearly so) from one an- other but may be geographically or ecologically distinct. As noted above, lineage diversity across the geographic distribution of a morphological or [Vol. 47 | | | | } | t } 2000] biological species, i.e., phylogeographic diversity, has been widely reported in vertebrates but has not been extensively studied in plants (see Soltis et al. 1997; Avise 2000; Schaal and Olsen 2000). In Cal- _ifornian groups of annuals, members of my lab | } have found various examples of cryptic diversity associated with geography. Two examples are par- ticularly important for illustrating groups that are not only morphologically cryptic but, based on multiple lines of evidence, must be recognized as distinct taxa because the well-supported but mor- phologically indistinct lineages are evidently not most closely related to one another. Downingia yina.—Schultheis (2000) examined relationships throughout Downingia (Lobeliaceae) with special attention to a lineage corresponding to three morphological species: D. bacigalupii, D. ele- gans and D. yina. Earlier work by Weiler (1962) and Foster (1972) established that D. yina is cyto- logically highly unusual, with a broad dysploid se- ries of chromosome numbers, 1.e., 2n = 6, 8, 10, and 12 Il. Chromosome “‘races”’ of D. yina are mostly geographically distinct but could not be dis- tinguished reliably on the basis of morphology us- ing multivariate analyses and other approaches (Schultheis 2000). Schultheis (2000) extensively sampled D. baci- galupii, D. elegans, and D. yina throughout their geographic ranges in an attempt to discern the evo- lutionary significance of chromosomal variation in the group. She found strong phylogenetic signals from sequences of both chloroplast DNA (cpDNA) and nuclear rDNA for three major lineages with cytological and geographic integrity that each in- clude populations of D. yina. One lineage corre- sponds to populations west of the Cascade Range, all with chromosome numbers of 2” = 6, 8, or 10 II, 1.e., D. elegans (2n = 10 ID) and populations of D. yina with 2n = 6, 8, or 10 II. The second lineage corresponds to a pocket in the southern Cascade Range of Oregon wherein populations of D. yina with 2n = 10 II are found. The third lineage cor- responds to populations east of the Cascades, with 2n = 12 II, 1.e., D. bacigalupii and D. yina. Schultheis (2000) concluded that the three well- supported evolutionary lineages in D. yina warrant taxonomic recognition despite being only crypti- cally distinct morphologically. Congruence be- tween two lines of molecular data leave minimal doubt that D. yina represents an example of diver- gent evolutionary lineages that remained morpho- logically static while closely related lineages (cor- responding to D. bacigalupii and D. elegans) un- derwent considerable morphological change. Dif- ferences among the three major groups of D. yina in geographic distribution and nuclear genomic ar- rangements conceivably extend to physiological differences of fundamental importance to survivor- ship in distinct ecological settings. Lasthenia californica—The goldfield genus Las- thenia (Compositae) has been the subject of a re- BALDWIN: DISCOVERY AND CONSERVATION OF PLANT LINEAGES 223 cent molecular phylogenetic study by Chan (2000), who sampled widely across populations of each taxon recognized by Ornduff (1966, 1993). Chan found strong evidence from cpDNA and _ nuclear rDNA sequences for morphologically cryptic lin- eages in L. californica, the most widespread species recognized by Ornduff (1993). Chan’s (2000) cpDNA and nuclear rDNA data led him to propose the hypothesis that one set of populations of L. californica sensu Ornduff (1993) is most closely related to the outer coastal, endemic Californian taxa L. macrantha subsp. macrantha and L. m. subsp. bakeri. Chan concluded that the three taxa constitute a well-supported group exclu- sive of the Pacific Northwest endemic L. macrantha subsp. prisca and other populations of L. califor- nica sensu Ornduff (1993). Chan (2000) also found that the two lineages corresponding to L. califor- nica sensu Ornduff (1993) have somewhat distinct (but overlapping) geographic distributions and mi- nor pappus differences, although some individuals of both groups are epappose and cannot be reliably distinguished morphologically. Lasthenia california sensu Ornduff (1993) appears to represent another example, parallel to Downingia yina, of lineages that do not constitute a natural group but have re- mained morphologically similar while related lin- eages have undergone substantial morphological change. Taxonomic recognition of cryptic plant groups.—Morphologically indistinct evolutionary groups such as the two examples discussed above (within Downingia yina and Lasthenia californica) present a special challenge to plant systematists and the botanical community. Cryptically distinct lin- eages that together constitute a monophyletic group may or may not be viewed as warranting taxonomic recognition. In the interests of accurate biodiversity assessment (which typically relies on taxa as units), I suggest that cryptic groups that are: (1) well-sup- ported by different lines of molecular or other ev- idence, and (2) geographically or ecologically dis- tinct should be recognized as taxonomically dis- tinct. Well-supported, cryptically-distinct groups that are not even most closely related to one anoth- er (e.g., groups in D. yina and in L. californica) leave systematists committed to natural classifica- tion with no choices other than to recognize the cryptic groups as taxonomically distinct or to treat all members of the minimal monophyletic group encompassing the cryptic lineages and the related group(s) as a minimal-rank taxon. I believe that the second option is undesirable because it under-rep- resents biodiversity. Practicality of classification is a concern for plant systematists and other botanists, especially for those faced with accurately identifying plant taxa with minimal time and resources. A system of clas- sification that does not allow some plant taxa to be identified on the basis of macroscopic morpholog- ical characteristics alone generally has been resisted 224 by vascular plant systematists (but not bryologists or other botanists). Insofar as geography or habitat characteristics often aid identification of cryptic lin- eages, practical problems arising from giving for- mal taxonomic status to morphologically indistin- guishable groups may be limited. Eventually, ad- vances in DNA analysis and computer technology may trivialize the process of screening for diagnos- tic genetic markers, even in the field, thereby al- lowing botanists to be less reliant on morphological characteristics for plant identification. I recognize, however, that for plant groups wherein morphology has evolved even faster than DNA sequences in commonly examined, rapidly-evolving gene regions, e.g., in some examples of insular adaptive radiation (see Baldwin et al. 1998), morphology may provide a finer-scale perspective on evolution- ary lineages than most easily obtained DNA data. Strict adherence to a criterion of macroscopic diagnosability for all vascular plant taxa places un- acceptable and unnatural limits on the information content of our taxonomy and potentially jeopard- izes important segments of biodiversity because of a human bias toward recognizing only what can be seen with minimally-assisted human eyesight. From a biological standpoint, cryptic differences between plant groups, e.g., in characters associated with ecophysiology, can be more important to the integ- rity and fitness of plant groups than differences that humans can perceive visually. From a conservation perspective, taxonomic recognition of cryptically distinct natural groups may be important to ensure their legal protection (e.g., only formally-named plant lineages are eligible for protection under the U.S. Endangered Species Act). Taxonomic status for cryptic groups also may help to ensure their protection from misguided restoration efforts that result in combining germplasm from different evo- lutionary lineages treated within the same minimal- rank taxon, with consequent loss of lineage integ- rity and possible outbreeding depression (see Mo- ritz 1999). From a more general perspective, ad- herence to the belief that plant systematics is a science that seeks to discern real entities of nature, 1.e., evolutionary groups, dictates that plant taxon- omy should reflect rigorous hypotheses of relation- ship rather than convenient but artificial or oversim- plistic assemblages. Based on available evidence, I suspect that widespread recognition of cryptic taxa would result in only a modest refinement, not a major overhaul, of plant taxonomy. Ill. Conservation prioritization. Mishler (1995) and others have discussed the potential value of phylogenetic data on the age and position of lin- eages for prioritizing conservation efforts (re- viewed by Soltis and Gitzendanner 1998). On an even more fundamental level, modern systematic data can help to resolve whether rare taxa of ques- tionable naturalness truly represent evolutionary lineages worthy of conservation attention and re- MADRONO we | [Vol. 47 | sources. Skinner et al. (1995) identified over 150 examples of rare, minimal-rank taxa of Californian | vascular plants that needed systematic attention. | The two rare taxa discussed below are examples of | groups that were studied systematically in part to | determine whether they warrant continued recog- | nition and, for Blepharizonia, to determine whether | gene flow between species represents a conserva- tion concern. Blepharizonia plumosa: Rare species or minor | morphological variant?—Baldwin et al. (2001) ex- | amined biosystematic and phylogenetic data for Blepharizonia to assess whether the common big | tarweed, B. plumosa subsp. viscida, warrants tax- | onomic distinction from the rare big tarweed, B. © plumosa subsp. plumosa. Keck (1959) regarded the © two taxa as allopatric but recent field work by Rob- | ert Preston established that the two taxa are mosa- ically sympatric in the northern Mt. Hamilton Range, California, where Baldwin et al. (2001) | sampled the big tarweeds for crossing and molec- ular studies. Phylogenetic analysis of rDNA se- quence data yielded evidence for ancient diver- | gence between the two taxa of Blepharizonia rela- | tive to timing of divergence between taxa in the sister-genus, Hemizonia (Fig. 3). Low interfertility of artificial hybrids corroborated phylogenetic evi- dence for greater divergence between the big tar- weed taxa than implied by Keck’s (1959) taxono- my. Baldwin et al. (2001) concluded that the two taxa of Blepharizonia should continue to be recognized and warrant treatment as separate species, B. laxa (=B. plumosa subsp. viscida) and B. plumosa (=B. plumosa subsp. plumosa). Baldwin et al. (2001) also concluded from DNA and artificial hybridiza- tion data that natural hybridization between B. laxa and B. plumosa probably does not pose a threat to the continued existence of the rare big tarweed, B. plumosa. Preliminary evidence for phylogeographic diversity uncovered within the rare B. plumosa (Baldwin et al. 2001) should serve as a caution against any conceivable restoration efforts that in- volve moving germplasm of B. plumosa between populations (especially north or south of the Liv- ermore Valley), at least until continuing studies of lineage diversity in Blepharizonia are completed. Sidalcea keckii: A minor variant of S. diploscy- pha or a distinct, rare lineage?—Phylogenetic stud- ies of Sidalcea (Malvaceae) by Andreasen (in prep.; see also Andreasen and Baldwin, 2001) helped to clarify evolutionary lineages in the genus, a group previously regarded as highly variable, taxonomi- cally difficult, and in need of systematic attention (Hill 1993). Among the issues of conservation con- cern examined by Andreasen was the evolutionary status of S. keckii, a narrowly endemic species from Tulare County, California, long thought to be ex- tinct until rediscovered in 1992 (see Skinner and Pavlik 1994). Assigning conservation priority to S. keckii has been complicated by uncertainty about 2000] Hemizonia congesta subsp. Juzulifolia Hemizonia congesta subsp. Jutescens Viscida (Parkfield Grade, Monterey Co.) Viscida (LLNL, San Joaquin Co.) Viscida (nw of Black Butte, San Joaquin Co.) Viscida (Tesla Road, Alameda Co.) Plumosa (Marsh Creek Road, Contra Costa Co.) Plumosa (LLNL, San Joaquin Co.) —— 0.005 substitutions/site Plumosa (Tesla Road, Alameda Co.) Fic. 3. Chronogram of one of two maximally parsimo- nious trees from phylogenetic analysis of nuclear riboso- mal DNA sequences of Blepharizonia and Hemizonia (modified from Baldwin et al. 2001). Branch lengths were optimized by maximum-likelihood to conform to an hy- pothesis of evolutionary rate constancy, which could not be rejected using a tree-wide likelihood-ratio test. Note the extensive divergence between the two, minimally in- terfertile, mosaically sympatic taxa of Blepharizonia rel- ative to divergence between the two representatives of Hemizonia. Biosystematic and phylogenetic data led Bald- win et al. (2001) to conclude that the two taxa of Ble- pharizonia each correspond to natural groups that warrant treatment as distinct species. Abbreviations: Plumosa = B. plumosa sensu stricto [=B. plumosa subsp. plumosa); Viscida = B. laxa [=B. plumosa subsp. viscida]. See Baldwin et al. (2001) for tree statistics and support values. distinctiveness of the species from the morpholog- ically similar, widespread species S. diploscypha. Andreasen (in prep.; see also Andreasen and Baldwin, 2001) sampled both species in a genus- wide phylogenetic analysis of rDNA spacer se- quences in Sidalcea. She found that S. diploscypha and S. keckii were most closely related to one an- other, as expected, but that representatives of each species constituted highly divergent lineages. Based on her findings, Andreasen concluded that S. keckii is worthy of continued taxonomic recognition and conservation attention. RECOMMENDATIONS FOR SYSTEMATIC STUDIES To promote further progress by systematists in the discovery of plant diversity corresponding to minimal-rank taxa and in the refinement of plant BALDWIN: DISCOVERY AND CONSERVATION OF PLANT LINEAGES 225 classification at the lowest taxonomic levels, the following recommendations are presented for plan- ning systematic studies: @ Sample widely within accepted taxa. To test tax- onomic hypotheses and to detect cryptic lineage diversity, sampling within taxa across the range of phenotypic variation and across the geograph- ical and ecological distribution has been produc- tive (see above). Examining only one exemplar per taxon cannot reveal hidden diversity or prob- lems in circumscription at the taxonomic level of sampling. @ Study herbarium collections. Apart from yielding valuable data on morphological, ecological, and geographic variation within minimal-rank taxa, studies of herbarium specimens may reveal un- described diversity more readily than new field exploration. Feasibility of extracting sufficient DNA for genetic analyses from small fragments of dried plant material may allow both morpho- logical and molecular characterization of new species discovered in herbaria (e.g., Vargas et al. 1993): @ Take seriously the old taxonomic literature. A sampling focus on taxa recognized only in the most recent taxonomic revision of a plant group may ensure a repeat of errors made in that sys- tematic treatment, especially if sampling within taxa is minimal. In addition to taking a fresh look at variation within a group, systematists may find that taxa no longer recognized in modern treat- ments represent evolutionary lineages warranting recognition (e.g., Chan 2000). @ Voucher all specimens examined. Vouchering specimens for systematic studies of groups cor- responding to minimal-rank taxa is perhaps even more essential than for studies at higher taxo- nomic levels to ensure that the identities of sam- pled plants are not misinterpreted by others. Doc- umentation of detailed collection data is also crit- ical for studies involving fine-scale sampling within minimal-rank taxa (see Huber 1998). @ Examine multiple lines of systematic evidence. A single line of systematic evidence (e.g., one gene) can be potentially misleading about rela- tionships within a group (see Wendel and Doyle 1998). Lineage sorting and hybridization are more likely to affect evolutionary patterns in young plant groups than in old lineages. Exam- ining multiple, unlinked gene regions or molec- ular and other lines of data (e.g., morphology or cytology) should increase the likelihood of achieving an accurate understanding of natural plant groups. @ Sample the rare taxa. The potential value to con- servation biology of gaining additional system- atic data on rare plants makes the efforts required to sample rare taxa worthwhile. Most modern molecular systematic methods involve use of the polymerase chain reaction (PCR), which requires 226 only minimal DNA (see Hillis et al. 1996). Mo- lecular data from rare plants can be obtained from minute amounts of fresh or dried (e.g., her- barium) tissue without impacting populations or significantly damaging voucher specimens. Study biological characteristics of the plants. \n- cluding an experimental biosystematic compo- nent (e.g., from artificial hybridizations or com- mon gardens) and field component (e.g., polli- nation ecology, demography) in modern system- atic studies can yield valuable biological data for resolving fine-scale diversity within a group and may lead to insights into evolutionary processes affecting diversification (see Baldwin 1995). Studies of population-genetic structure within lineages can provide critical biological data for resolving microevolutionary dynamics of popu- lations and for refining conservation strategies (e.g., Bushakra et al. 1999). Communicate with other field botanists. Close communication and cooperation with profession- al and amateur field botanists is especially valu- able for promoting discovery and conservation of plant diversity. The reciprocal flow of knowl- edge that can develop between systematists and other field-immersed plant biologists enriches botany in general and can lead to a more inten- sive, well-focused effort toward detecting and conserving diversity than would be otherwise possible (see Ertter 2000a, b). Publish findings and follow through on taxonom- ic changes. Other biologists, especially those in- volved in biodiversity assessment and conserva- tion (e.g., Skinner and Pavlik 1994), rely on for- mal taxonomic treatments and other publications by the systematic community. Translating perti- nent results of systematic studies into taxonomic changes is a potentially tedious but necessary step to ensure that newly discovered evolution- ary lineages and new understanding of the cir- cumscriptions and positions of monophyletic plant groups in general are recognized by others. RECOMMENDATIONS TO THE CONSERVATION COMMUNITY Based on the evidence from phylogenetic studies that circumscriptions of some minimal-rank taxa misrepresent or under-represent biodiversity, I sug- gest that the following recommendations be adopt- ed by the conservation community in the interests of preventing loss of natural plant groups: Regard taxa as hypotheses of natural groups subject to change. Some refinements to our un- derstanding of the composition and position of natural plant groups are inevitable and desirable to ensure that conservation efforts are well di- rected: Accept and encourage taxonomic changes based on solid evidence of natural groups. Although taxonomic changes create difficulties in data- MADRONO base management and communication, changes that reflect an improved understanding of natural | groups are valuable and worth adopting. From a | conservation perspective, names are expendable; , natural plant groups are irreplaceable. Bear in mind that recognized species or infra- | specific taxa are not necessarily minimal units of biodiversity. As noted above, unrecognized, ev- olutionarily distinct lineages may exist within a | species, subspecies, or variety (also see Soltis and Gitzendanner 1998). Research efforts to dis- | cern any undetected diversity within minimal- | rank plant taxa using modern systematic ap- proaches (e.g., phylogeographic studies) have | been minimal (see above). Available data sug- gest that cryptic lineages often show some de- gree of geographic distinction (see Avise 2000). Efforts to protect taxa throughout their geograph- ical and ecological ranges are therefore warrant- ed not only to ensure survival of locally adapted populations and overall allelic diversity within a group (Endler 1977; Chesser 1983) but also to preserve potentially distinct evolutionary lin- eages. Resist proposals to use non-local germplasm in- discriminately in restoration efforts. Use of non- local germplasm in restoration efforts may result in extensive hybridization between evolutionarily distinct but cryptic lineages and consequent loss of biodiversity (see Storfer 1999). This concern is especially important given the increasing prev- alence of mitigation efforts seeking to augment rare plant populations in protected areas with propagules or mature plants translocated from other populations slated for destruction. The well-intentioned practice of augmentation may be justifiable to prevent or overcome inbreeding depression, e.g., if the populations involved are remnants of a more continuous metapopulation fragmented by human-related activities or if ge- neological and population genetic data indicate that declining populations are of a common re- gional lineage and share similar genetic structure (see Moritz 1999). Indiscriminate translocation of plants from one population to another has po- tential to do much harm to biodiversity and to our prospects for understanding the evolution or population-genetic structure of natural plant pop- ulations. In the absence of genealogical and population- genetic data, proposals for augmentation of nat- ural populations with non-local seed should be viewed with the same skepticism as the univer- sally objectionable idea of intermixing germ- plasm from unquestionably distinct but interfer- tile, naturally allopatric species. Even if the pop- ulations to be intermixed do not represent highly divergent evolutionary lineages, potential still exists for outbreeding depression from loss of lo- cal adaptation or breakdown of co-adapted gene complexes (see Templeton 1986; Slatkin 1987; [Vol. 47 | 2000] Moritz 1999; Storfer 1999). Planting of wild- flowers along roads and highways is another widespread practice with similar potential for re- ducing biodiversity and confounding scientific investigation of natural plant populations. @ Consider cryptic biodiversity in conservation planning. Even if systematists decline to recog- nize cryptic lineages as taxa, conservationists can plan for preservation of cryptic groups in the interests of preserving unnamed, as well as named, biodiversity. Under the U.S. Endangered Species Act, evolutionarily significant but mor- phologically indistinct and unnamed lineages of vertebrates are eligible for protection; similar protection for cryptic plant groups may be pos- sible to achieve. Geographical and ecological cri- teria have been used effectively for recognizing cryptic vertebrate lineages (e.g., salmonids) and also may be useful for identifying various cryptic plant groups. CONCLUSIONS The potential for modern systematics to play an important role in the discovery and conservation of fine-scale plant biodiversity is enormous and most- ly untapped. The movement of systematics toward use of molecular and phylogenetic methods has been perceived by some botanists as an alarming diversion from the urgent business of finding and describing previously undetected and, usually, en- dangered plant diversity (“‘fiddling while Rome burns’’). I suggest that the use of modern system- atic approaches, far from posing a threat to ad- vancing our knowledge of fine-scale biodiversity, can be an invaluable means of achieving rapid progress in the discovery and conservation of plant lineages. ACKNOWLEDGMENTS Special thanks to Katarina Andreasen, Raymund K.-G. Chan, Staci Markos, and Lisa M. Schultheis for gener- ously allowing me to present examples from their pub- lished dissertations or (for KA) from an in-press postdoc- toral research article for this symposium presentation. Thanks also to Daniel J. Crawford, Brent D. Mishler, and John L. Strother for helpful comments on the manuscript. The research discussed here was funded in part by the National Science Foundation (DEB-9458237), by gifts from Roderic B. Park and other generous Friends of the Jepson Herbarium, and by the Lawrence R. Heckard En- dowment Fund of the Jepson Herbarium. LITERATURE CITED ANDREASEN, K. AND B. G. BALDWIN. 2001. Unequal evo- lutionary rates between annual and perennial lineages of checker mallows (Sidalcea, Malvaceae): Evidence from 18S-26S rDNA internal and external transcribed spacers. Molecular Biology and Evolution 18:936— 944. ANGIOSPERM PHYLOGENY Group. 1998. An ordinal classi- fication for the families of flowering plants. Annals of the Missouri Botanical Garden 85:531—553. BALDWIN: DISCOVERY AND CONSERVATION OF PLANT LINEAGES 224 ARNOLD, M. L. 1997. Natural hybridization and evolution. Oxford University Press, New York, NY. AVISE, J. C. 2000. Phylogeography: The history and for- mation of species. Harvard University Press, Cam- bridge, MA. AXELROD, D. I. 1992. Miocene floristic change at 15 Ma, Nevada to Washington, U.S.A. Palaeobotanist 41: 234-239. BABCOCK, E. B. AND H. M. HALL. 1924. Hemizonia con- gesta: A genetic, ecologic, and taxonomic study of the hay-field tarweeds. University of California Pub- lications in Botany 13:15—100. BALDWIN, B. G. 1995. A new prospect for California bot- any: Integrating biosystematics and phylogenetics. Madrono 42:154—167. 1999a. Deinandra bacigalupii (Compositae— Madiinae), a new tarweed from eastern Alameda County, California. Madrono 46:55—57. . 1999b. New combinations and new genera in the North American tarweeds (Compositae—Madiinae). Novon 9:462-—471. , D. J. CRAWForRD, J. FRANCISCO-ORTEGA, S.-C. Kim, T. SANG, AND T. E Stugssy. 1998. Molecular phylogenetic insights on the origin and evolution of oceanic island plants. Pp. 410—441 in P. S. Soltis, D. E. Soltis, and J. J. Doyle (eds.), Molecular system- atics of plants Il: DNA sequencing. Kluwer Academic Publishers, Boston, MA. , R. E. PRESTON, B. L. WESSA, AND M. WETHER- WAX. 2001. A biosystematic and phylogenetic assess- ment of sympatric taxa in Blepharizonia (Composi- tae—Madiinae). Systematic Botany 26:184—194. Boyb, S. AND T. S. Ross. 1997. Sibaropsis (Brassicaceae), a new monotypic genus from southern California. Madrono 44:29—47. BUSHAKRA, J. M., S. A. HopGEs, J. B. COOPER, AND D. D. KASKA. 1999. The extent of clonality and genetic di- versity in the Santa Cruz Island ironwood, Lyono- thamnus floribundus. Molecular Ecology 8:471—475. CHAN, R. K.-G. 2000. Molecular systematics of the gold- field genus Lasthenia (Compositae: Heliantheae sensu lato). Ph.D. dissertation. University of California, Berkeley. CHESSER, R. K. 1983. Isolation by distance: Relationship to management of genetic resources. Pp. 66—77 in C. Schonewald-Cox, S. Chambers, B. MacBryde, and W. Thomas (eds.), Genetics and conservation: A refer- ence for managing wild animal and plant populations. Benjamin Cummings Publishing Company, London. CLAUSEN, J. 1951. Stages in the evolution of plant species. Cornell University Press, Ithaca, NY (reprinted by Hafner Publishing Co., New York, NY). Coates, D. J. 2000. Defining conservation units in a rich and fragmented flora: Implications for the manage- ment of genetic resources and evolutionary processes in south-west Australian plants. Australian Journal of Botany 48:329-—339. ENDLER, J. A. 1977. Geographic variation, speciation and clines. Princeton University Press, Princeton, NJ. ERTTER, B. 2000a. Our undiscovered heritage: Past and future prospects for species-level field exploration. Madrono 47:237-252. . 2000b. Floristic surprises in North America north of Mexico. Annals of the Missouri Botanical Garden 87:81—-109. FAITH, D. P. 1996. Conservation priorities and phyloge- netic pattern. Conservation Biology 10:1286—1289. Foster, R. I. 1972. Explosive chromosome evolution in 228 Downingia yina. Ph.D. dissertation. University of California, Davis. Fu, N. 1997. Phylogeographic studies in Japanese alpine plants based on intraspecific chloroplast DNA varia- tions. Bulletin of the Biogeographical Society of Ja- pan 52:59-69. GRAHAM, A. 1999. Late Cretaceous and Cenozoic history of North American vegetation: North of Mexico. Ox- ford University Press, New York, NY. Hivuis, D. M., C. Moritz, AND B. K. MABLE (eds.). 1996. Molecular systematics. Sinauer Associates, Sunder- land, MA. HENNIG, W. 1966. Phylogenetic systematics. University of Ulinois Press, Urbana, IL. HICKMAN, J. C. (ed.). 1993. The Jepson manual: Higher plants of California. University of California Press, Berkeley, CA. HILL, S. R. 1993. Sidalcea. Pp. 755-760 in J. C. Hickman (ed.), The Jepson manual: Higher plants of California. University of California Press, Berkeley, CA. HuBer, J. T. 1998. The importance of voucher specimens, with practical guidelines for preserving specimens of the major invertebrate phyla for identification. Journal of Natural History 32:367—385. Keck, D. D. 1959. Subtribe Madiinae. Pp. 1106-1129 in P. A. Munz, A California flora. University of Cali- fornia Press, Berkeley, CA. MAppIison, W. P. 1997. Gene trees in species trees. Sys- tematic Biology 46:523-—536. Markos, S. 2000. Evolutionary patterns in Lessingia (Compositae, Astereae) using morphological, molec- ular, and biosystematic data. Ph.D. dissertation. Uni- versity of California, Berkeley. AND B. G. BALDwIN. 2001. Higher-level relation- ships and major lineages of Lessingia (Compositae, Astereae) based on nuclear rDNA internal and exter- nal transcribed spacer (ITS and ETS) sequences. Sys- tematic Botany 26:168—183. MIsHLER, B. D. 1995. Plant systematics and conservation: Science and society. Madrono 42:103—113. . 2000a. Deep phylogenetic relationships among ‘plants’ and their implications for classification. Taxon 49:661—683. . 2000b. The need for integrated studies of the Cal- ifornia flora. Madrono 47: this issue. Moritz, C. 1994. Defining ‘“‘Evolutionarily Significant Units”’ for conservation. Trends in Ecology and Evo- lution 9:373—375. . 1995. Uses of molecular phylogenies for conser- vation. Philosophical Transactions of the Royal So- ciety of London. Series B, Biological Sciences 349: 113-118. 1999. Conservation units and_ translocations: Strategies for conserving evolutionary processes. Hereditas 130:217—228. , AND D. P. Faitu. 1998. Comparative phylogeog- raphy and the identification of genetically divergent areas for conservation. Molecular Ecology 7:419— 429. OLSEN, K. M. AND B. A. SCHAAL. 1999. Evidence on the origin of cassava: Phylogeography of Manihot escu- lenta. Proceedings of the National Academy of Sci- ences of the United States of America 96:5586—5591. ORNDUFF, R. 1966. A biosystematic survey of the gold- field genus Lasthenia. University of California Pub- lications in Botany 40:1—92. . 1993. Lasthenia. Pp. 298-300 in J. C. Hickman MADRONO (ed.), The Jepson manual: Higher plants of California. | University of California Press, Berkeley, CA. RAVEN, P. H. AND D. I. AXELROD. 1978. Origin and rela- | tionships of the California flora. University of Cali- | fornia Publications in Botany 72:1—134. RIDDLE, B. R., D. J. HAFNER, L. EF ALEXANDER, AND J. R. JAEGER. 2000. Cryptic vicariance in the historical as- sembly of a Baja California Peninsular Desert biota. Proceedings of the National Academy of Sciences of | the United States of America 97:14438—-14443. RIESEBERG, L. H., D. E. SOLTIS, AND J. D. PALMER. 1988. A molecular reexamination of introgression between Helianthus annuus and H. bolanderi (Compositae). Evolution 42:227—-238. SCHAAL, B. A. AND K. M. OLSEN. 2000. Gene genealogies and population variation in plants. Proceedings of the National Academy of Sciences of the United States of America 97:7024—7029. , D. A. HAyworTuH, K. M. OLSEN, J. T. RAUSCHER, AND W. A. SMITH. 1998. Phylogeographic studies in plants: Problems and prospects. Molecular Ecology 7:465—474. SCHULTHEIS, L. M. 1999. Systematics of Downingia (Cam- panulaceae) and population structure of D. bicornuta. Ph.D. dissertation. University of California, Berkeley. SHAW, A. J. 2000. Molecular phylogeography and cryptic speciation in the mosses, Mielichhoferia elongata and M. mielichhoferiana (Bryaceae). Molecular Ecology 9:595-608. SKINNER, M. W. AND B. M. PAVLIK (eds.). 1994. California Native Plant Society’s Inventory of rare and endan- gered vascular plants of California, 5th ed. California Native Plant Society, Sacramento, CA. , D. P. Tipor, R. L. BITTMAN, B. ERTTER, T. S. Ross, S. Boyp, A. C. SANDERS, J. R. SHEVOCK, AND D. W. TayLor. 1995. Research needs for conserving California’s rare plants. Madrono 42:211-—241. SLATKIN, M. 1987. Gene flow and the geographic structure of natural populations. Science 236:787—792. SmocovitTis, V. B. 1997. G. Ledyard Stebbins, Jr. and the evolutionary synthesis (1924-1950). American Jour- nal of Botany 84:1625—1637. SoLtTis, D. E., P. S. SOLTIS, AND B. G. MILLIGAN. 1992. Intraspecific chloroplast DNA variation: Systematic and phylogenetic implications. Pp. 117-150 in P. S. Soltis, D. E. Soltis, and J. J. Doyle (eds.), Molecular systematics of plants. Chapman and Hall, New York, NY: , PS. SOLTIS, AND J. J. DOYLE (eds.). 1998. Mo- lecular systematics of plants II: DNA sequencing. Kluwer Academic Publishers, Boston, MA. , M. A. GITZENDANNER, D. D. STRENGE, AND P. S. SoLTis. 1997. Chloroplast DNA intraspecific phylo- geography of plants from the Pacific Northwest of North America. Plant Systematics and Evolution 206: 353-373. SoLTis, P. S. AND M. A. GITZENDANNER. 1998. Molecular systematics and the conservation of rare species. Con- servation Biology 13:471—483. STEBBINS, G. L., JR. 1950. Variation and evolution in plants. Columbia University Press, New York, NY. STORFER, A. 1999. Gene flow and endangered species translocations: A topic revisited. Biological Conser- vation 87:173—180. SWoFForRD, D. L., G. J. OLSEN, P. J. WADDELL, AND D. M. HILLis. 1996. Phylogenetic inference. Pp. 407—514 in D. M. Hillis, C. Moritz, and B. K. Mable (eds.), Mo- [Vol. 47 | 2000] lecular systematics, 2nd ed. Sinauer Associates, Sun- derland, MA. TANOwITZ, B. D. 1982. Taxonomy of Hemizonia sect. Ma- diomeris (Asteraceae: Madiinae). Systematic Botany 7:314-339. TEMPLETON, A. 1986. Coadaptation and outbreeding de- pression. Pp. 105—121 in M. Soulé (ed.), Conserva- tion biology: The science of scarcity and diversity. Sinauer Associates, Sunderland, MA. TREMBLAY, N. O. AND D. J. SCHOEN. 1999. Molecular phy- logeography of Dryas integrifolia: Glacial refugia and postglacial recolonization. Molecular Ecology 8: 1187-1198. VANE-WRIGHT, R. I., C. J. HUMPHRIES, AND P. H. WIL- LIAMS. 1991. What to protect? Systematics and the BALDWIN: DISCOVERY AND CONSERVATION OF PLANT LINEAGES 229, agony of choice. Biological Conservation 55:235— 254. VARGAS, P., L. CONSTANCE, AND B. G. BALDWIN. 1998. A new species of Sanicula (Umbelliferae/Apiaceae) en- demic to Baja California, Mexico. Brittonia 50:122— 126. WEILER, J. H., JR. 1962. A biosystematic study of the ge- nus Downingia. Ph.D. dissertation: University of Cal- ifornia, Berkeley. WENDEL, J. E AND J. J. DOYLE. 1998. Phylogenetic incon- gruence: Window into genome history and molecular evolution. Pp. 265—296 in D. E. Soltis, P. S. Soltis, and J. J. Doyle (eds.), Molecular systematics of plants II: DNA sequencing. Kluwer Academic Publishers, Boston, MA. MADRONO, Vol. 47, No. 4, pp. 230—236, 2000 THE NEED FOR INTEGRATED STUDIES OF THE CALIFORNIA FLORA BRENT D. MISHLER Jepson Herbarium, University Herbarium, and Department of Integrative Biology, University of California, Berkeley, CA 94720-2465 ABSTRACT The general field of systematic biology is an inclusive discipline that has taken great steps forward in the last two decades. New theories and methods have been developed for using character data to recon- struct phylogenetic relationships and thus improve classifications. Copious new sources of character data have become available from the molecular level. New analytical methods have been developed for using phylogenies to quantify biogeographic, ecological, and evolutionary patterns and test hypotheses about process. These new developments need to be integrated with traditional strengths in systematics such as collections-based research, floristics, and morphological/developmental studies, through greater commu- nication and collaboration. Because of its unique geology and biology, its long and intense history of study, and its outstanding botanical institutions, California can serve as an ideal model for a newly integrated approach to systematic biology. There is a danger, in this time of specialization and information overload, to become too narrowly entrenched in one’s main activities. This happens in the botanical community of California: an in-the- trenches conservationist can get frustrated with an herbarium researcher because of taxonomic name changes that may on the surface seem aggravating and pointless. A consultant doing field inventories might not see the need for all the theory and heavy- duty computation applied in academic ecology these days. A systematist comfortable with mor- phological characters in a large part of the flora might be suspicious of the proliferation of molec- ular characters and cladograms. A molecular sys- tematist might lose touch with the important mor- phological characters that should be useful for their work. Some professors in academia may get too far removed from the realities and training needs in the outside botanical community. The list could go on, but I hope the main take-home message is clear: we need some mutual enlightenment and under- standing. The study of plant systematics and evolution is an integrated whole. We need the floristic special- ists with their good field knowledge of geography and variation in plants. We need to add molecular characters to the mix, while not losing the ability to gather, evaluate, and use morphological and an- atomical characters. We need a strong, well-justi- fied theoretical framework within which to test hy- potheses. We need the powerful new analytic tools available in ecology and systematics to look at community structure and phylogenetic trees in more sophisticated ways. And most of all, we need to develop the best understanding we can of the flora at all levels, communicate it clearly to the rel- evant communities, and apply it to the urgent con- servation needs facing California (see also Baldwin 2000). Thanks to Jepson’s foresight in creating a practical botanical institution in the center of an academic hotbed, the Jepson Herbarium is perfectly | placed to span these different approaches and en- | courage much needed integrated studies, which we | attempted to do with this symposium. | The field of systematic biology has taken great — steps forward in the last two decades, in three ma- jor areas. One major step forward has come about © through the introduction of new theories and meth- ods for using character data to reconstruct phylo- genetic relationships and thus improve classifica- | tions. The second major step forward was the in- troduction of copious new sources of character data from the molecular level. The third major step for- ward was in the development of analytical methods to use phylogenies to quantify biogeographic, eco- logical, and evolutionary patterns and test hypoth- eses about process. I will go into each of these new developments in turn, with the goal being to show how they can be integrated with traditional strengths in systematics such as collections-based research, floristics, and morphological/developmen- tal studies to yield a truly integrated approach. Phylogenetic Systematics The need for phylogenetic classification. The main developments in systematic theory stem from the phylogenetic approach developed by Hennig (1966). I summarized the new developments in Ma- dronro in 1995, and they have been well outlined elsewhere in texts and reviews (Farris 1983; Sober 1988; Forey et al. 1992; Maddison and Maddison 1992; Mishler 1994, 2000). A quick outline should thus suffice here: the fundamental idea is that phy- logenetic branching events among lineages should be reconstructed using shared, homologous markers (Mishler 2000). The markers are characters that changed state along a lineage, thus serving as evi- dence in the future that the lineage once existed. The derived state of a marker when shared among 2000] contemporaneous taxa (called a synapomorphy) can thus be used to postulate the existence of a once- shared lineage uniting them. This hypothesis can be tested by other putative synapomorphies, and so on. In Hennigian systematics, classification follows the reconstruction of a well-supported phylogeny, and only hypothesized monophyletic groups should be named—these are groups that consist of all and only descendents of a common ancestor. Thus in current thinking there is a fundamental isomor- phism among synapomorphy, monophyly, and named taxa. Why has this three-fold parallelism been so widely accepted across the community? Phyloge- netic classifications may not always be the most practical, that is the synapomorphic characteristics may sometimes be hard to detect easily. There is an element of human choice in what we use our formal Linnaean nomenclatorial system for, so why choose phylogeny as a basis? To answer this ques- tion requires an examination of what classifications are for. Evolution is real, as are organisms (physi- ological units), lineages (phylogenetic units), and demes (interbreeding units), for example. On the other hand, our classification systems are obviously human constructs, meant to serve certain purposes of our Own: communication, data storage and re- trieval, predictivity, and studies of process. While the last purpose seems perhaps the most esoteric, the most important function of a classification is its theoretical meaning, as discussed by Mishler (2000). A classification should capture units partic- ipating in the most important causal processes op- erating in the system. In this way the classification will be maximally predictive and useful for prac- tical and theoretical purposes. In biology, our best understanding is that evolution is the most impor- tant process organizing biological diversity, and thus the field of systematics in general has settled on restricting the use of formal taxonomic names to represent phylogenetically natural, monophyletic groups. While this is a widely understood standard for modern systematics, recent developments in two areas deserve further, more detailed, discus- sion: species concepts and rank-free classification. Species. Given the above arguments, what should the species rank represent in the Linnaean classifi- cation system? There are two necessary parts to any species definition. The criteria by which organisms are grouped into taxa must be specified, as well as the criteria by which a taxon is ranked as a species rather than some other hierarchical level. Following the arguments given above supporting a Hennigian phylogenetic system of classification in general, the grouping criterion that should be used at the species level, as at all other levels, is monophyly (Mishler and Theriot 2000). Under this view, apomorphies are considered to be the necessary empirical evi- dence for unambiguous phylogenetic species, as for phylogenetic taxa of all ranks. It follows that re- MISHLER: INTEGRATED STUDIES OF THE CALIFORNIA FLORA 2351 productive criteria cannot be used to group organ- isms into phylogenetic species. The fundamental inappropriateness of using breeding compatibility in cladistic analysis is because the ability to inter- breed (potential or actual) is a symplesiomorphy (i.e., shared primitive characteristic) by definition, thus not a phylogenetically valid grouping criterion. The species ranking decision is currently forced be- cause systematists have legislatively constrained themselves to use a ranked Linnaean hierarchy (see the following section for a possible alternative). The ranking criteria for species should involve practical criteria such as the amount of character support for a group; the species could then be viewed as the smallest hypothesized monophyletic group with good support (the minimum-rank tax- on—see Baldwin 2000, this issue). The species ranking decision may also involve biological cri- teria in better known organisms, including repro- ductive criteria, e.g., the origin of a distinctive mat- ing system at a particular node or the acquisition of exclusivity (a condition in which each allele in a lineage is more closely related to another allele in the lineage than it is to an allele outside the lineage; Baum and Shaw 1995; Graybeal 1995). There are, of course, difficulties applying the concept of monophyly at this level. There are a number of different sources of homoplasy (incon- gruence between certain character distributions and the cladogram based on maximum parsimony), such as adaptive convergence, gene conversion, de- velopmental constraints, mistaken coding, lineage sorting, reticulation, etc. The last named factor is the most problematical because it involves the fun- damental model of reality underlying cladistic anal- ysis—the other factors are cases of mistaken hy- potheses of homology, whereas ‘“‘homoplastic”’ character distributions due to reticulate evolution involve true homologies whose mode of transmis- sion is not tree-like. As less inclusive levels in the genealogical hi- erarchy are considered there is an increasing prob- ability that reticulating (““hybridizing”’) events will occur, rather than the diverging phylogenetic rela- tionships assumed by the cladistic approach. How- ever, the problem of reticulation is not confined to the species level; indeed, reticulation can occur throughout the hierarchy of life, and so is a problem of more general difficulty, and one that is receiving more attention by systematists (e.g., McDade 1990, 1992). It is becoming clear that while a certain amount of reticulation does not preclude cladistic reconstructions of phylogeny, extensive reticulation can cause major problems. We can reconstruct re- lationships in the face of some amount of reticula- tion (how much is not yet established, but is ame- nable to study, e.g., McDade 1992). As with con- vergence, where the application of cladistic analysis provides the only rigorous basis we have for iden- tifying homoplasy and thus demonstrating non-par- simonious evolution, the only way we can identify 232 reticulation on the basis of character analysis alone is through the application of cladistic parsimony, followed by examination of homoplasy to attempt to discover its source (see discussion by Vrana and Wheeler 1992; Mishler and Theriot 2000). How modes of reticulation actually affect character dis- tributions on cladograms is a productive avenue for empirical and theoretical investigations. To summarize, a phylogenetic species concept (Mishler and Theriot 2000; not to be confused with the different phylogenetic species concepts of Cra- craft 1983; Nixon and Wheeler 1990) can be de- fined based on the following considerations. First, organisms should be grouped into taxa at all levels (including the lowest level, species) on the basis of evidence for monophyly; breeding criteria in par- ticular are not useful for grouping purposes. Sec- ond, criteria used to assign species rank to certain monophyletic groups must vary among different or- ganisms, but might well include ecological criteria or presence of breeding barriers in particular cases (see Mishler and Brandon 1987; Mishler and Ther- iot 2000 for elaboration). The need for rank-free classification. The above discussion assumes that the current Linnaean sys- tem of ranked classifications is to remain in place, thus the species ranking decision is forced because systematists have constrained themselves to use a ranked Linnaean hierarchy. An intriguing possibil- ity has arisen through recent suggestions for re- forming the Linnaean system by removing the con- cept of ranks (De Queiroz and Gauthier 1992). This proposed change would keep the hierarchy of named phylogenetic groups, but remove the ranks (including species) associated with the names. This would remove the arbitrariness of ranking decisions at the “‘species level’? (Mishler 1999). As the community has applied phylogenetic anal- ysis broadly at many levels, it has become clear that the ranks in the Linnaean system (orders, families, genera, etc.) are problematic for classification. The many changes that are needed to bring classifica- tion into line with our understanding of phylogeny, plus the sheer number of levels being resolved in the tree of life, have made the current system of nomenclature appear a bit outdated. There are not nearly enough ranks to suffice in classifying the tree of life, with its millions of branches. Furthermore, there are practical problems with the use of ranks. Most aspects of the current code, including priority, revolve around the ranks, which leads to instability of usage. The need to maintain the hierarchy of the ranks leads to names being changed without good reason. For example, when a change in relationships is discovered, say a cur- rent family is found to be nested cladistically inside another family, several names often need to be changed to adjust, including the names of groups whose circumscription has not changed. Frivolous changes in names often occur under the current MADRONO code, when authors merely change the rank of a | group without any change in postulated relation- | ships at all. The most important problem with ranked classi- | fications are that they lead to bad science, if a user | of a classification naively assumes that taxa at the | same rank are comparable in some way. The exist- | ing, ranked Linnaean nomenclatorial system is | based on a non-evolutionary world-view (with taxa | at the same rank being somehow equivalent in the | mind of the creator). Under an evolutionary world- | view, the ranks don’t make sense. Practicing sys- | tematists know that groups given the same rank | across biology are not comparable in any way (i.e., | in age, size, amount of divergence, diversity within, — etc.), but many users of classifications do not know | this. For example, ecologists or paleobiologists of- ten count numbers of taxa at a particular rank as an erroneous measure of “‘biodiversity.”’ I have argued (Mishler 1999) that the formal [Vol. 47 | ranks should be abandoned (including the species | rank), for efficient and accurate representation of phylogenetic relationships. Instead, names of clades | should be hierarchically nested uninomials regard- ed as proper names (although current usage should be followed as much as possible to retain links to | the literature and collections). A clade should retain | its name regardless of whether new knowledge | might change its phylogenetic position, thus in-— creasing nomenclatorial stability. Furthermore, since clade names would be presented to the com- munity without attached ranks, users would be en- | couraged to look at the actual attributes of the clades they compare, thus improving research in comparative biology. In the future, I hope that | ‘‘rank-free’”’ phylogenetic taxonomy will allow ef- ficient presentation of theoretically justified, maxi- mally useful classifications that will unify biology by providing a single, consistent framework for the | study of evolutionary and ecological processes at all levels. Molecular Data Many new data sets have been added to system- atics because of the availability of technology from molecular biology that allows relatively easy com- parative sequencing of genes (Soltis et al. 1998). It is important to note at the outset that these new molecular data are not meant to replace traditional morphological and anatomical characters in sys- tematics. On the contrary, a number of recent stud- ies in the field have shown that molecular charac- ters, while a very useful addition to systematics, are complementary to the traditionally used characters. These new characters should be added to appropri- ate morphological characters and used to improve our knowledge of plant relationships, including species circumscriptions as well as relationships among species, genera, and families. In turn, im- proved understanding of relationships is necessary 2000] to inform our practical uses of plants and our con- servation efforts. Various laboratory techniques have been devel- | oped for molecular systematic studies. Some of these, such as DNA hybridization and restriction fragment length polymorphisms (RFLPs) yield only distance information, which is difficult to apply to _ phylogenetic tree reconstruction since information about individual homologies is missing. Other tech- niques, including mapping of restriction enzyme sites and direct DNA sequencing methods, yield in- formation about specific characters. The latter tech- niques are thus more heavily favored, because char- acter-based methods (based on explicit evolution- ary models of homology) provide markers suitable for phylogenetic analysis, as described above. Speaking generally, molecular data do have a number of advantages for systematic studies (Mish- ler 1994). A large number of molecular characters is available for any given level of phylogenetic in- ference, which has proven to lead in many cases to increasingly better-supported hypotheses of rela- tionships. This advantage seems particularly true at low taxonomic levels, even within species, where morphological characters tend to be subtle and hard to define. On the other hand, molecular data have some disadvantages as well. There are problems with sampling at the molecular level—it is time consuming and expensive to sample within study taxa (to check for polymorphism) at the level that is possible for many morphological characters, and, of course, fossil taxa generally cannot be included. Point mutations in DNA are simple characters with few possible character-states, subject to parallel changes that can’t be detected easily except through their congruence with other characters on a clado- gram. Morphological characters have a number of ad- vantages of their own (Mishler 1994). They are of- ten complex in structure and development, with many possible character-states, thus allowing better supported initial hypotheses of homology. Sam- pling within study taxa to understand polymor- phism is often easier and cheaper than with molec- ular data. Many key morphological characters can be seen in well-preserved fossils, thus allowing in- clusion of completely extinct lineages, which can be essential to getting the correct tree. Morpholog- ical characters are, of course, subject to their own difficulties of interpretation, as compared to molec- ular data. There are usually many fewer characters, variation patterns can often be difficult to organize into discrete character-states, and convergence can lead to mistaken hypotheses of homology (of course, congruence can plague molecular characters as well). The generalized strengths and weaknesses of mo- lecular and morphological data are complementary to a large extent. Thus, the best approach, of course, is to apply appropriate characters from all levels of organization to some specific problem of MISHLER: INTEGRATED STUDIES OF THE CALIFORNIA FLORA 233 relationships. Molecular characters will remain es- sential as the bulk of available evidence, particu- larly for shallower branching events. Morphologi- cal characters will also remain critical pieces of ev- idence for many branch points in evolutionary his- tory, particularly the deeper ones; plus they are essential for integrating fossils into evolutionary trees and of course for identification purposes in practical applications of systematics such as florist- ics. Thus, the future clearly lies in studies integrat- ing both sources of data. Comparative Biology The interplay and contrast between phylogenetic and functional/structural groupings has ushered in a new era of scientific rigor in comparative biology with the development of explicit and testable hy- potheses of phylogenetic relationships. Many ad- vances have been made in improving evolutionary model building as a route to understanding; “‘tree- thinking” is now central to all areas of systematics and evolution. The central importance of phylogeny reconstruction in systematics, ecology, and evolu- tionary biology has become widely realized in re- cent years (Donoghue 1989; Funk and Brooks 1990; Wanntorp et al. 1990; Brooks and McLennan 1991; Harvey and Pagel 1991; Miles and Dunham 1993; Martins 1996). Explicit cladistic phylogenies now provide a critical basis for classification as well as for studies of speciation, biogeography, ecology, and behavior (among many other areas). The area of phylogenetic comparative methods is one of considerable controversy and rapid concep- tual development. Virtually every issue of major journals and each new book on systematics and evolution contains something of interest on this subject. The general working procedure is to first carefully define the causal hypothesis to be tested, then specify a null hypothesis (what you would ex- pect if the hypothesized cause is not working), and finally design a phylogenetic test that would let you reject the null hypothesis if it is indeed false. The large number of comparative methods can best be summarized by placing them into categories corresponding to the types of hypotheses meant to be tested, as addressed below. Comparing cladograms. These methods are meant for comparing different phylogenetic trees in the study of coevolution. Coevolution can be broad- ly defined as congruence between two or more sys- tems undergoing tree-like evolution (1.e., evolution by descent with modification). This is a generaliza- tion of the phylogeny/homology relationship (..e., the “‘coevolution” of organism lineages and char- acters discussed above). Coevolution comes in many forms: vicariance biogeography (organism/ earth coevolution), host/parasite relationships, com- munity evolution (e.g., symbionts, pollinator/plant coevolution, or other long-term ecological associ- ations). 234 Biogeography can serve as an example of the concept of comparing trees for their mutual infor- mation content. Historical biogeography has a long tradition in biology, and was indeed a major source of evidence for Darwin. After evolution became widely established as a principle, the initial ap- proach to biogeography was to look for areas of origin and dispersal patterns based on stable world geography (see Wiley 1981). The work of Hennig (1966) led to the development of phylogenetic bio- geography, which examined the distribution of one group at a time in relation to a cladogram. One famous outcome was Hennig’s “progression rule,” the observation that more derived species often tend to occur further from the initial area of a lin- eage following speciation by peripheral isolation. The phylogenetic examination of many groups at a time can be traced to the works of Croizot follow- ing a method he called “panbiogeography’’—a search for generalized distributional “‘tracks.”’ This search for matching geographic patterns led to the approach called vicariance biogeography, a search for sister groups sharing the same pattern across many cladograms (Nelson and Platnick 1981). The basic idea is to look for common patterns (and causes) of distribution—evidence from other organismal distributions can be relevant to under- standing the distribution of a particular group! Con- gruence is taken as evidence of shared biogeo- graphic history (vicariance); incongruence as evi- dence of separate history (dispersal). Methodolo- gies have diversified to compare cladograms in coevolutionary studies, including consensus tech- niques (Funk and Brooks 1990), tree-to-tree dis- tance metrics (Penny and Hendy 1985), and parsi- mony techniques (such as Brooks parsimony; see Brooks 1990; Brooks and McLennan 1991). Comparing clades within a cladogram. These methods are meant to detect whether there are im- balances in symmetry between sister clades in the same cladogram, in order to address various ques- tions in both micro- and macro-evolution. First of all, however, what is the null expectation? Intuitive- ly, one might expect balanced trees, perhaps, based on some sort of false analogy to coin flips. But is this a correct assumption? “‘Random”’ trees can be generated in many ways (Maddison and Maddison 1992), and include equiprobable trees (picked out of a set of all possible trees—bias towards asym- metry), random joining trees (models a random speciation process—intermediate bias), or a random partition (bias towards symmetry). Using a Yule “pure birth’? Markovian model to grow random trees, Slowinski and Guyer (1989) showed a non- intuitive result: the probability of each way of par- tioning taxa at a bifurcating node is equal [for n terminal taxa, the probability of generating any di- vision of species above a node into sister lineages of unequal size is 2/(n — 1); the probability is 1/(n — 1) for evenly divided sister lineages]. Thus, even MADRONO [Vol. 47 | a node in which one species is the sister taxon to | 39 other species is not significantly unbalanced at the P = 0.05 level (P > 0.051). | This work has lead to the realization that real — trees should be expected to be quite asymmetrical even under a random model. Furthermore, even if trees are judged significantly asymmetric, how can | we associate that judgement with some specific fac- | tor postulated to be the cause of that asymmetry? That leads to the hot topics of ‘‘key innovations” | and ‘‘adaptive radiations.’”’ There have been many, | often conflicting definitions of adaptive radiations (Givnish and Sytsma 1997). Decomposing the term is best, and suggests that ‘‘adaptation’’ needs to be | established separately from ‘“‘radiation.”’ The rapid | diversification of lineages (caused by a postulated © ‘key innovation’’) should be accompanied by eco- | logical, morphological, and/or genetic diversifica- _ tion. A number of methods have been developed to | deal with the required time estimation problem, | which involves two questions: Can we assume a | molecular clock? If we can, how do we calibrate it | (Sanderson and Wojciechowski 1996; Sanderson | 1997; Baldwin and Sanderson 1998)? | Discrete-state character comparisons on a cladogram. These methods are meant for examin- ing how discrete-state characters evolve on a tree individually and together. Such characters can be mapped onto cladograms using parsimony, so as to minimize the number of character-state changes. In this way, suites of characters are built up for Hy- pothetical Taxonomic Units (HTU’s). Specific types of hypotheses that can be tested include polarity of character-state changes in one character, and the as- sociation of state changes in two characters, either undirected (Ridley’s test; Ridley 1983) or directed (Maddison’s test; Maddison and Maddison 1992). Most of these studies are motivated by the search for adaptation. There is a long-standing observa- tion that organisms tend to match their environ- ment. Darwin and many Darwinians thought that all structures must be adaptive for something. But this assumption has come under severe challenge in recent years (Gould and Lewontin 1979). Not all structures and functions are adaptive. In fact, there are very few completely demonstrated examples of adaptations. The definition of adaptation in a formal sense requires fulfillment of four different criteria (Mish- ler 1988; Brandon 1990): 1. Engineering. Structure must indeed function in hypothesized sense. Requires functional tests. 2. Heritability. Differences between organisms must be passed on to offspring, at least proba- bilistically. Requires heritability tests (parent- offspring correlations; common garden studies). 3. Natural Selection. Difference in fitness must oc- cur because of differences in possession of the hypothesized adaptation in a common environ- ment. Requires fitness tests. | f | } | | 4. Phylogeny. Hypothesized adaptive state must have evolved in the context of the hypothesized cause. Requires phylogenetic tests. Only something that passes all these tests is an adaptation. If it passes tests 1—3, it can be called an aptation. If it then fails test 4 it can be called an exaptation (Gould and Vrba 1982). Thus, a phy- logenetic test, while not sufficient in itself, is nec- essary as part of a complete adaptive explanation (Coddington 1988; Mishler 1988; Donoghue 1989). Continuous character comparisons on a clado- gram. These methods are meant for examining how quantitatively varying characters are associated on phylogenies. Note that these are characters that do not meet the ‘discrete-state’ criteria for taxonomic characters. The “‘bad old way’? to compare two such characters was through direct correlations of species values (using species as data points). How- ever, as pointed out by Felsenstein (1985) and oth- ers, this treats species as if they are all equally re- lated to each other. The advent of quantitative com- parative approaches was motivated by attempting to ‘“‘remove’”’ the influence of history, for example using ANOVA and ANCOVA (Harvey and Pagel 1991), autocorrelation (Cheverud and Dow 1985), independent contrasts (Felsenstein 1985; Burt 1989), and general linear model approaches to par- tition variance and “‘subtract”’ the phylogenetic ef- fects (Martins 1996). Conversely, other methods explicitly describe variation due to phylogeny by tracing the quantitative characters on a phylogenet- ic tree, reconstructing values for nodes, and looking at direction of change by comparing ancestors and descendants (e.g., Huey and Bennett 1987). The Integrative Approach These diverse sources of data, complex theories, mathematically complicated algorithms, and multi- ple approaches to analysis have reinvigorated the field of plant systematics, yet at the same time they have made the field more complex and harder to master. No one person can keep ahead of all these parts of the whole endeavor. Thus, there will be an increasing need for mutual understanding among Specialists, increased collaborative research, and more sharing of expertise. Training of students must continue to diversify into all the new ap- proaches, while at the same time not losing sight of older, still valuable approaches. Botanical insti- tutions need to adapt and expand their vision and capabilities. Our ultimate goal for the next 50 years of California botany should be to serve as a model by developing integrated studies that combine all these approaches and presenting this information in easily accessible ways to the public. ACKNOWLEDGEMENTS I thank Bruce Baldwin for comments on the manuscript, and the National Science Foundation (PEET grant DEB- MISHLER: INTEGRATED STUDIES OF THE CALIFORNIA FLORA 235 9712347) for supporting the development of some of these ideas. LITERATURE CITED BALDWIN, B. G. 2000. (this issue). AND M. J. SANDERSON. 1998. Age and rate of di- versification of the Hawaiian silversword alliance (Compositae). Proceedings of the National Academy of Sciences of the United States of America 95:9402— 9400. Baum, D. A. AND K. L. SHAW. 1995. Genealogical per- spectives on the species problem. Pp. 289-303 in P. C. Hoch and A. G. Stevenson (eds.), Experimental and molecular approaches to plant biosystematics. Monographs in systematics, Vol. 53. Missouri Botan- ical Garden, St. Louis. BRANDON, R. N. 1990. Adaptation and environment. Princeton University Press, Princeton, NJ. Brooks, D. R. AND D. MCLENNAN. 1991. Phylogeny, ecol- ogy, and behavior. University of Chicago Press. . 1990. Parsimony analyisis in historical biogeog- raphy and coevolution: Methodological and theoreti- cal update. Systematic Zoology 39:14—30. Burt, A. 1989. Comparative methods using phylogeneti- cally independent contrasts. Oxford Surveys in Evo- lutionary Biology 6:33—53. CHEVERUD, J. AND M. Dow. 1985. An autocorrelation anal- ysis of the effect of lineal fission on genetic variation among social groups. American Journal of Physical Anthropology 67:113—121. CODDINGTON, J. 1988. Cladistic tests of adaptational hy- potheses. Cladistics 4:3—22. CRACRAFT, J. 1983. Species concepts and speciation anal- ysis. Current Ornithology 1:159—187. DE QUEIROZ, K. AND J. GAUTHIER. 1992. Phylogenetic tax- onomy. Annual Review of Ecology and Systematics 23:449—480. DONOGHUE, M. J. 1989. Phylogenies and analysis of evo- lutionary sequences, with examples from seed plants. Evolution 43:1137—1156. FARRIS J. S. 1983. The logical basis of phylogenetic anal- ysis. Pp. 7-36 in N. Platnick and V. Funk (eds.), Ad- vances in cladistics, Vol. 2. Columbia Univ. Press, NY. FELSENSTEIN, J. 1985. Phylogenies and the comparative method. American Naturalist 125:1—15. Forey, P. L., C. J. HUMPHRIES, I. L. KITCHING, R. W. SCoT- LAND, D. J. SIEBERT, AND D. M. WILLIAMS. 1992. Cla- distics. The Systematics Association Publication No. 10. Oxford University Press, NY. Funk, V. A. AND D. R. Brooks. 1990. Phylogenetic sys- tematics as the basis of comparative biology. Smith- sonian Institution Press, Washington, DC. GIvNISH, T. J. AND K. J. SYTSMA (eds.) 1997. Molecular evolution and adaptive radiation. Cambridge Univer- sity Press, Cambridge, England, UK. GouLb, S. J. AND R. LEWONTIN 1979. The spandrels of San Marco and the Panglossian Paradigm: a critique of the adaptationist programme. Pp. 147—164 in J. Maynard Smith and R. Holliday (eds.), The evolution of adaptation by natural selection. Royal Society of London. AND E. S. VRBA 1982. Exaptation—a missing term in the science of form. Paleobiology 8:4—15. GRAYBEAL, A. 1995. Naming species. Systematic Biology 44:237-250. Harvey, P. H. AND M. D. PAGEL. 1991. The comparative 236 method in evolutionary biology. Oxford University Press. HENNIG, W. 1966. Phylogenetic systematics. University of Illinois Press, Urbana. Huey, R. B. AND A. E BENNETT. 1987. Phylogenetic stud- les of co-adaptation: preferred temperature versus op- timal performance temperatures of lizards. Evolution 41:1098-115. MAaAppison, W. P., AND D. R. MApDpISON. 1992. MacClade: analysis of phylogeny and character evolution (ver- sion 3). Sinauer Associates, Sunderland, MA. MartTINS, E. P. 1996. Phylogenies, spatial autoregression, and the comparative method: a computer simulation test. Evolution 50:1750—1765. McDapbk, L. 1990. Hybrids and phylogenetic systematics. I. Patterns of character expression in hybrids and their implications for cladistic analysis. Evolution 44: 1685-1700. . 1992. Hybrids and phylogenetic systematics. I. The impact of hybrids on cladistic analysis. Evolution 46:1329-1346. Mites, D. B. AND A. E. DUNHAM. 1993. Historical per- spectives in ecology and evolutionary biology: the use of phylogenetic comparative analyses. Annual Review of Ecology and Systematics 24:587—-619. MISHLER, B. D. 1998. Reproductive ecology of bryo- phytes. Pp. 285-306 in J. Lovett Doust and L. Lovett Doust (eds.), Plant reproductive ecology. Oxford Uni- versity Press. NY. . 1994. The cladistic analysis of molecular and morphological data. American Journal of Physical Anthropology 94:143—156. . 1995. Plant systematics and conservation: science and society. Madrono 42:103—113. . 1999. Getting rid of species? Pp. 307-315 in R. Wilson (ed.), Species: new interdisciplinary essays. MIT Press. Cambridge, MA. . 2000. Deep phylogenetic relationships among “plants” and their implications for classification. Taxon 49:661—683. AND R. N. BRANDON. 1987. Individuality, plural- ism, and the phylogenetic species concept. Biology and Philosophy 2:397—414. AND E. THERIOT. 2000. The phylogenetic species MADRONO [Vol. 47 | concept sensu Mishler and Theriot: monophyly, apo- | morphy, and phylogenetic species concepts. Pp. 44—| 54, 119-132, 179-184 in Q. D. Wheeler and R. | Meier (eds.), Species concepts and phylogenetic the- | ory: a debate. Columbia University Press, NY. [three | chapters] NELSON, G. AND N. PLATNICK. 1981. Systematics and bio- | geography. Cladistics and vicariance. Columbia Uni- | versity Press, NY. Nixon, K. C. AND Q. D. WHEELER. 1990. An amplification of the phylogenetic species concept. Cladistics 6:21 1-— | 223; PENNY, D. AND M. D. HENDy. 1985. The use of tree com- | parison metrics. Systematic Zoology 34:75-82. | RIDLEY, M. 1983. The explanation of organic diversity: | the comparative method and adaptations for mating. | Oxford University Press, NY. SANDERSON, M. J. 1997. A nonparametric approach to es- — timating divergence times in the absence of rate con- | stancy. Molecular Biology and Evolution 14:1218- 1231) AND WOICIECHOWSKI, M. F 1996. Diversification | rates in a temperate legume clade: are there ““so many _ species” of Astragalus (Fabaceae)? American Journal | of Botany 83:1488—1502. SLOWINSKI, J. B. AND C. GUYER, 1989. Testing the sto- chasticity of patterns of organismal diversity: an im- proved null model. American Naturalist 134:907-__ O21 SoBER, E. 1988. Reconstructing the past. MIT Press, Cam- | bridge, MA. SoLtis, D. E., P. S. SOLTIS, AND J. J. DOYLE (eds.). 19983) Molecular systematics of plants II: DNA sequencing. Kluwer Academic Publishers, Boston. VRANA, P. AND W. WHEELER. 1992. Individual organisms as terminal entities: laying the species problem to rest. Cladistics 8:67—72. : WANNTORP, H.-E., D. R. BROOKS, T. NILSSON, S. NYLIN, FE | Ronguist, S. C. STEARNS, AND N. WEDELL. 1990. — Phylogenetic approaches in ecology. Oikos 57:119— 132, WILEY, E. O. 1981. Phylogenetics: the theory and practice of phylogenetic systematics. John Wiley and Sons, New York. HMADRONO, Vol. 47, No. 4, pp. 237-252, 2000 OUR UNDISCOVERED HERITAGE: PAST AND FUTURE PROSPECTS FOR SPECIES-LEVEL BOTANICAL INVENTORY BARBARA ERTTER University and Jepson Herbaria, University of California, Berkeley, CA 94720-2465 ABSTRACT On-going botanical field exploration and the synthesis of resultant data into species-level plant distri- bution information in the United States has been handicapped by multiple assumptions: results of such an effort would have little or no pragmatic implication; all necessary work has already been completed, and the resultant information just needs to be compiled within a modern informatics framework; herbarium vouchers are not only already sufficient but become peripheral once label data are captured; further contribution by the systematics community is likewise peripheral, except for a trickle of new species descriptions that can be readily accommodated; species-level field exploration within the United States is neither science nor fundable; and a comprehensive species-level inventory is simply too big a project to tackle. To address these assumptions, a brief account of botanical surveys contemporaneous with topo- graphic mapping efforts of the U.S. Geological Survey is presented, with parallels drawn where possible. Botanical inventory efforts of the University of California at Berkeley are likewise presented, including involvement in Wieslander’s Vegetation Type Mapping Project and Bailey and Bailey’s project to map the vegetation of Western National Parks. The cumulative result of these and other efforts, however, leave us with an estimated 5% of the national vascular flora still to be described, and distributional information of the known species falling far short of what is needed for informed decision-making. Simple accretion of additional distributional reports is not sufficient, but needs to be based on vouchered reports that have been critically evaluated within taxonomic models by members of the systematics community. The con- clusion is therefore that standing assumptions are unjustified, and that a large-scale biodiversity counter- part to the topographic efforts of the U.S. Geological Survey is in fact a realistic and desirable goal. During this past century, organized field explo- ration of botanical diversity and the synthesis of resultant data into species-level plant distribution information in the United States have become somewhat passé, at least in scientific realms. At best, earlier cataloguing of species has been tran- scended by vegetation mapping efforts undertaken by plant ecologists, often completely decoupled from plant systematists who are expected to focus on phylogenetic analyses. As addressed by subse- quent papers in this symposium (1.e., Baldwin [2000], Charlet [2000]), both phylogenetic analysis and vegetation mapping represent extremely pro- ductive arenas of research, having conservation and land management significance well beyond pure science. The question nevertheless needs to be an- swered: should the age of species-level botanical surveys indeed be properly relegated to the past, or is there instead not only a legitimate opportunity but a crying need for seriously supported species- level field exploration and the synthesis of resultant information within a scientifically valid frame- work? To answer this question, one must first analyze the real and perceived obstacles to species-level bo- tanical inventory in the United States. Prominent among these would be the following assumptions: @ The results of such an effort would have little or no pragmatic implication. @ All necessary work has already been completed, and the resultant information just needs to be compiled within a modern informatics frame- work. @ Herbarium vouchers are not only already suffi- cient, but become peripheral once label data are captured. @ Further contribution by the systematics commu- nity is likewise peripheral, except for a trickle of new species descriptions which can be readily accommodated by the environmental sciences and informatics communities. @ Species-level plant inventory in the United States is not sufficiently scientific, innovative, or oth- erwise high-profile to merit funding. @ A comprehensive species-level inventory is sim- ply too big a project to tackle (hence the short- cuts of vegetation mapping, indicator species, umbrella species, etc.) Just how valid are these assumptions? Are they supported by either the historical record or modern- day realities? And if not valid, what are the impli- cations for such modern-day issues as biodiversity conservation, which relies heavily on comprehen- sive, accurate distributional information as the basis for critical land-management decisions? The pur- pose of the current paper is to address these ques- tions, beginning with (and leaning heavily on) the historical context and precedents. i) eS) oe) LESSONS FROM THE PAST The fate of national surveys. When a modern bi- ologist is presented with the question, ‘What new- ly-established systematic survey of the United States was dealt a serious set-back by a hostile Con- gress on the grounds of being unrealistically am- bitious, too costly and long-lasting, scientifically suspect, of limited (or negative) value, too much power concentrated in a single federal bureaucracy, contrary to the ideals of private property and free- market enterprise, and a threat to western state’s rights?’ the abortive attempt to establish a National Biological Survey in the early 1990’s readily comes to mind. This ambitious proposal, outlined by the National Research Council of the National Acade- my of Sciences (1993), was a primary target of the heavily conservative 104th Congress’s ‘Contract with America,’ surviving on a significantly dimin- ished scale within the United States Geological Sur- vey (USGS) (Wagner 1999). Although the resultant Biological Resources Division, established in 1996, has certainly generated its share of solid biological research, the discrepancy between the original vi- sion and the eventual outcome left the concept of a full-fledged biodiversity survey of the nation dis- credited, and many of its proponents disheartened. What has been forgotten, however, is that the de- scription applies equally well to the founding of that venerable institution, the USGS itself, under the directorship of John Wesley Powell. Although Powell is best known nowadays for his heroic ex- ploration of the Grand Canyon, his biography by Stegner (1953) illuminates clearly the pivotal role that Powell played in the development of land man- agement principles and governmental agencies that went hand-in-hand with the setthement of the west- ern United States. Powell advocated comprehensive topographic mapping of areas being opened for set- tlement as an essential precursor to rationale land- use planning. He focused particularly on water rights and grazing allocations, recommending Mor- mon-style cooperative irrigation districts and 2500 acre grazing units (Goetzmann 1966). These pro- posals, however, were not well received by the ma- jority of Westerners at the time, giving rise to pro- tracted political battles that remain part of our pres- ent legacy. Nevertheless, in spite of vehement opposition to Powell’s vision and efforts, the long-term result in the topographic realm was success beyond Powell’s most ambitious dream. As noted by Stegner (1953), seven decades after the initiation of the USGS over half of the United States had been topographically mapped, but only about one-fourth on the scale needed for contemporary planning. As a result, in 1953 there were more than two dozen government bureaus engaged wholly or partially in the prepa- ration and printing and use of maps. Stegner also noted that this success resulted in spite of Powell’s gross under-estimation of the task (1.e., 24 years at MADRONO [Vol. 47 a cost of $18 million), and that even though ‘Some members of Congress a little later were ready to bet him that he couldn’t do it in a hundred years for a hundred million ... they ignored what was palpably true, that the maps were worth anything they cost, and more’ (p. 280). If a parallel effort had been undertaken for biological mapping over the same span of time, would we now likewise take for granted the worthiness of the effort and value of the results? The California geological survey. As it happens, various state and federally sponsored surveys that served as precursors to the USGS had generally included a biological component. As a prime ex- ample, the California Geological Survey (CGS) of | 1860-1873, under the direction of J. D. Whitney, included in its mission ‘a full and scientific descrip- tion of its rocks, fossils, soils, and minerals, and of | its botanical and zoological productions, together | with specimens of the same, which specimens shall be properly labeled and arranged, and deposited in such place as shall be hereafter provided for that purpose by the legislature’ (quoted in Brewster 1909, p. 185). In contrast to the USGS, state sur- | | veys in general enjoyed widespread support as ‘a | hallmark of enlightened state administration, a source of local cultural pride, and the means where- by exploitable resources might be cheaply located and advertised to would-be investors’ (Goetzmann 1966, p. 355). The CGS likewise enjoyed public acclaim initially, but popular support quickly dwin- dled when the anticipated flood of economic ben- efits did not immediately materialize. As bemoaned by Whitney: friends if I would be their confidential adviser in their interest in claims and stocks, but as it is, I do ‘State officers would be my best | not know one of them who cares a rye-straw for | the work [of the Survey]’ (quoted in Leviton and Aldrich 1997, p. 66). The prolonged decline of the CGS, as a result of political and economic interests independent of sci- | entific value, left Whitney increasingly dispirited | and disillusioned. This progression is wonderfully captured in a series of letters from Whitney to his brother, with a political cynicism that still resonates strongly (quoted in Brewster 1909, pp. 264—266): 26 February 1868: The prospects of the survey remain as uncertain as ever. Two committees have been at the office and exhibited even more | than their usual amount of stupidity and igno- rance. Since the Yosemite Valley bill passed over | the Governor’s veto, I feel so disgusted with Cal- ifornia that I can hardly stand it much longer. Still I am running the survey along in a small way at my own expense, waiting to see what the jackasses at Sacramento will do. 29 March 1868: We have had a nice little time of it in the legislature. The petroleum and other swindlers made a dead set against the survey and 2000] killed it, having malleable material to work with in the Democratic legislature. ... We were es- pecially unfortunate in having in the Senate . a former United States Surveyor General, under whose administration the fraudulent surveys in the southern part of the state were made, and the character of which is being exposed as fast as our work covers the ground. Of course he found against us with all his might. And finally (pp. 289-290): 19 March 1874: The survey has succumbed to the stupidity and malignity of the legislature, backed by the same characteristics on the part of the Governor. ... My own feelings are decidedly those of relief at getting the survey off my hands, with no fault of laches of my own, for it is hard work making a creditable thing of it on a small amount of money. I have always got more curses than coppers out of it. Also working against the CGS, and as a parallel to Powell, Whitney himself had drastically under- estimated the scale, to the extent that it was im- possible for him to follow through on what he had originally promised: I have found out that the State of California is a prodigiously large one. Not that I did not know it before; but now I have a realizing sense of it. It is as big as Great Britain, Ireland, Belgium, Hanover, and Bavaria put together! If I had a complete map of the state, a corps twice as large as I now have, and worked as fast (on the geol- ogy only) as the English government surveyors do, I should finish in just 150 years. Having our own maps to make, our labor is tripled; and con- sequently we shall be through in 450 years or thereabouts. (quoted in Brewster 1909, pp. 197-198) With various ups-and-downs and overall dwin- dling support, the CGS nevertheless struggled along until finally giving up the ghost in 1873. The bio- logical component, however, had been eliminated some years earlier. The initial survey crew included William H. Brewer as both botanist and Whitney’s second-in-command. When Brewer departed in 1864 to accept a professorship in Yale, he was nev- er replaced in kind. Instead, Henry N. Bolander was hired on a contract basis, as funding allowed, to do botanical surveys in parts of the state unvisited by Brewer (Jepson 1898; Ertter 2000b). The several thousand specimens accumulated by Brewer, Bo- lander, and others formed the basis for the first complete flora of California (Brewer et al. 1876; Watson 1880), compiled at Harvard University by Brewer, Asa Gray, and Sereno Watson, with treat- ments of specific groups (e.g., mosses) provided by an appropriate specialist. ERTTER: UNDISCOVERED HERITAGE P18) Although the original state legislation mandated that the reports of the CGS be copyrighted and sold for the benefit of the common school fund, no funds were allocated, so that Brewer’s efforts took the form of a labor of love: I received no pay whatever after the closing of my connection with the Survey of California, neither for the time nor the expense in working up results. I spent an aggregate of two years time, a litthe more rather than less, and over two thousand dollars in cash, besides deducting an- other one thousand dollars from my salary from college because of time taken out from my work, that is, absence during term time at work on my plants at the Cambridge Herbarium. (quoted in Farquhar 1930, p. xxiii) Whitney was eventually able to secure additional State funding for the publication of several Survey reports, but not for botany. Instead, a select group of California’s wealthier citizens, including Leland Stanford, provided the necessary funds from their own pockets (Brewer et al. 1876). The key lesson to be learned from this look at the history of the CGS, and the fate of the biolog- ical component in particular, is that termination of support resulted not from completion of the scien- tific goals, but because of unmet expectations, spe- cial interests, and pure politics. We can only won- der what legacy of critical biodiversity information would have accrued if the California Geological Survey had continued to the present, as has been the situation in several other states (e.g., [linois, New York). Perhaps, as noted by Brewster (1909), the collapse of the CGS was inevitable at the time: California, in 1860 when the survey began, looked to a future of unlimited growth and pros- perity, and cut its coat according to the cloth it expected to own. Its actual lot was flood and drought, and the Civil War. Under these changed conditions, there were many well-intentioned persons who felt that elaborate, hand-colored monographs on birds and land-shells were not the things the young state needed most. As it turned out, the California Survey, on the scale on which Whitney planned it, was distinctly premature. (p. 301) Valid as Brewster’s evaluation may have been to California of the late 19th century, it rings hollow in the face of early 21st century realities. Political and economic interests remain, but one can no lon- ger argue that a full-fledged biological survey of California, as envisioned by Whitney and as called for by the state legislature in 1860, would still be ‘premature.’ National biological surveys contemporaneous with the USGS. Federally supported surveys of 240 western lands began with the Lewis and Clark ex- pedition of 1804—1806 and reached a heyday in the mid-1800’s with a broad selection of surveys for railroad routes, boundary delimitations, and general exploration. Most of these surveys included a bo- tanical component and are an important part of the historical record. The focus here, however, is on biological surveys that were contemporaneous with the USGS and that could potentially have served as counterparts. As well described by Goetzmann (1966), the USGS resulted from the coalescence of three com- peting federally-sponsored surveys in the 1870’s: that of Lieutenant George Montague Wheeler, rep- resenting the last attempt by the military to hold onto its former domination of western exploration; those led by Ferdinand Vandiveer Hayden under the aegis of the fledgling Interior Department; and the early efforts of Powell himself, backed by a diver- sity of sources, both private and government, and including the scientific community as represented by the Smithsonian Institution. Each of these three pre-USGS surveys contained a botanical compo- nent, with that of Powell’s being weakest, and none providing more than the scantiest opportunity for collecting. This can be seen in the report by Town- shend Stith Brandegee, whose botanical career be- gan as part of the Hayden expeditions: Attached to the division of the San Juan as as- sistant topographer, as much time as possible was given to the botany of the country through which our work obliged us to pass. Under such circum- stances, it was impossible to make a complete botanical collection of the district to our division; therefore no plants were gathered excepting such as seemed to be additions to the flora of Colo- rado, as published by the Survey in Miscella- neous Publications, No. 4 [Porter & Coulter, 1874]. The collections and notes were almost all made while riding from one topographical station to another. (Brandegee 1876) J. T. Rothrock, who joined Wheeler’s expedition as botanist-surgeon in 1873 (Kelly 1914), not only collected plants for systematic analysis, but also helped initiate a new dimension of botanical survey by addressing economic aspects of the vegetation. Rothrock’s reports included a forerunner of a con- servation ethic (at least within the context of the time), quotes by Muir, and (similar to Powell) sug- gestions that governmental involvement might be appropriate: e.g., ‘In view ... of the acknowledged fact that in our older and more densely populated States we have an impending dearth of timber, would not a wise political economy endeavor to obviate such a result in our Western regions? Tree destruction began with us as a necessity, but it has been matured into an instinct’? (Rothrock 1878, p. 34). According to Goetzmann (1966), similar atti- MADRONO [Vol. 47 tudes characterized both Whitney and Wheeler, pre- saging the battles fought by Powell. George Vasey, who accompanied Powell’s 1868 expedition (Canby and Rose 1893), was in 1872 appointed Botanist to the U.S. Department of Ag- | riculture (USDA), which at that time housed the | botanical collections that had accumulated from | various exploring expeditions. These collections had been in the custody of John Torrey at Columbia College in New York, with additional oversight provided by Asa Gray as a Regent of the Smith- sonian Institution. A few years before his death, Torrey relinquished custodianship, and, in lieu of — suitable facilities in the Smithsonian building itself, the collections were deposited with the USDA in | 1868. They were not turned over to the Smithson- | ian Institution until 1894, to be merged with a sep- | arate plant collection that had been initiated by Les- — ter Ward, paleobotanist for USGS, giving rise to the © U.S. National Herbarium (Morton and Stern 1966). Vasey replaced Charles Christopher Parry, one of the premier field botanists associated with several earlier federally sponsored expeditions (e.g., ry’s abrupt dismissal in 1871 sheds considerable | light on the attitudes behind the declining status that descriptive botany and accompanying herbari- um specimens had already attained by this time. According to Frederick Watts, Parry’s superior as Commissioner of Agriculture, in a letter to Torrey, Gray, Brewer, and D. C. Eaton (reprinted in Gray 1871): . [NJothing at all had been done by Dr. Parry the | Mexican Boundary Survey). The rationale for Par- | beyond his attention to the preservation of the | herbarium. This Department is designed to ren- der the developments and deductions of science directly available to practice, that farmers and horticulturalists may be benefited by them. The principles of vegetable physiology, their relations to climate, soils, and food of plants, and the dis- eases of plants, which are principally of fungoid > origin, it is clearly the duty of a botanist to in- vestigate. If possible, he should throw some light upon the origin and condition of growth of the . lower orders of cryptogamic botany. This is a> domain into which I could not discover that Dr. Parry had ever entered, so far as his practical | work here gave any indication. The routine op- | erations of a mere herbarium botanist are prac- tically unimportant. Further prodding by Gray yielded the information | that insubordination, at least as perceived by Watts, may have provided the true grounds for dismissal. In any event, Gray’s subsequent recommendation © helped Vasey get the vacated position, and with © ‘patient effort? Vasey managed to overcome ‘the lack of appreciation of those in high office who | thought it a waste of time and money to advance | the sciences which wait upon and promote true ag- | riculture’ (Canby and Rose 1893, p. 173), and pro- I | | | |) ) | | 2000] ' ceeded to build up the collection extensively. Vasey was joined by Frederick V. Coville in 1888, who was then recruited to participate in a new federally sponsored initiative to survey the bi- ological resources of the nation. Perhaps Coville’s inclusion was spurred in part by an 1887 editorial in Botanical Gazette (12:197—198), which decried the cessation of federal support for botanical explo- ration following the coalescence of competing sur- veys into the USGS: ‘Millions have been spent in increasing our knowledge of the other riches of our domain, but the plants have been left to private en- terprise ... a few thousand dollars from an over- flowing treasury could be made to yield an ample return in our better knowledge of one of the noblest and (in a public way) most neglected sciences.’ As summarized by Coville (1893, p. 1): In 1886 and subsequent years appropriation was made by Congress for a study of the geographic distribution of animals, to be conducted by the Division of Ornithology and Mammalogy, Unit- ed States Department of Agriculture. In the year 1890 the scope of the work was enlarged by an act of Congress so as to include the distribution of plants as well as animals, and in accordance with this provision the writer was temporarily de- tailed from the Division of Botany as botanist of the Death Valley Expedition, the first of the bi- ological surveys under the new act. The work was planned and conducted under the direction of Dr. C. Hart Merriam, chief of the Division of Ornithology and Mammalogy. The _ botanical work undertaken by the writer was to collect and identify the plants of the region traversed by the expedition, to collate those data which had ref- erence to the range of species, and to arrange this accumulated material in such form that it would be useful in studying the facts and problems of geographic distribution. Subsequent appropriations were made annually ‘for botanical exploration and the collecting of plants in littlhe known districts of America in con- nection with the U.S. National Herbarium’ (Coville 1890), eventually resulting in the first state floras for Washington (Piper 1906), New Mexico (Woo- ton and Standley 1915), and Nevada and Utah (Ti- destrom 1925), among multiple other publications on the botany of North America and other parts of the world. However, initial support for botanical Survey, probably minimal at best, appears to have quickly declined, as evidenced in the correspon- dence of one collector, John B. Leiberg: During the past three summers I have been for- tunate enough to obtain a commission from the Dept. of Agri. for field work in the Columbia basin. As the routes are long, one obtains a pretty good field knowledge of many species over a ERTTER: UNDISCOVERED HERITAGE 241 considerable area. For this reason the position is desirable. From a pecuniary standpoint of view it is not. A commission is only given for a lim- ited period of each year and the expense involved in providing transportation and the details of one’s outfit absorb from 50% to 80% of the total salary that the commission carries. ... Whether these explorations will be continued I do not know. So long as there is any money available for field work there seems no good reason why they should not.... A great deal of our territory is so difficult and expensive to explore that un- less some Gov’t aid is afforded we will never know the complete flora of these regions. Sheep and cattle are rapidly destroying the native plants and by the time private explorations reach these regions the flora will have been totally extermi- nated by such agencies. (Leiberg to C. V. Piper, 5 July 1896 [printed in Sage Notes (Idaho Native Plant Society) 21(4): pp. 6—7]) Piper, to whom this correspondence was addressed, likewise received minimal support for his efforts, with field work ‘carried on in chance hours of lei- sure and in occasional summer vacations’ (Coville in Piper 1906, p. 5). Leiberg’s 1895 instructions represented a shift from general floristic survey to a comprehensive overview of topography, climate, timber resources, and aboriginal uses of native plants (Coville in Lei- berg 1897, p. 1). The timber focus was tied to the controversial Forest Reserve Act of 1891, which gave the President authority to establish forest res- ervations from public domain lands. This act was passed in response to the devastation that was being wrought by unregulated exploitation of western re- sources, as well detailed in Leiberg’s report: The next and last stage in the destruction of the forests, which is still in active operation, came when the great ore deposits in the Couer d’Alenes [in northern Idaho] were discovered. Thousands of prospectors flocked into the coun- try then, and the forest fires raged in hundreds of localities to clear away the dense growth of timber and shrubs, which very materially inter- fered with the work of the prospectors seeking mineral-bearing lodes. As the mines began to de- velop, fuel and timber were needed. The choice parts of the forest were cut into, debris took the place of the green tree, and fire coming later, finished what the axe had spared. (Leiberg 1897, p. 3) The resultant Forest Reserves, initially established within the Department of the Interior in 1897, were transferred to the USDA in 1905, and in 1907 were renamed National Forests in order to counter the impression that they had been completely with- drawn from use. The responsible agency was like- 242 wise renamed, in 1905, from the Bureau of Forestry to the U.S. Forest Service (McClure and Mack 1999). During this period the Division of Ornithology and Mammalogy, within which botanical surveys associated with the U.S. National Herbarium had been initiated, also went through several metamor- phoses, being renamed the Division of Biological Survey in 1896 and, in 1905, the Bureau of Bio- logical Survey. In 1939 it was furthermore trans- ferred from USDA to the Department of the Inte- rior, and the following year consolidated with the Bureau of Fishes to form the U.S. Fish and Wildlife Service. By this time, however, whatever support might have once existed to undertake a comprehen- sive botanical inventory at the national level had essentially disappeared. On the other hand, impetus for scattered species- level inventories, at least for a selection of rare spe- cies, was triggered with the passage of a diversity of environmental legislation at both federal and state levels (e.g., Endangered Species Act, Califor- nia Environmental Quality Act), beginning in the late 1960’s. One result has been a flurry of species- level surveys done as part of environmental impact statements, often limited to the target species but sometimes more comprehensive, with quality rang- ing from superb to dubious. As a broad generality with many exceptions, these have largely resulted in unvouchered species lists scattered throughout the ‘gray’ literature of environmental documenta- tion, or in the file cabinets of governmental agen- cies and consulting firms. Efforts to compile this massive accumulation of potentially invaluable dis- tributional information have begun (e.g., CalFlora [www.calflora.org]), though the complications in doing so have proven to be daunting. Furthermore, the majority of these surveys have been largely de- coupled from the systematics community who formed the core of earlier botanical survey efforts, and who continue to have primary responsibility for the comprehensive synthesis of floristic informa- tion. Various limitations and pitfalls resulting from this decoupling are discussed later in this paper. PAST AND FUTURE ROLE OF THE UNIVERSITY AND JEPSON HERBARIA The California Geological Survey not only served as a precursor to the USGS and many of the contemporaneous biological survey efforts high- lighted in the preceding section, but also set the stage for the long involvement of the University of California at Berkeley (UCB) in botanical surveys of California and the western United States in gen- eral. Whitney, as both director of the CGS and chairman for the commission that drafted plans for the future State University (Brewster 1909), claimed that: [T]he establishment of the Geological Survey was in fact the first step towards the production MADRONO [Vol. 47 | of a State University. Without the information to. be obtained by that Survey, no thorough instruc- | tion was possible on this coast, either in geog- raphy, geology, or natural history; for the student | of these branches requires to be taught in that | which is about him, and with which he is brought | into daily contact, as well as that which is distant | and only theoretically important. (quoted in Stadtman 1970, p. 27) Perhaps because of Whitney’s influence (and/or for fear that the collections would otherwise remain at Harvard University, which Whitney had been ac- cused of acting for the benefit of [Brewster 1909]), | the 1868 Organic Act establishing the University of | California specified that: The collections made by the State Geological | Survey shall belong to the University, and the | Regents shall, in their plans, have in view the | early and secure arrangement of the same for the | use of the students of the University, so soon as | the geological survey shall be completed, and of | giving access to the same to the public at large | and to visitors from abroad; and shall in every — respect, by acts of courtesy and accommodation, | encourage the visits of persons of scientific tastes — and acquirements from other portions of the | United States and of Europe, to California. The | said collections shall be arranged by the resident | Professors of the University in a building by | themselves, which shall be denominated the | ‘Museum of the University.’ 2 (California Assembly Bill No. 583, Sect. 24) | Tradition has it that an initial set of CGS botanical specimens was received by the University in 1872, thereby establishing the University Herbarium (in | fact if not in name), though no records have been | located to confirm this (Ertter 2000b). In any event, there is evidence that in-house col- lecting activities began within the first few years of the University’s existence, as evidenced in a printed report submitted by Joseph LeConte (1875), hired as the first professor of Geology, Natural History and Botany when the University opened its doors: ‘In accordance with my promise I hereby make a brief report of the results of the recent excursion made by the University Scientific Party. The party as you know was organized for the purpose of uti- lizing the Spring recess of a week, in giving some practical instruction in Geology, Lithology, and Surveying; but expected also to make some collec- tions of plants, rocks, fossils, etc., for the Museum.’ The party spent a full week to travel from Berkeley to Black Diamond Mines and Mount Diablo and back: ‘As our time was very limited we stopped but little until our objective points were reached.’ About 150 plants were collected by ‘our young bot- anist’ Franklin P McLean: ‘Whether any of these are new or not remains to be determined.’ McLean, 2000] a graduate of the University’s College of Pharmacy ‘in 1875, also accompanied LeConte’s expedition to Yosemite in the summer of 1875 and collected else- ‘where in California, in the process unfortunately assigning the wrong locality to several collections ‘(Jepson notebooks: Calif. Bot. Expl. I: pp. 99, (118-119, 128, 188). _ This early collecting tradition received new im- petus with the arrival in 1885 of Edward Lee Greene as the first full-time Professor of Botany, assisted by Marshall Avery Howe and attracting such dedicated students as Willis Linn Jepson and Ivar Tidestrom. In 1891, the Chamisso Botanical ‘Club was organized at the University ‘by officers and students interested in botanical work. The pro- moters of the club had especially in view the col- lection of material upon which to found local plant- lists’ (Jepson 1894, p. 171). Different members staked out territories, in which trespassing by rivals was discouraged, with one exception: ‘Professor Greene as the Great Chief was of course free from all restrictions. We had too much to gain from his friendship to object to his hunting on our grounds’ (Frederick Theodore Bioletti, quoted in Ewan 1955, p. 35). Tidestrom would go on to be one of the botanists working for the U.S. National Herbarium, whose efforts resulted in the first flora of Utah and Nevada (Tidestrom 1925). Jepson himself remained at Berkeley, amassing the extensive collection that would ultimately be donated to the University as the core of the Jepson Herbarium (Beidleman 2000). Jepson’s personal activities were greatly supplemented by the extensive network he culti- vated throughout California, ranging from lawyers to farmers to high-school teachers (Ertter 2000c). Beyond the extensive collecting activities of fac- ulty, staff, students, and others connected with UCB, which followed these early beginnings, there has been a strong tradition of large-scale collabo- rative survey efforts with various federal and state land management agencies. The most extensive was the Vegetation Type Mapping Project (VTM), with the U.S. Forest Service acting as lead agency (Wieslander 1935). The concept was purportedly inspired by a course taught by Jepson (Jepson et al. 2000). The original scope of the VTM entailed 220 map units (Wieslander 1935), coinciding with to- pographic quadrangles, but only 23 were published before further activities were disrupted by the out- break of World War II (Wieslander et al. 1932— 1943). In addition to maps, the VWTM resulted in over 23,000 vascular plant collections that are now housed in the University Herbarium: [The VIM collection] includes many plants in addition to those required for authenticating the maps and sample plots. Very complete field notes accompany each specimen, comprising informa- tion as to collector, date, elevation, location, also notes as to size and character of the plant, the slope exposure, the formation in which it grows, ERTTER: UNDISCOVERED HERITAGE 243 and the names of the more common associated species. The primary purpose of the herbarium is to serve as a check upon field identifications, and to afford a permanent record of the plants collected in each quadrangle. Probably its great- est value, however, will le in the wealth of ma- terial from all parts of the region, and in the de- tailed information, as to the range, habitat, and associated plants that will be available for each species. (Wieslander 1935, p. 142) The VTM vouchers have been an invaluable addi- tion to the University Herbarium collections, and have been used as types of at least twelve taxa: Arctostaphylos glutinosa B. Schreib., A. morroensis Wiesl. & Schreib., A. oftayensis Wiesl. & Schreib., A. rudis Jepson & Wiesl., A. pilosula Jepson & Wiesl., A. silvicola Jeps. & Wiesl., Ceanothus otay- ensis McMinn, Galium andrewsii A. Gray var. gd- tense Dempster, Githopsis pulchella Vatke subsp. campestris Morin, Helianthemum suffrutescens B. Schreib., Melica californica Scribn. var. nevadensis Boyle, and Sidalcea hickmanii Greene subsp. anomala C. L. Hitchc. Several of these were de- scribed by researchers unconnected to the VTM, long after the survey had ended. As an apparent offshoot of the VTM, and thereby likewise owing a debt to Jepson, a massive effort to map the vegetation of the Western National Parks was initiated by Harold E. Bailey (H. Bailey and V. Bailey 1941; V. Bailey and H. Bailey 1949). As recalled by his wife and coworker, Virginia Long Bailey: I met Harold E. Bailey, who had just come up from U.C.L.A., planning to work toward a Ph.D. degree under Lee Bonar. The Ph.D. degree was finally completed in 1935 just prior to the start of the Vegetation Type map Survey of the West- ern National Parks. During this three-year project (1935-1937, extending into Olympic Nat. Mon. thru 1938 with us) the winter periods were spent in Berkeley and my work with Dr. Jepson was continued part time.... I think the summer of 1933 must have been the time that Harold had done some veg. type map work with the Sequoia crew [of the VTM], led by Theodore Plain. (V. Bailey to R. Beidleman, 11—20 May 1996) Mapping activities were conducted by crews re- cruited from a variety of sources, with participants often going on to higher level positions within the parks and other agencies. As recalled by John Rut- ter, former Assistant Superintendent of Rocky Mountain National Park: I had to quit school to work awhile in 1934. I went to work as a helper in a type map crew for A.E. Wieslander in the California Forest and Range Experiment Station. I was loaned to Yo- semite N.P. for 90 days to map much of the Park 244 north of the Valley. I didn’t ever go back to the Forest Service. ... I knew Harold as a teaching assistant before he became project leader for the type map. (Rutter to R. Beidleman, 7 November 1996) Even more than the VIM, the National Park mapping project involved a close collaboration be- tween federal land-management agencies, in this case the U.S. National Park Service, and UCB, tak- ing advantage of depression-relief funding: The plant collecting activities in which we were involved were in connection with a vegetation type map survey (of the western national parks) carried out under a government sponsored ‘Emergency Conservation Works’ project under the direction of the western regional office of the Division of Forestry of the National Park Service during a three-year period, 1935-1937. . Headquarters was on the University of California campus in Berkeley and an agreement was reached with the University of California her- barium to identify the plant collections. They were to retain a duplicate set of the collections identified and send a list of the identifications to each park area involved. Duplicates should have been left at park headquarters in each case, but if not, then the herbarium was to send a set along with the list of identifications. (Bailey and Bailey to Wm. M. Lukens, Supt., Chiricahua Nat. Mon., 3 September 1974) As with the VIM, this productive collaboration was disrupted with the onset of World War I and the end of Emergency Conservation Works fund- ing. Not only was mapping work discontinued, but budget cuts within the University Herbarium pre- cluded further processing of the resultant speci- mens. Several thousand unidentified, unlabelled, and unmounted specimens languished as backlog until the 1990’s, when National Science Foundation funding (BSR-8417804) finally allowed the com- pletion of the University Herbarium’s contribution to one of the most exemplary collaborations it has ever been involved in. One further collaborative survey of California plants took place in the intervening years, involving the UCB Department of Botany, the California De- partment of Fish and Game, and the U.S. Fish and Wildlife Service. The focus was California’s wet- land flora, in particular the feeding and resting areas for migratory aquatic birds. Federal funds provided for five year’s of intensive field work by a team of assistants working under the direction of Herbert L. Mason, resulting in both the authoritative reference to California’s wetland flora (Mason 1957) and thousands of invaluable herbarium specimens de- posited in the University Herbarium. This wetland survey, along with the VTM and the Western Na- tional Parks mapping effort, serve as exemplary MADRONO [Vol. 47, models for comparable undertakings at a time when. accurate information on plant distributions has be- | come increasingly critical. | IsN’T IT DONE YET? oR, You GET WHAT You PAY FoR | The preceding historical accounts highlight scat- tered examples of state and federally supported bi- ological surveys that were contemporaneous with the development of the USGS, as well as the in-| volvement of the University of California at Berke-_ ley in such activities. In addition to presenting an’ overview that has not previously been summarized, | this synopsis is intended to emphasize the minimal. support allocated to species-level botanical surveys | during the period that the USGS topographic map- | ping effort was in full swing, resulting in the full) suite of topographic maps that are now taken for granted. This divergence in support undoubtedly | was tied to perceived economic importance, with : what botanical component there was increasingly | shifted to timber and rangeland resources of im-_ mediate and obvious economic significance. As a result, species-level inventories became increasing- | ly dependent on scattered individual efforts outside. of any organized framework (Ertter 1995, 2000a). With the advent of endangered species legisla- | tion, however, it suddenly became important to | have accurate, comprehensive information on past | and present distributions of all plants in the United | States. Not only does such information serve as the raw data from which rarity status is initially deter- | mined, but it also forms the basis on which in- formed decision-making depends. The negative consequences of basing critical land-management decisions on incomplete or inaccurate species-level distribution information can cut both ways, increas- ing the risk of misplaced (and expensive) mitigation efforts as well as the unanticipated extinction of overlooked species (Ertter 2000a). In other words, information that had been treated as primarily of peripheral scientific interest suddenly took on sig- nificant socioeconomic importance, over which lawsuits have been fought. Going beyond rare and endangered species, com- prehensive distributional information for all plants is increasingly needed for burgeoning restoration efforts. The importance of such for post-fire resto- ration is described by Charlet (2000), and infor- mation on historical distributions of plants has also played a role in formulating restoration goals for the San Francisco Bay (Goals Project, 1999). On an even grander scale, how much might we depend on comprehensive and reliable baseline information on current plant distributions against which to eval- uate the predicted impact of global warming? Given the current importance of comprehensive botanical inventory and plant distribution informa- tion, the question quickly arises: did earlier orga- nized survey efforts, as highlighted previously, sup- 2000] ‘plemented by subsequent scattered efforts, leave us ‘with a legacy of the necessary information? As pre- ‘sented in Ertter (2000a), the answer is a resounding ‘No!’ Nearly 60 vascular plant taxa per year are still being described from North American north of ‘Mexico, at a remarkably steady rate (Hartman and Nelson 1998). Recent discoveries, many by envi- ronmental consultants, range from distinctive shrubs along a well-traveled highway (Neviusia cliftonii Shevock, Ertter, & D. W. Taylor [1992]) to anew monotypic genus in the largely agricultural San Joaquin Valley (Twisselmannia californica A\|- Shehbaz [1999]). Not only is an extrapolated five percent of the national flora yet to be described (Taylor in Ertter 2000a), and therefore subject to extinction from ignorance alone, but the level of distributional information on currently known spe- cies is well below that needed for informed deci- sion-making. Charlet’s work in Nevada, for exam- ple, showed that the distribution of conifers, prob- ably the best-mapped of all plants (e.g., Little 1971), was less well-known than had been as- sumed, with 40% of the conifer-bearing mountain ranges in Nevada harboring at least one more spe- cies than had previously been recorded (Charlet 1996, 2000). Even where historical distribution has been ad- equately documented, information on current range is often insufficient to determine rarity status, es- pecially for formerly abundant plants that have lost most of their range to development. The once-com- mon Horkelia cuneata Lindley subsp. puberula (Greene) Keck, for example, was largely eradicated from the Los Angeles basin before anyone even became aware of its plight (Ertter unpublished data). In the opposite direction, the appearance and spread of non-native plants has been historically under-documented, leading to a massive catch-up effort as the economic impact of invasive species has become evident (e.g., the Sierra Nevada Co- operative Yellow Starthistle Mapping and Assess- ment Project [Yacoub and Schoenig 2001]). In es- sence, far from the days of field exploration being well behind us, the need for on-going, organized botanical inventory is both urgent and wide-reach- ing. How did the present situation come about, where the gap between available floristic information and what is needed for informed decision-making reached the magnitude it has?) Some blame can be laid on the systematics community itself, which has been guilty of seriously underestimating the task and overestimating what had already been accom- plished (Ertter 2000a). Whitney’s introduction to the botanical report of the California Geological Survey (Brewer et al. 1876) is a prime example: ‘The total number of species thus included was es- timated at two thousand and it was thought that the work of determining and describing them would not occupy more than a year or two.” As previously noted, the work took considerably more than two ERTTER: UNDISCOVERED HERITAGE 245 years, and the final tally of 3500 species was nearly twice the original estimate. Even this, however, was only half the number of vascular plants currently recorded from California (Hickman 1993), and the actual number is a matter of speculation. A parallel is readily seen with Powell’s underestimation in carrying out his vision of comprehensive topo- graphic mapping, as noted earlier, in the contrast between initial expectations of the task involved with the actual magnitude of effort required. An even greater determining factor, however, has been the support (or lack thereof) provided for on- going species-level botanical inventory by society in general and the scientific community in partic- ular, which controls funding, hiring, and promo- tions based on what is perceived to be a suitably appropriate scientific undertaking. Watts’ negative evaluation of Parry’s contribution to science, quot- ed previously, shows how deep-rooted the resis- tance to botanical inventory is. Paradoxically, my impression is that society-at-large, far from believ- ing that the generation of species-level distribution- al information is undeserving of institutional sup- port, instead takes for granted that such support has existed all along, fully parallel to the topographic mapping effort of the USGS. The resultant assump- tion is that comprehensive species-level distribution maps should already be available as needed, for all of the above-cited purposes. Instead, as the preced- ing historical account demonstrates, as a society we've simply gotten what we’ve paid for. THE CONTINUING ROLE OF VOUCHER SPECIMENS To the extent that the desirability of comprehen- sive, reliable, species-level plant distribution infor- mation is acknowledged, two somewhat contradic- tory stances have been adopted: either that all es- sential information already exists and simply needs to be compiled (the informatics approach); or that such a goal is completely unrealistic, and that var- ious short-cuts must therefore be pursued (the in- dicator species, gap analysis, and/or vegetation mapping approaches). These alternate approaches are unquestionably valuable, both for their own sakes and as components of a larger undertaking, but none can sufficiently take the place of a com- prehensive species-level inventory involving both voucher specimens and the systematic community. The limitations of vegetation mapping divorced from species-level information are addressed ad- mirably elsewhere in this symposium volume (Charlet 2000). Some limitations of the compilation approach have been elegantly analyzed by D. W. Taylor, mostly as work-in-progress. A key limitation of the compilation approach is its dependence on the adequacy of existing data sources. Figures | and 2, generated by Taylor, il- lustrate the inadequacy of existing documentation of species-level plant distributions in California, based on herbarium specimens in the University Fic. 1. County map of California showing density dis- tribution of UC/JEPS specimens (sheets/km7), based on a total of over 280,000 sheets. Herbarium (UC) and Jepson Herbarium (JEPS) at the University of California at Berkeley. Although these collections are obviously only a subset of the total number of herbarium specimens in existence available, they are nevertheless representative enough to serve as the basis for initial rough anal- yses, as presented here. Figure 1, showing collec- tion density per unit area (averaged throughout a county), illustrates the non-uniformity of documen- tation coverage among the different counties of California. Some of the non-uniformity can be readily explained (e.g., the highest densities in counties surrounding Berkeley; J. P. Tracy’s intense collecting efforts in Humboldt County; exchange from herbaria in various southern counties), but the overall pattern of irregular coverage is irrefutable. Furthermore, evidence from other sources under- scores how much remains to be documented even in high-density counties. Recent work on the Mount Diablo flora of Contra Costa County (Bowerman and Ertter in press), for example, increased the pre- viously documented flora (Bowerman 1944) by 26%, over half native. Several fully established non-natives were even additions to The Jepson Manual (Hickman 1993): e.g., Dittrichia graveo- lens (L.) Greuter and Trifolium tomentosum Willk. ex Nyman. In that the recently collected vouchers for the Mount Diablo study have not yet been ac- cessioned, they represent material beyond that in- cluded in Taylor’s analysis in which Contra Costa County already has one of the highest collection densities. Figure 2 carries the California-wide analysis a MADRONO [Vol. 47. 400 | 90 80 70 60 50 40 30 Percent of Asteraceae Taxa Vouchered 0 00 O1 O2 O08 o4 oof Asteraceae Collection Density (specimens/km2), Fic. 2. Relationship between the collection density of. Asteraceae (sheets/km?’) and the proportion of Asteraceae county records vouchered. Symbols: © = counties treated by a local flora; A = other counties. The line shown was. selected from amongst a variety of model forms tested based on overall goodness-of-fit (R? = 0.37, P < 0.001). San Francisco County was excluded from the plot (cf. Fig. 1). (D. W. Taylor, unpublished data). step further, attempting to correlate collections den-, sity of each county (x axis) with completeness of. species documentation (y axis), as calculated by: comparing predicted occurrence of species of As-i teraceae (extrapolated from multiple sources)' against the holdings of UC/JEPS. To the extent that’ this admittedly preliminary analysis is informative, | it may be that only 80% of the vascular plants have: been documented from even the most heavily col-: lected counties. | Of course, one question that begs to be addressed: is, why limit distribution reports to those docu-/ mented by herbarium vouchers? There is indeed le-: gitimacy in supplementing documented distribu-* tions with unvouchered reports (such as the huge’ number of species lists resulting from various en-' vironmental surveys referred to previously), to the: extent that an acceptable level of reliability can be! determined. Unfortunately, the limitations of deter-, mining reliability without a voucher quickly be- come apparent, underscored by the frequency with. which determinations of vouchered occurrences are ‘ changed over time for a variety of reasons. Al-| though some changes result from outright initial: misidentification, the majority reflect altered taxo- nomic circumscriptions as our understanding of species boundaries and relationships improves. Ex- | amples of both kinds of changes are represented in. | | 2000] ‘the updated Mount Diablo flora (Bowerman and Ertter in press), verifiable because of the profuse vouchers cited in the original flora (Bowerman 1944). Both voucher specimens for Prunus emar- ginata (Hook.) Walp. in the 1944 edition, for ex- ample, have been reidentified as escaped cultivated species, so P. emarginata has been eliminated from the more recent edition. Alternatively, vouchered references in the 1944 edition to Oenothera hirtella var. jonesii, which had subsequently been split among several taxa (Raven 1969), could be updated to the correct taxa as now circumscribed, something that could not be done with unvouchered citations. Echoing Wieslander’s previously quoted com- ments on the VIM collections, as well as argu- ments by Goldblatt et al. (1992) and Ferren et al. (1995), the importance of voucher specimens was clearly emphasized in the report, A Biological Sur- vey for the Nation, prepared by the National Re- search Council of the National Science Foundation (1993, p. 68): Collections of specimens are a critical compo- nent of the [National Partnership for Biological Survey]. In all but a few well-known taxa, iden- tifications of species must be based on voucher specimens, without which frequent misidentifi- cations are certain to be made. Faulty manage- ment decisions are likely to result from incorrect identifications. Collections are repositories for most of what we know about species diversity and are constantly pressed into use for new and often unexpected purposes. The critical role played by vouchered documenta- tion of species-level distributions, and the limita- tions of the purely compilation approach to distri- butional information, is further emphasized when the extent of rejected reports is realized. Although the value of indicating excluded species (i.e., taxa that at one time or another had been included with- in the group but which are now treated as members of other groups) is well-established in monographic works, the need for comparable lists of excluded or rejected species in floristic works has not generally been appreciated. This has not been a significant problem in monographic floristics, which largely rely on voucher specimens, other than increasing the likelihood of redundant effort anytime the source of the excluded report resurfaces. Keeping track of erroneous or dubious reports becomes crit- ical, however, now that mass compilation of spe- cies-level distribution reports from multiple sources has become popular. The magnitude of the potential error can be seen in some floristic examples that have attempted to indicate rejected reports; e.g., 97 in the East Bay flora (Ertter 1997), 66 in the Mount Diablo flora (Bowerman and Ertter in press), equiv- alent to 6% and 8% respectively of accepted taxa in each flora. Excluded reports include misappli- cations, confirmed misidentifications, and vouchers ERTTER: UNDISCOVERED HERITAGE 247 with suspect localities, but mostly represent un- vouchered reports of dubious nature, often far out- side known distributions. Although it has been in- sufficiently acknowledged, critical evaluation and decision to exclude reported occurrences has in fact been among the primary responsibilities and con- tributions of the systematics community to species- level distributional informatics. ‘ORGANIZED’ FLORISTICS AND THE SYSTEMATICS COMMUNITY The significance of critical evaluation by the sys- tematics community stands as a key distinction be- tween the compilation approach to biodiversity in- formatics, whether electronic or printed, and that employed in established floristics, in which the con- tents are carefully evaluated, filtered, and synthe- sized. This distinction underlies Jepson’s character- ization of compiled, accreted, and organized flo- ristic works, expressed in a recently unearthed letter to Wieslander (3 April 1939, JEPS archives): There are three kinds of manuals. First, a manual that is compiled. Second, a manual that is ac- creted. Third, a manual that is organized. A com- piled manual, for example, is such as Coulter’s [1885] Manual of the Rocky Mountain Flora (not Nelson’s [Coulter and Nelson 1909], but Coul- ter’s). Taken wholly from the literature, nothing is left out, nothing omitted. It is philosophically speaking, perfect and complete. But no real bot- anist, I think, ever looked within its pages. It is to him useless. ... Then there is the manual that is accreted. In this case everything is put in, not only from books but also from plants. It, too, leaves nothing out. It adds everything that comes along, both from plants and the literature. It 1s, also, philosophically speaking, perfect and com- plete. And, finally, there is the Manual that is organized. My Manual of Botany [Jepson 1923— 1925] is organized. It is not perfect nor complete, nor can ever be in a thousand years. The whole treatise is, however, organized into a single unit, every part depending and related and associated with every other part. And it is made up basically from research on plants.’ Jepson’s concept of an ‘organized’ floristic work, with ‘every part depending and related and asso- ciated with every other part,’ is equivalent to the argument in Ertter (2000a) that floras and other kinds of taxonomic treatments are best understood as complex models, encompassing multiple units whose exact identities depend on their relation to other units within the larger context. A prime ex- ample is provided by Fig. 3, contrasting four alter- nate taxonomic models that had been developed to circumscribe taxa within the Juncus triformis com- plex. Although this type of situation has sometimes been disparaged as evidence of the systematic com- munity’s purported inability to agree on standards, [Vol. 47) 248 MADRONO COMPARATIVE MODELS OF JUNCUS TRIFORMIS COMPLEX pre-Hermann Hermann, 1948 Cronquist, 1977 Ertter, 1986 J. triformis J. triformis J. megaspermus J. triformis var. (not addressed) stylosus J. leiospermus J. leiospermus var. leiospermus var. ahartii J. kelloggii J. triformis var. | J. kelloggii J. luciensis brachystylus J. tiehmii J. capillaris J. capillaris J. bryoides J. Kelloggii J. bryoides J. triformis J. uncialis | J. uncialis var. uniflorus | J. her hemidentyus J. hemindytus var. hemiendytus J. abjectus var. abjectus Fic. 3. Comparison of four taxonomic models (monographic treatments) of the Juncus triformis complex. Note in| particular the dramatically different circumscriptions of J. kelloggii between models. it is actually a straight-forward case of science in action, with earlier hypotheses and models giving way to new ones in the face of additional evidence. In this particular example, Cronquist (1977) hy- pothesized that the series of species proposed by Hermann (1948) did not meet the accepted criteria for recognition as distinct species, but rather ‘ap- pear to be mere technical variants, often locally constant as in self-pollinated groups in other gen- era, but with widely overlapping ranges and similar habitat requirements.’ In that Cronquist himself was aware of the limited evidence on which his model was based, he encouraged one of his students to put it to the test and was fully accepting of the alternate model that resulted (Ertter 1986), which was based on five years of focused field work, common gar- den studies, chromosome counts, and seed coat mi- cromorphology. The importance of this particular example in the present context is to illustrate the pitfalls associated with attempts to deal with taxo- nomic units as free-standing entities divorced from a specific model, as is generally the case in mass compilations. For better or worse, the nomenclatur- al system adopted by the international systematics community ties the name to a type specimen, not to a circumscription. As a result, the binomial Jun- cus kelloggii Engelm., rather than serving as a unique identifier, can code for three very different entities, depending on whether it is in the context of Hermann’s, Cronquist’s, or Ertter’s model. Com- pilation efforts that are unable to take this into con- sideration will inevitably end up generating the most inclusive circumscription (e.g., that of Cron- quist) even if this is not the currently accepted cir- cumscription. This can be seen, for example, in the distribution map generated for J. kelloggii in the PLANTS database __(http://plants.usda.gov:80/. plants/), which shows a range significantly larger. than the documented range published in 1986. Another example illustrating the nature of an ‘or-. ganized’ taxonomic work is provided by the recent. description of Deinandra bacigalupii B. G. Bald-. win (1999b), based on what had previously been treated as a disjunct northern population of Hemi-. zonia increscens (D. D. Keck) Tanowitz subsp. in- crescens (e.g., Tanowitz 1982). Not only did pub- lication of this new species provide impetus for, Baldwin to publish his emerging generic realign- ment of tarweeds that had resulted from morpho- | logical and molecular phylogenetic analysis (Bald- | win 1999a), but publication of D. bacigalupii also. created a new circumscription of Hemizonia/Dein- | andra increscens. As a result, D. bacigalupii cannot | simply be added to existing floristic treatments (e.g., Hickman 1993) without simultaneously mod- | ifying the description and distribution of D. incres- | cens to reflect its reduced circumscription. | The purpose of the preceding paragraphs is to clarify that critical analysis by members of the sys- | tematics community, rather than being peripheral, | is an essential component of on-going botanical in- | ventory. This is by no means intended to downplay | the equally critical involvement of agency biolo- — gists, environmental consultants, and avocational © enthusiasts, who are in fact currently responsible for generating the bulk of new field-gathered infor- | mation (Ertter 1995, 2000a). The point is that our modeling of biodiversity is still very much a work- | in-progress, such that even the seemingly mundane | aspects of plant distribution information are often | clues to the undescribed 5% of the North American flora, or to the ‘cryptic’ diversity that is also a crit-_ 2000] ical component of biodiversity (Baldwin 2000). As one example, the revision of the Mount Diablo flora (Bowerman and Ertter in press), as localized as it was, nevertheless involved numerous interactions with taxonomic specialists to address discrepancies between local variation (1.e., plants that “hadn’t read the book’) and treatments in The Jepson Manual (Hickman 1993), often resulting in changes to the latter. This is in part what Jepson (cited above) meant by a flora ‘organized into a single unit, every part depending and related and associated with ev- ery other part,’ and what he expanded on in the same letter: One of my students opened a bundle of plants [in my collection] and exclaimed: ‘Why, Dr. Jep- son, here are species new to California from the eastern Mohave borders collected by yourself. Why did you not put them in the Manual?’ I had to explain that these were critical species which would have taken a long time to determine; and, even after determination, would require a long time for organization into the manuscript. It was not possible to delay the Manual further. In his inexperience the student imagined species could be added just like adding another stick to a pile of cordwood. He had no conception of the hun- dreds of comparisons involving detailed analysis that must be made in the case of every species added to a systematic account. Even botanists in general have no notion of the mass of work in- volved in a large systematic treatise. LOOKING AHEAD: THE HARVEST TO COME Given the preceding discussion on the historical and current status of species-level botanical inven- tory in the United States, it is evident that most of the perceived obstacles to on-going efforts are based on false assumptions. Instead: _@ The results of such efforts have significant prag- matic implication and potential economic impact, primarily as a critical component of informed land-management decision-making. As a result, properly done survey efforts prove their worth in the long run and have even received significant support from far-sighted private donors on that account (e.g., Stanford’s support of the California Geological Survey). @ Federal- and state-funded survey efforts were terminated by politics, special interests, and mis- conceptions, not because the scientific goals were completed or unimportant. @ The essential fieldwork and critical taxonomic evaluations therefore remain far from finished, and can by no means be offset by simple com- pilation of existing data, even within a modern informatics framework. ERTTER: UNDISCOVERED HERITAGE 249 @ Herbarium vouchers remain an integral part of scientific documentation, with many more need- ed to document species-level distributions com- prehensively and reliably. @® On-going involvement of the systematics com- munity is likewise integral, not only to address the numerous undescribed species (an estimated 5% of the North American vascular flora) but to ensure that the resultant informatics framework is fully ‘organized.’ This leaves the following two assumptions: @ Species-level inventory within the United States is not sufficiently scientific, innovative, or oth- erwise high-profile to merit funding. @ Comprehensive species-level inventory is simply too big a project to tackle (hence the short-cuts of vegetation mapping, gap analysis, indicator species, umbrella species, etc.) The first assumption appears to be deeply rooted, at least within the American academic community, such that floristic work has long since fallen out of favor as a suitable topic for graduate work, in spite of Jepson’s lifelong efforts to develop floristics as sound science. In Europe, on the other hand, an entire field of chorology has developed around a Committee for Mapping the Flora of Europe, given a recent boost by advances in electronic approaches (e.g., Lahti and Lampinen 1999). This touches on the irony of the exploding prestige and popularity of geographic information systems, often taking place at the same institutions that scorn floristic work by systematists. Most efforts (and funds) to develop essential plant distribution information lay- ers, however, are completely decoupled from the systematics community, relying instead on compi- lation approaches, with the resultant pitfalls and shortcomings that have been discussed. Of course, biodiversity informatics as a whole is a favored topic, including within the systematics community itself, spawning a veritable alphabet soup of acronyms at state, federal, and international levels (e.g., as highlighted in ASC Newsletter 28[5], October 2000). At present, however, support for these efforts has been largely directed thus far to massive compilations, perhaps in fact the real- istic and appropriate starting points in an absolutely essential and long-overdue undertaking. Existing projects nevertheless appear to be a far cry from fully involving and providing the concomitant sup- port for the systematics community at large, con- sisting of the multitude of field collectors and mo- nographers who generate the raw data, critically evaluate the results, and synthesize the taxonomic models on which bioinformatics depends. Complementing such umbrella approaches to bioinformatics, there are a diversity of innovative 250 approaches that could be capitalized on to increase the availability and reliability of new species-level plant distribution data. Charlet (2000), for example, argues for the coupling of documented species-lev- el information with vegetation mapping. One also wonders how far various funds currently being al- located for studies on individually targeted rare or invasive species could go towards comprehensive mapping of all plant species in an area, minimizing the need for redundant surveys over the same ground when yet one more species becomes of in- terest. A parallel exists with Jepson’s advice to Wieslander to expand his proposed mapping effort beyond economically important woody species, on the grounds that “New economic aspects developed so rapidly that it was proven repeatedly that an eco- nomic map was and must be from its nature tran- sient and insufficient’ (Jepson et al. 2000). If this advice had been followed from the beginning, a ‘considerable appropriation’ could have been saved that was subsequently needed to re-map much of the area already covered. Tapping into the private sector, Ferren et al. (1995) note that the bulk of undocumented (and under-reported) field observations in the United States currently result from legally required envi- ronmental assessments prior to development. How- ever: Without vouchers deposited in institutional her- baria, the scientific and even legal credibility of these reports is suspect at best, and their long- term value is minimal in spite of the large sums of money spent in producing the documents. In southern California, it is not uncommon for ap- proximately $1 million to be spent for a specific plan and associated [environmental impact re- view] for larger development projects. ... For a little extra money, a much more worthwhile re- view effort could be undertaken. A client’s mon- ey would be more wisely spent if vouchers were collected and deposited in a formal herbarium than if the environmental review was not docu- mented professionally ... since the overall bud- gets for environmental review studies and doc- uments are substantial, it would take only a mod- est addition to the budget to cover the costs of collecting and depositing voucher specimens. (pp. 198, 202) Beyond and above these and other innovative ways to increase support for on-going botanical in- ventory, the most fundamental requirement is a change in our understanding of the situation. Rather than being intimidated by the scope of the chal- lenge, I propose that we have not been thinking big enough! We do not have to justify the initiation of a Big Science project; rather, we need to acknowl- edge that this is exactly what the systematics com- munity has been doing for the last 250 years: a massive international collaboration to model spe- MADRONO cies-level biodiversity, including distribution, that | will remain a work-in-progress for decades, perhaps centuries, to come. We are in this for the long run; | the challenge now is to assemble the scattered piec- | es together in a new collaborative framework, com- | bining the best of the systematics and informatics communities, governmental agencies, conservation organizations, avocational enthusiasts, and private landowners, all within a coordinated, mutually prof- | itable, scientifically valid framework. If this seems daunting, recall again the seemingly | impossible challenge faced by Powell in getting the | USGS off the ground, and its subsequent vindica- tion beyond his wildest dreams. In his 1886 defense | of the USGS (quoted in Stegner 1953, p. 289), | Powell provided this stirring testimony: If the work thus begun can be continued through | the labors of this Commission, and all of the sci- entific operations of the Government placed un- der efficient and proper control, scientific re- search will be established in America upon such a basis that the best and greatest results will ac- | crue there from. The harvest that comes from , well-directed and thorough scientific research has no fleeting value, but abides through the years, as the greatest agency for the welfare of man- kind. What would we have now if a true biological sur- vey had existed parallel to the USGS for the last hundred years? What might the next hundred years’ harvest be? ACKNOWLEDGMENTS Significant acknowledgements are due to Richard Bei- dleman, for freely sharing the invaluable documentation on the VTM that he had accumulated, including personal letters, and for many stimulating conversations on botan- ical history in general; to Dean W. Taylor, for freely shar- ing his ingenious floristic analyses; to Richard Moe, for generating various statistics and lists from UC/JEPS spec- imen databases; to Peter Raven, for introducing me to Stegner’s opus and other critical references; to David Charlet, for thought-provoking discussion on the impli- cations of vegetation mapping; to Bruce Baldwin, for the Deinandra bacigalupii story; to Cindi Wolff of the US Dept. of Interior Library, for tracking the fate of the Di- vision of Biological Survey; and to an anonymous review- er for providing valuable criticisms. Credit is also due the Jepson Herbarium archives for use of quotes from Jep- son’s notebooks and other archival material. LITERATURE CITED AL-SHEHBAZ, I. 1999. Twisselmannia (Brassicaceae), a re- markable new genus from California. Novon 9:132— 135; BaliLey, H. E. AND V. L. BaiLey. 1941. Forests and trees of the western National Parks. National Park Service, Conservation Bulletin No. 6. U.S. Government Print- ing Office, Washington, DC. BaILey, V. L. AND H. E. BaiLey. 1949. Woody plants of the western National Parks. American Midland Nat- [Vol. 47 | | 2000] uralist Monography No. 4., Pt. 1. University Press, Notre Dame, IN. BALDWIN, B. G. 1999a. New combinations and new genera in the North American Tarweeds (Compositae—Ma- diinae). Novon 9:462—471. . 1999b. Deinandra bacigalupi (Compositae—Ma- diinae), a new tarweed from eastern Alameda County, California. Madrono 46:55—57. . 2000. Roles for modern plant systematics in dis- covery and conservation of fine-scale biodiversity. Madrono 47:219—229. BEIDLEMAN, R. 2000. Willis Lynn Jepson—‘‘The Botany Man.”’ Madrono 47:273—286. BOWERMAN, M. L. 1944. The flowering plants and ferns of Mount Diablo, California. Gillick Press, Berkeley, CA. , AND B. ErtTer. In Press. The flowering plants and ferns of Mount Diablo, California, 2nd ed. Cali- fornia Native Plant Society, Sacramento, CA. BRANDEGEE, T. S. 1876. The flora of southwestern Colo- rado. Bulletin of the Geological and Geographical Survey of the Territories 2:227—248. BREWER, W. H., S. WATSON, AND A. GRAY. 1876. Botany [of California] Vol. 1. Welch, Bigelow, & Co., Uni- versity Press, Cambridge, MA. BREwSTER, E. T. 1909. Life and letters of Josiah Dwight Whitney. Houghton Mifflin Company, Boston and New York. CANBY, W. M. AND J. N. Rose. George Vasey: a biograph- ical sketch. Botanical Gazette 18:170—183. CHARLET, D. A. 1996. Atlas of Nevada conifers: a phy- togeographic reference. University of Nevada Press, Reno. . 2000. Coupling species-level inventories with vegetation mapping. Madrono 47:259—264. COULTER, J. M. 1885. Manual of the botany of the Rocky Mountain region. Ivison, Blakeman, Taylor, & Co., New York, Chicago. , and A. Nelson. 1909. New manual of botany of the Central Rocky Mountains. American Book Co., New York, Cincinnati, Chicago. CovILLe, EF V. 1890. Prefatory Note. Contributions from the U.S. National Herbarium 1:11. . 1893. Botany of the Death Valley Expedition. Contributions from the U.S. National Herbarium 4: 1— 362. CrONQUIST, A. 1977. Juncaceae. Pp. 47—67 in A. Cron- quist, A. H. Holmgren, N. H. Holmgren, J. L. Reveal and P. K. Holmgren (eds.), Intermountain flora: vas- cular plants of the Intermountain West, U.S.A., Vol. 6. Columbia University Press, NY. ERTTER, B. 1986. The Juncus triformis complex. Memoirs of the New York Botanical Garden 39:1—90. . 1995. The changing face of California botany. Madrono 42:114—122. . 1997. Annotated checklist of the East Bay Flora: native and naturalized vascular plants of Alameda and Contra Costa counties, California. Special Publication No. 3 of the California Native Plant Society, East Bay Chapter. 114 pp. . 2000a. Floristic surprises in North American north of Mexico. Annals of the Missouri Botanical Garden 87:81—109. . 2000b. People, plants and politics: the develop- ment of institution-based botany in California, 1853— 1906. In M. T. Ghiselin and A. E. Leviton (eds.), Cultures and institutions of natural history. California Academy of Sciences Memoir 25:203—248. ERTTER: UNDISCOVERED HERITAGE 254 . 2000c. A mulewrangler, a lawyer, and a lone woman: the continuing legacy of Jepson’s collecting network. The Jepson Globe 11(2):3. Ewan, J. 1955. San Francisco as a mecca for nineteenth century naturalists. Pp. 1-63 in A Century of Prog- ress in the Natural Sciences, 1853-1953. California Academy of Sciences, San Francisco. FARQUHAR, E P. (ed.). 1930. Up and down California in 1860-1864: the journal of William H. Brewer. Yale University Press, New Haven, CT: FERREN, W. R., JR., D. L. MAGNEY, AND T. A. SHOLARS. 1995. The future of California floristics and system- atics: collecting guidelines and documentation tech- niques. Madrono 42:197—210. GOALS PROJECT. 1999. Baylands Ecosystem Habitat Goals. A report of habitat recommendations prepared by the San Francisco Bay Area Wetlands Ecosystem Goals Project. U.S. Environmental Protection Agency, San Francisco/S.E Bay Regional Water Quality Control Board, Oakland, California. GOLDBLATT, P., P. C. HocH, AND L. M. McCook. 1992. Documenting scientific data: the need for voucher specimens. Annals of the Missouri Botanical Garden 79:969-970. GOETZMANN, W. H. 1966. Exploration & empire: the ex- plorer and the scientist in the winning of the Ameri- can West. Random House, Inc., New York (Vintage Books edition). GRAY, A. 1871. Correspondence relating to the dismissal of the late Botanists to the Department of Agriculture, at Washington. American Naturalist 6:3-8. HARTMAN, R. L. AND B. E. NELSON. 1998. Taxonomic nov- elties from North America north of Mexico: a 20-year vascular plant diversity baseline. Monographs in Sys- tematic Botany from the Missouri Botanical Garden 67:1-59. HERMANN, E J. 1948. The Juncus triformis group in North America. Leaflets of Western Botany 5:109—120. HICKMAN, J. C. (ed.). 1993. The Jepson Manual: higher plants of California. University of California Press, Berkeley, CA. JEPSON, W. L. 1894. Chamisso Botanical Club. Erythea 2: 171-172. . 1898. Dr. Henry N. Bolander, botanical explorer. Erythea 6:100—107. . 1923-1925. A manual of the flowering plants of California. Associated Students Store, University of California, Berkeley. , R. BEIDLEMAN, AND B. ERTTER. 2000. Willis Linn Jepson’s ‘Mapping in Forest Botany.’ Madrono 47: 269-272. KELLY, H. A. 1914. Some American medical botanists commemorated in our botanical nomenclature. South- worth Co., Troy, NY. LAnTI, T. AND R. LAMPINEN. 1999. From dots to bitmaps: Atlas Florae Europaeae goes digital. Acta Botanica Fennica 162:5—9. LECONTE, J. 1875. Report of the results of excursion of the scientific party of the university. [printed report submitted to R. E. C. Stearns; copy in University Scrapbook, 1864-1887, Bancroft Library (Call #: 308h.C153)] LEIBERG, J. B. 1897. General report on a botanical survey of the Couer d’ Alene Mountains in Idaho during the summer of 1895. Contributions from the U.S. Na- tional Herbarium 5:1—85. LEVITON, A. E. AND M. L. ALDRICH, (eds.). 1997. Theodore Henry Hittell’s The California Academy of Sciences: a narrative history: 1853-1906. California Academy of Sciences, San Francisco. LittLe, E. L. JR. 1971. Atlas of United States trees, Vol. 1. Conifers and important hardwoods. U.S. Dept. of Agriculture, Forest Service. Miscellaneous Publica- tions 1146. Government Printing Office, Washington, Dc. Mason, H. L. 1957. A flora of the marshes of California. University of California Press, Berkeley, CA. McC ure, R. AND C. Mack. 1999. Gifford Pinchot Na- tional Forest: early history. Heritage Program, Gifford Pinchot National Forest. http://www.fs.fed.us/gpnf/ heritage/early | .html Morton, C. V. AND W. L. STERN, 1966. The United States National Herbarium. Plant Science Bulletin 12(1):1— 4. (also at http://nmnhwww.si.edu/botany/colls/ us-hist. html) NATIONAL RESEARCH COUNCIL, NATIONAL ACADEMY OF SCIENCES. 1993. A biological survey for the Nation. National Academy Press, Washington, D.C. Piper, C. V. 1906. Flora of the State of Washington. Con- tributions from the U.S. National Herbarium 11:1-— 637. PORTER, T. C. AND J. M. COULTER. 1874. Synopsis of the flora of Colorado. U.S. Geological and Geographical Survey of the Territories, Miscellaneous Publica- tions—No. 4. Government Printing Office, Washing- ton; D:C. RAVEN, P. H. 1969. A revision of the genus Camissonia (Onagraceae). Contributions from the U.S. National Herbarium 37:161—396. ROTHROCK, J. T. 1878. Reports upon the botanical collec- tions made in portions of Nevada, Utah, California, Colorado, New Mexico, and Arizona, during the MADRONO [Vol. 47 | years 1871, 1872, 1873, 1874, and 1875. Volume 6— © Botany of Wheeler, G. M., Report upon United States Geographical Surveys west of the one hundredth me- | ridian. Government Printing Office, Washington, D.C. | SHEVOCK, J. R., B. ERTTER, AND D. W. TAYLor. 1992. | Neviusia cliftonii (Rosaceae: Kerrieae), an intriguing — new relict species from California. Novon 2:285—289. | STADTMAN, V. A. 1970. The University of California, | 1868-1968. McGraw Hill Book Company, New York. STEGNER, W. 1953. Beyond the hundredth meridian: John | Wesley Powell and the second opening of the West. Penguin Books USA Inc., New York. TANOWITZ, B. D. 1982. Taxonomy of Hemizonia sect. Ma- | diomeris (Asteraceae: Madiinae). Systematic Botany © 7:314-339. TIDEsTROM, I. 1925. Flora of Utah and Nevada. Contri- | butions from the U.S. National Herbarium 25:1—665. ~ WAGNER, FE H. 1999. Whatever happened to the National | Biological Survey? BioScience 49:219—222. | WATSON, S. 1880. Botany [of California]. Vol. 2. Little, | Brown, and Company, Boston. | WIESLANDER, A. E. 1935. A vegetation type map of Cal- ifornia. Madrofio 3:140—144. L (in charge) AND Forest Service Staff. 1932-1943. | Vegetation type maps of California and Western Ne- | vada. U.S. Department of Agriculture, Forest Service. | California Forest and Range Experiment Station (in — cooperation with the University of California). Wooton, E. O. AND P. C. STANDLEY. 1915. Flora of New | Mexico. Contributions from the U.S. National Her- | barium 19:1—794. YACOUB, R. AND S. SCHOENIG. 2001. Cooperative Yellow | Starthistle Mapping Project. Noxious Times 3(3):9. MaprRONO, Vol. 47, No. 4, pp. 253-258, 2000 FLORISTIC STUDIES IN CONTEMPORARY BOTANY THEODORE M. BARKLEY Botanical Research Institute of Texas, 509 Pecan Street, Fort Worth, TX 76102-4060 ABSTRACT This paper outlines the traditional procedures for monographic and floristic studies, and points out that floristic studies are a link between the producers and the consumers of botanical information. This paper is derived from a talk titled ““System- atics, Informatics, and Floristics”’ that was present- ed at the Jepson 50th Anniversary Celebration and Scientific Symposium in June, 2000, at the Univer- sity of California, Berkeley. The purpose of this paper is to review how botanical information flows from creators to consumers, and the central role of floristics in that process. Mankind’s age-long study of plants has produced an amazing legacy, which is evident in any schol- arly library (Barkley 1993). Systematists who add to the accumulating knowledge believe that it all has meaning, and presumably it is useful to people outside of the bounds of botany. The diagrams pre- sented here show how the information of system- atic botany is accumulated and how it reaches the consumer in useful form. Monographs and revisions (1.e., ““monography’”’ ) are the “‘soul of systematics,’ as pointed out so neatly by Stuessy (1975). Those studies are taken to be the fundamental syntheses of systematic knowledge and are the roadmaps for subsequent studies. The term “‘monograph”’ has been used for the larger and more sumptuous studies, while “‘re- vision”’ has meant studies presented in less detail. The central goals of both are similar, however, and, as with many species, there is no sharp distinction between the two. In this paper, all such studies are called *“‘monographs.”’ Traditional monographs fo- cus on some natural group, such as a genus or a section of a genus, and they include, among other things, delimitations and descriptions of the enti- ties, keys, hard data on ranges and habitats, and an ordered nomenclature. Modern monographs are also expected to include information on the biology of the group and on natural relationships. A how- to outline for a monographic study was presented by A. S. Hitchcock, the noted agrostologist, in a _ remarkable book called Descriptive Systematic Bot- any (Hitchcock 1925). The book is the product of an earlier era, and even though the author had few of the techniques that are available to us now, it is still well worth revisiting, for the author laid out a clear recipe for the standard procedures of mon- ography. Arguably, contemporary monography rests upon the procedures outlined by Hitchcock, to which have been added many new sources of in- formation and schemes for interpretation. Two matters were particularly important in the first half of the 20th century for the development of monography. The first was the intentional incor- poration of evolutionary questions. Beginning about 1930, a monograph was regarded as incom- plete if it did not offer some understanding of evo- lutionary relationships among the entities under consideration. One of the first and clearest mono- graphs to be published in the USA that had evo- lutionary relationships as a chief goal was the treat- ment of the genus Haplopappus by H. M. Hall (1928). Therein the author tried mightily to develop phylogenies as he understood them. It is not im- portant that Hall’s techniques are now seen as in- adequate and that many of the monograph’s con- clusions are no longer tenable; what is important 1s that he acted upon the assumption that good mon- ography must be centered upon evolutionary rela- tionships. The second matter was the advent of new labo- ratory techniques, which boosted systematic studies into an experimental science with garden and greenhouse studies, cytogenetics, comparative cy- tology, the analysis of secondary metabolites, etc. These studies came to be called “‘biosystematics”’ and have produced imaginative and detailed mon- ographic treatments. Biosystematics coupled com- fortably with ecological studies such as pollination, seed dispersal, population biology, and geohistory, thereby further enriching the content of monogra- phy. A new vista in monography was introduced with the arrival of cladistic theory and new information from molecular studies. Cladistic theory supplied a workable tool for showing evolutionary relation- ships, resulting in phylogenetic trees that could be objectively tested. Molecular studies have proved to be particularly compatible to cladistic analyses, and the two have created a vital subset of system- atic studies that focuses on evolutionary relation- ships, rather than on species delimitation. There is a rich literature on cladistic theory and derived phy- logenies. The application of the phylogenetic ap- proach based upon cladistics is comfortably treated 254 in the recent textbook Plant Systematics: A Phylo- genetic Approach by Judd et al. (1999). Contemporary phylogenetic studies have recog- nized that the traditional Linnaean concept of spe- cies iS imprecise at best and may be no longer jus- tifiable (1.e., species are indeed specious). From the early 1990’s to the present there has been a shower of literature on the creation of a new taxonomic scheme to reflect phylogenetic relationships, and indeed there was a symposium on the topic at the XVI International Botanical Congress in St. Louis, MO in 1999 (cf. Cantino 2000, and Cantino et al. 1999, for an introduction into the literature). It is doubtless true that changes are coming in how we conceive of “‘species,’’ but the proposed phyloge- netic classifications are yet to be elaborated, and are yet to be taken into the thinking of the consum- ers of botanical information. For the present, a con- servative approach is prudent, and so the treatment of floristic botany rests upon the standard, albeit flawed, Linnaean notions of species. Floristic studies account for all of the plants that occur in a particular region. Usually this is taken to mean the vascular plants, although the currently active Flora of North America project also includes the bryophytes. Hitchcock (1925) also includes a discussion on the methods of floristics, but without the notion of floras as encompassing summaries. The products of floristic studies are floras or man- uals. The two are similar and intergradant, but as with monographs and revisions, the former are more sumptuous, often in several volumes, while the latter are stripped-down for convenient use. Flo- ristic botanists derive their information from mono- graphs and revisions, but when no monographic studies have been done, they must prepare nonce- treatments with the information at hand. If a flo- ristic program required that all groups be treated at equal levels of sophistication, the flora would never be written. Floristics are best done by botanists with field- familiarity in their region who also have good her- barium and library resources. The techniques for synthesis have been largely intuitive, based upon the botanist’s memory and ability to organize great amounts of detail. But, just as cladistics and mo- lecular data added a huge new approach to mon- ography, electronic information management (“‘in- formatics’’) is changing floristics. It is now think- able that a floristic project can account for vast amounts of information that effectively lie fallow, and that, through floristic programs and their com- puter links, this buried information can be brought to the surface. To be certain, floristic projects that are based on informatic techniques are in their in- fancy, but the future impact is already evident. Three notable computer-based programs come to mind (but there are others, not mentioned here): (1) The magnificent summary of information on the North American Flora as compiled by John Kartesz in his Biota of North America Program (BONAP) MADRONO | | [Vol. 47. and distributed on a CD-ROM that was prepared. by Kartesz and Meacham (1999). (2) The detailed | Flora of Florida project centered at the University | of South Florida and prepared by Richard P. Wun- | derlin and assisted by Bruce F Hansen (a manual | was published in 1998). (3) The theoretical works. of Hugh Wilson at Texas A&M University. The ap-| plication of informatics technologies to floristic projects is not easy. The Flora of North America made an effort to incorporate informatics theory, which proved to be administratively difficult. EXPLANATIONS OF THE FIGURES | Figure 1 simply notes the cascade of information from monographs and revisions through the floras | and on to the consumers. The consumers are a mixed lot; here they are called “‘primary,”’ ‘‘indi- rect,”’ and “‘ultimate.’’ The primary consumers are scientific and academic professionals whose exper- | tise is not in systematic botany but whose experi- ence gives them the ability to judge the accuracy | of the information. These are the botanists’ col-. leagues. The indirect consumers are a large group | who use what is in floristic treatments essentially on faith. It is this group for whom the accumulated | wisdom in the herbaria and libraries is likely to be of greatest interest and least accessible. Floristic. projects have an awesome opportunity to connect this group with botanical information. The ultimate consumer is simply the person who needs infor- mation about a plant, e.g., the person who asks, “‘Is_ this crabgrass in my lawn? What do I do about it?”’ | In many states, the Cooperative Extension Services | are geared to accumulating information from pri- | mary consumers and delivering it to the ultimate consumer. Figure 2 summarizes the preparation of a mono- | graph, starting with the definition of the problem and the early survey work. Items 3 and 4 are crit- ical, for here the monographer’s experience (or if a graduate student, the experience of the student’s mentor) calls for the building of hypotheses and. expressing them as testable models. Items 5—8 are. the chief sources of information useful in monog- , raphy; they are not mutually exclusive, and some techniques have elements of two or more of these items. Clearly, comparative morphology is of great, importance because it is easily accessed in the her- | barium, there is a lot of it, and the techniques for using morphological information are of long tradi-. tion. Items 6 and 8 include such matters as polli- | nation studies, populational studies, introgression, | the role of climate change, etc. The last item has. become increasingly significant with the advent of | readily accessible Geographic Information Systems | (GIS). Item 9 is legalistic, mechanical, and utterly essential, for it is how the entities are given their correct names. Information from items 5—8 are as- | similated and the results are compared to the hy- — potheses and models generated in items 3 and 4. — BARKLEY: CONTEMPORARY BOTANICAL STUDIES N N Nn The Cascade of Systematics Monographs and Revisions EVAL EL OES TEAL een Ve os aan oh) ap Maal MT ila ee | 3. Primary Consumers: e.g., Ecology, Agronomy, Forestry, Conservation Biology, Geography, Pharmacology, History Books on Wildflowers, Natural History, Trees, Weeds Ee Teas INSTR The Ultimate Consumers "The Person on the Street" 2000] OREN MULLET EIR ILS Male SOILD: TLL a LLL LLL MLL LOL DAME 4. Indirect Consumers: Popular Works, e.g., Fic. |. Explanation in text. During the assimilation stage, phylogenetic (cladis- tic) techniques are applied, which yield justified evolutionary trees showing the current understand- ing of natural relationships. It is noteworthy that many phylogenetic studies that are based on mo- lecular data focus on higher groups, such as genera or families, and that species-centered phylogenetic studies often rest upon large components of mor- phological data. Item 12, integration, is the aligning of the information into the customary format for monographic studies. Keys, descriptions, specimen citations, sources of data, and conclusions drawn are presented in traditional ways, making the mono- graph (item 13) a readily understandable and use- able document, whether published as hard copy or on a website. Figure 3 notes the assimilation of information from monographs and other sources into a floristic treatment. Items 1 and 2 are obvious; the parame- ters of the project must be understood to account for the biological complexities of the region and the expectations of the intended users of the flora. Pri- mary information is taken from monographs and revisions as much as possible, but when no mon- ographic works are at hand, it is necessary to create treatments as best as possible; this step essentially incorporates items 3, 4, and 9 of Figure 2. Item 4, preparation of the treatments, is demanding and most easily accomplished by botanists with at least some monographic experience. Computer-assisted techniques are potentially very useful in item 4 (e.g., DELTA) but to date these techniques have long and steep learning curves. Many floristic bot- anists are not ready to embrace the computer as a tool to prepare keys, descriptions, and other textual matters. However, computer-based programs for generating maps are clearly with us. Text matters generated in item 4 can be entered into a website WN N O MADRONO Preparation of a Monograph or Revision Comparative Morphology and Geography (Herbarium) "Biosystematics" Breeding and Behavioral Studies, (Field and Garden) Cytological, Biochemical, Anatomical and Molecular Studies (Laboratory) Biogeography and Geohistory ees and Field) 132 Fic. 2. Explanation in text. to facilitate the following steps and the production of the final flora. Item 6 is also particularly de- manding, for here is where the tentative product is critically tested and edited for accuracy. Taxonomic reviews treat the botanical matters; regional re- views account for distributions and regional varia- tion. Item 7, amplifications, is the addition of in- formation needed by the intended users of the flora, e., the consumers. Item 8 1s where the manuscript is treated for editorial consistency, where the gen- eral keys are created and tested, and where the in- troductory essays are prepared and incorporated. The product may be published as hard copy (item 10) or posted on a website (item 11). A flora that is conveniently available on a website is easy for a primary consumer to consult when addressing broad questions (item 12), e.g., questions that were Systematic Monograph or Revision with Phylogenetic Interpretation Assimilation | is pecs not in the minds of the botanists who did the var-. ious studies that led to monographs or floras. A flora has a wealth of information relevant to distri- butions, variations, phenologies, etc., chaeology, medicine, and other areas not yet con- ceived. The point is that the hard data of the core | of systematic botany are translated for use by others through floras. CONCLUSIONS [Vol. 47 | " i that may be | coupled with soil types, geohistorical matters, ar- The abiding points are simple: There is a huge . body of literature in libraries and specimens in her- | baria that are the products of botanical enterprise. — Monographic studies are done to determine what entities exist, how to distinguish among them, how | 2000] BARKLEY: CONTEMPORARY BOTANICAL STUDIES 207 Preparation of a Flora or Manual Primary Treatments; 4. Amplifications, e.g., Horticultural Uses, Legal Status, Toxicity, Fic. 3. Explanation in text. they are related to each other, and how they behave. Monographic studies have spawned very interesting derivatives that relate to sophisticated understand- ings of evolution, but that still fall within the shad- ow of monographic studies. Floristic studies filter and assimilate the accumulated wisdom of the plants of a region and couple it to those who use the information. It is not for nothing that we recall a botanical beatitude attributed to the late Lloyd H. Shinners: “‘Blessed be those who write floras, for they shall discharge the botanists’ responsibilities to the public.” Note: A review of the growth of taxonomic con- cepts over the past half-century was recently pub- lished by P. F Stevens in a series of ‘‘Jubilee Pa- pers” in the journal Taxon (Stevens 2000). It ap- peared too late to impact the presentation of this paper at the Jepson Symposium in June, 2000. 1. Definition of the Project and its Application f vee Field, Herbarium and Library Surveys : 3. Monographs and Revisions Keys, Descriptions, Ecological and Distributional Data, Nomenclature 6. Taxonomic and Regional Reviews Y Weediness, Colloquial Names A 8. Integration and General Keys - 9. The Flora or Manual - 11. Flora on Website; & Interactive Database [ 12. "Broad Questions" : LITERATURE CITED BARKLEY, T. M. 1993. Synthesis: a historical perspective. Annals of the Missouri Botanical Garden 80:292— 296. CANTINO, P. D. 2000. Phylogenetic nomenclature: address- ing some concerns. Taxon 49:85—93. , R. G. OLMSTEAD, AND S. J. WAGSTAFF. 1997. A comparison of phylogenetic nomenclature with cur- rent systems: a botanical case study. Systematic Bi- ology 46:313-331. HALL, H. M. 1928. The genus Haplopappus: a phyloge- netic study in the Compositae. Publ. Carnegie Insti- tution (Washington) 389:1—-vili, 1-391. Hitcucock, A. S. 1925. Methods of descriptive systematic botany. John Wiley & Sons, New York. Jupp, W. S., C. S. CAMPBELL, E. A. KELLOGG, AND P. FE STEVENS. 1999. Plant systematics: a phylogenetic ap- proach. Sinauer Associates, Sunderland, MA. KARTESZ, J. T. AND C. A. MEACHAM. 1999. Synthesis of 258 MADRONO [Vol. 47. the North American Flora. CD-ROM. North Carolina StTugssy, T: EK 1975. The importance of revisionary studies Botanical Garden, Chapel Hill, NC. in plant systematics. Sida 6:104-113. STEVENS, P. E 2000. Botanical systematics 1950—2000; | WUNDERLIN, R. P. 1998. Guide to the vascular plants of, change, progress, or both? Taxon 49:635—659. Florida. University of Florida Press, Gainesville, FL. MaApDRONO, Vol. 47, No. 4, pp. 259-264, 2000 COUPLING SPECIES-LEVEL INVENTORIES WITH VEGETATION MAPPING DaAvip A. CHARLET Department of Biology-S2B, Community College of Southern Nevada, 3200 East Cheyenne Avenue, North Las Vegas, NV 89030 ABSTRACT The importance of high quality vegetation maps for land management is rapidly gaining recognition. Unfortunately, most vegetation maps in the western USA are old, have coarse resolution, or are not ground-truthed. Vouchers for these maps, even of the dominant species, are lacking. This makes natural resource management decisions, including those made during disasters such as fire, difficult or sometimes damaging because managers lack the basic information they need to make these decisions. In an effort to fill the information gaps, many vegetation mapping projects have been implemented across the nation, but those that want the maps often do not think of including species inventories in the mapping activities. At the same time, botanists continue to have difficulty finding funding for complete species inventories. This situation represents an opportunity to combine the skills of botanists with the needs of land managers. I present examples of vegetation projects I was involved in, in which I convinced the project leaders to incorporate plant species inventories in the mapping activities. The addition of species distribution infor- mation increased the quality, usefulness, and accuracy of these projects. Funding for species inventories can be found in restoration budgets. Botanists should take it upon themselves to involve themselves in mapping projects. Further, if botanists are willing to make their case, they should be able to convince the public and funding authorities to spend a little restoration money on species surveys before the need to restore arises. It remains extremely difficult to obtain funding for floristic surveys in the United States. This is true in spite of a steady, high rate of new plant species discovery for the past 100 years in Califor- nia alone (D. W. Taylor in Ertter 2000). Remark- ably, most floristic work is performed pro bono by both professionals and amateur enthusiasts (Ertter 2000). However, this situation is far from ideal and seriously slows the work. As botanists interested in plant distribution patterns, we know that our knowl- edge is far from complete. To increase our knowl- edge of these distributions at the necessary pace, funding must support species inventories. There- fore, we must be creative about how we present proposals to do this work. We need to learn how to convince management and funding agencies that species inventories are beneficial; not only inventories of vulnerable spe- cies, but also of the dominant, common, and un- common species. We can seize opportunities to fur- ther this goal by finding vegetation projects in our area and presenting to project leaders reasons why making voucher collections will improve the proj- ect’s usefulness. I here present examples from my experience of different situations where species in- ventories were included in, and improved the qual- ity of, other projects. VEGETATION MAPS A vegetation map was probably the first graphic display of plant distributions. Vegetation maps orig- inated with the military in order to provide basic information concerning the structure of vegetation pertinent to the movement of troops, maintenance of supply lines, cover, and other logistical concerns (Kuchler 1967). In the western USA, the first wave of vegetation maps were made for the purpose of resource extraction, beginning with Merriam’s life zones map (Merriam 1898). Through the 1950’s, most of the nation’s vegetation maps were econom- ic, a trend that Jepson resisted for decades (2001). Other important uses of vegetation maps began to emerge into the national arena in 1993 when the Gap Analysis Program (GAP) (Scott et al. 1993) was launched as a national project (Scott and Jen- nings 1997). However, plant species distributions are only rarely integrated into modern vegetation maps, in spite of rapidly developing Geographic In- formation System (GIS) technology. Even the GAP project concerns itself mainly with dominant plant species in order to predict the distribution of wild- life habitat (Scott et al. 1993), and species-specific distribution data are usually restricted to vulnerable species. Further, because of growing concerns about climate change, the abandonment of species- level surveys is considered prudent and these are being replaced by “‘plant functional type”’ classifi- cations (Smith et al. 1997). Today, most agency scientists know they need vegetation maps to serve as baseline data to man- age public lands. These maps must provide both ecosystem level and species-specific information. One layer should include structural information, that is, the distribution of the kinds (e.g., needleleaf evergreen, broadleaf deciduous) of forests, wood- lands, shrublands, grasslands, and riparian systems that occur on landscapes and in regions. Physiog- nomic and structural information is needed to man- 260 age for vital ecosystem functions (Kuchler 1967; Smith et al. 1997) upon which we depend. This information is also essential when planning for wildlife, recreation, and emergency procedures dur- ing natural disasters such as fire and flood. In ad- dition to managing for wildlife habitat and vulner- able species, accurate vegetation maps with spe- cies-specific data are needed to plan and conduct restoration projects, and to permit and monitor commercial activities. It is difficult to assign a dol- lar value to species-level surveys because they pro- duce irreplaceable basic information. However, many existing vegetation maps are fraught with problems, even at the structural, functional level. Technical aspects. Many of the most utilized maps are old (e.g., Wieslander 1940) or have coarse spatial resolution (e.g., Kuchler 1964). Most mod- ern maps lack ground-truthing, have little species distribution data, and only a few recent maps have vouchers. Our vegetation maps need this species- specific distribution data and ground-truthing. We can easily correct all these deficiencies at one time with some planning and a modest budget. Tying the species information to the vegetation map is easiest when the vegetation classes on the map are delimited by polygons in a GIS. I use the term polygon here to mean an irregular shape on a map with a sharp boundary that corresponds to an area on the ground of relatively homogeneous veg- etation (whose boundary on the ground is rarely so sharp). Polygon-based mapping allows for a spe- cies-level inventory and the collection of vouchers during the ground-truthing phase of map production because collections and observations can be made within and recorded for individual polygons. The polygons can be identified from aerial photography, as I did (Charlet 2000) by using the methods of Kuchler (1967), or from satellite imagery as Hogg et al. (1999) did by using the image segmentation approach of Ma et al. (2001). Once the polygons are in digital form, it is a simple matter to include species data in the attribute table when the voucher location data are precisely recorded. In this way, each polygon gets a species list. The hardest part of these surveys and mapping activities on the ground is actually getting to the sites. Once there, it is a simple matter of making voucher collections and adhering to strict record keeping standards, such as those recommended by Ferren et al. (1995). Species-specific distributions are easy to add to a polygon-based GIS map, even after the map is complete, by simply adding these distributions to the attribute table. SPECIES INVENTORIES CONCURRENT WITH OTHER PROJECTS Nevada wildlife map. In 1993, a research group at the University of Nevada was mapping wildlife habitat in Nevada. Since I spent much time in Ne- vada’s outback while conducting my Master’s and MADRONO [Vol. 47 | Ph.D. research, this group approached me to fill in | the details concerning the distribution of trees throughout the state and to make a vegetation map | of the state. I set about to construct a 1:1,000,000 | scale map of the eight vegetation zones of Billings | (1951) for Nevada. | In mapping the vegetation of Nevada, I included | the distribution of the different conifer species and mapped their occurrences. It was easy to collect vouchers after going to the trouble of getting to and | climbing these mountains, and so I did. Once the | time allotted for field work was complete and I looked at my list of collections, it appeared that I — had more than 100 cases of species in mountain | ranges not accounted for in the literature (e.g., Lit-_ fe 1971), | I was troubled by this result, and wondered how | many of my “‘range extensions’? were in herbaria | but had not been compiled. I went to 15 western herbaria with large Nevada collections, and found | even more conifer distributions that were neither | mapped by Little (1971) or Griffin and Critchfield | (1972), nor used in previous analyses (e.g., Wells — 1983). The changes were significant enough to war- | rant a new analysis, the results of which demanded | strikingly different conclusions (Charlet 1995). | Careful scrutiny of my collections led to other dis- | coveries, such as extensive gene flow between sev- | eral juniper taxa in the region (Terry et al. 2000). | Further, since publication of my conifer data for Nevada (Charlet 1996), others and I have found 6 | new county records for 4 species, and 12 new range | extensions. Altogether, information regarding the distribution of 8 of these 22 species and 9 different | mountain range conifer floras have changed since | 1996. There are more than 4000 vascular plant spe- | cies in the Great Basin/Mojave Desert region, but Nevada’s conifers represent less than 0.6% of that | flora. Clearly, we have only begun to map the dis- | tribution of the flora in detail. In fact, we are still | mapping the dominant species in the region. Lake Tahoe vegetation and wildlife maps. The | New Year’s Flood of 1997 was a harbinger of a | year filled with startling events in the eastern Sierra | Nevada (Horton 1997). The world-famous trans- | parent waters of Lake Tahoe had lost 8 m of clarity | in the previous 32 years (C. Goldman in Elliot-Fisk | et al. 1997), leading to a serious examination of the causes. In its final report to Congress, the Sierra | Nevada Ecosystem Project cited loss of water clar- | ity, drought, disease, and threat of catastrophic fire | (Elliot-Fisk et al. 1997). President Clinton then con- vened a Presidential Summit at Lake Tahoe in July 1997, a result being the declaration of Lake Tahoe | a national treasure. The President initiated a large | cooperative effort between the federal government, | California, and Nevada that would preserve the | lake (Clinton 1997). The federal portion of the | $900 million funding for 10 years of monitoring and restoration projects received final congressional | | 2000] approval in November 2000 (Las Vegas Review- Journal 2000a). To understand the ecosystems of Lake Tahoe, it is necessary to have a reliable vegetation map. For- tunately, by the time of the Lake Tahoe Presidential Summit, the GAP projects in both California (Davis - et al. 1998) and Nevada (Edwards et al. 1996) were complete or nearly so. It seemed simple to splice _ the maps, which the GIS technicians at a laboratory _ at the University of Nevada did. The resultant hy- _ brid map of the Carson Range had serious problems _ that led the Director to call me to see if I could devise a quick fix. I began by collapsing the two different classifi- cation schemes used on the maps to a simpler set of fewer classes (14) that were held in common by both maps. However, in the best reclassification scheme that I could devise, only 40% of vegetation classes along the edges matched. Even at the struc- tural level of forest, shrubland, and meadow, only 62% of the vegetation across the state boundary agreed. The only solution was to start over, and so I began developing new vegetation and wildlife maps for the Carson Range. I had one assistant and two months in the field to map 55 cover classes across 1340 square kilometers at 1:48,000 scale. Given so little time and so much ground to cover, the distributions I was concerned with were mainly the woody species. Nevertheless, this work yielded 7 new county records for 5 conifer species; this in a world-famous area within 100 miles of 2 major universities. The maps I produced for the Carson Range cov- ered only about half of entire Lake Tahoe basin, but their total cost was a mere $36,000, or 0.04% of the $900 million earmarked for Lake Tahoe res- toration. Further, this database is versatile, serving multiple functions simultaneously due to how the data were structured. The complete set of cover classes can be converted easily to structural classes or to wildlife classes according to the wildlife hab- itat types of California (Mayer and Laudenslayer 1988). In the attribute table, each forest polygon has a species list, in descending order, of the ar- boreal species. Species with an attendant collection in the polygon are noted. Further, the design of the map and accompanying attribute table lend them- selves to further augmentation with species-specific information in the future. Nevada Science Teacher Enhancement Project (N-STEP). What better way to promote high school science education than to introduce teachers and their best students to the construction of a scientific vegetation map in a remote Nevada wilderness? I thought this was my idea, but I learned later that Jepson (1934, 1935, and 2001, this volume) was doing something similar at U.C. Berkeley 80 years ago. It was Jepson’s student who made the vege- tation map I admired the most (Wieslander 1940), CHARLET: SPECIES INVENTORIES AND VEGETATION MAPPING 261 and this project collected more than 20,000 vouch- ers now at U.C. and elsewhere (Ertter 2001). My teams’ efforts in 2000 resulted in the discov- ery of a new Nevada record, Disporum trachycar- pum (David Charlet 2649 and Orne Grant UNLV, RENO). A key to this find was that our group had a visible presence in and demonstrated respect for the local community. In fact, this find was in Jar- bidge Canyon, merely 5 weeks after and 5 miles from the Jarbidge Shovel Brigade protest (San Jose Mercury-News 1999, Times-News 2000). The iron- ic twist is that we were led to the plant by a pro- tester and resident who had lived there for decades. GAP and re-GAP projects. GAP maps exist for all the states, and some states are beginning re-GAP projects (Scott and Jennings 1997). Regardless of the status of the GAP map of your state, GAP proj- ects are opportunities to conduct species inventories while mapping vegetation. We should grab this op- portunity and either improve the map during the re- GAP project, or ground-truth the existing GAP map. At the same time, we can conduct species inventories and collect vouchers, thereby improving the map, our herbaria, and our floristic database. WHERE IS THE MONEY? Big money is spent on our wildlands in two rel- evant areas: fire and restoration following fire. For example, Nevada’s first fire in the 2000 season, the Buck Springs Fire, conveniently occurred in the Spring Mountains in sight of my house as I was preparing this manuscript. I was shocked to learn that it cost $1 million to fight this 2000 acre fire (Las Vegas Review-Journal 2000b). One helicopter alone costs $53,000/day + $4000/hr. I admit it oc- curred to me that the daily fee is greater than my annual salary as a community college professor. But that was just one little fire in an ongoing firestorm. In early July 1999, a Nevada official de- clared the fire season was “‘of Biblical proportions” (Reno Gazette-Journal 1999a) and the season ulti- mately consumed 1.8 million acres in Nevada (Los Angeles Times 1999). Fire-fighting costs for the 1999 fire season in Nevada included $6 million in- curred by the state and $225 million by federal agencies (Reno Gazette-Journal 1999b). The crisis led Nevada Governor Guinn to announce what is probably the largest restoration project in the his- tory of the world (Reno Gazette-Journal 1999c), with $15 million in restoration costs anticipated. So for the 1999 cost of fires in Nevada alone, there 1s a price tag of $246 million. Another 660,000 Ne- vada acres burned in 2000 (Western Great Basin Interagency Fire Center 2000) and restoration plans are proceeding (Las Vegas Sun 2000). Nationwide, the Secretaries of Interior and Agriculture recom- mended to the President in September 2000 that $2.8 billion be spent for wildland fire programs, including $150 million for post-fire stabilization and restoration (USDA Forest Service 2001). 262 Nevertheless, I must ask myself, what species will be seeded and what seed will be used? These questions lead me to an answer to another question: How can species-level inventories fit into this spending? The answer is the seeds. Due to high demand and low availability, sagebrush seed now sells for up to $100 per pound, up from $30 per pound in normal years (Las Vegas Sun 2000). Here is a way to help floristic surveys to pay for them- selves during the short term: include seed collecting activities with voucher collecting and ground-tru- thing. In addition to our ignorance of the flora and its precise distribution, there is much disagreement about what “restoration” is (Billings Gazette 2000; Elko Free Press 2000). This situation has led the western USA to what I have called elsewhere a ‘biogeographic crisis’’ (Charlet 1999). This crisis, as relevant here, involves species distribution, rel- ative population levels, and the use of non-local or non-native seed in restoration projects. Even in areas where we have a good idea of the flora’s distribution, when large fire and flood emer- gencies occur, there can be no consideration for re- planting the region with seed from local popula- tions. We use what seed can be bought, no matter what the source. The introduction of other gene pools into a breeding population is background noise to the biogeographic signal present in the population’s genetic diversity, and may compro- mise the population’s long-term stability in the area. Further, these introductions threaten our ability to use our powerful new molecular biology techniques that allow us to look at a population’s DNA and to examine the nuts and bolts of its evolution. Great Basin ecosystems are reacting to recent changes in fire frequency and timing (Tausch and Nowak 2000) as well as water diversions and development (Castelli et al. 2000). Biogeographic patterns are clouding, and these changes may be irreversible. It is essential that we use the correct seed in the cor- rect places, and we cannot do this without baseline data and an established seed bank, both organized at the population level. CONCLUSIONS We can combine vegetation mapping with inven- tories by embedding polygons with species data into a GIS, and these species distributions should be documented, whenever practical, with vouchers. We also should collect seeds from the areas where we do our inventories and vegetation mapping. To be successful, the efforts of academia, agencies, and the public need to be coordinated and comple- mentary. Our knowledge deficiencies include ground-tru- thing and species inventory. To correct this, we must convince the public, legislators, and agencies that knowledge of this kind is inherently valuable. We need to take it upon ourselves to persuade ev- MADRONO eryone that this basic knowledge is valuable, and | spending money to obtain basic knowledge is a | good investment. Clearly, our restoration can only | be as good as the information available. There is | money: a mere 1% of $15 million restoration costs | 1999 Nevada fire season could yield | $150,000 for species inventories and vegetation — for the maps. Nationwide, only 1% of the $150 million earmarked for restoration following the 2000 fire season could represent $1.5 million for a large na- tional survey. The health and management of our ecosystems has captured the attention of both the public and its elected representatives, especially since the 2000 fire season consumed 7.3 million acres in the USA | (National Interagency Coordination Center 2000). These fires cost hundreds of millions of dollars to fight and hundreds of millions more in lost revenue. This public interest led the Western Governor’s As- sociation to have wildfires as the topic of their Win- ter 2000 meeting (Billings Gazette 2000). Through- out these meetings and plans, agencies must act as if ecosystem processes are understood and the dis- tribution of all species is known, and the public expects that the right decisions are made. But these things are not known. We are only now learning where the dominant species are, much less all the species in the flora. It is in the public’s interest that we obtain the basic information on the distribution of the flora, but it is up to us to convince the public that this is so. Vegetation maps, species data, and local, native seeds: all are needed for good resource management. With a little more effort than required for the vegetation map alone, we can include spe- cies inventories and seed collections and so en- hance these projects. Stimulating collaborations and powerful consen- sus can only arise when all parties are involved. We botanists, regardless of our affiliation or “‘am- ateur’’ status, need to cultivate relationships with every group. If we do, we will probably be sur- prised at what a tremendous pool of expertise and knowledge to which we have access. Local citizens are botanist’s allies. They live on the land, have intimate knowledge of their landscapes, and can take us to their special places. Agency land man- agers and scientists are also botanist’s allies: they got involved because they love the land and they love to serve the public. Outdoor recreationists are our allies too, as indicated by their choice to play outside on the land rather than in the gym. I would be remiss if I did not mention that more than once recreationists saved my crippled vehicle and me. Academicians are allies, especially if you come bearing good data and fine collections. Surely all parties will find common ground in the need to know what is on the ground and why, before we spend public money to restore it. ACKNOWLEDGMENTS My thanks to Barbara Ertter, Orne Grant, Paul Buck, Erin Noonan, Heidi Walters, Virginia Moran, and Jeanne Chambers, all of whom helped make this paper better. [Vol. 47 | 2000] LITERATURE CITED BILLINGS GAZETTE. 2000. Western governors weigh wild- fire risk. 12 December. Billings, MT. _ BILLINGS, W. D. 1951. Vegetational zonation in the Great Basin of western North America. Pp. 101—122 in Les bases écologiques de la régénération de la végétation des zones arides. Union of Biological Sciences, Series B, No. 9. Paris, France. - CASTELLI, R. M., J. C. CHAMBERS, AND R. J. TAUSCH. 2000. Soil-plant relations along a soil-water gradient in Great Basin riparian meadows. Wetlands 20(2):251-— 266. CHARLET, D. A. 1995. Great Basin montane and subalpine conifer diversity: Dispersal or extinction pattern? Un- published Ph.D. dissertation. University of Nevada, Reno, NV. . 1996. Atlas of Nevada conifers: A phytogeo- graphic reference. University of Nevada Press, Reno, NV. . 1999. Great Basin biogeography in the 21st cen- tury. Paper presented at the Great Basin Biological Research Conference, University of Nevada, Reno, NV. . 2000. Vegetation Map of the Carson Range, Cal- ifornia and Nevada. Biological Resources Research Center, University of Nevada, Reno, NV. ftp: biodiversity/pub/charlet CLINTON, W. J. 1997. Opening remarks by the President and the Vice President at Lake Tahoe Forum. The White House, Office of the Press Secretary, Incline Village, NV, 26 July. http://www.ceres.ca.gov/ HyperNews/get/forums/tcsf/5/1.html Davis, E W., D. M. Stoms, A. D. HOLLANDER, K. A. THO- MAS, P. A. STINE, D. OpIon, M. I. BORCHERT, J. H. THORNE, M. V. GRaAy, R. E. WALKER, K. WARNER, AND J. GRAAE. 1998. The California Gap Analysis Project—Final Report. University of California, Santa Barbara, CA. [http://www.biogeog.ucsb.edu/projects/ gap/gap_rep.html] Update (4/26/2000) Epwarops, T. C., JR., C. G. HOMER, AND S. D. BASSETT. 1996. Nevada Gap Analysis: an environmental infor- mation system. Unpublished technical report 96-5, Utah Cooperative Fish and Wildlife Research Unit, Utah State University, Logan, UT. ELKO FREE PREss. 2000. Angry ranchers confront BLM at fire meeting, by Gary Begin. 9 August, pl. Elko, NV. ELLIOT-Fisk, D. L., T. C. CAHILL, O. K. Davis, L. DUAN, C. R. GOLDMAN, G. E. GRUELL, R. HARRIS, R. KAT- TELMANN, R. LACEY, D. LEISz, S. LINDSTROM, D. Ma- CHIDA, R. A. ROWNTREE, P. RucKS, D. A. SHARKEY, S. STEPHENS, AND D. S. ZIEGLER. 1997. Lake Tahoe Case Study. Sierra Nevada Ecosystem Project: Final Report to Congress, Addendum. University of Cali- fornia, Davis, Centers for Water and Wildland Re- sources. ERTTER, B. 2000. Floristic surprises in North America north of Mexico. Annals of the Missouri Botanical Garden 87(1):81—109. . 2000. Our undiscovered heritage: future pros- pects for field exploration. Madrofio. 47:237—252. FERREN, W. R. Jr., D. L. MAGNEY, AND T. A. SHOLARS. 1995. The future of California floristics and system- atics: collecting guidelines and documentation tech- niques. Madrono 42(2):197—210. GRIFFIN, J. R. AND W. B. CRITCHFIELD. 1972. Distribution of forest trees in California. USDA Forest Service Research Paper PSW-82/1972. Washington, DC. CHARLET: SPECIES INVENTORIES AND VEGETATION MAPPING 263 Hoaa, J. T., N. S. WEAVER, J. J. CRAIGHEAD, M. L. Po- KORNY, B. M. STEELE, R. L. REDMOND, AND E B. FISH- ER. 1999. Abundance and spatial distribution of griz- zly bear plant-food groups in the Salmon-Selway Ecosystem: a preliminary analysis and report. Craig- head Wildlife-Wildlands Institute, Missoula, MT. Horton, G. A. 1997. The flood of 1997, final report: an analysis of snowpack water content and precipitation changes in the water basins of western Nevada and the effects on runoff and stream flows, December 16, 1996-January 6, 1997. Nevada Division of Water Planning, Department of Conservation and Natural Resources. Carson City, NV. http://www.state.nv.us/ cnr/ndwp/flood-97/floodana.htm JEPSON, W. L., R. BEIDLEMAN, AND B. ERTTER. 2000. Map- ping in forest botany. Madrono. 47:269—272. Jepson, W. L. 1934. Archives of the Jepson Herbarium, Jepson Field Book volume 54:28. . 1935. Archives of the Jepson Herbarium, Jepson Field Book volume 54:80. KUCHLER, A. W. 1964. Potential natural vegetation map of the conterminous United States: manual to accom- pany the map. America Geographical Society, New York, NY. . 1967. Vegetation mapping. Ronald Press, NY. LAS VEGAS REVIEW-JOURNAL. 2000a. Ceremony hails Ta- hoe preservation effort. 1 December, p. B2. Las Ve- gas, NV. . 2000b. Firefighters gain ground. 7 June. Las Ve- gas, NV. LAS VEGAS Sun. 2000. Rehab of 2 million acres ongoing in Nevada. 14 November. Las Vegas, NV. LITTLE, E. L., JR. 1971. Atlas of United States trees. Vol- ume 1: conifers and important hardwoods. USDA Forest Service Miscellaneous Publication Number 1146, Washington, DC. Los ANGELES TIMES. 1999. Arid Nevada scarred by its worst fire season, by Bettina Boxall. 24 October, p. Al. Los Angeles, CA. Ma, Z., M. M. HART, AND R. L. REDMOND. 2001. Mapping vegetation across large geographic areas: integration of remote sensing and GIS to classify multisource data. Photogrammetric Engineering and Remote Sensing 67 (in press). MAYER, K. E. AND W. E LAUDENSLAYER, JR. 1988. A guide to wildlife habitats of California. California Depart- ment of Forestry and Fire Protection. Sacramento, California. MERRIAM, C. H. 1898. The life zones and crop zones of the United States. USDA Division of Biological Sur- vey, Bulletin 10. Washington, DC. NATIONAL INTERAGENCY COORDINATION CENTER. 2000. In- cident management situation report. 24 November, 1000 MST. http://vwww.vita.org/disaster/wildfire/ 0011/0015.html RENO GAZETTE-JOURNAL. 1999a. Nevada wildfires burn record area, by Kevin Miller. 7 September. Reno, NV. . 1999b. Officials call this a fire season of ‘‘Bib- lical proportions,” by Kevin Miller. 3 July. Reno, NV. . 1999c. Unprecedented reseeding task lies ahead, by Tim Anderson. 31 August. Reno, NV. SAN JOSE MERcuRY-NEws. 1999. Town draws line in the land, by Brandon Bailey. 25 November, p. Al. San Jose, CA. Scott, J. M., EF DAvis, B. CsuTi, R. Noss, B. BUTTERFILED, C. Groves, H. ANDERSON, S. Caicco, EF D’ERcHIA, T. Epwarps, J. ULLIMAN, AND G. WRIGHT. 1993. Gap analysis: a geographic approach to the protection of 264 biological diversity. Journal of Wildlife Management 57(1) supplement, Wildlife Monographs No. 123. AND M. D. JENNINGS. 1997. A description of the National Gap Analysis Program. USGS Biological Resources Division. http://www.gap.uidaho.edu/ SMITH, T. M., H. H. SHUGART, AND E I. WOODWARD (eds.). 1997. Plant functional types: their relevance to eco- system properties and global change. Cambridge Uni- versity Press, Cambridge, United Kingdom. TAUSCH, R. J. AND C. L. Nowak. 2000. Influences of Ho- locene climate and vegetation changes on present and future community dynamics. Journal of Arid Land Studies 10S:5-8. TERRY, R. G., R. S. NOWAK, AND R. J. TAuscH. 2000. Genetic variation in chloroplast and nuclear riboso- mal DNA in Utah juniper (Juniperus osteosperma, Cupressaceae): evidence for interspecific gene flow. American Journal of Botany 87(2):250—258. MADRONO [Vol. 47 TIMES-NEws. 2000. Feds take on Shovel Brigade. Asso- ciated Press. 11 August, p. 1. Twin Falls, ID. USDA Forest SERVICE. 2001. Managing the impact of | wildfires in communities and the environment: a re- | port to the President in response to the wildfires of | 2000. Update: 5 January 2001. http://www.fs.fed.us/ | fire/wfp/president.shtml. WELLS, P. V. 1983. Paleobiogeography of montane islands | i | i i} \ in the Great Basin since the last glaciopluvial. Eco- logical Monographs. 53(4):341-—382. WESTERN GREAT BASIN INTERAGENCY FIRE CENTER. 2000. | 2000 Fire season overview. http://www.nv.blm.gov/ | 2wegbcc/summary/Overview2000.pdf WIESLANDER, A. E. 1940. Vegetation type and condition | maps of California and western Nevada. USDA For- | est Service, California Forest and Range Experiment | Station, Berkeley, California, in cooperation with — University of California, Berkeley, CA. ! ! | | ELECTRONIC ACTIVITIES OF THE UNIVERSITY AND JEPSON HERBARIA RICHARD MOE University and Jepson Herbaria, 1001 Valley Life Sciences Building, University of California, Berkeley, CA 94720-2465 This review treats computer activities that are carried out as part of the institutional agenda of the | University and Jepson Herbaria. Of course, individ- ual workers in the Herbaria depend on a variety of programs daily: e-mail has replaced to some degree letters and telephone calls and has gradually claimed an increasing part of our time. Word pro- cessors and spreadsheets have nearly completely re- placed typewriters. Programs that manipulate DNA sequences and generate hypothetical phylogenies are used by several of the staff and their students. Institutional use has as its nucleus the database de- veloped by the Specimen Management System for California Herbaria (SMASCH), the continuation and maintenance of The Jepson Manual, the revival and furtherance of Jepson’s Flora, publication of information dealing with nomenclature (including typification), and publication of information dealing with the history of the Herbaria. The electronic activities can be grouped into three broad classes (which overlap considerably): Curation—concerns of day-to-day operations of the Herbaria; Publication/Education—information made avail- able to colleagues and the general public Interactive outreach—uses which allow the Herbar- ia to benefit from the knowledge of the general public, as well as outside botanists, both amateur and professional. HISTORY Computerization of the Herbaria began in 1982 when Thomas Duncan, a pioneer in computerized taxonomy, assumed directorship. Supported by a succession of grants from the National Science Foundation, computerization was fostered by the Specimen Management System for California Her- baria (SMASCH) which began in 1991. As SMASCH got underway, Duncan moved to found the University Museum Informatics Project, which was closely associated with the development of SMASCH. Although SMASCH was conceived as a confederation that would include many western herbaria, only the University (UC) and Jepson Her- baria (JEPS) were able to obtain sufficient funding to proceed. SMASCH developed protocols for or- ganizing the tremendous variety of specimen infor- mation into databases and for automating herbari- um administrative procedures. The project consist- ed of a coordinator (Thomas J. Rosatti), a software specialist (Mickey Ellinger), and a data-entry crew, some of whom remained for nearly the whole proj- ect and others of whom were transient. The SMASCH software comprises a Sybase re- lational database that is accessed by data-entry rou- tines incorporating the X-window graphical user in- terface. The original goal of SMASCH was to cap- ture all label and annotation information for each vascular plant specimen from California in UC and JEPS, and to supplement this with a high-resolution image. It became apparent early on that these goals were too ambitious, and as a result, imaging was restricted to specimens of special importance, and recording complete annotation history was aban- doned. By the end of its final funding period in 1999, SMASCH had computerized more than 300,000 specimens—all of the California accessions and all of the North American holotypes (which were among the specimens imaged). During the project, emphasis was placed on optimizing efficiency of retrospective data entry and as a result, optimal pro- cedures for data retrieval were deferred. At present, data can be accessed via X-window screens, via direct Structured Query Language (SQL) queries against the Sybase data tables, or via a web inter- face. The first two methods are not usable directly by the public at large, but the web interface is avail- able to everyone: http://www.mip.berkeley.edu/ www-apps/smasch/. This interface, which is not yet complete, now allows users to produce a list of all specimens that: have a certain scientific or common name, or occur in a selected county (or counties), or were collected by a given collector, or were collected on a given date or range of dates, or contain designated “‘voucher”’ information, e.g., indication of flower color, chromosome count, habitat information. seat ae N The criteria can be combined so that it is possible, for instance, to request specimen data of all speci- mens of Pinus sabiniana Douglas collected by Jep- son between 1895 and 1900 in Napa or Solano counties. The interface also allows queries by col- lecting event (each collecting event being a unique combination of collector, date, and location). Thus one could retrieve all collecting events by Jepson in Amador County, and from that one could obtain any or all of the specimens collected at a given location. Web queries are made not against the 266 main database, but against tables extracted from the main database and optimized for retrieval speed. Therefore, the extract that is available on the web may lag behind the main database and not imme- diately reflect updates. It is also possible to retrieve lists of names grouped by county for which the Herbaria have vouchers. This feature was added recently at the request of people constructing county-level floras and is an example of how we hope to modify our web publications to serve the needs of the botanical community. In addition, the raw data have been made avail- able to the CalFlora Database http://www.calflora. org/, where they constitute most (more than eighty percent according to figures on the CalFlora *“‘In- formation about Datasets’”’ page) of the specimen data available at the site. The future of SMASCH. The database that was established by the SMASCH project will be main- tained by the staff of the Herbaria and will be ex- panded as time and funding permit. The following database tasks are now part of Herbaria routine: Modification of tables to reflect revisions of iden- tification or nomenclature; data entry of newly ac- cessioned California material or of returned loans; corrections of inconsistencies in the data. The orig- inal methods of data entry were designed to capture information from sheets in the Herbaria—retro- spective data capture. In the future we will have the chance to computerize specimens that are not yet accessioned, and for this we have explored new methods of data entry. Most collectors now make labels for specimens that they eventually deposit in UC/JEPS from databases or other computer files. When we receive new acquisitions in lots of several hundred, it works well to convert these databases or files to an intermediate format from which se- lections can be bulk loaded into the database. This reduces data entry to associating the collector’s number with a barcoded accession number. This is being done now with new accessions from Dean Taylor, Lowell Ahart, and Vernon Oswald, as well as with several lots of specimens in our “‘backlog.” We anticipate being able to handle the more than 100,000 bryophyte specimens deposited by Daniel Norris and specimens deposited by James Shevock similarly. We are also experimenting with data in- put via web forms. We have made available a label- printing form that can store data sent to it so that if the specimens in question are ultimately received, the corresponding data can be retrieved and trans- ferred to the main database. We also have a web form that will allow curators to enter information from any previously unrecorded type specimens that they encounter in the main collection. We are working to enrich the web query interface to the specimen database in order to allow a greater range of queries and to permit users to provide feedback automatically keyed to the specimen or name they MADRONO [Vol. 47 are dealing with. There may be users who are will- ing to help us rectify inconsistencies in the data- base, if we make the process convenient. For in- stance, a user might be able to add location infor- mation to a specimen by looking up another spec- imen collected by the same person on the same day. Electronic products relating to The Jepson Man- ual. The Jepson Manual (Hickman 1993) is a wide- ly used reference book that could be converted into an electronic product in numerous ways. Because the copyright is held by the University of California Press, however, use of the Manual other than in the form in which it was published has not been pur- sued. The electronic files from which camera-ready copy of the Manual was printed have been trans- lated in part into Extensible Markup Language (XML), and from this version we have extracted the names of the taxa and a variety of associated data, including distribution. Distribution is indicat- ed in the Manual by citing the bioregions in which a taxon occurs. Bioregions are hierarchical (Hick- man 1993, pp. 37—48): e.g., the Great Central Val- ley (GV) comprises the Sacramento Valley (ScV) and the San Joaquin Valley (SnJV), and is itself part of the California Floristic Province (CaFP). We have made a web application (http://ucjeps-herb. berkeley.edu/jeps-list.html) that uses the base map of bioregions from the Manual. The distributions are displayed on the map by expanding the com- posite regions and coloring in each smallest unit. The distribution records are modified as new infor- mation becomes available, and since the maps are constructed on request, they reflect current under- standing of ranges. Because of this dynamic gen- eration of the maps, there are no static pages for outside sites to link to directly. Furthermore, the URL for the page of a given taxon incorporates a compressed representation of the distribution, so the URL changes when the distribution changes. To circumvent this problem we maintain a simplified, but slower, access procedure for external links, whereby an incoming request that includes just the taxon name can be associated with other informa- tion. The tremendous quantity of taxonomic infor- mation available on the web often makes easy what was previously difficult or impossible. Much of the value of the web arises from hypertext links, but links are not easy to maintain—URL’s change, out- of-date URL’s remain in caches of indexing sites, methods of generating dynamic pages change with software modification. One of the challenges of website development is making pages flexible enough that they may be used in ways that the authors haven’t anticipated— without causing security problems. It is important to make each page independent of the pages to which it is linked from the main site, because con- text may be lost when visitors come from an un- intended page, or from an index cache. Therefore, we try to identify each page—whether it is dynam- | | | 2000] | ically or statically generated—with a title that is / meaningful in any context, and that has links back | to the Herbaria home page. | The Jepson Interchange. We have for some time | been considering the possibility of using web tech- _ nology to simultaneously track changes in Califor- _ nia floristics, make available expanded treatments _ of California plants (i.e., more extensive than those in The Jepson Manual), prepare for the second edi- tion of The Jepson Manual, and communicate with _ amateur and professional botanists interested in the | California flora (see http://ucjeps.berkeley.edu/far_ | westinitiative.html for an exposition by Barbara | Ertter of a distributed information system for native _ and naturalized plants). This project is now under- _ way, having been enabled by a grant from the Wil- liam R. Hewlitt Revocable Trust. The new project, called the On-line Interchange for Advances in Cal- _ ifornia Systematics, or the Jepson Interchange (http: //ucjeps.berkeley.edu/jepson-_flora_project. html), will provide a continuously updated author- itative list of California vascular plants, provide treatments for taxa not covered in The Jepson Man- ual, account for names not included in the list (syn- onyms, misidentifications), and most importantly, provide a convenient forum for all interested per- sons to assist in cataloging the California flora and understanding California plants. We now have web forms which can be used to report new records, call attention to publications pertinent to the flora, re- vise distributions, suggest hyperlinks, or contribute other information. The contents of the forms will be stored automatically in a database and e-mailed to an editor who will be the first stage of an au- thorization filter that may also include Jepson Man- ual editors and authors and other specialists. All submitted information will be available on the web, but only information authorized by the editorial committee will be incorporated into the list of taxa. As an adjunct to the Interchange, and in collabo- ration with the Digital Library Project of the Uni- versity of California and Xerox PARC, we will make available scanned images from Jepson’s ‘‘A Flora of California,” which contain a wealth of details on California plants. OTHER WEB PUBLICATIONS Index Nominum Algarum (INA). The INA (http: //ucjeps.herb.berkeley.edu/rlmoe/) is a card file maintained by Paul Silva at the Herbarium of the University of California. It contains nearly 200,000 names of algae (in the broad sense). Associated with the INA is a separate card file containing bib- liographic references pertaining to algal taxono- my—the Bibliographia Phycologia Universalis, or BPU. Cards that have been added since 1988 have been printed from a database, and the data are available on the web. As a preliminary step towards entering the remaining cards (pre-1988) in a data- base, and to provide archival protection, the cards MOE: UC/JEPS ELECTRONIC ACTIVITIES 207 have been scanned as digital images. These digital images can be used in conjunction with indexes ap- proximately like the physical cards can be used, but they are available from more than one site. We are making indexes to the images in two ways: via op- tical character recognition, and via forms that allow users to help by entering index entries directly. Purpus site. Carl Albert Purpus was a plant col- lector in western North America with an unpaid curatorial appointment at Berkeley. Barbara Ertter and Tom Schweich have innovatively combined the wealth of archival material in the Herbaria with specimen information to present historical, floristic, and related data about the North American collec- tions of Carl Albert Purpus in a globally accessible and informative manner for use by students, histo- rians, botanists, and interested laypersons (http:// ucjeps.berkeley.edu/Purpus/). We hope to have a variety of similar web publications in the future. Indian Ocean Catalogue. The Indian Ocean Cat- alogue (Silva et al. 1996) is a compilation of all published records of species and infraspecific taxa of benthic marine algae from the Indian Ocean. Published by the University of California Publica- tions in Botany in 1996, it was converted to a web version during the reviewing process (http://ucjeps. herb.berkeley.edu/rlmoe/tioc/ioctoc.html). The web version, which was generated by filtering some 75 files marked up for the troff typesetting program, allows a variety of searches, and is updatable by user input. CONCLUSIONS Because we are dedicated to increased comput- erization in the Herbaria, it is well to consider some of the assumptions and consequences of the pro- cess. The advantage of computerization is not that the traditional mission of the Herbaria can be ac- complished more cheaply and more rapidly, al- though this is sometimes assumed. In fact, com- puterization probably makes routine tasks more time-consuming and cumbersome. The real advan- tage is that tasks can be accomplished that were previously impossible, with concomitant increase of the value of the specimens and specimen data to the Herbaria, to other institutions, and to the public. As a simple example, during the production of The Jepson Manual (Hickman 1993), it was not possi- ble for contributors to check the vouchered distri- butions of species they were responsible for without borrowing all the specimens or visiting the Herbar- ia. When the second edition is assembled, it will be simple to provide all contributors with electronic reports of all the UC/JEPS specimens pertinent to their treatments, with distributionally noteworthy specimens flagged for their attention. A specimen database is institutional. It, like the rest of the Herbaria, needs to function in perpetuity. Resources need to be allocated to it forever. It must transcend hardware, software, and personnel. A sig- 268 nificant cost of computerization is the requirement for vigilance—not with respect to privacy or se- curity issues, though these are important—but with respect to changes in hardware, software, and per- sonnel. All of the changes must be accommodated, and neither too rapidly, which would lead to con- stant turmoil, nor too slowly, which might cause intermittent large disruptions. The Herbaria depend on the University for net- working, database servers, and expertise. As our computer applications come to be used by outside MADRONO [Vol. 47 | users, those users will be similarly dependent. As | Internet applications become more common, we | likewise depend directly on outside institutions. LITERATURE CITED HICKMAN, J. C. 1993. The Jepson manual: higher plants of California. University of California Press, Berke- | ley, CA. SILVA, P. C., P. W. BASSON, AND R. L. Moe. 1996. Cata- logue of the benthic marine algae of the Indian | Ocean. University of California Publications in Bot- | any 79. | MApRONO, Vol. 47, No. 4, pp. 269-272, 2000 WILLIS LINN JEPSON’S “MAPPING IN FOREST BOTANY” WILLIS LINN JEPSON, RICHARD BEIDLEMAN, AND BARBARA ERTTER Jepson Herbarium, University of California, Berkeley, CA 94720-2465 ABSTRACT A previously unpublished manuscript written by Willis Linn Jepson in 1938 describes the preparation and value of vegetation mapping as a field exercise in a forest botany course at the University of California at Berkeley. The resultant maps, of different sites in the Oakland—Berkeley hills, not only represent an invaluable baseline for charting vegetation changes since the early 20th century, but also provided the initial impetus for A. E. Wieslander’s Vegetation Type Mapping Project of California. INTRODUCTION RICHARD BEIDLEMAN AND BARBARA ERTTER Throughout his professional lifetime, botanist Willis Linn Jepson was dutiful about maintaining field books, as he called them. These pocket-sized, leather-bound diaries were used primarily for the purpose of recording the plants, often with detailed descriptions, that he encountered in his innumera- ble field excursions, especially throughout the state of California. But he would often include tidbits about people he met, article and books that he read, and anecdotes about other scientists, both old and new. And because he was continually thinking them up, he would write down suggestions for stu- dent projects under a heading “‘Subjects for Stu- dents,’ because he was a firm believer in “‘hands- on”’ science teaching. Jepson’s projects weren’t restricted to just plant collecting, classifying, and preserving. As a result of his own field encounters, his professional inter- ests stretched far beyond such narrow activities to involve plant physiology, genetics, evolution, ecol- ogy, and phytogeography. Because he believed in studying nature, not just books, he actually took many individual students and even entire classes out in the field. His suggested projects often in- volved field pursuits such as determining the most common plants in the Berkeley area, measuring the surface position of leaves of native species out-of- doors, or recording when native plants developed their flowers and leaves. Probably the most provocative field project idea that Jepson came up with, usually for students in his botany courses, was for small teams of students to pick a natural site in the Bay Area and map in acceptable detail all of the vegetation. Frederic Cle- ments, in his Research Methods in Ecology (1905) and his textbook Plant Physiology and Ecology (1907), described the use of quadrats and transects, and he even described ‘‘Formation Maps,”’ the lat- ter involving ‘‘an outline map in which the various zones, consocies, Communities, etc., are shown.”’ But few researchers, and even fewer students, were involved in the early 20th century producing ‘‘For- mation Maps” until Jepson actually set his teams of students to work on these projects just a few years after Clements published his ideas. Jepson’s students were becoming involved in more than mere plant listing. The mapping projects were definitely ecological in format and indirectly of greater significance and potential than even Jep- son might have initially predicted. When retirement in 1937 provided Jepson with more time for reflec- tion, he not only started going back over all of his field books and adding embellishments, but he also began writing about some of his research and teach- ing activities. Among these manuscripts was one that dealt with the vegetation mapping project. Be- cause of its outstanding historic value at a time when vegetation mapping and classification is re- ceiving much attention (e.g., Sawyer and Keeler- Wolf 1995), Jepson’s manuscript is published here for the first time. In a sense Jepson was correct in claiming that he was responsible for initiating, through senior pre- forestry student A. Everett Wieslander, forest map- ping as a widespread U.S. Forest Service program (Ertter 2001; Wieslander 1935; Wieslander et al. 1932-1945). However, there are chronological problems with the article, which Jepson wrote some 25 years after the pertinent events. Jepson initially stated that Wieslander and his classmate Frank Her- bert approached him in 1912 about teaching a course in forestry for students in the College of Agriculture. This seems unlikely inasmuch as Jep- son was on leave in 1912—1913. But at the end of his article Jepson wrote that the preparation of for- est vegetational mapping in “‘its final character flowed directly from the work of my botanical lab- oratory back in the year 1914.”’ Wieslander had had to take his upper division botany course from Har- vey Monroe Hall because Jepson was on leave in 1912-1913. But during the next academic year Wieslander and Herbert were seniors (Class of 1914), when they reportedly got the idea of doing a timber inventory at Muir Woods and were sent by Dean Hunt over to talk to Jepson, who had re- turned to teaching by this time. This is unquestion- ably the correct academic year for the discussion ZY between Jepson and the two pre-forestry students about mapping at Muir Woods, and it is relevant that Jepson was in the Muir Woods area three times during January and February of 1914 (as indicated by his field notes). In an oral history, Wieslander (1985) reflected that when he and Herbert approached Jepson about getting academic credit for a forest inventory and management project in Muir Woods National Mon- ument, Jepson “‘became hysterical and he started to sob and cry like a baby,’ apparently because the College of Agriculture was “‘trying to make a for- estry professor of me, and I’m a botanist.’” Anyone familiar with Jepson would have difficulty visual- izing him in a sobbing mode, although he did have a quick temper. It is true that some 20 years earlier Jepson had been irritated when it appeared that bot- any would be put under the College of Agriculture. But he had always been sympathetic with forestry, both from an academic and commercial standpoint, having spent much time with lumbermen in the field during preparation of his Trees of California (1909) and Silva of California (1910), both slanted towards forestry. Jepson was furthermore an insti- gator, adviser, and favored lecturer for the U.C. For- estry Club, which started in 1912 (Wieslander, in- cidentally, was president of the club in 1913). He was instrumental in having forestry established as an academic program at the university, with the Di- vision of Forestry becoming an entity in 1914 (Cas- amajor 1965). Indeed, for some time from 1911, Jepson’s academic rank was Associate Professor of Dendrology. He taught many a forestry student over the years and enjoyed running into and pro- viding help for these men when they became pro- fessionals. One could appreciate that Jepson might indeed have been a bit disgusted with two young students who wanted to do an economic forestry management plan in a national monument, which showed some immature poor judgment (as Wies- lander later appreciated). Jepson’s recommendation that the two young men conduct an overall woody plant mapping project at Muir Woods, with his help but without an economic focus, made sense and paved the way toward more comprehensive forest mapping and management, which would later be spearheaded by none other than Wieslander. MAPPING IN FOREST BOTANY WILLIS LINN JEPSON (from an unpublished 1938 manuscript in the Jep- son Herbarium archives; minor editing by B. Ertter) We first survey the plot, then draw the model. —Shakespeare, King Henry IV, Part 2 Mapping as a forest botany course exercise. From a very early period in the history of the flow- ering plant work in the Department of Botany there was at intervals occasional assignment of field MADRONO [Vol. 47. mapping as part of the work in Botany. The idea was original with me. Such assignments were made | to students doing independent work who expressed | a wish for a field exercise of this kind. As the years ran on and the value of the work became more and | more apparent, mapping was introduced into the. Forest Botany course as part of the regular require- ment. While field work had always been a regular | part of the course, this matter of mapping natural | areas was something quite new and unfamiliar. Because new and unfamiliar, many students mis- | trusted on the first day of a term their ability to do | any such thing, although it was in fact quite simple | though requiring pains and a large amount of field | observation. The students worked in pairs, some-_ times in threes. Two students were thus assigned to | each of various natural areas in the Berkeley Hills, - sometimes in the Oakland Hills, infrequently in | Marin County. Such a natural area was a small | drainage unit, a gulch, a canyon, a creek basin, or | a ridge. The first duty was to make on a manila sheet, 36 X 48 inches or somewhat smaller or somewhat larger, a preliminary sketch map of the | area, plotting its natural boundaries, creeks, and riv- | ulets, contour lines being drawn in by the eye. All | landmarks were indicated, such as large trees, | prominent rock outcrops, or knolls. If the area were | a canyon, the student moved around the summit of | the bounding ridge checking the position of his — landmarks and the flow of the contour lines which | indicated slope or elevation. A more elaborate method was used by students | who desired to take special pains. In the laboratory — a large sheet of white paper was fastened to the | wall; on this was projected by a lantern, enlarged | as required, a section from the topographic sheet of the United States Geological Survey showing the | area Selected. The student then drew in on the wall- sheet the lines of his map with a pencil and after-— ward finished it in ink. Or, yet again, the student could make a free-hand enlargement of his area | from a topographic sheet. Having made his preliminary map, the pairs of students were now ready for detailed field work, the object being to map the plant formations (grass- land, chaparral, or woodland), and after that the as- sociations within each formation (Fig. 1). The oc- currence of notable individuals were often recorded on the map, as well as various special biological features. A report upon the area, a description and discussion of the formations, and an annotated sys- tematic list of species was prepared by the student to accompany the map. This assignment had great training value for the following reasons: 1. The student was required to make a complete list of the woody species of his area, and he was, thus, called upon to perform an intensified bit of work. 2. In order to make a list of the species, he had to if = < \ 34 ieee 1 ee lb My ie" '7 onunent ai. aS is (ee SURE ara Sd —— DR. JEPSON ) BOTANY MOG ACs a. Fic. 1. JEPSON ET AL.: MAPPING IN FOREST BOTANY pg k| PECEIID 1 SALIX -LASIOLEDIS Zz PHYSOCARPUS- CAPITATUS 3, UMBELLULARIA-CAL1FORNICA 4 SDAMBUCUS-RACEMOSA 9. LONICERA- H1SPIDULA 6 DIRCA-OCCIDENTALIS 7 QUERCUS- AGRIFOLIA 8. RUBUS- PARVIELORUS 9 LONICERA - 1NVOLUCRATA 10 CORYLUS- ROSTRATA IL RIBES- MENZIES512 12. MOLODISCUS- DISCOLOR 132 RHAMNUS- CALIFORNICA 4. STACHYS- BULLATA 15, BACCHARIS-PILULARIS la ARTEMESIA-CALIFORNICA ¥ 17 LuPlLNUS- LitTORALIS 16: DiPLACUS- GLUTLNOSUS 19 ZAUSCHNERIA-CALIBORNICA 20 CASTILLBA- PARVIFUORA 2] RUBUS-= VITIFOLLUS 22, LATHRYS -VESTITUS 2 ARALIA- CALIFORNICA 24 RIBES- SANGUINEUN 2, SYMPHORICARPOS- RACEMOSUS 26. CORNUS - PUBESCE-NS 27. RIBES= GLUTINOSUM 76. ROSA—~GY MNOCARPA 2. HETEROMELES-ARBUTIFOLIA 30, SOLANUM ~UMBELLIFERUM 31 CREAN OTHUS- SOREDIATUS 32,.RHUS DIVERSILOBA. 33 ACER MACROPHYLLUM 34. [MBPLANTATIONS Se 25 aes PLANTED (9-16-1832 7 BERRY TE AGS AS SEI Example of map produced as a field mapping exercise in Jepson’s Forest Botany course: Head of Strawberry Creek, by Lila Bassett, undated (original in Jepson Herbarium Archives). Although this particular example lacks a date, other maps in the collection were prepared 1918—1923. Strawberry Creek is in the Oakland—Berkeley hills due east of the main campus of the University of California at Berkeley. Although still largely undeveloped, large areas have been significantly altered by the introduction and spread of non-native vegetation, including various pines, Eu- calyptus, and French broom (Genista monspessulana). Some of this vegetation is marked on Bassett’s ” (34). plantations learn what a species is in the field. It was nec- essary to compare hundreds and hundreds of in- dividuals of a given species and make sure that the differences amongst them were immaterial and that they truly belonged to one species. All species in the area had to be tested in this way. This was a highly practical exercise in the field recognition of species. 3. For his report, the student was shown how to observe various phenomena of the plant in order to determine as much as possible of its biology and life-history. He, therefore, became thor- oughly imbued with the idea of the plant as a living thing, since no plant was ever quite alike from season to season, not even from week to week. 4. The student pairs being assigned to different canyons or ridges, there was thus cultivated a spirit of independence and self-reliance. 5. The area being given to only two students, they developed in it a sense of personal pride and possession. Even after the final examination at Christmas, it was sometimes noted that students continued to study their area. When asked why map as “‘Im- they made yet another trip to it, the reply was: ‘“We wished to see what it is doing now.”’ So striking and significant, therefore, were the pro- gressive changes in the life history to the close observer. The sets of maps drawn by the Forest Botany students have been preserved (e.g., Fig. 1). They represent a valuable record of vegetation conditions in the hills at the time they were made. When hand- ed in by a class just before Christmas, the entire set was posted on the laboratory walls and in the cor- ridors of the Botany Building, where each year’s exhibition created much interest amongst visiting botanists, zoologists, and agriculturalists. Dr. Jo- seph Grinnell [Director, Museum of Vertebrate Zo- ology] declared these plant survey maps had sig- nificance in relation to the localized distribution of mammals and birds in the hills. Genesis of the Vegetation Type Mapping Project. In 1912 two students, E B. Herbert and A. E. Wies- lander, asked that I give a course in forestry for the benefit of the students in the College of Agriculture. There was at that time no division of forestry in 22 that college nor in the University. I refused to give such a course; the grounds of my refusal were that I was not a forester but a botanist in the College of Letters, and that my duties as a botanist more than consumed all my time and energies. But I added that I would give them work in botany which would be of the utmost value to them in all their lives as foresters and give them a real advantage over most foresters. So I put them at the task of making a botanical map of the Muir Woods basin on the south slope of Mount Tamalpais. Since these two students be- longed to the College of Agriculture, they wished to make an economic map of the area. I replied: ‘‘No. This must be a scientific map. To include only economic species would be ineffective because spe- cies of biological importance will be omitted; non- economic species tie in ecologically with economic species. The biological associations and interrela- tionships must be worked out; they are of funda- mental importance. Moreover, a woody species thought useless today may tomorrow be highly val- ued economically. An economic map is partial and temporary; a scientific map is sound in concept and of permanent value.’’ After a good deal of discus- sion, the two acceded to my view and went to work with great enthusiasm on their project. The final result was one of the most important reports on a natural area ever completed by students under the direction of my laboratory. Nor did the significance of it end here. Both men won scientific positions in the United States De- partment of Agriculture, Herbert in the Division of Entomology, Wieslander in the Forest Service. In the course of time, the California Forest Experi- ment Station was established, and Wieslander was called to its staff. A highly important project was undertaken, that of a woody (vegetation) type map of all California, showing all the important associ- ations in colors. The project was of the greatest scientific importance. Nothing on this scale had ever been done elsewhere in the New World, per- haps not even in the Old World. The central governmental authority at Washing- ton, contrary to Wieslander’s recommendation, de- creed after its fashion and custom that the map should be an economic map. Field work was pushed forward in southern California, and very soon, under Wieslander’s effective driving methods, a considerable number of quadrangles were mapped and made ready for use. Economic use of the maps MADRONO [Vol. 47 | oy other branches of the Forest Service, especially | by the various national forests, soon developed | such that in certain particulars the maps were de- | ficient though complete as to the original economic | conception. The explanation was simple. New eco-. nomic aspects developed so rapidly that it was_ proven repeatedly that an economic map was and | must be from its nature transient and insufficient. The federal authorities were finally prevailed | upon after many years to reverse their decision, and | Wieslander was authorized to carry out his original — plan, that of a scientific map. Moreover, he was | given a considerable appropriation to re-map the quadrangles in southern California. The map now | became a scientific map, that is to say a botanical - type map of the woody cover, and as such not was | only of importance to foresters but to California | botanists and to botanists everywhere. It appears to | be, in reality, the most important and comprehen- | sive botanical map of a large area ever undertaken — anywhere on the earth’s surface. It is naturally a satisfaction to consider that its final character flowed directly from the work of my botanical lab- | oratory back in the year 1914. LITERATURE CITED CASAMAJOR, P. 1965. Forestry education at the University © of California: The first fifty years. California Alumni | Foresters, Berkeley, CA. CLEMENTS, F. E. 1905. Research methods in ecology. Uni- | versity Publishing Co. Lincoln, NE. . 1907. Plant physiology and ecology. H. Holt & | Company. New York, NY. | ERTTER, B. 2000. Our undiscovered heritage: past and fu- ture prospects for species-level botanical inventory. | Madrono 47:237-—252. JEPSON, W. L. 1909. The trees of California. Cunningham, | Curtis, & Welch. San Francisco, CA. | . 1910. The silva of California. Memoirs of the | University of California 2:1—480. | SAWYER, J. O., AND T. KEELER-WOLF. 1995. A manual of California vegetation. California Native Plant Soci- ety. Sacramento, CA. WIESLANDER, A. E. 1935. A vegetation type map of Cal- | ifornia. Madrofio 3: 140-144. . 1985. California forester: Mapper of wildland vegetation and soils. (an oral history conducted in 1985 by Ann Lage). Regional Oral History Office, Bancroft Library. University of California, Berkeley, CA. | (in charge) and Forest Service Staff. 1932-1943. Vegetation type maps of California and western Ne- vada. U.S. Department of Agriculture, Forest Service. California Forest and Range Experiment Station (Gn | cooperation with the University of California). MApRONO, Vol. 47, No. 4, pp. 273-286, 2000 WILLIS LINN JEPSON—“THE BOTANY MAN” RICHARD G. BEIDLEMAN University and Jepson Herbaria, University of California, Berkeley, CA 94720-2465 ~ Near the tranquil summit of Cemetery Hill in the bustling valley community of Vacaville, California, stands a white Carrara marble headstone set upon a gray Sierran granite base. Carved on the tomb- stone is the name of Willis Linn Jepson, with the following inscription: ‘“‘Profound Scholar, Inspiring Teacher, Indefatigable Botanical Explorer, born in Vaca Valley August 19, 1867; died in Berkeley Nov. 7, 1946. In the ordered beauty of nature he found enduring communion.”’ It was east of the Araquipa Hills, near Alamo Creek, where Willis Jepson’s parents, William and Martha Ann, settled in 1857, coming out of Mis- souri in an adventure-fraught journey by bullock- drawn covered wagon, though earlier William had already experienced the gold rush days in Califor- nia. At the Vacaville homestead, which later would be christened “‘Littlke Oak”? for a seedling Valley Oak that sprang up in the woodpile, three daughters were born to the farming couple before Willis, and then a younger brother who later would die in a riding accident. About his boyhood Willis reflected that, ““We had things. We had horses and cows and chickens. And flour from our wheat and fruits from the orchard. But we never had any money.”’ How- ever, as the family grew up it became a tradition to celebrate May Day with an outing. Throughout his lifetime, when his schedule permitted, Jepson would always try to continue the tradition, to be out in the field around May 1, “‘the best days in the year for the open.”’ Two things which Willis yearned for as a lad were a gun and books. Willis’s father favored hard farm work and thought that reading was a terrible waste of time. The Jepson youngsters had one chil- dren’s book of California pioneer stories, and Willis would sit by the pot-bellied stove reading the tales. He envied the neighbor boy who had copies of Youth’s Companion. When he was able to save up a little money, the first thing Willis bought was a worn set of Charles Dicken’s novels, which would remain part of his library, as did the tattered pa- perback about pioneers. This personal library bur- geoned to the end of his life and eventually became part of the nucleus of today’s Jepson Herbarium Library. In those early days the Vaca Valley was sparsely populated beyond the village with farms and de- veloping orchards, but much of the surrounding countryside was natural, with nearby unspoiled wooded canyons and chaparral-covered ridges, marshes and meadows, tree-lined creekbottoms. At an early age Willis, encouraged especially by his mother, developed an interest in natural history. As he later reflected, it was she who had provided him with an “‘almost inordinate love of dancing, of the songs of birds, of the sound of running water, of the flowing wind waves rippling the field of sum- mer wheat—and all manner of joyous and pagan things.” But most formative, Jepson came to realize, was the opportunity as a teenager to attend the newly established Normal and Scientific School in Vaca- ville, which attracted able high school students from near and far. Under the tutelage of Wyandotte J. Stevens, A.M., the principal, young Jepson en- countered chemistry, geology, zoology, astronomy, mathematics, Latin, literature, philosophy, and his- tory. And then there was botany, which even in- cluded field trips and collecting. Here Willis was introduced to his first plant book, which may have been Volney Rattan’s new A Popular California Flora, written especially “‘for beginners.’’ One of his first preserved plant specimens was a Lepidium nitidum (Common Peppergrass) collected on March 6, 1884, probably around Vacaville. When his growing interest in botany occasioned a visit by the youthful Jepson to the California Academy of Nat- ural Sciences in San Francisco, his reception as a budding naturalist by the venerable botanist Albert Kellogg and the younger Edward Lee Greene was something he would never forget. At summer’s end in 1885 Willis Jepson rode the train to Berkeley to take the entrance examinations for the freshman class at “‘the great U.C.”’ Remark- ably, he was the first young person from Vacaville who had ever applied to the university, and he found the initial experience so trying that he was ‘*sick bodily and mentally”’ for days after the exam. Stress followed by relapse would unfortunately be- come, for Jepson, a familiar pattern throughout his lifetime. But there had been no reason for concern about the entrance exam, because he was readily accepted as a member of the Class of 1889, which four years later would graduate 42, including seven coeds. In a sense, the four years at Berkeley provided a release from farming-town bondage. After a some- what homesick freshman year, Willis came to life as one of the “‘Jolly Sophomore Boys.’’ He became a leader in the Class of 1889 fights and yells, spending evenings and even afternoons at whist and 274 cassino and attending coeducational parties, plays, musicals, readings, and art exhibitions. It is histor- ically significant that it was the Class of 1889 that initiated the university’s first yell and set a prece- dent for future classes with: “‘Rah! Rah! Rah! Cal- i-for-ni-a! U! C! Berkeley! Zip! Boom! Ah!”’ To be sure, the academic performance of Willis Jepson his first year had left something to be de- sired, with his poorest grades in geometry and al- gebra. But as a sophomore he proved somewhat more scholarly, at least outside of class, doing some botanical collecting for Edward Lee Greene, who had become Instructor in Botany in 1885, and among other pursuits Jepson initiated a recording of earthquake vibrations, noting that they generally occurred after eight in the evening. The high point of the second academic year, however, involved the wild sophomore celebration of Thanksgiving, which resulted in the kidnapping of several fresh- men, running up and down Bancroft singing and shouting, and keeping “‘the town awake in diverse way[s], having three separate bonfires.” At the beginning of his junior year Jepson com- menced an invaluable activity, namely keeping a diary. He prophetically promised himself to “‘re- cord everything that is worth recording.” “‘Not just a collection of dates—but rather a collection of thoughts.’ Although it started as what he called ‘“‘a student’s note-book,”’ this activity within a decade would become a routine. Over his lifetime his en- tries, including innumerable plant collections and detailed descriptions, field trip experiences, anec- dotes about fellow scientists, teaching and research suggestions, photography data, addresses, literature citations, would result in more than 50 compact, black morocco “‘Jepson Field Books,” an invalu- able record of a scientist’s professional lifetime. But the first diary, Jepson’s “‘student’s note- book,”’ was concerned with his final two years at the University of California, when he became an extra-curricular achiever. He worked on the Blue and Gold year book, soliciting advertisements far and wide, and in his senior year he went on to become business manager, then energetic editor of the literary magazine The Occident. In his diary he demonstrated his literary and artistic talents by writing descriptively and repeatedly about the beau- ty of San Francisco Bay sunsets. On campus he enthusiastically attacked fraternities and the univer- sity administration. On the other hand, he attacked his course work with less enthusiasm, especially Political Economy and U.S. History and Constitu- tional Law. His grade in Zoology was a disappoint- ing C, and at one time he was close to flunking chemistry and German. He worried constantly about examinations, and once, after getting only 14 hours of sleep during 48 hours of cramming, he was “utterly broken down.”’ The academic bright spots were his occasional botanical excursions with classmates, including Victor Chestnut, who would later become an im- MADRONO portant government scientist. But more influential | was his continuing relationship with botany profes- | sor Greene. During a week-long illness (“‘sauer- bawled”’ in student parlance), Greene saw to his | disciple’s well-being, plying him with medicine and | hot tea. And as Jepson reflected, “‘How many profs, | I wonder, would walk a mile on a cold morning to build a fire for a sick student?”’ Despite being ill the last two months of his senior year, Jepson still graduated with his favorite Class of 1889. Among interesting graduation statistics, — Jepson at 160 pounds was 1|5 pounds heavier than | the class average; at 5’10”, he was 1% inches taller | than the class average, and after graduation he | would eventually exceed 6 feet. His college occu- | pation, as cited in the Blue and Gold year book, | was “‘fire eater,’’ his future occupation “‘Lobbyist,”’ — and his favorite beverage “‘Calves’ Blood.’’ Hmm. | His most famous classmate was Lincoln Steffens, | whose future occupation was forecast as ‘“‘Would- | be Aristocrate,’’ but instead Lincoln became an as- | tute muckraker. Decades later, Jepson, demonstrat- | ing his continued dedication to that memorable | Class of 1889, would actually serve as its class sec- | retary, soliciting news of his remaining alums. Encouraged by Greene, Jepson now embarked on | a graduate program, was appointed Assistant in | Botany in 1891, and on April 20 commenced the field work at the Sutter Buttes that resulted in his | first publication, “‘Botany of the Marysville Buttes” (Torrey Botanical Club Bulletin 18, 1891), — in which he reported 110 species of plant. When | Greene read Jepson’s rough draft of the article, he told Jepson that the manuscript was too neat, that [Vol. 47 | “The New York people would remark over its pe- — dantic preciseness.”’ So Jepson roughened it up a | bit before submitting it. That fall Jepson became president of the newly formed Chamisso Botanical Club on campus, whose initial purpose was to generate a list of plants growing within 20 miles of the foot of San | Francisco’s Market Street, and which held frequent meetings into 1897. In 1892 he joined a small, se- lect group of individuals, headed by John Muir, to create the Sierra Club. The following year in early April Professor Greene and Jepson were in the field together, investigating the plants of the Marin Pen- insula and northward. Jepson designed a spring gar- den of native plants at the University, and with Greene’s help he started the botanical journal Ery- thea, which brought the young editor to the atten- tion of a wide array of scientists. At Greene’s bot- any laboratory and herbarium in South Hall Jepson was busily pursuing taxonomic work, with partic- ular emphasis on Umbelliferae, Polygonaceae, and Chenopodiaceae, together with Marshall Howe, Jo- seph Burtt-Davy, and Ivar Tidestrom, all members of this botanical quartet destined for distinguished futures. In 1894 Jepson first formulated plans for his magnum opus, The Flora of California. As early as 2000] - 1887 he had begun numbering his plant specimens, but the early efforts proved inadequate. He finally _ formalized the numbering in 1899 and set aside numbers 13,334—15,095 for those earliest speci- mens. He always claimed that he was responsible for starting the university herbarium. Eventually he would almost frenetically build up his own personal herbarium, separated about 1905 from the univer- sity herbarium, by himself and through a growing cadre of collectors around the state, as almost ex- clusive reference material for his writing of the Flo- ra. Surprisingly enough, Jepson would later em- phatically aver that “‘The Herbarium will be my monument, more than the Flora.”’ And today the Jepson Herbarium at the University of California does indeed constitute the ““monument,”’ associated with, but still distinct from the University Herbar- ium. The mid-1890’s became a period of ferment at the University. In 1895 Greene left Berkeley for Catholic University in Washington, D.C., taking with him not only his own herbarium but many of Jepson’s specimens. University of California fac- tions were divided between obtaining a new head for the department of botany or, as Dean of Agri- culture Eugene Hilgard urged, putting botany back under agriculture. Greene himself felt it better to let his “‘young team”’ handle the existing botany department with- out a chairman for a year. In February Jepson was granted a leave of absence with salary to study at Cornell University and was away from Berkeley until July. Meanwhile he somewhat shocked Greene by reporting that he had entered his name for the chairmanship; but as backup Jepson had also made an application to Oregon State College (Cor- vallis) “‘to fill the professorship of botany”’ at that institution. On August 12 the San Francisco Examiner car- ried a blurb about Greene’s leaving and went on to report that ““Willis Linn Jepson, who was Professor E. L. Greene’s assistant in the botanical department, will succeed his old chief.’” There were others, in- cluding Professor Willard Rowlee of Cornell Uni- versity, who anticipated that Jepson would indeed become chairman of botany. Greene, however, pointedly reminded Jepson that all he had promised was to request an instructorship for Jepson, not any chairmanship. By late summer, although a separate botany de- partment had been retained by the university, the new chairman and full professor was a Dr. William Setchell, who had earned an undergraduate degree from Yale and a Ph.D. from Harvard. Actually, Hil- gard had hoped to entice the distinguished Charles Bessey from the University of Nebraska for the po- sition, but California’s offer had fallen short. Two decades earlier, when Bessey had taught a three- week-long course on economic botany at the new university in Berkeley, he had been perfectly happy with a stipend of $300. BEIDLEMAN: JEPSON—**THE BOTANY MAN” O18 Dr. Setchell was slightly older than Jepson, was from the eastern Ivy League establishment, with a well-to-do New Haven family background, sported a Phi Beta Kappa key, already had his Ph.D., had been assistant professor at Yale, and was a “‘bom- bastic’’ extrovert. The more introspective Jepson received an appointment only as an instructor in botany from the Regents on August 13, 1895, with a salary of $1,200. Congratulations for this appoint- ment shortly arrived from his former classmate Vic- tor Chestnut, who was familiar with Setchell and optimistically commented that “‘He will speed up the biological side & will need you badly to bolster up the Systematic side. Hold your fort.’’ But the western farm boy from Vacaville, although with an egotism of his own, undoubtedly felt outclassed by Setchell; and in academia Jepson, at least in his own mind, would increasingly “‘suffer’’ beneath that domineering shadow until Setchell’s retirement in 1934. However, Jepson too would later sport a Phi Beta Kappa key, just like Setchell. In the fall of 1896 Willis Jepson received another semester’s leave of absence with salary, this time to work up his California collections at Harvard’s Gray Herbarium under curator Benjamin Robin- son’s direction. At the beginning of his trip, he re- ceived a solicitous letter from his new department chairman to take care of his health and be sure to look up the Setchell family when he passed through New Haven, Connecticut. When Jepson’s produc- tive stint at Harvard came to an end, he did visit New Haven, saw the Setchells, and enjoyed a din- ner with the famous William Brewer of Yale’s Shef- field Scientific School. Although Brewer’s plant collections from his “cup and down” days with the early California Geological Survey would not even come close to what Jepson would collect in his own lifetime, the number of California miles which Brewer covered in the California wilderness during the Civil War years may well have exceeded Jep- son’s eventual three-score years of extensive tramp- ing. By this time Jepson had already been on his own exploratory California treks. His first of many jun- kets to Yosemite National Park was in August of 1890, then two years later he was in the Yuba River Sierra, and in Mount Shasta in 1894. During July of 1896 he, alias ‘““Dusty Roads,”’ and Setchell, ali- as ‘“‘Weary Wilie,’? had a companionable botanical excursion by horse and wagon to the Santa Cruz Mountains and then east across the San Joaquin Valley to Yosemite. The summer of 1897 found Jepson in the wild Yolla Bolly and redwood coun- try of northwestern California, with former Cal football left tackle Loren Hunt as his able assistant, collecting new plants and recording the trip with his camera. Also, this was the first year that Jepson ever gave a public lecture, before a small audience isolated in the Santa Cruz Mountains. He talked, somewhat timidly, about the colorful massing of flowers in the springtime, illustrated with some of 276 the finest herbarium sheets he could find. He was well received, invited to stay overnight; and thus commenced a lifetime of popular lecturing. Indeed, 40 years later he would even be delivering a 30- minute radio broadcast from KLX, Oakland, on “The Message of John Muir’ to celebrate that great naturalist’s birthday. On May 11, 1898 Willis Linn Jepson took the final examination for his degree of Doctor of Phi- losophy, with a major in systematic botany, minors in plant physiology and paleophytology. His thesis was on the flora of Western Middle California, and comprising his Sub-Committee were Setchell and Dr. John A. Merriam, paleobotanist and future pres- ident of the Carnegie Institution. More than mere years now separated the Jepson of the Ph.D. exam from the Jepson of the freshman entrance exam, as he answered such questions as “‘Development of taxonomic knowledge of the California flora,’’ the ‘‘De Candollean System as modified by Bentham and Hooker,” “California Plant Areas and Com- munities,’ “‘Periodicity of growth” in plants, and ‘Floral changes during the Mesozoic Period.’’ His Ph.D. degree would be only the 10th granted by the University of California and its first in botany! It might well have seemed demeaning for Jepson to complete his degree under Professor Setchell. But that notwithstanding, ““The Botany Man” was at the gateway of his professional career. And there was no better way to launch it than with a series of 6 lectures on botany, illustrated with lantern slides, sponsored by the University of California and presented on Friday evenings in March and April of 1899 at the California Academy of Sci- ences. There would shortly be more than a purely aca- demic experience ahead. A tradition was develop- ing in the department of botany to have staff mem- bers spend summers out in the field. And 1899 was especially appropriate because it was the summer during which the Harriman Expedition, with its shipload of eminent scientists, was sailing into Alaskan waters. Thus it was that an intrepid quartet of U.C. young men including Setchell, Jepson, Hunt (now instructor with the U.C. Civil Engineer- ing Department), and the new assistant in the bot- any department, Anstruther Lawson (later to be- come botany professor at the University of Sydney in Australia), set out from San Francisco aboard the Bertha on their own Alaskan adventure. The group photographs of the quartet on shipboard show Jep- son reading Kipling while the other three were con- sorting with some of the female passengers. Once in Unalaska, the U.C. group moved into the bishop’s old home in the Aleut village of Iliu- liuk next door to the cathedral and Indian boys’ school. Non-smoker and non-swearer Willis was initiated into matesmanship by his devilish com- patriots with a baptism of cigar and pipe smoke and volleys of vulgarities. But out in the field it was time for botanizing, with the four busying them- MADRONO [Vol. 47 selves collecting mountain higher plants and coast- line lower plants, not to mention teaching the local Native American boys the fight yells of the Uni- versity of California, and socializing with Harri- man’s scientists, including John Muir and John Burroughs, when the latter were briefly at Dutch | Harbor. At summer’s end the quartet was back in Berkeley. Setchell delivered his lantern-slide lec- ture, ““A Botanical Trip to Alaska,”’ for a university audience, while Jepson received from the university | his appointment as assistant professor of botany. So | would a new century be ushered in. Assistant Professor Jepson was beginning a pro- , fessing career which would extend until his retire- | ment at summer’s end in 1937, becoming an asso- ciate professor in 1911, a full professor in 1918. He | embarked on an annual schedule that involved | teaching much of the academic year, with summers | and any other free time preferably out in the field. | He developed strong ideas about botanical field | work, asserting that it was not for weaklings. In the | earliest days he traveled on foot, on horseback or with mule, and for longer trips by railroad where | possible and otherwise with his camp wagon. When | automobiles became the public’s mode of traveling, | he warned that ““You must still go afoot if a real | botanist. No field botanist should become soft and travel only in an auto.” He would sleep under the stars, using a tent only to dry his plants in, cook his meals over a fire, trespass and collect where he wished, even in na- tional parks, and with nary a reprimand, bring along three presses (16” X 11”, of oak frame for lightness and strength, with straps of harness leath- er and nickel buckles and a leather handle), one of them for collecting and two for drying, a vasculum and a pick which he had designed himself, his knife attached by a buckskin string; and always mark in the field each collection folder with locality and specimen number. Eventually he prepared a collec- tor’s manual, which unfortunately he never pub- lished, commencing thus: “‘Exploring for plants in California. A handbook for making records, pre- paring specimens and guiding beginners in the ways of camp and trail.”’ It was fitting to start off 1900 with a new orga- nization at the University of California, the Field Club, intitiated on January 20 and devoted to ‘prompting tramping trips into the region about the Bay,” with Jepson being elected President. The summer of 1900 found Jepson and some colleagues far from just tramping around the bay, as they hiked into the southern Sierra towards the tallest moun- tain of them all, Mt. Whitney, accompanied by two dutiful mules, Hot Haste and Sierra, and a copy of Volume 1 of the Botany of California by Brewer and Watson, hardly a field-guide sized volume. It proved a small world in the high mountains, be- cause who should be encountered but the Univer- sity of California’s Joseph LeConte, long-time pro- fessor of natural history and author of A Journal of i] 1 \ ——————————— § 2000] | Ramblings through the High Sierra, which recount- ed his ramblings with the new 1870’s University _ Excursion Party. Just a year later LeConte would _ pass away during a Sierra Club hike in Yosemite. Although the Flora of California was Jepson’s | perpetual goal, ‘“‘a symbol of my life,’’ he began pursuing a variety of other involvements. His Ph.D. thesis, A Flora of Western Middle California, was published in 1901, and the next year he completed a small School Flora for the Pacific Coast. With a growing interest in forestry, that was the summer he spent in the lumber country of northwest Cali- fornia, especially investigating the mall oak indus- try but falling in love with the redwoods. When, in 1903, an influential group of Los An- geles citizens, concerned about forests and water- shed protection, approached the University of Cal- ifornia, President Benjamin Wheeler called upon Jepson and Professor Arnold Stubenrauch of the agriculture department(?) to conduct what turned out to be a very successful forestry summer camp at Idyllwild in the San Jacinto Mountains. The 10 lectures which Jepson delivered dealt with “‘Life- history of a Tree,”’ “‘Classification of Forest Trees,”’ and ‘Forests of California.” Pursuing this forestry interest, then, Jepson on his own began gathering pertinent information out in the field, becoming acquainted with lumbermen, and taking photographs related to the lumber in- dustry in California, which culminated with the publication in 1909 of his singular The Trees of California, and next year the definitive Silva of Cal- ifornia. He also incorporated more material on for- estry in his botany courses, was instrumental in en- couragement for a forestry school at the university, and when some of the students organized the For- estry Club in 1912, he, then as associate professor of dendrology, became a popular advisor and lec- turer for the group. Meanwhile, the first two fascicles of his Flora appeared in 1909, including gymnosperms, the wil- low family, oaks, and several other small families. The publications were illustrated with a few fine photographs by Jepson, and some line drawings, the three initialed by an inconspicuous M.H.S. be- ing drawn by Mary H. Smith. Incidentally, only rarely, as they were to note, did any of Jepson’s many artists over the years receive much, if any, recognition. The next fascicle, copyrighted in 1912, would include a lengthy section on Gramineae by U.S. Department of Agriculture grass authority A. S. Hitchcock. In future fascicles there would not only be detailed descriptions but extensive locality data for each species based upon herbarium vouch- er specimens. The academic year of 1905-1906 was Jepson’s first sabbatical leave as a professor, and in early July he was off for Europe as one of four American delegates to the Second International Congress for Agricultural Education, which met on July 28-29, 1905, at Liege, Belgium. Taking full advantage of BEIDLEMAN: JEPSON—**THE BOTANY MAN” wg his sabbatical, he then spent the fall and early win- ter in Great Britain, researching and recreating, with extended time at Kew Gardens where he ac- tually encountered historic plant collections from California, including type specimens gathered by von Chamisso in 1816. Then in early February he returned to the Euro- pean mainland, making the continental tour from Paris to Italy to Switzerland, and finally to Ger- many, where he settled down in Berlin, hired a Ger- man conversation teacher, and interacted with em- inent German taxonomists, especially at the Botanic Garden. In mid-May he was back at his favorite, Kew Gardens, and finally home to Vacaville in ear- ly August by way of Yellowstone National Park and the Columbia River. On every occasion abroad he was received as a distinguished visitor, dined and wined and regaled. Small wonder that in 1906 he was included among the 100 leading botanists in the United States in a report by Jaques McKeen Cattell. During the summer of 1909 Jepson, for the first time, joined an extended Sierra Club summer ex- cursion, this time in Yosemite and Hetch Hetchy. Jepson was in his element, hiking, collecting plants, taking photographs, giving a lecture on Sierra co- nifers and providing the farewell invocation. And since one of his hiking companions was John Mutt, there was many a conversation between these two, which were recorded in Jepson’s field book, includ- ing conversations about the proposed flooding of Hetch Hetchy Valley. Continuing for a number of summers Jepson would accompany Sierra Club trips, collecting, lecturing, reciting poetry along the trail, getting acquainted with a variety of outdoor- loving people from around the world, the groups sometimes including more than 200 individuals, covering many miles of the Sierra Nevada, and of- ten taking up most of a summer. Before Jepson knew it, another sabbatical leave year arrived, 1912-1913. Initially he had grandiose plans which included not only Kew Gardens and mainland Europe again, but followed by an around- the-world collecting tour. However, with his mother’s health in question, he settled initially for his first retreat to the Mohave Desert, at the Water- man Ranch near Barstow with the late Governor Waterman’s daughter Abby as his hostess. Abby had graduated from Berkeley in the Class of 1904, a president of Prytanean, the Women’s Honor So- ciety, during her senior year. Berkeley had been her home, but eventually she had to move to her fa- ther’s desert ranch because of her health. During this early June of 1912 at the Waterman Ranch, Jepson, alone and with Abby, botanized in the area, sometimes on foot, by horseback, or in Abby’s rig drawn by “‘the Blacks.” This became a Mohave Desert field tradition which would repeat itself many a time. In years to follow, when Jep- son’s Mohave-bound Santa Fe train, and later his automobile, would drop down from the Tehachapi 278 Summit, reach the first of the Joshua Trees and swing into the open desert of Creosote Bush, White Bursage, and distant vistas, the university professor was able to cast off his academic robes, so to speak. As he once wrote in his field book: “*... it caused my spirits to rise and at once I felt better in body as well as cheered and sustained in mind.” This first visit to the Waterman Ranch was for only 7 days, and then it was back to Berkeley. However, by the end of June 1912, Jepson was once more in southern California, this time to join the Sierra Club’s 5 week excursion through the moun- tainous Upper Kern River Sierran countryside. Again Jepson was in his element, hiking through wilderness with 200 club members and 40 assorted attendants, botanizing, lecturing, and in mid-July garnering alpine flowers on Army Pass at 12,000 feet, just south of Mt. Whitney. This hegira finally terminated on the east side of the Sierra at Lone Pine the end of July. The best was yet to come, during the autumn of his sabbatical. In mid-October Jepson was once more in the Mohave Desert, first at his tent site on the Waterman Ranch, and then joining his Sierra Club hiking companion James Rennie at Needles. There the two men bought a rowboat for $15, named it The Lotus, and proceeded down the un- tamed Colorado River to Yuma on a 15-day col- lecting adventure, a journey along the edge of the new state of Arizona replete with white water, a sunken boat, a presidential election, and Giant Sa- guaro photo ops. The remainder of Jepson’s leave was spent close to home, but filled with excitement. Since 1902 he had been dreaming about a botanical society for California. So on April 12, 1913, at his call some 20 people gathered in the meeting room of the Oak- land Public Museum to discuss the creation of such a society. Two weeks later at a general organiza- tional meeting in Oakland the California Botanical Society was established, and Dr. Jepson not only was elected its first president but would serve as editor of its publication Madrono for many years, and before summer’s end would make arrange- ments for its first annual banquet speaker. During late spring of 1913 Willis Jepson was headed once more for the Mohave. Abby Waterman picked him up at the train station with her road team, and he stayed again in his tent above the ranch. Soon out into the desert, at Calico Wash, he paused to write in his field book: ‘‘Sitting here on the ground studying flower parts under the lens is a pleasant occupation. When one’s eyes tire there are the desert ranges stretching one beyond another, and a soft breeze blowing from the west.’’ Even- tually he traveled on to the New York Mountains, Needles, and finally to San Bernardino where he went on a brief collecting trip with his friend Sam- uel Parish, the premier amateur botanist in southern California, before returning to Berkeley. It was now time to think seriously about a Cal- MADRONO [Vol. 47 | ifornia Botanical Society banquet speaker, and Jep- | son had candidates. The International Phytogeo- | graphical Excursion, with its entourage of impor- | tant European and American botanists, had been touring the Middle West and southern Rockies dur- | ing the summer of 1913, and in September was | coming to California as the state’s first organized | botanical excursion. Here Jepson would guide the group through Yosemite National Park. After a | week enjoying the incomparable Yosemite under | Jepson’s leadership, the excursion traveled by train | to Oakland, visited Luther Burbank at Santa Rosa, then proceeded to Muir Woods with Jepson, where | Alice Eastwood of the California Academy of Sci- | ences joined the assembly. Finally, the congrega- — tion went down to the Monterey Peninsula, where | the botanists were the guests at the Carnegie Desert © Laboratory’s Coastal Laboratory in Carmel. When the train arrived in Oakland from Yosem- | ite on September 12, there awaited an evening’s © entertainment. It was the inaugural banquet of the | California Botanical Society, with eminent and un- doubtedly exhausted phytogeography speakers for- | tuitously arranged by Jepson: Professor C. von Tu- | beuf from the University of Munich and Dr. Adolf Engler, Professor of Botany and Director of the | Royal Botanical Gardens at Berlin, where Jepson had sojourned in 1906. Jepson’s teaching assignments during the aca- | demic year of 1913-1914, including a routine Bot- » any | course, must have seemed a letdown after all | the events that had transpired the previous 2 years. | But he was back to the Mohave by late April and throughout May. This time Jepson, accompanied by Abby Waterman, her sister Dr. Helen Waterman from Berkeley, Mrs. Rice, and Herbert Manson (the | | ) Waterman Ranch foreman) embarked upon a _. month-long desert circuit from Barstow to the Old | Dad Mountains, Ord Mountains, Twenty-Nine Palms, Indio, Palm Springs, San Bernardino, and back to Barstow. Then in August Jepson made a collecting trip up the Sacramento River to Duns- muir. With the arrival of fall Jepson’s most disastrous | academic year commenced, which rightly or | wrongly would forever imprint its apparent conse- | quences upon Professor Jepson’s mind. In 1914 | William Setchell suddenly announced that he would be taking a sabbatical leave starting in the autumn, and, as Jepson expressed it, “‘thrust the Department | load almost wholly on me.”’ There had already been increasing antagonism between Jepson and Setchell in recent years. Jepson was heavily preoccupied with his work on the Flora, continuing to build up his herbarium by collecting trips, preferably long ones, in every available moment; had what seemed to him like a demanding teaching load; and had never cared for administrative duties though he might secretly have wanted to be department chair- man. Despite the ensuing travail, Jepson did manage 2000] to survive the academic-year demands while Setch- ell was absent, and was actually able to spend pro- ductive time in the field from late May until late July of 1915, especially doing field work in the central Sierra around Columbia and the middle fork of the Stanislaus River. By mid-September of 1915, however, he “‘snapped,”’ and went into the sanitar- ium at St. Helena, near the home of his sister Mary Elizabeth and her husband Frank Pellet. This sani- tarium would become a frequent retreat for Jepson because of its “‘tranquility.”” ““One finds there an atmosphere of peace, quiet, good cheer and hope- fulness.”> When Jepson requested another leave of absence from the university to recover from his breakdown, the dean reminded him that he had just had a leave with pay 2 years earlier, so this re- quested leave would have to be without salary. By mid-December of 1915 Jepson left the sani- tarium to continue his convalescence in Barstow at, naturally, the Waterman Ranch, where he remained until the end of March. The Waterman Ranch, ob- viously, had become Jepson’s favored retreat when he sought prolonged respite. And for those who have wondered, this explains the dedication which Jepson in 1936 so poetically penned for the second volume of his Flora: ‘“‘Abby Louise Waterman. Daughter of the desert and of a race of sages, pen- etrating observer of the arid wastes of mesa and playa, protector of the desert men against the errant flow of circumstance, to her is inscribed this dedi- cation page of the second volume of the Flora of California by the botanical traveler, who, driven from pitiless ranges and stone-dry hidden valleys in the year nineteen hundred and fifteen, found el- emental shelter at Waterman Ranch in the heart of the Mohave.” Much of the Barstow stay, this time, was not productive for botanizing, but during the winter months Jepson kept busy observing birds, prepar- ing what was probably Barstow’s first winter bird inventory. By April Jepson had recovered suffi- ciently so that he could make a trip to the San Di- ego Exposition, which he felt was a disappoint- ment, “‘unless for persons who seek merely to be amused.’” Then he carried out some field work in southern California before returning to Barstow, then Berkeley and Vacaville. Although Jepson would always look back on that Setchell sabbatical period as one of the two major calamities in his professional career, the other to come in the mid-1930’s, he did actively return to his teaching and field research routine by 1917, spending May of that year in Death Valley and from mid-July into early August in the White Mountains, busy collecting more plants. His inter- est, however, began to focus on coastal redwoods, with which he had become enamored during earlier field trips in the northwestern California redwood country. There was growing statewide interest in redwoods, including a drive for a redwood state park; in July of 1919 the Save-the-Redwoods BEIDLEMAN: JEPSON—‘‘THE BOTANY MAN” pa he) League was established, with Willis Jepson as a member of its Executive Committee. Immediately Willis Jepson embarked on more than a decade of public lectures around the state in support of the League and preservation of redwood stands. During the early 1920’s Jepson was at work on a new book, which would be published in 1923— 1925 as A Manual of the Flowering Plants of Cal- ifornia a 1200-page tome in a single volume (unlike Leroy Abram’s four-volume J//lustrated Flora, which began appearing at the same time, published by the Stanford University Press). Although the book was put out by the Associated Students Store at the University of California, the entire cost was borne by Jepson; and through the years he would frequently complain that while other faculty works were published at University expense, his never were. The Manual was not intended to replace Jep- son’s projected Flora, but with its comprehensive keys, many line drawings, and detailed species de- scriptions, the Manual would be the sole California botanical bible until the publication of A California Flora by Philip Munz and David Keck in 1959. It was in September of 1925 that Jepson finally moved into a home of his desire, at 11 Mosswood, a several-storied Mediterranean style mansion with red tile roof, on a prominence looking down into the lower end of Strawberry Canyon and the uni- versity stadium, well beyond the academic campus. Largely designed by Berkeley’s famous architect Julia Morgan and beautifully landscaped, with sev- eral attractive gateways into its walled enclosure, the home was embellished by Jepson inside and out with ornamentation both floral and faunal. In the downsloping first floor, paneled in redwood, was the fine large library and herbarium drawers. The key to the cabinet which held his type specimens was labeled “‘Holy of Holies.”’ Above, on the main floor, was the front room, with comfortable chairs and scattered Persian rugs, and three large overflow bookcases. The great west window looked out across San Francisco Bay from San Mateo north past Tamalpais to San Pablo Bay, a vista which Jepson especially enjoyed in early evening, with the twinkling lights of the metropolis. Among the many pictures on the wall was a large beautiful oil painting of the Suisun Marshes, with the mountains beyond and “‘magnificant clouds at which I look a great deal!’” Above the brick fire- place in the living room was a redwood panel with an engraving of a California Quail on one side, and on the other side Jepson’s boyhood favorite, the Acorn Woodpecker (which he incorrectly called a ‘California Red-headed Woodpecker”’), and in the center an engraving of ““Golden Eggs,” a yellow Oenothera (Oenothera ovata) that used to carpet the campus when Jepson was a student. On either side of the entryway into the dining room were ad- ditional recessed bookshelves. Even in the sparse bedroom, with its spartan wood-frame bed there was a bookstand by the bedside. Engraved into 280 wood paneling on doors were images of plants and birds, while at the front entrance, which faced the upper hillside, there were in the framing of the tall entrance door California Quail amidst twining run- ners of oak. It was rumored among the graduate students at the time that Jepson had prepared the home for a prospective wife, but alas she apparently married someone else. There is no question that Willis Jep- son, Over a span of many, many decades, was both a romantic figure and eligible bachelor in the imag- ination of many women of his acquaintance, wheth- er he had met them as one of his students on cam- pus, on a Sierra Club trek, at a scientific meeting back east, or even in London during a sabbatical. He was tall and lean, with craggy features, “‘rug- ged, like a tree”’ as one of his former women stu- dents reminisced. His personal correspondence abounds with appealing notes. But, of course, Jep- son would remained forever a bachelor. In December of 1925 Professor Jepson was off on a major sabbatical trip, this time to the Middle East with William Bade, dean at the Pacific School of Religion in Berkeley and a biblical scholar. This would turn into an extensive journey encompassing essentially all of the countries surrounding the east- ern portion of the Mediterranean Sea, with special emphasis on Egypt and Palestine. Jepson toured fa- mous landmarks from the pyramids and Karnak, to Mary’s Well in Nazareth and the Dead Sea, col- lected plants in the desert and among the cedars of Lebanon, visited institutes and universities, met Muslim botanists, Arab shepherds, camel drivers, bedouins, biblical scholars, British soldiers, geolo- gists, and archaeologists, and took enough photo- graphs to prepare a hand-tinted lantern slide pro- gram upon his return to Berkeley. In early June of 1926 Jepson went on to England from the Middle East, working again at his favorite Kew Gardens Herbarium until mid-July, as usual focusing on early West Coast plant collections. Then he scheduled a memorable trip to the Liver- pool area on the trail of one of America’s greatest frontier naturalists, Thomas Nuttall, the first bota- nist to cross from the Atlantic to the Pacific in 1834 and who, on his way back east in 1836, collected along the California coast from Monterey south to San Diego. Because of provisions of an uncle’s will, Nuttall had to return to his native England in 1841, occupying Nutgrove estate near Liverpool until his death in 1859. Jepson was ecstatic at ac- tually encountering Nuttall’s grand nephew, Dixon Nuttall, who as a young lad had met Nuttall; and then locating Nuttall’s old Nutgrove Hall (in 1926 it housed a girl’s school) and his grave at nearby Christ Church. In 1934 Jepson would publish an article in Madrono about Nuttall’s botanical excur- sion across America, the botanist heralded by the Missouri Botanical Garden as ‘““The Father of West- ern Botany’; and on October 25, 1935, he would MADRONO [ Vol. 47 | | deliver a dedication talk at Spring Valley Lake fom a Nuttall memorial there. Leaving England in early August of 1926 on the . Orca, Jepson stopped over in Ithaca, New York, to. attend the Fourth International Botanical Congress at Cornell University. He observed that ‘ a bewildering lot of men here,’’ many of them, of , course, old acquaintances, the ‘‘most terrible bore | of the Congress”’ being P. A. Rydberg of the New | York Botanical Garden, who always seemed to dwell on trivial points. During the 1920s the University of California botany department would move into the remodeled Palmer House and expand both its faculty, its roster | of graduate and undergraduate students, and its lab- oratory courses. Jepson had developed the basic | teaching philosophy that “every educated person | should know, at least broadly, the native forests, | shrubs and flowering plants in his own state, so that elsewhere he may be an intelligent traveler.”’ He | further felt that ‘“‘the object of botanical investiga- tion was to learn as much as possible about the plant and that every phase of the life-history pos- | sible should be represented in the record either by specimens or field notes... . Intelligent and critical observations with real things in view should be the © student’s aim, should be the things held in mind, | not, as I say, merely to collect, but to study plants.” © In his actual teaching, Jepson followed the peda- | gogical pattern of Louis Agassiz. Especially with , his graduate students, he would give adequate pre- | liminary instructions, then pay no further attention | until the student was ready to raise questions. Pro- fessor Jepson was always thinking up and writing down questions for students to contemplate, proj- ects for them to pursue. One invaluable field ex- | ercise initiated by Jepson for botany seniors was to have them do a detailed vegetational map of some | local site, an experience that in part provided the | model for development of our modern vegetational mapping programs. | By now, Jepson had taught long enough to have | generated a lengthy roster of former students, and | often he would coincidentally run into one of them | in the field, or hear from them by letter, sometimes with a query but often with a note of appreciation © for a remembered course. He not only taught in the | classroom but had the audacity on occasion to take | his students, from beginning classes to graduate | seminars, out into the field, even across San Fran- cisco Bay by ferry to the Marin Peninsula. And during a pause in a field trip, he would often bring | forth a favorite book, which he always had wrapped — protectively in paper or from a book bag, and read | aloud to the students, the likes of ‘““The Ode to True Romance”’ or ‘“‘The Old Three Decker’”’ (‘Full thir- ty feet she towered, from waterline to rail ...’’), Kipling and Stevenson ranking high among his | choices. There were some students who appreciated such an exposure to the liberal arts, while others must have felt a bit embarrassed. | ‘One meets | 2000] Among the botany faculty there were increasing schisms. Professor Setchell, as someone once not- ed, ‘“‘liked to draw people around him.”’ He was at ease with students, liked to be known as ‘‘Papa Setchell’’ or the ‘““Old Boy”’ and in turn called his favorites “‘nieces and nephews.” Jepson, on the other hand, “‘needed serenity and quiet in which to work.’ He received no informal nicknames from students, at least to his face, though behind his back Thomas Howell called him ‘“‘Jeppie’”’ and Helen- Mar Wheeler, “Linn.’”? Jepson zealously guarded his personal privacy, often keeping his office door closed to intruders, and even totally disappearing from sight for long periods, his whereabouts not even known to the botany department secretary. In the academic setting he assumed a formal per- sona. In his view most of his teaching colleagues were neither very qualified nor working hard enough, while he seemingly was working overly hard and being underappreciated. He was much bet- ter out in the field and had a faithful cadre of non- academic acquaintances, but in the field he might still appear in what we today would call somewhat formal attire, right down to a fedora with hat band, a white shirt with black tie, and a black vest, on occasion even a black suitcoat. He was, obviously, no longer the young collegian who had enjoyed dancing with Misses Spohn and Cora Smith and who, with classmate Wharff, sang with fervor, “‘So when a maiden kisses me, I’ think that I the Sultan Wee...” With respect to friends, an astute graduate stu- dent observed that Jepson had ‘‘at least 10 or prob- ably 13 circles of friendships,’ with very few in the inner circle, especially among the faculty. One of Jepson’s measures of friendship was complete loyalty to Jepson. For example, if a student or col- league gave plant specimens or attention to other than Jepson, the individual was quickly moved to an outer circle. Jepson himself admitted that he had a quick and violent temper, like his father, and it was readily directed towards those he came to dis- like. Yet with the select few in that innermost circle the friendship was mutual and would continue for years. Surprisingly, there were few individuals, as his graduate students Herbert Mason and Lauramay Tinsley had observed, who actually disliked Jepson as much as he increasingly seemed to dislike oth- ers. Despite the onset of the Great Depression, 1930 was a momentous year for the natural sciences at Berkeley, with the completion of the major Life Sciences Building at the lower end of the campus, the move commencing on January 5. Jepson should have been delighted. In the new building he had a beautiful office and a large private herbarium, sit- uated next to a seminar room and the secretary’s office. But after only the winter months on campus, Jepson was again afield, to the Kettleman Plains at the southwest end of the San Joaquin Valley in mid- April and back to Barstow and the Mohave Desert BEIDLEMAN: JEPSON—‘**THE BOTANY MAN” 281 for a week in early May. Then on May 21 he was off for another International Botanical Congress, this time in Cambridge, England, with a week at the Gray Herbarium and several days at the New York Botanical Garden before he took ship to Eu- rope. Since the Congress didn’t convene until mid- August, he inevitably spent most of the two inter- vening months at Kew. “‘It is a great delight to be here again,”’ he wrote in his field book. “‘I feel a real thrill at being once more within the bounds of the Garden. The collections are so vast, so rich, apparently inexhaustible are the botanical treasures stored here. And one is given so kind a welcome by all the staff that it warms one.”’ Jepson had once written that his favorite herbarium was his own in Berkeley, but his second fondest was the herbarium at Kew. Finally, with mid-August at hand, Jepson spent a short time in Oxford and then on to Cam- bridge University for the meetings, where he deliv- ered a paper on “‘The Role of Fire in Relation to the Differentiation of Species in the Chaparral,” and finally back to Kew. This time Jepson’s return to America was by Montreal and across Canada to Vancover by train. The next year Herbert Mason, who had essen- tially been Jepson’s assistant since 1925 and whom Jepson viewed as an accomplished and loyal dis- ciple, was appointed instructor, going on to teach the systematics course and supervising labs for Jep- son. As time went by, Professor Jepson came to handle only a graduate seminar, for which he would pick a particular topic such as Age and Area, Life Zones, etc., the topics often remaining the same every time Jepson taught the seminar, a pattern which students were quick to recognize. As Jepson frequently was absent for health rea- sons, Mason would periodically take over respon- sibility for Jepson’s classes. Mason received his Ph.D. degree in 1932 with a thesis on the paleo- botany of conifers and 2 years later assumed charge of the University Herbarium. Mason was a good and popular instructor, and he was an active partic- ipant in the departmental Calypso Club field trips, which is more than could be said for Jepson. Ulti- mately, in Jepson’s view this up-and-coming young botanist was beginning to lure away Jepson’s grad- uate students. And, indeed, he successfully did, starting with Carl Sharsmith. Furthermore, Jepson was incensed when Mason “took over’? Madrono, which Jepson viewed as his own. Thus commenced an academic situation which would equal the Setch- ell-Jepson relationship, until at long last Jepson swore that he never even wanted to hear the name ‘*Mason”’ mentioned in his presence; and indeed in Jepson’s will it was emphatically emphasized that Mason not share in any of the benefits or endow- ments. Meanwhile, in 1934 William Setchell, now a world authority on marine algae as well as being Jepson’s enigma, retired, although he continued carrying on research for some time. As if in cele- 282 bration Willis purchased a new car, this time a very attractive sporty roadster that attracted excited at- tention wherever the professorial botanist drove. Jepson now had only 3 more years to retirement. He was the oldest member of the botany depart- ment and was not without fame himself, in America and in Europe as well. He might personally antic- ipate being named chairman of botany, at long last, or at least being consulted; but other pressures and personalities were abroad. A confidential commit- tee of the university had recommended, once again, that botany be incorporated into the College of Ag- riculture. As a matter of fact, Dennis Hoagland, a plant nutrition authority, became chairman of the amalgamation of botany with the Division of Plant Nutrition. For Jepson, this was the ultimate treach- ery, with botany for the second time in his academ- ic career losing its specific identity. This mid-1930’s period for Jepson came to rank with that terrible period in 1915. Worse now, in a sense, because he was older, much more taciturn and peevish, not in particularly good health, and increasingly driven by the necessity to complete the Flora before time ran out. Yet he was honored by the university with the prestigious invitation to de- liver the University Research Lecture in Wheeler Auditorium on March 20, 1934. He entitled his ad- dress *““The Content and Origin of the Californian Flora: A Demonstration of Scientific Methods,” concluding with . the joy of science is in never ending explo- ration and discovery and acquisition. I thought in the ardor of youth to build me a temple to my science. As I cleared the ground there in full view were the foundations of a previous temple, and under that the foundation of another and still others builded by the men who had gone before me. And in some parts the walls were sound, and in other parts only ‘the ruined footings ran.’ And I knew that in due time my temple too would be but as rubble and as ruin. As I thought of the future, of the far horizons that will open to view in the greater days to come, in the words of the great poet of Sussex, in his very words, I carved upon the lintel stone: ‘After me, cometh a Build- er, Tell him, I too have known.’ It was this year when the first edition of his Shrubs & Flowers of the Redwood Region was is- sued by the Save-the-Redwoods League, a popular publication still in print. Even amidst the Depres- sion Jepson was receiving university research funds for his Flora work and was able to hire both artists and clerical help. As mentioned earlier, the second volume of the Flora would appear in 1936 (Cap- paridaceae to Cornaceae). Little could Jepson an- ticipate that the last part of his uncompleted Flora Vol. 4, Part 2, on the Rubiaceae, would be pub- lished in 1979 and would be written by Laura May Dempster, who had been not only one of his grad- ”) MADRONO [Vol. 47 | uate students but artist and personal assistant for the Flora into the 1940’s. In the early fall of 1935, | having been granted a 6-week leave of absence, | Professor Jepson attended his last International Bo- | tanical Congress (the Sixth), this time in Amster- | dam, with the usual stay at Kew Gardens. At the | Congress he presented a paper on “‘Centers of Plant | Endemism in California in Relation to Geological History.”’ Back in Berkeley, just before Christmas, he was honored by delivering the keynote address | for the University of California Forestry Alumni | Dinner. | Jepson took delight in proclaiming that he was the only American to attend all of the International | Botanical Congresses, but in truth he did not attend — the First, in Genoa in 1892. True, he was in Europe | in 1905, where the Second Congress took place in | Vienna during June, but he didn’t arrive there until July, and the meeting for which he was a delegate | was an international congress for agricultural edu- - cation. However, Jepson did take part in the Third, | Fourth, Fifth, and Sixth, and he was an official del- | egate for the last two. The academic year of 1936-1937 would be Jep- son’s final association as an active member of the | University of California faculty, and his major. voiced regret was that he was retiring just when he | had encountered his two most talented graduate stu- | dents, Robert Hoover and Joseph Ewan. Inciden- | tally, Hoover received his Ph.D. in 1937 and actu- | ally worked as an assistant for Jepson until 1942. | He was recognized by Jepson as ‘“‘one of the ablest | collectors in California. He had a trained eye for ; material of importance.’? Hoover would eventually | go on to his own successful career in academia. | Ewan, on the other hand, left for a teaching position | at the University of Colorado and would never | complete his Ph.D. But moving eventually to Tu- | lane University, he pursued the subject of early | American naturalists throughout his lifetime, some- thing in which Jepson had interested him, eventu- | ally becoming an eminent historian of American natural history. | Willis had spent the summer of 1936 in his fa- . vorite redwood country of northwestern California. During autumn he taught his advanced graduate | seminar. But after Christmas he was frequently in | poor health, and was in and out of the sanitarium | at St. Helena until the end of April, 1937, with Her- | bert Mason handling his classes. Then he recovered | sufficiently to make a collecting trip to the Sierra | around Jacksonville, and for several weeks in May © back to the Mohave Desert and Barstow. ! Although the academic year of 1936—1937 didn’t | end until August 19, the University granted Jepson | a leave of absence from July | to that date. During the latter part of July, 1937, he drove north to visit — two former graduate students, Helen Gilkey at Or- | egon State in Corvallis, and Lincoln Constance, who was on the faculty at Washington State, whom he warned not to take the offered position at Berke- — _—~——-— i} i | | | ) } } | ~ 2000] ley. Then on July 31, near Crescent City on his way home, Jepson stopped at a roadside spring to get a drink, caught his foot on a shrub branch, and seri- ously fractured his ankle. He went into the Knapp - Hospital in Crescent City until August 18, and then was at the St. Helena Sanitarium until the end of - September. This was not an enjoyable introduction to retirement, and to compound matters he had to have his immense herbarium moved down into the basement of his home on Mosswood, where it shortly began to suffer from insect attack. Now no longer with academic obligations, Jep- son was free to get into the field when he pleased, health permitting. He spent almost two weeks in April of 1938 in the northern Sierra, going up the Sacramento River Valley to the Red Bluff country- side in June after a May of intermittent botanizing in some of his favorite old haunts around Vacaville and the Napa Valley, with a brief stay at the sani- tarium. Undoubtedly the high moment this partic- ular year was when he was invited to deliver the banquet address for the Silver Jubilee of the Cali- fornia Botanical Society, which he, of course, had started; and his fitting though perhaps self-serving topic was the Society’s beginning years. Jepson had been involved on and off with Ran- cho Santa Ana Botanic Garden since 1926 when he supported its establishment. He consulted with Mrs. Bryant and her staff and became a councilor in 1933. During April of 1939 he combined a council meeting there with an extended collecting trip in southern California. In June he attended the Amer- ican Association for the Advancement of Science (Pacific Section) meetings at Stanford, spending three weeks later in the summer up in the Feather River region of the northern Sierra. The next year his field schedule would have exhausted a botanist much younger than he when he spent almost a month afield in the Death Valley region, then three weeks in June and early July for a junket to Seattle for a AAAS meeting, driving up and back through the Sacramento Valley and central and western Washington. During October he was back at Ran- cho Santa Ana for another council meeting. Jepson’s last extended excursion of his career was during April of 1941, toward the end of his 74th year, although he would continue to take short field trips after that time. He had driven from Berkeley down to Rancho Santa Ana for another meeting of the Botanic Garden Council on April 19. But this time the meeting was a “‘peaceful”’ event for Jepson because he had ‘‘no duties, no obligations,”’ and initially only the staff was aware that he would be present. But after Dr. Carl Wolf’s public lecture on conifers in the auditorium, Jepson was absolutely delighted upon being recognized, because ‘‘a little procession of visitors lined up’’ to get his autograph. He and Wolf had planned for an excursion after the Council to the Old Dad Mountains (actually the Granite Mountains) in the eastern Mohave, the pair BEIDLEMAN: JEPSON—**THE BOTANY MAN”’ 283 to meet for the trip at the Van Dyke Ranch in Dag- gett, which Jepson had often visited over the years. This was the ranch started by “‘Judge’’ Theodore Van Dyke, brother of the author of the classic The Desert, John C. Van Dyke. The Judge was a writer in his own right as well as Daggett’s longtime jus- tice of the peace. The Judge had died in 1923, and Jepson’s host in 1941 was his son Dix, also a “‘des- ert writer.’ Still in residence at the ranch was Abby Waterman’s friend Mary Beal, with whom Jepson first became acquainted during his prolonged stay in the Mohave three decades earlier. Suffering from tuberculosis, Mary, on the recommendation of John Muir (whose daughter Helen was recovering from the disease at the Van Dyke Ranch), had left her librarian job in Riverside and moved to the desert about 1911. Over the years Mary became an au- thority on the desert vegetation, photographing and writing about the flora, and was one of Jepson’s prime collectors and collaborators. As he observed, she “‘knows every plant in her desert that has any- thing of popular interest.”’ Leaving Rancho Santa Ana, Jepson drove over Cajon Pass to Barstow, where he received the sad news that his old friend Abby Waterman had passed away just two days earlier, on April 19, in Berkeley. Continuing east to Daggett, Jepson reached the Van Dyke Ranch where sleeping quarters were found for him. At the ranch Mary Beal invited him to use the enclosed front porch of her cottage as his lab- oratory. Also, because Jepson had never been en- thusiastic about the Dix cooking, Mary fed him his meals. The two took some local collecting trips, and back at the cottage Mary changed his plant driers, spreading damp specimens out in the sun. Meanwhile, Carl Wolf showed up on April 27, and he and Willis departed in Wolf’s field vehicle for a thorough investigation of the Granite Mountains and the Kelso Dunes. This was Jepson’s second desert excursion with Wolf, the two having been together during late April of 1935, when Jepson had been so impressed not only by this up-and- coming young Rancho Santa Ana botanist, who proved to be ‘‘a capital companion,”’ but by the innumerable mounts for plant driers festooning the top and sides of Wolf’s automobile. It was on his return from this 1941 expedition that Jepson was invested by the University of Cal- ifornia with an LL.D. degree, an honor which in part made up for what Jepson felt were years of neglect. However, when he was first told about the proposed honor, he actually wrote a letter (unsent), in essence declining the degree as an honor too late and too little. On December 7, 1941, of course, World War II commenced, which irritated Jepson on several scores. He lost his research assistants and research funds to the war effort, and as a staunch Republican he had to suffer under a Com- mander-in-Chief by the name of Franklin Delano Roosevelt. During Willis Linn Jepson’s last decade, his ob- 284 session with completing the Flora, now focusing on the demanding Scrophulariaceae-to-Compositae, was shared with a desire to elucidate his lifelong accomplishments. With a sense of manifest destiny, from his earliest days Jepson had felt a compulsion to save everything pertinent “‘as part of his life and autobiography’’—his extensive correspondence, field notes, photographs, honors, mementos, pass- ports, business cards, programs, rough drafts of ar- ticles, and so on. As he jokingly observed, ‘You would think ... all this old stuff of mine were as valuable as pearls—real pearls!’’ Secretaries and assistants were busied organizing the correspon- dence, preparing letters so that they could be bound into volumes, which eventually would total 51 tomes, including a comprehensive index. Each bound volume includes 300—400 pages of epistles, with the total exceeding 15,000 items, all of which are now part of the Jepson Herbarium Library. In addition, there are reams of correspondence which have never been bound. The year before he died, Jepson perceptibly wrote with respect to this vast collection of correspondence that “Such a file will be, in the future, consulted by many persons and should be available to any one.”’ Some of Jepson’s correspondents were remark- ably faithful through the years. His Vacaville boy- hood chum Ralph H. Platt sent chatty, homespun letters from 1888 until his death in 1928, often with a jocular salutation such as “‘Dear Billious.’’ Equal- ly delightful were the dozens of humorous letters and postcards between Jepson and Harry Dutton, a San Francisco businessman, Stanford graduate, and amateur botanist, between 1908 and 1945. The pair frequently greeted each other in correspondence with localities which they had visited: “‘Dear Val- lecito,”’ “‘Yours, Mission San Luis Rey.’ Willis Jepson’s dearest relative throughout his adult life was his niece Dorris Pellet, daughter of his sister Mary and Frank Pellet of St. Helena. Dorris had a varied professional career in public service around the world, never married, and was considered a mirror image of her uncle in terms of personality, profile, and interests. Their letter writing extended over the years, and during Jepson’s last few years Dorris wrote him lengthy epistles every Friday al- most without fail. Not to be overlooked are the number of letters and notes starting in 1941 that went between Jepson and his latter-day unofficial “*secretarial assistant,’’ geneticist Helen-Mar Whee- ler, whose father and Jepson had been on campus together as students. Helen-Mar would eventually become executrix of Jepson’s estate. With respect to Jepson’s five dozen field books, he began going back over them, enhancing many of the entries from his remarkable though on oc- casion faulty memory, and he had compiled 20 ad- ditional volumes which brought together his notes on California botanical collectors (four volumes), observations on systematic botany, plant common names, field records, addresses of botanists and oth- MADRONO er individuals, three indices, etc. These, too, now | make up part of the Jepson Herbarium Library col- | lection. He began penning a series of reminiscences on / early Vacaville days for the Vacaville newspaper | and continued to write articles on California’s bot- | anists, his last one on his old friend and lily enthu- | siast Carl Purdy being published after Jepson’s | death. But the book that he had long planned, The | History of Botanical Exploration in California, would never be completed, nor the work on Cali- | fornia plant geography, nor his booklet on plant | common names, nor, fortunately, his many-chapter- | ed, often venomous diatribe, “‘Man and Manners,”’ | which mixed “‘many delightful things in the manner | of John Adams” with what Jepson caustically termed the “‘gangsterism’’ in academia. that he and graduate students could make more suc- [Vol. 47 | Other | things, however, were being taken care of. As Jep- | son over the years had built up his herbarium, he | had generally not put the plants on sheets, feeling © cessful use of them if they were loose. Now at last | many of the specimens were being mounted. Jepson began reflecting on his botanical collect- | ing experiences throughout California and contem- | plated his favorite habitats. The White Fir belt of the Sierra Nevada had ‘‘a strong appeal,’’ and the | alpine slopes had a “‘strong grip’? upon him. But | better was the “‘magnificent’”? redwood country of | the northwest. The deserts had “‘irresistible fasci- | nation” for Jepson and he came “back to them | 29 again and again, season. But he had ‘‘no native affection for the des- always regretting missing a fine | ert ranges because of their ruthless implacable de- | fiance.”’ The Great Valley plains were “‘magnifi- cent’? but their immensity was overpowering. As | he told fellow botanist Carl Wolf, “‘I like best the | Coast Range valleys and their bounding ridges © where still primitive or nearly so,” County south to Parkfield, as well as the San Diego from Lake } backcountry. The Vacaville foothills and St. Helena | with nearby Mt. Howell had a special affection. At | the bottom of his list was the ocean shore line, “‘the utter cruelty and hopelessness of the sea”’ depress- ing his spirit both with respect to collecting and | traveling across the ocean aboard ship. Inevitably, Jepson had early developed a percep- | tive ecological, as well as a biogeographical sense, resulting from his extensive field work in California and abroad, broadening experiences which many | taxonomists of the day seemed to lack. As early as © 1902 he contemplated taking up “a treatise on the | 29 plant geography of the state,’ and he actually did delineate major areas and even local plant com- munities such as those of spring (“‘vernal’’) pools. In his writing he frequently included ecological — considerations. Collecting during July atop Mt. Ly- | ell in Yosemite he noted that more plants were blooming on the summit than below on the north- facing slope. He made a point of saying that from an ecological standpoint the diversified flora of 2000] - Moraga Ridge, over the hills beyond Berkeley, was his favorite example. Somewhat ahead of his time, a dozen small-type pages in the first section of Jepson’s 1925 Manual dealt with an ‘Outline of Geographic Distribution of Seed Plants in California,’ discussing floral dis- tribution related to Merriam’s life zones, with a sec- tion devoted to “Irregularities in the Life-Zones,”’ a consideration of plant distribution and geologic history, and several pages about ““The Endemic Populations”’ with a map showing some endemism areas in California. And although Jepson had an antipathy towards introduced species, he included a section on “The Alien Populations.”’ For com- parison, in Abram’s contemporary first Flora vol- ume only three pages discussed such topics. Jepson rightly chided American ecologists for paying too little attention to correct identification of plants, and he viewed with interest mixed with skepticism the increasing botanical interest in cy- togenetics on the part of systematists, but rational- ized that with the demands of completing the Flora he could not, despite his mild interest, afford the time at his age to become accomplished in this emerging botanical field. But Jepson did anticipate that eventually genetics would get around to con- sider all organisms, ‘‘and then we would really know how many there were.’ Professor Robert Ornduff, looking back on Jepson’s accomplish- ments 40 years after his death, concluded that *‘I continue to marvel at Jepson’s insights into matters just now being explored by botanists.” As mentioned earlier, Jepson had begun number- ing his plant collections definitively in the year 1899. His professional career’s last numbered spec- imen, No. 27,571, in his final field book, was, of all things, Salsola kali L. var. tenuifolia Tausch (Salsola tragus L.), collected at the Antioch Sand- hills. The herbarium sheet is in the Jepson Herbar- ium, incorrectly dated November 11, 1945. In ac- tuality, Helen-Mar Wheeler had driven Jepson to the sandhills on October 28. Later Rimo Bacigalupi had added the following note to the sheet: ‘“This is the last specimen collected by Dr. W. L. Jepson.” Yet even for this ignominious exotic weed Jepson not only penned a detailed description, as had often been his wont through the years for collected spec- imens, but added a sketch of the pistil and stamens. Jepson was in his 78th year. In mid-April of 1945 he had suffered a heart attack, overstrained by cut- ting down a dead almond tree at his Little Oak Ranch. He would never completely recover, spend- ing much of his time in the sanitarium at St. Helena and later in a hospital in Alameda. Finally, on No- vember 7, 1946, at his beloved home on Mosswood in Berkeley, Willis Linn Jepson passed away peace- fully. Dr. Jepson, that son of Vacaville pioneers, would express in his darker moments the fear that ‘‘one is always without honor in his own country.” Yet Pro- fessor Jepson has been memorialized for the Jepson BEIDLEMAN: JEPSON—‘**THE BOTANY MAN” 285 Herbarium, which now with more than 90,000 specimens celebrated its 50th anniversary in 2000, and for his diversity of publications, about 230 ranging from major books and scientific treatises to a three-page popular article in Sunset Magazine on ‘“Where Ducks Dine,” and indirectly including the new Jepson Manual. To Lincoln Constance, on one occasion, Jepson well expressed the obligatory re- lationship between plant collections and publishing: ‘It matters not how much knowledge may be ac- cumulated about a given species, how many mono- graphs discuss it—always botanists wish to go back to the plant, to authentic specimens. A flora which cites no specimens whatsoever may be a useful flo- ra but it is not a scientific flora.”’ David Keck in his 1948 obituary for Jepson in Madrono tabulated who had named the largest number of California plant species by that time, and Jepson ranked number nine. Only one other erst- while Californian outranked him, his mentor Ed- ward Lee Greene, who had a reputation for gener- ating scientific names. Speaking of appropriate names, there is the saxifrage genus Jepsonia, des- ignated by John K. Small of the New York Botan- ical Garden. And the scientific names of innumer- able species and subspecies of California plants honor Jepson, as well as those having been coined by him. Finally, although Jepson claimed he dis- liked the ‘“‘folk names” for plants, he was always quick to point out the ones he had popularized. In fact, he once proudly claimed that “‘I have invented more common names of native plants for nonbo- tanists than any one else in the New World.” Among his favorites were undoubtedly Mountain Misery (hike through it for miles and you’d appre- ciate why Jepson, as he emphasized, called it that), Johnny-tuck, and Red Maids, or “*Kisses”’ as they were known in Sonoma County. Jepson named Johnny-tuck after an elderly hired man on the Little Oak Farm, whom he remembered as a child stand- ing in a field of those flowers in his Sunday-going- to-meeting finery. When Jepson queried a little farm girl why she called Red Maids ‘‘Kisses,”’ she shyly replied ““You don’t always know why, you just do!”’ Jepson, by the way, was quick to chide authors who published his common names without crediting the source, yet he didn’t seem concerned that the country folk who gave him many of his names seldom received any specific recognition from him. At the western edge of his boyhood town there now stands the Willis Jepson Middle School, ded- icated on May 23, 1960, with appropriate floral plantings and colorful mural, bordered by streets named Jepson Way and Jepson Court. At Tomales Bay State Park there is the Jepson Trail, leading to the Willis Linn Jepson Memorial Grove of Bishop Pines, dedicated on November 8, 1952. There is a Jepson preserve on the Klamath River in northern California, one of his collecting haunts, and the Jepson Prairie in the Suisun Marsh country, another 286 Jepson botanizing locale. In 1902 it was Jepson who discovered the world’s largest Coast Madrone (now recently fallen) in Humboldt County and named it the Council Madrone, after coast and in- terior Native American tribes that used to parlay there. Two years later, Jepson planted a redwood beside the Vacaville cottage into which his mother moved from Little Oak Ranch after her husband’s death, and that redwood now towers over the cot- tage in this new century. Jepson would serve on the Council of Save-the-Redwoods League from its establishment until his death. He was also the hon- orary vice-president of the Sierra Club the last five years of his life. On April 15, 1923, members of the California Botanical Society gathered near Lower Crystal Valley Reservoir south of Millbrae to dedicate California’s second largest Bay Tree as the Jepson Laurel. In 1972, a quarter of a century after Willis Jep- son’s death, the United States Geological Survey designated a jagged 13,390-foot wilderness peak as Mt. Jepson, in King’s Canyon National Park near Mt. Whitney and close to Willis Jepson’s old Sierra Club hiking route. In silhouette Mt. Jepson, rising MADRONO [Vol. 47. shaped glacial gorge, remarkably resembles the dis- | tant skyline peak portrayed in Jepson’s book plate, | with its inscription “‘Something lost behind the | ranges—over yonder—Go you there.”’ This appro- | priate alpine monument in the high Sierra would indeed have pleased *“‘The Botany Man.”’ LITERATURE CITED This biographical treatise on Willis Linn Jepson is based upon an extensive variety of original pri- mary and a limited number of secondary sources, the most important of the former being the volumes dramatically far above timberline beyond a U- | | i of Jepson Correspondence, Jepson Field Books, and — the Jepson/Helen-Mar Wheeler Collection, all in © the University of California Jepson Herbarium Ar- chives, with other valuable sources including the — various published obituaries of Jepson, and the oral — interviews with Laura May Dempster, Joseph | Ewan, and Lincoln Constance in the Jepson Ar- — chives. Minor points of clarification were derived | from a number of published books and articles, but | most quotations are from original documents at the | Jepson Herbarium Archives. | | -Maprono, Vol. 47, No. 4, p. 287, 2000 PRESIDENT’ S REPORT FOR VOLUME 47 As the California Botanical Society’s program year draws to a close, I have the pleasure of reviewing this year’s accomplishments and activities. I also take this op- portunity to thank the officers and council members for their dedicated efforts to keep the Society a vital force in west American botany. On behalf of the Society, I offer special thanks to Editor Kristina Schierenbeck for her con- tinuing progress toward bringing Madrono back onto pub- lication schedule and elevating further the stature of the journal. I am delighted to announce that Kristina has very generously agreed to extend her invaluable service as Ed- itor of Madrono for a fourth year, through Volume 48. We are now searching for a new Editor of Madrono, for Vol- umes 49-51, and I encourage an aspiring editor to contact me or Kristina soon so that we can begin planning for a smooth transition. Incoming First Vice-President Rodney Myatt helped the Society off to a strong start in fall 2000 by organizing an outstanding slate of speakers for our monthly meetings at U.C. Berkeley. We heard excellent presentations by John Battles, Chris Brinegar, Tina Carlsen, Susan Harrison, Dan Norris, Dan Potter, and Maureen Stanton on a wide range of botanical subjects, each followed by a post-seminar re- ception in the University and Jepson Herbaria. Many thanks to Rodney for his excellent planning and to our graduate student representative, Kirsten Johannes, for faithfully coordinating the invitations and receptions for each of our monthly meetings during the past year. On 17 Feb 2001, the Society held one of its most im- portant functions, the biennial graduate student scientific meeting, in conjunction with the annual banquet. The two events were held this year on the campus of the California State University at Chico. The graduate student meeting, organized by Chico graduate student representative Leah Mahan and CBS graduate student representative Kirsten Johannes, was by all accounts a great success, and both of these hard-working students deserve thanks and con- gratulations. Scientific papers on proposed, ongoing, and completed botanical research were presented throughout the morning and afternoon by 35 student speakers from within and outside California. Graduate student research is responsible for much of what we know about west American botany and I highly recommend regularly at- tending this highly educational event. Second Vice-President Rob Schlising did an excellent job of hosting the Society’s well-attended 2000-2001 an- nual banquet. Past President Wayne Ferren graciously stepped in to officiate over the evening’s events while I was fighting off a flu in Berkeley (thank you, Rob and Wayne!). Attendees of this year’s annual banquet were treated to an after-dinner lecture by Professor Emeritus Arthur Kruckeberg (University of Washington), the pre- mier authority on serpentine endemism in plants, on the role of geology in molding the California flora. On behalf of the Society, I thank Dr. Kruckeberg for making this year’s banquet such a special and memorable occasion. A priority of the Council during the past year has been to elevate visibility and membership of the Society and, especially, to increase circulation of Madrono and citation of articles therein. Inclusion of Madrono in BIOSIS has been a positive step toward increasing article citation and we are actively pursuing incorporation of Madrono into other on-line databases. Thanks to incoming Society web- masters Curtis Clark and John LaDuke, the California Bo- tanical Society now has an active and attractive web-site (www.calbotsoc.org), which will soon include abstracts of Madrono articles (beginning with Volume 48) and should help to alert more botanists to the value of Madrono and membership in CBS. Recording Secretary Dean Kelch spearheaded a membership drive during the past year that resulted in wide dissemination of information about the Society (and Madrofo in particular) to academic and agency plant scientists throughout the West. I thank Dean and the other members of the Council for their important help in the membership drive. I ask all members of the Society to help promote our membership by encouraging your non-member collegues to join us. Two hard-working members of the Council that deserve special thanks for their efforts to ensure continuity of the Society’s membership and general maintenance of the So- clety’s finances are incoming Treasurer Roy Buck and Corresponding Secretary Susan Bainbridge. Thanks in part to Roy’s and Sue’s conscientiousness, the Society’s membership base is growing. I also thank Council mem- bers Diane Elam, Jim Shevock, and Bian Tan for their thoughtful contributions toward helping to chart the future of the Society and Recording Secretary Dean Kelch for reliably keeping minutes of our monthly Council meet- ings. Of course, none of our activities would be possible without the contributions of the members of the California Botanical Society. On behalf of the Council, I thank each of you for your critical support of the Society’s important goals and look forward to your participation in our new program year! —Bruce G. Baldwin. July 2001. MADRONO, Vol. 47, No. 4, p. 288, 2000 EDITOR’S REPORT FOR VOLUME 47 This report serves to inform members of the California Botanical Society the status of Madrono from manuscripts submitted to papers published. Since the previous editor’s report (see Madronrio 46[4]) the journal received 73 manu- scripts for review, including Articles, Notes, and Noteworthy Collections; 64 of these have since been accepted for pub- lication. The average time from article submission to publi- cation has been remained stable at approximately six months. Very few manuscripts were rejected after review. Authors of Madrono articles are generally quite responsive to reviewer and editorial suggestions. The publication schedule has been returning, albeit slow- ly, to a regular publication schedule. We are currently with- in six months of returning the journal to an on-time sched- ule, as publication submissions have increased in frequency. Noteworthy collections continue to be a valuable contri- bution to the journal but have suffered from reduced atten- tion due to editorial efforts with other manuscripts. There are many individuals who contribute to the edi- | torial process; Jon Keeley, who continues to serve as book | review editor; Steve Timbrook, who continues to assemble | the Index and Table of Contents; Dieter Wilken and Mar- | griet Wetherwax, who edit the Noteworthy Collections; | David Parks and Jeannie Trizzino, my editorial assistants; | Michael Abruzzo, Chair of the Department of Biological © Sciences at California State University Chico, who pro- | vides the funds to support David; Karen Ridgway at Allen ' Press; and members of the CBS executive council who | enthusiastically support Madrono in every aspect. I con- tinue to rely on the council of Robert Patterson, Wayne | Ferren, Jon Keeley, John Strother, and Beth Painter for guidance about the editorial process. On behalf of the so- | ciety, I thank the volunteer reviewers and the Board of | Editors on whom we all depend to make the peer review | process work for this valuable regional journal. REVIEWERS OF MADRONO MANUSCRIPTS 2000 James Agee Geraldine Allen Ihsan Al-Shehbaz Michael Barbour Theodore Barkley Matthew Beaty Bruce A. Bohm Lynn Bohs Kenneth Cameron Curtis Clark Kenneth Cole J. Travis Columbus Steve Darbyshire Frank Davis Richard S. Dodd Aron Fazekas Wayne Ferren Bruce Ford John Freudenstein Candice Galen Barbara Gartner Carol Goodwillie Craig W. Greene Howard Griffiths E Thomas Griggs James Hamrick Gary Hannan Klaus Helenurm Noel Holmgren Sara Hoot David C. Jarrell Leigh Johnson Steve Junak Jon Keeley David Keil Elizabeth Kellogg Job Kuyt John LaDuke David Lemke Yan Linhart Aaron Liston Timothy Lowry Scott Martens Robert Mathiasen Steve McLaughlin John McNeill Kathy Merrifield Richard Minnich Brent Mishler James Morefield Guy Nesom Steven O’ Kane V. Thomas Parker Robert Patterson Mark Porter James Pushnik James Reveal Ronald Robberecht Gregory Saenz John O. Sawyer Rob Schlising Joanna Schultz James Shevock Scott Stephens Nathan Stephensen Michael Stieber Robert Stockhouse John Strother Stanley Welch Dieter Wilken Paul Wilson David Wood Frederick Zechman -“Maprono, Vol. 47, No. 4, pp. 289-291, 2000 INDEX TO VOLUME 47 Classified entries: major subjects, key words, and results; botanical names (new names are in boldface); geographical | areas; reviews, commentaries. Incidental references to taxa (including most lists and tables) are not indexed separately. Species appearing in Noteworthy Collections are indexed under name, family, and state or country. Authors and titles are listed alphabetically by author in the Table of Contents to the volume. | Abies magnifica, mixed conifer and red fir forest in Sierra Nevada, CA, 1899 structure and uses, 43. _ Aliciella sedifolia, noteworthy collection from CO, 142. Alliaceae (see Triteleia) Ambrosia pumila, noteworthy collection from CA, 139. Apiaceae (see Cympoteris and Eryngium) Arabis pinzliae, noteworthy collection from CA, 209. Araceae (see Zantedeschia) Arizona: Tsegi Canyon floristics, 29. Noteworthy collections: Hexalectris revoluta, 138; Vauquelinia californica subsp. sonorensis, 211. Artemisia michauxiana, A. ludoviciana subsp. candicans, noteworthy collections from CA, 209. Asclepiadaceae (see Asclepias and Cynanchum) Asclepias involucrata, noteworthy collection from CO, 143. Asteraceae: Corethrogyne, revision of genus, 89; Lasthen- ia glabrata subsp. coulteri, effect of climatic variability on growth, reproduction and population viability, 174. New taxa: Corethrogyne filaginifolia var. californica, 89; Hedosyne, new genus for Iva, 204. Noteworthy collections: CA: Ambrosia pumila, 139; Artemisia michauxiana, A. ludoviciana subsp. can- dicans, 209; Brickellia knappiana, 141; B. multiflora, 142; Dicoria canescens, 140; Helenium microce- phalum, Heterosperma pinnatum, 143; Senecio aphanactis, 138. Astrolepis integerrima, noteworthy collection from CO, 143. Berberidaceae (see Berberis) Berberis darwinii, noteworthy collection from WA, 214. Boehmeria cylindrica, noteworthy collection from CA, 138. Boraginaceae (see Plagiobothrys) Bothriochloa_ springfieldii, noteworthy collection from CO, 143. Brassicaceae: Draba chromosome counts and taxonomic notes, 21; Streptanthus nickel tolerance and hyperac- cumulation, 97. Noteworthy collection: Arabis pinzliae trom CA, 209. Brickellia knappiana, 141, B. multiflora, 142, noteworthy collections from CA. Cactaceae (see Opuntia) California: Annual xylem water potential variation in oaks, 106; chaparral seed banks, composition, germi- nation and fire survival from long-unburned stands, 195; Delphinium variegatum floral variation, 116; Jep- son Herbarium 50th anniversary scientific symposium proceedings, 217—286; Lasthenia glabrata subsp. coul- teri, effect of climatic variability on growth, reproduc- tion and population viability, 174; mixed conifer and red fir forest in Sierra Nevada, 1899 structure and uses, 43; coastal redwood old-growth forest associations, 53; Sequoiadendron giganteum, ages, 61; crown structure, 127; yellow pine forest: genetic variation in Festuca idahoensis, Pinus ponderosa and Purshia tridentata, 164. New taxon: Eriogonum spectabile, 134. Noteworthy collections: Ambrosia pumila, 139; Arabis pinzliae, Artemisia michauxiana, A. ludoviciana subsp. candicans, 209; Boehmeria cylindrica, 138; Brickellia knappiana, 141; B. multiflora, 142; Cam- issonia pterosperma, Cornus glabrata, Cynanchum utahense, Glycera occidentalis, 142; Dicoria canes- cens, Erodium malacoides, 140; Eryngium constan- cel, 139; Eschscholzia rhombipetala, 138; Koeleria phleoides, 140; Liquidambar styraciflua, 209; Mon- ardella pringlei, Nama stenocarpum, 140; Nicotiana acuminata var. multiflora,, 142; Ononis alopecuro- ides, 139; Quercus palmeri, 141; Sapium sebiferum, 210; Senecio aphanactis, 138; Sesbania punicea, 210: Calluna vulgaris, noteworthy collection from WA, 214. Camissonia pterosperma, noteworthy collection from CA, 142. Caprifoliaceae (see Lonicera) Carex v. subsp. v. var. viridula, Eurasian origin, 147. Noteworthy collections: C. chordorrhiza, 144, C. lon- gii, 213, 214. Chaparral seed banks, composition, germination and fire survival from long-unburned stands, 195. Chenopodiaceae (see Chenopodium) Chenopodium cycloides, noteworthy collection from CO, 143. Chromosome counts: Corethrogyne filaginifolia, 91; Dra- ba 21. Clusiaceae (see Hypericum and Triadenum) Colorado: Noteworthy collections: Aliciella sedifolia, As- clepias involucrata, Astrolepis integerrima, Bothrioch- loa springfieldii, Chenopodium cycloides, Diplachne dubia, Eleocharis xyridiformis, Festuca subulata, He- lenium microcephalum, Heterosperma pinnatum, Rev- erchonia arenaria, Triteleia grandiflora, 142-144. Compositae (see Asteraceae) Corethrogyne: C. filaginifolia var. californica, new com- bination and revision of genus, 89. Cornaceae (see Cornus) Cornus glabrata, noteworthy collection from CA, 142. Cotoneaster: Noteworthy collections: C. dielsianus, C. franchetii, 214; C. horizontalis, 213; C. lacteus, 213, 214; C. rehderi, 214; C. simonsii, 213, 214. Cruciferae (see Brassicaceae) Cupressaceae (see Juniperus) Cymopteris beckii, new report for AZ, 29. Cynanchum utahense, noteworthy collection from CA, 142. Cyperaceae (see Carex and Eleocharis) Delphinium variegatum floral variation, 116. Dicoria canescens, noteworthy collection from CA, 140. Diplachne dubia, noteworthy collection from CO, 143. Draba chromosome counts and taxonomic notes, 21. 290 Eleocharis: Noteworthy collections: E. quadrangulata from OR, 213; E. xyridiformis from CO, 143. Eragrostis spicata, noteworthy collection from Sonora, Mexico, 211. Ericaceae (see Calluna and Vaccinium) Eriogonum spectabile, new sp. from northeastern CA, 134. Erodium malacoides, noteworthy collection from CA, 140. Eryngium constancei, 139, Escallonia rubra, noteworthy collection from OR, 144. Eschscholzia rhombipetala, noteworthy collection from CA, 138. Euphorbiaceae (see Reverchonia and Sapium) noteworthy collection from CA, Fabaceae (see Ononis and Sesbania) Fagaceae (see Quercus) Festuca: F. idahoensis, genetic variation in northwestern CA yellow pine forest, 164; F. subulata, noteworthy collection from CO, 143. Fire survival of seeds from long-unburned maritime chap- arral stands, 195. Floristic studies: Tsegi Canyon, AZ, 29. Fuchsia magellanica, noteworthy collection from OR, 144. Geraniaceae (see Erodium) Glyceria: Noteworthy collections: G. Glycera occidentalis, 142. Gramineae (see Poaceae) Grossulariaceae (see Escallonia) canadensis, 214; Haloragaceae (see Myriophyllum) Hamamelidaceae (see Liguidambar) Hedosyne, new genus for Iva, 204. Helenium microcephalum, noteworthy collection from CoO, 143. Helleborus foetidus, noteworthy collection from WA, 214. Heterosperma pinnatum, noteworthy collection from CO, 143. Hexalectris revoluta, noteworthy collection from AZ, 138. Hydrophyllaceae (see Nama) Hypericaceae (see Hypericum) Hypericum: Noteworthy collections: H. boreale, 144; H. canadense, H. ellipticum, 215; H. majus, 213; H. mu- tilum, 215. Iva (see Hedosyne) Jepson Herbarium 50th anniversary scientific symposium proceedings, 217-286. Jepson, Willis Linn (see Jepson Herbarium) Juncaceae (see Juncus and Luzula) Juncus: Noteworthy collections: J. brevicaudatus, 144; J. canadensis, 144, 215; J. diffusissimus, 215; J. pelocar- pus, 144, 215. Juniperus californica Pleistocene macrofossils from Ne- otoma (wood rat) middens in the northern Vizcaino Desert, Baja California del Norte, México, 189. Koeleria phleoides, noteworthy collection from CA, 140. Labiatae (see Lamiaceae) Lamiaceae (see Monardella and Salvia) Lasthenia glabrata subsp. coulteri, effect of climatic vari- ability on growth, reproduction and population viability, 174. MADRONO Leguminosae (see Fabaceae) | Liquidambar styraciflua, noteworthy collection from CA,! 209. | Lonicera pileata, noteworthy collection from WA, 215. Luzula arcuata subsp. unalaschensis, L. forsteri, note- worthy collections from OR, 213. México: Locations of Picea chihuahuana, P. martinezii\ and P. mexicana, 71; Neotoma (wood rat) midden Pleis- tocene macrofossils of Juniperus californica and Pinus quadrifolia in the northern Vizcaino Desert, Baja Cal- | ifornia del Norte, 189. Noteworthy collections: Eragrostis spicata, Portulaca johnstonii, Salvia similis, Vauquelinia californica , subsp. sonorensis, Ximenia parviflora var. glauca, ° PALE Monardella pringlei, noteworthy collection from CA, 140. Mount St. Helens, WA, seed rain during early primary | succession, |. Myriophyllum ussuriense, noteworthy collections from | OR and WA, 212. Nama stenocarpum, noteworthy collection from CA, 140. | Neotoma (wood rat) midden Pleistocene macrofossils in | the northern Vizcaino Desert, Baja California del Norte, México, 189. Nevada: Plagiobothrys glomeratus, taxonomic history, identity and distribution, 159. Nickel hyperaccumulation (see Streptanthus) Nicotiana acuminata var. multiflora, noteworthy collec- | tion from CA, 142. | Olacaceae (see Ximenia) Onagraceae (see Camissonia and Fuchsia) Ononis alopecuroides, noteworthy collection from CA, 139. Opuntia prolifera, molecular evidence for hybrid origin, 109. Orchidaceae (see Hexalectris). Oregon: Coastal redwood old-growth forest associations, 53; Noteworthy collections: Carex chordorrhiza, 144; C. longii, Cotoneaster horizontalis, C. lacteus, C. si- monsii, Eleocharis quadrangulata, 213; Escallonia rubra, Fuchsia magellanica, Hypericum boreale, 144; H. majus, 213; Juncus brevicaudatus, J. cana- densis, J. pelocarpus, 144; Luzula arcuata subsp. un- alaschensis, L. forsteri, 213; Myriophyllum ussur- iense, 212; Polygonum sagittatum, 213; Spiraea to- mentosa, 144; Zantedeschia aethiopica, 213. Papaveraceae (see Eschscholzia) Picea chihuahuana, P. martinezii and P. mexicana, loca- tions in México, 71. Pinaceae: Abies magnifica, mixed conifer and red fir forest in Sierra Nevada, CA, 1899 structure and uses, 43; Pic- ea chihuahuana, P. martinezii and P. mexicana, \oca- tions in México, 71; Pinus ponderosa, genetic variation in northwestern CA yellow pine forest, 164; P. quadri- folia Pleistocene macrofossils from Neotoma (wood rat) middens in the northern Vizcaino Desert, Baja Califor- nia del Norte, México, 189. Pinus: P. ponderosa, genetic variation in northwestern CA yellow pine forest, 164; P. quadrifolia Pleistocene ma- crofossils from a Neotoma (wood rat) midden in the northern Vizcaino Desert, Baja California del Norte, México, 189. 2000] | Plagiobothrys glomeratus, taxonomic history, identity and | distribution, 159. Plant communities: Coastal redwood old-growth forest as- - sociations, 53; environmental gradients and vegetation structure on south Texas coastal clay dunes, 10; mixed conifer and red fir forest in Sierra Nevada, CA, 1899 structure and uses, 43. Poaceae: Festuca idahoensis, genetic variation in north- western CA yellow pine forest, 164. Noteworthy collections: Bothriochloa springfieldii, Di- plachne dubia from CO, 143; Eragrostis spicata from Sonora, Mexico, 211; Festuca subulata from CO, 143; Glycera occidentalis, 142, Koeleria phleo- ides from CA, 140. Polemoniaceae (see Aliciella) Polygonaceae (see Eriogonum and Polygonum) Polygonum sagittatum, noteworthy collection from OR, 23: Portulaca johnstonii, noteworthy collection from Sonora, Mexico, 211. Portulacaceae (see Portulaca) Pteridiaceae (see Astrolepis) Purshia tridentata, genetic variation in northwestern CA yellow pine forest, 164. Quercus: Annual xylem water potential variation in CA oaks, 106; Q. palmeri, noteworthy collection from CA, 141. Ranunculaceae (see Delphinium) Reverchonia arenaria, noteworthy collection from CO, 143. Reviews: A Natural History of the Sonoran Desert ed. by S. J. Phillips and P. W. Comus, 68; 2nd Interface Be- tween Ecology and Land Development ed. by J. E. Kee- ley, M. Baer-Keeley and C. J. Fotheringham, 206. Rosaceae: Purshia tridentata, genetic variation in north- western CA yellow pine forest, 164. Noteworthy collections: Cotoneaster horizontalis, C. lacteus, C. simonsii from OR, 213; Spiraea tomen- tosa from OR, 144; Vauquelinia from AZ and So- nora, Mexico, 211. Ranunculaceae (see Helleborus) Salicaceae (see Salix) Salix purpurea, noteworthy collection from WA, 216. Salvia similis, noteworthy collection from Sonora, Mexi- co, 211; San Clemente Island, CA (see Delphinium) INDEX TO VOLUME 47 Zo) Sapium sebiferum, noteworthy collection from CA, 210. Senecio aphanactis, noteworthy collection from CA, 138. Sequoia sempervirens, old-growth forest associations, 53. Sequoiadendron giganteum, ages, 61; crown structure, 1277. Serpentine soils (see Streptanthus) Sesbania punicea, noteworthy collection from CA, 210. Solanaceae (see Nicotiana) Spiraea tomentosa, noteworthy collection from OR, 144. Streptanthus nickel tolerance and hyperaccumulation, 97. Succession: Seed rain on Mount St. Helens WA, I. Taxodiaceae (see Sequoia and Sequoiadendron) Texas: environmental gradients and vegetation structure on coastal clay dunes, 10. Triadenum fraseri, noteworthy collection from WA, 216. Triteleia grandiflora, noteworthy collection from CO, 144. Umbelliferae (see Apiaceae) Urticaceae (see Boehmeria) Vaccinium corymbosum, V. macrocarpon, noteworthy col- lections from WA, 216. Vauquelinia californica subsp. sonorensis, noteworthy collections from AZ and Sonora, Mexico, 211. Washington: Mount St. Helens, seed rain during early pri- mary succession, 1. Noteworthy collections: Berberis darwinii, Calluna vul- garis, Carex longii, Cotoneaster dielsianus, C. fran- chetii, C. lacteus, C. rehderi, C. simonsii, Glyceria canadensis, Helleborus foetidus, Hypericum boreale, H. canadense, H. ellipticum, H. mutilum, Juncus can- adensis, J. diffusissimus, J. pelocarpus, Lonicera pt- leata, 215; Myriophyllum ussuriense, 212; Salix pur- purea, Triadenum fraseri, Vaccinium corymbosum, V. macrocarpon, 216. Ximenia parviflora var. glauca, noteworthy collection from Sonora, Mexico, 211. Xylem water potential (see Quercus) Yellow pine forest: Genetic variation in Festuca idahoen- sis, Pinus ponderosa and Purshia tridentata in north- western CA, 164. Zantedeschia aethiopica, noteworthy collection from OR, 213. MApRONO, Vol. 47, No. 4, pp. 292-300, 2000 ROBERT ORNDUFF 1932—2000 Dr. Robert Ornduff, Professor Emeritus at the University of California, Berkeley, died on Septem- ber 22, 2000, at the age of sixty-eight, in Berkeley from complications of metastatic melanoma. ““Bob was one of the treasures of the botanical world, a green-thumb botanist who delighted in growing plants and disseminating his interest to the general public. He was a ‘rara avis’ in botany these days, who could operate brilliantly in both natural history and in ‘ivory tower’ plant biosystematics,”’ accord- ing to Art Kruckeberg, professor emeritus at the University of Washington, Seattle, Bob’s mentor and friend. Bob’s distinguished career spanned some thirty-seven years at the university. He leaves an impressive legacy of research in plant evolu- tionary biology, of mentoring highly distinguished students, of major contributions to the development of the University Botanical Garden, and of a life- time of effective participation in the botanic com- munity outside the university. Born in Portland, Oregon, on June 13, 1932, Ornduff grew up in the suburbs near a golf course, Where he collected and took home to raise numer- ous found creatures such as giant Pacific salaman- ders, fish, snails, and baby birds, a habit he retained throughout his life. Following graduation from Washington High School in Portland, he attended nearby Reed College and graduated in 1953 with a major in biology. Bob next obtained a Master’s de- gree at the University of Washington (1956) study- ing under the direction of Art Kruckeberg. He had met Art four years earlier, when Art was an instruc- tor in a nine-week University of Washington sum- mer field course, and the two developed a lifelong friendship. Bob claimed that, while he knew from childhood that he wanted to be a biologist, it was Art who drew him into a career in botany. During his studies with Art, he received a Fulbright schol- arship and spent a year in New Zealand, looking at the distribution and systematics of the puzzling and polymorphic New Zealand Senecio lautus complex. Kruckeberg commented that “His study of the group was certainly Ph.D. worthy in quality and extent. It was later published in a New Zealand journal.” In 1955, Bob moved to the University of Cali- fornia, Berkeley, commencing studies for his Ph.D. degree. One of us (PR) met him there as a Berkeley undergraduate, and commenced a lifetime friend- ship. Like many other Berkeley students, Bob was greatly influenced by Herbert Mason’s course in plant systematics, an outstanding inquiry-oriented course in which Don Stone was the teaching assis- tant and Job Kuijt, Galen Smith, Howard Arnott,' Jean Langenheim, and Peter Raven were among the, fellow students. Evidence of all kinds for arriving! at a proper understanding of plant classification and evolution was considered in discussions. Field trips, hours of discussion and analysis, and good com-! panionship made the course a memorable learning experience for all who were associated with it. Bob’ commenced his graduate studies under the direction | of Professor Mason, taking a year out to fill in at. Reed College, and obtained his degree in 1961. | During Mason’s course, one of us (PR) suggested | that the group of composites known then as Baeria | might be an interesting subject for Bob’s disserta- | tion research. Earlier, K. L. Chambers had found | different base chromosome numbers within the ge- | nus, and it seemed likely to be a suitable subject for biosystematic inquiry. Bob’s thorough investi- gation of the genetics, morphology, ecology, and _ evolutionary relationships resulted in the realign- | ment of generic limits and the incorporation of Baeria into the genus Lasthenia, with many inter- | esting relationships revealed in the course of the | study. The taxonomic treatment resulting from his | dissertation (1961) was published by the University | Press as a monograph, A Biosystematic Survey of | the Goldfield Genus Lasthenia, in 1966. After ap- | plying various approaches to the further elaboration | of evolutionary relationships within Lasthenia over | the years, Bob took particular delight in the work of Raymund Chan, his last graduate student, who | is using molecular systematic to deepen our under- | standing of the genus Lasthenia. These studies have also resulted in the recognition of a new species of Lasthenia, which Raymund intends “‘to name in memory of Dr. Ornduff.”’ Bob Ornduff’s first academic position after his graduation was at Duke University (1961-1963), but when invited to fill Mason’s chair at U.C. Berkeley found the offer to be irresistible. At his alma mater he served with distinction in a number of capacities: director of the Jepson and University Herbaria from 1967 to 1982, director of the Uni- versity Botanical Garden from 1973 to 1991, and chairman of the Botany Department from 1986 to 1989, when it merged into the Department of In- tegrative Biology. He served as chairman of the Editorial Committee at the University of California Press from 1975 to 1989, during the years when August Frugé was director and then for a requested holdover year when Jim Clark became director. Bob was executive director of the Miller Institute for Basic Research in Science from 1984 to 1987, ROBERT ORNDUFF— 1932-2000 293 Fic. 1. a program that awards grants both to visiting in- vestigators and students and to campus faculty for work off-campus. At the time of his death, Ornduff was actively involved with Jepson Herbarium, the University Botanical Garden, and his own research on the population dynamics of the genus Villarsia (Menyanthaceae). He also was co-editor of the Nat- ural History Series at the University of California Press with one of us (PF). Ornduff wrote more than 100 scientific papers and over 50 less formal papers on horticultural and related topics. He edited two books and wrote one, the popular Introduction to California Plant Life (U.C. Press 1974), which is still in print and has introduced generations of students to California’s unique flora. Following his retirement in 1993, he continued teaching his popular California Plant Life course through the U.C. Extension program to many hundreds of students. Ornduff’s research interests were broad but in general focused on the evolution of species and Species diversity. According to Steve Weller (now a professor in the Department of Ecology and Evo- lutionary Biology at U.C. Irvine), “‘Bob was one of the first bridges from biosystematics and evolu- tionary biology to plant population biology and was an inspiration to many.’’ A seminal event in Bob’s research career was his discovery of the unique fall-blooming saxifragaceous genus Jepsonia while on a field trip in 1956. He soon realized that Jep- Dr. Robert Ornduff botanizing at New Anguin, May 1989. Photo by Ken Wilson. sonia was heterostylous, where some individuals bore flowers that had long styles and short stamens (pin), and others had the opposite condition (thrum). Jepsonia comprises three species, confined to small areas of California, adjacent Mexico, and the offshore islands. In 1961, Ornduff published his first paper on the genus, reporting heterostyly in it and in the family Saxifragaceae for the first time. Thus he began a lifelong fascination with hetero- styly and its relationship with dioecism, investigat- ing the phenomenon in several other groups includ- ing Oxalis (Oxalidaceae), Pontederia (Pontederi- aceae), Lythrum (Lythraceae), Hypericum (Guttif- erae), Leucocrinum (Liliaceae), Villarsia (Men- yanthaceae), and Amsinckia (Boraginaceae). For Nymphoides (Menyanthaceae), he published evi- dence that heterostyly has been replaced by dioe- cism and analyzed the selective forces involved. Ornduff was also interested in the population dy- namics resulting from isolation and edaphic con- ditions. Early on in his career, in 1961, he visited the Farallon Islands off the California coast, where the sparse soil developed largely from seabird gua- no, publishing a florula of South Farallon Island. In 1981, he visited several granite rock outcrops in Western Australia (likely ones that Darwin had vis- ited) and identified them as “islands in a sea of Eucalyptus,” analyzing their endemic plants. He re- turned in 1983 to undertake floristic studies that were later published and described in the Harold L. 294 iIGe2. rector of the University of California Botanical Garden, 1991. Photo courtesy of the U.C. Botanical Garden. Dr. Robert Ornduff at his retirement party as Di- Lyon Arboretum Lecture Series and publications at the University of Hawaii. His papers are lucidly written and often pose interesting questions for fur- ther research. Notable in Bob Ornduff’s career was his work with graduate students. Two students, Steve Weller (University of California, Irvine) and Spencer Bar- rett (University of Toronto), fondly remember field trips with Bob to see his beloved Jepsonia popu- lations in the Sierra Nevada foothills. They describe riding in his sleek Mercedes-Benz sports car ac- companied by a gallon of ‘‘fine red wine.”” Spencer writes, “Bob always combined the best of botany with the gourmet delights of a field picnic in the finest style. This was one of the most important lessons I learned from him.’’ Steve writes, ‘‘This trip highlighted Bob’s love for the California land- scape, his unerring ability to find interesting plants and research projects, and his generosity in sharing this with his students.”’ Former graduate student Jim Eckenwalder, now Associate Professor at the University of Toronto, provided a synopsis of Bob’s approach to supervis- ing graduate students. “‘In contrast to the increas- ingly common practice of slotting Ph.D. students MADRONO i [Vol. 47, into an established program, Bob was a hands-off | supervisor who encouraged our independence as scientists from the start. Even so he was always | there for you when you needed him (at least when | he wasn’t doing fieldwork in some exotic locale). | Most importantly, he never dropped his standards and always provided effective and instructive crit- icism, whether of the thesis, manuscripts, or oral | presentations. Most of us worked on projects be- yond his immediate area of interest, so his well honed and sometimes devastatingly sharp critical | faculties were for us his greatest legacy. Mentor and friend, I will sorely miss ‘Our O.’ ”’ | Doug James (currently the owner of a small mi-_ cropropagating facility in Berkeley), longtime | friend and three-time teaching assistant for Orn- duff’s California Plant Life course (Botany 125) in the 1970’s recalls what a pleasure it was to be a teaching assistant for this class as Bob was such an | extraordinarily good teacher. ‘‘Although majors. would take his courses, he got supreme joy illu-— minating the subject for neophytes.... He never. missed an opportunity to expand a discussion with examples from his own experiences or observations — in the field. He made the subject come alive for his students.” For eighteen years from 1973 to 1991, Ornduff was director of the University Botanical Garden in Strawberry Canyon, Berkeley. His association with the garden brought him some of his greatest pleas- ures because of his lifelong interest in the diversity of plants and in cultivating them. He collected seed | wherever he traveled and greatly enriched the col- lections in the garden in this way. Visiting the other four areas of the world with mediterranean cli-— mates, the Mediterranean itself, Western Australia, | Chile, and South Africa, often with his close friend Bill Culberson, Bob built substantial living collec- tions from each of them. Holly Forbes, curator at — the garden, reports that Bob donated over 1100 col- lections from twenty-three countries, with over 700 still growing there. As Steve Weller, a former grad- uate student, observed ‘‘Bob’s accessions for the garden were remarkable in their number and diver- sity. Certainly, it is hard to imagine a more inter- esting botanical garden during the era when he was contributing to the development of the collections. It was a labor of love for him, and he had a lifelong commitment to the garden, long after he ceased to be director.”’ His successful development of a ‘“‘Friends of the Garden” docent program, however, perhaps provid- ed Bob the greatest satisfaction. In an interview shortly before his death he commented that the Friends program’s growth and evolution was one of his greatest pleasures. He liked getting to know people in the community, had a great admiration for many of the people in the program, and thought the docent program was “‘superb.”’ As a result of his efforts, the Garden gradually increased its size and scope, and became a world-class collection of 2000] plants from California and beyond. In recommend- ing Ornduff for an Award of Merit by the Botanical ‘Society of America (given in 1993), Sherwin Carl- 'quist, a long-time friend and research botanist at ‘the Santa Barbara Botanic Garden, commented that ‘even though small in size, the University Botanical Garden ‘‘is, without a doubt, the most significant botanical garden in the United States, acre for acre, | » when compared to other gardens. In 1992, Ornduff became director of the Ameri- can branch of the Stanley Smith Horticultural Trust, a charitable foundation with a sister branch in Brit- ain, both providing support for education and re- search in horticulture in botanical gardens, arboreta, and related institutions. He held this position until his death, refining the Trust’s focus to include small gardens as well as great ones and thereby helping many fledgling operations establish viable pro- grams. Unlike many professional botanists who shun association with horticulture, Ornduff always sought it out as yet another expression of his love of plants. With his ready wit and humor and profound knowledge of the California flora, Bob was a pop- ular speaker at garden clubs, California Native Plant Society meetings, and campus events. He last spoke at the 50th Anniversary Jepson Symposium banquet in June, 2000 with his lively talk entitled ‘Piss and Vinegar: Some Early California Bota- nists.”” As a founding board member of the Cali- fornia Native Plant Society, he played an important role in the formation of the organization by serving as vice-president from 1969 to 1971, and president from 1972 to 1973 in the era before chapters were formed. From 1971 to 1975 he was chairman of the editorial committee that produced the CNPS news- letter and founded its journal. Bob’s proposed name for the journal, Fremontia, won a naming contest and for twenty-seven years (ten years for Marge Hayakawa and seventeen for Faber) he reviewed and commented on almost all submitted manu- scripts for the journal. As editor, one of us (PF) was always profoundly grateful for his fast turn- around of material and quick eye for plant names. Steve Weller commented on the same quality. “‘He was a wonderful editor, and read everything I wrote within twenty-four hours!”’ Ornduff received numerous honors for his work and contributions, among them: Award of Merit from the American Association of Botanical Gar- dens and Arboreta (1993); Merit Award from the Botanical Society of America (1993); EF Owen Pearce Award of Horticulture, Strybing Arboretum Society (1994); Fellow, American Association for the Advancement of Science (1966); Fellow, Cali- fornia Academy of Sciences (1967); Fellow, Cali- fornia Native Plant Society (1997); and member of the team that prepared the master plan for the San Luis Obispo Botanical Garden, which received a 1999 Merit Award of Planning from the American Society of Landscape Architects. ROBERT ORNDUFF— 1932-2000 295 Bob Ornduff had a remarkably full career doing what he loved best, being closely associated with plants in many diverse ways. He arrived at his of- fice at the university at around six a.m. “‘because he liked being there.’’ His weekends were spent at his house in Inverness, for him a haven from which he explored the Point Reyes Peninsula from such places as an old limestone kiln in the Olema Valley to his favorite view of Ten Mile Beach near the lighthouse. He regularly kept track of the vegetative recovery beyond his house from the 1995 Mt. Vi- sion fire and wrote on the subject twice. He liked picking huckleberries for jam and collecting apples from his own orchard for pies. For Bob, all plants had stories he enjoyed sharing. He collected myriad books about plants, and maintained friendships with botanists from many places around the world. His was a life devoted to championing, better un- derstanding, and enjoying the world of plants. —PHYLLIS M. FABER, University of California Press, Berkeley and PETER H. RAVEN, Missouri Botanical Gar- den, St. Louis. Robert Ornduff Curriculum Vitae Born June 13, 1932, Portland, Oregon. Educational Background B.A. Reed College, 1953 (Biology). Fulbright Scholar, Victoria University, Wellington, New Zealand, 1954 (Botany). M.Sc. University of Washington, 1956 (Botany). Ph.D. University of California, Berkeley, 1961 (Botany). Professional Experience Reed College: Assistant Professor, Biology, 1961— 1962. Duke University: Assistant Professor, Botany, 1962-1963. University of California at Berkeley, Botany: Assistant Professor, Botany, 1963—1966 Associate Professor 1966—1969 Professor 1969-1993 Professor Emeritus 1993-2001 Assistant Curator of Seed Plants, University Her- barium, 1965—1966 Associate Curator, University Herbarium, 1966— 1969 Curator, University Herbarium, 1969-1993 Director, University Herbarium, 1967-1982 Director, Jepson Herbarium and Library, 1968-— 1982 Director, University Botanical Garden 1973- 1991 Faculty Curator, 1991—2001 Chairman, Department of Botany, 1986-1989 Professional Service President, American Society of Plant Taxonomists, 1975 Chairman, Editorial Committee, University of Cal- ifornia Press, 1975-1989 Associate Editor, Evolution, 1975—1977 Council member, Society for the Study of Evolu- tion, 1977-1979 Membership Committee, American Society of Nat- uralists, 1978—1979 President, California Botanical Society, 1981-1982 Executive Director, Miller Institute for Basic Re- search in Science, 1984-1987 Program Committee, Society for Economic Botany, 1987-1988 Board of Councilors, Save-the-Redwoods League, 1988-2001 Board of Directors, Pacific Horticultural Founda- tion, 1992-2001 Grants Director, The Stanley Smith Horticultural Trust. 1992—2001 Editorial Advisor, Fremontia, 1997—2001 Board of Trustees, Center for Plant Conservation, 1997-2001 Awards and Honors Award of Merit, American Association of Botanical Gardens and Arboreta, 1993 Merit Award, Botanical Society of America, 1993 E Owen Pearce Award of Horticulture, Strybing Arboretum Society, 1994 Fellow, California Native Plant Society, 1997 Author of over 100 scientific papers and about 50 papers on various horticultural and related topics. Publications 1960. An interpretation of the Senecio lautus com- plex in New Zealand. Transactions of the Royal Society of New Zealand 88:63-—77. Distributional notes on plants of the Warm Springs Area, Oregon. Madrono 15:225— 230. [With D. H. French. ] The Farallon flora. Leaflets of Western Bot- any 9:139-142. Heterostyly in Jepsonia. In Recent Advances in Botany 1:885—887. Senecio lautus complex. Pp. 1027—1030 in H. H. Allan, Flora of New Zealand. Interspecific hybridizations in Senecio glau- cophyllus Cheesem. Transactions of the Roy- al Society of New Zealand (Botany) 1:225— 229, Chromosome numbers in Compositae. III. Senecioneae. American Journal of Botany 50:131-139. [With P. H. Raven, D. W. Ky- hos, and A. R. Kruckeberg. | Experimental studies in two genera of Hel- eiaea (Compositae): Blennosperma and Las- thenia. Quarterly Review of Biology 3:141-— 150: 1960. 1961. 1961. 1961. L962. 1963. 1963: MADRONO 1964. 1964. 1964. 1964. [9G5. 1966. 1966. 1967. I9G7- 1967. 1967. 1968. 1968. 1969: 1969. 1969. 1969: 1969. 1970. 1970. 1970. [Vol. 47. The breeding system of Oxalis suksdorfii. | American Journal of Botany 51:307—314. Biosystematics of Blennosperma (Composi- ! tae). Brittonia 16:289-—295. | Reproductive biology of Prigueta carolini-| ana (Turneraceae). Rhodora 66:100—109.. [With J. D. Perry.] Evolutionary pathways of the Senecio lautus alliance in New Zealand and Australia. Evo-. lution 18:349—360. Ornithocoprophilous endemism in Pacific. basin angiosperms. Ecology 46:864-876. The goldfield genus Lasthenia (Compositae: | Helenieae): a biosystematic survey. Univer- sity of California Publications in Botany 40: | 1-92. | The origin of dioecism from heterostyly in- Nymphoides (Menyanthaceae). Evolution | 20:309-3 14. The breeding system of Pontederia cordata | L. (Pontederiaceae). Bulletin of the Torrey Botanical Club 93:407—416. Chromosome numbers in Compositae. VI. | Senecioneae. II. American Journal of Botany | 54:205—-213. [With T. Mosquin, D. W. Ky- | hos, and P. H. Raven. |] | Hybridization and regional variation in Pa-— cific northwestern /mpatiens (Balsamina- ceae). Brittonia 19:122—128. Papers on plant systematics. Little, Brown and Co., Boston. 429 pp. Numerical taxonomy of Limnanthaceae. American Journal of Botany 55:173-182. [With T. J. Crovello.] The systematics of populations in plants. Pp. 104-128 in Systematic Biology. National Academy of Sciences Press, Washington, DC. Limnanthes vinculans: a new California en- demic. Brittonia 21:11-14. Reproductive biology in relation to system- atics. Taxon 18:121-133. Ecology, morphology, and systematics of Jepsonia (Saxifragaceae). Brittonia 21:286— 298. Neotropical Nymphoides (Menyanthaceae): Meso-American and West Indian species. Brittonia 21:346—352. The origin and relationships of Lasthenia burkei (Compositae). American Journal of Botany 56:1042—1047. The systematics and breeding system of Gel- semium (Loganiaceae). Journal of the Ar- nold Arboretum 51:1-17. Variation in the spectral qualities of flowers in the Nymphoides indica complex (Meny- anthaceae) and its possible adaptive signifi- cance. Canadian Journal of Botany 48:603-— 605. [With T. Mosquin. ] Rhododendrons at the University of Califor- 1971. 1972. O72. 1973. O73: 1973: FOS: 1974. 1974. 1974. 1974. ROBERT ORNDUFF— 1932-2000 nia Botanical Garden. California Horticul- tural Journal 31:23—24, 33. . Pathways and patterns of evolution: a dis- cussion. Taxon 19:202—204. . Relationships in the Piriqueta caroliniana— P. cistoides complex (Turneraceae). Journal of the Arnold Arboretum 51:492—498. . Heteromorphic incompatibility in Jepsonia malvifolia. Bulletin of the Torrey Botanical Club 97:258-—261. . Incompatibility and the pollen economy of Jepsonia parryi. American Journal of Bota- ny 57:1036—-1041. . Cytogeography of Nymphoides (Menyantha- ceae). Taxon 19:715—719. . The reproductive system of Jepsonia heter- andra. Evolution 25:300—311. . A new tetraploid subspecies of Lasthenia (Compositae) from Oregon. Madrono 21:96— 98. . Systematic studies of Limnanthaceae. Ma- drono 21:103-—111. . The impact of science on society. South Af- rican Journal of Science 67:542—546. . Flavonoids of Lasthenia conjugens and Las- thenia fremontii. Phytochemistry 10:61 1-14. [With N. A. M. Saleh and B. A. Bohm.] . Genetics and taxonomy: a perspective. Pp. 11-13 in Proc. 4th Congress South African Genetic Soc. Tentative key to the poeciliid fishes of Costa Rica. Jn C. E. Schnell, 24-8:1—2 O.T.S. Data, keys, tables, figures, miscellaneous. The breakdown of trimorphic incompatibil- ity in Oxalis section Corniculatae. Evolution 26:52-65. Plant life in California. University of Cali- fornia Extension, Berkeley. 110 pp. The flavonoids and affinities of Blennosper- ma and Crocidium (Compositae). Taxon 22: 407-412. [With N. A. M. Saleh and B. A. Bohm. ] Menyanthaceae. (Taxonomy). /n S. Nilsson (ed.), World Pollen and Spore Flora 2:1—2, se A new Oxalis from the Western Cape. Jour- nal of South African Botany 39:201—203. Flavonoids of artificial interspecific hybrids of Lasthenia. Biochemical Systematics 1: 147-151. [With N. A. M. Saleh and B. A. Bohm. ] The flavonoids of Lasthenia. American Jour- nal of Botany 61:551—561. [With N. A. M. Saleh and B. A. Bohm.] Flavonoid races in Lasthenia. Brittonia 26: 411-420. [With B. A. Bohm and N. A. M. Saleh. ] Heterostyly in South African plants: a con- spectus. Journal of South African Botany 40: 169-187. Cytotaxonomic observation on Villarsia 1974. 1974. [O7S: Eo TS: 1975. 1975. 1975. 1976. 1976. 1976. 1976. 1976. OTs O77. Lew 197s: 1978. 1978. LOTS: 1978. 297, (Menyanthaceae). Australian Journal of Bot- any 22:513-516. Introduction to California plant life. Univer- sity of California Press. 152 pp. Disc-electrophoresis of albumin and globu- lin fractions from dormant achenes of Las- thenia (Compositae). Biochem. Systematics and Ecology 2:67—72. [With I. Altosaar and B. A. Bohm. ] Relationships of Rigiopappus and Tracyina. Madrono 23:53—55. [With B. A. Bohm. ] Geography of pollen and chromosomal het- eromorphism in Leucocrinum montanum (Liliaceae). Madrono 23:65—67. [With Mar- ion S. Cave.] Complementary roles of halictids and syr- phids in the pollination of Jepsonia heter- andra (Saxifragaceae). Evolution 29:371-— 373: Pattern diversity of incompatibility groups in Jepsonia heterandra (Saxitragaceae). Evo- lution 29:373-—75. [With S. G. Weller. ] Pollen flow in Lythrum junceum, a tristylous species. New Phytologist 75:161—166. Heterostyly and pollen flow in Hypericum aegypticum (Guttiferae). Botanical Journal of the Linnean Society 71:51—57. The reproductive system of Amsinckia gran- diflora, a distylous species. Systematic Bot- any 1:57—66. Speciation and oligogenic differentiation in Lasthenia (Compositae). Systematic Botany 1:91—96. Sympatry, allopatry, and interspecific com- petition in Lasthenia. In S. Jain (ed.), Vernal pools, their ecology and conservation. Insti- tute of Ecology (U.C. Davis) Publication No. 9:46-50. Plants and man. W. H. Freeman and Co., San Francisco. An unusual homostyly in Hedyotis caerulea (Rubiaceae). Plant Systematics and Evolu- tion 127:293-297. The annual habit. Fremontia 5:3—5. Cryptic self-incompatibility in Amsinckia grandiflora. Evolution 31:47—51. [With S. G. Weller. ] Floral enantiomorphy and the reproductive system of Wachendorfia paniculata (Hae- modoraceae). New Phytologist 80:427—431. [With R. Dulberger. ] Chromosome number in Lachenalia (Lili- aceae). Journal of South African Botany 44: 387-390. [With P. J. Watters) Features of pollen flow in dimorphic species of Lythrum section Euhyssopifolia. Ameri- can Journal of Botany 65:1077—1083. Using living collections: the problems. Bul- letin of the American Association of Botanic Gardens and Arboreta 12:113—117. Lasthenia californica (Compositae), another 298 1978. Oyo: 1979. 1979: 1979. 17: 1979: 1979: 1979: 1979. 1979: 1979. 1980. 1980. 1980. 1980. 1980. 1980. 1981, OS 1982. MADRONO name for a common goldfield. Madrofio 25: 227. [With D. E. Johnson. ] Chemotaxonomic studies in the Saxifraga- ceae s.1. 9. Flavonoids of Jepsonia. Madrono 25:39—43. [With B. A. Bohm.] The morphological nature of distyly in Lyth- rum section Euhyssopifolia. Bulletin of the Torrey Botanical Club 106:4-8. Pollen flow in a population of Primula vul- garis Huds. Botanical Journal of the Linnean Society 78:1—10. Reproductive characters and taxonomy. Sys- tematic Botany 3:420—427. Features of pollen flow in Gelsemium sem- pervirens (Loganiaceae). Journal of the Ar- nold Arboretum 60:377—381. Heterostyly in Oplonia (Acanthaceae). Jour- nal of the Arnold Arboretum 60:382—385. The genetics of heterostyly in Hypericum aegypticum. Heredity 42:271—272. Systematic Botany Resources in America. Part 2: The cost of services. New York Bo- tanical Garden, Cary Arboretum, Millbrook, NY. 116 pp. [With others. ] South African bulbs at home. Pacific Horti- culture. 40:28—31. California native plants: a distinctive flora. Pp. 29-38 in Lester Rowntree, California native plants woman. Bancroft Library, Uni- versity of California, Berkeley. Chromosome numbers and relationships of certain African and American genera of Haemodoraceae. Annals of the Missouri Bo- tanical Garden 66:577—580. Chromosome numbers in Cyanella (Teco- philaeceae). Annals of the Missouri Botani- cal Garden 66:581-—583. The University of California Botanical Gar- den today. Pacific Horticulture 41:9—10. Heterostyly, population composition, and pollen flow in Hedyotis caerulea. American Journal of Botany 67:95—103. The probable genetics of distyly in Gelsem- ium sempervirens (Loganiaceae). Canadian Journal of Genetics and Cytology 22:303- 304. Pollen flow in Primula veris (Primulaceae). Plant Systematics and Evolution 135:80—93. Joseph Burtt Davy: a decade in California. Madrono 27:171—176. Floral morphology and reproductive biology of four species of Cyanella (Tecophilae- aceae). New Phytologist 86:45—56. [With R. Dulberger. ] Leaf flavonoids and ordinal affinities of Cor- lariaceae. Systematic Botany 6:15—26. [With B. A. Bohm. ] The University of California Botanical Gar- den. Bulletin of the American Association of Botanic Gardens and Arboreta 15:6—8. Heterostyly and incompatibility in Villarsia 1982. 1983. 1983. 1984. 1985. 1935. 1985. 1985: 1985. 1986. 1986. 1986. 1987. 1987. 1987. 1988. 1988. capitata (Menyanthaceae). Taxon 31:495— 497. Interpretations of sex in higher plants. Pp.| 21-33 in W. J. Meudt (ed.), USDA/BARC) strategies of plant reproduction. | Heteromorphic incompatibility in Oxalis nel-' sonii. Bulletin of the Torrey Botanical Club} 110:214-216. Studies on the reproductive system of Niv-| enia corymbosa (Iridaceae), an apparently! androdioecious species. Annals of the Mis- souri Botanical Garden 70:146-148. | Darwin’s botany. Taxon 33:39—47. | Comparative fecundity of morphs in a mixed monoclinous-andromonoecious population. of Wurmbea dioica subsp. alba (Liliaceae) in Western Australia. Plant Systematics and | Evolution 149:299—302. | Flavonoids and affinities of the Cephalota-, ceae. Biochemical Systematics and Ecology 13:261—263. [With K. W. Nichols and B. Ag Bohm. ] Allozyme variation within and between Las- | thenia minor and its derivative species, L. | maritima (Asteraceae). American Journal of. Botany 72:1177—1184. [With D. J. Crawford | and M. C. Vasey. |] | Review of V. H. Heywood and D. M. Moore (eds.), Current concepts in plant taxonomy. | Quarterly Review of Biology 60:507. Review of Meeuse, Bastiaan and Sean Mor-_ ris, The sex life of flowers. Systematic Bot- | any 10:236. | Male-based sex ratios in the cycad Macro- zamia riedlei (Zamiaceae). Bulletin of the. Torrey Botanical Club 112:393-397. Flavonoids of the Menyanthaceae: intra- and interfamilial relationships. American Journal of Botany 73:204—213. [With B. A. Bohm and K. W. Nicholls. ] | Comparative fecundity and population com- position of heterostylous and non-heterosty- lous species of Villarsia (Menyanthaceae) in Western Australia. American Journal of Bot- any 73:282-—286. Islands on islands: plant life on the granite outcrops of Western Australia. University of Hawaii Harold Lyon Arboretum Lecture 15: 5-28. Reproductive systems and chromosome rac- es of Oxalis pes-caprae L. and their bearing on the genesis of a noxious weed. Annals of the Missouri Botanical Garden 74:79—84. Sex ratios and coning frequency in the cycad Zamia pumila L. (Zamiaceae) in the Domin- ican Republic. Biotropica 19:361—364. Distyly and incompatibility in Villarsia con- gestiflora (Menyanthaceae), with compara- tive remarks on V. capitata. Plant System- atics and Evolution 159:81-83. Chromosome numbers of Western Australian 1988. n989. 1989. 1989. 1989. 1989. £990. 1990. #990. 1991. 1991. 1991. HOOT. OO 1: IO 2: ROBERT ORNDUFF— 1932-2000 species of Villarsia (Menyanthaceae). Plant Systematics and Evolution 161:49—52. [With T. I. Chuang. ] Distyly and monomorphism in Villarsia (Menyanthaceae): some evolutionary con- siderations. Annals of the Missouri Botanical Garden 75:761-—767. Incompatibility in Amsinckia grandiflora (Boraginaceae): distribution of callose plugs and pollen tubes following intermorph and intramorph crosses. American Journal of Botany 76:277—282. [With S. G. Weller.] A six-year study of flavonoid distribution in a population of Lasthenia californica (As- teraceae). American Journal of Botany 76: 157-163. [With B. A. Bohm, A. Herring, K. W. Nicholls, and L. R. Bohm. ] Enzyme electrophoresis and evolutionary re- lationships among three species of Lasthenia (Asteraceae: Heliantheae). American Journal of Botany 76:289-296. [With D. J. Craw- ford. ] Size distribution and coning behavior of the Australian cycad Lepidozamia peroffskyana. Australian Journal of Ecology 14:241—245. Cyanella. In A. H. Halevy (ed.), Handbook of flowering, vol. 6:275—280. [With R. Dul- berger. | Geographic variation in reproductive behav- ior and size structure of the Australian cycad Macrozamia communis (Zamiaceae). Amer- ican Journal of Botany 77:92—99. A new species of Villarsia (Menyanthaceae from the Porongurup Range, Western Aus- tralia. Systematic Botany 15:216—220. Genetic structure of the Australian cycad Macrozamia communis (Zamiaceae). Amer- ican Journal of Botany 77:677—681. [With N. C. Ellstrand and J. M. Clegg.] Coning phenology of the cycad Macrozamia riedlei (Zamiaceae) over a five-year period. Bulletin of the Torrey Botanical Club 118: 6-11. South African bulbs for California gardens. Pp. 65-69 in J. Glattstein (ed.), Gardner’s world of bulbs. Brooklyn Botanic Garden. Size classes, reproductive behavior, and in- sect associates of Cycas media (Cycadaceae) in Australia. Botanical Gazette 152:203-— 207. Pollen-tube growth and inbreeding depres- sion in Amsinckia grandiflora (Boragina- ceae). American Journal of Botany 78:801— 804. [With S. G. Weller. ] The role of botanical gardens in resource conservation. Proceedings of the conference on landscaping with wildflowers and native plants. Wildflower 4:22—25. Features of coning and foliar phenology, size classes, and insect associates of Cycas arms- trongii (Cycadaceae) in the Northern Terri- f992. 1992. 1922. 1992. Oo? 1993: 1993. 1993: 1993, 1994. 1994. 1994. 1994. 1994. 1994. 1994. 190): 299 tory, Australia. Bulletin of Torrey Botanical Club 119:39—43. Historical perspectives on heterostyly. Pp. 31-39 in S. C. H. Barrett (ed.), Evolution and function of heterostyly. Springer-Verlag. Love and hate among the composites. Uni- versity of California Botanical Garden Newsletter 17:6—7. Intrapopulation variation in the breeding sys- tem of Villarsia lasiosperma (Menyantha- ceae), a distylous species. Plant Systematics and Evolution 180:227—233. Heterostyly. Wildflower 5:18—22. Seed morphology and systematics of Meny- anthaceae. American Journal of Botany 79: 1396-1406. [With T. I. Chuang. ] Intramorph and intermorph compatibility in tristylous populations of Decodon verticil- latus. Bulletin of the Torrey Botanical Club 120:19-22. Limnanthaceae [pp. 736-738]; Oxalidaceae [pp. 808-809]; Lasthenia [pp. 298-300]; Blennosperma [p. 214]; Rigiopappus [p. 334]; Tracyina [p. 354]. In J. C. Hickman (ed.), The Jepson manual: higher plants of California. University of California Press, Berkeley. Review of Arthur R. Kruckeberg, The nat- ural history of Puget Sound country. Madro- no 40:186—-187. Studies of the reproductive biology of Mac- rozamia species and Cycas media in Austra- lia. Pp. 121-124 in D. W. Sevenson and K. J. Norstog (eds.), Proceedings of CYCAD 90, second international conference on cycad biology. Palm and Cycad Societies of Aus- tralia. The UCB Chiapas plant collection: conser- vation and public education. Pp. 71—73 in J. White (ed.), Innovations in landscape pest management. University of California Divi- sion of Agriculture and Natural Resources. American plants. Science 263:838—839. Ringing the changes. Pacific Discovery 47: 50-51. Villarsia cambodiana (Menyanthaceae) in southeastern Asia. Nordic Journal of Botany 14:647—-648. Herbert L. Mason, 1896-1994. Madrono 41: 235-238. [With L. Constance. | A botanist’s view of the big tree. Pp. 11—14 in Proceedings of the symposium on giant sequoias: their place in the ecosystem and society. USDA Forest Service Gen. Tech. Rep. PSW-GTR-I51. Lasthenia minor (DC.) Ornduff. Madrono 41:332—333. [With M. C. Vasey and E. N. Kozloff. | Menyanthaceae. Pp. 140—142 in Wu Zheng- yi and P. H. Raven (eds.), Flora of China, 300 1095: 1996, 1996. 1996, 1996. 1997. eae LOOT: 1997. 1997. 127, 1998. 1998: LOOS. POOS. 1O°8. 1005. MADRONO Vol. 16. St. Louis: Missouri Botanical Gar- den. [With Ho Ting-nung. | The flora and vegetation of the Marin Is- lands, California. Madrono 42:358-—365. [With M. C. Vasey. ] The breeding system of Villarsia exaltata (Menyanthaceae), a distylous species. Telo- pea 6:805-811. Gender performance in a cultivated cohort of the cycad Zamia integrifolia (Zamiaceae). American Journal of Botany 83:1006—1015. A Californian’s commentary on plant life in Mediterranean climates. Pp. 81-89 in S. D. Hopper et al. (eds.), Gondwanan heritage: past, present and future of the Western Aus- tralian biota. Surrey Beatty & Sons, Chip- ping Norton. An unusual floral monomorphism in Villar- sia (Menyanthaceae) and its proposed origin from distyly. Pp. 212—222 in S. D. Hopper et al. (eds.), Gondwanan heritage: past, pres- ent and future of the Western Australian bi- ota. Surrey Beatty & Sons, Chipping Norton. Editorial “‘plants’’? Letter in Science 276: 1276, Rush to judgement. Letter in Pacific Discov- ery 50:49. Marion Stilwell Cave. University of Califor- nia, in memoriam, pp. 45—47. Academic Senate, University of California, Berkeley. [With L. Constance and D. R. Kaplan. ] Rebirth of a bishop pine forest: first year af- ter the Mount Vision fire. Fremontia 25:22— 28. [With Virginia Norris. ] San Francisco Bay region. Pp. 68—69 in P. M. Faber (ed.), California’s wild gardens. California Native Plant Society. Marion Stilwell Cave (1904—1995). Madro- no 44:211-—213. [With D. R. Kaplan and L. Constance. ] The Sequoia sempervirens (coast redwood) forest of the Pacific coast, USA. Pp. 221- 236 in Aimlee D. Laderman (ed.), Coastally restricted forests. Oxford University Press, New York. Botanical Mother Lode. [Review of P. M. Faber (ed.), California’s wild gardens: a liv- ing legacy.]California Wild 51:50—51. Review of Jack Hobbs and Terry Hatch, Best bulbs for temperate climates. Review of Bri- an Mathew, Growing bulbs: the complete practical guide. Pacific Horticulture 59:7-8. Review of P. M. Faber (ed.), California’s wild gardens: a living legacy. Fremontia 26: 36-37. Oxalidaceae: Oxalis family. Journal of the Arizona-Nevada Academy of Science 30: 115-119. [With Melinda Denton. ] Review of A. J. Richards, J. Kirschner, J. 1998. 1998. 1998. 1995. 1998. 199°. 1992. 1999. 1999) O°. (Vol. 47. Stepanek, and K. Marthold (eds.), Apomixis. and taxonomy. Systematic Botany 22:727-. 7128: | Preface. Jn P R. Dallman, Plant life in the, world’s Mediterranean climates. California Native Plant Society, Sacramento, and Uni- versity of California Press, Berkeley and “ Angeles. Ecology, conservation, and management of | vernal pool ecosystems. Proceedings of a 1996 conference. California Native Plant So-. ciety, Sacramento. [With C. W. Witham, E.. T. Bauder, D. Belk, and W. R. Ferren, Jr.] Goldfields revisited: phylogenetic analysis of the genus Lasthenia (Asteraceae) using mor- phology and DNA sequence data from the | nuclear ribosomal ITS and chloroplast trnK | intron regions. American Journal of Botany. 85:119. [With R. Chan. ] Three years after the Vision Fire in Point Reyes. Fremontia 26:26—27. Spirit of Mary Austin. Letter in California Wild 51:3. | Review of Peter Goldblatt and John Man- | ning, Gladiolus in southern Africa. Pacific | Horticulture 60:10—11. | Letter to editor re: Bishop Museum funding | | Honolulu Star Bulletin, April 23. Letter to editor re: Bishop Museum vital for. Pacific research, Honolulu Advertiser, April 29, Limnanthes macounii Trel. (Limnanthaceae). Madrono 45:184. [With Eva Buxton. ] A new species of Villarsia (Menyanthaceae) from South Africa. Novon 9:407—409. Dr. Robert Ornduff, major advisor for the following students: Lloyd, Robert Michael Niehaus, Theodore F Van Horn, Gene Stanley Kalin, Mary Therese Carter, Lynne Carter Hinton, William Frederick Ganders, Fred Russell Weller, Stephen Gregory Eckenwalder, James Emory Johnson, Dale Edwin Hannan, Gary Louis Morin, Nancy Ruth Schoen, Daniel Jay Lowrey, Timothy Kemper Fruchter, David Edward Thompson, David Malcolm Price, Robert Allen Mayer, Stephanie Susan Buck, Roy Ernest Ray, Martin Forbes Appleby, Walter Robert Botany 02/16/1967 Botany 03/02/1967 Botany 05/10/1968 Botany 06/12/1969 Sesame 03/02/1971 Botany 03/10/1972 Botany 06/21/1972 Botany 11/13/1973 Botany 06/19/1974 Botany 08/02/1974 Botany 03/19/1976 Botany 11/29/1977 Botany 02/21/1978 Botany 12/01/1978 Botany 12/11/1978 Botany 03/12/1982 Botany 03/18/1982 Botany 01/03/1984 Botany 02/17/1988 Botany 08/01/1990 Botany 07/29/1993 SUBSCRIPTIONS—-MEMBERSHIP Membership in the California Botanical Society is open to individuals ($27 per year; family $30 per year; emeritus $17 per year; students $17 per year for a maximum of 7 years). Late fees may be assessed. Members of the Society receive Maprono free. Institutional subscriptions to MADRONO are available ($60). Membership is based on a calendar year only. Life memberships are $540. Applications for membership (including dues), orders for sub- scriptions, and renewal payments should be sent to the Treasurer. Requests and rates for back issues, changes of address, and undelivered copies of Maprono should be sent to the Corresponding Secretary. INFORMATION FOR CONTRIBUTORS Manuscripts submitted for publication in Maprono should be sent to the editor. It is preferred that all authors be members of the California Botanical Society. Manuscripts by authors having outstanding page charges will not be sent for review. Manuscripts may be submitted in English or Spanish. English-language manuscripts dealing with taxa or topics of Latin America and Spanish-language manuscripts must have a Spanish RESUMEN and an English ABsTRACT. Manuscripts and review copies of illustrations must be submitted in triplicate for all articles and short items (NOTES, NOTEWORTHY COLLECTIONS, POINTS OF VIEW, etc.). Follow the format used in recent issues for the type of item submitted. Allow ample margins all around. Manuscripts MUST BE DOUBLE-SPACED THROUGHOUT. For articles this includes title (all caps, centered), author names (all caps, centered), addresses (caps and lower case, centered), abstract and resumen, 5 key words or phrases, text, acknowledgments, literature cited, tables (caption on same page), and figure captions (grouped as consecutive paragraphs on one page). Order parts in the sequence listed, ending with figures. Each page should have a running header that includes the name(s) of the author(s), a shortened title, and the page number. Do not use a separate cover page or ‘erasable’ paper. Avoid footnotes except to indicate address changes. Abbreviations should be used sparingly and only standard abbrevia- tions will be accepted. Table and figure captions should contain all information relevant to information presented. All measurements and elevations should be in metric units, except specimen citations, which may include English or metric measurements. Authors of accepted papers will be asked to submit an electronic version of the manuscript. Microsoft Word 6.0 or WordPerfect 6.0 for Windows is the preferred software. Line copy illustrations should be clean and legible, proportioned to the MADRONO page. Scales should be in- cluded in figures, as should explanation of symbols, including graph coordinates. Symbols smaller than 1 mm after reduction are not acceptable. Maps must include a scale and latitude and longitude or UTM references. In no case should original illustrations be sent prior to the acceptance of a manuscript. Illustrations should be sent flat. No illustrations larger than 27 X 43 cm will be accepted. Presentation of nomenclatural matter (accepted names, synonyms, typification) should follow the format used by Sivinski, Robert C., in MaproNno 41(4), 1994. Institutional abbreviations in specimen citations should follow Holmgren, Keuken, and Schofield, Index Herbariorum, 8th ed. Names of authors of scientific names should be abbreviated according to Brummitt and Powell, Authors of Plant Names (1992) and, if not included in this index, spelled out in full. Titles of all periodicals, serials, and books should be given in full. Books should include the place and date of publication, publisher, and edition, if other than the first. All members of the California Botanical Society are allotted 5 free pages per volume in Maprono. Joint authors may split the full page number. Beyond that number of pages a required editorial fee of $40 per page will be assessed. The purpose of this fee is not to pay directly for the costs of publishing any particular paper, but rather to allow the Society to continue publishing MADRONO on a reasonable schedule, with equity among all members for access to its pages. Printer’s fees for illustrations and typographically difficult material @ $35 per page (if their sum exceeds 30 percent of the paper) and for author’s changes after typesetting @ $4.50 per line will be charged to authors. At the time of submission, authors must provide information describing the extent to which data in the manu- script have been used in other papers that are published, in press, submitted, or soon to be submitted elsewhere. MADRONO A WEST AMERICAN JOURNAL OF BOTANY VOLUME XLVI 2001 BOARD OF EDITORS Class of: 2001—ROBERT PATTERSON, San Francisco State University, San Francisco, CA PAULA M. SCHIFFMAN, California State University, Northridge, CA 2002—-NORMAN ELLSTRAND, University of California, Riverside, CA CARLA M. D’ ANTONIO, University of California, Berkeley, CA 2003—-FREDERICK ZECHMAN, California State University, Fresno, CA JON E. KEELEY, U.S. Geological Service, Biological Resources Division, Three Rivers, CA 2004—Davipb Woop, California State University, Chico, CA INGRID PARKER, University of California, Santa Cruz, CA Editor—KRISTINA A. SCHIERENBECK California State University, Chico Department of Biology Chico, CA 95427-0515 kschierenbeck @csuchico.edu Published quarterly by the California Botanical Society, Inc. Life Sciences Building, University of California, Berkeley 94720 Printed by Allen Press, Inc., Lawrence, KS 66044 TABLE OF CONTENTS Allen-Diaz, Barbara, Randall D. Jackson and Catherine Phillips, Spring-fed plant communities of California’s Bast: Bay Elilis: Oak: WWOOGUAIN GS fess am ca eh Anderson, Paul D., James L. J. Houpis, David J. Anschel and James C. Pushnik, Among- and within-provenance variability of Pinus ponderosa (Pinaceae) seedling response to long-term elevated CO, exposure Anschel, David J. (see Anderson, Paul D.) Ashworth, Vanessa E. T. M., Bart C. O’Brien and Elizabeth A. Friar, Survey of Juniperus communis (Cupres- saceae) L. varieties from the western United States using RAPD fingerprinting Asp, Roxann (see Guralnick, Lonnie J.) Baldwin, Bruce G., Harmonia guggolziorum (Compositae: Madiinae), a new tarweed from ultramafics of south- | em: Mendocino County, California) cxs. thas ole i he ee ent eae 293 - Baldwin; Bruce, G,, President's report Lor Volare: 4B. casa ree ee ee 303 | Barbour, M. G., P. A. Castelfranco, M. Reymanek and R. W. Pearcy, A tribute to the contributions of Professor | AG IC IANO 2 ose eth otek A nt a ehh se AO Gil 1s wie nt aN cy et Ak cae 2158 Battaglia, Robyn, and Robert Patterson, A morphometric analysis of the Leptosiphon androsaceus complex (Polemoniaceae) im the Centralsand: South Coast Ranges: cc 2xc oe ee 62 | Bissell, Erin K. (see Gregory, Steven D.) Boyd, Steve, Noteworthy collection from California Brady, Timothy J., The significance of population successional status to the evolution of seedling morphology in Pinus contoria’ var. datijolia: (PINAaCeae)) <.2:222222 sem ssc 20 Dace Se ee ds i, 138 | Burgess, Rick, and Trisha Burgess, Noteworthy collections from California 211 | Burgess, Trisha (see Burgess, Rick) Burk, Jack H. (see Dorsett, Deborah K.) Calvin, Clyde L. (see Ruhland, Beverlee M.) Carlsen, Tina M. (see Gregory, Steven D.) Castelfranco, P. A. (see Barbour, M. G.) Chan, Raymund, A new section in the goldfield genus Lasthenia (Compositae: Heliantheae sensu lato) 38 | Chan, Raymund, Taxonomic changes and a new species in Lasthenia sect. Amphiachaenia (Compositae: He- WiAritlaeaeS Cus U At) ee Bee A a a Neue on ee ne ene gene 32 i pe 205 © Corbin, Jeffrey D. (see Reynolds, Sally A.) Davila-Aranda, Patricia (see G6mez-Sanchez, Maricela) D’ Antonio, Carla M. (see Reynolds, Sally A.) Dorsett, Deborah K., C. Eugene Jones and Jack H. Burk, The pollination biology of Eriastrum densifolium spp. Sanctorum (Polemoniaceac),.an’ endangered: plant) <.2222 1.20.20 as eg ie ee 265 Elle, Elizabeth (see Hare, J. Daniel) Espeland, Erin K. (see Gregory, Steven D.) Espeland, Erin K., and Rodney G. Myatt, Evidence for a sand hills ecotype of Eschscholzia californica (Papav- o} P10 a2} |e eas Pe oe eee et eee tee eases Eee eee eee ee Sere Ter Mat! Met We meer Sane ee Cah Sea eh eked v5) Everett, Richard G. (see Minnich, Richard A.) Friar, Elizabeth A. (see Ashworth, Vanessa E. T. M.) Gomez-Sanchez, Maricela, Patricia Davila-Aranda and Jess Valdés-Reyna, Estudio anat6mico de Swallenia (Poaceae: Eragrostideae: Monanthochloinae), un género monotipico de Norte América — 152 Gregory, Steven D., Erin K. Espeland, Tina M. Carlsen and Erin K. Bissell, Demography and population biology of a rare tarplant, Blepharizonia plumosa (Asteraceae), a California summer annual forb 212 Griffin, James R. (see Van Dyke, Eric) Guralnick, Lonnie J., Chad Marsh, Roxann Asp, and Aaron Karjala, Physiological and anatomical aspects of CAM-cycling in Lewisia cotyledon var. cotyledon (Portulacaceae) __....--..-- 22 131 Hare, J. Daniel, and Elizabeth Elle, Geographic variation in the frequencies of trichome phenotypes of Datura wrightii and correlation with annual water deficit 2222222222222 33 Holl, Karen D. (see Van Dyke, Eric) Holstein, Glen, Pre-acriculturalerassland in central California: 505. se ee 253 Houpis, James L. J. (see Anderson, Paul D.) Jackson, Randall D. (see Allen-Diaz, Barbara) Jacobson, Arthur L., Frederick C. Weinmann and Peter E Zika, Noteworthy collections from Washington PA'S) Jennings, W. Bryan, Comparative flowering phenology of plants in the western Mojave Desert 162 Johnson, Leigh A. (see Weese, Terri L.) Jones, C. Eugene (see Dorsett, Deborah K.) Karjala, Aaron (see Guralnick, Lonnie J.) Keeler-Wolf, Todd, Review of Terrestrial Ecoregions of North America: A Conservation Assessment by Taylor RICK CES EE ATs, Re ctrsrscchicee re cee foe te hm et ete Ute tices ta eae nee eed eos eee ce ee 45 Kephart, Paul (see Stromberg, Mark R.) Knight, Teri (see Pritchett, Daniel) _ 2001] TABLE OF CONTENTS Koontz, Jason A., and Pamela S. Soltis, Polyploidy and segregation analyses in Delphinium gypsophilum (Ran- TINGS UC A) em emer SIR Sarre ee SO eee ocean oes as G aan as sagectasSesiea yachwidute ses arenau leva cet Psp cabcoe ee ecoe sees eeeu Lesica, Peter, Recruitment of Franxinus pennsylvanica (Oleaceae) in eastern Montana woodlands — Marsh, Chad (see Guralnick, Lonnie J.) Minnich, Richard A., and Richard G. Everett, Conifer tree distributions in southern California Myatt, Rodney G. (see Espeland, Erin K.) Merrifield, Kathy, Bryophyte flora of William L. Finley National Wildlife Refuge, Willamette Valley, Oregon O’Brien, Bart C. (see Ashworth, Vanessa E. T. M.) Odion, Dennis C., Review of Savannas, Barrens and Rock Outcrop Plant Communities of North America edited by Roger C. Anderson, James S. Fralish and Jerry M. Baskin Patterson, Robert (see Battaglia, Robyn) Pearcy, R. W. (see Barbour, M. G.) Phillips, Catherine (see Allen-Diaz, Barbara) Pritchett, Daniel, Frank J. Smith and Teri Knight, Noteworthy collection from Nevada — Pushnik, James C. (see Anderson, Paul D.) Reeder, John R., Noteworthy collections from Mexico and Arizona Rejmanek, Marcel, Review of Trees and Shrubs of California by John D. Stuart and John O. Sawyer Rejmanek, Marcel (see Barbour, M. G.) Reynolds, Sally A., Jeffrey D. Corbin and Carla M. D’Antonio, The effects of litter and temperature on the germination of native and exotic grasses in a coastal California grassland Ruhland, Beverlee M., and Clyde L. Calvin, Morphological aspects of seedling establishment in four temperate nesion Phoradendron (VisCaceac) SpeCies 2-2 te ee Russell, William H., Review of Ecosystems of the World 16: Ecosystems of Disturbed Ground edited by Law- ROH CC sine ALK petinice, Scenes euaNz SMM Ar nacre 6h AT gk eed See A ahaa Raed Tic tiate SSA atal a peace a Rear cease Sno Re Be eee Sanderson, Stewart C. (see Stutz, Howard C.) Schierenbeck, Kristina A., Editor’s report for Volume 48 — Shevock, James R., and David Toren, A catalogue of mosses for the City and County of San Francisco, California Smith, Frank J. (see Pritchett, Daniel) Soltis, Pamela S. (see Koontz, Jason A.) Soreng, Robert J., A new species of Poa L. (Poaceae) from Baja California, Mexico Stromberg, Mark R., Paul Kephart and Vern Yadon, Composition, invasibility, and diversity in coastal California STAs Gol ZU) CS MORN Si RAP aL en fa N.S See Nc ater I ee, sc Stutz, Howard C., Mildred R. Stutz and Stewart C. Sanderson, Atriplex robusta (Chenopodiaceae), a new pe- Fennial species from northwestern Utah: 2.22.22 Stutz, Mildred R. (see Stutz, Howard C.) Tisch, Edward L., Corallorhiza maculata var. ozettensis (Orchidaceae), a new coral-root from coastal Washington Toren, David (see Shevock, James R.) Valdés-Reyna, Jestis (see G6mez-Sanchez, Maricela) Van Dyke, Eric, Karen D. Holl and James R. Griffin, Maritime chaparral community transition in the absence CET eM eM Nea ae SS Ae SaaS eyes Se tN Se ech gy aetna ah tn Ce ee os Se aan Walters, Gretchen, Noteworthy collection from Arizona Weese, Terri L., and Leigh A. Johnson, Saltugilia latimeri: A new species of Polemoniaceae — Weinmann, Frederick C. (see Jacobson, Arthur L.) Wilken, Dieter, A new /pomopsis (Polemoniaceae) from the southwest USA and adjacent Mexico Yadon, Vern (see Stromberg, Mark R.) Zander, Richard H., A new species of Didymodon (Musci) from California Zika, Peter F (see Jacobson, Arthur L.) DATES OF PUBLICATION OF MADRONO, VOLUME 48 Number 1, pages 1—50, published 31 August 2001 Number 2, pages 51—130, published 15 January 2002 Number 3, pages 131—220, published | February 2002 Number 4, pages 221—308, published 17 April 2002 ill 44 221 198 298 Statement of Ownership, Management, and Circulation (Required by 39 USC 3685) UNITED STATES B POSTAL SERVICE. Madrono 4. Issue Frequency Quarterly 5. Number of Issues Published Annually |6. Annual Subscription Price $27.00 7. Complete Mailing Address of Known Office of Publication (Not printer) (Street, city, county, state, and Zip+4) California Botanical Society, Inc.; Herbaria, Life Sciences Building University of California, Berkeley, CA 94720 Contact Person Roy Buck Telephone 510-848-4169 8. Complete Mailing Address of Headquarters or General Business Office of Publisher (Not printer) California Botanical Society, Inc. ; Herbaria, Life Sciences Building University of California; Berkeley, CA 94720 9. Full names and Complete Mailing Addresses of Publisher, Editor, and Managing Editor (Do not leave blank) Publisher (Name and complete mailing address) California Botanical Society, Inc., Herbaria, Life Sciences Building University of California; Berkeley, CA 94720 Editor (Name and complete mailing address) Kristina A. Schierenbeck; California State University, Chico; Dept. of Biology; Chico, CA 95429-0515 Managing Editor (Name and complete mailing address) 10. Owner (Do not leave blank. If the publication is owned by a corporation, give the name and address of the corporation immediately followed by the names and addresses of all stockholders owning or holding 1 percent or more of the total amount of stock. If not owned by a corporation, give the names and addresses of the individual owners. If owned by a partnership or other unincorporated firm, give its name and address as well as those of each individual owner. If the publicaiton is published by a nonprofit organization, give its name and address ) Full Name Complete Mailing Address Herbaria, Life Sciences Building California Botanical Society, Inc University of California Berkeley, CA 94720 77, Known Bondholders, Mortgages, and Other Security Holders Owning or Holding 1 Percent or More of Total Amount of Bonds, Mortgages, or Other Securities, If none, check box Full Name Complete Mailing Address 12. Tax Status (For completion by nonprofit organizations authorized to mail at special rates) (Check one) The purpose, function, and nonprofit status of this organization and the exempt status for federal income tax purposes. Has Not Changed During Preceding 12 Months Has Changed During Preceding 12 Months (Publisher must submit explanation of change with this statement) a PS Form 3526, September 1998 (See Instructions on Reverse) Computerized Facsimile [< 13. Publication Title 14. Issue Date for Circulation Data Below Madrono Oct-Dec 2001 15 Actual No. Copies of Single Issue Published Nearest to Filing Date ae Average No. Copies Each Issue Extent and Nature of Circulation During Preceding 12 Months a. Total Number of Copies (Net press run) Paid/Requested Outside- County Mail Subscriptions stated on Form 3541. (/nclude advertiser's proof and 874 908 exchange copies) Paid In-County Subscriptions (/nclude advertser's 0 b. Paid and/or proof and exchange copies) Requested Circulation Sales Through Dealers and Carriers, Street Vendors, 58 59 Counter Sales, and Other Non-USPS Paid Distribution fig see) : d. Free Outside-County as Stated on Form 3541 (je eal 1 Distribution Mail paula In-County as Stated on Form 3541 0 (Samples,complim entary, and other free) Other Classes Mailed Through the USPS 2 2 e Free Distribution Outside the Mail (Carriers or other means) 1 { Total Free Distribution (Sum of 15d and 152) Ss ee 4 g. Total Distribution (Sum of 15¢ and 15f) > 932 971 h. Copies not Distributed 174 179 Percent Paid and/or Requested Circulation 996 996 15c divided by 15g times 100 16. Publication of Statement of Ownership Publication required. Will be printed in the 48/4 issue of this publication. C1 Publication not required 17. Signature and Title of Editor, Publisher, Business Manager, or Owner Date & Rud 23 Marek 2 | certify that all information furnished on this form is true and complete. | understand that anyone who furnishes false or misleading information on this form or who omits material or information requested on the form may be subject to criminal sanctions (including fines and imprisonment) and/or civil sanctions (including civil penalties), Rsuvee Instructions to Publishers 1. Complete and file one copy of this form with your postmaster annually on or before October 1. Keep a copy of the completed form for your records. 2. In Cases where the stockholder or security holder is a trustee, include in items 10 and 11 the name of the person or corporation for whom the trustee is acting. Also include the names and addresses of individuals who are stockholders who own or hold 1 percent or more of the total amount of bonds, mortgages, or other securities of the publishing corporation. In item 11, if none, check the box. Use blank sheets if more space is required 3. Be sure to furnish all circulation information called for in item 15. Free circulation must be shown in items 15d, e, and f 4 Item 15h., Copies not Distributed, must include (1) newsstand copies originally stated on Form 3541, and returned to the publisher, (2) estimated returns from news agents, and (3), copies for office use, leftovers, spoiled, and all other copies not distributed 5. If the publication had Periodicals authorization as a general or requester publication, this Statement of Ownership, Management, and Circulation must be published, it must be printed in any issue in October or, if the publication is not published during October, the first issue printed after October 6. In item 16, indicate the date of the issue in which this Statement of Ownership will be published 7 Item 17 must be signed Failure to file or publish a statement of ownership may lead to suspension of second-class authonzation PS Form 3526. September 1998/Reverse) Computerized Facsimile VOLUME 48, NUMBER | JANUARY-MARCH 2001 Fi : fed BO Ts ite oH My \\i “td al y Y a ' | ‘ONTENTS A CATALOGUE OF MOSSES FOR THE CITY AND COUNTY OF SAN FRANCISCO, CALIFORNIA James R. Shevock ANd David TOren ...cc.ccccccccccccuccuccuccuccuccuccuccuscuscuccuceuceacs ] BRYOPHYTE FLORA OF WILLIAM L. FINLEY NATIONAL WILDLIFE REFUGE, WILLA- METTE VALLEY, OREGON A WEST AMERICAN JOURNAL OF BOTANY TCU IVI CY FUL crater neta ca Baeianleamemea ten eet en sec PONE SASRE ET AeA DANEN AMEE 17 EVIDENCE FOR A SAND HILLS ECOTYPE OF ESCHSCHOLZIA CALIFORNICA (PAPAVERACEAE) Erin K. Espeland and Rodney G. Myatt...» Peete E re ont are eee 29 GEOGRAPHIC VARIATION IN THE FREQUENCIES OF TRICHOME PHENOTYPES OF DATURA WRIGHTII AND CORRELATION WITH ANNUAL WATER DEFICIT ; J. Daniel Hare and Elizabeth Elle oan i, ne ee eRe Bo A New SECTION IN THE GOLDFIELD GENUS LASTHENIA (Compositar: HeLa SENSU LATO) ra C o Re “ K “fe eden silhjie) (Give) eee eee pane ne CSR ae Bibs SS 38 CORALLORHIZA MACULATA VAR. Ozerrensts ( (ORCHIDACEAR), A New Caled. Roor FROM COASTAL WASHINGTON LH Deal ahd Edward L. Tisch .... AG MEA aL ES, IRS 7 OE 40 & rs by a ay Ya B y ae Co OS pe V dig ‘ ~~ . + ct 4 re & at by és vi , i = at TEWORTHY INEVADA Scie hc A....\ih Ser a ate are Pedy KR Cie vee 43 : HAW, Abe (Nye 4 . 4. i , ye J )LLECTIONS LZ WINNS aN = 4) / / / 4 ey! \ x fi i "id WV I \t \ MA } Mix G7 \ Nik OK REVIEWS SAVANNAS, Bar RENS A aN D Rock O Dutcror PLAN Communrries OF NorTH AMERICA. EDITED 8 ROGER foi ANpeRson, Ji AMES FRALISH, ain M. BaAsKIN Dennis Odion Mig AY vous MLD SSS e ben soe eee Seer 44 \ \ TERRESTRIAL ECOREGIONS 0 ORT AMERICA: A ONSERVATION - Seg By AYLOR RICKETS, ERIC INERSTEIN, AVID OLSEN, OLBY OUCKS ET AL. T R s, Eric D D O Cou Todd Keeler-Wolf.............00606 Gy h She. ee 3 iy eA ceOn ee eer 45 \ : \ ' OUNCEMENTS BOTANY GRADS MEET AT CHICO STATE 2... MDW. .scsscssescsssessssessessescnsevecsesessecsenssaeees 48 TALKS PRESENTED AT THE 19TH CALIFORNIA BOTANICAL SOCIETY GRADUATE STUDENT: IVIBETINGSiicsauva nats costa casdueatianactes does ores ie use aeons ate eee eee 48 INE Wes PUIB ET OATION Seesee re a aati ss ptiela soe seat tee ea oe aura unt ter hs oacte ee A men ee 50 id PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL SOCIETY Maprono (ISSN 0024-9637) is published quarterly by the California Botanical Society, Inc., and is issued from the office of the Society, Herbaria, Life Sciences Building, University of California, Berkeley, CA 94720. Subscription information on inside back cover. Established 1916. Periodicals postage paid at Berkeley, CA, and additional mailing offices. Return requested. Postmaster: Send address changes to Maprono, Roy Buck, % University Herbarium, University of California, Berkeley, CA 94720. Editor—KrisTINA A. SCHIERENBECK California State University, Chico Department of Biology Chico, CA 95929-0515 kschierenbeck @csuchico.edu Editorial Assistant—JEANNIE TRIZZINO Book Editor—Jon E. KEELEY Noteworthy Collections Editors—DIETER WILKEN, MARGRIET WETHERWAX Board of Editors Class of: 2001—RobBert PATTERSON, San Francisco State University, San Francisco, CA PauLa M. ScHIFFMAN, California State University, Northridge, CA 2002—NorMAN ELLSTRAND, University of California, Riverside, CA Cara M. D’ Antonio, University of California, Berkeley, CA 2003—-FREDERICK ZECHMAN, California State University, Fresno, CA Jon E. KEELEY, U.S. Geological Service, Biological Resources Division, Three Rivers, CA 2004—-Davww M. Woon, California State University, Chico, CA INGRID PARKER, University of California, Santa Cruz, CA CALIFORNIA BOTANICAL SOCIETY, INC. OFFICERS FOR 2000-2001 President: BRUCE BALDwin, Jepson Herbarium and Dept. of Integrative Biology, 1001 Valley Life Sciences Bldg. #2465, University of California, Berkeley, CA 94720. First Vice President: Rop Myatt, San José State University, Dept. of Biol. Sciences, One Washington Square, San José, CA 95192. rmyatt @email.sjsu.edu Second Vice President: Ros ScuHLisinG, California State University, Chico, Dept. of Biol. Sciences, Chico, CA 95424. rschlising @csuchico.edu Recording Secretary: DEAN KeLcu, Jepson and University Herbarium, University of California, Berkeley, CA 94720. dkelch @sscl.berkeley.edu Corresponding Secretary: | SUSAN BAINBRIDGE, Jepson Herbarium, University of California, Berkeley, CA 94720. suebain @SSCL.berkeley.edu Treasurer: Roy Buck, % University Herbarium, University of California, Berkeley, CA 94720. The Council of the California Botanical Society comprises the officers listed above plus the immediate Past President, R. JoHN LitrLe, Sycamore Environmental Consultants, 6355 Riverside Blvd., Suite C, Sacramento, CA 95831; the Editor of MApRONo; three elected Council Members: Bian Tan, Strybing Arboretum, Golden Gate Park, San Francisco, CA 94122; James SHEvock, USDI National Park Service, Pacific West Region, 600 Harrison Street, Suite 600, San Francisco, CA 94107; DIANE ELaM, U.S. Fish and Wildlife Service, 3310 El Camino Avenue, Sacramento, CA 95825; Graduate Student Representative: KiRSTEN JOHANUS, Jepson Herbarium, University of California, Berkeley, CA 94720. This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). MADRONO, Vol. 48, No. 1, pp. 1-16, 2001 A CATALOGUE OF MOSSES FOR THE CITY AND COUNTY OF SAN FRANCISCO, CALIFORNIA JAMES R. SHEVOCK AND DAVID TOREN Department of Botany, California Academy of Sciences Golden Gate Park, San Francisco, CA 94118-4599 ABSTRACT The moss component for the bryoflora for the City and County of San Francisco contains 128 species representing over 22 percent of the California moss flora. Several mosses are suspected as being intro- duced since the mid-1900s and have subsequently become established and naturalized. A large component of the moss flora of San Francisco represents cosmopolitan species having wide geographical distributions. All but twelve of the 42 species based on the historic Bolander moss collections attributed to San Fran- cisco from the 1860s were relocated during this study. Sixty-five mosses are reported for the first time from the City and County of San Francisco. Of these, the rare species Triquetrella californica (Lesq.) Grout is also documented for San Francisco. Bryum pyriferum Crundw. & H. Whiteh., Campylopus pyriformis (E Schultz) Bridel, Heterocladium dimorphum (Brid.) Schimp. in B.S.G., Sematophyllum ad- natum (Michaux) Britton and Zygodon menziesii (Schwaegr.) Arnott are reported new for California and Zygodon menziesii is reported new for North America. What is today the city of San Francisco began at Mission Dolores when Spanish priests and soldiers established a mission on June 29, 1776. Since the California gold rush of 1849, San Francisco has functioned as a center for immigration, a focal point of commerce and as a financial center. Botanists too generally entered California via San Francisco (Slack 1993). Surrounded on three sides by the Pa- cific Ocean and San Francisco Bay, the City and County of San Francisco is contained at the tip of the San Francisco Peninsula, a compact rectangular area of 128 sq. km. (47 sq. mi.). Built on 43 hills, the City and County elevation ranges from sea level to 282.5 m. (927 ft.) at the summit of Mt. David- son. Twin Peaks at 281 m. and 275 m. along with Mt. Sutro at 277 m. are the highest points near the geographic center of the City and County of San Francisco. With a population of 770,000, San Fran- cisco is the second most densely populated Amer- ican city after New York. The Bay Area metropol- itan area is the Nation’s fifth largest and the work- day population of San Francisco swells to over 1.5 million (San Francisco Convention and Tourism Bureau). During the past 150 years San Francisco has un- dergone phenomenal development with an associ- ated loss of native habitats (Eastwood 1945, Howell 1934 and Howell, Raven & Rubtzoff 1958). As ear- ly as 1891, these landscape changes were already evident. In her introduction to the flora of San Fran- cisco, Brandegee (1891) states ‘‘that the number of plants is not greater is due not only to the destruc- tion of native species, caused by the cutting down of hills, filling of swamps and burying streams in- cident to the growth of a large city, but even more to the lack of variety in climate and conditions, the city being surrounded on three sides by water and swept by ocean winds.”’ While several portions of the City and County remained undeveloped up until the 1940s, it was in the aftermath of WW II that many natural wild areas were rapidly developed for residential use in San Francisco (Howell, Raven & Rubtzoff 1958). While much of the original land- scape was altered or converted to residential and high rise office buildings in subsequent decades, the City and County of San Francisco nonetheless still contain several areas of open space from land- scaped parks to undeveloped hills where remnants of native vegetation still persist. In other areas, non- native plantations dominated by Eucalyptus globu- lus Labill. and Cupressus macrocarpa Gordon (al- though the latter is native to other coastal areas of California) contributed to the replacement of the natural vegetation. These plantations, now ap- proaching 100 years old, provide many microhab- itats for the establishment of moss populations. History of moss collecting in San Francisco. Yer- ba Buena (the earlier name for San Francisco), was the northernmost expansion of Spanish rule in North America. There appear to be no moss col- lections obtained during the Spanish era between 1776 and 1821. Neither were moss collections made during the Mexican period from 1821 to 1846 when the United States declared war on Mexico. Yerba Buena was renamed San Francisco in 1847 and Alta California was transferred by treaty from Mexico to the United States in 1848. California was admitted to the Union as the 31st state in 1850. Moss collecting in San Francisco apparently began in the 1860s when Brewer collected Bryum argen- teum near Mission Dolores in 1862, Kellogg col- lected Alsia californica from Lone Mountain in 1866 and Bolander collected several mosses at Mis- sion Dolores and Fort Point in 1868. Earlier collec- tors of mosses in California include Sullivant in 2 MADRONO 1853 and Lesquereux in 1865 but label data are absent to conclude that any of their collections were obtained from San Francisco (Thiers & Emory 1992). Bolander, who was employed for a short time as a botanical assistant for the State Geologi- cal Survey, was the principal collector of mosses in California in the 1860s. His collections added significantly to the catalogue of mosses known from California (Lesquereux 1868 & Watson 1880), now evident in several taxa named to commemo- rate him. Brandegee (1891) consolidated the Bolander col- lections housed at the University Herbarium, Uni- versity of California and added a list of mosses to her San Francisco vascular plant flora. The moss flora was based on species concepts of that time and comprised 42 taxa. Bolander’s label informa- tion, although standard for that era, was very sparse. It is difficult to determine exactly where many of the San Francisco specimens were ob- tained beyond the few references to Mission Do- lores and Fort Point. There appears to be no pub- lished record to indicate the level of effort Bolander invested in his moss collecting endeavors within San Francisco. In addition to these early field workers, several other botanists, known nearly exclusively for their flowering plant collecting and floristic publications, also collected mosses in San Francisco during their careers. Collections were made by Alice Eastwood in 1923, by Peter Raven in 1949, 1950 and 1954, by Frederick J. Hermann in 1962 and by John Tho- mas Howell in 1969. Among Peter Raven’s first botanical collections were mosses obtained from Golden Gate Park when he was fourteen years old which he brought to the Academy of Sciences. Leo Koch and Fay MacFadden, the prominent field bryologists during this period, identified them. We undertook this study to determine how many of the mosses obtained by Bolander, as reported by Brandegee (1891), Lesquereux (1868), and Watson (1880), and from subsequent collectors, could be re-located within the City and County of San Fran- cisco after 130 years of substantial land-use changes. We also wanted to know if we could doc- ument mosses that were overlooked by Bolander and subsequent collectors and determine if addi- tional mosses, such as introduced exotics, are now becoming a naturalized component of the bryoflora of California. Our field collecting began in the fall of 1999 and concluded in the fall of 2000. Together we obtained 668 moss collections representing 112 species within the City and County of San Francis- co. Sixteen taxa reported by earlier collectors from San Francisco were not relocated during this study. Of this total, twelve were Bolander collections dat- ing from 1868. We believe that potential habitat still remains within the City and County for most of these taxa, even though we did not encounter them during our study. Five of these taxa are soil ephemerals that can be easily overlooked. Only [Vol. 48 three species, Fontinalis neomexicana, Orthotri- chum rivulare and Pohlia wahlenbergii appear to be extirpated from the City and County of San Francisco through loss of suitable habitat. Several mosses were routinely encountered on a variety of habitats and substrates while others appear to be locally uncommon or restricted. The number of col- lections cited for each species is a good approxi- mation of whether it is a common or a rare com- ponent of the moss flora in the City and County of San Francisco. Fragmentation of habitats and associated sub- strates available for moss establishment. Today, remnants of free-flowing creeks and fragments of native vegetation ranging from sand dunes and northern coastal scrub to serpentine grasslands can still be found in San Francisco. Salt and fresh water marshes have been drained and filled decades ago but are beginning to re-establish on fill land. In the areas where small lakes (ponds) remain, the natural flow of water has been altered along with great changes in water chemistry. Most of these bodies of water are no longer viable for the establishment of mosses, primarily a result of algal concentrations and/or presence of pollutants such as those in storm run-off from streets. Nearly all of the forested hab- itat is the result of tree plantations or trees that be- came established and spread from initial plantings that date from the 1870s to the 1920s (McClintock & Moore 1965). Several city parks are located throughout the City and County. The largest is Golden Gate Park at 407.5 hectares (1007 ac.). The Presidio and Golden Gate National Recreation Area, managed by the National Park Service provides a federal contribu- tion to open space in the City and County of San Francisco. The most protected landscape within the County of San Francisco is the Farallon Islands Na- tional Wildlife Refuge, managed by the U.S. Fish & Wildlife Service. These isolated, barren, wind- swept granitic islands are home to thousands of breeding oceanic birds and pinnipeds and are closed to public access. Today several conservation efforts are ongoing to either maintain or restore na- tive plant habitats from sand dunes and northern coastal scrub to serpentine grasslands. Management for endangered species is a catalyst for these vari- ous restoration efforts in San Francisco. Public de- bate will also focus on the long-term management of the extensive mature Eucalyptus plantations on Mt. Sutro, Mt. Davidson, and Monterey cypress plantations in the Presidio. Floristic Analysis. Most of the scientific names applied to the Bolander collections have since been modified based on either replacement of misapplied European moss names or changes in species con- cepts as well as generic placement. When the spe- cies circumscription differs from the name refer- enced by Brandegee (1891) Lesquereux (1868) or Watson (1880), we have provided the synonomy as | 2001) | SHEVOCK AND TOREN: SAN FRANCISCO MOSSES 3 far as we can determine it for that specimen. We have attempted to locate all of the Bolander collec- tions obtained from San Francisco. We also cite the herbaria and collectors where other San Francisco specimens are deposited. One feature of this spec- imen-based catalogue is to list for each taxon the earliest collection date found among herbarium re- cords examined. The California Academy of Sci- ences began operation in 1853. We speculate, how- ever, that some moss specimens from the Bolander era may also have been lost in the 1906 earthquake and fires that ravaged the City and damaged the California Academy of Sciences building which was at that time located on Market Street. Distribution of mosses within the City and Coun- ty of San Francisco is directly related to substrate preference and habitat availability (Table 1). The number of mosses occurring within the study area is, in our view, relatively species rich for such a highly urbanized area. We were, however, surprised by the number of common coastal species of the California bryoflora found in adjacent Marin and San Mateo counties that were not encountered dur- ing this study. We attribute these “‘missing taxa’’ to several factors. San Francisco seems to have lacked a native coast redwood forest similar to that found in the vicinity of Mt. Tamalpais in Marin County to the north or in the Santa Cruz Mountains region in Santa Cruz and San Mateo counties to the south. According to historic accounts trees were an un- common feature of the San Francisco landscape (Howell, Raven & Rubtzoff 1958). This lack of na- tive forests in San Francisco is most likely the re- sult of the strong winds that regularly channel through the San Francisco Bay and the extensive sand dune system that covered the western portion of San Francisco. Whatever native forest stands did exist were restricted to the higher peaks and then probably heavily utilized for both lumber and fuel wood consumption during the early years of growth of both Mission Dolores and the Presidio of San Francisco. Compared with other tree species along the coast, such as Acer macrophyllum Pursh, Alnus ru- bra Bong., Quercus spp., and Umbellularia cali- fornica (Hook & Arn.) Nutt., the bark of both Eu- calyptus and Cupressus appears to be limiting as a primary substrate supporting moss colonization. Where Eucalyptus dominates, its bark shedding characteristic coupled with a thick leaf litter can be quite an impediment to moss colonization that pre- fer soil or bark. Even where moss occurs on the trunks and bases of Eucalyptus, it is not widespread within a particular stand of trees. There appears to be a considerable amount of unoccupied habitat. The bark of Cupressus macrocarpa is well fur- rowed and thick, but we speculate that its density and composition does not readily permit moisture absorption. Even when wetted, the bark dries quickly. This would seem to explain why mosses are infrequently encountered on Monterey cypress except at the very base, among buttresses, or ex- posed roots. Bark chemistry may also be a factor in limiting moss colonization on Monterey cypress. Aside from these factors, the cypress canopies are nearly impervious to all but the heaviest rain, and this coupled with a outwardly descending branch arrangement moves moisture away from the trunk. The area beneath cypress trees is therefore dry much of the year. Stands of horticultural pines dom- inated by Monterey pine (Pinus radiata D. Don), while covered in lichens, generally lack mosses al- together. Again we speculate that the bark is not a suitable substrate for moss colonization within the climatic parameters present in San Francisco. Many wooded or forested areas are also choked with a dense understory of non-natives dominated by French broom (Genista monspessulana (L.) L. Johnson), English ivy (Hedera helix) L., Cape ivy (Delairea odorata Lem.), nasturtium (Tropaeolum majus L.) and/or Himalayan blackberry (Rubus dis- color Weihe & Nees). Only grasslands provide rel- atively open patches of gravelly or rocky soils where mosses have a chance to compete. Another major factor influencing moss colonization relates directly to the geological parent rock material. Franciscan chert, the most common rock type in San Francisco appears to be a poor substrate for the establishment of mosses due to its flaking and frac- turing characteristics. Other rock types are rare to uncommon within the City and County and that limits the potential for establishment of species that prefer a different rock or soil chemistry. For ex- ample, granitic rocks are only present at the Far- allon Islands and limestone is completely lacking. Only one rocky outcrop was found to contain enough lime to support a colony of Gymnostomum aeruginosum, an obligate calciphile. Besides rock and tree substrates that influence moss colonization, desiccation is also a factor due to the strong winds that frequently channel across the City from the Pa- cific Ocean. Several cosmopolitan species are quite successful in the City and County and can be found on a wide assortment of substrates such as concrete retaining walls, bricks, sidewalks, and even asphalt. Bryum, Didymodon, Grimmia, Syntrichia and Tortula spe- cies are relatively common on such substrates. Be- sides the widespread taxa, a few species in San Francisco are clearly elements of other bryofloras that we view as representing recent introductions to California. Exactly when these taxa were intro- duced remains speculative, but a few have been around for several decades and can be considered naturalized in California. These introduced mosses include Bryum pyriferum, Calliergonella cuspidata, Campylopus_ introflexus, Campylopus pyriformis, Campylopus subulatus, Heterocladium dimorphum, Sematophyllum adnatum, Vesicularia vesicularis and Zygodon menziesii. Of these taxa, Bryum pyr- iferum, Campylopus pyriformis, Heterocladium di- morphum, Sematophyllum adnatum and Zygodon 4 TABLE 1. GENERALIZED HABITAT PREFERENCES FOR MOSSES IN SAN FRANCISCO. Soul and gravels Atrichum undulatum Barbula convoluta Bryum argenteum Bryum bicolor Bryum canariense Bryum capillare Bryum gemmascens Bryum gemmilucens Bryum lisae Bryum pseudotriquetrum Bryum torquescens Ceratodon purpureus Claopodium whippleanum Dicranella heteromalla Dicranella howei Dicranella varia Didymodon insulanus Didymodon umbrosus Didymodon vinealis Ditrichum ambiguum Ditrichum schimperi Epipterygium tozeri Fissidens bryoides longifolius Fissidens crispus Fissidens curvatus Fissidens minutulus Fissidens sublimbatus Funaria hygrometrica Funaria muhlenbergii Homalothecium arenarium Kindbergia praelonga Pohlia nutans Polytrichastrum alpinum Polytrichum juniperinum Polytrichum piliferum Pseudocrossidium obtusulum Pseudotaxiphyllum elegans Scleropodium californicum Scleropodium cespitans Scleropodium julaceum Syntrichia amplexa Syntrichia bolanderi Timmiella anomala Timmiella crassinervis Tortula obtusifolia Triquetrella californica Weissia controversa Ephemeral mosses on soil (observed primarily during winter months) Acaulon rufescens Chenia leptophylla Entosthodon bolanderi Ephemerum serratum Fissidens bryoides longifolius Fissidens crispus Fissidens curvatus Fissidens minutulus Fissidens sublimbatus Funaria muhlenbergii Hennediella heimii Hennediella stanfordensis Microbryum starkeanum MADRONO TABLE |. CONTINUED. Phascum cuspidatum Pleuridium acuminatum Pleuridium sublatum Pseudocrossidium obtusulum Rock outcrops Anacolia menziensii Bartramia stricta Bryum canariense Bryum capillare Bryum pyriferum Campylopus introflexus Dicranella heteromalla Dicranoweisia cirrata Didymodon vinealis Grimmia laevigata Grimmia lisae Grimmia montana Grimmia pulvinata Grimmia trichophylla Homalothecium arenarium Homalothecium nuttallii Isothecium cristatum Kindbergia praelonga Plagiothecium laetum Polytrichum juniperinum Polytrichum piliferum Porotrichum bigelovii Pterogonium gracile Scleropodium californicum Syntrichia amplexa Syntrichia princeps Triquetrella californica Leaf litter Dicranum howellii Eurhynchium hians Homalothecium arenarium Kindbergia oregana Kindbergia praelonga Tree trunks, branches or exposed roots Alsia californica Amblystegium serpens Antitrichia californica Bryolawtonia vancouveriensis Bryum argenteum Bryum canariense Campylopous subulatus Dicranoweisia cirrata Dicranum fuscescens Heterocladium dimorphum Isothecium cristatum Isothecium myosuroides TIsothecium spiculiferum Orthotrichum consimile Orthotrichum diaphanum Orthotrichum lyellii Orthotrichum tenellum Pterogonium gracile Syntrichia pagorum Syntrichia princeps Tortula papillosa Tortula plinthobia Zygodon menziesii Zygodon rupestris | | | | f / 2001) |TABLE 1. CONTINUED. } \ Rotten wood and logs Amblystegium serpens Aulacomnium androgynum Dicranoweisia cirrata Isothecium cristatum Isothecium myosuroides Isothecium spiculiferum Kindbergia oregana Kindbergia praelonga Orthodicranum tauricum Plagiothecium laetum Sematophyllum adnatum Seeps and wet places Amblystegium juratzkanum Amblystegium serpens Brachythecium asperrimum Brachythecium rutabulum Brachythecium salebrosum Bryum capillare Crumia latifolia Didymodon tophaceus Drepanocladus aduncus Fontinalis neomexicana Gymnostomum aeruginosum Hygroamblystegium tenax Kindbergia oregana Kindbergia praelonga Leptobryum pyriforme Leptodictyum riparium Orthotrichum rivulare Pohlia wahlenbergii Porotrichum bigelovii Scleropodium cespitans Scleropodium colpophyllum Scleropodium julaceum Scleropodium obtusifolium Scleropodium touretii Vesicularia vesicularis Lawns Amblystegium serpens Brachythecium asperrimum Brachythecium rutabulum Calliergonella cuspidata Campylopus pyriformis Kindbergia praelonga Leptodictyum riparium Rhytidiadelphus squarrosus Scleropodium californicum Scleropodium cespitans Scleropodium colpophyllum Concrete walls, bricks, sidewalks, buildings, asphalt Bryum argenteum Bryum capillare Ceratodon purpureus Didymodon australasiae Didymodon nicholsonii Didymodon rigidulus Didymodon umbrosus Didymodon vinealis Funaria hygrometrica Grimmia pulvinata Homalothecium nuttallii Isothecium cristatum SHEVOCK AND TOREN: SAN FRANCISCO MOSSES 5 TABLE 1. CONTINUED. Orthotrichum diaphanum Orthotrichum tenellum Philonotis capillaris Syntrichia ruralis Syntrichia princeps Tortula atrovirens Tortula brevipes Tortula latifolia Tortula muralis Tortula obtusifolia Tortula papillosa menziesii are new additions to the California bry- oflora. Zygodon menziesii, native to Chile, Australia and New Zealand is reported here for the first time in North America. Considering the number of po- tential introduction sites from the horticultural trade and the diversity of cultivated plants imported from all over the world to San Francisco, it 1s somewhat remarkable that so few exotic mosses have actually become established and naturalized. It seems likely that both habitat and substrate availability coupled with climatic factors are indeed real barriers to moss establishment in San Francisco. Mosses not accustomed to a Mediterranean-type environment with growth primarily restricted to the cool wet winter season followed by an extended summer drought period are unlikely to become a naturalized component of the California bryoflora. The four highest peaks in the City also form a rain shadow effect that is quite evident during the summer fog periods. The Eucalyptus plantations on Mt. Sutro and Mt. Davidson during the summer months are commonly engulfed in fog while the eastern portion of the City remains relatively sunny and dry. The associated fog drip may account for the diversity of mosses encountered on these two peaks within the City and County of San Francisco. ACKNOWLEDGMENTS We thank the San Francisco Recreation and Parks De- partment, Presidio Trust, Golden Gate National Recreation Area, National Park Service and Farallon National Wild- life Refuge, U.S. Fish & Wildlife Service for providing collecting permits to sample various habitats for this study. Vouchers of our moss collections are placed in the herbarium of the Department of Botany, California Acad- emy of Sciences, Golden Gate Park (CAS). In addition, we have also cited the herbaria where duplicates have been deposited, primarily MO, NY, and UC. Jake Sigg and Peter Holloran provided habitat comments to an early draft of this paper and Steve Jessup and Brent Mishler, as peer reviewers, provided several comments that greatly improved the final product. We also thank several bryol- ogists for providing specimen identifications or verifying our determinations. Special thanks to Drs. Jan-Peter Frahm (Campylopus), Brent Mishler (Syntrichia & Tor- tula), Jesus Munoz (Grimmia), Ronald Pursell (Fissidens), John Spence (Bryum), Dale Vitt (Orthotrichum & Zygo- don) and Richard Zander (Pottiaceae). These various con- tributions are referenced directly with the specimen cita- 6 MADRONO tion in this catalogue. Finally, we thank Dan Norris who has been extremely helpful and supportive throughout this project. He provided determinations or confirmations for many specimens in this catalogue and he has inspired us to work on moss floristics. LITERATURE CITED BRANDEGEE, K. 1891. A list of the mosses of San Fran- cisco, California. [an appendix to flowering plants and ferns of San Francisco]. Zoe 2: 384-386. Eastwoop, A. 1945. The wildflower gardens of San Fran- cisco in the 1890s. Leafl. West. Bot. 4: 153-156. Howe LL, J. T. 1934. Field days in San Francisco. |. Leafl. West. Bot. 1: 89-91. , P. H. RAVEN AND P. RustTzorr. 1958. A flora of San Francisco, California. Wasmann J. of Biol. 16(1): 1-157. LESQUEREUX, L. 1868. Catalogue of Pacific Coast mosses. Mem. Cal. Acad. Sci. 1 part 1. San Francisco. 38 pgs. McCLINTOCK, E. AND V. Moore. 1965. Trees of the Pan- handle, Golden Gate Park, San Francisco. Cal. Acad. Sci., Misc. Paper No. 1. 56 pps. San Francisco. SLACK, N. G. 1993. Botanical exploration of California, chapter 10. In: Miller, S.M. (ed.), John Muir Life and Work. Univ. New Mexico Press. STEERE, W. C., L. E. ANDERSON AND V. S. BRYAN. 1954. Chromosome studies in California mosses. Mem. Torr. Bot. Club 20(4) 1-75. [chromosome counts made for two mosses obtained from San Francisco]. SYED, H. 1973. A taxonomic study of Bryum capillare Hedw. and related species. J. Bryol. 7(3): 265-326. [cites a historic collection of Bryum torquescens ob- tained from San Francisco]. THIERS, B.M. AND K.S.G. Emory. 1992, The history of bryology in California. The Bryologist 95: 68—78. WATSON, S. 1880. Botany of California. Musci. In: Vol. II. Geol. Survey of California, pgs. 353—423. Wiley & Sons, Univ. Press, Cambridge. WELCH, W. H.. 1960. A monograph of the Fontinalaceae. Martins Nijhoff. The Hague. 357 pps. [cites a historic collection of Fontinalis neomexicana from San Fran- cisco]. CATALOGUE OF MOSSES Acaulon rufescens Jaeg. [= Sphaerangium muticum (Brandegee 1891 & Lesquereux 1868)]. First record: 1868. On exposed soil. Grounds and fields of meadows, Mission Dolores, Bolander s.n. (UC) [confirmed by Toren]. Although habitat for this species still remains in San Francisco, we did not encounter it during this study. Alsia californica (Hook. & Arnott) Sull. First rec- ord: 1866. On branches and trunks of both hard- wood and conifer trees, especially Quercus, Pit- tosporum, Salix, Eucalyptus and Cupressus. Mis- sion Dolores, Kellogg s.n. (CAS); The Dell, Lone Mountain, Kellogg s.n. (CAS); San Francisco, Bolander s.n. (UC); Golden Gate Heights Park, Toren 7942 (CAS) & Shevock 19200 (CAS, UC); Glen Canyon, San Miguel Hills, Shevock ISS878 & 18880 (CAS, UC); Golden Gate Park: Botanical Garden, Hermann 17445 (CAS), near middle lake of Chain of Lakes, Shevock 19210 (CAS, UC), near Stow Lake, Shevock 18761 & | | [Vol. 48) 18765 (CAS, MO, UC); near Elk Glen Lake, Shevock 18894 (CAS, MO, UC), Cross, Shevock 18804 (CAS, UC); Panhandle,, Toren 7801 (CAS) and near Conservatory of. Flowers, Toren 7747 (CAS) & Shevock 18802| (CAS, UC); Brotherhood Way near Lake Mer-' ced, Shevock 19286 (CAS, UC); Mt. Sutro, Shev-| ock 19156 (CAS, UC); Lafayette Park, Shevoem 19249 (CAS, UC). | Amblystegium juratzkanum Schimp. First record:) 1953. Golden Gate Park, Japanese Tea Garden, V. S. Bryan 33 (herb. not stated) [specimen cited. in Steere, Anderson & Bryan 1954]. | Amblystegium serpens (Hedw.) Schimp. in B.S.G.! First record: 1962. On lawns, base of tree trunks | and occasionally concrete retaining walls. Gold-. en Gate Heights Park, Toren 7946 (CAS); El Po-. lin Loop, Presidio of San Francisco, Shevock 19230 (CAS, UC); Lafayette Park, Shevock 19238, 19239 & 19248 (CAS, UC); Jefferson Square, Shevock I1S898 (CAS, UC); Alta Plaza, | Shevock 19255 (CAS, UC); Mission Dolores | Park, Shevock 18989 & 19991 (CAS, UC); Wal-. ton Park, Shevock 19305 (CAS, UC); Embarcad- | ero near Pier 29 & 31, Shevock 19105 (CAS, | MO, UC); Lone Mountain, University of San. Francisco, Shevock 19339 (CAS, UC); Golden | Gate Park: Koch 1635 & 2308 (UC), Wagner s.n. (UC), Norris 97033 (UC); Strybing Botanical | Garden, Hermann 17444 (CAS) & Shevock | 19514 (CAS, UC), Conservatory of Flowers, | Shevock 19307 & 19308 (CAS, UC), Panhandle, | Toren 7800 (CAS), Shevock 18764 (CAS, UC), Shevock 18774 (CAS, MO, UC), Shevock 18933 (CAS, UC), De Laveaga Dell, Aids Memorial Grove, Shevock 19524 (CAS, UC); Golden Gate | National Recreation Area: Fort Mason, Shevock 19091 (CAS, UC) and Aquatic Park, Shevock | 19096 (CAS, UC); Third at Burke Streets, Shev- ock 20204 (CAS, UC). | Anacolia menziesii (Turn.) Par. First record: 2000. | On soil over rock. Open space (future park) at Castro and 30" Streets. Toren SO20 (CAS). Antitrichia californica Sull. in Lesq. First record: 2000. On concrete retaining wall. Generally found on tree trunks or boulders. Yerba Buena Island, Shevock 19111 (CAS, UC); Alcatraz Is- land, Golden Gate National Recreation Area, Shevock, Toren & Thomas 20210a (CAS). Atrichum undulatum (Hedw.) P. Beauv. [Lesquer- eux 1868]. First record: 1868. On exposed min- eral soils. Mission Dolores, Bolander s.n. (not located at UC). Although habitat exists for this moss in the City especially on Mt. Davidson and Mt. Sutro, we did not encounter it during our study. Aulacomnium androgynum (Hedw.) Schwaegr. First record: 2000. On rotten wood and logs. Golden Gate Park near Strawberry Hill, Shevock 19161 (CAS, MO, UC) and Chain of Lakes, Shevock 19207 (CAS, UC); Mt. Sutro, Toren 7749 (CAS). 2001] Barbula convoluta Hedw. [Lesquereux 1868]. First record: 1868. On exposed soil. In gardens, San Francisco, Bolander s.n. (UC); Stow Lake, Gold- en Gate Park, Koch 2310 (UC); De Laveaga Dell, Aids Memorial Grove, Shevock 19517 (CAS, UC); along California Academy of Sci- ences building, Shevock 18752 (CAS, UC); Twin Peaks, Eastwood 57 (MO) [det. by Zander] and Eastwood 59 (CAS) [det. by Bartram]; Mt. Sutro, Toren 7773 (CAS); Cesar Chavez Street at high- way 101, Toren 7816 (CAS); Below overpass of highway 101 near Pet Cemetery, Presidio of San Francisco, Shevock 19365 (CAS, UC). Barbula unguiculata Hedw. First record: 2000. On exposed soil. Grand View below Twin Peaks, To- ren 7743 (CAS); Golden Gate Park and 41st Av- enue and Lincoln Way, Toren S023 (CAS); West Pacific Street near Arguello Gate, Presidio of San Francisco, Shevock 19360 (CAS, UC). Bartramia stricta Brid. [Lesquereux 1868]. First record: 1868. On over rock. On moist rocks and ground around San Francisco, Bolander §s.n. (UC); Glen Canyon, San Miguel Hills, Shevock 18884 (CAS, UC); O’Shaughnessy Blvd., Glen Canyon, Toren 7812 (CAS). Brachythecium asperrimum (C. Mull.) Sull. First record: 1975. On lawns and grassy areas. Mary Ward Hall, San Francisco State University, Showers 2852 (SFSU) [confirmed by Toren]; Golden Gate Park near Chain of Lakes, Shevock 19163 (CAS, UC); Mt Sutro, Shevock 19185 (CAS, UC); Turk and Mason Streets, J. 7. How- ell s.n. (CAS). Brachythecium rutabulum (Hedw.) Schimp. in B. S. G. First record: 2000. On lawns and grassy areas. Near Pet Cemetery below highway 101, Presidio of San Francisco, Shevock 19367 (CAS, UC) [det. by Norris]. Brachythecium salebrosum (Web. & Mohr.) Schimp. in B. S. G. First record: 2000. On soil with seep or wet areas over rock. Cesar Chavez Blvd. At highway 101, Shevock 19276 (CAS, UC) [det. by Norris]. Bryolawtonia vancouveriensis (Kindb. in Mac.) Norris & Enroth First record: 2000. On base of tree trunks or exposed roots. Golden Gate Heights Park, Toren 7943 (CAS) & Shevock 19202 (CAS, MO, NY, UC); Golden Gate Park, De Laveaga Dell, Aids Memorial Grove, Shev- ock 19521 (CAS, UC). Bryum argenteum Hedw. [Lesquereux 1868]. First record: 1862. On exposed disturbed soils, con- crete walls or base of tree trunks. Between Mis- sion Dolores and the sea, Brewer 898 (CAS, UC); Common around San Francisco, Bolander s.n.(UC); near Sutrowood and Stanyan Street, Eastwood 55 (CAS); near Lake Merced, Koch 1633 (UC); Jefferson Square, Shevock 18903 (CAS); Bay View Park west of Candlestick, To- ren 7834 (CAS); Aqua Vista Park, Shevock 18994 (CAS, MO, UC); Balboa Park, Shevock SHEVOCK AND TOREN: SAN FRANCISCO MOSSES 7 19280 (CAS, UC); Lafayette Park, Shevock 19237 (CAS, MO, UC); Third and Burke Streets, Shevock 20205 (CAS, UC). Bryum bicolor Dicks. [= Bryum californicum (Brandegee 1891 & Lesquereux 1868), B. occi- dentale (Brandegee 1891) and = Bryum dicho- tomum|]. First record: 1868. On soil. Common around San Francisco, on the ground in grassy places, Bolander s.n. (UC); sand dunes, Olson- Seffer s.n. (UC) [det. by Koch; confirmed by To- ren]; Mt. Sutro, Toren 7774 (CAS); Bay View Park west of Candlestick, Toren 7832 (CAS); Alta Plaza, Shevock 19256 (CAS, UC); Lafayette Park, Shevock 19234 (CAS, UC). Bryum canariense Brid. First record: 2000. On soil, occasionally on rock outcrops or base of tree trunks. Park Blvd., entrance of Fort Winfield Scott, Presidio of San Francisco, Shevock 19369 (CAS, UC); Golden Gate National Recreation Area: Fort Miley, Shevock 19372 (CAS, UC) & Fort Funston, Shevock 19159 (CAS, UC); Cesar Chavez Blvd. At highway 101, Shevock 19277 (CAS, UC); Twin Peaks, Shevock 18864 (CAS, UC); Southeast Farallon Island, Farallon Nation- al Wildlife Refuge, Shevock 20147 & 20151 (CAS, UC); Corona Heights, Shevock 1882] (CAS, UC); Bay View Park, Shevock 19020 (CAS, UC); Twin Peaks, Shevock 18869 (CAS, UC); Yerba Buena Island, Shevock 19117 (CAS, MO, UC); Alamo Square, Shevock 18906 (CAS, UC). Bryum capillare Hedw. First record: 1868. On soil, rock outcrops or concrete retaining walls. Sandy soils near coast, Bolander s.n. (UC); Golden Gate Park, Wagner 2363 (UC) [det. by Koch; confirmed by Toren]; Alamo Square, Shevock 18913 (CAS, UC); Paramount Terrace off Stan- yan Street, Toren 7752 (CAS); Douglass Play- ground, Toren 7782 (CAS); between Marietta Drive and O’Shaughnessy Blvd., Toren 7807 (CAS); Bay View Park west of Candlestick, To- ren 7829a (CAS) & Shevock 19014 (CAS, UC); Potrero Hill, Toren 7825 (CAS); Brotherhood Way near Lake Merced Blvd., Toren 7946b (CAS); Glen Canyon, San Miguel Hills, Shevock 18889 (CAS, UC); Bernal Heights Park, Shevock 19126 (CAS, UC); 14th Avenue and Ortega Street, Shevock 19144 (CAS, UC); Stern Grove and Pine Lake Park, Shevock 19158 (CAS, MO, UC); Mt. Sutro, Shevock 19172 (CAS, UC); Alta Plaza, Shevock 19259 (CAS, UC); Golden Gate Heights Park, Shevock 19196 (CAS, UC); Pre- sidio of San Francisco: Fort Point, Shevock 18941 (CAS, UC) and near Pet Cemetery below highway 101, Shevock 19222 (CAS, UC); Alca- traz Island, Golden Gate National Recreation Area, Shevock, Toren & Thomas 20210 (CAS, UC). Bryum gemmascens Kindb. First record: 2000. On sou, rock and gravels. Bernal Heights Park, To- ren 7792 (CAS); Cesar Chavez Blvd. at highway 8 MADRONO 101, Toren 7818 (CAS); Castro and 30th Streets, Toren 8017 (CAS); Glen Canyon Park, San Mi- guel Hills, Shevock 18887 (CAS, UC); Aqua Vis- ta Park, Shevock 18995 (CAS, UC); Fort Point, Presidio of San Francisco, Shevock 18937 (CAS, UC). Bryum gemmilucens Wilcz. & Dem. First record: 2000. On soil and gravels. Aqua Vista Park, Shevock 18993 (CAS, UC) [det. by Norris]. Bryum lisae De Not. First record: 2000. On rocky soils, retaining walls and bricks. Turk and Fill- more Streets, Shevock 19343 (CAS, UC); Stern Grove and Pine Lake Park, Shevock 19153 (CAS, UC); Twin Peaks, Shevock 18870 (CAS, UC); Bay View Park, Shevock 19011 (CAS, UC); Pacific Street, Presidio of San Francisco, Shevock 19353 (CAS, UC); Golden Gate Nation- al Recreation Area: China Beach, Shevock 19274 (CAS, UC) and between Painted Rock Cliffs and Sutro Park, Shevock 18807 (CAS, UC). Bryum pseudotriquetrum (Hedw.) Gaertn., Meyer & Scherb. First record: 2000. On soil. Golden Gate Park between Middle and South Lake of Chain of Lakes west of the Polo Field, Shevock 19206 (CAS, UC). Bryum pyriferum Crundw. & H. Whiteh. First rec- ord: 2000. On rock wall. Alcatraz Island, Golden Gate National Recreation Area, Toren, Shevock & Thomas 8459 (CAS) [det. confirmed by John Spence]. This collection represents the first doc- umented occurrence of this species for Califor- nia. Bryum torquescens Bruch First record: 1947. On soil. Golden Gate Park, Koch s.n. (MICH) [cited by Syed 1973]; Lone Mountain, University of San Francisco, Shevock 19334 (CAS, UC) [det. by Norris]. Calliergonella cuspidata (Hedw.) Loeske First rec- ord: 1969. In lawns and grassy areas. Fort Ma- son, Golden Gate National Recreation Area, J. T. Howell s.n. (CAS); A lawn weed at San Francis- co State University, Showers 450 (SFSU, UC); Mt. Sutro, Toren 7771 (CAS); McLaren Park, Shevock 19136 (CAS, MO, UC); Potrero del Sol Park, Shevock 19278 (CAS, UC); Walton Park, Shevock 19304 (CAS, UC). Campylopus introflexus (Hedw.) Brid. First record: 2000. On rock outcrops with soil. Mt. Davidson, Toren 7760 (BONN, CAS) [det by J-P. Frahm]. Campylopus pyriformis (FE Schultz) Bridel First rec- ord: 2000. On soil of lawn. Embarcadero at Chestnut Street across from Pier 29 & 31, Shev- ock 19102 (BONN, CAS, MO, NY, UC) [det. by J-P. Frahm]. This collection represents the first documented occurrence of this species for Cali- fornia. Campylopus subulatus Schimp. in Rabenh. First record: 2000. Base of tree trunks and on clayey soils. Panhandle, Golden Gate Park, Shevock 18769 (BONN, CAS, MO, UC) & 18772 (BONN, CAS, NY, UC); Mt. Davidson, Shevock [Vol. 48 19291 (BONN, CAS, MO, UC) [dets. by J-P Frahm]. Ceratodon purpureus (Hedw.) Brid. First record: 1868. On soils and gravels, rock outcrops and occasionally at base of tree trunks. Mission Do- lores, Bolander s.n. (UC); Sutrowood and Stan- yan Street, Eastwood 57 (CAS); Tank Hill, Shev- ock 18919 (CAS, MO, UC); Twin Peaks, Shev- ock 18562 & 19142 (CAS, UC), Toren 7745 (CAS) and Raven 21 (CAS); between Marietta Drive and O’Shaughnessy Blvd., Toren 7813 (CAS); Bay View Park, Shevock 19009 (CAS, MO, UC); Mt. Davidson, Shevock 19289 (CAS, UC); Mt. Sutro, Shevock 19194 (CAS, UC) & 19183 (CAS, UC); 14th Avenue at Ortega Street, Shevock 19148 (CAS, UC); Golden Gate Heights Park, Shevock 19203 (CAS, UC); Corona Heights Park, Shevock 18824 (CAS, UC); Em- barcadero near Piers 29 & 31, Shevock 19104 (CAS, UC); McLaren Park, Shevock 19135 (CAS, MO, UC); Coastal Bluff Trail, Golden Gate National Recreation Area, Shevock 18805 (CAS, UC) & 18810 (CAS, MO, UC); Walton Park, Financial District, Shevock 19301 (CAS, UC); Stanyan & Fulton Streets, Golden Gate Park, Shevock 18837 (CAS, UC); Cesar Chavez Blvd. At highway 101, Toren 782] (CAS). Chenia leptophylla (C. Mull.) Zand. First record: 1977. On exposed soils. Golden Gate Park near 41st Avenue and Lincoln Blvd., Toren 2921] (CAS, SFSU). Claopodium whippleanum (Sull. in Whipple & Ives) Ren. & Card. First record: 2000. On ex- posed soils and rock underhangs. Mt. Sutro, Shevock 19167 (CAS, MO, UC) & Shevock 19179a (CAS, UC); Mt. Davidson, Shevock 19289 (CAS, MO, UC) & 19297 (CAS, UC). Crumia latifolia (Kindb. in Mac.) Schof. First rec- ord: 2000. On calcareous seeps. Coastal Bluff Trail, Golden Gate National Recreation Area, Shevock I8815 (CAS, UC); Sutro Bath Site, Golden Gate National Recreation Area, Shevock 19214 (CAS, UC). Dicranella heteromalla (Hedw.) Schimp. First rec- ord: 2000. On clayey soils and crevices of rock outcrops. Mt. Davidson, Shevock 15928 (CAS, MO, UC), Shevock 19258 (CAS, UC), 19292 (CAS, MO, UC) and Toren 7765 (CAS); Mt. Su- tro, Shevock 19171 (CAS, UC); Yerba Buena Is- land, Toren 8054 (CAS); Fort Point, Presidio of San Francisco, Shevock 18936 (CAS, UC); Southeast Farallon Island, Farallon National Wildlife Refuge, Shevock 20148 (CAS, UC). Dicranella howei Ren. & Card. First record: 1868. On clayey soils. Bay of San Francisco, Bolander s.n. (UC) [det. by Toren]; Coastal Bluff Trail, Golden Gate National Recreation Area, Shevock 18806 (CAS, MO, UC); McLaren Park, Shevock 19138 (CAS, UC). Dicranella varia (Hedw.) Schimp. First record: | 2001] 2000. On clayey soils. Bay View Park west of Candlestick, Toren 7839 (CAS). Dicranoweisia cirrata (Hedw.) Lindb. ex Mulde First record: 1954. On tree trunks, exposed roots and rock outcrops. Mt. Davidson, Toren 7770 (CAS); Golden Gate Heights Park, Toren 7941 (CAS); 14th Avenue and Ortega Street, Shevock 19149 (CAS, UC); Lincoln Park, Golden Gate National Recreation Area, Shevock 19268 (CAS, MO, UC): Mt. Sutro, Shevock 19180 (CAS, UC): Presidio of San Francisco, Shevock 19355 (CAS, UC); Panhandle section of Golden Gate Park, Shevock 20287 (CAS, UC) and horseshoe court near Stanyan and Fulton Sts., Raven s.n. (CAS) [det. by MacFadden; confirmed by Toren]. Dicranum fuscescens Turn. First record: 2000. On tree trunks. Mt. Sutro, Shevock 19/92 (CAS, UC). Dicranum howellii Ren. & Card. First record: 2000. On soil and leaf litter. Mt. Davidson, Shevock 19290 (CAS, MO, NY, UC) and Toren 7767 (CAS). Didymodon australasiae (Grev. & Hook.) Zand. First record: 2000. On concrete wall of buildings. Golden Gate Park, South Windmill, Shevock 18896 (BUF CAS, UC) [confirmed by Zander]. Didymodon insulanus (De Not.) M. Hill First rec- ord: 2000. On sandy soils. 14th Avenue and Or- tega Street, Shevock 19145 (CAS, UC) [det. by Norris]. Didymodon nicholsonii Culm. First record: 2000. On cement brick wall. Golden Gate Park at Wil- lard North and Fulton Streets, Shevock 18833 (CAS, UC) [det. by Norris]. Didymodon rigidulus Hedw. First record: 2000. On asphalt walkways. Golden Gate Park, De Lav- eaga Dell, Aids Memorial Grove, Shevock 19515 & 20283 (CAS, UC). Didymodon tophaceus (Brid.) Lisa [= Trichosto- mum tophaceum (Lesquereux 1868)]. First rec- ord: 1868. On wet soils and seeps generally with some salts. Fort Point (base of the Golden Gate Bridge), Presidio of San Francisco: Bolander s.n. (UC) & Howe 55 (UC), Shevock 18516 & 15934 (CAS, UC), Toren 7742 (CAS), Baker Beach, Shevock 18950 & 19152 (CAS, UC) and El Polin Spring, Shevock 19231 (CAS, UC); Bay View Park west of Candlestick, Toren 784] (CAS) and Shevock 19007 (CAS, MO, NY, UC); McLaren Park, Shevock 19137 (CAS, UC); Lake Merced, Raven 3 (CAS); Sutro Baths site, Golden Gate National Recreation Area, Shevock 19211 & 19213 (CAS, UC), China Beach, Shevock 19271 & 19272 (CAS, UC) and Lands End, Eastwood s.n. (CAS) [det. by Koch; confirmed by Toren]; Stow Lake, Golden Gate Park, Shevock 18756 (CAS, UC). Didymodon umbrosus (C. Mull.) Zand. First record: 2000. On compacted soils over concrete or as- phalt. Aqua Vista Park, Central Basin, Shevock SHEVOCK AND TOREN: SAN FRANCISCO MOSSES 2 18997 (BUE CAS, MO, UC) [det. by Zander] & Shevock 18992 (CAS, UC). Didymodon vinealis (Brid.) Zand. [= Barbula cy- lindrica, B. flexifolia, B. subfallax, B. vinealis & B. virescens (Brandegee 1891 & Lesquereux 1868)]. First record: 1868. On soil, rock out- crops, concrete retaining walls and buildings. Presidio of San Francisco: Fort Point, Bolander g.n. (UC) & Shevock 18938 (CAS, UC), Baker Beach, Shevock 18951 (CAS), Shevock 18946 & 18947 (CAS, UC), near Pet Cemetery below highway 101, Shevock 19223 (CAS, UC) and Ju- lius Kahn Playground near Locust Street, Shev- ock 19351 (CAS, UC); Mission Dolores, Howe 5OI (CAS) and Shevock 20282 (CAS, UC); Franklin and Pacific Streets, Raven s.n. (CAS); Turk & Fillmore Streets, Shevock 19341 (CAS, UC); Mt. Davidson, Shevock 19296 (CAS, UC); Mt. Sutro, Shevock 1918S (CAS, UC); Buena Vista Park, Shevock 18914 (CAS, MO, UC); Po- trero Hill, Toren 7826 (CAS), Shevock 19001 & 19004 (CAS, UC); Twin Peaks, Shevock 18865 (CAS, UC); Douglass Playground, Toren 7781] (CAS); Bernal Heights Park, Toren 7788 (CAS) and Shevock 19127 (CAS, UC); Bay View Park west of Candlestick, Toren 7838 (CAS); open space at 30th and Castro Street, Shevock 19346 (CAS, UC); Yerba Buena Island, Shevock 191715, I91IS & 20248 (CAS, UC); Kite Hill Park, Shevock 19140 (CAS, UC); Stern Grove and Pine Lake Park, Shevock 19150 & 19155 (CAS, UC); Golden Gate Heights Park, Shevock 19197 (CAS, UC); Near Stanyan and Fulton Streets, Raven s.n. (CAS); Golden Gate Park: Lily Lake, Shevock 18820 (CAS, UC); Mallard Lake, Ob- erlander s.n. (SFSU) [det. by Toren]; De Lav- eaga Dell, Aids Memorial Grove, Shevock 19525 (CAS, UC), California Academy of Sciences building, Shevock 18751 (CAS, UC); Golden Gate National Recreation Area: Fort Miley, USS San Francisco Memorial, Shevock 19370 (CAS, UC); Lincoln Park, Shevock 19265 & 19269 (CAS, UC), Fort Mason, Shevock 19093 (CAS, UC); Alcatraz Island, Golden Gate National Rec- reation Area, Steen s.n. (UC), Toren 5469 (CAS) and Shevock, Toren & Thomas 20211 & 20215 (CAS, UC). Ditrichum ambiguum Best First record: 2000. On soul over rock. Mt. Davidson, Toren 7SOS (CAS). Ditrichum schimperi (Lesq.) Kuntze First record: 2000. On soil over rock. Mt. Sutro, Toren 7748 (CAS); Corona Heights, Shevock 18882 (CAS, UC); Mt. Davidson, Shevock 15926 & 19299 (CAS, UC); Glen Canyon, San Miguel Hills, Shevock I8890 (CAS, MO, UC); Corona Heights, Shevock 18822 (CAS, MO, UC); Yerba Buena Island, Toren SOS7 (CAS). Drepanocladus aduncus (Hedw.) Warnst. [= Hyp- num aduncum (Brandegee 1891 & Lesquereux 1868)]. First record: 1868. On wet lawns and sat- urated soils. Swamps near San Francisco, Bolan- 10 MADRONO der s.n. (UC); Stonestown off of 19th Avenue, Toren SO6I (CAS); Diamond Heights, 28th and Douglass Streets, Toren 8458 (CAS). Entosthodon bolanderi Lesq. [Lesquereux (1868)] First record: 1868. On soil. Clayey soil banks of the bay of San Francisco, Bolander s.n. (UC). Although habitat for this species still remains in San Francisco, we did not encounter this ephem- eral species during our study. Ephemerum serratum (Hedw.) Hampe [Lesquereux (1868)]. First record: 1868. On exposed soil. On the grounds and fields of meadows, Mission Do- lores, Bolander s.n. (DUKE, UC) [confirmed by Toren]. Although habitat for this species still re- mains in San Francisco, we did not encounter this ephemeral species during our study. Epipterygium tozeri (Grev.) Lindb. [= Webera tozert (Brandegee 1891); Bryum tozeri (Les- quereux 1868)]. First record: 1868. On shaded clayey soils and rock underhangs. Borders of roads and ditches, San Francisco, Bolander s.n. (UC); Corona Heights, Shevock 18831 (CAS, MO, UC); Golden Gate Park, Shevock 18836 (CAS, MO, UC); between Marietta Drive and O’Shaughnessy Blvd., Toren 78/11 (CAS); Douglass Playground, Toren 7757 (CAS); Yer- ba Buena Island, Shevock 19124 (CAS, MO, UC); Bernal Heights Park, Shevock 19131 (CAS, UC); Fort Mason, Golden Gate National Recreation Area, Shevock 19094 (CAS, MO, UC); Arguello at Washington Blvd., Presidio of San Francisco, Shevock 19228 (CAS, MO, UC); Mt. Davidson, Shevock 19298 (CAS, UC); Mt. Sutro, Shevock 19178 (CAS, MO, UC); Alcatraz Island, Golden Gate National Recreation Area, Toren S468 (CAS). Eurhynchium hians (Hedw.) Sande Lac. First rec- ord: 2000. On damp soil with leaf litter. Lone Mountain, Shevock 19337 (CAS, UC) [det. by Norris]. Fissidens bryoides Hedw. var. longifolius (Brid.) Hampe [= F. bryoides var. viridulus (Swartz) Brotherus]. First record: 1962. On shaded clayey soils. Golden Gate Park, De Laveaga Dell, Aids Memorial Grove, Hermann 17449 (CAS) [det. by Pursell] Fissidens crispus Mont. [|= Fissidens limbatus Sull. (Brandegee 1891 & Lesquereux 1868)]. First record: 1868. On shaded clayey soils. Common around San Francisco, Bolander s.n. (UC) [det. confirmed by Pursell]; Golden Gate Park near Conservatory of Flowers, Toren 7746 (CAS) and Raven 13 (CAS) [confirmed by Pursell]; Presidio of San Francisco: Coastal Bluff Trail, Shevock 18943 (CAS, PAC, UC) and Arguello Blvd. at Washington Blvd., Shevock 19227 (CAS, PAC, UC) [dets. by Pursell]; Golden Gate National Recreation Area: Coastal Bluff Trail near Sutro Park, Shevock 18814 (CAS, PAC, UC) and Lin- coln Park near Palace of the Legion of Honor, Shevock 19263 (CAS, PAC, UC) [dets. by Pur- [Vol. 48 sell]; Glen Canyon, Shevock 18873 (CAS, PAC, — UC) [det. by Pursell]; Bay View Park, Shevock | 19018 (CAS, PAC, UC) [det. by Pursell]; Mt. | Sutro, Shevock 19173 & 19175 (CAS, PAC, UC) | [det. by Pursell]; Stern Grove, Shevock 19156 | (CAS, PAC, UC) [det. by Pursell]; Mt. Davidson, | Toren 7761 (CAS); Lone Mountain, Shevock | 19332 (CAS, MO, PAC, UC) [det. by Pursell]; Castro at 30th Street, Shevock 19349 (CAS, PAC, UC) [det. by Pursell]. Fissidens curvatus Horsch. [= Fissidens milo-bak- eri Koch]. First record: 2000. On clayey soils. Coastal Bluff Trail, Golden Gate National Rec- reation Area, Shevock 18809 (CAS, MO, PAC, UC) [confirmed by Pursell]; along | O’Shaughnessy Blvd. adjacent to west edge of © Glen Canyon, Shevock 18886 (CAS, PAC, UC) | [confirmed by Pursell]; Buena Vista Park, Shev- ock 18915 (CAS, PAC, UC) [confirmed by Pur- sell]; Douglass Playground, Toren 7758 (CAS); Mt. Sutro, Toren 7772 (CAS); Bernal Heights Park, Toren 7793 (CAS); Yerba Buena Island, Shevock 19125 (CAS, PAC, UC) & Toren 8052 (CAS)[confirmed by Pursell]; Corona Heights, Shevock 18830 (CAS, PAC, UC) [det. by Pur- sell]; Twin Peaks, Shevock IS8863 & 18872 (CAS, PAC, UC) [confirmed by Pursell]. Fissidens cf. minutulus Sullivant First record: 2000. Golden Gate Park, De Laveaga Dell, Aids Me- morial Grove, Shevock 20284 (CAS, PAC, UC) [det. by Pursell] Fissidens sublimbatus Grout First record: 2000. On shaded clayey soils. Coastal Bluff Trail, Presidio of San Francisco, Shevock 18945 (CAS, PAC, UC); Bay View Park, Toren 7831 (CAS); Potrero Hill, Toren 7824 (CAS) [dets. by Pursell]; Yerba Buena Island, Shevock 20251 (CAS, PAC, UC) [det. by Pursell]. Fontinalis neomexicana Sull. & Lesq. First record: 1875. Attached to rocks in creek. Vasey s.n. (US) [cited by Welch 1960]. This species is likely ex- tirpated from San Francisco due to lack of suit- able habitat. Funaria hygrometrica Hedw. First record: 1949. On moist or dry soils, retaining walls and other disturbed areas. Common on disturbed soils throughout the City. Golden Gate Park: Califor- nia Academy Building, Shevock 18753 (CAS, UC), Quarry Lake, Raven 7 (CAS); Civic Center, J. T. Howell s.n. (CAS); Fillmore at Vallejo Street, Raven s.n. (CAS); Webster at Fillmore Street, Raven s.n. (CAS); east end of 22nd Av- enue, J. T. Howell s.n. (CAS); above Lake Mer- ced, Raven 4 (CAS); Baker Beach, Golden Gate Recreation Area, Raven s.n. (CAS); Aqua Vista Park, Shevock 18996 (CAS, UC); Bay View Park, Shevock 19015 (CAS, UC); Corona Heights, Shevock 18827 (CAS, UC); Yerba Bue- na Island, Shevock 19123 (CAS, MO, UC); Em- barcadero across from Pier 29 & 31, Shevock 19103 (CAS, UC); Aqua Vista Park, Shevock 2001] 18996 (CAS, UC); Golden Gate Heights Park, Shevock 19195 (CAS, UC); Lake Merced, Shev- ock 19287 (CAS, UC); Sutro Bath site, Shevock 19216 (CAS, UC) and Fort Funston, Golden Gate National Recreation Area, Shevock 191/60 (CAS, UC); Lone Mountain, Shevock 19338 — (CAS, UC); Corona Heights above the Castro, | Shevock 18829 (CAS, UC); Alcatraz Island, Golden Gate National Recreation Area, Toren | 8467 (CAS). -Funaria muhlenbergii Turn. [= Funaria mediter- ranea (Brandegee 1891); Funaria calcarea (Les- quereux 1868)]. First record: 1868. On soil. Mis- sion Dolores, Bolander s.n. (UC). Although hab- itat for this species still remains in San Francisco, we did not encounter this ephemeral species dur- ing our study. Grimmia laevigata (Brid.) Brid. First record: 2000. On boulders and rock outcrops. Entrance to Hall of Man, California Academy of Sciences, Golden Gate Park, Toren 7822 (CAS). Grimmia lisae De Not. First record: 2000. On boul- ders, rock outcrops and concrete retaining walls. Twin Peaks, Toren 7753 (CAS); Yerba Buena Is- land, Toren S056 (CAS) & Shevock 19109 (CAS, MA, UC) [det. by Munoz]. Grimmia montana Bruch & Schimp. in B.S.G. [Brandegee (1891); Lesquereux (1868)]. First record: 1868. On boulders and rock outcrops. San Francisco, Bolander s.n. (UC). Although habitat exists for the species in San Francisco, we did not encounter it during our study. Grimmia pulvinata (Hedw.) Sm. First record: 2000. On boulders, rock outcrops and concrete retain- ing walls. Alamo Square, Shevock 1891/2 (CAS); Clipper Street near Douglass Playground, Toren 7779 (CAS); Bay View Park west of Candlestick, Toren 7533 (CAS); Yerba Buena Island, Shevock 19113 (CAS, MA, UC) [det. by Munoz]. Grimmia trichophylla Grev. [= Grimmia califor- nica (Brandegee 1891 & Lesquereux 1868) & G. watsoni (Brandegee 1891)]. First record: 1868. On boulders, rock outcrops and concrete retain- ing walls. Common around San Francisco, Bo- lander s.n. (UC); Golden Gate Heights Park, To- ren 7940 (CAS); Twin Peaks, Shevock 18871 (CAS, UC) and Toren 7744 (CAS); Mt. Sutro, Shevock 19174 (CAS, UC); Glen Canyon, San Miguel Hills, Shevock 18883 (CAS, MO, UC); Mt. Davidson, Toren 7769 (CAS); Yerba Buena Island, Shevock 19116 (CAS, UC); open space (Future park), corner of Castro and 30th Streets, Toren 8015 (CAS); 14th & Ortega Streets, Shev- ock 19147 (CAS, MA, UC) [det. by Mufioz]. Gymnostomum aeruginosum Sm. [= Gymnosto- mum calcareum var. perpusillum (Brandegee 1891 & Lesquereux 1868)]. First record: 1868. On shaded, moist calcareous soil over rock. San Francisco, Bolander s.n. (UC); Cesar Chavez Street at highway 101, Toren 7817 CAS). Hennediella heimii (Hedw.) Zand. First record: SHEVOCK AND TOREN: SAN FRANCISCO MOSSES 1] 2000. On exposed soil. Letterman Complex, Pre- sidio of San Francisco, Shevock 20207 (CAS, UG), Hennediella stanfordensis (Steere) Blockeel First record: 2000. On exposed soil. Potrero Hill, To- ren 7827 (CAS); Cesar Chavez Street at highway 101, Toren 7815 (CAS); Golden Gate Park, To- ren s.n. (CAS, SFSU); Wayland and University Streets near McLaren Park, Toren 7743b (CAS); Bernal Heights Park, Toren 7784 (CAS); Aqua Vista Park, Central Basin, Shevock 18999 (CAS, UC), McKinley Square, Shevock 19275 (CAS, UC); southeast end of Lake Merced. Shevock 20286 (CAS, UC). Heterocladium dimorphum (Brid.) Schimp. in B. S. G. First record: 2000. On tree trunks. Lafayette Park, Shevock 19242 (CAS, NY, UC) & 19245 (CAS, MO, UC); Balboa Park, Shevock 19279 (CAS, UC) [dets. by Norris]. These collections represent the first documented occurrences of this species for California. Homalothecium arenarium (Lesq.) Lawt. [= Hyp- num arenarium (Brandegee 1891 & Lesquereux 1868)]. First record: 1868. On soil, leaf litter and rock outcrops. Covering the sand among bushes near the shore, Bolander s.n. (UC); Twin Peaks, Shevock IS8SS868 (CAS, UC) and Toren 7755 (CAS); Bay View Park, Shevock 19013 (CAS, MO, UC); Bernal Heights Park, Toren 7791] (CAS) & Shevock 19130 (CAS, MO, UC); open space (future park) at Castro and 30th Streets, Toren SO18 (CAS); Douglass Playground, Toren 7776 (CAS); Kite Hill Park, Shevock 191/41] (CAS, UC); Baker Beach, Presidio of San Fran- cisco, Shevock 18948 (CAS, MO, UC). Homalothecium nuttallii (Wils.) Jaeg. First record: 2000. On bases of tree trunks, rock outcrops and concrete and brick-like retaining walls. Buena Vista Park, Shevock 18916 (CAS, UC); Bernal Heights Park, Toren 7786 (CAS); Paramount Terrace off Stanyan Street, Toren 7751 (CAS); Golden Gate Park, Panhandle, Toren 7SO02 (CAS) & Shevock 18768 (CAS, MO, UC); Yerba Buena Island, Toren SOS5S8 (CAS) & Shevock 19119 (CAS, MO, UC); Mt. Sutro, Shevock 19182 (CAS, MO, UC); near Crissy Field, Presidio of San Francisco, Shevock 19224 (CAS, MO, NY, UC); Julius Kahn Playground, Presidio of San Francisco, Shevock 19350 (CAS, MO, UC); Aquatic Park, Golden Gate National Recreation Area, Shevock 19097 (CAS, MO, NY, UC); Alta Plaza Park, Shevock 19262 (CAS, UC); Bay View Park, Shevock 19019 (CAS, MO, NY, UC). Hygroamblystegium tenax (Hedw.) Jenn. [= HAy- groamblystegium irriguum in Steere, Anderson & Bryan 1954] First record: 1953. Golden Gate Park, Japanese Tea Garden, L.E. Anderson 32 (herb. not cited) [specimen cited in Steere, An- derson & Bryan 1954]. Isothecium cristatum (Hampe) Robins. [= Hypnum brewerianum (Brandegee 1891 & Lesquereux MADRONO [Vol. 48 1868)]. First record: 1868. On exposed roots, logs and base of tree trunks, gravelly soils and rock outcrops and concrete retaining walls. On metamorphic sandstone around San Francisco, Bolander s.n. (UC); Glen Canyon, San Miguel Hills, Shevock 18875 & 18891 (CAS, MO, UC); Corona Heights, Shevock 18825 (CAS, UC); Mt. Sutro, Shevock 19177 (CAS, UC), Shevock 19169 (CAS, MO, UC) & J9I87 (CAS, UC); Open space (future park) corner of Castro and 30th Streets, Toren SO21] (CAS) & Shevock 19345 (CAS, UC); North Lake of Chain of Lakes, Golden Gate Park, Shevock 19205 (CAS, UC); Presidio of San Francisco: Pacific Street near playground, Shevock 19354 & 19357 (CAS, UC) and Presidio Golf Course, Shevock 19363 (CAS, UC); Yerba Buena Island, Shevock 20246 & 20249 (CAS, UC); Mt. Davidson, Shevock 19293 (CAS, UC); between Painted Rock Cliffs and Sutro Park, Golden Gate National Recreation Area, Shevock 18811 (CAS, MO, UC) [dets. by Norris]. Isothecium myosuroides Brid. [= Hypnum myosu- roides (Brandegee 1891 & Lesquereux 1868)]. First record: 1868. On tree trunks, logs and rock outcrops. In dry woods, San Francisco, Bolander s.n.(UC); Mt. Davidson, Shevock 18924 & 18927 (CAS, MO, UC) and Toren 7766 (CAS); Mt. Su- tro, Shevock 19170, 19176, 19190, & 19193 (CAS, UC); Golden Gate Heights Park, Toren 7945 (CAS); Brotherhood Way near Lake Mer- ced, Shevock 19285 (CAS, MO, UC); southeast side of Lake Merced, Shevock 20285 (CAS, UC); Golden Gate National Recreation Area: Lincoln Park, Palace of the Legion of Honor, Shevock 19264 (CAS, UC) and Coastal Bluff Trail be- tween Painted Rock Cliffs and Sutro Park, Shev- ock 18812 (CAS, UC) and Alcatraz Island, Shev- ock, Toren & Thomas 20209 (CAS, UC). Isothecium spiculiferum (Mitt.) Ren. & Card. First record: 2000. On tree logs. Mt. Sutro, Shevock 19181 (CAS, UC); Golden Gate Park, Chain of Lakes, Shevock 19204 (CAS, UC) [dets. by Nor- ris]. Kindbergia oregana (Sull.) Ochyra [= Eurhyn- chium oreganum (Sull.) Jaeg.] First record: 2000. On soil with leaf litter and logs. Mt. Davidson, Shevock 18923 (CAS); Mt. Sutro, Shevock 19184 (CAS, MO, UC); Fort Point, Presidio of San Francisco, Shevock 18939 (CAS, MO, UC); Lone Mountain, Shevock 19336 (CAS, MO, NY, UC); Glen Canyon, San Miguel Hills, Shevock 18877 (CAS, UC); Yerba Buena Island, Shevock 20247 (CAS, UC). Kindbergia praelonga (Hedw.) Ochyra [= Eurhyn- chium praelongum (Hedw.) Schimp. in B.S.G.]. First record: 1923. On soil with leaf litter, lawns, rock outcrops and tree trunks. Mt. Davidson, Shevock 18925 (CAS, MO, UC), Shevock 19294 (CAS, MO, UC), Shevock 19295 (CAS, UC) & Oberlander s.n. (SFSU); Lone Mountain, Shev- | ock 19333 (CAS, MO, UC); Brotherhood Way, near Lake Merced Blvd., Toren 4946d (CAS);! Walton Park, Shevock 19302 (CAS, UC); Lafa-, yette Park, Shevock 19243 (CAS, UC); Alamo, Square, Shevock 18907 (CAS, UC); Alta Plaza,| Shevock 19253 (CAS, UC); Golden Gate Heights’ Park, Shevock 19199 (CAS, UC); Corona) Heights, Shevock 18828 (CAS, UC); Stern Grove! and Pine Lake Park, Shevock 19154 & 19157| (CAS, UC); McLaren Park, Shevock 19134| (CAS, UC); Mt. Sutro, Shevock 19168 (CAS, UC); Burke at Third Street, Shevock 19006. (CAS, UC); Golden Gate Park: Koch 1636 (UC), Wagner 2365 (UC), Koch 1636 (UC); Lily Lake, Shevock I88I17 (CAS, UC) & Shevock 18818 (CAS, MO, UC), Mallard Lake, Shevock 18893. (CAS, MO, UC), between Stow Lake and De-. Young Museum, Shevock 18759 (CAS, MO, UC), Quarry Lake, Raven 8 & Hermann 17448, (CAS), Rhododendron Grove, Shevock 18762. (CAS, UC), Redwood Grove, Fulton Street near 12th Avenue, Raven s.n. (CAS), Conservatory of Flowers, Shevock 18819 (CAS, UC), Panhandle, | Shevock 18770 (CAS, UC); hill near Cole Street, Eastwood 58 (CAS), De Laveaga Dell, Aids Me- | morial Grove, Golden Gate Park, Shevock 19516, 19518, & 19523 (CAS, UC), California Acade- my of Sciences, Shevock 18754 (CAS, UC), Middle Chain of Lakes, Shevock 19208 (CAS, | UC), Strawberry Hill, Shevock 19162, 19164, & 19166 (CAS, UC); Presidio of San Francisco: near Crissy Field, Shevock 19226 (CAS, UC) &| Fort Point, Shevock 18935 (CAS, MO, UC), near. Arguello Gate, Shevock 19359 (CAS, UC), Let- | terman Complex, Shevock 19217, 19218, & 19220 (CAS, UC), Pacific Street near play- ground, Shevock 19358 (CAS, UC); Golden Gate | National Recreation Area: Fort Mason, Shevock | 19092 (CAS, UC), Lincoln Park, Shevock 19266 (CAS, UC). Leptobryum pyriforme (Hedw.) Brid. First record: | 1867. On soil of seepy areas. The Dell, Lone Mountain, Kellogg s.n. (CAS); Ewing Terrace, Toren 7783 (CAS); Bay View Park, Shevock 19016 (CAS, UC); Presidio of San Francisco near Arguello Gate, Shevock 19361 (CAS, MO, UC) and just west of Pet Cemetery, Shevock 19225 & 19366 (CAS, MO, UC). Leptodictyum riparium (Hedw.) Warnst. First rec- ord: 2000. On saturated soils and lawns. Alamo Square, Shevock 18909 (CAS, MO, UC); Em- barcadero near Pier 29 & 31, Shevock 19107 (CAS, UC). Microbryum starkeanum (Hedw.) Zand. [= Pottia starkeanum (Brandegee 1891) & as Anacalypta starkeana (Lesquereux 1868)] First record: 1868. On bare soil. Mission Dolores, Bolander s.n. (UC); Golden Gate Park near 41st Avenue and Lincoln Blvd., Toren 8022 (CAS). Orthodicranum tauricum (Sapehin) Smirnova First | ( » | 2001] SHEVOCK AND TOREN: SAN FRANCISCO MOSSES 13 record: 2000. On rotten logs. Mt. Sutro, Shevock | 19191 (CAS, UC). Orthotrichum consimile Mitten First record: 2000. Plagiothecium laetum Schimp. in B.S.G. First rec- ord: 2000. On logs and on moist soul with rock. Near Stow Lake, Golden Gate Park, Shevock On Pittosporum and Myoporum tree trunks. Di- amond Heights, Duncan Street and Cameo Way, Toren 8160 (CAS); Panhandle, Golden Gate Park, Shevock 18932 (CAS, UC) [det. by Norris]. Orthotrichum diaphanum Brid. First record: 2000. ~ On tree trunks and concrete retaining walls. Golden Gate Park, Panhandle, Toren 7795 (CAS) & Shevock 18931 (CAS, MO, UC); Balboa Park, Shevock 19283 (CAS, UC); Lafayette Park, Shevock 19240 (CAS, MO, UC). Orthotrichum lyellii Hook. & Tayl. First record: 2000. On tree trunks and branches. Golden Gate Park, Panhandle, Shevock ISS38 & I8932a (CAS, UC) and De Laveaga Dell, Aids Memorial Grove, Shevock 19519 (CAS, UC); Mt. Sutro, Shevock 19300 (CAS, UC); Alamo Square, Shev- ock 18905 (CAS, UC); Glen Canyon, San Miguel Hills, Shevock 1S88/] (CAS, UC); Lafayette Park, Shevock 19242 (CAS, UC); Embarcadero near Piers 29 & 31, Shevock 19106 (CAS, UC); Let- terman Hospital Complex, Presidio of San Fran- cisco, Shevock 19219 (CAS, UC) [det. by Vitt]; Third Street between Burke and Custer Streets, Shevock 19005 & 20203 (CAS, UC) [det. by Vitt]. Orthotrichum rivulare Turn. First record: 1868? On rocks along stream. San Francisco, Bolander s.n. no date (UC). Although this Bolander collection attributed to San Francisco was not referenced by Brandegee (1891), the specimen is nonethe- less accurately identified [det. confirmed by Nor- ris]. This species is most likely extirpated from San Francisco with the elimination of free-flow- ing streams with bedrock. Orthotrichum tenellum Bruch ex Brid. First record: 2000. On tree trunks and rock outcrops and con- crete retaining walls. Panhandle of Golden Gate Park, Toren 7794 (CAS), Shevock 18763 & Shevock 18930 (CAS, UC); Yerba Buena Island, Toren SOS1I (CAS); Lafayette Park, Shevock 19242 (CAS, UC); Jefferson Square, Shevock 18902 (CAS, UC); Pershing Square, Presidio of San Francisco, Shevock 19368 (CAS, UC); Holly Park, Shevock 19132 (CAS, UC); Third and Burke Streets, Shevock 20202 (CAS, UC); Southeast Farallon Island, Farallon National Wil- idlife Refuge, Shevock 20149 (CAS, UC). Phascum cuspidatum Hedw. [Brandegee (1891) & Lesquereux (1868)] First record: 1868. On ex- posed soil. On the grounds and fields and mead- ows, Mission Dolores, Bolander s.n.(UC) [con- firmed by Toren]; Potrero Hill, Toren 7823 (CAS). Philonotis capillaris Lindb. in Hartm. First record: 2000. On moist soil over rock. Tank Hill, Shev- ock 18918 (CAS, UC); Clipper Street near Doug- lass Playground, Toren 7780 (CAS); Bernal Heights Park, Toren 7790 (CAS). 18760 (CAS, MO, UC) & 19/65 (CAS, MO, NY, UC) and near Chain of Lakes, Shevock 1916la (CAS); Mt. Davidson, Toren 7763 (CAS). Pleuridium acuminatum [= Pleuridium bolanderi C. Muell ex Jaeg.]. First record: 1868. On ex- posed soil. Near San Francisco, Bolander s.n.(UC); Mt. Davidson, Shevock 1892] (CAS); Marietta Drive, Toren 7810 (CAS); on hill near Sutrowood and Stanyan Street, Eastwood 54 (CAS); Wayland at University Street near Mc- Laren Park, Toren 7743c (CAS); Bernal Heights Park, Toren 7787 (CAS); open space (future park) at corner of Castro and 30th Streets, Toren SOI6 (CAS); Diamond Heights, Toren 8457 (CAS). Pleuridium subulatum (Hedw.) Rabenh. First rec- ord: 1904. On exposed soil. Glen Canyon, San Miguel Hills, Shevock 1888S (CAS, UC); Ar- guello Blvd. At Washington Blvd., Presidio of San Francisco, Shevock 19229 (CAS, UC); Gold- en Gate Park, Gardner s.n. (UC) [dets. by Nor- ris]. Pohlia nutans (Hedw.) Lindb. First record: 2000. On moist soils with rock. Mt. Davidson, Toren 7/62 (CAS). Pohlia wahlenbergii (Web. & Mohr.) Andrews [= Bryum albicans (Brandegee 1891 & Lesquereux 1868)]. First record: 1868. On moist soil. On rocks watered by springs, Bolander s.n,. (UC) [det. confirmed by Shaw]. This species is likely extirpated from San Francisco due to a lack of suitable habitat. Polytrichastrum alpinum (Hedw.) G. L. Sm. First record: 2000. On soil with rock. Mt. Davidson, Toren 7764 (CAS). Polytrichum juniperinum Hedw. First record: 1950. On soil among rock outcrops. Glen Canyon, San Miguel Hills, Shevock I1SSS2 & 18892 (CAS, MO, UC); Mt. Davidson, Shevock 18922 (CAS, UC), Toren 7750 (CAS) and Raven 18 (CAS); Twin Peaks, Shevock 18867 (CAS, UC); between Marietta Drive and O’Shaughnessy Blvd., Toren 7808 (CAS); Douglass Playground, Toren 7777 (CAS); Bay View Park, Shevock 19010 (CAS, MO, UC); open space (future park) at corner of Castro and 30th Streets, Toren SOl14 (CAS). Polytrichum piliferum Hedw. First record: 2000. On soil among rock outcrops. San Miguel Hills be- tween Marietta Drive and O’Shaughnessy Blvd., Toren 7809 (CAS); Diamond Heights, Duncan and Newburg Streets, Toren S456 (CAS). Porotrichum bigelovii (Sull.) Kindb. First record: 1962. On shaded moist soil and rocks. De Lav- eaga Dell, Aids Memorial Grove, Golden Gate Park, F. J. Hermann 17451 (CAS) & Shevock 19526 (CAS, MO, NY, UC). Pseudocrossidium obtusulum (Lindb.) Crum & An- derson First record: 2000. On soil. Cesar Chavez 14 MADRONO [Vol. 48 Street at highway 101, Toren 7820 (CAS); Bay View Park, Toren 7835 (CAS). Pseudotaxiphyllum elegans (Brid.) Iwats. First rec- ord: 1973. On shaded clayey soil. Mt. Davidson, Toren & Wong 840 (CAS, SFSU) & Toren 7804 (CAS); Mt. Sutro, Shevock 19179 (CAS). Pterogonium gracile (Hedw.) Sm. [Lesquereux (1868)]. First record: 1868. On rock outcrops and tree trunks. On ground and rocks, Bolander s.n.(UC). Although habitat for this species re- mains in San Francisco we did not encounter it during this study. Rhytidiadelphus squarrosus (Hewd.) Warnst. First record: 1976. On lawn. Japanese Tea Garden, Golden Gate Park, Toren, Showers, & Smith 2811 (SFSU). Scleropodium californicum (Lesq.) Kindb. [= Hyp- num californicum (Lesquereux 1868)]. First rec- ord: 1868. On soil with leaf litter, rock outcrops and concrete retaining walls. Near the Bay of San Francisco, Bolander s.n.(UC); Golden Gate Na- tional Recreation Area, Coastal Bluff Trail be- tween Painted Rock Cliffs and Sutro Park, Shev- ock 18808 (CAS, MO, UC) & Shevock 18813 (CAS, UC), Fort Miley area, Shevock 19371 (CAS, UC) and Alcatraz Island, Shevock, Toren & Thomas 20214 (CAS, UC); South Windmill, Golden Gate Park, Shevock 18897 (CAS, MO, UC); Fort Point, Presidio of San Francisco, To- ren 7741 (CAS); Potrero Hill, Toren 7828 (CAS); Twin Peaks, Toren 7756 (CAS); Bernal Heights Park, Toren 7785 (CAS); Bay View Park, Toren 7830 & 7840 (CAS); Golden Gate Heights Park, Toren 7944 (CAS); Lake Merced, Toren, Showers & Halling 1460 (SFSU); Clar- endon Heights, Howe 523 (CAS); Potrero Hill Park, Shevock 19000 (CAS, UC); open space (fu- ture park) at Castro and 30th Streets, Toren SO19 (CAS); Yerba Buena Island, Toren SOS9 (CAS), Shevock 19112 (CAS, MO, UC), Shevock 19120, 20245, 20250 & 20252 (CAS, UC). Scleropodium cespitans (C. Mull.) L. Koch First record: 2000. On clayey soil, base of tree trunks and exposed roots, rock outcrops and concrete retaining walls. Yerba Buena Island, Toren SOS53 & S060 (CAS); Walton Park, Shevock 19303 (CAS, UC); below Coit Tower, Shevock 19108 (CAS, MO, UC); Corona Heights, Shevock 18823 & 18826 (CAS, UC); Golden Gate Heights Park, Shevock 19198 (CAS, UC); Lafa- yette Park, Shevock 19244 & 19247 (CAS, UC); Holly Park, Shevock 19133 (CAS, UC); Grand View Park, Shevock 191/43 (CAS, UC); Glen Canyon, San Miguel Hills, Shevock 18885 (CAS, UC); Golden Gate Park: California Academy of Sciences building, Shevock 18755 (CAS, MO, UC), Panhandle, Shevock 18771 (CAS, UC); Southeast Farallon Island, Farallon National Wildlife Refuge, Shevock 20150 & 20152 (CAS, UC); Presidio of San Francisco: Letterman Com- plex, Shevock 19221 (CAS, UC), Fort Point, Scleropodium colpophyllum (Sull.) Grout. First rec-; Scleropodium julaceum Lawt. First record: 2000., Shevock 18942 (CAS, UC), Baker Beach, Shem ock 18949 (CAS, UC). | ord: 2000. On soil and leaf litter, base of tree, trunks and exposed roots, and concrete and brick retaining walls. Brotherhood Way near Lake’ Merced Blvd., Toren 7946a (CAS); Mission Do-| lores Park, Shevock 18990 (CAS, UC); Alamo! Square, Shevock 18908 (CAS, UC); Glen Can- yon, Shevock 18876 (CAS, MO, UC); Turk &!| Fillmore Streets, Shevock 19342 (CAS, UC); Stern Grove & Pine Lake Park, Shevock 1915] & 19152 (CAS, UC); Golden Gate Park: near Conservatory of Flowers, Shevock 18803 (CAS, UC), Panhandle, Shevock 18766 (CAS, UC), De Laveaga Dell, Aids Memorial Grove, Shevock 19520 (CAS, UC) & between Stow Lake and DeYoung Museum, Shevock 18758 (CAS, MO, UC); Aquatic Park, Golden Gate National Rec-| reation Area, Shevock 19101 (CAS, UC); Alca- traz Island, Golden Gate National Recreation, Area, Toren 5461 & 8470 (CAS). On clay soils and rock outcrops. Along Park Pre- sidio (highway 1) between Lake and Sacramento. Streets, Shevock 19232 (CAS, UC); Potrero Hill. Park, 22nd Street between Arkansas and Missou-. ri Streets, Shevock 19003 (CAS, UC). Scleropodium obtusifolium (Mitt.) Kindb. in Mac. & Kindb, First record: 2000. On moist clayey soils over rocks. Lincoln Park, Golden Gate Na- tional Recreation Area, Shevock 19267 (CAS, UC); Red Rock Hill, southeast corner of Duncan | St. and Diamond Heights Blvd., Toren 8504. (CAS). Scleropodium touretii (Brid.) L. Koch [= Hypnum illecebrum (Brandegee 1891)]. First record: 1868. On clayey soil over rocks. Shady sandy ground, San Francisco, Bolander s.n. (UC); Glen Canyon, San Miguel Hills, Shevock 18874 (CAS, MO, UC); open space (future park) at Castro and 30th Streets, Toren SOI2 (CAS); Golden Gate Park, Oberlander s.n. (SFSU) [det. by Toren]; Twin Peaks, Shevock 18866 (CAS, UC); Bernal Heights Park, Shevock 19128 (CAS, UC); Yerba Buena Island, Shevock 19121 (CAS, UC); Bay View Park, Shevock 19012 (CAS, UC); Lone Mountain, University of San Francisco, Shevock 19337 (CAS, UC).. Sematophyllum adnatum (Michaux) Britton First record: 2000. On rotten logs. Mt. Davidson, Shevock 18929 (CAS, MO, NY, UC) [det. con- firmed by Norris]. This collection represents the first documented occurrence of this species for California. Syntrichia amplexa (Lesq.) Zand. [= Barbula am- plexa (Brandegee 1891)]. First record: 1868. On clayey soil and rocks. Near the Bay of San Fran- cisco, Bolander s.n.(UC) and Blasdale s.n. (UC); Bay View Park, Toren 7836 (CAS); China 2001 ] SHEVOCK AND TOREN: SAN FRANCISCO MOSSES 15 Beach, Golden Gate National Recreation Area, Shevock 19270 (CAS, UC). Syntrichia bolanderi (Lesq. & James) Zand. [= Barbula bolanderi (Brandegee 1891 & Lesquer- eux 1868)]. First record: 1868. On soil. In gar- dens, San Francisco, Bolander s.n.(UC). Al- though habitat for this species still remains in San Francisco, we did not encounter it during this study. Syntrichia pagorum (Milde) Amann First record: 2000. On tree trunks. Panhandle section, Golden Gate Park, Toren 7799 (CAS); Lafayette Park, Shevock 19246a (CAS); Alamo Square, Shevock 18910 (CAS, UC). Syntrichia princeps (De Not.) Mitt. First record: 1868. On tree trunks, soil over rocks and con- crete retaining walls. Mission Dolores, Bolander s.n. (UC) [det. by Norris]; Buena Vista Park, Shevock 18917 (CAS, UC); Twin Peaks, Toren 7754 (CAS); Yerba Buena Island, Toren SOSS (CAS), Shevock 19110 & 19114 (CAS, MO, UC); Pacific Street, Presidio of San Francisco, Shevock 19352 & 19356 (CAS, UC); Panhandle, Golden Gate Park, Shevock 18767 (CAS, MO, UC); Alta Plaza, Shevock 19260 (CAS, UC); Balboa Park, Shevock 19284 (CAS, UC); Lafa- yette Park, Shevock 19233 (CAS, UC); Bay View Park, Shevock 19008 (CAS, UC); Alcatraz Is- land, Golden Gate National Recreation Area, Shevock, Toren & Thomas 20213 & 20216 (CAS, UC). Syntrichia ruralis (Hedw.) Web. & Mohr First rec- ord: 1868? On concrete retaining wall. Sand Hills, E. L. Greene s.n. (UC); Alta Plaza Park, Shevock 19258 (CAS, UC) [dets. by Norris]. Timmiella anomala (Bruch & Schimp. in B. S. G.) Limpr. [= Trichostomum anomalum (Brandegee 1891)]. First record: 1868. On soil with rock. Near San Francisco, Bolander s.n.(UC); Bernal Heights Park, Shevock 1925S (CAS, MO, UC) [det. by Norris]. Timmiella crassinervis (Hampe) L. Koch [= Tri- chostomum flexipes (Brandegee 1891)]. First rec- ord: 1868. On soil with rock. Common on shaded ground and decaying trunks, Bolander s.n. (UC); Wayland Street at University Street near Mc- Laren Park, Toren 7743a (CAS); Mt. Sutro, To- ren 7775 (CAS); Bernal Heights Park, Toren 7789 (CAS); Bay View Park, Toren 7837 (CAS); Yerba Buena Island, Shevock 19/22 (CAS, UC): McLaren Park, Shevock 19139 (CAS, UC). Tortula atrovirens (Sm.) Lindb. [= Desmatodon nervosus var. edentulus (Brandegee 1891) & Desmatodon californicus (Lesquereux 1868)]. First record: 1868. On wall of buildings. Decay- ing ground of old walls of clay (adobe), San Francisco, Bolander s.n. (UC). Although habitat for this species still remains in San Francisco, we did not encounter it during this study. Tortula brevipes (Lesq.) Broth. [= Barbula brevi- pes (Brandegee 1891 & Lesquereux 1868)]. First record: 1868. On buildings. Mud walls, Mission Dolores, in mats an inch broad or more, Bolan- der s.n. (UC). Although habitat for this species still remains in San Francisco, we did not en- counter it during this study. Tortula latifolia Bruch ex Hartm. First record: 2000. On concrete retaining walls. Yerba Buena Island, Toren SOSO (CAS); Lafayette Park, Shev- ock 19236 (CAS, MO, UC) [det. by Norris]. Tortula muralis Hedw. First record: 1923. On soil, base of tree trunk, boulders, rock outcrops, con- crete retaining walls and buildings. San Francis- co, Sudiffe 49 (CAS); Francisco and Hyde Streets, L. S. Rose s.n. (CAS); University of San Francisco, Fulton at Parker Street, Raven. s.n. (CAS); Balboa Street near Arguello, Raven s.n. (CAS); Filbert and Van Ness Avenue, J. 7. How- ell s.n. (CAS); 26th Avenue near El] Camino del Mar, Raven s.n. (CAS); between Fillmore and Laguna Street, Raven s.n. (CAS); Alamo Square, Shevock 18913 (CAS, UC); Yerba Buena Island, Shevock 19118a & 20244 (CAS, UC); Potrero Hill Park, Shevock 19002 (CAS, UC); Jefferson Square, Shevock 18899 & 18904 (CAS, UC); Golden Gate Park: South Windmill, Shevock 18895 (CAS, MO, UC), Fulton near Ist Avenue, J. T. Howell s.n. (CAS), Stanyon Street between Fulton & Hayes Streets, Shevock 18834 & 18835 (CAS, UC), Quarry Lake, Raven 5 (CAS), Stow Lake, Shevock 18757 (CAS, UC) and De Lav- eaga Dell, Aids Memorial Grove, Shevock 19522 (CAS, UC); Cesar Chavez Street at highway 101, Toren 7814 (CAS); Lookout Point, Halling 600 (SFSU) [confirmed by Toren]; Glen Canyon, San Miguel Hills, Shevock 18879 (CAS, UC); Turk & Fillmore Streets, Shevock 19340 (CAS, UC); Aqua Vista Park, Shevock 18998 (CAS, UC); Alta Plaza Park, Shevock 19261 (CAS, UC); La- fayette Park, Shevock 19235 (CAS, UC); Presi- dio of San Francisco: Coastal Bluff Trail near Golden Gate Bridge, Shevock 18940 (CAS, UC) and near Arguello Gate, Shevock 19362 (CAS, UC);.Golden Gate National Recreation Area: Su- tro bath site, Shevock 19215 (CAS, UC) and Aquatic Park, Shevock 19099 (CAS, UC). Tortula obtusifolia (Schwaegr.) Math. [as Desma- todon flavicans, not in Brandegee 1891]. First record: 1868. On soil? Mission Dolores, Bolan- der s.n. (UC) [det. confirmed by Toren]. Al- though habitat for this species remains in San Francisco we did not encounter it during this study. Tortula papillosa Wils. in Spruce First record: 2000. On tree trunks and concrete retaining wall. Twin Peaks, Toren 7245 (CAS); Alta Plaza Park, Shevock 19254 & 19257 (CAS, UC); Balboa Park, Shevock 19281 (CAS, UC) & Shevock 19282 (CAS, MO, UC); Jefferson Square, Shev- ock 18900 (CAS, MO, UC); Lafayette Park, Shevock 19251 (CAS, MO, UC); Presidio Golf Course, Presidio of San Francisco, Shevock 16 MADRONO 19364 (CAS, UC); Third Street at Burke Street, Shevock 20201 (CAS, UC). Tortula plinthobia (Sull. & Lesq.) Broth. First rec- ord: 2000. On bark of poplars. Third Street at Burke and Custer Streets, Shevock 20206 (CAS, UC). Triquetrella californica (Lesq.) Grout First record: 2000. On soils and rock outcrops. Tank Hill, Shevock 18920 (BUF CAS, UC)[confirmed by Zander]; Clipper Street at Douglass Playground, Toren 7778 (CAS); open space (future park) at Castro and 30th Streets, Toren SO13 (CAS); Di- amond Heights, Duncan and Newburg Streets, Toren 8455 (CAS). Vesicularia vesicularis (Schwaegr.) Broth. First record: 1923. Wet areas of lawn. Golden Gate Park, Bradshaw s.n.(CAS), Wagner s.n.(CAS) and the Conservatory, Toren 688 (CAS, SFSU). Weissia controversa Hedw. [= Weissia viridula (Brandegee 1891 & Lesquereux 1868)]. First record: 1868. On soil. At and around San Fran- cisco, Bolander s.n.(UC); Mt. Davidson, Toren 7768 (CAS); Brotherhood Way near Lake Mer- ced Blvd., Toren 7946c (CAS); Aquatic Park, | Golden Gate National Recreation Area, Shevock | 19095 (CAS, UC); open space at 30th & Castro | St., Shevock 19348 (CAS, UC). Zygodon menziesii (Schwaegr.) Arnott First record: 2000. On trunks of Eucalyptus and Myoporum. | Panhandle section of Golden Gate Park, Toren 7797 (CAS) and Shevock 18773 (ALTA, CAS, MO, UC); Jefferson Park, Shevock 18901 (ALTA, CAS, NY, UC) [dets.confirmed by Vitt]; Diamond Heights, Duncan Street and Cameo Way, Toren 8159 (CAS). These collections rep- resent the first documented occurrences of this species for California and North America. [Vol. 48 | Zygodon rupestris (Lindb. ex Hartm.) Lindb. ex | Britt. First record: 1976. On trunks of Eucalyptus | and Cupressus. St. Francis Blvd., Toren 5001 (CAS); Mt. Davidson, Toren 7759 (CAS); Gold- | en Gate Park, Panhandle, Toren 7796 (CAS); Golden Gate Heights Park, Toren 7939 (CAS) & | Shevock 19201 (ALTA, CAS, MO, UC); Lafa- | yette Park, Shevock 19246 (CAS, NY, UC) & 19250 (ALTA, CAS, MO, UC). “Maprono, Vol. 48, No. 1, pp. 17-24, 2001 BRYOPHYTE FLORA OF WILLIAM L. FINLEY NATIONAL WILDLIFE | REFUGE, WILLAMETTE VALLEY, OREGON KATHY MERRIFIELD | Department of Botany and Plant Pathology, Oregon State University, | 2082 Cordley Hall, Corvallis, OR 97331-2902 ABSTRACT The Willamette Valley in northwestern Oregon is a mosaic of plant communities, some of which have become rare following European settlement. William L. Finley National Wildlife Refuge preserves ex- amples of many of these historic communities, which provide diverse substrates for bryophytes. Mosses and liverworts were collected at the refuge from 1993 through 1999. Eighty-four moss and 24 liverwort species were identified and their substrates cataloged. The moss Physcomitrella patens (Hedwig) Bruch, | Schimper & Gumbel is newly reported for Oregon. The rarity of some common Pacific Northwest species at Finley Refuge may be a function of the drier climate in the Willamette Valley than in the surrounding mountain ranges. Land at Finley Refuge is managed primarily for wildlife species, but the protection of natural and pre-settlement plant communities has resulted in conditions facilitating a rich bryophyte flora. The Willamette Valley in northwestern Oregon provides a diversity of bryophyte substrates that differ from those in the surrounding mountain for- ests because of its contrasting climate and vegeta- tion. The presence of hardwoods as well as coni- fers, open as well as closed canopies, wetlands as well as uplands, and human-altered as well as nat- ural landscapes provide a plethora of substrates supporting a rich bryoflora. Willamette Valley bryophytes have been variously surveyed (Sanborn 1929; Chapman and Sanborn 1941; Pike 1973; Pike et al. 1975), but no intensive collections of all sub- strates within a defined geographic area have been made. This study of a representative Willamette Valley location appears to be the first inventory of all substrates within such a defined area. Warmer and drier than the surrounding moun- tains, the Willamette Valley is a broad depression between the Coast and Cascade Ranges, extending from the Columbia River south to the convergence of the two ranges at Cottage Grove (Franklin and Dyrness 1973). While average yearly rainfall in the Cascades and Coast Range averages 200 to 340 cm, only about 100 cm falls in the Willamette Valley, and potential evapotranspiration far exceeds winter moisture buildup due to hot, dry summers (Habeck 1961; Franklin and Dyrness 1973). The Willamette Valley is a mosaic of deciduous and coniferous forests, savannahs, grasslands, and wetlands (Johannessen et al. 1971: Franklin and Dyrness 1973). Oak forests and savannahs are dom- inated by Quercus garryana Hook. Acer macro- phyllum Pursh, Pseudotsuga menziesii (Mirbel) Franco, and Arbutus menziesii Pursh may be co- dominant. Pseudotsuga menziesii dominates in co- niferous forests, but Abies grandis (Douglas) Lind- ley and A. macrophyllum are also widespread. Fraxinus latifolia Benth. forests are common in seasonally flooded areas, especially along streams. Prairies occupied extensive tracts of the Willamette Valley before it was settled; they are still wide- spread but now harbor many exotic species (Ha- beck 1961; Franklin and Dyrness 1973). These pre- settlement Willamette Valley prairies and savan- nahs may have been seral communities created and maintained by fire (Habeck 1961; Johannessen et al. 1971). Fire control has permitted the develop- ment of oak forests, and nearly all of the remaining prairie is currently being used for agriculture or grazing (Habeck 1961; Johannessen et al. 1971). METHODS Study Area. William L. Finley National Wildlife Refuge comprises 2156 hectares (5325 acres) about 14 km south of Corvallis, Benton County, OR, in the Willamette Valley. Elevations vary from 77 to 189 m (255 to 620 ft). Several pre-setthement Wil- lamette Valley plant communities are preserved within the refuge. Forests, oak savannahs, native wet prairies, swamps, and marshes are interspersed with agricultural fields of wildlife food and cover crops, which together make up 25% of the total area. Both pre-settlement and settler-altered vege- tation are represented. The refuge was established primarily as habitat for dusky Canada geese, the southern population of which winters almost exclu- sively in the Willamette Valley (Palmer 1976). All other native plant and animal species are protected as well. Human perturbations are uncommon out- side of agriculture and of road building and main- tenance and are strictly controlled. Several Willamette Valley bryophyte substrates are especially well preserved at the refuge. Willam- ette Valley native wet prairie is a substrate now unique to Finley Refuge and only a few other sites. In these prairies, soil accumulation around grami- noid hummocks forms vertical surfaces saturated during the winter but desiccated in summer. Forest 18 MADRONO [Vol. 48 : : substrates include the bark of oaks and maples up to 250 years old, rotting logs and stumps, and forest floor litter and soil. Weathered rocks and cervid bones and antlers remain undisturbed in forests. Freshly dug forest, edge, and upland prairie soil is often present because moles, gophers, and voles are not artificially controlled. The bark of isolated oaks in savannahs and agricultural fields support a bryo- phyte assemblage distinct from that of forests. Ver- nal pool conditions are common in agricultural fields because wildlife food crop fields are worked in late summer or fall, rather than in spring as com- mercial agricultural crops are, leaving poorly drained areas undisturbed through spring and early summer. In addition, swamp mud, wet rocks or branches in and near streams, seasonally flooded shaded soil along streams, and upland grasslands are well represented. Because some roads are rarely used or maintained, roadbeds and compacted gravel remain undisturbed for long periods. Pigeon Butte, a sandstone hill bounded on the south by a basalt dike, provides both naturally exposed and quarried basalt. Scattered basalt boulders are also present on other hilltops. Bryophyte Inventory. From February 1993 through August 1999, Finley Refuge was traversed on foot for the purpose of collecting bryophytes to make as complete an inventory as possible. Speci- mens were collected from all known substrates in all natural and human-influenced systems. The vas- cular plant community and substrate of each spec- imen was recorded. Except as otherwise noted, col- lections were identified by K. Merrifield using stan- dard taxonomic works (Howe 1899; Schuster 1969; Lawton 1971; Schuster 1974; Schuster 1977; Crum and Anderson 1981; Smith 1990; Christy and Wag- ner 1996). Taxa newly or rarely reported for Ore- gon were confirmed or corrected by one or more of the following: W. E Schofield and J. Harpel of the University of British Columbia, J. Christy of the Oregon Natural Heritage Program, and V. Bry- an of Duke University. Voucher specimens of each species, and of each variety where applicable, as well as additional specimens from varying sub- strates, were placed in the Oregon State University Herbarium (OSC). Duplicate specimens were placed in the cryptogamic collection at Finley Ref- uge. RESULTS The bryophyte flora of Finley Refuge consists of 108 species, including 84 moss and 24 liverwort species, and 67 genera, including 52 moss and 15 liverwort genera. At least one hornwort, Anthocer- os sp. sensu latu, was present. The largest moss family was Brachytheciaceae, represented by 14 species, followed by Bryaceae (8 species), Potti- aceae (8 species), Grimmiaceae (7 species), Dicran- aceae (6 species), and Mniaceae (6 species). The largest liverwort family was Lophocoleaceae, rep- resented by 5 species, followed by Jubulaceae and { Porellaceae (3 species each.) | Physcomitrella patens (Hedwig) Bruch, Schim- | : per & Gumbel is reported for the first time in Or- \ egon in this study. Collections of Syntrichia laevi- | pila Bridel var. meridionalis (Schimper) Juratzka { made during this study along with collections from | additional Willamette Valley localities were newly | reported for Oregon earlier (Merrifield 2000). Phys- | comitrium immersum Sullivant and Ephemerum | serratum (Hedwig) Hampe are both reported here | for the second time in Oregon (Conard 1944; Chris- ty 1980). As of 1982, the known moss flora of Oresoml| comprised 411 species in 134 genera (Christy et al. | 1982). About 20% of the statewide assemblage of | moss species representing 39% of Oregon genera > reside at Finley Refuge. Chapman and Sanborn ! (1941) documented 114 moss species in 60 genera | in the entire Willamette Valley; 74% as many spe- | cies and 87% as many genera were identified in this | study. The known liverwort flora of Oregon com- — prises 168 species in 57 genera (D. H. Wagner un- ! published) About 14% of the statewide assemblage | of liverwort species representing 26% of Oregon | genera were documented at Finley Refuge. Sanborn (1929) listed 116 liverwort species in 35 genera — throughout western Oregon; 21% as many species and 43% as many genera were identified at Finley Refuge in this study. DISCUSSION Bryophyte diversity increases as the number of | suitable substrates increases (Slack 1977). In west- ern Oregon Cascade and Coast Range sites, in- creased forest floor bryophyte diversity was posi- tively correlated with the abundance of rocks, coarse woody debris, stand openings, and hard- woods (Rambo and Muir 1998). Hardwoods such as red alder and bigleaf maple often support rich epiphytic bryofloras that differ from those on co- nifers (Christy and Wagner 1996). Rainfall dripping through hardwoods is richer in nutrients than that of conifers, and hardwoods allow greater light transmittance than do conifers (Rambo and Muir 1998). In forests alone at Finley Refuge, therefore, several factors encouraging bryophyte diversity are present: deciduous forests are more extensive than coniferous, rocks are present, variously decom- posed coarse woody debris is plentiful, and stand openings as well as edges are common. Oak savan- nahs, wetlands, and riparian communities contrib- ute additional diverse substrates under a variety of conditions. Forests. All 6 bryophyte species occurring on forest floor litter and soil were common at Finley Refuge. Rhytidiadelphus triquetrus was the most abundant, and it dominated the herbaceous layer in many areas. In addition, the branches of Trachy- 2001 ] TABLE 1. MERRIFIELD: BRYOPHYTES OF FINLEY REFUGE 19 TREE SUBSTRATES FOR BRYOPHYTES AT FINLEY REFUGE. Latin name Common name Acer circinatum Pursh Acer macrophyllum Pursh Alnus rubra Bongard Crataegus douglasii Lindley Fraxinus latifolia Bentham Physocarpus capitatus (Pursh) Kuntze Populus balsamifera L. spp. trichocarpa (Torrey & Gray) Brayshaw Pseudotsuga menziesii (Mirbel) Franco Quercus garryana Hooker Salix spp. bryum megaptilum adhered to litter, although plants originated on freshly dug forest soil. Rotting logs harbored 37 species. Among the most abundant were Antitrichia californica, Aula- comnium androgynum, Brachythecium frigidum, Dicranoweisia cirrata, Eurhynchium praelongum var. stokesii, Homalothecium fulgescens, Hypnum subimponens, Isothecium myosuroides, Plagiom- nium venustum, Cephalozia bicuspidata, and C. lunulifolia. The rotting log bryophyte assemblage is expected to contain a high proportion of the liv- erwort flora in the spruce-fir biome of North Amer- ica (Schuster 1969, 1974, 1977; Christy and Wag- ner 1996). Accordingly, 8 liverwort species in the genera Cephalozia, Cephaloziella, Lepidozia, Lo- phocolea, and Scapania, comprising 33% of the he- patic flora at Finley Refuge, were observed solely on rotting logs and branches. The bark of living trees (Table 1) provided sub- strate for 35 species. Antitrichia californica, Den- droalsia abietina, Dicranoweisia cirrata, Homal- othecium fulgescens, H. nuttallii, Hypnum subim- ponens, Metaneckera menziesii, Neckera douglasii, Orthotrichum consimile, O. lyellti, Plagiomnium venustum, Porella navicularis, and P. roellii were the most common. Bryophytes on shaded bark usu- ally formed a continuous mat, while those that oc- curred on more exposed bark usually grew in small groups or discreet tufts. A solid mat of M. menziesii was especially characteristic of large forest oaks. Isolated Oaks. The bryophyte assemblage of ex- posed, often sparsely shaded bark of isolated oaks comprised an unidentified, nonsporulating Didy- modon species, Dicranoweisia cirrata, Syntrichia ruralis, S. latifolia, S. laevipila var. meridionalis, and Zygodon viridissimus, the latter three of which are gemmiparous (Merrifield 2000). In some col- lections in which S. laevipila var. meridionalis ga- metophores were not found, its gemmae were pres- ent among the other mosses. Areas where some or all of these 6 taxa coalesced were colonized by mosses more characteristic of shaded bark, such as Homalothecium nuttallii, Metaneckera menziesii, and Dendroalsia abietina. Upland Soils. Thirty-eight bryophyte species oc- curred on soils. Disturbed, packed upland soil and Vine maple Bigleaf maple Red alder Western hawthorne Oregon ash Ninebark Black cottonwood Douglas fir Oregon white oak; Garry oak Willow soil freshly disturbed by small mammals each har- bored 9 species; Tortula atheroides and Scleropo- dium touretii were common to both. Trachybryum megaptilum grew on mammal-dug forest soil. Only three bryophyte species, Brachythecium albicans, Bryum lisae var. cuspidatum, and Dicranum sco- parium, inhabited upland prairie soils. The occur- rence of the first two were limited. In contrast, extensive collections from many prairie localities indicated that Brachythecium albicans is a wide- spread and profuse grassland community compo- nent. Because agricultural fields occupy about 25% of Finley Refuge land, and because some fields lie fal- low for long periods, exposed agricultural soil is extensive. Of the three moss species inhabiting up- land agricultural soils, Bryum dichotomum was by far the most abundant, covering vast expanses of soil unworked in spring. Wetland Soils. Cumulative losses of bryophyte taxa in western Oregon, Washington, and California have probably been greatest for the wet soil species occurring on floodplains, where agricultural im- pacts have been concentrated (Christy and Wagner 1996). Due to preservation of several native wet- land habitats and to crop management for wildlife rather than for market consumption, including leav- ing poorly drained field soils comparatively undis- turbed, Finley Refuge continues to harbor elements of this bryophyte assemblage. Several soil substrates were seasonally flooded. Primary succession on such recently disturbed soil includes many opportunistic bryophyte species (Christy and Wagner 1996). Both flat streamside soils and undrained soils in agricultural fields har- bored vascular plants characteristic of vernal pools. Physcomitrella patens, Leptodictyum riparium, and Riccia fluitans were unique to streamsides. Dicra- nella heteromalla, Pohlia annotina, and Fossom- bronia wondraczekii were unique to low, undrained areas in cultivated fields, and Physcomitrium pyri- forme occurred on both streamsides and undrained field soils. Several Riccia species from undrained soils remain to be identified (J. Wheeler personal communication). Riccia fluitans and Ricciocarpus natans were the only two free-floating species, and 20 MADRONO they were also found stranded on streamside mud. Ditrichum schimperi, Ephemerum serratum, and a depauperate form of Philonotis fontana were unique to seasonally flooded native wet prairie ver- tical soil surfaces of graminoid hummocks. Pla- giomnium medium and P. ellipticum were unique to the seasonally flooded mud of ash swamps. Permanently shaded vertical soil surfaces along streams supported a unique bryophyte assemblage, including Atrichum selwynii, Eurhynchium praelon- gum var. praelongum, Fissidens bryoides, Junger- mannia rubra, and Riccardia multifida. Rocks, Bones, and Antlers. Basalt provided sub- strate for 30 species. The most common included Dicranoweisia cirrata, Didymodon vinealis, Grim- mia pulvinata, G. trichophylla, Homalothecium nuttallii, Orthotrichum lyellii, Racomitrium canes- cens, and R. heterostichum. While a distinctive as- semblage of Pacific Northwest bryophyte species occurs on rocks (Christy and Wagner 1996), some rock-dwelling species at Finley Refuge also oc- curred on other substrates. Overlap occurred among rotting log-, bark-, and basalt-dwellers, but only Antitrichia californica, D. cirrata, and I. myosuro- ides occurred on all three. None of the moss species on basalt were among the 5 on weathered sandstone, which included Fis- sidens bryoides and Scleropodium touretii, also in- habitants of disturbed upland soils. Fifteen moss species occurred on concrete, and this was the only substrate on which Amblystegium serpens and Tor- tula muralis were found. No species were found on both concrete and sandstone. Only Brachythecium frigidum, Eurhynchium praelongum var. praelon- gum, and Sanionia uncinata occurred on bone and antler. Bryophytes Expected but Uncommon or Absent. Antitrichia curtipendula, Claopodium crispifolium, Scapania bolanderi, and Isothecium myosuroides were less common than expected judging from ac- counts of Pacific Northwest bryoflora (Lawton 1971, Schofield 1976, Vitt et al. 1988). The first three were found only on forested north-facing slopes in isolated patches rather than broad mats. Antitrichia curtipendula requires humidity and 1s considered closely associated with late successional or old-growth forests (Schofield 1976, FEMAT 1993, Christy and Wagner 1996), but Chapman and Sanborn (1941) collected A. curtipendula as well as C. crispifolium throughout the Willamette Val- ley. Scapania bolanderi has often been collected in coastal regions (Howe 1899, Vitt et al. 1988) and Cascade slopes (Sanborn 1929) and is also consid- ered closely associated with late-successional or old-growth forests (FEMAT 1993). Isothecium myosuroides was widespread at Finley Refuge but did not dominate forest epiphytes as it does at some Coast Range sites (Peck 1997). Two mosses that were expected but not found at Finley Refuge were Rhytidiadelphus loreus (Hed- [Vol. 48 wig) and Hylocomium splendens (Hedwig) Bruch, Schimper & Gumbel. Most of Chapman and San- born’s (1941) collections of R. Joreus were from the humid margins of the Willamette Valley, but their collections of H. splendens were on substrates and at elevations comparable to those at Finley Refuge. The absence of these species in this study may be a function of the relatively dry climate of the Wil- lamette Valley or of this particular locality. At Finley Refuge, land managed primarily for wildlife and for protection of natural systems pre- serves a variety of native and settler-altered plant communities that provide a wide range of sub- strates, resulting in a diverse and abundant bryo- phyte assemblage. ANNOTATED SPECIES LIST All substrates on which each taxon was found are listed. Moss nomenclature follows that of An- derson et al. (1990), except for that of Pottiaceae, which follow Zander (1993), and Eurhynchium, which follows Lawton (1971). Liverwort nomen- clature follows that of Stotler and Crandall-Stotler (1977). Numbers following taxa and descriptions are Merrifield collection numbers retained as vouchers in OSC. The number followed by RH is a Richard Halse collection number. Class Musci Amblystegium serpens (Hedwig) Bruch, Schimper & Gumbel var. juratzkanum (Schimper) Rau & Hervey [Amblystegiaceae]. Concrete. 932c. Antitrichia californica Sullivant in Lesquereux [Leucodontaceae]. Bark of Oregon white oak, willow sp., and black cottonwood; rotting decid- uous log, burned log, exposed basalt, shaded ba- salt, concrete. 605, 610, 1179. Antitrichia curtipendula (Hedwig) Bridel [Leuco- dontaceae]. Bark of Douglas fir and red alder, rotting vine maple; unidentified rock; all deeply shaded. //8/. Atrichum selwynii Austin [Polytrichaceae]. Stream- cut bare soil bank. /383. Aulacomnium androgynum (Hedwig) Schwaegrich- en [Aulacomniaceae]. Bigleaf maple log, uniden- tified deciduous log, basalt. 632, 65/. Brachythecium albicans (Hedwig) Schimper in Bruch, Schimper & Gumbel [Brachytheciaceae]. Soil among grass in savannah, upland grasslands, and lawn; packed soil; shaded basalt. 966, 977, 1700c. Brachythecium frigidum (C. Muller) Bescherelle [Brachytheciaceae]. Rotting bigleaf maple log, butt of rotting bigleaf maple, damp log in stream- bed, weathered unidentified rock, permanently shaded basalt in depression below soil line, shad- ed basalt above ground, clay soil among emer- gents, mud among Carex under Oregon white oak and Oregon ash, clay soil among emergents, fallen branch in mud, deer skull and antlers, 2001] Gandoderma-complex fruiting structure on log. 871, 979, 1187, 1418, 1692. Brachythecium rivulare Schimper in Bruch, Schim- per & Gumbel [Brachytheciaceae]. Moist, shad- ed basalt; dry, exposed rock; rock in flowing wa- ter; mud among Carex under Oregon white oak and Oregon ash; rotting bigleaf maple log; white- rotted standing bigleaf maple butt. 674, 677, 974, 975, 1385b. Bryum argenteum Hedwig [Bryaceae]. Totally un- shaded basalt in natural outcrop. /428. Bryum caespiticum Hedwig [Bryaceae]. Agricultur- al field soil. 1224. Bryum canariense Bridel [Bryaceae]. Basalt, con- crete, soil over concrete, soil between rocks in gravel road. 79/. Bryum capillare Hedwig [Bryaceae]. Rotting log, weathered stump roots, soil between uprooted tree roots, concrete, shaded basalt, packed soil, hummock in wet prairie. 649, 943, 1700b. Bryum dichotomum Hedwig [Bryaceae]. Packed soil at agricultural field edge, disturbed soil among grass at field edge, packed road _ soil. IZED wl D2 7D: Bryum lisae De Notaris var. cuspidatum (Bruch, Schimper & Gumbel) Margot [Bryaceae]. Con- crete, packed soil among grasses. 932b. Ceratodon purpureus (Hedwig) Bridel [Ditricha- ceae]. Exposed basalt, shaded basalt, packed soil. 789, 962. Claopodium crispifolium (Hooker) Renauld & Car- dot [Leskeaceae]. Bigleaf maple bark and weath- ered sandstone, both deeply shaded. //75. Dendroalsia abietina (Hooker) Britton [Leucodon- taceae]. Rotting deciduous and Douglas fir logs; bark of Oregon white oak, Douglas fir and black cottonwood bark. 609. Dicranella heteromalla (Hedwig) Schimper [Di- cranaceae]. Drying undrained agricultural soil. 1459b. Dicranoweisia cirrata (Hedwig) Lindberg ex. Mil- de [Dicranaceae]. Oregon white oak bark, rotting Oregon white oak log, burned log, split wood fenceposts, basalt, packed road soil. 620, 623, 1227a. Dicranum fuscescens Turner [Dicranaceae]. Rotting Douglas fir bark. 66/7, 68/. Dicranum howellii Renauld & Cardot [Dicrana- ceae]. Rotting Oregon white oak, rotting bigleaf maple stump, bigleaf maple bark. 62/, 1178. Dicranum scoparium Hedwig [Dicranaceae]. Rot- ting bigleaf maple stump, soil among grasses. 660a, 1687. Dicranum tauricum Sapehin [Dicranaceae]. Sides and cross section of rotting bigleaf maple and Douglas fir stumps. 68/a. Didymodon vinealis (Bridel) Zander [Pottiaceae]. Concrete, freshly disturbed soil, exposed basalt. 934b, 951. Didymodon sp. {Pottiaceae]. Bark of isolated Ore- gon white oaks. 1676b. MERRIFIELD: BRYOPHYTES OF FINLEY REFUGE P| Ditrichum shimperi (Lesquereux) Kuntze [Ditricha- ceae]. Shaded soil of native wet prairie hum- mock. /572. Ephemerum serratum (Hedwig) Hampe [Ephemer- aceae]. Soil of hummock in native prairie. 7570 [det. by V. Bryan]. Eurhynchium oreganum (Sullivant) Jaeger [Brach- ytheciaceae]. Vine maple bark, rotting Douglas fir log, Douglas fir butt, basalt, soil over rock. 617. Eurhynchium praelongum (Hedwig) Bruch, Schim- per & Gumbel var. praelongum [Brachytheci- aceae]. Woody debris among Carex, creekbank soil, damp upland soil, ash swale soil, shaded weathered sandstone, concrete, gravel in flowing stream, lawn, branches in stream splash zone, ninebark root in dry streambed, butt of bigleaf maple log, deer skull and antlers. 634, 9S4b. Eurhynchium praelongum (Hedwig) Bruch, Schim- per & Gumbel var. stokesii Turner [Brachythe- ciaceae]. Rotting Oregon white oak, butt of big- leaf maple log, forest floor litter. 6/3, /2/2. Fissidens bryoides Hedwig [Fissidentaceae]. Clay stream bank, freshly disturbed soil, weathered sandstone, ditch at field edge. 874, 877, 1186, 1257. Fontanalis antipyretica Hedwig var antipyretica [Fontinalaceae]. Rocks in flowing water, sub- merged tree root, ninebark root in dry streambed. LIS5. Funaria hygrometrica Hedwig [Funariaceae]. Burned soil, burned wood, weathered stump roots, agricultural field soil, mud. 835. Grimmia incurva Schwaegrichen [Grimmiaceae]. Exposed basalt. /424. Grimmia pulvinata (Hedwig) Smith [Grimmi- aceae]. Basalt, concrete. 752, 1048, 1682b. Grimmia trichophylla Greville [Grimmiaceae]. Ba- salt, concrete. 753, 942, 952, 1046, 1057, 1431. Homalothecium aeneum (Mitten) Lawton [Brachy- theciaceae]. Exposed basalt. 790. Homalothecium fulgescens (Mitten ex C. Muller) Lawton [Brachytheciaceae]. Bark of bigleaf ma- ple, hawthorne, Oregon ash, and Oregon white oak; fallen bigleaf maple branches, rotting un- identified log, rotting lumber, exposed rock, shaded basalt, concrete. 643, 666, 673a, 1054, 1690. Homalothecium nuttallii (Wilson) Jaeger [Brachy- theciaceae]. Bark of western hawthorne, black cottonwood and Oregon white oak; burned log, exposed unidentified rock, basalt, concrete, freshly disturbed soil. 6/6, 673b, 680. Hypnum circinale Hooker [Hypnaceae]. Rotting Oregon white oak and unidentified stump, Ore- gon ash bark, rotting stump, exposed unidentified rock, 625, 12/0; Hypnum subimponens Lesquereux [Hypnaceae]. Basalt, rotting Oregon ash and unidentified logs and branches, Oregon ash bark, bigleaf maple butt, shaded basalt. //83, //S8, 1691. £9) MADRONO Isothecium cristatum (Hampe) Robinson [Brachy- theciaceae]. Douglas fir bark; Douglas fir and un- identified stumps; rotting Oregon white oak, big- leaf maple, and unidentified logs. 644, 662, 1189, L210, Isothecium myosuroides Bridel [Brachytheciaceae]. Bark of Oregon ash, Douglas fir and Oregon white oak, rotting Oregon white oak and bigleaf maple logs, rotting branches of Douglas fir and Oregon ash, unidentified rock. 6/2, 1172. Leptobryum pyriforme (Hedwig) Wilson [Bry- aceae]. Freshly disturbed soil. 988d. Leptodictyum riparium (Hedwig) Warnstorf [Am- blystegiaceae]; soil in and near stream at agri- cultural field edge. 669. Leucolepis acanthoneuron (Schwaegrichen) Lind- berg [Mniaceae]. Oregon ash roots, rotting big- leaf maple and Douglas fir logs, basalt, shaded weathering sandstone, shaded soil. 635, /060. Metaneckera menziesii (Hooker in Drummond) Steere [Neckeraceae]. Bark of bigleaf maple, Or- egon ash, and Oregon white oak. /231/. Neckera douglasii Hooker [Neckeraceae]. Oregon white oak bark, Douglas fir branches, unidenti- fied rock. 6/17. Orthotrichum consimile Mitten [Orthotrichaceae]. Recently fallen Oregon ash and_ unidentified branches, burned log, basalt, concrete. 633, 11 D/G, 11S, I286, Orthotrichum lyellii Hooker and Taylor [Orthotri- chaceae]. Bark of bigleaf maple, western haw- thorne, Oregon ash, black cottonwood, Oregon white oak, burned log, basalt, shaded basalt. 6/5, 1058, 2157, 1205. Orthotrichum speciosum Nees ex Sturm [Orthotri- chaceae]. Black cottonwood bark. ///93. Philonotis fontana (Hedwig) Bridel [Bartrami- aceae]|. Side of mud hummock in wet prairie; one small collection of depauperate specimen. /569. Physcomitrella patens (Hedwig) Bruch & Schimper in Bruch, Schimper & Gumbel [Funariaceae]. Shaded mud under bridge. //56a [det confirmed by J. Christy]. Physcomitrium immersum Sullivant [Funariaceae]. Shaded mud along streams and under bridge, streamside mud. //56b [det. confirmed by W. B. Schofield]. Physcomitrium pyriforme (Hedwig) Hampe [Funar- iaceae]. Packed mud, drying undrained agricul- tural soil. /4/0. Plagiomnium ellipticum (Bridel) T. Koponen [Mni- aceae]. Clay soil under Oregon ash. 872. Plagiomnium insigne (Mitten) T. Koponen [Mni- aceae]. Rotting bigleaf maple, Douglas fir, and unidentified logs, unidentified rock near stream, forest soul, gravity-disturbed soil bank. 655c, 071,060,932, Ji 17, JI90. J279. Plagiomnium medium (Brusch & Schimper in Bruch, Schimper & Gumbel) T. Koponen [Mni- aceae]. Clay soil under Oregon ash. //59, 1/68. Plagiomnium venustum (Mitten) T. Koponen [Mni- [Vol. 48 aceae]. Bark of Oregon ash and Oregon white oak, rotting bigleaf maple and Oregon white oak logs, exposed basalt, shaded basalt, unidentified rock. 606, 622, 1050. Plagiothecium laetum Schimper in Bruch, Schim- per & Gumbel [Plagiotheciaceae]. Rotting big- leaf maple and unidentified deciduous log. 63/, 659. Pleuridium subulatum (Hedwig) Rabenhorst [Ditri- chaceae]. Freshly disturbed soil. 98Sb. Pohlia annotina (Hedwig) Lindberg [Bryaceae]. Drying undrained agricultural soil. 1/455, 156/. Polytrichum juniperinum Hedwig [Polytrichaceae]. Overgrown roadbed, exposed basalt, soil in de- pression in shaded basalt. 780, 1045. Polytrichum piliferum Hedwig [Polytrichaceae]. Exposed basalt. /434. Pterogonium gracile (Hedwig) Smith [Anomodon- taceae]. Oregon white oak bark, fallen angio- sperm bark, barkless Oregon white oak log, un- identified rotting log. 6/2, 648, 1416, 1673. Ptycomitrium gardneri Lesquereux [Ptychomitri- aceae]. Basalt. 1/043, 1OS/. Racomitrium canescens (Hedwig) Bridel [Grimmi- aceae]. Overgrown gravel roadbed, rocky soil, basalt. 779, 1044. Racomitrium heterostichum (Hedwig) Bridel [Grimmiaceae]. Dry, exposed basalt; shaded ba- salt, and unidentified rock. 672, 7858, 1385, 1433, 1696. Racomitrium occidentale (Renauld & Cardot) Re- nauld & Cardot [Grimmiaceae]. Partially shaded basalt. /053. Rhizomnium glabrescens (Kindberg) T. Koponen [Mniaceae]. Unidentified rotting log. ///4. Rhytidiadelphus triquetrus (Hedwig) Warnstorf [Hylocomiaceae]. Deciduous sapling bark, Ore- gon ash bark, forest floor litter, shaded basalt, soil over rock. 6/9. Sanionia uncinata (Hedwig) Loeske [Amblystegi- aceae]. Rotting Oregon ash branch, shaded wooden bridge support, shaded deer skull and antlers on forest floor. 1/67, 1382. Schistidium apocarpum (Hedwig) Brusch & Schim- per in Bruch, Schimper & Gumbel [Grimmi- aceae]. Shaded basalt. /689. Scleropodium cespitans (C. Muller) L. Koch [Brachytheciaceae]. Oregon white oak upper branch and butt bark. /284b, 1678, 1681, 1683. Scleropodium obtusifolium (Jaeger) Kindberg in Macoun & Kindberg’ [Brachytheciaceae]. Streambed gravel. 9S4a, 972, 1220. Scleropodium touretii (Bridel) L. Koch var touretii [Brachytheciaceae]. Packed soil, disturbed loose forest soil, shaded weathered sandstone, soil among road rocks. 960, 976, 985, 1195. Syntrichia laevipila Bridel var. laevipila [Potti- aceae]. Bark of Oregon white oak in small, open group in agricultural fields and savannahs. With the var. laevipila, 1198, 1684a, 1685a, 1685b. Syntrichia laevipila’ Bridel var. meridionalis (Schimper) Juratzka [Pottiaceae]. Bark of isolat- | ed Oregon white oak in agricultural fields and | savannahs, with var. laevipila. 1198, 1684a, — 1685a, 1685b [det. confirmed by J. Harpel]. Syntrichia latifolia (Hartman) Hubener [Potti- aceae]. Oregon white oak bark, shaded concrete. _ 946, 1197, 1684b, 1685b. ‘Syntrichia princeps (De Notaris) Mitten [Potti- ' aceae]. Oregon white oak bark, weathered stump roots, basalt, burned log, concrete. 930, 951, 1049, 1284a. Syntrichia ruralis (Hedwig) Weber & Mohr [Pot- ' tlaceae]. Basalt, Oregon white oak bark, es- _ pecially in savannahs. 1/059, 1677b, 1685c, 1686. Tetraphis pellucida Hedwig [Tetraphidaceae]. Rot- _ ting bigleaf maple log. 658. Timiella crassinervis (Hampe) L. Koch _ [Potti- aceae]. Freshly disturbed partially shaded soil. 988c. Tortula atheroides Zander [Pottiaceae]. soil, freshly disturbed soil. 963, 990. Tortula muralis Hedwig [Pottiaceae]. Concrete. 954. Trachybryum megaptilum (Sullivant) Schofield [Brachytheciaceae]. Disturbed forest soil, forest floor litter, exposed rotting wood and adjacent disturbed soil, exposed concrete. 950, 98/a. Zygodon viridissimus (Dickson) Bridel var. viridis- simus [Orthotrichaceae]. Savannah oak _ bark. 1675a, 1677a. Packed Class Hepaticae Cephalozia bicuspidata (L.) Dumortier [Cephalo- ziaceae]. Rotting unidentified log, among mosses on basalt, on and among Didymodon sp. on sa- vannah oak bark. 882. Cephalozia lunulifolia (Dumortier) Dumortier [Ce- phaloziaceae]. Rotting bigleaf maple log. 655a. Cephaloziellaceae divaricata (Smith) Schiffner [Cephaloziellaceae]. Rotting bigleaf maple log. 654a. Chyloscyphus polyanthos (L.) Corda var. polyan- thos [Lophocoleaceae]. Bare soil in Oregon ash swale, on soil among Carex sp. under Oregon white oak and Oregon ash. 870. Chyloscyphus polyanthos (L.) Corda var. rivularis (Schrader) Nees [Lophocoleaceae]. Streambed. 668. Fossombronia wondraczekii (Corda) Dumortier [Codoniaceae]. Drying undrained agricultural soil. 1560. Frullania bolanderi Austin [Jubulaceae]. Oregon white oak bark. 624. Frullania californica (Austin) Evans [Jubulaceae]. Oregon white oak bark. 925. Frullania tamarisci (L.) Dumort ssp. nisquallensis (Sullivant) Hattori [Jubulaceae]. Oregon ash bark, Douglas fir bark. 636, 1115. Jungermannia rubra Gottsche ex Underwood [Lo- phoziaceae]. Streambank mud. 878. MERRIFIELD: BRYOPHYTES OF FINLEY REFUGE 2S Lepidozia reptans (L.) Dumortier [Lepidoziaceae]. Rotting stump. /209. Lophocolea bidentata (L.) Dumortier [Lophocole- aceae]. Rotting bigleaf maple; rotting unidenti- fied logs and branches. 653, /38/. Lophocolea cuspidata (Nees) Limpricht [Lopho- coleaceae]. Rotting unidentified log. S8/. Lophocolea heterophylla (Schrader) Dumortier [Lophocoleaceae]. Rotting bigleaf maple log, rotting unidentified stump. 655b, 12/3. Marchantia polymorpha L. |Marchantiaceae]. Per- petually wet concrete, soil in ash swale, moist disturbed agricultural soil. 722/, RH 4729. Porella cordeana (Hubener) Moore [Porellaceae]. Among Rhytidiadelphus triquetrus among rocks over soil, concrete in ephemeral streambed. 987, 1384. Porella navicularis (Lehmann et. Lindenberg) Lindberg [Porellaceae]. Bark of Oregon white oak, Oregon ash, and Douglas fir; exposed ba- salt; shaded basalt. 603, 1/16. Porella roellii Stephani [Porellaceae]. Oregon white oak bark, rotting bigleaf maple log. 604, 613. Radula bolanderi Gottsche [Radulaceae]. Douglas fir bark. 645. Radula complanata (L.) Dumortier [Radulaceae]. Oregon ash bark; vine maple bark; rotting branch. 640, 685, 968a. Riccardia multifida (L.) S. Gray [Aneuraceae]. Crumbling undercut creekbank soil. 876. Riccia spp. [Ricciaceae]. Undrained agricultural soils. 1565a, 1565b. Riccia fluitans L. [Ricciaceae]. Streamside mud. 1d 53: Ricciocarpus natans (L.) Corda [Ricciaceae]. Free- floating, shaded mud. 869, S84. Scapania bolanderi Austin [Scapaniaceae]. Rotting stump on north-facing forested slope. /208. Scapania umbrosa (Schrader) Dumortier [Scapani- aceae]. Rotting Oregon white oak branch on north-facing forested slope. 1437. Class Anthocerotae Anthoceros sensu latu sp. [Anthocereotaceae]. North underhand of Juncus hummock in native wet prairie. 1568. ACKNOWLEDGMENTS R. Halse and J. Wheeler tirelessly curated Finley Ref- uge specimens for OSC. S. Settles, T. Melanson, G. Hughes, S. Crandall, J. Houk, and the especially helpful M. Naughton and L. Devaney of the U. S. Fish and Wild- life Service Western Oregon Refuge Complex, granted collecting permits and were gracious hosts. D. H. Wagner provided the current list of Oregon liverworts. J. Christy and D. H. Wagner provided encouragement. J. Christy, W. Schofield, V. Bryan, J. Wheeler, and J. Harpel confirmed or corrected difficult identifications. R. Halse, J. Shevock, K. Schierenbeck, and anonymous reviewer provided con- 24 structive comments on earlier drafts of this paper. I am grateful to them all. LITERATURE CITED ANDERSON, L. E., CRUM, H. A., AND Buck W. R. 1990. List of the mosses of North America north of Mexico. The Bryologist 93:448—499. CHAPMAN, C. J. AND E. I. SANBORN. 1941. Moss Flora of the Willamette Valley, Oregon. Oregon State Mono- graphs. Studies in Botany Number 4. Oregon State College, Corvallis, OR. Curisty, J. A. 1980. Additions to the moss flora of Ore- gon. The Bryologist 83:355—358. , J. H. Lyford, and D. H. Wagner. 1982. Checklist of Oregon mosses. The Bryologist 85:22—36. AND D. H. WAGNER. 1996. Guide to the identifi- cation of rare, threatened and sensitive bryophytes in the range of the northern spotted owl, western Wash- ington, western Oregon, and northwest California. Bureau of Land Management, Portland, OR. CONARD, H. S. 1944. Physcomitrium immersum in Oregon. The Bryologist 47:41. CruM, H. A. AND ANDERSON, L. E. 1981. Mosses of East- ern North America, Vols. | and 2. Columbia Univer- sity Press, NY. FEMAT (FOREST ECOSYSTEM MANAGEMENT ASSESSMENT TEAM). 1993. Forest ecosystem management: an eco- logical, economic, and social assessment. U. S. Gov- ernment Printing Office, Washington, DC. FRANKLIN, J. EF AND C. T. DyRNEss. 1973. Natural Vege- tation of Oregon and Washington. U. S. Forest Ser- vice. HABECK, J. R. 1961. The original vegetation of the mid- Willamette Valley, Oregon. Northwest Science 35: 65-77. Howe, M. A. 1899. The Hepaticae and Anthocerotes of California. Memoirs of the Torrey Botanical Club, Vol. VII. JOHANNESSEN, C. L., W. A. DAVENPORT, A. MILLET, AND S. MCWILLIAMS. 1971. The vegetation of the Willam- ette Valley. Annals of the Association of American Geography 61:286—302. LAWTON, ELVA. 1971. Moss Flora of the Pacific North- west. Hattori Botanical Laboratory. 3888 Honmachi, Nichinan-shi, Miyazaki-ken, Japan. MERRIFIELD, K. 2000. Bryophytes on isolated Quercus garryana trunks in urban and agricultural settings in MADRONO [Vol. 48 | the Willamette Valley, Oregon. The Bryologist 103: | 720-724. PALMER, R. P. 1976. Handbook of North American Birds. Vol. 2. Waterfowl (first part). Yale University, New | Haven, CT. | Peck, J. E. 1997. Commercial moss harvest in northwest- | ern Oregon: Describing the epiphyte communities. | Northwest Science 71:186—195. Pike, L. H. 1973. Lichens and Bryophytes of a Willamette | Valley Oak Forest. Northwest Science 46:149-158. 402. RAMBO, T. R. AND P. S. Muir. 1998. Forest floor bryo- | phytes of Pseudostuga menziesii-Tsuga heterophylla | stands in Oregon: Influences of substrate and oversto- | ry. The Bryologist 101:116—130. SANBORN, E. I. 1929. Hepaticae and Anthocerotes of West- | ern Oregon. University of Oregon Plant Biology Se- | ries Vol. I, No. 1. University Press, Eugene, OR. Benton County, OR. The Bryologist 44:111—114. SCHOFIELD, W. B. 1976. Bryophytes of British Columbia | III: habitat and distributional information for selected | mosses. Syesis 9:317—354. SCHUSTER, R. M. 1969. The Hepaticae and Anthocerotae of North America east of the hundredth meridian, | Vol. If. Columbia University Press, NY. lumbia University Press, NY. mer, Vaduz. SLACK, N. G. 1977. Species diversity and community | , W. C. DENISON, D. M. Tracy, M. A. SHERWOOD, | AND F M. RHOADES. 1975. Floristic survey of epi- | phytic lichens and bryophytes growing on old-growth | conifers in western Oregon. The Bryologist 78:389- . 1941. The bryophyte flora of Alsea Mountain, | . 1974. The Hepaticae and Anthocerotae of North America east of the hundredth meridian, Vol. III. Co- | . 1977. Boreal Hepaticae: A Manual of the Liv- erworts of Minnesota and Adjacent Regions. J. Cra-_ structure in bryophytes: New York State studies. New | York State Museum Bulletin 428. SmiTH, A. J. E. 1990. The Liverworts of Britain and Ire- land. Cambridge University Press, Cambridge. STOTLER, R. AND B. CRANDALL-STOTLER. 1977. A checklist © of the liverworts and hornworts of North America. ° The Bryologist 80:405—428. Vitt, D. H., J. E. MARSH, AND R. B. Bovey. 1988. Mosses, | Lichens, and Ferns of Northwest North America. Lone Pine Publishing, Edmonton, Alberta; University — of Washington Press, Seattle, WA. LANDER, R. H. 1993. Genera of the Pottiaceae: mosses of | harsh environments. Bulletin of the Buffalo Society of Natural Sciences. Vol. 32. Buffalo, NY. EVIDENCE FOR A SAND HILLS ECOTYPE OF ESCHSCHOLZIA CALIFORNICA (PAPAVERACEAE) ERIN K. ESPELAND! AND RODNEY G. MYATT Department of Biological Sciences, San Jose State University, San Jose, CA 94192 ABSTRACT The Eschscholzia californica Cham. growing in the Ben Lomond sand hills in the Santa Cruz mountains appears different from other populations in the surrounding areas. To quantify these differences, popu- lations in coastal, sand hills, and serpentine soils were compared. Phenotypic data were collected from the field in 1998 and 1999. A common garden experiment was also performed. Differences in torus rim width, flower color, flower number, and leaf coloration were examined in both the field and the common garden. The sand hills population was different from both the coastal and the serpentine populations for leaf coloration. Population differentiation occurred among all three sites for at least some of the floral characters examined. The common garden experiment indicates that some of the differences among the populations are phenotypically stable. INTRODUCTION The Ben Lomond sand hills are a textbook ex- ample of the geographic factors that give rise to endemism (Mayer et al. 1994; Kruckeberg 1986). Isolated intrusions of dry sandy soil and associated drought-adapted plant communities are in stark contrast to the damp redwood forest surrounding the sand hills. Edaphic habitat disjunction can cause parapatric speciation (Kruckeberg 1954; Proctor and Wodell 1975). In the case of Eschscholzia cal- ifornica Cham., a species that is found in open, disturbed habitats, populations in the sand hills may be isolated by large regions of dense forest habitat, eventually leading to allopatric speciation. Isolation and soil-specific adaptation can produce a great de- gree of population differentiation in a short period of time (Proctor and Wodell 1975). A more common example of edaphic factors as- sociated with endemism in the California flora is the presence of serpentine-adapted communities throughout the state. Serpentine adaptation is not treated consistently in taxonomic terms. In some cases, a serpentine-adapted group of populations is recognized as a separate subspecies (e.g., Streptan- thus insignis Jepson ssp. lyonii Kruckeb. & J. Mor- rison) and, in other cases, it is grouped with ser- pentine-intolerant populations (e.g., Streptanthus glandulosus Hook.) (Kruckeberg 1986). Eschscholzia californica is known for its local variation. By the early part of this century, over 90 varieties of had been described (Greene 1905). The 90 varieties were reduced to four by Munz and Keck (1968). They recognized coastal, central, southern, and dune varieties in California. Es- chscholzia californica is known for its plasticity ' Present address: Environmental Restoration Division, Lawrence Livermore National Laboratory, Livermore, CA 94551. and is now found in open, semi-disturbed habitats all over the globe. Cook (1961) performed a survey of E. californica over the state of California and found local differentiation in self-compatibility, flower fertility, seed production, and stamen num- ber. He found a graded mosaic pattern in the dis- tribution of nearly all the phenotypic characters he measured. Although his work focused on popula- tions west of the Central Valley, E. californica in Santa Clara and Santa Cruz counties were not in- cluded. This study examines local differentiation among 3 populations in 3 different habitats in the south San Francisco Bay Area: a serpentine habitat, an inland sand hill habitat, and a coastal meadow habitat. This study looks at differentiation among the populations in two ways: field phenotypic mea- surement (how the plants appear in their own hab- itat) and common garden phenotypic measurement (controlling environmental influence). The plants at the sand hills were of particular interest due to their vibrant purple leaves with large white spots. While the species is widely recognized to be variable in terms of floral morphology and growth habit, leaf color variation has not been formally described in this species before. Local differentiation is a much-studied phenom- enon because of its contribution to evolutionary and conservation theory (Waser and Price 1985; Mon- tagnes and Vitt 1991; Mayer et al. 1994; Kindell et al. 1996; Linhart and Grant 1996). Finding how much field-observed population differences are in response to environmental factors requires some ex situ investigation. Common garden experiments are effective in determining the strength of the adap- tation of each population to its own soil environ- ment. By performing a common garden experiment rather than a reciprocal transplant experiment, we focused purely on soil environment and controlled confounding factors such as unequal responses to %G MADRONO differences in interspecific competition and preda- tion that may have existed within each population. Also, by performing a common garden experiment rather than a reciprocal transplant experiment, we avoided the likelihood of polluting a possibly sen- sitive habitat-adapted genome (at the sand hills) with detrimental alleles. Recognition of special, locally-adapted popula- tions of plants has proved valuable in conservation of fragile ecosystems. Most rare plant occurrences in California are in mixed chaparral, grasslands, coastal scrub, and valley-foothill woodland, yet in terms of the percentage of total habitat preserved (wilderness areas, research reserves, national and state parks, wildlife refuges, and recreational sites) alpine and sub-alpine areas are afforded propor- tionally the most protection (Pavlik and Skinner 1994). While the identification of a rare plant in a previously unknown area can ensure some protec- tion (Bartel et al. 1994), Pavlik and Skinner (1994) recommend promoting habitat-based conservation plans, particularly in serpentinite, rocky, and sandy substrates at non-alpine elevations that harbor the highest degrees of overall endemism. The sand hills are certainly a candidate for this type of protection: California Department of Fish and Game has ex- pressed interest in this area, and 9 of the 97 species found in the sand hills are thought to be ecotypes, with 8 additional species probably warranting ad- ditional taxonomic study (Lee 1996). METHODS Study sites. In 1998, one coastal, one sand hills, and one serpentine population were sampled. In 1999, two coastal, two sand hills, and one serpen- tine population were sampled. Both coastal sites were located at Wilder Ranch State Park, just north of Santa Cruz. The sand hills sites were located in a watershed ravine just off Mt. Hermon Road in Scotts Valley. The property is privately owned and previously contained a sand quarry. The serpentine site 1s located on Tulare Hill in the Santa Teresa Hills in the city of San Jose. The property is leased for cattle grazing. The coastal site sampled in both 1998 and 1999 (W 1) is a tabletop meadow 0.4 km southwest of the historic ranch buildings. The site is mowed every summer, late in the flowering season. Bromus hord- caccus L., Raphanus sativus L., and Carduus pyc- nocephalus L. dominate this site. The coastal site sampled in 1999 only (W2) was located 0.4 km east of the historic ranch buildings. The species com- position of this south-facing community was dom- inated by Bromus diandrus Roth and C. pycnoce- phalus. Although the sites are coastal, the E. cali- fornica at this location does not fit the description of the coastal ‘“‘race”’ of E. californica (described in Munz and Keck 1968), but rather the inland va- riety. Due to the proximity of the sites to the ranch house and gardens, it is possible that these popu- lations are descendants of once-planted individuals. [Vol. 48, The sand hills E. californica population sampled| in 1998 and 1999 (SH1) occurred on both the south-facing and north-facing sides of the ravine. The south-facing slope community contained a grassland understory (Briza maxima L. dominant) beneath scattered ponderosa pine. The north-facing, slope community consisted of spring flowering an-| nuals (Lupinus bicolor Lindley, Gilia tenuiflora’ Benth., Castilleja exserta [A A Helbr.] Chuang &| Heckard) spread thinly over bare sand. The sand hills site sampled in 1999 only (SH2) was on the, lower sloping face of a sand quarry scar 0.2 km south of the first sand hills population. The scar’ bends in a semi circle from south-facing to west-. facing slopes. Plants at SH2 flowered 5 weeks ear- lier than those at SH1, and the community consisted. almost entirely of E. californica and Lupinus albi- frons Benth. : The E. californica population on serpentine soil . (SERP) was sampled in both 1998 and 1999. The’ population area is located on north- and east-facing | slopes in a serpentine grassland community con- taining species such as Avena fatua L., Hordeum murinum L. ssp. leporinum and Lasthenia platyg-' lossa. | Populations located on the same soil type were close enough together that gene flow between them. is a distinct possibility. Differences in plant com-. munity composition and extremely low densities of E. californica plants between the populations indi- cated that separate treatment might be appropriate. In all cases, the entire population of E. californica was not measured, but rather a high-density lobe within the population was sampled. Individual plants of E. californica could be found at low den- sities for thousands of meters from the sampling. areas. Sampling areas were defined by high E. cal- ifornica concentration (generally more than | plant per m’), and also by artificial (fences, drainage ditches) and geologic (cliffs and other drastic changes in slope) barriers. Field data collection. In 1998, 3 populations were sampled for plant phenotypic characteristics. Population W1 was sampled on June 26; SERP was sampled on May 1; SHI was sampled on July 1. An effort was made to sample each population when the largest numbers of plants were flowering. Between 45 and 65 plants were randomly sampled at each population. Data collected for each plant were the number of floral units, leaf color, torus rim width, and flower 1|-color or bicolor. Floral units were recorded as the number of flowers plus the number of buds plus the number of capsules. Torus rim width and flower color were measured on the tallest flower. Torus rim width was taken at the wid- est point and was recorded as either 1, 2, or 3 mm. Flower color was recorded by comparing the base to the tip of the petals on the flower on the tallest branch. Obvious differences between the two were considered to be evidence of the “‘bicolor”’ 2001 | character in the flower. If there appeared to be only a slight difference between the color of the tip of the petal compared to the base of the petal, or no difference at all, the flower was recorded as “I- color.” Leaf color data were collected by observing the color characteristics of the leaf: presence of white spots, visible presence of green pigmentation, vis- ible presence of purple pigmentation, and presence of red tips on the leaf. Red-tipped leaves appear to be fairly common in E. californica, and almost all living leaves of the plant can be characterized as green. However, some leaves of E. californica have a purplish tinge that is found throughout the leaf. Where green was visible and tinged with purple, both the green leaved and the purple leaved char- acter were marked as present. In some plants this purple color is so vibrant that no green color can be seen. In these plants, the purple-leaf character was recorded as present, and the green-leaf char- acter was recorded as absent. In 1998, each char- acter was noted as either present or absent for the third leaf from the top of the longest branch of each randomly selected plant. Stanton Cook (1961) used flower color and torus rim width in his investigations of statewide vari- ability in E. californica. Clark and Charest (1992) used number of floral units in their study of popu- lation differentiation in the Antelope Valley. Leaf color was of interest because of the apparent dif- ference between the sand hills population for this character and the other two populations. In 1999, 15 plants per population were sampled for floral characters to see if the differences ob- served among populations were consistent and if they would be present in two very different climatic years: El] Nino and La Nina winters (NOAA 2000). Number of flower units, torus rim width and flower color were measured over time at sites W1, SHI, and SERP to make sure that differences observed in 1998 would be consistent within the season. Fif- teen plants were sampled in each population every 2 weeks. The same sampling scheme was used for each data collection. Although it is unlikely that the same plants were measured on each sampling date, the plants were located in roughly the same areas. Sampling dates were April 1, April 15, April 28, May 12, May 27, and June 10. Floral and leaf character data were collected at the estimated peak of flowering for W1 (April 15), W2 (May 13), SH1 (May 27), SH2 (April 22), and SERP (April 28). Instead of just looking at a single leaf of the plant, all of the leaves of the plant were examined for leaf color characters in 1999. Because no differences were found for this character in 1998, data on the red-tipped leaf character were not collected in 1999. Common garden experiment. Seeds were col- lected from W1 and SHI in July 1998 from 30 randomly selected plants in each population. The ESPELAND AND MYATT: E. CALIFORNICA ECOTYPE rH TABLE |. DESIGN OF THE COMMON GARDEN EXPERIMENT. Soil source Seed source # pots # seeds/pot SH1 SH1 16 9 Wi SH1 16 9 SH1 W | 1] 6 Wi WwW! 1] 6 seeds, identified by parent plant, were stored in pa- per envelopes until February 1999, at which time the seeds were placed in 10 cm X 10 cm pots filled with soil from the W1 and SHI population areas. Only ripe seeds were used; collected seeds that were unripe were not used in the experiment. A total of 210 seeds from 11 plants from SHI and 210 seeds from 6 plants from W1 were used in the experiment. In all but 2 pots, seeds from only 1 parent were sown per pot. Due to seed from some SH parents not germinating, common garden data were collected from progeny of 8 (out of 11) SH parents and all 6 W parents. Soil was collected from locations near, but not within the SH! and W1 populations. This limited the possibility of contamination with seed from population seed banks. In order to identify whether contamination occurred despite our efforts, seeds were planted in an X-shaped pattern in each pot. Soil was collected within a week of seed potting to minimize biotic changes in the soil that might occur as a result of storage method. Neither SERP soil nor SERP seed was used in the common garden experiment because of the lack of differentiation of this population from the W1 population in the 1998 results. A reciprocal planting design allowed for seed from each population to be planted in its own and the other’s soil (Table 1). Each pot was labeled with a unique 3-digit randomly generated number to pre- vent bias when measuring plant characters. Record- ed measurements were matched to the plant parent population only at the completion of the experi- ment. A wick-based watering system was used, where the pots were placed on moist quilt batting dipped in tap water. This kept the flow of water to the base of the pots consistent, and pots were able to take up as much water as was transpired or evaporated. No extra fertilization was used. The pots were placed on the rooftop of Duncan Hall at SJSU for the months of February through April. In late April, the rooftop was so hot and sunny that the watering system was unable to keep up with the plants’ water needs without daily attention. The pots were moved to a northwest-facing patio in Pacifica for the re- mainder of the experiment. Germination and leaf color were monitored on a weekly basis. Pots were thinned to one plant per pot by retaining the plant growing closest to the center of the pot. In cases where several plants were 28 MADRONO 100% .-——-—__—- at ~ | o | a | n x © Purple xk eo | mGreen * s 13% White Spots 4 | S°) a ise) a oe) _— x < 50% = 5 a3 S = om | 5B 25% + —_— | oO 1S) = >) ae 0% SERP 98-99 SH 98-99 W 98-99 | Fic. 1. Percent presence of leaf characters for all populations, 1998 and 1999 combined. *SH populations different | [Vol. 48 from W and SERP populations (P < 0.01). **AII 3 soil types different from each other (P < 0.01). equidistant from the center of the pot, the largest plant was kept. Leaf color data were collected from all the leaves of each plant. When plants flowered, torus rim width data were collected from all flow- ers. Data analysis. The majority of statistical tests were performed in SAS (1990 Ver. 6.0) using the general linear model. G-fit tests performed in Mi- crosoft Excel (1998 Ver 7.0) were used to test for differences among populations in frequencies of discrete characters (Sprinthall 1987). Discrete char- acter frequencies were normalized before analysis. The Bonferroni correction was used in deciding the critical value (adjusted from alpha = 0.05) to min- imize the risk of coming upon a chance difference between populations because of the number of characters being compared. A repeated measures MANOVA (SYSTAT 1992 Ver. 5.2.1) was used to define differences over time in 1999 data for torus rim width and number of floral units per plant. All percentage data were arcsine transformed prior to analysis. RESULTS Field data collection. To show the total variabil- ity in expression of leaf color characters in 2 very different climatic years, leaf color data was com- bined over 1998 and 1999 for analysis. All 3 pop- ulations differed from each other in the purple-leaf character. Purple leaves were much more prevalent in SH populations than in either the SERP or W populations (Fig. 1), but the SERP population had significantly higher number of leaves with purple than the W populations (P < 0.01). All plants at SERP and W sites in both years had leaves with visible green, but many plants at SH sites lacked visible green pigment. Almost all plants at SH sites had white-spotted leaves, but less than a third of | plants at the other sites had this leaf coloration character. While SERP and W populations were similar for green-leaf and white-spotted-leaf char- acters, SH populations were different from both for these characters (P < 0.01). No differences were found in the distribution of the red-tipped leaf char- acter in 1998. Differences in floral characteristics were found among populations in 1998 (Table 2). The torus rim was much narrower in the SH populations than in the W and SERP populations (P < 0.01). The bi- color flower character was less prevalent in the SH population in 1998 than in the other two popula- tions (P < 0.01). In 1998, SH sites also tended to have more floral units than the other two site types (R= O:01). Floral measurements were taken over time in 1999 to test the validity of a single-date sampling scheme in 1998. If the characters changed over time, differences observed in 1998 could be attrib- uted to the date of sampling. No differences were found in torus rim width over time (P = 0.14). There was a statistical interaction between mea- surement date and population site for the bicolor flower character (P < 0.002, Fig. 2): the number of bicolor flowers increased over time in the SERP population (P < 0.01), but remained about the same in the other two populations, although the dip in 2001 J # Flowers Bicolor (/15) Day1 Day15 Fic. 2. Day28 Number of bicolor flowers (7 = 15) for W1, SH1, and SERP populations over time in 1999. Normalized ESPELAND AND MYATT: E. CALIFORNICA ECOTY PE 29 1998 q population B 4 values mo ’ , —-@--W1 —a— SHI ---O1-- SERP Day 42 Day 54 values from 1998 are included. Change over time significantly different among all 3 populations (P < 0.01). the W population in week 2 is not a random effect (P < 0.01). The number of floral units per plant increased over time in all populations, but increased more slowly in the SERP population versus the oth- er two populations (Fig. 3). Since torus rim width did not change over time in 1999, all collection dates and years were lumped to confirm the differences among populations. There was a year by site statistical interaction for this character (P < 0.0001). Torus rims were larger in 1999 compared to 1998, but the populations kept their size difference relative to each other: SH rims were the smallest, W rims were the largest, and SERP rims were intermediate (Table 2). Common garden experiment. No E. californica germinated outside the X-shaped seed planting pat- tern. This does not prove the soil was uncontami- nated with an existing seed bank, but such contam- ination is unlikely. Leaf phenotype remained con- stant throughout the experiment: once true leaves emerged, they did not change color over time. All SH progeny had purple leaves, regardless of the soil type in which they were grown (Fig. 4). Two W progeny had purple leaves when grown in SH soil, but this distribution was not very different from W in the field. Green leaves were more often absent from SH progeny, again regardless of soil type. All SH progeny had white-spotted leaves, re- gardless of soil type, where this character was only sometimes present in W progeny. All plants planted in W soil flowered, regardless of parent type, but only 7 of the 11 SH plants flow- ered in their own soil, and only | of the 7 W plants flowered in SH soil. Torus rim width (Table 3) is influenced by parent population (P < 0.001), indi- vidual parent plant (P < 0.001), and soil type (P < 0.0162): rim width for SH plants was less than for W plants in both soil types, however, torus rims from both parent sources were smaller in SH soil compared to W soil. DISCUSSION The differences observed between the popula- tions in terms of the bicolor characteristic and the number of floral units in 1998 were due to the sin- gle-date sampling scheme. As shown from the 1999 data, the populations do differ in their expression of these traits, but with the single-date sampling scheme in 1998, lack of difference could just as easily have been observed. This change in some traits over time should serve as a cautionary note to other researchers who plan to sample populations only once per year. The change in flower color and number of floral units over time is not a similarly- expressed trait: the type of change over time varies among populations. It is difficult to say what affects this change over time in some populations but not others. Different types of drying patterns and dif- ferent soil chemistries at the population sites could be responsible, or these differences could be due to differential responses among the populations to the same environmental factors. Since E. californica 1s 30 MADRONO Average number of flower heads Day 1 Day 15 Fic. 3. --@- WI —a— SHI ----- SERP 1998 population values Day28 Day 42 Average number of flowers per plant for W1, SH1, and SERP populations over time in 1999. Single values from 1998 are included. Change over time significantly different among all 3 populations (P < 0.01). a plant that has an indeterminate flowering system, the dip seen at Day 15 in number of floral units in W and SH populations is puzzling. We would ex- pect the number of floral units to increase through- out the season. It is possible that a subpopulation of early-flowering plants was sampled at Day | and that Day 15 represented the start of the season for a later-flowering subpopulation. Torus rim width does not appear to change over the flowering season and differences observed be- tween the populations in 1998 were confirmed when the data for the two years were analyzed to- gether. Torus rim width did change between years, and while this indicates some plasticity for this character, it also indicates that the variability is con- strained differently within each population. In the common garden experiment, SH plants kept their purple and white coloration even when planted in Wilder Ranch soil. All SH plants had purple and white coloration even though, while the expression of white coloration is close to 100% in TABLE 2. the field, the expression of purple coloration in the field is lower. Progeny from only 8 SH plants were used in the common garden experiment, so it is likely that this difference between common garden and field is due to small sample size. It is clear that the expression of leaf coloration characters that contribute to population differentiation in E. cali- fornica is phenotypically stable. The purple-leaf, white-spotted characteristic of sand hills plant leaves does not appear to be a direct response to soul type. Although the differences in leaf color are phenotypically stable, they may not be genetic. Ef- fects of the maternal environment in which the seed ripens have been known to include everything from seed germination rates to progeny plants’ tolerance to saline environments (see Rossiter 1995 for re- view). Although leaf color in particular has not been shown to be determined by maternal environ- ment, seed gathered from controlled pollinations of the potted plants should be grown to determine the FLORAL CHARACTERISTICS. All averages + 1 SD. !' For 1998 and 1999 data combined, both populations from each soil type were included. *>* For each row, values with different letters are statistically different (P < 0.01). # floral units 1998 W'! 5.90 + 5.052 % big color 1998 92,.0# Torus rim (mm) 1998 42 e072 1998 and 1999 combined 2.34 + 1.042 1998 n = 50 141 _SH! SERP 12.78 + 15.9» ALS 3.273 65.3° 05.5% 1.04 + 0.2° 1.41 + 0.062 114 O47" 5222 O62" n= 49 n = 66 n= 142 n= 143 1998 and 1998 i = [Vol. 48 | Percent of plants with leaf color character present ESPELAND AND MYATT: E. CALIFORNICA ECOTY PE 31] aN N 7 : Y 7] ] 7 4 Green Purple White spots SN SH W soil plant - SH plant - SH soil W plant - W soil W plant - SH soil Plant parent population - soil type Fic. 4. Leaf color characteristics in the common garden experiment. strength of any effects of maternal environment upon leaf phenotype. Population differentiation is a relatively common phenomenon, but differentiation among populations is expected to be less strong when the flowers are large and the plants outcross than when the flowers are small and the plants are autogamous (Linhart and Grant 1996). Floral characteristics are less like- ly to be divergent among populations than vegeta- tive ones (Slentz et al. 1999), as usually even the most disparate populations still have the same pol- linator species, and thus floral characters tend to be uniformly selected. On the other hand, the charac- ters measured in this study may not be undergoing selection at all. While a few leaves with purple coloration and white spots were found at W and SERP locations, no plants at these locations had the vibrant purple TABLE 3. TORUS RIM WIDTH IN COMMON GARDEN EXPER- IMENT BY PARENT POPULATION AND SOIL TyPE. All averages = 1] SD. Parent popula- Torus rim width tion Soil type (mm) Total SH W 1:07 = 0.53 fi SH SH 0.54 + 0.48 7 W W 2.91 + 0.79 5 W SH 2.00 coloration masking the visibility of green pigments in their leaves. The sand hills are host to many endemic species (Lee 1996), probably adapted to the particular edaphic environment. The phenotypic stability of leaf-color differences and the influence of parent population on torus rim width indicate that the sand hills population is probably geneti- cally differentiated from the other two populations in this study. Proximate causes of this differentia- tion are still a mystery: is this differentiation main- tained by lack of gene flow into the population, or solely by strong selective forces at the population site? Even though E. californica 1s a near-obligate out- breeder, pollinated by insects and wind which can carry pollen over a considerable range, Cook (1961) found that EF. californica populations can be differentiated in as small a scale as hundreds of feet. Differentiation has also been found at distanc- es of less than 3 km (Clark and Charest 1992). The balance between phenotypic plasticity and differ- ential adaptation is not known in this species. Both Cook’s and Clark and Charest’s results were from field observations: only self-compatibility was test- ed in a common garden experiment by Cook. It is certainly possible that the species is not as plastic as once thought (i.e., the same genes reacting dif- ferently to different environments), but instead adapts locally (different genes in different environ- ments) to exhibit the great variety we observe. A 32 MADRONO confounding factor to further investigation of local adaptation is the widespread use of EF. californica cultivars in landscaping and the success of the spe- cies at colonization. Future studies may wish to at- tempt to determine the extent of cultivar introgres- sion for older populations (Clark and Charest 1992); This study was designed to determine if the pop- ulations at the sand hills at the very least represent an ecotype of the EF. californica species (sensu Toresson 1922 and Kruckeberg 1951). A sand hills ecotype may be indicated by these results. Testing the interfertility of sand hills with surrounding pop- ulations and examining the possibility and efficacy of gene flow to the sand hills population will give more information on how the leaf coloration of the E. californica at the sand hills is maintained. ACKNOWLEDGMENTS Drs. Chris Brinegar, Wayne Savage, Curtis Clark, and an anonymous reviewer gave helpful comments on the manuscript. Other support was received from Kathy Hyde, Mark Fisher, Hilair Chism, Peg Edwards, and California State Parks. The SJSU Nelson Research Fellowship and the California Native Plant Society Santa Clara Valley chapter’s Dudleya-Serpentine Habitat Scholarship pro- vided financial assistance. LITERATURE CITED BARTEL, J. A., M. W. SKINNER, AND J. C. KNIGHT. 1994. The Endangered Species Act and rare plant protection in California. Pp. 30—33 in M. W. Skinner and B. M. Pavilk (eds.), California Native Plant Society Inven- tory of Rare and Endangered Vascular Plants of Cal- ifornia, 5th ed. California Native Plant Society, Sac- ramento, CA. BEATTY, A. V. 1936. Genetic studies on the California poppy. Journal of Heredity 27:331-338. CLARK, J. C. AND N. A. CHAREST. 1992. Comparison of annual and perennial Eschscholzia californica (Pa- paveraceae) at the Antelope Valley California Poppy Reserve. Crossosoma 18:19—26. Cook, S. A. 1962. Genetic system, variation, and adap- tation in Eschscholzia californica. Evolution 16:278— 299. GREENE, E. L. 1905. 205-292. KINDELL, C. E., A. A. WINN, AND T. E. MILLER. 1996. The effects of surrounding vegetation and transplant age on the detection of local adaptation in the perennial grass Aristida stricta. Journal of Ecology 84:745— 754. Revision of Eschscholzia. Pittonia 5: } KRUCKEBERG, A. R. 1986. An essay: the stimulus of und usual geologies for plant speciation. Systematic Bot-, any 11:455—463. . 1954. The ecology of serpentine soils III. Plant species in relation to serpentine soils. Ecology 35: 267-74. | . 1951. Intraspecific variability in the response of certain native plant species to serpentine soil. Amer- ican Journal of Botany 38:408-18. LINHART, Y. B. AND M. C. GRANT. 1996. Evolutionary significance of local genetic differentiation in plants. Annual Review of Ecology and Systematics 27:23 Tag 2a de MAYER, M. S., P. S. SOLTIS, AND D. E. SOLtis. 1994. Thel evolution of the Streptanthus glandulosus complex’ (Cruciferae): genetic divergence and gene flow in ser-) pentine endemics. American Journal of Botany 81(1): 1288-1299. MicrosorT ExceL. 1998 Version 7.0. Microsoft Corpora-| tion, Redmond, WA. Monrtacnes, R. J. S. AND D. H. Virt. 1991. Patterns of. morphological variation in Meesia triquetra (Bryop- sida: Meesiaceae) over an arctic-boreal gradient. Sys-' tematic Botany 16(4):726—735. | Munz, P. A. AND D. D. Keck. 1968. A California Flora! and Supplement. University of California Press, Berkeley, CA. NOAA. 2000. Center for Environmental Prediction: Recent El Nino and La Nifa events. http://www.pmel.noaa. gov/ toa-tao/la-nina-story.html#recent Accessed on 4 Dec} 2000. PAVLIK, B. M. AND M. W. SKINNER. 1994. Ecological char- acteristics of California’s rare plants. Pp. 4-6 in M.. W. Skinner and B. M. Pavilk (eds.), California Native; Plant Society Inventory of Rare and Endangered Vas-| cular Plants of California. 5th ed. California Native Plant Society, Sacramento, CA. Proctor, J. AND S. R. J. WODELL. 1975. The ecology o serpentine soils. Advances in Ecological Research. 9: 255-365. RossITER, M. C. 1996. Incidence and consequences of in- herited environmental effects. Annual Review of! Ecology and Sytematics 27:451—476. SAS. 1990 Ver. 6.0. SAS Institute Inc. Cary, NC. SLENTz, S. A., E. Boyp, AND L. A. McDape. 1999. Mor-) phological differentiation among Madrean sky island | populations of Castilleja austromontana (Scrophular- | iaceae). Madrono 46(2):100—111. SYSTAT. 1992 Ver. 5.2.1. SPSS, Chicago, IL. TURESSON, G. 1922. The genotypical response of the plant species to the habitat. Hereditas 3:211-—350. WASER, N. M. AND M. V. Price. 1985. Reciprocal trans- plant experiments with Delphinium nelsonii (Ranun- | culaceae): evidence for local adaptation. American’ Journal of Botany 72(11):1726—1732. | i } \ Maprono, Vol. 48, No. 1, pp. 33-37, 2001 GEOGRAPHIC VARIATION IN THE FREQUENCIES OF TRICHOME PHENOTYPES OF DATURA WRIGHTIT AND CORRELATION WITH ANNUAL WATER DEFICIT J. DANIEL HARE! AND ELIZABETH ELLE? Department of Entomology, University of California, Riverside, CA 92521 ABSTRACT The perennial plant species, Datura wrightii Regel (Solanaceae) is dimorphic for leaf trichome type. Some plants produce almost exclusively short, non-glandular trichomes, while other plants within the same population produce almost exclusively longer, glandular trichomes. In a previous survey of 19 southern California populations, the frequency of plants producing glandular trichomes ranged from O— 82%, and plants with glandular trichomes were absent from desert populations. Here we expand our studies to a total of 56 D. wrightii populations from central and southern California. We also examined the relationship between the frequency of glandular trichomes and two factors that broadly determine the availability of water at each site. The first was mean annual rainfall, while the second was mean annual evapotranspiration rate. The frequency of plants with glandular trichomes increased with increasing mean rainfall and decreased with increasing mean annual evapotranspiration rate. Combined, these two envi- ronmental variables accounted for about one-third of the variation in the proportion of plants with glan- dular trichomes. Results suggest that the production of a water-based exudate by plants with glandular trichomes may impose an additional demand for water on those plants compared to plants with non- glandular trichomes. Because of this additional water demand, the frequency of plants with glandular trichomes may decline relative to that of plants with non-glandular trichomes as available water becomes more limiting. INTRODUCTION Trichomes, or plant hairs, are found on aerial parts of plants in a multitude of forms. This diverse group of structures can be arbitrarily subdivided into glandular, secretory trichomes and non-glan- dular trichomes (Levin 1973). Among the suggest- ed ecological functions of trichomes are mainte- nance of the water balance in the leaves, deflection of intense solar radiation, and protection against herbivores (Levin 1973; Ehleringer 1984; Duffey 1986). Both glandular and non-glandular trichomes have been described in several genera within the Solanaceae (Luckwill 1943; Lemke and Mutschler 1984; Gregory et al. 1986; Ogundipe 1992). Individual plants of Datura wrightii Regel (So- lanaceae) produce mostly (>95%) glandular tri- chomes or mostly (>95%) non-glandular trichomes (van Dam et al. 1999). Plants that produce glan- dular trichomes feel sticky when touched, while plants with non-glandular trichomes feel velvety. The difference in trichome morphology is under the control of a single Mendelian gene, with the glan- dular condition dominant to the non-glandular con- dition (van Dam et al. 1999). Hereafter, we refer to plants with non-glandular trichomes as sticky plants and plants with glandular trichomes as velvety plants. Sticky plants produce an exudate composed ' Author to whom correspondence should be addressed. E-mail: harejd @citrus.ucr.edu. * Present Address: Dept. of Biological Sciences, Simon Fraser University, Burnaby, BC V5A 1S6 Canada. of acyl sugar esters in water, while velvety plants do not (van Dam and Hare 1998a). These plant types grow adjacently in populations in which they co-occur, often with their branches interdigitated, indicating that microsite specialization of the types is unlikely. The trichome dimorphism is important in determining plant susceptibility to herbivores. Velvety plants are more susceptible to whiteflies and the tobacco hornworm, Manduca sexta (Jo- hannson) (van Dam and Hare 1998a), but sticky plants are more susceptible to a mirid bug, Tupio- coris notatus (Distant) (van Dam and Hare 1998b; Elle and Hare 2000). Previously, van Dam et al. (1999) surveyed the distribution of velvety and sticky D. wrightii phe- notypes in 19 southern California plant populations and found that the frequency of the sticky pheno- type varied from 0% to 82%. They suggested that the production of a water-based exudate by glan- dular trichomes might be especially costly in arid environments, thus possibly accounting for the rel- ative scarcity of sticky plants in the deserts (van Dam et al. 1999). Here, we expand that initial sur- vey to include a total of 56 plant populations not only from southern California, but also from coastal and inland central California as well. We also ex- plore in more detail the potential interaction be- tween water availability and glandular trichome production by analyzing the frequency of sticky plants as a function of overall water availability, as indexed by mean annual rainfall and mean annual evapotranspiration demand. 34 MADRONO 100 Fic. 1. PP —_———— 3S Kilometers Sampling locations for sticky and velvety D. wrightii. Frequencies of sticky plants are shown in gray, and) frequencies of velvety plants are shown in black. The number within each graph refers to numbered populations) described in Table 1. MATERIALS AND METHODS Datura wrightii is a perennial plant species found in sandy or gravelly dry places in the southwestern U.S. and Mexico (Avery et al. 1959; Munz 1973). Despite the production of large trumpet-shaped flowers, self-pollination predominates, and _ out- crossing rates are no more than 30% (Snow and Dunford 1961; E. Elle, personal observation). The phenotype of plants was determined within 56 plant populations of varying size in southern and central California (Fig. |, Table 1). The phenotype of at least 25 plants in each population was clas- sified by touch and sight into either sticky or vel- vety categories. The presence or absence of glan- dular trichomes was confirmed with a hand lens (10x). Most populations were sufficiently small that every individual was examined. The southern California populations were censused in 1997, the San Joaquin Valley and Sierra foothill populations were censused 1998, and the central coast popula- tions were censused in 1999. Sampling was con- ducted in August or September of each year to en- sure that plants were expressing their adult tri- chome phenotype (van Dam et al. 1999). Mean annual rainfall at each sampled location was obtained from a map of California annual pre- ! | cipitation published by the National Resources Conservation Services PRISM Climate Mapping Project. This map shows lines of equal mean annual | rainfall (isohyets) based upon records taken from) 1961 through 1990 plotted at 5 cm. Mean annual rainfall was interpolated between adjacent isohyets. to the nearest 2.5 cm. Thus, mean annual rainfall is used as an index of water “‘supply”’ to all sam-! pled sites. An index of mean annual water “‘demand”’ that) is widely used in irrigation management is “‘refer-. ence evapotranspiration,” and is based on the amount of water that would evaporate from an open pan. Values for mean annual reference evapotrans- piration (ET,) were obtained from a map published) by the California Irrigation Management Informa- | tion System (CIMIS) that divides California into 18 | evapotranspiration zones. An index of overall water deficit was calculated for each site by subtracting the mean annual rainfall | from the mean annual ET, for each site. A value_ of zero would indicate that mean annual rainfall equaled mean annual ET,, while positive values _ would indicate that mean annual ET, exceeded mean annual rainfall. The relationship between the percent of plants within populations expressing the 2001] HARE AND ELLE: TRICHOME VARIATION AND WATER AVAILABILITY ies) N TABLE |. LOCATIONS OF D. WRIGHTII POPULATIONS SAMPLED FOR THE FREQUENCY OF TRICHOME TyPEsS. Populations are numbered as in Figure |. All localities are in California. ie 155 & 65. Roadside population near the intersection of State Highways 155 and 65, near Mile Marker # 20 on State Highway 155. 33 & 152. Roadside population at the intersection of State Highways 33 and 152 near Dos Palos. Arroyo Trabuco. Natural population within O’ Neill Regional County Park, Orange Co. . Avenue 22 & Road 191/2. Natural population in an abandoned field bordering this intersection in Madera Co. . Banning. Roadside population south of Banning on Old Banning Road. Barker Dam. Natural population, Joshua Tree National Park. Bell Canyon. Natural population, Caspers Regional Wilderness Park, Orange Co. . Bonsall. Roadside population along State Highway 76 south of Bonsall. . Buttonwillow. Roadside population near the intersection of Elk Grove Road and State Highway 158. . Carrizo Canyon. Natural population off of State Highway 74. . Casino Morongo. Roadside population near this Indian casino on State Highway 62. . Coyote Pass. Natural population within the Lake Perris State Recreation Area. Estrella Road. Roadside population north of the intersection of Estrella Road and State Highway 46. Fillmore. Roadside population near the eastern city limits on State Highway 126. . Firebaugh. Roadside population at the intersection of State Highway 33 and Douglas Avenue. Gilman Springs Road. Roadside population south of State Highway 60, approx. 0.8 mi. south on Gilman Springs Road. Gorman. Roadside population on Gorman Road, approximately 2 miles east of Interstate 5. Idyllwild. Natural population off State Highway 243 approximately two miles south of Idyllwild. Kaweah Lake. Natural population at the Lemon Hills Recreation Area on State Highway 198. Kern River Canyon. Roadside population on State Highway 178, near Call Box # 178-128. . La Palma. Roadside population at the intersection of La Palma Road and Huxford Avenue at the east end of Yorba Regional Park. . Lake Elsinore. Roadside population near the intersection of Riverside Avenue and Collier Road. . Lemoore. Roadside population near the intersection of Idaho Street and 19th Avenue. . McFarland. Weedlot population on Highway 43 directly west of McFarland in waste ground near an abandoned railroad siding. . McKittrick. Weedlot population near the intersection of State Highways 58 and 33. . Mill Creek. Natural Population at the Mill Creek Ranger Station, State Highway 38. . Millerton Lake. Natural population around the Millerton Historic Courthouse. . Milo Ranger Station. Roadside population on Yokohl Road near the Milo Ranger Station. . Moreno Valley. Weedlot population near the intersection of Moreno Beach Drive and Ironwood Avenue. . Motte. Natural population within the Motte Rimrock UC Reserve along Pictograph Trail. . Neenatch. Roadside population on State Highway 138 approximately midway between Pearblossom and Little Rock, 4 mi. east of the California Aqueduct. . Ortega Flats. Natural population in Caspers Regional Wilderness Park, Orange Co. . Pacheco Pass. Weedlot population off Highway 152 at Dinosaur Point boat launching ramp, San Luis Reservoir. . Palm Springs. Weedlot population in the vicinity of the parking lot for the Palm Springs Tram off State Highway a . Porterville Road. Roadside population north of Glenville, on Jack Ranch Road. . Rich Bar. Weedlot population along State Highway 178 near the Rich Bar Overflow parking lot. . Riley Wilderness. Natural population within Riley Wilderness Park, Orange Co. . Route 62. Roadside population 2 miles east of the boundary of Joshua Tree National Park. . San Onofre. Roadside population along Interstate 5. . Sequoia National Forest. Roadside population along State Highway 180 approximately | mile inside National Forest boundary. . Simi Valley. Weedlot population at Los Angeles Avenue and Angus Road. . Three Rivers. Natural population in a field on North Fork Road 4.8 miles from State Highway 198. . Tollhouse Grade. Roadside population on Tollhouse Road at the Sierra National Forest boundary. . Tollhouse Road. Roadside population between Humphrey’s Station and Tollhouse. . Trimmer Springs Road & Belmont Avenue. Roadside Population at intersection. . Twentynine Palms. Roadside population near the intersection of Utah Trail and Underhill Road near the north entrance to Joshua Tree National Park. . UC Irvine. Weedlot population in an undeveloped field near University Avenue and Beech Tree Road. . UC Riverside. Natural population within the UC Riverside Botanic Gardens grounds. . Val Verde. Population in an abandoned field on San Martinez Road off State Highway 126 near Val Verde County Park. . White Tank. Natural population within Joshua Tree National Park. . Whitesbridge Road. Roadside population on State Route 180 near Mendota. . Wild Horse Road. Weedlot population in an abandoned field south of King City, east of U.S. 101. . Wilson Canyon Wash. Natural population within Joshua Tree National Park. . Winchester Road. Weedlot population on undeveloped ground near the intersection of Winchester Creek Road and State Highway 79. . Woodlake. Natural population along a stream on Highway 245 directly east of the Woodlake Airport. . Yorba Linda. Weedlot population on a vacant lot near Weir Canyon Road and Savi Ranch Road. 36 MADRONO sticky phenotype and mean annual water deficit was determined by linear regression analysis. RESULTS AND DISCUSSION The frequency of sticky plants varied from 0% in six populations, mostly from the Mojave Desert, to 93% in a population near Lake Elsinore. Other plant populations having more than 75% sticky in- dividuals included two populations from the coastal mountains of Orange County (Ortega Flats and Riley Wilderness State Park), one from the moun- tains of the central Coast Range (Estrella Road), and another from the foothills of the central Sierra Nevada range (Sequoia National Forest). Mean annual precipitation from these sites ranged from a low of 10 cm annually at the Twen- tynine Palms population in the Mojave Desert to a high of nearly 69 cm annually at the Idyllwild pop- ulation in the San Jacinto Mountains. Sites where rainfall also averaged 15 cm or less include all of the Mojave Desert populations as well as popula- tions in the southwestern portion of the San Joaquin Valley (McKittrick, Buttonwillow). Other popula- tions where mean annual rainfall was relatively high (SO cm or more) were limited to other areas in the foothills of the Sierra Nevada mountain range (Porterville Road, Milo Ranger Station, Tollhouse Road, Tollhouse Grade, and Sequoia National For- est) or the base of the San Bernardino Mountains (Mill Creek). Highest mean annual ET) (183 cm, Zone 18) oc- curred in the two populations in the Coachella Val- ley (Carrizo Canyon, Palm Springs), followed by the three populations in Joshua Tree National Park (Barker Dam, White Tank, Wilson’s Creek, 168 cm, Zone 17), and the two populations from the floor of the San Joaquin Valley in Fresno Co. (Whites- bridge Road, Lemoore, 157 cm, Zone 16). Lowest mean annual ET, occurred in the San Onofre pop- ulation (Zone 1, 84 cm), followed by the popula- tions on the coastal plain of Orange and San Diego Counties (UC Irvine, Bonsall), and at the base of the Santa Ana Mountains in Orange County (Ar- royo Trabuco, Bell Canyon, Ortega Flats, Riley Wilderness, Zone 4, 118 cm). Overall, the percentage of sticky plants was pos- itively correlated with increasing mean annual rain- fall (r = 0.455, P = 0.004, n = 56) and negatively correlated with increasing mean annual ET) (r = —0.474, P = 0.0002, n = 56). Mean annual ET, and mean annual rainfall were significantly, but im- perfectly negatively correlated (r = —0.32, P = 0.016, n = 56), so that when the two variables were combined into the new variable, ‘““mean annual wa- ter deficit,’ the percentage of sticky plants declined with increasing water deficit, (rk = —0.572, P < 0.0001, n = 56), and water deficit accounted for 33% of the variation in the percentage of sticky plants (Fig. 2). These results suggest that sticky plants may be at a selective disadvantage in rela- [Vol. 48 100 5 | @ | @ gy 807 e .° (= a Oo. > < 2 a) “” a = ® © hes ® a He aI a ee 40 60 80 100 120 140 160 180 Mean Annual Water Deficit Fic. 2. Percent of sticky plants in each population as a function of water deficit. Water deficit was calculated as the mean annual rainfall at each site subtracted from the mean annual evapotranspiration (ET)) at each site. Re- gression equation: Percent sticky = 98.90 — (0.6370 * water deficit); P < 0.0001, n = 56). tively dry environments because glandular tri- chomes may impose an additional water demand on sticky plants, as has been found for at least one other plant species (Lauter and Munns 1986). The low proportion of variance in the frequency of sticky plants that is accounted for by water sup- ply (rainfall) and demand (ET ) suggests that other factors may also influence the frequency of sticky plants in particular locations, and we know from previous studies that trichome morphology strongly influences the susceptibility of plants to insect at- tack (van Dam and Hare 1998a, 1998b; Elle et al. 1999; Elle and Hare 2000). Within sites of equiv- alent water availability, natural selection may favor sticky plants over velvety plants when the herbi- vore community is dominated by species such as whiteflies, flea beetles, and M. sexta and disfavor sticky plants when the herbivore community is dominated by species like 7. notatus that are par- ticularly well adapted to feed on plants with glan- dular trichomes (Elle and Hare 2000). Thus, a por- tion of the remaining variation in trichome fre- quencies could be accounted for by variation in the structure of herbivore communities attacking plant populations in areas of similar water availability. Such variation in herbivore community structure has already been shown for different populations within southern California habitats (Elle and Hare 2000). Additionally, because these plant populations also are relatively small, and self-pollination pre- dominates (Snow and Dunford 1961), founder ef- fects and limited gene flow among plant popula- tions may also contribute to variation in phenotype frequencies among plant populations with similar water availability. In order to completely account for all of the variation in trichome frequencies among these plant populations, it would be neces- sary to consider the actual site-specific water avail- 2001 | ability, the pattern and magnitude of damage by insect species differentially adapted to trichome type, and the ability of plant populations to respond genetically to natural selection by these factors. Nevertheless, the data presented here suggest that variation in the availability of water, as indexed by variation in mean annual rainfall and mean annual evapotranspiration, may provide a broad gradient of resource availability upon which more specific in- teractions between D. wrightii trichome types and both biotic and abiotic components of the plant’s local environments are displayed. ACKNOWLEDGMENTS We thank W. Helms of the UCR Science library for assistance in producing the digital base map for Figure 1. The source for the map showing county lines was pub- lished by the U. S. Geological Survey in 1999 titled “County Boundaries of the United States.’’ Elevational relief was added from the source, “Digital Chart of the World,” 1993 version published by the Environmental Systems Research Institute, Inc. This study was funded in part by the National Science Foundation (NSF DEB 96- 15134 to JDH). LITERATURE CITED AVERY, A., S. SATINA AND J. RIETSEMA. 1959. Blakeslee: the genus Datura. Ronald Press Company, NY. DuFFey, S. S. 1986. Plant glandular trichomes: their par- tial role in defence against insects. /n B. E. Juniper and T. R. E. Southwood (eds.), Insects and the Plant Surface. E. Arnold, London. EHLERINGER, J. 1984. Ecology and ecophysiology of leaf pubescence in North American desert plants. Jn E. Rodriguez, P. L. Healey and I. Mehta (eds.), Biology and Chemistry of Plant Trichomes. Plenum Press, New York. ELLE, E., N. M. VAN Dam, AND J. D. HARE. 1999. Cost of glandular trichomes, a “‘resistance” character in Da- HARE AND ELLE: TRICHOME VARIATION AND WATER AVAILABILITY OF tura wrightii Regel (Solanaceae). Evolution 53:22— 35. AND J. D. HARE. 2000. No benefit of glandular trichome production in natural populations of Datura wrightii? Oecologia 123:57—65. GREGORY, P., D. A. AVE, P. Y. BOUTHYETTE, AND W. M. TINGEY. 1986. Insect-defensive chemistry of potato glandular trichomes. /n B. E. Juniper and T. R. E. Southwood (eds.), Insects and the Plant Surface. E. Arnold, London. LAUTER, D. J. AND D. N. Munns. 1986. Water loss via the glandular trichomes of chickpea (Cicer arietinum L.). Journal of Experimental Botany 37:640—649. LEMKE, C. A. AND M. A. MUTSCHLER. 1984. Inheritance of glandular trichomes in crosses between Lycoper- sicon esculentum and Lycopersicon pennellii. Journal of the American Society for Horticultural Science 109:592-596. Levin, D. A. 1973. The role of trichomes in plant defense. Quarterly Review of Biology 48:3-15. LuckwiLL, L. C. 1943. The genus Lycopersicon, an his- torical, biological, and taxonomic survey of the wild and cultivated tomatoes. University Press, Aberdeen. Munz, P. A. 1973. A Californian flora (with supplement), 1973 Edition. University of California Press, Berke- ley, CA. OGUNDIPE, O. T. 1992. Leaf epidermal studies in the genus Datura Linn. (Solanaceae). Phytomorphology 42: 209-217. SNow, R. AND M. P. DUNFORD. 1961. A study of inter- change heterozygosity in a population of Datura me- teloides. Genetics 46:1097—1110. VAN DAM, N. M. AND J. D. HARE. 1998 a. Biological ac- tivity of Datura wrightii glandular trichome exudate against Manduca sexta larvae. Journal of Chemical Ecology 24:1529-—1549. AND J. D. HARE. 1998b. Differences in distribution and performance of two sap-sucking herbivores on glandular and non-glandular Datura wrightii. Ecolog- ical Entomology 23:22-32. ; , AND E. ELLe. 1999. Inheritance and dis- tribution of trichome phenotypes in Datura wrightii. Journal of Heredity 90:220—227. MApRONO, Vol. 48, No. 1, pp. 38-39, 2001 A NEW SECTION IN THE GOLDFIELD GENUS LASTHENIA (COMPOSITAE: HELIANTHEAE SENSU LATO) RAYMUND CHAN Jepson Herbarium and Department of Integrative Biology, University of California, 1001 Valley Life Sciences Building # 2465, Berkeley, CA 94720-2465 ABSTRACT Lasthenia Cass. sect. Ornduffia R. Chan is a new section in the goldfield genus Lasthenia (Compositae: Heliantheae sensu lato). The goldfield genus Lasthenia Cass. (sensu Orn- duff 1966, 1971, 1993) comprises 20 species and subspecies in six sections. Five of Ornduff’s (1966) six sectional circumscriptions are monophyletic based on results from a recent molecular phyloge- netic study using nuclear and chloroplast DNA se- quences (Chan 2000, Chan et al. in press): L. sect. Baeria (Fisch. & Mey.) Ornduff, L. sect. Burrielia (DC.) Ornduff, L. sect. Hologymne (Bartling) A. Gray in Torr. and A. Gray, L. sect. Lasthenia, and L. sect. Platycarpha (Hall) Ornduff. Based on the same data, L. sect. Ptilomeris (Nutt.) Ornduff (sen- su Ornduff 1966), which comprises six species [L. burkei (Greene) Greene, L. conjugens Greene, L. coronaria (Nutt.) Ornduff, L. fremontii (Torr. ex A. Gray) Greene, L. maritima (A. Gray) M. Vasey, and L. minor (DC.) Ornduff], is strongly resolved as two well-supported monophyletic groups. Lasthen- ia burkei, L. conjugens, and L. fremontii form an unresolved monophyletic lineage; L. coronaria, L. maritima, and L. minor form another monophyletic group. The relationship between these two groups of species is unresolved. Disparity among these species in chromosome numbers, flavonoid chem- istry (Bohm et al. 1974; Ornduff et al. 1974), and morphological features further challenge the mono- phyly of L. sect. Ptilomeris. A review of all avail- able data support the separation of L. sect. Ptilom- eris into two monophyletic sections. The two groups of species can be distinguished by fruit sizes, chromosome numbers, habitat pref- erences, and, to some extent, by geographic distri- bution. Lasthenia burkei, L. conjugens, and L. fre- montii have cypselae that are less than 1.5 mm long, have chromosome numbers of 2n = 12, and are commonly associated with vernal pools. They have distributions mostly limited to interior Cali- fornia. Both L. burkei and L. conjugens are listed as endangered species in the federal list of endan- gered and threatened wildlife and plants (Tibor 2001). Lasthenia coronaria, L. maritima, and L. mi- nor have cypselae more than 1.5 mm long, have chromosome numbers of 2” = 8 or 10, and are not usually associated with vernal pools although L. minor has been found in vernal pools (D. Keil pers. comm.). They have wide distributions that include | coastal habitats in California (and, for L. coronaria and L. maritima, elsewhere along the Pacific coast — of North America). Based on phylogenetic results, I propose that the members of L. sect. Ptilomeris sensu Ornduff (1966, 1993) be relegated to two sections: L. sect. | Ptilomeris sensu stricto with L. coronaria, L. mar- | itima, and L. minor and L. sect. Ornduffia with L. | burkei, L. conjugens, and L. fremontii. Lasthenia Cass. sect. Ptilomeris (Nutt.) Ornduff, emend. R. Chan Plants not associated with vernal pools, leaves entire, irregularly lobed, or pinnatifid, involucres hemispheric to obconic, phyllaries free, receptacles conic, corollas of disc florets 5-lobed, floral pig-— ments remaining yellow in dilute alkali, tips of an- thers ovate to obovate, cypselae greater than 1.5 mm long, pappose or epappose, scales of pappi erose, lance-aristate, and/or subulate-aristate, 2n = 8, 10. A new section is erected for L. burkei, L. con- jugens, and L. fremontii, and is named for Professor Emeritus Robert Ornduff, in recognition of his out- standing contributions to the understanding of the evolution of Lasthenia and other groups in the Cal- ifornia flora. Lasthenia Cass. sect. Ornduffia R. Chan, sect. nov. Type species: Dichaeta fremontii Torr. ex A. Gray = Lasthenia fremontii (Torr. and A. Gray) Greene Plantae in consortio lacunarum vernalium et fo- liis plerumque pinnatifidis, involucris hemisphaeri- cis vel obconicis, phyllariis libris (ex parte connatis in una specie), receptaculis conicis vel tholiformi- bus, corollis flosculorum discorum 5-lobatis, pig- mentis floralibus remanentibus flavis in solutioni- bus dilutis alcalinis, apicibus antherarum linearibus vel anguste ovatis, cypselis <1.5 mm longis, pap- posis vel epapposis, squamellis papparum erosis vel subulatis-aristatis, 2n = 12. | 2001] Plants associated with vernal pools, leaves usu- ally pinnatifid, involucres hemispheric or obconic, phyllaries free (partly fused in L. conjugens), re- ceptacles conic or dome-shaped, corollas of disc florets 5-lobed, floral pigments remaining yellow in dilute alkali solution, tips of anthers linear to nar- rowly ovate, cypselae less than 1.5 mm long, pap- pose or epappose, scales of the pappi erose or su- bulate-aristate, 2n = 12. ACKNOWLEDGMENTS I thank Bruce G. Baldwin, David J. Keil, Robert W. Patterson, John L. Strother, and John W. Taylor for re- viewing the manuscript. This paper constitutes part of a doctoral dissertation submitted to the Department of In- tegrative Biology, University of California, Berkeley. LITERATURE CITED Boum, B. A., N. A. M. SALEH, AND R. ORNDUFF. 1974. The flavonoids of Lasthenia (Compositae). American Journal of Botany 61:551—561. CHAN, R., B. G. BALDWIN, AND R. ORNDUFF. Goldfields CHAN: LASTHENIA SECT. ORNDUFFIA 59 revisited: A molecular phylogenetic perspective on the evolution of Lasthenia (Compositae: Heliantheae sensu lato). International Journal of Plant Sciences (In press). CHAN, R. K.-G. 2000. Molecular systematics of the gold- field genus Lasthenia (Compositae: Heliantheae sensu lato). Ph.D. dissertation. University of California, Berkeley. OrRNDuFF, R. 1966. A biosystematic survey of the gold- field genus Lasthenia (Compositae: Helenieae). Uni- versity of California Publications in Botany 40:1—92. . 1971. A new tetraploid subspecies of Lasthenia (Compositae) from Oregon. Madrono 21(2):96—98. 1993. Lasthenia Pp. 298-299. In The Jepson manual: Higher plants of California, J. C. Hickman (ed.). University of California Press, Berkeley, CA. , B. A. BOHM, AND N. A. M. SALEH. 1974. Flavo- noid races in Lasthenia (Compositae). Brittonia 26: 411-420. Trpor, D. (ed.). 2001. Inventory of rare and endangered vascular plants of California. California Native Plant Society Special Publication No. 1, 6th ed. California Native Plant Society, Sacramento, CA. MADRONO, Vol. 48, No. 1, pp. 40—42, 2001 CORALLORHIZA MACULATA VAR. OZETTENSIS (ORCHIDACEAE), A NEW CORAL-ROOT FROM COASTAL WASHINGTON EDWARD L. TISCH Biology Department, Peninsula College, Port Angeles, WA 98362 ABSTRACT Corallorhiza maculata var. ozettensis is a newly described mycoheterotrophic orchid from western Washington. It occurs in foggy rainforests bordering the Pacific coast of the north Olympic Peninsula. Unlike typical C. maculata, its flowers are consistently non-spotted, with a narrow, white labellum bearing two apical undulations and low, non-rugose basal lamellae. Stem cross sections show epidermal cells mostly tangentially elongate, each bearing 4—10 delicate cuticular ridges bounded laterally by narrow sinuses. In June 1967, I collected a unique, white-lipped Corallorhiza near the Ozette Indian Reservation of coastal Washington. Subsequent collections and ob- servations revealed that populations of this coral- root, referred eventually to C. maculata (Raf.) Raf. (Buckingham and Tisch 1979), extended northward and inland at least 27 km and 1.5 km, respectively. The type collections are remarkably uniform in col- or, morphology, cell anatomy, and ecological fidel- ity, and in this locality occur to the apparent exclu- sion of contrasting varieties of C. maculata. While these plants exhibit homogeneity suggestive of re- productive isolation and are not distributed random- ly within populations of spotted C. maculata, as are many of its recognized color forms, their structural parameters lie within the limits established for C. maculata (Luer 1975, Freudenstein 1997), and I have relegated them to varietal status under that species. Corallorhiza maculata (Raf.) Raf. var. ozettensis E. Tisch, var. nov. (Fig. 1 in part)—-TYPE: USA, Washington, Clallam Co., forested bluffs above Cape Alava, 48°10’N 124°44'W, T31N R16W sect. 26, ca. 100’ (30 m) elev., 28 June 1967, E.L. Tisch 689A & 689B (holotype, UC; isotype, OSC). Caulis erectus, 20—60 cm altus, pallidus, roseus- violescens vel brunneo-violescens. Inflorescentia 3—20-flora, 5-17 cm longa, 2—3 cm lata. Pedicelli 1—2 mm longi, erecti vel penduli, bracteati. Bractae ovatae vel lanceolatae, 0.5—1.5(1.8) mm _ longae, acutae, obtusae, truncatae vel emarginatae. Flores 1.0-1.5 cm longi, 6-12 mm lati; sepala superior oblongo-oblanceolata, obtusa vel emarginata, (6.2)7.0—9.5(9.8) mm longa, (2.0)2.2—2.3(2.5) mm lata, apex purpureus, basis flavus; sepala laterala oblongo-oblanceolata, obtusa vel acuta, (5.8)6.5— 9.0(9.4) mm longa, (1.8)2.2—2.3(2.5) mm lata, apex purpureus, basis flavus; petala oblongo-lanceolata, obtusa vel acuta, (5.5)6.0—7.0(7.3) mm _ longa, (1.7)1.9—2.1(2.3) mm lata, flava; labellum oblan- ceolatum vel obovatum, trialobatum, trinerviatum, album immaculatum, obtusum, (5.0)5.5—7.5(8.0) | mm longum, (2.6)3.1—3.5(3.6) mm latum, apex bi- undulatum; mentum 1.0—2.0 mm longum, 0.4—1.0 mm altum; columna arcuata, 3.5—4.8 mm longa; stigma ca. 0.9-1.3 mm lata. Fructus elliptico-ob- longus, purpureus vel brunneus, pendulus, pauci- verrucosus, 1.0—1.5 cm longus, 3—4 mm crassus. Stems erect, 20—60 cm tall, pale pinkish violet or brownish violet. Racemes 3—20-flowered, 5—17 cm long, 2—3 cm wide. Pedicels 1—2 mm long, erect at anthesis to pendent in fruit. Floral bracts ovate to lanceolate, 0.5—1.5(1.8) mm long, acute, obtuse, truncate, emarginate, or bluntly tridentate. Flowers ascending at anthesis, 1.0—1.5 cm in length, 6—12 mm wide; dorsal sepal forward facing, oblong-ob- lanceolate, obtuse to obliquely emarginate, (6.2)7.0—-9.5(9.8) mm long, (2.0)2.2—2.3(2.5) mm wide, the apex purplish, often with translucent mar- gins, basal portions yellowish; lateral sepals usually spreading, oblong-oblanceolate, obtuse to nearly acute, (5.8)6.5—9.0(9.4) mm long, (1.8)2.2—2.3(2.5) mm wide, the apex purplish, basal portions yellow- ish; petals forward facing, oblong-oblanceolate, ob- tuse to acute, (5.5)6.0—7.0(7.3) mm long, (1.7)1.9— 2.1(2.3) mm wide, yellowish; labellum oblanceo- late or obovate, 3-lobed, nearly always 3-nerved, pure white at early anthesis (darkening with age), obtuse, (5.0)5.5—7.5(8.0) mm long, (2.6)3.1-— 3.5(3.6) mm wide at the widest part of the median lobe, the apex slightly dilated but rarely crenate- undulate or involute, usually bi-undulate at the tip (Fig. 1), the basal lamellae 1.7—2.2 mm long, aris- ing within 2 mm of the labial attachment and ex- tending to within 2.2—2.5 mm of its apex, non-ru- gose; mentum yellowish, 1.0—2.0 mm long, 0.4—1.0 mm high, yellow; column yellowish, often arcuate- ascending, 3.5—4.8 mm long; stigma ca. 0.9-1.3 mm wide. Capsule ellipsoidal, purplish to brown, slightly warty, 1.0—1.5 cm long, 3—4 mm thick. Paratypes. USA, Washington, Clallam Co.: coastal forests at Cape Flattery, 48°23'N 124°44'W, T33N RI16W sect. 1, ca. 75’ (23 m) elev., 23 July 1984, E.L. Tisch 2653 (WTU), 2654 (ORE), 2655 100 pm Fic. 1. ozettensis TISCH: C. MACULATA VAR. OZETTENSIS 4] occidentalis 1cm Illustrations comparing three varieties of C. maculata. A. Anterior views taken of living flowers with perianths folded back. B. Lateral views of immature capsules (perianths removed) showing frequent column orientations. C. Portions of stem cross sections taken | cm below the inflorescence. Illustrations by Karen Lull-Butler. (WS); shady coastal forest, ca. 90 m inland, Cape Flattery, ca. 100’ (30 m) elev., 27 July 1984, E.L. Tisch 2688 (UC), 2689 (OSC); forested bluffs, Portage Head, 48°17'N 124°41'W, T32N RISW sect. 7, ca. 80’ (25 m) elev., 18 June 1988, ELL. Tisch 3256 (MO), 3257 (V). Distribution, habitat, and phenology. Corallo- rhiza maculata var. ozettensis grows in moist, fog- gy, very shady to moderately illuminated forests bordering the northwestern coastline of the Olym- pic Peninsula. The collection sites, all within 300 m of the Pacific Ocean, are overstoried by mixtures of Picea sitchensis (Bong.) Carr., Thuja plicata D. Don, Tsuga heterophylla (Raf.) Sarg., and Alnus rubra Bong. A sparse understory of Malus fusca (Raf.) Schneid. and Rhamnus purshiana DC. is of- ten present, while the medium-shrub layer includes Vaccinium alaskense Howell, V. ovatum Pursh, V. parvifolium Smith, and Menziesia ferruginea Smith. The common herb associates are Blechnum spi- cant (L.) Smith, Polystichum munitum (Kaulf.) Presl, Maianthemum dilatatum (Alph. Wood) Nel- son and J. EK Macbr., Tiarella trifoliata L., Listera caurina Piper, and L. cordata (L.) R. Br. In its typ- ical habitats C. maculata var. ozettensis is incon- spicuous and rare. It is mycoheterotrophic and has knobby rhizomes embedded 1—2 dm in moist hu- mus. Depending on weather conditions, it blooms from about mid-June through late July. This is con- sidered ‘‘late’’ flowering for Corallorhiza as de- scribed by Freudenstein and Doyle (1994) and Freudenstein (1997). A? MADRONO Taxonomic relationships. Luer (1975) called C. maculata the most common and variable coral-root in the conterminous United States, and suggested that its color forms, while sometimes clustering in communities, tend to lack morphological identity separate from that of associated spotted individuals. He did not clearly differentiate between forms and varieties. Kartez (1994) synonymized all of the C. maculata variants under that single specific epithet. After years of research, Freudenstein (1986, 1992, 1997) narrowed the C. maculata complex, north of Mexico, to two intergradient varieties: maculata and occidentalis (Lindl.) Ames. Brown (1998), however, in his orchid checklist, recognized 8 in- fraspecific segregates, including forms, within that same complex. The var. maculata, a narrow-lipped, late-blooming variant, appears to be uncommon on the Olympic Peninsula, and is often intergradient here with the broad-lipped, early-blooming var. oc- cidentalis, which is larger and quite conspicuous, flowering as early as May 5 in the Olympic low- lands. Variety ozettensis has a narrow, white label- lum bearing two closely adjacent, upward undula- tions, one to either side of the mid-apex, and low, non-rugose basal lamellae. The labellum tapers to its attachment which is usually less than 1 mm wide. Its lateral and apical margins are semi-entire, contrasting with the crenate-undulate margins of the other two varieties. The narrow cuticular ridges on the stem epidermis number 4—10 per cell, nearly twice as many as the low, rounded ridges bordering comparable cells of vars. maculata and occidentalis (Fig. 1C). Also, the cauline epidermal cells of var. ozettensis, seen in cross sections taken | cm below the inflorescence, are >65% tangentially elongate, while those from the two spotted varieties of this region have <50% positioned in that plane. These diagnostic microscopic features were encountered consistently in living stems from 10 specimens of var. ozettensis, 15 of var. occidentalis, and 7 of var. maculata. The latter two varieties, at least on the Olympic Peninsula, have columns that often align with the floral axis, while those of var. ozettensis tend to ascend at angles >25° (Fig. 1B), but these tendencies are not entirely reliable. Superficially, var. ozettensis appears to be close- ly allied with forma immaculata (Peck) Howell, a white-lipped variant described from Linn Co., Or- egon (Peck 1954), and currently referred to var. oc- cidentalis in Brown’s (1998) checklist. Actually, this form of C. maculata is readily separable from var. ozettensis. The immaculata holotype (OSC!) has spreading perianth parts, and a crenulate, dis- tally expanded labellum with multiple levels of ve- nation. The latter half of the following key is mod- eled after Freudenstein’s (1997) key to the varieties of C. maculata. [Vol. 48 | KEY TO THREE VARIETIES OF CORALLORHIZA MACU- LATA AS REPRESENTED IN COASTAL WASHINGTON 1. Labellum white at early anthesis (darkening with age), its lateral nerves usually simple; the margins of its, central lobe sub-enure 2 ke eee Serer eee re C. maculata var. ozettensis 1. Labellum usually white, spotted with purple, its lat- eral nerves often prominently branched; the central lobe distally crenate-undulate .............. 2 2. Central lobe of labellum distinctly expanded, its broadest distal portion > 1.5 times wider than its base; labial apex broadly rounded to retuse ...... ee ett eee ee C. maculata var. occidentalis | 2. Central lobe of labellum slightly if at all expanded, its broadest distal portion < 1.5 times wider than its base; labial apex narrowly rounded to acute ee eee ee ee ee C. maculata var. maculata © The Ozette coral-root is named after the Ozette | band of Makah Indians that occupied the original — collection site for hundreds of years. ACKNOWLEDGMENTS I wish to thank Karen Lull-Butler for the drawings, Vince Murray for reviewing my Latin, and Kristina A. Schierenbeck for her expeditious editing of this manu- script. Kenton L. Chambers and Paul M. Brown gave many useful suggestions for which I am grateful. LITERATURE CITED Brown, P. M. 1998. Checklist of the orchids of North America north of Mexico. North American Native Orchid Journal 4(1):61—99. BUCKINGHAM, N. M. AND E. L. TiscH. 1979. Vascular plants of the Olympic Peninsula, Washington (a cat- alogue). Natl. Park Serv., Univ. Wash. Coop. Parks Studies Rep. B-79-2, Seattle, WA. FREUDENSTEIN, J. V. 1986. A preliminary study of Coral- lorhiza maculata (Orchidaceae) in eastern North America. Contributions from the University of Mich- igan Herbarium 16:145—153. . 1992. Systematics of Corallorhiza and the Cor- allorhizinae (Orchidaceae). Ph.D. dissertation, Cor- nell University, Ithaca, NY. . 1997. A monograph of Corallorhiza (Orchida- ceae). Harvard Papers in Botany 10:5—51. AND J. J. DOYLE. 1994. Plastid DNA, morpholog- ical variation, and the phylogenetic species concept: the Corallorhiza maculata (Orchidaceae) complex. Systematic Botany 19(2):273—290. KarTEZ, J. T. 1994. A synonymized checklist of the vas- cular plants of the United States, Canada, and Green- land, 2nd ed. Timber Press, Portland, OR. LUER, CARLYLE A. 1975. The native orchids of the United States & Canada excluding Florida. New York Bo- tanical Garden, Bronx, NY. Peck, M. E. 1954. Notes on certain Oregon Plants with descriptions of new varieties. Leaflets of Western Botany 7:177—200. MaApRONO, Vol. 48, No. 1, p. 43, 2001 NOTEWORTHY COLLECTIONS NEVADA ABIES CONCOLOR (Gordon & Glendinning) Hildebrand var. concolor (PINACEAE).—Nye Co., Belted Range, Nellis Air Force Bombing and Gunnery Range (NAFBGR), UTM Zone 11 579411E 4143156N (NAD 27), on steep NW slope ca. | km south of Indian Spring, 2287 m (7503’), 18 May 1996, F. Smith and D. Pritchett 3970 (RENO, UNLV, UTC); Nye Co., Belted Range, NAFBGR, UTM Zone 11 582031E 4150875N (NAD27), on steep N slope ca. 5.8 km north of Wheelbarrow Peak, 2417 m (7930’), 15 September 1996, F. Smith, D. Pritchett and E. Watkins 3982 (RENO, UNLV, UTC); Nye Co., Belted Range, NAFBGR, UTM Zone 11 581185E 4145466N (NAD27), northwest slope of Wheelbarrow Peak ca. 0.5 km from summit, 2349 m (7707'), 28 September 1996 F. Smith and D. Pritchett 3984 (RENO, UNLV). Specimens were ex- amined by Ronald M. Lanner, Utah State University, Lo- gan, Utah. Forms woodlands with Pinus monophylla gen- erally above 2286 m (7500’) on summits and ridges of the central portion (i.e., UTM Zone 11 ~ Northing 4151000 to ~ Northing 4138000) of the Belted Range, especially on N/NW slopes. Characteristic understory spp. include Chrysothamnus viscidiflorus, Ribes cereum, Ar- temisia tridentata and Poa fendleriana. Previous knowledge. Abies concolor var. concolor (Rocky Mountain white fir) occurs in Idaho, Utah, Colo- rado, Arizona, New Mexico, Nevada, and southeastern California while A. concolor var. lowiana (Gordon) Lem- mon (California white fir) is found in California, Oregon, and western Nevada. The nearest known occurrence of A. concolor var. concolor is on the summit of Bald Mountain in the Groom Range, about 32 km to the east (D.A. Char- let, 1996, Atlas of Nevada conifers, University of Nevada Press, Reno, NV). The nearest known occurrence of A. concolor var. lowiana 1s in the southern Sierra Nevada in California, about 250 km due west (R.M. Lanner, 1999, Conifers of California, Cachuma Press, Los Olivos, CA). Significance. This is the westernmost known occurrence of Abies concolor var. concolor in Nevada and may also be the westernmost occurrence in the Great Basin (D.A. Charlet, 1996, Atlas of Nevada conifers, University of Ne- vada Press, Reno, NV). A large portion (3.1 million acres) of central-southern Nevada is occupied by the Nellis Air Force Bombing and Gunnery Range and has been closed to the public since 1940. Janice Beatley, one of the few botanists who had access to the area, reported Abies concolor to be absent from Wheelbarrow Peak (Belted Range) and high peaks elsewhere in the region (J. Beatley, 1976, Vascular plants of the Nevada test site and central-southern Nevada: eco- logic and geographic distributions, U.S. Department of Commerce, National Technical Information Service). Bea- tley must have approached Wheelbarrow Peak from the southeast (probably from the road to Johnnies Water); we approached from the northwest where Abies concolor is abundant. The fact that extensive stands of a species of this size were only documented for the first time in 1997 (T. Knight, E Smith, and D. Pritchett, 1997, An inventory for rare, threatened, endangered and endemic plants and unique communities on Nellis Air Force Bombing and Gunnery Range, Clark, Lincoln and Nye Counties, Ne- vada, The Nature Conservancy of Nevada, Las Vegas) is an indication of how little botanical inventory work has been done in the Belted Range and suggests that further such work will prove fruitful. Work was supported by funds from the Department of Defense’s Legacy Resource Management Program awarded to The Nature Conservan- cy of Nevada. We are grateful to Cols. Thomas Lillie and Douglas Ripley for support and review of manuscripts. —DANIEL PRITCHETT, University of California White Mountain Research Station, 3000 E. Line St., Bishop, Cal- ifornia, 93514; FRANK J. SMITH, Western Ecological Ser- vices, Inc., PO. Box 422, Millville, UT 84326; TERI KNIGHT, Director of Science and Stewardship, Nevada Field Office, The Nature Conservancy, 1771 E. Flamingo #B111, Las Vegas, NV 89119. MADRONO, Vol. 48, No. 1, pp. 44-45, 2001 REVIEWS Savannas, barrens and rock outcrop plant com- munities of North America. Edited by Roger C. An- derson, James S. Fralish, and Jerry M. Baskin. 1999. Cambridge University Press, Cambridge UK. 470 p. Hardcover $110.00. ISBN 0-521-57322-X. This book synthesizes information on a number of North American plant community types found in environments that restrict tree and other plant growth due to harsh substrata and or other factors. Why combine savannas with edaphic complexes? I did not find a satisfactory answer to this. I think there is sufficient logic, and material, especially since some additional community types could have been included, for two somewhat smaller books. This would have better served readers who want a less expensive book on either topic, but not both. Despite this opinion and some other criticisms, I found much to like about this book. It is divided into 26 chapters, each describing an individual community, or in some cases a broad vegetation type (e.g., Southeastern Pine Savannas, Ponderosa Pine and Subarctic Woodlands). Forty-seven lead- ing experts on the subjects selected contribute to create an authoritative treatment. There are three indexes; plant, animal, and topic. The chapters on strongly edaphic or extreme-soil-condition com- munities comprise the portion of the book that is the most valuable in terms of types that have not previously received enough attention. While infor- mation on, for example, Serpentine barrens of Western North America, is readily available (e.g., Kruckeberg’s book), it is not for communities such as mid-Appalachian Shale Barrens, Granite Out- crops of the Southeastern United States or Southern Ontario, and especially Niagra Escarpment, Great Lake Alvar (limestone/marble substrata), and Sand Shinnery Oak communities. There is no mention of these in my 1988 edition of North American Ter- restrial Vegetation, and the authors of the chapter on Sand Shinnery Oak (S. S. Dhillion and M.H. Mills) mention that theirs is the first ever review of the ecology and future conservation of this surpris- ingly (at least it was to me) extensive community. To get a sense of how thoroughly community types are treated, note that North American Terres- trial Vegetation has seven pages on which Serpen- tine is mentioned compared to 2 chapters and 29 pages on the topic in this more specialized book. Thus, the book contains much enlightening infor- mation, even for ecologists with an encyclopedic level of knowledge of North American vegetation. Savannas and relatively widespread vegetation types discussed in this book are treated in reviews elsewhere, but not typically with as great an em- phasis on conservation and management issues, nor the effort to integrate plant and animal ecology found throughout this text. Unfortunately, savanna is a nebulous term. As a result there are chapters | on vegetation often thought of as woodland (i.e., | Pinyon-Juniper) that some might not expect to find. In fact, three chapters use woodland rather than sa- | vanna in their title (e.g., Subarctic). Conversely, some savanna or woodland types are left out (e.g., Garry Oak, sub-tropical). There are some edaphic complexes omitted as well, such as, in Western | North America, Sierran granite outcrops, and scab- land associated with lava flows in the Cascades and on the Modoc Plateau, etc. However, these subjects have been studied relatively little by ecologists and others, there may be insufficient literature from which to prepare a review. There is not a consistent set of topics covered in | each chapter, which is partly understandable con- sidering how different some of the chapter subjects are. For example, while fire is a keystone process maintaining the open tree spacing in many savanna communities, fire effects in rock outcrop and some barrens communities are nil. Nonetheless, there could be greater consistency among chapters. While most authors provided useful, concise sum- maries, they are missing from 6 chapters. In addi- tion, the use and effectiveness of maps and photos is variable. Some chapters lacked sections on con- servation and management despite these being im- portant overall themes. I was particularly interested in the chapter on California Oak Savanna, as I know more about this vegetation than the other types in the book. The floristic information is less detailed compared to many chapters. Other authors provided species lists, ordinations and/or other summary information, which I liked. This chapter will not serve as a re- placement for Jim Griffin’s excellent treatment in Terrestrial Vegetation of California, at least among more botanically-oriented readers. The Savanna chapter has a range management perspective, and there is considerable detail on this important, prac- tical topic. However, the related, complex subject of oak regeneration is not treated in enough detail. Answers to the cause of apparent failure of blue and valley oaks to regenerate saplings, and even the significance of the apparent failure are multifaceted, and still unclear. But I think more evidence from additional studies that have been undertaken should have been mentioned in the Chapter. For example, the work of K. Danielsen suggesting improved re- generation of Valley Oak with germination in na- tive versus annual grassland, and the extensive re- generation studies performed for the State of Cali- fornia by T. Swieki. It would also have been ap- propriate to include Garry Oak woodland, even | 2001) though it extends well north of California. Finally, I found 3 typo’s/mistakes, which makes me wonder how many went unnoticed in other chapters. I think the most significant contributions of this ‘book are the descriptions of unique, edaphic com- “munity types that many vegetation ecologists are unfamiliar with, the integration of plant and animal ecology, and the conservation and management considerations. Compared to typical community de- scriptions, there is greater discussion of subjects such as endemism, animal interactions, and the widespread management problem of how to deter- mine and achieve an appropriate fire regime. It is interesting and worthwhile to look at these phe- nomena from the many different aspects and per- spectives found in this book. —DENNIS C. ODION. Marine Science Institute, Univer- sity of California Santa Barbara, Santa Barbara, CA 93106. MADRONO, Vol. 48, No. 1, pp. 45-47, 2001 Terrestrial ecoregions of North America: A conser- vation assessment. By Taylor Rickets, Eric Diner- stein, David Olsen, Colby Loucks et al. 1999. Is- land Press. Covelo, California. 485 pp. When I first saw this book my initial feeling was that this was yet another way to classify ecosystems similar to the U.S. Forest Service ECOMAP (Bai- ley et al. 1994) or Omernik (1995) approach. How- ever, as I probed the pages I found the book was indeed not only an attempt to divide up the conti- nent into ecologically based units, but most impor- tantly, a true analysis of biological and conserva- tion traits using the 116 ecoregions as study units. By defining the ecoregions and having specialists in various taxonomic groups address basic conser- vation attributes of each ecoregion, the editors have come up with the first uniform treatment for the continent of these principal building blocks of eco- logical assessment. Within the last two years this book and The Na- ture Conservancy’s Precious Heritage (Stein et al. 2000) have arrived and addressed similar issues of national/continental scale conservation. Both are valuable additions to the literature. However, Ter- restrial Ecoregions of North America is by far the more scholarly and amenable for use as an actual conservation biology tool for the North American Continent. Precious Heritage, with its glossy format and beautifully illustrated examples of the biota, is pitched to the neophyte who needs to be educated on the urgency of conservation needs and on an overall awareness of the biotic distinctiveness of the United States. Terrestrial Ecosystems is a more utilitarian (and less costly) book. The photos of an- imals and plants are minimal and not particularly REVIEW 45 well reproduced. However the numerous multi-col- or maps of the continent are clean and effective. One of the greatest attributes of the Terrestrial Ecosystems book is its adherence to the standard- ized and uniform division of ecoregions. These ecoregions were defined in a systematic way that is well described in the first chapters. The rational for defining them anew, in lieu of adopting an existing ecoregional classification, is based on the lack of existing uniform treatments for ecoregions covering the full geographic extent of North America. How- ever, adherence to previously defined units where possible is strong, particularly Omernick’s (1995). Using the ecoregions, instead of a combination of ecological and political (states, counties) geograph- ic criteria for judging conservation importance, as is the case in Precious Heritage, the book provides a even-handed and readily comperable method of assessing the entire area. The standardized assessment approach involves a hierarchical division of the continent into realms (3), biogeographical zones (5), 10 major habitat types, and 116 ecoregions. Each of these is given a set of discriminators broken into biological dis- tinctiveness criteria and conservation status criteria. The former include species richness, endemism, rare ecological or evolutionary phenomena, and rare habitat types. The latter include habitat loss, remaining habitat blocks, degree of fragmentation, degree of protection, and future threat. Derived from these criteria are: 1) a biological distinctive- ness index, and 2) a conservation status index, which are integrated into five main categories rang- ing from “globally outstanding ecoregions requir- ing immediate protection” to “‘bioregionally and nationally important ecoregions requiring protec- tion of representative habitat blocks and proper management elsewhere for biodiversity conserva- tion.”’ Richness and endemism is analyzed for rep- resentative taxa for which sufficient information ex- ists. These include amphibians, birds, butterflies, mammals, vascular plants, reptiles, and land snails. Subdivisions of these major taxonomic groups are also treated including conifers, trees, and tiger bee- tles. Special features such as subterranean karst bio- diversity are also discussed. The biological assessment elucidates some very interesting facts. Though it is brought out that if politically defined, California leads all other con- terminous states in species richness and endemism, the division of the general California area into 12 distinct but relatively small ecoregions has moved the individual ecoregions within or partially within the Golden State into the second tier with respect to several criteria. These include total endemism of all taxa (leaders are the southeastern conifer forests of Florida and adjacent states, and the Colorado Plateau). Bird richness and endemism are lead by southwestern US ecoregions including the Chihua- 46 MADRONO huan and Sonoran Deserts. The southeast U.S. is leader in amphibian and snail richness and diver- sity. Mammal richness is also centered on the Col- orado Plateau and the Chihuahuan Desert, while surprisingly mammal endemism is greatest in the Sierra Nevada and the California interior chaparral and woodlands. Many patterns reflect the general trends of tropical diversity. For example, the south- western border ecoregions lead in butterfly and rep- tile diversity and endemism. Perhaps most interesting to you may be the vas- cular plant patterns, which show the southeastern mixed forests lead the nation in richness with over 3100 species, while the most diverse ecoregion in California is the Mojave Desert with about 2300— 2400 species. Vascular plant endemism is lead by the Colorado Plateau and the southeastern conifer forests with over 200 species, while the California ecoregions having highest endemism include the Klamath Province and the California interior chap- arral and woodlands, both with between 111 and 150 species. Note this analysis is based on full spe- cies (no subspecific taxa) as determined by John Kartesz. Turning to the conservation status of ecoregions, portions of California do rank among the most threatened including the Great Valley—deemed to have no remaining large blocks of habitat. The Great Valley, the Northern California coastal for- ests, and the California Coastal sage and chaparral all rank among the most critically imperiled of ecoregions in the conservation snapshot analysis. California also has the unfortunate distinction of containing the greatest number of introduced vas- cular plant taxa in any ecoregion in its interior chaparral and woodlands ecoregion (879 species). In the synthesis of biological distinctiveness and conservation status California contains six ecore- gions (more than any other state) that are consid- ered globally outstanding requiring immediate pro- tection of remaining habitat and extensive restora- tion. These are: Klamath-Siskiyou forests, Northern California coastal forests, Sierra Nevada forests, California interior chaparral and woodlands, Cali- fornia montane chaparral and woodlands (includes the Transverse and Peninsular ranges), and the Cal- ifornia coastal sage and chaparral. The main body of the book concludes with a list of the dozen highest-ranking ecoregions in need of immediate attention. The authors compare the rel- ative amount of attention that the Florida Ever- glades has received with these additional ecore- gions, and suggest that all of these are as worthy of attention as the Everglades. Three of these are in California; Coastal sage and chaparral, Klamath- Siskiyou, and Sierra Nevada forests. Recommendations for the protection of these ecoregions are listed in a 10-point plan. These points, though inherently reasonable and rational, [Vol. 48. will have different likelihood of success unless po- | litical climates change. For example, completing | networks of last remaining habitat in a system of. reserves iS a easier goal to achieve than allowing fire to play its critical role in maintaining biodiver- | sity or restricting livestock grazing in a number of | ecoregions. | Although the core of the book is a conservation assessment (104 pages), the largest portion of the | book is devoted to six appendices, the largest of | which is a detailed account of each of the 116) ecoregions written by local experts. Other appen- | dices detail the methods of the calculations used to | arrive at the biological distinctiveness and conser- | vation ranks, and also include specific lists of rich- | ness, endemism for each of the taxonomic groups _ treated by ecoregions. Appendix F is more than > twice as lengthy as the main portion of the book. | It contains individual summaries of each ecoregion | and is written and/or edited by a group of local | experts. (Robin Cox, David Olsen, Bob Holland, and John Sawyer have co-authored a number of the , California ecoregions). My criticisms of the book are relatively minor. — As the book is titled a conservation assessment of North America, one would expect to see a detailed treatment that includes not only the US and Canada, but Mexico down to the isthmus of Tehuantapec. In fact although a good map of Mexican ecoregions is displayed, there is insufficient information to af- ford a detailed analysis of the Mexican ecoregions. Although the treatments of each ecoregion in Ap- pendix F are valuable, they are somewhat uneven in scope and content. I was generally satisfied with those I am most familiar with. However, some in- consistencies remain. For example, some of the best treatments include detailed point-by-point descrip- tions of what conservation actions need to be taken (including specific locations that need protection), while other areas are not specifically addressed in this way. As with many good hierarchical treatments, the next step becomes clear. For each of these ecore- gions a similar local level assessment needs to be done. The Nature Conservancy is doing this work on an ecoregional level throughout much of Cali- fornia and the rest of the United States. Although the hierarchy in this book stops at the 116 ecore- gions, the Bailey ecoregions (1994) do have a more complete nesting of hierarchies down to very local level a geographic scale down to the sub-watershed or so-called ecological land unit. The ecological subsection map and descriptions produced for Cal- ifornia (Miles and Goudey 1997) has great concep- tual valuable in this regard, yet similar detailed treatments have not been done for all of the coun- try. I recommend this book for a lucid, scientific ap- proach to conservation at the continental scale. I 2001] ‘intend to take it with me whenever I travel through- out the continent, as it not only affords a clear strat- egy, but also is a valuable biogeographical sum- mary of information that stands alone in its own right. —Topp KEELER-WOLF. Senior Vegetation Ecologist, Department of Fish and Game, 1416 9" St. 12™ Floor, Sacramento, CA 95814. REFERENCES BAILEY, R. G. P. E. Avers, T. KING, AND W. H. MCNAB. (eds.). 1994. Ecoregions and sub-ecoregions of the REVIEW 47 United States. Washington D.C., USDA Forest Ser- vice. Mies S. R. AND C. B. GoubDEy. 1997. Ecological subre- gions of California: section and subsection descrip- tions. USDA Forest Service Pacific Southwest Re- gion R5-EM-TP-00S. OMERNIK, J. M. 1995. Ecoregions: a framework for man- aging ecosystems. George Wright Forum 12(1):35— 51. STEIN, B. A., L. S. KUTNER, AND J. S. ADAMS. 2000. Pre- cious Heritage: The Status of Biodiversity in the United States. Oxford University Press. 416 pp. BOTANY GRADS MEET AT CHICO STATE The 19th California Botanical Society Graduate Student Meetings, for students studying any aspect of plant sciences, was hosted by the Department of Biological Sciences at Chico State University on 17 February 2001. Local Botany graduate student Leah Mahan, with some support by Rob Schlising, planned and organized talk sessions and oversaw this bienniel event where graduate students share their research and ideas in a day of formal talks and discussions with their colleagues. There were 31 abstracts submitted by students from Claremont Graduate University/Rancho Santa Ana Botanical Garden; California State University, Fullerton; Uni- versity of California, Santa Barbara; California Polytechnic State University, San Luis Obispo; San Francisco State University; University of California at Berkeley; Humboldt State University; University of Kentucky; and Chico State University. One morning session was followed by two concurrent sessions in the afternoon, with talks categorized as research that was 1) completed, 2) in progress, or 3) proposed. Graduate students from Chico State | and elsewhere chaired all sessions, or served as} judges to evaluate talks in each of the three cate- | gories. | Awards of books and memberships in the Cali-_ fornia Botanical Society were presented at the an-— nual meeting and banquet of the Society, held on campus that evening. First place winners for talks | were Christopher Adams, University of Kentucky | (completed—“‘Seed dormancy in Aristolochia cal- | ifornica’’); Justin Whittal, UC Santa Barbara (in progress—‘“‘Key innovation and rapid radiation. among the columbines: insights from another chlo- | roplast region and introns of three nuclear genes’’ ); and Gavin Blosser, Chico State (proposed—‘“Di- | versity and characterization of arbuscular mycor- | rhizal fungi in the soils of vernal pools in northern | California’). Serpentine plant expert, Dr. Arthur | Kruckeberg, of the University of Washington, gave. the address after the banquet on “The Influence of | Geology in Shaping the California Flora.” TALKS PRESENTED AT THE 19TH CALIFORNIA BOTANICAL SOCIETY GRADUATE STUDENT MEETINGS Adams, Christopher A. School of Biological Sciences, University of Kentucky, Lexington, KY 40506. SEED DORMANCY IN ARISTOLOCHIA CALIFORNICA (ARISTOLOCHIACEAB). (Completed) Bell, Hester L. Department of Botany, Claremont Grad- uate University and Rancho Santa Ana Botanical Gar- den, 1500 North College Ave., Claremont, CA 91711. RESPONSE OF SPOROBOLUS VIRGINICUS (GRA- MINEAE) TO SALINITY. (Completed) Blosser, Gavin D. Department of Biological Sciences, California State University, Chico, CA 95929. DIVER- SITY AND CHARACTERIZATION OF ARBUSCU- LAR MYCORRHIZAL FUNGI IN THE SOILS OF VERNAL POOLS IN NORTHERN CALIFORNIA. (Proposed) Bradford, Darhl L. Department of Biological Sciences, California State University, Chico, CA 95929. THE HYBRIDIZATION OF CALIFORNIA SYCAMORE (PLATANUS RACEMOSA) AND THE LONDON PLANE TREE (PLATANUS X ACERIFOLIA) IN CAL- IFORNIA’S RIPARIAN WOODLAND. (In Progress) Bushakra, Jill M. Department of Biological Sciences, California Polytechnic State University, San Luis Obis- po, CA 93407. GENETIC DIVERSITY AND PHY- LOGENY OF CIRSIUM OCCIDENTALE. (In Progress) Cerros-Tlatilpa, Rosa. Department of Botany, Claremont Graduate University and Rancho Santa Ana Botanic Garden, 1500 North College Ave., Claremont, CA 91711. SYSTEMATICS OF ARISTIDA (GRAMINE- AE). (In Progress) Douhovnikoff, Vladimir. Department of Environmental Science, Policy & Management, University of Califor- nia at Berkeley, 4927 Happy Valley Rd., Lafayette, CA 94549. THE IMPORTANCE OF CLONING IN SALIX EXIGUA (In Progress) Ellberg, Sherry R. Department of Biological Sciences, California State University, Chico, CA 95929, QUAN- TITATIVE TRAIT LOCI ANALYSIS IN CLARKIA LINGULATA AND CLARKIA BILOBA SSP. AUSTRAL- IS. (in Progress) Fry, Danny, L. Department of Biological Sciences, Cal- ifornia Polytechnic State University, San Luis Obispo, CA 93407. EFFECTS OF A PRESCRIBED FIRE ON OAK WOODLAND STAND STRUCTURE. (in Prog- ress) Ganong, Constance K. Department of Biology, San Fran- cisco State University, 1600 Holloway Ave., San Fran- cisco, CA 94132. PHYLOGENETIC ANALYSES OF PHACELIA SECTION MILTITZIA USING MORPHO- LOGICAL AND MOLECULAR DATA. (In Progress) Gehrung, Loren. Department of Biological Sciences, California State University, Chico, CA 95929. INVES- TIGATIONS IN THE GENUS VACCINIUM, SEC- TION MYRTILLUS, IN NORTHERN CALIFORNIA AND SOUTHERN OREGON, INCLUDING RESO- LUTION OF THE TAXONOMIC STATUS OF VAC- CINIUM COCCINEUM PIPER BY RAPD ANALY- SIS. (Completed) Griffith, Patrick M. Department of Botany, Claremont Graduate University and Rancho Santa Ana Botanical Garden, 1500 North College Ave., Claremont, CA 91711. NATURAL INTERSPECIFIC HYBRIDIZA- TION IN OPUNTIA OF THE NORTHERN CHIHUA- HUAN DESERT REGION. (Completed) | 2001) Hendrick, Mike B. Department of Biological Sciences, California State University, Chico, CA 95929. ENVI- RONMENTAL FACTORS AND THEIR EFFECTS ON FLORAL COMMUNITY STRUCTURE IN THREE MONTANE MEADOWS IN BUTTE COUN- TY, CALIFORNIA. (Completed) _Honer, Michael. Department of Botany, Claremont Grad- uate University and Rancho Santa Ana Botanical Gar- den, 1500 North College Ave., Claremont, CA 91711. A FLORA OF THE GLASS MOUNTAIN REGION, MONO COUNTY, CA. (In Progress) Kashani, Nasser. Department of Environmental Science, Policy & Management, University of California at Berkeley, 4927 Happy Valley Rd., Lafayette, CA 94549. EXTENT OF GENETIC DIFFERENTIATION BETWEEN QUERCUS PARVULA VAR. SHREVEII AND @Q. WISLIZENIT. (Completed) Karr, Stephen J. Department of Biological Sciences, Cal- ifornia State University, Chico, CA 95929. INFLU- ENCE OF CARBON TO NITROGEN RATIO ON THE GENETIC CONTROL OF XYLOGENESIS IN ARABIDOPSIS THALIANA. (Proposed) Kirk, Paul. Department of Biological Sciences, California State University, Chico, CA 95929. AFLP ASSESS- MENT OF PUTATIVE JUGLANS HINDSII IN RIPAR- IAN FORESTS OF NORTHERN CALIFORNIA. (Pro- posed) LaDoux, Tasha. Department of Botany, Claremont Grad- uate University and Rancho Santa Ana Botanical Gar- den, 1500 North College Ave., Claremont, CA 91711. COALESCENCE OF S-ALLELES IN PAYSALIS CI- NERASCENS (DUNAL) A.S. HITCHC. (Proposed) Mahan, Leah M. Department of Biological Sciences, California State University, Chico, CA 95929. FAC- TORS AFFECTING THE SURVIVAL OF HEMIZON- IA FITCHIT (ASTERACEAE) IN THE NORTHERN SACRAMENTO VALLEY OF CALIFORNIA. (In Progress) McDill, Joshua R. Biology Department, San Francisco State University, 1600 Holloway Ave, San A COM- PARATIVE ANATOMICAL STUDY OF LINANTHUS AND RELATED GENERA (POLEMONIACEAE). (In Progress) McGlaughlin, Mitchell. Department of Botany, Clare- mont Graduate University and Rancho Santa Ana Bo- tanical Garden, 1500 North College Ave., Claremont, CA 91711. GENETIC VARIABILITY OF REINTRO- DUCED POPULATIONS OF PINK SAND VERBE- NA. (Completed) McGraw, Jodi M. Department of Integrative Biology, University of California, Berkeley, CA 94720. FIRE 49 SUPPRESSION, TREE ENCROACHMENT, AND THE SPREAD OF EXOTIC SPECIES INFLUENCE THE PERSISTENCE OF TWO ENDANGERED PLANTS IN CALIFORNIA. (In Progress) Menke, Marck. Department of Biology (Ecology and Systematics), San Francisco State University, 1600 Hol- loway Ave., San Francisco, CA 94132. A MOLECU- LAR BASED PHYLOGENETIC ANALYSIS OF THE WOODY HYDROPHYLL CLADE. (Proposed) Parks, David T. Department of Biological Sciences, Cal- ifornia State University, Chico, CA 95929. EVOLU- TIONARY RELATIONSHIPS OF ARCTOSTAPHY- LOS MEWUKKA AND ASSOCIATED SPECIES. (In Progress) Phipps, Frances A. Department of Biological Sciences, California State University, Chico, CA 95929. POPU- LATION GENETIC ANALYSIS OF HOWELLIA AQUATILIS (CAMPANULACEAEB). (In Progress) Rentz, Erin D. Biology Department, San Francisco State University, 1600 Holloway Ave, San Francisco, CA 94132. EFFECTS OF BURNING ON THE ANATOM- ICAL STRUCTURE OF CORYLUS CORNUTA AND XEROPHYLLUM TENAX, PLANTS COMMONLY USED IN CALIFORNIA INDIAN BASKETRY. (Pro- posed) Stewman, Casey J. Department of Biological Sciences, Humboldt State University, Arcata, CA 95521. EN- CROACHMENT PATTERNS OF DOUGLAS-FIR INTO OAK WOODLANDS IN THE CENTRAL WESTERN KLAMATH REGION. (Completed) Timme, Ruth E. Department of Biology (Ecology and Systematics), San Francisco State University, 1600 Hol- loway Ave., San Francisco, CA 94132. A MOLECU- LAR PHYLOGENY OF THE GENUS POLEMONIUM (POLEMONIACEAE). (In Progress) Wendel, Heather E. Department of Biological Science, California State University, Fullerton 800 N. State Col- lege Ave., Fullerton, CA 92831. COMPETITION BE- TWEEN INVASIVE EXOTIC ANNUALS AND NA- TIVE ENDANGERED ERIASTRUM DENSIFOLIUM SSP. SANCTORUM (POLEMONIACEAE) SHRUBS. (Proposed) Willyard, Ann M. Department of Biological Sciences, California State University, Chico, CA 95929. MAP- PING A DISEASE-RESISTANCE GENE IN SUGAR PINE AND WESTERN WHITE PINE. (In Progress) Whittall, Justen, B. Ecology, Evolution, and Marine Bi- ology, University of California, Santa Barbara, CA. KEY INNOVATION AND RAPID RADIATION AMONG THE COLUMBINES: INSIGHTS FROM ANOTHER CHLOROPLAST REGION AND _IN- TRONS OF THREE NUCLEAR GENES. (In Progress) NEW PUBLICATIONS Northwest Plant Hunters Series: Life of Botanist Louis F. Henderson The Native Plant Society of Oregon proudly an- nounces the publication of NPSO Occasional Paper Number 2, “‘Louis E Henderson (1853-1942): the Grand Old Man of Northwest Botany,”’ by Dr. Rho- da M. Love of Eugene. The peer-reviewed paper has been formatted as a 64-page booklet with 56 historic and modern im- ages—many never before published. It is carefully researched, with 133 notes. Also included are a chronology of Henderson’s life, notes on many of his important collections, a list of his publications, and a list of plants named for Henderson. The re- search took nearly three years and extended throughout the Pacific Northwest as well as to Mis- sissippi, Cornell University, The Chicago Field Mu- seum, the Smithsonian Institution, and the Jepson Herbarium at Berkeley. The Occasional Paper is a much-expanded version of Dr. Love’s earlier essay on Henderson which appeared in Pacific Northwest Quarterly last year. Henderson lived through the Civil War in Mis- sissippi only to see his lawyer father murdered in New Orleans during the Reconstruction period. Young Louis was educated at Cornell, studying bot- any under David Starr Jordan, later President of Stanford. He came west in 1874 and moved to Port- land in 1877 to take up a teaching post. He began his botanizing in Washington and Oregon at that) time. Soon after, Henderson married fellow teacher Kate Robinson and the couple had two daughters. Henderson had several careers in botany in the} Northwest including that of Professor of Botany at the University of Idaho from 1893 to 1908. It was. during this time that his herbarium burned, destroy- ing an estimated 85,000 specimens. At the age of seventy-one he became Curator of the Herbarium | of the University of Oregon and remained for 15. years, greatly increasing the collection. Approxi-. mately 30 species were named for Henderson; 16 bear his name today. The Native Plant Society of Oregon, a non- profit organization, has advanced funds for the printing | and mailing of 200 copies of the Henderson Oc-| casional Paper. The Native Plant Society wishes to, recoup its investment in a timely fashion, thus mail orders will be accepted starting immediately. The cost is $10.00 per copy which includes mailing and handling. Orders will be filled as soon as received. | Now Available Free of Charge Proceedings volume—2” Interface Benweam Ecology and Land Development In California. Ed-. ited by J.E. Keeley, M.B. Keeley and C.J. Fother-. ingham. 2000. U.S. Geological Survey. Request a copy from either: Jon_Keeley @usgs. gov. or Sequoia National Park, 47050 Generals Highway, | Three Rivers, CA 93271. | Volume 48, Number 1, pages 1—50, published 31 August 2001 SUBSCRIPTIONS—MEMBERSHIP Membership in the California Botanical Society is open to individuals ($27 per year; family $30 per year; emeritus $17 per year; students $17 per year for a maximum of 7 years). Late fees may be assessed. Members of the Society receive Maprono free. Institutional subscriptions to MADRONO are available ($60). Membership is based on a calendar year only. Life memberships are $540. Applications for membership (including dues), orders for sub- scriptions, and renewal payments should be sent to the Treasurer. Requests and rates for back issues, changes of address, and undelivered copies of Maprono should be sent to the Corresponding Secretary. INFORMATION FOR CONTRIBUTORS Manuscripts submitted for publication in MADRONO should be sent to the editor. It is preferred that all authors be members of the California Botanical Society. Manuscripts by authors having outstanding page charges will not be sent for review. Manuscripts may be submitted in English or Spanish. English-language manuscripts dealing with taxa or topics of Latin America and Spanish-language manuscripts must have a Spanish RESuMEN and an English ABSTRACT. Manuscripts and review copies of illustrations must be submitted in triplicate for all articles and short items (NOTES, NOTEWORTHY COLLECTIONS, POINTS OF VIEW, etc.). Follow the format used in recent issues for the type of item submitted. Allow ample margins all around. Manuscripts MUST BE DOUBLE-SPACED THROUGHOUT. For articles this includes title (all caps, centered), author names (all caps, centered), addresses (caps and lower case, centered), abstract and resumen, 5 key words or phrases, text, acknowledgments, literature cited, tables (caption on same page), and figure captions (grouped as consecutive paragraphs on one page). Order parts in the sequence listed, ending with figures. Each page should have a running header that includes the name(s) of the author(s), a shortened title, and the page number. Do not use a separate cover page or ‘erasable’ paper. Avoid footnotes except to indicate address changes. Abbreviations should be used sparingly and only standard abbrevia- tions will be accepted. Table and figure captions should contain all information relevant to information presented. All measurements and elevations should be in metric units, except specimen citations, which may include English or metric measurements. Authors of accepted papers will be asked to submit an electronic version of the manuscript. Microsoft Word 6.0 or WordPerfect 6.0 for Windows is the preferred software. Line copy illustrations should be clean and legible, proportioned to the Maprono page. Scales should be in- cluded in figures, as should explanation of symbols, including graph coordinates. Symbols smaller than 1 mm after reduction are not acceptable. Maps must include a scale and latitude and longitude or UTM references. In no case should original illustrations be sent prior to the acceptance of a manuscript. Illustrations should be sent flat. No illustrations larger than 27 X 43 cm will be accepted. Presentation of nomenclatural matter (accepted names, synonyms, typification) should follow the format used by Sivinski, Robert C., in Maprono 41(4), 1994. Institutional abbreviations in specimen citations should follow Holmgren, Keuken, and Schofield, Index Herbariorum, 8th ed. Names of authors of scientific names should be abbreviated according to Brummitt and Powell, Authors of Plant Names (1992) and, if not included in this index, spelled out in full. Titles of all periodicals, serials, and books should be given in full. Books should include the place and date of publication, publisher, and edition, if other than the first. All members of the California Botanical Society are allotted 5 free pages per volume in MAprono. Joint authors may split the full page number. Beyond that number of pages a required editorial fee of $40 per page will be assessed. The purpose of this fee is not to pay directly for the costs of publishing any particular paper, but rather to allow the Society to continue publishing Maprono on a reasonable schedule, with equity among all members for access to its pages. Printer’s fees for illustrations and typographically difficult material @ $35 per page (if their sum exceeds 30 percent of the paper) and for author’s changes after typesetting @ $4.50 per line will be charged to authors. At the time of submission, authors must provide information describing the extent to which data in the manu- script have been used in other papers that are published, in press, submitted, or soon to be submitted elsewhere. MA vet BOT INN iN hs ONTENTS W SPECIES )K REVIEWS VOLUME 48, NUMBER 2 MADRONO A WEST AMERICAN JOURNAL OF BOTANY AMONG- AND WITHIN-PROVENANCE VARIABILITY OF PINUS PONDEROSA (PINACEAE) SEEDLING RESPONSE TO LONG-TERM ELEVATED CO, Exposure Paul D. Anderson, James L. J. Houpis, David J. Anschel and James C. LES TATAVK ae nent ee anes ke Aae Soke Tse Rae re ee sarees occa oe A MorPHOMETRIC ANALYSIS OF THE LEPTOSIPHON ANDROSACEUS COMPLEX (POLEMONIACEAE) IN THE CENTRAL AND SOUTH COAST RANGES Robyn E. Battaglia and Robert PAtterSon ....ccccccccccccccccccccccesccccesceecseeseeeees MorPHOLOGICAL ASPECTS OF SEEDLING ESTABLISHMENT IN Four TEMPERATE REGION PHORADENDRON (VISCACEAE) SPECIES r Dy Beverlee M. Ruhland and Clyde L. Calvin ” s eeeeeree peer er’. omeereee er POLYPLOIDY AND SEGREGATION ANALYSES IN Deveniiuy oresopaitum (RANUNCULACEAE) 5 ‘| F BY v Jason A. Koontz and Pamela S. Soltis .... ye AGP, > sees ge er SPRING-FED PLANT COMMUNITIES OF CALIFORNIA’ S East Bay Huts Oak WOODLANDS Barbara Allen-Diaz, Randall ne Jackson and Catherine BALA ee ATRIPLEX ROBUSTA Con ORI 7 Ne EW v PERENNIAL Species FROM Nortu- ] 4 hi WESTERN UTAH A [NS >) Pe” CF ex ae Howard C. iz, Mildred R. Stutz and Stewart Cc WERNdersohe Adee... A New IPomopsis pf Oona FROM THE SOUTHWEST USA. AND ABEER Mexico | A A / | RY a hi Nis oN Fa )) a , \ \ Dieter Wilke bseeaesecdvedaaveeaseessuenssscenesepeesease NOH Nese BSS ere A NEw SPECIES OF Pos (POACEAE) FROM BAJA CALIFORNIA, Mexico — Robert J. Soreng.. AA /; ee WM BSc PPA, \3)) Neen aA: ne y V// \ ~— WAN | x ARIZONA i sscsesecsrcosdssesese MO scscssesseses MEME S25) BRP || )\ SS: \ i eee GATIFOR INIA nee eae eee sssstee ersten EI) SS) ree. SS Ee eee er ae TREES AND SHRUBS OF CALIFORNIA, BY JOHN D. STUART AND JOHN O. SAWYER VIO C CLAN CTUMGMOK gece sexs atten tes oe 2 we a Recta ome ene cc eaaaiee saeco ee New EpItor BEGINNING WITH VOLUME 49 CRITE St cere ageesieee is eo i tery, APRIL-JUNE 2001 79 90 116 PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL SOCIETY Maprono (ISSN 0024-9637) is published quarterly by the California Botanical Society, Inc., and is issued from the office of the Society, Herbaria, Life Sciences Building, University of California, Berkeley, CA 94720. Subscription information on inside back cover. Established 1916. Periodicals postage paid at Berkeley, CA, and additional mailing offices. Return requested. PosTMASTER: Send address changes to Maprono, Roy Buck, % University Herbarium, University of California, Berkeley, CA 94720. Editor—KrisTINA A. SCHIERENBECK California State University, Chico Department of Biology Chico, CA 95929-0515 kschierenbeck @csuchico.edu Editorial Assistant—JEANNIE TRIZZINO Book Editor—Jon E. KEELEY Noteworthy Collections Editors—DIETER WILKEN, MARGRIET WETHERWAX Board of Editors Class of: 2001—RobBErtT PatTErSON, San Francisco State University, San Francisco, CA PAuLa M. ScuIFFMAN, California State University, Northridge, CA 2002—NorMAN ELLSTRAND, University of California, Riverside, CA Cara M. D’ Antonio, University of California, Berkeley, CA 2003—FREDERICK ZECHMAN, California State University, Fresno, CA Jon E. Keetey, U.S. Geological Service, Biological Resources Division, Three Rivers, CA 2004—Davwp M. Woon, California State University, Chico, CA INGRID PARKER, University of California, Santa Cruz, CA CALIFORNIA BOTANICAL SOCIETY, INC. OFFICERS FOR 2000—2001 President: Bruce BALDwin, Jepson Herbarium and Dept. of Integrative Biology, 1001 Valley Life Sciences Bldg. #2465, University of California, Berkeley, CA 94720. First Vice President: Rop Myatt, San José State University, Dept. of Biol. Sciences, One Washington Square, San José, CA 95192. rmyatt @email.sjsu.edu Second Vice President: Ros SCHLIsING, California State University, Chico, Dept. of Biol. Sciences, Chico, CA 95424. rschlising @csuchico.edu Recording Secretary: DEAN KELCcHu, Jepson and University Herbarium, University of California, Berkeley, CA 94720. dkelch @sscl.berkeley.edu Corresponding Secretary: SUSAN BAINBRIDGE, Jepson Herbarium, University of California, Berkeley, CA 94720. suebain @SSCL.berkeley.edu Treasurer: Roy Buck, % University Herbarium, University of California, Berkeley, CA 94720. The Council of the California Botanical Society comprises the officers listed above plus the immediate Past President, R. Joun Litt_e, Sycamore Environmental Consultants, 6355 Riverside Blvd., Suite C, Sacramento, CA 95831; the Editor of MAprono; three elected Council Members: BiAN Tan, Strybing Arboretum, Golden Gate Park, San Fran- cisco, CA 94122; James SHEvock, USDI National Park Service, Pacific West Region, 600 Harrison Street, Suite 600, San Francisco, CA 94107; DIANE Exam, U.S. Fish and Wildlife Service, 3310 El Camino Avenue, Sacramento, CA 95825; Graduate Student Representative: KirSTEN JoHANus, Jepson Herbarium, University of California, Berkeley, CA 94720. This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). \ } “Mapron0o, Vol. 48, No. 2, pp. 51-61, 2001 AMONG- AND WITHIN-PROVENANCE VARIABILITY OF PINUS PONDEROSA (PINACEAE) SEEDLING RESPONSE TO LONG-TERM ELEVATED CO, EXPOSURE PAUL D. ANDERSON Department of Forest Resources, University of Minnesota, Grand Rapids, MN 55744 JAMES L. J. HOUPIST Department of Biological Sciences, Southern Illinois University, PO. Box 1099, Edwardsville, IL 62026. DAvID J. ANSCHEL Health and Ecological Assessment Division, Lawrence Livermore National Laboratory, Livermore, CA 94551 JAMES C. PUSHNIK* Department of Biological Sciences, California State University, Chico, CA 95929-0515 ABSTRACT Among- and within-provenance variability in growth and physiological performance were investigated in Pinus ponderosa Dougl. ex P. Laws. seedlings subjected to ambient or elevated carbon dioxide (ambient + 175 pLL™! or ambient + 350 wLL™! CO,) for 16 months. Among-provenance variability was studied with bulk-collection sources from 5 different physiographic regions of California. Within-provenance variability was investigated with three half-sibling families from a common physiographic locale. Re- gardless of source, stem volume increased at ambient + 175 pLL~!' CO,, but further increase in CO, to ambient + 350 wLL™! resulted in a variety of stem volume responses with about equal numbers of sources showing either no change or slight increases. Physiological responses to elevated CO,, including decreased efficiencies of photochemical transfer (F,/F,,), no change in stomatal conductance, and increased photo- synthesis and water-use efficiency, were consistent among half-sibling families. Thus, for this limited survey, there was little evidence for within-provenance variation in physiological response to elevated CO,. Among- and within-provenance variability in growth response to CO, suggests differing genetic control of carbon acquisition and allocation mechanisms among sources of P. ponderosa. Understanding the extent and sources of intraspecific variation in growth and physiological responses to elevated CO, is a critical need in developing management strategies that account for future altered environments. INTRODUCTION The potential impacts of increasing atmospheric CO, concentration, related increasing mean tem- peratures, and changing precipitation patterns on the sustained functioning and productivity of forest ecosystems are not well known. After many years of research on forest tree responses to elevated CO, concentrations, the level of uncertainty in estimat- ing these impacts remains high as experimental ex- posures of tree species to elevated atmospheric CO, has revealed a diverse range of responses (Gunder- son and Wullschleger 1994). Elevated CO, generally results in increased car- bon assimilation and increased growth of young trees. Growth enhancement results from physiolog- ical adaptations that optimize photosynthetic carbon acquisition and allocation processes (Acock and AI- + Present Address: Department of Biological Sciences, California State University, Chico, CA 95929-0515. * Corresponding author. len 1985; Eamus and Jarvis 1989; Pushnik et al. 1995). In general, growth responses consist of 1n- creased biomass allocated among different plant structures (e.g., foliage, stems and roots) and may result in shifts in root:shoot ratios (Callaway et al. 1994; Walker et al. 1995). Growth responses vary widely among genera (Tolley and Strain 1984), among species within the same genus (Rogers et al. 1994), and intraspecifically (Houpis et al. 1995). Differences among taxa suggest that multiple fac- tors, possibly under genetic control, may be oper- ating to determine the physiological and growth re- sponses to elevated atmospheric CO, concentra- tions. There have been few investigations of intraspe- cific variation in CO, responses for forest trees. Published studies have addressed intraspecific vari- ation in terms of response differences among eco- types (e.g., Pinus ponderosa, Callaway et al. 1994; DeLucia et al. 1994; Houpis et al. 1995), among provenances (e.g., Picea mariana, Johnsen and Ma- 52 MADRONO jor 1998) and within stands (e.g., Betula alleghan- iensis, Wayne and Bazzaz 1995). Typically, studies to date have addressed relatively narrow sources of variation by either considering relatively few ge- notypes or relatively distinct ecotypes. The works of Johnsen and others with P. mariana are among the most robust studies of provenance X CO, in- teractions based on the number and range of genetic families considered (Johnsen and Seiler 1996; Johnsen and Major 1998). Pinus ponderosa Dougl. ex Laws. is common throughout much of the forested western United States, and as such is of great ecological and eco- nomic importance. Pacific and Rocky Mountain va- rieties of P. ponderosa differ in growth and phys- iological responses to elevated concentrations of CO, (Surano et al. 1986; Houpis et al. 1988; Surano and Kercher 1993). Sierra Nevada seedlings dem- onstrated substantially greater increases in stem height and stem volume at CO, concentrations of 150 and 300 wLL™! above ambient than did Rocky Mountain seedlings. Following two years of growth under elevated CO,, seedlings of Rocky Mountain origin had shed most of their two-year-old needles and a large percentage of their 1-year-old needles. In contrast, seedlings of Sierra Nevada origin main- tained more age classes of foliage at elevated CO, concentrations but displayed extensive mid-needle abscission in older age classes and a twisting de- formation of current-year needles. In California, P. ponderosa is found in several diverse physiographic/climatic regions including the north-south oriented coastal and Sierra Nevada mountain ranges, the Klamath Mountains of the northwest, the Modoc Plateau in the northeast, and in the transverse mountain ranges of southern Cal- ifornia. Climates vary from temperate and humid along the northern Pacific coast, to cold and semi- arid in the eastern Sierra Nevada (Bailey 1994). Pinus ponderosa occurs aS a component of four major California forest types in which species com- position varies with physiographic and climatic site characteristics: Pacific ponderosa pine, interior pon- derosa pine, Sierra Nevada mixed conifer, and Pa- cific ponderosa pine—Douglas-fir (Eyre 1980). Seed zone and breeding zone stratification in the U.S.ES., Pacific Southwest Region breeding pro- gram (Kitzmiller 1976) implies that physiographi- cally or climatically adapted sub-populations of P. ponderosa may have evolved. It is unknown if sub- populations differentially adapted to climate and physiography will respond to future atmospheric carbon dioxide concentrations similarly or differ- ently. Our study surveyed variability in P. ponderosa seedling responses to elevated CO,, both among provenances (geographic locations of natural ori- gin) representative of the species occurrence in sev- eral major forest types of California, and within a provenance using genotypes of P. ponderosa hav- ing high growth potential. Physiological and growth [Vol. 48 | \ parameters were used to assess response variability | at multiple scales of plant structure and function. | Elucidation of underlying sources of intraspecific | variability at both physiological and whole-plant scales is a critical step in the development of pro- cess and ecological models for the assessment of | climate change impacts on forest community dy- namics. Predictions of genotype and population re- | sponse to climate change can be used to develop | genetic resource and silvicultural management strategies to ensure the maintenance of genetic di- | versity, ecosystem integrity, and forest productivity — in a future environment. | MATERIALS AND METHODS Plant material and growth conditions. Pinus pon- derosa seedlings from seed sources of different geographic origin were exposed to various atmo- spheric CO, concentrations at the Lawrence Liv- | ermore National Laboratory (LLNL) exposure fa- | cility. Among-provenance variability was evaluated | with seedling sources originating as bulked seed | collections from stands in each of 5 different phys- | iographic regions in California. Each bulk collec- | tion was made up of seed from 25 to 50 individuals | per stand, with stands typically being less than 25 | ha in size. These 5 sources included two coast | range provenances (Mendocino, north coastal | range, and Santa Clara, central coastal range), a_ provenance from the southern California transverse | mountain ranges (San Bernardino), and two prov- | enances from the Sierra Nevada (El Dorado, west- | side Sierra, and Tahoe, east-side Sierra). Although © ponderosa pine is a common species in forests of | each, the 5 physiographic regions have distinct cli- © matic, geologic and vegetative characteristics that are reflected in different ecological classifications (Table 1; Fig. 1). Within-provenance variability was assessed using seedlings of 3 half-sib families (fam- ilies 3087, 3088, and 3399), obtained from maternal | parents located within a 2-km radius at ca. 1500-m elevation in the central Sierra Nevada of California — (El Dorado National Forest). Growth and physiol- ogy of the 3 half-sib families have been intensively characterized in previous studies examining geno- — typic variation in ponderosa pine response to ozone and acid deposition (Benes et al. 1995; Anderson et al. 1997; Momen et al. 1997). Bulk-collection seedlings were obtained as 1-0 bare-root stock from the California Department of Forestry nursery in Davis, California. Half-sib seedlings were obtained as one-year-old container stock from the U.S.ES. Genetic Resource Center in Chico, California. Upon receipt at LLNL, the seed- lings were transplanted to 12.8-L pots containing a loam soil mix formulated to optimize pine seedling growth (American Soil Co., Berkeley, CA). The soil mix, which had good aeration and high nutrient and water-holding capacity, consisted of 2 parts clay (pulverized expanded shale), 3 parts red lava 2001] ANDERSON ET AL.: INTRASPECIFIC VARIABILITY IN RESPONSE TO CO, Ss “TABLE 1. ECOLOGICAL CLASSIFICATION AND CLIMATIC CHARACTERISTICS OF SEEDLING PROVENANCES AND THE COMMON- /GARDEN TEST SITE. Seed zone designations are those of the USDA Forest Service, Region 5 Tree Improvement Program. ‘Ecological classification is according to the USDA Forest Service ECOMAP program (McNab and Avers 1994). Climatic descriptions apply to the ecological classification subsections from which the provenance collections were -made and are based on subsection descriptions by Goudey and Smith (1994) and Miles and Goudey (1997). Mean annual climatic Location conditions of subsection Gis SINGS Ecological classification Winds Big nce seed CCL) eee ea ee, S Piecipita- ature period Source zone (m) Section Subsection tion (cm) (°C) (d) Bulk-collections (among provenance) ' Mendocino 351 610 Northern California M261Ba 100-305 5.4—12.3 100—200 Coastal Range Santa Clara 097 150 Central California M261 Af 50-150 10.0—14.4 250—300 Coastal Range San Bernardino 994 1525 Southern California M262Bh 75—100 4.4-10.0 150—200 | Mountains and Valleys Tahoe 2 1525 Sierra Nevada (east- M2615; 50-100 1.7-7.2 25-75 side) El Dorado 526 915 Sierra Nevada (west- M26lEg 100—205 7.2-12.8 100-150 side) Half-siblings (within provenance) Families 3087, 526 1370 Sierra Nevada (west- M261Eg 100—205 7.2-12.8 100—150 3088, 3399 side) LLNL Common Garden Test Site 30 Great Valley 262Aq 20-41 15.0—-16.7 250-275 Ecological Subsection Classification for Provenances M262Bh - San Bernardino M261 Ej - Tahoe M261 Ba - Mendocino 261 Af - Santa Clara M261Eg - El Dorado, Half-siblings 262Aq - Study Site Ecological Subsections of California ? il a | ay iy 2 = ‘ ACS A oN Lae e a. Fic. 1. Ecological subsections of California. Shaded areas represent subsections from which bulk-collection or half- sibling seedlings originate, and the subsection in which the study site exists. 54 MADRONO rock (ground to particle sizes less than 2 mm di- ameter, resulting in a coarse sand texture), | part colma sand, and 3 parts bark. The seedlings were grown for 2 years prior to CO, fumigation. At the start of the experiment, seedling were distributed to 18 standard outdoor open-top exposure chambers (3-m diameter and 3-m height; Rogers et al. 1983). All seedlings were watered at 2-3 day intervals and were fertilized at 4—6 week intervals during the growing season using a one-half strength Hoag- land’s solution. CO, treatments. Seedlings were exposed to 3 CO, concentrations including ambient (ca. 350 wLL~! CO,), ambient + 175 wLL7! CO,, and am- bient)+ 350 wll! CO, to cover the range from existing mean concentration through the doubling in mean concentration expected to occur within the 21st century. The open-top chamber CO, concen- trations were monitored using dedicated CO, ana- lyzers (Horiba Model PIR-2000). Analyzers were zero- and span-checked daily and under went a multi-point calibration monthly. Chamber atmo- spheres were sampled 12 times per hour at canopy height, at the center of the chamber. The chamber CO, concentrations were maintained within +5% of the treatment concentration, 24 hr per day from April through completion of second season stem elongation in July for a total of 16 months. Seedling growth. Seedling total height and basal diameter were measured during September and July, following cessation of growth, in the first and second seasons. Pre-experimental height was deter- mined from leaf scars at the base of the stem seg- ment. Estimates of total height and diameter were used to calculate an index of stem volume (7 * radius’ * height). Total height growth was estimated as the difference between the final and pre-experi- mental total heights. Stem volume growth was es- timated as the difference between first and second season’s main stem volumes. Light harvesting system. The relative photochem- ical efficiency of electron transport reactions for photosystem II (PSII) was estimated by chlorophyll fluorescence (Long and Drake 1992). Dark-adapted fluorescence measurements were made in situ for both current-year and one-year-old foliage using a portable chlorophyll fluorescence measurement sys- tem (CF-1000, Morgan Scientific, Andover, MA). Fluorescence was induced with an excitation light intensity of 750-~mol m~? s~! and fluorescence ki- netics was monitored for the subsequent 20 sec- onds. An index of PSII quantum efficiency was cal- culated as the ratio of the variable fluorescence component to maximal fluorescence (Fv/Fm; Genty et al. 1989). Measurements were made in July of the second growing season of experimental expo- sures, in conjunction with gas-exchange sampling. The concentrations of chlorophyll a and b, and carotenoids were determined for current-year and one-year-old foliage. Following determination of [Vol. 48 } foliar surface area, pigments from foliage samples were extracted in 5 ml of N,N-dimethylformamide in the dark and at 4°C for a period of 14 days (Mor- | an and Porath 1980). Concentrations of the three! pigments in the solution were calculated according to Wellburn and Lichtenthaler (1983). | Gas exchange. CO, and H,O vapor flux of one- year-old foliage were measured in July following | cessation of current-year needle elongation with a closed-loop photosynthesis system consisting of a portable infrared gas analyzer and microprocessor | controller (LI-6200, Licor Inc., Lincoln, NE) cov- | pled to a 0.25 L cuvette. All measurements were | made mid-morning (O900—1030) inside the open- | top chambers at the growth CO, concentration +— 15 pLL™'. Light intensity within the cuvette was _ maintained at 1000 pmol m~ s~! (above the satu- | rating light intensity for P. ponderosa) using a LED | array with peak spectral radiation of 670 nm wave- | length (QB-2001, Quantum Devices Inc., Barne- | veld, WI). Mean (+ | SE) leaf temperature over all | gas exchange measurements was 30.1 + 0.4°C. The | mean leaf-air vapor pressure deficit VPD over all | gas exchange measurements was 1.48 + 0.14 kPa. | Estimated gas-exchange parameters included net photosynthesis (P,), stomatal conductance to water | vapor (g,,), leaf internal CO, concentration (C,), and photosynthetic water-use efficiency (WUE, net photosynthesis rate/transpiration rate). Experimental design. The study was conducted using a split-plot design. Three levels of atmospher- | ic CO, concentration, the main plot factor, were randomly assigned to 18 open-top chambers to pro- vide 6 replications of each CO, concentration. Within each chamber, seedlings of the 3 half-sibling families and 5 bulk-collection sources represented a sub-plot factor. Three seedlings per bulk-collec- tion source and one seedling per half-sib family were grown in each chamber. Growth, fluorescence, and pigmentation responses were measured for all 8 families. Gas-exchange responses to elevated CO, were assessed in the half-sibling families only. Sep- arate statistical analyses were performed for among- and within-provenance evaluations. Significance of genetic source, CO, treatment ef- fects, and their interaction on fluorescence and gas exchange parameters were evaluated using ANO- VA. Differences with a type I error probability of 0.05 or less were considered significant. Seedling growth responses may be dependent on seedling size at the initiation of CO, exposures. Therefore, ANACOV was used to analyze treatment effects on total height growth and stem volume growth with pre-experiment seedling height as a covariate to ac- count for initial variation in seedling size. For those parameters having significant among-provenance or within-provenance effects, linear contrasts were used to determine significant differences between: 1) west-side and east-side Sierra Nevada sources; 2) coast range and Sierra Nevada sources; 3) Sierra 55) ANDERSON ET AL.: INTRASPECIFIC VARIABILITY IN RESPONSE TO CO, an =) =) I ‘7Z0'0 I ‘C110 LT 8S0 I ‘L100 I ‘r00'0 I ‘ZOO I ‘1000 b T1160 1 ‘890°0 [ceo °H < qolg 9S + 98h PS + e88s Co + 86h Sv © 2261S ZS + «S99 [S-= ucis C7 + wbtr O€ FH «xb8S [Oo TV) ZS = 609 L8 % v6S ZB = 86S SII = reo 6€ + OT9 L9 = 009 86 + IL8 Cou VGC tbr + €8P 6€ = 79 ose + quy Or + 6SS [G2 = 657 Ill = LE9 69 = €9¢ CE = OGG S6 BTS 6b = 98S SII + 6S ve = ISP Cy =:079 CLI + quy 9S + Ob TEI + SBE 96 + O€S CC. ER6C St + Ler L9 = 6CP 6b + BES Se + OIr O€ = IVE 99 = TIS JUSIQqUIY (,-WI9) YIMOIS SUINJOA UWIIIS b *Z¢0'0 I ‘8700 b ‘OOL'0 | LELO0 I ‘7S9'0 I ‘80 b ‘o9r'0 b ‘OSL'0 I “$790 b ‘geto °H < qOlg OC Sas CE 1 £20169 - 6:0 = 6 16 LC Hebel OC aCe “Cea Sh Uo ee00R To eS “OD IIV CC = 8165, PS 2.086 ve x TSO Coz. 9CC ITH E8b WP + BOS li = 2S 67 = Cr Srp +O8E Of + S05 Ose + quiy Cie 9 Ob = 95 00 O'S) 60 = Iss C64 Lor ¢€F5 = 005 iS oe AS (ea) Oh OG 2 Olbp- Sc 2 ORC CLI + quiv FEES CSE Lr eS + 799 O€ + ISP CS Lor moo cL CSCC eG So = OCP OW 44566 DES = Gils JUSIqUIY (Wd) MOIS JYSIOH, SoT[TWWe 66EL S80E L8OE SUOTJDV[[OD Oopelod IH BIC) eyURS OUTIpIeUIS | SOURL OUTDOpUs|[J JUIUTCOIT Bul|qIs -xING ITV BIIDIS “AA [BIsvO09 “D ues BLIDIS “A [B1svoo “Ny ‘OO “Jey TV a a ee = = fd a ie a i = a ° > eon DdOINOS PIasS UOT]Sa][OO-y[Ng 9OINOS PIes QIS-J[eH YIMOIS SUITPIaS “sJSBI]UOD [BUOBOYIIO Aq POUTUIDJap se ddINOS YORA IO} $}D9JJ2 “OD (b) oNeIpenb Jo (1) IVSUT] JO DOUROYTUSIS oY) SIVOIPUT saNTIqeqolg (SQ’'O > d) ATfeousNels Josip jou op Jae] UOWUTUIOD & Aq payeUsIsap s[aag] “OOD [[B J9AO suv sdUvUDdAOId JO ATTUTe “IYStoY [210 JUDUNeaN)-a1d UI UOTeLIeA JOJ JUOUNSN[pe Ue BUNeIOdIOOUI AS F suvoU Sarenbs-j}sea] aie saNTeA SUINIOA Wa}g “UOHRUIqWIOS JUDUTIRAN “OD x 90I1NOS Jed sUOTIBAJaSgO 9 = U JOJ YS F SuvoU oie son[eA “SHOANOS ONIIGHAS NOILOATIOD-N1InNg GNV dIS-dTIVH YOd “OD ANV AOMNOS Ad HLMOUD ANNTIOA GNV HLMOUD LHOIGH ONITGHAS °7 ATaAVL 56 MADRONO [Vol. 48) | TABLE 3. FOLIAGE PIGMENT CONCENTRATIONS AND PHOTOSYSTEM II EFFICIENCY BY SOURCE AND CO, TREATMENT FOR 5! BULK-COLLECTION (AMONG-PROVENANCE) AND 3 HALF-SIB (WITHIN-PROVENANCE) SEEDLING SOURCES. Values are means | one standard error of the mean. Source X CO, treatment means are based on n = 10—12 observations. Source means. averaged over all CO, treatments denoted by a common letter do not differ at the P = 0.05 level of significance. Probabilities indicate the significance of linear (1) or quadratic (q) CO, effects for each source as determined by, orthogonal contrasts. Foliage pigment concentrations and photosystem II efficiency Bulk-collection seed source San CO, N. coastal E. Sierra Bernardino C. coastal W. Sierra Treatment BCS BC6 BC7 BC8 BC9 Total chlorophyll (wg cm?) Ambient 195e2e1-01 17.8 + 1.8 207 15.6% 1-2 14.8 + 1.7 Amb + 175 OSes o-6 Was ie? |e eo gana bess) 13.6° 22.12 12.352 el Amb + 350 loo 213 13-57. 16 Wed eras i) 15.3 + 1.4 13.592 110 All CO, NS ofS ase 0) 16.67 = 0.9 16,9225. 1.0 14.85 + 0.7 13.52 = 038 Prob. > Ho 0.021, q O225.11 0.033, q 0.893, | 0.625, | Carotenoids (wg cm?) Ambient 94+ 0.4 ropa 0} =) 9.1 = 0:8 (ides cers Ue) 7.4 + 0.7 Amb + 175 8.4 + 0.8 8.0 = O16 6.6 + 0.7 6.6 + 0.4 6.1 + 0.4 Amb + 350 825° 2205 7.3 + 0.6 8.1 = 0.6 T.1 = 0.6 6.7 + 0.4 All CO, Sy) a0S 8.0% + 0.4 7.9% + 0.4 TAP F038 6.7" 21013 Prob. > Hy 0.186, | 0.086, 1 0.004, q O55554 0.452, 1 Photosystem II efficiency (F,/F,,) Ambient 0.780 + 0.070 0.780 + 0.008 0.767 + 0.005 O22 01022 0.749 + 0.007 Amb + 175 0.748 + 0.010 0.759 + 0.008 0.722 + 0.009 0.682 + 0.026 0.688 + 0.002 Amb + 350 O776 = 0.011 0.740 + 0.009 0.765 + 0.009 0.707 + 0.015 0.714 + 0.017 All CO, 0.7684 + 0.005 0.760? + 0.005 O77 52 = 0,005 0.718 + 0.012 0.704° + 0.010 Prob. > Hy 0.079, q 0.040, 1 0.010, q 0.060, q 0.015, q Nevada sources and the San Bernardino source; and 4) coast range sources and the San Bernardino source. Response to CO, concentration as a main effect, or within-provenance or family was tested using orthogonal contrasts. Orthogonal contrasts discern responses having either significant linear or quadratic trends over the range of CO, treatments. All analyses were performed using PROC GLM of SAS v. 6.11 (SAS 1989). RESULTS Foliage-age class variation. Photosystem II effi- ciencies did not differ among foliage age-classes. In half-sib seedlings, foliage pigment concentra- tions were significantly greater (P < 0.001) in one- year-old foliage than in current-year foliage. In bulk-collection seedlings, pigment concentrations were slightly greater (P = 0.05—0.09) in one-year- old foliage. Interaction effects of foliage age-class by CO, and by genotype were non-significant. From this point forward, results and discussion of PSII and pigmentation results refer to means aver- aged over both foliage age-classes. Among-provenance variation. Substantial among- provenance variation in growth was detected. When averaged over all CO, treatments, height growth varied from 40 cm for the Tahoe provenance to 54 cm for the Mendocino provenance, and volume growth ranged from 424 cm? for the Tahoe prove- ' nance to 665 cm? for the Santa Clara provenance (Table 2). Stem volume growth, averaged over all provenances, was 39% greater under ambient + 350 CO, than under ambient CO,. However, stem volume growth responses to elevated CO, were only significant for the El Dorado (P = 0.001) provenance (Table 2). In contrast to all other prov-— enances, the stem volume growth of the San Ber- nardino provenance was greatest in the ambient + 175 CO, treatment and tended to decline with fur-— ther CO, increase to ambient + 350 (Table 2). Height growth response to CO, was generally non- significant (P = 0.652 over all provenances) and patterns of response to increasing concentration were inconsistent among provenances (Table 2). Pre-treatment height did not contribute significantly to the explanation of variance in height growth and was not included as a covariate in the analysis pre- sented here. Foliage pigmentation and PSII efficiency also differed significantly among provenances. Total chlorophyll and carotenoid concentrations ranged from 13.5 and 6.7 wg cm~?, respectively, for the El Dorado provenance to 19.9 and 8.7 wg cm~ for the Mendocino provenance. Photosystem II efficiency (F,/F,, ratio) ranged from 0.704 for the El Dorado provenance to 0.768 for the Mendocino provenance (Table 3). As with height growth, pigmentation and | Bulk-collection seed source ANDERSON ET AL.: INTRASPECIFIC VARIABILITY IN RESPONSE TO CO, 7 TABLE 3. EXTENDED. Foliage pigment concentrations and photosystem I efficiency Half-sib seed source All BC 3087 3088 3399 All HS Total chlorophyll (wg cm~’) 725° 0.7 15.8 + 1.3 | oveey es ® [Be L522 186 16.3 + 0.8 1631 201.3 1333; 25143 159 1d 16.8 + 1.0 |Leies acne 009 ibsyopaans OF c! Sea 14.8 + 1.4 VA ee 1D 37 Ow 13165227037. 15.47 + 0.7 lo:3) 07, 0.146, 1 0.002, | 0.361, 1 0.055, 1 0.022, | Carotenoids (wg cm?) 8.4 + 0.3 Toe 00S 1,6 22016 $33 2.07 18 210.3 Yee 03 6.8 + 0.5 T6,22205 8.0 + 0.4 I-03 These moO) Sey OS TOezEAOS 12 Oey 6.7 + 0.3 CO 20 gay 20.38 7.8 + 0.3 0.128, 1 0.002, | 0.239, | 0.061, | 0.030, | Photosystem II efficiency (F,/F,,) 0.760 + 0.006 0.750 + 0.006 0.775 + 0.009 0.778 + 0.008 0.768 + 0.005 0.720 + 0.009 0.735 + 0.011 0.740 + 0.013 0.756 + 0.012 0.744 + 0.007 0.740 + 0.006 0.726 + 0.014 0.736 + 0.012 0.744 + 0.015 0.735 + 0.008 0.7592 + 0.006 0.7502 + 0.007 0.737" + 0.007 0.039, | 0.115, 1 0.012, 1 0.029, | i 0.025, 1 PSII efficiency response to CO, was inconsistent. Total chlorophyll content of the Mendocino prov- enance decreased with increasing CO, from ambi- ent + 175 to ambient + 350. In contrast, total chlo- rophyll and carotenoid concentrations of the San Bernardino provenance were lowest under the am- bient + 175 treatment (Table 3). Pigment concen- trations of the other bulk-collection provenances were not significantly impacted by CO, concentra- tion. In general, PSII efficiency of the bulk-collec- tion provenances were greatest under ambient CO,, and minimal under the ambient + 175 treatment; the only exception being the El Dorado provenance for which F,/F,, decreased linearly with increasing CO, (Table 3). Contrast analysis indicated that height growth and physiological performance differed significant- ly between Tahoe and El Dorado provenances; the former having greater pigmentation and PSII effi- ciency but lesser height growth than the latter (Ta- bles 2, 3, and 4). Growth and physiology of the Sierra provenances could not be distinguished from that of the San Bernardino provenance. In contrast, the coastal provenances had growth rates that were TABLE 4. CONTRAST ANALYSES OF AMONG-PROVENANCE SEEDLING GROWTH AND PHYSIOLOGY. Among-provenance contrast East-side Sierra Sierra vs. Coastal vs. VS. Coastal vs. Sierra San Bernardino San Bernardino West-side Sierra (Mendocino, Santa (Tahoe, El Dorado (Mendocino, Santa (Tahoe vs. Clara vs. Tahoe, vs. San Clara vs. Parameter El Dorado) El Dorado) Bernardino) Santa Barbara) Probability > H, Height growth 0.018 0.001 0.550 0.005 Volume growth 0.837 0.001 0.701 0.008 Total chlorophyll 0.063 0.403 0.163 0.831 Carotenoids 0.008 0.104 0.149 0.882 Photosystem II <0.001 0.786 0.224 0.150 efficiency 58 clearly superior to those of the Sierra provenances and the San Bernardino provenance (Tables 2 and 4). Within-provenance variation. Averaged over all CO, treatments, height growth of family 3088 was approximately 32% greater than that of families 3087 and 3399 (Table 2). Volume growth by family 3088 was approximately 18 and 21% greater than that of family 3087 and 3399, respectively (Table Pay With increasing CO, concentration, both families 3087 and 3399 demonstrated a tendency for in- creased volume growth (linear effect, P = 0.02 and 0.11, respectively, Table 2). In contrast, volume growth by family 3088 was not substantially influ- enced by CO, treatment (Table 2). Pigment concentrations were greatest for family 3087 and least for family 3399 with the differences among families being significant (P < 0.05). In contrast, PSII efficiency was greatest for family 3399 and least for family 3087 (P < 0.05). In gen- eral, pigment concentrations declined with increas- ing CO, concentration from ambient to ambient + 350, but the effect was pronounced in families 3087 and 3399 (13-27% decline) but not in family 3088 (4-8% decline) (Table 3). PSII efficiency ranged from 0.737 for family 3399 to 0.759 for family 3087. However, all families demonstrated a de- crease in PSII efficiency with increasing CO, con- centration from ambient to ambient + 350 that av- eraged 4% (P = 0.12, 0.01, and 0.03 for families 3087, 3088, and 3399, respectively). Averaged over all CO, concentrations, net pho- tosynthetic rates ranged from 3.9 pmol m~? s7! for family 3088 to 4.5 wmol m~?’ s7! for family 3399, but differences among families were not significant (Figure 2). Similarly, non-significant family differ- ences in water-use efficiency were also observed. Both net photosynthesis and water-use efficiencies increased significantly with increasing CO, in all three families from 71 to 121% (P = 0.01 to 0.05). Thus, gas exchange responses to CO, concentration varied less among half-sib families than did pig- mentation, PSII efficiency, or growth responses. DISCUSSION The substantial among-source variation in ele- vated CO, growth enhancement we observed is consistent with earlier long-term studies comparing Sierra Nevada and Rocky Mountain varieties of P. ponderosa (Surano et al. 1986; Houpis et al. 1988). In contrast, greenhouse studies of either 3 black spruce provenances from the species range in Can- ada or 20 families of black spruce from the New Brunswick province concluded that there were no significant CO, X family effects on growth (John- sen and Seiler 1996; Johnsen and Major 1998) and that such interactions are more likely to occur under field conditions where other environmental stresses co-occur (Johnsen and Major 1998). MADRONO [Vol. 48 | | —@— Family 3087 —w— Family 3088 —H-— Family 3399 Photosynthesis (mol mol" s”) = (b) E 0.05 i = E @ 0-04 Oo — Ss 5 0.03 To = e) O 0.02 8 (c) > O — 6 wo i it © 4 ”) a} fe ® w 2 = 0 Amb Amb+175 Amb+350 CO, Treatment Fic. 2. Foliage gas exchange characteristics by family and CO, treatment for three half-sib (within-provenance) seedling sources: a) net photosynthesis, b) stomatal con- ductance, and c) water-use efficiency. Values are means + 1 SE of the mean. Source X CO, treatment means are based on n = 3 observations. Net photosynthesis response to CO, (linear effect) was significant for families 3087 and 3399 (P = 0.010 and 0.012, respectively). WUE efficiency response to CO, was significant for families 2087, 3088, and 3399 (P = 0.023, 0.032, and 0.005, respectively). The effect of CO, on stomatal conductance was not significant for any family. 2001] The climate of the study site, located in the Great Valley ecological subregion, is characterized by -hot, dry summers and mild winters (Table 1). The mean annual temperature is between 14 and 17°C. The higher temperatures and the longer growing _season at the common garden site provide reason- _able conditions for testing the influence of elevated CO, given the coupling between expected global increases in temperature and atmospheric CO, con- centration. Being a potted seedling study, finite soil water and nutrient resources necessitated irrigation and fertilization to maintain seedling survival and growth. Because of the difficulties in mimicking natural variation in soil resource availability in pots, we chose to supply ample water and nutrients to decrease the potential confounding of limitations of these resources among individual seedlings or among provenances. How potted seedlings would perform relative to seedlings in the ground is dif- ficult to predict without knowledge of soil moisture depletion. At the leaf-level, the high evaporative demand coupled with stomatal aperture being sen- sitive to leaf-water deficits (Anderson 1991; An- derson and Helms 1994) would probably lead to reduced gas-exchange for seedlings growing either in pots or in the ground. Based on previous studies with planted seedlings of families 3087, 3088, and 3399, the effect of the relatively high leaf temper- atures in this study would be to decrease net pho- tosynthetic rates by ca. 5—15% from maximum rates at leaf temperatures of 25 to 26°C (Helms et al. 1994). The five bulk-collection sources in this study originate in distinct physiographic regions of Cali- fornia and therefore have different degrees of ad- aptation to the common garden environment. The two sources demonstrating significant or nearly sig- nificant response to elevated CO,, the Tahoe and the Santa Clara sources, respectively, originate in distinctly different environments. Callaway et al. (1994) observed differences in CQO,-induced changes in biomass allocation to foliage and roots of P. ponderosa seedlings of 4 geographic origins. As with the variation in volume growth response we observed in bulk-collection sources, the among- population differences they observed were not con- sistently related to the contrasting montane or des- ert origins. This implies that response to CO, may not be strongly driven by environment of origin, but rather by the genetic potential of the particular populations sampled. The limited number of bulk- collection sources in this study serves to illustrate potential among-population variation across the species range in California, but is not sufficient to characterize variation among populations within geographic regions, and is therefore inadequate for making comparisons of mean population perfor- mance between geographic regions. Further, vari- ability in relative growth response to elevated CO, among the bulk-collection sources may reflect not only intraspecific variation in CO, response mech- ANDERSON ET AL.: INTRASPECIFIC VARIABILITY IN RESPONSE TO CO, 59 anisms, but also intraspecific variation in adaptation to the common garden climate. In spite of this po- tentially confounding effect, the large variation in growth response among bulk-collection sources grown at the common garden site provides a mea- sure of the variability in CO, response that exists among provenances of California ponderosa pine. The varied growth responses to elevated CO, by three half-sib families indicate that substantial ge- netic variation exists independent of potential con- founding influences of geographic origin. Future ef- forts to characterize genotype or provenance X CO, interaction would benefit by increased numbers of genetic families per geographic region and repeti- tion of the common garden planting at multiple lo- cations over all geographic regions of interest (We- ber et al. 1996). The scale of provenance testing may determine the extent to which genotype x CO, interactions are observed. After failing to detect significant provenance X CO, interaction among three families of black spruce representing very diverse geograph- ic origins, Johnsen and Seiler (1996) hypothesized that CO, X genotype interactions would become more significant as the genetic background of the families became more similar. Our study supports this hypothesis, as provenance X CO, interactions were not evident in comparisons among ponderosa pine bulk-collection sources originating in distinct- ly different environments, while significant geno- type xX CO, interaction effects were detected for within-provenance comparisons of growth. At the local scale, all three half-sibling families had similar physiological responses to increasing CO, concentration: a decrease in foliage pigmen- tation; a decrease in photosystem II efficiency; an increase in net photosynthetic rate; and increased water-use efficiency. Thus, processes associated with carbon assimilation were responding to CO, concentration similarly, but differences in the effi- ciencies of assimilate conversion to biomass or dif- ferences in the allometric patterns of biomass al- location resulted in significant among-family dif- ferences in above-ground growth response. In summary, carbon assimilation was enhanced with elevated CO, for all sources of P. ponderosa studied. Enhanced assimilation was accompanied by enhanced volume growth and in some cases en- hanced stem elongation. The effect of elevated CO, on stem height growth varied both among- and within-provenances. Those sources having more vigorous height growth under ambient CO, condi- tions tended to demonstrate greater enhancement of height growth under elevated CO,. Assimilation rates were enhanced under elevated CO, in spite of decreased light harvesting capacity, suggesting an increase in Overall photosynthetic efficiency. Geo- graphic patterns of provenance growth response to elevated CO, were not apparent. Understanding genotypic variability in response to elevated CO, is essential to the development of 60 MADRONO forest management strategies. Although this study is limited to demonstrating a potential range of variation present in a limited sample, it serves as a model that, if applied with greater sampling inten- sity, can be used to characterize the adaptive ge- netic population structure of ponderosa pine in Cal- ifornia. Such knowledge can be used by ecologists to better predict the performance of forest tree pop- ulations and will guide forest geneticists and re- source managers in the breeding and deployment of genotypes that foster genetic diversity and ecosys- tem resilience in the face of a changing global cli- mate. ACKNOWLEDGMENTS We gratefully acknowledge Thorpe Loeffler, Roland Ng, and Ramford Ng for their collection and processing of data. We also acknowledge Dr. Marian Smith and two anonymous reviewers for their comments on an earlier version of this manuscript. This research was performed under the auspices of the U.S. Department of Energy at Lawrence Livermore National Laboratory under contract W-7405-Eng-48 and partially supported by a grant from National Institute for Global Environmental Change West- ern Region W/GEC 92-037. LITERATURE CITED ANDERSON, P. D. 1991. Water use by Pinus ponderosa and Arctostaphylos patula in response to environment. Unpublished Ph.D. dissertation. Wildland Resource Science, University of California, Berkeley, CA. AND J. A. HELMS. 1994. Tissue water relations of Pinus ponderosa and Arctostaphylos patula exposed to various levels of soil moisture depletion. Canadian Journal of Forest Research 24:1495—1502. , J. L. J. Houpis, J. A. HELMS, AND B. MOMEN. 1997. Seasonal variation of gas-exchange and pig- mentation in branches of three grafted clones of ma- ture ponderosa pine exposed to ozone and acid rain. Environmental Pollution 97:253—263. Acock, B. AND L. H. ALLEN. 1985. Crop responses to elevated carbon dioxide concentration. Pp. 53-97 in B. Strain and J. Cure (eds.), Direct Effects of Increas- ing Carbon Dioxide on Vegetation. U.S. Department of Energy, Washington, D.C. BAILEY, R. G. 1994. Descriptions of the ecoregions of the United States, 2nd ed. Forest Service, U.S. Depart- ment of Agriculture. Misc. Publ. 1391. BENES, S. E., T. M. Murpuy, P. D. ANDERSON, AND J. L. J. Houpts. 1995. Relationship of antioxidant enzymes to ozone tolerance in branches of mature ponderosa pine (Pinus ponderosa) trees exposed to long-term, low concentration, ozone fumigation and acid precip- itation. Physiologia Plantarum 94:124—134. CALLAWAY, R. M., E. H. DeLucta, E. M. THOMAS, AND W. H. SCHLESINGER. 1994. Compensatory responses of CO, exchange and biomass allocation and their effects on the relative growth rate of ponderosa pine in different CO, and temperature regimes. Oecologia 98:159—-166. DeLuca, E. H., R. M. CALLAWAY, AND W. H. SCHLESING- ER. 1994. Offsetting changes in biomass allocation and photosynthesis in ponderosa pine (Pinus ponder- osa) in response to climate change. Tree Physiology 14:669-676. \ EAMus, D. AND P. G. JARVIS. 1989. The direct effects of increase in the global atmospheric CO, concentration | on natural and commercial temperate trees and for- | ests. Advances in. Ecological Research 19:1—55. Eyre, EF H. (ed.). 1980. Forest cover types of the United | States and Canada. Society of American Foresters. Washington, DC. GENTY, B., J. BRIANTAIS, AND N. R. BAKER. 1989. The relationship between the quantum yield of photosyn- | thetic electron transport and quenching of chlorophyll | fluorescence. Biochimica et Biophysica Acta 990:87— o2. | GOUDEY, C. B. AND P. W. SMITH (eds.). 1994. Ecological | units of California: Subsections (map). U.S. Depart- | ment of Agriculture, Forest Service. GUNDERSON, C. A. AND S. D. WULLSCHLEGER. 1994. Pho- | tosynthetic acclimation in trees to rising atmospheric _ CO,: A broader perspective. Photosynthesis Research | 39:369-388. | HELMs, J. A., P. D. ANDERSON, AND J. L. J. Houpts. 1994. | Final report for the project ““Gas Exchange by Pinus ponderosa in Relation to Atmospheric Pollutants.”’ California Air Resources Board. Sacramento, CA. Houpis, J. L. J., K. A. SuRANO, S. COWLES, AND J. H.. SHINN. 1988. Chlorophyll and carotenoid content of — two varieties of Pinus ponderosa seedlings subjected — to long-term elevated carbon dioxide. Tree Physiol- | ogy 4:187-193. , J. C. PUSHNIK, D. J. ANSCHEL, AND P. D. ANDER- SON. 1995. Variability in the intraspecific response of Pinus ponderosa seedlings subjected to long-term ex- posure to elevated CO,. Air and Waste Management | Association, Pittsburgh, PA. JOHNSEN, K. H. AND J. R. SEILER. 1996. Growth, shoot phenology and physiology of diverse seed sources of black spruce: I. Seedling responses to varied atmo- spheric CO, concentrations and photoperiods. Tree Physiology 16:367—373. AND J. E. MAsor. 1998. Black spruce family growth performance under ambient and elevated at- mospheric CO,. Tree Physiology 15:271—281. KITZMILLER, J. H. 1976. Tree improvement master plan for the California region. U.S. Department of Agricul- ture, Forest Service, San Francisco, CA. Lona, S. P. AND B. G. DRAKE. 1992. Effect of long-term elevation of CO, concentration in the field on the quantum yield of photosynthesis of the C, sedge, Scirpus olneyi. Plant Physiology 96:221-—226. McNas, W. H. AND P. E. Avers. (eds.). 1994. Ecological subregions of the United States: Section descriptions. U.S. Department of Agriculture, Forest Service, Pub- lication WO-WSA-5. MILES, S. R. AND C. B. GOuDEY (eds.). 1997. Ecological subregions of California: Section and subsection de- scriptions. U.S. Department of Agriculture, Forest Service, Publication R5-EM-TP-5. Momen, B., P. D. ANDERSON, J. A. HELMS, AND J. L. J. Hovupis. 1997. Effects of increased rain acidity and ozone on gas exchange of field-grown seedlings and mature trees of Pinus ponderosa. International Journal of Plant Science 158:617—621. Moran, R. AND D. PoRATH. 1980. Chlorophyll determi- nation in intact tissues using N,N dimethylforma- mide. Plant Physiology 65:478—479. PuSHNIK, J. C., R. S. DEMAREE, J. L. J. Houpis, W. B. FiLory, S. M. BAUER, AND P. D. ANDERSON. 1995. The effect of elevated carbon dioxide on a Sierra-Nevadan | | | 2001] dominant species: Pinus ponderosa. Journal of Bio- geography 22:249-254. , G. B. REUNION, AND S. V. KRupA. 1994. Plant response to atmospheric CO, enrichment with em- phasis on the rhizosphere. Environmental Pollution 83:155-189. SAS INSTITUTE, INC. 1989. SAS/STAT User’s Guide, Ver- sion 6, Vol. 2, 4th ed. SAS Institute, Inc. Cary, NC. SuURANO, K. A., P. E DALEY, J. L. J. Houpis, J. H. SHINN, J. A. HeLMs, R. J. PALASSOU, AND M. P. COSTELLA. 1986. Growth and physiological responses of Pinus ponderosa Dougl. ex P. Laws. to long-term elevated CO, concentrations. Tree Physiology 2:243—259. AND J. R. KERCHER (eds.). 1993. Response of veg- etation to carbon dioxide: Effects of elevated CO, Concentrations on Pinus ponderosa. University of California, Lawrence Livermore National Laboratory, Livermore, CA. UCRL-ID-115638. pp. 72-87. _ToLiey, L. C. AND B. R. STRAIN. 1984. Effects of CO, ~Maprono, Vol. 48, No. 2, pp. 61, 2001 ANDERSON ET AL.: INTRASPECIFIC VARIABILITY IN RESPONSE TO CO, 61 enrichment and water stress on growth of Liquidam- ber styraciflua and Pinus taeda seedlings. Canadian Journal of Botany 62:2135—2139. WALKER, R. FE, D. R. GEISINGER, D. W. JOHNSON, AND J. T. BALL. 1995. Interactive effects of atmospheric CO, enrichment and soil N on growth and ectomycorrhizal colonization of ponderosa pine seedlings. Forest Sci- ence 41:491—S00. WEBER, W. E., G. WRICKE, AND T. WESTERMANN. 1996. Selection of genotypes and prediction of performance by analysing genotype by environement interactions. Pp. 354-371 in: M. S. Kang and H. G. Gauch, Jr. (eds.) Genotype by environment interaction, CRC Press, Boca Raton, FL. WELLBURN, A. R. AND H. K. LICHTENTHALER. 1983. De- terminations of total carotenoids and chlorophylls a and b of leaf extracts in different solvents. Pp. 415 in Abstracts of the 6th International Congress on Pho- tosynthesis, Brussels, Belgium. NOTEWORTHY COLLECTIONS ARIZONA ENCHYLAENA TOMENTOSA R. Br. (CHENOPODIACE- AE).—Maricopa Co., Phoenix, Arizona National Guard Papago Park Military Reservation, northern boundary with Oak Street, disturbed roadside between desert habitat and residential area, found in association with Salsola tra- gus, Isocoma acradenia, and Chamaesyce_ polycarpa, 33°27'30"N, 111°57'30"W, 354 m elev., 26 November 1999, G. Walters 266 (ASU, NSW). Determined by B. Wiecek (NSW). Previous knowledge. E. tomentosa was known from a single 1938 specimen from the Tucson Arizona Soil Con- servation Service Nursery, L. Gooding and L. Brinkerhoff 2809 (ARIZ) and this is the first time it has been collected outside of cultivation in Arizona; the second time ever collected in the state. It has been seen growing along the Central Arizona Project canal system and is currently sold at specialty nurseries in the Phoenix Metropolitan area. Significance. This plant is endemic to Australia where it is drought tolerant but does well in a variety of habitats. The fruits are animal dispersed and germinate prolifically. This combination of characters in a non-native shrub in the Sonoran Desert could potentially render this plant in- vasive. —GRETCHEN WALTERS, Herbarium, Department of Plant Biology, Box 871601, Arizona State University, Tempe, AZ 85287-1601 MADRONO, Vol. 48, No. 2, pp. 62-78, 2001 A MORPHOMETRIC ANALYSIS OF THE LEPTOSIPHON ANDROSACEUS COMPLEX (POLEMONIACEAE) IN THE CENTRAL AND SOUTH COAST RANGES ROBYN E. BATTAGLIA AND ROBERT PATTERSON Department of Biology, San Francisco State University, San Francisco, CA 94132 ABSTRACT The taxonomy of the Leptosiphon androsaceus Benth. complex has been troublesome because of re- markable morphological similarity among species. During the past 160 years, members of this complex have been classified in 4 different genera, and numerous specific and infraspecific names have been applied. Despite numerous treatments written by early taxonomists, analytical studies were not performed on these species until recently. We examined morphometric relationships among 1264 individuals from 51 populations, from the Central and South Coast Ranges. We focused on populations from San Francisco to Santa Barbara County because much of the variability in flower color occurs in this region, and color has been used by previous authors to distinguish species and subspecies. We investigated morphological variation using an array of multivariate analyses, including cluster analysis, principal components analysis, and discriminant analysis. Our analyses show six species of the L. androsaceus complex occur in this region of California: L. acicularis (Greene) Jeps., L. androsaceus, L. bicolor Nutt., L. croceus (Eastw.) J. M. Porter & L. A. Johnson, L. parviflorus Benth., and L. rosaceus. Leptosiphon croceus and L. rosaceus were described nearly 100 years ago, but have not been included in recent treatments. Our results offer strong support for recognition of L. croceus and L. rosaceus at the species level. The Leptosiphon (=Linanthus; see below) an- drosaceus group is a monophyletic lineage of Po- lemoniaceae (Bell and Patterson 2000) character- ized morphologically within the genus by sessile flowers borne in terminal, bracteate heads, salver- form corollas with long filiform tubes, and calyces with narrow intercostal hyaline membranes con- necting the lobes. Although visited and presumably pollinated primarily by long-tongued flies, the spe- cies within the L. androsaceus complex exhibit a variety of breeding systems (Goodwillie 1997, 1999a, b). Within the salverform-tubed leptosiphons, rela- tive breadth of calyx membranes and lobes distin- guishes two well marked groups (Table 1): the L. androsaceus group, characterized by membranes clearly narrower than the lobes, and the L. ciliatus group with membranes broader than the lobes. This distinction is also supported by molecular data (Bell and Patterson 2000). All are small, spring-blooming annuals occurring in grassland and woodland areas from the Sierra Nevada foothills to the Pacific Coast in western North America. The remarkable morphological similarity among these species has hampered resolution of species limits and relationships. Furthermore, the nomen- clature is extensive (Jepson 1943; Mason 1951; Munz 1959) and the task of sorting and assigning names is challenging. Hooker (1870) referred to Linanthus as ‘“‘one of the most variable genera of hardy annuals, the limits between the species of which are as difficult to draw from living speci- mens as from herbarium ones.”’ There are few mor- phological characters available to distinguish these similar species, and compounding the taxonomic confusion is the fact that these characters often ex- | hibit a high degree of variability within a species. ' Leptosiphon parviflorus, for example, is an es- > pecially variable species with regard to corolla tube. length and corolla color, two characteristics tradi- | tionally used to identify species. : Taxonomic background. In his monograph of Po- : lemoniaceae, Grant (1959) recognized 6 sections of | Linanthus based on several morphological features. | Among these was his sect. Leptosiphon. The ear- : liest recognized species in this section were de- | scribed originally by Bentham (1833) as members of the genus Leptosiphon. Greene (1889-1892) | combined several genera into a single genus, Lin- | anthus, based largely on the presence of opposite, | palmately lobed leaves. Grant’s (1959) sections | largely represent the genera that were combined into Linanthus by Greene. Porter and Johnson — (2000) presented a revision of the entire family | with the goal of recognizing only monophyletic groups. Their revision, supported by morphological and molecular data (Johnson et al. 1996; Porter 1996; Bell et al. 1999; Bell and Patterson 2000) divides Linanthus sensu Greene into two distinct, non-sister genera, Linanthus and Leptosiphon; the latter genus includes, but is not limited to, all of the L. androsaceus group. We follow Porter and Johnson’s taxonomy in this paper (Table 1). Although many treatments involving the L. an- drosaceus group have been provided by earlier tax- onomists (Bentham 1833, 1845, 1849; Endlicher 1836-1840; Nuttall 1848; Bentham and Hooker 1876; Gray 1870, 1886; Greene 1889-1892; Jepson 1901, 1925, 1943: Danforth 1945; Mason 1951), no 2001] | ‘TABLE 1. TAXONOMY OF LONG-TUBED LEPTOSIPHON. | \L. ciliatus group—calyx membranes wider than calyx lobes . breviculus (A. Gray) J. M. Porter & L. A. Johnson . ciliatus (Benth.) Jeps. . montanus (Greene) J. M. Porter & L. A. Johnson . nudatus (Greene) J. M. Porter & L. A. Johnson . oblanceolatus (Brand) J. M. Porter & L. A. Johnson Sal oll all oll LL. androsaceus group—calyx membranes narrower than | calyx lobes L. acicularis (Greene) Jeps. L. androsaceus Benth. L. bicolor Nutt. L. croceus (Eastw.) J. M. Porter & L. A. Johnson L. jepsonii (Schemske and Goodwillie) J. M. Porter & L. A. Johnson L. latisectus (E. G. Buxton) J. M. Porter & L. A. John- son L. minimus (H. Mason) R. Battaglia L. parviflorus Benth. L. rosaceus (Greene) R. Battaglia L. serrulatus (Greene) J. M. Porter & L. A. Johnson unidentified populations MRM _ Morgan Meadow, Santa Cruz Co. PIN Pinnacles National Monument, San Benito Co. analytical studies have been performed. The first explicit analyses include two recent morphometric studies that sampled populations in northern Cali- fornia where the distribution ranges of these taxa overlap (Buxton 1993; Schemske and Goodwille 1996). Each of these studies revealed the presence of previously unrecognized taxa and provided sta- tistical support for the recognition of the other members of this complex. Nevertheless, these stud- ies sampled only a fragment of the variation present in the entire complex. Only through quantitative data analyses can spe- cies limits within the complex be resolved. Deter- mination of species limits is a prerequisite for un- derstanding phylogenetic relationships among members of this group. Our study continues to clar- ify taxonomic relationships in the L. androsaceus complex by sampling from populations in the southern portions of its range. In order to determine which taxa occur in this geographic area, we used a number of multivariate analyses (PCA, Cluster, DA) to group specimens based on morphological similarities. We were specifically interested in de- termining whether any of the various L. parviflorus color morphs merited taxonomic recognition. Once the taxonomic groups were identified, we used PCA and DA analyses to identify the morphological characteristics most responsible for distinguishing the taxa. METHODS Sampling. Quantitative morphological data for this study were gathered from fresh specimens col- lected during spring of 1997 and 1998. We collect- BATTAGLIA AND PATTERSON: LEPTOSIPHON ANDROSACEUS COMPLEX 63 ed 1264 samples (individual plants) from 51 pop- ulations from San Francisco to Santa Barbara County (Table 2; Fig. 1). This is the region in which L. parviflorus, a particularly troublesome taxon, ex- hibits the greatest variation in color and color pat- tern. In addition, plants were grown in the greenhouse from field-collected seed to estimate whether ob- served character differences were influenced by en- vironmental conditions. Cotyledon measurements were also made on these plants. Characters. We obtained a range of leaf and flo- ral measurements and observations to represent the overall form of the plant (Table 3). A total of 44 measurements was taken on each specimen. Many of these characters have been used to distinguish among species in this complex. To ensure that mea- surements were comparable, specimens for mea- surements were prepared as follows: From each plant, one leaf and one flower were mounted on an overhead transparency using clear packing tape. The first leaf below the inflorescence was chosen to represent the leaves of the plant. The calyx and corolla were dissected prior to being examined. The calyx was cut between the lobes to flatten it and obtain a clear image. The corolla was dissected by peeling three or four of the limb lobes, including the throat, from the tube and mounting them inde- pendently. The stigmas and style were removed and mounted. Measurement. Qualitative characters, counts, and two length measurements were scored by hand pri- or to mounting specimens. Schemske and Good- willie (1996) showed patterns of calyx pubescence in this group fall into two categories; fewer than 100 or greater than 100 trichomes per lobe. We counted the number of trichomes per calyx lobe if they numbered fewer than 100. The remaining continuous characters were mea- sured by digitizing the contours of the mounted specimens using a computerized image capturing system. We used the software program MorphoSys ver. 1.26 (Meacham and Duncan 1989), which al- lows the contour of a specimen to be drawn, land- marks selected, and data saved. This allowed for a relatively rapid and accurate means of collecting the large amount of data necessary for a morpho- metric analysis. Width measurements for the corol- la and leaf lobes were spaced proportionally throughout the proximal region of the lobe, at 0.5, 0.7, 0.8, and 0.9 of the total length of the lobe, because this region appeared to be the most vari- able. Analytical techniques. We used several multivar- iate methods per Pimentel’s (1993) recommenda- tion. If the results of several different analyses agree, then violations of assumptions such as non- linearity and heteroscedacity are minimized and the results of the analyses are robust. The multivariate Statistical methods we used included: Cluster Anal- 64 MADRONO [Vol. 48) TABLE 2. COLLECTION LOCALITIES OF LEPTOSIPHON SAMPLES. All collections were made in 1998 except EDG, MPT, MTH, PIN, RSA, and WSK, which were made in 1997. See Table 4 for color code translation. Number sampled Color Acronym Location (N) code L. acicularis BFX Bolinas Fairfax Rd., Marin Co. 14 9 PLR Pleasanton Ridge Regional Park, Alameda Co. 25 9 SNB Sinbad Canyon, Pleasanton Ridge Regional Park, Alameda Co. 25 9 L. androsaceus AQS Almaden Quicksilver County Park, Santa Clara Co. 25 10 BFR Bolinas Fairfax Rd., Marin Co. 25 10 BNK Bunker Hill, Highway 280, San Mateo Co. 25 10 DUN East Dunne Rd., Santa Clara Co. Jas) 10 JPR Jasper Ridge Biological Preserve, San Mateo Co. 25 10 LMP Reynolds Rd., Stanton Ranch, Santa Clara Co. 25 10 MHM Eastern side of Mount Hamilton, Santa Clara Co. 2D 10 MIN Mines Rd., Alameda Co. 25 10 MTD Mount Diablo State Park, Contra Costa Co. 2S 10 MTH Mt. Hamilton, Santa Clara Co. 24 10 FF RSA Rancho San Antonio County Park, Santa Clara Co. NS) 10s RSN Rancho San Antonio County Park, Santa Clara Co. 25 10 UVA Uvas Rd., Santa Clara Co. Das, 10 | L. bicolor | ADL Paso Robles, San Luis Obispo Co. 25 11 | CHI Red Hill Rd., Chinese Camp, Tuolumne Co. 25 5 COL Coalinga Road, Monterey Co. 25 11 DPC Del Puerto Canyon Rd., Santa Clara Co. ap 3 PRB Parkfield, Monterey Co. 25 11 RDH Red Hill Rd., Chinese Camp, Tuolumne Co. 2D 11 STR Reynolds Rd., Stanton Ranch, Santa Clara Co. D9 11 VNY Vineyard Canyon Rd., Monterey Co. Das) 3 WSK Whiskey Falls, Madera Co. 2D 11 L. croceus MSB Moss Beach, San Mateo Co. 25 8 L. latisectus CUT Potter Valley, Mendocino Co. 25 Lt EEL Potter Valley, Mendocino Co. 25 11 P. parviflorus ALQ Almaden Quicksilver County Park, Santa Clara Co. 25 1 CAC Cachagua Rd., Monterey Co. 29 1 CLG Coalinga Rd., Monterey Co. 25 3 CRZ Highway 58, San Luis Obispo Co. 25 2 CVR Carmel Valley Road, Monterey Co. 25 2 DLP Del Puerto Canyon Rd., Santa Clara Co. 25 3 DNN East Dunne Rd., Santa Clara Co. 29 1 EDG Edgewood County Park, San Mateo Co. 24 1 EGW Edgewood County Park, San Mateo Co. 25 l EHT Carmel Valley Road, Monterey Co. 25 2 FGM Figueroa Mountain Rd., Santa Barbara Co. Ds) 2 FGS Happy Canyon Rd., Santa Barbara Co. 25 2 FIG Figueroa Mountain Rd., Santa Barbara Co. 25 2 FTO Impossible Canyon, Fort Ord, Monterey Co. 25 6 HST Hastings Natural History Reservation, Monterey Co. 25 4 JSP Jasper Ridge Biological Preserve, San Mateo Co. 25 1 JSR Jasper Ridge Biological Preserve, San Mateo Co. 25 5 LPD Los Padres National Forest, San Luis Obispo Co. je) 6 LCU Lucile’s Court, Boulder Creek, Santa Cruz Co. 25 8 PKS Parkfield, Monterey Co. 25 2 PNC Pinnacles National Monument, San Benito Co. 25 3 PRK Parkfield, Monterey Co. 25 6 QHL Quail Hollow Ranch County Park, Santa Cruz Co. 25 7 2001] BATTAGLIA AND PATTERSON: LEPTOSIPHON ANDROSACEUS COMPLEX 65 TABLE 2. CONTINUED Number sampled Color Acronym Location (N) code SAZ Near Sierra Azul County Park, Santa Clara Co. 25 | SND Sandhill Rd., Santa Cruz Co. 25 Ds | TRK Turkey Flat Campground, San Luis Obispo Co. 25 6 L. rosaceus | MPT Mori Point, San Mateo Co. 24 3 MRI Mori Point, San Mateo Co. 2a 3 "unidentified MRM Morgan Meadow, Santa Cruz Co. 29 10 PIN Pinnacles National Monument, San Benito Co. 9 2 PNN Pinnacles National Monument, San Benito Co. Os 2 ysis of a dissimilarity matrix (UPGMA), Principal Components Analysis (PCA) using a correlation matrix of the standardized data sets, and Discrimi- nant Analysis (DA) on a variance-covariance ma- trix. Several data matrices were used in our analyses. A summary of each data matrix with regard to iden- tity, number of groups, number of specimens, and number of variables is shown in Table 4. Matrices 1, 2, and 2a were used to determine how many taxa (species) occured in the study region. Matrix | con- tains all data collected from field specimens during spring and summer 1998, and was used for screen- AQS, ALQ SAZ, STR, LMP /{ 7 Fic. 1. study. Location of Leptosiphon samples included in this ing variables to be included in the final analyses. Matrix 2 is a subset of Matrix | resulting from data evaluation, and containing representative samples from all species in the L. androsaceus complex ex- cept for L. jepsonii, L. minimus and L. serrulatus (Greene) J. M. Porter & L. A. Johnson comb. nov. These species do not occur in the geographical range covered by this study. Matrices 3 and 4 were used to evaluate whether L. parviflorus color morphs merited infraspecific recognition. Matrix 3 is a subset of Matrix 1, con- taining representative samples from only popula- tions identified as L. parviflorus and the L. croceus population from Moss Beach (MSB). Leptosiphon croceus was included in this data set to assess its relationship to L. parviflorus. It is the final data set for L. parviflorus resulting from data evaluation. Matrix 4 contains only populations identified as L. parviflorus, and was used to determine whether any of the various color morphs deserved taxonomic recognition. Data evaluation. Multivariate techniques require a rigorous examination of data, because the effects of an ill-conditioned data set can be compounded across several variables and have quite substantial effects (Tabachnick and Fidell 1996). To minimize these risks, data from Matrix 1 were screened for accuracy, precision, missing data, and co-linearity. Several qualitative variables were excluded due to either difficulty in consistent interpretation of char- acter ranks or a high degree of variability observed within a single population. These include calyx tri- chome length, corolla tube pubescence and glan- dularity, corolla throat pubescence, leaf pubescence and glandularity, degree of branching, conivent ver- sus spreading stamens, number of bracts, number of internodes, and number of open flowers. Because corolla color and color pattern were consistent within populations, the variables for tube, abaxial lobe, adaxial lobe, and throat ring col- Or were summarized into one variable describing the 11 different color morphs (Table 5) observed in this study. These data were used only in the anal- 66 MADRONO [Vol. TABLE 3. MORPHOLOGICAL CHARACTERS ANALYZED. All distance measurements are in mm. Used in Qualita- Hand- multivariate tive MorphoSys — scored analysis Calyx: Length of lobes Total length of calyx Width of calyx lobe: at 0.5 length of lobe at base of lobe Location of calyx pubescence xX (absent/ciliate/throughout) Degree of calyx pubescence (# trichomes/lobe) Length of calyx trichomes xX (short/medium/long) Glandular/not glandular KK KK xx KM Matrix 2 x K~ K Corolla: Length of corolla lobe Width of corolla lobe: at base at 1/10 from lobe base at 5/10 from lobe base at 7/10 trom lobe base at 8/10 from lobe base at 9/10 from lobe base Length of throat Length of tube Width of tube Tube color Tube pubescence (absent/sparse/dense) Tube trichomes glandular/not glandular Throat trichomes present/absent Lobe color: Abaxial surface Adaxial surface Throat ring color KKK KKM MOK x xxx x KKM KKK MK KKK RK K OK Androecium and gynoecium: Length of filament x Length of stigma Length of style Leaf: Length of palm Width of palm Length of middle lobe Width of middle lobe: at base of lobe at % from lobe base at %4 from lobe base at % from lobe base Number of lobes per leaf Pubescence (absent/ciliate/throughout) Glandularity (present/absent) Matrix 2 xx x ve) & aN Matri xxx KM xx x XK x x x Kx General: Number of internodes on longest stem Number of open flowers per inflorescence Number of bracts subtending inflorescence Branching (none/above/below/throughout) Stamens connivent/spreading Total height of plant xx KKK KKK 2001] i BATTAGLIA AND PATTERSON: LEPTOSIPHON ANDROSACEUS COMPLEX 67 TABLE 4. DATA MATRICES USED IN ANALYSES. The number of groups refers to the number of groups used for multi- ' variate analyses. Matrices 5—8 were not used in the multivariate analyses. Popula- tions © Number Number | Data Groups in repre- of speci- of vari- matrix Description analysis sented mens ables | Matrix 1 All 1998 field data 313) 53 1314 44 | Matrix 2 Results of data analysis, all species included 51 (DA2-1) 51 1264 P21 | 9 (DA2-2) | Matrix 2a Variable means for each population in Matrix 2 ] 51] il 2h _ Matrix 3 Results of data analysis, only L. croceus and L. par- 26 26 650 20 | viflorus populations included Matrix 3a Variable means for each population in Matrix 3 ] 26 26 20 Matrix 4 Results of data analysis, only L. parviflorus popula- 25 (DA4-1) DS 625 20 tions included 7 (DA4-2) Matrix 5 All field data from 1997 N/A 6 118 2 Matrix 6 1997 Greenhouse data N/A 3 Te 26 Matrix 7 1998 Greenhouse data N/A Pe 276 26 Matrix 8 Cotyledon data from 1998 greenhouse plants N/A 19 447 py) ysis involving populations of L. parviflorus. Most species in the L. androsaceus complex have two color morphs: white, and either pink, lavender, or yellow. Leptosiphon parviflorus is unique in that all the above color morphs occur. In addition, throat color varies, and markings on the limb lobes may be present (Fig. 2). Eight different color morphs of L. parviflorus were observed in this study. Measurements taken using MorphoSys were first checked for precision (repeatability) by randomly choosing one population and remeasuring for each variable. Measurements that showed significant dif- ferences (P > 0.05) between measurement sessions were eliminated from our analysis. The remaining 27 variables (23 metric continu- ous, two metric counts, one multistate, and one bi- nary) were then checked for near-perfect correla- tion to reduce the risk of co-linearity. Multi-co-lin- earity problems occur when r > 0.9 (Tabachnick and Fidell 1996); therefore we eliminated selected variables from pairs with Pearson’s correlation scores of r > 0.9. Calyx pubescence and glandularity posed prob- lems with data scoring. Degree of calyx pubescence could be measured as number of trichomes per lobe, but the trichome location and glandularity had to be coded (multistate and binary respectively). However, all three characters were highly correlat- ed (all r values > 0.92), thus the two coded vari- ables were omitted. Style exsertion has been used in the past (Ben- tham 1833; Greene 1889—1892; Mason 1951; Munz 1959; Buxton 1993) to aid in characterizing mem- bers of this complex. We observed that style length, and therefore style exsertion, increased with age of the flower in L. parviflorus. In addition, tube length and style length were highly correlated, r = 0.944. Despite its previous taxonomic use, we excluded style length from our analysis. Statistics. Cluster analysis was used to suggest similarity among populations. Results from the cluster analyses allowed us to define groups in later analyses that require group identity. Population TABLE 5. COROLLA COLOR PATTERNS RECORDED IN THE LEPTOSIPHON ANDROSACEUS COMPLEX. Color code Lobe color Throat color 1 white yellow 2 white and/or lavender-pink — violet or yellow 3 white yellow 4 white violet 5) pink yellow 6 white orange 7 deep yellow-orange orange 8 deep yellow-orange orange 9 light buttery yellow light yellow 10 white or lavender-blue distally _ — pink ring distally Additional markings 2 red spots at base of lobes (“‘two-spot’’) striations on lobes (“‘candystripe”’ ) 2 red spots at base of lobes (“‘two-spot”’) 1 red bar at base of lobes (“‘bullseye’’) iw) red spots at base of lobes (“‘two-spot’’) violet at base, yellow yellow with white 68 MADRONO [Vol. 48 BiG: 2. at base of lobe, b) ‘‘2-spot’’, two red spots at base of lobe, c) “‘candystripe’’, lavender striations on lobe. means (group centroids) for each variable in Matrix 2 and Matrix 3 were calculated. New data matrices (Matrix 2a and Matrix 3a) were created using vari- able means for each population, and each was sub- jected to cluster analysis. The group average meth- od for linkage (UPGMA) using a dissimilarity ma- trix generated by Euclidean distances was used (Sneath and Sokal 1973). Principal components analysis (PCA) was pri- marily used to analyze the variables on a correla- tion matrix of the data sets Matrix 2 and Matrix 3. Data were standardized prior to the analysis as part of the SPSS (SPSS 1997) protocol for the PCA method. Discriminant analysis (DA) was performed on a variance-covariance matrix of the Matrix 2 and Ma- trix 4 data sets. Discriminant analyses were run on the two data sets using several a priori grouping arrangements. Three analyses (DA2-1, DA2-2, DA2-3) were run on Matrix 2. DA2-1 used the 51 collection populations as predefined groups. DA2- 2 used the results from the UPGMA and PCA anal- yses to assign individuals to the following groups: PNN, MRM, L. androsaceus, L. acicularis, L. bi- color, L. latisectus, L. parviflorus, L. croceus, and L. rosaceus. Because the sizes of the groups in DA2-2 were not equal (ranging from 25 to 625 in- dividuals per group), the analysis was repeated us- ing equal group sizes. DA2-3 used a subset of 25 individuals selected randomly from each group used in DA2-2. The Matrix 4 data set was subjected to two DA analyses, DA4-1 and DA4-2. A priori groups were defined by color morphology and based on the results from the PCA and cluster anal- yses: “*2-spot” (all corolla colors), “‘bullseye,”’ ““candystripe,”’ “‘yellow,”’ “‘white with violet,” and ‘‘white with yellow.”’ The sizes of the groups based on color morphology were not equal, ranging from 25 to 200 individuals. The same reduction proce- dure was performed by randomly selecting 25 in- dividuals from each group, and the analysis was run again (DA4-2). Examples of the various corolla markings observed in L. parviflorus: a) “‘bullseye’’, red crescent shaped bar RESULTS Because of their large size, matrices generated (correlation, component, structure) and Geisser classification summaries are not included in this pa- per, but are in Battaglia (1999) or are available from the first author upon request. Results of analyses using all species are discussed first, results from the L. parviflorus color morph analyses are discussed second. Four of the 51 populations (MRI, MSB, PNN, MRM) were not identifiable using the current taxonomy (Patterson 1993; Buxton 1994; Schem- ske and Goodwillie 1996). Two of these were later identified as L. croceus (MSB) and L. rosaceus (MRI), species synonomized with L. parviflorus and L. androsaceus respectively. The PNN and MRM populations remained unidentifiable. Cluster analysis—all species. The UPGMA clus- ter analysis of Matrix 2a (Fig. 3) is in general ac- cord with the conventional taxonomy of the group, with several noteworthy exceptions. Five of seven species cluster together; however, L. acicularis and L. parviflorus do not. Principal components analysis—all species. Re- sults of PCA on Matrix 2 showed the total variance was generally well spread among variables, with only 77% of the total variance explained by the first 6 components. This indicates variables were gen- erally independent of each other, with little corre- lation or covariation. Graphical representation of regression factor scores for each individual (not shown) indicate seven distinct clusters representing L. acicularis, L. androsaceus, L. rosaceus, L. cro- ceus, L. parviflorus, L. latisectus, and L. bicolor. Results from the component matrix indicate there was no single variable contributing to the observed variation among groups. More interestingly, corolla tube length had component scores of 0.232, —0.008, and 0.203 for the first three components. This indicates corolla tube length explained little of the observed variation. Although corolla tube length is traditionally used as a character to distin- 2001] androsaceus (11) bicolor (6) rosaceus (1) MRM (1) PNN (1) (1) acicularis (2) latisectus (2) parviflorus (25) croceus (1) Fic. 3. Dendrogram of UPGMA cluster analysis for all species (Matrix 2a). MRM (Morgan Meadow) and PNN (Pinnacles) are two unidentifiable populations. Numbers in parentheses refer to the number of populations repre- sented by the branch. guish among species in this complex, it does not correlate with segregation on the first three com- ponents. Discriminant function analysis—all species. (DA2-1). When using a priori groups based on the 51 collection sites, all Wilks’ A values were small (0.040—0.575), indicating strong differences among group centroids for each variable. ANOVA’s show significant differences (P < 0.001) among all vari- able means for each of the 51 populations. Results of the Geisser classification summary for the 51 collection sites indicate three populations have 100% classification success: MSB (L. cro- ceus), CHI (L. bicolor), and LUC (L. parviflorus). Ten populations have greater than 90% classifica- tion success: AQS, RSN, UVA (L. androsaceus), COL, STR (L. bicolor), CLG, JSR, QHL (L. par- viflorus), MRI (L. rosaceus), and PNN (undeter- mined). One group, TRK (ZL. parviflorus), has a low score of 54%, and the remaining groups range from 64-88%. Of the 1264 individuals, 1053 (83.3%) were Classified correctly based on collection site, indicating that all groups are unique. Misallocations to collection sites in geographical proximity occurred frequently. In nearly every case, misallocations were within species groups that re- sulted from the UPGMA and PCA analyses. The exceptions involve three of the four unusual pop- ulations identified in the previous analyses. MRI (L. rosaceus) had 24 correct hits and | misallocation to BFR (L. androsaceus); PNN had 24 correct hits and | misallocation to ADL (L. bicolor); and MRM BATTAGLIA AND PATTERSON: LEPTOSIPHON ANDROSACEUS COMPLEX 69 had 16 correct hits, 4 incorrect classifications to JSR and CAC (L. parviflorus) and 5 misallocations to RSN (L. androsaceus). The unusual population from MRM was the only one to have members classified into two other species groups. Results from DA2-3 (a priori groups of equal numbers defined by UPGMA and PCA, respective- ly) appear below. We do not discuss in detail results from DA2-2 because general patterns regarding group discrimination were the same, and graphical interpretation in two dimensions is complicated with the larger sample size. In each case one-way ANOVAS indicated variable means for all groups differed significantly (P < 0.001). The first three canonical functions contained 74.4% of the variation in DA2-3, and the variance was evenly distributed among the functions. The first function in DA2-3 contained 33.3% of the variation. Strong group differences for all variables in DA2-3 was indicated by generally low Wilks’ A scores (0.168—0.674). The majority of variables scored lower than 0.5. Patterns are revealed by graphing the first two canonical discriminant functions (Fig. 4). Group separation is distinct: L. croceus and L. rosaceus are clearly separated from L. parviflorus and L. an- drosaceus respectively, L. bicolor and PNN are not separated, and MRM is closely associated with L. androsaceus, although individuals are still scattered between L. parviflorus and L. androsaceus. A three-dimensional depiction (Fig. 5) of the first three discriminant function scores for individuals in DA2-3 reveals 7 distinct clusters, each correspond- ing to one of the 7 species. Individuals from the PNN and MRM populations do not form coherent groups. Geisser classification results for each analysis are similar. In DA2-2 all but MRM (76%) have above 90% successful predicted group membership. In DA2-3, all but L. androsaceus (94%) have 100% successful classification. Cluster analysis—L. parviflorus color morphs. We performed a UPGMA cluster analysis on the Matrix 3a data set to evaluate support for grouping L. parviflorus populations based on corolla color pattern (Fig. 6). Leptosiphon croceus (MSB) 1s clearly separated from L. parviflorus. Within L. parviflorus, all but one of these populations dis- playing two spots at the base of the petal lobe clus- ter together. Three populations with the “‘bullseye”’ pattern form a cluster, but this cluster and the ex- cluded population LPD are nested deeply within the remaining populations of L. parviflorus. No other color morphs form discrete clusters. Discriminant analysis—L. parviflorus color morphs. We performed a DA on the Matrix 3a data set. Based on the results of the previous analyses, all of the ‘*2-spot’”? color morphs (white or pink) were grouped together for this analysis. The results from DA4-1 and DA4-2 were nearly identical, in- 70 MADRONO Function 2 [Vol. 48 PNN MRM croceus rosaceus parviflorus latisectus bicolor androsaceus acicularis Function | Fic. 4. Discriminant analysis results from all species (Matrix 2), equal group sizes (DA2-3). Graph of first two canonical functions for each individual. dicating the unequal group sizes within DA4-1 did not adversely affect the analysis. Only the results from DA4-1 (all L. parviflorus individuals) will be discussed. Wilks’ A values for the variables were mostly greater than 0.8, demonstrating weak differences among group centroids for each variable; however, ANOVAS revealed that variable means for all groups were significantly different (P < 0.001). Tube length was the only variable for which there Function 1 Fic. 5. was a strong group difference (Wilks’ A = 0.385). The three petal lobe measurements (tube width, plant height, and leaf mid-lobe width) had Wilks’ A scores ranging from 0.518 to 0.673. Geisser classification showed that “‘candystripe”’ morphs had the lowest predicted group membership scores, with 71% being classified correctly. Misal- locations for this color morph were made to each other color group, with the greatest number (25) being classified into the “‘white with violet”’ group. + acicularis Xx latisectus @ androsaceus @ parviflorus © bicolor V = rosaceus A croceus O MRM O PNN Discriminant analysis results from all species (Matrix 2), equal group sizes (DA2-3). a) Three dimensional graph of the first three discriminant scores for each individual. b) Same graph with axes rotated. 2001] white/2 spot (ALQ) white/2 spot (SAZ) white/2 spot (EGW) white/2 spot (DNN) white/2 spot (JSP) pink/2 spot (JSR) yellow/2 spot (LUC) candystripe (SND) candystripe (FGM) white/2 spot (CAC) bullseye (FTO) bullseye (PRK) bullseye (TRK) white w/ violet (HST) candystripe (EHT) candystripe (PKS) bullseye (LPD) candystripe (FIG) candystripe (CVR) candystripe (FGS) candystripe (CRZ) white w/ yellow (CLG) white w/ yellow (PNC) white w/ yellow (DLP) yellow (QHL) Fic. 6. Dendrogram of UPGMA cluster analysis for L. parviflorus color morphs (Matrix 3a). Acronyms refer to population locations. The ‘“‘white with violet’? flowers were allocated correctly 92% of the time; however these speci- mens came from one population, and the two mis- classified individuals were allocated to the ‘“‘can- dystripe’’ group. The *‘2-spot’”’ morphs had 87.5% correct classi- fication, which was lower than expected based on results from PCA but consistent with prior PCA and cluster analyses. Of 200 specimens, 175 were classified correctly and 25 were misallocated to each of the other color morphs. Case-wise exami- nation of these 25 misallocations showed 22 were from the CAC population, the same population that did not cluster with other ‘‘2-spot’? populations in the cluster analysis. Taking this into consideration, the remaining individuals of the ‘‘2-spot’? morph show 98% correct classification. The “white with yellow” and the ‘‘yellow”’ morphs had high classification rates, 97.3% and 100% respectively. The two misallocations of the “white with yellow’ morph were to the ‘‘yellow group,” a single population. The ‘‘white with yel- low” morph shows general affinity for the ‘‘yel- low” morph based on misallocations, but with BATTAGLIA AND PATTERSON: LEPTOSIPHON ANDROSACEUS COMPLEX 71 ° 100% classification of “yellow,” affinities of “‘yel- low” with other morphs cannot be assessed. The correctly predicted group membership for ‘“‘bullseye’’> was 86%, but no affinities to other groups could be determined because misclassified individuals were allocated to each of the other color morphs. In addition, misallocations did not share similar patterns observed in the PCA and cluster analyses. In the UPGMA tree, FTO, TRK, and PRK clustered together, while LPD was excluded. Case- wise analysis of the misallocations revealed nine of 14 misallocations were from the FTO population, rather than the LPD population. Results from dis- criminant analysis are consistent with geographical distributions of these populations: FTO occurs near the coast while PRK, TRK and LPD are closer to each other in the inner Coast Ranges. We extracted five canonical discriminant func- tions in our analysis; 93% of the total variation was explained in the first three axes (55.7%, 24.5%, and 12.8%, respectively). Corolla tube length and width scored high on the first axis (0.677 and 0.494). Co- rolla tube length is not particularly informative in distinguishing species in the Matrix 2 analyses, but it iS important in distinguishing the ‘‘2-spot”’ L. parviflorus morphs from the remaining color morphs. Width of corolla lobe at the tip contributes most to the second axis, separating the narrower lobed ‘“‘bullseye”’” morphs from the more rounded ‘“‘white with yellow’? morphs (Fig. 7). Group differences based on individual variables. Results from the PCA and DA demonstrated that most variables were necessary for distinguishing among groups, whether they were groups based on species or color morphs of L. parviflorus. Discrim- ination among species was primarily based on ca- lyx pubescence, corolla lobe length, corolla lobe width, and leaf lobe width. When only L. parviflo- rus populations were examined, PCA indicated most all variables were necessary for discrimina- tion. On the other hand, DA showed corolla tube length and width, along with corolla lobe tip width, to be the most important variables in distinguishing the ‘‘2-spot’’? morphs from the remaining color morphs of L. parviflorus. Greenhouse data. Nearly all measurements (means for each variable) from the plants grown in the greenhouse were larger than those collected from the field populations, but all variables fell within the range of measurements observed in the field populations. This indicates that there is a ge- netic basis for the observed differences, and envi- ronmental conditions do not greatly effect the vari- ables used to distinguish among the various taxa. The potential for cotyledon characters in helping to distinguish species of Leptosiphon should be of interest to students of the genus. We noticed im- portant patterns in our analysis of cotyledon width and length measurements. There were two basic cotyledon morphologies: long and linear, versus 72 MADRONO Function 2 Fic. 7. canonical functions for each individual. shorter, rounded, and more or less obovate. The cot- yledons of L. acicularis are linear and extremely long, with a length-to-width ratio of 7.1. Leptosi- phon parviflorus also has linear cotyledons, with a ratio of 5.1; however two populations of L. parvi- florus (LUC and JSP) had rounded cotyledons. These are each *‘‘2-spot’’? populations. The remain- ing species surveyed had oval to obovate cotyle- dons. Leptosiphon androsaceus had a _ length-to- width ratio of 1.8; L. bicolor, 1.9; L. croceus, 1.5, and L. rosaceus, 1.4. The length-to-width ratio of L. latisectus was 1.6. The unidentifiable population, MRM, had coty- ledons consistent with L. androsaceus. They were short and rounded, with a length-to-width ratio of 1.6. The cotyledons of the other unusual popula- tion, PNN, were longer and narrower than those of L. bicolor to which it was affiliated, with a ratio of 2.5, but they were not as linear as the cotyledons of L. parviflorus. DISCUSSION In closely related species that show far more sim- ilarities than differences, it is important to examine carefully all characters and to seek disjunctions among character states that may help define taxa and elucidate their relationships. Because the L. an- drosaceus group has a long taxonomic history ac- companied by an abundance of nomenclatural ac- tivity, it is critical that we begin our analysis by examining morphological characters and identify- ing which of those best reflect relationships in the group. Despite the taxonomic age of the group, it has remained difficult to delineate species within it. Identification of taxa and the morphological char- acters that delimit them is a necessary precursor to yellow yx bullseye white w/violet white w/yellow candy stripe 2-spot 2 4 6 Function 1 Discriminant analysis from L. parviflorus color analysis (Matrix 3), all specimens (DA4-1). Graph of first two further research on evolutionary relationships among these species. We focused on morphological characters because they are the most practical means of identification and without proper identi- fication, further systematic research is compro- mised. Our results should be used as a working hypothesis of the taxonomic structure within the group, providing a framework for future research on evolutionary relationships among its species. Results from our analysis support recognition of six species in central California (L. acicularis, L. androsaceus, L. bicolor, L. parviflorus, L. croceus, and L. rosaceus) bringing the number of species within the L. androsaceus complex to nine (includ- ing L. latisectus and L. jepsonii from northern Cal- ifornia, and L. serrulatus from the southern Sierra Nevada). Leptosiphon croceus and L. rosaceus were described nearly 100 years ago, but were syn- onomized with L. parviflorus and L. androsaceus by later authors (Milliken 1904; Jepson 1925). Nei- ther author offered any explanation for their action; however, the omission of these species may be ex- plained by several factors, including the morpho- logical similarity among all members of this com- plex, the large amount of variation observed in the few characters used for distinguishing them, and the relative rarity of both species. Each is known from only a few populations collected from coastal bluffs in the San Francisco Bay region. These bluffs have undergone severe disturbance from increased developmental activities in the last hundred years, and it is likely that of the populations of these spe- cies that were known, few remain. For example, L. croceus was originally described from a population near Pt. San Pedro, in San Mateo Co. In her de- scription of the species, Eastwood (1904) observed [Vol. 48 | 2001 | that ‘‘it covered the ground for several acres . . . the | great masses almost monopolized the ground.”’ De- | spite its historical presence, L. croceus no longer - occurs in the Pt. San Pedro area. Today only one _ population of L. croceus is known. In addition to the 6 species mentioned above, this study also identified 2 unusual populations. One _ relatively invariable population, from Pinnacles Na- tional Monument (PNN), consists of plants that most closely resemble L. bicolor, but its limb and | stigmas are somewhat larger and the calyx shorter | than typical L. bicolor. The coloration is also un- _ usual, being light lavender (darker on the margins _ of the lobes fading to white near the throat) instead _ of the typical white or pink. This population is par- _ ticularly interesting because L. bicolor is generally the least morphologically variable species in the L. androsaceus complex. In addition to the morpho- logical similarity of PNN to L. bicolor evidenced by the multivariate analyses, the PNN plants grown in the greenhouse readily set seed, indicating it is autogamous. The only other plants to set seed in the greenhouse were from populations of L. bicol- or. The other unusual populations is the highly vari- able population from the Santa Cruz Mountains (MRM), a mosaic of L. parviflorus and L. andros- aceus. Limb size was smaller than L. androsaceus and more like that of L. parviflorus. Flower color and limb lobe shape were like the white form of L. androsaceus (white lobes with a throat that is violet at the base and yellow above). Calyx pubescence is a stable character, yet the calyces from this pop- ulation ranged from completely glabrous, to ciliate, to densely pubescent throughout, although never glandular. The within-population variance for most characters was high. Morphological relationships among the L. an- drosaceus complex in the Central and South Coast Ranges. Leptosiphon acicularis. This species is most clearly defined from others in this complex by the long, narrow, nearly needle-like leaf lobes. The size of the leaf palm is also the smallest in the complex, thus the leaves appear to be very finely dissected. The length of the calyx is long (7—9 mm), with the calyx lobes being narrow and much longer than the fused portion. Leptosiphon acicularis most closely resembles L. bicolor, both of which are the small- est-flowered members in this group. The limb of L. acicularis is always yellow, but the short corolla tube may be yellow or a light tannish pink (i.e., not always yellow as cited by Patterson [1993]). The shape of the tube is reminiscent of L. parviflorus, being very thin (0.5 mm), but in contrast to L. par- viflorus, which may have extremely long tubes, the tube of L. acicularis is the shortest in the complex (11-17 mm). Although the limb of L. bicolor may also be light yellow, it is clearly distinguished from L. acicularis by a suite of other characters (calyx BATTAGLIA AND PATTERSON: LEPTOSIPHON ANDROSACEUS COMPLEX 73 pubescence, shape of the leaf lobes, stigma length, stamen length). Aside from its unique leaf morphology, L. aci- cularis also has a densely glandular pubescent ca- lyx, whereas L. bicolor is pubescent only on the lobe margins and the trichomes are nonglandular. The size of the stamens and stigmas are also very different between the two species. Leptosiphon aci- cularis has some of the longer filaments in the com- plex (only those of L. croceus are as long), yet it has the one of the smallest limbs. Thus, the stamens of L. acicularis are well exserted, reaching *% the length of the petal lobe. The stigmas are also large, being 2-4 mm long. The stamens of L. bicolor barely exceed the throat, and the stigmas are gen- erally less than | mm long. As with most other species in this complex, the corollas open and close daily until senescence. The corollas of both species close for the night by mid-afternoon, earlier than the other species in the complex. Leptosiphon androsaceus. This is the largest- flowered member of the L. androsaceus complex, and its floral characteristics most closely resemble L. rosaceus and L. latisectus. The limb lobes are the longest (8—11 mm) and differ from L. rosaceus and L. latisectus in being more oblong to oval, of- ten with an apiculation at the tip. In comparison, the lobes of L. rosaceus and L. latisectus are very rounded. The limb is typically white or pale lav- ender, and the throat is commonly violet at the base, turning to yellow just as the throat flares into to limb. Buxton (1993) found populations in northern California with pink limbs, but none of the popu- lations in this survey of central California had limbs with this color. The stigmas are generally long (2—4 mm) and the filaments are short in rela- tion to the size of the limb. The corolla tube is typically moderate in length (19—26 mm), although this character is rather variable, and populations with longer tubes were observed. Similar to L. ro- saceus and L. latisectus, the width of the tube is relatively wide (1 mm). Other characters differen- tiating this species include, leaf shape, plant height, calyx size, and most importantly calyx pubescence. The calyx is non-glandular and sparsely pubescent, with trichomes only on the margins of the lobes. As with the other moderate- to short-tubed mem- bers, the total length of the calyx is moderately long (4-6 mm), with long calyx lobes relative to the fused portion. Nearly all leaf measurements for L. androsaceus are large for the complex. This species has the largest palm and longest middle lobe, al- though the width of the lobes is less than other species, giving it a less rounded appearance. The plants are typically the tallest growing members of the complex, and can occasionally be found grow- ing on serpentine soil. Grant (Grant and Grant 1965) observed a cyrtid fly, Eulonchus smaragdinus, visiting flowers of L. androsaceus. Although as a general rule he dis- 74 counted beetles as potential pollinators, we often observed beetles visiting flowers, probably not so much seeking out nectar as consuming pollen. Grant (Grant and Grant 1965) proposed L. andros- aceus to be self-compatible and partially autoga- mous, with protandry being incomplete (overlap- ping stages). However, Goodwillie (1999b) showed it to be a self-incompatible, obligate outcrosser. Grant (Grant and Grant 1965) also suggested that L. androsaceus and L. parviflorus may hybridize locally, although he did not offer any evidence to support his statements. Hybridization is often in- voked to explain unusual forms, but without careful study this explanation remains conjectural. Leptosiphon bicolor. This species 1s morpholog- ically the least variable member of the L. andros- aceus complex. Its limb is small (3—4 mm), and its corolla tube is moderate in length (17—26 mm) and width (0.8 mm). It has by far the longest tube rel- ative to the size of the limb. The limb is typically either pink or white with a yellow throat. Its reproductive structures are small, with the stigmas | mm and stamens only one-half the length of the limb lobe, as might be expected in an autog- amous species. The plants are relatively small (5— 13 cm), and rarely is there more than one open flower per inflorescence. The flowers close by mid afternoon, opening again the following day. The calyx is relatively long (7-9 mm), especially the length of the calyx lobes compared to the fused portion. The calyx is also ciliate and non-glandular, but, in contrast to L. androsaceus and L. rosaceus, the density of trichomes per lobe is generally great- er in L. bicolor (30-50 trichomes per lobe). The leaves of L. bicolor are small, with short lobes and large palms. Buxton (1993) reported L. bicolor to have the greatest number of leaf lobes in the com- plex. In the southern populations, we found the variance of this character to be high both within and among populations, and we found no signifi- cant differences among the number of lobes in L. bicolor, L. androsaceus, L. acicularis, L. parviflo- rus, or L. croceus. Goodwillie and Stiller (2001) recently elevated Linanthus bicolor (Nutt.) Greene subsp. minimus to species rank. While the scope of our study does not involve this species, following the taxonomy of Porter and Johnson (2000) the following combina- tion is made: Leptosiphon minimus (H. Mason) R. Battaglia, comb. nov. Linanthus bicolor var. minimus H. Mason. Madrono 9:249—255, 1948. Linanthus minimus Goodwillie and Stiller, Systematic Bot- any (2001). In press. Leptosiphon croceus. First described by East- wood (1904), L. croceus was later synonomized as varieties of L. parviflorus (Milliken 1904) and L. androsaceus (Jepson 1925). Our analysis supports its recognition as a distinct species. As with L. la- MADRONO [Vol. 48 | | tisectus, L. croceus shares morphological characters | with both L. androsaceus and L. parviflorus and | many of the characters are intermediate between > the two (e.g., limb size). Like L. latisectus, its leaf | lobes are characteristically rounded at the tip, al- | though its leaves are generally smaller. In addition, | its leaves are thick and somewhat succulent. It is extremely low growing, being the shortest of all the | species (2-6 cm). Although plants grown from seed | of this population were slightly larger (6-8 cm) | when raised in the greenhouse, they remained sig- | nificantly shorter than any other species. Likewise, | greenhouse grown plants also remained somewhat | succulent, a likely response to conditions experi- — enced directly on coastal bluffs. Leptosiphon cro- | ceus is often branched at the base with each branch | having many closely spaced internodes. The close | spacing of internodes makes the leaves appear ‘‘as — if whorled’”’ (Eastwood 1904). Its calyx is similar — to that of L. latisectus in that the lobes and fused | portion are nearly equal in length, and that it is | densely glandular pubescent. The distinction lies in | the size of the calyx and the width of the lobes. Leptosiphon croceus has a much larger calyx (7— | 9 mm) than L. latisectus, and the width measure- | ments are one-half to two times that of any other © species. The limb is also similar in shape to that of © L. latisectus, although the lobes are slightly larger | (6—8 mm) and more rounded. The width of corolla lobes, both at middle and at the tip, are the largest » in the complex. The corolla tube is also very long (29-37 mm), and thus distinguishes it from L. la- tisectus. The tube is generally much wider (0.9 — mm) than that of L. parviflorus, more closely re- | sembling L. androsaceus or L. latisectus. The limb is a bright, vibrant yellow, with an orange throat, and commonly has two red spots at the base of the lobes. The tube is yellow to yellowish-pink. In con- trast to L. latisectus, the stigmas are relatively large (2—4 mm) and the filaments are long with the sta- mens exserted. The length of the filaments is sim- ilar to that of L. acicularis. Its cotyledons are rounded, like all the other species except L. aci- cularis and L. parviflorus. Leptosiphon latisectus. This species has features of L. androsaceus and L. parviflorus, and closely resembles L. croceus in some features. As with L. acicularis, leaf measurements are important distin- guishing characteristics. There are few leaf lobes, and they are wide at the tip, appearing more or less spatulate. Buxton (1993) showed palm lengths to be large, but variation in the two populations we sampled was too great to make conclusions. Its ca- lyx is the smallest of any species in this group (5— 7 mm). Its lobes are also small, nearly equal to the length of the fused portion. Its calyx is also densely glandular pubescent. Its limb shape and corolla tube width are similar to that of L. croceus. The limb lobes are moderate in size (5-7 mm), between L. androsaceus and L. parviflorus, but as with L. cro- ceus they are especially wide at the tip, the lobes Wi | | | | 5 | ' being rounded to obovate. Unlike L. croceus, the tube is moderate in length (19-24 mm), similar to -L. androsaceus and L. bicolor. The tube is also wide (= 1 mm) as is seen in L. androsaceus and _L. croceus. A smaller limb with rounded lobes, a _densely glandular calyx, and spatulate leaf lobes distinguish L. latisectus from L. androsaceus, while -a comparatively larger limb, wider tube, and spat- ‘ulate leaf lobes distinguish it from L. parviflorus. | Corolla lobes are either dark pink or white, and the | throat is yellow. The two pink populations sampled also had a white ring present at the top of the throat. _ Stigma length (1-2 mm) was among the smallest of any species (only L. bicolor had smaller stig- _mas), and filament length was short, barely exceed- ing the throat. Buxton (1993) suggested this might indicate possible autogamy, however Goodwillie (1999b) states L. latisectus is an obligate outcrosser. Individuals grown in the greenhouse had flowers that remained open even at night and in cooler tem- peratures, unlike any other species in the complex. Leptosiphon rosaceus. This species was first rec- ognized as Leptosiphon parviflorus var. rosaceus (Hooker 1870). Hooker considered this taxon a va- riety rather than a new species because he could find “‘no other difference”’ than the color and size of the flower (more than some authors would use). As to the flower, he stated it ““was of a pale deep rose color, with a white or yellow eye”’ and that it “agrees with L. androsaceus one of the largest flowered of all,’ but that the lobes of the corolla had ‘‘a very different shape,’’ being orbicular, ver- sus narrower in L. androsaceus (Hooker 1870). Greene (1889-1892) elevated Hooker’s Leptosi- phon parviflorus var. rosaceus to species level, Lin- anthus rosaceus. With regard to L. rosaceus, Greene observed that it was the “‘most beautiful plant of the Leptosiphon group” having stoutish short internodes, decumbent branches, obovate- spatulate leaf segments, the flower with a rose-red limb and an ample orange throat. He stated the ‘*specific characters are as good as are found in this subgenus.’’ He noted there was an albino form of this species as well. The population we observed during the course of this study was white. Another population discovered after the completion of the study was pink. Each of these observed populations had a yellow throat, not orange, as Greene de- scribed. Occurring on coastal bluffs, its height and habit is similar to L. croceus, being densely branched and low growing (6-15 cm). It is significantly shorter than L. androsaceus, which is generally tall (17-31 cm) and not as densely branched. The leaf lobes of L. rosaceus are spatulate and more or less succu- lent. They are larger than those of L. croceus and even more rounded at the tip than either L. croceus or L. latisectus. The calyx is glabrous to sparsely pubescent, with the fewest number of trichomes per lobe than any other species in this complex (gen- 2001] BATTAGLIA AND PATTERSON: LEPTOSIPHON ANDROSACEUS COMPLEX is erally less than 10). If present, the trichomes are found only on the margins of the lobes and are non- glandular, another difference between it and both L. latisectus and L. croceus. The calyx lobes are mod- erately long (7—8 mm), but very wide at the base, appearing more or less deltoid. The limb of L. rosaceus is moderate in size (7— 9 mm) and similar in shape to L. croceus and L. latisectus, being smaller and more rounded than L. androsaceus. In contrast to L. croceus, the limb is white (or pink) with a pale yellow throat, and the corolla tube is not as long (21—26 mm). Tube length and width (1 mm) is more like that of L. andros- aceus and L. latisectus. Unlike L. androsaceus, there is no violet coloration in the throat. The stig- mas are generally long, being 4—5 mm, and well exserted beyond the throat. Porter and Johnson (2000) did not include L. ro- saceus in their recent revision; therefore the follow- ing new combination is made: Leptosiphon rosaceus R. Battaglia, comb. nov. Leptosiphon parviflorus var. rosaceus Hooker. Curtis’ Botanical Magazine 96, 1870. Tab 5863. Leptosiphon parviflorus. Based on the analyses in this study L. parviflorus is taxonomically discrete at the species level, despite the observed variation. This species is variable in leaf and corolla size and shape, and especially so in corolla color pattern. In general, measurements for most morphological characters of L. parviflorus fell in the mid-range of the other species in the L. androsaceus complex (larger than L. bicolor or L. acicularis, smaller than L. androsaceus or L. rosaceus). The calyx is long (7-9 mm) and the calyx lobes are nearly equal in length to the fused portion. They are longer than L. latisectus, and not as wide as L. croceus. Like L. latisectus and L. croceus, an important identifying characteristic is its densely glandular pubescent ca- lyx. The limb lobes are 4—6 mm long and range from elliptic to oval or obovate. Lobe color may be white, pale yellow, yellow-orange, lavender-pink, or dark pink, with or without red spots or darker pink striations. The throat may be violet, yellow, or orange. In contrast to other species the throat is narrowly constricted at the base. Unlike L. latisec- tus or L. croceus, which have wider corolla tubes, the corolla tube of L. parviflorus is 0.6 (—1) mm, and 18-33 (—45) mm long. Tube length is highly variable, the standard deviation for tube length is twice that of the other species in this complex. Likewise, stigma length and style exsertion are also variable, as they lengthen over time in unfertilized flowers (Goodwillie 1999a). Stamens are well ex- serted, extending one-half the length of the corolla lobes. Goodwillie (1997) has shown L. parviflorus to be a self-incompatible, obligate outcrosser, with limited ability for wind pollination. 76 Corolla color patterns within L. parviflorus. Clearly different color patterns can be distinguished in this species, and some of these morphs correlate with other morphological features; however, the re- sults of the multivariate analyses show distribution of these color patterns across L. parviflorus do not support sublevels of groupings. Within L. parviflo- rus the *‘2-spot’’ morph has the strongest support, and there is some weak support for the “‘bullseye”’ and ‘‘white with yellow” morphs based on the UPGMA clustering (Fig. 6) and the DA (Fig. 7), but examination of the variables fails to reveal dis- tinct differences among the groups based on color. Tube width and tube length are the characters responsible for the greatest separation between ‘‘2- spot’’ and the other color forms. The corolla tubes of the *‘2-spot’’ morph are generally very long (28— 40 mm) and somewhat wider (0.8 mm) than those of the remaining color morphologies (0.5 mm). Tube length in the “‘bullseye”’ morph is also longer than average (22—29 mm), though not as long as the *‘2-spot,’”’ and, like the other color morphs, is much narrower (0.5 mm). The limbs of ‘‘2-spot’’ and “‘bullseye”’ are also different, and provide a clearer separation than tube length. The *‘2-spot’? color morphs have limb lobes that are larger than “‘bullseye”’ and are similar in shape to most other color morphs. In fact, all co- rolla measurements are consistently higher for the **2-spot”’ than the ‘bullseye’? morphs. This is par- ticularly true of the tip of the limb lobe, which is the largest in ‘‘2-spot’’ and the smallest in “‘bulls- eye.’ The limb lobes of “‘bullseye’’ are conse- quently narrower and more elliptic than the ‘‘2- spot,’’ which are generally wider and more round- ed, or flattened (Figure 2). The differences between the **2-spot”’ and “‘bulls- eye’’ color morphs is interesting because these are the only color morphs with red markings on the corolla lobes. The presence of two red dots could seemingly evolve to become one large bar, or vice versa, and we would have expected their morphol- ogies to be more similar. Also, the *‘2-spot’’ flowers have a yellow throat, whereas the “‘bullseye”’ have a dark yellow-orange throat. Remaining Questions Of the species in the L. androsaceus complex, L. parviflorus in particular raises many questions and merits further study. How is the variation in color pattern maintained? There is no geographic com- ponent to the color morphologies observed in our analysis, and only limited morphological separa- tion. With such a highly modified corolla (nectar guides and long nectar tubes) and obligate outcross- ing, it seems natural to expect specific pollinators are involved. Zebell (1993) examined a similar pat- tern of corolla color variation in Calochortus ven- ustus Benth. He hypothesized that “flower color and pattern are ‘released from tight pollinator se- MADRONO lection pressures’... as long as enough flowers get pollinated to maintain the species.’’ Perhaps the limited wind pollination in L. parviflorus, as re- ported by Goodwillie (1997), is enough of a selec- | tion release to allow for such variation in color morphology. Based on the recent analytical work of Buxton | (1993), Schemske and Goodwillie (1996), and Bat- taglia (1999), L. parviflorus should be approached | with a careful eye. It is highly variable, not only in | regard to color patterns, but in other vegetative and | floral characters. Further taxonomic investigations, particularly populations of L. parviflorus from the | Sierra Nevada foothills and south of the range cov- ered in our analysis, are warranted to fully assess [Vol. 48 | the variation observed in this species. Only after | the fundamental morphological field work is com- — pleted to circumscribe the taxa can we begin to ad- | dress other questions concerning phylogenies, pop- ulation genetics, or the evolutionary history of the ! complex in California. TAXONOMIC KEY TO THE LEPTOSIPHON ANDROSACEUS COMPLEX 1. Corolla tube <10 mm L. serrulatus 1.’ Corolla tube >10 mm. 2. Calyx densely pubescent throughout, trichomes glandular or non glandular. 3. Leaf and calyx lobes acerose, corolla lobes generally <4 mm, corolla lobes yellow. BE ee ee ea Sim on CR ae L. acicularis 3.' Leaf and calyx lobes not acerose, corolla lobe color variable. 4. Corolla lobes <3 mm, tube <14 mm, northern Puget Sound, Vancouver Is- land, possibly Coastal N CA 4.’ Corolla lobes >3 mm. 5. Calyx lobes deltoid, width at mid- dle of lobe 1 mm, corolla lobes rounded at apex, always yellow, generally <7 cm tall, coastal bluffs San Mateo Co L. croceus Calyx lobes narrowly acute, width at middle of lobe 7 cm tall. 6. Corolla tube width generally >1 mm, leaf lobes commonly spatulate, >2 mm wide, corolla lobes generally >6 mm and rounded L. latisectus Corolla tube width generally <1 mm, leaf lobes not com- monly spatulate, <2 mm wide, corolla lobe generally <6 mm and shape varied elliptic to ob- ovate L. parviflorus 2.’ Calyx not densely pubescent, trichomes ciliate and non glandular. 7. Corolla lobes <4 mm long, stigmas <1 L. bicolor 2001] 7.' Corolla lobes >4 m long, stigmas >1 mm 8. Corolla lobes 4—6 mm long, stigmas generally <2 mm L. jepsonii 8.’ Corolla lobe 6—14 mm long, stigmas generally >2 mm 9. Leaf lobes spatulate and more or less fleshy, corolla lobes generally 6-8 mm long and rounded, low growing, coastal San Francisco Bay area L. rosaceus 9.’ Leaf lobes not as above, corolla lobes generally >8 mm long, often with an apiculation at tip L. androsaceus ACKNOWLEDGEMENTS We are especially indebted to Toni Corelli, Mike Vasey and Randy Morgan for their keen field observations, and for guiding us to the very important populations from Moss Beach and Mori Point, which ultimately resulted in the resurrection of Leptosiphon croceus and L. rosaceus. We are also grateful to Jasper Ridge Biological Preserve, San Mateo County Parks and Recreation, Santa Clara County Parks and Recreation Department, East Bay Re- gional Park District, Pinnacles National Monument, Los Padres National Forest, and the State of California De- partment of Parks and Recreation for permits and access to plant material. We wish to thank Dr. Greg Spicer for his assistance with the preparation of this manuscript, Dr. Stan Williams for his advise on multivariate statistical analyses, and Dr. Goodwillie for her help with the addition of L. minimus to the taxonomic key. Leigh Johnson and one anonymous reviewer provided thoughtful criticisms on an earlier version of this paper. Financial support for this project was provided by a grant from the California Native Plant Society and a scholarship from the East Bay Chapter of the California Native Plant Society. Literature Cited BATTAGLIA, R. E. 1999. A taxonomic analysis of the Lin- anthus androsaceus complex (Polemoniaceae) using morphological data. Masters thesis. San Francisco State University, San Francisco, CA. BELL, C. D. AND R. PATTERSON. 2000. Molecular Phylog- eny and Biology of Linanthus (Polemoniaceae). American Journal of Botany 87(12):1857—1870. » ——., AND L. A. HAMILTON. 1999. Sec- tional integrity in Linanthus (Polemoniaceae): a mo- lecular phylogeny of section Dianthoides. Systematic Botany 24:632—644. BENTHAM, G. 1833. Polemoniaceae. Edward’s Botanical Register 19:plate 1622. London. . 1845. Polemoniaceae. pp. 302—322 in A. P. De Candolle (ed.), Prodromus systematis naturalis regni vegetabilis 9 Paris. . 1839-1957 (Aug. 1849). Planta Hartwegiana. G. Pamplin, London. AND J. D. HOOKER. 1876. Polemoniaceae. Gen- era Plantarum 2:820—824. Buxton, E. 1993. A morphometric study of the Linanthus androsaceus (Polemoniaceae) complex. Masters the- sis. San Francisco State University, San Francisco, CA. . 1994. Linanthus latisectus (Polemoniaceae), a BATTAGLIA AND PATTERSON: LEPTOSIPHON ANDROSACEUS COMPLEX w newly described species from the California North Coast Ranges. Madrono 41(3):178—184. DANFORTH, E 1945. The genus Linanthus. A concept of the genus and a taxonomic study of its species. Mas- ters thesis. Stanford University, Stanford, CA. EAstwoop, A. 1904. Some new species of western Pole- moniaceae. Botanical Gazette 37:442—443. ENDLICHER, S. 1836-1840. Polemoniaceae. Genera plan- tarum secundum ordines naturales deposita 2:657. GOODWILLIE, C. 1997. The genetic control of self-incom- patibility in Linanthus parviflorus (Polemoniaceae). Heredity 79:424—4372. . 1999a. Wind pollination and reproductive as- surance in Linanthus parviflorus (Polemoniaceae), a self-incompatible annual. American Journal of Bota- ny 86:948. . 1999b. Multiple origins of self-compatibility in Linanthus Section Leptosiphon (Polemoniaceae): Phylogenetic evidence from internal-transcribed- spacer sequence data. Evolution. 53(5):1387—1395. GOODWILLIE, C. AND J. W. STILLER. 2001. Evidence for polyphyly in a species of Linanthus (Polemoniaceae): convergent evolution of self-fertilizing taxa. System- atic Botany (in press) GRANT, V. 1959. Natural history of the Phlox family. Mar- tinus Nijhof, The Hague. AND K. A. GRANT. 1965. Pollination in the Phlox family. Columbia University Press, New York, NY. GRAY, A. 1870. Revision of the North American Pole- moniaceae. Proceedings of the American Academy of Arts and Sciences 8:247-—282. . 1886. Synoptical flora of North America, edi- tion 2:139. Ivison, Blakeman, Taylor & Co., New York; Ni. GREENE, E. L. 1889-1892. Some American Polemoni- aceae. Pittonia 2:251—260. shoe Eryinea. 3.121. Hooker, J. D. 1870. Leptosiphon parviflorus var. rosa- ceus. Curtis’ Botanical Magazine 96. Tab 5863. Lon- don. JEPSON, W. L. 1901. Flora of western middle California. p. 431. . 1925. A manual of the flowering plants of Cal- ifornia. University of California Press, Berkeley, CA. . 1943. Polemoniaceae. A flora of California, vol. 3(2):131—222. Associated Students Store, Uni- versity of California, Berkeley, CA. JOHNSON, L. A., J. L. SCHULTZ, D. E. SOLTIS, AND P. S. SoLTis. 1996. Monophyly and generic relationships of Polemoniaceae based on matK sequences. American Journal of Botany 83:1207—1224. Mason, H. L. Some additional notes of Polemoniaceae. Madrono 9:249—255. 1951. Polemoniaceae. Pp. 396—474 in L. Abrams (ed.), Hlustrated flora of the Pacific states, Vol. 3. Stanford University Press, Stanford, CA. MEACHAM, C. A. AND T. DUNCAN. 1989. MorphoSys, Ver- sion 1.26. University of California, Berkeley, CA. MILLIKEN, J. 1904. A review of California Polemoniaceae. University of California Publications in Botany 2:1— aly. Munz, P. A. 1959. A California Flora. University of Cal- ifornia Press, Berkeley, CA. NuTTALL, T. 1848. Descriptions of plants collected by Wil- liam Gambel, M. D., in the Rocky Mountains and Upper California. Journal of the Academy of Natural Sciences of Philadelphia 2:156—157. 78 MADRONO PATTERSON, R. 1993. Polemoniaceae. Jn J. C. Hickman (ed.), The Jepson Manual: Higher plants of Califor- nia. University of California Press, Berkeley, CA. PIMENTEL, R. 1993. BIOSTAT II: a multivariate statistical toolbox tutorial manual. Sigma Soft, Placentia, CA. PorRTER, J. M. 1996 [1997]. Phylogeny of Polemoniaceae based on nuclear ribosomal internal transcribed spac- er DNA sequences. Aliso 15:57-—77. AND L. A. JOHNSON. 2000. A phylogenetic clas- sification of Polemoniaceae. Aliso (in press). SCHEMSKE, D. W. AND C. GOODWILLIE. 1996. Morpholog- ical and reproductive characteristics of Linanthus jep- MADRONO, Vol. 48, No. 2, p. 78, 2001 [Vol. 48 sonii (Polemoniaceae), a newly described, geograph- ically restricted species from Northern California. Madrono 43(4):453—463. SNEATH, P. H. A. AND R. R. SOKAL. 1973. Numerical tax- onomy. W. H. Freeman & Co., San Francisco, CA. SPSS 1997. SPSS for Windows version 8.0. SPSS Inc. TABACHNICK, B. G. AND L. S. FIDELL. 1996. Using multi- variate statistics, 3rd ed. Harper-Collins College Pub- lishers. ZEBELL, R. 1993. Systematic evaluation of three species of Calochortus (Liliaceae): C. venustus, C. simulans, C. argillosus. Masters thesis. San Francisco State University, San Francisco, CA. NOTEWORTHY COLLECTIONS CALIFORNIA CHORIZANTHE PARRYI VAR. FERNANDINA (POLYGONA- CEAE)—Ventura Co., S. slope of Laskey Mesa, Ahmanson Ranch; NE 3.2 Km Mureau Road and 101 Freeway, aban- doned roadbed on compacted soils, San Andreas sandy loam, | May 1999, R. E. Riefner & T. Bomkamp 99-271 (RSA); same location as above, 3 Jun 1999, R. E. Riefner 99-283 (RSA); same location as above, 19 Jun 1999, R. E. Riefner 99-299 (RSA). Previous knowledge. Historically known from sandy places along drainages of the San Fernando Valley, north- eastward into the Castaic Creek and Lake Elizabeth areas of the Liebre Mountains, and southward along the Los Angeles coastal plain into Orange County near Santa Ana (Reveal & Hardham 1989, Phytologia 66:98—198). Reveal and Hardham (loc. cit.) also report a J. G. Lemmon col- lection from San Bernardino County, but without other locality data, and several C. C. Parry collections without any locality information. Reveal and Hardham (loc. cit.) report 32 collections of the taxon, the most recent being Hoffmann’s 1929 specimen from Elizabeth Lake. To this may be added an undated A. Davidson collection (s.n.) from Toluca (RSA 392509), and another A. Davidson specimen (s.n.) dated 11 May 1890 from Burbank (RSA 392787), both in Los Angeles County. These two speci- mens bear Reveal’s annotation labels and were apparently omitted inadvertently from the 1989 paper. Significance. Widely thought extinct (e.g., Skinner & Pavlik, C.N.PS. Inventory of Rare and Endangered Vas- cular Plants of California, 5th ed., 1994; Hickman, ed., The Jepson Manual: Higher Plans of California, Univer- sity of California Press, 1993; Reveal & Hardham loc. cit.). The type locality is San Fernando Canyon, Los An- geles County. The Laskey Mesa population represents a first record for Ventura County, and at present, the only known extant population. Seed collected from a number of the stands comprising the Laskey Mesa population have been placed into long-term storage at Rancho Santa Ana Botanic Garden (M. Wall, personal communication). —STEVE Boyb, Herbarium, Rancho Santa Ana Botanic Garden, 1500 N. College Avenue, Claremont, CA 91711. MApDRONO, Vol. 48, No. 2, pp. 79-89, 2001 MORPHOLOGICAL ASPECTS OF SEEDLING ESTABLISHMENT IN FOUR TEMPERATE REGION PHORADENDRON (VISCACEAE) SPECIES BEVERLEE M. RUHLAND! AND CLYDE L. CALVIN Department of Biology, Portland State University, Portland, OR 97207 ABSTRACT Morphological aspects of seedling establishment were studied in four species of the mistletoe genus, Phoradendron (Viscaceae). Marked differences occurred between species. In P. densum and P. villosum the plumular shoot developed as the main shoot of the plant, whereas in P. juniperinum the plumular shoot failed to develop in about 10% of seedlings. When the plumular shoot failed to develop a nonplu- mular seedling shoot displaced the original seedling axis to a lateral position near the base of the devel- oping nonplumular shoot. The undeveloped seedling axis may remain attached and living for several years. In P. californicum the plumular shoot never developed in two of the populations studied, but did develop in a small percentage of individuals in a third population. In both species nonplumular shoots developed from a mass of tissue, termed the haustorial cushion, that forms immediately beneath the holdfast. A haustorial cushion was also observed in P. densum, but the shoot buds that formed on the cushion rarely developed into shoots. Cotyledon characters also varied between species, but in all four species the small cotyledons were persistent. Contrary to earlier reports, the cotyledons do not later develop into foliage leaves. In agreement with Kuijt (1990) we regard the cryptocotylar condition ob- served in P. californicum and, on rare occasion, in P. juniperinum, both squamate species, to be advanced. The morphological, ecological and evolutionary implications of these observations are illustrated and discussed. Phoradendron and six closely related genera comprise the mistletoe family, Viscaceae. It has re- cently been suggested that the seven viscoid genera be transferred to the Santalaceae (APG 1998). The large genus Phoradendron is broadly divided into two subgroups, based on the presence or absence of scale-like appendages, the cataphylls (Wiens 1964). The acataphyllous species are predominantly northern in distribution, and all but one of the U.S. species are from this group. Monographs of the ge- nus include those by Trelease (1916) and Wiens (1964) (acataphyllous species only). A revision of the entire genus is now in progress (Kuijt personal communication). Ashworth (2000) has analyzed phylogenetic relationships in the tribe Phoraden- dreae, which includes both Phoradendron and the sympatric and morphologically similar genus, Den- drophthora. The seven viscacean genera are considered to have relatively uniform seedlings (Kuijt 1982), ex- cept for the genus Arceuthobium. Typically, the elongating root axis is the first organ to emerge from the seed (Calvin 1966; Kuijt 1969; Tubeuf 1923). As the tip of the root contacts the host branch, a disc-shaped holdfast is formed. Subse- quently, the seedling becomes erect, with its coty- ledons and shoot tip still enclosed within the fruit mass. Eventually the fruit vestiges are shed and the elongating plumular shoot emerges. Developed seedlings of both Viscum album L. (Tubeuf 1923) ' Current address: 23670 Shepard Road, Clatskanie, OR 97016 and Phoradendron macrophyllum (Engelm.) Cock- erell (Calvin 1966) have an elongate hypocotyl, two small cotyledons and a plumular shoot. While the first aerial shoot is usually plumular in origin, shoots may also arise from the haustorial disc re- gion and/or from the endophytic system, particu- larly when the plumular shoot is damaged or lost (York 1909). Information on seedlings of the tra- ditional santalacean mistletoes is sparse (Kuyt 1990). The pattern of seedling establishment in the highly specialized genus Arceuthobium is strikingly different than that described for Viscum and Phor- adendron. In Arceuthobium the cotyledons are only rudimentary, and the shoot apex is poorly devel- oped (Cohen 1963). At germination epicotylar growth is suppressed in favor of endophytic estab- lishment (Kuit 1969). All aerial shoots are root- borne, arising from the endophytic system two or more years after the initial infection (Hawksworth and Wiens 1996). Phoradendron californicum Nutt. is reported to have a pattern of establishment sim- ilar to that of Arceuthobium in that, like Arceu- thobium, aerial shoots are said to arise exclusively from the endophytic system (Kuijt 1989). In Vis- cum minimum the embryonic shoot apex aborts, and the initial seedling shoots arise from the en- dophytic system or shoot buds that form on the holdfast or directly below it (Kut 1986). The shoot apex of P. macrophyllum, a large-leaved species, appears to be more highly developed than that of Arceuthobium (compare Cohen 1963, fig. 3, with Calvin 1966, fig. 9). Major changes from the usual 80 MADRONO sequence of seedling growth, such as reduction or suppression of embryonal apical differentiation, seem to be an evolutionary trend in parasitic flow- ering plants (Teryokhin and Nikiticheva 1982). For mistletoes in general an insufficient data base exists on seedling structure (Kuijt 1982). This lack of basic information can lead to misinterpretation of life cycle sequences important to physiological, developmental, and taxonomic determinations. During another study (Calvin et al. 1993) we noted a number of unusual seedling features in P. juni- perinum. This observation, coupled with Kuijt’s (1982) comments on the paucity of information on mistletoe seedlings in general, prompted this study. Three additional species of Phoradendron were in- cluded to provide a comparative dimension to the study and to broaden its focus. Our field and lab- oratory studies indicate that a greater variation in seedling morphology exists within the genus than has been reported. MATERIALS AND METHODS The four species of Phoradendron chosen for this study were: P. californicum (desert mistletoe), P. densum Trel. (dense mistletoe), P. juniperinum A. Gray (juniper mistletoe), and P. villosum (oak mistletoe). For P. juniperinum, the primary focus of the study, several hundred seedlings were ex- amined. Fewer specimens (100—200) were analyzed for each of the other species studied. More than one population was utilized for 3 of the species, as in- dicated below. The 4 species studied occur at the northern limit of the Phoradendron range. Two of the species, P. californicum and P. juniperinum, are squamate; that is, their leaf blades are scale-like. Phoradendron densum has small leaves, and P. vil- losum has leaves of medium size (as compared to P. macrophyllum). Specimens of the four Phoradendron species were collected at intervals during the years 1984— 1987. The materials of P. californicum were col- lected in proximity to Mesquite, NV, Oatman, AZ, and on the Cabezon Indian Reservation near Indio, CA. The host trees were primarily Prosopis, but specimens were also collected from Acacia, Cer- cidium and Larrea. Phoradendron densum was col- lected 6.5 miles downstream from the J. C. Boyle Powerhouse in Klamath Canyon, OR, and in the vicinity of Weed and Henley, CA. The host tree was Juniperus. Phoradendron juniperinum was col- lected in and near Bend, OR, also growing on Ju- niperus. Phoradendron villosum was collected in two widely separated localities: near Corvallis, OR, and 30 miles east of Red Bluff, CA, along State Highway 36. Herbarium specimens of the four spe- cies studied are deposited in the Portland State Uni- versity Herbarium (HPSU). For all species, collect- ed specimens were brought to the laboratory and examined while fresh, using a Zeiss dissecting mi- croscope. Some specimens were preserved in al- [Vol. 48 | \ cohol for further study. Description of gross shoot | morphology centered on the nature of the cotyle- | dons (number, size, fused or free, persistence, vis- | ibility), the origin of aerial shoots (plumular, haus- | torial disc region, endophytic system), the fate of | aerial shoots, internode extension, and length of the hypocotyl. In this study three categories of aerial shoots are | recognized: 1) Plumular shoots, arising from the | shoot pole of the embryo; 2) Nonplumular seedling — shoots, arising from the haustorial disc or cushion | region; and, 3) Root-borne shoots, arising from the | endophytic system. Nonplumular seedling shoots probably represent a type of root-borne shoot; but | they are segregated here since a major focus of this — report is on the contrasting origins of aerial shoots | in the species studied. Further, since the morpho- | logical nature of the holdfast and cushion regions | remains problematic, a distinction is justified. | ——_ RESULTS In the Bend, OR, area the fruits of Phoradendron | jJuniperinum reach maturity in early winter. Several | bird species feed on these fruits and void seed on | nearby host branches (Fig. 1). Shortly thereafter _ germination begins. In P. juniperinum (Figs. 1-12) | and the other Phoradendron spp. examined the rad- | icle is the first structure to emerge from the seed (Fig. 1). The specimen shown was collected in mid- May and reflects the slow but continuous growth of the green hypocotyl-root axis. Eventually the root tip reaches a penetration site, which is typi- cally, but not always, beneath a scale leaf, and form — a holdfast (Fig. 2). Throughout this stage the cot- yledon tips remain more or less embedded within the endosperm, which is in turn covered by the sclerophyllous ‘“‘seed”’ coat (Fig. 2). As growth continues, the seedling becomes erect, shedding the remnants of the endosperm, and the plumular shoot develops (Fig. 3). At this stage of seedling growth striking differ- ences between seedlings become evident. In many seedlings an expanded region of tissue, here des- ignated the haustorial cushion, develops beneath the holdfast raising it above the host surface (Fig. 4). Of 254 seedlings studied for this feature 52% developed a cushion. Typically, 2—4 vegetative buds arise on the cushion (Fig. 4), but as many as 11 were observed. These buds may remain dormant for a time (Fig. 4) or initiate shoot growth almost as soon as they appear (Fig. 5). Seedlings that failed to form a cushion did not develop nonplumular seedling shoots at the initial infection site. Root- borne shoots may develop later, however, directly from the endophytic system. The development of shoots from the endophytic system is a common phenomenon in some viscacean genera and has been observed by several workers (Bray 1910; Cannon 1901; Kuijt 1969; York 1909). In approximately 20% of specimens studied, 2001] RUHLAND AND CALVIN: PHORADENDRON SEEDLING MORPHOLOGY 81 Fics. 1-6. P. juniperinum. Fig. 1, germinating seed, s, with elongate hypocotyl-root axis, r. Fig. 2, seedling with elongate hypocotyl, h, and holdfast label? Fig. 3, seedling with cotyledons, c, and plumular shoot. Fig. 4, young plant showing holdfast, hf, haustorial cushion, hc, and vegetative shoot bud, sb. Fig. 5, seedling with curved hypocotyl, prominent holdfast, young plumular shoot, ps, and developing nonplumular shoot, ns. Fig. 6, young plant with well- developed plumular and nonplumular seedling shoots; note basal cup, be, present at base of nonplumular seedling shoot; endosperm remains, e; host branch, hb. All x1. 82 MADRONO nonplumular seedling shoots originating on the haustorial cushion equaled (Fig. 6) or surpassed plumular shoots in size by the time the latter had five extended internodes. Nonplumular shoots can be identified by a basal, cup-like structure, presum- ably the lowermost leaf pair. Plumular shoots, in contrast, have a holdfast, a hypocotyl and cotyle- dons (Fig. 6). A particularly interesting feature observed in some P. juniperinum seedlings is the virtual ab- sence of growth from the plumular apex. This phe- nomenon was noted in about 10% of the seedlings examined. In this event, the growth of nonplumular shoots displaces the original seedling to a lateral position near the base of the developing nonplu- mular seedling shoots (Figs. 7—9). The remnants of the initial seedling are persistent and have been seen on infections in which developing shoots had ten or more extended internodes. When first ob- served these small, undeveloped “‘basal’’ seedlings were interpreted to be autoparasites, but as devel- opmental stages were examined the true nature of these seedling structures became evident. Nonplumular seedling shoots often appear more vigorous than shoots developing from the embry- onic shoot apex. This vigor may be related to sev- eral factors. Frequently nonplumular shoots have a darker green color than plumular shoots (Fig. 9), possibly reflecting a higher content of chlorophyll. A constriction, not seen in other shoots, is often evident in the hypocotyl region of plumular shoots immediately above the holdfast (Fig. 10). Addition- ally, extreme curvature of the hypocotyl can occur in this region (Fig. 11). These features may dimin- ish the quality and quantity of the vascular connec- tion between shoot and root. Constriction and/or curvature may also represent a weak point, since seedlings broken off in this region were observed in the field. Phoradendron juniperinum generally has two cotyledons (Fig. 3) although the number ranges from one to three. Usually the cotyledons are free and spreading; that is, plants display phanerocotyly (Figs. 3, 6). In some cases the cotyledons may be partially or wholly fused along one edge; rarely they are fused to such an extent that the demarca- tion between the individual cotyledons is difficult to discern (Fig. 12). The cotyledons display a set of morphological characteristics that distinguishes them from subsequent leaves. First, their extreme tips (approximately | mm) are seen to be pointed and chlorotic upon removal of the endosperm. Later the points appear brown and withered. These min- ute tips are nearly always present, and can be seen when cotyledons are examined closely. Second, the cotyledons join smoothly and without interruption with the hypocotyl region (Fig. 3), whereas the sub- sequent scale leaves often have a definite line of demarcation at their junction with the stem (sl in Fig. 9). Third, the cotyledons are not deciduous, whereas in species with foliage leaves (but not in [Vol. 48 P. juniperinum) the latter are deciduous. Although} the shape of cotyledons changes as stem diameter | increases, they can still be recognized on specimens | such as that seen in Figure 12, which has over 20! extended internodes. Using these criteria the pres-! ence of cotyledons can be used to distinguish be- | tween plumular and nonplumular shoots in the! field. | Phoradendron densum (Figs. 13-18) follows | closely the sequence of events described for early | seedling growth of P. juniperinum. Following ger-_ mination and holdfast formation, the seedling be- comes erect (Fig. 13). A haustorial cushion may | also form, but not as commonly as in P. juniper- | inum. Of 121 specimens examined, 37% had a_ haustorial cushion. Generally one or two shoot buds | are initiated on the cushion, but 90% of these had not developed into shoots even after seven extend- | ed internodes were visible on the plumular shoot. | It was noted, however, that all buds produced | shoots in the six cases where the original seedling had been damaged by herbivory or mechanical in- | jury. This species frequently produces root-borne | shoots directly from the endophytic system, and large clusters of these shoots are found at varying distances from the initial infection site. In none of the plants examined did we observe arrested plu- mular shoot growth as illustrated in Figures 7—9 for P. juniperinum. The cotyledons of P. densum, typically two or three in number, display the distinct morphological - features described for P. juniperinum (Fig. 14). Fu- sion of the cotyledons is common, occurring at their tips (Fig. 15), along their margins (Fig. 16), or a combination of these (Fig. 17). In any event, the cotyledons do not appear to enlarge very much and remain smaller, although somewhat more suc- culent, than subsequent foliage leaves. With contin- ued lateral expansion of the main stem, the coty- ledons may assume a position perpendicular to the stem axis, and they persist even after foliage leaves above them are lost (Fig. 18). As in P. juniperinum, the presence of cotyledons is a reliable indicator of plumular shoots. Seedling development in P. californicum (Figs. 19-24) differs from that of both P. densum and P. juniperinum. Upon germination the extending rad- icle, which is distinctly reddish in color, forms a holdfast from which the endophyte is established (Fig. 19). Seedlings may then become erect with respect to the host branch (Fig. 19) or remain in the same plane as the host branch (Figs. 20-22). At about the same time a thin to somewhat thick- ened haustorial cushion forms immediately below the holdfast, elevating the holdfast above the branch surface (Figs. 20, 22). From the cushion region an average of seven, but as many as 20, shoot buds are formed (Fig. 23). When a large number of shoot buds are formed, they may be present around the entire periphery of the cushion, eventually forming shoots that radiate outward in » 2001] - RUHLAND AND CALVIN: PHORADENDRON SEEDLING MORPHOLOGY 83 Fics. 7-12. P. juniperinum. Fig. 7, nonplumular seedling shoot, ns, with undeveloped plumular axis at base as evidenced by seed, s, with developed hypocotyl-root axis, h. Fig. 8, plant with 2 nonplumular shoots and _ basal, undeveloped plumular axis, pa. Fig. 9, young plant with developing nonplumular shoots and basal plumular axis. Fig. 10, seedling with developed plumular shoot, ps; note constricted hypocotyl and adjacent root-borne shoot, rs. Fig. 11, portion of young plant showing minute cotyledons, c, and contorted hypocotyl. Fig. 12, older plant; fused cotyledons still visible; scale leaf sl. All <1. all directions from the initial infection site (Fig. 24). Regardless of their number and distribution, the vast majority of vegetative buds develop, re- Sulting in infections having thick clusters of shoots. Shoots may also arise from the endophytic system, particularly when the initial infection site is at a node. In all but one population of P. californicum 100% of plumular shoots failed to develop. How- ever, the original seedling remained, often for 84 MADRONO [Vol. 48 Fics. 13-18. P. densum. Fig. 13, seedling with cotyledons, c; note second seedling at unlabeled arrow. Fig. 14, older seedling with succulent cotyledons and foliage leaves, fl. Fig. 15, seedling with cotyledons united at tip, ct, and emergent foliage leaves. Fig. 16, two seedlings, seedling at left has fused cotyledons and emergent foliage leaves. Fig. 17, seedling with three cotyledons fused both at tips and margins. Fig. 18, older plant with persistent cotyledons, note leaf scar, Is, at adjacent node; holdfast, hf. All <1. years, as a usually living hypocotyl-root axis in amongst the developing nonplumular seedling shoots (Fig. 24). In some cases the elongate hypo- cotyl-root axis becomes erect, and remnants of the endosperm may remain visible, obscuring the cot- yledons (Fig. 22). Although not always readily vis- ible, the location of the failed plumular shoot clear- ly identifies the initial site of infection. In the Cab- ezon Indian Reservation population, in contrast, the plumular shoot had developed in 5 of 17 newly 2001] RUHLAND AND CALVIN: PHORADENDRON SEEDLING MORPHOLOGY 85 Fics. 19-24. P. californicum. Fig. 19, germinated seed, s, with holdfast, hf. Fig. 20, germinated seed that has entered host tissue; note elevated haustorial cushion, hc, beneath holdfast. Fig. 21, young plant with undeveloped plumular axis, pa, at base of nonplumular shoot. Fig. 22, young plant with erect, undeveloped plumular axis at base of nonplu- mular shoot. Fig. 23, germinated seed with elongate hypocotyl, h, and several vegetative shoot buds, sb, radiating outward from the haustorial cushion. Fig. 24, older plant with seven nonplumular shoots, ns; note plumular shoot axis at center of radiating shoots. All x1. 86 established seedlings examined. In these seedlings the endosperm mass had been shed and the coty- ledons were clearly visible. In size and form, the cotyledons resembled those of P. juniperinum. All of the seedlings had two cotyledons, and in none were the cotyledons fused. Seedling establishment in P. villosum resembles closely establishment events described previously for P. macrophyllum (Calvin 1966). Dispersed seed lie flat upon the host branch. At germination the elongating hypocotyl-root axis makes contact with the host branch and a holdfast is formed (Fig. 25). Subsequently seedlings become erect (Figs. 26, 27) and the plumular shoot begins its development. In all seedlings and juvenile plants examined two sep- arate cotyledons were evident. As in the other spe- cies described herein the cotyledons were small with brownish, pointed tips. Also, the cotyledons were persistent, remaining on the plant long after later-formed leaves had abscissed (Figs. 26, 27). Fi- nally, as in the other species, the internode imme- diately above the cotyledons frequently showed lit- tle or no elongation (Fig. 26). In none of the more than 25 seedlings examined was a haustorial cush- ion evident, and in none of the specimens were shoot buds present at the infection site. Presumably, root-borne shoots can develop from the endophytic system of P. villosum, as occurs in other Phora- dendron species, but none were observed in the specimens examined. Instead, younger shoots were seen to have two small but persistent cotyledons (Fig. 27). The seedling establishment events described above for four Phoradendron species share features in common with seedling establishment in other viscoid genera, as well as in Tupeia, a member of the Loranthaceae. In Figure 28 is shown a seedling of Viscum rotundifolium collected in South Africa. Visible are the holdfast, elongate hypocotyl, and two small, persistent cotyledons with pointed ends (shown enlarged in Fig. 29). Note also that the in- ternode directly above the cotyledons did not elon- gate, as is common in Phoradendron. In Figure 30 is shown a seedling of Notothixos subaureus col- lected near Wisemans Ferry, New South Wales, Australia. The seedling has a holdfast, an erect, elongate hypocotyl and two minute cotyledons. Im- mediately above the cotyledons are two shoots. It appears that both of these shoots arose from axil- lary buds following injury to the main, plumular shoot. Finally, more than 100 seedlings of Tupeia antarctica, growing on tree lucerne on the Banks Peninsula near Christchurch, New Zealand, were examined for seedling characters. Remarkably, more than one third of these seedling showed small, arrested plumular shoots near the base of their de- veloped nonplumular seedling shoots, a situation almost identical to that illustrated for P. juniperin- um (Figs. 7-9) and P. californicum (Figs. 21—24). MADRONO DISCUSSION Dispersed and germinating seed of the Phora- | dendron species studied lie flat on the host branch; | that is, their smallest dimension is perpendicular to | the branch surface. A similar placement is found in | the viscoid genera Arceuthobium (Hawksworth and Wiens 1996), Notothixos (McLuckie 1923) and Vis- cum (Salle 1983; Kuijt 1986). Dispersed and ger- | minated seed of many loranth genera also lie flat | against the host branch (Polhill and Wiens 1998; | McLuckie 1923; C. L. Calvin personal observa- | from the host branch, and the elongating hypocotyl- root axis curves 180 degrees to contact the host. This contrasting orientation apparently relates to the distribution of viscin in fruits. In both Alepis | and Peraxilla the viscin is positioned in a ring near | one end of the seed (Ladley et al. 1997), whereas | tion). This stands in marked contrast to the dis- persed and germinating seed of the loranth genera © Alepis and Peraxilla which stand on end (Ladley et > al. 1997). In these genera the root tip points away | in the viscoids studied the viscin is dispersed more | evenly around the seed. The formation of a haustorial cushion is de- | scribed for three of the species studied. This cush- | ion arises at the base of the holdfast and is the site of origin of nonplumular seedling shoots. In P. jun- iperinum, which has the most pronounced cushion, its development may raise the holdfast well above the surface of the host branch. Kuijt (1986) noted that in V. minimum seedlings aerial shoots may also arise from directly beneath the attachment disc, as well as from the margin of the disc. York (1909) found that in P. macrophyllum “the aerial shoots which are first formed usually arise from buds, which develop on the attachment disc. ...”’ In the present study shoot buds were not seen to arise on the attachment disc (holdfast), but rather on the cushion of tissue formed directly beneath the hold- fast. The shoots of V. minimum that arise beneath the holdfast (Kuijt 1986) presumably also arise from a cushion-like region. The first aerial shoots of Phoradendron seedlings have generally been regarded as being plumular in origin (Cannon 1901; York 1909; Calvin 1966). In the present study the initial shoots of P. densum and P. villosum were exclusively plumular in ori- gin, whereas those of P. californicum were almost entirely nonplumular. Seedling shoots in P. juni- perinum were either plumular or, somewhat less commonly, nonplumular in origin. Where plumular shoot growth is arrested shoots arise from shoot buds initiated on the haustorial cushion formed be- neath the holdfast. Additionally, root-borne shoots may be formed in proximity to the infection site (see rs in Fig. 11), particularly in P. californicum. These contrasting patterns of shoot origin represent a greater diversity than previously recognized. The situation in P. californicum approaches that in Ar- ceuthobium spp. (Cohen 1963; Hawksworth and 2001] RUHLAND AND CALVIN: PHORADENDRON SEEDLING MORPHOLOGY 87 Fics. 25-30. P. villosum (Figs. 25-27), V. rotundifolium (Figs. 28, 29) and N. subaureus (Fig. 30). Fig. 25, germinated seed with well developed holdfast, hf. Figs. 26, 27, seedling with two cotyledons; note leaf scar, Is, above and directly opposite cotyledons. Figs. 28, 29, young plant with elongate hypocotyl-root axis, holdfast and free cotyledons (one of the two leaves attached at the node directly above cotyledons was removed to enhance visibility of cotyledons). Fig. 30, seedling with two minute cotyledons and small plumular shoot; cotyledon, c, hypocotyl-root axis, h, seed mass, s. All X1. 88 Wiens 1996) and in Tristerix aphyllus (Mauseth et al. 1984) where the initial shoot and all succeeding shoots are root-borne. In P. californicum, however, the first aerial shoots arise from an aerial portion of the seedling, not from the endophytic system as has been reported (Kuijt 1989). In Arceuthobium and 7. aphyllus, on the other hand, the initial shoots arise from the endophytic system and do not appear for some time after penetration of the host branch; about two years in Arceuthobium spp. (Hawksworth and Wiens 1996). The cotyledons of the Phoradendron species studied varied in both number and form. Most seed- lings of P. juniperinum had two cotyledons, but a few had either one or three. In P. densum either two or three cotyledons were present, whereas in P. villosum all seedlings had two cotyledons. The few plumular shoots of P. californicum observed in the Cabezon population also had two cotyledons. The cotyledons of P. densum were the largest and most succulent in appearance, those of P. juniper- inum and P. villosum were intermediate in size and those of P. californicum the smallest. Fused coty- ledons were common in P. densum, but rare or ab- sent in the other Phoradendron species. Observa- tions of earlier workers on the fate of cotyledons are in conflict. Bray (1910) states that in P. macro- phyllum the cotyledons become erect and slowly expand as the first pair of green leaves. York (1909), on the other hand, states that the cotyledons either wither or become slightly enlarged but never form foliage leaves. Our observations are more similar to those of York. Of interest, in all 4 species the cotyledons were persistent, as are the scale leaves of the squamate species. In contrast, the sub- sequently formed foliage leaves of P. densum and P. villosum are deciduous. Persistent cotyledons that do not develop into foliage leaves are consid- ered unusual, and Sporne (1974) cites only two ex- amples of this phenomenon, both in the family Ges- neriaceae. Phoradendron densum and P. villosum were con- sistently phanerocotylous. The vast majority of specimens of P. juniperinum were also phanero- cotylous, but in occasional specimens, such as those shown in Figures 7 and 8, the cotyledons nev- er emerge. In contrast, in only a single population of P. californicum were seedlings observed in which the cotyledons were visible and spreading. It is probable that the ancestor of the viscoids was phanerocotylous, as are the majority of modern vis- coid species. In agreement with Kutt (1990) we regard the cryptocotylar condition observed in highly specialized, squamate species of Phoraden- dron, illustrated here by P. juniperinum and P. cal- ifornicum, to be advanced. Cryptocotyly is regard- ed by some as the ancestral germination pattern in angiosperms as a whole (Gifford 1991). The four Phoradendron species examined show a progressive reductional trend in leaf size. Leaves of P. villosum are of intermediate size (compared MADRONO to those of P. macrophyllum, for example), those of P. densum are small and those of P. juniperinum | and P. californicum are reduced to scales. The four , also display reductional trends in seedling devel- | opment. In this series plumular shoots give way to | nonplumular seedling shoots developing on a haus- torial cushion, and phanerocotyly is replaced by cryptocotyly. This trend is carried even further in | Arceuthobium species. In all members of this genus aerial shoots are root-borne, and neither plumular | shoots nor nonplumular seedling shoots are ever | formed (Hawksworth and Wiens 1996). The signif- | icance of these reductional trends is difficult to as- | certain. One possible advantage is the formation by | seedlings of a larger number of seedling shoots—a _ sort of tillering—at an early age through the for- | mation of shoot buds either on the attachment disc — or on the proliferative tissue formed beneath the | attachment disc. Also, as noted earlier, nonplumular seedling shoots frequently appear more vigorous | than plumular shoots when both occur together in a seedling. Finally, species capable of forming non- | [Vol. 48 | plumular seedling shoots and/or root-borne shoots | have much greater vegetative and reproductive ver- | satility than those relying solely on plumular . shoots. Reduction in or suppression of plumular shoot | growth is common in the viscoid mistletoes. It is - illustrated here for P. juniperinum and P. califor- nicum, and described elsewhere for V. minimum (Kuijt 1986) and Arceuthobium (Kuijt 1969; Hawk- | sworth and Wiens 1996). Reduction in or suppres- sion of plumular shoot growth also occurs in some | Loranthaceae. In the New World loranth T. aphyl-— lus, all seedling aerial shoots are reported to be root-borne (Mauseth et al. 1984), a condition anal- ogous to that in Arceuthobium species. The fre- quent absence of plumular shoot growth is also re- ported here for the monotypic New Zealand lor- anth, Tupeia antarctica. It is probable that the sup- pression of plumular shoot growth occurs to a lesser or greater degree in other genera of viscoids and loranths as well. In Phoradendron and other viscoids the suppression of plumular shoot growth iS positively correlated with the squamate habit. It is interesting that the suppression of plumular shoot growth is common in species that are considered specialized within their respective clades when it represents a basic growth pattern. LITERATURE CITED ANGIOSPERM PHYLOGENY Group. 1998. An ordinal classi- fication for the families of flowering plants. Annals of the Missouri Botanical Garden. 85:531—553. ASHWorRTH, V. E. T. M. 2000. Phylogenetic relationships in Phoradendreae (Viscaceae) inferred from _ three regions of the nuclear ribosomal cistron. I. Major lin- eages and paraphyly of Phoradendron. Systematic Botany 25:349—370. Bray, W. 1910. The mistletoe pest in the Southwest. Bull. 166, Bureau of Plant Industry. USDA. | 2001) } | CALVIN, C. L. 1966. Anatomy of misletoe (Phoradendron | flavescens) seedlings grown in culture. Botanical Ga- | zette 127(4):171-183. | , C. L., C. A. WILSON, AND G. VARUGHESE. 1991. | Growth of longitudinal strands of Phoradendron jun- iperinum in shoots of Juniperus occidentalis. Annals of Botany 67:153-161. CANNON, W. A. 1901. The anatomy of Phoradendron vil- losum Nutt. Torrey Botanical Club Bulletin. 28:374— | 390. | COHEN, L. I. 1963. Studies on the ontogeny of the dwarf mistletoes, Arceuthobium. I. Embryogeny and histo- genesis. American Journal Botany 50:400—407. GiFFORD, H. T. 1991. Germination patterns in dicotyle- dons. Aliso. 13:207—213. HAWkKSWworTH, FE G. AND D. WIENS. 1996. Dwarf Mistle- toes: Biology, Pathology, and Systematics. USDA Forest Service Ag. Handbook 709, Washington DC. KuutT, J. 1964. Critical observations on the parasitism of New World mistletoes. Canadian Journal of Botany 42:1243-1278. 1969. In the Biology of Parasitic Flowering Plants. University of California Press, Berkeley, CA. . 1982. Seedling morphology and its systematic significance in Loranthaceae of the New World, with supplementary comments on Eremolepidaceae. Bo- tanische Jahrbiicher Fiir Systematik 103:305—342. . 1986. Observations on establishment and early shoot emergence of Viscum minimum (Viscaceae). Acta Botanica Neerlandica 35:449—456. . 1989. A note on the germination and establish- ment of Phoradendron californicum (Viscaceae). Ma- drono 36:175—179. . 1990. Correlations in the germination patterns of Santalacean and other mistletoes. /n P. Baas, K. Kalk- man, and R. Geesink (eds.), The Plant Diversity of RUHLAND AND CALVIN: PHORADENDRON SEEDLING MORPHOLOGY 89 Malesia, 63—72. Kluwer Academic Publishers, Dor- drecht. LADLEY, J. D. KELLY AND D. A. Norton. 1997. A guide to hand-planting of New Zealand mistletoes (Loran- thaceae). New Zealand Botanical Society Newsletter. 49:16-19. MausetH, J. D., G. MONTENEGRO, AND A. M. WALSKO- WIAK. 1984. Studies of the holoparasite Tristerix aphyllus (Loranthaceae) infecting Trichocereus chi- lensis (Cactaceae). Canadian Journal of Botany 62: 847-857. McLuckie, J. 1923. Studies in parasitism. A contribution to the physiology of the Loranthaceae of New South Wales. Botanical Gazette 75:333—369. POLHILL, R. AND D. WIENS. 1998. Mistletoes of Africa. The Royal Botanic Gardens, Kew. SALLE, G. 1983. Germination and establishment of Viscum album L. Chapter 9, In M. Calder and P. Bernhardt (eds.), The Biology of Mistletoes, Academic Press, New York, NY. SPORNE, K. R. 1974. The Morphology of Angiosperms. St. Martin’s Press, New York, NY. TERYOKHIN, E. S. AND Z. I. NIKITICHEVA. 1982. Biology and evolution of embryo and endosperm in parasitic flowering plants. Phytomorphology 32(4):335—339. TRELEASE, W. 1916. The Genus Phoradendron: A Mon- ographic Revision. University of Hlinois Press, Ur- bana, IL. TUBEUF, K. 1923. Monographie der Mistel. R. Oldenbourg Verlag, Munich and Berlin. WIENS, D. 1964. Revision of the acataphyllous species of Phoradendron. Brittonia 16:11—54. York, H. 1909. The anatomy and some of the biological aspects of the ““American mistletoe,’ Phoradendron flavescens (Pursh) Nutt. Bulletin of the University of Texas 20, Science Series 13. MADRONO, Vol. 48, No. 2, pp. 90-97, 2001 POLYPLOIDY AND SEGREGATION ANALYSES IN DELPHINIUM GYPSOPHILUM (RANUNCULACEAE) JASON A. KOONTZ! AND PAMELA S. SOLTIS? School of Biological Sciences, Washington State University, Pullman, WA 99164-4236 ABSTRACT Delphinium gypsophilum Ewan consists of both diploid and tetraploid individuals, but the type of polyploidy (allo- vs. autopolyploidy) has not been documented. Cytotyping using flow cytometry indicated that some populations were 2n, others were 4n, and several had mixed ploidy. Triploid individuals rep- resented approximately 20% of the sampled plants. There appears to be little geographic structuring of cytotypes. Progeny arrays from controlled crosses provided evidence favoring tetrasomic inheritance, and Delphinium gypsophilum has allozyme banding patterns that are consistent with autotetraploidy. Genetic data indicate that polyploids may have formed recurrently, but the exact number of origins and specific progenitor-derivative relationships remain uncertain. Conservation efforts should manage the two cyto- types separately, as they represent potentially different evolutionary units. INTRODUCTION Polyploidy is an important phenomenon in the evolution of many plant species. Approximately 47-52% of angiosperms and 44-95% of pterido- phytes may have polyploid origins (Grant 1981; Vida 1976). Given the pervasiveness of polyploidy, much remains to be learned concerning its evolu- tionary consequences. Polyploidy has often been considered to be an evolutionary dead end; how- ever, recent research is revealing that polyploidy is a dynamic process and does not necessarily lead the species toward extinction (reviewed in D. Soltis and P. Soltis 1993, 1999; P. Soltis and D. Soltis 2000). Two types of polyploids exist. Allopoly- ploids combine the genomes of two diploid species via hybridization and subsequent chromosome dou- bling. Autopolyploids arise intraspecifically from a diploid progenitor. Because of the differences as- sociated with their origin, allopolyploids and auto- polyploids have different modes of inheritance and different genetic attributes. Allopolyploids exhibit fixed heterozygosity and disomic-digenic inheri- tance. Autopolyploids do not exhibit fixed hetero- zygosity, but have increased levels of heterozygos- ity due to polysomic inheritance (often detected as unbalanced heterozygotes; reviewed in D. Soltis and P. Soltis 1993). Taxa that exist as both diploid and polyploid individuals offer a natural laboratory in which further studies can be developed to ex- amine the evolutionary consequences of polyploidy (e.g., shifts in mating systems, Cook and Soltis 'Current address and for correspondence: [linois Natural History Survey, Center for Biodiversity, 607 E. Peabody Dr., Champaign, IL 61820-6970, jkoontz @inhs.uiuc.edu * Current address: Florida Museum of Natural History and the Genetics Institute, University of Florida, Gaines- Ville, FL 32614 1999, 2000). This paper presents data on the nature : of polyploidy in Delphinium gypsophilum Ewan. Delphinium gypsophilum gypsum-loving lark- spur, grows in open grasslands in the southern San > Joaquin Valley of central California. Two subspe- cies of D. gypsophilum have been described based | on size, and to a lesser degree, flower color and | range (Lewis and Epling 1954; Warnock 1997). | Delphinium gypsophilum subsp. gypsophilum typi- | cally has white flowers that are larger than those of » D. gypsophilum subsp. parviflorum Lewis & Epling | (Warnock 1997). Delphinium gypsophilum subsp. | parviflorum has flowers that are usually white, but — are often lavender or pink, and this subspecies is | typically located closer to the Pacific coast (Lewis © and Epling 1954). Delphinium gypsophilum subsp. | 16) and © tetraploid (2n = 32) individuals (Lewis et al. 1951), | and the two cytotypes are morphologically indistin- | guishable (Lewis and Epling 1959; J. Koontz pers. | obs.). The morphological similarity between diploid | and tetraploid individuals of D. gypsophilum sug- « gypsophilum consists of both diploid (2n = gests autotetraploidy. Delphinium gypsophilum subsp. parviflorum is apparently only diploid (War- nock 1995). Previous genetic work on D. gypso- philum (Koontz 2000) does not support the recog- nition of subspecies, therefore, we treat both sub- species together as D. gypsophilum in this study. In their study of chromosome numbers of most — Californian larkspurs, Lewis et al. (1951) cytotyped | 378 individuals from 35 populations of D. gypso- philum (the subspecies were not recognized in their — study). Sixteen populations were diploid, and 19 were tetraploid. Neither populations of mixed ploi- — dy nor triploid individuals were documented (Lewis | et al. 1951). Their survey of D. gypsophilum sam- pled populations from throughout the range of the species as it occurred in the 1950’s; however, since © that time, many populations have been lost because 1 | | | | | | | 2001] grown at WSU. KOONTZ AND SOLTIS: POLYPLOIDY IN D. GYPSOPHILUM Oi | TABLE 1. POPULATIONS OF D. GYPSOPHILUM STUDIED, THEIR PLOIDY, METHOD OF DETERMINATION, AND SAMPLE SIZE (N). 1 Numerical population designations are from Koontz (2000). * Full locality data available from first author. * Population Population! Locality? Ploidy/method/N ] Kern Co., Hwy 58 3n-4n/flow/6 2 Kern Co., Hwy 58 2n-3n/flow/6 3 San Luis Obispo Co., Soda Lake 2n-3n-4n/flow/9 5 Kern Co., Elk Hills 4n/flow/6 6 Kern Co., Elk Hills 3n-4n/flow/7 7 Kern Co., Elk Hills 2n-4n/flow/4 9 Kern Co., Elk Hills 2n-3n/flow/5 11 San Luis Obispo Co., Hwy 41 3n-4n/flow/7 12 San Luis Obispo Co., Hwy 41 2n-3n/flow/4 13 San Luis Obispo Co., Cypress Mtn. Rd. 2n/flow/4 14 San Luis Obispo Co., G14 2n/flow/3 15 Monterey Co., New Pleyto Rd. 2n/flow/4 16 Monterey Co., Vineyard Canyon Rd. 3n/flow/1 ii Merced Co., Los Bafios Reservoir 4n/squash/5 20 Kern Co., A. D. Edmonston Pumping Plant 2n/flow/1 1 HR? Merced Co., Howard’s Ranch, Aqueduct mi. 65 4n/squash/5 0? Merced Co., O’ Neill Forebay 4n/squash/5 of extensive development in the Central Valley of California. Delphinium gypsophilum continues to co-exist with human development, but the Califor- nia Native Plant Society (CNPS) has placed this species on its List 4, a “‘watch list’”’ of species that may become threatened or endangered. We reex- amined the distribution of ploidy in D. gypsophilum because the range of D. gypsophilum has changed since 1951 and because additional populations of D. gypsophilum are now known. If the range of D. gypsophilum continues to change and the species becomes threatened or endangered, these data on ploidy will be useful in management plans to en- sure the protection of both cytotypes. Additionally, determining the type of polyploidy in D. gypso- philum is important for understanding the evolu- tionary history of this species. The differences be- tween the genetic attributes of allopolyploids and autopolyploids could affect how the cytotypes are treated taxonomically and ultimately managed in conservation efforts. This study was designed to determine (1) the number, distribution, and similarity of diploid and tetraploid populations, and (2) the mode of inheri- tance, and therefore the type of polyploid, in the tetraploid cytotype. Flow cytometry and mitotic root-tip squashes were used to determine ploidy of the samples, and their ploidal distributions were plotted along with data from Lewis et al. (1951). To determine the type of polyploid, controlled crosses were performed to generate progeny arrays to test for inheritance patterns of allozyme markers. MATERIALS AND METHODS Number, distribution, and similarity of diploid and tetraploid populations. Fifteen populations of D. gypsophilum were sampled (Table 1). Leaf material from these populations was sampled in April and early May, 1999. Two or three leaves were removed from up to 15 individuals per population. These leaves were wrapped in dry paper towels, placed in labeled plastic sandwich bags and shipped on ice overnight to the University of Arizona (UAZ) for flow cytometry analyses following Galbraith et al. (1997). Briefly, protocols for an arc lamp-based flow cytometer (Partec CCAII, Partec GMbH, Munster, Germany) were used with the Galbraith Homogenization Buffer I and DAPI fluorescent stain. Fluorescent microbeads (Alignflow, Molecu- lar Probes, Inc., Eugene, OR) were used to align the instrument, and then samples of Nicotiana ta- bacum cy. Xanthi were run to set up the instrument. Leaf material from known 4n D. gypsophilum in- dividuals grown at Washington State University (WSU) were sent to UAZ to use as Delphinium standards. The samples of known ploidy (either the Nicotiana or 4n Delphinium) were also run at the beginning of each day and rerun at intervals during the day to ensure that the alignment had not drifted. Each sample was run for 10,000—30,000 events. Flow cytometry is an indirect measure of ploidy and is an effective and efficient technique for esti- mating the ploidy of natural populations (e.g., Bur- ton and Husband 1999; Greilhuber and Obermayer 1999; Husband and Schemske 1998; Keeler 1992; Thompson et al. 1997). Samples from three addi- tional populations of D. gypsophilum were main- tained in greenhouse culture at WSU. These pop- ulations were started in September, 1996, from seed collected in April and May, 1996, from natural pop- ulations (Table 1). Mitotic root-tip squashes follow- ing Soltis (1980) were performed to determine the ploidy of these populations. Root tips were har- vested from five actively growing plants per pop- ulation. Voucher specimens from each population were collected and are deposited in the Marion 92 San Joaquin Alameda e - Santa Clara Fic. 1. MADRONO Merced Fresno General location of D. gypsophilum populations and their ploidy. Inset: Map of California with counties shaded. Populations from Table | are indicated by a connecting line to the ploidy of the samples: 2 = diploid population; 2* = 2n and 3n individuals present; 3 = triploid population (but based on only | sample); 4 = tetraploid population; 4* = 4n and 3n individuals present; M = 2n and 4n individuals present; M* = 2n, 3n, and 4n individuals present; open circles = 4n populations surveyed by Lewis et al. (1951); closed circles = 2n populations surveyed by Lewis et allo): Ownbey Herbarium (WS). Ploidal levels of sam- pled populations are listed in Table 1. Figure 1 shows the relative locations of the populations stud- ied. We investigated the similarities among extant populations of D. gypsophilum using 12 allozyme loci reported in a previous study (Koontz 2000). These data were originally used to test the hypoth- esis of hybrid origin of D. gypsophilum (Lewis and Epling 1959). The analyses of allozyme data re- ported here are original. We computed genetic iden- tities (Nei 1972) that were then subjected to clus- tering analysis (UPGMA) using BIOSYS-1 (Swof- ford and Selander 1989) to explore which popula- tions were more similar to one another, to determine if the number of origins of tetraploid D. gypsophilum could be inferred, and if more than one diploid population was involved. Mode of inheritance in the tetraploid cytotype. We tested inheritance patterns by crossing cultivat- ed plants of known allozyme genotype to generate progeny arrays. The crossing design, number of progeny scored, and allozyme genotypes are listed in Table 2. Controlled crosses were performed in the Steffen Center Greenhouses, WSU, during Feb- [Vol. 48 | Fi \ \ i i \ | % | 1 i} Ny } i i 2001] KOONTZ AND SOLTIS: POLYPLOIDY IN D. GYPSOPHILUM 93 TABLE 2. CROSSING DESIGN FOR D. GYPSOPHILUM TO TEST FOR TETRASOMIC INHERITANCE. No. of Source population progeny Cross (individual) Locus Cross genotype scored la O’Neill (30) X O’ Neill (42) Pgm bbbc X bbbc GA lb Aat-1 bbbc X bbbb 84 2, O’Neill (35) X O’Neill (37) Lap aacc X aaaa 81 3a Los Banos (26) X O’Neill (30) Lap aacc X aaaa 88 3b Aat-1 bbbb X bbbc 82 4a Los Banos (9) X Los Banos (11) Pgm bbbd X bbbb 51 4b Lap aacc X aaaa a7 ruary—March, 1997. Seven plants were selected based on their previously determined allozyme ge- notype (Table 2). For each cross, 10 flowers per plant were cross-pollinated in the following man- ner. Flowers were emasculated just prior to open- ing. The removal of the anthers triggered the stig- matic surface to become receptive two days later, when pollen was transferred to the stigma. Each cross was performed reciprocally. Fruits were har- vested at maturity, just before or at dehiscence (ap- proximately 3—4 weeks after pollination). The fruits and seeds were stored in paper coin envelopes at room temperature. Previous work on this species indicated that the seeds remain viable for many years, but only germinate when planted in the early fall (JAK pers. obs.). The seeds were planted in plastic flats using regular potting soil in September, 1999. As soon as the seedlings had produced their first true leaves, they were harvested for allozyme electrophoresis. Allozyme procedures followed Soltis et al. (1983), with the exceptions listed below. Up to 96 individuals per cross were harvested. The fresh leaves were ground, and the wicks were frozen as described by Cook and Soltis (1999). All starch gels were 12.5% (w/v). Buffer system 6 was used to resolve aspartate aminotransferase (AAT). Sys- tem 8, as modified by Haufler (1985), was used to resolve leucine aminopeptidase (LAP) (stain recipe in McDonald 1985). The morpholine system (Odr- zykoski and Gottheb 1984) at pH 6.1 was used to resolve phosphoglucomutase (PGM). Isozymes were numbered sequentially starting with the most anodal as /. Alleles were designated alphabetically, the most anodal as a. A x’ test of significance was used to determine if the frequency of progeny genotypes deviated from expected ratios of disomic-digenic and tetra- somic inheritance. RESULTS AND DISCUSSION Geographic distribution of cytotypes. Although 15 individuals were sampled from each population for flow cytometry, the leaf material for some in- dividuals was unusable when the flow cytometry was conducted (Table 1). Low sample sizes for some populations may therefore fail to reflect the proportions of diploid and tetraploid individuals ac- curately. Populations were either 2n, 4n, or of mixed ploi- dy (Table 1), and 18 individuals (approximately 20% of the 92 samples cytotyped) were interpreted from the flow cytometric data as 3n (in population #16, the only individual was 3n). In an abstract, Lewis (1946) reported the occurrence of natural triploid hybrid individuals in areas of contact be- tween diploid and tetraploid cytotypes of D. gyp- sophilum. However, Lewis et al. (1951) sampled between | and 48 individuals per population (mean = 10, standard deviation = 12) from 35 popula- tions, but they did not detect any 3n individuals or populations of mixed ploidy in that study. The high frequency of triploids we detected by flow cyto- metry is surprising given that Lewis et al. (1951) detected no triploid individuals with broader sam- pling. Other studies have detected triploids, but at lower frequencies [e.g., 1.4% in Heuchera grossu- lariifolia Rydb. (Thompson et al. 1997), 9% in Chamerion angustifolium (L.) Holub (Husband and Schemske 1998), and 11% in Galax urceolata (Poin) Brumintt (Burton and Husband 1999)]. The coefficients of variation for the Delphinium samples measured on the flow cytometer ranged from 3.3 to 41, the average being 10.5 + 0.74 (SEM). The CVs were high because some of the field-collected sam- ples had started to degrade and were therefore less than optimal for flow cytometry. In assigning ploidy to the samples, the values set for each ploidy class were arbitrary, but were cal- ibrated on the values obtained from the known 4n samples. To determine the effect of changing the boundaries of the 2” and 4n ploidy classes on the number of 3” samples inferred, we broadened the range of 2m and 4n classes by 10%, but this change only reduced the triploid frequency to 15%. Given the discrepancy in the number of 3n individuals ob- served here and by Lewis et al. (1951), future work is needed using both flow cytometry and mitotic or meiotic squashes. Geographic structure among related diploids and polyploids has been commonly reported (e.g., Hus- band and Schemske 1998; Ness et al. 1989; Soltis 1984). Polyploids often have broader ecological amplitudes, in part due to their increased levels of 94 MADRONO genetic variation, that allow them to occupy habi- tats that are inhospitable to their diploid progeni- tors. This structuring leads to the successful estab- lishment of the polyploid race by ensuring individ- uals of the same ploidy do not co-occur. The dis- tributions of the cytotypes reported here show little geographic structure. However, the tetraploid pop- ulations are clustered in Merced County and in cen- tral San Luis Obispo County to western Kern Coun- ty (Fig. 1). Diploid populations (containing no 3n or 4n individuals) appear on both sides of the Mon- terey-San Luis Obispo County line. Many of the populations analyzed by Lewis et al. (1951) no lon- ger exist, though attempts were made to locate them for use in this study. A tetraploid population from San Joaquin County was Lewis et al.’s (1951) northernmost sample, well separated from the other populations of D. gypsophilum they surveyed (Fig. 1). Despite several attempts to locate this popula- tion, it appears to have been destroyed; neverthe- less, the historical presence of this population makes the tetraploid populations discovered in Merced County less isolated. The data from Lewis et al. (1951) indicate that the tetraploids generally clustered around southwestern Fresno County into Kings County, San Luis Obispo County, and in both western and central Kern County. Other species that exist as both diploid and poly- ploid populations often display broad geographic structuring of ploidy, but these species all have a much larger geographic range than D. gypsophilum. For example, the tetraploid cytotypes of Heuchera micrantha Doug]. ex Lindl. occur in the central part of the range, with diploid populations occurring to the north and south (Ness et al. 1989). A distinct north-south distribution is found in Tolmiea men- ziesii (Pursh) Torn & Groy, in which the tetraploid cytotype occurs from southeastern Alaska to central Oregon and the diploid cytotype occurs from cen- tral Oregon into northern California (Soltis 1984). Diploid and tetraploid cytotypes of Chamerion an- gustifolium are also distributed latitudinally, with the diploids occurring at higher latitudes (Husband and Schemske 1998). The geographic structure of cytotypes of Galax urceolata is less defined be- cause the diploid and polyploid cytotypes overlap; in general, the frequency of diploids decreases north to south, while tetraploids increase (Burton and Husband 1998). The diploid cytotype of Heu- chera grossulariifolia occurs throughout river sys- tems in Idaho and western Montana, but the tetra- ploids are more limited in distribution across north- central Idaho into western Montana (Segraves et al. 1999; Wolf et al. 1990). Multiple origins of polyploid species and cyto- types have been detected in almost all cases that have been investigated (reviewed in Soltis et al. 1992; D. Soltis and P. Soltis 1993, 1999; P. Soltis and D. Soltis 2000). Some of the tetraploid popu- lations in this study cluster geographically with one or more diploid populations (Fig. 1, pops. 1 and 2; [Vol. 48, 5, 6, and 9; 11 and 12). Additionally, some popu-! lations are of mixed ploidy, containing both dip-, loids and tetraploids, as well as some triploids (Fig. 1). These distributions suggest the possibility of: multiple origins of the tetraploid cytotype from dif- ferent diploid progenitor populations. | In previous work, allozyme analyses of multiple populations of D. gypsophilum (Koontz 2000; raw | data available from first author by request) indicate | few differences among tetraploid populations, and DNA sequence divergence in the nuclear ribosomal internal transcribed spacer (ITS) regions between the two diploids and one tetraploid sampled is low | (~0.17%) (Koontz 2000). Comparisons of the al-— leles reported in Koontz (2000) found that neigh- boring diploid and tetraploid populations e.g., pops. | 1 and 2; 5, 6, 7, 8, and 9; and 11 and 12. show. similar allele frequencies at most loci. The genetic , identities (Nei 1972) among the populations sam- | pled are high (~88—99%); however, a UPGMA © clustering diagram reveals two groups (Fig. 2), one : composed mainly of the populations found in east- | ern San Luis Obispo County, eastern Monterey County, and easternmost Kern County, and the oth- | er composed of those populations in western Kern | County, northwestern San Luis Obispo County, | southwestern Monterey County, and Merced Coun- ty. Tetraploids and diploids occur in both clusters, | and those that occur together geographically [i.e., pops. 1 (3n/4n) and 2 (2n/3n), 6 (3n/4n) and 9 (2n/ | 3n), 11 (3n/4n) and 12 (2n/3n)] generally occur in | the same group in the UPGMA phenogram (Fig. 2), consistent with recurrent formation of tetraploid — populations from neighboring diploid populations. — Other populations from the same geographic area occur in separate groups. Population 7 (2n/4n) from — the Elk Hills area of Kern County occurs in a clus- ter separate from other populations from this area [pops. 5 (4n), 6 (3n/4n), and 9 (2n/3n), where pop. | 5 is more similar to pops. | and 2 than to pops. 6 and 9]. Both populations 3 and 7 contain mixed cytotypes, suggesting that the 4n cytotype could have arisen within each of these two populations. These data do not provide conclusive evidence of specific progenitor-derivative relationships; how- ever, they are consistent with more than a single origin of the tetraploid cytotype. To test the hy- pothesis of multiple origins of polyploidy thor- oughly, additional populations will need to be sam- pled both cytologically and genetically. Segregation analyses. Although no differences in seed set were observed between reciprocal crosses, the seed produced from parent individuals 9 and 11 (crosses 4a and 4b) had low germination, and only 68 progeny were harvested. The numbers of prog- eny scored for each cross are lower than the total harvested (Table 2) because some individuals did not express well and could not be scored with con- fidence. Allopolyploids are characterized by fixed hetero- KOONTZ AND SOLTIS: POLYPLOIDY IN D. GYPSOPHILUM 75) Kern, 3n/4n Kern, 2n/3n 1 pi 5 Kern, 4n 6 Kern, 3n/4n Q Kern, 2n/3n 13 SLO, 2n 17 Merced, 4n 14 SLO, 2n 15 Mont., 2n 3 SLO, 2n/3n/4n 1] SLO, 3n/4n 16 Mont., 3n 12 SLO, 2n/3n 7 Kern, 2n/4n 20 Kern, 2n aeeaahaustaanaaaeia aes ot 0.87 0.90 Os: Fic. 2. O97 1.00 UPGMA phenogram of genetic identity (Nei 1972) for the populations of D. gypsophilum, their county, and ploidy. SLO = San Luis Obispo County; Mont. = Monterey County. zygosity at allozyme loci (e.g., Roose and Gottlieb 1976). Fixed heterozygosity was not observed at any allozyme loci during this or previous work on D. gypsophilum (Koontz 2000). At several loci (e.g., Pgi, Aat, Pgm, Lap) the tetraploids exhibited unbalanced heterozygotes, indicating dosage effects that may result from either tetrasomic segregation or disomic-digenic segregation with shared alleles at the two loci. Genotypes from progeny arrays were compared with expectations for both tetraso- mic and disomic-digenic segregation. Crosses la-b, 3b, and 4a (Table 2) yielded prog- eny arrays that were consistent with both disomic- digenic and tetrasomic inheritance models; there- fore, x? scores could not distinguish between the two models. For crosses 2, 3a, and 4b (Table 3), progeny were obtained that could only be expected under the tetrasomic model; however, ratios of the observed progeny did not fit the expected tetraso- mic ratios (Table 3). Cross 2 had one progeny with an unexpected genotype of cccc. These crosses also had a higher proportion of aaaa (2), aacc (3a), or both aacc and aaaa (4b) genotypes than expected, and crosses 2 and. 4b also had fewer aaac progeny than expected. The progeny arrays from the crossing experi- TABLE 3. EXAMPLES OF THE EXPECTED AND OBSERVED PROGENY FREQUENCIES UNDER DISOMIC-DIGENIC AND TETRASOMIC MODELS OF INHERITANCE FOR CROSSES 2, 3A, AND 4B. All are Lap aacc X dada. aa,ac = genotype aa at one disomic locus and ac at the second disomic locus. NA = genotypes present that are not possible under the given model, making a significance test not appropriate. * The cccc genotype was not included in the x? computation, but it would actually make the value ‘“‘NA’’ for the tetrasomic model. Progeny aa,cc X dc,ac X genotype daa,aa daa,aa Tetrasomic daaaa 0.5 0 0.167 aaac 0 ] 0.666 aacc 0.5 0 0.167 cCccc 0) 0) 0 x’ NA NA see column under each cross 2 3a 4b Observed Observed Observed 0.516 0.159 0.474 0.234 0.5 0.193 0.247 0.341 0:333 0.012 0) O 82.54* 19.718 41.974 P < 0.0005 P < 0.0005 P < 0.0005 96 ments do not offer a clear answer for the mode of inheritance in the tetraploid cytotype. In some cases, the crosses could not distinguish between tet- rasomic and disomic-digenic inheritance; these progeny arrays are therefore consistent with both models. In other crosses, the disomic-digenic model could be ruled out because multiple genotypes were recovered that were impossible under the disomic- digenic model without invoking a high frequency of chromatid segregation. However, these same crosses did not statistically fit the tetrasomic model, and one genotype that was not expected under ei- ther model appeared in the progeny of one cross. The occurrence of a novel genotype in low fre- quency in the progeny of a known cross may rea- sonably be attributed to chromatid segregation (re- viewed in Wolf et al. 1989). Additionally, gametic selection has been implicated where progeny arrays derived from the same parents alternately fit per- fectly or deviate significantly from expectations when produced and grown in different environ- ments (e.g., Henningsen et al. 1993). Future work should address the possible role of gametic selec- tion. CONCLUSIONS Using flow cytometry and root-tip preparations, we mapped the cytotypes of D. gypsophilum from throughout its current known range. Unlike the pre- vious study (Lewis et al. 1951), we detected mixed ploidy within some populations, as well as triploid individuals. The evidence from allozyme data and the segregation analyses presented here point to tet- raploid D. gypsophilum as an autotetraploid. Al- though diploid and tetraploid populations cluster to- gether both geographically and genetically, sug- gesting recurrent formation of the tetraploid cyto- type, the genetic data do not provide conclusive evidence of specific progenitor-derivative relation- ships among populations. Conservation implications. The data presented here and elsewhere (Koontz 2000) do not support the subdivision of D. gypsophilum into two subspe- cies. Populations 13, 14, and 15 (all 2m) are from localities that Warnock (pers. comm.) has identified as subsp. parviflorum. Population 13 1s more sim- ilar genetically to populations of subsp. gypsophil- um (Fig. 2, populations 9, 6, and 17;) than to pop- ulations 14 and 15. Furthermore, ploidy does not distinguish the two subspecies. Although all of the tetraploids detected occur in D. gypsophilum subsp. gypsophilum (Lewis and Epling 1954; this study), five 2n populations are also recognized as subsp. gypsophilum. The combination of genetic and ploidy data sug- gest that two subspecies should no longer be rec- ognized, even though the subspecies may be distin- guished to some degree by range and flower color (but not size). Both subspecies are currently placed on the CNPS List 4. Because the range of D. gyp- MADRONO [Vol. 48 | sophilum sensu lato continues to be affected by hu- man development, D. gypsophilum should remain a | List 4 species. Both cytotypes and all morphologi- , cal variants should be included in any future con- | servation efforts for this species; the cytotypes should be managed separately, as they represent po- | tentially different evolutionary units. ACKNOWLEDGMENTS This study was supported by grants from the B. W. | Higinbotham Trust to the Department of Botany, WSU; a | G. R. Hardman Native Plant Award; and a WSU College , of Sciences Minigrant award, all to JAK. Many thanks to | G. Lambert (UAZ) for conducting and troubleshooting the | flow cytometry analyses, and D. Galbraith (UAZ) for al- lowing the analyses to be conducted in his lab. C. Cody | maintained the D. gypsophilum plants at WSU. L. Pater- | son and K. Brown, California Department of Water Re- sources, aided in collection of plant material along DWR aqueducts. E. Cypher, Endangered Species Recovery Pro- gram, helped locate and sample many other populations. Thanks also to Occidental Petroleum Corporation for al- lowing JAK and EC onto the Elk Hills Naval Petroleum Reserve to sample D. gypsophilum populations there. JAK also thanks M. Olson, Univ. of Alaska, Fairbanks, for dis- { | cussions on tetrasomic inheritance, D. Soltis for help and | advice on root-tip preparations, J. Thompson for editorial | help, and Frank Hutto for map-drawing assistance. This © work represents a portion of the Ph.D. dissertation by © JAK. LITERATURE CITED Burton, T. L. AND B. C. HUSBAND. 1999. Population cy- totype structure in the polyploid Galax urceolata (Diapensiaceae). Heredity 82:381—390. Cook, L. M. AND P. S. Sottis. 1999. Mating systems of diploid and allotetraploid populations of Tragopogon (Asteraceae). I. Natural populations. Heredity 82: 237-244. AND P. S. SoLtis. 2000. Mating systems of diploid and allotetraploid populations of Tragopogon (Aster- aceae) II. Artificial populations. Heredity 84:410-— 415. GALBRAITH, D. W. 1989. Analysis of higher plants by flow cytometry and cell sorting. International Review of Cytology 116:165—228. , G. M. LAMBERT, J. MACAS, AND J. DOLEZEL. 1997. Analysis of nuclear DNA content and ploidy in high- er plants. Jn J. P. Robinson, Z. Darzynkiewicz, P. N. Dean, A. Orfao, P. S. Rabinovitch, H. J. Tanke, and L. L. Wheeless, (eds.), Current Protocols in Cyto- metry, 7.6.1-7.6.22. John Wiley & Sons, New York, NY. GRANT, V. 1981. Plant Speciation, 2nd ed. Columbia Uni- versity Press, New York, NY. GREILHUBER, J. AND R. OBERMAYER. 1999. Cryptopolyplo- idy in Bunias (Brassicaceae) revisted—A flow-cyto- metric and densitometric study. Plant Systematics and Evolution 218:1—4. HAvuFLeR, C. H. 1985. Enzyme variability and modes of evolution in Bommeria (Pteridaceae). Systematic Bot- any 10:92—104. HENNINGSEN, K. W., J. E. BOYNTON, AND D. VON WETT- STEIN. 1993. Mutants at xantha and albina Loci in Relation to Chloroplast Biogenesis in Barley (Hor- 2001] TETSBAND, B. C. AND D. W. SCHEMSKE. i deum vulgare L.). Biologiske Skrifter 42. The Royal Danish Academy of Sciences and Letters, Copenha- gen. 1998. Cytotype distribution at a diploid-tetraploid contact zone in Chamerion (Epilobium) angustifolium (Onagraceae). American Journal of Botany 85:1688—1694. KEELER, K. H. 1992. Local polyploid variation in the na- tive prairie grass Andropogon gerardii. American Journal of Botany 79:1229—1232. Koontz, J. A. 2000. Tests of hybrid origin of two Cali- fornian Larkspurs (Delphinium, Ranunculaceae). Ph.D. dissertation. Washington State University, Pull- man, WA. Lewis, H. 1946. Polyploidy in the Californian Delphini- ums. American Journal of Botany 33:842. (abstract) AND C. EPLING. 1954. A taxonomic study of Cal- ifornian Delphiniums. Brittonia 8:1—22. AND . 1959. Delphinium gypsophilum, a diploid species of hybrid origin. Evolution 13:511- 25: : , G. A. L. MEHLQUIST, AND C. G. WYCK- oFF. 1951. Chromosome numbers of Californian Del- phiniums and their geographical occurrence. Annals of the Missouri Botanical Garden 38:101—117. McDonaLp, J. H. 1985. No Bad Gels: Starch Gel Elec- trophoresis for the Masses. State University of New York, Stony Brook, NY. Nel, M. 1972. Genetic distance between populations. American Naturalist 106:283—292. Ngss, B. D., D. E. SOLTIS, AND P. S. SOLTIS. 1989. Auto- polyploidy in Heuchera micrantha (Saxifragaceae). American Journal of Botany 76:614—626. ODRZYKOSKI, I. J. AND L. D. GOTTLIEB. 1984. Duplications of genes coding 6-phosphogluconate dehydrogenase in Clarkia (Onagraceae) and their phylogenetic im- plications. Systematic Botany 9:479—489. Roose, M. L. AND L. D. GorTTLiesB. 1976. Genetic and biochemical consequences of polyploidy in Trago- pogon. Evolution 30:818—830. SEGRAVES, K. A., J. N. THompson, P. S. SOLTIS, AND D. E. SoLTis. 1999. Multiple origins of polyploidy and the geographic structure of Heuchera grossulariifolia. Molecular Ecology 8:253-262. SoOLTis, D. E. 1980. Karyotypic relationships among spe- cies of Boykinia, Heuchera, Mitella, Sullivantia, Ti- arella, and Tolmiea (Saxifragaceae). Systematic Bot- any 5:17-29. . 1984. Autopolyploidy in Tolmiea menziesii (Sax- KOONTZ AND SOLTIS: POLYPLOIDY IN D. GYPSOPHILUM oT, ifragaceae). American Journal of Botany 71:1171— 1174. , C. H. HAUFLER, G. J. GASTONY, AND D. C. DAR- ROW. 1983. Starch gel electrophoresis of ferns: a com- pilation of grinding buffers, gel and electrode buffers, and staining schedules. American Fern Journal 73:9— Dike AND P. S. Sortis. 1993. Molecular data and the dynamic nature of polyploidy. Critical Reviews in Plant Sciences 12:243-—273. AND . 1999. Polyploidy: recurrent forma- tion and genome evolution. Trends in Ecology and Evolution 14:348—352. SoLTis, P. S., J. J. DOYLE, AND D. E. SOLTis. 1992. Mo- lecular data and polyploid evolution in plants. Jn P. S. Soltis, J. J. Doyle, and D. E. Soltis (eds.) Molec- ular Systematics of Plants, pp. 177-201. Chapman and Hall, New York, NY. AND D. E. SOLTIsS. 2000. The role of genetic and genomic attributes in the success of polyploids. Pro- ceedings of the National Academy of Sciences USA 97:7051-—7057. SWOFFORD, D. L., AND R. K. SELANDER. 1989. BIOSYS- 1, release 1.7. Hlinois Natural History Survey, Cham- paign, IL. THOMPSON, J. N., B. M. CUNNINGHAM, K. A. SEGRAVES, D. M. ALTHOFF, AND D. WAGNER. 1997. Plant polyploidy and insect/plant interactions. American Naturalist 150:730-743. VIDA, G. 1976. The role of polyploidy in evolution. Jn V. J. A. Novak and I. Pacltova (eds.), Evolutionary Bi- ology, 267-304. Czechoslovak Academy of Sciences, Prague. WARNOCK, M. J. 1995. Documented chromosome reports and supplementary information on previously report- ed chromosome counts of North American Delphin- ium. Phytologia 78:102—118. . 1997. Delphinium. In Flora of North America Ed- itorial Committee (ed.), Flora of North America North of Mexico, Vol. 3, 196-240. Oxford University Press, New York, NY. WoLF, P. G, P. S. SOLTIS, AND D. E. SOLTIS. 1989. Tetra- somic inheritance and chromosome pairing behaviour in the naturally occurring autotetraploid Heuchera grossulariifolia (Saxifragaceae). Genome 32:655— 659. , D. E. SOLTIS, AND P. S. Sottis. 1990. Chloroplast- DNA and allozymic variation in diploid and autote- traploid Heuchera grossulariifolia (Saxifragaceae). American Journal of Botany 77:232—244. MADRONO, Vol. 48, No. 2, pp. 98-111, 2001 SPRING-FED PLANT COMMUNITIES OF CALIFORNIA’S EAST BAY HILLS OAK WOODLANDS BARBARA ALLEN-DIAZ! AND RANDALL D. JACKSON Ecosystem Sciences Division, University of California, Berkeley, 151 Hilgard Hall, Berkeley, CA 94720-3110 CATHERINE PHILLIPS Pacific Southwest Research Station, Stanislaus National Forest, USDA Forest Service, 19777 Greenley Rd. Sonora, CA 95370 ABSTRACT Sixty-eight spring ecosystems containing 207 plant species were sampled in 1991 and 1992 on the | East Bay Hills of Alameda and Contra Costa Counties, California. Using TWINSPAN, we identified and ' described four plant community types: 1) Ryegrass-Herbaceous, 2) Rush-Herbaceous, 3) Willow-Poison | oak, and 4) California bay-Poison oak. Spring types were related to environmental gradients and a binary disturbance variable (livestock grazing presence/absence) using canonical correspondence analysis (CCA). Extraction of 2 CCA axes proved useful in relating spring types to productivity (using litter as proxy) and elevation gradients. A third CCA axis indicated that the presence or absence of grazing was somewhat helpful in discriminating among plant community types although gradients found with detrended corre- spondence analysis (DCA), which uses only the species matrix, were much stronger than could be gen- erated by combining measured environmental and management factors (1.e., CCA). While community type, which was defined by taxonomic abundance, was apparently independent of grazing presence/ absence, the grazing axis (CCA;) was effective in dispersing willow size classes such that overstory willows scored higher on the ungrazed end and shrub willows scored higher on the grazed end of CCA;. We conclude that grazing can affect vegetation structure of some spring-fed plant communities, but that composition is controlled by site variables. Classification of California vegetation has a long and varied history. General descriptions of major vegetation types can be found in Jepson (1975), Munz and Keck (1973), as well as Ornduff (1974). More recently, The Jepson Manual (Hickman 1993) described major climatic and geographic zones in which plants can be found in California. Sawyer and Keeler-Wolf (1995) described California plant communities based on quantitative analysis of field plot data, or relied heavily on Holland’s (1986) qualitative descriptions of communities if plot data were unavailable. Many vegetation classification systems have been developed for specific purposes. For example, oak woodlands were described to subseries to pro- vide baseline, quantitatively-based descriptive in- formation for the Integrated Hardwood Range Man- agement Program (Allen et al. 1989). Ferren et al. (1994) provided a detailed framework for classifi- cation of wetlands of the central and south coast of California for the Environmental Protection Agen- cy. Gordon and White (1994), Fites (1993), Smith (1994) and other USDA Forest Service ecologists have intensively sampled plant community diver- sity on their forests to provide fundamental infor- mation for improving communication between di- verse resource specialists as well as providing eco- ' E-mail: ballen@nature.berkeley.edu logical information on potential community re- sponses to management. The oak woodlands of California have been of particular interest to ecologists and managers over the past 15 years. Political awareness arose out of perceived threats to the oak woodland from urban development, wood cutting, lack of oak regenera- tion and recruitment, and livestock grazing (Muick and Bartolome 1987; Bolsinger 1988). In a political response, a research, extension, and management program was forged; the Integrated Hardwood Range Management Program (IHRMP) was devel- oped to direct and fund research and extension ac- tivities in California’s oak woodlands (Passof and Bartolome 1985). The IHRMP has supported a number of studies on oak ecology, oak woodland ecology and management, wildlife, invertebrates, regeneration and recently water quality (Swiecki and Bernhardt 1991; Allen-Diaz and Holzman 1993; Davis et al. 1995; Campbell and Allen-Diaz 1997; Allen-Diaz et al. 1998; Allen-Diaz and Jack- son 2000). A concern of many government water quality agencies and private conservation groups are live- stock effects on wetland and riparian systems. Pre- dicting the effects of livestock grazing on riparian areas such as spring-fed wetlands remains elusive (Clary 1995; Larsen et al. 1998; Belsky et al. 1999; Clary 1999; Rinne 1999; Allen-Diaz and Jackson 2000). These systems respond to grazing treatments in ways that do not always correspond with anec- dotal and observational evidence (Kauffman and Krueger 1984). Disparity among results may lie in: 1) the confounding of grazing histories and other land-uses both past and present, 2) the application of grazing treatment levels that are not reflective of actual livestock use, 3) ambiguous or vague grazing system definitions, intensities, and seasons, 4) char- acteristics inherent to riparian ecosystems (i.e., non- equilibrium or chaotic dynamics), and/or 5) aver- aging of variability from site-specific responses. It follows that adequate description and understand- ing of riparian areas is lacking. Nonetheless, management mandates abound. The 1987 renewal of the 1977 Clean Water Act amend- ment to the Federal Water Pollution Control Act of 1972 (PL 92-500; 33 U.S.C. s/s 1251 et seq.) shift- ed nonpoint-source pollution control emphasis from “direct threat to human health or safety”’ to “‘threat to ecosystems and habitat”? (Sec 319). Revision of the Coastal Zone Management Act of 1992 (as amended by PL 92-583; 16 U.S.C. 1451 et seq.) required states to develop and implement nonpoint- source pollution programs and establish manage- ment measures for implementation. Moyle et al. (1996) contains appendices listing riparian protec- tion guidelines and prescriptions on federal and pri- vate lands. The East Bay Regional Park District (EBRPD) has recently conducted hearings to collect public 2001] ALLEN-DIAZ ET AL.: SPRING-FED PLANT COMMUNITIES he) | Bose Poy Tose RRR RRR xX MERE es aie | Boon eer Ee CN eg aD Pc: BLACK _ BRIONES DIAMOND WILDCA “dp MORGAN TERRITORY 7 Yi Mi a MISSION PEAK Fic. |. East Bay Regional Parks and East Bay Municipal Utility District areas where springs were located. and expert input to guide future management on park landscapes. Hence, we sought coarse-scale patterns that might indicate whether community types at spring-fed ecosystems were influenced by the presence or absence of livestock grazing. To- ward this end, we described and classified spring- fed plant community types on the hardwood range- lands of California’s East Bay Hills and then related these vegetation types to environmental and man- agement factors. These results should inform future survey stratification and experimental design. STUDY SITES The East Bay Municipal Utility District (EB- MUD) owns and manages approximately 11,330 ha in the East Bay (Fig. 1). EBMUD’s reservoirs store high-quality drinking water for approximately 1.2 million users. Domestic livestock grazing has been a significant component on these rangelands for at least 100 years. At the time of this study, cattle were managed on 7285 ha, with 22,000 AUM’s (animal unit months) grazed annually (EBMUD oS): The EBRPD manages 36,834 ha (20,457 ha in Alameda County; 16,377 ha in Contra Costa Coun- ty) in 50 regional parks (EBRPD 1996). In 1992, about 15,785 ha were leased to grazing cattle for an authorized use of 24,000 AUM’s (EBRPD 1992). Ungrazed spring sites were determined by 100 MADRONO TABLE 1. of sites. Site Grazed EBRPD Black Diamond Briones Chabot Diablo Foothills Garin Las Trampas Mission Peak Morgan Territory Ohlone Sunol Wauhab Wildcat Canyon EBMUD North South Totals ONNADWNRFK ONN — GN — A Be assessing their physical accessibility (fences, shrub cover, topography) and evidence of livestock ab- sence (no fecal material, untrampled soil, ungrazed vegetation). Specific dates of livestock exclusion were not available, but all fencing material had ap- parently aged substantially and exclusion by shrubs and topography would have existed for decades to millennia (or at least as long as the development of the spring itself). The goals of the EBMUD and EBRPD livestock grazing programs are to manage livestock in order to maintain and enhance the health of the grassland ecosystem, remove fuels where fire poses a signif- icant hazard, maintain a healthy agricultural econ- omy where consistent with other district goals, pro- tect and enhance water quality, and generate reve- nue (EBRPD 1992, EBMUD 1995). Spring head 10-m diameter shrub and overstory sampling plot 10-m line-point herbaceous vegetation transects Fic. 2. Spring site sampling schematic. EAST BAY REGIONAL PARK DISTRICT AND EAST BAY MUNICIPAL UTILITY DISTRICT SAMPLING SITES. ! Numben Ungrazed 'Means of grazing exclusion 0 0 1 fence 0 l fence 2 topography (2) 2 fence (2) 1 fence 0 3 fence (2), topography (1) 0 2 fence (1), topography (1) 1 fence | 13) | METHODS Data were collected March through June of 1991 - and 1992 from 68 (55 grazed and 13 ungrazed) | spring sites (Table 1). Sites were found by locating | them on 7.5” quadrangle topographic maps. Sites | were then visited and selected if vegetation indi- | cated an active spring and no recent disturbance to | develop the spring for livestock use was evident. | Sampled sites included those maintaining old weir | boxes (indicating past development) but exhibiting | established vegetation, as well as springs that had | apparently never been developed. | A sharp ecotone was present between spring vegetation and surrounding upland areas, which al- | lowed for point-source designation along the up- | slope boundary, i.e., the point from which water | flowed. Where weir boxes had been previously in- | stalled, the downslope side of the spring box or. pond was considered the point-source. | Elevation, aspect, slope, slope position, and soil | parameters were estimated for each site. Vegetation — data were collected from two 10-m point-intercept | transects (Bonham 1989) run parallel to streamflow — and emanating from a randomly located point on > either side of the spring head but confined to the “green zone’’ delineating spring vegetation (Fig. 2). Herbaceous plant species that were intercepted by a sharpened point lowered at 10 cm intervals along each transect were recorded for a total of 200 point-observations per site. A 5-m radius circular plot centered on the spring head was used to ocu- larly estimate relative cover of trees and shrubs. Bias and measurement error were minimized by en- suring that all plots in each year were estimated by a single individual (Elzinga et al. 1998). The dif- ficulty of ocular estimation of cover increases with the number of taxa present (Elzinga et al. 1998), 1) \ i 2001] ‘however most of our plots were dominated by only ‘two or three tree and/or shrub species. Fine-scale differences in these vegetation layers by species combinations were not instrumental in the subse- quent classification. Unknown species were col- lected and keyed to species using Munz and Keck (1973) and then updated according to Hickman 1(1993). | We delineated plant communities by subjecting ‘the combined herbaceous, shrub, and tree layer ‘cover data to the classification software TWINSPAN \(PC-ORD version 4, Hill 1979; McCune and Mef- ford 1999). TWINSPAN uses cover classes delimited according to cut-levels that specify class ranges. ‘Default cut-levels were used resulting in absolute ‘cover classes of >0 to 2%, 3 to 5%, 6 to 10%, 11 ‘to 20%, and >20%. TWINSPAN uses each cover class 'X species combination to create pseudospecies, e.g., Bromus hordeaceus L. 6—10% is considered a dif- ferent taxon than B. hordeaceus 11-20%. Pseudos- pecies are then used to drive a divisive classifica- ‘tion, each level of which is the result of bifurcating groups produced by previous divisions. The relative ‘strength of a division, hence the resultant 2 groups, was denoted by an eigenvalue (A) showing increas- ‘ing strength from 0.00 to 1.00 (Gauch 1982). Ei- | genvalues approximate the percentage of pseudos- /pecies not common to each group, so, 4 = | de- notes 2 groups with no pseudospecies overlap (Jongman et al. 1995). Eigenvalues provide an ob- jective criterion for determining the merit of each ‘division, although, the \ at which further splits are ignored (the critical \) is a subjective choice de- pendent on the research question. We sought a rel- atively broad scale classification so X = 0.30 was used as our critical 2. Canonical correspondence analysis (CCA, PC- “ORD version 4, McCune and Mefford 1999; Ter Braak 1986; Ter Braak 1987) was employed to re- late environmental and management factors to the vegetative groupings determined with TWINSPAN. CCA finds the linear combination of these factors that maximizes species dispersion along an ordi- nation or canonical axis. As with TWINSPAN, the strength of this dispersion is indicated by an eigen- value (A). The correlation of environmental vari- ables to canonical axes was assessed by examining intraset correlations (Palmer 1993). The correlation matrix of environmental variables indicated no multicollinearity problems. Monte Carlo tests of significance were run with 99 iterations for each canonical axis. Three of the 68 sites were eliminated from the CCA analysis because they contained missing val- ues. Eight environmental variables were entered into the CCA. The variable ASPECT was created by taking the absolute value of 180 minus the azi- muth reading resulting in values ranging from O for south and 180 for north (sensu Stohlgren et al. 2000). Only 3 textural classes were evident at these sites: loam, clay loam, and clay. Hence, binary ALLEN-DIAZ ET AL.: SPRING-FED PLANT COMMUNITIES 101 dummy variables were created as LOAM or CLAY- LOAM with all other sites classified as CLAY. The presence or absence of livestock grazing was also coded as a binary dummy variable (GRAZED). The remaining variables were continuous: ELEVATION (m), SLOPE (%), LITTER (% cover of all dead organic matter), and BARE (% cover of bare ground). Finally, DCA (PC-ORD version 4, McCune and Mefford 1999) was performed on the species ma- trix to assess the ability of CCA-generated ordina- tion axes to depict important underlying gradients. DCA extracts ordination axes from the species ma- trix as does CCA, but DCA ordination is not con- strained by specified environmental variables. DCA simply maximizes species dispersion using a 2-way weighted averaging algorithm (Jongman et al. 1995). Hence, DCA provides some indication of the total amount of dispersion or variability in a species matrix, while CCA shows how a combi- nation of environmental or management variables can emulate this dispersion. If DCA and CCA gen- erate gradients of similar magnitude (1.e., Apc, ~ Acca) We would conclude that the environmental variables provide a well-specified model of gradi- ents in the species matrix. Conversely, an under- specified CCA model is one where the combination of environmental and management variables do not approximate gradients generated from the species matrix alone (1.e., Apc, => Acca). In practice, Apca will always be greater than Acca, SO Comparisons must remain qualitative and used in an exploratory manner as we have done here. A distinction must be made between using these ordination techniques for testing hypotheses (i.e., where manipulations of treatment variables are made) and for exploring structure or pattern in a dataset. A rigorous exper- imental design would include a priori grazing con- trasts that were randomly assigned to each site. In this case, a balanced design where equal numbers of grazed and ungrazed sites were sampled would be ideal. However, ours was a heuristic use of or- dination to search for possible relationships be- tween environmental and management variables and community types and to examine how sites were distributed along these gradients. Our unbal- anced design (more grazed than ungrazed sites) does not affect this technique because it was not an experimental design, but an exploratory analysis. RESULTS A total of 207 plant species were found on the springs, including 16 trees and 4 shrubs. Four oak species were found at the springs, coast live oak (Quercus agrifolia Nee), valley oak (Q. lobata Nee), blue oak (Q. douglasii Hook. & Arn.), and interior live oak (Q. wislizeni A. DC.). In addition, willow (Salix spp.) and California bay (Umbellu- laria californica [Hook. & Arn.] Nutt.), and some- times alder (Alnus rhombifolia Nutt.), California aqyeeiq i an a ee ee ee (4) MOy[IM Se eS Se ee ae SS OS (0) MoqIM = eS 2S 8a er Stee =e SNA Sy & BS ae TSP a ene Se 2 ee Se “SS (y) eo uostod = Sia ee sy Ee OR Sere ices Pe eee Se eS aS See eS (s) yeo uosiod Swe Sig Ss Rea te = (y) Auiaqyoe[q a a ec i (y) Aeq erusosteD PG Se ea lee ae SS ee Se (0) Aeq ermsoyie> Se es Oe ee ee Fae a (0) BO ddl] ySeOo 1st SG ser Sao CaaS rs er) sl et et hr ssoul Se eee Sy ee eS eee Cee eee es Gnas jreqas10y CSC See Se S58 Te. 6 eS eo SS ee See Se SS eora ana MADRONO NS ™N ! I Ns ‘ a ! > ~m N N con) ~m a ~m ~ sv N Ww fo] N a a ol a | ' | | ysmu peo} | d | N foal ' I ' N ' a — bs ' ~m ' N ! | ' =f t ' i) ' ' ' ! | | ssayo Yos Tee ssei3oAl N SSIISIOJEM ie fos) | c ws ~N N iS (re Ww ive) ve) N aed im) wo re fag} uw) N ~m ON ar wo wr Seat ive) a Mm a ~ MN a (og) wr Car Glo St ve) JamoyjAoaxuoul UOUILUOS N (0.3 emp] Ds (.@) ive) =! NS YsT poaro[-sil N N p Vie rags ee ysru 30q I |]oMpseds ueduaury Ses = ysni uoulw09 - - - [— - Z saspos exe va Lo: ot SD (S) ) oO ~ fon} (Ss) = c , ) N a N ) iS yeo uosiog-Aeq Bluse) Yeo UOSIOg-MOTITAA snoadeqsey-ysny snoaseqsay-ssessoAy <— add} Ajrunutuio N o = ot <- apoo ays LTS’ 8 1901 0 1103 3001 2603 1905 0 1681 1601 2 is). 1106 1101 8 1906 NOR OSHS 3001 3086 3084 308 1707 alata: 1114 2084 Tabs al ) CN 4 ‘snosoeqiey = (Y) ‘qniys = (Ss) ‘AIO}SIOAO = (0) ‘'d'T ‘SIOART UOTRIIBIA [BOTADA AQ POZIIOSIIVO Je VXBL “MHOT< (S) ‘%WOZ-I1 (+) ‘%OI-9 (€) *%S—-€ (Z) *%IZ-L (1) “at ‘sassejo 1aA09 sarsadsopnasd ayeorpul santeA “NOLLVOWISSWID NVdSNIM.L WO'd TEV OLLGONAS "7 ATAVE 102 { 2001] ALLEN-DIAZ ET AL.: SPRING-FED PLANT COMMUNITIES 103 | : TABLE 3. RYEGRASS-HERBACEOUS TYPE DESCRIPTIVE STATISTICS. Constancy equals the number of species occurrences/ ‘total number of plots; cover equals the mean cover (%) of a species for a given type; and range equals the range of cover values (T = trace or <1%). {| | ‘Characteristic plant taxa Con- stancy Cover Range Tree Cover (%) Shrub Cover (%) Grass Cover (%) Forb Cover (%) Abiotic environment Elevation (m) 49 (1-100) n = 14 13 (1-50) n = 13 43 (1-85) n = 36 13 (1-65) n = 36 376 (128-817) Lolium multiflorum Lam. ryegrass 94 12 T—44 Juncus bufonius L. toad rush 69 4 T-22 Rorippa nasturium-aqutica (L.) Hayek watercress 67 6 T-—39 Polypogon monspeliensis (L.) Desf. rabbitfoot grass 44 3 T-33 Bromus diandrus Roth ripgut brome 44 l T-8 Juncus xiphioides E. Meyer iris-leaved rush 39 4 T-31 Bromus hordeaceus L. soft chess 36 l T-7 Mimulus guttatus DC. common monkeyflower 31 2 T-29 Geranium molle L. common geranium 28 | T-4 Hordeum marinum Hudson Mediterranean barley 28 | T-10 Juncus effusus L. bog rush 28 10 1—29 moss moss 28 a T-1 Biotic environment Total Vegetative Cover (%) 71 (35-100) Slope (%) 21 (0-65) Aspect SW primarily Position Mid-slope, upper/lower slopes, draws Bare (%) 25 (1-62) n = 36 Rock (%) 5 (1-21) n = 20 Soil Series Los Osos, various Texture Clay loam, sandy clay loam, loam Coarse Fragments (%) Rootability Soil Drainage buckeye (Aesculus californica [Spach] Nutt.), or big-leaf maple (Acer macrophyllum Pursh), were common on springs with a tree overstory. Willows were identified to genus because of poor catkin years. Spring herbaceous species were diverse and were significant in identifying spring types. No rare, threatened or endangered species were identified in the spring samples. Spring plant communities. We identified 4 plant community types for the East Bay Hills oak wood- lands (Table 2). The initial split by TWINSPAN divided the tree-dominated types from the herba- ceous-dominated plots (A = 0.608). The herbaceous types were distinguished from one another based on the amount of sedges (Carex spp.), rushes (Juncus spp.), and horsetail (Equisetum arvense L.) present, all indicative of wetter sites (A = 0.418). Two tree- dominated types were distinguished when TWIN- SPAN separated California bay from willow sites (A = 0.531). Ryegrass-herbaceous type.—The Ryegrass-her- baceous type (Table 3) was dominated by ryegrass 22 (1-60) n = 14 Hard & massive, hard & fractured Somewhat-poorly to poorly drained (Lolium multiflorum Lam.), watercress (Rorippa nasturium-aquaticum [L.] Hayek), and toad rush (Juncus bufonius L.). This type averaged 43% total grass cover and 13% total forb cover. Total vege- tation cover averaged 71%, ranging between 35% and 100% cover (herbaceous + shrubs + trees). Only 39% of the springs in this type maintained a tree overstory, and only 36% of the plots in this type maintained a shrub component. Elevation averaged 376 m (128-817 m), textural classes were primarily somewhat poorly drained clay loams. Thirty-three of the 36 sample springs in the Ryegrass-herbaceous type were grazed by cattle. Rush-herbaceous type.—The Rush-herbaceous type (Table 4) was characterized by the presence of two rush species: common rush (Juncus patens E. Meyer) and J. xiphioides E. Meyer. Ryegrass was still a common component, and willows occurred on about 5 of the 11 plots in the type. Sedges (Car- ex spp.), American speedwell (Veronica americana [Raf.] Benth.), and field horsetail (Eguisetum av- ense L.) also occurred frequently. The Rush-her- 104 TABLE 4. RUSH-HERBACEOUS TYPE DESCRIPTIVE STATISTICS. Constancy equals the number of species occurrences/total| number of plots; cover equals the mean cover (%) of a species for a given type (T = the range of cover values. Characteristic plant taxa Salix spp. Rhamnus californica Eschsch. Rubus ursinus Cham. & Schldl. Lolium multiflorum Lam. Juncus patens E. Meyer Juncus xiphoides E. Meyer Veronica americana (Raf.) Benth Carex spp. Equisetum arvense L. Juncus effusus L. Mimulus guttatus DC. Conium maculatum L. Juncus bufonius L. Rorippa nasturium-aquatica (L.) Hayek Picris echioides L. Bromus diandrus Roth Cynara cardunculus L. Biotic environment Total Vegetative Cover (%) Tree Cover (%) Shrub Cover (%) Grass Cover (%) Forb Cover (%) Abiotic environment Elevation (m) MADRONO Con- stancy Cover trace or <1%); and range equals, Range [Vol. 48 \ } willow 45 46 1-100 California coffeeberry OF, 14 3-24 California blackberry a5 10 5-15 ryegrass 7 3 T-8 common rush 73 3 T-10 iris-leaved rush 73 5) T-16 American speedwell 64 6 2-14 sedge 64 y 2-16 field horsetail 55 17 T-35 bog rush 45 8 3-14 common monkeyflower 45 5 1-18 poison hemlock 36 3 2-6 toad rush 36 | T-3 watercress 2H 4 2-6 bristly oxtongue 27 1 T-2 ripgut brome oF 4 2-8 artichoke thistle Dy. 9 5-14 83 (55-100) 57 (1-100) n = 8 22 (1-57) n = 8 29 (10-50) n = 10 22 (1-65) n = 11 253 (128-402) Slope (%) 21 (10-32) Aspect NW primarily Position Upper and mid slopes, draws | Bare (%) 19 (3-50) n = 10 Rock (%) 3 (2-4) n = 2 Soil Series Los Gatos primarily Texture Loams | Coarse Fragments (%) 11 (4-15) n = 3 : Rootability Hard & massive, hard & fractured Soil Drainage baceous spring type was distinguished from the Ryegrass-herbaceous type by the presence of sedges and horsetail, as well as the occurrence of California blackberry (Rubus ursinus Cham. & Schldl.) and/or a woody overstory. Total vegetation cover averaged 83%, which is similar to the Rye- grass-herbaceous type. Tree and/or shrub cover was found on this type about 73% of the time, and when found, averaged 57% and 22% cover, respectively. The Rush-herbaceous type was found at an av- erage elevation of 253 m (128-402 m) on loamy textured soils. Three of the 11 plots classified as Rush-herbaceous were ungrazed. Willow-Poison oak type.—The Willow-Poison oak type (Table 5) was dominated by willows. On plots without willow, blue elderberry (Sambucus mexicana C. Presl) was often present. Poison oak (Toxicodendron diversilobum [Torrey & A. Gray] E. Greene) was common, occurring on 73% of the Somewhat well to poorly drained 11 sample plots classified in this type. The herba- | ceous understory was sparse with total graminoid | cover averaging 19%, and forb cover averaging 8%. Shrub cover averaged 40% and was found on all plots classified in this type. Tree cover averaged | 49% and occurred on all but one plot in this type where the shrub coffeeberry (Rhamnus californica’ Eschsch.) was found. | The Willow-Poison oak type occurred on sites averaging 254 m (116-536 m) elevation. Soils were predominantly loams, and 5 of 11 sites in this type were ungrazed. California bay-Poison oak type.—The California bay-Poison oak type (Table 6) was dominated by California bay. Coast live oak was present as the overstory species at the one site in this type that did not contain California bay. This type also con- tained Poison oak, blackberry, mosses, and occa- sionally chainfern (Woodwardia fimbriata Smith). i 2001] ALLEN-DIAZ ET AL.: SPRING-FED PLANT COMMUNITIES 105 “TABLE 5. WILLOW-PoISON OAK TyPE DESCRIPTIVE STATISTICS. Constancy equals the number of species occurrences/ | total number of plots; cover equals the mean cover (%) of a species for a given type; and range equals the range of ‘cover values (T = trace or <1%). i! "Characteristic plant taxa | Salix spp. Sambucus mexicana C. Pres] Toxicodendron diversilobum (Torrey & A. Gray) E. Greene Baccharis pilularis DC. Rubus ursinus Cham. & Schldl. Lolium multiflorum Lam. Bromus diandrus Roth Carex spp. Biotic environment Total Vegetative Cover (%) Tree Cover (%) Shrub Cover (%) Grass Cover (%) Forb Cover (%) Abiotic environment Elevation (m) Slope (%) Aspect Position Bare (%) Rock (%) Soil Series Texture Coarse Fragments (%) Rootability Soil Drainage Total vegetation cover averaged 91%. Graminoid and forb cover was very low, 5% and 11% respec- tively. Tree cover averaged 83% and was found on all plots in this type. Seventy-five percent of the plots in this type maintained shrub cover which av- eraged 32% when present. Elevation ranges for this type were 122 to 658 m with an average of 394 m. Grazing occurred at two of the 12 sites in this type. The textural class of these soils was primarily loams. Environmental and management factors. CCA ordination axes showed that at least 3 significant orthogonal gradients could be created by taking the linear combination of environmental variables (Ta- ble 7). Axis 1 (CCA,) was most closely correlated with the variable LITTER. The California bay-Poi- son oak type scored high on this vector compared to the other 3 vegetation groups (Fig. 3a). CCA, was correlated with ELEVATION primar- ily; however, the two categorical variables LOAM and CLAYLOAM were also useful in creating this axis (Table 7). Though CCA, helped disperse the species matrix as a whole, indicating a strong un- derlying gradient, it provided minimal insight into the separation of vegetation types. Willow-Poison Oak sites generally scored positively while the Con- stancy Cover Range willow 75) 42 25-56 blue elderberry 2. 9 4-18 poison oak 73 26 1-64 baccharis 45 20 T-59 California blackberry 45 23 2-60 ryegrass 36 2 T-4 ripgut brome ZI 7 3-13 sedge Deh Z T-3 85 (50-95) 49 (5-93) n = 10 40 (5-80) n = 11 19 (1-50) n = 11 8 (1-35) n = 11 286 (116-536) 21 (5-58) SE, SW Upper, mid, lower slopes 33 (5-60) n = 10 14 (1-20) n = 3 Los Osos, various Loams 19 (1-60) n = 4 Primarily hard & massive Primarily well drained Rush-herbaceous type appeared at mainly below- average elevations. Finally, CCA, separated sites within all four veg- etation types (Figs. 3b, c) apparently for the pres- ence or absence of grazing (Table 7). However, only 2 California bay-Poison oak sites scored pos- itively (indicating no grazing) on CCA, (Fig. 3c), contrasting CCA, and CCA,. Three DCA axes were also extracted from the Species matrix whose very high eigenvalues indi- cated 3 underlying gradients that were not com- pletely explained by linear combinations of our en- vironmental and management variables, 1.e., DCA found even stronger gradients than CCA (Table 7). This indicated an under-specified explanatory mod- el for this species matrix. Classical indirect gradient analysis would continue from this point by inferring causes of these gradients. Figure 4a shows that DCA, arrays our TWINSPAN-derived vegetation groupings from overstory to herbaceous types with little overlap, while DCA, separates the higher-el- evation California bay-Poison oak sites from the middle-elevation Willow-Poison oak and Ryegrass- herbaceous types. Rush-herbaceous sites tended to be located at lower elevations, which is also reflect- ed in Figure 4a. There is no evidence from Figures 106 MADRONO [Vol. 48, TABLE 6. CALIFORNIA BAY-POISON OAK TYPE DESCRIPTIVE STATISTICS. Constancy equals the number of species occur- | rences/total number of plots; cover equals the mean cover (%) of a species for a given type; and range equals the | range of cover values (T = trace or <1%). Con- stancy Cover Range Characteristic plant taxa Umbellularia californica (Hook. & Arn.) Nutt. California bay 92 57 16—100 Quercus agrifolia Nee coast live oak 50 44 10-95 Acer macrophyllum Pursh big leaf maple 25 17 1-32 Toxicodendron diversilobum (Torrey & A. Gray) E. Greene poison oak 75 23 T-20 Rubus ursinus Cham. & Schldl. California blackberry 42 6 3-17 moss moss 42 4 1-17 Galium aparine L. common bedstraw 2» 1 T-3 Woodwardia fimbriata Smith chaintfern 25 1] 4—25 Biotic environment Total Vegetative Cover (%) 91 (75-100) Tree Cover (%) Shrub Cover (%) Grass Cover (%) Forb Cover (%) Abiotic environment Elevation (m) 83 (45-100) n = 12 32 (1-60) n = 9 S115) no 10 14 (1-40) n = 11 394 (122-658) Slope (%) 30 (10—65) Aspect All Position Upper, mid, lower slopes Bare (%) 27 (7-65) n = 12 Rock (%) 3 (2-7) n = 6 Soul Series Los Osos, various Texture Loams Coarse Fragments (%) 34 (20-60) n = 4 Rootability Primarily fractured Soul Drainage 4b and 3c that DCA, is a grazing gradient as de- termined for CCA, based on its high correlation with GRAZED. Caution must be used when inter- preting correlations between ordination axes and bi- nary variates. Hence we examined a scatterplot of grazed and ungrazed sites distributed along DCA, TABLE 7. ORDINATION RESULTS. ! 180—azimuth. Axis | Axis 2 Axis 3 CCA Results Eigenvalue (\) 0.487 0.415 0.363 P 0.02 0.01 0.01 Intraset correlations ELEVATION —O0:312. —0554 —0:436 ASPECT! —().156 0.486 —0.221 SLOPE =—0. 342 + —0.207 —0:01% LOAM —04416) -—0:445 0.341 CLAYLOAM 0.321 0.428 —0.579 LITTER —0.834 0.062 0.156 BARE —0.211 0.276 —0.153 GRAZING 0.192 —0.006 —O.852 DCA Results Eigenvalue (\) 0.765 0.636 0.547 Excessively well to poorly drained that showed the fewer number of ungrazed sites to | be randomly distributed with a narrow range among the many grazed sites that exhibited a much wider range of distribution along DCA, (Fig. 5). Figure 6 shows selected taxa arrayed along CCA, and CCA,. The dotted line indicates where in spe- cies space ungrazed sites were found vis-a-vis grazed sites; there was no overlap. Overstory wil- lows (oOSALIX), blackberry shrub (SRUUR), and herbaceous layer Poison oak (hTODI) scored high- ly on CCA, concomitant with ungrazed_ plots. While willows were also found on the grazed plots, they tended to be found as sSsALIX—willows in a shrub state. Other overstory taxa were found on grazed plots as California bay (OQUMCA) and Coast live oak (OQUAG). Herbaceous taxa were distrib- uted along the negative side of the CCA, axis with Italian ryegrass (hHLOMU) and common monkey- flower (hMIGU) scoring highest for grazed plots and sedges (hCAREX) and rushes (hJUXI, hJUEF) scoring moderately. DISCUSSION DCA axes displayed three much stronger under- lying gradients than CCA axes exposed, making it ALLEN-DIAZ ET AL.: SPRING-FED PLANT COMMUNITIES 107 CCA | COMMUNITY TYPE A Ryegrass-herbaceous © Rush-herbaceous @ Willow Poison oak ° MM California bay Poison oak 2001] CCA 2 a A b CCA 3 ® — o & a © & A ee ungrazed grazed iTTER ro) & Va ry a =| A a a a . CCA3 z coo og Fic. 3. Canonical correspondence analysis (CCA) joint-plots contrasting each of 3 orthogonal ordination axes. All sites above dotted line were ungrazed; all sites below were grazed. Note: polygons not inclusive of all sites from a given community type. clear that the CCA model was under-specified, i.e., one or more driving factors were not included. CCA,’s correlation with litter quantity indicated a productivity gradient from the relatively low-pro- ductivity annual grass-dominated Ryegrass-herba- ceous type to the high biomass tree types—Cali- fornia bay-Poison oak and Willow-Poison oak. This spread seems indicative of a gradient driven by wa- ter availability. Ryegrass is an annual species that fares well on dry uplands as well as on areas that undergo periodic inundation throughout California, while rushes tend to be hydrophilic taxa existing in topographic depressions, seeps, and springs (Bow- erman 1944; Keator 1994). At lower elevation spring sites, water availability is likely to be more constant than at higher elevation sites due to greater upslope catchment area. Hence, these lower eleva- tion sites with even greater water availability and 108 MADRONO [Vol. 48) COMMUNITY TYPE A Ryegrass-herbaceous Rush-herbaceous b DCA 3 @ Willow-Poison oak MB California bay-Poison oak DCA 2 a Fic. 4. Detrended correspondence analysis (DCA) joint-plots contrasting each of 3 orthogonal ordination axes. Note: — polygons not inclusive of all sites from a given community type. | 2001] ALLEN-DIAZ ET AL.: SPRING-FED PLANT COMMUNITIES 109 | GRAZED ee re | | he.) ie | << UNGRAZED a 4 << J FT DCA, ‘Fic. 5. Distribution of ungrazed and grazed sites along the third detrended correspondence analysis gradient (DCA,). probably higher flows support willows. Willows are known to use stream water in addition to soil water to compete effectively with herbaceous vegetation, | ' Poison oak type scored high for greater litter levels since California bay is known for recalcitrant, slow- which uses soil water only (Alstad et al. 1999). Fi- nally, it was not surprising that the California bay- ‘ly decomposing leaves-due to its relatively high concentrations of phenolic secondary compounds, (Goralka and Langenheim 1995). California bay- | Poison oak sites also often maintained a coffeeber- _ry or blackberry shrub component adding to the high productivity. CCA, indicated a strong gradient, orthogonal to the productivity gradient discussed above (CCA,), that appeared tied to elevational differences among _ sites. However, the categorical soil texture variables LOAM and CLAYLOAM each showed high cor- relations with CCA, as well. Visual examination of _ Figures 3a and 3c showed that while CCA, strongly separated some of the individual species, it did not disperse our four plant groupings very well. Per- haps some species wax and wane with temperature fluctuations that vary with elevation while the dom- inant taxa (which defined the vegetation types) re- main. Also, the coarseness of our soil texture vari- ables may play a role in the relative ambiguity of the gradient represented by CCA,. The correlation coefficients for LOAM and CLAYLOAM showed opposite signs with respect to CCA,. LOAM corresponded roughly to higher elevations and CLAYLOAM to lower sites indica- tive of greater soil weathering via periodic inun- dation. CCA, appeared to represent a species ordination related to the presence or absence of livestock graz- ing. Interestingly, three plots from each type were ungrazed and all scored highly on CCA, indicating that these sites all had something in common veg- etatively. A popular notion is that livestock grazing degrades riparian areas and especially remove the willow component. The presence of willow at both grazed and ungrazed sites argues against this notion at first glance. Of the ten sites classified as Willow- Poison oak, six were grazed. What seems clear from Figure 6 is that while willows were present at both grazed and ungrazed sites, their size-classes were likely affected so that grazed sites maintained more of a willow shrub component, while ungrazed sites were more likely to maintain willows in the CCA 3 sRUUR hTODI ungrazed ee ee oSALIX grazed enna sTODI ELEVATION hMIGU Fic. 6. cca? hJUXT WEF hCAREX GRAZED oUMCA sSALIX hLOMU Canonical correspondence analysis (CCA) joint-plot (same as Figure 3c) showing selected taxa with associated vertical layer prefix, i.e., herbaceous (h), shrub (s), or overstory (0). Taxa are: RUUR—Rubus ursinus, TODI—Toxi- _ codendron diversiloba, SALYX—Salix spp., QUAG—Quercus agrifolia, JUXI—Juncus xiphioides, JUEF—J. effusus, ~ CAREX—Carex spp., UMCA—Umbellularia california, MIGU—Mimulus guttatus, LOMU—Lolium multiflorum. 110 overstory. These results align with Peinetti et al. (2001) who showed willow productivity was un- affected by large herbivores, but that overall mor- phology shifted to more prostrate growth forms with grazing. They further concluded that the tem- poral distribution of grazing was important in de- termining these effects. Another conspicuous pattern emerging from the 4 ungrazed Willow-Poison oak sites was the high cover of blackberry in the shrub layer (often >40% cover) and the herbaceous layer (~5—20% cover). We have also observed the conversion of herba- ceous spring sites to blackberry shrub dominance that completely eliminates the herbaceous compo- nent at Sierra Nevada oak woodland springs (B. Allen-Diaz unpublished data). Interestingly, both the Ryegrass-herbaceous and the California bay-Poison oak types scored posi- tively for grazing (negatively on CCA,) showing that both life-form types (herbaceous and oversto- ry) were extant under this management scenario. More detailed information about grazing manage- ment (intensity and temporal distribution) under ex- perimental designs needs to be applied before con- clusions about grazing effects on spring types can be verified. Livestock grazing has been implicated in general riparian area degradation (Fleischner 1994; Belsky et al. 1999), but has also received credit for ameliorating streambank slumping (My- ers and Swanson 1992) and freshwater fish habitat (Knapp et al. 1998). Allen-Diaz and Jackson (2000) showed that light to moderate grazing intensity (based on upland residual dry matter estimates) re- sulted no compositional shifts on Sierra Nevada oak woodland springs. Certainly, overgrazing (unsus- tainable grazing pressure), will induce a cascade of deleterious effects on ecosystems, 1.e., bank ero- sion, vegetation loss, reduced stream water quality (Belsky et al. 1999), however, there is neither ca- sual nor scientific indication that either of the two land management agencies discussed here practice overgrazing. Grazing presence/absence does not appear to in- fluence the community type overall (i.e., composi- tion) but may have important within-type effects on the vegetative structure when willows are present. Other overstory taxa did not show this pattern. By using quantitative, quasi-objective classifica- tion analysis, we were able to delimit four spring- fed plant community types for the East Bay Hills, California. Finer scale description would certainly provide a greater number of community types; these four appeared to repeat strongly across these landscapes. These results should not be taken as evidence that livestock grazing has no important effects on these systems—only well designed field experiments will inform these questions. However, these results should provide useful information to those designing and implementing future experi- ments and surveys. Other California landscapes and regions containing similar and disparate spring MADRONO types also need delineation. For instance, we did not observe the ubiquitous cattail type (Typha spp.) [Vol. 48 so often observed in Sierra Nevada foothill oak | woodlands. Community types should be described | on a site-specific basis in order to more fully char- | acterize variability within and among these patch | ecosystems. ACKNOWLEDGMENTS Thanks to Ray Budzinski (EBRPD), Rod Tripp (EB- | MUD), and other park managers for their help. Peter Hop- | kinson’s comments greatly improved early versions of this manuscript. The East Bay Regional Park District and the East Bay Municipal Utility District funded this study. LITERATURE CITED ALLEN, B. H., R. R. Evett, B. A. HOLZMAN, AND A. J. MARTIN. 1989. Report on rangeland cover type de- scriptions for California hardwood rangelands. Forest and Rangeland Resources Assessment Program, De- partment of Forestry and Fire Protection, Sacramento, CA. Contract #8CA63912. ALLEN-DIAZ, B. AND R. D. JACKSON. 2000. Grazing effects on spring ecosystem vegetation of California’s hard- wood rangelands. Journal of Range Management 53: 215-220. , AND J. S. FEHMI. 1998. Detecting channel morphology change in California’s hardwood range- land spring ecosystems. Journal of Range Manage- ment 51:514-518. AND B. A. HOLZMAN. 1993. Resampling WTM plots in blue oak cover type series. Forest and Range- land Resources Assessment Program, Department of Forestry and Fire Protection, Sacramento, CA. Con- tract #8CA63912. ALSTAD, K. P., J. M. WELKER, S. A. WILLIAMS, AND M. J. TRLICA. 1999. Carbon and water relations of Salix monticola in response to winter browsing and changes in surface water hydrology: an isotopic study using delta C-13 and delta O-18. Oecologia 120:375-— So: BELSKY, A. J., A. MATZKE, AND S. USELMAN. 1999. Survey of livestock influences on stream and riparian eco- systems in the western United States. Journal of Soil and Water Conservation 54:419—431. BOLSINGER, C. L. 1988. Hardwoods of California’s tim- berlands, woodlands and savannas. Portland, OR: USDA Pacific Northwest Forest and Range Experi- ment Station, Forest Service. PNW-RB-148. BONHAM, C. D. 1989. Measurements for terrestrial vege- tation. Wiley: New York. BOWERMAN, M. L. 1944. The flowering plants and ferns of Mount Diablo, California: their distribution and association into plant communities. Gillick Press, Berkeley CA. CAMPBELL, C. G. AND B. ALLEN-DIAz. 1997. Livestock grazing and riparian habitat water quality: an exam- ination of oak woodland springs in the Sierra Foot- hills of California. USDA Pacific Southwest Forest and Range Research Station. PSW-GTR-160. CLARY, W. P. 1995. Vegetation and soil responses to graz- ing simulation on riparian meadows. Journal of Range Management 48:18-—25. . 1999. Stream channel and vegetation responses to late spring cattle grazing. Journal of Range Man- agement 52:218—227. 2001] Davis, E W., R. W. Brown, B. BUYAN, AND J. KEALY. 1995. Vegetation change in blue oak and blue oak/ foothill pine woodland. Forest and Rangeland Re- sources Assessment Program, Department of Forestry and Fire Protection, Sacramento, CA. Contract #8CA06673. | EBMUD. 1995. Proposed East Bay watershed master plan. East Bay Municipal Utility District, Oakland, CA. EBRPD. 1992. Wildland management policies and guide- lines. East Bay Regional Park District, Oakland, CA. . 1996. East Bay Regional Park District master plan. East Bay Regional Park District, Oakland, CA. _ Exzinca, C. L., D. W. SALZER, AND J. W. WILLOUGHBY. 1998. Measuring and monitoring plant populations. US Department of the Interior Bureau of Land Man- agement National Applied Resource Sciences Center, Denver, CO. ’ FEeRREN, W. R., P. L. FIEDLER, AND R. A. LeIDy. 1994. Wetlands of the central and southern California coast and coastal watersheds. Unpublished report. Environ- mental Protection Agency, Region 9, San Francisco, CA. Fites, J. 1993. Ecological guide to mixed conifer plant associations. USDA Forest Service, Pacific South- west Region, San Francisco, CA. FLEISCHNER, T. L. 1994. Ecological costs of livestock graz- ing in western North America. Conservation Biology 8:629—644. Gaucu, H. G. 1982. Multivariate analysis in community ecology. Cambridge University Press, Cambridge. GoRALKA, R. J. L. AND J. H. LANGENHEIM. 1995. Analysis of foliar monoterpenoid content in the California bay tree, Umbellularia Californica, among populations across the distribution of the species. Biochemical Systematics and Ecology 23:439—448. Gorbon, H. AND T. C. WuirTe. 1994. Ecological guide to southern California chaparral plant series. USDA For- est Service, Pacific Southwest Region, San Francisco, CA. R5-ECOL-TP-005. HickMaN, J. C. 1993. The Jepson manual. University of California Press, Berkeley, CA. Hitt, M. O. 1979. TWINSPAN: A FORTRAN program for arranging multivariate data in an ordered two-way table by classification of the individuals and _attri- butes. Cornell University, Ithaca, NY HOLLAND, R. E 1986. Preliminary description of the ter- restrial natural communities of California. California Resources Agency, Department of Fish and Game. JEPSON, W. L. 1975. A manual of the flowering plants of California. University of California Press, Berkeley, CA. JONGMAN, R. H., C. J. EF TER BRAAK, AND O. E R. VAN TONGEREN. 1995. Data analysis in community and landscape ecology. Cambridge University Press, Cambridge. KAUFFMAN, J. B. AND W. C. KRUEGER. 1984. Livestock impacts on riparian ecosystems and streamside man- agement implications: a review. Journal of Range Management 37:430—438. KeEaATor, G. 1994. Plants of the East Bay parks. Roberts Rinehart Publishers, Niwot, CO. Knapp, R. A., V. T. VREDENBURG, AND K. R. MATTHEWS. 1998. Effects of stream channel morphology on gold- ALLEN-DIAZ ET AL.: SPRING-FED PLANT COMMUNITIES 111 en trout spawning habitat and recruitment. Ecological Applications 8:1104—1117. LARSEN, R. E., W. C. KRUEGER, M. R. GEORGE, M. R. BARRINGTON, J. C. BUCKHOUSE, AND D. E. JOHNSON. 1998. Viewpoint: livestock influences on riparian zones and fish habitat—literature classification. Jour- nal of Range Management 51:661—664. McCune, B. AND M. J. MEFForD. 1999. PC-ORD multi- variate analysis of ecological data. Version 4. MjM Software Design. Geneden Beach, OR. Moy Leg, P. B., R. KATTELMANN, R. ZOMER, AND P. J. RAND- ALL. 1996. Sierra Nevada ecosystem project, Final Report to Congress, Vol. 3. Centers for Water and Wildland Resources University of California Davis, CA. Mulick, P. C. AND J. W. BARTOLOME. 1987. An assessment of natural regeneration of oaks in California. Califor- nia Dept. Forestry and Fire Protection, Sacramento, CA. Munz, P. A. AND D. D. Keck. 1973. A California flora: with supplement. University of California Press, Berkeley, CA. Myers, T. J. AND S. SWANSON. 1992. Variation of stream stability with stream type and livestock bank damage in northern Nevada. Water Resources Bulletin 28: 743-754. OrNbuFF, R. 1974. An introduction to California plant life. University of California Press, Berkeley, CA. PALMER, M. W. 1993. Putting things in even better order— the advantages of canonical correspondence analysis. Ecology 74:2215—2230. Passor, P. AND J. BARTOLOME. 1985. An integrated hard- wood range management program. A proposal for re- search and extension by the UC Division of Agricul- ture and Natural Resources. PEINETTI, H. R., R. S. C. MENEZES, AND M. B. COUGHEN- our. 2001. Changes induced by elk browsing in the aboveground biomass production and distribution of willow (Salix monticola Bebb): their relationships with plant water, carbon, and nitrogen dynamics. Oec- ologia 127:334-—342. RINNE, J. N. 1999. Fisheries habitat—fish and grazing re- lationships: the facts and some pleas. Fisheries 24: 12-21. SAWYER, J. O. AND T. KEELER-WOLF. 1995. A manual of California vegetation. California Native Plant Soci- ety, Sacramento, CA. SmiITH, S. 1994. Ecological guide to eastside pine plant associations. USDA Forest Service, Pacific South- west Region, San Francisco, CA. R5-ECOL-TP-004. STOHLGREN, T. J., A. J. OWEN, AND M. Leg. 2000. Moni- toring shifts in plant diversity in response to climate change: a method for landscapes. Biodiversity and Conservation 9:65—86. SwiEckI, T. J. AND E. A. BERNHARDT. 1991. Minimum in- put techniques for restoring valley oaks on hardwood rangeland. Forest and Rangeland Resources Assess- ment Program, Department of Forestry and Fire Pro- tection, Sacramento, CA. Contract #8CA84965. TER BRAAK, C. J. E 1986. Canonical correspondence anal- ysis: a new eigenvector technique for multivariate di- rect gradient analysis. Ecology 67:1167—1179. . 1987. The analysis of vegetation-environment re- lationships by canonical correspondence analysis. Ve- getatio 69:69-78. MADRONO, Vol. 48, No. 2, pp. 112-115, 2001 ATRIPLEX ROBUSTA (CHENOPODIACEAE), A NEW PERENNIAL SPECIES FROM NORTHWESTERN UTAH HOWARD C. STUTZ AND MILDRED R. STUTZ Department of Botany and Range Science Brigham Young University, Provo, UT 84602 STEWART C. SANDERSON USDA Forest Service Shrub Sciences Laboratory Provo, UT 84601 ABSTRACT Atriplex robusta is a newly described species from northwestern Utah. It is abundant along the shoul- ders of roads and highways and in slightly elevated terrain within and bordering saline playas. It appears to be morphologically closest to A. tridentata Kuntze but differs in its larger stature, more woody, caes- pitose habit, larger urn-shaped fruiting bracts and broader leaves. On the shoulders of Highway I-80 between Grantsville and Wendover, Tooele County, Utah in northwestern Utah, there are numerous small pop- ulations of a distinctive form of perennial Atriplex. Because of its robust habit we refer to it as A. ro- busta, described below as a new species. It is par- ticularly abundant about 65 km east of Wendover near the railroad siding at Knolls. In this area Atri- plex canescens (Pursh) Nutt. and A. tridentata Kun- tze come together, hybridize and produce a variety of hybrid segregants including plants of A. robusta which appear to have spread, as a new species, for several miles along the freeway (I-80) shoulders, in both directions. Atriplex robusta H. C. Stutz, M. R. Stutz, and S. C. Sanderson, sp. nov. (Fig. 1).—TYPE: USA, Utah, Tooele Co., 1 mi W of Knolls, T15 R13W S15, shoulder of highway I-80, 1280 m eleva- tion, 16 Sep 1977, H. C. Stutz 814] (Holotype: BRY; Isotypes, BRY, CA, CAS, GH, MO, NY, RM, UC). Frutices caespitosi, 40-80 cm alti. Caules erecti vel ascendentes, ramosi a basi ad apicem, dense furfuraceus, 1-8 mm diam., fragilis. Folia oblonga, ascendentia usque appressa, dense furfuracea; folia ephemera verna et aestiva 15-30 mm longa, 5—10 mm lata; folia serotina aestiva et hiberna 3—10 mm longa, 2—5 mm lata, anatomia foliaris Kranz-typi. Plantae dioeciae, raro monoeciae. Flores staminati sessiles, ad brevi-ramulus axillares in angustipani- culas terminales; perianthium campanulatum, 5- partitum ad medium, dense furfuraceum, segmentis Ovatis usque ellipticis, 2 mm longis, | mm latis; stamina 5, filamentis | mm longis, antheris ca. 2 mm longis, | mm latis. Flores pistilati solitarii, ses- siles, in plerumque sine foliis confertas paniculas terminales. Bracteae fructiferae furfuraceae, com- pressae, urceolatae, latissimae infra media, 5 mm latae, 7-8 mm longae, exappendiculatae, cum 3—10 marginalibus dentibus, 0.5—2 mm longis, qui me- dianus longissimus. Utriculus orbiculatus, pericar- pio membranceo pellucido. Semena 5 mm diam., | testa membranacea, brunnea; radicula supera. Perennial caespitose shrub, 40—80 cm tall (Fig. 2). Stems erect or ascending, ramified from base to top, densely furfuraceous, 1-8 mm in diameter, brittle. Leaves oblong, obtuse, ascending to ap- pressed, densely furfuraceous, Kranz-type anatomy, ephemeral spring and summer leaves 5—10 mm wide, 15—30 mm long, late summer and winter leaves 2—5 mm wide, 3—10 mm long. Plants dioe- cious, rarely monoecious; staminate flowers sessile, on short axillary branches in terminal narrow pan- icles, calyx campanulate, sepals 5, united halfway, densely scurfy, ovate to elliptical, | mm wide, 2 mm long, stamens 5, yellow, filaments 1 mm long, anthers 2 mm long, | mm wide; pistillate flowers solitary, sessile, in dense, mostly non-leafy, termi- nal panicles, bracteoles furfuraceous, laterally com- pressed, urceolate, widest below the middle, 5 mm wide, 7-8 mm long, unappendaged, united to near the apical margin, marginal lips slightly divergent, marginal teeth 3—10, conspicuous, 0.5—2 mm long, central tooth largest. Utricles orbicular, pericarp membranaceous, transparent. Seeds 5 mm in di- ameter, testa membranaceous, brown, radicle su- perior. Flowering period: June—July. Chromosome number: 2n = 54. Additional Collections: USA, Utah, Tooele Co.: road shoulder W of Knolls, 30 Apr 1975, K. Harp- er; shoulders of highway I-80, Knolls, T15 R13W S15, 25 Sep 1975, H.C. Stutz 7842; Knolls, 26 Jul 1977, H.C. Stutz 8068; 1 mi S of Knolls, T15 R1I3W S23, 17 Aug 1978, H.C. Stutz 8338; 8 mi W of Knolls, T15 RI4W S16, 12 Jul 1979, H.C. Stutz 8461; 3 mi W of Knolls on shoulders of old abandoned highway, T15 R13W S21, 20 Sep 1994, H.C. Stutz 9670; 1 mi W of Knolls on old highway, 9 Sep 1995, AC. Smiz 9S3T. | | Sem 5mm Fic. 1. Atriplex robusta. a. Habit. b. Fruiting bract. c. Seed. (Illustrations by Loretta Orgill.) Atriplex robusta is common in Tooele county, Utah, particularly along the shoulders of Highway I-80 (Fig. 3). Its nearest relative appears to be A. tridentata Kuntze, from which it differs in several significant features. The fruiting bracts of A. ro- busta usually have trident or polydent apical mar- gins like those in most A. tridentata plants, but they are considerably larger (5 mm wide X 6—8 mm long vs. 3 mm wide X 4 mm long). The fruiting bracts of A. robusta are urn-shaped, non-appendaged, whereas those of A. tridentata are cuneate and usu- ally appendaged. A. robusta plants are much taller than A. tridentata (40—80 cm vs. 20—50 cm) and woodier, and in contrast to A. tridentata plants, A. STUTZ ET AL.: ATRIPLEX ROBUSTA 113 robusta plants do not form root-sprouts. The leaves of A. robusta are much wider than those of A. tri- dentata (5-10 mm vs. 2—5 mm). Atriplex robusta differs from A. canescens in its non-winged fruiting bracts, more herbaceous habit, smaller stature, broader leaves and urn-shaped utricles. Associated Species. The principal associates of A. robusta are Allenrolfea occidentalis (Wats.) Kuntze and Suaeda torreyana Wats. Each of these species appears to be well adapted to the conditions along the shoulders of the roads and railroads in these areas. Allenrolfea and Suaeda plants also oc- cur in some abundance in many of the surrounding saline playas, but A. robusta appears to be mostly restricted to the limited habitat along the shoulders of roadways that provides increased amounts of fresh water and improved leaching of salts, and in elevated areas within the saline playas. Near Knolls, where a sizeable population of A. canescens occurs, hybrids between A. robusta and A. canes- cens are common. Although A. robusta is hexaploid and the putative A. canescens parent is tetraploid, the resulting pentaploid hybrids are partially fertile and segregant progeny, displaying a wide array of intermediate phenotypes, are fairly common. A POSSIBLE ORIGIN OF ATRIPLEX ROBUSTA As reported by Stutz et al. (1979), A. robusta appears to have been derived from hybrids between tetraploid A. canescens (2n = 36) and hexaploid A. tridentata (2n = 54). Near Knolls, Tooele C., Utah, where A. robusta is abundant, both A. canescens and A. tridentata are present, as well as putative hybrids between them and also several plants that appear to be segregants from the hybrids including plants described as A. robusta. In this area, A. ca- nescens is abundant on sand dunes that extend to the south shoulder of Highway I-80 and a few A. tridentata plants are on the adjacent lower, saline flats. Sizeable populations of A. tridentata occur a few miles to the east and a few miles to the west, on bottomland clay soils. The contrast in phenotypes of A. canescens, A. tridentata and A. robusta plants in this area, is strik- ing. Atriplex canescens plants are tall (1—2 meters), caespitose, woody shrubs with large leaves (ca. 5 mm wide X 30 mm long), and large 4-winged fruit- ing bracts (20 X 20 mm). Atriplex tridentata plants are short (ca. 20—5O cm), herbaceous, subshrubs with vigorous root sprouts, and leaves that are mostly small and linear (ca. 5 X 20 mm). Atriplex robusta plants are intermediate, in several features supporting the conjecture that it is a derivative from A. tridentata X A. canescens hybrids. They are larger-statured (40—80 cm) than A. tridentata plants and smaller than those of A. canescens, more woody than A. tridentata plants, less woody than A. canescens plants and their distinctive, mostly smooth surfaced, urn-shaped fruiting bracts are much larger than those of A. tridentata and smaller 114 MADRONO ith \ % : "lh Fic. 2. Atriplex robusta. A single plant in a population near Knolls, Tooele Co., Utah. uw boy Uk Fic. 3. Population of Atriplex robusta, near Knolls, Tooele Co., Utah. Bar = 80 cm for plants in foreground. } | H } 2001] than those of A. canescens. Collectively, their caes- pitose habit, intermediate stature, large, broad, ob- tuse leaves, numerous upright stems and large, un- _ appendaged, urn-shaped fruiting bracts, clearly set _ A. robusta plants apart from those of both putative | parents. Hybrids between A. tridentata and A. canescens are common whenever they occur together in na- ture, and although such hybrids are highly sterile pentaploids, fertile derivatives are sometimes de- rived from them (Stutz et al. 1979; Stutz and San- derson 1998). Since, at Knolls, there is a large pop- ulation of A. canescens, but only a few A. tridentata plants where the hybrids and hybrid derivatives are common, A. tridentata plants were probably the fe- male parents of the hybrids. Their relative scarcity may have enhanced the production of hybrid off- spring. If this interpretation of the origin of A. robusta is correct, A. robusta is probably of rather recent vintage because its preferred habitat, along the shoulders of Highway I-80 and neighboring road- ways, is also very young (I-80 was completed in 1969). An early count of A. robusta plants on the STUTZ ET AL.: ATRIPLEX ROBUSTA |B) shoulders of the west-bound lane of I-80 near Knolls, in 1975, by the authors, showed approxi- mately 17,500 plants on the shoulders of the west- bound lanes. Assuming the same density on the shoulders of the east-bound lanes, there were ap- proximately 35,000 plants present in 1975; today there appear to be millions. Their rapid increase was probably enhanced by their high seed produc- tion; some individual plants produce more than 25,000 fruits. ACKNOWLEDGMENTS We thank Professor Ge-lin Chu for technical assistance, and Broken Hill Proprietary and Brigham Young Univer- sity for financial support. LITERATURE CITED Stutz, H. C., C. L. Pope, AND S. C. SANDERSON. 1979. Evolutionary studies of Atriplex: Adaptive products from the natural hybrid, 6n A. tridentata X 4n A. canescens. American Journal of Botany 66:1181— G3: AND S. C. SANDERSON. 1998. Taxonomic clarifi- cation of Atriplex nuttallii (Chenopodiaceae) and its near relatives. Sida 18:193—212. MApRONO, Vol. 48, No. 2, pp. 116-122, 2001 A NEW I[POMOPSIS (PALE MONIACEAE) FROM THE SOUTHWEST USA AND ADJACENT MEXICO DIETER WILKEN Santa Barbara Botanic Garden, 1212 Mission Canyon Road, Santa Barbara, CA 93105 ABSTRACT Ipomopsis longiflora subsp. neomexicana, a new subspecies, is distinguished from subsp. longiflora (Torr.) V. E. Grant and subsp. australis R. A. Fletcher and W. L. Wagner by its annual habit, ovaries with a modal number of 30 ovules, minutely winged seeds, and ellipsoid capsules that are about twice the length of the calyces. Subspecies neomexicana occurs in parts of the Great Basin and Chihuahuan Desert, ranging from southeastern Utah and southern Colorado through eastern Arizona and New Mexico into northern Mexico. The 3 subspecies are self-compatible, strongly self-pollinating, but have flowers that attract hawkmoths. Ipomopsis longiflora (Torr.) V. E. Grant is based on a specimen (Holotype: NY!) collected by Edwin James ‘‘on the Canadian River,” probably what is now the North Canadian River in Oklahoma (McKelvey 1956). Descriptions of Ipomopsis lon- giflora were given by Wilken (1986), Welsh et al. (1987), and Cronquist et al. (1984), the latter two under the name Gilia longiflora (Torr.) G. Don. The species is distinguished from all members of /pom- opsis by its sparsely glandular-puberulent herbage, pinnatifid leaves with 5—7 linear to narrowly oblong segments, open, paniculate inflorescences, glandu- lar-puberulent calyx tubes, and white to pale blue salverform corollas that are 25—45 mm long. /pom- opsis longiflora is confused sometimes with the re- lated I. laxiflora (J. M. Coult.) V. E. Grant, which has a smaller stature and corolla tubes between 8 and 20 mm long (Wilken 1986). Subspecies lon- giflora is a biennial of the western Great Plains, distributed from South Dakota to central Texas (Wilken 1986). It is characterized by narrowly el- lipsoid capsules that are about the length of the ca- lyx. Fletcher and Wagner (1984) applied the name J. longiflora subsp. australis to plants with ovoid to broadly ellipsoid capsules that are equal to or that only slightly exceed the calyx. Fletcher and Wagner also reported differences in seed shape and calyx lobe vestiture. Subspecies australis occurs from ‘‘Mohave Co., Arizona, south and west of the Mo- gollon Plateau, to southwestern New Mexico as far east as Deming in Luna Co., New Mexico, and to northern Sonora’ (Fletcher and Wagner 1984). However, Fletcher and Wagner did not study spec- imens of the species east of the Rocky Mountains. While reviewing specimens and conducting field work for the Flora of the Great Plains and Flora of North America projects, I noticed differences be- tween plants of the Great Plains and those treated as subsp. longiflora by Fletcher and Wagner. Plants from the Great Plains are all assignable to the typ- ical subspecies, but plants referred to subsp. lon- | giflora by Fletcher and Wagner are recognized here | as a new, morphologically and geographically dis- tinct subspecies. Ipomopsis longiflora (Torr.) V. E. Grant subsp. neomexicana Wilken, subsp. nov. (Fig. 1)— TYPE: USA, New Mexico: Sierra Co., Mitchell Point Road, 1.6 mi N of junction with Interstate 25, 14 mi N of Truth or Consequences. 15 May 1985, Wilken 14323 (Holotype UNM; isotypes ARIZ, CAS, CS, RM, RSA, UC). Herba annua foliosa, 20-52 cm alta. Folia cau- lina pinnatifida, 12—35 cm longa, supra glabri infra secus costas sparsim pubescentes; partes 5—7, an- guste lineares, laterales distantes 5-19 mm longae, terminales 9-29 mm longa. Inflorescentia laxa sub- corymbosa vel aliquantum rotundata; bracteae simplices vel subulatae, 7-15 mm longae. Calyces in statu florendi 6—9 mm longae, extus minute glan- dulosus, interne prope apices loborum puberulus. Corollae hypocrateriformes, 30-47 mm _ longae; lobis late obovatis vel rotundati, persaepe caesii, 7— 10 mm longi. Capsulae ellipsoidae, 10—15 mm lon- gae, calyce 1.5—2plo longiorae. Semina anguste el- lipsoidea, pallide flavo-brunea, 2.5—3 mm longa, ad apices minute alati. Annual herb 20-52 cm tall, with one to four erect to ascending branches arising from below middle of central axis. Stems glabrous to sparsely invested with short, uniseriate, eglandular and glan- dular trichomes. Leaves 12—35 mm long, pinnatifid, with (3)5—7(9) linear segments ca. 1-2 mm wide, lateral segments 5-19 mm long, distal segments 9— 29 mm long, petioles 4-13 mm long; basal leaves in a rosette, senescent at anthesis; cauline leaves 18-35 mm long, gradually reduced upwards, sub- glabrous above, sometimes with minute trichomes bearing a terminal, globose cell, sparsely pubescent on lower surface of rachis and lateral segments, tri- chomes uniseriate and mostly eglandular. /nflores- 2001] WILKEN: JPOMOPSIS LONGIFLORA SSP. NEOMEXICANA Fic. 1. Tle/ Ipomopsis longiflora, with details of diagnostic characters that separate the subspecies. A-C. Calyx and fruit. A. ssp. australis. B. ssp. neomexicana C. ssp. longiflora. D-E Seeds. D. ssp. australis. E. ssp. neomexicana FE. ssp. longiflora. G. Generalized habit of ssp. longiflora. H. Generalized habit of ssp. neomexicana. cences indeterminate, paniculate, corymbiform to somewhat rounded, with flowers in loose, terminal clusters of 2—3; bracts simple, linear to aculeate, 7— 15 mm long; pedicels slender, 9-24 (30) mm long, sparsely glandular-puberulent, trichomes uniseriate with a terminal globose cell. Calyces 6—9 mm long, tube 3-6 mm long, lobes 2.5—4 mm long, narrowly acuminate, tube sparsely to moderately glandular- puberulent externally, lobe apices eglandular pu- berulent within, trichomes uniseriate and straight or slightly curled. Corollas salvertorm, 30-47 mm long, glabrous, tube white, lobes elliptic to subro- tund, 7-10 mm long, usually pale bluish or bluish lavender, sometimes white, apices rounded to apic- ulate. Ovary 5—6 mm long, style 25—44 mm long, stigmas included in the tube, sometimes slightly ex- serted, stigmatic branches 2—3 mm long; ovules 8— 10 per locule. Stamens inserted unequally above mid-tube, 3—4 included, 1—2 slightly exserted, low- er filaments subsessile to 4 mm long, upper fila- 118 ments 1—3 mm long, anthers 2—3 mm long. Pollen white, sometimes bluish, grains subspheroidal, zon- ocolporate, colpi 6—8, exine striate or striate-retic- ulate near colpi. Capsules 10-15 mm long, 1.5—2 times longer than the fruiting calyx; seeds 2.5—3 mm long, yellowish brown, narrowly ellipsoid, dor- sal margins and apices minutely winged. Paratypes. USA, Arizona: Apache Co., 10 mi N of Springerville, Pase 1508 31 Aug 1965 (ASU), Coconino Co., Flagstaff, Hanson 128 11 Jun 1922 (LL, NY, TX), Mohave Co., Mociac Ranch, Cottam 4171 6 Jun 1929 (NY), Navajo Co., 5 mi N of Snowflake, Deaver 6490 12 Jul 1963 (ARIZ, ASU), Yavapai Co., Cornville, Jones 266 19 Sep 1922 (ARIZ); Colorado: Alamosa Co., 25 mi NE of Ala- mosa, J//tis 4352 11 Jul 1953 (RSA, WI), Costilla Co., N of Alamosa, Bethel, Willey & Clokey 4250 27 Jun 1921 (NY, POM, RM, US, WS), La Plata Co., Animas River, 2 mi S of Bondad, Ownbey 1445 17 Aug 1937 (NY, RM, UTC, WS), Monte- zuma Co., McElmo Canyon, Weber 7948 1 Sep 1952 (COLO, WS), Montrose Co., La Sal Creek, Gierisch 1381 21 Jun 1942 (RM), Saguache Co., Crestone, Ramaley 12089 18 Jul 1928 (RM); New Mexico, Bernalillo Co., Rio Puerco, Ripley & Bar- neby 2379 7 Oct 1939 (NY), Catron Co., E of Re- serve, Clark s.n. 17 Aug 1942 (UNM), Dona Ana Co., 3 mi E of Las Cruces, Ward 66 9 Apr 1981 (LL, NMC, NY, TX), Eddy Co., 22 mi SW of Carlsbad, Waterfall 3750 14 Aug 1942 (NY), Grant Co., 8 mi W of Silver City, Barkley 14694 24 Sep 1944 (OKL), Guadalupe Co., | mi E of Santa Rosa, Clark 10016 18 Jun 1951 (UNM), Lincoln Co., 4 mi NW of Capitan, Stephens 25897 29 Jul 1968 (KANU), Luna Co., 5 mi S of Deming, Sands s.n. 6 Jun 1962 (UNM), McKinley Co., 7 mi NE of Borrego Pass, Marley 1575 21 Jun 1978 (UNM), Otero Co., 11 mi W of Cloudcroft, Waterfall 12951 22 Aug 1956 (ARIZ, US), Rio Arriba Co., near Espanola, Correll 50851 26 Jul 1979 (NY), San Juan Co., 4 mi N of La Plata, Spellenberg & Ward 6155 11 Aug 1981 (NMC), Sandoval Co., 12 mi N of San Ysido, Shultz & Shultz 1304 (NY, UTC), Santa Fe Co., 13 mi W of Pojoaque, Gentry 23037 6 Sep 1972 (ASU), Socorro Co., 5 mi W of Bing- ham, Dunn 5201] 6 Oct 1948 (UNM), Taos Co., Ojo Caliente, Smith s.n. 30 Jul 1894 (WS), Torrance Co., 8 mi S of Estancia, Martin 4374 12 Sep 1960 (UNM), Valencia Co., El Morro Nat. Mon., Kayes 42 9 Jul 1978 (UNM); Texas, Brewster Co.: 2 mi E of Lajitas, Higgins 2763 25 May 1970 (NY), Cul- berson Co., near Van Horn, Warnock 388 20 Apr 1938 (TX), El Paso Co., Hueco Tanks, Powell & Powell 3005 12 Jul 1976 (TX), Hudspeth Co., 8 mi W of Sierra Blanca, Hitchcock 6775 25 Jun 1940 (NY, POM, UTC, WS), Presidio Co., 11 mi S of Marfa, Correll & Rollins 23651 19 Apr 1961 (NY TX); Utah, Garfield Co., 11.4 mi E of Escalante, Holmgren & Holmgren 4719 10 Aug 1970 (KANU, NY, US, UTC), Grand Co., Dead Horse Point, Car- MADRONO unless otherwise noted. Subsp. longiflora. Colorado, Logan Co., US Highway 6, | 2.8 mi E of junction with Interstate 76, 29 Sep 1982, | Wilken 13887, 2n = 14; Nebraska, Sheridan Co., 0.8 mi N of Niobrara River, 12 Aug 1983. Wilken 13988, 2n = | 14. Subsp. neomexicana. Arizona, Apache Co., Hwy 666, 2 - mi N of Zuni Wash, 21 May 1985, Wilken 14422, 2n = | 14, Coconino Co., between Wupatki Ruin and Sunset Cra- ; ter, 8 Aug 1987, Wilken 14844, 2n = 14; Colorado, Ala- | mosa Co., 5 mi E of Alamosa, 7 Jul 1986, O’Kane 2464; - New Mexico, Sierra Co., Mitchell Point Rd, 1.6 mi N of | junction with Interstate 25, 15 May 1985, Wilken 14323, | 2n = 14; Texas, El Paso Co., El Paso, 28 Mar 1983, Wor- | thington 9651 2n = 14, (UTEP). | Subsp. australis. Arizona, Gila Co., Hwy 60, N side of | Gila River Bridge, 21 May 1985, Wilken 14416, 2n = 14, | Hwy 77, 12 mi N of Globe, 20 May 1985, Wilken 14414, | 2n = 14. ter 1561 27 May 1940 (UTC), San Juan Co., near La Sal, Flowers 2084 7 Jun 1939 (NY). MEXICO, Chihuahua, 4.8 mi S of Samalayuca, Hendrickson 5783 19 Aug 1971 (TX), Coahuila, 10 mi s of La- guna del Rey, Hendrickson 14158 20 Sep 1974 CLS): METHODS 677 specimens from 21 herbaria were studied (ARIZ, ASU, BRY, COLO, CS, KANU, KSU, LL, NEB, NMC, NY, OKL, POM, RM, RSA, TX, UNM, US, UTC, WI, WS). Eight quantitative char- acters were chosen for detailed study (Table 2). Measurements were made on a subset of specimens that represented the geographic range of each tax- on. Sample sizes for each character varied, depend- ing on the condition of specimens. Field studies were conducted in one to three populations of each subspecies (Table 1) to study population structure and to obtain seeds for assessing life history char- acteristics under controlled conditions, the breeding system, and chromosome number. Seeds were ger- minated in a greenhouse at Colorado State Univer- sity, Fort Collins, Colorado in late November and transplanted to 6-inch diameter pots in early Janu- ary. Plants of subsp. australis and subsp. neomex- icana, which began bolting in March, completed their life cycles (April-May) under greenhouse conditions. Plants of subsp. longiflora continued to produce basal leaves throughout the spring and ear- ly summer months, but did not flower. Ten plants from each of two populations were retained under greenhouse conditions; a total of 21 additional plants were transplanted to a garden in July. Self- compatibility and autogamous seed production rel- ative to manipulative self-pollination were assessed in plants by enclosing inflorescences in loose-fitting bags constructed from cheesecloth. Self-compati- 2001] WILKEN: JPOMOPSIS LONGIFLORA SSP. NEOMEXICANA We) TABLE 2. COMPARATIVE TABLE OF DIAGNOSTIC CHARACTERS OF [POMOPSIS LONGIFLORA SUBSPECIES. The mean, standard deviation, and sample size (in parentheses) are given for each quantitative character. Midcauline leaf segment number -and number of ovules per ovary are reported as modes, followed by range and sample size in parentheses. Character I. Ll. australis I. 1. longiflora I. 1. neomexicana Plant height (cm) 24.1 + 9.4 (45) 61.0 + 15.6 (71) 2A 95+ 9383) Basal stem diameter (mm) 2.6 + 0.1 (45) LO O2.0 1) 2.5: 221015083) Midcauline leaf segment number 5 (3-7; 51) 7 ©-9; 71) 5 3-7; 93) Midcauline leaf length (mm) DOW) ea 3a(oih) ee pase 0 5) 237512s 5:00(93) Calyx lobe length (mm) Ol sO (23) 2a 270352) 505 2705152) Calyx lobe apex (internal) pubescence dense glabrous to sparse moderate to dense Modal ovule number 24 (17-24; 50) 30 (22-30; 50) 30 (23-30; 50) Capsule length (mm) 8.8 + 1.1 (30) Mes ea les(34) IDS 17636) Capsule length/fruiting calyx length deca Osi 30) 1.8 + 0.3 (34) 9° 0236) Seed apex wing length (mm) 0.1 + 0.0 (50) 0.3 + 0.1 (50) 0.6 =70,2 (50) bility was assessed by emasculating flower buds prior to anthesis and anther dehiscence and directly applying pollen to stigmas using dehisced anthers from the same plant. Flowers were also permitted to complete anthesis without manipulation. Mature capsules were dissected and the number of seeds in each and counted (Table 3). The ovary of a third flower on the same plant was dissected to estimate ovule number. Chromosome counts were made by studying actively growing root tips and flower buds, using orcein in proprionic acid (modified from Smith 1974). MORPHOLOGICAL RELATIONSHIPS, CHROMOSOME NUMBER, AND DISTRIBUTION Several diagnostic characters separate the three subspecies (Table 2). Plants of subsp. /ongiflora are generally more than twice as tall as those of subsp. australis and ssp. neomexicana. The lower 2—10 cm of the stems are thick, sclerified, and about twice as wide at the base than in either subsp. australis and subsp. neomexicana. Cauline leaves are longer and tend to have more segments. The lowest inflo- rescence branches in subsp. longiflora usually oc- cur above the middle, whereas in most plants of subsp. australis and subsp. neomexicana they occur well below the middle (Fig. 1). Calyx lobe length in subsp. australis and subsp. neomexicana aver- ages about | mm longer than in subsp. longiflora. All three taxa have calyx lobes that are minutely glandular on the outer surface. However, the inner surfaces of lobe apices are moderately to densely eglandular puberulent in subsp. australis and subsp. neomexicana. Those of subsp. longiflora are either glabrous or bear only a few eglandular trichomes. Twenty-four ovules per ovary are modal in subsp. australis, as compared to 30 ovules in subsp. /on- giflora and subsp. neomexicana (Table 3). The fruits of subsp. neomexicana resemble those of subsp. /ongiflora in length and shape (Fig. 1; Table 3). Capsules are ellipsoid and about twice the length of the fruiting calyx in both taxa. Seeds of subsp. longiflora and neomexicana are narrowly el- lipsoid, with a flat to slightly convex dorsal surface, and minutely winged along the margin and at the tips (Fig. 1). In subsp. australis, the capsules are ovoid to broadly ellipsoid, shorter in length, and scarcely exceed the calyx (Fig. 1; Table 2), and the seeds are essentially wingless (Fig. 1; Table 2). The typical subspecies is a western Great Plains endemic, ranging from southwestern South Dakota (Bennett and Todd Cos.) south through western Ne- braska, eastern Colorado, western Kansas, western Oklahoma, eastern New Mexico, and northwestern Texas (Fig. 2). It often occurs on sandy soils, es- pecially sand hills in the northern plains and sandy alluvials along principal tributaries of the upper Ni- obrara, Platte, Republican, Arkansas, Canadian, Red, and Brazos rivers. TABLE 3. COMPARISON OF AUTOGAMOUS VERSUS MANIPULATED SELF-POLLINATION IN [POMOPSIS LONGIFLORA. Mean ovule number, mean seed number per capsule, and mean seed set, expressed as a proportion of ovule number, + standard error. Sample size is uniformly 50 capsules. I. 1. australis Ovule number Mean ovule number 19.8 + 0.4 Autogamous self-pollination Mean seed number (1322, 0:6 Seed/ovule ratio Diez es) Y Manipulated self-pollination Mean seed number I 04 Seed/ovule ratio SOIC Ty I. 1. longiflora I. 1. neomexicana 211 = 04 2012105 LO 09 13a 208 38.0 222358 50.4 + 3.1 18.7 + 0.4 197 = 0D ©7297 2 Vhops | fungal Bee) 120 MADRONO / | es ssp. australis A ssp. longiflora ssp. neomexicana Fic. 2. Geographic distribution of Ipomopsis longiflora. The distribution of subsp. australis coincides with the northeastern portion of the Sonoran Sub- province, as defined by Thorne (1993). It occurs on sandy soils of open sites in desert shrublands, grasslands, and woodlands, generally south of the Mogollon Rim in Arizona, and extending south to Sonora and northwestern Chihuahua, Mexico (Fig. 2). Subspecies neomexicana occurs in similar hab- itats, but its distribution coincides with the south- eastern Great Basin Province and Chihuahuan Sub- province (Thorne 1993). It occurs from southeast- ern Utah and southwestern Colorado, through east- ern Arizona and New Mexico as far south as northern Coahuila. The distribution of each subspe- cies is largely allopatric, although some populations of subsp. australis and subsp. neomexicana are less than 40 km apart (Fig. 2; east-central Arizona, [Vol. 48 | southwestern New Mexico, and northern Chihua- hua). Grant (1959) reported 2n = 14 for plants now treated as subsp. longiflora (Colorado, Weld Co., Grant 9503, RSA) and subsp. neomexicana (New Mexico, Socorro Co., Grant 8813, RSA). Addition- al counts of the same number were obtained from representative populations of each subspecies (Ta- ble 1), VEGETATIVE AND REPRODUCTIVE BIOLOGY Field and common garden studies suggest that subsp. longiflora is at least a biennial, whereas subsp. australis and neomexicana are annuals. Ob- servations of subsp. /ongiflora in the sand hills of western Nebraska (Sheridan Co.) and eastern Col- orado (Logan Co.) suggest that it lives through at | | } | 2001) | least two growing seasons, separated by one winter, prior to reproduction. Germination occurs during the spring (April-June) following snowmelt and late spring storms. Plants develop into vegetative rosettes that expand by additional growth during the first summer, remain dormant during the winter, followed by renewed rosette growth, shoot elon- gation, and reproduction during the late spring and summer months (May—September) of the second year. Plants grown to maturity from 6-month-old transplants in a common garden in Ft. Collins, Col- orado required 18 months to flower. Plants grown under greenhouse conditions did not flower unless they were subjected to at least 10-12 weeks of cool, short-day conditions. In contrast, plants of subsp. australis and neomexicana, whose germi- nation may coincide with either spring or summer monsoon precipitation, apparently complete their life cycle within 1 year. Plants of subsp. australis (north of Globe, Gila Co.) were recruited as seed- lings during the spring and reproduced and died prior to the winter of the same year. Plants recruited during monsoon rains (July-August) over-wintered as rosettes and flowered in the succeeding spring and summer (March—September). Similar observa- tions were made of subsp. neomexicana in the Rio Grande Valley of south-central Colorado (Alamosa Co.), northwestern Arizona (Apache and Coconino Cos.), and central New Mexico (Sierra Co.). Ipomopsis longiflora is apparently pollinated by the hawkmoth, Hyles lineata (Grant and Grant 1965; Grant 1983). I observed unidentified hawk- moths in two populations (Table 1. subsp. longiflo- ra, Logan Co., Colorado; subsp. australis, Gila Co., Arizona). Hyles lineata and Manduca were fre- quent visitors in experimental populations of all three taxa in Ft. Collins, Colorado. No other polli- nator was observed on flowers of the three subspe- cies. Nectar is relatively high in sucrose, consistent with that expected in hawkmoth-pollinated flowers (Baker and Baker 1983; Freeman et al. 1985; Free- man and Wilken 1987). Autogamous (unmanipulated) pollination result- ed in seed/ovule ratios ranging from 38.0 in subsp. longiflora to 57.8 in subsp. australis and 50.4 in subsp. neomexicana (Table 3). Self-pollination of emasculated flowers resulted in seed/ovule ratios of 80.1 in subsp. australis, 67.9 in subsp. longiflora, and 76.1 in subsp. neomexicana (Table 3). These data provide further evidence of self-compatility re- ported by Grant and Grant (1965) and suggest that I. longiflora may experience a mixed mating system favoring selfing, resulting from self-compatibility, close proximity of style branches and upper an- thers, and pollination by hawkmoths. KEY TO JPOMOPSIS LAXIFLORA AND THE SUBSPECIES OF [POMOPSIS LONGIFLORA 1. Corolla tube 8-20 mm long, lobe 4—6 mm long; avules;4+—6 per locule 275. 2304 4.054. I. laxiflora WILKEN: JPOMOPSIS LONGIFLORA SSP. NEOMEXICANA Pa 1’ Corolla tube 25—45 mm long, lobe 6—11 mm long; ovules G=l10i per locule..7..4 e254 + I. longiflora 2. Mature capsule ovoid to broadly ellipsoid, slightly exceeding the fruiting calyx; seed tip scarcely winged (wings < 0.1 mm long) ... subsp. australis 2' Mature capsule ellipsoid, about twice the length of the fruiting calyx; seed tip winged (wings > 0.2 mm long) 3. Inflorescence branches arising above the middle of the central axis; base of stem 4—9 mm in dia- meter; most cauline leaves with 7 segments.... er tes a ar SNe Ce eee eh Re subsp. longiflora Inflorescence branches arising from throughout central axis; base of stem 1—4 mm in diameter; most cauline leaves with 5 segments......... subsp. neomexicana ~ 5) DISCUSSION The three taxa comprising [pomopsis longiflora share a combination of characters that are unique within the genus, including open, paniculate inflo- rescences, white to pale bluish, salverform corollas with tubes that exceed 20 mm in length, and pin- natifid leaves with 5—7 linear to narrowly oblong segments. Diagnostic differences among the three taxa are quantitative (Table 2), analogous to other infraspecific taxa within Jpomopsis (Grant and Wilken 1988). Ipomopsis longiflora ssp. neomexi- cana has no apparent autapomorphies and is sepa- rated by a combination of characters shared with either ssp. australis (e.g., habit, annual life form, inflorescence branching pattern, calyx and leaf morphology) or ssp. longiflora (ovule number, seed morphology, capsule length and shape). Its geo- graphical distribution lies between that of the other subspecies. Other closely related taxa treated as species within /pomopsis differ by one or more qualitative differences (e.g., leaf and inflorescence architecture, corolla morphology, anther position, filament length; see Grant 1959; Grant and Wilken 1988). Preliminary studies based on matK and ITS sequences show that the three taxa comprise a monophyletic lineage embedded within a largely unresolved polytomy of related taxa treated as spe- cies (M. Porter unpublished). Consequently, these taxa are treated as subspecies rather than as distinct species. Like Fletcher and Wagner (1984) and Freeman et al. (1985), this study did not reveal any evidence of hybridization, parapatry (sensu Endler 1977), or intermediacy at the population level. Nectar-sugar composition, however, was significantly higher for sucrose in both subsp. australis and subsp. neo- mexicana populations near their “‘contact zone”’ in southwestern New Mexico (Freeman et al. 1985), suggesting some form of interaction. Although se- lection for increasing sucrose levels may be hy- pothesized as a consequence of gene flow, cross- compatibility among the three subspecies is un- known. Studies of the late Quaternary (Spaulding and Graumlich 1986; Van Devender et al. 1987; Wright 1976) suggest that the modern climate and 122 MADRONO vegetation of the desert Southwest and western Great Plains developed during the last 10,000— 15,000 years. Thus, the close relationships among the three subspecies and their distribution suggest a relatively recent origin and/or expansion of range, coincident with relatively rapid development of arid interior southwestern biomes. Some populations of subsp. australis and subsp. neomexicana occur within 10—15 km of each other without apparent or significant differences in alti- tude, habitat, or vegetation. Notable areas of close proximity (Fig. 2) include the vicinities of Hol- brook (Navajo Co., AZ) Silver City and Lordsburg (Grant Co., NM), and Samalyuca (northern Chi- huahua, Mexico). Populations of subsp. neomexi- cana and subsp. longiflora occur within 30—40 km of each other in Chaves Co., NM. These potential contact zones may provide an opportunity to inves- tigate the proximate causes of their respective dis- tributions. ACKNOWLEDGEMENTS I thank Reggie Fletcher, Ed Freeman, Jim Henrickson, Steve O’Kane, and Warren L. Wagner for their advice, specimens, and discussions concerning this study and Bob Patterson and Mark Porter for their critical reviews. Jan Beckert skillfully composed and prepared the illustrations. LITERATURE CITED CRONQUIST, A., A. HOLMGREN, N. HOLMGREN, J. REVEAL, AND P. HOLMGREN. 1984. Intermountain flora. Jn Vas- cular plants of the intermountain west, U.S.A., Vol. 4. New York Botanical Garden, Bronx, NY. ENDLER, J. 1977. Geographic variation, speciation, and clines. Princeton University Press, NJ. FLETCHER, R. A. AND W. L. WAGNER. 1984. A new sub- species of Ipomopsis longiflora (Polemoniaceae) from Arizona, New Mexico, and northern Mexico. Madro- no 31:20-23. FREEMAN, C., W. REID, AND R. WORTHINGTON. 1985. Pat- terns of floral nectar-sugar composition of [pomopsis longiflora (Polemoniaceae) near the contact zone of its subspecies Jongiflora and australis. American Journal of Botany 72:1662—1667. [Vol. 48 AND D. WILKEN. 1987. Variation in nectar sugar | composition at the intraplant level in Ipomopsis lon- giflora (Polemoniaceae). American Journal of Bota- | ny. 74:1681—1689. | GRANT, V. 1959. Natural history of the phlox family. Mar- | tinus Nijoff, The Hague. . 1983. The systematic and geographical distribu- | tion of hawkmoth flowers in the temperate North American flora. Botanical Gazette 144:439—449. AND K. GRANT. 1965. Flower pollination in the phlox family. Columbia University Press, New York. | AND D. WILKEN. 1986. Taxonomy of the Jpom- opsis aggregata group (Polemoniaceae). Botanical Gazette 147:359-371. | AND . 1988. Racial variation in Ipomopsis | tenuituba (Polemoniaceae). Botanical Gazette 149: 443-449. | McKELvey, S. 1956. Botanical exploration of the trans- | Mississippi west, 1790-1850. Arnold Arboretum of | Harvard University, Jamaica Plain, MA. Situ, B. 1974. Cytological evidence. Pp. 237-258. In_ Radford, A., W. Dickison, J. Massey, and C. R. Bell, | Vascular plant systematics, Harper and Row, NY. | SPAULDING, W. AND L. GRAUMLICH. 1986. The last pluvial | episodes in the deserts of southwestern North Amer- — ica. Nature. 320:441—444. | THORNE, R. E 1993. Phytogeography. Pp. 132—153. In Flo- ra of North America Editorial Committee: Flora of | North America, Vol. 1. Oxford University Press, NY and Oxford. VAN DEVENDER, T., R. THOMPSON, AND J. BETANCOURT. 1987. Vegetation history of the deserts of southwest- ern North America: The nature and timing of the late Wisconsin-Holocene transition. Pp. 323-352. In W. Ruddiman and H. Wright (eds.), The geology of North America. Geological Society of America, Boulder, Co. WELSH, S., N. ATwoop, S. GOODRICH, AND L. HIGGINS. 1987. A Utah flora. Great Basin Naturalist Memoirs. 9:1-894. WILKEN, D. 1986. Polemoniaceae. Pp. 666—678. In The Great Plains Flora Association, Flora of the Great Plains, University Press of Kansas, Lawrence, KS. WRIGHT, H. 1976. The dynamic nature of Holocene veg- etation, a problem in paleoclimatology, biogeography, and stratigraphic nomenclature. Quaternary Research. 6:581-596. | “Maprono, Vol. 48, No. 2, pp. 123-127, 2001 A NEW SPECIES OF POA L. (POACEAE) FROM BAJA CALIFORNIA, MEXICO ROBERT J. SORENG Department of Botany, National Museum of Natural History, Smithsonian Institution, Washington, DC 20560-0166 ABSTRACT A new perennial grass species, Poa bajaensis, is described from Baja California, Mexico. It is compared to P. strictiramea Hitchc, and the type of P. orcuttiana Vasey. The species has long been known by the latter name in herbaria. It may be distinguished from these by its short upper leaf blades, more closed sheaths, and abaxially smoother leaves and smoother panicle branches. It is endemic and fairly frequent in the upper elevations of the Sierra San Pedro Martir. The new species is tentatively considered to be related to P. strictiramea, and to be near to P. sect. Homalopoa Dumort. RESUMEN Se discribe una nueva graminea perenne, Poa bajaensis, de Baja California, Mexico. Se compara con P. strictiramea Hitchc. y el tipo de P. orcuttiana Vasey. Este especies, por mucha tiempo se denomin6 P. orcuttiana en los herbarios. Se puede distinguir de las otras dos, por las cortas hojas de caulinares de la parte superior, por las vainas mas cercanas, y por las hojas abaxialmente mas lisas y las ramas de paniculas mas lisas. Es aparantemente endemica, y mas 0 menos frequente en las elevacfones superiores de da Sierra San Pedro Martir. La nueva especie tentativemente esta relacionada con P. strictiramea, y parece estar cercana a P. sect. Homalpoa Dumort. A new and distinctive species of Poa is here named after the Baja California Peninsula of Mex- ico. The new species is known only from the upper elevations of the Sierra San Pedro Martir. It has been collected there in the spring on several occa- sions, and seems to be fairly frequent. Specimens included here were treated as P. orcuttiana Vasey by Gould and Moran (1981), and have since passed under that name. So far as I am aware, J. D. Olm- sted, in 1962, was the first person to collect the new species. I realized it was probably new when I first examined a specimen of it while working on my dissertation at New Mexico State University in the early 1980s, but needed to see more material and to compare it with the type specimen of P. orcut- tiana. Poa bajaensis Soreng, sp. nov. (Fig. 1)—-TYPE: Mexico, Baja California, Sierra San Pedro Mar- tir, east rim above Yerba Buena, 31°01’'N, 115°26'W, elev. 2700 m, common in duff under trees, | Jun 1968, Reid Moran 15070 (Holotype, US-259736; isotype: SD-69304). A Poa Strictiramea similis sed in paniculis 4—13 cm longis (versus 10—30 cm) foliorum vaginis no- dorum superiorum connatis 29-36% (versus 10% vel minor) laminis nodorum superiorum abbrevia- tis, 0.1—-1.5(4) cm longis multo brevioribus quam vaginis (versus plerumque longioribus) laminis abaxialiter laevibus (versus plerumque scabris) dif- fert. Perennial, densely tufted, without rhizomes. Ba- sal tuft of leaves 5—10(15) cm tall. Basal sheaths often persisting into the next growth season, papery (not anastomosing). New vegetative shoots emerg- ing intravaginally; prophylls 1—2 cm long, split abaxially, scabrous and pilose on the 2 keels. Culms 20-50 cm tall, top 1—2 leaf nodes well exposed or barely enclosed, all lower nodes enclosed in their sheaths. Leaves: sheaths slightly keeled, very sparsely to moderately (rarely densely), evenly, finely scabrous, uppermost culm sheaths 8—15 cm, margins fused 29-36% of the length; collars not noticeably more scabrous than their sheaths; ligules of vegetative and lower culm leaves membranous, 0.25—0.5 mm, of upper culm leaves 0.5—1.5(2) mm, apex truncate to obtuse, backs and apical margins densely scabrous; culm and vegetative leaf-blades simular, flat or folded, moderately thick, with in- volute margins, abaxially smooth or very sparsely and finely scabrous, to moderately scabrous apical- ly, veins weakly expressed adaxially and abaxially, adaxially smooth or moderately to densely finely scabrous, longest blades less than 5 cm, uppermost culm blade 0.1—1.5(4) cm, blades 1.5—2.75 mm wide (expanded), narrowly and abruptly prow- tipped. Panicles 4—13 cm, not or slightly to highly anthocyanic, erect, open, sparse, lower axis smooth, lowest internodes 1.8—3.9(5.2) cm; branches 2—3(5) per lower node, widely spreading to reflexed, fairly strict, smooth or very sparsely finely scabrous prox- imally, smooth or moderately (rarely densely) fine- ly scabrous distally, terete or weakly angled, hooks not confined to rows on angles, the longest branch- es 3-7 cm, with 5—15 spikelets. Spikelets 3.75—8 mm; glumes and lemmas distinctly keeled, glumes thin, smooth or sparsely scabrous on the upper keel 124 |e) (Cree Spikelet. and sometimes on the lateral veins, and sometimes between them, % to %4 the length of adjacent lem- mas, the first 2.5—-3 mm, 1(—3)veined, the second 2.8—3.5 mm, 3-veined, more than two times as wide as the first; rachillas smooth, longest internodes . ily \ Ws j MADRONO S| \Y y SY Y 2. Sy): Y > WF ~ oy <= Ss > \ // ; ica a \ ss —“ N Tee N WV y X**¥ \ NN BN SS =) Wy pe | y ; hy & yA \/ \ \ ) \\ (/ V/ \ UN Y) Me } | g INT RE Jf y Poa bajaensis Soreng, illustrated from the holotype collection, Reid Moran 15070 (US-259736). A. Habit. B. 1.25—2 mm; florets (1)2—4(6); calluses of some plants all glabrous, of other plants infrequently to frequently with 1—several woolly dorsal hairs up to 2 mm long; lemmas 3.2—4.2 mm, 5-veined, dis- tinctly keeled, apically sparsely scabrous on the [Vol. 48) | 2001) SORENG: POA BAJAENSIS NEW SPECIES 125 keel, lateral veins weakly expressed, the body and other veins smooth, entirely glabrous in one plant, “or softly puberulent, hairs to 0.5 mm on the keel and marginal veins, to 0.2 mm on the sides between ' the veins and sparsely to densely covering the low- sandy to rocky to clayey, sometimes duff-covered, granitic soils, 1450—2950 m elev. Flowering May to June. Discussion. Herbarium specimens of the new er 4-45, apex acute, entire; palea approximately equaling the lemma in length, keels nearly smooth | to closely and finely scabrous, glabrous or sparsely | puberulent near the middle on and between the _keels. Flowers perfect; anthers |.7—3.2 mm (rarely sterile, but then ca. 1.7—1.8 mm). Paratypes. Mexico, Baja California [Norte]: Si- erra San Pedro Martir: Crest of range N of obser- vatory, head of Canada el Copal and S slope of Cero Venado Blanco, 2500—2700 m, 3 Jun 1988, S. Boyd, T. Ross, K. McCulloh 2311 (RSA); La Con- cepcion, 31°01’N—115°37'W, ca. 1450 m, 31 May 1968, Reid Moran 15006 (SD); open W slope of Cerro 2828, ca. 31°02’N—115°27'W, ca. 2800 m, 31 May 1968, Reid Moran 15060 (BH, SD); 2 mi W of Vallecitos, 31°00’N—115°29’'W, ca. 2250 m, 2 Jun 1968, Reid Moran 15083 (SD); 3 km NE of El Alto de Corona, 31°00’N—115°41'W, ca. 2400 m, 20 Aug 1977, Reid Moran 24555 (SD); W slope below summit of El Picacho, 30°59'30"N— 115°22'30"W, ca. 2950 m, 5 May 1978, Reid Moran 25611 (SD); end of road into high end of northern sierra, ca. 64 mi. from end of paved road to Ensen- ada, 7200 ft (2210 m), 6 Jun 1962, J. D. Olmsted 4561 (RSA; somewhat intermediate to P. secunda); Central [region], ca. 3 mi. ESE of Prado del Co- rona, ca. | mi. up canyon from southernmost aspen colony, tributary of Rio San Rafael, 8100 ft (2490 m), 9 Jun 1962, J. D. Olmsted 4711 (RSA); S of Vallecitos near Cerro la Botella Azul, 30°57'’20”N— 115°25'26"W, ca 2440 m, 27 Jun 1998, J. Rebman & A. Russell 5384 (US); near crest of mountain range, approx. 2 mi SE of the observatory, 31°14’N-115°64'W, ca. 2985 m, 28 Jun 1998, J. Rebman & A. Russell 5384 (SD); ‘‘Corral Mead- ow”, 7.5 km NW (340°) of the observatory, 31°06'45”"N—115°29'S50”"W, 16 Jun 1988, A. C. Sanders, R. Minnich, E. Franco M. Salazar 7895 (RSA, SD); Vallecitos, ca. 31°02’N—115°28'W, ca. 2430 m, 18 Jun 1985, R. F. Thorne, R. Dahlgren, S. Boyd & D. Charlton 60858 (RSA, SD); Valle- citos, ca. 31°02’N—115°27.5'W, ca. 2430 m, 1 Sep 1985, R. F. Thorne, M. Z. Thorne, L. Thorne & T. Petrella 61394 (RSA); just above observatory liv- ing quarters, ca. 31°02’N—115°28'W, ca. 2600 m, 7 May 1986, R. F. Thorne, T. S. Elias & Paulino Rojas 61967 (MO-3333160, RSA); near gate to UNAM Observatory, 31°02’N—115°29'30"W, 2520 m, 29 May 1982, G. Yatskievych, S. Forbes, M. Gallagher, J. Evans & A. Kelley 82-190 (SD). Habitat characteristics. Mountain slopes, flats and drainages, in Salix and Populus tremuloides thickets, Quercus-Pinus jeffreyi forests, and Quer- cus-P. jeffreyi-P. lambertiana-Abies concolor for- ests, and open meadows of Pinus jeffreyi park, in species had been determined as: P. fendleriana (Steud.) Vasey (P. sect. Madropoa Marsh ex So- reng); P. interior Rydb. (P. sect. Stenopoa Du- mort.); or most commonly as P. orcuttiana Vasey (=P. secunda J. Presl subsp. secunda; P. sect. Se- cundae Marsh ex Soreng). Although it seems to have little in common with any of these taxa, it is superficially similar in a few respects to the type of P. orcuttiana. Poa orcuttiana was placed in synonymy under P. scabrella Vasey by Hitchcock (1951) and Keck (1959), and (along with P. scabrella) in P. secunda J. Pres] by Kellogg (1986). The type of P. orcut- tiana is typical of material that has been recognized as P. scabrella (Hitchcock 1951; Keck 1959; Gould and Moran 1981). It fits well within the range of morphological variation found in P. secunda subsp. secunda in which P. scabrella is included by So- reng (1994; P. orcuttiana was considered a minor synonym of P. scabrella sensu Hitchcock or Keck, and thus was not included in the brief synonymy published there). However, Gould and Moran (1981) recognized both P. scabrella and P. orcut- tiana in their treatment of grasses of Baja Califor- nia, Mexico. Gould and Moran’s description of P. orcuttiana, its habitat, range (except for the stated range in southern California), and key, match the new species perfectly. To clear up this contradiction it is necessary to compare the type of P. orcuttiana with the new species in detail. The type specimen of P. orcuttiana Vasey (Ho- lotype: USA, California, San Diego, Chollas [a site below 150 m elevation], 26 May 1884, C. R. Orcutt 1070 (US-556833), West. Amer. Scientist 3: 165. 1887.), has long, narrow (O.5 mm), very thin, and sparsely to moderately densely and coarsely sca- brous blades, the basal tuft 20 cm tall, upper culm blades 4.5—7 cm long (25 to % the length of their sheaths), with acute or lacerate ligules 2.75—3.25 mm long. The sheaths are open to near the base and are moderately densely, coarsely scabrous. The panicles are contracted, or, on one culm in anthesis (probably lending to past confusion with the new species), open with somewhat spreading branches. (It is usual for P. secunda panicles to open up at anthesis and later contract, though they remain open in some ecotypes of sheltered habitats.) The panicle branches are densely and coarsely scabrous throughout. The spikelets have 5—8 florets, with sparsely scabrous rachillas, first glumes with 3 veins, calluses glabrous or with a short crown of hairs surrounding the lemma base, and lemmas rounded on the back and quite scabrous, with short sparse pubescence mainly on the keel and marginal veins. The new species differs in several respects from 126 TABLE 1. MADRONO of the P. scabrella form of P. secunda subsp. secunda are indicated by * Characters Culm length Blade shape and thick- ness Blade abaxial scabrosity Basal leaf tuft height Leaf blade and sheath persistence Top leaf blade length Top blade-to-sheath length ratio Top ligule length Top ligule apex Sheath closure from base Panicle length, shape, and branch angle of divergence Panicle branch scabrosity Spikelet no. of florets Rachilla vestiture First glume no. of veins Callus pubescence Lemma shape Lemma scabrosity Species P. bajaensis 20-50 cm flat or folded with involute margins, moderately thick smooth or very sparse throughout to moderate apically, fine 5—10 (-15) cm persisting 0.1—-1.5 (—4) cm < 1:5 (rarely more) 0.5—1.5 (-2) mm truncate to obtuse 29-36% 4—13 cm, open, branches widely spreading to re- flexed smooth or sparse proximally, smooth or moderate to dense distally, fine (1—) 2—4 (-6) smooth, glabrous 1 (—3 faint) glabrous, or with a sparse dorsal web keeled smooth Type specimen P. orcuttiana 60—70 cm* flat*, thin* sparse to moderately-dense*, coarse* 20 cm* soon withering* 4.5—7 cm* 219-1227 2.75—3.25 mm* acute or lacerate* closed only near the base or up to 10%* 14 cm*, contracted*, branch- es ascending, or during an- thesis open with branches somewhat spreading* dense* and coarse* through- out* 5—8* sparsely scabrous*, glabrous 3% glabrous, or with a crown of short hairs* rounded on back* distinctly scabrous* [Vol. 48, COMPARISON OF POA BAJAENSIS WITH P. STRICTIRAMEA AND THE TYPE OF P. ORCUTTIANA. Character states typical | ! Species P. strictiramea 30-90 cm flat or folded with involute margins, moderately thick sparse to dense, coarse 15-30 cm persisting 7—15 cm [complete data un- available] mostly > 1:1 0.5—4 (-—6) mm truncate to acute closed only near the base or up to 10% (7—) 10—30 cm, open, branches spreading moderate to dense on the an- gles throughout, fine to coarse 2-5 smooth or scabrous, some- times sparsely hirtellous 1-3 glabrous, rarely with a sparse dorsal web keeled smooth or sparsely to dense- the type of P. orcuttiana (Table 1). The scabrosity of the leaves is much finer, and sparse or absent on the abaxial surface and denser on the adaxial sur- face; the sheaths are closed over a greater portion of their length; the leaf-blades are shorter (the up- per culm blades mostly less than 1/5 the length of their sheaths), broader, thicker, firmer, and flat or more often folded; and the ligules are shorter and truncate. The inflorescence remains open and the branches are widely spreading to reflexed and are smooth or only sparsely scabrous distally (moder- ately scabrous in Olmsted 4561, which also has a 4-cm-long upper blade [25 the sheath in length], sterile anthers, and is perhaps intermediate between P. bajaensis and P. secunda). The spikelets usually have fewer (2-4) florets; the first glumes are mostly 1-veined; the rachillas are smooth; the lemmas are keeled; and the calluses sometimes have a dorsal web (never a crown of hairs). Poa scabrella, as treated by Keck (1959) and by Gould and Moran (1981), is said to occur mainly below 1500 m, and is not known to reach into high- er elevations of the Sierra San Pedro Martir. The new species occurs mainly above 2200 m. ly scabrous The degree of closure of the sheaths, the smooth- ness of the branches, and the presence of 1-veined first glumes exclude this species from P. sect. Sten- opoa. In addition, the keeled spikelets and lemmas and the occurrence of a dorsally isolated web on the callus exclude this species from P. sect. Secun- dae. There is no evidence of dicliny in P. bajaensis, eliminating the possibility it belongs to P. sect. Madropoa, P. sect. Dioicopoa E. Desv., or the P. nervosa complex of P. sect. Homalopoa Dumott., all of which exhibit pistillate flowers (with rudi- mentary anthers) in many plants (Soreng 2000). The new species is most similar to Poa strictir- amea Hitche. (Table 1; syn. P. involuta Hitchc.) of the mountains in and around the Chihuahuan Desert and the eastern Sierra Madre Occidental. The near- est site I am aware of for the latter species is in northwest Chihuahua, near Babicora, over 850 km ESE. The new species is tentatively considered to be allied to P. strictiramea and its rhizomatous rel- ative P. ruprechtii Peyr. of the Sierra Madre Ori- ental. It is placed incertae sedis, near to the latter two species, these near to P. sect. Homalopoa Du- mort. 2001) ACKNOWLEDGEMENTS My thanks are extended to Paul Peterson and anony- | mous reviewers for careful review of the manuscript, Har- ‘old Robinson for correcting the Latin diagnosis, Ximena ' Lofidono for correcting the Spanish, Nancy Soreng for the | illustration, and the curators of RSA and SD for specimen ) _ loans. LITERATURE CITED ~GOULD, E W. AND R. Moran. 1981. The grasses of Baja California, Mexico. San Diego Society of Natural History, Memoir 12:1—140. Hitcucock, A. S. 1951. Manual of the Grasses of the United States (2nd ed. revised by A. Chase). SORENG: POA BAJAENSIS NEW SPECIES 127 U.S.D.A. Misc. Pub. 200, United States Government Printing Office, Washington, DC. Keck, D. D. 1959. Poa. In P. A. Munz (in collaboration with D. D. Keck), A California Flora, and Supple- ment. University of California Press, Berkeley, CA. KELLOGG, E. A. 1985. A biosystematic study of the Poa secunda complex. Journal of the Arnold Arboretum 66:201—242. SORENG, R. J. 1994. Poa. In J. C. Hickman (ed.), The Jepson Manual, Higher Plants of California, Univer- sity of California Press, Berkeley, CA. . 2000. Apomixis and amphimixis comparative biogeography: A study in Poa (Poaceae). Pp. 294— 306. In S. W. L. Jacobs and J. Everett (eds.), Grasses: Systematics and Evolution, CSIRO, Melbourne. MADRONO, Vol. 48, No. 2, pp. 128-129, 2001 REVIEW Trees and Shrubs of California, by John D. Stuart and John O. Sawyer. Illustrated by Andrea J. Pick- art. 2001. 467 pp. University of California Press, Berkeley. Hardcover $45.00 ISBN 0-520-22109-5, softcover $22.50 ISBN 0-520-22110-9. So far, being seriously interested in woody taxa in California meant carrying a subset of the below listed publications plus The Jepson Manual to the field. Will this new guide to trees and shrubs of California replace all that? Yes, to some extent. This will be a useful book for beginners. Almost all native California trees are here and many com- mon shrubs. Some commonly naturalized woody species are here as well. Keys are friendly, based on readily available vegetative characters. Repro- ductive structures are needed only exceptionally. Technical terms are kept to minimum; all of them are explained in the glossary. Nomenclature fol- lows The Jepson Manual, with some justifiable ex- ceptions (e.g., rehabilitation of the genus Chamae- cyparis). About *%3 of the included species are illus- trated by line drawings. Unfortunately, some, like those of Cercis occidentalis, Genista monspessu- lana, or Salix gooddingii, are of rather marginal quality to say the least. Thirty nine color photo- graphs are, in general, excellent. The distribution of 313 species in California is illustrated by small range maps. For many of the species, these are the first sketchess of distributional maps ever complet- ed. Some more or less relevant references are listed; several really useful ones are missing (e.g., Benson and Darrow 1981; McMinn and Maino 1980; Pe- trides and Petrides 1992; Sampson and Jespersen 1981; Sudworth 1967). I do not expect any complaints about coverage of California trees. Only very rare species like Lyonothamnus floribundus are missing. However, if you do not know some less common shrubs, like Forestiera pubescens, Fouquieria splendens, Lotus scoparius, Peraphyllum ramosissimum, Ribes mal- vaceum, or Romneya coulteri, you will be lost. They are not included. Some genera with common woody species like Amorpha, Brickellia or Penste- mon are not treated at all. Nevertheless, Neviusia cliftonii is here. If you collect one of the more than 40 excluded species of Arctostaphylos, you should retreat to The Jepson Manual or, even better, to Wells’ (2000) book on manzanitas. If it happens that you find Quercus palmeri, or Q. tomenlella, you will certainly be more successful with Pavlik et al. (1991) or Roberts’ (1995) manuals. Stuart and Sawyer’s guide, as the authors themselves admit, is clearly less useful in southern California where oth- er sources will have to be consulted (e.g., Benson and Darrow 1981; Conrad 1987; Elmore and Janish 1976). Several invasive woody species, even if cur- | rently not really widespread in California, could be | included in a manual like this (e.g., Acacia deal- | bata, Alhagi pseudalhagi, Catalpa bignonioides, | Cotoneaster spp., Crataegus monogyna, Elaeagnus — angustifolia, Fraxinus uhdei, Ligustrum spp., My- | oporum laetum, Nicotiana glauca, Sapium sebifer-— um, Schinus molle, Sesbania punicea). Their early — detection can be critical for their successful control | or eradication. | There are regions, like Great Britain, south-east- ern Australia, New Zealand, or Kenya with a long | tradition of excellent complete (or almost complete) — field guides to trees and shrubs. Recently published | Brayshaw’s (1996) guide to woody species of Brit- ish Columbia also belongs in this category. Stuart | and Sawyer’s guide is just a first step in the right — direction. —MARCEL REJMANEK, Section of Evolution and Ecology, University of California, Davis, CA 95616 LITERATURE CITED BENSON, L. D. AND R. A. DARROW. 1981. The Trees and Shrubs of the Southwestern Deserts. 3rd ed. Univer- sity of Arizona Press, Tucson, AZ. BRAYSHAW, T. C. 1996. Trees and Shrubs of British Co- lumbia. UBC Press, Vancouver, BC. ConrAD, C. E. 1987. Common Shrubs of Chaparral and Associated Ecosystems of Southern California. USDA Forest Service General Technical Report PSW-99. Pacific Southwest Forest and Range Exper- iment Station, Berkeley, CA. ELMorE, E H. AND J. R. JANISH. 1976. Shrubs and Trees of the Southwest Uplands. Southwest Parks and Mon- uments Association, Tucson, AZ. McMinn, H. E. 1939, 1964. An Illustrated Manual of Cal- ifornia Shrubs. University of California Press. Berke- ley, CA. AND E. MAINO. 1980. Pacific Coast Trees. 2nd ed., 13th printing. University of California Press. Berke- ley, CA. PAVLIK, B. M., P. C. Mutck, S. JOHNSON, AND M. POPPER. 1991. Oaks of California. Cacuma Press, Los Olivos, CA. PETRIDES, G. A. AND O. PETRIDES. 1992. A Field Guide to Western Trees. Houghton Miffin Co., Boston, MA. RAVEN, P. H. 1966. Native Shrubs of Southern California. University of California Press. Berkeley, CA. Roserts, FE M. 1995. Illustrated Guide to the Oaks of the Southern Californian Floristic Province. F M. Roberts Publications, Encinitas, CA. SAMPSON, A. W. AND B. S. JESPERSEN. 1963, 1981. Cali- fornia Range Brushlands and Browse Plants. Division 2001] REVIEW 129 i of Agricultural Sciences, University of California, © SUDWoRTH, G. B. 1967. Forest Trees of the Pacific Slope. H} Berkeley, CA. Dover Publications, New York, NY. ‘Tuomas, J. H. AND D. R. PARTELL. 1974. Native Shrubs WELLS, P. V. 2000. The Manzanitas of California. Pub- | of the Sierra Nevada. University of California Press, lished by the Author. ISBN: 0-933994-22-2. Berkeley. Volume 48, Number 2, pages 51-130, published 15 January 2002 SUBSCRIPTIONS—-MEMBERSHIP Membership in the California Botanical Society is open to individuals ($27 per year; family $30 per year; emeritus $17 per year; students $17 per year for a maximum of 7 years). Late fees may be assessed. Members of the Society receive Maprono free. Institutional subscriptions to MADRONO are available ($60). Membership is based on a calendar year only. Life memberships are $540. Applications for membership (including dues), orders for sub- scriptions, and renewal payments should be sent to the Treasurer. Requests and rates for back issues, changes of address, and undelivered copies of MApDRONO should be sent to the Corresponding Secretary. INFORMATION FOR CONTRIBUTORS Manuscripts submitted for publication in MADRONO should be sent to the editor. It is preferred that all authors be members of the California Botanical Society. Manuscripts by authors having outstanding page charges will not be sent for review. Manuscripts may be submitted in English or Spanish. English-language manuscripts dealing with taxa or topics of Latin America and Spanish-language manuscripts must have a Spanish RESUMEN and an English ABsTRACT. Manuscripts and review copies of illustrations must be submitted in triplicate for all articles and short items (NOTES, NOTEWORTHY COLLECTIONS, POINTS OF VIEW, etc.). Follow the format used in recent issues for the type of item submitted. Allow ample margins all around. Manuscripts MUST BE DOUBLE-SPACED THROUGHOUT. For articles this includes title (all caps, centered), author names (all caps, centered), addresses (caps and lower case, centered), abstract and resumen, 5 key words or phrases, text, acknowledgments, literature cited, tables (caption on same page), and figure captions (grouped as consecutive paragraphs on one page). Order parts in the sequence listed, ending with figures. Each page should have a running header that includes the name(s) of the author(s), a shortened title, and the page number. Do not use a separate cover page or ‘erasable’ paper. Avoid footnotes except to indicate address changes. Abbreviations should be used sparingly and only standard abbrevia- tions will be accepted. Table and figure captions should contain all information relevant to information presented. All measurements and elevations should be in metric units, except specimen citations, which may include English or metric measurements. Authors of accepted papers will be asked to submit an electronic version of the manuscript. Microsoft Word 6.0 or WordPerfect 6.0 for Windows is the preferred software. Line copy illustrations should be clean and legible, proportioned to the MAprono page. Scales should be in- cluded in figures, as should explanation of symbols, including graph coordinates. Symbols smaller than 1 mm after reduction are not acceptable. Maps must include a scale and latitude and longitude or UTM references. In no case should original illustrations be sent prior to the acceptance of a manuscript. Illustrations should be sent flat. No illustrations larger than 27 X 43 cm will be accepted. Presentation of nomenclatural matter (accepted names, synonyms, typification) should follow the format used by Sivinski, Robert C., in MaproNo 41(4), 1994. Institutional abbreviations in specimen citations should follow Holmgren, Keuken, and Schofield, Index Herbariorum, 8th ed. Names of authors of scientific names should be abbreviated according to Brummitt and Powell, Authors of Plant Names (1992) and, if not included in this index, spelled out in full. Titles of all periodicals, serials, and books should be given in full. Books should include the place and date of publication, publisher, and edition, if other than the first. All members of the California Botanical Society are allotted 5 free pages per volume in Maprono. Joint authors may split the full page number. Beyond that number of pages a required editorial fee of $40 per page will be assessed. The purpose of this fee is not to pay directly for the costs of publishing any particular paper, but rather to allow the Society to continue publishing MApRONo on a reasonable schedule, with equity among all members for access to its pages. Printer’s fees for illustrations and typographically difficult material @ $35 per page (if their sum exceeds 30 percent of the paper) and for author’s changes after typesetting @ $4.50 per line will be charged to authors. At the time of submission, authors must provide information describing the extent to which data in the manu- script have been used in other papers that are published, in press, submitted, or soon to be submitted elsewhere. VOLUME 48, NUMBER 3 JULY-SEPTEMBER 2001 MADRONO A WEST AMERICAN JOURNAL OF BOTANY PHYSIOLOGICAL AND ANATOMICAL ASPECTS OF CAM-CYCcLING IN LEWISIA COTYLEDON VAR. COTYLEDON (PORTULACACEAE) Lonnie J. Guralnick, Chad Marsh, Roxann Asp, and Aaron Karjala ...... 131 THE SIGNIFICANCE OF POPULATION SUCCESSIONAL STATUS TO THE EVOLUTION OF SEED- LING MORPHOLOGY IN PINUS CONTORTA VAR. LATIFOLIA (PINACEAE) LEEETEO LUNA) GIST UALS) Titers nesters te aa eda tass tes a ates es or eR ne RE HN 138 Estupio ANATOMICO DE SWALLENIA (POACEAE: ERAGROSTIDEAE: MONANTHOCHLOINAE), UN GENERO MOoNoriPICco DE NorTE AMnicn y Maricela Goémez-Sdnchez, Patricia Davila. -Aranda, and Jesus Valdés- TRON TLR arate etendetoms sie caanen cures tie ce meen esos ee : i = 4 es eee re eee 152 COMPARATIVE FLOWERING PHENOLOGY OF PLANTS IN ‘tHe WESTERN Monve DESERT W. Bryan Jennings ...ccccccccccccssssccceeeeenens af pone tones ¥ SY) fees eee 162 SURVEY OF JUNIPERUS COMMUNIS L. (CupREssacEAE) VARs 0M THE WESTERN UNITED States UsiInG RAPD FINGERPRINTS — pe Ga Vanessa E. T. M. Ashy Ose SS Bart Cc ao Brien, and Elizabeth A. Friar..... 172 ConiIFER TREE DISTRIBUTIONS IN SO OUTHERN C CALIFORNIA DG Richard A. Minti ad Go Richard G. Everett oe oe Te Pesce leigh NEW SALTUGILIA LATIMERI: A NEW Me, Pou WW SER <0 ¥ , ASI SPECIES Terri L. Wee ese and Leigh A. Johnsor Fe ao Ww pesheteineasecest. wast 198 TAXONOMIC CHANG 4 ES i A New SPECIES IN ae NIA SECT. “AMPHIACHAENIA “f ELIANT AE Ay SENSU LATO, (Composttak: » eee errr rere eeccc ccc reeeseececesesessssssesescessseess ‘“EWORTHY CALIFORNIA .........000004es 4 )) ee | Qs LECTIONS IME XIC OMe tM sounds: oD ARGS. Sk INRIZON AR rereen hen cae ea thee aie eee x yy j SS SPR. ee WASHINGTON «.cecsecccscseceseceesseesaesssesoces QUEM ccscccseseteseeeseten Woe ee EVGA ORs eres sects cate ere ee atic ee eee gece meee ee eras Dali YUNCEMENTS EIGHTY- YEAR INDEX TO MADRONO ........sssseseeeeeececcccceccccecceseeeesererstsscsscccccccececeeeeeeens 219 INGE VG IEUTD IT vcore eae Pet ace ge areca ae ce seats hrs Reet Ci at sien Sak Mada ak Tae ines aes TOON 219 ae PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL SOCIETY Maprono (ISSN 0024-9637) is published quarterly by the California Botanical Society, Inc., and is issued from the office of the Society, Herbaria, Life Sciences Building, University of California, Berkeley, CA 94720. Subscription information on inside back cover. Established 1916. Periodicals postage paid at Berkeley, CA, and additional mailing | offices. Return requested. Postmaster: Send address changes to MApDRONO, Roy Buck, % University Herbarium, | University of California, Berkeley, CA 94720. Editor—KRrisTINA A. SCHIERENBECK California State University, Chico Department of Biology Chico, CA 95929-0515 kschierenbeck @csuchico.edu Editorial Assistant—JEANNIE TRIZZINO Book Editor—Jon E. KEELEY Noteworthy Collections Editors—DIETER WILKEN, MARGRIET WETHERWAX Board of Editors Class of: 2001—Robert Patterson, San Francisco State University, San Francisco, CA PAULA M. SCHIFFMAN, California State University, Northridge, CA 2002—NorMAN ELLSTRAND, University of California, Riverside, CA Cara M. D’ Antonio, University of California, Berkeley, CA 2003—-FREDERICK ZECHMAN, California State University, Fresno, CA Jon E. KEELEY, U.S. Geological Service, Biological Resources Division, Three Rivers, CA 2004—Davip M. Woon, California State University, Chico, CA INGRID Parker, University of California, Santa Cruz, CA CALIFORNIA BOTANICAL SOCIETY, INC. OFFICERS FOR 2001—2002 President: Bruce BALDwin, Jepson Herbarium and Dept. of Integrative Biology, 1001 Valley Life Sciences Bldg. #2465, University of California, Berkeley, CA 94720. First Vice President: Rov Myatt, San José State University, Dept. of Biol. Sciences, One Washington Square, San José, CA 95192. rmyatt@email.sjsu.edu Second Vice President: PETER Fritscu, Dept. of Botany, California Academy of Sciences, Golden Gate Park, San Francisco, CA 94118-4599. pfritsch @calacademy.org Recording Secretary: DEAN KELcu, Jepson and University Herbarium, University of California, Berkeley, CA 94720. dkelch @sscl.berkeley.edu Corresponding Secretary: SUSAN BAINBRIDGE, Jepson Herbarium, 1001 VLSB #2465, University of California, Berkeley, CA 94720-2465, 510/643-7008. suebain@SSCL.berkeley.edu Treasurer: Roy Buck, % University Herbarium, University of California, Berkeley, CA 94720. The Council of the California Botanical Society comprises the officers listed above plus the immediate Past President, R. Joun LittLe, Sycamore Environmental Consultants, 6355 Riverside Blvd., Suite C, Sacramento, CA 95831; the Editor of MAprono; three elected Council Members: BIAN Tan, Strybing Arboretum, Golden Gate Park, San Fran- cisco, CA 94122; James SHEvockK, National Park Service, 1111 Jackson St., Suite 700, Oakland, CA 94607-4807, 510/ 817-1231; ANNE BRADLEY, USDA Forest Service, Pacific Southwest Region, 1323 Club Drive, Vallejo, CA 94592, abradley @fs.fed.us; Graduate Student Representative: KirsSTEN JOHANNUS, Jepson Herbarium, University of California, Berkeley, CA 94720. This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). Maprono, Vol. 48, No. 3, pp. 131-137, 2001 PHYSIOLOGICAL AND ANATOMICAL ASPECTS OF CAM-CYCLING IN LEWISIA COTYLEDON VAR. COTYLEDON (PORTULACACEAE) LONNIE J. GURALNICK!, CHAD MARSH, ROXANN ASP, AND AARON KARJALA Division of Natural Science and Mathematics, Western Oregon University, Monmouth, OR 97361 ABSTRACT We investigated leaf morphology, CO, uptake, and the diurnal acid fluctuation of the montane species, Lewisia cotyledon Robinson. This species belongs to the Portulacaceae, a small family with ~ 425 species, which exhibit photosynthetic flexibility. Lewisia cotyledon was found in the Siskiyou Mts. in Josephine County, OR along with other montane species that utilize the CAM pathway. Plants were collected and returned to the greenhouse to study the photosynthetic metabolism of this species under both well-watered and drought conditions. Lewisia cotyledon showed primarily daytime CO, gas exchange. Well-watered plants exhibited a significant diurnal acid fluctuation during the day. Phosphoenolpyruvate (PEP) carbox- ylase (EC 4.1.1.31) activity was low in both control and water-stressed plants. Leaf anatomy showed a CAM-like spongy mesophyll tissue with palisade mesophyll tissue atypical of C, plants. The percentage of leaf air space was 16.8%, which is characteristic of CAM and CAM-cycling plants. Water stress did not induce CAM metabolism. Our results suggest that: 1) L. cotyledon exhibited the CAM cycling pathway that would better enable the plant to withstand water stress and high hght environments; 2) CAM is more widespread in the Portulacaceae than was previously thought; and 3) the evolution of the CAM pathway in the Portulacaceae appears to require the acquisition of both physiological and anatomical modifications. Crassulacean acid metabolism (CAM) is a pho- tosynthetic pathway that is characterized by net nighttime carbon dioxide (CO,) uptake (Kluge and Ting 1978; Ting 1985; Winter and Smith 1995). The CO, is stored as malic acid in the plant’s cen- tral vacuole during the night period. The malate is then decarboxylated to release CO, to be used in the PCR (photosynthetic carbon reduction) cycle in the subsequent light period. During the day, due to high CO,, CAM plants close their stomata to main- tain high internal CO, levels for photosynthesis. The closure of the stomata subsequently reduces the amount of water loss due to transpiration. Thus, these plants are usually found in environments where water availability becomes restricted tem- porarily or seasonally, such as in deserts or rock outcrops, or as in epiphytes. There are two criteria for easily distinguishing CAM plants from C, plants which exhibit daytime CO, uptake; one cri- terion is nocturnal CO, uptake, and the other is a diurnal acid fluctuation where organic acid levels are high in the morning and low in the evening. Variations of CAM have been shown in many other plants. One modification is called CAM-cy- cling (Ting 1985). In CAM-cycling, plants do not completely shut their stomata during the day and fix little or no exogenous CO, during the night. Plants still use malic acid to store CO, but it comes primarily from recycling of CO, from respiration at night (Ting 1985). CAM-cycling has been noted in many plants, for example, Talinum spp. (Martin and Zee 1983; Martin et al. 1988a), Peperomia spp. ' Author for correspondence e-mail: guralnl @ wou.edu. (Sipes and Ting 1985) and Portulacaria afra (L.) Jacq. (Guralnick et al. 1992). Another variation of CAM is called CAM-idling. CAM-idling plants close their stomata day and night in response to low water availability but still show slight diurnal acid fluctuations. It is hypoth- esized that these plants maintain low levels of met- abolic activity and can respond quickly to rainfall when the opportunity arises (Ting 1985). CAM- idling has been described in many plants, for ex- ample, Opuntia basilaris Engelm. & Bigelow (Sza- rek and Ting 1974), Xerosicycos danguyii H. Humb. (Rayder and Ting 1983), and P. afra (Gur- alnick and Ting 1987). Lewisia cotyledon is a small plant with thick fleshy leaves. Lewisia belongs to the small but widely distributed family Portulacaceae. The west- ern North America taxa of Portulacaceae (Lewisia, Claytonia, Calyptridium, and Montia) have been considered to be primarily C, plants (Nyananyo 1985). In a phylogenetic study, the western North America taxa were grouped together taxonomically as C, plants (Hershkovitz and Zimmer 1997). We have studied the photosynthetic pathways of many genera in this family, including Montia, Portula- caria, Talinopsis, Portulaca, Anacampseros, Cal- andrinia, Talinum, and Calyptridium (Guralnick and Jackson 2001). The photosynthetic diversity of the Portulacaceae is considerable despite the small size of the family. Some are considered typical C, plants (Nyananyo 1985). Other members show CAM-cycling (Martin et al. 1988a, b) while others are C, plants and also show CAM-cycling (Koch and Kennedy 1980, 1982; Kraybill and Martin 1996). Still other members of the family are suc- 132 MADRONO culent species that are facultative CAM plants and shift between C, and CAM depending on environ- mental conditions (Ting and Hanscom 1977; Gur- alnick et al. 1984; Guralnick and Ting 1988). Lewisia cotyledon is a montane species suspected of being CAM. Lewisia is considered to be drought- tolerant, occurring in rocky outcrops or on gravelly ridges that do not retain water well (Hohn 1975). This montane species undergoes significant water stress during the course of the summer due to the lack of rainfall during the months of May through September (Hohn 1975). The species grows in close association with Sedum spp. in which photo- synthetic pathway flexibility is better known. Based on these observations the investigators hypothesize that this montane species may possess metabolic or structural adaptations typical of CAM plants. This would enable Lewisia to withstand the typical sum- mer droughts of the Siskiyou Mountains. The objectives of this study were to determine the type of photosynthetic pathway utilized by Lew- isia since the photosynthetic flexibility of the genus has not been studied. This project was designed to determine what, if any, attributes of the CAM path- way Lewisia exhibited. This current study has in- dicated that attributes of the CAM-cycling pathway are more widespread in the Portulacaceae than was previously reported. MATERIALS AND METHODS Plant material. Plants of L.c. var. cotyledon were collected from the Siskiyou Mountains in Josephine Co. of southern Oregon and transplanted into 2-L pots in a glasshouse on the campus of Western Or- egon University. Additional plants were purchased from the Rare Plant Research Institute (Portland, OR). All plants were irrigated prior to the studies with 1/2-strength Hoagland’s solution. The plants were grown under natural light conditions with a light intensity of 600 pmol m~’ s~! during the sum- mer months. During the winter months, the plants were grown with a 14-hour light/10-hour dark cycle with diffuse sunlight of 200 wmol m~ s~! supple- mented with artificial light of 400 wmol m~ s"!. To study potential shifts in photosynthetic metab- olism, water was withheld from the plants for up to 10 days. Experiments were repeated four sepa- rate times with typical results of one experiment reported. Titratable acidity. Four leaves were collected in the morning and evening and were frozen (—20°C) until assayed a week later. Leaf punches of ca. 0.8 cm?’ were taken with a cork borer and weighed. The punches were ground in glass distilled water and titrated with 0.01 N-KOH to a pH 7 endpoint. CO, uptake. CO, uptake was measured using a CID-301 PS portable photosynthesis system (CID, Inc., Vancouver, WA). Four to six intact leaves were measured per timepoint over a 24-hr cycle in a typical experiment. [Vol. 48 | PEP carboxylase activity. Leaf samples for PEP- Case activity were collected in triplicate in the late | afternoon when inhibition from malate would be | the lowest. Approximately | g of tissue was col- | lected for each sample and ground in 10 ml of 100 | mM Hepes-KOH, 10 mM MegCl,, 10 mM DTT, 1% | (w/v) polyvinlypyrolidone, and 1% (v/v) Triton X- 100 adjusted to pH 7.8. An aliquot of 1 ml was taken for chlorophyll determination as described by Guralnick and Ting (1988). The sample was cen- | trifuged at 10,000 rpm for 15 min at 4°C. The crude | extract was used to assay enzyme activity spectro- | photometrically by following the oxidation of | NADH at 340 nm. The assay mixture contained | 100 mM Hepes-KOH (pH 7.8), 10 mM MgCl, 1 | mM NaHCoO,, 0.2 mM NADH, 3 mM PEP, and 200 wl crude extract in a total volume of 3 ml. Mesophyll succulence. Mesophyll succulence (S,,) was calculated from the ratio of total tissue water to total chlorophyll in fresh leaf samples (Kluge and Ting 1978). Tissue water content was determined by drying samples in an oven and sub- tracting the dry weight from the fresh weight. Chlo- rophyll was determined by grinding fresh leaf tis- sue in 5 ml of 80% acetone and then centrifuging the sample at 2500 rpm. A 3 ml aliquot was taken and measured spectrophotometrically (Guralnick et al. 1986). Leaf anatomy. Mature leaves were collected and fixed in 2% paraformaldehyde and 1.25% glutar- aldehyde in 50 mM PIPES at pH 7.2 for 24 hr. at 4°C. The tissue was sequentially dehydrated in an ethanol series, infiltrated with LR White acrylic resin, and | wm sections were placed on gelatin- coated slides. Sections were stained with Stevenel’s blue for 1 min, and then with 0.5% safranin O for 30 sec. Stereological methods were used to quantify the percentage of mesophyll airspace in five differ- ent leaves according to Guralnick et al. (1986). RESULTS CO, uptake. Irrigated plants of Lewisia were found to have primarily daytime CO, uptake with no nighttime CO, uptake (Fig. 1) though some leaves of Lewisia did show a positive nocturnal CO, uptake. When water was withheld from Lew- isia for three days, daytime CO, uptake was re- duced and there was no concomitant increase in nighttime CO, uptake (Fig. 1). Withholding water for seven days did not increase the nighttime CO, uptake (Fig. 1). Plants that were stressed for 10 days also showed reduced day and night CO, up- take, which was indicative of CAM-idling (data not shown). Nocturnal release of respiratory CO, was usually greater but not different during the night in irrigated plants compared to water-stressed plants for all treatments (Fig. 1). Titratable acidity. A slight diurnal acid fluctua- tion of 20 weq g ' FW was observed in control GURALNICK ET AL.: CAM-CYCLING IN LEWISIA 153 —u- Control --O- -H20 (3 days) -x---H20 (7 days) 1200 Time of Day ©@ ae) = gS 5 wu £ on oO S Osx Q —_ O © Fic. 1. Diurnal variation in CO, gas exchange in control and water-stressed plants for three and seven days in Lewisia cotyledon. Black bar indicates the night period and error bars represent | SE for all figures. Typical results of one experiment are shown. plants with morning acid levels significantly differ- ent than afternoon acid levels (Fig. 2, t-test, P < 0.05, n = 4). Water stress for 10 days dampened the day/nighttime differences in acidification rela- tive to the control plants (Fig. 2). Morning and eve- ning acid levels were not significantly different in 10-day water-stressed plants (Fig. 2, t-test, P > 0.05, n = 4). Plants stressed for seven days showed reduced morning and evening acid levels with sig- Titratable Acidity (eq gFW") 7 days Treatment Fic. 2. Diurnal fluctuation of total titratable acids. Sam- ples were taken at sunrise and sunset. Control and 7 day water-stressed plants showed a significantly higher a.m. acid levels than p.m. acid levels (P < 0.01, n = 4). = es nificant diurnal acid fluctuation half as large when compared to control plants (Fig. 2). PEP. carboxylase activity. PEPCase activity in control plants was 1.75 pmol mg chil”! min"! throughout the experimental period. PEPCase activ- ity in the water-stressed plants was 2.00 pmol mg chl-' min“! and did not show any increase after ten days of water stress. The activity of the water stress plants was not statistically significant from the con- trol PEPCase activity levels. Leaf anatomy and mesophyll succulence. Meso- phyll succulence was 6.33 g H,O mg chl"'! and in the range of other CAM and CAM-cycling plants (Fig. 3). Lewisia leaves did not have the typical C, appearance in which the mesophyll was well or- ganized and differentiated into palisade and spongy mesophyll tissue (Figs. 4, 5). A very prominent spongy mesophyll was observed which was typical of CAM plants. The palisade parenchyma layer showed some organization but was not organized into rows of cells. The palisade cells were elongat- ed and rounded at the ends. The palisade cells also contained a very large central vacuole (Figs. 4, 5). The percentage of airspace in the mesophyll tissue was 16.8 + 4.5% (Table 1). ae (ma | eran) || |i Prati yr. DISCUSSION In a study of the Portulacaceae, Lewisia cotyle- don was described as a C, plant (Nyananyo 1985). In addition, Hershkovitz and Zimmer (1997) in their phylogenetic analysis of the Portulacaceae, in- dicated that Lewisia fell into the C, group of North American taxa. Our results indicate that Lewisia shows aspects of CAM-cycling. Well-watered plants showed gas exchange patterns typical of C, photosynthesis: exogenous CO, uptake during the day and little, if any, nocturnal CO, uptake. Lewisia had higher titratable acid levels and a small but significant diurnal acid fluctuation not observed in typical C, plants and indicates CAM-cycling (Klu- ge and Ting 1978; Ting 1985; Guralnick et al. 1992;). The acid fluctuation was similar in magni- tude to other CAM-cycling species (Martin and Zee 1983; Harris and Martin 1991a; Guralnick et al. 1902), The physiological data did not support the hy- pothesis that L. cotyledon was a facultative CAM species. Well-watered plants showed no nocturnal CO, uptake and low PEPCase activity when com- pared to other CAM species such as Portulacaria (Guralnick and Jackson 2001). Drought did not in- duce CAM in this species even after withholding water for up to 10 days whereas 10 days of drought in Talinum triangulare produced a slight increase in nocturnal CO, uptake and an increase in PEP- Case activity (Herrera et al. 1991). Nocturnal CO, output in irrigated Lewisia was comparable to other CAM-cycling species such as Talinum (Martin and Zee 1983; Harris and Martin 1991b), Peperomia (Sipes and Ting 1985), Sedum (Martin et al. 1988a), and Portulaca (Kraybill and Martin 1996). There was also a comparable in- crease in titratable acids that we attribute to a re- cycling of respiratory CO, (Ting 1985). This wide- spread occurrence of CAM-cycling appears to have some adaptive significance. Martin and Zee (1983) and Martin et al. (1988b) have calculated that CAM-cycling species could show a significant re- duction in water loss through decreased transpira- tion because of the endogenous source of CO,. An- other consequence of CAM-cycling is that the ac- cumulation of malic acid in the vacuole may aid in osmotic water uptake (Eller and Reuss 1984, 1986). These adaptations may contribute to the survival of Lewisia in highly exposed areas such as rock out- crops, where water availability is reduced for pro- longed periods. The leaf structure of Lewisia was found to be typical in some aspects to other CAM and CAM- cycling species. Leaves of CAM plants are char- acterized by tightly packed chlorenchyma with large vacuoles where the malic acid accumulates 135 CAM-CYCLING IN LEWISIA GURALNICK ET AL. 200 pm. , bar = ion de mesophyll Mature leaf cross sect 4 igure 100 pm FE ia cotyledon leaves. f Lewis f leaf cross ions o sect Cross- FIGURES 4 and 5. spongy SM = ’ isa pal PM bar sections, guard cell. vascular bundle, GC = Figure 5. Higher magnification o mesophyll, VB 136 TABLE 1. RELATIVE PERCENT OF LEAF VOLUME OCCUPIED BY AIR SPACE IN THE MESOPHYLL TISSUE OF SELECTED SPE- cles. * The airspace data from Peperomia is the relative percent of airspace in the spongy mesophyll which is the CAM like tissue (Gibeaut, D. M. and W. W. Thomson. 1989. Stereology of the internal structures of leaves in Peperomia obtusifolia, P. camptotricha, and P. scandens. Botanical Gazette 150(2): 115-121). ° The number in pa- rentheses is the SE of the mean (n = 5). Percent of Photosynthetic | mesophyll Species mode air space Peperomia obtusifolia® C, 212 Peperomia camptotricha* CAM-cycling Lo Lewisia cotyledon CAM-cycling 16.8 (4.5)? Peperomia scandens CAM 6.4 (Gibson 1982). In addition, the mesophyll is gen- erally not well differentiated into palisade and spongy parenchyma layers (Kluge and Ting 1978). Cross sections of Lewisia leaves showed no well- organized rows of palisade parenchyma in the me- sophyll while typical C, plants have a very well organized palisade parenchyma layer. The palisade cells were elongated but the ends of the cells were rounded which is more typical of spongy parenchy- ma. The palisade cells also contained a very large central vacuole. These anatomical features were similar to the facultative CAM species, P. afra, in having a very prominent spongy mesophyll tissue and no well-organized palisade parenchyma (Gur- alnick 1987). The percentage of mesophyll airspace found in Lewisia places it in the range similar to other CAM and CAM-cycling species, such as Codonanthe crassifolia (Focke) C.V. Morton and Peperomia camptotricha Miq. which were in the range of 10— 20% mesophyll airspace (Smith and Heuer 1981; Gibeaut and Thomson 1989; Guralnick et al. 1986, Table 1). The C,; P. obtusifolia (L.) A. Dietr. had 27% by airspace volume in the spongy mesophyll tissue, which is the CAM-like tissue in Peperomia species (Nishio and Ting 1987; Gibeaut and Thom- son 1989). The reduced amount of airspace in Lew- isia compared to a typical C, species may be an important component in CAM-cycling by facilitat- ing the capture of CO, before it can diffuse out of the leaf. The mesophyll succulence results for Lew- isia placed it in the range typical of true CAM spe- cies such as P. afra and Ceraria fruticulosa Pearson & E.L. Stephens but the diurnal acid fluctuations were much smaller. This report indicates that the species L. cotyledon shows attributes of the CAM pathway, such as leaf anatomy with primarily a spongy mesophyll pres- ent, mesophyll airspace in the range typical of other CAM-cycling species, and a diurnal organic acid fluctuation. The magnitude of the diurnal acid fluc- tuation places Lewisia in the group that performs CAM-cycling. Lewisia when compared to other MADRONO [Vol. 48. genera showed fewer physiological attributes than | other CAM-cycling species. The results presented | here correlates with the phylogenetic data presented | by Nyananyo (1985) and Hershkovitz and Zimmer | (1997) which showed the genus Lewisia more | closely aligned with the genera of the western | North American taxa, Montia and Calyptridium. | Species in these genera show only slightly fleshy, succulent leaves, and low diurnal organic acid fluc- tuations indicating primarily a C, pattern of pho- | tosynthesis (Guralnick and Jackson 2001). The re-— sults presented in this paper with Lewisia would | tend to support this hypothesis that anatomical at- | tributes appear earlier with physiological attributes of CAM arising later as indicated by the low levels of PEPCase later which is supported by the phy- logenetics of CAM presented by Guralnick and Jackson (2001). In conclusion, the results of this paper indicate that L. cotyledon performs CAM-cy- cling and that CAM is not induced under water- stress conditions. ACKNOWLEDGMENTS The authors would like to thank Dr. Vince Franceschi, Nathan Tarlyn, and Simi Watson for their work on the leaf anatomy and photographs. This work was supported in part by NSF-ILI award #92-51652 and NSF-ROA award #99-41717 to Lonnie J. Guralnick. Additional support was received from a faculty development award and a Western Foundation grant to Lonnie J. Guralnick. LITERATURE CITED ELLER, B. M. AND B. R. Rugss. 1984. CO, dark fixation in CAM enhances water uptake during the night. South African Journal of Botany 3:347—-349. AND . 1986. Modulation of CAM and wa- ter balance of Senecio medley-woodii by environmen- tal factors and age of leaf. Journal of Plant Physiology 125:295—309. GIBEAUT, D. M. AND W. W. THOMSON. 1989. Stereology of the internal structures of leaves in Peperomia ob- tusifolia, P. camptotricha, and P. scandens. Botanical Gazette 150:115—121. GIBSON, A. 1982. Anatomy of succulence. Pp. 1-17 in I. P. Ting and M. Gibbs (eds.) Crassulacean Acid Me- tabolism. Proceedings of the Fifth Symposium in Bot- any. Waverly Press, Baltimore, MD. GURALNICK, L. J. 1987. The effect of drought on the sea- sonal shift from C3 to CAM photosynthesis in Por- tulacaria afra (L.) Jacq. Ph.D. dissertation. Univer- sity of California, Riverside, CA. AND M. D. JACKSON. 2001. The occurrence and phylogenetics of crassulacean acid metabolism in the Portulacaceae. International Journal of Plant Science 162(2):257-262. AND I. P. TING. 1987. Physiological changes in Portulacaria afra (L.) Jacq. during a summer drought and rewatering. Plant Physiology 70:85—91. AND . 1988. Seasonal patterns of water re- lations and enzyme activity of the facultative CAM plant Portulacaria afra (L.) Jacq. Plant, Cell, and En- vironment 11:811—818. , P. A. RORABAUGH, AND Z. HANSCOM. 1984. Sea- - 2001] sonal shifts in photosynthesis in Portulacaria afra (L.) Jacq. Plant Physiology 76:643—646. , L. PR. Ting, and E. M. Lord. 1986. Crassulacean acid metabolism in the Gesneriaceae. American Jour- nal of Botany 73:336—345. , R. L. HEATH, G. GOLDSTEIN, AND I. P. TING. 1992. Fluorescence quenching in the varied photosynthetic modes of Portulacaria afra (L.) Jacq. Plant Physiol- ogy 99:1309-1313. Harris, E S. AND C. E. MARTIN. 1991a. Correlation be- tween CAM-cycling and photosynthetic gas exchange in five species of Talinum (Portulacaceae). Plant Physiology 96:1118—1124. AND . 1991b. Plasticity in the degree of CAM-cycling and its relationship to drought stress in five species of Talinum (Portulacaceae). Oecologia 86:575—584. HERRERA, A., J. DELGADO, AND I. PARAGUATEY. 1991. Oc- currence of inducible crassulacean acid metabolism in leaves of Talinum triangulare (Portulacaceae). Journal of Experimental Botany 42:493—499, HERSHOKIVITZ, M. A. AND E. A. ZIMMER. 1997. On the evolutionary origins of the cacti. Taxon 46:217—232. Houn, J. E. 1975. Biosystematic studies of the genus Lew- isia, Section Cotyledon (Portulacaceae). Ph.D. disser- tation. University of Washington, Seattle, WA. KLuGE, M. AND I. P. TING. 1978. Crassulacean acid me- tabolism. Springer-Verlag, Berlin. Kocu, K. E. AND R. A. KENNEDY. 1980. Characteristics of crassulacean acid metabolism in the succulent C, di- cot, Portulaca oleracea L. Plant Physiology 65:193— Owe AND . 1982. Crassulacean acid metabolism in the succulent C, dicot, Portulaca oleracea L. under natural environmental conditions. Plant Physiology 69:757-761. | KRAYBILL, A. A. AND C. E. MARTIN. 1996. Crassulacean acid metabolism in three of the C, genus Portulaca. International Journal of Plant Science 157:103—109. AND A. K. ZEE. 1983. C, Photosynthesis and cras- sulacean acid metabolism in a Kansas rock outcrop GURALNICK ET AL.: CAM-CYCLING IN LEWISIA [37 succulent, Zalinum calycinum Engelm. Plant Physi- ology 73:718-—723. Martin, C. E., M. HIGLEY, AND W-Z WANG. 1988a. Re- cycling of CO, via crassulacean acid metabolism in the rock outcrop succulent Sedum pulchellum Michx. (Crassulaceae). Photosynthesis Research 18:337—343. ’ . 1988b. Ecophysiological signit- icance of CO,-recycling via crassulacean acid metab- olism in Talinum calycinum Engelm. (Portulacaceae). Plant Physiology 86:562—568. NisHio, J. N. AND I. P. TING. 1987. Carbon flow and met- abolic specialization in the tissue layers of the cras- sulacean acid metabolism plant, Peperomia campto- tricha. Plant Physiology 84:600—604. NYANANYO, B. L. 1985. Tribal and generic relationship and classification of the Portulacaceae (Centrosper- mae). Ph.D. dissertation. University of Reading, En- gland. RAYDER, L. AND I. P. TING. 1983. Shifts in the carbon metabolism of Xerosicyos danguyi H. Humb. (Cucur- bitaceae) brought about by water stress. Plant Physi- ology 72:606—610. SIPEs, D. L. AND I. P. TING. 1985. Crassulacean acid me- tabolism and crassulacean acid metabolism modifi- cations in Peperomia camptotricha. Plant Physiology 77:59-63. SMITH, J. A. C. AND S. HEUER. 1981. Determination of the volume of intercellular spaces in leaves and some val- ues for CAM plants. Annals of Botany 48:915—917. SZAREK, S. R. AND I. P. TING. 1974. Seasonal patterns of acid metabolism and gas exchange in Opuntia basi- laris. Plant Physiology 52:76-81. TING, I. P. 1985. Crassulacean acid metabolism. Annual Review of Plant Physiology 36:595—622. AND Z. HANSCOM. 1977. Induction of acid metab- olism in Portulacaria afra. Plant Physiology 59:51 1— 514. WINTER, K. AND J. A. C. SMITH. 1995. Introduction to crassulacean acid metabolism. biochemical principles and ecological diversity. Pp. 1-13 in K. Winter and J. A. C. Smith (eds.), Crassulacean Acid Metabolism: biochemistry, ecophysiology and evolution. Springer, Berlin. MADRONO, Vol. 48, No. 3, pp. 138-151, 2001 THE SIGNIFICANCE OF POPULATION SUCCESSIONAL STATUS TO THE EVOLUTION OF SEEDLING MORPHOLOGY IN PINUS CONTORTA VAR. LATIFOLIA (PINACEAE) TIMOTHY J. BRADY! Department of Biology, Harney Science Center 342, University of San Francisco, 2130 Fulton Street, San Francisco, CA 94117-1080 ABSTRACT The objective of this research was to test the hypothesis that the successional role of a plant population, because of its implications for the nature of the selective regime experienced by regeneration cohorts, determines, in part, the course of autecological evolution in a lineage of populations. A provenance study, which involved the raising of seedlings of lodgepole pine (Pinus contorta Loudon var. latifolia) from seed under uniform conditions in a greenhouse, provided a test of this hypothesis. The seeds came from seral, climax, and persistent lodgepole pine populations indigenous to the Blue Mountains region of northeastern Oregon and southeastern Washington. Data about shoot and root system features, collected at the end of the first season of growth, proved useful in evaluating competitive competence, the relative ability of a plant, or group thereof, to compete successfully for essential resources such as light, water, and mineral nutrients. Analyses of variance and discriminant function analysis facilitated the search for correlations between population successional status and seedling morphology. The total leaf area, pho- tosynthetic potential, degree of subdivision of the root system, and total root length of a typical seedling derived from a seral population are smaller than such quantities for the other population types. In climax populations, seedlings tend to exhibit the largest total leaf areas and total photosynthetic potentials among lodgepole pine seedlings. They are equipped with more elaborate and larger root systems. Despite their exceptional heights, seedlings belonging to persistent populations not only have slightly lower total pho- tosynthetic potentials than those from climax populations, they also possess reduced lateral root densities and total root lengths. Greater competitive competences, with reference to life in the subcanopy, char- acterize seedlings from climax lodgepole pine populations compared to seedlings from seral or persistent populations. The results of this study support the hypothesis that successional status influences the evo- lution of autecological attributes in a population lineage. Forest succession is the undisputed exemplar of vegetation change through time. The formation of a canopy as the crowns of adjacent trees expand in size represents the most consequential structural change occurring during forest succession. In co- niferous forests, canopy closure may not occur for Over a century after stand establishment (Peet 1981). The appearance of a canopy results in the elimination of an open site and the creation of a subcanopy. Gradual alterations in stand structure and abiotic environmental conditions occur as an open site gives way to a canopied stand. An open site, though it is an area characterized by physical extremes, is a place where a plant, due to the absence or paucity of other plants with sim- ilar requirements, participates in little, if any, com- petition for essential resources such as light, water, and mineral nutrients. In contrast, the canopy in- sulates tree seedlings and other inhabitants of the subcanopy from the severe abiotic environmental conditions that prevail on an open site. Neverthe- less, because resources are highly accessible to, and vigorously exploited by, the fully developed trees composing the canopy, at least one resource ordi- narily limits the survival of subcanopy plants. Con- ' E-mail: brady @usfca.edu sequently, a plant almost invariably must partici- pate in intense intra- and interspecific competition when a resident of the subcanopy. Any new aut- ecological trait appearing through mutation or gene flow that improves the “‘competitive competence”’ of a subcanopy plant will increase its fitness, 1.e., the likelihood that it will become a part of the can- opy and contribute to the genetic constitutions of future generations. Competitive competence is a relative expression of the ability of a plant, or group thereof, given the features that characterize it (including many aspects of genetics, physiology, anatomy, morphology, and breeding behavior), to compete successfully for those resources that limit maintenance and growth activities. It is an heuristic tool that provides a means of comparing plants according to their abil- ities to survive, grow, and reproduce under a com- petitive regime. Competitive competence recalls the general version of the concept of tolerance em- braced by most silviculturists during the first half of the twentieth century (e.g., Biihler 1918; Baker 1937, 1950; Toumey and Korstian 1937). Compet- itive competence permits the comparison of plants growing in naturally complex settings, not under simplified garden, greenhouse, or laboratory con- ditions; it avoids the artificiality of the tendency to rank individuals, populations, or species by single- factor tolerances. | To ensure the continued existence of its popula- ‘tion lineage (a temporal sequence of conspecific populations related as ancestors and descendants), the members of a population must beget juveniles that are able to compete successfully for resources within the subcanopy, or they must produce prop- -agules that disperse and give rise to individuals ca- pable of surviving, growing, and reproducing in a different location (seed dormancy and cone serotiny may permit the continuity of a population lineage in the absence of emigration). The autecology of a population represents the product of evolutionary history, acting as a phyletic constraint that limits the types of successional role that a population can assume in a given environment. In view of its con- sequences for regeneration dynamics, successional status certainly contributes significantly to the se- lective milieu experienced by the members of a population. Consequently, succession potentially influences the course of autecological evolution within a population lineage. Populations of lodge- pole pine indigenous to the Blue Mountains region of the Pacific Northwest, which occupy an immense habitat island, represent consummate candidates for studying the relationship between successional sta- tus and the evolution of competitive competence. The successional roles of Blue Mountain lodge- pole pine populations. The range of Pinus contorta var. latifolia (hereafter referred to by its vernacular name “lodgepole pine’’) encompasses the Rocky Mountains, the Washington Cascades, the area be- tween the Rockies and Coast Range in Canada, and the Blue Mountains. Due to its rapid early growth rate and its tolerance of exposed conditions and poorly developed substrates, lodgepole pine estab- lishes a foothold on many open sites throughout its range (Pfister and Daubenmire 1973; Volland 1985). Hence, lodgepole pine often is a colonizer. In most cases, because of a dearth of nearby seed sources, it becomes only a minor constituent of the stands on these sites. Lodgepole pine adopts a seral role in such situations: It does not regenerate suc- cessfully beneath the canopy; species having great- er shade tolerances, viz., grand fir (Abies grandis (Douglas) Lindley) or subalpine fir (A. lasiocarpa (Hook.) Nutt.), replace it in 5O—200 years. On some sites, lodgepole pine populations achieve domi- nance and assume unique successional positions. Pfister and Daubenmire (1973) identified three gen- eral successional roles for populations of lodgepole pine in stands that it dominates: (dominant) seral, climax, and persistent. These three population types do not represent different temporal components of a common sere. Rather, each is a particular element of a unique sere. E C. Hall (USDA Forest Service, unpublished) classified lodgepole pine populations according to successional role in communities that they dominate in the Blue Mountains. Franklin and BRADY: SEEDLING EVOLUTION IN LODGEPOLE PINE 139 @ Spokane COLUMBIA a BASIN MOUNTAINS PROVINCE cane © HIGH LAVAN ~ PLAINS 14.5 cm) and be- ! come susceptible to attack by mountain pine beetles | (Dendroctonus ponderosae). The deaths of these | trees produce a high fuel load within the stand, | which inevitably leads to a hot fire 5-10 years later. - The fire consumes all live and dead woody material on the site, including canopy members and poten- | tial regeneration in the subcanopy. In 10—20 years, | the disturbed site becomes suitable for the estab- | lishment of lodgepole pine seedlings. Following | dispersal from some external seed source, the se-— quence begins anew. Succession proceeds until a_ beetle kill and fire again destroy the stand. As a_ persistent population cannot replace itself in situ, the appearance of lodgepole pine after each fire de- pends upon seed dispersal from another site. The temporal continuity of a persistent population lin- eage depends upon the success with which the seedlings of its constituent populations can colonize open sites. Hall’s view of the successional role of a persistent Blue Mountain lodgepole pine popu- lation is controversial. The details of his proposed successional scenario will likely change once stand development studies are complete. I assume, pro- visionally, that seed vagility and seedling tolerance of harsh abiotic environmental conditions, not seed- ling competitive competence, contribute most sig- nificantly to the fitness of a persistent lodgepole pine tree. MATERIALS AND METHODS Cone collection. A lodgepole pine population represented a dominant canopy element in 61 stands inspected by F C. Hall (USFS unpublished data) as part of a plant association analysis of the Blue Mountains region. Twenty of these popula- tions proved appropriate for inclusion in this study: In August 1994, each had at least one dominant, codominant, or intermediate lodgepole pine tree that was cone-bearing, accessible, and apparently free of insect damage (Table 1). Using the tripartite scheme introduced above, Hall identified six of these populations as seral (species with superior competitive competences present with height growth rates that exceed those of lodgepole pine trees), seven as climax (topographic depressions; species with superior competitive competences ab- sent), and seven as persistent (species with superior competitive competences present, but lodgepole pine trees exhibit greater height growth rates). Within each successional type, the populations are widely distributed throughout the Blue Mountains with respect to latitude, longitude, and elevation. To reduce the chances that gene flow, not common site-specific selective pressures, might produce a correlation between successional status and com- petitive competence, I admitted no contiguous pop- ulations possessing identical successional roles into this study. 2001] TABLE |. BLUE MOUNTAIN LODGEPOLE PINE POPULATIONS SAMPLED. The number of participating families indicates ‘those submitted to analyses of variance and discriminant function analysis. Several assistants and I collected cones from one or more lodgepole pine trees belonging to each population during August 1994. Since minimum distances of 40-50 m separated the sampled trees, each most likely resides within a distinct genetic neighborhood. We recovered a minimum of 50 cones from each tree. We picked the cones off branches removed from each crown by gunshot or through the use of loppers. To maximize the pro- portion within each family of progeny derived from crosses between members of the same _ genetic neighborhood, we generally collected cone-bearing branches from the sides of crowns (the infiltration of pollen grains from afar to branches below the tops of crowns probably is limited, especially for trees belonging to seral and climax lodgepole pine populations where crowns frequently are contigu- ous or overlapping). We used positional criteria so as to collect only those cones that matured in re- sponse to pollination events occurring in the spring of 1992. Although most lodgepole pine trees in the Blue Mountains possess nonserotinous cones (Lo- tan and Critchfield 1990), we were careful to avoid older closed (serotinous) cones. At the time of col- lection, most cones were light brown in color, which is indicative of ripeness (Krugman and Jen- kinson 1974). Many were beginning to open. We stored the cones in paper bags, loosely packed in cardboard boxes, in the open bed of a truck for the duration of the fieldwork (up to four weeks prior to seed extraction). BRADY: SEEDLING EVOLUTION IN LODGEPOLE PINE Participating Latitude Longitude Elevation Population families (N) (W) (m) | Seral populations Mount Pisgah 2 44°28’ 12014" 2,100 Little Kelsay Creek | 44°54’ 118°45’ 1,900 North Fork Wolf Creek 2 45°08’ PS O87 1,900 Bingham Spring | 44°30’ 12030 1,900 Thompson Spring 2 44°28’ 120) 1S" 2,200 Little Phillips Creek 5 45°42! 118°03' 1,700 Climax populations Bingham Prairie 2 44°31’ 1Z0n32) 1,900 Jackson Creek 2 44°27' 119°58' 1,800 Crowsfoot Creek Edge | 43°54’ 119°30' 1,800 Summit Prairie Edge 2. 44°] 1’ 118°30' 1,800 Wickiup Creek 2 44°11’ 119°14’ 1,800 Ditch Creek Edge | 45°07' 119°2 1 1,600 Myrtle Creek 2 43°59! 119°05’ 1,900 Persistent populations Stove Spring 2 44°31" 12038) 1,800 Summit Prairie Slope 3 44°11’ 118°30' 1,900 Camp Creek Z 44°03’ 119°07’ 1,800 Tribble Creek | 45°10’ IO O2? 1,800 Indian Springs Butte I 44°15’ 118°42’ 2,100 Dixie Butte I 44°33’ 118°37' 2,000 Winom Creek I 45°O1' 118°38' 1,700 Seed extraction. Cone processing took place in September 1994 at the Wind River Nursery Seed Extractory operated by the USFS near Carson, Washington. After transferring the cones from the paper bags to loose-weave nylon sacks, we im- mersed them in hot water (about 60°C) for 4—5 minutes. We dried the cones in a kiln dryer at 35°C for 24 hours. This treatment effectively opened nearly all of the cones. We shook the seeds out of the open cones by use of a manually-driven tum- bler. A Clipper cleaner permitted us to partially de- wing the seeds and remove debris from each seed lot. We used an x-ray machine and Polaroid film to generate images of the contents of 19—200 seeds per cleaned sample. Seedling production. After soaking 39 seed lots (each representing a single family) in water for 72 hours, I stratified them without media in polyeth- ylene bags at 2°C for 33 days (Bonner et al. 1974; Krugman and Jenkinson 1974; Owens and Molder 1984; J. McGrath, USDA Forest Service, personal communication). We sowed the seeds immediately after completion of the stratification process in con- tainers in a fiberglass, unheated greenhouse at the University of California, Berkeley. We sowed 2—5 seeds in a 1:1 mixture of sphagnum peat moss and vermiculite in each of 50 Ray Leach Pine Cell Cone-tainers per family. The sowing rate, calculat- ed by reference to the x-ray image of a subsample of seeds produced at the time of seed extraction, 142 MADRONO varied in direct proportion to the percentage of filled seeds. We subsequently placed each container in arandomly chosen slot in a cluster of rectangular trays in the center of the greenhouse. We raised over 1,000 lodgepole pine seedlings through a sin- gle three-stage season (151 days) under uniform conditions in the greenhouse. To encourage seed germination (1—24 days after sowing), we main- tained high air temperatures and relative humidities, suspended a 50% shadecloth near the greenhouse roof to reduce photon flux densities, and kept the growing medium in each container constantly wet. As necessary, we thinned each container to a single seedling. During the free growth phase (25-115 days after sowing), we operated an evaporative cooling system for 12 hours per day, illuminated the seedlings with fluorescent lights for 18 hours per day, and injected a Plantex 20:20:20 macro- and micronutrient solution (100 ppm of nitrogen) into the irrigation system during every other watering (about once each week). We created drought stress conditions and terminated the use of daylength ex- tension lights at the beginning of the budset stage (116-151 days after sowing). Throughout this final period of seedling growth, we used an exhaust fan for 24 hours per day to encourage lower tempera- tures and relative humidities, and we halved the nitrogen content of each fertilizer application (50 ppm of nitrogen). Collection of data on seedling morphology. The- oretical predictions and empirical evidence suggest that, regardless of sibling relatedness and heritabil- ity, an acceptable approximation of a family mean for a given quantitative morphological feature is obtainable by sampling 10—20 progeny (Brady un- published). Consequently, within three weeks of termination of the growth period (1.e., the end of the bud-set stage), we harvested ten randomly-cho- sen seedlings per family on which terminal resting buds had developed and acquired information for 12 distinct morphological attributes. A description of each trait appears in Table 2. Statistical procedures. Following the recommen- dations of Gould and Johnston (1972) for identi- fying patterns of geographical variation within and among species, I adopted different approaches, viz., analyses of variance and discriminant function analysis, in my attempt to discover a correlation between lodgepole pine seedling morphology and population successional status. The analysis of vari- ance (ANOVA) furnishes a way to evaluate the ob- served differences among three or more (statistical) population means (Winer 1971; Tabachnick and Fi- dell 1983; Lindman 1992; Bogartz 1994). In the context of the present investigation, successional status functions as the sole independent variable with three levels, or groups (seral, climax, and per- sistent). The morphological features act as depen- dent variables. Discriminant function analysis in- cludes an array of multivariate techniques that TABLE 2. DESCRIPTIONS OF 12 ATTRIBUTES PERTAINING TO | SHOOT AND ROOT SYSTEM MORPHOLOGY ASSESSED FOR | LODGEPOLE PINE SEEDLINGS ON WHICH TERMINAL RESTING | BUDS HAD DEVELOPED WITHIN 151 DAYS OF GROWTH IN A | GREENHOUSE. | SHL (shoot length) The distance between the cotyledonary node and the base | of the terminal resting bud on the main stem as measured with a ruler to the nearest millimeter. | | STC (stem caliper) The maximum diameter of the main stem at its midlength | (SHL/2), as measured with a vernier caliper to the nearest — 0.025 mm. NPL (number of needle-like primary leaves) The number of needle-like primary leaves attached to the main stem, disregarding cotyledons. | NSS (number of axillary short-shoots) The number of axillary short-shoots attached to the main stem. NLS (number of axillary long-shoots) The number of axillary long-shoots attached to the main stem. BLL (blade length) Pertaining to the intact and fully developed primary leaf closest to the midlength of the main stem (SHL/2), the distance between the base and tip of the leaf blade along its midrib, as measured with a ruler to the nearest milli- meter. TPC (taproot caliper) The maximum diameter of the taproot at its midlength (actually, half the distance between the cotyledonary node and the tip of the taproot), as measured with a vernier caliper to the nearest 0.025 mm. NLR (number of lateral roots) The number of lateral (secondary) roots attached to the proximal half of the taproot (the region between the cot- yledonary node and the midlength of the taproot). LRL (lateral root length) The length of the lateral root attached closest to the mid- length of the taproot, as measured with a ruler to the near- est millimeter. NTR (number of tertiary roots) The number of tertiary (absorbing) roots connected to the lateral root that is attached closest to the midlength of the taproot (we considered a tertiary root and all of its branch- es, if present, as a single unit). SHB (shoot biomass) The weight of the air-dried shoot system (epicotyl), as obtained with an electronic balance to the nearest 0.01 g. ROB (root biomass) The weight of the air-dried root system (hypocotyl), as obtained with an electronic balance to the nearest 0.01 g. make use of a set of independent variables to dis- cover the dimensions along which the differences among groups are greatest, to test the statistical sig- nificance of those differences, to predict group '2001] membership, and to interpret the (biological) mean- ing of each dimension (Kendall and Stuart 1966; Lachenbruch 1975; Gnanadesikan 1977; Karson 1982; Tabachnick and Fidell 1983; Reyment et al. 1984; Morrison 1990). In this study, successional status is the dependent variable with three groups (seral, climax, and persistent). Morphological attri- butes act as independent variables. I performed all quantitative analyses for this study on a MacIntosh Quadra 950 using JMP 3.1 application software or programs written by myself and executed with the Microsoft QuickBASIC 1.00B Interpreter. RESULTS AND DISCUSSION I calculated the means and standard deviations of family means by successional class (seral, climax, and persistent) on each of 12 morphological traits for a total of 390 lodgepole pine seedlings (39 fam- ilies). Examination of these statistics as well as his- tograms depicting the frequency distributions of families for these features (not shown) indicated that five families (one seral, two climax, and two persistent) represent outliers. I made no further use of the data for these five families. Table 3 gives summary Statistics on each of the 12 morphological features by successional class for the remaining 34 families. The mean of family means on three of the attributes each exhibits very little variation among successional classes: NLS (number of axillary long- shoots), TPC (taproot caliper), and SHB (shoot bio- mass). Consequently, I withdrew them from further statistical consideration. Analyses of variance. The validity of the results of a series of ANOVAs depends upon the assump- tion that the scores on a particular dependent vari- able within each group are approximately normally distributed, and upon the assumption that each group possesses a common variance on a given de- pendent variable. I evaluated the normality assump- tion separately for the nine morphological traits within each of the three successional classes in two ways: I examined a histogram showing the distri- BRADY: SEEDLING EVOLUTION IN LODGEPOLE PINE 143 TABLE 3. THE MEAN AND STANDARD DEVIATION OF THE FAMILY MEAN OF EACH OF 12 MORPHOLOGICAL FEATURES OF ' LODGEPOLE PINE SEEDLINGS BY SUCCESSIONAL CLASS FOR 34 FAMILIES. Units of measurement appear in parentheses. All ‘numbers were rounded to two digits to the right of the decimal point for display purposes. SHL STC BLL TPC ERE SHB ROB (mm) (mm) NPL NSS NLS (mm) (mm) NLR (mm) NTR (g) (g) seral (total number of families = 11) mean = 58.36 1.87 106.86 4.66 20? 35.56 0.66 ZIL20 92.38 26.49 35.56 28.05 'sd = 8.17 0) 10.41 4.10 0.64 240) 0.11 2.62 Doo 3.79 7.40 4.69 climax (total number of families = 12) mean = 61.00 1.95 117.54 1.48 2.26 36.68 0.65 29.66 98.23 29.14 B73 33°33 sd = 8.00 0.09 16.35 91 0.23 1.32 OA 2.44 7.74 4.01 Sy ay AAG persistent (total number of families = 11) mean = 72.26 1.87 iS.23 2.60 2.26 37.14 0.69 26.68 95.42 27.69 38.75 32.01 sd = 7.58 0.12 11.86 3.25 0.23 1.87 OOF 3.08 6.90 By I/9) Sal 3.87 bution of family means, which facilitated the visual detection of skewness, and I tested the null hypoth- esis that this distribution is normal using the Sha- piro-Wilk W-statistic. An assumption of normality was supported for eight of the attributes in every class (a = 0.05). The distribution of one seedling feature, NSS (number of axillary short-shoots), ex- hibited severe positive skewness in the climax and persistent groups. Therefore, I did not probe NSS further using ANOVA. I performed four different statistical tests (O’ Brien’s, Brown-Forsythe, Levene F, and Bartlett’s tests) to check the homogeneity of variance among the three successional classes for each of eight morphological characteristics. None of the tests detected statistically significant (a = 0.05) differences in variance among groups for any of these traits. Eight ANOVAs revealed that the differences among group means are Statistically significant (a = 0.05) for three morphological attributes: SHL (shoot length), NLR (number of lateral roots), and ROB (root biomass). In each case, t-tests identified the particular group differences responsible. The re- sults of the significant tests appear in Table 4. Year-old seedling shoots are, on average, longer in families derived from persistent populations (mean SHL = 72.26 mm) than in families belong- ing to either seral or climax populations (58.36 mm and 61.00 mm, respectively). However, no real dif- ference in seedling shoot length exists between ser- al and climax groups. The t-tests detected signifi- cant differences in the number of lateral roots be- tween seral and climax and between climax and persistent groups. In fact, the largest mean number of lateral roots per seedling (mean NLR = 29.66) characterize climax families. However, no evidence exists of a genuine difference in the number of lat- eral roots between seral and persistent families (27.20 and 26.68, respectively). The biomasses of seedling root systems are significantly smaller in seral families (mean ROB = 28.05 g), but the cli- max and persistent groups are not statistically dis- 144 MADRONO TABLE 4. RESULTS OF SIGNIFICANT (a = 0.05) ONE-WAY ANALYSES OF VARIANCE AND T-TESTS OF THE DIFFERENCES| AMONG SUCCESSIONAL CLASSES IN MEAN VALUES ON EIGHT LODGEPOLE PINE SEEDLING TRAITS. df = degrees of freedom, ! SS = sum of squares, MS = mean square. I rounded all numbers to two digits to the right of the decimal point for| display purposes. SHL (shoot length) Source df successional status 2 error Sih ‘seral:persistent = 4.14 df = 20 P= 0.00 ‘climax:persistent = —3.46 df = 21 P = 0.00 NLR (number of lateral roots) Source df successional status 2 error 31 tseral:climax = 2.33 di — 2 BP =.0:03 ‘climax:persistent = 2.58 df=21 P< 0.02 SS MS F P 1208.22 604.11 962 0.00 1946.33 62.79 [Vol. 48 { | SS MS F P 58.82 29.41 3.98 0.03 228.85 7.38 ROB (root biomass) Source df successional status 2 error 31 tseral:climax = 3.43 df= 21 P= 0.00 ‘seral:persistent = 2.16 df = 20 P< 0.04 tinguishable on the basis of this trait (33.33 g and 32.01 g, respectively). A series of ANOVAs yields the maximum amount of information about the importance of each attribute to the determination of group affili- ation only if none of those traits covary. A matrix of Pearson product-moment correlation coefficients (Table 5) indicates that, in fact, every pair of mor- phological attributes is, to some degree, correlated. Thus, the claim that differences among successional classes on the mean values of SHL, NLR, and ROB are statistically significant incorrectly implies that successional status affects three independent phe- nomena. A closer examination of the relationship between seedling morphology and successional sta- tus demands a multivariate perspective. Discriminant function analysis. Justification of a multivariate normality assumption (within each group, the sampling distribution of the mean on each independent variable and all linear combina- TABLE 5. SS MS F P N71.28 85.64 6.08 0.01 436.90 14.09 tions of them exhibit normality) is a prerequisite for the use of discriminant function analysis. With small, unequal sample sizes (as in this study), val- idation of the multivariate normality assumption is largely a matter of judgment. By discarding the at- tribute NSS (number of axillary short-shoots), shown previously to possess a highly skewed dis- tribution within each successional class, the validity of the assumption of multivariate normality is like- ly. A discriminant function analysis of 34 lodgepole pine family means on each of eight morphological characteristics, which accounts for about 68.2% of the total variation in seedling morphology, created two discriminant functions (Table 6). The first func- tion effectively ordinates all three successional classes by partitioning approximately 66.5% of the variation among successional classes to achieve group separation (Fig. 2). The second discriminant function appropriates about 33.5% of the variation CORRELATION MATRIX OBTAINED BY COMPUTING PAIRWISE PEARSON PRODUCT-MOMENT CORRELATION COEFFI- CIENTS ACROSS ALL THREE SUCCESSIONAL CLASSES FOR EIGHT MORPHOLOGICAL FEATURES. SHL SiC NPL BLL NLR LRL NTR ROB SHL 1.0000 SEC —0.0049 1.0000 NPL ORS 9 5/ O:11093 1.0000 BLL 0.2631 0.0554 0.2300 1.0000 NLR —0.3149 0.4287 0.0759 20: 0122 1.0000 LRL 0.3808 O2128 0.2941 0.4289 0.0791 1.0000 NTR 0.1109 —0.0596 0.1189 0.4347 0.3571 0.6601 1.0000 ROB 0.3650 OZ161 O3255 0.4426 0.2874 0.4326 0.4246 1.0000 EIGHT FEATURES PERTAINING TO SEEDLING MORPHOLOGY. BRADY: SEEDLING EVOLUTION IN LODGEPOLE PINE 145 _ TABLE 6. DISCRIMINANT FUNCTION ANALYSIS OF A DATA SET CONSISTING OF 34 LODGEPOLE PINE FAMILY MEANS ON standardized discriminant function coefficients Discriminant function SHL STC NPL BLL NLR LRL NTR ROB | +0138 0.1790 0.2367 = 27 SO 0.0476 0.5094 aU Oi59 0.2698 2 OSnly 0.1089 0.2082 0.1081 0.2483 0.0600 i327 0.5586 Comparison of predicted and actual group membership Predicted Actual group membership group membership Seral Climax Persistent Seral 8 l 0 Climax l 1] 2, Persistent 2 O 9 Teeny Loading matrix function SHEL STC NPL BLL NLR ERE NTR ROB l —O7501 0.3860 0.0434 = 10) AIAG} Onss7 0.1818 0.1757 0.0939 2 0.5628 0.3780 0.5765 0.4998 0.3649 0.5492 0.4326 0.8950 among successional classes to distinguish seral from the other groups (Fig. 2). A jackknife technique failed to expose any un- usually large Mahalanobis’ distances (the distances in multivariate space from family means to their group centroids); and plots of standardized family scores for the seral, climax, and persistent groups revealed roughly equal dispersions. These findings verify the absence of multivariate outliers and sup- 2.5 climax - 15 Ao) e = 20S ye ic © £ ; = age persistent Oo 2 O 45 -2.5 -2.5 “1.5 -0.5 0.5 1.5 2.5 Discriminant function 2 Fic. 2. The location of each of three lodgepole pine suc- cessional classes in multivariate space as defined by a dis- criminant function analysis of eight attributes pertaining to seedling morphology. Each axis consists of standard- ized scores on the indicated discriminant function. A dot marks each group centroid (mean standardized discrimi- nant function score). The area enclosed by a circle cor- responds to the 95% confidence region around a given centroid. port the assumption of homogeneous variance-co- variance matrices, which validate the use of dis- criminant function analysis in this study. Two re- sults confirm the statistical significance of this mul- tivariate inquiry: Firstly, an approximate F-ratio justified rejection of the null hypothesis that the centroids associated with the three successional classes are equal. Secondly, in a comparison of ac- tual group memberships and those based on pos- terior probabilities obtained from Mahalanobis’ dis- tances, 82% of the 34 family predictions proved correct (Table 6). The large value (0.73) of the as- sociated Kappa statistic, which measures the agree- ment between predicted and actual group affinity, connotes that the results of the present discriminant function analysis are, indeed, very reliable. I referred to the loading matrix (Table 6) to ex- plain the differences among seral, climax, and per- sistent population types on each of the two discrim- inant functions. Following statistical convention, | deemed only those loadings of at least 0.50, which implies an overlap in variance of about 25% be- tween an independent variable and a discriminant function, as eligible for interpretation. A seedling’s growth polarity. The first, and most information-laden, discriminant function, or dimen- sion of variation in seedling morphology, reflects a change in the pattern of asset allocation, not the overall amount of growth. In morphometric terms, most of the variation among successional classes, as explained by the first dimension, is attributable to the alteration of seedling “‘shape’’, not “‘size’’. This finding differs markedly from the results of most multivariate morphometric studies, which hopelessly confound the genetic determination of form and phenotypic plasticity, and where the first axis of variation corresponds to a generalized size dimension (Reyment et al. 1984). All of the inde- 146 MADRONO Stem surface area (mm2) 0 10 #20 30 40 #50 60 70 80 490 4100 Stem height (mm) Fic. 3. Surface area of a cylindrical stem as a function of stem caliper and height. The three points indicate the average stem surface areas of lodgepole pine seedlings derived from seral, climax, and persistent populations. pendent variables that contribute substantially to a size dimension must exhibit discriminant function coefficients, correlations among themselves, and loadings of like sign (Jolicoeur and Mosimann 1960; Reyment et al. 1984). In the present case, the two morphological attributes having the largest ab- solute loadings on the first discriminant function (SHL and NLR) possess discriminant function co- efficients, correlations, and loadings of opposite sign. The first dimension of variation clearly por- trays an anisotropic pattern of change in seedling growth among seral, climax, and persistent lodge- pole pine populations. The trait SHL (shoot length) represents a measure of the total amount of seedling height growth dur- ing the first season of development, or a height growth rate. Internode elongation, the primary component of height growth, may lead to the pro- duction of new leaves in regions of higher photon flux densities, it may move existing leaves into sun- light, and it tends to reduce self-shading by increas- ing the distances among appendages attached to the stem (e.9:, Hom 1971; Halle et al. 1975; Fisher 1986; Givnish 1986, 1988, 1995; Sakai 1986; Ko- hyama 1987; Kiippers 1989). Since their epicotylar stems remained green during the first season of de- velopment, height growth via internode elongation may further contribute to the carbon economies of lodgepole pine seedlings by promoting stem pho- tosynthesis (see Nilsen 1995). The total green stem surface area provides a means of comparing the stem photosynthetic potentials of different groups of seedlings. As the epicotylar stems observed in this study lacked taper, I calculated the surface area of a stem by assuming that it is cylindrical in shape. The surface area of a stem increases with both height and caliper (Fig. 3). Because long-shoots were relatively small and nearly constant in number across all lodgepole pine seedlings (NLS in Table 3), the exclusive application of a simple cylindrical model to the main stem furnishes an adequate com- [Vol. 48. parative measure of the total stem photosynthetic | potential. The three points in Fig. 3 mark the av- | erage main stem surface areas for the three lodge- | pole pine successional classes (based on mean STC | and mean SHL from Table 3). Due to their excep- | tional heights (mean SHL = 72.26 mm), seedlings | from persistent families possess the greatest stem | photosynthetic potentials. Seedlings from seral and climax populations have comparable stem photo- | synthetic potentials (mean SHL = 58.36 mm and | 61.00 mm for seral and climax groups, respective- | ly). NLR (number of lateral roots) is the second mor- | phological attribute making a large contribution to — the differentiation of successional classes along the — initial dimension of variation. As it provides an es- | timate of the degree of branching, or subdivision of | the root system, NLR relates the thoroughness with | which a seedling can extract water and mineral nu- | trients from a given volume of soil (Fitter 1985, 1991, 1994; Caldwell and Richards 1986). While | only its tip functions in uptake, a lateral root pro- vides a “‘platform”’ for numerous absorbing tertiary roots. Because the most efficient zone of absorption — occurs near the tip of any root, the total number of root tips in a specified volume of soil acts as a measure of the absorptive capacity of a root system (Kramer and Kozlowski 1960; Caldwell and Rich- ards 1986). Imagine a lodgepole pine seedling root system that is completely embedded in a cylindrical mass of soil. The height and vertical centerline of the cylinder correspond to the length and position of the taproot, and its radius equals the length of a lateral root. All lateral roots are identical in length, and each possesses the same number of attached tertiary roots. The following equation gives the to- tal number of root tips characteristic of the root system per mm? of soil: Cea aU l d . Tw (linea) 7 ee taproot Preeiiaenns en tertiary roots P+! (1) where P is the number of taproot tips (usually, P = 1)3 Laproor 18 taproot length; diaic,ais 1S a density, the number of lateral roots produced per unit length of taproot; Jjien; 1S the length of a lateral root; and Qrertiary roots 1S the density of tertiary roots along a lateral root. The product J,,,,,0,d gives the total number of lateral root tips. The product lij,,coQiater aple a. yields the total number of tertiary root tips. The denominator is the volume of the reference cylinder of soil. Based on estimates of P (1), taproor (160 mmm) and 4,327, room (O29 per mule meter of lateral root length) from this study (all are essentially constant among the seedlings analyzed), Fig. 4 presents 7 as a function of lateral root density for three different lateral root lengths. The number of root tips composing a root system per unit vol- ume of soil increases in response to both an aug- mentation in the number of lateral roots attached to laterals tertiary roots 3.50 K 3 climax ‘5 3.25 seral o . Es persistent ES 300 gx ” = 2.65 re) a 2.50 2.25 2.00 0.250 0.275 0.300 0.325 0.350 0.375 0400 0.425 0.450 Lateral root density, djsterais (roots mm") Fic. 4. The number of root tips composing a lodgepole pine seedling root system per mm? of soil (7) as a func- tion of lateral root density (djorerqis) and lateral root length (J),,c,q,). The three labelled points indicate the av- erage values of T for seedlings belonging to the seral, climax, and persistent groups. the taproot and a decrease in the lengths of lateral roots (the radius of the cylindrical mass of soil). The three points in Fig. 4 indicate the average val- ues of T for lodgepole pine seedlings belonging to the seral, climax, and persistent groups (dporerais 1S twice the value of mean NLR from Table 3 divided by a taproot length of 160 mm; /,,,.,,, is mean LRL from Table 3). Bearing in mind the simplifying as- sumptions associated with the calculation of 7, the absorptive capacities of seedlings from climax pop- ulations only slightly exceed those of seral seed- lings, though, given their longer lateral roots, the former may have access to water and mineral nu- trients from larger volumes of soil than either seral or persistent seedlings. Seedlings derived from per- sistent populations possess distinctively lower ab- sorptive capacities, due mainly to the production of fewer lateral roots (mean NLR = 27.20, 29.66, and 26.68 for seral, climax, and persistent families, re- spectively). A seedling’s resource acquisition potential. Dis- criminant function analysis identified a second, less explanatory, dimension that highlights the variation in the ability of Blue Mountain lodgepole pine seedlings to obtain essential resources from their surroundings. It utilizes information about overall seedling size and the sizes of individual organs in- volved in the interception/uptake of light, water, and mineral nutrients to separate the three succes- sional classes in multivariate space. In accord with the interpretation of the second dimension as a size vector, all five morphological attributes with high loadings (20.50) on the second discriminant func- tion (SHL, NPL, BLL, LRL, and ROB) possess stan- dardized discriminant function coefficients, corre- BRADY: SEEDLING EVOLUTION IN LODGEPOLE PINE 147 10.0 9.5 E€ E 9.0 oy f@)) ce 85 ie SS — climax {o) 12) : = 8.0 a persistent Ko) 2 seral 7.5 7.0 85 90 95 100 105 Lateral root length, /)tera; (mm) Fic. 5. Total root length (R) as a function of lateral root length (jierq)- The three labelled points mark the average total root lengths for the three lodgepole pine successional classes. lations among themselves, and loadings of similar sign (Tables 5 and 6). ROB (root biomass), the morphological charac- teristic having the greatest loading on the second discriminant function, grades the overall size of the seedling root system. ROB does not convey infor- mation about shape, 1.e., the pattern of subdivision of the root system. Instead, it incorporates, and in- evitably confounds, two aspects of development: total root length (the sum of the lengths of all roots) and average root caliper. Consider a lodgepole pine seedling in which all lateral roots are identical in length, each lateral root possesses the same number of attached tertiary roots, and all tertiary roots are equal in length. Under these conditions, the follow- ing equation expresses the total root length in mm: [— Uapee a Ciera ater Taal Teves Ube Cerne epic Urepap root ( 2) where L,,,,o01 18 taproot length; 11,,,.,1, 18 the number of lateral roots attached to the taproot; /,,,.,,; 18 the length of a lateral root; d,.nicry roors IS a density, the average number of tertiary roots that arise per unit length of lateral root; and L,.,.:0- oor 1S the length of a tertiary root. The quantity Tenet icomees is the sum of the lengths of all lateral roots. The product Win teraiUiaprour Greeny Gee eRe oC DE Coe Mes ane SUT of the lengths of all tertiary roots. The results of this study supplied constant values for /,,,,., (160 MM), Njprerals (95.8, twice the mean of 34 family means on NER irom able 3), and do. 00-29 mm~'). The approximate median of the range of mature tertiary root lengths of forest trees reported by Sutton and Tinus (1983) provided a reasonable value for Liviary roo, (2-0 mm). Figure 5 shows that the total length of a lodgepole pine seedling root system (R) increases in direct proportion to lateral 148 MADRONO (x 103) persistent Total primary leaf area, L (mm?) 34.0 34.5 35.0 35.5 36.0 36.5 37.0 37.5 38.0 Blade length, b (mm) Fic. 6. Graph of total primary leaf area (L) for various values of p (number of primary leaves) and b (blade length). The three labelled points correspond to the total primary leaf areas of average lodgepole pine seedlings belonging to the seral, climax, and persistent groups. root length (/,,,.,;). The three points in Figure 5 de- note the values of R for typical seedlings belonging to the seral, climax, and persistent groups (based on mean LRL from Table 3). The typical specific root lengths (calculated from R and mean ROB in Table 3) associated with seral, climax, and persis- tent seedlings are 296.06, 264.64, and 244.62, re- spectively. An increase in total root length allows the root system to explore the soil and locate un- exploited resources (Fitter 1985, 1991, 1994; Cald- well 1987, 1994). Seral seedlings possess the small- est total root lengths. The total root lengths of seed- lings from climax populations surpass those of seedlings from other population types. Hence, su- perior abilities to find water and mineral nutrients within the soil distinguish climax seedlings. Three shoot system attributes (SHL, NPL, and BLL) help to isolate the seral group along the sec- ond dimension of variation (Table 6). As discussed above, SHL (shoot length) reflects, among other things, the surface area of the green epicotylar stems. Leaf morphology significantly influences the amount of light that a plant can intercept for use in photosynthesis (e.g., Taylor 1975; Givnish 1979, 1988; Givnish and Vermeij 1976). Specifically, NPL (number of needle-like primary leaves) and BLL (blade length) jointly provide information about total primary leaf area, a useful comparative measure of leaf photosynthetic potential among lodgepole pine seedlings. All primary leaves ana- lyzed in this study were dorsiventrally flattened and approximately rectangular in outline. By assuming that the leaves are identical in length and width (1 mm), the following equation gives the total primary leaf area (the sum of the single-sided areas of all primary leaves) in mm? of a lodgepole pine seed- ling: L = pb G) where p is the number of primary leaves and b is blade length. As shown in Figure 6, L increases [Vol. 48 | with both blade length and number of leaves. Based | on the mean values of NPL and BLL from Table 3, | the three points in Figure 6 betoken the total pri- mary leaf areas of average lodgepole pine seedlings | belonging to the seral, climax, and _ persistent | groups. Since they produce fewer and shorter pri- mary leaves, seedlings derived from seral popula- tions are distinguished by having substantially smaller total primary leaf areas than seedlings of other successional affinities. The average total pri- mary leaf area of climax seedlings barely exceeds | that of persistent seedlings. Despite their long > blades, the modest production of primary leaves by | persistent seedlings restricts their total primary leaf | areas. ! | | The competitive competences of Blue Mountain | lodgepole pine seedlings. In a comparative sense, seedlings from seral lodgepole pine populations in | the Blue Mountains are unremarkable in exhibiting | no conspicuous preference for either shoot or root | system growth (Fig. 2; Tables 3 and 6). However, they are notably smaller than seedlings belonging to climax and persistent populations in overall and individual organ size (Fig. 2; Tables 3 and 6). Due to restricted height growth and the production of few and small primary leaves, relatively low pho- tosynthetic potentials characterize seral seedlings (Figs. 3 and 6). Although seedlings derived from seral populations manufacture modest numbers of root tips per unit volume of soil (Fig. 4), root ex- tension growth is anemic. Their small total root lengths (Fig. 5) imply that seral seedlings can ex- plore only limited regions of the soil for water and mineral nutrients. Compared to the juvenile mem- bers of climax populations, seedlings taken from seral populations possess morphological traits that would place them at distinct competitive disadvan- tages wherever light or soil resources are scarce, i.e., within the subcanopy (even in a light or water/ mineral nutrient gap). With reference to life in the subcanopy, seedlings from seral lodgepole pine populations exhibit relatively low competitive com- petences. Allocation patterns in climax seedlings favor root system elaboration over shoot growth (Fig. 2; Ta- bles 3 and 6). In addition, in terms of both general size and the sizes of organs involved in light inter- ception and water/mineral nutrient uptake, climax populations have the largest seedlings of all suc- cessional groups (Fig. 2; Tables 3 and 6). Climax seedlings exhibit unimpressive stem photosynthetic potentials (Fig. 3). However, compared to seedlings from seral and persistent populations, climax seed- lings exhibit superior primary leaf photosynthetic potentials, principally as the result of the fabrica- tion of greater numbers of leaves (Fig. 6). Seedlings from climax populations produce relatively large numbers of root tips per unit volume of soil (Fig. 4) as well as long roots (Fig. 5). Climax seedlings have morphological traits, especially those reflect- | 2001] | ing their abilities to find and absorb extremely scarce soil water and mineral nutrients, that would _ give them clear advantages over other lodgepole pine seedlings within the subcanopy. Seedlings from climax populations possess greater competi- _ tive competences, as regards subcanopy life, than those from seral or persistent populations. Development in seedlings from persistent lodge- - pole pine populations is aimed at expansion of the shoot system to the detriment of root growth (Fig. 2; Tables 3 and 6). Persistent seedlings are roughly comparable in size to those belonging to climax populations, but considerably larger than seedlings from seral populations (Fig. 2; Tables 3 and 6). Be- cause of their incredible heights, the stem photo- synthetic potentials of persistent seedlings far ex- ceed those of seedlings from other population types (Fig. 3). However, since they produce fewer leaves, the primary leaf photosynthetic potentials of per- sistent seedlings are slightly lower than those of seedlings from climax populations (Fig. 6). Al- though moderate root lengths characterize seedlings from persistent lodgepole pine populations (Fig. 5), they bear relatively few root tips per unit volume of soil (Fig. 4). Persistent seedlings have “‘stunted”’ root systems that impair their abilities to locate and extract water and mineral nutrients from the soil. Hence, they probably could not survive competition with climax seedlings in a canopied stand, except, perhaps, in a water/mineral nutrient gap. Seedlings derived from persistent lodgepole pine populations display intermediate competitive competences, as concerns life in the forest subcanopy. Possible sources of unexplained variation in seedling morphology. Error variation accounts for 31.8% of the total variation among 34 Blue Moun- tain lodgepole pine family means on eight attributes pertaining to seedling morphology considered by discriminant function analysis. Here, error variation is equivalent to the variation among families within successional classes. Discriminant function analysis cannot identify specific sources of error variation. However, several factors, including both experi- mental inadequacies and evolutionary forces, rep- resent plausible causes. While failings in the ex- perimental procedure undermine the assumption that all observed morphological variation reflects underlying genetic variation, selection, gene flow, and drift may yield error variation with a genetic basis. A lack of spatial uniformity in the environmental conditions within the greenhouse could account for some proportion of error variation. During devel- opment, the seedlings were randomly distributed within the cluster of trays on the greenhouse bench- es. They were not blocked by family or succes- sional class. Unless it produced skewed distribu- tions of counts or measurements, this factor would benignly promote within-family variability, not variation among family means within a succession- BRADY: SEEDLING EVOLUTION IN LODGEPOLE PINE 149 al group. No visible spatial pattern of variation in above-ground growth performance had emerged by the time of harvest. Therefore, any unintended sys- tematic variation in photon flux density, watering, fertilizer application, temperature, or relative hu- midity probably played a relatively minor role in the generation of morphological variation among families. In forested settings, lodgepole pine seedlings usually form ectomycorrhizae with various basidio- mycetes and ascomycetes following the develop- ment of tertiary roots (Castellano and Molina 1989). Because we sowed the seeds in artificial growing medium, the seedlings used in this study could become naturally ectomycorrhizal only through wind dispersal of spores from fruiting bod- ies on infected trees near the greenhouse. As ec- tomycorrhizae influence the growth of lodgepole pine by enhancing the uptake of water and mineral nutrients (Cline and Reid 1982; Ekwebelam and Reid 1983), the differential infection of seedlings could generate error variation. In fact, careful, but non-microscopic, examinations failed to provide evidence (fungal mantles or trailing hyphae) of a single infected seedling. Diagnostic mistakes also could produce error variation. Since the order in which we diagnosed the seedlings was random, inaccuracies in counting and measuring would tend to increase within-fam- ily variability rather than error variation (unless they rendered skewed data). To minimize system- atic diagnostic inconsistencies, the same individual assessed a given morphological attribute for all seedlings. Selective pressures induced by environmental variables unrelated to successional status (e.g., cli- mate, soil parent material composition, topographic position, and elevation) doubtless are partly respon- sible for the creation of error variation. Although they probably are not independent of population successional status, such symbiotic relationships as herbivory, parasitism, and mutualism may influence the evolution of seedling morphology in ways not directly affecting the outcome of competition for resources (e.g., Coley 1983). In the most general sense, the observed pattern of genetic variation in lodgepole pine seedling morphology represents a compromise among phyletic constraints, concur- rent, but often conflicting, selective pressures im- posed by numerous environmental factors, gene flow, and drift. Conclusions. The purpose of this research was to carry out a test of the hypothesis that the succes- sional role of a plant population, because of its im- plications for the nature of the selective regime ex- perienced by regeneration cohorts, determines, in part, the course of autecological evolution within a lineage of populations. The results demonstrate that, in accordance with predictions deduced from the hypothesis, a rather strong and statistically sig- 150 nificant correlation exists between a heritable pat- tern of variation in competitive competence, as re- vealed by seedling morphology, and the succes- sional status (seral, climax, or persistent) of a lodgepole pine population in the Blue Mountains region of northeastern Oregon and southeastern Washington. ACKNOWLEDGMENTS E C. Hall supplied unpublished data and assisted in population selection. M. L. Bradbury, G. A. Brady, J. J. Brady, and C. J. Landen helped in the fieldwork, green- house operations, and seedling diagnosis. C. Carter, J. Jones, and J. McGrath provided advice regarding seed ex- traction and handling. R. Schmid, L. J. Feldman, F C. Hall, and two anonymous reviewers provided comments on earlier versions of the manuscript. Phi Beta Kappa, Sigma Xi, and the Provost’s Research Fund, University of California, Berkeley, provided funding. LITERATURE CITED BAKER, FE S. 1934. Theory and practice of silviculture. McGraw-Hill, New York. 1950. Principles of silviculture. McGraw-Hill, New York. BOGARTZ, R. S. 1994. An introduction to the analysis of variance. Praeger, Westport, CT. BONNER, E T., B. EK MCLEMORE, AND J. P. BARNETT. 1974. Presowing treatment of seed to speed germination. Pp. 127-135, in C. S. Schopmeyer (technical coor- dinator), Seeds of woody plants in the United States. USDA Forest Service, Agriculture Handbook 450. BUHLER, A. 1918. Der Waldbau nach wissenschaftlicher Forschung und praktischer Erfahrung, Band 1. Ver- lagsbuchhandlung Eugen Ulmer, Stuttgart. CALDWELL, M. M. 1987. Competition between root sys- tems in natural communities. Pp. 167-185, in P. J. Gregory, J. V. Lake, and D. A. Rose (eds.), Root de- velopment and function (Society for Experimental Bi- ology, Seminar Series 30). Cambridge University Press. 1994. Exploiting nutrients in fertile soil micro- sites. Pp. 325-347, in M. M. Caldwell and R. W. Pearcy (eds.), Exploitation of environmental hetero- geneity by plants: ecophysiological processes above- and belowground. Academic Press, San Diego. and J. H. Richards. 1986. Competing root sys- tems: morphology and models of absorption. Pp. 251-273, in T. J. Givnish (ed.), On the economy of plant form and function. Cambridge University Press. CASTELLANO, M. A. AND R. MOLINA. 1989. Mycorrhizae. Pp. 101-167, in T. D. Landis, R. W. Tinus, S. E. McDonald, and J. P. Barnett (eds.), The container tree nursery manual. USDA Forest Service, Agriculture Handbook 674. CLINE, M. L. AND C. P. P. REID. 1982. Seed source and mycorrhizal fungus effects on growth of container- ized Pinus contorta and Pinus ponderosa seedlings. Forest Science 28:237—250. CoLey, P. D. 1983. Herbivory and defensive characteris- tics of tree species in a lowland tropical forest. Eco- logical Monographs 53:209—233. EKWEBELAM, S. A. AND C. P. P. REID. 1983. Effect of light, nitrogen fertilization, and mycorrhizal fungi on growth and photosynthesis of lodgepole pine seed- MADRONO [Vol. 48. lings. Canadian Journal of Forest Research 13:1099— | 1106. FISHER, J. B. 1986. Branching patterns and angles in trees. | Pp. 493-523, in T. J. Givnish (ed.), On the economy | of plant form and function. Cambridge University | Press. FITTER, A. H. 1985. Functional significance of root mor- | phology and root system architecture. Pp. 87-106, in A. H. Fitter, D. Atkinson, D. J. Read, and M. B. Usher (eds.), Ecological interactions in soil: plants, mi- crobes and animals. Blackwell Scientific, Oxford. . 1991. The ecological significance of root system | architecture: an economic approach. Pp. 229-243, in | D. Atkinson (ed.), Plant root growth: an ecological | perspective (Special Publication Number 10 of the | British Ecological Society). Blackwell Scientific, Ox- ford. 1994. Architecture and biomass allocation as components of the plastic response of root systems to soil heterogeneity. Pp. 305-324, in M. M. Caldwell and R. W. Pearcy (eds.), Exploitation of environmen- tal heterogeneity by plants: ecophysiological process- es above- and belowground. Academic Press, San Di- ELO: FRANKLIN, J. EF AND C. T. DyRNEss. 1988. Natural vege- tation of Oregon and Washington. Oregon State Uni- versity Press, Corvallis. GIVNISH, T. J. 1979. On the adaptive significance of leaf form. Pp. 375-407, in O. T. Solbrig, S. Jain, G. B. Johnson, and P. H. Raven (eds.), Topics in plant pop- ulation biology. Columbia University Press, New York. . 1986. Biomechanical constraints on crown ge- ometry in forest herbs. Pp. 525-583, in T. J. Givnish (ed.), On the economy of plant form and function. Cambridge University Press. . 1988. Adaptation to sun and shade: a whole plant perspective. Australian Journal of Plant Physiology 15:63-92. . 1995. Plant stems: biomechanical adaptation for energy capture and influence on species distributions. Pp. 3-50, in B. L. Gartner (ed.), Plant stems: physi- ology and functional morphology. Academic Press, San Diego. and G. J. Vermeij. 1976. Sizes and shapes of liana leaves. American Naturalist 110:743-778. GNANADESIKAN, R. 1977. Methods for statistical data anal- ysis of multivariate observations. John Wiley and Sons, New York. GOULD, S. J. AND R. E JOHNSTON. 1972. Geographic vari- ation. Annual Review of Ecology and Systematics 3: 457-498. HALLE, F, R. A. A. OLDEMAN, AND P. B. TOMLINSON. 1978. Tropical trees and forests: an architectural analysis. Springer-Verlag, Berlin. Horn, H. S. 1971. The adaptive geometry of trees. Prince- ton University Press, Princeton, NJ. JOLICOEUR, P. AND J. E. MOSIMANN. 1960. Size and shape variation in the Painted Turtle, a principal component analysis. Growth 24:339-—354. Karson, M. J. 1982. Multivariate statistical methods: an introduction. Iowa State University Press, Ames, IA. KENDALL, M. G. AND A. STUART. 1966. The advanced the- ory of statistics. Griffin, London. KoHyYAMA, T. 1987. Significance of architecture and al- lometry in saplings. Functional Ecology 1:399—404. KRAMER, P. J. AND T. T. KozLowski. 1960. Physiology of trees. McGraw-Hill, New York. 2001] | KRUGMAN, S. L. AND J. L. JENKINSON. 1974. Pinus L. Pine. | Pp. 598-638, in C. S. Schopmeyer (technical coor- dinator), Seeds of woody plants in the United States. | USDA Forest Service, Agriculture Handbook 450. | Ktppers, M. 1989. Ecological significance of above- ground architectural patterns in woody plants: a ques- tion of cost-benefit relationships. Tree 4:375—379. | LACHENBRUCK, P. A. 1975. Discriminant analysis. Hafner, | New York. _ Linpman, H. R. 1992. Analysis of variance in experimen- | tal design. Springer-Verlag, New York. _ LotAan, J. E. AND W. B. CRITCHFIELD. 1990. Pinus contorta Dougl. ex. Loud., lodgepole pine. Pp. 302-315, in R. M. Burns and B. H. Honkala (technical coordinators), Silvics of North America, Vol. 1, Conifers. USDA Forest Service, Agriculture Handbook 654. Morrison, D. EF 1990. Multivariate statistical methods (3rd ed.). McGraw-Hill, New York. NILSEN, E. T. 1995. Stem photosynthesis: extent, patterns, and role in plant carbon economy. Pp. 223-240, in B. L. Gartner (ed.), Plant stems: physiology and func- tional morphology. Academic Press, San Diego. OweENs, J. N. AND M. Movper. 1984. The reproductive cycle of lodgepole pine. British Columbia, Ministry of Forests, Forestry Division, Research Branch, Vic- toria, British Columbia. PeeET, R. K. 1981. Changes in biomass and production during secondary forest succession. Pp. 324-338, in D. C. West, H. H. Shugart, and D. B. Botkin (eds.), Forest succession: concepts and applications. Spring- er- Verlag, Berlin. PFISTER, R. D. AND R. DAUBENMIRE. 1973. Ecology of lodgepole pine. Pp. 27—46, in D. M. Baumgartner BRADY: SEEDLING EVOLUTION IN LODGEPOLE PINE 151 (ed.), Management of lodgepole pine ecosystems. Washington State University Cooperative Extension Service, Pullman. REYMENT, R. A., R. E. BLACKITH, AND N. A. CAMPBELL. 1984. Multivariate morphometrics (2nd ed.). Aca- demic Press, London. SAKAI, S. 1986. Patterns of branching and extension growth of vigorous saplings of Japapese Acer species in relation to their regeneration strategies. Canadian Journal of Botany 65:1578—1585. STUART, J. D., J. K. AGEE, AND R. I. GARA. 1989. Lodge- pole pine regeneration in an old, self-perpetuating forest in south central Oregon. Canadian Journal of Forest Research 19:1096—1 104. SUTTON, R. E AND R. W. TINuS. 1983. Root and root sys- tem terminology. Forest Science (supplement) 29:1— WS 7: TABACHNICK, B. G. AND L. S. FIDELL, 1983. Using multi- variate statistics. Harper and Row, New York. TAYLOR, S. E. 1975. Optimal leaf form. Pp. 73-86, in D. M. Gates and R. B. Schmerl (eds.), Perspectives of biophysical ecology (Ecological Studies 12). Spring- er-Verlag, New York. TOUMEY, J. W. AND C. F KorsTIAN. 1937. Foundations of silviculture upon an ecological basis (2nd ed.). John Wiley and Sons, New York. VOLLAND, L. A. 1985. Ecological classification of lodge- pole pine in the United States. Pp. 63-75, in D. M. Baumgartner, R. G. Krebill, J. T. Arnott, and G. EF Weetman (eds.), Lodgepole pine: the species and its management. Washington State University Coopera- tive Extension Service, Pullman. WINER, B. J. 1971. Statistical principles in experimental design (2nd ed.). McGraw-Hill, New York. MADRONO, Vol. 48, No. 3, pp. 152-161, 2001 ESTUDIO ANATOMICO DE SWALLENIA (POACEAE: ERAGROSTIDEAE: MONANTHOCHLOINAE), UN GENERO MONOTIPICO DE NORTE AMERICA MARICELA GOMEZ-SANCHEZ Licenciatura en Biologia, Universidad Aut6noma de Querétaro, Cerro de las Campanas s.n., 76010 Querétaro, Qro, México PATRICIA DAVILA-ARANDA Unidad de Biotecnologia y Prototipos, Escuela Nacional de Estudios Profesionales, Universidad Nacional Autonoma de México, Ave. de los Barrios s.n., Los Reyes, Iztacala, Tlalnepantla, 54090 Edo. de México, México JESUS VALDES-REYNA Departamento de Botanica, Universidad Aut6noma Agraria Antonio Narro, Buenavista, Saltillo 25315 Coahuila, México RESUMEN Swallenia Soderstr. & H. KF Decker es un género endémico de California, Estados Unidos de América, extremadamente raro, tiene afinidades taxon6micas inciertas y se encuentra en peligro de extincidn. Se presentan observaciones de morfologia, anatomia foliar y micromorfologia de la lema. Muestra una es- tructura ““Cloridoide”’, anatomia Kranz tipo PCK o PEP-ck y subtipo X,MS+. La lamina foliar, vista en seccion transversal, es nodular debido a que ambas superficies, abaxial y adaxial, tienen surcos profundos. El clorénquima es continuo entre haces vasculares contiguos. Las células de parénquima incoloro, co- munes en la subtribu Monanthochloinae, estan ausentes. La vaina externa del haz es redonda. La epi- dermis, en vista superficial, carece de papilas y los estomas estan confinados a los surcos y estan cubiertos por aguyones y una célula “‘papiliforme” intercostal. Las células cortas costales se distribuyen en largas hileras de hasta 15 células o mas. Los caracteres observados en la lema tienen correlaci6n con los de la lamina foliar, sin embargo, no muestran un patr6n definido. Se describe su habitat y se hacen algunas consideraciones adaptativas. Las caracteristicas anatOmicas no apoyan la ubicacion del género Swallenia en la subtribu Monanthochloinae. ABSTRACT Swallenia Soderstr. & H. EF Decker is a rare genus, with uncertain taxonomic affinities, endemic to California, USA and endangered. Morphology, leaf blade anatomy and lemma micromorphology are presented. It shows a “Chloridoid” structure, Kranz anatomy PCK or PEP-ck type and X,MS+ subtype. The leaf blade, as viewed in transverse section, is nodular because abaxial and adaxial surfaces have deep grooves. Chlorenchyma is continuous between adjacent vascular bundles. Colourless parenchyma cells, that are common in the Monanthochloinae subtribe, are absent. Outer bundle sheath is round. In the epidermis, as seen in surface view, papillae are absent and stomata occur in the grooves and are obscured by prickles and an intercostal “‘papilliform” cell. Costal short cells are arranged in rows to I5 cells or more. Lemma micromorphology characters are correlated to those of the leaf blade, but they do not show a characteristic pattern. Habitat and adaptative considerations are described. The anatomical characters do not support the placement of Swallenia genus as belonging to Monanthochloinae subtribe. Swallenia Soderstr. & H. KF Decker, con su tnica especie (S. Alexandrae (Swallen) Soderstr. & H. E Decker) es una planta rizomatosa, perenne; con la ligula en una linea de pelos; inflorescencia una panicula contraida y escasa; glumas casi tan largas como la espiguilla, membranosas, 7—11 nervios; le- mas 5—7 nervios, villosas en los margenes, mu- cronadas, palea villosa en el margen. Swallenia fue colectado por Annie M. Alexander y Louise Kellogg en Mayo de 1949 en la localidad del Valle de Eureka, Condado de Inyo en Califor- nia. Swallen (1950) reconoci6 algunos rasgos Uuni- cos y lo describid6 como Ectosperma, un género monotipico. El nombre genérico se invalid6 porque ya se habia usado, en 1903, para un género algal. Mas tarde, Soderstrom y Decker (1963) lo renom- braron en honor a Jason R. Swallen (1903-1991), distinguido agrost6logo americano que contribuy6 de manera destacada al conocimiento de las gra- mineas del Nuevo Mundo, y designaron como es- pecie tipo a Swallenia alexandrae (Swallen) Sod- erstr. y H. EK Decker. Swallen (1950) consider6 este género como miembro de la tribu Festuceae pero con base en caracteristicas (glumas tan largas como las espi- guillas) afines a la tribu Aveneae. Pilger (1954) también lo ubico en la tribu Festuceae junto con los géneros Melica L., Schizachne Hack., Lycochloa | Samuels, Vaseyochloa Hitche., Anthochloa Nees y 'Meyen, Neostapfia Davy y Ramosia Merr. No ob- stante, con base en caracteres morfolégicos y aque- llos anat6micos generados por Metcalfe (1960), | Stebbins y Crampton (1961) lo incluyeron en la tri- bu Monanthcohloinae (=Aeluropodineae) al lado 'de géneros como Aeluropus Trin., Distichlis Raf., _Monanthochloé Engelm., Jouvea Fourn. y Vasey- -ochloa. Gould y Shaw (1983) reconocieron la tribu | Aeluropodeae e incluyeron los géneros Distichlis, Allolepis Soderstr. y H. KF Decker, Monanthochloé -y Swallenia. Diversos autores como Clayton y Ren- -voize (1986), Peterson et al. (1995, 1997) y Watson y Dallwitz (1992) reconocen a Swallenia alexan- drae como integrante de la subtribu Monanthoch- loinae y tribu Eragrostideae. Peterson et al. (1995) proponen una clasificacion tribal para las Eragros- tideas del nuevo Mundo y reconocen la subtribu Monanthochloinae con seis géneros americanos (Distichlis, Jouvea, Allolepis, Monanthochloe, Reederochloa y Swallenia) excepto Aeluropus que se distribuye en el Viejo Mundo. Observaciones previas de la anatomia foliar de Swallenia (G6mez y Koch, 1993) y micromorfo- logia de la lema de Monanthochloinae (G6mez 1998) revelaron ciertas caracteristicas notables y unicas. Estas caracteristicas anat6micas, sus afini- dades taxonomicas inciertas y su distribucion res- tringida combinadas con el pobre conocimiento acerca de su habitat, morfologia, micromorfologia de lema, etc., hacen de Swallenia un género im- portante y merecedor de un estudio exhaustivo y detallado. Este género esta probremente represen- tado en los herbarios y, hasta ahora, se conocen solamente los ejemplares tipo. Las primeras colec- tas datan de los anos 50’s y el género se estableci6 en 1963. No se registran mas colectas sino hasta el 22 de Mayo de 1976 y posteriormente hasta el 30 de Marzo de 1986. Esta ultima es una fotografia de la planta en su medio natural y esta depositada en la colecci6n del Herbario de la Universidad de Ca- lifornia (UC). El objetivo de este trabajo es proveer informacion amplia y detallada que servira para en- tender las afinidades taxonomicas del género y pos- teriormente evaluar sus relaciones filogenéticas. MATERIALES Y METODOS Se revisaron 23 especimenes provenientes de los herbarios CAS, RSA, UC y US (Cuadro 1), acr6- nimos segtin Holmgren et al. (1990). Para la ana- tomia foliar se consideraron laminas basales y de su parte media se tomo un fragmento de | cm. Para las observaciones con microscopia Optica, se hicie- ron preparaciones anatOmicas de la epidermis ab- axial en vista superficial y la estructura interna vista en secciOn transversal siguiendo las técnicas de ras- pado directo y secciones manuales de Metcalfe (1960) y Gémez y Koch (1998). El montaje se hizo en jalea glicerinada, sin tinci6n. Para las observa- ciones al microscopio electrénico de barrido 2001] GOMEZ-SANCHEZ ET AL.: ANATOMIA DE SWALLENIA, GENERO MONOTIPICO 153 CUADRO 1. COLECTORES Y COLECCIONES DONDE ESTAN DEPOSITADOS IOS ESPECIMENES EXAMINADOS. Swallenia alexandrae (Swallen) Soderstr. y H. EK Decker Annetta carter 2784, 8.Jun.1950, topotipo (RSA, UC, US). Annie M. Alexander y Louise Kellogg 5655, 24.May.1949, isotipo (UC). Christopher Davidson 4015, 22.May.1976 (RSA). Douglas Powell s.n., 21.Abr.1952, topotipo (UC, US). D. E. Anderson. 2406, 6.Abr.1963 (UC). D. W. Taylor s.n., 17.Nov.1979 (UC). Jack L. Reveal y Arlene H. Reveal 37, 13.May.1962, to- potipo (RSA, UC). James D. Morefield y Douglas H. McCarty 3297, 30.Mar.1986, topotipo (RSA, UC). John C. Roos 6352, 6355 y 6365, 13.May.1955, topotipo (CAS, RSA, UC). John & Lucille Roos 6175, 28.Jun.1954. topotipo (CAS, RSA UG): John & Lucille Roos 6467, 21.Jun.1955 (RSA). John y Lucille Roos 6320, 10.Abr.1955 (RSA, UC). Mary de Decker 1457, 2.Abr.1961, topotipo (RSA). (MEB), las muestras se trataron previamente con cloroformo por 2—6 horas para remover la cuticula y algunas impurezas. La micromorfologia de la lema se estudid en ejemplares de herbario con ayuda del MEB. Las muestras se tomaron del primero y segundo fl6s- culo de espiguillas maduras y se trataron previa- mente con cloroformo de 2—4 horas para remover la cuticula. Después de lavar con agua, se montaron en portaobjetos, se Ilevaron a un bano de aluminio y se tomaron las fotomicrografias con un micros- copio Hitachi S-2460N. En algunos casos (Figs. 3, 4A-B) se hizo el bafo en oro y se empleo un mi- croscopio JEOL. Para las descripciones de anatomia se sigui6 la terminologia propuesta por Ellis (1976, 1979) con algunas modificaciones. Las ilustraciones de ana- tomia foliar, para microscopia Optica, se hicieron con ayuda de una camara clara. La epidermis aba- xial esta orientada de tal forma que el eje longitu- dinal de la lamina aparece en posici6n horizontal en la ilustraci6n y el apice hacia la derecha. RESULTADOS Swallenia Soderstr. & H. E Decker, Madrono 17: 88. 1963. Tipo: Swallenia alexandrae (Swallen) Soderstr. and H. EF Decker. Este género es endémico, extremadamente raro e interesante, de afinidades inciertas y adaptado a so- brevivir en sitios secos y arenosos. Se han generado diferentes criterios acerca de su ubicaci6n taxono- mica lo que ha hecho dificil el entendimiento de sus afinidades filogenéticas. Morfologia (Fig. 1). Planta perenne, rizomatosa, rizomas__ robustos. Culmos amacollados, erectos, 10—60 cm de altura (formando colonias 0 masas | 154 MADRONO [Vol. 48. ZZ LEE \ COpUHE Mao. Fic. 1. Swallenia alexandrae (Swallen) Soderstr. & H. E Decker. A. Habito de la planta. B. Panicula. C. Espiguilla. | D. Ligula. E. Flésculo. E Cariopsis. (Ilustracién de Reveal & Reveal a7 UC): / GOMEZ-SANCHEZ ET AL.: ANATOMIA DE SWALLENIA, GENERO MONOTIPICO Ie) Fic. 2. Swallenia alexandrae (Swallen) Soderstr. & H. EK Decker. Lamina foliar vista con microscopio Optico (Ilus- tracion de D. Powell s/n, US). A. Epidermis abaxial. B y C. Secci6n transversal. RC = regi6n costal, RI = region intercostal, a = aguij6n, cb = células buliformes, ccc = liforme’’, e = célula corta costal, cl = clorénquima, cp = célula “‘papi- estoma, eab = epidermis abaxial, ead = epidermis adaxial, es = esclerénquima, hvp = haz vascular primario, hvs = haz vascular secundario, ve = vaina externa del haz. extensas de hasta dos metros de diametro), cespi- tosos, pilosos. Vainas mas largas que los entrenu- dos, margenes lisos, ciliadas (villosas en la garganta y el collar), las maduras basales fibrosas, perennes (en la base de las plantas), color ocre-amarillentas. Articulaci6n vaina-lamina conspicua. Ligula una li- nea de pelos, 1.0—1.5 mm de largo. Auriculas au- sentes. Lamina foliar linear; plana a convoluta; con- spicuamente nervada, nervios 30—40 en la super- ficie abaxial; pulverulenta; rigida; punzante; Apice atenuado; 3—13 cm de largo; 3—5 mm de ancho. Inflorescencia una panicula contraida; 4—10 cm de largo; 0.5—1.0 cm de ancho; ramas ligeramente ad- presas; exerta 0 parcialmente incluida en las vainas Superiores; eje principal escabroso; ramas primarias adpresas al eje principal; pedicelos pilosos, esca- brosos. Desarticulaci6n arriba de las glumas, lema y palea caen como una unidad. Callo piloso. Espi- guillas (5)10—15(20) mm de largo, casi tan largas como anchas, solitarias, lateralmente comprimidas, hialinas a verde-amarillentas, ovadas, pedicelos 0.5—1.0 mm de largo. Glumas tan largas como la espiguilla o ligeramente mas largas, lisas, glabras, membranosas a coriaceas. Primera gluma 6—7 ner- vios, 10 mm de largo, apice acuminado. Segunda gluma, 7—11 nervios, 9-10 mm de largo, apice acu- minado. Raquilla conspicuamente marcada entre los fl6sculos, entrenudos 0.3—0.5 mm de largo, pu- bescentes a escabrosos. Fl6sculos 3—7 por espigui- lla. Fldsculos estériles presentes, arriba de los fl6s- culos fértiles. Lema entera; 7-8 mm de largo; sin arista; papiracea; densamente pilosa a villosa; 5—7 nervada, nervios conspicuos y pilosos. Palea pilosa; 5—6 mm de largo; papiracea; 3—4 nervada, nervios conspicuos; margenes no envolviendo el fruto, ci- liados, lisos. Lodiculas truncadas. Estambres 3. An- teras 2.0 mm de largo, rojizas a purpura. Estigmas 2. Cariopsis redondeado a ovoideo elipsoidal, a ve- ces dorsoventralmente comprimido, con las bases de los estilos persistentes. Cariopsis con las bases 156 MADRONO Fic. 3. Swallenia alexandrae (Swallen) Soderstr. & H. E Decker. Lamina foliar vista al microscopio electrénico de barrido (Ilustraci6n de D. Powell s/n, US). A. Epidermis abaxial. B y C. Epidermis adaxial. a = aguij6n, cp = célula papiliforme, e = estoma. de los estilos relativamente cortos. Nimero bdsico de cromosomas X = 10, 2n = 20. Anatomia Seccion transversal (Figs. 2B, C). Involucion de la lamina en forma de “‘V’’, muy amplia, nodular. Cara abaxial con surcos profundos y costillas pro nunciadas; apice de las costillas cuadrado, redon- deado, o plano. Cara adaxial con surcos profundos y costillas pronunciadas; dpice de las costillas cua- drado, redondeado, o plano. Raz6n profundidad del surco:grosor de la lamina 0.4—0.5. Haces vasculares 30-32; usualmente haces vasculares primarios y se- cundarios, a veces dos terciarios alternando con los haces vasculares primarios. Vaina externa del haz completa, redondeada en todos los haces vascula- res. Extension de la vaina externa del haz ausente. Esclerénquima abaxial y adaxial; copioso; el aba- xial en grandes hebras frente a los haces vasculares, 10-14 células de ancho y 3-6 células de alto; el adaxial en grandes hebras sobre los haces vascu- lares, 8-15 células de ancho y 4-8 células de alto; ausente sobre los haces vasculares terciarios cuando éstos existen. Esclerénquima abaxial intercostal au- sente. Clorénquima radial alrededor de los haces vasculares; sus células alargadas y estrechas; con- tinuo entre haces vasculares sucesivos. Células in- coloras ausentes. Células buliformes en grupos adaxiales discretos, regulares; célula central ligera- mente inflada; 0.05 mm de largo; casi a la mitad del grosor de la lamina o mas corta. Epidermis abaxial (Figs. 2A, 3A) Regiones cos- tal e intercostal bien diferenciadas. Células largas de 20 wm de largo; paredes anticlinales horizonta- les sinuosas en la regi6n costal y onduladas en la intercostal; paredes periclinales con punteaduras simples abundantes, conspicuas. Estomas abundan- tes, 15 estomas por campo visual; r6mbicos (vistos en superficie), a veces en domo; en 1-2 hileras ha- cia los margenes de la regién intercostal; usual- mente cubiertos por una célula ‘‘papiliforme’’ in- tercostal, a veces ocultos por aguijones costales e intercostales. Células cortas costales en hileras lar- gas (de hasta 15 células), cuadradas a redondeadas. Células cortas intercostales ausentes; cuando pre- sentes entonces solitarias, altas y estrechas, rara vez GOMEZ-SANCHEZ ET AL.: ANATOMIA DE SWALLENIA, GENERO MONOTIPICO 157 Fic. 4. Swallenia alexandrae (Swallen) Soderstr. & H. EK Decker. Micromorfologia de lema vista al microscopio electronico de barrido. A y B. Parte basal (Ilustraci6n de A. Carter 2784, UC). C y D. Parte media (Ilustracion de D. Powell s/n, US). a = aguij6n, cc = célula corta. cuadradas. Cuerpos de silice siempre en la region costal y ausentes en la intercostal; redondeados, cuadrados, o en forma de silla de montar; 3.0—4.5 wm de largo. Papilas ausentes. Aguijones abundan- tes, usualmente intercostales, cubriendo los esto- mas, a veces costales. Micropelos no vistos. Ma- cropelos ausentes. Micromorfologia de la lema (Fig. 4). Base de la lema pilosa. Regiones costal e intercostal no difer- enciadas. Células largas con paredes anticlinales horizontales fuertemente onduladas a sinuosas. Papilas ausentes. Células cortas abundantes, usual- mente solitarias, a veces en pares, altas y estrechas, cuadradas o redondeadas. Cuerpos de silice redon- deados a ligeramente cuadrados. Aguijones abun- dantes. Micropelos no vistos. Macropelos ausentes. DISTRIBUCION Norteamérica. Nativa del Nuevo mundo. Solo se le conoce de los Estados Unidos de América. Crece en las dunas arenosas de Eureka al sureste de Ca- lifornia (Fig. 5). Especie endémica del Valle de Eu- reka, en el Condado de Inyo, en California. Habitat Swallenia alexandrae es una planta xerofitica y se desarrolla en vertientes 0 laderas bajas de dunas arenosas, en sitios secos y a una altitud entre los 2900 y 3400 m. Crece en grandes colonias 0 masas de hasta un metro o mas de diametro y profunda- mente incrustada en arenas siliceas (Henry 1979). Se le encuentra asociada con Chaetadelpha whee- leri, Cleome sparsifolia, Coldenia plicata, Dalea polyadenia, Dicoria clarkae, Ligodesmia juncea, Oenothera deltoides y Stanleya pinnata spp. In- yoensis, entre otras. Estatus de Conservacion Esta especie se encuentra en las listas rojas de la Union Internacional de Conservacion de la Natur- aleza (Walter y Gillett 1998). Se considera en pe- ligro de extinci6n en los Estados Unidos y especie rara en California (Smith y York 1984; Hickman O86). DISCUSION Y CONCLUSIONES La ubicacion de Swallenia alexandrae, en la sub- tribu. Monanthochloinae, hasta ahora habia sido apoyada con base en su morfologia externa. Los estudios anatOmicos en este género son escasos. Metcalfe (1960) y Watson y Dallwitz (1992) sena- lan parcialmente algunas caracteristicas de la lami- na foliar. Esta investigacioOn proporciona informa- cidn anat6mica amplia, adiciona algunos aspectos 158 MADRONO 130° 100° Fic. 5. de la micromorfologia de la lema y detalles impor- tantes que permiten entender mejor sus afinidades. Swallenia comparte algunos caracteres morfol6- gicos con los otros miembros de la subtribu Mon- anthochloinae, tales como la lema papiracea y con 5—7 nervios y la ligula definida por una linea de pelos. No obstante, este género se diferencia del resto de los componentes, por poseer cariopsis con las bases de los estilos persistentes y cortos y porque sus flores son perfectas o hermafroditas. Esta observaci6n coincide con la opinion de Peter- son et al. (1997). La informacion citolé6gica se reduce al conoci- miento del numero cromosémico. La especie tiene un numero cromosomico X = 10 (Peterson et al. 1997) 0 el diploide 2n = 20 (Hickman 1996). Este numero coincide con el de algunos de los géneros restantes de la subtribu. Las caracteristicas anat6micas que se manifiestan en la secciOn transversal de la lamina foliar de Swallenia alexandrae evidencian una estructura ‘““cloridoide”’ segun lo propuesto por Brown (1958). Presenta todos los haces vasculares rodeados por una vaina externa parenquimatica conspicua, provista de grandes cloroplastos, abundantes y dis- 30° 70° Mapa de distribucion de Swallenia alexandrae (Swallen) Soderstr. y H. E Decker. tintos de los del clorénquima. Los haces vasculares primarios poseen ademas una vaina interna fibrosa, conspicua, que rodea completamente al xilema y floema. El clorénquima se compone de células de pared delgada y esta dispuesto de manera radial al- rededor de los haces vasculares. E] numero maximo de células clorénquimaticas, existentes entre haces vasculares contiguos, es de 2 a 3, considerandose bajo (Hattersley y Watson 1975). Estos caracteres permiten ademas inferir que, por tener esta estruc- tura anat6mica denominada Kranz, Swallenia per- tenece a las plantas C,, es decir, que en su via fo- tosintética sigue el ciclo de Hatch y Slack (1966). Esta estructura anatomica Kranz es conocida como tipo PCK y la especie es PEP-ck, Tiene una vaina del haz doble y cloroplastos especializados dis- puestos de manera centrifuga, esto es, estan hacia la pared celular externa de la célula de la vaina Kranz. Asimismo, es evidente el subtipo anat6mico X,MS-+ por la presencia de células (de la vaina interna) entre los elementos del metaxilema y la vaina parenquimatica (Kranz), sobre todo en los ha- ces vasculares primarios (Brown 1975, 1977; Ellis 1977; Hattersley y Watson 1976; Watson y Dallwitz 1992). [ 159 Z 7 ANATOMIA DE SWALLENIA, GENERO MONOTIPICO : < = 1) N na ani O Z < a N a) = fo) ) SOSBOSO BSO][IA-OUu soqyuasoid soquosne opriduimn.4419jut epeao epeAo eUUIN[OS ud seiqoy O seqey soperxepe Je[Npou-ou QUILIOJIARTO soquosne soquosoid ojugsne soJyuosne so]uosne Ssofenxostun ao] yIoyJUuvUuopy SOSBOSO BSOT[IA-OU soquosoid sojuasne opriduin.9utr epeAao epeAao euuN[od ud seiqoay oO seqen Soyerxepe Je[npou-ou DUIIOJIARTO soquosne soquasoid oyuosne sojuosne soyuosne sojenxosiun DO]YIOAIPIIY sajuondaly BSO][IA-OU soquosoid SOSBOSO opidum.19}utr epeao epeAao evuuIN[Oo ud seqe.n sayerxepe Je[npou-ou DULIOFIARTO sajuasne saqqjuasoid ojyUuUIsSNne SsoJuosne sojuosne SeLIPOIJeULIOY sndoanjayv SOSROSO BSO][IA-OU soquosoid sosevosa optduini19}ul epeAO eBpeAo euuN[OS ud seiqoy O seqriy soyerxepe Je[npou-ou QUIIOJIARIO sajuasne sojuasaid ojJuosne sgjuosne soyuosne Soyenxostun sida]0]]1V saquandady sojuepunqe BSO][IA-OU soqyuosoid soseosa oprduinii9}ul epeAO epeao euuN[os ud seqen SoleIxepe Je]npou-ou DULIOJIARTO sojuasne sojuosaid oyuosne soyuosne SoJUISNeS soyenxasiun SIYOUSIG saquondaly BSO][IA-OU soquosoid SOSBOSO opriduinii9qut epeAO PPeAO euUUN[OS Ua seiqoy O seqeiy soyerxepe Je|npou-ou DULIOJLARTO soqjuosne soquasoid ojJUuIsSne soyuosne soJuosne sofenxostun paanor sajuepunqge BSOTIIA sojuasne sajuepunqe DOI[IS ap sodiang aseg sejideg souolinsy VWAT 4d VIDO IOAAOWORDIN onunuos epuopol epuopal sajuasne seiqoy sopuris sojerxepe A soyerxeqe Ieynpou sajuosne soyuepunqe ajuasne ajuasoid sajuasoid sojuasoid SPUPOIJLULIOY DIUI]]DMG sons -Wu0d AH anus eullnbusi0j9 zey [OP AA BP] Op euUlIOy AH SO] op eulioy se -IO[OOUL Se[N][Id Se] op OSILUIY evulimnbuaraposq CUIWIL] BP] Op SOdINS [esIoAsued} UOIDIIS Ud BISIA PUIWILT] sojedorotu ap ePUIO{ souolinsy sayeysooroqur sepideg BUIO}SA [9 aiqnd onb ourojsiided veinjag SPINIID CI op seis -[IY Ud S3TB}JSOD SPIO sK[N[ID avil1Od VINOLVNV sauajsisiod sopysa SO] Op saseq uOod sIsdolIeED SOOT VIDO 1IOdYOW SHaALIVAVO ‘ZeYy [OP BUIO}XO BUIVA = AA ‘Sole[NOSeA SoOVY = AH “AVNIOTHOOHLNVNO|Y NAIMLENS V1 AG OLSAY TA VINATIVMGS OYANAD TV NANONILSIG ANO SAMALOVUVD “TJ OAAVAD 160 MADRONO La lamina foliar es nodular por la presencia de costillas pronunciadas y surcos profundos en ambas superficies (Fig. 2C). La presencia de los surcos profundos y los estomas y aguijones (Fig. 3A, B) confinados a dichos surcos, son caracteristicas que permiten, a la planta, su adaptacion al habitat ex- tremo donde crece. Las grandes hebras de escle- rénquima y los surcos profundos confieren a la lamina rigidez y evitan su colapso. En la epidermis no se observaron micropelos, sin embargo, Watson y Dallwitz (1992) senalan para el género micropelos alargados tipo “‘cloridoide”’. Es- tos, son claramente bicelulares y las células tienen paredes delgadas lo que los [leva al colapso. Esto ultimo explica su ausencia en las ilustraciones. Los estomas estan cubiertos por aguijones 0 por una célula “‘papiliforme”’ intercostal (Figs. 2A, 3A— C). Sus células subsidiarias tienen forma triangular o redonda lo que le confiere al estoma las formas r6mbica 0 en domo respectivamente. Las células “‘papiliformes’”’ intercostales, presen- tes en la epidermis, son tnicas en el género Swal- lenia. Estas células estan en los margenes de los surcos y cubren parcial o totalmente los estomas como lo senalan también Watson y Dallwitz (1992). Sin embargo, la epidermis no tiene las papilas tipi- cas ni las cantidades caracteristicas que se observan en las especies de ambientes secos y salinos (Figs. 2A, 3B-C). Estas células ‘“‘papiliformes”’ intercostales, probablemente funcionan de manera similar y sus- tituyen a las células buliformes y a aquellas de pa- rénquima incoloro que ocurren en el resto de la subtribu, sin que éstas sean homologas. Por su pos- icion y estructura, las células “‘papiliformes”’ po- siblemente sirven como canales para la entrada de luz e intervienen en la apertura y cierre de los sur- cos cuando la lamina esta abierta 0 cerrada respec- tivamente. Esto ultimo [eva a la planta a controlar la evapotranspiracion. La micromorfologia de la lema (Fig. 4A—C) muestra correlaci6n con las caracteristicas de la lamina foliar. Ademas de ser pilosa en la base, po- see abundantes aguijones y células cortas. También carece de papilas y micropelos, mientras que, en el resto de los géneros de la subtribu Monanthochloi- nae las papilas estan presentes, los aguijones son escasos y los cuerpos de silice estan con menor frecuencia 0 son escasos. Los cuerpos de silice no muestran un patron definido como lo establecen Valdés y Hatch (1991) en su estudio de la lema de las Eragrostideas. Todos los géneros de la subtribu Monanthoch- loinae comparten la presencia de surcos profundos en la cara adaxial de la lamina y el esclerénquima adaxial distribuido en hebras. Sin embargo, [lama la atenci6n la separaci6n de Swallenia por 16 de los 18 caracteres diagnésticos que se muestran en el Cuadro 2 (células cortas costales en hileras de hasta 15 células de largo, células papiliformes cu- briendo los estomas, papilas ausentes, aguijones abundantes, micropelos ausentes, lamina nodular, surcos abaxiales y adaxiales, esclerénquima copio- so distribuido en grandes hebras, células incoloras! ausentes, clorénquima continuo entre haces vascu-| lares sucesivos y por la forma de los haces vascu-| lares y de la vaina externa del haz redonda). Esta observacion coincide con la opinidn de Renvoize (1983) en el sentido de que Swallenia presenta al- gunas anomalias con respecto al patr6n anat6mico: general de la tribu Eragrostideae. | Las caracterfsticas anat6micas de la lamina foliar. y la lema no convalidan los criterios de Clayton y. Renvoize (1986) y Peterson et al. (1995, 1997), quienes consideran a Swallenia como un género in-_ tegrante de la subtribu Monanthochloinae. Las ob- | servaciones del presente trabajo permiten inferir que las afinidades de Swallenia son inciertas, no obstante, es claro que este género no esta relacio- | nado con Dizistichlis, Allolepis, Jouvea, Monan- | thochlde, Reederochloa (del Nuevo Mundo) y Ae- | luropus (del Viejo Mundo), que son los géneros que | mejor conforman la subtribu. Las caracteristicas | morfold6gicas, anatomicas y de ultraestructura su-_ gieren que Swallenia debe separarse de la subtribu_ Monanthochloinae. AGRADECIMIENTOS Agradecemos a los curadores de los herbarios CAS, GH, RSA, UC y US por el préstamo de ejemplares. A Sara Fuentes del laboratorio de microscopia electr6énica del Instituto de Biologia, UNAM y a Jorge Valdés del Colegio de Postgraduados por las facilidades brindadas para el uso del MEB. A Cuahutémoc Gonzalez de Leén por la ilustraci6n de la planta. A los revisores an6nimos agradecemos su opinion critica al manuscrito. Este trabajo fue financiado por el Consejo Nacional de Ciencia y Tec- nologia (CONACYT/29106-N) y un apoyo parcial para su publicaci6n se obtuvo del proyecto FOMES/9623-04. LITERATURA CITADA Brown, W. V. 1958. Leaf anatomy in grass systematics. Botanical Gazette 119:170—178. . 1975. Variations in anatomy, associations, and origins of Kranz tissue. American Journal Botany 62(4):395—402. —. 1977. The Kranz syndrome and its subtypes in grass systematics. Bulletion of the Torrey Botanical Club 23(3):1-97. CLAYTON, W. D. y S. A. RENVOIZE. 1986. Genera gramin- um: grasses of the world. Kew Bull. Additional Series XI. Her Majesty’s Stationery Office, Royal Botanic Garden, Kew. London. 389 pp. ELuis, R. P. 1976. A procedure for standardizing leaf anat- omy in the Poaceae. I. The leaf-blade as viewed in transverse section. Bothalia 12(1):65—109. . 1977. Distribution of the Kranz syndrome in the southern african Eragrostideae and Panicoideae ac- cording to bundle sheath anatomy and cytology. Agroplantae 9:73—110. . 1979. A procedure for standardizing leaf anatomy in the Poaceae. II. The epidermis as seen in surface view. Bothalia 12(4):641—671. GOMEZ-SANCHEZ, M. 1998. Micromorfologia de la lema de Monanthochloinae (Poaceae: Chloridoideae: Era- grostideae). Resumen. 7° Congreso Latinoamericano de Botanica y XIV Congreso Mexicano de Botanica. México, D. EK México. y S. D. Koch. 1993. Anatomia foliar de Swalenia alexandrae (Poaceae: Chloridoideae. Aeluropodeae) y sus relaciones taxonomicas. Resumen. XII Congre- so Mexicano de Botanica. Mérida, Yuc. México. Gomez-Sanchez, M. y S. D. Koch. 1998. Estudio an- atOmico comparativo de la lamina foliar de Eragrostis (Poaceae: Chloridoideae) de México. Acta Botanica | Mexicana 43:33—56. GouLp, E W. yY R. B. SHAW. 1983. Grass systematics. Sec- ond edition. Texas A & M University Press, College Station, Texas. 397 pp. HatcH, M. D. y C. B. SLACK. 1966. Photosynthesis by sugar-cane leaves. Biochemical Journal 101:103—111. HATTERSLEY, P. W. y L. WATSON. 1975. Anatomical pa- rameters for predicting photosynthetic pathways of grass leaves: “the maximum, lateral cell count”? and the “‘maximum cells distant count’. Phytomorphol- ogy 25(3):325-—333. : . 1976. C, Grasses: an anatomical crite- rion for distinguishing between NADP-malic enzyme species and PCK or NAD-malic enzyme species. Australian Journal Botany 24:297—308. Henry, M. A. 1979. A rare grass on the Eureka Dunes. Freemontia 7(2):3—6. HIcKMAN, J. C. (ed.). 1996. The Jepson Manual. Higher Plants of California. Thirdth edition. University of California Press, Berkeley, California. 1400 pp. HOLMGREN, K. P, N. H. HOLMGREN, Y L. C. BARNETT. 1990. Index Herbariorum. Part I. The herbaria of the world. Eighth edition. New York Botanical Garden, New York. 693 pp. METCALFE, R. C. 1960. Anatomy of Monocotyledons. I. Gramineae. Clarendon Press, Oxford. 731 pp. GOMEZ-SANCHEZ ET AL.: ANATOMIA DE SWALLENIA, GENERO MONOTIPICO 161 PETERSON, P. M., R. D. WEBSTER, J. VALDES-REYNA 1995. Subtribal Classification of the New World Eragrosti- deae (Poaceae: Chloridoideae). Sida 16(3):529—544. ; ; . 1997. Genera of New World Eragrostideae (Poaceae: Chloridoideae). Smithsonian Contribution to Botany 87:1—SO. PILGER, R. 1954. System der Gramineae unter Auschlus der Bambusoideae. Botanische Jahrbicher Fir Syste- matik, Pflanzengeschichte und Pflanzengeographie 76(3):28 1-384. RENVOIZE, S. A. 1983. A Survey of the leaf-blade anatomy in grasses IV. Eragrostideae. Kew Bulletin 38(3):469— 478. SMITH, J. P. JR. y R. YORK. 1984. Inventory of rare and endangered vascular plants of California. Thirdth edi- tion. 178 pp. California Native Plant Society, Special Publication No. 1. SODERSTROM, J. R. Y H. EF DECKER. 1963. Swallenia a new name for the California genus Ectosperma (Gramin- eae). Madrono 17(3):88. STEBBINS, G. L. y B. CRAMPTON. 1961. A Suggested re- vision of the grass genera of temperate North Amer- ica. In Recent Advances in Botany (IX Int. Bot. Con- gress, Montreal, 1959) 1:133—245. SWALLEN, J. R. 1950. Ectosperma, a new genus of grasses from California. Journal of the Washington Academy of Sciences 40(1):19-—21. VALDES-REYNA, J. Y S. L. HATCH. 1991. Lemma Micro- morphology in the Eragrostideae (Poaceae). Sida 14(4):531-549. WALTER, K. S. y H. J. GILLETT (EDS.). 1998. 1997 IUCN red list of threatened plants. Compilado por el World Conservation Monitoring Centre. [UCN—The World Conservation Union, Gland, Switzerland y Cam- bridge, UK. 862 pp. WATSON, L. y M. J. DALLWITz. 1992. The grass genera of the world. C. A. C. International, Wallingford. 1081 Pp. MApDRONO, Vol. 48, No. 3, pp. 162-171, 2001 COMPARATIVE FLOWERING PHENOLOGY OF PLANTS IN THE WESTERN MOJAVE DESERT W. BRYAN JENNINGS Section of Integrative Biology, University of Texas at Austin, Austin, TX 78712 ABSTRACT Above-average precipitation fell on the western Mojave Desert in February and March 1991, which resulted in the flowering of most plant species that spring. This mass germination event occurred despite the almost total lack of rainfall between September and December 1990, a finding inconsistent with previous observations from the Mojave Desert. The 58+ annual species observed on the study site flowered in a sequential manner from early March to mid-June, 1991. Climatic phenomena were similar the following year, as several winter storms produced large amounts of precipitation in February and March 1992. Consequently, at least 63 annual and 15 perennial plant species bloomed over spring 1992. Like the previous year, plants flowered sequentially from early March to late June. An analysis of flowering dates between years suggests that timing of flowering for these species is highly consistent from year to year. Some phylogenetic biases in flowering dates were evident among some of the locally-dominant taxa. Species in the families Brassicaceae and Boraginaceae flowered in early spring, whereas species in the Fabaceae, Asteraceae, and Polygonaceae, bloomed from mid- to late-spring. However, Polemoniaceous species flowered throughout spring. The Mojave Desert has a remarkably diverse ephemeral plant flora consisting of over 250 species and varieties (Shreve and Wiggins 1964). As a re- sult, this region has long attracted biologists inter- ested in germination and flowering phenomena of desert plants (Went 1948, 1949; Juhren et al. 1956; Tevis 1958a, b; Beatley 1967, 1969, 1974; Johnson et al. 1978). These studies from various localities in the Mojave Desert have reached similar conclu- sions as to environmental requirements for suc- cessful germination and community-wide flowering phenology. Germination of Mojave Desert winter annuals may be a straightforward consequence to particular environmental conditions with regards to the quan- tity and timing of precipitation. In general, it is thought that a minimum of 20—25 mm of rain is required for successful germination of winter an- nuals (Went 1948; Juhren et al. 1956; Tevis 1958a; Beatley 1967, 1969, 1974). Seasonal timing of pre- cipitation may also be a critical factor, as Beatley (1974) concluded that successful germination of winter annuals in the Mojave Desert is contingent upon a “‘critical autumn rain.” Another common finding among previous studies is the observation that different plant species flower in a temporal sequence during the spring growing season. Beatley (1974), who extensively studied plants at the Nevada Test Site in the eastern Mojave Desert (Fig. 1), defined this phenomenon as fol- lows: ‘“‘flowering and fruiting are phenomena of April and May, with the precise time of anthesis, fruiting, and death varying among the species, and with the elevation and the season, but in any case proceeding in an orderly overlapping species se- quence once in progress.”’ Tevis (1958b) also noted sequence-like flowering times for plants growing at a Colorado Desert site. The “Colorado Desert’’ rep- resents a subdivision of the Sonoran Desert biogeo- graphic province located in southeastern California’ and is bounded by the Mojave Desert, Little San Bernardino Mountains, Peninsula Ranges, and Col- orado River (Turner and Brown, 1982; Fig. 1). At another Colorado Desert site, Burk (1982) also doc- umented interspecific variation in flowering phe- nology among 18 ephemeral species within and be- tween the spring and summer growing seasons. Yet another example comes from the southern Atacama Desert of southern Chile where Vidiella et al. (1999) described the sequential flowering of 25 an- nual and perennial species during the spring season. Thus although deserts are defined by their scant and unpredictable precipitation, the within-season tim- ing of flowering may be a more predictable occur- FENCE. In early 1991 and again in 1992 an El Nifo/ Southern Oscillation (ENSO) event produced large amounts of precipitation across the California de- serts. This afforded the author an opportunity to study flowering phenologies of plants found in the western Mojave Desert. Here, I describe the flow- ering phenologies for the majority of annual and perennial species coexisting on a single site during spring 1991 and 1992. I also show that the temporal sequence of flowering within a single growing sea- son is predictable. MATERIALS AND METHODS This study was conducted at the Desert Tortoise Natural Area (DTNA) in eastern Kern County, Cal- ifornia. This 100 km? nature preserve, which is lo- cated along the western edge of the Mojave Desert (Fig. 1), contains a variety of vegetation commu- nities including creosote bush scrub, Joshua tree 120° Desert Tortoise Natural Area 30° Mojave Desert Sonoran Desert 1202 JENNINGS: PHENOLOGY OF WESTERN MOJAVE DESERT PLANTS 163 NV Nevada Test Site } Me ete ee s vi + * ee ey = ?. antes 2 pee Fy ae oe 30° Fic. 1. Map showing the Mojave and Sonoran Deserts in the southwestern United States and adjacent Mexico (after MacMahon, 1988). The locations of the Desert Tortoise Natural Area (Kern County, California) and Nevada Test Site (Nye County, Nevada) are also shown. woodland, saltbush scrub, and sandy wash. The flo- ra of the DTNA is rich by Mojave Desert standards, as it is comprised of at least 126 annual and 57 perennial species. Flowering phenologies of annuals and perennials were studied in the northwestern corner of the DTNA (elevation ~900 m) between 1 March—12 June 1991 and 1 March—21 June 1992. Local to- pography consisted of flat sandy areas interspersed by sandy washes and low rolling, rocky, hills. 164 MADRONO [Vol. 48. a Average (1937-1999) (A) Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 50 1990 (B) 25 fo) Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 125 1991 100 NJ Ol (C) Monthly Precipitation (mm) NO O17 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 1992 125 100 (D) 75 50 Za 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Month Fic. 2. Monthly precipitation in the western Mojave Desert measured at Randsburg, California 1937-1999: (A) average monthly precipitation for years 1937-1999, (B) total monthly precipitation during 1990, (C) total monthly precipitation during 1991, (D) total monthly precipitation during 1992 (data from NCDC 2000). {\ \ i | \ | | ] 2001] JENNINGS: PHENOLOGY OF WESTERN MOJAVE DESERT PLANTS 165 SPECIES FAMILY MARCH | APRIL MAY JUNE “Tropidocarpum gracile Brassicaceae @ | Erodium cicutarium Geraniaceae @ - Gilia minor Polemoniaceae @ -_Phacelia fremontii Hydrophyllaceae eS Lasthenia californica Asteraceae @ _ Lepidium flavum Brassicaceae e Guillenia lasiophylla Brassicaceae @ Amsinckia tessellata Boraginaceae @ Linanthus dichotomus Polemoniaceae @ Pectocarya spp. Phacelia tanacetifolia Coreopsis bigelovii Cryptantha angustifolia Pholistoma membranaceum Cryptantha nevadensis Eschsholzia minutiflora Menizelia eremophila Caulanthus inflatus Oxytheca perfoliata Chorizanthe watsonii Chorizanthe brevicornu Mentzelia spp. Gilia latiflora Syntrichopappus fremontii Caulanthus cooperi Camissonia palmeri Cryptantha pterocarya Lupinus odoratus Descurainia pinnata Camissonia campestris Eriogonum pusillum Salvia columbariae Plantago ovata Salvia carduacea Chaenactis fremontii Calycoseris parryi Cryptantha circumcissa Malacothrix glabrata Nama demissum Lotus humistratus Centrostegia thurberi Malacothrix coulteri Eriophyllum pringlei Eriogonum gracillimum Astragalus didymocarpus Linanthus parryae Glyptopleura marginata Chaenactis carphoclinia Loeseliastrum schottii Eriogonum nidularium Nemacladus spp. Langloisia setosissima ssp. punctata Camissonia boothii Chorizanthe rigida Prenanthella exigua Eriastrum eremicum Eriogonum angulosum Eremocarpus setigerus Fic. 3. Dates of first flowering for 58+ species of annual plants at the Desert Tortoise Natural Area, eastern Kern County, California during spring 1991. Boraginaceae Hydrophyllaceae Asteraceae Boraginaceae Hydrophyllaceae Boraginaceae Papaveraceae Loasaceae Brassicaceae Polygonaceae Polygonaceae Polygonaceae Loasaceae Polemoniaceae Asteraceae Brassicaceae Onagraceae Boraginaceae Fabaceae Brassicaceae Onagraceae Polygonaceae Lamiaceae Plantaginaceae Lamiaceae Asteraceae Asteraceae Boraginaceae Asteraceae Hydrophyllaceae Fabaceae Polygonaceae Asteraceae Asteraceae Polygonaceae Fabaceae Polemoniaceae Asteraceae Asteraceae Polemoniaceae Polygonaceae Campanulaceae Polemoniaceae Onagraceae Polygonaceae Asteraceae Polemoniaceae Polygonaceae Euphorbiaceae 166 SPECIES Lepidium lasiocarpum Tropidocarpum gracile Pholistoma membranaceum Pectocarya spp. Erodium cicutarium Gilia minor Cryptantha nevadensis Guillenia lasiophylla Cryptantha pterocarya Mentzelia spp. Amsinckia tessellata Eschscholzia minutiflora Lepidium flavum Lasthenia californica Phacelia tanacetifolia Phacelia fremontii Caulanthus inflatus Caulanthus cooperi Linanthus dichotomus Uropappus lindleyi Astragalus acutirostris Mentzelia eremophila Malacothrix coulteri Descurainia pinnata Syntrichopappus fremontii Cryptantha circumscissa Camissonia campestris Coreopsis bigelovii Camissonia palmeri Gilia latiflora Linanthus parryae Lupinus odoratus Monoptilon bellioides Centrostegia thurberi Eriogonum pusillum Astragalus didymocarpus Oxytheca perfoliata Cryptantha angustifolia Eriophylium pringlei Malacothrix glabrata Salvia carduacea Chaenactis fremontii Nama demissum Nemacladus spp. Calycoseris parryi Plantago ovata Chorizanthe watsonii Lotus humistratus Glyptopleura marginata Mucronea perfoliata Loeseliastrum schottii Chaenactis carphoclinia Salvia columbariae Chorizanthe brevicornu Eriogonum nidularium Eriogonum gracillimum Prenanthella exigua Camissonia boothii Chorizanthe rigida Eriastrum eremicum Langloisia setosissima ssp. punctata Eriogonum angulosum Eremocarpus setigerus MADRONO FAMILY Brassicaceae Brassicaceae Hydrophyllaceae Boraginaceae Geraniaceae Polemoniaceae Boraginaceae Brassicaceae Boraginaceae Loasaceae Boraginaceae Papaveraceae Brassicaceae Asteraceae Hydrophyllaceae Hydrophyllaceae Brassicaceae Brassicaceae Polemoniaceae Asteraceae Fabaceae Loasaceae Asteraceae Brassicaceae Asteraceae Boraginaceae Onagraceae Asteraceae Onagraceae Polemoniaceae Polemoniaceae Fabaceae Asteraceae Polygonaceae Polygonaceae Fabaceae Polygonaceae Boraginaceae Asteraceae Asteraceae Lamiaceae Asteraceae Hydrophyllaceae Campanulaceae Asteraceae Plantaginaceae Polygonaceae Fabaceae Asteraceae Polygonaceae Polemoniaceae Asteraceae Lamiaceae Polygonaceae Polygonaceae Polygonaceae Asteraceae Onagraceae Polygonaceae Polemoniaceae Polemoniaceae Polygonaceae Euphorbiaceae MARCH | APRIL MAY | 2001] JENNINGS: PHENOLOGY OF WESTERN MOJAVE DESERT PLANTS 167 SPECIES GROWTH FORM FAMILY MARC APRIL MAY JUNE | Astragalus layneae herbaceous perennial Fabaceae ee | Xylorhiza tortifolia woody shrub Asteraceae Dichelostemma capitatum herbaceous perennial Liliaceae Delphinium parishit herbaceous perennial Ranunculaceae Salazaria mexicana woody shrub Lamiaceae Psorothamnus fremontii woody shrub Fabaceae herbaceous perennial woody shrub Mirabilis bigelovit Larrea tridentata Ericameria cooperi woody shrub Asteraceae Acamptopappus sphaerocephalus — woody shrub Asteraceae Stephanomeria parry! herbaceous perennial Asteraceae Eriogonum fasiculatum woody shrub Polygonaceae Stephanomeria pauciflora herbaceous perennial Asteraceae Eriogonum inflatum herbaceous perennial Polygonaceae Chamaesyce albomarginata herbaceous perennial Euphorbiaceae Fic. 5. Nyctaginaceae Zygophyllaceae Flowering phenologies for 15 species of perennial plants at the Desert Tortoise Natural Area, eastern Kern County, California during spring 1992. Horizontal bars illustrate the flowering phenology of each species as follows: shaded bar to left of white bar = “‘first flowering” stage, white bar = “peak flowering”’ stage; and shaded bar to right of white bar = “‘past-peak flowering”’ stage. Vertical dashed line shows the final day of observations, June 21, 1992. Dominant species of shrubs in this area were Lar- rea tridentata (DC.) Cov., Ambrosia dumosa (A. Gray) Payne, and Acamptopappus sphaerocephalus A. Gray. Daily and monthly precipitation data were obtained from climatic records of Randsburg, Cal- ifornia (22 airline km away), a NOAA station since 1937 (National Climatic Data Center [NCDC], 2000). Observations were generally made several times per week throughout the study period. However, as this study was in a pilot stage during 1991, phe- nological data for that year only consists of dates of first flowering for each species, whereas the 1992 data set contains additional phenological data (see below) for each species. Field identification of spe- cies of Nemacladus, Pectocarya, and Mentzelia (except for M. eremophila (Jepson) H. J. Thompson & Joyce Roberts) proved difficult therefore I only considered these taxa at the level of genus and fam- ily in all analyses. Nomenclature follows Hickman (1993). Flowering phenology of each species was subjectively partitioned into four stages: “‘first flowering’: minority of individuals in flower; “peak flowering’: most individuals in flower; ““past-peak flowering”’’: majority of individuals still succulent but few still in flower; and “‘dried’’: ma- jority of individuals in dried state or, in the case of shrubs, are completely in seed. I analyzed seasonal variation in flowering dates by comparing the 1991 and 1992 first flowering dates for each annual species. Concordance of rel- ative flowering dates would suggest that species’ flowering times are predictable. — RESULTS Following 1990, a period of well-below normal precipitation, winter storms across the western Mo- jave Desert produced well-above average precipi- tation during February and March 1991 and again during the same period in 1992 (Fig. 2A—D; NCDC, 2000). Not only were climatic conditions nearly duplicated both years, but the diversity of plants that underwent flowering was also quite sim- ilar. Annuals in 1991 flowered between early March and middle June (Fig. 3). Observations ended in middle June 1991 so it is unknown if any plants bloomed over summer. However, it is doubtful that many could have flowered because few annuals were even succulent in June. An examination of first flowering dates for 58+ annual species in 1991 reveals substantial temporal variation in flowering with some species beginning to flower in early March, while others didn’t start blooming until April, May, or even June (Fig. 3). Complete flowering-phenology data were ob- tained for 63+ annual and 15 perennial species dur- ing spring 1992. Like 1991, much seasonal varia- tion in flowering was apparent, as annuals and pe- rennials apparently flowered in a sequential manner throughout spring (Fig. 4, 5). It remains unknown if any plants bloomed over summer 1992 but, again, few annuals were even green by June so it is doubtful the flowering season extended much further. Most annuals remained in their peak flow- ering stage for only I—2 weeks but some main- Fic. 4. Flowering phenologies for 63+ species of annual plants at the Desert Tortoise Natural Area, eastern Kern County, California during spring 1992. Horizontal bars illustrate the flowering phenology of each species as follows: shaded bar to left of white bar = ‘“‘first flowering” stage, white bar = “‘peak flowering” stage; and shaded bar to right of white bar = “‘past-peak flowering” stage. Vertical dashed lines shows the final day of observations, June 21, 1992. 168 MADRONO tained this stage for up to a month (Fig. 4). Follow- ing the peak flowering stage very few annuals re- mained in flower beyond a month (Fig. 4). While some annuals began flowering in early March, no perennials bloomed until April (Fig. 5). However, like the annuals the perennials displayed consider- able temporal variation in flowering (Fig. 5). A comparison of first flowering dates for over 58 annual species between 1991 and 1992 indicates that between year flowering dates were remarkably similar (Fig. 6). This suggests that the observed temporal variation in flowering is mostly sequential and predictable. Furthermore, this sequence in flowering times occurred irrespective of growth form, as species of annuals, herbaceous perennials, and woody shrubs flowered throughout spring (Fig. 3-5). An examination of the phenological sequence from the level of plant family suggests a possible phylogenetic connection, as species within the fam- ilies Brassicaceae and Boraginaceae bloomed early in the growing season while the Fabaceae, Astera- ceae, and Polygonaceae flowered relatively late (Fig. 7). However, not all families displayed sea- sonal specificity in flowering, as members of the Polemoniaceae flowered throughout spring (Fig. 7). DISCUSSION The variable nature of desert rainfall was evident in the western Mojave Desert during the early 1990s. Little rain fell in 1990 while relatively large amounts fell in early 1991 and again in 1992. These substantial rainfall events promoted the flowering of nearly every plant species known from the study site each year. The duration of the flowering season as well as the general flowering sequence of species was also strikingly similar both years. Beatley (1967, 1969, 1974) hypothesized that successful germination of Mojave Desert winter an- nuals is contingent upon a >25 mm rainfall event occurring sometime between late September and mid-December. Yet apparently most winter annuals at the DTNA flowered throughout spring 1991 even though almost no rain fell in the area the previous autumn and early winter period. Therefore, germi- nation of DTNA annuals must have occurred in re- sponse to ENSO-driven rains falling in early 1991. The first winter storm of 1991 passed through the area between January 3—5 leaving an accumulation of 19 mm of precipitation (Table 1). While some plants may have flowered in response to this minor rainfall event (see Went 1948; Beatley 1974), it seems much more likely that the storm(s) of Feb- ruary 28 through March 5 caused the massive flow- ering of plants, as over 67 mm of rain fell during this brief period (Table 1). Another series of storms passed through the area between March 19—28 re- sulting in a total of 73 mm of precipitation (Table by Climatic and flowering phenomena during early | [Vol. 48° 1992 were extraordinarily similar to the same pe- riod the year before. The autumn and early winter’ period of 1991 was very dry with only 22.5 mm of| precipitation from five different rainfall events (Ta- ble 1). Flowering in early 1992 must have been triggered by the 47.1-mm rainfall event of late De- cember to early January or the 134.5-mm event in early February (Table 1). Two additional storms passed through the area in early and late March’ leaving accumulations of 23.7 and 64 mm of rain respectively (Table 1). These results are noteworthy because Beatley (1967, p. 746) noted that native winter annuals at, the Nevada Test Site never germinated during the | months January—September regardless of rainfall during this period. Why then did native winter an-. nuals in the western Mojave Desert respond to late winter rainfall and in such dramatic fashion? The | annual plant floras in the eastern and western | regions of the Mojave Desert are roughly similar, especially at the level of family (Beatley 1967, | 1974). This raises the possibility that winter annu-_ als in the eastern Mojave Desert have different ger- mination requirements than their relatives in the western Mojave Desert. | Indeed, geographic variation in climate may help | explain this paradox. The eastern Mojave Desert. experiences substantial rainfall events during sum-. mer and autumn (Beatley 1974; Turner 1982). Win- | ter annuals in this region must have restrictive ger-_ mination requirements otherwise a heavy downpour | in the middle of summer could trigger germination, which might be maladaptive for these C,; annuals (Mulroy and Rundel 1977). On the other hand, win- | ter annuals in the western Mojave Desert are not. faced with this problem owing to the lack of sum-. mer rainfall (Turner 1982) so more flexible germi- nation requirements may be a more advantageous strategy. Cool and moist conditions during the late winter-spring period of 1991 and 1992 in the west- ern Mojave Desert seemed to be ideal for winter annual growth and reproduction. Perhaps these plants would have ‘“‘missed out” had they not re- sponded to late winter precipitation. Thus, geo-— graphical differences in climate may select for re- gion-specific germination strategies but more re-— search is needed to precisely determine what en-— vironmental variables other than rainfall influence germination in these plants. Sequential flowering of desert plants seems to be > a general phenomenon and involving species of various growth forms and phylogenetic lineages (Beatley 1974; Tevis 1958b; Vidiella et al. 1999; this study; also see Poole and Rathcke 1979; Stiles 1979; Cole 1981; Rathcke 1984). While Beatley (1974) suggested that species’ flowering times are predictable, to my knowledge there has been no multi-year comparison of flowering dates for any | Mojave Desert community to corroborate this hy- pothesis. In the present study, I have demonstrated that first flowering dates for an winter annual com- | \ SPECIES _ Tropidocarpum gracile | Erodium cicutarium Gilia minor _ Phacelia fremontii _ Lasthenia californica _ Lepidium flavum — Guillenia lasiophylla _ Amsinckia tessellata Linanthus dichotomus Pectocarya spp. Phacelia tanacetifolia Coreopsis bigelovii Cryptantha angustifolia Pholistoma membranaceum Cryptantha nevadensis Eschsholzia minutiflora Mentzelia eremophila Caulanthus inflatus Oxytheca perfoliata Chorizanthe watsonii Chorizanthe brevicornu Mentzelia spp. Gilia latiflora Syntrichopappus fremontii Caulanthus cooperi Camissonia palmeri Cryptantha pterocarya Lupinus odoratus Descurainia pinnata Camissonia campestris Eriogonum pusillum Salvia columbariae Plantago ovata Salvia carduacea Chaenactis fremontii Calycoseris parryi Cryptantha circumcissa Malacothrix glabrata Nama demissum Lotus humistratus Centrostegia thurberi Malacothrix coulteri Eriophyllum pringlei Eriogonum gracillimum Astragalus didymocarpus Linanthus parryae Glyptopleura marginata Chaenactis carphoclinia Loeseliastrum schottii Eriogonum nidularium Nemacladus spp. Langloisia setosissima ssp. punctata Camissonia boothii Chorizanthe rigida Prenanthella exigua Eriastrum eremicum Eriogonum angulosum Eremocarpus setigerus FAMILY Brassicaceae e Geraniaceae Polemoniaceae Hydrophyllaceae Asteraceae Brassicaceae Brassicaceae Boraginaceae Polemoniaceae Boraginaceae Q Hydrophyllaceae Asteraceae Boraginaceae Hydrophyllaceae ¢) Boraginaceae Papaveraceae Loasaceae Brassicaceae Polygonaceae Polygonaceae Polygonaceae Loasaceae Polemoniaceae Asteraceae Brassicaceae Onagraceae Boraginaceae Fabaceae Brassicaceae Onagraceae Polygonaceae Lamiaceae Plantaginaceae Lamiaceae Asteraceae Asteraceae Boraginaceae Asteraceae Hydrophyllaceae Fabaceae Polygonaceae Asteraceae Asteraceae Polygonaceae Fabaceae Polemoniaceae Asteraceae Asteraceae Polemoniaceae Polygonaceae Campanulaceae Polemoniaceae Onagraceae Polygonaceae Asteraceae Polemoniaceae Polygonaceae Euphorbiaceae e @ poo"s* ©8909, 60° O MARCH Scee O0O00 O 2001] JENNINGS: PHENOLOGY OF WESTERN MOJAVE DESERT PLANTS APRIL MAY ® O O Oo @ © © Oo O O O O O O O @ O @ ‘ @) @ O O O O O O & C0 ,0g@ee0ee O O Ogee oe”? 169 Fic. 6. First flowering dates for 58+ species of annual plants in 1991 (black circles) and 1992 (white circles) at the Desert Tortoise Natural Area, eastern Kern County, California. 170 MADRONO 3 Brassicaceae ) 2 : i q Boraginaceae 0) ® 2 : Oo Polemoniaceae @ 0 2. ” 2 5 | i @ Fabaceae he 0 ® 2 4 e Asteraceae = ) 4 Polygonaceae 0 12 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 March April May June Study Period (weeks) Fic. 7. Distribution of peak flowering dates among annuals and perennials within six selected families throughout spring 1992. Horizontal axis represents the particular week when each species entered its “‘peak-flowering”’ stage. Vertical axis is the number of species within a given family. Only families with at least six species found at the study — site were included. TABLE 1. RAINFALL EVENTS AT RANDSBURG, CALIFORNIA BETWEEN SEPTEMBER 1990 AND JUNE 1992. RAINFALL ON CONSECUTIVE DAYS OR WITHIN A 3-DAY PERIOD IS CONSID- ERED A SINGLE RAINFALL EVENT (DATA FROM NCDC 2000). Rainfall (mm) September 20, 1990 Pipe) November 19, 1990 5.3 December 20, 1990 0.3 January 3—5, 199] 19.0 January 10, 1991 4.5 February 17, 1991 1.8 February 28—March 5, 1991 67.0 March 14, 199] 1.0 March 19-28, 1991 73 May 21, 199] 1.0 September 5, 1991 6.0 September 22, 1991 3.8 October 26, 1991 5.3 November 14, 1991 0.5 December 8-11, 1991 7.0 December 28—January 8, 1992 47.1 February 6—16, 1992 P3425 March 1-7, 1992 23.8 March 21—April 2, 1992 64.8 May 7, 1992 5.8 [Vol. 48. munity were remarkably consistent between years thereby supporting Beatley’s observations. As to the evolution of species’ specific flowering times, this could be driven either by ecological in- teractions (Mosquin 1971; Stiles 1977; Waser 1978; Gleeson 1981; Rathcke and Lacey 1985) or by phy- logenetic relatedness (Vidiella et al. 1999; this _ study). Previous investigations have largely taken the ecological approach such as looking at the in- fluence of limited pollinator availability (Mosquin 1971; Stiles 1977; Waser 1978; Gleeson 1981; Rathcke and Lacey 1985), while few studies have taken phylogeny into account (but see Kochmer and Handel 1986; Vidiella et al. 1999). This study provides weak evidence that the timing of flowering for some plants may be due to phylogeny rather than solely to ecological factors. Additional re- search should be undertaken in an effort to eluci- date the relative roles of ecology and phylogeny in arranging the temporal aspect of plant community structure. ACKNOWLEDGMENTS I am grateful to my former botany professor Dr. J. Rob- ert Haller for formally introducing me to the California flora. I would also like to thank Dr. Kristin Berry and Dr. Matt Brooks for assistance and encouragement throughout 2001] ‘this project, and the two anonymous reviewers for their very helpful comments on the manuscript. Funding was ‘provided by the Bureau of Land Management under con- tract No. B950-C2-0014. | LITERATURE CITED BEATLEY, J. C. 1967. Survival of winter annuals in the | northern Mojave Desert. Ecology 48:745—750. | . 1969. Biomass of desert winter annual plant pop- ulations in southern Nevada. Oikos 20:261—273. . 1974. Phenological events and their environmen- tal triggers in Mojave Desert ecosystems. Ecology 55: 856-863. Burk, J. H. 1982. Phenology, germination, and survival of desert ephemerals in Deep Canyon, Riverside County, California. Madrono 29:154—163. Cote, B. J. 1981. Overlap, regularity, and flowering phe- nologies. American Naturalist 117:993—997. GLEESON, S. K. 1981. Character displacement in flowering phenologies. Oecologia 51:294—295. HICKMAN, J. C. 1993. The Jepson manual: higher plants of California. University of California Press, Berke- ley. JOHNSON, H. B., E C. VASEK, AND T. YONKERS. 1978. Re- sidual effects of summer irrigation on Mojave Desert annuals. Bulletin Southern California Academy of Science 77:95—108. JUHREN, M., E W. WENT, AND E. PHILLIPS. 1956. Ecology of desert plants. IV. Combined field and laboratory work on germination of annuals in the Joshua Tree National Monument, California. Ecology 37:318— 330. KOcHMER, J. P. AND S. N. HANDEL. 1986. Constraints and competition in the evolution of flowering phenology. Ecological Monographs 56:303-—325. MaAcManuon, J. A. 1988. Warm deserts. Pp. 231—264 in M. G. Barbour and W. D. Billings (eds.), North American terrestrial vegetation, Ist ed. Cambridge University Press, NY. Mosauin, T. 1971. Competition for pollinators as a stim- ulus for the evolution of flowering time. Oikos 22: 398-402. Mutroy, T. W. AND P. W. RUNDEL. 1977. Annual plants: adaptations to desert environments. Bioscience 27: 109-114. NATIONAL CLIMATIC DATA CENTER. 2000. Monthly precip- itation data (Randsburg Station, California 1937-99). JENNINGS: PHENOLOGY OF WESTERN MOJAVE DESERT PLANTS ri NOAA, U.S. Department of Commerce, Asheville, NC. POOLE, R. W. AND B. RATHCKE. 1979. Regularity, random- ness, and aggregation in flowering phenologies. Sci- ence 203:470—471. RATHCKE, B. 1984. Patterns of flowering phenologies: test- ability and causal inference using a random model. Pp. 383-393 in D. R. Strong, D. Simberloff, L. G. Abele, and A. B. Thistle (eds.), Ecological commu- nities: conceptual issues and the evidence. Princeton University Press, Princeton. AND E. P. LAcEy. 1985. Phenological patterns of terrestrial plants. Annual Review of Ecology and Sys- tematics 16:179—214. SHREVE, FE AND I. L. WIGGINS. 1964. Vegetation and flora of the Sonoran Desert. Stanford University Press, Stanford. STILES, EF G. 1977. Coadapted competitors: the flowering seasons of hummingbird-pollinated plants in a tropi- cal forest. Science 198:1177—1178. . 1979. Response to Poole and Rathcke. Science 203:471. Tevis, L, Jr. 1958a. Germination and growth of ephemerals induced by sprinkling a sandy desert. Ecology 39: 681-688. . 1958b. A population of desert ephemerals ger- minated by less than one inch of rain. Ecology 39: 688-695. TURNER, R. M. 1982. 153.1 Mohave desertscrub. Pp. 157— 168 in D. E. Brown (ed.), Biotic communities of the American southwest—United States and Mexico. University of Arizona Press, Tucson. AND D. E. BRown. 1982. 154.1 Sonoran deserts- crub. Pp. 181—221 in D. E. Brown (ed.), Biotic com- munities of the American southwest—United States and Mexico. University of Arizona Press, Tucson. VIDIELLA, P. E., J. J. ARMESTO, AND J. R. GUTIERREZ. 1999, Vegetation changes and sequential flowering after rain in the southern Atacama Desert. Journal of Arid Environments 43:449—458. WASER, N. M. 1978. Competition for hummingbird pol- lination and sequential flowering in two Colorado wildflowers. Ecology 59:934—944. WENT, E W. 1948. Ecology of desert plants. I. Observa- tions on germination in the Joshua Tree National Monument, California. Ecology 29:242—253. . 1949. Ecology of desert plants. II. The effect of rain and temperature on germination and growth. Ecology 30:1—13. MApDRONO, Vol. 48, No. 3, pp. 172-176, 2001 SURVEY OF JUNIPERUS COMMUNIS (CUPRESSACEAE) L. VARIETIES FROM THE WESTERN UNITED STATES USING RAPD FINGERPRINTS VANESSA E. T. M. ASHWORTH!, BART C. O’BRIEN AND ELIZABETH A. FRIAR? Rancho Santa Ana Botanic Garden, 1500 North College Avenue, Claremont, CA 91711-3157 ABSTRACT RAPD fingerprints were generated from seven wild populations of Juniperus communis L. to assess whether molecular data support subdivision into vars. saxatilis, jackii and sibirica, members of California Floristic Province, and depressa, a component of the Great Basin Floristic Province. Results from UPGMA and Neighbor Joining cluster analyses showed little correspondence between RAPD-derived distances and varietal boundaries. Juniperus communis var. jackii, in particular, was highly heterogeneous, lending support to the hypothesis that the characteristic growth habit of this serpentine dweller (elongated, sparsely branched lateral branches) is environmentally induced. In contrast to the RAPD results, nucleotide se- quences of the ITS1 region of nuclear ribosomal DNA were identical in four of five var. jackii individuals sequenced, and the fifth exhibited three base substitutions. Juniperus communis L. is a circumboreal species of juniper (Franco 1962) characterized by acicular leaves. Two varieties of J. communis (Cronquist et al. 1972; Adams 1993) are encountered in the west- ern United States. Juniperus communis var. depres- sa Pursh is native to the Great Basin Floristic Prov- ince, extends northward into Alaska and eastward across much of Canada and the Great Lakes region, arching south along the east coast to North Caro- lina. Juniperus communis var. saxatilis Pallas oc- curs from British Columbia southward into Cali- fornia in the Cascade Ranges, North Coast Ranges, and Sierra Nevada, but also has a circumboreal dis- tribution outside North America (Adams 1993). The two varieties differ primarily in habit, leaf size and shape and width of the glaucous stomatal band on the adaxial leaf surface. Although both are low-growing, variety depressa develops a some- what erect main stem whereas variety saxatilis is entirely prostrate. Leaf dimensions are ca. 1.0—1.6 mm broad X (6) 10—18 mm long (depressa) and (1.2) 1.5—-1.8 mm broad X 5—10 (12) mm long (sax- atilis) (Cronquist et al. 1972), and the glaucous sto- matal band is as broad as, or narrower, than each green margin (depressa) or 2—3 times as broad as each green margin (saxatilis; Franco 1962). In California, two other varieties are occasionally distinguished. Juniperus communis var. jackii Reh- der (Rehder 1940) differs from var. saxatilis by having longer, more sparsely branched lateral branches and is encountered on serpentinite sub- strates of inland coastal areas in northern California and Oregon. Juniperus communis var. sibirica Rydb. is described as a very prostrate, almost mat- like, form found on coastal bluffs and in the ex- ' Present address: Department of Botany and Plant Sci- ences, University of California, Riverside, CA 92521. * Author for communication. treme northwest of California and southwestern Or- | egon, and as a disjunct population at Ebbett’s Pass | in the Sierra Nevada. According to Roof (1973), | this variety is characterized by leaves that are more | incurved, making it less prickly to the touch than — J. communis vars. jackii or saxatilis. Adams (1993) | and Cronquist et al. (1972) placed varieties jackii | and sibirica in synonymy under J. communis var. montana, a name recently placed in synonymy un- | der var. saxatilis (Farjon 1998). Our previous paper > (Ashworth et al. 1999) used the older varietal epi- | thet. | The purpose of this study was to make a prelim- — inary assessment of genetic variability among the > four varieties of J. communis in the Western United States and to examine whether molecular data fa- vors one of the taxonomic schemes over another. Specifically, do the data support a subdivision into vars. depressa and saxatilis, and/or is there evi- dence supporting the recognition of varieties sibir- ica and jackii? A second goal was to ascertain whether the mats formed by these creeping junipers are genetically uniform (i.e., clonal) or harbor dis- tinct genotypes. RAPD analysis was chosen as a quick and relatively inexpensive means of getting a fingerprint of the genome of plants from each of the native populations. This technique has been ap- plied successfully to interspecific studies in Junip- erus (Adams and Demeke 1993). Additionally, se- quences of the ITSI spacer region of the nuclear ribosomal DNA were generated for a subset of sev- en samples. | METHODS Plant material. Plant material was gathered from seven wild Juniperus communis populations repre- senting vars. saxatilis (saA—saC, saG) and depressa (deD—deF). Under the alternative taxonomic scheme, populations saA and saG correspond to J. i | | i | | 2001] ASHWORTH ET AL.: SURVEY OF JUNIPERUS COMMUNIS VARIETIES ls | |TABLE 1. COLLECTION LOCALITIES AND ALTERNATIVE VARIETAL DELIMITATIONS, WITH JUNIPERUS COMMUNIS SEGREGATED ‘INTO Two VARIETIES (DEPRESSA AND SAXATILIS, AS IN FLORA OF NORTH AMERICA COMMITTEE (1993) AND CRONQUIST ET AL. | (1972); A) OR FOUR VARIETIES (DEPRESSA AND SAXATILIS, AS WELL AS JACKII SENSU REHDER AND “‘SIBIRICA”” SENSU RYDBERG; Collection locality OR, Curry County: Cape Sebastian CA, Del Norte County: Gasquet Toll Road; two sites ca. | mile apart CA, Humboldt County: Onion Mountain/Onion Lake intersection CA, Alpine County: Ebbett’s Pass, Sierra Nevada OR, Hood River County, Mt. Hood OR, Curry County, ‘Gold Beach’ CA, Del Norte County, ‘Point St. George’ UT, Iron County: between Cedar Breaks National Monument and Pan UT, Iron County: Cedar Breaks National Monument iB). Varietal Varietal _ Sample delimitation delimitation | designation A B _saAl—saA7 saxatilis sibirica ' saBl—saB6 saxatilis Jackii | saCl—saC2 saxatilis jackii ~saG1l—saG3 saxatilis sibirica -CVv2 saxatilis saxatilis CV5 saxatilis sibirica CV 11 saxatilis sibirica deD depressa depressa guitch deE depressa depressa deF depressa depressa NV, White Pine County: Wheeler Mtn., Great Basin National Park communis var. sibirica and populations saB and saC to J. communis var. jackii. Table 1 summarizes collection details and taxonomic designations of each of the native populations (see Ashworth et al. 1999 for more complete information), as well as for three cultivated accessions originating from Mount Hood, OR (CV2), Gold Beach, OR (CV5), and Point St. George, CA (CV11), that were in- cluded in this study. These three plants grow at Rancho Santa Ana Botanic Garden but were estab- lished from cuttings collected in the wild. In a pre- vious study that included both native and non-na- tive Juniperus species (Ashworth et al. 1999) they clustered with the native J. communis varieties. CV2 represents J. communis var. saxatilis under all taxonomic systems presented here. CV5 and CV11 are var. saxatilis sensu Adams (1993) and Cron- quist et al. (1972) and var. sibirica sensu Rydberg. DNA analysis. Information on DNA extraction method, PCR reaction conditions and RAPD primer sequences are detailed in Ashworth et al. 1999. Bands were scored as present or absent by the first and last author. Average taxonomic distances gen- erated from these binary scores were analyzed us- ing the clustering algorithm UPGMA (Unweighted pair group method with arithmetic averages; Sneath and Sokal 1973) and Neighbor-Joining (NJ; Saitou and Nei 1987) available on PAUP* version 4.0 B1 (Swofford 1998). Effects of alternative measures of distance/similarity on clustering were explored us- ing NTSYS version 2.0 (Rohlf 1993). Jaccard co- efficients of similarity were calculated using the NTSYS ‘SIMQUAL module, and the cophenetic correlation coefficient was determined via the COPH and MXCOMP modules. Sequences of the ITS1 spacer region were gen- erated using the forward primer ITSS (GGAAG- TAAAAGTCGTAACAAGG) and reverse primer ITS4 (TCCTCCGCTTATTGATATGC; both prim- ers from White et al. 1990). Amplification condi- tions consisted of 40 cycles, each with three suc- cessive phases of (1) 97°C for 1 min, (2) 48°C for 1 min, and (3) 72°C for 2 min, followed by a final extension time of 7 min at 72°C. Double-stranded template was purified by precipitation in polyeth- ylene glycol/2.5 M NaCl (Morgan and Soltis 1993; Johnson and Soltis 1995) with a 70% and 95% EtOH wash. Single-stranded DNA template was generated by cycle sequencing with incorporation of dye terminators (PRISM™ Dye Terminator Cycle Sequencing Kit with AmpliTaq®; Perkin-Elmer, CT). Settings were 25 cycles of 0.5 min at 95°C, 0.25 min at 50°C and | min at 60°C. The resulting product was purified by ethanol precipitation (Sam- brook et al. 1989) and electrophoresed on a 6% polyacrylamide gel (Sequagel®) in an Applied Bio- systems Model 373A Automated Sequencer. Se- quences were proofed and assembled using Se- quencher 3.0 (Gene Codes Corporation, Inc., Ann Arbor, MI). RESULTS Of 65 primers screened for RAPD analysis, five primers showing scorable and reproducible banding patterns were entered into the final analysis. Scor- able bands per primer ranged from one (UBC-329) to nine (UBC-244), with a total of 27 bands scored for 24 individuals. Identical banding patterns were found for saAl, saA3 and saA4, with saA2 differ- ing by a single band. Figure 1 shows the UPGMA and NJ phenograms generated from distance matrices derived from the RAPD scores. The UPGMA phenogram reveals six main clusters (#1—6), ranging in average within- cluster distance from 0.065 (cluster 4) to 0.273 (cluster 2). CV2 is the most distant accession. Clus- ters 6 and 5 are linked at a distance of 0.244, with cluster 4 attaching next (0.319), then cluster 3 (0.357), cluster 2 (0.370) and cluster 1 (0.395). Cluster 6 comprises mostly var. sibirica (saA1—4, plus CV5, CV11) but also saB4, cluster 5 includes the remaining two members of population saA 174 MADRONO [Vol. 48) Cluster 6 Cluster 5 Cluster 4 Cluster 3 Cluster 2 Cluster 1 sa Al sa AS Neighbor sa A4 ae sa A2 Joining avis Cluster 4 CV 11 sa A5 sa A6 Cluster 5 Re sa B5 Cluster 1 Cluster 3 Cluster 2 Fic. 1. UPGMA and Neighbor Joining phenograms generated from distances derived from RAPD data of 24 Juniperus communis accessions from the western United States. | 2001) (saA6, saA7) and also saB5. Cluster 4 comprises population saG, the Sierra Nevadan representatives of var. sibirica. Members of population saB (var. jackii) appear in five of six clusters, including clus- ter 2 (saB3) which contains all three accessions of var. depressa. Only clusters 1 and 3 contain exclu- sively var. jackii. UPGMA clustering based on Jaccard coefficients of similarity resulted in identical cluster composi- tion but an altered cluster hierarchy: cluster 3 is the most distant (0.36), followed in order of increasing similarity by CV2, cluster 4, cluster 1, cluster 2, cluster 5 and cluster 6. The matrix correlation co- efficient indicates a good fit of distances derived from the phenogram to the original distance matrix. Five of six clusters present in the UPGMA phen- ograms are also identified by the NJ algorithm. Three main differences emerge from a comparison of the UPGMA and NJ phenograms: (1) saB4 re- sides in cluster 6 (predominantly var. sibirica) in the UPGMA tree but near cluster 3 (var. jackii) in the NJ tree; (2) population saG, which forms a sep- arate cluster below the bifurcation of clusters 5 and 6 in the UPGMA tree, inserts within cluster 6 in the NJ tree; and (3) clusters 1 and 5 are closest to each other in NJ but placed most distantly in the UPGMA analysis. Of the seven ITS | sequences, identical sequenc- es were found for saBl, saB3—5, and saG1. Only CV2 and saB6 each exhibited three autapomorphic base substitutions, and CV2 additionally had an in- sertion of three nucleotides. DISCUSSION AND CONCLUSION Regardless of the clustering algorithm or dis- tance measure used, our RAPD fingerprint data are unable to clarify relationships among the four J. communis varieties depressa, jackii, saxatilis or si- birica. This is a consequence primarily of the mark- edly heterogeneous population saB (saxatilis/jack- li), Which suggests that the jackii morphology (sparsely branched, elongated branches) is an en- vironmentally induced growth form. Our data thus support Adams (1993) and Cronquist et al. (1972) who place the variety in synonymy under var. sax- atilis on the grounds that the jackii habit disappears under common garden conditions (p. 15, Adams (1993)). Kruckeberg (1967) cites J. communis as an example of a substrate-indifferent (“‘bodenvag”’ sensu Unger 1836) serpentine dweller but makes no mention of morphological differences between ser- pentine and non-serpentine plants. It is well docu- mented that the serpentine environment has a major impact on plant growth and adaptation, although the soil substrate is no longer seen as the only fac- tor responsible. Instead, indirect effects, such as greater light availability, also exert a strong selec- tive force (Baskin and Baskin 1988; Gankin and Major 1964). The elongated, sparsely branched habit of var. jackii may thus be the result of reduced ASHWORTH ET AL.: SURVEY OF JUNIPERUS COMMUNIS VARIETIES 7S competition from other vegetation and_ plentiful light. The integrity of var. saxatilis is contradicted by the fact that the saxatilis accessions in this study are never united in a single cluster distinct from var. depressa. The proximity of clusters 4, 5 and 6 in the UPGMA analysis lends some support to var. sibirica, although this is weakened by the presence of saB4 and saBS. In the NJ analysis cluster 5, with its two sibirica representatives saA6 and saA7, is more similar to non-sibirica cluster | than to the other sibirica accessions. Interestingly, the NJ analysis causes population G to cluster with sibirica representatives, consistent with its sibirica-like growth habit and in contrast to its geographic origin (Sierra Nevada). Although geographically close to the Great Basin variety de- pressa, none of the analyses presented here show a close association between var. depressa and popu- lation saG. Nucleotide substitutions and an insertion in the ITS! region were revealed only in saB6 and CV2, corroborating their basal placement on the UPGMA phenogram. By contrast, saGl and four members of the heterogeneous saB population (saBl and saB2—4) exhibited identical sequences, showing a lack of concordance between RAPD-derived dis- tances and ITSI sequence divergence. The absence of support from our RAPD data for a distinction between vars. saxatilis and depressa is Surprising but may be a function of relatively few markers in relation to the number of genotypes studied. A higher marker to genotype ratio and a greater sampling density might clarify some of the variation encountered. Although our data are unable to provide answers to our taxonomic questions, they nonetheless give insight into the genetic composition of juniper mats. Individuals of var. saxatilis population A originated from various positions around the pe- riphery of a large mat. The identical fingerprints of individuals saAl, saA3 and saA4 suggest that this part of the mat is clonal (saA2 differs only by a single band), but individuals saA5—saA7 have dis- tinct fingerprints. This mat is therefore a combina- tion of clonally-spread and seed-derived individu- als. Population B was collected from two nearby mats. This makes the great diversity of distinct fin- gerprints even more surprising and we speculate whether individuals from this population constitute a hybrid swarm. LITERATURE CITED ADAMS, R. P. 1993. Juniperus in Flora North America, Vol. 2., Pteridophytes and Gymnosperms. Oxford University Press, New York. AND T. DEMEKE. 1993. The relationships in Juniperus based on random amplified polymorphic DNAs (RAPDs). Taxon 42:553—571. ASHWORTH, V. E. T. M., B. C. O’ BRIEN, AND E. A. FRIAR. 176 MADRONO 1999. Fingerprinting Juniperus communis L. cultivars using RAPD markers. Madrono 46:134—141. BASKIN, J. M. AND C. C. BASKIN. 1988. Endemism in rock outcrop plant communities of unglaciated eastern United States: an evaluation of the roles of the edaph- ic, genetic and light factors. Journal of Biogeography 15:829-840. CRONQUIST, A., A. H. HOLMGREN, N. H. HOLMGREN, AND J. L. REVEAL. 1972. Intermountain Flora. Vascular plants of the Intermountain West, U.S.A., Vol. 1. Haf- ner Publishing Company, Inc., New York. FARJON, A. 1998. World checklist and bibliography of co- nifers. Kew, London. FRANCO, J. DO AMARAL. 1962. Taxonomy of the common juniper. Boletim da Sociedade Broteriana 36:101— 120. GANKIN, R. AND J. MAsor. 1964. Arctostaphylos myrtifol- ia, its biology and relationship to the problem of en- demism. Ecology 45:792—808. JOHNSON, L. A. AND D. E. SoLtis. 1995. Phylogenetic in- ference in Saxifragaceae sensu stricto and Gilia (Po- lemoniaceae) using matK sequences. Annals of the Missouri Botanical Garden 82:149-175. KRUCKEBERG, A. R. 1967. Ecotypic response to ultramafic soils by some plant species of the northwestern Unit- ed States. Brittonia 19:133-151. MorGan, D. R. AND D. E. SOLTis. 1993. Phylogenetic re- lationships among Saxifragaceae sensu lato based on rbcL sequence data. Annals of the Missouri Botanical Garden 80:631—660. [Vol. 48 REHDER, A. 1940. Manual of cultivated trees and shrubs hardy in North America. The MacMillan Company New York. ROHLF, E J. 1993. NTSYS-pc, numerical taxonomy anal mutltivariate analysis system, version 2.0. Setauket, New York: Exeter Software. Roor, J. B. 1973. Native dwarf junipers. The Four Sea-, sons 4(1):2-9. SAITOU, N. AND M. NEI. 1987. The neighbor-joining meth- od: a new method for reconstructing phylogenetic) trees. Molecular Biology and Evolution 4:406—425. SAMBROOK, J., E. E FRITSCH, AND R. MANIATIS. 1989. Mo-. lecular cloning, a laboratory manual, 2nd edition. Cold Spring Harbor Laboratory Press, Plaviea New York. SNEATH, P. H. A. AND R. R. SOKAL. 1973. Principles of numerical taxonomy. Freeman, San Francisco. SWOFFORD, D. L. 1998. PAUP*. Phylogenetic analysis us- ing parsimony (*and other methods). Version 4.0 61. | Sinauer Associates, Sunderland, MA. UNGER, E 1836. Uber den Einfluss des Bodens auf die! Verteilung der Gewiichse, nachgewiesen in der Veg- | etation des nordéstlichen Tirols. Rohrmann ung Schweigerd, Vienna. Waite, T. J., T. BRUNS, S. LEE, AND J. TAYLor. 1990. Am plification and direct sequncing of fungal ribosomal | RNA genes for phylogenetics. Pp. 315-322 in M. In- | nis, D. Gelfand, J. Sninsky, and T. White (eds.), PCR | Protocols: a Guide to Methods and Applications. Ac | ademic Press, San Diego. CONIFER TREE DISTRIBUTIONS IN SOUTHERN CALIFORNIA RICHARD A. MINNICH AND RICHARD G. EVERETT Department of Earth Sciences, University of California, Riverside, CA 92521] ABSTRACT Maps and text describe current distributions of 19 conifer tree species in southern California. Distri- butions are related to climate, geology, terrain, and other aspects of the environment. Corrections to earlier published descriptions and maps are highlighted. Although species are discussed as components of six vegetation types (closed-cone conifer forest, foothill woodland, mixed evergreen forest, pinyon-juniper woodland, mixed-conifer forest, and subalpine forest), the distribution maps are for species, not vegetation. Species limits correspond to a density as low as | tree of the species of concern per 10 ha. Some changes in distribution over the past century, primarily due to fire, can best be interpreted and understood when a species is placed correctly in one of three models: shifting populations, standing populations, and patch mosaic populations. Because of their value for lumber and fuelwood, the distribution of coniferous trees in southern Cal- ifornia has been described since Europeans first set- tled the region in the late 18" century (Minnich 1988; Minnich and Franco-Vizcaino 1998). Early accounts include descriptions in Spanish diaries, the Pacific Railroad Survey of the 1850s, the US—Mex- ico Boundary survey of 1894, early explorers, and reports of the U.S. Forest Service. The first maps published over 100 years ago by Kinney, Grinnell, and in Forest Reserve Reports were impressionistic or based on life zone theory (Minnich 1988). While forest distributions were realistic, boundary criteria were not explicitly defined. The first objective maps were produced by the Vegetation Type Map Survey (VTM) in 1929—34 (Colwell 1988) on the basis of physiognomy and dominance criteria. Map accu- racy was compromised because boundaries were drawn in the field without the advantage of aerial photographs, and their precision was reduced by planometric error of triangulated 1900 topographic base maps. Still, the VTM maps show high quality and are consistent with the earliest aerial photo- graphs in 1928—1939. Griffin and Critchfield (1976) reproduced the ranges of California trees based largely on maps of the VTM survey. The population size and extent of species of course are dynamic, so it is essential to update maps to ascertain the impacts of natural distur- bances and land management. Using aerial photo- graphs, this study produces maps of coniferous trees in southern California south of lat. 35° and west of the Mojave and Colorado Deserts. In south- ern California’s mediterranean climate, wildland fire is strongly integrated into the ecological func- tion of conifer species. Because conifers are largely nonsprouters, stand-replacement fires have _ pro- duced important changes in species ranges since the VTM survey. The objectives of this study are to produce a modern baseline of distributions for com- parison with the VTM survey, to correct taxonomic errors of the VIM survey, describe species ranges in relation to climate, geology, terrain, and other aspects of the environment, and to evaluate species changes over the 20th century. Among the various reasons for conducting the VTM Survey, Wieslander (1935) stated that it was important to produce a baseline of vegetation for purposes of replication at some time in the future. Likewise, the hope of this survey of southern Cal- ifornia conifer forests is to encourage resurveys to evaluate their broad-scale dynamics over long time scales. Coastal Southern California. Most conifers grow in the Transverse and Peninsular Ranges that divide the Mojave and Sonoran Deserts from the coastal plains and valleys (Fig. 1). The Transverse Ranges extend from Point Conception to Joshua Tree Na- tional Park. The Santa Ynez Mountains join the San Rafael and Sierra Madre Mountains of the South Coast Range of central California. From this junc- tion, a series of parallel ridges that include the Or- tega, Santa Paula, and Topatopa Mountains, and Pine Mountain Ridge merge into a single chain that includes the Liebre, San Gabriel, and San Bernar- dino Mountains, and the plateau of Joshua Tree Na- tional Park (place names are located in Table 1). The Tehachapi Mountains of the Sierra Nevada ter- minate at the Lockwood Valley plateau that con- tains isolated undissected ranges, including Frazier Mountain, Alamo Mountain, and Mt. Pinos. The plateau is bordered on the north by the San Emig- dio Mountains. The north-south Peninsular Ranges include the San Jacinto Mountains that extend from San Gorgonio Pass to the Santa Rosa Mountains. To the west is a series of coastal ranges including the Santa Ana, Palomar, Volcan, Cuyamaca, and Laguna Mountains. Elevations in Transverse Rang- es frequently exceed 2000 m, and reach 3080 m at Mt. San Antonio and 3499 m at Mt. San Gorgonio. The more subdued Peninsular Ranges are typically 1000-1800 m, although summit altitudes reach 3273 m at Mt. San Jacinto. 178 MADRONO [Vol. 48 TABLE |. LOCATION OF PLACE NAMES USED IN TEXT. Place name Latitude (°N) Longitude (°W) Aguanga Shoes| LG 752 Alamo Mountain 34 40 118 56 Azusa 34 09 Ne fies 5) Banner Canyon 33 O5 L16433 Barton Flats 34 10 HIG 753 Bautista Canyon 33 40 116 47 Beechers Bay, Santa Rosa Island 33° (59. 120 10 Big Bear 34 15 Lio 756 Big Pine Mountain 34 42 1195 39 Big Tujunga Canyon 34 20 1iS.205 Bluff Lake 34 13 116 58 Butler Peak 34 16 117 OO Cattle Canyon 34 15 117 41 Cerro Nordeste 34 50 Os City Creek 34 07 1t7. 10 Cobblestone Mountain 34 36 118 52 Coxcomb Mountains 34 02 115 20 Crestline 34 15 117 15 Cucamonga Peak 34 14 i 35 Cuddy Valley 34 50 119-02 Cuyama River 34 45 119 20 Deep Creek 34 18 7a OF: Doone Vallye 5320 116 55 Eagle Mountains 33 49 115 40 Figueroa Mountain 34 45 LOSI, Frazier Mountain 34 47 S59 Garner Valley 337 37 117 40 Gavilon Hills 33 48 117 20 Government Peak 34 09 i705 Grapevine Canyon 34 23 117 04 Guatay 32 OL 116 34 Gypsum Canyon 33 50 117 43 Hagador Canyon 33 48 37 Hesperia 34 24 117 20 Holcomb Creek 34 18 117 00 Holcomb Valley 34 19 Gr 5S Hurricane Ridge 34 46 119 47 Icehouse Canyon 34 15 117 40 Idyllwild 33 45 116 43 Japacha Peak 32 56 116 43 Keller Peak 34 12 i 03 La Brea Creek 34 57 LO Ds Lake Arrowhead 34 15 17 lt Lake Hemet 33 40 116 42 Lake Mathews 33.51 Di ae26 Lake Piru 34 29 118 45 Lakeview Mountains 33 46 117 04 Lazaro Canyon 34 49 119 50 La Rumerosa 32, 30 116 03 Lockwood Valley 34 45 119102 Los Pinos Mountain 32 46 116 36 Lytle Creek 34 14 Lg 229) Manzanita Creek 34 47 119 54 McCain Valley 32 45 11620 Mill Creek 34 05 116 56 Mt. Baden-Powell 34 22 117 46 Mt. Grinnell 34 O07 116 49 Mt. Laguna 32 52 116-25 Mt. San Antonio 34 17 59 Mt. Waterman 34 21 Lies 6 Mt. Wilson 34 13 118 04 Ojai 34 27 LO tS Ontario Peak 34 14 ART) Shs) Onyx Summit 34 11 116 43 Ortega Hill 34 34 tone 23 i } 1 } i) | | Dy i {i i } | 2001] TABLE |. CONTINUED. Place name | Otay Mountain | Palmdale Palm Divide Perris Phelan Pinyon Ridge Pinyon Flat Pine Valley | Pleasants Peak Pleasant View Ridge Plunge Creek Prairie Fork Purisima Hills Rabbit Peak Raywood Flat Redlands Riverside-Perris Plain Rock Creek Running Springs San Bernardino Ridge San Emigdio Mountains San Gorgonio Mountain San Rafael Mountain Santa Paula Peak Santa Susana Mountains Sawmill Mountain Sespe Creek Sisquoc River Soledad Canyon Sugarloaf Mountain Sulphur Spring Canyon Tahquitz Peak Tecate Peak Temescale Canyon Thomas Mountain Throop Peak Tip Top Mountain Topatopa Mountains Torrey Pines State Beach Vandeventer Flat Warm Springs Mountain Waterman Canyon Whale Peak Wright Mountain Wrightwood Zaca Ridge The climate is mediterranean with winter precip- itation and summer drought. Mean winter temper- atures range from 10°C in the coastal and desert areas to O°C at 2500 m in the mountains. Mean temperatures in summer range from 38°C in the desert to 15°C at 2500 m. A shallow marine layer results in mean temperatures of 18—22°C at the coast increasing to 27°C in inland valleys. Large differences in local conifer species zona- tion result from strong gradients in average annual precipitation (AAP, Fig. 2). Most precipitation oc- curs during cold fronts arriving from the North Pa- cific Ocean. Because winds aloft are southwest to southerly (Minnich 1984), physiographic lift is MINNICH AND EVERETT: CONIFER DISTRIBUTIONS vo Latitude (N) Longitude ((W) 32 35 PG 952 34 35 LAss 05 33 42 LG ¥39 33 48 117 14 34 26 117 34 33 Il 116 26 5357 16 27, 32,790) 116: 31 33 47 117 38 34 25 5) 34 09 Lie 209 34 20 117 42 34 44 120: 27 33 26 116 14 34 03 116 50 34 03 get 53500 ANS) S45 25 Lie 750 34 13 117 06 34 08 116 54 34 50 119 OO 34 06 116 50 34 43 119 48 34 26 119 OO 34 20 LES eS) 34 42 118 34 By ah LTO bee 34 51 Eels) 34 27 118 18 34 12 lop aay 34 45 1 dhe ys. 33 45 116 40 32 35 116 43 33 46 117 30 eh) Sirs) 116 42 34 21 117 48 34 15 116 42 34 32 119 OO Boao Lat Bis) s)e) lGwes3 34 36 118 35 34 12 MIS ele7, 33) 02 Ge as 34 20 ys 39) 34 21 DN ETA SNe ee 34 46 120 Ol strongest on the south front of the Transverse Rang- es, with the highest AAP in coastal escarpments of the “upwind”? Topatopa, San Gabriel, and San Ber- nardino Mountains (AAP, 90-110 cm). Amounts decrease toward downwind ranges and water- sheds—regardless of altitude—due to depletion of storm air mass moisture and descending airflow in rain shadows, including the San Rafael Mountains, Pine Mountain Ridge, Big Bear, and upper Santa Ana River (AAP, 60—80 cm). Farther inland, the San Emigdio, Tehachapi, and Liebre Mountains, Mt. Pinos, and lee slopes near Big Bear receive only 35-60 cm AAP. The AAP is only 40—60 cm in the Peninsular Ranges because slopes parallel 180 MADRONO [Vol. 48 | 5 1118Ww . Kain "Srp \ ot j na? Mojave Desert Magy. Mt. Pinos. Ten#° ; ; “ew, Pine Mtn. Ridge © ims, Santa Ynez Mtns: us San Gabr; Santa Barbara Brie Mountains San Bernardino Mountains Santa Cruz Is. Santa Monica Mountains : ee a fel S4N Los Angeles ; Santa Rosa Is, Nez San Jacinto Ans Mountains Gp, >. Mp Santa Rosa Mt. Mtn. Palomar Hot Springs Min. Mt. Cuyamaca vg Buns y : taj San Diego fee == Baja California Mexico cy 40 0 40 Kilometers ee See Fic. |. Location map and place names in southern California. Topographic maps and elevations were developed from United States Geological Survey 1:250,000 Digital Elevation Models (USGS 2000). storm winds. Amounts locally reach 80—100 cm on local southern escarpments of the Santa Ana Moun- tains, Mt. Palomar, and Cuyamaca Peak. The San Jacinto Mountains receive 40—70 cm and the Santa Rosa Mountains receive <50 cm because they are leeward of the Santa Ana and Palomar Mountains. The AAP decreases to 10—20 cm in the Mojave and Sonoran Deserts. METHODS The large size and distinct morphology of coni- fers permits accurate mapping to species level us- ing aerial photographs. Maps were interpreted from aerial photographs using a roll film stereoscope (on file in the Department of Earth Sciences, University of California, Riverside). Trees were identified on the basis of their crown perimeter and apex config- urations, vertical structure, shadows, and color. We field verified for tree occurrence in all southern Cal- ifornia ranges. Procedures are summarized in Min- nich (1987a) where maps of many of the same spe- cies in northern Baja California were presented. Distributions were interpreted from a variety of aerial photograph coverages of the entire region taken mostly between 1971-1980 including the Mission 164 coverages of California and photo- graphs taken in 1938—1939. Site-specific georefer- encing of repeat aerial photographs on a Zoom Transfer Scope (ZTS) reveals that virtually the same mature trees can be matched on the earliest 40 59 30. 20 20 403020 LB 10 SG “70° = 10. SQ 505 Do =” a0 S35 (ONES Sat T09 SN 40 SS SS RY | 40 WS . Wik E 34N By 30 | ayy mI 10) 30 S30 CLEP SENS 30 30 SSN \ ¢ \\\ 60. 10 50~ ABSPSh ea ee ‘> |%9920 30 40 ® 40 40 0 40 80 Kilometers 8 _ Fic. 2. Mean annual precipitation (after California 1980). Pinus muricata 34.N 40 80 Kilometers » ® 4 Fic. 3. botanical collections or field vouchers. and most recent photograph coverages. However, rapid changes occurred in all species from stand- replacement burns. Recent flights (> 1990) were ex- amined for stand-replacement burns to update co- nifer species distributions since the Mission 164 flights. Since the objective is to map broad species rang- es, rather than forest dominance patterns, we de- fined the limit of species as a minimum threshold of one tree per 10 ha. This criterion allows for more detailed range maps than depicted from Griffin and Critchfield’s definition of “‘group of stands more than two miles (3 km) across, or stands less than two miles (3 km) across or of unknown size.’’ It also excludes rare outposts that would result in ex- cessive detail. Botanical collections are included for the occurrence of outposts outside our species rang- eS. Boundary data were georeferenced and trans- ferred onto 1:250,000 scale topographic sheets (USGS 2000) using a ZTS. Contour creation using DTED Level O data (NIMA 2001) and subsequent spatial manipulations were performed by Geo- graphical Information Systems using ESRI (2000) software. Graphic versions of the species maps are accessible from our website (http://spotfire/ucr.edu/ socaltree). Changes in forest distributions can be evaluated by comparing website maps with pub- lished and unpublished VWTM quadrangle maps un- der curation at the University of California, Santa Barbara. DISTRIBUTIONS Conifer tree ranges are presented under six veg- etation types, as recognized in Barbour and Major (1988): closed-cone conifer forest, foothill wood- land, mixed evergreen forest, pinyon-juniper wood- MINNICH AND EVERETT: CONIFER DISTRIBUTIONS 181 118 W Pinus attenuata « The distribution of Pinus attenuata and P. muricata. Shaded areas mapped from aerial photographs. x = land, mixed-conifer forest, and subalpine forest. Species ranges are given from northwest to south- east. Total stand area for each species was obtained from the GIS software. Most conifers grow else- where in California and western North America (Critchfield and Little 1966; Griffin and Critchfield 1976; Hickman 1993) and extend into northern Baja California (Minnich 1987a; Minnich and Fran- co-Vizcaino 1998). Statements on substrate are based on the Southern California Areal Mapping Project (SCAMP 2000). Place names are given in Table 1. Closed-cone conifer forest. Closed-cone conifer forest includes three taxonomic groups: serotinous pines in Pinus subsect. oocarpae (Pinus attenuata Lemmon, P. muricata D. Don; taxonomy and vari- ation reviewed in Millar 1986), the partially serot- inous Pinus sect. sabinianae (big-cone pines, P. coulteri D. Don and P. torreyana Carriere), and two cypresses, Cupressus arizonica E. Greene ssp. ar- izonica of the C. arizonica complex and C. forbesii Jepson (Rehfeldt 1997). Pinus coulteri and P. tor- reyana are treated here due their similar fire ecol- ogy to species in Pinus subsect. oocarpae (Haller 1986) and cypresses (Vogl et al. 1988). However, the big-cone pine P. sabiniana Douglas is treated as a member of foothill woodland. Closed-cone co- nifer forests grow in chaparral with stand-replace- ment fires resulting in spatially synchronized even- aged, short-lived stands. Distributions in southern California are fragmented, with stand boundaries frequently unrelated to local environmental gradi- ents due apparently to stochastic recruitment fail- ures after fire (Vogl et al. 1988). Nearly all closed- cone conifers are found in monospecific stands. Pinus muricata D. Don (937 ha, Fig. 3).—Small, compact stands of bishop pine are found within 20 182 MADRONO 34N 40 0 40 80 Kilometers Cupressus forbesii 1118 W Cupressus arizonica var. arizonica Fic. 4. The distribution of Cupressus arizonica var. arizonica and C. forbseii. Shaded areas mapped from aerial photographs. x = botanical collections or field vouchers. km of the Pacific Coast in the Purisima Hills and the west end of Santa Ynez Mountains. Offshore stands include two colonies on Santa Rosa Island and larger stands on Santa Cruz Island. Mainland stands grow on diatomaceous shales or Careaga sandstones, which produce shallow, highly acid soils with high water-holding capacity (Schoenherr 1992). Island stands grow on a variety of substrates (Vogl et al. 1988). While the AAP ranges from 35— 55 cm, fog drip is important to the seasonal distri- bution of soil moisture (Vogl et al. 1988). Bishop pine grows between 200 and 400 m within the range of coastal low clouds in summer (DeMarrais et al. 1965). The concentration of stands on ex- posed ridgetops and western exposures optimize fo- liar fog condensation in strong northwesterly winds of the marine layer. Low temperatures reduce sum- mer transpiration. Bishop pine has recently sus- tained heavy mortality from needle blight caused by Dothistroma septospora (Ades et al. 1992). Most stands on Santa Cruz Island have been de- graded or extirpated by grazing and rooting of feral sheep and pigs since 1855 (Brumbaugh 1980; Min- nich 1980). The designation of P. remorata Mason for trees on Santa Cruz Island has been merged with P. muricata (Millar 1986). Pinus attenuata Lemmon (721 ha, Fig. 3).— Scattered even-aged stands occur in the San Ber- nardino Mountains from City Creek to Government Peak between 1000 and 1400 m. A VTM error showing a large stand on Keller Peak was a con- fusion for juvenile P. coulteri stands regenerating from fire about 1905 as seen in 1938 aerial photo- graphs. Knobcone pine grows on concave summits and ridges of granitic substrate with thin chaparral dominated by Adenostoma fasciculatum Hook. & Arn., Quercus wislizeni A.DC., and Arctostaphylos glandulosa Eastw. (Minnich 1999). While almost all stands were burned in 1938, 1956, and 1970, there was abundant post-fire regeneration in the same areas. Two colonies grow in the Santa Ana Mountains at 1000 m: one on the west slope near Pleasants Peak (Vogl 1973), and the other 2 km NE in Hagador Canyon. Both grow on fine-grained, acidic soils associated with hydrothermally altered volcanics (Vogl et al. 1988). The AAP ranges from 50 to 80 cm. Supplemental soil moisture from fog drip is seasonally phased with wet soils from winter storms and may not reduce drought stress in sum- mer. Coastal stratus at the elevations of P. attenuata is most frequent in late spring when fog drip totals reach 10 cm per month (Vogl 1973). The climate is warm and dry in summer because stands lie above the marine layer (normally <600 m, De- Marrais et al. 1965). Cupressus forbesii Jepson (268 ha, Fig. 4).— This Baja California species grows on north-facing slopes between 200 and 1200 m at four localities: an isolated stand in Gypsum Canyon in the northern Santa Ana Mountains, and at Guatay, Tecate Peak, and Otay Mountain in San Diego Co. (AAP, 40-70 cm). Tecate cypress was reported from Hot Springs Mountain in 1880 (Griffin and Critchfield 1976). The Tecate Peak and Otay Mountain stands grow on alkaline clay soils derived from gabbro basalts rich in iron and magnesium (Zedler 1981; Schoen- herr 1992). The stand at Guatay grows on granite. The Gypsum Canyon stand grows on Paleocene sandstone, where trees over 200 years old were re- corded on canyon floors (Schoenherr 1992). It was not shown in Griffin and Critchfield (1976), al- though the stand was recorded on VTM quadrangle maps. Gene frequency data shows that genetic dis- tance was not significantly associated with geo- graphic distance between populations (Truesdale and McClenaghan 1998). Variable fire intervals 2001] Pinus torreyana 40 0 40 _ 80 Kilometers Fic. 5. botanical collections or field vouchers. over long-time scales encourage small-scale extinc- tions and reinvasions (Zedler 1981). Stand degra- dation may be rare because only 6.0% of the chap- arral sustained fire intervals shorter than 20 years during 1920-1972 (Minnich and Chou 1997). The Gypsum Canyon stand regenerated well after a stand-replacement fire in 1948 (Griffin and Critch- field 1976). In northern Baja California, hundreds of even-aged groves grow in a fine-grained chap- arral patch mosaic of numerous small burns (Min- nich and Chou 1997). Cupressus arizonica E. Greene ssp. arizonica Reyfeldt (5.3 ha, Fig. 4).—A single stand grows on a chaparral-covered west-facing slope of Mt. Cu- yumaca near Japacha Peak at an altitude of 1400 m (AAP, 75 cm). Discriminant analysis of 15 genetic traits align the Cuyumaca cypress population in the C. arizonica spp. arizonica complex which occurs in Arizona and northwestern Mexico; the Cuya- maca population may also occur among inbred lines (Rehfeldt 1997). The nearest stand occurs in the southern Sierra Juarez of northern Baja Cali- fornia (Broder 1963). This tree was formerly clas- sified as C. stephensonii C.B. Wolf and C. arizonica var. stephensonii (C. B. Wolf) Little. Pinus coulteri D. Don (33,587 ha, Fig. 5).—A near endemic to California, Coulter pine is most widespread in the southern part of the state. It grows between 1000 and 2300 m on steep, concave slopes in association with dense chaparral and scrubby thickets of Quercus chrysolepis Liebm.., mostly on windward escarpments (AAP, 40-100 cm). Coulter pine responds to stand-replacement burns with high colonizing ability and reproductive effort, with mass post-fire recruitment from seed release from partially serotinous cones, leading to even-aged stands (Borchert 1985; Vogl et al. 1988; Minnich 1999). Mixed-aged stands with arboreal MINNICH AND EVERETT: CONIFER DISTRIBUTIONS 183 118 W Pinus coulteri The distribution of Pinus coulteri and P. torreyana. Shaded areas mapped from aerial photographs. x = subcanopy of Q. chrysolepis or Q. kelloggii Newb. occur on gentle slopes >1300 m (cf. Sawyer et al. 1988). Trees exhibit parallel decreasing cone sero- tiny with increasing elevation as stands shift from chaparral to oak woodlands (Borchert 1985). It is common in the northern Sierra Madre and San Ra- fael Mountains, and locally in the Santa Ynez Mountains. After a gap of 150 km, it recurs at Big Tujunga Canyon in the western San Gabriel Moun- tains and extends eastward into desert slopes at Rock Creek and coastal slopes to Crystal Lake. In the San Bernardino Mountains, Coulter pine is scat- tered across the southern escarpment east of Wa- terman Canyon. A continuous belt covers desert drainages from Lake Arrowhead to Holcomb Creek. It is extensive in the San Jacinto, Santa Ana, Hot Springs, Volcan, Palomar, and Cuyamaca Mountains but is not known from the Santa Rosa Mountains. The southernmost populations grow on Los Pinos Mountain. Stands tend to occur on gra- nitic and sandstone substrates that weather into deep soils having secure moisture retention for post-fire recruitment. In drier ranges, where there is greater chance of reproductive failure from drought, Coulter pine frequently survives fire on bedrock slopes, similar to stands in Baja California (Minnich 1987). In spite of comparable climate and topography, Coulter pine does not occur in per- meable Precambrian metamorphic gneisses and schists, including the Pelona Schist, that are exten- sively exposed in the Liebre, Frazier, and Alamo Mountains, as well as the eastern San Gabriel and western San Bernardino Mountains. It is also absent from extensive Eocene marine shales in Ventura County. Coulter pine sustained heavy mortality from insect predations during drought in 1989— 1991. Mortality rates as high as 50 ha“! were re- corded in the San Jacinto Mountains (Savage 184 MADRONO 34.N Pinus sabiniana 40 0 40 _ 80 Kilometers [Vol. 48 | 118 W Fic. 6. The distribution of Pinus sabiniana. Shaded areas mapped from aerial photographs. x = botanical collections or field vouchers. 1994). There were few changes in its range since the VTM survey. Pinus torreyana Carriére (391 ha, Fig. 5).—The rarest pine in the world (Griffin and Critchfield 1976), Torrey pine occurs at only two localities near the ocean. One includes two stands spanning 6.0 km on rapidly eroding coastal bluffs incised into Eocene sandstone in San Diego Co.: a compact forest at Torrey Pines State Reserve and a frag- mented population near Torrey Pines State Beach. The other locality is a 1.0 km forest growing in several ravines cut into upper Miocene sandstone above Beechers Bay on Santa Rosa Island, 270 km to the NW. Prescribed burn experiments shows that severe or fatal fire damage is followed by mass re- cruitment associated with seed dispersed from se- rotinous cones, similar to P. coulteri (Wells and Ge- tis 1999). The low AAP of 25-35 cm at both lo- calities is supplemented by fog drip (Schoenherr 1992) and is offset by low summer transpiration rates. Haller (1986) proposes that the island popu- lation be named P. torreyana ssp. insularis on the basis of gross morphology and garden plantings. Pollen grains in a sediment core taken from an es- tuary on Santa Rosa Island indicate that P. torrey- ana has been continuously present there since the middle Holocene (Cole and Liu 1994). Foothill woodland. Foothill woodland consists of open parks of deciduous and evergreen oaks (Quer- cus lobata Nee, Q. douglasii Hook & Arn., Q. agri- folia Nee, Q. wislizeni A. DC.), scattered shrub cover, and exotic annual grassland (Griffin 1988). The California endemic Pinus sabiniana has an ex- tensive range on foothill slopes encircling the Cen- tral Valley that are characterized by hot summers and an AAP of 30—60 cm. Pinus sabiniana Douglas (3,931 ha, Fig. 6).— Gray pine grows in the San Rafael Mountains from Lazaro Canyon to Zaca Ridge. Colonies extend in- land along Manzanita Creek and Sulphur Spring Canyon to Hurricane Deck where it grows on upper Cretaceous bedrock exposures. It is common in the Liebre Mountains. Botanical collections near Lake Piru mark the southern limit of the species. In the San Rafael Mountains, gray pine is found in mixed- aged stands, the trees apparently surviving under- story fires. In the Liebre Mountains, many stands are even-aged due to stand-replacing chaparral burns, the recruitment establishing from seed re- leased by serotinous cones. The distribution of gray pine is similar to that on VWTM quadrangle maps. Mixed evergreen forest. Mixed evergreen forest, which comprises long-lived evergreen coniferous and hardwood trees, is extensive in the coastal ranges of northern and central California (Sawyer et al. 1988). In southern California, this type con- sists of the Pseudotsuga macrocarpa—Quercus chrysolepis phase. One outpost of P. menziesii grows near Lompoc. Pseudotsuga menziesii (Mirbel) Franco var. men- ziesii (4.2 ha, Fig. 7).—A single colony occurs in a moist canyon in the Purisima hills (Griffin 1964). It is surrounded by chaparral and Pinus muricata D. Don stands. This population receives abnormally low precipitation for the species (AAP, 40 cm), and apparently survives on cool summers and the high water table of the watercourse. The nearest stand lies 150 km NW in the Santa Lucia Mountains. Pseudotsuga macrocarpa (Vasey) Mayr (22,232 ha, Fig. 7)—A southern California endemic, this long-lived tree (Bolton and Vogl 1969; McDonald and Littrell 1976; Haston and Michaelson 1994) has a fragmented distribution of compact groves in steep canyons, north-facing slopes, and cliff faces MINNICH AND EVERETT: CONIFER DISTRIBUTIONS 185 2001] 118W x R cow ee RS Pseudotsuga Cae Sate WE menziesii TRE a 34N Pseudotsuga macrocarpa 40 0 @ 40 _80 Kilometers IG. 7. x = botanical collections or field vouchers. between 1000 and 2200 m. Big-cone Douglas fir grows largely on windward escarpments, especially on slopes with high rates of erosion associated with fault-shattered substrate (AAP, 50-110 cm). The small-scale map in Griffin and Critchfield (1976) portrays a more continuous distribution than de- picted on VTM quadrangle maps and Figure 7. Rare colonies in La Brea Creek in the northern Sierra Madre Mountains are 20 km disjunct from abundant stands on Figueroa and Big Pine Moun- tains of the San Rafael Mountains. Numerous groves occur Over an extensive area between the San Rafael Mountains and Lake Piru. Between Pine Mountain Ridge and Mount Pinos, big-cone Doug- las fir is surprisingly abundant in semiarid badlands eroding into Quaternary alluvial terrace deposits and Pliocene sedimentary rocks, where it is asso- ciated with Pinus monophylla Torrey & Frémont. It is scarce on the undissected slopes of Mt. Pinos, Frazier Mountain, and Alamo Mountain, but com- mon in canyons along the San Andreas fault in the San Emigdio and Liebre Mountains. Extraordinary stands grow at 600 m in the eastern Santa Susana Mountain badlands on Pliocene and upper Miocene marine sandstones and shales. Big-cone Douglas fir is extensively distributed in deeply incised canyons cutting into fault-shattered substrate of the San Ga- briel Mountains. Continuous forests (stands >500 ha) are found near Mt. Wilson and Mt. San Anto- nio. In the San Bernardino Mountains, colonies span the southern escarpment and desert drainages west of Lake Arrowhead. In the Peninsular Ranges, big-cone Douglas fir is mostly restricted to escarp- ments of active faults. Along the Elsinore fault, it is common in the Santa Ana, Palomar, and Volcan Mountains. It is uncommon along the San Jacinto fault from Bautista Mountain to Thomas Mountain. Pseudotsuga macrocarpa is not known from the The distribution of Pseudotsuga macrocarpa and P. menziesii. Shaded areas mapped from aerial photographs. Santa Rosa, Hot Springs, Cuyamaca, and Laguna Mountains. Although big-cone Douglas fir survives recurrent fire in association with arboreal Quercus chrysole- pis in convex canyons, it has recently sustained widespread extirpations from wind-driven chaparral fires. In the eastern San Gabriel and San Bernardino Mountains, stand-replacement burns have exceeded recolonization rates, resulting in a net extirpation of 18% of forests since the 1938 aerial photographs (Minnich 1999). Current losses may be a natural perturbation due to variability in the fire process. However, fires should not act to synchronize stand age structure, nor produce extensive changes in the spatial extent of stands over short time scales be- cause subcanopy burns result in continuous recruit- ment and high tree survivorship over multiple fire cycles. The extirpation of P. macrocarpa is not due to excess fuel build-up because regional fire inter- vals have changed little under fire suppression (Minnich and Chou 1997; Minnich 1999). Alter- natively, the elimination of stands may be related to the nonrandomization of large fire occurrences to the severest weather conditions by suppression, notably Santa Ana winds (Minnich and Chou 1997). In the 19th century, forests were burned by low intensity fires persisting for months (Minnich 1987b). VTM stands have been reduced or extir- pated in many areas of the Cobblestone and Topa- topa Mountains, the eastern Transverse Ranges, Lake Hemet, Mt. Palomar, and Volcan Mountain. Pinyon-juniper woodland. Pinyon-juniper wood- land, which comprises P. monophylla, Juniperus osteosperma (Torrey) Little and J. californica Car- riére, is widespread on leeward mountain slopes and along the margin of the Mojave and Sonoran Deserts. Trees grow with open shrub cover of Great 186 Pe . CNS ty 34N Pinus monophylla 40 _0 40 _ 80 Kilometers FIG. 8. or field vouchers. Basin sage scrub, desert chaparral, and Mojave Desert scrub. Pinus quadrifolia occurs in chaparral in inland coastal slopes. Pinus monophylla Torrey & Frémont (156,815 ha, Fig. 8).—Single-leaf pinyon is found on lee- ward escarpments from 1300-2700 m (AAP, 20— 50 cm). Extensive stands occur in the plateaus of the upper Cuyama River, Lockwood Valley, Sis- quoc River, and Sespe Creek, including a few sand- stone outcrops in the Topatopa Mountains. It is widespread on the northern escarpments of the San Gabriel and San Bernardino Mountains. A small population grows on fire-resistant cliffs in Cattle Canyon on the coastal slope of Mt. San Antonio (Thorne 1988). It is common in Joshua Tree Na- tional Park, especially on granites. In the Peninsular Ranges, P. monophylla is widespread only in the Santa Rosa Mountains from Pinyon Flat to Rabbit Peak. Small stands occur in the northeastern San Jacinto Mountains and the summits of Pinyon Ridge and Whale Peak. Outliers grow on Mt. La- guna and east of McCain Valley. Single-leaf pinyon is extensive above 1300 m in the northeast Mojave Desert. In the Colorado Desert, small stands cover two summits of the Eagle Mountains and another grows on the summit of the Coxcomb Mountains. Based on needle resin duct morphology, it had been proposed that P. monophylla be divided into two varieties, P. monophylla var. monophylla in the northern Mojave Desert and P. monophylla var. cal- iforniarum in the Sierra Nevada and southern Cal- ifornia, but The Jepson Manual does not recognize varieties of single-leaf pinyon (Lanner 1999). Since the VTM survey, stand-replacement burns have removed single-leaf pinyon stands at Rock Creek and near Wrightwood in the San Gabriel Mountains, in the northern San Bernardino Moun- tains, Pinyon Flat in the Santa Rosa Mountains, and MADRONO The distribution of Pinus monophylla. Shaded areas mapped from aerial photographs. x = botanical collections northwestern Joshua Tree National Park. However, the distribution of single-leaf pinyon appears to be stable because low primary productivity rates as- sociated with low AAP limit fire intervals in the order of centuries, and gradual recolonization is seen in burns dating to the early 19th century (Wangler and Minnich 1996). Establishment does not appear to be limited by fire size due to efficient seed dispersal and caching by birds and rodents (VanderWall 1997). Juniperus californica Carriére (149,464 ha, Fig. 9).—California juniper is common between 800 and 1700 m on alluvial bajadas extending onto the Mojave Desert from the Sierra Nevada and the north slope of Transverse Ranges to Joshua Tree National Park. It is locally abundant on the leeward flank of the Peninsular Ranges, with extensive stands in the Santa Rosa Mountains. Outposts occur in semiarid coastal drainages, including the Cuya- ma River, upper Soledad Canyon, alluvial fans of the Transverse Range (Azusa, Lytle Creek, Red- lands), and in the Riverside-Perris Plain (Temescal Canyon, Lake Mathews, Gavilon Hills, Perris, Lakeview Mountains, Aguanga). A few trees occur near the coast in the Santa Monica Mountains (Ra- ven and Thompson 1966) and in the Santa Ana Mountains. The AAP ranges from 20—40 cm. Extensive stands mapped by VTM workers in the Mojave Desert have been denuded by fires carried by exotic grassland dominated by Bromus madri- tensis L. and Schismus barbatis L. (Lovich and Bainbridge 1999). Stands have also been cleared for development. Juniperus osteosperma (Torrey) Little (1,619 ha, Fig. 10).—Utah juniper occurs in the eastern Trans- verse Ranges between 1500 and 2200 m (AAP, 20— 35 cm). In the San Gabriel Mountains, a few stands grow near Phelan. It is common on the northern 2001] [34 N Juniperus californica 40 0 40 80 Kilometers Fic. 9. lections or field vouchers. escarpment of the San Bernardino Mountains from Grapevine canyon to Tip Top Mountain, with stands concentrating on carboniferous calcite and dolomite marbles. Large stands have been extirpat- ed by the development of limestone quarries. Bo- tanical collections of EF C. Vasek (UCR) indicate that most specimens are intergrades with J. occi- dentalis Hook. var. australis and J. californica Car- riere. Utah juniper is common in the higher ranges of the northeast Mojave Desert. Pinus quadrifolia Parl. (668 ha, Fig. 11).—An endemic to the Peninsular Ranges of southern Cal- ifornia and northern Baja California, four-needle pinyon grows in chaparral from 1300-1800 m (AAP, 35-55 cm). The largest stands occur at Tho- 34.N Juniperus osteosperma 40 0 40 80 Kilometers MINNICH AND EVERETT: CONIFER DISTRIBUTIONS The distribution of Juniperus californica. Shaded areas mapped from aerial photographs. x = 187 118 W botanical col- mas Mountain, southern Garner Valley, and Van- deventer Flat in the southern San Jacinto Moun- tains. Small stands occur 55 km south on Mt. La- guna and in McCain Valley. Time-series aerial pho- tographs in the Sierra San Pedro Martir show that stand-replacement burns are followed by rapid re- colonization in mature chaparral. Since the VTM survey, large populations have established in 80-yr old chaparral at the south end of Thomas Mountain. Pinyon pines are not known for cone serotiny, and recruitment appears to be dependent on seed cached by birds and mammals. Lanner (1999) proposes that P. quadrifolia is a hybrid between P. monophylla and P. juarezensis (a five-needle pinyon). The designation is based on 118 W Fic. 10. The distribution of Juniperus osteosperma. Shaded areas mapped from aerial photographs. x = botanical collections or field vouchers. 188 MADRONO 34.N Pinus We 40 0 40 80 Kilometers Fic. 11. or field vouchers. large variation in needle number, leaf resin canal number, twig hairiness and stomate position in pop- ulations of the La Rumerosa Plain near the Mexican border. However, P. monophylla and P. quadrifolia have strong ecologic and geographic segregation, with P. monophylla forming open stands on the desert-facing slopes and P. quadrifolia occurring in chaparral on coastal escarpments, their ranges broadly overlapping only on the La Rumerosa plain (Minnich 1987a). Monospecific P. quadrifolia woodlands, such as those on coastal slopes of the Sierra San Pedro Martir or on Thomas Mountain, should be investigated for their morphological and genetic properties. Pinus juarezensis is not recog- nized in The Jepson Manual (Hickman 1993). Mixed-conifer forest. While the modest eleva- tions of the southern California mountains limit the spatial extent of mixed-conifer forest, the broad dis- tribution of dominant tree species resembles those in the Sierra Nevada (Barbour and Minnich 2000). Mesic “‘westside”’ forests of P. ponderosa, P. lam- bertiana, Calocedrus decurrens and Abies concolor grow on windward slopes (AAP, 75—115 cm). Drier ‘“‘eastside”’ forests of P. jeffreyi, A. concolor and J. occidentalis grow on leeward slopes (AAP, 30—75 cm). “‘Eastside’’ forests also cover thin soils on windward flanks of dissected ranges. Mixed-conifer forest grows in areas with low combustible shrub biomass, mostly at higher altitudes above the chap- arral belt (>1500—2200 m) or in shrub-free basins within the chaparral belt (Minnich 2001). Subcan- opy consists of open cover of montane chaparral dominated by Arctostaphylos pringlei C. Parry, A. patula E. Greene, C. cordulatus Kell., Ceanothus integerrimus Hook. & Arn., Cercocarpus ledifolius Nutt., Chrysolepis sempervirens (Kell.) Hyjelmg., and Rhamnus californica Eschsch. Important hard- [Vol. | 118 W The distribution of Pinus quadrifolia. Shaded areas mapped from aerial photographs. x = botanical collections wood associates are Q. chrysolepis and Q. kellog- gli. Mixed-conifer forest is not found in the Santa Ana Mountains despite summit altitudes of 1700 m. Pinus ponderosa Laws (22,841 ha, Fig. 12).—In southern California, WTM workers identified Jef- frey pine as P. ponderosa var. jeffreyi, resulting in ambiguous differentiation between P. ponderosa and P. jeffreyi (Griffin and Critchfield 1976). VIM maps show confusion in the Transverse Ranges where P. ponderosa was erroneously shown to be extensive. In our survey, ponderosa pine was dis- tinguished from Jeffrey pine on aerial photographs based on deeper yellow-green (blue-green) foliage and more cylindrical (conical) shape of canopies of ponderosa (Jeffrey) pine. In the field we noted for foliage color and yellow (brown) immature cones of ponderosa (Jeffrey) pine. The reduction in the spatial extent of ponderosa pine in Fig. 12 com- pared to VTM maps is consistent with its decreas- ing importance southward in California. Ponderosa pine is most abundant in areas with high AAP and deep soils between 1400 and 2100 m. Small colonies cover the San Rafael Mountains, and it is common along Pine Mountain Ridge. Pon- derosa pine appears to be absent from Mt. Pinos, Frazier Mountain, and Alamo Mountain (Vogl and Miller 1968). We saw a few stems on groundwater seeps along the San Andreas fault in Cuddy Valley and north of Mt. Pinos. A monotypic stand covers Sawmill Mountain in the Liebre Mountains. Pon- derosa pine covers small basins or gentle summits in the San Gabriel Mountains. The largest stands grow from Mt. Waterman to Mt. Islip and near Crystal Lake (cf. Thorne 1988). The only extensive stand is a 30-km belt along an old erosion surface of weathered granite in the western San Bernardino Mountains. Trees grow as low as 1100 m, compa- 34.N Pinus ponderosa 40 0 40 80 Kilometers Fic. 12. or field vouchers. rable to other “‘westside”’ stands in the Sierra Ne- vada. Large stands grow at Barton Flats and Mill Creek. In the Peninsular Ranges, ponderosa pine occurs on the west flank of Mt. San Jacinto. To the south, a small colony grows at Doone Valley on Mt. Palomar, and a single tree 30 m tall straddles an arroyo on Hot Springs Mountain at 1600 m. Stands on Mt. Cuyamaca represent the southern limit of the species along the Pacific Coast. Calocedrus decurrens (Torrey) Florin (13,952 ha, Fig. 13).—The range of incense cedar is similar to P. ponderosa, except that it concentrates on stream courses from 1100 to 2200 m. It occurs lo- cally in the San Rafael Range and in one north- facing canyon of the northern Sierra Madre. Stands SCX 34.N Calocedrus decurrens 40 0 40 80 Kilometers Fic. 13. collections or field vouchers. MINNICH AND EVERETT: CONIFER DISTRIBUTIONS 189 118 W The distribution of Pinus ponderosa. Shaded areas mapped from aerial photographs. x = botanical collections are common from Pine Mountain Ridge to Alamo Mountain, and from the San Emigdio Mountains to the Tehachapi Mountains. It grows in canyons throughout the San Gabriel Mountains. In the San Bernardino Mountains, incense cedar forms wide- spread understory thickets in dense ponderosa pine forest from Crestline to Running Springs, in Barton Flats, and in Mill Creek, with stands concentrating near watercourses in open forests (Minnich et al. 1995). Subcanopy thickets are also widespread in the northern San Jacinto, Palomar, and Cuyamaca Mountains. Local mature stands grow in the drier Santa Rosa, Hot Springs, Volcan, and Laguna Mountains. Pinus lambertiana Douglas (53,477 ha, Fig. 118 W ae) a. The distribution of Calocedrus decurrens. Shaded areas mapped from aerial photographs. x = botanical 190 MADRONO 34N Pinus lambertiana 40 0 40 80 Kilometers [Vol. 48 118 W x Fic. 14. The distribution of Pinus lambertiana. Shaded areas mapped from aerial photographs. x = botanical collec- tions or field vouchers. 14).—Sugar pine is common on steep, mostly north-facing slopes between 1700 and 2700 m. Iso- lated colonies grow in the San Rafael Mountains, but it is extensive on Pine Mountain Ridge, Cob- blestone Mountain, and north-facing cliffs of the Topatopa Mountains. It is rare in the semiarid rang- es to the north, except locally on Mt. Pinos. Sugar pine is common across the San Gabriel and San Bernardino Mountains, but is absent on leeward slopes near Wrightwood and Big Bear. In the Pen- insular Ranges, it is widespread only on Mt. San Jacinto. Local stands grow on the Santa Rosa, Hot Springs, and Cuyamaca Mountains. It is not known from the Palomar, Volcan, and Laguna Mountains. Abies concolor (Gordon & Glend.) Lindley bo" we wie, SPL 34. N Abies concolor 40 0 40 __ 80 Kilometers BIG): or field vouchers. (107,415 ha, Fig. 15).—Vasek (1985) provides ev- idence that southern California white fir is the Rocky Mountain variety A. concolor var. concolor, whereas the Pacific coast variety A. concolor var. lowenia grows in the Sierra Nevada southward to the Tehachapi Mountains. White fir is often domi- nant on north-facing slopes from 1500 to 2800 m. It grows in the San Rafael Mountains, along Pine Mountain Ridge, and in “‘eastside’’ forests covering the Mt. Pinos, San Emigdio, Alamo, Frazier, and Tehachapi Mountains. White fir is found throughout the San Gabriel and San Bernardino Mountains with extensive subcanopy thickets growing in the dense mixed-conifer forests at Lake Arrowhead and Barton Flats (Minnich et al. 1995). The tree is 118 W ta abe The distribution of Abies concolor. Shaded areas mapped from aerial photographs. x = botanical collections bez ar MES ae, SC 34.N Pinus jeffreyi 40 0 40 80 Kilometers Fic. 16. The distribution of Pinus jeffreyi. Shaded areas or field vouchers. widespread on Mt. San Jacinto, and an invasive subcanopy tree near Idyllwild. Small populations occur on Palm Divide, Santa Rosa Mountain, Tho- mas Mountain, and Hot Springs Mountain. White fir is invasive in the Palomar and Cuyamaca Moun- tains. A VIM record of white fir in the Volcan Mountains cannot be confirmed, and it is not known from the Laguna Mountains. The Rocky Mountain variety grows in the Kingston and Clark Mountains of the northeast Mojave Desert. Pinus jeffreyi Grev. & Balf. (124,551 ha, Fig. 16).—Jeffrey pine is the forest dominant >2000 m in southern California. It also covers local shrub- free basins in the chaparral belt to as low as 1100 m in monospecific stands. Jeffrey pine typically re- places P. ponderosa along decreasing precipitation gradients. The confusion of P. jeffreyi and P. pon- derosa by VTM workers resulted in maps errone- ously showing P. jeffreyi in areas where only P. ponderosa occurs. Jeffrey pine is extensive on Pine Mountain Ridge, Mt. Pinos, Frazier Mountain, Alamo Moun- tain, and the San Emigdio Mountains. Monotypic stands extend downslope into Lockwood Valley and along the Cuyuma River plain to 1300 m. A few trees grow in the San Rafael Mountains. It spans the San Gabriel Mountains and the San Ber- nardino Mountains east of Lake Arrowhead. Jeffrey pine is common on Mt. San Jacinto, Santa Rosa Mountain, and the summit of Thomas Mountain. It is extensive in nearby Garner Valley at 1400 m, and a large VTM stand to the east was mistaken for P. coulteri. A single stand grows on Hot Springs Mountain (it was shown as P. ponderosa on VTM maps), but a few trees near Palomar Mountain Ob- servatory appear to be planted. Jeffrey pine is com- mon in the Cuyamaca Mountains and forms mono- typic stands on the Laguna Mountain plateau and 2001] MINNICH AND EVERETT: CONIFER DISTRIBUTIONS ot 118 W mapped from aerial photographs. x = botanical collections adjoining basins. The southernmost population is at Pine Valley at an altitude of 1100 m. Juniperus occidentalis Hook. var. australis (Va- sek) A. Holmgren & N. Holmgren (23,011 ha, Fig. 17).—Sierra Nevada western juniper is found lo- cally on Wright Mountain, Mt. San Antonio, and in Icehouse Canyon in the eastern San Gabriel Moun- tains. In the San Bernardino Mountains, it is wide- spread with Jeffrey pine >2100 m on the semiarid plateaus near Big Bear, with stands frequently ex- tending into pinyon-juniper woodland. Ancient trees with dbh >2-3 m are seen from Sugarloaf Mountain to Onyx Summit. Isolated trees occur on San Gorgonio Mountain, the southern limit of the species. While members of mixed conifer forest exhibit extensive range overlap, species boundaries are in- variably congruent at the edge of recent stand-re- placement burns. Congruent boundaries also occur at the chaparral ecotone. These trends reflect the similar adaptations of these conifers (tall stature, thick bark) to survive subcanopy burns recurring 2—3 times per century (Minnich et al. 2000). All mixed-conifers are selectively eliminated by stand- replacement fires in chaparral. Since the VTM sur- vey, fire suppression has led increasing fire inter- vals and stand-densification, with an age-specific trend away from dominance by P. ponderosa or P. Jeffreyi, and toward dominance by juvenile, pole- size classes of A. concolor and C. decurrens (An- sley and Battles 1998; Minnich 1988; Minnich et al. 1995; Roy and Vankat 1999; Minnich et al. 2000). Increasing fuel loads and stand-densification have led to widespread stand-replacement burns (Minnich 1999; Barbour and Minnich 2000; Min- nich et al. 2000), similar to that in the Sierra Ne- vada (Weatherspoon et al. 1992; McKelvey and Johnston 1992: SNEP 1996). Stands that were 192 MADRONO 34.N Juniperus occidentalis 40 0 40 80 Kilometers Fic. 17. collections or field vouchers. logged during the late 19th century, mostly at Lake Arrowhead, Idyllwild, Mt. Laguna and Mt. Cuya- maca (Minnich 1988; Pryde 1984) have densities exceeding 300—500 stems ha™', or are 1.5 to 6 times that recorded in forests sampled by VTM workers in 1929-1934, and in the Sierra San Pedro Martir, Mexico, where open parklike forests are produced by intense subcanopy burns at intervals of 50 years without fire control (Minnich et al. 1995, 2000). In southern California, open forests persist on drier leeward slopes with low primary productivity (Minnich et al. 1995) or in steep, dissected ranges with thin soils such as the San Gabriel Mountains. Crown fires have denuded extensive stands (>500 ha units) mapped by VWITM workers, with stands being replaced by successional montane shrub- lands, Q. chrysolepis and Q. kelloggii (cf. Kauff- man and Martin 1990, 1991). Portions of VTM for- ests were extirpated at Pine Mountain Ridge, Ala- mo Mountain, Frazier Mountain, as well as Pleasant View Ridge, Prairie Fork, Ontario Peak, and Cu- camonga Peak in the San Gabriel Mountains. Ex- tirpations also occurred near Big Bear Lake and Raywood Flat in the San Bernardino Mountains, Mount San Jacinto, Mount Cuyamaca, and the south edge of the Laguna Mountain plateau. Subalpine forest. Subalpine forest grows on 1so- lated summits >2500 m (AAP, 40—100 cm). Com- pared to the Sierra Nevada, southern California subalpine forests are floristically depauperate, with only two pine species occurring in the region. Pinus contorta Loudon ssp. murrayana (Grev. & Balf.) Critchf. (11,696 ha, Fig. 18).—In the San Gabriel Mountains, lodgepole pine occurs from Throop Peak to Mt. Baden-Powell, and on summits from Mt. San Antonio to Cucamonga Peak. In the San Bernardino Mountains, it grows on slopes and | 118 W The distribution of Juniperus occidentalis. Shaded areas mapped from aerial photographs. x = botanical valley floors at Butler Peak, Sugarloaf Mountain, Holcomb Valley, Bluff Lake, Big Bear, and the up- per Santa Ana River. It dominates subalpine forests on San Gorgonio Mountain. The southernmost stands occur on Mt. San Jacinto and Tahquitz Peak. Pinus flexilis James (6,642 ha, Fig. 19).—Small colonies grow on Mt. Pinos and nearby Cerro Nordeste. In the San Gabriel Mountains, the only major population extends from Throop Peak to Mt. Baden-Powell; a few trees occur 10 km E on Wright Mountain. It is strangely absent from lodge- pole pine forests near Mt. San Antonio. In the San Bernardino Mountains, limber pine is common on Sugarloaf Mountain, Onyx Summit, and other semiarid peaks in the upper Santa Ana River where it grows with ‘‘eastside” stands of P. jeffreyi, A. concolor, and J. occidentalis. At Onyx Summit (AAP, 35 cm), limber pine forms ecotones with sin- gle-leaf pinyon woodlands, similar to forest zona- tion in the Great Basin. It is common >2,800 m on Mt. San Gorgonio. In the San Jacinto Mountains, stands grow on Mt. San Jacinto and Tahquitz Peak. Maps of Griffin and Critchfield (1976) show them 15 km too far east. A small population on Toro Peak in the Santa Rosa Mountains is the southern limit of the species. While subalpine forests in southern California experience numerous small burns initiated by light- ning, site-specific mean fire intervals are in the or- der of centuries (Sheppard and Lassoie 1998). Stands show local patchiness from 19th century stand-replacement burns on the San Bernardino Ridge, Mt. Grinnell (Minnich 1988), and the east face of Mt. San Jacinto. Since the VIM survey, stand-replacement burns have occurred in the upper Whitewater River near Mount San Gorgonio, near Butler Peak, and on Ontario Peak. All burns are regenerating well. 34.N Pinus contorta ssp. murrayana 40 OO 40 _ 80 Kilometers Fic. 18. or field vouchers. 34.N Pinus flexilis 40 0 40 _ 80 _Kilometers MINNICH AND EVERETT: CONIFER DISTRIBUTIONS The distribution of Pinus contorta. Shaded areas mapped from aerial photographs. x = [93 118 W Ke botanical collections 118 W : d, Fic. 19. The distribution of Pinus flexilis. Shaded areas mapped from aerial photographs. x = botanical collections or field vouchers. DISCUSSION While distributions viewed statically provide in- sight into ecological relationships, how a species adapts to an environment can also be evaluated from a synoptic time-series frame of reference. In addition, a regional approach permits the exami- nation of the broadscale status of populations as seen in time- and space-averaging of local scale population dynamics. The assessment of stand dy- namics is not straightforward with respect to whether species changes reflect normal biomass re- moval and accumulations from fire and postfire successions, or whether they reflect long-term di- rectional change. This distinction cannot be made here from a time-series comparison of forest maps only 70 years apart. However, for the purpose of hypothesis testing of population stability, we pro- pose that vegetation baselines in southern Califor- nia be interpreted using three fire disturbance mod- els which address key processes or patterns that in- fluence stability. Shifting population model (closed-cone conifer forests, Pinus quadrifolia woodland). The large size of chaparral stand-replacement burns (two to three per century) relative to the size of individual co- nifer populations tends to synchronize stand age distributions across entire stream drainages, result- ing in large temporal variation in their spatial extent at local scales. Stands “‘disappear’’ in recent burns, but emerge as even-aged stands in older chaparral 194 patches, “‘following”’ old-growth patch mosaics. To evaluate the stability of shifting populations, it may be best to search for postfire recruitment failures (stands which fail to reestablish after burns, e.g., Zedler 1981; Vogl et al. 1988) because individual fires can eliminate both the adult population and the seed bank, resulting in local extinctions. Pinus quadrifolia forests have these characteristics, but individual stands are mixed-aged because recruit- ment is continuous through successions. Standing population model (mixed-conifer forest, mixed-evergreen big-cone Douglas fir forest, gray pine foothill woodland). The canopy layer persists through recurrent subcanopy fires (two to three per century), resulting in “‘standing”’ or fixed distribu- tions with small local fluctuations. Although trees recruit continuously, stands undergo intense selec- tive elimination of sapling and polesize trees from subcanopy fires, with a few mid-size trees incre- mentally joining the canopy layer between fire se- quences, producing mixed-aged overstory. Discrete age mosaics in the canopy layer are rare because stand-replacement burns are rare. Stands occupy sites safe from dense shrub subcanopy that nor- mally produce stand-replacement fires, especially chaparral. Stability can be evaluated from the long- term expansion or contraction of “‘standing”’ tree populations over multiple fire cycles. A key factor may be the spatial extent of individual stand-re- placement burns, and whether stand attrition is bal- anced by recolonization. Since the initiation of fire Suppression, stand-replacement burns have fre- quently exceeded 1000 ha, whereas patches created by crown fires rarely exceed 10 ha without fire con- trol in mixed-conifer forests of the Sierra San Pedro Martir of Baja California (Minnich 1999; Minnich Cleal, 2000); Patch mosaic population model (pinyon-juniper woodland, subalpine forest). Low productivity rates and stand structure result in long interval subcan- opy to canopy fires (return intervals <1 per cen- tury). Stand mortality is high because the flame front both consumes tree canopy, and the thin bark of partially burned conifers results in fatal cambium damage. Stands exhibit discrete inter-stand age structures arrayed in a patch mosaic. Individual stands are mixed-aged due to continuous recruit- ment, especially in shade-tolerant species such as P. monophylla and P. flexilis. Trees exhibit little change in broadscale distributions, but patches ap- pear in the form of recent stand-replacement burns. These patches fade after 50-100 years as recruit- ment gradually establishes closed canopy. Older age class boundaries are subtle because fire return intervals scale at several per millennia. Trees are longer-lived than members of closed-cone conifer forests and stability is harder to establish over time scales of a century. The potential for destabilization of forests may again be reflected in the size of stand-replacement burns. MADRONO [Vol. 48 How can these population models be applied to! long-term species changes? While the distribution of conifer forests has been altered by fire over the past century, the stability of these ecosystems can- not be judged alone by the historical severity and spatial extent of fires. Fire regimes in California forests are an outgrowth of cumulative fuel build- up scaling from several decades to centuries, the time lag between fuel accumulation and burning making fire self-limiting, and time-dependent (Min- nich and Chou 1997; Minnich et al. 2000). The site- specific properties of fire (intervals, intensities, re- moval of biomass) vary with climate, primary pro- ductivity, and fuel accumulation rates and exert profound selection in tree species distributions de- pending on their life history traits (Veblen et al. 1991; Christensen 1993; Johnson and Gutsell 1994). Forest ecosystems tend to burn most fre- quently in areas with highest productivity (cf. Knight 1987; Veblen et al. 1991). In southern Cal- ifornia growth rates are proportional to mean an- nual rainfall, except at highest elevations (>ca. 2300 m) where productivity is limited by short growing seasons. Fire return intervals range from two to three times per century in closed-cone, mixed-evergreen, and mixed-conifer forest on moist coastal windward slopes (Minnich and Chou 1997; Minnich et al. 2000) to <1 per century in pinyon-juniper woodland on semiarid leeward slopes (Wangler and Minnich 1996) and subalpine forests on highest summits. Fire severity and co- nifer mortality rates also vary because conifer spe- cies are associated with divergent subcanopy veg- etation and vertical distribution of fuels (Minnich 2001). The role of climate variability on fire regimes and long-term directional change of forests is a gradual process because fire outcomes are an out- growth of long-term vegetation successions and fuel build-up. Directional vegetation changes also lag behind climatic perturbations because selection processes that result in changes in recruitment and successions require several generations to translate into mature phases of the vegetation. Fire suppression, which is unprecedented in eco- logical history, may produce rapid change, es- pecially where fires have been excluded from for- ests for extended periods. In addition, the intensity of large fires may have increased because suppres- sion both encourages excessive fuel build-up where fire intervals have lengthened and selectively re- stricts uncontrolled fires to the severest weather conditions (Minnich and Chou 1997). Forests ex- periencing shifting or patch mosaic population dy- namics normally experience high fire severity and mortality, and how changes in fire intensity with suppression would change the stand-replacing fire regime is unclear. Fire intervals in chaparral and closed-conifer forests have been stable with or without suppression (Minnich and Chou 1997). Un- productive patch mosaic model ecosystems may be MINNICH AND EVERETT well within presuppression fire-free periods. In spite of local removal of forests by stand-replace- ment burns, the spatial extent of conifer species having shifting and patch mosaic population dy- namics has changed little during the 20th century. The demonstration of long-term change in these ecosystems would require evidence that replace- ment recruitment rates are dependent on the sur- vivorship of the adult population. Reduced repro- ductive potential in short interval fire recurrences may be infrequent because the turnover of patch mosaics is dependent on cumulative fuel build-up. Invasive exotic annual grasses that produce abun- dant cured fuel, including Bromus madritensis L., B. diandrus Roth, and Avena barbata Link, may increase the frequency of short-interval burn se- quences and degradation of coastal sage scrub and chaparral (Freudenberger et al. 1987; Minnich and Dezzani 1998). However, invasive species are gen- erally limited to coastal valleys and lower foothills far removed from conifer ecosystems. Directional change in closed-cone conifers may arise from post-fire recruitment failure due to short-term ex- treme environmental conditions, such as drought, but on the basis of chance extreme conditions sel- dom coincide with fire cycles over long time scales. Recent fire history suggests that the greatest po- tential for directional vegetation change may occur in ecosystems having “‘standing” population dy- namics due to increasing fire severity and stand mortality (Minnich 2001). The maintenance of standing forests (mixed-conifer forest, bigcone Douglas fir forest) with recurrent subcanopy fires may be compromised by increasing fire intervals (e.g., Swetnam 1993; Minnich et al. 2000), stand- densification (Minnich et al. 1995; Albright 1998; Roy and Vankat 1999) and increasingly extensive stand-replacement burns. Over long time scales, the transformation of these forests from “‘standing”’ population dynamics to either “‘shifting” or “patch mosaic’” dynamics may result in progressive re- gional extirpations, a trend seen in the 20th century. LITERATURE CITED ADES, P. K., J. A. Simpson, K. G. ELDRIDGE, AND R. H. ELDRIDGE. 1992. Genetic variation in susceptibility to Dothistroma needle blight among provenances and families of Pinus muricata. Canadian Journal of For- est Research 22:1111—1117. ALBRIGHT, D. 1998. Vegetation change in second-growth forests of the western San Bernardino Mountains. MS thesis. University of California, Riverside, CA. ANSLEY, J. S. AND J. J. BATTLES. 1998. Forest composition, structure, and change in an old-growth mixed conifer forest in the northern Sierran Nevada. Journal of the Torrey Botanical Society 125:297—308. BarBour, M. G. AND J. MAsor. 1988. Terrestrial vegeta- tion of California, expanded edition. California Na- tive Plant Society Press. Sacramento, CA. AND R. A. MINNICH. 2000. Californian upland for- ests and woodlands. Pp. 161—202 in M. G. Barbour : CONIFER DISTRIBUTIONS 195 and W. D. Billings (eds.), North American Terrestrial Vegetation. Cambridge University Press, NY. BoLTon, R. B. JR. AND R. J. VoGL. 1969. Ecological re- quirements of Pseudotsuga macrocarpa (Vasey) Mayr in the Santa Ana Mountains, California. Journal of Forestry 67:112—116. BORCHERT, M. 1985. Serotiny and cone-habit variation in populations of Pinus coulteri (Pinaceae) in the south- ern coast ranges of California. Madrono 32:29—49. BRUMBAUGH, R. W. 1980. Recent geomorphic and vegetal dynamics on Santa Cruz Island. Pp. 139-158 in D. M. Power (ed.), The California Islands: proceedings of a multidisciplinary symposium. Santa Barbara Mu- seum of Natural History, Santa Barbara, CA. CALIFORNIA. 1980. California rainfall summary: monthly total precipitation, 1849-1984. Department of Water Resources. Sacramento, CA. CHRISTENSEN, N. L. 1993. Fire regimes and ecosystem dy- namics. Pp. 233-244 in P. J. Crutzen and J. G. Gol- dammer (eds.), Fire in the environment: the ecologi- cal, atmospheric, and climatic importance of vegeta- tion fires. John Wiley and Sons. CoLe, K. L. AND G. W. Liu. 1994. Holocene paleoecology of an estuary on Santa Rosa Island. Quaternary Re- search 41:326—335. COLWELL, W. L. 1988. The status of vegetation mapping in California today. Pp. 195—210 in M. G. Barbour and J. Major (eds.), Terrestrial vegetation of Califor- nia, expanded edition. California Native Plant Society Press, Sacramento, CA. CRITCHFIELD, W. B. AND E. L. LITTLE. 1966. Geographic distribution of the pines of the world. U.S. Depart- ment of Agriculture Miscellaneous Publication 991. Washington DC. DEMarralis, G. A., G. C. HOLZWORTH, AND C. R. HOSLER. 1965. Meteorological summaries pertinent to the at- mospheric transport and dispersion over southern California. Technical Paper 54. U.S. Weather Bureau, Washington DC. ESRI. 2000. Arc/Info Geographical Information System, ver. 8.0.1. Environmental Systems Research Institute, Inc. Redlands, CA. . 2001. Arcview GIS 3.2. Environmental Systems Research Institute, Inc. Redlands, CA. FREUDENBERGER, D. O., B. E. FISH, AND J. E. KEELEY. 1987. Distribution and stability of grasslands in the Los Angeles Basin. Bulletin of the Southern Califor- nia Academy of Science 86:13—26. GRIFFIN, J. R. 1964. A new Douglas-fir locality in southern California. Forest Science 10:317-319. . 1988. Oak woodlands. Pp. 383-415 in M. G. Barbour and J. Major (eds.), Terrestrial vegetation of California, expanded edition. California Native Plant Society Press. Sacramento, CA. AND W. B. CRITCHFIELD. 1976. The distribution of forest trees in California. USDA Forest Service Re- search Paper PSW-82. Pacific Southwest Forest and Range Experiment Station, Berkeley, CA. HALLER, J. R. 1986. Taxonomy and relationships of the mainland and island populations of Pinus torreyana (Pinaceae). Systematic Botany 11:39—50. HASTEN, L. AND J. MICHAELSEN. 1994. Long-term central coastal California precipitation variability and rela- tionships to El Nifio-Southern Oscillation. Journal of Climate 7:1373-1387. HickMaN, J. C. (ed.). 1993. The Jepson Manual: higher plants of California. University of California Press, Berkeley, CA. 196 MADRONO JOHNSON, E. A. AND S. L. GUTSELL. 1994. Fire frequency models, methods and interpretations. Advances in Ecological Research 15:239—287. KAUFFMAN, J. B. AND R. E. MARTIN. 1990. Sprouting shrub response to different seasons and fuel consumption levels of prescribed fire in Sierra Nevada mixed co- nifer ecosystems. Forest Science 36:748—764. AND . 1991. Factors influencing the scari- fication and germination of three montane Sierra Ne- vada shrubs. Northwest Science 65:180—187. LANNER, R. M. 1999. Conifers of California. Cachuma Press, Los Olivos, CA. Lovicn, J. E. AND D. BAINBRIDGE. 1999. Anthropogenic degradation of the southern California desert ecosys- tem and prospects for natural recovery and restora- tion. Environmental Management 24:309—326. McDONALD, P. M. AND E. E. LITTRELL. 1976. The bigcone Douglas fir—canyon live oak community in southern California. Madrono 23:310—320. McKELvey, K. S. AND J. D. JOHNSTON. 1992. Historical perspective on forests of the Sierra Nevada and the Transverse Ranges of southern California: forest con- ditions at the turn of the century. Pp. 225-246 in J. Verner, K. S. McKelvey, B. R. Noon, R. J. Gutiérrez, G. I. Gould, and T. W. Beck (tech. coords.), The Cal- ifornia spotted owl: a technical assessment of its cur- rent status. USDA Forest Service General Technical Report PSW-GTR-133. MILLAR, C. I. 1986. The Californian [USA] closed cone pines (subsection Oocarpae Little and Critchfield): a taxonomic history and review. Taxon 35:657—670. MINNICH, R. A. 1980. Vegetation of Santa Cruz and Santa Catalina Islands. Pp. 123-138 in D. M. Power (ed.), The California Islands: proceedings of a multidisci- plinary symposium. Santa Barbara Museum of Nat- ural History, Santa Barbara, CA. . 1982. Pseudotsuga macrocarpa in Baja Califor- nia? Madrono 29:22-31. . 1984. Snow drifting and timberline dynamics on Mt. San Gorgonio, California. Arctic and Alpine Re- search 16:395—412. . 1987a. The distribution of forest trees in northern Baja California. Madrofio 34:98—127. . 1987b. Fire behavior in southern California chap- arral before fire control: the Mount Wilson burns at the turn of the century. Annals of the Association of American Geographers 77:599-618. . 1988. The biogeography of fire in the San Ber- nardino Mountains of California: a historical survey. University of California Publications in Geography 28:1—120. . 1999. Vegetation, fire regimes, and forest dynam- ics. Pp. 44-78 in P. R. Miller and J. R. McBride (eds.), Oxidant air pollution impacts in the montane forests of southern California: a case study of the San Bernardino Mountains. Springer, NY. . 2001. Fire and elevational zonation of transborder mediterranean chaparral and conifer forests in the San Jacinto Mountains, California, and the Sierra San Pe- dro Martir, Baja California. Pp. 120-142 in G. Web- ster and C. J. Bahre (eds.), Changing Plant Life of La Frontera: observations of Vegetation in the United States/Mexico Borderlands. University of New Mex- ico Press, Albuquerque, AZ. , M. G. BARBour, J. H. BURK, AND R. FE FERNAU. 1995. Sixty years of change in Californian conifer forests of the San Bernardino Mountains. Conserva- tion Biology 9:902-914. [Vol. 48 ’ : , AND J. SOSA-RAMIREZ. 2000. Californian conifer forests under unmanaged fire re- gimes in the Sierra San Pedro Martir, Baja California, Mexico. Journal of Biogeography 27:105—129. AND Y. H. CuHou. 1997. Wildland fire patch dy- namics in the chaparral of southern California and northern Baja California. International Journal of Wildland Fire 7:221—248. AND E. FRANCO-VIZCAiNO. 1998. Land of chamise and pines: historical accounts and current status of northern Baja California’s vegetation. University of California Publications in Botany 80:1—166. | AND R. J. DEZZANI. 1998. Historical decline of coastal sage scrub in the Riverside-Perris Plain, Cal- ifornia. Western Birds 29:366—391. NIMA. 2001. DTED level 0 data. National Imagery and Mapping Agency. http://164.214.2.59/geospatial/ digital_products.htm. PRYDE, P. R. 1984. San Diego, an introduction to the re- gion: an historical geography of the natural environ- ments and human development of San Diego County. Kendall/Hunt, Dubuque, IA. RAVEN, P. H. AND H. J. THOMPSON. 1966. Flora of the Santa Monica Mountains, California. Dissertation. University of California, Los Angeles, CA. REHFELDT, G. E. 1997. Quantitative analyses of the genetic structure of closely related conifers with disparate distributions and demographics: the Cupressus ari- zonica (Cupressaceae) complex. American Journal of Botany 84:190—200. Roy, D. G. AND J. L. VANKAT. 1999. Reversal of human- induced vegetation changes in Sequoia National Park, California. Canadian Journal of Forest Research 29: 39-412. SAVAGE, M. 1994. Anthropogenic and natural disturbance and patterns of mortality in a mixed conifer forest stand in California. Canadian Journal of Forest Re- search 24:1149—1159. SAWYER, J. O., D. A. THORNBURGH, AND J. R. GRIFFIN. 1988. Mixed evergreen forest. Pp. 359-381 in M. G. Barbour and J. Major (eds.), Terrestrial vegetation of California, expanded edition. California Native Plant Society Press. Sacramento, CA. SCAMP. 2000. Digital Geologic scamp. wr.usgs.gov/ Schoenherr, A. A. 1992. A natural history of California. University of California Press. SHEPPARD, P. R. AND J. P. LAsSsorg. 1998. Fire regime of the lodgepole pine forest of Mt. San Jacinto, Califor- nia. Madrono 45:47—56. SNEP SCIENCE TEAM (eds.). (1996). Status of the Sierra Nevada, final report to Congress of the Sierra Nevada Ecosystem Project. Wildland Resources Center Re- port 36, University of California, Davis, CA. SWETNAM, T. W. 1993. Fire history and climatic change in giant Sequoia groves. Science 262:885—890. THORNE, R. FE 1988. Montane and subalpine forests of the Transverse and Peninsular Ranges. Pp. 537—557 in M. G. Barbour and J. Major (eds.), Terrestrial vegetation of California. Special Publication, California Botani- cal Society, Davis, CA. VANDER WALL, S. B. 1997. Dispersal of singleleaf pinon pine (Pinus monophylla) by seed-caching rodents. Journal of Mammalogy 78:181—191. VEBLEN, T. T., K. S. HADLEY, AND M. S. RErb. 1991. Dis- turbance and stand development of a Colorado sub- alpine forest. Journal of Biogeography 18:707—716. TRUESDALE, H. D. AND L. R. MCCLENAGHAN. 1998. Pop- Maps. _ http:// 2001] ulation genetic structure of tecate cypress (Cupres- saceae). Southwestern Naturalist 43:363—373. VASEK, F.C. 1985. Southern California white fir. Madrono | 32:65-77. UNITED STATES GEOLOGICAL SURVEY. 2000. 1:250,000 | Scale Digital Elevation Models (DEMs). http:// | edcwww.cr.usgs.gov/doc/edchome/ndcdb/ndcdb. html. VoGcL, R. J. 1973. The ecology of the knobcone pine in the Santa Ana Mountains, California. Ecological Monographs 43:125—143. , W. P. ARMSTRONG, K. L. WHITE, AND K. L. COLE. 1988. The closed-cone pines and cypresses. Pp. 295— 357 in M. G. Barbour and J. Major (eds.), Terrestrial Vegetation of California, expanded edition. California Native Plant Society Press, Sacramento, CA. AND B. C. MILLER. 1968. The vegetational com- position of the south slope of Mt. Pinos, California. Madrono 19:225—234. WANGLER, M. AND R. A. MINNICH. 1996. Fire and succes- MINNICH AND EVERETT: CONIFER DISTRIBUTIONS 197 sion in pinyon-juniper woodland in the San Bernar- dino Mountains, California. Madrono 43:493—5 14. WEATHERSPOON, C. P., S. J. HUSARI, AND J. W. VAN WAG- TENDONK. 1992. Fire and fuels management in rela- tion to owl habitat in forests of the Sierra Nevada and southern California. Pp. 247—260 in J. Verner, K. S. McKelvey, B. R. Noon, R. J. Gutiérrez, G. I. Gould, and T. W. Beck (Tech Coords), The California spotted owl: a technical assessment of its current status. USDA Forest Service General Technical Reports, PSW-GTR- 133. WIESLANDER, A. E. 1935. A vegetation type map of Cal- ifornia. Madrono 3:140—144. WELLS, M. L. AND A. GETIS. 1999. The spatial character- istics of stand structure in Pinus torreyana. Plant Ecology 143:153—-170. ZEDLER, P. H. 1981. Vegetation change in chaparral and desert communities in San Diego County, California. Pp. 406—424 in D. C. West, H. H. Shugart, and D. B. Botkin (eds.), Forest succession: concepts and appli- cation. Springer-Verlag, New York, NY. MApRONO, Vol. 48, No. 3, pp. 198-204, 2001 SALTUGILIA LATIMERI: A NEW SPECIES OF POLEMONIACEAE TERRI L. WEESE AND LEIGH A. JOHNSON Department of Botany and Range Science, Brigham Young University, Provo, UT 84602-5181 ABSTRACT Saltugilia latimeri, a new species of Polemoniaceae, is endemic to dry, rocky to sandy slopes and foothills, primarily in San Bernardino and Riverside Counties, California. Saltugilia latimeri superficially resembles S. australis and has, until now, been considered conspecific with this taxon based on the small corolla size shared by both species. The two species differ in several morphological characters including stature, corolla throat coloration and exertion, and degree of glandularity. Saltugilia latimeri, species, Polemoniaceae, taxonomy Gilia is the historical catchall genus within Po- lemoniaceae for species of uncertain phylogenetic affinity (Wherry 1940, Mason and Grant 1948). Within this polyphyletic genus, Mason and Grant (1948) described a new taxon, Gilia splendens ssp. australis H. Mason & A. D. Grant, with stamen insertion and corolla throat and tube proportions similar to those of Gilia splendens H. Mason & A. D. Grant, but with a smaller corolla and a propor- tionally longer limb, similar in these respects to Gi- lia caruifolia Abrams. As characters delimiting putative natural groups within Gilia were identified, such groups have been recognized with sectional or generic status. In one such case, Grant and Grant (1954) constructed Gi- lia section Saltugilia and included therein five spe- cies divided into two species groups. The section was considered to be natural from observations of corolla morphology and ecology of the members. Within their newly erected section Saltugilia, Grant and Grant (1954) elevated G. splendens ssp. aus- tralis to the species level, as Gilia australis (H. Mason & A. D. Grant) V. E. Grant & A. D. Grant. Grant and Grant (1954) distinguished G. australis from G. splendens and G. caruifolia by its smaller stature, whitish flowers (pale violet to white), and simpler leaves. The authors also noted minimal geographic overlap, partial to full breeding barriers, and apparent ecological isolation between G. aus- tralis and these other species. Johnson (in Porter and Johnson 2000) elevated Saltugilia to generic status within Polemoniaceae. Following rules of priority, Porter and Johnson also restored Brand’s (1907) epithet, grinnellii, in place of the more recently used epithet, splendens, for the largest-flowered species. As circumscribed by Por- ter and Johnson (2000), the genus Saltugilia com- prises four closely related taxa: Saltugilia australis (H. Mason & A. D. Grant) L. A. Johnson, S. ca- ruifolia (Abrams) L. A. Johnson, S. grinnellii (Brand) L. A. Johnson subspecies grinnellii and S. grinnellii subspecies grantii (Brand) L. A. Johnson. A fifth taxon in Saltugilia has been identified. Our observations of greenhouse-grown plants and herbarium specimens indicate that two distinct forms have been referred to S. australis (as Gilia australis). Herbarium collections at RSA indicate that Latimer (1958) recognized, in his unpublished dissertation, these two forms as distinct subspecies within G. australis: a typical subspecies australis, and a geographically restricted subspecies deserto- rum. Based on observations of additional distin- guishing characters, we here describe this geo- graphically restricted taxon as a new species in the genus Saltugilia. Saltugilia latimeri T. L. Weese & L. A. Johnson, sp. nov. (Fig. 1)—TYPE HERE DESIGNATED: USA, California, Riverside County, canyon in pass between Whitewater and Morongo Valley, 0.4 miles south of the San Bernardino county line, 9 April 1950, Verne & Alva Grant 8840 (holotype, RSA!). Species nova ab Saltugilia australis (H. Mason & A.D. Grant) L.A. Johnson differt tubo corollae violaceo et exserto (nec albo et incluso), lobi cor- ollae acutis (nec cuspidatis), et calyce dense glan- duloso (nec glabro, glabrescenti, vel sparse glan- duloso). Annual herbs, to 30 cm in height, scapiform with a central stem and several basally branching axes subequal to equal with the main axis. Lower stems with long, uniserriate transparent trichomes com- monly terminating in transparent glands (Fig. II; stalk cells becoming flattened and chain-like upon drying). Mid to upper stems densely glandular with the glands multicellular, flat-topped, and translucent (Fig. 1J; chloroplasts present in glands, but these nei- ther darken beyond amber nor appear anthocyanic, as is common in Gilia). Lower leaves persistent, 5— many, forming a loose to dense basal rosette. Basal and lower cauline leaves generally 2<—3 x pinnately divided with 3-10 sub-opposite lobes; leaf blade 20— 45 mm long (~70 mm on greenhouse grown plants), 6-16 mm wide (—-40 mm on greenhouse grown plants; Fig. 1C). Rosette and lower cauline leaves with transparent, glandular trichomes of varying lengths (Fig. 1H). Upper cauline leaves glandular (Fig. 1K), reduced, bract-like, entire or occasionally pinnately divided, 1.5-14 mm long (-30 mm on ‘greenhouse grown plants) and 0.2—6 mm wide (—17 -mm on greenhouse grown plants; Fig. 1D). Inflores- ‘cence cymose, with flowers borne singly, or paired on ultimate stem extensions. Pedicels 2-16 mm long, occasionally longer, densely glandular (cf. Fig. 1L). ‘When paired, pedicel of terminal (first maturing) flower usually shorter than pedicel of lateral (second maturing) flower (Fig. 1E). Calyx 2.6—3.5 mm long in flower, densely glandular (Fig. 1L—N), with an average of 20 glands per calyx lobe on herbarium sheets (range from 6 to 35 glands per calyx lobe). Calyx lobes dull green, sometimes purple spotted, 0.8 mm wide, united 3/5 length of calyx by an hyline membrane. Corolla 7.5—10 mm long, the fused por- tion 3—6.5 mm long, tube (=unexpanded portion of the fused corolla) exserted from the calyx. Tube dark lavender-purple, fading to blue (or yellow with ex- treme age) when dried. Lower throat (=expanded portion of fused corolla) yellow spotted; upper throat and lobes pinkish-lavender, but also fading to pale blue or yellowish on herbarium sheets. Corolla lobes 2.0—3.5 mm long and 1—3.5 mm wide, acute. Sta- mens inserted equally in the sinuses of the corolla lobes. Filaments 0.5—1.1 mm long to point of inser- tion, anthers 0.65—0.85 mm long and 0.35—0.50 mm wide. One stamen frequently extends at right angle to the throat. Style 7-8 mm long, extending slightly beyond the orifice. The three lobes of the stigma 0.8—1.0 mm long with tips curling downward. Cap- sule 3.5—5.0 mm long and 2.0—3.5 mm wide, typi- cally 1.4—1.8 times as long as wide, more or less equaling the length of the calyx. Capsule dehiscent in three sections from the top to the base, 6—9 ovules per locule. Seeds + 0.85 mm long, honey gold in color. Testa verrucate with anticlinal epidermal cell walls defined as low ridges, the ridges lost and cell wall boundaries obscure when this outer epidermal layer is shed upon wetting. Seeds mucilaginous via expansion of spiricals when wet. Pollen grains blue, +40 wm in diameter, 5—6 zonocolpate, with lirae radiating from the apertures in magnetic-field fash- ion. n = 9. Paratypes. U.S.A. California, Inyo County: Han- aupah Canon, Panamint Mountains, 16 May 1917, Jepson 7091 (JEPS). Riverside County: Box Can- yon, 16 Mar 1937, Winblad s.n. (CAS); Palms to Pines Hwy, Mount San Jacinto, 23 Apr 1937, Win- blad s.n. (CAS); Palm Springs, 11 Apr. 1920, Spen- cer 1569 (POM); Pinyon Flat, 1/2 way between Palm Canyon Drive and Jeraboa Road, 15 May 1992, Johnson 92-021 (BRY, RSA, WS). San Ber- nardino County: Sandy-rocky place at base of foot- hills of Little San Bernardino Mountains, south of Yucca Valley, 6 May 1957, Grant & Latimer 9986 (RSA); Cactus Flat, San Bernardino Mountains, 25 Jun 1926, Munz 10514 (POM 96484 in part); Gran- WEESE AND JOHNSON: SALTUGILIA LATIMERI 199 ite Mountains, Mojave Desert, 13 May 1939, Jae- ger s.n. (CAS, POM); About 3 miles East of Joshua Tree off road to Twentynine Palms, 3 May 1964, Thorne 33975 (RSA). Comparative morphology and relationships. Un- til now, Saltugilia latimeri has been regarded as conspecific with S. australis. Both species possess small flowers, tend toward a smaller stature, and are ecologically disposed to drier habitats relative to S. grinnellii and S. caruifolia. Because conver- gence in these features may result from selection imposed by similar habitat types and available pol- linators, they are not necessarily useful indicators of phylogenetic affinities or taxonomic boundaries. Flower size, in particular, has received dispropor- tionate emphasis as an identification tool in Saltu- gilia (Day 1993). As a result, small flowered rep- resentatives of S. caruifolia and S. grinnellii have been misidentified as S. australis (e.g., CAS 627511, POM 48851). Similarly, it seems likely that S. latimeri has escaped previous delimitation by over reliance on flower size, to the exclusion of other characters that clearly distinguish S. latimeri from S. australis. Beyond flower size, S. latimeri differs from S. australis in other floral features (Table 1). The co- rolla of S. australis is whitish overall. The lobes are cusped at the apex and may be suffused with pink or bluish-lavender, but the tube and throat (ex- clusive of the yellow spots, characteristic of Sal- tugilia) remain white. Furthermore, the tube 1s more or less included within the calyx, the throat flares widely, and the calyx is mostly glabrescent at ma- turity. These features are consistent with the type (UC!) and populations examined from throughout S. australis’ range. In contrast, S. /atimeri has sat- urated pink to lavender, tapering acute corolla lobes, a colored and more narrowly funnelform throat, a purple, exserted tube, and a more intensely glandular calyx, with the glands large and as abun- dant as on the pedicel. Although corolla coloration provides a definitive means of distinguishing S. latimeri from S. aus- tralis on fresh and most herbarium specimens, color may fade from Saltugilia flowers over time and can be a less reliable diagnostic character on some her- barium specimens of extreme age. On living plants, flowers senesce quickly (l—2 days after opening, personal observation), and there is insufficient time for the distinctive coloration to fade while the flow- ers remain on the plant, thus identification of S. latimeri relative to S. australis is straightforward. When color has faded on herbarium specimens, the exserted tube, narrower throat, and glandular calyx are useful secondary characteristics for distinguish- ing between S. /atimeri and S. australis. Relative to other Saltugilia, S. latimeri is readily distinguished from S. caruifolia in stamen features (short filaments inserted in the sinus of the corolla lobes vs. long exserted stamens inserted mid- 200 MADRONO [Vol. 48 050m (E-G, R) O01 mm (H-L) 50 ym (M,N) 100um (O) 10 um (P) 10 um (Q) | 2001] | Saltugilia australis | WEESE AND JOHNSON: SALTUGILIA LATIMERI Saltugilia carutfolia TABLE 1. MORPHOLOGICAL CHARACTERS OF SALTUGILIA SPECIES. Saltugilia grinnellii Saltugilia latimeri corolla lobe color white/pale pink lavender/blue lavender/bright pink/lavender | pink throat color white white/lavender lavender/pink pink/lavender tube color white purple purple/pink purple petal lobe shape cusped obtuse/rounded obtuse acute, tapering tube position relative to included exserted exserted exserted calyx expansion of throat out- flares widely wide flare/cam- narrow flare narrow flare side calyx panulate stamen insertion sinus of corolla mid throat sinus of corolla sinus of corolla lobes lobes lobes average number of glands 3 (0-7) 4 (O-11) 7 (O—27) 21 (6-35) per calyx lobe (min.— max. observed) habit one main central leader leader one main central one main central branches near base leader iI I throat), and from S. grinnellii that tends to have rounder corolla lobe apices and a minutely glan- dular exterior corolla (Table 1). Saltugilia latimeri corollas most nearly approach the relative propor- tions of S. grinnellii corollas, although without the minute glands mentioned above. As implied by morphology, Saltugilia are closely allied species differing primarily in corolla features. In addition to these floral characters, the capsule in S. latimeri is more or less equal to the calyx, where- as it frequently (but not always) greatly exceeds the calyx in S. australis. Differing from all other Sal- tugilia, S. latimeri also typically branches near the base so that several sub-equal lateral branches soon approach the central stem in length and diameter. In contrast, other Saltugilia usually have dominant central axes. Although the basal leaves of S. lati- meri have broader, and fewer, lobe segments than typical S. australis, the leaves of S. australis show considerable variation across its range with some specimens possessing broader lobes. Close relationships among Saltugilia species are also evidenced by hybridization studies. Grant and Grant (1954) concluded that species now placed in Saltugilia are interfertile, although these studies re- vealed that sterility to partial sterility exists be- tween some populations (from 5 to 28% inviable individuals following interspecific crosses), leading Grant and Grant to hypothesize the presence of semi-lethal gene combinations (see also Latimer 1958). Of relevance here, a cross between S. Jati- meri and S. australis produced viable F1 with pol- len viability of 11% (Latimer 1958), compared to 58% viable pollen in F1 progeny of a cross between two populations of S. latimeri. This reduced fertility in interspecific crosses of S. latimeri * S. australis parallels our own results. A cross of S. latimeri (Johnson 92-021) with S. australis (Johnson 92- 011, BRY) produced vigorous Fl offspring, but even hand pollination of the Fl produced only | to 7 seeds per capsule (average = 2.7; average number of ovules per capsule = 25) compared to an average of 23 seeds (24 ovules) and 19 seeds (21 ovules) in self pollinated capsules of the parental S. latimeri and S. australis, respectively. Pollen viability in hy- brid progeny from other crosses reported by Grant and Grant (1954) averaged 14% between S. carui- folia and both subspecies of S. grinnellii, and 12% between extreme forms of the S. grinnellii. The ability of S. latimeri and S. australis to cross with S. grinnellii differed substantially (Latimer 1958). These hybridization studies provide inconclusive evidence regarding species boundaries and sister re- lationships within Saltugilia, but they do indicate levels of differentiation between S. /atimeri and S. australis at least as great as between pair-wise com- binations of other Saltugilia species. Were Saltugilia latimeri itself the product of hy- — Fic. 1. Features of Saltugilia latimeri alone and in comparison with selected features of other Saltugilia. A. habit in early flower. B. inflorescence branch in mid flower. C. basal leaf. D. cauline leaf. E, F flower. G. capsule. H—L. trichomes from living plant. H. basal leaf. I. Lower stem. J. mid stem (gland on right with secretion present). K. flower bract (secretion present). L. calyx (secretion present). M. calyx gland cleared in choral hydrate. N. paraffin section of calyx gland. O. scanning electron micrograph of unhydrated seed. P. scanning electron micrograph of unhydrated seed surface showing verrucae and ridges formed from anticlinal walls. Q. scanning electron micrograph of untreated pollen grain. R. capsule of S. australis. S—W. relative size and shape of Saltugilia flowers. S. S. latimeri. T. S. australis (Johnson 97-049, BRY). U. S. caruifolia (Johnson 97-099, BRY). V. S. grinnellii subspecies grinnellii (Johnson 93-098, BRY). W. S. grinnellii subspecies grantii (Johnson 94-057, BRY). A, Q. Grant & Grant 8840, RSA. B-P, S. progeny of Johnson 92-021, BRY. 202 bridization, progeny of self-pollinated plants should demonstrate character segregation. In each genera- tion, less intermediacy and more individuals with parental phenotypes would be observed. Four gen- erations of greenhouse grown progeny from a sin- gle population have produced no segregation of morphological characters, supporting the autonomy of this species. Additionally, greenhouse grown plants readily set full capsules of seeds upon self- ing, suggesting that Saltugilia latimeri is autoga- mous. Further evidence for autogamy is provided by allozyme analyses (26 individuals; Johnson 92- 021) that reveal complete homozygosity at 28 sur- veyed putative loci representing 19 enzyme sys- tems: AAT, ACN, ACPH, ALD, BGAL, CAT, EST, GDH, G3PDH-1, G3PDH-2, IDH, MDH-1, MDH- 2, MDH-3, ME-1, ME-2, MNR-1, MNR-2, 6PGD- 1, 6PGD-2, PGI-—1, PGM-1, PGM-2, PGI-2, SKDH, SOD, TPI-1, TPI-2 (Following methods of Soltis et al. 1983 with or without minor modifica- tions). Six of these 28 loci showed polymorphism with at most two alleles. Sequences of the chloroplast matK region from S. latimeri (Johnson 92-021) have been included in molecular studies (as Gilia australis; Johnson and Soltis 1995, Johnson et al. 1996) and analyses of these data provide strong support for placing S. la- timeri with other sampled representatives of Saltu- gilia. Additional sampling of multiple populations of all Saltugilia taxa (Weese and Johnson unpub- lished data) with sequences of the chloroplast trnL intron—trnF spacer and nuclear ITS regions strongly support the monophyly of Saltugilia, but provide insufficient resolution of taxa within the genus to use these markers alone as the basis of species rec- ognition. Grant and Grant (1954) proposed that species now included in Saltugilia evolved from an ances- tral type, represented by S. grinnellii or a similar large flowered ancestor, occupying mild climatic conditions similar to those found along the Cali- fornia coast. Arising from this ancestral type were taxa adapted to extreme conditions, including the dry, arid habitats occupied by S. australis. This im- plies that S. australis represents an autogamous de- rivative from S. grinnellii or another similar large flowered ancestor. This hypothesis applies equally well to S. latimeri. We suggest S. latimeri repre- sents a second, independent lineage of small-flow- ered, autogamous desert annuals derived from an ancestral Saltugilia species with features generally approaching those of S. grinnellii. Ecology and distribution. Saltugilia latimeri grows in dry, desert slopes of coarse sandy to rocky soils at elevations of 400-1900 m. Flowering times typically range from mid April to early June, al- though collections have been made as early as mid March and as late as the end of June. Our survey of herbaria (CAS, JEPS, POM, RSA, SD, UCR, UC) reveals that fewer than 20 independent collec- MADRONO [Vol. 48 tions of S. latimeri exist. These were obtained from the Joshua Tree Wilderness Area at the border be-. tween San Bernardino and Riverside Counties, in the foothills of the Little San Bernardino, San Ber- nardino, and Santa Rosa Mountains, in the Granite Mountains of central San Bernardino County, and one disjunct collection in the Panamint Mountains of Inyo County (Fig. 2). This latter population is approximately 120 miles from other known S. la- timeri populations. Several collections of S. latimeri from the Gran-| ite Mountain range pose some taxonomic difficulty because they possess a central leader habit and cus- pidate corolla lobes—two features typical of S. aus- tralis, but not characteristic of the holotype and paratypes listed above for S. latimeri. However, these collections fall within the limits of S. latimeri and can be identified based on their exserted, purple corolla tube, narrow corolla throat, capsule that is subequal to the calyx, and propensity for glandular calyces. The Granite and adjacent Old Dad Moun- tains (Fig. 2) represent a locality for Saltugilia sep- arated by 50 miles from the nearest portion of Sal- tugilia’s range in the desert slopes of the San Ber- nardino and Santa Rosa Mountains. Specimens re- ferred to as S. grinnellii, S. australis, and S. latimeri have all been collected in the Granite Mountains, but only infrequently. Further study of this material, including new accessions, is warranted. The pres- ence of these minimally intermediate S. latimeri collections in a geographically restricted location in no way interferes with the ability to differentiate S. latimeri from S. australis in other portions of these species’ ranges. In addition to the Granite Mountain area, S. latimeri is in close proximity to S. australis in the Santa Ana and Little San Bernardino Moun- tains, and with S. grinnellii in the Morongo Canyon area of the Littlke San Bernardino Mountains (Fig. 2). Mixed collections of S. latimeri with other Sal- tugilia on herbarium sheets have not been ob- served. Inferred from the small number of specimens present among Saltugilia collections acquired from seven herbaria (CAS, JEPS, POM, RSA, SD, UCR, UC), S. latimeri may merit conservation concern. Though almost certainly restricted in distribution, it is not known whether this species is truly rare or simply overlooked by collectors because of its small stature, inconspicuous habit, and short flow- ering duration. The habitat of S. /atimeri includes desert scrub and chaparral communities that can be dense and difficult to navigate, perhaps leading to under-representation of this species in herbaria. However, more extensive collections of S. australis are available in several of these same areas, sug- gesting that S. /atimeri is less abundant in nature. With human development encroaching in some of the known locations for this species, further study is warranted to establish the rarity of S. latimeri. Whereas S. australis is frequently collected on first year burns within its range and its seeds re- 2001] @ S. latimeri + S. australis A S. grinnellii Cl S. caruifolia 100 km Fic. 2. WEESE AND JOHNSON: SALTUGILIA LATIMERI 208 Distribution of Saltugilia species in California, USA. The ranges of Saltugilia australis and S. caruifolia extend into Baja Mexico, but these populations are not shown. Populations are represented by symbols: @ Saltugilia latimeri. + Saltugilia australis. A Saltugilia grinnellii. (] Saltugilia caruifolia. Locations mentioned in the text are indicated by numbers: 1. Little San Bernardino Mountains. 2. San Bernardino Mountains. 3. Santa Rosa Mountains. 4. Old Dad/Granite Mountains. 5. Panamint Mountains. 6. Santa Ana Mountains. 7. Cushenbury Grade. spond positively to charcoal as a germination cue (Keeley and Keeley 1987), no mention of fire is made on any specimen labels for S. latimeri. We routinely add charcoal when germinating seeds of all Saltugilia with great success, but have not rig- orously determined whether S. latimeri responds more positively to this treatment than it would without the addition of charcoal. Taxonomic evaluation of earlier studies. Because Saltugilia latimeri previously has not been distin- guished from S. australis, earlier literature may in- clude either of these species under the name “*Gilia australis’’. Johnson and Soltis (1995) and Johnson et al. (1996) both include, as Gilia australis (John- son 92-021), a population of S. latimeri in compar- ative DNA sequencing studies of the chloroplast matK gene. In both studies, Saltugilia is incom- pletely sampled (represented by S. latimeri, S. ca- ruifolia and S. grinnellii ssp. grantii in Johnson and Soltis 1995 and S. latimeri and S. grinnellii ssp. grantii in Johnson et al. 1996), yet forms a well- supported clade distinct from Gilia. In their initial circumscription of Gilia section Saltugilia, Grant and Grant (1954) performed chro- mosome counts as well as extensive crossing stud- ies within and among species of Saltugilia. The population of G. australis collected by Grant from Morongo Canyon in the San Bernardino Mountains (Fig. 2) used for these studies, and cited by Grant (1956), is S. latimeri. Additionally, specimens cited under the description of G. australis by Grant and Grant (1954) include both S. australis and S. lati- meri. This is significant because the “‘desert race”’ of Gilia splendens discussed by Grant and Grant (1965) could be misconstrued to be our S. latimeri. This clearly is not the case, however, because this desert race of G. splendens is also discussed in Grant and Grant (1954), Latimer (1958) and Grant (1966). Latimer’s (1958) unpublished thesis recognized 204 S. latimeri at the subspecies level (as G. australis ssp. desertorum). In Grant’s (1966) citation of Lat- imer’s work, the populations of Gilia australis con- sidered sympatric with G. splendens (collected in the Morongo Valley and Cushenbury Grade of the San Bernardino Mountains) are in fact S. latimeri, while those populations considered allopatric with G. splendens (collected inthe Santa Ana Mountains and Cajon Pass) are S. australis (Fig. 2). Based on these observations, the recognition of Saltugilia latimeri as a new species in Polemoni- aceae iS warranted. This recognition is a starting point for further species level studies of Saltugilia latimeri and larger scale investigations both within Saltugilia and Polemoniaceae. To facilitate the cor- rect identification and incorporation of S. latimeri in future works, the following key is provided. KEY TO GENUS SALTUGILIA (BEGINNING WITH COUPLET 55 OF PORTER AND JOHN- SON 2000) 1. Trichomes on basal leaves eglandular white arach- niod (present at least in axils), or white geniculate, or translucent; if translucent, inflorescence glands long stalked, diameter of gland less than length of stalk and generally equally abundant on pedicel and calyx, or inflorescence trichomes including Vome: Darestye. ct eae oe eee ear, eee ad ee ae Gilia 1.’ Trichomes on basal leaves translucent, minutely gland tipped; inflorescence glands subsessile, di- ameter of gland > length of stalk, more abundant on pedicel than calyx; pedicels and calyx without lone stalked ams = a ee ee 2 (Saltugilia) 2. Corolla tube and throat white (throat yellow spot- ted), adaxial lobe surface white to white suffused with blue, lavender, or pink; corolla tube included in calyx, corolla lobes cuspidate ..... S. australis | Yh 2.’ Corolla tube and throat pigmented (magenta to pink to purple and throat yellow spotted), adaxial lobe surface saturated blue, lavender, or pink; co- rolla tube exserted from calyx, corolla lobes round- ed to acute (cuspidate only in Granite Mountains of San Bernardino County, California, but then possessing pigmented and exserted corolla tube) 3. Stamens exserted well beyond the corolla lobes, inserted equally mid throat; corolla lobes often re- flexed, throat widely flaring, nearly campanulate, with purple marks visible on the interior of the throat at the base of each corolla lobe — S. caruifolia 3." Stamens less than or equal to corolla lobes, insert- ed equally in the sinus of the corolla lobes; corolla lobes not reflexed, throat narrowly flaring, lacking purple marks on interior of throat 4. Corolla less than 11 mm long, corolla lobes pale pink to lavendar, tube purple; corolla tube eglan- dular and’calyx slandular ~ .22...55. «2: S. latimeri 4.’ Corolla generally more than 11 mm long, corolla lobes pale to bright pink, tube pink to magenta; corolla tube minutely glandular, calyx eglandular to glabrescent 5 (S. grinnellit) 5. Corolla tube 4-10 mm long, I—2 X calyx Se ee CA UR ey: subspecies grinnellii MADRONO [Vol. 48 5.’ Corolla tube 7-18 mm long, 2-5 X calyx ..... subspecies grantii ACKNOWLEDGEMENTS We thank the herbarium staff at CAS, JEPS, POM, RSA, SD, UCR, and UC for loans of herbarium material, J.M. Porter for discussion, G. Baird for proofreading and improving the Latin diagnosis, the Botany Department at North Carolina State University and the Botany and Range Science Department at BYU for supporting re- search that aided in this contribution, and two anonymous reviewers for their helpful comments. The population used | in our greenhouse studies (Johnson 92-021) was obtained from a collecting trip supported by a grant from the Cal- ifornia Native Plant Society, and the enzyme electropho- retic survey funded by NSF grant DEB 9321788 to D. E. Soltis and LAJ. LITERATURE CITED BRAND, A. 1907. Polemoniaceae. Pp. 1—203 in A. Engler (ed.), Das pflanzenreich [V. 250. Engelmann, Leipzig. Day, A. G. 1993. Gilia. Pp. 828-836 in J. Hickman, (ed.), The Jepson manual: higher plants of California. Uni- versity of California Press, Berkeley, CA. GRANT, V. 1956. The genetic structure of races and species in Gilia. Advances in Genetics 8:55-—87. . 1966. The selective origin of incompatibility bar- riers in the plant genus Gilia. American Naturalist 100:99-118. AND A. GRANT. 1954. Genetic and taxonomic studies in Gilia. VII. The woodland gilias. Aliso 3: 59-91. AND K. GRANT. 1965. Flower pollination in the phlox family. Columbia University Press, New York. JOHNSON, L. A. AND D. E. SOLtTis. 1995. Phylogenetic in- ference in Saxifragaceae sensu stricto and Gilia (Po- lemoniaceae) using matK sequences. Annals of the Missouri Botanical Gardens 82:149—-175. JOHNSON, L.A. J. L. ScHuirz, D. EY SoOrtis; AND P'S: SoLTis. 1996. Monophyly and generic relationships of Polemoniaceae based on matK sequences. American Journal of Botany 83:1207—1224. KEELEY, J. E. AND S. C. KEELEY. 1987. Role of fire in the germination of chaparral herbs and _ suffrutescents. Madrono 34:240—249. Latimer, H. L. 1958. A study of the breeding barriers between Gilia australis and Gilia splendens. Unpub- lished Ph.D. dissertation, Claremont Graduate School, Claremont, CA, USA. Mason, H. L. AND A. D. GRANT. 1948. Some problems in the genus Gilia. Madrono 9:201—220. PorTER, J. M. AND L. A. JOHNSON. 2000. A phylogenetic classification of Polemoniaceae. Aliso 19: 55-91. SoLtis, D. E., C. H. HAUFLER, G. J. GASTONY, AND D. C. Darrow. 1983. Starch-gel electrophoresis of ferns: a compilation of grinding buffers, gel and electrode buffers, and staining schedules. American Fern Jour- nal 73:9—27. WEESE, T. L. AND L. A. JOHNSON. 2000. Taxonomic limits and phylogenetic affinities of Saltugilia: molecular and morphological support for generic recognition apart from Gilia. American Journal of Botany, 87 (supplement):166. [Abstract. | Wuerry, E. T. 1940. A provisional key to the Polemon- iaceae. Bartonia 20:14—17. Maprono, Vol. 48, No. 3, pp. 205-210, 2001 TAXONOMIC CHANGES AND A NEW SPECIES IN LASTHENIA SECT. AMPHIACHAENIA (COMPOSITAE: HELIANTHEAE SENSU LATO) RAYMUND CHAN Jepson Herbarium and Department of Integrative Biology, University of California, 1001 Valley Life Sciences Building #2465, Berkeley, CA 94720-2465 ABSTRACT | A molecular phylogenetic study of the goldfield genus Lasthenia has clarified the relationships among taxa in the group. It has become desirable to make taxonomic and nomenclatural changes in order to reflect the relationships. Lasthenia sect. Amphiachaenia is the correct name for L. sect. Baeria sensu Ornduff (1966). The section now comprises six species and subspecies. In consequence of a molecular phylogenetic analysis of the goldfield genus Lasthenia (Chan 2000; Chan et al. 2001), it has become desirable to revise the taxonomy of L. sect. Baeria sensu Orn- duff (1966) (see Table 1 for a summary) to reflect the relationships of taxa within. The study utilized DNA sequence data from the internal and external transcribed spacers of 18S—26S nuclear ribosomal DNA and the 3’ ¢rnK intron of chloroplast DNA from multiple populations of taxa in L. sect. Baeria sensu Ornduff (1966) and close relatives. Relation- ships shown by molecular data are supported by morphology, chromosome numbers, and flavonoid chemistry (Bohm et al. 1974; Ornduff et al. 1974). Ornduff (1966) included Burrielia hirsuta in his circumscription of L. sect. Baeria sensu Ornduff (1966) and did not account for the earlier sectional name Burrielia DC. sect. Amphiachaenia Nutt. The older sectional name is used here: Lasthenia Cass. sect. Amphiachaenia (Nutt.) R. Chan, comb. nov. = Burrielia DC. sect. Amphia- chaenia Nutt., Trans. Amer. Phil. Soc. n.s. 7: 381. 1841.—TYPE SPECIES: Burrielia hirsuta Nutt. = Lasthenia gracilis (DC.) Greene. Baeria Fisch. & Mey., Index Sem Hort. Petrop. 2: 29s 1856, linnaca, MiGente. (Berm): 96. 1837.4 — Lasthenia Cass. sect. Baeria (Fisch. & Mey.) Ornduff, Univ. Calif. Publ. Bot. 40: 56. 1966, nom. illegit—TYPE SPECIES: Baeria chrysos- toma Fisch. & Mey. = Lasthenia californica DC. ex Lindl. Leaves entire. Involucres bell-shaped, depressed- hemispheric, hemispheric, or obconic. Phyllaries free. Disc florets 10—100+, corollas 5-lobed, floral pigments turning bright red in dilute alkali. Anther tips deltate or sublanceolate to subulate, broadened above the base. Cypselae pappose or epappose; pappi monomorphic. 2n = 16, 32, 48. Maximum parsimony analysis has yielded results showing that L. sect. Amphiachaenia is monophy- letic only with the inclusion of L. leptalea from L. sect. Burrielia. More importantly, this study shows that L. californica (sensu Johnson and Ornduff 1978 and sensu Ornduff 1993; both = L. chrysos- toma sensu Ornduff 1966) is cryptically diverse and is resolved into two robustly supported clades. One clade includes L. macrantha pro parte; the other may be sister to all other members of L. sect. Am- phiachaenia. Each clade can be morphologically diagnosed by pappus morphology (see Fig. 1) and has a partially distinct distribution. No other mor- phological characters were found that could reliably distinguish the two clades. Although Rajakaruna and Bohm (1999) concluded that two geographical races of L. californica sensu Ornduff (1993) cor- respond to two edaphically distinct groups, I found that members of both clades span the habitat types to which the edaphic races appear to be restricted (N. Rajakaruna pers. comm.). In light of this, I have recognized each clade of L. californica sensu stric- to as different species in L. sect. Amphiachaenia and have included L. /eptalea in the section. The smallest clade comprising L. macrantha also includes one clade of L. californica sensu Ornduff (1993). Lasthenia macrantha subsp. bakeri and L. macrantha subsp. macrantha and this clade of L. californica sensu Ornduff (1993) form a monophy- letic group with L. macrantha subsp. prisca emerg- ing as sister to this robustly supported clade. These results support previous interpretations of a close relationship between L. californica sensu Ornduft (1993) and L. macrantha. Close relationship be- tween these two species was first recognized by Gray (1857), who treated them as varieties of the same species. Ornduff (1966) discussed the rela- tionships in detail and Ornduff (1971) stated that L. californica (sensu Ornduff 1993) “‘may be viewed as an annual version of L. macrantha.”’ The close relationship between a clade of L. cal- ifornica sensu Ornduff (1993), L. macrantha subsp. bakeri, and L. macrantha subsp. macrantha should now be reflected in their taxonomy by treating these 206 MADRONO [Vol. 48 TABLE 1. COMPARISON BETWEEN LASTHENIA SECTS. BAERIA AND BURRIELIA SENSU ORNDUFF (1966, 1971, 1993) AND THE REVISED TAXONOMY. Corresponding species and subspecies are shown on the same line. *—sect. Amphiachaenia is the correct name for sect. Baeria sensu Ornduff (1966). **—L. gracilis was previusly recognized as part of L. californica sensu Ornduff (1966, 1993). Lasthenia Cass. sensu Ornduff (1966, 1971, 1993) sect. Baeria (Fisch. & Mey.) Ornduff L. californica DC. ex Lindl. L. macrantha (A. Gray) Greene subsp. macrantha subsp. bakeri (J. T. Howell) Ornduff subsp. prisca Ornduff sect. Burrielia (DC.) Ornduff L. leptalea (A. Gray) Ornduff L. debilis (Greene ex A. Gray) Ornduff L. microglossa (DC.) Greene Lasthenia Cass. | sect. Amphiachaenia (DC.) R. Chan* L. gracilis (DC.) Greene** L. californica DC. ex Lindl. subsp. californica subsp. macrantha (A. Gray) R. Chan subsp. bakeri (J. T. Howell) R. Chan L. ornduffii R. Chan L. leptalea (A. Gray) Ornduff sect. Burrielia (DC.) Ornduff L. debilis (Greene ex A. Gray) Ornduff L. microglossa (DC.) Greene three taxa as members of L. californica sensu stric- to. 1 am provisionally continuing to recognize these three taxa (as subspecies) pending further studies to assess relationships in the clade. The molecular results, coupled with differences in morphology (D. Keil pers. comm.) also suggest that L. macrantha subsp. bakeri and L. macrantha subsp. macrantha populations at the southernmost parts of their rang- es may warrant treatment as distinct taxa in their own right. Fic. 1. I have assigned L. macrantha subsp. prisca to species rank as L. ornduffii based on chromosomal, morphological, biogeographical, physiological, and molecular evidence (Chan 2000; Chan et al. 2001: Ornduff 1971). Lasthenia macrantha subsp. prisca is a tetraploid Oregonian endemic with fleshy roots and narrower leaves compared to the hexaploid Californian subspecies with tap roots and broader leaves. Ornduff (1966) also reported that the Ore- gon populations of L. macrantha are long-lived pe- Different pappus types present in L. californica subsp. californica and L. gracilis. A—clear to brown linear awns (L. californica subsp. californica only); B—opaque, white, ovate-lanceolate scales, each tapering to an awn (ZL. gracilis only); C—clear to brown subulate awns (L. californica subsp. californica only); D—epappose (both L. cali- fornica subsp. californica and L. gracilis). Dimensions are given in descriptions in the text. 2001] rennials whereas their Californian counterparts may flower during the first year and behave as annuals under prolonged drought conditions. Key TO LASTHENIA SECT. AMPHIACHAENIA 1. Plants annual; coastal or inland. 2. Stems glabrous proximally; receptacles narrowly conic; phyllaries usually 3—6, glabrous except at tips; anther tips subulate; interior western Cali- fornia (Monterey and San Luis Obispo counties) nr eee eee 3. L. leptalea 2. Stems pubescent; receptacles conical; phyllaries usually 6-13, pubescent; anther tips deltate; coastal or inland. 3. Pappi, when present, of 1—7, clear to brown, linear to subulate awns (see Figs. 1A and 1C); northern California and Oregon la. L. californica subsp. californica 3. Pappi, when present, of (2—)4(—6) opaque, white, ovate-lanceolate scales, each tapering to an awn (see Fig. 1B); California (including the Channel Islands), central Arizona, and Baja California (including Guadalupe Island) 2. L. gracilis 1. Plants perennial; coastal. 4. Roots fleshy, clustered; stems erect, Q—2+- branched; leaves usually 1-2 mm wide; Califor- Tall Meee eens ag lb. L. californica subsp. bakeri 4. Roots not fleshy, not clustered; stems decum- bent, usually O—5+-branched at base; leaves usually 2—5.5 mm wide; California or Oregon. 5. Leaves 28-88 mm long; laminae of ray co- rollas 6-18 mm long; California lc. L. californica subsp. macrantha 5. Leaves usually 20—40 mm long; laminae of ray corollas 5-9 mm long; southern Oregon 4. L. ornduffii eo 6 © © © © 8 eo 6 ee © © © « «© © @ © © © © © © © © © © © «6 1. Lasthenia californica DC. ex Lindl., Edwards’s Bot. Reg. 21: facing pl. 1780. 1835.—TYPE: ‘‘HHS [Hort. Horticultural Society of London],”’ grown in London from seeds collected in Cali- fornia by David Douglas, J. Lindley s.n. (holo- type: CGE!). Plants annual or perennial, 5-40 cm. Roots fi- brous, from taproot or fleshy, clustered. Stems erect or decumbent, simple or 1—5(—20+)-branched, + pubescent or hirsute. Leaves linear to oblanceolate, 8—210 mm long, 1—5.5(—15) mm wide, entire or with 3—5+ short, lateral teeth, + fleshy in coastal forms, glabrous or sparsely to densely pubescent, or + hirsute. Involucres bell-shaped, depressed- hemispheric, or hemispheric. Phyllaries 4—16, 5—14 mm long, pubescent. Receptacles conic, muricate, usually glabrous. Ray florets 6—16, laminae of ray corollas 5-18 mm long. Anther tips deltate to sub- lanceolate. Style tips deltate with apical tufts of hairs and subapical fringe of shorter hairs. Cypselae linear to + club-shaped, 2—4 mm long, glabrous or pubescent, pappose or epappose; pappi of 1-7, clear to brown, linear to subulate awns. 2n = 16, 32, 48. CHAN: LASTHENIA SECT. AMPHIACHAENIA 207 la. Lasthenia californica DC. ex Lindl. subsp. cal- ifornica Baeria chrysostoma Fisch. & Mey., Index Sem. Hort. Petrop. 2329; 1836; Linnaea 1 )Guitt) Ber): 96. 1837. = Burrielia chrysostoma (Fisch. & Mey.) Torr. & A. Gray, Fl. N. Amer. 2: 379 1842. = Lasthenia chrysostoma (Fisch. & Mey.) Greene, Man. Bot. San Francisco 205. 1894.— LECTOTYPE (Ornduff, 1966, p. 57): California, Sonoma Co., vicinity of present-day Fort Ross, protologue: “‘Hab. circa coloniam Ruthenorum Ross, in sinu Bodega, Nova California,” 1832, collector unknown (L; isolectotype: BM!). Baeria gracilis A. Gray var. aristosa A. Gray, Proc. Amer. Acad. Arts 19: 21. 1883. = Baeria aris- tosa (A. Gray) Howell, Fl. N. W. Amer. 1: 354. 1900.—TYPE: illustration in Bot. Mag. 66(13 n.s.): pl. 3758. 1840. Lasthenia hirsutula Greene, Man. Bot. San Fran- cisco 206. 1894. = Baeria hirsutula (Greene) Greene, Fl. Fran. 438. 1897. = Baeria chrysos- toma Fisch. & Mey. subsp. hirsutula (Greene) Ferris, Contr. Dudley Herb. 5: 99. 1958.—LEC- TOTYPE (Ornduff, 1966, p. 57): California, Monterey Co., Pt. Lobos, protologue: “‘Along the seacoast from Marin Co. southward,”’ | Jul 1891, E. L. Greene s.n. (ND!). Plants annual. Roots fibrous, from taproot. Stems erect or decumbent, simple or 1-—6(—10+)- branched, + hirsute. Leaves 8-70 mm long, 1-3 mm wide, + hirsute. Involucres bell-shaped or hemispheric. Phyllaries 4-13, 5-10 mm long. Ray florets 6-13, laminae of ray corollas 5—10 mm long. Anther tips deltate. Style tips deltate with apical tufts of hairs and subapical fringe of shorter hairs. Cypselae + club-shaped, 2—3 mm long, glabrous or pubescent, pappose or epappose; pappi of 1-7, clear to brown, linear to subulate awns. 2n = 16, 32, 48. Lasthenia californica subsp. californica grows in a variety of habitats in southwestern Oregon and northern California. Within California, L. califor- nica subsp. californica is found from northern Monterey, Santa Clara, Merced, and Madera coun- ties northwards. Lasthenia californica subsp. cali- fornica was previously circumscribed together with L. gracilis as L. californica sensu Ornduff (1993). Lasthenia californica subsp. californica is morpho- logically very similar to L. gracilis but it can be distinguished from L. gracilis by its clear to brown, linear to subulate pappus awns (see Fig. 1A and 1C) and more northern distribution. In sympatric populations, epappose plants cannot be easily dis- tinguished morphologically. Elevation O—1,500 m. Flowering Feb—Jun. Lasthenia hirsutula is a polyphyletic taxon com- prising the maritime populations of Baeria chry- sostoma. Plants matching the pappus morphology of both L. californica subsp. californica and L. gracilis are known to have been included in the 208 original circumscription of L. hirsutula (D. Keil pers. comm.). The type specimen of L. hirsutula possesses the pappus morphology of L. californica subsp. californica and is thus included here. lb. Lasthenia californica DC. ex Lindl. subsp. bakeri (J. T. Howell) R. Chan, comb. nov. = Baeria bakeri J. T. Howell, Leafl. W. Bot. 1: 7. 1932. = Baeria macrantha (A. Gray) A. Gray var. bakeri (J. T. Howell) Keck, Aliso 4: 101. 1958. = Lasthenia macrantha (A. Gray) A. Gray subsp. bakeri (J. T. Howell) Ornduff, Univ. Calif. Publ. Bot. 40: 62. 1966.—TYPE: California, Mendocino Co., Pt. Arena, protologue: “‘mead- Owy opening in the forest on the coastal plain, six miles south of Pt. Arena,’ 26 Jun 1931, M. S. Baker 5283 (holotype: CAS; isotype: US). Plants perennial (rarely annual or flowering first year). Roots fleshy, clustered. Stems erect, simple or 1—2(—4+)-branched, = pubescent. Leaves 20— 210 mm long, 1—2 mm wide, glabrous or sparsely to densely pubescent, basally clustered. Involucres bell-shaped to depressed-hemispheric. Phyllaries 13-16, 9-14 mm long. Ray florets 8—16, laminae of ray corollas 5-16 mm long. Anther tips deltate to sublanceolate. Style tips deltate with apical tufts of hairs and subapical fringe of shorter hairs. Cyp- selae linear to narrowly club-shaped, 2—4 mm long, usually glabrous, pappose or epappose; pappi of 1— 4, clear to brown, subulate awns, variable or miss- ing in some florets of a head. 2n = 48. Lasthenia californica subsp. bakeri grows in grasslands and woods along the coast in Mendocino and Sonoma counties. Populations of this rare sub- species appear to be increasingly more difficult to locate because of habitat destruction. Elevation O-— 500 m. Flowering year round, mostly May—Jun. lc. Lasthenia californica DC. ex Lindl. subsp. ma- crantha (A. Gray) R. Chan, comb. nov. = Bur- rielia chrysostoma (Fisch. & Mey.) Torr. & A. Gray var. macrantha A. Gray in J. Torrey, Pacif. Railr. Rep. 4(5): 106. 1857. = Baeria macrantha (A. Gray) A. Gray, Proc. Amer. Acad. Arts 19: 21. 1883. = Lasthenia macrantha (A. Gray) Greene, Man. Bot. San Francisco 205. 1894. = Baeria macrantha (A. Gray) A. Gray var. littor- alis Jeps., nom. illegit., Man. Fl. Pl. Calif. 1112. 1925.—TYPE: California, Marin Co., Pt. Reyes, protologue: “‘Punta de los Rey[e]s,’’ Apr 1854, Bigelow s.n. (holotype: GH; isotypes: K, NY). Baeria macrantha (A. Gray) A. Gray var. pauciar- istata A. Gray, Proc. Amer. Acad. Arts 19: 21. 1883.—LECTOTYPE (Ornduff, 1966, p. 59): California, Mendocino Co., protologue: ‘“‘sea shore,”’ 4 Aug 1882, C. G. Pringle s.n. (GH; iso- lectotypes: K, NY). Baeria macrantha (A. Gray) A. Gray var. thalas- sophila J. T. Howell, Leafl. W. Bot. 5: 108. 1948.—TYPE: California, Marin Co., Dillons Beach, protologue: “‘on ocean bluffs just above MADRONO [Vol. 48 the high-tide line,” 30 Apr 1947, J. T. Howell 23108 (holotype: CAS; isotypes: UC!, US). Plants perennial (rarely annual or flowering first year). Roots fibrous, from taproot. Stems usually decumbent, simple or 1—5(—20+ )-branched at base, + pubescent. Leaves 28-88 mm long, 1.5—5.5(—15) mm wide, glabrous to densely pubescent. Involu- cres bell-shaped to depressed-hemispheric. Phyllar- ies 9-16, 9-14 mm long. Ray florets 8—16, laminae of ray corollas 6-18 mm long. Anther tips deltate to sublanceolate. Style tips deltate with apical tufts © of hairs and subapical fringe of shorter hairs. Cyp- selae linear to narrowly club-shaped, 2—4 mm long, — usually glabrous, pappose or epappose; pappi of 1— 4, clear to brown, subulate awns, variable or miss- ing in some florets of a head. 2n = 48. Lasthenia californica subsp. macrantha grows in grasslands or on dunes along the immediate coast in Humboldt, Mendocino, Sonoma, Marin, San Ma- teo, and San Luis Obispo counties, California. Las- thenia californica subsp. macrantha is morpholog- ically very similar to L. ornduffii; their ranges are allopatric. Elevation 0-500 m. Flowering year round, mostly May—Aug. 2. Lasthenia gracilis (DC.) Greene, Man. Bot. San Francisco 206. 1894. = Burrielia gracilis DC., Prodr. 5: 664. 1836. = Baeria gracilis (DC.) A. Gray, Proc. Amer. Acad. Arts 9: 196. 1874. = Baeria chrysostoma Fisch. & Mey. var. gracilis (DC.) H. M. Hall, Univ. Calif. Publ. Bot. 3: 170. 1907. = Baeria chrysostoma Fisch. & Mey. subsp. gracilis (DC.) Ferris, Contr. Dudley Herb. 5: 100. 1958.—TYPE: California, protologue: ‘In Nova-California legit cl. Douglas,’ D. Douglas s.n. (holotype: G!; isotypes: BM!, GH, Ke NY). Burrielia tenerrima DC., Prodr. 5: 664. 1836. = Baeria tenerrima (DC.) A. Gray, Proc. Amer. Acad. Arts 9: 196. 1874. = Baeria gracilis (DC.) A. Gray var. tenerrima (DC.) A. Gray, Syn. FI. N. Amer. 17: 326. 1884. = Baeria chrysostoma Fisch. & Mey. f. tenerrima (DC.) H. M. Hall, Univ. Calif. Publ. Bot. 3: 171. 1907.—TYPE: California, protologue: “‘In Nova-California legit cl. Douglas,’ D. Douglas s.n. (holotype: G!; is- otypes: BM!, K, NY). Burrielia hirsuta Nutt., Trans. Amer. Phil. Soc. n.s. 7. 381. 1841.—TYPE: California, Santa Barbara Co., protologue: “‘Hab. Santa Barbara,”’ 7. Nut- tall s.n. (holotype: BM!; isotype: GH). Burrielia longifolia Nutt., Trans. Amer. Phil. Soc. n.s. 7: 380. 1841.—TYPE: California, Santa Bar- bara Co., protologue: “‘near Santa Barbara,” 7. Nuttall s.n. (holotype: BM!). [This specimen is also the type of Baeria gracilis (DC.) A. Gray var. paleacea A. Gray.] Burrielia parviflora Nutt. Trans. Amer. Phil. Soc. n.s. 7: 381. 1841.—TYPE: California, Santa Bar- bara Co., protologue: “near Santa Barbara,” T. 2001] Nuttall s.n. (holotype: BM! [label states “*St. Di- ego’’]; isotypes: GH, NY). Baeria palmeri A. Gray, Bot. Calif. 1: 376. 1876. = Baeria chrysostoma Fisch. & Mey. var. pal- meri (A. Gray) J. T. Howell, Leafl. W. Bot. 3: 152. 1942.—LECTOTYPE (Ornduff, 1966, p. 57): Mexico, Guadalupe Island, 1875, E. Palmer 45 (PH; isolectotypes: BM, EK K, L, MBG, NY). Baeria gracilis (DC.) A. Gray var. paleacea A. Gray, Proc. Amer. Acad. Arts 19: 21. 1883. = Baeria chrysostoma Fisch. & Mey. f. paleacea (A. Gray) H. M. Hall, Univ. Calif. Publ. Bot. 3: 171. 1907.—LECTOTYPE (designated here): California, Santa Barbara Co., protologue: “‘near Santa Barbara,” 7. Nuttall s.n. (BM!). [This specimen is also the type of Burrielia longifolia Nutt. ] Baeria clevelandii A. Gray, Proc. Amer. Acad. Arts 19: 22. 1883.—TYPE: California, San Diego Co., protologue: “‘near San Diego,” 1874, D. Cleveland s.n. (holotype: GH). Baeria curta A. Gray, Proc. Amer. Acad. Arts 19: 21. 1883. = Baeria chrysostoma Fisch. & Mey. f. curta (A. Gray) H. M. Hall, Univ. Calif. Bot. Publ. Bot. 3: 172. 1907.—LECTOTYPE (Orn- duff, 1966, p. 57): California, San Bernardino Co., protologue: “‘near San Bernardino,”’ 1880, J. G. Lemmon 135, (GH; isolectotype: UC!). Baeria palmeri A. Gray var. clementina A. Gray, Syn. Fl. N. Amer. ed. 2, 17: 452. 1886. = Baeria chrysostoma Fisch. & Mey. f. clementina (A. Gray) H. M. Hall, Univ. Calif. Bot. Publ. Bot. 3: 172. 1907.—TYPE: California, San Clemente Is- land, 1885, J. C. Nevin and D. Lyon s.n. (holo- type: GH; isotypes: DS, ND). Baeria chrysostoma Fisch. & Mey. f. nuda H. M. Hall, Univ. Calif. Bot. Publ. Bot. 3: 170. 1907.— TYPE: California, Los Angeles Co., protologue: “San Francisquito Canon,’ 3 May 1902, H. M. Hall 3100 (UC!). Baeria chrysostoma Fisch. & Mey. f. crassa H. M. Hall, Univ. Calif. Bot. Publ. Bot. 3: 172. 1907.— TYPE: California, San Diego Co., protologue: ““Ocean Beach near San Diego,’? May 1906. K. Brandegee (holotype: UC!; isotypes: UC!, DS). Plants annual, 5—40 cm. Roots fibrous, from tap- root. Stems erect or decumbent, simple or 1—6(— 10+)-branched, = strigose. Leaves linear to oblan- ceolate, 8-70 mm long, 1-3 mm wide, entire or with 3—5+ short, lateral teeth, + fleshy in coastal forms, glabrous or + strigose. Involucres_ bell- shaped or hemispheric. Phyllaries 4-13, 5-10 mm long, = strigose. Receptacles conic, muricate, gla- brous. Ray florets 6—13, laminae of ray corollas 5— 10 mm long. Anther tips deltate. Style tips deltate with apical tufts of hairs and subapical fringe of shorter hairs. Cypselae + linear, 2-3 mm long, gla- brous or pubescent, pappose or epappose; pappi of 2—6, usually 4, opaque, white, ovate-lanceolate scales, each tapering to an awn. 2n = 16, 32. CHAN: LASTHENIA SECT. AMPHIACHAENIA 209 Keck (1959) said of L. gracilis (as Baeria chry- sostoma subsp. gracilis) ““The most abundant com- posite in the state [of California].”’ It grows in a wide variety of soils and habitats throughout Cali- fornia, central Arizona, the Channel Islands, Gua- dalupe Island, and Baja California. It was circum- scribed together with L. californica subsp. califor- nica as L. californica by Ornduff (1993). It is mor- phologically very similar to L. californica subsp. californica and to L. leptalea but can be distin- guished from L. californica subsp. californica and L. leptalea by its opaque, white, ovate-lanceolate scales, each tapering to an awn (see Fig. 1B). El- evation O—1,500 m. Flowering Feb—Jun. 3. Lasthenia leptalea (A. Gray) Ornduff, Univ. Calif. Publ. Bot. 40: 63. 1966. = Burrielia lep- talea A. Gray, Proc. Amer. Acad. Arts 6: 546. 1865. = Baeria leptalea (A. Gray) A. Gray, Syn. Fl. N. Amer. 17: 325. 1884.—TYPE: California, Monterey Co., Santa Lucia Mountains, proto- logue: “‘on very dry hillside along the Naciso- mento [Nacimiento] River,’ 2 May 1861, W. AH. Brewer 548 (holotype: GH; isotypes: K, UC!, US). Plants annual, 5—15 cm. Roots fibrous, from tap- root. Stems erect, simple or 1—5+-branched, gla- brous proximally, densely villous in peduncular re- gion. Leaves linear, 3—20 mm long, entire, sparsely pubescent. Involucres obconic to bell-shaped. Phyl- laries 3-6, 4—6 mm long, glabrous but for pubes- cent tips. Receptacles narrowly conic, glabrous. Ray florets 6—9, laminae of ray corollas 2.5—5 mm long. Anthers tips subulate. Style tips + deltate, with long apical pubescence. Cypselae linear to narrowly club-shaped, ca. 2 mm long, sparsely pu- bescent; pappi of 1—4, white to yellowish, narrowly tapered awns, missing in some florets of a head. 27 = 16. Lasthenia leptalea usually grows in open areas of oak woodlands, interior southern Monterey and northern San Luis Obispo counties, California. Las- thenia leptalea is morphologically very similar to L. gracilis; it can be consistently distinguished from L. gracilis by its subulate anther tips and phyl- laries pubescent at the tips. Elevation O—650 m. Flowering Feb—Apr. 4. Lasthenia ornduffii R. Chan, sp. nov. Based on: Lasthenia macrantha (A. Gray) Greene subsp. prisca Ornduff, Madrono 21: 96. 1971.—TYPE: Oregon, Curry Co., protologue: “‘very abundant on Cape Blanco,” 16 Jul 1929, L. F. Henderson 11400 (holotype: UC!; isotypes: ORE, PH). Plants perennial (rarely annual or flowering first year), 5-40 cm. Roots fibrous, from taproot. Stems usually decumbent, 1—3+-branched at base, + pu- bescent. Leaves linear to oblong, 20—40 mm long, 1.8-—3 mm wide, glabrous or densely pubescent. In- volucres bell-shaped to depressed-hemispheric. 210 Phyllaries 8-14, 9-14 mm long, pubescent. Recep- tacles conic, muricate, usually glabrous. Ray florets 8—15, laminae of ray corollas 5—9 mm long. Anther tips deltate to sublanceolate. Style tips deltate, hair- tufted. Cypselae linear to narrowly club-shaped, 2.5—4 mm long, usually glabrous, pappose or epap- pose; pappi of 1—4, clear to brown, subulate awns, often variable or missing in some florets of a head. Li 32) Lasthenia ornduffii is known from six or so pop- ulations in grasslands along the immediate coast in Curry Co., southern Oregon. Elevation 0-500 m. Flowering year round, mostly May—Aug. Lasthenia ornduffii was originally described as a tetraploid subspecies of L. macrantha. The species name is intended to honor the late Professor Emer- itus Robert Ornduff, a native Oregonian, in appre- ciation of his outstanding contributions to our un- derstanding of the evolution of Lasthenia and other groups in the California flora. Excluded names (L. californica sensu Ornduff 1993): Baeria punctata Greene ex C. FE Baker, W. Amer. Plants 2: 8. 1903, nom. nudum. Based on C. F. Baker 2962, Lakeport, Lake Co., California. Baeria subcilata Greene ex C. E Baker, W. Amer. Plants 2: 8. 1903, nom. nudum. Based on C. F. Baker 2857, Lake Merced, San Francisco Co., California. ACKNOWLEDGMENTS I thank Bruce G. Baldwin, David J. Keil, John L. Strother, John W. Taylor, and Michael C. Vasey for re- viewing the manuscript. Caroline Stromberg deserves spe- cial thanks for making the illustrations of cypselae. I also thank the curators and staff of these herbaria for assistance in making material available for this study: BM, CGE, G, JEPS, ND, and UC. This paper constitutes part of a doc- MADRONO [Vol. 48 toral dissertation submitted to the Department of Integra- tive Biology, University of California, Berkeley. LITERATURE CITED Boum, B. A., N. A. M. SALEH, AND R. ORNDuFF. 1974. The flavonoids of Lasthenia (Compositae). American Journal of Botany 61:551—561. CHAN, R. K.-G. 2000. Molecular systematics of the gold- field genus Lasthenia (Compositae: Heliantheae sensu lato). Ph.D. dissertation. University of California, Berkeley, CA. CHAN, R., B. G. BALDWIN, AND R. ORNDurFF. In Press. Goldfields revisited: a molecular phylogenetic per- spective on the evolution of Lasthenia (Compositae: Heliantheae sensu lato). International Journal of Plant Sciences 162: 1347-1860. GRAY, A. 1857. Compositae. Jn Description of the general botanical collections, J. Torrey. Reports of explora- tions and surveys, to ascertain the most practical and economical route for a railroad from the Mississippi River to the Pacific Ocean 4:61—182. JOHNSON, D. E. AND R. ORNDUFF. 1978. Lasthenia califor- nica (Compositae), another name for a common gold- field. Madrono 25:227. Keck, D. D. 1959. Baeria, Crockeria, and Lasthenia. In A California flora, P. A. Munz. University of Cali- fornia Press, Berkeley, CA. ORNDUFF, R. 1966. A biosystematic survey of the gold- field genus Lasthenia (Compositae: Helenieae). Uni- versity of California Publications in Botany 40:1—92. . 1971. A new tetraploid subspecies of Lasthenia (Compositae) from Oregon. Madrono 21:96—98. . 1993. Lasthenia. In J. C. Hickman (ed.), The Jep- son manual: higher plants of California. University of California Press, Berkeley, CA. , B. A. Bohm, and N. A. M. Saleh. 1974. Flavo- noid races in Lasthenia (Compositae). Brittonia 26: 411-420. RAJAKARUNA, N. AND B. A. BOHM. 1999. The edaphic fac- tor and patterns of variation in Lasthenia californica (Asteraceae). American Journal of Botany 86:1576— 1596. | | | | | | MADRONO, Vol. 48, No. 3, pp. 211-214, 2001 NOTEWORTHY COLLECTIONS CALIFORNIA CASTILLEJA TENUIS (A. A. Heller) Chuang & Heckard (SCROPHULARIACEAE).— Ventura Co., headwaters of Little Mutau Creek, TON R20W sec. 1, alt. 2042 m, 18 Jun 2000, in moist swales among Pinus jeffreyi, associated with Deschampsia danthonioides, Lotus purshianus, and Scutellaria siphocampyloides. R. Burgess and T. Burgess 3940 (JEPS, SBBG), det. M. Wetherwax (JEPS). Previous knowledge. Known from Alaska south to the Klamath Mountains, Cascade Range, Klamath Range, and Sierra Nevada of California, and disjunctly in the San Ber- nardino and Palomar mountains of southern California (as Orthocarpus hispidus Benth. in P. A. Munz 1963, 1974). Significance. First records for Ventura County and the western Transverse Ranges, representing a disjunction of ca. 140 km southwest of nearest known records in the Sierra Nevada (e.g., Bartolas Creek, Kern County, Twis- selmann 12452, JEPS) and ca. 180 km west of nearest known records in the San Bernardino Mountains (e.g., east of Bluff Lake, Thorne et al. 47843, RSA, UC). TRICHOSTEMA MICRANTHUM A. Gray (LAMIACEAE).— Ventura Co., near Cuddy Ranch, R. Hoffmann s.n., 21 Sep, 1927 (SBBG), det H. Lewis (LA); Mutau Road, ca. 1 mi south of junction with Lockwood Valley Road, T7N R21W sec. 17, alt. 1646 m., 14 July 2000, in dry vernal pool with Epilobium densiflorum, Myosurus minimus, and Psilocarphus tenellus var. globiferus. R. Burgess and T. Burgess 4025 (LA, SBBG), det. H. Lewis (LA), also col- lected by Steve Junak at Yellow Jacket Meadows approx- imately 3.5 km east of this site (S. Junak, personal com- munication); dry sandy floodplain, associated with Mim- ulus pilosus, Iva axillaris var. robustior, Cardaria pubes- cens, and Salix exigua, south side of Lockwood Valley Road, T8N R21W sec. 25, alt. 1494 m., 22 July 2000, R. Burgess and T. Burgess 4048 (LA, SBBG), det. Harlan Lewis (LA). Previous knowledge. Known from margins of mead- ows, drying lakes, and meadows in the San Bernardino Mountains, and the mountains of Baja California del Nor- te (Munz 1963, 1974; Lewis, Brittonia 5:276—303. 1945). Significance. First record for Ventura County and the western Transverse Ranges, representing a disjunction of ca. 180 km from nearest known sites in the San Bernar- dino Mountains (e.g., Big Bear Lake, H. Lewis 1689, LA). SENECIO IONOPHYLLUS E. Greene.—Ventura Co., on north-facing, granitic slopes associated with Pinus lam- bertiana, Galium jepsonii, Arabis repanda, and Hulsea vestita ssp. gabrielensis, west flank of Seward Mountain, TON R19W sec. 5, alt. 2027 m., R. Burgess and T. Burgess 3948 (SBBG), det. D. H. Wilken (SBBG). Previous knowledge. Known from dry, rocky conifer- ous forests in the southern Sierra Nevada, San Gabriel, and San Bernardino mountains of California (Munz 1963, 1974). Significance. First record for Ventura County and the western Transverse Ranges, a disjunction of ca. 100 km from nearest known sites in the San Gabriel Mountains (e.g., Kratka Ridge, R. Bacigalupi 6423, JEPS). PYROLA ASARIFOLIA Michaux ssp. asarifolia (ERICA- CEAE).—On vernally moist, n-facing slope, associated with Pinus lambertiana, Galium jepsonii, Arabis repanda Senecio ionophyllus, west flank of Seward Mountain, T6N R1I9W sec. 5, alt. 2072 m., R. Burgess and T. Burgess 3947 (SBBG), det. D. H. Wilken (SBBG). Previous knowledge. Known from Asia, western North America from Alaska south to California, and northeastern North America. Significance. First record for Ventura County and the western Transverse Ranges, representing a disjunction of ca. 190 km southwest of nearest known sites in the Sierra Nevada (e.g., Whitney Meadows, Tulare Co., Purpus in 1895, UC) and ca. 180 km west of nearest known sites in the San Bernardino Mountains (e.g., Vivian Creek, Munz 7593, RSA). —RICK AND TRISHA BURGESS, Ventura County Flora Project, 221 Juneau Place, Oxnard, CA 93030. MEXICO SETARIA ARIZONICA Rominger (GRAMINEAE).—Baja California Sur, mpio. de La Paz, area del Borrego Cimar- ron, Sierra “‘E] Mechudo’’, cerca del Rancho Las Animas, 25 km al N de San Juan de la Costa. 24°42'’N 110°42’W. Alt. 375 m. Veg. Matorral Xerofilo. 20 Oct 1996. Rey- mundo Dominguez C. 1665 (HCIB, ARIZ, ASC). Previous knowledge. Known only from Pima and Santa Cruz Cos., AZ, and Sonora, Mexico. The species was de- scribed by J. Rominger in his monograph of North Amer- ican Setaria (Illinois Biol. Monogr. No 29, pp 66-68. 1962). The type is L. N. Goodding 3754 from the Babo- quivari Mts., Pima Co., AZ (Holotype ARIZ). Four other specimens are cited, all from Pima and Santa Cruz Cos., AZ. The cited collections were named S. liebmannii E. Fourn., a species which is superficially very similiar. Rominger considered his new species to be rare. Subse- quent to publishing his monograph, Rominger made a few collections, which did not alter the known range. Origi- nally he stated that all collections were within a 100 mile radius of Nogales, but he cited no gatherings from Mex- ico. Apparently he intended the “‘radius of 100 miles’”’ to refer to the United States only. In recent years I have observed the species to be common in the area of Brawley Wash, SW of Tucson. It is frequently growing in the shade of mesquites, as mentioned by Rominger in the proto- logue. I have also seen it in some abundance in the Bue- nos Aires National Wildlife Refuge, near Arivaca, Pima Co. Although not rare, as Rominger thought, the species is not often collected. Aside from gatherings by Rominger, and several by my wife and me, the ARIZ herbarium has only one recent collection from Pima County: R. S. Felger 97-16, from just north of Tucson. Two recent collections from Santa Cruz Co. are L. J. Toolin 2262 and T. R. Van Devender 91-951, both from the Tumacacori Mts. As doc- umentation of S. arizonica in Mexico, ARIZ has Beetle M-6969, T. R. Van Devender 90-489, 91-654, 92-1081 and 94-700, all from Sonora, the latter two some 500 km south of the US-Mexico border. Most collectors since Rominger have commented on the abundance of the species in the areas where they found it. Significance. First record from the Baja California pen- 212. insula. The Dominguez collection also extends the known range of the species some 200 km farther south. ARIZONA SETARIOPSIS AURICULATA (E. Fourn.) Scribner (GRAMIN- EAE).—Pima Co., Altar Valley, 26 km S of Robles Junc- tion. In a broad sandy wash draining into Brawley Wash. Rather local on a north-facing shady bank with Setaria arizonica and S. grisebachii. Elev. 800 m. 23 Aug 1990. J. R. REEDER & C. G. REEDER 8553 (ARIZ, US, CAS, TEX). Previous knowledge. There is no publshed record of this genus occurring in Arizona—nor in the US. Mc Vaugh (Flora Novo-Galiciana, 1983) gives the range as: ““B.C., Son., Chih., Sin., Nay., Gto., Jal., Col., Gro., Méx., Mor., Oax., Pue., S.L.P., Chis., Camp., Yuc.; Centr. Amer.”’ At ARIZ there is a specimen with the following label data: J. J. Thornber s.n. (ARIZ-38003) Nogales, Arizona, among shrubs. 10—12 Sep 1930. It was originally named Chaetochloa grisebachii var. ampla Scribn. & Merr., but in 1958 J. M. Rominger annotated it (correctly) as Setar- iopsis auriculata (Fourn.) Scribner. I find no record that this was published by Rominger, and the name does not appear in Kearney & Peebles’ Arizona Flora (Supplement 1960), nor in Lehr’s Catalogue of the Flora of Arizona (1978). Although one finds the name in the key to Setaria and allied genera in Rominger’s monograph of North American Sefaria (Illinois Biol. Monogr. No. 29. 1962), the author makes no further mention of the genus. Significance. Setariopsis was collected in Arizona in 1930, but this failed to get into literature on Arizona grass- es. The 1990 collection reported here documents that Se- tariopsis 1s, indeed, a part of the Arizona flora. It is also the first published report of the genus from the US. More- over, it occurs about 50 km north of the US-Mexico bor- der. I have visited this site several times during the past decade, and can report that the population, although small, is thriving. It is interesting that among the several Sonoran specimens of this species at ARIZ, the one nearest to Ar- izona 1s T. R. Van Devender et al. s.n. [ARIZ] collected in Palm Canyon, SE of Magdalena some 75 km south of the International Border. ALOPECURUS ARUNDINACEUS. Poir. (GRAMINEAE).—Co- conino Co. N end of Mormon Lake in a marshy area. A common species here, with other grasses; strongly rhizo- matous. Elev. 2200 m. 21 June 1995. J. R. Reeder & C. G. Reeder 9309 (ARIZ, US, CAS). Same area, one of the dominant grasses in marshy ground; plants strongly rhi- zomatous; associated here with Phalaris arundinacea and Poa pratensis. 24 June 1998. J. R. Reeder & C. G. Reeder 9675 (ARIZ, RSA). Previous knowledge. This grass is a native of Eurasia. In Hitchcock’s Manual (1951) it is not included in the key to Alopecurus, but at the end of the treatment for that genus one finds the name Alopecurus arundinaceus Poit., along with a short description. As reason for its inclusion in the Manual, there is the statement: “‘Adventive in hay meadows, Labrador; North Dakota; Eurasia.”? According to a map kindly provided by Mary Barkworth, this species is now established in all northern states west of, and in- cluding, the Dakotas. The most southern records, accord- ing to Barkworth, are Cache Co., Utah and Garfield Co., Colorado. Significance. First record of the species from Arizona; also apparently the most southerly locality in which it is recorded as growing spontaneously in the U.S.A. MADRONO TRIDENS ALBESCENS (Vasey) Wooton & Standley (GRA- MINEAE).—Pinal Co., a well-established local popula- tion in a riparian habitat in a tributary of the Santa Cruz River in the environs of Casa Grande—near the Mormon Batallion Monument on Hwy. 387, ca. 2 miles N of its jet. with Hwy. 84. Elev. 1400 ft. [325 m]. 21 Oct 1991. Dan James s.n. (ARIZ, US, CAS). Same area. Many clumps in a shallow tributary of the Santa Cruz River. 26 June 1997. J. R. & C. G. Reeder 9598 (ARIZ, ASU, NMCR). The James specimen was brought to the ARIZ Herbarium by Horace Miller. Previous knowledge. Hitchcock’s Manual (1951) gives the range as CO, KS, NM, OK, TX. The ARIZ herbarium has specimens grown in Tucson in Experimental grass gar- dens in 1936, 1938, and 1939, but no material from plants which were growing spontaneously in Arizona. Significance. First record of this species for Arizona. Also, it constitutes a considerable range extension. Prior to the Arizona collections the most westerly records for Tridens albescens were Sandoval and Dofia Ana counties in New Mexico. That this species is well established in at least one area in Arizona is attested to by the fact that it was recollected in the same location 6 years later. BRACHIARIA PLATYPHYLLA (Munro ex Wright) Nash (GRA- MINEABE).—Santa Cruz Co., San Rafael Valley, pond on Ki- He-Kah Ranch, T23S R1I7E, Sec. 11 SW%. Common pros- trate annual in mud. Elev. 4850 ft. [ca. 1500 mJ]. 24 Aug 1991. T. R. & R. K. Van Devender et al 91-689 (ARIZ). Same location: Abundant with other weedy vegetation along margin of pond. 22 Sep 1992. J. R. & C. G. Reeder 8930 (ARIZ, ASU, US). San Rafael Valley, ca. 3.5 km SSW of Canelo Pass in vicinity of Little Outfit Ranch. Charco in grassland with scattered oaks and manzanitas. Abundant Marsilea, Heteranthera, and weedy grasses sur- rounding the pond. Fairly common along the pond margin with other dense vegetation. Apparently grazed by cattle. Elev. 1550 m. 22 Sep 1992. J. KR. & C_ G. Reeder 8929 (ARIZ, NMCR). The two localities cited above are sepa- rated by some 8 km. Previous knowledge. A native species from Florida to TX, OK; Mex., Cuba. Easily distinguished from grasses with a similar inflorescence by the spikelet orientation, the first glume turned toward the rachis. Significance. First record from Arizona. Previously not known W of Texas. NOTE: R. D. Webster [The Australian Paniceae (Poaceae). Stuttgart: J. Cramer 1987] argued that the traditional character of spikelet orientation (with lower glume turned toward, or away from the rachis) used to separate Brachiaria from Urochloa is unreliable and, moreover, of minor significance. He proposed restricting the genus Brachiaria to a single species, B. erucaeformis (Sm.) Griseb., and relegating other traditional members of the genus to Urochloa. This suggestion has some merit and has been adopted by some agrostologists. Following Webster, the name of the plant under discussion would be Urochloa platyphylla (Munro ex Wright) R. D. Webster. (For an explanation of the author citations for these bi- nomials, see Kartesz & Ghandi in Phytologia 69(4):303. 1990). UROCHLOA PANICOIDES P. Beauv. (GRAMINEAE).— Maricopa County, Phoenix, in horse pastureland west of lumberyard at Precision Components, Inc., 1820 S 35th Ave. (at Durango). Grass lies flat to ground, radiating from center, flowering stalks rising upward. Formerly (at least 5 yrs. ago) grazed by Chaolais cattle, originally from France via Mexico. Grass made sudden appearance after first discing of pasture. 24 June 1988. D. J. PINKAVA 14365 (ARIZ, ASU, CAS, TEX, US) 2001] Previous knowledge. A native of e and s Africa, India, and Pakistan, and now adventive in many localities in warmer parts of the world. It was not mentioned in Hitch- ~cock’s Manual (1951), and McVaugh (Flora Novo-Gali- ~ciana, 1983) commented that the two collections he cited were apparently the first records of the species in America. In the United States it has now been reported from several localities in Texas and also from New Mexico. It is an aggresive annual which is listed in the Federal Noxious Weed Act. Significance. First record of Urochloa panicoides in Ar- izona; also first report of the species in the U.S.A. west of Doma Ana Co., NM. SCLEROCHLOA DURA (L.) P. Beauv. (GRAMINEAE).— Maricopa Co., Phoenix, Encanto Golf Course. Thriving on the fairway, where it has persisted for two or three years. | May 1988. Robert Lytle s.n. (ARIZ, ASU, US). Previous knowledge. A rather inconspicuous annual, native to southern Europe and the Middle East. Adventive in the U.S. in several scattered locations in western states, usually as a weed in lawns, golf courses, and roadsides. Reported from WA, OR, ID, CO, UT, NM, TX. Also known from CA see Hickman 1993, Jepson Manuel. Significance. First record for Arizona. ENNEAPOGON CENCHROIDES (Licht.) C. E. Hubbard (GRA- MINEAE).—Pima County, Santa Catalina Mountains, Molino Basin, south side of road, elevation 4500 ft [1370 m]. September 1976. EF. Schmutz s.n. (ARIZ); along Mt. Lemmon Hwy. in the Molino Basin area, common along the highway, elevation 1280 m, 22 September 1980. J. R. Reeder & C. G. Reeder 7329 (ARIZ). Tucson Mountains, in sandy soil of Oeste Wash, Tucson Mountain Park; T14S R12E S13 NE%, elevation 2650 ft. [800 m] 29 October 1989 P. D. Jenkins 89-70 (ARIZ). Previous Knowledge. This species is a native of the Old World. Renvoize (Kew Bull. 22:393—402. 1968), in his study of the genus, gives the range of Enneapogon cen- chroides as: “‘From Sudan southwards to the Cape Prov- ince of South Africa; through Arabia to India; also on Ascension Island.”’ I have been unable to find that it is anywhere recorded as growing spontaneously in _ the U.S.A. Moreover, I have no information regarding when, nor why, it became established in the Santa Catalina and Tucson Mountains of Arizona. The ARIZ Herbarium has several other collections from the southern Santa Catalina Mountains between 600 and 1400 m, the latest with the date 1999. ENNEAPOGON CENCHROIDES is a robust annual, which can attain a height of a meter. Although the spikelets are sim- ilar to those of our native EF. desvauxii P. Beauv., it is readily distinguished by its much coarser stems, and the larger, somewhat open inflorescence which is often as much as 20 cm or more in length. The plant is quite at- tractive and conspicuous when in flower. Significance. Although it has been documented by spec- imens at ARIZ since 1976, curiously there seems to be no record in the literature that it is established anywhere in the United States. —JOHN R. REEDER, Herbarium, University of Arizona, 113 Shantz Building, Tucson, AZ 85721. W ASHINGTON AMARANTHUS BLITUM L. (AMARANTHACEAE ).—King Co., Juanita Beach Park, wet sandy shore of Lake Wash- NOTEWORTHY COLLECTIONS 213 ington, T26N RSE S30, elev. 4 m, 5 Sep 1998, Weinmann 42 (WTU); S shore of Lake Sammamish, mouth of Issa- quah Cr., sand and gravel shore of small island and ad- jacent park beach, T24N R6E S17, elev. 8 m, 25 Aug 1999, Zika 14131 (WTU); S shore of Lake Sammamish, mouth of Laughing Jacobs Cr., gravelly shore, with Lyth- rum salicaria, Cyperus bipartitus, T24N R6E S16, elev. 8 m, 7 Oct 1999, Zika 14560, Jacobson & Weinmann (WTU); Bellevue, N shore of Phantom Lake, damp gravel near lawn, with Hypericum mutilum, Portulaca oleracea, T24N RSE S2, elev. 75 m, 15 Oct 1999, Zika 14604, & Weinmann (WTU). Previous knowledge. Guernsey pigweed is introduced from the Mediterranean, and found as a weed in eastern North America, west to Utah. Significance. First report for Washington. This and the following taxa are all from the Seattle metropolitan area. BALLOTA NIGRA L. subsp. fOETIDA (Vis.) Hayek (LAMI- ACEAE).—King Co., Seattle, weed in shade, Univ. of Washington campus, T25N R4E S16, elev. 25 m, 26 Oct 1999, Zika 14655 & Jacobson (WTU). Previous knowledge. Black horehound is native to Eu- rope, and adventive in eastern North America, west to Nebraska. Significance. First report for Washington. BRIZA MINOR L. (POACEAE ).—Kitsap Co., Restoration Point, SE end of Bainbridge Is., Puget Sound, weed in meadow near golf course, with Perideridia gairdneri, Hy- pochaeris radicata, Agrostis capillaris, T24N R2E S12, elev. 4 m, 20 Aug 1999, Zika 14085 & Jacobson (WTU). Previous knowledge. Little quaking grass is a common weed in the Willamette River valley of northern Oregon, 200 km to the south. Significance. First report for Washington. CAREX PENDULA Huds. (CYPERACEAE).—King Co., Washington Park arboretum, naturalized along small creek, with Ranunculus repens, Equisetum telmateia, T25N R4E° S821, elev. 1s m, 11 Oct 1999° Zika 14576 (MICH, WTU). Previous knowledge. Pendulous sedge is native to Eu- rope and occasionally planted as an ornamental in western Washington. Known as an adventive in the arboretum for the last decade. Significance. First report as an escape from cultivation in Washington. CAREX PROJECTA Mack. (CYPERACEAE).—King Co., near NE shore of Rattlesnake Lake, just above high water line, in partial shade, with Malus fusca, Alnus rubra, Salix sitchensis, Phalaris, T23N R8E S34, elev. 275 m, 26 Jul 1996, Weinmann 30, 31 (WTU); same site, 29 Sep 1999, Zika 14428 & Weinmann (WTU). Previous knowledge. Necklace sedge is native to east- ern North America, west to Saskatchewan. Significance. First record as an adventive in Washing- ton. CAREX SYLVATICA Huds. (CYPERACEAE).—King Co., SE end of Mercer Is., Lake Washington, Clarke Beach Park, weed along shaded, paved trail, with Lapsana, Hed- era, Carex deweyana, Acer macrophyllum, T24N RSE S30, elev. 10 m, 13 June 1998, Weinmann 35 (WTU); same site, 6 Oct 1999, Zika 14523 & Weinmann (WTU). Previous knowledge. Wood sedge is native to Europe, and has been reported as an adventive in southern British Columbia and eastern North America. Significance. First report for Washington. CRASSULA TILLAEA Lest.-Garl. (CRASSULACEAE).— King Co., Shilshole Bay, Seattle waterfront, 0.7 km S of Meadow Point, common weed in gravel and bare ground, 214 with Poa annua, P. pratensis, Aira caryophyllea, T25N R3E S3, elev. 2 m, 8 May 1999, Jacobson s.n. (WTU); same site, 25 May 1999, Zika 13758 & Jacobson (WTU). Previous knowledge. Mossy stonecrop is native to Eu- rope and adventive on the west coast, N to Lane Co., OR, 370 km to the S. Significance. First report for Washington. CYPERUS ODORATUS L. (CYPERACEAE ).—King Co., West Point, Seattle waterfront, weed in wetland, with Mentha pulegium, Cyperus eragrostis, T25N R3E S9, elev. 2 m, 20 Oct 1999, Jacobson s.n. (EIU, WTU). Previous knowledge. Rusty flat sedge is a pantropical weed, and has been collected in Multnomah Co., OR, 200 km to the S. Significance. First report for Washington. DATURA WRIGHTII Regel (SOLANACEAE).—King Co., Queen Anne, Seattle, Queen Anne Ave. near Boston St., weed in gravel parking lot, T25N R3E S24, elev. 120 m, 20 Oct 1999, Zika 14632 & Jacobson (WTU). Significance. First report for Washington. ERAGROSTIS CURVULA (Schrad.) Nees (POACEAE).— King Co., S side of West Point, Seattle waterfront, dis- turbed ground near path, T25N R3E S16, elev. 3 m, 4 August 1999, Jacobson s.n. (WTU). Previous knowledge. Weeping lovegrass is native to Af- rica, and has been collected as an adventive in Multnomah Co., OR, 200 km to the S. Significance. First report for Washington. Known from the site since 1998, and increasing. GEUM URBANUM L. (ROSACEAE).—King Co., Island Crest Park, Mercer Is., Lake Washington, shaded trailside, and in wetland below suspension bridge, with Hedera, Geum macrophyllum, T24N RSE S19, elev. 90 m, 6 Oct 1999, Zika 14532 & Weinmann (WTU); Seattle, Univ. of Washington campus, T25N R4E S16, elev. 25 m, 26 Oct 1999, Zika 14659 & Jacobson (WTUV); Seattle, Interlaken Park, shaded roadside, T25N R4E S20, elev. 40 m, 7 Oct 1999, Zika 14554 (WTU); Seattle, arboretum, common along paths, T25N R4E S21, elev. 20 m, 4 Sept 1998, Zika 13544 (WTU); Seattle, Lakeview Park, Harrison Ridge, bare ground, partial shade, T25N R4E 27. Previous knowledge. Wood avens is native to Europe, and was first observed in the arboretum in 1978. It has been known as a weed in Portland, OR, since 1993, 230 km to the south. Significance. First report for Washington. PARIETARIA JUDIACA L. (URTICACEAE).—King Co., Capitol Hill, Seattle, near Aloha St., weed in cracks in concrete, T25N R4E S828, elev. 110 m, 20 Oct 1999, Zika 14629 & Jacobson (WTU); Pigeon Point, Seattle, near 19th St., weed on shaded ground near concrete steps, from top of bluff to base of West Seattle Bridge, T24N R3E S13, elev. 15—45 m, 20 Oct 1999, Zika 14633 & Jacobson (WTU). MADRONO [Vol. 48 Previous knowledge. Pellitory-of-the-wall is native to Africa and Eurasia. It is weedy in coastal California, 1000 km to the south. Significance. First report for Washington. PARIETARIA OFFICINALIS L. (URTICACEAE).—King Co., Seattle, Univ. of Washington campus, T25N R4E S16, elev. 25 m, 26 Oct 1999, Zika 14657 & Jacobson (V, WTU). Previous knowledge. Eastern pellitory-of-the-wall is na- tive to central and southern Europe. Cultivated at the me- dicinal herb garden of the University, it is now an occa- sional weed in the area. Significance. First report for Washington as an escape from cultivation. POTENTILLA INCLINATA Vill. (ROSACEAE).—King Co., Seattle, Univ. of Washington campus, T25N R4E S16, elev. 25 m, 26 Oct 1999, Zika 14651 & Jacobson (WTU). Previous knowledge. Cultivated in the medicinal herb garden at the University for a decade, and readily reseed- ing in adjacent areas. Removed from the gardens ca. 1990, but persisting as a rare weed in the area. Significance. First report for Washington as an escape from cultivation. SCROPHULARIA NODOSA L. (SCROPHULARIACEAE).— King Co., Seattle, Univ. of Washington campus, T25N R4E S16, elev. 25 m, 26 Oct 1999, Zika 14666 & Jacob- son (WTU); Seattle, Good Sheperd Center, NE 50th St., waste ground, T25N R4E S8, elev. 90 m, 16 May 2000, Zika 14983 (WTU). Previous knowledge. Common figwort is native to Eu- rope and cultivated in the medicinal herb garden at the University. Now an occasional weed in the area. Significance. First report for Washington. VERBASCUM PULVERULENTUM Vill. (SCROPHULARIA- CEAE).—King Co., Seattle, Washington Park arboretum, rare weed along path, with Dactylis, Lapsana, Poa pra- tensis, Taraxacum, T25N R4E S21, elev. 20 m, 15 Sep 1999, Zika 14338 & Jacobson (WTU); arboretum, adven- tive by storm grate, T25N R4E S21, elev. 25 m, 17 Nov 1999, Zika 14739 & Jacobson (WTU). Previous knowledge. Hoary mullein is native to Europe, and has not been reported as a wild plant in our area. Significance. First report for Washington. VERBENA OFFICINALIS L. (VERBENACEAE).—Seattle, Univ. of Washington campus, weed on waste ground, T25N R4E S16, elev. 25 m, 26 Oct 1999, Zika 14649 & Jacobson (WTUV). Previous knowledge. Vervain is native to Europe, and has been reported as a weed on ballast in Multnomah Co., Oregon, 200 km to the south. Significance. First report for Washington. —ARTHUR L. JACOBSON, FREDERICK C. WEINMANN, and PETER FE ZIKA, Herbarium, Dept. of Botany, Box 355325, Univ. of Washington, Seattle, WA 98195-5325. MApRONO, Vol. 48, No. 3, pp. 215-218, 2001 A TRIBUTE TO THE CONTRIBUTIONS OF PROFESSOR JACK MAJOR M. G. BARBouR, P. A. CASTELFRANCO, M. REJMANEK, AND R. W. PEARCY University of California, Davis, CA 95616 Jack Major, Professor Emeritus of Plant Ecology at the University of California, Davis (UCD), died 13 February 2001 in Davis at the age of 83. Professor Major had a profound impact on the direction of plant ecology in the United States during the second half of the 20th century. His contributions to ecologists and land managers in Cal- ifornia are particularly important, and those contributions were highlighted by the editors of Madrono in a 1982 issue dedicated to him (Parsons 1982; Anonymous 1982). Jack’s academic home for most of his career was the UCD Botany Department, where he taught from 1955 un- til retirement in 1981. His spiritual home, however, was in the mountains: the Uinta Mountains of Utah, the Sierra Nevada of California, the Grand Tetons of Wyoming, the Brooks Range and the Juneau ice fields of Alaska, and the Himalayas of Nepal. This was the environment that he most often shared with graduate students and those undergraduates fortunate enough to take his plant ecology classes. He truly was the ideal scientist described by Poin- care (1958), as someone who “... does not study Nature because it is useful to do so. He studies it because he takes pleasure in it ... [and] because it is beautiful.”’ Jack was born 15 March 1917 in Salt Lake City, UT and completed high school there in 1935. He went on to Utah State Agricultural College (now Utah State Univer- sity) and received a BS in Range Management in 1942. For the next several years he served in the Army’s 10th Mountain Division, the justifiably famous unit of 1000 skiers and alpinists who trained hard in the mountain west before participating in the Italian campaign (Fig. 1). After the war, a number of men from that Division went on to become conservationists, ecologists, and leaders in the promotion of recreational skiing. Between 1946 and 1953, Jack attended graduate school at the University of Cali- fornia, Berkeley, obtaining a PhD in Soil Science under the direction of Professor Hans Jenny. During this time he also met and married Mary Cecil, thanks to an introduc- tion from brother Ted who had met Mary by chance on a rock climbing expedition in the Grand Tetons. She, too, had a love for the mountains. Mountain landscapes and vegetation remained lifetime passions for both of them and for their sons, as celebrated in Paul Castelfranco’s 1988 poem, “‘Voices of the mountains:”’ I listened to voices/The innumerable voices/Of the mountains. . . Voices of red lava pinnacles/Voices of grey gran- ite boulders. . . And up the ridge/Above the forest/Above the meadow/ Through lodgepole, hemlock/And timberline. From the crest I could see/Vast basins of granite/ And blue silhuoettes/ In the distance.../ I listened/To all these voices/To the one complex chorus/Of the mountain. Jack was hired as a member of a young weed science group in the Botany Department at UCD (Fig. 2). His strong interest in the ecology of un- disturbed mountain vegetation, however, conflicted with the weed group’s focus on plants in agronom- ic, low-elevation settings. This habitat bias gradu- ally distanced him from weed science, and a 1964 Fulbright Fellowship to Innsbruck, Austria was to cement a lifetime’s focus on vegetation science. He had a driving curiosity that made him an ex- tensive reader of, and correspondent with, scientists who specialized in a wide range of topics, including those who wrote in other languages. As a result, he was far ahead of his time. For example, we have correspondence in 1948 between Jack and Sewal Wright, one of the major contributors to the syn- thesis of Darwinism and Mendelism. Wright re- sponded to Major’s query of how to determine the relative importance of multiple interacting factors that explain a plant community’s distribution limits, by describing his own original statistical method, path analysis. Path analysis has only been used reg- ularly in the ecological literature for the past dozen years, but it was part of Jack’s education 40 years earlier. Another example: Inspired by his major professor’s book (Jenny 1941), The factors of soil formation, he wrote a paper (Major 1951) that pro- posed to use differential equations to describe the sum of vegetation-environment relationships for any given plant community. Not for another quarter of a century, however, did any ecologist actually begin to use differential equations in models of plant communities. Several aspects of the Jenny/Major approach may now appear to be naive. Today—instead of relating attributes of soils or vegetation directly to the factors of soil formation—a clear distinction be- tween, processes and factors seems now to be the more productive way to go (Humphreys and Paton 1998). But we must remember that Major’s 1951 paper was written in a pre-ordination era before ad- equate canonical multivariate techniques with per- mutation tests and computers were available to test his hypotheses. One measure of Professor Major’s vision and 1m- pact is the fact that several of his earliest papers are still cited today, in some cases more often now than originally. According to the ISI Web of Sci- ence, “‘A functional, factorial approach to plant ecology’’ has been cited 91 times in the past 25 years. His superb synthesis of the California flora, geology, and ecology (“‘Endemism and speciation 216 Fic. |. Corporal Jack Major when a member of the 10th Mountain Divisiion, 1944. Photo courtesy of (then Ser- geant) Don Bothwick. in the California flora,’’ Stebbins and Major 1963) has been cited 102 times in the same period, and his most-often cited paper, ““‘Buried viable seeds in California bunchgrass sites and their bearing on the definition of flora,” (Major and Pytott 1956) has been cited 138 times, and is still being cited at the rate of seven times per year for the past 5 years. His work on primary succession following glacial retreat (Crocker and Major 1955) is a classic, cited and described in many textbooks nearly a half-cen- tury later (e.g., Barbour et al. 1999; Begon et al. 1996; Krebs 2001) and in recent reviews on suc- cession (Wali 1999). Jack was one of very few Americans to practice the phytosociological protocols widely used in Eu- rope (and throughout the non-English-speaking world) for the sampling and classification of veg- etation. Consequently, releve sampling and syntax- onomy were employed by most of Jack’s students in their dissertations (e.g., Neilson 1961; Pemble 1970; Taylor 1976; Burke 1979; Benedict 1981). Jack’s gentle leadership in pulling reluctant Amer- ican ecologists across a then-narrow bridge of com- munication into the rest of the world, was without doubt of seminal help later to Robert Whittaker in the 1970s when his travels and publications wid- ened that bridge. Only now—20-—30 years after his students have finished their graduate degrees—are phytosociological papers becoming accepted and publishable in the US. Professor Major was the opposite of a bandwag- MADRONO [Vol. 48 rN \ ’ \ | 1 i a N Wane ipo a. - Th Lr Assistant Professor Jack Major in 1956 as a new faculty member at UC Davis while on a departmental field trip at Calaveras Big Trees State Park. Photo courtesy of Roman Gankin. Fic. 2. on scientist. He preferred to go in his own sense of an appropriate direction, even when he was so far ahead of others that few understood his choice or took the same route. Paul Castelfranco’s 1991 poem, “‘Epitaph,’’ captures this aspect of his per- sonality: “‘And on his grave some kindly person wrote/Never did he jump on a bandwagon ... /He preferred to walk.”’ Throughout his career, Dr. Major was as well- known for his reviews of ecological books written in other languages as for his own research. The journal Ecology alone publshed 158 of his book reviews, most of them of works written in French, German, and Russian. These detailed reviews brought foreign news and ideas to the attention of otherwise ethnocentric and linguistically chal- lenged American ecologists. In 1975 the Ecological Society of America gave him its first Distinguished Service Citation specifically for his his prodigious reviewing activity, judged to be an outstanding ser- vice to Society members. According to then-Presi- dent Richard Miller (1975), “‘Major’s reviews have consistently pointed out gaps in our own knowl- edge of American ecosystems and have indicated directions for fruitful new research ... [We] would be immeasurably poorer without his dedicated ef- forts.’’ Unfortunately, his encyclopedic knowledge of the literature has only partially been preserved in his papers, reviews, and bibliographies (Major and Rejmanek 1988/9). Unfortunately, also, is the fact that this kind of selfless scholarly work is poor- ly rewarded by the usual academic promotion pro- cess. Dr. Major’s tenure promotion was repeatedly delayed and finally achieved long after those who 2001] Fic. 3. Jack Major crossing an Alaskan stream in 1982, on his way to visit the research area of (then student) Ann Odaz. Photo courtesy of Dr. Odaz. understood the value of his work would have awarded it. He was a gentleman scholar: learned but soft- spoken and modest to the point of self-effacement. If presented in conversation with an opinion con- trary to his own, he was sincerely quizzical and would quite innocently ask why one thought that way, rather than offering a defensive or challenging counter-statement. In this manner, Jack made those around him feel equally learned. Even when he dis- agreed with them, his own contrary opinions were delivered so delicately and non-confrontationally (usually ending with his traditional phrase, “‘Is this alright?’’) that the recipients might not realize their logic had been shredded until reflecting on it some days later. His forte in teaching was with small groups. His low-key manner was not well suited to large lecture sections or busloads of fieldtrip students. On hikes in the field, a student had to be self-motivated enough to keep up and crowd close around him while he pointed out species and talked of their indicator value. Those who hung back missed a great education. His method of teaching was So- OBITUARY PANG Fic. 4. Jack and Mary Major hiking in the Grand Tetons in 1992. Photo courtesy of Ted Major. cratic, inviting questions and asking questions back, usually including his stock phrase, “Is this al- right?”’ because he didn’t want to lose anyone. His classes and his research interests were reflected in theses, dissertations, and publications: alpine plant communities (Burke 1979; Neilson 1961; Major and Taylor 1977, 1988), biogeography (Taylor 1977), California vegetation (Barbour and Major 1977, 1988), gradient analysis (Waring and Major 1964), plant ecophysiology (Macdonald 1981; Bar- ry 1968), plant-soil relations (Myatt 1968), system- atics (e.g., Gankin 1957), the history of ecological concepts (Major 1969), and vegetation change (VanKat 1970). He was mentor to more than 20 graduate students of his own and to many more via correspondence or by way of serving as a member on their thesis/ dissertation committees (Fig. 3). We join his wife Mary and sons Paul, John, and James, and brother Ted in their sorrow at his phys- ical absence among us now; but the memories of his delight in the high country remain with us (Fig. 4). A modified version of this memorium recently appeared in the Bulletin of the Ecological Society of America (Barbour et al. 2001). LITERATURE CITED ANONYMOUS. 1982. Symposium papers in honor of Jack Major, and a dedication. Madrono 29:145—219. BARBOUR, M. G., J. H. BuRK, W. D. Pitts, E S. GILLiIAM, AND M. W. SCHWARTZ. 1999. Terrestrial plant ecology, 3rd ed. Addison Wesley Longman, Menlo Park, CA. —., P. A. Castelfranco, R. W. Pearcy, and M. Rejma- nek. 2001. Resolution of respect: Jack Major, 1917— 2001. Bulletin of the Ecological Society of America 82:174—-176. and J. Major (eds.). 1977. Terrestrial vegetation of California. Wiley, New York, NY. [Revised in 1988 and published by the California Native Plant Society, Sacramento, CA.] Barry, W. J. 1968. The ecology of Populus tremuloides, a monographic approach. Ph.D. dissertation. Univer- sity of California, Davis. 218 BEGon, M., J. L. HARPER, AND C. R. TOWNSEND. 1996. Ecology, 3rd ed. Blackwell, Cambridge, MA. BENEDICT, N. B. 1981. The vegetation and ecology of sub- alpine meadows of the Sierra Nevada, California. Ph.D. dissertation. University of California, Davis. BurGEss, R. L. 1996. American ecologists: a biographical bibliography. Huntia 10:5—116. BurKE, M. T. 1979. The flora and vegetation of the Rae Lakes Basin, southern Sierra Nevada: an ecological overview. M.S. thesis. University of California, Da- VIS. CASTELFRANCO, P. A. 1991. Pebbles and flints. Rock Crys- tal Press, Davis, CA. . 1988. A basket of straw. Rock Crystal Press, Da- vis, Ca. GANKIN, R. 1957. The variation patern and ecological re- strictions of Arctostaphylos myrtifolia Parry (Erica- ceae). M.A. thesis. University of California, Davis. HumPHREYS, G. S. AND T. R. PATON. 1998. On the relations between complex systems and the factorial model of soil formation—discussion. Geoderma 86:34—41. JENNY, H. 1941. The factors of soil formation. McGraw- Hill, New York, NY. Kress, C. J. 2001. Ecology, 5th ed. Benjamin Cummings, Palo Alto. MACDONALD, R. 1981. Water relations of woody plants in riparian, chaparral, and foot-hill woodland vegetation types of the Inner Coast Ranges, California. Ph.D. dissertation (incomplete). University of California, Davis. Mayor, J. 1951. A functional, factorial approach to plant ecology. Ecology 32:392-—412. . 1969. Historical development of the ecosystem concept. pp. 9-22 in G. M. Van Dyne (ed.), The eco- system concept in natural resource management, Ac- ademic Press, New York, NY. , and W. T. Pyott. 1966. Buried, viable seeds in two California bunchgrass sites and their bearing on the definition of a flora. Vegetatio 13:253-282. , and M. Rejmanek. 1988/9. Bibliographic review on the vegetation of California and its ecology, parts III nd IV. Excerpta Botanica, Section B 25:279—320 and 26:1—125. MADRONO [Vol. 48 , and D. W. Taylor. 1977. Alpine. Pp. 601—675 in M. G. Barbour and J. Major (eds.), Terrestrial vege- tation of California, Wiley, New York, NY. [Revised in 1988 and published by the California Native Plant | Society, Sacramento, CA.] MILLER, R. S. 1975. Distinguished service citation for Jack Major. Bulletin of the Ecological Society of America 56(4):24. Myatt, R. G. 1968. The ecology of Eriogonum apricum Howell. M.S. thesis. University of California, Davis. NEILSON, J. A. 1961. Plant associations on glaciated gran- ite at Sterline Lake, Nevada County, California. M.S. thesis. University of California, Davis. Parsons, D. J. 1982. The role of plant ecological research in Sierran park management: a tribute to Jack Major. Madrono 29:220—226. PEMBLE, R. H. 1970. Alpine vegetation in the Sierra Ne- vada of California as lithosequences and in relation to local site factors. Ph.D. dissertation. University of California, Davis. POINCARE, A. 1958. The value of science. Dover, New York, NY. STEBBINS, G. L. AND J. MAJor. 1965. Endemism and spe- ciation in the California flora. Ecological Monographs 35:1-35. TAYLOR, D. W. 1976. Ecology of the timberline vegetation at Carson Pass, Alpine County, California. Ph.D. dis- sertation. University of California, Davis. , 1977. Floristic relationships along the Cascade- Sierran Axis. American Midland Naturalist 97:333— 349. VANKaT, J. L. 1970. Vegetation change in Sequoia Na- tional Park, California. Ph.D. dissertation. University of California, Davis. , and J. Major. 1978. Vegetation changes in Se- quoia National Park, California. Journal of Biogeog- raphy 5:377—402. WaLl, M. K. 1999. Ecological succession and the reha- bilittion of disturbed terrestrial ecosystems. Plant and Soil 213:195—220. WARING, D. AND J. MAJor. 1964. Some vegetation of the California coastal redwood region in relation to gra- dients of moisture, nutrients, light, and temperature. Ecological Monographs 34:167—215. Mapbrono, Vol. 48, No. 3, p. 219, 2001 ANNOUNCEMENT ‘The Eighty-Year Index to Madrono (Volumes 1—43, 1916—1996)”’ now available. “The ‘“‘Eighty-Year Index to Madrono,” a publication of the California Botanical Society (ISBN 0-9961882-0- 9), is now available from Sycamore Press. Through 1996, 43 volumes of Madrono were published consisting of 282 numbers, 13,554 numbered pages and 234 supplemental pages. Information in “‘The Eighty-Year Index to Madro- fio”’ is presented in four sections, Authors & Titles (2,277 entries); Subject Index (11,812 entries); Noteworthy Col- lections (857 entries); and Reviews of Books and Articles (929 entries). All sections are cross-referenced. The ‘“‘In- dex”’ is a rich resource of scientific and historical infor- mation for those interested in any aspect of California botany. A table listing each volume, year(s) of publication, numbers per volume, and pages per number is included. To order send a check or P.O. for $37.20 to Sycamore Press, 6355 Riverside Blvd., Suite C, Sacramento, CA 95831 (916/427-0703), attention ‘“‘Madrono Index.’ The price includes tax and shipping to USA addresses. Add $10.00 for shipping to foreign addresses. Credit card pur- chase not available. NEW EDITOR There will be a new editor for Madrono beginning with volume 49. Please send all manuscript submissions to: Dr. John Callaway Dept. of Environmental Science University of San Francisco 2130 Fulton Street San Francisco, CA 94117-1080 email: callaway @usfca.edu Volume 48, Number 3, pages 131-219, published 1 February 2002 = "& é ‘ SUBSCRIPTIONS—-MEMBERSHIP Membership in the California Botanical Society is open to individuals ($27 per year; family $30 per year; emeritus $17 per year; students $17 per year for a maximum of 7 years). Late fees may be assessed. Members of the Society receive MAproNo free. Institutional subscriptions to MApRONO are available ($60). Membership is based on a calendar year only. Life memberships are $540. Applications for membership (including dues), orders for sub- scriptions, and renewal payments should be sent to the Treasurer. Requests and rates for back issues, changes of address, and undelivered copies of Maprono should be sent to the Corresponding Secretary. INFORMATION FOR CONTRIBUTORS Manuscripts submitted for publication in Maprono should be sent to the editor. It is preferred that all authors be members of the California Botanical Society. Manuscripts by authors having outstanding page charges will not be sent for review. Manuscripts may be submitted in English or Spanish. English-language manuscripts dealing with taxa or topics of Latin America and Spanish-language manuscripts must have a Spanish RESuMEN and an English ABSTRACT. Manuscripts and review copies of illustrations must be submitted in triplicate for all articles and short items (NOTES, NOTEWORTHY COLLECTIONS, POINTS OF VIEW, etc.). Follow the format used in recent issues for the type of item submitted. Allow ample margins all around. Manuscripts MUST BE DOUBLE-SPACED THROUGHOUT. For articles this includes title (all caps, centered), author names (all caps, centered), addresses (caps and lower case, centered), abstract and resumen, 5 key words or phrases, text, acknowledgments, literature cited, tables (caption on same page), and figure captions (grouped as consecutive paragraphs on one page). Order parts in the sequence listed, ending with figures. Each page should have a running header that includes the name(s) of the author(s), a shortened title, and the page number. Do not use a separate cover page or ‘erasable’ paper. Avoid footnotes except to indicate address changes. Abbreviations should be used sparingly and only standard abbrevia- tions will be accepted. Table and figure captions should contain all information relevant to information presented. All measurements and elevations should be in metric units, except specimen citations, which may include English or metric measurements. Authors of accepted papers will be asked to submit an electronic version of the manuscript. Microsoft Word 6.0 or WordPerfect 6.0 for Windows is the preferred software. Line copy illustrations should be clean and legible, proportioned to the MaprRoNo page. Scales should be in- cluded in figures, as should explanation of symbols, including graph coordinates. Symbols smaller than 1 mm after reduction are not acceptable. Maps must include a scale and latitude and longitude or UTM references. In no case should original illustrations be sent prior to the acceptance of a manuscript. Illustrations should be sent flat. No illustrations larger than 27 X 43 cm will be accepted. Presentation of nomenclatural matter (accepted names, synonyms, typification) should follow the format used by Sivinski, Robert C., in MADRONO 41(4), 1994. Institutional abbreviations in specimen citations should follow Holmgren, Keuken, and Schofield, Index Herbariorum, 8th ed. Names of authors of scientific names should be abbreviated according to Brummitt and Powell, Authors of Plant Names (1992) and, if not included in this index, spelled out in full. Titles of all periodicals, serials, and books should be given in full. Books should include the place and date of publication, publisher, and edition, if other than the first. All members of the California Botanical Society are allotted 5 free pages per volume in Maprono. Joint authors may split the full page number. Beyond that number of pages a required editorial fee of $40 per page will be assessed. The purpose of this fee is not to pay directly for the costs of publishing any particular paper, but rather to allow the Society to continue publishing MaproNo on a reasonable schedule, with equity among all members for access to its pages. Printer’s fees for illustrations and typographically difficult material @ $35 per page (if their sum exceeds 30 percent of the paper) and for author’s changes after typesetting @ $4.50 per line will be charged to authors. At the time of submission, authors must provide information describing the extent to which data in the manu- script have been used in other papers that are published, in press, submitted, or soon to be submitted elsewhere. VOLUME 48, NUMBER 4 OCTOBER—DECEMBER 2001 MADRONO A WEST AMERICAN JOURNAL OF BOTANY “ONTENTS MarITIME CHAPARRAL COMMUNITY TRANSITION IN THE ABSENCE OF FIRE PriC Van Dy ke-and Karen Dill Olle ere eeiseee sneer ese ee 221 THE EFFECTS OF LITTER AND TEMPERATURE ON THE GERMINATION OF NATIVE AND Exotic GRASSES IN A COASTAL CALIFORNIA GRASSLAND Sally A. Reynolds, Jeffrey D. Corbin, and Carla M. D’Antonio ............... 230 CompPosITION, INVASIBILITY, AND DIVERSITY IN COASTAL CALIFORNIA GRASSLANDS Mark R. Stromberg, Paul Kephart, and Vern YAdON ........0ccccccccccesceseeeees 236 PRE-AGRICULTURAL GRASSLAND IN CENTRAL Cough GVCTL TELOVSECTIO aca rneeee eset ee ese eri asses DS THE POLLINATION BIOLOGY OF ERIASTRUM DENSIFOLIUM S ssp. SANCTORUM ; (POLEMONIACEAE), AN ENDANGERED PLANT x x? Deborah K. Dorsett, C. Eugene Jones, and | Jack Jal Burk .. eT: 265 DEMOGRAPHY AND POPULATION BIOLOGY OF A RARE TARPLANT, 4 LEPHARIZONIA PiuMosa, (ASTERACEAE) A CALIFORNIA SUMMER ANNUAL Fors Steven D. Gregory, Erin “s Be i see M. Carlsen, and Erin K. Bissell ...:...000000.0000- @AaoG AR a Bal Pei Stroh oT. WA eas ee ore 272 RECRUITMENT OF FRAXINUS PENNSYLVANICA (Gteaceas) IN V EASTERN ‘Montana WOODLANDS ae izes. P/\> ‘| ae iS 3 P, FID, Peter lésictf/........./,caa, Wha s® wees =h CT dye...- AIR... 286 NEW HaARMONIA Guaaoizionum (Composirar- -MablINak), A. NEW TARWEED FROM SPECIES ULTRAMAFICS OF SOUTHERN MENDOCINO. Count CALIFORNIA Bruce G. Baldwin... a4 Me AMAL; Nostale ASS eee ere 293 A New SPECIES OF Dipymopon (Musct) FROM Ciro YA ) Richard H. Zander .. ry eee Ws oS... ES othe eraeecene 298 | OK REVIEW ECOSYSTEMS OF THE WorLD 16: ECosYsTEMs OF DISTURBED Sark, EDITED BY LAWRENCE R. WALKER ; v William H. RusSell.............0....0004. 9M ines aaah RI 2 Sati tin hades, em 301 JUNCEMENTS PNIN@ WIN CEMIENTES Nar Meat ete ca sd AcE sas piecemeal ake ahs oe a 302 PRESIDENT: S-REPORTFOR® VOLUME 4-8: s2s;ccsssuscoves cases ssashap coders acaesdibocssacouascceisesaerteeeses 303 EDITOR'S REPORT FOR VOLUME 48 .cc0.cs5005.4.ae3500005c ee een veces seneeecoasasocseece 304 REVIEWERS OF MANUSCRIPTS 22..202.02-000-ceseteeuts gee” ETE Bere Manso sat ces teevenee 305 INDEX. TORY, OLUME 4 8x shen caat escent ercdede enue Oe le Nees isos 306 DEDICATION AND BIOGRAPHY—JUNE MCCASKLL ..... BIG) -te..arcgesssscseccossncneenedtpasensenncenes 308 TABLE OF CONTENTS FOR VOLUME 48 ........eceeeeeeseeeeees wed she Sage nga es: il DAMES OF IPUBIIGATIONG eter aseiin scncstel amen estan enn an en ese me aanSs ere ene il PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL SOCIETY Maprono (ISSN 0024-9637) is published quarterly by the California Botanical Society, Inc., and is issued from the office of the Society, Herbaria, Life Sciences Building, University of California, Berkeley, CA 94720. Subscription information on inside back cover. Established 1916. Periodicals postage paid at Berkeley, CA, and additional mailing offices. Return requested. POSTMASTER: Send address changes to MApRONO, Roy Buck, % University Herbarium, University of California, Berkeley, CA 94720. Editor—Dr. JOHN CALLAWAY Dept. of Environmental Science University of San Francisco 2130 Fulton Street San Francisco, CA 94117-1080 callaway @usfca.edu Book Editor—Jon E. KEELEY Noteworthy Collections Editors—DiETER WILKEN, MARGRIET WETHERWAX Board of Editors Class of: 200 1—RobBEeRT PaTTERSON, San Francisco State University, San Francisco, CA PAuLA M. ScuirFMAN, California State University, Northridge, CA 2002—NorMAN ELLSTRAND, University of California, Riverside, CA Carta M. D’ Antonio, University of California, Berkeley, CA 2003—FREDERICK ZECHMAN, California State University, Fresno, CA Jon E. Keetey, U.S. Geological Service, Biological Resources Division, Three Rivers, CA 2004—Davip M. Woop, California State University, Chico, CA INGRID Parker, University of California, Santa Cruz, CA CALIFORNIA BOTANICAL SOCIETY, INC. OFFICERS FOR 2001—2002 President: BRucE BALDwin, Jepson Herbarium and Dept. of Integrative Biology, 1001 Valley Life Sciences Bldg. #2465, University of California, Berkeley, CA 94720. First Vice President: Rop Myatt, San José State University, Dept. of Biol. Sciences, One Washington Square, San José, CA 95192. rmyatt@email.sjsu.edu Second Vice President: PETER Fritscu, Dept. of Botany, California Academy of Sciences, Golden Gate Park, San Francisco, CA 94118-4599. pfritsch @calacademy.org Recording Secretary: DEAN KELCH, Jepson and University Herbarium, University of California, Berkeley, CA 94720. dkelch @sscl.berkeley.edu Corresponding Secretary: | SUSAN BAINBRIDGE, Jepson Herbarium, 1001 VLSB #2465, University of California, Berkeley, CA 94720-2465. 510/643-7008. suebain@SSCL.berkeley.edu Treasurer: Roy Buck, % University Herbarium, University of California, Berkeley, CA 94720. The Council of the California Botanical Society comprises the officers listed above plus the immediate Past President, R. JoHN LittLe, Sycamore Environmental Consultants, 6355 Riverside Blvd., Suite C, Sacramento, CA 95831; the Editor of Maprono; three elected Council Members: Bian Tan, Strybing Arboretum, Golden Gate Park, San Fran- cisco, CA 94122; JAMEs SHEvock, National Park Service, 1111 Jackson St., Suite 700, Oakland, CA 94607-4807. 510/ 817-1231; ANNE BRADLEY, USDA Forest Service, Pacific Southwest Region, 1323 Club Drive, Vallejo, CA 94592. abradley @fs.fed.us; Graduate Student Representative: KirsTEN JOHANNUS, Jepson Herbarium, University of California, Berkeley, CA 94720. This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). -Maprono, Vol. 48, No. 4, pp. 221-229, 2001 MARITIME CHAPARRAL COMMUNITY TRANSITION IN THE ABSENCE OF FIRE ERIC VAN DYKE! AND KAREN D. HOLL Department of Environmental Studies, University of California, Santa Cruz, CA 95064 JAMES R. GRIFFIN Hastings Reservation, University of California, Carmel Valley, CA 93924 ABSTRACT Maritime chaparral stands on California’s central coast are dominated by a number of endemic Arc- tostaphylos species and are the habitat for several other species of concern. Although chaparral is a fire- adapted vegetation type, maritime chaparral occurs in densely populated regions where fire suppression prevents most stands from burning. In 2000, we re-sampled vegetation at six locations in north Monterey County’s Prunedale sandhills that were sampled in 1975—1976 by Griffin (1978); this allowed us to document changes in community composition, canopy cover, and seedling abundance over a 25-year interval after more than 70 years of fire exclusion. Although species richness in the tree and shrub layers changed little between 1975-1976 and 2000, combined tree and shrub cover increased from 86 to 99%. Cover of Arctostaphylos pajaroensis J. Adams increased from 58 to 82%. Cover of Quercus agrifolia Nee and Heteromeles arbutifolia (Lindley) Roemer also increased significantly, whereas percent cover for most shrub species decreased, often dramatically. Species richness in the herb layer was markedly lower in the 2000 survey. Seedlings were rare under the dense canopy, although seedling abundance for Q. agrifolia and Mimulus aurantiacus Curtis increased. These results suggest that the long absence of fire in maritime chaparral stands may lead to dominance by one or two species and a gradual transition from chaparral to oak woodland. Land managers should consider the reintroduction of wildfire, or prac- tices that mimic the effects of fire, to assure the long-term survival of maritime chaparral vegetation communities. INTRODUCTION Large areas of California’s central coast are re- ported to have been covered with dense chaparral at the end of the nineteenth century (Cooper 1922). Today, only small, isolated fragments of northern and central maritime chaparral can be found grow- ing in well-drained sandy soils along ridgelines and on coastal terraces between Sonoma and Santa Bar- bara counties (Holland 1986). Each of these stands is dominated by one or more Arctostaphylos spe- cies, including about 20 that are narrowly distrib- uted endemics (Hickman 1993). Although chaparral is widely reported to be de- pendent on periodic burning for renewal (e.g., Wells 1962; Hanes 1988), the cool and foggy cen- tral coast has one of the lowest rates of lightning- caused fire in California (Greenlee and Langenheim 1990). Estimates of historic fire return intervals for the Monterey Bay area range from as short as 10 to as long as 100 years or more (Greenlee and Lan- genheim 1990; Moritz 1997), but none of these es- timates are presented with much confidence. Mod- ern fire suppression practices have greatly reduced the size and frequency of wildfires in these heavily populated areas (Greenlee and Langenheim 1990). 'Current address: Elkhorn Slough Foundation, P.O. Box 267, Moss Landing, CA 95039. E-mail: vandyke @ aromas.org Deviation from the natural fire frequency may alter the relative proportions of shrubs in the chap- arral canopy by favoring obligate seeding Arcto- staphylos species over crown sprouters (Keeley and Zedler 1978) and taller, longer-lived Arctostaphylos over Ceanothus, Adenostoma, or Salvia (Davis 1972; Davis et al. 1988). The long absence of fire may eventually favor crown sprouting species such as Quercus and Heteromeles over obligate seeders (Keeley 1992b; Zammit and Zedler 1993). Fire fre- quency also affects the composition of the chapar- ral understory, both through the direct effects of heat, smoke, and ash, and indirect effects such as reduced competition and herbivory (Sweeney 1956, Christensen and Muller 1975a; Keeley and Keeley 1987; Tyler 1996). Chaparral remnants in the coastal sandhills of north Monterey County between the communities of Pajaro and Prunedale are dominated by Arcto- staphylos pajaroensis J. Adams and include several other uncommon species (Table 1). These stands continue to be fragmented and degraded by agri- cultural conversion and residential development, and their preservation is considered a high priority by Monterey County and by conservation organi- zations (Monterey County Planning Department 1981; Elkhorn Slough Foundation 1999). Unfortu- nately, little is known about the long-term effects of changing disturbance regimes on this unusual vegetation community. ae TABLE 1. UNCOMMON PLANTS OF PRUNEDALE HILLS MARITIME CHAPARRAL. | California Department of Fish and Game | 2001: Skinner and Pavlik 1994. Species Arctostaphylos hookeri ssp. hookeri Arctostaphylos pajaroensis Ceanothus cuneatus var. rigidus Chorizanthe pungens var. pungens Ericameria fasciculata Piperia yadonit The objective of this study was to identify changes in community composition, canopy cover, and seedling abundance that occur in maritime chaparral stands during long periods of fire sup- pression and habitat fragmentation. The existence of field data from a 25-year old survey of maritime chaparral in the Monterey Bay region (Griffin 1978) provided a unique opportunity to compare historical vegetation data with current conditions. Most previous studies of chaparral dynamics in the absence of fire have relied on a chronosequence of sites (e.g., Zammit and Zedler 1988; Keeley 1992a, b) that may vary along abiotic or biotic gradients. Present extent of Arctostaphylos pajaroensis ++ 36°52'30" Elkhorn Slough Monterey Bay ¢ Prunedale + 36°45' 121°45' 121°37'30" Fic. |. Location of stands surveyed and present extent of Arctostaphylos pajaroensis. For descriptions of num- bered stands see Table 2. MADRONO [Vol. 48 Rarity! CNPS List 1B (rare, threatened, or endangered) CNPS List 1B CNPS List 4 (watch list) Fed. threatened; CNPS List 1B CNPS List 1B Fed. endangered; CNPS List 1B STUDY AREA Griffin (1978) surveyed maritime chaparral stands in the Monterey Bay region between Octo- ber 1975 and March 1977 using the Braun-Blan- quet minimal-area relevé method (Mueller-Dom- bois and Ellenberg 1974). Between five and ten plots were sampled in the least disturbed portions of each of these stands. Seven stands were within the range of Arctostaphylos pajaroensis in the Prunedale sandhills. In spring and summer 2000, we re-surveyed the seven Prunedale area stands (Fig. 1, Table 2). Grif- fin’s field data and sketch maps for five of the stands, along with historic aerial photographs, al- lowed us to re-locate 50 plots to within a few me- ters of the original. Of these, 21 no longer con- tained maritime chaparral due to development. In the two stands where there was insufficient infor- mation to permit exact re-location, we selected 18 plots with slope, aspect, and substrate matching the original. Plots were identified on topographic maps to avoid selection bias and adjusted in the field only as necessary to avoid edge effects from roads or other human disturbance. Davis (1972) estimated that approximately 50 years had elapsed since the last major fire in the Prunedale hills region by counting annual growth rings in mature A. pajaroensis. We verified that none of our study plots have burned during the past 70 years by examining historic vertical aerial pho- tographs of each stand taken at intervals between 5 and 12 years beginning in May 1931. Maritime chaparral stands in the Prunedale hills occur within a matrix of oak woodland and coastal sage scrub. The chaparral comprises two distinct plant associations. The 1932 Vegetation Type Map survey (US Forest Service 1932) distinguished a shorter ““dwarfed chamise type” and a taller “‘chap- arral type’’. Exposed plateaus, ridges, and sand- stone outcrops support a patchy chaparral that rare- ly exceeds one meter in height and where A. hook- eri G. Don ssp. hookeri and Adenostoma fascicu- latum Hook & Arn. are dominant. On slopes and in depressions below the ridgelines, a tall, dense A. pajaroensis canopy predominates. Boundaries be- tween these two chaparral types are typically quite abrupt. The majority of plots sampled for this study 2001 | VAN DYKE ET AL.: MARITIME CHAPARRAL TRANSITION 223 TABLE 2. PRUNEDALE HILLS MARITIME CHAPARRAL SAMPLING LOCATIONS. Number of plots Stand Description 1975-6 2000 | McGuffie Road 10 6 original plots; 4 plots lost to development 2 Vierra Canyon 10 3 original plots; 7 plots lost to development 3 Castroville Boulevard a 10 approximate plot locations 4 Manzanita Circle 10 10 original plots 5 Prunedale 10 10 original plots 6 Lewis Road 10 10 plots lost to development qi Hidden Valley 5 8 approximate plot locations are composed of the A. pajaroensis dominated as- sociation. Soils at the Prunedale hills are Arnold-Santa Ynez complex, a mixture of deep, excessively drained, slightly acid loamy sand derived from old marine dunes and soft, weathered sandstone out- crops (US Department of Agriculture 1978). Slopes vary from flat on plateaus and ridges up to 40 per- cent. Elevation ranges from 50 to 150 m. The dis- tance of Monterey Bay is between 7 and I1 km. Mean precipitation is 50 cm, falling mainly during the winter months (Monterey County Water Re- sources Agency). Summers are dry, moderated by frequent fog. METHODS We employed Griffin’s (1978) sampling tech- niques to facilitate comparison. Forty-seven 10 X 10 m square plots were delineated and all vascular plant species identified. Three height strata were distinguished: tree layer (rising above the shrub canopy), shrub layer (the chaparral canopy), and herb layer (the understory). At every plot, we re- corded an estimate of the average canopy height and the percentage of vegetation cover for all three layers. We ranked every species in each layer ac- cording to the Braun-Blanquet cover-abundance scale: “‘r’? = | individual, “*+’’ = few individuals, “1” = 1-5% cover, “2”. = 5—25% cover, “3” = 25-50% cover, “4”? = 50-75% cover, “‘5” = 75— 100% cover (Mueller-Dombois and Ellenberg 1974). A single species might occupy more than one layer. Adult Quercus, for example, contribute to the tree layer, saplings contribute to the shrub layer, and seedlings to the herb layer. In both sur- veys, all species of annual and perennial grass were combined into two categories. Additional species observed nearby but outside the plots were record- ed separately; these species are included in Appen- dix 1, but not in quantitative analyses. Nomencla- ture is according to The Jepson Manual (Hickman 1993). We calculated a “‘coefficient of community sim- ilarity’’ according to Jaccard (Mueller-Dombois and Ellenberg 1974) to compare species present in the tree and shrub layers among the plots of each sur- vey and between the 1975—1976 and 2000 surveys at individual plots. We also calculated differences in mean percent cover for the combined tree and shrub layers and for individual tree and shrub spe- cies between the plots of the two surveys. Because a Braun-Blanquet scale value represents a range of percentages, we used the median of each class (e.g., 62.5% for rank “4”, which represents S50—75% cover). The two lowest Braun-Blanquet scale val- ues represent number of individuals rather than cover; we chose cover percentages of 0.1% for rank “r’? and 0.5% for “*+’’. We applied paired f-tests to arcsine-transformed percentages to determine the significance of each change in tree and shrub cover (Sokal and Rohlf 1995). As an index for comparing community similarity based on percent cover, we calculated “‘percent similarity”? (Wolda 1981) be- tween the plots of each survey and between the two surveys. We calculated changes in the herb layer using Braun-Blanquet scale ranks rather than esti- mates of percent cover because converting abun- dance to cover for the two lowest classes could be misleading where overall percentages are low. For the 2000 data, we tested whether the number of species present at each stand was dependent on average canopy height or mean percent cover using linear regression. The 1975—1976 data did not in- clude canopy height, so comparison between the two surveys was not possible. Nearly all plots tend to south facing, although their slopes vary consid- erably. We used linear regression to test whether changes in species composition or canopy cover were dependent on slope or aspect. Variations in soil type, relative elevation, and distance inland were minimal between plots, so the effects of these variables were not tested. RESULTS A total of 20 plant species were present in the tree and shrub layers of all plots in the 1975-1976 survey; 19 species were present in 2000. All spe- cies encountered both in and near the sample plots are listed in Appendix |. Three shrub species that were uncommon in the earlier survey, Ceanothus B. pilularis © > e) oO = © a H. arbutifolia o 1975-6 2000 Fic. 2. Change in mean percent cover between 1975-6 and 2000 for selected species. * = P < 0.05; ** = P < 0.01; *** = P < 0.001 from paired t-test on arcsine-trans- formed percentages. Error bars indicate +1 standard error. Note different y-axis scales. dentatus Torrey & A. Gray, C.. thyrsiflorus Eschsch., and Ericameria ericoides (Less.) Jepson, were absent from the 2000 survey. The shrub Vac- cinium ovatum Pursh and the introduced tree Pinus radiata D. Don were not encountered in 1975-1976 and were rarely present in 2000. The average num- ber of species per plot was nearly unchanged, in- creasing from 6.2 to 6.4. The Jaccard index, which compares the number of species in common, showed greater similarity for plots between the two surveys (0.81) than among the 1975-1976 plots (0.65) or among the 2000 plots (0.69). Jaccard in- dices were lowest at the two stands where re-lo- MADRONO [Vol. 48 | cation was approximate, suggesting that these lo-| cations were somewhat mismatched. Mean percent cover of the combined tree and | shrub layers increased from 86 to 99% during the | 25-year period (t = 6.5, P < 0.0001). This increase | in canopy cover was chiefly due to a growing dom- | inance by Arctostaphylos pajaroensis in nearly ev- | ery plot. Mean cover of A. pajaroensis increased + from 58 to 82% (t = 7.34, P < 0.0001, Fig. 2). | Large increases in percent cover were also recorded | for Quercus agrifolia, Heteromeles arbutifolia, | Rhamnus californica Eschsch., and Garrya elliptica | Lindley, but the total contribution of these oak | woodland-associated sclerophylls remained small. Percent cover for all other shrubs decreased. Lead- | ing this decline was Salvia mellifera E. Greene, — which dropped from 6.0 to <0.5% (t = 4.48, P < | 0.0001). The percent similarity index confirmed a dramatic increase in homogeneity among the plots, | from 0.61 in 1975-6 to 0.92 in 2000. | In the herb layer, overall species richness de- | creased from 27 species in 1975-6 to 18 in 2000, and the average number of species per plot de- | creased from 2.8 to 1.7 (t = 2.54, P = 0.015, Table | 3). All five tree and shrub species that had seedlings | present in the earlier survey also had seedlings © present in 2000, and three of these, Mimulus au- rantiacus, Q. agrifolia, and Toxicodendron diver- silobum (Torrey & A. Gray) E. Greene, were more | numerous and widespread (Table 4). | The Braun-Blanquet cover-abundance rankings for most subshrubs, herbs, and grasses decreased, as did the number of plots where they were found | (x? = 6.1, P = 0.047 for all species, Table 4). The © fern Pteridium aquilinum (L.) Kuhn var. pubescens L. Underw. showed the largest decline, from up to 15 percent cover in each of 12 plots down to just a few individuals in four plots. More than half of the annual and perennial herbs counted in 1975— 1976 were no longer present in 2000. The number of species present in a plot was in- versely related to canopy height (F = 8.87, R? = 0.16, P = 0.005) and to percent cover (F = 9.14, R? = 0.17, P = 0.004). Regression revealed no sig- nificant relationship between number of species or percent cover and slope or aspect. Non-native species were not major constituents of intact maritime chaparral in the Prunedale hills. Introduced annual grasses such as Bromus spp. TABLE 3. HERB LAYER: TOTAL NUMBER OF SPECIES AND MEAN NUMBERS OF SPECIES PER PLOT. * = P < 0.05; *** = P < 0.001 from paired ¢-test. +1 standard error in parentheses. For species included in each category see Appendix 1. Total species Category 1975-6 Tree and shrub seedlings > Perennial subshrubs, herbs, and grasses 18 Annual herbs and grasses 4 All species 27 Species per plot 2000 Common 1975-6 2000 5 2) 0.27 (0.09) 0.73 (0.13)* 1] 8 2234 (O30) 0.94 (0.18)*** 2 | 0.15 (0.07) 0.04 (0.04) 18 14 2.78 (0.39) 17. (0224) 2001) VAN DYKE ET AL.: MARITIME CHAPARRAL TRANSITION 225 TABLE 4. HERB LAYER: NUMBER OF PLOTS IN EACH BRAUN-BLANQUET COVER-ABUNDANCE CLASS. For additional species included in each category see Appendix |. ° ° La | ’ ’ Category and selected species Tree and shrub seedlings Baccharis pilularis Mimulus aurantiacus Quercus agrifolia Rhamnus californica Toxicodendron diversilobum Perennial subshrubs, herbs, and grasses Gnaphalium spp. Helianthemum scoparium Lotus scoparius Marah fabaceus Pteridium aquilinum Annual herbs and grasses All species N COO OWODWOrF WKN ~ ey) were occasionally present in plots in the earlier study, as they were in 2000. No exotic trees, shrubs, or herbs were counted in 1975-1976, although Griffin (1978) noted that Carpobrotus edulis (L.) N.E.Br., Cortaderia jubata (Lemoine) Stapf, and Genista monspessulana (L.) L. Johnson were in- vading nearby disturbed areas. In 2000, C. jubata was widespread near all of the sampling locations, although only one individual appeared within the study plots. Introduced Pinus radiata grew near three of the sites, and a single sapling was present in one plot. Large numbers of Eucalyptus globulus Labill. saplings were present in chaparral near three of the stands in 2000. DISCUSSION Maritime chaparral stands in the Prunedale hills have undergone significant changes in community composition, canopy cover, and seedling abundance between 1975-1976 and 2000, a period during which fire has been excluded. In the 1970’s, the vegetation was patchy. Trees large enough to rise above the shrub layer were uncommon. Arctosta- phylos pajaroensis competed with several other shrubs for dominance. A variety of grasses and forbs contributed to a sparse but widespread herb layer under the broken canopy. Today, the tree and shrub layers approach 100 percent cover forming a dense, closed canopy. Arctostaphylos pajaroensis is now the overwhelming dominant, although Quer- cus agrifolia cover has also increased significantly. The understory is generally bare except for occa- sional Q. agrifolia and Mimulus aurantiacus seed- lings. Most herb layer species are restricted to the few remaining canopy gaps. All three of these trends, increased dominance by A. pajaroensis, loss of species diversity, and invasion by Q. agrifolia, may be attributed to the long absence of fire in the Prunedale hills. 1975-6 2000 : 19 ; cop peo” 4 0 10 25 0 2 0 2 0) 0) 0 0 4 7 0 0 0 2 13 0 0) 0 | 0 0 2 0 | 5 0 66 8 Ls 30 0 6 0 ps 0 0 2 0 O 0 0 9 0 2 3 0 0 0 ) 6 0 4 8 0) 4 0) 6 0 0 2 0 76 8 ie 57 O Arctostaphylos pajaroensis dominance. The dra- matic increase in A. pajaroensis cover, and similar- ly dramatic decreases for several other shrubs, like- ly result from the greater relative height of this long-lived species when freed from periodic de- struction by wildfire. Davis (1972) noted that, with sufficient time and in the absence of fire, the stature of A. pajaroensis exceeds that of all other associ- ated species except Q. agrifolia. This competitive advantage is largely due to the adaptation of “‘bark striping’, where the amount of living tissue on stems in the lower, shaded portions of the shrub is minimized while providing structure to support the growth of new leaves and branches in full sunlight above (Davis 1973). We encountered a tangle of dead Salvia, Adenostoma, Ceanothus, and other shrubs in plots wherever the canopy exceeded two meters in height, suggesting the fate of these shorter species as they become overtopped and shaded by A. pajaroensis. McPherson and Muller (1967) de- scribed a similar competition for light in mature coastal chaparral in which shorter Salvia were pro- gressively killed by taller shrubs. The dominant in this case was Ceanothus cuneatus (Hook.) Nutt., a species that also exhibits the bark striping strategy (Keeley 1975). Interestingly, in the Prunedale hills C. cuneatus var. rigidus (Nutt.) Hoover is one of the species that is overtopped and killed by taller A. pajaroensis. Herb layer composition. Declining species rich- ness and abundance in the understory are probably consequences of greater canopy height and density, and the resulting shade and litter accumulation, rather than differences in precipitation. Griffin (1978) suggested that the herb layer did not develop fully in 1975-1976, a drought year, yet he recorded a wider diversity of species growing under and among the trees and shrubs than we encountered in 2000, a normal rainfall year (Monterey County Wa- ter Resources Agency). 226 MADRONO [Vol. 48 | Many studies have commented on the absence of herbs and shrub seedlings under mature, undis- turbed chaparral (e.g., Sampson 1944; Christensen and Muller 1975b; Hanes 1988). Competition for light, moisture, and nutrients and high levels of her- bivory are common explanations (e.g., McPherson and Muller 1969; Schlesinger and Gill 1980; Swank and Oechel 1991; Tyler 1996; Keeley 2000). Arc- tostaphylos may also be a source of allelopathic substances, inhibiting the establishment of seed- lings under the chaparral canopy (Muller et al. 1968, Chou and Muller 1972). The only herbaceous species that showed a significant increase in the shade under dense A. pajaroensis was Marah fa- baceus (Naudin) E. Greene, a vine that resprouts annually from a large underground tuber and can quickly reach sunlight in the shrub canopy (Schlis- ing 1969). Two annual herbs, Chorizanthe pungens Benth. var. pungens and Navarretia hamata E. Greene, and three perennial shrubs or subshrubs, Eriophyllum confertiflorum (DC.) A. Gray, Helianthemum sco- parium Nutt., and Lotus scoparius (Nutt.) Ottley, were restricted to openings in the canopy in 2000. Canopy gaps are important for seed germination and seedling establishment and for maintaining the seed banks of many chaparral species (Davis et al. 1989; Zammit and Zedler 1994; Odion and Davis 2000). As tree and shrub layer cover has increased, gaps have grown increasingly rare. Quercus agrifolia invasion. An increase in Q. agrifolia canopy cover and seedling abundance in Prunedale hills maritime chaparral stands is consis- tent with Keeley’s (1992a, b) conclusion that spe- cies like Q. agrifolia that are capable of regenera- tion from root crowns will only produce seedlings after a long fire-free period, and only under a dense canopy in a heavy accumulation of leaf litter. Cal- laway and D’ Antonio (1991) found that shrubs fre- quently serve as “‘nurse plants’’, providing micro- habitat conditions that facilitate the establishment of Q. agrifolia seedlings that will eventually over- top and kill their hosts. We frequently encountered mature Arctostaphylos skeletons in the shaded un- derstory of oak woodland immediately adjacent to chaparral stands, an observation also reported by Davis (1972). Boundaries between chaparral, coastal sage scrub, and oak woodland communities are dynamic and highly dependent on fire frequency (Gray 1983; Callaway and Davis 1993). Live oak woodland is widely characterized as the successional climax for maritime chaparral stands in the absence of fire (e.g., Wells 1962; Davis 1972; McBride and Stone 1976; Griffin 1978). Several studies suggest that a gradual transition to oak woodland is underway at various locations on the central coast as Q. agrifolia invades long unburned areas (Davis et al. 1988; Callaway and Davis 1993; Mensing 1998; White 1999). Various studies have also proposed that the segregation of maritime chaparral and oak wood- fire. As was the case in 1975-1976, no recruitment | land communities is at least partially due to edaphic | differences (e.g., Wells 1962; Cole 1980; Davis et | al. 1988). Davis (1972) and Griffin (1978) both conclude that Q. agrifolia is successional to the more mesic A. pajaroensis association in the ab- | sence of fire, while conversion from chaparral to | oak woodland will progress more slowly, and per- | haps even be arrested, in the more xeric A. hookeri association. Even on the harshest sandstone ridges, Q. agrifolia is occasionally present, but only aq seedlings and shrub-sized individuals. Long-term vegetation changes in the absence of of obligate seeding Arctostaphylos or Ceanothus | species was observed in any of our study plots in > 2000. The only seedlings of these taxa that we en- | countered anywhere during this survey were a few dozen young A. pajaroensis and C. dentatus that had established near one of the plots in a small area | that had burned two years previously. These results | are not surprising as seeds of several species in | these two genera are reported to require heating or | exposure to smoke or charate to stimulate germi- | nation (Keeley 1987; Keeley and Keeley 1987; Keeley and Fotheringham 1998). The effect of long fire-free periods on chaparral | is a topic of considerable discussion. Keeley | (1992a) has persuasively argued that pejorative | terms such as decadent and senescent that are often | applied to long-unburned chaparral stands (e.g., | Sampson 1944; Hanes 1988) are inappropriate for describing a gradual successional shift from obli- | gate seeding Arctostaphylos and Ceanothus species | to crown sprouting Quercus and Heteromeles. Keeley (1992a) also notes the importance of | variable fire regimes to maintain equilibrium in spe- | cies composition. The long absence of fire may lead | to local extinction of certain species if soil seed | banks become exhausted. The length of time that | seeds remain viable is unknown for most maritime | chaparral species (Tyler and Odion 1996). Further | study of seed bank longevity is needed to under- stand the risk to species of concern such as C. cu- neatus var. rigidus, Chorizanthe pungens var. pun- gens, Ericameria fasciculata (Eastw.) J.R Macbr., and Piperia yadonii R. Morgan & J. Ackerman. Because the majority of Prunedale hills maritime chaparral has not burned for at least 70 years, we feel that concern for the future of A. pajaroensis, C. cuneatus var. rigidus, and the other plants that characterize this unusual vegetation community is warranted. Neither they nor their seeds can survive forever. If wildfire continues to be excluded, the composition of these stands will undoubtedly be very different in the future. Management implications. Griffin (1978) con- cluded with this warning: “Pressures for develop- ment are so great around Monterey Bay that mari- time chaparral stands need legal protection to sur- vive. No adequate sample of chaparral near Prune- dale has formal protection now.” Since this writing, several important Prunedale hills stands have re- ceived protection as conservation lands. At the same time, additional chaparral acreage is lost ev- ery year and development pressures continue to grow. Of the seven original 1975-1976 sampling areas, three are highly modified with only frag- ments of undisturbed chaparral remaining, two had remained relatively intact but currently have sub- division plans underway, and two are within the boundaries of Manzanita County Park where sports facility expansion is proposed. Areas dominated by the low A. hookeri chaparral association are generally unsuited for agriculture because of their shallow soils and remain as frag- ments along ridgelines throughout the Prunedale hills. Monterey County land use policies discourage development on ridges and encourage the dedica- tion of scenic and conservation easements on un- buildable portions of subdivisions that contain these maritime chaparral fragments (Monterey County Planning Department 1981). Stands of the taller A. pajaroensis association, because they occur on the gentler south-facing slopes and deeper soils favored for cultivation, have been lost to agricultural con- version over a period of many decades. In recent years, a shortage of land suitable for residential de- velopment in north Monterey County has acceler- ated the destruction of this chaparral type. Alter- native management strategies are needed for these two different chaparral types. Loss of species diversity caused by shading is associated with canopy height, thus with the A. pa- jaroensis dominated chaparral type. Invasion by Q. agrifolia is also rapid in these more mesic sites. For these areas, the introduction of prescribed burning, or perhaps mechanical disturbance with smoke or charate treatment, may be necessary to open the canopy, facilitate seedling establishment, and slow the advance of oaks. Enhancement of the seed bank with stockpiled chaparral soil, in conjunction with burning, could be necessary in degraded areas (Odion 1995). Unfortunately, non-native species of- ten follow disturbance in chaparral (Zedler and Scheid 1988; D’Antonio et al. 1993; Tyler and Odion 1996; Holl et al. 2000), so a control program would likely be required. In the low A. hookeri chaparral type, where gaps are frequent and Quer- cus grow slowly, prevention of any kind of distur- bance might be the more appropriate management strategy. Land protection is the essential first step toward conserving increasingly rare maritime chaparral communities. Conservation efforts should focus on stands that include both chaparral associations. Ac- tive land stewardship will also be necessary in or- der to conserve the full complement of native plant species. Management strategies should attempt to maximize diversity by maintaining a variety of suc- cessional stages and canopy heights including bare rock and soil, patchy mixed chaparral, closed Arc- tostaphylos canopy, and mixed chaparral/oak VAN DYKE ET AL.: MARITIME CHAPARRAL TRANSITION 220 woodland. The effects of a modified disturbance cycle in the Prunedale hills will need to be under- stood in order to ensure the survival of this unusual vegetation and to minimize the loss of endemic spe- GieS. ACKNOWLEDGMENTS We thank Mark Stromberg for locating and loaning Griffin’s field notes and maps; Mark Brown, Bill Crane, Pete Jacobsen, Mark Silberstein, and Ron Whitehead for access to chaparral stands; and Michael Loik and Kerstin Wasson for helpful comments. LITERATURE CITED CALIFORNIA DEPARTMENT OF FISH AND GAME. 2001. Nat- ural diversity database version 2.1.2 (March 5, 2001). California Department of Fish and Game, Sacramento CA. CALLAWAY, R. M. AND C. M. D’ANTONIO. 1991. Shrub facilitation of coast live oak establishment in central California. Madrono 38:158—169. AND FE. W. Davis. 1993. Vegetation dynamics, fire, and the physical environment in coastal central Cal- ifornia. Ecology 74:1567—1578. CuHou, C. AND C. H. MULLER. 1972. Allelopathic mecha- nisms of Arctostaphylos glandulosa var. zacaensis. American Midland Naturalist 88:324—347. CHRISTENSEN, N. L. AND C. H. MULLER. 1975a. Relative importance of factors controlling germination and seedling survival in Adenostoma chaparral. American Midland Naturalist 93:71—78. ——— AND C. H. MULLER. 1975b. Effects of fire on fac- tors controlling plant growth in Adenostoma chapar- ral. Ecological Monographs 45:29—55. CoLe, K. 1980. Geologic control of vegetation in the Pur- isima hills, California. Madrono 27:79—-89. Cooper, W. S. 1922. The broad-sclerophyll vegetation of California: an ecological study of the chaparral and its related communities. Carnegie Institution of Wash- ington Publication Number 319, Washington, DC D’ ANTONIO, C. M., D. C. ODION, AND C. M. TYLer. 1993. Invasion of maritime chaparral by the introduced suc- culent Carpobrotus edulis: the roles of fire and her- bivory. Oecologia 95:14—21. Davis, C. B. 1972. Comparative ecology of six members of the Arctostaphylos andersonii complex. Ph.D. dis- sertation. University of California, Davis. . 1973. “Bark striping” in Arctostaphylos (Erica- ceae). Madrono 22:145—149. Davis, EF W., D. E. HICKSON, AND D. C. ODION. 1988. Composition of maritime chaparral related to fire his- tory and soil, Burton Mesa, Santa Barbara County, California. Madrono 35:169—195. ———., M. I. BORCHERT, AND D. C. ODION. 1989. Estab- lishment of microscale vegetation pattern in maritime chaparral after fire. Vegetatio 84:53—67. ELKHORN SLOUGH FOUNDATION/THE NATURE CONSERVANCY. 1999. Elkhorn Slough watershed conservation plan. Elkhorn Slough Foundation, Moss Landing, CA. GRAY, J. T. 1983. Competition for light and a dynamic boundary between chaparral and coastal sage scrub. Madrono 30:43—49. GREENLEE, J. M. AND J. H. LANGENHEIM. 1990. Historic fire regimes and their relation to vegetation patterns in the Monterey Bay area of California. American Midland Naturalist 124:239—253. 228 GRIFFIN, J. R. 1978. Maritime chaparral and endemic shrubs of the Monterey Bay region, California. Ma- drono 25:65—112. HANgEsS, T. L. 1988. California chaparral. Pp. 418—469 in M. G. Barbour and J. Major (eds.), Terrestrial vege- tation of California. California Native Plant Society, Sacramento, CA. HICKMAN, J. C. 1993. The Jepson manual: higher plants of California. University of California Press, Berke- ley, CA. Ho i, K. D., H. N. STEELE, M. H. FUSARI, AND L. R. Fox. 2000. Seed banks of maritime chaparral and aban- doned roads: potential for vegetation recovery. Jour- nal of the Torrey Botanical Society 127:207—220. HOLLAND, R. F 1986. Preliminary descriptions of the ter- restrial natural communities of California. California Department of Fish and Game, Natural Heritage Di- vision, Sacramento, CA. KEELEY, J. E. 1975. Longevity of nonsprouting Ceano- thus. American Midland Naturalist 93:504—507. . 1987. Role of fire in seed germination of woody taxa in California chaparral. Ecology 68:434-—443. . 1992a. Recruitment of seedlings and vegetative sprouts in unburned chaparral. Ecology 73:1194- 1208. 1992b. Demographic structure of California chaparral in the long-term absence of fire. Journal of Vegetation Science 3:79—90. . 2000. Chaparral. Pp. 204—253 in M. G. Barbour and W. D. Billings (eds.), North American terrestrial vegetation. Cambridge University Press, New York, NY. AND C. J. FOTHERINGHAM. 1998. Smoke-induced seed germination in California chaparral. Ecology 79: 2320-2336. AND S. C. KEELEY. 1987. Role of fire in the ger- mination of chaparral herbs and suffrutescents. Ma- drono 34:240—249. AND P. H. ZEDLER. 1978. Reproduction of chap- arral shrubs after fire: a comparison of sprouting and seeding strategies. American Midland Naturalist 99: 142-249. McBRIDE, J. R. AND E. C. STONE. 1976. Plant succession on the sand dunes of Monterey Peninsula, California. American Midland Naturalist 96:118—132. McPHERSON, J. K. AND C. H. MULLER. 1967. Light com- petition between Ceanothus and Salvia shrubs. Bul- letin of the Torrey Botanical Club 94:41-—55. AND . 1969. Allelopathic effects of Aden- ostoma fasciculatum, “‘chamise’’, in the California chaparral. Ecological Monographs 39:177—198. MENSING, S. A. 1998. 560 years of vegetation change in the region of Santa Barbara, California. Madrono 45: 1-11. MONTEREY COUNTY PLANNING DEPARTMENT. 1981. Mon- terey County, California North county land use plan: local coastal program. Monterey County Planning Department, Salinas, CA. Moritz, M. A. 1997. Analyzing extreme disturbance events: fire in Los Padres National Forest. Ecological Applications 7:1252—1262. MUELLER-DomMBo!ISs, D. AND H. ELLENBERG. 1974. Aims and methods of vegetation ecology. John Wiley and Sons, N.Y. MULLER, C. H., R. B. HANAWALT, AND J. K. MCPHERSON. 1968. Allelopathic control of herb growth in the fire cycle of California chaparral. Bulletin of the Torrey Botanical Club 95:225-—231. MADRONO [Vol. 48 | OpIoNn, D. C. 1995. Effects of variation in soil heating i during fire on patterns of plant establishment and re- | growth in maritime chaparral. Ph.D. dissertation. Uni- | versity of California, Santa Barbara. AND E W. Davis. 2000. Fire, soil heating, and the formation of vegetation patterns in chaparral. Ecolog- ical Monographs 70:149—-169. SAMPSON, A. W. 1944. Plant succession on burned chap- | arral lands in northern California. University of Cal- ; ifornia Agricultural Experiment Station Bulletin 685. | SCHLESINGER, W. H. AND D. S. GILL. 1980. Biomass, pro- ) duction, and changes in the availability of light, wa-_ ter, and nutrients during development of pure stands | of the chaparral shrubs, Ceanothus megacarpus, after fire. Ecology 61:781-789. | SCHLISING, R. A. 1969. Seedling morphology in Marah | (Cucurbitaceae) related to the Californian Mediterra- nean climate. American Journal of Botany 56:556— | 560. SKINNER, M. W. AND B. M. PAVLIK. 1994. CNPS inventory | of rare and endangered vascular plants of Califor- | nia—Sth ed. CNPS Press, Sacramento, CA. SOKAL, R. R. AND E J. ROHLF. 1995. Biometry. W. H. | Freeman, N.Y. SWANK, S. E. AND W. C. OECHEL. 1991. Interactions among the effects of herbivory, competition, and re- | source limitation on chaparral herbs. Ecology 72: 104-115. SWEENEY, J. R. 1956. Responses of vegetation to fire: a study of the herbaceous vegetation following chap- arral fires. University of California Publications in Botany 28:143-216. TyYLer, C. M. 1996. Relative importance of factors con- tributing to postfire seedling establishment in mari- time chaparral. Ecology 77:2182—2195. AND D. C. ODION. 1996. Ecological studies of Morro Manzanita (Arctostaphylos morroensis). Re- port prepared for the California Department of Fish and Game, Endangered Plant Program. Sacramento, CA. US DEPARTMENT OF AGRICULTURE. 1978. Soil survey of Monterey County, California. US Department of Ag- riculture, Soil Conservation Service. Washington, DC US Forest SERVICE. 1932. Vegetation type map of Cali- fornia. Sheet 105A, San Juan Bautista Quadrangle. US Forest Service, Forest and Range Experiment Sta- tion. Berkeley, CA. WELLS, P. V. 1962. Vegetation in relation to geological substratum and fire in the San Luis Obispo quadran- gle, California. Ecological Monographs 32:79-103. White, K. L. 1999. Revisiting native Pinus radiata forests after twenty-nine years. Madrono 46:80—87. Wo pA, H. 1981. Similarity indices, sample size and di- versity. Oekologia 50:296—302. ZAMMIT, C. A. AND P. H. ZEDLER. 1988. The influence of dominant shrubs, fire, and time since fire on soil seed banks in mixed chaparral. Vegetatio 75:175—87. AND P. H. ZEDLER. 1993. Size structure and seed production in even-aged populations of Ceanothus greggii in mixed chaparral. Journal of Ecology 81: 499-511. AND P. H. ZEDLER. 1994. Organisation of the soil seed bank in mixed chaparral. Vegetatio 111:1—16. ZEDLER, P. H. AND G. A. SCHEID. 1988. Invasion of Car- pobrotus edulis and Salix lasiolepis after fire in a coastal chaparral site in Santa Barbara County, Cali- fornia. Madrono 35:196—201. ! | 2001] VAN DYKE ET AL.: MARITIME CHAPARRAL TRANSITION ] | | ; APPENDIX |. SPECIES ENCOUNTERED AT PRUNEDALE HILLS MARITIME CHAPARRAL SAMPLING LOCATIONS. [1] = present in plots 1975-6, [2] = present in plots 2000, [3] = present in plots both surveys. Species without numbers in brackets were observed near but never in plots. Trees: Eucalyptus globulus (non-native) Pinus radiata (introduced) [2] Quercus agrifolia [3] Quercus wislizenii Shrubs: Adenostoma fasciculatum [3] Arctostaphylos hookeri ssp. hookeri [3] Arctostaphylos pajaroensis [3] Arctostaphylos tomentosa ssp. crustacea [3] Artemisia californica Baccharis pilularis ssp. consanguinea [3] Ceanothus cuneatus var. rigidus [3] Ceanothus dentatus [1] Ceanothus thyrsiflorus [1] Chrysolepis chrysophylla Dendromecon rigida [3] Ericameria ericoides [1] Ericameria fasciculata [3] Eriophyllum confertiflorum [3] Garrya elliptica [3] Genista monspessulana (non-native) Heteromeles arbutifolia [3] Lepechinia calycina Mimulus aurantiacus [3] Pickeringia montana [3] Rhamnus californica [3] Salvia mellifera [3] Solanum umbelliferum Symphoricarpos mollis Toxicodendron diversilobum [3] Vaccinium ovatum [2] Perennial subshrubs, herbs, and grasses: Achillea millefolium [1] Calochortus albus Cardionema ramosissimum Carex spp. [1] Carpobrotus edulis (non-native) Castilleja foliolosa Chlorogalum pomeridianum Cortaderia jubata (non-native) [2] Dichelostemma capitatum Dudleya lanceolata [3] Eriogonum nudum Galium californicum [1] Gnaphalium spp. [3] Helianthemum scoparium [1] Horkelia cuneata |1] Tris douglasiana Lessingia filaginifolia Lomatium sp. [1] Lotus scoparius [3] Lupinus spp. Marah fabaceus [2] Oxalis albicans ssp. pilosa Pedicularis densiflora [3] Pellaea mucronata {1} Pentagramma triangularis [3] perennial grass [3] Piperia yadonii Polygala californica [3] Prteridium aquilinum var. pubescens [3] Rosa gymnocarpa [2] Rubus ursinus [1] Rupertia physodes Scrophularia californica [1] Scutellaria tuberosa Sisyrinchium bellum Solidago sp. [1] Zigadenus fremontii var. fremontii Annual herbs and grasses: annual grass [1] Camissonia spp. Chorizanthe pungens var. pungens [2] Cryptantha sp. Hemizonia sp. [1] Linaria canadensis Lupinus spp. Madia sp. [1] Navarretia hamata |3| N MADRONO, Vol. 48, No. 4, pp. 230-235, 2001 THE EFFECTS OF LITTER AND TEMPERATURE ON THE GERMINATION OF NATIVE AND EXOTIC GRASSES IN A COASTAL CALIFORNIA GRASSLAND SALLY A. REYNOLDS, JEFFREY D. CORBIN* AND CARLA M. D’ ANTONIO Department of Integrative Biology, 3060 VLSB University of California, Berkeley CA, 94720 ABSTRACT Through their effects on seed germination, accumulation of plant litter and temperature may play a role in the invasion of coastal California grasslands by exotic annual and perennial grasses. Germination of native and exotic grasses was examined as a function of both litter cover and temperature. When species were grouped by life form (native perennial grass vs. exotic perennial grass), exotic species germinated at consistently higher rates than native species. Individual species, however, varied in their response to litter addition. While one exotic perennial species, Festuca arundinacea, maintained germi- nation rates significantly higher than native species’ across three levels of litter cover, the other exotic perennial species, Holcus lanatus, showed no advantage over native species in the presence of a heavy litter layer. Exotic annual grasses had significantly higher germination rates than native perennial or exotic perennial grasses in laboratory growth chambers. Decreasing the average fall temperature in laboratory growth chambers by 5°C significantly reduced the germination percentages of Bromus diandrus and F. arundinacea relative to other species. The remaining two exotic annual grasses, Avena barbata and Vulpia myuros, were consistently the first seeds to germinate. Grouping species according to life form masked germination responses of individual species that otherwise provide insight into the potential role of ger- mination conditions in community composition of coastal grasslands in California. INTRODUCTION The invasion of exotic grasses has substantially altered the species composition of California grass- land ecosystems. Currently, seven million hectares in California are vegetated by European annual grasses, yet it is thought that, prior to European settlement, much of this area was dominated by pe- rennial bunchgrasses (Burcham 1970; Heady et al. 1988; Baker 1989). Native perennial-dominated grasslands that have resisted invasion by European annual species are currently rare in California, though several native grasslands have persisted in northern coastal California (Ornduff 1974; Hektner and Foin 1977; Dremann 1988). Recently, invasion by exotic perennial grass species such as Holcus lanatus L. and Festuca arundinacea (Schreber) has further threatened these remnant native grasslands (Elliot and Wehausen 1974). Exotic perennial spe- cies may be of equal or greater threat than exotic annual species since their larger investment in veg- etative structures enables the maintenance of dom- inance once they become established (Jackson and Roy 1986). The importance of early life stages in exotic in- vasions has been demonstrated in a variety of eco- logical systems (Parker 2000; Ruiz et al. 2000). Germination and seedling establishment are likely to be especially crucial phases in the invasion of grasslands by long-lived perennial species (Fowler 1986), though little is known about the factors that * Corresponding Author: Corbin @socrates.berkeley.edu influence the seed germination of exotic perennial species in California. Climatic conditions, especial- ly temperature and rainfall at the time of seed ger- mination, have been shown to play a role in year- to-year variation in species composition of exotic annual grass- and exotic forb-dominated habitats in California (Heady 1958; Pitt and Heady 1978; Jackson and Roy 1986; Young and Evans 1989). Soil disturbances, such as those created by gopher activities, provide open spaces that facilitate both native and exotic seedling establishment in Califor- nia (Platt 1975; Hobbs and Mooney 1985; Peart 1989; Kotanen 1996). The accumulation of plant litter, which has been shown to have both positive and negative effects on seed germination, also af- fects species composition of grasslands (Young et al. 1971; Hamrick and Lee 1987; Facelli and Pick- ett 1991la, b; Foster and Gross 1998). These factors, climatic conditions, availability of openings at the soil surface, and litter quantities, are likely to vary across the landscape and from year to year, and could influence the ability of exotic perennial spe- cies to invade grassland habitats. To examine the response of exotic perennial grass species to variation in micro-environmental and climatic conditions, we compared the germi- nation rates of native and exotic grass species under three levels of litter addition and at three different temperatures. Our experimental design allowed us to compare the germination responses of each spe- cies across a range of environmental conditions, and to examine conditions influencing germination and establishment of exotic species. We used seeds 3 | | 2001] REYNOLDS ET AL.: PERENNIAL GRASS SPECIES GERMINATION 251 | from four native perennial grass species, three com- -mon exotic annual species and two of the most | abundant exotic perennial grasses in this part of California, H. lanatus and F. arundinacea. treatment plot contained six replicate subplots for each species, plus six unplanted subplots to control for germination from the seed bank or residual seed rain. The number of grass seeds emerging from the METHODS Study site. The field portion of this study was conducted at Tom’s Point, a private nature preserve adjacent to Tomales Bay in northern Marin County (38°13'N, 122°57'W). The vegetation of this coastal prairie community consists of European annual grasses and forbs with interspersed stands of native perennial bunchgrasses. The site is located on sandy loam soils and has been free of livestock grazing for at least 30 years. The climate is Medi- terranean with an average annual temperature of 16°C, dropping only slightly in January and Feb- ruary. Rainfall totals are approximately 800 mm/ year, falling predominantly between October and April. Coastal fog present in the summer months moderates the summer drought. Deschampsia caes- pitosa ssp. holciformis (C. Pres] and W.E. Lawr), Festuca rubra L., Calamagrostis nutkaensis (C. Presl and Steudel), and Nassella pulchra (A. Hitch. and Barkworth) are the most common native spe- cies at the site. The most common exotic perennial species are F. arundinacea and H. lanatus, and the most common exotic annual species are Avena bar- bata L., Bromus diandrus (Roth), and Vulpia myu- ros var. myuros (c. Gmelin). All seeds for the experiment were collected in either July 1998 or July 1999. Seeds collected in 1998 were stored at room temperature for one year. Seeds collected in 1999 were stored at room tem- perature for one month. Lab trials prior to experi- mentation confirmed the viability of both sets of seeds. We used AH. lanatus leaf and culm litter be- cause current invasion of this species into both ex- otic annual and native perennial-dominated grass- lands is resulting in large quantities of previously non-existent litter that appears to be relatively per- sistent in coastal grasslands. Litter was collected in September 1999. Litter experiment. Seeds were added to three | m X 1.5 m experimental plots in October 1999 in an area of the grassland that had been cleared of background vegetation (predominantly Conium ma- culatum L.) in 1998 and again in 1999 just before planting. There was no gradient in soil moisture content (10 cm depth), inorganic nitrogen concen- trations, or annual grass germination rates in the area where our plots were established (Corbin and D’ Antonio, unpublished data). Germination rates of the four native and two exotic perennial species were compared across three litter treatment plots: bare soil, 1 cm of H. lanatus litter (200 g spread evenly over the plot), and 3 cm of A. lanatus litter (450 g spread evenly over the plot). Within each plot, 30 seeds of each of the six perennial species were planted in 10 cm X 10 cm subplots. Each control plots was very low in all treatments (mean + SD: 1.0 + 1.0 seeds), and we concluded that there was no substantial seed bank for any of the species. Six soil cores (2 cm diameter = — | | ! al | > a xX) Animas RSC Barker 2 Bird Rock | Bird Rock 2 Danthonia RSC Diablo SLH Olson Hill Fish Ranch | Fish Ranch 2 Ft. Ord Torro Jade Flat Molera Piedras LH Piedras | Piedras 2 Plaskett Ridge 240 Point Lobos (Heuer 1994). In 1991, 80 inland Nas- sella prairies in the Sierra de Salinas were sampled (Stromberg and Griffin 1996). Species richness de- pends on the scale of sampling (Gross et al. 2000) sO comparisons to other studies were made at sim- ilar scales (1 sq. m). METHODS Sampling was done at the peak of the flowering season, from mid-April through May in 1996 and 1997. At each grassland stand (Fig. 1), we chose a homogenous area and flagged a 50 m X 20 m plot (O.1 ha) with the long axis parallel to topographic contours. We recorded presence of all plant species we could discover on the larger plots. Starting from the midpoint of the short axis, we placed steel quadrats (20 cm X 50 cm; 0.1 m7?) at 2.5 m intervals along the 50 m midline of the 0.1 ha plot (20 quad- rats/plot). For each quadrat, we recorded a cover class for canopy cover of all plant species we ob- served (Daubenmire 1959). Quadrats were painted to facilitate recognition of Daubenmire cover class- es. A cover value of 0.01 was assigned to each species seen only in the larger plot and not seen in any smaller quadrat. Cover for each species at each stand (site) was calculated by averaging midpoints of the cover classes assigned to each of species in the 20 quadrats. Stands were at least 200 m apart, and more often, many kilometers apart. For each stand, we recorded aspect, elevation, distance in- land from the coast, latitude and longitude (UTM), total number of plant species (Hickman 1993), soil type, land form, number of grasses and forbs, and numbers of perennials and annuals, native and ex- otic. Species were counted based on scores in suc- cessive quadrats on original field sheets; one coastal terrace data sheet was inadvertently lost after tran- scribing summary numbers so the species tally by area for one coastal terrace is missing. Nine soil series were included (Cook 1978; Ernstrom 1984). Land forms of coastal terrace prairies include: 1) coastal terraces immediately adjacent to the ocean that are almost level; 2) grasslands on the sides of isolated bald hills arising inland and up at least 10 m from the terraces, sometimes locally known as ‘““potreros’’; and 3) drier, inland ridges well over 100 m above the coastal terraces and bases of the inland mountain ranges. Staview 5.0 (SAS) was used for statistical comparisons. Bonferroni/Dunn post-hoc tests were included to show individual dif- ferences in pairwise comparisons (P = 0.05). We used several methods to order stands based on spe- cies composition and with regard to measures of diversity, including CCA, DCA and Bray-Curtis or- dinations (PC-ORD ver 4) (ter Braak 1987a, b; McCune and Mefford 1999). Stand coordinates in our ordination were based on species data. We dropped species that occurred in only one stand with a cover value less than 5% in order to reach a numerical solution for CCA. Computational prob- MADRONO lems (Tausch et al. 1995) have been addressed, and | the method we used is inherently robust (Leps and | Hadincova 1992). Species composition of the herb layer was ob- | served using similar methods in 46 stands of Mon- | terey pine forests. Discovery of archival records (K. | White, unpublished data, Hastings archives) al- | lowed us to include observations from 40 to 80 | quadrats (20 cm X 50 cm; 0.1 m?’) that were read | as above for plant cover. A larger area of 0.1—1 ha | was then searched for additional species present and each was recorded (K. White, unpublished data; Hastings archives). Discovery of additional comparable data allowed [Vol. 48 | | us to include comparisons to coastal shrub com- | munities (Heuer 1994). Coastal shrub communities, | often adjacent to coastal terrace prairies, were sam- | pled with 16 sq. m quadrats at 141 locations, again | at the peak of the flowering season, in 1993. No larger sampling areas were surveyed for additional | species (Heuer 1994) and only the larger (16 sq. m) quadrats were used. Inland Nassella prairies were described in detail previously (Stromberg and Griffin 1996). We in- cluded inland prairies in this study to examine the larger scale differences across the landscape as in- land Nassella prairies share Nassella and other spe- cies with coastal terrace prairies, but occur at a dri- er, inland part of an environmental continuum oc- cupied by native grasslands in coastal California. Data sets from this study will be made available (ESA Ecological Archives or NRS archives). RESULTS Average species richness varies with the area sampled (Fig. 2). For individual sites, most reach an asymptote by about 2 m (Fig. 2a). No definitive asymptote is reached for the average coastal terrace prairie/Monterey pine forest (CTP) or (MPF). Spe- cies counts at 0.1 ha represents the best estimates for total species richness. Mean species numbers between all pairs are significantly different (paired t-tests, P < 0.001) for comparisons at | sq. m. and at O.1 ha (Table 1). This comparison of species richness with area leads to an interesting observation on the effects of a major human-directed use of the ecosystem. In a previous analysis of inland Nassella prairie stands with and without active grazing by domestic cattle, significantly fewer plant species were observed in grazed stands (Stromberg and Griffin 1996) based on areas of 0.1 ha. In this analysis of species num- ber at a smaller sampling scale (1 sq. m), this pat- tern in species richness was reversed and is clearly dependent on scale (Fig. 2). Coastal grasslands have much greater species richness in comparison to inland Nassella grass- lands, coastal pine, or coastal scrub plant commu- nities. A total of 82 species of grasses or sedges and 258 species of forbs (340 total) were found in | | | 2001] TABLE 1. STROMBERG ET AL.: COASTAL CALIFORNIA GRASSLANDS 241 COMPARISONS OF SPECIES DIVERSITY BETWEEN COASTAL TERRACE PRAIRIES (CTP), INLAND NASSELLA PRAIRIES (INP) AND MONTEREY PINE FORESTS (MPF) BASED ON FIRST TEN 0.1 SQ. M OBSERVATIONS (1 SQ. M) IN EACH SAMPLE. Coastal scrub species density was derived from 16 sq. m. quadrats at each stand, expressed here as species/sq. meter using S = cA’ where z = 0.25. Number of species per square meter Count Coastal Terrace Prairies a2 Inland Nassella Prairies 80 Monterey Pine Forests 46 Coastal Scrub 141 Comparison of means (INP, CTP, MPF) DF Category 3) Residual 155 33 coastal terrace prairie stands. In 80 inland Nas- sella prairie stands, a total of 194 species were found; 136 plant species were found in 48 coastal Monterey pine stands (Vogl et al. 1988). Many few- er plant species (65 total) were found in 141 coastal scrub stands. Species richness (species/sq. m) in coastal prairies is about 3.5 times greater than in adjacent Monterey pine forests (Table |) and nearly 4 times greater than in adjacent coastal scrub. Coastal prairies also have a much greater species richness compared to other grasslands in California 3D Inland Nassella Prairie TT] LEE YD 96 —2+ Ungrazed "Oo —e Grazed o =F Y 20 Ces (e) o B 15 = mm) Z. 10 5 a! Fic. 3. Mean Std. err. 22.6 1.58 14.7 0.78 6.2 0.58 aes) 0.20 Mean square F value P-value 2615.2 62.9 <0.0001 41.6 and North America (Table 2) when each is pre- sented at a comparable scale (1 sq. m). Preliminary Bray-Curtis ordinations of the 33 stands revealed that both stands at Bird Rock and both forest stands in Pebble Beach (Poppy Hills, 24 Padre Lane-Table 4) were outliers. Although elim- inated, they offer insights to dynamics of coastal prairies. Samples at Bird Rock had unusually high cover of Carex pansa L. Bailey (68% and 18%, respectively). Carex pansa was only found in two other stands, and there it was not abundant (cover 1.5 Le |e, 1000 Average number of species present as sampling areas are added in 43 ungrazed (1937-1991) inland Nassella prairies and 37 grazed (1880-1991) inland Nassella prairies (Stromberg and Griffin 1996). 242 TABLE 2. MADRONO COMPARISON OF SPECIES RICHNESS (SPECIES/SQ. M, = SD or + SE) IN CALIFORNIA GRASSLANDS AND OTHER | ARID WESTERN PLANT COMMUNITIES. Species codes: Agr spi, Agropyron spicatum; Agr smi, Agropyron smithii; Amo can, Amorpha canescens; Art tri, Artemesia tridentata, Fes ida, Festuca idahoensis Elmer; Sti let, Stipa lettermanii; | Poa pra, Poa pratensis. Community name Coastal Terrace Prairies Napa Co.—L. Berryessa, CA Serpentine Meadows Napa Co.—L. Berryessa, CA Non-Serpentine Meadows Monterey Co., Sierra de Salinas, CA Annual Calif. Grassland Bighorn Basin, C. WY Agr spi/Art tri Grant Teton, NW WY Agr spi/Art tri Gunnison, W. CO Sti let/Art tri Wind Cave, W. SD Poa pra/Amo can Charles Russell NWR, E. MT Agr smi/Art tri Yellowstone, W. WY Fes ida/Art tri Pipestone N.M., S. MN Tallgrass Prairie Cheyenne, SE. WY Mixed-grass Prairie Pawnee Butte, NE CO Short-grass Prairie <5%). Bird Rock 2 was the only stand to have more than 10% Lolium multiflorum Lam. with 34.4%. Padre Lane was the only stand where we found Phalaris californica Hook & Arn. and Hier- ochloé occidentalis Buckley and was one of only two stands with Pinus radiata (11%). Poppy Hills also had some Pinus radiata (2%), but was the only stand with Rubus ursinus Cham. & Schldl. or to have >1% Arctostaphylos hookeri G. Don (10%). Because these four stands have such exceptional composition, they confound analysis of the other open coastal prairies and were subsequently dropped from more detailed analysis of coastal ter- race prairies. Ordinations of species composition data from the 29 coastal terrace prairies showed that different land forms (“‘inland ridges,’ “coastal terraces,” and “bald hills’’ above coastal terraces) each have a relatively distinct species composition. This was supported by results of the analysis of variance for individual measures. Comparisons of the three land forms of coastal terrace prairies showed that bald hills have more species of grasses and sedges over- all (Table 3-a), more native grasses and sedges (Ta- ble 3-b) and more native forbs at both scales of sampling (Table 3-c, d) and so more species of na- tive plants overall (Table 3-e). The cover of the native plants show a similar pattern—Bald Hills has more cover of natives (Table 3-f), particularly more cover of native perennial forbs (Table 3-g). Inland Species density 22.62, 12.3; 10.3, 14.7, 10.2, 8.9, 8.8, Sts 4.6, 9.0, ee 8 10.7, S.), Reference + 8.9 This study. 23 (Harrison 1999) 2 (Harrison 1999) ae OG (Stromberg and Griffin 1996) + 0.5 (Stohlgren et al. 1999a) + 0.6 (Stohlgren et al. 1999a) = 04 (Stohlgren et al. 1999a) HOD: (Stohlgren et al. 1999a) 0.5 (Stohlgren et al. 1999a) + 0.6 (Stohlgren et al. 1999a) OZ (Stohlgren et al. 1999b) + OS (Stohlgren et al. 1999b) + 0.4 (Stohlgren et al. 1999b) ridges, as will be seen in a pattern extending to inland Nassella prairies (below) show more domi- nance by annual exotic forbs (Table 3-h) and other | exotics in general (Table 3-1) than the more diverse, bald hill coastal terrace prairie. Both inland Nassella prairies and coastal terrace prairies show an increase in the number of native species with total species richness in 0.1 ha plots (Figure 6-a, b). The number of exotic species also increases with species richness in inland Nassella prairies but not in coastal terrace prairies (Figure 6- c, d). The relative cover of exotics decreases in both inland Nassella prairies and coastal terrace prairies with species richness (Fig. 6-e, f). One rel- ative outlier (Fig. 6-f) is instructive. This stand, Pie- dras Blancas-1, is unusually low in both relative cover of exotics and species richness because it is the only stand with 95% cover of one native, the perennial grass (Deschampsia cespitosa var. holci- formis). In some cases then, relative cover of ex- otics can be low (~1%) even with low species rich- ness (~34) if the total cover of the site is high. Including all coastal terrace prairies and inland Nassella prairies, species richness (0.1 ha) and total cover were positively correlated (R*,,; = 0.18, F = 24.8, P < 0.001, residual df = 107) and relative cover of exotic species is negatively correlated with total cover (R?,, = 0:19, F = 25.6,,P=— 000m residual df = 107). Native species richness (0.1 ha) and exotic species richness were positively corre- [Vol. 48 | 2001] TABLE 3. BASED ON OPEN COASTAL GRASSLANDS, DIFFER- ENCES BETWEEN MEAN VALUES FROM COASTAL TERRACES (CT, N = 10), COASTAL BALD HILLS (BH, N = 10 AND INTERIOR Dry RIDGES (IR, N = 9) ARE SHOWN BASED ON ANALYSIS OF VARIANCE. Bonferroni/Dunn post-hoc com- parison (e.g. BH, CT) are shown only with a significance level of 0.05. Residual df = 26 in all 2-way ANOVA. Arcsin-square root transformation of ratio data were done before ANOVA; results expressed below in ratios. a. Species of Grasses and Sedges 0.1 ha F = 3.61, P= 0.04 mean std. error CT 14.5 2 BH 15.4 1.4 BH, IR P = 0.05 IR 11.0 0.8 b. Species of Native Grasses and Sedges 2 sq. m F = 6.1, = (0.006 mean std. error CT 14.5 1.2 BH 15.4 1.4 BH, IR P = 0.05 IR 11.0 0.8 c. Species of Native Forbs 0.1 ha F = 3.92, P = 0.032 mean std. error GT LG 2.6 BH 28.9 3.4 BH, CT P = 0.05 IR 20.4 3.0 d. Species of Native Forbs 2 sq. m F = 3.45 P = 0.04 mean std. error (Gay 9.7 1.9 BH 17.1 225 BH, CT P = 0.05 IR ISA 1.9 e. Species of Native Plants 0.1 ha F = 4.39 P = 0.02 mean std. error cr Z| 2.9 BH 36.7 Sis) BH, IR P = 0.05 IR 24.6 2.9 f. Cover—Native Plants F = 7.02, P = 0.036 mean std. error GE 677 8.8 CT, IR P <= 0.01 BH Tid 5.3 BH, IR P < 0.01 IR 42.9 4.6 g. Cover—Native Perennial Forbs F = 3.89, P = 0.035 mean std. error CT [22 27 BH Dae 4.7 BH, IR P = 0.05 IR 8.6 1.6 h. Cover—Annual Exotic Forbs F = 6.84, P = 0.004 mean std. error CT 9.7 25 CT, IR P <= 0.05 BH 6.9 1.7 BH, IR P < 0.05 IR 20.6 3.9 1. Ratio of Exotic Species/Total Forb Cover F = 4.86, P = 0.016 mean std. error CT 0.48 0.069 BH 0.37 0.048 BH, IR P < 0.05 IR 0.63 0.025 STROMBERG ET AL.: COASTAL CALIFORNIA GRASSLANDS 243 lated for inland Nassella prairies (R*,4; = 0.23, F = 24.6, P < 0.001, residual df = 78) but not for coast- al terrace prairies (R? = 0.02, F = 0.67, P < 0.001, residual df = 27). Considering just the 29 open coastal terrace prai- ries, grasses and forbs differ in how they vary with species richness at 0.1 ha. On this scale, neither total vegetative cover nor cover of native species is related to species richness. However, cover of all exotic species decreases significantly with increas- ing species richness (R*,,, = 0.15, F = 6.0, P < 0.05, residual df = 27). This decrease in cover of exotics with an increase in species richness was largely due to the decrease in cover of exotic grass- es. Cover of exotic grass was significantly nega- tively correlated with species richness (R?,4, = 0.12, F = 4.96, P < 0.05, residual df = 27). Neither the cover of exotic forbs or the number of exotic forbs was correlated with species richness in coastal ter- race prairies. The degree to which the grasses are native increases with species richness; the number of native grasses is correlated with species richness (Rg — O14, F = 5.56, P= 0.05, residual dt — 27) although the cover of native grass is not cor- related. The degree to which forbs are native also increases with increasing species richness. In this case both cover and number of native forbs (0.1 ha) were significantly correlated with species richness (respectively, R*,4 = 0.29, F = 12.9, P =< 0.001, residual df = 27 and R’,, = 0.83, F = 136, P < 0.001, residual df = 27). The summary of the 25 most frequent and dom- inant species of grasses and sedges in the coastal prairies (Fig. 4) reveals that coastal prairies are heavily invaded with non-natives. The widespread fescue, Vulpia spp. is in most samples and is often very abundant. Danthonia californica, Nassella pulchra, and Festuca rubra are the native grasses that define this community. A summary of the 25 most frequent and dominant species of forbs in coastal prairies (Fig. 5) show a similar dominance by non-natives; in this case by widespread Plan- tago, Erodium, and Hypochaeris, all of which ex- tend well inland. Most important natives include the very similar Plantago erecta E. Morris and Baccharis pilularis. These natives, along with Vi- ola, Sidalcea, Cammisonia, and Acaena are the na- tive forbs that complete the definition of these coastal prairies. In this community, the dominance of grasses drops off rapidly (Fig. 4), while the cover and frequency of the forbs is far greater for more species (Fig. 5). Dominance of exotic grasses and forbs in coastal terrace prairie increases inland coastal terraces, to bald hills, and further to the drier ridges (Table 3). For coastal prairies, the average ratio of the number of exotic grass or sedge species to total species present is 0.44, and the average ratio of exotic forb species is 0.33 (n = 33). Corresponding average ratios of exotic species in inland Nassella prairies are 0.79 and 0.19 (n = 80). 244 MADRONO [Vol. 48 | TABLE 4. STAND NAMES AND LOCATIONS (UTM) INCLUDED IN SAMPLING OF CALIFORNIA COASTAL GRASSLANDS. No. Stand name Abbreviation East North Soil type | Barker Ranch, Laureles Ridge Barker | 614.291 4041.710 Sandy Loam 2 Barker Ranch, Laureles Ridge Barker 2 613.835 4041.865 Sandy Loam A Bird Rock, Inland terrace Bird Rock 1 593.202 4050.157 Sand 4 Bird Rock, Inland terrace Bird Rock 2 293.177 4050.126 Sand =) Canada Woods Site No. 4 CW No. 4 603.776 4051.868 Clay Loam 6. Canada Woods, Garage Site CW Garage 604.100 4047.500 Clay Loam 7 Canada Woods, Lower Pine Tree CW Low Pine 603.500 4048.200 Clay Loam 8 Canada Woods, Pine Tree CW Pine Tree 606.660 4049.100 Clay Loam 9 Canada Woods, Swale CW Swale 603.800 4046.900 Clay Loam 10 Canada Woods, Big Pool CW Big Pool 604.085 4046.295 Clay Loam ile Fish Ranch, above Entrance Rd. Fish Ranch 1 597300 4042.694 Silty Loam 12. Fish Ranch, above Entrance Rd. Fish Ranch 2 597515 4042.692 Silt Loam 13. Fort Ord, opposite Toro Park Ft. Ord Toro 615.941 4052.110 Sandy Loam 14. Jade Flat, Central Terrace Jade Flat 638.283 3975.370 Serpentine Clay is Laureles Grade, Laureles Ridge Laureles 611.870 4043.319 Sandy Loam 16. Molera State Park, N. Terrance Molera 602.546 4016.610 Sand Loam lige Olson Hill, Diablo Canyon Olson Hill 698.859 3896.149 Loam 18. Pebble Beach, 24 Padre Lane Pebble For 592.366 4048.055 Loamy Sand 19. Plaskett Ridge, Above Camp Plaskett Rdg 638.277 3975:359 Serpentine Rock pA0R Point Lobos, Escobar Flat Pt. Lobos Flat 397-250 4042.194 Sandy Loam Zl: Point Lobos, Mima Mounds Pt. Lobos Mima 584.700 4041.300 Clay Loam 22. Poppy Hills, Opposite Golf Club Poppy Hills 595.225 4048.980 Sandy Loam 23, Pt. Piedras Blancas Light House Piedras LH 646.262 3948 .422 Sand 24. Pt. Piedras Blancas Terrace | Piedras | 654.425 3951.349 Loam 255 Pt. Piedras Blancas Terrace 2 Piedras 2 654.347 3951.341 Loam 26. Rancho San Carlos, Animas Entr. Danthonia RSC 600.086 4040.213 Clay Loam Zi Rancho San Carlos, Animas Rdg. Animas RSC 600.071 4041.197 Clay Loam 28. San Bruno Mtn., Great Meadow San Bruno | 548.059 4171.974 Loamy Sand 2; San Bruno Mtn., Great Meadow San Bruno 2 549.274 4171.733 Loamy Sand 30. San Louis Hill, Diablo Canyon San Luis Hill 703.599 3893.915 Clay Loam ol Soda Springs, Bluff W. of Hwy 1 Soda Spring 646.266 3965.330 Serpentine Rock 32. Spruance Meadow, Spruance Dr. Spruance 595.662 4047.953 Loamy Sand 33. Work Ranch, Hwy 68 Work Ranch 603.586 4048.821 Sandy Loam We used CCA to sort out which individual com- parisons of community characteristics (measures of diversity and the degree of invasion by exotic spe- cies) are most closely associated with vegetation composition across the landscape gradient from in- land Nassella prairies to coastal terrace prairies. Some of these measures were highly correlated, so we proceeded with only 19 (Table 5). Plant species with total cover of less than 0.05% and which oc- curred in 5 or fewer stands were eliminated, leaving 192 species. With these simplifications (tolerance set to 0. LOOOOOE-12), PCord reached a CCA solu- tion after fewer than 100 iterations. Inland Nassella prairies were clearly grouped to the left (Fig. 7) and coastal terrace prairies were on the right. The high- est “‘intraset correlation”? (ter Braak and Smilauer 1998) with this first axis was the cover of native, perennial grass and on the second axis, the cover of exotic species (Fig. 7). Coastal terrace prairies are Characterized as having both more species and higher cover of native perennial grasses than any of the tightly grouped (thus similar) inland Nassella prairies that are clearly placed to the left of the first axis where stands can be described as having higher cover of annual, exotic forbs (for example, Ero- dium cicutarium (L.) L. Hér). The second axis dis- tinguishes between coastal terrace prairies into those with relatively more dominance by exotic species (Piedras Blancas Light House, Fish Ranch, etc.) and those with relatively high cover of native, | perennial forbs (San Bruno, Soda Springs, Piedras Blancas 1, etc.) (Fig. 7—joint plot). By restricting our focus to only coastal terrace prairies, we can examine the pattern of species composition and floristic characteristics at a smaller scale. By dropping the 80 inland Nassella prairies, — the number of plant species (total cover >0.5%) included in the CCA dropped to 149. Correspond- ingly, more of the floristic variables were highly correlated (Table 5) and only 12 were relatively in- dependent. A joint plot (Fig. 8) shows many coastal terrace prairies in a group with relatively high cover of exotic species. The highest intra-set correlation with the first axis was with cover of exotic species. The highest intra-set correlation with the second axis was with the number of native forbs. Stands such as San Bruno, Soda Springs, the stands near Avila Beach (Olson Hill, San Luis Hill) are ex- emplary coastal terrace prairies with fewer exotics and higher cover of native forbs. The Point Lobos Mima mound prairie and Piedras Blancas | are coastal terrace prairies with fewer exotics, higher 2001] STROMBERG ET AL.: COASTAL CALIFORNIA GRASSLANDS Dominant Grasses/Sedges - Coastal Grasslands —_— a) (ee) Vulpia myuros myuros Danthonia californica Bromus hordeaceus Importance Value- log (Freq. x Cover) | 0.1 Elymus glaucus Juncus bufonius Bromus diandrus Deschampsia cespitosa cespitosa BM Native [__]Non-Native Lolium multiflorum Nassella pulchra Avena barbata 10 Aira caryophyllea Nassella cernua Briza minor Hordeum branchyantherum Festuca rubra Deschampsia cespitosa holciformis Avena fatua Bromus carinatus Lolium perenne Hordeum marinum Juncus pheocephalus Distichlis spicata Leymus triticoides Briza maxima Nassella lepida Species Importance Order Fic. 4. Importance value (frequency X average percent cover) of the 25 most dominant species of grasses and sedges, averaged from 29 coastal terrace prairies in central, coastal California. cover of native forbs and more native grasses (Fig. 8—joint plot) while stands clustered on the left side relatively more cover of exotic species. Soils and locations of the various sites are de- scribed in Table 4. Fourteen of the coastal terrace prairies (44%) occurred on soils with a hardpan from 10—20 cm beneath the surface. An additional three coastal terrace prairies occurred on serpentine rock or clay with limited drainage. A hardpan that provides standing water during the winters was of- ten present, but apparently not required. Coastal terrace prairies also occurred on sands, loams and clays. DISCUSSION Diversity of plant species in coastal terrace prai- ries is among the highest in grasslands of North America (Stohlgren et al. 1999b). County and wild- land planners often have selected coastal terrace prairies for development, perhaps not recognizing the biodiversity of coastal terrace prairies or be- cause political support to protect forests or coastal scrub has been comparatively well organized. If protection of biodiversity is a goal, then coastal ter- race prairies should be protected and development should be focused on relatively species-poor plant communities. Invasions of natural communities by exotic spe- cles may occur more readily in areas of low species diversity than in areas of high species diversity (Darwin 1859). Plant communities with low species diversity (and total cover) may use resources less completely, allowing invasion by similar species (MacArthur and Wilson 1967; Pimm 1991; Tilman et al. 1997). Evidence for this relationship between diversity and invasions in grasslands depends on scale, and at a landscape scale, may be reversed (Stohlgren et al. 1999b). Based on the number of species, our studies do not support the theory that exotics are more abundant where species diversity is relatively low; inland Nassella prairies (but not coastal terrace prairies) with the highest diversity have more exotic species (Fig. 6c, d). But, numbers of species probably do not reflect ecosystem func- tion. Species occupying more space intercept more light, and presumably are more important in nutri- ent capture and storage. In both inland Nassella prairies and coastal terrace prairies, Most species in our grasslands have cover <5%. Two species may be equally counted as present, but one may occupy far more cover in the community. Based on relative cover (Figs. 6e, f, 7) both diverse inland Nassella prairies and coastal terrace prairies have reduced exotic dominance, supporting the theory that less diverse communities are more likely to be invaded. Ordinations of the species and stands agree with Dominant Forbs — = Plantago lanceolata Erodium botrys Hypochaeris glabra Plantago coronopus Baccharis pilularis Plantago erecta — © Geranium dissectum 0.1 Rumex acetosella Madia exigua Ranunculus Importance Value- log (Freq. x Cover) MADRONO Hypochaeris radicata [Vol. 48 - Coastal Grasslands Viola pedunculata Anagallis arvensus Sidalcea malvaeflora Camissonia ovata BM Native [_]Non-Native Trifolium dubium Silene gallica Acaena pinnatifida californica Geranium molle Eryngium armatum Chlorogallum pomeridian ‘ : Gnaphalium purpureu californicus P saa a Daucus pusillus Sisyrinchium bellum Lotus corniculatus Species Importance Order Fic. 5. Importance value (frequency X average percent co from 29 coastal terrace prairies in central, coastal California. our general understanding of the biology of indi- vidual species. The ordinations are also consistent with patterns seen by making pairwise statistical comparisons (Table 3, Fig. 7). All coastal terrace prairies have been invaded to some degree by ex- otics, but the importance of exotic species is far greater in inland Nassella prairies. Two of the three grasses with the most cover (Fig. 4) and the four forbs with the most cover in coastal terrace prairies are exotics (Fig. 5). Although the effects of grazing by large domes- tic herbivores on these grasslands were not specif- ically addressed, a re-analysis of inland Nassella prairie data at various scales of sampling (Figs. 2, 3) supports observations elsewhere that the loss in plant species diversity in grasslands grazed by her- bivores is only seen at a scale larger than about 100 m? (Chaneton and Facelli 1991, Olff and Ritchie 1998). Grazing, or its removal, probably has little effect on species diversity in other California grass- lands because grazing has been continuous for cen- turies following European settlement (Harrison 1999). At some scales, grazing may have little ef- fect on native species richness in other North American grasslands (Stohlgren et al. 1999a). All grasslands in this study were probably grazed since European settlement. There is no evidence that | ver) of the 25 most dominant species of forbs, averaged | herds of large herbivores co-evolved with the Cal- | ifornia coastal grasslands (Painter 1995). In gener- al, domestic livestock grazing has had severe im- | pacts on grassland ecosystems in western North America (Painter and Belsky 1993) and livestock | removal has been suggested at various scales (Bock et al. 1993). However, grazing has been present for | so long that careful consideration must be given before livestock are removed from coastal grass- lands. On Santa Cruz Island, grasslands formerly grazed by cattle now support near monocultures of Foeniculum vulgare Miller, an exotic plant former- ly held in check by year-long grazing (M. Strom- berg pers. obs., Mayfield et al. 2000). On other coastal parklands where grazing has been entirely removed after many years of year-round grazing (Andrew Molera State Park, Santa Clara County Parks, San Mateo County’s Mid-Peninsula Open Space District) we have seen extensive, rapid ex- pansions of F. vulgare, B. pilularis, and Dipsacus spp. where these plants were formerly relatively un- important. Most coastal terrace prairies were clearly open grasslands, but the composition of several stands included trees and shrubs (e.g., Poppy Hills, Padre Lane). The presence of otherwise typical coastal terrace prairies grasses and forbs in these brushy or 2001 | STROMBERG ET AL.: COASTAL CALIFORNIA GRASSLANDS 247 oO on —) Inland Nassella Grasslands Coastal Grasslands nN = & = N i) Rho = .96, P< .001 Number of Native Species 3 Rho = .94, P< 0.001 0 10 = 2030 40 50670 0 10 «=6©20—:—i830'—s —“( aésS— eis Total Species Richness (1000 sq. m) Total Species Richness (1000 sq. m) >) oO. Inland Nassella Grasslands Rho = 0.08, P = 0.65 Number of Exotic Species Number of Exotic Species Rho = 0.56, P < 0.001 (n.S.) 0 10 20 #30 40 £450 = ©6©60~)©670 0 10 20 30 40 £50 ©=©60.~©6©70 Total Species Richness (1000 sq. m) Total Species Richness (1000 sq. m) e f 1.2- , Inland Nassella Grasslands = | O oS 6 0.8 -H 80.6 z E g ‘a 0.4 (9.34 Coastal Grasslands . ol xe © 9.24 Rho = -.44, P= 0.02 ae g 8 5 9.1 Rho = -.47, P < 0.001 ) 10 20 30 £40 50 60 70 9 0 10 20 30 40 50 60 70 Total Species Richness (1000 sq. m) Total Species Richness (1000 sq. m) Fic. 6. Spearman rank correlation between species richness and number of all native plant species for a.) 80 inland Nassella prairies and b.) 29 coastal terrace prairies. Spearman rank correlation between species richness and number of all exotic plant species for c.) 80 inland Nassella prairies and d.) 29 coastal terrace prairies. c. Spearman rank correlation between species richness and relative cover of exotic plant species for e.) 80 inland Nassella prairies and f.) 29 coastal terrace prairies. 248 MADRONO [Vol. 48 | TABLE 5. FLORISTIC VARIABLES CONSIDERED IN ANALYSIS OF SPECIES RICHNESS IN CALIFORNIA GRASSLANDS. Highly | correlated variables were dropped, leaving 19 variables (middle column) with sufficiently minimal correlation required | to compute canonical correlation analysis (Fig. 6). When only 29 coastal terrace prairies are compared, 12 floristic | variables were included in the analysis (Fig. 7). For example, in a CCA including 80 INP and 29 CTP, variable 1 (No. of Species in 2 sq. m) was dropped as it had a high correlation with variable 21 (No. of all Natives). Correlated variable(s) no., R? No. Variable name 80 INP and 29 CTP 29 CTP is No. of Species in 2 sq. m OA ere 2b, 22 a No. of Species in 0.1 ha 25, .87 25, .9le 27. 295 2 Total Vegetative Cover included included 4. Cover of all Exotics included included ap Cover of all Natives 8, .78; 17, .79; 6, .86 6, .80 6. Cover of Native, Perennial Grass included included Vs Cover of Native, Perennial Forbs included included 8. No. of Native, Perennial Grasses (0.1 ha) 92 25.99 23,99 9. No. of Native, Perennial Forbs (0.1 ha) included 25, .91 10. Cover of all Exotic Grasses (0.1 ha) included 4, .9] 11. Cover of all Exotic Forbs (0.1 ha) included included (2. Cover of Annual, Exotic Grasses 4, .99 4, .90 3 No. of Annual, Exotic Grasses (0.1 ha) included 16, .84; 22, .94 14. Cover of Annual, Exotic Forbs included included 15. No. of Annual, Exotic Forbs (0.1 ha) included 240292 16. No. of Exotic Grasses (2 sq. m) included 22, .88 iba No. of Native Grasses (2 sq. m) 8, .94 Os OZ). 255.00 18. No. of Exotic Forbs (2 sq. m) included 20, .86 19, No. of Native Forbs (2 sq. m) included included 20. No. of all Exotics (2 sq. m) included included 2 le No. of all Natives (2 sq. m) included included 22. No. of all Exotic Grasses (0.1 ha) 13, .96 13, .94 23% No. of all Native Grasses (0.1 ha) included included 24. No. of all Exotic Forbs (0.1 ha) included 26, .86 207 No. of all Native Forbs (0.1 ha) included included 20. No. of all Exotics (0.1 ha) included included 20 No. of all Natives (0.1 ha) 25, .94 25295 Piedras | CvrNtPrnFb #NatFb 0.1 ha #NtFbs 2 sq m BB #NatSp 2 sq m BANtPrnFr 0.1 ha CvrNtPrnGs #NaGs O.lha #AnExFbs 0.1 ha #ExFbs 2 sq m AnExGrs 0.1 hé #ExFb 0.1 ha #Ex Sp 2 sqm #Ex(Qrs.2 m sq fm Total Cover #ExSp 0.1 ha San Luis Hill ® Olson Hill Jade Flat I e Piedras 2 CvrExFrb i | CvrAnExG -0.4 Canonical Variate Two Danthonia RSC Fish Ranch | | @ TL eee Piedras LH Fish Ranch 2 Cyr ExS Ge vr ExSp 15-1 05 0 05 1 15 2 25 06-04-02 0 02 04 06 08 1 Canonical Variate One Fic. 7. Canonical correspondence analysis (ter Braak 1994) of 80 inland Nassella prairies and 29 coastal terrace prairies, based on cover of 197 plant species and 19 floristic variables. Inset: joint plot of stands and correlated floristic variables for all stands; length of vector associated with each variable is related to correlation with position of stand on canonical axes; highly correlated floristic variables are plotted near each other. CoE @ J) © 3 i Fish Ranch | al Danthonia RSC Jade a = 0 5 Molera @® @| Piedras 2 z — i] 5 (pPt- Lobos Plat CW Garage | : @ nimas = Fort ofl Toro@ @e CW No. 4 CW Pine Tree 2 CW Lower Pine W Big Pool Piedras LH 5 Op ee CW Swale4 ; -1.5 D ae ie -1 -05 0 O05 1 15 2 2.5 Canonical Axis One FIG. 8. STROMBERG ET AL.: COASTAL CALIFORNIA GRASSLANDS 249 CvrNtPrn Total Co #NaGs .1 Canonical correspondence analysis of 29 coastal terrace prairies based on cover of 149 plant species and 12 floristic variables. Inset: joint plot of floristic variables for axes one and two. forested stands supports the concept of a dynamic tension between forest and grassland vegetation mediated by occasional fire (or grazing) in the coastal communities (Greenlee and Langenheim 1990). Currently, central California coastal grass- lands are gradually seeing increased dominance by the native coastal shrub (B. pilularis) (McBride and Heady 1968) or oak woodland (Callaway and Davis 1993). Indeed, B. pilularis is a co-dominant native of the coastal prairies. California’s human population will double by 2040, and coastal development is much faster than that in interior California (Medvitz and Sokolov 1995). Although prescribed fires are the most cost- effective way to maintain the grasslands on a large scale (Kephart 2000), current and planned devel- opment almost precludes this option. Small-scale prescribed burns, mowing, and controlled grazing during the dry season should be included in man- agement strategies to sustain the long-term viability of California’s coastal prairies. Although some may attempt to assign names or define units of vegetation (Sawyer and Keeler-Wolf 1995) this may be impossible (Zedler 1997) be- cause vegetation occurs On a continuum in the en- vironment where each species has an individual distribution on the gradient from coastal to inland. Even if we could find identical environments, spe- cies composition would probably vary due to other factors that have undoubtedly influenced the abun- dance of individual species in a given year (Fox and Fox 1986). Gradients may also be based on competition; from wet (coastal) to dry (inland) (Lane et al. 2000) or disturbance (fire frequency, grazing duration and intensity, gopher abundance, €tc.). Gopher tailings probably sustain a disturbance regime of inland California annual grasslands and old fields (Stromberg and Griffin 1996) where go- pher density can be very high. On coastal terrace prairies, however, gopher tailings rarely observed. We did not sample all known high-quality or rel- ict stands in this study area. Terraces on the San Simeon Ranch, the grasslands just north of Santa Cruz adjacent to Wilder Ranch State Park, those on the San Mateo coast, and those north of Bodega Bay need more investigation. Coastal prairies support a number of state or fed- erally designated “‘rare’”’ species and are often man- aged for protection of rare animals (Launer and Murphy 1994). Rare animals include several but- terflies; the Mission Blue and San Bruno Elfin (McClintock et al. 1990, Weiss 1993). Conserva- tion of the coastal terrace prairie on San Bruno Mountain includes the first “‘habitat conservation plan”’ approved by the federal government. Species considered as “‘special plants’” by various regula- tory agencies and observed in this study include: Arctostaphylos hookeri G. Don, Astragalus tener A. Gray var. titi (Eastw.) Barneby, Allium hickmanii Eastw. (in 23% of the coastal terraces sampled), Sanicula maritima S. Watson, Trifolium polyodon E. Greene, Psilocarphus tenellus Nutt. var. globi- ferus (DC.) Morefield, Cirsium occidentale (Nutt.) 250) Jepson var. compactum Hoover, Perideridia gaird- neri (Hook & Arn.) Mathias, and Arabis blephar- ophylla Hook & Arn. Each of these officially rare species occurred in only one stand, and then only sparsely in the larger plot (SO m X 20 m, 0.1 ha). We also found Ophioglossum californicum Prantl at Spruance Meadow, not seen in Monterey County since its original collection in 1910. A. ¢. var. titi is federally listed as endangered and occurs only in one location (Bird Rock). We purposely included this site in the sampling as it occurs on an excep- tional relict stand including Danthonia and Des- champsia cespitosa var. holciformus. Land manag- ers who can identify the species assemblages de- scribed here (Figs. 4, 5) should expect other asso- ciated rare species. ACKNOWLEDGMENTS We dedicate this paper to the memory of Oren Pollak who arranged funding for this study through The Nature Conservancy, California. Thanks to various landowners or managers who allowed us to study the grasslands under their care; Tom Gray—Santa Lucia Preserve, Diane Fish—Fish Ranch, Bill Barker and Danny Hoag—Barker Ranch, Pacific Gas and Electric Co., Ted Horton—Pebble Beach Company, Alan Williams—Canada Woods, Galen Rathbun and Norm Scott, USFWS—Piedras Blancas & Cambria, who provided helpful access to land. Mike Markkula and Helen Johnson provided logistical support. Support from the Museum of Vertebrate Zoology, UC- Berkeley is gratefully acknowledged. Susan Harrison and Carla D’ Antonio provided helpful advice and we appre- ciate Walter Koeing’s critical reviews. LITERATURE CITED BARBOUR, M., AND J. MAsor. 1995. Terrestrial Vegeta- tion of California, Special Publication No. 9, 1995 ed. California Native Plant Society, Sacramento, CA. BARRY, W. J. 1972. The Central Valley prairie. California State Parks and Recreation Department, Sacramento, CA. BISWELL, H. H. 1956. Ecology of California grasslands. Journal of Range Management 9:19—24. Bock, C. E., J. H. Bock, AND H. M. Situ. 1993. Proposal for a system of federal livestock exclosures on public rangeland in the western United States. Conservation Biology 7:731-—733. BuRCHAM, L. T. 1957. California range land. California Department of Forestry, Sacramento, CA. CALLAWAY, R. M., AND E W. Davis. 1993. Vegetation dy- namics, fire, and the physical environment in coastal central California. Ecology 74:1567—1578. CHANETON, E. J., AND J. M. FAcCELLI. 1991. Disturbance effects on plant community diversity: Spatial scales and dominance hierarchies. Vegetation: 143—156. Cook, T. D. 1978. Soil Survey of Monterey County, Cal- ifornia. Soil Conservation Service, U.S. Forest Ser- vice, and University of California Agriculture Exper- iment Station., Berkeley, CA. CYLINDER, P. D. 1995. The Monterey ecological staircase and subtypes of Monterey pine forest. Fremontia 23: 7-13. DARWIN, C. 1859. Origin of species. Reprinted by Penguin Books, London, UK. MADRONO DAUBENMIRE, R. F 1959. Canopy coverage method of veg- etation analysis. Northwest Science 33:43—64. ErnstTRoM, D. J. 1984. Soil survey of San Luis Obispo County, California, Coastal Part. Soil Conservation Service, Forest Service, and University of California Agriculture Experiment Station, Berkeley, CA. Fox, M. D., AND B. J. Fox. 1986. The susceptibility of natural communities to invasion. Pp. 57-60 in R. H. Groves and J. J. Burdon (eds.). Ecology of biological invaions. Cambridge University Press, Cambridge, UK. GREENLEE, J. M., AND J. H. LANGENHEIM. 1990. Historic fire regimes and their relation to vegetation patterns in the Monterey bay area of California. American Midland Naturalist 124:239—253. Gross, K. L., M. R. WiLiGc, L. GOUGH, R. INOUYE, AND S. B. Cox. 2000. Patterns of species density and pro- | [Vol. 48 | ductivity at different spatial scales in herbaceous | plant communities. Oikos 89:417—427. HAMILTON, J. G. 1998. Changing perceptions of the pre- European California grasslands. Madrono 44:31 1— 333. HARRISON, S. 1999. Native and alien species diversity at the local and regional scales in a grazed California | grassland. Oecologia 121:99—106. HARRISON, S., K. RICE, AND J. MARON. 2001. Habitat | patchiness promotes invasion by alien grasses on ser- pentine soil. Biological Conservation 100:45—53. HEADY, H. F, J. W. BARTOLOME, M. D. Pitt, G. D. Sav- ELLE, AND M. C. Stroup. 1992. California Prairie. Pp. 313-335 in R. T. Coupland (ed.). Natural grasslands: Introduction and Western Hemisphere. Elsevier, Am- | sterdam, Netherlands. G. BARBOUR, AND W. J. BARRY. 1988. Valley grass- land. Pp. 491-514. in M. G. Barbour and J. Major | (eds.). Terrestrial vegetation of California. John Wiley and Sons, NY. HEvuER, P. 1994. Vegetationskundliche Untersuchungen im | Kustenbusch Mittelkaliforniens. Diepholz angefertigt im Systematisch. Geobotanischen Institut der Georg August Universitat zu Gottingen, Gottingen, Germa- ny. HICKMAN, J. C. (ed.). 1993. The Jepson Manual: Higher | Plants of California. University of California Press, | Berkeley, CA. HOLLAND, V. L., AND D. J. Ket. 1995. California Vege- tation. Kendall/Hunt Publishing Company, Dubuque, IA. HUENNEKE, L. E 1989. Distribution and regional patterns of California grasslands. Pp. 1-12 in L. KF Huenneke and H. A. Mooney (eds.). editors. Grassland structure and function: California annual grassland. Kluwer Academic Press, Dordrecht, Netherlands. Huston, M. A., L. W. AARSSEN, M. P. Austin, B. S. CADE, J. D. FRIDLEY, E. GARNIER, J. P. GRIME, J. HODGSON, W. K. LAUENROTH, K. THOMPSON, J. H. VANDERMEER, D. A. WARDLE, A. HEcTOR, B. SCHMID, C. BEIER- | KUHNLEIN, M. C. CALDEIRA, M. DIEMER, P. G. DIMI- TRAKOPOULOS, J. A. FINN, H. FREITAS, P. S. GILLER, J. Goop, R. HARRIS, P. HOGBERG, K. Huss-DANELL, J. JOSHI, JUMPPONEN, A., C. KORNER, P. W. LEADLEY, M. LorEAu, A. MINNS, C. P. H. MULDER, G. O’ DONOVAN, S. J. Orway, J. S. PEREIRA, A. PRINZ, D. J. READ, M. SCHERER-LORENZEN, E.-D. SCHULZE, A.-S. D. SIAMAN- TZIOURAS, E. SPEHN, TERRY, A. C., A. Y. TTROUMBIS, FE I. Woopwarpb, S. YACHI, AND J. H. LAwtTon. 2000. , LC. Forn, M. M. HEKTNER, D. W. TAYLOR, M. | | 4 ] | | 2001) . . No consistent effect of plant diversity on productivity. Science 289:1255—1255a. KAISER, J. 2000. Ecology: Rift over biodiversity divides ecologists. Science 289:1282—1283. KEELEY, J. E. 1989. California Valley Grassland. Pp. 2— 23 in A. A. Schoenherr (ed.). Endangered plant communities of Southern California. Southern Cal- ifornia Botanists, California State University, Ful- lerton, CA. KEPHART, P. 1993. Sensitive resource inventory for Diablo Canyon lands. Biosystems Inc., Diablo Canyon Pow- er Plant, Avila Beach, CA. . 2001. Resource management demonstration at Russian Ridge preserve. Grasslands 11(1):1—8. KUCHLER, A. W. 1964. Potential natural vegetation of the conterminous United States. Special Publication 36, American Geographic Society, NY. LANE, D. R., D. P. COFFIN, AND W. K. LAUNROTH. 2000. Changes in grassland canopy structure across a pre- cipitation gradient. Vegetation Science 11:359- 368. LaAuNER, A. E., AND D. D. MurpHy. 1994. Umbrella spe- cies and conservation of habitat fragments: a case of a threatened butterfly and a vanishing grassland eco- system. Biological Conservation 69:145—153. Leps, J., AND V. HADINCOVA. 1992. How reliable are our vegetation analyses? Journal of Vegetation Science 3: 119-124. MACARTHUR, R. H., AND E. O. WILSON. 1967. The theory of island biogeography. Princeton University Press, Princeton, NJ. Mack, R. N. 1989. Temperate grasslands vulnerable to plant invasions: characteristics and consequences. Pp. 155-179 in J. A. Drake (ed.), Biological in- vasions: a global perspective. John Wiley and Sons, Ltd., NY. MAYFIELD, R. L., D. VAN VUREN, AND M. L. JOHNSON. 2000. Demography of an insular endemic rodent, Peromyscus maniculatus santacruzae, on Santa Cruz Island. Southwestern Naturalist 45:508—513. McBripDE, J., AND H. EK HeEADy. 1968. Invasion of grass- land by Baccharis pilularis DC. Journal of Range Management 21:106—108. McCLINTOCK, E., P. REEBERG, AND W. KNIGHT. 1990. Flora of the San Bruno Mountains, San Mateo County, Cal- ifornia. California Native Plant Society, Sacramento, CA. McCune, B., AND M. J. MEFForD. 1999. PC-ORD Multi- variate analysis of ecological data. in MjM Software Design, Gleneden Beach, OR. Mepvitz, A. G., AND A. D. SOKOLOV. 1995. Population growth threatens agriculture, open space. California Agriculture 49:1 1—17. Myers, N., R. A. MITTERMEIER, C. G. MITTERMEIER, G. A. B. DA FONSECA, AND J. KENT. 2000. Biodiversity hot- spots for conservation priorities. Nature (London) 403 (6772):853-858. OLFF, H., AND M. E. RITCHIE. 1998. Effects of herbivores on grassland plant diversity. Trends in Ecology and Evolution 13:261—265. PAINTER, E. L. 1995. Threats to the California flora: un- gulate grazers and browsers. Madrono 42:180—188. , AND A. J. BELSKY. 1993. Application of herbivore optimization theory to rangelands of the western United States. Ecological Applications 3:2—9. PETERS, R. L., AND R. E Noss. 1995. America’s endan- gered ecosystems. Defenders 70:16—27. STROMBERG ET AL.: COASTAL CALIFORNIA GRASSLANDS 251 Primm, S. L. 1991. The balance of nature? University of Chicago Press, Chicago, IL. ROBINSON, R. H. 1971. An analysis of ecological factors limiting the distribution of a group of Stipa pulchra associations. Korean Journal of Botany 14:61—80. SAWYER, J. O., AND T. KEELER-WOLF. 1995. A Manual of California Vegetation. California Native Plant Soci- ety, Sacramento, CA. SCHIFFMAN, P. A. 1994. Promotion of exotic weed estab- lishment by endangered giant kangaroo rats (Dipo- domys ingens) in a California grassland. Biodiversity and Conservation 3:524—537. SCHIFFMAN, P. M. 2000. Mammal burrowing, erratic rainfall, and the annual lifestyle in the California prairie: is it time for a paradigm shift? Pp. 153- 160 in J. Keeley, M. Baer-Keeley, and C. J. Foth- eringham (eds.), Interface between ecology and land development in California. US Geologial Sur- vey Open-File Report 00-62, Occidental College, Los Angeles, CA. STOHLGREN, T., M. B. COUGHENOUR, G. W. SHON, D. BINK- LEY, M. A. KALKHAN, L. D. SCHELL, D. J. BUCKLEY, AND J. K. BERRY. 1997. Landscape analysis of plant diversity. Landscape Ecology 12:155—170. , D. BINKLEY, G. W. CHONG, M. A. KALKHAN, L. D. SCHELL, K. A. BULL, Y. OTSUKI, G. NEWMAN, M. BASHKIN, AND Y. SON. 1999b. Exotic plant species invade hot spots of native plant diversity. Ecological Monographs 69:25—46. STOHLGREN, T. J., L. D. SCHELL, AND B. V. HUEVEL. 1999a. How grazing and soil quality affect native and exotic plant diversity in Rocky Mountain grasslands. Eco- logical Applications 9:45—64. STROMBERG, M. R., AND J. R. GRIFFIN. 1996. Long-term patterns in coastal California grasslands in relation to cultivation, gophers and grazing. Ecological Appli- cations 6:1189—-1211. SymsTAD, A. J. 2000. A test of the effects of functional group richness and composition on grassland invasi- bility. Ecology 81:99—109. TAuscH, R. J., D. A. CHARLET, D. A. WEIXELMAN, AND D. C. AAMupIo. 1995. Patterns of ordination and classification instability resulting from changes in input data order. Journal of Vegetation Science 6: 897-902. TER BRAAK, C. FE, AND P. SMILAUER. 1998. CANOCO Ref- erence Manual and User’s Guide for Canoco for Win- dows: Software for Canonical Community Ordination (ver. 4). Microcomputer Power, Ithaca, NY. TER BRAAK, C. J. E 1987a. The analysis of vegetation- environment relationships by canonical correspon- dence analysis. Vegetatio 69:69—77. . 1987b. Ordination. Pp. 91-173 in R. H. G. JONG- MAN, C. J. E TER BRAAK, AND O. FL R. TONGEREN (eds.), Data analysis in community and landscape ecology. Pudoc, Wageningen, Netherlands. . 1994. Canonical community ordination. Ecosci- ence 1:127—140. TILMAN, D., J. KNops, D. WEDIN, P. REICH, M. RITCHIE, AND E. SIEMANN. 1997. The influence of functional diversity and composition on ecosystem processes. Science 277:1300—1302. VocL, R. J.. W. P. ARMSTRONG, K. L. WHITE, AND K. L. CoLe. 1988. Closed-cone pines and cypress. Pp. 295— 358 in M. G. Barbour and J. Major (eds.), Terrestrial vegetation of California. University of California, Da- vis, CA. WEISS, S. B. 1993. The San Bruno Elfin, /ncisalia mossi 252 MADRONO bayensis (Brown). Pp. 141—142 in T. R. e. New (ed.), Conservation biology of Lycaenidae. IUCN, Gland, Switzerland. WESTMAN, W. E. 1975. Edaphic climax pattern of the pyg- my forest region of California. Ecological Mono- graphs 45:109-135. Waite, K. L. 1966a. Unpublished data. in Hastings archives. . 1966b. Old-field succession on Hastings Reser- vation, California. Ecology 47:865—886. [Vol. 48 . 1967. Native bunchgrass (Stipa pulchra) on'| Hastings Reservation, California. Ecology 48: 94 955) YADON, V. 1995. A checklist of the vascular plants of | Monterey County, California. California Native Plant Society, Monterey Bay Chapter, CA. ZEDLER, P. H. 1997. Review: A manual of California veg: | etation by John O. Sawyer and Todd Keeler-Wolf. Madrono 44:214—220. -Maprono, Vol. 48, No. 4, pp. 253-264, 2001 PRE-AGRICULTURAL GRASSLAND IN CENTRAL CALIFORNIA GLEN HOLSTEIN Zentner & Zentner 1509 Pacific Drive, Davis CA 95616 ABSTRACT An increasingly dogmatic paradigm maintains that central California’s pre-agricultural grasslands were once entirely dominated by the bunchgrass Nassella pulchra. Evidence from early records and current relict vegetation, however, indicates they were spatially diverse. In moderate precipitation areas Nassella pulchra frequently dominated grasslands in foothills and occasionally also on sandy valley floors, but grassland on heavier soils in valleys and on many hillslopes was dominated by the rhizomatous graminoids Leymus triticoides, Carex barbarae, and C. praegracilis. Dominance shifted to spring-active annual forbs in low precipitation areas and probably to summer-active annual forbs (tarweeds) on infertile old terrace soils. INTRODUCTION When I was seeking remnant examples of native perennial prairie in 1994 as models for grassland restoration in California’s Central Valley, it soon became evident that there were at least an order of magnitude more remnants dominated by the rhi- zomatous native perennial grass Leymus triticoides (Buckley) Pilger than by native bunchgrasses like Nassella pulchra (A. Hitche.) Barkworth. This ob- servation was surprising since conventional botan- ical wisdom at that time assumed herbaceous valley vegetation was formerly dominated by bunchgrass- es before nineteenth century land use changes caused their almost total replacement by exotic an- nual grasses (Heady 1977). Examination of such apparent paradoxes has gradually led to a reevaluation of long-standing as- sumptions about the valley’s original natural vege- tation. In 1981, for example, two Madrono papers questioned the then generally accepted bunchgrass dominance paradigm (BDP). Bartolome and Gem- mill (1981) first showed N. pulchra is not well- adaptated for dominating pristine herbaceous veg- etation. Then, three months later, Wester (1981) provided historical evidence that well before the post-1850 cattle introduction traditionally blamed for the demise of bunchgrasses, they were either rare or entirely absent in the San Joaquin Valley, a significant part of the area they were traditionally assumed to dominate. The BDP remained one of California botany’s most dominant paradigms through the 1990’s, how- ever, despite these and numerous other studies. Late in the decade Hamilton (1997) systematically dem- onstrated its creation was motivated by pre-existing theory rather than strong evidence, but it is too soon to determine if his work significantly impacted the surprisingly durable BDP. Meanwhile many California Floristic Province vegetation surveys un- critically claim its grasslands were once covered by bunchgrass prairies dominated by N. pulchra (Bar- bour and Christensen 1993; Schoenherr 1992; Sims 1988), while others treat the BDP with only mild scepticism and fail to suggest alternative hypothe- ses (Barbour et al. 1993). One excellent survey of the state’s vegetation in that period (Holland and Keil 1989) acknowledged Wester’s work and ques- tioned BDP, but local publication limited its impact until a new edition (1995) was issued by a national publisher. California native grassland restoration ef- forts consequently still emphasize creating bunch- grass prairies even where their former presence is highly unlikely (Harker et al. 1993; Dremann 1995). Evidence for BDP rests on Clements’ (1920, 1934) observation of N. pulchra stands along rail- roads in the San Joaquin Valley near Fresno and his subsequent conclusion they were relicts of a for- merly widespread bunchgrass prairie that once cov- ered most California Floristic Province valleys and foothills. Hamilton (1997) amply demonstrated, however, that this conclusion was heavily influ- enced by and made to fit grassland theories Cle- ments had previously developed outside California. As discussed above, Bartolome and Gemmill (1981) also demonstrated that N. pulchra is a semi- ruderal species adapted to disturbed fire-prone hab- itats like those seen by Clements along railroads because its abundant seeds can readily germinate on bare ground and thus permit rapid colonization of sites where disturbance has temporarily reduced competition from other species. Ironically Cle- ments’ BPD is alive and well even though his more famous climatic monoclimax theory from which it arose was discredited and subsequently discarded (Daubenmire 1968; Krebs 1972: Mueller-Dombois and Ellenberg 1974). The fully developed BDP has several key as- sumptions: 1. “Stipa (= N.) pulchra, beyond all doubt, dom- inated the valley grassland”’ (Heady 1977). 2. Following Kiichler (1964), the largest single area of former valley grassland in California in- cluded much of the Central Valley. 3. Replacement of native bunchgrasses by the ex- 254 otic annuals that presently dominate California grasslands started with overgrazing in the 1850’s and accelerated during severe drought in the 1860's. OBJECTIVES AND METHODS To determine if BDP’s assumptions are accurate I looked for evidence about the nature of central California’s pre-agricultural grasslands in two plac- es: 1. Historical accounts describing central Califor- nia grassland during the period their domination by bunchgrasses was assumed by BDP. 2. Vegetation surveys of central California grass- land areas dominated by native species. These start- ed along the Cosumnes River in southern Sacra- mento County where my interest in the BDP prob- lem began but ultimately included several other ar- eas of interest. Each survey consisted of a sample stand (relevé) in essentially homogeneous vegeta- tion and used Braun-Blanquet’s cover-abundance scale (Mueller-Dombois and Ellenberg 1974) to es- timate cover of all vascular plant species in the stand. In the scale 5 = >75% cover, 4 = 50-75%, 3 = 25-50%, 2 = 5—25%, 1 = numerous but <5%, and + = few. The scale also includes r for solitary individuals but all such records were assigned to +. RESULTS Historical records Central California bunchgrass prairies dominated by N. pulchra were first clearly described by Fre- mont (1848) in 1845 and Bryant (1985) in 1846, well before they were purportedly replaced by ex- otic annual grasses in the Central Valley. The bunchgrass prairies described by Fremont and Bry- ant were located, respectively, in the Sierra Nevada foothills and the inner Coast Ranges, however, and both authors clearly distinguished these upland prairies from quite different more continuous grass- lands they found on the nearby Central Valley floor. While traveling through the Sierra Nevada foot- hills in 1845, for example, Fremont reported that ‘““Emerging from the woods, we rode about sixteen miles over an open prairie, partly covered with bunch grass, the timber reappearing on the rolling hills of the River Stanislaus in the usual belt of evergreen oaks.”’ (Fremont 1848). A year earlier, in contrast, while camped at the site of the present city of Sacramento he found that “Here the grass is smooth and green, and the groves very open; the large oaks throwing a broad shade among sunny spots...” (McKelvey 1955). In my opinion Fre- mont’s contrasting observations distinguish be- tween Leymus triticoides prairie, which has a smooth aspect and numerous relict stands near Sac- ramento, and Nasella-dominated bunchgrass prai- rie, which has a rough aspect and many relict stands in the Sierra foothills. MADRONO [Vol. 48 Bryant similarly clearly distinguished central California prairies observed just four days and about 45 miles apart on an 1846 journey between what are now Sacramento and Alameda counties: | **September 14.—We crossed the Coscumne riv- | er [sic] about a mile from our camp, and trav- | elled over a level plain covered with luxuriant grass and timbered with evergreen oak, until | three o’ clock, when we crossed the Mickelemes | .. The soil of the bottom | fiver [Sic], .. 3. witere %.. appears to be very rich, and produces the finest qualities of grass. ‘September 18.—... From this plain we en- | tered a hilly country, covered to the summits of | the elevations with wild oats and tufts or bunchs | of a species of grass, which remains green | through the whole season.” (Bryant 1985). It is clear Bryant on September 18 is describing | his first observation of a Nasella species on hills near the present city of Livermore, where bunch- grass prairie relicts are still frequent. If the “‘luxu- { ) ( ) \ | 1 | riant grass”’ he had seen four days earlier along the | Cosumnes and Mokelumne rivers was also a bunchgrass, he would have said so. More likely it | was Leymus triticoides, which still dominates nu- merous prairie relicts along those rivers and is the | only common native or non-native Central Valley _ grass associated with oaks which would be “‘luxu- | riant’”’ in September before irrigation was widely | introduced to the Central Valley in the 1860’s and 1870’s (Hundley 1992). Contrary to popular impression, the grassland vegetation of central California valley floors was | clearly described well before 1850. Early accounts © by travelers leave little doubt domination of valley | grasslands by bunchgrasses like N. pulchra was generally absent in the period BDP assumes it oc- curred. Jedediah Smith, for example, camped | twelve miles south of the American-Sacramento © river confluence on February, 1828, and wrote: ‘*The whole face of the country is a most beautiful green, resembling a flourishing wheat field”’. (Bur- cham 1957). The hummocky tussocks of bunch- grass prairies never produce smooth grass canopies characteristic of wheat fields, but relict native prai- ries near the site of Smith’s camp (see below) still resemble wheat fields because of their domination by L. triticoides, a rhizomatous native perennial grass whose name reflects its strong resemblance to wheat. Two years earlier, in 1826, Captain Beechey re- ported that three members of his expedition (Collie, Marsh, and Evans) described how a valley floor near what is now the border between San Mateo and Santa Clara counties ‘“‘opened out on a wide country of meadow land, with clusters of fine oak free from underwood. It strongly resembled a no- bleman’s park’? (McKelvey 1955). The European 2001] parks familiar to members of this British expedition are dominated by stoloniferous or rhizomatous grasses like Agrostis tenuis Sibth. and Poa praten- sis L. (Hessayon 1990; Jenkins 1994) resembling _L. triticoides much more than non-rhizomatous N. pulchra. The site of this observation is now covered by highly urbanized ‘‘Silicon Valley’’, but relict na- tive prairies dominated by L. triticoides are still present nearby at a sea level ecotone with saltmarsh located just across San Francisco Bay at Coyote Hills Regional Park (personal observation). Relict bunchgrass prairies dominated by N. pulchra do oc- cur in hills east and west of the bay but are absent from plains near sea level. A striking common feature of all these early ac- counts is the great similarity between what they re- ported and vegetation present in the same places today wherever landscapes have not been altered by agriculture or urbanization. That is not what the BDP predicts since it imagines California valleys were covered by Nasella pulchra before overgraz- ing in the 1850’s and 1860’s caused its replacement by exotic weedy species from Eurasia (Heady 1977). Traveler’s reports from earlier periods make it clear, however, that two exotic Eurasian annuals, Avena fatua L. and Erodium cicutarium (L.) L Heér., were common and dominant before 1850. Bryant’s 1846 description, cited above, of wild oats and bunchgrass codominating a Coast Range hillside is a typical but not particularly early example of such reports (Bryant 1985). Oceanic pollen deposits of E. cicutarium in the Santa Barbara Basin demon- strate its abundance in the California Floristic Prov- ince by 1751-1765, well before California’s first European settlement in 1769 and possibly the result of its introduction to the Baja Californian part of the province in the early 1750’s by the Jesuit ex- plorer Consag and subsequent explosive natural spread northward facilitated by openings created by pre-European burning. (Mensing and Byrne 1999). Its presence with A. fatua in adobe bricks of Cali- fornia’s oldest European buildings (Burcham 1957) is consequently explained. Vegetation surveys Numerous small relict prairies are still extant in central California, where they are particularly fre- quent in Sacramento County near the places they were seen and described in the nineteenth century by Jedediah Smith, John C. Fremont, and Edwin Bryant. Most frequently these prairie remnants are dominated by Leymus triticoides, but two grami- noid sedges, Carex barbarae Dewey and C. prae- gracilis W. Boott, are also often very important el- ements. Relict prairies near Sacramento occur both in the open and as groundcover in valley oak (Quercus lobata Nee) savannas, woodlands, and forests, which are all present at The Nature Con- servancy’s Cosumnes River Preserve in the south- ern part of the county. If “‘savanna’’ is interpreted HOLSTEIN: PRE-AGRICULTURAL GRASSLAND IN CENTRAL CALIFORNIA 25) broadly enough to include open valley oak wood- land, Griffin’s (1977) statement ‘Relatively undis- turbed savannas of Quercus lobata have not been available on prime alluvial soils for over a century ... fortunately does not accurately describe the Cosumnes Preserve. A variety of plant communities occur at the re- serve in a sequence which changes with increasing distance from its streams. Riparian forest dominat- ed by Populus fremontii S. Watson occurs in the immediate vicinity of the Cosumnes River and its associated sloughs but is replaced by closed-canopy valley oak forest a short distance away from them. As distance from the river and sloughs increases, space between valley oak crowns also progressively enlarges so a vegetation sequence from closed for- est through open woodland to savanna is formed. This sequence appears to result from a correlation between distance from rivers and sloughs and in- creasingly unfavorable soil conditions that may in- clude greater competition for groundwater (Walter 1979) rather than from historic land uses since it is most evident where signs of disturbance are scarc- est. In the sequence at the preserve from closed forest to savanna, valley oak density per hectare declines from 540 to 2.7 and soil shifts from Cos- umnes silt loam to Dierssen sandy clay loam. The former is an Aquic Xerofluvent occurring on the Cosumnes River floodplain, and the latter, an Argic Durixeroll that is consequently much less permea- ble to both water and air, is on the rims of basins more distant from the river (Tugel 1991). The Cos- umnes is one of California’s few undammed rivers, and in most years large areas of the preserve are covered by winter flood waters, which often do not recede for several months. At the preserve prairie vegetation is best devel- oped as an understory in open valley oak forest. Cover estimates from twelve such stands are pro- vided below (Table 1). All are on Cosumnes silt loam except | and 6, which respectively are on Di- erssen sandy clay loam and Columbia sandy loam (Tugel 1991). All the above stands were sampled between Au- gust and December, 1994, and are representative of open oak forest in the western and oldest part of the preserve. Some differences among them are ev- ident, however. The lower valley oak cover of stand 3 indicates it is located at the open forest’s outer edge ecotone with savanna, where soils become heavier and water stress greater. In contrast, signif- icant Vitis and Fraxinus cover in stands 5—8 indi- cate their location is at open forest’s inner edge ecotone with oak- and cottonwood- (Populus) dom- inated closed riparian forest, where soils are lighter and water stress reduced. Similar open oak forest vegetation occurring on two recent additions to the preserve located east of its original oldest portion was sampled in May— June, 1995. The Orr Ranch addition, located im- mediately east of the original preserve, had been 256 TABLE |. MADRONO BRAUN-BLANQUET COVER-ABUNDANCE VALUES AND MEANS FOR SPECIES IN 12 OPEN OAK FOREST-RELICT PRAI- RIE SAMPLE STANDS AT THE NATURE CONSERVANCY’S CONSUMNES RIVER PRESERVE. The cover class mean (with + assigned a value of 0.1) of each species is given at right. — Sample stand: Quercus lobata Leymus triticoides Carex barbarae Carex praegracilis Rosa californica Lolium multiflorum Lactuca serriola Toxicodendron diversilobum Rumex crispus Cichorium intybus Convolvulus arvensis Lotus corniculatus Phyla nodiflora Lotus purshianus Lepidium latifolium Carduus pycnocephalus Atriplex triangularis Raphanus sativus Rumex conglomeratus Cirsium vulgare Picris echioides Cordylanthus pilosus Gylcyrrhiza lepidota Lathyrus jepsonit Asclepias fascicularis Foeniculum vulgare Plantago lanceolata Plantago major Barbarea vulgaris Cyperus eragrostis Fraxinus latifolia Epilobium brachycarpum Vitis californica Aster chilensis Mentha pulegium Euthamia occidentalis Rubus ursinus Acer negundo Cornus glabrata Rubus discolor Xanthium strumarium Salix exigua Asparagus officinalis a5 Rumex pulcher BWM] Ne N — =a DPS We +etewe ++H++4++4+ ++eReNN + +e +e + acquired a short time before the survey, while the Valensin Ranch, located somewhat farther east near Highway 99, was acquired by the preserve after the survey. The superior floristic richness of the addi- tion surveys (Table 2) reflects their spring-summer (rather than summer-fall) timing and the outstand- ing condition of open oak forest at Valensin Ranch. The Orr Ranch sample is on Dierssen sandy clay loam and the Valensin on Liveoak sandy clay loam (Tugel 1991). Not all vegetation at the Cosumnes Preserve is open oak forest. Three stands representative, re- spectively, of open savanna (OS), oak-dominated closed riparian forest (OR), and cottonwood-dom- = 6 7 8 Y 10 i 5 Nn N N Nn N ReSeNON WN = = WN Nn — MN —_— © N —— +4 oS NO — N N N — — — — — — N — inated closed riparian forest (CR) sampled in Sep- tember, 1994, are provided below in Table 3. The open savanna sample stand is on Dierssen sandy clay loam, and the two riparian forest sample stands are on Columbia sandy loam (Tugel 1991). The great difference between savanna and closed riparian forest vegetation is evident in Table 3 since valley oak is their sole species in common. The oak and cottonwood dominated types of closed riparian forest are much more similar, however, and have many species in common since both are subject to frequent extended flooding. Such flooding also pre- vents development of a graminoid understory, and its absence readily distinguishes them from open [Vol. 48 2001] TABLE 2. BRAUN-BLANQUET COVER-ABUNDANCE VALUES AND MEANS FOR SPECIES IN 2 OPEN OAK FOREST-RELICT PRAIRIE SAMPLE STANDS LOCATED, RESPECTIVELY, AT ORR AND VALENSIN RANCHES IN THE NATURE CONSERVANCY’S COSUMNES RIVER PRESERVE. Valensin Sample stand: Orr Ranch Ranch x Quercus lobata =) > 5.0 Leymus triticoides 2, 2 2.0 Anthemis cotula 2 a 1.1 Cichorium intybus 2 1.0 Lolium multiflorum 2 l 1.5 Bidens frondosa | Si 0.6 Stellaria media 2 + 1.1 Vicia sativa ] 0.5 Conium maculatum l +p 0.6 Lythrum hyssopifolium ] 1 1.0 Carex praegracilis 2 l less: Carex barbarae 2 4 3.0 Cyperus eragrostis 7 1 0.6 Lactuca serriola + ae 0.1 Rumex conglomeratus l ] 1.0 Raphanus sativus ae + 0.1 Rumex crispus 1 1.0 Toxicodendron diversilobum l + 0.6 Sisymbrium officinale ] oP 0.6 Solanum americanum ] a 0.6 Chenopodium ambrosioides ] 0.5 Sonchus oleraceus “F 0.1 Hordeum murinum + si 0.1 Bromus diandrus + ar 0.1 Cirsium vulgare + 0.1 Melilotus indica ++ 0.1 Chlorogalum pomeridianum 1 1.0 Polygonum punctatum 1.0 Rumex pulcher 0:5 Rubus discolor 0.1 Xanthium strumarium O5 Ranunculus muricatus Sonchus asper Geranium dissectum Rosa californica Atriplex triangularis Phyla nodiflora Convolvulus arvensis Stachys ajugoides Brodiaea elegans Ammi visnaga Amaranthus albus Anthriscus caucalis Polypogon monspeliensis Dipsacus fullonum Maclura pomifera Mimulus gutatus Medicago arabica Plantago major Juncus xiphioides t+Ht+H++HHe He eee +eee¢et¢ene Sesesesessesessesesesesorsesese Cee eee en NO ee ee eee oak forest. Most species largely confined to the cot- tonwood-dominated phase of closed riparian forest, including Populus fremontii itself and Cephalan- thus occidentalis L., are particularly adapted to very frequent flooding and associated coarse allu- vial soils (Holstein 1984). HOLSTEIN: PRE-AGRICULTURAL GRASSLAND IN CENTRAL CALIFORNIA 207 TABLE 3. BRAUN-BLANQUET COVER-ABUNDANCE VALUES FOR SPECIES IN REPRESENTATIVE OPEN SAVANNA (OS), OAK-DOMINATED CLOSED RIPARIAN FOREST (OR), AND COT- TONWOOD-DOMINATED CLOSED RIPARIAN FOREST (CR) SAM- PLE STANDS AT THE CONSERVANCY’ S COSUMNES RIVER PRE- SERVE. O n OR CR 5 3 Sample stand: Quercus lobata Lolium multiflorum Distichlis spicata Lactuca serriola Rumex crispus Hirschfeldia incana Centaurea solstitialis Bromus hordeaceus Hordeum marinum Leymus triticoides Grindelia camporum Rumex pulcher Rubus ursinus Vitis californica Acer negundo Cyperus eragrostis Rosa californica Rubus discolor Fraxinus latifolia Cirsium vulgare Picris echioides Rumex conglomeratus Populus fremontii Euthamia occidentalis Plantago major Oenanthe sarmentosa Cephalanthus occidentalis Polygonum punctatum Artemisia douglasiana Urtica dioica +t+eReENNNKFNYK NN UND + tee Ree EP NNWUY WWeN +ten+teENUNGE The lower cover of the native prairie grass Ley- mus triticoides in open savanna relative to open oak forest at the Cosumnes Preserve appears to contra- dict the early reports discussed above that suggest it once dominated many of California’s open areas. Evidence of prairie remnants in open areas farther north in Sacramento County at the new Stone Lakes National Wildlife Refuge supports the reports, how- ever, and suggests present scarcity of native prairie grass in the Cosumnes savannas results from for- mer cultivation, which did not occur in the open forests. The preserve is now planting trees to re- store extensive areas where they were removed to facilitate farming, but cultivation often occurred in the savannas without general removal of their wide- ly scattered large valley oaks. In June-July, 1995, two prairie remnants were located at Stone Lakes refuge in open grassland lacking valley oaks or other woody plants and sub- sequently sampled. The larger, which covered 930 m? and is on Dierssen sandy clay loam, is repre- sented in Table 4 below by sample SL1, while the smaller, SL2, covered 56 m? and is on Clear Lake clay (Tugel 1991). Three contemporary samples 258 MADRONO [Vol. 48 TABLE 4. BRAUN-BLANQUET COVER-ABUNDANCE VALUES FOR SPECIES AT FIVE SACRAMENTO COUNTY SAMPLE STANDS. Three (MC 1-3) are in oak forest along Morrison Creek and two (SL 1-2) are in relict prairies at Stone Lakes National | Wildlife Refuge. = @ Sample stand: Quercus lobata Leymus triticoides Rubus ursinus Sambucus mexicanus Toxicodendron diversilobum Foeniculum vulgare Juglans hindsii Rubus discolor Cornus glabrata Rosa californica Vitis californica Prunus cerasifera Lactuca serriola Bromus diandrus Lolium multiforum Stellaria media Sisymbrium officinale Sonchus oleraceus Rumex conglomeratus Raphanus sativus Convolvulus arvensis Rumex crispus Bidens frondosa Urtica urens Chenopodium berlandieri Atriplex triangularis Chenopodium murale Xanthium strumarium Asparagus officinalis Galium aparine Solanum americanum Lepidium latifolium Malva nicaeensis Cyperus eragrostis Avena fatua Vicia villosa Trifolium hirtum Hemizonia pungens Rorippa palustris Hirschfeldia incana Helianthus annuus Phyla nodiflora Gnaphalium luteo-album Anthemis cotula Polygonum arenastrum totaeere +t nr ne Nn from oak forest along Morrison Creek a short dis- tance north of the refuge are provided for compar- ison. These resemble similar forest at the Cosumnes Preserve, and consist of one oak-dominated closed riparian forest sample, MC1, and two open oak for- est samples, MC2 and MC3. All are on Egbert clay (Tugel 1991). In July-August, 1995, five additional native prai- rie remnants varying in area from 230 to 470 m? and resembling the Stone Lakes remnants in having few woody plants were located and sampled on Coast Range hillslopes in northwestern Contra Cos- ta County. They are represented below by samples MC2 MC3 SLI SL2 5 5 4 5 5 5) 2 1 2 + 1 + 1 2 2 ar 2 2 a 3 Es 2 1 | 1 oa 1 1 + ] — ot i I + l at ae + 1 + + 4 + of + 2 z + + + + + 2 2 + + 1 + 1 + + 1 CC 1-5 in Table 5. CC 1 and 2 are on Sehorn clay, CC 3 and 4 are on Millsholm loam, and CC 5 is on Clear Lake clay (Welch 1977). DISCUSSION Several theoreticians important in the first half of the twentieth century like J.C. Willis, Sigmund Freud, and Frederic Clements later fell from favor when it became clear their theories did not describe reality. That happened for Clements in at least two areas. He was a leading American opponent of Dar- winian evolution (Hagen 1992), the central para- fe a ' 2001 ] HOLSTEIN: PRE-AGRICULTURAL GRASSLAND IN CENTRAL CALIFORNIA 259 TABLE 5. BRAUN-BLANQUET COVER-ABUNDANCE VALUES AND MEANS FOR SPECIES AT 5 RELICT PRAIRIE SAMPLE STANDS ON COAST RANGE HILLSLOPES IN NORTHWESTERN CONTRA COSTA COUNTY. Sample stand: CCl CC2 Leymus triticoides 3 4 Silybum marianum 1 Carduus pycnocephalus l | Sonchus oleraceus | Lactuca serriola ] ] Avena fatua 2 | Brassica nigra 4 Toxicodendron diversilobum 2 Aira caryophyllea 3 Vulpia myuros 2 Conium maculatum Epilobium brachycarpum Picris echioides Brassica rapa Chlorogalum pomeridianum Hirschfeldia incana Clarkia unguiculata Madia gracilis Phleum pratense Sonchus asper Bromus madritensis Scrophularia californica Amsinckia menziesii Gnaphalium californicum Geranium dissectum Bromus diandrus digm of biology, and developed monoclimax the- ory, which assumed vegetation in each climate zone converges toward a common type (Hamilton 1997). Monoclimax was long influential but has gradually failed as evidence accumulated that climate is one of the least stable environmental factors. In Cali- fornia, for example, the climate 150 years ago in the Little Ice Age was significantly colder and wet- ter than at present during the lifetimes of many in- dividual trees in old growth forests (Bradley 1999; Fagan 2000). Theories inaccurately describing reality are clear- ly social constructions even when arising within science, but recently some elements in the human- ities have claimed all science and even reality itself is a social construct (Gross and Levitt 1994; Sokal and Bricmont 1998). Such claims may just be new weapons in an old war for academic influence, but science still needs to police itself and eliminate any lingering social constructs it still contains. The BDP appears to be a good example, but Clements should not entirely be blamed for that. He eventu- ally recognized Leymus triticoides formerly domi- nated much of the Central Valley (Clements and Shelford 1939), but by that time his academic in- fluence had waned (Hagen 1992) and the world was more preoccupied by war than science. It was later figures like Heady (1977) who provided the dog- matic character of today’s BDP. Before then opinion regarding California grassland ecology was more eclectic. The conclusions of Biswell (1956) and CC3 CES x N+N s) ie) NN a) —_ N N o ') Burcham (1957), for example, are generally com- patible with my field observations; those of Heady are not. So what were California’s pre-agricultural grass- lands really like? The vegetation samples reported above and numerous other field observations sug- gest sufficient relict evidence remains to reasonably reconstruct their basic nature. Leymus triticoides dominated most central California grasslands on sites with clay or loam soil, flat to moderately slop- ing topography, precipitation above 250 mm per year, and moderate to high fertility. Relict stands of L. triticoides are frequent on such sites, but more often they are dominated by Lolium multiflorum Lam., Bromus diandrus Roth, or Avena fatua. Sim- ilarly frequent relict stands of Nassella pulchra sug- gest pre-agricultural dominance shifted to it on sites with steeper slopes, coarser soils, and lower fertil- ity; which today are most frequently dominated by Bromus hordeaceus L., Avena barbata Link, Cy- nosurus echinatus L., and Taeniatherum caput- medusae (L.) Nevski. Wester (1981) previously ar- gued that arid valleys like the southern San Joaquin with precipitation <250 mm per year were pre-ag- riculturally dominated by annual rather than peren- nial species. Such annuals were largely spring-ac- tive forbs. Three interesting questions remain: 1. Is L. triticoides always L. triticoides? Stebbins and Walters (1949) concluded from 260 their chromosome and hybridization studies of L. triticoides and L. condensatus (C. Presl) A. Love that much of the grass traditionally called L. triti- coides in central California is actually a largely sterile hybrid between that species and L. conden- satus. They strongly implied the hybrid, which they called Elymus triticoides Buckl. ssp. multiflorus Gould and distinguished from typical L. triticoides by its possession of 3 to 7 rather than 2 spikelets per central spike node, is much more common than either of its parents and includes all hillslope pop- ulations like those from northwestern Contra Costa County sampled above. I don’t doubt their conclu- sions since the senior author’s contributions to Cal- ifornia plant evolution remain unequaled and partial sterility arising from natural hybridization accounts for the rarity of viable seed in the central California grass currently most often called L. triticoides. Nevertheless, the current name for E. triticoides ssp. multiflorus is L. Xmultiflorus (Gould) Bark- worth, a grass far too large (Barkworth 1993) to match Stebbins and Walters’ abundant hybrid or any of the Leymus populations sampled above. Re- cent central California local floras (Best et al. 1996: Ertter 1997; Matthews 1997; Oswald and Ahart 1994) also fail to follow Stebbins and Walters’ tax- onomy since all treat L. triticoides as common and L. Xmultiflorus as rare to completely absent. Hy- brid or not, the common central California Leymus, because of its small size and strongly rhizomatous habit, is physiognomically close to typical L. triti- coides and distant from L. condensatus, which is only weakly rhizomatous and one of California’s largest native upland grasses. All stands sampled above matched L. tritcoides in size and were strongly rhizomatous; so much so that stands and single clones were probably often equivalent. Con- sequently the name L. triticoides is applied here to all samples even though, following Stebbins and Walters (1949), it is likely their fertility is de- pressed through introgression from the L. conden- satus genome. 2. If the BDP is wrong, why is Nassella pulchra at Jepson Prairie? Nassella pulchra is common on the Central Val- ley floor in at least one place, the Jepson Prairie Preserve located in Solano County at the northern edge of the Montezuma Hills. Since the preserve has generally flat topography, the above discussion might lead to the expectation it is dominated by Leymus triticoides, which is actually rare to absent there. How is that accounted for? One explanation, of course, is that the BDP is true despite arguments against it provided here. Wester (1981) provided another when he concluded “‘Areas of bunchgrass [at Jepson Prairie] occupy relatively moist sites in- fluenced by the cool, humid, maritime air able to penetrate to this part of the Central Valley through the San Francisco Bay gap. These conditions are not typical over the remainder of the Valley.’ His MADRONO [Vol. 48 conclusion has merit since the distribution of Nas- sellla pulchra in California closely matches oak woodland (Beetle 1947; Dremann 1987), and sev- eral species associated with that vegetation type oc- cur on the Central Valley only where the flow of maritime air is strongest. Examples are the occur- | rence of the trees Quercus agrifolia Nee and Aes- culus californica (Spach) Nutt. (Griffin and Critch- field 1972; personal observation) and the mammals Sylvilagus bachmani riparius (riparian brush rabbit) and Neotoma fuscipes riparia (riparian woodrat) (Zeiner et al. 1990) in a localized area of the Cen- tral Valley floor that includes San Joaquin and southern Sacramento counties. What Wester’s hy- pothesis does not explain is why L. triticoides rath- er than N. pulchra is dominant in northwestern Contra Costa County, where flow of maritime air is even stronger than at Jepson Prairie. As discussed above, L. triticoides is favored by heavy clay and loam soils and N. pulchra by lighter ones since bunchgrasses are better adapted for drought stress than rhizomatous grasses (Grime 1979). The clay favored by L. triticoides holds more water than the light, frequently sandy soils hosting N. pulchra (Kramer 1969), and the latter’s usual occurrence on slopes also accelerates runoff and consequent drought stress. In northwestern Contra Costa County, for example, L. triticoides oc- curs on a series of relatively heavy soils: Clear Lake, Diablo, and Sehorn clays; Conejo, Lodo, and Los Osos clay loams; and Los Gatos, Millsholm, and Tierra loams (Welch 1977). At Jepson Prairie, in contrast, N. pulchra occurs on a light soil, San Ysidro sandy loam (Bates 1977); an environmental factor accounting for the vegetational difference be- tween the two relict prairie areas better than mari- time air flow. Light soils may also account for Cle- ments’ momentous observation of N. pulchra near Fresno since a large area of sandy soil derived from Kings Canyon’s glacial outwash occurs near there (Storie and Weir 1951). Even on generally flat val- ley floors aeolian movement of sand creates hum- mocky microtopography (Selby 1985) containing small versions of the slopes favored by N. pulchra. Grassland on sandy soil is among California’s least known vegetation types since its ease of cultivation made it particularly attractive to early valley farm- ers. Consequently a high percentage of California’s extinct plant species (e.g., Eriogonum truncatum Torrey and A. Gray and Monardella leucocephala A. Gray) were associated with sandy grassland (Skinner and Pavlik 1994). N. pulchra at Jepson Prairie now faces a different threat, overprotection. It has steadily declined there since preserve estab- lishment eliminated even the light grazing required to maintain grass diversity by causing release from exotic annual competition (Howe 1999). Ironically grazing elimination at Jepson Prairie was motivated by BDP myths. | | NORTH OF FAIRFIELD IN SOLANO COUNTY (L). _ Taeniatherum caput-medusae 2001] HOLSTEIN: PRE-AGRICULTURAL GRASSLAND IN CENTRAL CALIFORNIA 261 TABLE 6. BRAUN-BLANQUET COVER-ABUNDANCE VALUES FOR SEVERAL RELICT PRAIRIE SAMPLE STANDS: FIVE (WITH MEANS) NEAR FOLSOM IN SACRAMENTO COUNTY (F 1-5), ONE NEAR GREEN ISLAND IN NAPA County (N1), AND ONE Sample stand: Holocarpha virgata Bromus hordeaceus | Leontodon taraxacoides Lolium multiflorum _ Briza minor Trifolium microcephalum Castilleja attenuata Hemizonia congesta Centaurea calcitrapa Vulpia myuros Xanthium strumarium Picris echioides Polypogon monspeliensis Rumex crispus Rumex pulcher Convolvulus arvensis Cichorium intybus Centaurea solstitialis Cirsium vulgare Lythrum hyssopifolium Hypochaeris radicata Hordeum marinum Bellardia trixago Kickxia elatine Eremocarpus setigerus Polygonum arenastrum Raphanus sativus Agrostis avenacea Plantago lanceolata Epilobium brachycarpum Sonchus oleraceus Juncus bufonius Malvella leprosa Lactuca saligna Crypsis schoenoides Lactuca serriola Bromus diandrus Lepidium latifolium Foeniculum vulgare Phalaris aquatica Eryngium aristulatum Hirschfeldia incana Leymus triticoides Distichlis spicata Amaranthus hybridus Avena barbata Hordeum murinum Hemizonia corymbosa Malva nicaeensis Cynodon dactylon Ammi majus Erodium cicutarium Atriplex triangularis Carduus pycnocephalus Hordeum brachyantherum Lupinus formosus Fl NK NNW HE F2 4 3 ++Neer 3 NNN WH NR NRK WE NeENeEWA + WN ++tHtHHH+Ht+HHHH+tH+HHH ttt ete eee eee Hee EH HHH HH EDA L N 262 3. What was on the old terraces? A significant area of California mapped as bunchgrass prairie in accordance with the BDP (Kuchler 1964) has precipitation too high to fit Wester’s ephemeral annual model (Wester 1981) and lacks both modal environmental conditions and relict examples of L. triticoides and N. pulchra. Along the eastern edge of the Central Valley a near- ly continuous band of old terraces (Wahrhaftig and Birman 1965) is characterized by flat topography and infertile duripan soils like Redding loam (Tugel 1991). Old terraces have received attention because most California vernal pools occur there, but up- lands around the pools are largely ignored. As in other California lowlands, exotic annual grasses like B. hordeaceus, T. caput-medusae, and Aegilops triuncialis L. are important elements of current old terrace vegetation, but it is also frequently domi- nated by a native plant remaining photosyntheti- cally active throughout long, dry Central Valley summers. BMD predicts this old terrace dominant will be a bunchgrass, but it isn’t. It isn’t even a grass. The native dominating many square miles of California’s old terraces is, Holocarpha virgata (A. Gray) Keck, and other tarweeds like Hemizonia congesta DC. dominate extensive areas regularly mapped as grassland elsewhere in California. De- spite their dominance of much California vegeta- tion, however, the first of these tarweeds is not mentioned in two of the most extensive recent treat- ments of state vegetation (Barbour and Major 1977; Sawyer and Keeler-Wolf 1995), and the second ap- pears just once (as Hemizonia multicaulis H. & A., a synonym) without comment in one of the for- mer’s tables. There is little doubt these vegetation- ally important native tarweeds were ignored be- cause of the BDP. They went unnoticed because they aren’t bunchgrasses or even grasses despite their extensive dominance of “‘grassland’’. At this time It can’t be proved conclusively these native tarweeds were as dominant pre-agriculturally as they are now, but the burden of proof they weren’t rests on advocates of the BDP. The dominant tarweeds resemble low-precipita- tion area dominants in being annual forbs but differ in being more productive in summer than spring because they can tap summer soil water unavailable to most other plant species (Walter 1979). Centau- rea solstitialis L., an exotic weed, uses a similar adaptive strategy to destructively invade rangeland throughout northern California (Thomsen et. al. 1996). In valley and foothill prairie remnants with soils similar to those most suitable for bunchgrasses an- other forb, Lupinus formosus E. Greene, is frequent that differs from tarweeds in being perennial. It oc- curs at Stone Lakes refuge away from Leymus tri- ticoides on somewhat sandier sites, is frequent on Delhi sands in Merced County, and also occurs on steep Coast Range foothills north of Fairfield in So- lano County. MADRONO [Vol. 48 Table 6 presents five July, 1999, sample stands of tarweed-dominated vegetation on Argonaut-Au- burn complex soils (Tugel 1991) along East Bid- well Road near Folsom, Sacramento County (F 1— 5); a single September, 1998, sample stand on Haire loam (Lambert and Kashiwagi 1978) near Green | Island Road in Napa County (N1); and a single Au- gust, 1999, lupine prairie sample stand on Dibble- Los Osos loams (Batel 1977) north of Fairfield in | Solano County (L). The greater diversity of the Napa County sample is only partially real since it © covered a larger and less homogeneous area. CONCLUSIONS Clements should not be blamed for the BDP’s current dogmatism. His observations in California were limited, and his last publication dealing with its grasslands (Clements and Shelford 1939) before his 1945 death is reasonably accurate. The BDP hardened into dogma only later. Criticism of Cle- ments and the BPD does not mean the end of knowledge about pre-agricultural California grass- lands or resurrection of bizarre theories that they were once covered by chaparral (Hamilton 1997, Cooper 1922). Early descriptions and current relict vegetation provide a congruent picture of diverse grassland vegetation in pre-agricultural California reflecting similar diversity of climate, geology, and soils. Some important elements are becoming known. It is time to identify the rest in an atmo- sphere free from dogma. ACKNOWLEDGEMENTS Special thanks to John Zentner for his generous en- couragement and support, to Linda Bailey for her valuable assistance, to The Nature Conservancy and especially Greg Elliott for permission to conduct research at their Cosumnes River Preserve, to 3 reviewers for helpful sug- gestions that improved the initial draft, and to Jeff Hart for access to the Stone Lakes Refuge and for his stimu- lating discussions about California grassland history. LITERATURE CITED BARBOUR, M. G. AND N. L. CHRISTENSEN. 1993. Vegeta- tion. Pp. 97-131 in Flora of North America editorial committee (eds.), Flora of North America north of Mexico, Vol. 1, Introduction. Oxford, UK. AND J. Masor. 1977. Terrestrial vegetation of Cal- ifornia. Wiley, New York. Barsour, M. G., B. PAVLIK, EF DRYSDALE, AND S. LIND- STROM. 1993. California’s changing landscapes: di- versity and conservation of California vegetation. California Native Plant Society, Sacramento. BaRKwortTH, M. E. 1993. Leymus. Pp. 1268-1270 in J. Hickman (ed.), The Jepson Manual: higher plants of California. UC Press, Berkeley. BARTOLOME, J. W. AND B. GEMMILL. 1981. The ecological status of Stipa pulchra (Poaceae) in California. Ma- drono 28:172—184. Bates, L. A. 1977. Soil survey of Solano County, Cali- fornia. United States Department of Agriculture Soil Conservation Service, Washington, DC. } ( | | 2001] BEETLE, A. A. 1947. Distribution of the native grasses of California. Hilgardia 17:309-—357. BEST, C., J. T. HOWELL, WALTER AND I. KNIGHT, AND M. WELLS. 1996. A flora of Sonoma County. California Native Plant Society. Sacramento. BISWELL, H. H. 1956. Ecology of California grasslands. Journal of Range Management 9:19—24. BRADLEY, R. S. 1999. Paleoclimatology: reconstructing climates of the Quaternary, 2nd ed. Academic Press, New York. BRYANT, E. 1985. What I saw in California. Univ. of Ne- braska Press, Lincoln. BurcuHaM, L. T. 1957. California rangeland. California De- partment of Natural Resources. Sacramento. CLEMENTS, F. E. 1920. Plant indicators. Carnegie Institu- tion of Washington, Washington, DC. . 1934. The relict method in dynamic ecology. Journal of Ecology 22:39—68. AND V. E. SHELFORD. 1939. Bio-ecology. Wiley, New York. Cooper, W. S. 1922. The broad-sclerophyll vegetation of California: an ecological study of the chaparral and its related communities. Carnegie Institution of Wash- ington, Washington, DC. DAUBENMIRE, R. 1968. Plant communities: a textbook of plant synecology. Harper & Row, New York. DREMANN, C. C. 1987. Grasses of California: distribution maps. Redwood City Seed Company, Redwood City, California. . 1995. Grasses of European origin and their im- pact upon California bunchgrass (California). Grass- lands 5(3):7—10. ERTTER, B. 1997. Annotated checklist of the East Bay flo- ra: native and naturalized plants of Alameda and Con- tra Costa counties, California. East Bay Chapter of California Native Plant Society in association with Jepson Herbarium, Berkeley. FAGAN, B. 2000. The Little Ice Age: how climate made history 1300—1850. Basic, New York. FREMONT, J. C. 1848. Geographical memoir upon upper California. 30th Congress, Ist Session, Senate Misc. Doc. 148. GRIFFIN, J. R. 1977. Oak woodland. Pp. 383-415 in M. Barbour and J. Major (eds.), Terrestrial vegetation of California. Wiley, New York. AND W. B. CRITCHFIELD. The distribution of forest trees in California, USDA Forest Service Research Paper PSW-82. USDA Forest Service, Berkeley. GRIME, J. P. 1979. Plant strategies and vegetation process- es. Wiley, New York. Gross, P. R. AND N. Levirr. 1994. Higher superstition: the academic left and its quarrels with science. Johns Hopkins Press, Baltimore. HAGEN, J. B. 1992. An entangled bank: the origins of eco- system ecology. Rutgers University Press. New Brunswick, NJ. HAMILTON, J. G. 1997. Changing perceptions of pre-Eu- ropean grasslands in California. Madrono 44:311— seek HARKER, D., S. EVANS, M. EVANS, AND K. HARKER. 1993. Landscape restoration handbook. Lewis, Boca Raton, Florida. Heapy, H. FE 1977. Valley grassland. pp. 491-514 in M. Barbour and J. Major (eds.), Terrestrial vegetation of California. Wiley, New York. HEssayon, D. G. 1990. The lawn expert. pbi Publications, Waltham Cross, UK. HOLLAND, V. L. AND D. J. KeIL. 1989. California vegeta- HOLSTEIN: PRE-AGRICULTURAL GRASSLAND IN CENTRAL CALIFORNIA 263 tion. Biological Sciences, California Polytechnic State University, San Luis Obispo, California. . 1995. California vegetation. Kendall/Hunt, Du- buque, Iowa. HOLSTEIN, G. 1984. California riparian forests: deciduous islands in an evergreen sea. Pp. 2—22 in R. Warner and K. Hendrix (eds.), California riparian systems: ecology, conservation, and productive management. UC Press, Berkeley. Howe, H. E 1999. Dominance, diversity, and grazing in tallgrass restoration. Ecological Restoration 17:59— 66. JENKINS, V. S. 1994. The lawn: a history of an American obsession. Smithsonian, Washington, DC. HUNDLEY, JR., N. 1992. The great thirst: Californians and water, 1770s—1990s. UC Press, Berkeley. KRAMER, P. J. 1969. Plant and soil water relationships: a modern synthesis. McGraw-Hill, New York. Kress, C. J. 1972. Ecology: the experimental analysis of distribution and abundance. Harper & Row, New York. KUCHLER, A. W. 1964. Potential natural vegetation of the conterminous United States, Special Publica- tion No. 36. American Geographical Society, New York. LAMBERT, G. AND J. KASHIWAGI. 1978. Soil survey of Napa County, California. United States Department of Agriculture Soil Conservation Service, Washing- ton, DC. MATTHEWS, M. A. 1997. An illustrated field key to the flowering plants of Monterey County and ferns, fern allies, and conifers. California Native Plant Society, Sacramento. McKELvEY, S. D. 1955. Botanical exploration of the trans- Mississippi west 1790-1850. Arnold Arboretum of Harvard University, Jamaica Plain, MA. MENSING, S. AND R. ByRNE. 1999. Invasion of Mediter- ranean weeds into California before 1769. Fremontia. 27(3):6-9. MUELLER-DOMBOIS, D. AND H. ELLENBERG. 1974. Aims and methods of vegetation ecology. Wiley, New York. OSWALD, V. H. AND L. AHART. 1994. Manual of the vas- cular plants of Butte County, California. California Native Plant Society, Sacramento. SAWYER, J. O. AND T. KEELER-WOLF. 1995. A manual of California vegetation. California Native Plant Soci- ety, Sacramento. SCHOENHERR, A. A. 1992. A natural history of California. UC Press, Berkeley. SELBY, M. J. 1985. Earth’s changing surface. Oxford, Ox- ford, UK. Sims, P. I. 1988. Grasslands. Pp. 265-286 in M. Barbour and W. Billings (eds.), North American terrestrial vegetation. Cambridge, Cambridge, UK. SKINNER, M. W. AND B. M. PAVLIK. 1994. California Na- tive Plant Society’s inventory of rare and endangered vascular plants of California, 5th ed. California Na- tive Plant Society, Sacramento. SOKAL, A. AND J. BRICMONT. 1998. Fashionable nonsense: postmodern intellectuals’ abuse of science. Picador. New York. STORIE, R. E. AND W. W. WEIR. 1951. Generalized soil map of California. University of California Division of Agricultural Sciences, Berkeley. THOMSEN, C., W. WILLIAMS, M. VAYSSIERES, C. TURNER, AND W. LANINI. 1996. Yellow starthistle biology and control. University of California Division of 264 Agriculture and Natural Resources Publication 21541. TuGEL, A. J. 1991. Soil survey of Sacramento County Cal- ifornia. United States Department of Agriculture Soil Conservation Service, Washington, DC. WAHRHAFTIG, C. AND J. H. BIRMAN. 1965. The Quaternary of the Pacific Mountain System in California. Pp. 299-340 in H. Wright, Jr. and D. Frey (eds.), The Quaternary of the United States. Princeton Univ. Press, Princeton, NJ. MADRONO [Vol. 48 | WALTER, H. 1979. Vegetation of the Earth and ecological | systems of the geo-biosphere, 2nd ed. Springer-Ver- lag, New York. : WELCH, L. E. 1977. Soil Survey of Contra Costa County, — California. United States Department of Agriculture Soil Conservation Service, Washington, DC. WESTER, L. 1981. Composition of native grasslands in the — San Joaquin Valley, California. Madrono 28:231—241. ZEINER, D., W. LAUDENSLAYER; K. MAYER, AND M. WHITE. 1990. California’s wildlife: Volume IJ, mammals. Cal- ifornia Department of Fish and Game, Sacramento. MApRONO, Vol. 48, No. 4, pp. 265-271, 2001 THE POLLINATION BIOLOGY OF ERIASTRUM DENSIFOLIUM SSP. SANCTORUM (POLEMONIACEAE), AN ENDANGERED PLANT DEBORAH K. DorsettT!, C. EUGENE JONES, AND JACK H. BURK Department of Biological Science, California State University, Fullerton, CA 92834-6850 ABSTRACT An analysis of the reproductive biology of the Santa Ana Woolly Star, Eriastrum densifolium ssp. sanctorum (Milliken) H. Mason shows that the primary pollinators of this rare and endangered plant include the Giant Flower-loving Fly, hummingbirds, bumble bees, halictid bees, and digger bees. In comparison to previous studies (Munoz 1991; Erickson 1993; Stone 1995), fewer individuals of each pollinating species were present in this study suggesting that overall fruit set may be declining. However, seed set per fruit appears to be similar to values found in earlier studies indicating that pollinator efficiency is not a limiting factor. In fact, the three principal pollinators that were tested were all reasonably efficient pollen vectors. Only a single simulated pollination event was usually required to effect normal fruit and seed set, indicating that this plant is not dependent on any single pollinator for its reproductive success. The flowers produced an average of nearly 2 wl of nectar per flower, which is in the range of insect pollinated plants. Eriastrum densifolium ssp. sanctorum (Milliken) Mason (Polemoniaceae), the Santa Ana River Woolly Star, (Eds), is one of the 12 most endan- gered plants in California (York 1987). Research toward understanding and preserving this plant has been conducted for more than a decade by faculty and students at California State University, Fuller- ton (CSUF). Eds is distinguished from the other subspecies by the presence of a very long floral tube that normally exceeds 25 mm in length (Bru- nell, 1999). This short-lived, perennial subshrub with woolly pubescence occupies sand _ terraces within the floodplain of the Santa Ana River at the base of the San Bernardino Mountains in San Ber- nardino County, California (Munoz 1991, Brunell 1996, 1999). The restricted distribution of Eds and loss of habitat led to the listing of the subspecies as endangered by the United States Fish and Wild- life Service (USFWS) in 1987. The completed Sev- en Oaks Dam is expected to further imperil the sur- vival of this species by reducing the deposition of new sand that is required for the establishment of new populations (Burk et al. 1988). A management plan for the Santa Ana River Woolly Star (Burk and Jones 1993) was approved in November of 1993 with the goal of maintaining a healthy and geneti- cally viable Eds population. A detailed knowledge of the pollination biology of any species is valuable, but this knowledge often becomes critical when trying to prevent rare and endangered species from becoming extinct. It is of little use to preserve habitat for an endangered plant species if, at the same time, the habitat for its pollen 'This paper is dedicated to the memory of Deborah K. Dorsett, who died 29 April 1999 in a tragic automobile accident. Deborah was an outstanding biologist, writer, and human being. vectors is destroyed. The intent of these studies was to augment the existing knowledge of Eds present- ed elsewhere (see Zembal and Kramer 1985; Burk et al. 1988; Burk et al. 1989; Munoz 1991; Burk and Jones 1993; Erickson 1993; Stone 1995; Jones and Burk 1996). In addition to pollination biology, studies were conducted on the breeding biology of Eds (in prep.) and the life history of two of its principle pollinators, Rhaphiomidas acton (Stein- berg et al. 1998) and Halictus tripartitus (in prep- aration). Although other studies were completed on the pollination biology of this species, this is the first study in which the research was conducted si- multaneously over the entire extant range of Eds in the Santa Ana River Wash north of Redlands, Cal- ifornia, in the country of San Bernardino. METHODS Study areas. Five study sites, located within the Santa Ana River floodplain north of Redlands, San Bernardino County, California, were selected to cover the range of habitats and habitat ages within the floodplain that currently support all known pop- ulations of Eds (see Fig. 1). Three plots, where 100 or more Eds individuals formed a continuous pop- ulation, were established in the immediate vicinity of each of the 5 study sites for a total of 15 study plots. All work presented here was completed with- in these plots. Diurnal observation. To determine the major pol- linators of Eds at each of the 5 Study Sites, a series of 9 “‘dawn to dusk”’ observations were conducted between 19 June and 6 July 1995. At each of the 5 study sites, three sets of three-day sequences of dawn to dusk observations were conducted for a total of 15-day sequences. The first sequence was done during the early blooming period, June 19 to 117°12'30"W 117°15'W ® SAN BERNARDINO FOOTHILL BLVD = SAN BERNARDINO INTERNATIONAL AIRPORT °o (FORMERLY NORTON AIR FORCE BASE) vt o Pe ee Ps es in Caer ea i . waa? ANNSAV JONVOAddIL . ween e tts is VWwvav iv O€ 3LNOY 3LVLS 1S JONVYO 117°15'W 117°12'30"W Fic. 1. MADRONO “ee. [Vol. 48 117°10'W ® HIGHLAND 117°07'30"W N.80.ve GREENSPOT ROAD Perennial Channel ste! (dashed centerline) ey Seasonal Channel =——S NT N.20.E ee . of * ot ate * avost ay on iit} * . on oot wie PY Tea eece . wn nee ant . COS) (See cee ana Paani eye fev ba Yd © el ealete et Te gay N.90.vE LANDS MUNICIPAL AIRPORT ® REDLANDS N.SO.vE RL. Allen 2001 117°10'W 117°07'30"W Range and study locations for Eriastrum densifolium ssp. sanctorum. Dashed lines indicate tributary creeks and the perennial channel of the Santa Ana River. Dotted lines indicate the seasonal channel of the river. Gray shading shows the entire known modern range of Eds. The numbers 1, 2, 3, 4, and 5 indicate study locations. 22 1995, when approximately 25% of the Eds plants in a given study plot were in bloom. The second sequence was completed during the mid blooming period, June 26—28 1995, when about 50% of the Eds plants in a given study plot were in bloom. The final sequence was done during the late blooming period, July 3—6 1995, when approx- imately 75% of the plants had finished blooming. Observations were conducted for 12 daylight hours beginning at 0700 and continuing until 1900 hours on each of the study days. At each study site, observations of approximately 15—20 Eds plants at plot number | started on the hour and continued until 10 minutes after the hour. Next the observer would move to plot number 2 within the same study site and observe approximately 15-20 Eds plants from 20 minutes after the hour until 30 min- utes after the hour. Finally the observer would move on to plot number 3, the last plot in each study site, where 15—20 Eds plants were observed from 40 minutes after the hour until 50 minutes after the hour. This observation sequence continued hourly throughout the day from 0700 until 1900. Observations were completed simultaneously at all 5 of the study areas by 5 different observers. Therefore, in each 12 hour observation period, 6 hours of pollinator observations were recorded for each site or a total of 30 hours of observation of pollinator activity on Eds plants each day through- out the distribution of this subspecies. All flower visitors that made contact with the anthers, the stig- ma branches, or both were recorded. Insect visitors were identified using a reference collection devel- oped by Douglas Stone during his work on Eds (Stone 1995). All insects were identified and/or ver- ified by Roy Snelling of the Natural History Mu- seum of Los Angeles County. Nectar rewards. Average nectar volume pro- duced per flower was determined by using Drum- mond ‘“‘Microcaps”’ microcapillary tubes. Five flowers on each of 10 plants in each of the 3 plots at each of the 5 study sites were sampled between 20 June and 26 June 1995 for a total of 750 flowers sampled. The flowers sampled were bagged prior to anthesis and nectar was sampled and measured during the morning hours once the flower buds had fully opened (usually within two days of bagging the buds). Pollinator constancy. Tests to determine polli- nator constancy were conducted by collecting pol- len-laden specimens of each of the primary Eds in- sect pollinators (see Table 2 for the species sam- pled). All insects used to determine pollinator con- stancy were collected between 3 July and 6 July 1995 and between 7 June and 18 June 1996 as they visited Eds. The insects were killed and brought back to the laboratory where the purity of their pol- len loads was determined. All insects were pinned for identification. No hummingbirds were sampled. Pollen samples were carefully removed from each of the pollinators collected. Pollen was then placed on a microscope slide, covered by a drop of glycerol, and then covered by a glass cover slip. The slides were examined using a compound mi- croscope, and the number and frequency of each different type of pollen was determined. For each pollen load, the number of Eds pollen grains and the number of non-Eds pollen grains were counted. The percent constancy was determined by dividing the number of Eds pollen grains by the total num- ber of pollen grains of all types counted. The num- ber of different pollen types present provides an estimate of the different flowering plant species each flower visitor to Eds is visiting. Thus, percent constancy data can be used as an estimate of the relative constancy of a given pollinator to Eds. For the purpose of this paper, a pollinator will be con- sidered to be a ‘“‘constant’’ pollinator for a partic- ular species when that pollinator visits only that plant species during a single foraging flight. 2001 | DORSETT ET AL.: POLLINATION OF SANTA ANA RIVER WOOLLY STAR 267 36.5 DAWN TO DUSK OBSERVATIONS - EDS 34 - 32 | 30 - Early Season Data ( Hummingbirds 98 4 19 - 22 June 1995 — Bombus sp. 26 I Halictids | Fa Syrphids 24 1 @ Butterflies/Moths 22 | Ed Apis mellifera Number of Visits © 1000 1100 PIG. 22. observations were recorded during this period. Pollinator efficiency. Pollinator efficiency is a measure of how effective a given flower visitor is at pollinating a given plant. Using the data from the ‘““dawn to dusk”’ observations, we determined that the three most important pollination vectors visiting Eds during the 1996 flowering season were bum- blebees, (Bombus spp.), hummingbirds and the Gi- ant Flower-loving Fly, (Rhaphiomidas acton (see Steinberg, et al. 1998, for a discussion on the sub- specific designation of this fly). We determined how each of these principal pollinators effects pol- lination by carefully observing which portions of the pollinators bodies came into contact with the anthers and stigmas in each flower. Six inflorescences on each of 5 plants at each of the 5 Study Sites were emasculated in the bud stage and covered with a pollinator exclusion bag. Pol- linator exclusion bags were obtained by purchasing brown (suntan) nylon anklet stockings and cutting approximately 20 cm lengths including the toe por- tion. The nylon bags were placed over an Eds in- florescence and secured with a twist tie. The flow- ers were marked by placing a drop of Elmer’s flu- orescent glue on a sepal of the flower bud to be used. Approximately two days after bagging was completed, and when three flowers on each of the 5 plants at each of the 5 study sites had opened and were ready to be pollinated, freshly killed speci- mens of Bombus californicus and Rhaphidiomidas acton and the head of a stuffed Anna’s humming- bird from the teaching collection at CSUE were in- 1200 Time #4] Melissodes sp. 41 Micranthophora flavocincta C1] Rhaphiomidas acton fi Others Sr SS eS i 1400 1300 1500 1600 1700 1800 Observations of pollinators on literally hundreds of flowers on 45—60 large plants of Eds—270 hours of serted into an open flower of a nearby plant in a manner that mimicked, as closely as possible, the way a living individual approaches a flower. This procedure resulted in pollen deposition in the nor- mal location for each of the principal pollination vectors. Each individual was then inserted into one of the bagged and emasculated test flowers. This procedure was repeated for each of the 3 principal pollination vectors until all the emasculated test flowers had received a single “‘visit’” by each prin- cipal pollinator. To determine the possible effect of repeated pol- linator visits, the entire procedure was repeated with a second set of flowers, but instead of a single insertion, each pollinator was inserted twice into each emasculated test flower. The fruit and seed set per the number of flowers pollinated in all proce- dures was determined and provides an estimate of the relative pollinator efficiency for the three prin- cipal pollinators of Eds in 1996. In addition, three fruits were collected from each of the study plants. These fruits were used as controls to determine the average seed set per pollinated flower in open-pol- linated, non-manipulated situations. RESULTS Diurnal observations. The dawn to dusk polli- nator observations are presented in Figs. 2, 3, and 4. All plots in all 5 Study Sites were summed and are presented for the early season (Fig. 2), mid sea- 26 Number of Visits ee EE EP EP PE Ie Ceres] a a a a 900 1000 1100 PIG: 3: observations were recorded during this period. son (Fig. 3), and late season (Fig. 4). Data were collected and maintained for most species individ- ually. However, for clarity of presentation, polli- nators have been lumped as follows: all humming- birds, all species of bumblebees (Bombus spp), all halictid bees, all syrphid flies, and all butterflies and moths. The minor visitors, including various beetles and wasps, have been grouped together as ‘“‘other.”’ The remaining pollinators are presented as single- species categories. Overall, the most prevalent pol- linators included hummingbirds, species of bumble- bees, halictid bees, digger bees (Melissodes spp.), and the Giant Flower-loving Fly, Rhaphiomidas ac- ton. Variation did occur among the sampling periods. The most common pollinators during the early sea- son observations were halictid bees, primarily Hal- ictus tripartitus. The halictid bees continued to be important pollen vectors throughout the blooming season of Eds, but their numbers declined. Digger bees, (Melissodes spp.), were relatively common throughout the blooming period, but were particu- larly prevalent during the mid season. Humming- birds were important pollinators throughout the en- tire flowering period. They were typically the pre- dominant pollinators during the morning hours and usually had a smaller, but significant burst of visi- tation activity in the late afternoon and early eve- ning hours. The Giant Flower-loving Fly, Rha- phiomidas acton, became a major contributor to the pollination success of Eds during the mid season of MADRONO DAWN TO DUSK OBSERVATIONS - EDS Mid Season Data - 28 June 1995 1200 [Vol. 48 Hummingbirds | Bombus sp. Halictids Ei Syrphids @ Butterflies/Moths Apis mellifera J Melissodes sp. 4 Micranthophora flavocincta 1 Rhaphiomidas acton fH Others FLEE LILLE ELE LEE L ROLE ELL a | a a a 1400 1300 1500 1600 Time Observations of pollinators of literally hundreds of flowers on 45—60 large plants of Eds—270 hours of blooming. This species was one of the dominant flower visitors in the afternoon hours. Nectar rewards. Eds nectar rewards varied con- siderably within and among the Study Sites (see Table 1). When all the data were combined, the average quantity of nectar was 1.91 wl per flower. Pollinator constancy. Pollen loads were exam- ined on eight different insect pollinators of Eds to determine the relative constancy of each of these insect pollination vectors. Data were pooled for all sites and are presented in Table 2. Using our defi- nition of floral constancy, the only truly constant pollinator was the Giant Flower-loving Fly. The other pollinators show tremendous variability in constancy among the sampled individuals. Note that one of the more important groups of pollina- tors, the hummingbirds, was not sampled. Pollinator efficiency. Pollinator efficiency in terms of fruit set for a single touch and for two touches by each of the three primary pollinators, the Giant Flower-loving Fly, bumblebees, and hum- mingbirds is presented in Table 3. Using a t-test, there were no significant differences found between a single touch and two touches for any of the three principal pollinators tested. Pollinator efficiency in terms of seed per fruit for a single touch and for two touches by each of these same three primary pollinators is presented in Table 4. Seed set per capsule was the same regardless of | DAWN TO DUSK OBSERVATIONS - EDS Number of Visits @ gl EE EE EE EP PPP PP PPI \= = IN: : 4 S AE : nN : nN: = i: : 2 : LL AN: : nD i: : i N= = | h: = A ld Oe ~ =a 700 1000 1100 Fic. 4. observations were recorded during this period. treatment and was not significantly different from the control seed set. DISCUSSION A variety of organisms have been shown to be effective pollinators of Eds: hummingbirds, bum- blebees, halictid bees, digger bees and the Giant Flower-loving Fly (see Tables 3 and 4). Other spe- cies have proved to be significant pollinators in pre- vious years including a small bee, Micranthophora flavocincta (Munoz 1991) and hawkmoths (Erick- son 1993). All of the effective pollinators have ‘“‘tongues”’ long enough to reach the nectar reward at the base of the 25-35 mm long floral tubes of Eds. With the diversity of pollinators visiting Eds, it would appear that no particular pollinator should be a limiting factor in the reproductive success of Eds. However, overall numbers of pollinators can influence reproductive success. Although the spec- trum of potential pollinators observed in the most recent studies remained very similar to past obser- vations (See Munoz 1991; Erickson 1993; Stone, DORSETT ET AL.: POLLINATION OF SANTA ANA RIVER WOOLLY STAR Late Season Data 3, 5 - 6 July 1995 1200 Hummingbirds & Bombus sp. Halictids Fs] Syrphids @ Butterflies/Moths Apis mellifera ZA] Melissodes sp. 4 Micranthophora flavocincta 1) Rhaphiomidas acton Others Ye a a a oe oe LLL L LS EES le | 1300 1400 1500 1600 1800 Time 1700 Observations of pollinators on literally hundreds of flowers on 45—60 large plants of Eds—270 hours of 1995; Atallah and Jones unpubl. data), the numbers of individuals of each of these species declined from the numbers observed in dawn to dusk obser- vations in the previous studies. This reduction in the number of individuals of each pollinating spe- cies is correlated with a dramatic reduction in seeds per fruit in the 1995 blooming season. The ovary of each flower of Eds has three locules and each locule normally contains five, rarely to six or seven ovules (Harrison 1972; Munz 1974; Hick- man 1993). Therefore, each fruit (capsule) has the potential of normally having up to 15 seeds. How- ever, Erickson (1993) found that seed per fruit in open pollinated Eds controls at Study Sites | and 5 in 1991 averaged 5.74 and 5.8, respectively. An examination of seeds per fruit in open pollinated controls at these same two sites in 1992 resulted in an average seeds per fruit of 3.89 and 4.21, respec- tively. In contrast, during the 1995 season, average seeds per fruit in open pollinated control flowers, which had been marked in the bud stage (OT), was only 0.94 and 0.91 at Study Sites | and 5. Even when compared to a second set of open pollinated TABLE 1. AVERAGE NECTAR QUANTITIES PER FLOWER FOR THE EDS STUDY SITES. Site | Site 2 Site 3 Site 4 Site 5 N 150 150 150 150 150 Range (in pl) 0.02—3.8 1 0.05—5.38 0.04—5.14 0.01—3.97 0.01—4.17 Average pl Nectar Per Flower 1.28 2.99 2.59 1.36 1.49 270 MADRONO [Vol. 48 TABLE 2. POOLED POLLINATOR CONSTANCY DATA FOR 8 MAJOR INSECT POLLINATORS FOR ALL SITES. N = sample size. s = standard deviation. * Sample size too small for valid sample mean or standard deviation calculations. Species Bombus californicus Melissodes communis Apis mellifera Micranthophora flavocincta Rhaphiomidas acton Melissodes sp. Halictus farinosus Halictus triparitus — NEFENANONDA Z unmarked, control flowers that had set fruit and were selected at the time the fruits were collected (ONT), seeds per fruit at Study Sites | and 5 for the 1995 season averaged only 3.13 and 1.75 seeds per fruit respectively (Atallah and Jones in prepa- ration). What effect this reduction in reproductive success will have on Eds is unknown, but it clearly needs to receive further investigation (Attallah and Jones manuscript in preparation). An understanding of pollinator constancy helps determine how interdependent a flowering plant is with its pollinator(s). The vast majority of Eds pol- linators did not show uniform pollinator constancy (see Table 2). Some individuals of most of the pol- linators investigated had very pure loads of Eds pollen and were quite obviously very constant to Eds. Other individuals of the same species, which were also collected while visiting Eds, were just as obviously visiting several species of flowering plants on a single foraging flight. The Giant Flower- TABLE 3. POLLINATOR EFFICIENCIES OF 3 MAJOR INSECT VECTORS IN TERMS OF FRUIT SET (CALCULATED AS PERCENT POLLINATED FLOWERS SETTING FRuIT). Data from all five sites were pooled. ONT: controls that were established using non-tagged flowers from the same inflorescence containing the experimental flowers. 1TF: one touch by a freshly killed, pollen laden Giant Flower-loving Fly. 1TB: one touch by a freshly killed, pollen-laden bumblebee (Bombus sp). 1TH: one touch from a pollen-laden head of a stuffed Anna’s Hummingbird. 2FT: two touches by a freshly killed pollen laden Giant Flower-loving Fly. 2TB: two touches by a freshly killed, pollen-laden bumblebee. 2TH: two touches from a pollen-laden head of a stuffed Anna’s Hummingbird. SD = Standard Deviation. 2SE = Two standard error of the mean. CI = 95% Confidence Interval. n = 25, (exceptions: 1TB, n = 22, for 1TF and 2TB, n = 23, and for 2TF and 2TB, n = 24). Treatment Mean (%) SD +SE CI (%) ONT 94.7 20.8 8.3 86.4—100.0 ITF 125 35.8 14.9 57.6—87.4 ITB 80.4 24.5 10.5 69.9-90.8 1TH 54.1 33.9 13.8 40.3—68.0 2TF 63.9 32.6 13.3 50.6—77.2 2TB 82.7 Zane 9.2 73.5-91.9 2TH 74.8 30.9 123 62.4—87.1 Mean % Range constancy (%) Ss 40.0 1.4—91.7 36.4 48.2 9.2-89.5 25.9 76.2 2.2—100.0 35.0 74.9* 52.0—97.8 - 94.8 89.1—100.0 3.81 21.6% 6.5—48.7 es * 72.6 > OK 6 ky 76.4—82.7 = loving Fly, Rhaphiomidas acton, is the only prin- cipal pollinator that uniformly exhibited a very high degree of constancy as measured by the purity (94.8%) of the pollen loads that it was carrying. It appears from our data that the Giant Flower-loving Fly may be monotrophic, meaning that they pri- marily visit only a single species of flowering plant for all their nectar and/or pollen needs. We have occasionally observed individuals of the Giant Flower-loving Fly visiting flowers of other species of plants in the Eds habitat, but this is not a com- mon event. We were unable to sample any hummingbirds to determine their constancy, but many observations of hummingbirds in the Eds habitat indicate that they routinely visit many different flowering plants at the same time that they are visiting Eds plants in bloom. These observations lead us to believe that they are also not particularly constant. Nectar volume per flower varied considerably among the sites and within each site (see Table 1). As nectar volume per flower increases, the length of time spent at a flower by a given pollinator may increase (Montgomerie 1984; Harder 1986; Mitch- ell and Paton 1990). The inverse should reduce the time spent at a flower by a given pollinator and thus, increase the number of flowers visited. There- fore, it may be advantageous for plants to either reduce the amount of nectar per flower or, at the very least, exhibit considerable variation in nectar production per flower on the same plant and within TABLE 4. POLLINATOR EFFICIENCIES OF 3 MAJOR INSECT VECTORS IN TERMS OF SEED SET PER FRuit. Data from all five sites were pooled. See Table 3 for an explanation of abbreviations and sample size information. Mean seeds Treatment per fruit SD +2SE Cl ONT 4.75 yas 0.9 3.85—5.65 ITF 4.50 peo) 1.2 3.31-5.69 ITB 528 29 D2 4.04—6.52 1TH ai2N Zod 1.1 2.09—4.33 2TF 3.07 2 Tel 2.77-4.97 2TB Ses: 25 Ll 4.47-6.59 20H Aa/S 2.4 1.0 2.77—4.69 2001] the population of the species (the latter is certainly the case for Eds). Such a strategy should result in an increase in the number of flowers a pollinator visits to achieve a full nectar load. This should in- crease the number of interplant visits and be ad- vantageous for an obligate outcrosser like Eds (Pyke 1978, 1981; Zimmerman 1983). The three principal pollinators we tested are all reasonably efficient pollinators (see Tables 3 and 4). In terms of fruit and seed set, there were no sig- nificant differences between single and dual polli- nator visits (touches) for any of these test pollina- tion vectors. This indicates that any one of these principal pollinators could effectively pollinate a flower of Eds with a single visit and, as a result, effect full or nearly full fruit and/or seed set when compared to the controls. This further indicates that Eds reproduction is not dependent on any single pollination vector. ACKNOWLEDGEMENT We would like to thank Psomas and Associates, Inc., for their support of this research. Special thanks to all the students and staff who helped in the data collection and in editing this manuscript. Specifically, we thank the fol- lowing: Bob Anderson, Barbara Hanlon, Annalisa Miller, David Moskovitz, Beth Pearson, Chan Phommasaysy, Chirag Shah, Frances Shropshire, Leo Song, Margaret Steinberg, Doug Stone, Frank Wegscheider, and Beth Wiese. A special thanks to Bob Allen for Figure 1. LITERATURE CITED BRUNELL, M. 1996. Biosystematics of Eriastrum densifol- ium (Bentham) Mason (Polemoniaceae). Ph.D. dis- sertation. University of California, Riverside. BRUNELL, M. AND R. WHITKUus. 1999. Assessment of Mor- phological Variation in Eriastrum densifolium (Pole- moniaceae): Implications for Subspecific Delimita- tion and Conservation. Systematic Botany 23(3):351— 368. Burk, J. H., C. E. JONEs, J. WHEELER AND S. DESIMONE. 1988. The Ecology of Eriastrum densifolium sanc- torum (Milliken) Mason: Final Report. Prepared for U.S. Army Corps of Engineers, Los Angeles District by Environmental Audit Inc., Placentia, CA. Burk, J. H., C. E. JONES AND J. WHEELER. 1989. New Information on the Rare Santa Ana River Woolly Star. Fremontia 17(3):20—21. Burk, J. H. AND C. E. JoNEs. 1993. Management Plan for the Santa Ana River Woolly Star, Eriastrum densi- folium ssp. sanctorum. Final. Prepared for the U.S. Army Corps of Engineers, Los Angeles District, Los Angeles, CA by Chambers Group, Inc. DORSETT ET AL.: POLLINATION OF SANTA ANA RIVER WOOLLY STAR 271 ERICKSON, M. 1993. Optimal Outcrossing and Pollinator Foraging Distance in the Santa Ana River Woolly Star, Eriastrum densifolium ssp. sanctorum (Milliken) Mason (Polemoniaceae). Master’s thesis, California State University, Fullerton, CA. HARDER, L. D. 1986. Effects of nectar concentration and flower depth on flower handling efficiency of bumble bees. Oecologia (Berlin) 69:309-—319. HARRISON, H. K. 1972. Contributions to the study of the genus Eriastrum: Il Notes concerning the type spec- imens and descriptions of the species. Brigham Young University Science Bulletin 16:1—26. HICKMAN, J. C. (Ed.) 1993. The Jepson Manual, Higher Plants of California. University of California Press, Berkeley, California. JONES, C. E. AND J. H. BURK. 1996. Evriastrum densifolium ssp. sanctorum Report on 1995 Field Studies. Pre- pared for Psomas and Associates. MITCHELL, R. J. AND D. C. PATON. 1990. Effects of nectar volume and concentration on handling times for three species of honeyeater (Meliphagidae). Oecologia (Berlin) 83:238—246. MONTGOMERIE, R. D. 1984. Nectar extraction by hum- mingbirds: response to different floral characters. Oecologia (Berlin) 63:229—236. Munoz, A. A. 1991. Reproductive Biology of the Endan- gered Santa Ana River Woolly Star, Eriastrum den- sifolium ssp. sanctorum (Milliken) Mason (Polemon- iaceae). Master’s thesis, California State University, Fullerton, CA. Munz, P. A. 1974. A Flora of Southern California. Uni- versity of California Press, Berkeley, California. Pyke, G. H. 1978. Optimal nectar production in hum- mingbirds: testing the marginal value theorem. Amer- ican Zoologist 18:739-—752. . 1981. Optimal nectar production in a humming- bird pollinated plant. Theoretical Population Biology 20:326-343. STEINBERG, M., D. DORSETT, C. SHAH, C. EUGENE JONES, AND J. BuRK. 1998. Pupal case of Raphiomidas acton Coquillett (Diptera: Mydidae) and behavior of newly- emerged adult. Pan-Pacific Entomologist 74(3):178— 180. STONE, D. R. 1995. Pollinator Effectiveness and Assem- blages in Three Populations of Eriastrum densifolium (Benth.) Mason (Polemoniaceae). Master’s thesis, California State University, Fullerton, CA. U.S. FISH AND WILDLIFE SERVICE. 1989. Biological Opin- ion and Conference on the Corps of Engineers Santa Ana River Project in San Bernardino, Riverside, and Orange Counties, California. (Number 1-6-88-F-6). YorK, R. 1987. California’s Most Endangered Plants. pps. 109-120. in Elias, T. Editor. Conservation and Man- agement of Rare and Endangered Plants. California Native Plant Society, Sacramento. ZEMBAL, R. AND K. J. KRAMER. 1985. The status of the Santa Ana River Woolly-Star. Fremontia 13:19—20. MapRONO, Vol. 48, No. 4, pp. 272-285, 2001 DEMOGRAPHY AND POPULATION BIOLOGY OF A RARE TARPLANT, BLEPHARIZONIA PLUMOSA, (ASTERACEAE) A CALIFORNIA SUMMER ANNUAL FORB STEVEN D. GREGORY, ERIN K. ESPELAND, AND TINA M. CARLSEN Environmental Restoration Division, Lawrence Livermore National Laboratory, Livermore, CA 94551 ERIN K. BISSELL Associated Western Universities, Salt Lake City, UT 84124 ABSTRACT Blepharizonia plumosa (Kellogg) E. Greene, also known as the big tarplant, is a rare, summer-flowering, annual tarplant found in California grasslands. Although rare throughout its range, B. plumosa is numerous at the Lawrence Livermore National Laboratory (LLNL) experimental test facility, Site 300. We conducted a common garden experiment, a reciprocal transplant study, and a laboratory germination study to com- pare B. plumosa to its more common relative, B. laxa Greene (which is rare at Site 300), and to gather data on the basic biology of B. plumosa in preparation for a possible mitigation project. Three populations of B. plumosa were compared to a single population of B. laxa. Little population differentiation was found in B. plumosa in terms of transplant success. Blepharizonia plumosa expresses much more extreme dimorphism with respect to seed germination: ray seeds have low germination percentages (<4%) com- pared to disc seeds (50-78%, P < 0.017). Blepharizonia laxa did not show any significant differences between ray (18%) and disc (17%) seed germination (P > 0.05). Blepharizonia plumosa produced three times more ray seeds than disc seeds (ratio: 3.2), while B. /axa produced little over half as many ray seeds as disc seeds (ratio: 0.65). Blepharizonia plumosa plants grown from ray seeds have a lower biomass accumulation (<5 g per plant compared to 28 g per plant for B. axa, P < 0.017) and none survived to flower production. Among ray-derived B. /axa plants, 70% survived to flower production. This difference in ray seed production and performance points to a possible basis for ecological differences between the two species. Until the mechanisms controlling dormancy are more clearly understood, only disc seeds should be used in the creation of new populations of B. plumosa. Annual plants within California grassland com- munities are often divided into functional groups based on their flowering period, e.g., early grasses, late grasses, early forbs, and summer forbs (Pen- dleton et al. 1983). Due to the changes in climate throughout the growing season, plants within these groups are confronted with a wide range of envi- ronmental conditions, e.g., changing soil moisture content and photoperiod. Compared to the early spring flora, relatively little ecological information appears to be available on the annual summer forbs. These summer forbs, or late forbs, include species of the tarplant subtribe Madiinae (Compositae), e.g., Hemizonia, Madia, and Calycadenia, as well as Blepharizonia. Information on species in the summer flora is important for conservation and habitat restoration activities. For example, 110 late- flowering (July—September) taxa listed by the Cal- ifornia Native Plant Society (CNPS) are of conser- vation concern. Of these, eight taxa are state or fed- erally listed as threatened or endangered (Skinner and Pavlik 1994). The flowering period of summer forbs, such as Hemizonia (Chiariello 1989) and Blepharizonia (personal observation), starts in late July or early August and can extend into November. Because of their late-flowering period, summer forbs spend much of their lifetime in the harsh California sum- mer drought. They are able to survive drought con- ditions by developing deep root systems and are able to escape the higher temperatures that occur at the soil surface due to their height (Morse 1988). In water usage studies conducted by Gulman et al. (1983), it was found that a spring-flowering plant (Clarkia) and a summer-flowering plant (Hemizon- ia) differ in their access to stored water in the post- precipitation season. Hemizonia has the ability to root into deeper non-nutritive substrate for water, while Clarkia cannot. In addition to being able to access additional water, summer flowering plants postpone water stress through hydrophilic, pecti- naceous, extracellular polysaccharides concentrated in leaf tissue (Morse 1988) and are able to flower at lower water potentials (Chiariello 1989). Schoen- herr (1992) points out that by postponing the flow- ering stage into the summer months, summer forbs have less competition for pollinators. Later-blooming plants accumulate an order of magnitude more biomass than _ earlier-blooming plants (Mooney et al. 1986) and may also have some additional colonization advantages. First, nat- urally disturbed sites, such as gopher mounds, can be ideal for colonization. The later a species flow- ers, the larger the number of mounds available. Second, since late-flowering species tend to be taller than spring-flowering species, their seeds — 2001] _ have a wider dispersal radius. This allows them to ' colonize mounds and other disturbed areas farther from their immediate neighborhood (Hobbs and ' Mooney 1985). Blepharizonia plumosa (Kellogg) E. Greene is an _ extremely rare late-flowering plant included on the 'CNPS List 1B (Skinner and Pavlik 1994, under the _ former moniker Blepharizonia plumosa subsp. plu- mosa'), which includes plants that are rare, threat- ened, or endangered. The CNPS R-E-D code (rar- ity-endangerment-distribution) for B. plumosa is 3- 3-3, which indicates that this plant is limited to one or several restricted populations, is endangered throughout its range, and is endemic to California. The CNPS also noted that possibly the only re- maining populations exist on private property in the _ hills near Livermore, California. Populations have been previously identified in Alameda, Contra Cos- ta, San Joaquin, Stanislaus, and Solano Counties (Skinner and Pavlik 1994). Preston (1996) noted that a population was discovered at Contra Loma Regional Park, south of Antioch in 1979, but that surveys conducted by the East Bay Regional Park District in 1991 were unable to relocate the species. In 1994, several more populations were discovered on private property southwest of Brentwood (CNDDB 1996). Another small population was found at Chaparral Springs near Mount Diablo (Preston 1996). Current status of these populations is unknown. Several populations of B. plumosa were identi- fied during a habitat survey in 1996 at Lawrence Livermore National Laboratory’s (LLNL) experi- mental test facility (Site 300). A few populations of the more common big tarplant, Blepharizonia laxa Greene, were also found. Blepharizonia laxa, although also endemic to California, exists in plen- tiful numbers and has a much larger range which extends farther south into the inner South Coast Ranges, including San Benito County (Hickman 1993). Site 300 is a high-explosive test site oper- ated by the University of California for the De- partment of Energy. The site is closed to the public and has had no agricultural activity since its estab- lishment in the 1950s. The botanical diversity of the site may be due to its lack of public access, lack of agricultural activity, and high burn frequen- cy. Large stands of the native perennial bunchgrass, Poa secunda J.S. Presl, occur in the northern por- tion of the site and are thought to be maintained by the annual late spring/early summer controlled burns conducted for wildfire control (Carlsen et al. 2000). The large-flowered fiddleneck, Amsinckia grandiflora A. Gray (Boraginaceae), which is on the state and federal endangered species lists 'The Blepharizonia genus has recently been revised (Baldwin et al. 2001): Blepharizonia plumosa was for- merly B. plumosa subsp. plumosa (Hickman 1993), while Blepharizonia laxa was formerly B. plumosa subsp. vis- cida (Hickman 1993). GREGORY ET AL.: DEMOGRAPHY AND POPULATION BIOLOGY 2a (Schoenherr 1992), and the diamond-petaled poppy (Eschscholzia rhombipetala E. Greene), which was until recently thought to be extinct (Hickman 1993), are both found at Site 300. While historical occurrences of many rare plants probably have been extirpated by agricultural activities and non- native plant colonization (Skinner and Pavlik 1994), these factors may be reduced at Site 300. While the rarer taxon, B. plumosa, is quite com- mon at Site 300, occurring most plentifully in areas that are routinely burned, B. /axa is uncommon at Site 300, but exists in both unburned and burned areas. The two species are known to co-occur at only two locations, one of which is_ routinely burned. That the two species appear to differ in their habitat requirements may indicate some eco- logical differences between them. Comparison of rare and common congeners can provide important information for rare plant management (Pantone et al. 1995) and can illuminate differences that affect comparative abundance (Byers 1998). For conservation and management purposes, a thorough understanding of the population dynamics and the feasibility of population restoration of B. plumosa are critical. Mitigation of a B. plumosa population may need to be performed at Site 300 in the future, but few reintroduced populations of endangered species appear to be self-sustaining (Pavlik 1994; Parsons and Zedler 1997; Pavlik and Espeland 1998). Allen (1994) stated that only four of the 45 reintroduction projects undertaken in Cal- ifornia during the past decade were successful when judged by survival and reproduction. In many instances, reintroduction failure was attributed to poor planning or lack of information about the spe- cies (Hall 1986; Pavlik 1994). Population restora- tion is often performed without regard to the level of microhabitat adaptation in source populations (Knapp and Rice 1994). Small-scale adaptation in source populations can prevent success in restora- tion attempts (Dyer and Rice 1997). However, in situ population differentiation is not necessarily negatively correlated to transplant success (Meagh- er et al. 1978); thus each of the three B. plumosa populations were compared for transplantability in our experiment. We initiated a study in November of 1996 to col- lect basic demographic and population biology data on B. plumosa. The study was needed to determine if population differentiation in B. plumosa occurs with respect to transplant success and also to ex- amine differences between B. plumosa and _ the more common big tarplant, B. /axa. The study in- cluded a common garden experiment followed by a field reciprocal transplant study to investigate field germination rates. In addition, a laboratory germination study was conducted to investigate the relationship between seed type, seed age, and ger- mination/dormancy. Because so little is known about population biology of B. plumosa, and be- cause ongoing activities at Site 300 could poten- 274 tially impact the populations there, these data could be used both to improve management practices and to prepare for a possible onsite mitigation project at Site 300. MATERIALS AND METHODS Study species. Both B. plumosa and B. laxa are dicots within the family Asteraceae, and members of the tribe Helenieae (Karis and Ryding 1994). They are both summer annual forbs, which germi- nate with the onset of the first substantial fall/winter rains and flower July—October. One of the common attributes to this family is heterocarpy, or the pro- duction of two or more single-seeded fruit types (achenes) on a single plant. Although the term ‘**seed’”? should be used when addressing germina- tion and dormancy, and fruit or achene should be used when addressing morphology, the term “*seed”’ will be used solely throughout this paper. In the case of Blepharizonia, the plants have di- morphic flowers within the same inflorescence. Seed dimorphism is very common in the plant kingdom, found in the families such as Asteraceae, Chenopodiaceae, Poaceae, Brassicaceae, and is of- ten associated with arid or disturbed environments (Harper 1977; van der Pijl 1972). As discussed by Bremer (1994), seed morphology is among the most important and useful features in Asteraceae classification at the generic and species levels. The two species can be distinguished from one another by seed morphology and leaf color (Hick- man 1993; Baldwin et al. 2001). The most distinc- tive characteristic of B. plumosa is the pappus on the disc seed that is 1.5—3 mm in length. The disc seed pappus is sometimes described as plumose (thus the specific epithet plumosa) and contrasts with the very minute pappus of the ray seeds (Fig. 1). The plants also have a pale green color, as their foliage is sparsely glandular below the inflores- cence. Older plants have many inflorescences on side branches. The disc and ray seeds of B. /axa appear much more similar to the naked eye and have a short pappus from O—1 mm in length. Blepharizonia laxa is much more glandular than B. plumosa giving the plant a more yellow-green color and a much stron- ger scent. Blepharizonia laxa also tend to be slight- ly taller than B. plumosa (personal observation). Older plants have inflorescences mostly terminal on slender wand-like, bracted peduncles (Hickman 1993). Study site. Corral Hollow is a valley of the Inner Coast Ranges that contains examples of native plants and animals characteristic of the Great Cen- tral Valley. This area has been used primarily for grazing sheep and cattle (Schoenherr 1992). Site 300 covers 2711 ha and is located in the Altamont Hills of the Diablo Range separating the Livermore and San Joaquin Valleys. The southern portion of Site 300 (T3S R4E sect. 29 SW %) extends into MADRONO | [Vol. 48 | Corral Hollow and consists of rugged north/south- | trending canyons of elevations ranging from 150 m | to over 500 m. It is primarily a grassland-dominated | site, with sparse populations of coastal sage scrub | and blue oak woodland in the southwest corner of the site. | Three populations of B. plumosa, designated | B834 Berm, B834 Drainage, and B850, and one population of B. laxa, designated Middle Canyon, | were used in this study. Figure 2 shows Site 300 | and the locations of the study populations. Table 1 | includes some habitat characteristics of all four — populations. i] Common Garden Experiment Intact inflorescences containing seeds were col- lected in October and November 1996 from the three populations of B. plumosa and the single pop- ulation of B. laxa. Following collection, they were segregated into disc and ray seeds. A large con- tainer of soil (approximately 60 liters) was also col- lected in November 1996 from each of the four population sites. No attempt was made to insure the soil was free from existing seeds. A fifth container was filled with potting soil. Each container was di- vided into four quadrants, each of which was then divided in half. Each quadrant was assigned a source population and its two sub-quadrats delin- eated disc and ray seeds. In late November 1996, 25 disc and 25 ray seeds from each population were planted into adjacent sub-quadrants in each con- tainer, as shown in Fig. 3. Germination was allowed to occur naturally following rainfall events, al- though all containers were later outfitted with a drip irrigation system to ensure even watering. Germi- nation and growth were closely monitored for ap- proximately 10.5 months, until October 1997. Oth- er species that germinated within the pots were re- moved. Any B. plumosa or B. laxa plants that emerged in locations not associated with the plant- ing arrangement were also removed. Seeds were collected from plants as they matured and biomass was collected after senescence. Data analysis. Percent germination for disc and ray seeds was calculated by population in each of the soil types. To determine if differential germi- nation was occurring, the average percent germi- nation for all seeds planted in native soil (n = 4) was compared to that of all seeds planted in non- native soil (n = 12). Because no significant rela- tionship was found between soil type (native vs. non-native) and the amount of germination (P > 0.95) or percent survivorship (P > 0.19), each pot was treated as a replicate for all further data anal- yses. Percent germination for seed type was cal- culated by dividing the final number of germinated seeds by the total number planted (n = 25). Then, the average percent germination was calculated by taking the mean across all replicates (n = 5). The percentage of plants surviving to flower production | 2001] Disc GREGORY ET AL.: DEMOGRAPHY AND POPULATION BIOLOGY Ray Common Tarplant, Blepharizonia laxa Disc Fic. 1. was calculated by dividing the number of flower- producing plants by the number of plants that ger- minated, then averaging across replicates (n = 5). Average seed production per plant is the total seed production per replicate divided by the number of plants of the parent seed type in that replicate that produced mature seeds. All five replicates were then averaged. Statistical analysis of seed produc- tion was performed on the ratio of ray to disc seeds Comparison of pappus on seeds of Blepharizonia plumosa and B. laxa. that each plant produced. Average biomass was cal- culated by summing the dry weights (in grams) of the plants collected in each replicate divided by the number of plants, then taking the mean across all replicates (n = 5). Statistical analyses were performed using the general linear model in SAS, version 6 (SAS 1990). All percentage data were arcsine trans- formed prior to analyses. Single degree of free- 216 Francisco Pacific — - Ocean | 0 5 10 15 20 Kilometers yokued aipPuN yoaued semoL 600 Fic. 2. dom f-tests (orthogonal contrasts) were per- formed to ascertain if differences occurred be- tween disc and ray seeds within each species and then if differences occurred between the two spe- cies within each seed type. Tukey’s separation of means was performed to determine if populations of B. plumosa were different from each other. Be- MADRONO Sacramento Legend ceeeee Ridge line Blepharizonia populations A B834 berm B. plumosa population * B850 B. plumosa population V Middle Canyon B. laxa population Scale : feet 0 1000 2000 — B834 drainage B. plumosa population Site 300 and locations of Blepharizonia plumosa populations. cause of the multiple tests performed on the same data sets, a corrected alpha value was used to determine significance. Using a Bonnferroni cor- rection, we divided our starting alpha value of 0.05 by three (the number of comparisons per- formed upon each dataset), so a test would have to have a P value of less than 0.017 to be signif- 2001] HABITAT CHARACTERISTICS OF THREE BLEPHARIZONIA PLUMOSA POPULATIONS AND ONE B. LAXA POPULATION AT SITE 300 (ADAPTED FROM PRESTON 1996). TABLE 1. Management Elevation Number of plants Slope (%) Soil type practices Aspect (m) eat a ls. Plant community Population B834 Berm not burned, berm 8—30 clay, Alo-Vaquero com- north la A exotic annual grassland, Avena sp., Gutierre 200 a a with low grass cover annually burned plex californica, Eriogonum angulosum, Bromus diandrus, Holocarpha obconica exotic annual grassland, Bromus hordeaceus, (B. plumosa) GREGORY ET AL.: DEMOGRAPHY AND POPULATION BIOLOGY eg | sandy to clay loam, Wis- 50-75 north =215 500-1500 B834 Drainage flat-Arburua-San Timo- teo complex rocky sand to clay loam, B. diandrus, Amsinckia intermedia, B. mad- (B. plumosa) ritensis ssp. rubens, Grindelia camporum. disturbed annual grassland, Avena sp., Bromus annual burned north 30—50 ~400 ~100 B850 Wisflat-Arburua-San Timoteo complex sandy to clay loam, Wis- madritensis ssp. rubens, Vulpia myuros (B. plumosa) not burned 50-75 east ~=400 exotic annual grassland, Avena spp., Bromus not determined Middle Canyon flat-Arburua-San Timo- teo complex diandrus, B. rubens, B. hordeaceus, Hor- (B. laxa) deum marinum, Silybum marianum, Marah fabaceus, Gutierre distans ia californica, Phacelia A icant. This correction decreases our chance of re- jecting the null hypothesis erroneously. Reciprocal Transplant Study Four 100-m? plots were established within each of the three B. plumosa populations. Disc and ray seeds from the common garden experiment were used in this study. As shown in Fig. 4, seeds were planted five across in four rows, each row corre- sponding to one of the four populations. Each plot was subdivided; one subdivision was assigned to disc seeds and the other to ray seeds. The seeds were spaced 11—12 cm apart within each row, each row was spaced six to seven cm apart, and each subdivision was spaced 11-12 cm apart. Seeds were planted in all four plots in each of the three B. plumosa populations. At the time of planting, seed placement was marked using colored plastic straws 10 cm in length. Upon germination, the seedlings were marked by placing a circle of string loosely around the base of the plant. The germi- nation rate, growth, and survival of all marked plants were monitored throughout the season. Data analysis. There was no difference in ger- mination between locations for any parent popula- tion (P = 0.544), so data from all locations were pooled for analysis. Percent germination was cal- culated for each plot by source population and seed type. The individual plot values were then averaged (n = 4). Finally, the mean for the three populations or transplant locations was found (n = 3). Statisti- cal analyses were performed in the same manner as those for the common garden study. Laboratory Germination Study Inflorescences were collected from the three B. plumosa populations and the B. /axa population in 1996, 1997, and 1998, and from the common gar- den experiment pots in 1997. All were collected in autumn (September—November) at the predispersal stage. The material was sorted into disc and ray seeds and stored at room temperature. Only seeds that appeared to be undamaged and fully mature were selected for use in the germination study. The seeds selected from the 1996 and 1998 collections were taken from native populations. However, due to the limited number of suitable seeds collected from the native populations in 1997, seeds from the common garden experiment were used for this year. For the B. plumosa populations, three replicates of 20 ray seeds and three of 20 disc seeds were used. Six replicates of 20 ray seeds and six of 20 disc seeds were used for the B. /axa population. Ger- mination was initiated on 30 Nov 1998 for seeds collected in 1996 and on | Dec 1998 for seeds col- lected in 1997 and 1998. Seeds were placed in plas- tic Petri plates on Whatman 80 filter paper moist- ened with deionized water. The plates were then sealed with parafilm to reduce water evaporation 278 Middle Canyon Common Tarplant Population B850 Rare Tarplant Population Fic. 3. and placed in a dark cooler kept at room tempera- ture. Germination data were collected for approxi- mately six weeks (43 days) following initial wetting of the plates. All plates were wet again at the end of Week 1 (Day 8) and Week 2 (Day 16). At the end of Week 4 (Day 36), 60 percent of the plates were rewet as necessary. Germinules were removed as they germinated to control for any possible post- germination effects. Data analysis. Percent germination was found by dividing the number of germinated seeds by the to- tal number of seeds in the plate. Because some of the seeds selected were later determined to be im- mature, the number of seeds per plate ranged from 13 to 20. The seeds that did not germinate were examined under a dissecting microscope at the end of the experiment and designated as immature if they lacked an embryo. The individual plate values were then averaged by seed type, population, and year (n = 3 for B. plumosa, n = 6 for B. laxa). Cumulative germination represents final percent germination values. Final germination percentages were arcsine transformed and analyzed for age and subspecies effects using the general linear model in SAS, version 6 (SAS 1990). Fruit age (0, 1, or 2 years) was plotted against final percent germination for ray and disc seeds in B. plumosa and B. laxa MADRONO Planting arrangement of tarplant seeds within each of five containers in the common garden experiment. B834 Berm Rare Tarplant Population B834 Drainage Rare Tarplant Population using a least-squares linear regression (SAS 1990). Because three analyses were run on the same da- taset (linear regression by species and by seed type plus a general linear model), our corrected alpha value was 0.05 divided by 3 to determine our sig- nificant P value (0.017). RESULTS Common Garden Experiment Table 2 presents the results and statistical anal- yses of the germination rates, survivorship, and biomass production for the common garden exper- iment. A dramatic difference was observed in per- cent germination between the disc seeds and ray seeds for B. plumosa. Germination of the B. plu- mosa disc seeds ranged from 50 to 78%, while ray seed germination was below 4%. No statistical dif- ference was observed between disc and ray ger- mination for B. laxa. Although B. plumosa disc seeds appeared to have a higher level of germina- tion than those of B. /axa disc seeds, and the in- verse was true for ray seeds, due to the small sam- ple size and high amount of variability within the B. laxa, these differences were not significant. A priori comparisons showed that there were no dif- ferences between B. plumosa populations for ray and disc germination. 2001] Disc Seeds Ray Seeds Pilot ©®0O9OO|O000 #1 SK Qggegges @Se80 0108000 Disc Seeds Ray Seeds @@e@ee0000 0 Plot ®®@QO0O0|\0000 #4 SSSR SES](SESESES 2008 0|\96 @8@@ @ B834 Berm Seeds B834 Drainage Seeds GREGORY ET AL.: DEMOGRAPHY AND POPULATION BIOLOGY 279 Disc Seeds Ray Seeds Plot #2 Disc Seeds Plot #3 © B850 Seeds @ Middle Canyon Seeds Fic. 4. Diagram of reciprocal transplant study within each of three tarplant field populations. When comparing the percentage of tarplants sur- viving to flower production in the common garden experiment, a survivorship difference was apparent between the disc- and ray-seed derived plants for the B. plumosa. No ray-derived B. plumosa plants survived to flowering. However, because of the large difference in sample size (n = 7 ray-seed de- rived plants compared to n = 231 disc-derived plants) statistically significant differences were not found. No significant difference in survivorship was found between disc- and ray-derived B. laxa plants. Also, no difference in survivorship was found between B. plumosa and B. laxa disc-derived plants. However, a significantly lower percentage of B. plumosa ray-derived plants survived to flower production when compared to B. laxa ray-derived plants. As already mentioned, none of the B. plu- mosa plants from ray seeds survived, while 70% of B. laxa plants from ray seeds survived to flower production (n = 9). A priori comparisons showed that there were no differences in survivorship be- tween B. plumosa populations. When comparing biomass production, no differ- ence per plant between disc- and ray-derived plants was found in either B. laxa or B. plumosa. The B. laxa plants, however, achieved much higher bio- mass (average 28 g) for ray-derived plants than B. plumosa ray-derived plants were able to accumulate (each plant less than 5 g). The average seed production per plant is shown in Figure 5. Statistical analyses found an interaction between parent morph and seed morph ratio for B. plumosa and B. laxa (P < 0.0005). The ratio of ray to disc seed production from disc-derived B. plu- mosa plants was 3.2, while the ratio from the disc- derived B. laxa plants was 0.65, meaning that more similar numbers of disc and ray seeds were pro- duced from the B. laxa. Again, no seeds were pro- duced by the ray-derived B. plumosa (none sur- vived to reproductive maturity). Ray-derived B. laxa produced many more disc seeds (average = 185) than ray seeds (average = 59). Several of the disc-derived B. laxa plants died prior to complete maturation of the seeds, which may account for the low numbers of seeds produced by these plants. Reciprocal Transplant Study The results and statistical analyses for germina- tion rates in the reciprocal transplant study are pre- sented in Table 2. As in the common garden ex- periment, a significantly higher amount of germi- nation was observed for disc seeds than for ray seeds for B. plumosa, but no difference was ob- served between disc and ray germination for B. | 280 MADRONO [Vol. 48 | TABLE 2. GERMINATION, SURVIVORSHIP, AND BIOMASS RESULTS OF THE RECIPROCAL TRANSPLANT AND COMMON GARDEN | EXPERIMENTS (AVERAGE + | SE). Differential shading between ray and disc categories indicates significant difference at P < 0.017 in the rare tarplant (B. plumosa). *.*» Different lowercase letters indicate significant differences among | B. plumosa populations for disc seeds at P < 0.017. “8.48 Different uppercase letters indicate significant differences among B. plumosa populations for ray seeds at P < 0.017. * ** Different symbol repetitions indicate significant differences between species for disc seeds at P < 0.017.* ** Different symbol repetitions indicate significant differences | between species for ray seeds at P < 0.017. Reciprocal Common garden transplant % % Average % Parent population germination survivorship biomass (g) germination B834 Berm B. plumosa Dis 33.4414.0° 2.28+0.54* Ray” 16+40.98"" Os02" 0+0" 334167" Ratio (Ray:Disc) 0.021 0 0 0.094 B834 Drainage B. plumosa Dis +098" 49.8415.27 3.0440.73" | 65,0042.897 Ray 98+08° O20 ATi Oe 17 +167" Ratio (Ray:Disc) 0.016 0) 1.549 0.026 B850 B. plumosa Dis 48.84+18.07 4.00+ 1.59% Ray 32415 0+0* 0.36 + 0.09%" O40" Ratio (Ray:Disc) 0.056 0 0.090 0) Middle Canyon B. laxa Disc 23.2+0.8° §8.2+14.7 7.144+2.04 16.704+6.01° Ray 7.2+2.65" 70+20 27.824+10.97" 18.30+7.26" Ratio (Ray:Disc) 0.310 1.203 3.896 1.096 laxa. Once again, although higher levels of germi- nation were apparent for B. laxa ray seeds when compared to B. plumosa ray seeds, and the inverse seemed to be true for disc seeds, these differences were not significant (P = 0.049 and 0.057, respec- tively). The level of germination of disc seeds from the B834 Berm population was significantly greater than was disc germination from the Drainage pop- ulation. Otherwise, no population differences were found in germination for B. plumosa. All of the plants in the B850 and B834 Drainage populations were lost during the annual spring con- trolled burns at Site 300. B834 Berm was _ not burned, but only five B. plumosa and three B. laxa plants survived to the flowering stage. All eight were disc-derived plants. Because of the limited sample size, we did not perform statistical analyses of the growth and survivorship data. Laboratory Germination Study Germination appears to be related to seed age (Fig. 6). A statistical interaction was found among seed age, seed morph, and parent species (P < 0.0001). Overall disc germination was greater than overall ray germination, and germination in B. plu- mosa tended to be higher than for B. laxa. Linear regression performed for each seed type of each species yielded a significant correlation between seed age and germination for disc seeds in both B. plumosa and B. laxa (Table 3). The regression for B. plumosa disc seeds shows a negative slope: ger- mination decreases slightly as seed age increases. Blepharizonia laxa disc seeds show the opposite trend where increasing seed age correlated to a small increase in germination. Blepharizonia plu- mosa ray seeds also show a slight increase in ger- mination with increasing seed age, while germina- 2001 | GREGORY ET AL.: DEMOGRAPHY AND POPULATION BIOLOGY 281 200 1 a ae te +— T ae ae | alia a {|+— - T i T | c oe Seki ron pice leis ool 5 Seeds O. Disc Beedle ork Ray parents (DS/RP) L Ray Seeds from Ray Parents (RS/RP) hes o 150 + ©. r I Por) Oo | 2 L re) 100 - hen Qa = o L ® i) . cd) 50 5) © o > [ i 1 = : t 2 | O aD oc (a) jam @) 1 ores 2 | ! B834 Berm B. plumosa B834 Drainage B. plumosa B 850 B. plumosa Middle Canyon B. laxa Seed Parent Populations Fic. 5. = | SE): tion for B. laxa ray seeds did not show any signif- icant changes with increasing age. DISCUSSION Germination rates and dormancy. The common garden, reciprocal transplant and laboratory ger- mination experiments showed extreme dimorphism in germination percentages of disc and ray seeds within B. plumosa, while little dimorphism was found for B. laxa. This dimorphism seems to be a common phenomenon within Asteraceae, e.g., in Heterotheca subaxillaris var. subaxillaris (Baskin and Baskin 1976), Heterotheca grandiflora Nutt. (Flint and Palmblad 1978), Holocarpha macraden- ia (DC.) E. Greene (Palmer 1982), Senecio jaco- baea (McEvoy 1984), Heterotheca latifolia (Vena- ble and Levin 1985), Hemizonia increscens (Keck) B.D. Tanowitz (Tanowitz et al. 1987), Heterosper- ma pinnatum (Venable et al. 1987), and Hedypnois rhagadioloides (Kigel 1992). In all of these studies, disc seeds germinated more rapidly and in higher final proportions than did ray seeds. However, to our knowledge, we are the first to report differences in germination from ray and disc seeds between two closely related species within Asteraceae. These different germination responses are fairly Average number of mature disc and ray seeds produced from each parent seed type plant (n = 5, error bars subtle, and the implications are not fully known. Although seed dimorphism is often associated with different germination responses (Esashi and Leo- pold 1968; Baskin and Baskin 1976; Flint and Palmblad 1978; McEvoy 1984), few have attempt- ed to determine how laboratory results correspond with germination characteristics in the field (Tan- owitz et al. 1987). It has been demonstrated that disc seeds exhibit either no dormancy or significantly shorter dorman- cy periods in comparison to ray seeds, which ex- hibit dormancy in all species mentioned above. Several causes of delayed germination for ray seeds have been examined. The most prominent effects on ray seed germination are observed when the pericarp and seed coat are damaged or removed (Palmer 1982; McEvoy 1984; Tanowitz et al. 1987). In these studies, no significant differences were observed in germination rates and final ger- mination proportions between disc seeds and ex- cised ray embryos. McEvoy (1984) suggested that the thick pericarp of the ray seeds acts to physically inhibit germination. However, Palmer (1982) noted that if left intact, the inner nucellar layer (a clear, membranous material) would inhibit germination, potentially acting “‘in a metabolically active way.”’ 282 MADRONO 100 (ma [ae i i] Deiat | Lie T =n i ea | T | [|] B834 Berm B. plumosa B834 Drainage B. plumosa B850 B. plumosa Middle Canyon B. laxa 5 80 ~~ © ‘= = i= Vd) O 60 Pe] ® a) i ® 0. 40 cd) = _ oe E 5 20 O Disc Ray Disc Ray Disc Ray Zero years One year Two years Seed Age Fic. 6. Blepharizonia plumosa, n = In all cases, the embryos of ray seeds are not dor- mant, but rather have dormancy imposed on them by maternal tissues (Palmer 1982; McEvoy 1984; Tanowitz et al. 1987). Multiple environmental factors have been found to influence germination in ray seeds to a greater extent than they influence germination in disc seeds. In several studies, germination of ray seeds was inhibited by darkness, while light availability had little or no effect on disc seed germination (Baskin and Baskin 1976; Flint and Palmblad 1978; Venable and Levin 1985; Venable et al. 1987; Kigel TABLE 3. RESULTS OF GERMINATION BY SEED AGE LINEAR ** Regression is significant at P < 0.017. Germination of disc and ray seeds in the laboratory germination study from three source years (n = 6 for B. laxa, error bars = 1 SE). 3 for 1992). Venable and Levin (1985) tested this re- sponse further and found that germination of ray seeds was negatively correlated with burial depth, and as such they are an important part of the seed bank. In seed bank studies, only ray seeds were found in the soil (Palmer 1982; Venable et al. 1987; Tanowitz et al. 1987). This light response raises questions about the representativeness of the ob- served germination of ray seeds in our laboratory germination study. Since the germinations were conducted in the dark, except when checked for germination, this could have affected germination REGRESSION OF LABORATORY GERMINATION EXPERIMENT. [Vol. 48 Species Morph Regression equation R? n F P B. plumosa disc % germination = —0.33 seed age + 0.779** 0.55 25 32. <0.0001 B. plumosa ray % germination = 0.017 seed age + 0.002** 0.29 25 La 7 0.0021 B. laxa disc % germination = 0.077 seed age — 0.099%** 0.42 16 13.3 0.0022 B. laxa ray % germination = (9.25 X 107'8) seed age + 0.03 0.06 16 0) ] 2001 | rates. We may have seen a higher percentage of germination from ray seeds had they been kept in a lighted environment. Temperature also has an influence on ray seed germination in some species. Baskin and Baskin (1976) found that ray seed after-ripening and sub- sequent germination in H. s. var. subaxillaris was inhibited at low temperatures and promoted at high temperatures. They interpreted this response in the context of natural environmental conditions for the winter annual, concluding that ray seeds over-win- ter for at least one season following dispersal and possibly more, while disc seeds germinate during the first autumn following dispersal. Since our store of seeds used for the laboratory germination study were kept at room temperature, this may also have affected ray seed germination. Seed production and survivorship. Blepharizonia laxa consistently produced more disc seeds than ray seeds. Blepharizonia plumosa derived from disc seeds produced more ray seeds than disc seeds, and ray-seed derived plants produced no seeds at all. As presented in Fig. 5, the majority of the mature seeds produced by B. plumosa are ray seeds, which, as discussed above, are less likely to germinate, and if they do germinate, are less likely to produce seeds. Although most of the seeds produced by the B. laxa originated from ray-derived plants, this was due to the death of several of the disc-derived B. laxa plants just prior to seed maturation. Had this mortality not occurred, the disc-derived B. laxa may also have outpaced disc-derived B. plumosa plants in seed production since the biomass of the disc-derived B. laxa plants was comparable to that of the ray-derived B. laxa plants. Also, the disc- derived B. laxa plants that died were very large and contained a large number of inflorescences. Al- though there does not appear to be a survivorship difference between the disc-derived plants of the two species, the survivorship of ray-derived B. plu- mosa plants was lower than that of B. laxa ray- derived plants. This difference in survivorship was mirrored by the differences in biomass between the two species. Interspecific differences and evolutionary con- sequences. The main differences between the two species appear to be related to the production of dimorphic seeds and the correlation of dimorphism to germination. Blepharizonia laxa produces more disc seeds than ray seeds, while this ratio is in- verted, and is more extreme, for B. plumosa. Ger- mination rates differed between disc and ray seeds in B. plumosa but did not differ for B. laxa. Ble- Pharizonia laxa disc seed germination increased with age, while B. plumosa disc seed germination decreased with age. Blepharizonia plumosa ray seed germination increased with age, while B. laxa ray seed germination remained the same regardless of age. The morphology of disc seeds is different be- GREGORY ET AL.: DEMOGRAPHY AND POPULATION BIOLOGY 283 tween the two species: disc seeds have only a small pappus in B. laxa, but have a large pappus in B. plumosa, but the ray seed morphology between the two species 1s more similar (small to nonexistent pappus). The function of seed dimorphism in the ecology of each species may be different. Pappus presence is usually associated with greater dispersal ability and hence differences in pappus presence may be related to differences in seed bank dynam- ics (Palmer 1982; Venable et al. 1987; Tanowitz et al. 1987) and bet-hedging strategies (Westoby 1981) between the two species. The evolutionary value of the poor ray seed per- formance of B. plumosa in the field is a puzzle. A question important to many species in Asteraceae is why would a species put so many resources into production of seeds that do not germinate? One ex- planation could be that ray seeds form the primary seed bank supply, buffering the population for fu- ture poor production years. Our study did not fully determine what factors affect ray seed dormancy. As discussed above, light and temperature may play an important role in ray seed germination, and this should be evaluated in future studies. Areas at Site 300 where B. plumosa occurs are subject to routine annual burning. Fire may aid the survival of B. plumosa by promoting the growth of native bunch grasses, which may provide a more favorable habitat. However, as we observed, fire also caused direct mortality of seedlings found in the path of a controlled burn. Populations that are routinely burned may therefore depend upon ray seeds, assuming they are a significant part of the seed bank. Reasons for the difference in the prev- alence of the two species at Site 300 still elude us. Blepharizonia laxa seed exhibits lower germination percentages than B. plumosa seed. However, B. laxa plants produce more seed for compensation. It is likely that fire plays a differential role in facili- tating germination between the two species. Population differentiation and mitigation. Little differentiation was found between populations of B. plumosa. The fact that some differences were seen in the reciprocal transplant experiment germination percentages indicates that some populations may respond better to mitigation than others. Additional experiments with larger sample sizes should be per- formed prior to any mitigation work. However, we do have a good foundation upon which future ex- periments relating to population relocation can be performed. The high disc seed germination rates observed for each parent population at each trans- plant site indicate that seeds from multiple, local populations could be used for this new population. Since the low ray seed germination rate was not mitigated by transplant site/soil type, and not much is known about factors affecting dormancy, ray seeds would not be used as a seed source for this population. At present, it is unknown whether suc- cessful long-term establishment of new populations 284 would be possible. This can only be determined through long-term monitoring and experimental manipulation of both new and existing populations. Future Studies Continued monitoring of populations of B. plu- mosa and B. laxa within controlled burn areas and comparing them to populations that are not rou- tinely burned will help elucidate the impacts of fire on these populations and the reasons for the differ- ent distributions of these two species at Site 300. Factors affecting ray seed dormancy in B. plumosa and the role of ray seeds in populations of this spe- cies deserve further examination in order to devel- op the most informed management strategy for this rare plant. ACKNOWLEDGEMENTS We would like to thank S. Eric Walter for his dedication and technical assistance, Kim Heyward for artwork, Jim Lane of Site 300 Management for funding support, John Ziagos and Valerie Dibley of the Environmental Resto- ration Division (ERD) at LLNL for funding support, and ERD for infrastructure support. We would also like to thank Drs. Bruce Baldwin and Robert Preston for their technical input and review of the manuscript. The sug- gestions of Dr. Kristina Schierenbeck and an anonymous reviewer improved the manuscript immensely. Work per- formed under the auspices of the U. S. Department of Energy by Lawrence Livermore National Laboratory un- der Contract W-7405-Eng-48. LITERATURE CITED ALLEN, W. H. 1994. Reintroduction of endangered plants. BioScience 44:65—68. BALDWIN, B. G., R. E. PRESTON, B. L. WessA, AND M. WETHERWAX. 2001. A biosystematic and phylogenetic assessment of sympatric taxa in Blepharizonia (Com- positae—Madiinae). Systematic Botany 26:184—194. BASKIN, J. M. AND C. C. BASKIN. 1976. Germination di- morphism in Heterotheca subaxillaris var. subaxillar- is. Bulletin of the Torrey Botanical Club 103:201— 206. BREMER, K. 1994. Asteraceae Cladistics and Classifica- tion. Timber Press, Inc., Portland, OR. Byers, D. L. 1998. Effect of cross proximity on progeny fitness in a rare and a common species of Eupatorium (Asteraceae). American Journal of Botany 85:644— 653. CARLSEN, T. M., J. W. MENKE AND B. M. PAVLIk. 2000. Reducing competitive suppression of a rare annual forb by restoring native California perennial grass- lands. Restoration Ecology. (in press). CHIARIELLO, N. R. 1989. Phenology of California grass- lands. Pp. 47-55 in L. EK Huenneke and H. A. Moo- ney (eds.), Grassland structure and function. Kluwer Academic Publishers, Dordrecht, Netherlands. CNDDB (California Natural Diversity Database). 1996. File information on Blepharizonia plumosa b. plu- mosa. State of California, Department of Fish and Game, Natural Heritage Program, Sacramento, CA. Dyer, A. R. AND K. J. Rice. 1997. Evidence of spatial genetic structure in a California bunchgrass popula- tion. Oecologia 112:333-339. MADRONO [Vol. 48 ESASHI, Y. AND C. A. LEOPOLD. 1968. Physical forces in dormancy and germination of Xanthium seeds. Plant Physiology 43:871-—876. FLINT, S. D. AND [. G. PALMBLAD. 1978. Germination di- morphism and development in the ruderal weed Het- erotheca grandiflora. Oecologia 36:33-43. GULMAN, S. L., N. R. CHIARIELLO, H. A. MOONEY AND C. C. Cuu. 1983. Phenology and resource use in three co-occurring grassland annuals. Oecologia 58:33—42. HALL, L. A. 1986. Transplantation of sensitive plants as mitigation for environmental impacts. Pp. 413-420 in T. S. Elias (ed.), Conservation and management of rare and endangered plants. California Native Plant Society, Sacramento, CA. HARPER, J. L. 1977. Population biology of plants. Aca- demic Press, San Diego, CA. HICKMAN, J. C. (ed.) 1993. The Jepson manual: higher plants of California, University of California Press, Berkeley, CA. Hosss, R. J. AND H. A. Mooney. 1985. Community and population dynamics of serpentine grassland annuals in relation to gopher disturbance. Oecologia 67:342— S51: Karis, P. O. AND O. RypbING. 1994. Tribe Helenieae. Pp. 521-556 in K. Bremer (ed.), Asteraceae Cladistics and Classification. Timber Press, Inc. Portland, OR. KIGEL, J. 1992. Diaspore heteromorphism and germination in populations of the ephemeral Hedypnois-rhagadi- oloides (L) Schmidt, E W. (Asteraceae) inhabiting a geographic range. Acta Oecologica 13:45—53. KNApP, E. E. AND K. J. Rice. 1994. Starting from seed: genetic issues in using native grasses for restoration. Restoration Ecology Notes 12:40—45. McEvoy, P. B. 1984. Dormancy and dispersal in dimor- phic achenes on tansy ragwort, Senecio jacobaea L. (Compositae). Oecologia 61:160—168. MEAGHER, T. R., J. ANTONOVICS AND R. PRIMACK. 1978. Experimental ecological genetics in Plantago. III. Genetic variation and demography in relation to sur- vival of Plantago cordata, a rare species. Biological Conservation 14:244—257. Mooney, H. A., R. J. HoBss, J. GORHAM AND K. WIL- LIAMS. 1986. Biomass accumulation and resource uti- lization in co-occurring grassland annuals. Oecologia 70:555-—558. Morse, S. 1988. Drought tolerance and reproduction in annual plants. Unpublished Ph.D. dissertation. Uni- versity of California, Berkeley, CA. PALMER, R. E. 1982. Ecological and evolutionary patterns in Holocarpha (Compositae, Madiinae). Unpublished Ph.D. dissertation. University of California, Davis, CA. PANTONE, D. J., B. M. PAVLIK AND R. B. KELLEY. 1995. The reproductive attributes of an endangered plant as compared to a weedy congener. Biological Conser- vation 71:305—311. Parsons, L. S. AND J. B. ZEDLER. 1997. Factors affecting reestablishment of an endangered annual plant at a California salt marsh. Ecological Applications 7:253-— 267. PAVLIK, B. M. 1994. Demographic monitoring and the re- covery of endangered plants. Pp. 322-350 in M. Bowles and C. Whalen (eds.), Restoration of endan- gered species. Blackwell Scientific, London. and E. K. Espeland. 1998. Demography of natural and reintroduced populations of Acanthomintha dut- tonii, an endangered serpentinite annual in northern California. Madrofio 45:31—39. 2001] PENDLETON, D. F, J. W. MENKE, W. A. WILLIAMS, AND R. G. WOODMANSEE. 1983. Annual grassland ecosystem model. Hilgardia 51:1—44. PRESTON, R. E. 1996. Status of Blepharizonia plumosa (Big Tarplant), its occurrence on LLNL Site 300, and potential project impacts on the species. Report to LLNL Environmental Impact Report (EIR) Update Project File, October 3, 1996. SAS. 1990. SAS/STAT user’s guide, Pp. 891—996. Version 6, 4th ed., vol. 2. SAS Institute Inc., Cary, NC. SCHOENHERR, A. A. 1992. Natural history of California. University of California Press, Berkeley, CA. SKINNER, M. AND B. M. PAVLIK. 1994. Inventory of rare and endangered vascular plants of California, 5th ed. California Native Plant Society, Sacramento, CA. GREGORY ET AL.: DEMOGRAPHY AND POPULATION BIOLOGY 285 TANOwITZ, B. D., P. EF SALOPEK, AND B. E. MAHALL. 1987. Differential germination of ray and disc achenes in Hemizonia increscens (Asteraceae). American Jour- nal of Botany 74:303-—312. VAN DER PUL, P. 1972. Principles of dispersal in higher plants. Springer-Verlag, Berlin, Germany. VENABLE, D. L., A. BURQUEZ, G. CORRAL, E. MORALES, AND E EspPINosA. 1987. The ecology of seed hetero- morphism in Heterosperma pinnatum in central Mex- ico. Ecology 68:65-—76. and D. A. Levin. 1985. Ecology of achene di- morphism in Heterotheca latifolia: 1. Achene struc- ture, germination, and dispersal. Journal of Ecology 73:133-145. WesToBy, M. 1981. How diversified seed behavior is se- lected. American Naturalist 118:882—885. MADRONO, Vol. 48, No. 4, pp. 286-292, 2001 RECRUITMENT OF FRAXINUS PENNSYLVANICA (OLEACEAE) IN EASTERN MONTANA WOODLANDS PETER LESICA Conservation Biology Research 929 Locust Missoula, MT 59802 ABSTRACT Fraxinus pennsylvanica Marsh. (Oleaceae) woodlands are an important habitat for conservation of biological diversity on the Northern Great Plains. Many F. pennsylvanica woodlands appear to lack tree regeneration; however, little is known about the mode of recruitment in these stands. I sampled species composition, seedling density and the extent of vegetative recruitment of F. pennsylvanica for 17 stands in east-central Montana. Stand age distributions indicate that recruitment during the past 50 years has been greatly reduced compared to the previous 50 years. Detrended Correspondence Analysis identified a strong gradient in species composition positively associated with canopy cover of tall shrubs and mean annual precipitation and negatively associated with exotic grasses. Seedling recruitment of F. pennsyl- vanica was more common in stands with high DCA scores. The frequency of multi-stem tress indicated that at least 30% of F. pennsylvanica trees arose vegetatively. Rejuvenating sparse F. pennsylvanica woodlands should include both vegetative and seedling tree recruitment. Burning and/or cutting old, diseased trees with sparse canopies could create more vigorous stands once the sprouts have matured. Shadier habitat of such stands may encourage seedling recruitment by reducing the vigor of sod grasses. Fraxinus pennsylvanica Marsh woodlands are found along small-order drainages and on moist, cool slopes throughout much of the eastern third of Montana as well as most of the Northern Great Plains. Although F. pennsylvanica woodlands com- prise only a small proportion of this prairie land- scape, their aesthetic, economic and biological val- ues are large compared to their aerial extent (Noble and Winokur 1984). Many species of plants and animals occur only in habitat provided by F. penn- sylvanica woodlands. For example, Faanes (1984) recorded 47 species of breeding birds in western North Dakota woodlands, 22 of which were neo- tropical migrants. Of the 81 species of birds ob- served in F. pennsylvanica woodlands by Rumble and Gobeille (1998), 65 require woodland habitat. Lesica (1989) lists several vascular plant species occurring in eastern Montana only or primarily in F. pennsylvanica woodlands. Unfortunately, evidence from throughout the Northern Great Plains suggests that tree recruitment has declined in many F. pennsylvanica woodlands (Boldt et al. 1978; Lesica 1989). Woodlands with dense, stratified canopies and undergrowth domi- nated by native shrubs, graminoids, and forbs are being replaced by open canopy communities with few tall shrubs in the understory and a ground layer dominated by introduced sod-forming grasses (Hansen and Hoffman 1988). Many native animals decline as woodlands become more open. Bird den- sity and diversity was lower in these open-canopy stands in northwestern South Dakota (Hodorff et al. 1988; Rumble and Gobeille 1998). Deer mice (Peromyscus maniculatus), white-footed mice (Peromyscus leucopus) and woodrats (Neotoma ci- nerea) occurred more commonly in closed-canopy stands, while no mammalian species was more common in the open stands (Hodorff et al. 1988). It is imperative to maintain existing closed can- opy F. pennsylvanica woodlands and restore open stands in order to maintain biological diversity of the Northern Great Plains. Management and resto- ration of these woodlands hinges on understanding tree recruitment. Fraxinus pennsylvanica reproduc- es from seed and by vegetative sprouting from the tree base (Hansen et al. 1984; Uresk and Boldt 1986; Lesica 1989; Sieg and Wright 1996). How- ever, the past and present importance of these two modes of recruitment in native woodlands is not known. In this study I measured current levels of vegetative and seedling recruitment in relation to current stand condition to help determine what en- vironments are conducive to recruitment. I also ob- tained age-class distributions from increment core data to gain insight into past recruitment regimes. Results are used to develop restoration and man- agement strategies. STUDY SITES I conducted my study at six sites in Custer, Daw- son, Fallon, McCone, Prairie and Wibaux counties in east-central Montana (Fig. 1). Elevations of the study sites range from 745 m at Fort Peck to 1005 m at Cedar Creek. Soils are derived from soft sand- stones, siltstones and claystones of the late Creta- ceous Fort Union Formation (Veseth and Montagne 1980). Climate of the study area is semi-arid and continental. Mean annual precipitation ranged from 29 cm at Fort Peck to 41 cm between Baker and Wibaux (USDA-SCS 1981). Mean January mini- mum and July maximum temperatures at Terry in the center of the study area were —18° and 31°C respectively (NOAA 1982). Mean annual precipi- tation for the study areas was derived from a map 2001] ns ecm os: Missouri River Fic. 1. Location of six study sites in east-central Mon- tana: (1) Rattlesnake Butte, (2) Cabin Creek, (3) Wibaux, (4) Bible Camp, (5) Fort Peck, (6) Cedar Creek. of precipitation isohyetal lines for Montana for the period 1941-1970 (USDA-SCS 1981). Upland vegetation of the study area is steppe dominated by Artemisia tridentata ssp. wyomingen- sis Beetle and Young and perennial grasses includ- ing Stipa comata Trin. and Rupr., Agropyron smithii Rydb. and Bouteloua gracilis (H. B. K.) Griffiths Woodland vegetation dominated by F. pennsylvan- ica occurs in narrow bottomlands along ephemeral drainageways dissecting the uplands. These ‘‘woody draws’”’ often occur at the head of drainage basins but may be found along channels farther downstream as well. SPECIES DESCRIPTION Fraxinus pennsylvanica is a small, dioecious, de- ciduous tree to 20 m tall and 50 cm in diameter in Montana. It is shallow rooted and common along streams and in floodplains and other bottomland habitats from Nova Scotia to Alberta south to Texas and Florida (Great Plains Flora Association 1986; Farrar 1995). Fraxinus pennsylvanica is reported to have a short-lived seed bank (Farrar 1995). Seed- lings grow equally well in sun or shade (Borger and Kozlowski 1972). Wood is hard and was extensive- ly cut for firewood, especially in the prairie states (Peattie 1953). Maximum age is thought to be ca. 100 years (Farrar 1995) Montana plants are some- times considered to be var. lanceolata or var. sub- integerrimus, but these varieties are now considered of little taxonomic value (Great Plains Flora As- sociation 1986). LESICA: RECRUITMENT IN FRAXINUS WOODLANDS 287 METHODS Field methods. 1 sampled 17 stands among the six study sites. Stands were stratified by county and selected randomly from a list of stands managed by the USDI Bureau of Land Management. Only stands more than 2 ha in extent with Fraxinus pennsylvanica canopy cover reported to be at least 40% were included in the original pool of stands. After thorough reconnaissance, I subjectively lo- cated a 50 X 20 m sample plot to represent each stand. I estimated total tree canopy cover with a spherical densiometer at 12.5 and 37.5 m along the center line of each plot. I also recorded ocular es- timates of canopy cover of all woody plants and all herbaceous plants with cover of 1% or more. Vas- cular plant nomenclature follows Great Plains Flora Association (1986). In each plot I recorded the total number of trees equal to or greater than 2 m high into five 10-cm size Classes by species. I tallied all trees in the sam- ple plot into size classes and noted the presence of multiple stems and trunk sprouts (shoots arising vegetatively from the tree base) for all trees. Den- sity of tree seedlings (stems <2 m tall and <10 yrs old not arising from an older plant) was estimated from four 20 m? or 50 m? circular subplots equi- distant along the macroplot center line. Larger sub- plots were used when seedling density was low. I obtained the age, diameter and height of one randomly chosen sample tree of each species in each size class in each quarter-section of each plot. I estimated the height of sample trees to the nearest 0.5 m with a 3-m gauging pole. Diameter was mea- sured to the nearest 1 cm with a tape. Age was obtained from increment cores taken at 0.5 m above ground level. The number of annual rings was counted using a 10—20X microscope with cross- dating to help assure accuracy (Stokes and Smiley 1996). This method could underestimate the true age if tree stems were less than 0.5 m for one or more years. Several sample trees had rotten centers, and age could not be determined. Diameter of mul- tiple-stem trees was calculated as the diameter of a single-stem tree of equivalent basal area. Data analysis. Stand-level basal area of tree spe- cies was estimated by assigning the midpoint of each diameter size class to all trees in that class. I used the coefficient of variation (cv, standard de- viation/mean) as a measure of how evenly distrib- uted tree ages were within stands. I ordinated common species and stands using Detrended Correspondence Analysis (DCA; Gauch 1982; Rasmus 2000) to elucidate environmental gradients important to recruitment. Canopy cover estimates of all species present in at least five stands were used as input (McCune and Mefford 1997). I used regression analysis to test the significance of associations between precipitation and vegeta- tion (DCA score) and the proportions of trees with 288 TABLE 1. QUENCY >259%) SPECIES IN 17 SAMPLE STANDS. MADRONO [Vol. 48 | FREQUENCY OF OCCURRENCE, MEAN PERCENT CANOPY COVER AND DCA Axis 1 SCORES FOR COMMON (FRE- | Frequency % cover DCA score Trees Acer negundo i 2 233 Fraxinus pennsylvanica 17 7 120 Shrubs Amelanchier alnifolia 7 ] 21 Prunus americana 6 226 Prunus virginiana ee) 16 243 Ribes setosum 12 | 20 Rosa woodsii 10 ] —18 Symphoricarpos occidentalis 14 12 26 Toxicodendron rydbergii 6 ] 138 Graminoids Agropyron repens 5 3 81 Bromus inermis 7 9 91 Bromus japonicus 6 8 24 Carex sprengelii 5 14 323 Poa pratensis 15 26 —14 Forbs Achillea millefolium 5 <1 —26 Arctium minus 6 l 269 Cystopteris fragilis 6 <] 288 Disporum trachycarpum 6 | 169 Galium aparine 11] | 289 Galium boreale 6 | 161 Hackelia deflexa 6 <1 162 Monarda fistulosa 5 <1 ay, Ranunculus abortivus 6 | 273 Smilacina stellata 10 pi 158 Smilax herbacea v <=] 120 Taraxacum officinale 8 4 12 Thalictrum spp. 10 a | 252 Viola canadensis 5 <] 69 trunk sprouts and multiple stems and between seed- ling density and precipitation, vegetation and cv of tree age. I used Analysis of Variance (ANOVA) to test the effect of the presence of multiple stems and trunk sprouts on tree age and size. Mixed models included site and an interaction term as factors. The main effect was tested against the interaction term when the latter was significant (P <= 0.10). RESULTS Stand descriptions. Fraxinus pennsylvanica woodlands had total tree canopy cover of 18% to 73% with a mean of 45%. Fraxinus pennsylvanica was the dominant tree in all stands. Basal area ranged from 5.6 m’/ha to 37.5 m?’/ha with a mean of 14.9 m’/ha, and mean canopy cover was 37% (SE = 4%). Acer negundo L. (boxelder) occurred in six stands with a mean basal area of 2.2 m?/ha and mean canopy cover of 7%. Ulmus americana L. (American elm) occurred in three of the four easternmost stands with a mean basal area of 5.7 m’/ha and a mean canopy cover of 22%. Fraxinus pennsylvanica trees were 2.0 to 14.5 m tall with a mean of 7.0 m (N = 204, SE = 0.2 m) and basal diameters of 2.5 to 65 cm with a mean of 23 cm (n = 227, SE = 1 cm). Ulmus americana occurred in three sample stands at the Wibaux and Bible Camp sites. Diameter of U. americana ranged from 7 cm to 68 cm with a mean of 25 cm (N = 20, SE = 3 cm). Height varied from 5 m to 14 m with a mean of 9 m (SE = 0.5 m). I sampled only 10 Acer negundo trees, all of which were at the Cabin Creek and Wibaux sites. Diameter ranged from 20 cm to 50 cm with a mean of 30 cm (n = 10, SE = 3 cm). Height ranged from 6 m to 13 m with a mean of 8 m (n = 8, SE = 1 m). Detrended Correspondence Analysis (DCA) identified one strong gradient; axis | accounted for 64% of the variation in species composition. Spe- cies with low scores for DCA axis 1 included in- troduced rhizomatous grasses, Bromus inermis Leyss. and Poa pratensis L. as well as the low shrubs, Symphoricarpos occidentalis Hook. and Rosa woodsii Lindl. The weedy forbs, Achillea mil- lefolium L. and Taraxacum offiicinale Weber also had low scores (Table 1). Species at the high end LESICA: RECRUITMENT IN FRAXINUS WOODLANDS 60 60 60 60 60 60 289 Wibaux N=26 70 80 90 100 170-4120 Fort Peck N=15 70 80 90 100 110 120 Bible Camp N=13 70 80 90 100 110 120 Rattlesnake N=34 70 80 90 100 110 120 Cabin Creek N=37 70 80 90 100 110 120 Cedar Creek N=45 70 80 90 100. TG, “+20 Age of trees 2001 | 6 4 2 @) a 6 610"~620-0:«(8SO (40 ~~ 50 e — 6 oO 4 - 2 2 0 E 10 20 30 40 #50 Ze 6 4 2 on O Pieolo:. 20 <0 40 50 AIG Oo 4 = 2 Tapee E 10 20 30 40 #2450 — 5 4 2 a6 o 10 20 30 40 50 eG Oo 4 any) isc E 10 20 30 40 #250 z Fic. 2. of DCA axis | include the tall shrubs Prunus vir- giniana L., P. americana Marsh and Amelanchier alnifolia Nutt. The highest score was assigned to Carex sprengelii Dewey, a native graminoid. Com- mon forest species, Galium aparine L., Cystopteris fragilis (L.) Bernh., Ranunculus abortivus L. and Thalictrum spp. and the mesic-loving exotic, Arc- tium minus Bernh., also had high scores (Table 1). Stands with high DCA scores had higher mean an- nual precipitation (R? = 0.26, P = 0.036). Canopy cover of tall Prunus spp. ranged from 0% to 65% and was negatively correlated with the canopy cov- er of introduced rhizomatous grasses which ranged from 0% to 86% (r = 0.58, P = 0.016). Higher DCA scores were associated with lower canopy cover of exotic species (R* = 0.80, P < 0.001), and Age distribution of Fraxinus pennsylvanica at six study sites in east-central Montana. a model with both precipitation and cover of ex- otics explained 86% of the variation in vegetation (P < 0.001). Recruitment. Fraxinus pennsylvanica, Ulmus americana and Acer negundo all reproduced both vegetatively and from seed in study plots. However, sample sizes were large only for F. pennsylvanica, so results and discussion will be limited to this spe- cles. Age distributions of F. pennsylvanica trees in- dicate that most stands were not even-aged, and recruitment was sporadic (Fig. 2). A disproportion- ately large number of F. pennsylvanica trees in sample stands were 60—75 years old. Over all sites, 39% of the F. pennsylvanica trees regenerated in 290 the 15-year period of 1926-1940. Only 27% (n = 173) of the F. pennsylvanica trees sampled were 50 years or younger, and only 3% were 100 years or older. An average of 33% of F. pennsylvanica trees in study stands had live sprouts at their base, and 30% had more than one bole, indicating that they arose as basal trunk sprouts. Sprouting trees were larger than trees without sprouts with a mean diameter of 24 cm (SE = 1 cm) compared to 20 cm (SE = 1 cm; F, 59, = 8.7, P = 0.003). However, the mean age of sprouting and non-sprouting trees did not differ (P = 0.296). Multi-stem trees averaged 64 years (SE = 3 yrs) compared to 58 years (SE = 2 yrs) for single-stem trees (F, 53 = 4.0, P = 0.05). The abundance of multi-stem and sprouting trees was not associated with vegetation (P > 0.62). Fraxinus pennsylvanica seedlings were uncom- mon in most stands. Mean number of seedlings per 100 m* was 10 (SE = 5). Nine of 17 stands (53%) had one or fewer seedlings per 100 m’, and seed- lings were entirely absent from sample plots in five stands. There was no association between mean an- nual precipitation and seedling density (R’* = 0.02, P = 0.58). However, higher seedling density was associated with more mesic vegetation (high DCA axis 1 scores; R? = 0.23, P = 0.05), and this as- sociation was due more to a positive correlation with canopy cover of Prunus spp. (r = 0.71) than a negative correlation with exotic rhizomatous grasses (r = —0.39). Stands in which tree ages were more evenly distributed (as measured by the coef- ficient of variation for age) tended to have more seedlings (R* = 0.22, P = 0.06). DISCUSSION The age structure of Fraxinus pennsylvanica stands indicates that recruitment in the second half of the past century has been low relative to the first half. Nearly 75% of F. pennsylvanica trees in sam- ple stands were 50 years or older. Fraxinus penn- sylvanica stems rarely persist more than 100 years on the Northern Great Plains (Butler and Goetz 1984; Girard 1985; Hansen et al. 1984; Hansen and Hoffman 1988; Farrar 1995; Sieg 1991) so density of F. pennsylvanica stems will decline by 50% over the next 50 years under current levels of recruit- ment. Sprouting from the base of the trunk is an im- portant mode of reproduction in F. pennsylvanica. A minimum of 30% of the trees in our sample stands arose as basal trunk sprouts, and vegetative reproduction occurred regardless of differences in associated vegetation. Although sprouting ability may be a function of tree size, younger trees were not more likely to sprout than older trees. More than 90% of F. pennsylvanica sprouted after being cut in an experiment in North Dakota (Uresk and Boldt 1986), although it is not known what pro- portion of these survived to become trees. MADRONO [Vol. 48 | The large pulse of F. pennsylvanica recruitment | that occurred between 1926 and 1940 may well | have been due to trunk sprouting. Many trees were | cut down by the large influx of homesteaders dur- | ing the years of 1900-1918 (Malone and Roeder | 1976). Starting around 1920 a decline in farm pric- | es and a series of severe droughts led to a rapid | reduction in the rural population (Malone and Roe- der 1976) and a concomitant lessening of wood- | cutting and livestock numbers (Lee and Williams | 1964), undoubtedly leading to lower woodcutting © and grazing pressure. Interviews with long-time | residents indicate that fire was not a significant fac- | tor (Lesica and Atthowe 2001). However, F. penn- | sylvanica may have responded directly to the © drought conditions of the mid-1920’s and 1930’s by dying back to the ground and resprouting (Albert- son and Weaver 1945). It is conceivable that whole stands rejuvenated during this time by trunk sprout- ing alone. However, it seems unlikely that F. penn- sylvanica woodlands can persist over the long term relying solely on vegetative reproduction to provide a tree canopy. Fraxinus pennsylvanica stands with higher den- sities of tree seedlings tended to have a greater ar- ray of tree ages, suggesting that steady recruitment from seed produced more uneven age structures than sporadic vegetative recruitment following drought, woodcutting or fire. However, F. pennsyl- vanica seedlings were uncommon in most stands and entirely absent from five of 17 sample plots. Lack of recruitment from seed during the past 50 years or longer has probably contributed to the skewed age distributions of most sample stands. Fraxinus pennsylvanica woodland vegetation oc- cupied a gradient from stands dominated by sun- loving, exotic grasses, grassland forbs and low shrubs to those with more closed understories dom- inated by Prunus virginiana, P. americana, Ame- lanchier alnifolia, Carex sprengelii and shade-lov- ing forbs. Several other researchers have observed the same gradient in Montana and the Dakotas (Butler and Goetz 1984; Girard et al. 1987; Hansen and Hoffman 1988; Hodorff et al. 1988; Lesica 1989; Vorhees and Uresk 1992). Recruitment of tree seedlings was higher beneath more closed un- derstories. The gradient defined by DCA axis 1 was associated with increasing mean annual precipita- tion, and decreasing canopy cover of exotic species. Abundance of exotics is frequently associated with level of disturbance, especially by livestock (Parker et al. 1993; Kotanen et al. 1998; Smith and Knapp 1999). Woodlands receiving more precipitation may be more resilient to grazing disturbance (Fah- nestock and Detling 1999). Assuming that canopy cover of exotic species is a surrogate for distur- bance, these results suggest that drought stress and overgrazing disturbance work in concert to favor stands with a more xeric, meadow-like understory, less conducive to tree seedling recruitment. The association between density of seedlings and 2001] more mesic, less disturbed stands characterized by a high canopy cover of Prunus spp. and lower abundance of rhizomatous grasses suggests that re- cruitment of F. pennsylvanica from seed may de- pend on facilitation by a tall shrub understory. Re- duced vigor of sod grasses associated with shading by a healthy shrub layer would likely mean more safe sites for tree seedlings (Albertson and Weaver 1945; Petranka and McPherson 1979; Van Auken and Bush 1997). In addition, F. pennsylvanica seedlings are very shade-tolerant (Borger and Ko- zlowski 1972), so interference from the P. virgini- ana canopy may be minimal. Tree seedlings may also experience a more humid environment and less herbivory under a shrub canopy (Callaway 1992, Werner and Harbeck 1982). The relationship be- tween P. virginiana and tree recruitment could be pivotal to succession from a meadow/low shrub community to F. pennsylvanica woodland. Facili- tation of F. pennsylvanica recruitment by Prunus species in these woodlands is plausible but requires experimental study for verification. Recruitment of trees both vegetatively and from seed has been important in F. pennsylvanica wood- land dynamics in the past. Rejuvenating open green ash woodlands is likely best accomplished by en- couraging tree recruitment through both of these modes. Uresk and Boldt (1986) rejuvenated west- ern North Dakota F. pennsylvanica woodlands by cutting decadent trees. Nearly all the trees sprouted after cutting. Prescribed fire may also be useful in encouraging vegetative recruitment of trees; low- intensity experimental burns induced sprouting of F. pennsylvanica in northwestern South Dakota (Sieg and Wright 1996). Burning and/or cutting old, diseased trees with sparse canopies could eventu- ally create stands with greater canopy leaf area once the sprouts have matured. Tall shrub densities may also increase in stands exposed to fire (Zim- merman 1981). Shadier habitat of stands rejuvenat- ed by cutting or fire should encourage seedling re- cruitment by reducing the vigor of rhizomatous sod grasses. However, fire could increase the abundance of exotic grasses in the short term by decreasing shade and increasing nutrient availability (Blair 1997). These proposed restoration methods need to be tested in controlled experiments with livestock grazing excluded. Cattle will use woodland habitat heavily during the growing season (Boldt et al. 1978), resulting in lower canopy cover of tall shrubs (Butler and Goetz 1984; Hansen and Hoff- man 1988). Tree sprouts grew taller, and survival of planted F. pennsylvanica, Prunus virginiana and P. americana was higher in ungrazed stands com- pared to stands grazed by cattle (Uresk and Boldt 1986). Further studies on the effects of tall shrubs and fire on seedling recruitment are needed. ACKNOWLEDGEMENTS Helen Atthowe helped with data collection. Kent Bow- en and Bruce Smith helped locate study sites. Louise de LESICA: RECRUITMENT IN FRAXINUS WOODLANDS 29] Montigny, Larry Rau and Hal Vosen provided logistical support. Ray Bannister allowed me to sample woodlands on his ranch. The manuscript was improved by comments from two anonymous reviewers. Funding was provided by UDSI Bureau of Land Management. LITERATURE CITED ALBERTSON, EK W. AND J. E. WEAVER. 1945. Injury and death or recovery of trees in prairie climate. Ecolog- ical Monograph 15:395—433. BLAIR, J. M. 1997. Fire, N availability, and plant response in grasslands: a test of the transient maxima hypoth- esis. Ecology 78:2359—2368. BoLpT, C. E., D. W. URESK, AND K. E. SEVERSON. 1978. Riparian woodlands in jeopardy on the northern High Plains. Pages 184-189 in R. R. Johnson and J. FE McCormick (eds.), Proceedings of the national sym- posium on strategies for protection and management of floodplain wetlands and other riparian ecosystems. USDA Forest Service General Technical Report WO- 12. Washington D.C. BorGER, G. A. AND T. T. KozLowskI. 1972. Effects of light intensity on early periderm and xylem development in Pinus resinosa, Fraxinus pennsylvanica, and Ro- binia pseudoacacia. Canadian Journal of Forest Re- search 2:190—197. BUTLER, J. AND H. Goetz. 1984. Influence of livestock on the composition and structure of green ash commu- nities in the Northern Great Plains. Pp. 44—49 in D. L. Noble and R. P. Winokur (eds.), Wooded draws: characteristics and values for the Northern Great Plains. South Dakota School of Mines and Technol- ogy, Rapid City. CALLAWAY, R. M. 1992. Effect of shrubs on recruitment of Quercus douglasii and Q. lobata in California. Ecology 73:2118—2128. FAANES, C. A. 1983. Breeding birds of wooded draws in western North Dakota. Prairie Naturalist 15:173—187. FAHNESTOCK, J. T. AND J. K. DETLING. 1999. Plant re- sponses to defoliation and resource supplementation in the Pryor Mountains. Journal of Range Manage- ment 52:263—270. FARRAR, J. L. 1995. Trees in Canada. Fitzhenry & White- side, Markham, Ontario Canada. 502 pp. Gaucu, H. G. 1982. Multivariate analysis in community ecology. Cambridge University Press, Cambridge, GIRARD, M. M. 1985. Native woodland ecology and hab- itat type classification of southwestern North Dakota. Ph.D. dissertation, South Dakota State University, Brookings. , H. GOETZ, AND A. J. BJUGSTAD. 1987. Factors influencing woodlands of southwestern North Dakota. Prairie Naturalist 19:189—198. Great Plains Flora Association. 1986. Flora of the Great Plains. University Press of Kansas, Lawrence. HANSEN, P. L. AND G. R. HOFFMAN. 1988. The vegetation of the Grand River/Cedar River, Sioux, and Ashland Districts of the Custer National Forest: a habitat type classification. USDA Forest Service General Techni- cal Report RM-157, Fort Collins. , AND A. J. BJUGSTAD. 1984. The vegeta- tion of Theodore Roosevelt National Park, North Da- kota: a habitat type classification. USDA Forest Ser- vice General Technical Report RM-113. Fort Collins, CO. Hoporrr, R. A., C. H. SIEG AND R. L. LINDER. 1988. Wild- PAS BE life response to stand structure of deciduous wood- lands. Journal of Wildlife Management 52:667—673. KOTANEN, P. M., J. BERGELSON AND D. L. HAZLETT. 1998. Habitats of native and exotic plants in Colorado shortgrass steppe: a comparative approach. Canadian Journal of Botany 76:664-—672. LEE, B. AND D. WILLIAMS. 1964. Last grass frontier. The South Dakota stock grower heritage. Black Hills Pub- lishers, Sturgis, SD. LesicA, P. 1989. The vegetation and condition of upland hardwood forests in eastern Montana. Proceedings of the Montana Academy of Sciences 49:45—62. , AND H. E. ATTHOWE. 2001. Tree recruitment, dis- ease and insect pests of green ash woodlands in east- central Montana. Unpublished report, USDI Bureau of Land Management, Miles City, MT. MALONE, M. P. AND R. B. ROEDER. 1976. Montana a his- tory of two centuries. University of Washington Press, Seattle. McCune B. AND M. J. MEFFoRD. 1997. PC-ORD. Multi- variate analysis of ecological data, version 3.0. MjM Software Design, Gleneden Beach, OR. NATIONAL OCEANIC AND ATMOSPHERIC ASSOCIATION. 1982. Monthly normals of temperature, precipitation and heating and cooling degree days. Montana, 1951— 1980. National Climate Center, Ashville, NC, USA. Nose, D. L. AND R. PB. WINOKUR (eds.). 1984. Wooded draws: characteristics and values for the Northern Great Plains. Great Plains Agricultural Council Pub- lication No. 11, South Dakota School of Mines and Technology, Rapid City. PARKER, I. M., S. K. MERTENS AND D. W. SCHEMSKE. 1993. Distribution of seven native and two exotic plants in a tallgrass prairie in southeastern Wisconsin: the im- portance of human disturbance. American Midland Naturalist 130:43-—55. PEATTIE, D. C. 1953. A natural history of western trees. Houghton Mifflin Co., Boston MA, USA. PETRANKA, J. W. AND J. K. MCPHERSON. 1979. The role of Rhus copallina in the dynamics of the forest-prairie ecotone in north-central Oklahoma. Ecology 60:956— 963: Rasmus, E. 2000. Can we trust gradients extracted by de- MADRONO [Vol. 48 trended correspondence analysis? Journal of Vegeta- tion Science 11:565—572. RUMBLE, M. A. AND J. E. GOBEILLE. 1998. Bird commu- nity relationships to succession in green ash (Fraxi- nus pennsylvanica) woodlands. American Midland Naturalist 140:372-381. S1eG, C. H. 1991. Ecology of bur oak woodlands in the foothills of the Black Hills, South Dakota. Ph.D. dis- sertation, Texas Tech University, Lubbock. , AND H. A. WRIGHT. 1996. The role of prescribed fire in regenerating Quercus macrocarpa and associ- ated woody plants in stringer woodlands in the Black Hills, South Dakota. International Journal of Wild- lands Fire 6:21—29. SmiTH, M. D. AND A. K. Knapp. 1999. Exotic plant species in a cy,-dominated grassland: invasibility, disturbance and community structure. Oecologia 120:605—612. STOKES, M. A. AND T. L. SMILEY. 1996. An introduction to tree-ring dating. University of Arizona Press, Tuc- son. Uresk, D. W. AND C. E. BoLpbT. 1986. Effects of cultural treatments on regeneration of native woodlands on the northern High Plains. Prairie Naturalist 18:193-— 202. USDA-SCS. 1981. Average annual precipitation. Mon- tana. USDA Soil Conservation Service, Bozeman, MT. VAN AUKEN, W. AND J. K. BusH. 1997. Growth of Pro- sopis glandulosa in response to changes in above- ground and belowground interference. Ecology 78: 1222-1238. VESETH, R. AND C. MONTAGNE. 1980. Geologic parent ma- terials of Montana soils. Montana Agricultural Ex- periment Station Bulletin 721, Bozeman, MT. VOORHEES, M. E. AND D. W. URESK. 1992. Relating soil chemistry and plant relationships in wooded draws of the Northern Great Plains. Great Basin Naturalist 52: 35-40. WERNER, P. A. AND A. L. HARBECK 1982. The pattern of tree seedling establishment relative to staghorn sumac cover in Michigan old fields. American Midland Nat- uralist 108:124—132. ZIMMERMAN, G. M. 1981. Effects of fire upon selected plant communities in the Little Missouri Badlands. M.S. Thesis, North Dakota State University, Fargo. Mapbrono, Vol. 48, No. 4, pp. 293-297, 2001 HARMONIA GUGGOLZIORUM (COMPOSITAE-MADIINAE), A NEW TARWEED FROM ULTRAMAFICS OF SOUTHERN MENDOCINO COUNTY, CALIFORNIA BRUCE G. BALDWIN Jepson Herbarium and Department of Integrative Biology, 1001 Valley Life Sciences Building #2465, University of California, Berkeley, CA 94720-2465 ABSTRACT Harmonia guggolziorum is a new tarweed from ultramafic (serpentine) soils of southern Mendocino County, California. Unlike other species of Harmonia, H. guggolziorum combines the following morpho- logical characteristics: primary stems usually longer than branches of the subumbelliform capitulescences, leaves unevenly distributed but not densely congested, heads erect in bud and fruit, phyllaries irregularly hirsute, disc florets bisexual, ray cypselae weakly arcuate, ray pappi present, and disc pappi of linear, fimbriate scales 0.6—0.8 mm long. Based on molecular phylogenetic data, I suggest that H. guggolziorum is the only known representative of a lineage that predates diversification of the other serpentine endemic species of Harmonia (1.e., H. doris-nilesiae, H. hallii, and H. stebbinsii). The apparent phylogenetic relationships and geographic location of H. guggolziorum lead me to hypothesize that Harmonia origi- nated in the southern North Coast Ranges and has undergone more extensive diversification on ultramafics than previously suspected for the genus or any other lineage in Madiinae. California’s exceptionally rich serpentine flora is especially well represented in the northwestern part of the state (Kruckeberg 1984, Harrison et al. 2000), where botanical exploration has continued to reveal previously unknown ultramafic endemics. In the tarweed genus Harmonia B. G. Baldwin [=Madia Molina sensu Keck (1959) pro parte (i.e., the yellow-anthered, pappose annuals, with 2n = 9 II); see Baldwin (1999)], two ultramafic endemics (H. doris-nilesiae and H. stebbinsii) have been de- scribed from the North Coast Ranges of California since 1980 (Nelson and Nelson 1980, 1985). Here- in, I describe yet another species of Harmonia from the North Coast Ranges. Harmonia guggolziorum B. G. Baldwin, sp. nov. (Fig. 1)—TYPE: USA, California, Mendocino Co., on serpentine on the north side of Feliz Creek Road (County Road 109), 2.1 miles west of Hopland (T13N, R1I2W, S23, NWI1/4 of NE1/4), ca. 150—200 plants associated with Pla- tystemon californicus, Gilia capitata, and Cryp- tantha clevelandii, 164 m, 30 April 2000, Jack and Betty Guggolz 1635 (holotype, JEPS; iso- type, CAS). Ab species ceteris Harmoniae characteribus com- binatis differt caulibus primariis plerumque ramis subumbelliformium capitulescentiarum longioribus; foliis distributis impariter, non dense congestis, plerumque proximalibus in caulibus primariis et ad basibus capitulescentiarum; capitulis plerumque er- ectis ante, per, et post anthesin; phyllariis irregular- iter et saepe sparsim hirsutis cum pilis prope mar- gines mollibus saepe implicitis; flosculis discorum bisexualibus; cypselis radiorum leniter arcuatis; squamis papporum radiorum fimbriatis, ca. 0.5 mm longis; squamis papporum discorum linearibus fim- briatis ca. 0.6—0.8 mm longis. Annual herbs. Stems erect, branched mostly in distal half, slender, mostly reddish-purple, to 3 dm high, sparsely to densely hirsute proximally, dense- ly stipitate-glandular distally, the glands dark-pur- plish (or yellowish). Leaves opposite proximally, alternate distally, sessile, mostly cauline, unevenly distributed, mostly proximal on primary stems and at bases of capitulescence branches (otherwise sparse or absent), ascending or usually widely spreading, often with reflexed apices; blades linear to filiform, 5-50 mm long (mostly 20—25 mm long on primary stems), 1—3 mm wide, entire or sparsely and shallowly toothed, slightly revolute, hirsute, eglandular or (mostly in capitulescence) stipitate- glandular (especially near apices), the glands dark- purplish (or yellowish). Capitulescences subumbel- liform, branches often 5—7 cm long (max. 15 cm long) and overtopping the nearly sessile head of the primary stem. Peduncles 2—12 mm long, stipitate- glandular, the glands dark-purplish (or yellowish). Heads usually erect in bud, anthesis, and fruit. /n- volucres obovoid, ca. 3—4 mm diam. (4—5 mm diam. in pressed specimens). Phyllaries (3—) 5 (—6) (1 per ray floret), uniseriate, herbaceous, linear, 4— 5 mm long, each completely enveloping a ray ova- ry, the free apices purplish, erect or spreading, flat or involute, <1/5 the length of enfolded basal por- tion of phyllaries; abaxial faces irregularly and of- ten sparsely hirsute with broadly arching or some- what appressed hairs, often with soft, matted hairs near margins, ciliate, irregularly stipitate-glandular, the glands dark-purplish (or yellowish). Ray florets 294 MADRONO [Vol. 48 | t =o r Ls 4 a Y x oS Gh Y,, , %. Sibe if * ANE AY : aa ei - 4 q \. + ar a 4 ji " \g Nut, SS | Ae Ob NEE 7 Se’ yi Y) Se ee 1 ae 7 apoctepetaeltgbeles <2 01 om EAR R No a, 2mm Fic. 1. Harmonia guggolziorum. (A) habit; (B) head; (C) phyllary, ray floret, palea, and disc floret (right to left); (D) adaxial view of ray floret and associated phyllary; (E) disc floret; (F) palea; (G) ray cypsela and pappus; (H) disc cypsela and pappus. 2001 | 10 changes Putative serpentine- endemic ancestor Fic. 2. BALDWIN: HARMONIA GUGGOLZIORUM 225 H. nutans H. guggolziorum H. hallii ; Serpentine clade H. stebbinsii H. doris-nilesiae Relationships in Harmonia based on 18S/26S nuclear ribosomal DNA sequences of the external and internal transcribed spacer regions (Baldwin, unpublished). The tree is rooted with sequences from the other diploid members of the ‘“‘Madia”’ lineage (Baldwin 1996). (3—) 5 (-6), pistillate, corollas bright yellow, tubes ca. 1.5 mm long, sparsely hirtellous, laminae broad- ly overlapping in head, flabelliform, 4-5 mm long, 5—7 mm wide, 3-lobed to ca. half length, glabrous. Disc florets 8-13, bisexual, corollas bright yellow, 2—3.5 mm long, tubes much shorter than the nar- rowly funnelform throats, lobes 5, glabrous abaxi- ally, densely bristly adaxially. Anthers yellow. Style branches acuminate, hispidulous. Receptacles flat, glabrous. Paleae not persistent, ca. 8, in one pe- ripheral series, linear, 4-5 mm long, herbaceous near apices or throughout, sometimes chartaceous proximally (especially along margins), flat or with margins partially enveloping a disc ovary, sparsely hirsute and densely ciliate near apices, sparsely stipitate-glandular near apices, the glands dark-pur- plish (or yellowish), the margins of adjacent bracts free or weakly fused proximally. Ray cypselae black, slightly laterally compressed, abaxially rounded, adaxially angled, clavate, weakly arcuate, ca. 3—3.5 mm long, glabrous, beakless. Ray pappi of ca. 10—12 stramineous, linear, fimbriate scales ca. 0.5 mm long. Disc cypselae black, +terete to clavate, straight or weakly arcuate, ca. 3—3.5 mm long, with antrorse hairs. Disc pappi of ca. 9-11, stramineous or purplish, linear, +flat (not crisped), fimbriate scales, 0.6—0.8 mm long. Chromosome number 2n = 9 II [reported here from B. G. Bald- win 1140 (JEPS)]. Paratype. USA, California, Mendocino County, on serpentine hillside at junction of Feliz Creek Road (County Road 109) and County Road 110, west of Hopland, 8 May 2001, B. G. Baldwin 1140 (JEPS). KEY TO SPECIES OF HARMONIA . Heads usually reflexed in bud and fruit; ray pappi 0; disc pappus elements lance-attenuate, fimbrillate, 2—3.7 mm long H. nutans (Greene) B. G. Baldwin Heads usually erect in bud and fruit; ray pappi present (often rudimentary); disc pappus elements subulate, linear, oblong, or quadrate, fimbriate or pulmose, 0.2—3.5 mm long. . Leaves + evenly distributed along stems; ray cypselae gibbous (bowed out abaxially), distinctly beaked (beaks <1 mm long); disc florets functionally staminate eee eee ce ee eee eee H. doris-nilesiae (T. W. Nelson & J. P. Nelson) B. G. Baldwin Leaves unevenly distributed, mostly restricted to proximal stems and bases of subumbelliform capitulescences; ray cypselae weakly arcuate, beakless; at least some disc florets bisexual. Phyllaries pilose near margins; disc pappus elements subulate, 1.2—3.5 mm long, plumose H. stebbinsii (T. W. Nelson & J. P. Nelson) B. G. Baldwin Phyllaries with inconspicuous, soft, often matted hairs near margins; disc pappus elements linear, oblong, or quadrate, 0.2—0.8 mm long, fimbriate. Primary stems usually shorter than branches of the subumbelliform capitulescences; distal leaves of primary stem densely congested; disc pappus elements oblong or quadrate, 0.2—0.5 mm long.................... H. hallii (D. D. Keck) B. G. Baldwin Primary stems usually longer than branches of the subumbelliform capitulescences; distal leaves of primary stem not densely congested; disc pappus elements linear, 0.6—0.8 mm long .... AH. guggolziorum B. G. Baldwin i | Relationships. Based on phylogenetic analyses of 18S/26S nuclear ribosomal DNA sequences of the external and internal transcribed spacers (Baldwin, unpublished), H. guggolziorum represents a basally divergent lineage in a monophyletic group com- prising all serpentine endemics in Harmonia (Fig. 2). Origin of the lineage represented by H. guggolz- iorum apparently predates divergence of H. doris- nilesiae, H. hallii, and H. stebbinsii from a common ancestor; H. guggolziorum is the sister group of the lineage corresponding to H. doris-nilesiae, H. hall- ii, and H. stebbinsii. Support for the hypothesis that H. guggolziorum represents an ancient, divergent lineage in Harmonia rather than a recent product of hybridization comes from unique character-states at eight rDNA nucleotide sites in H. guggolziorum 296 and from four rDNA mutations shared by the other three serpentine species of Harmonia but not by H. guggolziorum, H. nutans, or any other diploid spe- cies of the “‘Madia”’ lineage (1.e., diploid species of Anisocarpus Nutt., Carlquistia B. G. Baldwin, Jensia B. G. Baldwin, Kyhosia B. G. Baldwin, or Madia Molina; see Baldwin 1996, 1999). Biogeographic and evolutionary history of Har- monia. Discovery of H. guggolziorum has allowed for refined perspectives on the history of edaphic endemism and overall pattern of diversification in Harmonia. Based on phylogenetic analyses of rDNA sequence data (Fig. 2, Baldwin unpublished), the four species of ultramafic endemics in Harmon- ia (H. doris-nilesiae, H. guggolziorum, H. hallii, and H. stebbinsii) represent a well-supported mono- phyletic group that is sister to H. nutans, an endem- ic of volcanic-ash exposures in Napa and Sonoma counties, California. In light of the phylogenetic data, I propose a simple hypothesis to explain pat- terns of edaphic endemism in Harmonia: diver- gence of the ultramafic and volcanic-ash lineages from a common ancestor preadapted (preapted) to ‘“‘*harsh”’ edaphic conditions, followed by extensive diversification on serpentines in the ultramafic lin- eage, that is, descent of H. doris-nilesiae, H. gug- golziorum, H. hallii, and H. stebbinsii from a com- mon, ultramafic-endemic ancestor. Lack of diver- sity in the volcanic-ash lineage may be attributable in part to the limited geographic distribution of vol- canic exposures in northwestern California com- pared to the wide distribution of serpentine expo- sures in the region (Kruckeberg 1984; Fox et al. 1985). Phylogeographic considerations lead me to sug- gest a general history for Harmonia of wide dis- persal and allopatric diversification among edaphic “islands”? (see Raven 1964; Kruckeberg 1991). Harmonia guggolziorum and H. nutans, represent- ed by two lineages that diverge in succession at the base of the Harmonia rDNA tree (Fig. 2), and H. hallii are allopatric or parapatric taxa that are en- demic or largely restricted to the southern North Coast Ranges (H. nutans extends south into the northern San Francisco Bay area). The two species of the northern North Coast Ranges and southern Klamath Ranges (H. doris-nilesiae and H. stebbins- ii) are apically nested in the rDNA tree among the southern North Coast Range lineages and therefore are suggested to be products of dispersal from the south. Based on the rDNA tree topology, H. doris- nilesiae and H. stebbinsii are not sister species and may represent independent south-to-north dispersal events. Alternatively, the two species may have de- scended from the same northerly-dispersed ances- tor, with HA. hallii representing an instance of north- to-south dispersal. Harmonia doris-nilesiae and H. stebbinsii are highly divergent in morphology and molecular sequences and are to my knowledge the only taxa in Harmonia that provide an example of MADRONO [Vol. 48 | sympatry [V. Parker 757 (JEPS) and V. Parker 759 | (JEPS), at a site southwest of Dubakella Mountain, | Trinity Co., California]. In summary, members of the ultramafic clade of | Harmonia appear to be outstanding examples of | serpentine neoendemics, that is, groups that | evolved on ultramafics, rather than relicts or pa- | leoendemic taxa secondarily restricted to serpen- tines (see Stebbins 1942; Kruckeberg 1954, 1984; Stebbins and Major 1965; Raven and Axelrod 1978; Mayer and Soltis 1994a, b). Rarity. Discovery of H. guggolziorum along a paved, public road less than 3 miles from US High- way 101 at Hopland, near a University of Califor- nia field station, probably reflects extreme rarity of the species and insufficient access by botanists to serpentines in the vicinity. Smith and Wheeler (1990-1991) explored some nearby ultramafic sites in Mendocino County and did not report any spe- cies referable to Harmonia in their flora of Men- docino County. I did not find collections of H. gug- golziorum at CAS, CHSC, DAV, JEPS, PUA, ROPA, UC, or the herbarium of the University of California Hopland Research and Extension Center and am unaware of any collections of the species from anywhere other than the holotype and para- type localities. Harmonia guggolziorum is probably naturally rare, based on the paucity of documented localities for other serpentine harmonias; H. doris- nilesiae, H. hallii, and H. stebbinsii are all listed as rare or endangered (List 1B) by the California Na- tive Plant Society (2001). Exploration for new pop- ulations of A. guggolziorum on ultramafics of southern Mendocino County and adjacent counties is needed. I am pleased to name this species for Jack and Betty Guggolz of Cloverdale, California, who col- lected the first specimens of Harmonia guggolzior- um and who have contributed significantly to con- servation of California’s North Coast Range flora through years of dedicated effort. ACKNOWLEDGMENTS Special thanks to Jack and Betty Guggolz for sending me specimens of Harmonia guggolziorum and other in- teresting tarweeds and for sharing their enthusiasm and knowledge about native plants of California’s North Coast Ranges. I also thank Lesley Randall for preparing the il- lustration of H. guggolziorum; John L. Strother and Alan R. Smith for assistance with the Latin diagnosis; JLS, Theodore Barkley, and Kenton L. Chambers for helpful comments on the manuscript; JLS, Susan J. Bainbridge, and Charles F Quibell for field assistance; Bridget L. Wes- sa for lab assistance; and the curators of the following herbaria for loans of specimens of Harmonia: HSC, NDG, and the Shasta-Trinity National Forest herbarium in Redd- ing, California. LITERATURE CITED BALDwin, B. G. 1996. Phylogenetics of the California tar- weeds and the Hawaiian silversword alliance (Madi- 2001 | inae; Heliantheae sensu lato). Pp. 377-391 in D. J. N. Hind and H. Beentje (eds.), Compositae: System- atics. Proceedings of the International Compositae Conference, Kew, 1994 D. J. N. Hind, (ed.), Vol. 1. Royal Botanic Gardens, Kew, UK. . 1999. New combinations and new genera in the North American tarweeds (Compositae-Madiinae). Novon 9:462—471. CALIFORNIA NATIVE PLANT SocteETy. 2001. Inventory of rare and endangered vascular plants of California, 6th ed. Rare Plant Scientific Advisory Committee, D. Ti- bor, Convening Editor. California Native Plant Soci- ety, Sacramento, CA. rox, K..F,. JR: R. J, Fueck, G. H. Curtis; AND C. E. MEYER. 1985. Implications of the northwestwardly younger age of the volcanic rocks of west-central California. Geological Society of America Bulletin 96:647-654. HARRISON, S., J. H. VIERS, AND J. E QUINN. 2000. Climatic and spatial patterns of diversity in the serpentine plants of California. Diversity and Distributions 6: 153-161. Keck, D. D. 1959. Madia. Pp. 1113-1117 in P. A. Munz, A California flora. University of California Press, Berkeley, CA. KRUCKEBERG, A. R. 1954. The ecology of serpentine soils. III. Plant species in relation to serpentine soils. Ecol- ogy 35:267—-274. . 1984. California serpentines: Flora, vegetation, BALDWIN: HARMONIA GUGGOLZIORUM 297 geology, soils, and management problems. University of California Press, Berkeley, CA. . 1991. An essay: Geoedaphics and island bioge- ography for vascular plants. Aliso 13:225—238. MAYER, M. S. AND P. S. Sottis. 1994a. The evolution of serpentine endemics: A chloroplast DNA phylogeny of the Streptanthus glandulosus complex (Crucifer- ae). Systematic Botany 19:557—574. AND . 1994b. The evolution of the Strep- tanthus glandulosus complex (Cruciferae): Genetic divergence and gene flow in serpentine endemics. American Journal of Botany 81:1288—1299. NELSON, T. W. AND J. P. NELSON. 1980. A new Madia (Compositae) from northwest California. Brittonia 32:323=325. AND . 1985. A new Madia of section An- isocarpus. (Compositae: Helantheae) from Trinity County, California [USA]. Brittonia 37:394—396. RAVEN, P. H. 1964. Catastrophic selection and edaphic en- demism. Evolution 18:336—338. AND D. I. AXELROD. 1978. Origin and relationships of the California flora. University of California Pub- lications in Botany 72:1—134. SMITH, G. L. AND C. R. WHEELER. 1990-1991. A flora of the vascular plants of Mendocino County, California. Wasmann Journal of Biology 48/49:1—387. STEBBINS, G. L. 1942. The genetic approach to problems of rare and endemic species. Madrono 6:241—258. AND J. MAJor. 1965. Endemism and speciation in the California flora. Ecological Monographs 35:1—35. MADRONO, Vol. 48, No. 4, pp. 298-300, 2001 A NEW SPECIES OF DIDYMODON (MUSCI) FROM CALIFORNIA RICHARD H. ZANDER Buffalo Museum of Science, 1020 Humboldt Parkway, Buffalo, NY 14211 USA ABSTRACT A new moss species, Didymodon eckeliae R. H. Zander, is described from San Diego County in southern California. It is distinguished from its closest relatives in Didymodon sect. Vineales mainly by scalloped and bistratose leaf margins. The Mediterranean climate of southern Califor- nia supports an assemblage of arid-adapted mosses, of which Didymodon is a major element (Harthill et al. 1979; Koch 1950). A member of the harsh- environment family Pottiaceae (Zander 1993), Di- dymodon in North America and Mexico (Zander 1981, 1994, 1998) is composed of a number of complexes that are uncommonly difficult to identify to species with certainty. When a distinctive new species, as is here described, is discovered, it is a matter of amazement and gratification. Didymodon eckeliae R. H. Zander, sp. nov. Type: USA, California, San Diego Co., 13 km NE of Lakeside, Barona [Rancheria] Indian Reser- vation, trunk of Quercus agrifolia, 1. L. Wiggins, April 9, 1954 (holotype, NY, segregated as ‘*sub- packet A’). Mixed with Grimmia_ pulvinata (Hedw.) Sm. & Sowerb. Plantae in parte distali atrovirentes. Folia caulina mucronata, longi-lanceolata, in parte distali cari- nata, 2-3 mm longa, late crenata, in parte foli dis- tali “4—%4 in margine bistratosa, cellulis eis laminae similibus praedita; costa brevi-excurrens e cellulis irregulariter subisodiametricis vel quadratis com- posita; cellulae basales foliares juxta costam sub- distinctae, brevi-rectangulares, 11-14 pm latae, 1— 3:1; parietes cellulares basales aeque incrassati vel tenues; sinus foliares crenulationum § subfragiles. Lamina in KOH rubra reagens. Plants growing in cushions, dark green above, tan below. Stems to 1.5 cm, branching often; round- ed-pentagonal in transverse section, hyalodermis absent, sclerodermis weakly developed, diameter of central cylinder cells 20-25 wm, central strand present, strong; sparsely radiculose; axillary hairs 4—5 cells in length, basal cell thicker-walled or brownish. Cauline leaves incurved, appressed, somewhat twisted about the stem when dry, patent to spreading-recurved when moist; long-lanceolate, adaxial surface keeled, 2-3 mm long; base not dif- ferentiated or short-rectangular, sheathing; margins weakly recurved in proximal 4—% of leaf, evenly and broadly crenate and bordered by 1—2 rows of bistratose cells similar to the laminal cells in distal 4-4 of leaf; apex long-acuminate; costa short-ex- current as a mucro of quadrate or irregular nearly isodiametric cells, adaxial cells quadrate distal to leaf base, in 4 rows, abaxial cells quadrate distal to leaf base; transverse section semicircular, adaxial epidermis present, adaxial stereid band absent, guide cells 6 (4+2) in 2 layers, hydroid strand ab- sent, abaxial stereid band present, lunate in cross section, abaxial epidermis present, weakly differ- entiated; basal cells weakly differentiated at leaf base near the costa, short-rectangular, 11-14 um wide, 1—3:1, walls of basal cells evenly thickened to thin-walled; distal laminal cells quadrate-hex- agonal, essentially homogeneous, 7—9 wm wide, 1: 1, abaxial to adaxial wall width ratio 1:1, lamina 13-15 pm thick medially, thickness ratio of mul- tistratose to unistratose portions of leaf 2:1, papillae multiplex, poorly defined, as thick, irregular caps over the lumens, cell walls evenly thickened, con- vex on both sides of lamina. Specialized asexual propagation: leaf somewhat fragile at sinuses of crenulations. Sexual condition: apparently dioi- cous, archegonia alone present, terminal on stem. Sporophyte unknown. KOH laminal color reaction red. The new species is named for Patricia M. Eckel in gratitude. This is the second new species of Didymodon Hedwig (Pottiaceae, Musci) to be discovered re- cently for California, USA (Zander 1999), though the present find is from taxonomically long-neglect- ed herbarium material. It joins the species D. nor- risii R. H. Zander and D. nevadensis R. H. Zander, from Nevada (Zander et al. 1995) as new western species of the genus. Two Asian species, D. anser- inocapitatus (X.-j. Li) R. H. Zander (Zander and Weber 1997) and D. tectorum (Mill. Hal.) K. Saito (Zander and Ochyra 2001) have also been discov- ered in the American West. Given the acute and persistent activity of bryologists in California and elsewhere in the American West, it may confidently be predicted that additional new and exotic species of Pottiaceae, if not Didymodon, will detected. The new species is reminiscent of D. sinuosus (Mitt.) Delogne of Europe in its broadly crenulate leaf margins (notches averaging about 8—10 cells apart), but that species has distinct teeth at the apex of at least the immature leaves, and the distal leaf margins are not bistratose or only rarely so in small 2001] ZANDER: A NEW SPECIES OF DIDYMODON O > Dat DIOL) O O O 0 a yay ox 02; Pree OD SOtSC OQ OU ; ie wae <> Us SOGO8a0 COVERSOS ORS [oe RAR y OQ) lta SURES QSONS CoH AOUSBSYLOSS One : DAS SG fintucseeastess. alUguaantessscy Go) C2 Cc) == C20 OO poe, a WAG Fic. 1. Didymodon eckeliae. |. Habit. 2. Perichaetium. 3—5. Cauline leaves. 6. Leaf apex. 7. Leaf base. 8. Leaf cross section at mid-leaf. 9. Stem cross section. Scale bars: A = 4 mm (fig. 1); B = 0.5 mm (figs. 3-6); C = 70 pm (figs. 6-9). patches. There are several moss species of the American Southwest and adjacent Mexico that have bistratose leaf margins that may be confused with the new species, but none have broadly crenulate (scalloped) leaf margins. It differs, additionally, from D. rigidulus var. subulatus (E. B. Bartram) R. H. Zander by the latter’s long-subulate apex, and smooth leaf cells. Didymodon australasiae (Hook. & Grev.) R. H. Zander has a much flattened costal section. Species of the pottiaceous genus Mironia R. H. Zander have strongly differentiated half- sheathing leaf bases. Rhexophyllum subnigrum (Mitt.) Hilp., likewise in the Pottiaceae, has deeply cleft leaf margins, but these are also dentate, and the distal portion of its leaves are bistratose in patches throughout. The new species is closely related to Didymodon vinealis (Brid.) R. H. Zander, sharing such distinc- tive characters as weakly differentiated leaf base, transverse section of the costa at leaf base rather 300 MADRONO concave, and the presence of a short, deep groove with the appearance of a long-elliptical window on the adaxial surface of the costa near the apex. This last distinctive feature is lacking in D. sinuosus, which, by its minutely crenulate leaf margins may be more closely related to Trichostomum tenuiros- tre (Hook. & Taylor) Lindb. It is similar to D. ni- cholsonii Culm. in many characters, but that also closely related species has broadly elliptical leaves with smooth margins and the costa is subpercurrent or percurrent. The common species D. vinealis is quite variable, and specimens with irregularly bis- tratose margins or unevenly notched leaves are not uncommon in California, but the combination, and regularity of the scalloping and evenness of the marginal band of bistratose cells is distinctive in the new species. The original collection at NY was labeled only “Trichostomopsis,’> a taxon presently recognized as Didymodon sect. Asteriscium (Mill. Hal.) R. H. Zander, probably because of the bistratose distal laminal margins and the lack of an adaxial costal stereid band. The new species, however, is imme- diately distinguished from species of that section by its weakly differentiated basal cells. The follow- ing key is based on that of Zander (1999). KEY TO DIDYMODON SECT. VINEALES MORPHOLOGI- CALLY SIMILAR TO D. ECKELIAE 1. Leaves short- to long-lanceolate or long-triangular, to 4.0 mm, margins recurved near base or up to proximal % of leaf, propagula rare. 2. Leaves unistratose or bistratose in very small patches marginally Didymodon vinealis 2. Leaves bistratose marginally or medially. 3. Leaves long-lanceolate, long-acuminate, mar- gins evenly and broadly crenulate above leaf base, bistratose in 1—2 rows ene Gece hae oN ee Didymodon eckeliae 3. Leaves long-ovate to broadly lanceolate, apex blunt to broadly acute, margins smooth, bis- tratose marginally in 1—several rows in distal leaf half or occasionally only in patches . . dt sence 8 Bavaro ak ei ee ENS 2, Me os Didymodon nicholsonii 1. Leaves deltoid to short-lanceolate or ovate, to 1.5 or rarely to 2.0 mm, margins recurved or revolute to near apex, propagula sometimes present. 4. Costal section showing adaxial epidermal cells thin-walled, remainder of costa thick-walled; costa blunt apically, costa wider at midleaf than below, with a bulging adaxial surface forming a long-elliptic unistratose pad of cells, guide cells in 2(—3) layers, leaf margins loosely revolute, gemmae absent or at least rare, tubers occasional on proximal rhizoids .... Didymodon nevadensis [Vol. 48 4. All cells of costal section about equally thick- ened; costa often with an apical conical cell or costa short-excurrent, costa gradually narrowing distally, adaxial surface usually nearly flat (but costa occasionally thickened and bulging adax- ially), not forming a wide pad of cells, guide cells usually in 1| layer, leaf margins narrowly to loosely recurved, small spherical gemmae often present in leaf axils, rhizoidal tubers absent. 5. Leaves ovate or ovate-lanceolate, 0.7—1.0 mm, base ovate or weakly differentiated, apex cucullate or weakly concave, margins weakly recurved, costa percurrent or very weakly excurrent from an obtuse or acute apex in 1-3 cells; lamina red in nature and with KOH, seldom green and KOH negative Ee ee eee eee Didymodon brachyphyllus 5. Leaves deltoid to deltoid-lanceolate, base squared, 1.0—1.5(—2.0) mm, apex flattened, or keeled, often somewhat reflexed, margins strongly recurved to revolute, costa excurrent from an obtuse apex as a several-celled blunt mucro; lamina green in nature and with KOH ee ee ee te he Didymodon tectorum LITERATURE CITED HARTHILL, M. P., D. M. LONG AND B. D. MISHLER. 1979. Preliminary list of Southern Californian mosses. Bryologist 82:260—267. Kocnu, L. E 1950. Mosses of California: an annotated list of species. Leaflets of Western Botany 6:1—40. ZANDER, R. H. 1981 [1982]. Didymodon (Pottiaceae) in Mexico and California: taxonomy and nomenclature of discontinuous and nondiscontinuous taxa. Cryp- tog., Bryol. Lichénol. 2:379—422. ZANDER, R. H. 1993. Genera of the Pottiaceae: Mosses of harsh environments. Bulletin of the Buffalo Society of Natural Science. 32:i—vi, 1-378. . 1994. Didymodon. Pp. 299-319, in A. J. Sharp, H. A. Crum and P. M. Eckel (eds.), Moss Flora of Mexico. Memoirs of the New York Botanical Garden, Vol. 69. 2 vols. . 1998. A phylogrammatic evolutionary analysis of the moss genus Didymodon in North America North of Mexico. Bulletin of the Buffalo Society of Natural Science: 36:81—115. . 1999. A new species of Didymodon (Bryopsida) in western North America and a regional key to the taxa. Bryologist 102:112—115. AND R. OcHyRA. 2001. Didymodon tectorum and D. brachyphyllus (Musci, Pottiaceae) in North Amer- ica. Bryologist 104:372—377. , L. R. STARK AND G. MARRS-SMITH. 1995. Didy- modon nevadensis, a new species for North America, with comments on phenology. Bryologist 98:590-— 595; AND W. WEBER. 1997. Didymodon anserinocapt- tatus (Musci, Pottiaceae) new to the New World. Bryologist 100:237—238. Mapbrono, Vol. 48, No. 4, p. 301, 2001 REVIEW Ecosystems of the World 16: Ecosystems of Dis- turbed Ground. Edited by Lawrence R. Walker. 1999. 868 pages. Elselvier Science Press, Amster- dam, The Netherlands. For many years ecologists have examined the concept of ecological disturbance in relation to nat- ural processes and community dynamics. Some have argued that much of what has been defined as disturbance is in fact an intrinsic part of natural systems. This book is a survey of contemporary topics relating to natural and anthropogenic distur- bance, ecological succession, and environmental is- sues associated with disturbance. There are thirty- three individually authored chapters covering a broad range of interests. A general reading of the text gives the reader not only a sense of the variety of disturbance mechanisms, but also the range of approaches used to study them. The diversity of topics addressed in this volume makes this collec- tion a valuable reference for researchers. The editor suggests that the contributions be di- vided into the following categories: natural distur- bance, anthropogenic disturbance, processes, and human response. After reading through the chapters it was difficult to delineate between those focusing on anthropogenic versus natural disturbance. Though some processes, such as volcanic and wind disturbance fall clearly under the heading natural disturbance, processes such as fire and erosion can be both natural and anthropogenic in origin. Inter- esting discussions on the categorization of distur- bance as various types including natural and an- thropogenic, endogenous and exogenous, and in- herent and foreign could be found in many of the chapters. Central issues such as the definition of the term disturbance in an ecological context are left up to each author resulting in an opportunity for a com- parison of divergent views. Many authors cited well-established definitions. The most common of which were based on the ‘‘any relatively discreet event in time’’ concept put forth in the seminal work by Pickett and White “The Ecology of Nat- ural Disturbance and Patch Dynamics.”’ Other au- thors suggested their own definitions. Both I. K. Bradbury and L. R. Walker suggest definitions that included any process resulting in the loss of bio- mass. Several authors also suggested definitions based on new growth and colonization opportuni- ties for individuals. An in-depth discussion of the use of the term disturbance can be found at the beginning of the chapter on disturbance in deserts authored by J. A. MacMahon. MacMahon suggests that the use of the “‘discrete event in time”’ defi- nition does not adequately address extended cli- matic disturbances. The ambiguity of the term “‘dis- turbance,”’ in relation to spatial and temporal scales, is also discussed by S. T. A. Pickett et al. in their chapter on patch dynamics. The topics presented are generally discussed in sufficient detail. However, the rationale on how top- ics were selected for inclusion is unclear. Several chapters are dedicated to specific mechanisms of disturbance including glaciation, erosion, wind, volcanism, and mining. Other ecologically signifi- cant disturbance mechanisms such as fire and flood did not receive treatment in their own chapters. Similarly, several ecosystem types were discussed individually including deserts, boreal forests, Med- iterranean shrublands, grasslands and savannas, North American wetlands, temperate forests, and urban areas, while others were omitted. A discus- sion of the criteria for selection in the introduction chapter would have improved an otherwise com- mendable work. This book is useful as a tool for exploring di- vergent points of view on basic ecological ques- tions. It has great value as a reference on a variety of subjects related to ecological disturbance, and would be useful as a supplementary text for stu- dents studying environmental issues and basic ecol- ogy. —WILLIAM H. RusseELL, USGS Western Ecological Re- search Center, Golden Gate Field Station. Fort Cronkhite Bldg 1063, Sausalito, CA 94965. Maprono, Vol. 48, No. 4, p. 302, 2001 ANNOUNCEMENTS NEw EDITOR Beginning with Volume 49, the new editor for Madrono will be: Dr. John Callaway Department of Environmental Science 2130 Fulton Street San Francisco, CA 94117-1080 email: callaway @usfca.edu All new submissions should be sent to Dr. Cal- laway. DONATIONS IN JUNE MCCASKILL’S MEMORY The University of California at Davis is building a new herbarium facility, and the plant identifica- tion laboratory will be named for June. An exhibit of photos of June as well as information on her contributions to botany at UC Davis will be on dis- play in the room. For more information, you can contact Ellen Dean at the UC Davis Herbarium Contributions in June McCaskill’s memory can be made to the UC Davis Foundation and sent to: Ellen Dean, Plant Biology, One Shields Ave., UC Davis, Davis, CA 95616. Mapbrono, Vol. 48, No. 4, p. 303, 2001 PRESIDENT’S REPORT FOR VOLUME 48 Thanks to the tireless efforts of outgoing Editor Kristina Schierenbeck, I have the pleasure of providing this report for Volume 48 less than six months after my report for Volume 47. Yes, Kristina achieved her goal of bringing Madrono back onto publication schedule, to the great ben- efit of the California Botanical Society and to west Amer- ican botany in general. Return to a normal publication schedule should have a positive effect on subscriptions, manuscript submissions, and article citations, in part by brightening the prospects for inclusion of Madrono in more on-line databases (in addition to BIOSIS). On behalf of the Society, I offer heartfelt thanks to Kristina for her hard work, dedication, and accomplishments. She now concludes her outstanding, extended service as Editor with Madrono in excellent health. Our new Editor, John Callaway, is busily handling new submissions to Madrofo (in coordination with Kristina Schierenbeck) and I am pleased to report that the transi- tion between editorships is proceeding smoothly. John is an Assistant Professor of Environmental Science at the University of San Francisco, where he researches plant and soil ecology, especially in wetland systems. His role aS a university instructor extends well outside San Fran- cisco; John regularly visits Budapest, where he teaches in a Masters program in environmental management. The Society is indeed fortunate to have another conscientious, well-respected, and active scientist serving as Editor of Madrono. Many thanks to John for accepting this vitally important responsibility for the Society. Attendance at our monthly meetings at UC Berkeley continues to rise, thanks in no small measure to First Vice- President Rodney Myatt, who once again organized an exciting lecture series for our program year. This fall, we have heard excellent presentations by Donald Strong, Richard Dodd, and Kevin Rice on research in the areas of plant ecology, evolution, genetics, and conservation bi- ology. We are looking forward to lectures on diverse bo- tanical topics by Bruce Mahall, Ron Amundson, Ingrid Parker, and John Callaway in winter and spring, 2002 (see www.calbotsoc.org). Thanks to Graduate Student Repre- sentative Kirsten Johannes, our monthly meetings have been more widely advertised than in past years and con- tinue to be capped by lively post-lecture receptions in the University and Jepson Herbaria. Second Vice-President Peter Fritsch is organizing the Society’s annual banquet for 2002, which will be held at the U. C. Berkeley campus on 16 February. Our distin- guished speaker for the occasion will be the world-re- nowned plant ecologist Chris Field, who will discuss his extensive research on the effects of global change on Cal- ifornia ecosystems. The annual banquet is always a major highlight of our program year and promises to be es- pecially enjoyable and educational in 2002, thanks to Pe- ter Fritsch’s efforts and Chris Field’s participation. The Council continues to pursue efforts to increase membership and visibility of the California Botanical So- ciety and our journal, Madrofio. A major challenge faced by the Council has been to find ways to promote timely membership renewals. Late renewals are much more wel- come than lost memberships but do present financial and logistical difficulties for the Society (e.g., mailings of back issues). We can no longer afford to send Madrono to un- renewed members beyond the first issue of a new (unpaid) volume and I urge all members to respond as quickly as possible to renewal notices. We have begun offering reg- ular members the option of renewing memberships for multiple years (at a discounted price) as a means of pro- moting continuity of memberships from year to year. Now that Madrono is back on publication schedule, we antici- pate fewer late renewals. Special thanks to Treasurer Roy Buck and Correspond- ing Secretary Sue Bainbridge for their dedicated hard work attending to the Society’s membership services and financial concerns. Roy’s and Sue’s continual efforts are crucial to the Society and are responsible in part for our growing membership base. I also thank webmasters Curtis Clark and John LaDuke for recent improvements to our web site (www.calbotsoc.org), which no doubt will play an increasingly important role for the Society. Recording Secretary Dean Kelch and Council members Anne Brad- ley, Jim Shevock, and Bian Tan deserve thanks for their important contributions to planning and guiding the activ- ities and future of the Society. I welcome nearly 100 new members to the California Botanical Society and thank all members for your contri- butions and continuing support. I ask all of you to help us recruit new members and to encourage your library to continue or initiate subscription to Madrono. On behalf of the Council, I wish all of you a great year for botany in 2002! —BRUCE G. BALDWIN 28 NOVEMBER 2001 MADRONO, Vol. 48, No. 4, p. 304, 2001 EDITOR’S REPORT FOR VOLUME 48 This report serves to inform members of the California Botanical Society the status of Madrono from manuscripts submitted to papers published. Since the previous editor’s report (see Madrono 47[4]) the journal received 62 manu- scripts for review, including Articles, Notes, and Note- worthy Collections; 44 of these have been accepted for publication. The average time from article submission to publication has remained stable at approximately six months. Very few manuscripts were rejected after review. Authors of Madrofo articles are generally quite respon- sive to reviewer and editorial suggestions. There are many individuals who contribute to the edi- torial process; Jon Keeley, who continues to serve as book review editor; Steve Timbrook, who continues to assemble the Index and Table of Contents; Dieter Wilken and Mar- griet Wetherwax, who edit the Noteworthy Collections; Jeannie Trizzino, my editorial assistant; the Department of Biological Sciences at California State University Chico, that provides the funds to support Jeannie; Karen Ridgway at Allen Press; and members of the CBS executive council who enthusiastically support Madrofo in every aspect. On behalf of the society, I thank the volunteer reviewers and the Board of Editors on whom we all depend to make the peer review process work for this valuable regional jour- nal. It has been a privilege to serve as editor of Madrono for four volumes, but I have passed the baton into the capable hands of Dr. John Callaway. Early indications give me the confidence that Dr. Callaway will keep a tight publication schedule and be devoted to the quality of the journal. I have appreciated the authors, reviewers, and readership who patiently overlooked my shortcomings and allowed me to grow into the position of editor. The edi- torial experience was a growth experience for me at many levels; I particularly enjoyed the opportunity to corre- spond with many colleagues with whom I might have not otherwise. Thank you, California Botanical Society, for the opportunity to serve the botanical community of west- ern North America. MApDRONO, Vol. 48, No. 4, p. 305, 2001 REVIEWERS OF MADRONO MANUSCRIPTS 2001 Robert Adams John Bailey Harvey Ballard Michael Barbour Carol Baskin John Battles Ellen Bauder Brian Beckage David Bogler Robert Christian Richard Dodd Andrew Dyer Andrew Eckert Gordon Fox Susan Galatowitch Nir L. Gil-Ad Carol Goodwillie Lisa Graumlich Craig Greene FE Thomas Griggs James Henrickson Diane Horton Steven Jessup Leigh Johnson Jon Keeley David Keil Peter Kotanen Joannes Knops David Kuntz Job Kuijt John LaDuke John Little Richard Lis Aaron Liston Andrea Lloyd Jeff Lovich John Maron Brent Mishler Pamela Muick V. Thomas Parker Robert Patterson Mark Porter James Pushnik Jon Rebman James Reveal Ronald Robberecht Darren Sandquist John O. Sawyer Rob Schhising Simon Shaumon James Shevock Rich Spellenberg Michael Steiber Scott Stephens Nathan Stephensen John Strother John Tappeiner Bart van der Kamp Susan Verhoek Alan Whittemore Dieter Wilken Barbara Wilson Paul Wilson Stephen Yool MApRONO, Vol. 48, No. 4, pp. 306-307, 2001 INDEX TO VOLUME 48 Classified entries: major subjects, key words, and results; botanical names (new names are in boldface); geographical areas; reviews, commentaries. Incidental references to taxa (including most lists and tables) are not indexed separately. Species appearing in Noteworthy Collections are indexed under name, family, and state or country. Authors and titles are listed alphabetically by author in the Table of Contents to the volume. Abies concolor, noteworthy collection from NV, 43. Alopecurus arundinaceus, noteworthy collection from AZ, O12. Amaranthaceae (see Amaranthus) Amaranthus blitum, noteworthy collection from WA, 213. Arizona: Noteworthy collections: Alopecurus arundina- ceus, Brachiaria platyphylla, 212; Enchylaena tomen- tosa, 61; Enneapogon cenchroides, Sclerochloa dura, Setariopsis auriculata, Tridens albescens, Urochloa panicoides, 212. Asteraceae: Blepharizonia plumosa, demography and pop- ulation biology, 272. New taxa: Harmonia guggolziorum, 293; Lasthenia sect. amphiachaenia, 205; L. sect. ornduffia, 38; L. californica subsp. bakeri, L. californica subsp. ma- crantha, L. ornduffii, 205. Atriplex robusta, new sp. from UT, 112. Ballota nigra subsp. foetida, noteworthy collection from WA, 213. Blepharizonia plumosa, demography and population bi- ology, 272. Brachiaria platyphylla, noteworthy collection from AZ, 212, Briza minor, noteworthy collection from WA, 213. Bryophytes: Bryophyte flora of William L. Finley Nation- al Wildlife Refuge, Willamette Valley, OR, 17; Didy- modon eckeliae, new species of moss from CA, 298; moss flora of San Francisco, CA, 1. California: Blepharizonia plumosa, demography and pop- ulation biology, 272; comparative flowering phenology in western Mojave Desert, 162; composition, invasibil- ity and diversity in coastal grasslands, 236; conifer tree distribution in southern CA, 177; Datura wrightii, geo- graphic variation in trichome phenotypes correlated with annual water deficit, 33; Delphinium gypsophilum, polyploidy and segregation analyses, 90; effects of litter and temperature on germination of native and exotic grasses in a coastal grassland, 230; Eriastrum densifol- ium subsp. sanctorum pollination biology, 265; Es- chscholzia californica sand hills ecotype, 25; fire, mar- itime chaparral community transition in absence of, 221; Leptosiphon androsaceus, morphometric analysis in central and south coast ranges, 62; moss flora of San Francisco, 1; plant communities, spring-fed, of East Bay Hills oak woodland, 98; pre-agricultural grass- lands, 253; Swallenia, taxonomic implications of ana- tomical studies, 152. New taxa: Didymodon eckeliae, 298; Harmonia gug- golziorum, 293; Lasthenia sect. ornduffia, 38; L. californica subsp. bakeri, L. californica subsp. ma- crantha, 208; Leptosiphon minimus, 74; L. rosa- ceus, 75; Saltugilia latimeri, 198. Noteworthy collections: Castilleja tenuis, 211; Chori- zanthe parryi var. fernandina, 78. CAM (crassulacean acid metabolism) (see Lewisia) Carex pendula, C. projecta, C. sylvatica, noteworthy col- lections from WA, 213. Castilleja tenuis, noteworthy collection from CA, 211. Chaparral, maritime, community transition in absence of fire, 221. Chenopodiaceae (see Atriplex and Enchylaena) Chorizanthe parryi var. fernandina, noteworthy collection from CA, 78. Chromosome count: Ipomopsis longiflora subsp. neo- mexicana, 116. Compositae (see Asteraceae) Conifer tree distribution in southern CA, 177. Corallorhiza maculata var. ozettensis, new var. from WA, 40. Crassula tillaea, noteworthy collection from WA, 213. Crassulaceae (see Crassula and Cyperus) Cupressaceae (see Juniperus) Cyperaceae (see Carex) Cyperus odoratus, noteworthy collection from WA, 214. Datura wrightii, geographic variation in trichome pheno- types correlated with annual water deficit, 33; notewor- thy collection from WA, 214. Delphinium gypsophilum, polyploidy and segregation analyses, 90. Didymodon eckeliae, new species of moss from CA, 298. Editor’s report, 304. Enchylaena tomentosa, noteworthy collection from AZ, 61. Enneapogon cenchroides, noteworthy collection from AZ, 212. Eragrostis curvula, noteworthy collection from WA, 214. Eriastrum densifolium subsp. sanctorum pollination biol- ogy, 265. Eschscholzia californica sand hills ecotype, 25. Fire, maritime chaparral community transition in absence Ol, 221. Fraxinus pennsylvanica, recruitment in eastern MT wood- lands, 286. Germination: Effects of litter and temperature on native and exotic grasses in a coastal CA grassland, 230. Geum urbanum, noteworthy collection from WA, 214. Gramineae (see Poaceae) Grasslands: Composition, invasibility and diversity in coastal California, 236; effects of litter and temperature on germination of native and exotic grasses in coastal CA, 230; pre-agricultural in central CA, 253. Harmonia guggolziorum, new sp. from ultramafics in CA; 293. Ipomopsis longiflora subsp. neomexicana, new subsp. from southwestern U.S. and adjacent Mexico, 116. 2001] Juniperus communis, RAPD fingerprint survey of varie- ties from western U.S., 172. Labiatae (see Lamiaceae) Lamiaceae (see Ballota) Lasthenia: Taxonomic changes, 205. New taxa: L. sect. amphiachaenia, 205; L. sect. orn- duffia, 38; L. californica subsp. bakeri, L. califor- nica subsp. macrantha, L. ornduffii, 205. Leptosiphon: L androsaceus, morphometric analysis in central and south coast ranges of CA, 62. New taxa: L. minimus, 74; L. rosaceus, 75. Lewisia cotyledon var. cotyledon, physiological and ana- tomical aspects of CAM-cycling, 131. Linanthus (see Leptosiphon) Major, Jack, tribute to, 215. McCaskill, June: Dedication of Volume 48 to, 308; do- nations in memory of, 302. México: Poa bajaensis, new sp. from Baja California, 123: Noteworthy collection: Setaria arizonica, 211. Mojave Desert, comparative flowering phenology in west- ernmost extension, 162. Montana (see Fraxinus) Mosses (see Bryophytes) Nassella pulchra in pre-agricultural central CA grass- lands, 253. Nevada (see Abies) Oleaceae (see Fraxinus) Orchidaceae (see Corallorhiza) Oregon: Bryophyte flora of William L. Finley National Wildlife Refuge, Willamette Valley, 17; Pinus contorta var. latifolia, relationship of successional status to seed- ling morphology, 138. New taxon: Lasthenia ornduffii, 205. Papaveraceae (see Eschscholzia) Parietaria judiaca, P. officinalis, noteworthy collections from WA, 214. Phoradendron, seedling establishment in four temperate species, 79. Pinaceae (see Abies and Pinus) Pinus: P. contorta var. latifolia, relationship of succes- sional status to seedling morphology, 138; P. ponder- osa, variability of seedling response to elevated CO, exposure, 51. Plant communities, spring-fed, of East Bay Hills, CA, oak woodland, 98. Poa bajaensis, new sp. from Baja California, Mexico, 123. Poaceae (see also Grasslands): Nassella pulchra in pre- agricultural central CA grasslands, 253; Poa bajaensis, new sp. from Baja California, Mexico, 123; Swallenia, taxonomic implications of anatomical studies, 152. Noteworthy collections: Alopecurus arundinaceus, Bra- chiaria platyphylla from AZ, 212; Briza minor from WA, 213; Enneapogon cenchroides from AZ, 212; Eragrostis curvula from WA, 214; Sclerochloa dura from AZ, 212; Setaria arizonica from Mexico, 211; Setariopsis auriculata, Tridens albescens, Urochloa panicoides from AZ, 212. INDEX TO VOLUME 48 307 Polemoniaceae: Eriastrum densifolium subsp. sanctorum pollination biology, 265; Leptosiphon androsaceus, morphometric analysis in central and south coast ranges of CA, 62. New taxa: Ipomopsis longiflora subsp. neomexicana, 116; Saltugilia latimeri, 198. Pollination biology of Eriastrum densifolium subsp. sanc- torum, 265. Polygonaceae (see Chorizanthe) Portulaceae (see Lewisia) Potentilla inclinata, noteworthy collection from WA, 214. President’s report, 303. Ranunculaceae (see Delphinium) RAPD fingerprinting of Juniperus, 172. Reviews: Ecosystems of the World 16: Ecosystems of Dis- turbed Ground ed. L. R. Walker, 301; Savannas, Bar- rens and Rock Outcrop Plant Communities of North America, eds. R. C. Anderson, J. S. Fralish and J. M. Baskin, 44; Terrestrial Ecoregions of North America: A Conservation Assessment by T. Rickets, et al., 45; Trees and Shrubs of California by J. D. Stuart and J. O. Sawyer, 128. Rosaceae (see Geum and Potentilla) Saltugilia latimeri, new sp. from CA, 198. Sclerochloa dura, noteworthy collection from AZ, 212. Scrophularia nodosa, noteworthy collection from WA, 214. Scrophulariaceae: Noteworthy collections: Castilleja ten- uis from CA, 211; Scrophularia nodosa from WA, 214; Verbascum pulverulentum from WA, 214. Serpentine (see ultramafic soils) Setaria arizonica, noteworthy collection from Mexico, ZA. Setariopsis auriculata, noteworthy collection from AZ, 22ND. Solanaceae (see Datura) Swallenia, taxonomic implications of anatomical studies, 152. Tridens albescens, noteworthy collection from AZ, 212. Ultramafic soils (see Harmonia) Urochloa panicoides noteworthy collection from AZ, 212. Urticaceae (see Parietaria) Utah (see Atriplex) Verbascum pulverulentum, noteworthy collection from WA, 214. Verbena officinalis noteworthy collection from WA, 213. Verbenaceae (see Verbena) Viscaceae (see Phoradendron) Washington (see Corallorhiza) Noteworthy collections: Amaranthus blitum, Ballota ni- gra subsp. foetida, Briza minor, Carex pendula, C. projecta, C. sylvatica, Crassula tillaea, Cyperus odoratus, Datura wrightii, Eragrostis curvula, Geum urbanum, Parietaria judiaca, P. officinalis, Potentilla inclinata, Scrophularia nodosa, Verbascum pulveru- lentum, Verbena officinalis, 213. MADRONO, Vol. 48, No. 4, p. 308, 2001 DEDICATION June McCaskill June McCaskill’s career as the Curator of the UC Davis Herbaria spanned more than 37 years. She joined the Botany Department in 1953 and retired in 1991. June was born in Pasadena, California, in 1930, where her parents operated and owned a small nursery. Her father hybridized and sold ca- mellias, naming one in honor of June; Camellia ja- ponica ‘June McCaskill.’ In 1951 June graduated from Mills College in Oakland with a degree in Botany. Soon after she began working for Professor Howard McMinn at Mills College, assisting in the small teaching her- barium and greenhouses. In 1953, she joined the herbarium at UC Davis. At Davis, June established herself as one of the best taxonomist in the state, identifying thousand of specimens sent or brought in by faculty, Farm Advi- sors, land managers, students, and the general public. She was particularly good at identifying agricultural weeds and poisonous plants sent to her from the School of Veterinary Medicine at UC Davis. One of June’s remarkable skills was her ability to identify plant fragments, including plant parts in hay bales, aquatic species left too long is sealed bags, and hair- balls found in the throat of dead animals. She assisted in many criminal court cases, identifying seeds and plant fragments important to prosecutors. During her tenure at Davis, June co-authored the Growers’ Weed Identification Handbook, which is the most widely used weed guide in California. She was a founding member of the Friend of the Davis Arboretum, and even after her retirement in 1991 help to start the Davis Herbaria Society. Over the years June won many awards for her accomplishments, including a two time recipient of the Outstanding Performance Award by the Botany Department at UC Davis, Award of Excellence pre- sented by the California Weed Science Society, and the Award of Distinction by UC Davis. The latter award is the highest honor of the College of Agri- cultural and Environmental Sciences award at UC Davis. In 1988 she was selected for the Women in Botany Oral History Project at the Bancroft Library of UC Berkeley. One of June’s passions was to travel around the world examining plants and nature. In addition to her many trips throughout California, she organized and led expeditions to the mid-Atlantic states, Tex- as, Hawaii, Canada, Costa Rica, Norway, Sweden, New Zealand and Greece. On her trip to Greece, June collected and mounted approximately 2,500 specimens of historical significance. To those many people who had the good fortune to know June, we will remember her humor, friend- ship, and unselfishness. June was a favorite among the many undergraduates she employed and the nu- merous graduate students she assisted. She had an uncanny way of correcting the mistakes of others without quenching their enthusiasm for using tax- onomic keys. In her spirit and the enthusiasm and friendship she extended to everyone she met, we dedicate Madrofio volume 48 to June McCaskill (1930—2001), botanist and friend. SUBSCRIPTIONS—MEMBERSHIP Membership in the California Botanical Society is open to individuals ($27 per year; family $30 per year; emeritus $17 per year; students $17 per year for a maximum of 7 years). Late fees may be assessed. Members of the Society receive MAprono free. Institutional subscriptions to MADRONO are available ($60). Membership is based on a calendar year only. Life memberships are $540. Applications for membership (including dues), orders for sub- scriptions, and renewal payments should be sent to the Treasurer. Requests and rates for back issues, changes of address, and undelivered copies of MADRONO should be sent to the Corresponding Secretary. INFORMATION FOR CONTRIBUTORS Manuscripts submitted for publication in MAproNo should be sent to the editor. It is preferred that all authors be members of the California Botanical Society. Manuscripts by authors having outstanding page charges will not be sent for review. Manuscripts may be submitted in English or Spanish. English-language manuscripts dealing with taxa or topics of Latin America and Spanish-language manuscripts must have a Spanish RESUMEN and an English ABsTRACT. Manuscripts and review copies of illustrations must be submitted in triplicate for all articles and short items (NOTES, NOTEWORTHY COLLECTIONS, POINTS OF VIEW, etc.). Follow the format used in recent issues for the type of item submitted. Allow ample margins all around. Manuscripts MUST BE DOUBLE-SPACED THROUGHOUT. For articles this includes title (all caps, centered), author names (all caps, centered), addresses (caps and lower case, centered), abstract and resumen, 5 key words or phrases, text, acknowledgments, literature cited, tables (caption on same page), and figure captions (grouped as consecutive paragraphs on one page). Order parts in the sequence listed, ending with figures. Each page should have a running header that includes the name(s) of the author(s), a shortened title, and the page number. Do not use a separate cover page or ‘erasable’ paper. Avoid footnotes except to indicate address changes. Abbreviations should be used sparingly and only standard abbrevia- tions will be accepted. Table and figure captions should contain all information relevant to information presented. All measurements and elevations should be in metric units, except specimen citations, which may include English or metric measurements. Authors of accepted papers will be asked to submit an electronic version of the manuscript. Microsoft Word 6.0 or WordPerfect 6.0 for Windows is the preferred software. Line copy illustrations should be clean and legible, proportioned to the MADRONO page. Scales should be in- cluded in figures, as should explanation of symbols, including graph coordinates. Symbols smaller than | mm after reduction are not acceptable. Maps must include a scale and latitude and longitude or UTM references. In no case should original illustrations be sent prior to the acceptance of a manuscript. Illustrations should be sent flat. No illustrations larger than 27 X 43 cm will be accepted. Presentation of nomenclatural matter (accepted names, synonyms, typification) should follow the format used by Sivinski, Robert C., in MApRONo 41(4), 1994. Institutional abbreviations in specimen citations should follow Holmgren, Keuken, and Schofield, Index Herbariorum, 8th ed. Names of authors of scientific names should be abbreviated according to Brummitt and Powell, Authors of Plant Names (1992) and, if not included in this index, spelled out in full. Titles of all periodicals, serials, and books should be given in full. Books should include the place and date of publication, publisher, and edition, if other than the first. All members of the California Botanical Society are allotted 5 free pages per volume in Maprono. Joint authors may split the full page number. Beyond that number of pages a required editorial fee of $40 per page will be assessed. The purpose of this fee is not to pay directly for the costs of publishing any particular paper, but rather to allow the Society to continue publishing MaproNo on a reasonable schedule, with equity among all members for access to its pages. Printer’s fees for illustrations and typographically difficult material @ $35 per page (if their sum exceeds 30 percent of the paper) and for author’s changes after typesetting @ $4.50 per line will be charged to authors. At the time of submission, authors must provide information describing the extent to which data in the manu- script have been used in other papers that are published, in press, submitted, or soon to be submitted elsewhere. Ih 8288 LIBRARIES iN 111 UTION | | Will INSTIT' 3 9088 0 | | | | | ae ee te Pentre ae tare ANN At ne TRL Froese sen POTS ET te we bse, TN eee te coe eee ae Pere herons Wikeere tases Ie Beg, a arr ate ee mar, pene ae ns SRR eas Weare mee a yeiee B z Z Qe ear wee : . OU Oe BERRY cat ota rete rere) ryt te tas py Cre ee 2 ONO & YN PRR ky PASS eee PE : ARE HE Aga te tty ere th oe a ae GD Lee TF We de Ye reve VED