We Wh alietvbe hoes belt Ma Bg ied Vin thay Hin ta Mig Be Bin tte BN ye Shae Hite Nagy RAAB aaleas Marita aha aM asia baie dre Meeihy Oat Mpg tie Be PWS ded sthte dhe the Mas Mintre Pe vba ee dhe ands ty a that: Pea iledty dates ‘A he a! Dag ag Baw Ldap Fea Hm ihe thai ite he eit CO ‘ Matin Hee Hig te Be athe nN rR ARY I yh tg Ps Boe tak tT ee De Ne sie ee pla ROME: faite MO BhE Foe Debate 6 vin Mae) oe Aerhavielinde Hal ete ie le Ns ithe rela val ale nn aren tacky oe Nie Nae tga he te onli Uh Baw Man The Hi PS Raia! eo Ma es Tis Hh Berit Mig Ite Benits thy tee Fe Ta Lg aR Pat Rally eM fy Mytia Te ya Oy mien Pa Ma Aaa ie IE nse ieee Melle sete Fignera ertoh eT aiee ss Foghton ie Maltin Stet atin Simtha lhe evn ite Ind Ha lig de it, Lee ete alee Padiadhe O10 re Baste itae! pip lpi iherite hatha Dare sass her de e im Tape sete ee der ae ei RM Fig tees a tha Lim one Ha asin et pet LD a dg IR Ny Kigali Beg Ua Na ity Pe eh Be te Poel a te aa as om Neti th ‘op the ity dew fi city Way endow 4a tae 2s Fu Me rh hae tala ee INT he te Pathe artaniiag ante Oe cae ems jain A 2 Nat Merah 8A Sha Mowe oa amen Fahehetece ta Nahi tne My Aas tes om igi* sMplig treats ener he terete arama ars estes ny ae te gener « ens! hed ay ase arrears yuan asa feted a ate Fie Deemed a whe Foes » WE Cet nate eet ant wt Va No oe Shad thesis hal wih sid att ee BEER GIR a eamepiies Gabe AS Wm sol wiv D BH Ae SS el oF aoa ahah ceywit PASM AAR ay Nee ave ae a at ate ea ata ong AGRA WH ata Beate et trea TNaeeatte a ea Tes een sang, Aen TR eR tea cai ae Naren Oe —_ 2 Ae Ve | Oo - = 4 MAL yO: 7: FF a a a x . : oe — - . aC ; _ VOLUME 55, NUMBER | JANUARY-MARCH 2008 MADRONO A WEST AMERICAN JOURNAL OF BOTANY By We jie = ‘ vat) Hi, SHAW ey ites aa RAN Weide ¥ / Woy jt wa aha guar” % riage j pf Me % @> *. ae pty : ‘>; Se, é Uy F ig yi a ce ai 'CONTENTS VASCULAR FLORA OF THE LOWER SAN FRANCISCO VOLCANIC FIELD, COCONINO COUNTY, ARIZONA ISG COU SUDO aes eccranere Se 0 se Dosic, reds paalsees ane Cee eae ee aOR RTE Cae tone I l A LATE HOLOCENE RECORD OF VEGETATION AND CLIMATE FROM A SMALL WETLAND IN SHASTA COUNTY, CALIFORNIA R. Scott Anderson, Susan J. Smith, Renata B. Jass, And W. Geoffrey ESO CLCUULE TUG aa asuerecneat talatac the chanel ata Bata eee cad arene ea necee eenta mre aa nees Maan arnaeaies eee ES) LOCAL SCALE VEGETATION MAPPING AND ECOTONE ANALYSIS IN THE SOUTHERN COAST RANGE, CALIFORNIA Robert J. Steers, Michael Curto, And V. L. Holland....cccccccccccecccceeccceeecenee 26 EcoLoGy AND ECOPHYSIOLOGY OF A SUBALPINE FELLFIELD COMMUNITY ON MOunr PINOS, SOUTHERN CALIFORNIA Arthur C. Gibson, Philip W. Rundel, And M. Rasoul Sharifi.............00.0+- 4] MORPHOLOGICAL TRAITS AND INVASIVE POTENTIAL OF THE ALIEN EUPHORBIA TERRACINA (EUPHORBIACEAE) IN COASTAL SOUTHERN CALIFORNIA Erin C. Riordan, Philip W. Rundel, Christy Brigham, And John ES ACT ciudad ee neste he ania stale A feet fa a eG ee ceeded este eee ee THE REDISCOVERY AND STATUS OF DISSANTHELIUM CALIFORNICUM (POACEAE) ON SANTA CATALINA ISLAND, CALIFORNIA Jenny Le MceCune Ajid DEnISCA ARRGDD scsccidsassecssses seen sae 60 RESURRECTION OF ASCLEPIAS SCHAFFNERI (APOCYNACEAE, ASCLEPIADOIDEAE), A RARE, MEXICAN MILKWEED Mark Fishbein, Veronica Judrez-Jaimes and Leonardo O. PANU GOO = OF COTS ae Soca sec tine ne pease Bea ON en taN ooeaee nee I acatewA eae ace te 69 HABITAT AND DISTRIBUTION OF CRYPTANTHA CRINITA GREENE (BORAGINACEAE) Brian A, Elliott And Samantha S. D. MACKEY vi.cccccccccccc cece cece eeeeetttteeeeeeeees 76 LIVESTOCK TRAMPLING AND LILAEOPSIS SCHAFFNERIANA VAR. RECURVA (BRASSICAEAE) Jacob W. Malcom And William R. RadKe............cccccccccccessseeccccesseecccseeeues 8 | NEw TAXA FOLLOWING A REASSESSMENT OF ERIASTRUM SPARSIFLORUM (POLEMONIACEAE) DAV LEG OVC mentee ie ON IIe ee cei Beat Ae eR falc diene Nes Bae me 82 PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL SOCIETY MAbDRONO (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 MADRONO, Susan Bainbridge, Jepson Herbarium, University of California, Berkeley, CA 94720-2465. Editor—POLLY SCHIFFMAN Department of Biology California State University Northridge, CA 91330-8303 madrono @csun.edu Book Editor—JON E. KEELEY Noteworthy Collections Editors—DIETER WILKEN, MARGRIET WETHERWAX Board of Editors Class of: 2008—ELLEN DEAN, University of California, Davis, CA ROBERT E. PRESTON, Jones & Stokes, Sacramento, CA 2009—DOoNoOvVAN BAILEY, New Mexico State University, Las Cruces, NM Mark BorcuHert, USFS, Ojai, CA 2010—FReED Hrusa, California Department of Food and Agriculture, Sacramento, CA RICHARD OLMSTEAD, University of Washington, Seattle, WA 2011—JAMIE KNEITEL, California State University, Sacramento, CA KEVIN Rice, University of California, Davis, CA CALIFORNIA BOTANICAL SOCIETY, INC. OFFICERS FOR 2007—2008 President: Dean Kelch, Jepson Herbarium, University of California, Berkeley, CA 94720, dkelch@sscl.berkeley. edu First Vice President: Andrew Doran, University and Jepson Herbaria, University of California, Berkeley, CA 94720, andrewdoran @ berkeley.edu Second Vice President: Dan Harder, Arboretum, University of California, Santa Cruz, CA 94064 Recording Secretary: Staci Markos, Friends of the Jepson Herbarium, University of California, Berkeley, CA 94720-2465, smarkos @ socrates.berkeley.edu Corresponding Secretary: Susan Bainbridge, Jepson Herbarium, University of California, Berkeley, CA 94720- 2465, sjbainbridge @calmail.berkeley.edu Treasurer: Susan C. Dunlap, Aerulean Plant Identification Systems, Inc., Menlo Park, CA 94025, susancdunlap @ hotmail.com The Council of the California Botanical Society comprises the officers listed above plus the immediate Past President, Michael Vasey, Department of Biology, San Francisco State University, 1600 Holloway Avenue, San Francisco, CA 94132, mvasey @sfsu.edu; the Editor of Madrono; three elected Council Members: James Shevock, National Park Service, Cooperative Ecosystems Studies Unit, 337 Mulford Hall, University of California, Berkeley 94720- 3114, jshevock @nature.berkeley.edu; Roy Buck, Jepson Herbarium, University of California, Berkeley, CA 94720, roybuck @email.msn.com; Diane Ikeda, USDA Forest Service, Vallejo, CA 94592, dikeda@fs.fed.us; Graduate Student Representatives: Abby Moore, Jepson Herbarium, University of California, Berkeley, CA 94720, ajmoore @ berkeley.edu. Webmasters: Curtis Clark, Biological Sciences Department, California State Polytechnic University, Pomona, CA 91768-4032, jcclark @csupomona.edu. This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). MADRONO, Vol. 55, No. 1, pp. 1-14, 2008 VASCULAR FLORA OF THE LOWER SAN FRANCISCO VOLCANIC FIELD, COCONINO COUNTY, ARIZONA KYLE CHRISTIE Deaver Herbarium, Department of Biological Sciences, Northern Arizona University, Flagstaff, AZ 86011-5640 kylechristiel @hotmail.com ABSTRACT The San Francisco Volcanic Field les near the southern edge of the Colorado Plateau in north- central Arizona, and is dominated by an extensive Pinyon-Juniper woodland. During 2004 and 2005, a floristic inventory vouchered 487 taxa from 74 families and 268 genera, including eight species endemic to Arizona. The Asteraceae, Poaceae, Fabaceae, Brassicaceae, and Scrophulariaceae comprised 51% of the total flora. Eriogonum, Muhlenbergia, Penstemon, Aristida, Astragalus, and Cryptantha were the best represented genera. Nonnative taxa accounted for ten percent of the total flora. Cryptantha minima and Suckleya suckleyana were vouchered as new records for Arizona. Key Words: Floristics, flora, Pinyon-Juniper woodlands, San Francisco Volcanic Field, volcanic endemism. Arizona displays the third highest diversity of vascular plants in the United States with over 3,500 known taxa; however, it is also the fifth most “‘at-risk’’ state with over 15% of its plant species at risk of extinction due to rarity or habitat loss (Stein 2002). While many floristic inventories document vascular plant diversity around the state, no vascular plant inventory exists for the lower San Francisco Volcanic Field (SFVF), an expansive Pinyon-Juniper woodland in north-central Arizona (Moore and Cole 2004). Inventories in adjacent areas have been primarily restricted to National Park Service units and scenic landscapes such as deep canyons or mountain peaks (e.g., McDougall & Haskell 1960; Joyce 1976; Hazen 1978; Schilling 1980; Kierstead 1981; Gilbert and Licher 2005; Moir 2006). Pinyon-Juniper woodlands have often been overlooked or ignored during the creation of inventories, and thus the relatively vast lower SFVF represents a significant gap in floristic knowledge. The SFVF lies at the convergence of several biotic communities, in the middle of a large elevational gradient, and at the junction of several known physiographic and floristic zones; and was therefore anticipated to have a high diversity of vascular plants (Brown and Lowe 1980; McLaughlin 1992). Cinder ecosystems resulting from the recent Sunset Crater volcanic eruption are also known to host several edaph- ically-limited, rare, and endemic plants (AGFD 2005a, b, c). The distributions of Pinyon-Juniper wood- lands, and potentially the understory species associated with this community type, have varied historically with changing climate variables (Be- tancourt et al. 1990). Shifts in vegetation distributions due to global climate change, especially in semiarid environments and at ecotones, are a definite possibility in upcoming years (Allen and Breshears 1998). The ecophysio- graphic nature of the lower SFVF (it encompass- es a woodland community in a semi-arid climate, and has a widespread ecotonal component) lends greater urgency to a floristic assessment of the area. The findings from this study will provide a benchmark of local floristic diversity and serve as a basis for future comparisons. Study Area The SFVF hes near the southern edge of the Colorado Plateau physiographic province in north-central Arizona; south-central Coconino County (Bailey et al. 1994; Bailey 1998). The lower SFVF, and synonymously the study area, was defined as the contiguous Pinus edulis/ Juniperus monosperma dominated portion of the SFVF. It is elevationally bound by montane coniferous forests above and grasslands below (Brown and Lowe 1980). Inclusions of other vegetation on north-facing slopes, drainages, canyons, or unusual microsites were also included as part of the study area. The study area lies between 35°10'47”" and 35°41'30" latitude and between —111°21'3" and —112°10'12” longitude, and encompasses approx- imately 1134 km* (Fig. 1). Elevations range from 1700 to 2400 m; however, approximately 84% of the study area falls between 1829 and 2134 m. Roughly 73% of the study area lies within the Coconino and Kaibab National Forests, while the remainder occurs on essentially undeveloped private (16%) and State (11%) land. Approxi- mately 70 widely-spaced cinder cones dot the Nw Colorado River CA N A 0 20 «40 80 Kilometers Fic. 1. otherwise flat plateau of the lower SFVF. The study area lacks major river systems, bodies of water or streams, and perennial water is essen- tially absent. Biotic communities are solely restricted to Great Basin Conifer Woodlands with several small inclusions of Petran Montane Conifer Forests and Desert Grasslands at anom- alous sites (Brown and Lowe 1980). The SFVF sits near the junction of the Colorado Plateau and Apachian floristic areas (McLaughlin 1989). The lower SFVF has a semi-arid climate and receives approximately 38 cm of precipitation annually; however, precipitation can vary dra- matically, often ranging between 25 and 65 cm from year to year (Bailey et al. 1994; Bailey 1998; WRCC 2005). Precipitation is bi-modal, as very dry late springs and early summers punctuate winter and monsoonal moisture. Approximately 50% of the annual precipitation comes from summer monsoons. The winter of 2004-2005 received 200% of the historical precipitation average, while the spring of 2005 was 140% wetter than normal. Summer rains of 2005 brought 75% of the historical precipitation average (WRCC 2005). The SF VF was created by at least seven major eruptive events, and is one of the major basaltic volcanic fields of the Colorado Plateau (Cooley 1962; Tanaka et al. 1986). It includes over 600 Pliocene, Pleistocene, and Holocene volcanic vents, volcanoes, and cinder cones, and _ their associated sheet deposits and lava flows (Tanaka et al. 1986). Local basaltic volcanism began roughly six millions yrs ago and has occurred MADRONO Lower SFVF Flagstaff ARIZONA [Vol. 55 NM Location of the lower San Francisco Volcano Field (lower SF VF). continuously for the past three million yrs (Moore et al. 1976). Paleomagnetic data, Potas- sium-Argon dating, and thorough field mapping, suggest that substrates of the SFVF range from less than one-thousand yrs old at the eastern edge of the volcanic field to over six-million yrs old at the southwestern edge of the volcanic field, only 100 km away (Tanaka et al. 1986; Moore and Wolfe 1987; Wolfe et al. 1987) Due to its volcanic history, igneous rocks and their associated soils are the dominant substrates of the SFVF. Basaltic rocks, cinders, and lava from the Holocene to Middle Pliocene cover about 80% of the of the study area. Quaternary alluviums (3.5%), Holocene to Middle Pliocene rhyolites and andesites (3%), Pleistocene alluvi- | ums (2%), and Permian sandstone and limestone | (11%) comprise the additional surficial geology | (Richard et al. 2000). ) METHODS Vascular plants were collected for 55d (ca. 400 hr) between March and October of 2004 and | 2005. An effort was made to collect every taxon, | as well as to collect from the entire geographic and habitat range of the study area. Special effort. was made to target areas and habitats that harbored greater diversity or locally unusual taxa. Voucher collections included plant descrip-. tions, habitat descriptions, locality descriptions, | and lists of associated species. UTM coordinates | and elevations were taken from a Garmin eTrex: Legend GPS unit (Garmin International; Olathe, 2008] (ABLE 1. Divisions and Classes Families Genera Polypodiophyta 2 3 Pinophyta Pinopsida 2 2 Gnetopsida | l Magnoliophyta Magnoliopsida a7 212 Liliopsida [2 50 Totals 74 268 KS). Voucher specimens were pressed, dried, mounted, and deposited in the Deaver Herbari- um (ASC) at Northern Arizona University following typical protocols (Weber 1976). Spec- imens were identified primarily using the /nter- mountain Flora (Cronquist et al. 1972+), the Flora of North America (Flora of North America Editorial Committee 1993+), treatments from the Manual of Vascular Plants of Arizona as published in the Journal of the Arizona-Nevada Academy of Science (Vascular Plants of Arizona Editorial Committee 1992+), and Seed Plants of Northern Arizona (McDougall 1973). All nomen- clature follows the USDA PLANTS database (USDA 2005). Digital searches of the ASC, ASU, ARIZ, DBG, and NAVA herbaria based upon all of the place names within the study area (gleaned from United States Geological Survey 7.5’ Quadrangle maps) were conducted via the Southwestern Environmental Information Network’s query tools (SEINet 2005). A geographical extraction based upon georeferenced specimens from the ASU and ARIZ herbaria was also conducted. Various species richness estimates for the study area were determined in an attempt to access the completeness of the inventory. Nonparametric estimators, functional extrapolation, and predic- tive regressions were used due to ease of implementation and compatibility with previous- ly gathered data. Species occurrences in twenty randomly selected 0.25 Ha plots were used to input the statistical models. Longino et al. (2002) provide a discussion of the merits of various species richness estimation techniques. RESULTS A total of 873 plant collections were made during two field seasons, representing 456 distinct taxa. Herbaria searches revealed an additional 31 taxa, thus the inventory identified 487 taxa from the lower SFVF. This inventory vouchered 313 previously uncollected taxa, representing a 280% increase in the documented flora of the lower SFVF. Seventeen taxa represented by historical collections were not encountered in the field during this survey. (These taxa are indicated by an abundance classification of ‘‘0’ in Appen- CHRISTIE: VASCULAR FLORA OF THE LOWER SFVF 3 TAXONOMIC COMPOSITION OF THE LOWER SAN FRANCISCO VOLCANIC FIELD. Species Infraspecific taxa Total Taxa s) 0 3 6 5 0 5 l 0 I 478 a2 6 378 98 2 100 479 8 487 dix A.) This absence may suggest that these species are locally uncommon; as approximately one-third reach the extent of their ranges near the SFVF, and another one-third lack much suitable habitat in the SFVF. On average it has been 25 yrs since one of these species was collected from the SFVF, perhaps because these species simply no longer occur or have decreased in abundance locally. Fourteen additional common taxa were encountered in the field but were not collected due to documentation by extant her- barium collections. These taxa are indicated with a pound sign (#) in Appendix A. The annotated catalog gives the scientific name, life form, abundance, habitat, and a list of voucher collections for each vascular plant of the lower SFVF (Appendix A). Seventy-four (74) families, 268 genera, 479 species, and 8 infraspecific taxa were vouchered from within the study area (Table 1). The Asteraceae (97 taxa), Poaceae (76 taxa), Fabaceae (33 taxa), Brassicaceae (22 taxa), Scrophularia- ceae (20 taxa), and Polygonaceae (18 taxa) were the best represented families. The five most species-rich families comprised approximately 51% of the flora, while 28 families were represented by a single taxon. Eriogonum (12 taxa), Muhlenbergia (10 taxa), Bromus (8 taxa), Penstemon (8 taxa), Aristida (7 taxa), and Astragalus (7 taxa) were the best represented genera. Perennial and annual herbs accounted for 79% of the total flora; while trees, biennial herbs, and annual graminoids accounted for only 12%. A total of 49 non-native plant species com- prised 10% of the total flora. The Poaceae (12 taxa), Asteraceae (9 taxa), Brassicaceae (5 taxa), Chenopodiaceae (3 taxa), and Fabaceae (3 taxa) represented the most common non-native fami- lies. Non-native species were most commonly annual (54%) or perennial herbs (25%). This inventory vouchered 38 new non-native taxa from the lower SFVF. Populus alba was the only escaped horticultural species that was discovered. Salsola tragus, Marrubium vulgare, Bromus tec- torum, and Thinopyrum intermedium seemed to be the most widespread non-natives; while Carduus nutans, Centaurea diffusa, and Sisymbrium altis- simum seemed to have the highest potential for displacing native vegetation. 4 MADRONO TABLE 2. Technique Ist order Jackknife 2nd order Jacknife ICE (incidence-based coverage estimator) Chao I upper 95% C.I. Chao II upper 95% C.I. Bootstrap Michaelis-Menten Function* (means) Cryptantha minima and Suckleya suckleyana represent new records for Arizona (Christie 2006), while collections of Peteria scoparia and Panicum mohavense represent new records for Coconino County. This inventory vouchered seven species which are endemic to Coconino and adjacent counties (Penstemon clutei, Mentzelia collomiae, Phemer- anthus validulus, Chrysothamnus molestus, Phlox amabilis, Lotus mearnsii var. mearnsi, and Ivesia multifoliolata); and Rorippa microtitis which 1s endemic to Arizona. Phacelia serrata, and Ca- missonia gouldii are nearly endemic to Arizona, but also occur in similar barren cinder habitats in New Mexico and Utah, respectively. Overall species richness estimates suggest that approximately 250—300 species of vascular plants occur within the lower SFVF (Table 2). DISCUSSION The underestimation of actual species richness of the lower SFVF by several techniques (Ta- ble 2) probably results from habitat patchiness, and illustrates some of the differences in habitat dynamics between the Southwest and Eastern hardwood forests (where several of the richness estimate techniques originated). While the land- scape of the lower SF VF is quite homogeneous as a whole, small landscape anomalies host dispro- portionate numbers of plant species. For exam- ple, a single small spring or wetland encompasses an infinitesimal portion of the study area, but (as with Law’s Natural Tank in the case of this inventory) might host 10 species (or 2% of the total flora) found nowhere else within the study area. While 20 sample plots used for the richness estimation analysis were randomly selected, the vast majority occurred on flat, open slopes and represented the most homogeneous portions of the local landscape. Statistically-driven tech- niques seemed to estimate the overall species richness fairly accurately for the homogeneous portions of the landscape, however as a result of habitat patchiness and random plot selection, they failed to take into account localized habitats of high plant diversity. Another approach to estimating the complete- ness of a flora is to apply one’s data to a Species Estimate [Vol. 55 OVERALL SPECIES RICHNESS ESTIMATIONS. * - functional extrapolation technique. Reference for Technique Smith & van Belle 1984 Palmer 1991 Chazdon et al. 1998, Chao et al. 2000 Chao 1987 Chao 1987 Smith & van Belle 1984 Colwell et al. 2004 regression based upon similar checklists. In western floras, species diversity is strongly positively-correlated with the elevational range of a study area. Greater ranges of elevation host a wider degree of precipitation and temperature regimes, as well as greater topographic heteroge- neity, and thus a greater variety of habitats and microhabitats (Bowers and McLaughlin 1996). Using the findings from 24 floras from Arizona and New Mexico, Bowers and McLaughlin (1996) regressed the number of native species found within an area by its elevational range and created a predictive tool for estimating inventory completeness in the Southwest. The lower SFVF has an elevational range of approximately 700 m (although over 85% of the study area actually occurs within a 350 m range), and based on the predictive tool of Bowers and McLaughlin (1996) is expected to harbor roughly 450 native plant species. This inventory vouchered 440 native plant taxa, which is over 95% of the 450 taxa expected. Areas with a high diversity of commu- nity types, canyon environments, permanent water, and a high degree of topographic rough- ness typically harbor additional plant species than the regression would indicate, while more homogeneous areas typically harbor fewer species (Bowers and McLaughlin 1982; Bennet and Kunzmann 1992; Bowers and McLaughlin 1996). This again suggests the floristic inventory of the lower SF VF is relatively complete as it has a single community type, lacks any major canyon environments or permanent water, and displays a high level of topographic homogeneity. The most emblematic species of the lower | Camissonia_ gouldii, - SFVF, Phacelia serrata, Penstemon clutei, and Mentzelia collomiae, are | soil endemics that inhabit barren cinder ecosys- | tems mostly associated with the Sunset Crater. eruption. The eruption of Sunset Crater roughly | 900 yrs ago has left a barren, moon-like land- scape devoid of much vegetation or any true soil (Wolfe et al. 1983). Little research has been done on the evolutionary history of these species, but | environmental stochasticity followed by subse- | quent speciation events provides a likely hypoth- i esis for their origin. Restrictive structural com- | ponents of the cinders and the associated lack of | competition, as opposed to nuances of soil 2008] chemistry, seem to influence local endemism in the lower SFVF (see Kelso et al. 2003). Phacelia serrata and C. gouldii are annuals that germinate after summer rains. Precipitation patterns and the presence of somewhat undis- turbed, large expanses of fine-textured, barren cinders seem to influence population dynamics. Populations of both species are fairly common within a restricted local range and are seemingly secure. A cursory comparison of Arizona mate- rial of C. gouldii to Utah material, suggests subspecific variation between the geographically separated populations and warrants further inquiry. P. clutei, a perennial species, occurs in similar habitats of fine-to-medium textured barren black cinders, and seemingly disturbance and fire regimes influence population dynamics (AGFD 2005b). Unchecked off-road-vehicle traffic poses the greatest threats to populations of these species. The narrowly restricted M. collomiae is perhaps the most fascinating of the local endem- ics. It occurs only within the lower SFVF and populations are uncommon, sporadic, and con- sist of only several plants. The plant was only recently described, is under-researched, and seems to be untracked by conservation agencies (Christy 1997; see NatureServe 2005). It inhabits medium to coarse-textured, oxidized red cinders, on moderate to steep sloughing slopes, in the vicinity of Red Mountain and Sunset Crater. Although the species was described as an annual, it is seemingly a biennial with a stout taproot, which often diverges at right angles from the stem. This taproot confers a degree of stability to the plant, on the shifting, sloughing slopes which preclude the establishment of other species. This unique plant certainly deserves additional re- search and monitoring. The presence of appro- priate habitat seems to be the primary limitation to the health and distribution of the species, and no human impacts on the populations were observed. In addition to the volcanic endemics, Phemer- anthus validulus and Panicum mohavense were found on a single limestone bench at the northern edge of the SFVF. Phemeranthus validulus 1s intolerant of competition but can inhabit a variety of substrates. While populations are known from Yavapai and Coconino counties, the collection from this inventory represents a new, intermediate locality for the plant (AGFD 2005d). Only several plants were seen, and this species is seemingly very uncommon locally. The critically imperiled (G1/S1) P. mohavense was previously known only from Mohave Co., AZ, and from a single small population in Socorro Co., NM that has not been relocated in over 15 yrs (NatureServe 2005; NMRPTC 2005). This species is a diminutive annual, and active grazing within its narrow local habitat could threaten the CHRISTIE: VASCULAR FLORA OF THE LOWER SFVF 5 species. Chrysothamnus molestus, Phlox amabilis, and Lotus mearnsii var. mearnsii are also endemic to northern Arizona. These species grow in a variety of soil types and habitats throughout a several-county range, and seem somewhat toler- ant of competition. They occur at lower eleva- tions within the SFVF, and perhaps can only persist in somewhat xeric climatic regimes. Potentially this group of endemics represents a relictual component of a past flora. C. molestus occurs on the northern edge of the volcanic field in shrub-dominated openings of the surrounding woodlands. Some synergistic qualities of open, low-lying habitats within Pinyon-Juniper wood- lands seem to influence its distribution. P. amabilis occurs primarily on the western side of the volcanic field in fine-textured clay soils with some interspersed rock. This perennial flowers very early in the season, perhaps as a temporally- based strategy to cope with infraspecific compe- tition. Within the study area, the high levels of early season soil moisture associated with clay- based soils, combined with the structural nuances of shallow rocky soil, seem to influence the distribution of P. amabilis. L. mearnsii var. mearnsii prefers fine-textured to sandy soils on flat, open slopes. It occurs, mostly singly or in small groups, in small patches of open soil at lower elevations. It is unclear what influences the distribution of this species, however the minimal sizes of populations suggest specific recruitment or pollination requirements. Ivesia multifoliolata and Rorippa microtitis, also Arizona endemics, are found in wet habitats within the lower SFVF. These species are typically found in either higher or wetter areas. I. multifoliolata grows in rocky drainage bottoms, while R. microtitis prefers moist or wet soils near ponds, fields, or meadows. These species seem secure regionally, albeit quite restricted within the lower SFVF. CONCLUSION This inventory has contributed significantly to the floristic knowledge of the San Francisco Volcanic Field by vouchering 313 new records from the study area, and through the contribu- tion of 873 collections to the Deaver Herbarium. These collections provide both a snapshot of the flora at a time of potentially intensified climate change, and also support future botanical and ecological research. This inventory gathered crucial data on the area’s rare and endemic plants, which will hopefully to be used in conservation decisions by management agencies. This flora, in conjunction with the flora of the higher elevation San Francisco Peaks (Moir 2006), documents a remarkable hotspot of plant diversity in Arizona. The greater San Francisco Volcanic Field encompasses only about 0.5% of 6 MADRONO the state’s area, but hosts almost 30% of its vascular plant species. ACKNOWLEDGMENTS I would like to thank my major advisor, Tina Ayers, for her exceptional guidance throughout this project. The staff of the Deaver Herbarium: H. David Hammond, Daniela Roth, Kristin Henningsen, and Mar-Elise Hill provided invaluable daily support and assistance. The Babbitt Ranches allowed me to collect plants on their property, the Arizona HDMS provided locality information on rare taxa, and the staff of the ASU and ARIZ herbaria accommodated me on several visits. LITERATURE CITED ALLEN, C. D. AND D. D. BRESHEARS. 1998. Drought- induced shift of a forest-woodland ecotone: Rapid landscape response to climate variation. Proceed- ings of the National Academy of Sciences 95: 14839-14842. ARIZONA GAME AND FISH DEPARTMENT (AGFD). 2005a. Camissonia gouldii. Unpublished abstract compiled and edited by the Heritage Data Man- agement System, Arizona Game and Fish Depart- ment, Phoenix, AZ. ARIZONA GAME AND FISH DEPARTMENT (AGFD). 2005b. Penstemon clutei. Unpublished abstract compiled and edited by the Heritage Data Man- agement System, Arizona Game and Fish Depart- ment, Phoenix, AZ. ARIZONA GAME AND FISH DEPARTMENT (AGFD). 2005c. Phacelia serrata. Unpublished abstract compiled and edited by the Heritage Data Man- agement System, Arizona Game and Fish Depart- ment, Phoenix, AZ. ARIZONA GAME AND FISH DEPARTMENT (AGFD). 2005d. Talinum validulum. Unpublished abstract compiled and edited by the Heritage Data Man- agement System, Arizona Game and Fish Depart- ment, Phoenix, AZ. BAILEY, R. G. 1998. Ecoregions map of North America. USDA Forest Service Miscellaneous Publication No. 1548. , P. E. AveRS, T. KING, AND W. H. MCNAB. 1994. Map of the ecoregions and subregions of the United States. 1:7,500,000. With supplementary table of map unit descriptions, compiled and edited by W. H. McNab and R. G. Bailey. U.S. Forest Service, Department of the Agriculture, Washing- ton, D.C. BENNETT, P. S. AND M. R. KUNZMAN. 1992. Factors affecting plant species richness in the Madrean Archipelago north of Mexico. Pp. 23-26 in A. M. Barton and S. S. Sloane (eds.), Chiricahua Moun- tains Research Symposium. Proceedings March 16-17, 1992. Southwest Parks and Monuments Association, Tuscon, AZ. BETANCOURT, J. L., P. S. MARTIN, AND T. R. VAN DEVENDER. 1990. Packrat middens: The last 40,000 yr of biotic change. University of Arizona Press, Tucson, AZ. Bowers, J. E. AND S. P. MCLAUGHLIN. 1982. Plant species diversity in Arizona. Madrono 29:227—233. AND 1996. Flora of the Huachuca Mountains, a botanically rich and _ historically significant sky island in Cochise County, Arizona. [Vol. 55 Journal of the Arizona-Nevada Academy of Science 29(2): 66-107. BROWN, D. E. AND C. H. LOWE. 1980. Biotic communities of the Southwest. Map _ at 1:1,000,000 scale. USDA Forest Service General Technical Report RM-78. CHAO, A. 1987. Estimating the population size for capture-recapture data with unequal catchability. Biometrics 43:783-791. , W. H. HWANG, Y. C. CHEN, AND C. Y. Kuo. 2000. Estimating the number of shared species in two communities. Statistica Sinica 10:227—246. CHAZDON, R. L., R. K. COLWELL, J. S. DENSLOW, AND M. R. GUARIGUATA. 1998. Statistical methods for estimating species richness of woody regeneration in primary and secondary rain forests of NE Costa Rica. Pp. 285-309 in F. Dallmeier and J. A. Comiskey (eds.), Forest biodiversity research, monitoring and modeling: Conceptual background and Old World case studies. Parthenon Publishing, Paris. CHRISTIE, K. 2006. Noteworthy Collections - Arizona. Madrono 53:409. CHRISTY, C. M. 1997. A new species of Mentzela Section Bartonia (Loasaceae) from Arizona. Novon 7:25—26. COLWELL, R. K., C. X. MAO, AND J. CHANG. 2004. Interpolating, extrapolating, and comparing inci- dence-based species accumulation curves. Ecology 85:2717-2727. CooLey, M. E. 1962. Geomorphology and the age of volcanic rocks in northeastern Arizona. Arizona Geological Society Digest 5:97—115. CRONQUIST, A., A. H. HOLMGREN, N. H. HOLMGREN, AND J. L. REVEAL. 1972+. Intermountain Flora: Vascular plants of the Intermountain West, U.S.A. The New York Botanical Garden and Hafner Publishing, New York, NY. FLORA OF NORTH AMERICA EDITORIAL COMMITTEE. 1993+. Flora of North America north of Mexico. Oxford University Press, New York, NY. GILBERT, E. AND M. LICHER. 2005. Flora and Vegetation of the West Fork of Oak Creek Canyon, Coconino County, Arizona. Desert Plants 21(1): 3-59. HAZEN, J. M. 1978. The flora of Bill Williams Mountain. M.S. thesis. Northern Arizona Univer- sity, Flagstaff, AZ. Joyce, J. F. 1976. Vegetation analysis of Walnut Canyon, Arizona. Arizona Academy of Science 11(3): 127-133. KELSO, S., N. W. BOWER, K. E. HECKMANN, P. M. BEARDSLEY, AND D. G. GREVE. 2003. Geobotany of the Niobrara chalk barrens in Colorado: A study of edaphic endemism. Western North Amer- ican Naturalist 63(3): 299-313. KIERSTEAD, F. R. 1981. Flora of Petrified Forest | National Park, Arizona. M.S. Thesis. Northern Arizona University, Flagstaff, AZ. | LONGINO, J. T., J. CODDINGTON, AND R. K. COLWELL. | 2002. The ant fauna of a tropical rain forest: | Estimating species richness three different ways. Ecology 83:689—702. | McDouGAL, W. B. 1973. Seed plants of Northern | Arizona. The Museum of Northern Arizona, Flagstaff, AZ. AND H. S. HASKELL. 1960. Seed plants of Wupatki and Sunset Crater National Monuments. 2008] Bulletin 37, Museum of Northern Arizona, Flag- staff, AZ. McLAUGHLIN, S. P. 1989. Natural floristic areas of the western United States. Journal of Biogeography 16:239-248. 1992. Are floristic areas hierarchically ar- ranged? Journal of Biogeography 19:21—342. Morr, W. 2006. Vascular Plants of the San Francisco Peaks area. An unpublished checklist. Deaver Herbarium, Northern Arizona University, Flag- staff, AZ. Moore, D. B. AND J. B. COLE. 2004. Arizona’s local floras and plant lists: A bibliography with locations and maps. Journal of the Arizona-Nevada Acad- emy of Science 37(1): 1-55. Moore, R. B. AND E. W. WOLFE. 1987. Geologic map of the east part of the San Francisco volcanic field, north-central Arizona. Miscellaneous Field Studies Map (MF-1960). Department of the Interior. U.S. Geologic Survey, Washington, D.C. , AND G. E. ULRICH. 1976. Volcanic rocks of the eastern and northern parts of the San Francisco volcanic field, Arizona. Journal of Research U.S. Geologic Survey 4(5): 549-560. NATURESERVE. 2005. NatureServe Explorer: a query of the NatureServe central databases. Available at: http://www.natureserve.org/explorer/. Accessed on 12/2005. NEW MEXICO RARE PLANT TECHNICAL COUNCIL (NMRPTC). 2005. Taxon report for Panicum mohavense. Available at: http://nmrareplants.unm. edu. Accessed on 12/2005. PALMER, M. W. 1991. Estimating species richness: the second-order jackknife reconsidered. Ecology (21912-1515. , G. L. WADE, AND P. NEAL. 1995. Standards for the writing of floras. BioScience 45:339—345. RICHARD, S. M., S. J. REYNOLDS, J. E. SPENCER, AND P. A. PEARTHREE. 2000. Geologic Map of Arizona (a revision of AZGS Map 26). Arizona Geologic Survey, Tuscon, AZ. SCHILLING, M. A. 1980. A vegetational survey of the Volunteer and Sycamore Canyon Region. MS thesis. Northern Arizona University, Flagstaff, AZ. SMITH, E. P. AND G. VAN BELLE. 1984. Nonparametric estimation of species richness. Biometrics 40:119—-129. SOUTHWEST ENVIRONMENTAL INFORMATION NET- WORK (SEINET). 2005. Arizona biological data network. Arizona State University Center of Environmental Studies, Phoenix, AZ, Available | at: http://seinet.asu.edu/. Accessed on 12/2005. STEIN, B. A. 2002. States of the Union: Ranking America’s Biodiversity. NatureServe, Arlington, VI. TANAKA, K. L., E. M. SHOEMAKER, G. E. ULRICH, AND E. W. WOLFE. 1986. Migration of volcanism | in the San Francisco volcanic field, Arizona. Geological Society of America Bulletin 97:129—141. | ‘UNITED STATES DEPARTMENT OF AGRICULTURE (USDA). 2005. Natural Resource Conservation Service. The PLANTS Database. National Plant Data Center, Baton Rouge, LA, Available at: http://plants.usda.gov. Accessed 12/2005. VASCULAR PLANTS OF ARIZONA EDITORIAL COMMIT- TEE. 1992 (+). A new flora for Arizona. Journal of the Arizona-Nevada Academy of Science 26:1. CHRISTIE: VASCULAR FLORA OF THE LOWER SFVF 7 WEBER, W. A. 1976. Rocky Mountain Flora. Colorado Associated University Press, Boulder, CO. WESTERN REGIONAL CLIMATE CENTER (WRCC). 2005. Historical Climate Information. Available at: http://www.wrcc.dri.edu/index.html. Accessed in December 2005. WOLFE, E. W., G. E. ULRICH, AND R. B. MOORE. 1983. San Francisco volcanic field, Arizona. Volcano News 13:13. , AND C. G. NEWHALL. 1987. Geologic map of the northwest part of the San Francisco volcanic field, north-central Arizona. Miscellaneous Field Studies Map (MF-1957). Department of the Interior. U.S. Geologic Survey, Washington, D.C. APPENDIX A ANNOTATED CHECKLIST OF THE VASCULAR PLANTS OF THE LOWER SAN FRANCISCO VOLCANIC FIELD Species are arranged first by division and class, then alphabetically by family, genus, species, and infraspe- cific rank. Nomenclature, author names, nativity status, and life form descriptions follow the United States Department of Agriculture PLANTS database (USDA 2005). Life form descriptions include: annual (A), biennial (B), perennial (P), shrub (S), and tree (T). Abundance classifications and habitat descriptions are subjective observations and pertain only to the lower SFVF. The abundance classifications: dominant or co- dominant (1), frequent (2), occasional (3), infrequent (4), very infrequent (5), and unseen, but previously vouchered from within the study area (0), follow Palmer et al. (1995). Habitat descriptions include: barren soil (BS), barren black cinders (BBC), canyons (CA), cinders (CI), cliffs (CL), cindery soil (CS), disturbed areas (DA), ditches (DI), drainages (DR), flat open slopes (FOS), fine-textured soil (FTS), high elevations (HE), low elevations (LE), mesic areas (MA), north-facing slopes (NFS), rocky areas (RA) roadsides (RO), sandy areas (SA), steep slopes (SS), sedimentary soils (SSO), tanks (TA), ubiquitous (UB), various habitats (VH), and wet areas (WA). All specimens were collected by the author, and are deposited at the Deaver Herbarium (ASC) unless otherwise noted. Taxon entries include: an asterisk (*) to indicate non-native, a pound sign (#) to indicate documentation from a previous herbarium collection, scientific name with authority, life form code, abundance code, general habitat descriptions, pertinent notes, and a list of voucher collections. PTERIDOPHYTA Dryopteridaceae Woodsia oregana D.C. Eat. ssp. cathcartiana (B.L. Robins.) Windham - P; 5; CL, NFS, RA; 1228. Pteridaceae Cheilanthes feei T. Moore - P; 4; CL; 357. Pellaea wrightiana Hook. - P; 5; CL; 725. PINOPHYTA — PINOPSIDA Cupressaceae Juniperus deppeana Steud. - T; 4; CSL; 1259. Juniperus monosperma (Engelm.) Sarg. - T; 1; S127 530, 1204-1308; 1322. Pinaceae #- Pinus edulis Engelm. - T; 1; UB; M. Porter 11. UB; 8 MADRONO Pinus ponderosa P.& C. Lawson var. scopulorum Engelm. - T; 3; CA, NFS, HE; 987. Pseudotsuga menziesii (Mirbel) Franco var. glauca (Beissn.) Franco - T; 5; NFS, CA; 599. PINOPHYTA — GNETOPSIDA Ephedraceae Ephedra viridis Coville - S; 3; BBC, 614. MAGNOLIOPHYTA — MAGNOLIOPSIDA Aceraceae Acer negundo L. var. interius (Britt.) Sarg. - T; 5; CA, MA; 1264B. Amaranthaceae Amaranthus albus L. - A; 3; DA, FOS; 1073, 1261. Amaranthus powellii S. Wats. - A; 2; DA, CLS; 761, 1038, 1233. * Amaranthus retroflexus L. - A; 3; DA, CLS; 888. Amaranthus torreyi (Gray) Benth. ex S. Wats. - A; 3; VH, CLS; 1146, 1202B, 1210. Anacardiaceae #4 Rhus trilobata Nutt. var. trilobata - S; 3; FOS, DR, RA; H.C. Sanchez 36. Toxicodendron rydbergii (Small ex Rydb.) Greene - P; 4; MA, RA, CA; 977. Apiaceae Cymopterus purpurascens (Gray) M.E. Jones - P; 3; PIS, FOS; 513,597, Lomatium foeniculaceum (Nutt.) Coult. & Rose ssp. macdoualii (Coult. & Rose) Theobald - P; 3; FTS, RA, FOS; 572, 647. Lomatium nevadense (S. Wats.) Coult. & Rose var. parishii (Coult. & Rose) Jepson - P; 3; FTS, RA; 551, 1G; 71zZ. Apocynaceae Apocynum cannabinum L. - P; 4; DR; 783, 1245. Asclepiadaceae Asclepias asperula (Dene.) Woods. ssp. asperula - P; 4: VH, RA, DR; 643. Asclepias engelmanniana Woods. - P; 4; DR, RA; 777. Asclepias involucrata Engelm. ex Torr. - P; 5; FOS, Is. LE: 730. Asclepias latifolia (Yorr.) Raf. - P; 4; VH, BBC, LE; 1293. Asclepias subverticillata (Gray) Vail - P; 2; VH, RO; 480, 1132. Asteraceae Achillea millefolium L. var. occidentalis DC. - P; 4; MA, DR; 1114. Acourtia wrightii (Gray) Reveal & King - P; 5; RA, LE A176: Ageratina herbacea (Gray) King & H.E. Robins. - P; 3; DR, NFS; 934, 1181. Agoseris glauca (Pursh) Raf. var. /aciniata (D.C. Eat.) Smiley - P; 3; FTS, FOS; 546, 575. Ambrosia acanthicarpa Hook. - A; 2; RO, DA; 1039, O67, 1318. Ambrosia psilostachya DC. - P; 3; DR, DI; 1142. Ambrosia tomentosa Nutt. - P; 4; FTS, DI, RO; 1077. Antennaria rosulata Rydb. - P; 4; HE; 1260. * Artemisia biennis Willd. var. biennis - B; 5; WA; 1236. Artemisia bigelovii Gray - S; 0; FOS, LE; H.C. Sanchez 42. [Vol. 55 Artemisia campestris L. ssp. borealis (Pallas) Hall & Clements var. scouleriana (Hook.) Crongq. - P; 3; FOS, CS; 896, 1206. Artemisia carruthii Wood ex Carruth. - P; 2; UB; 507, 920, 944, 1000, 1321. # Artemisia dracunculus McLaughlin 2294 (ASU). Artemisia frigida Willd. - P; 3; VH; 952. Artemisia ludoviciana Nutt. ssp. ludoviciana - P; 4; DR, CA, NFS; 936. Artemisia tridentata Nutt. ssp. tridentata - S; 4; RO; 1024, 1098. Baccharis pteronioides DC. - 8; 5; RA, CA, DR; 343. Bahia dissecta (Gray) Britt. -B; 2; CSL; 1314. Bidens heterosperma Gray - A; 4; CI, RO; 872, 1101. Brickellia californica (Torr. & Gray) Gray var. californica - S; 1; BBC, CA; 1165. Brickellia eupatorioides (L.) Shinners var. chlorolepis (Woot. & Standl.) B.L. Turner - P; 3; VH, RA; 1092. Brickellia grandiflora (Hook.) Nutt. - P; 4; NFS, DR; 1186. Brickellia oblongifolia Nutt. var. linifolia (D.C. Eat.) B.L. Robins. - S; 4; CI, LE; 394, 623. * Carduus nutans L. - P; 5; DA, WA; 699. * Centaurea diffusa Lam. - P; 3; DA, RO; 958. Chaetopappa ericoides (Torr.) Nesom - P; 2; VH; 329, 379, 569, 667. Chrysothamnus depressus Nutt. - S; 3; RA, FOS; 931, O97. Chrysothamnus molestus (Blake) L.C. Anders. - S; 4; FOS, SSO, N. AZ endemic; 1151, 1310, 1312. Cirsium arizonicum (Gray) Petrak - P; 4; CA, RA; 780. Cirsium calcareum (M.E. Jones) Woot. & Standl. - P; 3; RO,CE 481.658: Cirsium neomexicanum Gray var. neomexicanum - P; 32 Vids 737,858: Cirsium ochrocentrum Gray -P; 3; DI, RO, DR.; 490, 810, 839, 859. Cirsium undulatum (Nutt.) Spreng. var. undulatum - P; 3; CSL, DR, RO; 669, 884, 930, 1297. * Cirsium vulgare (Savi) Ten. - B; 4; DA; 1093, 1307. Cirsium wheeleri (Gray) Petrak -P; 3; HE, DR; 879, O27, 1078. Conyza canadensis (L.) Cronq. var. glabrata (Gray) Cronq. - A; 4; DA, RO; 922. Coreopsis tinctoria Nutt. var. atkinsoniana (Dougl. ex — Parse PDN Ae See _ Lindl.) H.M. Parker ex E.B. Sm. - A; 5; WA; 1196. Dyssodia papposa (Vent.) A.S. Hitche. - A; 5; SSO, | LE: 1299, | Ericameria nauseosa (Pallas ex Pursh) Nesom & Baird ssp. consimilis (Greene) Nesom & Baird - S; 1; UB, FOS; 1040, 1178, 1232. Ericameria nauseosa (Pallas ex Pursh) Nesom & | Baird ssp. nauseosa - S; 1; UB, FOS; 1138, 1168. | Erigeron bellidiastrum Nutt. var. bellidiastrum - A; 4; ; CA, RA; 837. Erigeron canus Gray - P; 5; FOS, HE; 805. Erigeron colomexicanus Gray - B; 2; RA; 345, 674. Erigeron concinnus (Hook. & Arn.) Torr. & Gray var. | concinnus - P; 4; FOS, FTS, SSO, LE; 745, 1303, 1309. | Erigeron divergens Torr. & Gray - B; 2; VH; 331, 571. Erigeron oreophilus Greenm. - P; 4; RA, CA; 849. Gaillardia pinnatifida Torr. var. pinnatifida - P; 3; FOS, CSL; 397, 509, 866, 943. Gnaphalium exilifolium A. Nels. - A; 4; WA; 1123, I191.1280: | 2008] | Grindelia nuda Wood var. aphanactis (Rydb.) Nesom | - P; 3; FOS, RO; 1099. | Grindelia squarrosa (Pursh) Dunal var. serrulata | (Rydb.) Steyermark - P; 0; RO, LE; S.P. McLaughlin 6475 (ARIZ). | Gutierrezia sarothrae (Pursh) Britt. & Rusby - S; 1; HUB, FOS; 1037. Helianthus annuus L. - A; 3; RO, DS; 873, 1036. Helianthus ciliaris DC. - P; 3; TA, RA; 648, 1080. Helianthus petiolaris Nutt. ssp. fallax Heiser - A; 3; | RO, DA; 863, 1313. | Heliomeris longifolia (Robins. & Greenm.) Cockerell _var. annua (M.E. Jones) Yates - A; 1; FOS; 1054, 938. Heterosperma pinnatum Cav. - A; 4; CI, DI; 1109, 211, 1250. Heterotheca villosa (Pursh) Shinners - P; 4; DR, HE; 1965. | Hieracium fendleri Schultz-Bip. var. fendleri - P; 4; HeS, HE; 1107. _ Hymenopappus filifolius Hook. var. lugens (Greene) | Jepson - P; 3; CS, DR, FOS; 332, 670. — Hymenothrix loomisii Blake - A; 4; RO; 1129. #4 Hymenoxys richardsonii (Hook.) Cockerell - P; 2; FOS, FTS; J. Bandoli s.n. | * Lactuca serriola L. -A; 3; FOS, FTS; 924, 998. _ Laennecia schiedeana (Less.) Nesom - A; 3; RA, DA; | 932. _ Laennecia sophiifolia (hunth) Nesom - A; 5; RA, CL; } 1126. _ Layia glandulosa (Hook.) Hook. & Arn. - A; 3; FOS; 579, 660. Machaeranthera canescens (Pursh) Gray ssp. canes- _cens - B; 2; VH; 386, 870, 1130. — Machaeranthera canescens (Pursh) Gray ssp. glabra (Gray) B.L. Turner - P; 0; FOS, CS; H.D. Hammond 11804, 11805. ~— Machaeranthera gracilis (Nutt.) Shinners - A; 3; FOS, | LE; 510, 940. Machaeranthera tanacetifolia (Kunth) Nees - B; 3: VH; 1004, 1147. Malacothrix torreyi Gray - A; 3; FOS, LE; 407, 587. ~ * Onopordum acanthium L. - B; 4: DA, RO; 939. Packera multilobata (Torr. & Gray ex Gray) W.A. Weber & A. Love - P; 3; VH, CS; 685. _ Pericome caudata Gray -S; 3; CA, RA, BBC; 950. ! Pseudognaphalium canescens (DC.) W.A. Weber ssp. canescens - B; 4; RA, CL; 1127. | Psilostrophe cooperi (Gray) Greene - P; 0; FOS, LE: JM. Rominger s.n. Psilostrophe sparsiflora (Gray) A. Nels. - P; 3: CSL: | 386, 620. Ratibida columnifera (Nutt.) Woot. & Standl. - P: 4: | RO; 689. Sanvitalia abertii Gray - A; 3; FOS, BS: 1059. Schkuhria multiflora Hook. & Arn. - A; 4; SA, DR; 948, 1266. Senecio flaccidus Less. var. flaccidus - P; 3; FOS, CS; 1 619. ## Senecio spartioides Torr. & Gray var. multi- _ capitatus (Greenm. ex Rydb.) Welsh - P; 4; FOS, CS: J. Bandoli s.n. Senecio spartioides Torr. & Gray var. spartioides - P: 3; FOS, CS; 941, 1169. _ Solidago canadensis L. var. gilvocanescens Rydb. - P; 14 SS, DR, HE; 1185. _ Solidago velutina DC. - P; 3; RA, DR: 929, 968, 1104. * Sonchus asper (L.) Hill - A: 4; DA, WA: 697. | | CHRISTIE: VASCULAR FLORA OF THE LOWER SFVF 9 Stephanomeria pauciflora (Torr.) A. Nels. - P; 2; FOS, CS: 391, 488,625, 821, 1053. Stephanomeria spinosa (Nutt.) S. Tomb - P; 3; FOS, CS; 499, 1075. Stephanomeria thurberi Gray - P; 4; FOS; 823. *# Taraxacum officinale G.H. Weber ex Wiggers - A; 4; DA; J.M. Rominger 2015. Tetradymia canescens DC. - S; 2; FOS, CSL; 504. Tetraneuris acaulis (Pursh) Greene var. arizonica (Greene) Parker - P; 4; SSO, RA; 338, 634, 833. Thelesperma subnudum Gray var. subnudum - P; 5: SSO, LE; 842. Townsendia incana Nutt. - A; 3; FOS; 385, 528, 744, 983, 1152. * Tragopogon dubius Scop. - A; 4; DA, RO; 757. Verbesina encelioides (Cav.) Benth. & Hook. f. ex Gray ssp. exauriculata (Robins. & Greenm.) J.R. Coleman - A; 3; DA, RO; 828, 986. Xanthium strumarium L. var. canadense - A; 4: RA, MA, DR, DI; 1128, 1143. Zinnia grandiflora Nutt. - P; 3; FOS, LE; 383, 830. Berberidaceae Mahonia fremontii (Torr.) Fedde - S; 2; VH, FOS; 409. Mahonia haematocarpa (Woot.) Fedde - S: 0; FOS, RA; R.E. Bodley s.n. Mahonia repens (Lindl.) G. Don - S; 4; DR, MA, NES; 835: Boraginaceae Cryptantha cinerea (Greene) Cronq. var. cinerea - P; 3; FOS, CS; 484, 663, 729. ¢# Cryptantha cinerea (Greene) Cronq. var. jamesii Cronq. - P; 3; FOS, CS; H.D. Hammond 11100, 12028. Cryptantha fendleri (Gray) Greene - A; 2; CS; 502, 959; 1026; 1319, 1205, Cryptantha gracilis Osterhout - A; 4; CSL; 584. Cryptantha minima Rydb. - A; 5; FOS, FTS; new AZ record; 641. Cryptantha pterocarya (Torr.) Greene var. pterocarya = A; 3: CL LES 566, 585,61 1. Cryptantha setosissima (Gray) Payson -P; 4; FOS, HE, DR; 803. Lappula occidentalis (S. Wats.) Greene var. cupulata (Gray) Higgins - A; 3; DA, FOS; 341, 601, 664. Lappula occidentalis (S. Wats.) Greene var. occiden- talis - A; 3; DA, FOS; 561. Lithospermum incisum Lehm. - P; 2; FOS, CS; 359, 3/0, 390. # Lithospermum multiflorum Torr. ex Gray - P; 4: FOS, DR, RA; M.L. Carson 19. Tiquilia nuttallii (Hook.) A. Richards. - A; 3; BBC, LE: 605,919,982; Brassicaceae * Alyssum minus (L.) Rothm. var. micranthum (C.A. Mey.) Dudley - A; 4; DA, RO; 754, 854. Arabis fendleri (S. Wats.) Greene var. fendleri - P; 3; RAS CA: 3505893. Arabis gracilipes Greene - P; 4; RA, FTS; 577. Arabis perennans S. Wats. - P; 3; CSL, SS, CS; 537, 563,596, 753. * Camelina microcarpa DC. - A; 5; DA, RO; 853. Descurainia obtusa (Greene) O.E. Schulz ssp. adeno- phora (Woot. & Standl.) Detling - A; 3; VH, CS; 857. Descurainia pinnata (Walt.) Britt. - A; 2; VH, CS; 526, 509; 7 39) 7472 7359: * Descurainia sophia (L.) Webb ex Prantl - A; 4; FOS; 406, 552. 10 MADRONO Draba cuneifolia Nutt. ex Torr. & Gray var. cuneifolia = is 3) FOS, LE: 3153:535,.553:. Erysimum capitatum (Dougl. ex Hook.) Greene var. purshii (Dur.) Rollins - P; 5; FOS; 1012. * Erysimum repandum L. - A; 5; TA, FTS; 401. Lepidium densiflorum Schrad. var. densiflorum - A; 5; DA: 755. Lepidium montanum Nutt. var. montanum - P; 3; FOS, FTS; 479, 841, 1047, 1058. Lesquerella intermedia (S. Wats.) Heller - P; 3; SSO; 339, 633. Lesquerella rectipes Woot. & Standl. - P; 5; FOS, FTS; 1061. Physaria newberryi Gray - P; 4; BBC; 533, 568. Rorippa microtitis (B.L. Robins.) Rollins - A; 4; WA, TA; AZ endemic; 1121, 1156. Rorippa sphaerocarpa (Gray) Britt. - A; 4; WA, TA; 1243, 1281. Schoenocrambe linearifolia (Gray) Rollins - P; 3; CS, CI; 482, 487, 506. * Sisymbrium altissimum L. - B; 3; FOS, CS, DA; 586, 602. Thelypodium wrightii Gray ssp. wrightii - B; 4; CA, RA, CL, NFS; 937, 1298. 4# Thlaspi montanum L. var. montanum - P; 5; HE; M. Porter 13. Cactaceae Echinocereus fendleri (Engelm.) F. Seitz - P; 3; FOS; 688, 746. Escobaria vivipara (Nutt.) Buxbaum var. arizonica (Engelm.) D.R. Hunt - P; 4; RA, FOS; 806. Opuntia fragilis (Nutt.) Haw. - P; 5; FOS, FTS; 1050. Opuntia macrorhiza Engelm. - S; 3; FOS,GA, SA, FTS; 692, 740, 832. Opuntia phaeacantha Engelm. - S; 0; FOS, LE; J. Bandoli s.n. Opuntia whipplei Engelm. & Bigelow - S; 2; UB; 609, 827. Callitrichaceae Callitriche heterophylla Pursh ssp. heterophylla - P; 5; TA, WA; 715. Capparaceae Cleome lutea Hook var. lutea - A; 0; RO, LE; R.J. Barr 216 (ARIZ). Cleome serrulata Pursh - A; 2; RO, DA; 489. Polanisia dodecandra (L.) DC. ssp. trachysperma (Torr. & Gray) Iltis - A; 3; BBC; 372, 1216. Wislizenia refracta Engelm. ssp. refracta - A; 5; FOS, LE; 984. Caprifoliaceae Symphoricarpos rotundifolius Gray var. parishii (Rydb.) Dempster - S; 5; NFS, RA, CA; 344, 836. Caryophyllaceae Arenaria eastwoodiae Rydb. var. adenophora Kearney & Peebles - P; 4; RA, CA; 348, 574. Arenaria lanuginosa (Michx.) Rohrb. ssp. saxosa (Gray) Maguire - P; 4; DR, HE; 724, 874, 1230. Drymaria glandulosa K. Presl - A; 3; WH, CS; 1096, 1108. Drymaria leptophylla (Cham. & Schlecht.) Fenzl ex Rohrb. - A; 0; BS, BBC, C.M. Christy 1868 (ASU). Drymaria molluginea (Lag.) Didr. - A; 3; BS, BBC; 501B, 1184, 1211. Chenopodiaceae Atriplex canescens (Pursh) Nutt. var. canescens - S; 4; LE, RA, FOS; 378. * Atriplex rosea L. - A; 3; DA, RO; 1068, 1069, 1070, 1194. [Vol. 55 Chenopodium atrovirens Rydb. - A; 4; VH; 1029. Chenopodium fremontii S. Wats. var. fremontii - A; 2; UB, CS, DA; 1046, 1085, 1093, 1161, 1183, 1269. Chenopodium graveolens Willd. - A; 2; CI, BS, CS; 1034. Chenopodium leptophyllum (Mogq.) Nutt. ex S. Wats. - A; 3; DA, FOS; 942. Chenopodium neomexicanum Standl. var. neomexica- num - A; 4; FOS, CS; 1203. Chenopodium salinum Standl. - A; 4; TA, RA, FTS; 403, 1264. Chenopodium watsonii A. Nels. - A; 3; FOS, FTS; 1045, 1072, 1134. * Kochia scoparia (L.) Schrad. - A; 3; DA, RO; 1234. Krascheninnikovia lanata (Pursh) A.D.J. Meeuse & Smit - S; 3; FOS, FTS, LE; 978. * Salsola tragus L. - A; 3; DA, FOS; 608. Suckleya suckleyana (TYorr.) Rydb. -A; 5; MA, WA, TA; new AZ record; 1193. Convolvulaceae * Convolvulus arvensis L. - P; 4; DA, RO; 656. Ipomoea costellata Torr. - A; 5; DR, RO; 1160, 1180. Cucurbitaceae Cucurbita foetidissima Kunth - P; 4; DR, RO; 767. Cuscutaceae Cuscuta pentagona Engelm. var. pentagona - A; 5; parasite on various hosts, LE; 1190. Elaeagnaceae * Elaeagnus angustifolia L. - T; 5; RO, DI; 1249. Elatinaceae Elatine brachysperma Gray - A; 5; WA, TA; 1197. Euphorbiaceae Acalypha neomexicana Muell.-Arg. - A; 5; BS, CS; 1255: Chamaesyce albomarginata (Torr. & Gray) Small - P; 3; FOS, CS, RO; 651, 1179. | Chamaesyce chaetocalyx (Boiss.) Woot. & Stand. var. chaetocalyx - P; 5; FOS, FTS, LE; 1149. | Chamaesyce fendleri (Torr. & Gray) Small - P; 2; VH, CS; 371,589, 683;. L090: | Chamaesyce glyptosperma (Engelm.) Small - A; 5; CSL, RO; 921. Chamaesyce revoluta (Engelm.) Small - A; 3; BBC; 1008, 1097, 1163, 1208. Chamaesyce serpyllifolia (Pers.) Small ssp. serypllifo-— lia - A; 2; UB, BS, DA; 483, 495, 902, 1084, 1122, 1164, LESS: 123 11262: Euphorbia bilobata Engelm. - A, 5, SS, BBC; 1217. Euphorbia brachycera Engelm. - P; 4; VH, CS, HE; — 503. Tragia ramosa Torr. - P; 3; CSL, CI, RA; 389, 624. Fabaceae Astragalus allochrous Gray var. playanus Isely - A; 2; | UB, CS; 337, 392, 560, 610, 666, 1060. Astragalus brandegeei Porter - P; 4; FTS, RO, DI; | 749, 819. Astragalus calycosus Torr. ex S. Wats. var. scaposus | (Gray) M.E. Jones - P; 4; SSO, LE; 843. Astragalus humistratus Gray var. humistratus - P; 4; RA, DR; 826. Astragalus lentiginosus Dougl. ex Hook. - B; 3; FOS, ° LE; 521, 640, 742, 868. Astragalus mollissimus Torr. var. mogollonicus | (Greene) Barneby - P; 4; FOS, FTS; 736. Astragalus sabulonum Gray -A; 5; RO, DI, BBC; 630. | Caesalpinia drepanocarpa (Gray) Fisher - P; 5; CS, FOS, LE; 384. | Dalea albiflora Gray - P; 4; DR, RA; 1103. 2008] Dalea candida Michx. ex Willd. var. oligophylla (Torr.) Shinners - P; 4; DR, RA; 491. Dalea leporina (Ait.) Bullock - A; 4; RO, BBC; 760, 995, 1248. Dalea polygonoides Gray -A; 5; BBC, BS; 1158, 1212, 1254. Dalea urceolata Greene - A; 5; BBC, BS; 1159, 1213. Desmanthus cooleyi (Eat.) Trel. - P; 4; FOS, FTS, SA; 831, 999. Desmodium rosei Schub. - A; 5; SS, BBC; 1218. Glycyrrhiza lepidota Pursh - P; 5; DR, SA; 775. Hedysarum boreale Nutt. ssp. boreale - P; 0; FOS; G.A. Goodwin 1227. Lathyrus eucosmus Butters & St. John - P; 4; DR, RA; 970, 1089, 1141. Lotus mearnsii (Britt.) Greene var. mearnsii - P; 4; FOS, LE, FTS, N. AZ endemic; 637, 732. Lotus wrightii (Gray) Greene - P; 3; VH, FOS, DR; 334. Lupinus brevicaulis S. Wats. - A; 3; BS, LE; 330, 582, 638, 668. Lupinus hillii Greene - P; 0; FOS, DR, HE; E. Van Winkle 110, 143. Lupinus kingii S. Wats. var. kingii - A; 3; BS; 645, 662, 875: * Medicago minima (L.) L. - A; 5; RO; 657. * Medicago sativa L. ssp. sativa - P; 5; DA; 1201. * Melilotus officinalis (L.) Lam. - B; 3; RO, DA; 686, £315. Oxytropis lambertii Pursh var. bigelovii Gray - P; 3; FOS, CS; 615. Peteria scoparia Gray - P; 5; CS; new Coconino County record; 801. Phaseolus angustissimus Gray - P; 3; BBC; 616. # Psoralidium lanceolatum (Pursh) Rydb. - P; 4; BBC, SA, DR; C.F. Deaver 473 (UA). Psoralidium tenuiflorum (Pursh) Rydb. - P; 3; RA, IR, FOS, CS; 782, 877. Robinia neomexicana Gray var. neomexicana -S; 5; CA, RA; 346. Sophora nuttalliana B.L. Turner - P; 0; FOS; G.A. Goodwin 1224. Fumariaceae Corydalis aurea Willd. - A; 4; DA, RO; 400, 543, 603. Geraniaceae * Erodium cicutarium (L.) L’Hér. ex Ajit. ssp. cicutarium - A; 4; DA, CSL; 527, 554. Geranium caespitosum James var. caespitosum - P; 4; FOS, HE, CS; 802. Grossulariaceae Ribes cereum Dougl. - S; 2; WH, CS; 410, 524, 531, 536, 598. Hydrophyllaceae Nama dichotomum (Ruiz & Pavon) Choisy - A; 4; CSL; SOLA. Phacelia alba Rydb. - A; 3; FTS, DI, TA; 336, 646, 713, 1062. Phacelia crenulata Torr. ex S. Wats. var. crenulata - A; 4; CSL, LE; 388, 617, 867. Phacelia serrata J. Voss -A; 3; BBC; endemic to N. AZ and N. NM; 971, 988, 1015, 1032, 1215. Juglandaceae Juglans major (Torr.) Heller - T; 5; MA, DR; 774. Lamiaceae Dracocephalum parviflorum Nutt. - A; 3; BS, DA; 710. Hedeoma drummondii Benth. - A; 3; RA; 492, 613, 844, 918. CHRISTIE: VASCULAR FLORA OF THE LOWER SFVF 1] Hedeoma oblongifolia (Gray) Heller - P; 5; MA, RA, DE: Llp7. * Marrubium vulgare L. - P; 3; DA, DR, TA; 335, 679. Monarda fistulosa L. ssp. fistulosa var. menthifolia (Graham) Fern. - P; 5; MA, RA, DR; 773, 778. Monardella odoratissima Benth. - P; 5; SS, NES, BBC, HE; 889. Salvia reflexa Hornem. - A; 3; RO, DI; 818, 871, 993. Salvia subincisa Benth. - A; 4; RO, MA, RA; 991, 1145. Stachys rothrockii Gray - P; 3; FTS, DI, TA; 846, 1079. Linaceae Linum australe Heller var. glandulosum Rogers - A; 3; FOS; 382, 876, 1041, 1071, 1086, 1110. Linum lewisii Pursh var. lewisii - P; 4; FOS, RO; 766, 1044. Linum puberulum (Engelm.) Heller - P; 5; FOS, FTS; 103; Loasaceae Mentzelia albicaulis (Dougl. ex Hook.) Dougl. ex Torr. & Gray - A; 3; FOS, BBC, CI; 326, 583. Mentzelia collomiae C.M. Christy - B/P; 4; BBC, SS; narrowly endemic to the SFVF; 626, 891, 895, 1017, 1: Mentzelia montana (A. Davids.) A. Davids. - A; 3; FOS, BBC, CI; 591, 701. Mentzelia multiflora (Nutt.) Gray - P; 1; CSL, RO; 390, 485, 768, 807, 865, 946, 989, 1131, 1209, 1292. Mentzelia rusbyi Woot. - B; 4; RO; 856, 883. Malvaceae * Malva neglecta Wallr. - A; 4; DA; 1076. Sphaeralcea fendleri Gray - P; 3; VH; 678, 711, 809, 816, 1202. Sphaeralcea hastulata Gray - P; 3; LE, FOS, DR; 333, 735, 829. Sphaeralcea parvifolia A. Nels. - P; 3; LE, FOS, CS; 627, 981, 1296, 1306. Molluginaceae * Mollugo cerviana (L.) Ser. - A; 4; BBC; 1031, 1162, 1320; 1291, Nyctaginaceae Mirabilis albida (Walt.) Heimerl - P; 0; FOS; C.F. Deaver 3355 (ASU). Mirabilis coccinea (Torr.) Benth. & Hook. f. - P; 4; FOS, RA; 847. Mirabilis decipiens (Standl.) Standl. - P; 5; HE, CS; 1227. Mirabilis linearis (Pursh) Heimerl - P; 3; VH, RO, RA; 741, 899, 990, 1035. Mirabilis multiflora (Torr.) Gray - P; 3; FOS, CS; 370. Mirabilis oxybaphoides (Gray) Gray - P; 3; RA; 1005. Oleaceae Forestiera pubescens (Gray) Gray var. pubescens - S; 3; DR, CA, BBC; 342, 374, 974. Menodora scabra Gray - P; 3; FOS, FTS, CS; 700, 7315 101K Onagraceae Camissonia gouldii Raven - A; 3; BBC; endemic to N. AZ and S. UT; 607, 915, 985, 1014, 1166. Epilobium brachycarpum K. Presl - A; 4; FOS; 886, 1100. Epilobium ciliatum Raf. ssp. ciliatum - P; 5; WA, DA; 1237; Gaura coccinea Nutt. ex Pursh - P; 3; FOS, FTS; 644, 728, 881. 12 MADRONO Gaura hexandra OrteA ssp. gracilis (Woot. & Standl.) Raven & Gregory - P; 3; FOS, DR, CS; 960, 992. Gaura mollis James - A; 4; RO; 758. Gayophytum ramosissimum Torr. & Gray - A; 5; FOS, FTS; 639, 734. Oenothera caespitosa Nutt. ssp. marginata (Nutt. ex Hook. & Arn.) Munz - P; 4; BBC; 539, 594, 892. Oenothera coronopifolia Torr. & Gray - P; 5; CS, RO; 659. Oenothera flava (A. Nels.) Arrett ssp. flava - P; 5; HE, FTS; 1049. Oenothera pallida Lindl. ssp. runcinata (Engelm.) Munz & W. Klein - P; 3; RO, CS; 1002. Orobanchaceae Orobanche fasciculata Nutt. - A; 5; SS, CI; 622. Orobanche ludoviciana Nutt. ssp. /udoviciana - A; 4; Di, PiS:-733.815: Oxalidaceae Oxalis sp. L. - A; 5; MA, RA, DR; 1117. Papaveraceae Argemone_ pleiacantha Greene ssp. ambigua G.B. Ownbey - P; 4; LE, RO, FOS; 704. Plantaginaceae Plantago argyraea Morris - A; 3; FOS, RA; 500, 727, 845, 1048, 1251. Plantago patagonica Jacq. - A; 4; fFOS, LE; 738. Polemoniaceae Gilia leptomeria Gray - A; 4; BBC, LE; 581. Gilia ophthalmoides Brand - A; 3; FOS, CS; 399, 523, 525, 542, 593, 702. Ipomopsis aggregata (Pursh) V. Grant ssp. formosis- sima (Greene) Wherry - B; 3; CSL; 1317. Ipomopsis arizonica (Greene) Wherry - B; 3; CSL;562, 808. Ipomopsis longiflora (Torr.) V. Grant - B; 3; FOS, CS; 612. Ipomopsis multiflora (Nutt.) V. Grant - P; 3; CSL; 498, 994. Ipomopsis polycladon (TYorr.) V. Grant - A; 5; LE, CSL; 629. Phlox amabilis Brand - P; 3; RA, FTS; endemic to N. and C. AZ; 545, 547, 573, 1253. Phlox gracilis (Hook.) Greene ssp. gracilis - A; 3; FOS, BA; 541, 578, 604. Phlox longifolia Nutt. - P; 4; FTS, RA; 529. Polygonaceae Eriogonum alatum Torr. var. alatum - P; 4; FOS, HE, BIS: /63. Eriogonum cernuum Nutt. var. cernuum - A; 3; FOS, CS; 869, 945, 1016, 1135. Eriogonum corymbosum Benth. var. corymbosum - S; 4; BBC; 1174. Eriogonum corymbosum Benth. var. glutinosum (M.E. Jones) M.E. Jones - S; 4; BBC; 972, 1175. Eriogonum deflexum Torr. var. deflexum - A; 4; CI; 1295: Eriogonum hookeri S. Wats. - A; 4; Cl; 917, 1033. Eriogonum jamesii Benth. var. flavescens S. Wats. - P; 5; FOS, FTS, CS; 1018. Eriogonum palmerianum Reveal - A; 5; SS, CI, CS; 1182. Eriogonum pharnaceoides Torr. var. pharnaceoides - A; 3; Cl; 486. Eriogonum polycladon Benth. - A; 3; FOS, CS; 961, 1027, 1173. Eriogonum racemosum Nutt. - P; 4; FOS, FTS, HE; 1001. [Vol. 55 Eriogonum wrightii Torr. ex Benth. var. wrightii - P; 4; FOS, FTS, RA; 1136, 1140, 1177. Polygonum amphibium L. - P; 5; WA; 1198. * Polygonum aviculare L. - P; 3; DA, MA, TA; 1238. Polygonum douglasii Greene ssp. johnstonii (Munz) Hickman - A; 4; FOS, FTS; 709. Polygonum lapathifolium L. - A; 5; WA; 1235. * Rumex crispus L. - P, 5; WA; 862, 1242. Rumex salicifolius Weinm. var. mexicanus (Meisn.) C.L. Hitche. - P; 4; WA; 402, 825, 1082. Portulacaceae Phemeranthus validulus (Greene) Kiger - P; 5; SSO; N. AZ endemic; 1302. Portulaca halimoides L. - A; 4; DR, BS, CI; 1167, 1267B, 1273. Portulaca oleracea L. - A; 4; DR, CSL; 979, 1267. Primulaceae Androsace septentrionalis L. ssp. puberulenta (Rydb.) G.T. Robbins - A; 5; CS, HE, DR, CA; 1229. Ranunculaceae * Ceratocephala testiculata (Crantz) Bess. - A; 4; BS, DA; 514, 544. Clematis ligusticifolia Nutt. var. ligustisifolia -P; 5; DR, RA; 955. Delphinium scaposum Greene - P; 3; FOS; 395, 654, 684, 751. Myosurus apetalus C. Ay var. montanus (Campbell) Whitemore - A; 5; WA; 714. Thalictrum fendleri Engelm. ex Gray - P; 4; DR, CA, NES; 680. Rosaceae Amelanchier utahensis Koehne var. utahensis - S; 4; DR, SA; 897. Chamaebatiaria millefolium (Torr.) Maxim. - S; 4; RA, CL, SSO; 412. # Fallugia paradoxa (D. Don) Endl. ex Torr. - S; 2; VH, BBC; H.C. Sanchez 31 Holodiscus dumosus (Nutt. ex Hook.) Heller - S; 5; NFS, CL, HE; 1257. #- Ivesia multifoliolata (Torr.) Keck - P; 5; MA, RA, DR, endemic to N. and C. AZ, not collected due to lack of flowering material; S.J. Pinkerton s.n. Potentilla biennis Greene - A; 5; DR, RA; 676. Potentilla norvegica L. ssp. monspeliensis (L.) Aschers. & Graebn. - P; 5; WA, MA; 861, 1241. Potentilla penslyvanica L. var. pensylvanica - P; 5; RA, CL, HE; 1256. * Potentilla recta L. - P; 5; FOS, MA, DR; 1087. Purshia stansburiana (Torr.) Henrickson - S; 3; CS, DR, RA; 898. Rosa woodsii Lindl. var. woodsii - S; 5; MA, RA, DR; 973. Rubiaceae Galium wrightii Gray - P; 3; SS, CSL, CA, RA; 890, 1010, 1207, 1226. Rutaceae Ptelea trifoliata L. ssp. angustifolia (Benth.) V. Bailey | - 8; 5; RA, DR; 772. Salicaceae * Populus alba L. - T; 5; RO, DA; 852. Populus angustifolia James - T; 5; MA, RA, CA, DR; | 7s 'e Populus tremuloides Michx. - T; 5; NFS, RA, CA; | 838. Salix exigua Nutt. - S; 4; DR, CA, WA; 588, 1244, 1272, 1288. ) Salix gooddingii Ball - T; 5; WA, DI, RO; 1043, — 1286. 2008] Saxifragaceae Heuchera rubescens Torr. var. rubescens - P; 4; NFS, CL; 779, 894, 1222. Scrophulariaceae Castilleja austromontana Standl. & Blumer - P; 5; CSL; 687. Castilleja integra Gray var. integra - P; 3; CSL; 549, 595, 764. Castilleja linariifolia Benth. - P; 3; FOS, BBC; 497, 964. Cordylanthus parviflorus (Ferris) Wiggins - A; 4; FOS, LE; 851, 1088. Cordylanthus wrightii Gray ssp. tenuifolius (Pennell) Chuang & Heckard - A; 3; FOS; 923. Limosella acaulis Sessé & Moc. - A; 5; WA; 1195. * Linaria dalmatica (L.) P. Mill. ssp. dalmatica - P; 4; DA, CSL; 621. Mimulus rubellus Gray -A; 3; CSL; 516, 522, 532, 564, 565. Orthocarpus purpureoalbus Gray ex S. Wats. - A; 3; FOS, FTS; 850, 1003. Penstemon barbatus (Cav.) Roth ssp. torreyi (Benth.) Ieck = Ps 3: CSL; 822, 962, 963. Penstemon caespitosus Nutt. ex Gray var. desertipicti (A. Nels.) N. Holmgren - P; 4; FOS, SSO, LE; 635, 743. Penstemon clutei A. Nels. - P; 4; BBC; endemic to N. AZ; 618. Penstemon linarioides Gray ssp. linarioides - P; 4; VH, CSL; 665. Penstemon ophianthus Pennell - P; 3; RA, FOS, CS; 358, 632, 671. Penstemon palmeri Gray var. palmeri - P; 4; RO; 408. Penstemon rostriflorus Kellogg - P; 3; RA, CA, CL; 848. Penstemon strictus Benth. - P; 5; RO, HE; 878. * Verbascum thapsus L. - B; 3; DA; 804. Veronica anagallis-aquatica L. - P; 5; WA; 698. Veronica peregrina L. ssp. xalapensis (Kunth) Pennell - A; 5; WA; 723. Solanaceae Chamaesaracha coronopus (Dunal) Gray - P; 4; DA, FTS, RA; 653. Datura wrightii Regel - P; 4; RA, LE; 631. Lycium pallidum Miers var. pallidum - S; 3; FOS; 377, 600. Nicotiana attenuata Torr. ex S. Wats. - A; 4; FOS, FTS, DI; 493. Nicotiana obtusifolia Mertens & Aleotti var. obtusi- folia - P; 4; CI, DI; 398, 567. Physalis hederifolia Gray var. fendleri (Gray) Cronq. - P; 3; FOS, RA; 880, 933, 1028. Solanum jamesii Torr. - P; 4; FOS, FTS; 511, 1133. Solanum triflorum Nutt. - A; 4; CSL; 887. Tamaricaceae * Tamarix ramosissina Ledeb. -S; 5; DR, RO; 769, 1294. Ulmaceae * Ulmus pumila L. - T; 5; RO, DI; 380. Verbenaceae Glandularia bipinnatifida (Nutt.) Nutt. - P; 3; FOS, TA, RO; 381, 705, 1111. Phyla cuneifolia (Torr.) Greene - P; 4; TA, FTS, RA; 649, 1083. Verbena bracteata Lag. & Rodr. - P; 3; TA, FTS; 494, T92. Viscaceae Arceuthobium divaricatum Engelm. - P; 4; parasitizing Pinus edulis; 901. CHRISTIE: VASCULAR FLORA OF THE LOWER SFVF [3 Phoradendron juniperinum Engelm. ex Gray - P; 4: parasitizing Juniperus monosperma; 900. Vitaceae Parthenocissus vitacea (Knerr) A.S. Hitchc. - P; 5; DR, RA; 776. Vitis arizonica Engelm. - P; 5; DR, RA; 774B. Zygophyllaceae * Tribulus terrestris L. - A; 4; RO, DA; 817. MAGNOLIOPHYTA - LILIOPSIDA Agavaceae Yucca angustissima Engelm. ex Trel. var. angustis- sima - 8S; 2; VH, CSL; 655, 677. Yucca baccata Torr. var. baccata - S; 4; RA; 834. Yucca baileyi Woot. & Standl. - S; 0; FOS, CS; H.J. Fulton 8220 (ARIZ). Alismataceae Alisma triviale Pursh - P; 5; WA, TA; 1285. Commelinaceae Commelina dianthifolia Delile var. longispatha (Torr.) Brashier - P; 3; SS, CI; 1188, 1214. Tradescantia occidentalis (Britt.) Smyth var. occiden- talis - P; 4; BBC; 1013. Cyperaceae Carex athrostachya Olney - P; 0; WA; H.D. Hammond 12019. Carex brevior (Dewey) Mackenzie - P; 5; WA; 1153, 1283. Carex duriuscula C.A. Mey. - P; 5; RA, CA; 935. Carex geophila Mackenzie - P; 4; RA, CA, NFS; 925. Carex occidentalis Bailey - P; 3; RA, DR, CA, MA; 340, 719, 840. Cyperus esculentus L. var. leptostachyus Boeckl. - P: 3; DR; SA; 770, 1271. Cyperus fendlerianus Boeckl. - P; 4; RA; 505, 1091. Cyperus squarrosus L. - A; 5; DR, SA; 947. Eleocharis palustris (L.) Roemer & J.A. Schultes - P: 5; WA; 720, 1199. Schoenoplectus acutus (Muhl. ex Bigelow) A.& D. Love var. occidentalis (S. Wats.) S.G. Sm. - P; 5: WA; 1287. Hydrocharitaceae Elodea bifoliata St. John - P; 5; WA; 1200. Iridaceae Tris missouriensis Nutt. - P; 5; MA, DR; 1115. Juncaceae Juncus tenuis Willd. - P; 4; WA, DR; 721, 1120, 1278. Liliaceae Calochortus ambiguus (M.E. Jones) Ownbey - P; 3; FOS, FTS, RA; 650, 707. Echeandia flavescens (J.A. & J.H. Schultes) Cruden - P; 5; HE, FTS, RA; 1106. Poaceae Achnatherum hymenoides (Roemer & J.A. Schultes) Barkworth - P; 4; FOS, LE, BBC; 373. Achnatherum speciosum (Trin. & Rupr.) Barkworth - P; 5; RA; 1006. * Aegilops cylindrica Host - A; 4; DA, RO; 693. Agropyron cristatum (L.) Aertn. ssp. pectinatum (Bieb.) Tzvelev - P; 4; FOS, FTS; 691, 1102. Alopecurus carolinianus Walt. - A; 5; WA; 718. Andropogon gerardii Vitman - P; 3; BBC, DR; 375, 969. Aristida adscensionis L. - A; 5; RA; 1268. # Aristida arizonica Vasey - P; 4; BBC, CI; R.A. Darrow s.n. (UA). 14 MADRONO Aristida divaricata Humb. & Bonpl. ex Willd. - P; 3; BBC; 1025, 1170, 1247. Aristida havardii Vasey - P; 4; BBC; 916. Aristida purpurea Nutt. var. fendleriana (Steud.) Vasey - P; 3; VH, CSL; 606, 1171. Aristida purpurea Nutt. var. /ongiseta (Steud.) Vasey - P; 3; FOS; CSL; 661, 706. Aristida purpurea Nutt. var. purpurea - P; 3; FOS; CSL; 606, 1172. Bouteloua aristidoides (Kunth) Griseb. - A; 4; CSL, LE, RA; 1290. #4 Bouteloua curtipendula (Michx.) Torr. - P; 3, RA; C. Jass s.n. Bouteloua eriopoda (Torr.) Torr. - P; 3; FOS, LE; 328, 387. Bouteloua gracilis (Willd. ex Kunth) Lag. ex Griffiths - P; 1; UB; 996, 1304. Bouteloua simplex Lag. - A; 5; BS; 1064. Bromus carinatus Hook. & Arn. - P; 0; TA; L.C. Moore s.n. Bromus ciliatus L. - P; 5; RA; 1223. * Bromus commutatus Schrad. - A; 3; DA, RO; 708, 102, 8). * Bromus inermis Leyss. ssp. inermis - P; 4; DA, RO; 690, 1042. * Bromus japonicus Thunb. ex Murr. - A; 4; DA, MA; WN Bromus lanatipes (Shear) Rydb. - P; 4; CA, DR, NFS, Hi; O75, 1125. * Bromus rubens L. - A; 5; DA, LE; 628. * Bromus tectorum L. - A; 3; DA, CSL; 534, 540, 592. Dasyochloa pulchella (Kunth) Willd. ex Rydb. - P; 5; LE, FOS, BS; 980. * Echinochloa crus-galli (L.) Beauv. - A; 4; WA; 1240. Echinochloa muricata (Beauv.) Fern. var. microsta- chya Wieg. - A; 4; WA; 1118, 1154, 1239. Elymus canadensis L. - P; 5; MA, DR, RA; 976, 1274. Elymus elymoides (Raf.) Swezey ssp. brevifolius (J.G. Sm.) Barkworth, comb. nov. ined. - P; 3; FOS; 405. Elymus trachycaulus (Link) Gould ex Shinners ssp. trachycaulus - P; 3; DI, RO, FTS; 722, 748, 820. Eragrostis lutescens Scribn. - A; 4; MA, DR, SA; 12651282, Eragrostis mexicana (Hornem.) Link ssp. mexicana - A; 4; VH, BS, MA; 1124, 1305. Eragrostis pectinacea (Michx.) Nees ex Steud. var. pectinacea - A; 3; BS, RA, DR, RO, DA; 957, 1052, 1139. Festuca arizonica Vasey - P; 5; HE, RO; 1019. Hesperostipa comata (Trin. & Rupr.) Barkworth ssp. comata - P; 3; FOS, CS, RA; 393, 682. Hesperostipa neomexicana (Thurb. ex Coult.) Bark- worth - P; 5; RA, LE; 1030. Hordeum jubatum L. ssp. jubatum - P; 3; MA, TA, DA, RO; 681, 695, 716, 1119. * Hordeum murinum L. ssp. murinum - A; 5; DA, WA; 694. Koeleria macrantha (Ledeb.) J.A. Schultes - P; 3; RA, CAs 642,.6/2,.0/3. Lolium pratense (Huds.) S.J. Darbyshire - P; 5; RO; 1023. Lycurus setosus (Nutt.) C.G. Reeder - P; 5; RA; 1007. Monroa squarrosa (Nutt.) Torr. - A; 4; FOS, BS, DA; 1148. Muhlenbergia depauperata Scribn. - A; 5; LE, RA, BS; 1270, 1301, [Vol. 55 Muhlenbergia minutissima (Steud.) Swallen - A; 3; VH, BS... CS; 1095, 12352, Muhlenbergia montana (Nutt.) A.S. Hitche. - P; 3; RA, DR, HE; 356, 1105, 1187. Muhlenbergia pauciflora Buckl. - P; 0; RA, CS; B. Chaney s.n. Muhlenbergia porteri Scribn. ex Beal - P; 0; RA, LE; B. Chaney s.n. Muhlenbergia racemosa (Michx.) B.S.P. - P; 4; RA, DR, SS, NFS; 951, 1224, 1275. Muhlenbergia ramulosa (Kunth) Swallen - A; 5; SS, CSL; HE: 1217. Muhlenbergia repens (J. Presl) A.S. Hitche. - P; 5; LE, FTS, DI; 1066. Muhlenbergia rigens (Benth.) A.S. Hitchc. - P; 3; DR; 953, 1113, 1144. Muhlenbergia torreyi (Kunth) A.S. Hitche. ex Bush - P; 4; FOS; 1057, 1074, 1112. Muhlenbergia wrightii Vasey ex Coult. - P; 3; RA, DR; 726, 752, 926, 1022, 1051. Panicum bulbosum Kunth - P; 5; MA, DR, RA; 1116. Panicum capillare L. - A; 4; MA, DR, WA; 956, 1155. Panicum hirticaule J. Presl var. hirticaule - A; 5; SS, CI; 1220. Panicum mohavense J. Reeder - A; 5; LE, RA, SSO, endemic to N. AZ and C. NM; new record for Coconino Co.; 1300. Panicum virgatum L. var. virgatum - P; 3; RA, DR; 781, 966, 1021, 1276. Pascopyrum smithii (Rydb.) A. Love - P; 4; RO, DI, DA; 1055. Piptatherum micranthum (Trin. & Rupr.) Barkworth - P; 3; RA; 347, 411, 675. Pleuraphis jamesii Torr. - P; 3; FOS, LE; 327. Poa fendleriana (Steud.) Vasey - P; 3; VH, RA, DR, HE; 538, 548, 555. Poa pratensis L. ssp. pratensis - P; 4; MA, DA, TA; 696. * Psathyrostachys juncea (Fisch.) Nevski - P; 5; FOS, DA; 750, 1150. Schedonnardus paniculatus (Nutt.) Trel. - P; 3; FOS, FTS; 404, 652, 1065, 1081, 1137, 1263. Schizachyrium scoparium (Michx.) Nash var. scopar- ium - P; 4; DR, CA, RA; 954,. * Secale cereale L. - A; 4; DA, RO; 756. * Setaria viridis (L.) Beauv. var. viridis - A; 4; RA, DR, RO, DI; 1277, 1311. | Sorghastrum nutans (L.) Nash - P; 4; RA, DR; 967, | 1289. | Sporobolus airoides (Torr.) Torr. - P; 5; RO, LE; | 1020. | Sporobolus contractus A.S. Hitche. - P; 4; VA, LE, , RO; 949, 1246, 1316. Sporobolus cryptandrus (Torr.) Gray - P; 3; WH, DR; | 508, 864, 928, 1009. * Thinopyrum intermedium (Host) Barkworth & D.R. | Dewey - P; 3; RO, DI, DA; 824, 1056. Vulpia octoflora (Walt.) Rydb. var. hirtella (Piper) | Henr. - A; 5; LE, FOS, FTS; 636. Potamogetonaceae Potamogeton pusillus L. ssp. tenuissimus (Mert. & | Koch) Haynes & C.B. Hellquist - P; 5; WA; 1279. Typhaceae Typha latifolia L. - P; 5; WA; 860. MADRONO, Vol. 55, No. 1, pp. 15-25, 2008 A LATE HOLOCENE RECORD OF VEGETATION AND CLIMATE FROM A SMALL WETLAND IN SHASTA COUNTY, CALIFORNIA R. ScotrT ANDERSON!, SUSAN J. SMITH’, RENATA B. JASS*** AND W. GEOFFREY SPAULDING® 'Center for Environmental Sciences & Education, & Quaternary Sciences Program, Northern Arizona University, Flagstaff, AZ 86011; Scott.Anderson@nau.edu > Bilby Research Center, Northern Arizona University, Flagstaff, AZ 86011 7>CH2M Hill, Inc., 2285 Corporate Circle, Suite 200, Henderson, NV 89074 ABSTRACT A long-term history of water table fluctuations, from alternating periods of drought and abundant precipitation, can be preserved in the stratigraphy of wetland sediments. We examined the middle to late Holocene history of vegetation and climate change from a small wetland on the Modoc Plateau in Shasta County, northeastern California. This site is at a transition between the Great Basin and the Californian Floristic Provinces, and the paleoecological record from Flycatcher Basin exhibits affinities to both. Although the sedimentary record extends back to ca. 8300 cal yr BP, organic sediment did not form until ca. 4500 cal yr BP, indicating that water was probably absent in the basin during the middle Holocene. Pollen and plant macrofossils deposited after 4500 cal yr BP suggests a mixed conifer — Quercus forest grew around Flycatcher Basin. Charcoal is abundant in these sediments, indicating periodic forest fire. Distinctly modern forests developed by about 2200 cal yr BP, when Pinus became the dominant conifer with Quercus, in a more closed forest, perhaps with more frequent fire. The record from Flycatcher Basin provides no evidence for change in the boundaries between the Great Basin and California (Cascadian) floristic provinces during the period of record. The late Holocene is interpreted as a generally increasingly mesic sequence, with a long-term increase in groundwater recharge, yet interspersed by extended drought during the last 2000 yr. Extended droughts occurred from ca. 1125 AD to 1450 AD, with an earlier protracted dry period from ca. 100 AD to ca. 900 AD. Generally wetter periods occur from ca. 1000 to 1125 AD, and after ca. 1450 AD. The paleoenvironmental changes in the Flycatcher Basin wetland are a local expression of a much broader climatic pattern, as shown by several studies of higher resolution proxies. The record from Flycatcher Basin wetland is important in demonstrating the centennial to millennial-scale fluctuations in water availability in a region of rapidly expanding human population, with an increasing need for water resources. Key Words: California, drought reconstruction, fire history, paleoecology, pollen analysis, wetland hydrology, Modoc Plateau. The Modoc Plateau of northeastern California is a region of steep vegetation gradients from interior basin to upland forest. Located at the transition between the Great Basin, the southern extension of the Cascade Range and the northern extent of the Sierra Nevada, the Plateau is presently home to vegetation types allied to Sierran — Cascade coniferous forests, as well as to Great Basin sagebrush steppe (Kuchler 1988). To the west, on the flanks of the Cascade Range, are the Sierran montane forests, which include ponderosa and Jeffrey pine (Pinus ponderosa Laws., P. jeffreyi Grev. & Balf.), white fir (Abies concolor (Gordon & Glend.) Lindley)), California black oak (Quercus kelloggii Newb.), and western juniper (Juniperus occidentalis Hook.), among others (Rundel et al. 1988). At somewhat lower elevations, Oregon oak (Q. garryana Hook.) woodland dominates. South-facing xeric slopes *Present Address: 4014A Lewis Lane, Austin, TX 78756. support gray pine (P. sabiniana Douglas) and chaparral, including buckbrush (Ceanothus pros- tratus Benth.) and manzanita (Arctostaphylos spp.). To the east is Great Basin vegetation, dominated by sagebrush (Artemisia tridentata Nutt.) and Juniperus occidentalis, with additional shrubby perennials. Occurring within this transitional region, less than 10 km east of the Great Basin, is a series of small ephemeral wetlands, known informally as the Flycatcher Basins (Fig. 1). The small wet- lands lie within a canyon created by the Pit River’s course through the southern Cascade Range. The Flycatcher Basin wetlands occupy hollows in jumbled terrain of probable landslide origin (Page 1995), are unusual in this relatively dry environment, and preserve a paleoenviron- mental record of mid- through late Holocene. This is important since our knowledge of former environments of this region is limited to just a few sites. Prior to 1990, only three studies had been conducted on or near the Modoc Plateau, including Klamath Lake (Hansen 1942), Tulelake 16 MADRONO [Vol. 55 Flycatcher Basin se 30 Meters 619000 620000 621000 Aerial Photograph % 4 ah Trench 4544100 Nominal Center of Basin Perimeter of Peat High Shoreline Benchmark 4543400 4542700 622000 4542000 » Flycatcher Basin & 4542000 4541300 4541300 3 EE % California, U.S. | 8 43 5 » Flycatcher Basin | 2 “_ es & % = rr 617000 618000 Projection: Universal Transverse Mercator (UTM) Zone 10 N Ss Datum: North American Datum (NAD) of 1983 Ss 0 100 200 Kilometers Fic. 1. Location of Flycatcher Basin, Shasta County, northeastern California. The center image clearly shows the | slump terrain along the Pit River that produced the basins. The small inset in the upper left shows the basin itself and coring site. The inset in the lower right shows sites mentioned the text (1 = Little Willow Lake [West 2004]; 2 = | McCoy Flat [Wigand and Rhode 2002]; 3 = Tulelake [Adam et al. 1989]; 4 = Lower Klamath Lake [Hansen 1942]; | 5 = Medicine Lake [Starratt et al. 2003]; 6 = Klamath Mountains lakes [Mohr et al. 2000; Daniels et al. 2005]; 7 = Sierra Nevada sites [Anderson 1990; Smith and Anderson 1992; Anderson and Smith 1994; Edlund 1996; Brunelle and Anderson 2003; Potito et al. 2006]; 8 = Southern Oregon lakes [Bradbury et al. 2004; Hakala and Adam 2004)]). | 2008] (Adam et al. 1989), and Diamond Craters (Mehringer and Wigand 1990). More recently, additional long records of vegetation change have come from Upper Klamath and Grass Lakes in southern Oregon (Bradbury et al. 2004; Hakala and Adam 2004), Medicine Lake (ca. 50 km north of Mt. Shasta; Starratt et al. 2003), Little Willow Lake (Lassen National Park; West 2004), and McCoy Flat (eastern fringe of Modoc Plateau; Wigand and Rhode 2002). To the west, Mohr et al. (2000) and Daniels et al. (2005) determined postglacial vegetation and fire history from the Klamath Mountains, while several studies come from the Sierra Nevada to the south (e.g., Anderson 1990; Smith and Anderson 1992: Anderson and Smith 1994; Edlund 1996; Brunelle and Anderson 2003; Potito et al. 2006). Though the Klamath Lake record appears to be largely Holocene, radiocarbon dates are absent and its exact chronology is unknown. The Tulelake record encompasses much of the last three million years, but few Holocene samples were analyzed. Thus additional records from the region are needed to understand past vegetation changes that have occurred here. Recent studies of the late Holocene of the region using multiple proxies point to periodic, but widespread and persistent dry conditions over the last 2000 yr that are greater in magni- tude than those experienced during the historic period (e.g., the 1930’s and 1950’s droughts [Schubert et al. 2004]). These include lower lake levels (Stine 1994) and reduced runoff (Benson et al. 2002), higher fire frequencies (Mohr et al. 2000; Brunelle and Anderson 2003; Daniels et al. 2005), reduced tree growth (Hughes and Graum- lich 1996), and lowered treelines (Lloyd and Graumlich 1997) in and near the California mountains. These droughts are a regional expres- sion of widespread periodic drought in western North America (e.g., Grissino-Mayer 1996; Laird et al. 1996; deMenocal 2001; Fye et al. 2003; Mason et al. 2004) during the late Holocene. The Site Flycatcher Basin is one of several small closed basins supporting wetlands, located within a large landslide known as the Flycatcher Embay- ment, ca. 1.2 km northeast of the Pit River, in Shasta County, California (Fig. 1). Bedrock along the Pit River here is Pliocene diatomite overlain by basalt. The diatomite is a structurally weak rock prone to landslides (Page 1995). The basin is located at ca. 924 m elevation and has a small drainage area within the local forest with no tributary streams or channels. Although dry during the period of excavation, a cracked peat- mud floor, pond turtle (Chlemys marmorata) bones, and a wave-cut shoreline indicated that it recently held water to a depth of at least 1.1 m. ANDERSON ET AL.: FLYCATCHER BASIN PALEOECOLOGY 17 Uplands surrounding the basin are typically closed-canopy Oregon oak (Q. garryana) wood- land with Pinus ponderosa. Dense P. ponderosa forest occurs on the north-facing slope immedi- ately south of the basin. P. sabiniana and Ceanothus prostratus occur within an open Quercus woodland on the adjacent south-facing slope. Other trees and shrubs occurring locally include rare Douglas-fir (Pseudotsuga menziesii (Mirbel) Franco), scattered Juniperus occidenta- lis, wild rose (Rosa sp.), waxberry (Symphoricar- pos sp.), squawberry (Rhus trilobata Torry & A. Gray), buck brush (Ceanothus cuneatus (Hook.) Nutt.), and antelope bush (Purshia tridentata (Pursh) DC). A variety of herbs and subshrubs, including grasses (Poaceae), composites (Achillea millefolium L., Artemisia cf. dracunculus L., Cirsium sp., Xanthium strumarium L.), Epilobium brachycarpum C. Presl and others occur within the basin. Woody phreatophytes such as willow (Salix sp.) and alder (Alnus viridis (Chaix) DC ssp. sinuata (Regel) A. Love & D. Love) do not occur along the margins of the basin, but are found along the Pit River to the south. Plants growing within the basin itself include sedges (Cyperaceae), dock (Rumex sp.), mullein (Ver- bascum thapsus L.), and plantain (Plantago sp.), with patches of tule (Scirpus sp.) and rushes (Juncus sp.) in wetter areas. METHODS Sediments from Flycatcher Basin were collected during September 1992. A 1.5 * 2 m trench was hand excavated to a depth of 2.1 m near the middle of the desiccated basin (Fig. 1). The top 40.5 cm of sediment was retrieved in consecutive blocks of 4 to 22 cm length. Below this a 10-cm diameter pipe with a sharpened end was driven vertically into the exposed section to 170.5 cm depth. From 170.5 to 208.5 cm blocks of sediment were cut from the exposure in the same manner as the top 40.5 cm. The multiple methods were used to obtain as much of the sedimentary column intact as possible. Until analyzed in 1994, these sediments were kept in cold storage at the Laboratory of Paleoecology (LOP), Northern Arizona University. For pollen and charcoal analysis, 42 subsam- ples (2 cm® each) were taken at intervals of 2 to 15 cm. Three additional surface pollen samples were obtained from deposits elsewhere in the basin. Pollen and charcoal were extracted using a Fegri and Iversen (1989) chemical separation technique, additionally treated with sodium pyrophosphate and sieved through an 8-um screen following acetolysis. Pollen preservation was variable, so the number of pollen grains identified per sample was not uniform (see below). We analyzed the microscopic charcoal fraction here, since macroscopic charcoal oc- curred less commonly in the record. 8 MADRONO [Vol. 55 TABLE 1. RADIOCARBON AND CALENDAR AGES FOR THE FLYCATCHER BASIN SEDIMENTS. * median probability age Laboratory number Depth (cm) Age ('4C yr BP) Age (cal yr BP)* Age (AD/BC)* Beta-64384 15—20 430+ 60 473 1477 AD Beta-62111 40.4—S0.5 1240+70 1159 791 AD Beta-68 120 65—70 2080+ 80 2054 105 BC Beta-62112 125-135 3580+ 80 3876 1927 BC Beta-62310 192—208.5 7420+ 170 S217 6268 BC Pollen types were identified by comparison to the reference collection at the LOP. Most grains were identified to genus (e.g., Pinus), less often to species or species-type. Some pollen types can only be distinguished reliably to family (e.g., Asteraceae, Apiaceae, Poaceae). Though most of the Cupressaceae pollen was undoubtedly Juni- perus, we retain the designation of the family. All of the Polygonum pollen belonged to an aquatic species (1.e., P. amphibium L.). We used the TGView 2.0 program (Grimm 1993) to calculate sums and graph data (the pollen data are available from the North American Pollen Data- base (NAPD. [http://www.ngdc.noaa.gov/paleo/ napd.html]), but pollen zones were created by visual inspection. In order to highlight potential hydrological changes in the basin we calculated a ratio of pollen percentage of “‘moist” taxa to a sum of “moist” and “‘dry”’ pollen types, such as: (Polygonum amphibium + Alnus + Potamogeton + Cyperaceae) / (Polygonum amphibium + Alnus + Potamogeton + Cyperaceae) + (Asteraceae + Cheno — am). Charcoal fragments were identified on pollen slides, and the area of each individual charcoal particle was measured, then the total area (mm7) of charcoal was calculated for each sediment level analyzed. Plant macrofossils were recovered by extract- ing subsamples (ca. 300 cm’*® each) from the sediments at ca. 10-cm intervals. The volume of each subsample was estimated by the amount of water displaced in a beaker. Sediments were disaggregated in water, and sieved by gentle water washing. Each fraction was hand-sorted and picked for plant and animal macrofossils. Macrofossils were compared with the LOP reference collection. RESULTS Radiocarbon Dates and Sediment Accumulation Rates Five bulk sediment radiocarbon dates were obtained (Table 1), and radiocarbon ages were converted to calendar ages (cal yr BP and AD/ BC) using CALIB 5.0.2 (Stuiver et al. 1998). For age-depth construction, we chose the median probability age (Telford et al. 2004) from the output of CALIB 5.0.2, with a linear interpola- tion between ages. Sediment accumulation rate (SAR) was slowest (0.02 cm/yr) from the profile bottom to ca. 130cm depth (Fig. 2). Above 130 cm, SAR varied from 0.03 to 0.04 cm/yr. These SAR’s are comparable to those obtained from other lakes in California (Anderson 1990), despite the fact that this is a small basin with no tributaries. Sediment Stratigraphy The section top from 0 to 10 cm is composed of dry, largely unconsolidated, dark gray-brown silty peat (Fig. 2). From 10—110 cm sediments are compact, prismatic, dark brown clay with inter- spersed flecks of charcoal. These sediments contain very little organic material or sand. At ca. 110 cm depth, this fine-grained sediment grades gradually into a reddish-brown, silty- sandy clay with occasional gravel inclusions. Increasing amounts of sand were found below 122 cm. Below 175 cm sediments are sandy, brown, friable clays. Modern Pollen Assemblages Three pollen samples provide modern pollen | assemblages for comparison with the fossil _ pollen. Samples were taken from a patch of. Scirpus, near a former shoreline of the pond, and | from the top of a trench dug in the southern > portion of the basin (Fig. 3). Each sample was_ dominated by Pinus, varying from ca. 55-75% of | the pollen sum. Other important pollen types include Cupressaceae (probably Juniperus), Quer- cus, and Abies. Cupressaceae appears over- | represented based on its near-absence from the local vegetation, and Quercus appears under- | represented compared to its dominance in the local woodland. Very small amounts (ca. 5%) of | Artemisia were also recovered. The occurrence of both Abies and Artemisia in sediments suggests | pollen sources from both the west (the dominant wind direction resulting in the transport of Abies pollen from the southern Cascade Range) and the | east (the direction of Artemisia-dominated Great 2008] Age (cal yr BP x 10°) 0 Z 4 6 8 0 | | =I | | [es | 50 i E — < 100 — ja a 150 200 +4 250: —! B38 Gray-brown silty clay Ee] Transition to below Gray-brown silty peat fe] Reddish-brown silty- sandy clay Fic. 2. Age-Depth curve for the five radiocarbon dates obtained from the Flycatcher Basin sediments, along with a sediment description of the profile. Ages are given in Table 1. The age model is based upon a linear interpretation of the median probability age for the sample. Basin steppe vegetation). The occurrence of dwarf mistletoe (Arceuthobium) in two samples is consistent with the presence of a dense Pinus ponderosa forest on the south side of the basin, since pollen of this conifer parasite is not widely dispersed (Anderson and Davis 1988). Though the basin was dry at the time of fieldwork, pollen of disturbed ground plants (e.g., Cheno-Ams — only a single grain recovered from the four samples) was very rare. Pollen of damp-ground or aquatic plants (Brassicaceae, Polygonum am- phibium, Thalictrum, Alnus, Potamogeton, and Cyperaceae) was consistently found, indicating recent basin wetting. Spores of quillwort (/soetes) and aquatic algae were recovered from the south trench and Scirpus sites, but not from the old shoreline. This is probably a reflection of the greater time that water covered the central portion of the basin, suggesting that the old Shoreline is an ephemeral feature, or that considerable time has passed since the highest water level was attained. Fossil Pollen and Macrofossil Assemblages Forty-two fossil pollen samples were analyzed from Flycatcher Basin, representing an average of 195 yr between samples for the entire record, but 105 yr between samples for the late Holocene (last 4500 yr) only. We identified 31 pollen and spore types. Pollen preservation varied through- out the core. Preservation was very good in the modern surface samples and in the upper 40 cm of the core, where deteriorated grains averaged less than 10%. Below this to ca. 140 cm, ANDERSON ET AL.: FLYCATCHER BASIN PALEOECOLOGY 19 Flycatcher Basin, CA, Modern Pollen @ & @ & cas rod mo P oO So 0 S ~~ NS As ee (e) Wo © GGL SA SS Some SP AP WRN, Pel O Fk YH ORG S. Trench Old Shore Scripus Top Sect. 20 40 60 80 20 Fic. 3. Dominant pollen types in the modern pollen samples from Flycatcher Basin. Samples were obtained from four locations in the modern basin. deteriorated pollen comprised approximately 20 to 30% of the sum; pollen was essentially absent in sediments below 140 cm. We believe that the decline in pollen preservation with depth is probably related to greater incidence of drying of the sediment in older portions of the core, as evidenced by the oxidized appearance of the sediment, and as discussed below. As expected, pollen concentrations were greatest in the modern samples and near the core top, and declined with depth (Fig. 4). Vegetation and climatic changes are interpret- ed from the terrestrial pollen percentage (Fig. 4) and wetland and aquatic remains (Fig. 5) dia- grams. Pollen and macrofossils were essentially absent in the bottom 50—60 cm of the core. Pollen was well-preserved in sediments deposited subse- quent to approximately 4500 cal yr BP. However, pollen concentrations and macrofossil abundanc- es increased dramatically in sediments deposited over the last ca. 2000 yr (Figs. 4 and 5). Dominant pollen types were those common in the modern surface samples, including Pinus, Cupressaceae, Quercus, Abies and Asteraceae. Small amounts of Artemisia were found in virtually all samples. The macrofossil samples contained ca. 25 plant and animal taxa. Most taxa were aquatic or semi-aquatic; of terrestrial species, only the remains of Pinus ponderosa, Arctostaphylos cf. patula, and seeds of a Rosaceae were recovered. Pollen Zone FB-I (ca. 8300 to ca. 4500 cal yr BP). Pollen and macrofossils were essentially absent from this zone (Fig. 4), and charcoal concentrations were also low. Sediments were reddish-brown, indicating persistent exposure and greater oxidation compared to overlying sediments. Oxidation promotes intense decom- 20 MADRONO [Vol. 55 Flycatcher Basin, CA, Pollen Percentages cw Xe) xs x? oe we S ¥ < g & RG 2 x &< 2) 5 we PF SO Vv ao & o& we R Rey > xo & OR A YY’ Ona: & — 80-250 } Bw; 10 300 ao FB-II € 120; 0 > e240 350 ro) ” 2 ~ 400! pas = o a | Saee = 8 B 140) D @ 140) > 450 o ie a Oo -_ O 500 2 550 1 16 600 FB-I 650 1 120000 20 40 20 40 60 20 20 40 20 Fic. 4. Pollen concentration, proportion of degraded pollen, and terrestrial pollen percentages for common | terrestrial pollen types in the Flycatcher Basin sediments. Little pollen was recovered below 140 cm depth (presence only denoted by a dot). Silhouette = pollen percentage x 10. Flycatcher Basin, CA, Wetland & Aquatic Remains ~~) S £ S NO e) \ @ S & R oe ce oe eS eS NJ of ro} & xO re RAS AS S” x9 Zw sd OF PP _ Oe & es eg SEP PFO FF FS Fr LB | O \ @ : & rv 3 3F ENG e or i Sg sre xz oe e gs. ws ce Ss Ss a Ww Q° we we QF Q ey) co SS oe yY ww y} ey »s XO L TREAT a FB-I Ney wey Ww a a |, 180 7000 20 40 100 200 300 a Fic. 5. Occurrence of aquatic pollen (continuous curves) and plant and animal macrofossils (dots) from the’ Flycatcher Basin record. Silhouette = pollen percentage x 10. 2008] position, which probably accounts for the lack of preserved organics. Pollen Zone FB-II (ca. 4500 to ca. 2200 cal yr BP). Pollen percentages in this zone were dominated by Pinus, Cupressaceae (likely Juni- perus), and Asteraceae, with smaller amounts of Poaceae, Artemisia and Quercus. Poaceae, Cu- pressaceae and Asteraceae pollen declined toward the end of the zone. Degraded pollen was less than in Zone FB-I, averaging 20%. Pollen concentrations were lower than in overlying sediments (see below). The ratio of arboreal to non-arboreal pollen was generally low, indicating dominance by taxa such as Asteraceae, Poaceae, and Apiaceae. Charcoal was present, but the quantity was lower than in the subsequent zone. Macrofossils were entirely of aquatic taxa, dominated by Jsoetes and Chara, with occurrence of freshwater sponge spicules and Chrysophyte cysts (Fig. 5). Pollen Zone FB-III (ca. 2200 cal yr BP to present). Two subzones (FB-IIIa, ca. 2200 to 1100 cal yr BP; FB-IIIb, ca. 1100 cal yr BP to present) were identified. Pollen concentration for the entire zone increased fivefold over Zone FB- II. The percentage of degraded pollen declined toward the surface. Conifer pollen, particularly Pinus, increased during this zone and that of Arceuthobium was more common than below. Abies pollen was first recovered in FB-IIIb, while values of Poaceae and Asteraceae were lower in this subzone. Charcoal concentrations generally increase during FB-III over the previous zone. The richness of aquatic taxa increased sub- stantially in Zone FB-III. Types common in FB- II continued here. Pollen and macrofossils of Potamogeton, sedges (both Carex and Cyperus), and tule (Scirpus) became abundant early in subzone FB-IIIa. Other rooted, floating aquatics such as Najas cf. guadalupensis (Sprengel) Mag- nus, Brasenia schreberi J. Gmelin, along with Nitella, became common in FB-IIIb. Ephippia of Daphnia were abundant, indicating more persis- tent standing water. DISCUSSION The paleoecological record from Flycatcher Basin is important in understanding Holocene climatic events and their impact on the vegetation of the Modoc Plateau and vicinity, the stability of a major vegetation ecotone, and potentially the history of drought conditions there. Preservation anomalies are undoubtedly important in deter- mining the characteristics of the fossil assemblag- es, particularly during Zone I time (ca. 8200 to 4500 cal yr BP). The generally coarser clastic sediment deposited during this period may indicate greater slopewash at that time, and the reddish-brown color is consistent with strong ANDERSON ET AL.: FLYCATCHER BASIN PALEOECOLOGY 21 oxidation under subaerial conditions. This prob- ably accounts for the lack of organic preserva- tion. We interpret the mid-Holocene local envi- ronment of the basin as being a small, mostly dry depression. After ca. 4500 cal yr BP fossil preservation and concentration increases, and these fossils suggest a mixed conifer — Quercus forest grew around Flycatcher Basin. This forest may have been similar to the modern forest, but more open (greater Asteraceae, Poaceae) and with less Pinus than today. Abundant sedimentary charcoal (Fig. 4) suggests regular forest fire, though our technique cannot estimate any potential changes in fire episode frequencies over earlier times. Distinctly modern forests developed by ca. 2200 cal yr BP when Pinus became the dominant conifer, growing with Quercus, in a more closed forest, perhaps with more frequent fire (Fig. 4). The record does not show any marked long-term changes in the importance of sagebrush (Artemi- sia) pollen over the last 4500 yr, and therefore suggests little if any change in the position of the ecotone between Californian (Cascadian) and Great Basin floristic provinces. Wetland and aquatic species, including perhaps Alnus, became established within and near the site at this time also. A/nus does not presently exist in the basin, although it does grow along the Pit River. Because the modern surface samples (Fig. 3) yield amounts of A/nus pollen compara- ble to those found in samples from the strati- graphic column (Fig. 4) it is possible that these grains are not locally derived. The occurrence of TIsoetes, Chara, Chrysophyte cysts and others indicate that water must have covered portions of the basin for much of the year, but a lack of floating aquatics suggests shallow water depths. Pollen evidence from other sites within the region also suggests warmer and/or drier climates during the middle Holocene than today, which could have led to a regional lowering of water tables. During the middle Holocene, Pinus, Quer- cus, and Cupressaceae dominated forests of the Klamath (West 1989; Mohr et al. 2000; Daniels et al. 2005) and mid-elevations of the Sierra Nevada (Smith and Anderson 1992; Anderson and Smith 1994; Edlund 1996) regions. Chironomid-inferred temperatures suggest that the mid-Holocene was the warmest period in the Sierra Nevada (Potito et al. 2006). Although individual characteristics prevailed at each site, after ca. 4500-5500 cal yr BP the more mesic species Abies increased at all sites. Further north in southern Oregon, the middle Holocene at Diamond Pond was dry prior to 6200 cal yr BP, but after this groundwater tables rose, forming a permanent lake (Wigand 1987). These data are consistent with a regional increase in effective annual precipitation, leading to higher groundwater tables, and establishment of persis- tent wetlands. O91 MADRONO Anderson et al. (this study) Stine (1990) (Mono Lake) Stine (1994) (Mono Lake) Yuan et al. (2004) (Walker Lake) No record ~ Mensing et al. (2004) Lloyd & Graumlich (1997) . . warm? Southern Sierra Treeline Lloyd (1997) Sequoia No record ¢ National Park Treeline Y Meko et al. (2001) Sacramento River Flow No record ¢ 1750 1550 1350 FIG. 6. Treeline , [Vol. 55 YEAR AD lower lower Treeline 4 Treeline 4 dry cold Death > Births Birth > Deaths dry wetter Driest 20 year x f Driest 50 year period period —_ 1150 950 750 Approximate CAL YR BP Comparison of climate proxies for the last 2000 yr: a. Calculation of relative wet and dry periods using pollen function (Polygonum amphibium + Alnus + Potamogeton + Cyperaceae)/(Polygonum amphibium + Alnus + Potamogeton + Cyperaceae + Asteraceae + Cheno-am) (this study); b. Calculated elevation (m) of Mono Lake, California, levels (Stine 1990); c. Calibrated radiocarbon ages of submerged tree stumps at Mono Lake (Stine 1994): d. Period of lowest lake levels at Walker Lake, Nevada (Yuan et al. 2004); e. Calibrated ages of lowered lake levels at Pyramid Lake, Nevada (Mensing et al. 2004); f. Treeline fluctuations in the southern Sierra Nevada, California (Lloyd and Graumlich 1997); g. Relationship between tree death and germination at treeline in Sequoia National Park, California (Lloyd 1997); and h. Flow of the Sacramento River, California, from tree ring evidence (Meko et al. 2001). The Latest Holocene The development of modern conditions oc- curred during the last 2200 yr. Pinus increased within the area, while Quercus remained impor- tant and Juniperus declined. Indicators of open- ground such as taxa of Poaceae and Asteraceae also declined. The first occurrence of Abies pollen probably represents expansion of A. concolor in the Cascades to the west (Mohr et al. 2000; Daniels et al. 2005). Research here and from the Sierra Nevada to the south (Davis et al. 1985; Anderson 1990; Anderson and Smith 1994; Edlund 1996) established a continuous, incre- mental expansion of Abies during the late Holocene, suggesting a long-term increase in effective precipitation and soil moisture. During this period, the aquatic flora at Flycatcher Basin became particularly rich, indicating that the basin held permanent water, and that water depths were sufficient in most years to support floating aquatic plants, especially after ca. 1100 yr ago (Fig. 5). However, using our equation to compare wet and dry pollen indicators over the last 2000 yr (see methods), our analysis suggests wet condi- tions periodically alternated with more arid climates during the latest Holocene (Fig. 6a). Although the pollen sampling interval varies over this time period, the ratio suggests generally dry: conditions from ca. 1125 to 1450 AD, with another possible protracted dry period from ca. 100 to ca. 900 AD. Generally wetter periods occur from ca. 1000 to 1125 AD, and after ca. 1450 AD. | This reconstruction is very similar to the lake- level reconstruction for Mono Lake over the last. 2000 yr (Fig. 6b; also Stine 1990). Lake levels at 2008] Mono Lake are controlled by runoff from the Sierra Nevada, and were generally low prior to 1000 AD, peaked by 1100 AD, and declined between ca. 1150 and 1450 AD. After ca. 1450 AD, lake levels fluctuated, but generally trended upward. This latter period may have been driven by cooler temperatures and greater precipitation during the Little Ice Age, dated in the Sierra Nevada as after ca. 650 cal BP (Matthes glaciation; see references in Clark and Gillespie 11997). We also compared the Flycatcher Basin record with several higher resolution records of climate variation for northeastern California and the Sierra Nevada, and for western Nevada (Fig. 6c— h.). The comparisons are particularly striking for the last ca. 1000 yr, with fewer data available prior to that. For instance, in a study of runoff of the Sacramento River in California, Meko et al. (2001) determined the driest 50-year and 20-year periods to be from 1140 to 1160 AD, and 1350 to 1400 AD, respectively. Similarly, numerous dates on submerged stumps from Mono Lake (1025 to 1250 and 1275 to 1400 AD; Stine 1994) suggest lowest lake levels then. Yuan et al. (2004) determined from 60" analysis that the driest late Holocene period at Walker Lake, Nevada, was from 1200 to 1360 AD. At Pyramid Lake, Nevada, Mensing et al. (2004) determined major dry periods at ca. 1150 to 1225 and 1350 to 1500 AD. Each of these reconstructions suggests particularly dry conditions contemporaneously with the 12th to 15th century dry period at Flycatcher Basin. Tree-ring and stand-age data from high elevation foxtail pine (Pinus balfouriana (Grev. & Balf.)) stands in the Sierra Nevada also confirm the Flycatcher Basin interpretations. Decline in upper elevation treeline due to drought occurred from ca. 1000 to 1400 AD (Lloyd and Graumlich 1997), with maximum drought be- tween ca. 1250 to 1400 AD (Graumlich 1993), while tree deaths exceeded births from 1100 to 1450 AD (Lloyd 1997). However, contempora- neously with our reconstruction of wetter condi- tions at Flycatcher Basin after ca. 1450 AD, treeline data suggest cooler and wetter conditions | in the Sierra Nevada (Fig. 6f. and g.). CONCLUSIONS Occurring at the transition between the Great Basin and the Californian-Cascadian Floristic Provinces, the paleoecological record from Fly- catcher Basin exhibits affinities to both. Antevs (1948) defined a paleoclimatic sequence for the Great Basin as consisting of three distinct stages: the Anathermal (ca. 11,000~-7000 yr ago; early Holocene), the Altithermal (ca. 7000-4500 yr ago; middle Holocene), and the Medithermal (ca. 4500 yr ago to present; late Holocene). Based ANDERSON ET AL.: FLYCATCHER BASIN PALEOECOLOGY 23 upon the presence or absence of lakes in the Great Basin, this sequence defined a regime of effective precipitation changing from subhumid to arid conditions, followed by semiarid climates. Subsequent research has confirmed this sequence. The middle Holocene was warmer and drier than the early Holocene (Wigand and Rhode 2002; Tausch et al. 2004); Lake Tahoe was below its rim during this period (Lindstrom 1990; Benson et al. 2002), Pyramid Lake may have been at its lowest level (Born 1972), and Walker Lake was essentially dessicated (Bradbury et al. 1989). These conditions may have lasted until about 3500 yr BP in parts of the Great Basin (Tausch et al. 2004). But subsequent cooler temperatures, and increases 1n winter precipitation during the Neoglacial led to generally higher groundwater tables and greater persistence of lakes and wetlands in the western Great Basin (Wigand and Rhode 2002; Tausch et al. 2004). By way of contrast, studies of high elevation vegetation change to the south in the Sierra Nevada defined an early Holocene xerothermic (Davis et al. 1985; Davis and Moratto 1988; Anderson 1990; Brunelle and Anderson 2003) ending by about 7800—6800 cal yr BP, after which effective precipitation increased, as shown by expansion of Abies and other upper montane species. Intensification of cooling and increased effective precipitation occurred after ca. 4500 cal yr BP (Davis et al. 1985), while additional evidence for cooling at ca. 2600 and 900 cal yr BP is evident at several locations (Anderson 1990). But, vegetation sequences from middle elevation sites in the northern Sierra show no evidence for expansion of Abies until after ca. 3900 to 4500 cal BP (Edlund 1996). However, the origination of a semi-permanent wetland, as determined by aquatic macroscopic remains, in Flycatcher Basin after ca. 4500 cal yr BP initially is consistent with its position intermediate between the Great Basin and Sierran sequences. The record from Flycatcher Basin # 2, though covering only a portion of the Holocene, 1s important in our understanding of the late Holocene paleoenvironmental changes that have occurred in northeastern California. The ca. 4500-year pollen and plant macrofossil record is interpreted as a generally increasingly mesic sequence, with a long-term increase in ground- water recharge, yet interspersed by extended drought during the last 2000 yr. The paleoen- vironmental changes in the Flycatcher Basin wetland are a local expression of a much broader climatic pattern, as shown by several studies of higher resolution proxies. The Fly- catcher Basin demonstrates the long-term, cen- tennial-to-millennial fluctuations in water avail- ability in a region of rapidly expanding human population, with an increasing need for water resources. 24 MADRONO ACKNOWLEDGMENTS We thank Janet L. McVickar for essential assistance in the field and laboratory. We also thank Jamie Cleland, Liz Manion, and Mike Kelly, all of the former Dames & Moore, Inc., for assistance, and Pacific Gas & Electric for financial support for this project. We thank Frank Bayham for identification of the turtle bones from the site; Kirsten Ironside for Figure 1; Ron Redsteer for Figure 6; and Andrea Brunelle, Scott Mensing, John Hunter and an anonymous reviewer for very helpful reviews. Laboratory of Paleoecology Contribution # 90. LITERATURE CITED ADAM, D. P., A. M. SARNA-WOJCICKI, H. J. RIECK, J. P. BRADBURY, W. E. DEAN, AND R. M. FORESTER. 1989. Tulelake, California: the last 3 million years. Palaeogeography, Palaeoclimatol- ogy, Palaeoecology 72:89-103. ANDERSON, R. S. 1990. Holocene forest development and paleoclimates within the central Sierra Nevada, California. Journal of Ecology 78:470-489. AND O. K. DAVIS. 1988. Contemporary pollen rain across the central Sierra Nevada, California: relationship to modern vegetation types. Arctic and Alpine Research 20:448—460. AND S. J. SMITH. 1994. Paleoclimatic interpre- tations of meadow sediment and pollen stratigra- phies from California. Geology 22:723—726. ANTEVS, E. 1948. Climatic changes and pre-white man. Bulletin of the University of Utah 38:168—-191. BENSON, L., M. KASHGARIAN, R. RYE, S. LUND, F. PAILLET, J. SMOOT, C. KESTER, S. MENSING, D. MEKO, AND S. LINDSTROM. 2002. Holocene multi- decadal and multicentennial droughts affecting Northern California and Nevada. Quaternary Science Reviews 21:659-682. BoRN, S. M. 1972. Late Quaternary history, deltaic sedimentation, and mudlum formation at Pyramid Lake, Nevada. Desert Research Institute, Univer- sity of Nevada, Reno. BRADBURY, J. P., R. M. FORESTER, AND R. S. THOMPSON. 1989. Late Quaternary paleolimnology of Walker Lake, Nevada. Journal of Paleolimnol- ogy 1:249-267. , S. M. COLMAN, AND W. E. DEAN. 2004. Limnological and climatic environments at Upper Klamath Lake, Oregon during the past 45,000 yr. Journal of Paleolimnology 31:167—188. BRUNELLE, A. AND R. S. ANDERSON. 2003. Sedimen- tary charcoal as an indicator of late-Holocene drought in the Sierra Nevada, California, and its relevance to the future. The Holocene 13:21—28. CLARK, D. H. AND A. R. GILLESPIE. 1997. Timing and significance of late-glacial and Holocene cirque glaciation in the Sierra Nevada, California. Qua- ternary International 38/39:21—38. DANIELS, M. L., R. S. ANDERSON, AND C. WHITLOCK. 2005. Vegetation and fire history since the Late Pleistocene from the Trinity Mountains, north- western California, USA. The Holocene 15:1062—1071. DAvis, O. K., R. S. ANDERSON, P. FALL, M. K. O’ROURKE, AND R. S. THOMPSON. 1985. Palyno- logical evidence for early Holocene aridity in the southern Sierra Nevada, California. Quaternary Research 24:322—332. [Vol. 55 AND M. J. MORATTO. 1988. Evidence for a warm dry early Holocene in the western Sierra Nevada of California: pollen and plant macrofossil analysis of Dinkey and Exchequer Meadows. Madrono 35:132-149. DEMENOCAL, P. B. 2001. Cultural responses to climate change during the Late Holocene. Science 292:667-673. EDLUND, E. G. 1996. Late Quaternary environmental history of montane forests of the Sierra Nevada, California. Ph.D. dissertation. University of Cali- fornia, Berkeley, CA. FAGRI, K. AND J. IVERSEN. 1989. Textbook of Pollen Analysis. 4th edition. Chichester, New York, NY. FYE, F. K., D. W. STAHLE, AND E. R. Cook. 2003. Paleoclimatic analogs to twentieth-century mois- ture regimes across the United States. Bulletin of the American Meteorological Society 84(7): 901-909. GRAUMLICH, L. J. 1993. A 1000-year record of temperature and precipitation in the Sierra Ne- vada. Quaternary Research 39:249-255. GRIMM, E. 1993. Tilia and Tilia Graph versions 2.0.b.4. Illinois State Museum, Research and Collections Center, Springfield. GRISSINO-MAYER, H. D. 1996. A 2129-year recon- struction of precipitation for northwestern New Mexico, U.S.A. In J. S. Dean, D. M. Meko, and T. W. Swetnam (eds.), Tree Rings, Environment and Humanity. Radiocarbon 20:191—204. HAKALA, K. J. AND D. P. ADAM. 2004. Late Pleistocene vegetation and climate in the southern Cascade Range and the Modoc Plateau region. Journal of Paleolimnology 31:189-215. HANSEN, H. P. 1942. A pollen study of peat profiles from Lower Klamath Lake of Oregon and California. Pp. 103-114 in L. S. Cressman (ed.), Archaeological Researches in the Northern Great Basin. Carnegie Institution of Washington Publi- cation 538, Washington, DC. HUGHES, M. K. AND L. J. GRAUMLICH. 1996. Multimillennial dendroclimatic studies from the western United States. Pp. 109-124 in P. D. Jones, R. S. Bradley, and J. Jouzel (eds.), Climate Variations and Forcing Mechanisms of the Last 2000 yr. Springer-Verlag. KUCHLER, A. 1988. Terrestrial Vegetation of California (Map). California Native Plant Society, Special Publication 9, Sacramento, CA. LAIRD, K. R., S. C. FRITZ, K. A. MAASCH, AND B. F. CUMMING. 1996. Greater drought intensity and frequency before A.D. 1200 in the northern Great Plains, U.S.A. Nature 384:552—554. LINDSTROM, S. 1990. Submerged tree stumps as indicators of mid-Holocene aridity in the Lake Tahoe Basin. Journal of California and Great Basin Anthropology 12:146—157. | LLoyp, A. M. 1997. Response of tree-line populations | of foxtail pine (Pinus balfouriana) to climate variation over the last 1000 yr. Canadian Journal of Forest Research 27:936—942. | AND L. J. GRAUMLICH. 1997. Holocene dynamics of treeline forests in the Sierra Nevada. | Ecology 78:1199—1210. | MASON, J. A., J. B. SWINEHART, R. J. GOBLE, AND D. B. Loope. 2004. Late-Holocene dune activity linked to hydrological drought, Nebraska Sand | Hills, USA. The Holocene 14(2): 209-217. 2008] MEKo, D., M. THERREL, C. BAISAN, AND M. HUGHES. 2001. Sacramento River flow reconstructed to A.D. 869 from tree rings. Journal of the American Water Resources Association 37:1029—1039. MENSING, S. A., L. V. BENSON, M. KASHGARIAN, AND S. LUND. 2004. A Holocene pollen record of persistent droughts from Pyramid Lake, Nevada, USA. Quaternary Research 62:29—38. MEHRINGER, P. W., JR., AND P. E. WIGAND. 1990. Comparison of Late Holocene environments from woodrat middens and pollen: Diamond Craters, Oregon. Pp. 294-325 in J. L. Betancourt, T. R. Van Devender, and P. S. Martin (eds.), Packrat Middens: the Last 40,000 yr of Biotic Change. University of Arizona Press, Tucson, AZ. Monr, J. A., C. WHITLOCK, AND C. N. SKINNER. 2000. Postglacial vegetation and fire history, eastern Klamath Mountains, California, USA. The Holocene 10(4): 587-601. NAPD. NORTH AMERICAN POLLEN DATABASE. [http:// www.ngdc.noaa.gov/paleo/napd.html]). PAGE, W. D. (ed.). 1995. Quaternary geology along the boundary between the Modoc Plateau, southern Cascade Mountains, and northern Sierra Nevada. Friends of the Pleistocene, Pacific Cell. Pacific Gas & Electric Company, San Francisco, CA. PotTitTo, A. P., D. F. PORINCHU, G. M. MACDONALD, AND K. A. MOSER. 2006. A late Quaternary chironomid-inferred record from the Sierra Ne- vada, California, with connections to northeast Pacific sea surface temperatures. Quaternary Re- search 66:356—363. RUNDEL, P. W., D. J. PARSONS, AND D. T. GORDON. 1988. Montane and subalpine vegetation of the Sierra Nevada and Cascade Ranges. Pp. 559-600 in M. G. Barbour and J. Major (eds.), Terrestrial Vegetation of California. California Native Plant Society, Special Publication 9, Sacramento, CA. SCHUBERT, S. D., M. J. SUAREZ, P. J. PEGION, R. D. KOSTER, AND J. T. BACMEISTER. 2004. On the cause of the 1930’s dust bowl. Science 303:1855—59. SMITH, S. J. AND R. S. ANDERSON. 1992. Late Wisconsin paleoecologic record from Swamp Lake, Yosemite National Park, California. Quaternary Research 38:91—102. STARRATT, S. W., J. A. BARRON, T. KNEESHAW, R. L. PHILLIPS, J. L. BISCHOFF, J. B. LOWENSTERN, AND J. A. WANKET. 2003. A Holocene record from Medicine Lake, Siskiyou County, California: pre- liminary diatom, pollen, geochemical, and _ sedi- mentological data. Pp. 131-147 in G. J. West and N. L. Blomquist (eds.), Proceedings of the 19th Annual Pacific Climate Workshop. Interagency ANDERSON ET AL.: FLYCATCHER BASIN PALEOECOLOGY 25 Ecological Program for the San Francisco Estuary, Technical Report 71. STINE, S. 1990. Late Holocene fluctuations of Mono Lake, eastern California. Palaeogeography, Palaeo- climatology, Palaeoecology 78:333—381. 1994. Extreme and persistent drought in California and Patagonia during mediaeval time. Nature 316:530—532. STUIVER, M., P. J. REIMER, E. BARD, J. W. BECK, G. S. BURR, K. A. HUGHEN, B. KROMER, F. G. McCorRMAc, J. v.D. Plicht, AND M. SpuRK. 1998. INTCAL98 radiocarbon age calibration 24,000—0 cal BP. Radiocarbon 40:1041—1083. TAUSCH, R. J., C. L. NOWAK, AND S. A. MENSING. 2004. Climate change and associated vegetation dynamics during the Holocene: the paleoecological record. Pp. 24-48 in J. C. Chambers and J. R. Miller (eds.), Great Basin Riparian Areas, Ecology, Management and Restoration. Island Press, Wa- shington, DC. TELFORD, R. J., E. HEEGAARD, AND H. J. B. BIRKS. 2004. The intercept is a poor estimate of a calibrated radiocarbon age. The Holocene 14:296-298. WEST, G. J. 1989. Late Pleistocene/Holocene vegetation and climate. Pp. 36—55 in M. E. Basgall and W. R. Hildebrandt (eds.), Prehistory of the Sacramento River Canyon, Shasta County, California. Center for Archaeological Research, Davis, CA. . 2004. A late Pleistocene — Holocene pollen record of vegetation change from Little Willow Lake, Lassen Volcanic National Park, California. Pp. 65-80 in S. W. Starratt and N. L. Blomquist (eds.), Proceedings of the 20th Annual Pacific Climate Workshop. Interagency Ecological Pro- gram for the San Francisco Estuary, Technical Report 72. WIGAND, P. E. 1987. Diamond Pond, Harney County, Oregon: vegetation history and water table in the eastern Oregon desert. Great Basin Naturalist 47:427-458. AND D. RHODE. 2002. Great Basin vegetation history and aquatic systems: the last 150,000 yr. Pp. 309-368 in R. Hershler, D. B. Madsen, and D. R. Currey (eds.), Great Basin Aquatic Systems History. Smithsonian Contributions to the Earth Sciences, No. 33. Smithsonian Institution Press, Washington, D.C. YUAN, F., B. K. LINSLEY, S. LUND, AND J. P. McCGEFHIN. 2004. A 1200 yr record of hydrologic variability in the Sierra Nevada from sediments in Walker Lake. Geochemistry Geophysics Geosys- tems 5:1—13. MADRONO, Vol. 55, No. 1, pp. 26—40, 2008 LOCAL SCALE VEGETATION MAPPING AND ECOTONE ANALYSIS IN THE SOUTHERN COAST RANGE, CALIFORNIA ROBERT J. STEERS!, MICHAEL CURTO AND V. L. HOLLAND Biological Sciences Department, California Polytechnic State University, San Luis Obispo, CA 93407 ABSTRACT Plant communities in the southern Coast Range of California form a mosaic with discrete to gradual transitions between multiple vegetation types. To accurately portray this pattern and to quantify the areal coverage of ecotonal space, a new method of mapping vegetation was developed. Vegetation stands were classified and mapped in separate GIS layers to the full extent of their respective suite of indicator species. Since all stands were mapped in this way, the overlap of different communities in the GIS represents ecotonal space. Vegetation mapping was entirely ground-based using a GPS receiver. Vegetation classification followed the Holland and Keil scheme. Eleven plant communities were identified within the 92.6 ha study area. This mapping method revealed that 32% of the total area was ecotonal and that the majority of plant communities exhibited a greater portion of their total area as ecotone than as discrete space. This finding suggests that typical vegetation maps depicting discrete boundaries between all vegetation types may misrepresent a nontrivial proportion of the area mapped. In addition, because ecotones are ecologically significant and important to conservation, the portrayal of transitional space between communities is worth consideration in the future creation of vegetation maps within California. Key Words: ecotone, full extent, fuzzy boundary, multi-layer mapping, semi-stand, serpentine, vegetation classification, vegetation map. The study of ecological boundaries has played an important role in developing the field of ecology. Research in this topic 1s diverse and has ranged from exploring small-scale boundaries at the root soil interface (Belnap et al. 2003) to large-scale boundaries across continents (Thomp- son et al. 2005; Peinado et al 2007). One of the most common terms used to express ecological boundaries is the ecotone. The liberal usage of ecotone in the literature has spurred many attempts at reclassification and introduction of new vocabulary (Kent et al. 1997; Holland 1988; Strayer et al. 2003). For the purposes of this study, ecotone is the transition between adjacent plant communities, as first defined by Clements (1905). Because ecotones are the product of adjacent plant communities, the plant community concept is central to the concept of the ecotone (Kent et al. 1997). While Gleason (1926) used ecotones as part of his argument against the existence of plant communities and while ongoing debate over plant community concepts still exist (reviewed in Tansley 1920; Austin 1985; Mucina 1997), most vegetation scientists at least acknowledge the usefulness of recognizing plant communities and the narrow to broad transitional zones between them (Barbour et al. 1999). Ecotones may be the result of various phenomena (Lloyd et ‘Current address: Department of Botany and Plant Sciences, University of California, Riverside, CA 92521, Email: rstee001@ucr.edu al. 2000) such as anthropogenic and natural disturbances (Cadenasso et al. 2003), abiotic and biotic environmental gradients (Walker et al. 2003), and biological invasion fronts (Hoffman et al. 2004). Much has been learned about basic ecology through the investigation of ecotones (Austin 1985; Gosz 1993; Smith et al. 1997; Kark and van © Rensburg 2006). Nevertheless, these transitional spaces have often been overlooked by vegetation scientists who tend to focus on discrete, repeat- — able vegetation types (Risser 1995; Mucina 1997). This focus on discrete vegetation is also reflected | in the field of vegetation mapping (Kuchler | 1988a; Goodchild 1994). Nearly all paper or | digital vegetation maps depict two-dimensional orthographic canopy cover with complete cover- — age by non-overlapping polygons. An obvious | drawback of this typical approach is that vegetation stand boundaries are depicted more | discretely than they actually are in the field. As a> result, information about the extent and compo- | sition of ecotones is lost and this renders the | vegetation map less accurate. The goal of this study was to describe and | analyze the plant communities and ecotones of a | nature reserve in Poly Canyon, located in the | southern Coast Range of California (Fig. 1), | through the use of a high-resolution, multi-layer approach to local ground-based vegetation map- ping. This detailed approach resulted in several noteworthy mapping unit categories: those being » full extent, discrete, ecotone, semi-stand, and total 2008] STEERS ET AL.: VEGETATION MAPPING AND ECOTONE ANALYSIS 27 Pacific Ocean 5 es Kilometers Fic. 1. Location of study area in San Luis Obispo County, CA. overlap (Fig. 2). Full extent corresponds to a complete stand, including the discrete portion, and if present, the ecotone portion. The discrete portion of a stand represents the distinct and definable area that clearly embodies a classified plant community or vegetation type. Ecotone represents the area of overlap between two or more plant communities, as mentioned previous- ly (Fig. 3). Semi-stands represent a _partial/ incomplete “‘stand’”’ of one plant community, lacking any discrete space of its own, that is entirely within the matrix of a different plant community (Figs. 2 and 3). A semi-stand togeth- er with the matrix community resembles an ecotone in structure and species composition. The term matrix is used here to represent the background vegetation with its own unique structure or composition (Forman and Godron 1981) from the semi-stand that is within it. Total overlap represents all areas that are not one discrete vegetation type and is calculated by the sum of ecotone and semi-stand areas. Specifical- ly, we wanted to determine the amount of full extent, discrete, ecotone, semi-stand, and total overlap area that each vegetation type occupied, discover which vegetation types shared the greatest amount of ecotone and total overlap area with other vegetation types, and measure the ecotonal and total overlapping space of the study area. STUDY SITE The 93 ha study area is centered near 35°19’'N, 120°39'W (WGS84) within Poly Canyon, a 510 ha natural area NNE of and adjacent to the core campus of California Polytechnic State University, San Luis Obispo, in San Luis Obispo County (Fig. 1). Poly Canyon lies along the southwest foothills of the southern Santa Lucia Range, part of the larger southern Coast Range. The canyon is formed from two northeast-to- southwest trending ridgelines flanking Brizzolara Creek, a seasonal tributary of Stenner and San Luis Obispo Creeks. Many hillside springs and seeps feed the seasonal flow. Elevations range from around 120 to 345 m. The general slope of both canyon sides 1s about 20° (36%) with steeper local inclines to 45° (100%). Soils within the study area are mostly of the Los Osos Loam series, Lodo—Diablo Clay Loam complex, Los Osos— Diablo complex, Rock Outcrop—Lithic Haploxe- rolls complexes (serpentine), and Obispo—Rock outcrop complexes (serpentine) sensu Ernstrom (1984). Climate is a cool summer phase of the dry- summer subtropical (“mediterranean”) type of humid mesothermal climates (Trewartha 1968; Yahr 1961). Winter high temperatures average near 18°C, lows average around 6°C. Summer high temperatures average near 25°C, with average lows near 11°C. The lowest temperature recorded on the adjacent core campus was —12.7°C and the highest was 44.4°C. Precipita- tion falls as rain primarily from October through April, and averages about 558 mm per year. Typically, less than 25 mm of precipitation is recorded from 1 May to 30 September, but overnight and morning fog with near 100% humidity occurs nearly every night unless drier, MADRONO Discrete A. 0.90 ha B. 0.90 ha C. 0.72 ha 28 Vegetation Full Extent Map ‘JA. 0.90 ha (2.80 ha) FIG. 2. down-sloping winds descend from the Salinas Valley over the Santa Lucia Range to overwhelm the onshore flow of marine air (Felton 1965; WRCC 2006) Poly Canyon exhibits high vascular plant diversity with over 400 species collected thus far (De Rome 1997). Within the study area rare plants are present, such as the local serpentine endemic Calochortus obispoensis (De Rome 1997). Typical serpentine indicators (Safford et al. 2005), such as Quercus durata var. durata, are also frequent. The vegetation of the study area 1s = FIG. 3. Hypothetical ecotone and semi-stand scenar- ios and corresponding representative GIS polygon overlap. A. Ecotone between Mixed Chaparral (left- side w/corresponding gray GIS polygon) and Southern Coastal Scrub (right-side w/corresponding striped GIS polygon). B. A semi-stand of Mixed Chaparral occurring in a Southern Coastal Scrub matrix. Hypothetical vegetation map, composed of three plant communities (A, B, and C), depicting defined map units utilized to describe the vegetation of the study area. composed of numerous plant communities, which are described later in this study. METHODS The relatively low cost and high precision of using GPS (Geographic Positioning System) and GIS (Geographic Information System) technolo- gies (see Foster 1993) is now ideal for a high resolution, multi-layer approach to vegetation mapping where stands (patches of a particular plant community) are defined by a suite of indicator species that are mapped to their fullest extent in separate layers. Since plant communities integrate, forming gradual to abrupt ecotones across the landscape, they can be individually separated into layers of a GIS. This approach avoids the arbitrary or inconsistent definitions of stand boundaries that can result from creating one integrated map that portrays stands as entirely discrete and not overlapping. Portraying plant communities and ecotones in this way is precise and feasible at the local scale (e.g., hillside, small nature reserve, rancho, etc.). Distinct plant communities were classified and mapped in this study following the Holland and Keil (1995) plant community classification scheme and this multi-layer, full extent mapping approach. Dominant species from the following categories: tree, shrub, forb, and grass, were also recorded for each individual stand mapped and used to further describe the communities and to provide a cross-reference to vegetation series in the Manual of California Vegetation (Sawyer and [Vol. 55 2008] Keeler-Wolf 1995). Ecotones among the eleven distinct communities were mapped as the overlap of their respective polygons in the different GIS layers (Figs. 2 and 3). The areas of overlap were then used to quantify community and ecotone characteristics within the study area using a GIS. Methodology of the vegetation classification, mapping, and ecotone analyses are described in the following sections. Vegetation Classification The Holland and Keil (1995) plant community classification scheme was used to classify the vegetation of the study area. This classification scheme distinguishes plant communities primarily on physiognomy and secondarily on species composition. Habitat characteristics are also incorporated when they will increase usefulness (e.g., coastal sand dune communities, marine aquatic communities, riparian, vernal pool). This classification scheme is not all-inclusive of the plant communities found in California but does provide a logical framework that is helpful for classifying vegetation. During spring 2001, the study area was traversed by foot to produce a list of plant communities present. Stands encountered were examined for physiognomy and dominant species composition by carefully walking throughout the discrete portions of each stand. Dominant species were defined as those plant species that contrib- uted the greatest cover (Barbour et al. 1999) based on ranked percentage cover estimates (Daubenmire 1959; Mueller-Dombois and Ellen- berg 1974). Once dominant and subordinate species were recorded for a stand they were compared to community descriptions in Holland and Keil (1995) and classified accordingly. The list of plant communities was then formatted as a data dictionary and uploaded into the GPS datalogger for use in mapping and classifying vegetation polygons in the field. Vegetation Mapping Area Subdivision. Mapping began spring 2001 and was completed by spring 2002. The study area was split into eight subareas that roughly followed the four major slopes (N-, E-, S-, and W-facing slopes) of the two prominent hills/ ridgelines of the study area. Mapping priority fell to stands located on the perimeter of the study -area within each subarea since these stands typically had less integrating neighboring stands, and because it was easier to keep track of species composition and reference vegetation patches upslope than downslope. Interior (uphill) stands within a subarea were subsequently mapped until the subarea was completed. All stands within a I STEERS ET AL.: VEGETATION MAPPING AND ECOTONE ANALYSIS 22 subarea were mapped before moving on to an adjacent subarea. Mapping Units. No exact minimum mapping unit was defined before mapping began but general guidelines were established based on two factors: the physiognomy of the stand, and the context it occurred in. In regard to physiog- nomy, minimum vegetation units generally in- creased in size by herb-, shrub-, and _ tree- dominated stands, respectively. For example, the minimum mapping unit of Valley and Southern Coastal Grassland was smaller than that for Coast Live Oak Woodland because some stands of Valley and Southern Coastal Grassland could be the size of the canopy of one large coast live oak tree. On the other hand, a single large coast live oak tree in the middle of Valley and Southern Coastal Grassland could not be con- sidered Coast Live Oak Woodland even if it occupied the same or a bigger areal extent as a patch of grassland because it was only one individual and not an assemblage. In regard to context, if a patch of vegetation appeared to form a distinct stand based on physiognomy and species composition compared to adjacent patch- es, then it was mapped. Indicator Species. Stand-specific indicator spe- cies were determined before mapping in order to delineate the boundary of the stand (1.e., its full- extent). Indicator species were species with high cover in the stand, were distributed throughout the stand, and usually had a life form that corresponded to the physiognomy of the stand itself (e.g., shrub species in a shrubland). In a few instances, species of dissimilar life form than the stand physiognomy were also used (e.g., Rhamnus californica, a shrub, used as an indicator for Coast Live Oak Woodland). If an indicator species was also common within an adjacent stand it was removed from the list of indicator species. Thus, suites of indicator species (differential species) collectively exhibit high fidelity (strong or exclusive correlation), within-stand constancy (continuous presence), and area-of-occupancy within discrete vegetation associations (sensu Braun-Blanquet 1965; Kent and Coker 1992), and are uncommon or absent in adjacent associations. Sometimes, only one indicator species met these criteria, but often several species were needed. The suite of indicator species was stand specific and not all stands classified as the same plant community were necessarily repre- sented by the same suite of indicator species. Semi-stands. Semi-stands are a low density patch, lacking any discrete space of their own, and composed of one or more species typical of one community that is located entirely within another community. The resulting mix of species 30 MADRONO from both the community representing the sem1- stand and the community representing the matrix resembles an ecotone in structure and species composition. In these cases, the low density patch of species (semi-stand) was mapped so that where it overlapped with the matrix community, it would represent the ecotone-like assemblage that occurred there. For example, a number of hard- stemmed sclerophylous shrub species common to the Mixed Chaparral of the study area were within a stand of Southern Coastal Scrub that was dominated by soft-stemmed drought decid- uous shrubs of a relatively lower stature. Collectively, these Mixed Chaparral species did not form a discrete stand of Mixed Chaparral because their cover and density was too low, but the patch of chaparral shrubs embedded in the scrub matrix resembled the ecotone between Mixed Chaparral and Southern Coastal Scrub (Fig. 3). In this example there was no discrete area belonging to the matrix community or belonging to the community representing the semi-stand, yet quantifying discrete space was a goal for this study, so these areas could not be overlooked. Two main possibilities for mapping this situation were considered. Either one could map the area as discrete space that represented a unique community (in this case a stand of Mixed Chaparral\Southern Coastal Scrub co-dominated by Cercocarpus betuloides and Artemisia califor- nica) with no ecotonal space between the surrounding Southern Coastal Scrub, or one could map the low density patch of Mixed Chaparral shrubs and overlay it on the matrix patch of Southern Coastal Scrub to represent the ecotone-like assemblage. The later alternative was chosen since it was conceptually similar to the method used to map ecotonal space, because it would avoid the classification of potentially numerous new communities formed from the combinations of the eleven vegetation types found in Poly Canyon, and because it seemed likely in most of these situations that the semi- stands together with the matrix vegetation represented a successional stage from one vege- tation type to another and not a static or stable vegetation type (further monitoring would be required in order to confirm this). Thus, semi- stands were classified and mapped where physi- ognomy and species composition within a dis- crete matrix patch approached characteristics of a separate plant community, mimicking an ecotone but lacking the full transition to a discrete stand of the semi-stand vegetation type. Field Mapping Sessions. Once the plant com- munity and indicator species had been deter- mined for a stand or semi-stand, it could then be mapped. Master lists of plant communities and species based on field data gathered during [Vol. 55 previous inventories (De Rome 1997; Curto 2000) and this project were formatted as a data dictionary uploaded into a GeoExplorer® III (Trimble Navigation Limited, Sunnyvale, CA) mapping-grade GPS receiver used to map vege- tation polygons and assign their dominant species attributes in the field. Vegetation polygons were created by slowly walking the receiver around the border of each stand while GPS positions were logged at three-second intervals. Stand borders were based on the full extent of respective suites of indicator species (Figs. 2 and 3). Mapping sessions were planned around times in the day when the GPS precisional dilution of position (PDOP) was lowest (<4) resulting in the highest positional accuracy. In a few instances, topogra- phy or a dense canopy would obstruct the GPS unit from satellite view enough that the desired PDOP was not achieved. GPS and GIS' Data Processing. GPS data were differentially corrected (horizontal accuracy +1 m) using Trimble® GPS Pathfinder® Office 2.80 (Trimble Navigation Limited, Sunnyvale, CA) before import to the GIS (ArcView®, ESRI, Redlands, CA) relative to the nearest base station at Vandenberg Air Force Base. Differentially corrected stand polygons were edited in the GIS to correct any points determined to be outliers by comparison to other points in the polygon while. overlaid on background orthophotographs (which had 1 m resolution). The few stands that: were partially mapped at a higher PDOP than 4. were carefully scrutinized. | Relational Species Lists. After all mapping was completed in spring 2002, lists of the top three species with the highest cover in each of the following categories: tree, shrub, forb, and grass, were created for every stand and semi-stand mapped. All polygons were revisited and the species were determined based on ranked per- centage cover estimates (Daubenmire 1959; Mueller-Dombois and Ellenberg 1974). The. species were recorded in rank order within each growth form category and were linked to their) respective polygon on the GIS using the attribute’ table. Nomenclature followed Hickman (1993). This was done to establish a baseline of the dominant growth forms in each stand for future: reference. In addition, the lists were used to. provide basic floristic descriptions of each plant: community mapped and to reconcile the plant communities classified in Holland and Keil (1995) with their respective vegetation series in. the Manual of California Vegetation (Sawyer and. Keeler-Wolf 1995). Ecotone Analysis Geoprocessing functions of the GIS (ArcGIS®, ESRI, Redlands, CA) were used to ascertain the: 2008] amount of total overlap (ecotone plus semi- stand) among communities. Specifically, the union, intersect, and dissolve processes were used to create layers of discrete vegetation and total overlap for each individual stand, plant commu- nity, and for Poly Canyon as a whole, based on the eleven layers of plant communities. Where stands and semi-stands of different plant com- munities overlapped, the intersect process would create a new layer representing that overlap (or intersection) between the two communities. This was performed for every combination of plant communities. In addition, all of the overlap layers created were combined for each community, and for the entire study area, with the union process. The dissolve process was then used to remove boundaries within contiguous areas to form single polygons representing total overlap of the entire study area. The end products represented total overlap (ecotone plus semi-stand) for the entire study area, by plant community, and by individual polygons. Discrete space was then calculated by subtracting total overlap from the original data to determine values for the entire study area and for each plant community. To determine ecotonal area, all semi-stand polygons were deleted from copies of the shape files representing all of the plant communities. Then, the same procedure as described above was implemented with the geoprocessing functions of the GIS to obtain the amount of ecotonal area. The end product represented ecotone space for the entire study area, by plant community, and by individual polygons. RESULTS Vegetation Classification Eleven plant communities were identified using the Holland and Keil (1995) classification scheme. Descriptions of each plant community are listed below. The numbers of stands with discrete area for each plant community are written in parentheses following the name of the plant community. Species information within each description was derived from the surveys of the three most dominant species within each growth form (1.e., tree, shrub, forb, and grass), which were recorded for each stand mapped. The species data used in the descriptions were based solely on stands containing discrete area (i.e., species information from semi-stands was not used). Woody perennials that exhibited a suffru- _tescent or vine-like growth form were included in the forb category. _ Several communities, such as Yucca\Bunch- grass Scrub, Serpentine Chaparral, and Califor- nia Bay\Leather Oak Mosaic, were suspected of Indicating serpentine soil in the study area based /on the consistent presence of serpentine indicator | M | STEERS ET AL.: VEGETATION MAPPING AND ECOTONE ANALYSIS 31 species (Safford et al. 2005) and the appearance of the substrate found within their stands. While no soil samples were collected to analyze for serpentine characteristics, when stands of these vegetation types were overlaid on a soil map (Ernstrom 1984), all three were found on serpentine soils. Most Native Bunchgrass Grass- land stands also overlapped with serpentine soils (serpentine bunchgrass sensu CNDDB 2003) but a few stands were also found in non-serpentine soils. Finally, the corresponding Manual of Califor- nia Vegetation (MCV) (Sawyer and Keeler-Wolf 1995) vegetation series are listed at the bottom of each description for cross-reference purposes, and are designated by “MCV”. Some of the series encountered were not in the manual but were still named using the format described in the MCV. Asterisks (*) indicate those vegetation types that have not been previously described by Holland and Keil (1995), Sawyer and Keeler- Wolf (1995), or by other classifications or studies (i.e., Epling and Lewis 1942; Munz and Keck 1949; Thorne 1976; Kirkpatrick and Hutchinson 1977; Paysen et al. 1980; Westman 1983; Holland 1986; Barbour and Major 1988; Desimone and Burk 1992; Rodriquez-Rojo et al. 2001; CNDDB 2003). Valley and Southern Coastal Grassland (11) — Dominated by various nonnative annual grass species from the genera Avena, Brachypodium, Bromus, Hordeum, and Lolium. Nassella pulchra was recorded as exhibiting high cover in several stands but never was the dominant. Nonnative forbs included Foeniculum vulgare, Hirschfeldia incana, Rumex crispus, and Vicia villosa, among others. Native forbs included Eschscholzia cali- fornica, Ranunculus californicus, and Sisyrinchium bellum, among others. MCV = California Annu- al Grassland. Native Bunchgrass Grassland (7) — Dominated by two species of native perennial bunchgrass species, either Melica imperfecta or Nassella pulchra. Other native grasses included Nassella lepida. Annual grasses included Vulpia microsta- chys and nonnatives typical of Valley and Southern Coastal Grassland. Forbs included Bloomeria_ crocea, Calochortus clavatus subsp. clavatus (List 4.3 - CNPS 2007), Cryptantha clevelandii, Galium porrigens, Grindelia hirsutula, Layia platyglosa, Plantago erecta, Sisyrinchium bellum, Stachys bullata, and Trifolium willdenovii. MCV = Purple Needlegrass (5); “Onion Grass (2). *Yucca\Bunchgrass Scrub (2) — Co-dominated by Yucca whipplei and Nassella lepida. In addition, soft-stemmed shrubs characteristic of Southern Coastal Scrub collectively contributed high cover, notably Artemisia californica, Lotus scoparius, and Mimulus aurantiacus. Forbs with the highest cover included Chorizanthe palmeri 32 MADRONO (List 4.2 - CNPS 2007), Eschscholzia californica, Plantago erecta, Selaginella bigelovii, and Stachys bullata. Other grasses included Melica imperfecta, Nassella pulchra, and Bromus madritensis. MCV = *Chaparral Yucca\Purple Needlegrass. Southern Coastal Scrub (17) — Dominated by soft-stemmed shrubs, including Artemisia califor- nica, Mimulus aurantiacus, Salvia mellifera, or Toxicodendron diversilobum. One stand on the southwest corner of the study area near the core campus was dominated by non-native Opuntia ficus-indica. Other shrubs included Baccharis pilularis, Hazardia squarrosa, Lotus scoparius, Lupinus albifrons, and Rhamnus crocea. Forbs with the highest cover exhibited a vine or vine- like growth form, such as Calystegia macrostegia, Galium californicum, Keckiella cordifolia, and Senecio mikanioides. Other forbs included A chil- lea millefolium, Carduus pycnocephalus, Conium maculatum, Gnaphalium californicum, and Salvia spathacea, among others. Grasses with the highest cover were nonnative annuals typical of Valley and Southern Coastal Grassland, notably Brachypodium distachyon. MCV = California Sagebrush (6); *Sticky Monkey Flower (2); Black Sage (5); *Poison Oak (3); *Indian-Fig (1). Chamisal Chaparral (1) — Dominated by Adenostoma fasciculatum. Cercocarpus betuloides and Salvia mellifera were also present. No forbs were found. Grasses with the highest cover were nonnative annuals typical of Valley and Southern Coastal Grassland. MCV = Chamise. Mixed Chaparral (1) — Codominated by Adenostoma fasciculatum and Cercocarpus betu- loides. Rhamnus crocea had the third highest shrub cover. One individual of Arctostaphylos luciana (List 1B.2 - CNPS 2007) was also found. Forbs with the highest cover were Keckiella cordifolia, Salvia spathacea, and Symphoricarpos mollis. Grasses with the highest cover were Bromus diandrus, Leymus condensatus, and Nas- sella lepida. MCV = *Birchleaf Mountain- Mahogany — Chamise. Serpentine Chaparral (61) — Dominated by the strict serpentine endemic Quercus durata var. durata (Holland and Keil 1995; Safford et al. 2005). Cercocarpus betuloides and Rhamnus crocea were occasional, and Garrya veatchii was rare. Forbs with the highest cover were Calyste- gia macrostegia, Galium californicum, and Stachys bullata. Grasses with the highest cover were Bromus diandrus, Bromus madritensis, Leymus condensatus, Melica imperfecta, and Nassella pulchra. MCV = Leather Oak. *California Bay\Leather Oak Mosaic (1) — Co- dominated by Umbellularia californica and Quer- cus durata var. durata. Quercus berberidifolia and Rhamnus crocea had the second and third highest cover in this stand, respectively. Forbs and grasses with the highest cover were similar to those found in Serpentine Chaparral. This [Vol. 55 community exhibited a bi-modal physiognomy appearing as an equal and even mixture of an open-canopied, reduced form of Central and Southern Mixed Evergreen Forest (see Holland and Keil 1995) and Serpentine Chaparral. The California Bay\Leather Oak Mosaic was restrict- ed to seeps on steep slopes in serpentine soils dominated by Yucca\Bunchgrass Scrub, and was therefore always mixed with Yucca whipplei in the understory or dripline. Only one stand exhibited discrete vegetation apart from the Yucca\Bunch- grass Scrub matrix. This community might be similar to the “Leather Oak-California Bay- Rhamnus spp. Mesic Serpentine NFD Super Alliance” found in Napa County by Thorne et al. (2004). MCV = *California Bay\Leather Oak. *“Toyon Woodland (6) — Although toyon (Heteromeles arbutifolia) is commonly recorded as a shrub in California vegetation (Holland and Keil 1995), within this study area it exhibited both a shrub and tree form. Heteromeles arbutifolia trees were the dominant component of the Toyon Woodland. Prunus illicifolia had the second highest cover in this community and Quercus agrifolia and Sambucus mexicana were also important tree components. Shrubs with the highest cover included Holodiscus discolor and also species common to the study area’s Southern Coastal Scrub. Forbs with the highest cover were Carduus pycnocephalus, Stachys bullata, and Torilis arvensis. Grasses with the highest cover included Brachypodium distachyon, Leymus con- densatus, and Melica imperfecta. MCV = *Toy-. on. Coast Live Oak Woodland (11) — Dominated by Quercus agrifolia. Heteromeles arbutifolia and Umbellularia californica were common trees with high cover. Shrubs with high cover included. species common to the study area’s Southern Coastal Scrub, notably Toxicodendron diversilo-. bum. Forbs with high cover included Carduus pycnocephalus, Galium porrigens, Salvia spatha- cea, Solidago californica, and Stachys bullata. Grasses with high cover included Elymus glaucus, Melica imperfecta and nonnative grasses indica-_ tive of Valley and Southern Coastal Grassland. | MCV = Coast Live Oak. Valley and Foothill Riparian (11) — In the study area, ten Valley and Foothill Riparian stands exhibited a woodland physiognomy while one stand exhibited an open shrubland physiognomy. | The ten Valley and Foothill Riparian woodlands, were dominated by Platanus racemosa, Q. agri-. folia, Salix lasiolepis, or U. californica. Other. trees, such as Salix laevigata and Heteromeles arbutifolia, were occasional. Baccharis pilularis, Rhamnus californica, and Toxicodendron diversi-\ lobum had the highest shrub cover in these, woodlands. Forbs with high cover included Carex senta, Juncus patens, Helenium puberulum, | Mimulus guttatus, Rorripa nasturtium-aquaticum, | 2008] and Rumex crispus. Grasses with the highest cover included Agrostis viridis, Elymus glaucus, Phalaris aquatica, Piptatherum milliaceum, and nonnative annuals indicative of Valley and Southern Coastal Grassland. The one shrub stand of Valley and Foothill Riparian was located adjacent to fenced-off cattle pasture on the western edge of the study area. It appeared that cattle had grazed this stand in the past based on numerous ruts found along the contours of sloped sections. The stand was dominated by Baccharis pilularis while Ricinus communis had the second-highest cover. Typha latifolia and a Juncus sp. had the highest forb cover, and Lolium multiflorum had the highest grass cover. MCV = California Sycamore (2); Coast Live Oak (2); Arroyo Willow (2); *Bay Laurel (4); *Coyote Bush (1). Vegetation Map Eleven plant community layers and an anthro- pogenic disturbance layer were created (Fig. 4). Total area mapped was 92.6 ha and was made up of 229 vegetation polygons. Anthropogenic areas covered about 3% (21 polygons) of the total area mapped and represented plantings, such as a Eucalyptus stand, and severely disturbed locales such as roads, irrigated pastures, and a small landfill/quarry. Anthropogenic coverage was excluded from all analyses. The three most extensive plant communities as a function of full extent areal coverage were, in descending order, Yucca\Bunchgrass Scrub, Native Bunchgrass Grassland, and Valley and Southern Coastal Grassland (Table 1). When ranking the three most extensive plant communities as a function of discrete areal coverage, the only change was that Valley and Foothill Riparian had more coverage than Valley and Southern Coastal Grassland. The three plant communities with the least full extent areal coverage were, in descending order, Chamisal Chaparral, Serpen- tine Chaparral, and Toyon Woodland. When ranking the plant communities with the least discrete areal coverage, the ranking became _ California Bay\Leather Oak Mosaic, Toyon _ Woodland, and then Serpentine Chaparral. | Serpentine Chaparral had the highest number of stands with discrete space, at 61, while California Bay\Leather Oak Mosaic, Chamisal _ Chaparral, and Mixed Chaparral were represent- | ed only by one stand with discrete area (Table 1). | The largest mapped stand with discrete space was of Yucca\Bunchgrass Scrub, at 192,676 m’, while _ the smallest stand mapped was of Serpentine _ Chaparral, at 9 m*. Serpentine Chaparral also _ had the highest number of semi-stands at 45 while _ Chamisal Chaparral and Valley and Foothill _ Riparian had none. The largest semi-stand was | of California Bay\Leather Oak Mosaic, at STEERS ET AL.: VEGETATION MAPPING AND ECOTONE ANALYSIS Bis) 13,680 m’*, while the smallest semi-stand was Yucca\Bunchgrass Scrub, at 4 m’. Ecotone Analysis Out of the 89.9 ha of mapped vegetation (not including anthropogenic cover), 54.8 ha (61%) was discrete, non-overlapping vegetation, and 35.1 ha (39%) was overlap (ecotone plus semi- stand). When all semi-stands were removed and the analyses repeated, the study area was found to have 32.5 ha (36%) of ecotone (Fig. 5). Of all vegetation coverages, Yucca\Bunchgrass Scrub and Native Bunchgrass Grassland occupied the largest total area (discrete plus overlap) and discrete coverage, respectively. The plant com- munity with the highest amount of ecotone (176,399 m?) and highest amount of total overlap (179,942 m*) was Native Bunchgrass Grassland. Only Chamisal Chaparral, Valley and Foothill Riparian, and Yucca\Bunchgrass Scrub exhibited discrete areal coverages more than 50% of their respective total areas (Fig. 6). The other eight plant communities had more of their area represented as ecotone than as discrete areal coverage. Thus, in general, a greater percentage of each community’s full extent areal coverage was ecotonal, even though the majority of the study area was discrete space. Finally, each plant community’s ecotone and total overlap were analyzed to determine which other communities contributed or shared the majority of that area (Table 2). Southern Coastal Scrub was found to have the greatest areal contribution with the most number of plant communities. Valley and Southern Coastal Grasslands had the second greatest areal contri- bution to both ecotone and total overlap space with the most plant communities. Among the 79 semi-stands, 51.9% were found within Yucca\Bunchgrass Scrub, 22.1% within Native Bunchgrass Grassland, 20.2% within Valley and Southern Coastal Grassland, 2.9% within Southern Coastal Scrub, 1.9% within Valley and Foothill Riparian, and 1.0% within Mixed Chaparral. All of the semi-stands in Yucca\Bunchgrass Scrub and Native Bunchgrass Grassland were on serpentine soil based on the overlay of a soils map (Ernstrom 1984). On non- serpentine soils, semi-stands were relatively less common but they were most often found in a matrix of Valley and Southern Coastal Grass- land, then Southern Coastal Scrub. DISCUSSION Vegetation Classification Based on the Holland and Keil classification scheme, 11 visually distinct plant communities were mapped. If a more detailed classification [Vol. 55 ~ MADRONO 34 “Toe i — NS yb Center for Embedded Networked Sensing, University of California, Los Angeles, CA 90095 rundel@biology.ucla.edu ABSTRACT Mount Pinos at the western margin of the Transverse Ranges of Southern California. reaches 2692 m in elevation at its peak. The broad summit of the mountain supports an area of subalpine fellfield vegetation, as well as an unusual low elevation occurrence of limber pine (Pinus flexilis). We describe the summit area of the mountain and characterize the community structure and ecophysiology of the fellfield community and associated pines. The fellfield community, dominated by low mats and cushions of Eriogonum kennedyi, Phlox diffusa, and Lupinus breweri var. bryoides, has a mean plant cover of 44%. These fellfield species have amphistomatic leaves with abundant, slender palisade cells on both sides of the mesophyll. These traits represent a strategy shared with desert plants. Amphistomaty helps to increase assimilation by maximizing stomatal conductance to CO, during the gas phase of diffusion, and the isolateral mesophyll maximizes diffusion of CO, by exposing a very high mesophyll membrane area per leaf surface area. However, the mean maximum rates of assimilation measured for Phlox diffusa, Eriogonum kennedyi, Astragalus purshhii var. tinctus, and Lupinus breweri var. bryoides were 6.8 to 13.9 umol CO; m ’ s_'", placing them in the lower range of those measured in alpine areas of the world. Key Words: fellfield, Massenerhebung effect, Mount Pinos, photosynthetic rate. Mount Pinos, with a summit reaching 2692 m, forms the western end of the Transverse Ranges of California. The mountain holds much botan- ical interest because it lies at a junction between the Coast Range to the north, the Transverse Ranges to the east, the San Joaquin Valley to the northeast (Fig. 1), and with a_ well-developed desert community of single-leaf pinyon pine (Pinus monophylla) and sagebrush (Artemisia tridentata) on its eastern flanks. Having high elevation, its flora is strongly related to that of the Sierra Nevada but it is well isolated across the Mojave Desert from this range. With the exception of a single community study of Jeffrey pine (Pinus jeffreyi) forest stands on the south slope of Mount Pinos (Vogl and Miller 1968), there have not been ecological studies of this remarkable area. Our interests have focused on two aspects of the summit area where subalpine plants occur well below their normal range of distribution. One of these is the _ presence of a large fellfield community of _ herbaceous perennial cushions and low herbs on thin granitic soil (Fig. 2a, b), while the other is _the unexpected occurrence of the subalpine conifer Pinus flexilis. Our objectives for this study were to describe the summit area of the _mountain and to characterize the community _ Structure and ecophysiology of the fellfield community and associated pines. Species taxon- - omy in this paper follows Hickman (1993). STUDY AREA Mount Pinos along with nearby Frazier Mountain (2365 m) are prominent mountain massifs immediately south of the San Andreas fault in the western Transverse Ranges of Southern California. These mountains are sepa- rated from the west-east trending San Emigdio and Tehachapi Ranges to the northeast by the trough of the San Andreas Fault, which is responsible for the steep north slope of Mount Pinos, while the Big Pine Fault separates Frazier Mountain from Mount Pinos (Dibble 1982). Geologic maps of these features can be seen in Kellogg (2004). These mountains, which have been variously considered the southeastern cor- ner of the California Coast Ranges and the western margin of the Transverse Ranges, are characterized by a low profile and broad (several km?’), nearly flat-topped summits. Both have unusual geology resulting from the thrusting of granitic and gneissic basement rock over sedi- mentary rock as young as the Pliocene. Although it was once felt that the summits of Mount Pinos and Frazier Mountain were rem- nants of old erosional surfaces (Dibble 1982), recent studies suggest a different tectonic history (Kellogg 2004). Crystalline rocks underlying both mountain massifs are thoroughly fractured and crushed, being thrust above easily deformable Miocene and Pliocene sandstone, mudstone, and 42 Mount Pinos @Fr SANTA BARBARA COUNTY VENTURA COUNTY pwn 7 Anacapa Island Fic. 1. shale formations below. Late Pliocene and early Quaternary thrusting associated with regional uplift and ongoing seismic activity of the San Andreas Fault have resulted in deep shattering of the granitic bedrock. Field studies were conducted within the 206-ha Botanical Research Area comprising the broad summit area of Mount Pinos (lat. 34.48.7°N, long. 119.08.4°W) in the Los Padres National Forest. Our work was carried out in a subalpine fellfield community and adjacent open pine stands at 2650-2670 m on north- and south- facing slopes adjacent to the dirt access road to the summit. We consider this to be a subalpine community because of the fellfield community and presence of Pinus flexilis. The highest point of the mountain houses a U.S. Air Force microwave facility for communication between Vandenberg and Edwards Air Force Bases. The summit area supports an open subalpine forest, with a mixed conifer community heavily dominated by Jeffrey pine (Pinus jeffreyi). Also present are scattered individuals of white fir MADRONO ‘\ KERN COUNTY [Vol. 55 LOS ANGELES COUNTY \ Los Angeles Location of Mount Pinos on the western end of the Transverse Ranges on the boundary line between Ventura and Kern Counties. (Abies concolor) and single-leaf pinyon (Pinus monophylla). A surprising member of this conifer community is limber pine (Pinus flexilis), which grows along ridges in the summit area. These pines grow with an open understory of scattered low shrubs of Ribes cereum, Symphoricarpos rotundifolius var. parishii, and Ceanothus cordu- — latus. On other slopes with deeper soils, cushion plants commonly grow with the low shrub Chrysothamnus viscidiflorus subsp. viscidiflorus. Several associations of fellfield community, much like typical fellfields at higher alpine locations in the Sierra Nevada, occur on the slopes where soils are thin sandy loam over | fractured granite rock (Fig. 2a). These fellfield | communities contain a relatively moderate diver- — sity of herbaceous species, including cushion | plants, low-growing mats, upright dicot perenni- | als, perennial graminoids, and geophytes | (Fig. 2b). A checklist of the flora of the summit area of Mount Pinos above 2590 m includes 138 native » vascular plants plus five non-native species — GIBSON ET AL.: FELLFIELD PLANTS OF MOUNT PINOS 43 RIG. 2. (Muns 1994). The seven largest families present in this flora, excluding non-native species, are the Asteraceae (18 spp.), Scrophulariaceae (12 spp.), Fabaceae (11 spp.), Liliaceae, sensu Jepson Manual (9 spp.). Poaceae (8 spp.), Polemoniaceae (7 spp), and Polygonaceae (7 spp.). The flora is dominated by low-growing herba- ceous perennials (hemicryptophytes) that annu- ally die back to ground level. Only four tree species, as described above, and nine shrub Species are present. In addition to the three shrub species listed above, shrubs present are Chry- _sothamnus nauseosus, Holodiscus discolor, Ribes Fellfield habitat and study species growing at about 2650 m in the summit area of Mount Pinos, California. a) General aspect of fellfield community; b) Typical cover of the fellfield community, with Eriogonum kennedyi, Phlox diffusa, and Lupinus breweri as dominants; c) Astragalus purshii var. tinctus; d) Eriogonum kennedyi; e) Lupinus breweri var. bryoides; f) Phlox diffusa. montigenum, R. velutinum, Rhamnus tomentella subsp. cuspidata, and Sambucus mexicana. The summit region is rich in geophytes, with nine species - Jris missouriensis, Sisyrinchium bellum, Allium burlewii, A. campanulatum, A. denticula- tum, A. fimbriatum, A. howellii var. clokeyi, Calochortus venustus, Fritillaria pinetorum, Muilla maritima, and Veratrum californicum. Annuals are relatively abundant, with 27 native species reported. We know of no records for precipitation for the summit area of Mount Pinos. Twisselmann (1967) suggested a mean annual precipitation of 44 about 430 mm, with a range from about 280 mm during a drought year to as much as 890 mm ina wet year. The USDA Forest Service Chuchupate Ranger Station, at 1603 m elevation and about 30 km by road from Mount Pinos, has a mean annual rainfall of 233 mm for a limited record of three years. North-facing slopes of the mountain hold snow well into June in wet years as in 2005, while south-facing slopes melt out far earlier. There was 1.5 to 2 m or more of snow on Mount Pinos at the end of January 2005 (Mork 2005). During dry years when snowfall is light, these summit slopes are exposed to high solar irradi- ance and wind for much of the year. MATERIALS AND METHODS Phenology Multiple field trips to the study site in 1998 to 2005 allowed qualitative observations to be made of variable patterns of community vegetative and flowering phenology. These patterns were related to snow conditions that varied greatly between years. Community Structure Line transects 50m in length along slope contours were used to sample four topographic areas of the fellfield community on June 13, 2000 to determine cover of each species, total vegeta- tion, and exposed large rocks for each plot. Canopy interception along each transect was recorded to the nearest 5 cm for all perennial plant species. When canopies of two species overlapped, the subcanopy presence of a species was recorded separately to allow determination of total ground cover and whether any species had a characteristic understory habit. Cover of exposed large rocks (>20 cm diam.) was also measured to the nearest 5 cm along the transect. Leaf Form and Anatomy of Cushion Plants Anatomical characteristics of blades were first observed with a light microscope from thin trans- sections of several fresh leaves cut with a razor blade, and measured with an optical micrometer. Leaves were also liquid-preserved in 70% forma- lin-acetic acid-alcohol, dehydrated in an ethanol series, critical point dried and coated with 200 nm of gold-palladium, and then examined with an ETEC Autoscan scanning electron microscope. Ecophysiological Characteristics Physiological parameters of selected summit species were sampled under conditions of new growth (full leaf expansion and peak flowering), MADRONO [Vol. 55 midsummer, and fall conditions. Gas exchange measurements were made midmorning through midday on July 8, 1999, on four common cushion-forming species of the fellfield: Eriogo- num kennedyi var. kennedyi (Polygonaceae), Phlox diffusa (Polemoniaceae), Lupinus breweri var. bryoides (Fabaceae), and Astragalus purshii var. tinctus (Fabaceae) (Fig. 2c—2f). Also mea- sured were the photosynthetic traits of the two common pine species in the summit area, Pinus Jeffreyi and P. flexilis. On November 8, 1999, we repeated these sets of measurements under late growing season conditions, but the aboveground tissues of E. kennedyi and P. diffusa were dormant on this date. Key gas exchange param- eters measured were mean maximum assimilation rate (Amax, UMol m~” s_'), stomatal conductance to water vapor (g,, mmol m~’ s‘'), transpiration (E, mmol H,0 m-~’s''), and internal versus ambient CO, concentration (cj:c, ratio). These measurements were obtained in situ using a LI- 6200 portable photosynthesis system equipped with a 250-ml sample chamber (LI-COR, Inc., Lincoln, NB). Sample sizes on each date were typically 4-5 individuals. These measurements were made on days when the sky was cloudless, but a portable light source was used to provide saturating blue and red wavelengths of the visible spectrum (Quantum Devices, Inc. Barnsveld, WI) at photon flux density exceeding 1500 umol photons m~’ s'' (PFD,,;) to achieve the highest possible photosynthetic rates. Maximum PFD levels at this site exceeded 2100 umol photons m *s_'. Reported photosynthetic rates are based on the the projected leaf area for both angio- sperms and pines. On August 5, 1999, we measured light response curves on Astragalus purshii var. tinctus and Phlox diffusa at the field site using the portable light source. These values were measured at ambient temperatures of 22—23°C and relative humidity of 22-25%, and VPD was 2.02.5 kPa. The sample chamber was covered with aluminum foil and allowed to equilibrate to obtain mea- surements of dark respiration. Shoots enclosed in the sample chamber were | harvested and kept in moisturized plastic bags, | for determining area measured using a LI-3100 | leaf area meter (LI-COR, Inc., Lincoln, NB). | Manual measurements of leaf geometry were also | used to verify these values. Leaves were dried | for 24 hr at 85°C to determine dry weight for calculating leaf specific area (LSA, m° leaf area | kg! dry leaf tissue). Midday shoot water potentials of small. branchlets were measured on the July and> November sample dates using a Scholander- | type pressure chamber (PMS Instruments). Replicate measurements were made for at least | three individuals per species to calculate a mean value. 2008] TABLE 1. GIBSON ET AL.: FELLFIELD PLANTS OF MOUNT PINOS 45 PERCENT MEAN COVER OF ALL PERENNIAL SPECIES IN A SUBALPINE FELLFIELD SAMPLED ALONG PAIRS OF 25-M LINE TRANSECTS FROM FOUR REPRESENTATIVE SLOPES AT THE BOTANICAL RESEARCH AREA OF MOUNT PINOS. Line transects followed contours of the slope. Upper slope, SW-facing, Taxa slope 10% Astragalus purshii var. tinctus 0.2 Astragalus whitneyi 0.1 Castilleja applegatei subsp. martini 6.1 Chrysothamnus viscidiflorus 2.1 Eriogonum kennedyi 22.8 Ivesia santolinoides 0.0 Leptodactylon pungens 0.1 Lupinus breweri var. bryoides 0.0 Lupinus lepidus var. confertus 3.0 Pedicularis semibarbata 0.2 Penstemon speciosus 0.0 Phlox diffusa 5.0 Silene bernardina 1.6 Poaceae 2D Other Asteraceae 02 Alliaceae 0.2 Overlap,% 29 Total plant cover, % 41.2 Exposed rocks, % 6.5 RESULTS Phenology Variable snowfall conditions strongly impact the phenology of herbaceous perennials and subshrubs at the summit area. In the spring of 1998, although regional precipitation was below average, snowbanks nevertheless persisted in the pine forests adjacent to the study site into early July. At the same site in 2000, the summit received little winter snow and the study site was much drier in early July. Spring growth of the fellfield perennials and the neighboring woody shrubs Ribes cereum and Chrysothamnus viscidiflorus had begun by mid-April in 1999 and 2000, when midday air temperatures at the site characteristically reach or exceed 15°C. With the heavy snowfall of 2005, new growth of fellfield perennials and leafing of shrubs at the beginning of June was limited to warmer exposed south- facing slopes, and was there only just beginning. Flowering peaked in late July 2005. Community Structure Mean line transect cover of the subalpine fellfield vegetation varied from 39-51% of ground area with a mean for the four plots of 44%, whereas exposed large rocks were relatively ‘infrequent on the four slopes (Table 1). Approx- imately half of the exposed ground area was soil, _a sandy loam with subsurface rocks and scattered Mean cover, % Lower slope, Upper slope, Upper slope, SW-facing, NW-facing slope S to SE-facing, slope <5% 5—10% slope 10% 0.0 0.0 0.6 0.0 0.0 0.0 0.7 0.2 3.0 10.5 3.2 3.0 17.9 9.1 19:0 0.1 1.4 0.2 0.2 0.0 0.8 0.0 8.9 0.0 0.3 0.0 0.0 1.0 1.7 0.7 0.2 0.6 0.1 15.0 12.5 7.1 1.7 0.8 {2 2.8 3.7 1.7 0.5 0.1 0.5 03 03 0.3 4.4 3.8 2.2 46.8 38.7 37.1 23 2.3 1.8 total cover of all perennials (1.e., the sum total of canopy and subcanopy cover) and canopy ground cover in the transect measurements was only 3.2%, indicating that few individuals were growing beneath the canopies of taller plants. Sampled vegetation, including 20 species of sampled perennial plants, was largely less than 10cm in height. The gray-white evergreen cushion plant Eriogonum kennedyi had_ the highest cover, a mean of 17.2%. It comprised half the cover in plots | and 4. Plants of E. kennedyi, which often were present next to exposed large rocks, were typically 2—5 cm tall, and the largest mounds were up to 50 cm across. The bright green cushion plant Phlox diffusa was also common and a codominant in plots 2 and 3 (Table 1). It was 2-8 cm tall, and large individuals were 30-40 cm across. The silvery- leaved, low cushion Lupinus breweri was common in only plot 3 and the bluish gray cushion Astragalus purshii was encountered only infre- quently. Low shrubs (<10—15 cm in height) of Chrysothamnus viscidiflorus were present but widely scattered in the fellfield plots, whereas fairly dense and much taller stands of this species occurred on adjacent slopes in deeper soils. Erect herbs, such as Silene bernardina, the hemipar- asites Castilleja applegatei subsp. martinii and Pedicularis semibarbata, Penstemon speciosus, and perennial grasses (e.g., Elymus elymoides and Poa spp.) were often encountered on the slopes but comprised very low total cover. Perennial grasses sometimes appeared to use Eriogonum kennedyi as a nurse plant, and /vesia 46 MADRONO a Fic. 3. [Vol. 55 SEM photograph of the leaf trans-section of Lupinus breweri var. bryoides, illustrating the typical amphistomatic structure with isolateral mesophyll layers. The bar indicates 200 um. santolinoides with a basal rosette of leaves also commonly occurred on the edges of Eriogonum cushions. Geophytes (A//ium species and Calo- chortus venustus) were emerging from bare soil patches, but individuals were less common in June 2000 than were observed in early summer 1999 and July 2005 when these geophytes were abundant. Leaf Form and Anatomy of Cushion Plants Leaf blades of the cushion plants were very small, and mean blade area for the four species ranged from 4.2 to 13.6 mm’. Blades of Phlox diffusa were dark green, linear and somewhat needle-like with an acicular tip, and possessed widely spaced uniseriate trichomes, which were most conspicuous on the adaxial surfaces. In contrast, leaves of Eriogonum kennedyi had more vertically oriented, elliptic blades that were covered with whitish vestiture, consisting of twisted unicellular trichomes, and the blade was TABLE 2. often enrolled somewhat to the underside. The lanceolate or narrowly ovate to elliptic leaflets of the two leguminous cushion plants, Lupinus breweri (palmately compound) and Astragalus purshii (pinnately compound), were folded along the midvein and cupped upward. Although the canopy of L. breweri appeared very reflectant, the adaxial blade surface was green because the reflectivity was caused by the long, straight, dense abaxial trichomes. In contrast, the leaflets of A. purshii were uniformly covered with softer, straight unicellular trichomes on both adaxial and abaxial surfaces. All four cushion plant species had amphisto- matic leaves with isolateral organization of the mesophyll, and each possessed at least two layers of abaxial palisade chlorenchyma (Fig. 3, Ta- ble 2). Mean blade thickness ranged from 209 um — in E. kennedyi to 497 um in the thicker but extremely narrow leaf of P. diffusa. Primary | phloem fibers were present in midveins of all | species, but in addition P. diffusa had a zone of LEAF ANATOMICAL AND MORPHOLOGICAL TRAITS OF EIGHT COMMON FELLFIELD PERENNIALS FROM | THE SUMMIT AREA OF MOUNT PINOS. ND = no data collected. Leaf thickness (Um) Astragalus purshii var. tinctus 22) Castilleja applegatei 270-442 Eriogonum kennedyi 209 Lupinus breweri var. bryoides 298 Pedicularis speciosus 1728 Penstemon speciosus 432 Phlox diffusa 497 Silene bernadina 583 Leaf specific area | Stomatal form Mesophyll (m? kg" ') amphistomatic isolateral 92 amphistomatic isolateral ND amphistomatic isolateral 40.5 amphistomatic isolateral 70.9 amphistomatic isolateral ND amphistomatic isolateral ND amphistomatic isolateral 62:1 amphistomatic isolateral ND 2008] fibers along the two leaf margins and the fibrous leaf apex was produced as an extension of the midvein fiber. These fibers were highly developed and contributed greatly to leaf thickness. More- over, leaves of P. diffusa become somewhat spine- like when they senesce. Leaf specific area (LSA) was high for leaflets of A. purshii (92.5 m* kg~'). Means of LSA were very slightly lower for L. breweri (70.9 m?* kg™') and P. diffusa (62.1 m* kg™'), while the leaves of E. kennedyi had the lowest LSA (40.5 m* kg™'). In comparison, pines, with a very different anatomical structure, had low LSA values of 28.5 and 24.1 m’ kg! for Pinus flexilis and P. Jeffreyi, respectively. Ecophysiological Characteristics In early July 1999, the two legume species, A. purshii and L. breweri, had assimilation rates (Amax) Of 13.4 and 13.9 umol CO, m”’s"', respectively (Table 3). Stomatal conductance (g,) was 293 and 263 mmol CO, ms‘! for the two species. Phlox diffusa had a relatively low Amax Of 6.8 umol CO, m~’s“', but proportion- ally high value of 212 mmol CO, m~’ s‘! for g,. Eriogonum kennedyi had an assimilation rate of 8.2 umol CO, m~’s"', but a low g, rate of 74 mmol CO, m’ s_''. Higher values of A:E and A:g, and lower values of c;:c, ratio indicate higher water-use efficiency. Eriogonum kennedyi demon- strated the highest water-use efficiency by all three of these measures (Table 3). In early November 1999, the aboveground tissues of both P. diffusa and E. kennedyi as well as many other species were no longer active. Two other two fellfield cushion plants had positive but low rates of assimilation. Astragalus purshii had an assimilation rate of 3.8 umol CO, m’s'|, 28% of its July rate, while Lupinus breweri with an assimilation rate of 5.0 umol CO; m~’ s“! was at 36% of its July level. For the two pine species in early July, P. Jeffreyi was the most active with an assimilation rate of 11.2 umol CO, m ’ s"! (Table 3). Photo- synthesis was also moderately high in P. flexilis with a rate of 8.1 umol CO, m~’ s_'. There was little change in these rates four months later in early November. Pinus jeffreyi had an assimila- tion rate of 91% of its July rate while P. flexilis was at 96% of its previous rate (Table 3). Water- use efficiency did not increase significantly over this seasonal period in the pines. Overall, the water-use efficiencies by all three measures were significantly higher on both dates in the pines than in three of the four fellfield cushions. Only E. kennedyi had a broadly comparable level of _ water-use efficiency (Table 3). _ Light response curves measured in early August 1999 showed that P. diffusa and A. _purshii both achieved light saturation for photo- GIBSON ET AL.: FELLFIELD PLANTS OF MOUNT PINOS 47 synthesis at high levels of about 1500 umol photons m’s'! (Fig. 4a,b). At PFD gat, Amax for the silvery leaves of A. purshii was 18.71 umol CO, m-’s'!' when g, was 189 mmolm~®s‘', whereas Ayax Of the bright green, fibrous leaves of P. diffusa was only 6.54 umol CO; m~*’s'! when g, was 110 mmol m ’ s'!. Xylem water potentials of the two fellfield legumes, A. purshii and L. breweri, dropped to lower levels (higher water stress) between the July and November sample dates (Fig. 5). Eriogonum kennedyi maintained relatively constant water potentials of about —2.5 MPa over the two sample dates. Phlox diffusa (not photosyntheti- cally active in November) and the two pine species showed a reverse pattern of higher water potentials (lower water stress) in November. DISCUSSION Although the flora of the summit area of Mount Pinos is strongly Sierra Nevadan in its flora (Muns 1994), this massif has been climat- ically isolated from the Sierra Nevada since the Pleistocene. Moreover, the current climate regime of Mount Pinos makes the summit region much more seasonal in climate and more arid than comparable sites in the Sierra Nevada. This isolation and relative aridity can be seen in the presence of distinct subspecies and a number of desert elements. Examples of these are non- Sierran taxa such as Eriogonum kennedyi var. kennedyi, Phacelia mohavensis, Allium howellii var. clokeyi (a local endemic), Castelleja applega- tei subsp. martini, and Astragalus whitneyi var. whitneyi. The presence Pinus flexilis on Mount Pinos is unexpected based on its typical habitat distribu- tion. Whereas this species also can be found in small numbers on nearby Brush Mountain to the northwest of Mount Pinos and Frazier Peak to the east (Twisselman 1967), the nearest popula- tions to these occur at high elevations 130 km to the east in the San Gabriel Mountains of the Transverse Ranges (Thorne 1977). Pinus flexilis does occasionally occur as low as 2560 m in the higher Transverse and Peninsular Ranges, but is more typical of higher elevations of 2750 to 3300 m. Although the rainfall at the summit of Mount Pinos is far lower than that of comparable sites with P. flexilis in the Sierra Nevada and in the high Transverse and Peninsular Ranges, it is similar to that present in the White Mountains of California where P. flexilis and P. monophylla can be found growing in close proximity at about 3050 m elevation (Rundel, personal obser- vation). Pinus contorta subsp. murrayana (lodgepole pine), commonly associated with P. flexilis in the Sierra Nevada and higher Transverse and Penin- sular Ranges, is absent from Mount Pinos. Two [Vol. 55 ~ MADRONO 48 (L€0'0+) 6070 (p+) OTT (900+) 98°€ (O€ OF) 89°T (C+) 06 (660+) 67 01 TSQUISAON ¢ (p70';0+) L8€°0 (¢+) SOI (O1L'0+) €6°7 (TEOF) 88'E (O1+) 801 (680+) TET Aqn¢ g idadffal snug (9700+) 60¢°0 (C+) TTI (610+) ves (LO'0) 9¢'I (C=) 79 (1e0+) LLL TOQUIsAON ¢ (S1O'O+) TIS‘0 (C+) 06 (LI'0+) 97 (OC OF) TEE (c=) 06 (TEO+) 80°8 Aynf g syixa]f{ snuig (SIO'O+) S8L°0 (1+) ZE (90'°0+) TI (1S'0+) O19 (Sl) Cie (cr'O+) LL'9 Ane g psnffip Xo[Yd (1v0'0+) OIL'0 (p=) Iv (LTO+) I6'l (70+) €L°7 (8+) v1 (IS 0+) 70'S TOQUISAON ¢ (8100+) 9L9°0 (p=) SS (OL'O+) 68'1 (CT OF) EPL (pT+) €9T (670+) C6 El AIne 8 saplodsg “eA llamadsg SnuldnT (6700+) S90 (8=) IIT (97 0+) 917 (TCO) OBE (10+) vl (190+) €7'8 Ane 8 1Apauuay WUNUOSO] (TSO'0+) 689°0 (OI) 79 (cE 0+) 777 (610+) 8L'I (L’0+) 79 (8S'O+) 6L'E TOQUISAONT ¢ (S10';0+) $89°0 (C=) OV (90'0+) 871 (€8'0+) ES Ol (TT+) €6T (S8'0+) SEE] Ane g SNJIUL] “eA 11Yysand SnjDSDAISP oned °9:'0 S3:V ae (,_S -_W Q°H JowU) 7 (,_S -_UWI JourU) Ss (,_S -_W “QD Joum) **"P so1sadg ‘ONTVA YOR 19}Jv SosayJUsIed UI UMOYS 1B SIOILIO piepuels > suvspy ‘(Ones °9:'9 pure 3: fp) AOUDIOIJS ISN-19}BM SISULIVUL “(72 f) AQUDIOIJJS ISN-19]BM SNOsUPUPIsSUI “(7) 9}3e1 UONPIIdsuPd “(S3) DOURIONPUODS [ReUIO}s *(*PF) Gad porveinjzes ye 9}81 UONRIIUMISSe LUNTUTXPUL UBd|Y “AVIPTUUNY SANLIAI %ZI[—-O] YUM D.8I—-9] SBM JIN} eAdduID} JUSIQUI ¢ IOQUIDAON ‘ALIPIUINyY dANPIaI Wpp—-Op YUM D.TE —-LTZ SBM din eJadw9} JUSIQUIe g ANE :puIM OU ATTeIVUADSSO ‘VGY 77 HOYap oinssaid 1odea ‘SAYS ssa]pNoy]d pur uns ][[NJ JO SUONIPUOD JUSIQUIY “SONId LNNOP 4O LINWOAS FHL UVAN GNV CTaIS TIA ANIUTVANS V NIW 0L97-0€97 LV ‘AONVWYOG YALNIM ONIYALNA 666] “S YAAWAAON ANV ‘HLMOUD FJALLVLADAA JALLOY GNV ONIYAMOT MVdad ONIYNdG 6661 ‘8 ATAL NO GaANSVAWP SHAIOddS ANIG OME GNV SLNVIg NOIHSND AO SdIOddS UNOY YOA SOLLSIYALOVUVHD TVOIDOTOISAHGQ ‘€ ATIVE 2008] 20 15 10 Astragalus purshii var. tinctus “o € Ke) E = = 20 S 2 is = 15 . £ Phlox diffusa Q = 40 0 500 1000 1500 2000 PFD (umol m* s") Fic. 4. Light response curves of net photosynthesis in relation to photon flux density (PFD) irradiance for P. diffusa and A. purshii var. tinctus. common montane conifers from the Sierra Nevada and higher Transverse Ranges, Pinus lambertiana (sugar pine) and Calocedrus decur- rens (incense cedar), are present but limited to isolated local populations on Mount Pinos despite the apparent presence of appropriate habitats. The rates of assimilation measured for the four fellfield species on Mount Pinos were in the lower portion of the range of those measured in alpine areas in other parts of the world (Korner 1999). These values ranged from 6.8 to 13.9 umol CO; m ~s 'on Mount Pinos, can be compared with a range of 11.5 to 25.5 umol CO, m ’ s_' reported for eight fellfield species in the arid White Mountains of California (Rundel et al. 2005). With the exception of Eriogonum kennedyi, which had a very low value of c;:c, ratio in our study, White Mountain species generally had lower values of c;:c, ratio than those reported here for Mount Pinos, indicating higher water-use effi- ciency. The assimilation rates and stomatal conductances of the two pine species on Mount Pinos fall within the ranges of values previously reported for pines in the literature (Rundel and Yoder 1998). The November measurements of water poten- tial showed an expected decrease in water potential from July values for Astragalus purshii and Lupinus breweri following dry summer GIBSON ET AL.: FELLFIELD PLANTS OF MOUNT PINOS 49 0.0 = > e July 8 a o November 8 = -1.0 re a = 5 s = cy D 2.0 a g ¢ ¢ S & 3.0 : | = @ > 2 40 » NN \ & YO oe 2 oe ae” ve a Rw ow ae 500° O_o Lye Ro) oe Q 4 Y> of Fic. 5. Midday xylem water potential on July 8, and November 8, 2005, for Lupinus breweri var. bryoides, Astragalus purshii var. tinctus, Eriogonum Kennedyi, Phlox diffusa, Pinus flexilis, and P. jeffreyi in the summit study site at 2650 m on Mount Pinos. conditions. Water potential, however, remained relatively constant in Eriogonum kennedyi and increased in Phlox diffusa and the two pine species. Both E. kennedyi and P. diffusa were dormant in November and thus an absence of transpiration reduced water stress. We do not have sufficient data to fully interpret the reduced water stress in the pine species at a time when photosynthetic rates were not significantly differ- ent from July rates, but this likely relates to cooler temperatures in November that reduced transpirational water loss and increased water- use efficiency as seen in ratios of assimilation to transpiration. Many accounts of California plant life have noted that numerous genera present on moun- taintops also have congeners occurring in low- land deserts (Went 1948), and so desert-alpine comparisons should also be considered. Like nonsucculent dicotyledons of the nearby Mojave Desert (Rundel and Gibson 1996), the four fellfield species used for gas exchange measure- ments have leptophylls, i.e. minute leaves or leaflet blades. For desert plants, such microphyl- lous leaves have the adaptive benefit of staying close to ambient temperature, hence never reaching lethal high temperatures, because such leaves have thin boundary layers and high conductive heat exchange (Gibson 1996, 1998; Smith et al. 1997). At alpine and subalpine elevations, such leaves that track cool daytime temperatures would be below expected optimal temperatures for photosynthesis. Although cush- ion plants with full sun exposure throughout the day never experience high air temperature, the cushion canopy may become heated via radiation and conduction from warm soil surfaces and thus have leaf temperatures substantially above air 50 MADRONO temperatures. Temperatures of low, ground hugging dense mats can be 20°C or more above air temperature (Korner 1999; Rundel et al. 2005). Many such mats and cushions have densely packed leaves, making the operational boundary layer for gas exchange that of the canopy itself rather than individual leaves. Fellfield plant species on Mount Pinos have amphistomatic leaves with abundant, slender palisade cells on both side of the mesophyll. These traits appear to be those present in many desert plants worldwide (Gibson 1996, 1998). Amphistomaty helps by increasing instantaneous rates of assimilation by maximizing stomatal conductance to CO, during the gas phase of diffusion (Mott et al. 1982), and isolateral mesophyll tends to maximize diffusion of CO, in liquid phase by exposing a very high mesophyll cell membrane area per leaf surface area (Nobel 2005). When leaves are not severely water stressed (e.g. midmorning during weeks of peak growth under saturated PFD), maximal instan- taneous assimilation rates of desert annuals and perennials often reach or exceed 25 umol CO, m*s' (Gibson 1998), levels generally higher than those reported for alpine fell field species (Rundel et al. 2005). Critical studies are needed to determine why desert plants have higher assim- ilation rates than related alpine plant species, and also how congeners from the two habitats may differ in allocation of photosynthate to vegetative and reproductive growth, and to tissue mainte- nance. The causal factors explaining the relatively low elevation occurrence of subalpine fellfield and limber pine forest communities on Mount Pinos remain an open question for investigation. Fell- field plant communities are generally found on high elevation sites near or above treeline. These sites are typically exposed and dry, often windy and snow-free summits and ridges where snow- melt occurs rapidly and wet soils do not persist (Bliss 1985). Subalpine fellfield and treeline conifers rarely occur as low as 2800 m in the southern Sierra Nevada but are more character- istic of elevations well above 3500 m. The summit of Mount Pinos may well have these traits in most winters because of the relatively low amounts of snowfall and the exposed position of the mountain near the coast. However, in cold and wet winters, snowfall covers much of the summit area through the late spring into early summer, as occurred in 2005. One component of an explanation for the subalpine communities of Mount Pinos lies with the Massenerhebung, or mass-elevation, effect, which has been documented for the Alps of Central Europe (Barry 1982). Core areas of the central Alps have a higher elevation treeline than those present on peaks at the outer margins of the range. Greater solar radiation inputs in these core [Vol. 55 areas have been hypothesized to lead to warmer summer temperatures with reduced duration of snowfields, and thus longer growing seasons. This effect also predicts lower elevation treelines on isolated mountains compared with elevations on larger mountain ranges with greater mass. The summit of Mount Pinos is relatively isolated and sufficiently close to the coast to be influenced by maritime conditions and cloud cover, moderating solar radiation and temperatures at the summit. Moreover, growing season length likely is also reduced by the extreme summer aridity on Mount Pinos. Thus, moderate summer temperatures and drought-shortened growing season conditions would act to lower timberline. As noted above, scattered individuals of Pinus flexilis do occur at similar elevations in the San Gabriel, San Bernardino, and San Jacinto Mountains, but the primary range of the species in these ranges is above 2750 m. Detailed climatological measure- ments on the summit of Mount Pinos could help resolve this interesting issue. ACKNOWLEDGMENTS We are grateful to District Rangers Mark Bethke and John Kelly of the Los Padres National Forest for allowing us to perform this research at the summit of Mount Pinos, and Judy King for her assistance in the vegetation measurements. REFERENCES CITED BARRY, R. G. 1982. Mountain weather and climate, 2nd ed. Routledge, London. BILLINGS, W. D. AND H. A. MOONEy. 1968. The | ecology of arctic and alpine plants. Biological Review 43:481—529. BLiss, L. C. 1985. Alpine. Pp. 41-65 in B. F. Chabot and H. A. Mooney (eds.), Physiological ecology of North American plant communities. Chapman and Hall, London. DIBBLE, D. W. 1982. Regional geology of the Transverse Ranges, Province of Southern Califor- | nia. Pp. 7—26 in D. L. Fife and J. A. Minch (eds.), Geology and Mineral Wealth of the California Transverse Ranges. South Coast Geological Soci- ety, Santa Ana. GIBSON, A. C. 1996. Structure-function relations of warm desert plants. Springer, Berlin. . 1998. Photosynthetic organs of desert plants. BioScience 48:91 1—920. HALL, C. A. (ed). 1991. Natural history of the White- Inyo Range, eastern California. University of. California Press, Berkeley. HICKMAN, J. C. (ed). 1993. The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. | KELLOGG, K. S. 2004. Thrust-induced collapse of mountains — an example from the “Big Bend” region of the San Andreas Fault, Western Trans- verse Ranges, California. USGS Science Investiga- tions Report 2004-5206. KORNER, C. 1999. Alpine plant life: functional ecology of high mountain ecosystems. Springer-Verlag, | Berlin. 2008] MAJor, J. AND D. W. TAYLOR. 1977. Alpine. Pp. 601— 675 in M. G. Barbour and J. Major (eds.), Terrestrial vegetation of California. John Wiley, New York. Mork, B. 2005. Summary of the Month. California Climate Watch. January 2005. . Mott, K. A., A. C. GIBSON, AND J. W. O’LEARY. 1982. The adaptive significance of amphistomatic leaves. Plant, Cell and Environment 5:455—460. Muns, B. 1994. Flora of Mount Pinos summit area. Ventura and Kern Counties, California, . NOBEL, P. S. 2005. Physicochemical and environmental plant physiology. Elsevier Academic, Amsterdam. RUNDEL, P. W. AND A. C. GIBSON. 1996. Ecological communities and processes in a Mojave Desert ecosystem. Cambridge University Press, Cam- bridge. ; , AND M. R. SHARIFI. 2005. Plant functional groups in alpine fellfield habitats of the GIBSON ET AL.: FELLFIELD PLANTS OF MOUNT PINOS a White Mountains, California. Arctic Antarctic and Alpine Research 37:358—365. AND B. J. YODER. 1998. Ecophysiology of Pinus. Pp. 296-323 in D. M. Ricrdson (ed.), Ecology and biogeography of Pinus. Cambridge University Press, Cambridge. SMITH, S. D., R. K. MONSON, AND J. E. ANDERSON. 1997. Physiological ecology of North American desert plants. Springer, Berlin. THORNE, R. F. 1977. Montane and subalpine forests of the Transverse and Peninsular Ranges. Pp. 537-557 in M. G. Barbour and J. Major (eds.), 7errestrial vegetation of California. John Wiley, New York, NY. TWISSELMANN, E. C. 1967. A flora of Kern County, California. Wasmann Journal of Biology 25:1— 395; VOGEL, R. J. AND B. C. MILLER. 1968. The vegeta- tional composition of the south slope of Mt. Pinos, California. Madrono 19:225—234. WENT, F. W. 1948. Some parallels between desert and alpine floras in California. Madrono 9:241—249. MADRONO, Vol. 55, No. 1, pp. 52-59, 2008 MORPHOLOGICAL TRAITS AND INVASIVE POTENTIAL OF THE ALIEN EUPHORBIA TERRACINA (EUPHORBIACEAE) IN COASTAL SOUTHERN CALIFORNIA ERIN C. RIORDAN Department of Ecology and Evolutionary Biology, University of California, Los Angeles, CA 90095 eriordan@ucla.edu PHILIP W. RUNDEL Department of Ecology and Evolutionary Biology and Center for Embedded Networked Sensing, University of California, Los Angeles, CA 90095 rundel@biology.ucla.edu CHRISTY BRIGHAM AND JOHN TISZLER National Park Service, Santa Monica Mountains National Recreation Area, 401 West Hillcrest Drive, Thousand Oaks, CA 91360 ABSTRACT Euphorbia terracina L., also known as terracina spurge, is a Mediterranean Basin perennial that has recently become invasive in southern California and is actively spreading at a virtually exponential rate along coastal areas in Los Angeles County. The National Park Service (NPS) is undertaking measures to treat and control further spread of current populations, but little is known about the plant’s ecology and impact on native plant communities. This study reviewed the existing information on E. terracina and investigated populations established in Solstice Canyon in coastal Los Angeles County. Populations of E. terracina were compared in two different habitats in Solstice Canyon: in an open site along an old road and a shaded riparian site subject to past disturbance. Both open- disturbed and shaded sites had high aboveground biomass densities, with the highest density in the open, disturbed site. Sites differed in biomass allocation and specific leaf area (SLA) between sites, with plants at the open site having significantly lower specific leaf area than those at the shaded site. Open site plants also had high SLA compared to native coastal sage scrub species. Euphorbia terracina produces large quantities of seeds that do not show dormancy. Seeds germinated well under low light intensities without mechanical or chemical treatment. Euphorbia terracina possesses numerous traits — success in disturbed sites, phenotypic plasticity, high SLA compared to native species, high reproductive output, and seeds lacking dormancy — that have been associated with invasive species and likely contribute to both its success and the difficulty in treatment and control of established populations. Key Words: biomass allocation, Euphorbia terracina, germination, invasive, phenotypic plasticity, specific leaf area. Euphorbia terracina is an invasive herbaceous perennial from the Mediterranean Basin that has recently established in coastal areas of Los Angeles County in southern California. The plant’s prior invasive history and wide distribu- tion in Australia, where —. terracina has natural- ized to and invaded disturbed areas of nutrient- poor sandy soils in the coastal heath of Victoria, South Australia, and New South Wales (Randall and Brooks 2002), suggest E. terracina 1s likely to be a successful invader elsewhere in the world (Reichard and Hamilton 1997). In south- ern California, however, FE. terracina is not restricted to coastal sandy soils and has natural- ized to both coastal bluff and sage scrub habitats. Although multiple points of origin are possible, a major center of its spread appears to be the central Malibu coast. Euphorbia terracina has since expanded its range west and east along the Malibu coast and up into canyons of the Santa Monica Mountains with remarkable aggressive- | ness, often forming dense monocultures in disturbed areas. It has been reported along Highway 150 in Ventura County and from. coastal areas of the Palos Verdes Peninsula in southern Los Angeles County (P. Rundel, personal observation). Despite its current abundance in coastal southern California, E. terracina is not listed in the Jepson Manual (Hickman 1993) and little is known about the species. The first known collection of E. terracina was made at the El Segundo Dunes near the Los Angeles Interna- tional Airport in 1987 (Hrusa et al. 2002) and the > National Park Service (NPS) only became aware of its spread in the late 1990’s. The objective of this study is to provide a review of existing literature on current E. terracina distribution and 2008] to develop a better understanding of the ecology of E. terracina to aid in the control and management of populations in coastal southern California. Field and laboratory studies were implemented to investigate the invasive potential of E. terracina in southern California. Previous studies have proposed that morpho- logical plasticity in response to varying environ- mental conditions enables non-native species to successfully invade a range of habitats (Baker 1974; Schweitzer and Larson 1999; Daehler 2003; Parker et al. 2003). Euphorbia terracina plants were compared between two sites with differing disturbance and light conditions in Solstice Canyon, Malibu California: an open-disturbed habitat along roadsides and trails and a riparian woodland habitat. Successfully established plants were expected to differ in biomass allocation and specific leaf area (SLA) between these sites. In addition, reproductive characteristics of E. terra- cina were investigated to better understand the plant’s success as an invasive. Invasive species often lack seed dormancy, readily germinating without chemical treatments (Reichard and Hamilton 1997). Thus, Euphorbia terracina was expected to require little or no pre-germination treatment for successful germination. MATERIALS AND METHODS Species Ecology Euphorbia terracina, also called terracina spurge, false caper, and Geraldton carnation weed (in Australia), is an herbaceous perennial in the Euphorbiaceae. Its native habitat is dry, sandy soils along the coasts of islands and adjacent mainland of the Mediterranean Basin, extending northwest from the Iberian Peninsula to the eastern Mediterranean in Croatia and Greece (Tutin et al. 1968). Mature plants grow to 1 m and contain a noxious, milky sap common to many Euphorbia species. Plants are simple, branched with 0—5 auxiliary shoots arising from the base of the plant, and grow from a root crown with a long taproot (Tutin et al. 1968). The leaves are dark green, minutely serrate, and alternate. Each shoot produces 4 to 5 equal-length pedicels arranged in clusters at the stem apex. Euphorbia terracina has. small, inconspicuous yellow to green flowers, which are surrounded by five pointed, oval bracts. Each flower produces a small three-lobed fruit with each lobe containing one seed. Euphorbia terracina has a great reproductive capacity, as each shoot is capable of producing up to 200 seeds (C. Brigham, personal communication). Seeds are dispersed explosively from fruits, traveling up to 5 m from the plant (C. Brigham, personal communication). Flowering begins in spring and seeds are pro- duced in summer. Although plants may die back RIORDAN ET AL.: INVASIVE EUPHORBIA TERRACINA Bs, with water stress in late summer, new stems surface from the original root crowns at the start of the next growing season (P. Rundel, personal communication). Seeds may germinate in any month of the year and young plants are capable of branching repeatedly to form dense clusters before spring flowering. While E. terracina does not reproduce vegetatively, plants cut at their base prior to setting seed rapidly grow new stems using carbohydrates stored in their taproot (P. Rundel, personal communication). Field Sites Solstice Canyon in Malibu, California (34°S, 118°W) appears to be a center of introduction of E. terracina into coastal southern California and has large E. terracina populations. It receives a mean annual rainfall of about 330 mm, with considerable interannual variation. Temperature conditions are mild throughout the year, with mean maximum summer temperatures of about 21°C and mean winter minimum temperatures of about 10°C. Field sites were selected within Solstice Canyon to compare populations of E. terracina in an open-disturbed and a shaded riparian woodland habitat. The open site (a 20 X 30 m plot) was located on a gentle, dry slope in full sun, having high disturbance from past mowing and bordering hiking trails. The shaded study site was a 25m X 5m plot in a riparian woodland corridor about 400 m west of the open site; 1t was located 10 m from the creek and approximately 25 m from any trail. (The study plots differed in size because E. terracina populations differed in size between sites.) All field data and observations were collected be- tween May 10 and June 23, 2004. Biomass Allocation Comparisons Euphorbia terracina aboveground biomass was harvested from ten randomly selected 0.3 X 0.3m plots at each site. The allocation of E. terracina aboveground shoot biomass to repro- ductive, photosynthetic, and structural tissue was then determined for both sites using five repre- sentative plants from each of the 0.30 < 0.30 m samples. These five plant samples were pooled to generate one sample per plot (n = 10). Each shoot was separated into three organ types. All fruits, developed buds, and flowers were classified as reproductive organs; flower bracts and leaves as photosynthetic organs; and remaining stems as structural organs. The number of flowers, devel- oped buds, and fruits were counted for each plant. All plant material was dried at 85°C for 48 hrs prior to weighing. The relative biomass allocation of each organ component per shoot was calculated for open and shade populations. Means for sun and shade sites are reported as 54 MADRONO mean +1 standard deviation (SD). Student’s t- tests were used to determine if there was a significant difference in the allocation of biomass between plants in the open and shade popula- tions. Student’s t-test was also used to compare estimations of total aboveground biomass _ be- tween sites. The experiment-wide error rate was adjusted using the Bonferroni method to take into account elevated error caused by three comparisons on the biomass data (Gotelli and Ellison 2004). Based on the Bonferroni correc- tion, P-values of 0.01667—0.05 were considered suggestive, rather than significant. All statistics were preformed using SAS Institute, Inc. Stat- view® (version 5.0) software. Leaf morphology also was compared between open and shade plants. On May 24, 2004, all leaves from five shoots were randomly harvested from each site (Open site n = 60, shade site n = 64). The leaf area for both open and shade plants was measured using a LI-COR, Inc. Leaf Area Meter, model LI-3100. The leaves were then oven-dried and weighed to determine their specific leaf area (SLA). A LI-COR Solar Monitor, model LI-1776 was used to measure light intensities at E. terracina canopy height and at ground level below the canopy for both sites. Measurements were taken at 14:20 for the open site and 14:40 for the shade site on June 23, 2004. The open site had a light intensity of 1800+190 umol m-’ s'' (mean +1 SD) at cano- py height and 720+340 umol m’ s'' at ground level. The shade site had lower light intensities of 70+31 umolm~’s'' at canopy height and 41+17 umol m ’ s‘'! at ground level. Germination Experiments Mature seeds were collected directly from representative plants in the open site to investi- gate seed germination. Germination experiments were performed within one week of seed collec- tion. Seeds were germinated in the lab under low- light intensity of 2 umol m ’ s ' at a temperature of 24°C. Light intensities were measured using a LI-COR Solar Monitor, model LI-1776. To determine if E. terracina seeds possess seed dormancy preventing germination, aside from the mechanical barrier of the seed coat, two seed treatments were prepared; a scarification treat- ment and a control treatment with no scarifica- tion. Seeds were scarified immediately prior to the germination experiments using a fine, sharp needle to puncture a small hole in the seed coat. There were five replicates of 20 seeds each per scarification treatment. Seeds were placed on wet filter paper in 9-cm diameter petri dishes. Moisture in the dishes was maintained constant with de-ionized water. Seeds were recorded as having germinated once the radicle emerged from the seed coat. The number of germinated seeds [Vol. 55 per replicate was recorded every 24 hr for up to 21 d, or until the number of germinated seeds per treatment remained constant for over 48 hr. Mean percentage germination for each treatment (n = 5) were compared using a t-test (SAS Institute Statview® [version 5.0]), and values reported as mean +1 SD. RESULTS While E. terracina had high aboveground biomass at both open and shade sites, above- ground biomass was significantly higher at the open site (Fig. 2) (t = 11.379, DF = 98, P < 0.001). The mean sum of structural, photosyn- thetic, and reproductive biomass was 324.2+ 71.8 gm? (n = 10) in the open site compared to 118.8+43.1 gm? (n = 10) in the shade site (t = —0.661, DF = 18, P < 0.001) (Fig. 2). The considerable difference in the mean sum of biomass between open and shade sites can be contributed to the large difference in structural biomass between sites (open: 239.8+56.8 g m * and shade 76.0+35.5 g m ’) (Fig. 2). The differ- ence in biomass of photosynthetic organs between open and shade sites, however, was only sugges- tive (t = —2.275, P = 0.0362 (adjusted a = 0.0167)). Euphorbia terracina also showed differ- ences in the relative biomass allocation to different organs between sites (Fig. 3). Shade site plants had a significantly greater relative photo- synthetic tissue biomass than open site plants. Sun plants had significantly higher relative biomass allocation to structural (t = —5.372, DF = 98, P < 0.001) and reproductive (t = —7.124, DF = 98, P < 0.001) tissues than the shade plants because of their taller stature. There was not a significant difference in the number of flowers (P > 0.05), but the relative allocation of biomass to reproductive tissues in the sun population was three times higher than that in shade populations. Open site © plants also had a significantly lower specific leaf area of 0.192+0.014 cm* mg ' compared to shade | plants with 0.464+0.046 cm? mg"! (t = 44.193, | DF = 122, P < 0.001) (Fig. 4, Table 1). The germination experiments showed that E. : terracina seeds readily germinate under con- | trolled laboratory conditions. Scarification had | no significant effect on final percent seed | germination under low-light (2 umol m~ s“‘) | laboratory conditions (t = 1.00, DF = 8, P > 0.1). Control treatments had a final germination percentage of 100+0.00% and scarified treat-— ments have a final germination of 98+4.47%. Scarification did have an effect on the rate of germination, as scarified seeds germinated faster | than the control seeds (Fig. 5). The number of | days required to reach maximum percentage of germination was 6.6+0.89 d in scarified treat-— ments (n = 5) and 12.6+1.52d in control : treatments (n = 5). RIORDAN ET AL.: INVASIVE EUPHORBIA TERRACINA my) Los Angeles County Palos Verdes Peninsula 2008] Santa Barbara County Ventura County ~_ santa Monica Mountains Fic. 1. Distribution of Euphorbia terracina in Southern California. DISCUSSION In Solstice Canyon, E. terracina has become widely and successfully established in both an open, disturbed and a shaded, riparian woodland 400 (7 Open Gm Shade Mean Biomass (g m~2) NO w [o) oO [o) [o) = fo) (o) Structural Photosynthetic Reproductive Total Fic. 2. The mean total, structural (stems), photosyn- thetic (leaves and bracts), and reproductive (flowers and fruits) biomass per m? of Euphorbia terracina plotted + 1 standard error for open (n = 10) and shade (n =10) sites. *** Indicates P < 0.001 significant difference between open and shade values. site, indicating a capacity to invade a range of habitats in southern California. The notably higher E. terracina total and reproductive bio- mass in the open site suggest high light availabil- ity and disturbance provide better conditions for E. terracina success. Disturbance often promotes the invasion of non-native species that are well adapted to high light environments associated with disturbance (Hobbs and Huenneke 1992; Burke and Grime 1996). Frequent mowing has been employed to control a variety of weed problems at the open site, potentially stimulating E. terracina biomass production. Low relative allocation to leaves and high reproductive output, as exhibited in E. terracina at the open site, have been interpreted as responses to high light environments (Givinish 1988; Rice and Bazzaz 1989). However, because biomass collection 1s pseudoreplicated (Hurlbert 1984) between single open and shade sites, the study cannot definitive- ly determine if light availability was responsible for the greater E. terracina establishment and success at the open site. For example, a competitive advantage for other resources, such as nutrients or water, compared to native species could also influence E. terracina success (Crawley et al. 1999; Callaway and Aschehoug 2000; Baruch et al. 2000). The ability of E. terracina to form dense monocultures given optimal high light conditions makes it a considerable threat to native flora in 56 MADRONO (] Open Gls Shade 0.8 0.6 0.4 Relative aboveground biomass 0.2 0.0 Structural Photosynthetic Reproductive Fic. 3. The mean relative biomass + | standard error per shoot in Euphorbia terracina, plotted for both open (n = 50) and shade (n = 50) plants. *** Indicates P < 0.001 significant difference between open and shade values. coastal southern California. Before recent control efforts by NPS, E. terracina covered five of the 90 ha in Solstice Canyon, forming dense mono- cultures in disturbed areas and along the terraces of the riparian corridor, excluding native plants from reoccupying disturbed sites, and thereby altering these costal sage scrub and riparian woodland communities (C. Brigham, personal communication). When compared to a classic study in re-colonization and productivity of Lotus scoparius, a native subshrub that forms a significant element of post-fire successional veg- etation in California shrublands (Nilsen and Schlesinger 1981), open site E. terracina popula- tions in Solstice Canyon had greater total structural, photosynthetic, and reproductive aboveground biomass per square meter (Ta- ble 2). Impacts on species richness and biodiversity of native flora were not surveyed in this study, but field observations suggest that E. terracina had a large negative impact once established in the area, strongly altering the natural plant community composition. Although competition was not addressed in this study, biomass can be predictive of competitive ability (Gaudet and Keddy 1988). A greater biomass of E. terracina compared to L. scoparius may indicate an ability to establish and propagate well in competition with native species, however, biomass data from additional species and manipulative experiments are necessary to evaluate this hypothesis. Phenotypic plasticity, as illustrated by the considerable difference in phenology, biomass allocation, and leaf morphology between open [Vol. 55 | 35 30 25 o c 20 a) S oO © 15 LL 10 5 0 L 3 0.15 020 025 030 035 040 045 050 055 060 Specific leaf area (cm? mg”) Fic. 4. Histogram of specific leaf area (cm* mg “') of Euphorbia terracina for open (n = 60) and shade sites (n | = 64). and shade plants, may explain the success of E. terracina in two strikingly different habitats. Field observations established that open site plants began flowering and fruiting 3-weeks earlier in the season, illustrating plasticity in phenology that may have contributed to the | difference in relative reproductive biomass allo- cation. Greater relative biomass allocation to photosynthetic structures and higher specific leaf area in shade plants likely reflect a growth | strategy for light capture in a low light environ- ment (Baruch et al. 2000; Monaco et al. 2005). In © contrast, low SLA, as evident in the open site | plants, has been correlated with a longer leaf life | span, lower photosynthetic rates, low soil mois- ture availability, and high light environments, and thus stress tolerance (Givnish 1988; Reich et | al. 1992; Reich et al. 1999). This suggests E. | terracina is able to respond to both the low light of the riparian woodland corridor and the water | stress and higher light intensity characteristic of a sunny coastal slope. As this study is limited to one sun and shade site, further investigation 1s needed to determine to what extent phenotypic plasticity in response to light may account for the observed differences in populations. Nonetheless, phenotypic plasticity in response to environmen- | tal conditions has been proposed to explain an | exotic species’ successful invasion into a broad | range of habitats (Baker 1974; Schweitzer and — Larson 1999: Daehler 2003; Parker et al. 2003). — Euphorbia terracina represents a_ substantial threat to a range of habitats in coastal southern © California. High SLA compared to native species has also been associated with invasive species (Baruch and Goldstein 1999; Smith and Knapp 2001; Durand | and Goldstein 2001; Grotkopp et al. 2002; Lake and Leishman 2004). Specific leaf area can be » thought of as the investment per unit of light- capture surface deployed, and is a critical trait in 2008] TABLE 1. RIORDAN ET AL.: INVASIVE EUPHORBIA TERRACINA aT EUPHORBIA TERRACINA SPECIFIC LEAF AREA OF OPEN AND SHADE PLANTS. Values are reported as mean +1 standard deviation, n = 60 for open, n = 64 for shade. Leaf type Area (cm7’) Weight (mg) Open 1.140.34 6.0+1.7 Shade 5.3+1.50 123726 the carbon fixation strategy of plants. Species with a high specific leaf area have a shorter investment return rate and greater potential for fast growth. The specific leaf area of sun populations of E. terracina (0.192 cm* mg!) is relatively high in comparison to coastal sage scrub species such as Salvia mellifera (0.106 cm? mg_') Eriogonum cinereum (0.069 cm? mg‘), Mirabilis californica (0.114 cm? mg™'), and En- celia californica (0.181 cm? g-' mg™') (R. Sharifi [UCLA], unpublished data). In addition to leaf traits, the relatively low canopy height and herbaceous growth form of E. terracina are traits associated with a short juvenile phase, which is characteristic of invaders in disturbed sites (Reymanek 1996; Lake and Leisham 2004). Euphorbia terracina also displays a number of reproductive traits typical of invasive weeds. High reproductive output and small seed mass are associated with rapid colonization and are especially indicative of invaders in_ physically disturbed, open sites (Rejmanek and Richardson 1996; Reichard and Hamilton 1997; Lake and Leishman 2004). Euphorbia terracina is capable of producing over 200 seeds per shoot (C. Brigham, personal observation). The relative allocation of biomass to reproductive tissues is approximate 10%, a value quite close to that reported for Lotus scoparius, an effective native colonizer of burned chaparral slopes (Nilsen and Schlesinger 1981). While E. terracina does not spread through —e— Scarified -O-- Control S Cc Ao) o £ E oO O Time (days) Fic. 5. Mean germination rate (%) plot for lab low —%7 light (2 umol m~? s~') scarification and control treat- ments. The mean percent is plotted +1 standard error for each day (n = 5). Specific leaf area (cm? mg ') 0.192+0.014 0.464+0.046 vegetative reproduction, it 1s capable of rapid resprouting and growth in response to cutting, another trait that is associated with invasives (Baker 1974). Soil movement by vehicles and earth moving equipment has probably played a significant role in E. terracina’s spread because of its ability to colonize disturbed sites. Informal observations (P. Rundel, personal observation) suggest that the infestation of seeds into potted plants at local nurseries may be responsible for spread to private property in Los Angeles County. Ant dispersal may also be a factor in the local spread of E. terracina, but was not addressed in this study. Many Euphorbia species, including E. terracina, have seeds with elaiosomes that are ant dispersed (Espadaler and Gomez 1994; Espadaler and Gomez 1996). While this trait is not usually associated with invasiveness (Lake and Leishman 2004), the closely related E. esula is ant dispersed in its introduced range (Pemberton 1988). Euphorbia terracina seeds also display a germi- nation pattern typical of many weedy species. Seeds requiring little or no pre-germination treatment can also be indicative of invasive species (Reichard and Hamilton 1997). The high rates of germination in both control and scarified treatments suggest that EF. terracina has no deep seed dormancy. Even though scarification was not necessary for germination to occur, it influenced the germination rate, scarified seeds germinated faster than control seeds. The slight lag in germination of control seeds is probably the result of the physical resistance of the seed coat and not related to seed dormancy. Euphorbia terracina may also have a substantial soil seed bank, as field observations suggest seed life in the soil is as long as 5—8 yrs or more (C. Brigham, personal communication). With its high number of invasive traits and successful introduction into various habitats, E. terracina poses substantial threat to the native flora of southern California. Phentoypic plastic- ity in biomass allocation and SLA suggest E. terracina could be capable of invading a range of habitats in southern California. Its high SLA compared to native species, herbaceous growth form, reproductive output, rapid regeneration, and success in disturbed sites are all characteristic of weedy, invasive species. Further studies are needed to determine the extent of the impact E. terracina has on the native flora and should address its competitive ability for resource 58 MADRONO TABLE 2. 5 [Vol. 55 | ABOVEGROUND BIOMASS OF EUPHORBIA TERRACINA COMPARED TO LOTUS SCOPARIUS. Biomass for — each species is measured as mean g m * and L. scoparius data is compiled from Nilsen and Schlesinger (1981). Percent of total biomass is shown in parentheses. Species Structural Lotus scoparius 105.6 (64.2) E. terracina (open) 239.8 (73.9) E. terracina (shade) 76.0 (64.0) acquisition compared to native species. The recent and rapid expansion of E. terracina suggests that control may not be possible without immediate attention by management agencies. The ability of E. terracina to thrive under high disturbance conditions, however, makes it par- ticularly difficult to eradicate as many control treatments, such as mechanical removal and mowing, create conditions that favor further establishment. For E. terracina, suppression of the seed germination from the soil seed bank or mulching could be potential control methods, but their success is unknown. Effective management needs to incorporate information gained from ecological studies and must address E. terracina’s ability to readily germinate and thrive under high light conditions. Coastal southern California’s combination of unique native vegetation, high anthropogenic disturbance, and high number of successful non-native invasions makes invasive species management a high priority. The establishment of invasives in areas of high disturbance has inhibited restoration efforts of native plant communities (Eliason and Allen 1997). Further- more, urbanization in the Santa Monica Moun- tains is expected to increase dramatically in the next 25 yrs (Swenson and Franklin 2000), poten- tially increasing both non-native plant invasions into coastal scrub habitats and the need for ecological restoration in areas disturbed by development. It is therefore important to develop control methods for invasive species. The suc- cessful treatment of currrent E. terracina popu- lations may prove significant in managing other invasive plant species in the area and protecting already threatened habitats. ACKNOWLEDGMENTS The senior author would like to thank the National Park Service for allowing her to carryout this research in the Santa Monica Mountains National Recreation Area. Rasoul Sharifi of UCLA aided greatly in the experimental work on this project and provided unpublished data on specific leaf areas of coastal sage scrub species. LITERATURE CITED BARUCH, Z. AND G. GOLDSTEIN. 1999. Leaf construc- tion cost, nutrient concentration, and net CO, assimilation of native and invasive species in Hawaii. Oecologia 121:183—192. Photosynthetic Reproductive Total 42.1 (25.6) 16.7 (10.2) 164.4 51.8 (16.0) 32.7 (04) 324.3 39.1329) 3.7 (3.1) 118.8 , R. R. PATTISON, AND G. GOLDSTEIN. 2000. Responses to light and water availability of four invasive Melastomataceae in the Hawaiian islands. International Journal of Plant Sciences 161: 107-118. BASKIN, J. M. AND C. C. BASKIN. 1980. Ecophysiology of seed dormancy in seeds of Ambrosia artemisii- folia. Ecology 61:475-480. BAKER, H. G. 1974. The evolution of weeds. Annual Review of Ecology and Systematics 5:1—24. | BURKE, M. J. W. AND J. P. GRIME. 1996. An experimental study of plant community invisibility. — Ecology 77:776—790. CALLAWAY, R. M. AND E. T. ASCHEHOUG. 2000. Invasive plants versus their new and old neighbors: © a mechanism for exotic invasion. Science 290:521—523. CRAWLEY, M. J., S. L. BROWN, M. S. HEARD, AND G. R. EDWARDS. 1999. Invasion-resistance in | experimental communities: species richness or species identity? Ecology Letters 2:140—148. DAEHLER, C. C. 2003. Performance comparisons of co- | occurring native and alien plants: implications for conservation and restoration. Annual Review of Ecology Evolution and Systematics 34:183-—211. DURAND, L. Z. AND G. GOLDSTEIN. 2001. Photosyn- © thesis, photoinhibition, and nitrogen use efficiency | in native and invasive tree ferns in Hawaii. Oecologia 126:345—354. ELIASON, S. A. AND E. B. ALLEN. 1997. Exotic grass competition in suppressing native shrubland re- establishment. Restoration Ecology 5:245—255. ESPADALER, X. AND C. GOMEZ. 1994. Seed dispersal by ants in Euphorbia characias and Euphorbia | nicaeensis (Euphorbiaceae). Ecologia Mediterranea 20:51-59. AND . 1996. Seed production, predation and dispersal in the Mediterranean myrmecochore Euphorbia characias (Euphorbiaceae). Ecography 19:7-15. GAUDET, C. L. AND P. A. KEDDy. 1988. A compar- ative approach to predicting competitive ability from plant traits. Nature 334:242—243. GIVNISH, T. J. 1988. Adaptation to sun and shade-a whole plant perspective. Australian Journal of. Plant Physiology 15:63—92. GOTELLI, N. J. AND A. M. ELLISON. 2004. A primer of | ecological statistics. Sinauer Associates, Inc, Sun- derland, MA. | GROTKoppP, E., M. REJMANEK, AND T. L. Rost. 2002. Toward a causal explanation of plant invasiveness: seedling growth and life history strategies of 29. pine (Pinus) species. American Naturalist 159: 396-419. HICKMAN, J. C. (ed.). 1993. The Jepson manual: higher plants in California. University of California Press, Berkeley, CA. 2008] Hosss, R. J. AND L. F. HUENNEKE. 1992. Disturbance, diversity and invasion: implications for conserva- tion. Conservation Biology 6:324—337. Hrusa, F., B. ERTTER, A. SANDERS, G. LEPPIG, AND E. DEAN. 2002. Catalogue of non-native vascular plants occurring spontaneously in California be- yond those addressed in The Jepson Manual—Part I. Madrono 49:61—98. HURLBERT, S. H. 1984. Pseudoreplication and the design of ecological field experiments. Ecological Monographs 54:187—211. LAKE, J. C. AND M. R. LEISHMAN. 2004. Invasion success of exotic in natural ecosystems: the role of disturbance, plant attributes and freedom from herbivores. Biological Conservation 117:215—226. Monaco, T. A., D. A. JOHNSON, AND J. E. CREECH. 2005. Morphological and physiological responses of the invasive weed /satis tinctoria to contrasting light, soil nitrogen and water. 2005. Weed Research 45:460—466. NILSEN, E. T. AND W. H. SCHLESINGER. 1981. Phenology, productivity and nutrient accumulation in the post-fire chaparral shrub Lotus scoparius. Oecologia 50:217—224. PARKER, I. M., J. RODRIGUEZ, AND M. E. LOIK. 2003. An evolutionary approach to understanding the biology of invasions: local adaptations and general- purpose genotypes in the weed Verbascum thapsus. Conservation Biology 17:59—72. PEMBERTON, R. W. 1988. Myrmecochory in the introduced range weed leafy spurge Euphorbia esula. American Midland Naturalist 119:431—435. RANDALL, R. AND K. BROOKS. 2002. Geraldton carnation weed. Euphorbia terracina. L. Euphor- biaceae. Environmental Weed Action Network, Perth. REICH, P. B., D. S. ELLSWORTH, M. B. WALTERS, J. M. VOSE, C. GRESHAM, J. C. VOLIN, AND W. D. RIORDAN ET AL.: INVASIVE EUPHORBIA TERRACINA a BOWMAN. 1999. Generality of leaf trait relation- ships: A test across six biomes. Ecology 80:1955—1969. , M. B. WALTERS, AND D. S. ELLSWORTH. 1992. Leaf life-span in relation to leaf, plant and stand characteristics. Ecological Monographs 62:365— 392. REICHARD, S. H. AND C. W. HAMILTON. 1997. Predicting invasions of woody plants introduced into North America. Conservation Biology 11: 193-203. REJMANEK, M. 1996. A theory of seed plant invasive- ness: The first sketch. Biological Conservation 78:171-181. AND D. M. RICHARDSON. 1996. What attributes make some plant species more invasive? Ecology 77:1655-1661. Ric—E, S. A. AND F. A. BAZZAZ. 1989. Growth consequences of plasticity of plant traits in response to light conditions. Oecologia 78:508—512. SCHWEITZER, J. A. AND K. C. LARSON. 1999. Greater morphological plasticity of exotic honeysuckle species may make them better invaders than native species. Journal of the Torrey Botanical Society 126:15—23. SMITH, M. D. AND A. K. KNAppP. 2001. Physiological and morphological traits of exotic, invasive exotic, and native plant species in tallgrass prairie. International Journal of Plant Science 162: 785-792. SWENSON, J. J. AND J. FRANKLIN. 2000. The effects of future urban development on habitat fragmenta- tion in the Santa Monica Mountains. Landscape Ecology 15:713—730. TUTIN, T. G., V. H. HEUWoob, N. A. BURGEs, D. M. Moore, D. H. VALENTINE, S. W. WATERS, and D. A. WEBB. (eds.). 1968. Flora Europaea vol. 2. Cambridge University Press, New York, NY. MADRONO, Vol. 55, No. 1, pp. 60—68, 2008 THE REDISCOVERY AND STATUS OF DISSANTHELIUM CALIFORNICUM (POACEAE) ON SANTA CATALINA ISLAND, CALIFORNIA JENNY L. MCCUNE AND DENISE A. KNAppP! Catalina Island Conservancy, Conservation Department, P.O. Box 2739, Avalon, CA 90704 ABSTRACT Dissanthelium californicum (Nutt.) Benth. (Poaceae) is an annual grass known only from Santa Catalina and San Clemente Islands, off the coast of Southern California, USA, and Guadalupe Island, off the coast of Baja California, Mexico. It had not been recorded since Blanche Trask collected it in 1903 on San Clemente Island, and was therefore considered to be extinct, possibly as a result of overgrazing by introduced goats on these islands. During monitoring by the Catalina Island Conservancy from March through July 2005, D. californicum was discovered growing in seven diverse and widely spread locations on Santa Catalina Island. The rediscovery may be a result of the removal of feral goats and pigs from the island, record-breaking rainfall, and increased exploration of remote areas. A revised description of the species and a summary of its habitat preferences are provided. Thorough monitoring and surveying of known and potential locations will be necessary to better determine the conservation status of the species. RESUMEN Dissanthelium californicum (Nutt.) Benth. (Poaceae) es una graminea anual conocida solo de las islas Santa Catalina y San Clemente, que se encuentran cerca de la costa del Sur de California, USA y de la isla Guadalupe, localizada cerca de la costa de Baja California, México. Esta especie no ha sido reportada desde que Blanche Trask la colectd en 1903 en la isla San Clemente, y por lo tanto ha sido considerada como extinta, posiblemente como resultado del sobre pastoreo de cabras introducidas a las islas. Durante el monitoreo realizado por la organizacion Catalina Island Conservancy en los meses de marzo a julio del 2005, D. californicum fue redescubierto creciendo en siete diversas y muy distantes localidades de la isla Santa Catalina. E] redescubrimiento podria haberse dado debido a la eliminacion de las cabras y cerdos de la isla, a la Iluvia sin precedentes y el aumento de exploracion en areas remotas. Una descripcion revisada de la especie y un resumen de su preferencia en habitat son proporcionadas. Riguroso monitoreo e inspecciOn de localidades conocidas y potenciales seran necesarios para una mejor determinacion del estado de conservacion de la especie. Key Words: California, Dissanthelium, extinct, Guadalupe Island, Poaceae, rediscovery, San Clemente Island, Santa Catalina Island. Dissanthelium californicum, an annual grass first collected and described in the mid-1800s, is known from three islands off the coast of southern California and Baja California, Mexico. Last collected in 1903 and subject to severe grazing pressures on all three islands, it had been presumed extinct until its re-discovery in 2005 on Santa Catalina Island (hereafter Catalina Island). In this paper, the history of this species is discussed and new population information is presented, including a revised description. Cur- rent conservation actions and concerns are discussed. TAXONOMIC AND COLLECTION HISTORY Thomas Nuttall, in his presentation to the Academy of Natural Sciences in 1848, described over one hundred plant species collected in the American West (Jercinovic 2004). Among the ' Author for correspondence, e-mail: dknapp@lifesci. ucsb.edu species was an annual grass that had not yet been described, collected by William Gambel on Catalina Island, off the southern coast of | California. Nuttall placed the annual grass in a | new genus named Stenochloa, and gave it the | specific epithet californica (Nuttall 1848). Ben- | tham later transferred it to the genus Dissanthe- lium, giving it the specific epithet californicum | (Hooker 1881). The genus Dissanthelium was described by C.B. | Trinius in 1836, and comprises approximately 20 | species (Nicora 1973; Swallen and Tovar 1965; Tovar 1985; Soreng 1998; Soreng et al. 2003). | Most species grow at high altitudes, usually | ~4000—5000 m, in the central Andes; one of | these species also grows on the high volcanoes of » central Mexico (Swallen and Tovar 1965; Tovar 1985). Dissanthelium californicum is the only species in the genus found north of the equator — at elevations near sea level. A recent phyloge- netics study by Refulio-Rodriguez has deter- mined that Dissanthelium forms a clade nested | within the genus Poa (Refulio-Rodriguez 2007). 2008] The placement of D. californicum in the genus was questioned by Hitchcock (1923), who wrote that ‘“‘this species does not appear to be closely related to the other two [(Andean) species, the only ones recognized at the time], but it does not seem to be sufficiently different to constitute a distinct genus.”’ Oscar Tovar, an authority on the genus, has also questioned the position of D. californicum in the genus (personal communica- tion). Unfortunately, low DNA sequence varia- tion within the Dissanthelium clade prevents a full analysis of the relationships among its species (Refulio-Rodriguez 2007). Dissanthelium californicum was collected only twice more following Gambel’s collection; in 1875 by Edward Palmer on Guadalupe Island, 261 km off the coast of Baja California, Mexico (Moran 1996), and in 1903 by Blanche Trask on San Clemente Island, about 34 km south of Catalina Island (Jepson 1912; Raven 1963). There has been some uncertainty about the dates of the collections made by Gambel and Trask. The isotype of D. californicum and other specimens collected by Gambel deposited in the Gray (GH23589, isotype) and Kew herbaria (Phillips personal communication) lack a date of collection. According to Miullspaugh and Nuttall (1923), Gambel visited Catalina Island in February 1847. Elsewhere, however, Gambel is reported to have arrived in California in Novem- ber 1841 (McKelvey 1955; Graustein 1967), and visited Catalina Island in 1842 (McKelvey 1955). He remained on the West Coast until 1845, when he returned to Philadelphia in order to enter medical school (Jercinovic 2004). He had sent his plant specimens by whaling ship, which did not arrive until nearly a year after he had returned. From Philadelphia, he sent the collection to Nuttall in England for identification, which explains the delay between Gambel’s collection and Nuttall’s paper (Graustein 1967; Jercinovic 2004). Thus D. californicum was first collected between 1841 and 1845. In Smith’s treatment of D. californicum in The Jepson Manual: Higher Plants of California, the last known collection of the species, taken by Trask, is said to be 1912 (Smith 1993). However, Raven (1963) does not mention any trips by Trask to San Clemente after 1903. We examined one of Trask’s specimens (Trask 324, PH 469621), and the year 1903 is penciled in on the label, along with the word ‘“‘common.” We could find no evidence that Trask or anyone else collected D. californicum after 1903, and conclude that the 1912 sighting reported by Smith (1993) is an error. Dissanthelium californicum is known from Catalina and San Clemente Islands off the coast of California, USA, and Guadalupe Island off the coast of Baja California, Mexico (Millspaugh and Nuttall 1923; Raven 1963; Thorne 1967, MCCUNE AND KNAPP: DISSANTHELIUM REDISCOVERY ON CATALINA ISLAND 61 1969; Moran 1996; Ross et al. 1997). Both Eastwood (1941) and Dunkle (1950) list D. californicum as occurring on Santa Cruz and Santa Rosa Islands in addition to the three southern islands, while Dunkle (1950) adds a northern mainland distribution. However, exten- sive searches for herbarium specimens reveal that these reports are unsubstantiated (Wallace 1985; S. Junak, Santa Barbara Botanic Garden, per- sonal communication). Despite the fact that Catalina Island, San Clemente and Guadalupe Islands have been relatively well surveyed by botanists (e.g., East- wood 1941; Millspaugh and Nuttall 1923; Moran 1996; Raven 1963; Thorne 1967, 1969), D. californicum has not been recorded since Trask’s collection, and therefore, has been presumed extinct (e.g., CNPS 2001; Smith 1993). It has been suggested that intense herbivory by feral goats (Capra hircus) was the cause of D. californicum’s disappearance and the decline of other native plants growing on these islands (Thorne 1967; Moran 1996). Palmer, following work on Guadalupe, wrote that it was very succulent and that goats were very fond of it (Watson 1876). According to Moran (1996), “‘the plants were succumbing fast to the cresting goat population, and he [Palmer] was just in time to find eight natives never found on the island again: Ceanothus cuneatus, Dissanthelium californicum, Hesperelaea palmeri, Micropus californicus, Planta sp., Pogogyne tenuiflora, Ribes sangui- neum, and Silene antirrhina.” REDISCOVERY In the following description of our rediscovery of D. californicum, the term population refers to a distinct cluster of individual plants. An occur- rence, aS per the California Natural Diversity Database (Bittman 2001), consists of all popula- tions found within 0.4 km of each other. Biologists from the Catalina Island Conser- vancy (i.e., “‘the Conservancy’’) have been studying and monitoring plant communities on the island in an effort to learn how they are recovering from many years of overgrazing and how they may be further protected and restored. On March 29, 2005, during yearly monitoring of recently burned Island chaparral/coastal sage scrub habitat near Catalina Island’s airport (the ‘‘Airport Burn’’), a single individual of D. californicum was found growing within a large patch of Malacothamnus fasciculatus (Torrey & A. Gray) E. Greene (Occurrence 1, Figs. 1 and 2). In the spring of 2005, we also began a project characterizing the Quercus pacifica Nixon & C.H.Mull. communities on the island, which brought us to many seldom-visited areas of the island. During these surveys in April, we discov- ered three additional populations of D. californi- [Vol. 55 62 MADRONO | | a] In mid-July, we found three additional occur- | owe] rences within Quercus pacifica plots. The first | ie ee ae consisted of four populations in the eastern | en 2 oa section of the island near the town of Avalon, | am totaling 75 individuals (Occurrence 5, fig. 2). The | Biome | second consisted of one population of four | —_— individuals, and the third consisted of one lone | —) individual (Occurrences 6 & 7, Fig. 2); both | ans ee] occurrences are near the isthmus in the north- | ass — western section of the island. These finds brought ] the total to approximately 678 individuals in 17 el populations and seven occurrences (Fig. 2). PR oa, | Dissanthelium californicum occured on a variety amd : | Siete of aspects, elevations, and slopes and had a range | foe) ——] of associated species (Table 2). The aspect of D. | —_.__} californicum locations ranged from northerly — — through easterly to south-southeasterly slopes, — ee a and in elevation ranged from approximately | ae 137 m to 290 m. It was found predominantly in | Fic. 1. Detail of the inflorescence fragment collected Island chaparral habitat, but also in areas where | this habitat intergrades with coastal sage scrub. on first noticing D. californicum in Occurrence |. Photo by Jenny L. McCune. REVISED DESCRIPTION cum: one of approximately 60 plants on the eastern side of the island (Occurrence 2, Fig. 2), and two separated by about 80m in the southwestern portion of the island, numbering 100 and 25 individuals (Occurrence 3, Fig. 2). Upon hearing the news of our discovery, Nancy Refulio-Rodriguez, J. Travis Columbus, and Susan Jett (Rancho Santa Ana Botanic Garden, hereafter RSA), and Steve Junak (Santa Barbara Botanic Garden, hereafter SBBG) conducted a collecting trip to the island from May 2nd to 4th, 2005, during which additional plants, populations, and occurrences were discovered. At Occurrence I, we counted approximately 140 plants in the vicinity of the first discovery (Fig. 2), within three populations. Columbus and colleagues also dis- covered a population about 500 m away (Gust outside of the burn area), consisting of approxi- mately 70 individuals (Occurrence 4, Fig. 2). Additionally, we discovered 110 individuals in Dissanthelium californicum shows some plastic- _ ity in its characteristics. The plant height ranges _ from 10 cm or less to a maximum of 60 cm. In shade, it is usually a bright green, but in the open sun it tends to have a reddish tinge in the leaves, _ stems, and inflorescences. Each individual plant — has from one to several flowering stalks. In general, the stalks were erect, but some, especially in shade, tended to be decumbent. | Dissanthelium californicum was flowering at the | end of March when we first observed it, and by the beginning of May some seed was already © mature. By mid- to late- May, most florets had | fallen, and the plants discovered after May had — turned tan in color with only empty glumes | remaining. The new collections of this plant — (along with inspection of previously existing © collections) have given us an opportunity to produce a revised description, below. Dissanthelium californicum (Nutt.) Benth. two new populations at Occurrence 2 and two more small populations at Occurrence 3, totaling 23 plants. Small amounts of seed were collected by the Conservancy (Accession numbers 1337-1340) and RSA (Accession numbers 21890-21894) to use for germination testing, storage, and restoration. We collected voucher specimens from Occur- rences 1-3 (McCune 1, 2 and 3 respectively), CALIFORNIA DISSANTHELIUM. Annual Stem 6-46 cm. Leaf: blade flat, 6-20 cm, 1-4 mm wide, smooth, ligule membranous. Inflorescence < 1.6 dm; panicle narrow to open, branches in fascicles, some of them floriferous to the base. Spikelet florets 2(3), green to occ. reddish-purple, 2.5-5 mm, equal; glumes narrow, acute, glabrous or minutely scabrous on the keel toward the tip, duplicates of which are deposited at the Wrigley Botanical Garden on Catalina Island, RSA, and SBBG respectively. Refulio-Rodriguez collected a voucher from Occurrence 4 (Refulio 238) and deposited it at RSA. Searches of Guadalupe and San Clemente Islands in 2005 did not reveal any D. californicum populations (Refulio-Rodriguez per- sonal communication). A summary of all known herbarium specimens is presented in Table 1. nearly equal, lower glume 3-veined, upper glume l-veined; glumes > florets; lemma 1.5—2 mms hairy, obtuse to acute. Open to full shade of Island chaparral and coastal sage scrub; <500 m.. sChI (Santa Catalina, San Clemente Islands); Baja CA (Guadalupe Island). Refs. —- DISSANTHELIUM CALIFORNICUM (Nutt.) Benth. 2008] .MCCUNE AND KNAPP: DISSANTHELIUM REDISCOVERY ON CATALINA ISLAND Catalina Island UTM 374000 E Occurrence 5: 4 populations, 145 individuals atte % Occurrence /: 1 population, UTM 3699000 N Le Occurrence 1. 3 populations, Ev 140 individuals Occurrence 2: 3 populations, 170 individuals 4 individuals Pm Se Bo A ama 63 iFiG. 2. Map showing all seven occurrences of D. californicum found on Catalina Island between the end of March and July of 2005. Numbers of individuals are estimates. 64 TABLE Il. MADRONO [Vol. 55 FLORISTIC DOCUMENTATION OF D. CALIFORNICUM. The 1912 date listed in Hickman (1993) appears to be an error, and is revised here as 1903. Locality USA: CA, Santa Catalina Island MEXICO: Guadalupe Island. Guadalupe Guadalupe Guadalupe San Clemente Santa Catalina Santa Catalina Santa Catalina Santa Catalina TABLE 2. SITE CHARACTERISTICS AT D. CALIFORNICUM LOCATIONS ON SANTA CATALINA ISLAND. Elevations for Voucher W. Gambel s.n. (GH) . Palmer 96 (NY) . Palmer 96 (GH) . Palmer 96 (MO) . Trask 324 (US #469621) McCune 1-3 (RSA) N. Refulio-Rodriguez 238 (RSA) J. McCune 1-3 (SBBG) J. McCune 1-3 Wrigley Memorial E E E. Palmer 96 (CM #268852) Is B J. Date ca. 1841-1845 1875 1875 1875 1875 1903 May May May May Botanical Garden, Avalon, CA. Frequency and distribution Unknown On warm rocky slopes in the middle of the island; not very abundant; very succulent, and the goats are very fond of it (Watson 1876) “common” Infrequent. Plant found in seven locations on the island (17 populations, 678 individuals) 2005 2005 2005 2005 all occurrences are approximate. Slope for Occurrence 4 is estimated. Occ # 1 UTM 366902E 3695747N S7IZTSE 3695273N 363801E 3691475N 367374E 3695702N 362599E 3700121N 374542E 3691983N 362947E 3700121N ASP. Slope 87° a7 130° 27° 340° = 334° ei > 10°. 22° 145° 32° 0 26° Elev.(m) 260 137 182 274 213 Habitat Island chaparral Island chaparral/ coastal sage scrub Island chaparral/ coastal sage scrub Island chaparral/ coastal sage scrub Island chaparral Island chaparral Island chaparral Substrate and associated species Bare ground in part shade of Malacothamnus i fasciculatus, with Quercus pacifica, Bromus madritensis, Polypogon interruptus, Chenopodium californicum, Opuntia littoralis. Leaf litter or soil with large rock outcrops in full or part shade of Q. pacifica or Salvia mellifera, with Brachypodium distachyon, Bromus madritensis, Eucrypta chrysanthemifolia, Melica imperfecta, Opuntia littoralis. Soil or near large rock outcrops in the open or | in part to full shade of Heteromeles arbutifolia, with Salvia mellifera, Polypodium californicum, Trifolium willdenovii, Melica imperfecta, Bromus madritensis, Bromus hordeaceous, Daucus pusillus, Galium nuttallii, Rhus integrifolia, Avena barbata, Vulpia sp. Moist soil in full or part shade of Rhus integrifolia, with Salvia mellifera, Malacothamnus fasciculatus, Antirrhinum nutallianum, Phacelia cicutaria, Gnaphalium sp., Solanum sp., Chenopodium | californicum, Opuntia littoralis, Polypogon sp., Desmazeria rigida, Brachypodium distachyon, Piptatherum miliaceum. Leaf litter or soil with rock outcrops in full or | part shade of Q. pacifica, Rhus integrifolia, — or Crossossoma californica, with Bromus madritensis, Gastridium ventricosum, Melica imperfecta. Moist soil, quite bare, in full shade of Q. pacifica, Rhus integrifolia and Cercocarpus betuloides, with Melica imperfecta. Leaf litter in part shade of Q. pacifica, with Opuntia littoralis, Bromus madritensis, Brachypodium distachyon, Sonchus asper. 2008] Swallen & Tovar. Phytologia 11: 361—376. 1965. Benth. in Hook. Icon. Pl. HI. 4: 56. pl 1375. 1881. Based on the next. Stenochloa californica Nutt. Journ. Acad. Phila. II. 1:189. 1848. Type from Santa Catalina Island, Gambel. DISSCUSSION Three factors may have contributed to the rediscovery of D. californicum on Catalina Island after having been unrecorded for more than a century. First, it seems likely that D. californicum, like many other island plants, has responded to the record rainfall in 2004-2005, making it more abundant and conspicuous than in previous years. Second, the removal of feral goats and pigs has relieved grazing and rooting pressures. Finally, the initiation of our project to charac- terize the oak communities on Catalina Island brought us to many rarely explored areas. The island received 68.1 cm of rainfall during the 2004-2005 season (averaged across Avalon, Middle Ranch, Airport-in-the-Sky, and Two Harbors), which was significantly greater than the mean (Conservancy data: 29.8 cm, averaged across four to six sites from 1948-2005) and the highest rainfall recorded to date. This factor likely contributed to the germination, abundance, and robustness of D. californicum that year. The response of D. californicum populations to rainfall was further indicated by 2006 survey data: after below-average rainfall (24 cm from July 2005 through June 2006), surveys of four out of seven of the known population sites did not reveal any plants. It is unclear if D. californicum has seed that can persist in the soil for years. Generally, grass seed does not persist as long in the soil as seeds of forbs, but annuals are more likely than perennials to form a seed bank (Rice 1989 and references cited therein; Thompson et al. 1998). Although many annual grasses have very short-lived seeds (i.e., 1-3 yr) (Zorner et al. 1984; Roberts 1986; Thompson 1987; Russi et al. 1992; Masin et al. 2006), multiple species of Poa (both annual and perennial) have been found to have more persistent seeds (Roberts 1986; Thompson 1987). Results from germination trials by RSA suggest that D. californicum seeds do not possess physical dormancy mechanisms, as they obtained 35% germination in 2006 by pre-treating the seeds with only a 24-hour water soak (M. Wall personal communication). Introduced herbivores may have greatly re- duced the abundance and distribution of D. californicum. Introduced ungulates have particu- larly severe impacts on island ecosystems, where endemic plants lack defenses against herbivory, and are a major cause of island extinctions (e.g., Coblentz 1978; Vitousek 1988: Adsersen 1989: MCCUNE AND KNAPP: DISSANTHELIUM REDISCOVERY ON CATALINA ISLAND 65 Atkinson 1989; Schofield 1989; Bowen and Van Vuren 1997; Oberbauer 2005). Native plants and animals may recover rapidly with the release of this pressure (Wehtje 1994; Laughrin et al. 1994; Chess et al. 2000), depending on environmental factors such as water availability (Donlan et al. 2002). On Guadalupe Island, six plant species presumed extinct and other rare pine seedlings reappeared following goat removal (Krajick 2005). Feral goats were reported on San Clemente and Catalina Islands by the early 1800s (Dunkle 1950; Raven 1963), and thrived until systematic eradication programs began in the 1970s on San Clemente Island (Moran 1996) and in 1990 on Catalina Island (Schuyler et al. 2002). Currently, few or no goats remain on either island. It is unknown when goats were first introduced on Guadalupe, but the goat population peaked at more than 100,000 in 1870. By 1994, the population had been reduced to 7000 (Moran 1996), and today, nearly all goats have been removed (Island Conservation 2007). Other introduced animals on Catalina Island that have posed a threat are feral pigs (Sus scrofa), mule deer (Odocoileus hemionus), and American bison (Bison bison). Rooting distur- bance by feral pigs decreases plant productivity and may disturb over 65% of the land in areas with high pig densities (Sweitzer and Van Vuren 2002); they have contributed to the extinctions of numerous oceanic island species (Waithman et al. 1999). Although mule deer are primarily brows- ers, grasses comprise between 21% (December— January) and <1% (June—July) of their diet on Catalina Island, and their population on the 48,000-acre island is estimated at 2341 individuals (range: 1682-3259) (Manuwal and Sweitzer 2007; T. Manuwal, University of North Dakota, unpublished data). Introduced bison on Catalina Island alter plant communities by grazing, foraging, trampling, and wallowing, and have been found to facilitate the dispersal of non- native plants (Sweitzer et al. 2003; Constible et al. 2005). Trampling by bison and deer was found to negatively impact oak seedlings on the island (Manuwal and Sweitzer 2007). Their diet is composed of approximately 84% grasses; they appear to have the positive effect of reducing non-native annual grass cover, although their input of nutrients may also encourage annual grass growth (Sweitzer et al. 2003). Management of these introduced herbivores is being addressed by the Conservancy. The major- ity of the feral pigs were eradicated by the end of 2004; there are currently 1—4 animals estimated remaining, and the Conservancy continues its efforts. Bison herds are managed to between 150 and 200 individuals. Mule deer have been harvested almost annually on Santa Catalina Island from 1949 through the present, currently 66 MADRONO as part of the Private Lands Management program of the California Department of Fish and Game. Final results from a two-year mule deer study will be submitted by the end of 2007, and management options are under discussion. With the removal of feral goats from all three islands where D. californicum has been known to occur, the remaining threat of largest concern 1s competition with invasive annual and perennial grasses. These introduced species, including Bromus spp., Avena spp., Lolium spp., and Vulpia spp. have been shown to interfere with native shrubs, forbs, and perennial grasses (Young and Evans 1973; Da Silva and Bartolome 1984; Gordon et al. 1989; Danielsen and Halvorson 1991; Ehason and Allen 1997; Dyer and Rice 1999; Hamilton et al. 1999; Brown and Rice 2000; Kolb et al. 2002). Invasive perennial grasses occurring on Catalina such as Piptatherum miliaceum (L.) Cosson, Festuca arundinacea Shreber and Phalaris aquatica L., and other non-grass invasive species are also of concern (Corbin et al. 2004; Bossard et al. 2006). Introduced grasses are listed as associated species at five of the seven occurrences of D. californi- cum. Although invasive annual grasses are too ubiquitous on the island to control except in very targeted areas, the Conservancy has initiated an ambitious Invasive Plant Management Program, targeting 25 species for eradication and five for reduction based on extensive mapping data and ecological impact considerations, and targeting watersheds containing high numbers of rare species and habitats (Knapp and Knapp 2005). To better understand the status of the species in the wild, we need to know more about its range, frequency and annual variability on the island, its growth requirements, and any addi- tional threats. Studies of seed germination and plant growth requirements are particularly im- portant. The Catalina Island Conservancy con- tinues to monitor known populations of D. californicum and survey additional high potential locations; it is our hope that it will regain the ‘common’ status attributed to it by Trask more than a century ago. ACKNOWLEDGMENTS The authors would like to thank Dr. J. Travis Columbus and Nancy Refulio-Rodriguez for their enthusiasm and assistance in researching and reviewing this manuscript, Steve Junak for his expert knowledge and advice, and Lauren K. Danner for her assistance in the field. We also thank Ms. Refulio-Rodriguez for her translation of the abstract to Spanish, as well as John Knapp, John Hunter, and two anonymous reviewers for their helpful suggestions for improving the article. LITERATURE CITED ADSERSEN, H. 1989. The rare plants of the Galapagos Islands and their conservation. Biological Conser- vation 47:49—77. [Vol. 55 ATKINSON, I. 1989. Introduced animals and extinctions. Pp. 54-75 in D. Western and M. C. Pearl (eds.), Conservation for the 21st century. Oxford Univer- sity Press, New York, NY. BITTMAN, R. 2001. The California natural diversity database: a natural heritage program for rare species and vegetation. Fremontia 29:57—62. BOSSARD, C. C., M. Brooks, J. DiI TOMASo, J. RANDALL, C. ROYE, J. SIGG, A. STANTON, AND P. WARNER. 2006. California invasive plant inventory. California Invasive Plant Council, Berkeley, CA. BOWEN, L. AND D. VAN VUREN. 1997. Insular endemic plants lack defenses against herbivores. Conserva- tion Biology 11:1249—-1254. Brown, C. S. AND K. J. RICE. 2000. The mark of Zorro: effects of the exotic annual grass Vulpia myuros on California native perennial grasses. Restoration Ecology 8:10—17. CNPS (CALIFORNIA NATIVE PLANT SOCIETY). 2001. Inventory of rare and endangered plants of California (sixth edition). Rare Plant Scientific Advisory Committee, D. P. Tabor, Convening Editor. California Native Plant Society, Sacra- mento, CA. CHESS, K., K. Mc EACHERN, AND D. H. WILKEN. 2000. Island jepsonia (Jepsonia malvifolia) demog- raphy on Santa Rosa and Santa Cruz Islands, California. Pp. 155-157 in D. R. Browne, K. L. Mitchell, and H. W. Chaney (eds.), Proceedings of the Fifth California Islands Symposium. U.S. Department of the Interior, Minerals Management Service, Pacific OCS Region, Camarillo, CA. COBLENTZ, B. E. 1978. The effects of feral goats (Capra hircus) on island ecosystems. Biological Conserva- tion 13:279—286. CONSTIBLE, J. M., R. A. SWEITZER, D. H. VAN VUREN, P. T. SCHUYLER, AND D. A. KNAPP. 2005. Dispersal of non-native plants by introduced bison in an island ecosystem. Biological Invasions 7: 699-709. CORBIN, J. D., M. THOMSEN, J. ALEXANDER, AND C. M. D’ANTONIO. 2004. Out of the frying pan: invasion of exotic perennial grasses in coastal prairies (Abstract). Pp. 27-28 in C. Pirosko (ed.), Proceedings of the California Invasive Plant Council Symposium, Vol. 8. California Invasive Plant Council, Berkeley, CA. DANIELSEN, K. C. AND W. L. HALVORSON. 1991. Valley oak seedling growth associated with selected grass species. Pp. 9-13 in Proceedings of the symposium on oak woodlands and hardwood rangeland management. General Technical Report PSW-12, U.S. Forest Service, Albany, CA. DA SILVA, P. G. AND J. W. BARTOLOME. 1984. Interaction between a shrub, Baccharis pilularis subsp. consanguinea (Asteraceae) and an annual grass, Bromus mollis (Poaceae), in coastal Califor- nia. Madrono 31:93—101. DONLAN, C. J., B. R. TERSHY, AND D. A. CROLL. 2002. Islands and introduced herbivores: Conservation action as ecosystem experimentation. Journal of | Applied Ecology 39:235—246. DUNKLE, M. B. 1950. Plant ecology of the Channel Islands of California. University of Southern California Press, Los Angeles, CA. | DYER, A. R: AND K.. Je RICE. 1999: Effects sof competition on resource availability and growth | of a California bunchgrass. Ecology 80:2697—2710. 2008] EASTWOOD, A. 1941. The islands of southern California and a list of the recorded plants. Leaflets of Western Botany 3:54-78. ELIASON, S. A. AND E. B. ALLEN. 1997. Exotic grass competition in suppressing native shrubland re- establishment. Restoration Ecology 5:245—255. GorDON, D., J. M. WELKER, J. W. MENKE, AND K. J. RIcE. 1989. Competition for soil water between annual plants and blue oak (Quercus douglasii) seedlings. Oecologia 79:533-41. GRAUSTEIN, J. E. 1967. Thomas Nuttall, Naturalist: Explorations in America 1808—1841. Harvard University Press, Cambridge, MA. HAMILTON, J. G., C. HOLZAPFEL, AND B. E. MAHALL. 1999. Coexistence and interference between a native perennial grass and non-native annual grasses in California. Oecologia 121:518—526. Hitrcucock, A. S. 1923. Dissanthelium, an American genus of grasses. Journal of the Washington Academy of Sciences 13:223—225. Hooker, J. D. (ed.). 1881. Hooker’s [cones Plantarum: or, figures, with descriptive characters and re- marks, of new and rare plants, selected from the Kew herbarium. Third Series. Vol. XIV, Part III. Williams and Norgate, London. ISLAND CONSERVATION. 2007. Guadalupe Island restoration project. Available at: http://www. islandconservation.org/guadalupe.html JEPSON, W. L. 1912. A flora of California. Volume 1, Parts 1-7. University of California, Berkeley, CA. JERCINOVIC, G. 2004. William Gambel: New Mexico plant specimens. The New Mexico Botanist 30:2—4. KNAPP, D. A. AND J. J. KNAPP. 2005. Ecosystem protection through watershed-level prioritization on Catalina Island. Pp. 39-46 in G. Skurka (ed.), Proceedings of the California Invasive Plant Council Symposium, Vol. 9. California Invasive Plant Council, Berkeley, CA. KOLB, A. P., A. D. ENTERS, AND C. HOLZAPFEL. 2002. Patterns of invasion within a grassland community. Journal of Ecology 90:871—881. KRAJICK, K. 2005. Winning the war against island invaders. Science 310:1410—-1413. LAUGHRIN, L., M. CARROLL, A. BROMFIELD, AND J. CARROLL. 1994. Trends in vegetation changes with removal of feral animal grazing pressures on Santa Catalina Island. Pp. 523-530 in W. L. Halvorson and G. J. Maender (eds.), The fourth California Islands symposium: Update on the status of resources. Santa Barbara Museum of Natural History, Santa Barbara, CA. MANUWAL, T. AND R. SWEITZER. 2007. Advance report, Catalina Island mule deer project: Implica- tions of introduced large ungulates on island scrub oak habitats on Santa Catalina Island, California. Unpublished preliminary report prepared for Catalina Island Conservancy, Avalon, CA. MASIN, R., M. C. ZUIN, S. OTTO, AND G. ZANIN. 2006. Seed longevity and dormancy of four summer annual grass weeds in turf. Weed Research 46:362-370. MCKELVEY, S. D. 1955. Botanical exploration of the trans-Mississipp1 West 1790-1850. Arnold Arbore- tum of Harvard University, Jamaica Plain, MA. MILLSPAUGH, C. F. AND L. W. NUTTALL. 1923. Flora of Santa Catalina Island (California). Field Muse- um of Natural History, Chicago, IL. MCCUNE AND KNAPP: DISSANTHELIUM REDISCOVERY ON CATALINA ISLAND 67 MORAN, R. 1996. The flora of Guadalupe Island, Mexico. California Academy of Sciences, San Francisco, CA. NIcORA, E. G. 1973. Novedades agrostdlogicas pata- gonicas. Darwiniana 18(1—2):80—106. NUTTALL, T. 1848. Descriptions of plants collected by William Gambel, M.D., in the Rocky Mountains and Upper California. Journal of the Academy of Natural Sciences of Philadelphia vol. 1 second series:149—189. OBERBAUER, T. 2005. A comparison of estimated historic and current vegetation community struc- ture on Guadalupe Island, Mexico. Pp. 143-151 in D. K. Garcelon and C. A. Schwemm (eds.), Proceedings of the Sixth California Islands Sympo- sium. National Park Service Technical Publication CHIS-05-01, Institute for Wildlife Studies, Arcata. RAVEN, P. H. 1963. A flora of San Clemente Island, California. Aliso 5(3):289-347. REFULIO-RODRIGUEZ, N. F. 2007. Systematics of Dissanthelium Trin. (Poaceae: Pooideae). Ph.D. dissertation. Claremont Graduate University, Clar- emont, CA. RIcE, K. 1989. Impacts of seed banks on grassland community structure and population dynamics. Pp. 211—230 in M. A. Leck, V. T. Parker, and R. L. Simpson (eds.), Ecology of soil seed banks. Academic Press, San Diego, CA. ROBERTS, H. A. 1986. Persistence of seeds of some grass species in cultivated soil. Grass and Forage Science 41(3):273—276. Ross, T. S., S. BoyD, AND S. JUNAK. 1997. Additions to the vascular flora of San Clemente Island, Los Angeles County, California, with notes on clarifi- cations and deletions. Aliso 15(1):27—40. Russ, L.. P. 8S. Cocks, AND BE. H. ROBERTS, 1992. Seed bank dynamics in a Mediterranean grassland. Journal of Applied Ecology 29:763—771. SCHOFIELD, E. K. 1989. Effects of introduced plants and animals on island vegetation: Examples from the Galapagos Archipelago. Conservation Biology 3:227-238. SCHUYLER, P. T., D. GARCELON, AND S. ESCOVER. 2002. Control of feral goats (Capra hircus) on Santa Catalina Island, California, USA. p. 412 in C. R. Veitch and M. N. Clout (eds.), Turning the tide: The eradication of invasive species. Proceed- ings of the international conference on eradication of island invasives, February 2001. Occasional Paper of the IUCN Species Survival Commission No. 27. IUCN, Gland, Switzerland and Cam- bridge, UK. SMITH, J.P: 1993. Poaceae. Pp. 1218-1303 mJ. C. Hickman (ed.), The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. SORENG, R. J. 1998. An infrageneric classification for Poa in North America, and other notes on sections, species, and subspecies of Poa, Puccinellia, and Dissanthelium (Poaceae). Novon 8:187—202. , P. M. PETERSON, G. DAVIDSE, E. J. JUDZIE- WICZ, F. O. ZULOAGA, T. S. FILGUEIRAS, AND O. MORRONE. 2003. Catalogue of New World grasses (Poaceae): IV. Subfamily Pooideae. Contr. U.S. Natl. Herb. 48:1—730, Available at: http://mobot. mobot.org/W3T/Search/nwgc.html SWALLEN, J. R. AND O. TOVAR. 1965. The grass genus Dissanthelium. Phytologia 11:361—376. 68 MADRONO SWEITZER, R. A. AND D. H. VAN VUREN. 2002. Rooting and foraging effects of wild pigs on tree regeneration and acorn survival in California’s oak woodland ecosystems. Pp. 219-231 in R. B. Stan- diford, D. McCreary, and K. L. Purcell (tech. cords.) (eds.), Proceedings of the fifth symposium on oak woodlands: Oaks in California’s changing landscape. General Technical Report PSW-GTR- 184. Pacific Southwest Research Station, U.S. Forest Service, Albany, CA. —_—, J. CONSTIBLE, AND D. H. VAN VUREN. 2003. Population ecology and ecological effects of bison on Santa Catalina Island, California. Unpublished report prepared for Santa Catalina Island Conser- vancy, Avalon, CA. THOMPSON, K. 1987. Seeds and seed banks. Phytologist 106:23—34. ——, J. P. BAKKER, R. M. BEKKER, AND J. G. HODGSON. 1998. Ecological correlates of seed persistence in soil in the north-west European flora. Journal of Ecology 86:163—169. THORNE, R. F. 1967. A flora of Santa Catalina Island, California. Aliso 6:1—77. 1969. A supplement to the floras of Santa Catalina and San Clemente Islands, Los Angeles County, California. Aliso 7:73-83. TOVAR, O. 1985. Ocho especies nuevas de Gramineas del Peru. Publicaciones del Museo de Historia Natural “Javier Prado’’. Serie B. Botanica 33:1—16. VITOUSEK, P. M. 1988. Diversity and_ biological invasions of oceanic islands. Pp. 181-189 in E. O. New [Vol. 55 Wilson (ed.), Biodiversity. World Resources Insti- tute, Washington, D.C. WAITHMAN, J. D., R. A. SWEITZER, D. VAN VUREN, J.D. DREw, A. J. BRINKHAUS, I. A. GARDNER, AND W.M. Boyce. 1999. Range expansion, population sizes, and management of wild pigs in California. Journal of Wildlife Management 63:298—308. WALLACE, G. W. 1985. Vascular plants of the Channel Islands of southern California and Guadalupe Island, Baja California, Mexico. Contributions in Science, Number 365. Natural History Museum of Los Angeles County, Los Angeles, CA. WATSON, S. 1876. Botanical Contributions. Proceed- ings of the American Academy of Arts and Sciences 11:105—148. WEHTJE, W. 1994. Response of a Bishop pine (Pinus muricata) population to removal of feral sheep on Santa Cruz Island, California. Pp. 331—340 in W. L. Halvorson and G. J. Maender (eds.), The fourth California Islands symposium: Update on the status of resources. Santa Barbara Museum of Natural History, Santa Barbara, CA. YOUNG, J. A. AND R. A. EVANS. 1973. Downy brome — intruder in the plant succession of big sagebrush communities in the Great Basin. Journal of Range Management 26(6):410-415. ZORNER, P. S., R. L. ZIMDAHL, AND E. E. SCHWEIZER. 1984. Sources of viable seed loss in buried dormant and non-dormant populations of wild oat (Avena fatua L.) seed in Colorado. Weed Research 24(2):143—150. MADRONO, Vol. 55, No. 1, pp. 69—75, 2008 RESURRECTION OF ASCLEPIAS SCHAFFNERI (APOCYNACEAE, ASCLEPIADOIDEAE), A RARE, MEXICAN MILKWEED MARK FISHBEIN Portland State University, Department of Biology, P.O. Box 751, Portland, OR 97207 mfish@pdx.edu VERONICA JUAREZ-JAIMES AND LEONARDO O. ALVARADO-CARDENAS Herbario Nacional, Instituto de Biologia, Universidad Nacional Aut6noma de Mexico, A.P. 70-367, Del. Coyoacan, 04510, Mexico, D. F. ABSTRACT Relocation and further study of the local endemic Asclepias zacatecana McVaugh showed that it is conspecific with 4. schaffneri A. Gray, a distinct species that previously had been synonymized with A. quinquedentata A. Gray. However, we consider a recently described species, A. rzedowskii W.D. Stevens, to be conspecific with A. quinquedentata. Asclepias schaffneri is lectotypified. RESUMEN El redescubrimiento y estudios consecuentes de Asclepias zacatecana McVaugh, una especie endemica de Zacatecas de distribucion restringida, muestra que es sinonimo de A. schaffneri A. Gray, la cual a su vez ha sido incluida como parte de A. guinquedentata A. Gray. Se discute porque A. shaffneri es diferente de A. guinquedentata y es el nombre correcto para reconocer A. zacatecana. Sin embargo, se considera que otra especie recientemente descrita, A. rzedowskii W. D. Stevens es a su vez sinonimo de A. guinquedentata. Se lectotipifica A. schaffneri. Key Words: Apocynaceae, Asclepiadoideae, Asclepias, lectotype, Mexico, milkweed, taxonomy. Asclepias zacatecana McVaugh (Apocynaceae) was described from a single collection made by Salvador Correa in 1975 (Correa 25 [MICH]) in the Sierra de Morones of Zacatecas, Mexico (McVaugh 1978). This species has been thought to be a rare, local endemic; prior to the fieldwork reported here, the only other collection known to the authors was made in the same general area 1n 1971 (Diaz 2356 [ENCB]). In the summer of 2006, we located a population attributable to A. zacatecana in this area (Fishbein et al. 5846 [HPSU, MEXU, MO], Judrez-Jaimes et al. 813 [MEXU]), and also discovered populations of what is clearly the same species north of Juventino Rosas, Guanajuato (Judrez-Jaimes 519 [MEXU], Fishbein et al 5856 [HPSU]). This apparent range extension of 300 km to the southeast prompted further investigation into the distribution and affinities of A. zacatecana, the results of which are presented here. Asclepias zacatecana is most similar to other diminutive, narrow leaved, and small-flowered milkweeds of Mexican highlands, such as 4A. fournieri Woodson, A. quinquedentata A. Gray, A. rosea Kunth, and A. scaposa Vail, but is readily distinguished by floral and vegetative characteristics. However, specimens identical in every respect to A. zacatecana were described by Gray (1880) as A. schaffneri A. Gray (Schaffner 56 [syntype GH, isosyntype MO] and Parry & Palmer 582 [syntype GH]), but this name was later treated as a synonym of A. guinquedentata by Woodson (1954) in his monograph of North American Asc/epias. It is worth noting that Parry, who may have been the collector of flowering specimens of milkweeds attributed to “Parry & Palmer” from the vicinity of San Luis Potosi prior to Palmer’s arrival in late July 1878 (McVaugh 1956), recognized two distinct species in these collections when he assigned different catalogue numbers to specimens referable to A. schaffneri and A. quinquedentata (S582 and 583, respectively). Parry assembled and numbered these collections at his home in Davenport, Iowa in early 1879 (McVaugh 1956), suggesting that his separate numbers may have been based as much on a perceived specific difference as on distinct gatherings in the field. Gray (1880) also explicitly distinguished Parry & Palmer 582, one of his syntypes of A. schaffneri, from Parry & Palmer 583, which he determined to be A. quinquedentata. In the protologue of A. schaff- neri, Gray compared his new species not to his A. quinquedentata, but to A. coulteri A. Gray, a plant of larger stature and flowers that 1s endemic to calcareous substrates in the Sierra Madre Oriental (Gray 1880). Gray’s recognition of A. schaffneri and A. quinquedentata as distinct species was followed by Hemsley (1881), who observed duplicates of the Parry & Palmer and Schaffner collections at Kew. Vail (1898) added another name based on 70 MADRONO these collections by describing A. palmeri Vail with Parry & Palmer 583 as type. She explicitly compared her new species to both A. guinque- dentata and A. schaffneri, claiming that A. palmeri was clearly distinct and intermediate between the other two species, but in an unspecified manner. However, her description of A. palmeri is wholly commensurate with the range of variation found in A. quinquedentata. Woodson (1954) followed Gray (1880) in consid- ering Parry and Palmer 583 to belong to A. quinquedentata, a determination with which we are in agreement. Although not referring specif- ically to the syntypes of A. schaffneri, Woodson (1954:90) noted that specimens collected in San Luis Potosi (the state in which Gray’s syntypes were collected) were atypical of A. quinquedentata and perhaps the result of hybridization with A. coulteri. Woodson’s reasoning is not explicit, though his inference may have been based on the corona of A. schaffneri (Fig. 1A), which is more similar in form (but not size) to that of A. coulteri (Fig. 1C) than that of A. guinquedentata (Fig. 1D). However, we have found that the three species are easily distinguished by corona morphology alone, in addition to other floral and vegetative characters (Table 1). Other than some similarity in corona form, there is little reason to attribute the distinctive traits of A. schaffneri to hybridization between A. gquinquedentata and A. coulteri, species that are almost completely allopatric and with very different habitat preferences (Fig. 2). Asclepias quinquedentata is found in grass-dominated openings in oak woods on a variety of substrates, but typically not on limestone, whereas A. coulteri is found almost exclusively on limestone outcrops, usually in submontane scrub. A. schaffneri is known from an apparent gap in the distribution of A. guinquedentata between Dur- ango and San Luis Potosi, outside the range of A. coulteri (Fig. 2). Other than Parry and Palmer’s rather vague collecting localities in San Luis Potosi, there is no indication that any pair of these species co-occurs. It should be noted that Woodson (1954) made other unwarranted infer- ences of hybridization involving poorly known Mexican Asclepias (cf. Fishbein and Lynch 1999). Most recently, Asclepias rzedowskii W.D. Stevens, a species described from Estado de Mexico (Stevens 1983), was distinguished from A. quinquedentata by an entire, rather than toothed, upper hood (corona) margin. We have examined specimens of A. rzedowskii from the general region of the holotype (the holotype is missing; see below) and conclude that the distinguishing character 1s, at best, inconstant. We provisionally reduce A. rzedowskii to the synonymy of A. guinguedentata. Our study of the types of A. quinquedentata, A. schaffneri, and A. zacatecana, in addition to a [Vol. 55 large proportion of the few collections of these entities represented in herbaria, supports the recognition of both A. quinquedentata and A. schaffneri, the conspecificity of A. zacatecana and A. schaffneri, and the conspecificity of A. rzedowskii and A. quinquedentata. Characters useful in distinguishing A. schaffneri, A. quinque- dentata, and A. coulteri are presented in Table 1 and known localities of these species are shown in Big.22: TAXONOMY ASCLEPIAS SCHAFFNERI A. Gray, Proc. Amer. Acad. Arts 16:103—104. 1880.—TY PES: MEX- ICO, San Luis Potosi, mountains near Mo- rales, July 1876, .Schajffner’ S6oCectotype, designated here; GH!, “isolectotypes, K!, MEXU!, MO!), MEXICO; San Luis Potosi, chiefly in the region of San Luis Potosi, 22° N Lat., 6000—8000 ft., 1878, Parry & Palmer 582 (syntype, GH!, isosyntypes, K!, MO, 2 sheets!). Fig. 3A. Asclepias zacatecana McVaugh, Contrib. Univ. Mich. Herb. 11:289. 1978. Syn. nov.—TYPE: MEXICO, Zacatecas, near summits between Jalpa and Tlaltenango, 2400-2500 m, 22 July 1975, Corea 25 (holotype, MICH!). Additional specimens seen. MEXICO, GUA- NAJUATO, Mpio. de Juventino Rosas, Juventino Rosas-Guanajuato road, 23 km north of Juven- tino Rosas, 2220 m, 6 July 2006, Judrez-Jaimes et al. 819 (MEXU); Mpio. de Juventino Rosas, 25 km north of Juventino Rosas, 2290 m, 6 July 2006, Fishbein et al. S856 (HPSU, MEXU, MO). ZACATECAS, Jalpa-Tlaltenango road, 2100 m, 2 July 1971, Diaz 2356 (ENCB); Mpio. de Tlalte- nango, Sierra de Morones, Tlaltenango-Jalpa road, 27 km southeast of Tlaltenango, 2525 m, 5 July 2006, Judrez-Jaimes et al. 813 (MEXU), Fishbein et al. 5846 (HPSU, MEXU); Mpio. de Chalchihuite, southeast end of Sierra Prieta, 3 km north of La Colorada mining area, 30 km south of Sombrerete along gravel roads, 16 km (by air) southeast of Chalchihuite, above 2900 m, 25 July 1982, Diggs and Nee 3008 (F). Asclepias schaffneri is endemic to mountains of the central plateau of Mexico, in the states of Guanajuato, San Luis Potosi, and Zacatecas (Fig. 2). The species is documented from _ pine- oak forest from 2100 to 3000 m. It is apparently quite rare, occurring in small, widely dispersed populations. The plants can easily be overlooked because of their small stature and narrow leaves. Asclepias schaffneri should be sought in_ the adjacent states of Aguascalientes, Durango, Jalisco, and Querétaro where similar habitat occurs. Gray (1880) cited two syntypes in describing A. schaffneri. We have selected the first cited collection, Schaffner 56, as a lectotype, because i t 2008] FISHBEIN ET AL.: RESURRECTION OF ASCLEPIAS SCHAFFNERI 71 C Fic. 1. Comparison of the flowers of Asclepias schaffneri, A. coulteri, and A. quinquedentata. A. Flower of A. schaffneri from a population in the Sierra de Morones, Zacatecas, México (Fishbein et al. 5846). B. Inflorescence of A. schaffneri from the same population in Zacatecas. C. Flower of A. coulteri from a population in the Sierra Gorda, Querétaro, México (Fishbein et al. 5172). D. Flowers of A. quinquedentata from a population in the Galiuro Mountains, Arizona, USA (Fishbein & King 2850). Photos by M. Fishbein (A, C, D) and L. Alvarado-Cardenas (B). it is represented in four major herbaria and were surely collected at a later date than the because of the epithet chosen by Gray. The other flowering specimens with the same catalogue Syntype is Parry & Palmer 582. One of the two number, which are more commonly represented sheets of Parry & Palmer 582 housed at MO | in herbaria. It is quite possible that the fruiting consists of two mature, fruiting specimens. These specimens were collected by Palmer, rather than 72 TABLE lI. also FIGs. 1, 3). Character Growth form Inflorescence Corolla color Corona color A. coulteri Sub-shrub with multiple, branched shoots that persist for more than l yr Flowers laxly spreading; a few may be pendent Green White MADRONO A. quinquedentata Herbaceous, annually renewed shoots that are basally branched with one to several fertile stems and several short, sterile stems Flowers all pendent Green, often with reddish or brownish tinge Brown or purplish at base, [Vol. 55 DISTINGUISHING CHARACTERS OF ASCLEPIAS COULTERI, A. QUINQUEDENTATA AND A. SCHAFFNERI (see A. schaffneri Herbaceous, annually renewed shoots that are unbranched, with 1—2 fertile stems Flowers laxly spreading; a few may be pendent Pink or magenta Pink or magenta, paler at white at apex apex Corona height More than twice as tall as Slightly surpassing anthers More than twice as tall as anthers anthers Corona appendage — Strongly incurved over style Strongly incurved over Erect to shghtly incurved (“horn’’) posture apex style apex Parry, who had left San Luis Potosi by August 3, 1878 (McVaugh 1956). The fruits of A. schaffneri and A. quinquedentata are not sufficiently known to permit definitive determination of the speci- mens bearing fruits that are mounted on a single sheet at MO. The architecture of the plant on the left side of the sheet is consistent with that of A. schaffneri. However, the plant on the right bears two short, sterile branches more suggestive of A. guinquedentata. Vhe probability that this sheet represents a mixed collection of the two species 1s bolstered by the existence of Parry & Palmer 583, which consists of bona fide flowering specimens of A. quinquedentata. It is conceivable that Parry erred in compiling the catalogue for the collec- tions when he associated fruiting specimens, which may have been collected by Palmer after Parry’s departure from the field, with flowering specimens that Parry may have collected earlier in — the year. Other than this putative mixed collec-_ tion, there is no evidence that these two species ever occur in sympatry. ASCLEPIAS QUINQUEDENTATA A. Gray, Proc. Amer. Acad. Arts 12:71. 1877.—TYPE: USA, West Texas [sic], on or near the San Pedro FIG. 2. Approximate collection localities of representative specimens of A. coulteri (squares), A. quinquedentata | (triangles), and A. schaffneri (circles). Type localities are indicated as follows: A. coulteri (MB), A. quinquedentata | (A), A. rzedowskii (%), A. schaffneri (@), A. zacatecana (@). 2008] Herbario Nacional de México (MEXU) Coleccién de Tipos ASCLEPIADACEAE INTIPO s Asclepias schaffneri A. Gray Proc, Amer, Acad, Arts 16: 103-104, 1881. Venificado por Verdnica Juarez daimes & LO Alvarado Cardenas Oct 2004 APD. Ase lepias schaffneri 4,: HERBARIO NACIONAL DE MEXICO aye (Mexu) reAcad. Arts Sc. 162104.1¢ Bd “afael Uerndndez Nagata 1987 Asclapias Zainguedentete A. sor TseotyPe of A. scheffner; A. Grey Determined by W.D. Stevens 198 sarden Missouri Botanical € BIG. 3. the Herbario Nacional, Universidad Nacional Autonoma de México (MEX). B. Isotype of A. guinquedentata A. Gray in the Missouri Botanical Garden Herbarium (MO). Photo used by permission of the Missouri Botanical Garden. River, 1851-2, C. Wright 1689 (holotype, GH, not seen, isotypes K!, MO!, US!). Fig. 3B. Asclepias quinquedentata A. Gray var. neomex- icana Greene ex A. Gray, Proc. Amer. Acad. Arts 16:103. 1880.—TYPE: USA, New Mex- ico, rocky mountain-side east of Pinos Altos, 22 June 1880, Greene s. n. (holotype, GH, not seen, isotype, MO!). Asclepias palmeri Vail, Bull. Torr. Bot. Club 25:171-172. 1898.—TYPE: MEXICO, San Luis Potosi, chiefly in the region of San Luis Potosi, 22° N Lat., 6000-8000 ft.,1878, Parry & Palmer 583 (holotype, GH!, isotypes, MO!, NY!). Asclepias amsonioides Standl., Field Mus. Nat. Hist., Bot. Ser. 22:44. 1940.—TYPE: MEX- ICO, Chihuahua, El Cima, 29 June 1936, LeSueur 848 (holotype, F!, isotypes, ARIZ!, GH!, TEX!, US)). ANA STATE UNIVERSITY IN SHREVEPORT FISHBEIN ET AL.: RESURRECTION OF ASCLEPIAS SCHAFFNERI 73 WALANG 2761964 MISSOURI BOTANICAL CARDEN HERE ARIOM {hla mn suilit ili vant vil ili 1H isotype of Asclegixs quaquededats A Gay Ps | | | 1 | I y Ret, Poe Amer Acad. Sey, 17: FL. 187% | | Miazous| Botanmal Garden (MOY Po Ween J SOR RRI o R n corae g otcaaoe Sraae goa actos Moses Botecte GHeOEM, A Type Specimen oS a HERB, MAG, aS Type specimens of Asclepias schaffneri and A. quinquedentata. A. Isolectotype of A. schaffneri A. Gray in Asclepias rzedowskii W.D. Stevens, Phytologia 53:402-403. 1983. Syn. nov.—TYPE: MEX- ICO, Estado de Mexico, parte baja de ladera sur del Cerro Sincoque, 2200 m, 15 April 1981, Romero & Rojas 34-1193 (holotype, ENCB, not found). Additional specimens seen. MEXICO, CHI- HUAHUA, Sierra Madre, Cusihuiriachic, 30 Au- gust 1887, Pringle s.n. (MEXU, MO); Sierra Madre, 1899, Nelson 6130 (US); Sierra Madre, Colonia Garcia, 11 July 1899, Townsend & Barber 125 (US, 2 sheets); Madera, 1908, Palmer 321 (ULS.)> Mipio, 2008] representing the size of the corolla and stamens. Both E. tracyi and E. brandegeeae were described as new (Mason 1945), each with a limited distribution in the inner North Coast Range of California. Craig had treated collections of these two taxa as “intergrades”. Although Mason described the range of E. sparsiflorum as ‘“‘East base of Cascades and Sierra Nevada, Tehachapi Mountains, and north on the west slope of the Sierra Nevada to Fresno County, California’, the only representative specimens from California that he cited were two from Mt. Pinos in Ventura County. Both of these (Hall 6580, Dudley & Lamb 4685) are what I’ve treated here as E. signatum. Mason did include his own collection from Nevada (Mottsville, Mason 12362), which is one of the few collections matching the type of E. sparsiflorum. Following Mason’s work, most of the major floras (e.g., Abrams 1951; Munz 1959; Hickman 1993), as well as a treatment of the genus by Harrison (1959, 1972), have echoed Mason with only minor changes. Harrison’s (1959) treatment of E. sparsiflorum included a longer list of representative collections covering similar regions as had Craig, but additionally included Howell 16142, and Rose 40675, both made in the 1940’s. These two are from near the type area in Fresno Co. The Howell collection is E. sparsiflorum, but the Rose collection is E. tracyi. The drawings of dissected corollas that Harrison (1959, 1968) presented to illustrate E. sparsiflorum are from material collected from Nevada just east of Lake Tahoe (Mason 12169). This collection however, is E. signatum, not E. sparsiflorum. What is evident from the works of these authors is that their concept of E. sparsiflorum was based primarily on a different plant than is represented by Eastwood’s type collection. This can also be observed by the annotations made by these authors on herbarium collections. None of the treatments prepared by these authors includ- ed Eastwood’s description of the plant as being “glandular-pubescent’’. ERIASTRUM SPARSIFLORUM Eriastrum sparsiflorum (Eastw.) H. Mason, Ma- drono 8: 86. 1945.—Gilia sparsiflora Eastw., Proc. Calif. Acad. III, 2: 291. 1902. Navarretia filifolia Kuntze subsp. sparsiflora Brand in Engler, Pflanzenreich IV, 250: 167. 1907. —Gilia filifolia Nutt. var. sparsiflora J. F. Macbr, Contr. Gray Herb. 49:57. 1917. — Hugelia filifolia var. sparsiflora Jepson, Man. Fl. Pl. Calif. 792. 1925. —Type: U.S.A., California, Fresno Co., Bubbs Creek, South Fork of Kings River, 1-13 July 1899, Alice Eastwood s.n. (holotype: CAS!). As described by Eastwood (1902), E. sparsi- florum is “minutely glandular-pubescent”. The GOWEN: NEW ERIASTRUM TAXA 85 floral heads tend to have a narrow appearance, with few flowers per head, as the name implies. The whole plant, which is generally much taller than E. signatum, appears sparse and elongated. Before seeing these plants in the field, I had surmised that Eastwood’s collection had possibly grown in the shade and was not representative of the plant in more natural conditions. I have since seen the plant growing in exposed sites and the flowers are still few and the floral heads are narrow. The bracts and leaves are often without lobes, or the lobes are short and barely angled away from the axis. An excellent drawing that illustrates the distinctive habit of E. sparsiflorum compared to E. signatum can be found in Harrison (1972). This drawing shows the type of E. sparsiflorum (Eastwood s.n.) as well as two collections of E. signatum, one from the Mt. Pinos region (7wisselmann 2316), the other northeastern California (Bacigalupi & Robbins 5417). The corollas of Eriastrum sparsiflorum are uniformly pale whitish with a slight tint of blue or lavender. They have the appearance of being white, compared to E. signatum, which generally appears light blue. (On several flowers I have found a barely perceptible dark fleck or two at the base of the corolla lobe that hints of the more obvious pattern of E. signatum, but such flowers are rare, and thus are unlikely to be encoun- tered.) The throat is slightly widened and the stamens are exserted about two anther lengths beyond the sinus of the corolla lobes in E. sparsiflorum. Eastwood described the flower as ‘“‘white with some purple dots in the funnel-form throat”. Eriastrum sparsiflorum H. Mason has been generally thought to be a widespread species, found from central Oregon to southern Califor- nia and western Nevada. However, only a handful of herbarium collections identified as E. sparsiflorum match the type of the species. (Most specimens that I examined were E. signatum.) Based on specimens matching the type specimen, the distribution of Eriastrum sparsiflorum in- cludes scattered locations in the southern Sierra Nevada of Fresno, Tulare, and northeastern Kern counties. It also occurs at the eastern base of the Sierras in Inyo County, the northwestern base of the White Mountains in Mono County, then north along eastern Alpine County and the western edge of Nevada to southeastern Lassen County. Most often it is found in granitic sand, frequently in sagebrush scrub. Representative collections: California. Alpine Co.: near Woodfords, Gowen 648 (JEPS). Fresno Co.: Kings River Canyon, Zumwalt Meadows, J.T. Howell 16142 (CAS). Inyo Co.: 7 mi. south of Bishop, Duran 1466 (UC); Keough Hot Springs Road, Gowen 64] (JEPS). Kern Co.:: 8 mi. west of Walker Pass, Hwy 178, Shevock 999 86 MADRONO (CAS). Lassen Co.: Red Rock Road, south of Constantia, Gowen 657 (JEPS). Mono Co.: near mouth of Queen Dick Canyon, Morefield & McCarty 3767 (UC); mouth of Marble Creek 4.4 mi. S 56° E of Benton, Morefield & McCarty 4028 (UC); Benton, Gowen 646 (JEPS). Tulare Co.: about 1 1/4 mi. north of Soda Spring on Freeman Creek, near trail to Jerky Meadow, C.N. Smith 1222 (JEPS); base of Church Dome, along FS 34E08, Shevock 8991 (CAS); Kennedy Meadows, Twisselmann 7860 (CAS, SBBG). Nevada. Ormsby Co.: Eagle Valley, C. F. Baker 1403 (UC). Douglas Co.: Mottsville, Mason 12362 (UC); above Mottsville, near Hwy 207, Gowen 651 (JEPS). ERIASTRUM TRACYI Eriastrum tracyi H. Mason, Madrono 8: 87. 1945. —Type: U.S.A., California, Trinity Co.: Hay- fork Valley, 30 June 1923, J.P. Tracy 6463 (holotype: UC!; isotypes JEPS! UC!). Recently E. tracyi has been considered a synonym of E. brandegeeae (Patterson 1993). Patterson did not elucidate his reason for synonymizing E. tracyi, but he was likely following comments made by Harrison (1959, 1972). Harrison (1959) maintained E. tracyi as a distinct entity “‘with considerable hesitancy”’ until ‘“‘a second collection is made in the area that may substantiate the distinction”. At the time of writing, Harrison believed that only the type collection, and possibly the disjunct Sharsmith 3299 collection from Santa Clara County, repre- sented E. tracyi. Eriastrum tracyi is frequently misidentified to E. sparsiflorum, especially when found in the southern Sierra Nevada, where it previously was not known to occur. Several collections annotated as E. sparsiflorum by Harrison are E. tracyi (e.g., Rose 40675, L.R. Short S-208). The distinctions between E. tracyi and E. brandegeeae are subtle and additional work is needed to better understand their relationship. Eriastrum tracyi has shorter stamens and wider corolla lobes. Collections from their type loca- tions, as well as garden plants grown from those collections can be separated on this basis. Plants found geographically between these two areas (i.e., in Colusa and Glenn counties) are more problematic. Several distinct forms are found, but they do not seem to show a definable gradation between the two species. Future molecular work might provide a better understanding of this group. The small anther size, short, mostly included stamens (they just reach the corolla sinus), and corolla lobe shape unite plants from the south- western Sierra with E. tracyi found in the type area of Trinity County. Plants found in Santa Clara County (e.g., Sharsmith 3299), as well as [Vol. 55 northeastern Shasta County also appear to belong here. In the southwestern Sierra Nevada it occurs in Kern, Tulare, and Fresno counties, most often associated with chaparral communi- ties. In the Kings River Canyon it is found not far from the type area of E. sparsiflorum. The corollas of Eriastrum tracyi are blue to almost white. Most often they have a yellow to white throat above the blue in the upper portion of the tube. The corolla is less widened in the throat and the lobes are shorter and wider compared to E. signatum and E. sparsiflorum. There is no spot at the corolla lobe base. Eriastrum tracyi usually develops only a single seed per locule, although the Sierra plants occasionally will develop two. Both E. sparsi- florum and E. signatum have two seeds (or rarely one) per locule. Representative collections: California. Fresno Co.: Kings River Canyon, junction of Middle & South Fork, Rose 40675 (CAS, UC); along California Hwy 180, 0.1 mi. west of Ten Mile Creek, Sequoia National Forest, Shevock 8774 (CAS); Bell Fire Rd., east of Auberry, Gowen 480 (JEPS). Kern Co.: Poso Creek near Poso Mine, C.N. Smith 1 (UC); on Greenhorn, about 6 mi. from head of Eugene Grade, C.N. Smith 1043 (JEPS); Havilah Ranger Station, L.R. Short S-208 (UC); hill just east of Bodfish Gap, Twisselmann 12149 (CAS). Santa Clara Co.: Arroyo Bayo, 17 June 2003, Gowen s.n. (JEPS). Shasta Co.: Fall River Mills-Cassell Rd., Oswald 9561 (JEPS); same area, Gowen 783 (JEPS). NEW COMBINATION IN ERIASTRUM Eriastrum harwoodii (T. T. Craig) D. Gowen, © stat. et comb. nov. —Gilia filifolia var. — harwoodii T. T. Craig, Bull. Torrey Bot. Club | 61: 424-425. 1934. Hugelia diffusa var. harwoodii Jepson, Fl. Calif. 3: 167. 1943. — , Eriastrum diffusum subsp. harwoodii (YT. T. | Craig) H. L. Mason, Madrofno 8: 77. 1945. — Eriastrum sparsiflorum subsp. harwoodii (T. T. — Craig) H. K. Harrison, Phytomorphology 18: | 401. 1968. —Type: U.S.A., California, River- | side Co., sandy desert 1200 ft., Blythe Junc- tion, 2 April 1920, Munz & Harwood 3589, © (holotype: POM!). Eriastrum harwoodii was treated as a variety or | subspecies of Gilia filifolia, E. diffusum (A. Gray) © H. Mason, and E. sparsiflorum, before being placed in synonomy under E. sparsiflorum by Patterson (1993). As noted in the discussion | above, previous workers based their understand- ing of E. sparsiflorum on E. signatum. Rather than now moving E. harwoodii to either E. signatum or | E. sparsiflorum, it is elevated to species level | because it is morphologically, ecologically, and geographically distinct from the other two species. This proposal parallels the treatment of E. 2008] diffusum and E. sparsiflorum, also formerly treated as varieties under Gilia filifolia, and subsequently elevated to species status. The entire plant of Eriastrum harwoodii 1s conspicuously woolly, in marked contrast to E. sparsiflorum, which is minutely glandular-pubes- cent, or E. signatum, which is sparsely floccose. The flower is straw yellow or whitish with apiculate corolla lobes. The corolla lobes are shorter and wider than E. sparsiflorum or E. GOWEN: NEW ERIASTRUM TAXA 87 signatum. Eriastrum harwoodii occurs outside the ranges of both, being known only from sandy desert areas of eastern San Bernardino and Riverside counties. Representative collections: California. River- side Co.: 3 mi. south of Rice on Blythe Road, Wolf 3119 (UC). San Bernardino Co.: Kelso, June 1915 Brandegee s.n. (UC); Dale Dry Lake, P. A. Munz 15690 (UC); slopes south of Dale Dry Lake, Bacigalupi 6240 (JEPS). KEY TO ANNUAL ERIASTRUM WITH STAMENS EXSERTED LESS THAN 1/2 COROLLA LOBE LENGTH 1. Anthers exserted beyond sinus of corolla. 2. Stamens unequal, longest filament twice or more the length of the anther; leaves often 3—5-lobed Pe i a E. wilcoxii 2' Stamens equal, filaments less than twice the length of the anther; leaves often entire. 3. Plant minutely glandular-pubescent....... Dene ier es ene ere Se Acting eee ra ens E. sparsiflorum 3’ Plant floccose to woolly, not minutely glandular-pubescent. 4. Corolla lobes light blue, with dark colored spot at base, rounded apically; mainly of pinyon- juniper woodland E. signatum 4’ Corolla lobes straw yellow or whitish, without dark colored spot at base, apiculate apically; mainly of creosote bush scrub 1’ Anthers below or just at the sinus of corolla. Se ahaa Meclhisiek a: Shit i aed Ga Ge oles E. harwoodii 5. Upper leaves and bracts pinnately 3—7-lobed; inflorescences many flowered, densely bracteate, dense wool matted at base.................... wr Verdes dee ee cae, oe ce, Gah cease he Gch rs ca E. abramsii 5’ Upper leaves and bracts 1—3-parted, or palmately 5-lobed; inflorescences few flowered, bracts few, loosely woolly. 6. Corolla 7-10 mm long, blue to lavender or white; seeds 1(—2) per locule. 7. Corolla light blue to white, lobes 1.2—1.5 mm wide (length = 2 width); stamens 0.75—1.5 mm long, filaments equal anther length .... Ssh te clas ttet eee oe, Gace heel hates aa ay 2 ey eee ee ges ee E. tracyi 7’ Corolla light blue to lavender, lobes 0.9—1.1 mm wide (length = 3X width); stamens 1.5—2 mm long, filaments twice anther length .... das KE ws ey Sd eee eee aes E. brandegeeae 6’ Corolla 5-7 mm long, white; seeds 2-4 per locule............. 20.0.0... 000002 e eee E. hooveri ACKNOWLEDGMENTS First and foremost I wish to thank Barbara Ertter, a true mentor and friend. Dr. Ertter not only provided encouragement and guidance, but also obtained the grant to fund the illustration. Thanks also to the entire staff at UC/JEPS, but especially to the amazingly patient and helpful Kim Kersh. Much appreciation is also extended to the curators and staff at CAS, RENO, and SBBG for providing hospitality and access. John Strother graciously translated the Latin diagnosis. Linda Vorobik brought life to a very dead flattened plant with her illustration, which was funded by the Lawrence R. Heckard Endowment Fund of the Jepson Herbarium. I thank Sarah De Groot, John Hunter, Leigh Johnson, Jim Reveal, and an anonymous reviewer for many comments that have substantially improved this paper. LITERATURE CITED ABRAMS, L. 1951. Illustrated flora of the pacific states, Washington, Oregon, and California. Vol. III. Geraniaceae to Scrophulariaceae. Stanford Uni- versity Press, Stanford, CA. CRAIG, T. 1934. A revision of the subgenus Hugelia of the genus Gilia (Polemoniaceae). Bulletin of the Torrey Botanical Club 61:385—428. EAsTwoop, A. 1902. New species from the Sierra Nevada mountains of California. Proceedings of the California Academy of Sciences, 3™ Series 2:285—-293. HARRISON, H. K. 1959. Morphological and taxonomic studies in the genus Eriastrum. Ph.D. Thesis. University of California, Berkeley, CA. . 1968. Contributions to the study of the genus Eriastrum. 1. The corolla and androecium. Phyto- morphology 18:393—402. . 1972. Contributions to the study of the genus Eriastrum 2. Notes concerning the type specimens and descriptions of the species. Brigham Young University Science Bulletin, Biological Series 16:1—26. HICKMAN, J. C. (ed.). 1993. The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. MASON, H. L. 1945. The genus Eriastrum and the influence of Bentham and Gray upon the problem of generic confusion in Polemoniaceae. Madrono 8:65—91. Munz, P. A. 1959. A California flora. University of California Press, Berkeley, CA. PATTERSON, R. W. 1993. Eriastrum. Pp. 826-828 in J.C. Hickman (ed.), The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. TWISSELMANN, E. C. 1967. A flora of Kern County, California. Wasmann Journal of Biology 25:308. MADRONO, Vol. 55, No. 1, pp. 88-92, 2008 NOTEWORTHY COLLECTIONS ARIZONA DACTYLOCTENIUM RADULANS (R. Br.) P. Beauv. (POACEAE) [button grass].—Pima Co., Marana, ca. 20 km NNW of Tucson, Continental Ranch community and the adjacent floodplain (west bank) of the Santa Cruz River, Sonoran desertscrub on farmland fallow for over 10 yr, summer ephemeral occurring in low areas and drainages on silty-clay substrate. Locally common around a mudhole on floodplain, with Atriplex polycarpa, Bouteloua aristidoides, B. barbata, Chloris virgata, and Pennisetum ciliare, 32°21'45.5"N, 111°06'35.7”’W, elev. 648 m, 9 Sept 2006, J.-F. Wiens 2006-157 (ARIZ, ASDM). Rare near the top of concrete embankment of the Santa Cruz River, with Bouteloua aristidoides, B. barbata, Chloris virgata, Eragrostis echinochloidea, Iso- coma teniusecta, Larrea divaricata, Panicum coloratum, Pennisetum ciliare, and Prosopis velutina, 32°22'07.3"N 111°06’43.1"W, elev. 646 m, 9 Sept 2006, J.F. Wiens 2006-160 (ARIZ, ASDM). Locally abundant along shallow drainage on the floodplain, with Atriplex polycarpa, Chloris virgata, Larrea divaricata, Pennisetum ciliare, and Prosopis velutina, 32°22) 10.2"N 111°06’50.4"W, 641 m, 9 Sept 2006, JF. Wiens 2006- 162 (ARIZ, ASDM). Rare in concrete-sided drainage within the community, with Bouteloua aristidoides, Eragrostis lehmanniana, and Sorghum _ halepense, 32°22'18.4"N, 111°08’09.6"W, elev. 640 m, 24 Sept 2006, M.F. Hanson 2006-01(ARIZ). Rare in concrete- sided drainage within the community, with Bouteloua aristidoides, Cynodon dactylon, Eragrostis lehmanniana, and Prosopis velutina, 32°23'08.9"N, 111°07'36.6"W, elev. 636m, 24 Sept 2006, M.F. Hanson 2006-02 (ASDM). Locally common in concrete-sided drainage within the community, near Silverbell and Cortaro Farms Rds., with Amaranthus palmeri, Bouteloua aris- tidoides, Chloris virgata, Dactyloctenium aegyptium, and Eragrostis lehmanniana, 32°20'59.1"N, 111°06'05.0"W, elev. 651 m, 25 Sept 2006, J.F. Wiens 2006-183 (ARIZ, ASU, UT, US, ASDM), det. J. R. Reeder. Rare (solitary plant) found on a graded site 12 km WSW of Tucson (on Eagle Cove Drive, just E of the intersection of AZ Hwy. 86 and W. Valencia Rd.), with Prosopis velutina, Pennisetum ciliare, Chloris virgata, and Salsola tragus, 32°08'15.5”N, 111°07'58.0"W, elev. 750 m, 30 Aug 2007, C. Hemingway 2007-01 (ASDM). Previous knowledge. Dactyloctenium radulans is native to tussock grasslands to open-woodland areas of New South Wales and Queensland in Australia. In agricul- tural areas of that region it is found as a weed in pastures and along irrigation ditches and roads (Queensland Environmental Protection Agency, 2003, Regional Ecosystems of the Desert Uplands; http//www. epa.qld.gov.au/media/nature_conservation/biodiversity/ desert_uplands/Factsheets/100901.htm). Three U.S. col- lections from 1957 to 1960 were from Jamestown, Berkeley Co., South Carolina (H.E. Ahles 30800, 35612, 53777, MASS). These were all from the waste ground surrounding the Santee Wool Combing Mill on SC Rte #45 (K.B. Searcy, MASS, personal communi- cation). This fits a likely introduction scenario, in that the seed could easily have been transported in wool imported from Australia (J.R. Reeder, personal com- munication). The USDA website (http://plants.usda. gov/java/profile?symbol=DARA2) mentions its occur- rence (authors unable to confirm) in Massachusetts and Florida. The only previous Arizona specimens were from cultivated greenhouse stock at the Plant Materials Center (Southwest Nurseries, Soil Conservation Ser- vice, U.S. Department of Agriculture) formerly on the University of Arizona campus, Tucson (A4.R. Purchase A-751, 11 July 1939, and J. Burrell s.n., 12 Sept 1959, ARIZ). The Burrell collection label reads “‘Introduced in 1933, as plant introduction no. 106469, from Australia, now established at Plant Material Center, Tucson.” The current manager of the Arizona Plant Materials Center, USDA — NRCS in Tucson (now located near Campbell Road and the Rillito Wash) finds no archival mention of the species ever being at the PMC, although early records are incomplete (R. Garner, personal communication). Significance. First collections of wild plants in the western United States. The two earlier Arizona collections were from cultivated plants in a nursery | (no longer existing), and the species has not been seen in Arizona for over 40 yr. An extensive search of similar | habitat 20 km upstream from the 2006 collection sites on the Santa Cruz River, and two major tributaries, the Rillito Wash and the Canada del Oro as well as likely habitat for 85 km downstream, yielded no additional D. radulans. Channelized washes draining the southern | Tucson Mountains through the Continental Ranch area | were checked thoroughly. The survey led us to believe | D. radulans was contained within the Continental Ranch neighborhood and adjacent Santa Cruz flood- plain, an area of ca. 790 ha, bounded on the west by Silverbell Rd. (Hanson 2006-01), the south by Cortaro | Rd. (Wiens 2006-183), an un-named, side drainage to the north (Hanson and 2006-02), and the east by the Santa Cruz River (Wiens 2006-160). There D. radulans seems to prefer shallow washes where seasonal flood- | waters move slowly and don’t scour the substrate. Due to the many sites within the neighborhood, D. radulans must have been present locally for at least one © season previous to summer 2006, although not observed until this year. It appears not to be particularly | competitive with other plant species, especially peren- nial grasses, but its seed are likely spread by mechanical | disking that is used for weed control in the neighbor- hood drainages. The Hemingway (2007-01) collection came from the Brawley Wash watershed (running parallel to the Santa Cruz River Watershed), approx- imately 45 km SSE of the 2006 Wiens and Hanson | collections. Due to the discovery of this plant on a disturbed construction site so far from the larger | population, we feel it may have been introduced to this area via dirty grading equipment. We see no reason that — it could not spread to favorable river floodplain habitat, — agricultural fields, and construction sites in both watersheds. —JOHN F. WIENS, Department of Botany, Arizona- © Sonora Desert Museum, 2021 N. Kinney Road, Tucson, AZ 85743; MARILYN F. HANSON, 7105 W. | Deserama Drive, Tucson, AZ 85743; and CARROLL 2008] HEMINGWAY, 6207 S. Eagle Cove Drive, Tucson, AZ 85757. CALIFORNIA CALOCEDRUS DECURRENS (Torrey) Florin (CUPRES- SACEAE).—San Diego Co., Bucksnort Mountain, two seedlings that established after a wildfire in 2003 were located on a concave west-facing slope of 13°, at the uppermost extent of an unnamed drainage, at an elevation of 1820 m, about 350 m S of Combs Peak on the Bucksnort Mountain USGS 7.5 min topographic quadrangle, NW 1/4 NE 1/4, Sec. 18, T9S, R4 E, San Bernardino Base Meridian, 33°23’29"N 116°36'18"W, 23 March 2007, Goforth & Minnich s.n. (UCR). Previous knowledge. Extensive field mapping across the southern California Peninsular range in the Vegeta- tion Type Map (VTM) survey (Griffin and Critchfield 1976, Research Paper PSW-82, U.S. Forest Service) and subsequent searches using aerial photography (Minnich and Everett 2001, Madrono 49:177—197) recorded C. decurrens on the San Jacinto Mountains, Palomar Mountain, Hot Springs Mountain, Volcan Mountain, and the Cuyamaca-Laguna Mountains. C. decurrens was not known to occur on Bucksnort Mountain (Beauchamp 1986, A flora of San Diego County, California, Sweetwater River Press). Significance. This is the first collection of C. decurrens from Bucksnort Mountain, documenting long-distance seed dispersal and successful establishment at a site burned in July 2003 (3 growing seasons prior to the collection). The nearest stands occur on the summits of Hot Springs Mountain (10 km S), Palomar Mountain (25 km W) and Santa Rosa Mountain (30 km NE; Minnich and Everett 2001, Joc. cit.). Intervening basins and mountains separating these widely disjunct stands are covered by chaparral, southern oak woodland, nonnative annual grassland, and creosote brush scrub. The seedlings of C. decurrens were discovered 9 m apart while surveying stands of fire-killed Pinus coulteri D. Don at Bucksnort Mountain. They were 24.0 cm and 25.5 cm in height, with basal diameters of 4.0 mm and 5.0mm. Subsequent ground searches across Bucksnort Mountain failed to detect additional seed- lings, or a parent tree. Bucksnort Mountain (1880 m) is the highest coastal range mountain in southern California lacking mixed conifer forest, and is arid (average annual precipitation ~40 cm) due to its rain-shadowed position NE of Palomar Mountain. Summit locations only support stands of P. coulteri, while Hot Springs Mountain located south of the Palomar Mountain rainshadow has mixed conifer forest at the same elevation (Minnich and Everett 2001, Joc. cit.). Soils are derived from colluvium of weakly-metamorphosed granitic rock, with an A horizon having sandy-loam texture. Associated species include abundant seedlings of P. coulteri (350 ha~') ina fire-killed stand with 310 trees ha~', as well as Arcto- staphylos glandulosa Eastw., Ceanothus leucodermis E. Greene, Quercus chrysolepis Liebm, Rhamnus californica ssp. tomentella (Benth.) C.B. Wolf, Solidago californica Nutt., and Turricula parryi (A. Gray) J.F. Macbr. Colonization of Bucksnort Mountain by C. decurrens was facilitated by wildfire and subsequent near-record rainfall. The 2003 burn cleared dense shrub cover which diminished evapotranspiration and exposed bare min- NOTEWORTHY COLLECTIONS 89 eral soil for germination. The 2004-2005 winter season was one of the wettest on record (~250% of normal rainfall). We believe on the basis of chance that a cone was dispersed to the site rather than individual seeds. C. decurrens usually has 4 fertile seeds per cone, and the lightweight cones are small (<2.5 cm in length; Hick- man 1993, The Jepson Manual: Higher Plants of California, University of California Press, Berkeley, CA). We speculate that seed was dispersed to the site after the fire by a bird that could carry a seed cone, such as a Scrub Jay, Nuttall’s Woodpecker, or Raven. It is doubtful that strong winds could carry winged seed to Bucksnort Mountain from the nearest stand. It is also unlikely that the seedlings were planted because the site is far from the nearest road or trail. A species distribution may expand if seed dispersal results in successful establishment of new outlying populations (Sauer 1988, Plant Migration: The Dynam- ics of Geographic Patterning in Seed Plant Species, University of California Press, Berkeley, CA). However, events of long-distance plant migration as this collec- tion documents are rarely witnessed. The seedlings have survived a most hazardous phase of their ontogeny, 1.e., establishment, and it is possible they will grow to reproductive maturity. Since many P. coulteri trees at Combs Peak exhibit basal fire scars, indicating survival of past wildfire, mature C. decurrens trees could also endure such burn events at the site, and perhaps reproduce subsequent generations. On the other hand, colonization may fail if the seedlings die due to the site’s aridity or a future stand-replacement fire; in which case, their establishment would illustrate how plant distribu- tions are maintained in equilibrium between an inherent tendency to disperse and constraining environmental conditions. —BRETT R. GOFORTH and RICHARD A. MINNICH, Department of Earth Sciences, University of California, Riverside, CA 92521. CALIFORNIA CAREX LONGI MACK. (CYPERACEAE).—Shasta Co., along the north side of Rt. 299, about 100 ft W. of the intersection with Jim Harvey Road and about 3.8 mi E. of the Interstate-5 overpass. Elev. 650 ft. 40°37'35.5"N, 122°17'36.1".W. UTM 10 559765E, 449754IN (NAD83/WGS84). Uncommon herbs to 2 ft tall, growing in mud along a small seasonal creek and moist swale in an open woodland on loamy soil (Redding gravelly loam, 0 to 3% slopes; fine, mixed, active, thermic Abruptic Durixeralfs), with Aira car- yophyllea, Anthoxanthum aristatum, Avena barbata, Briza maxima, B. minor, Bromus secalinus, Cyperus eragrostis, Dichelostemma sp., Eleocharis macrostachya, Eryngium articulatum, Glyceria occidentalis, Juncus tenuis, Lolium multiflorum, Mimulus guttatus, Odontos- tomum hartwegii, Pinus sabiniana, Quercus douglasii, Q. wisilzenii, Rumex crispus, Taeniatherum caput-medusae & Trifolium hirtum. 16 May 2005, D. Goldman 3316 (BH, GH, MICH, OSC, UC). The site was revisited during the dry season in late July 2005 and the following species were also found: Hypericum perforatum, Lactuca serriola, Lotus purshia- nus, Mentha pulegium, and Paspalum dilatatum. 90 This site was a visually attractive but an otherwise unremarkable habitat. Several common vascular plant species were collected at this site, but this Carex proved the most difficult to identify. The specimen was then sent for expert identification and it was determined to be Carex longii. Previous knowledge. Carex longii is a widespread wetland and wetland-margin species native to eastern North America (Ontario and Nova Scotia south to Texas and Florida; J. Mastrogiuseppe et al., 2002, Carex Linneaus sect. Ovales Kunth, pp. 332—378 in Flora of North America Editorial Committee (ed.), Flora of North America, north of Mexico, vol. 23, Cyperaceae, Oxford University Press, Oxford, U.K.), central Mexico through Central America (A.A. Rezni- cek, 1993, Carex L., pp. 243-267 in R. McVaugh & W.R. Anderson (eds.), Flora Novo-Galiciana, vol. 13, Limnocharitaceae to Typhaceae, University of Michi- gan Herbarium, Ann Arbor, MI; A.O. Chater, 1994, Carex L., pp. 464-473 in G. Davidse, M. Sousa & A.O. Chater (eds.), Flora Mesoamericana, vol. 6, Alismata- ceae a Cyperaceae, Universidad Nacional Autonoma de Mexico, Mexico, D.F., México), Hispaniola & Ber- muda (Mastrogiuseppe et al., ibid.), to south-central South America (G.A. Wheeler, 1987, A new species of Carex (Cyperaceae) from western South America and a new combination in the genus, Aliso 11: 533-537), plants in the latter region representing var. meridionalis. A good morphological summary of the species is provided by P.E. Rothrock (1991, The identity of Carex albolutescens, C. festucacea and C._ longii (Cyperaceae), Rhodora 93: 51-66). This species has also been reported as introduced in cultivated cranberry bogs in western Oregon and Washington (P.F. Zika, 2000, Noteworthy collections, Oregon & Washington, Madrofio 47:213—216), and moist to wet areas in Hawaii (M.T. Strong & W.L. Wagner, 1997, New and noteworthy Cyperaceae from the Hawaiian Islands, Bishop Museum Occasional Papers 48:37—-50) and New Zealand (A.J. Healy & E. Edgar, 1980, Flora of New Zealand, vol. 3, adventive cyperaceous, petalous and spathaceous monocotyle- dons, Government Printer, Wellington, New Zealand). The rather common habitat of this species in California, a roadside swale that is seasonally wet and alternatively very dry and hot, could suggest that it may be more widespread in the western United States. It has been regarded by some authors as somewhat weedy or favoring disturbed habitats (R.K. Godfrey & J.W. Wooten, 1979, [as C. albolutescens misapplied] Aquatic & wetland plants of southeastern United States, University of Georgia Press, Athens, GA; J. Gomez- Laurito, 2003, Cyperaceae, pp. 458-551 in B.E. Ham- mel, M.H. Grayum, C. Herrera & N. Zamora (eds.), Manual de plantas de Costa Rica, vol. 2, gimnospermas y monocotiledéneas (Agavaceae — Musaceae), Missouri Botanical Garden Press, St. Louis, MO; A.A. Reznicek, ibid.; P.F. Zika, ibid.). Because it is something of a generalist in its habitat preferences, it is not surprising that it has become established well outside of its native range and therefore it seems have the potential to become widely naturalized elsewhere. Although Zika (2000, ibid.) suggests that it may have been introduced to Oregon and Washington via the transport of Vacci- nium macrocarpon between agricultural areas, how it was introduced to other regions remains unknown. Significance. First record for California. MADRONO [Vol. 55 —DOUGLAS GOLDMAN, Harvard University Herbar- ia, 22 Divinity Avenue, Cambridge, MA 02138. dgoldman@fas.harvard.edu. ACKNOWLEDGMENTS Tony Reznicek for generously helping to identify this Carex specimen, Mary Diehl for Hawaiian geography information and Hobbes Goldman for general support | throughout. CALIFORNIA RYTIDOSPERMA CAESPITOSUM (Gaudich.) Conner & | Edgar (POACEAE).—San Diego County, City of San | Diego: Lusardi Creek MSCP parcel, ridges south of | Lusardi Creek, 0.6 km NNE Low South Survey Marker (33°00'26"N, 117°05'58”"W; T13S, R3W, SW/4 SW/4 sect. 26), alt. 93 m; locally common on rocky ridge on cobbly clay in openings of chamise chaparral. F.M. Roberts 5434, 18 Apr 2001 (RSA); City of San Diego: near Fairbanks Ranch, ridge south | of Lusardi Creek, ca. 0.2 km north San Dieguito Rd. and 1.3 km NE _ Fairbanks’ Lake. 117°09'59"W (T13S, R3W NW/4 Sec 35) UTM Z1I1S; 04 84 468 mE, 36 51 734mN, NAD 27, alt. 83 m; 33°00'20’"N, | locally abundant on southeast-facing slope on clay soil | in mixed grassland bordering coastal sage scrub at edge of fresh grading. F. M. Roberts 6136, 29 Mar 2005 | (RSA, SD); City of San Diego: near Fairbanks Ranch, | ridge south of Lusardi Creek, just north of San | Dieguito Rd. and 1.7km NE _ Fairbanks (SD). Previous knowledge. Rytidosperma caespitosum is native to southern Australia. Darbyshire and Connor | (2003, Flora of North America 25: 310) stated that the / species has been grown experimentally in the United | States and Canada but is not known to have escaped or persisted. However, specimens deposited at UC (Clifton | s. n., 4 Aug 1992; Ertter 17815a, 16 Feb 2002; Ertter 1823la, 26 Jun 2003) suggest the species is possibly © in Alameda and San Mateo counties, | naturalized California. Significance. First wild collection of this non-native perennial grass for southern California. When first located on the ridge above Lusardi Creek, the grass was | known from only a few clumping individuals growing | along roads and open areas on a cobbly ridge in> It was also observed in small | numbers on the opposite (north) ridge in disturbed : areas bordering coastal sage scrub. The grass was noted - during a 2001 base-line rare plant survey of the newly | established Lusardi Creek parcel within San Diego> County’s Multiple Species Conservation Plan (MSCP) | area. By 2005 the plant had spread significantly to the | southeast and in some places formed dense stands mixing with Stipa pulchra, Avena barbata, and Vulpia — myuros. The origin of this grass is uncertain but based | on the spreading observed and the difference in- distribution in 2001 and 2005, Rytidosperma caespito-. chamise chaparral. sum probably reached this area during the 1990’s. It can be expected to be found in adjacent areas of San Diego | Lake | (33°00'19"N, 117°09'53”W), alt. 124 m; fairly frequent © on hilltop along dirt road on clay soil in non-native | annual grassland. F. M. Roberts 6137, 29 Mar 2005 | 2008] County. This grass could prove invasive and contribute to the decline of native grassland systems within the MSCP preserve system. —TFRED M. ROBERTS, P.O. Box 517, San Luis Rey, California, 92068; and J. TRAVIS COLUMBUS, Rancho Santa Ana Botanic Garden, 1500 North College Avenue, Claremont, CA 91711-3157. MONTANA RANUNCULUS JOVIS A. Nels. [R. digitatus Hook.] (RANUCULACEAB), Carbon County, Montana; East Pryor Mountain along USFS Road 2849 as on Big Ice Cave, Mt. USGS Quad. 45°10'N, 108°25’W. In places abundant. Collections made by the authors on 19 May 2005 with the plant in flower and on 03 June 2005 with the plant in fruit. Voucher specimens held at MONT, Montana State University, Bozeman, MT and at Rocky Mountain College, Billings, MT. Prior to our discovery, Ranunculus jovis was not known to be in the Pryor Mountains. The Montana Natural Heritage program lists R. jovis as S1/G4. Exploring East Pryor Mountain in 2006 we found eleven sites separated into two populations 2.5 km apart. The total estimated for the two populations is 221,700 plants. Previous knowledge. Ranunculus jovis is found in mountains of Idaho, Utah and Wyoming. Heretofore collections had been made in Montana _ locations immediately north and west of Yellowstone National Park. There are no known reports of R. jovis along the rims of the Bighorn Basin of which the Pryor Mountains are the northern terminus. Significance and comment. This finding of R. jovis in the Pryor Mountains is a significant 130 km extension northeast of the plant’s known range, over the Beartooth — Absaroka mountain ranges from the nearest previ- ously known population in the northeast corner of Yellowstone National Park. Within the Pryors, R. jovis is an ephemeral plant that emerges with Claytonia lanceolata from melting snow- bank communities in early spring. The snowbanks are sufficiently deep to support subnivean activity of pocket gophers (Thomomys talpoides). This association was consistent at every site we found. This association has not been noted in collections done elsewhere; however, whenever the site has been described it is often noted as being at the foot of a melting snowbank. As with other spring ephemeral/deep snowbank plants, R. jovis has evolved storage roots. Ranunculus species have relatively long-tapered roots occasionally described as somewhat fleshy. In R. jovis those roots are decidedly fleshy, and are best described as clavate, resembling a club or better, a baseball bat. R. jovis occurs in various soil types and plant communities. Within the Pryors the collection sites varied from Artemisia tridentata/grasslands with loamy clay soil among limestone cobbles at 2134 m elevation to openings within the Pseudotsuga menziesii forest at 2438 m with soil richly organic and overlain by mucky duff. Our two populations of R. jovis in the Pryors had not before been noticed by botanists because, in the past, weather and road conditions prohibited explora- NOTEWORTHY COLLECTIONS 91 tion of these mountains in early spring. Warm climate/ drought conditions have occurred during the last seven years and with the diminished snow pack we can reach the populations as the plants emerge from the snowbanks. We are now conducting studies in the field on pollination, root development, and the relationship of R. jovis to Ranunculus glaberrimus, an often sympatric species that has similar above-ground morphology but very different roots. —JENNIFER LYMAN, Ph.D., Rocky Mountain Col- lege, Dept of Botany and Environmental Sciences, 1511 Poly Drive Billings, Mt 59102; and CLAYTON McCCRACKEN, M.D., 3227 Country Club Circle, Billings, MT 59102. PERU GENTIANELLA CALANCHOIDES (Gilg) Fabris (Gen- tianaceae).—Huancavelica: Tayacaja Province, 40 km from Colcabamba, 4070 m, 31 Jul 1978, Aronson & Berry 567 (HAM, MO). Previous knowledge. Known from Deptos. Huanuco and Junin. Previous knowledge of the distribution of the species reported here is based on my studies for L. Brako and J.L. Zarucchi, Catalogue of the Flowering Plants and Gymnosperms of Peru, Monographs in Systematic Botany, Missouri Botanical Garden 45, 1993, and P.M. Jorgensen and S. Leon-Yanez, Catalogue of the Vascular Plants of Ecuador, Monographs in Systematic Botany, Missouri Botanical Garden 75, 1999. Significance. Extends the known range to Depto. Huancavelica. GENTIANELLA ERICOTHAMNA (Gilg) Zarucchi (Gen- tianaceae).—Pasco: Prov. Oxapampa, Distr. Huanca- bamba, P.N., Yanachaga-Chemillen, Sector Santa Barbara, bosque montano con abundante chusquea, 10°20’06"S, 75°38'42”"W, 3340 m, 11 Mar 2004, Vasquez & Monteagudo 29981 and 29983 (both HAM, MO). Previous knowledge. Known only from the type from Depto. Huanuco, Prov. Huamalies, Berge stidwestlich von Monzon, 3300-3500 m, collected prior to 1906. Significance. Indicates that the species 1s extant and is distinct from G. radicata (Griseb.) J.S.Pringle, which also occurs in Depto. Pasco. Extends the known range to Depto. Pasco. GENTIANELLA GILIOIDES (Gilg) Fabris (Gentiana- ceae).—Cajamarca: Prov. San Ignacio, San José de Lourdes, cerro Picorana, bosque enano, 4°58'17"S, 78°53'00"W, 2830 m, 17 Aug 1998, Campos et al. 5547 (HAM, MO). Previous knowledge. Known only from Provs. Loja and Zamora-Chinchipe, Ecuador. Significance. This specimen, from near the Ecuador- ean border, is the first record of the species from Peru. GENTIANELLA HERRERAE (Gilg) Zarucchi (Gen- tianaceae).—Ayacucho: Prov. Huanta: road from Quinua to Tambo, S12°59'W074°05’, 4300 m, 19 Feb 2000, Weigend & Weigend 2000/373 (NY). Previous knowledge. Known only from the type from Depto. Cusco, Andes del Paucartambo, 3900 m, col- lected in 1924. Significance. Confirms that G. herrerae is a distinctive species, encourages the hope that it remains extant, and extends its known range to Depto. Ayacucho. 92 MADRONO Comment. The corolla of G. herrerae was originally described as ‘“‘apparently violet with a yellow base” (translation). In this collection the corollas, as seen in this study and as described by the collectors (in sched.), are bright yellow. Forms differing in having, respec- tively, violet and yellow corollas occur in other South American species of Gentianella. GENTIANELLA POLYANTHA J.S.Pringle (Gentiana- ceae).—Cajamarca: Prov. Jaén, Sallique, localidad El [Vol. 55 Paramo, 05°40’50"S, 079°16'20”"W, 3200 m, 23 Jun 1998, Campos et al. 5079 (HAM, MO). Previous knowledge. Known only from Prov. Loja, Ecuador. Significance. First record for Peru. —JAMES S. PRINGLE, Royal Botanical Gardens, P.O. Box 399, Hamilton, Ontario, Canada L&8N 3H8. Volume 55, Number 1, pages 1—92, published 16 April 2008 SUBSCRIPTIONS — MEMBERSHIP Membership in the California Botanical Society is open to individuals ($35 per year; family $40 per year; emeritus $27 per year; students $27 per year for a maximum of 7 years). Late fees may be assessed. Members of the Society receive MADRONO free. Institutional subscriptions to MADRONO are available ($70). Membership is based on a calendar year only. Life memberships are $750. Applications for membership (including dues), orders for subscriptions, and renewal payments should be sent to CBS c/o Jepson Herbarium, University of California, Berkeley, CA 94720-2465. Requests and rates for back issues, changes of address, and undelivered copies of Ma- DRONO 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, five 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 abbreviations 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 are encouraged to include the names, addresses, and e-mail addresses of two to four potential reviewers with their submitted manuscript. 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 included 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 MADRONO. Joint au- thors 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 manuscript have been used in other papers that are published, in press, submitted, or soon to be submitted elsewhere. TINT SONAN INSTITUTION LIBRARIES VOLUME 55, NUMBER 2 APRIL-JUNE 2008 MADRONO A WEST AMERICAN JOURNAL OF BOTANY ‘CONTENTS CATALOGUE OF NONNATIVE VASCULAR PLANTS OCCURRING SPONTANEOUSLY IN CALIFORNIA BEYOND THOSE ADDRESSED IN THE JEPSON MANUAL — Part Il Ellen Dean, Fred Hrusa, Gordon Leppig, Andrew Sanders, CGMS CV GIT oases ees ee 93 THE SALSOLA TRAGUS COMPLEX IN CALIFORNIA (CHENOPODIACEAE): CHARACTERIZATION AND STATUS OF SALSOLA AUSTRALIS AND THE AUTOCHTHONOUS ALLOPOLYPLOID SALSOLA RYANII SP. Nov. Gol AP USAGES. EWG O SRI ato enacho te watt eee ee css Ose 113 GENETIC EVIDENCE OF HYBRIDIZATION BETWEEN OENOTHERA WOLFII (WOLF’S EVENING PRIMROSE) AND O. GLAZIOVIANA, A GARDEN ESCAPE Jennifer DeWoody, Leonel Arguello, David Imper, Robert D. Westfall, Gnd VAVCHICTD TAKING Je cetes.cesesaasaav sees sat oe tean eine e e 132 LEAF ANATOMY OF ORCUTTIEAE (POACEAE: CHLORIDOIDEAE): MORE EVIDENCE OF C, PHOTOSYNTHESIS WITHOUT KRANZ ANATOMY Laura M. Boykin, William T: Pockman, and Timothy K. Lowrey.........00++- 143 DIRECT AND INDIRECT EFFECTS OF HOST PLANTS: IMPLICATIONS FOR REINTRODUCTION OF AN ENDANGERED HEMIPARASITIC PLANT (CASTILLEJA LEVISECTA) Beth A. Lawrence and IhomasN oR ave 23h en hatte 151] MUHLENBERGIA ALOPECUROIDES (POACEAE: MUHLENBERGIINAE), A NEW COMBINATION Paul M. Peterson and J.. TVGViS COMMIMUUS vvcccscsatssocssasehnsienvesscosasdiescascesees 159 DISTRIBUTION OF DWARF MISTLETOES (ARCEUTHOBIUM SPP., VISCACEAE) IN DURANGO, MEXICO Robert L. Mathiasen, M. Socorro Gonzalez Elizondo, Martha Gonzdlez Elizondo, Brian E. Howell, I. Lorena Lépez Enriquez, Jared'Se ott, and Soreen. Jeng PACES 35.35. eee ee ee 161 [EW SPECIES New RosA (ROSACEAE) IN CALIFORNIA AND OREGON ; Barbara Eriter and Walter He Lewis ..6.ccccd. chev cadesszceohottcdh eee ceveddesaceteseeseee 170 YOK REVIEWS CALIFORNIA NATIVE PLANTS FOR THE GARDEN BY CAROL BORNSTEIN, DAVID FROSS, AND BART O’ BRIEN WICIGNIC TR GCT- 1, CEL Viaccscdteehe cee tat ea oe reas oe haaie vi onan cata eee 169 PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL SOCIETY MADRONO (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 MADRONO, Kim Kersh, Membership Chair, Jepson Herbarium, University of California, Berkeley, CA 94720-2465. Editor—JOHN HUNTER Center for Plant Diversity Plant Sciences MS#7 One Shields Avenue University of California Davis, CA 95616-8780 madrono @lists.berkeley.edu Book Editor—JON E. KEELEY Noteworthy Collections Editors—DIETER WILKEN, MARGRIET WETHERWAX Board of Editors Class of: 2008—ELLEN DEAN, University of California, Davis, CA RoBERT E. PRESTON, Jones & Stokes, Sacramento, CA 2009—DoNovAN BAILEY, New Mexico State University, Las Cruces, NM MARK BorcHERT, USFS, Ojai, CA 2010—FRED Hrusa, California Department of Food and Agriculture, Sacramento, CA RICHARD OLMSTEAD, University of Washington, Seattle, WA 2011—JAMIE KNEITEL, California State University, Sacramento, CA KEVIN RIcE, University of California, Davis, CA CALIFORNIA BOTANICAL SOCIETY, INC. OFFICERS FOR 2007-2008 President: Dean Kelch, Jepson Herbarium, University of California, Berkeley, CA 94720, dkelch@sscl.berkeley. edu First Vice President: Andrew Doran, University and Jepson Herbaria, University of California, Berkeley, CA 94720, andrewdoran @ berkeley.edu Recording Secretary: Staci Markos, Friends of the Jepson Herbarium, University of California, Berkeley, CA 94720-2465, smarkos @ socrates.berkeley.edu Corresponding Secretary: Heather Driscoll, University Herbarium, University of California, Berkeley, CA 94720- 2465, heather.driscoll @nature.berkeley.edu Treasurer: Susan C. Dunlap, Aerulean Plant Identification Systems, Inc., Menlo Park, CA 94025, susancdunlap @ hotmail.com The Council of the California Botanical Society comprises the officers listed above plus the immediate Past President, Michael Vasey, Department of Biology, San Francisco State University, 1600 Holloway Avenue, San Francisco, CA 94132, mvasey @sfsu.edu; the Editor of Madrofio; three elected Council Members: James Shevock, National Park Service, Cooperative Ecosystems Studies Unit, 337 Mulford Hall, University of California, Berkeley 94720-3114, jshevock @nature.berkeley.edu; Roy Buck, Jepson Herbarium, University of California, Berkeley, CA 94720, roybuck @email.msn.com; Chelsea Specht, Department of Plant and Microbial Biology, University of California, Berkeley, CA 94720-2465, cdspecht @ nature.berkeley.edu. Graduate Student Representatives: Ben Carter, Department of Integrative Biology and University Herbarium, University of California, Berkeley, CA 94720-2465, bearter@berkeley.edu. Webmaster: Susan Bainbridge, Jepson Herbarium, University of California, Berkeley, CA 94720-2465, sjbainbridge @ berkeley.edu. © This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). MADRONO, Vol. 55, No. 2, pp. 93-112, 2008 CATALOGUE OF NONNATIVE VASCULAR PLANTS OCCURRING SPONTANEOUSLY IN CALIFORNIA BEYOND THOSE ADDRESSED IN THE JEPSON MANUAL ~— PART II ELLEN DEAN UC Davis Center for Plant Diversity, Department of Plant Sciences Mail Stop 7, One Shields Avenue, Davis, CA 95616 eadean@ucdavis.edu FRED HRUSA California Department of Food and Agriculture, Plant Pest Diagnostics Center, 3294 Meadowview Rd., Sacramento, CA 95832-1448 fhrusa@cdfa.ca.gov GORDON LEPPIG California Department of Fish and Game, Coastal Conservation Planning, 619 Second Street, Eureka, CA 95501 ANDREW SANDERS Department of Botany and Plant Sciences, University of California, Riverside, CA 92521-0124 BARBARA ERTTER University and Jepson Herbaria, University of California, Berkeley, CA 94720-2465 ABSTRACT We present Part II of a catalogue documenting nonnative vascular plant taxa occurring spontaneously in California beyond those addressed in The Jepson Manual: Higher Plants of California (Hickman 1993). Here we document an additional 117 taxa occurring spontaneously in California that were not accounted for in Part I (Hrusa et al. 2002) or in The Jepson Manual. The catalogue was compiled from new collections by the authors and others, previously existing herbarium specimens, peer-reviewed publications, other printed reports, and direct communications with field botanists. Only reports backed by herbarium vouchers are accepted as adequately documented. Of the 117 taxa, 42 are fully or sparingly naturalized in relatively undisturbed wildland habitats, 14 are naturalized in non-wildlands (roadsides, fallow fields, croplands, other disturbed areas), 13 are tenuously established or locally persisting, 22 are weeds of greenhouse or other horticultural environments, 7 are presumed to be non-persisting casuals (waifs), and for 19 there is no current information. Taxa recorded as already being widely naturalized and/or potentially significant pests include Brachypodium sylvaticum, Cuscuta japonica, Danthonia decumbens, Glyceria declinata, Juncus usitatus, Melaleuca viminalis, Rytidosperma penicillatum, Verbena incompta and Zostera japonica. Key Words: California, invasive plants, nonnative, pest plants, voucher specimens, weeds. By reducing the productivity of farm, pasture, and rangelands; choking waterways; altering fire regimes; and outcompeting or hybridizing with native plants or both; invasive nonnative plants have profoundly altered natural ecosystems and significantly impacted agriculture and commerce (Anttila et al. 1998; Brooks et al. 2004; Dudley 1998; Enserink 1999; Ayres and Strong 2003; Pimental et al. 2005; Vitousek et al. 1997). A recent estimate (Pimental et al. 2005) puts the economic cost of invasive plants in the United States at $35 billion per year (and that estimate does not take into account the cost of fighting invasive plants in most rangelands and natural areas). In California, nonnative plant invasions began over two centuries ago with the arrival of the Spanish in the 18" century. Indeed, with few exceptions, the best definition of ‘California native plant” is “‘a plant that 1s assumed to have been present prior to 1769, the date of the first Spanish mission” (Ornduff 1991; Randall et al. 1998). If a species was introduced to the state by human activities after that date, either intention- ally or unintentionally, it can be considered nonnative. The introduction of nonnative plants in California has had dramatic consequences. Dur- ing the latter half of this past century, we have seen changes in the floristic composition of nearly every low to mid-elevation California plant community due to invasive species (see the broad range of articles in Pirosko 2004). The floristic changes that occurred in California grasslands soon after the arrival of the Spanish are well 94 documented and are ongoing (Holstein 2001; Stromberg et al. 2001; Corbin et al. 2004). More recently, California’s northern desert regions are undergoing floristic and physiognomic modifica- tions due to the increasing dominance of Bromus tectorum, and in sandy parts of the low deserts Brassica tournefortii 1s a severe problem (D’An- tonio and Vitousek 1992; Kemp and Brooks 1998; E. A. Dean personal observation; A. C. Sanders personal communication). During the past decade, botanists have observed that inva- sive species such as Lepidium latifolium (perennial pepperweed) and Glyceria declinata (Eurasian waxy mannagrass) are now found in the deeper reaches of vernal pools in the Central Valley (McCarten 2005; Barkworth and Anderton 2007; Whipple et al. 2007). This means that even very specialized California plant communities are now threatened by invasive plants. Because of these ecological and economic impacts, the decisions of California land manag- ers and government officials that are related to invasive species have increasingly far-reaching ecological and economic consequences. Decisions regarding monitoring and control efforts are particularly important. Early detection and con- trol is paramount because eradication of new nonnative plant occurrences is usually only possible when populations are small (Bayer 1999; Wittenberg and Cock 2001; Rejmanek and Pitcairn 2002). However, informed decisions are difficult without up-to-date taxonomic and distributional information. Often the first reports of new species introduc- tions are made by taxonomists who are sent specimens for identification. In 2002, we pub- lished a catalogue of 315 nonnative vascular plant taxa that have been documented by herbaria as occurring spontaneously in California but were not included in The Jepson Manual: Higher Plants of California (Hrusa et al. 2002). Since that publication we have continued to compile new instances of nonnative vascular plants documented in California. We present Part II of this catalogue in the interest of providing the California botanical community, land managers and government officials, with more taxonomic and distributional information on the nonnative plants of California. This information is especial- ly timely, given that a revision of the Jepson Manual: Higher Plants of California is due out in the next few years. METHODS Data sources were as described in Hrusa et al. (2002), but with a smaller number of taxa derived from older published sources and “‘lost’”? herbar- ium specimens, although recent publications were an important source. The primary source of new taxa was submissions of plant samples to MADRONO [Vol. 55 TABLE |. CURRENT STATUS (CS) SUMMARY. Expla- nation and definition of individual categories provided in the text and Hrusa et al. (2002). Catalogue Definition Abbrev. Total Naturalized in wildlands NW 42 Naturalized outside of wildlands N 14 Tenuous/locally persisting TEN 13 Casual Cc 7 Greenhouse, nursery, garden weed GH/C 22 No current information NCI 19 herbaria for identification; almost as important | were specialist identifications of specimens sent for study by authors contributing to the Flora of North America project (Flora of North America Editorial Committee 1993+) and the second edition of the Jepson Manual project (Jepson | Herbarium 2007). Indeed, many taxa (specimens) not included here are awaiting determinations and remain with specialists; these will be reported later. The Consortium of California Herbaria | (2007) online searchable database of specimens © held in most major California herbaria was an | important source of “‘lost’’ herbarium specimens, | duplicate specimens held in different herbaria, | and especially, additional distributions. However, | the criteria used for inclusion of taxa follows | those outlined in Hrusa et al. (2002) in which a » reported taxon had to be verified as to identity by | one of the authors, or by a specialist. Some > Consortium records represent specimens not yet confirmed as to identity and therefore the specimen details for these are not included in Appendix 2. However, if additional records from | the Consortium added County or physio-geo- | graphic regions to the species distribution, or if. additional collections exist for a documented region, these are indicated in the NOTES following each taxon report. The Consortium | should be consulted by anyone interested in additional records for any given taxon on this. list, or on the list in Part I (Hrusa et al. 2002). Moreover, the database is constantly being. updated and new records not currently available . will become so in the near future. Naturalization | categories have not been modified in order to— allow both comparison and combination of these two datasets. They are: Naturalized in wildlands (NW); naturalized outside of wildlands (N);. persistence tenuous (TEN); casual (waif) (C); greenhouse/cultivation (GH/C) and no current. information (NCI). The extirpated (EXT) cate-— gory in Part I, had no members in Part II. RESULTS A total of 117 non-native vascular plant taxa (Table 1) meeting the criteria for inclusion are listed in Appendix 1 and itemized in detail in| 2008] Appendix 2. Both lists are organized as in Hrusa et al. (2002) and The Jepson Manual (Hickman 1993). Introduced taxa in Part II are documented in 47 of California’s 58 counties and in every biogeographic region in Hickman (1993). DISCUSSION The taxa cataloged here, like those document- ed in Part I (Hrusa et al. 2002), represent a diverse assemblage of recent and not so recent documented introductions with wide variation in distributions, current or potential invasiveness, and ecological and economic threat. Zea mays for instance, despite its wide cultivation and high potential to occur as a widespread waif, poses virtually no economic or ecological threat be- cause it is barely capable of reproduction in the wild. Twenty-two taxa (15%) are entirely green- house or garden waifs with, as yet, little indication of persistence or spread outside of cultivated settings (Table 1). Though many taxa are documented from only one occurrence, other taxa (e.g., Glyceria declinata, Rytidosperma peni- cillatum, Verbena incompta) are widespread and documented from numerous occurrences. Fifty-six taxa (48%) are documented here as fully naturalized, with 42 (36%) naturalized in wildlands. The taxa included here as “‘naturalized in wildlands” occur in a wide variety of habitat types. Terrestrial habitats with these plants include: mixed chaparral-oak woodland (Grevil- lea robusta), redwood forest (Brachypodium sylvaticum), and pine forest (Coronilla_ varia, Gnaphalium coarctatum). Wetland habitat exam- ples include: tidal mud flats (Zostera japonica), tidal salt marshes (Juncus gerardii), coastal peat- lands (Danthonia decumbens), rivers and ponds (Pontederia cordata), riparian scrub (Melaleuca viminalis), and various freshwater wetlands in- cluding vernal pools (Glyceria declinata). These already naturalized taxa pose the greatest eco- logical and economic threats, though the poten- tial for the eventual spread of casual or horticul- tural weeds cannot be underestimated. In some cases, introduced plant species appear to be harmless ornamentals or remain as_ localized introductions for years, and then, for various reasons, such as the introduction of a pollinator (or for entirely inexplicable reasons) they become aggressive invaders (Schierenbeck et al. 1998; Enserink 1999; Rejmanek 2000). We have no current information on the persistence or spread of 19 (16%) of the taxa documented here. Habitat Invasions We highlight the ecological importance of the taxa documented here by elaborating on the threat posed by three taxa in three different wetland habitats. DEAN ET AL.: CATALOGUE OF NON-NATIVE VASCULAR PLANTS 3) Glyceria declinata 1s a Eurasian grass that commonly occurs in the Central Valley and North Coast region and is known to invade vernal pools, one of California’s most threatened habitats (Hrusa and Leppig (personal observa- tion and vouchers at CDA, HSC); Barkworth and Anderton 2007). California’s vernal pools are habitat to one of California’s largest assemblage of state and federally listed taxa (California Native Plant Society 2007; California Natural Diversity Database 2007); through competition and habitat alteration, G. declinata could pose a direct threat to these species. This species already has a wide distribution in 15 counties and has been established in California for decades. It was included in Munz and Keck (1968) but omitted from Hickman (1993); as a result, specimens of this species collected since 1993 were usually identified as the native Glyceria occidentalis (Piper) J.C. Nelson. This has been a source of confusion for vernal pool investigators and has prevented California botanists from understand- ing the degree to which this species has invaded vernal pool habitat. This case highlights the importance of correct species determination and the inclusion of nonnatives in regional floras. Danthonia decumbens, another Eurasian grass, also poses a significant ecological threat. Popu- lations of this species are documented at the Crescent City Marsh, Del Norte County, and Big Lagoon Bog, Humboldt County. Both sites are unusual coastal peatlands and rare plant hot- spots, each with at least five California Native Plant Society-listed plant populations (California Natural Diversity Database 2007; Leppig 2004). Crescent City Marsh also has the largest Cali- fornia population of the State and-federal endan- gered western lily (Lilium occidentale) (U.S. Fish and Wildlife Service 1998). Danthonia decumbens is now regularly found in coastal grasslands in Del Norte and Humboldt counties (Leppig, personal observation) and it likely occurs in, or will soon spread to other regions of the state. Zostera japonica, a rhizomatous, mat-forming seagrass, was likely introduced to the Pacific Northwest from Japan through the oyster indus- try (Harrison 1976; Harrison and Bigley 1982) and occurs in Oregon, Washington, and southern British Columbia estuaries. This species colonizes subtidal estuarine sediments and generally occurs at a higher tidal elevation than the native Z. marina. One small population of Z. japonica was discovered in 2002, in Humboldt Bay, Califor- nia’s second largest estuary. Despite ongoing eradication efforts since then (Schlosser et al. 2005), another population, located approximately eight kilometers from the first occurrence, was detected in 2006. Thus the ability to successfully eradicate this species remains uncertain. Zostera colonies slow water-currents and entrap fine particles while their rhizome-root system binds 96 and stabilizes substrates (Phillips 1984). Over time, they affect the mean grain size, sorting, skewness, shape of sediment particles, and parameters that influence the redox potential and nutrient cycling processes of the substrate in which they occur (Phillips 1984). Consequently, if this species were to become dominant in large areas of its potential habitat, it could affect not only Humboldt Bay’s important fisheries and wildlife values (Barnhart et al. 1992), but also its physical processes, including large-scale sediment distribution patterns, channel morphology, and the Bay’s overall bathymetry and _ shoreline erosion patterns. Zostera asiatica poses a similar risk where it becomes established. Hybridization Threat The introduced taxa documented here also highlight the concern over invasive genotypes mixing with native taxa. While hybridization 1s, of course, an important phenomenon in the evolutionary biology of plants, a diversity of introduced taxa together with widespread habitat alterations, increases the potential for at least two genetic impacts within native floras: |) extinction through genetic assimilation and outbreeding, and 2) an increase in invasiveness and expansion of the ecological amplitude of formerly non- invasive or narrowly distributed taxa (Cox 2004). The Wolfs evening primrose (Oenothera wol- fii), for example, is listed as threatened by the state of Oregon, and both the California Native Plant Society and the Oregon Natural Heritage Program list this species as endangered through- out its range (Imper 1997). An endemic to California’s Northcoast and coastal Oregon, it is threatened by habitat loss as well as by extinction through hybridization with the escaped garden plant Oenothera glazioviana (DeWoody et al. in press). Even regionally common native taxa, such as Spartina foliosa and Lupinus littoralis face local extinction due to hybridization with introduced taxa (Anttila et al. 1998; Ayres and Strong 2003; Wear 2004). It is worth noting that 32 of the genera (43 of the species) listed in our Appendix 2 are congeners of native Califor- nia species. Horticultural Escapes Worldwide, escaped horticultural plants are the principal cause of invasive plant introductions (Rozefelds et al. 1999; Mack and Lonsdale 2001; Reichard and White 2001; Mack and Erneberg 2002). While the vast majority of horticultural taxa serve their desired purpose without escaping or becoming a naturalized nuisance, many clearly do not (Bell et al. 2007). In most cases, the mode of introduction of the species listed here cannot be definitively determined and with some species, MADRONO [Vol. 55 more than one pathway may have led to their introduction. That notwithstanding, 49 of the species documented here are either available in the horticultural trade, documented as spreading away from homesites where they were intention- ally cultivated, or known to be cultivated by Californians for food or medicine. An additional eight species were first documented in California as contaminants in nurseries. Clear horticultural escapes that are now naturalized in wildlands include Cordyline australis, Cupaniopsis anacar- dioides, Grevillea robusta, Rhamnus alaternus, and Solanum mauritianum. CONCLUSION New invasive plant species continue to be detected at a surprisingly high rate through discovery of previously established populations as well as fresh invasions. As discussed in Part I (Hrusa et al. 2002), we reiterate the critical need for documentation of unrecorded invasions. We call on all field workers, including restoration biologists, biological consultants, land managers, naturalists and others, many of whom have an intimate knowledge of a particular region of | California, to take the simple steps necessary to document plants that look “‘out of place.” It is likely that most readers of this paper have had an introduction to the use of a plant press at some point in their careers, and we would deeply appreciate your applying that knowledge to document plants that you believe to be new to an area. Any unknown plant can be pressed and dried and sent to one of the authors of this paper. The most efficient action to take is to collect the plant when you first see it. Even if the material is | not optimal, collect it, put it in a plastic bag, and . when possible, press the best sheet you can in a fold of paper. More complete material can be acquired later. Flattened plants can be placed | between two sturdy cardboards secured with tape, placed in a padded envelope, and mailed to one of us. Speed is of the essence; the sooner material is delivered, identified and control action taken, if necessary. the sooner it can be | Doubtless, there are many more invasive species — waiting to be detected in the California flora — and others will be arriving shortly. But, verbal reports are of limited value and digital photo- | graphs are only slightly better. Pressing plant | specimens is easy, and you may even find it motivating to know that your specimen, once accessioned into a herbarium, will have a label | with your name on it and will last for hundreds of © years. ACKNOWLEDGMENTS We thank: the collectors of the specimens included in this article for carefully documenting their floristic studies with herbarium specimens; the BRIT, CAS, 2008] CDA, CHSC, DAV/AHUC, HSC, MO, RSA, SD, UCR, and UC/JEPS herbaria for use of their specimens and/or staff assistance; I. Al-Shebaz, M. Barkworth, L. Bohs, H. Conner, D. Correll, S. Darbyshire, L. Gross, D. Kelch, P. Michael, A. Murdock, M. Nee, G. Nesom, C. Reeder, S. G. Smith, R. Spellenberg, J. Strother, A. Tucker, G. Tucker, S. White, K. Wilson, P. Zika for taking the time to examine and identify specimens cited here; and Kai Neander and Tony LaBanca for field assistance. We also thank John Hunter, Marcel Rejmanek, and an anonymous reviewer for editorial assistance that much improved the contents and flow of this article. LITERATURE CITED ANDERSON, E. N. 2002. Some preliminary observations on the California black walnut (Juglans califor- nica). Fremontia 30:12—19. ANTTILA, C. K., C. C. DAEHLER, N. E. RANK, AND D. R. STRONG. 1998. Greater male fitness of a rare invader (Spartina alterniflora), Poaceae, threatens a common native (Spartina foliosa) with hybridiza- tion. American Journal of Botany 85:1597—-1601. AYRES, D. R. AND D. R. STRONG. 2003. Spartina foliosa (Poaceae) — a common species on the road to rarity? Madrono 50:209-213. BARKWORTH, M. E. AND L. K. ANDERTON. 2007. Glyceria R. Br. Pp. 68-88 in M. E. Barkworth, K. M. Capels, S. Long, L. K. Anderton, and M. B. Piep (eds.), Flora of North America north of Mexico, Vol. 24. Oxford University Press, New York, NY. BARNHART, R. A., M. J. BOYD, AND J. E. PEQUEGNAT. 1992. The ecology of Humboldt Bay, California: an estuarine profile. U. S. Fish and Wildlife Service, Biological Report 1. BAYER, D. 1999. Biological, economic and public policy aspects. Pp. 17-22 in R. H. Coppick and M. Krieth (eds.), Symposium proceedings; exotic pests and diseases, biology, economics and public policy. Agricultural Issues Center, University of California, Davis, CA. BEAUCHAMP, R. M. 1986. A flora of San Diego County, California. Sweetwater River Press, Na- tional City, CA. BELL, C. E., J. M. DITOMASO, AND C. WILEN. 2007. Invasive plants. Pest Notes, University of Califor- nia Agriculture and Natural Resources publication 74139. University of California Statewide IPM Program, University of California, Davis, CA. BoyD, S. 1999. Flora of the Liebre Mountains, Western Transverse Ranges, California. Aliso 18:93—139. . 2004. Report of a new introduced species and revised key to the annual Crassula (Crassulaceae) of California. Madrono 51:393-394. BROMILOw, C. 1995. Problem Plants of South Africa. Briza Publications, Arcadia, South Africa. BROOKs, M. L., C. D’ANTONIO, D. M. RICHARSON, J. B. GRACE, AND J. E. KEELEY. 2004. Effects of invasive alien plants on fire regimes. BioScience 54:677-688. CALIFORNIA NATIVE PLANT SOCIETY. 2007. Inventory of Rare and Endangered Plants (online edition). California Native Plant Society. Sacramento, CA. (http://www.cnps.org/inventory). CALIFORNIA NATURAL DIVERSITY DATABASE. 2007. Special plants list. California Department of Fish and Game Natural Diversity Database, July 2007. DEAN ET AL.: CATALOGUE OF NON-NATIVE VASCULAR PLANTS 7 CONSORTIUM OF CALIFORNIA HERBARIA. 2007. Avail- able at: http://ucjeps.berkeley.edu/consortium/. CORBIN, J., M. THOMSEN, J. ALEXANDER, AND C. D’ANTONIO. 2004. Out of the frying pan: invasion of exotic perennial grasses in coastal prairies. Pp. 27-28 in C. Pirosko (ed.), Proceedings of the California Invasive Plant Council symposium Vol. 8. Cox, G. W. 2004. Alien species and evolution, the evolutionary ecology of exotic plants, animals, microbes, and interacting native species. Island Press, Washington, DC. D’ANTONIO, C. AND P. VITOUSEK. 1992. Biological invasions by exotic grasses, the grass/fire cycle, and global change. Annual Review of Ecology and Systematics 23:63—87. DANDY, J. E. AND D. P. YOUNG. 1959. Oxalis megalorrhiza. Proceedings of the Botanical Society of the British Isles 3:174-175. DARBYSHIRE, S. J. AND H. E. CONNOR. 2003. Rytidosperma Steud. Pp. 309-312 in M. E. Bark- worth, K. M. Capels, S. Long, L. K. Anderton, and M. B. Piep (eds.), Flora of North America north of Mexico, Vol. 25. Oxford University Press, New York, NY. DEWoopy, J., L. ARGUELLO, D. IMPER, R. WEST- FALL, AND V. D. HIPKINS. In Press. Genetic evidence of hybridization between Oenothera wolfii (Wolfs evening primrose) and O. glazioviana, a garden escape. Madrono, 55. (in press). DUDLEY, T. 1998. Exotic plant invasions in California riparian areas and wetlands. Fremontia 26:24—29. ENSERINK, M. 1999. Biological invaders sweep in. Science 285:1834-1836. EVENHUIS, N. L. AND L. G. ELDREDGE. 2004. Records of the Hawaii Biological Survey for 2003. Part II. Notes. Occasional Paper 79. Bishop Museum, Honolulu. FLORA OF NORTH AMERICA EDITORIAL COMMITTEE, (eds). 1993+. Flora of North America north of Mexico. 12+ vols. New York and Oxford. Avail- able at: http://hua.huh.harvard.edu/FNA/. HARRISON, P. G. 1976. Zostera japonica Ascher. and Graebn. in British Columbia, Canada. Syesis 9:359—360. AND R. E. BIGLEY. 1982. The recent introduc- tion of the seagrass Zostera japonica Ascher. and Graebn. to the Pacific coast of North America. Canadian Journal of Fisheries and Aquatic Scienc- es 39:1642-1648. HICKMAN, J. C. (ed.). 1993. The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. HOLSTEIN, G. 2001. Pre-agricultural grassland in central California. Madrono 48:253—264. HRUSA, F., B. ERTTER, A. SANDERS, G. LEPPIG, AND E. DEAN. 2002. Catalogue of non-native vascular plants occurring spontaneously in California be- yond those addressed in the Jepson Manual — Part I. Madrono 49:61—98. HYLAND, B. P. M. 1983. A revision of Syzygium and allied genera (Myrtaceae) in Australia. Australian Journal of Botany, Supplementary Series No. 9: 1-164. IMPER, D. K. 1997. Ecology and conservation of Wolf's evening primrose in northwestern California. Pp. 34-40 in T. N. Kaye, A. Liston, R. M. Love, D. L. Luoma, R. J. Meinke, and M. V. Wilson 98 (eds.), Conservation and Management of Native Plants and Fungi. Native Plant Society of Oregon, Corvallis, OR. JANEWAY, L. P. 1992. The Cyperaceae of Butte County. Part 1: Carex. Studies from the Herbarium, California State University, Chico, No. 9. . 2005. Noteworthy Collections. Madrono a2e12 5: JEPSON HERBARIUM. 2007. Second Edition of the Jepson Manual: Vascular Plants of California. Available at: http://ucjeps.berkeley.edu/jepsonmanual/ review/. Last updated 22 August 2007. Accessed September 6, 2007. Kemp, P. AND M. BROOKS. 1998. Exotic species of California deserts. Fremontia 26:30—34. LEPPIG, G. 2004. Rare plants of northern California coastal peatlands: patterns of endemism and phytogeography. Pp. 43—50 in M. B. Brooks, S. K. Carothers, and T. LaBanca (eds.), The ecology and management of rare plants of northwestern Cali- fornia: proceedings from a 2002 symposium of the North Coast Chapter of the California Native Plant Society. California Native Plant Society, Sacramento, CA. Mack, R. N. AND M. ERNEBERG. 2002. The United States naturalized flora: largely the product of deliberate introductions. Annals of the Missouri Botanical Garden 89:176—189. AND W. M. LONSDALE. 2001. Humans as global plant dispersers: getting more than we bargained for. BioScience 51:95—102. 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, CA. MCCARTEN, N. 2005. Final Conservation and Man- agement Plan for Yolo County Grasslands Park. CALFED at-risk species, habitat restoration and recovery, and non-native species management ERP-02-P46. CALFED Ecosystem Restoration Program. Sacramento, CA. MICHAEL, P. W. 1995. A new name for a widespread and misunderstood species of Verbena (Verbena- ceae). Telopea 62:181—183. Munz, P. A. AND D. D. KECK. 1968. A California flora and supplement. University of California Press, Berkeley, CA. NESOM, G. L. 2002. Euchiton (Asteraceae: Gnapha- lieae) in North America and Hawaii. Sida 20:515—521. . 2004. New Distribution records for Gamo- chaeta (Asteraceae: Gnaphalieae) in the United States. Sida 21:1175—1185. ORNDUFF, R. 1991. Introductory lecture for Integrative Biology 102, California Floristics, given at Univer- sity of California Berkeley on January 23, 1991. PHILLIPS, R. C. 1984. The ecology of eelgrass meadows in the Pacific Northwest: a community profile. U.S. Fish and Wildlife Service. FWS/OBS-84/24. AND S. WYLLIE-ECHEVERRIA. 1990. Zostera asiatica Miki on the Pacific coast of North America. Pacific Science 44:130—134. PIGNATTI, S. 1972. Limonium. Pp. 38-50 in T. G. Tutin, et al. (eds.). Flora Europaea, Vol. 3. Cambridge at the University Press, Cambridge. PIMENTEL, D., R. ZUNIGA, AND D. MORRISON. 2005. Update on the environmental and economic costs MADRONO [Vol. 55 associated with alien-invasive species in the United States. Ecological Economics 52:273—288. PIROSKO, C. (ed.). 2004. Proceeding of the California Invasive Plant Council symposium Vol. 8: 2004. The California Invasive Plant Council, Berkeley, CA RANDALL, J.. M. REJMANEK, AND J. HUNTER. 1998. Characteristics of the exotic flora of California. Fremontia 26:3—12. REBMAN, J. P. AND M. G. SIMPSON. 2006. Checklist of the Vascular Plants of San Diego County. 4th edition. San Diego Natural History Museum and San Diego State University, San Diego, CA. REICHARD, S. H. AND P. WHITE. 2001. Horticulture as a pathway of invasive plant introductions in the United States. BioScience 51:103—113. REIFNER, R. E., JR. AND G. F. HRUSA. Noteworthy Collections: California. Madrofo 55 (in press). , J. T. COLUMBUS, AND: S. BOYD. ' 2003; Noteworthy Collections. Madrono 50:312—313. REJMANEK, M. 2000. Invasive plants: approaches and predictions. Austral Ecology 25:497—506. exotic pest plants a realistic goal? Pp. 249-253 in C. R. Veitch and M. N. Clout (eds.), Turning the tide: the eradication of invasive species. IUCN, Gland, Switzerland and Cambridge, UK. RICHARDSON, D. M. AND M. REJMANEK. 1999. Metrosideros excelsa takes off in the fynbos. Veld & Flora 85:107—111. ROBERTS, F., S. D. WHITE, A. C. SANDERS, D. E. BRAMLET, AND S. BoypbD. 2004. The Vascular Plants of Western Riverside County, California, An Annotated Checklist. F.M. Roberts Publs., San Luis Rey, CA. ROZEFELDS, A. C. F., L. CAVE, D. I. MORRIS, AND | 1999. The weed invasion in | Tasmania since 1970. Australian Journal of Botany © A. M. BUCHANAN. 47:23-48. SANDERS, A. C. 1996. Noteworthy collections. Madro- no 43:524-532. SCHIERENBECK, K., K. GALLAGHER, AND J. HOLT. 1998. The genetics and demography of invasive plant species. Fremontia 26:19—23. SCHLOSSER, S., N. KALSON, J. MELLO, V. FREY, AND | J. ROBINSON. 2005. Detection and attempted eradication of the non-indigenous eelgrass, Zostera japonica, from Humboldt Bay, California, USA AND M. PITCAIRN. 2002. When is eradication of | and restoration of Zostera marina. Abstract. Pacific | Estuarine Research Society Annual Meeting, | March 2005, Charleston Oregon, USA. SMITH, J. P., JR. AND J. O. SAWYER. 1986. A checklist | of the vascular plants of northwestern California. | Humboldt State University Herbarium Miscella- | neous Publication #2, 8th edition. Humboldt State © University Herbarium, Arcata, CA. STROMBERG, M., P. KEPHART, AND V. YADON. 2001. Composition, invasibility, and diversity in coastal California grasslands. Madrono 48:236—252. ULMER, T. AND J. M. MACDOUGAL. 2004. Passiflora. — Timber Press, Portland, OR. UNITED STATES FISH AND WILDLIFE SERVICE. 1998. Recovery plan for the endangered western lily | (Lilium occidentale). Portland, OR. VITOUSEK, P., C. D’ANTONIO, L. LOOPE, M. REJMA- NEK, AND R. WESTBROOKS. 1997. species: a significant component of human-caused Introduced | 2008] global change. New Zealand Journal of Ecology 21:1-16. WEAR, K. S. 2004. Natural hybridization between native and introduced Lupinus on the north coast of California. Pp. 23-30 in M. B. Brooks, S. K. Carothers, and T. LaBanca (eds.), Proceedings of the 2002 symposium of the North Coast Chapter of the California Native Plant Society. California Native Plant Society, Sacramento, CA. WHIPPLE, I. G., M. E. BARKWORTH, AND B. S. BUSHMAN. 2007. Molecular insights into the taxonomy of Glyceria (Poaceae: Meliceae) in North America. American Journal of Botany 94:551—557. WITTENBERG, R. and M. J. W. Cock (eds.). 2001. Invasive alien species: a toolkit of best management practices. CAB International, Wallingford, Oxon, UK. ZIKA, P. 2005. Noteworthy Collections: California. Madrono 52:208. ZULOAGA, F., J. PENSIERO, AND O. MORRONE. 2004. Systematics of Paspalum Group Notata (Poaceae- Panicoideae-Paniceae). Systematic Botany Mono- graphs 71:1—75. APPENDIX | LIST OF TAXA WITH NATURALIZATION (CURRENT STATUS) CLASS As in Part I (Hrusa et al. 2002), family circumscrip- tions and organization follow The Jepson Manual (Hickman 1993). Current Status subcategories in brack- ets: NW = naturalized in wildlands, N = naturalized (outside of wildlands), TEN = persistence tenuous, C = casual (waif), GH/C = greenhouse/cultivation, NCI = no current information. See Table | and Appendix 2. Details and descriptions of these categories are in the Methods section and Hrusa et al. (2002). FERNS AND ALLIES Psilotaceae Psilotum nudum (L.) P. Beauv., [GH/C]. ANGIOSPERMS: DICOTS Asteraceae Ageratum conyzoides L., [GH/C]. Centipeda minima (L.) A. Braun & Asch., [GH/C]. Crassocephalum crepidioides (Benth.) S. Moore, [GH/C]. Crepis pulchra L., [C]. Gnaphalium antillanum Urb., [GH/C]. Gnaphalium calviceps Fern., [N]. Gnaphalium coarctatum Willd., [NW]. Gnaphalium involucratum G. Forst., [NCI]. Gnaphalium pensylvanicum Willd., [N]. Gnaphalium stachydifolium Lam., [NW]. Senecio cineraria DC., [TEN]. Youngia japonica (L.) DC., [GH/C]. Balsaminaceae Impatiens capensis Meerb., [TEN]. Impatiens glandulifera Royle, [NCI]. Brassicaceae Berteroa incana (L.) DC., [NW]. DEAN ET AL.: CATALOGUE OF NON-NATIVE VASCULAR PLANTS 99 Carrichtera annua (L.) DC., [NW]. Matthiola longipetala (Vent.) DC. ssp. bicornis (Sibth & Sm.) P. W. Ball., [NCI]. Teesdalia nudicaulis (L.) R. Br., [TEN]. Campanulaceae Laurentia fluviatalis (R. Br.) E. Wimm., [GH/C]. Caprifoliaceae Viburnum rigidum Vent., [NW]. Caryophyllaceae Cerastium tomentosum L., [C]. Silene armeria L., [C]. Silene coeli-rosa (L.) Godron, [NCI]. Chenopodiaceae Atriplex glauca L., [N]. Atriplex vesicaria Heward in Hook. f., [NCI] Beta macrocarpa Guss., [NW]. Crassulaceae Crassula colligata Toelken ssp. lamprosperma Toelken, [NW]. Cucurbitaceae Ecballium elaterium (L.) A. Rich., [GH/C]. Cuscutaceae Cuscuta japonica Choisy, [N]. Euphorbiaceae Euphorbia pulcherrimia Willd. ex Klotzsch, [NCI]. Phyllanthus caroliniensis Walt., [GH/C]. Fabaceae Coronilla varia L., [NW]. Crotalaria capensis Jacq., [GH/C]. Cytisus battandieri Maire, [NCI]. Dorycnium hirsutum L., [NCI]. Lathyrus nissolia L., [NCI]. Medicago muricata All., [NW]. Medicago scutellata (L.) P. Mill., [NW]. Sophora secundiflora Lag., [NCI]. Hamamelidaceae Hamamelis sp., [NW]. Juglandaceae Carya illinoiensis K. Koch, [NW]. Lamiaceae Leonurus sibiricus L., [GH/C]. Malvaceae Urena sinuata L., [GH/C]. Melianthaceae Melianthus major L., [NW]. Moraceae Ficus rubiginosa Desf. ex Vent., [TEN]. Myrtaceae Eucalyptus pulchella Desf., [NW]. Melaleuca citrina (Curtis) Dum.Cours., [NW]. Melaleuca viminalis (Sol. ex Gaertn.) Byrnes, [NW]. Metrosideros kermadecensis W.R.B. Oliv., [C]. 100 MADRONO Myrtus communis L., [NW]. Syzygium australe (Wendl. ex Link ) B.Hyland, [NW]. Nyctaginaceae Boerhavia diffusa L., [NCI]. Nymphaeaceae Nelumbo lutea Willd., [N]. Oxalidaceae Oxalis carnosa Molina, [GH/C]. Oxalis corymbosa DC., [GH/C]. Passifloraceae Passiflora ‘Coral Seas’, [NW]. Plumbaginaceae Limonium binervosum (G.E. Sm.) Salmon, [NW]. Limonium gougetianum Kuntze, [TEN]. Polygonaceae Polygonum aubertii Henry, [N]. Portulacaceae Portulacaria afra (L.) Jacq., [GH/C]. Proteaceae Grevillea robusta Cunn., [NW]. Ranunculaceae Ranunculus trilobus Desf., [N]. Rhamnaceae Rhamnus alaternus L., [NW]. Rosaceae Aphanes arvensis L., [GH/C]. Cotoneaster simonsii Baker, [NW]. Rosa sempervirens L., [TEN]. Rubiaceae Diodia virginiana L., [N]. Sapindaceae Cardiospermum halicacabum L., [TEN]. Cupaniopsis anacardioides (A. Rich.) Radlk., [NW]. Saxifragaceae Francoa ramosa D. Don, [TEN]. Scrophulariaceae Veronica biloba L., [NW]. Solanaceae Nierembergia frutescens Durieu., [TEN]. Nierembergia hippomanica Miers., [TEN]. Solanum cardiophyllum Lindl., [TEN]. Solanum chrysotrichum Schltdl., [NCI]. Solanum jasminoides Paxt., [NCI]. Solanum mauritianum Scop., [NW]. Verbenaceae Verbena incompta Michael, [NW]. Violaceae Viola conspersa Reichenb., [N]. [Vol. 55 ANGIOSPERMS: MONOCOTS | Arecaceae [Palmae] Washingtonia robusta H. Wendl., [NW]. Bromeliaceae Billbergia nutans H. Wend1., [C]. Commelinaceae | Callisia repens (Jacq.) L., [GH/C]. Cyperaceae Carex pendula Huds., [NW]. Cyperus gracilis R. Br., [NCI]. Cyperus regiomontanus Britt., [GH/C]. Cyperus retrorsus Chapm., [GH/C]. Eleocharis lanceolata Fernald, [NCI]. Iridaceae Sisyrinchium pruinosum Bickn., [GH/C]. Juncaceae Juncus gerardii Loisel., [NW]. Juncus usitatis L.A.S. Johnson, [NW]. Liliaceae (sensu lato) Allium ampeloprasum L., [NCI]. Aloe striatula Haw., [NW]. Asparagus densiflorus (Kunth) Jessop, [NW]. Asparagus setaceous (Kunth) Jessop, [NCI]. Bulbine semibarbata (R. Br.) Haw., [N]. Cordyline australis (G. Forst.) Endl., [NW]. Hyacinthoides hispanica (Mill.) Rothm., [C]. Narcissus papyraceus Ker Gawl., [GH/C]. Muscari armeniacum Leicht. ex Baker, [NCI]. Poaceae Brachypodium sylvaticum (Huds.) Beauv., [NW]. Danthonia decumbens (L.) DC., [NW]. Dinebra retroflexa (Vahl) Panz var. retroflexa, [NW]. Eragrostis tenella (L.) Roem. & Schultes, [GH/C]. Glyceria declinata Bresbiss., [NW]. Paspalum notatum Flugge var. saurae Parodi, [NCI]. Rytidosperma caespitosum (Gaudich.) Connor & Edgar, | [NW]. Rytidosperma penicillatum (Labill.) Connor & Edgar, | [NW]. Rytidosperma racemosum (R. Br.) Connor & Edgar, © [N]. Rytidosperma richardsonii (Cashmore) Connor & Ed- | gar, [N]. Sporobolus creber De Nardi, [N]. Stipa papposa Nees, [C]. Zea mays L., [C]. Pontederiaceae Pontederia cordata L., [N]. Zingiberaceae Hedychium flavescens N. Carey ex Roscoe, [TEN]. Zosteraceae Zostera asiatica Miki, [NW]. Zostera japonica Ascher & Graebner, [NW]. 2008] APPENDIX 2 ANNOTATED CATALOGUE As in Part I (Hrusa et al. 2002), family circumscrip- tions and organization follow The Jepson Manual (Hickman et al. 1993). Generic and specific applications reflect published treatments by specialists, modified only if clearer information was gained by utilizing an alternative nomenclature. Abbreviations are as follows: DISTRIBUTION (DIST) with geographic subdivisions as used in the Jepson Manual; CURRENT STATUS (CS); Current Status subcategories: NATURALIZED IN WILDLANDS (NW); NATURALIZED OUT- SIDE OF WILDLANDS (N); PERSISTENCE TEN- UOUS (TEN); CASUAL (waif) (C); GREENHOUSE/ CULTIVATION (GH/C); NO CURRENT INFOR- MATION (NCI); DOCUMENTATION (DOC). De- tails and descriptions of these categories are provided in the Methods section and Hrusa et al. (2002). Specimens cited only from the Consortium of California Herbaria (Consortium 2007) have not been verified as to identity. FERNS AND ALLIES Psilotaceae Psilotum nudum (L.) P. Beauv.: DIST: SCo: CS: GH/ C: DOC: Los Angeles Co.: Los Angeles, Hollywood area nr. the int. of Gower St. & Melrose Ave., Paramount Studios lot, elev. 111 m/365 ft, 34°05'01"”N, 118°19'17”"W. Weed in landscape bed under a star-jasmine shrub, previously seen on lot in potted Podocarpus plants and in shady sites on north side of buildings, a persistent weed that has been present for over 3 yr, Nov. 26, 2002, H. Williamson s.n. (UCR); Riverside Co.: Riverside, E side of town at foot of Box Springs Mtns., at 266 Frost Ct., cultivated ground in 1/ 4 acre home citrus orchard, shaded W-facing slope, fairly common perennial at base of most trees, near irrigation emitters; present for several years as a weed. Elev. 396 m/1300 ft, 33°58’N, 117°19’W, Jan. 9, 1999, A. C. Sanders 22375 (UCR); Riverside, Riverside Comm. College campus near library, weed under shrubs in planter, 33°58'16’"N, 117°22’49"W, Feb., 1999, A. C. Sanders & M. Provance 22450 (UCR); Riverside: UCR campus, along East Campus Dr beside greenhouses at intersection with Eucalyptus Dr, urbanized area in city, uncommon rhizomatous perennial in deep leaf litter under a hedge. Elev. 335 m/1100 ft 33°58’1S’N, 117°19’15’"W, July 12, 2000, A. C. Sanders & J. M. DiTomaso 23587 (UCR); San Diego Co.: Growing in ornamental planting at San Diego Zoo. 32°42'53’N; 117°09'26"W, April 21, 2006, C. Bell PDR 1317790 (CDA); Carlsbad, Legoland, in landscape, gen. nr. sprinklers and leaks. 33°07'24’N; 117°18’'46’W, elev. 30 m/100 ft, Oct. 12, 2007, K. Silrum PDR 1317236 (CDA); Santa Barbara Co.: Growing in ornamental planting at 105 E Anapamu St. Santa Barbara, 34°25'29"N, 119°42'12”W, April 4, 2006, M. Guy PDR 1269752 (CDA): NOTES: Reported also in Orange Co. (Consortium 2007). Apparently reproduc- ing in orchards and flower beds. Seems always to be found in deep leaf litter under shade of shrubs and trees, where it is often associated with drip irrigation emitters _ or other sources of moisture. DEAN ET AL.: CATALOGUE OF NON-NATIVE VASCULAR PLANTS 101 ANGIOSPERMS—DICOTS Asteraceae Ageratum conyzoides L.: DIST: SCo: CS: GH/C: DOC: Santa Barbara Co.: Weed in cycad at Island View nursery, Carpenteria. April 22, 2002, S. Bryants s.n. (CDA): NOTES: First documented report for Califor- nia. Reported earlier by Smith and Sawyer (1986) for northwestern California, but the report source not specified. This latter is also the source of the BONAP and USDA Plants record. A worldwide weed of the moist tropics and subtropics, it has also been grown out from unidentified contaminants in seed shipments from Laos to California, and thus is to be expected more widely. Centipeda minima (L.) A. Braun & Asch.: DIST: SCo: CS: GH/C: DOC: San Diego Co.: Weed in private bonsai collection, Alto Dr., La Mesa. Dominating greenhouse floor, spreading rapidly by seed. Sept. 4, 2005, G. F. Arusa s.n. (CDA, UC, UCR): NOTES: Seedlings sent by M. Jarcho in April, 2005, grown to maturity in greenhouse at CDA. A common weed in eastern Asia. Crassocephalum crepidioides (Benth.) S. Moore: DIST: SnFrB: CS: GH/C: DOC: Contra Costa Co.: Growing out of cracks in residence patio in San Pablo at Stanton & San Pablo Ave. Many here. Jan. 18, 2006, G. F. Hrusa 1684la (CDA); Riverside Co.: Weed in potted plants brought from Hawai, Cardiff Growers, Anza Rd, Temecula. July 15, 1986. D. Domenigoni PDR 631917 (UCR, OBI); Greenhouse at UCR, grown from plants collected in potted stock from Hawaii. Sept. 15, 1986, A. C. Sanders 6960 (UCR): NOTES: Garden and nursery weed. Common as a weed in Florida and Hawaii nursery stock. Crepis pulchra L.: DIST: NCoRI: CS: C: DOC: Napa Co.: West of the city of Napa. Along Mt. Veeder Rd, near perennial drainage that crosses road. A few individuals. Elev. ca. 200m. May 28, 2007, A. Solomeshch s.n. (DAV): NOTES: Reported also from Napa Co. (Consortium 2007). Gnaphalium antillanum Urb.: DIST: SnJV: CS: GH/ C: DOC: Stanislaus Co.: 2 mi SW of La Grange, yard of Allen residence. May 25, 1969. P.S. Allen 347 (DAV). NOTES: Sometimes treated as Gamochaeta antillana (Urb.) Anderb. Det. by G. L. Nesom (BRIT) (as Gamochaeta). Gnaphalium calviceps Fern.: DIST: DMoj, SCo, SnFrB: CS: N: DOC: Contra Costa, San Diego Cos: Nesom (2004 pg. 1180); Fresno Co.: City of Fresno, June, 1998, B. Fischer s.n. (DAV); San Bernardino Co.: Rancho Cucamonga, N Day Creek Blvd. under power- lines below mouth of Day Creek Canyon, alluvial fan sage scrub, 34°09'57’"N, 117°31'57’"W, elev. 655 m, June 7, 2003, K. Stockwell s.n. (UCR): NOTES: Sometimes treated as Gamochaeta calviceps (Fern.) Cabrera or Gamochaeta falcata (Lam.) Cabrera; occasionally mis- spelled as G. claviceps. Similar to the native plant gen. identified as G. purpureum L. (which is a distinctive form best treated as G. ustulatum Nutt., see Nesom 2004), differing by its narrower cauline leaves and more or less glabrous capitula. Also reported from Stanislaus Co. [Damas s.n. (DAV)], but material immature acc. to Nesom. Dets. by G. L. Nesom (BRIT) (as Gamochaeta). Gnaphalium coarctatum Willd.: DIST: NCo, n SNF, ScV, SnJV: CS: NW: DOC: Placer Co.: Dutch Flat Area, Alta Exit N of Hwy 80. Drum Powerhouse Rd, +/— 1.5 rd mi NE of intersection with Main St. Moist 102 road banks in Ponderosa pine forest. May 17, 2003. E. Dean 1758 (DAV); Humboldt, Sacramento, Stanislaus Cos.: Nesom (2004, pg. 1181): NOTES: Sometimes treated as Gamochaeta coarctata (Willd.) Kerguélen. Dets. by G. Nesom (BRIT), 2005 (as Gamochaeta). Gnaphalium involucratum G. Forst.: DIST: NCo: CS: NCI: DOC: Humboldt Co.: 1948 collection cited in Nesom (2002, pg. 518): NOTES: Sometimes treated in the segregate genus Euchiton as E. involucratus (G. Forst.) Holub, 1974. The name E. involucratus (G. Forst.) Anderb. (1991, Opera Botanica 104) is super- fluous. Gnaphalium pensylvanicum Willd.: DIST: DSon, SCo, SnJV: CS: N: DOC: Fresno Co.: Fresno?, June 1998, B. Fischer s.n. (DAV); Los Angeles Co.: Occasional annual in dry, bare areas. Glendale, at E end of Glenoaks Ave., nr. Scholl Canyon Park. Elev. 1000 ft April 21, 1978, D. Koutnik 405 (DAV); Riverside Co.: Cathedral City, Cathedral Plaza, Hwy 111 at date Palm Dr., 33°46'42’N, 116°27'27"W, elev. 300 ft, March 17, 1996, A. C. Sanders 17996 (UCR); Riverside, UCR, field crop area south of Pennsylvania Ave., 33°57'53"N, 117°20'19"W, elev. 980 ft, Feb. 11, 1971, O. F. Clarke s.n. (UCR); San Bernardino Co.: Highland, 34°08’09’N, 117°11'42”W, elev. 1500 ft, May 16, 1950, J. C. Roos 4814 (UCR); San Diego Co.: San Diego, Balboa park, near entrance to San Diego Zoo, 32°44’40’N, 117°08'54”W, elev. 300 ft, Apr. 19, 2003, A. C. Sanders 26083 (UCR); Stanislaus Co.: Near Ceres and Turlock, 2 mi. WSW of Keyes, 6524 Moffett Rd., near lateral # 3, c. % mile south of Keyes Rd., 37°32'35’N, 120°56'47’"W, elev. 80 ft, July 8, 2000, A. C. Sanders 23530 (UCR): NOTES: Sometimes treated as Gamo- chaeta pensylvanica (Willd.) Cabrera. This is a common urban weed in southern California. Additional records for most of the counties above are in the Consortium of California Herbaria (2007). Dets. by G. L. Nesom (BRIT), 2005 (as Gamochaeta). Gnaphalium stachydifolium Lam.: DIST: SNF, SnJV: CS: NW: DOC: Amador, Butte Cos: Nesom (2004, pg.- 1180); Stanislaus Co.: Nr. barn just W of Morgan Gulch, along La Grange Dam Rd, E of La Grange. May 13, 1969, P.S. Allen 310 (DAV): NOTES: Sometimes treated as Gamochaeta stachydifolia (Lam.) Cabrera. Dets. by G. L. Nesom (BRIT), 2005 (as Gamochaeta). Senecio cineraria DC.: DIST: NCoR: CS: TEN: DOC: Humboldt Co.: Roadside ditch, Hwy 36 nr. Carlotta. eight perennial plants. TO2N, ROIE, Sec. 18, H. Sept. 1, 2002, P. Haggard PDR P070860 (CDA) Det. by D. G. Kelch (CDA). NOTES: Similar to Senecio Jacobaea but heavily tomentose and about twice as tall. Reported also in Riverside and Santa Barbara Cos. (Consortium of California Herbaria 2007). Youngia japonica (L.) DC.: DIST: SW: CS: GH/C: DOC: Orange Co.: Irvine, weed in UCI Arboretum nursery, Tustin 7.5’ Q., 33°39'52"N, 117°54’14’W, elev. 15 m, Feb. 22, 2004, M. A. Elvin 3148 (UCR, IRVC); Riverside Co.: Riverside, garden weed at Canyon Crest Drive at Via Zapata. Oct. 22, 1995, D. Koutnik s.n. (CDA, UC/JEPS, RSA, UCR); Riverside, restaurant on University Ave. west of Iowa St., weeds in flowerbeds beside building, 33°58’30"N, 117°20’30"W, elev. 302 m, March 8, 2002, A. C. Sanders 24895 (UCR): NOTES: Dets. by J. L. Strother (UC) and A. C. Sanders. This species arrives regularly as a weed in nursery stock from the SE States and is more solidly established in southern California than these few records would suggest. MADRONO [Vol. 55 Balsaminaceae Impatiens capensis Meerb.: DIST: NCo: CS: TEN: DOC: Alameda Co.: Strawberry Canyon Ecological Study Area in Berkeley Hills E of UC campus, edge of Strawberry Creek where it enters culvert near parking area. Aug. 29, 1996, B. Ertter 15271 (UC): NOTES: Det. by P. Zika (WTU); native to the eastern U.S. Not seen at this site in recent years. Impatiens glandulifera Royle: DIST: NCo: CS: NCI: DOC: Marin Co.: Abundantly naturalized on Vision Rd., in wet place, Inverness. June 27, 1970, L. McHoul s.n. (CAS, CDA). Brassicaceae Berteroa incana (L.) DC.: DIST: CCo, MP: CS: NW: DOC: Lassen Co.: Anderson Ranch, southern Lassen Co., dryish, sandy soils, several acres, 39°51.332’N, 120°05.711’W, elev. 4850 ft, July, 2005, C. Battis s.n. (CDA); Loc. cit. July 18, 2005, D. McDonald PDR 1307852 (CDA and to be distributed); Monterey Co.: Single plant on old Fort Ord army base, Seaside. On burn of previous season where straw applied for erosion control. Eucalyptus Rd at summit of trail 50. July 9, 2004, V. Yadon s.n. (CDA, PGM): NOTES: Lassen Co. population stable for many years, but beginning to expand after wet 2004/5 winter season (C. Battis, personal communication to Hrusa). Additional Mon- terey Co. records available at the Consortium of California Herbaria (2007). Carrichtera annua (L.) DC.: DIST: SCo: CS: NW: DOC: San Diego Co.: Ecological Reserve, east side of El Camino Real, c. 2 km south of Palomar Airport Rd, south-facing slope in clay soil grassland in mixed chapparal and coastal | sage scrub, edge of cleared area, locally common on exposed clay soil, 33°07’06’N, 117°15'48’W, elev. 52 m/ | 170 ft, April 6, 2007, F. M. Roberts 6512 (CDA, DAV, UC, UCR); Loc cit. NOTES: Det. by A. C. Sanders (UCR). Matthiola longipetala (Vent.) DC. ssp. bicornis (Sibth | & Sm.) P. W. Ball.: DIST: DMoj: CS: NCI: DOC: Kern | Co.: Infrequent on roadside, Tehachapi. June 17, 1964, L. B. Krauter s.n. (CDA). Det. by T. C. Fuller (CDA): | NOTES: Reported also in San Diego Co. (Consortium | of California Herbaria 2007). Teesdalia nudicaulis (L.) R. Br.: edge of Parker Flats Rd. BLM Section. May 12, 2006, V. Yadon & D. Styer s.n. (CDA): NOTES: Similar — outwardly to T. coronipifolia but with longer petals and visible style. Det. by I. Al-Shebaz (MO). Campanulaceae Laurentia fluviatalis (R. Br.) E. Wimm.: DIST: nSNF, SW: CS: GH/C: DOC: El Dorado Co.: Lake Oaks Mobile Home Park, Diamond Springs, garden plots | around homes, occasional, growing as weed in lawns | with Duchesnia indica and Oxalis corniculata, 38°41'14"N, 120°49'49’"W, ca. 524m, May 11, 2007, G. K. Helkamp 11992 (DAV, UCR); Riverside Co.: Riverside, 11'* St. betw. Almond and Chestnut. Local | lawn weed escaped from cultivation. Elev. 259 m/850 ft — Feb. 14, 2004, A. C. Sanders 27388 (CDA, DAV, UCR). Caprifoliaceae Viburnum rigidum Vent.: Trail and Wildcat Gorge south of Lake Anza. Multiple F. M. Roberts 6513 (UCR): | DIST: CCo: CS: | TEN: DOC: Monterey Co.: Fort Ord, single spot at | DIST: SnFrB: CS: NW: DOC: Contra Costa Co.: Tilden Regional Park, Selby | Carlsbad, La Costa Greens © 2008] shrubs scattered in forest understory. Oct. 29, 2001, B. Ertter 17810 (UC): NOTES: Viburnum rugosum Pers. 1s a taxonomic synonym. Caryophyllaceae Cerastium tomentosum L.: DIST: KR, SNE: CS: C: DOC: Mono Co.: City of Mammoth Lakes. Grassy area in yellow pine forest. Rather common in an area some distance from dwellings. Probably deliberately intro- duced. Elev. 7900 ft 37°39'N; 118°58.5'W. July 18, 2000, G. Helmkamp 5848 (UCR). Det. by A. C. Sanders, 2002; Siskiyou County: Yreka, on ditch banks, June 10, 1958, M. V. Maxwell s.n. (AHUC). Silene armeria L.: DIST: CCo, NCoRI, SCo, SN, SnGB.: CS: C: DOC: Los Angeles Co.: San Gabriel Mtns., foothills along Hwy 2 at La Canada, turnoff to SCE Gould substation, Gould Canyon watershed. Scarce annual on weedy road edge, fls. Purple. Elev. 2000 ft/610 m. Pasadena 7.5’ quad. TO2N, R13W, Sec. 25. SB. 34°13.5'N, 118°1L.5'W, June 22, 1998, A. C. Sanders 21986A (UCR); Mono Co.: City of Mammoth Lakes. Roadside, under pines. Uncommon. 37°38.8'N, 118°58.5’'W. Elev. 8100 ft Sept. 26, 1996, G K. Helmkamp 1222 (UCR); Napa Co.: Volunteer in residence garden, (city of) Napa. Aug. 29, 1980, L. Darby s.n. (CDA); San Luis Obispo Co.: Nipomo Mesa, Mesa Dunes Mobile Home Park along Mesa View Dr., 0.2 mi. S of jet. of Halcyon Rd., SE of Oceano, open grassy lot, not related to an established garden, 35°05.5'N, 120°34.6’W, elev. 76 m, Aug. 15, 2004, G. K. Helmkamp 8&788B (UCR): NOTES: Assoc. with weeds and locally exotic “‘wildflowers’’ at Los Angeles Co. site, including Gilia tricolor and multipetaled Clarkia. Perhaps present in a sown “wildflower” mixture? Dets. by A. C. Sanders. Reported also for San Diego and Mono Cos. (Consortium 2007). Silene coeli-rosa (L.) Godron: DIST: SnFrB: CS: NCI: DOC: Santa Clara Co.: Palo Alto Foothills Park, Toyon Trail. July 7, 1967, F. P. Cronemiller 3180 (CDA). Chenopodiaceae Atriplex glauca L.: DIST: SCo: CS: N: DOC: Los Angeles Co.: Boyd (1999, pg. 118); Reifner and Hrusa (in press); Orange Co., Riverside Co. and San Diego Co.: Reifner and Hrusa (in press): NOTES: The seed is sold on the internet. This taxon is (among others) promoted for use as a “‘revegetation”’ species in alkaline areas. See the Consortium of California Herbaria (2007) for additional collections. Atriplex vesicaria Heward in Hook.f.: DIST: SCo, CCo.: CS: NCI: DOC: Los Angeles Co.: Munz, P. A. and D. D. Keck (1968, pg. 75); Weed, Northridge. Dec. 9, 1964, D. Williams s.n. (CDA); Monterey Co.: Railroad ROW, Greenfield. T18S, RO7E, Sec. 07, MD. July 9, 1974, B. Oliver & C. Twohy s.n. (CDA); Loc. cit. Aug. 19, 1974, C. Twohy s.n. (PGM): NOTES: Williams s.n. (Los Angeles Co.) the source of citation in Munz and Keck (1968). Beta macrocarpa Guss.: DIST: DSon, NChI, SnJV: CS: NW: DOC: Imperial Co.: Grain field. Aug. 1938, Henson s.n. (CDA), det. by M. K. Bellue; Weedy in sugar beet field 1.6 km west of Imperial County airport, Imperial. T15S, R13E, Sec. 22, SB. Feb. 27, 1974, G. D. Barbe 1756 (CDA, DAV); Roadside, desert slope, W of Hwy 99 to Indio, 24.7 mi NW of Westmoreland. April 4, 1962, T. C. Fuller 8008 (CDA); Hwy 115 at Streiby Rd., 6 mi east of Brawley, 32°57'36"N, 115°24’00’"W, elev. —27m, May 6, 1998, D. Bartsch 22 (SD, UC, UCR); Hwy 98 at Pullma Rd., 10 mi east of Calexico, DEAN ET AL.: CATALOGUE OF NON-NATIVE VASCULAR PLANTS 103 32°40'12”N, 115°39’00"W, elev. 7 m, May 6, 1998, D. Bartsch 01 (SD, UC, UCR); Kings Co.: Newton Bros. prop. Stratford. June 5, 1964, F. H. Surber s.n. (CDA), det. by T. C. Fuller, Feb. 1974; Santa Barbara Co.: Santa Cruz Island, NE end, just south of Cavern Point at top of coastal road, elev. 98 m, March 28, 1992, S.A. Junak SC-3111 (UCR, SBBG), det. by Detlef Bartsch, 1998: NOTES: Reported also from Los Angeles Co. (Santa Catalina Island) and San Diego Co. (Consortium of California Herbaria 2007). Crassulaceae Crassula colligata Toelken ssp. lamprosperma Toelken: DIST: SCo: CS: NW: DOC: Los Angeles Co.: Boyd (2004, pg.393-4). Cucurbitaceae Ecbhallium elaterium (L.) A. Rich.: DIST: CCo: CS: GH/C: DOC: Alameda Co.: Berkeley, volunteer in dry, shady, neglected corner of garden at 1856 Catalina Ave., no seed source evident. Aug. 21, 2004, B. Ertter G-584 (UC): NOTES: A garden volunteer also confirmed from Santa Clara Co., but a voucher not available. Cuscutaceae Cuscuta japonica Choisy: DIST: ScV, SnFrB, SnJV: CS: N: DOC: Contra Costa Co.: San Pablo, at Stanton and San Pablo Ave, on Ligustrum japonicum, Hedera canariensis, Malus, tomato, Kalanchoe. Flwrs. present, few. Jan. 18, 2006, G. F. Hrusa s.n. (CDA); Shasta Co.: Redding, Arizona St., on citrus. July 29, 2004, C. Moen PDR 1360556 (CDA); Yuba Co.: Olivehurst, Western St. Parasite on cultivated grape (Vitis). Oct. 21, 2005, Roush et al. PDR 1295325 (CDA); Olivehurst, Western Ave., TI4N, RO4E, Sec. 5, MD, on grape and plum, Oct. 4, 2005, B. Umino PDR 1295324 (CDA). NOTES: Many specimens, in addition to the above, have been seen, mostly in vegetative condition only. Apparently intentionally introduced. Capable of establishing on a wide range of host species, this dodder is used as both a vegetable and a medicinal by certain Asian cultures. Uncited locations known only from sterile material have been found in Alameda, Fresno, Los Angeles, Sacramento, Shasta, and Stanislaus Counties, and the plant is probably present in many others. The species flowers late in the season and has not yet been confirmed to set seed in California, and thus the current locations may not persist. However, the potential to reproduce sexually was confirmed through successful artificial crosses between living accessions at CDA. This pest has been known to kill or terminally damage large trees. See also http://www.cdfa.ca.gov/ phpps/ipc/noxweedinfo/noxweedinfo_jdodder.htm for more information. Euphorbiaceae Euphorbia pulcherrima Willd. ex Klotzsch: DIST: SCo, sChI: CS: NCI: DOC: Ventura Co.: Ventura, persisting in waste area south of Southern Pacific R. R. Station (near site of buildings destroyed about ten years ago), Dec. 21, 1967, H. Pollard s.n. (DAV); San Nicolas Island, bottom of Tule canyon, near pumping area. July 28, 1965, R. E. Foreman et al. 151 (UC). Phyllanthus caroliniensis Walt.: DIST: SCo: CS: GH/ C: DOC: Los Angeles Co.: Weedy in container grown ornamentals, Long Beach. March 17, 2003, Burton s.n. (CDA): NOTES: Reported also for San Diego Co. where it appears to be an established garden weed (Consortium of California Herbaria 2007, A. C. Sanders personal observation). 104 MADRONO Fabaceae Coronilla varia L.: DIST: CaR, SCo, SNH: CS: NW: DOC: Butte Co.: West side of Doe Mille Road, about 1 mi SE of bridge across Butte Creek. Elev. 2341 ft, T24N RO3E, Sec. 34, MD, dry rocky soil, common, June 22, 2004, L. Ahart 11121 (CDA, CHSC, UC); Nevada Co.: N side of Lake Van Norden, bottom of Pinus contorta woods on dry, silty flats, elev. 7100 ft, 39°19'43"N, 120°22'26’W, July 15, 2001, G. F. Arusa 15896 (CDA, CHSC); ...Lake Van Norden, dry sites on old dam (breached), elev. 7100 ft, July 15, 2001, G. F. Hrusa 15872 (CDA, CHSC). Dets. by L. Janeway (CHSC); Placer Co.: North Shore of Lake Tahoe, Stateline Road, on decomposed granite, June 23, 1985, B. Crampton 9997 (DAV); Santa Barbara Co.: Road- side, Goleta. July 19, 1968, M. E. Cravens s.n. (CDA). Det. by T. C. Fuller; Siskiyou Co.: Siskiyou Avenue exit from I-5 N of Dunsmuir, driveway to Dunsmuir Elementary School, naturalized in drainage area below road, roadbanks seeded to control erosion, June 13, 1972, T. C. Fuller & G. D. Barbe 943 (CDA, DAV): NOTES: Reported also for El Dorado, Monterey and San Diego Cos. (Consortium of California Herbaria 2007); additional records from Butte Co are also available in the Consortium database. Crotalaria capensis Jacq.:. DIST: SCo: CS: GH/C: DOC: Orange Co.: UC Irvine arboretum. Weedy near fence by bird enclosures. “Not cultivated [in the arboretum] currently or recently”. Oct. 12, 2003, A. C. Sanders 27318 (CDA, UCR). Cytisus battandieri Maire: DIST: SnGB: CS: NCI: DOC: Los Angeles Co.: Betw. La Canada and Big Pines on the Angeles Crest Highway, woody shrub to app. 15 ft, June 22, 1976, J. Cooper s.n. (CDA): NOTES: Sometimes treated as Argyrocytisus battandieri (Maire) Raynaud. Dorycnium hirsutum L.: DIST: KR, SnFrB: CS: NCI: DOC: Alameda Co.: Scattered plants in waste ground, abandoned sect. of old SCS (Soil Conservation Service) nursery, Alameda County Fairgrounds, Pleasanton. Sept. 10, 1975, T. C. Fuller 20104 (CDA); Shasta- Trinity Co.: (boundary): 1—2 mi S of Lewiston, granitic soil, Oct. 28, 1971, W. H. Brooks IIT s.n. (DAV), det. by B. Crampton, 1971 (AHUC): NOTES: Freely reseeding in garden setting in Alameda Co., including adjacent pavement crevices, suggesting high potential for inva- siveness. Lathyrus nissolia L.: DIST: NCoR: CS: NCI: DOC: Sonoma Co.: Garden weed in Santa Rosa, Mar. 7, 2003, J. Raisner PDR 1279785 (CDA). Medicago muricata All... DIST: SCoRI: CS: NW: DOC: San Luis Obispo Co.: Chimineas Unit, Carrizo Plain Ecological Preserve, along 3 Dam Rd., 0.6 mi NW of Scale House, common in grasslands, elev. 2500 ft, Apr. 27, 2005, G. Butterworth s.n. (CDA); Loc. cit. May 12, 2005, G. Butterworth s.n. (CDA): NOTES: Probably arrived via rangeland improvement program. Vouchers at CDA from several experimental plantings in mid-20" century. Medicago scutellata (L.) P. Mill.: DIST: CCo: CS: NW: DOC: Monterey Co.: Rana Creek Ranch app. 20 mi E Hwy | along Carmel Valley Rd. April 29, 2002, V. Yadon s.n. (CDA, PGM); Loc. cit. May 6, 2002, V. Yadon s.n. (CDA). Sophora_ secundiflora Lag.: DIST: SCo: CS: NCI: DOC: Riverside Co.: Several plants found growing in the upper San Jacinto Valley, slopes below hills on the [Vol. 55 east side, coastal sage scrub, elev. 1500 ft, Aug. 25, 1976, L. B. Ziegler 552 (UCR). Hamamelidaceae Hamamelis sp.: DIST: n SNF: CS: NW: DOC: Butte Co.: Cyn. N. Fk. Feather R. (Lake Oroville) betw. Big Bend Mtn. & Swayne Hill, 1.8 km NNW of mouth of French Ck. Large colony on steep, NNE-facing slope. 39°42'42"N, 121°24'43’"W, elev. 299 m, July 11, 2003, L. P. Janeway 7898 (CDA, CHSC, JEPS): NOTES: Acc. to D. G. Kelch (CDA) this colony is not referable to any of the four known species, and may represent a hybrid form persisting from cultivation and spreading by root sprouts. It is provisionally identified in California Consortium of Herbaria specimen records as H. vernalis Sarg.(Consortium of California Herbaria 2007). Further study needed. Juglandaceae Carya illinoiensis K. Koch: DIST: SCo, SnJV: CS: NW: DOC: Riverside Co: Temescal Valley, Lawson Dr. near Glen Ivy, in a wetland reserve area for the Trilogy Development, in riparian forest dominated by willows, common 15—20 m trees, 33°46'15’N, 117°29’'28’W, elev. 335 m, May 21, 2003, A. C. Sanders 26778 (RSA, SD, UCR); Riverside, west edge of UCR campus, lower drainage of Big Springs Canyon, near intersection of | Canyon Crest Dr. and University Ave., common in | riparian forest with Salix and Populus, 33°58'36’N, 117°19'56’W, elev. 320 m, May 6, 2003, A. C. Sanders | 26558 (RSA, UC, UCR); Stanislaus Co.: Edge of pond | along Hwy. 132, +/— 200 yds E of Basso Bridge, 2 mi SW of La Grange, May 27, 1969, P. S. Allen 354 (DAV): NOTES: Anderson (2002, pg. 14,) cites a | naturalized location in Ventura Co., but this remains unverified by vouchers. Besides the records for River- side Co. above, this species is also naturalized along the Santa Ana River at Riverside (Sanders, pers. obs.). Naturalized reports from San Diego Co. are also | available (Consortium 2007). Widely planted by pio- | neers and often persisting in old homestead sites. Lamiaceae | Leonurus sibiricus L.: DIST: SCo: CS: GH/C: DOC: | San Diego Co.: Rebman, J. and M. Simpson (2006): | NOTES: Record from electronic version viewable at http://www.sdnhm.org/ research/botany/sdplants/check- | list_19.html. Used in Asian and Latin American | cultures as a medicinal. Present as a weed in the | California State University, Fullerton, Arboretum (A. Sanders, pers. obs.) in Orange Co. Malvaceae Urena sinuata L.: DIST: SCo: CS: GH/C: DOC: Santa Barbara Co.: Weed in rose-growing operation in | Carpinteria, growing in coco fiber imported from Sri ] Lanka. Sept., 1998, 7. Watson s.n. (CDA): NOTES: Plants grown to maturity in CDFA Sacramento | greenhouse. | Melianthaceae Melianthus major L.: DIST: NCo: CS: NW: DOC: | Humboldt Co.: PALCO industrial timberlands, edge of disturbed logging road in second growth redwood © forest, shoulder of Road 35 at Fox Creek gate, single — shrub persisting for at least two years, UTM Zone 10, } 413888E, 4509069N, Jan. 6, 2005, G. Leppig 2201 4} (CDA, HSC); Los Angeles Co: South Coast, El Segundo | Dunes, immediately west of L.A. International Airport » and Pershing Drive, and east of Vista Del Mar Blvd., | 2008] 33.933°N, 118.433°W, coastal dunes formerly largely occupied by residential areas, the houses removed c. 15 yr ago, associated with Abronia umbellata, Camisso- nia cheiranthifolia, and Croton californicus, a scarce persisting ornamental, May 18, 1988, A. C. Sanders 7817 (RSA, UCR); Orange Co.: Upper Newport Bay, above Back Bay Drive in the vicinity of the Narrows, c. 0.8 km north of Big Canyon, west-facing slope at border of coastal sage scrub and disturbed vegetation, erect shrub, c. 1.5 m tall, a few scattered individuals, first observed on this slope in 1987, there was only a single individual covering much less area, 33°38'18"N, 117°53'12"W, Nov. 24, 2007, R. DeRuff s.n. (UCR): NOTES: At the Los Angeles Co. site, the species had been persisting with no care for 15 yr at the time of collection. At the Orange Co.site, the species has been present and increasing for 20 yr. This species can be invasive in favored climates. Moraceae Ficus rubiginosa Desf. ex Vent.: DIST: SCo: CS: TEN: DOC: Orange Co.: Irvine. University Dr along tidal creek just E of I-73. Brackish marsh and border vegetation. Uncommon shrub to 1.5m in stream channel, nr. 33°39'03”"N, 117°51'18”’W, elev. 9 m/30 ft, Dec. 9, 2006, D. G. Kelch 06.565-B (CDA, SD). Det. by D. G. Kelch (CDA); San Diego Co.: Batiquitos Lagoon Ecological Reserve, north side, east of the I-5 Fwy, along the trail past the nature center, solitary 3 x 3 m shrub, epiphytic, c. 4m up, on trunk of naturalized Phoenix canariensis, an incipient strangler?, 33.091°N, 117.296°W, elev. 0-6 m, June 22, 2005, A. C. Sanders 31177 (RSA, UC, UCR): NOTES: Report from Los Angeles Co. (Consortium of California Herbaria 2007), is of uncertain naturalization. Myrtaceae Eucalyptus pulchella Desf.: DIST: SnFrB: CS: NW: DOC: Alameda Co.: Joaquin Miller Park in Oakland Hills, locally dominant along Sequoia-Bayview trail just south of horse arena as a nearly solid stand, with some Acacia mixed in, all size classes, Aug. 19, 2001, B. Ertter 17805 (UC); Loc. cit., Oct. 21, 1993, D. Keil 24454 (OBI, UC): NOTES: Determination confirmed by M. Ritter (UC Santa Cruz Arboretum). Synonyms include Eucalyptus linearis Dehnh. and Eucalyptus amygdalina var. angustifolia Burtt Davy. This is only the second report on naturalization of this species worldwide (after New Zealand). Melaleuca citrina (Curtis) Dum.Cours.: DIST: SCo: CS: NW: DOC: San Bernardino Co.: San Bernardino Valley, Fontana/Bloomington, property on south side of Slover Ave., c. 0.5 mi E of corner Slover Ave. and Sierra Blvd. Ca. 10 plants ‘“‘evidently naturalized”’ along an arroyo. Fontana 7.5’ quad. 34°01'00’N, 117°25’00’"W, elev. 329 m/1080 ft, Sept. 11, 2000, Provance 2224 (UCR): NOTES: Det. by L. Craven (CANB). This species has previously been most often treated as Callistemon citrinus (Curtis) Skeels. Other localities from San Bernardino Co. appear to be cultivated plants persisting near old homesites. Natu- ralized reports also from Orange and San Diego Cos. (Consortium 2007). Melaleuca viminalis (Sol. ex Gaertn.) Byrnes: DIST: SCo: CS: NW: DOC: Orange Co.: Newport Beach, Big Canyon near Jamboree Rd. [off Newport Back Bay]. 2 to 3 plants seen. Near 33°37'42”N, 117°52'40’W, July 18, 1987, R. DeRuff 301 (UCR); City of Irvine, San Diego Ck. vicinity of Campus Ave. at University Dr, DEAN ET AL.: CATALOGUE OF NON-NATIVE VASCULAR PLANTS 105 naturalized in disturbed riparian scrub and along dirt access rd. in alkaline soil, UTM (NAD 83) 11S 0421602E, 3724316N, elev. ca. 15 ft, Nov. 14, 2006, R. E. Reifner 06-674, 06-675 (CDA); Newport Beach, Irvine Coast Trail at San Diego Ck., nr. SR-73 crossing, S of Jamboree Rd., common in alkaline soil in Atriplex scrub and disturbed riparian scrub along creek, many small saplings, UTM (NAD 83) I1S 0419948E, 3723890N, elev. ca 11 ft, Nov. 12, 2006, R.E. Reifner 06-559 (CDA); City of Newport Beach, urban creek draining into San Diego Ck. at confluence with Newport Bay nr. SR-73 overpass, disturbed riparian woodland, UTM (NAD 83) 11S 04198895E, 3723901N, elev. ca 19 ft, Nov. 12, 2006, R. E. Reifner 06-669 (CDA); San Bernardino Co.: San Bernardino Valley, Fontana/Bloomington, S side of Slover Ave., ca. 0.5 mi east of corner Slover Ave. and Sierra Blvd, solitary tree growing in arroyo, 34°01’00"N, 117°25’00’"W, elev. 329 m/1080 ft, Sept. 11, 2000, M. Provance 2235A (UCR); San Diego Co.: Carlsbad, S edge of Carlsbad Highlands Ecological Reserve, 1 km south of Cerro de la Calavera, two trees, 1 ca. 3 m tall, the other 5—6 m, clearly spontaneous in moist soil nr. Spring, 33°09'28"N, 117°16'54"W, elev. 30-61 m/100—200 ft, Sept. 3, 2004, A. C. Sanders 28793 (CDA, UCR); Buena Vista Ecological Reserve, lower section between Pacific Coast Highway and the railroad, solitary 2 m shrub, nr. 33°10’06’N, 117°21'29"W, elev. 2—3 m/5— 10 ft, April 16, 2004, A. C. Sanders 27453 (UCR); North Park/Golden Hill area, Cedar Ridge Park Canyons along stream beside I-15, several escaped trees, reproducing with saplings, 32°43’N, 117°07 W, June 30, 2000, J. Rebman 6933 (UCR); Encanto area, canyon just N of Market St. and E of Malcolm X Library and Euclid, tree to 5m tall, 32°42’39’N, 117°04'46"W, Mar. 31, 2000, Rebman 6362 (UCR): NOTES: This species has been most often treated as Callistemon viminalis (Sol. ex Gaertn.) Cheel. All UCR duplicates determined by L. Craven (CANB). Addi- tional records (as either Melaleuca or Callistemon) from Los Angeles, San Bernardino and especially San Diego Cos. are available at the Consortium of California Herbaria (2007). Metrosideros kermadecensis W.R.B. Oliv.: DIST: SCo: CS: C: DOC: Los Angeles Co.: La Piedra State Beach, just above the highest tide line, May 30, 2004, M. O’Brien MOB3-04 (CDA), det. D. G. Kelch, G. F. Hrusa (CDA). NOTES: Population likely not planted, but not obviously reproducing. Separable from M™. excelsa by its shorter filaments (1—2 cm vs. 3-4 cm), smaller leaves (2-5 cm vs. 5-10 cm) and slightly smaller capsules (ca. 6 mm vs. 7-9 mm). A variegated form is commonly cultivated, but this specimen is not varie- gated. Metrosideros excelsa has been reported escaped in South Africa (Richardson and Rejmanek 1999), and M. kermadecensis has escaped in Hawaii (Evenhuis and Eldredge 2004). Additional record from Los Angeles Co. (Consortium of California Herbaria 2007) may or may not be naturalized. Myrtus communis L.: DIST: cSNF: CS: NW: DOC: Stanislaus Co.: At edge of dredge tailings, % mile W of La Grange, north across a grassy field from State Hwy 132, in woodland with Quercus lobata, Ficus carica, Cydonia oblonga, Rubus procerus, and Carex nebras- censis, Nov. 30, 1968, P. Allen 36 (DAV); Shrub beside marsh S of old Dredge Camp, 2.5 mi SW of La Grange, Dec. 1, 1968, P. Allen 37 (DAV): NOTES: Reported 106 also from San Diego and Sonoma Cos. (Consortium of California Herbaria 2007). Syzygium australe (Wendl. ex Link) B. Hyland: DIST: SCo: CS: NW: DOC: San Diego Co.: Mission Valley Preserve, San Diego, disturbed riparian forest and scrub with many introduced species, Aug. 3, 2004, L. R. Landrum & J. Rebman 10961 (ASU, UC): NOTES: Syzygium australe has been segregated from the closely related and commonly cultivated S. paniculatum Gaertn. (Hyland 1983). Separation appears difficult, and California specimens of Syzygium often cannot be determined to species. The relative frequency of cultivation or escape of these two species in California is not known, and situations where they are found are often equivocal as to naturalization. Records of specimens labeled as S. paniculatum are available for Los Angeles, Orange and San Diego Cos. (Consortium of California Herbaria 2007). Nyctaginaceae Boerhavia diffusa L.: DIST: ScV: CS: NCI: DOC: Yolo Co.: West Sacramento, off West Capitol Avenue, Y2 mi. E of E levee of Yolo Bypass, disturbed sandy soil with Salsola and Polygonum aviculare, flat mat on ground, 0.5 m across, | plant, Aug. 26, 2005, A. M. Shapiro s.n. (DAV): NOTES: Det. by A. G. Murdock and R. Spellenberg, 2007. Nymphaeaceae Nelumbo lutea Willd.: DIST: SCo, ScV: CS: N: DOC: Tehama Co.: Table Mountain Ranch, Red Bluff. T28N, RO3W, Sec. 2, MD, July 19, 2005, Stoffel PDR 1411565 (CDA); San Diego Co.: Corte Madera Ranch, Pine Valley, Aug. 30, 1980, H. van der Werff 4136 (CDA, SD): NOTES: Tehama Co. site is a large population in an artificially maintained lake. There is potential for spread of the species at this site, but survival of this population is dependent upon lake managers keeping the water level high. The San Diego Co. location is also apparently a private lake, but the current status of this population is not known. Oxalidaceae Oxalis carnosa Molina: DIST: SnFrB: CS: GH/C: DOC: Contra Costa Co.: Weedy in nursery, Interna- tional Succulent Institute, betw. Moraga and Orinda, March 25, 1976, G. D. Barbe 2117 (CDA): NOTES: Oxalis megalorrhiza Jacq. misapplied (Dandy and Young 1959). Oxalis corymbosa DC.: DIST: SCoRO, SnFrB: CS: GH/C: DOC: Santa Barbara Co.: Nr. Lompoc, Frank Costa Farm, weed in agricultural field, common, flowers white to purple, June 21, 2005, S. Fennimore s.n. (DAV); Santa Clara Co.: Hayward (Mt. Eden), Mt. Eden Nursery (West Jackson St.), greenhouse rose beds, grown out from bulbs sent in by L. Pyeatt, bulb scales with orange stripes, flowers rose-purple, tube yellow- green with purple stripes, Jan. 14, 1977, J. McCaskill 783 (DAV): NOTES: Field workers at the Lompoc locality dig up the rhizomes and suck on them because they are sweet and taste like caramel. Passifloraceae Passiflora ‘Coral Seas’: DIST: CCo: CS: NW: DOC: Monterey Co.: Mission Trail Nature Preserve, Carmel, near 11'* St. entrance, spreading over 0.25 ha area, no fruits found, June 17, 2004, J. Randall s.n. (DAY): NOTES: Probably the ‘cultivar’ ‘Coral Seas’, a sterile hybrid between P. manicata and an unknown species or MADRONO [Vol. 55 several species (Ulmer and MacDougal 2004). The origin of this occurrence is unknown. Plumbaginaceae Limonium binervosum (G.E. Sm.) Salmon: DIST: SCo: CS: NW: DOC: Orange Co.: Newport Back Bay Dr., in side canyon, disturbed coastal scrub, riparian forest, and brackish marsh, common clumping herb to 30 cm in upper marsh, fls. sordid white, nr. 33°37'52"N, 117°53'02’"W, elev. 6 m/20 ft, Dec. 9, 2006, D. G. Kelch 06.547 (CDA, DAV, SD): NOTES: Det. by D.G. Kelch (CDA). This name covers a complex of closely related species. As we cannot determine which of these may be represented by this population, we are using the broader name. Limonium gougetianum Kuntze: DIST: ScV: CS: TEN: DOC: Alameda Co.: Livermore, Marina Ave., betw. road edge and ditch, site floods temporarily in winter, 30°39'24"N, 121°45'30’W, elev. 170 ft, June 20, 2007, D. Petersen s.n. (CDA); Yolo Co.: Roadside weed, County Rd. 96, Woodland, elev. 30 m, 38°42’40’N, 121°50'22’”E, July 24, 2006, B. Lyon PDR 1340754 (CDA). NOTES: A difficult genus taxonomically, these determinations are based on descriptions in Flora Europaea (Pignatti 1972). Comparison to authentic material may ultimately result in modification to the identities used here. Polygonaceae Polygonum aubertii Henry: DIST: KR, NCo, SCo, ScV: CS: N: DOC: Los Angeles Co.: Beside signal tower, P.E.RR. crossing in East Pasadena, Nov. 7, 1956, G. C. Fleischman s.n. (CDA); Santa Barbara Co.: Bank of Mission Creek nr. San Andreas and San Pasqual St., Santa Barbara, Nov. 5, 1961, H. M. Pollard s.n. (CAS, CDA). Cited on this specimen label as also at Micheltorena St. S of S.P.RR crossing; Sonoma Co.: Sonoma Creek bank next to Sonoma Creek Apts. in urban Sonoma (city) limits, June 5, 1998, S. Schoenig | s.n. (CDA); Ventura Co.: Persistent in long-abandoned garden site, S. Montgomery St., Ojai, June 21, 1962, H. M. Pollard s.n. (CAS, CDA): NOTES: Differing from | Polygonum baldschuanicum Regel in its scabrous | inflorescence axis and smaller, white to greenish-white | flowers (as opposed to glabrous infl. axis and larger | pink flowers). Possibly all references to P. baldschuani- cum in California are actually this species which is | naturalized in numerous sites in North America. A survey of herbarium specimens has not been undertak- | en. Reported also from Glenn and Siskiyou Cos. (Consortium of California Herbaria 2007). Portulacaceae Portulacaria afra (L.) Jacq.: DIST: SCo: CS: GH/C: | DOC: Los Angeles/Ventura Co.: Leo Carrillo State | Beach, 3 clumps (plants?), 12 ft tall, probably persistent | or waifs from previous settlement, flowers pink/ — magenta, Jan. 29, 2003, M. O’Brien s.n. (CDA): | NOTES: Commonly cultivated but rarely flowering. | Highly persistent in mild coastal climates. Spreads by | dispersal of broken stem joints, and several individuals | from the site above were establishing at the base of the cliff on which the mature plants were found. Reported | also for San Diego Co. (Consortium of California Herbaria 2007) but naturalization status uncertain. Proteaceae Grevillea robusta Cumn.: DIST: CCo: CS: NW: DOC: San Luis Obispo Co.: Nat. grove, 2.5 mi N of Arroyo Grande on Hwy 227. T31S, R13E, Sec. 33, MD, Jan. — 2008] 27, 1977, G. D. Barbe 2242 (CDA); Near summit along Hwy 227 between San Luis Obispo and Arroyo Grande, mosaic of coast live oak woodland, chaparral, and coastal scrub, well established on hillside and in canyon drainage, more than 100 individuals in all stages from seedlings to mature trees, elev. 600—700 ft, Nov. 12, 1988, D. J. Keil 21194 with L. M. Kelly, K. Bay, and F. Wertman (OBI); Orange Co.: Newport Beach, upper Newport Bay, beyond the east end of Mesa Dr., marsh level in disturbed coastal sage scrub, with Brazilian pepper and annual grassland mixture, 33.656°N, 117.872°W, elev. 3m, a single small tree, 3 m tall, growing near salt marsh, probably originated from two trees planted within irrigated landscaping at top of bluff, May 2005, R. DeRuff, 345 (UCR); Ventura Co.: Spontaneous as seedlings in crevice at side of rd., intersection of Signal and Eucalyptus Sts., Ojai, Dec. 9, 1967, H. M. Pollard s.n. (CAS, CDA, SBBG): NOTES: Both San Luis Obispo Co. records are the same population. Reported also for Los Angeles and San Diego Cos. (Consortium of California Herbaria 2007). Ranunculaceae Ranunculus trilobus Desf.: DIST: CCo: CS: N: DOC: San Mateo Co.: Along Hwy 101 in Half Moon Bay, growing in former agricultural field on coastal flat terrace, clay soils, 1—2 ft tall, April 17, 2002, P. Greer s.n. (DAV). Rhamnaceae Rhamuus alaternus L.: DIST: SCo, SnFrB, SnJV: CS: NW: DOC: Contra Costa Co.: NW slope of Lime Ridge Open Space on S side of Ygnacio Valley Rd., nr. culvert for main swale, with Baccharis and Juglans, Oct. 18, 1998, B. Ertter & T. Morosco 16431 (UC); Loc. cit., Apr. 14, 2006, Ertter & Gowen 18667 (CDA, UC); Los Angeles Co.: Los Angeles area, undeveloped mesa between Sybil Brand Institute and Sheriff Academy, 1 km SSW of 710/10 freeway intersection, Monterey Park, three plants ca. 2m tall naturalized in brushy area with Baccharis pilularis, 34°03'N, 118°09.5'W, elev. 165 m, June 3, 1998, D. S. Cooper s.n. (UCR); Orange Co.: Newport Back Bay Dr. in side canyon, disturbed coastal scrub, riparian forest, and brackish marsh, rare shrub, 1.3 m, 33°37'52”N, 117°53'02’"W, elev. 6 m/20 ft, Dec. 9, 2006, D. G. Kelch 06.546 (CDA, SD); Riverside Co.: Riverside, NE side of the UCR campus, two large shrubs naturalized at the edge of an arroyo beneath a dead blue gum, assoc. with Schinus polygamus and S. molle, weedy shrubs at margins of campus, scarce weed in hedge of Pittosporum and Xylosma on N side of parking lot 13, 33°58.5’N, 117°19’W, elev. 335 m, March 14, 1996, A. C. Sanders 17972 (UCR); Riverside, E side of UCR campus, 33°58'N, 117°19'W, elev. 335 m, March 1, 1997, 4. C. Sanders 19726 (UCR): NOTES: Seeds apparently dispersed by birds (A. Sanders pers. obs.). The plants on the UCR campus are doubtless dispersing from a cultivated plant somewhere in the area, but no cultivated plants of this species appear to be present on the campus. Additional records for the documented counties above and reports for Merced and Monterey Cos. are available from the Consortium of California Herbaria (2007). Rosaceae Aphanes arvensis L.:. DIST: DSon, SCo: CS: GH/C: DOC: Los Angeles Co.: Huntington Gardens, San Marino, weed in lawn and edge of lawn, March 13, DEAN ET AL.: CATALOGUE OF NON-NATIVE VASCULAR PLANTS 107 1965, J. C. Roos s.n. (UCR); Pasadena, Campbell Seed Company (sent in for identification), May, 1927, P. Kennedy s.n. (DAV); Riverside Co.: Coachella Valley, Palm Desert, lawn weed in condominium complex at Mountain Shadows Golf Course, April 6, 1992, J. Di Bernardo 92-1 (UCR); Loc. cit. April 16, 1992, J. Di Bernardo 92-2 (UCR): An additional location in Riverside Co. is in the Consortium of California Herbaria (2007). As per the label data, this species is also naturalized in wildlands. Cotoneaster simonsti Baker: DIST: NCo: CS: NW: DOC: Del Norte Co.: Zika (2005, pg. 208). Rosa sempervirens L. cv. ‘‘Felicite et Perpetue’’: DIST: SnFrB: CS: TEN: DOC: Contra Costa Co.: Along railroad west of Crockett at west edge of Carquinez Straits, mid-slope above abandoned railroad station, April 2, 1993, B. Ertter 11481 (UC): NOTES: Det. by D. G. Kelch (CDA); the site is close to old habitations where a number of cultivated plants persist. This individual is spreading by root sprouts. Rubiaceae Diodia virginiana L.: DIST: GV: CS: N: DOC: Shasta Co.: Lawn and sidewalk weed on Leonard St., city of Redding, T31N RO5W, Sec. 23, MD, Sept. 10, 2003, E. Finley s.n. (CDA): NOTES: First record for this species west of the Rocky Mountains. Additional specimens from Shasta Co. are reported at the Consortium of California Herbaria (2007). Sapindaceae Cardiospermum halicacabum L.: DIST: GV: CS: TEN: DOC: Sacramento Co.: Just N of Elk Grove city limit, N side of Dwight Road opposite end of Portofino St. growing spontaneously in drainage ditch, partly ascending into a Himalayan blackberry thicket, Sept. 9, 2006, A. M. Shapiro s.n. (DAV); Solano Co.: Putah Creek, E of low water bridge, Aug. 13, 1943, M. Ensign s.n. (DAV): NOTES: Consortium of California Her- baria (2007) record from San Diego Co. may or may not be naturalized. Cupaniopsis anacardioides (A. Rich.) Radlk.: DIST: SCo: CS: NW: DOC: San Diego Co.: Batiquitos Lagoon Ecological Reserve, north side of lagoon, east of the I-5 Fwy, along the trail past the nature center, disturbed uplands along trail, scattered Eucalyptus groves, two saplings seen, 1.5 and 4m tall, in moist upland areas, clearly spontaneous, nr. 33°05'33’N, 117°17'25’"W, elev. 0-6 m/0—20 ft, April 22, 2005, A. C. Sanders 29658 (CDA, DAV, UCR): NOTES: Additional records from San Diego Co. and reports for Los Angeles and Orange Cos. are available at the Consortium of California Herbaria (2007). Saxifragaceae Francoa ramosa D. Don: DIST: SnFrB: CS: TEN: DOC: Alameda Co.: Cragmont Rock Park in north Berkeley Hills, sparingly naturalized in untended area with Algerian ivy, July 24, 2004, B. Ertter 18443 (UC). Scrophulariaceae Veronica biloba L.. DIST: MP: CS: NW: DOC: Modoc Co.: Fox Mountain Spring, ca. 10 km NW of Adin, Pinus ponderosa forest with Calocedrus and Abies, T40N, RO8E, Sec. 35, MD, annual, flowers blue with white at base of throat, elev. 1620 m, June 20, 1991, B. Bartholomew 5790 (CAS, RSA): NOTES: Determina- tion and report by L. Gross (RSA). Similar to Veronica persica L., but with bracts shorter than the leaves and 108 style < 1 mm. Sometimes treated as Pocillia biloba (L.) W. A. Weber. Solanaceae Nierembergia frutescens Durieu.: DIST: SCo: CS: TEN: DOC: Los Angeles Co.: Glendale, Chevy Chase Country Club, Small bush along seventh fairway, disturbed area of a sub-division, Adenostoma fascicula- tum chaparral on rocky, dry soil, elev. 1000 ft, June, 1971, T. A. Zink s.n. (UCR); Riverside Co.: Jurupa Mountains, Sunnyslope area, ca. 0.5 mi E of 34th St. and Armstrong Rd., seepy disturbed vacant lot, uncommon herb along dry margins of the seep, 33.0133°N, 117.4305°W, elev. 274 m/900 ft, July 23, 1998, M. Provance 931 (UCR): NOTES: Dets. by S. White and A. C. Sanders, 2002. Nierembergia hippomanica Miers.: DIST: DMoj/ DSon boundary: CS: TEN: DOC: Riverside Co.: Dry Morongo Wash, 100 yards downstream from the bridge at the San Bernardino County line, growing among willows in the wash, May 20, 1981, G. K. Helmkamp s.n.(UCR). Solanum cardiophyllum Lindl.: DIST: ScV: CS: TEN: DOC: Yolo Co.: University of California at Davis, Vegetable Crops Dept. Campbell Tract, weed in tomato field, source not certain, probably of Mexican origin, Aug. 21, 1969, P. Smith s.n. (DAV); About a dozen plants scattered over an acre of tomato variety test plots, Veg. Crops Dept., Univ. Calif., Davis, Aug. 27, 1969, T. C. Fuller 18855 (CDA). Det. by D. S. Correll, 1969. Solanum chrysotrichum Schitdl.: DIST: SCo: CS: NCI: DOC: San Diego Co.: Balboa Park, San Diego, naturalized along drainageway of an unlandscaped canyon S of the Museum, Oct. 21, 1969, 7. C. Fuller 19011 (CDA, DAV); Loc. Cit., April 22, 1969, 7. C. Fuller 18245 (CDA). Dets. by M. Nee (NY), 1988 (orig. det. as S. hispidum Pers.). Solanum jasminoides Paxt.: DIST: CCo: CS: NCI: DOC: San Luis Obispo Co.: 1.7 mi NNW of Cayucos, T28S, RIOE, Sec. 32, MD, elev. 50 m. Mar. 28, 1936, C. M. Belshaw 1751 (DAY). Solanum mauritianum Scop.: DIST: SCo, SnFrB: CS: NW: DOC: Alameda Co.: Joaquin Miller Park. south terminus of Sunset Trail below curve in Bayview Trail, single plant on shaded bank, 5 Dec 2005, B. Ertter & D. Gowen 18658 (UC); Los Angeles Co.: Commonly naturalized in cyn. bottom and adjacent slopes, Nichols Cyn. Rd. at Dresden Dr., Santa Monica Mts, Los Angeles. TOIS, R14W, Sec 5 SB. March 9, 1976, 7.C. Fuller 20120 (CDA); Orange Co. Residential yard weed, Mayfair Ave., Orange. May 1, 2000, A. C. Helm PDR1223083 (CDA),det. G.F. Hrusa; Santa Barbara Co.: Spontaneous in oranges and open fields, NE corner East Valley & Hot Springs Rds, Montecito. Oct. 19, 1960, 7. C. Fuller 5333 (CDA); N side of RR tracks 200 yds W of San Ysidro Rd. Montecito. March 23, 1971, 7. C. Fuller 19823 (CDA); Hwy 101 and Glen Annie Road, NE corner of intersection in riparian area along creek. Oct. 27, 2000. J. DiTomaso s.n. (DAY). Det. by L. Bohs, 2001): NOTES: Collected in Pasadena (Los Angeles Co.) from cultivated material as early as 1949, and noted then by G. C. Fleishman (CDA) as an “‘undesireable spreader.” A_ difficult pest in South Africa (Bromilow 1995). Also known from Riverside and San Bernardino Cos. (Sanders 1996) and reported from San Diego Co. (Consortium of California Herbaria 2007). MADRONO [Vol. 55 Verbenaceae Verbena incompta Michael: DIST: CCo, GV, SNF (likely also in SCo and SW): CS: NW: DOC: Fresno Co.: Lincoln & Brawley Rds., Oct. 20, 1959, T. C. Fuller 3141 (CDA); Perennial weed along a ditchbank, corner of Dickensen and Jensen Avenues, between Fresno and Kerman, June 16, 1982, B. Fischer s.n. (DAV); Marin Co.: Waste places in southern part of Mill Valley, elev. 25 ft, plants 5—6 ft tall, Sept. 5, 1943, J. Howell 19323 (CDA, DAV, UC, UCR); Merced Co.: Ditchbank, 0.8 mi north of Livingston, Cressey St., Aug. 15, 1954, B. Crampton 2262 (AHUC, UC,UCR); Placer Co.: Folsom Lake, American River bike trail, uncommon here, elev. 300 ft 38°42’N, 121°10’W, May 27, 2002, M. I. Wibawa 268 (CDA); Sacramento Co.: Colony at base of American River Bluffs, in +/— wet spot along | lakeshore bike trail, 38°37’N, 121°02’W, elev. 75 m, July 29, 1990, G. F. Hrusa 8171 (CDA, DAV): Grant Line Road, 0.6 mi S of junction with Douglas Ave., roadside, Aug. 14, 1978, B. Crampton 9657 (DAV); Cosumnes River Preserve (multiple locations and collectors) 1994-2003; Sacramento Delta, Georgiana Slough, middle part, elev. 5m. 38°09'04’N, 121°35'199W, Oct. 11, (CDA); San Joaquin Co.: Caswell Memorial State Park, 1998, G. FL Hrusa 14864 | common along moist riverbanks, especially below the | easternmost campgrounds, July 8, 1992, G. F. Hrusa 9160 (DAV); Caswell Memorial State Park, Middle . reaches of Fence Line trail, 37°41’N, 121°11’W, Oct. . 14, Perennial up to four feet high, corolla lavender, on moist or flooded ground along drainage ditches, 1998, G F. Hrusa 14894 (CDA): Yuba Co.: | common, occurring intermittently along the road for ' seven miles, elev. ca. 200 ft, road to Grass Valley, four mi. E of Marysville city limits, Sept. 25, 1938, A. Carter | 3721 (DAV, JEPS): NOTES: Native to South America. This species remained unnamed and was consistently (mis)identified as Verbena bonariensis L. worldwide until described from Australian material (Michael | 1995). While this name is thus a general replacement for V. bonariensis in California, a single specimen from | Glenn Co. (marshy area, E side County Rd. “R’’, S of | Hwy. 162 nr. Willows, May 15, 1980, J. R. Keck s.n., CDA) represents true V. bonariensis — in a wildland — setting. It has also been collected as a garden “‘weed”’ | from Sonoma and San Diego Cos. but these cannot be distinguished as “‘weedy” or “‘cultivated”’ instances. True V. bonariensis also occurs in the eastern US. It is | separated from V. incompta by its broader corolla limb | 1mm), with flowers which — (+/— 45mm vs. +/— overtop the apex of the inflorescence spike, inflores- cence parts clearly with stalked glands, stamens not | inserted above the middle of the tube, and slightly longer nutlets. Violaceae Viola conspersa Reichenb.: DIST: CCo: CS: N: DOC: | Monterey Co.: Cultivated plants in Pacific Grove, | reseeding prolifically; originally collected from a. driveway crack along 17 Mile Drive, Del Monte Forest, | Pebble Beach, Sept. 15, 2003, V. Yadon s.n. (CDA), det.» by G. F. Hrusa, M. Kerr (CDA), 2003: NOTES: Plants sold in local nurseries as Viola labradorica are indistin- | guishable from this, which may be a multi-species hybrid | involving V. labradorica, V. conspersa and V. waltheri; however, the morphological evidence is equivocal and | the plants fit within the range of V. conspersa. 2008] ANGIOSPERMS—MONOCOTS Arecaceae [Palmae] Washingtonia robusta H. Wendl.: DIST: SCo: CS: NW: DOC: Los Angeles Co.: Malibu Bluffs State Park, Marie Cyn., many individuals, to 15 ft tall, June 19, 2004, M. O’Brien MOB8-04 (CDA): NOTES: Numer- ous specimens have been collected of this species indicating that it is also naturalized in Orange, Riverside, San Diego, and Ventura Counties (Consor- tium of California Herbaria 2007; Roberts et al. 2004). It is a common escape along rivers and drains and is common along the Santa Ana River at Riverside. Bromeliaceae Billbergia nutans H. Wendl.: DIST: CCo: CS: C: DOC: Monterey Co.: Persisting at roadside dump along Ronda Rd. 200 yds W of jct. with Sunridge Rd., Del Monte Forest, Pebble Beach, Oct. 9, 2003, V. Yadon s.n. (CDA). Commelinaceae Callisia repens (Jacq.) L.: DIST: SCo: CS: GH/C: DOC: Riverside Co.: Riverside, yard of house at 422 Campus View Dr., irrigated garden, spontaneous, invading from neighboring yard, 33°58'45’"N, 117°19'W, elev. 341 m/1120 ft, Nov. 20, 2001, A. C. Sanders 24886 (CDA, RSA, UCR); San Bernardino Valley region between Pedley and Mira Loma, Sims prop- erty at NE corner of Limonite Ave. and Bain St., weedy disturbed area beside a dirt vehicle path, 33°58’30’N, 117°30'15”"W, elev. 210 m/690 ft, Nov. 28, 2001, A. C. Sanders 24889 (CDA, RSA, UCR): NOTES: Also seen elsewhere spreading away from cultivated sites, but apparently always in irrigated areas. This may be new to the area, as it has only been seen in the last few years. Cyperaceae Carex pendula Huds.: DIST: ScV: CS: NW: DOC: Butte Co.: Janeway, L. (2005, pg. 125): NOTES: Orig- inally reported as Carex spissa (Janeway 1992, pg. 58). Cyperus gracilis R. Br.: DIST: SCo: CS: NCI: DOC: Los Angeles Co.: Santa Monica, weed in the alley behind the house at 460 23rd St., Aug. 15, 1983, 7. Yutani s.n. (CDA, UCR); Loc. cit., Aug. 8, 1983, 7. Yutani s.n. (RSA): NOTES: Det. by G. Tucker (EIU). Cyperus regiomontanus Britt.: DIST: SCo: CS: GH/ C: DOC: Los Angeles Co.: Pasadena, M. R. Alexander residence, weed in garden, apparently annual, Nov. 5, 1991, O.F. Clarke s.n. (UCR): NOTES: Det. by G. Tucker (EIU), but with a question mark. The material seems reasonably mature, but the plant base is poorly represented and the “‘apparently annual” comment may have created uncertainty. Cyperus retrorsus Chapm.: DIST: SCo: CS: GH/C: DOC: Riverside Co.: Riverside, UCR campus Botanic Garden, naturalized along the trail in the chaparral section just west of (below) the lath house, Sept. 13, 1994, S. Morgan s.n. (UCR): NOTES: Det. by G. Tucker (EIU). First seen about 1990 as a weed in a pot in the lath house; allowed to grow to see whether it proved to be of interest. It is now escaped and well- established outside. Eleocharis lanceolata Fernald: DIST: ScV: CS: NCI: DOC: Butte Co.: Weedy rice field, ca. 5 mi. W of Richvale, July 29, 1949, M. Nobs & S. G. Smith 1084 (UC): NOTES: UC specimen annotated by S. G. Smith, 2000, as “Undoubtedly introduced, only California record!” DEAN ET AL.: CATALOGUE OF NON-NATIVE VASCULAR PLANTS 109 Iridaceae Sisyrinchium pruinosum Bickn.: DIST: s SnJV, SCo: CS: GH/C: DOC: Kern Co.: Weed in _ turfgrass operation, Bakersfield, July 20, 1998, 7. Prather s.n. (CDA); Los Angeles Co.: Locally common lawn weed, Montebello Hills in Montebello, 34°01'36’N, 118°06'34’"W, elev. 122 m, May 17, 2001, A. C. Sanders 24313 (CDA, UCR). Juncaceae Juncus gerardii Loisel.: DIST: deltaic GV, SnFrB: CS: NW: DOC: Contra Costa Co.: Martinez Regional Shoreline, brackish marsh along Pickleweed Trail west of Arch Bridge, 38°1.324'N, 122°8.575'W, May 31, 2003, Ertter & Matson 18175 (UC); Solano Co.: Benecia State Recreation Area, Southampton Marsh, growing in tidal salt marsh, June 25, 2002, B. Grewell s.n. (DAV); Loc. cit. June 20, 2004, B. Grewell s.n. (CDA, DAY). Juncus usitatis L.A.S. Johnson: DIST: GV, CaRF: CS: NW: DOC: Butte Co.: About 2 mi W of Oroville Dam, near Longs Bar Monument, on WN side of Thermalito Diversion Pool, Lake Oroville Recreation Area, July 23, 1992, L. Ahart 6828 (UC). Merced Co.: Merced National Wildlife Refuge, Merced Unit, West Farm Field, irrigated pasture, Aug. 10, 2004, S. Winter 1286 (JEPS); Stanislaus Co. (E corner): Tuolumne River NW of junction Hwy 132 and La Grange Road (J59), July 24, 2003, B. Ertter 18259a & 18261 (UC); Tehama Co.: Wild Cat Road between Manton Road and Battle Creek, ca 20-22 airmiles NE of Red Bluff, shallowly sloping seepage area on lava, 40°24.986'N, 121°58.785'W, elev. 1070 ft, July 20, 2003, B. Ertter 18254 (UC); Yuba Co.: Damp red soil, edge of a small pond, about 100 yds. N of Long Ravine Rd, ~300 yds SE of the intersection of Camp Far West Rd and Long Ravine Rd, north of Camp Far West Res., uncommon, elev..398 ft, 39°05 19.9"N, 121°17'33.3" W, Oct. S; 2002, L. Ahart 9947 (UC). Det. by P. Zika (OSU). Originally determined as J. effusus var. exiguus: NOTES: This rapidly spreading Australian species is easily confused with J. effusus s.1., but easily distinguished by the blunt tepals that are notably shorter than the hardened globose capsule. The determination of Ahart 9947 was confirmed by K. Wilson of NSW (personal communi- cation to B. Ertter, 2005). Liliaceae (sensu lato) Allium ampeloprasum L.: DIST: ScV, SnFrB: CS: NCI: DOC: Marin Co.: Tiburon Peninsula, shoulder of road below Old St. Hilary’s Chapel, July 8, 1961, J. Penalosa 1951 (DAV); Sacramento Co.: Escaped from cultivation along roadside, El Centro and San Juan Rds, TOON, RO4E, Sec. 21, MD, June 5, 1974, J. Yamaguchi s.n. (CDA). Det. By G. D. Barbe (CDA): Reported also for Orange Co. (Consortium of Califor- nia Herbaria 2007). Aloe striatula Haw.: DIST: SnFrB: CS: NW: DOC: Alameda Co.: Claremont Canyon Regional Preserve, annual grasslands along trail heading uphill from upper intersection of Panoramic Way and Dwight Way to Panoramic Place, localized but well-established colony stretching along trail, just coming into bloom, bush- sized, much branched, with succulent deep green leaves, flowers tubular, yellowish with green tinge, orangish at base, May 20, 2007, B. Ertter 19027 (UC). Asparagus densiflorus (Kunth) Jessop: DIST: CCo: CS: NW: DOC: Monterey Co.: Plants growing in leaf litter under Pinus radiata and Quercus agrifolia on undeveloped ridge...Pebble Beach, Aug. 8, 2004, V. 110 Yadon s.n. (CDA, PGM): NOTES: Well-established at Monterey Co. site acc. to V. Yadon, but site slated for development. Reported also from Orange and San Diego Cos. (Consortium of California Herbaria 2007). Asparagus setaceous (Kunth) Jessop: DIST: SCo: CS: NCI: DOC: Ventura Co.: Spontaneous by seed or root- propagating on waste ground (site of former garden), Junipero St. betw. Santa Clara and Main Sts., Ventura, July 15, 1966, H.M. Pollard s.n. (CAS, CDA, SBBG). Det. by G. D. Barbe (CDA); originally determined as A. plumosus Baker. Bulbine semibarbata (R. Br.) Haw.: DIST: CCo: CS: N: DOC: Monterey Co.: Roadside near Carmel Highlands, April 11, 1990, V. Yadon H-3864 (PGM): NOTES: Identified as B. semibarbata only on the basis that it has 3 bearded and three glabrous stamens. Also reported from San Luis Obispo Co. (Cambria, V. Yadon personal communication to Hrusa). Cordyline australis (G. Forst.) Endl.: DIST: NCo, SnFrB: CS: NW: DOC: Alameda Co.: Berkeley yard, 1303 Albina Ave., volunteer (doubtless from plants in neighborhood), 6+ m tall, flowered 2 seasons without watering or other special care, old leaves persisting 2+ years, May 22, 2004, J. L. Strother 1366 (UC/JEPS); Humboldt Co.: One tree on Cooper Slough S of Myrtle Ave. at edge of brackish marsh, Eureka, May 4, 2004, G. Leppig 2153 (CDA, HSC); Sporadic trees 3-4 m tall, along RR tracks at 2nd & Commercial Ave., Eureka, June 8, 2004, G. Leppig 2165 (CDA, HSC): NOTES: A Cordyline, fully-naturalized and spreading in the vicinity of Plantation (coastal Sonoma Co.), appears to be this species but reproductive material for confirmation not yet at hand. Reported also from Monterey and San Diego Cos. (Consortium of California Herbaria 2007). Hyacinthoides hispanica (Mill.) Rothm.: DIST: SnFrB: CS: C: DOC: Alameda Co.: Waif in forested part of Albany Hill at end of Madison Street, NW Albany, Feb. 11, 1995, B. Ertter 13930 (UC): NOTES: Cultivated under several generic assignments such as Scilla hispanica and Endymion hispanica: NOTES: Reported also from Humboldt Co. (as Scilla hispanica) (Consortium of California Herbaria 2007). Narcissus papyraceus Ker Gawl.: DIST: SCo: CS: GH/C: DOC: Los Angeles Co.: Boyd, S. (1999, pg. 133): NOTES: Listed also for San Diego Co. (Beauchamp 1986) but naturalization there uncertain. Often persis- tent from ornamental plantings. Muscari armeniacum Leicht. ex Baker: DIST: NCo: CS: NCI: DOC: Humboldt Co.: Samoa Cookhouse, 3 mi NW of Eureka, moist weedy field, escaped orna- mental, March 28, 1976, J. M. DiTomaso 344 (DAV). Poaceae Brachypodium sylvaticum (Huds.) Beauv.: DIST: CCo: CS: NW: DOC: San Mateo Co.: Dark shady area under redwoods at 4506 Oak Knoll Dr., Emerald Hills, Nov. 10, 2003, K. Melo 1387519 (CDA); Large colony beneath deep shade of coast redwoods, Hwy 84 at Grandview Terrace and Schilling Lake, Martin Creek drainage, Dec. 1, 2003, K. Melo PDR 1387956 (CDA, UC, UCR, UT); Along Hwy. 84 at Grandview Dr. and into adjacent redwood forest, TO06S, RO4W, Sec. 12, MD, Dec. 12, 2003, J Beall PDR 1387963 (CDA): NOTES: Additional records from San Mateo Co. are available at the Consortium of California Herbaria (2007). A rapidly spreading weed in central Oregon forests. These are the first records for California. The distribution in California is centered on the site of the MADRONO [Vol. 55 old Schilling estate, but has spread widely in the local area, esp. in the Martin Creek drainage. Danthonia decumbens (L.) DC.: DIST: NCo: CS: NW: DOC: Del Norte Co.: NW of Crescent City McNamara Airport near pond in grassy opening, July 28, 2000, C. Golec s.n. CHSC); Crescent City, Crescent City Marsh Wildlife Area, common in Sitka spruce forest and coastal scrub on edge of marsh, UTM Zone 10, 403364E, 462176N, June 24, 2003, G. Leppig and T. Labanca 1988 (CDA, HSC); Humboldt Co.: Big Lagoon County Park, Big Lagoon Bog, rare in peatland, UTM Zone 10, 405321E 4557202N, elev. 3 m, Sept. 28, 2003, G. Leppig 2057 (HSC): NOTES: Also known as Sieglingia decumbens (L.) Bernh. Recorded erroneously in California by Matthews (1997). See Hrusa et al. (2002, pg. 64) for discussion. Dinebra retroflexa (Vahl) Panz var. retroflexa: DIST: SCo: CS: NW: DOC: Orange Co.: Yorba Linda, Yorba Regional Park, Santa Ana River bank near intersection _ of E. La Palma Ave. and Fairmont Blvd., edge of rip- rap at margin of river channel, wet mud with many | herbs, including Leptochloa uninervia, Ludwigia pe- — ploides, Eleusine, Eclipta prostrata, Bidens laevis, etc., 33°51'52’N, 117°46'26"W, elev. 91 m, 23 Aug. 2004, O. F. Clarke s,n, (UCR). Det. by T. Columbus, 2004. Riverside Co.: Reifner et al. (2003, pg. 312): NOTES: This species is occasionally encountered by the Cali- fornia Department of Food and Agriculture Seed | Laboratory in flower-seed lots imported from Africa (J. Effenberger personal communication to Hrusa), and | it is more established than the few cited records indicate | (A. Sanders pers. obs.). Eragrostis tenella (L.) Roem. & Schultes: DIST: SCo: | CS: GH/C: DOC: Santa Barbara Co.: Weed growing in | coconut fiber used for root support in hydroponic | roses, glasshouse in Santa Barbara, seedling grown to maturity in greenhouse, Sacramento, June 3, 1998, 7. Watson s.n. (CDA): NOTES: Determination somewhat | tentative; authentic material not yet compared. Glyceria declinata Bresbiss.:. DIST: CaR, GV, SNF, | NcoR, SCo: CS: NW: DOC: Amador Co.: Colony in | now-dry vernal pool, app. 3 mi. E of Sacramento Co. | line on Michigan Bar Rd., May 5, 2004, G. F. Hrusa | 16267 (CDA); Sutter-Ione Rd. 0.6 mi. E of Hwy 124 | and +/— 3/4 mi. N of Ione, disturbed, moist drainage below artificial pond, 38°23’03”N, 120°54'04”W, June 7, | 2001, G. F. Hrusa 15868 (CDA); Butte Co.: Dry disturbed clay soil, slickens from the Cherokee Mine, | on the E side of Clark Rd. (Hwy 191) about 200 yds S | of Dry Creek, May 21, 2000, L. Ahart 8377 (CDA, CHSC); Red Hill Ranch, Nugent Rd. Gridley, growing / in rice stubble in field, 39°22’38’N, 121°44’32”W, April 17, 2006, M. Stewart PDR 1290445 (CDA); Calaveras . Co.: Wet creekbed of tributary of Black Creek, +/— | 3 mi. S of Copperopolis, dry bottom of pool formed by small dam, 37°56'18”N, 120°36'56’W, June 12, 2000, G. | F. Hrusa 15478 (CDA); Fresno Co.: 2 mi. E of Hildreth — Rd., Kennedy Table Mountain top, April 17, 2001, C. | Witham and J. Buck 80 (DAV); Humboldt Co.: Arcata, roadside ditch on Community Center Drive, TOSN, | ROIE, Sec. 33, H, May 5, 1996, G. Leppig 329 (HSQ); | Heavy in ditches, Arcata Bottom area, July 16, 1968, E. C. Whitney s.n. (CDA); Nevada Co.: Wet soil in bottom of drying swale, E side of meadow on S side of McCourtney Rd. about 300 yds E of inters. of McCourtney Rd.and White Oak Dr., May 28, 2002, L. Ahart 9673 (CDA, CHSC); Mendocino Co.: Union | Lumbert Co. Fort Bragg. May 3, 1928, C. S. Myaska — 2008] s.n. (CDA); Marin Co.: Marshall Beach Rd. ca. 1 mi NW of Tomales Bay State Park. Shallow water, margin of small pond. March 31, 1979, G.H. True 8470 (CAS, CDA); Riverside Co: Morongo Indian Reservation, SE edge of Burro Flats in vicinity of natural pond and bog, 33°59'32'"N, 116°50'41"W, elev. 1098-1159 m, April 24, 1996, A. C. Sanders 18088 (UCR); Morongo Indian Reservation, seep in Wood Canyon, 33°59'50’"N 116°49'48’”W, elev. 1220 m, April 23, 1998, 7. B. Salvato MOR- 905 (UCR); Sacramento Co.: Vallensin Ranch, wet bottom of deep clayey pool at corner of Dillard Rd. & Ranch Rd. 38°22'N; 121°18’W, April 13, 1996, G F. Hrusa 12796 (CDA); Lawn in Rancho Cordova. March 17, 2003, J. Davidek PDR 1278116 (CDA); South Sacramento, vernal pool of S side of Calvine Rd, ~ 1/4 mi. E of Strawberry Creek crossing, large colony in formerly wet part of pool, May 11, 1993, G. F. Hrusa 10935 (CDA); Vernal pools and intervening grasslands immediately E of Jaeger Rd. and +/— 1 mi. E of Sunrise Blvd., 38°32'47"N, 121°13'39"W, April 25, 2002, GF. Hrusa 16012 (CDA); Open grasslands immediately N of Meiss Rd. ca 3 mi SE of Dillard Rd., wet drainage on E side of stock pond, 38°28'N, 121°43’W, April 10, 1994, G. F. Hrusa 11764 (CDA); disturbed, leveled, former vernal pool field to W of Plant Pest Diagnostics Center at 3294 Meadowview Rd., Sacramento, March 31, 1999, G. FL. Hrusa 14932 (CDA); Large colony, prostrate or decumbent, ephem- erally wet old roadbed at corner of Sunrise Blvd. and Keifer Rd., site still with ponded water, elev. 50 ft, May 20, 2005, G. F. Hrusa 16459 (CDA); 3 mi NW of Galt, in Ladino Clover seed field. April 28, 1975, Walker s.n. (DAV); San Joaquin Co.: Replacing clover in a pasture, Flood Rd., Linden, TO2N, RO8E, Sec. 19, MD, Sept. 16, 1974, J. B. Gianelli s.n. (CDA); Shasta Co.: Munz and Keck (1968, pg. 1481); Sonoma Co.: In lawn. May 21, 1982, C. Elmore s.n. (DAV); Abundant in low areas of pasture, Santa Rosa district, April 3, 1968, H. F. McCracken s.n. (CDA); Weed in pasture, Mill Station Rd., Sebastopol, May 22, 1969, Ag. Commissioner s.n. (CDA); Stanislaus Co.: Bed of a vernal pool along the RR 2 miS of Oakdale on the Waterford Road, May 11, 1953, B. Crampton 1247 (DAV); Sutter Co.: Growing in fallow rice field, clay soil, March 24, 2004, M. Stelmok PDR 1411715 (CDA); Calpine electrical generation facility. Disturbed field to W of E Township Rd. 39°0S5'N, 121°43’W, April 15, 1997, G. F. Hrusa 13681 (CDA); Tehama Co.: Springy area on N side of Mill Ck. 1.9 km ENE of Black Rock and Ponderosa Way, April 21, 2000, L. P. Janeway 6669 (CDA, CHSC); Seep on N side of Mill Ck. 3.4 mi. W of Black Rock Campground and Ponderosa Way, July 9, 2004, L. Ahart 11183 (CDA, CHSC); Vestal Rd., 4.9 mi. W of ject. with Cannon, Pettyjohn, and Reeds Ck. Rds., W of Red Bluff and S of Hwy 36, T27N, RO6W, Sec. 7, MD, April 27, 1998, V. Oswald & L. Ahart 8978 (CDA, CHSC); Yolo Co.: grown from seed collected May 11, 1953 in vernal pool bed 2 mi. S of Oakdale, April 15, 1954, B. Crampton 2243 (AHUC); Yuba Co.: Weed in rice field previously a pasture, found prior to planting, March, 1984, JF. Williams s.n. (CDA, DAV); Damp soil in a rice field, San Shintaffer Farm, S of Woodruff Lane, Sept. 17, 1999, L. Ahart 8248 (CDA, CHSC): NOTES: Included in Munz and Keck (1968) in which it was distinguished from G. occidentalis based on narrower leaf width and shorter stature. In the treatment in Flora North America (Barkworth and Anderton 2007), a key characteristic is the obvious protrusion of the palea DEAN ET AL.: CATALOGUE OF NON-NATIVE VASCULAR PLANTS 111 beyond the lemma in at least some florets. Most deter- minations above by M. Barkworth (UTC). Reported also from Lake, Mariposa, Monterey and Nevada Cos. (Consortium of California Herbaria 2007). Additional records for many of these and the above documented counties are also available in the Consortium database. Paspalum notatum Flugge var. saurae Parodi: DIST: DSon, SCo, SnFrB, ScV: CS: NCI: DOC: Alameda Co.: Pleasanton, USDA-SCS Plant Materials Center, lawn weed by headquarters building, Sept. 29, 1965, B. Crampton 7520 (AHUC); Riverside Co.: Mira Loma, lawn weed at 317 Swan Lake Mobile Home Park, Sept. 19, 1976, C. Tilforth 1224 (DAV, RSA); Blythe, Aug. 1, 1972, L. Ede s.n. (AHUC); Sacramento Co.: Turfgrass, 3501 Mignon St. Sacramento, Sept. 5, 1969, 7. C. Fuller 18882 (CDA, DAV, UCR); Ventura Co.: Large patch in lawn on W side of Loma Dr. nr. Cruzero St., Ojai Valley, Oct. 12, 1963, H. M. Pollard s.n. (CAS, CDA, DAV): NOTES: Although listed in The Jepson Manual as Paspalum notatum (sine var.), the description is of var. notatum as defined by Zuloaga et al. (2004, pg. 46— 47). Paspalum notatum var. saurae is comprised of those forms with narrower blades (0.2—0.4 cm vs. 0.4—-1 cm) and shorter, narrower spikelets (2.8—3.2 mm vs. 3.2— 4.0mm). While the two specimens cited are from cultivated situations, the relative naturalized distribu- tion of these two forms in California is not known, and thus var. saurae could be expected anywhere var. notatum has been reported. Rytidosperma caespitosum (Gaudich.) Connor & Edgar: DIST: SCo, SnFrB: CS: NW: DOC: Alameda Co.: NW edge of Strawberry Canyon on E edge of main UC-Berkeley campus, eucalyptus woodland, with R. richardsonii, (see below). June 26, 2003, B. Ertter 1823la (UC); San Diego Co.: Near Fairbanks Ranch, ridge south of Lusardi Creek, ca. 0.2 km N of San Dieguito Rd and 1.3 km NE of Fairbanks Lake, elev. 83 m, March 29, 2005, F. M. Roberts 6136 (UCR); San Mateo Co.: Cascade Ranch State Park, betw. White- horse Ck. and Cascade Ck., Aug. 4, 1992, G. Clifton s.n. (UC): NOTES: Dets. by H. E. Connor, 2004. Addi- tional records from San Diego Co. are available at the Consortium of California Herbaria (2007). Rytidosperma penicillatum (Labill.) Connor & Edgar: DIST: KR, NCo, SCo, SnFrB: CS: NW: DOC: Alameda Co.: Berkeley, lawn of Cragmont Park, “‘an aggressive weed, hard to control, particularly where lawn is watered little,” Aug. 15, 1941, G. L. Stebbins 3202 (AHUC). Acc.to B. Ertter, not found in Cragmont Park in 2005.; Berkeley, Cragmont Park, from lawn, where growing as weed, June 1950, G. L. Stebbins 020 (UC). Prev. det. by Vickery as Danthonia racemosa, 1951.; North Berkeley, residential area, July 19, 2000, B. Ertter 17419 p.p. (mixed coll. with P. racemosum) (UC); Humboldt Co.: Loleta (coast), Aug. 6, 1938, B.A. Madson s.n. (AHUC); Bayside Golf Links, “‘probably introduced in grass seed mixture,” June 1941, Mrs. Wm. Grotzman s.n. (UC); Sheep pasture near “Hungry Hollow”’, Bald Mountain, locally abundant “‘as if a recent introduction,” Aug 17, 1941. J. P. Tracy 16994 (AHUC, UC); Loc. cit. June 28, 1942, J. P. Tracy 17267 (UC); Loc. cit. Aug. 11, 1949, J. P. Tracy 18451 (UC); Parrott’s Ranch, Table Bluff, ca 1.5 mi N of Loleta, Jun 1947, A. H. Murphy s.n. (AHUC); near Loleta, Jul 1947, R. Tofsrud 200 (AHUC); Bear River, 3 mi E of Capetown, July 19, 1956, R. Evans s.n. (AHUC); Mendocino Co.: Point Arena, July 15, 1941, D. Jensen s.n. (DAV); Point Arena, Aug. 1949, W. H. Brooks 111 Pi2 (AHUC); Hwy | 1/2 mi N of Albion, July 26, 1955. B. Crampton 3041 (AHUC); Mendocino. July 27, 1955, B. Crampton 3128 (AHUC); Hwy 1, 4.5 mi S of Mendo- cino, June 29, 1966, B. Crampton 7792 (AHUC); Sinkyone Wilderness State Park, 1/8 mi SE of Jones Beach, July 7, 1989, F. Bowcutt 1366 (DAV); Santa Barbara Co.: Santa Barbara, upper Hillside Park, lawn grass, May 13, 1947, C. F. Smith 2209 (DAV); Loc. cit. July 22, 1949, C. F. Smith 2406 (DAV); Santa Clara Co.: Palo Alto, parking area on University Ave. at Chaucer St., June 29, 1960, J. 7; Howell 35482 (AHUC, CDA); Siskiyou Co.: Yreka. Aug. 15, 1940, H. B. Shontz s.n. (AHUC). Sonoma Co.: Hwy 1, 2 mi N of Stewarts Point, July 22, 1957, B. Crampton 4358 (AHUC, UC); Hwy 1, 11.7 mi S of Stewarts Point, July 22, 1957, B. Crampton 4392 (AHUC); Sea Ranch, near Hwy 1, June 24, 1974, M. M. Hektner et al. 068 (DAV); Annapolis Rd nr landing strip at Sea Ranch, Nov. 30, 1980, R. D. Stone 332 (JEPS); Bodega Bay, behind community center on W side Hwy 1, July 19, 2003, B. Ertter 18246 (CDA, DAV, UC): NOTES: Dets. by H. E. Connor and S. J. Darbyshire, 2004. Acc. to these authorities, plants reported for California as Danthonia pilosa R. Br. (=Rytidosperma pilosum (R. Br.) Connor & Edgar) are actually this species (see Darbyshire and Connor 2003). Rytidosperma racemosum (R. Br.) Connor & Edgar: DIST: KR, NCo, SCo, SnFrB: CS: N: DOC: Alameda Co.: Berkeley at corner of Hearst and Oxford, May 1951, A. Haig s.n. (AHUC, DAV); Campus of UC- Berkeley, under Eucalyptus trees north of Life Sciences Bldg., Sept. 11, 1952, R. W. Pohl 7201 (UC); Hearst Ave. sidewalk, Oxford Tract, UC Campus, Berkeley, Aug. 25, 1953, B. Crampton 1609 (AHUC); West side Albany Hill, eucalyptus forest, June 13, 1998, B. Ertter 16207 (UC); North Berkeley, residential area, July 19, 2000, B. Ertter 17419 p.p. [mixed coll. with P. penicillatum| (UC); Multiple collections from North Berkeley residential area, June 19-20, 2003, B. Ertter 18216-18223 (UC); Albany, San Gabriel Ave., S of Brighton Ave., June 21, 2003, B. Ertter 18224 (CDA, DAV, UC); Marin Co.: Angel Island, perimeter road just north of West Garrison, locally common weed, June 26, 1978, G. H. True 8468 (AHUC); Yolo Co.: Davis, June 11, 1951, B. Madson s.n. (AHUC): NOTES: Dets. by S. J. Darbyshire and H. E. Connor, 2004. Additional records from Alameda Co. are available at the Consortium of California Herbaria (2007). Rytidosperma richardsonti (Cashmore) Connor & Edgar: DIST: SnFrB: CS: N: DOC: Alameda Co.: NW edge of Strawberry Canyon on E edge of main UC- Berkeley campus, eucalyptus woodland, June 26, 2003, B. Ertter 18231 (CDA, DAV, UC) [with R. caespito- sum]: NOTES: Det. by S. J. Darbyshire, 2004. Additional records from Alameda Co. are available at the Consortium of California Herbaria (2007). Sporobolus creber De Nardi: DIST: ScV: CS: N: DOC: Glenn Co.: Ranch 4 mi S of Willows and 5 mi E on County Rd. 60, just N of Sacramento Valley Wildlife Refuge, elev. ca. 100 ft Cattle pasture, present for several years, Sept. 20, 1995, R. Holzapfel 1 (CDA, UC, UCR). Det. by C. G. Reeder (ARIZ), Jan. 2006: NOTES: Reported also from Butte Co. (Consortium of California Herbaria 2007). Similar to and easily confused with Sporobolus indicus (L.) R. Br., a species common in the same region of the Sacramento Valley. Stipa papposa Nees: DIST: SnFrB: CS: C: DOC: Alameda Co.: City of Berkeley, Ashby Ave., across MADRONO [Vol. 55 street from Alta Bates Hospital, Nov. 9, 1983, Blumler s.n. (UC): NOTES: Determined by M. Barkworth, Sept. 1983 from plant in same population. Escape from cultivation. Synonyms include Achnatherum papposum (Nees) Barkworth; Jarava plumosa (Spreng.) S. W. L. Jacobs & J. Everett; Stipa ichu Kunth. Zea mays L.: DIST: NCo, SCo: CS: C: DOC: Humboldt Co.: Fay Slough Wildlife Area, two vagrant plants on edge of vernal ponds managed for waterfowl, UTM Zone 10, 407378E, 4517441N, Oct. 17, 2003, G. Leppig 2078 (HSC); Riverside Co: Lake Skinner Reserve area, Diamond Valley, canyon along Rawson Road going into hills toward Crown Valley, a solitary plant c. 1 m tall along wash on canyon bottom — doubtless a waif, 33.666°N, 117.008°W, Nov. 20, 1997, A. C. Sanders 21609 (UCR). Pontederiaceae Pontederia cordata L.: DIST: CCo, SCo: CS: N: | DOC: Alameda Co.: Ohlone College, edge of stagnant | pond, Oct. 11, 2001, K. Peek PDR P199016 (CDA); Loc. cit. Nov. 5, 2001, K. Peek PDR P147746 (CDA); Fremont, Stivers Lagoon, S of Lake Elizabeth, under gazebo, Sept. 1, 2003, L. Ellis s.n. (JEPS); Monterey Co.: (Apparently clonal) colony in Salinas River bed, at Davis Rd. crossing, Oct. 21, 2004, V. Yadon s.n. (CDA, PGM); Riverside Co.: Margin of 3 acre pond, Thun- derbird Country Club Golf Course, Rancho Mirage, | June 23, 1988, C. Butsch s.n. (CDA); San Joaquin Co: | San Joaquin Delta, 14 Mile Slough, outside of Stock- | ton, E of I-5, July 1, 2005, F. Zarate PDR 1349506 — (CDA): NOTES: This species is known to occur in San | Diego Co. but no vouchers have been collected yet (A. Sanders, pers. obs.). Some naturalized locations may be | derived from intentional plantings; Monterey Co. site is a frequently used dump for garden waste. Zingiberaceae Hedychium flavescens N. Carey ex Roscoe: DIST: CCo: CS: TEN: DOC: Santa Cruz Co.: Gully tributary | to Aptos Creek, Nisene Marks State Park boundary, below (west) of Redwood Drive, plants in the colony | are spreading along creek by vegetative fragmentation, no flowers have been produced during 5 yrs of observation, Aug. 30, 1997, D. W. Taylor 16194 (JEPS, [ UC): NOTES: Reported also from Monterey Co. (Consortium of California Herbaria 2007). Zosteraceae Zostera asiatica Miki: DIST: NCo: CS: NW: DOC: | Sonoma, Los Angeles, Marin, Monterey, San Francisco, — San Luis Obispo, San Mateo, Santa Barbara, Santa Cruz, Ventura Cos.: Phillips and Wyllie-Echeverria © (1990): NOTES: Marine angiosperm. Zostera japonica Ascher & Graebner: DIST: NCo: CS: NW: DOC: Humboldt Co.: Humboldt Bay, Indian ~ Island, SW shore, 40°48'54.7"N, 124°10'68.5”"W, 41.2. MLW (mean low water), mid-intertidal zone on sand and mud, July 8, 2002, G. Leppig 1782, 1783, 1784 | (CDA, HSC); Arcata Bay, mudflat W of City of Arcata | oxidation ponds, upper intertidal zone, 2 m from dike, NAD 27 UTM Zone 10, 4522557N, 408068E, Nov. 17, 2006, G. Leppig & D. Couch 2396 (HSC, UC/JEPS): NOTES: Established in bays and estuaries in Oregon, | Washington, and British Columbia; ongoing eradica- tion efforts in Humboldt Bay directed by California Dept. of Fish and Game Marine Division. Population still present in 2007. | MADRONO, Vol. 55, No. 2, pp. 113-131, 2008 THE SALSOLA TRAGUS COMPLEX IN CALIFORNIA (CHENOPODIACEAEB): CHARACTERIZATION AND STATUS OF SALSOLA AUSTRALIS AND THE AUTOCHTHONOUS ALLOPOLYPLOID SALSOLA RYANIT SP. NOV. G. F. HRUSA California Department of Food and Agriculture, Plant Pest Diagnostics Branch, 3294 Meadowview Road, Sacramento, CA 95832-1448 FHrusa@cdfa.ca.gov J. F. GASKIN USDA Agricultural Research Service, Northern Plains Agricultural Research Laboratory, 1500 N. Central Avenue, Sidney, MT 59270 ABSTRACT Over the past century in California, the invasive weed Salsola tragus (russianthistle) has become a widespread and troublesome pest plant. Early attempts at biological control of russianthistle achieved only partial success. Efforts to improve effectiveness of renewed biocontrol efforts revealed that two distinct, often sympatric, genetic entities comprise what has been called Salsola tragus: Salsola tragus and Salsola ‘type B’. Efforts to identify and characterize ‘type B’ resulted in recognition of a third form, ‘type C’. We present a taxonomic and morphological examination of Salsola tragus, Salsola ‘type B’, Salsola ‘type C’ and Salsola paulsenii using discriminant analysis with DNA sequence genotypes as the taxonomic framework. Sa/sola tragus and ‘type B’ were morphologically distinct; ‘type C’ was morphologically intermediate between them and contained DNA sequence genotypes that were an additive mixture of haplotypes mostly exclusive to tetraploid S. tragus and others exclusive to diploid ‘type B’. “Type C’ 1s a fertile allohexaploid that originated via hybridization between S. tragus and ‘type B’. We provide a pre-existing name, Salsola australis, for ‘type B’, and propose Salsola ryanii sp. nov. for ‘type C’. Morphological variation, habitats, and dispersal behaviors among these Sa/sola taxa were examined in the herbarium and in the field. These are compared and discussed. Key Words: Chenopodiaceae, discriminant, PEPC, polyploidy, russianthistle, Salso/a, speciation, tumbleweed. Open almost any floristic account that includes California, and the non-native plant known by the common name russianthistle or tumbleweed is referred to by one of several different scientific names. In 1996, S. L. Mosyakin showed that Salsola tragus L. was the correct name to use for the widespread North American tumbleweed. Thus, at least the following six names should be considered synonyms of or misapplications to Salsola tragus L.: Salsola australis R. Br., S. iberica (Sennen & Pau) Botsch., S. kali L. subsp. ruthenica (Ijin in Keller et al.) Sod & Jav. in Sod et Javorka, S. kali L. var. tenuifolia Tausch ex Mogq., S. pestifer A. Nelson (S. ‘pestifera’), and S. ruthenica Iljin in Keller et al. Some of these may be actual synonyms of S. tragus and others distinct or segregate taxa, but their application to genuine Salsola tragus is, as presently under- stood, uncertain or incorrect. And so despite the considerable variation within and among popu- lations of S. tragus in North America, all were previously referenced in whole by any one of the six names above, or by S. tragus itself. Salsola tragus sensu lato will here be used in reference to this morphologically variable group of plants. In North America, S. tragus sensu lato appeared first in the mid-19'"" century (Ryan and Ayres 2000); in California it arrived around 1890 in the Mojave Desert near Lancaster in San Bernardino Co., and had been collected widely by 1911 (Jepson 1914). In California and Arizona, Salsola tragus sensu lato is distributed at elevations below about 2500 m where it grows in semi-alkaline, open, and usually disturbed habitats. In California, it occurs along the south coast, in the mountain foothills, the low and high deserts and the Central Valley; it is also on the Modoc Plateau, throughout the eastern Sierra Nevada valleys, and in the Mojave Desert. The largest California populations occur from the southern Sacramento Valley south to Tehachapi Pass, with extensive stands in the northern San Joaquin Valley. There is considerable among-year variation and large stands frequently appear in the Mojave Desert and California South Coast Ranges. Other than as a casual, it is absent only from the mesic North Coast Ranges, the high Sierra Nevada and the Cascade Mountains. Wherever stands occur, tumbling plant ‘skele- tons’ in the late fall and winter pile against intercepting fences, fill drains and ditches and are 114 a moving hazard on roads. Biological control efforts in the 1960’s resulted in the introduction of two Coleophorid moths (Coleophora kli- meschiella Toll and Coleophora parthenica Meyr- ick), but these failed to significantly reduce plant abundance (Goeden and Pemberton 1995). Dur- ing a second biocontrol research effort in the late 1990’s, field observations suggested that rus- sianthistle populations expressed differential sus- ceptibility to herbivory or infection. Bruckart et al. (2004) found differential establishment of the fungi Colletotrichum gloeosporioides (Penz.) Penz. and Uromyces salsolae Reich. among local populations of russianthistle and similar differ- ential establishment of insect biocontrol agents was documented by Sobhian et. al. (2003). With these biocontrol inconsistencies in mind, Ryan and Ayres (2000) examined allozyme variation within California russianthistle and found that two distinct genetic forms were sometimes present within populations that were otherwise thought to be monotypic. Although both forms had the outward appearance of the common tumbleweed, experienced observers found they were more or less visually distinguish- able. Moreover, a chromosome count showed that one was diploid (2n = 18), the other tetraploid (2n = 36) (Ryan and Ayres 2000). Subsequently, a qualitative assessment by the first author revealed that the tetraploid formed rounded, ascending-branched tumbleweeds with a dense spicate inflorescence and wingless mature tepals at the lower nodes. The diploid in contrast, had a more upright habit, was more horizontally branched, had less condensed inflorescences, and mature fruits with winged tepals at both the upper and lower nodes. At the time, it was not known which type, if either, was nomenclaturally typical Salsola tragus, and the tetraploid was arbitrarily referred to as ‘type A’, the diploid as ‘type B’ (Ryan and Ayres 2000). Mosyakin concluded, based on his familiarity with the S: tragus lectotype, that the tetraploid ‘type A’ was the typical form (S. L. Mosyakin, Kholodny Insitute of Botany, Kiev, Ukraine, personal communication), and several years later, Gaskin et al. (2006) confirmed that application using molecular markers. The indigenous range of ‘type A’ (Salsola tragus sensu stricto) was therefore known (Mo- syakin 1996; Rilke 1999). However, plants referable to ‘type B’ were not described by Mosyakin (1996, 2003), and Rilke (1999) dis- cussed them only as a minor variant of S. tragus, adventive in southern Africa and possibly Aus- tralia. In contrast, a cluster analysis of RAPDs data for Sal/sola tragus sensu lato, grouped ‘type B’ not with ‘type A’, but with Salsola paulsenii Litv. (Ryan and Ayres 2000). During Ryan and Ayres’ examination of Salsola tragus, a third allozyme profile in MADRONO California Sa/sola was recognized (F. Ryan USDA Agricultural Research Service, Fresno, CA personal communication). The tested mate- rial, from near Coalinga in Fresno County, was given the working title ‘type C’. Gaskin et al. (2006) identified a series of nuclear DNA sequence haplotypes that distin- guished Salsola ‘type A’ and ‘type B’. Here we provide a morphologic and taxonomic examina- tion of Salsola tragus, ‘type B’, and ‘type C’ based on the differential presence of those haplotypes, with a morphologic-only examination of Salsola | [Vol. 55 | paulsenii for context. We discuss these plants’ | variation, and provide names for ‘type B’ and | ‘type C’. We conclude with discussion of their observed ecological behaviors and comment on adaptation in ‘type C’. METHODS Molecular Following extraction of DNA by standard | methods (as described in Ryan and Ayers 2000), amplification of the intron between the fourth © and fifth exon of the phosphoenolpyruvate | carboxylase gene utilized the primer pair ppcx4f — (S’-ACTCCACAGGATGAGATGAG-3’) and | ppex5r (5’'-GCAGCCATCATTCTAGCCAA- | 3’) designed by J. F. Gaskin from the sequences of other taxa of the Caryophyllales found in | GenBank. Amplification was conducted after a | 2 min denaturation at 95°C and consisted of 30 cycles of 95°C (1 min), 52°C (1 min) and 72°C | (2 min); followed by 5 min at 32°C. The two PCR products (one band approximately 500 bases in length and the other approximately 400 bases in length) were present in all samples. These | bands were separated by electrophoresis on a 2% agarose gel and the shorter band was excised (the | identity of the longer band is unknown, and its sequence variation was not useful for this | analysis). DNA was purified with the QIlAquick Gel Extraction Kit (Qiagen Inc., Valencia, CA). , The resultant template was sequenced on a. Beckman CEQ 2000XL (Beckman Coulter Inc., | Fullerton, CA) using reagents and protocols | supplied by the manufacturer and the same | primers mentioned above. Each heterozygotic genotype was cloned and sequenced to determine | the haplotypes involved. Clones were created | using the Promega pGEM-T Vector System II (Promega Corp., Madison, WI), then sequenced using the protocol above. Sequences were aligned | by hand using SE-Al software (Rambaut 1996) | and are available in GenBank (accession numbers | are in Gaskin et al. 2006). Haplotypes were | arranged manually into a most parsimonious network (Fig. 1; also see Gaskin et al. 2006, Fig. 2). --- J/A/S ez:- oo" - -- J] ‘type B’ S. australis Fia. 1. HRUSA AND GASKIN: SALSOLA TRAGUS AND SALSOLA AUSTRALIS - dataset and new functions calculated. Discrimi- | nation accuracy was reduced to 97.8% (4 misclassifications out of 196). A discriminant — score plot showed that the ‘type C’ specimens | were interpolated between S. tragus and ‘type B’. (Fig. 5), due to their intermediate-width minor | wings and mid-length anther sacs. The four misclassifications were three specimens previous- | ly classifying as S. tragus (‘type A’) that now — classified as ‘type C’, and one a known ‘type C’ ~ that re-classified as S. tragus. The stepwise addition of a fourth potential discriminator, the | major tepal-wing width, regained 100% accurate classification. Although the major tepal-wing | width and the overall winged-fruit diameter were correlated overall (r? = the overall winged-fruit diameter, the major wing is broad in S. tragus and S. paulsenii, while it is — 0.90), the correlations | were different among the four groups. Relative to | 2008] 00°0Z1 HRUSA AND GASKIN: SALSOLA TRAGUS AND SALSOLA AUSTRALIS 119 4 Salsola ‘type C’ Salsola ‘type B’ genotype 3/3 WH Salsola ‘type B’ genotype 4/4 ¥%& Salsola ‘type B’ undetermined genotype California distribution of 2/3 Salsola ‘type B’ S jo) O° j=) CALIFORNIA 2/5/4 N 100 KM SCALE 500 KM Fic. 3. Distribution of Salsola ‘type B’ and Salsola ‘type C’ in California and the US southwest. California specimens of known genotype are individually plotted. Arrows indicate the single known genotype for each population of ‘type C’. The dark line surrounds the known distribution of ‘type B’. Sa/sola tragus occurs throughout California; only along the immediate coast, the high Sierra and in the Klamath/Siskiyou regions is it rarely encountered. Outside of the outlined region Salsola ‘type B’ distribution is imperfectly known. narrower in ‘type B’ and ‘type C’ (compare Figs. 6C and 6D; also Fig. 8). After completion of the above analyses, the 25 individuals of known genotype but not used for the function calculations, were scored for the above four discriminating characters. Their discriminant scores are plotted in Fig. 7 and the specimens are listed in Table 1 and Appendix III. Among these, five S. tragus and one ‘type B’ misclassified as ‘type C’. Among those with S. tragus genotypes, Moran 21152 from Baja California had most of the features and the habit of Salsola tragus sensu stricto (see Table 3) but had unusually spatulate minor wings, similar to those of ‘type C’. Hrusa 16172 (Kern Co.) was thought perhaps an F' hybrid as it had relatively narrow minor wings, a smaller maximum calyx appendage diameter, and a more condensed than usual inflorescence for ‘type B’, but it had a 3/3 ‘type B’ genotype. This individual occurred where both Sa/lsola tragus and ‘type C’ were mixed with ‘type B’, but a second plant of similar appearance was found mixed in a nearly pure population of ‘type B’ in the San Joaquin Valley north of Fresno, and it may be that this is simply an unusual ‘type B’ variant. The four other misclas- sified specimens were similar in general appear- ance to ‘type C’ yet contained only S. tragus haplotypes. Akers 65-1, 56-1 (Kern Co.) and 42-/, 42-3 (Stanislaus Co.) approached ‘type C’ in tepal-wing form, the relative persistence of mature fruit and the degree of wing development from plant base to summit, differing from ‘type C’ only by their slightly longer anther sacs. [Vol. 55 sovyjd ur jusyststod — /+ opIm UY} JoT[e} ‘Ie[NSoLI1 jenuue AyLINyeUL 1B sastosqe poeiq jusoe[pe UeY} 19}10Ys Bore QGii-1oJyUur pue QI Wojs a}eT[Ided wur 6¢'0 = x uu 39°0-tV'0 Iepnsoi AYsys 0} ‘yJoours ouTyM A[aIVI O} Ies]O ‘JUsONySUeI} ~ MADRONO jueld jnoysnoiy} UOSBIS SULIOMOTJ JO pus PoISSAUI JOU ‘pasodxa jINIy asoulds ‘snoy[Aydo.syos Ieouly A[MOIIeU IO a}a;Osqo wu /°g = x LOL9 9 sovyd ul yUd\sISIod —/+ opm ueY) AD][e) “eNO O} Ie[Nso1 Jeltuustad poalj-J10Ys 0} [enuue AVLINJeU 7B SasIosqe yov1q yusoe[pe uevy} JosUuOo] ‘ud3 SQII UWId}s SOULOYSI} JOSUOT YM pidsiy Ajorey “WU 7°) > SOULOYSLI} YM IO snoIgERys Wu so 0 = = wut /'0-Sv'0 Ie[nso1it APWYSISs 0} Yous yuid yyst] 07 o1yM ‘uDON[sSUevI}-TUIaS 07 onbedo yuvjd jnoysnoiy} o1nj}eul SIOMOTJ [ENPIAIPUI Se ‘°d'T ‘UOSBOS SULIOMOTJ JO SUIUUIZOQ POISSAUI JOU ‘pasodxoa jINIy osoulds jou ‘jos ‘ourpeAy 318A0gO A[PpeoIg wu 9°9 = Xx WU 6'L-8'V sould ul JUa}sIS1od —/+ 0} SUTTQUIN} stUOS ‘Te SB OpIm se —/+ ‘uodo "Udd ‘Ie[NBIIII 0} Ie[NsaI jenuue }U9}SISIod-1uW9s yoRig jUus0e[pe uvy} I9sUOC] 0} 19}10Ys Sq wiajs A[}SOU snoiqe]s — /+ 0} pidsty Ajosieds 6 Os CE SO Iepnsoii ATYSITs yuid ISI] 01 9114M ‘onbedo SOpouU sd1Y4}—-OM} jeseq }daoxa Jnoysnoiyy uoseas SULIOMO]J JO I[Pprut pojsoaur A[daop }0u ‘posodxo Apjsour yinay asourds jou ‘jos ‘ourpeAy o}e[nyeds 0} 9]78A0qgO A[MOLIeU UL 2 = mur | 8-97 poomoayquin} [[e@} Wey} JspIMm ‘uds ‘Ie[NsoI jenuue yueld uo jua\sisiad jov1g yusoR[pe ueY} 19,10Ys SqII UI9}s A[}SOUW snoiqe[s — /+ JO ‘WU 79 < sawoyory} yum pidsiy Ajasuap 07 Ajasieds PO CO Gur ¢’I-9'0 YyOousS A[aIVI 0} Ie~NSai A[poysreu por/yuid 0} 9114m ‘anbedo yueld jo wed ¢/] 01 S/p taddn~ UOSBdS SULIOMOTJ JO pua pojsoautr A[doop ynay osoulds jou ‘1yos ‘ourpeAy onda yuNTG 0} IvoUTT WU 6p = xX wu 786°C IOTABYOIQ JUDOSOUDS-]SOdg odeys jurl[g uedsojl] URI g IOIAevyoq WI sInyepy YIsus] 9pousgjul SuNnmM1y (juosoid usyM) uontsod o1njyWs3a Was OINJSSA UW91§ ysug] 1oylUy UISIeU yUWUUNs BuIM-[edo] (YsolJ) JOJO 2 Aytoedo Suim-jeday W1d}S UO UOTIISOd JINIJ-posurAA usoyed JUsWIdO[SANp JINAJ-Pposul AA, diysuoneyjor joviq/in1{ jeda} sunini{ adeys SUuIM-IOUIY SUIM “[OUT 1d}DUIVIP INI luasjnpd pjosjps dh as ‘ ax x & ia} Confirmed S. tragus genotypes nie a o 8 e - x ° Confirmed ‘type B' genotypes Discriminant Score 2 Discriminant Score 1 FiG. 5. Four-group discrimination plot for the Fig. 4 ‘core’ groups plus ‘type C’. Scores are based on four characters—the three Fig. 4 characters plus major-wing width. The ‘type C’ specimens were of known haplotype, and ‘type C’ from the common-garden is displayed separately from wild ‘type C’. Confirmed genotypes are plotted separately from specimens identified morphologically; these are juxtaposed com- pletely, indicating the accuracy of the morphological discrimination. Discriminant classification was 100% accurate with these morphological characters. HRUSA AND GASKIN: SALSOLA TRAGUS AND SALSOLA AUSTRALIS 121 08 09 07 08 ne 07 =a 06 B a: a 2 = 05 oar 2 2 = 04 £ & = o 9 9 = No aw S. tragus co— o 9 9° = nN w +om— iH ‘type B type C S. paulsenii S. tragus type B “type C" S. paulsenii 0.9 0.9 08 08 0.7 07 2 06 sg 08 & % o Q 05 = 05 5 § w 04 5 0.4 § 3S s £ 03 0.3 - q 0.2 0.2 0.1 0.1 0 0 2 & b ¢F $ @ ob ¢€ a o a o £ 2 ® gL G 2 2 2 oF PB § ae eee 7) “ Fic. 6. Univariate comparisons among the four features discriminating California Sa/sola tragus, S. ‘type B’, S. ‘type C’ and S. paulsenii. Cross bar within boxes represents the mean; boxes equal one standard deviation from the mean; T-bars indicate the range. Actual sizes are scaled to between 0 and 1 using Gower’s ranging algorithm (Sneath & Sokal 1973). County in the northern San Joaquin Valley, were found isolated from the other known ‘type C’ populations, and Akers 65-1] was collected in the far western Mojave Desert, an area where neither ‘type B’ nor ‘type C’ are established. Similar specimens have been seen from Washington State and the Modoc Plateau. That this is a recurring form could support the hypothesis that it represents an unrecognized polyploid. Plants of ‘type C’ morphology (of any genotype) appear to be relatively rare and, with the exception of the first known collection, have been found only when looked for specifically. While Salsola tragus is readily separable by eye from ‘type B’, morphologically separating this putative hexa- ploid form of S. tragus from the similar ‘type C’ 1s generally unreliable. Quantitatively ‘type B’ is recognizable by its combination of laterally expanded minor wings and short anthers (+/—0.5 mm) (Fig. 4; Table 1). Its winged fruits are not invested behind the ° r O ; eee 5 2 ho o & ae O ne | Oo % S = @) 3S o Oo fe) 2 Qo £ O00 o\O oO = O fe) 5 ca i 0 a ri 7 O S. tragus 'type A' 2 rN A O ‘Type B' yy A dp ‘Type C' a ee Aa, A S. paulsenii 4 @ Test, S. tragus genotypes A A @ Test, ‘type B' genotypes - x Test, 'type C' genotypes 6 T T T ails ea | a | T T 1 6 -4 -2 0 2 4 a 8 10 12 Discriminant Score 1 Fic. 7. Plot of scores for 25 test specimens of known genotype using the four discriminant characters in Fig. 6. Specimens classifying morphologically outside their genotype group are labeled as follows: A) Akers 43-3, genotype 1/2 (‘type A’). B) Akers 42-1, genotype 2/ 5 (type A’). C) Akers 56-1 genotype 2/5 (‘type A’). D) Akers 65-1, genotype 1/2/5 (‘type A’). E) R. Moran 21152, genotype 1/2 (‘type A’). F) Arusa 16172 genotype 3/3 (‘type B’). G) M. Rejmanek SA1, Republic of South Africa, genotype 3/3 (‘type B’). H) C. Borger 01, Australia, genotype 3/3 (‘type B’). Collection details in Appendix ITI. subtending bract, and they are non-persistent. The stem vestiture usually appears glabrous, but at higher magnification, (+/—60X), the stem ribs may have sparse and nearly microscopic (<0.2 mm) epidermal trichomes that are absent from the inter-rib areas (Fig. 9B). Trichomes like these are not known in S. tragus, S. paulsenii or ‘type C’. “Type B’ is also unique within Salsola sect. Kali in that individuals sometimes overwin- ter and resprout the following spring. These plants do not, or rarely, grow into full size individuals, but their capacity to live past a single season 1s unique in Salsola sect. Kali. Among ‘type B’ with 3/3 or 4/4 genotype there were no morphological correlations to either combination. However, our sample of ‘type B’ with known genotype is not large, and corre- sponding physical or behavioral patterns may not yet be visible. Short anthers as in ‘type B’ are also found in Salsola paulsenii (see Figs. 6B and 8), but that species combines reduced or obsolete minor wings with a broader overall winged-fruit diam- eter and a stem epidermal covering of dense, short, +/— columnar papillae on both the stem ribs and inter-rib areas (compare Fig. 9B and 9D). Its variability centers on the minor wings; these may be both obsolete, one may develop a narrowly oblanceolate wing, or rarely both may MADRONO be so developed. Its most prominent identifying | feature is the sharp spine which, at maturity, develops from the perianth tips (Fig. 8). Among these russianthistles, Salsola tragus is the most polymorphic. It has anthers often twice the length of those in ‘type B’ or S. paulsenii | (Table 3; Figs. 6 and 8); it has reduced, but not obsolete, minor tepal-wings and a stem rib vestiture of long trichomes (Figs. 8, 9A), or none at all. Individuals may have fruiting tepal-wings similar in diameter to those of ‘type B’ or be half | that size, produce fully winged fruits at all but the | lowermost nodes, have them only at the upper- most nodes, or show a continuous enlargement of | the wings from stem base to summit. Its tepal- wing margins are generally irregular and the - major wing commonly has a notched tip. — Wingless fruits are obscured by the subtending | bract, and winged fruits are closely appressed beneath the bract and against the stem rachis. All are persistent after maturity. ee : : : | As to variation among S. tragus in California, one is a coarse-textured, short-statured, dense, suff, spiny and rounded tumbleweed with broad | calyx wings present on fruits at nodes well down | the stems; the winged-fruits are invested within the bracts and are not exposed as in ‘type B’. The | lowermost fruits are consistently wingless. This is the dominant form in the high and low deserts and mountainous areas. It is not generally sympatric with ‘type B’. The second form is a low-elevation, late-fruiting, relatively leafy plant — with small and irregular calyx-wings formed only on the top 1/5 or less of the stems; sometimes ' [Vol. 55 | ——— = there are no winged fruits formed at all. It is. common in the low elevation Central Valley, and is scattered in the Salinas Valley and South Coast region. In all these areas, it may be sympatric with ‘type B’. These distinctive and common extremes probably represent part of what is actually continuous variation in a variable— species. The winged fruits of these two types are compared in Fig. 8. Characterization of Salsola tragus sensu stricto is further complicated by the presence of plants that bridge essentially the entire morphological gap between S. tragus and Salsola paulsenii. This intermediate is mostly in transmontane regions, and is rather common in both the field and herbarium, where specimens are identified equal- , ly frequently as S. tragus or S. paulsenii. It 1s found throughout the intermountain west, the: Sonoran and Mojave deserts, and has a disjunct occurrence in the southern San Joaquin Valley, — centering in the disturbed areas about the oil fields near Taft (Kern Co., CA). Their interme- diacy is most apparent in the usual presence of stem papillae combined with a non-spinose or ‘lax’ mature perianth (see Mosyakin 2003), similar to S. tragus or ‘type B’. The fruit wing texture is variable, but commonly approaches the > 2008] Salsola ‘type C’ FIG. 8. HRUSA AND GASKIN: SALSOLA TRAGUS AND SALSOLA AUSTRALIS 123 Salsola paulsenii SCALE A — 0.5 mm B 5 mm Tepal-wing, mature tepals, anther, and stem vestiture comparisons among Salsola tragus, S. ‘type B’, S. ‘type C’ and S. paulsenii. Anthers (A) and overall tepal-wing diameter (B) are scaled to facilitate comparison. A: mature anther. B: winged fruit from above. C: perianth form at fruit maturity. D: cross-section of stem showing approximate trichome size and position. E: side view of stem surface. F: lower node un-winged mature fruit in Salsola tragus and Salsola ‘type C’. hyaline form found in S. paulsenii. Sometimes longer trichomes, like those in Salsola tragus and absent from S. paulsenii, are mixed with the papillae typical of S. paulsenii, while less fre- quently these papillae are absent and the S. tragus type trichome is exclusively present. Their growth habit is also more like the conical form of S. paulsenii than the rounded form of S. tragus. Mosyakin (2003) considered them a match to Salsola gobicola Ijin from central Asia and similar plants hypothesized as derived from hybridization between S. tragus and S. paulsenii 124 MADRONO FIG. 9. Salsola tragus, S. paulsenii, S. ‘type B’ and S. ‘type C’ stem trichomes. A: Salsola tragus. B: Salsola ‘type | B’. C: Salsola ‘type C’. D: Salsola paulsenii. Trichome density varies considerably among and within individuals. | For illustration clarity trichome density displayed in Salsola ‘type B’ is greater than is typical; the remainder | are representative. are known from throughout their contact zone in that region (Rilke 1999). Its ploidy level (or levels) in North America is not known, nor is it known if this form is autochthonous here, in full or in part. The Identity of ‘Type B’ Among the types and regular specimens examined and photographed by S. L. Mosyakin at LE, several appeared superficially similar to ‘type B’; examination of the digital images indicated that all of these were forms of S. tragus sensu stricto, usually plants with slightly broader than typical minor wings. Mosyakin examined specimens from central Asia, southeast Middle. Asia, Tajikistan, Kyrgyzstan, and SW Kazakh-. stan, but none matched ‘type B’. Material at. MPU from southern Europe was likewise refer- | able to either S. kali sensu stricto, or S. tragus | sensu stricto as were all host plant biocontrol. vouchers from Europe, Asia and North Africa held by USDA in Albany, CA or deposited at CDA. Only Australian and South African specimens photographed at K or observed at the University of California Herbarium (UC) matched ‘type B’ closely. “Type B’ was compara- | ble to the holotype image of Salsola kali subsp. austroafricana from South Africa at Munich (M), and the BM lectotype of S. australis also matched | 2008] ‘type B’ closely. It was clear that both represent Salsola ‘type B’, matching in all visible charac- teristics, particularly the broad minor tepal-wings and the exposed winged fruits present at both upper and lower nodes. The Salsola australis lectotype had also short anthers and a sparse, near microscopic, vestiture like ‘type B’ (Table 3, Figs. 8 and 9), but these latter characters could not be seen on the image of S. kali subsp. austroafricana. Specimens at K indicated that S. australis is widespread in southern Australia. Our sequencing of two distinctive Salsola forms from that region (C. Borger 03, 18 and 05 PERTH) revealed both were genotype 3/3 (‘type B’). Thus, there are two distinct Sa/sola species in California to which Salsola tragus L. has been applied; genuine Salsola tragus L. and Salsola australis R. Br. Salsola australis (‘type B’) was reported in Gaskin et al. (2006) as S. kali L.subsp. austroafricana Aellen but that report was produced before the type of S. australis had been seen. Sa/lsola australis R. Br. has nomencla- tural priority (Prodromus Florae Novae Hollan- diae 411. 1810) and therefore S. kali L. subsp. austroafricana Aellen (Mitteilungen der Bota- nischen Staatssammlung Munchen 4. 1961) is a taxonomic synonym. Salsola australis California distribution.—Salsola australis is found throughout the San Joaquin Valley, the South Coast Ranges, the south coast, all the Channel Islands, and is apparently localized in the Colorado Desert. These areas all have warm winters without regular hard frost. Ryan and Ayres (2000) cited unspecified specimens at The California Academy of Sciences Herbarium (CAS) documenting S. australis in the inter- mountain region, but they apparently mistook intermediates between S. tragus and S. paulsenii (=S. gobicola) for S. australis. Both S. australis and S. gobicola have fruits with wings at some or all lower nodes, but as discussed previously, differ in other characteristics. In the southern Sacramento Valley Salsola australis may be near the northern margin of its potential range, and populations there appear ephemeral. It is scattered along highways from about the latitude of Monterey southwards, with largest contiguous populations in the interior South Coast Ranges from southern Monterey to San Luis Obispo Cos. North of Santa Barbara, Salsola australis, like S. tragus, appears absent from the immediate coast. Southward, S. aus- tralis occurs right up to the intertidal zone. Salsola australis plants on Santa Catalina and the northern Channel Islands are similar to mainland plants, while collections from San Nicolas and San Clemente Islands are distinctive in appear- ance. The latter are more succulent, densely leafy and have smaller and fewer winged-fruits that are HRUSA AND GASKIN: SALSOLA TRAGUS AND SALSOLA AUSTRALIS 125 hidden within the fleshy leaves and thus possibly not so readily dispersed than are those of the mainland forms. Although the voucher specimen has not been found, the report by Mosyakin (1996, and cited in Hrusa et al. 2002) of S. kali subsp. pontica (Pall.) Mosyakin on San Nicholas Island is probably also this form of S. australis. Among the mainland populations, there were two minor, more or less allopatric morphs of Salsola australis. Plants from the South Coast Ranges were semi-succulent, irregular in habit, brittle, often blue-glaucous in color, and tended to have smaller winged-fruit diameter than did plants from the San Joaquin Valley and south coastal region, although not so small as those on San Nicolas and San Clemente Islands. Uncor- related to these two morphs was a minor bimodal distribution of the minor-wing width that was also geographically uncorrelated. Distribution outside of California.—As far as currently known, in western North America outside of California, herbarium specimens of Salsola australis have been seen from Arizona, southwestern Texas and low elevations in Mex- ico. A survey of specimens from the remainder of North America has not been undertaken, and, given the confusion with S. tragus, it is clear that its non-California North American distribution is scarcely known. Outside of North America, we have seen specimens from Australia, Namibia and the Republic of South Africa. Because species of Salsola sect. Kali have been thought indigenous only to the northern hemisphere (Mosyakin 1996; Rilke 1999), its presence in the southern hemi- sphere was assumed adventive (Botschantzev 1974; S. L. Mosyakin, Kholodny Insititute of Botany, Kiev, Ukraine, personal communica- tion). This led to the interpretation that S. australis was a form of Eurasian S. tragus. Examination and molecular evaluation of addi- tional Australian and southern Africa Salsola materials would be useful, both to confirm the presence or absence of Sa/sola tragus sensu stricto in those regions, but also to elucidate the diversity of S. australis genotypes there. Our limited sampling revealed only haplotype 3 in African and Australian S. australis. However, if the taxon had a long association with one or more of those regions additional haplotypes would be expected, including haplotype 4, common in North Amer- ican S. australis, but not yet found elsewhere. Australian specimens at K appear to represent both morphological forms of S. australis found in California. The lectotype compared most favor- ably with succulent material from the California inner South Coast Ranges, while the holotype of S. kali subsp. austroafricana 1s apparently more like the San Joaquin, South Coast, and Arizona forms. In addition, two Australian specimens (C. 126 Borger 06, 18 PERTH) of what has been called S: australis var. strobilifera (Benth.) Domin, a plant distinct from S. australis sensu stricto in its densely congested ‘strobilus’-like inflorescences, but with similar tepal-wings, had genotype 3/3. A third Australian form (C. Borger 05 PERTH), was a prostrate sand dune plant, similar in habit to S. kali from European and North African coastal sands. This genotype 3/3 plant has calyx wings similar to S. australis in form and texture, but with narrower minor-wings. It is of note that in Flora of Australia (Wilson 1984) the ‘Salsola kalv illustrated is S. australis, which is listed as a synonym, while S. tragus is not mentioned nor listed in the synonymy. Our untested hypothesis is that Salsola sect. Kali of Australia represents a disjunct species complex and S. australis as treated here is just the typical form. Habitat differences.—Sa/sola australis occurs in California from sea level to about 400 m, and reaches at least 500 m in Arizona. Contrary to Salsola tragus, which is common in the Mojave Desert, S. australis is there only a rare casual. It has recently been collected in the low warm desert of California near the Mexican border (R. Riefner 07-64 CDA), and the few specimens so far seen from Mexico were also collected at low eleva- tions. Above about 400 m in California Salsola tragus replaces S. australis; in general Salsola tragus’ adaptation seems to favor more severe environments and its widespread occupation of interior North America (see Mosyakin 2003) is further evidence. Distribution (and often morphological) pat- terns within or among widespread indigenous taxa commonly correspond to some ecologic or geographic condition, while widespread adven- tive plants rarely display such coherence, occu- pying more or less disturbed areas in many types of habitat. Their distributions often reflect better the relatively random movement of specific, usually human, vectors. Salsola tragus and Salsola australis overlap in range and are often sympatric at low elevations, but when growing adjacent they generally occupy different habitats. For example, in favorable years on the open shale hills of the South Coast Ranges east of San Lucas in Monterey Co., Salsola australis occupies thousands of acres with S. tragus restricted to patches along valley bottom fences and immedi- ate roadsides. Some of the microhabitat separation may be the result of different dispersal mechanics. The fruits of S. tragus are invested by the subtending bract, tightly appressed to the rachis and persis- tent after maturity. The mature fruits do not abscise readily and are dispersed as the plant tumbles, or as it decomposes post-tumbling. This pattern has been documented for Salsola tragus (reported as S. iberica) in the Great Basin MADRONO [Vol. 55 (Stallings et al. 1995). Conversely, the fruits of Salsola australis generally abscise readily at maturity (see Table 3), and the plant skeletons do not tumble to any extent. The result is that Salsola australis is often found on hillslopes and steep roadcuts where the parent plants grew. This dispersal behavior may also hold for Australia, where, according to C. Borger (University of Western Australia, Perth, personal communica- tion) their ‘russianthistle’ does not tumble. Commonly one can also recognize the two species when pressed and dried by the accumulated pile of loose separated fruits in mature S. australis, something not seen with Salsola tragus specimens. A Name for “Type C’ All but four of the individuals morphologically referable to ‘type C’ had combined S. tragus and ‘type B’ haplotypes. Those four with S. tragus genotypes only, have been discussed above. The remaining ‘type C’ individuals with three different haplotypes were almost certainly hexaploid with two chromosome sets from tetraploid S. tragus | and one from diploid ‘type B’. Without additional | data it could not be confirmed that the Sutter County genotype 2/3 and Coalinga genotype 1/4 | plants were also hexaploid; their Sa/sola tragus , parents may have been tetraploid homozygotes and thus only two distinguishable haplotypes are | present in the hybrid. In our view, genuine ‘type C’ is a fertile | allohexaploid derived via hybridization between diploid Salsola australis and tetraploid Salsola | tragus. These allohexaploid derivatives would be | inter-fertile among themselves, yet inter-sterile to | the parent diploids or tetraploids. At least two | persisting, true-breeding populations of ‘type C’ | are present in California and appear to represent at least a locally adapted entity. This inter- | fertility among ‘type C’ individuals and the | offspring sterility expected if a hexaploid is backcrossed to tetraploid S. tragus or to diploid © S. australis supports the hypothesis that “type C’ is a biologically distinct species. It has originated | at least three times, with each instance a_ formation of the same species. We hereby propose the following specific | epithet be applied to ‘type C’. | Salsola ryanii G. F. Hrusa and Gaskin sp. nov. © (Fig. 8). TYPE: USA, California, Kern Co., . Hwy 119, 0.5 mi W of Old River. Genotype 2/ 5/4, 35.267°N, 119.153°W. El. 315 ft. 01 Nov. ff 2002, P. Akers 69-3 (Holotype CDA; isotypes — DAV, RSA). | Salsola ryanii sp. nov. Planta annua usque ad 2m alta. Caules cylindrici, sparse hispidi vel glabri. Fasciculi collenchymatis plerumque vir- ides, interdum rubentes. Fructuum alae tepalum 2008] albae vel subroseae, venis visibilibus sed non prominentibus. Alae minores tepalum plus min- usve spathulatae, infra angustatae supra expan- sae, limbo ungue latiore. Fructus, alis inclusis, 5.3-7.2 mm in diametro. Fructus superiors alati, basales partim alati vel non alati. Tepala in maturitate hyalina remanentia, mollia. Fructus in maturitate semipersistens, bractea subtendenti partim investitus. Antherae 0.5—1.3 mm longae. Annual plant up to 2m tall. Stems cylindric, sparsely hispid or glabrous. Stem ribs usually green, sometimes reddening. Tepals in maturity remaining hyaline, soft. Wings of tepals of fruits white or pinkish, with veins visible but not prominent. Minor-wings of tepals more or less spatulate, below narrowed, above expanded, with limb broader than claw. Fruit, wings included, 5.2-7.2 mm in diameter; upper fruits winged, basal partly winged or not winged. At maturity semipersistent, partly invested by subtending bract. Anthers 0.5—1.3 mm long. The epithet honors Fred Ryan of the USDA/ ARS, who first collected and recognized this entity. Visual recognition Salsola ryanii. Figures 8 and 9 illustrate tepal-wing, anther, and stem vestiture morphologies and Table 3 compares taxonomi- cally useful morphological traits of S. ryanii (‘type C’), S. australis, S. paulsenii and S. tragus. Salsola ryanii 1s morphologically intermediate between S. tragus and S. australis in nearly all features, and this intermediacy makes it a challenge to recognize on its own. Adding to the difficulty are those few S. tragus plants that are similar morphologically to S. ryanii. The general texture of ‘type C’ is difficult to assess on a dried herbarium sheet; it is a thinner-stemmed and less brittle plant than the often thick- stemmed and semi-succulent S. australis, and more brittle than the wiry and tough Salsola tragus. It does not form the long, dense, spiciform inflorescence common in mature Salsola tragus, nor does it have the loose, more or less uncondensed inflorescence of most Sa/lsola aus- tralis. When growing side by side with S. tragus, Salsola ryanii may be distinguished by the relatively long internodes and winged fruits at lower nodes. Identification is improved by becoming familiar with the parents. Distinguish- ing genuine ‘type C’ plants from hexaploid S. tragus failed using our discriminant characters, as discussed previously. The latter are apparently less hispid and more robust than S. ryanii, but further study will be necessary before these can be reliably separated using morphology alone. How- ever, they are readily distinguished using molec- ular markers. Distribution. Salsola ryanii is known from three localized small to extended medium-sized popu- HRUSA AND GASKIN: SALSOLA TRAGUS AND SALSOLA AUSTRALIS 127 lations (Fig. 3A). All occur below 300 m and occur where Sa/sola australis and Salsola tragus are sympatric. Two populations are along regional highways; at the intersection of Hwy. 166 and Hwy. 33 near Maricopa (Kern Co.) and alongside Hwy. 33 immediately north and south of Coalinga (Fresno Co.). In both locations Salsola ryanii grows mixed with Salsola tragus. The largest populations are along Highways 166 and 119 from immediately east of Maricopa east to Bakersfield (Kern Co.). It has also been collected near Famoso and Shafter in that region. The area from south of the Fresno Co. line, north of Hwy. 166, east of the Elk Hills and west of approximately Interstate 5 has scattered small to medium-sized patches of S. ryanii. In this region, it is sympatric with S. tragus, S. australis and sometimes S. gobicola, although the two latter are often growing on sands or shales while Salsola ryanii is usually on the deeper soils S: tragus occupies. The two small populations in Sutter County have not been relocated, and as with the S. australis in that area, may be ephemeral. The haplotype combinations in Sal- sola ryanii are geographically localized and point to regional dispersal from either a local origin or introduction; as all these sites are along or close by roadsides, it is conceptually easy see their dispersal as predominantly human-mediated. Will it spread? Most novel hybrids are not selectively advantageous, but evidence is growing that hybrids or novel genotypes are involved in many invasions (Ellstrand and Schierenbeck 2000). Infamous recent examples include the marsh grass Spartina spp. (Ayres et al. 1999), the freshwater aquatic plant Myriophyllum spp. (Moody and Les 2002), saltcedar shrubs (Ta- marix spp.) (Gaskin and Schaal 2002), the marine alga Caulerpa spp. (Durand et al. 2002), and a mustard (Rorippa spp.) (Bleeker 2003). In addi- tion to potentially stimulating invasiveness, novel hybrids also may affect control efforts by presenting phenotypes that are novel to potential classical biological control agents. Analysis of invasion identities and population dynamics using genetic markers has contributed much information about species that are biological control targets (Sakai et al. 2001; Roderick and Navayjas 2003). Abbott (1992) summarized four known allo- polyploid speciation instances, two between pairs of non-native taxa and two between native and non-native taxa. Both instances of non-natives crossing with other introductions are in North America; two species of Tragopogon, T. dubius and T. pratensis, have given rise to 7Tragopogon miscellus Ownbey (Amer. J. Bot. 37: 498. 1950) that has spread beyond the range of one parent. The other, 7. mirus Ownbey (Amer. J. Bot. 37: 498. 1950), a product of hybridization between 7; 128 MADRONO dubius and T. pratensis, has not spread beyond the vicinity of its parent species. In Great Britain, Senecio cambrensis Rosser (Watsonia 3: 228. 1955) has been derived at least twice between the native Senecio squalidus and introduced Senecio vulgaris, and is always found where both parents are present. Salsola ryanii appears to behave similarly to the 7ragopogon mirus and Senecio models; it has originated at least three times, but each known independent derivative remains in the region of the parent species. In this case, although Salsola sect. Kali travels readily in concert with humans and the current populations of Salsola ryanii may be distant from their sites of origin, they still occur within the parents adaptive range. Salsola ryanii may be restricted in distri- bution at least in part due to its intermediacy in the differing ecological adaptations characteriz- ing the parents; fruits do not abscise readily, nor are they persistent; plants tumble to some small extent, but are not dense and orbicular like S. tragus. We feel, based on its current distribution pattern of isolated or extended roadside patches that S. ryanii will likely spread only locally by non-anthropogenic means, but will attain longer distance dispersal via human traffic. Its present adaptive deficiencies will, to some extent, prob- ably restrict its subsequent movement, and the species will unlikely become a widespread future pest. ACKNOWLEDGMENTS We dedicate this report to Dr. Sergei L. Mosyakin, whose encouragement and kind words in addition to providing photographs, specimens and herbarium data from Russia and the Ukraine, both began and finished this study. Dr. Mike Pitcairn and Dr. Pat Akers of the Integrated Pest Control Branch Biocontrol Program of the California Dept. of Food and Agriculture were also instrumental in the first author’s access to Salsola materials. Dr. Pitcairn arranged for and managed the common garden, while Dr. Akers performed field surveys both alone and in the author’s company. His keen eye was responsible for most Sa/sola ryanii finds. Molecular analyses were supported in part by funding from the U.S. Department of the Interior, Bureau of Land Management and Bureau of Indian Affairs. Molecular work was performed by K. Mann, J. Lassey, and J. Londo. Additional USDA collaborators includ- ed Dr. Lincoln Smith who arranged for Old World vouchers collected during surveys for biocontrol agents to be routed to the first author, collected specimens himself, performed herbarium surveys and _ photo- graphed Old World specimens. Dr. Fred Ryan provided photographs of Russian specimens, translated German text, and freely shared his Sa/so/a materials and data. M. Irene Wibawa provided data to the authors ranging from greenhouse grown Asian materials to photographs from Kew. Dr. Catherine Borger (PERTH) made available several useful Australian specimens. Curators at the following herbaria provided photographs, specimens or welcomed the authors’ or collaborators visits; BM, CAS, DAV, K, KW, LE, M, MONTP, PERTH, RSA, and UC. R. K. Riefner collected [Vol. 55 specimens in southern California, studied herbarium | material and arranged for the first author’s access to | specific RSA specimens. Dr. Hans-Joachim Esser of M kindly provided the holotype photograph of S. k. ssp. austroafricana. Scott Kinnee of CDFA Plant Pest | Diagnostics handled many technical presentation issues | and walked the first author through the SEM photog- | raphy. A special thanks to Genevieve Walden for her extra effort in providing the line drawings. Thank you also to Dr. Mark Garland for the Latin diagnosis. Additional help came from J. Naughton, B. Villegas, Dr. W. Bruckart, Dr. R. Price, Dr. M. Rejmanek, M. O’Brien, Dr. D. Ayres, and numerous others to whom | we apologize for forgetting them at this moment. Thank — you also to the anonymous reviewers and especially to | Madrono editor Dr. John Hunter for his tolerance and patience. LITERATURE CITED ABBOTT, R. J. 1992. Plant invasions, interspecific hybridization and the evolution of new plant taxa. Trends in Ecology & Evolution 7:401—404. AYRES, D. R., D. GARCIA-RossI, H. G. DAVIS, AND | D. R. STRONG. 1999. Extent and degree of hybridization between exotic (Spartina alterniflora) | and native (S. foliosa) cordgrass (Poaceae) in — California, USA determined by random amplified | polymorphic DNA (RAPDs). Molecular Ecology | 8:1179-1186. BLEEKER, W. 2003. Hybridization and Rorippa aus- triaca (Brassicaceae) invasion in Germany. Molec- ular Ecology 12:1831—1841. BOTSCHANTZEV, V. P. 1974. A synopsis of Salsola | (Chenopodiaceae) from south and south-west Africa. Kew Bulletin 29:597—614. BRUCKART, W. L., C. A. CAVIN, R. SOBHIAN, L. | VAINA, I. SCHWARCZINGER, F. J. RYAN, AND S. HASAN. 2004. Differential susceptibility of Salsola tragus to Colletotrichum gloeosporides and Uro- | myces salsolae. Biological Control 30:306—311. DURAND, C., M. MANUEL, C. F. BOUDOURESQUE, A. | MEINESZ, M. VERLAQUE, AND Y. LE PARCO. 2002. Molecular data suggest a hybrid origin for the invasive Caulerpa racemosa (Caulerpales, Chloro- | phyta) in the Mediterranean Sea. Journal of | Evolutionary Biology 15:122—133. ELLSTRAND, N. C. AND K. A. SCHIERENBECK. 2000. Hybridization as a stimulus for the evolution of invasiveness in plants? Proceedings of the National Academy of Sciences of the United States of America 97:7043—7050. GASKIN, J. F., F. J. RYAN, G. F. HRUSA, AND J. P. LONDO. 2006. Genotype diversity of Salsola tragus | and potential origins of a previously unidentified | invasive Salsola from California and Arizona. Madrono 53:246—253. AND B. A. SCHAAL. 2002. Hybrid Tamarix widespread in US invasion and undetected in native Asian range. Proceedings of the National Academy | of Sciences of the United States of America. 99:11256—-11259. GOEDEN, R. D. AND R. W. PEMBERTON. 1995. Russianthistle. Pp. 276-280 in J. R. Nechols (ed.), Biological Control in the Western United States. Publication 3361, Division of Agriculture and Natural Resources, University of California, Oak- land, CA. 2008] Hrusa, G. F., B. ERTTER, A. C. SANDERS, G. LEPPIG, AND E. A. DEAN. 2002. Catalogue of non-native taxa occurring spontaneously in California beyond those addressed in the Jepson Manual. Madrono 48:61—-98. JEPSON, W. L. 1914. A Flora of California Part IV. H.S. Crocker & Co., San Francisco, CA. Moopy, M. L. AND D. H. LES. 2002. Evidence of hybridity in invasive watermilfoil (Myriophyllum) populations. Proceedings of the National Academy of Sciences of the United States of America 99:14867-14871. MOSYAKIN, S. L. 1996. A taxonomic synopsis of the genus Salsola (Chenopodiaceae) in North America. An- nals of the Missouri Botanical Garden 83:387—395. . 2003. Salsola (Chenopodiaceae). Pp. 398-403 in N. Morin (ed.), Flora of North America North of Mexico Vol. 4. Oxford University Press, Oxford, New York, NY. RAMBAUT, A. 1996. Se-Al Sequence Alignment Editor. Oxford, U.K. Available at: http://evolve.zoo.ox.ac. uk/software.html RILKE, S. 1999. Revision der sektion Salsola s.l. der gattung Sal/sola (Chenopodiaceae). Bibliotheca Botanica. Heft 149. RODERICK, G. K. AND M. NAVAJAS. 2003. Genes in new environments: Genetics and evolution in biological control. Nature Reviews Genetics 4:889—899. RYAN, F. AND D. AyRES. 2000. Molecular markers indicate two cryptic, genetically divergent popula- tions of Russian thistle (Sa/sola tragus) in Califor- nia. Canadian Journal of Botany 78:59-67. SAKAI, A. K., F. W. ALLENDORF, J. S. HOLT, D. M. LODGE, AND J. MOLOFSKY. 2001. The population biology of invasive species. Annual Review of Ecology and Systematics 32:305—332. SNEATH, P. H. A. AND R. R. SOKAL. 1973. Numerical Taxonomy. W.H. Freeman & Co., San Francisco, CA. SOBHIAN, R., F. J. RYAN, A. KHAMRAEV, M. J. PITCAIRN, AND D. E. BELL. 2003. DNA phenotyp- ing to find a natural enemy in Uzbekistan for California biotypes of Salsola tragus L. Biological Control 28:222—228. STALLINGS, G. P., D. C. THILL, C. A. MALLORY- SMITH, AND L. W. LAss. 1995. Plant movement and seed dispersal of Russian thistle (Sa/sola iberica). Weed Science 43:63—69. WILSON, P. G. 1984. Salsola (Chenopodiaceae). Pp. 313-316 in Alexander S. George (ed.), Flora of Australia, Vol. 4. Australian Government Pub- lishing Service, Canberra, Australia. APPENDIX I Specimens used in the morphometric analysis. Some of these are also cited in Gaskin et al. (2006). Specimens with genotypes determined specifically for this study are also listed in Appendix II. ““Common garden”’ collec- tion data are seed collection provenances; vouchers for the original parent material are not available. Offspring vouchers are at CDA except where specified. Common Garden Materials Salsola ‘type B’ USA: CALIFORNIA: FRESNO CO.: Fresno, HCRL RR tracks, 36°43.3’N, 119°44.1'W, Row A-6/ HRUSA AND GASKIN: SALSOLA TRAGUS AND SALSOLA AUSTRALIS 129 inds. 1-6; row B-6/inds. I—5:; row C-S/inds. 1-6: MERCED CO.: Santa Nella, water tower on Hwy 33, 37°4.3'N, 121°1.2’W, Row B-2linds. 1—5; row A-4linds. I-8. SAN DIEGO CO.: Mission Valley, Hazards Center Dr. off I-8. 32° 46.1’N, 117°9.2’W, Nov. 1999. Row A-S5/inds. 1—7; row C-4/ inds. 1-6; row B-7linds. 1-6. Salsola tragus (‘type A’) USA: CALIFORNIA: FRESNO CO.: Fresno, 30 acre field adjacent to HCRL RR, 36°43.3’N, 119°44.1'W, Nov. 1999. Row A-I/inds. 1-4; row B-5/ inds. 1-6; row C-10/inds. 1-4: Big Field in Fresno, 36°48.9'N, 119°44.1'W, Dec. 1999; row A-3/inds. 1-6: row B-Ilinds. 1-4; row C-Ilinds I1—5: Coalinga, Phelps Rd. at Hwy 33, 36°9.9'N, 120°21.2'W, row B-4/inds. 2— 4; row C-6/inds. 1-4; row A-Y/inds. 1, 3, 4. YOLO CO.:: Davis, Koso St. at Cowell Blvd. 38°32.1'N, 121°44.1’W, Dec. 1999. Row A-S/inds. 1-6; row C-3/ inds. 1-5; row B-9/inds. 1-4. Salsola paulsenii USA: CALIFORNIA: SAN BERNARDINO CoO.:: Barstow, downtown Business I-15 and E. White St. 34°54’01"N; 117°01'22"W, 650 m, row B-S/inds. 1-3; row A-2/inds. 1-2; row C-9/1 ind. Salsola ‘type C’ USA: CALIFORNIA: FRESNO CO.: Coalinga, Phelps Rd. at Hwy 33, 36°9.9'N, 120°21.2'W, row C- 2linds. 2, 3; row A-10linds. 2, 3; row B-3linds. 1, 3, 4. Wild Materials Salsola ‘type B’ USA: CALIFORNIA: FRESNO CoO.: Fuller 18732 § side Hwy 149 (Hwy 198) 2.1 mi. W of Coalinga. 7/24/ 1969. IMPERIAL CO.: RE. Reifner 07-64, disturbed alkaline flats on Ross Rd. ca. 0.2 mi. E of Dogwood St. N. of Hwy 8, nr. El Centro. UTM: (NAD 83) 11S 063755E, 362805 N. El. minus 28 ft. 02/17/2007. KERN CO.: Hrusa 16169, 1/4 mi. W of Maricopa on Hwy 166/ 33, 35.056°N, 119.405°W, 925 ft. el. 10/12/2003. Hrusa 16173 Hwy 166/33 10 mi. E of Maricopa. 35.058°N, 119.258°W. 600 ft. el. 10/12/2003. H.L. Green s.n., near Buena Vista Lake, 11/10/61. KINGS CO.: Fuller 12935 E slope Pyramid Hills, 14 mi. S of Avenal. 12/17/1964. LOS ANGELES CoO. B. Villegas B247-3; B247-4; B247- 5, all haplotype 4. Plant DNA#s in Gaskin et al. 2006. Calabasas nr. Ventura Co. line. 34°9.93'N, 118°40.67', 8/21/2002; M. O’Brien s.n. Figueroa Blvd., Los Angeles. 9/13/2004. M. O’Brien s.n. Pescadero State Beach. 12/ 23/2004. MONTEREY CO.: Hrusa 16182 RR tracks at Monterey Rd. on E side US 101 (Rancho del Salinas). 36.675°N, 120.694°W. 683 ft. el. 10/14/2003. Arusa 16183A. Plant DNA# in Gaskin et al. 2006; /6/83B; 16183C; 16183D. Plant DNA#s in Gaskin et al. 2006. All haplotype 3. Hwy 198 betw. San Lucas & Coalinga +/— 1 mi. E US 101. 36.149°N, 121.002°W. 492 ft. el. 10/14/2003. Hrusa 16188. Gonzales, W side US 101 at corner Folletta & Atta Rds. 36.524°N, 121.466°W, 100 ft. el. 10/14/2003. Hrusa 16193, haplotype 3. Plant DNA# in Gaskin et al. 2006. Jolon Rd. at Lockwood Post Office. 35.937°N, 121.072°W, 970 ft. el. 10/14/ 2003. SAN BENITO CoO.: Hrusa 16186, Hwy 25, S of Paicines, vineyard edge. 36.677°N, 121.256°. 782 ft. el. 10/14/2003. SAN JOAQUIN CoO.: Hrusa 16134 Hwy 33 exit 1/2 mi. S of Interstate 5. 37.672°N, 121.333°W. 130 MADRONO 75 ft. el. 10/10/2003. SAN LUIS OBISPO CO.: Hrusa 16179, Cuyama Cyn. along Hwy 166 betw. Cuyama & Santa Maria. 35.047°N, 120.183°W, 896 ft. el. 10/13/ 2003. Hrusa 16181, Templeton, US 101 1/4 mi. N of Paso Robles Ck. Bridge. 35.536°N, 120.711°W, 775 ft. el. 10/13/2003. Hrusa 16196, Hwy 46 E of Paso Robles, immed. E of Whitley Gardens. 35.659°N, 120.480°W 1050 ft. el. 10/14/2003. Salsola tragus (‘type A’) USA: CALIFORNIA: INYO CO: Hrusa 16709 CDFA IPC seed accession RTO5-3; 16710; Hwy 6 roadside, 6 mi. N of Bishop, 37°24'49’N, 118°21'11'W, 4100 ft. el. 9/27/ 2005. KERN CO: Akers RT-056-3. N of Hwy 58, int. Edison Hwy & Tower Line Rd. along RR. 35.332°N, 118.807°W, 794 ft. el. 10/30/2002. Hrusa 16787 CDFA IPC seed accession RT05-14A genotype 1/2. Hwy 43 S of Shafter, open silty field. 35°26'10’N, 119°15’19'W, 375 ft. el. 9/29/2005. Hrusa 16813 CDFA IPC seed accession RTO5-18, genotype 1/2; Hrusa 16814 CDFA IPC seed accession RT05-19, genotype 2/5, sag pond on San Andreas fault, S end Carrizo Plain. 34°59'55’N, 119°29'19'W, 9/29/2005. LOS ANGELES CO: O’Brien s.n. Roadside, Gorman. 11/3/2005. MONO CoO.: Hrusa 16697 Corner of Hwy 395/Hwy 89. 38°38'55'N, 119°31'50'W, 5100 ft. el. 9/27/2005. MONTEREY CO.: Fuller 14184 N side Hwy 198 1.5 mi. W of Coalinga. 10/7/1965. SAN LUIS OBISPO CO.: Fuller 14179, 14180, Hwy 466 (Hwy 46) 1.3 mi. W of Shandon. 10/7/ 1965. Fuller 14181 S side Hwy 466 (Hwy 46) N of Shandon. 10/7/1965. Hrusa 16817; 16818; 16819; 16820; 16821; 16822, S end of Elkhorn Plain. 35°01'36'N, 119°29’'27'W. 2500 ft. el. 9/29/2005. SAN BERNAR- DINO CO.: Akers RT-065-3 Old AFB, Phantom West Rd. and Aviation Dr., 34.590°N, 117.366°W. 2850 ft. el. 10/31/2002. NEVADA: NYE CoO.: Hrusa 16727 geno- type 2/5; Hwy 95 roadside at Esmeralda Co. line. 37°27'58'N, 117°10'07'W.4500 ft. el. 9/27/2005. WA- SHOE CoO.: Hrusa 16691; 16692; 16693; Mogul exit off Interstate 80, +/— 10 mi W of Reno. 39.517°N, 119.923°W, 4800 ft. el. 9/26/2005. Salsola paulsenii USA: CALIFORNIA: INYO Co.: Arusa 16736, 16737 CDFA IPC seed accession RTO5-5, 16738, Tecopa, Death Valley, Corner Old Spanish Trail Rd. & Furnace Lake Rd. on rd. to China Rch. 35°50’S51'N, 116°12’08'W. 1500 ft. el. 9/27/2005. KERN CO: Fuller 18837 S side Hwy at W limits of Boron. El. 2450 ft. 8- 21-1969. SAN BERNARDINO CO.: Fuller 12644 S side Hwy 66 to Daggett, E limits of Barstow. 9/22/1964; Fuller 12645 Hwy. To Barstow, 2.6 mi. W of Daggett. 9/22/1964; Fuller 15216, disturbed ground, 2 mi. E of Barstow on I-40, 9/28/1966. Fuller 15217, S of I-40 to Daggett, 2 mi. E of Barstow. 9/28/1966. Hrusa 16740 CDFA IPC seed accession RTO5-7, Dumont Dunes, Death Valley, 9/27/2005; Hrusa 16744 CDFA IPC seed accession RTOS-8, 16745 CDFA IPC seed accession RT05-9, NE edge of Silver Lake, edge Hwy 127, 35°22'34'N, 116°07'23'W, 900 ft. el. 9/27/2005. Hrusa 16757 CDFA IPC _ seed accession RTOS-11, 16756, 16758, 16759, Mojave R. bed in BLM Afton Cyn Preserve. 35°02'51'N, 116°23'54’W, 1400 ft. el. 9/28/ 2005. Hrusa 16761 CDFA IPC seed accession RTO5-12, 16762, 16763, 16765, 16766 Daggett, sandy bank of Mojave River. 34°52'34'N, 116°53'27'W 2000 ft. el. 9/ 28/2005. Hrusa 16775 CDFA IPC seed accession RTOS- 13, 16776, 16777, Hwy 58 betw. Kramer Jtn. And [Vol. 55 Barstow on Harpur Lake Rd. at RR track crossing. 34°57'03'N, 117°19'59'W. 2700 ft. el. 9/28/2005. Salsola ‘type C’* USA: CALIFORNIA: KERN CoO.: Akers RT-056-2, genotype 2/5/4, N of Hwy 58, int. Edison Hwy & Tower Line Rd. along RR. 35.332°N, 118.807°W, 794 ft. el. 10/30/2002. Akers RT-069-3, genotype 2/5/4, Hwy 119, 0.5 mi. W of Old River. 35.267°N, 119.153°W. 315 ft. el. 11/01/2002 (holotypus). Akers RT-72-1, genotype 2/ 5/4, RT-072-2, genotype 2/5/4, Elk Hill Rd., NE base of Elk Hills. 35.333°N, 119.463°W, 395 ft. el. 11/01/2002. SUTTER CO.: AHrusa 15999, genotype 2/3, Yuba City downtown rd. edge. 10/01/2001. YUBA CO.: Akers RT- 010-2, genotype 2/3, RT-10-3 genotype 2/3, Smartville Rd., 0.3 mi. N of Daugherty. 39.186°N, 121.304°W. 100 ft. el. 9/19/2002. APPENDIX IIT ‘Type C’ specimens whose PEPC intron genotypes | were determined for this study. CG = ‘“‘common garden”. See Appendix I for additional specimen data. The 14 specimens marked with an asterisk were confirmed ‘type C’ genotypes and used in the discrim- inant function calculations. The remainder were used in | tests of those functions (Fig. 7, Table 1 and Appendix -— III). | Plant Genotype Specimen County DNA # 2/5/4 Akers RT-069-2 Kern 6l11 2/5/4 *Akers RT-069-3 Kern 6112 2/5/4 *Akers RT-072-2 Kern 6116 2/5/4 Akers RT-055-1 Kern 6106 2/5/4 *Akers RT-072-1 Kern 6115 2/5/4 *Akers RT-056-2 Kern 6108 1/4 CG: C-2/1 Fresno 6125 1/4 *CG: C-2/2 Fresno 6126 1/4 *CG: C-2/3 Fresno 6127 1/4 CG: A-10/1 Fresno 6117 1/4 *CG: A-10/2 Fresno 6118 1/4 *CG: A-10/3 Fresno 6119 1/4 CG: A-10/4 Fresno 6120 1/4 *CG: B-3/1 Fresno 6121 1/4 CG: B-3/2 Fresno 6122 1/4 *CG: B-3/3 Fresno 6123 1/4 *CG: B-3/4 Fresno 6124 2/3 *Akers RT-010-2 Yuba 6101 2/3 *Akers RT-010-3 Yuba 6102 2/3 *Hrusa 15999 Sutter 6133 APPENDIX III Salsola tragus, S. ‘type B’ and S. ‘type C’ specimens used to test the efficacy of the discriminant functions. Groupings below are based on correlation to those in Fig. 5. Vouchers at CDA unless specified. See also Table 1. Salsola ‘type B’ AUSTRALIA: C. Borger 01, genotype 3/3. Gerald- ton, Western Australia (PERTH). CALIFORNIA: KERN CO.: Hrusa 16172 genotype 3/3. Hwy 166/33 * Plant DNA#s in Appendix II. 2008] 10 mi. E of Maricopa. 35.058°N, 119.258°W. 600 ft. el. 10/12/2003; Akers RT-074-1 genotype 3/3. E side Hwy 46/I-99 interchange. 35.601°N, 119.208°W, 415 ft. el. 11/01/2002. LOS ANGELES CoO: B. Villegas B247-1; B247-2; genotype 4/4, Calabasas nr. Ventura Co. line. 34°9.93'N, 118°40.67', 8/21/2002. MERCED CO.: Akers RT-030-2 genotype 4/4. Billie Wright Rd. ~0.4 mi. W of I-5 bridge. 37.04°N, 120.96°W, 246 ft. el. 10/10/2002; Akers RT-013-1 genotype 4/4. S_ side Hwy 152. 36.983°N, 121.374°W, 255 ft. el. 09/23/2002; REPUBLIC OF SOUTH AFRICA: M. Rejmanek SA1, genotype 3/3. Kimberley region. 10/28/2005. Salsola tragus (‘type A’) MEXICO: BAJA CALIFORNIA: Moran 21152 genotype 1/2. Four mi. W of Ojos Negros. 31°53’N, 116°22'’W, 600 m el. 11/10/1973. UNITED STATES: CALIFORNIA: BUTTE CO.: Ahart 8295 genotype 1/ 2. E side Sacramento River 1 mi. SE of Ord Ferry. 105 ft. el. 10/3/1999. FRESNO CO.: Common garden row A-7/2-4, all genotype 1/2. Akers RT-014-] genotype 2/5. Kearney Research Center, Rio Vista Rd. 0.1 mi. N HRUSA AND GASKIN: SALSOLA TRAGUS AND SALSOLA AUSTRALIS 131 of Manning. 36.606°N, 119.473°W, 353 ft. el., 09/23/ 2002. KERN CO: Akers RT-056-1 genotype 2/5. Akers 65-1 genotype 1/2/5. Old AFB, Phanton West Rd. and Aviation Dr. 34.59°N, 117.366°W, 2850 ft. el. 10/31/ 2002. SOLANO CO: Akers RT-O005-3 genotype 1/2 (Gaskin et al. 2006). Corner of Robben & Hackman Rds. 38.453°N, 121.179°W. 43 ft. el. 09/18/2002. STANISLAUS CO.: Akers RT-042-1, genotype 2/5, Akers RT-042-3, genotype 1/2, Radio tower, ~6 mi. E of Turlock at Main & Central. 37.491°N, 120.957°W. 46 ft. el. 10/28/2002. Salsola ‘type C’ CALIFORNIA: FRESNO CO.: Common garden row C-2/ind. 1; row A-10linds. 1, 4; row B-3/ind. 2, all genotype 1/4. Coalinga, Phelps Rd. at Hwy 33, 36°9.9'N, 120°21.2",W. KERN CO.: Akers RT-055-1 genotype 2/5/4, Shafter, vacant lot to W of Santa Fe Way ~1000 ft, N of Hagemon. 35.400°N, 119.151°W. 10/30/2002. Akers RT-069-2, genotype 2/5/4, Hwy 119, 0.5 mi. W of Old River. 35.267°N, 119.153°W. 315 ft. el. 11/01/2002. MADRONO, Vol. 55, No. 2, pp. 132-142, 2008 GENETIC EVIDENCE OF HYBRIDIZATION BETWEEN OENOTHERA WOLFII (WOLF’S EVENING PRIMROSE) AND O. GLAZIOVIANA, A GARDEN ESCAPE JENNIFER DEWoopy*', LEONEL ARGUELLO’*, DAVID IMPER*, ROBERT D. WESTFALL* AND VALERIE D. HIPKINS! 'USDA Forest Service, Pacific Southwest Research Station, National Forest Genetics Lab, 2480 Carson Road, Placerville, CA 95667 > U.S. Department of the Interior, National Park Service, Redwood National and State Parks, P.O. Box 7, Orick, CA 95555 7U.S. Fish and Wildlife Service, 1655 Heindon Road, Arcata, CA 95521 *USDA Forest Service, Pacific Southwest Research Station, P.O. Box 245, Berkeley, CA 94701-0245 ABSTRACT Isozyme analysis of the rare Oenothera wolfii (Wolfs evening primrose) and the garden escape, O. glazioviana, indicates that hybridization between these species may be more widespread than morphological evidence indicates. Although both species contained low amounts of genetic variation, unique alleles were identified in both taxa. Analysis of 22 populations, including pure populations of each species, identified eight populations as containing putative hybrid individuals. Four of these putative hybrid populations were considered pure O. wo/fii based on morphological analysis. This study confirms that the native O. wo/fii may be at risk not only from habitat destruction, but potentially from genetic swamping where it co-occurs with O. glazioviana. These results can be used as baseline information for future genetic monitoring efforts. Key Words: complex heterozygote, genetic swamping, hybridization, isozymes, Oenothera. Although habitat loss usually poses the great- completely describe hybrid swarms of individuals, | particularly if second-generation hybrids or back- | cross individuals occur frequently. Genetic infor- | mation may provide greater power to identify | hybrids if unique alleles occur in either or both | pure species. Even in the absence of unique alleles, given sufficient variation in neutral, bi-parentally © est threat to a rare species’ survival, there is evidence that hybridization with widespread related taxa poses an immediate threat to some species (Rhymer and Simberloff 1996). A rare species may become functionally extinct through genetic swamping after repeated hybridization and backcrossing with a more common species (Levin et al. 1996). Management efforts to minimize hybridization in order to protect a rare species may not be justified if hybridization results from natural processes. By contrast, artificial hybrid zones arising from human- mediated habitat modification or species intro- duction may require management action to minimize the potential loss of a rare species (Rhymer and Simberloff 1996; Allendorf et al. 2001). In order to minimize the effects of artificial inherited, genetic markers (for example, iso- | zymes), statistical methods exist to identify not | only first-generation hybrid individuals, but also | second-generation hybrids and introgressed indi- | viduals resulting from backcrosses with either parental species (Rannala and Mountain 1997; — Rieseberg et al. 1998; Nason et al. 2002). Oenothera wolfii [Munz] Raven, W. Dietr. | Stubbe (Onagraceae) (Wolf's evening primrose) is a biennial to short-lived perennial native to the coastal areas of northern California and southern | hybrid zones, managers must be able to distin- guish between pure populations of a rare species and hybrid swarms where the two species coexist. Frequently, hybrids display phenotypes interme- diate to either parent species, although hybrid Oregon. Populations of this species are rare and | patchy in distribution, found on moderately > disturbed sites, including the upper margin of | beach strand and coastal bluffs (Imper 1997). While disturbance resulting from continued | morphology may be extreme to either parent (Schwarzback et al. 2001). Due to these varia- tions, morphology alone may be insufficient to * Author for correspondence. Current address: Un1i- versity of Southampton, School of Biological Sciences, Southampton, SO16 7PX, United Kingdom. +44-(0)79- 8363-0570 (voice), +44-23-8059-4459 (fax), }.dewoody@ soton.ac.uk. development and recreation along the coast have | created new habitat for O. wolfii in some. instances, the net effect of human encroachment has been negative for existing populations (Imper 1997). As a result, O. wolfii is listed as threatened by the state of Oregon, and both the California Native Plant Society and the Oregon Natural | Heritage Program list this species as endangered throughout its range (Imper 1997). 2008] While habitat loss is affecting O. wo/fii, hybridization with a common congener, O. glazioviana Micheli, may prove the more imme- diate threat (Imper 1997). As a garden escape (i.e., a horticultural species of hybrid origin that has become established in natural areas), O. glazioviana may be interfertile with O. wolfii, producing artificial hybrid zones where the species coexist. Several factors support this hypothesis. First, introgression is common be- tween many members of this genus. Greenhouse experiments have shown that hybridization be- tween O. wol/fii and other members of the genus readily occurs (Wasmund and Stubbe 1986). Second, individuals of hybrid origin have been identified at the California-Oregon border area based on morphological traits (Carlson et al. 2001). Hybrids are fertile, vigorous, and display a greater fitness than either parent species (Imper 1997). Although genetic typing of hybrid individ- uals indicates that hybrids tend to breed true, there is evidence of hybrids back-crossing with O. wolfii (Imper 1997). Third, O. wo/fii is potentially susceptible to genetic swamping by O. glazioviana based on the mating systems of each species. O. wolfii is self-compatible and produces the major- ity of its seed via self-pollination (Carlson et al. 2001). This breeding system is a consequence of O. wolfii’s complex heterozygous genome, which is maintained through self-fertilization and bal- anced lethals, and results in approximately half of the mature pollen grains being sterile (Wasmund and Stubbe 1986). In contrast, O. glazioviana 1s an outcrossing species (Imper 1997). Given the asymmetry of available pollen between these parent species asymmetric gene flow might occur as O. glazioviana pollen swamps O. wol/fii stigmas at sympatric sites. Together, this evidence sug- gests that hybridization likely occurs between this rare endemic and the widespread garden escape. This study reports an investigation into the extent and structure of hybrid zones between O. wolfii and O. glazioviana using isozymes, which are putatively neutral, bi-parentally inherited, molecular markers. Three questions were ad- dressed: First, does sufficient genetic variation exist to discriminate between pure O. wo/fii and O. glazioviana populations? Second, can hybrid populations be identified using these molecular markers? Third, what is the frequency of hybrid individuals in natural populations of O. wolfii? Ultimately, these genetic findings provide greater insight and guidelines for management plans and conservation objectives. METHODS O. wolfii is a complex heterozygote (Wasmund and Stubbe 1986), or complex hybrid (Bussell et al. 2002), where a diploid individual contains not two copies of a single genome, but a single copy DEWOODY ET AL.: HYBRIDIZATION IN OENOTHERA 133 of two distinct genomes. Multiple reciprocal translocations across the genome have produced a single linkage group consisting of both sets of chromosomes at meiosis. As a result, the 14 chromosomes in a diploid individual form a single ring instead of seven bivalents. This phenomenon persists through self-pollination coupled with balanced lethals, with gametophytic and sporophytic lethals persisting in the hetero- zygous state. Although self-fertilization may produce embryos homozygous for either genome, only heterozygous embryos survive to produce viable seed, since those embryos homozygous for one genome will also be homozygous for either the sporophytic or gametophytic lethal allele. As a result, alleles are not independently assorted, and populations are not randomly mating. This mechanism explains the lack of genotypic diver- sity and recombination observed in the data set (see Results), and prevents the use of statistical analyses typical of co-dominant genetic data (e.g., admixture analyses or population assignment tests). Samples were collected from populations at 22 sites whose taxonomy was determined by mor- phological traits (Tables | and 3, Fig. 1). Typical O. wolfii plants produce small (<5 cm) pale- yellow corollas having petals that do not overlap, with sepals covered in dense long-spreading pubescence, both villous and glandular pubes- cence occur on fruits, and plants have a reddish upper stem (Imper 1997). By contrast, typical O. glazioviana plants produce larger (>8 cm) bright yellow flowers having substantial overlap in the petals, minimal pubescence on either sepals or fruit, green upper stems, and more wrinkled and lighter green foliage than observed on O. wolfii (Imper 1997). Field observations identified four populations as O. glazioviana (nos. 1 to 4; the populations of the garden escape most proximate to O. wolfii), 14 populations as O. wo/fii (nos. 6 to 19), and three populations as intermediates or putative hybrids (nos. 20 to 22). Field observa- tions could not distinguish between O. glaziovi- ana and O. elata, a common congener at one site (no. 5), and one population appeared to be O. wolfii, but occurred in a novel location (no. 15). A single leaf was collected from between four and 25 individuals in each population for subsequent genetic analyses. Tissue was prepared for isozyme analysis following the liquid nitrogen procedure using Gottleib (1981) extraction buffer, as described in NFGEL Standard Operating Procedures (USDA Forest Service 2003). Samples were frozen at —70°C until electrophoresis. Electrophoresis took place on three buffer systems (adapted from Wendel and Weeden 1989): a tris-citric acid gel buffer (pH 8.3) with a lithium hydroxide-boric acid tray buffer (pH 8.3; LB), a tris-citric acid gel buffer 134 MADRONO TABLE lL. POPULATION NUMBER, NAME, LOCATION (LATITUDE, LONGITUDE), AND SPECIES COMPOSITION OF 22 SITES SAMPLED FOR THIS STUDY. Species composition was determined by field observations and genetic analysis, and is indicated by: WO = O. wolfii, GL = O. glazioviana, HY = intermediate morphology potentially due to hybridization, and UN = unknown taxonomy. Number Name Location Species l Charleston, Coos Co., OR 43.3397N, 124.3308W GL 2 Crescent City, Del Norte Co., CA 41.7486N, 124.2022W GL 3 Manila, Humboldt Co., CA 40.8483N, 124.1650W GL 4 Trinidad, Humboldt Co., CA 41.0353N, 124.1058W GL 5 Junction City, Trinity Co., CA 40.7378N, 123.0575W UN 6 Port Orford City Park, Curry Co., OR 42.8320N, 124.5020W UN Z Houda Point, Humboldt Co., CA 41.0359N, 124.1187W WO 8 Port Orford Beach, Curry Co., OR 42.7318N, 124.4825W UN 9 Port Orford Bridge, Curry Co., OR 42.7318N, 124.4825W UN 10 Luffenholtz, Humboldt Co., CA 41.0353N, 124.1247W WO 11 Pistol River, Curry Co., OR 42.2717N, 124.4051W WO 12 Point St. George, Del Norte Co., CA 41.7778N, 124.2405W WO 13 Devil’s Gate, Humboldt Co., CA 40.4055N, 124.3914W UN 14 Davis Creek, Humboldt Co., CA 40.3765N, 124.3725W WO 15 McKerricher State Park, Mendocino Co., CA 39.5146N, 123.7769W UN 16 Freshwater Spit, Humboldt Co., CA 41.2667N, 124.1058W WO 17 Crescent Beach, Del Norte Co., CA 41.7194N, 124.1447W WO 18 False Klamath Cove, Del Norte Co., CA 41.6027N, 124.1064W WO 19 Crescent Overlook, Del Norte Co., CA 41.7048N, 124.1447W WO 20 Klamath, Del Norte Co., CA 41.5151N, 124.0298W HY 21 Lucky Bear Casino, Del Norte Co., CA 41.9529N, 124.2022W HY 22 Fruit Station, Curry Co., OR 41.9984N, 124.2124W HY (pH 8.8) with a sodium hydroxide-boric acid tray buffer (pH 8.0; SB), and a citric acid-N-(3- aminopropyl)-morpholine gel and tray buffer (pH 8.0; MC8). A total of 15 loci were examined. Four loci were resolved on the LB system: phosphoglucose isomerase (PGI2), phosphoglu- comutase (PGM1), and two loci in leucine aminopeptidase (LAPI and LAP2). Four loci were also resolved on the SB system: aspartate aminotransferase (AAT1), superoxide dismutase (SOD1), triosephosphate isomerase (TPI1), and uridine diphosphoglucose pyrophosphorylase (UGPP1). Seven loci were resolved on the MC8& system: two loci in esterase (EST1 and EST2), fluorescent esterase (FEST1), isocitrate dehydro- genase (IDH), malate dehydrogenase (MDH), and two loci in 6-phosphogluconate dehydroge- nase (6PGD1 and 6PGD2). All stain recipes were adapted from Conkle et al. (1982). Banding patterns were consistent with published protein structure (Crawford 1989). As a consequence of complex hybridity in O. wolfii, the isozyme data resolved in this study violate two assumptions common to most statis- tical analyses designed for genetic data: indepen- dent assortment of loci and random mating. Multivariate analyses can often be used to analyze co-dominant genetic data, but these analyses require that errors (or residuals) are normally distributed. This assumption is gener- ally satisfied under independent assortment and random mating, and experience shows that even binary-scored markers can fit the assumption. But, portions of the Oenothera isozyme data violate these assumptions. Given the unique nature of the genetic system of O. wolfii, and [Vol. 55 | the lack of statistical procedures available to > account for complex hybridity as a mode of inheritance, an ad hoc approach involving two statistical analyses was employed to determine if hybridization occurs between O. wo/fii and O. | glaziovianna: multivariate analyses over popula- tions and individuals, and a maximum likelihood analysis over populations. While acknowledging that neither approach is statistically ideal, we contend that given the unique nature of the. genome of Oenothera which may bias results of a single statistical test, combining statistical analy- ses with careful examination of the electropho- retic patterns provides an informative approach to describe the genetic similarities and potential for hybridization between these species. For the multivariate analysis, we scored each allele as 1.0, 0.5, or 0.0 for homozygous, | heterozygous, or homozygous for another allele, respectively (Westfall and Conkle 1992). These > scored data were submitted to a canonical discriminant analysis. We first ran the analysis on populations, without respect to species classi- fication to determine if populations grouped by species identity. We then contrasted these results with a classification based on the apriori group- | ings. Based on the canonical discriminant plot, each population was classified as either “‘pure” (that is, a parental species) or ‘““Cunknown”’ (that is, either putative hybrid or unknown taxonomy) 2008] DEWOODY ET AL.: HYBRIDIZATION IN OENOTHERA 135 sie 44° 42 42° 40° 40° FIG. 1. Location of 22 populations sampled for genetic analysis. Numbers correspond to populations in Table 1. 136 TABLE 2. MADRONO [Vol. 55 ISOZYME DIVERSITY SUMMARY STATISTICS FOR 22 POPULATIONS DESCRIBED IN TABLE 1. MEANS OVER SPECIES INCLUDE ONLY “‘PURE”’ POPULATIONS. NV = number of samples, P = percent polymorphic loci, A = mean alleles per locus, Ap = mean alleles per polymorphic locus, H, = observed heterozygosity, F = fixation index. Variance reported in parentheses. Population N ya A Ap ta F Mean over species: O. wolfii 137 13.33 1.200 (0.293) 2.500 0.021 (0.005) =().571 O. glazioviana 61 6.67 1.067 (0.062) 2.000 0.067 (0.062) — 1.000 l 25 0.067 1.067 2.000 0.067 — 1.000 Ds 17 0.067 1.067 2.000 0.067 — 1.000 3 1] 0.067 1.067 2.000 0.067 — 1.000 4 8 0.067 1.067 2.000 0.067 — 1.000 5 pA | 0.133 b.133 2.000 0.133 =1.000 6 6 0.133 1.133 2.000 0.078 =(0.7 50 7 12 0.067 1.067 2.000 0.011 —0.048 8 9 0.067 1.067 2.000 0.067 — 1.000 9 9 0.067 1.067 2.000 0.067 — 1.000 10 13 0.067 1.067 2.000 0.015 —0.091 11 25 0.000 1.000 n.a. 0.000 0.000 12 19 0.067 1.067 2.000 0.018 =O 125 13 10 0.200 1.200 2.000 0.073 0.214 14 9 0.133 1.133 2.000 0.067 —0.385 15 10 0.000 1.000 n.a. 0.000 0.000 16 25 0.000 1.000 n.a. 0.000 0.000 i 5 0.000 1.000 n.a. 0.000 0.000 18 25 0.067 1.067 2.000 0.067 — 1.000 19 4 0.000 1.000 n.a. 0.000 0.000 20 10 0.067 1.067 2.000 0.067 — 1.000 21 10 0.067 1.067 2.000 0.067 — 1.000 22 5 0.067 1.067 2.000 0.067 — 1.000 (Table 1). The allele frequencies observed over all pure populations of each species were then used as the basis to classify each unknown population using the methods described above. These anal- yses were done in JMP (SAS Institute, Inc, 2004). This software’s canonical discriminant analysis 1s based on Bayesian probabilities, whereby, in well- differentiated species, individuals of one species will have a probability of 1.0, those of the other species, 0.0, and those of hybrid or backcross types will have probabilities between 1.0 and 0.0. The second analysis estimates the frequency of six genealogical classes (each parental class, first- and second-generation hybrids, and first genera- tion backcross to each parent species) in each population based on the maximum likelihood estimates of the multilocus genotypes observed in a population arising from the allele frequencies observed in each of the pure parental species (Nason et al. 2002). While this method assumes both independent assortment of alleles and random mating within each population, assump- tions that are violated here, it includes all loci in the data set in the maximum likelihood estima- tions, without requiring unique alleles in each parent species (Nason et al. 2002). Finally, the conclusions from each statistical analysis were considered in the context of the alleles observed in each genotype in the nine unknown populations, with particular attention given to those loci displaying alleles unique to at least one parental species. RESULTS Six of the 15 loci examined were polymorphic: 6PGD2, AAT1, UGPP1, FEST1, EST1, and EST2 (Appendix A). Four loci displayed varia- tion within or among populations of the pure species, and two loci contained variation in unknown populations not observed in either pure species. Low levels of genetic variation were observed over all populations surveyed (Table 2). Based on the mean over species, O. wolfii contained higher levels of polymorphism (percent polymorphic loci, P = 13.3), alleles per locus (A = 1.20), and alleles per polymorphic locus (Ap = 2.50), but lower levels of heterozygosity (observed proportion of heterozygotes, H, = 0.021) than O. glazioviana (P = 6.67, A = 1.07, Ap = 2.00, H, = 0.067). Many populations displayed an excess of heterozygotes, as indicated by fixation indices, which is consistent with complex hybridity (Table 2). Multivariate analyses over populations indi- cated sufficient genetic differentiation exists to distinguish between O. wo/fii and O. glazioviana (Fig. 2). In general, populations grouped by species identity as defined in field observations, with the majority of populations being separated 2008] WO e HY no. 20 UN no. 15 = GC) S UN no. 13 5° ; UN nos. 6, 8,9 UN no. 5 HY nos. 21, 22 >| T T 30 -20 -10 0 CAN1 Fic. 2. Distribution of populations along the first two canonical variables produced by a discriminate coordi- nate analysis. Populations classifications and numbers correspond to Table |. Can! = first canonical variable, Can2 = second canonical variable. by species identification along the second canon- ical axis (Fig. 2). However, the first two canonical coefficients revealed greater genetic differentia- tion than predicted among populations classified as O. wolfii from morphological characteristics. In particular, three populations from Oregon, nos. 6, 8, and 9, were genetically distinct from the other populations considered pure O. wolfii (Fig. 2). Additionally, population 13 was inter- mediate to the two parental species (Fig. 2). As a result of these observations, these four outlying populations were classified as unknown taxono- my for the remaining statistical tests, reducing the number of O. wo/fii populations to those listed in Table 1. No evidence of hybridization (or admixture) was found using the Bayesian classification analysis of individuals. Bayesian tests classified TABLE 3. DEWOODY ET AL.: HYBRIDIZATION IN OENOTHERA r7 all individuals as either pure O. wo/fii or pure O. glazioviana (Table 3); no intermediate probabili- ties were observed. Samples from three popula- tions identified as pure O. wol/fii based on morphological observations were classified as O. glazioviana (nos. 6, 8, and 9). Of the three remaining populations of unknown taxonomy, all individuals from one population were classi- fied as O. glazioviana (no. 5), all from another as O. wolfii (no. 15) and the final population contained a mixture of individuals classified as both pure species (no. 13). Of the three popula- tions classified as hybrid based on morphological observations, genetic analyses classified samples from one as O. wo/fii (no. 20), and those from the other two as O. glazioviana (nos. 21 and 22). Genealogical class frequency estimates were also inconsistent with morphological predictions (Table 3). Unlike the Bayesian classifications, however, some genotypes were identified as consistent with hybrid origin. Of the populations identified as pure O. wo/fii a priori, three were classified as hybrids (nos. 6, 8, and 9), and a fourth (no. 13) was classified as a mixture of O. wolfii and hybrids. Consistent with the Bayesian classifications, two of the putative hybrid popu- lations were classified as O. glazioviana (nos. 21 and 22), although the third was classified as backcross to O. wolfii (no. 20). DISCUSSION Does sufficient genetic variation exist to discriminate between species? In genetic studies of hybridization, the genetic variation in each species is often defined by identifying “pure” populations from morpholog- ical observations and assaying each for genetic markers. Alternatively, multivariate analyses such as the canonical discriminant analysis described above, can identify genetically similar or distinct populations without a priori classifi- cation. In this study, the canonical discriminant CLASSIFICATION OF NINE POPULATIONS OF UNKNOWN OR HYBRID ORIGIN BASED ON 6 VARIABLE ISOZYME Loct. See text for details of the Bayesian classifications and genealogical class freuqencies. Isozyme phenotypes are classified by the presence of alleles found to be unique to either parental species. Field Population Observations Bayesian classification Genealogical class frequency Isozyme phenotype =) Unknown O. glazioviana Backcross to O. Neither species or Hybrid glazioviana 6 O. wolfii O. glazioviana Hybrid Hybrid 8 O. wolfii O. glazioviana Hybrid Hybrid 9 O. wolfii O. glazioviana Hybrid Hybrid 13 O. wolfii Mix of pure parental Mix of O. wolfii and Mix of O. wolfii and Hybrid individuals Hybrid 15 O. wolfii O. wolfii O. wolfii O. wolfii 20 Hybrid O. wolfii Backcross to O. wolfii Hybrid 21 Hybrid O. glazioviana O. glazioviana O. glazioviana 22 Hybrid O. glazioviana O. glazioviana O. glazioviana 138 MADRONO analysis indicated that four populations which were identified as O. wo/fii in field observations were genetically distinct from the other O. wolfii populations (Table 3, Fig. 2). Given the striking genetic differences between these populations, the four outliers were treated as “‘unknown”’ taxon- omy for the remaining data analyses and interpretation. The multivariate analysis also indicates that sufficient genetic differentiation exists between the nine O. wol/fii populations and four O. glazioviana populations to discriminate between the parental species (Fig. 2). Although isozyme markers revealed low levels of variation in O. wolfii and O. glazioviana (Table 2), greater variation was observed in O. wolfii (O—20% polymorphic loci) than O. glazioviana (6.7% polymorphic loci), and all samples from “‘pure”’ O. glazioviana populations (nos. 1-4) shared a common genotype: heterozygous at AAT1, and monomorphic at all other loci. O. wolfii con- tained a greater number of alleles per locus (1.20 compared to 1.07 in O. glazioviana), and dis- played greater levels of fixation (—0.57 compared to —1.00 in O. glazioviana). However, had four populations initially con- sidered O. wo/fii (nos. 6, 8, 9, and 13) been included in the description of the parental species, the genetic differentiation would have been much less pronounced. Specifically, analyzing these populations as O. wolfii would affect the distri- bution of alleles at locus 6PGD2, making allele 6PGD2-2 no long unique to O. glazioviana, but shared between the species. Allele AAT-1 would remain unique to O. glazioviana, however. This difference would have likely reduced but not removed the ability of the multivariate and genealogical class frequency analyses to distin- guish between pure and hybrid individuals. These analyses are complicated, however, by the occurrence of alleles in several populations that are not observed in either pure species (Appendix A). There are three possible explana- tions for these observations. First, these alleles may be present in other populations of either or both parental species that were not sampled for this study. Second, considering populations 6, 8, and 9 as O. wolfii as per field observations would make alleles EST1-2, EST2-2, and FEST1-2 unique to O. wolfii. As population 20 has consistently been considered of putative hybrid origin, such a change in classification of other populations would not explain the origin of allele FEST 1-3. Third, the model we are testing, that all populations are either pure O. wo/fii, pure O. glazioviana, or an admixture of the two, may not explain the genetic structure observed. Past hybridization and introgression between O. wolfii and a third, unidentified species (possibly O. elata) may explain the high frequency of alternate alleles observed in some test populations. As no [Vol. 55 data were collected from other Oenethera species, we cannot test this alternate model. Analyzing genetic data from these species and conclusively identifying hybrid individuals is further complicated by the recombination system displayed by O. wolfii. As a complex hybrid, putative diploid individuals contain not two copies of a single genome, but one copy each of two distinct genomes. Wasmund and Stubbe (1986) showed that O. wolfii maintains this heterozygosity through self-fertilization coupled with balanced lethals. This recombination system causes species to be functional apomicts, typically displaying little genetic variation and heterozy- gosity (Russell and Levin 1988). The low levels of allelic diversity and near lack of genotypic diversity observed in O. wol/fii are consistent with these expectations. Although sufficient genetic variation exists to allow differentiation of pure species and identification of hybrid individuals, the lack of recombination and independent assortment at meiosis means that these data violate the assumptions common to most statis- tical analyses. Thus, standard statistical methods of identifying and monitoring hybrid swarms may not be applicable to O. wo/fii. In order to appropriately interpret genetic data without loosing information due to violations of model assumptions, comparing the results of multiple statistical analyses coupled with phenotypic descriptions of the multilocus genotypes provides insight into the origin of unknown or putative hybrid populations. As a final caveat, interpretation of this data set, as well as its application in future studies, must be considered in the context of the small sample sizes | at some populations and the small number of pure O. glazioviana populations sampled. Anal- | ysis of additional “‘pure’’? populations of O. | glazioviana may identify additional unique alleles | or reveal alleles thought to be unique to O. wolfii to be shared by the two species. Either observa- | tion could change the classification of unknown | samples and the conclusions herein. Ultimately, this data set represents a fraction of the Oenothera genome, and may not completely | represent the levels of variation or hybridization | in these species. Can hybrid populations be identified using 1sozymes? The genetic differences observed between the | nine populations of O. wo/fii and the four | populations of O. glazioviana are sufficient to allow identification of hybrid populations. Re- | sults of the two statistical analyses are inconsis- — tent, but indicate that hybridization may occur at © a rate greater than that expected from morpho- logical observations. The multivariate classifica- | tion of the six unknown and three putative hybrid | 2008] populations identified each collection as either parental species or a mixture of the two (Table 3). The genealogical class frequency estimates, by contrast, only identified three populations as either parental species, and the remaining popu- lations as some hybrid origin (Table 3). This lack of consensus between analyses may be a conse- quence of the violations of the statistical assump- tions these data present. Three general conclu- sions can be made when the field observations and statistical analyses are considered together. First, populations 6, 8, 9, 13, and 20 are distinct from either parental species. Second, populations 5,21, and 22 more closely resemble O. g/azioviana than O. wo/fii. Third, population 15 is consistent with being O. wo/fii both morphologically and genetically. What is the frequency of hybrid individuals in natural populations? Although population-level analyses to detect hybridization produced inconsistent results, care- ful consideration of the multilocus genotypes demonstrates that plants from six populations display alleles unique to both parent species, and are thus consistent with hybrid origin (Appen- dix B). Samples from four populations consid- ered O. wolfii from field observations (nos. 6, 8, 9, and 13) contained one allele unique to O. glazioviana (6PGD2-2) as well as one unique to O. wolfii (UGPP1-2). Had these populations been considered pure O. wo/fii for the classification tests, and allele 6PGD2-2 would consequently be shared between the parental species. However, populations 6, 8, and 9 also displayed three alleles not observed in either parental species, EST1-2, EST2-2, and FEST1-2. These alleles may be unique to either parental species but not detected in the pure populations, or it may be the result of introgression with another Oenothera species (e.g. O. elata). As no other species was included in this study, no conclusions can be made regarding the origin of these alleles from these data. The genotype observed in population 5 is consistent with hybrid origin irrespective of the classification of populations 6, 8, and 9, as it contains the alternate allele unique to O. gla- zioviana, AAT1-2, as well as the allele unique to O. wolfii, UGPP1-2. Similarly, the genotype observed in population 20 is also consistent with a hybrid origin, containing the O. glazioviana allele AAT1-2 as well as the O. wolfii allele 6PGD2-1. However, the genotype in population 20 also contains two alleles observed in popula- tions 6, 8, and 9 (EST1-2 and EST2-2), as well as an allele unique to its population (FEST1-3). Again, given the absence of these alleles in either parental species and the lack of other Oenothera species in this study, the origin of these alleles cannot be determined. DEWOODY ET AL.: HYBRIDIZATION IN OENOTHERA 139 Despite violations of assumptions in each statistical analysis, this genetic study reveals evidence of hybridization between the rare endemic O. wo/fii and the garden escape O. glazioviana. A number of genotypes contain alleles found to be unique to each pure species (Appendix B), an observation most easily ex- plained as evidence of hybrid origin. These results indicate hybridization may occur at a greater rate than expected based on morphological observa- tions alone. Although the genetic structure of population 20, a putative hybrid population found to contain an intermediate genotype, demonstrates that not all hybridization events will lead to the genetic swamping of the rare species, timely removal of O. glazioviana plants from sympatric sites may be warranted to prevent further loss of the endemic genotype. While the isozyme loci used here provide alleles unique to each parent species, and thus the ability to identify hybrid individuals, the direction of hybridization and introgression cannot be as- sessed due to their bi-parental inheritance. Assaying these species for variation at maternal- ly-inherited markers (e.g., chloroplast haplo- types), and combining data from those markers with data from isozyme or other nuclear markers may provide the power to determine which species 1S serving as the seed donor in each hybridization, and thus determine if O. wolfii flowers are being swamped by O. glazioviana pollen. In addition, including O. e/ata in future studies would be prudent given the high rates of hybridization between many members of this genus. ACKNOWLEDGMENTS The authors would like to thank J. D. Nason for insightful discussion about this study as well as R. K. Dumroese and colleagues and two anonymous review- ers for thoughtful comments on the manuscript. LITERATURE CITED ALLENDORF, F. W., R. F. LEARY, P. SPRUELL, AND J. K. WENBURG. 2001. The problems with hybrids: setting conservation guidelines. TRENDS in Ecol- ogy and Evolution 16:613—622. BUSSELL, J. D., M. WAycotTT, J. A. CHAPPILL, AND S. H. JAMES. 2002. Molecular phylogenetic analysis of the evolution of complex hybridity in /sotoma petraea. Evolution 56:1296—1302. CARLSON, M. L., R. J. MEINKE, AND A. WIERCK. 2001. Wolfs Evening Primrose (Oenothera wolfii) hybridization, reproductive ecology, seed germina- tion, and cultivation. Oregon Department of Agriculture, Plant Conservation Biology Program, Salem, OR. CONKLE, M. T., P. D. HODGSKISs, L. B. NUNNALLY, AND S. C. HUNTER. 1982. Starch gel electropho- resis of conifer seeds: a laboratory mManual. General Technical Report PSW-64. USDA Forest Service, Pacific Southwest Forest and Range Experiment Station, Berkeley, CA. 140 CRAWFORD, D. J. 1989. Enzyme electrophoresis and plant systematics. Pp. 146-164 in D. E. Soltis and P. S. Soltis (eds.), Isozymes in plant biology. Dioscorides Press, Portland, OR. GRANT, V. 1971. Plant Speciation. Columbia Univer- sity Press, New York, NY. GOTTLIEB, L. D. 1981. Gene number in species of Asteraceae that have different chromosome num- bers. Proceedings of the National Academy of Sciences USA 78:3726—3729. IMPER, D. K. 1997. Ecology and conservation of Wolf's evening primrose in northwestern California. Pp. 34-40 in T. N. Kaye, A. Liston, R. M. Love, D. L. Luoma, R. J. Meinke, and M. V. Wilson (eds.), Conservation and management of native plants and fungi. Native Plant Society of Oregon, Corvallis, OR. LEVIN, D. A., J. FRANCISCO-ORTEGA, AND R. K. JANSEN. 1996. Hybridization and the extinction of rare plant species. Conservation Biology 10:10—16. NASON, J. D., S. B. HEARD, AND F. R. WILLIAMS. 2002. Host-associated genetic differentiation in the goldenrod elliptical-gall moth, Gnorimoschema gallaesolidaginis (Lepidoptera: Gelechiidae). Evo- lution 56:1475—-1488. RANNALA, B. AND J. L. MOUNTAIN. 1997. Detecting immigration by using multilocus genotypes. Pro- ceedings of the National Academy of Sciences USA 94:9197-9201. MADRONO [Vol. 55 RHYMER, J. M. AND D. SIMBERLOFF. 1996. Extinction by hybridization and introgression. Annual Review of Ecology and Systematics 27:83—109. RIESEBERG, L. H., S. J. E. BAIRD, AND A. M. DESROCHERS. 1998. Patterns of mating in wild sunflower hybrid zones. Evolution 52:713—726. RUSSELL, J. AND D. A. LEVIN. 1988. Competitive relationships of Oenothera species with different recombination systems. American Journal of Bot- any 75:1175—1180. SCHWARZBACK, A. E., L. A. DONOVAN, AND L. H. RIESEBERG. 2001. Transgressive character expres- sion in a hybrid sunflower species. American Journal of Botany 88:270—277. USDA FOoRESsT SERVICE. 2003. National Forest Genet- ic Electrophoresis Laboratory standard operating procedures. USDA Forest Service Pacific South- west Research Station, Placerville, CA. WASMUND, O. AND W. STUBBE. 1986. Cytogenetic investigations on Oenothera wolfii (Onagraceae). Plant Systematics and Evolution 154:79-88. WENDEL, J. F. AND N. F. WEEDEN. 1989. Visualization | and interpretation of plant isozymes. Pp. 5—45 in D. E. Soltis and P. S. Soltis (eds.), Isozymes in plant biology. Dioscorides Press, Portland, OR. WESTFALL, R. D. AND M. T. CONKLE. 1992. Allozyme | markers in breeding zone designation. New Forests _ | 6:279—309. 14] 000 I 000 I 000° I 000'| 00SO0 O00S0 = O000'T (HE ne 000 I 000 I 000 'I 000 I 00S0 O00S'0 = O000'T IZ Q 000'1 000 I 0001 000 'T 0OS'O0 O00S'0 000 I OC = 000 | 000 | 000° I 000° 000 I 000 I 61 S 00S 0 00S 0 000 | 000° 000 I 000° I 000° | 81 =, 000'1 000'| 000'1 000 I 000 I 000 I | 3 000 I 000 I 000 I 000 I 000 I 000 I | 7. 000 ' I O00 I 000 | 000 I 000 I 000° I SI = 00S'0 00S 0 000 I ITT 0 688 0 000 I 000 | 000 '| vl Z 00S 0 O08 0 000 I 00c 0 008°0 000 I 000 I OsvO Oss 0 el © CELO: - $980 000 I 000 'I 000 I 000 I 000 I cl 2 000 I 000 I 000 I 000 I 000 | O00 '| II N 000 I 000 I SIT O S880 000 | 000° I 000° T Ol a 00S'0 O0S'0 000° I 000' I 000 I 000 I 000° 1 6 v 00S'0 O00 000 I 000 I 000 I 000 I 0001 8 = 000 I 000 I 800 L916 0 000° I 000 I 000 I Ls a 00S'O0 =O00S'0 000 I 000° I L160 800 000 I 000 I 9 me 00S'0 = O00S'0 000 'I 000° I 000° I 00SO0 O0S0 O00'l S — 000° 000 I 000 I 000° I 00SO0 O0S0 O000'T v — 000'l 000 I 000 I 000° 00S0 O0S'0 = O00'l c = 000 'I 000 I 000 I 000° 00SO0 0080 O00'l C Se 000 I 000'1 000° I 000° I 00SO0 O0S0 O00'T I 5 000° 000 I 000 I 000° I 00S0 O00S0 O000'T DUDIAOI=D]B O © CVI O 8S8°0 000 I L000 8100 SL60 000 I 000 I 000 I 1f]OM “O - sa1oads IdA0 URd/Y A C I © c I v c C I ‘9 I c I fc [ uonrndog IddON [.LSdH CLSA LLSA ILLVV cd0Dd9 =o V/sh0'] ‘| ATAVE NI GadIyOsdd SNOILV1INdOd TT AHL NI GHANASIO IOOT AWAZOST ATAVINVA XIS AHL YOS SAIONANOAAY ATATTYV “VY XIGNdddV 2008] 142 MADRONO [Vol. 55 APPENDIX B. GENOTYPES AT SIX VARIABLE ISOZYME IOCI OBSERVED IN THE 22 POPULATIONS DESCRIBED IN TABLE |. Genotype Code identifies each unique genotype observed in the study. * Genotype contains alleles unique to both parent species. Population Genotype Code 6PGD2 AATI1 EST1 EST2 FESTI1 UGPPI1 1 A 2d 12 11 ip) 1] 1] 2 A 22 12 11 11 1] 11 3 A De 2 1] 1] 1] 11 4 A 22 12 ie) 11 11 1 5 B* 22 12 11 1] 11 12 6 CF 22 1] 22 pp 22 2 IB 2 1] 12 22 22 12 7 I 11 11 i! 1] 1] 11 J 1] | i) 13 11 Ht 8 C* 22 11 22 ep) 22 2 9 C* 22 al 22 De 22 1 10 I 11 11 11 1] 1] 11 J 1] |g 1] 3 1] 11 i I 1] 11 11 ul iat 1] 12 I 1] 1] 11 1] 11 1] K 11 iM 11 1] 11 1 13 K 1] lat 11 1] Is 12 Gig 12 11 11 1] i 12 E* 22 1] 11 ial 11 12 F* 22 1] 1] 44 [et [2 14 K 11 1] 1] 1] ie 12 L 1] 1] iB 44 11 12 15 I 1] 11 1] 1] 11 1] 16 I 1] 11 1] 11 1] 11 LZ i 1] 1] 1] 1] im it 18 K 11 1] 1] 11 11 12 19 I 1] 1] 1 1] i 1] 20 H* 11 12 22 Ze 33 1] 21 A 22 {2 11 11 Te I 22 A 22 | We 11 11 11 1] MADRONO, Vol. 55, No. 2, pp. 143—150, 2008 LEAF ANATOMY OF ORCUTTIEAE (POACEAE: CHLORIDOIDEAE): MORE EVIDENCE OF Cy, PHOTOSYNTHESIS WITHOUT KRANZ ANATOMY LAURA M. BOYKIN!?", WILLIAM T. POCKMAN! AND TIMOTHY K. LOWREY! ‘Department of Biology, University of New Mexico, Albuquerque, NM 87131 ABSTRACT C, photosynthesis without Kranz anatomy (single-cell C4 photosynthesis) occurs in only 0.003% of known species of C4 flowering plants. To add insight into the evolution of C4 photosynthesis, we studied the tribe Orcuttieae (Poaceae: Chloridoideae), which has species that can grow under both aquatic and terrestrial conditions, and utilize single-cell C4 photosynthesis when growing submerged. Carbon isotope ratios from aquatic, floating, and terrestrial leaves were in the range —12.25 to —14.31, suggesting that all species carry out C4 photosynthesis. Using light microscopy, we examined the anatomy of aquatic, floating and terrestrial leaves from eight of the nine species in the tribe to assess the pattern of evolution of C4 photosynthesis and Kranz anatomy among these vernal pool grasses. Kranz anatomy was present in all floating and terrestrial leaves of Orcuttia californica, O. inaequalis, O. pilosa, O. tenuis, O. viscida,Tuctoria greenei, T. mucronata, and Neostapfia colusana. Although carbon isotope data indicated C4 photosynthesis, aquatic leaves of all members of Orcuttia lacked Kranz anatomy, while aquatic leaves of Tuctoria and Neostapfia possessed Kranz anatomy. When considered in a phylogenetic context, these findings support previously proposed hypotheses suggesting that Orcuttieae are derived from a terrestrial ancestor and are now becoming more specialized to an aquatic environment. Key Words: C4 photosynthesis, Chloridoideae, Kranz anatomy, Neostapfia, Orcuttia, Orcuttieae, Poaceae, Tuctoria, single-cell C4 photosynthesis, vernal pool. Since the discovery of C4 photosynthesis in the 1960’s, there has been great interest in document- ing the biochemical and anatomical features of the process since it has a much greater photo- synthetic capacity as compared to C3 photosyn- thesis from which it has been derived (Leegood 2002). In the chloroplast, the C3; pathway assimilates CO, to form phosphoglycerate (PGA), catalyzed by the enzyme ribulose bispho- sphate carboxylase oxygenase (RUBISCO). The Cy, pathway, observed in 19 plant families (Kellogg 1999; Sage 2004), couples the C3; pathway with a prior carboxylation step cata- lyzed by phosphoenolpyruvate carboxylase (PEP- case), producing four-carbon organic acids such as malate and aspartate. Initial carboxylation by PEPcase and subsequent decarboxylation of the C4, products, acts as a carbon concentrating mechanism that effectively minimizes the oxy- genase activity of RUBISCO and minimizes -photorespiration, increasing quantum yield at the low CO, concentrations (Ehleringer et al. 1991; Sharkey 1988). In most C4 species, the carbon concentrating activity depends upon the presence of Kranz anatomy in leaves. Kranz anatomy is the wreath of radially arranged mesophyll cells surrounding the bundle sheath Haberlandt (1882, 1914). In these species, PEP- *Current Address: USDA, ARS, USHRL, South Rock Road, Fort Pierce, FL 34945. *Author for correspondence (E-mail Iboykin@ macworld.com) 2001 case activity is restricted to mesophyll cells while RUBISCO activity occurs in the bundle sheath cells, in the high CO, microenvironment created by the decarboxylation of C4 products. It has been generally accepted that Kranz anatomy is essential for C4 photosynthesis (Kellogg 1999), but several studies have shown otherwise for submerged monocots (Keeley 1998), submerged dicots (Casati et al. 2000; de Groote and Kennedy 1977; Holaday and Bowes 1980; Lara et al. 2002; Salvucci and Bowes 1983; Spencer et al. 1996), and terrestrial dicots (Freitag and Stichler 2002; Sage 2002; Voznesenskaya et al. 2002; Voznesenskaya et al. 2001). With the exception of a few terrestrial dicot taxa, the majority of species that lack Kranz anatomy are aquatic (Bowes et al. 2002). The aquatic plants that have Cy, photosynthesis without Kranz anatomy are: Hydrilla verticillata (Holaday and Bowes 1980; Magnin et al. 1997; Salvucci and Bowes 1983; Spencer et al. 1996), Egeria densa (Casati et al. 2000; Lara et al. 2002), Elodea canadensis (de Groote and Kennedy 1977), and the tribe Orcuttieae (Keeley 1998). Orcuttieae is a small tribe of semi-aquatic grasses (three genera and nine species) restricted to vernal pools in California and Baja California. Vernal pools result from an unusual combination of soil conditions, summer-dry Mediterranean climate, topography, and hydrology. These pools form in small depressions that are filled by winter precipitation and retain moisture longer than surrounding grasslands before evaporation dur- 144 TABLE 1. MADRONO [Vol. 55 LOCALITY INFORMATION PROVIDED FOR ORCUTTIEAE SAMPLES INCLUDED IN THE LEAF ANATOMY AND CARBON ISOTOPE ANALYSIS. Voucher specimens are deposited at University of New Mexico Herbarium (UNM) (except for Tuctoria fragilis which is located at University of Arizona [ARIZ]). Collectors of plant material were Laura M. Boykin (LMB), Paula M. Hall’ (PMH), and John Reeder. Species Location Orcuttia californica Vasey Riverside, Co. Orcuttia inaequalis Hoover Merced, Co. Orcuttia viscida (Hoover) J. Reeder Orcuttia pilosa Hoover Tehama, Co. Orcuttia tenuis A.S.Hitche. Shasta, Co. Tuctoria greenei (Vasey) J. Reeder Tehama, Co. Tuctoria fragilis (Swallen) J. Reeder B.C.S., Mexico Tuctoria mucronata (Crampton) J. Yolo, Co. Reeder Neostapfia colusana Davy Yolo, Co. ing the hot, dry summer causes the pools to shrink and eventually disappear. As a result, vernal pool habitat represents a continuously changing environment, which contains a special- ized biota and relatively large numbers of threatened and endangered species (Holland and Jain 1977) including the Orcuttieae. Orcut- tieae are unusual among Cy, plants because they spend a large portion of their life cycle as submerged aquatics (Ehleringer and Monson 1993). The presence of C4 photosynthesis in this group may have been favored because it reduces water loss at anthesis when the pools are dry in late summer and plants are subjected to drought and high temperatures (Keeley 1998). C, plants are divided into three biochemical subtypes that differ mainly in the Cy, acid transported into the bundle sheath cells (malate and aspartate) and in the way in which it is decarboxylated; they are named, based on the enzymes that catalyze their decarboxylation, NADP-dependent malic enzyme (NADP-ME) found in the chloroplasts, NAD-dependent malic enzyme (NAD-ME) found in the mitochondria, and phosphoenolpyruvate (PEP) carboxykinase (PCK), found in the cytosol of the bundle sheath cells (Edwards and Walker 1983; Ghannoum et al. 2001; Hatch 1987; Hatch et al. 1975; Jenkins et al. 1989). Anatomically, the three subtypes of C4 photosynthesis differ in chloroplast position and cell outline of the bundle sheath cells (Hattersley and Browning 1981; Hattersley and Watson 1976; Hattersley and Watson 1992; Prendergast and Hattersley 1987). Both NADP-ME and PCK have an uneven outline of the bundle sheath cells, while NAD-ME has an even outline. The chloroplasts in the bundle sheath cells of NADP-ME and PCK have a centrifugal position, while NAD-ME has a centripetal position. Keeley (1998) provided photosynthetic and ana- tomical data for four of the nine species of Orcuttieae: Orcuttia californica, O. viscida, Tuc- toria greenei, and Neostapfia colusana. In addi- tion, he found that all species examined in his Sacramento, Co. Voucher Latitude Longitude LMB & PMH 44 33°43'356" 117°03'3045” LMB & PMH 32 37° 14.818” 120°13.521" LMB & PMH 42 39°31 291" 121°11'345” LMB & PMH 16 39°54.407” 122°58.942” LMB & PMH 62 40°17'115" 122-07 185° LMB & PMH 28 39°54.124” 121 58.963" J. Reeder 7141 | Baja California Sur, Llanos de Hiray LMB & PMH 19 3S: 29509" 121°41.157” LMB & PMH 21 37°14.852” 1207371307 study had the NADP-ME (C, photosynthetic subtype based on chloroplast position and bundle sheath cell outline. The exception, N. colusana, had high activities of both NADP-ME and NAD-ME in pulse-chase experiments indicating a mixed subtype (Keeley 1998). The most unexpected finding of Keeley (1998) was that aquatic leaves of Orcuttia californica and O. viscida carry out C4 photosynthesis but lack Kranz anatomy while terrestrial and floating | leaves of these species carry out C4 photosynthe- sis with Kranz anatomy. The other members of | the tribe included in his study, Tuctoria greenei | and Neostapfia colusana, carry out C4 photosyn- thesis but possess Kranz anatomy in both aquatic | and terrestrial leaves. The photosynthetic and | anatomical diversity that exists among such closely related taxa provides a unique opportu- | nity to study the evolution of single-cell Cy, | photosynthesis. Single-cell C4 photosynthesis in | aquatic plants has generated great interest as the © system that may hold promise for genetic engineering in C3 plants (Leegood 2002). To assess the anatomical and photosynthetic | pathway variation of all species in the Orcuttieae | (excluding Tuctoria fragilis), we surveyed leaf | anatomy at the aquatic, floating and terrestrial | stages using light microscopy and carbon isotope | ratio of leaf tissue. Leaf anatomy and carbon > isotope data for four species of Orcuttieae, O. | inaequalis, O. tenuis, O. pilosa, and T. mucronata | are presented for the first time. METHODS Seed and Soil Collection Seeds, and associated soil, of eight species in — Orcuttieae were collected from vernal pools throughout the geographic distribution of Orcut- tia, Tuctoria, and Neostapfia (Table 1) (federal | permit # TE-029387-0 and California state permit # 00-04). Collections were made from. different populations of the species (Orcuttia | 2008] californica, O. viscida, Tuctoria greenei, and Neostapfia colusana) than those previously sam- pled by Keeley (1998). Exact locality data for all collections are provided in Stone et al. (1988) and the California Natural Diversity Database (CNDDB [CDFG 2007]). The ninth species in Orcuttieae, Tuctoria fragilis, could not be sampled due to the lack of living plants in the only known population in Baja California Sur. Cold stratification of the soil collected from the Llanos de Hiray, the only known location for 7. fragilis, produced no seed germination. Terrestrial leaves for carbon isotope analysis were obtained from herbarium speci- mens provided by J. Reeder. Seed Germination Leaves for anatomy and carbon isotope measurements were obtained for eight species by growing plants from seed modifying the seed germination procedures of Griggs (1974, 1980). Orcuttia tenuis, O. pilosa, and Tuctoria greenei germination requires an initial period of cold stratification in cool, wet conditions followed by a gradual increase in daily temperature to a maximum of 32°C and a minimum of 15°C (Griggs 1974). Experiments with all Orcuttia species and Tuctoria greenei showed that naked seeds cannot be germinated in the laboratory, regardless of treatment (Griggs 1980). A germi- nation rate of 90—-100% is possible when sub- merged inflorescences containing ripe seeds are allowed to become completely engulfed by an aquatic fungus (Griggs 1980). After unsuccessful attempts using calcium carbonate-rich local clay soil in New Mexico, germination of seeds of all species except Tuctoria fragilis was achieved using soil collected from natural vernal pools at the time of seed collection. Whole inflorescences from field-collected ma- terial were placed in petri dishes containing 50 mL of deionized (DI) water and placed at 20°C for three weeks. Soil was placed in 4-inch pots, saturated with DI water, and placed at 20°C for three weeks. Whole inflorescences were buried in the cold-stratified soil after the three-week cold stratification. The pots were submerged in tubs filled with DI water and placed in a greenhouse at the Department of Biology, University of New Mexico. DI water was added to the tubs daily. Approximately four months after the soil and seeds were submerged, the DI water was allowed to evaporate completely and the plants were exposed to the air. Collection of Leaf Material Once plants were established in the greenhouse, leaves were sampled as the different stages became available (aquatic, floating and terrestri- BOYKIN ET AL.: LEAF ANATOMY OF ORCUTTIEAE 145 al). Aquatic leaves were collected three months after the seeds and soil were submerged. Floating leaves were collected the first week after they reached the surface of the water. Terrestrial leaves, which developed entirely in air, were collected approximately two weeks after the tubs had completely dried down. All leaf material was cut into small pieces and fixed for one week in formalin acetic acid (FAA) (Berlyn and Miksche 1976). Dehydration, Infiltration, and Sectioning of Leaf Tissue Fixed leaf material was removed from the FAA and placed in 70% ethanol overnight and changed five times in the 24 hr period. The leaf material was put through an ethanol dehydration series: 95% ethanol for 20 min, twice; and 100% ethanol for 20 min, three times. Ethanol dehy- dration was followed by three washes of pure propylene oxide for 20 min each. Leaf samples were then infiltrated with Epon LX112 resin (Ladds Scientific, Williston). An epon mixture of 4:6 (epon A:epon B) was used in all the infiltration steps after mixtures of 3:7 and 2:8 produced unsatisfactory hardness for cross-sec- tioning. The epon mixture of 4:6 was diluted for the first infiltration step in a 1:1 ratio with propylene oxide for three hours. The tissue was then transferred to a 3:1 mixture of epon (4:6) to propylene oxide and infiltrated overnight. The samples were then placed in molds (Pelco, 105) filled with only epon 4:6 and left for ten hours. The final step was to transfer the leaf material into a new mold filled with epon 4:6 and incubate at 60°C for 18 hr. Sections 7 um thick were cut with an ultra-microtome (Sorvall MT2-B) and stained with a 1:1 mixture of 1% azure II and 1% methylene blue. Light microscopy was used to determine presence of Kranz anatomy. All leaf sections were photographed at final magnifica- tions of 20 and 40 with a Nikon Coolpix 990 attached to a Nikon Eclipse E400 microscope. C/"C Isotope Values 'SC/PC isotope ratios were determined in leaf tissue to assess the photosynthetic pathway. C/"C isotope ratios of Orcuttieae aquatic and terrestrial leaves were obtained using a Delta E mass spectrometer in the Earth and Planetary Sciences Department at the University of New Mexico. 6'°C values, a measure of the carbon composition, were determined on plant samples using standard procedure relative to PDB (Pee Dee Belemnite) limestone as the carbon isotope standard (Bender et al. 1973). Isotope studies have demonstrated that Cy has less negative 6'°C values than those found in C; plants (Bender 1968; Bender 1971; Smith and Epstein 1971). This 146 Aquatic O. inaequalis ah t x * A\, ‘as eres fs T. mucronata Fic. 1. MADRONO Floating [Vol. 55 Terrestrial Transverse sections of aquatic, floating and terrestrial leaves of Orcuttieae. Aquatic leaves were taken at a 20 magnification while floating and terrestrial leaves are a 40 magnification. difference in isotopic composition has become a standard mechanism for distinguishing plant tissues from these two groups, with C3 plants having a ratio of —20 to —35% and C, plants having values of —9 to —14% (Bender 1971). RESULTS Leaf Anatomy Among the eight species of Orcuttieae sampled, Kranz anatomy was absent in aquatic leaves of O. pilosa, O. tenuis, and O. inaequalis but present in floating and terrestrial leaves of these species (Fig. 1). In addition, we confirmed this pattern, previously observed in aquatic and terrestrial leaves by Keeley (1998), for the remaining two species in the genus, O. viscida and O. californica (data not shown, see Keeley 1998 for photos). Kranz anatomy was present in aquatic and terrestrial leaves of 7. mucronata (Fig. 1), and we also confirmed this pattern in Tuctoria greenei and Neostapfia colusana (data not shown, see Keeley 1998 for photos). Considerable morphological and anatomical | variation was present in the aquatic leaves of | Orcuttia. Orcuttia viscida, O. californica, and O. pilosa had linear leaves, while O. inaequalis and O. tenuis had terete leaves, due to folding of the | leaves so the margins nearly touch (Fig. 1). Extensive lacunae were evident in the Orcuttia | aquatic leaves while they were lacking in aquatic | leaves of Tuctoria and Neostapfia species. Chloroplast arrangement was variable in the . terrestrial leaves of Orcuttia (Fig. 1). Orcuttia | inaequalis, O. viscida (not shown), and O. tenuis | exhibited centrifugal arrangement of chloroplasts | in terrestrial leaves, while O. californica (not | shown) and O. pilosa had chloroplasts arranged | around the outer edge of the cell. Floating leaves | of Orcuttia all have centrifugal chloroplast distribution in the bundle sheath cells (Fig. 1 and Keeley 1998). Neostapfia colusana and T. | greenei have centrifugal distribution of chloro- | plasts in the bundle sheath cells (Keeley 1998). Tuctoria mucronata aquatic leaves have the | chloroplasts in the bundle sheath cells distributed | centripetally (Fig. 1). 2008] TABLE 2. STABLE CARBON ISOTOPE RATIOS IN AQUATIC, FLOATING AND TERRESTRIAL LEAVES OF ORCUTTIEAE. Values shown are the mean of five replicates + 0.3 standard error. All dried material excluding 7. fragilis was obtained from greenhouse material. Tuctoria fragilis material was obtained from voucher specimens supplied by Dr. John Reeder. Values represent 61°C values (%o). Species Aquatic Floating Terrestrial Orcuttia californica =12.7 127) lee Orcuttia inaequalis 13.0 = 1238 180) Orcuttia pilosa ae Pe) = 12:6 =12:6 Orcuttia tenuis —12.6 —12.4 =—12,3 Orcuttia viscida = 13.6 =13./ =13:7 Tuctoria greenei =13:.1 n/a —13.4 Tuctoria fragilis =13.0 n/a al We 10 Tuctoria mucronata = 13.6 n/a 136 Neostapfia colusana _—14.3 n/a —14.2 Carbon Isotope Ratio The carbon isotope values for all Orcuttieae from aquatic, floating (only present in Orcuttia) and terrestrial leaves (Table 2) are within the range exhibited by Cy, plants (—9 to —14%). Terrestrial, floating and aquatic leaves of all species of Orcuttia had 6'°C values between —12.3 and —13.7%o. Tuctoria aquatic and terres- trial leaves had 6'°C values between —13.0 and —13.6%o. Neostapfia colusana aquatic and terres- trial leaves had the most negative values (— 14.3 and —14.2%o respectively). DISCUSSION Our findings revealed that Cy, photosynthesis occurs throughout Orcuttieae (Table 2) but with considerable leaf anatomical variation (Fig. 1). Aquatic leaves in all species of Orcuttia carry out C, photosynthesis (Table 2) in the absence of Kranz anatomy (Fig. | and Keeley 1998). The aquatic environment might result in less negative isotope ratios because of diffusion limitations (Osmond et al. 1981; Raven et al. 1982). However, the consistency of the isotopic ratios between aquatic and terrestrial leaves provide strong circumstantial evidence that the underly- ing biochemical pathway is C4. Floating and terrestrial leaves of these same species utilize the C4 pathway with Kranz anatomy. On the other hand, both aquatic and terrestrial leaves of Tuctoria mucronata, T. greenei and Neostapfia carry out Cy, with Kranz anatomy (Fig. 1). Orcuttieae is unique among plants exhibiting single-cell Cy photosynthesis because it produces both an aquatic and a terrestrial phase of growth. A single plant can be submerged for three months with the aquatic leaves lacking Kranz anatomy while three months later (after the water in the pool has evaporated), the same plant produces leaves with Kranz anatomy after exposure to the BOYKIN ET AL.: LEAF ANATOMY OF ORCUTTIEAE 147 terrestrial environment. Aquatic single-cell Cy, photosynthesis has previously been documented from several aquatic species in addition to Orcuttieae: Hydrilla verticillata (Holaday and Bowes 1980; Salvucci and Bowes 1983; Spencer et al. 1996; Magnin et al. 1997), Egeria densa (Casati et al. 2000; Lara et al. 2002), and Elodea canadensis (de Groote and Kennedy 1977). Single-cell C4, photosynthesis also occurs in several terrestrial species most notably Bienertia cycloptera and Borszczowia aralocaspica (Cheno- podiaceae) (Voznesenskaya et al. 2001; Vozne- senskaya et al. 2002; Edwards et al. 2004). Orcuttieae is the only documented group of species that is both aquatic and terrestrial in a given year with attendant photosynthetic and anatomical diversity. The variation observed in chloroplast arrange- ment (Fig. 1) suggests that the biochemical C, subtype according to C4 acid decarboxylases (NADP-malic enzyme, NAD-malic enzyme, phosphoenolpyruvate carboxykinase) may differ within Orcuttieae. Such variation, which is common in plant families and genera (Kellogg 1999), has been documented previously in terres- trial leaves of Orcuttia viscida (Keeley and Rundel 2003) and O. tenuis (Sage et al. 1999) which have a centrifugal (NADP-ME) and centripetal (NAD-ME) chloroplast arrangement, respectively. Additionally, our data showed centrifugal (NADP-ME) and centripetal (NAD- ME) chloroplast arrangement in terrestrial leaves of Tuctoria greenei and T. mucronata, respective- ly. Further research on Orcuttieae should include analysis for type of C4 acid decarboxylases (enzyme assays and western blots) to determine the Cy, photosynthetic subtypes (Edwards et al. 2004). Although there are no apparent performance advantages between NAD-ME and NADP-ME (Voznesenskaya et al. 1999), the chloroplast variation and inferred C4 subtype variation in Orcuttieae (Fig. 1) indicates that there may have been multiple independent origins of the NAD- ME and NADP-ME sub-types in the tribe. A switch in subtype could occur if the C4, pathway were controlled by few regulatory genes, which would make evolutionary changes in the pathway expression relatively easy (Ehleringer and Mon- son 1993; Ku et al. 1996; Monson 1996; Monson 1999). Variation in environmental conditions such as, depth and length of inundation of the vernal pools might also have led to variation in chloroplast position. However, this is not a likely explanation for our results because all samples in this study were exposed to the same conditions during growth and development. The vernal pool environment in which Orcuttia thrives, has apparently favored different patterns of leaf development across leaf types (aquatic, terrestrial and floating) while preserving the 148 carbon-concentrating effects of C4, photosynthe- sis (Keeley 1998). Aquatic plants that utilize the C4 pathway are adapted to reduced levels of COs (Keeley 1998; Keeley and Rundel 2003) that are largely unrelated to long-term patterns of atmo- spheric CO, (Keeley and Rundel 2003). For example, Hydrilla verticillata switches between C3 and C, photosynthesis in response to changes in CO; in the surrounding water driven by biogenic depletion of carbon in dense mats of vegetation (Salvucci and Bowes 1983). Similarly, the occur- rence of Kranz anatomy in floating and terres- trial leaves but not in aquatic leaves suggests that leaf anatomical development responds to the relative abundance of CO; and O; during growth of individual leaves. Besides the relative amount of CO, and O>, around aquatic versus terrestrial leaves, the high diffusive resistance to gases in the aquatic environment may affect leaf anatomical development. Manipulating the CO, and Os; concentrations in water or air during develop- ment of aquatic and terrestrial leaves, respective- ly, could test this hypothesis. Although the advantages of C4 in aquatic leaves are known, the relative benefits associated with the shift in leaf anatomy from aquatic to floating and terrestrial leaves of Orcuttia remain unclear. The dense vegetation, high light and high temperatures in the shallow seasonal pools inhabited by Orcuttieae lead to CQO>-depletion and QO, supersaturation (Keeley and Busch 1984). C4, photosynthesis, with or without Kranz anatomy, is favored because its COs pumping mechanism concentrates CO, around the active site of the enzyme Rubisco, maintaining a high CO;:O, ratio and eliminating photorespiration (Ehleringer and Monson 1993) relative to C3 plants under the low CO;:O, conditions in the pools. Although the reduction in photorespira- tion with Cy, photosynthesis represents a clear advantage over C3 plants in the same environ- ment, any differences in performance between C4 leaves with and without Kranz anatomy await additional studies. One possibility is that C4 without Kranz achieves similar rates of COs fixation with a reduced construction cost com- pared to Kranz anatomy (Sage 2002). The origin of aquatic C4 species varies among taxonomic groups. The presence of C4 photosyn- thesis in Hydrilla and Egeria, which comprise separate lineages of Hydrocharitaceae, indicates more than one origin of C4 within this family. All taxa in the family are aquatic, and Cy, is absent from the nearest sister group (Kellogg 1999), supporting an independent aquatic origin for C4. However, the presence of C4 photosynthesis in aquatic leaves of Orcuttieae show a very different origin, one from a terrestrial C4 ancestor (Keeley 1998). Orcuttieae are one of four tribes in the Chloridoideae, a subfamily of 1400 species, all possess Cy photosynthesis; thus, Orcuttieae likely MADRONO [Vol. 55 were derived from upland Cy, ancestors, and the monophyly of the Orcuttieae are well supported (Boykin 2003; Boykin et al. in revision; Hilu and Alice 2001). Since Orcuttia has many aquatic specializations, including the loss of Kranz anatomy, Reeder (1982) and Keeley (1998) suggest that it is the most derived genus in the tribe. In a study of phylogenetic relationships in the tribe, our data support the derived status of Orcuttia (Boykin 2003). With the data presented here together with that from previous studies (Boykin 2003; Keeley 1998; Reeder 1982) we believe Orcuttieae originated on land and has subsequently become more specialized to an aquatic environment. ACKNOWLEDGMENTS The authors thank Paula Hall for valuable laboratory and field assistance, Jon Keeley and John Reeder for plant material, Tom Turner and Gerry Edwards for valuable comments on the manuscript, Joy Avritt and Jeffrey Lucero for greenhouse assistance, Sergio Flores for field assistance in Baja California. Steve Stricker, Toni Smythe, and Anna Tyler were instrumental in helping process leaf material for cross-sectioning. Zach — Sharp and Viorel Autoderi analyzed leaf samples for carbon isotope data. This study was funded by a P.E.O. fellowship, California Native Plant Society, American — Society of Plant Taxonomists, Botanical Society of America and various sources at University of New Mexico (Springfield and Grove fellowships, GRAC, SRAC, and RPT). This research represents a partial fulfillment of the requirements for the degree of Doctor of Philosophy in Biology at University of New Mexico. — LITERATURE CITED BENDER, M. M. 1968. Mass spectrometric studies of carbon-13 variations in corn and other grasses. Radiocarbon 10:468—472. . 1971. Variations in °C/'?C ratios of plants in relations to the pathway of carbon dioxide fixation. Phytochemistry 10:1239—-1244. , |. ROUHANI, H. M. VINES, AND C. C. BLACK. | 1973. C/?C ratio changes in Crassulacean acid | metabolism plants. Plant Physiology 52:427—-430. BERLYN, G. P. AND J. P. MIKSCHE. 1976. Botanical microtechnique and cytochemistry. Iowa State University Press, Ames, IA. BowEs, G., S. K. RAO, G. M. ESTAVILLO, AND J. B. REISKIND. 2002. C4 mechanisms in aquatic angio- sperms: comparisons with terrestrial C4 systems. Functional Plant Biology 29:379-392. BOYKIN, L. M. 2003. Phylogeny and evolutionary biology | of Orcuttieae (Poaceae: Chloridoideae): analysis of | radiation into a unique amphibious environment. | Ph.D. Dissertation. Biology Department University | of New Mexico, Albuquerque, NM. ——-, L. A. SALTER, AND T. K. LOwREYy. In revision. | Success of Bayesian methods for rooting phyloge- netic trees: a case study using Orcuttieae (Poaceae: Chloridoideae). Systematic Biology. CASATI, P., M. V. LARA, AND C. S. ANDREO. 2000. Induction of a Cy-like mechanism of CO; fixation in Egeria densa, a submerged aquatic species. Plant Physiology 123:1611—1621. 1 ~ 2008] CDFG (CALIFORNIA DEPARTMENT OF FISH AND GAME). 2008. California Natural Diversity Data- base. Biogeographic Data Branch, California Department of Fish and Game, Sacramento, CA. Available at: http://www.dfg.ca.gov/biogeodata/ enddb/. DE GROOTE, D. AND R. A. KENNEDY. 1977. Photo- synthesis in Elodea canadensis Michx. Plant Phys- iology 59:1133—1135. EDWARDS, G. E., V. R. FRANCESCHI, AND E. V. VOZNESENSKAYA. 2004. Single-cell C4 photosyn- thesis versus the dual-cell (Kranz) paradigm. Annual Review of Plant Biology 55:173—196. AND O. A. WALKER. 1983. C3 C4 mechanisms and cellular and environmental regulation of photosynthesis. Blackwell Scientific Publications, London, United Kingdom. EHLERINGER, J. R. AND R. K. MONSON. 1993. Evolutionary and ecological aspects of photosyn- thesis pathway variation. Annual Review of Ecology and Systematics 24:41 1—439. , R. F. SAGE, L. B. FLANAGAN, AND P. W. PEARCY. 1991. Climate change and the evolution of C4 photosynthesis. Trends in Ecology and Evolution 6:95—99. FREITAG, H. AND W. STICHLER. 2002. Bienertia cycloptera Bunge ex Boiss., Chenopodiaceae, an- other C, plant without Kranz tissues. Plant Biology 4:121-132. GHANNOUM, O., S. WAN CAEMMERER, AND J. P. CONROY. 2001. Plant water use efficiency of 17 Australlian NAD-ME and NADP-ME Cz, grasses at ambient and elevated CQO, partial pressure. Australian Journal of Plant Physiology 28:1207— A We GRIGGS, F. T. 1974. Systematics and ecology of the genus Orcuttia (Gramineae). M.A. Thesis. Califor- nia State University, Chico, CA. . 1980. Population studies in the genus Orcuttia (Poaceae). Ph.D. Dissertation. University of Cali- fornia, Davis, CA. HABERLANDT, G. 1882. Vergleichende anatomie des assimilatorischen gewebesystems der pflazen. Jahr- bucher fur Wissenschaftliche Botanik 13:74-188. . 1914. Physiological plant anatomy. Macmil- lian, London, United Kingdom. HATCH, M. D. 1987. C4 photosynthesis: a unique blend of modified biochemistry, anatomy and ultrastruc- ture. Biochemical and Biophysics Acta 895:81—106. , T. KAGAWA, AND S. CRAIG. 1975. Subdivision of C4-pathway species based on differing C4 acid decarboxylating systems and ultrastructural fea- tures. Australian Journal of Plant Physiology 2:111-128. HATTERSLEY, P. W. AND A. J. BROWNING. 1981. Occurrence of the suberized lamella in leaves of grasses of different photosynthetic types. I. In parenchymatous bundle sheaths and PCR (“Kranz”) sheaths. Protoplasma 109:371—401. — AND L. WATSON. 1976. C4 grasses: an anatomical criterion for distinguishing between NADP-malic enzyme species and PCK or NAD- ME enzyme species. Australian Journal of Botany 24:297-308. AND . 1992. Diversification of photosyn- thesis. Pp. 38-116 in G. P. Chapman (ed.), Grass evolution and domestication. Cambridge Universi- ty Press, Cambridge, United Kingdom. BOYKIN ET AL.: LEAF ANATOMY OF ORCUTTIEAE 149 HILu, K. W. AND L. A. ALICE. 2001. A phylogeny of Chloridoideae (Poaceae) based on matK sequences. Systematic Botany 26:386—405. HOLADAY, A. S. AND G. BOWES. 1980. C4 acid metabolism and dark CO; fixation in a submerged aquatic macrophyte (Hydrilla verticillata). Plant Physiology 65:331—335. HOLLAND, R. F. AND S. K. JAIN. 1977. Vernal pools. Pp. 515-553 in M. G. Barbour and J. Major (eds.), Terrestrial vegetation of California. Wiley Intersci- ence, New York, NY. JENKINS, C. L. D., R. T. FURBANK, AND M. D. HATCH. 1989. Mechanism of C4 photosynthesis. Plant Physiology 91:1372—1381. KEELEY, J. E. 1998. C4 photosynthetic modifications in the evolutionary transition from land to water in aquatic grasses. Oecologia 116:85—97. AND G. BUSCH. 1984. Carbon assimilation characteristics of the aquatic CAM plant, /soetes howellii. Plant Physiology 76:525—530. AND P. W. RUNDEL. 2003. Evolution of CAM and Cy, carbon-concentrating mechanisms. Inter- national Journal of Plant Sciences 164:S55—S77. KELLOGG, E. A. 1999. Phylogenetic aspects of the evolution of C4 photosynthesis. Pp. 411-444 in R. F. Sage and R. K. Monson (eds.), C4 plant biology. Academic Press, San Diego, CA. M. S. B., Y. KANO-MURAKAMI, AND M. MATSUOKA. 1996. Evolution and expression of C, photosynthesis genes. Plant Physiology 111:949-957. LARA, M. V., P. CASATI, AND C. S. ANDREO. 2002. CO>-concentration mechanisms in Egeria densa, a submersed aquatic plant. Physiologia Plantarum 115:487-495. LEEGOOD, R. C. 2002. C4 photosynthesis: principles of CO, concentration and prospects for its introduc- tion into C3 plants. Journal of Experimental Botany 53:581—590. MAGNIN, N. C., B. A. COOLey, J. B. REISKIND, AND G. BOwEs. 1997. Regulation and localization of key enzymes during the induction of Kranz-less C,- type photosynthesis in Aydrilla verticillata. Plant Physiology 115:1681—1689. MONSON, R. K. 1996. The use of phylogenetic perspective in comparative plant physiology and developmental biology. Annals of the Missouri Botanic Garden 83:3—16. . 1999. The origins of C4 genes and evolutionary pattern in the C4 metabolic phenotype. Pp. 551-584 in R. F. Sage and R. K. Monson (eds.), C4 Plant Biology. Academic Press, San Diego, CA. OSMOND, C. B., N. VALAANE, S. M. HASLAM, P. UOTILA, AND Z. ROKSANDIC. 1981. Comparisons of '°C values in leaves of aquatic macrophytes from different habitats in Britain and Finland; some implications for photosynthetic processes in aquat- ic plants. Oecologia 50:117—124. PRENDERGAST, H. D. AND P. W. HATTERSLEY. 1987. Australian C4 grasses (Poaceae): leaf blade ana- tomical features in relation to C4 acid decarboxyl- ation types. Australian Journal of Plant Physiology 35:355-382. RAVEN, J., J. BEARDALL, AND H. GRIFFITHS. 1982. Inorganic C-sources for Lemanea, Cladophora and Ranunculus in a fast-flowing stream: measurements of gas exchange and of carbon isotope ratio and their ecological implications. Oecologia 53:68—78. KU, 150 REEDER, J. R. 1982. Systematics of the tribe Orcuttieae (Gramineae) and the description of Tuctoria, new genus. American Journal of Botany 69:1082—1095. SAGE, R. F. 2002. C4 photosynthesis in terrestrial plants does not require Kranz anatomy. Trends in Plant Science 7:283—285. . 2004. The evolution of C4, photosynthesis. New Phytologist 161:341—370. , M. Li, AND R. K. MONSON. 1999. The taxonomic distribution of C4, photosynthesis. Pp. 551-584 in R. F. Sage and R. K. Monson (eds.), C4 plant Biology. Academic Press, San Diego, CA. SALvuccI, M. E. AND G. BOWES. 1983. Two photo- synthetic mechanisms mediating the low photo- respiratory state in submerged aquatic angiosperm. Plant Physiology 73:488—496. SHARKEY, T. D. 1988. Estimating the rate of photo- respiration in leaves. Physiologia Plantarum 73:147-152. SMITH, B. N. AND S. EPSTEIN. 1971. Two categories of C/"C ratios for higher plants. Plant Physiology 47:380—-384. SPENCER, W. E., R. G. WETZEL, AND J. TERRI. 1996. Photosynthetic phenotype plasticity and the role of MADRONO [Vol. 55 phosphoenolpyruvate carboxylase in AHydrilla ver- ticillata. Plant Science 118:1-9. STONE, R. D., W. B. DAVILLA, D. W. TAYLOR, G. L. CLIFTON, AND J. C. STEBBINS. 1988. Status survey of the grass tribe Orcuttieae and Chamaesyce hooveri (Euphorbiaceae) in the Central Valley of California. U.S. Fish and Wildlife Service, Office of Endangered Species, Sacramento, CA., Technical Report. VOZNESENSKAYA, E. V., V. R. FRANCESCHI, O. KiIRATS, E. G. ARTYUSHEVA, H. FREITAG, AND G. E. EDWARDS. 2002. Proof of C4 photosynthesis without Kranz anatomy in Bienertia cycloptera (Chenopodiaceae). The Plant Journal 31:649—-662. , H. FRETZAG, AND G. E. EDWARDS. 2001. Kranz anatomy is not essential. for terrestrial C4, plant photosynthesis. Nature 414:543-546. , V. I. PYANKOV, AND G. E. EDWARDS. 1999. Anatomy, chloroplast structure and com-— partmentation of enzymes relative to photosyn- thetic mechanisms in leaves and cotyledons of | species in the tribe Salsoleae (Chenopodiaceae). | Journal of Experimental Botany 50:1779-1795. MADRONO, Vol. 55, No. 2, pp. 151-158, 2008 DIRECT AND INDIRECT EFFECTS OF HOST PLANTS: IMPLICATIONS FOR REINTRODUCTION OF AN ENDANGERED HEMIPARASITIC PLANT (CASTILLEJA LEVISECTA) BETH A. LAWRENCE*! AND THOMAS N. KAYE!” ‘Oregon State University, Department of Botany and Plant Pathology, Corvallis, OR 97330 * Institute for Applied Ecology, 563 SW Jefferson St., Corvallis, OR 97333 ABSTRACT Rare, parasitic plants pose an interesting challenge to restoration practitioners. In addition to the problems associated with small population size, rare parasites may also be limited by their host requirements. We examined how the performance of a rare Pacific Northwest hemiparasite, Castilleja levisecta, was affected by the availability of different host combinations in the greenhouse and in the field. Castilleja levisecta individuals were grown with two individuals of the grass Festuca roemeri, two individuals of the aster Eriophyllum lanatum, one individual of each of these species (a “‘mixed”’ treatment), or without any host. We did not find support for the complimentary diet hypothesis, which predicts that parasites grown with multiple host species perform better than individuals grown alone or with a single host. In the greenhouse, C. /evisecta individuals grown in the mixed treatment had greater stem growth than those planted with F. roemeri, but did not differ from E. /anatum or no- host treatments. In the field, vole activity had indirect effects on C. /evisecta survival mediated through host species: vole tunneling and C. Jevisecta mortality were strongly associated with host treatments including E. /anatum. Vole tunneling and C. /evisecta mortality were strongly associated with host treatments including E. /anatum. Field survival of no-host and F. roemeri treatments were significantly higher than those grown with E. l/anatum. Our results emphasize the importance of basing conservation decisions on experimental research conducted under conditions similar to those of the intended application, as greenhouse results were a poor predictor of field performance. For restoration of endangered hemiparasitic plants, we recommend planting with hosts that are not attractive to herbivores. Key Words: Castilleja levisecta, complimentary diet hypothesis, host-use, rare hemiparasite, reintroduction. Parasitic plants are a dynamic component of many plant communities capable of altering productivity (Marvier 1998b; Matthies 1997), competitive interactions (Gibson and Watkinson 1991; Matthies 1996), and community structure (Gibson and Watkinson 1992; Press 1998). Although many parasitic plants are agricultural pests, some are of conservation concern and pose an interesting challenge to restoration practition- ers (Marvier and Smith 1997). In addition to the diversity of obstacles typically encountered dur- ing reintroduction, rare parasites may also be limited by host requirements. Uncertainties asso- ciated with parasite host specificity and the availability and quality of hosts at restoration sites are likely to impede parasitic plant reintro- duction efforts (Marvier and Smith 1997). Therefore, successful management of rare para- sites necessitates consideration of their unique biology. We conducted greenhouse and field experiments with Castilleja levisecta Greenman (golden paintbrush), a rare hemiparasite endemic to the prairies of the Pacific Northwest of the * Current address: University of Wisconsin-Madison, Botany Department, 430 Lincoln Drive, Madison, WI 53706; email: balawrence@wisc.edu United States, to evaluate its host preferences in support of recovery actions. Although facultative hemiparasites are photo- synthetic and do not require a host plant, they often form haustoria (1.e., physical connections with other root systems) through which nutrients, water, and secondary compounds are obtained from the host (Kuyt 1969; Press 1989). In natural systems, unattached mature facultative parasites are uncommon, and attachment to a host generally stimulates parasite fitness and growth (Kuit 1969). Most members of the genus Castilleja are considered generalist hemiparasites, capable of parasitizing multiple host species (Dobbins and Kuyt 1973; Heckard 1962). How- ever, the degree to which a host stimulates hemiparasite fitness varies considerably among host species (Chuang and Heckard 1971; Gibson and Watkinson 1992; Marvier 1998b; Matthies 1996, 1997; Seel and Press 1993). Interactions between plant parasites and host species can have direct and indirect effects both on host and parasite performance, as well as their pollinators (Adler et al. 2001), and herbivores (Adler 2002, 2003; Adler et al. 2001; Marko 1996; Marvier 1996). Parasitic plants can acquire secondary compounds from host species (Govier et al. 1967; Schneider and Stermitz 1990; Stermitz and Harris 152 1987), which in turn can alter species interactions. For example, acquisition of alkaloids from the host Lupinus albus directly reduced insect herbiv- ory of Castilleja indivisa, and indirectly increased pollination (Adler et al. 2001). In the field, hemiparasitic plants often parasit- ize several hosts simultaneously (Gibson and Watkinson 1989; Matthies 1996). Generalist hemiparasites may perform better on a mixed diet relative to a homogenous diet due to improved nutrient balance and/or dilution of toxic secondary plant compounds (Marvier 1998a). Many taxa benefit from multiple food sources, including some insects (Bernays et al. 1994), gastropods (Pennings et al. 1993), and reptiles (Bjorndal 1991). Therefore, we propose that providing multiple nutrient sources increases individual fitness (the complimentary diet hy- pothesis). We test this hypothesis by comparing the size and survival of a rare hemiparasitic plant in mixed host, single host, and no-host plantings. Castilleja levisecta is a federally threatened species and is currently restricted to eleven populations in the Pacific Northwest. The species is extinct in the southern portion of its historic range, including the Willamette Valley, Oregon. Federal recovery criteria for C. /evisecta call for the existence of 20 populations composed of 1000 flowering individuals (USFWS 2000). However, the species has limited capacity for natural dispersal and colonization of new sites, necessi- tating ex situ conservation techniques to meet recovery goals. Thus, a strategic reintroduction plan has been prepared to support the long-term viability of C. /evisecta and requires the estab- lishment of new populations within its historic range (Caplow 2004). Although several studies have investigated C. /evisecta host use (Pearson and Dunwiddie 2006; Wayne 2004; Wentworth 2001), clarification of its host dynamics in a restoration context is necessary before large scale reintroduction efforts are pursued. While C. levisecta does not require a host to reproduce in a greenhouse environment and does not appear to be host specific (Wentworth 2001), evidence suggests planting C. /evisecta in the field with a perennial host increases size and repro- ductive output (Pearson and Dunwiddie 2006; Wayne 2004). Greenhouse observations suggest that C. /evisecta can form haustorial connections with several perennial prairie species [e.g., Leu- canthemum vulgare Lam., Eriophyllum lanatum (Pursh) Forbes, Festuca roemeri (Pavlick) Alex- eev, and Fragaria vesca L.], and with itself when grown alone (Kaye 2001; Wentworth 2001). Field experiments indicate that outplanting C. /evisecta with F. roemeri increases the number of inflores- cences produced compared to no-host controls, although host presence did not affect field survival rates (Wayne 2004). In addition, C. levisecta is frequently eaten by small mammals MADRONO [Vol. 55 (Caplow 2004; Wayne 2004), but host-mediated effects of herbivory on the species has not previously been evaluated. Here, we use greenhouse and field studies to test the complimentary diet hypothesis and examine how host-interactions affect herbivory by rodents under field conditions, as well as provide recommendations for future C. levisecta recovery efforts. METHODS Study species Castilleja levisecta (Orobanchaceae, formerly classified in Scrophulariaceae) is a short-lived (5— 6 yr), multi-stemmed, perennial endemic to the native grasslands of the Western Pacific North- west United States. It is an out-crossing species primarily pollinated by Bombus spp. and is known only to reproduce by seed (Kaye and Lawrence 2003; Wentworth 2001). The eleven remaining C. levisecta populations are concen- trated in the San Juan Archipelago of the Puget Trough eco-region, and are found on sandy, well drained soils of glacial origin (Chappell and . Caplow 2004). Despite the rarity of this species, the remaining populations maintain unusually high levels of genetic diversity compared with © other endemic species and members of the | Orobanchaceae (Godt et al. 2005). Greenhouse experiment To test for differences in C. levisecta perfor- mance when grown with different host combina- tions, we randomly assigned individuals to one of | four host treatments, including no-host (control), | two E. lanatum (Asteraceae) individuals, two F. roemeri (Poaceae) individuals, or one individual » of each of these host species (““mixed’’). We used plant material from two C. Jevisecta source | populations located on Whidbey Island, WA (Ebey’s Landing: 48°13'35’N, 122°46'00"W and Forbes Point: 48°16'15”N, 122°37'35”’W). Ap- proximately twenty host treatment replicates from each of these source populations were used | to test our hypotheses, for a total of 39 replicates » per host treatment (n lanatum and F. roemeri were used as host plants 156). Eriophyllum © because C. Jevisecta forms haustorial connections ‘ with these native perennials (Wayne 2004, Beth Lawrence, pers. obs.), they are common at southern extant populations (Chappell and Ca- | plow 2004), and are likely to be present at. reintroduction sites. Castilleja levisecta capsules were collected from | 17 maternal plants from each of the two source » populations in August 2003 to provide seeds for this experiment and were germinated using the methods outlined in Lawrence and Kaye (2005). | 2008] On 1 December 2003, C. Jevisecta germinants were planted into cell flats in a well-drained medium amended with slow release micro- and macro- nutrients and were placed in a greenhouse with 400 watt high pressure sodium lights and temperature fluctuating every 12 hr (12°C /18°C). A randomized block design was implemented to assign host treatments to C. /evisecta individuals, with source population and maternal line serving as the blocking factors. Two maternal lines from Ebey’s Landing and three from Forbe’s Point were assigned to two blocks, because extra plants from these maternal lines were available. Plants were repotted into 3.8 L pots with their assigned host treatment on 28 January 2004. Castilleja levisecta individuals and potential hosts were planted in a triangle with all plants 10 cm apart; C. levisecta individuals assigned the no-host treatment were planted in the center of the pot. We used a no-host control rather than planting three C. Jevisecta individuals together because we had a limited number of plants. Eriophyllum lanatum plants were rooted cuttings from Will- amette Valley genetic stock provided by Heritage Seedling Co., Salem, OR. We used F. roemeri individuals grown from Willamette Valley seed that were one year old when paired with C. levisecta. We attempted to equalize above- and below-ground biomass of provided hosts by trimming them with shears. Plants were random- ized on greenhouse benches and fertilized bi- weekly with a liquid 15-30-15 fertilizer to encourage growth and establishment. We recorded total stem length, stem number, and number of flowers produced by each C. levisecta individual in May 2004, approximately 15 wk after potting the hemiparasites and hosts together. Flowering had finished at this time, so Our measurements are considered estimates of total flower production. Plants were moved to a shade-house in June 2004 and received supple- mental water throughout the summer. Field experiment To test our host and herbivore hypotheses under field conditions, we transplanted the same potted plants with hosts used in the greenhouse study to an upland prairie on 1 December 2004. Our field site was located at Pigeon Butte, Finley National Wildlife Refuge, OR (44°23'54’N, 123°19'11”W), in habitat likely to be used for future C. /evisecta recovery efforts in the Wil- lamette Valley. The site had a high diversity of native perennials and abundant non-native pas- ture grasses (e.g., Festuca arundinacea Schreb. and Arrhenatherum elatius (L.) P. Beauv. ex J. Pres] & C. Presl). It was situated on the shoulder of a butte at 150 m elevation, dominated by silty- clay-loam soils. Average annual precipitation in this region is approximately 115 cm, with average LAWRENCE AND KAYE: CASTILLEJA LEVISECTA HOST USE 153 annual minimum and maximum temperatures of 5°C and 17°C, respectively (WRCC 2005). We randomly planted host-parasite replicates (each pot was a replicate) into the center of a 1 m* plot within a 10 X 15-m grid fenced to exclude deer. Deer are frequent herbivores of the species and threaten extant populations; fences to exclude deer have been built at two of the extant C. levisecta populations (Beth Lawrence, pers. obs.). Castilleja levisecta individuals and hosts were dormant at the time of outplanting, and senesced material was removed. A_ balanced design could not be executed in the field because some greenhouse plants died during the previous summer. However, at least 22 replicates of each of the four host treatments were transplanted into the field (no-host, n = 39; F. roemeri, n = 31; E. lanatum, n = 26; mixed, n = 22). Field transplants were monitored in early June 2005 because surveys conducted in a companion 2004 field study revealed that transplants were at their maximum size and peak inflorescence at this time (Lawrence 2005). Vole abundance was unusually high throughout the Pacific Northwest during the 2005 growing season and all surviving C. levisecta individuals at the field site were subjected to herbivory, most likely from grey- tailed voles (Microtus canicaudus) (Beth Law- rence, pers. obs.). Stem length and/or number, as well as flower and/or seed production were not reliable measures of C. /evisecta performance, as herbivory appeared to stimulate resprouting, alter plant morphology, and prevent individuals from flowering (Beth Lawrence, pers. obs.). Therefore, we used C. Jevisecta survival as the response variable for the field component of our study. Vole tunneling was also very frequent, indicating herbivore pressure occurred in the root zone as well as above ground. Tunnels were unevenly distributed throughout the study area, so herbivore pressure by voles was measured as presence or absence of tunnels within 15 cm of the transplant root crown. Statistical analyses We used multivariate analysis of variance (MANOVA) with a Wilks’ lambda multivariate F test to simultaneously test for differences in C. levisecta greenhouse response variables (stem length, stem number, and flower number) among host treatments. Prior to MANOVA analysis, stem number was log-transformed to improve homoscedasticity. Castilleja levisecta source pop- ulation and maternal effects were used as blocking factors in this analysis because differential growth among populations and individuals from different maternal lines has been observed in this species (Kaye 2001). However, we focus our analysis on host treatment effects. Following MANOVA, univariate ANOVAs were conducted 154 TABLE 1. MADRONO [Vol. 55 RESULTS FROM UNIVARIATE ANOVAS TESTING THE EFFECT OF HOST TREATMENT (HOST), SOURCE POPULATION (SOURCE), MATERNAL LINE (MATLINE), AND THE HOST TREATMENT*SOURCE POPULATION INTERACTION (HOST*SOURCE) ON THE NUMBER OF C. LEVISECTA FLOWERS PRODUCED (FLOWER #), NUMBER OF STEMS (STEM #), AND TOTAL STEM LENGTH (STEM LENGTH) IN THE GREENHOUSE. Significant effects at « = 0.05 denoted with *. RESPONSE EFFECT DF Flower # host 3 source ] matline 32 host*source 3 Stem # host 3 source 1 matline 32 host*source 3 Stem Length host 3 source 1 matline 32 host*source BS on the three greenhouse response variables. Signif- icant ANOVAs were followed by Tukey’s HSD for pairwise comparisons among host treatments. We used binary logistic regression to test for differences among host treatments in C. /evisecta field survival and vole tunnel presence. Signifi- cance was measured by drop in deviance (DEV) with a chi-square distribution. Dunn-Sidak cor- rections were used to adjust alpha levels for all pair-wise comparisons among host treatments. We used linear regression to determine if the proportion of transplants with vole tunnels was associated with the proportion of C. /evisecta individuals surviving. Finally, we calculated an odds ratio to compare C. /evisecta survival when planted with EF. /anatum versus survival when not planted with this species (1.e., alone or with F- roemeri). All analyses were conducted using S- PLUS v. 6.2 (Insightful 2000). RESULTS Greenhouse experiment According to MANOVA analyses, Castilleja levisecta greenhouse performance differed among host treatments (Wilkes = 0.84, F3 116 = 2.37, P = (0.014) as well as source populations (Wilkes = 0.55, Fi 116 = 31.47, P = 0.001), but no differences were observed among maternal lines (Wilkes <= 0.42, F32 116 = 1.19, P = 0.13). Host treatment effects were consistent among C. levisecta individuals from the two source popu- lations used in this study, as the interaction between source population and host treatment was not significant (Wilkes = 0.93, F3 116 = 0.92, P = 0.51). While C. levisecta stem number and total stem length differed among host treatments, the number of flowers did not (Table 1). Univar- late ANOVA statistics for the three response variables are presented in Table 1. Post-hoc pair wise comparisons of univariate ANOVAs re- MS F Pr 474 0.84 0.47 sot3 32:32 <0.001* 877 1:55 0.048* 38 0.067 0.98 0.24 2.87 0.039* 2.61 31.76 <0.001* 0.097 119 0:25 0.0069 0.084 0.97 TAZ 3.69 0.014* 197 0.10 0.75 3689 1.9 0.0067* 1879 0:97 0.41 vealed that individuals grown with mixed hosts had a greater number of stems and total stem length compared to those grown with F. roemeri, but did not differ from those grown without a host or with E. lanatum. Field experiment Field survival of C. /evisecta differed among host treatments (DEV33; = 44.65, P < 0.001), but neither source population (DEV, 3; = 0.089, P = 0.77) nor maternal line (DEV32.8; = 34.43, P = (0.40) accounted for a significant portion of the residual deviance. A higher proportion of no-host C. levisecta individuals survived compared to those planted with either E. /anatum or mixed hosts, but did not differ from plants with F- roemeri hosts (Fig. 1). Also, C. levisecta planted with F. roemeri hosts had significantly higher survival than those planted with E. /anatum (Fig. 1). Rodent tunnel presence near transplant root crowns differed significantly among host treat- ments (DEV3 114 a 50.17, P= 0.001). Castilleja levisecta individuals planted with F. roemeri or without a host had fewer rodent holes near their root crowns compared to those planted with either E. lanatum or mixed hosts (Fig. 1). In addition, we measured a strong inverse relation- ship between C. /evisecta survival and the presence of tunneling within the vicinity of the root crown (F,.> = 23.07, P = 0.04, R* = 0.92) (Fig. 2). The odds of a C. /evisecta transplant surviving in the field when planted without an E. lanatum host were 11.25 (95% C.I. = 4.29, 28.78) times greater than when co-planted with an E. lanatum host. DISCUSSION We did not find support for the complimentary diet hypothesis, which predicts that individuals with multiple nutritional sources will perform 2008] 2 2 © © oOo +-& Ha @ proportion surviving oO no host F. roemeri E lanatum mixed ii) 1 a a b b ry E 08 = S 0.6 Cc s 04 rc o o 0.2 Qa. 0 no host F. roemeri E lanatum mixed Fic. 1. 1) Castilleja levisecta field survival by host treatment. 11) Proportion of C. Jevisecta transplants located within 15 cm of rodent tunnels. Host treatments not sharing a common letter were significantly different (P = 0.05) after Dunn-Sidak corrections. better than those provided with a limited diet. Mixed hosts improved some measures (i.e., stem number and total stem length) of C. Jevisecta greenhouse performance compared to those paired with F. roemeri, but did not confer an advantage over no-host or E. /anatum treatments. Likewise, mixed hosts did not promote C. levisecta field survival. In fact, no-host controls had greater field survival than both mixed and E. lanatum treatments (Fig. 1). The complementary diet hypothesis may not be the most appropriate theory to apply to hemiparasite nutrition, as this hypothesis has primarily been tested in animal systems. Other studies addressing hemiparasite fitness using multiple hosts have also found mixed results. During greenhouse studies, Mel- ampyrum arvense did not benefit from mixed hosts (a legume and a grass) (Matthies 1996), though Castilleja wightii growth and reproductive output were improved by simultaneous attach- ment to two host species (a legume and an aster) (Marvier 1998a). However, greenhouse studies may oversimplify field dynamics and should be extrapolated to the field with caution. For example, we observed strong indirect effects of herbivory mediated by host species in the field, which has important consequences for C. levi- secta recovery efforts. While we observed improved C. levisecta stem performance when grown with mixed hosts relative to F. roemeri in the greenhouse, the number of flowers produced did not differ among host treatments. Mixed hosts may have improved C. levisecta nutrition by providing complimenta- ry resources, thereby improving stem growth LAWRENCE AND KAYE: CASTILLEJA LEVISECTA HOST USE 155 R*=0.92, P=0.04 proportion surviving 0 0.2 0.4 0.6 0.8 1 proportion near tunnel FIG. 2. Scatterplot and trendline from linear regres- sion of the average proportion of C. levisecta trans- plants within 15cm of a vole tunnel and average transplant survival for each host treatment (no host = @, F. roemeri = *, mixed = #, E. lanatum = ©). relative to F. roemeri hosts. Alternatively, root competition may explain why F. roemeri is a poor host in pots, as pots with F. roemeri were generally more root bound than other host treatments (Beth Lawrence, pers. obs.). This is consistent with our previous work that found C. levisecta grown in pots with F. roemeri were smaller and flowered less frequently in the second growing season compared to those potted with E. lanatum (Kaye 2001). This explanation is more plausible, as C. Jevisecta performed similarly among all other treatments, but did poorly when paired just with F. roemeri. Our greenhouse results may also have been confounded by our judicious use of fertilizer, as attachment to hosts may not confer fitness benefits in the presence of abundant nutrients. Vole activity had strong indirect effects on C. levisecta field survival mediated by host species. Populations of the grey-tailed vole (Microtus canicaudus) were larger than average in the Willamette Valley during the 2005 field season due to a mild winter in 2004-05, increasing herbivore pressure on C. /evisecta transplants and impacting the region’s grass seed crop. Nine Oregon counties were declared agricultural disas- ter areas by the U.S. Department of Agriculture due to large crop losses from voles (A.P. 2005). Although population sizes in 2005 were atypically large, voles are ubiquitous in Pacific Northwest prairies and are major herbivores contributing to grassland dynamics (Wilson and Carey 2001). Further, global warming may increase the frequency of mild winters in the Pacific North- west (Leung and Ghan 1999) and result in greater regularity of vole outbreaks. Selective herbivory by voles in other grassland systems has been shown to dramatically alter species composition and diversity (Batzli and Pitelka 1970; Howe and Lane 2004). Using exclusion experiments, Howe 156 and Lane (2004) observed that meadow voles eliminated otherwise common plants due to preferential herbivory. Castilleja levisecta field survival also did not support the complimentary diet hypothesis, possibly as a result of indirect effects from herbivore activity. While herbivory was evident on all surviving C. /evisecta individuals at the study site, vole tunneling and field mortality were strongly associated with host treatments that included E. /anatum, whose roots may have been particularly palatable to voles. Castilleja levisecta plants paired with two E. /anatum individuals had higher field mortality than those planted with a single E. lanatum individual (mixed host), al- though these effects were not strictly additive (Fig. 1). The mechanism contributing to high mortality of C. /evisecta individuals associated with E. Janatum is unclear, but root system disturbance, direct grazing of C. levisecta roots, or the indirect effect of reduced host vigor/ survival likely contributed to this observation. Meanwhile, C. /evisecta individuals planted with- out a host or with F. roemeri had much higher survival rates and less rodent tunneling. This indicates that voles did not just target potting soil or areas with low root density to tunnel in, but were specifically attracted to E. lanatum. Foliage and roots of plants in the genus Eriophyllum contain sesquiterpene lactones (Bohlmann et al. 1981), an extremely diverse group of compounds that may be desirable to herbivores due to anti- fungal, anti-bacterial, anti-tumourgenic, or anti- inflammatory properties (Picman 1986), and may have contributed to increased vole tunneling in the vicinity of E. lanatum. Although we have provided evidence that vole activity mediated C. /evisecta survival through host species in the field (Fig. 2), an alternative process could be responsible for the observed field patterns. Due to a malfunction of the automatic watering system, we observed differ- ential survival of the potted plants during the 2004 summer in the shadehouse. Survival was greater among no-host (100%) and F. roemeri (79.5%), than among E. Janatum (66.7%) and mixed hosts (54.4%), similar to the pattern of differential survival we observed in the field. Thus it is possible that our field survival rates were only spuriously correlated with vole activity and that survival of C. /evisecta is directly influenced by host treatment, rather than indirectly via herbivore activity. Results from our greenhouse and field studies suggest that planting C. /evisecta with a host may not be absolutely necessary, but may confer some advantages to field plantings. Although our findings are likely context dependent, no-host controls performed as well or better than all other host treatments in both greenhouse and field environments. Host plants can provide water and MADRONO nutrients to hemi-parasites during periods of critical environmental stress (Kuijt 1969; Press 1989). However, under horticultural growing — conditions with ample water, nutrients, and light, © C. levisecta individuals produced abundant bio- | mass and had high reproductive output without hosts. In our field study, no-host C. levisecta individuals had the highest proportion surviving (x = 0.78), although at the time of monitoring these plants had yet to experience summer [Vol. 55 drought conditions typical of the region. Natural — populations of C. /evisecta emerge in early March, flower in May, and senesce in July in response to dry conditions (Caplow 2004). Summer drought is a strong selective force > resulting in substantial C. Jevisecta transplant — mortality, as field survival is typically high the | first growing season, but is generally reduced the © second growing season (Lawrence 2005; Pearson and Dunwiddie 2006; Swenerton 2003; Wayne 2004). Results from a companion common > garden experiment indicate that planting a_ perennial host with C. /evisecta transplants improves second year survival (Lawrence 2005). | Second year survival at a site where individuals | were planted with F. roemeri was particularly high (X = 0.75), compared to the average proportion surviving at the other nine common > gardens (X = 0.21), that were not provided a host. Pearson and Dunwiddie (2006) observed greater C. levisecta flower production when grown with | E. lanatum compared with F. roemeri, but field | survival was greater with F. roemeri. Another field experiment also observed greater C. levisecta | survival when outplanted with F. roemeri relative — to no-host and EF. /anatum treatments (S. Reich- ard 2005, University of Washington, pers. comm.). Castilleja hispida Benth., which can | hybridize with C. /evisecta, also had higher field survival when planted with F. roemeri than when planted with no host (Schmidt 1998). These observations suggest that planting a perennial > host with C. Jevisecta in the field is beneficial, and may allow the parasite to take advantage of host roots to exploit nutrients and water from a larger volume of soil during periods of environmental stress. Further, this suggests that C. Jevisecta | survival is higher when planted with F. roemeri than E. /anatum, and this may be due, at least in part, to preferential vole herbivory of E. /anatum. | Other native perennial species, including le- | gumes and showy angiosperms that can attract | pollinators, may also be appropriate hosts for C. levisecta. Leguminous hosts are commonly better | hemiparasite hosts than grass species because of their capacity to fix nitrogen (Adler 2003; Gibson and Watkinson 1991; Matthies 1997; Seel and Press 1993). Additionally, alkaloid uptake from | leguminous hosts can confer resistance to herbiv- | ory (Adler 2002), and increase pollinator visita- | tion (Adler et al. 2001). Although the mycorrhizal 2008] status of C. levisecta has not been investigated, many hemiparasites in the Orobanchaceae are considered non-mycorrhizal (Harley and Harley 1987). The mycorrhizal status of the host plant however, can influence the performance of the hemiparasite. Studies have shown that hemipar- asites attached to mycorrhizal hosts have greater biomass and flower production than those growing with non-mycorrhizal hosts (Davies and Graves 1998; Salonen et al. 2001). We suggest that new C. /evisecta potential host species and mycorrhizal inoculation of hosts should be examined experimentally in the field to further examine the complimentary diet hypothesis and improve the success of large-scale reintroductions of this endangered species. Implications for practice ¢ Conservation decisions should be based on experimental research conducted under condi- tions similar to those of the intended applica- tion; C. levisecta greenhouse performance was a poor predictor of field survival. Extrapola- tion of greenhouse results to natural systems can oversimplify the complex biotic interactions that species are exposed to in the field, and worse, suggest inappropriate management actions. e Greenhouse propagation of endangered hemi- parasites like C. levisecta may not require a host, but growth and survival after field planting may be improved by planting with additional species. See (Lawrence and Kaye 2005) for details on propagation techniques for C. levisecta. e Failure to find support for the complimentary diet hypothesis with C. levisecta suggests that outplanting rare hemi-parasites with multiple hosts may not be necessary. ¢ We recommend against planting hemiparasites with hosts that are attractive to herbivores when and where these animals are present. We suspect planting C. Jevisecta with a perennial host will increase future field performance and recommend using F. roemeri over E. lanatum as a host for C. levisecta recovery efforts. © Herbivore management should be an integral part of rare hemiparasite recovery and manage- ment. Herbivore control may involve the same actions as prairie habitat management, such as mowing or burning to reduce the accumulation of thatch. Large fences can be erected to exclude ungulate browsers from an outplanting, while small wire cages dug into the ground can prevent rodent grazing of individual plants. ACKNOWLEDGMENTS We would like to thank the USFWS Castilleja _ levisecta Recovery Team, Institute for Applied Ecology, Jock Bealle, and Amy Bartow, for advice and logistical LAWRENCE AND KAYE: CASTILLEJA LEVISECTA HOST USE Loy support. Bruce McCune and Manuela Huso provided statistical guidance. Drafts of this work were reviewed by Matt Blakely-Smith, Dave Pyke, and Aaron Liston, as well as two anonymous reviewers. This research was funded by a Native Plant Society of Oregon field grant, Bonnie Templeton Award for Plant Systematics, a National Science Foundation Graduate Research Fel- lowship, and USFWS contract 101813M465S. LITERATURE CITED A.P. 2005. Voles head for Oregon vineyards, orchards. Available at: http://www.wtopnews.com/index. php?nid=220&sid= 594057; accessed 15 November 2005. ADLER, L. S. 2002. Host effects on herbivory and pollination in a hemiparasitic plant. Ecology 83: 2700-2710. . 2003. Host species affects herbivory, pollina- tion, and reproduction in experiments with para- sitic Castilleja. Ecology 84:2083—2091. , R. KARBAN, AND S. Y. STRAUSS. 2001. Direct and indirect effects of alkaloids on plant fitness via herbivory and pollination. Ecology 82:2032—2044. BATZLI, G. O. AND F. A. PITELKA. 1970. Influence of meadow mouse populations on California grass- land. Ecology 51:1027—1039. BERNAYS, E. A., K. L. BRIGHT, N. GONZALEZ, AND J. ANGEL. 1994. Dietary mixing in a generalist herbivore: tests of two hypotheses. Ecology 75: 1997-2006. BJORNDAL, K. A. 1991. Diet mixing: nonadditive interations of diet items in an omnivorous fresh- water turtle. Ecology 72:1234-1241. BOHLMANN, F., C. ZDERO, J. JAKUOVIC, H. ROBIN- SON, AND R. M. KING. 1981. Eriolanolides, eudesmanolides and rearranged sesquiterpene from Eriophyllum species. Phytochemistry 20:2239—2244. CAPLOW, F. 2004. Reintroduction plan for golden paintbrush (Castilleja levisecta). Washington Nat- ural Heritage Program, Department of Natural Resources, Olympia, WA. CHAPPELL, C. B. AND F. CAPLOwW. 2004. Site charac- teristics of golden paintbrush populations. Wash- ington Natural Heritage Program, Department of Natural Resources, Olympia, WA. CHUANG, T. I. AND L. R. HECKARD. 1971. Observa- tions on root-parasitism in Cordylanthus (Scroph- ulariaceae). American Journal of Botany 58:218—228. DAvIES, D. M. AND J. D. GRAVES. 1998. Interactions between arbuscular mycorrhizal fungi and the hemiparasitic angiosperm Rhinanthus minor during co-infection of a host. New Phytologist 139: 555-563. DosBINs, D. R. AND J. Kut. 1973. Studies on the haustorium of Castilleja (Scrophulariaceae). I. The upper haustorium. Canadian Journal of Botany 51:917-922. GIBSON, C. C. AND A. R. WATKINSON. 1989. The host range and selectivity of a parasitic plant: Rhi- nanthus minor L. Oecologia 78:401—406. AND . 1991. Host selectivity and the mediation of competition by the root hemiparasite Rhinanthus minor. Oecologia 86:81—87. AND . 1992. The role of the hemiparasitic annual Rhinanthus minor in determining grassland community structure. Oecologia 89:62—68. 158 GopT, M., F. CAPLOW, AND J. L. HAMRICK. 2005. Allozyme diversity in the federally threatened golden paintbrush, Castilleja levisecta (Scrophular- iaceae). Conservation Genetics 6:87—99. GOVIER, R. N., M. D. NELSON, AND J. S. PATE. 1967. Hemiparasitic nutrition in angiosperms. I. The transfer of organic compounds from host to Odontites verna (Bell.) Dum. (Scrophulariaceae). New Phytologist 66:285—297. HARLEY, J. L. AND E. L. HARLEY. 1987. A check-list of mycorrhiza in the British flora. New Phytologist 105:1—102. HECKARD, L. R. 1962. Root parasitism in Castilleja. Botanical Gazette 124:21. Howe, H. F. AND D. LANE. 2004. Vole-driven succession in experimental wet-prairie restorations. Ecological Applications 14:1295—1305. INSIGHTFUL. 2000. S-PLUS 2000 for Windows. In- sightful, Inc., Seattle, WA. KAYE, T. N. 2001. Restoration research for golden paintbrush (Castilleja levisecta), a threatened spe- cies. Institute for Applied Ecology, Corvallis, OR. AND B. A. LAWRENCE. 2003. Fitness effects of inbreeding and outbreeding on golden paintbrush (Castilleja levisecta): implications for recovery and reintroduction. Institute for Applied Ecology, Corvallis, OR. KUIJUT, J. 1969. The biology of parasitic flowering plants. University of California Press, Berkeley, CA. LAWRENCE, B. A. 2005. Studies to facilitate the reintroduction of golden paintbrush (Castilleja levisecta) to the Willamette Valley, Oregon. M.S. thesis, Oregon State University, Corvallis, OR. AND T. N. KAYE. 2005. Growing Castilleja for restoration and the garden. Rock Garden Quarter- ly 63:128-134. LEUNG, L. R. AND S. J. GHAN. 1999. Pacific Northwest climate sensitivity simulated by a regional climate model driven by a GCM. Part II: 2x CO2 simulations. Journal of Climate 12:2031—2053. MARKO, M. D. 1996. Hemiparasitism by Castilleja sulphurea: alkaloid incorporation and herbivore response. M.S. thesis, Colorado State University, Fort Collins, CO. MARVIER, M. A. 1996. Parasitic plant-host interac- tions: plant performance and indirect effects on parasite-feeding herbivores. Ecology 77:1398—1409. . 1998a. A mixed diet improves performance and herbivore resistance of a parasitic plant. Ecology 79:1272-1280. . 1998b. Parasite impacts on host communities: plant parasitism in a California coastal prairie. Ecology 79:2616—2623. AND D. L. SMITH. 1997. Conservation implica- tions of host use for rare parasitic plants. Conservation Biology 11:839-848. MATTHIES, D. 1996. Interactions between the root hemiparasite Melampyrum arvense and mixtures of host plants: heterotrophic benefit and parasite- mediated competition. Oikos 75:118—124. . 1997. Parasite-host interactions in Castilleja and Orthocarpus. Canadian Journal of Botany 75:1252-1260. MADRONO [Vol. 55 PEARSON, S. F. AND P. DUNWIDDIE. 2006. Experimen- tal seeding and outplanting of golden paintbrush (Castilleja levisecta) at Glacial Heritage, Mima Mounds, and Rocky Prairie, Thurston County, WA. USS. Fish and Wildlife Service, Olympia, WA. PENNINGS, S. C., M. T. NADEAU, AND V. J. PAUL. 1993. Selectivity and growth of the generalist herbivore Dolabella auricularia feeding upon com- plementary resources. Ecology 74:879-890. PICMAN, A. K. 1986. Biological activities of sesquiter- pene lactones. Biochemical Systematics and Ecol- ogy 14:255-—281. PREss, M. C. 1989. Autotrophy and heterotrophy in root hemiparasites. Trends in Ecology & Evolution 4:258-263. . 1998. Dracula or Robin Hood? A functional role of root hemiparasites in nutrient poor | ecosystems. Oikos 82:609-611. SALONEN, V., M. VESTBERG, AND M. VAUHNKONEN. 2001. The effect of host mycorrhizal status on host | plant-parasitic plant interactions. Mycorrhiza 11:95—100. SCHMIDT, D. 1998. Restoration of a prairie ecosystem | at the Yellow Island preserve and the propagation of Castilleja hispida by vegetative cutting. M.S. | thesis, University of Washington, Seattle, WA. SCHNEIDER, M. J. AND F. R. STERMITZ. 1990. Uptake © of host plant alkaloids by root parasitic Pedicularis — species. Phytochemistry 29. SEEL, W. E. AND M. C. PRESS. 1993. Influence of the | host on three sub-Artic annual facultative root | hemiparasites. New Phytologist 125:131—138. STERMITZ, F. R. AND G. H. HARRIS. 1987. Transfer of pyrrolizidine and quinlizidine alkaloids to Castil- leja (Scrophulariaceae) hemiparasites from com- | posite and legume host plants. Journal of Chemical | Ecology 13:1917—1925. SWENERTON, K. K. 2003. Soil suitability and site | preparation techniques for Castilleja levisecta ° restoration on Whidbey Island, Washington, M.S. | thesis, University of Washington, Seattle, WA. USFWS. 2000. Recovery plan for golden paintbrush (Castilleja levisecta). U.S. Fish and Wildlife Service, Portland, OR. WAYNE, W. C. 2004. Factors affecting the reintroduc- tion of golden paintbrush (Castilleja levisecta), a threatened plant species. M.S. thesis, University of | Washington, Seattle, WA. WENTWORTH, J. B. 2001. The demography and | population dynamics of Castilleja levisecta, a. federally threatened perennial of Puget Sound Grasslands. Pp. 49-51 Plant Society, Seattle, WA. WILSON, S. M. AND A. B. CAREY. 2001. Washington. Northwest Science 75:342-—349. WRCC. 2005. Western Regional Climate Center. | Available at: http://www.wrcc.dri.edu/index.html; | accessed 15 September 2005. in R. S. Reichard, P. | Dunwiddie, J. Gamon, A. Kruckeberg, and D. | Salstrom (eds.), Conservation of Washington’s | Native Plants and Ecosystems. Washington Native | Small | mammals in oak woodlands in the Puget Trough, | MADRONO, Vol. 55, No. 2, pp. 159-160, 2008 MUHLENBERGIA ALOPECUROIDES (POACEAE: MUHLENBERGIINAB), A NEW COMBINATION PAUL M. PETERSON' Department of Botany, MRC-166, National Museum of Natural History, Smithsonian Institution, Washington, DC 20013-7012 J. TRAVIS COLUMBUS Rancho Santa Ana Botanic Garden, 1500 North College Avenue, Claremont, California 91711-3157 ABSTRACT A new combination, Muhlenbergia alopecuroides (Griseb.) P. M. Peterson & Columbus, is made based on published and ongoing molecular phylogenetic studies. RESUMEN Se propone una nueva combinacion, Muhlenbergia alopecuroides (Griseb.) P. M. Peterson & Columbus, basada en estudios filogenéticos moleculares publicados y en curso. Key Words: Lycurus setosus, Muhlenbergia alopecuroides, Muhlenbergiinae, Poaceae, taxonomy. The grass subtribe Muhlenbergiinae (Chlori- doideae: Cynodonteae) recently has comprised 10 genera: Aegopogon Humb. & Bonpl. ex Willd., Bealia Scribn., Blepharoneuron Nash, Chabois- saea E. Fourn., Lycurus Kunth, Muhlenbergia Schreb., Pereilema J. Presl, Redfieldia Vasey, Schaffnerella Nash, and Schedonnardus Steud. (Duvall et al. 1994; Peterson 2000; Peterson et al. 1995, 1997, 200la, 2007). Members of Muhlen- bergiinae are characterized by ligules usually membranous and eciliate; the inflorescence re- branched or of spicate primary branches, branch axes persistent or falling entire; spikelets bisexual, staminate, or sterile, solitary, rarely paired or in triplets, occasionally secund, glumes awned or unawned, floret usually 1; lemmas 3-nerved, awned or unawned; cleistogenes occasionally present in the leaf sheaths; and a base chromo- some number of x = 8-10. A phylogenetic study of Chloridoideae based on trnL—F (cpDNA) and ITS (nrDNA) sequences (Columbus et al. 2007) has revealed that Muh- lenbergiinae are monophyletic but Muhlenbergia is paraphyletic with respect to the other genera in the subtribe. We and colleagues are conducting a phylogenetic study focused on Muhlenbergiinae in which all species are targeted for sequencing. TrnL—F and ITS sequences from most species have been obtained and analyzed (J. T. Colum- bus et al. unpubl. data). The monophyly of Muhlenbergiinae and paraphyly of Muhlenbergia are consistent outcomes in all analyses of this large and expanding data set (Peterson et al. 2001b, 2004). We are proposing taxonomic ' Author for correspondence, email: peterson@si.edu changes based on these results. Herein, in preparation for the second edition of The Jepson Manual (Baldwin et al. in prep.), we transfer Lycurus setosus (Nutt.) C. Reeder, which was sampled in the Columbus et al. (2007) study, to Muhlenbergia. Muhlenbergia alopecuroides (Griseb.) P.M. Peter- son & Columbus, comb. nov. Lycurus alope- curoides Griseb., Abh. Konigl. Ges. Wiss. Gottingen 19:255—256. 1874. Type: Argentina, Prov. Catamarca, ca. Belén, en el altivalle de Las Granadillas, Feb 1872, P. G. Lorentz s.n. (Holotype: GOET; Isotypes: BA!, BAA, CORD!, SI!, US-996080 fragm. ex GOET!). Pleopogon setosum Nutt., Proc. Acad. Nat. Sci. Philadelphia 4:25. 1848. Lycurus setosus (Nutt.) C. Reeder, Phytologia 57:287. 1985. Type: U.S.A., New Mexico, Santa Fe Co., mountains near Santa Fe, 1841 or 1842, W. Gambel s.n. (Holotype: K; Isotypes: PH, US- 610839 fragm. ex K!). Comments. The new combination cannot be based on the basionym Pleopogon setosum because the epithet in Muhlenbergia [as M. setosa (Kunth) Trin.] was previously used. Therefore, we are using the next available name, Lycurus alopecuroides, as the basis for the new combina- tion. In California, there are 19 species of Muhlen- bergia including M. alopecuroides (Peterson 1993, 2002, in press). In the first edition of The Jepson Manual (Hickman 1993), M. alopecuroides was mistakenly treated as Lycurus phleoides Kunth instead of L. setosus (Smith 1993, 2002). Muh- 160 lenbergia alopecuroides can be distinguished from L. phleoides by having leaf blades with fragile, awnlike tips (3—)4-7(—12) mm long (blades acute or with bristles 1-3 mm long in L. phleoides) and longer ligules that are (2—)3—10(-12) mm long (igules 1.5-3 mm long in L. phleoides) [Reeder 2003]. Muhlenbergia alopecuroides has an amphi- tropical distribution, occurring in the southwest- ern United States, northern Mexico, southern Bolivia, and northern Argentina. Based on allozyme variation, it seems likely that the species recently dispersed to South America because populations there contain less genetic variation (Peterson and Morrone 1998; Peterson 2000). ACKNOWLEDGMENTS We would like to thank the Smithsonian Institution and the National Geographic Society for financial support, and Patricia Gomez Bustamonte for correcting the Spanish resumen. LITERATURE CITED BALDWIN, B. G., S. BOYD, B. J. ERTTER, D. J. KEIL, R. W. PATTERSON, T. J. ROSATTI, AND D. H. WILKEN (eds.) In prep. The Jepson manual of higher plants of California, 2nd edition. University of California Press, Berkeley, CA. COLUMBUS, J. T., R. CERROS-TLATILPA, M. S. KIN- NEY, M. E. SIQUEIROS-DELGADO, H. L. BELL, M. P. GRIFFITH, AND N. F. REFULIO-RODRIGUEZ. 2007. Phylogenetics of Chloridoideae (Gramineae): a preliminary study based on nuclear ribosomal internal transcribed spacer and chloroplast trnL—F sequences. Aliso 23:565—579. DUVALL, M. R., P. M. PETERSON, AND A. H. CHRISTENSEN. 1994. Alliances of Muhlenbergia (Poaceae) within New World Eragrostideae are identified by phylogenetic analysis of mapped restriction sites from plastid DNAs. American Journal of Botany 81:622—629. HICKMAN, J. C. (ed.). 1993. The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. PETERSON, P. M. 1993. Muhlenbergia. Pp. 1272-1274 in J. C. Hickman (ed.). The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. 2000. Systematics of the Muhlenbergiinae (Chloridoideae: Eragrostideae). Pp. 195-212 in S. W. L. Jacobs and J. Everett (eds.). Grasses: systematics and evolution. CSIRO, Melbourne. . 2002. Muhlenbergia. Pp. 586-588 in B. G. Baldwin, S. Boyd, B. J. Ertter, R. W. Patterson, T. J. Rosatti, and D. H. Wilken (eds.). The Jepson desert manual: vascular plants of Southeastern MADRONO [Vol. 55 California. University of California Press, Berke- ley, CA. . In press. Muhlenbergia. in B. G. Baldwin, S. Boyd, B. J. Ertter, D. J. Keil, R. W. Patterson, T. J. Rosatti, and D. H. Wilken (eds.). The Jepson manual of higher plants of California, 2nd edition. University of California Press, Berkeley, CA. , J. T. COLUMBUS, R. CERROS T., AND M. S. KINNEY. 2001b. Phylogenetics of Muhlenbergia and relatives (Poaceae: Chloridoideae) based on internal transcribed spacer region sequences (nrDNA). Botany 2001 Abstract available at: www.botany2001.org/section12/abstracts/33.shtml , N. F. REFULIO-RODRIGUEZ, R. CERROS- TLATILPA, AND M. S. KINNEY. 2004. A phylogeny of the Muhlenbergiinae (Poaceae: Chloridoideae: Cynodonteae) based on ITS and trnL—F sequences. Botany 2004 Abstract available at: www.2004.botanyconference.org/engine/search/ index.php?func= detail&aid=38 , AND S. J. PENNINGTON. 2007. Classi- fication and biogeography of New World grasses: Chloridoideae. Aliso 23:580—594. AND O. MORRONE. 1998. Allelic variation in the amphitropical disjunct Lycurus setosus (Poaceae: Muhlenbergiinae). Madronio 44:334—-346. AND R. J. SORENG. 2007. Systematics of California grasses (Poaceae). Pp. 7-20 in M. R. Stromberg, J. D. Corbin, and C. M. D’Antonio (eds.). Ecology and management of California grasslands. University of California Press, Berke- ley, CA. , G. DAVIDSE, T. S. FILGUEIRAS, F. O. ZULOAGA, AND E. J. JUDZIEWICZ. 2001la. Cata- logue of New World grasses (Poaceae): II. Sub- family Chloridoideae. Contributions from the United States National Herbarium 41:1—255. , R. D. WEBSTER, AND J. VALDES-REYNA. 1995. Subtribal classification of the New World Eragros- tideae (Poaceae: Chloridoideae). Sida 16:529—544. , AND . 1997. Genera of New World Eragrostideae (Poaceae: Chloridoideae). Smithsonian Contributions to Botany 87:1—50. REEDER, C. G. 2003. Lycurus. Pp. 200-203 in M. E. Barkworth, K. M. Capels, S. Long, and M. B. Piep (eds.). Magnoliophyta: Commelinidae (in part): Poaceae, part 2. Flora of North America north of Mexico, volume 25. Oxford University Press, New York, NY. SMITH, J. P. JR. 1993. Lycurus. Pp. 1270 in J. C. Hickman (ed.). The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. . 2002. Lycurus. Pp. 584 in B. G. Baldwin, S. Boyd, B. J. Ertter, R. W. Patterson, T. J. Rosatti, and D. H. Wilken (eds.). The Jepson desert manual: vascular plants of Southeastern California. University of California Press, Berkeley, CA. | MADRONO, Vol. 55, No. 2, pp. 161-169, 2008 DISTRIBUTION OF DWARF MISTLETOES (ARCEUTHOBIUM SPP., VISCACEAE) IN DURANGO, MEXICO ROBERT L. MATHIASEN School of Forestry, Northern Arizona University, Flagstaff, AZ 86011 Robert.Mathiasen@nau.edu M. SOCORRO GONZALEZ ELIZONDO Herbario CIIDIR, Instituto Politécnico Nacional, Durango, Durango, México MARTHA GONZALEZ ELIZONDO Herbario CIIDIR, Instituto Politécnico Nacional, Durango, Durango, México BRIAN E. HOWELL Forest Health Protection, USDA Forest Service, Lakewood, CO 80522 I. LORENA LOPEZ ENRIQUEZ Herbario CIIDIR, Instituto Politécnico Nacional, Durango, Durango, México JARED SCOTT School of Forestry, Northern Arizona University, Flagstaff, AZ 86011 JORGE A. TENA FLORES Herbario CIIDIR, Instituto Politécnico Nacional, Durango, Durango, México ABSTRACT The dwarf mistletoes (Arceuthobium spp., Viscaceae) are an ecologically and economically important group of hemi-parasitic flowering plants that parasitize members of the Pinaceae in both the Old and New Worlds. There are two areas of maximum diversity for the dwarf mistletoes: northern California and Durango, Mexico. Here we provide additional information on the distribution of dwarf mistletoes in Durango based on field observations and herbarium collections. This information supplements data provided in a monograph on Arceuthobium published in 1996. Needed research on the dwarf mistletoes in Durango is also discussed. Key words: Arceuthobium, distribution, dwarf mistletoes, Durango, Mexico. RESUMEN Los muérdagos enanos (Arceuthobium spp., Viscaceae) son un grupo de plantas hemiparasitas ecologica y econoOmicamente importantes que parasitan a miembros de la familia Pinaceae, tanto en el Viejo como en el Nuevo Mundo. Existen dos areas de maxima diversidad para los muérdagos enanos: el norte de California y Durango, México. En este trabajo se provee de informacion adicional sobre los muérdagos enanos en Durango con base en observaciones de campo y datos de material de herbario. Esta informaciédn complementa los datos presentados en una monografia sobre Arceuthobium publicada en 1996. Se discute también la necesidad de mas investigaciOn sobre los mueérdagos enanos en Durango. Palabras clave: Arceuthobium, distribucion, muérdagos enanos, Durango, México. DISTRIBUTION OF DWARF MISTLETOES (ARCEUTHOBIUM SPP., VISCACEAE) IN DURANGO, MEXICO Dwarf mistletoes (Arceuthobium spp., Visca- ceae) are among the most economically and ecologically important parasitic flowering plants in North America (Hawksworth and Wiens 1996; Mathiasen 1996; Hawksworth et al. 2002; Shaw et al. 2004). They are common and abundant parasites of the Pinaceae and affect nearly all of the commercially important coniferous species in the western United States, Canada, and Mexico (Hawksworth and Shaw 1984; Hawksworth and Wiens 1996; Geils and Hawksworth 2002). The geographic regions with the greatest diversity and concentration of dwarf mistletoes are northern California and southern Oregon, United States with 12 taxa and Chihuahua and Durango, Mexico with 13 taxa (Hawksworth and Wiens 1996; Mathiasen and Marshall 1999: Hawks- worth et al. 2002). Eleven of the taxa found in the Sierra Madre Occidental of Chihuahua and Durango occur in Durango. Two additional taxa, A. abietinum Engelm. ex Munz and A. vaginatum (Willd.) Presl subsp. cryptopodum (Engelm.) 162 Hawksw. & Wiens, occur north of Durango in Chihuahua, giving this area of the Sierra Madre Occidental the most diverse assemblage of dwarf mistletoes in North America. Information on the distribution and hosts of dwarf mistletoes in Durango, Mexico has been primarily based on the work of Hawksworth and Wiens (1965, 1970, 1972, 1977, 1989, 1996). Hawksworth and Wiens (1996) provided distri- bution maps for each dwarf mistletoe found in Durango, primarily based on their collections obtained from 1961—1987. More recent collections, however, provide additional information on dwarf mistletoe distri- butions and hosts in the region. The senior author has been collecting additional data on the distribution of dwarf mistletoes in Durango since 1997. In addition, other investigators have deposited many specimens of Arceuthobium collected throughout Durango with the Herbario CIIDIR, Instituto Politecnico Nacional located in Durango City, Mexico (CIIDIR). Based on our extensive field observations and these her- barium records we now have a great deal of additional information on the distribution of these economically and ecologically key parasitic plants in Durango. These data illustrate that there are three areas of the state that have been intensively sampled: west-central Durango near Altares, east and west of El Salto along Mexico Route 40, and southeast of Durango City. These areas have been intensively sampled because they are transected by major roads, making access to them much easier than for the more remote regions of the state. Areas in Durango that have not been adequately sampled for dwarf mistletoes include the regions south of La Flor and south and west of Topia. These areas are dominated by mountainous terrain high enough in elevation to support extensive pine forests, and hence dwarf mistletoes, and should be surveyed. Eastern Durango is dominated by grasslands and xero- phytic scrub with few pines, and thus far, no dwarf mistletoes have been reported from this part of the state. In addition, although the state has extensive pinOn woodlands (Garcia and Gonzalez 2003), no dwarf mistletoes have been reported on pinones in the state (Hawksworth and Wiens 1996). Because the pine forests of the Sierra Madre Occidental are among the most productive commercial forests in Mexico, studies of the distribution and host range of dwarf mistletoes and the damage associated with severe infection by these parasites should continue there. The taxa of Arceuthobium occurring in Dur- ango are listed below and their distributions based on a compilation of published literature, herbarium specimens, and the author’s field observations discussed under each taxon. Dwarf mistletoe distributions are illustrated in Figs. 1 and 2 and include information from Hawksworth MADRONO [Vol. 55 and Wiens (1996), data from CIIDIR, and our field observations and collections. Most of the senior author’s collections of Arceuthobium from Durango are deposited at the Deaver Herbarium, Northern Arizona University, Flagstaff (ASC). Principal hosts of each species, based on data in Hawksworth and Wiens (1996) and our field observations, are listed in Table 1. Pine taxono- my follows Farjon and Styles (1997), except for Pinus chihuahuana Engelm. and P. cooperi C.E. Blanco being recognized at the species level following Garcia and Gonzalez (2003) and Almaraz et al. (2006). The taxonomic status of several pines occurring in the Sierra Madre Occidental is still in debate, particularly the classification of pines in subgenus Strobus; we have followed the classification used by Garcia and Gonzalez (2003) for this subgenus. In addition, we discuss needed research on the taxonomy, distribution, and pathology for some of the common dwarf mistletoes in Durango. ARCEUTHOBIUM BLUMERI HAWKSwW. & WIENS This parasite of white pines occurs more commonly in Durango than originally reported by Hawksworth and Wiens (1996). It is distrib- uted in the Sierra Madre Occidental from northwestern to southern Durango (Fig. 1). Because it occurs in far southern Durango and its hosts are distributed in adjacent states, it probably occurs in northern Nayarit, northern Jalisco, and western Zacatecas, but it has not been reported in these Mexican states thus far. Although this is the only dwarf mistletoe that parasitizes white pines in Durango, the plants of | A. blumeri are small and gray, so their presence on infected trees may be overlooked. Hawks- worth and Wiens (1972, 1996) reported that this dwarf mistletoe seldom induces the formation of | witches’ broom on infected trees. However, our field observations from throughout its range in | Mexico and in southern Arizona (Huachuca Mountains) indicate it often induces witches’ broom formation, even on lightly infected trees. Therefore, infected trees can usually be identified using the presence of witches’ brooms (Mathiasen 1979). Hawksworth (1991) suggested that the pattern of parasitism of the Mexican white pines by Arceuthobium may be useful when interpreting | the taxonomy and distribution of these pines in | Mexico. The classification of the white pine populations in northern Mexico has been debated for many years (Shaw 1909; Martinez 1948; — Critchfield and Little 1966; Perry 1991; Farjon | and Styles 1997; Garcia and Gonzalez 2003). Hawksworth (1991) interpreted the selective parasitism of white pines in the Sierra Madre | Occidental by A. blumeri as evidence that these — populations should be classified as Pinus ayaca- | 2008] MATHIASEN ET AL.: ARCEUTHOBIUM DISTRIBUTION IN DURANGO, MEXICO \ Goméz \ Palatio 1 \ Tepehuanes va pi __ Santiago ~~" Papasquiaro : a eae : ‘\ \ Durango “ft “~Du NG b V Se EI Salto) yy—“s'v ° BB Va t/ya V c~., o, Mar BB} pusDuDup 6 Vaya Va eal “B Va C Va \B Du Va 100 Kilometers Fic. 1. Distribution of Arceuthobium blumeri (B), A. strictum (S), A. vaginatum subsp. vaginatum (Va), A. vaginatum subsp. durangense (Du), A. verticilliflorum (V), and A. yecorense (Y) in Durango, Mexico. Distributions are based on data in Hawksworth and Wiens (1996), specimens deposited at the Herbario CIIDIR, Durango, Mexico, the Deaver Herbarium, Flagstaff, AZ, and field observations. Populations designated with large, bolded letters are specifically discussed in the text. 164 MADRONO [Vol. 55 a ee a RR £ —— $ g 3 ? % “Saag, go, ss a 3 e PRE a 5 td Ane RR # ® £ CCBA _ eet OCA @ = Tepehuanes | Gi N D : a Santiago 4 Papasquiaro f G Altares 0 n ve ri6i G Durango Pept OR NUDES ESTELLE, i 100 Kilometers Fic. 2. Distribution of Arceuthobium douglasii (D), A. gillii subsp. gillii (Gi), A. gillii subsp. nigrum (N), A. globosum (G), and A. rubrum (R) in Durango, Mexico. Distributions are based on data in Hawksworth and Wiens (1996), specimens deposited at the Herbario CIIDIR, Durango, Mexico, the Deaver Herbarium, Flagstaff, AZ, and field observations. Populations designated with large, bolded letters are specifically discussed in the text. 2008] TABLE 1. MATHIASEN ET AL.: ARCEUTHOBIUM DISTRIBUTION IN DURANGO, MEXICO 165 PRINCIPAL HOSTS OF DWARF MISTLETOES THAT OCCUR IN DURANGO, MEXIco. Host classification of the pines follows Farjon and Styles, 1997 (in part) and Garcia and Gonzalez 2003. Dwarf mistletoe . blumeri . douglasii . gillii subsp. gillii . gillii subsp. nigrum globosum rubrum . strictum vaginatum subsp. vaginatum . vaginatum subsp. durangense . verticilliflorum . pecorense Se oe Oe huite Ehrenb. ex Schltdl. var. brachyptera Shaw, a view supported by Eguiluz Piedra (1991). Hawks- worth’s conclusion was influenced by the selective parasitism of Arceuthobium apachecum Hawksw. & Wiens of the white pine populations in Arizona, New Mexico, and northeastern Mexico that he classified as Pinus strobiformis Engelm. Hawksworth (1991) argued that parasitism by A. blumeri was confined to P. ayacahuite var. brachyptera 1n Mexico and southern Arizona and that A. apachecum was the only parasite of P. strobiformis in the southwestern United States and northeastern Mexico. Pinus ayacahuite is not parasitized by dwarf mistletoes in central Mexico, but Arceuthobium guatemalense Hawksw. & Wiens parasitizes P. ayacahuite starting in central Oaxaca extending through Chiapas and into western Guatemala (Hawksworth and Wiens 1996; Mathiasen et al. 2003). The absence of A. blumeri and A. guatemalense from central Mexico is noteworthy. Perry (1991) considered that both P. strobiformis and P. ayacahuite var. brachyptera co-exist in the Sierra Madre Occidental of northern Mexico, a view shared by Garcia and Gonzalez (2003). Based on our field observations of the white pines in Durango, we also consider both P. strobiformis and P. ayacahuite to be present in the state. These white pines are both parasitized by A. blumeri there (Mathiasen 1979). Therefore, we don’t consider the parasitism by A. blumeri as a consistent character that can be used to separate these white pines from each other in northern Mexico. Additional taxonomic studies are needed for the white pine populations throughout northern Mexico. ARCEUTHOBIUM DOUGLASIT ENGELM This important parasite of Pseudotsuga men- ziesii (Mirb.) Franco has now been collected from seven locations in northwestern Durango and from one location in southern Durango (Fig. 2). Therefore, it is probable that it will be discovered from additional locations in west-central Du- Principal hosts Pinus strobiformis; P. ayacahuite var. brachyptera Pseudotsuga menziesii Pinus chihuahuana; P. herrerai; P. leiophylla; P. lumholtzii Pinus chihuahuana; P. leiophylla; P. lumholtzii; P. teocote Pinus cooperi, P. durangensis; P. engelmannii Pinus cooperi, P. durangensis; P. engelmannii; P. herrerai; P. teocote Pinus chihuahuana Pinus arizonica; P. cooperi; P. durangensis; P. engelmannii Pinus coopert; P. devoniana; P. douglasiana; P. durangensis; P. engelmannii Pinus arizonica; P. cooperi; P. durangensis; P. engelmannii Pinus chihuahuana; P. durangensis; P. herrerai; P. leiophylla; P. lumholtzii rango. Because stands of Pseudotsuga have only persisted in the Sierra Madre Occidental where site conditions favor their regeneration and growth, this tree species occurs in scattered stands (Guerra-De la Cruz 2001), and 4. douglasii has evidently persisted in some of these areas and not others. It is probable that as more stands of Pseudotsuga are examined in Durango, this dwarf mistletoe will be discovered in many new areas. Although A. douglasii forms very small, olive-green plants that are not easily observed, it induces the formation of large, distinctive witches’ brooms on infected trees, which are easy to observe in stands of Pseudot- suga (Hawksworth and Wiens 1996). ARCEUTHOBIUM GILLIT HAWKSW. & WIENS SUBSP. G/JLLIT AND SUBSP. NIGRUM HAWKSW. & WIENS These parasites of Pinus leiophylla Schiede ex Schltdl., P. /umholtzii B. L. Rob. & Fernald, and P. chihuahuana are common in the Sierra Madre of Durango (Fig. 2). Although Hawksworth and Wiens (1989, 1996) recognized A. gillii subsp. nigrum at the species level (A. nigrum (Hawksw. & Wiens) Hawksw. & Wiens), we prefer to apply the earlier classification of this mistletoe by the same authors (1965, 1972) as a subspecies of A. gillii because recent molecular data indicate these taxa are very closely related (Nickrent et al. 2004). The molecular data and the morphological similarities between A. gillii and A. nigrum strongly support the earlier classification of A. nigrum as a subspecies of A. gillii. Although Hawksworth and Wiens (1996) only reported two locations for A. gi/lii subsp. gi//ii in Durango, collections at CIIDIR and our field observations indicate it is widely distributed from northern to southeastern Durango. Arceuthobium gillii subsp. nigrum is also distributed from northern to southern Durango and is one of the most common dwarf mistletoes in the state. Although it also parasitizes P. teocote Schltdl. and Cham., P. lumholtzii, and P. chihuahuana in 166 Durango, it 1s most common on P. /eiophylla. Hawksworth and Wiens (1996) reported that these two dwarf mistletoes occur in the same mountain range near Tepehuanes, Durango, but they indicated these taxa were separated by elevation differences; A. gillii occurring below 2200 m and A. gillii subsp. nigrum occurring above 2600 m. Plant color was considered by Hawksworth and Wiens (1996) as one of the distinguishing charac- ters separating A. gillii from A. nigrum: plants of A. gillii are greenish-brown and those of A. nigrum are dark brown to black. However, Hawksworth and Wiens (1989) mentioned in their discussion of A. nigrum that shoots of this species could also be dark green. Our field observations in Durango indicate that plants of A. gillii subsp. nigrum are usually greenish-brown, dark green, or dark brown, and only occasionally black. Plants of A. gilli subsp. nigrum in central Mexico are dark green to black. However, these plant characteris- tics are also representative of a morphologically similar, but phenologically and genetically distinct dwarf mistletoe that has now been reported from southern Mexico, A. hondurense Hawksw. & Wiens (Mathiasen et al. 2001, 2003). The mor- phological similarities between these dwarf mis- tletoes are demonstrated by the earlier identifica- tions of A. hondurense in Chiapas and Oaxaca as A. gillii subsp. nigrum (Hawksworth and Wiens 1972) and later as A nigrum by Hawksworth and Wiens (1989, 1996). Further studies are warranted to determine the taxonomic status of the popula- tions of A. gillii subsp. nigrum in central Mexico. Furthermore, because of the similarity in plant color of subsp. gi//ii and subsp. nigrum in Durango, the identification of these taxa is often difficult. We have primarily relied on the size of plants as the key characteristic to distinguish these taxa; plants of A. gillii subsp. nigrum being larger than those of A. gillii subsp. gillii (Hawksworth and Wiens 1989, 1996). Further taxonomic studies of these taxa are also needed in Durango. ARCEUTHOBIUM GLOBOSUM HAWKSwW. & WIENS SUBSP. GLOBOSUM This dwarf mistletoe is one of the most widespread in Durango occurring through the Sierra Madre Occidental from northern to southern Durango (Fig. 2). It has been reported to parasitize three common and economically important pines in Durango (Table 1). It is most common on Pinus cooperi and P. durangensis Martinez in this region. Although this dwarf mistletoe does not usually induce the formation of witches’ brooms on infected pines, it is easily identified by its bright yellow plants that form large, round masses of shoots in infected trees. Its extensive distribution, its large plants, and its distinctive yellow color have all contributed to its MADRONO [Vol. 55 being extensively collected in Durango. Because it is so widespread and locally abundant its effects on the growth of its hosts should be investigated. Populations of A. globosum approximately 50 km west of El Salto near Buenes Aires (Fig. 2) have some very large (>30 cm), green plants which are morphologically similar to A. globosum Hawksw. & Wiens subsp. grandicaule Hawksw. & Wiens, a common dwarf mistletoe on pines in central Mexico (Hawksworth and Wiens 1996). However, these populations also have plants that are characteristic of typical A. globosum subsp. globosum. Therefore, we have classified these populations as A. globosum subsp. globosum, but further taxonomic studies of these populations are warranted. ARCEUTHOBIUM RUBRUM HAWKSw. & WIENS This distinctively red dwarf mistletoe that has shiny fruits has only been reported in the Sierra Madre Occidental of Durango and extreme eastern Sinaloa (Hawksworth and Wiens 1996). It is widely distributed in Durango (Fig. 2) and parasitizes five pines as principal hosts (Table 1). It is probably that it will eventually be found in Chihuahua, Nayarit, northern Jalisco, and west- ern Zacatecas because its principal hosts are widely distributed in these states also. Popula- tions of A. rubrum near Altares in Durango have plants that are much larger than those in other areas of Durango, and thus further taxonomic study of these populations is warranted (Hawks- worth and Wiens 1996). Little is currently known about the effects A. rubrum has on its hosts, and because this mistletoe 1s so widely distributed in Durango, and occurs on some of the most economically important pines in this region, studies of its effects on host mortality and growth should be initiated. ARCEUTHOBIUM STRICTUM HAWKSW. & WIENS This dwarf mistletoe is endemic to Durango, but because it occurs very close to the borders with Nayarit, Jalisco, and Zacatecas it will undoubtedly be discovered in these states (Fig. 1). This is particularly true for western Zacatecas because several populations of A. strictum have been collected near its border with Durango. Although A. strictum principally par- asitizes Pinus chihuahuana which is distributed throughout the Sierra Madre Occidental, this dwarf mistletoe has only been found on the far eastern side of these mountains in central and southern Durango. It is noteworthy that the other dwarf mistletoes parasitizing P. chihua- huana, A. gillii subsp. gillii and A. gillii subsp. nigrum, occur more extensively through the Sierra Madre Occidental. Why A. strictum appears to be restricted to a small part of the geographic range 2008] of its principal host is an interesting question. It appears to be limited by elevation, occurring below 2500 m, but its hosts commonly occur below 2500 m throughout the Sierra Madre Occidental. This dwarf mistletoe is a distinctive species that can easily be identified by its flowers, which have up to 7 perianth lobes, and by its staminate plants, which are typically a single spike when it flowers in the late summer and early fall (Hawksworth and Wiens 1965, 1996). Our field observations support those of Hawksworth and Wiens (1996) which indicate that A. strictum is often associated with mortality of severely infected trees in southern Durango. ARCEUTHOBIUM VAGINATUM (WILLD.) PRESL AND A. VAGINATUM (WILLD.) PRESL SUBSP. DURANGENSE HAWKSwW. & WIENS Arceuthobium vaginatum subsp. vaginatum 1s one of the most widespread and locally abundant dwarf mistletoes in Durango (Fig. 1) where it parasitizes several species of pines (Table 1). It is probably the most economically damaging dwarf mistletoe in the state (Hawksworth and Wiens 1996). Arceuthobium vaginatum subsp. vaginatum is not only the most widely distributed dwarf mistletoe in Durango, but is common throughout northern and central Mexico. Because of its local abundance, widespread distribution, and parasit- ism of more than 10 species of pines, it is considered to be the most economically damaging dwarf mistletoe in Mexico (Hawksworth and Wiens 1996). Additional research is needed on its economic impact throughout the country and particularly in Durango. Hawksworth and Wiens (1972, 1996) reported that Arceuthobium vaginatum subsp. durangense only occurs in the extreme western regions of Durango on the west side of the Sierra Madre Occidental, but we have found it north of El Salto parasitizing both P. engelmannii and P. cooperi (Fig. 1) (Mathiasen 2007). Both of these hosts are severely parasitized and we have classified them as principal hosts (Table 1). The distributions of the two subspecies of A. vaginatum overlap in Durango northeast of El Salto where they occur within 6 km of each other southeast of San Miguel de las Cruces (Fig. 1). The distribution of A. vaginatum subsp. durangense is still poorly known, but it appears to be more common in Sinaloa along the western slopes of the Sierra Madre at elevations <2000 m (Hawksworth and Wiens 1996). It is likely that it will be found in Nayarit because it has been collected in western Jalisco (Hawksworth and Wiens 1996). Although Hawksworth and Wiens (1989, 1996) treated subsp. durangense at the specific level (A. durangense (Hawksw. & Wiens) Hawksw. & Wiens), we prefer their earlier classification of this mistletoe as a subspecies of A. vaginatum MATHIASEN ET AL.: ARCEUTHOBIUM DISTRIBUTION IN DURANGO, MEXICO 167 (Hawksworth and Wiens 1965, 1972). Again, recent molecular data indicate that A. vaginatum and A. durangense are closely related; these subspecies can not be distinguished using the molecular markers examined thus far (Nickrent et al. 2004). These two subspecies are only distinguished by their plant color (bright orange versus dark brown to black) and perhaps their host range because Pinus teocote 1s reported to be immune to infection by A. vaginatum subsp. durangense, but P. teocote is a secondary host for A. vaginatum subsp. vaginatum (Hawksworth and Wiens 1996). Because of the molecular data currently available and the morphological simi- larity of these taxa, we believe the classification of the bright orange populations of A. vaginatum in Durango as subspecies durangense is the most appropriate treatment. ARCEUTHOBIUM VERTICILLIFLORUM ENGELM This distinctive dwarf mistletoe is more widely distributed in Durango than previously reported. At one time it was thought to be extremely rare in Durango (Hawksworth and Wiens 1972), but it has now been collected from several areas of the state (Fig. 1). We are certain it will continue to be collected throughout Durango and possibly in northern Nayarit and southern Chihuahua be- cause its principals hosts are widespread in these areas. This dwarf mistletoe is easily identified by its whorled arrangement of flowers (usually 6 flowers per whorl), its very large mature fruits (>1 cm in length), and its thick staminate spikes which are 4-6 mm in width (Hawksworth and Wiens 1965, 1972, 1996). It causes the formation of large witches’ brooms that often have main branches with much larger diameters than uninfected branches in the same whorl, particu- larly on P. engelmannii. The distribution of A. verticilliflorum within pine stands is interesting in that it is often only found on one or a few trees in small groups, and these will be scattered at distances supporting the suggestion that this dwarf mistletoe may only be bird disseminated (Hawksworth and Wiens 1996). We have also observed dense stands of small P. cooperi growing directly under infected, large trees with no infection whatsoever on the understory pines ca. 20 km northeast of El Salto along the road to San Miguel de las Cruces. This also suggests this dwarf mistletoe is completely, or at least primarily, bird disseminated. The biology of this dwarf mistletoe warrants further study because of its possible dependence on birds for seed dispersal. ARCEUTHOBIUM YECORENSE HAWKSW. & WIENS The distribution and abundance of this dwarf mistletoe is still poorly known. Currently it is 168 only known from one area of Durango, south- west of Altares (Fig. 1), and we have not observed it in any other locations in Durango thus far. It is also known from a small region on the Chihuahua/Sonora border near the town of Yecora, Sonora (Hawksworth and Wiens 1989, 1996). The Sonora and Durango populations are approximately 400 km apart and this species probably occurs in additional populations in northern Durango through western Chihuahua and eastern Sonora because it occurs on several pine species that occur in these states (Table 1). ADDITIONAL RESEARCH NEEDS Field observations indicate that several of the dwarf mistletoes in Durango are associated with increased mortality of severely infected trees, particularly A. blumeri, A. globosum, A. strictum, and A. vaginatum. Furthermore, these mistletoes may be associated with significant reductions in the growth of severely infected trees, but this relationship has not been studied in Mexico to the extent it has been in the United States and Canada (Geils and Hawksworth 2002). A coop- erative research project funded by the USDA Forest Service, the Comision Nacional Forestal, and CIIDIR has been initiated in Durango to examine the effects of A. vaginatum subsp. vaginatum on the growth of Pinus cooperi (which is treated as a variety of P. arizonica by Farjon and Styles 1997) using stem analysis techniques. This study should provide data that can be used to guide management decisions related to dwarf mistletoe mitigation efforts in Durango. Addi- tional research on the effects dwarf mistletoes have on their pine hosts in the Sierra Madre is needed as is research on the ecological relation- ships of the dwarf mistletoes with insects, birds, and other animals throughout Mexico. LITERATURE CITED ALMARAZ ABARCA, N., M. S. GONZALEZ ELIZONDO, J. A. TENA FLores, J. A. AviILA REYES, J. HERRERA CORRAL, AND N. NARANJO JIMENEZ. 2006. Foliar flavonoids distinguish Pinus leiophylla and Pinus chihuahuana (Coniferales, Pinaceae). Proceedings of the Biological Society of Washing- ton 119:426-436. CRITCHFIELD, W. B. AND E. L. LITTLE. 1966. Geographic distribution of the pines of the world. Miscellaneous Publication 991, USDA, Washing- ton, DC. EGUILUZ PIEDRA, T. 1991. Biosystematics of the Mexican and Central American white pines. Pp. 1-7 in P. W. Garrett (ed.), Proceedings of a Symposium on White Pine Provenances and Breeding. General Technical Report NE-155, USDA Forest Service, Radnor, PA. FARJON, A. AND B. STYLES. 1997. Pinus (Pinaceae). Flora Neotropica, Monograph 75, New York Botanical Gardens, New York, NY. MADRONO [Vol. 55 GARCIA AREVALO, A. AND M. S. GONZALEZ ELI- ZONDO. 2003. Pinaceas de Durango. Ed. IPN- Instituto de Ecologia-CONAFOR, México, D.F. GEILS, B. W. AND F. G. HAWKSWORTH. 2002. Damage, effects, and importance of dwarf mistletoes. Pp. 57— 65 in B. W. Geils, J. Cibrian Tovar, and B. Moody (tech. coords.), Mistletoes of North American conifers. General Technical Report RMRS-GTR-98, USDA Forest Service, Fort Collins, CO. GUERRA-DE LA CRUZ, V. 2001. Stand structure and dynamics of isolated Pseudotsuga forests in south- ern North America. Ph.D. dissertation, School of Forestry, Northern Arizona University, Flagstaff, AZ. HAWKSWORTH, F. G. 1991. Coevolution of Mexican white pines and their dwarf mistletoe parasites. Pp. 23-28 in P. W. Garrett (ed.), Proceedings of a Symposium on White Pine Provenances and Breeding. General Technical Report NE-155, USDA Forest Service, Radnor, PA. AND C. G. SHAW. 1984. Damage and loss caused by dwarf mistletoes in coniferous forests of western North America. Pp. 120-129 in R. K. S. Wood and G. J. Jellis (eds.), Plant Diseases: infection, damage, and loss. Blackwell Scientific Publications, Oxford, U.K. AND D. WIENS. 1965. Arceuthobium in Mexico. Brittonia 17:213—238. AND . 1970. New taxa and nomenclatural changes in Arceuthobium (Viscaceae). Brittonia 22:265—269. AND . 1972. Biology and taxonomy of dwarf mistletoes (Arceuthobium). Agricultural Handbook 401, USDA Forest Service, Washing- ton, D. C. AND . 1977. Arceuthobium in Mexico: additions and range extensions. Brittonia 29:41 1— 418. AND . 1989. Two new species, nomen- clatural changes, and range extensions in Mexican Arceuthobium (Viscaceae). Phytologia 66:3—11. AND 709, USDA Forest Service, Washington, D. C. HAWKSWORTH, F. G. D. WIENS, AND B. W. GEILS. | 2002. Arceuthobium in North America. Pp. 29-56 in B. W. Geils, J. Cibrian Tovar, and B. Moody (tech. coords.), Mistletoes of North American conifers. | General Technical Report RMRS-GTR-98, USDA | Forest Service, Fort Collins, CO. MARTINEZ, M. 1948. Los Pinos Mexicanos. Second | Edition. Universidad Autonoma de México, Mex- ico City. MATHIASEN, R. L. 1979. Distribution and effect of. dwarf mistletoes parasitizing Pinus strobiformis in | Arizona, New Mexico, and northern Mexico. | Southwestern Naturalist 23:455—461. Northwest Science 70:61—71. subsp. durangense on Pinus cooperi and Pinus | engelmannii in Mexico. Plant Disease 90:1201. . 1996. Dwarf Mistletoes: Biology, | pathology, and systematics. Agriculture Handbook 1996. Dwarf mistletoes in forest canopies. . 2007. First report of Arceuthobium vaginatum - \ AND K. MARSHALL. 1999. Dwarf mistletoes in| the Siskiyou-Klamath Mountain Region. Natural | Areas Journal 19:379—385. , J. MELGAR, J. BEATTY, C. PARKS, D. L.| NICKRENT, S. SESNIE, C. DAUGHERTY, B. HOWELL, | 2008] AND G. GARNETT. 2003. New distributions and hosts for mistletoes parasitizing pines in southern Mexico and Central America. Madrono 50:115—-121. MATHIASEN, R., D. NICKRENT, C. PARKS, J. BEATTY, AND S. SESNIE. 2001. First report of Arceuthobium hondurense in Mexico. Plant Disease 85:444. NICKRENT, D. L., M. A. GARCIA, M. P. MARTIN, AND R. L. MATHIASEN. 2004. A phylogeny of all species of Arceuthobium (Viscaceae) using nuclear and chloroplast DNA sequences. American Journal of Botany 91:125—138. MADRONO, Vol. 55, No. 2, p. 169, 2008 MATHIASEN ET AL.: ARCEUTHOBIUM DISTRIBUTION IN DURANGO, MEXICO 169 PERRY, J. P. 1991. The pines of Mexico and Central America. Timber Press, Portland, OR. SHAW, D. C., D. A. WATSON, AND R. L. MATHIASEN. 2004. Comparison of dwarf mistletoes (Arceutho- bium spp., Viscaceae) in the western United States with mistletoes (Amyema spp., Loranthaceae) in Australia - ecological analogs and _ reciprocal models for ecosystem management. Australian Journal of Botany 52:481—-498. SHAW, G. R. 1909. The pines of Mexico. Publications of the Arnold Arboretum 1. Harvard University, Cambridge, MA. REVIEW California Native Plants for the Garden. By CAROL BORNSTEIN, DAVID FROSS AND BART O’BRIEN. 2005. Cachuma Press, Los Olivos, CA. 2? Printing February 2006. 271 pp. ISBN 0-9628505-8-6, $27.95, paperback California Native Plants for the Garden is a truly valuable contribution to the somewhat limited, although increasingly available references on the topic of cultivating California native plants. The science of horticulture is critical to the success of landscape projects, including not only residential landscape design, but wildlands restoration and management. California Native Plants for the Garden capably and thoroughly addresses the many aspects of growing native plants well. This book was written by three renowned California horticulturists: Carol Bornstein (Hor- ticulture Director at Santa Barbara Botanic Garden) Dave Fross, (Landscape Architect, Cal Poly San Luis Obispo Instructor, and Native Sons Wholesale Nursery Owner) and Bart O’Brien (Senior Staff Research Associate at Rancho Santa Ana Botanic Garden, heralding previously from Yerba Buena Nursery in Palo Alto, CA). In combination, they offer consider- able expertise on the subject, although their background experience leads to a slight partiality towards coastal rather than interior regions of the State. Produced by California’s own, Cachuma Press, California Native Plants for the Garden 1s in a similar format to two of its other beautiful publications: Oaks of California and Conifers of California. Like its predecessors, the book is visually gripping with over 450 glossy color photographs, informative, thorough and well organized. While clearly its greatest appeal 1s for gardeners, landscape designers and native plant aficionados, it also benefits botanists and ecologists in their efforts to revegetate or restore land by providing horticultural essentials such as plant establishment, irrigation, care and mainte- nance. In the “Plant Profiles” section, it offers a depth of information demystifying the many genera as well as cultivated varieties available commercially today. Derivation, description, as well as cultural requirements are given for hundreds of plants. Each genus is illustrated with a colorful photograph—often depicted in an instructive landscape context. Additionally, concerns associated with wild- land-urban interfaces, such as exotic plant invasions, genetic contamination of endangered native plants, fire-safe landscaping, sustainability, drought, and erosion—are fundamentally, yet definitively addressed. —MELANIE BAER-KEELEY, Restoration Horticulturist, Sequoia & Kings Canyon National Parks, 47050 Generals Hwy, Three Rivers, CA 93271; Melanie_ Baer-Keeley@nps.gov. MADRONO, Vol. 55, No. 2, pp. 170-177, 2008 NEW ROSA (ROSACEAE) IN CALIFORNIA AND OREGON BARBARA ERTTER University and Jepson Herbaria, University of California, Berkeley, CA 94720-2465, and Missouri Botanical Garden, P.O. Box 299, St. Louis, MO 63166-0299, USA ertter@berkeley.edu WALTER H. LEWIS Washington University, Department of Biology, St. Louis, MO 63130-4899, and Missouri Botanical Garden, P.O. Box 299, St. Louis, MO 63166-0299, USA lewis@biology.wustl.edu ABSTRACT Two new roses are described from California and adjacent Oregon, with keys to distinguish each from the typical representative of their respective species. Rosa pisocarpa A. Gray subsp. ahartii Ertter & W. H. Lewis differs from typical R. pisocarpa by generally having fewer and larger leaflets, fewer flowers, and sepals that are most commonly eglandular, among other features. It is the characteristic member of the R. blanda Aiton complex in the northern Sierra Nevada and southern Cascade Range, where existing collections have been variously identified as R. pisocarpa, R. woodsii Lindl. subsp. ultramontana (S. Watson) Roy L. Taylor & MacBryde, R. californica Cham. & Schltdl., and R. bridgesii Crépin ex Rydb. The rank of subspecies is used to indicate a significant ecogeographic component of the species’ range. In contrast, varietal rank is used for a localized expression of R. gymnocarpa Nutt. occurring on open ultramafic substrates in northwestern California and adjacent Oregon, described here as Rosa gymnocarpa var. serpentina Ertter & W. H. Lewis. Diagnostic characters include relatively short stature, fewer and blunter leaflets, and pedicels that are frequently eglandular. Key Words: California, new taxa, Oregon, Rosa gymnocarpa, Rosa pisocarpa, serpentine. The taxonomy of the genus Rosa L. (Rosaceae) is notoriously difficult, complicated by extensive phenotypic plasticity, rampant hybridization, and polyploidy. These factors, when addressed by differing taxonomic philosophies, have resulted in radically different treatments of Rosa in North America. At one extreme, Rydberg (1918) recognized 129 native species in the most recent attempt to provide uniform monographic cover- age of the entire continent. In contrast, Erlanson (1932, 1934) accepted only 15 full species of native North American roses, plus seven additional ‘““ecotype species’’ and four putative hybrids. Her conclusions were based on cytogenetics, experimental hybridizations, common garden experiments, and extensive field observations. Erlanson did not summarize her conclusions in a monographic format, and no consensus treat- ment has been adopted by subsequent floras (e.g., Abrams 1944; Cronquist 1961; Cronquist and Holmgren 1997; Ertter 1993; Kearney and Peebles 1951; Munz 1959). One unifying feature has nevertheless been a conservative approach to Rosa taxonomy, with relatively few species recognized. There has also been a general reticence to describe new taxa and thereby risk following in the footsteps of Rydberg (1917, 1918) and his fellow “splitter”? Greene (1899, 1911, 1912), both of whom described a plethora of species that currently reside in synonymy. We agree that this caution has been well justified, especially given Erlanson’s (1934: 204) | observation that “‘It is not impossible to find on the same bush two of some of the species listed by | Rydberg.” Our respective studies (e.g., Ertter 1993; Lewis 1959a, 1959b, 1959c, 1962, 1965, | 2008; Lewis and Ertter 2007) have nevertheless led | us to conclude that there is taxonomic structure within North American Rosa beyond what has | recently been recognized. Our approach has been to focus on the general occurrence of unique suites | of morphological and ecological features, or at | least notably differing probabilities and/or ranges of variation in these features, that are associated | with definable ecogeographic settings. Taxa cir- | cumscribed by this approach are not always well differentiated from one another, due to the inherent complexity of the genus, but the alterna- tive is to gloss over ecogeographically meaningful components of biodiversity that have significance = = for conservation planning, restoration efforts, and _ horticultural purposes. To discern these ecogeographically based taxa, we have relied primarily on extensive field studies | and examination of herbarium specimens (espe- cially MO, UC/JEPS, RSA/POM, CAS/DS, and | CHSC), supplemented by common garden stud- ies. Many existing herbarium specimens are. inadequate for some of the most critical diagnos- — tic characters, such as the mid-stem prickles, so 2008] the accumulation of fully representative addi- tional collections (deposited in UC and MO) and detailed observations of field populations have been extremely important. Cytological studies are currently underway, but results are still prelimi- nary and ambiguous. Existing molecular analyses (e.g., Bruneau et al. 2007; Joly & Bruneau 2007) are likewise insufficient to address species-level questions in California, although intriguing questions are being raised. Our conclusions will be incorporated in pend- ing treatments for Flora of North America North of Mexico and The Jepson Manual. The present paper describes two new taxa occurring in California and adjacent Oregon needed for these publications. New combinations in Rosa nutkana C. Presl and R. woodsii Lindl. have recently been published elsewhere (Lewis & Ertter 2007); recent and future papers address variation within the Rosa carolina L. (Lewis 2008) and R. californica Cham. & Schltdl. complexes (Ertter in prep.). As will be seen, we have adopted a three-tiered approach to taxonomic ranks within North American Rosa. The species rank is used for the most consistently distinct entities, which are also those that have been most commonly recognized even in the more conservative treatments. Major ecogeographic components within species are treated as subspecies, while unique localized expressions are treated as varieties. A NEW SUBSPECIES OF ROSA PISOCARPA Rosa pisocarpa A. Gray is the westernmost member of the Rosa blanda Aiton complex, typically occurring west of the Cascade Range from southern British Columbia south to north- western California as far as Elk Mountain, Mendocino County. The typical expression is a moderately tall thicket-forming shrub with clus- ters of relatively small flowers that mature into round, pea-sized hips (hence the Latin epithet). Prickles are few, generally occurring as relatively stout, straight infrastipular pairs; leaflets tend to be ovate with flaring teeth; and the sepals generally have stipitate glands and elongate expanded tips. Although the species mostly occurs west of the Cascade Range, populations with pisocarpa characters can be found in interior regions along major rivers, such as the Columbia River Basin, where they evidently intergrade with another member of the R. blanda complex, R. woodsii subsp. ul/tramontana (S. Watson) Roy L. Taylor & MacBryde, which is the characteristic member of the complex in the lowlands and adjacent mountains between the Cascade Range and northern Rocky Mountains (Lewis and Ertter 2007). This latter taxon differs in generally having more numerous and more slender prick- les, more obovate leaflets with forward-oriented teeth, eglandular sepals with shorter tips, and ERTTER AND LEWIS: CALIFORNIA AND OREGON ROSES 17] more ovate hips. The vestiture of the rachis also tends to be more finely and evenly velutinous than in R. pisocarpa. Except for the aforementioned intermediate populations on the interior rivers, most members of the R. blanda complex in far western North America can be assigned to either R. pisocarpa or R. woodsii s.1. with some degree of confidence. The outstanding exception comprises populations from the northern Sierra Nevada and southern Cascade Range in northern California and adjacent Oregon, which tend to be relatively low open shrubs with few (or no) prickles, relatively large ovate leaflets with flaring teeth, few flowers, and ovate hips. Sepals are most commonly eglandular with relatively short tips, although populations with glandular sepals and/or expanded tips are interspersed. In addition to being variously identified as R. pisocarpa and Rosa woodsii var. ultramontana, collections have often been identi- fied as R. californica, since the prickles are sometimes curved, though typically more slender than is characteristic of R. californica. Identifica- tions as Rosa bridgesii Crépin ex Rydb. have also occurred. Duplicates of the same collection, or multiple collections from the same population, have occasionally been given different identifications. Given this level of problematic taxonomic assignment, and the relative (for Rosa) uniformi- ty within the geographic area of concern, we believe that the best solution is to recognize the populations in question as comprising a distinct taxon, described here as a new subspecies of R. pisocarpa. The closer relation to typical R. pisocarpa than to R. woodsii subsp. ultramontana or R. californica is deduced from leaflet shape, rachis vestiture, and the occasional presence of stalked glands on the sepals, as well as from the complex zone of intergradation in the Siskiyou Mountains of Siskiyou and Trinity counties, California. The rank of subspecies is used because the ecogeographic range is larger and more complex than that of the varieties we are recognizing in North American Rosa, although smaller than that of most other subspecies being recognized in the genus. Rosa pisocarpa A. Gray subsp. ahartii Ertter & W. H. Lewis, subsp. nov. (Fig. 1)—TYPE: USA, California: Butte Co., Black Bart Road 1.3 mi from Forbestown Road, ca. 10 air mi ESE of Oroville, roadside in Foothill Wood- land, 2150 ft elev., 39°29.060'N 121°21.634'W, 23 Jun 2007, B. Ertter & L. Ahart 19074 (Holotype: UC; Isotypes: K, LE, MO, MT, OSC, NY, RSA, US, + other duplicates to be distributed) Differt a subsp. pisocarpa floribus paucioribus sepalis plerumque eglandulatis foliolis paucior- ibus majoribus. 172 MADRONO [Vol. 55 FIG. 1. Rosa pisocarpa subsp. ahartii. A. Habit, showing frequent absence of prickles, leaves with 5 leaflets, and i few flowers, with enlargement of leaflet margin showing single eglandular teeth (Ahart 12144). B. Flower bud, © glabrous pedicel, subtending bract, and eglandular sepals with moderately prolonged tips (Ertter et al. 15678). C. Mature hip with sepals (Ertter & Ahart 18750). | 2008] Rhizomatous shrubs. Stems solitary to loosely clustered, sometimes forming thickets, erect, (0.2—) 0.4-1.4(-1.8) m tall, growing tips puberulent but soon glabrous, sometimes glaucous when young, with age dark reddish brown, outer layer sometimes exfoliating as thin ash-gray sheet; prickles usually sparse to absent at least distally, occurring primarily as 1—2 straight to somewhat curved infrastipular prickles 2-5 mm long X 1-— 2 mm wide at base, flaring to an attachment to 8 mm long, sometimes compressed, smaller internodal prickles usually absent. Leaves (5—) 6—-11(-13) cm long X (3—)4-7(-10) cm wide; stipules 10—20(—30) mm long (including 3—S5 mm long auricle), 3—5(—7) mm wide, the margins subentire to shallowly crenate to irregularly erose; petiole and rachis subglabrous to finely puberu- lent to pubescent with hairs to 1 mm, stipitate glands absent, pricklets sometimes present; leaf- lets 5—7, most commonly 5, sparsely pubescent; terminal leaflet with petiolule S—15(—20) mm long, the blade elliptic-ovate, usually widest at or below middle, basally obtuse (rounded), apically obtuse to acute, (1.5—)2—-4.5(—6) cm long x (1—)1.5—2(-3) cm wide, usually singly serrate with 10—15(—20) acute 1° teeth per side on distal 4/5 of blade. Inflorescences commonly 1-—3-flowered on side branches, rarely 10-flowered or more in terminal candelabras; bracts 6-20 mm long xX 2-5 mm wide, the margins irregularly gland-toothed; pedicels 1—2(—2.5) cm long, glabrous. Flowers 2.5—3.5 cm diam.; hypanthium glabrous, ovoid- urceolate, 4-5 mm long < 2.5-4 mm wide in bud, with a narrowed neck +2 mm wide; sepals subglabrous to puberulent, sometimes with but more often without stipitate glands, 10-17 mm long 50 cm), while shrubs and tree saplings (1— 3 m in height) rely on water from shallower layers (O—50 cm) at the beginning of the season, but as soils become drier the shrub’s and the sapling’s primary zone of uptake shifts downward in the soil profile. These findings imply that shrubs may inhibit the survival of establishing tree seedlings until the seedlings have a deep enough root system to extract soil moisture from soil below 50 cm. Our study suggests that tree seedling survival may depend on a seedling’s ability to compete with shrubs for scarce soil moisture in the near-surface soil layers. Key Words: Abies concolor, Arctostaphylos patula, mycorrhizae, Pinus lambertiana, root competition, stable isotopes, water uptake depth. Seedling establishment is a critical stage in the life history of most plant species, and in forests can have a significant influence on a stand’s species composition and development for centuries. In the forests of the Californian Sierra Nevada, managers often have problems regenerating trees on sites after they are treated for fuels reduction (McDo- nald 1976; Helms and Tappeiner 1996). Currently there is little information on the effects of important site factors on tree regeneration in mixed (shrub/tree) vegetation and under closed canopy conditions (Tappeiner and McDonald 1996). A century of fire suppression has significantly in- creased fuel loads, increasing stem densities and canopy cover in many western forests, reducing understory light and soil moisture conditions for establishing seedlings (Minnich et al. 1995; Gray et al. 2005). Mixed-conifer, the dominant forest type on the Sierran Nevada’s western slopes, has been characterized as having a mixture of closed canopy, ' Corresponding author, e-mail: tdawson@berkeley.edu shrub and gap patch conditions, each with significantly different microclimates (North et al. 2002; Gray et al. 2005). Sierran forests have a prolonged summer drought, typical of a Mediter- ranean climate (Parker 1994; Stephenson 1998), and shrub competition for scarce soil water 1s believed to influence seedling survival and growth rates. For example, a study of tree demography in the Sierran Nevada forest found that many Jeffrey pines (Pinus jeffreyi; nomenclature follows Hick- man 1993) and sugar pines (Pinus lambertiana), which often become established on drier micro- sites favored by shrubs, had become established in wet El Nino years (North et al. 2005). However, research suggests that shrubs may facilitate tree seedling establishment by ameliorating microcli- mate extremes (GOmez-Aparicio et al. 2004) or by providing an ectomycorrhizal network that estab- lishing tree seedlings can tap into (Horton et al. 1999). Thus, further research into the influence of shrubs on conifer regeneration and microclimate conditions is needed to improve the management of Sierran forests. 192 MADRONO One factor that strongly influences the estab- lishment of tree seedlings in Sierran forests 1s likely to be the availability of below-ground resources, which may make a significant contribution to the patch conditions and gaps that characterize mixed-conifer forests. Royce and Barbour (2001) found that shrubs generally depleted soil moisture more rapidly and ultimately extracted a greater proportion of the available soil water than conifers. Shrubs, however, also provide shade, thereby reducing surface temperatures and the evaporative demand for establishing seedlings. Recent research in drought-stressed chaparral also suggests that conifer seedlings may benefit from shrub mycorrhizae. Horton et al. (1999) studying Pseudotsuga seedling survival in sites dominated by Arctostaphylos spp. suggested that the ecto- mycorrhizal fungi associated with Arctostaphylos spp. contributed to the establishment of Pseudo- tsuga menziesii seedlings. It has been postulated that the sharing of ectomycorrhizal fungi between conifers and Arbutus or Arctostaphylos spp. may play a major role in plant community develop- ment (Molina and Trappe 1982; Perry et al. 1989). However, there have been no studies, to date, in which the possible transfer of water has been rigorously excluded from the mechanisms where- by shrubs facilitate seedling establishment in dry chaparral ecosystems, or in Sierran forests where prolonged drought occurs during most of the growing season. In this study, we examined the microclimate, water use, and conifer seedling survival rates in manzanita shrub (Arctostaphylos patula) and mature forest patches in Sierran mixed-conifer forests. The objective was to determine whether (and if so how) seedling dynamics vary between two dominant patch types; areas dominated by shrubs with shallow roots, and closed canopy forest with deep rooting systems. Our hypotheses were (1) survival rates of tree seedlings would be higher for seedlings where mycorrhizal connec- tions were not blocked; (11) shrubs and overstory trees extract soil water from different depths (as measured by '8O/'®O and ?H/'H ratios of xylem water) and (iil) seedling survival rates would be higher in forest than in manzanita patches. To elucidate mechanisms potentially influencing these responses we also examined microclimatic (air temperature and soil moisture) differences between patch types, and the effects of an irrigation treatment on seedling transpiration rates. Water availability was also manipulated in the irrigation study to examine tree seedlings survival. MATERIALS AND METHODS Field Site The study was conducted in Teakettle Exper- imental Forest, 80 km east of Fresno, California, [Vol. 55 in the north drainage catchment of the Kings River (36°44'N, 117°30'W). The experimental forest consists of 1300 ha of old-growth mixed- conifer. Elevation ranges from 1980 m along the eastern boundary to 2590 m at the top of Patterson Mountain on the western boundary. Annual precipitation averages 125 cm at 2100 m and falls mostly as snow between November and April. Mean, maximum, and minimum July temperatures are 17°C; 30°C, and 3° © respec- tively (North et al. 2002). In Teakettle Forest there is a gradation in the vegetation from a mixture of white fir (Abies concolor), sugar pine (Pinus lambertiana), 1ncense-cedar (Calocedrus decurrens), Jeffrey pine (Pinus jeffreyi), and red fir (Abies magnifica) at the lower elevations to red fir, lodgepole pine (Pinus contorta), and western white pine (Pinus monticola) at higher elevations. Soils are Dystric and Lithic Xeropsamments of loamy sand to sandy loam textures (Anonymous 1993), derived from granitic rock, with exposed weathered and un-weathered rock common throughout the study area. Treatments Three manzanita (Arctostaphylos patula) patches and three closed-canopy patches (Fig. 1) were selected subjectively in the autumn of 2000 for a study of the survival of sugar pine (Pinus lambertiana) and white fir (Abies concolor) seedlings. Patches were selected which were large enough (=70 m7’) for the field experiment, distinct (separated by gaps), and yet close enough to be on the same mapped soil type. Manzanita and closed-canopy patches were separated by at least 80 m to ensure that there were no below- ground connections between patch types. White fir seedlings were transplanted from the sur- rounding area in October 2000 into three different types of cylinders (Fig. 1), which were inserted into the ground in both the manzanita and closed-canopy patches. The above-ground height of the transplanted white fir seedlings averaged ~10cm and root length averaged ~15 cm, a size associated with 1-4 yr old seedlings. There were not enough sugar pine seedlings at Teakettle to transplant, so five seeds were sown within each cylinder at the end of October 2000 and covered with a steel mesh to exclude predation of the seeds. In the spring of 2001, 90 cylinders each contained three estab- lished white fir seedlings, while another 90 cylinders contained three to five sugar pine seedlings which had germinated from the sown seed. Each cylinder was 35 cm long, 15.5 cm in diameter, and 65% of its area was open to the surrounding soil (control treatment), covered with a 50 um mesh, or covered with a 5 um mesh (Fig. 1). The 50 um mesh was sized to prevent root penetration but provide openings 2008] PLAMBOECK ET AL.: CONIFER SEEDLING SURVIVAL IN THE SIERRA 193 Closed-canop 2 pis . aR ae Tae . Dae Sa Francisco Manz an tch - fe “t No mesh 50um mesh 5um mesh Fic. 1. Map of California showing the location of the Teakettle Experimental Forest, the manzanita and closed- canopy patches, and the design of the three different cylinder treatments; control, 50 um pore-size mesh, and 5 um pore-size mesh. 194 MADRONO large enough to allow mycorrhizae to establish across it. The 5 um mesh was designed to prevent penetration of both roots and mycorrhizae. In total, 180 cylinders (2 plant species 3 enclosure treatments X 5 replicates X 6 sites) were installed. The seedlings were allowed to establish until 2002 in order to allow adequate time for external hyphal colonization in the chambers open to the surrounding soil (control treatment) and the chambers covered with a 50 um mesh (root exclosure treatment). We stress that the cylinder treatments allow hyphal connections to occur, but we did not sample roots to determine whether actual connections were established. We applied an irrigation treatment to assess how seedlings in closed-canopy and shrub patches assimilate water. August was chosen because it is the driest, warmest month with dry upper soil layers that will affect the transpiration of the seedlings. During August 2002, nine cylinders from each treatment (36 cylinders in total) were selected at random and irrigated twice a week for three weeks. Eighteen cylinders received 250 ml dl of water at each irrigation, and 18 cylinders received 500 ml at each irriga- tion. Transpiration rates in seedlings were mea- sured to assess relative water-use efficiency in the different patch and root exclosure treatments. Field Sampling To infer the depth at which plants were taking up water, 6°H and 6'°O of water extracted from mature conifers, conifer saplings (1-3 m in height) and manzanitas was compared to 6°H and 6'SO of water extracted from soil taken at different depths. From May to September 2001 soil samples were collected from every 10 cm layer down to ~70cm depth and water was extracted from each sample and analyzed. One soil pit from each of the manzanita and the closed-canopy patches was sampled on May 19, June 26, August 9, September 5, and September 29. At the same time, wood cores were extracted from mature conifers, conifer saplings and manzanita shrubs for xylem water extractions. More than 10 plant samples were collected in the vicinity of each soil pit on every sampling occasion. The soil and wood samples were kept in airtight plastic bottles that were stored in plastic bags and refrigerated (~5°C) until cryo- genic distillation (Ehleringer et al. 2000). The temperature, relative humidity, and abso- lute humidity were recorded with eight HOBO data loggers (Onset Computer Corporation, MA), calculating averages over 30 min during 2001 and 2002 in the manzanita and closed- canopy patches at 20 cm and 150 cm above the ground. In 2002, the relative humidity and absolute humidity were used to calculate the vapor pressure deficit (VPD) using the formula: [Vol. 55 ab humidity VPD = | ——__.— (= humidity ) — ab.humidity. (1) Volumetric soil water content was estimated using time-domain reflectometers (TDR, model No. 1502C, Tektronix Inc., Beaverton, OR) in two soil pits; one in a manzanita shrub area and one in closed-canopy forest. Only two soil pits were excavated because while soil depth differs between the two patch types, within each patch the soils are fairly homogenous (Erickson et al. 2005). The computed volumetric water content was based on the “low-C” calibration described in Gray and Spies (1995). This method averages water content over the length of steel probes left in the soil. The stainless probes were 30 cm long and installed horizontally at depths of 5 cm, 10 cm, 20 cm, 30 cm, 40 cm, 50 cm and 60 cm, while at 70 cm the probes were installed vertically (giving an average for the 70-100 cm profile). Depth to bedrock for each study area was measured by Gasch & Associates (G&A, Sacra- mento, CA), using a seismic refraction (SR) method, which measures the velocity at which a seismic wave 1s propagated through a soil or rock medium. Higher seismic primary-wave velocities indicate material of higher density, typically indicating the strength or composition of the material, and wave velocity is used as an index of depth to bedrock. The seedling survival was checked during four occasions, starting mid-June, during 2002. The seedlings were considered dead if all needles were brown. To determine how irrigated seedlings in closed- canopy and shrub patches assimilated irrigation water, leaf conductance rates (mmol m ~~?’ s_') of 36 seedlings representing each of the irrigation treatments were measured at the beginning of September using a LI-1600 null-balance steady state porometer (LI-COR, Inc., Lincoln, NE). Eighteen non-irrigated seedlings were also mea- sured to allow for comparison between irrigated and non-irrigated seedlings. Porometer measure- ments were taken from 9:00 to 12:00 a.m. Preparation and Analysis Soil and wood cores were stored for one to four weeks before soil and xylem water was extracted from the soil and seedling samples, respectively, by cryogenic distillation. The ex- tracted water was used for isotope analyses. The °H/'H ratios of the water from the different soil depths were determined using an isotope mass spectrometer (IRMS; a Finnigan-MAT Delta Plus XL, coupled to a H/Device at the UC Berke- ley Center for Stable Isotope Biogeochemistry). Results are expressed in standard notation (6°H) in parts per thousand (%o) relative to V-SMOW, where 8°H = [(Rsampte/Rv-smow) — 1] X 1000%o 2008] and R is the molar ratio of heavy to light isotopes (D/H) with a sample precision of +1%bo. The '8O/'°O ratios of xylem sap and soil water were determined using the same IRMS, but coupled to a Gas Bench II; a universal on-line interface which allowed automated isotope ratio determinations of small gas samples after 50 to 200 ul of water had equilibrated with CO; after a set time (following Dugan et al. 1985). Results were expressed in standard notation (6'°O) in parts per thousand (%o) relative to V-SMOW, with a sample precision of +0.05%o. The proportions of deep water and surface water used by the shrubs and the conifers were calculated using the following equations: V surface a V deep aa V plant (2) where Vpiant 1S the total water taken up by the plant, Vsurface 18 the fraction of water from upper soil layers (0-50 cm) and Vgcep 1s the fraction of water from deeper soil layers (>50 cm). tl tl = tl surface ¥ surface + Caeep Vdeep _ plant ¥ plant (3) ce surface = ee V plant (4) t2 Vsurface + Caeep Veep plant where C'! and C” are the concentrations of the 6o°H and 46'%O tracers, respectively, and the subscripts surface, deep and plant refer to water from the upper soil layers, deeper soil layers and xylem sap of the plant, respectively. With one hydrological tracer for the source of xylem sap water, two mass balance equations can be written (Eqs. 2, 3), making it possible to determine the proportion of uptake from two sources. This study had two hydrological tracers, which could allow uptake from three sources to be deter- mined, but because of the covariance between '*O and °H due to kinetic effects (Dansgaard 1964) it was only possible to determine the proportion of uptake from two sources. For this reason, we divided the soil profile into just two layers, 0- 50 cm and >50 cm deep. Analysis of variance was used to compare differences between closed-canopy and manzani- ta shrub areas and between treatments using the GLM ANOVA procedure of the SPSS statistical package (version 10, 1999, SPSS Inc., Chicago, IL), and differences were assumed to be statisti- cally significant if P < 0.1. RESULTS Diurnal temperature differences during the growing season (June-August) were larger in the shrub-dominated areas than in the closed- canopy patches (Far) = 8110, P < 0.001; Fig. 2) and greater near the ground, at a 20 cm height, than 150 cm above the ground (Far, = 1625.4, P < 0.001; Fig. 2). The soil profile was significantly PLAMBOECK ET AL.: CONIFER SEEDLING SURVIVAL IN THE SIERRA 195 drier in the shrub than in the closed-canopy patches during the 2001 growing season (Far, = 6.728, P = 0.016). The increase in soil moisture at the end of 2001 was caused by an unusual rain storm in the early fall. During the 2002 growing season there was no significant difference in soil moisture between shrub and closed-canopy patches (Far) = 2.502, P = 0.145; Fig. 3). For all sites, the soil profile was drier during the 2001 growing season than in 2002 (Fap = 88.4, P < 0.001). The seismic data indicate that the active soil profile is deeper in the closed-canopy areas (2.8 + 0.5 m) than in manzanita areas (1.6 + 0.28 m), but the depth to the solid bedrock was ~3.3m in both the closed-canopy and shrub areas. In the beginning of the growing season, the soil profile had 6'°O and SH values of —13%o and 95%o, respectively, because of the water from the snow melt. As the upper layers in the soil profile gets drier and the 6'°O values in the soil become less negative, the 6°H values increase. Dwell water had 6'%O and 5°H values of —13 and 93%bo, respectively, during the whole growing season. Plant isotope values differed between shrubs, trees, and saplings. The overstory sugar pines and white firs had consistent 6'°O values over the whole growing season. Throughout the June— September sampling period, they extracted water with 6'°O values of —12 to —13%o similar to soil profiles >70 cm. Both shrub and conifer sapling water composition changed during the sample period. The shrubs having 6'°O of —12%o in early growing season (May-July 1*) after which the 6'8O values became less negative in September. Unfortunately, we discovered that the 50 um mesh did not prevent fine manzanita roots growing through the mesh as we had intended, though it did exclude tree roots from cylinders in the closed canopy-patch. Due to this compromise in our design, our seedlings in the manzanita patches had only two treatments; mycorrhizae exclosure and no exclosure. The closed-canopy patches, however, still had three treatments. Transpiration from seedlings in all three cylinder treatments under the closed-canopy increased with irrigation volume (Fars = 2.869, P = 0.035; Fig. 4). In closed-canopy conditions, there was also a significant difference between species, with sugar pine seedlings having greater transpi- ration rates (1.66 mmol m ’ s_') than white fir seedlings (1.4 mmol m~’ s"'; Far) = 3.925, P = 0.065). The seedlings in the 50 um pore-size mesh and 5 um pore-size meshes that were not irrigated had similar transpiration rates in both the manzanita and closed-canopy patches, while the transpiration rates from the seedlings in the control treatment were greater in the closed- canopy than in the manzanita patches (Far = 6.36, P = 0.086; Fig. 4). The vapor pressure deficit (VPD) was greater in the manzanita areas 196 MADRONO Manzanita areas eal My AVE Yah MN ALS A | | MAP| yy AIT | ed AERA PA FT yi a EA Ts | S| re ee A ae | Tce WWTP, VIN Ae | oe | | et Temperature (°C) HAIN LAMA A | HEE a) MY INT 05-16-01 07-16-01 09-16-01 11-16-01 01-16-02 03-16-02 05-16-02 07-16-02 09-16-02 [Vol. 55 Closed-canopy areas Le | 50cm ia, a a eA tr i a — a a 6 Eig C2 ees Fy AW VIVE yy ll MPR | aes | | ALAN TAT aR a | PF | | a A 20cm 05-16-01 07-16-01 09-16-01 11-16-01 01-16-02 03-16-02 05-16-02 07-16-02 09-16-02 Date FIG. 2 Mean daytime (solid line) and night time (dotted gray line) temperatures from two different heights above the ground, 150 cm (a, c) and 20 cm (b, d). Figures a and b show temperatures from the manzanita patches, while c and d show temperatures from the closed-canopy patches. than in the closed-canopy areas during the growing season (Fg) = 7.823, P = 0.031; Table 1), but it did not differ between the areas in June to August. Survival rates of both sugar pine and white fir seedlings at the end of 2002 were greater in the closed-canopy than in the manzanita patches (Far) = 9.359, P = 0.03). In the closed-canopy patches, the survival rates for both species did not differ between the different mesh treatments. However, in the manzanita patches, both white fir and sugar pine seedlings in the 5 um pore-size mesh treatment had greater survival rates than seedlings in cylinders which did not exclude manzanita roots (Far) = 9.365, P = 0.092; Fig. 5). DISCUSSION Our study suggests that tree seedlings may be water stressed in shrub patches until they develop a root structure that can access relatively deep soil water. We found no evidence suggesting that seedlings that could establish mycorrhizal con- nections were any less water stressed than seedlings growing in soil where tree roots were excluded. Low survival rates of tree seedlings and the strong positive response in transpiration rates when irrigation was provided suggest that dry summers and shrub competition for water resources may be strong influences on seedling survivorship. To facilitate seedling establishment and survival, managers may need to make greater use of prescribed fire because it is more effective at reducing moisture competing shrubs than thinning in mixed-conifer forests (Wayman and North 2007). The seismic information on the active soil — profile suggests that in our study area closed- | canopy forest was on deeper soils with a higher — water holding capacity than in the manzanita | areas (Meyer et al. 2007). Because of the dry | summers associated with the Mediterranean © climate of the study site, during the growing © season, plants depend on available water that has accumulated in the soil from snow melt in May. However, in a study based in an area with similar soils in southern California, Hubbert et al. (2001) found that calculations of plant-available water may need to include the weathered bedrock | profile, from which roots of Jeffrey pine extracted | a significant portion of late summer moisture. At Teakettle Forest, the depth to the solid bedrock | in both the closed-canopy and shrub areas was ~3.3m. We do not know, however, whether overstory white fir and sugar pine were extracting | water from this weathered bedrock layer. Isotope signatures suggest that overstory trees | rely on deep water during the whole growing 2008] Manzanita areas PLAMBOECK ET AL.: CONIFER SEEDLING SURVIVAL IN THE SIERRA 197 Closed-canopy areas Volumetric water content (%) >a panaeoecekclesaaeouDMmMteaelketlememlhlUCOOUC] SST S557 3535 09 D6 66 = = - we pp ZS TH WH i ee ES a MONI S'. , ue, Oy OE — i oN Oe OT 2 FIG. 3. during the 2001 and 2002 vegetation seasons. season, while shrubs rely on water from shallower horizons at the beginning of the growing season and as the soils become drier the primary zone of water uptake shifts to deeper horizons (Fig. 6). Tree saplings show a similar shift in their primary zone of water uptake to the shrubs; utilizing shallow horizons early in the season, then deeper layers as the soil becomes drier. This utilization pattern would put shrubs and tree saplings in No irrigation Irrigation > wanc flo BOUlBTlUCcUMDCUMOUCORUlUCUCUNlOlUCL.DOUUC/!hUCUCOO ee s55 5 S353 09088686 S=e2arrer 6 ft tMNE 5S Mr PF —-— NN DO WT WH DM aa a ee ON OK OCT © Volumetric water contents (%) in a manzanita and a closed-canopy patch at eight different soil depths direct competition for water early in the growing season (May and June). Furthermore, tree seedlings, which are known to be shallow-rooted, with roots down to ~20—-30 cm, would be competing with shrubs and tree saplings for water during the whole growing season. Rose et al. (2003) found similar differences in patterns of water between Jeffrey pine and manzanita shrubs. Their results suggest that 25-30 yr old Double irrigation Manzanita areas y) @ Closed-canopy areas Transpiration (mmol m°s"') 50um 5um control 50um control 5um control 5um FIG. 4. Transpiration in September 2002 from seedlings in the cylinder treatments in manzanita and closed— canopy areas that had been exposed to different irrigation treatments for three weeks. Bars show +1 SE (n = 3). 198 MADRONO TABLE 1. MONTHLY MEAN VALUES FOR THE VAPOUR PRESSURE DEFICIT IN THE MANZANITA AND THE CLOSED-CANOPY AREAS DURING THE VEGETATION SEASON OF 2002. Manzanita Forest Vapour pressure Vapour pressure Month deficit (mbar) + std deficit (mbar) + std May 6.8 + 1.0 fy areal a) June 11.3 2 1:36 | ee eae, July 16.7 + 1.9 16.6 + 3.2 August 14.9 + 0.7 [asia 7 September 15.8 0.9 | be ean) Oa October 99 + 0.9 a Vso | Jeffrey pine use more bedrock water and have a functional rooting depth that is slightly deeper than the manzanita, which rely on shallow soil water. Anderson and Helms (1994) showed that manzanita shrubs were able to extract more water from moisture-depleted soil profiles than pines, perhaps by achieving greater reductions in total water potentials. Osmotic adjustments and the consequent ability to use soil water held at relatively low water potentials are traits common to both Jeffrey pine and manzanita (Anderson and Helms 1994). Similar studies have not been Manzanita areas l SUrV1Va Percent (%) Fic. 5. [Vol. 55 conducted with white fir, although it is generally considered to be much less drought tolerant than Jeffrey pine (Minore 1979; Royce and Barbour 2001). Unlike pine, however, manzanita also has low cell elasticity (Anderson and Helms 1994). Consequently, small reductions in relative water content lead to relatively large reductions in turgor pressure in manzanita shrubs, allowing them to use water held at lower water potential than pine trees. Several investigators have reported that asso- ciated mycorrhizae increases drought tolerance and recovery from water stress in various woody species (Parke et al. 1983; Boyle and Hellenbrand 1991; Smith and Read 1997). Fungal hyphae and rhizomorphs extend the root system, entering soil pores too small to be penetrated by roots, and often have longer lifespans than fine roots (Parke et al. 1983). Several laboratory-based labeling experiments have also found indications that carbon is transported between plants connected via ectomycorrhizal and vesicular arbuscular networks (Read et al. 1985; Finlay and Read 1986; Watkins et al. 1996; Simard et al. 1997; Fitter et al. 1998). Furthermore, Querejeta et al. (2003) found that water may be transferred from oaks to their mycorrhizal symbionts during severe soil drying. In a laboratory-based experi- Closed-canopy areas iene < 2S a SS i a ings 100.0 90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 ea Re eee a ee eae ee | ugar pine seedling Ma ee ee je es = Pe a eas eS es 100.0 90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 Date Percentage survival of seedlings in the different treatments in manzanita areas and closed-canopy areas © during 2002. Triangles represent the control treatment, filled squares the 50 um pore-size mesh treatment and open squares the 5 um pore-size mesh treatment. Bars show +1 SE (n = 27-45). | 2008] 120 8 100 Adult trees s © 80 4) = o> 60 0) xe) oO 40 Cc © 5 20 oO 0 May June FIG. 6. PLAMBOECK ET AL.: CONIFER SEEDLING SURVIVAL IN THE SIERRA 199 Tree saplings Manzanita (Arctostaphylos spp.) July August Uptake depths of water for shrubs, overstory trees (>40 m tall) and tree saplings (1-3 m tall) during May to August 2001. Water from below 50 cm of depth in the soil profile is considered to be deep water. Bars show +1 SE (n = 3-8). ment, using tripartite mesocosms containing manzanita and young seedlings of sugar pine and Douglas-fir (Pseudotsuga menziesii), Plam- boeck et al. (2007) showed that water can be transported via mycorrhizal hyphae to conifer seedlings. If the transport of resources from shrubs or mature trees to seedlings was important for the survival of the seedlings in our system, the seedling survival rates in the control and the 50 um pore-size mesh treatment would have been higher than in the 5 um pore-size mesh treatment. However, seedling survival rates in the 5 um pore-size mesh treatment were not significantly less than survival rates in the two treatments that allowed associations with an established mycelial network (Fig. 5). Survival rates were actually greater for tree seedlings planted in the manzanita patches when they were enclosed in the 5 um pore-size mesh, possibly because this treatment successfully excluded manzanita roots. Our results show that conifer seedling survival rates were greater in the closed tree canopy than in patches dominated by manzanita shrubs. Isotope signatures suggest that in closed-canopy patches there may be soil-water partitioning between mature canopy trees (deep) and estab- lishing seedlings (shallow) that reduces competi- tion for critical water resources. In manzanita patches, however, we found that water competi- tion between tree seedlings and shrubs negatively affected seedling establishment. Treatments that allowed the establishing seedlings to tap into an already existing ectomycorrhizal network did not _ enhance their survival. Shrubs may help modify _ microclimate conditions compared to open gaps, _ but diurnal temperature fluctuations were still much greater than in closed-canopy forest. When roots were not excluded (the control treatment), seedling transpiration was very low and survival rates were significantly lower in manzanita patches than in closed-canopy forest. These results suggest that shrubs may inhibit the survival of tree seedlings until they establish roots deep enough to extract moisture from soil profiles below the depths at which the shrubs root. With fire suppression, shrub cover has signif- icantly increased in some mixed-conifer forests (Parson and DeBenedetti 1979; Minnich et al. 1995). Our study suggests that forest managers may need to reduce shrub cover with treatments such as prescribed fire to enhance survival of conifer seedlings and facilitate restoration of the historic forest composition. ACKNOWLEDGMENTS Financial support was provided by the Swedish Foundation for International Cooperation in Research and Higher Education, STINT (Dnr 99/666), the University of California and the Mellon Foundation. We thank Professors Tom Bruns and Tom Parker for their feedback on many aspects of this investigation, Teakettle’s site managers Nate Williams and Jim Innes, and all the field crews that worked at the Teakettle Experimental Forest from 2000-2002 for their assis- tance in the field. LITERATURE CITED ANDERSON, P. D. 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—150S. 200 ANONYMOUS. 1993. Soil survey of Sierran National Forest areas, California. USDA Forest Service Pacific Southwest Region, Clovis, CA. BOYLE, C. D. AND K. E. HELLENBRAND. 1991. Assessment of the effect of mycorrhizal fungi on drought tolerance of conifer seedlings. Canadian Journal of Botany 69:1764-1771. DANSGAARD, W. 1964. Stable isotopes in precipitation. Tellus 16:436—468. DUGAN, J. P., J. BORTHWICK, R. S. HARMON, M. A. GAGNIER, J. E. GLAHN, E. P. KINSEL, S. MACLEopD, J. A. VIGLINO, AND J. W. HEss. 1985. Guanidine hydrochloride method for deter- mination of water oxygen isotope ratios and the oxygen-18 fractionation between carbon dioxide and water at 25°C. Analytical Chemistry 57: 1734-1736. EHLERINGER, J. R., J. RODEN, AND T. E. DAWSON. 2000. Assessing ecosystem-level water relations through stable isotope ratio analyses. Pp. 181-198 in O. Sala, R. Jackson, and H. A. Mooney (eds.), Methods in ecosystem science. Academic Press, San Diego, CA. ERICKSON, H. E., P. SoTo, D. W. JOHNSON, B. ROATH, AND C. HUNSAKER. 2005. Effects of vegetation patches on soil nutrient pools and fluxes within a mixed-conifer forest. Forest Science 51:211—220. FINLAY, R. D. AND D. J. READ. 1986. The structure and function of the vegetative mycelium of ectomycorrhizal plants: I. translocation of 14C- labeled carbon between plants interconnected by a common mycelium. New Phytologist 103:143—156. FITTER, A. H., J. D. GRAVES, N. K. WATKINS, D. ROBINSON, AND C. SCRIMGEOUR. 1998. Carbon transfer between plants and its control in network of arbuscular mycorrhizas. Functional Ecology 12:406-412. GOMEZ-APARICIO, L., R. ZAMORA, J. M. GOMEZ, J. A. HODAR, J. CASTRO, AND E. BARZA. 2004. Apply- ing plant facilitation to forest restoration: a meta- analysis of the use of shrubs as nurse plants. Ecological Applications 14:1128—1138. GRAY, A. N. AND T. A. SPIES. 1995. Water content measurement in forest soils and decayed wood using time domain reflectrometry. Canadian Jour- nal of Forest Research 25:376—385. GRAY, A., H. ZALD, R. KERN, AND M. NortTHu. 2005. Stand conditions associated with tree regeneration in Sierran mixed-conifer forests. Forest Science 51:198—210. HELMS, J. A. AND J. C. TAPPEINER. 1996. Silviculture in the Sierras. Pp. 439-476 in Sierra Nevada Ecosystem Project: final report to Congress, Vol. II, assessments and scientific basis for management options. Center for Water and Wildland Resources, University of California, Davis, CA. HICKMAN, J. C. (ed). 1993. The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. HorTon, T. R., T. D. BRUNS, AND V. T. PARKER. 1999. Ectomycorrhizal fungi associated with Arc- tostaphylos contribute to Pseudotsuga menziesii establishment. Canadian Journal of Botany 77:93— 103. HUBBERT, K. R., J. L. BEYERS, AND R. C. GRAHAM. 2001. Roles of weathered bedrock and soil in seasonal water relations of Pinus jeffreyi and MADRONO [Vol. 55 Arctostaphylos patula. Canadian Journal of Forest Research 31:1947—1957. McDONALD, P. M. 1976. Forest regeneration and seedling growth from five major cutting methods in north-central California. Research Paper PSW-115. USDA Forest Service, Pacific Southwest Forest and Range Experimental Station, Berkeley, CA. MEYER, M., M. NorTH, A. GRAY, AND H. ZALD. 2007. Influence of soil thickness on stand charac- teristics in a Sierra Nevada mixed-conifer forest. Plant and Soil 294:113-123. MINNICH, R. A., M. G. BARBOUR, J. H. BURK, AND R. F. FERNAU. 1995. Sixty years of change in Californian conifer forest of the San Bernardino Mountains. Conservation Biology 9:902-914. MINORE, D. 1979. Comparative autecological charac- teristics of Northwestern tree species: a literature review. General Technical Report GTR-PNW-87. United States Department of Agriculture, Forest Service, Pacific Northwest Research Station, Port- land, OR. MOLINA, R. AND J. M. TRAPPE. 1982. Lack of mycorrhizal specificity by the ericaceous host Arbutus menziesii and Arctostaphylos uva-ursi. New Phytologist 90:495—509. NorTH, M., B. OAKLEY, J. CHEN, H. ERICKSON, A. GRAY, A. I[ZZo, D. JOHNSON, S. MA, J. MARRA, M. MEYER, K. PURCELL, T. RAMBO, B. ROATH, D. RIZZO, AND T. SCHOWALTER. 2002. Vegetation and ecological characteristics of mixed-conifer and red fir forests at the Teakettle Experimental Forest. USDA Forest Service General Technical Report PSW-186. United States Department of Agricul- ture, Forest Service, Pacific Southwest Research Station, Albany, CA. , M. HURTEAU, R. FIEGENER, AND M. BAR- BOUR. 2005. Influence of fire and El Nino on tree recruitment varies by species in Sierran mixed conifer. Forest Science 51:187—197. PARKE, J. L., R. G. LINDERMAN, AND C. H. BLACK. 1983. The role of ectomycorrhizae in drought tolerance of Douglas-fir seedlings. New Phytologist 95:300—-304. PARKER, A. J. 1994. Latitudinal gradients of coniferous tree species, vegetation, and climate in the Sierran- Cascade axis of northern California. Vegetatio © 115:145-155. PARSONS, D. J. AND S. H. DEBENEDETTI. 1979. Impact of fire suppression on mixed-conifer forest. Forest Ecology and Management 2:21-33. PERRY, D. A., M. P. AMARANTHUS, J. G. BORCHERS, S. L. BORCHERS, AND R. E. BRAINERD. 1989. Bootstrapping in ecosystems. BioScience 39:230— D3: PLAMBOECK, A. H., T. E. DAWSON, L. M. EGERTON- WARBURTON, M. NorTH, T. D. BRUNS, AND J. I. QUEREJETA. 2007. Water transfer via ectomycor- © rhizal fungal hyphae to conifer seedlings. Mycor- | rhiza 17:439-447. QUEREJETA, J. I., L. M. EGERTON-WARBURTON, AND | M. F. ALLEN. 2003. Direct nocturnal water | transfer from oaks to their mycorrhizal symbionts during severe soil drying. Oecologia 134:55—64. READ, D. J., R. FRANCIS, AND R. D. FINLAY. 1985. Mycorrhizal mycelia and nutrient cycling in plant | communities. Pp. 193-217 in A. H. Fitter (ed.), Ecological interaction in soil. Blackwell Scientific Publications, Oxford, United Kingdom. 2008] RosE, K. L., R. C. GRAHAM, AND D. R. PARKER. 2003. Water source utilization by Pinus jeffreyi and Arctostaphylos patula on thin soils over bedrock. Oecologia 134:46—54. Royce, E. B. AND M. G. BARBOUR. 2001. Mediterra- nean climate effects: I. conifer water use across a Sierra Nevada ecotone. American Journal of Botany 88:911-918. SIMARD, S. W., M. D. JONES, D. M. DURALL, D. A. PERRY, D. D. MYROLD, AND R. MOLINA. 1997. Reciprocal transfer of carbon isotopes between ectomycorrhizal Betula papyrifera and Pseudotsuga menziesii. New Phytologist 137:529—542. SMITH, S. E. AND D. J. READ. 1997. Mycorrhizal symbiosis. 2nd ed. Academic Press, London, United Kingdom. STEPHENSON, N. L. 1998. Actual evapotranspiration and deficit: biologically meaningful correlates of PLAMBOECK ET AL.: CONIFER SEEDLING SURVIVAL IN THE SIERRA 201 vegetation distribution across spatial scales. Jour- nal of Biogeography 25:855—870. TAPPEINER, J. C. AND P. M. MCDONALD. 1996. Regeneration of Sierra Nevada forests. Ch. 12 in Sierra Nevada Ecosystem Project: final report to Congress, Vol. III, assessments, commissioned reports, and background information. Centers for Water and Wildland Resources, University of California, Davis, CA. WATKINS, N. K., A. H. FITTER, J. D. GRAVES, AND D. ROBINSON. 1996. Carbon transfer between C3 and Cy, plants linked by a common mycorrhizal network, quantified using stable carbon isotopes. Soil Biology and Biochemistry 28:471-477. WAYMAN, R. AND M. Nor Tu. 2007. Initial response of a mixed-conifer understory community to burning and thinning restoration treatments. Forest Ecol- ogy and Management 239:32-44. MADRONO, Vol. 55, No. 3, pp. 202-215, 2008 THE ALPINE FLORA OF THE WHITE MOUNTAINS, CALIFORNIA PHILIP W. RUNDEL Department of Ecology and Evolutionary Biology and Center for Embedded Networked Sensing, University of California, Los Angeles, CA 90095 rundel@biology.ucla.edu ARTHUR C. GIBSON AND M. RASOUL SHARIFI Department of Ecology and Evolutionary Biology, University of California, Los Angeles, CA 90095 ABSTRACT The alpine zone of the White Mountains of California, defined as non-forested areas above 3500 m, includes 163 native species of vascular plants in an area of 106 km’. No invasive species have become well established. Nearly two-thirds of the native species occur in just seven families, led by the Asteraceae with 30 species. Six genera have five or more species, led by Carex with 14 species. Life forms of the flora are heavily dominated by broad-leaved herbaceous perennials (53%), followed in importance by graminoid perennials (22%) and mats and cushions (11%). Woody shrubs, chamaephytes (low subshrubs), and annuals are relatively few in number, and those species present are generally more characteristic of lower elevation communities. Fellfields form the characteristic habitat for 41% of the flora, while moist meadows and open slopes habitats characterize 24 and 22% of the flora, respectively. Only 31% of the flora is restricted in the White Mountains to the alpine zone, while nearly a third of the alpine flora has a range extending to lower elevations of the montane or cold desert zones below 2900 m. The alpine flora of the White Mountains shares over 70% of its species with the Sierra Nevada. Only three species are endemic to the alpine zone of the White Mountains: Draba californica, D. monoensis, and Potentilla morefieldii. Key Words: alpine biogeography, alpine flora, plant life forms, White Mountains. Generalizations about the patterns of biodi- versity and life forms in alpine floras have largely come from research conducted in relatively mesic alpine habitats with summer rainfall regimes, as for example in the Alps of central Europe and the Rocky Mountains of the western United States (see reviews by Tranquillini 1979; Chapin and Korner 1995; Korner 1999; Bowman and Seast- edt 2001). Unlike the majority of alpine regions in the Northern Hemisphere that share elements of a circumboreal arctic-alpine flora, the high mountain ranges in California have developed a unique alpine flora under the influence of mediterranean-climate conditions with relatively dry summers added to other alpine environmen- tal stresses. Although broad ecological studies have been carried out in the alpine elevations of the Sierra Nevada and White Mountains of California (Chabot and Billings 1972; Major and Taylor 1977; Morefield 1992), there is still a relatively poor knowledge of patterns of floristic diversity and life form distribution of the alpine flora for these ranges. The White Mountains of California and adjacent Nevada (Fig. 1) present a particularly interesting area for study. This mountain range is positioned at the interface between two major geomorphic provinces, the Sierra-Cascade Prov- ince and the arid Basin and Range Province, yet is isolated from direct contact with high eleva- tions of either sets of ranges. Moreover, warmer and more xeric climatic conditions during the Altithermal period of the early Holocene allowed an upward movement of subalpine conifers, restricting the area available for growth of alpine communities (Jennings and Elliot-Fisk 1991, 1993). Thus, the White Mountains present an example where both climate history and geo- graphic isolation have played significant roles in the evolution of the alpine flora. Lying in the rain shadow of the Sierra Nevada, the White Mountains receive only about one-third of the precipitation reaching similar elevations on the west slope of the Sierra Nevada. These arid. conditions, combined with the extremes of low. temperature, wind, low atmospheric pressure and | high ultraviolet radiation load that characterize. most temperate alpine conditions, produce unusu- | ally severe conditions of environmental stress for. plant growth. Adding to the habitat conditions of) the range are high levels of both topographic. diversity and geologic complexity, with the latter including granites, dolomites, shales, limestones, and metavolcanics (Nelson et al. 1991; Ernst et al.! 2003) that have strong influences on florisitic. composition (Lloyd and Mitchell 1973; Marchand | 1973; Rundel et al. 2005). | The White Mountains extend approximate- | ly 60 km from their northern end in Nevada across into California and Westgard Pass in the 2008] [s:Jarea above 3500m Mono County Inyo County Fic. 1. that portion of the range above 3500 m elev. south, or three times this length if the contigu- ous Inyo Range south of this pass is added (Fig. 1). The range is very narrow, however, with -a width averaging only 20-25 km. As a result, the range rises sharply from elevations of 1250 m in the Owens Valley to a high point of 4343 m at White Mountain Peak over a lateral distance of less than 10 km. This peak is tied with Mount Shasta as the third highest summit in California. There has been a long history of floristic and ecological studies in high elevations of the White Mountains where alpine habitats are present. The RUNDEL ET AL.: ALPINE FLORA OF THE WHITE MOUNTAINS 203 Regional view of the White Mountains of eastern California and adjacent Nevada. The stippled area is floristic diversity and relationships of the full flora of the White Mountains has been described in considerable detail (Lloyd and Mitchell 1973; Morefield et al. 1988; Morefield 1992). Although there have been ecological studies of plant-soil relationships in the alpine regions of the range (Mooney et al. 1962; Mitchell et al. 1966; Marchand 1973; Ernst et al. 2003), no floristic research has focused on the alpine region above 3500 m or analyzed its characteristics specifically. In this paper we present a broad overview of the alpine flora of the White Mountains by providing a detailed analysis of the floristic richness, 204 ecological diversity, and biogeographic relation- ships of the alpine flora present within this zone. MATERIALS AND METHODS Floristic Richness and Life Forms The alpine flora of the White Mountains in this study was considered to include all species with a known occurrence above 3500 m, an elevation limit roughly corresponding to upper treeline, although Pinus longaeva reaches elevations of up to about 3700 m in scattered locations. While there are certainly alpine-like communities and species assemblages below this elevational limit, an occurrence above the upper limit for growth of P. longaeva and P. flexilis indicates a definitive alpine habitat. Our listing of the alpine species to be considered was based on a careful examination of published material (Lloyd and Mitchell 1973; Morefield et al. 1988; Morefield 1992; Hickman 1993), herbarium records (Cal Flora, White Mountain Research Station), and our own field observations carried out on numerous visits each summer from 1999-2006 . Scientific names used here follow those of Hickman (1993), with the exception of Poa pattersonii which has been divided into Poa abbreviata subsp. pattersonii and subsp. marshii (Soreng 1991). Adapting the broad classification scheme set out by Morefield (1988, 1992) for the entire flora of the White Mountains, we developed a system to categorize the alpine species by life form, ecological habitat, elevational zone, and biogeographic distri- bution. Each alpine species was placed into one of six life forms in a modified Raunkiaer (1934) classification: phanerophytes (shrubs reaching 50 cm or more in height), chamaephytes (subshrubs lower in stature), mat or cushion plants (<10 cm in height and prostrate in growth, perennial grami- noids (i.e., Poaceae, Cyperaceae, and Juncaceae), broad-leaved herbaceous perennials (tussocks, ro- sette perennials, biennials, and geophytes), and annuals (therophytes). With very few exceptions, plant species in the alpine zone have very narrow leaves or leaflets with blades <10 mm wide, and would be classified as leptophylls with leaf surface areas <25 mm/? (Raunkiaer 1934). Designations of ecological habitats were adapt- ed from characterizations of Morefield (1988, 1992) and our experience to provide seven categories. Along a rough gradient of mesic to xeric these ecological habitats are aquatic sites, wet sites (areas with saturated soils and riparian habitats), moist sites (e.g., wet meadows and areas with snow melt accumulation), fellfields with seasonal moisture availability, talus slopes, open slopes, and dry rocky slopes. Where a species occurred broadly across more than one of these habitat categories, it was placed in what we considered its most typical habitat. MADRONO [Vol. 55 The classification of species by characteristic elevational zone was designed to separate obligate alpine species from those extending above 3500 m but also occurring at lower elevations. These five categories listed as the elevational belts of lowest occurrence in the White Mountains are cold desert (1220-1980 m), montane (1980-2900 m), subalpine (2900-3500 m), alpine (3500-4000 m), and high alpine elevations (4000-4332 m). The montane zone roughly corresponds to the pinyon- juniper zone and the subalpine belt to the upper pine zone of vegetation. The biogeographic range of each alpine species was classified into one of five categories. These were: widespread species present in many habitats or regions throughout the world or across North America, cordilleran species widespread in moun- tain regions of the western United States, Sierra/ Cascade species, intermountain species wide- spread across the Great Basin, and species endemic to the White Mountains. Climate Regimes Climatic data to characterize the alpine envi- ronment of the White Mountains was taken from — long-term records collected at the Barcroft— Station at 3801 m elev. (37°35’N lat., 18°15’W long.) for the period 1953-1973 (Pace et al. 1974; | Powell and Klieforth 1991). The mean monthly | maximum temperatures at Barcroft vary from a > high of 11.9°C in July to a low of —5.3°C in) February (Fig. 2). Record maximum tempera- : tures of 22°C have been reached in July and) August. Mean monthly maximum temperatures | remain below freezing for six months of the year, | from November through April. Mean minimum | temperatures range from a high of 2.4°C in July | to a low of —14.0°C in March. Mean minimum | temperatures drop below freezing for every > month of the year except July and August. Mean annual precipitation at Barcroft Station 1s 478 mm (Fig. 2). There are elements of a mediter- | ranean-type pattern of winter precipitation but | with a strong influence of summer convective . storms from the east that bring scattered precip- itation events throughout the growing season. | Mean monthly precipitation ranges from a high | of 56 mm in December to a low of 18 mm in| September, but year-to-year variation is high. The : extremes in annual precipitation over the record | period have ranged from 242 to 852 mm. With the | exception of the summer months of July through | September, all of this precipitation falls as snow. RESULTS Native Flora Defining the alpine zone as non-forested areas | occurring at or above 3500 m, the alpine flora of | RUNDEL ET AL.: ALPINE FLORA OF THE WHITE MOUNTAINS 205 60 50 40 30 20 Precipitation (mm) 10 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Time (date) 2008] 15 10 O oe S 6 SE oO 2 Ge = = > -10 Ay -15 -20 FIG. 2. Data from Pace et al. (1974). the White Mountains includes 163 native species occurring in an area of 106 km’. Rather than representing the proportional family relationships of the California flora overall, the alpine flora has just seven families that account for nearly two- thirds of its total species. Leading this group is the Asteraceae with 30 species, followed in order by the Brassicaceae (18 species), Poaceae (17 species), Cyperaceae (15 species), Rosaceae (9 species), Caryophyllaceae (9 species), and Poly- gonaceae (7 species). Genera with large numbers of alpine species in this flora include Carex (Cyperaceae, 14 species), Arabis (Brassicaceae, 7 species), Draba (Brassicaceae, 7 species), Poa (Poaceae, 6 species), Potentilla (Rosaceae, 5 species), and Eriogonum (Polygonaceae, 5 spe- cies). Several species just miss our lower eleva- tional limit of 3500 m, and thus are not included in our figures here, although they occur in alpine fellfield and open slope habitats. These include Antennaria dimorpha (Asteraceae), Tonestus peir- sonii (Asteraceae), and Eriogonum umbellatum var. covillei (Polygonaceae). The life form distribution of the native alpine flora of the White Mountains is strongly domi- hated by herbaceous perennials, as is typical of temperate alpine habitats elsewhere. Within herbaceous perennials, the assemblage of peren- “nial graminoids (most notably grasses and sedges) form 22% of the total flora, mats and cushions form 11%, and other herbaceous perennials such as broad-leaved tussocks, rosettes, and biennials form 53% (Fig. 3). Combined, herbaceous peren- | l -nials thus comprise 86% of the flora. _ There are 18 native species of mats and cushions, all low in stature and mostly relatively small in size, that make up a heterogeneous group Mean monthly climatic conditions at the Barcroft Station at 3801 m in the White Mountains, 1953-1973. that shares the characteristic of a prostrate growth form with either a central taproot or multiple points of rooting through layering. These form an ecologically significant component of plant cover on fellfield slopes and the margins of drainage swales. A different group of species is present, however, dominates on granitic versus dolomitic substrates. Common mats and cush- ions on the widespread granitic substrates include Draba oligosperma (Brassicaceae), Trifolium an- dersonii var. beatleyae (Fabaceae), Eriogonum ovalifolium var. nivale (Polygonaceae), and Pen- stemon heterodoxus var. heterodoxus (Scrophular- iaceae). On dolomitic substrates, there is a shift in dominance to species such as Erigeron pygmaeus and Stenotus acaulis (Asteraceae), Astragalus kentrophyta and Oxytropis parryi (Fabaceae), Linum lewisii (Linaceae), Eriogonum gracilipes (Polygonaceae), and Castilleja nana (Scrophular- iaceae). These dolomite species , however, may occur on other soil parent materials over their ranges. Shrubs (phanerophytes) and subshrubs (cham- aephytes) are few in the alpine flora of the White Mountains. Just three shrub species reaching heights of 50 cm have a range that extends above 3500 m: Ribes cereum (Grossulariaceae), Poten- tilla fruticosa (Rosaceae), and Salix orestera (Salicaceae). All of these shrub species are more typical of lower elevations. Only Ribes cereum 1s common in the alpine zone, and its distribution is highly correlated with large granite outcrops where thermal and hydrologic characteristics may be significant in providing favorable micro- sites for this species. Seven species of low woody subshrubs are present, representing just two families. These are 206 100 80 60 40 Number of species 20 MADRONO [Vol. 55 shrub — subshrub mat/ herb. | graminoid annual Cushion perennial perennial Fic. 3. Life form distribution of the native vascular plant species in the alpine zone above 3500 m in the White Mountains. Artemisia rothrockii, Chrysothamnus viscidiflorus subsp. viscidiflorus, Ericameria discoidea, and E. suffruticosa (Asteraceae), and Leptodactylon pun- gens, Linanthus nuttallii subsp. pubescens, and Phlox pulvinata (Polemoniaceae). These three Polemoniaceae are quite low in stature in their alpine habitats and approach becoming mats in life form. The highest elevation reached by any of these woody species occurred with Chrysotham- nus viscidiflorus, which extends to near 4000 m at the top of Barcroft Peak. Thirteen annual species are present, making up 8% of the flora. The majority of the annual species that extend their range into the alpine zone are more typical of open habitats at lower montane or subalpine elevations. Most of the annuals are relatively uncommon. The greatest numbers of alpine species in the White Mountain occur in habitats intermediate along a moisture gradient between wet sites with saturated soils and dry slopes (Fig. 4). Drainage courses and associated fellfield habitats with seasonal moisture availability are the typical habitat for 67 species (41% of the flora). Open slopes that become relatively dry in summer are second in significance as the characteristic habitat for 34 species (21% of the flora). Moist meadows are home to 39 species (24% of the flora), mostly graminoids. Only a small number of species typically occur on wet sites with saturated soils (11 species), aquatic habitats (1 species), or dry slopes (4 species). Plant life forms showed mixed patterns of correlation with alpine habitats across a moisture gradient. The number of species represented in many categories, however, is too small for relevant statistical tests of association. Broad- leaved herbaceous perennials, the most common | life form, have virtually the same proportional - distribution among habitats as does the entire flora (Table 1). Perennial graminoids, likewise, generally match the overall pattern but with an. Over-representation of species in moist and wet habitats and an under-representation in drier open habitats. Among the less common life forms | there is a much stronger association with specific — habitats. Annual plant species are strongly | associated with fellfield habitats (77% of species). ; Cushion plants and mats are over represented in | open spaces and rocky habitats, rare in moist . habitats, and absent from wet habitats. Both) chamaephytes and shrubs favor open slopes and moist sites. Separating species into categories of eleva- tional ranges over which they occur demonstrates that 51 species (31% of the flora) are alpine specialists in the White Mountains that are. largely restricted to sites above 3500 m (Fig. 5). However, a few of these alpine-restricted species in the White Mountains, as for example the aquatic Callitriche verna and the wet site Juncus bryoides, occur at lower elevations in wetter’ mountain ranges. Among the alpine species are seven high-elevation alpine specialists that occur only above 3960 m (Morefield et al. 1988). These species are Erigeron vagus (Asteraceae) Anelsonia eurycarpa (Brassicaceae), Cerastium beeringianum (Caryophyllaceae), Polemonium chartaceum (Po-. lemoniaceae), and three grasses; Elymus scribneri, Poa lettermanii, and P. suksdorfii. 2008] 70 60 50 40 30 Number of species 20 10 0 moist meadow aquatic wet FIG. 4. 3500 m in the White Mountains. Another large group of 59 species (36% of the flora) has a range extending into the alpine zone from subalpine forests and shrublands at eleva- tions above 2900 m. Forty-two species (26% of the flora) present in the alpine zone of the White Mountains can also be found in montane habitats as low as 1980 m. Finally, there are six species in the alpine that reach to cold desert elevations as low as 1220 m, as exemplified by Chrysothamnus TABLE 1. RUNDEL ET AL.: ALPINE FLORA OF THE WHITE MOUNTAINS fellfield 20) talus slope rocky slope open slope Lower elevational zone of occurrence for the native vascular plant species present in the alpine zone above viscidiflorus subsp. viscidiflorus that occurs com- monly from 1550-4000 elevation. It is instructive to combine the data on life form and elevational range of occurrence to assess if certain life forms are more likely than others to have broad elevational ranges. Among the alpine flora described here, broad-leaved herbaceous perennials, graminoids, and cushions and mats have 75%, 72% and 81%, respectively, RELATIONSHIP BETWEEN PLANT LIFE FORM AND HABITAT OF CHARACTERISTIC OCCURRENCE FOR THE ALPINE PLANT SPECIES OF THE WHITE MOUNTAINS, CALIFORNIA. Cushion/ Shrub Chamaephyte mat Number of species Aquatic 0 0 0 Wet 0 0 0 Moist 2 2 2 Tallus slope 0 ] 0 Fellfield 0 l 4 Open slope l 3 9 Rocky site 0 0 3 Total 3 q 18 Relative values (%) Aquatic os _ ~- Wet — —_ Moist 50 29 1] Tallus slope — 14 _- Fellfield — 14 22 Open slope 50 43 50 Rocky site — _— hey Total 100 100 100 Broad-leaved Perennial Total herbaceous perennial graminoid Annual flora 0 0 l l fi 4 0 1] 20 1] 2 39 4 2 0 fi 36 16 10 67 18 S 0 37 l 0 0 4 86 36 13 163 -— — 8 0.6 8 1] — 7 | 23 3] 15 24 > 6 — 4 42 44 7 4] 21 8 — 21 l — — 2 100 100 100 99.6 208 MADRONO [Vol. 55 70 60 wy 90 oO ‘Oo a a7 40 ro) oS 30 = | = 20 10 cold desert montane subalpine alpine high alpine (>1220 m) (>1980 m) (>2895 m) (>3500 m) (>3960 m) Fic. 5. Typical habitat occurrence of the native vascular plant species in the alpine zone above 3500 m in the White Mountains. of their species restricted to subalpine and alpine habitats (i.e., elevations above 2900 m). This is a far greater proportion than that present in phanerophytes, chamaephytes and annuals where only 50%, 38%, and 54% of species respectively are limited to these higher elevations. Obligately alpine species in the White Mountains fall almost entirely into just three life forms — broad-leaved herbaceous perennials (33 species), graminoid perennials (17 species), and cushions/mats (3 species). Only one chamaephyte, Phlox conden- sata, has this elevational restriction, and it is quite mat-like in its growth habit. The two other alpine specialists are the wetland annuals Callitriche verna and Juncus bryoides that would be expected to occur at lower elevations in the White Mountains if suitable habitats were present. These data are consistent with the generalization that upright woody plants and annuals have relatively low abundance and diversity in alpine habitats. The biogeographic affinities of the alpine flora show strong links throughout the mountains of the western United States. Only 26 species in the alpine flora are widespread or transcontinental species. The largest group, 55 species (34% of the flora), has broad linkages to western cordilleran regions including the Rocky Mountains and/or Pacific Northwest (Fig. 6). Another 36 species (22% of the flora) are linked directly to the Sierra Nevada/Cascade Ranges. Together then, more than 70% of the alpine flora of the White Mountains is shared with the Sierra Nevada. Strong biogeographic connections to the flora of the Intermountain Ranges are characteristic of 41 species (25% of the flora). Just three species are endemic to the alpine > zone of the White Mountains, while three other — species come close to being endemic. The | endemics are Draba californica, D. monoensis, | and Potentilla morefieldii. The high alpine Pole- | monium chartaceum occurs principally in the | White-Inyo Range with a disjunct occurrence in the Klamath Region of northern California. Recent studies have shown that these populations | are morphologically distinct and deserve more | careful study (Pritchett and Patterson 1998). | Draba subumbellata and Arabis pinzlae reach just | beyond the White Mountains into the eastern | slope of the Sierra Nevada in northwestern Inyo | County, and D. sierrae extends only slightly further into the Sierra Nevada (Price and Rollins 1988; Rollins and Price 1988; Constance-Shull | and Sawyer 2000). Introduced Flora Alpine habitats are generally inhospitable for. alien plant species, and the White Mountains are | no exception. Seven species of introduced plants have been reported from the alpine zone: Senecio | vulgaris (Asteraceae), Capsella bursa-pastoris, | Sisymbrium irio, S. orientale, and Descurainia sophia (Brassicaceae), Stellaria media (Caryo- | phyllaceae), and Chenopodium rubrum (Cheno- | podiaceae). The first six of these are alien to’ California, while the last in this list is now questionably considered to be native. All seven) 2008] 70 60 50 40 30 Number of species 20 10 FIG. 6. Mountains. species, however, owe their presence in the alpine zone of the White Mountains to unintentional human introduction. None of these species is well established, with only Capsella bursa-pastoris reported in multiple small colonies along the roadside. Six of these species are annuals attempting to survive near their upper elevational limits. The one exception is Chenopodium rubrum, which although an annual at lower elevations, takes on a biennial growth form in the alpine zone of the White Mountains (Spira and Wagner 1988). DISCUSSION Floristic Richness and Life Forms The total flora of the White Mountains includes 988 species (Morefield 1992), and thus the alpine flora as defined here thus includes 16% of this flora. Although conditions of the total and seasonal distribution of rains differ in alpine habitats from the Sierra Nevada eastward across the Great Basin to the Rocky Mountains, the relative dominance of a few life forms does not change dramatically across this gradient (Billings 1978, 2000). Herbaceous perennials represent the dom- inant life form in all of these alpine regions. This life form in alpine regions has the characteristic of maintaining large proportions of total biomass belowground where they play an important role RUNDEL ET AL.: ALPINE FLORA OF THE WHITE MOUNTAINS oO ‘> x& Ss Biogeographic affinities of the native vascular plant species of the alpine above 3500 m in the White in carbohydrate storage over the winter months (Mooney and Billings 1960; Billings 1974). The herbaceous perennials include species with a variety of ecological forms and life history strategies of carbon allocation to belowground, aboveground vegetative, and reproductive tissues (Rundel et al. 2005). Within this broadly defined category are perennial graminoids, broad-leaved tussocks, mats and cushions, rosettes, and biennials (Raunkiaer 1934). Many of these are relatively long-lived plants surviving for decades (Billings 1974; Pollak 1991), while others are biennials living for only two years. Perennial graminoids commonly dominate plant communi- ties of wet meadows that dry earlier than fellfield communities. In contrast, fellfield habitats exhibit a mixed dominance of mats and cushions, broad- leaved tussocks, perennial graminoids, and ro- sette plants (Rundel et al. 2005). The perennial graminoids themselves include a diverse set of ecological strategies. Whereas these species comprise a large portion of the species and heavily dominate the cover of wet areas with saturated soils and moist meadows, they also include a number of more drought-adapted fellfield species as well. The fellfield grass Muhlenbergia richardsonis, a species with C4 metabolism, reaches a remarkable elevation of nearly 4000 m at the summit of Mount Barcroft (Sage and Sage 2002). This is likely the highest elevation reached by any C, species in the United States or Canada. 210 Longer-lived species commonly allocate less to reproductive tissues than shorter-lived species. This can be illustrated by the alpine gentians of the White Mountains, for example, where relative allocation to reproduction in the broad-leaved tussock Gentiana newberryi is significantly lower than in the biennial Gentianella tenella and Gentiana prostrata (Spira and Pollak 1986). Such biennial growth forms, however, are relatively uncommon in the alpine flora White Moun- tains where harsh growing conditions select against plants with short life spans. Beyond these two gentians, the other six species of biennials are all members of the Brassicaceae. These are Arabis divaricarpa, A. holboelii var. peduncuocarpa, A. inyoensis, Descurainia incana, Draba albertina, and Halimolobus virgata (More- field et al. 1988). Annual plants, for the same reasons of short and severe growing conditions, are generally rare in the typical circumboreal arctic-alpine floras of the Northern Hemisphere, comprising only 1—-2% of the flora (Billings 2000). Although not abundant, annuals, nevertheless, are relatively more common in the summer dry alpine of the Sierra Nevada and White Mountains where they comprise about 8% of the floras (Jackson and Bliss 1982; Jackson 1985; Spira 1987; Spira and Wagner 1988). Annual plants are largely restrict- ed to specific habitats in the alpine zone of the White Mountains. Sandy drainage channels are the habitat for such annuals as Cryptantha glomeriflora (Boraginaceae), Gayophytum race- mosum (Onagraceae), Gymnosteris parvula (Po- lemoniaceae), and Mimulus suksdorfii (Scrophu- lariaceae) (Morefield 1988; Spira 1991). Such microsites provide relatively high growing tem- peratures and available water early in the growing season, a time critical for annual plants to complete their life cycle. Subshrubs and shrubs are less rich in species in the White Mountains alpine than typical circum- boreal arctic-alpine floras. Their unusually low diversity in the alpine flora of the White Mountains reflects perhaps the relatively dry conditions and low nutrient availability in this area. There has been little study to date of these species in the White Mountains. Biogeographic Patterns of Species Richness A number of checklists of regional alpine floras exist for the western United States, and these can aid in examining patterns of biogeographic relationships. It is interesting to compare patterns of species richness among these alpine areas, although some caution is necessary because of the differing manners in how alpine zones are defined. Alpine regions of the central Rocky Mountains that extend over an extensive area of hundreds of square kilometers include 609 MADRONO [Vol. 55 vascular plant species (Scott 1995). Although a specific enumeration of the alpine flora of the Sierra Nevada has not been made, it has been estimated to be about 600 species (Chabot and Billings 1972). Alpine floras of smaller ranges across the Great Basin vary in size depending on not only the area of habitat present, but also the degree of isolation of the mountain range. In general, these small ranges with relatively low rainfall have smaller alpine floras than comparable areas of alpine habitat in either of the larger and less xeric Sierra Nevada or Rocky Mountains. Ranges lying closer to the Rocky Mountains generally show strong florisitic linkages to this range, while those lying on the western margins of the Intermountain Region have floras more strongly linked to the Sierra Nevada. Loope (1969) reported 189 alpine species from the Ruby Mountains in northeastern Nevada, with this flora showing a strong affinity to alpine floras of the Rocky Mountains. The isolated San Fran- cisco Mountain in Arizona with only 5.2 km? of alpine habitat has 80 species and likewise shows — strong floristic relationships to the Rocky Moun- tains despite its separation (Schaak 1983). In > contrast, the Sweetwater Mountains lying 33 km > west of the Sierra Nevada supports a flora of 173 species in 16 km? of alpine habitat, with 94% of | this flora common to the Sierra Nevada (Bell | Hunter and Johnson 1983). The small alpine zone | on Mt. Grant to the north of the Sweetwater , Mountains in western Nevada is just 2.6 km? in > area supports a flora of 70 species dominated by Sierra Nevadan elements (Bell and Johnson | 1980). ignoring subspecific taxa), the alpine flora of the ' White Mountains exhibits a much stronger | biogeographic relationship to the Sierra Nevada than to the central Rocky Mountains. Thus, 87% | of the species in the alpine flora of the White Mountains are also found in the Sierra Nevada (Hickman 1993) compared with only 58% that occur in the ranges of the central Rocky Mountains (Scott 1995). These values are signif- icantly higher for both ranges than earlier 1 estimates made on incomplete data (Lloyd and Mitchell 1973). The alpine flora of the White Mountains, like the Sierra Nevada, lacks many of | the typical circumboreal alpine species commonly | present in other alpine regions of North America | (Billings 2000). ACKNOWLEDGMENTS We greatly acknowledge the financial and logistical | support offered to this project by the White Mountains Research Station. In particular we thank Catherine Kleier, Frank Powell, Dave Trydahl, and Mike: Morrison for the assistance with various parts of this. research. Two anonymous reviewers provided invalu-_ Comparing checklists at the species level (i.e., | ! | 2008] able data in their comments on an earlier draft of this manuscript. LITERATURE CITED BELL, K. L. AND R. E. JOHNSON. 1980. Alpine flora of the Wassuk Range, Mineral County, Nevada. Madrono 27:25-—35. BELL HUNTER, K. L. AND R. E. JOHNSON. 1983. Alpine flora of the Sweetwater Mountains, Mono County, Nevada. Madrono 30:89—105. BILLINGS, W. D. 1974. Adaptations and origins of alpine plants. Arctic and Alpine Research 6:129— 142. 1978. Alpine phytogeography across the Great Basin. Great Basin Naturalist Memoirs 2:105—-117. . 2000. Alpine vegetation. Pp. 536-572 in M. G. Barbour and W. D. Billings (eds.), North American terrestrial vegetation. 2nd ed. Cambridge Univer- sity Press, Cambridge, United Kingdom. BOWMAN, W. D. AND T. R. SEASTEDT. 2001. Structure and function of an alpine ecosystem: Niwot Ridge, Colorado. Oxford University Press, Oxford, Unit- ed Kingdom. CHABOT, B. F. AND W. D. BILLINGS. 1972. Origins and ecology of the Sierran alpine flora and vegetation. Ecological Monographs 42:163—199. CHAPIN, F. S. and C. KORNER (eds). 1995. Arctic and alpine biodiversity: patterns, causes and ecosystem consequences. Springer Verlag, Berlin, Germany. CONSTANCE-SHULL, H. A. AND J. O. SAWYER. 2000. Arabis pinzlae Rollins (Brassicaceae). Madrono 47:209. ERNST, W. G., C. M. VAN DE VEN, AND R. J. P. LYON. 2003. Relationships among vegetation, geology, and climatic zones in the central White- Inyo Range, eastern California. Bulletin of the Geological Society of America 115:1583— 1597. HICKMAN, J. (ed). 1993. The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. JACKSON, J. L. 1985. Floristic analysis of the distribu- tion of ephemeral plants in treeline areas of the western USA. Arctic and Alpine Research 17:251— 260. AND L. C. BLISS. 1982. Distribution of ephemeral herbaceous plants near treeline in the Sierra Nevada, California, USA. Arctic and Alpine Research 14:33—44. JENNINGS, S. A. AND D. L. ELLIOT-FISK. 1991. Late | Pleistocene and Holocene changes in plant com- munity composition in the White Mountain region. Pp. 1-17 in C. A. Hall, V. Doyle-Jones, and B. Widawski (eds.), Natural history of eastern Cali- fornia and high-altitude research. White Mountains Research Station, Symposium 3. University of California, Los Angeles, CA. AND . 1993. Packrat midden evidence of late Quaternary vegetation change in the White Mountains, California-Nevada. Quaternary Re- search 39:214-221. KORNER, C. 1999. Alpine plant life: functional plant ecology of high mountain ecosystems. Springer Verlag, Berlin, Germany. RUNDEL ET AL.: ALPINE FLORA OF THE WHITE MOUNTAINS 211 LLoybD, R. M. AND R. S. MITCHELL. 1973. A flora of the White Mountains of California. University of California Press, Berkeley, CA. LoopeE, L. L. 1969. Subalpine and alpine vegetation of northeastern Nevada. Ph.D. dissertation. Duke University, Durham, NC. MAJor, J. AND D. W. TAYLOR. 1977. Alpine. Pp. 601— 675 in M. G. Barbour and J. Major (eds.), Terrestrial vegetation of California. Wiley, New York, NY. MARCHAND, D. E. 1973. Edaphic control of plant distributions in the White Mountains of eastern California. Ecology 54:233—250. MITCHELL, R. S., V. C. LAMARCHE, AND R. M. LLoyb. 1966. Alpine vegetation and active frost features of Pellsier Flats, White Mountains, Cali- fornia. American Midland Naturalist 75:516—525. Mooney, H. A. AND W. D. BILLINGS. 1960. The annual carbohydrate cycle of alpine plants as related to growth. American Journal of Botany 47:594-598. , G. ST. ANDRE, AND R. D. WRIGHT. 1962. Alpine and subalpine vegetation patterns in the White Mountains of California. American Midland Naturalist 68:257—273. MOREFIELD, J. D. 1988. Floristic habitats of the White Mountains, California and Nevada: a local ap- proach to plant communities. Pp. 1-18 in C. A. Hall and V. Doyle-Jones (eds.), Plant biology of eastern California. Natural history of the White- Inyo range, Symposium Volume 2, White Moun- tains Research Station. University of California, Los Angeles, CA. 1992. Spatial and ecologic segregation of phytogeographic elements in the White Mountains of California and Nevada. Journal of Biogeogra- phy 19:33—SO0. , D. W. TAYLOR, AND M. N. DEDECKER. 1988. Vascular flora of the White Mountains of Califor- nia and Nevada: an updated synonymized working checklist. Pp. 310-364 in C. A. Hall and V. Doyle- Jones (eds.), Plant biology of eastern California. Natural history of the White-Inyo range, Sympo- sium Volume 2, White Mountains Research Station. University of California, Los Angeles, CA. NELSON, C. A., C. A. HALL, AND W. G. ERNST. 1991. Geologic history of the White-Inyo Range. Pp. 42— 74 in C. A. Hall (ed.), Natural history of the White- Inyo range. University of California Press, Berke- ley, CA. PACE, N., D. W. KIEPERT, AND E. M. NISSEN. 1974. Climatological data summary for the Crooked Creek Laboratory, 1949-1973, and the Barcroft Laboratory, 1953-1973. White Mountain Research Station Special Publication, Bishop, CA. POLLACK, O. 1991. Morphology and dynamics in alpine populations of Jvesia lycopodioides subsp. scandularis from the White Mountains of Califor- nia. Pp. 97-116 in C. A. Hall, V. Doyle-Jones, and B. Widawski (eds.), Natural history of eastern California and high-altitude research. White Mountains Research Station, Symposium 3. Uni- versity of California, Los Angeles, CA. POWELL, D. R. AND H. E. KLIEFORTH. 1991. Weather and climate. Pp. 3—26 in C. A. Hall (ed.), Natural history of the White-Inyo Range. University of California Press, Berkeley CA. 2A PRICE, R. A. AND R. C. ROLLINS. 1988. Draba (Cruciferae) in the White Mountains and neighbor- ing ranges. Pp. 52-58 in C. A. Hall and V. Doyle- Jones (eds.), Plant biology of eastern California. Natural history of the White-Inyo range, Sympo- sium Volume 2, White Mountains Research Sta- tion. University of California, Los Angeles, CA. PRITCHETT, D. W. AND R. PATTERSON. 1998. Mor- phological variation in California alpine Polemoni- um species. Madrono 45:200—209. RAUNKIAER, C. 1934. The life forms of plants and statistical plant geography..Clarendon Press, Ox- ford, United Kingdom. ROLLINS, R. C. AND R. A. PRICE. 1988. High-elevation Draba (Cruciferae) of the White Mountains of California and Nevada, USA. Aliso 12:17—28. RUNDEL, P. W., A. C. GIBSON, AND M. R. SHARIFI. 2005. Plant functional groups in alpine fellfield habitats of the White Mountains, California. Arctic, Antarctic and Alpine Research 37:358—365. SAGE, R. F. AND T. L. SAGE. 2002. Microsite characteristics of Muhlenbergia richardsonis (Trin.) Rydb., an alpine C4, grass from the White Mountains, California. Oecologia 132:501—508. SCHAAK, C. G. 1983. The alpine vascular flora of Arizona. Madrono 30:79-88. ScoTT, R. W. 1995. The alpine flora of the Rocky Mountains: Volume 1. The Middle Rockies. University of Utah Press, Salt Lake City, UT. MADRONO [Vol. 55 SORENG, R. J. 1991. Notes on new infraspecific taxa and hybrids in North American Poa (Poaceae). Phytologia 71:390-413. SPIRA, T. P. 1987. Alpine annual plant species in the flora of the White Mountains of eastern California. Madronio 34:314—-324. 1991. Population ecology of Gymnosteris parvula (Polemoniaceae): a leafless alpine annual plant in the White Mountains, California. Pp. 117— 131 inC. A. Hall, V. Doyle-Jones, and B. Widawski (eds.), Natural history of eastern California and high-altitude research. White Mountains Research Station, Symposium 3. University of California, Los Angeles, CA. AND O. D. POLLACK. 1986. Compara- tive reproductive biology of alpine biennial and perennial gentians, Gentiana (Gentianaceae) in California. American Journal of Botany 73:39-47. AND L. K. WAGNER. 1988. Weedy annuals in the alpine flora of the White Mountains, Califor- nia. Pp. 92-98 in C. A. Hall and V. Doyle-Jones (eds.), Plant biology of eastern California. Natural history of the White-Inyo range, Symposium Volume 2, White Mountains Research Station. University of California, Los Angeles, CA. TRANQUILLINI, W. 1979. Physiological ecology of the alpine timberline: Tree existence at high elevations with special reference to the European Alps. Springer Verlag, Berlin, Germany. 2008] RUNDEL ET AL.: ALPINE FLORA OF THE WHITE MOUNTAINS 213 APPENDIX 1. THE NATIVE VASCULAR PLANT FLORA OF ALPINE AREAS OF THE WHITE MOUNTAINS ABOVE 3500 M ELEV. Species are listed alphabetically by family and species, with family names abbreviated by their first four letters. Species follow Hickman et al. (1993). Life forms are shrub (phanerophyte, S), chamaephyte (subshrub, CH), cushion or mat (C/M), broad-leaved herbaceous perennial (HP), perennial graminoid (G), and annual (A). Elevation zones of lowest occurrence are cold desert (CD), montane (MON), subalpine (SUBA), alpine (ALP), and high alpine (HALP). Habitat categories are aquatic (A), areas with saturated soils (W), wet meadow (M), fellfield (FF), talus slope (T), open slope (OS), and rocky slope (R). Biogeographic distribution categories are widespread (W), western cordilleran (CORD), Sierra-Cascade (S-C), intermountain (INT), and endemic (END). See text for discussion of these categories. Species Family Life form Elevation zone Habitat Biogeography Pteridophyta Selaginella watsonii SELG HP SUBA FF INT Cystopteris fragilis DRYO HP SUBA FF INT Pellaea breweri PTER HP MON FF INT Angiospermae—Dicotyledonae Cymopterus cinerarius APIA HP SUBA OS INT Agoseris glauca var. laciniata ASTE HP SUBA FF WIDE Antennaria media ASTE C/M ALP M CORD Antennaria rosea ASTE HP MON FF WIDE Antennaria umbrinella ASTE C/M SUBA M WIDE Artemisia dracunculus ASTE LP MON FF INT Artemisia ludoviciana subsp. incompta ASTE HP CD FF INT Artemisia michauxiana ASTE HP SUBA M CORD Artemisia rothrockii ASTE CH SUBA M S-C Chaenactis alpigena ASTE HP ALP M S-C Chrysothamnus viscidiflorus subsp. ASTE CH MON a INT viscidiflorus Crepis nana ASTE HP ALP M CORD Ericameria discoidea ASTE CH SUBA M INT Ericameria suffruticosa ASTE CH SUBA FF INT Erigeron clokeyi ASTE HP MON OS INT Erigeron compositus ASTE C/M MON FF WIDE Erigeron pygmaeus ASTE C/M ALP FF S-C Erigeron vagus ASTE HP HALP FF INT Hulsea algida ASTE HP ALP FF INT Machaeranthera canescens var. ASTE HP MON FF CORD canescens Pyrrocoma apargioides ASTE HP SUBA M S-C Raillardella argentea ASTE HP ALP FF S-C Senecio integerrimus var. major ASTE HP MON FF CORD Senecio pattersonensis ASTE HP ALP PF INT Senecio scorzonella ASTE HP SUBA M S-C Senecio werneriaefolius ASTE HP ALP FF S-C Solidago multiradiata ASTE HP SUBA FF CORD Stenotus acaulis ASTE C/M MON R INT Townsendia condensata ASTE HP ALP OS CORD Townsendia leptotes ASTE HP ALP FF CORD Trimorpha lonchophylla ASTE HP MON M WIDE Cryptantha cinerea var. abortiva BORA HP MON OS INT Cryptantha glomeriflora BORA A SUBA FF S-C Cryptantha humilis BORA HP SUBA OS INT Cryptantha nubigena BORA HP SUBA FF INT Anelsonia eurycarpa BRAS HP HALP FE S-C Arabis X divaricarpa BRAS HP ALP OS CORD Arabis holboellii var. pendulocarpa BRAS HP ALP FF INT Arabis inyoensis BRAS HP MON OS INT Arabis lemmonii BRAS C/M SUBA OS INT Arabis lyalii var. nubigena BRAS HP SUBA OS INT Arabis pinzlae BRAS HP ALP FF S-C Arabis platysperma var. platysperma BRAS HP SUBA PF S-C Descurainia incana BRAS HP MON M CORD Draba albertina BRAS HP SUBA PE CORD Draba breweri BRAS HP ALP FF S-C Draba californica BRAS HP ALP M END Draba densifolia BRAS C/M SUBA FF INT 214 MADRONO [Vol. 55 APPENDIX 1. CONTINUED Species Family Life form Elevation zone Habitat Biogeography Draba monoensis BRAS HP ALP FF END Draba oligosperma var. oligosperma BRAS C/M SUBA OS CORD Draba subumbellata BRAS HP ALP a S-C Halimolobos virgata BRAS HP SUBA W CORD Lesquerella kingii subsp. kingii BRAS HP MON R INT Callitriche verna CALL A ALP A WIDE Arenaria kingii var. glabrescens CARY C/M SUBA OS S-C Cerastium beeringianum var. capillare CARY HP HALP M WIDE Minuartia nuttallii subsp. gracilis CARY HP ALP FF CORD Minuartia rubella CARY HP ALP FF CORD Minuartia stricta CARY HP ALP M CD Sagina saginoides CARY HP MON W WIDE Silene bernardina CARY HP MON OS CORD Silene sargentii CARY HP ALP FF S-C Stellaria umbellata CARY HP ALP W WIDE Chenopodium atrovirens CHEN A CD M CORD Chenopodium leptophyllum CHEN A MON FF CORD Monolepis nuttalliana CHEN A SUBA BE CORD Sedum roseum subsp. integrifolium CRAS HP ALP FF WIDE Astragalus kentrophyta var. tegetarius FABA C/M SUBA OS S-C Lupinus lepidus var. utahensis FABA HP MON M CD Oxytropis borealis var. viscida FABA HP ALP FF CORD Oxytropis parryi FABA C/M SUBA OS INT Trifolium andersonii var. beatleyae FABA C/M SUBA OS END Trifolium monanthum var. monanthum FABA HP MON M S-C Gentiana newberryi var. tiogana GENT HP ALP M S-C Gentiana prostrata GENT HP SUBA M WIDE Gentianella tenella subsp. tenella GENT HP MON W WIDE Ribes cereum GROS P MON OS INT Nama densum HYDR A MON FP INT Phacelia hastata subsp. compacta HYDR HP MON OS S-C Linum lewisii var. alpicola LINA HP ALP OS INT Gayophytum racemosum ONAG A SUBA FF CORD Gymnosteris parvula POLE A SUBA FF CORD Ipomopsis congesta subsp. montana POLE HP MON FF S-C Leptodactylon pungens POLE CH SUBA OS CORD Linanthus nuttallii subsp. pubescens POLE CH MON OS CORD Phlox condensata POLE C/M MON OS CORD Phlox pulvinata POLE CH ALP OS CORD Polemonium chartaceum POLE HP HALP OS END Eriogonum gracilipes POLY C/M SUBA R INT Eriogonum lobbii var. lobbii POLY HP ALP FF S-C Eriogonum ovalifolium var. nivale POLY C/M SUBA OS INT Eriogonum rosense POLY C/M MON R INT Eriogonum spergulinum vat. POLY A MON FF INT reddingianum Oxyria digyna POLY HP ALP M WIDE Rumex paucifolius POLY HP SUBA OS S-C Calyptridium roseum PORT A MON FF INT Calyptridium umbellatum PORT HP SUBA PE CORD Lewisia pygmaea PORT HP, ALP FF CORD Montia chamissoi PORT HP MON WwW S-C Androsace septentrionalis subsp. PRIM HP ALP OS CORD subumbellata Dodecatheon redolens PRIM FIP MON W INT Ranunculus alismifolius var. alismellus RANU EP SUBA WwW S-C Ranunculus eschscholtzii var. oxynotus RANU HP ALP a ly S-C Ranunculus glaberrimus var. ellipticus RANU HP SUBA M CORD Thalictrum alpinum RANU HP MON M WIDE Ivesia gordonii ROSA HP ALP OS INT Ivesia lycopodioides subsp. scandularis ROSA HP ALP M INT Ivesia shockleyi var. shockleyi ROSA C/M ALP OS S-C Potentilla fruticosa ROSA P MON M WIDE 2008] RUNDEL ET AL.: ALPINE FLORA OF THE WHITE MOUNTAINS 215 APPENDIX |. CONTINUED Species Family Life form Elevation zone Habitat Biogeography Potentilla glandulosa subsp. ROSA HP ALP OS CORD pseudorupestris Potentilla morefieldii ROSA HP SUBA OS END Potentilla pensylvanica ROSA HP SUBA PE WIDE Potentilla pseudosericea ROSA FP ALP FF CORD Sibbaldia procumbens ROSA HP ALP HF WIDE Galium multiflorum RUBI HP SUBA T INT Salix orestera SALI P SUBA M S-C Heuchera duranii SAXI HP SUBA an CORD Castilleja nana SCRO HP SUBA OS INT Mimulus mephiticus SCRO A SUBA ale S-C Mimulus primuloides subsp. primuloides SCRO HP MON M S-C Mimulus suksdorfii SCRO A SUBA FF INT Pedicularis attollens SCRO HP SUBA M S-C Penstemon heterodoxus var. heterodoxus SCRO C/M SUBA FF S-C Angiospermae—Monocotyledonae Carex albonigra CYPE G ALP M CORD Carex breweri var. breweri CYPE G ALP W S-C Carex capitata CYPE G ALP W WIDE Carex douglasii CYPE G CD M CORD Carex eleocharis CYPE G SUBA FF CORD Carex filifolia var. erostrata CYPE G SUBA M CORD Carex haydeniana CYPE G SUBA M CORD Carex helleri CYPE G ALP FF S-C Carex heteroneura var. heteroneura CYPE G SUBA WwW INT Carex microptera CYPE G MON M CORD Carex phaeocephala CYPE G SUBA FF CORD Carex straminiformis CYPE G SUBA M S-C Carex subnigricans CYPE G ALP M INT Carex vernacula CYPE G ALP M CORD Eleocharis pauciflora CYPE G MON W WIDE Juncus balticus JUNC G SUBA M CORD Juncus bryoides JUNC G-A ALP PF CORD Juncus nevadensis JUNC G MON M CORD Juncus parryi JUNC G ALP Pie CORD Luzula spicata JUNC G ALP FF WIDE Achnatherum pinetorum POAC G MON OS INT Calamagrostis purpurascens POAC G MON Pr WIDE Deschampsia cespitosa subsp. cespitosa POAC G ALP M WIDE Elymus elymoides POAC G MON OS CORD Elymus scribneri POAC G HALP T CORD Elymus trachycaulus subsp. subsecundus POAC G MON FF WIDE Festuca brachyphylla subsp. breviculmis POAC G SUBA FF WIDE Festuca minutiflora POAC G SUBA FF CORD Koeleria macrantha POAC G SUBA OS WIDE Muhlenbergia richardsonis POAC G MON la CORD Poa abbreviata subsp. pattersonii and POAC G HALP FF CORD subsp. marshii Poa cusickii subsp. epilis POAC G SUBA FF CORD Poa glauca subsp. rupicola POAC G ALP FF CORD Poa keckii POAC G ALP al S-C Poa lettermanii POAC G HALP FF CORD Poa secunda subsp. secunda POAC G CD FF CORD Trisetum spicatum POAC G SUBA FF WIDE MADRONO, Vol. 55, No. 3, pp. 216-222, 2008 THE ROLES OF FLOODS AND BULLDOZERS IN THE BREAK-UP AND DISPERSAL OF ARUNDO DONAX (GIANT REED) JOHN M. BOLAND Boland Ecological Services, 3504 Louisiana Street, San Diego, CA 92104 JohnBoland@sbcglobal.net ABSTRACT Arundo donax L. (Poaceae) is an invasive grass that severely degrades riparian habitats. It grows in many-stemmed clumps and, in California, spreads vegetatively only. Currently, A donax is thought to invade new areas through fragments broken from established clumps during flood events. But the role of flooding in generating fragments is based on anecdotal evidence only and has not been adequately studied. I examined A. donax clump break-up and reproduction via fragmentation in the Tijuana River Valley, California. I found that: (1) the majority of the new recruits from fragments grew from rhizome fragments (85% of 54) rather than stem fragments; (2) during the record rainfall of 2004— 2005, flood waters damaged the rootstock of only a small proportion of clumps in the flood zone (7%; n = 46 clumps), and relatively few recruits from fragments subsequently became established in the valley at large (0.048 recruits 100 m-*); and (3) during emergency channel maintenance along one tributary, bulldozers severely damaged the rootstock of all clumps growing on the channel bank (n = 3 clumps), and many recruits from fragments subsequently became established downstream of the bulldozer activity (2.92 recruits 100 m-’; 61 times the number in the valley at large). These results indicate that, in the Tijuana River Valley, flood events only rarely break up A. donax rootstock and wash rhizomes downstream, and bulldozers play an important, and overlooked, role in the break-up and dispersal of A. donax. To reduce the spread of A. donax via rhizome fragments, regulatory agencies should require appropriate management practices when bulldozers are used in the presence of A. donax, and land managers should not use bulldozers when attempting to eradicate A. donax. Key Words: Arundo donax, dispersal, flooding, giant reed, invasive species, non-indigenous species, plant fragments, rhizomes. Arundo donax L. (Poaceae), giant reed, is a - large, invasive grass from Eurasia that severely degrades riparian habitats in many temperate areas of the world (Dudley 2000; Global Invasive Species Database 2005). In California, a tremen- dous amount of effort has gone into controlling its spread (Katagi et al. 2002), but it is still common in most watersheds and even the dominant plant along many reaches (Neill and Giessow 2004). To determine the best strategy for control of any invasive plant, one needs detailed knowledge of its means of spread (Radosevich et al. 1997). Arundo donax spreads vegetatively in California, as no seedlings or viable seeds have been found (Perdue 1958; Else 1996; Johnson et al. 2006). The vegetative expansion of established clumps via lateral growth of rhizomes (Decruye- naere and Holt 2005; Boland 2006) and via layering of stems (Boland 2006) has received some recent attention, but there is little informa- tion on the formation of new clumps from vegetative fragments (cf. Else 1996; Boland 2006). The conventional wisdom regarding the spread of A. donax via fragmentation is simple: during flood events, fragments are broken from estab- lished clumps and dispersed downstream where they subsequently sprout and grow into new clumps. Bell (1993), for example, states that “(f)lood events break up clumps of Arundo and spread the pieces downstream. Fragmented stem nodes and rhizomes can take root and establish as new plant clones.’ This idea has been repeated many times and has become entrenched in the 2. donax literature (e.g., Else 1996; Bell 1997; DiTomaso 1998; Kelly 1999; McWilliams 2004). But the central premise — that flood events cause the break-up of clumps — is based on anecdotal evidence only and has not been adequately examined. No other method of fragmentation has yet been proposed. But there are anthropogenic forces at work in watersheds that can influence A. donax break-up and dispersal. In the Califor-_ nia wildlands where A. donax has become abundant, heavy equipment, such as bulldozers, loaders, excavators and tractors, are frequently | used to maintain river channels, to construct | flood-control berms beside agricultural fields, to’ maintain dirt roads, to dig quarries, and even to. control A. donax (personal observations). I have | observed bulldozers and other heavy equipment. undercut, break up and move clumps of A. donax, and I suggest that they play an important, | and heretofore overlooked, role in the break-up) and dispersal of A. donax. The purpose of this study was to examine. vegetative reproduction via fragmentation in A. | donax. First, I asked the question: What plant. 2008] parts (stems or rhizomes) most commonly become successful recruits? Then I conducted observations and surveys that focused on the roles of flood flows and bulldozers in the break- up of established clumps, the dispersal of fragments and the recruitment of new clumps. The results of this study provide a new view of the mechanism of break-up and dispersal of A. donax that elevates the role of mechanical disturbance and suggests that new management practices are urgently needed if A. donax is to be successfully controlled. STUDY SITE The study was conducted in the Tijuana River Valley, California, which is a coastal floodplain at the end of a large (448,000 ha) watershed that is partly in Mexico. The valley spans 1457 ha at approximately sea-level and includes a county regional park, a state park, and a national wildlife refuge (Concur 2000; Boland 2006). The valley contains prime riparian and salt marsh habitats that have been invaded by invasive, non- native species, including A. donax, salt cedar (Tamarix spp.), and castor bean (Ricinus commu- nis). Arundo donax is common in the valley (Southwest Wetlands Interpretive Association 2002) and in much of the rest of the watershed upstream (Woch 2005). This study was conducted between 2004 and 2006. The 2004-2005 rainfall year was the third wettest in San Diego history (57.2 cm of precip- itation; Western Regional Climate Center 2006), and the unusually heavy rainfall produced many days of high flows in the Tijuana River Valley— 172 days when the average daily flow rates were >1m?* sec ' (Boland 2006; International Bound- ary & Water Commission 2006). During January 2005, emergency channel maintenance was con- ducted in Smuggler’s Gulch, a narrow, sandy tributary leading to the main river flood plain. High flows threatened to breach the channel banks and flood neighboring farms, so bulldozers were used along an 800 m section to deepen the channel by working the bottom sediment into the flowing stream and to raise the banks by depositing sediment from the channel onto the banks. No other major bulldozing occurred in the flood zone in the valley between 2003 and 2006. METHODS Plant Fragments that Become Successful Recruits To determine which A. donax plant parts most often become successful recruits, extensive ‘searches for new recruits from plant fragments were conducted throughout the valley during the spring and summer of 2005. These searches included: (a) valley-wide belt transects that BOLAND: BREAK-UP AND DISPERSAL OF ARUNDO DONAX PANG covered 0.84 ha — described in more detail below and in Boland (2006); (b) localized surveys that covered 0.12 ha — also described below; and (c) additional searches that covered approximately 0.3 ha in selected areas where new recruits from fragments were expected to occur, e.g., areas containing debris piles. The total area surveyed was therefore >1.25 ha. When a sprouting fragment was found, the sprout was dug up and the fragment was identified as either a piece of rhizome or stem (1.e., culm). The Role of Flood Flows Observations of A. donax Clumps During Flooding Events. To determine whether flood flows broke up clumps of A. donax, I watched the behavior of A. donax clumps in the flood waters throughout the severe winter of 2004-2005. During more than 30 visits to the valley, I observed clumps from bridges while flooding was occurring. I looked for evidence of damage to clumps due to flood flows, e.g., breaking of rhizomes, undermining of rootstock, and wash- ing-away of whole clumps during the flooding. Estimation of Flood Damage to A. donax Rootstock. To assess the damage to clumps of A. donax caused by the exceptional winter of flood events, I examined clumps before and after the 2004—2005 floods. I had photographed 63 A. donax clumps in all parts of the valley, during summer 2004, as part of a separate study. During summer 2005, I reexamined these clumps. At each clump, I compared the clump with the photo taken before the flooding and estimated the amount of flood damage to the rootstock as either none, slight (area of <1 m?* missing), moderate (l—3 m* missing) or severe (>3 m° missing). During the visit, the clump was also determined to be inside or outside the 2004-2005 flood zone according to its position relative to flood debris. The clumps were located through- out the valley, both inside (46 clumps) and outside (17 clumps) the flood zone. Clumps outside the flood zone were not expected to be damaged by flood flows but their results are presented for comparative purposes. Abundance of Sprouting A. donax Fragments in the Valley After the Flooding. To determine the number of new recruits from fragments after the 2004—2005 flooding, I surveyed the valley during June 2005 using the same procedures as Else (1996). Eight transects across the river valley were chosen in a stratified-random manner (Boland 2006). The transects were 2 m-wide belts that ran perpendicular to the river channel and extended from the southern edge to the northern edge of the 2005 flood zone. The boundary of the flood zone was determined by the presence of debris indicating the highest flood level of the 2004— 218 2005 flood season. Transect lengths varied depending on the width of the flood zone (range = 97-865 m; n = 8). The total area surveyed was 0.84 ha. Within each transect, the number of new A. donax recruits from fragments was counted. Each new recruit was excavated and determined to be from a fragment if it was not attached to a parent plant (as opposed to a new recruit formed via stem layering, which would have a visible connection to the parent plant; Boland 2006). Recruits from fragments were further classified as being from a fragmented stem or rhizome. The average density of new recruits from fragments in the valley as a whole, 1.e., resulting from flood events only, was estimated from this survey and reported as the number per 100 m’. The Role of Bulldozers Observations of A. donax Clumps During Bulldozer Activity. The January 2005 emergency channel maintenance at Smuggler’s Gulch pro- vided an opportunity to observe the effects of bulldozers on A. donax. | watched for cutting of rhizomes, undermining of rootstock, pushing of live material into the channel flow, and deposit- ing of excavated A. donax onto channel banks. Estimation of Bulldozer Damage to A. donax Rootstock. To assess the damage caused by bulldozers to rootstocks of A. donax, | examined the three large clumps growing on the banks of Smuggler’s Gulch in January 2005, immediately after the maintenance was completed. I did a follow-up survey in June 2005 to further evaluate damages due to the maintenance activity. For each clump, I estimated the amount of bulldozer damage to the rootstock as either none, slight (area of <1 m° missing), moderate (1—3 m° missing) or severe (>3 m° missing) based on photos and previous knowledge of clump size from prior site visits. Abundance of Sprouting A. donax Fragments Downstream of a Bulldozed Channel. Immediately after the dredging at Smuggler’s Gulch in January 2005, I saw vegetation debris scattered over a cleared staging area located in the floodplain approximately 150 m downstream of Smuggler’s Gulch. To determine the number of new recruits from fragments occurring down- stream of the bulldozer activity, I surveyed this staging area for sprouting fragments in June 2005. I divided the staging area into two survey sites, each 100 m X 50 m and, within each site, searched for new A. donax recruits in six randomly-chosen, 2m X 50m belt transects (sensu Else 1996; Boland 2006). The total area surveyed was (0.12 ha. Within each transect, the number of new recruits from fragments was counted. Each new recruit was excavated and identified as a fragment of either stem or rhizome. MADRONO [Vol. 55 The average density of new recruits below Smuggler’s Gulch, 1.e., due to bulldozer activity, was estimated from this survey, reported as the number per 100 m’*, and compared to the density of new recruits from fragments in the river valley as a whole using the Chi-square Test with Yates’ correction. RESULTS Plant Fragments that Become Successful Recruits In extensive searches of the Tijuana River Valley, covering >1.25 ha, I found a total of 61 new A. donax recruits sprouting from vegetative fragments. For 54 of these, the fragment type could be unmistakably identified as rhizome or stem, but seven were too deeply buried to be identified. Of the 54 identifiable recruits, 46 (85%) were from rhizome fragments, and 8 (15%) were from stem fragments. These results indicate that, while fragments of both rhizomes and stems do sprout, rhizome fragments are the more likely to successfully sprout and give rise to new clumps. Therefore, in examining the roles of floods and bulldozers on the break-up of A. donax clumps, I have focused on the fragmenta- tion of rhizomes and not of stems. The Role of Flood Flows Observations of A. donax Clumps During Flooding Events. During the severe floods of 2004-2005, there was no evidence of A. donax clumps being broken-up or otherwise damaged by flood waters. During low flows, water flowed around the clumps without disturbing the stems or rootstock. During high flows, water flowed through and over the clumps, dead stems were swept away, and live stems swayed violently in the currents but remained attached to the rootstock. All rootstocks appeared to remain intact. Estimation of Flood Damage to A. donax = Rootstock. Examination of A. donax clumps after the floods of 2004-2005, revealed that clumps_ suffered relatively little damage to their rootstock | due to flooding. Of the 46 clumps growing inside , the flood zone, 43 (93%) showed no signs of, damage to the rootstock (Table 1). Only three clumps (7%) had damaged rootstocks where. flood flows had partially undermined and swept, away a portion of their rootstocks. In each case, the extent of damage was slight to moderate, with less than 3 m? of the rootstock lost. As expected, the 17 clumps growing outside the flood zone) showed no sign of damage to their rootstocks, (Table 1). | Abundance of Sprouting A. donax Fragments in the Valley After the Flooding. Extensive surveys | { { 2008] TABLE 1. BOLAND: BREAK-UP AND DISPERSAL OF ARUNDO DONAX 219 THE DEGREE OF DAMAGE TO ARUNDO DONAX CLUMPS PHOTOGRAPHED IN SUMMER 2004 AND REEXAMINED IN SUMMER 2005. Damage is described as slight (rootstock area of <1 m’ missing), moderate (1— 3 m? missing) or severe (>3 m? missing). TJRV = Tijuana River Valley. Degree of damage to Clumps rootstocks o, Site (n) Source of Damage None Slight Moderate Severe Damaged TJR V—Inside Flood Zone 46 flooding 43 2 ] 0 71% TJR V—Outside Flood Zone 17 none 17 0 0 0 0% Smuggler’s Gulch Bank 3 bulldozers and flooding 0 0 l 2 100% after the floods of 2004—2005 showed that new A. donax recruits growing from fragments were rare in the Tijuana River Valley. Only four recruits from fragments were encountered in the valley- wide surveys, i.e., 0.048 recruits per 100 m? (Table 2). The Role of Bulldozers Observations of A. donax Clumps During Bulldozer Activity. During the January 2005 channel maintenance in Smuggler’s Gulch, bull- dozers undermined A. donax clumps and easily cut their rootstocks. Front-loaders piled the dredge spoils on the banks nearby, and this spoil contained live A. donax pieces, which later developed into 15 new clumps, a five-fold increase in clumps on the bank. In addition, the bulldozers pushed living A. donax material into the channel where it was washed downstream. Estimation of Bulldozer Damage to A. donax Rootstock. Examination of the A. donax clumps on the banks of Smuggler’s Gulch after the channel maintenance activities confirmed that bulldozers had substantially reduced the root- stocks of all three large clumps and caused moderate to severe damage to each (Table 1). Abundance of Sprouting A. donax Fragments Downstream of a Bulldozed Channel. New recruits growing from fragments were abundant in the flood zone immediately downstream from the bulldozer work in Smuggler’s Gulch. A total of 35 recruits were present in the surveys, at a density of 2.92 per 100 m? (Table 2). This density was 61-times the density of new recruits in the } | _ TABLE 2. -entire valley. Therefore, the density of recruits downstream of the bulldozer activity was signif- icantly greater than the density of recruits in areas not influenced by the bulldozers (Chi- square Test with Yates’ correction; P < .005). DISCUSSION The Importance of Rhizome Fragments in the Dispersal of A. donax To understand dispersal in A. donax, one needs to know which plant part is responsible for most of the new recruits. In the Tijuana River Valley, the majority of the new fragment recruits (85%) were growing from rhizome fragments, and many fewer were growing from stem and _ branch fragments. This result is not surprising, as A. donax rhizomes have been the most viable fragment under both lab (Decruyenaere and Holt 2001) and field conditions (Else 1996), and farmers use rhizomes when propagating A. donax (Hoshovsky 2003). Arundo donax rhizomes are thick and solid, and designed for carbohydrate storage rather than for rapid expansion of the clump (Boland 2006). They provide the plant with a site for resource storage protected from fire, frost, grazers and desiccation. When dis- persed, they provide abundant resources for the successful establishment of a new clump. As for the other plant parts, main stems and branches are hollow and, although their fragments can sprout (Motamed and Wijte 1998), they are less likely to become successful recruits (Dudley 2000; this study). What these results show is that when studying the reproduction of A. donax by vegetative fragmentation one needs to focus primarily on rhizomes and the mechanisms that break live rhizomes from rootstocks. THE NUMBER AND DENSITY OF NEW ARUNDO DONAX RECRUITS FROM FRAGMENTS IN THE ENTIRE _ TIJUANA RIVER VALLEY AND IN THE FLOODPLAIN IMMEDIATELY DOWNSTREAM OF THE BULLDOZED CHANNEL, | SMUGGLER’S GULCH. Entire Tijuana River Valley Downstream of Smuggler’s Gulch Bulldozer-use upstream? Surveyed area (ha) Total no. of recruits from fragments _ Density of recruits from fragments (100 m °) Test of densities (chi-square) no yes 0.837 0.12 4 35 0.048 2.92 P < .005 N N i) The Role of Flood Flows Currently it is thought that flooding is the mechanism responsible for the break-up of A. donax (e.g., Bell 1993; McWilliams 2004). How- ever, when A. donax clumps were observed during an extremely wet year—when extensive fragmentation would be expected—fragmenta- tion and the production of new recruits were found to be rare. I found that only a small proportion of the clumps were undermined (7%), only a few rhizomes were removed by flooding (<3 m° at each clump) and relatively few recruits from fragments became established in the valley (0.048 recruits 100 m *). Furthermore, in years when flows are average or below average, one sees even fewer instances of fragmentation and fewer recruits (personal observations). Floods do not easily break off rhizomes because rootstocks and the soils they bind create effective barriers to water flows, and because rhizomes are not easily broken (personal observations). Hence, although the conventional wisdom gives the impression that A. donax clumps are frequently and easily fragmented by flood flows (Bell 1993), this is not the case. The Role of Bulldozers In contrast to flood events, bulldozers and other earthmoving equipment easily cut, under- mined, and moved large sections of A. donax rootstocks in the Tijuana River Valley. By doing so bulldozers influenced both the local and long- distance dispersal of A. donax. At Smugglers Gulch, bulldozers increased the number of clumps on the bank five-fold and increased the density of recruits downstream by 61-times. These results show that bulldozers can play a major role in the break-up, dispersal and propagation of A. donax. Bulldozers are a “‘disturbance” in the tradi- tional sense (e.g., Begon et al. 1996), in that they create gaps into which A. donax can invade. They also act as vectors that carry rhizomes relatively short distances, and act as dispersal facilitators that produce the propagules (rhizomes) and leave them to be dispersed over long distances via river flows. It is not unusual for mechanical equipment to facilitate the dispersal of invasive plants (e.g., USDA Forest Service 2001). Usually the equip- ment is the vector, carrying the plant from an infected area to an uninfected area. But, mechan- ical equipment can also be the agent that produces the dispersed material. When mowers are used to control any of several invasive waterweeds, e.g., leafy elodea (Egeria densa), parrot’s feather (Myriophyllum aquaticum), Eur- asian water-milfoil (Myriophyllum spicatum), they cut and release plant fragments that can drift into, and establish in, new habitats (Bossard MADRONO [Vole 55 2000; Godfrey 2000; Hoshovsky and Anderson 2000). Bulldozers and waterweed mowers there- fore play similar roles in the spread of their respective invasive plants. There have been some recent questions about the spread of A. donax in California. For instance, Johnson et al. (2006) state: ‘‘The invasion of California riparian areas by Arundo continues despite efforts to control its spread, and there remains some uncertainty as to how it is able to do so.” They determined that seed production was not the mechanism by which A. donax was invading. I suggest that the “continued inva- sions” of A. donax they describe are due to bulldozer activities in the watersheds. The find- ings in the Tijuana River Valley show that much of the recruitment of new A. donax clumps can be separated, in both space and time, from the bulldozer event that produced them. The dis- lodged rhizomes can be dispersed hundreds of meters, and the time between the bulldozer impact and the obvious growth of the new recruits can be up to ten months. Someone finding new recruits in a flood zone may not realize that the recruits came from a bulldozer disturbance many months earlier, and possibly many hundreds of meters upstream. This sepa- ration of cause and effect has probably contrib- uted to our slow appreciation of the role that bulldozers play in the spread of A. donax. Consequences for Management Bulldozers are used in A. donax areas to dredge channels, to raise channel banks, to cut dirt roads, to mine for sand and gravel, to cut and clear vegetation, etc. Now, with evidence that bulldozers promote the dispersal of A. donax, permitting agencies should insist on appropriate management practices for these kinds of activi- ties. The practices on-site could include spraying of A. donax clumps with herbicide before, during, and after earthmoving activities, and the instal- lation and maintenance of plant-debris catchers during the project. In addition, all soil and plant debris that is removed from the site should be treated appropriately to prevent the spread of A. donax. Another problem is that bulldozers are some- times used to eradicate A. donax (Bell 1997; Oakins 2001). Mechanical excavation would be. an acceptable option if the method were 100% efficient in removing A. donax rhizomes, but the method is not that efficient. Typically, at these A. donax-control sites, bulldozers excavate the rootstock and front-loaders load the material. into a tub-grinder on-site. The finely-ground | material produced by the tub-grinder is not able. to sprout (Boland unpublished data). But A. donax rootstocks are incompletely removed from the soil, rhizome pieces are dropped along the. | 2008] way to the tub-grinder, and other pieces are thrown out uncut by the tub-grinder (personal observations) and these rhizome pieces are capable of sprouting (Boland unpublished data). At one treatment site on the Santa Margarita River where some mechanical excavation and tub-grinding was conducted, Giessow and Gies- sow (1999) noted that “most of the Arundo resprouts that occurred resulted from small pieces of rhizome that broke off during the mechanical removal process.’ Therefore, even the well-inten- tioned use of heavy equipment as agents for A. donax-control can undermine an eradication effort by producing fragments that propagate A. donax on-site and downstream. Until safe meth- ods are developed, mechanical equipment should be limited to dealing with only the above-ground biomass of A. donax, and rhizomes should be left in place and treated chemically. It is time to recognize the threat posed by bulldozers and other earthmovers in the uninten- tional break-up and dispersal of A. donax, and to focus our efforts on preventing this method of spread. ACKNOWLEDGMENTS Partial support for this research came from South- west Wetlands Interpretive Association (SWIA) through a Proposition 13 grant from the California State Water Resources Control Board. I thank Carl Bell, Jeff Crooks, John Hunter, Chris Nordby, Bill Winans, Deborah Woodward and two anonymous reviewers for helpful comments on an early draft of this manuscript. LITERATURE CITED BEGON, M., J. L. HARPER, AND C. R. TOWNSEND. 1996. Ecology: Individuals, populations and com- munities. Blackwell Scientific Publications, Boston, MA. BELL, G. P. 1993. Biology and growth habits of giant reed (Arundo donax). Arundo donax Workshop Proceedings, Ontario, CA. November 19, 1993, pp. 1-6. Team Arundo del Norte. Available at: http://teamarundo.org/ecology_impacts/proc93/ proc93_index.html . 1997. Ecology and management of Arundo donax, and approaches to riparian habitat restora- tion in Southern California. Pp. 103-113 in J. H. Brock, M. Wade, P. Pysek, and D. Green (eds.), Plant invasions: studies from North America and Europe. Blackhuys Publishers, Leiden, The Nether- lands. BOLAND, J. M. 2006. The importance of layering in the rapid spread of Arundo donax (giant reed). Madrono 53:303-312. _ BOSSARD, C. 2000. Myriophyllum spicatum L. Pp. 254— 258 in C. Bossard, J. Randall, and M. Hoshovsky (eds.), Invasive plants of California’s wildlands. University of California Press, Berkeley, CA. CONCUR. 2000. Comprehensive management plan for Tiyuana River National Estuarine Research Re- serve and Tijuana Slough National Wildlife Ref- uge. Prepared for California Dept. of Parks and BOLAND: BREAK-UP AND DISPERSAL OF ARUNDO DONAX 221 Recreation, US Fish and Wildlife Service, and the National Oceanic and Atmospheric Administra- tion. DECRUYENAERE, J. AND J. HOLT. 2001. Seasonality of clonal propagation in giant reed (Arundo donax). Weed Science 49:760—767. AND . 2005. Ramet demography of a clonal invader, Arundo donax (Poaceae), in south- ern California. Plant and Soil 277:41—52. DITOMASO, J. M. 1998. Biology and ecology of giant reed. Pp. 1—5 in C. Bell (ed.), Arundo and saltcedar: the deadly duo. Proceedings of the Arundo and saltcedar workshop. University of of California Cooperative Extension Publication, Imperial County, CA. DUDLEY, T. L. 2000. Arundo donax L. Pp. 53-58 in C. Bossard, J. Randall, and M. Hoshovsky (eds.), Invasive plants of California’s wildlands. Universi- ty of California Press, Berkeley, CA. ELSE, J. A. 1996. Post-flood establishment of native woody species and an exotic, Arundo donax, in a southern California riparian system. M.S. Thesis. San Diego State University, San Diego, CA. GigEssow, J. A. AND J. H. GiEssow. 1999. 1999 Annual report: Arundo removal project, Santa Margarita River, CA. Prepared for Southwest Division of the Navy, Contract No. N68711-95-LT-C028. GLOBAL INVASIVE SPECIES DATABASE. 2005. 100 world’s worst alien species. Available at: http:// WwWw.iSSg.org GODFREY, K. 2000. Myriophyllum aquaticum. Pp. 249— 254 in C. Bossard, J. Randall, and M. Hoshovsky (eds.), Invasive plants of California’s wildlands. University of California Press, Berkeley, CA. HOSHOVSKY, M. 2003. Arundo donax: element steward- ship abstract. The Nature Conservancy, San Francisco, CA. AND L. ANDERSON. 2000. Egeria densa. Pp. 161— 164 in C. Bossard, J. Randall, and M. Hoshovsky (eds.), Invasive plants of California’s wildlands. University of California Press, Berkeley, CA. INTERNATIONAL BOUNDARY & WATER COMMISSION. 2006. Tijuana River flows at international bound- ary. International Boundary & Water Commission, El Paso, TX. Available at: http://www.ibwe.state. gov JOHNSON, M., T. DUDLEY, AND C. BURNS. 2006. Seed production in Arundo donax? Cal-IPC News 14:12-13. KATAGI, W., C. LOPER, AND N. E. JACKSON. 2002. Arundo removal protocol. Team Arundo website. Available at: http://www.sawpa.org/Arundo/index. html KELLY, M. 1999. Roundup of Arundo projects reveals commitment, strategic weakness. CalEPPC News 7:4-9, MCWILLIAMS, J. D. 2004. Arundo donax. [Online] in: Fire effects information system. U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station, Fire Sciences Laboratory. Avail- able at: http://www.fs.fed.us/database/feis/plants/ graminoid/arudon/all.html MOTAMED, E. R. AND A. H. WIJTE. 1998. Rooting by stem fragments from hanging and upright stems of giant reed (Arundo donax). Abstract only. Pp. 69— in C. Bell (ed.), Arundo and saltcedar: the deadly duo. Proceedings of the Arundo and _ saltcedar 222 workshop. University of of California Cooperative Extension Publication, Imperial County, CA. NEILL, B. AND J. GIESSOW. 2004. Distribution of Arundo donax in coastal watersheds of southern California. Available at:http://www.smslrwma.org/ invasives/Arundo/ADRegionalMap.html OAKINS, A. 2001. An assessment and management protocol for Arundo donax in the Salinas Valley Watershed. Senior Thesis. California State Univer- sity of Monterey Bay, Seaside, CA. PERDUE, R. 1958. Arundo donax: source of musical reeds and industrial cellulose. Economic Botany 12:368—404. RADOSEVICH, S. R., J. S. HOLT, AND C. M. GHERSA. 1997. Weed ecology: implications for management. John Wiley and Sons, New York, NY. MADRONO [Vol. 55 SOUTHWEST WETLANDS INTERPRETIVE ASSOCIATION. 2002. Tijuana River Valley invasive plant control program. Prepared by Tierra Environmental Ser- vices. Southwest Wetlands Interpretive Associa- tion, Imperial Beach, CA. USDA _ FOREST SERVICE. 2001. Guide to noxious weed prevention practices. Available at: http:// www.fs.fed.us/rangelands/ftp/invasives/documents/ Guideto Nox WeedPrevPractices_07052001.pdf WESTERN REGIONAL CLIMATE CENTER. 2006. Month- ly total precipitation at San Diego WSO airport, California. Available at: http://www.wrcc.dri.edu/ WocH, E. F. 2005. Dispersion patterns and associations of invasive exotic plants within the Tijuana River watershed. M.S. Thesis. San Diego State Univer- sity, San Diego, CA. MADRONO, Vol. 55, No. 3, pp. 223-237, 2008 VEGETATION CHANGE OVER SIXTY YEARS IN THE CENTRAL SIERRA NEVADA, CALIFORNIA, USA JAMES H. THORNE’, BRIAN J. MORGAN, AND JEFFERY A. KENNEDY Department of Environmental Science and Policy, University of California, Davis, CA, 95616, USA ABSTRACT In California, the Vegetation Type Map (VIM) project of the 1930’s has provided valuable historical vegetation data. Albert Wieslander led this effort to survey the forests of California in the 1930’s. His crews surveyed over 150,000 km’, drawing detailed vegetation maps, taking 3000 photos and 17,000 vegetation plots. We developed a technique to digitize the Placerville 30’ quadrangle VIM, rendering it to a Geographic Information System (GIS). The map covers 2408.8 km/° of the west slope of the Sierra Nevada. In this area VIM crews identified 59 dominant plant species and eight genera or land cover classes and mapped their distribution into 3422 polygons. They identified recently disturbed areas that covered 13.5% of the landscape. We compared the digital VITM quad to CALVEG, a satellite-derived vegetation map from 1996. Land cover change for California Wildlife Habitat Relationship (WHR) vegetation types had occurred on 42.1% of the area. WHR types with the largest gains were: Montane Hardwood, Douglas-Fir, and Annual Grassland. Low elevation hardwoods, particularly Blue Oak Woodland (dominated by Quercus douglasii, Fagaceae), chaparrals and upper elevation conifers were the types that lost the most area. Differences in mapping techniques are unlikely to be the cause of this change because the analysis used controlled for map-based errors. Potential causes of the observed change at these physiognomic levels of classification include human perturbation, succession, and climate change. Key Words: Climate change, conifer loss, historical maps, land cover change, Sierra Nevada, VTM Project. Historical landscape ecology interprets previous landscape conditions, which can provide a refer- ence for assessment of changes in dominant vegetation, which are of particular interest for evaluation of ecosystem condition. These studies tend to be unique, in that interpretations of change are dependent on the data available, and methods for interpreting a data source are usually devel- oped for each study. Notwithstanding, a wide variety of data have been used for these types of studies, for example: soil types, agricultural census, and weather station data have been analyzed to identify causes of the Dust Bowl (Cunfer 2002); historical distributions of tree species have been determined from locations of witness trees used for demarcating early ownership parcels (Cogbill et al. 2002); tree ring patterns have been used to establish historical climate and fire histories (e.g., Graumlich 1993); vegetation plot revisits have shown shifts in proportional species abundance along elevational gradients (Beckage et al. 2008; Kelly and Goulden 2008); and re-photography of historical photographs has been used to determine ecological trends (Veblen and Lorenz 1991). Historical Overview The Wieslander Vegetation Type Map (VIM) Project of California, USA was a United States ‘Email: jhthorne@ucdavis.edu Forest Service (USFS) effort to record the state’s vegetation between 1928 and 1940 (Wieslander 1935a, 1935b, 1986; Griffin and Critchfield 1972). Headed by Albert Wieslander, the group took over 3000 photographs of vegetation, surveyed over 17,000 vegetation plots, recorded field notes, and mapped vegetation across about 35% of the state (~155,000 km’; Colwell 1977). The study covered predominantly USFS lands, but three national parks (Lassen, Yosemite, and Sequoia/ Kings Canyon) were also surveyed using the same protocols (Griffin and Critchfield 1972; Wieslan- der 1986). The project also collected 25,000 plant specimens, housed at the Jepson Herbarium, University of California, Berkeley. These data collections form an important California legacy, and work is underway to systematically process them for preservation and state-wide analyses (Ertter 2000; Kelly et al. 2005; Thorne et al. 2006). The VIM project produced vegetation maps for 215 quadrangles (55 7.5-minute, 88 15-minute and 72 30-minute), though portions of some quadrangles are missing or were not fully surveyed. Twenty three of the 30-minute maps were published by the University of California Press (Colwell 1977). Most of the published maps were destroyed before sale, although 20 sets were saved by P. Zinke (Wieslander 1986). These published maps have extensive margin notes but contain reduced floristic detail compared to the original survey maps. N N a The VIM project data is the source for much of the current knowledge of tree and shrub species distribution in California. Elevational transect maps of vegetation (Critchfield 1971), and maps of the distribution of California trees (Griffin and Critchfield 1972), and range brush- lands and shrubs (Sampson and Jespersen 1963) are based heavily on VIM data. The vegetation plots have received the most research attention to date, having been used in dissertations and published studies relating to both community classification (Jensen 1947; Allen et al. 1991; Allen-Diaz and Holman 1991), vegetation change (Bradbury 1974; Allen-Diaz and Holzman 1991; Holzman 1993; Minnich et al. 1995; Minnich and Dezzani 1998; Bouldin 1999; Taylor 2000; Franklin et al. 2004; Taylor 2004), and a study of carbon sequestration in soils (Fellows and Goulden 2008). The plots have been fully digitized and are available to the public online (Kelly 2008). The photos have been scanned by the University of California Bancroft Library and are available for viewing (Bancroft Library 2008). They have been used in at least two studies: changes in San Diego County vegetation (Dodge 1975) and forest changes at Lassen National Park (Taylor 2000). At least three types of VIM maps were prepared: maps showing the locations of photo- graphs, the locations of vegetation plots, and vegetation distribution. Ancillary information sometimes was recorded on an additional set of maps that shows stands of individual trees too small to map to polygons, routes taken, fire boundaries, and sawmill locations. Wieslander intended that plot data and vegetation maps to be used together, with the maps providing the extent and the plots providing composition and structure of the vegetation (Colwell 1977; Wies- lander 1986). A number of early works presented such analyses (Weeks et al. 1934, 1943; Wies- lander and Jensen 1946). The VIM vegetation maps were used for land management analysis by USFS personnel, who would determine vegetation type extent by making a grid of points and overlaying it on the original VIM vegetation maps. Among their findings was that the lower edge of the western Sierra Nevada pine belt had moved upslope by an average of 305 m, representing a 16 km horizon- tal eastward displacement of the forest’s western edge, recorded along 54 km of forest edge in El Dorado County (Weeks et al. 1934; Wieslander 1935c). This change, attributed to logging and which occurred between the 1894 gold rush and 1934, left a deforested area of 65,560 ha, much of which had been replaced by tree and shrub species associated with lower elevations. Several studies used VIM vegetation maps as background information: Bradbury (1974) looked at changes in 134 vegetation plots and MADRONO [Vol. 55 25 landscape photos, with vegetation extents taken from the VIM vegetation maps for the Ramona Quadrangle in southern California; Freudenberger et al. (1987) examined changes in grasslands based on aerial photographs, with species identification based on VIM vegetation maps; and Davis et al. (1995; 1998) used scanned versions of the VIM maps to assist in attribution of satellite imagery-derived vegetations polygons for the Gap Analysis Program’s (GAP) vegeta- tion map of California. VIM species were used as GAP attributes if modern satellite image-derived shapes of vegetation polygons were similar to the original VIM polygon shapes. Walker’s (2000) dissertation assessed an early attempt to digitize VIM vegetation maps into a GIS. He used the VIM vegetation map from Yosemite National Park, which park scientist J. W. van Wagtendonk had digitized in the early 1980’s. The digital version was produced on a digitizing table; the exact methods used are not known. Walker compiled the digitized line files he received from the park. He determined that the topographic map edition the VIM was drawn on (the base map), had non-systematic topographic mis-registration errors of up to 250m _ when compared to a newer, digital version of the topography. He treated the problem by applying some 14,000 tie points to warp the historic maps (the park is spread over parts of four, 30-minute topographic quads) to the digital topography. Subsequently, species lists from modern surveyed vegetation plots were in good agreement with the vegetation of the VIM polygons they were located in. This paper presents an analysis of change in vegetation from a 30-minute quadrangle in California’s Sierra Nevada that was part of the Wieslander Vegetation Type Mapping (VTM) survey. We developed a technique for rendering the VIM vegetation maps to GIS, applied it to — the quadrangle, and compared dominant vegeta- tion type extents to a second vegetation map, | mapped in 1996, comprising a time step of 62 yr. METHODS We selected the VIM map for the Placerville | 30-minute quadrangle, located in the Sierra | Nevada (Fig. 1) and developed a method to | digitize it, with the intent of examining changes in | vegetation extent. The VIM map was originally | surveyed according to a protocol which included: | surveying from ridgelines; drawing vegetation | polygons on the topographic base map; and recording the dominant plant species as codes in | each polygon. Once species codes were recorded, | field crews colored the polygons to match a habitat-level classification, drew cross-hatch lines to indicate burned or logged areas, and fixed the | colors using a high quality benzene or gasoline ; 2008] Fic. 1. THORNE ET AL.: SIXTY YEARS SIERRA NEVADA LANDSCAPE CHANGE Kilometers Study area. Wieslander VIM crews surveyed the vegetation of the Placerville 30 min quadrangle, on the western slope of the Sierra Nevada, between 1931 and 1934. (Wieslander et al. 1933). The vegetation recorded on the Placerville VIM quadrangle was mapped between 1931 and 1934 (Weeks et al. 1943). It was drawn directly on a 1931 edition of a United States Geological Survey (USGS) topographic quadrangle, which had been surveyed in 1887 and first published in 1893. Aerial photographs were not used. When finalized, the VIM map was cut into 16 sections (tiles) and glued to a canvas backing in blocks of four tiles, with spaces between the tiles where the map could be folded when taken in the field, which protected the maps from data loss along the creases. Walker’s dissertation (2000) indicated that bringing the VIM vegetation maps into align- ment with modern topography would be time- intensive. Therefore, we used a different ap- proach: to produce a VIM map to the level of spatial accuracy at which it was originally produced. Since the base topographic map upon which the VTM was drawn was surveyed at the turn of the century, we would reproduce the -VTM map to those standards and not add the additional step of warping the image to get it to conform to later, improved, topographic accura- cy. There were several reasons for selecting this approach. First, there was concern about intro- ducing change to vegetation area estimates as the historic map would need to be warped to modern topography. Also, we were less concerned with identifying what change had happened on any 100 m’ than in assessing change in dominant vegetation across the whole 2400 km? represented on the quadrangle. Future research can apply the additional processing required to replicate Walk- er’s approach from our product, should that level of detail be needed, and funds are available. Each of the 16 VIM tiles composing the original VTM map was scanned on a flatbed scanner at a 300 dots per inch (dpi) resolution. This produced 16 20-megabyte images. To assemble these tiles in a georeferenced manner, the first step was to register each vegetation tile onto a copy of the original topographic map. Thus, we obtained a scanned version of the 1931 USGS topographic quad (sheet scanned at the UC Santa Barbara, Alexandria Digital Library at 300 dpi), the same topographic quadrangle edition used by the VIM mapping crew. Base Topographic Map Registration The original projection of the USGS _ base topographic map is polyconic and uses Clarke’s spheroid of 1866, map standards used by the USGS between 1886 and the late 1950’s (U:S. Department of Commerce, Coast and Geodetic Survey 1917; Snyder 1982). This projection, which runs a meridian through the center of a map, was used because of its high level of precision on any given quad, although the assembly of multiple quads leads to high levels 226 of distortion. USGS field survey crews were able to construct such polyconic maps in the field, on which they recorded the topographic features (Gannet 1904; Beaman 1928). Using ERDAS IMAGINE (Leica Geosystem 2004), the scan of the USGS topographic quadrangle was trans- formed into polyconic projection using 16 control points placed at the intersections of the quad’s latitude and longitude reference lines. VTM Tile Registration We re-assembled the VIM map to a single image by using ERDAS IMAGINE to register each of the 16 digital VIM tile images to the scanned topographic base map, using nine control points per tile. These control points were selected from features common to both maps, such as text, road intersections, and contour lines. Vectorization and Polygon Attribution Once the digital topographic base map and VTM tiles were in their native projection, it was possible to trace vegetation polygon boundaries. Several semi-automated vectorization techniques were tested, but these were not efficient due to the large number of other line features, such as elevation contours, on the map. Supervised image classification techniques also proved unsuitable as the coloring on the VIM tiles was smeared or deteriorated. We therefore used heads-up digitiz- ing, using a Wacom Cintiq 21UX digitizing tablet (Wacom 2004), a LCD flat screen that permits the use of a digital pen to draw lines. Used in streaming mode, the pen assigns multiple vertices as a line is drawn. Linework was traced from the digital images at a resolution of 1:7000 or finer, a scale at which the thickness of the line being produced was equal to the lines being traced. The line production required that every dot in the dotted lines demarking vegetation polygons be intersected by the hand-drawn line. Once a shapefile of the linework was completed, poly- gons were built, and attribute columns added (ESRI 2004). Species codes written in each polygon were recorded in the order in which they appeared. We used a species crosswalk (G. F. Hrusa, Senior Plant Taxonomist, CDFA, personal com- munication) to convert VTM-era species names to current scientific nomenclature (all authorities from Hickman 1993) for each species code on the map. Once the VTM species codes were entered, we used the ESRI ArcMap “Join” function to assign modern species names to the map codes. The resulting combinations of modern species names were then assigned to vegetation alliances, as documented in the Manual of California Vegetation (MCV) classification system (Sawyer and Keeler-Wolf 1995), and thence to California MADRONO [Vol. 55 Wildlife Habitat Relationships (WHR) types (California Department of Fish and Game 2004). Spatial extents for each habitat type were then developed. The VIM maps record dominant species as they occurred in each stand in order of percent cover, according to a standard set of cover thresholds specified in the VIM field methods manual (Wieslander et al. 1933). This recording of dominant species in whatever combinations they occurred provides data that can be translat- ed into multiple classification systems. WIM polygons labeled with a single dominant species contain a minimum cover of 80% for that species. A polygon attributed with multiple species specifies each co-dominant cover at least 20% of the polygon. Species in VTM polygons can be grouped in one of four recognized growth form strata: trees, shrubs, herbs and grasses. Species from these classes can co-occur or be separated when classification to vegetation type is assigned. One or two vegetation types occur in VIM polygons. Polygons with two types (mosaic polygons) contain contrasting growth forms, whose component sub-stands were either smaller than the minimum mapping unit, or in such a complex spatial pattern as to be impractical to map individually. In the case of a mosaic polygon, a species was listed if it covered a minimum of 20% of the mosaic sub-component, rather than 20% of the entire polygon. Mosaic polygons were labeled with species codes in a m1x of horizontal and vertical orientations, so that the component vegetation types could be identified in a mosaic type. Unfortunately, the relative cover of each component type was not indicated in the attribute label. Consequently, we assigned the area of vegetation in mosaic polygons, with primary types receiving 2/3 and secondary types 1/3 of a polygon’s area. The VIM vegetation maps contain a large number of unique species combinations. To. crosswalk these combinations to MCV alliances, | we used the dominant species and disturbance | information listed in each VIM map polygon, . together with the MCV type descriptions and our | knowledge of California vegetation patterns. We developed a set of rules to determine whether a | polygon’s vegetation was a mosaic or a single’ unit: | 1) Grass was included in woodland types, but | excluded from chaparral, forming a mosaic. vegetation component when found with chaparral types. 2) In most cases, if conifers were present with oaks, oaks were assumed to be subdomi- nants. | 3) Riparian species were always considered a’ secondary type with other dominants when co-occurring. | 2008] 4) Shrubs were assigned to (or excluded from) dominant conifer classes depending on disturbance, the species of shrub, and known general canopy density of the dominant trees. Generally, upper elevation conifer types include shrubs as sub-compo- nent (e.g., Arctostaphylos patula, Ericaceae), while the lower elevation pine types (all Pinus sabiniana, Pinaceae) exclude shrubs. 5) Species in the genus Quercus found in recently disturbed, chaparral shrub-domi- nated polygons were included as chaparral. If no disturbance was indicated, then chaparral and oak tree species were broken into mosaic types. The MCV vegetation types were then cross- walked to WHR types, using a defined crosswalk from California Department of Fish and Game (2004). The WHR classification is based on physiognomic characteristics associated with a named dominant plant species, creating a habitat type often used to model vertebrate distributions in California. Registration and Map Accuracy Assessment Root mean square (RMS) error in meters was calculated for the process of registering VTM tiles to the topographic base map. Line vectorization RMS error was assessed by comparing the drawn lines to the registered VIM tiles. An estimate of RMS error to modern topography was calculated by registering 73 road locations to a modern digital coverage of roads (U.S. Department of Commerce 2001). Comparison to Modern Maps Two modern digital vegetation maps were available for comparison: the California Gap Analysis vegetation map (Davis et al. 1998), and the 1996 CALVEG land cover map (Schwind and - Gordon 2001). We did not use the Gap Analysis vegetation map since many of its polygon attributes were derived from VIM maps. The CALVEG vegetation map source is independent, as its attributes were derived directly from 30 m _ resolution satellite imagery. CALVEG land cover classes are delineated through an automated process that is regionally calibrated and which produces | ha minimum size polygons labeled _with WHR class land cover attributes. Both VIM and CALVEG maps were re- projected to Albers Equal-Area projection, and the CALVEG map was clipped to the extent of the Placerville quadrangle. The vegetation type extent on each quad was summed by WHR type. Prior to comparison with the VTM map, the high polygon density in the CALVEG map was reduced by dissolving the borders between THORNE ET AL.: SIXTY YEARS SIERRA NEVADA LANDSCAPE CHANGE 227 adjacent polygons that had the same WHR attributes. Tabular summaries of WHR extents on each quad were compared in a non-spatial context to determine the major trends of land cover change. A graphic illustration of that change was produced by intersecting the maps, but the area extent reported represents independent measures of each map. We analyzed land cover change at the scale of the entire quad, developing a table showing the historic and current extent of each WHR type. RESULTS VTM Map Processing The VIM map covers 2408.8 km?’ and ranges from 200 m elevation in the southwest corner, on the Cosumnes River, to 1627 m in the northeast corner near Devil Peak and the Rubicon River. The registration of the topographic base map yielded a 7.4 m root mean square (RMS) error for control point error. The registration of the VTM tiles onto the base topographic map yielded a RMS error of 2.0 m. Therefore, geographic registration of the VIM map to the historic topographic map was completed with a RMS error increase of 667.7 km* of Upper Elevation Hard- | wood, and decreases of 156.6 km? of Lower. Elevation Hardwood (including Foothill Pine), | and 177 km* of Chaparral. Working Landscapes | (grasslands and agriculture) increased 25.1 km’. Six WHR classes mapped by the VIM survey | were not present in the CALVEG map, and three WHR classes mapped by CALVEG, were not in| the VIM (Table 4). | DISCUSSION The two possible over-arching reasons for the observed change are 1) map-based errors, in. 2008] TABLE 2. SPECIES IDENTIFIED IN THE WIESLANDER VTM PLACERVILLE QUADRANGLE. Wieslander vegeta- tion mappers identified 59 species (Authorities Hickman 1993), one genus, two habitats and five land cover classes on the 2408 km/? of the Placerville quadrangle. Species Abies concolor Acer macrophyllum Adenostoma fasciculatum Aesculus californica Aira caryophyllea Alnus rubra Arctostaphylos manzanita Arctostaphylos mewukka mewukka Arctostaphylos nissenana Arctostaphylos patula Arctostaphylos viscida Avena barbata Avena fatua Bromus carinatus carinatus Bromus diandrus Bromus hordeaceus Calocedrus decurrens Ceanothus cordulatus Ceanothus cuneatus Ceanothus integerrimus Ceanothus leucodermis Ceanothus parvifolius Ceanothus spinosus Ceanothus tomentosus Cercocarpus betuloides Cercocarpus ledifolius Chamaebatia foliolosa Cupressus macnabiana Eriodictyon californicum Erodium cicutarium Heteromeles arbutifolia Hypericum perforatum Lithocarpus densiflorus Lithocarpus densiflorus echinoides Pellaea mucronata Pinus attenuata Pinus contorta murrayana Pinus jeffreyi Pinus lambertiana Pinus monophylla Pinus ponderosa Pinus sabiniana Prunus emarginata Pseudotsuga menziesii menziesii Pteridium aquilinum pubescens Quercus berberidifolia Quercus chrysolepis Quercus chrysolepis nana Quercus douglasii Quercus durata Quercus kelloggii Quercus lobata Quercus wislizeni Quercus wislizeni frutescens Toxicodendron diversilobum Trifolium variegatum Umbellularia californica Vulpia myuros hirsuta THORNE ET AL.: SIXTY YEARS SIERRA NEVADA LANDSCAPE CHANGE 229 TABLE 2. Continued. Species Genera and Physiognomic Types Salix sp. Grass Wet Meadow Unidentified Code Hp Land Cover Types Barren Cultivated Residence River Rock which case the question is how the errors were introduced; and 2) the dominant vegetation has changed, in which case the question becomes what are the drivers of the change. If the changes are due to map-based errors, those could derive from the processing of the historic data, vegeta- tion type classification errors, or inaccuracies in the historic or modern maps. If the changes are real, they could be due to a variety of drivers, including human activity, climate, and their combined effects on fire, succession, and disease. This discussion reviews the possible sources of error first, followed by consideration of the measured vegetation change. Potential Map-Based Errors There has been much discussion of the accuracy of the VIM survey (Keeley 2004; Kelly et al. 2005; Kelly et al. 2008), and a legitimate question is to what degree can we rely on these maps to portray historical conditions? Two possible sources of error in the maps are registration error, due to inaccuracies in the topographic base map on which the VIM maps were drawn; and classification error, in which species combinations from the VIM maps are not correctly interpreted, or are not comparable to the CALVEG use of the WHR classification. The largest source of spatial error we could detect in the VIM vegetation maps comes from the underlying topographic maps, rather than from the vegetation mapping itself. We avoided geo- registration issues that would be introduced by intersecting the maps from the two time periods, through tabular compilation of the vegetation type extents from each time period and compar- ison of the resulting tables. This did not permit us to follow the progression of change at any given location, but the overall extents of each vegeta- tion type within the study area were comparable. However, we present map overlays to help visualization of where the change for some habitat types occurred, and to illustrate the [Vol. 55 ~ MADRONO N OL’9LL‘907 0€°890°0E 0L°996C %001 OL LI8‘OrZ OO'CCHE [¥}0 1 cv %00'0 cv I (Nd’T) 9Uld 2[odespoT] c9 %00'0 C9 I (DSO) qnog [eIsBoD 68 %00'°0 68 ! (aun) ueqin 6 OI %10°0 6 Ol I (Ndd) 2Uld spisiseq esl %10°0 ool S (XXX) UMoUyUS) VCC %10°0 VCC ss (ALLA) MOpRey 19 6 LS %CO 0 6 LS I (THA) ueiedry yrmoo, AayeA SOL %eO 0 SOL 8 CUVA) vssreg CSC %O1'O DESC C CUAM) A OHA [Tver SIs %0T 0) 9°S8L LI (MOA) PURTPOOM FRO ALBA €°S09 %ST'0 €°S09 ce (QdD) ssaidAD-auld 2U0D-peso[D 8'0C8 8 OSL MVS) 9°06cI fe (dqHA) Utd [Ty004 EL bot %08'0 VLI6l Ot (TWIN) ueledry surjuoyy L CLO L Soll 611 © 898C Sv (OHW) JestuoD-poompiey suvjuopy OCOD 61 WVS'C gS SOI9 O8t (dao) pueldoirD Vv 6LS9 OCLC vO6LS9 LC (AAC) HWA-sejsnog TP 1c0e EiOLLe MC8 CT 8 0089 tC (dOW) [etuedeyD ouviuoyy rsie Ol S| Dev I'rer ol 661 (OUD) [euedeyD Yyueyspoy-osrumeyD SOS TI Se86l TCO § EVES El OCI (MO) PUPTPOOM ARO enl_ 6 900C p9CT Ol CCLI %08'S €S96 ET L0c (HOW) [etiedeyD pox cor v1 TO1oe T0LC.L Tc0s Lt 6c (dOd) 9Uld []IYIOO4-ARO ong 7 £90 TI 6 V6t8 © OF %60'8 '88r 61 6¢9 (SOV) purisseip enuuy 8 LI98I t 6¢8 £ SCC YLI'8 v'789'6I tcV (MHW) poompiey sueju0opy Ecce Toss Cl IL7CO€ 00C (OWS) Joftuod Pex UPLIIS 9°988°98 16861 Y169€ L°SL8°88 DIL (Ndd) 9Uld Pso.spuod (ey) poinsevoul (ey) suing 0} (ey) SUISSO] 0} Bore [¥10} (ey) voy suosdjod odAi WHM s0uRgINIsSIpP ON onp [eios Apieq onp [esos Ape JO JUDIIIg jo 1oquUINN, ‘penb ot Jo WI'Fl potdns90 passo] us0q 10 pouing ATUD901 pey yey} sodA) YH AA JO WAXY peo] punoie o[suvipenb ayflAss0v"][g dy} UO UTe.IID} SOU OY) SuIAdNd90 9d Aq ADAOOpUR] JURUTWOp oY] SPM suid vSOIOpUOd ‘pe6l NI FATONVYdVNAYD ATIAYAOVTId AHL NO SLNIYd LOOY AONVEANLSICG| GNV SddA [ dIHSNOILVTAY LVLIGVH SAITATIA, SO LNALXY *¢ ATEV EE 2008] THORNE ET AL.: SIXTY YEARS SIERRA NEVADA LANDSCAPE CHANGE 25) 121°0'0"W 120°50'0"W 120°40'0"W 120°30'0"W oe Z S g Dp a ~ ~m = 2 2 S iB 2 99 a) i) isa) 2 2 iS E 5 8 ge Bo oe aa) = 2 2 =) a =) 8 ae ®% 8 120°50'0"W 120°40'0"W 120°30'0"W Kilometers f 0 5 10 15 20 25 30 3 WHR Types $B Ponderosa Pine (PPN) Montane Chaparral (MCP) oo Tree Dominated GM Closed-Cone Pine-Cypress (CPC) | Herbaceous Dominated / Blue Oak Woodland (BOW) HM Sierran Mixed Conifer (SMC) Annual Grassland (AGS) g ry (MB Blue Oak-Foothill Pine (BOP) HRM Douglas Fir (DFR) Other i, 1) Valey Foothill Riparian (VRI) HE White Fir (wFR) Cropland (CRP) } HB Montane Riparian (MRI) WHMM Eastside Pine (EPN) (urban (URB) Valley Oak Woodland (VOW) Shrub Dominated Barren (BAR) Montane Hardwood (MHW) Chamise-Redshank Chaparral (CRC) [J Unknown (xxx) N [9) Montane Hardwood-Conifer (MHC) Coastal Scrub (CSC) wiSiere bon a BR Lodgepole Pine (LPN) Mixed Chaparral (MCH) M Fic. 2. The extent of California Wildlife Habitat Relationship landcover types on the Placerville quadrangle in 1934, We classed the historical species distributions mapped by VTM field crews into 25 Wildlife Habitat Types. _ The original map (insert) has the species codes used to identify the WHR types written in each polygon. [Vol. 55 ~ MADRONO AN laa) “NA %0 9 %9 OE — Clo 8e= Tao. Ie 691°L8 YC CS (Ndd ‘OWS “YAC) J9sTuoD DS tlie %0° LS — Ef 8I= %L'S TLO'E Tacel (Odd ‘OUD “HOW “dOW) TestedeyD %S'9— %L 6b — 7999'S 1 — %9°9 198°C Mel (dOd ‘MO ‘MOA) poompieH uoNnPRAs[y Jomo] %L'O— YT S3— eSol= %10 98C %8'0 (IU ‘THA SWALAM) Uerredry %0'I %8'6 IISc YL II SOI87 %9 OI (uO “SOV) sodeospuryT sunpiom YL‘ LT %V'96T OLL‘99 YVLE OCE“68 Mv'6 (OHW ‘MHW) poompiey uonraslq todd (Jso]) 10 poutes (JSoO]) 10 (ey) (SO]) 10 BIB [10] JO (ey) vore voIe [B10} JO (sodA YHM) Set089}e9 pouIquIo0d deur jo jus010g MLO ToL eG = PLO 8 e— Do C= Tole 7 /o) OO ie %6 O— iL. 0— YT O— %VO— %1O— %0'0 %0'0 %0'0 %0'0 %0'0 %0'0 %0'0 %0'0 vO VO MSO) %8 I ToC %0'L YL'8 YL OT (Jso]) 10 poures deut Jo 1Ud010g poures 1U00190g YC VI— %S°T6— %6 L3— %LCO— 3 I= Des YI 91 — %6 93 — Vege Us ame VT — %0° 001 — %0' 001 — %0° 001 — %0'00T — %0 001 — %0° 001 — %0°0 8 LS %0'0 MSSLI %6' CITI %0'0 %T ELE SP %0'9T YL 16S I OLE YOEST (JSO]) 10 poures odAj Jo JUDDIOg poules voy %001 6£0 LS — MT El 16191 — MSO vLI6— MSO cOEO— MT 0 ClSC= SI OOS c— MSV LVCC— 06 V ie be) Ve %I1'0 OIn— %I1'0 SG MSS tSc— %0'0 8S — %0'0 Sl — %0°0 | %0'0 o> %0'0 v— %0'0 9 %0°0 el %0'0 Sl %0'0 cS8 %9'0 068 MVv'O S8Il %S'O 68CL %8'I £906 %Ce Ol IL6 91 C8 97017 STI 66L 6b %8°8T (ey) (soy) IO poules voly Bale [e101 JO jusdiod ONAATVO PUP OAATVOD Jo luad1sd WLA qusdied OAATVO = DAATVWO 8L8°0r7Z LES TE Clel 09CI 860 VSSE LCL UC 6IL'TI ISc S6l LITE! (ey) quaoied WELA %001 %S'0 M6 9E De L Dev %8°C MSC %9'T1 %8°S %8'0 YEO %I'S %1'0 %0'O %0'0 %0'0 %0'0 %0'0 %0'0 %0°0 %0'0 YT O %0'0 %0'0 %0'0 %1'8 YC I Yat, C C8 RaIe [P10} ZI8‘OrZ [BIO], 16cl (qHA) 2Uld []yI004 9L8°88 (Ndd) 2Uld eso1apuodg COS LI (dOd) 9Uld [[YI004-AVO sng rer Ol (DUO) [erredeyD yueyspoy-osrumeyD 1089 (dOW) [erredeyD suriwopy 9019 (quo) purldosp LCT OE (OWS) J9ftuoD pextfl UPLIIIg S96 ¢1 (HOW) [esredeyD pox LI61 (IIA) ueuedry ourjuopy S09 (OdD) ssoidAZD-dulg 9U0D-psso[D pes el (MO) PUPTPOOM APO eng CSC (CHAM) JT OU. 8S (TMA) Uerredry [[Mpoo4 AayeA aI (xx xX) UMoUyUL:) a (Ndd) SUI opisiseq 9 (OSD) qn4dg [Riseog v (Nd‘J) 9Uld 2[oda8po7T 0 (MOD) PURTPOOM YeO [eIsvOD cc (ALLA) MOpRre| 1PM 0 (Onda) smdAyeong 98P (MOA) PURTPOOM ARO APA Lz (UVa) voseg 0 (LVM) JRA 6 (quan) ueqiy 88r'6l (SOV) pursserH yenuuy S987 (OHW) JaJtUoD-poompiepY{ suevjUOpy 6LS9 (WAd) I4-sejsnoqg 78961 (MHW) poompiey ourju0jy (vy) od41 WHAM Bole WALLA ‘(sasayjuaied Ul sopod WHAA 24} Aq poledIpuUl UONISOdUIOS sadA} PAUIQUIOD ‘UI0}}0q 1B) MIIADI JO ISVd IOJ POUIQUIODS 319M SddA} YH AA SUIOS “1Op1o Surpusosoap ul poyues sem sodA} Jo penb afiAssov[g dy} UO SJUD1X9 YH M JO IsUPYD [[PISAC “9661 GNV Pe6l NAAIMLAG SLNALXY AdAL YAAOOGNV’ dIHSNOLLV19dY LVLIGVH JAITGTIA, NIFDNVHD AAGUO GAANVA ‘f ATAVL 2008] largest patterns of land cover change. A rough estimate of the spatial registration errors is therefore of interest. The RMS error developed from registering the topographic base map to the digital roads map (U.S. Department of Com- merce 2001) was 263 m. If we assume the spatial error to be spread equally among vegetation types, this approach permits an estimate derived by buffering VTM polygon boundaries with the RMS error, that 700 km? (34%) of the map is spatially precise. From a sampling perspective, 34% of the landscape is sufficient to capture the dominant trends. Note that in many areas of the map, the registration was considerably better than the average error. The taxonomic component of the Wieslander surveys is the other potential source of error. However, this error is not likely to have been introduced by the VIM mappers themselves. The Placerville VIM map is floristically remarkable in its scope and detail. Wieslander wrote, **... the reliability of all future assumptions and conclusions to which any analysis of the VIM data may lead is dependent directly upon the character and quality of the field work. Therefore it is of utmost importance that this phase of the job be done strictly according to the rules set up and with consistency in their application.” (Wieslander 1933). Accordingly, the species recorded are likely correct. Error would more likely be introduced in the classification of species to MCV and WHR types. However, the VIM species lists and color codes on the VIM map allowed for rule-based crosswalking of the component VTM species into modern, pre-defined, classifications. Classification error may also enter where the cover thresholds used to define CALVEG types differed from those used to define species in the VTM maps. CALVEG assigns transitions from grass to shrub, shrub to hardwood, and hard- - wood to conifer if >10% of the polygon contains the taller type, as opposed to >20% of a polygon being occupied by any co-dominant for a species to be listed in the VIM system. This generally means that a loss by a taller type to a shorter type _is a conservative measure, whereas a gain in a taller type at the expense of a shorter type is less _ certain. For example, some of the measured loss of chaparral between time periods may be due to _CALVEG classification of VTM Chaparral areas as hardwood. However, oaks commonly establish in chaparral, and so this potential error cannot be definitively identified because succession may actually be causing oak types to overtop chapar- ral on parts of the landscape. The large amount of the WHR type Montane Hardwood on the landscape is potentially partially due the CAL- VEG classification, which would favor this type, THORNE ET AL.: SIXTY YEARS SIERRA NEVADA LANDSCAPE CHANGE 253 if only 10% of a brush field was in oaks that were discernable. The indicated loss of conifers is particularly important to consider from the classification perspective. The loss appears real because the CALVEG map assigns conifer classes to less dense conifer stands than the VIM map. Therefore, the contemporary map is biased to producing more area in conifer than the VIM map, meaning a change showing reduction in conifers is more likely to be real. Since this is the biggest change on the landscape, we are confident it is real, where the shifts indicated in chaparral types are more problematic. Land Cover Change Loss of Q. douglasii extent (WHR types Blue Oak Woodland, Blue Oak-Foothill Pine, and Foothill Pine) and increase of grasslands (WHR type Annual Grassland) were the dominant trends at low elevation. The loss of Q. douglasii is also a transition from larger to shorter vegetation, making the estimate of change a conservative one. Potential explanations for these land cover changes include grazing, which impacts Q. douglasii recruitment (Hall et al. 1992) and which, in combination with pressure from wood cutting, contributes to QO. douglasii conversion to grassland. El Dorado County, which occupies most of the Placerville quad, has also experienced rapid population growth (10.6% between 2000 and 2004), with consequent devel- opment of low density rural residential areas (United States Census Bureau 2006) that could affect lower elevation vegetation. Chaparral types associated with lower elevations showed a reduc- tion in extent, which have potentially transitioned to Montane Hardwood or Montane Hardwood Conifer. The second major pattern of land cover change was at upper elevations, where the extent of P. ponderosa-dominated areas (WHR type Ponder- osa Pine) was reduced by 64%, a loss of 570 km? (23.7% of the whole map). Pinus ponderosa areas were replaced by WHR type Montane Hard- wood, whose dominant species is defined as Quercus kelloggii, Fagaceae, by WHR type Montane Hardwood-Conifer, where Q. kelloggii mingles with emerging conifers, and by WHR type Douglas-Fir (Fig. 3c for Montane Hard- wood-Conifer, 3d for Montane Hardwood). There are several possible explanations for the changes in the upper elevations. First, since Q. kelloggii is a subdominant tree under conifers, particularly P. ponderosa, the measured increase in hardwoods could be due to the loss of the conifer overstory canopy, where the understory remained in place. Second, all the oak tree species at this elevation stump sprout, and are often found co-mingled with chaparral shrubs. Hence, 234 MADRONO [Vol. 55 Me Gained (1996) Historic Remaining (1996) w WH Historic Lost (1934) FIG. 3. The major changes in the WHR extents on the Placerville quadrangle between 1934 and 1996. Changes in WHR habitat types fall into distinct elevation zones. At lower elevations, Blue Oak Woodland and Blue Oak- Foothill Pine (3a) were greatly reduced, while Annual Grassland and Cropland increased (3b). At mid elevations, Ponderosa Pine, the lowest growing of the Montane WHR conifer-dominated types, showed the greatest reduction (3c: red indicates Ponderosa Pine WHR type lost over time, blue indicates historic extent remaining and green the Kilometers 2008] some expansion in hardwoods could be the result of succession, where oaks in brush fields follow- ing fire or clear cutting of conifers overtop the shrubs they are growing with. The numbers suggest some of this is occurring due to decreasing chaparral extents. This pattern of lower P. ponderosa replaced by oaks was also observed in 1934, when (Weeks et al. 1934) estimated that the coniferous belt had previously (in 1850) extended as low as 305 m, but had retreated greatly uphill by the time of his survey. While P. ponderosa has been replaced by hardwoods at lower elevations, it has been replaced by Pseudotsuga menziesii var. menziesii, which has grown into it at higher elevations, changing the WHR type (Fig. 3c). Two overlapping processes may be driving pattern on this landscape: disturbance, which removes structure and vegetation types; and subsequent succession with potentially altered pathways reflecting new climatic and disturbance regimes, where a distinct change in potential vegetation 1s manifested as a different vegetation type in the current map, from what was present in the 1930’s. Succession may not ultimately result in the re-establishment to the vegetation mapped in the 1930’s, or it may take longer to return to the historic type due to climatic shifts. Temper- ature measurements from nearby Placerville indicate that minimum monthly temperatures there have been warming at the rate of 0.089° Cyr ' (R° = 0.71, P < 0.001) between 1960 and 2000 (data from National Climate Data Center 2005). Such a shift could potentially contribute to vegetation change by making re-establishment conditions post-disturbance less suitable for prior vegetation, or make the time required for successional regeneration longer, which in turn increase the probability of another disturbance before the historic type is reached. CONCLUSION This study demonstrates the utility and impor- tance of historical vegetation maps. While their development is technically challenging, the po- tential insights that can be derived are of lasting import. The Placerville VTM quadrangle con- _ tains information of about 59 species, which is _ better taxonomic resolution than any comparable _ modern vegetation map. While there are legiti- mate concerns over the use of VIM vegetation maps to detect change, at least at the landscape _ scale, they provide information available from no other source. It is hoped that this presentation of the map development methods and their appli- THORNE ET AL.: SIXTY YEARS SIERRA NEVADA LANDSCAPE CHANGE 239 cation will lead to interest in developing this data for the remaining regions available in California. ACKNOWLEDGMENTS This study was made possible by grant # PNV 03- JV-11261979-162 from the US Forest Service Pacific Northwest Research Station, the western regional science coordinator for the US Forest Service; and by grant # 500-02-004 from the California Energy Commission. Dr. Barbara Allen-Diaz, UC Berkeley, permitted access to the VIM collection and offered useful commentary. Dr. Fred Hrusa assisted with the development of the species lookup table, Dr. Maggi Kelly, Ken-Ichi Ueda, and Dave Shaari, UC Berkeley, assisted with technical problems, and staff of the UC Berkeley Geology Library and the UC Santa Barbara Alexandria Digital Library assisted with scanning maps. We thank Aaron Fulton, Camille Kustin, Alexa Callison-Burch, and Myra Kim for their patient work digitizing and attributing the original map. We thank Dr. Jim Quinn, UC Davis, for providing the space to house computers and host the work. LITERATURE CITED ALLEN, B. H., B. A. HOLZMAN, AND R. R. EVETT. 1991. A classification system for California’s hardwood rangelands. Hilgardia 59:1—45. ALLEN-DIAZ, B. H. AND B. A. HOLZMAN. 1991. Blue oak communities in California. Madrono 38:80—95. BANCROFT LIBRARY. 2008. Wieslander Vegetation Type Maps and Photographs website. Available at: http://www.lib.berkeley.edu/BIOS/vtm/. BEAMAN, W. M. 1928. Topographic instructions of the United States Geological Survey. Department of Interior, U.S. Geological Survey, United States Government Printing Office, Washington, DC. BECKAGE, B., B. OSBOURNE, D. G. GAVIN, C. PUCKO, T. SICCAMA, AND T. PERKINS. 2008. A rapid upward shift of a forest ecotone during 40 yr of warming in the Green Mountains of Vermont. Proceedings of the National Academy of Sciences 105:4197—4202. BOULDIN, J. R. 1999. Twentieth-century changes in forests of the Sierra Nevada, California. Ph.D. dissertation. University of California, Davis, CA. BRADBURY, D. E. 1974. Vegetation history of Ramona Quadrangle San Diego County, California (1931— 1972). Ph.D. dissertation. University of California, Los Angeles, CA. CALIFORNIA DEPARTMENT OF FISH AND GAME. 2004. The California wildlife habitat relationships sys- tem. California Department of Fish and Game, Sacramento, CA. Available at: http://www.dfg.ca. gov/whdab/html/wildlife_habitats.html COGBILL, C. V., J. BURK, AND G. MOTZKIN. 2002. The forests of presettlkement New England, USA: spatial and compositional patterns based on town proprietor surveys. Biogeography 29:1279—1304. COLWELL, W. L. 1977. The status of vegetation mapping in California today. Pp. 195-220 in M. G. Barbour and J. Majors. (eds.), Terrestrial — new extent of a combination of Ponderosa Pine, Douglas-Fir, Sierra Mixed Conifer, and Lodgepole Pine WHR _ types), while Montane Hardwood showed the greatest increase, replacing much of the Ponderosa Pine (3d). 236 vegetation of California. John Wiley & Sons, New York, NY. CRITCHFIELD, W. B. 1971. Profiles of California vegetation. Research Paper PSW-76. USDA Forest Service, Pacific Southwest Forest and Range Experiment Station, Berkeley, CA. CUNFER, G. 2002. Causes of the dust bowl. Pp. 93—103 in A. K. Knowles (ed.), Past time, past place, GIS for history. ESRI Press, Redlands, CA. DAVIS, F. W., P. A. STINE, D. M. STOMS, AND A. D. HOLLANDER. 1995. Gap analysis of the actual vegetation of California: 1. The southwestern region. Madrono 42:40—78. , D. M. Stoms, A. D. HOLLANDER, K. A. THOMAS, P. A. STINE, D. ODION, M. I. BORCHERT, J. H. THORNE, M. V. GREY, R. E. WALKER, K. WARNER, AND J. GRAAE. 1998. The California gap analysis project: final report. University of Cali- fornia, Santa Barbara. DODGE, J. M. 1975. Vegetation changes associated with land use and fire history in San Diego County. Ph.D. dissertation. University of California, Riv- erside, CA. ERTTER, B. 2000. Our undiscovered heritage: past and future projects for species-level botanical inventory. Madrono 47:237-252. ESRI. 2004. Arc Info. Geographic Information System software, Redlands, CA. FELLOwS, A. W. AND M. L. GOULDEN. 2008. Has fire suppression increased the amount of carbon stored in western U.S. forests? Geophysical Research Letters 35:L12404, doi:10.1029/2008GL03396S. FRANKLIN, J., C. L. COULTER, AND S. J. REY. 2004. Change over 70 yr in a southern California chaparral community related to fire. Journal of Vegetation Science 15:701—710. FREUDENBERGER, D. O., B. E. FISH, AND J. E. KEELEY. 1987. Distribution and stability of grass- lands in the Los Angeles Basin. Bulletin of Southern California Academy of Sciences 86:13— 26. GANNET, S. S. 1904. Geographic tables and formulas. United States Geological Survey, Department of Interior, Government Printing Office, Washington, DC. GRAUMLICH, L. J. 1993. A 1000-year record of temperature and precipitation in the Sierra Ne- vada. Quaternary Research 39:249—255. GRIFFIN, J. R. AND W. B. CRITCHFIELD. 1972. The distribution of forest trees in California. Research Paper PSW-82. USDA Forest Service, Pacific Southwest Forest and Range Experiment Station, Berkeley, CA. HALL, L. M., M. R. GEORGE, D. D. MCCREARY, AND T. E. ADAMS. 1992. Effects of cattle grazing on blue oak seedling damage and survival. Journal of Range Management 45:503—506. HICKMAN, J. C. (ed). 1993. The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. HOLZMAN, B. A. 1993. Vegetation change in Califor- nia’s blue oak (Quercus douglasii) woodlands 1932— 1992. Ph.D. dissertation. University of California, Berkeley, CA. JENSEN, H. A. 1947. A system for classifying vegetation in California. California Fish and Game 33:199— 266. MADRONO [Vol. 55 KEELEY, J. E. 2004. VIM plots as evidence of historical climate change: goldmine or landmine? Madrono 1372-378, KELLY, A. E. AND M. L. GOULDEN. 2008. Rapid shifts in plant distribution with recent climate change. Proceedings of the National Academy of Sciences 105:11823-11826. KELLY, M., B. ALLEN DIAZ, AND N. KOBZINA. 2005. Digitization of a historic dataset: the Wieslander California Vegetation Type Mapping Project. Madrono 52:191—201. . 2008. The Wieslander Vegetation Type Map- ping Webpage. http://vtm.berkeley.edu/. , K. I. UEDA, AND B. ALLEN DIAZ. 2008. Considerations for ecological reconstruction of historic vegetation: analysis of the spatial uncer- tainties in the California Vegetation Type Map dataset. Plant Ecology 194:37-94. LEICA GEOSYSTEM. 2004. ERDAS IMAGINE. Soft- ware. Norcross, GA. MINNICH, R. A., M. G. BARBOUR, J. H. BURKE, AND R. F. FERNAU. 1995. Sixty years of change in Californian conifer forests of the San Bernardino Mountains. Conservation Biology 9:902-914. AND R. J. DEZZANI. 1998. Historical decline of coastal sage scrub in the Riverside-Perris Plain, California. Western Birds 29:366—391. NATIONAL CLIMATE DATA CENTER. 2005. United States Historical Climatology Network (USHCN). Available at: http://www.ncdc.noaa.gov/oa/climate/ research/ushcn/ushen.html SAMPSON, A. W. AND B. S. JESPERSEN. 1963. California range brushlands and browse plants. University of California, Division of Agriculture and Natural Resources, Oakland, CA. SAWYER, J. O. AND T. KEELER-WOLF. 1995. A manual of California vegetation. California Native Plant Society, Sacramento, CA. SCHWIND, B. AND H. GORDON. 2001. CALVEG | geobook: a comprehensive information package describing California’s wildland vegetation, ver- sion 2. U.S.D.A. Forest Service, Pacific South- | west Region, Remote Sensing Lab, Sacramento, | CA. SNYDER, J. P. 1982. Map projections used by the U.S. | Geological Survey. Geological Survey Bulletin | 1532. United States Government Printing Office, — Washington, DC. TAYLOR, A. H. 2000. Fire regimes and forest changes in mid and upper forests of the southern Cascades, Lassen Volcanic National Park, California, U.S.A. Journal of Biogeography 27:87—104. TAYLOR, R. S. 2004. A natural history of coastal sage | scrub in southern California: regional floristic patterns and relations to physical geography, how — it changes over time, and how well reserves — represent its biodiversity. Ph.D. dissertation. Uni- | versity of California, Santa Barbara, CA. | THORNE, J. H., R. KELSEY, J. HONIG, AND B. MorGAN. 2006. The development of 70 yr old — Wieslander Vegetation Type Maps and an assess- — ment of landscape change in the central Sierra Nevada. California Energy Commission. CEC 500- | 2006-107. PIER Energy Related Environmental Program, Sacramento, CA. | UNITED STATES DEPARTMENT OF COMMERCE, COAST ] AND GEODETIC SURVEY. 1917. Geodesy: tables for a polyconic projection of maps, based upon © 2008] Clarke’s reference spheroid of 1866. Fourth edi- tion. Special Publication No. 5. Department of Commerce, Washington Government Printing Of- fice, Washington, DC. UNITED STATES DEPARTMENT OF COMMERCE. 2001. TIGER\Line files, roads. Department of Com- merce, Washington, DC. UNITED STATES CENSUS BUREAU. 2006. State and County Quick Facts. Census Bureau, Washington, DC. Available at: http://quickfacts.census.gov/qfd/ states/06/06017.html VEBLEN, T. T. AND D. C. LORENZ. 1991. The Colorado Front Range: a century of ecological change. University of Utah Press, Salt Lake City, UT. WAcOM. 2004. Digitizing screen that uses digital pen. Wacom Technology, Vancouver, WA. WALKER, R. E. 2000. Investigations in vegetation map rectification, and the remotely sensed detection and measurement of natural vegetation changes. Ph.D. dissertation. University of California, Santa Bar- bara, CA. WEEKS, D., A. E. WIESLANDER, AND C. L. HILL. 1934. The utilization of Eldorado County Land. Univer- sity of California Bulletin 572. Giannini Founda- tion, University of California, Berkeley, CA. THORNE ET AL.: SIXTY YEARS SIERRA NEVADA LANDSCAPE CHANGE P| : , H. R. JOSEPHON, AND C. L. HILL. 1943. Land utilization in the northern Sierra Nevada. Agricultural Experiment Station, University of California College of Agriculture, Berkeley, CA. WIESLANDER, A. E. 1935a. A vegetation type map for California. Madrono 3:140—-144. 1935b. First steps of the forest survey in California. Journal of Forestry 33:877—-884. . 1935c. The Forest Survey in California. R publication Series. California Forest and Range Experiment Station, Berkeley, CA. 1986. A.E. Wieslander, California forester: mapper of wildland vegetation and soils (an oral history conducted in 1985 by Ann Lange). Region- al Oral History Office, Bancroft Library, Univer- sity of California, Berkeley, CA. AND H. A. JENSEN. 1946. Forest areas, timber volumes, and vegetation types in California. Forest Survey Rel. No. 4. Forest and Range Experiment Station, Berkeley, CA. : , AND H. S. YATES. 1933. California vegetation type map: Instructions for the preparation of the vegetative type map of California. Unpublished USDA Forest Service report on the file in library at Yosemite National Park, Yosemite Valley, CA. MADRONO, Vol. 55, No. 3, pp. 238-243, 2008 A NEWLY DESCRIBED SPECIES OF ARCTOSTAPHYLOS (ERICACEAE) FROM THE CENTRAL CALIFORNIA COAST MICHAEL C. VASEY Department of Biology, San Francisco State University, 1600 Holloway Avenue, San Francisco, CA 94132 Department of Environmental Studies, University of California, Santa Cruz, 1156 High Street, Santa Cruz CA 95064 mvasey@sfsu.edu V. THOMAS PARKER Department of Biology, San Francisco State University, 1600 Holloway Avenue, San Francisco, CA 94132 ABSTRACT A newly described local endemic species, Arctostaphylos ohloneana M.C. Vasey and V.T. Parker, is found scattered within populations of another geographically restricted manzanita species, Arctostaphylos glutinosa Schreiber, from the “‘Lockheed Chalks” area on siliceous shale ridges, northern Ben Lomond Mountain, western Santa Cruz County. This species is found in at least four scattered occurrences within the distribution of A. glutinosa on private property owned by the Lockeed Martin Corporation at the end of Empire Grade Road. Arctostaphylos ohloneana superficially resembles A. pungens and A. manzanita, but it presents distinctive characters that separate it from these two species. Unlike the tetraploid A. manzanita, A. ohloneana is diploid, and it lacks the distinctive nascent inflorescence of A. pungens. Since neither A. pungens nor A. manzanita occurs in the Santa Cruz Mountains, A. ohloneana is all the more remarkable by virtue of its distinctiveness compared to other nearby species. Key Words: Arctostaphylos, Arctostaphylos ohloneana, central California coast, endemism, Ohlone manzanita, Santa Cruz Mountains. During the past 150-plus years, over 100 taxa in the shrubby genus Arctostaphylos have been recognized by a wide array of authors ranging from Arctostaphylos uva-ursi (L.) Spreng. origi- nally described by Linnaeus in 1700 to recent descriptions of new species, Arctostaphylos gabi- lanensis V.T. Parker and M.C. Vasey (Parker and Vasey 2004) and other new taxa (Keeley et al. 2007; Parker et al. 2007). The vast majority of manzanita taxa (~93%) are situated in California (Markos et al. 1999; Wells 2000; Boykin et al. 2005) where they exhibit an array of different habits (prostrate to arborescent shrubs) and occupy a wide diversity of habitats on nutrient- limited soils that are distributed in many different geographic settings. Although Arctostaphylos fossils date back to the middle Miocene (Stebbins and Major 1965; Raven and Axelrod 1978; Edwards 2004), diversification within Arctostaph- ylos is hypothesized to have occurred in the relatively recent past (Stebbins 1971; Raven and Axelrod 1978), perhaps as recently as the latest Pleistocene and Holocene. Various processes have most likely interacted in driving this remarkable radiation. Recent climate fluctuations (Raven and Axelrod 1978), intensifying fire regimes, evolution of the obligate seeder life history (Wells 1969; Keeley and Zedler 1978; Parker and Kelly 1989), and hybridization between species are likely to have played signif- icant roles (Wells 2000). Concentrations of manzanita endemism along the summer fog- moderated California coast (Cody 1986; Keeley 1992) suggest that a favorable water-energy balance may also be important (Richerson and Lum 1980; O’Brien 1998; Francis and Currie 2003). In this paper, we recognize a local endemic Arctostaphylos species from a _ fog-influenced coastal area in the Santa Cruz Mountains. THE OHLONE MANZANITA In the late 1980s, Randy Morgan, a consulting biologist, discovered an undescribed manzanita while surveying a unique area of private property owned by the Lockheed Corporation called the ‘‘Lockheed Chalks” located at the end of Empire Grade Road about 3 km southwest of Eagle Rock on Ben Lomond Mountain in western Santa Cruz County (Fig. 1). The property encompasses several hundred acres of siliceous shale ridges that constitute the watershed divide between the Scott Creek, Mill Creek, and Boyer Creek drainages that lead into the nearby Pacific Ocean approximately 6 km away. Ridgecrests at the Lockheed Chalks are dominated by a knobcone pine-maritime chaparral community that grades into redwood-tan oak forest in the 2008] Eagle Rock Lockhead Martin N Facility Empire Grade 37°08'24” Ried . . Mill-Cr * (Boyer Cr Type Locality SY lo te. Reservoir 122°13'12" 122 1216" Fic. 1. VASEY AND PARKER: ARCTOSTAPHYLOS OHLONEANA 239 @San Francisco 37°25’ e San Jose Gregorio Santa Cruz 37°00) 123-0: 122°0’ Locations for four known occurrences of Arctostaphylos ohloneana. Total population is found within 4 km? area on private property owned by Lockheed Martin Corporation at the northern end of Ben Lomond Mountain on Monterey shale ridges that drain into the Scott Creek watershed. Populations are at an elevation of about 500 m within 6 km of the ocean. upper arroyos (Fig. 2). The chaparral component of this area is rich in manzanita species, including Arctostaphylos crustacea Eastwood subsp. crinita (Gankin) V.T. Parker, M.C. Vasey, & J.E. Keeley, A. andersonii Gray, A. sensitiva Jepson, and the local endemic, A. glutinosa Schreiber. Morgan was surprised to find yet another manzanita mixed into this chaparral that is highly distinct from the other four taxa. As he continued his survey, he located two other small occurrences of the new manzanita in disjunct stands removed by several hundred meters from one another and well integrated into the local plant assemblage. Although difficult to access, the Lockheed Chalks is well known to manzanita savants because it is the type locality of Arctostaphylos glutinosa Schreiber, originally described by Beryl O. Schreiber based on a visit to this site in 1939 during a Weislander Vegetation Type Map project field survey sponsored by the U.S. Forest _ Service (Schreiber 1940). Two original photo- graphs of the site (#2145 and #2146) taken during this survey are available on-line at http:// www.lib.berkeley.edu/BIOS/vtm/search.html. Al- though a small stand of A. glutinosa extends beyond the Lockheed property limits onto the Swanton Ranch west of Lockheed (Roy Buck and Jim West, personal communication), the vast majority of the population is confined to the upland shale ridges encompassed by the Lock- heed property. Morgan contacted us to seek our interpretation of this manzanita. We were able to visit the Lockheed Chalks property on February 10, 1993. Morgan guided us to three local populations, and we examined individuals within these populations and, with the permission of Lockheed, made collections of flowering and vegetative material. Since our first visit, we have returned to this site on additional occasions to collect fruiting and nascent inflorescence material of Arctostaphylos ohloneana, to collect buds for chromosome analysis, to search for other populations, and to show the species to Jon Keeley, an Arctostaphylos expert from southern California. Keeley agreed that, despite the superficial similarity between the new manzanita and 4. pungens Kunth., the distinctive nascent inflorescence character distin- guishing A. pungens is absent in the new entity (Fig. 3). Immature buds of the new manzanita were sent to Kristina Schierenbeck at California State University Chico for a chromosome determina- tion. Schierenbeck got clear resolution for a chromosome count, which was determined to be n = 13, the diploid condition for Arctostaphylos. The diploid chromosome count further separates this entity from other similar species, such as the tetraploid A. manzanita, which occurs to the FIG. 2. MADRONO Habitat of Arctostaphylos ohloneana on siliceous-shales substrates in knobcone pine — maritime chaparral assemblage. Four other species of Arctostaphylos dominate this shrub matrix including A. crustacea subsp. crinita, A. sensitiva, A. glutinosa, and A. andersonii. interior in the Diablo Range, and north into the North Coast Ranges, considerably distant from the A. ohloneana population. Further, molecular analysis demonstrates that A. ohloneana has a distinctive ITS sequence that puts it in the smaller of two ITS clades (Boykin et al. 2005; Wahlert 2005). We are still investigating candidate species that are likely to be the closest living relatives to A. ohloneana, but our preliminary view is that A. ohloneana most likely represents a paleoendemic species that has been able to persist in the nutrient-poor Monterey shale barrens of north- ern Ben Lomond Mountain. Despite searches in the surrounding region, we have not been able to locate any other popula- tions of A. ohloneana other than one additional occurrence on Lockheed Chalks property bring- ing the total number of occurrences to four (Fig. 1). Although relatively few in estimated number (ca. 100 individuals), A. ohloneana 1s variable in leaf size and habit, suggesting the possibility of limited hybridization with adjacent taxa. Surprisingly, Schreiber, the botanist with the Vegetation Type Map project who described A. glutinosa, did not notice this new manzanita while sampling this area in 1939. On the other hand, A. glutinosa individuals are far more abundant than the new manzanita, and it is conceivable that Schreiber would not have discovered the much rarer new manzanita if it were roughly as abundant throughout the Lock- heed Chalks property in the late 1930’s as it is today (particularly since it tends to be located down slope rather on the ridge crests). Accord- ingly, on the basis of its distinctiveness, we are recognizing the new manzanita from the Lock- heed Chalks as Arctostaphylos ohloneana M.C. Vasey and V.T. Parker. TAXONOMIC TREATMENT OF ARCTOSTAPHYLOS OHLONEANA Arctostaphylos ohloneana M.C. Vasey and V.T. Parker, sp. nov. — TYPE: USA, California, Santa Cruz County, northern Ben Lomond Mountain, siliceous shale ridges located on slopes above Mill Creek drainage as part of the Lockheed Martin missile test site facility; 37°07 09.635" Ne 122°13'02.55" W;. 51> ms February «10%, 1993s Michael Vasey and Thomas Parker O111 (holo- type CAS; isotypes UC and SFSU). Additional small populations occur at 37°06'56.50” N, 122°1273 182" -W, S00mm 37°07220 SNe 122°13'06:76" W, ..528%m,. and .37°05'59: 5) Ne 122°12’05.6”, 451 m. Maritime chaparral associ- ates include A. crustacea, A. sensitiva, Adenos- toma fasciculata, Ceanothus cuneatus, Vaccinium ovatum, Pinus attenuata. 2008] FIG. 3. VASEY AND PARKER: ARCTOSTAPHYLOS OHLONEANA Arctostaphylos ohloneana in flower (a.) and with immature inflorescence (b.). Note the paniculate immature inflorescence, the round-ovate leaves, glabrous pedicels, and scale-like bracts. Leaves are green with stomata equally distributed on both leaf surfaces. Not shown are occasional long gland-tipped hairs found on twigs that are otherwise covered by short, dense hairs. Frutex erectus, 2—3 m. altus; lignotuber absens; ramorum cortex levis, ruber; ramulorum tricho- mata densa, brevia, et trichomatibus sparsa, longa, glanduliferis; laminis foliorum late ellipti- cis vel ovatus, circa 3 cm longis, 1—1.5 cm latis, viridis, stomatibus aequalibus supra et infra, apice plerumque mucronatis; inflorescentia pani- culata plerumque 3-5 ramosa, laxa, pedunculus planus, bracteae deltoidae, apiculus marcescens, trichomatibus glanduliferis; pedicelli glabrescens, corolla alba vel roseo-alba; ovarium glabrum; fructus depressus-globosus, 5-8 mm diametro, glaber, aurantiaco-ferrugineus; pyrenae librae. Erect to spreading shrub, 1—2 m high; stems with burl absent; bark red-brown, smooth; twigs covered in short, dense non-glandular hairs often with occasional long, gland-tipped hairs; /eaves green, surfaces alike, stomata equally dense on both surfaces, leaf surfaces sparsely short pubes- cent; ca. 3 cm long, 1—1.5 cm wide, elliptic to round-ovate in shape, tips mucronate, base rounded; petioles prominent, ca. 5 mm _ long; immature inflorescence 3—5 branched open panicle with somewhat flattened peduncles supporting terminal clusters of buds subtended by scale-like, appressed to mildly spreading, awl-shaped, mar- cescent bracts 1-2 mm long; bracts sparsely ciliate with some gland-tipped hairs; flowers conical-urceolate, whitish-pink supported by gla- brous pedicels; ovary glabrate, base of anther filaments glabrous; fruit depressed-globose, red- dish-brown, 5-8 mm wide, pyrenes separable: n=13 (per K. Schierenbeck). The specific epithet honors people of Native American Ohlone ancestry who occupied the Santa Cruz Mountains before European coloni- zation. We also honor the memory of C.H. Merriam, a pioneering ecologist who defied the nomenclatural tradition of his time by naming two species of manzanitas, A. mewukka and A. nissenana, in honor of Native American people he observed during his travels in the late 1800's. Table | provides a comparison of A. ohloneana morphology with three species to which it has been tentatively assigned in the past. Note that A. ohloneana is somewhat intermediate between 4A. manzanita C. Parry (a widespread arborescent shrub of the inner northern California coast range and Sierra Nevada foothills) and A. pungens (a widespread multi-stemmed shrub of the interior mountains of southern California, Arizona, Baja California, and mainland Mexico). It is quite different from the nearby A. hookeri G. Don in southern Santa Cruz and northern Monterey. Morphologically, A. ohloneana 1s more similar to A. manzanita than A. pungens, particularly in terms of its leaves, immature inflorescence and bracts. A unique character for A. ohloneana is the occasional long gland-tipped hairs that typically are found intermixed with short dense hairs on its twigs. Given the richly diverse assemblage of Arcto- staphylos taxa that occur in the Santa Cruz Mountains south of the San Gregorio watershed (central and southern region), we provide a Key to the manzanitas of this subregion that includes A. 242 MADRONO [Vol. 55 TABLE |. Comparison between Arctostaphylos ohloneana and three representative species with similar morphology. A. manzanita A. hookeri A. ohloneana subsp. manzanita subsp. hookeri A, pungens Ploidy diploid tetraploid diploid diploid Twig hairs occassional gland-tipped no glandular hairs no glandular hairs no glandular hairs hairs Leaf Hue green green bright green green Leaf Length 3-4 cm 3-5 cm 2-3 cm 1.5-4 cm Leaf Shape elliptic to round ovate ovate to obovate elliptic elliptic Immature Inflorescence panicle 3—5 branch panicle 5—7 branch raceme-1 branch raceme-1 branch Floral Bracts appressed to spreading, appressed, keeled scales spreading, awl-tipped recurved, awl-tipped scales scales awl-tipped scales Fruit width 5-8 mm 8—12 mm 4-6 mm 5-8 mm ohloneana and illustrates its distinctiveness com- small area includes twelve taxa of which five are pared to other nearby Arctostaphylos taxa. The local endemics. Nomenclature follows the new geographic area covered is approximately 1000 treatment of Arctostaphylos prepared for the 274 square miles (2590 km’). Note that this relatively Edition of the Jepson Manual. KEY TO ARCTOSTAPHYLOS IN THE CENTRAL AND SOUTHERN SANTA CRUZ MOUNTAINS 1. Plants with burls or woody platforms, usually resprouting after a fire. 2. Leaves with upper and lower surfaces appearing different in color, hairiness, and stomatal distribution; twigs typically with long stiff hairs and short dense tomentum. 3. (Lower leaf surfaces clabrous or sparsely Nairy ex-4 a 405s we be ak a ee vat gc Caos Can eyes ee A. crustacea Eastw. (Widespread member of chamise chaparral in the interior uplands on the east side of the range from Castle Rock Ridge to Mt. Madonna) (Thomas 1961). 3’ Lower leaf surfaces densely hairy, sometimes also upper leaf surfaces although less so (as on Ice Cream, CFTAGE)s oc & hee ae oy a he Se pe et ne a a es ee ee ee A. crustacea subsp. crinita V.T. Parker, M.C. Vasey, J.E. Keeley (Locally abundant in maritime chaparral on coastal uplands on west side of range from Butano Ridge south to hills near Soquel). 2’ Leaves with upper and lower surfaces essentially alike, similar in color and hairiness, with at least half as many stomata above as below; twigs short-hairy but lacking long stiff hairs .............. A. glandulosa Eastw. subsp. cushingiana Eastw. (Individuals from near the summit of Loma Prieta Peak may key to this subspecies. Generally, A. glandulosa is not found in the Santa Cruz Mountains). 1’ Plants lacking burls or woody platforms, not sprouting after fire. 4. Plants with grey-hued leaves. 5. Leaves smooth and waxy, lacking hairs on either surface; fruits spheric with pyrenes fused into a SOlld'S(ONG % 4.460 e ao See oak eh ees SRA Se See ee ee ee .. .A. glauca Lindley (serpentine chaparral on southern and eastern interior foothills of range). 5’ Leaves hairy on both surfaces, often densely; fruits depressed-spheric with pyrenes separable. 6. Leaf bases deeply lobed, apparently clasping twigs, twig and inflorescence hairs long and tipped with glands........... A. glutinosa Schreiber (Local maritime chaparral endemic of Lockheed Chalks growing on silicious shales on northern end of Ben Lomond Mountain). 6’ Leaf bases rounded to cuneate, not deeply lobed, lacking long glandular hairs. 7. ‘Ovaries: and fruus Jacking haits) cuca. 2 sees oe no i ee we ee ee A. silvicola Jeps. and Weisl. (Local endemic of maritime chaparral on ‘sand hill’ Santa Margarita sandstone formation in the southern part of range from Bonnie Doon to Felton). 7’ Ovaries and fruits densely hairy. 8. “Hairs on-pedicels and ovaries lacking plands«:..2.2%.c «de. p: pes Se se eee ee bee breed apa, Bataan, Bah fy ewan Straten Eocene A, canescens Eastw. (Southern disjunct popu- lation in chaparral along granite ridges between Loma Prieta and Mount Madonna). 8’ Hairs on pedicels and ovaries tipped with glands............ A, canescens subsp. sonomensis (Eastw.) Wells (Several individuals of this subspecies were found intermixed with main population along Summit Road in southern range near Loma Prieta Peak). 4’ Plants with green-hued leaves. 9. Leaves with distinct basal lobes, +/— clasping twigs, petioles obscure, leaves oblong in shape, ca. 4-7 cmlong........ A. andersonii Gray (Tall shrub to small tree in chaparral, often at forest edges, local endemic of central and southern Santa Cruz Mountains, mostly in redwood zone). 9’ Leaves with truncate or rounded base, petioles obvious, leaves oblanceolate, elliptic, or oval in shape, less than 3 cm long. 10. Leaves with upper and lower surfaces different, dark green above, light green below, stomata lacking on upper surface, flower parts (calyx and corolla) 4-merous.................. 2008] VASEY AND PARKER: ARCTOSTAPHYLOS OHLONEANA 243 ee ate: A. sensitiva Jeps. (In maritime chaparral on uplands near the coast in a variety of soil types, from Butano Ridge to Mount Herman). 10’ Leaves with stomata distributed equally on both leaf surfaces, leaf-hue similar on each surface, flower parts (calyx and corolla) 5-merous. 11. Immature inflorescence generally a compact, globose raceme terminated by a cluster of acuminate, spreading bracts; leaves bright green and lustrous; twigs lacking any long, glandular hairs :..2 2 6.22 .se5 4. es .. A. hookeri G. Don subsp. hookeri (A Monterey Bay maritime chaparral endemic, found in the Santa Cruz Mountains only in the south- western coastal foothills near Soquel, ranging farther south in low hills to Carmel Bay). 11’ Immature inflorescence a 3—5 branched open panicle with somewhat flattened peduncles supporting terminal clusters of buds subtended by scale-like, appressed, awl-shaped bracts; leaves dull green, not lustrous; twigs with occasional long gland-tipped hairs.......... de ended utah ae eee A. ohloneana M.C. Vasey and V. T. Parker (Local endemic of the coastal Lockheed Chalks in knobcone pine-maritime chaparral at the northern end of Ben Lomond Mountain, western Santa Cruz Mountains). ACKNOWLEDGMENTS We thank our colleague, Jon Keeley, for his examination of A. ohloneana in the field as well as his review and constructive comments on this manuscript. We also thank Kristina Schierenbeck for her assistance in determining the ploidy level for A. ohloneana. Bob Patterson was instrumental in helping with the latin descriptions and editorial comments. We also very much appreciate Randy Morgan for his botanical skills, discerning eye and invitation to examine the new manzanita at the Lockheed Chalks facility. We also appreciated the hospitality of Mr. David Murphy of Lockheed Martin Corporation who arranged for our visits and accompanied us during our field trips. Bret Hall of the UC Santa Cruz Herbarium has provided recent assistance to help provide access to the site. LITERATURE CITED BOYKIN, L. M., M. C. VASEY, V. T. PARKER, AND R. PATTERSON. 2005. Two lineages of Arctostaphylos (Ericaceae) identified using the internal transcribed spacer (ITS) region of the nuclear genome. Madrono 52:139-147. Copy, M. L. 1986. Diversity, rarity and conservation in Mediterranean-climate regions. Pp. 122-152 in M. E. Soulé (ed.), Conservation biology: the science of scarcity and diversity. Sinauer Associates, Sunderland, MA. EDWARDS, S. 2004. Paleobotany of California. The Four Seasons 12:1—75. FRANCIS, A. P. AND D. J. CURRIE. 2003. A globally consistent richness-climate relationship for angio- sperms. American Naturalist 161:523—536. KEELEY, J. E. 1992. A Californian’s view of fynbos. Pp. 372-378 in R. M. Cowling (ed.), The ecology of fynbos: nutrients, fire, and diversity. Oxford University Press, Capetown, South Africa. AND P. H. ZEDLER. 1978. Reproduction of chaparral shrubs after fire: comparison of sprout- ing and seeding strategies. American Midland Naturalist 99:142-161. , M. C. VASEY, AND V. T. PARKER. 2007. Subspecific variation in the widespread burl- forming Arctostaphylos glandulosa. Madrono 54:42-62. MARKOS, S. E., V. T. PARKER, L. HILEMAN, AND M. C. VASEY. 1999. Phylogeny of the Arctostaphylos hookeri complex (Ericaceae) based on nrDNA sequence data from the ITS region. Madrono 45:187-199. O’BRIEN, E. M. 1998. Water-energy dynamics, climate, and prediction of woody plant species richness: an interim general model. Journal of Biogeography 25:379-398. PARKER, V. T. AND V. R. KELLY. 1989. Seed banks in California chaparral and other Mediterranean climate shrublands. Pp. 231—255 in M. A. Leck, V. T. Parker, and R. L. Simpson (eds.), Ecology of soil seed banks. Academic Press, New York, NY. AND M. C. VASEY. 2004. Arctostaphylos gabilanensis, a newly described auriculate-leaved manzanita from the Gabilan Mountains, Califor- nia. Madrono 51:322—325. ‘ , AND J. E. KEELEY. 2007. Taxonomic revisions in the genus Arctostaphylos (Ericaceae). Madrono 54:148—155. RAVEN, P. H. AND D. I. AXELROD. 1978. Origin and relationships of the California flora. University of California Publications in Botany 72:1—134. RICHERSON, P. J. AND K. L. LUM. 1980. Patterns of plant species diversity in California: relation to weather and topography. American Naturalist 116: 504-536. SCHREIBER, B. O. 1940. The Arctostaphylos canescens complex. American Midland Naturalist 23: 617-631. STEBBINS, G. L. 1971. Chromosomal variation in higher plants. Arnold, London, United Kingdom. AND J. MAJor. 1965. Endemism and speciation in the California flora. Ecological Monographs 35:1-35. THOMAS, J. H. 1961. Flora of the Santa Cruz Mountains of California: a manual of the vascular plants. Stanford University Press, Stanford, CA. WAHLERT, G. A. 2005. A phylogeny of Arctostaphylos (Ericaceae) inferred from ITS sequence data. M.S. thesis, San Francisco State University, San Fran- cisco, CA. WELLS, P. V. 1969. The relation between mode of reproduction and extent of speciation in woody genera of the California chaparral. Evolution 23:264-267. ———. 2000. The manzanitas of California, also of Mexico and the world. Published by the author. MADRONO, Vol. 55, No. 3, pp. 244-247, 2008 NOTES ON TWO SOUTHERN AFRICAN ARCTOTIS SPECIES (ARCTOTIDEAE: ASTERACEAE) GROWING IN CALIFORNIA ALISON M. MAHONEY Department of Biology, Minnesota State University — Mankato, 242 Trafton Science Center S, Mankato, MN 56001 alison.mahoney@mnsu.edu ROBERT J. MCKENZIE Molecular Ecology and Systematics Group, Department of Botany, Rhodes University, P.O. Box 94, Grahamstown, 6140, South Africa ABSTRACT A literature review and determination of specimens performed in conjunction with treatment preparations for the Flora of North America North of Mexico and the second edition of The Jepson Manual indicates that the names in use for two Arctotis species (Arctotideae: Asteraceae) occurring in California need updating. Venidium fastuosum (Jacq.) Stapf, a rare escape from cultivation, should be Arctotis fastuosa Jacq. (Pl. hort. schoenbr. 2: 20, pl. 166; 1797) and A. venusta Norl. (Bot. Not. 118: 406-7; 1965) is the correct name for a naturalized species previously determined as A. stoechadifolia P.J. Bergius. Key Words: Arctotideae, Arctotis, Asteraceae, naturalization, Venidium. The only members of tribe Arctotideae (Aster- aceae) with established populations in North America occur in California and New Mexico (Mahoney 2006). All are southern African species introduced through horticulture. Among them, only one species is potentially invasive; the others range from occasional to naturalized garden escapes with limited distributions. Munz and Keck (1973) and McClintock (1993) treated four species: Gazania linearis (Thunb.) Druce [G. longiscapa DC.], Arctotheca calendula (L.) Le- vyns, Venidium fastuosum (Jacq.) Stapf, and Arctotis stoechadifolia P.J. Bergius. Naturalized populations of G. /inearis (treasure- flower) occur along roadsides, especially in urban coastal areas of Santa Barbara and Los Angeles Counties. Entities treated previously as Arctotheca calendula (capeweed) (Munz and Keck 1973; McClintock 1993) actually consist of two distinct species. Arctotheca prostrata (Salis.) Britten is a sterile perennial that spreads aggressively by prostrate stems along roadsides and in other disturbed sites in the North Coast (NCo), South Coast (SCo), Central West (CW), and Western Transverse Ranges (WTR) floristic province subregions of California. Arctotheca calendula, a fertile annual classified as invasive by the Cali- fornia Exotic Pest Plant Council (Brossard et al. 2000), occurs in a few coastal and disturbed urban habitats in the NCo, Central Coast (Cco), and Outer South Coast Ranges (SCoRO) subregions. The present paper is confined to updating names and identities of Venidium fastuosum (monarch-of-the-veld) and Arctotis stoechadifolia (blue-eyed African daisy). Venidium fastuosum The genus Arctotis is characterized principally by achenes with three + well-developed abaxial wings that create one or two distinct furrows or “cavities” (McKenzie et al. 2005). Venidium was segregated from Arctotis by Lessing (1831, 1832) based on the less well-developed achene wings, the absence or extreme reduction of pappus scales, and differences in achene pubescence. Beauverd (1915) and Lewin (1922) found the distinguishing characters used by Lessing unten- able and transferred Venidium species into Arctotis, but Stapf (1926) disagreed and trans- ferred A. fastuosa Jacq. to Venidium. Use of the name V. fastuosum has persisted in U.S. floral treatments (Munz and Keck 1973; Liberty Hyde Bailey Hortoritum 1976; McClintock 1993; Quat- trocchi 2000; Calflora 2007; USDA 2007). A morphological study of achenes and initial molecular analyses of subtribe Arctotidinae (McKenzie et al. 2005, 2006; Funk et al. 2007) support Beauverd’s and Lewin’s observations for some of the species previously placed in Veni- dium. Results from a larger molecular study of | Arctotidinae (McKenzie and Barker 2008) indi- cate that although Arctotis s.1. is polyphyletic and needs redefining, California’s Arctotis species will retain their names. Based on these new data, the name Arctotis fastuosa Jacq. should be adopted for the species previously known as Venidium fastuosum. Arctotis fastuosa is a hirsute to lanate, tap- rooted annual with heads up to 10cm in | diameter. Its rays are bright orange to yellow (a white-rayed form is in cultivation), usually — 2008] marked basally with a purple-black band; its discs are yellowish-brown. The species is native to the semi-arid, winter-rainfall Namaqualand re- gion of South Africa. Stapf (1926) referred to A. fastuosa as a weed of cornfields in its native range, where it is now common on roadsides, but in California it very rarely escapes from cultivation and does not appear to have natural- ized. We are not aware of the species being a problematic weed in any country. McKenzie et al. (2005) found that herbarium specimens determined as A. fastuosa can be divided into at least three distinct entities based on achene morphology. We have been able to examine achenes from two Californian specimens (Fuller 8191, CDA; R. Whitaker s.n., RSA); in both, achenes lack pappi, pubescence, and basal tufts of hairs and are readily referable to typical A. fastuosa. Specimens examined. USA, CA, Riverside Co., Thermal, establishing itself, 10 March 1949, Whitaker s.n. (RSA), 5.2 mi S of Beaumont, roadside sand, 17 February 1960, Fuller 3609 (CDA); San Bernardino Co., San Bernardino, NE of Central Ave. & 3" St, Gate 5, Norden AFB, waste ground, 11 April 1962, Fuller 8191 (CAS, CDA). Arctotis stoechadifolia and A. venusta During a 1939-40 study of Arctotis specimens, Norlindh (1964) found that in the last revision of Arctotidinae, Lewin (1922) had misinterpreted A. stoechadifolia P.J. Bergius, a prostrate, mat- forming perennial occurring in open sandy, seasonally wet areas and stabilized dunes in the southwestern Cape. Lewin applied this name to an erect, tap-rooted annual occurring in a wider range of habitats and with a broader distribution in inland southern Africa (Norlindh 1964). The native ranges of the two entities do not overlap. Although Norlindh (1965) subsequently de- scribed and named the annual species A. venusta Norl., Lewin’s interpretation has been adopted widely. Arctotis venusta is a popular horticultural plant and is often grown under the name A. stoechadifolia, A. stoechadifolia var. grandis (Thunb.) Less. or A. grandis Thunb. Sometimes the names A. stoechadifolia and A. venusta are treated as synonyms (Brickell and Zuk 1996). Mahoney’s (2006) FNA treatment maintained the name A. stoechadifolia but commented on its probable misuse. A full investigation subsequent to preparation of that treatment and availability of further data have clarified the identity of the Californian specimens. Arctotis stoechadifolia and A. venusta are morphologically distinct species (Norlindh 1964). Fresh plants cannot be confused. Arctotis stoechadifolia is a mat-forming perennial that produces long (up to + 1m), adventitiously rooting, prostrate stems from which the erect MAHONEY AND MCKENZIE: TWO ARCTOTIS SPECIES IN CALIFORNIA 245 flowering shoots arise. Its ray florets are white or pale yellow; its discs are black. The silver-grey leaves have a dense, tightly appressed woolly tomentum on both the upper and lower surfaces. The outer involucral bracts have a rigid base and a 4-7 mm long, acuminate, linear-cylindrical apical appendage. The obovoid-obconical achenes are densely tomentose and bear strongly incurved, + entire lateral wings that partially conceal the abaxial ‘cavities.’ Arctotis venusta 1s an erect, tap-rooted, summer-flowering annual. Its ray florets are white with a narrow yellow band at the base of the limb; its discs are a distinctive greyish-purple. The leaves are usually more thinly and loosely tomentose. The outer involucral bracts have a soft base and a short, 1— 3 mm long appendage with a rounded or obtuse apex. The oblong-obconical achenes are + glabrous to sparsely tomentose on the abaxial and tangential surfaces and the oblong-obovate cavities are obvious. Herbarium specimens can be more difficult to distinguish due to faded florets and missing lower-stem or below-ground parts, but the differences in leaf pubescence, involucral- bract and achene features allow discrimination of the two species. Our determination of Californian specimens from horticulture and naturally-occurring popu- lations previously determined as A. stoechadifolia confirms that they belong to A. venusta. No specimens of A. stoechadifolia have been ob- served. Norlindh (1964) noted that A. venusta 1s weedy even in its native range, often occurring in cultivated fields or along roadsides, while A. stoechadifolia does not readily tolerate competi- tion so that some populations have disappeared due to urban sprawl and the planting of erosion- preventing trees and shrubs. Wells et al. (1986) classified A. venusta as a ruderal, agrestal and pastoral weed in South Africa. Stated undesirable characteristics include: that it may be poisonous to stock, it taints milk, and is a crop-seed contaminant. Arctotis venusta is predicted as highly likely to become a weed, and further- more to become an agricultural weed, in Aus- tralia (Scott and Panetta 1993). It appears that A. venusta has been cultivated in California for at least 150 years but in that time has not naturalized extensively. However, all but one of the Californian specimens of A. venusta we examined were collected before 1971; whether this reflects its rarity or a lull in active collecting is unclear. It is interesting that naturalized Califor- nian populations of A. venusta occur in coastal, sandy habitats that might be favored by A. stoechadifolia, even though A. venusta 1s an inland species occurring in rangeland habitats far from the coast in its native range in southern Africa. The true Arctotis stoechadifolia has become naturalized along parts of the coastline of 246 Australia (Jeanes 1999; Rippey and Rowland 2004; Barker et al. 2005) due to its popularity as a garden ornamental plant and its use as a dune stabilizer. It inhabits the fore-dune through to open sites in the protected hind-dune area, as in its indigenous range in South Africa. The species has been assessed to be a minor problem weed in natural ecosystems warranting control at four or more locations within a state or territory in Australia (Groves et al. 2003). In parts of South Australia, it has a high Weed Risk Assessment rating, as the dense mats smother smaller indigenous plants by shading and competition for resources and can alter dune shape (Cording- ley and Petherick, 2005). Therefore, if A. stoechadifolia is cultivated in California we recommend assessment of its weed risk poten- tial and vigilance for naturalized plants to ensure it does not become a problem species in California. Specimens examined. USA, CA, Los Angeles Co., Glendora, Alosta Ave. at Glendora Ave., 750 ft., dry roadside, 24 June 1933, Wheeler 1889 (CDA, RSA, UCR); [county name illegible, probably Los Angeles], Ocean Park, sand lots near beach, 2 August 1934, Cohen [?] 515 (RSA); Orange Co., 2 mi NE of Huntington Beach near N end of Bolsa Chica Salt Marsh, disturbed roadside, infrequent escape, 25 July 1970, Hen- rickson 5091 (RSA), Newport Bay, 10 ft., sandy flat just back of high tide, 21 October 1933, Wheeler 2224 (RSA); Santa Barbara Co., Santa Maria, N side of W Jones St E of S Curver [sic. probably Curryer] St., naturalized in waste ground, 21 May 1968, Fuller 17065 (CDA), Santa Maria, 100 E Main, vacant lot, 22 September 1965, Jones & Allen s.n. (CDA), Lauro Canyon Dam Project, roadside, 30 May 1952, Pollard s.n. (CDA), Hwy between Santa Maria and Orcutt, sandy soil along highway, 22 November 1960, Smith 6332 (RSA, SBBG, UCR); Ventura Co., Ojai, Canada St., 12 July 1946, Pollard s.n. (CAS). ACKNOWLEDGMENTS We thank David Keil (OBI) and Andrew Sanders (UCR) for insights and assistance; the curators of the CAS, CDA, RSA, SBBG, SD and UC herbaria for the loan of, or access to, specimens; and the National Research Foundation of South Africa for a Postdoc- toral Fellowship to R. J. McK. LITERATURE CITED BARKER, W. R., R. M. BARKER, J. P. JEPSON, AND H. P. VoONow (eds.) 2005. Census of South Australian vascular plants. 5° ed. Journal of the Adelaide Botanic Gardens Supplement |. Botanic Gardens of Adelaide and State Herbarium, Ade- laide, Australia. BEAUVERD, G. 1915. Contribution a lTétude des Composées (suite X). Bulletin de la Société botanique de Geneve (série 2) 7:21—56. MADRONO [Vol. 55 BRICKELL, C. AND J. D. ZUK (eds.) 1996. American Horticultural Society: A-Z encyclopedia of garden plants. Dorling Kindersley Publishing, New York, NY. BROSSARD, C. C., J. M. RANDALL, AND M. C. HOSHOVSKyY (eds.) 2000. Invasive plants of Cali- fornia’s wildlands. University of California Press, Berkeley, CA. CALFLORA, 2007. Information on California plants for education, research and conservation [web appli- cation]. The Calflora Database, Berkeley, CA. Available at: http://www.calflora.org. Accessed March 03, 2007. CORDINGLEY, S. P. AND C. E. PETHERICK. 2005S. Vegetation management plan for the Henley South and West Beach Dune Reserve. City of Charles Sturt, Adelaide, Australia. FUNK, V. A., R. CHAN, AND A. HOLLAND. 2007. Cymbonotus (Compositae: Arctotideae, Arctotidi- nae): an endemic Australian genus embedded in a southern African clade. Botanical Journal of the Linnean Society 153:1-8. GROVES, R. H., J. R. HOSKING, G. N. BATIANOFF, D. A. CooKE, I. D. COWIE, R. W. JOHNSON, G. J. KEIGHERY, B. J. LEPSCHI, A. A. MITCHELL, M. MOERKERK, R. P. RANDALL, A. C. ROZEFELDS, N. G. WALSH, AND B. M. WATERHOUSE. 2003. Weed categories for natural and agricultural ecosystem management. Bureau of Rural Sciences, Canberra, Australia. JEANES, J. A. 1999. Asteraceae. Vol. 4, Pp. 652-984 in Walsh N. G. and T. J. Entwistle (eds.). Flora of Victoria. Inkata Press, Melbourne, Australia. LESSING, C. F. 1831. Synanthereae Rich. Pp. 83-170 in L. C. A. von Chamisso and D. F. L. von Schle- chtendal (eds.). De plantis in expeditione specula- toria romanzoffiana observatis disserere. Linnaea, 6:76-170. . 1832. Synopsis generum compositarum ear- umque dispositionis novae tentamen, monogra- phiis multarum capensium interjectis. Duncker and Humblot, Berlin, Germany. LEWIN, K. 1922. Systematische Gliederung und geo- graphische Verbreitung der Arctotideae-Arctotidi- nae. Repertorium specierum novarum regni veget- abilis, Beihefte 11:1—75. | LIBERTY HYDE BAILEY HORTORIUM. 1976. Hortus III. MacMillan Publishing Co., New York, NY. MAHONEY, A. M. 2006. Arctotis. Vol. 19, Pp. 198-199 in Flora of North America Editorial Committee (ed.). Flora of North America North of Mexico. Oxford University Press, New York, NY. MCCLINTOCK, E. 1993. Arctotis, Pp. 200; Venidium, p. | 356 in J. C. Hickman (ed.). The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. MCKENZIE, R. J. AND N. P. BARKER. 2008. Radiation | of southern African daisies: Biogeographic infer- | ences for subtribe Arctotidinae (Asteraceae, Arcto- tideae). Molecular Phylogenetics and Evolution 49:1-16. | ——, E. M. MULLER, A. K. W. SKINNER, P. O. KARIS, AND N. P. BARKER. 2006. Phylogenetic relationships and generic delimitation in subtribe | Arctotidinae (Asteraceae: Arctotideae) inferred by | DNA sequence data from ITS and five chloroplast regions. American Journal of Botany 93:1222— 1235: 2008] , J. SAMUEL, E. M. MULLER, A. K. W. SKIN- NER, AND N. P. BARKER. 2005. Morphology of cypselae in subtribe Arctotidinae (Compositae- Arctotideae) and its taxonomic implications. An- nals of the Missouri Botanical Garden 92:569—594. Munz, P. A. AND D. D. KECK. 1973. A California flora (with supplement by P.A. Munz). University of California Press, Berkeley, CA. NORLINDH, T. 1964. On the identity of Arctotis stoechadifolia Berg. Svensk Botanisk Tidskrift 58:193—203. . 1965. Arctotis venusta T. Norl. spec. nova, an ornamental plant from southern Africa. Botaniska Notiser 118:403-411. QUATTROCCHI, U. 2000. CRC world dictionary of plant names, vol. 4. CRC Press, Boca Raton, FL. RIPPEY, E. AND B. ROWLAND. 2004. Coastal plants: Perth and the south west region. 2"¢ ed. University of Western Australia Press, Perth, Australia. MAHONEY AND MCKENZIE: TWO ARCTOTIS SPECIES IN CALIFORNIA 247 SCOTT, J. K. AND F. D. PANETTA. 1993. Predicting the Australian weed status of southern African plants. Journal of Biogeography 20:87—93. STAPF, O. 1926. Venidium fastuosum. Curtis’s Botanical Magazine 152, tab. 9127. UNITED STATED DEPARTMENT OF AGRICULTURE, NATURAL RESOURCES CONSERVATION SERVICE (USDA). 2007. The PLANTS database. National Plant Data Center, Baton Rouge, LA. Available at: http://plants.usda.gov/. Accessed March 3, 2007. WELLS, M. J., A. A. BALSINHAS, H. JOFFE, V. M. ENGELBRECHT, G. HARDING, AND C. H. STIRTON. 1986. A catalogue of problem plants in southern Africa, incorporating the National Weed List of Southern Africa. Memoirs of the Botanical Survey of South Africa No. 53. Botanical Research Institute and Department of Agriculture and Water Supply, Pretoria, South Africa. MADRONO, Vol. 55, No. 3, pp. 248—250, 2008 NOTES ON EARLY COLLECTIONS AND THE DISTRIBUTION OF THE RED ALGA CUMATHAMNION SYMPODOPHYLLUM MICHAEL J. WYNNE Department of Ecology and Evolutionary Biology and Herbarium, University of Michigan, Ann Arbor, MI 48109 mwynne@umich.edu ABSTRACT Specimens (in FH and PC) of Cumathamnion sympodophyllum M.J. Wynne and K. Daniels (Delesseriaceae, Rhodophyta) made by C. G. Pringle at Cape Mendocino, northern California in 1882 are the first known collections of this rare monospecific genus. A collection (US) made by E.Y. Dawson from Trinidad Head, Humboldt County, and reported in 1965 as a “‘very compact, short- bladed” form of Delesseria decipiens J. Agardh is re-determined to be C. sympodophyllum. Records of C. sympodophyllum from Oregon, Washington, and British Columbia (Canada) are based on one collection from British Columbia, which cannot be verified. This taxon is most likely restricted to very exposed lower littoral rocky outcrops of Sonoma, Mendocino, and Humboldt counties in California. Key Words: Cumathamnion sympodophyllum, biogeography, endemic, marine algae, Rhodophyta. A recent visit to the National Museum d’Histoire Naturelle, Paris (PC) allowed me to find an interesting algal specimen held in the Bornet and Thuret Herbarium. The specimen [filed in the Preridium folder as ‘‘106-2-7” with the number “TA 21844’’] bore the following label data: “Delesseria pleurospora Harv. Cape Men- docino, Cal. leg. C. G. Pringle’, and in a different hand: ‘‘Farlow 1883” and ‘‘Herbarium G. Thuret”, with a handwritten note, apparently by Bornet: “Je ne reconnais ni l’espéce ni le genre de cette algue.”’ I next requested that staff at the Farlow Herbarium check for any other material of this Pringle collection in their holdings, and they located one dated “1882” (Fig. 1). These specimens, with densely congested branching that is pectinately arranged in the final orders and with thickened, fleshy (cartilaginous), denuded primary axes, are clearly identifiable as Cu- mathamnion sympodophyllum M.J. Wynne & K. Daniels (Wynne and Daniels 1966), and they represent the earliest known collection of this taxon. Cyrus Guernsey Pringle was one of the most prolific plant collectors of all time, having collected and distributed more than 500,000 specimens that he collected from the United States, Canada, and Mexico (Davis 1936). In 1880, Pringle made his first trip to the Pacific Slope, working as a botanical collector for the American Museum of Natural History, making general collections for Asa Gray, and exploring the forests of the region for the United States Census Department. He collected extensively in the Pacific states and in Mexico between 1880 and 1909 (Stafleu and Cowan 1983). His biogra- phy by Davis (1936) contains the diaries of his collecting trips to Mexico (1885-1909) but not of his earlier trips to California. However, according to information at the on-line site Consortium of California Herbaria (http://ucjeps.berkeley.edu/ consortium/), Pringle collected in Mendocino County during the first two weeks of August, 1882. The record of Pteridium (or Delesseria) pleur- ospora does not appear to have entered into the algal literature for California or the Pacific coast (Anderson 1891, 1894; Abbott and Hollenberg 1976). Delesseria pleurospora (Harvey 1855) was described from New Zealand and was variously transferred to Membranoptera (Kuntze 1891), Pteridium (Agardh 1898), Hydrolapatha (Kuntze — 1898), and Nitophyllum (Laing 1927). I regard it as conspecific with Schizoseris dichotoma (Hook- er f. et Harvey) Kylin, a species with a distribution restricted to the Southern Hemi- sphere (Papenfuss 1964; Ricker 1987; Ramirez and Santelices 1991; Adams 1994). In 1965 Ken Daniels and I were identifying algal specimens that we had collected from Mendocino City in northern California and were planning to describe some of our specimens as representing a new genus and species. At that time E. Yale Dawson, Curator of Algae at the Smithsonian Institution, published a manual to © the marine algae of Humboldt County (Dawson, © 1965). He had taught a course at Humboldt State | College [now Humboldt State University] that summer, and it seemed possible that he may have | come across the same alga that we were proposing to name as new to science. So on October 25, 1965, I wrote to him, telling him that © our purported undescribed genus at first glance | resembled Delesseria decipiens but had a distinc- | tive sympodial development. I asked if he might have obtained such an alga in his Humboldt | 2008] 3cm Cab. faverss OH. CC Pry. /yve2. FIG. 1. Cumathamnion sympodophyllum. Collection made by C. G. Pringle in 1882 from Cape Mendocino, California. (FH). County collections that were not yet determined that matched our alga. He responded (October 29, 1965), saying that he had not come across anything other than Delesseria decipiens and Membranoptera multiramosa N.L. Gardner, “‘al- though there is the possibility that something else may have gotten into these collections unno- ticed.”’ As it turned out, that proved to be the case. A few years later I was on a visit to the Herbarium of the Smithsonian Institution check- ing the collections of Delesseriaceae. In the folder of Delesseria decipiens, | came upon material of Cumathamnion sympodophyllum that he had misidentified as D. decipiens, namely, a collection from Trinidad Head, E.Y. Dawson 25297 (US). Interestingly, in his manual under Delesseria decipiens Dawson (1965, p. 39) referred to ‘“‘some very compact, short-bladed forms with coarse axes occur on heavily beaten cliff faces at Trinidad Head’’, while the label data include this phrase: “‘An atypical form on heavily dashed cliff face”. The specimens are indeed very compact and short but clearly represent C. sympodophyl- lum. Specimens Examined U.S.A., California. Humboldt Co., Trinidad Head, '%4 mile north of the head opposite Coon Island, 12 July, 1965, leg. Burnett = Dawson 25297 (US); about % mile west of Trinidad Head, 12 July 1972, ‘‘on rocks w/in Postelsia & Lessoniopsis”, De Cew s.n. (HSC, MICH, UC 1601559). College Cove, north of Trinidad Head, 1 August 1975, De Cew s.n. (UC 1601560). North Bidwell Point, Elkhead, Trinidad, 26 Sept. 1990, De Cew s.n. (UC 1601557). Mendocino Co., Cape Mendocino, 1882, C. G. Pringle s.n. (FH, PC); Mendocino City, 4 June 1965, leg. Daniels, McLaughlin & McLaughlin = Wynne 292 WYNNE: DISTRIBUTION OF RED ALGA CUMATHAMNION SYMPODOPH YLLUM 249 (MICH, UBC, UC); 30 July 1965, Daniels & Wynne (Wynne 430), Type collection: Holotype (UC 1318217); isotypes (MICH, UBC); 6 July 1966, on rocks among Lessoniopsis littoralis, Wynne 777 (MICH, UC, US); 2 June 1973, Young 624 (AHFH 84533 in UC, MICH), epiphytic on Lessoniopsis littoralis. Point Cab- rillo, north of Mendocino City, 20 June 1974, De Cew s.n. (UC 1601561). Fort Bragg, May 1988 (NCC). Sonoma Co., south side of Horseshoe Cove, Bodega Marine Laboratory, July 1978, De Cew s.n. (UC 1601558). Collecting History Cumathamnion sympodophyllum was first de- scribed from a single locality, Mendocino City, Mendocino County, CA. Waaland (1973) report- ed it from “Botany Beach” [= Botanical Beach] near Port Renfrew, Vancouver Island, British Columbia, Canada. In their marine algal flora of California, Abbott and Hollenberg (1976) re- ferred to this taxon as “Rare, low intertidal, Vancouver I., Br. Columbia; and on rocks exposed to heavy surf, Mendocino Bay, Calif. (type locality)’. They did not cite any additional collections. Based on Waaland’s (1973) report from Vancouver Island, Cumathamnion has entered into the literature for British Columbia and Washington (Widdowson 1975; Scagel et al. 1986). It is included in an on-line ‘‘working list of rare marine algae” for the State of Washington (Mumford 2004). In their lists of marine algal flora of Oregon, Phinney (1977) and Hansen (1997) included Cumathamnion in a list of species that occur south and north of Oregon but have yet to be found within Oregon. Citing the Hansen (1997) publication, Guiry and Guiry (2007) indicate Cumathamnion as occurring in Oregon. Waaland (1973) reportedly deposited voucher specimens in the herbaria of the Department of Botany and that of the Friday Harbor Labora- tories, University of Washington, but no such voucher specimens are now present in either herbarium. In 1995 Waaland wrote to me that in an office move the specimens were lost prior to his having deposited them in either of those herbaria. In light of the lack of any specimens of C. sympodophyllum collected north of northern California, it seems prudent to reconsider the distribution of this rare alga. According to Silva (2004), the southern end-point in the distribution of C. sympodophyllum is Bodega Head 1n Sonoma County, northern California. According to Gab- rielson et al. (2006), the presence of C. sympodo- phyllum in the local area (of Oregon, Washing- ton, British Columbia, and southeast Alaska) “requires confirmation”. I conclude that con- firmed collections of Cumathamnion sympodo- phyllum exist only for sites in Sonoma, Mendo- cino, and Humboldt Counties of northern 250 California and that the earliest known collection of this taxon was made in 1882 by C. S. Pringle from Cape Mendocino. One possible reason why this species may be under-reported in its range is that it is restricted to extremely exposed, lower littoral rocky outcrops, co-occurring with the cumaphytic kelps Lessoniopsis littoralis and Postelsia palmaeformis. Such habitats can be safely visited only at times of very low tides and reduced surf activity. ACKNOWLEDGMENTS I thank the following persons who have assisted me in this study: Bruno de Reviers (PC), Genevieve Lewis- Gentry (FH), Michael Hawkes (UBC), Robert Sims (US), Kathy Ann Miller (UC), Thomas Mumford (Washington Dept of Natural Resources, Olympia, WA), Frank Shaughnessy (HSC), Megan Dethier (Friday Harbor Laboratories, WA), Chris Kjeldsen (Sonoma State University, CA), and Craig Schneider (Trinity College, Hartford, CT). I am also grateful to Esther McLaughlin, David McLaughlin, and Ken Daniels for bringing their algal samples to my attention in June of 1965, which led to the description of Cumathamnion by Daniels and me. I wish to acknowl- edge the constructive criticism offered by an anony- mous reviewer. LITERATURE CITED ABBOTT, I. A. AND G. J. HOLLENBERG. 1976. Marine algae of California. Stanford University Press, Stanford, CA. ADAMS, N. M. 1994. Seaweeds of New Zealand: an illustrated guide. Canterbury University Press, Christchurch, New Zealand. AGARDH, J. G. 1898. Species genera et ordines algarum, 3(3), Lund, Sweden. ANDERSON, C. L. 1891. List of California algae, with notes. Zoe 2:217-225. 1894. Some new and some old algae but recently recognized on the California coast. Zoe 4:358—362. DAvIs, H. B. 1936. Life and work of Cyrus Guernsey Pringle. Free Press Printing Co., Burlington, VT. DAwson, E. Y. 1965. Marine algae in the vicinity of Humboldt State College. Humboldt State College, Arcata, CA. GABRIELSON, P. W., T. B. WIDDOWSON, AND S. C. LINDSTROM. 2006. Keys to the seaweeds and seagrasses of southeast Alaska, British Columbia, Washington and Oregon. Phycological Contribu- tion Number 7. PhycoID. University of British Columbia, Vancouver, BC, Canada. GuiIRY, M. D. AND G. M. GuIRy. 2007. AlgaeBase version 4.2. World-wide electronic publication, National University of Ireland, Galway, Ireland, Available at: http://www.algaebase.org. Accessed September 21, 2007. HANSEN, G. I. 1997. A revised checklist and prelimi- nary assessment of the macrobenthic marine algae and seagrasses of Orgeon. Pp. 175-200 in T. N. Kaye, A. Liston, R. M. Love, D. L. Luoma, R. J. Meinke, and M. V. Wilson (eds.), Conservation and management of native flora and fungi. Native Plant Society of Oregon, Corvallis, OR. MADRONO [Vol. 55 HARVEY, W. H. 1855. Algae. Pp. 211-266 and plates 107-121 in J. D. Hooker (ed.), The botany [of] the Antarctic voyage of the H.M. Discovery ships Erebus and Terror, in the years 1839-1843, under the commend of Sir James Clark Ross. II. Flora Novae-Zelandiae. Part II Cryptogamia. Reeve, London, United Kingdom. KUNTZE, O. 1891. Revisio generum plantarum vascu- larium omnium atque cellularium multarum secun- dum leges nomenclaturae internationales cum enumeratione plantarum exoticarum in itinere mundi collectarum. Pars I & II. Arthur Felix, Leipzig, Germany. . 1898. Revisio generum plantarum vascularium omnium atque cellularium multarum secundum leges nomenclaturae internationales cum enumer- atione plantarum exoticarum in itinere mundi collectarum. Pars III. Arthur Felix, Liepzig, Germany. LAING, R. M. 1927. A reference list of New Zealand marine algae. Transactions of the New Zealand Institute 57:126—185. MUMFORD, T. 2004. Working list of rare marine algae. Washington Natural Heritage Program, Washing- ton State Department of Natural Resources, Olympia, WA, Available at: http://wwwl.dnr.wa. gov/nhp/refdesk/lists/algae.html. Accessed October 1, 2008. PAPENFUSS, G. F. 1964. Catalogue and bibliography of Antarctic and Sub-Antarctic benthic marine algae. Antarctic Research Series Vol. 1. American Geo- physical Union, Washington DC. PHINNEY, H. K. 1977. The macrophytic marine algae of Oregon. Pp. 93-115 in R. W. Krauss (ed.), The marine plant biomass of the Pacific Northwest Coast: a potential economic resource. Oregon State University Press, Corvallis, OR. RAMIREZ, M. E. AND B. SANTELICES. 1991. Catalogo de las algas marinas bentonicas de la costa temperada del Pacifico de Sudamérica. Monogra- fias Bioldgicas 5:1—437. RICKER, R. W. 1987. Taxonomy and biogeography of Macquarie Island seaweeds. British Museum (Nat- ural History), London, United Kingdom. SCAGEL, R. F., D. J. GARBARY, L. GOLDEN, AND M. W. HAWKES. 1986. A synopsis of the benthic marine algae of British Columbia, northern Wash- ington and southeast Alaska. Phycological Contri- butions. University of British Columbia, Vancou- ver, BC, Canada. SILVA, P. C. 2004. Distribution of California seaweeds. Fremontia 32(1), 18—26. STAFLEU, F. A. AND R. S. COWAN. 1983. Taxonomic literature: a selective guide to botanical publica- tions and collections with dates, commentaries and types. Vol. IV: P-Sak. Second edition. Bohn, Scheltema & Holkema, Utrecht, The Netherlands. WAALAND, J. R. 1973. New records of red algae from Washington and British Columbia. Syesis 6:269. WIDDowsoON, T. B. 1975. The marine algae of British Columbia and northern Washington: revised list and keys. Part II. Rhodophyceae (red algae). Syesis 7:143-186. WYNNE, M. J. AND K. DANIELS. 1966. Cumathamnion sympodophyllum, a new genus of the Delesseriaceae (Rhodophyta). Phycologia 6:13—28. Volume 55, Number 3, pages 181-250, published 30 January 2009 Zs: SUBSCRIPTIONS — MEMBERSHIP Membership in the California Botanical Society is open to individuals ($35 per year; family $40 per year; emeritus $27 per year; students $27 per year for a maximum of 7 years). Late fees may be assessed. Members of the Society receive MADRONO free. Institutional subscriptions to MADRONO are available ($70). Membership is based on a calendar year only. Life memberships are $750. Applications for membership (including dues), orders for subscriptions, and renewal payments should be sent to CBS c/o Jepson Herbarium, University of California, Berkeley, CA 94720-2465. Requests and rates for back issues, changes of address, and undelivered copies of Ma- DRONO 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, five 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 abbreviations 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 are encouraged to include the names, addresses, and e-mail addresses of two to four potential reviewers with their submitted manuscript. 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 included 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 MADRONO. Joint au- thors 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 manuscript have been used in other papers that are published, in press, submitted, or soon to be submitted elsewhere. SMITHSONIAN INSTITUT! ON LIBRARIES VOLUME 55, NUMBER4 © OCTOBER-—DECEMBER 2008 MADRONO A WEST AMERICAN JOURNAL OF BOTANY UM iti Wi 7 < ae re Re 4 BaP et ae ae ae a is, { wa Baits 4 of hi oe eye re aot “ONTENTS THE GENUS KOCHIA (CHENOPODIACEAE) IN NORTH AMERICA | Gein: Cniiand SUC. SQUGCISON eosin nen 251 A SPATIAL AND TEMPORAL INVESTIGATION OF ELEOCHARIS MACROSTACHYA AND ORCUTTIA TENUIS Michelle Clark, Richard Lis, Dean Fairbanks, Gnd Kristind SCHICTENDE CK vvcsccccccceccoseseceuse teehee ascssacessevecnsecsDibscssevsesesease Poy | AN ERIOPHYLLUM LANATUM (ASTERACEAE) HYBRID ZONE IN OREGON JORN S! MOOTING voscccccassevesscsseesnnsesedas Webs Sdt sd oi vcctv pA aed ovssssaeaclibenes 269 MORPHOLOGICAL AND CYTOLOGICAL EVIDENCE FOR HOMOPLOID HYBRIDIZATION IN IOCHROMA (SOLANACEAE) Stacey DeWitt Smith, Vanessa J. Kolberg, and David A. Baum................ 280 GLUCOSE-6-PHOSPHATE ISOMERASE VARIATION AND GENETIC STRUCTURE IN YUCCA BREVIFOLIA (AGAVACEAE) Amy T. Toulson Wimmer and Robert Merritt .........cccccccccccceeeeesseseeseessenseees 285 IDENTIFICATION, DISTRIBUTION, AND FAMILY PLACEMENT OF THE PLEUROCARPOUS Moss BESTIA LONGIPES James R. Shevock, Daniel H. Norris, and A. Jonathan Shaw...........00..0064. 291 A TAXONOMIC REASSESSMENT OF CLARKIA CALIENTENSIS AND CLARKIA TEMBLORIENSIS PTORK Cy VGSOR ANGE svscsccon doc Qt PEO SRN sed enc k sk URES OS wtva casts cncaovPbawels 297 A REEXAMINATION OF THE ORIGIN OF FOREST DIFFERENCES AT A SUBALPINE LOCATION IN COLORADO SICVEM AS CHIIIGS cohsuiasescsecvsecdic ele Win viioxsieccacor ot seuenobanestndssiwnbatstenssteds 303 (OK REVIEW FLORA OF CHINA ILLUSTRATIONS. VOL. 22. POACEAE BY WU ZHENGYI, PETER H. RAVEN, AND HONG DEYUAN FLV US ACO LLELILIO IS won assis eccen sine societal RE Odea eA at io Sect am 296 TEWORTHY OETA) £2. Oe ee an en REN RON ca OA teow Ree een Lenn se mE 306 i\LLECTIONS DUNCEMENTS PRESIDENT’S REPORT FOR VOLUME 55 .....cccccccceeeeeceeeeeeeeeeeeeseeseessssssessssssssssssssssses OIA EDIMORS 4REPORT FOR VOLUMES 5-c.assctstecesiee gocecosepreocemn noe ceases chee ce taneous esl bs) RE Vile WERS “OF VIA NOSCRIPT S acs josecdcsctees ans sccacane. Gases sends eu eusenramaonhias wa Giete eae 316 PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL SOCIETY MADRONO (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 MADRONO, Kim Kersh, Membership Chair, Jepson Herbarium, University of California, Berkeley, CA 94720-2465. Editor—JOHN HUNTER Center for Plant Diversity Plant Sciences MS#7 One Shields Avenue University of California Davis, CA 95616-8780 madrono @lists.berkeley.edu Book Editor—JON E. KEELEY Noteworthy Collections Editors—DIETER WILKEN, MARGRIET WETHERWAX Board of Editors Class of: 2008—ELLEN DEAN, University of California, Davis, CA RoBERT E. PRESTON, Jones & Stokes, Sacramento, CA 2009—DOoNoOVAN BAILEy, New Mexico State University, Las Cruces, NM MARK Borcuerrt, USFS, Ojai, CA 2010—FReED Hrusa, California Department of Food and Agriculture, Sacramento, CA RICHARD OLMSTEAD, University of Washington, Seattle, WA 2011—JAmMIE KNEITEL, California State University, Sacramento, CA KEVIN Rice, University of California, Davis, CA CALIFORNIA BOTANICAL SOCIETY, INC. OFFICERS FOR 2007-2008 President: Dean Kelch, Jepson Herbarium, University of California, Berkeley, CA 94720, dkelch@sscl.berkeley. edu First Vice President: Andrew Doran, University and Jepson Herbaria, University of California, Berkeley, CA 94720, andrewdoran @berkeley.edu Recording Secretary: Staci Markos, Friends of the Jepson Herbarium, University of California, Berkeley, CA 94720-2465, smarkos @socrates.berkeley.edu Corresponding Secretary: Heather Driscoll, University Herbarium, University of California, Berkeley, CA 94720- 2465, heather.driscoll @ nature.berkeley.edu Treasurer: Susan C. Dunlap, Aerulean Plant Identification Systems, Inc., Menlo Park, CA 94025, susancdunlap@ hotmail.com The Council of the California Botanical Society comprises the officers listed above plus the immediate Past President, Michael Vasey, Department of Biology, San Francisco State University, 1600 Holloway Avenue, San Francisco, CA 94132, mvasey @sfsu.edu; the Editor of Madrofio; three elected Council Members: James Shevock, Department of Botany, California Academy of Sciences, 55 Music Concourse Drive, San Francisco, CA 94118, jshevock@ nature.berkeley.edu; Roy Buck, Jepson Herbarium, University of California, Berkeley, CA 94720, roybuck@ email.msn.com; Chelsea Specht, Department of Plant and Microbial Biology, University of California, Berkeley, CA 94720-2465, cdspecht@nature.berkeley.edu. Graduate Student Representatives: Ben Carter, Department of Integrative Biology and University Herbarium, University of California, Berkeley, CA 94720-2465, bcarter@ berkeley.edu. Webmaster: Susan Bainbridge, Jepson Herbarium, University of California, Berkeley, CA 94720- 2465, sjbainbridge @ berkeley.edu. This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). MADRONO, Vol. 55, No. 4, pp. 251-256, 2008 THE GENUS KOCHIA (CHENOPODIACEAE) IN NORTH AMERICA GE-LIN CHU Institute of Botany, Northwest Normal University, Lanzhou, Gansu 730070, China Gelinchu@yahoo.com.cn S. C. SANDERSON Shrub Sciences Laboratory, 735 N 500 E, Provo UT 84606 ssanderson@fs.fed.us ABSTRACT The genus Kochia and Bassia with which it has been combined, of Chenopodiaceae tribe Camphorosmeae, were at one time considered to include plants native to Eurasia, Australia, and North America, and included species of both C3 and C, photosynthetic types. This aggregate has been reduced in size by removal of a large group of C3 Australian genera and species. Because of their intercontinental disjunction, the presence of root sprouting, and the results of recent phylogenetic studies, it appears that the two North American species Kochia americana and K. californica, of C3 photosynthetic type, should be removed as well, and we designate Neokochia as a new genus for them. In agreement with a study by other authors, comparison of pubescence characters and ploidy levels within K. americana did not give support recognition of the segregate K. vestita. Key Words: Chenopodiaceae, Kochia, Kranz anatomy, Neokochia, polyploidy. Kochia Roth and Bassia Allen are shrubs, subshrubs or herbs with indehiscent fruits, and with embryos that are horizontally oriented in the flower and ring-like rather than spiral. Tribe Camphorosmeae, to which these genera belong, has narrow, hairy leaves, and, similar to other Chenopodiaceae, has apetalous small flowers. A disseminule consists of a seed and its pericarp enclosed within a persistent perianth (Uhlbrich 1934). The name Kochia has usually been applied to taxa with horizontal wings growing from the perianth, whereas the name Bassia has been applied to taxa with spines. The form of these appendages is apparently related to method of dispersal, by wind or animals, a characteristic likely to experience strong and variable selection. As might not be entirely surprising, therefore, the distinction has been found untrustworthy, lead- ing to a merger of these genera (Scott 1978). However, recent research indicates the need for a more extensive reorganization (Kadereit et al. 2005; Lee et al. 2005). _ KochialBassia was formerly a larger genus, but taxa native to Australia were portioned out as a number of separate genera, including Maeireana Mogquin and Sclerolaena R. Brown (Wilson 1975; Scott 1978). Recent studies have shown the Australia Camphorosmeae, which have C3 pho- ‘tosynthesis, to be a distinct clade from the Cy Eurasian group containing the majority of species of Kochia, Bassia, and related genera (Kadereit et al. 2003, 2005). In the same studies, North American Kochia came out either sister to the Australian Camphorosmeae or sister to a clade formed by the Australian Camphorosmeae plus the Cy, group. However, these relationships were all rather weakly supported. Eurasia does contain a few members of the Camphorosmeae that are C3 and are currently placed in Kochia and Bassia; these plants are related to the Australian C3 taxa or to North American Kochia (Kadereit et al. 2003, 2005). The type species of both Kochia and Bassia are in the C4 group. See Table 1 for a comparison of some of these taxa. Photosynthetic type has been identified as a character of considerable phylogenetic depth in Camphoros- meae, and in other parts of the Chenopodiaceae as well, but has been little utilized systematically (Pyankov et al. 2001; Kadereit et al. 2003, 2005). It appears that a large part of the present disorder in the tribe can be corrected by attention to photosynthetic type and its morphological corre- late, the presence or absence of Kranz anatomy. There are two species ascribed to Kochia that are native to North America, K. americana S. Watson, and K. californica S. Watson. Leaf morphology of these has been examined and found to be non-Kranz (Carolin et al. 1975). They are semi-shrubs growing in saline soils, the former widespread in saline deserts of the mountain west of North America, and the latter an uncommon plant of the Central Valley and the Mojave Desert in the state of California, and rarely of adjacent Nevada. In this paper we examine the species of North American Kochia in order to confirm their photosynthetic type. We also reconsidered the separation of K. vestita A. Nelson from K. americana. Kochia vestita was segregated on the basis of indumentum, but examination of repre- sentative specimens by Blackwell et al. (1978) failed to reveal consistent differences. However, as 252 MADRONO [Vol. TABLE 1. COMPARISON OF CHARACTERISTICS OF NEOKOCHIA WITH OTHER GENERA. Neokochia Maireana C4 Kochia & Bassia Leaf anatomy non-Kranz non-Kranz Kranz Leaf type Terete or half-terete Complanate to terete Complanate Life form Subshrubs Mainly subshrubs, with some Annuals, or (one) subshrub shrubs and perennial herbs Perianth Winglike, developing from Winglike, developing from the Winglike, developing from appendages the base of the lobes of tubular part of the perianth, thus __ the base of the lobes of the the perianth, thus the the wings are united into a single perianth, thus the wings wings are free orbicular structure, or if free, at are free least the bases of the wings are | united | Sex Bisexual or with some Bisexual to dioecious Bisexual | plants unisexual Root sprouting Present Absent Absent Perisperm Absent Present Present our preliminary observations showed both diploid and tetraploid populations existing within K. americana, it seemed desirable to examine corre- lation of the morphological characteristics with the ploidy level. We have to that end surveyed ploidy levels in K. americana and considered pubescence characters in relation to them. METHODS The C:3/C4 status of K. americana and K. californica were determined by measurement of SC/'°C ratios, averaged for two determinations per species (Hatch et al. 2006), and presence or absence of Kranz anatomy was confirmed by examination of leaf cross sections of the species (Carolin et al. 1975). Ploidy was determined by examination of meiosis in developing anthers and by flow cytometry of somatic leaf cells. Meiotic chromo- some counts were made by acetocarmine squash methods (Sanderson and Stutz 1994). Flow cytometry was carried out using fresh leaves or winter buds, which were chopped finely with a razor blade in 0.3 ml of a DAPI (4’ 6-diamidino- 2-phenylindole) solution (CyStain UV Ploidy, Partec GmbH). An additional 1.7 ml of the solution was added, and the suspension was filtered and introduced into the flow cytometer. Ploidy values for peaks at different positions were determined by comparison with plants whose ploidies had been determined cytologically. Pubescence in mid summer and fall was compared for diploid and tetraploid populations by sampling a branch from each of three plants per population for 23 populations. The plants were marked so that the same ones could be sampled at both times. Response of leaf swelling and pubescence loss to substrate salinity was studied using plants of K. americana, which were brought to the greenhouse from several locations in the field and watered initially with tap water made 0 M, 0.1 M, 0.2 M, and 0.5 M in sodium chloride. ) | Voucher specimens were deposited at the Institute of Botany, Northwest Normal Univer- sity, Lanzhou, China. RESULTS All six populations of K. californica that were’ examined for ploidy were diploid. Of 103 popu- lations of K. americana examined, 63 were diploid and 40 were tetraploid (Fig. 1). The tetraploids were mainly found in western Utah in the basin of Pleistocene Lake Bonneville, including the area of the present Great Salt Lake and Bonneville Salt Flats, with a few additional tetraploid popula- tions in eastern Nevada and at scattered locations in the Colorado Plateau. A triploid plant spread- ing by root sprouting was found growing within a diploid population at McElmo Creek, Montrose County, Colorado. Comparison of diploid DNA amounts of K. americana and K. californica by running samples of the species together in flow cytometry gave exactly overlapping peaks, and so their C-values (DNA contents of individual genomes) appear similar. Examination of internal morphology of leaves. of the American species K. americana and K. californica showed a radially organized palisade parenchyma and several vascular bundles with- out a bundle sheath or Kranz layer (Fig. 2), as was previously shown in diagrammatic form in Carolin et al. (1975). This may be contrasted with K. prostrata and K. scoparia, which had promi- nent bundle sheaths (Fig. 2). C!?/C'* ratios were —23.89 for K. americana and —28.16 for K. californica, indicating a C3 photosynthetic path- way for these species (Winter 1981). | Pubescence of K. americana consisted of elongate collapsed hairs attached to papillae on the leaf surface. Hairs of growing leaves were initially closely spaced but became more separat- ed as the leaf expanded during growth. Early summer pubescence was often widely variable within populations and also varied between populations or regions. It was perhaps more S id | = Bo ay ane ra fn ER N hee ees FIG. 2. MADRONO [Vol. vgn & = So LE FW « ty a 7 ae Scale Bar = 0.2 mm Cross sectional views of leaves, showing presence or absence of bundle sheaths and Kranz anatomy; A) | Neokochia californica S. Watson (non-Kranz). B) N. americana S. Watson (non-Kranz). C) Kochia scoparia (L.)| Schrader (Kranz). D) Kochia prostrata (L.) Schrader (Kranz). PM = palisade mesophyll, SP = parenchyma, BS = bundle sheath. prominent in tetraploids, although we found some diploid populations with as much pubes- cence as the tetraploids. Another factor affecting pubescence was appar- ently leaf swelling. Leaves may have swelled after maturity because of salt accumulation (Osmond et al. 1980), and patches of trichomes appeared to be then more easily lost by abrasion or other processes because swollen leaves in the field were mostly glabrous. All of the populations compared showed at least some loss of trichome patches by mid-autumn, but plants of eastern Utah and of the Escalante Desert in southwestern Utah started earlier and became more strongly glabrescent. The leaves of plants in the greenhouse that we watered with any of the salt solutions became grossly swollen and terete, whereas plants given water without the added salt remained slender and linear in shape. Leaves in the greenhouse experiments that had become swollen did not immediately lose their pubescence, however. DISCUSSION An alternative name for the native American taxa that we considered was ** Rhizomatosa,”’ from an herbarium annotation by Nuttall, referring to frequent vegetative reproduction by root sprout- ing. Unfortunately, Nuttall’s name is not techni- cally correct since sprouting in these species is from true roots, while rhizomes, horizontal stems with internodes and scalelike leaves, are absent. The presence of root sprouting in the North American Kochia species, as well as their inter- spongy | ; continental separation from Asian taxa, justify their designation as a separate genus. In addition, absence of Kranz anatomy also clearly differen- | tiates them from most of the old world taxa. According to results of Kadereit et al. (2003), the origin of C4 photosynthesis in the Camphor- | — . osma clade was Miocene, likely early Miocene. At) the present there are multiple named genera within the C3 and Cy, portions of Tribe Cam- : phorosmeae, and the C3 portion has radiated. from the Eurasian continent into Australia and North America. The combination of photosyn- | thetic pathway with Kranz morphology therefore. represents a taxonomic character of higher-than- | genus level within these taxa. American Kochia | should clearly be separated from Asian C4 Kochia and Bassia. It might be objected that if photo-_ synthetic type is used as a genus level character, a split within the chenopod genus Atriplex, affect-. ing many species, might also be required. While it | does appear that changes in Atriplex are neces- | sary, the time of origin for C4 photosynthesis | differs within the chenopod family (Kadereit et. al. 2003), and the appropriate taxonomic level for » the character might therefore best be evaluated on a case-by-case basis in conjunction with other evidence. | The Kochia americana—K. vestita Question Sereno Watson (1874) described var. vestita of K. americana from what probably was tetraploid | 2008] material from the shores of the Great Salt Lake. Tetraploids appeared to have somewhat longer early season pubescence on average, although there was often high variation within popula- tions, and some diploid populations were as pubescent as any of the tetraploids. Later in the season, loss of pubescence becomes more visible than trichome length. Aven Nelson (Coulter and Nelson 1909) raised var. vestita to species level, apparently with pubes- cence loss in mind. Rydberg’s Flora of the Rocky Mountains (1922) distinguishes the species K. vestita as plants that do not become glabrescent, as opposed to others that do regardless of pubescence that either may have had earlier in the season. Plants of western Utah, including both diploids and tetraploids, and of the Color- ado Plateau, mostly diploid, show a greater degree of swelling and more shedding of patches of trichomes, and the shedding begins earlier in the summer than for plants of most other locations. However, we found that all 23 of the populations we examined in the autumn shed their pubescence to some degree. The greenhouse experiment showed that leaf swelling can be induced artificially in response to salinity. There- fore we concur with Blackwell et al. (1978) that var. vestita not be recognized. TAXONOMY Neokochia (Ulbrich) G. L. Chu et S. C. Sanderson, gen. et stat. nov.—Type: Kochia americana S. Wats. Kochia Roth in Schrad. Journ. Bot. 1: 307. 1800 (1801) pro parte. Sect. Neokochia Ulbrich in Engl. et Prantl, Nat. Pflanzenfam. 2 Aufl. 16c: 535. 1934. p. p. Series Neokochia (Ulbrich) A.J. Scott, Feddes Repert. Spec. Nov. Regni Veg. 89: 108. 1978. Subfrutices vel frutices, interdum ex radices horizontales pullulans. Caules numerosi, erecti, initio dense-tomentosi. Folia linearia ad anguste oblonga, sessilia, leviter succulenta, teretia ad semi-teretia, absque Kranz anatomiis. Flores 2—5 in glomerulos, axillares, sub floribus plerumque cum singulari foliacea (breviter) vel squamosa bractea, bisexuales vel unisexuales, necnon versu dioecii; perianthia fere globosa, 5-partita, 5- nervata, evoluta 5 alato-appendicibus ex basi segmentorum fructificationum, saepissime dis- creta; alae patulae horizontaliter, tenuibus venis; stamina 5, evidenter disco; stigmata 2, raro 3, stylo brevissimo. Utriculus ovatus, pericarpio leviter carnoso. Semen horizontale, testa membranaceo, embryone prope annulari, radicula laterali, coty- ledone leviter ampliato et absque perispermate. Subshrubs or shrubs, spreading by root sprouting. Stems numerous, erect, at first densely silky-tomentose. Leaves linear, sessile, slightly CHU AND SANDERSON: NORTH AMERICAN KOCHIA 23) succulent, terete to half-terete, anatomy non- Kranz. Flowers 2—5 in glomerules, axillary, each flower usually subtended by a foliaceous or scale- like bract, bisexual or unisexual and tending towards dioecy. Perianth nearly globular, 5- parted, 5-veined, developing 5 free wing-append- ages from base of the segment at frutescence; the wings spreading, horizontal, with fine veins; stamens 5, filaments filiform, arising from a disc, anthers elliptic; stigmas 2, rarely 3, subulate or linear, style very short. Utricle depressed-ovate, pericarp slightly succulent. Seeds horizontal, testa pellucid-membranous, embryo near annular, radical lateral, cotyledons somewhat enlarged, without perisperm. X = 9. Plants of Neokochia have a multi-veined perianth, with the wing-appendages developing from the lobes. The wings are therefore separate and orbiculate, but fall together with the fruit because of the united perianth. A genus of two species endemic to western North America. Neokochia americana (S. Watson) G. L. Chu et S. C. Sanderson, comb. nov. Kochia americana S. Watson, Proc. Amer. Acad. 9: 93. 1874. K. vestita A. Nelson in J. M. Coulter & A. Nelson, New Man. Bot. Centr. Rocky Mt. 165. 1909.— Lectotype (Blackwell et al. 1978): ““Western Nevada,” J. Torrey 465 (GH). Distribution: USA: on saline soils, most abundant in Utah, Nevada, also present in parts of Arizona, California, Colorado, Idaho, Mon- tana, Oregon, and Wyoming. Neokochia californica (S. Watson) G. L. Chu et S. C. Sanderson, comb. nov. Kochia californica S. Watson, Proc. Amer. Acad. 9: 93. 1874. Kochia americana var. californica (S. Watson) M. E. Jones, Contr. West. Bot. I1: 19. 1903.— Lectotype (Blackwell et al. 1978): “Colton,” Parry 275 (GH). Distribution: USA: Central Valley and Mojave Desert of California and southern Nevada. ACKNOWLEDGMENTS Thanks to following herbaria: BRY, CAS, LZD, MO, NY, PC, RSA, UC, US, WUK, WZU for access to their collections; Dr. H. C. Stutz and Dr. E. D. McArthur for their support; and to Stephen T. Nelson, David T. Tingey, and Kent A. Hatch of Brigham Young University for carbon stable isotope measure- ments, and the Electron Microscopy Lab of BYU for preparation of leaf cross sections. We thank the reviewers of the manuscript and Blair L. Waldron and Leigh A. Johnson for reading an earlier draft, and Stanley L. Welsh for examining the Latin. LITERATURE CITED BLACKWELL, W. H., JR., M. D. BAECHLE, AND G. WILLIAMSON. 1978. Synopsis of Kochia (Chenopo- 256 diaceae) in North America. Sida, Contributions to Botany 7:248—254. CAROLIN, R. C., S. W. L. JACOBS, AND M. VESK. 1975. Leaf structure in Chenopodiaceae. Botanishe Jahr- bucher ftir Systematische Pflanzengeschichte 95:226-255. COULTER, J. M. AND A. NELSON. 1909. New manual of botany of the central Rocky Mountains. American Book Co., New York, NY. HATCH, K. A., M. A. CRAWFORD, A. W. KUNZ, S. R. THOMSEN, D. L. EGGETT, S. T. NELSON, AND B. ROEDER. 2006. An objective method for diagnos- ing anorexia nervosa and bulimia nervosa using 'SN/*N and °C/?C ratios in hair. Rapid Commu- nications in Mass Spectrometry 20:3367—3373. KADEREIT, G., T. BORSCH, K. WEISING, AND H. FREITAG. 2003. Phylogeny of Amaranthaceae and Chenopodiaceae and the evolution of C4 photo- synthesis. International Journal of Plant Sciences 164:959—-986. , D. GOTZEK, S. JACOBS, AND H. FREITAG. 2005. Origin and age of Australian Chenopodia- ceae. Organisms, Diversity and Evolution 5:59-80. LEE, B. S., M. Y. KIM, R. R.-C. WANG, AND B. L. WALDRON. 2005. Relationships among 3 Kochia species based upon PCR-generated molecular sequences and molecular cytogenetics. Genome 48:1104—1115. OSMOND, C. B., O. BJIORKMAN, AND D. J. ANDERSON. 1980. Physiological processes in plant ecology: towards a synthesis with Atriplex. Springer-Verlag, Berlin, Germany. MADRONO PYANKOV, V. I., E. G. ARTYUSHEVA, G. E. EDWARDS, | [Vol. 55. Cc: C.. BLACK, JR., AND: P.. S. -Sorris.. 2001 Phylogenetic analysis of tribe Salsoleae (Chenopo- | diaceae) based on ribosomal ITS sequences: | implications for the evolution of photosynthetic | types. American Journal of Botany 88:1189-! 1198. RYDBERG, P. A. 1922. Flora of the Rocky Mountains | and adjacent plains. P. A. Rydberg, New York, NY. | SANDERSON, S. C. AND H. C. Stutz. 1994. High | chromosome numbers in Mojavean and Sonoran | Desert Atriplex canescens (Chenopodiaceae). | American Journal of Botany 77:490-498. ScoTT, A. J. 1978. A revision of the Camphorosmioi- | deae (Chenopodiaceae). Feddes Repertorium | 89:101-119. ULBRICH, E. 1934. Chenopodiaceae. Pp. 16¢:375—584 in | A. Engler and H. Harms (eds.), Ed. 2, Die’ naturlichen Pflanzen-familien. Berlin, Germany. USDA. 2006. U.S. Department of Agriculture, Natural Resources Conservation Service. http://plants.nrcs. | usda. gov. WATSON, S. 1874. A revision of the North American } Chenopodiaceae. Proceedings of the American Academy of Arts and Sciences 9:82—126. | WILSON, P. G. 1975. A taxonomic revision of the | genus Maireana (Chenopodiaceae). Nuytsia 2:2— | 83. WINTER, K. 1981. C4 plants of high biomass in arid | regions of Asia-occurrence of C4, photosynthesis in | Chenopodiaceae and Polygonaceae from the Mid- | dle East and USSR. Oecologia 48:100—106. | MADRONO, Vol. 55, No. 4, pp. 257-268, 2008 A SPATIAL AND TEMPORAL INVESTIGATION OF ELEOCHARIS MACROSTACHYA AND ORCUTTIA TENUIS MICHELLE CLARK!, RICHARD LIS*, DEAN FAIRBANKS*, AND KRISTINA SCHIERENBECK*? Department of Geography and Planning, California State University, Chico, Chico, CA 95929-0425 ABSTRACT We investigated the possible spatial structure and temporal patterns that may determine the distribution and cover of Eleocharis macrostachya and Orcuttia tenuis within two vernal pools located in Tehama County, California. Rock cover, soil depth to hardpan, and basin elevation were compared with E. macrostachya and O. tenuis cover to investigate spatial structure. Yearly E. macrostachya and O. tenuis cover were compared with yearly precipitation and air temperature to assess temporal patterns. The spatial results suggest that soil depth to hardpan may determine E. macrostachya distribution. Rock cover and basin elevation appeared to have little effect on either E. macrostachya or O. tenuis distribution. Temporal analyses suggest that biotic interactions such as life history traits and competition may be important factors for E. macrostachya and O. tenuis distribution and density. Orcuttia tenuis cover 1s relatively stable at scale of 1.0 m? but varies at a 0.25 m’. Variability at a micro scale could be due to variations in annual air temperatures causing a possible shifting mosaic steady state. Orcuttia tenuis life history traits coincide with adaptations expected for warmer temperatures. Even though abundance is relatively stable for O. tenuis and E. macrostachya in both pools, there is significant negative correlation and spatial structure between them. Eleocharis macrostachya may dictate O. tenuis density within a pool through a combination of abiotic and biotic features. Key Words: Competition, Eleocharis macrostachya, Orcuttia tenuis, shifting mosaic steady state, soil depth, Tehama County, vernal pools. Changes in the distribution and density of populations over time and space provide the knowledge needed to understand the processes that regulate population size (Ricklefs 1997). Population parameters can be dictated by abiotic factors such as climate, topography, and soil depth, or biotic factors such as competition and life history traits. Understanding these processes is essential for the conservation and protection of vernal pool plants such as Orcuttia tenuis (CDFG 1991; USFWS 1997, 2003; CNPS 2001). Vernal pools provide an excellent opportunity to explore species distribution and density in relation to abiotic processes and biotic interac- tions. Vernal pools are isolated wetlands with unique morphologies that fluctuate seasonally "Author for correspondence, email: michelle_clark@ dot.ca.gov Current address: California Department of Trans- portation, PO Box 4906073, Redding, CA 96049-6073. *California Department of Fish and Game, 2440 Athens Ave, Redding, CA 96001. *Department of Geography and Planning, CSU Chico, Chico, CA 95929-0425. *Department of Biology, CSU Chico, Chico, CA 95929-0002. °Present address: USDA/ARS, Exotic and Invasive a Research Unit, 920 Valley Road, Reno, NV 9512. between periods of inundation and desiccation (Keeley and Zedler 1998). They support high levels of endemic species, both annual and perennials, and are considered a complex ecosys- tem (Holland and Dains 1990). The first reference to vernal pools and their distinctive plant associations can be attributed to Jepson (1925) and the large number of endemic species found within vernal pools to Hoover (1937). There are approximately 100 plants species commonly found in vernal pools, of which 90% are native and 55% endemic to California (Holland and Jain 1988; Keeler-Wolfe et al. 1998). Vernal pools occupy areas of conflicting land use and have suffered extensive habitat destruc- tion in the last century through urbanization and conversion to agriculture. California has lost between 93 and 97% of its vernal pools (Holland 1978). The continued destruction of vernal pool habitat has led to the designation of this unique ecosystem as “critical habitat’? in 2003 by the United States Fish and Wildlife Service (USFWS). Due to the high level of endemism and the loss of critical habitat, there are 15 ‘listed’? plants found in vernal pools, ten are listed as endangered, and five are threatened. Orcuttia tenuis A. Hitchce. (Poaceae), is a small, loosely tufted, blue-green annual, endemic to California vernal pools (Hickman 1993), which ts listed by the State of California as threatened and 258 by the federal government as endangered (CDFG 1991; USFWS 1997). Alterations in hydrology can affect both abiotic processes and biotic interactions for vernal pool plants. The greatest threat to O. tenuis is altered hydrology (USFWS 2003), shorter inundation periods prevent O. tenuis seeds from dispersing and germinating (Corbin and Schoolcraft 1989). Seeds are distributed within the pools by floating as the pools fill in the rainy season (Crampton 1976) and require enough standing water to allow the growth of a soil fungus over the seed in order to break dormancy (Corbin and Schoolcraft 1989). This adaptation insures that O. tenuis will germinate only when sufficient water is present in the pool to complete its lifecycle. Likewise, longer inundation periods could have a detrimental affect by promoting the growth of a perennial marsh species, such as Eleocharis macrostachya Britton (Crampton 1959; Stone et al. 1988; Bauder 1989; Corbin and Schoolcraft 1989). Stone et al. (1988) reported that FE. macrostachya competition combined with hydrological modifi- cation might have caused the loss of one O. tenuis population in Shasta County. Inundation periods are affected by basin mor- phology (Brooks and Hayashi 2002), basin depth (Zedler 1987; Bauder 1987, 2000), soil (Griggs 1981; Holland and Dains 1990; Williamson et al. 2005), and yearly precipitation (Zedler 1987; Bauder 2000). The shape and size of a basin determines the amount of subsurface water-flow into the pools and deeper basins have longer inundation periods (Keeley and Zedler 1998). In general, soils act as reservoirs for the water and moisture retention varies with soil depth (Miller and Donahue 1990). Changes in any of these abiotic factors affect the inundation period of a pool, and hence the density and distribution of O. tenuis. There is little information on the distribution of vernal pool plant species within an individual pool in relation to abiotic processes (basin morphology, basin elevation, and soil depth to hardpan) and biotic interactions (competition and life history traits). Our purpose is to gain information on the spatial structure and temporal patterns that affect E. macrostachya and O. tenuis distribution and density within individual vernal pools. We aim to accept or reject the following hypotheses: Spatial 1) O. tenuis and E. macrostachya have spatial structure and a significant negative correla- tion. O. tenuis and E. macrostachya cover estimates have a significant positive correlation with decreasing basin elevation (deep depths). 3) O. tenuis cover estimates have a significant positive correlation with decreasing soil depth to hardpan and increasing rock cover. MADRONO SG i a re ~~ Spring Branch Raad Fic. 1. Location of study pools in eastern Tehama County, California. 4) E. macrostachya cover estimates have a significant positive correlation with increas- ing soil depth to hardpan and decreasing | rock cover. Temporal Ly. 70; at both.a 0.25 m’ and 1.0 m’ scale. 2) O. temperatures. The results from these tested hypothesizes will guide management of O. tenuis and E. macrostachya cover esti- | mates vary significantly over sampling years | tenuis and E. macrostachya cover esti- | mates have a positive correlation with © increasing precipitation and decreasing air | tenuis through the | identification of critical abiotic and biotic habitat | preferences. METHODS Study Area Two large deep northern volcanic vernal pools were surveyed in the Northeastern Sacramento | Valley Vernal Pool Region (Keeler-Wolfe et al. | 1998). The first pool sampled, Sevenmile Lake (7- | Mile) is owned and managed by the Bureau of | Land Management (BLM) and is located ap- | proximately seven miles northeast of Red Bluff | on Highway 36 (Fig. 1). Sevenmile Lake is | approximately 2.3 acres. The second pool is. Spring Branch 2 (SB2), also owned by BLM, | located on Spring Branch Road and is approx- | imately 3.32 acres in size (Fig. 1). Resource and | geographic constraints limited this study to only two vernal pools. Data Collection All data points were recorded with a Trimble | GeoExplorer 3 GPS unit, differentially corrected, 2008] and exported to a resident ArcGIS shapefile format. Basin morphology. The perimeter of each pool was walked with a Trimble GPS unit and delineated by the lack of common shallow water wetland and vernal pool indicator species typical in these pools. In low gradient areas with more widely fluctuating inundation periods, the species most often used for delineation were Deschampsia danthonioides, Hordeum marinum ssp. gussonea- num, and in some areas, Lolium multiflorum. In higher gradient areas with more stable long inundation periods, typical species were Eryngi- um castrense, Downingia spp., Eleocharis macro- stachya, and Orcuttia tenuis. Hydrologic and geomorphologic features also were used exten- sively. Basin elevation. A laser level (Topcon RL-H3C 980’ Diameter- Servo Leveling) was used to map the basin elevation of the vernal pool basins in Fall 2005. Line transects were placed every ten meters, running in a general east to west pattern through the vernal pool basin and elevation points were collected every five meters along each transect. A total of 253 elevation points were collected along ten transects for 7-Mile and 268 elevation points were collected along 16 transects for SB2. Additional elevation points were col- lected in the upland mosaic to create a more accurate physiographic map. Relative cover estimates. Rock, E. macrosta- chya, and O. tenuis cover were estimated every five meters along the basin transects. Cover was determined by using two 0.5 m? quadrats placed side by side at each elevation point, and was read as one meter squared. A modified cover class system was created and used to quantify cover estimates in this study (Table 1). A total of 1601 m* quadrats were read for 7-Mile in October 2005 and 3001 m° quadrats were read for SB2 in July 2006. Soil depth to hardpan. The measurable depth of soil to hardpan was calculated using a soil probe at the same location and time as the vegetation quadrats within the basin. At this time the pool soils were completely dry and usually cracking due to desiccation. The soil probe was pushed into the ground, usually between three and seven times, within each | m quadrat. Each time the soil probe was pushed into the basin floor; the ground surface level was marked and measured to the end of the probe. The maximum soil depth was recorded in centimeters for each vegetation sampling point. Absolute cover estimates. As part of an earlier vernal pool study, permanent transects were established in 1999 from a stake (pivotal point) driven into a central area of a pool that traverses CLARK ET AL.: AN INVESTIGATION OF E. MACROSTACHYA AND O. TENUIS 259 TABLE 1. ORCUTTIA TENUIS AND E. MACROSTACHYA ESTIMATE RELATIVE COVER CLASSES. Class Percent Cover Absent 1-10% 11-20% 21-30% 31-40% 41-50% 51-60% 61-70% 71-80% 81-90% 91—-100% SODNDNBRWNK OO — to the pool margin (Lis and Eggeman 1999; Lis and Clark 2007). Sevenmile Lake has four transects and SB2 has two transects. A system of permanently marked quadrats, located on permanent transects was used with the permanent quadrats and transects marked with large steel spikes. At each spike, four quadrats (0.25 m°) were laid out, and the absolute cover of rock, bare ground, and indicator plant species record- ed, including O. tenuis and E. macrostachya. Data were collected in the years 1999, 2000, 2001, 2004, and 2006. Climate data. The mean maximum monthly air temperature and mean monthly precipitation data were obtained from the California Depart- ment of Water Resources (DWR) (http://cdec. water.ca.gov/queryTools.html) and Western Re- gional Climate Data Center (http://www.wrcc.dri. edu/). The data station used was Red Bluff Diversion Dam (RBD) for both queries. Data Analysis Basin morphology. The area and perimeter of each pool was automatically calculated during the creation of the polygon shapefile. The shape value was calculated using the following equa- tion: perimeter/(3.54 * sqrt (area)), where the value of 1.0 1s equal to a circle, and larger for a distended shape (Longley et al. 2001). Basin elevation and soil depth to hardpan. The elevation points were interpolated into a Digital Elevation Model (DEM), using Hutchinson’s (1989) Topo to Raster implemented in ArcGIS Ver. 9.1 (Environmental Systems Research Insti- tute), which interpolates a hydrologically correct surface from point, line, and polygon data. A two-group t-test was performed between the two pools to determine the significant difference between means for soil depth to hardpan and basin elevation. Spatial autocorrelation. Spatial dependence in O. tenuis and E. macrostachya was investigated in 260 several ways to understand both the overall spatial pattern of each species (global random vs. non-random structure) and to identify signif- icant patch distributions in each pool. These tests also helped us to anticipate any issues with conducting standard normal linear statistics. While spatial autocorrelation does not inflate the explained variance term (in r or r* relationship analysis), it can lead to inflated (artificially large) sample sizes due to non-independent samples, which may influence significance tests (P-values) and result in the inclusion of non-significant predictor variables (Legendre and Fortin 1989). The local spatial correlation was examined using isotropic derived Moran’s I correlograms and local indicators of spatial association (LISA) maps (Anselin 1995). Correlations. After spatial autocorrelation was examined, the point shapefiles representing cover estimates and soil data for each site were buffered by | m to create a 1 m diam. zone for each point and zonal statistics were calculated from the DEM. The zonal statistical result (the mean basin elevation) was added to the field data (soil depth to hardpan and cover estimates) to create a data analysis matrix for each pool. All statistical tests were performed in Systat 8.03 (SPSS Inc.) with « = 0.05. Since the cover estimates were ranked data, a Spearman’s rank correlation matrix was used to determine if there was a significant correlation between O. tenuis and E. macrostachya with rock, soil depth, and elevation (Sokal and Rohlf 1981). Linear regressions were calculated in order to determine variable predictability. Absolute cover estimates. Analyses of absolute cover estimates were conducted at two scales. The first scale was at the 0.25 m? level for which each of the four quadrats per sampling point were individually analyzed to examine micro-changes in distribution within the pool. The second scale was at 1.0 m? for which each of the four quadrats per sampling point were averaged to aggregate the data into one sampling point value to examine changes at the whole pool level. A one-way Analysis of Variance (ANOVA) was used at both scales (0.25 m* and 1.0 m7’) to determine if there was a significance difference in mean absolute cover estimates for O. tenuis and E. macrostachya over sampling years. A Bone- ferroni post-hoc comparison was used to deter- mine significant differences among years. Climate data. Air temperature and precipita- tion data were tested against O. tenuis and E. macrostachya cover using a Pearson’s Correlation Test (x = 0.05). Data were checked for normality before analysis and linear regressions were used to determine if air temperature and precipitation were significantly correlated. MADRONO TABLE 2. COMPARISON OF BASIN POOL. Sevenmile Lake Area = 9460 m? Perimeter = 479 m Shape = 1.15 Spring Branch 2 Pool Area = 13,838 m7? Perimeter = 440 m Shape = 1.27 RESULTS AND DISCUSSION Spatial Data Basin morphology. Spring Branch 2 pool is the | larger pool with a distended shape that has been caused by an artificial dirt dam on the southwest side of the pool (Table 2). It has been suggested that artificial dirt dams provide more area suitable for Orcuttia growth by prolonging | inundation (Griggs 1974). Basin elevation and soil depth to hardpan. In | addition to SB2 being larger than 7-Mile, it is | deeper with less soil. T-tests determined a_ significant difference between means (7-Mile = | 0.72 ft and SB2 = 1.16 ft) for basin elevation (df | = 458, P = 0.000) and between means (7-Mile = | 8.62 cm and SB2 = 6.79 cm) for soil depth to | hardpan (df = 458, P = 0.002) between the two | pools. Since the study pools are morphologically different, O. tenuis and E. macrostachya distribu- | tion and cover may vary between pools due to | microhabitat differences. Spatial autocorrelation. One of our larger goals | was to determine if there is significant spatial | tenuis and E. macrostachya | within the study pools. The value for Moran’s I | for global spatial autocorrelation indicate that | spatial pattern of O. tenuis and E. macrostachya | are not random but have significantly clustered | structure for O. patches in both pools (P = 0.01). [Vol. 55 | MORPHOLOGY | (AREA, PERIMETER, AND SHAPE) FOR EACH VERNAL | Moran’s I correlograms also detected spatial | structure in the populations between pools. If cover | diversity were randomly distributed, the correlo- grams would be statistically non-significant and would exhibit a more or less flat profiles (values | near O for most distance classes). The Moran’s I correlograms (Fig. 2) show O. tenuis and E. | macrostachya have significant positive autocorre- | lation detected in the pool at different distances depending on the pool. The spatial autocorrela- | tion pattern (size of patches) for E. macrostachya | is different (7-Mile = and O. tenuis is similar (7-Mile = 0-45 m, SB2 = 0-51 m), between study pools (Fig. 2). The distance differences for significant positive spatial autocorrelation detected for E. macro- stachya in 7-mi compared SB2 could be due to 0—52 m, SB2 = 0-32 m), | pool abiotic differences (soil depth to hardpan) | and/or biotic differences (life history reproduc- tion) (Sokal and Oden 1978). Eleocharis macro- 2008] CLARK ET AL.: AN INVESTIGATION OF E. MACROSTACHYA AND O. TENUIS 261 Sevenmile Lake Moran's I —@ 0. tenuis large shapes -statistically significant —--E. macrostachya Dispersed Distance (meters) Spring Branch 2 Clustered Moran's | —@- 0. tenuis large shapes -statistically significant --E. macrostachya Dispersed Distance (meters) Fic. 2. stachya is perennial and reproduces vegetatively from rhizomes and by seeds (DiTomaso and Healy 2003). Seeds typically germinate in stand- ing water during mid-spring through early summer (DiTomaso and Healy 2003) and rapid rhizomatous growth occurs in mid and late summer (Strandhede 1966). Orcuttia tenuis is an annual, and only reproduces by seeds, which are distributed within the pools by floating as the pools fill in the fall or winter (Crampton 1976). A correlogram for O. tenuis and E. macrostachya in Sevenmile Lake and Spring Branch 2 Pool. Because vegetative reproduction is not observed in O. tenuis, this autocorrelation at similar distance classes probably reflects the occurrence of similar patch area sizes (i.e., high probability areas for survival) for plant development in this annual. According to Sokal (1979), patch size can be estimated by the distance at which the Moran’s I correlogram first intercepts the abscissa, as this corresponds to the shortest dimension of an 262 MADRONO irregularly shaped patch. The Moran’s I correlo- grams (Fig. 2) also indicate that there is a linear gradient from significant to non-significant auto- correlation across each pool for both species (1.e., positive values are not seen at higher distance classes). This suggests there is not a circular gradient (Sokal and Oden 1978) of patches for these species in a pool (1.e., not like a typical vernal pool bathtub ring of species distribution) and that the patch distribution is not regular (Legendre and Fortin 1989; Radeloff et al. 2000). Figures 3 and 4 show the location of significant clusters of E. macrostachya and O. tenuis within the pools in relation to rock cover, soil depth, and basin elevations. Biotic correlations. A Spearman’s rank corre- lation matrix determined that O. tenuis and E. macrostachya clusters are significantly negatively correlated in both pools (7-Mile P = 0.00; SB2 P = 0.00; Figs. 4 and 5). A linear regression found a negative significant correlation between O. tenuis and E. macrostachya in both pools but that is a weak predictor for cover (7-Mile — df = 157, adjusted r* = 0.18, P = 0.00 and SB2 — df = 297, adjusted r°> = 0.13, P = 0.00). This result suggests that there is a better predictor variable for cover such as inundation period, which was not examined in this study. Even though the negative correlation between the species is weak, the relationship may be competition, which is supported by Stone et al. (1988), who observed that competition with FE. macrostachya rather than inundation directly limits the distribution of O. tenuis. Abiotic correlations — Orcuttia tenuis. Basin elevation.—A Spearman’s rank correlation matrix determined that the 7-Mile population of O. tenuis is significantly negatively correlated (df = 155, P = ().00) with basin elevation. A 7-Mile LISA map of O. tenuis cover and basin elevation shows that the significant patches of O. tenuis are found in shallower depths of the pool (Fig. 3). A linear regression found basin elevation to be a weak but significant predicting variable (df = 157, adjusted r = 0.18, P = 0.00). This result also suggests that another variable such as inundation period may be a better predictor for O. tenuis cover. A Spearman’s rank correlation matrix deter- mined that the SB2 population of O. tenuis had no significant correlation (df = 295, P = 0.24) with basin elevation. The lack of correlation between O. tenuis cover and basin elevation can be seen in the location of the significant patches of O. tenuis on the SB2 LISA map (Fig. 4). The significant clusters of O. tenuis were not found at shallow or deep depths in SB2 and suggest basin elevation does not affect O. tenuis density and distribution. The lack of positive correlations between O. tenuis cover and basin elevation is unexpected because previous research indicates O. tenuis [Vol. 55 | prefers deep portions of a pool (Corbin and | Schoolcraft 1989), but not surprising since we | observed O. tenuis patches in the shallow and |} medium depths of the pools. This result could be | due to the large size of the study pools, which | tend to have longer inundation periods (Brooks | and Hayashi 2002). The deep portions of the | study pools may have an exceptionally long inundation period that push O. tenuis to survive } in the shallow and medium depths of the pool. | These shallower depths must have an inundation | period long enough to meet the threshold of O. | tenuis (Holland 1987). Our findings may indicate | that inundation period not depth affect O. tenuis | abundance and distribution. Rock cover and soil depth to hardpan.—A | Spearman’s rank correlation matrix found that | O. tenuis has a significant positive correlation (7- | Mile df = 155, P = 0.00; SB2 df = 295 P = 0.00) } with rock and no significant correlation (7-Mile | df = 155, P = 0.69; SB2 df = 295, P = 0.53) with soil depth to hardpan in both study pools. For) SB2, a linear regression analysis found rock cover | to be an insignificant predicting variable for O. | tenuis cover. Rock cover may shape O. tenuis | distribution within a pool by allowing more | available microhabitat for O. tenuis that may. not be suitable for other vernal pool plant. species. Soil depth to hardpan appears to have. little to no affect on O. tenuis distribution and. abundance. Research on soil requirements for O. | tenuis 1s not currently available indicating a need | for more information. | Abiotic correlations — Eleocharis macrostachya. | Basin elevation.—A Spearman’s rank correlation | matrix found E. macrostachya to have no. significant ENS MLO) (7-Mile df = 155, P = 0.89; SB2 df = 295, P = 0.68) with basil elevation in both study pools. The locations of significant E. macrostachya clusters in relation to | elevation show this lack of correlation (Figs. 3. and 4). In these LISA maps, it appears the significant patches of E. macrostachya were not found in shallow or deep depths for both study pools. These results suggest that the inundation | requirements for E. macrostachya are being met across the topographic gradient, which are. standing water no deeper than | m with occa- sional fluctuations down to saturated conditions throughout the growing season (USDA 2006). Rock cover and soil depth to hardpan.—A Spearman’s rank correlation matrix found E. macrostachya has a significant negative correla- | tion (df = 295, P = 0.00) with rock cover and. significant positive correlation (df = 295, P =— 0.00) with soil depth to hardpan in SB2. There is | a significant negative correlation between soil and | rock cover indicating that rock cover probably | has little affect on E. macrostachya and was eliminated from the linear regression, which. 2008] CLARK ET AL.: AN INVESTIGATION OF £. MACROSTACHYA AND O. TENUIS N Nn ies) Sevenmile Lake LISA Map with DEM Significant Clusters (number in shape = cover value) [| So) O. tenuis E. macrostachya Water Depth in feet 1 centimeter equals 12 meters Sevenmile Lake LISA Map with Soil Depth Significant Clusters (number in shape = cover value) [| O. tenuis © E. macrostachya Soil Depth in cm 1 centimeter equals 12 meters Fic. 3. A LISA map of significant O. tenuis, E. macrostachya patches in Sevenmile Lake in relation to basin elevation and soil depth. Orcuttia tenuis has a significant negative correlation with E. macrostachya and basin elevation. Eleocharis macrostachya has a significant positive correlation with soil depth to hardpan. found soil depth to be a significant (df = 297, period may be a better predictor for E. macro- adjusted r° = 0.23, P = 0.00), but weak predicting stachya cover. In 7-Mile and SB2, a Spearman’s variable for E. macrostachya cover. This result rank correlation matrix found E. macrostachya suggests that another variable such as inundation — has a significant positive correlation (df = 295, 264 MADRONO [Vol. 55 1 centimeter equals 16 meters 1 centimeter equals 16 meters Fic. 4. A LISA map of significant O. tenuis, E. macrostachya patches in Spring Branch 2 Pool in relation to basin , elevation and soil depth. Orcuttia tenuis has a significant negative correlation with E. macrostachya and no | Spring Branch 2 LISA Map with DEM Significant Clusters (number in shape = cover value) [| O. tenuis >) E. macrostachya Water Depth in feet Spring Branch 2 LISA Map with Soil Depth Significant Clusters (number in shape = cover value) [| O. tenuis e E. macrostachya Soil Depth in cm correlation with basin elevation. Eleocharis macrostachya has a significant positive correlation with soil depth to hardpan. P = 0.00) with soil depth and no correlation with rock. The positive soil correlation can be seen in a LISA map of each pool, representing the significant patches of E. macrostachya in relation to soil depth to hardpan (Figs. 3 and 4). These results suggest that soil depth to hardpan may. determine E. macrostachya density and distribu- | tion in both study pools. | 2008] Sevenmile Lake Least Squares Means 41 < 33 = =) S) ro) 25 A B S O 7 1999 2000 2001 2003 2006 1B YEAR 1a T E Cc Sevenmile Lake Y Least Squares Means E R 10.0 7.8 ” be HG o13 e) mW 34 LJ 12 1999 2000 2001 2003 2006 YEAR Fic. 5. The mean absolute cover of O. tenuis and E. macrostachya (0.25 m*) in Sevenmile Lake and over sampling years. The positive correlation between E. macrosta- chya and soil depth to hardpan, may explain why -E. macrostachya patches were larger in SB2 compared to 7-Mile (Figs. 3 and 4). Spring Branch 2 pool has a greater mean for soil depth to hardpan compared to 7-Mile, indicating soil depth may determine cluster size. Soil depth to hardpan may play an important part in the significant spatial structure of E. macrostachya and O. tenuis in both study pools. Temporal Data Absolute cover estimates. A one-way ANOVA found a significant difference in mean absolute cover estimates for O. tenuis and E. macrostachya over sampling years for 7-Mile, at the micro scale of 0.25 m? (df = 315, O. tenuis P = 0.00; E. -macrostachya P = 0.01). The mean absolute cover of the 7-Mile population of O. tenuis is significantly different between 1999 and 2001 (df = 315, P = 0.01), and 2001 and 2006 (df = 315, P CLARK ET AL.: AN INVESTIGATION OF E. MACROSTACHYA AND O. TENUIS 265 = 0.01), 2001 having the highest mean (Fig. 5). The mean absolute cover of the 7-Mile popula- tion of E. macrostachya at the micro scale of 0.25 m* is significantly different between 1999 and 2003 (df = 315, P = 0.01), and 2000 and 2003 (df = 315, P = 0.03), with 1999 having the highest mean and 2003 the lowest (Fig. 5). In SB2, O. tenuis cover also varied at micro scale (0.25 m*) over sampling years. A one-way ANOVA found a significant difference (df = 155, P = 0.01) in mean absolute cover estimates for O. tenuis 1s significantly different between 1999 and 2000 (df = 155, P = 0.03), 2000 and 2003 (df = 155, P = 0.01), and 2000 and 2006 (df = 155, P = 0.02), with 2000 having the highest mean (Fig. 6). There is no significant difference (df = 155, P = 0.96) in mean absolute cover estimates at a micro scale (0.25 m°) for E. macrostachya over sampling years in SB2 (Fig. 6). These results indicate that O. tenuis mean absolute cover at a 0.25 m° scale changes over time. Sevenmile Lake and SB2 have different years for high cover estimates of O. tenuis, which could be caused by morphological differences between the pools. Since E. macrostachya cover only varied in 7- Mile, these results indicate E. macrostachya may be relatively stable over time. Eleocharis macro- stachya cover may have varied over sampling years in 7-Mile and not SB2 because of soil depth differences between the pools. Sevenmile Lake has a higher mean for soil depth to hardpan compared to SB2, which may allow for more variability in density and distribution for E. macrostachya. A one-way ANOVA found no significant difference in mean absolute cover estimates for O. tenuis and E. macrostachya over sampling years for both study pools at an aggregated 1.0 m° scale. This result is puzzling since O. tenuis cover varied at a 0.25 m° scale and unexpected because research has shown that climate (Crampton 1959; Griggs 1976, 1981; Griggs and Jain 1983) and possible competition between FE. macrostachya and O. tenuis (Crampton 1959; Stone et al. 1988) limited O. tenuis abundance. Our results suggest that E. macrostachya and O. tenuis cover 1s relatively stable at the whole pool level. Even though O. tenuis cover 1s relatively stable at a 1.0 m’ scale, changes on a micro scale (0.25 m’) indicate a possible “shifting mosaic steady state’. Bormann and Likens (1979) report ‘shifting mosaic steady state” is the result of yearly disturbances changing the location for a specific habitat, however, even though the actual locations change, the amount of each habitat over a large area may remain somewhat constant (steady state). Disturbances cause changes in habitat. Yearly changes in precipitation and air temperature, hence weather, may affect O. tenuis cover. 266 Spring Branch 2 Pool Least Squares Means 19 15 < 5 1 ) ne fe) v4 x B S 1999 2000 2001 2003 2006 O YEAR L U T E Spring Branch 2 Pool Cc Least Squares Means O V 25 E R 20 ” pen = © W 10 Lu 5 1999 2000 2001 2003 2006 YEAR Fic. 6. The mean absolute cover of O. tenuis and E. macrostachya (0.25 m*) in Spring Branch 2 over sampling years. Climate data. Eleocharis macrostachya and O. tenuis mean absolute cover at a 0.25 m* scale changes over time, and may be due to changes in weather. A Pearson’s correlation matrix deter- mined that there is a significant positive correla- tion (df = 109, P = 0.02) between air temperature and O. tenuis, and a significant negative correla- tion (df = 109, P = 0.02) between precipitation and FE. macrostachya at the 0.25 m? scale within 7-Mile. A linear regression (df = 109, adjusted r° = 0.03) for O. tenuis mean absolute cover found precipitation (P = 0.01) to be a negative predicting variable and (P = 0.00) air tempera- ture a positive predicting variable, with air temperature as the stronger predictor. For E. macrostachya mean absolute cover, a _ linear regression (df = 109, adjusted r*7 = 0.02, P = 0.02) determined precipitation is the only signif- icant predicting variable. In SB2, a Pearson’s correlation matrix found a significant positive correlation (df = 158, P = MADRONO 0.05) between O. tenuis and air temperature. A linear regression (df = 158, adjusted r? = significant but a weak predictor for O. tenuis cover. No climate correlations were found for E. | macrostachya cover in SB2. This result is anticipated since E. macrostachya cover did not vary over time in SB2. Our results suggest that O. tenuis cover is not | affected by the amount of rain per year, which is unexpected based on previous studies suggesting | there is relationship between precipitation and O. tenuis (Crampton 1959; Griggs 1974, 1981; Griggs and Jain 1983). The differences in correlation between O. tenuis and precipitation — could be related to the differences in correlation © since basin elevation | with basin elevation, (Zedler 1987; Bauder 1987, 2000) and yearly precipitation (Zedler 1987; Bauder 2000) are | connected to the inundation period of a vernal pool. The study pools are large and deep which | promote long inundation periods (Brooks and | Hayashi 2002). These longer inundation periods ! may meet the inundation threshold for O. tenuis | seeds (Reeder 1965, 1982; Holland 1987) across the topographic gradient, indicating that dura- tion not depth of inundation determine O. tenuis | distribution. The positive correlation between O. cover and air temperature suggests that cover will be higher in warmer years and possibly due | to the ability of O. tenuis to switch from a C-3 to | C-4 photosynthesis in the late spring (Corbin and | Schoolcraft 1989). This switch to C-4 photosyn- thesis allows O. tenuis to adapt to high intensive | sunlight and use water more efficiently. This may | also explain why O. tenuis significant clusters | were not found in the deep parts of the pools. Ifa | winter season extends into late spring, the lack of | sunlight and increased precipitation will hinder | the C-4 photosynthesis pathway of O. fenuis, | especially in deep parts of the pool. In this | tenuis cover will be higher in > shallower depths further supporting a hypothesis | for shifting mosaic steady state (Bormann and situation, O. Likens 1979). CONCLUSIONS It is important to understand the abiotic and biotic features that limit the distribution and > density of a species, especially for species and | habitats that have a special conservation status, | such as O. tenuis and vernal pools. Eleocharis macrostachya and O. tenuis abun- dance is relatively stable in both pools at a 1 m° scale. Orcuttia tenuis cover varied over time in | both pools at the 0.25 m?° scale, which could be | due to a combination of biotic (negative correla- tion with FE. macrostachya and life history | reproduction traits) and abiotic (positive correla- | [Vol. 55 | 0.02, Em = 0.05) determined that air temperature is a | tenuis | 2008] tion with air temperature) features. It 1s impor- tant to emphasize that although cover changes occur at a micro scale these changes were not reflected when the data were aggregated at a scale of 1.0 m*. Annual changes in air temperatures that allow O. tenuis to thrive in some areas of the pool better than others in different years may be explained by the shifting mosaic steady state theory. Although cover is relatively stable in both pools for O. tenuis and E. macrostachya at the 1.0 m scale, we found both a significant negative correlation between the two species. Temporal data show no significant changes in cover over time for both E. macrostachya and O. tenuis at a 1 m° scale. Thus, if competition is occurring between E. macrostachya and O. tenuis, E. macrostachya is not limiting the density of O. tenuis but rather dictating its distribution within a pool. Our data show that depth of soil to hardpan may drive E. macrostachya density and distribution. Eleocharis macrostachya 1s perennial species that thrives in pool areas with deep soils (soil depth >6 cm), which can prevent O. tenuis from establishing in these areas. Since O. tenuis cannot establish in areas with E. macrostachya, it is forced to live in the rocky areas of a pool that have shallow soils. Our results show that conservation strategies need to include monitoring sedimentation within the vernal pools by managing grazing activities and limiting anthropogenic disturbances that may increase erosion in the surrounding uplands. If pools fill with sediment, it will increase the available habitat for E. macrostachya; hence limit available habitat for O. tenuis. Our results show that soil plays an important role in conserving vernal pool plant species. Future research for O. tenuis conservation should include soil mapping, transplant experiments, and inundation periods. Mapping soils based on texture and chemistry may give insight on the soil types preferred by E. macrostachya and O. tenuis. Conducting transplant experiments in different soil depths would determine if there is positive correlation between of soil depth and possible completion between E. macrostachya and O. tenuis. ACKNOWLEDGMENTS _ This study was part of an ongoing research project conducted by the California Department of Fish & Game on the community interactions within Tehama ‘County vernal pools. Funded in Part by: U.S. Fish and Wildlife Service, Section 6 Grant # E-2-P-21, RO185002 (Receivable), and the California Dept. of Fish and Game, Northern Region FG 8610-R1, P0310730, P0310730 AmOI contracts with California State Uni- versity, CSU Chico Research Foundation. This work was extensively supported by the California Depart- ment of Fish and Game, Northern Region. CLARK ET AL.: AN INVESTIGATION OF E. MACROSTACHYA AND O. TENUIS 267 LITERATURE CITED ANSELIN, L. 1995. Local indicators of spatial associa- tion — LISA. Geographical Analysis 27:93—115. BAUDER, E. T. 1987. Species assortment along a small- scale gradient in San Diego vernal pools. Ph.D. dissertation. University of California, Davis, CA. . 1989. Drought stress and competition effects on the local distribution of Pogogyne abramsii. Ecology 70:1083—1089. . 2000. Inundation effects on small-scale distri- bution in San Diego, California vernal pools. Aquatic Ecology 34:43-61. BORMANN, F. H. AND G. E. LIKENS. 1979. Catastrophic disturbance and the steady state in northern hardwood forests. American Scientist 67:660—669. BROOKES, R. T. AND M. HAYASHI. 2002. Depth-area- volume and hydroperiod relationships of ephem- eral (vernal) forest pools in southern New England. Wetlands 22:247—255. CAMPBELL, N. 1993. Biology. The Benjamin/Cummings Publishing Company, Inc., Redwood City, CA. CORBIN, B. AND G. SCHOOLCRAFT. 1989. Orcuttia tenuis species management guide. Unpublished report to Lassen National Forest and Susanville District Bureau of Land Management. CALIFORNIA DEPARTMENT OF FISH AND GAME (CDFG). 1991. Annual report on the status of California State listed threatened and endangered animals and plants. California Department of Fish and Game, Sacramento, CA. CALIFORNIA NATIVE PLANT SOCIETY (CNPS). 2001. Inventory of rare and endangered plants of California, 6th ed. Rare Plant Scientific Advisory Committee, David P. Tibor, Convening Editor. California Native Plant Society. Sacramento, CA. CRAMPTON, B. 1959. The grass genera Orcuttia and Neostaphia: A study in habitat and morphological speciation. Madrono 12:225—256. 1976. Rare grasses in a vanishing habitat. Fremontia 4:22—23. DITOMASO, J. M. AND E. A. HEALY. 2003. Aquatic and riparian weeds of the West. Publication 3421. University of California, Davis, CA. GRIGGs, F. T. 1974. Systematics and ecology of the genus Orcuttia (Gramineae). M.A. thesis, Califor- nia State University, Chico, CA. 1976. Life history strategies of the genus Orcuttia (Gramineae).-Pp. 57-63 in S. Jain (ed.), Vernal pools, their ecology and conservation. Institute of Ecology Publication No. 9. University of California, Davis, CA. 1981. Life histories of vernal pool annual grasses. Fremontia 9:14—17. AND S. K. JAIN. 1983. Conservation of vernal pool plants in California. IT. Population biology of a rare and unique grass genus Orcuttia. Biological Conservation 27:171—193. HICKMAN, J. C. 1993. The Jepson Manual. University of California Press, Berkeley, CA. HOLLAND, R. F. 1978. The geographic and edaphic distribution of vernal pools in the Great Central Valley, California. California Native Plant Society, special publication number 4, Fair Oaks, CA. . 1987. What constitutes a good year for an annual plant? Two examples from the Orcuttieae. Pp. 329-333 in T. S. Elias (ed.), Proceedings of a California conference on the conservation 268 and management of rare and endangered plants. California Native Plant Society, Sacramento, CA. AND V. I. DAINS. 1990. The edaphic factor in vernal pool vegetation. Pp. 31-48 in D. H. Ikeda and R. A. Schlising (eds.), Vernal pool plants-their habitat and biology. Studies from the Herbarium No. 8. California State University, Chico, CA. AND S. K. JAIN. 1988. Vernal pools. Pp. 515— 533 in M. G. Barbour and J. Major (eds.), Terrestrial vegetation of California. Wiley-Intersci- ence, New York, NY. Hoover, R. F. 1937. Endemism in the flora of the Great Valley of California. Ph.D. dissertation, University of California, Berkeley, CA. HUTCHINSON, M. F. 1989. A new procedure for gridding elevation and stream line data with automatic removal of spurious pits. Journal of Hydrology 106:211—232. JEPSON, W. L. 1925. A manual of the flowering plants of California. Associated Students Store, Universi- ty of California, Berkeley, CA. KEELER-WOLF, T., D. R. ELAM, K. LEWIS, AND S. A. FLINT. 1998. California vernal pool assessment preliminary report. California Department of Fish and Game, Sacramento, CA. KEELEY, J. E. AND P. H. ZEDLER. 1998. Characteriza- tion and global distribution of vernal pools. Pp. 1— 14 in C. W. Witham, E. T. Bauder, D. Belk, W. R. FerrenJr, and R. Ornduff (eds.), Ecology, conser- vation, and management of vernal pool ecosystems. Proceedings from a 1996 conference. California Native Plant Society, Sacramento, CA. LEGENDRE, P. AND M. J. FORTIN. 1989. Spatial pattern and ecological analysis. Vegetatio 80:107—138. Lis, R. A. AND E. C. EGGEMAN. 1999. Community interactions and grazing impacts on seven rare vernal pool plants and macro-invertebrates in Tehama County. Study design and Methodology, California Department of Fish and Game, Redd- ing, CA. AND M. A. CLARK. 2007. Conservation and management of vernal pools through analysis of community interactions of grazing and fire impacts on five rare plants and two rare macro-inverte- brates in Tehama County. Final report. California Department of Fish and Game, Redding, CA. LONGELY, P. A., M. F. GOODCHILD, D. J. MAGUIRE, AND D. W. RHIND. 2001. Geographic information systems and science. John Wiley & Sons, Ltd., Chichester, U.K. MILLER, R. W. AND R. L. DONAHUE. 1990. Soils: An introduction to soils and plant growth, 6th ed. Prentice-Hall Inc., Englewood Cliffs, NJ. RADELOFF, V. C., T. F. MILLER, H. S. HE, AND D. J. MLADENOFF. 2000. Periodicity in spatial data and geostatistical models: autocorrelation between patches. Ecography 23:81—92. MADRONO [Vol. 55 REEDER, J. R. 1965. The tribe Orcuttieae and the sub- tribes of the Pappophoreae (Gramineae). Madrono 18:18—28. 1982. Systematics of the tribe Orcuttieae (Gramineae) and the description of a new segregate genus, Tuctoria. American Journal of Botany 69:1082—-1095. RICKLEFS, R. E. 1997. The economy of nature: a textbook in basic ecology, 4th ed. W.H. Freeman and Company, New York, NY. SOKAL, R. R. 1979. Ecological parameters inferred from spatial correlograms. Pp. 167-196 in G. P. Patil and M. L. Rosenzweig (eds.), Contemporary quantitative ecology and related ecometrics. Inter- national Cooperative Publishing House, Fairland, — MD. AND N. L. ODEN. 1978. Spatial autocorrelation in biology. Some biological implications and four applications of evolutionary and ecological inter- est. Biological Journal of the Linnean Society 10:229-249. AND F. J. ROHLF. 1981. Biometry, 2nd ed. W. H. Freeman and Co., San Francisco, CA. | STRANDHEDE, S. 1996. Morphologic variation and | taxonomy in European Eleocharis, subser. Palus- | tres. Pp. 1-187 in S. Strandhede (ed.), Opera | Botanica: 10. Stockholm, Sweden. | STONE, R. D., W. B. DAVILLA, D. W. TAYLOR, G. L. | CLIFTON, AND J. C. STEBBINS. 1988. Status survey > of the grass tribe Orcuttieae and Chamaesyce hooveri (Euphorbiaceae) in the Central alley of | California, 2 volumes. U.S. Fish and Wildlife Service Technical Report, Sacramento, CA. | UNITED STATES DEPARTMENT OF AGRICULTURE | (USDA), NATURAL RESOURCES CONSERVATION | SERVICE. 2006. PLANTS database (2006) [Online]. | Available: http://plants.usda.gov/. UNITED STATES FISH AND WILDLIFE SERVICE. (USFWS). 1997. Endangered and threatened wild- | | life and plants; endangered status for four plants | from vernal pools and mesic areas in northern | California. Federal Register 62:34029-34038. | . 2003. Endangered and threatened wildlife and | plants; final designation of critical habitat for four vernal pool crustaceans and eleven vernal pool plants in California and Southern Oregon. Federal | Register 68:46683—46867. WILLIAMSON, R. J., G. E. FoGG, M. C. RAINE, AND | T. H. HARTER. 2005. Hydrology of vernal pools at | three sites, Southern Sacramento Valley. Final | Report for Project F2001 IR 20. Developing a_ Floristic Statewide Vernal Pool Classification, and | Functional Model of Pool Hydrology and Water Quality. University of California, Davis, CA. ZEDLER, P. H. 1987. The ecology of southern California | vernal pools: a community profile. U.S. Fish and. Wildlife Service. Biology Report 85(7:11). MADRONO, Vol. 55, No. 4, pp. 269-279, 2008 AN ERIOPHYLLUM LANATUM (ASTERACEAE) HYBRID ZONE IN OREGON JOHN S. MOORING Biology Department, Santa Clara University, Santa Clara, CA 95053 jmooring@scu.edu ABSTRACT Eriophyllum lanatum vars. achilleoides and leucophyllum intergrade in southwestern Oregon. Some populations cannot be unequivocally assigned to either variety. Chromosome counts showed polyploid populations predominating where the varieties meet in southwestern Oregon. Pollen fertility, estimated by cotton blue-lactophenol staining, was the main criterion used to assess barriers to interbreeding. Artificial hybridizations between diploids revealed barriers to interbreeding between vars. achilleoides and leucophyllum, and between each of them and a morphologically intermediate population. The barriers to interbreeding are substantially less developed than those present between most of the other eight varieties of the FE. /anatum complex. Supernumerary chromosomes are postulated to be adaptive in Eriophyllum hybrid zones. Key Words: Artificial hybridizations, Asteraceae, barriers to interbreeding, Eriophyllum, hybrid zone, pollen fertility, species complex, supernumerary chromosomes. I dedicate this contribution to the memory of Harlan Lewis, mentor and friend. The Eriophyllum lanatum (Pursh) J. Forbes species complex extends from California to British Columbia and eastward to Utah, Mon- tana, and Wyoming. The base chromosome number is x = 8. One to four supernumerary chromosomes may be present. Diploid, tetra- ploid, hexaploid, and octoploid populations occur from coastal to subalpine communities. Artificial hybridizations showed significant differences in pollen fertility in the progeny of some intervarietal crosses, indicating barriers to interbreeding. The barriers were higher in dip- loids than in tetraploids. The “varieties” are geographic subspecies or semispecies. Many of the polyploid populations occur where the ranges of taxa overlap or nearly overlap (Mooring 2001). Assigning individuals or populations to taxa is often difficult in such areas. The morphological variability and the high frequency of polyploid populations in these areas strongly suggest that they are hybrid zones. Particularly good examples of taxonomic uncertainties occur in the Klamath Region of northwestern California and southwestern Ore- gon where E. lanatum vars. achilleoides (DC) Jeps., grandiflorum (A. Gray) Jeps., lanceolatum (J. T. Howell) Jeps., and leucophyllum (DC) W. R. Carter occur. Constance (1937, p. 89) wrote that “it seems quite impossible to assign many specimens definitely to any one subspecific category.” In particular, vars. achilleoides and leucophyllum intergrade smoothly (Constance 1937; Kenton Chambers, Dept. Of Botany and Plant Pathology, Oregon State University, oral -communication). The var. achilleoides-var. leucophyllum transi- tion region in southwestern Oregon offered an opportunity to plot the distribution of diploid and polyploid populations and to examine the extent of barriers to interbreeding between the two taxa. Mooring (2001) had reported three diploid and two tetraploid populations in Curry, Douglas, Jackson, and Josephine Counties. The Douglas County population, a tetraploid, was clearly var. leucophyllum, three other populations were var. achilleoides. The fifth population a diploid, was somewhat intermediate, but assigned to var. achilleoides. The only artificial hybridiza- tion between vars. achilleoides and leucophyllum did not use populations from southwestern Oregon (Mooring 2001, Appendix 1). A more intensive biosystematic study of both varieties in the Pacific Northwest could assess barriers to interbreeding and might suggest the extent of the hybrid zone. Here, I report the results of cytogeographic and artificial hybridization stud- ies in E. lanatum vars. achilleoides and leucophyl- lum in Oregon and Washington, and discuss the results in the context of a previous study (Mooring 2001). MATERIALS AND METHOD Plants Fruits or plants were collected from the populations listed in Appendices | and 2. Growth Conditions Black and thick fruits were germinated in vermiculite or in vermiculite-soil mixtures. Seed- lings were potted in commercial potting mixtures, 270 “UC Soil Mix” or Orchard Supply Hardware soil mix with fertilizer (12% N, 12% P, 6% K) added. The latter mix gave better results. Seedlings were grown in an unheated greenhouse at Santa Clara University. Most seedlings were transplanted to a local garden within 5—6 mo because they outgrew the pots and were damaged by white flies. Most garden plants were allowed to grow for at least two years (until they died of natural causes or had to be removed because space was needed for another crop). Meiotic Analyses Young capitula of wild, garden, or greenhouse plants were fixed in 1:3 acetic ethanol or in 1:3:6 acetic-chloroform-ethanol. Quickly placing col- lections in an ice-filled cooler usually improved fixation (Anderson 1966). Microsporocytes in acetocarmine or aceto-orcein squashes were examined by light or phase contrast microscopy at diakinesis or first metaphase. Aceto-orcein usually gave better preparations. Artificial Hybridizations Because some northwestern plants did not flower synchronously or at all, only 14 hybrid combinations were possible. Bagging experiments and decades of growing plants of this species complex show the plants to be highly self- incompatible. Pollination was done by rubbing together capitula of isolated greenhouse plants over 3-8 d. Pollen Fertility Estimates Pollen was sprinkled on a slide, a drop of cotton blue-lactophenol was added and the drop was swirled around before the coverslip was applied. The grains were stained overnight before being examined at 100 to 400 magnification. Plump, dark blue grains were scored as viable. Over 98% of the estimates rest on 300+ grains per sample, with each plant being sampled twice, on different days. No sample had fewer than 150 grains. Statistical Analyses Statistical significance of differences in percent- age pollen fertility among progenies was assessed by a one-tailed t-test using a Sigmastat program. RESULTS Cytogeography This study added chromosome counts for 16 populations from Oregon, | from Washington, and | from California (Appendix 1) to a previous report (Mooring 2001). All but three populations MADRONO [Vol. 55 were var. achilleoides or var. leucophyllum or intermediates between these taxa. The counts included two diploid and two tetraploid popula- tions of var. achilleoides, three diploid, three tetraploid, and one hexaploid population of var. leucophyllum, and one diploid, two tetraploid, and one hexaploid population of intermediates (Appendix 1). The exact number of chromosomes in the two hexaploid-level populations was unclear; determinations of 23—24 II and 23—25 II were entered for, respectively, Populations 334 and 346. The count for Population 346 is a first report of a hexaploid in var. leucophyllum. Appendix | also lists a count of 8 II in var. grandiflorum, a first report for the occurrence of that variety in Oregon, and counts of 8 II for two populations of var. lanceolatum. The diploid and polyploid populations found in the present study were not randomly distrib- uted. In respect to var. /eucophyllum populations, the diploids were in northernmost Oregon. Three of the four polyploid ones were in central and southwestern Oregon, interspersed with diploid and polyploid populations of var. achilleoides and populations intermediate between the two varieties. The intermediates are Populations 338, diploid, 339 and 335, tetraploid, and 334, hexaploid (Appendix 1). Some individuals of the intermediates averaged taller than plants of nearby populations of var. achilleoides (336, 350) and had ternately divided as well as pinnately divided leaves, similar to the description of Eriophyllum_ ternatum Greene (see Cronquist 1955, p. 196). Supernumerary Chromosomes Chromosomes in excess of the basic comple- ment, here called “‘supernumerary”? chromo- | somes, were present in diploid populations 338 | and 355 and in tetraploid populations 335 and. 337. When two supernumeraries were present | they remained unpaired or formed a bivalent. | Three of the plants analyzed in Population 338 | had one supernumerary chromosome; one plant | had two (Appendix 1). These chromosomes were | transmitted or increased in number in artificial | hybridizations (Table 1, crosses 10 and 11). Artificial Hybridizations Of the 15 hybridizations reported in this study, | aspects of those numbered 1, 2, 3, 4, 7, 9, 10, and | 11 (Table 1) have been published in parts of a longterm investigation of the E. Janatum complex | (Mooring 2001, 2007). Germination. Germination varied more in: diploid (3%-44%) than in tetraploid progenies (9%-22%) (Table 1). Mean germination, howev-. er, did not differ significantly between the diploid 2008] (23% + 14%) and the tetraploid progenies (18% + 8%). Likewise, mean germination did not vary significantly in the progeny of diploid varietal combinations: 29% + 9% for achilleoides xX achilleoides, 16% + 14% for leucophyllum xX leucophyllum, 29% + 22% for achilleoides xX leucophyllum, and 15% + 17% for achilleoides- leucophyllum intermediate xX achilleoides. Per- centage germination in the only achilleoides- leucophyllum intermediate < /eucophyllum prog- eny was 14%. Pollen fertility variability. Pollen fertility varied widely within most diploid progenies, up to 58 percentage points. Wide ranges were more frequent in Crosses 1, 2, 9, 10, and 11, which used parents from the Calistoga, California, or Dillard, Oregon, populations. Pollen fertility ranges in tetraploid progenies were much nar- rower, 3—20 percentage points (Table 1). Mean pollen fertility for the progenies of the diploid varietal combinations was 80% + 18% for achilleoides < achilleoides, 90% + 119% for leucophyllum X leucophyllum, 74% + 15% for achilleoides X leucophyllum, and 68% + 13% for the achilleoides-leucophyllum intermediate X var. achilleoides. Pollen fertility was 81% + 13% for the single achilleoides-leucophyllum intermediate xX var. leucophyllum progeny. The mean of means for these varietal combinations was approximate- ly the same as the means given above (Table 1). Mean pollen fertility for the progenies in the three tetraploid combinations (86%—98%) was much higher than the diploid combinations (Table 1, Crosses 12, 13, and 14). Mean pollen fertility in diploid progenies. With- in the var. achilleoides X var. achilleoides progenies and within the var. achilleoides X var. leucophyllum progenies, mean pollen fertilities did not differ significantly (Table 2, Crosses 1—4 and 7-8). In the var. leucophyllum X var. leucophyllum progenies, mean pollen fertility was significantly lower (P = 0.01) in the Rowena X Shepperd’s Falls progeny than in the Silver Falls X Rowena progeny (Tables 1, 2, Crosses 5 and 6). In the var. achilleoides-var. leucophyllum intermediate X var. achilleoides crosses, the mean pollen fertilities of the two progenies did not differ significantly (Table 1, Crosses 9 and 10), but the Dillard xX Calistoga one (Cross 9) differed significantly (P = 0.05) from three of the four var. achilleoides X var. achilleoides progenies (Tables 1, 2, Crosses 1, 3, and 4). In the single var. achilleoides-var. leucophyllum intermediate = var. leucophyllum cross, mean pollen fertility differed significantly (P = 0.01) from the var. achilleoides-var. leucophyllum intermediate X var. achilleoides progeny. The Population 338 parent in the two crosses came from different plants (Table 1, Crosses 9 and 11). MOORING: ERIOPHYLLUM HYBRID ZONE 274 Parent-progeny pollen fertility differences. In the eight crosses among diploids where parental pollen fertility was known, mean pollen fertility in five progenies was lower by 11%-—22%, and in three crosses it was higher by 1%-—11%. In the three tetraploid crosses, pollen fertility equaled or was higher (19%, 25%) than that of the parents (Table 1). Meiotic Pairing Tetraploids were not examined. Diploids formed 8 II at diakinesis or metaphase in the progeny of six of the ten crosses where microspo- rocytes could be studied. In Crosses 6 and 7, 8 II was the norm, but some 7 II + 2 I configura- tions were seen, as well as one apparent bridge- fragment in Cross 7 (Table 1). One or two supernumerary chromosomes were present in the progeny of Crosses 9 and 10, where parent S338-4 had one supernumerary chromosome, and Cross 11, where parent S338-6 had two (Table 1). DISCUSSION Systematics Eriophyllum lanatum var. leucophyllum (DC.) W.R. Carter extends from southwestern Oregon to the Columbia River, thence eastward to approximately Hood River, Oregon, north in the Cascade Mountains and coastal regions of Washington to southwestern British Columbia. To the east, var. /eucophyllum intergrades with var. lanatum. Constance (1937) included var. leucophyllum in var. lanatum, considering it to be a western expression of var. /anatum. Mooring (2001), relying on morphological and ecological information, and using results from artificial hybridizations, believed that the western popula- tions of var. /anatum (sensu Constance 1937) were best treated as var. leucophyllum (DC.) W. R. Carter. The results of artificial hybridizations using four populations indicated strong barriers to interbreeding between the var. /anatum and var. leucophyllum populations. Variety achilleoides (DC.) Jepson ranges from central Oregon and extreme northwestern Ne- vada south to San Luis Obispo County in coastal California. In the North Coast Ranges of California, it intergrades with var. arachnoideum (Fisch. & Avé-Lall.) Jeps. to the west and with var. grandiflorum to the east. In the North Coast Ranges, 36 of 66 populations of vars. achilleoides, arachnoideum, and grandiflorum or intermediates between vars. achilleoides and grandiflorum or between vars. achilleoides and arachnoideum were polyploid, including 18 of the 19 intermediates (Mooring 2001). Variety achilleoides consists of regional clusters of populations. Constance (1937, p. 87-88) ob- Ic [Vol. 55 ({ aseyd) esojsi[eD ‘eden +JI8‘ITI+I8 18 0¢ CG IT + 69 68-0S 96 08 (LZ) 6L/TT C-S-8ITIS' X #-8EES' ‘VO x preyid ‘sesnogq :yO ‘6 SAGCIOATUHOYV “AVA X ALVIGAINUALNI WATTAHdOONATSACIOATHAOV ‘QUO TdId Sl + ML %CCT + %OC = ub = UPI s][e4 JoATig ‘uolepP, -YO x Is Ol Ll 91 + 8L v6-0S C6 96 (C1) 6rI/6l UY I[-TPE X SLE (Z aseyd) Yyorog PlOH “A1mND “YO ‘8 JUSUIS BAY -a3pliq pue Q[IAouuog ‘yeuoulynp] -WO x I7T+IL MYC TITS § ¢ Il + 99 I8-IS GN “GN (pr) SC/TI CLUS X FIGS ({ eseyd) 9YRT JOMOT OYBT WO OL WATTAHdOONAT UVA X SACIOATTIHOV “AVA -AlOTdIG %I1 + %06 %bvl + MII = ub = ub sjje4 s.proddays ‘yemouynyy IT+I.Lpuells 9 cl + 18 06-£9 CN ‘66 (TE) ELL/IE COLE X CCrES “aO XK PUSMOY “OOSPM AO 9 s][ey Ioayig ‘Uoleyy CIN. — £ 0 L + V6 66-98 16 L6 (9) IS/€ SAVE XC Ore “AO X BUBMOY “OSSPM AO “AS VUIMOY ‘OSSBAA 10 Ee eee fo ¢ + 96 OOI-r8 L6 ‘16 (OI) 16/6 C-CVE X E-IVE “AO XK SRA A9ayig “UoeYy -YO “eS ‘o) WATTAHdOONAT AVA X WATTAHdUOONYT “AVA ‘CIOTdIG ia %8l F BOL 6 F %6e we = UPI] = urs (Z aseyd) yodsoig ‘uosyore -YO z. Ws ¢ 8t pl + $8 6-69 86 09 (LZ) ELTISL £-CLES X< I-C#CTS X(T BSPYd) PIPYYOWT] i v 2 SOIR) SOT “PIRID ae see Ws ¢ 8+ 18 68-89 CN °86 (81) pOr/ol SE X ESLES x (T aseYyd) Wodsoig ‘uosyoee YO '€ (T aseyd) aftAsosjayy “oxPT “WO GN Cl a= 61 + Ll 86-0F C8 £8 (6€) 9S/CT STO) KALSE x ({ aseyd) esorsyTRD “edeN :wO 7 (7 oseyd) Bsojsiped ‘eden :vO IL-3 cl Lo Ot + 08 86-0S 6°98 (O¢) Lr/rl og hata Mae 4 i. Wa x (T oseyd) esoisyTRD “edeN :WOD 1 SAGIOATUIHOV “AVA X SICIOATTIHOV “AVA :-GilOTdIG stsouryerp IO [TA u % ds | + ues osuey = URI (%) sa1suonbosy sJUdIe uoneoo] Aq — ssoip ye Sulsied ‘sisompy ‘sjuoied jo ———— UONRUIULIDD ueoul WOLy sJUsIed DOUdIII JIG IA uajod sjqvurejs o8v]Us0I0g ‘ugtjod Surystuimy Auosoid Jo JOQUINU 9} O} SIOJOI ..N,, “ATQR[IVAR UOTVULIOJUT IO BYep OU = CIN ‘SSOIS jeooid1oat = ..Y,, “ISI post] st yuored pods dU, ..’S,, UB YOR] siN}eU WHOL sjuURfdsues ‘s}UR[d UMOIB-posos 0} SAOJOI ..S,, UOHe[ndod jey) Jo uR]d vB puodsss dy} ‘uUoNLiNdod ay) sjuasoidos Joquunu Is ay} :s}usIed Ajuopt sioquinu pozioyel ‘saploayjiyop “rea ul suoneindod Jo siaysnjJo [RUOIBII O} SIOJOI ..OSPUd,, “KWID Jsasopo *AjUNOD (UOWUIYSePM = VA 'UOSAIOC = YO ‘PUIOFYeD Pe VD) 19181S dB SUONBIOT “UOISSNOSIP d}e}TTIOVJ OF POTOQUINU dIR SOSSOID “XATANOD WALVNVT WATIKHdOIYY AWL NI SNOILVZIGRIAAH TVIOMILYY "| ATavL MOORING: ERIOPHYLLUM HYBRID ZONE 2008] %% + M6 %8 + %8l = uvou plojdenay = uvow projdesay PUIPTYIVD “UN YeIYYeA “VM CGN § Sct L + 98 Lo-LL SL C9 (6) IvI/CI UY [-KrES' X L-SEES' x WO” [BUS “UOSyoeE -YO “FI WATTAHdOONAT UVA X ALVIGAINAAINI W2TTAHdOONATSACIOdTTTIAHOV “AVA ‘GIO TdV ALAL (T SCY) ISARY ansoy ‘ALND -WO X (7 wsveuyd) GN Cc 6[+ Cc + 96 L6-v¥6 66 “79 (CZ) SOP/L8 Ud Y S-8rES X L-SEES' jUIod [P1}UZ) “UOsyIPL YO ‘el (c aseYd) IOATY onsoy “AlIND :-YO x ({ Iseyd) GN 9 0 C + 86 OOI-F6 66 “86 (TC) ILI/8E U BY o-8SrEs xX T-6Fr lS ssuLidg Inq]iM “Psnjod -WO TI SAdCIOATUAOIY “AVA X SHCMIOATTIAOV “AVA -GlOTdVaELAL re BUIMOY “OSS AA +II8 TI+H8 18 YI EEE €Il + 18 b6-th S819 (pI) 8S 1/77 9-CPES X 9-8EES' ‘YO X Pslelid ‘sepsnogd -yo ‘TI WATTAHdOONAT AVA X ALVIGAINGALINI WATTAHdOONATSACIOATHUHOV -QlOTdId %el + %89 %LI + MSI = ura = urs (Z aseyd) s1[vH ‘ourydasor IT+H8 18 61 — Oc + 9 v6-IS 8Z ‘08 (€) S6S/SI UY €1-OS€S' X F-8EES' YO X pleyid ‘sesnog :yO ‘Ol Ssisouryerp 10 TA u % dS [ + ues osury = URI (%) saouanbary sJUdIvg uoT}ed0] AQ — SSOID ye Sutsited ‘sIsorypy ‘sjuoied jo oe UONVUIULIOD URS WOLY SJUSIed DOUdIATJIC] Id uajod s[qeuleys o8v}UD0I0g GaANNILNO,) ‘[ ATaV LE 274 TABLE 2. numbered as in Table 1. Cross comparison . Dillard < Calistoga vs. 1. Calistoga Calistoga . Dillard < Calistoga vs. 3. Prospect x Los Gatos . Dillard X Calistoga vs. 4. Litchfield < Prospect . Dillard < Calistoga vs. 11. Dillard x Rowena OOO OWN served that two major groups of populations were present, the “‘typical race’’ of the North Coast Ranges and a “‘somewhat different population” of the Klamath region of California and Oregon. My conclusions differ; three major and widespread intergrading regional phases can be distinguished in var. achilleoides, as well as one local phase. Phase 1 occurs from San Luis Obispo County to approximately Tehama County in California. Phase 2 occurs in northwestern California and southwestern Oregon (the Klamath area). Phase 3 occupies more arid regions than Phase 2, with which it smoothly intergrades, occurring in northeastern California, the northwestern corner of Nevada, and southeastern Oregon as far north as Harney County. Generally, Phase 3 popula- tions consist of smaller plants than those of Phase 2 (some populations appear to have been intro- gressed by var. integrifolium (Hook.) Smiley.) The local phase, including some populations treated as Eriophyllum Cusickii Eastw. by Eastwood and E. ternatum by Greene (Constance 1937, p. 88) centers in the Ashland to Medford area of southwestern Oregon. Cronquist (1955, p. 196) noted that taller plants, often with ternately divided leaves, replaced individuals with pinnately divided leaves, and commented they “‘are possibly best considered a distinct variety.””’ Morphologi- cally, these populations appear to provide a continuum between vars. dchilleoides and leuco- phyllum, being intermediate in respect to charac- ters that can be used to distinguish modal populations of the two varieties (Table 3). A massive root system, abundance of root buds, and the clumped nature of the plants TABLE 3. LEUCOPHYLLUM. Character Root system taprooted, slight Root buds absent Individuals single Life span annual?, biennial, short-lived perennial, often flowering the first year Leaves +crispate, gray- woolly beneath MADRONO . Rowena X Silver Falls vs. 6. Rowena < Shepperd’s Falls . Prospect X Los Gatos vs. 7. Lower Lake X Bonneville Var. achilleoides [Vol. 55 STATISTICALLY SIGNIFICANT DIFFERENCES IN POLLEN FERTILITY BETWEEN CROSSES. Crosses are P value <0.01 <0.05 <0.05 <0.05 <0.05 <0.01 characterized the natural populations of var. leucophyllum that I examined. Some of these populations appeared to consist of one or more clones. What appeared to be separate plants were connected by tough roots (not rhizomes), e.g., 337, 340. The characters noted in Table 3 largely persisted in cultivation. The massive root system was especially noticeable. The root system of a single plant, and the shoots that arose from root buds, usually filled a 5-inch pot so completely that it was difficult to remove the plant. In garden culture, individuals had a diameter of up to 1.5m, and spread by new shoots arising from root buds. Some garden plants produced 15-30 flowering stems, others had not flowered after three years. A massive root system was not evident in the natural populations of var. achilleoides that I looked at in southwestern Oregon (Appendix 1). Most were + taprooted. Massive root systems, however, developed in greenhouse and garden plants of var. achilleoides populations 348 and 349. A massive root system was present to some degree in nature and in cultivation in the populations I treated as intermediate between vars. achilleoides and leucophyllum. That root system was best developed in Population 334 (hexaploid) and 335 and 339 (tetraploid), less well developed in 338 (diploid), except in cultivation. Cytogeography The diploid and polyploid populations of vars. leucophyllum and achilleoides and the achilleoides- | leucophyllum populations identified in this study | MAJOR DIFFERENCES BETWEEN MODAL POPULATIONS OF VAR. ACHILLEOIDES PHASE | AND VAR. Var. leucophyllum fibrous, massive abundant often clumped longer-lived perennial, often not flowering the first year +plane, white- woolly beneath 2008] O 2x, @ 4x, A 2x, & 4x, A 6x, var.leucophyllum 0 2x, Mi 4x, ial 6x, var. achilleoides intermediates Fic. 1. Distribution of diploid, tetraploid, and hexa- ploid populations of Eriophyllum lanatum var. achil- leoides, var. leucophyllum, and var. achilleoides-var. leucophyllum intermediates in British Columbia, Wash- “ington, Oregon, and northern California. Not all the California populations are shown; see Fig. | and 2 of Mooring (2001) for more detail. MOORING: ERIOPHYLLUM HYBRID ZONE 213 and a previous one (Mooring 2001) were not randomly distributed. In respect to var. leucophyllum, the 12 north- ernmost cytodemes were diploid, extending from British Columbia to Pierce, Kittitas, and Yakima counties in Washington (Fig. 1, Appendices 1, 2). Southwestern Washington and northwestern Oregon had four tetraploids and five diploids. Most were close to the Columbia River. The easternmost diploid population (342) showed some features of var. /anatum, with which var. leucophyllum intergrades along the Columbia River. Two other var. Jeucophyllum diploid populations were farther south, in Marion County, Oregon. South of them were var. leucophyllum tetraploid populations in Lane, Linn, and Douglas counties and a morphologi- cally aberrant hexaploid in Coos County. Variety achilleoides cytodemes in Oregon and northernmost California were diploid except for three tetraploids in southwestern Oregon (Fig. 1, Appendices 1, 2). Differentiating between vars. achilleoides and leucophyllum becomes increasing difficult from approximately Roseburg to Ash- land, Oregon. In this region, a mixed pattern of diploid and polyploid populations of both varieties and their intermediates occurs. Here, in Douglas, Josephine, and the western portion of Jackson counties, six populations were polyploid (1S, 337, 335, 339, 334, 62) and five were diploid (92, 173, 338, 336, 350). Populations 334, 335, 336, 337, 338, 339, and 350 were grown in greenhouse or garden culture, or both. I regarded 334, 335, 338, and 339 as var. achilleoides-var. leucophyllum intermediates, 78 and 337 as var. leucophyllum, and 92, 173, 336, 62, and 350 as var. achilleoides (Appendices 1, 2; Fig. 1). The morphological intermediacy of the poly- ploids and their lack of multivalents suggest allopolyploidy. I recorded counts of 17 II in Population 335, 17-18 II in Population 337, 23— 24 II in Population 334, and 23-25 II in Population 346. The presence of two extra chromosomes in Population 335 may be an instance of aneuploid numbers generated by neopolyploids (see Ramsey and Schemske 2002, pp. 601-607). Alternatively, the chromosomes in excess of the basic complement may be supernu- meraries, as in the diploid populations 338 and 355 (Appendix 1). The variation in number of chromosomes in Populations 337, 334, and 346 may reflect errors in counting chromosomes in “‘difficult’” preparations. Alternatively, it may be a matter of aneuploid numbers or supernu- merary chromosomes as noted above for Popu- lation 335. Supernumerary Chromosomes Supernumerary chromosomes are known in six of the ten varieties of Eriophyllum lanatum. 276 Their frequency in individuals and populations of var. achilleoides was 8/85 (9%) and 8/61 (13%), respectively, in var. l/eucophyllum 6/37 (16%) and 6/29 (21%) (Mooring 2007). Artificial hybridizations have shown that the pollen and the seed parent can transmit or increase the number of supernumerary chromosomes in vars. achilleoides and leucophyllum (Mooring 2007). Three of the plants analyzed in Population 338, intermediate between vars. achilleoides and leucophyllum, had one supernumerary chromo- some; one plant had two. In artificial hybrid- izations these chromosomes were transmitted or increased in number (Table 1, Crosses 10 and gle P Eriophyllum lanatum populations with super- numerary chromosomes were not randomly distributed. Their frequency was higher in regions where the varieties are sympatric or nearly so (Mooring 2007). Mooring (2007) hypothesized that these supernumerary chromosomes could originate by hybridization between chromosom- ally differentiated populations, especially in regions where the varieties are sympatric or nearly so. Do these supernumerary chromosomes remain in populations because they are transmit- ted or increased in number by meiotic mecha- nisms? Neutral or even deleterious gene combi- nations could be maintained in this way. Alternatively, I speculate, that some supernumer- ary chromosomes could persist in FE. /anatum populations because they are the chromosomal equivalents of the transposable genetic elements found in maize and other species. Reserves of genetic variability are potentially adaptive, espe- cially in hybrid zones with a variety of environ- ments, including ecologically marginal ones, e. g., southwestern Oregon. Barriers to Interbreeding Components of interbreeding. In an elegant essay, Ornduff (1969) epitomized interbreeding as a succession of three related events: “relative crossability” (pollination, fertilization, and mat- uration of the seed), “‘progeny establishment” (germination, growth, and maturation of the offspring), and “‘fertility of the progeny.” Infor- mation obtained from greenhouse and garden studies cannot be uncritically applied to inter- breeding in natural populations. However, Orn- duff’s (1969) “‘relative crossability” and “‘progeny establishment” have a higher chance of success under greenhouse and garden conditions than under field conditions. Therefore, 1t seems likely that barriers to interbreeding are higher under natural conditions than under artificial ones. Criteria for assessing barriers to interbreeding. To assess differences in barriers to interbreeding, I relied mainly on percentage pollen fertility in MADRONO [Vol. 55 the progeny of the crosses and percentage change in pollen fertility between the means for the progeny and the means for their parents. Differences in percentage germination could be an important indicator of another isolating mechanism, but in the present study percentage germination could be misleading for two reasons. First, dark and thick fruits may not have a viable embryo or the embryo may not be able to penetrate the seed and fruit coats (probing the fruits to identify those with embryos improves the validity of the percentage germination criterion). Second, germination was staggered. Within a lot, most seedlings came up within 6-14 d; others required up to 30. The conditions of my study required seedlings be transplanted to pots within about three weeks after the first seedlings of that lot appeared. Other factors influencing inter- breeding are meiotic aberrations and F1 vigor. In this study, conspicuous differences among prog- enies in meiotic pairing and F1 vigor were usually not apparent. Caveats. 1) Assessing pollen fertility by staining techniques has limitations (see Stace 1980). Also, in this study some pollen grains in some hybrid combinations showed a continuum of intermedi- ate conditions of plumpness and staining. I evaluated some intermediate pollen grains as fertile. Therefore, pollen fertility in some hybrid combinations may be slightly lower than my estimates. 2) As to differences in pollen fertility between parents and progeny, low pollen fertility in one parent lowers the average fertility, thus decreasing the validity of this criterion. For example, in Crosses 4 and 11 the mean pollen | fertility of the progeny was higher than that of | the mean of the parents. But, the pollen fertility | of the parents in each cross differed substantially, | 60% and 98%, and 61% and 85%, respectively. 3) | Crosses other than 6 and 7 showed no evidence of | chromosomal structural changes at diakinesis or | first metaphase. Jackson (1984) has noted that | study of pachytene stages might reveal these | changes. Barriers to interbreeding within var. achilleoides. | No significant differences in mean pollen fertil- ities or other barriers to interbreeding were | observed in var. achilleoides X var. achilleoides progenies despite the distance and environmental and other differences between the Oregon and California populations (Tables 1, 2, Crosses | 1-4). Barriers to interbreeding within var. leucophyl-. lum. Only two interpopulation crosses were | possible within var. /Jeucophyllum. Both involved | Oregon populations, Rowena x Silver Falls, and | Rowena X Shepperd’s Falls (Tables 1, 2, Crosses | 5 and 6). The mean pollen fertilities of the pro- | genies, 95% + 5% and 81% + 12%, respectively, | 2008] differed significantly, P = 0.01. Interestingly, mean percentage germination in Rowena X Silver Falls was only 8%, contrasting with its 95% pollen fertility. Some cells of a Rowena X Shepperd’s Falls plant formed 7 II + 2 I, suggesting the possibility of chromosomal re- structuring (Table 1). The mean percentage pollen fertility of 95% seemed astonishingly high in the progeny of the Rowena X Silver Falls reciprocal cross. Eight of the ten plants had pollen fertilities of 95%—99%, a percentage that would be unusual even in intrapopulation crosses. The available evidence suggests that the high pollen fertility in this progeny did not result from frequent self pollination induced by human rather than natural pollinations. The mean pollen fertility of progenies of the reciprocal crosses were 96% and 94% (Table 1, Crosses 5a and 5b), and the lowest pollen fertility in any of the progenies was 84%. It is unlikely that high percentages of selfing would occur in both seed parents. Evidence supporting frequent selfing in Populations 34/ and 342 1s scanty. Plants of Population 34/ (Silver Falls) were seed parents in a total of three other interpopulation crosses; mean pollen fertilities in the progenies were 65%—78%. Plants of Popula- tion 342 (Rowena) were seed parents in a total of six other interpopulation crosses; mean pollen fertilities in the progenies were 49%-—88%. Of the total of 39 plants in both sets of progenies, only 5 had pollen fertilities of 90%—94%. Rather than frequent self pollination, the 95% mean pollen fertility in the Silver Falls X Rowena progenies most probably resulted from recombination. Recombinations yielding higher than average fertility would be advantageous under natural conditions, providing, of course, that such geno- types would be otherwise fit. See Arnold (1997, Chapter 6) for discussions of natural hybridiza- tions leading to new evolutionary lineages. Barriers to interbreeding between var. achil- leoides and var. leucophyllum. The presence of some 7 II + 2 I configurations and a bridge- fragment in the Lower Lake xX Bonneville progeny suggests the possibility of an inversion and perhaps other chromosome rearrangements. Percentage germination, however, was relatively high. The mean pollen fertilities of the two var. achilleoides X var. leucophyllum progenies did not differ significantly. The mean pollen fertility of the Lower Lake X Bonneville progeny, however, differed significantly (P = 0.05) from the var. achilleoides X var. achilleoides progeny, Prospect x Los Gatos (Table 2, Crosses 3 and 7). Barriers to interbreeding between a var. achil- leoides-var. leucophyllum intermediate and var. achilleoides. Population 338, Dillard, Ore- gon, is a var. achilleoides-var. leucophyllum ‘intermediate. Plant S338-4, with one supernu- MOORING: ERIOPHYLLUM HYBRID ZONE 254 merary chromosome, was used in the Dillard x Calistoga, California, and the Dillard * Galice, Oregon crosses. The progeny included plants with one or two supernumerary chromosomes, indicating that these chromosomes can be trans- mitted or increased in number (Mooring 2007). Mean pollen fertilities of the progenies were similar, 69% + 11% and 64% + 20%, respec- tively. These means were 22% and 19% lower, respectively, than the averages of their parents (Table 1, Crosses 9 and 10). Mean pollen fertility in the Dillard xX Calistoga progeny differed significantly (P = 0.05) from the following var. achilleoides X var. achilleoides progenies: Calis- toga xX Calistoga, Litchfield x Prospect, and Prospect < Los Gatos) (Table 2, Crosses 1, 3, 4, and 9). The Calistoga and Los Gatos populations are Phase | of var. achilleoides; the Litchfield and Prospect populations are Phase 2. In Cross 10, Dillard =X Galice, mean pollen fertility of the progeny did not differ significantly from the var. achilleoides X var. achilleoides progenies noted above (Table 2). Percentage germination, howev- er, was 3%, and only 4 of the 15 progeny flowered (Table 1). Barriers to interbreeding between a var. achil- leoides-var. leucophyllum intermediate and var. leucophyllum. Population 338, Dillard, Ore- gon, is a var. achilleoides-var. leucophyllum intermediate. Plant S338-6, with two supernu- merary chromosomes, was crossed to a var. leucophyllum plant from a Rowena, Oregon population. One or two supernumeraries were transmitted to five of the six progeny (Mooring 2007). Mean pollen fertility of the progeny (81% + 13%) was significantly higher (P = 0.01) than that of the Population 338 X var. achilleoides progeny (69% + 11%), suggesting that barriers to interbreeding are less well developed between Population 338 and var. leucophyllum than between Population 338 and var. achilleiodes (Tables 1, 2, Crosses 9, 10, and 11). Comparisons of pollen fertilities. Mean pollen fertilities in the progenies of diploid crosses involving var. achilleoides, var. leucophyllum, and Population 338, intermediate between the two varieties, were 64%-81%, averaging 78% (Table 1), much higher than mean pollen fertil- ities reported for other intervarietal crosses in the E. lanatum complex (Mooring 2001). In the 2001 study, ranges for low, medium, and high mean pollen fertilities were set at 22% 40%, 42%-58%, and 60%-—76%, respectively (Mooring 2001, Table 7). Barriers to interbreeding do not seem to be well developed among var. achilleoides, var. leucophyllum, and Population 338, if pollen fertility is used as a criterion for assessing such barriers. However, the possibility that the super- numerary chromosomes in Population 338 pro- moted higher pollen fertility has to be considered. CONCLUSION The geographic distribution of diploid and polyploid populations in E. /anatum vars. achilleoides and leucophyllum, and intermediates between these two taxa in southwestern Oregon portrays a familiar pattern in the FE. /anatum species complex: polyploid populations predom- inate in regions where infraspecific taxa meet. The relatively high pollen fertility in the progenies of artificial hybrids in this study suggests that barriers to interbreeding between vars. achilleoides and leucophyllum are relatively low in southwestern Oregon. The capacity to produce fertile hybrids from crosses between taxa accounts for some of the difficulty in deciding on taxonomic designations for inter- mediate populations. Also, low barriers to interbreeding increase the probability of produc- ing genotypes leading to new evolutionary lineages, the “‘evolutionary novelties’ of Arnold (1997, Chapter 6). The geological antiquity, relatively benign climate, and the topographic and edaphic diversity of southwestern Oregon (Hunt 1974) increase the chances of new and successful evolutionary lineages. The popula- tions characterized as “‘Eriophyllum ternatum” (see Cronquist 1955, p. 196) may be examples of such lineages. Polyploid populations in Oregon largely bridge the gap between the diploid var. /eucophyllum populations of British Columbia and Washing- ton, and the diploid var. achilleoides populations of northern California and southwestern Oregon. If the Oregon polyploid populations identify a history of natural hybridizations, then the hybrid zone may extend from the Columbia River to the Siskiyou Mountains of southwestern Oregon and northwestern California. Some features of what I have termed “‘Phase 2” of var. achilleoides may reflect the introduction of genes from var. leucophyllum. Molecular level studies might define the extent of the hybrid zone, indicate whether vars. grand- iflorum and lancelolatum are involved in the taxonomic complexity in the Siskiyou Moun- tains, and find if clines are present. ACKNOWLEDGMENTS The late Kathryn Halloran Mooring helped with field work, expert proof reading, and other assis- tance. Craig Stephens and Michelle Marvier provided laboratory facilities after retirement. An _ internal grant supported the project. I thank Robert Numan for his guidance with statistical procedures, Kenton Chambers for his knowledge of Oregon botany, Katherine Preston for suggestions on organizing the manuscript, John Strother for editorial expertise in pre-submission reviews, and Kenton Chambers and an anonymous reviewer for their post-submission comments. MADRONO [Vol. 55 LITERATURE CITED ANDERSON, L. 1966. Cytotaxonomic studies in Chry- sothamnus Astereae, Compositae). American Jour- nal of Botany 53:204-212. ARNOLD, M. 1997. Natural hybridization and evolu- tion. Oxford University Press, New York, NY. CONSTANCE, L. 1937. A systematic study of the genus Eriophyllum Lag. University of California Publica- tions in Botany 18:69—136. CRONQUIST, A. 1955. Vascular plants of the Pacific Northwest, Part 5: Compositae. University of Washington Press, Seattle, WA. HUNT, C. 1974. Natural regions of the United States and Canada. W. H. Freeman, San Francisco, CA. JACKSON, R. 1984. Chromosome pairing in species hybrids. Pp. 67-86 in W. F. Grant (ed.), Plant systematics. Academic Press, New York, NY. MOORNING, J. 2001. Barriers to interbreeding in the Eriophyllum lanatum (Asteraceae, Helenieae) spe- cies complex. American Journal of Botany 88:285—312. . 2007. Supernumerary chromosomes in Erio- Phyllum lanatum and E. confertiflorum var. con- fertiflorum (Asteraceae). Madrono. 54:30-41. ORNDUFF, R. 1969. The systematics of populations in plants. Pp. 104-128 in Systematic Biology, Pro- ceedings of an International Conference. National Academy of Sciences, Washington, DC. RAMSEY, J. AND D. SCHEMSKE. 2002. Neopolyploidy in flowering plants. Annual Review of Ecology and Systematics 33:589-639. STACE, C. Plant taxonomy and biosystematics. Edward Arnold, London, U.K. APPENDIX 1 UNPUBLISHED 2N CHROMOSOME COUNTS IN THE ERIOPHYLLUM LANATUM COMPLEX The county name follows the state name. The first number is that of the population, the second is my collection number. Voucher specimens have been deposited in SACL. Counts are for one plant, unless a number in parentheses shows more. Locations are approximate. Populations used in artificial hybridiza- tions (Tables 1, 2) are in bold print. E. lanatum (Pursh) Forbes var. achilleoides (DC.) Jepson OREGON: JOSEPHINE, 336, 3999, 8 II, Granite Hill Road, Grants Pass; 350, 4017, 8 II (2), near Galice. CURRY, 348, 4015, 16 II, hwy 33 between Gold Beach and Agness; 349, 4016, 16 II (2), junction of hwys 33 and 23 near Agness. E. lanatum (Pursh) Forbes var. leucophyllum (DC.) W. R. Carter | OREGON: MARION, 341, 4004, 8 II, near Silver Falls — State Park. MULTNOMAH, 355, 4030, 8 II, 8 If + 2 I, | near Shepperd’s Dell State Park. WASCO, 342, 4005, 8 II (3), near Rowena along old hwy 30. DOUGLAS, 337, | 4000, 16 II, 17-18 II, near Azalea. LANE, 340, 4003, 16 | II, between Elmira and Monroe. COOS, 346, 40/3, 23—25 II, near Bridge. WASHINGTON: WAHKIAKUM, 344, 4011, 16 il, | along Columbia River near Cathlamet. Var. achilleoides-var. leucophyllum intermediates. OREGON: DOUGLAS, 338, 4001, 8 I+ 11 (3), 8 + 2 Tor 9 II, near Dillard; 339, 4002, 16 II, near Yoncalla. | 2008] JACKSON, 335, 3998, 17 II, near Central Point; 334, 3997, 23-24 II, near Emigrant Lake. E. lanatum var. grandiflorum (A. Gray) Jepson OREGON, JOSEPHINE, 35/, 40/8, 8 Hf, near O’Brien on road to Happy Camp. E. lanceolatum (Howell) Jepson OREGON, JOSEPHINE, 352, 40/9, 8 II, ca. 15 mi N of Happy Camp. CALIFORNIA, SISKIYOU, 353, 4020, 8 II, 9 mi N of Happy Camp on road to O’Brien. APPENDIX 2 PREVIOUSLY PUBLISHED 2N CHROMOSOME COUNTS IN THE ERIOPHYLLUM LANATUM COMPLEX The name of the county follows that of state. The first number is that of the population, the second is my MOORING: ERIOPHYLLUM HYBRID ZONE 219 collection number. Voucher specimens have been deposited in SACL. Locations are approximate. Pop- ulations represented in artificial hybridizations (Ta- bles 1 and 2) are in bold print. E. lanatum (Pursh) Forbes var. achilleoides (DC.) Jepson CALIFORNIA: Colusa, 149, 2073, Wilbur Springs, 16 II. Lake, 120, 1913, Kelseyville, 8 Il; 727, 1916, Lower Lake, 8 II. Lassen, 242, 3125, Litchfield, 8 II. Napa, 118, 1908, Calistoga, 8 II. Santa Clara, 38, 1375, Los Gatos, 8 II (5). OREGON: Curry, 347, 4014, Gold Beach, 8 II. Jackson, 173, 2158, Mt. Ashland, 8 IT; 275, 2495, Prospect, 8 II. Josephine, 92, 1576, Sunny Valley, 8 II; 62, 2260, Selma, 16 I. E. lanatum (Pursh) Forbes var. leucophyllum (DC.) W. R. Carter OREGON: Douglas, /8, 1056, Whistlers Bend State Park, 16 Il, 16 If + 1 I. Multnomah, 167, 2147, Bonneville, 8 IT. MADRONO, Vol. 55, No. 4, pp. 280-284, 2008 MORPHOLOGICAL AND CYTOLOGICAL EVIDENCE FOR HOMOPLOID HYBRIDIZATION IN JOCHROMA (SOLANACEAE) STACEY DEWITT SMITH!, VANESSA J. KOLBERG, AND DAVID A. BAUM Department of Botany, University of Wisconsin, 430 Lincoln Drive, Madison, WI 53706 stacey.smith@duke.edu ABSTRACT Previous phylogenetic and biogeographic studies of Jochroma (Solanaceae) suggested that J. ayabacense S. Leiva is a hybrid between I. cyaneum (Lindl.) M. L. Green and I. lehmannii Bitter. Chromosome counts for these three taxa demonstrate that all have a haploid chromosome number of n = 12, and thus that the formation of 1 ayvabacense did not involve changes in ploidy level. A comparison of vegetative and floral morphology revealed that 1. ayabacense demonstrates a striking intermediacy between [. cyaneum and I. lehmannii. The results support the conclusion that /. avabacense 1s a homoploid hybrid between 1. cyaneum and I. lehmannii. RESUMEN Estudios filogenéticos y biogeograficos de Jochroma han sugerido que I. ayabacense S. Leiva es un hibrido entre 7. cyaneum (Lindl.) M. L. Green y I. lehmannii Bitter. Recuentos de cromosomas definen un numero haploide de 7 = 12 para estas especies y asi indican que la formacion de J. ayabacense no involucro un cambio en el nivel de ploidia. Una comparacion de la morfologia vegetativa y floral revelo que J. ayabacense tiene un forma intermedia entre 1. cyaneum and I. lehmannii. Estos resultos respaldan la conclusion que 1. avabacense es un hibrido homoploide entre 1. cyaneum y I. lehmannii. Key Words: Amotape-Huancabamba, chromosome number, homoploid hybridization, Jochroma, Solanaceae. Interspecific hybridization has long been rec- ognized as an important generator of diversity in flowering plants (Grant 1981; Rieseberg 1997). Hybrid speciation can occur via two modes: allopolyploidy, in which the hybrid exhibits twice the chromosome number of its parental species, and homoploidy, in which the hybrid has the same ploidy as both parents (Arnold 1997). Allopolyploidy is considered to be the more common mode of hybrid speciation because the change in chromosome number creates an instant reproductive barrier between the hybrid and its parents, facilitating persistence of the hybrid species (Ramsey and Schemske 1998; Chapman and Burke 2007). However, theoretical work suggests that homoploid hybridization can also lead to stable hybrid lineages (Buerkle et al. 2000), either though recombinational speciation (Stebbins 1959; Grant 1981) or through ecolog- ical selection favoring novel hybrid phenotypes (Rieseberg et al. 1999). While the number of clear examples of homoploid hybrid speciation re- mains small (Rieseberg 1997), combining molec- ular phylogenetic studies with traditional lines of evidence holds promise for detecting additional instances (e.g., Ferguson and Sang 2001; Ho- warth and Baum 2005). Here we apply cytolog- ical and morphological evidence to examine the 'Current address: Department of Biology, Box 90338, Duke University, Durham, NC 27705. case of Jochroma ayabacense, a species suspected of hybrid ancestry based on recent molecular phylogenetic studies (Smith and Baum 2006). Iochroma is a genus of approximately 25 species distributed throughout the northern Andes, from Colombia to Peru. Based on conflicting placement across gene trees, Smith and Baum (2006) identified three suspected hybrid taxa in Jochroma, of which I. ayabacense provided perhaps the most compelling case. In phylogenetic analyses of three nuclear regions (ITS, LFY intron II and waxy), I. avabacense was found to contain “‘divergent” alleles, which fell either in the [ cyaneum group or in the JL lehmannii-I. squamosum group (Smith and Baum 2006). We consider /. squamosum S. Leiva and Quipuscoa a synonym of J. lehmannii and will treat it as such in this paper. This phylogenetic information together with the geographic distri- bution of J. ayabacense strongly suggested that it) is the product of hybridization between J. lehmannii and I. cyaneum. In this paper, we examine the morphological features of 1. ayaba-. cense relative to its putative parents, and report: on the results of chromosome counts undertaken to determine whether the formation of J.) ayabacense involved a change in ploidy level. METHODS I. ayabacense occurs exclusively in the environs. of Ayabaca, Peru, near populations of both L 2008] FIG. 1. Geographic distribution of study taxa. Circles are sites where [ cyaneum has been collected and squares where /. l/ehmannii has been collected. / ayabacense is only found in Ayabaca, Peru, indicated by the star. cyaneum and I. lehmannii (Fig. 1). Buds for chromosome counts were collected from individ- uals of ZL ayabacense and I. lehmannii near Ayabaca, and buds from J. cyaneum were collected from a greenhouse-grown accession. Vouchers for these sources are given in Appendix 1. The buds, ranging from 4-20 mm in length, were fixed in Carnoy’s solution (3 parts chloro- form: 2 parts ethanol: 1 part acetic acid) for 12— 24 hr and then transferred to 70% ethanol for storage. Slides of meiocytes stained with aceto- carmine were prepared using the ‘“‘squash”’ technique outlined by Beeks (1955) and modified by Kowal (1975) and viewed with a Zeiss standard WL microscope with a 100 oil immersion objective. Twelve cells with chromosomes clearly visible in a single plane were counted for each species. Images of meiotic figures from each species were captured with an Axiocam Hrm mounted on a Zeiss Axioplan 2 microscope using the 100 oil immersion objective. Information on floral and vegetative morphol- ogy of I. ayabacense and its putative parents was compiled from taxonomic descriptions (Bitter 1918; Shaw 1998: Leiva et al. 2003; Leiva and _Lezama 2005) and observations of all available specimens (see Appendix 1). Fourteen morpho- logical traits (see Appendix 2) were scored for the available specimens of the three taxa, and these _data were analyzed using principal components ) analysis in JMP 7.0 (SAS Institute Inc., Cary, NC). We also measured pollen grain size, which is known to be positively correlated with ploidy level (Gould 1957; Kowal 1975). Measurements of mature pollen grains were taken from individ- uals used for chromosome counts. Pollen was SMITH ET AL.: HYBRIDIZATION IN JOCHROMA 281 stained with gelatin fuchsin, and twenty pollen grains from each individual were measured to the nearest 0.1 um using a Zeiss AxioSkop 2 microscope with a 40 objective. RESULTS Anthers at roughly two-thirds their mature length (2.2 mm for LZ cyaneum, 2.1 mm for J. ayabacense and 3.5 mm for I. lehmannii) provided the best material for chromosome counts. Within a single flower containing anthers at this size, we were able to observe all the stages of meiosis I and II. Some cells could be counted at prophase I and IT, but the most countable stage for all species was telophase II. Chromosome counts for all species including /. ayabacense were n = 12 (Fig. 2), consistent with previous counts in Jochroma [1 cyaneum, n = 12 (Ratera 1961; Mahadavian 1967; Mehra and Bawa 1969) and I. fuchsioides, n = 12 (Ratera 1961)]. Pollen size for I ayabacense fell within the range of the parents (Table 1), consistent with the fact that all three share the same ploidy level (Gould 1957; Kowal 1975). The vegetative and floral features of J. ayabacense appeared to be largely intermediate between its putative parents (Table 1). The plant habit, amount of pubescence, numbers of flowers per inflorescence, and the form of the flowers (Fig. 2) fell exactly between J. lehmannii and I. cyaneum. Size measurements (e.g., plant height) for I. avyabacense were within the ranges observed in the putative parents although, for a few traits, such as leaf length, 7. ayabacense exhibited a wider range of variation than the parents. Also, in some traits, 1. avabacense favored one of the putative parents. For instance, it largely lacks the stem scales characteristic of J. lehmannii, but its fruits are nearly included in the accrescent fruiting calyx, resembling 1. /ehmannii rather than I. cyaneum. Principal components analysis (PCA) provided additional evidence for the morphological inter- mediacy of 1. ayabacense. Variation across thel4 floral and vegetative traits (Appendix 2) was largely captured by the first few components, with components | and 2 accounting for 73% (Fig. 3). The components clearly separate the tree taxa into discrete clusters in morphospace with I. ayaba- cense falling between its putative parents (Fig. 3). DISCUSSION This cytological and morphological study, together with existing phylogenetic and biogeo- graphic information, shows that 1. ayvabacense is a homoploid hybrid between [. cyaneum and I. lehmannii. While the phylogenetic data (Smith and Baum 2006) provided clear evidence that /. ayabacense carries alleles derived from both the /. cyaneum lineage and the J. /ehmannii lineage, they a. METRIC Wa ideas METRIC METRIC FIG. 2. MADRONO vd rs a 2 ae al eee [Vol. 55 Flowers, meiotic figures and pollen grains from /. cyaneum, I. ayabacense and I. lehmannii (top to bottom). The haploid number of 7 = 12 chromosomes is visible in all three taxa. did not indicate whether /. ayabacense was a homoploid or polyploid hybrid. The chromosome counts and the pollen measurements presented here indicate that 7. ayabacense has the same ploidy level as its parents (1 = 12). Also, the morphological data demonstrate that I ayaba- cense 18 intermediate between its two parents in a wide array of floral and vegetative traits. Although the morphological distinctiveness and cohesiveness of [ ayvabacense was deemed sufficient to merit specific recognition (Leiva and Lezama 2005), it remains to be seen whether the species can maintain reproductive isolation from its parents. At present, populations of ayaba- cense are largely separate from populations of I. cyaneum and I. lehmannii, perhaps due to differences in microhabitat preferences (Smith and Baum 2006). However, extensive searching during a recent collecting trip revealed one apparent backcross individual (J. ayabacense X TABLE 1. MORPHOLOGICAL DIFFERENCES AMONG I. CYANEUM, I. AYABACENSE AND I. LEHMANNII. ' This range is larger than that given in the species description, which appears to have been based on immature fruits. I. cyaneum Shrub, 1.5—3 m Young stems green, densely covered in branched hairs Ovate to elliptic, 10-23 cm by 5—10 cm 10—20 flowers per inflorescence Calyx 5—9 mm, slighty to markedly inflated in flower Corolla tubular, 2.5—-4 cm long, deep purple, largely glabrous but with hairs around the mouth of the tube Filaments 2.5—4 cm, hairy toward the base Anthers 3-4 mm Pollen 25-33 um (mean = 28) in diam. Style 2-4 cm Berry conical, 1.8—2.5 cm long, bottom third enveloped in fruiting calyx I. ayabacense Shrub to small tree, 3-4 m Young stems green or purplish toward the tips, covered in branched hairs, with occasional small triangular scales Elliptic to lanceolate, 8—22.5 cm by 5—10.5 cm 3-8 flowers per inflorescence Calyx 5—7 mm, slightly inflated in flower Corolla tubular to funnel—shaped, 2.6-4.0 cm, yellow with purple, interior glabrous, exterior with few to many hairs Filaments 1.6—2.2 cm, with sparse hairs toward the base Anthers 3.5—6 mm Pollen 25—33 um (mean = 28) in diam. Style 1.8—2.8 cm Berry globose to slightly conical, 1-1.5 cm long’, largely enveloped in fruiting calyx I. lehmannii Small tree, 4-10 m Young stems purplish and covered in | triangular scales Elliptic to lanceolate, 4-18.5 cm by 2.5—10 cm 2—6 flowers per inflorescence Calyx 5—7 mm, not inflated Corolla funnel-shaped to campanulate, 2—3.5 cm, yellow- green, interior glabrous, exterior pubescent Filaments 1.5—2 cm, glabrous or with sparse hairs at the base Anthers 4-7 mm Pollen 28-33 um (mean = 30) in diam. Style 1.2—1.6 cm Berry globose to slightly conical, 1—1.6 cm, largely enveloped in fruiting calyx 2008] ttn etre —— HH _Component 2 ~ Componentl =” Fic. 3. Principal components analysis of morpholog- ical measurements from [. cyaneum (solid diamonds), J. ayabacense (open diamonds) and I. lehmannii (solid squares). Component 1 accounts for 60.7% of the variation, and component 2 for 12.7%. I. lehmannii), suggesting that introgression re- mains possible. Also, 1. cyaneum, I. lehmannii and I. ayabacense in Ayabaca are all pollinated by the same two species of hummingbird, Adelomyia melanogenys and Coeligena iris (Smith et al. 2008), suggesting that pollen flow could occur between the species when they occur in close proximity. However, individual hummingbirds appear to exhibit preference for particular Iochroma species, thereby limiting interspecific pollen flow (Smith et al. 2008). The frequency and evolutionary significance of interspecific hybridization in Jochroma merits additional investigation. Within the last fifteen years, nine new species of Jochroma have been described from the area of greatest overlap in species distributions, the Amotape-Huanca- bamba zone of Ecuador and Peru, and several new species have yet to be described (S. Leiva, HAO, pers. comm.). While a hybrid origin has been suggested for some of these taxa (Smith and Baum 2006), at least six of these new (described and undescribed) taxa have yet to be included in phylogenetic, morphological and biogeographic studies. Acknowledging the possibility of hybrid- ization between Jochroma species, it is imperative to consider the geographic distribution, the morphological distinctiveness and the stability of any newly-found form before granting it recognition as a new species. ACKNOWLEDGMENTS The authors thank S. Leiva for assistance in the field, R. R. Kowal for advice in preparing squashes, and F. Lutzoni, K. Pryer and G. Wray for use of their microscopes. We also thank the curators of the following herbaria for providing material on loan: MO, NY and WIS. S. D. S. gratefully acknowledges financial support from the National Science Founda- tion grant DEB-0309310. SMITH ET AL.: HYBRIDIZATION IN JOCHROMA 28 ies) LITERATURE CITED ARNOLD, M. L. 1997. Natural hybridization and evolution. Oxford University Press, Oxford, UK. BEEKS, R. M. 1955. Improvements in the Squash Technique for Plant Chromosomes. Aliso 3:131— 133" BITTER, G. 1918. Solanaceae quattor austro-ameri- canae adhuc generibus falsis adscriptae. Repertor- ium Specierum Novarum Regni Vegetabilis 15: 149-155. BUERKLE, C. A., R. J. MORRIS, M. A. ASMUSSEN, AND L. H. RIESEBERG. 2000. The likelihood of homo- ploid hybrid speciation. Heredity 84:441-451. CHAPMAN, M. A. AND J. M. BURKE. 2007. Genetic divergence and hybrid speciation. Evolution 61: 1773-1780. FERGUSON, D. M. AND T. SANG. 2001. Speciation through homoploid hybridization between allote- traploids in peonies (Paeonia). Proceedings of the National Academy of Sciences, USA 98:3915—3919. HOWARTH, D. G. AND D. A. BAUM. 2005. Genealog- ical evidence of homoploid hybrid speciation in an adaptive radiation of Scaevola (Goodeniaceae) in the Hawaiian Islands. Evolution 59:948—961. GRANT, V. 1981. Plant speciation. Columbia University Press, New York, NY. GOULD, F. W. 1957. Pollen size as related to polyploidy and speciation in the Andropogon saccharoides-A. barbinodis complex. Brittonia 9:71—75. KOWAL, R. R. 1975. Systematics of Senecio aureus and allied species on the Gaspé Peninsula, Québec. Memoirs of the Torrey Botanical Club 23:1—113. LEIVA G., S., P. LEZAMA A., AND V. Q. SILVESTRE. 2003. lochroma salpoanum y I. squamosum (Sola- naceae: Solaneae) dos nuevas especies Andinas del Norte del Peru. Arnaldoa 10:95—104. AND P. LEZAMA A. 2005. Iochroma albianthum e Iochroma ayabacense (Solanaceae: Solaneae): dos nuevas especies del Departamento de Piura, Peru. Arnaldoa 12:72-80. MADHAVADIAN, P. 1967. The cytology of Jochroma tubulosa Benth. Caryologia 20:309-315. MEHRA, P. N. AND K. S. BAWA. 1969. Chromosomal evolution in tropical hardwoods. Evolution 23: 466-481. RAMSEY, J. AND D. W. SCHEMSKE. 1998. Pathways, mechanisms and rates of polyploid formation in flowering plants. Annual Review of Ecology and Systematics 29:467—S01. RATERA, E. L. 1961. Estudios cariologicos en Solana- ceas. Revista del Instituto Municipal de Botanica 1:61—65. RIESEBERG, L. H. 1997. Hybrid origins of plant species. Annual Review of Ecology and Systematics 28:359-389. , M. A. ARCHER, AND R. K. WAYNE. 1999. Transgressive segregation, adaptation and specia- tion. Heredity 83:363—372. SHAW, J. M. H. 1998. Jochroma — a review. The New Plantsman 5:154—-191. SMITH, S. D. AND D. A. BAUM. 2006. Phylogenetics of the florally diverse Andean clade Iochrominae (Solanaceae). American Journal of Botany 93: 1140-1153. , S. J. HALL, P. R. IZQUIERDO, AND D. A. BAUM. 2008. Comparative pollination biology of sympatric and allopatric Andean Jochroma (Sola- 284 naceae). Annals of the Missouri Botanical Garden 95:596-613. STEBBINS, G. L. 1959. The role of hybridization in evolution. Proceedings of the American Philosoph- ical Society 103:231—251. APPENDIX I HERBARIUM SPECIMENS USED FOR COMPARATIVE MORPHOLOGY Vouchers for chromosome counts and pollen mea- surements are indicated with asterisks. Iochroma ayabacense *PERU, Piura, Ayabaca, carretera Ayabaca-Yacu- pampa, 4.6146°S 79.7118°W, S. D. Smith, S. Leiva G. 337 (HAO, F, MO, NY, USM, WIS), 15/1/04. PERU, Piura, Ayabaca, 4.6456°S 79.7192°W, S. D. Smith, S. J. Hall 351 (HAO, F, MO, NY, USM, WIS), 16/1/04. PERU, Piura, Ayabaca, Barrio San Jose Obrero, 4.6455°S 79.7178°W, S. D. Smith 500 (QCNE, WIS), 30/1/05. Iochroma cyaneum *USA, Wisconsin, Madison, Grown from seed from a plant grown by W. G. D’Arcy at the Missouri Botanical Gardens, original collection likely to be Plowman 4594 from Azuay, Ecuador, S. D. Smith 265 (WIS), 20/V/2003. PERU, Piura, Ayabaca, carretera Ayabaca-Piura, 4.6514°S 79.7362°W, S. D. Smith, S. Leiva G. 329 (HAO, F, MO, NY, USM, WIS), 14/1/04. PERU, Piura, Ayabaca, Yacupampa, 4.6194°S 79.7106°W, S. D. Smith, S. Leiva G. 336 (HAO, F, MO, NY, USM, WIS), 15/1/04. PERU, Piura, Ayabaca, Barrio San Jose Obrero, 4.6441°S 79.7185°W, S. D. Smith 502 (QCNE, WIS), 30/1/05. ECUADOR, Loja, Cerro Sozoranga, 4.3553°S 79.7022°W, P. M. Jorgensen, C. Ulloa, H. Vargas G. Abendano 628 (MO), 29/ITV/1994. MADRONO [Vol. 55 Iochroma lehmannii *PERU, Piura, Ayabaca, carretera Ayabaca-Piura, 4.6596°S 79.7404°W, S. D. Smith, S. Leiva G. 330 (HAO, F, MO, NY, USM, WIS), 14/1/04. PERU, Piura, Ayabaca, Bosque Cuyas. 4.6035°S 79.7110°W, S. D. Smith, S. Leiva G. 339 (HAO, F, MO, NY, USM, WIS), 15/1/04. PERU, Piura, Ayabaca, Bosque Cuyas. S. Leiva G., N. W. Sawyer, V. Quipuscoa 2028 (F), 6/V1/97. ECUADOR, Canar, Km 15 Chunchi-Zhud, 2.3480°S 78.9377°W, S. D. Smith, L. Lopez 486 (F, QCNE, MO, WIS), 8/I/0S. ECUADOR, Canar, Km 20 Chunchi-Zhud, 2.3555°S 78.9590°W, S. D. Smith, L. Lopez 487 (QCNE, MO, NY WIS), 8/1/05. ECUADOR, Chimborazo, Carretera Alausi-Baguil- Guamote, V. Zak 2375 (MO), 11/VIII/1987. ECUADOR, Canar, Km 21 Chunchi-Zhud, 2.4333°S 79.0333°W, A. Alvarez and M. Tirado 1475 (MO), 9/VIT/1995. APPENDIX 2 CHARACTERS SCORED FOR PRINCIPAL COMPONENTS ANALYSIS Leaf length Leaf width Pedicel length Calyx width Calyx length Corolla width at base Corolla width at mouth Corolla length Corolla pigmentation (0 = no purple, | = light purple, 2 = deep purple) Corolla pubescence (0 = glabrous or few hairs, 1 = pubescent) Style length Filament length Anther length Flowers per inflorescence MADRONO, Vol. 55, No. 4, pp. 285-290, 2008 GLUCOSE-6-PHOSPHATE ISOMERASE VARIATION AND GENETIC STRUCTURE IN YUCCA BREVIFOLIA (AGAVACEAE) AMY T. TOULSON WIMMER! AND ROBERT MERRITT Department of Biological Sciences, Smith College, Clark Science Center, Northampton, MA 01063 ABSTRACT We used electrophoretic analysis of isozymes to investigate genetic structure of the Mojave Desert endemic, Yucca brevifolia. To test the hypothesis that, because of geographic isolation, most variation in allele frequencies would be between populations we sampled three subpopulations in each of five distinct populations: Arizona, Joshua Tree, Mojave, Western Mojave, and Utah. Two of nine proteins had isozymes that resolved adequately. One of these, Superoxide dismutase (SOD) was monomorphic. For the other, glucose-6-phosphate isomerase (GPI), five alleles were identified, and the three most common alleles were present in all populations sampled. For GPI, allele frequencies differed significantly between all five populations, as well as between subpopulations in both the Joshua Tree and Arizona populations. We also tested the hypothesis that there would be strong localized genetic structure due to the high proportion of recruitment attributed to vegetative reproduction. Based on GPI data from an intensively sampled 1 ha plot, genotypes were randomly distributed, and thus sexual reproduction through outcrossing may be the principal mode of recruitment in Y. brevifolia at this site. This study describes genetic structure within populations of Y. brevifolia based on protein variation at a single locus. Future research, using additional markers (nuclear, cytoplasmic, or both) is necessary to understand the dynamics of gene flow, genetic variation, and recruitment within the species Y. brevifolia Key Words: Genetic structure, glucose-6-phosphate isomerase, Mojave Desert, spatial autocorrela- tion, Yucca brevifolia. The distribution of Yucca brevifolia Engelm., an endemic of the Mojave Desert, includes portions of California, Nevada, Utah, and Arizona. Within this distribution, the Joshua tree is found in isolated populations at altitudes between 600 and 2200 m. The limits of the species distribution, both altitudinally and latitu- dinally, appear to be a result of sensitivity to mean temperature maxima and minima (Gates 1966). These temperature limitations produced altitu- dinal and latitudinal changes in the distribution of the species over the geologic time scale as a result of climatic variation. Evidence of altitudi- nal range contractions has been observed in both the Panamint mountain range and Death Valley, where Y. brevifolia tissue has been identified in packrat middens from the Pleistocene era at elevations more than 125m lower than the current distribution (Grayson 1993). Changes in latitudinal distribution are demonstrated by Yucca brevifolia tissue identified in middens near the Mexican border approximately 240 km South of the species current limit (Laudermilk and Munz 1934; Grayson 1993). The current distribution of Y. brevifolia, comprised of isolated populations limited to the ' Author for correspondence; email address: atoulson@ alumnae.smith.edu Mojave Desert, has likely had strong effects on genetic variation within and among populations. Genetic studies of both fragmented and endemic plant species typically reveal limited genetic variation within populations, while genetic dif- ferentiation among these populations tends to be high as a result of genetic drift and natural selection (Chung et al. 2004; Fracaro and Echeverrigaray 2006; Meister et al. 2006; Zhao et al. 2006). In addition to the geographic distribution of a species, genetic variation within populations is also influenced by factors includ- ing pollination, rates of inbreeding, outcrossing, and asexual reproduction (Loveless and Hamrick 1984). Methods of recruitment within Y. brevifolia populations are poorly understood. Yucca brevi- folia is pollinated by the moth species Tegiticula synthetica and T. antithetica, but it is unknown whether, in addition to moth pollination, the flowers undergo self-pollination (Pellmyr and Segraves 2003). Further, the relative importance of sexual and asexual reproduction in popula- tions is also unknown. Clonal populations have been reported in southern California (Boyd 1999) and populations in other areas may also exhibit vegetative reproduction. While a high proportion of sexual reproduction through random mating would result in a random distribution of geno- types in a population, a high proportion of vegetative reproduction and/or inbreeding would 286 MADRONO [Vol. 55 Population Subpopulation CALIFORNIA A B Cc | Arizona ©} w 6) / ARIZONA Joshua Tree q q) (& / wae OQ un 2292 Ve 433°N, 17°W Western Mojave (D> q) q) Fic. |. Map of populations sampled in this study with Mojave Desert boundaries indicated by light grey and the distribution of Y. brevifolia overlaid in dark grey (Little 1976; Griffiths et al. 2006). Pie charts depict the frequency of the most common allele, GPI 3, in each subpopulation. result in localized genetic structure. For a general discussion of the biology of Yucca brevifolia see Gucker (2006). The present study employs glucose-6-phos- phate (GPI) allozymes to investigate genetic structure in Y. brevifolia. This enzyme, coded for by a single locus in Y. brevifolia, is highly polymorphic, so it is well suited as a marker to study outcrossing, inbreeding and cloning. How- ever, geographic variation in GPI allele frequen- cies may result from selection (Gillespi 1991), which could confound the effects of geographic isolation on gene flow and genetic drift. There- fore, geographic variation in GPI allele frequen- cies may reflect some combination of isolation, selection, and drift. Markers that are both highly variable and neutral have yet to be identified in Y. brevifolia. Regarding the genetic structure of Y. brevifolia we test the following hypotheses: First, allele frequency would vary significantly among geo- graphically isolated populations. Second, allele frequencies would be homogenous among sub- populations since moth aided pollination and seed dispersal by animal and wind should lead to unrestricted gene flow within populations. Third, that outcrossing is the predominant form of sexual reproduction within Y. brevifolia. And finally, that recruitment through vegetative re- production occurs at levels sufficient to result in spatial structuring within some populations. METHODS Study Sites We selected five populations for sampling that spanned the distribution of Y. brevifolia including a large central population found in the Mojave Preserve (MP) and four peripheral populations; Joshua Tree National Park (JT), Arizona (AZ), Utah (UT) and Western Mojave (WM). Three of the populations were disjunct (AZ, MP, UT) while two populations were contiguous (JT, WM). All populations were separated from one another by a minimum of 200 km (Fig. 1, Table 1). Both disjunct and nondisjunct popula- tions were included in this study to permit assessment of long distance gene flow between contiguous populations. The populations includ- ed both var. brevifolia (AZ, JT, WM) and var. jaegeriana (MP, UT). Within each of these five populations, three subpopulations were sampled that were a mini- mum of 1.6 km apart. Subpopulation samples consisted of 48 trees, except for the Western Mojave subpopulations where low density and restricted access limited sample sizes to 24 trees per subpopulation, selected haphazardly. Trees were selected from all size classes and were at least 10 m apart. From each tree, three recently developed leaves were collected and stored on dry ice. | In addition to the five populations, we also | intensively sampled a l-ha plot in Joshua Tree | National Park to assess localized genetic struc- | ture. In the plot, sampled stems were separated by at least 1 m. Where two or more stems were 0.05). The results of pair-wise tests for homogeneity indicated that within population allele frequen- cies in MP, UT and WM were homogenous across subpopulations. However, significant het- erogeneity was found among subpopulations at both JT and AZ (Table 3; P < 0.05). With the exception of three comparisons (AZC with JTB, P = 0.06; AZC with MPABC, P = 0.65: and JTB with MPABC, P = 0.14), all pairwise compari- sons of allele frequencies across populations revealed significant heterogeneity (P < 0.01) The expected heterozygosity of the intensively sampled one hectare plot was 0.65. The popula- 288 TABLE 2. MADRONO [Vol. 55 ALLELE FREQUENCIES OBSERVED AT EACH SUBPOPULATION WITHIN THE FIVE SAMPLED POPULATIONS (MOST COMMON ALLELE IN BOLD) ALONG WITH THE EXPECTED HETEROZYGOSITY (Hg) FOR EACH SUBPOPULATION. Pop. Subpop. GPI 1 GPI 2 AZ A B Cc JT A 0.08 B 0.01 ce 0.05 MP A 0.01 B Cc 0.01 wah A 0.06 B 0.07 Cc 0.05 WM A B Cc tion was in Hardy-Weinberg equilibrium. Mor- an’s I indicated a random distribution of genotypes across plants in the plot (Z = —1.12). DISCUSSION Genetic Structure Our data are consistent with the hypothesis that geographic isolation among populations of Y. brevifolia restricts gene flow below the level required for homogenization of allele frequencies at the GPI locus. However, given that some studies have suggested an adaptive role for GPI variation (Gillespi 1991; Katz and Harrison 1997; Wheat et al. 2006), the variation in allele TABLE 3. PAIR-WISE COMPARISON OF HOMOGENEITY OF ALLELIC DISTRIBUTION BETWEEN SUBPOPULATIONS WITHIN EACH POPULATION. P-values that are bold indicate subpopulations which differ significantly in allele frequencies. Subpopulation Pop. 2 P-value AZ <0.01 <0.01 0.01 0.03 0.01 <0.01 0.46 0.24 0.55 0.63 0.43 0.80 0.12 0.91 0.37 JT MP UT Brrr rrmrrrrrurPry|- ANRPANBWANARPANBNANY Allele GPI 3 GPI 4 GPI 5 He 1.00. 0.00 0.84 0.04 O12 0.28 0.69 0.18 0.13 0.47 0.48 0.38 0.06 0.62 0.53 0.32 0.14 0.60 0.27 0.59 0.08 0.57 0.68 0.25 0.06 0.47 0.63 0.26 0.12 0.52 0.67 0.19 0.13 0.50 0.75 0.17 0.02 0.40 0.68 0.20 0.05 0.49 0.71 0.17 0.07 0.46 0.58 0.40 0.02 0.50 0.40 0.58 0.02 0.50 0.54 0.44 0.02 0.51 frequencies may be a reflection of selection in this long-lived plant rather than the effects of geo- graphic isolation between the five populations. Our data also show heterogeneity in allele frequencies among subpopulations within both the AZ and JT populations. Clearly, our hypoth- esis that moth pollination coupled with wind and animal seed dispersal would produce sufficient gene flow to sustain homogeneity of allele frequencies across subpopulations is not support- ed. While subpopulation AZC is separated from both AZA and AZB by relatively large distances (29 km and 37 km respectively), such is not the case for AZA and AZB (7.4 km) or for the JT subpopulations (2—7.4 km). The AZ population is in the transition zone between the Mojave and Sonoran deserts in an area with a mix of communities associated with each desert. In addition, this population spans the largest altitu- dinal range (>400 m) and includes a subpopula- tion (AZC) 50m below the reported elevation range of Y. brevifolia. The JT population, however, is in a relatively homogeneous environ- ment. Whether the observed variation in allele frequencies is a result of selection (as suggested by | the environmental heterogeneity in AZ) or genetic: drift (as suggested by the apparent environmental homogeneity in JT), pollen and seed dispersal in’ Y. brevifolia appears restricted, in some locations, | even over small geographic distances. Inbreeding In populations where inbreeding is common, a heterozygote deficiency is observed compared to. Hardy-Weinberg expected zygotic frequencies. All subpopulations in this study were found to be in Hardy-Weinberg equilibrium, thus indicat- ing that inbreeding does not occur at any significant level in most populations of Y. | | | | 2008] brevifolia. It is still possible that self pollination events occur, but, perhaps the seeds are not viable as is the case in Yucca filamentosa (Marr et al. 2000). Recruitment A review of studies using Moran’s I by Heywood (1991) indicated that selfing species show pronounced local genetic structure, while outcrossing species with random sexual repro- duction and seed dispersal do not show any structured pattern. Vegetative reproduction would have an even stronger effect on population structure. Results of the spatial autocorrelation analysis indicate that genotypes in the I-ha intensively sampled plot were randomly distrib- uted, and thus that sexual reproduction through random mating, rather than vegetative reproduc- tion or self-pollination, has been the predominant form of recruitment at this site. Recruitment in Yucca brevifolia may be pri- marily episodic (Wallace and Romney 1972). A twenty-year study by Comanor and Clark (2000) of Y. brevifolia recorded no recruitment in their study areas. The combination of predominant outcrossing with long stretches of no measurable recruitment suggests that recruitment in Y. brevifolia 1s episodic, and likely reliant on rare environmental conditions. While this model of recruitment through outcrossing may be charac- teristic of this study plot, it may not be common for the species as a whole. Further, in this study an individual was defined as a stem or group of stems less than 1m from one another. In many cases, the individuals sampled were com- posed of more than one stem. While vegetative reproduction appears to have a small role in recruitment of new individuals at this site, it may play a significant role in the longevity of individuals. Further Research This project has developed the first model of genetic structure found within populations of Y. brevifolia based on protein variation. Future studies investigating variation at the molecular level are needed to increase understanding of genetic variation both within and across popula- tions. Through continued investigation into genetic structure of Y. brevifolia, biologists will develop a better understanding of the relation- ships of gene flow, genetic variation, and recruitment levels across the landscape of the Mojave Desert. ACKNOWLEDGMENTS Funding for this project came from the Blakeslee Grant at Smith College, the Howard Hughes Grant and the Elizabeth B. Horner Fund. Permits for specimen TOULSON AND MERRITT: GENETIC STRUCTURE OF GPIIN YUCCA BREVIFOLIA 209 collection were granted from Joshua Tree National Park and the Mojave National Preserve. Special thanks to the Sweeney Granite Mountain Desert Research Center for their support during the field portion of this project and to Laura Katz for her guidance in preparing this manuscript. LITERATURE CITED Boyp, S. 1999. Vascular flora of the Liebre Mountains, western Transverse Ranges, California. Aliso 18:93-139. CHUNG, J. D., T. P. LIN, Y. C. TAN, M. Y. LIN, AND S. Y. HWANG. 2004. Genetic diversity and bioge- ography of Cunninghamia konishii (Cupressaceae), an island species in Taiwan: a comparison with Cunninghamia lanceolata, a mainland species in China. Molecular Phylogenetics and Evolution 33:791-801. COMANOR, P. L. AND W. H. CLARK. 2000. Preliminary growth rates and the proposed age-form classifica- tion for the Joshua Tree, Yucca brevifolia (Agava- ceae). Haseltonia 7:37—46. FRACARO, F. AND S. ECHEVERRIGARAY. 2006. Genetic Variability in Hesperozygis ringens Benth. (Lamia- ceae), an Endangered Aromatic and Medicinal Plant of Southern Brazil. Biochemical Genetics 44:479—490. GATES, R. 1966. Climatic elements and the disjunct distribution of the Joshua Tree (Yucca brevifolia). Joshua Tree National Monument Library, Twenty- nine Palms, CA, unpublished. GILLESPIE, J. H. 1991. The causes of molecular evolution. Oxford University Press, New York, NY. GRAYSON, D. K. 1993. The desert’s past: a natural prehistory of the Great Basin. Smithsonian Insti- tution Press, Washington, D.C. GRIFFITHS, P. G., R. HEREFORD, AND R. H. WEBB. 2006. Sediment yield and runoff frequency of small drainage basins in the Mojave Desert, U.S.A. Geomorphology 74:232—244. GUCKER, C. L. 2006. Yucca brevifolia. Jn Fire Effects Information System (database). U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station, Fire Sciences Laboratory, http// www.fs.fed.us/database/feis/ (accessed April 5, 2009). HALDANE, J. B. S. 1954. An exact test for randomness of mating. Journal of Genetics 52:631—635. HEYwoop, J. 1991. Spatial analysis of genetic variation in plant populations. Annual Review of Ecology and Systematics 22:335—338. KATz, L. A. AND R. G. HARRISON. 1997. Balancing selection on electrophoretic variation of phospho- glucose isomerase in two species of field cricket: Gryllus veletis and Gryllus pennsylvanicus. Genetics 147:609-621. LAUDERMILK, J. D. AND P. A. MUNZ. 1934. Plants in the dung of Nothrotherium from Gypsum Cave, Nevada. Contributions to Paleontology from the Carnegie Institution of Washington 453:29-37. LEE, J. AND D. W. S. WONG. 2000. Statistical analysis with Arcview GIS, Vol. I. Wiley, John and Sons, Inc., New York, NY. LITTLE, E. L., JR. 1976. Atlas of United States trees, volume 3, minor Western hardwoods. Miscella- 290 neous Publication 1314, U.S. Department of Agriculture, Washington, D.C. LOVELESS, M. D. AND J. D. HAMRICK. 1984. Ecolog- ical determinants of genetic structure in plant populations. Annual Review of Ecology and Systematics 15:65—95. MARR, D. L., J. LEEBENS-MACK, L. ELMS, AND O. PELLMYR. 2000. Pollen dispersal in Yucca filamen- tosa (Agavaceae): the paradox of self-pollination behavior by Tegeticula yuccasella (Prodoxidae). American Journal of Botany 87:670—677. MEISTER, J.. M. A. HUBAISHAN, N. KILIAN, AND C. OBERPRIELER. 2006. Temporal and spatial diversi- fication of the shrub Justicia areysiana Deflers (Acanthaceae) endemic to the monsoon affected coastal mountains of the southern Arabian Penin- sula. Plant Systematics and Evolution 262:153—171. PELLMYR, O. AND K. SEGRAVES. 2003. Pollinator divergence within an obligate mutualism: two yucca moth species (Lepidoptera; Prodoxidae: Tegeticula) on the Joshua tree (Yucca brevifolia; Agavaceae). Annals of the Entomological Society of America 96:716—722. POULIk, M. D. 1957. Starch gel electrophoresis in a dis- continuous system of buffers. Nature 180:1477-9. ROWLANDS, P. G. 1978. The vegetation dynamics of the Joshua tree (Yucca brevifolia Engelm.) in the southwestern United States of America. Doctoral dissertation. University of California Riverside, Riverside, CA. MADRONO [Vol. 55 SELANDER, R. K., M. H. SMITH, S. Y. YANG, W. E. JOHNSON, AND J. B. GENTRY. 1971. Biochemical polymorphism and systematics in the genus Pero- myscus. I Variation in the old field mouse (Peromyscus polionotus). Studies in Genetics VI: 59-90. SOLTIS, D. E., C. H. HAUFLER, D. C. DARROW, AND G. J. GASTONY. 1983. Starch gel electrophoresis of ferns: a compilation of grinding buffers, gel and electrode buffers and staining schedules. American Fern Journal 73:9-27. WALLACE, A. AND E. M. ROMNEY. 1972. Radioecol- ogy and ecophysiology of desert plants in the Nevada Test Site. Rep. TID-25954. Office of Information Services, U.S. Atomic Energy Com- mission, Washington, D.C. WENDEL, J. F. AND C. R. PARKS. 1982. Genetic control of isozymes variation in Camellia japonica L. Journal of Heredity 73:197—204. WHEAT, C. W., W. B. WATT, D. D. POLLOCK, AND P. M. SCHULTE. 2006. From DNA to fitness differences: sequences and structures of adaptive variants of Colias phosphoglucose isomerase (PGI). Molecular Biology and Evolution 23:499— 512. ZHAO, A., X. CHEN, X. ZHANG, AND D. ZHANG. 2006. Effects of fragmentation of evergreen broad-leaved forests on genetic diversity of Ardisia crenata var. bicolor (Myrsinaceae). Biodiversity and Conserva- tion 15:1339-1351. MADRONO, Vol. 55, No. 4, pp. 291-296, 2008 IDENTIFICATION, DISTRIBUTION, AND FAMILY PLACEMENT OF THE PLEUROCARPOUS MOSS BESTIA LONGIPES JAMES R. SHEVOCK Department of Botany, California Academy of Sciences, 55 Music Concourse Drive, Golden Gate Park, San Francisco, CA 94118-4599 jshevock@calacademy.org DANIEL H. NORRIS University Herbarium, 1001 Valley Life Sciences, University of California, Berkeley, CA 94720-2465 dhnorris@berkeley.edu A. JONATHAN SHAW Department of Biology, Duke University, Durham, NC 27708-0338 shaw@duke.edu ABSTRACT Recent field work and examination of specimens at CAS, MICH, MO, NY, SFSU, UBC and UC lead us to conclude that Bestia longipes, contrary to some earlier literature reports, is endemic to coastal regions of north-central to southern California. Based on analyses of DNA nucleotide sequence data we recognize Bestia in the Lembophyllaceae. Key Words: Bestia, California, [sothecium, Lembophyllaceae, mosses. INTRODUCTION In examining the literature for Bestia longipes (Sullivant & Lesquereux) Brotherus, it becomes clear that this species endemic to the Pacific Coast of North America has been poorly understood. Part of this confusion can be attributed to various interpretations regarding the generic circumscription of Bestia and its taxonomic affinities at the family level (Lesquer- eux 1868; Lesquereux and James 1884; Grout 1928; Crum 1991). When first described in 1865 as Alsia longipes by Sullivant & Lesquereux, it became the third species in the genus Alsia (established by Sullivant in 1855) joining Al/sia californica (W. J. Hooker & Arnott) Sullivant (transferred from Neckera) and Alsia abietina (W. J. Hooker) Sullivant (transferred from Neckera). The genus Al/sia, with three species, was a morphologically heterogeneous assemblage of Pacific Coast endemics. Alsia abietina was recombined as Dendroalsia abietina in 1905 by Britton and Alsia longipes was transferred to Bestia in 1906 by Brotherus (Crosby et al. 2000). These three genera, each now monospecific, are today all placed in different families. _ Had Brotherus (1906) retained Bestia as a ‘monospecific genus, subsequent taxonomic and ‘nomenclatural confusion would have been nil. However, this was not to be. Two other taxa Usothecium obtusatulum Kindberg and Thamnium vancouveriense Kindberg) were placed in Bestia oy Brotherus (1925) further altering the generic circumscription. Over the years Bestia has been attributed to such families as the Neckeraceae, Thamnobryaceae, Cryphaeaceae, Hypnaceae, Brachytheciaceae (Brotherus 1925; Crum 1987; Buck and Goffinet 2000; Crosby et al. 2000), and most recently, the Lembophyllaceae (Goffinet and Buck 2004; Huttunen et al. 2004; Quandt et al. 2008). This diversity of family placements was in part based on which species of Bestia had been critically examined and what combination of gametophytic and sporophytic characters were considered most important for inferring relation- ships. Grout (1928) also listed three species in Bestia. His interpretation of the genus included the incorporation of Hypnum brewerianum Les- quereux as a new combination within Bestia, expanding yet again the generic circumscription. Bestia breweriana (Lesquereux) Grout is now regarded as a synonym of /sothecium cristatum (Hampe) H. Robinson (Andrews 1952). The other species that Grout retained in Bestia he called B. holzingeri (Renauld & Cardot) Broth- erus. This species also has a long nomenclatural history of being placed in 10 different genera containing 14 species synonyms (Norris and Enroth 1990). By the time Lawton (1971) published her moss flora of the Pacific Northwest it was generally known as Bestia vancouveriensis (Kindberg) Wijk & Margadant. Norris and Enroth (1990) resolved this part of the problem when they elevated Bestia vancouveriensis to generic rank as Bryolawtonia, and transferred it back to the Thamnobryaceae (Neckeraceae s.].). 292 Crum (1987) and Norris and Enroth (1990) provided a detailed overview of this history and we will not repeat it further except to say that Crum (1987) proposed Bestia longipes be trans- ferred to the Brachytheciaceae and viewed it to be closely related to [sothecium. Norris and Enroth (1990) also accepted this interpretation, as did Crosby et al. (2000). Crum (1991) further speculated that Bestia may not even need to be recognized as a monospecific genus at all, suggesting that B. /ongipes might be a mere form of Isothecium myosuroides Bridel. Recent DNA evidence supports the hypothesis that Bestia and Isothecium are indeed closely related (Goffinet and Buck 2004) yet distinct genera. Based on our herbarium studies and detailed examination of both gametophytic and sporophytic characters we conclude that Bestia is so distinctive as to require separate generic status. This conclusion is also supported by additional molecular data (Quandt et al. 2008). DISCUSSION Since the publication of a bryoflora of California (Norris and Shevock 2004a, b), a re- examination of material in California herbaria showed that several of the collections labeled and reported in the literature as Bestia longipes are in fact one or more species of /sothecium. Therefore, our earlier interpretation (Norris and Shevock 2004a) of this taxon was also clouded. This issue, however, became clear once we examined plants of Bestia longipes in the field from the general type locality in the Oakland Hills above Berkeley. Additional field work from coastal mountains in the San Francisco Bay region showed Bestia to be a rather common component of the flora on boulders along streams where it is often associ- ated with Prerogonium gracile (Hedwig) Smith and species of /sothecium, primarily I. cristatum. This freshly collected material became the base- line for comparative purposes during our review of the herbarium record. While bryologists in California can attest to the difficulties in naming species of Jsothecium many yet recognize the various growth form expressions that appear fairly constant in nature. The taxonomic and nomenclatural history of this genus has made working with Jsothecium here in California, as well as more broadly along the Pacific Coast of North America, difficult and complex (Allen 1983). Despite the recent publication of two analyses based on molecular data (Ryall et al. 2005; Draper et al. 2007), consensus among west coast bryologists on the delimitation of Jsothe- cium taxa and what to call them remains elusive. Notwithstanding, knowledge about the appear- ance of Bestia in the field along with a sense of its very specific microhabitat renders it easily separated from /sothecium, even without sporo- MADRONO [Vol. 55 phytes. One can recognize this very long-pendent, pinnately-branched moss by its branches being much shorter than the main axis. Smaller plants or those occurring in marginal habitat can have a main axis so short as to look more like an Isothecium. Dried Bestia upon close examination also looks different from Jsothecium species, especially [. cristatum with its more julaceous branches, but it can superficially resemble J. myosuroides s.1. In the northern portion of its range, Bestia is considerably harder to distinguish in the field from /. myosuroides because Isothe- cium dominates the suitable habitat and plants from the two genera can be intertwined. We speculate that /sothecium, especially I. myosur- oides, out competes and replaces Bestia for that particular ecological niche at the northern portion of its range. The sporophytes of Bestia are not like the shorter generally asymmetrical ‘hypnaceous’ cap- sules of Jsothecium. However, the perichaetial leaves in both Bestia and Jsothecium are greatly enlarged and sheathe the seta base. Developing sporophytes in Bestia with light colored setae are inserted ventrally on the main stem axis adjacent to the substratum. Because they are oriented upward along and parallel to the main stem axis, they can be easily overlooked during a casual inspection of plants in the field. We found the illustration of Bestia longipes (as Alsia L., plate 63) in Sullivant (1874) and reprinted in part (Broth- erus 1925) to be remarkably detailed and accurate. Jsothecium cristatum on the other hand is commonly associated with Bestia especially in the central portion of its range and generally produces abundant sporophytes with reddish setae that are erect and easily visible above the branches. When hydrated, Bestia has a rather plumose appearance and cascades downward) from the substrate. Colonies can cover square. meters of rock surface. | The following key should readily separate. Bestia from Isothecium species. | la. Leaves with serrulations restricted to the leaf apex (1/5 of the leaf), serrulations consisting uniformly of one cell; cells across leaf surface uniform in color; costa prominent, extending to near leaf apex with several spines on the distal 1/6 or more of the costa; median cells, even those near the costa, uniformly short (generally <4:1) and thick walled resembling cells near apex; foliose pseudoparaphyllia with truncate and crenulate apices of short cells (<2:1); capsules straight, long-cylindric, sym- metrical, lacking an abrupt constriction below | PATON UE Pacis “a 2 ce cc ate eer ee em Bestia lb. Leaves with serrulations extending from leaf apex to middle or even lower, at least some teeth composed of more than one cell; cells across leaf with an abrupt transition from the green alar region to a paler green throughout rest of leaf; costa prominent, extending rarely beyond 2/3 of leaf, ending in 1 to few spines; 2008] SHEVOCK ET AL.: juxtacostal region with elongate cells (up to 8:1) and much longer than cells at leaf apex; foliose pseudoparaphyllia irregularly lobed or toothed with elongate cells making up those teeth; capsules oblong-ovoid, suberect to cernuous, generally slightly asymmetrical with an abrupt constriction below mouth. . . Jsothecium DISTRIBUTION After examining collections labeled Bestia from several herbaria we conclude that this monospe- cific genus is a Californian endemic and con- firmed for the following counties: Alameda, Contra Costa, Lake, Los Angeles, Marin, Men- docino, Monterey, Napa, San Benito, San Luis Obispo, San Mateo, Santa Barbara, Santa Clara, Santa Cruz, Sonoma, and Ventura. The only coastal county within this range lacking Bestia 1s San Francisco, due to a lack of suitable riparian habitat (Shevock and Toren 2001). Although Bestia longipes was not encountered during our herbarium review as occurring in Orange and San Diego counties just south of Los Angeles County, we believe there is a high probability that Bestia will be discovered in this area because seemingly appropriate habitats occur in the coastal portions of the Peninsular Ranges. Moreover, the range of Bestia could perhaps extend yet farther south to Guadalupe Island, Baja California, Mexico where both Alsia and Dendroalsia reach their southernmost outposts (Schofield 2004). Reports of Bestia (Norris and Shevock 2004a) from the northwest corner of California in Del Norte, Humboldt, Shasta, Siskiyou counties and the northern Sierra Nevada in Placer County are erroneous and represent Jsothecium species. Two Oregon specimens identified as Bestia (NY) and cited in Chapman and Sanborn (1941) also proved to be Jsothecium. Bestia longipes was not reported from Oregon by Christy et al. (1982) or Lawton (1971). Koch (1950) reported Bestia for Oregon although no specimen was cited. Two historical specimens, correctly identified as Bestia _longipes, are attributed to Alaska and seem to be distributional anomalies. The specimen housed at NY was purchased sometime in the late 1800s. It lacks data except for being attributed to Alaska. The other collection (originally at CAS but now at UBC) states “collected by Kellogg in 1867 from Redout Bay, Alaska.” However, in both cases these historical collections seem geograph- ically unlikely to have been collected in Alaska. Also, both of these packets have no soil, litter or other mosses mixed with the specimen, but | rather, just a few ‘clean’ individual branches. _We view these two collections as likely some type | of herbarium processing error. Also, Bestia (as Alsia longipes) was not reported from Alaska by _Cardot and Thériot (1902, 1906). Considering the level of bryophyte collecting in the coastal Portions of the Pacific Northwest from Alaska, BESTIA LONGIPES 293 British Columbia, Washington and Oregon over the past century without encountering a modern- day collection of Bestia longipes leads us to conclude that it is in fact a Californian endemic. Bestia 1s most commonly encountered at sites with cool air drainage patterns and is therefore more or less restricted to narrow stream channels and canyons. Although Bestia prefers rock walls and massive boulders in partial shade, it can also infrequently occur on adjacent bases of tree trunks, primarily Umbellularia californica (Hook- er & Arnott) Nuttall. Many of the habitats of Bestia are influenced by summer fog and cooler temperatures compared to hotter inland valleys. The elevation range is from 50—2500 ft. SPECIMENS EXAMINED CALIFORNIA. Alameda Co.: Alameda, Gib- bons 14 (NY); San Antonio Creek, Kellogg s.n. (NY); Oakland, collector unknown (MO, NY): Oakland, Eaton s.n. (NY); Berkeley, Howe 27 (MO, NY, UBC) & 232 (CAS, MICH, NY, UC); Strawberry Creek east of Hilgard Hall, U.C. Berkeley, Norris 82503 (CAS, UC); Garin Re- gional Park southeast of Hayward, Whittemore 5295 (CAS, MO, UC); Grass Valley Creek, Anthony Chabot Regional Park, Shevock 29365 (CAS, COLO, H, MO, NY, S, UC, US) & 29485 (CAS, KRAM, MO, NY, UC). Contra Costa Co.: Gorge Trail toward Lake Anza, Tilden Regional Park, Norris 109512 (CAS, COLO, DUKE, H, MO, NY, OSU, UBC, UC, UNAM, UNLV, US) and Norris 109518 (UC); Havey’s Canyon Trail, Wildcat Canyon Regional Park, Norris & Hillyard 109589 (CAS, MO, UC); Castle Rock Park, Schofield 87551 (MO, UBC): East Fork Sycamore Creek, Mt. Diablo State Park, Shevock & Ertter 24528 (CAS, US) and Shevock & Ertter 24531 (CAS, DUKE, E, H, UC). Lake Co.: Anderson Springs, Toren 2529 (SFSU); southern shore of Clear Lake at Kelsey Creek Slough, Clear Lake State Park, Toren 6844 (CAS); Troutdale Creek, Mt. St. Helena, Robert Louis Stevenson State Park, Toren 8427 (CAS). Los Angeles Co.: near Pasadena, McClatchie 456 (NY); Monrovia Canyon, collector unknown (NY); San Gabriel Mts., McClatchie s.n. (UBC); White’s Landing, 5 mi north of Avalon, Santa Catalina Island, Steere s.n. (UBC). Marin Co.: Mill Valley, Blasdale s.n. (CAS, UBC, UC); Devils Gulch, Samuel P. Taylor State Park, Shevock 29824 (CAS, H, S, UC); Panoramic Drive, Mt. Tamalpais State Park, Shevock 29833 (CAS, DUKE, E, MO, NY, UC, US). Mendocino Co.: no locality given beyond county, Toren & Showers s.n. (CAS); Hendy Woods State Park, Toren 2772 (SFSU); Parsons Creek below Hunt- ley Peak, Heise 2230 (UC). Monterey Co.: divide 0.5 mi south of Santa Lucia Memorial Park, Los Padres National Forest, Shevock & Kellman 294 24811 (CAS, MO, UC); Alder Creek Campground, Los Padres National Forest, Shevock & Kellman 27729 (CAS, UC); Indians Road 4.52 mi north of Fort Hunter Liggett Military Reservation bound- ary, Los Padres National Forest, Kellman & Shevock 3602 (CAS); South Fork Devils Canyon Creek below Canogas Falls, Los Padres National Forest, Shevock & Kellman 27830 (CAS, MO, NY); Palisades off of Gabilan Road, Fort Hunter Liggett Military Reservation, Kellman, Shevock, & Robertson 3743 (CAS); Big Sur River, Kellman 3996 & 4000 (CAS); Little Sur River below Jackson Camp, Los Padres National Forest, Shevock & Kellman 29054 (CAS, MO, NY) and Kellman & Shevock 4873 (CAS); Big Creek, UC Landels-Hill Big Creek Reserve, Shevock & Kellman 28898 (CAS, UC) and Devil’s Creek at Redwood Camp, Shevock & Kellman 27759 & 27842 (CAS, MO, NY, UC) and Shevock & Kellman 27765 (CAS, MO, NY), Kellman & Shevock 4909, 5016 (CAS); South Fork McWay Canyon, Julia Pfeiffer Burns State Park, Kellman 4946 (CAS); Rocky Ridge Trail to Soberanes Creek, Garrapata State Park, Kellman 3358 (CAS); Grimes Canyon below highway 1, She- vock & Kellman 27739 (CAS, UC); Redwood Gulch along highway 1, Becking 670719 (UBC); 1 mi south of Castro Canyon, Schofield 29181 (UBC). Napa Co.: St. Helena Creek, highway 29 at milepost marker 47.62 about | mi south of Lake County line, Shevock 29852 (CAS, COLO, DUKE, H, KRAM, MO, NICH, NY, OSU, UC, UNAM, US, WTU). San Benito Co.: Juniper Canyon along stream, Pinnacles National Mon- ument, Kellman, Shevock, & Villaserior 4244 & 4251 (CAS). San Luis Obispo Co.: Cypress Mountain Road southern end of Santa Lucia Range, Carter 1325 (UC); Cerro Alto Trail, Cerro Alto Campground, Los Padres National Forest, Carter 1843 (UC). San Mateo Co.: Pescadero Creek about 3 mi west of San Mateo Park entrance, Pescadero Road, Schofield 11206 (MICH, MO, NY, UBC, UC) & 96269 (MO, NY); Purissima Creek near Searsville Lake, Schofield 12899 (MO, NY, UBC, UC) and Steere s.n. (UBC); Lake Pilarcitos, Koch 3366 (MICH, NY, UC); Highway 1, about 2 mi north of Santa Cruz County line, Koch 2083 (MICH, NY, UC); between Serpentine Trail and Sylvan Trail, Edgewood County Park, Whittemore & Sommers 5238 (CAS) and 5240 (MO); Sylvan Way, Redwood City, Whittemore 3114 (CAS) and 1407 (MO); San Mateo Creek, Howe s.n. (NY); Tafoni Sandstone Formation, El Corte de Madera Creek Open Space Preserve, Shevock 29927 (CAS, COLO, DUKE, E, GOET, H, KRAM, MA, MHA, MICH, MO, NICH, NY, OSU, PE, S, UBC, UC, UNAM, US). Santa Barbara Co.: above Santa Barbara, Smith King s.n. (UC); Cold Springs, Los Padres National Forest, Shevock 27875 (CAS, MO, NY, UC) and MADRONO [Vol. 55 Laeger 526 (CAS); Pelican Bay, Santa Cruz Island, Fosberg 208 (UC); Lady’s Harbor, Santa Cruz Island, Fosberg 361 (NY); Cherry Canyon, Santa Rosa Island, Channel Islands National Park, Shevock 20911 (CAS, UC) and Water Canyon east of Black Mountain, Santa Rosa Island, Channel Islands National Park, Shevock & Rodriguez 20817 (CAS, UC). Santa Clara Co.: Alum Rock Canyon, San Jose, Bradshaw 496 (CAS, MICH, NY, UBC, UC); Alum Rock Park, Schofield & Mueller 6923 (MO) & 36446 (UBC); Lower Braen Canyon and Hunting Hollow east of Gilroy Hot Springs Road, Henry W. Coe State Park, Whittemore & Briggs 6745 (CAS, MO, NY) and canyon at Woodchopper Creek, Whittemore & Briggs 6766 (CAS, MO, NY); Woods Trail from Mt. Umunhum Road near Guadalupe Creek, Sierra Azul Open Space Preserve, Whitte- more et al. 6105 (CAS, MO); Mine Hill Trail adjacent to new Almaden Trail, Almaden Quick- silver County Park, Whittemore 5323 (CAS, MICH, MO); Mt. Hamilton Road at milepost 8.5, entrance to Joseph D. Grant County Park, Whittemore 6058 (CAS, NY) & 6056 (MO); Jasper Ridge Nature Reserve, Stanford Univer- sity, Schofield & Thomas 71538 (MO, UBC) and 96794 (MO, UBC); New Grade Road, Geis 657 (MICH); Stevens Canyon Creek above reservoir, Steere s.n. (UBC); Los Altos Hills, Schofield 23051, 63731, & 112008 (UBC). Santa Cruz Co.: Granite Creek, Kellman 1105 (CAS, UC); Laguna Creek, Kellman 1854 (CAS); Baldwin Creek, Wilder Ranch State Park, Kellman 832 & 2579 (CAS); Majors Creek Canyon near Highway 1 between Santa Cruz and Davenport, Kel/man 1065 (CAS); Scott Creek, Swanton Road and milepost 4.15, Kellman 410 (CAS); confluence of Boyer and Big Creek, Kellman 1897 (CAS); Zayante Canyon Road north of Felton, Norris 56869 (UC). Sonoma Co.: Adobe Canyon near Kenwood, Koch 463 (MICH, NY, UC); Sugar- loaf Ridge State Park, Norris & Hillyard 109804, 109812, & 109814 (CAS, UC); Fife Creek, Armstrong Redwoods State Reserve, Shevock 29859 (CAS, DUKE, MO, NY, UC); Audubon Canyon Ranch Bouverie Preserve north of Sonoma, Carter 1874 (UC). Ventura Co.: Willard Canyon, McClatchie s.n. (NY); Carlisle Creek, Santa Monica Mountains., Wishner & Sagar 2005-09-23-1 (CAS, UC); Big Sycamore Canyon, Santa Monica Mountains., Point Mugu State - Park, Sagar 711 (UC). FAMILY PLACEMENT As mentioned previously, components within — Bestia have been assigned to a wide assortment of moss families since the late 1800s. Crum (1987) | aligned Bestia longipes in the Brachytheciaceae and this placement was generally accepted by | other bryologists (Crosby et al. 2000). DNA- 2008] SHEVOCK ET AL. based analyses, however, suggest that Bestia, along with /sothecium, belong in the Lembophyl- laceae (Goffinet and Buck 2004; Huttunen et al. 2004; Tangney 2007; Quandt et al. 2008). As part of our present review of the problem, we obtained a sequence for the plastid rps4 gene from a recent collection of B. Jongipes and found that sequence nearly identical to a sample included in earlier analyses. The DNA evidence from both collec- tions place Bestia longipes into a clade that includes other genera generally attributed to the Lembophyllaceae, corroborating earlier conclu- sions. Therefore, we view Bestia and Isothecium as the two North American representatives of the Lembophyllaceae. RARITY AND CONSERVATION IMPLICATIONS Although Bestia is an endemic monospecific genus restricted to coastal California, it is not now of conservation concern. Populations of Bestia are relatively small within a narrow band of suitable habitat but occurrences are numerous and geographically dispersed in the state. Ripar- lan areas also generally have layers of legal protections at both the state and federal levels, especially along those streams that contribute spawning habitat for anadromous fisheries. Bestia occurs on a wide variety of public lands from county and regional parks to state parks, national forests and national parks thus contrib- uting to its long-term conservation. Additional populations of Bestia are likely to be documented with ongoing exploration and collection within the range of this species. We suspect that populations have been overlooked when Bestia is inadvertently assumed to be Jsothecium in the field, and therefore, it is probably under-collect- ed. ACKNOWLEDGMENTS We thank the curators of CAS, MICH, MO, NY, SFSU, UBC, and UC for access to specimens. We also thank Ken Kellman, David Toren, Tarja Sagar, Carl -Wishner, Ben Carter, and Kerry Heise for providing us with their field insights and observations regarding Bestia longipes. Comments provided by Dave Wagner _and an anonymous reviewer improved the final version. ) LITERATURE CITED ALLEN, B. 1983. Isothecium myosuroides Brid. and I. | stoloniferum Brid. A quantitative study. The | Bryologist 86:358—364. ANDREWS, A. L. 1952. Taxonomic notes XI. The case | of Tripterocladium brewerianum. The Bryologist | 55:166-172. | BROTHERUS, V. F. 1925. Musci. In A. Engler, K. Prantl, and Die NatiirlichenPflanzenfamilien (eds.), ed. 2, Vol. 11. Engelmann, Leipzig. Buck, W. R. AND B. GOFFINET. 2000. Morphology and classification of mosses. Pp. 71-123 in A. J. Shaw and B. Goffinet (eds.), Bryophyte biology. Cambridge University Press, Cambridge, UK. : BESTIA LONGIPES 205 CARDOT, J. AND T. THERIOT. 1902. Papers from the Harriman Alaska Expedition. XXIX. The mosses of Alaska. Proceedings of the Washington Acad- emy of Sciences. Washington, D.C. AND . 1906. On a small collection of mosses from Alaska. University of California Publications in Botany 2:297—308. CHAPMAN, C. J. AND E. I. SANBORN. 1941. Moss flora of the Willamette Valley, Oregon. Monographic Studies in Botany No. 4, Oregon State, Corvallis. CHRISTY, J. A., J. H. LYFORD, AND D. H. WAGNER. 1982. Checklist of Oregon mosses. The Bryologist 83:355-358. CrosBYy, M., R. E. MAGILL, B. ALLEN, AND S. HE. 2000. A checklist of the mosses. Missouri Botanical Garden, St. Louis. CRUM, H. 1987. Bestia, Tripterocladium, and Isothe- cium: an explication of relationships. The Bryolo- gist 90:40-42. . 1991. A partial clarification of the Lembophyl- laceae. Journal of the Hattori Botanical Laborato- ry 69:313-—322. DRAPER, I., L. HEDENAS, AND G. W. GRIMM. 2007. Molecular and morphological incongruence in European species of /sothecium (Bryophyta). Mo- lecular Phylogenetics and Evolution 42:700—716. GOFFINET, B. AND W. R. BUCK. 2004. Systematics of the bryophyta (mosses): from molecules to a revised classification. Pp. 205—239 in B. Goffinet, V. Hollowell, and R. Magill (eds.), Molecular systematics of bryophytes. Monographs in System- atic Botany Vol. 98, Missouri Botanical Garden Press, St. Louis, MO. GROUT, A. J. 1928. Moss Flora of North America north of Mexico. Volume III, Part 1. Published by the author, Staten Island, New York. HUTTUNEN, S., M. S. IGNATOV, K. MULLER, AND D. QUANDT. 2004. Phylogeny and evolution in the three moss families Meteoriaceae, Brachythecia- ceae, and Lembophyllaceae. Pp. 328-356 in B. Goffinet, V. Hollowell, and R. Magill (eds.), Molecular systematics of bryophytes. Monographs in Systematic Botany Vol. 98, Missouri Botanical Garden Press, St. Louis, MO. KocHu, L. F. 1950. Mosses of California: an annotated list of species. Leaflets of Western Botany 6:1—40. LAWTON, E. 1971. Moss flora of the Pacific Northwest. Hattori Botanical Laboratory, Nichinan, Miya- zaki, Japan. LESQUEREUX, L. 1868. Catalogue of Pacific Coast mosses. Memoirs of the California Academy of Sciences, Vol. 1 Part 1. San Francisco. AND T. P. JAMES. 1884. Manual of the mosses of North America. Bradlee Whidden, Boston, MA. Norris, D. H. AND J. ENROTH. 1990. Description of Bryolawtonia Norris & Enroth (Thamnobryaceae), with observations on the genus Bestia. The Bryologist 93:328—-331. AND J. R. SHEVOCK. 2004a. Contributions toward a bryoflora of California. I. A specimen- based catalogue of mosses. Madrono 51:1—131. AND . 2004b. Contributions toward a bryoflora of California. II. A key to the mosses. Madrono 51:133—269. QUANDT, D., S. HUTTUNEN, R. TANGNEY, AND M. STECH. 2008. Back to the future? Molecules take us back to the 1925 classification of the Lembophyl- 296 laceae (Bryopsida). Systematic Botany 33 (in press). RYALL, K., J. WHITTON, W. SCHOFIELD, S. ELLIS, AND A. J. SHAW. 2005. Molecular phylogenetic study of interspecific variation in the moss Jsothecium (Brachytheciaceae). Systematic Botany 30:242—247. SCHOFIELD, W. B. 2004. Endemic genera of bryophytes of North America (north of Mexico). Preslia, Praha 76:255—277. SHEVOCK, J. R. AND D. TOREN. 2001. A catalogue of mosses for the City and County of San Francisco, California. Madrono 48:1-—18. SULLIVANT, W. S. 1864. Icones Muscorum or figures and descriptions of most of the mosses peculiar to MADRONO, Vol. 55, No. 4, p. 296, 2008 MADRONO [Vol. 55 eastern North America (facsimile printed in 1969 by Asher & Company, Amsterdam). 1874. Icones Muscorum Supplement or figures and descriptions of most of the mosses peculiar to eastern North America (facsimile printed in 1969 by Asher & Company, Amster- dam). TANGNEY, R. 2007. Biogeography of Austral pleuro- carpous mosses: distribution patterns in the Aus- tralasian Region. Pp. 393-407 in A. E. Newton and R. S. Tangney (eds.), Pleurocarpous mosses sys- tematics and evolution. The Systematics Associa- tion Special Volume Series 71. CRC Press, Boca Raton, FL. REVIEW Flora of China Illustrations. Vol. 22. Poaceae. By Wu Zhengyi, Peter H. Raven (editorial commit- tee co-chairs), Hong Deyuan (editorial committee vice co-chair). 2007. Science Press (Beiing, China) and Missouri Botanical Garden Press (St. Louis, MO). 937 pp. Hardcover. $140.00. ISBN 978-1-930723-61-0. China harbors an incredible diversity of plants, including ca. 31,000 fern and seed plant species, or one-eighth of the world’s vascular flora. By comparison, the United States and Canada—together twice the size of China in land area—have one-third fewer species (ca. 20,000). There are ca. 1,800 grass (Poaceae) species in China, including 534 species of woody bamboos. Flora Reipublicae Popularis Sinicae (FRPS), a Chinese-language Flora of the country, was published as 80 volumes in 125 books from 1959 until its completion in 2004. The Flora of China (FOC) is an English-language revision of FRPS that will comprise 49 volumes when complete, including 24 volumes devoted to illustrations. As of December 2008, 16 text volumes and 14 illustrations volumes had been published, the first in 1994. The Poaceae text volume was published in 2006. The FOC is unusual in having separate text and illustrations volumes. Unfortunately, for the Poaceae (and perhaps all) treatments, the text volume does not reference figures in the illustra- tions volume, which means that the user cannot tell from the text volume if a species is illustrated in the companion volume nor, if it is illustrated, on what page the figure appears. Because the genera in both volumes are arranged by tribe, not alphabetically, the index in the illustrations volume must be consulted to find out if a species is figured, and on what page. All of China’s 226 grass genera (including native and naturalized species and economically important exotics) in 28 tribes are illustrated and arranged as in the text volume, beginning with Bambuseae and ending with Andropogoneae. Seventy-one percent (1,271) of the species and 74 infraspecific taxa are illustrated as line drawings in 904 full-page plates, and a lovely color figure of Phyllostachys reticulata serves as the frontis- piece. Indeed, all of the line drawings, inked by 76 illustrators, are beautiful and detailed. Most were published previously in FRPS, although many of these were redrawn for FOC. Text in the plates is minimal—only scale bars, which are given for about half of the plates, and numbers identifying the figure parts which are explained in the captions. The captions also include the scientific, Chinese, and pinyin names of the taxa, page references to where the taxa are treated in the FOC Poaceae text volume and FRPS, the names of the artists, and the sources of illustrations not | published in FRPS. There are three indices of names—Chinese, pinyin, and scientific. | The Poaceae illustrations volume is a vital; companion to the text volume for those learning | and identifying Chinese grasses. In fact, as one-. sixth of the world’s grass species are found in» China, both volumes should be close at hand for | anyone interested in the diversity of this large, | ecologically and economically important plant. family. —J. TRAVIS COLUMBUS, Research Scientist, Rancho Santa | Ana Botanic Garden, and Associate Professor, Claremont | Graduate University, 1500 N. College Ave., Claremont, CA 91711-3157; j.travis.columbus@cgu.edu. MADRONO, Vol. 55, No. 4, pp. 297-302, 2008 A TAXONOMIC REASSESSMENT OF CLARKIA CALIENTENSTS AND CLARKIA TEMBLORIENSIS FRANK C. VASEK'” Department of Botany and Plant Sciences, University of California, Riverside, CA 9252] ABSTRACT The taxonomic relationships of Clarkia calientensis were re-evaluated in light of information accumulated subsequent to its designation as C. tembloriensis subsp. calientensis. Closeness of relationship between C. calientensis and C. tembloriensis was based on ease of crossing and on cytological pairing configurations in intertaxon hybrids. These interpretations require the assumptions that closely related species are easy to cross, and that maximum cytological configurations are known. However, extensive crossing data show that sympatric, closely related species of Clarkia are very difficult to cross but allopatric species cross rather readily. Furthermore, the cytological data are incomplete as maximum configurations are seldom observed, and some hybrids have not been produced. In both cases, the required assumptions are not tenable. Sterile hybrids between these two easily crossed taxa are strong evidence that they are not conspecific. Consequently, C. calientensis should be recognized as an independent species. Examination of floral variation in C. tembloriensis led to the recognition of large-flowered, protandrous forms as a new subspecies: C. tembloriensis subsp. longistyla. Key Words: Clarkia calientensis, C. tembloriensis, crossability, hybrid sterility, outcrossing rate, translocations. This paper is dedicated in honor of Professor F. Harlan Lewis, who passed away in December, 2008. Clarkia calientensis Vasek (Onagraceae) occurs only in the Caliente Hills, east of Bakersfield (Vasek 1977). Clarkia tembloriensis occurs in the arid, Inner Coast Ranges from Alameda and San Joaquin Counties (Corral Hollow) southward to San Luis Obispo and Kern Counties (Temblor Range), and in the Caliente Hills and Tejon Hills, east of Bakersfield (Vasek 1977). They are related to each other, and to C. exilis and C. spring- villensis of the Sierra foothills in Kern and Tulare Counties, as well as to the widespread C. unguiculata from which all four were probably derived (Vasek 1958, 1964a, 1968, 1977). The derivatives are allopatric with each other, but sympatric, or marginally so, with C. unguiculata -at the low elevation margin of its range. These derivative species are similar to one another but were first recognized in the field on morpholog- ical traits and subsequently subjected to extensive Studies of distribution, ecology, flowering, devel- opment, cytology, and breeding systems (Vasek 1977, and included references). Nevertheless, they ‘may be difficult to distinguish as herbarium specimens. _ Astudy was undertaken by Holsinger (1985) to determine how these species correspond with phenetic groupings. Upon completion of this phenetic study, a reinterpretation of then existing "Retired *Present address: 3418 Mono Pl., Davis, CA 95618. cytogenetics, crossing relationships, and ecological data resulted in the conclusion that C. temblor- iensis and C. calientensis had origins independent of C. unguiculata, and should be considered conspecific. Furthermore, C. calientensis Vasek was transferred to C. tembloriensis as a subspecies, C. tembloriensis subsp. calientensis Holsinger. Since then, new collections, and extensive re- search, particularly in population dynamics and breeding system functions have added new infor- mation and insights. The present paper reviews the available information, both old and new, and clarifies the relationships of these species. TAXONOMIC STATUS OF C. CALIENTENSIS The phenetic study by Holsinger (1985) in- cludes two main parts: a phenetic study proper, and a reinterpretation of existing information. The phenetic study is carefully detailed and well done. A first ordination, based on petal size, ovary pubescence and stigma exsertion against ovary size (length and width), determined three clusters corresponding to: Clarkia unguiculata, C. exilis, and a mixed group of C. springvillensis, C. tembloriensis, and C. calientensis. A second ordination of the latter group separated out C. springvillensis by ovary size and its purple sepals. The continued grouping of C. tembloriensis and C. calientensis could not be resolved. However, a secondary ordination of just the mixed subgroup (Fig. 3 in Holsinger 1985) showed essentially separate groupings for, not only C. springvillensis, but also for C. calientensis, and for large-flowered and small-flowered forms of C. tembloriensis. The 298 discussion indicates that petals of C. calientensis are larger than those of small-flowered C. tembloriensis and about the same as _ large- flowered C. tembloriensis. Further, C. calientensis differs from both forms of C. tembloriensis by a more slender ovary and slightly broader leaves. Despite methodological limitations imposed by reliance on character traits of flowering herbar- ium specimens, the phenetic study generally agrees with a larger phenotypic study (Vasek 1977) of the same species, which enjoyed the luxury of developmental traits and larger popu- lation samples and provided an estimate of phenotypic plasticity. Both studies show a close similarity, but not identity, of C. calientensis with small-flowered C. tembloriensis. The second part of the phenetic study is more speculative. On the basis that hybrids were readily produced between C. calientensis from the Caliente Hills and C. tembloriensis from the Temblor Range, it concludes that these two species are conspecific and did not have indepen- dent origins in C. unguiculata (as suggested by Vasek 1964a, 1968, 1977). This interpretation requires the assumption that closely related species are easily crossed. Closely related deriv- ative species generally occur adjacent to parental species (Lewis and Raven 1958) and probably originated about where they are found now. With sympatric or even marginally sympatric species, some barrier to crossing must be in place upon, or soon after, origin of a derivative species. Otherwise, the original species and its derivative would cross breed back to panmixia. One possible isolating mechanism might be based on control or selection of the pollen genotypes that actually function in particular stigmas and styles. A significant body of data has been accumulating recently on this and related topics (Bowman 1987; Jones 1994; Kerwin and Smith-Huerta 2000; Smith-Huerta 1996b, 1997; Smith-Huerta and Vasek 1984). Another possible isolating mechanism is hybrid sterility occasioned, per- haps, by chromosomal translocations (Vasek 1958, 1964a, 1968). Sympatric, or marginally sympatric species of Clarkia are actually very difficult to cross, and some hybrid combinations have not been produced despite numerous attempts. Thus, strong isolating mechanisms are in place. In addition, the few hybrids produced are usually highly sterile, often as a consequence of chromosome rearrangements. In contrast, allopatric species are rather easily crossed (Vasek 1958, 1964a, 1968), but their hybrids are also usually sterile. In any event, the required assumption regarding crossability and the close- ness of relationship is not supported. The second part of the phenetic study also calls upon cytological data to indicate that C. calien- tensis and C. tembloriensis are more closely related to each other than to any other taxon of MADRONO [Vol. 55 this group. Most interspecific hybrids in this group of Clarkia are heterozygous for a number of translocations, often 5 or 6 (Vasek 1964a, 1968). Four or more translocations are indicated (citing Vasek 1964a) as the difference between C. calientensis and C. unguiculata, compared with only three translocation differences between C. calientensis and C. tembloriensis (citing Vasek 1968 as 1967). An interpretation like this requires that maximum translocation configurations are known for the interspecific hybrid combinations being considered. However, the former hybrid discussed here has not been produced despite 30 attempts (Vasek 1968). In the latter hybrid, the maximum configuration may or may not have been observed, since only one cell, out of 55 analyzed, showed a closed ring of 8 chromo- somes. Maximum translocation configurations are seldom observed, and hence are often problematical. An attempt to construct detailed relationships based on the perception that one hybrid may have one more translocation than another exceeds the capability of the data available. The evidence, therefore, does not support the conclusion. The second part of the phenetic study also suggests derivation of C. calientensis in the Caliente Hills from C. tembloriensis in the Temblor Range. Given the distance and the unsuitable terrain between, such an event 1s geographically highly improbable. Clarkia tem- bloriensis (large flower type) does occur in the upper, eastern portion, and C. calientensis occurs in the lower western portion, of the Caliente Hills, but they are not sympatric. C. unguiculata occurs between and is marginally sympatric with both. Hence, the geography would favor either C. unguiculata or large-flowered C. tembloriensis as a possible species of origin. Whereas C. unguicu- lata is well known for chromosomal innovation, C. tembloriensis 1s cytologically uniform as far as is known. Therefore, C. unguiculata seems a more likely source of origin for C. calientensis, as it likely was for C. exilis, C. springvillensis, and C. tembloriensis. | Finally, the sterility of hybrids between C._ calientensis from the Caliente Hills and C.. tembloriensis from the Temblor Range is strong, evidence that the two are not conspecific. ; Accordingly, the evidence presented does not support the reinterpretation of data in the second part of the phenetic study. Therefore, it is. recommended, that Clarkia calientensis be recog- nized at the species level. TAXONOMY OF CLARKIA TEMBLORIENSIS Floral Variation Clarkia tembloriensis was described (Vasek. 1964b) primarily from material under study from | 2008] TABLE 1. VASEK: TAXONOMY OF CLARKIA 299 SYNOPTIC COMPARISON OF FLOWER-TYPES IN CLARKIA TEMBLORIENSIS. Entries for style and petal length are rounded progeny means for experimental plants grown in a greenhouse. CP = crinkled petal. ND = not determined. 'Estimate from a very small population in Cedar Canyon (Temblor Range) having a very high percentage of CP. *In table 3 (Vasek and Weng 1988), samples Tl and T2 are large-flowered out crossers, and samples T3—TS5 are small-flowered outcrossers. Flower type Large outcrosser Flower size index 78-70 Outcrossing rate (%) 87—67 P/O ratio? 160-148 Style length — mm (mean) 22-22 Petal length — mm (mean) 24-21 Kern and San Luis Obispo counties (Vasek 1964a). Plants farther north were scarcely known and not yet available for study, but a few plants from as far north as San Benito County contributed to the species description (Vasek 1964b). Later collections and field studies estab- lished the occurrence of large-flowered popula- tions in San Benito and Kern counties (Vasek and Sauer 1971; Vasek and Harding 1976) and introduced the concept of small-flowered and large-flowered syndromes in Clarkia (Moore and Lewis 1965; Vasek 1971, 1977; Holsinger 1985; Vasek and Weng 1988; Holtsford and Ellstrand 1989, 1992). A large flower syndrome collectively includes large hypanthia, sepals and petals and long styles in an out crossing breeding system. Development of the stigma is delayed for a number of days as the style and other flower parts continue to grow. Hence the flowers attain larger size, before the stigma expands and becomes receptive (Smith- Huerta and Vasek 1984, 1987). Often, a half dozen flowers or so may open before the stigma of the first flower expands in readiness for pollination. Separation of the stigma from the anthers, both in time and space, imposes a requirement for a pollen vector (Vasek 1968). A small flower syndrome collectively includes smaller hypanthia, sepals and petals and a shorter style associated in a breeding system capable of effecting unaided self-pollination. The stigma matures and expands, and is receptive for pollination the first or sometimes the second day of anthesis. Growth of the style stops upon stigma maturation, and growth of petals stops soon after (Vasek 1958). _ The large-flowered group within C. temblor- iensis includes only one prominent flower-type: _large-flowered out crossers (Fig. 1). However, the _small-flowered group includes three prominent _ flower-types: small-flowered out crossers, small- flowered selfers, and crinkled petals (Fig. 2). _ Together, the four flower-types differ in two major variable traits and their included compo- nents: flower size and out crossing rate. A flower size index was developed for small-flowered plants by summing measurements (in mm) of petal width and the lengths of ovary, hypanthi- Small outcrosser Small selfer Small CP 66—55 54-42 42—33 58—51 26-03 61'—08 164-141 73—30 ND 19-16 13-10 13—09 19-14 15—10 1109 um, sepal, style and petal (Vasek 1964a), and extended to large-flowered plants by adaptation of data in a study of phenotypic variation (Vasek 1977). Herbarium specimens examined for this paper fall into the same flower-size index classes, with the same ranges, as just described for experimental material. Outcrossing rates were taken directly from the estimates of Vasek and Harding (1976) based on phenotypic traits, and of Holtsford and Ellstrand (1989) based on isozyme polymorphisms. In addition, a general indication of breeding system characteristics 1s afforded by the pollen/ovule (P/O) ratio, and this information is taken directly from a study by Vasek and Weng (1988). A synoptic comparison of the four flower-types is summarized in Table 1. The type locality for Clarkia tembloriensis is Carneros Rocks (Vasek 1964b) on the east side of the Temblor Range (Kern County) at an elevation of about 1400 ft. This standard small- flowered self-pollinator (Fig. 2a) is common in the Temblor Range region of western Kern and eastern San Luis Obispo counties and northward to Fresno County (Cantua Creek), Merced County (Los Banos Creek), and to Alameda and San Joaquin counties (Corral Hollow). Standard small-flowered plants are fully capable of self-pollination, but some cross-pollination commonly occurs (Table 1). Two prominent variations occur. One variant form, termed small-flowered outcrossers (Smith- Huerta, personal communication), has_ styles somewhat longer than those in standard small- flowered selfers, and some degree of protandry may occur (Fig. 2c). The flowers are generally a little larger, and the outcrossing rate somewhat higher, than in standard small-flowered plants (Table 1). Several small-flowered outcrossing populations occur at elevations above 2300 ft in the Temblor Range, and at lower elevations in the Red Hills (Hughes Canyon), and in the Kettleman Hills and Corral Hollow. A second variant is termed crinkled petal (CP) in which petals are reduced to very narrow (l— 3 mm wide) sepal-like, unexpanded structures in which the blade is scarcely wider than the very short claw (Fig. 2b). CP is conditioned by a single 300 MADRONO [Vol. 55 Fic. 1. Clarkia tembloriensis subsp. longistyla. a) Five consecutive flowers (top to bottom) from flower opening through style and stigma development to pollination; b) a young flower with elongating style and immature stigma; c), an older flower with expanded stigma recently pollinated. Photographs courtesy of N.L. Smith-Huerta. The white line represents a length of 1 cm. recessive gene and occurs in both standard small- flowered populations and in small-flowered outcrossing populations (Vasek 1964a, 1966; Smith-Huerta 1992, 1996a), and therefore has little taxonomic significance. Nevertheless, some populations may have rather high CP frequencies and CP is apparently fixed in one population near Cholame. CP plants have not been observed north of San Luis Obispo County. The final major flower-type consists of large- flowered plants (Fig. la) with long styles (Fig. 1b, c), large flower parts, marked protandry and high outcrossing rates (Table 1). Large populations occur in the Griswold Hills (San Benito Co.) and Tumey Hills (Fresno Co.) and the Caliente Hills (Kern Co.). Protandrous, large- flowered plants have also been collected in Stanislaus Co. (Arroyo del Puerto: Sharsmith in 1935) and in Kern Co. (near Long Tom Mine: Smith in 1941; and near Bakersfield: Davy in 1896). Small-flowered plants had been intercrossed experimentally in numerous combinations, with no sign of any reduction in fertility or chromo- somal uniformity (Vasek 1964a). Large-flowered types had not yet been available for inclusion in the hybridization program, and subsequently were assumed to be completely interfertile with the small flowered types. This assumption has been ably documented by the elegant experimen- tal investigations of pollen germination and pollen tube growth in both types and_ their hybrids (Kerwin and Smith-Huerta 2000; Smith- Huerta 1996b). DISCUSSION AND CONCLUSION Variation in flower size in small-flowered selfers and small-flowered outcrosssers 1S essen- tially continuous (Table 1). Some populations may include both small-flowered types, at least at different times. Both types occur along the length of the species range and are commonly adjacent | in and around the Temblor Range. Interestingly, the small flowered outcrossers have high P/O ratios, comparable to those of large-flowered outcrossers, and higher than those of small- | flowered selfers (Table 1). The somewhat similar occurrence of slightly | longer styles and some degree of protandry was | reported in C. exilis (Vasek 1958), an otherwise | small-flowered selfer. It is tempting to think of small-flowered outcrossers as an intermediate step in the evolution of a self-pollinating system | (see Moore and Lewis 1965). The difference between large-flowered out crossers and small-flowered out crossers is essentially discontinuous, especially for flower size (Table 1; and fig. 3 in Holsinger 1985). The | two types have not been observed in mixed | populations and their ranges apparently do not | overlap. Occasionally, however, an odd plant with small flowers where large-flowered plants with long styles were collected in other years: Griswold Hills: Hesse | 2451 in 1958 and Raven 10866 in 1957; and | Tumey Hills: Vasek 620505-3 in 1962 and Sanders | 35150 in 2008. Similarly, a stunted plant with | very small flowers was collected by Smith-Huerta | s.n. in 2004 in an often-visited study population | of small-flowered outcrossers in the Temblor | Range. These examples of presumed phenotypic | plasticity may reflect adverse weather conditions, as is commonly observed for late season or poor has been collected in a |: population area of a larger flower type. Two — examples are known in which a small-flowered » plant with short styles was collected in a locality - season reduction in size of parts (Vasek 1977; . Holtsford and Ellstrand 1992). Problems of this | sort are few and only minimally inconvenient. On the basis of the evidence reviewed here, the | protandrous plants with large flowers and long, | exserted styles (Fig. la, b, c) are hereby proposed for recognition as a new subspecies, C. temblor- | 2008] VASEK: TAXONOMY OF CLARKIA 301 FIG. 2. Clarkia tembloriensis subsp. tembloriensis. a) A standard small flower from Cantua Creek, with a short style and a receptive stigma among anthers of about the same length; b) a flower with crinkled petals (CP) from Red Rocks near Cholame, with style and stamens about the same length; c) a small flowered outcrosser from McKittrick Road, with an elongating style and an immature stigma exceeding the anthers. Photographs courtesy of N.L. Smith-Huerta. The white line represents a length of 1 cm. iensis subsp. longistyla, and the small-flowered types are proposed to constitute the residual subspecies, C. tembloriensis subsp. tembloriensis. TAXONOMY Clarkia tembloriensis subsp. longistyla Vasek subsp. nov. (Fig. 1)—TYPE: USA CALIFOR- NIA, San Benito County, Griswold Hills, 4.6 mi south of Panoche Road. May 4, 2008. Sanders 35167 (holotype UCR; isotypes: CAS, DAV, MO, NY, RSA, SBBG, SD, UC, US). Sanders 35187 (paratypes: DAV, UCR); Sand- ers 35150 (paratypes: CAS, DAV, RSA, UC, WER)... -- Flores hypanthiis 2-4 mm; sepalis viridibus, interdum ruberis suffusis, 14-18 mm, connatis (in 4s), puberulentis, sine longis patentibus pilis; petalis lavandulis-roseis, interdum unumquidque cum purpurascenti macula, 17-25 mm, unguibus gracilibus integeris longitudibus I- vel 2-plo laminis, laminis plus minusve rhombicis; stame- nibus 8, extimis antheris ruberis vel lavandulis et intimis dilutioribus; stylis 17-25 mm, longiexser- tis, extensis ultra stamenibus, stigmatibus non fungentibus antea 3—6 diebus postanthesin; cap- sulis 1.5—3 cm. Stems erect, to 8 dm, glabrous, glaucous. Leaves lanceolate 2—7 cm, 0.5-2.5 cm _ wide, glaucous, gray-green: petioles 0-13 mm. Inflores- cence axis in bud straight; buds reflexed. Flowers with hypanthia 2-4 mm; sepals 14-18 mm long, green, sometimes tinged with red, connate, puberulent, without long spreading hairs; petals 17-25 mm long, lavender-pink, often with pur- plish spot, with claws slender, entire, lengths 1—2 times blades, with blades more or less rhombic; Stamens 8, with outer anthers red to lavender, the inner paler; styles 17-25 mm, long-exserted, extending beyond stamens, with stigmas not receptive before 3-6 d after anthesis; capsules 1.5—3 cm. Flowering April-May. California. Two disjunct areas: A) Interior Coast Ranges, San Benito, Fresno, and Stanislaus counties; and B) Sierra Foothills, Kern County. Representative collections: CALIFORNIA. Fresno Co.: Tumey Hills, Sanders 35150 (CAS, DAV, RSA, UC, UCR); Mercy Springs, Smith- Huerta s.n. (RSA); S. of Panoche Rd., Lewis & Lewis 911 (UC); Ciervo Hills, Crosby & Morin 14358 (RSA). Kern Co.: Caliente Hills, Vasek 670416-2 (UCR); Bakersfield, Davy 1711 (UC); Long Tom Mine, Smith 351 (JEPS, RSA). San Benito Co.: Griswold Hills, Sanders 35187 (DAV, UCR); Constance & Morrison 2267 (POM, UC); Raven 10866 (RSA, UC); McCaskill 440 (DAV): Holtsford 860506-6 (UCR). Stanislaus Co.: Ar- royo del Puerto, Sharsmith 1753 (UC). Clarkia tembloriensis subsp. tembloriensis Vasek (Fig:-2): Flowers smaller than those of Clarkia temblor- iensis subsp. longistyla, with the hypanthia 2— 3 mm; sepals 10-14 mm, petals 8—16 mm long, 5-8 mm wide; or, sometimes narrower, or re- duced and unexpanded, the claw scarcely distin- guishable from the blade, which may be only 1— 3 mm wide; style short, usually less than 16 mm, not exceeding the stamens, or only slightly so, the stigma receptive usually the first or second day the flower opens. 2n = 18 (Vasek 1964a). Flowering April-May. California. Interior Coast Ranges, western Kern County to Alameda and San Joaquin Counties. 400 to 3300 ft in elevation. Representative collections: CALIFORNIA. Alameda Co.: Corral Hollow, Eastwood & Howell 5291 (RSA). Fresno Co.: Tumey Hills, Vasek 620505-3 (UCR); Ciervo Hills, Preston 2070 (DAV); Cantua Creek, Holtsford s.n. (UCR); Vasek 620505-2 (UCR); W. of Coalinga, Vasek 620505-5 (UCR); Warthan Creek, Helmkamp s.n. (UCR); Coyote Canyon, Janeway 1529 (CHSC). Kern Co.: Carneros Rocks, Vasek SS8OS503-2 (UCR); Carneros Canyon, Holtsford sn. 302 (UCR). Kings Co.: Kettleman Hills, Hoover 3301 (UC); Sanders 22727 (UCR). Merced Co.: Los Banos Creek, Janeway 1597 (CHSC). San Benito Co.: Griswold Hills, Hesse 245] (UC). San Joaquin Co.: Corral Hollow, Hoover 3358 (UC), Taylor 8790 (JEPS, UC). San Luis Obispo Co.: Red Canyon (Cholame), Holtsford s.n. (UCR); Syncline Hill, Preston 1122 (DAV, UCR); Hughes Canyon, Holtsford s.n. (UCR); Temblor Range (McKittrick Rd.), Holtsford s.n. (UCR); Smith-Huerta s.n. (RSA); Temblor Range (3,000 ft), Johannsen 1141 (UC); W. Soda Lake, Vasek 620502-1 (UCR). ACKNOWLEDGMENTS I am grateful to John Strother (UC) for organizing and translating the new subspecies description to Latin, and Alan Smith for reviewing it. My thanks are due to Harlan Lewis (LA) for important suggestions regarding the study, James Shevock (UC) and another reviewer for guidance and corrections on the manuscript, Ellen Dean (DAV) for numerous corrections and clarifica- tions on the manuscript and for making the DAV herbarium facilities and staff available to me, Andrew Sanders (UCR) for making special collections at short notice, Steve Boyd (RSA) for finding a series of specimens not listed by the consortium, and to other individuals at CHSC, DAV, JEPS, RSA, UC and UCR for help with specimens, loans, information, herbarium facilities, or specific collections. LITERATURE CITED BOWMAN, R. N. 1987. Cryptic self-incompatibility and the breeding system of Clarkia unguiculata (Ona- graceae). American Journal of Botany 74:471—476. HOLSINGER, K. E. 1985. A phenetic study of Clarkia unguiculata Lindley (Onagraceae) and its relatives. Systematic Botany 10:155—165. HOLTSFORD, T. P. AND N. C. ELLSTRAND. 1989. Variation in outcrossing rate and population genetic structure of Clarkia tembloriensis (Onagra- ceae). Theoretical and Applied Genetics 78:480— 488. AND . 1992. Genetic and environmental variation in floral traits affecting outcrossing rate in Clarkia tembloriensis (Onagraceae). Evolution 46:216—225. JONES, K. N. 1994. Non-random mating in Clarkia gracilis (Onagraceae): a case of cryptic self incom- patibility. American Journal of Botany 81:195—-198. KERWIN, A. M. AND N. L. SMITH-HUERTA. 2000. Pollen and pistil effects on pollen germination and tube growth in selfing and outcrossing populations of Clarkia tembloriensis (Onagraceae) and _ their MADRONO [Vol. 55 hybrids. International Journal of Plant Sciences 161:895—902. LEwIs, H. AND P. H. RAVEN. 1958. Rapid evolution in Clarkia. Evolution 12:319—-336. Moore, D. M. AND H. LEwis. 1965. The evolution of self-pollination in Clarkia xantiana. Evolution 19:104-114. SMITH-HUERTA, N. L. 1992. A comparison of floral development in wild type and a homeotic sepaloid mutant of Clarkia tembloriensis Onagraceae). American Journal of Botany 79:1423—-1430. 1996a. A comparison of venation pattern development in wild type and a homeotic sepaloid mutant of Clarkia tembloriensis (Onagraceae). American Journal of Botany 83:712-715. 1996b. Pollen tube growth in selfing and outcrossing populations of Clarkia tembloriensis (Onagraceae). International Journal of Plant Sci- ence 157:228—233. . 1997. Pollen tube attrition in Clarkia temblor- iensis (Onagraceae). International Journal of Plant Science 158:519—524. AND F. C. VASEK. 1984. Pollen longevity and stigma pre-emption in Clarkia. American Journal of Botany 171:1183—1191. AND . 1987. Effects of environmental stress on components of reproduction in Clarkia unguiculata. American Journal of Botany 74:1-8. VASEK, F. C. 1958. The relationship of Clarkia exilis to Clarkia unguiculata. American Journal of Botany 45:150—-162. . 1964a. The evolution of Clarkia unguiculata relatives adapted to relatively xeric environments. Evolution 18:26—42. unguiculata. Madrono 17:219-221. Evolution 20:243—244. 1968. . 1964b. Two new species related to Clarkia’ | . 1966. A case of gene homology in Clarkia. | The relationship of two ecologically | marginal, sympatric Clarkia populations. Ameri- can Naturalist 102:25—40. . 1971. Variation in marginal populations of | Clarkia. Ecology 52:1046—-1051. . 1977. Phenotypic variation and adaptation in | Clarkia Section Phaeostoma. Systematic ae 2:251-279. AND J. HARDING. 1976. Outcrossing in natural | populations. V. Analysis of outcrossing, inbreed- . ing, and selection in Clarkia exilis and Clarkia | tembloriensis. Evolution 30:403—411. AND R. H. SAUER. 1971. Progression of | flowering in Clarkia. Ecology 52:1038—1045. AND V. WENG. 1988. Breeding Systems of | Clarkia sect. Phaeostoma (Onagraceae): I. Pollen- | Ovule Ratios. Systematic Botany 13:336—350. MADRONO, Vol. 55, No. 4, pp. 303-305, 2008 A REEXAMINATION OF THE ORIGIN OF FOREST DIFFERENCES AT A SUBALPINE LOCATION IN COLORADO STEVEN A. JENNINGS Department of Geography and Environmental Studies, University of Colorado at Colorado Springs, Colorado Springs, CO 80933 sjenning@uccs.edu ABSTRACT Baker (1991) proposed that forest differences observed on either side of a fence were attributable to differences in grazing. The study location is in a subalpine forest on the Pike National Forest adjacent to the fenced boundary of Colorado Springs watershed land. Grazing on the watershed land has been excluded for over a century, and U.S. Forest Service land has had moderate grazing over the same time period. The Forest Service land supports a relatively dense forest comprised primarily of Pinus aristata. The watershed land has a less dense cover composed primarily of Picea engelmannii. Baker (1991) attributed vegetation differences to differential grazing pressure. Additional information suggests an alternate explanation for the vegetation characteristics of this site. In the 1930's, the Forest Service began a program of monitoring areas where trees had been planted or where forest health was a concern. Included in these photographic records is a 1960 photograph of the study area that documents the differences in forest type and cover are related to tree planting activities on the Forest Service side of the fence. The evidence that these trees are planted 1s based on the linear pattern of trees and a general map that shows where tree planting was done. Key Words: Colorado Rocky Mountains, grazing, Pikes Peak, repeat photography, subalpine forests, tree planting. Baker (1991) published an analysis of a subalpine environment on the southern slope of Pikes Peak, Colorado (Range 68 W, Township 15 S, Section 5). He was interested in differences in forest composition related to the different land uses on either side of a fence (Fig. 1). Erected at the beginning of the twentieth century, the fence separates U.S. Forest Service land from Colorado Springs watershed land. The area on both sides of the fence had been affected by a wildfire in the late 1860’s and land on the Forest Service side of the fence had been used for cattle grazing. Cattle had been excluded from the adjacent watershed land for more than a century. Pinus aristata Engelm. (bristlecone pine) is the most common tree on the Forest Service land while Picea engelmannii (Parry) Engelm. (Engelmann spruce) is the dominate tree on the watershed side. Using tree rings Baker (1991) was able to determine that the P. engelmannii had established between 1886 and 1921 while the P. aristata stands were younger with establishment dates between 1934 and 1952. Baker (1991) concluded that the distinct differences between the two sides of the fence are the result of the differential grazing uses. He hypothesized that grazing on the Forest Service side of the fence had led to drier conditions because of reduced plant cover that favored P. aristata over P. engelmanniti. Forest Service records indicate that there is another reason for this asymmetric tree distribu- tion. Following the establishment of the US. Forest Service in 1905, large portions of the Pike National Forest were reforested through plant- ings. Stahelin (1941) reported that between 1906 and 1941 approximately 32,000 acres in Pike National Forest were planted with trees grown at the nearby Monument Nursery. To monitor the health of these trees, the Forest Service estab- lished a repeat photography program in the 1930’s. The goal of the program was to do repeat photography on a decadal basis. The program was abandoned in the 1960’s, and the records were stored by the Forest Service at the Pikes Peak Ranger District Office in Colorado Springs. One of these photographic sites is located near Baker’s (1991) site. Although the photographic site was positioned southwest of Baker’s site, Forest Service employee J.D. Grover, found the tree pattern along the fence to be of enough interest to photograph it 1960 (Fig. 2). There is no written documentation associated with the photograph that definitively described this loca- tion as being planted, but the linear tree pattern strongly suggests that trees on the Forest Service side of the fence were planted. Other well documented photographs of planted trees else- where on Pikes Peak show this same type of pattern, so it is reasonable to attribute the tree pattern to human intervention (Jennings 2003). This photographic evidence appears to refute Baker’s (1991) conclusion that grazing was the 304 MADRONO [Vol. 55 Fic. 1. the dense Pinus aristata trees behind the campsite. The Colorado Springs watershed land is on the right side and is characterized by Picea englemannii. cause of the differences across the fence. It appears more likely that tree planting has had a much greater impact on the plant distribution than grazing. The ages of the P. aristata trees corresponds with the time of planting and the high density of trees 1s commensurate with planting densities of that time (Jennings 2003). The differences that Baker (1991) attributes to grazing can also be attributed to tree planting. In some areas of the Western United States it is important to investigate the tree planting history of the area in order to understand the forest dynamics for that area (Show 1924; Flora 2003; Carnus et al. 2006). In many cases the documen- tation 1s not readily available. It is apparent that while Baker (1991) was diligent in examining Forest Service records, he never located the photographs. Only their recent discovery by a Forest Service employee provides a better under- standing of the Pikes Peak planting history. The author gained access to the photographic record only after asking the Forest Service about planting documents a year or two earlier. These documents turned out to be valuable sources of The study site in the fall 2007. U.S. Forest Service land is on the left of the fence which is on the edge of | information about the forest history of the Pike | National Forest. Lacking documentation, it. would be prudent for a researcher to examine a study site with the intent of determining if there | are indications that the trees at the site had been | previously planted. Characteristics of planted | forests in the Pike National Forest include trees organized in linear patterns, densely planted | trees, and single species stands. The logistics of tree planting would suggest that these character- | istics would be found in other regions of the — United States. For example, Stahelin (1941) documents that plantations of varying sizes are | located throughout Colorado. Researchers would | be prudent to keep in mind the likelihood of | historical reforestation activities when studying forest stands in Colorado. ACKNOWLEDGMENTS The author wants to thank S. Kelso and S. Cunha for | their insightful comments in reviewing this manuscript. | The staff of the Pikes Peak National Forest, past and present, has been very helpful in gathering information | about forests on Pikes Peak. 2008] JENNINGS: REEXAMINATION OF COLORADO SUBALPINE FOREST DIFFERENCES 305 FIG. 2. LITERATURE CITED BAKER, W. L. 1991. Livestock grazing alters succession after fire in a Colorado subalpine forest. Pp. 84-90 inS. C. Nodvin and T. A. Waldrop (eds.), Fire and the environment: ecological and cultural perspec- tives. USDA Forest Service General Technical Report SE-69, Southeastern Forest Experiment Station, Asheville, NC. CARNUS, J., J. PARROTTA, E. BROCKERHOFF, M. ARBEZ, H. JACTEL, A. KREMER, D. LAMB, K. O’HARA, AND B. WALTERS. 2006. Planted forests and biodiversity. Journal of Forestry 104:65—77. The same site as Fig. 1 taken in 1960. The linear pattern indicates that the trees on the left side of the photograph were planted. FLORA, D. F. 2003. Forest economics research at the Pacific Northwest Research Station, to 2000. USDA Forest Service General Technical Report PNW-GTR-562, Pacific Northwest Research Sta- tion, Portland, OR. JENNINGS, S. A. 2003. Unconsidered impacts of reforestation in the Pikes Peak region, Colorado. Papers of Applied Geography Conferences 26: 399-407. SHOw, S. B. 1924. Some results of experimental forest planting in Northern California. Ecology 5:83—94. STAHELIN, R. 1941. Thirty-five years of planting on the National Forests of Colorado. Rocky Mountain For- est & Range Experiment Station, Fort Collins, CO. MADRONO, Vol. 55, No. 4, pp. 306-313, 2008 NOTEWORTHY COLLECTIONS CALIFORNIA VIBURNUM EDULE (Michx.) Raf. (CAPRIFOLIA- CEAE).—Siskiyou Co., McCloud, in a_= spring-fed montane riparian/montane meadow habitat. Associated species include Salix lucida, Crataegus suksdorfiii, Spiraea douglasii, Ribes nevadense, Prunus virginiana, Rhamnus purshiana, Cornus sessilis, Carex spp., Juncus spp., Mimulus guttatus, Populus tremuloides, and Pinus ponderosa. Girard Ridge USGS 7.5’ quadrangle, T39N R3W NE” Sec. 12, UTM 10 0572283E 4568912N, elevation 945 m, 5 July 2007, L. Lindstrand IIT, s. n. (North State Resources Herbarium! [private], Shasta- Trinity National Forest Herbarium’, JEPS). Previous Knowledge. Viburnum edule has not previ- ously been recorded from California. The species is known to occur across Canada and the northern U.S. in moist forests and swamps. This finding represents the first record of the species in California, and also represents the southernmost-recorded extent of the species on the west coast of North America. At the McCloud site the species was first observed in vegetative and flowering condition on 12 June 2007, when a small amount of material was collected in the field. Following an initial examination, the plant was considered to be an undetermined species of Viburnum. Subsequently, the species was examined by the second author and tentatively identified as Viburnum edule. Additional plant material was collected in the field on 5 July 2007, and a voucher was sent to the Jepson Herbarium for annotation, where herbarium staff confirmed the species identification as Viburnum edule, noting that the determination “‘is obvious”’. Significance. This represents the first recorded obser- vation of Viburnum edule in California, and along with Viburnum ellipticum, 1s the second species of Viburnum known to occur in the state. This also represents a southern extension of the species in the west coast portion of its range. Viburnum edule occurs from Alaska to Newfoundland, south to Oregon, northern Idaho, Colorado, Minnesota, and Pennsylvania; including sixteen states and twelve provinces (NatureServe, 2007, NatureServe Explorer: An online encyclopedia of life [web application], Version 6.2, NatureServe, Arlington, Virginia, Available http://www.natureserve.org/explor- er, Accessed November 19, 2007). Viburnum edule 1s currently designated by the states of Michigan, New York, and Vermont as a threatened species, as endangered by the state of Wisconsin, and as a special concern species in Maine. In the northwestern U.S., Viburnum edule occurs in western, central and north- eastern Washington, northern Idaho, and north-central Oregon (Hitchcock, C. L., A. Cronquist, M. Ownbey, and J. W. Thompson [eds.], 1959, Vascular Plants of the Pacific Northwest, Vol. 4, Seattle, WA), and has no conservation status in those states. The nearest recorded locations are approximately 402 km north of the 'North State Resources, Inc. Herbarium, 5000 Bechelli Lane, Suite 203, Redding, CA 96002. °Shasta-Trinity National Forest Herbarium, 3644 Avtech Parkway, Redding, CA 96002. McCloud site in the Cascade Range of central Oregon at Mink Lake Basin, Lane County, at 1410 m elevation (Oregon Plant Atlas, Oregon Flora Project, [web application], Available http://www.oregonflora.org, [Accessed November 5, 2007]) (R. E. Brainerd #57, September 4, 1997; ZUMREBS7). —LEN LINDSTRAND III, Fisheries/Wildlife Biologist, Terrestrial Biology Program Manager, North State Resources, Inc., 5000 Bechelli Lane, Suite 203, Red- ding, CA 96002. lindstrand@nsrnet.com; and JULIE KIERSTEAD NELSON, Forest Botanist, Shasta-Trinity National Forest, 3644 Avtech Parkway, Redding, CA 96002. jknelson@fs.fed.us CALIFORNIA ATRIPLEX AMNICOLA Paul G. Wilson (CHENOPODIA- CEAE).—Los Angeles Co., City of Malibu, Malibu Lagoon at Pacific Coast Hwy., UTM (NAD 83) 11S 0344636E 3766992N, elev. 10 m, locally common on beach sands, in sandy scrub, edge of salt marsh, and along paths and access roads, 8 Jul 2006, Riefner 06-289 (CANB, CDA, RSA). Orange Co., City of Dana Point, beach south of Salt Creek near Ritz Carlton Hotel at Pacific Coast Hwy., UTM (NAD 83) 0433601E 3702917N, elev. 2 m, uncommon, 22 May 1999, Riefner 99-290 (RSA); same locality, 9 Sep 2005, Riefner 05-658 (RSA); City of Newport Beach, Upper Newport Bay, E of Bayside Dr. at confluence with Big Canyon drainage, UTM (NAD 83) 11S 0418191E 3721588N, elev. 6 m, locally common, edge of salt marsh and in Afriplex lentiformis scrub, 16 Sep 2005, Riefner 05-666 (CANB, CDA, RSA); City of San Clemente, along north-bound I-5 Freeway at Camino de Estrella exit, UTM (NAD 83) 11S 0439445E 3702126N, elev. 67m, frequent, annual grassland and planted Eucalyptus stand, 12 Feb | 2006, Riefner 06-21 (CANB, CDA, RSA); City of | Newport Beach, N shore of Newport Bay, near | Bayview Wy. and Bayview Pl., UTM (NAD 83) IIS | 0419132E 3723855N, elev. 7m, common, edge of | riparian scrub and in coastal sage scrub, 17 Apr 2006, © Riefner 06-155 (CDA, RSA); City of Newport Beach, E © of Newport Bay, Big Canyon Creek drainage near Back | Bay Dr., UTM (NAD 83) 11S 0417997E 3721493N, | elev. 9 m, common, edge of salt marsh, on alkaline flats, | and in disturbed scrub, 6 Aug 2006, Riefner 06-381 | (CDA, RSA); City of San Juan Capistrano, San Juan | Creek, ca. 0.2 mi W of intersection of Paseo Tirador | and Calle Arroyo St., UTM (NAD 83) 11S 0439010E | 3706660N, elev. 31 m, locally common, edge of riparian | woodland and in open field, 4 Aug 2007, Riefner 07-344 (RSA); City of San Juan Capistrano, San Juan Creek, | vicinity of San Juan Creek Rd. and Paseo Christina, | UTM (NAD 83) 11S 0440852E 3707792N, elev. 34 m, | locally common, dirt lot, disturbed creek bank, and | Baccharis salicifolia scrub, 11 Sep 2007, Riefner 07-378 | (RSA); City of San Clemente, vicinity of Calle Vicente | and Ave. Vaquero, UTM (NAD 83) 0440240E | 3702184N, elev. 40 m, locally common on roadside slopes, 26 Oct 2007, Riefner 07-455 (RSA); City of San 2008] Clemente, vicinity of Camino de Los Mares and Calle Nuevo, UTM (NAD 83) 0440303E 3702760N, elev. 45 m, large population on steep slope near residential community, spreading to roadsides and coastal sage scrub, 25 May 2008, Riefner 08-118 (RSA). San Diego Co., University City, Governor Dr. on-ramp at Hwy. 805, UTM (NAD 83) IIS 0482779E 3635380N, elev. 117 m, uncommon, roadside in ruderal vegetation, 19 Nov 2006, Riefner 06-681 (RSA). Previous knowledge. Atriplex amnicola (river salt- bush) was not included in The Jepson Manual or in recent publications listing non-native species or facul- tative exotic wetland plants recently established in California (Taylor and Wilken 1993, in Hickman, ed., The Jepson Manual: Higher Plants of California, University of California Press, Berkeley, CA; Hrusa et al. 2002, Madrono 46: 61—98; DiTomaso and Healy 2003, Aquatic and Riparian Weeds of the West, U.C. Agriculture and Natural Resources Publication 3421, Oakland, CA; DiTomaso and Healy 2007, Weeds of California and other Western States, Vol. 1, Aizoaceae— Fabaceae, U.C. Agriculture and Natural Resources Publication 3488, Oakland, CA; Grewell et al. 2007, Estuarine wetlands, in M.G. Barbour, T. Keeler-Wolf, and A.A. Schoenherr, eds., Terrestrial Vegetation of California, 3rd ed., University of California Press, Berkeley, CA; Riefner and Boyd 2007, J. Bot. Res. Inst. Texas 1: 709-730). Atriplex amnicola, a native of Australia, however, is well established on sea beaches and in coastal scrub near Malibu in Los Angeles County (Welsh 2004, in Flora of North America, Vol. 4, Oxford University Press, New York, NY). A small population has recently been reported from Newport Back Bay in Orange County, where it is spreading rapidly (Clarke et al. 2007, Flora of the Santa Ana River and Environs, Heyday Books, Berkeley, CA). In Western Australia, A. amnicola occurs in coastal regions and inland sites along creeks and the margins of salt lakes (Wilson 1984, in Flora of Australia, Vol. 4, Australian Government Publishing Service, Canberra, Australia.). It is one of a suite of saltbushes utilized for rehabilitation of saltland pasture and harvested for commercial seed production in Western Australia (Barrett-Lennard 2003, Saltland Pastures in Australia- A Practical Guide, Land, Water and Wool, Canberra, Australia; Tranen Revegetation Systems 2005, Austra- lian Native Seed Catalogue, Jolimont, Western Aus- tralia; Stevens et al. 2006, Australian J. Agricultural Res. 57: 1279-1289). Due to its tolerance to saline soils and waterlogged conditions, A. amnicola is also used for reclaiming salt-affected wasteland in developing coun- tries (Asad 2002, Communications in Soil Science and Plant Analysis 33: 973-989: Menzel and Lieth 2003, in Leith and Mochtchenko, eds., Cash Crop Halophytes: Recent Studies, Kluwer Academic Publishers, Dor- drecht). Saltbushes (Atriplex spp.), both native and exotic species introduced from Australia, South and North America are also utilized in large-scale arid land rehabilitation in the Mediterranean Basin (Le Houérou 1992, Agroforestry Systems 18: 107-148). Accordingly, Atriplex seed, including A. amnicola, is readily available on the worldwide market for use in restoring saline soil habitats (B & T World Seeds 2008, List 154—Plants for ‘Salty Conditions, accessed May 2008, ). Significance. First documented report of A. amnicola for San Diego County, and the first populations NOTEWORTHY COLLECTIONS 307 documented outside of the Santa Ana River watershed in Orange County (Roberts 1998, A Checklist of the Vascular Plants of Orange County, California, 2nd ed., F.M. Roberts Publications, Encinitas, CA; Hrusa et al. 2002 /oc. cit.; Rebman and Simpson 2006, Checklist of the Vascular Plants of San Diego County, 4th ed., San Diego Natural History Museum, San Diego, CA: Clarke et al. 2007 Joc. cit.). Atriplex amnicola has not been documented from western Riverside County (Roberts et al. 2004, The Vascular Plants of Western Riverside County, California: An Annotated Checklist, F. M. Roberts Publications, San Luis Rey, CA). Based on field observations in southern California, A. amnicola occupies habitats similar to big saltbush [Atriplex lentiformis (Torrey) S. Watson], which in- cludes saline to moderately alkaline soils of salt marsh transition zones, the coastal strand, and non-saline sage scrub, stream bank and riparian communities, and ruderal or roadside habitats. Atriplex amnicola may have been introduced via seed mixes used to restore/ enhance coastal lagoon ecosystems before more rigor- ous guidelines were adopted that prioritize the use of appropriately adapted native plants over exotic species for natural community revegetation programs (Rodgers and Montalvo 2004, Genetically Appropriate Choices for Plant Materials to Maintain Biological Diversity, University of California, Davis). Atriplex amnicola, however, 1s apparently still being utilized in seed mixes designed for habitat restoration and erosion control projects in coastal Orange County and possibly in San Diego County. It is spreading rapidly from landscaped slopes in urban environments to native plant commu- nities. Owing to its availability on the commercial seed market, ecological adaptability, and widespread success in rehabilitating degraded lands, it is expected elsewhere in the South Coast region. Atriplex amnicola appears to be an invasive shrub that could displace native plants and disrupt natural ecosystems in southern California. ATRIPLEX GLAUCA L. (CHENOPODIACEAE).—Los Angeles Co., Port of Los Angeles, City of San Pedro, Cabrillo Beach along Via Cabrillo Marina Rd. near Shoshonean Rd., UTM (NAD 83) 11S 0380990E 3731603N, elev. 10m, uncommon, growing with Atriplex semibaccata on unstable bluff soils, 21 Jan 2006, Riefner 06-11 (RSA); Palos Verdes Peninsula, vicinity of Via Alar and Palos Verdes Dr. West, UTM (NAD 83) IIS 0371250E 3741020N, elev. 64 m, locally common on brushy slopes, 10 Mar 2007, Riefner 07-121 (RSA); Palos Verdes Peninsula, Portuguese Bend, 0.5 mi S of Peppertree Dr. on Palos Verdes Dr. South, UTM (NAD 83) 11S 0374057E 3734168N, elev. 68 m, uncommon on roadsides and in annual grassland, 21 Apr 2007, Riefner 07-181 (RSA). Orange Co., City of Newport Beach, Upper Newport Bay, general vicinity of Bayview Wy. and Bayview PI., UTM (NAD 83) I1S 0419132E 3723855N, elev. 7 m, common, growing on edge of riparian woodland and in coastal sage scrub, 4 May 2002, Riefner 02-88 (CDA, RSA); City of Newport Beach, E of Newport Bay, E of Back Bay Dr. near Big Canyon Creek, UTM (NAD 83) I1S 0418138E 3721615N, elev. 5 m, common, growing on edge of salt marsh, 4 May 2002, Riefner 02-89 (CDA, RSA); City of Lake Forest, E side of SR-241 ca. 0.5 mi N of Alton Pkwy., UTM (NAD 83) 11S 0436667E 3727953N, elev. 230 m, locally common, growing with Atriplex semi- baccata on roadside and edge of avocado grove, 13 Dec 2005, Riefner 05-784 (CDA, RSA); City of Newport Beach, W side of MacArthur Blvd. near San Joaquin 308 St., UTM (NAD 83) 11S 0419715E 3719909N, elev. 75 m, locally common, growing with Atriplex semibac- cata in power line right-of-way and edge of coastal sage scrub, 1 Feb 2006, Riefner 06-30 (CDA, RSA); City of Newport Beach, near Avocado Ave. and San Nicolas Dr., W of MacArthur Blvd., UTM (NAD 83) 11S 0419405E 3719645N, elev. 77m, locally common, growing with Atriplex semibaccata on roadside and in annual grassland, 19 Mar 2006, Riefner 06-52 (CDA); City of Rancho Santa Margarita, W side of SR-241 near Los Alisos Blvd., UTM (NAD 83) I1S 0441877E 3724130N, elev. 292 m, uncommon in coastal sage scrub, 20 Jun 2006, Riefner 06-253 (CDA, RSA); City of Rancho Santa Margarita, vicinity of Towne Center Dr. and Alton Pkwy., UTM (NAD 83) 11S 0437713E 3726926N, elev. 291m, uncommon in coastal sage scrub and on roadside, 30 Jun 2006, Riefner 06-334 (RSA); City of San Juan Capistrano, San Juan Creek at Hwy. 74, UTM (NAD 83) I1S 0441820E 3709001N, elev. 56 m, locally common on roadsides and sparse sage scrub, 22 Jan 2007, Riefner 07-20 (RSA); Ladera Ranch/San Juan Capistrano area, Antonio Pkwy. ca. 0.6 mi S of Covenant Hills Rd., UTM (NAD 83) 11S 0442042E 3710740N, elev. 170 m, uncommon, roadside and non-native grassland, 22 Aug 2007, Riefner 07-370 (RSA). Riverside Co., City of Corona, along Magnolia Ave. off-ramp on SW side of I-15 Freeway, UTM (NAD 83) 11S O450110E 3747033N, elev. 216 m, uncommon, growing with roadside ruderal plants, 12 Dec 2002, Riefner 02-521 (RSA); City of La Sierra, Gramercy PI. near La Sierra Ave., UTM (NAD 83) 11S 0454476E 3754271N, elev. 227 m, rare, growing with Atriplex suberecta on disturbed lot, 30 Dec 2005, Riefner 05-792 (RSA); City of Corona, vicinity of Green River Rd. and Palisades Dr., UTM (NAD 83) 11S 0440802E 3749217N, elev. 16m, uncommon in coastal sage scrub, 22 Aug 2006, Riefner 06-399 (RSA). San Diego Co., City of La Jolla, E of I-5 Freeway near Town Center Dr., UTM (NAD 83) IIS 0479754E 3638753N, elev. 104m, uncommon in coastal sage scrub and margin of dirt road, 17 Aug 2003, Riefner 03- 339 (RSA, UCR); City of Carlsbad, Carlsbad Blvd. near Tierra del Oro St., UTM (NAD 83) 11S 0468532E 3666310N, elev. 18 m, locally abundant in coastal bluff scrub, 6 Dec 2003, Riefner 03-492 (RSA, UCR); City of Carlsbad, vicinity of Encina Power Station, Carlsbad Blvd., ca. 0.3 mi N of Tierra del Oro St., UTM (NAD 83) 11S 0468471E 3666435N, elev. 13m, _ locally abundant, coastal bluff scrub and coastal strand vegetation, 3 Jul 2005, Riefner 05-544 (CDA, RSA, SD, UCR); City of Carlsbad, Agua Hedionda, inlet channel at Carlsbad Blvd., UTM (NAD 83) 11S 0468013E 3667422N, elev. 4m, locally common on road bank and along fisherman trail with Atriplex semibaccata, 19 Sep 2005, Riefner 05-668 (RSA, UCR); City of Carlsbad, Carlsbad Blvd. at Shore Dr., UTM (NAD 83) 11S 0468823E 3665819N, elev. 14m, uncommon on disturbed lot with Atriplex semibaccata, 19 Sep 2005, Riefner 05-670 (RSA, UCR); City of Carlsbad, South Carlsbad, Carlsbad Blvd. ca 0.2 mi N of Island Wy. at Ponto Beach, UTM (NAD 83) 11S 0469682E 3664127N, elev. 9 m, uncommon, disturbed roadside and on ocean bluff, 23 Nov 2005, Riefner 05- 767 (CDA, RSA); City of Del Mar, SW ca. 0.2 mi from intersection of Carmel Valley Rd. and Via Mar Valle Rd., UTM (NAD 83) 11S 0475686E 3644616N, elev. 19m, common, spreading from roadsides to native coastal sage scrub vegetation with Encelia californica MADRONO [Vol. 55 and Rhus integrifolia, 16 Jan 2006, Riefner 06-1 (RSA, UCR); City of Oceanside, Buena Vista Lagoon, vicinity of Rue des Chateaux at Ocean St., UTM (NAD 83) 11S 0466584E 3669564N, elev. 3 m, locally established on sand dune, 15 Oct 2007, Riefner 07-430 (RSA). Previous knowledge. Atriplex glauca (gray saltbush or glaucous-leaved saltbush) is native to the Mediterra- nean Basin (Tutin et al., eds., 1980, in Flora Europaea, Vol. 1, Cambridge University Press, New York, NY). It has not been included in treatments of the Chenopo- diaceae in floras covering North America or other major publications addressing non-native plants estab- lished in California (Taylor and Wilken 1993 Joc. cit.; Hrusa et al. 2002 loc. cit.; Welsh 2004 loc. Cit.; DiTomaso and Healy 2007 Joc. cit.; Grewell et al. 2007 loc. cit.). Atriplex glauca, however, is well established outside of cultivation in coastal southern California, but has not been reported previously from western Riverside County or San Diego County (Roberts et al. 2004 Joc. cit.; Rebman and Simpson 2006 /oc. cit.). It was likely first introduced in seed mixes used for erosion control programs (Boyd 1999, Aliso 18: 93-139; Clarke et al. 2007 Joc. cit.). Atriplex glauca seed has been available on the commercial market in California since at least the early 1990’s, and is well known for its hardiness and tolerance to alkaline and harsh soil conditions (City of Riverside Planning Department 1994, Water Efficient Landscap- ing and Irrigation Ordinance Summary and Design Manual, Riverside, CA; USDA 1995, Commercial Suppliers of Tree and Shrub Seed in the United States, Miscellaneous Report M8-MR 33, Washington, DC; S & S Seeds 2005, Seed Selection Guide, Carpinteria, CA). Significance. First documented reports of A. glauca for Riverside and San Diego counties, and extensions of range for Los Angeles County outside the Liebre Mountains, and for Orange County outside of the — Santa Ana River watershed (Boyd 1999 loc. cit.; Rebman and Simpson 2006 Joc. cit.; Clarke et al. 2007 loc. cit.). This shrub is more common than herbarium | records indicate, and is spreading rapidly from land- | scaped areas to disturbed urban habitats and native | plant communities. Much like A. amnicola, A. glauca is | a well-known halophyte (Menzel and Lieth 2003 loc. | cit.). Additional populations of A. glauca are to be | expected throughout the South Coast region and in | many habitats, including the coastal strand, salt marsh- | upland transition zones, dune swales, coastal sage | scrub, non-native grassland, and ruderal or roadside | habitats. Based on field observations, A. glauca occupies ' habitats similar to Australian saltbush (Atriplex semi- | baccata L.), which is an aggressive invader that | displaces native plants and disrupts natural ecosystems | (Bossard et al. 2000, Invasive Plants of California’s — Wildlands, UC Press, Berkeley, CA). Atriplex glauca is | more robust and apparently more broadly adapted than | A, semibaccata, and is invading undisturbed coastal | sage scrub communities where 4. semibaccata does not. | Owing to its invasiveness, A. glauca should not be | planted outside of managed landscapes or for erosion control projects that are adjacent to native plant. communities or natural open space areas without | careful consideration (University of California Coop- | erative Extension 2000, Estimating the Irrigation Water Needs of Landscape Plantings in California, Depart- ment of Water Resources, Sacramento, CA). 2008] ATRIPLEX MUELLERI Benth. (CHENOPODIACEAE).— Imperial Co., E of El Centro, along Ross Rd. near Bass Cove Rd., N of Hwy. 8, UTM (NAD 83) 11S 0638008E 3628013N, elev. ca. —9 m, common along irrigation ditch and on roadsides, 9 Oct 2006, Riefner 06-513 (CANB, CDA, RSA); E of El Centro, Ross Rd. ca. 0.2 mi E of Dogwood St., N of Hwy. 8, UTM (NAD 83) 11S 0637549E 3628051N, elev. ca. —9 m, common on vernal alkaline flats with Suaeda moquinii, 9 Oct 2006, Riefner 06-515 (CANB, CDA, RSA); E of El Centro, Ross Rd. at Hawkeye Rd. off Hwy. 111, UTM (NAD 83) 11S 0640473E 3628049N, elev. ca. —8 m, common along roadside in ruderal vegetation, 9 Oct 2006, Riefner 06-519 (RSA); near Seeley, along Drew Rd. at Hwy. 8, UTM (NAD 83) 11S 0622670E 3627210N, elev. ca. —10m, uncommon, roadside depression with Malvella leprosa, 9 Oct 2006, Riefner 06-520 (RSA). Previous knowledge. Atriplex muelleri (Mueller’s saltbush) is native to Australia (Wilson 1984 Joc. cit.; Barker et al., eds., 2005, Census of South Australian Vascular Plants, Sth ed., J. Adelaide Bot. Gard. Supplement 1, Botanic Gardens of Adelaide and State Herbarium, Adelaide). It was not included in the treatment of Atriplex for California by Taylor and Wilken (1993 Joc. cit.), and Welsh (2004 Joc. cit.) was unable to verify its presence in North America. Atriplex muelleri has recently been reported, however, for Riverside County from specimens collected in 1965 near Blythe (Roos s.n.; COLO, UC, UCR!) (Hrusa et al. 2002 /oc. cit.). The Blythe population has apparently been extirpated (Hrusa et al. 2002 Joc. cit.). Significance. First documented reports of A. muelleri for Imperial County. Atriplex muelleri is a halophyte, and seed is available for purchase on the world market for use in restoring degraded saline soil habitats (Menzel and Lieth 2003 Joc. cit.; B & T World Seeds 2008 Joc. cit.). Atriplex muelleri is similar to A. semibaccata (Welsh 2004 Joc. cit.). Therefore, it may be overlooked in the field and could be more widespread than our records indicate. Atriplex muelleri is expected to occur elsewhere in disturbed and ruderal habitats, and in low-lying, vernally moist alkaline sites in the California deserts and the inland valleys of the South Coast region, especially the saline-alkali vernal plains of western Riverside County. ATRIPLEX STIPITATA Benth. (CHENOPODIACEAE).— Los Angeles Co., City of San Pedro, Cabrillo Beach Park near Salinas de San Pedro, UTN (NAD 83) 11S 0380844E 3730983N, elev. 13 m, local in coastal sage scrub exhibit garden and uncommon on nearby bluff slope, 21 Jan 2006, Riefner 06-11 (CANB, CDA). Previous knowledge. Atriplex stipitata (mallee saltbush or kidney saltbush) is native to the semi-arid regions of Southern Australia (Wilson 1984 Joc. cit.; Walsh 1996, Chenopodiaceae, in N.G. Walsh and T.J. Entwistle, eds., Flora of Victoria, Vol. 3, Inkata Press, Melbourne, Australia; Barker et al. 2005 Joc. cit.). It has not been reported previously for North America (Taylor and Wilken 1993 Joc. cit.; Welsh 2004 Joc. cit.). Atriplex Stipitata was also not included in recent publications addressing non-native species established in California or in local floras covering the South Coast region (Roberts 1998 Joc. cit.; Hrusa et al. 2002 Joc. cit.; Roberts et al. 2004 /oc. cit.; Rebman and Simpson 2006 /oc. cit.; Clarke et al. 2007 Joc. cit.; DiTomaso and Healy 2007 loc. cit.; Riefner and Boyd 2007 Joc. cit.). Atriplex Stipitata, a halophyte, is sold on the worldwide market NOTEWORTHY COLLECTIONS 309 for rehabilitation of saline soil habitats (Menzel and Lieth 2003 Joc. cit.; B & T World Seeds 2008 Joc. cit.). Significance. First documented report of A. stipitata for North America. Atriplex stipitata, a 1 m tall shrub, has narrowly elliptic to orbicular leaves and could be confused with small forms of A. /entiformis in southern California. Atriplex stipitata, however, is readily sepa- rated from A. /entiformis by its reniform, cordate bracteoles, generally 5 mm long and 10 mm wide, that are born on a slender stipe in open panicles (Wilson 1984 Joc. cit.). Atriplex stipitata is likely more wide- spread in urban environments and coastal native plant communities than our records indicate. It is one of a number of non-native unidentified Atriplex species now widely established in the South Coast region that could be easily mistaken for A. /entiformis. These undeter- mined Atriplex species are often vegetatively similar to A. lentiformis, but differ by the long-stipitate, often denticulate bracteoles that are cordate, reniform, or orbicular in shape. Since A. stipitata and several other unidentified taxa may be difficult to identify in the field, seed may be inadvertently collected and included in native seed mixes applied during native scrub restoration projects, thereby spreading exotics to new sites in southern California’s native plant communities. Accordingly, A. stipitata 1s expected to occur elsewhere on coastal bluffs, scrub, and lagoon habitats in the South Coast region. Botanists, restoration ecologists, seed collectors, and the resource agencies should be vigilant regarding the often locally abundant and widespread distribution of many of the non-native Atrip/ex in southern California. ATRIPLEX UNDULATA L. (CHENOPODIACEAE).—San Diego Co., City of Del Mar, San Dieguito Wetlands, Racetrack View Dr. ca. 0.25 mi W of Racetrack View Ct., UTM (NAD 83) IIS 04764484E 3647165N, elev. 20 m, locally common on alkaline sandy flats and in open coastal sage scrub around the lagoon, 1 Mar 2003, Riefner 03-65 (CDA, RSA); same locality, 19 Apr 2003, Riefner 03-218 (CANB, CDA, RSA); City of Del Mar, San Dieguito Wetlands, San Dieguito Dr. ca. 0.25 mi from Jimmy Durante Rd., UTM (NAD 83) IIS O475704E 3647724N, elev. 3 m, locally common on roadside and upland transition zone along salt marsh, 6 Aug 2006, Riefner 06-367 (CDA, RSA). Previous knowledge. Atriplex undulata (wavy-leaved saltbush) is native to Argentina (Patagonia and Buenos Aires) (Cabrera 1963/1970, Flora de la Provincia de Buenos Aires, Col. Cient. INTA, Buenos Aires, Argentina; Carretero 2001, Multequina 10: 67—74). It is widely planted for forage and is adventive in Western Australia (Hussey and Lloyd 2002, Western Weeds, Additions, Deletions and Name Changes, Department of Conservation and Land Management, Western Australia; Barrett-Lennard 2003 /oc. cit.). In| North America, A. undulata has not been included in treatments of the Chenopodiaceae in major floras that include non-native exotic plants established in Califor- nia or in local floras covering the South Coast region (Taylor and Wilken 1993 Joc. cit.; Roberts 1998 Joc. cit.; Hrusa et al. 2002 Joc. cit.; Roberts et al. 2004 /oc. cit.; Welsh 2004 Joc. cit.; Rebman and Simpson 2006 /oc. cit.; Clarke et al. 2007 Joc. cit.; DiTomaso and Healy 2003, 2007 loc. cit.; Grewell et al. 2007 Joc. cit.; Riefner and Boyd 2007 Joc. cit.). Atriplex undulata, a halophyte, is another species of a suite of saltbushes frequently utilized for rehabilitation of saltland pasture and harvested for commercial seed 310 production in Western Australia (Barrett-Lennard 2003, Joc. cit.; Menzel and Lieth 2003 Joc. cit.; Tranen Revegetation Systems 2005 Joc. cit.). Its seed is also available on the world market, specified for use in rehabilitating saline soils (B & T World Seeds 2008 Joc. cit.). Atriplex undulata, much like A. amnicola and A. stipitata, may have been imported for coastal lagoon enhancement projects before the consequences of introducing exotic species into salt marsh and transi- tional upland habitats were fully understood (Callaway and Zedler 2004, Urban Ecosystems 7: 107-124; Rodgers and Montalvo 2004 Joc. cit.). Significance. First documented reports of A. undulata for North America. Its unique wavy-margined leaves and small bracteoles make for easy identification. Owing to its availability on the commercial seed market, and use in rehabilitating degraded saline lands, it is likely that A. undulata will continue to be applied in erosion control and habitat enhancement projects in southern California. Atriplex undulata is expected to occur elsewhere in coastal scrub, salt marsh transition zones, and other lagoon habitats in southern California. The potential to utilize saline water for irrigation while incorporating halophytes for developing sustain- able agricultural production, animal fodder, and rehabilitation of degraded lands are promising, attain- able goals for the Mediterranean Region and subtrop- ical dry regions where freshwater resources are limited and conventional agricultural products are unsuited for strongly saline soils (Koyro 2003, in Leith and Mochtchenko, eds., Cash Crop Halophytes: Recent Studies, Kluwer Academic Publishers, Dordrecht). Atriplex, which is comprised of nearly 150 species identified as halophytes, will likely become an impor- tant component of the crops utilized during these efforts (Osmond et al. 1980, Physiological Processes in Plant Ecology: Towards a Synthesis with Atriplex, Springer- Verlag, Berlin; Asad 2002 Joc. cit.; Debez et al. 2003, in Leith and Mochtchenko, eds., Cash Crop Halophytes: Recent Studies, Kluwer Academic Publishers, Dor- drecht; Menzel and Lieth 2003 Joc. cit.). The dilemma now presented is how to balance the growing demand to cultivate non-native crops that provide economically beneficial uses in one region while restricting imports to minimize the risk of introducing potentially noxious weeds or invasive pests into native ecosystems elsewhere (Barney and DiTomaso 2008, BioScience 58: 64-70). Therefore, we should expect to find other non-native Atriplex species established in southern California’s alkaline coastal ecosystems. —RICHARD E. RIEFNER, JR., Research Associate, Rancho Santa Ana Botanic Garden, 1500 North College Avenue, Claremont, CA 91711. rriefner@ earthlink.net; G. FREDERIC HRUSA, California Depart- ment of Food and Agriculture, Botany Laboratory and Herbarium, Plant Pest Diagnostics Center, 3294 Mead- owview Road, Sacramento, CA 95832-1448; and DAVID MALLINSON and BRENDAN LEPSCHI, Australian Na- tional Herbarium, Centre for Plant Biodiversity Re- search, GPO Box 1600, Canberra, Australia, 2601. CALIFORNIA ELEUSINE CORACANA (L.) Gaertn. subsp. AFRICANA (Kenn.-O’Byrne) Hilu & De Wet (POACEAE).—Los MADRONO [Vol. 55 Angeles Co., Signal Hill, W of 405 Fwy., Temple Ave. at 29th St., UTM (NAD 83) 11S 0392713E 3741666N, elev. 24 m, rare, in pavement cracks and moist soil in roadside gutter, 23 Nov 2003, Riefner 03-472 (RSA); same locality and date, Riefner 03-474 (RSA); City of Pasadena, Arroyo Seco Creek, vicinity of Lower Arroyo Park, UTM (NAD 83) 11S 0392515E 3778822N, elev. 119 m, uncommon but widely scattered along creek in wet sand and in Salix-Baccharis riparian scrub, 22 Aug 2004, Riefner 04-380 (RSA); City of South El Monte, San Gabriel River at Thienes Ave., UTM (NAD 83) 11S 0405107E 3766943N, elev. 74 m, uncommon, wet sand on river bank with Hydrocotyle, Ludwigia, and Scirpus, and on disturbed clay soils along bike trail, 28 Aug 2004, Riefner 04-404 (RSA); near Griffith Park, E bank of Los Angeles River near River Ride Stables, ca. 0.5 mi S of Colorado St., UTM (NAD 83) 11S 0382604E 3777331N, elev. 120m, widely scattered, moist sandy soil in open riparian scrub, 2 Oct 2006, Riefner 06-464 (RSA); San Dimas, Pudding- stone Reservoir, Bonelli Region County Park at East Shore Dr., UTM (NAD 83) 11S 0426247E 3771609N, elev. 290 m, uncommon on receding lake shoreline, 3 Aug 2007, Riefner 07-341 (RSA); City of Malibu, along Malibu Rd. ca. 0.5 mi W of Pacific Coast Hwy., UTM (NAD 83) 11S 0340836E 3767080N, elev. 10 m, local, wet roadside depression, 10 Nov 2007, Riefner 07-498 (RSA). Orange Co., City of San Clemente, San Clemente State Beach, SW corner of park at camp- ground number 156, UTM (NAD 83) 11S 0444133E 3696183N, elev. 37 m, uncommon, disturbed ground in runoff from campground showers, 14 Oct 2003, Riefner 03-457 (RSA); City of Anaheim, Santa Ana River bottom, E of 57 Fwy. between Orangewood St. and Chapman Ave., UTM (NAD 83) IIS 0418311E 3739098N, elev. 41 m, locally abundant, disturbed wetlands, margin of active channels, and on sandbars in riparian scrub, 8 Aug 2004, Riefner 04-372 (RSA); City of Irvine, San Diego Creek at Woodbridge High School, S of intersection of Barranca Rd. and West Yale Loop Rd., UTM (NAD 83) 11S 0425279E 3727007N, elev. 27 m, common, native plant-dominat- ed wetlands and sandy wash habitats, 27 Aug 2004, | Riefner 04-386 (RSA); City of Irvine, San Diego Creek | at 405 Fwy., UTM (NAD 83) 11S 0429935E 3724259N, | elev. 50 m, uncommon, sandy wash and sandbars in | riparian scrub, 27 Aug 2004, Riefner 04-391 (RSA); City | of Fountain Valley, along Newland Ave. at Edinger St., | UTM (NAD 83) 11S E 409083E 3732691N, elev. 14 m, | rare, weed in moist soil along sidewalk, 28 Aug 2004, | Riefner 04-407 (RSA); City of Los Alamitos, Oak | Middle School, vicinity of Oak St. at Catalina St., , UTM (NAD 83) IIS 0400172E 3741242N, elev. 14 m, | common, moist depressions in ball field turf grass, 3 Sep | 2004, Riefner 04-423 (RSA); City of Los Alamitos, W of , Oak Middle School along San Gabriel River, UTM | (NAD 83) 11S 0400172E 3741242N, elev. 10 m, rare, | bank of drainage ditch with Leptochloa, 3 Sep 2004, | Riefner 04-424 (RSA); City of Huntington Beach, | Golden West College, along Golden West St. near Blue Bonnet Ave., UTM (NAD 83) 11S 0406757E | 3733198N, elev. 14 m, locally common, moist soil in- roadside gutter, 12 Sep 2004, Riefner 04-426 (RSA); | City of Irvine, Irvine Valley College, along Jeffrey Rd. | near Irvine Center Dr., UTM (NAD 83) 11S 0427758E | 3726570N, elev. 49m, uncommon, wet ditch along. parking lot and adjacent strawberry field, 12 Dec 2004, Riefner 04-552 (RSA); City of San Clemente, along El | 2008] Camino Real ca. 0.15 mi S of intersection with Los Molinos Rd., UTM (NAD 83) 11S 0441920E 3699403N, elev. 35m, common, moist soil on irrigated slope with ornamental shrubs, 20 Sep 2005, Riefner 05- 671 (RSA); City of Yorba Linda, near Horseshoe Bend along the Santa Ana River, UTM (NAD 83) 11S 0431309E 3749053N, elev. 102 m, widespread but uncommon, banks of urban runoff ditch and wet sand along active channel of Santa Ana River, 5 Oct 2005, Riefner 05-718 (RSA); City of Irvine, San Diego Creek between Alton Pwky. and 405 Fwy., UTM (NAD 83) 11S 0429336E 3724184N, elev. 51 m, uncommon and scattered on moist sandy banks and wet sand along creek channel, 8 Oct 2006, Riefner 06-474 (RSA); City of Los Alamitos, Arroyo Park, Heather St. at Lamson Ave., UTM (NAD 83) 11S 0402697E 3738566N, elev. 10 m, common and widespread, wet ditches and moist depressions in turf grass, 17 Oct 2005, Riefner 05-724 (RSA); City of Huntington Beach, along Slater Ave. near Goldenwest St., UTM (NAD 83) 11S 0406860E 3730121N, elev. 11m, uncommon, moist soil in Eucalyptus woodland and edge of drainage ditch, 4 Oct 2006, Riefner 06-477 (RSA); City of Irvine, Shady Canyon Rd. at Quail Hill, UTM (NAD 83) IIS 0427699E 3724253N, locally common in drainage ditch and irrigated landscapes, elev. 35m, 15 Nov 2006, Riefner 06-677 (RSA). Riverside Co., near Jurupa, intersection of Hellman Ave. at Pine Ave., UTM 11S 0443544E 3757824N (NAD 83), elev. 195 m, locally abundant in roadside drainage ditch, 20 Dec 2003, Riefner 03-548 (RSA); City of Norco, Sixth St. at Sierra Ave., UTM (NAD 83) 11S 0448759E 3755512N, elev. 200 m, locally abundant, weed in moist clay soil along urban equestrian trial, 30 Jul 2004, Riefner 04-282 (RSA). San Bernardino Co., City of Fontana, Merrill St. at Olive St.. UTM (NAD 83) IIS 0459593E 3772444N, elev. 381 m, uncommon in drainage ditch with Cyperus in vacant dirt lot, 5 Oct 2005, Riefner 05- 721 (RSA). San Diego Co., City of Carlsbad, near Laguna Riviera City Park, Kelly Dr. at Park Dr., UTM (NAD 83) 11S 0471025E 3667607N, elev. ca. 13 m, uncommon, drainage ditch and open riparian scrub, 14 Oct 2006, Riefner 06-542 (RSA); City of San Diego, San Diego River Trail, Robb Field at San Diego River, UTM (NAD 83) 11S 0477053E 3624151N, elev. 3 m, locally abundant in moist sandy soil, 13 Sep 2007, Riefner 07-385 (RSA). Ventura Co., City of Santa Paula, near intersection of Hallock Dr. and Lemon- wood Dr., UTM (NAD 83) 11S 0312334E 3803599N, elev. 85 m, rare, growing with Cyperus, Leptochloa, and — Baccharis in urban runoff ditch draining into the Santa Clara River wash, 27 Jun 2004, Riefner 04-215 (RSA); same location, 24 Sep 2004, Riefner 04-439 (RSA); City of Fillmore, Santa Clara River bottom, ca. 0.5 mi W of ~SR-123, UTM (NAD 83) 11S 0323579E 3807088N, _ elev. 133 m, scattered in wet sand along low-flow river channel, 5 Jul 2007, Riefner 07-295 (RSA). Previous knowledge. Eleusine species are predomi- nately African, but E. coracana (L.) Gaertn. subsp. _ africana (Kenn.-O’ Byrne) Hilu & De Wet and E. indica _ (L.) Gaertn. are widely distributed weeds that are easily misidentified (Phillips 1972, Kew Bull. 27: 251-270; Hilu 2003, in Flora of North America Editorial -~Committee, eds., Flora of North America, Vol. 25, Magnoliophyta: Commelinidae (in part): Poaceae, Part _ 2, Oxford University Press, New York, NY). Eleusine indica (goosegrass or wiregrass) is a highly successful _ Cosmopolitan annual that is often a troublesome weed NOTEWORTHY COLLECTIONS ee in many warm temperate and tropical countries (Holm et al. 1977, The World’s Worst Weeds: Distribution and Biology, University Press of Hawaii, Honolulu, HI). It grows in disturbed areas and lawns throughout the California Floristic Province and most of the contigu- ous United States (Smith 1993, in Hickman, ed., The Jepson Manual: Higher Plants of California, University of California Press, Berkeley, CA; Hilu 2003 /oc. cit.). Eleusine coracana subsp. africana (African finger millet, wild finger millet), an annual, is native to East Africa, but is also known from India where it was likely introduced with imported seeds of the cultivated finger millet (E. coracana [L.] Gaertn. subsp. coracana). In the Western Hemisphere, EF. coracana subsp. africana has been accidentally introduced to Coahuila (Columbus 2842; RSA) and Sonora (Columbus 2721, 3627; RSA), Mexico, and to Calhoun County, South Carolina, in the eastern United States (Neves et al. 2005, Molecular Phylogenetics and Evol. 35: 395-419). Eleusine coracana subsp. africana 1s also weedy in its native East Africa, where it frequently occupies disturbed ground of roadsides, villages, and cultivated lands (Phillips 1972 Joc. cit.; Van Wyk and Van Oudtshoorn 1999, Guide to Grasses of Southern Africa, Briza Publications, Arcadia, South Africa). It is thought to be the wild progenitor of the cultivated E. coracana subsp. coracana (finger millet or ragi) that was domesticated in East Africa over 5,000 yr ago and is now grown as a cereal crop in many parts of Africa, Asia, the Arabian Peninsula, China, India, and elsewhere (Phillips 1972 loc. cit.; De Wet et al. 1984, Am. J. Botany 71: 550-557; Brisht and Mukai 2002, Plant Syst. Evol. 233: 243-258; Hilu 2003 Joc. cit.; Neves et al. 2005 Joc. cit.). In North America, E. coracana subsp. coracana is often cultivated at agricul- tural experiment stations and occasionally escapes (Hilu 2003 loc. cit.). In the United States, it has been successfully grown as far north as Davis, California, but with difficulty owing to problems of photoperiod sensitivity (National Research Council 1996, Lost Crops of Africa, Vol. I, Grains, Board on Science and Technology for International Development, National Academy Press, Washington, DC). Smith (1993 Joc. cit.) and Hilu (2003 Joc. cit.), however, did not provide specific localities for adventive populations. In general, there has been little disagreement regard- ing the taxonomy of most Eleusine species and distinctive morphological features easily distinguish most taxa. However, the taxonomic status of E. coracana s.str., E. africana Kenn.-O’Byrne s.str., and E. indica has often been disputed (Phillips 1972 loc. cit.; Neves et al. 2005 Joc. cit.). Eleusine coracana subsp. coracana, a robust annual, is readily distinguished from other taxa in the genus by its large, almost globose seeds and upright habit, which explains the preference of some authors to treat it as a distinct species (Hilu 2003 Joc. cit., Neves et al. 2005 Joc. cit.). However, the large grain size and upright habit are likely the result of agricultural selection produced during domestication. The few genes that control these agriculturally desirable traits and the remaining genome are indistinguishable from its wild progenitor and genetically similar FE. coracana subsp. africana (Neves et al. 2005 Joc. cit.; Dida et al. 2006, Genetic Diversity in Finger Millet [E. coracana] and Related Wild Species, poster presented at the Plant & Animal Genomes XIV Conference, January 14-18, 2006, Town & Country Convention Center, San Diego, CA). Therefore, the rank of subspecies more adequately 312 reflects the ancestral relationship and genetic similarities of the E coracana taxa (Hilu and De Wet 1976, Econ. Bot. 30: 199-208: Neves et al. 2005 Joc. cit.). Morphologically, E. coracana subsp. africana is most similar to E. indica and the two have often been confused, making clear-cut identifications problematic (Phillips 1972 Joc. cit.; Brisht and Mukai 2002 Joc. cit.; Neves et al. 2005 Joc. cit.). However, ongoing taxo- nomic study and recent molecular analysis have provided important insights regarding the separation of the two taxa. The grains of Eleusine species are ornamented and have a unique surface pattern and shape that are useful for identifying and separating E. coracana subsp. africana from E. indica; 1.e., oblong seeds with the grain surface shallowly ridged and uniformly granular in E. coracana subsp. africana and elliptic to ovoid seeds with the grain surface obliquely striate in E. indica (Phillips 1972 loc. cit.; Hilu 2003 Joc. cit.). Eleusine coracana subsp. africana is also a considerably more robust plant and has lower glumes that are 2- or 3-veined, whereas E. indica is much more slender and has lower glumes that are 1-veined (Phillips 1972 loc. cit.; Hilu 2003 Joc. cit.). Dida et al. (2006 Joc. cit.), however, note that the length of the upper glume and broader leaf width of E. coracana subsp. africana are also reliable morphological characters that can be used to separate it from E. indica. In addition to morphological distinctions, molecular analysis has established that FE. indica, a diploid, is genetically isolated from the tetraploid E. coracana subsp. africana, and attempts to produce artificial hybrids between them have resulted in sterile plants (Chennaveeraiah and Hiremath 1974, Euphytica 23: 489-495; Hiremath and Salimath 1992, Theor. Appl. Genet. 84: 747-754). Accordingly, molecular analysis has discounted hybrid- ization as an explanation for the difficulties in separating the two taxa (Neves et al. 2005 Joc. cit.). Significance. First records of E. coracana subsp. africana reported for California. Eleusine coracana subsp. africana has not been included in recent treatments of the Poaceae dealing with exotic plants recently established in California or in local floras covering Orange, Los Angeles, Riverside, San Diego or Ventura counties (Smith 1993 /oc. cit.; Roberts 1998, A Checklist of the Vascular Plants of Orange County, California, 2nd ed., F.M. Roberts Publications, En- cinitas, CA; Hrusa et al. 2002, Madrono 46: 61—98; Hilu 2003 /oc. cit.; Roberts et al. 2004, The Vascular Plants of Western Riverside County, California: An Annotat- ed Checklist, F. M. Roberts Publications, San Luis Rey, CA; Rebman and Simpson 2006, Checklist of the Vascular Plants of San Diego County, 4th ed., San Diego Natural History Museum, San Diego, CA; Clarke et al. 2007, Flora of the Santa Ana River and Environs, Heyday Books, Berkeley, CA; DiTomaso and Healy 2007, Weeds of California and Other Western States, U.C. Agriculture and Natural Resourc- es Publication 3488, Oakland, CA; Riefner and Boyd 2007, J. Bot. Res. Inst. Texas 1: 709-730). In southern California, E. coracana subsp. africana 1s most common in disturbed, moist-soil urban habitats, including irrigated residential and commercial orna- mental landscapes, nuisance runoff and sediment debris flows in street gutters, in culvert outflow basins and drainage ditches, along roadsides and sidewalks, and in turf grass. However, unlike E. indica, which 1s largely restricted to urban environments, E. coracana subsp. africana has become established in native wetland and MADRONO [Vol. 55 riparian habitats and is dispersing rapidly via the numerous urbanized drainage systems located at the wildland-urban interface (WUI) in coastal southern California. New water sources generated by over irrigation of lawns and landscaping, discharge of municipal water treatment plant effluent into natural drainages, increased dry-season stream flows associated with decreased precipitation infiltration and increased hard-surface runoff following storm events, and chang- es in historic hydrologic regimes have facilitated the establishment of exotic plants and have significantly modified WUI riparian and wetland communities in coastal southern California (Burkhart 2006, Fre- montia 34: 14-19; White and Greer 2006, Landscape and Urban Planning 74: 125-138; Riefner and Boyd 2007 loc. cit.). Historically, summer drought and the seasonal nature of stream course flows in the South Coast region have acted as a barrier to colonization by many moisture-dependent exotic plants in the WUI (Brigham 2007, in Knapp, ed., Flora and Ecology of the Santa Monica Mountains, Southern California Botanists Special Publication No. 4, Fullerton, CA). Accordingly, increased moisture availability in an otherwise summer- dry Mediterranean climate has thereby promoted the establishment and spread of E. coracana subsp. africana and several other warm-season moisture-dependant exotic grasses in southern California, including Dinebra retroflexa (Vahl.) Panz var. retroflexa, Ehrharta erecta Lam., Panicum coloratum L., and Setaria adhaerans (Forssk.) Chiov. (Riefner et al. Madrono 50: 312-313; Clarke et al. 2007 Joc. cit.; Riefner and Boyd 2007 Joc. cit.). Although E. coracana subsp. africana is relatively widespread and can often be locally abundant, these new records suggest that detailed taxonomic and floristic studies are needed to identify and thoroughly document the establishment and dispersal of exotic plants in southern California (Riefner and Boyd 2007 loc. cit.). Additional occurrences of E. coracana subsp. africana should be sought in urban settings and native plant communities associated with summer-wet habitats — of sandy stream benches, in open riparian scrub, and — the edge of perennial wetlands at the WUI throughout © southern and central California. —RICHARD E. RIEFNER, JR., Research Associate, — Rancho Santa Ana Botanic Garden, 1500 N. College | Avenue, Claremont, CA 91711. rriefner@earthlink.net; | J. TRAVIS COLUMBUS, Claremont Graduate University, . Rancho Santa Ana Botanic Garden, 1500 N. College Avenue, Claremont, CA 91711; and STEVE BoyD, | Herbarium, Rancho Santa Ana Botanic Garden, 1500 | N. College Avenue, Claremont, CA 91711. CALIFORNIA SENECIO LINEARIFOLIUS A. Richard var. LINEARIFO- LiuS (ASTERACEAE ).—Orange Co., City of Irvine, | University of California campus, just north of Bonita > Canyon Dr. (i.e., directly west of T-intersection of | Shady Canyon Cr. and Bonita Canyon Dr. [=Culver | Dr.]), low-lying seasonal wetland along a mapped | intermittent ‘“‘blueline” stream, USGS Tustin topo- graphic map, ca. 33°38'N, 117°49'W, ca. 70 m elev.; abundant in a solitary patch about 5—10 m across: Scott | D. White 6783, 26 Jun 1998 (BRIT, RSA, UCR). At 2008] least four other collections from the same or nearly the same location: Scott D. White 9613, 7 Jul 2003 (RSA); R. Noll s.n., 20 Aug 2003 (UCR); Mark A. Elvin 3011, 24 Aug 2003 (UCR, CAS [not yet accessioned]); R. No// s.n., 15 Sep 2003 (RSA, UCR). Orange Co., City of Laguna Niguel, roadside and steep slope along Pacific Island Drive, USGS Dana Point topographic map, Township 8S, Range 8W, Section 9; ca. 33°29.8'N, 117°43.3'W, ca. 60-120 m elev. Locally common in roadside seeps, both sides of the road. Scott D. White and Michael Honer 10401, 17 May 2004 (RSA) and Scott D. White 10606, 7 July 2004 (CDA, RSA, UCR). San Diego Co., Chula Vista, canyon east of Highway 805 on East H Street, between Terra Nova Rd. and Rancho del Rey Parkway; ca. 32°39'39"N, 117°2'13"W; riparian, rocky, sandy loam soil, with Baccharis salicifolia, Salix lasiolepis, S. gooddingii and scrub, with Artemisia californica, Encelia californica; Brant Prim- rose 31, 16 Oct 2002 (BRIT, SD). Same location, Brant Primrose 32, 15 Jan 2003 (SD) and Jon P. Rebman and Brant Primrose 8298, 20 Jan 2003 (SD, UMO). San Diego Co., Mission Valley, ca. 32°48'21’"N, 117°6'2"W; near sea level; riparian drainage, with Salix laseolepis, S. goodingii, Baccharis salicifolia, B. saro- throides, Lonicera subspicata, Quercus agrifolia, Populus fremontii, Foeniculum vulgare, Picris echioides, and Phoenix canariensis; Brant Primrose 33, 12 Aug 2004 (SD). Identifying these specimens was a years-long inter- national project. Duplicates of the first California specimens (White 6783 and Primrose 31, above) were sent to Theodore Barkley at BRIT (Fort Worth, Texas) for determination. Barkley did not recognize them and was confident they were new for North America. He suggested that they may represent an undescribed taxon, though he evidently had no opportunity to compare them with Senecios of other continents before becoming ill. Primrose began work on a description of the material and some specimens (above) may be labeled “‘S. serenae ined.” After Barkley’s death in 2004 Guy Nesom (also of BRIT) compared the specimens with Australian material at MO and identified them as S. /inearifolius. That was the first correct determination, but it came too late for full inclusion in the Flora of North America Senecio treatment. Instead, the Editorial Committee added S. linearifolius as a note in the introduction (T. Barkley 2006, vol. 20 pp. 544-570, Flora of North America, Oxford Univ. Press). Leroy Gross (RSA) made the same determination soon after. There are several varieties of S. /inearifolius including S. Jinearifolius var. dangarensis, an endangered narrow-endemic of basaltic soils on Mt. Dangar in New South Wales (I. _ Thompson 2004, Muelleria 20:67—110). The determina- tion of S. Jinearifolius var. linearifolius was made by Randy Bayer at CANB (Canberra, Australia) in May 2006 from Rebman and Primrose 8298 (above, anno- tated specimen at UMO). | Previous knowledge. Senecio linearifolius is native to southeastern mainland Australia and Tasmania “‘within a few hundred km of coastlines in medium to high rainfall areas of woodlands and forests” (Thompson op _cit.). Thompson noted further that “it can be a dominant component of understory vegetation espe- cially on disturbed sites such as road verges”’ and that it NOTEWORTHY COLLECTIONS ee Be, ranges from “‘sea level to alpine altitudes.’ The variety, S. linearifolius var. linearifolius is widespread through- out much of the species’ range. Its common name is ‘““groundsel fireweed.” Senecio linearifolius var. linearifolius is a shrub, ca. 0.6—1.8 m tall, with linear to oblanceolate leaves. There are about 10-50 heads in compact corymbose arrays. The involucres are about 3 mm wide. The ray corollas are about 2—3 mm long; the ray and disc flowers are yellow. The young stems are weakly woody, straight, wand-like, and vertical. It spreads laterally as older stems lay onto the ground and sprout new vertical shoots. The lateral stems may grow to at least 2—3 cm diam., and may not be evident beneath the foliage and closely spaced erect stems or if buried by sediments. Superficially, S. /inearifolius var. linearifolius may resemble the native Euthamia occidentalis. Photographs of pressed specimens in Thompson (2004, op cit.) may be useful for identification. Significance. First records for North America. Similar plants were seen by Fred Roberts (pers. comm.) in 1985 along Oso Creek in San Juan Capistrano, near the Laguna Niguel site (above), but were never identified. This observation suggests that Senecio linearifolius var. linearifolius may have been established in that area without being documented for nearly 20 yr. There also is an unvouchered report from the Sweet- water River in San Diego County (M. Dodero, pers. comm.). In southern California, S. /inearifolius var. linearifolius grows in wetland margins at roadsides, seeps, alkaline flats, and stream channels. Its affinity for disturbed sites, mesic habitats, and its wide elevational range in Australia (Thompson op cit.) suggest that it could become invasive throughout cismontane southern California in riparian or seasonal wetlands, as Arundo donax and Tamarix ramosissina have. Its occur- rence in mesic Australian forests and woodlands (Thompson op cit.) implies that 1t could also invade disturbed upland sites in coastal central California and northward. It already is naturalized in New Zealand (Thompson op cit.) The California occurrences probably escaped from local cultivation. Senecio linearifolius is grown as an ornamental in mainland Australia and Tasmania (davesgarden.com, site ac- cessed 22 Oct 2007) and seed is available by mail order from a French dealer (b-and-t-world-seeds.com, site accessed 21 Oct 2007). It probably is being grown locally in specialty nurseries or traded among garden hobbyists. We are grateful for Ted Barkley’s interest and efforts with this plant. Many thanks to curators, staff and associates of BRIT, CANB, CAS, CDA, RSA, UMO, and UCR, including Randy Bayer, Steve Boyd, Judy Gibson, Leroy Gross, Fred Hrusa, Robin Kennedy, Guy Nesom, Jon Rebman, Andy Sanders, Debra Trock, and Leszek Vincent for their efforts in identifying and piecing together the identifications, history, and whereabouts of the specimens cited here. —ScoTT D. WHITE, Scott White Biological Consult- ing, 201 N. First Ave., No. 201, Upland, CA 91786. scottbioservices@verizon.net; and BRANT PRIMROSE, Jones and Stokes, 9775 Businesspark Ave., San Diego, CA. 92131 bprimrose@Jjsanet.com or bryophytes_2000@ yahoo.com. MADRONO, Vol. 55, No. 4, p. 314, 2008 PRESIDENT’S REPORT FOR VOLUME 55 Dear CBS member, As I write this letter, the world is going through a great deal of economic turmoil. At such times, it 1s comforting to dwell on those values and organizations that have shown their ability to abide over time. The California Botanical Society (CBS), founded in 1914, has pursued its mission of promoting Western American botany and natural history through two world wars and many economic upheavals, including the Great Depression. Of course, it can’t always have been easy for a volunteer run society to push forward against the inevitable obstacles. Nevertheless, due to the unflagging commitment of its members and dedicated volunteers, CBS continues to thrive. This past year, we have been especially privileged to have an excellent and dedicated board. I thank previous editors John Callaway and John Hunter for serving as interim editors for the issues 2 and 3. I also thank CBS board members Andrew Doran, Heather Driscoll, Staci Markos, Abigail Moore, Jim Shevock, Sue Bainbridge, and Mike Vasey for their efforts beyond the call of duty I welcome the new co-editors of Madrono Professor Richard Whitkus of Sonoma State University and Professor Tim Lowrey of University of New Mexico. They will bring to the editorship a long history of collaboration with each other, and a detail-oriented approach. In an age when academics face increasing professional demands on their time, we honor their commitment to publishing high quality studies of Western American botany in Madrojno. Recently, CBS held its annual banquet in collaboration with the California Native Plant Society (CNPS) during the CNPS Conservation Conference in Sacramento, California. This conference had an exciting program that brought together academic, professional, and amateur botanists, as well as naturalists and gardeners. Many members of CBS are also members of CNPS, and both organizations are committed to increasing knowledge of California plant life. The CBS 2010 banquet is planned to take place at San Jose State University in conjunction with the biannual graduate student meeting in February, 2010. This is less than a year away, so if you are a graduate student, or if you advise graduate students, consider planning to participate in this event. Dean G. Kelch December 2008 MADRONO, Vol. 55, No. 4, p. 315, 2008 EDITORS’ REPORT FOR VOLUME 55 We are please to report the publication of this volume of Madrono by the California Botanical Society (CBS) in 2008. This has been an “interesting” year for the journal with editorial duties dispersed among past editors and Board members. However, the spirit of cooperation and dedication to a quality publication has been realized with the current volume. Of the manuscripts reviewed this year, 44 were published in the current volume. The interval between submission to publication has remained approximately one year, although dispersed editorial duties did extend this period for several manuscripts. This volume also has 12 noteworthy collections and 3 book reviews. The journal maintains a solid tradition of publishing articles on the natural history of botanical organisms of western America, Central America and South America (includ- ing 1 new genus, 3 new species, and range revision for 26 species) while also providing ample opportunity for current and emerging issues such as outcomes of climate change and invasive species biology. We look forward to continuing this forum for members of the CBS and others in the coming year. Numerous reviewers have provided their service to the Society and to aid authors in seeing manuscripts reach successful publication. The Society and authors have also benefited from an editorial board, Jon Keeley (Book Reviews Editor), Dieter Wilken (Noteworthy Collections Editor), Margariet Wetherwax (Noteworthy Collections Editor), and Steve Timbrook (Compiler for Annual Index). Additionally, John Hunter and John Callaway oversaw main editorial duties for this volume, with additional input from Andrew Doran, Heather Driscoll, Dean Kelch, Kim Kersh, Staci Markos, James Shevock, Michael Vasey, Annielaurie Seifert and her colleagues at Allen Press, and the Executive Council of the California Botanical Society. All of these individuals deserve our thanks for their contribution to Madrojo. Finally, as the new co-editors, Dr. Tim Lowrey (‘‘corre- sponding” editor) and Dr. Richard Whitkus (“copy”’ editor) have contributed to seeing this volume come to fruition and they look forward to serving the Society and Madrojno. All who have contributed to this volume December 2008 MADRONO, Vol. 55, No. 4, p. 316, 2008 REVIEWERS OF MADRONO MANUSCRIPTS 2008 Scott Abella David Bainbridge Bruce Baldwin Matthew Brooks Matt Busse Lesley DeFalco Laurence Door Colleen Hatfield Becky Kerns Colin Long Duncan Patten Martin Ritchie James Shevock Michael Simpson Lloyd Stark David Wagner MADRONO, Vol. 55, No. 4, pp. 317-319, 2008 INDEX TO VOLUME 55 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 (see Conifer seedling survival) Agavaceae (see Yucca) Alpine flora of White Mts, CA, 202. Apocynaceae (see Asclepias) Arceuthobium distribution in Durango, Mexico, 161. Arctostaphylos: A. patula (see Conifer seedling survival) New taxon: A. ohloneana, 238. Arctotis, notes on two southern African spp. growing in CA, 244. ARGENTINA (see Muhlenbergia) Arizona: Lilaeopsis schaffneriana var. recurva, livestock trampling effects on, 81; vascular flora of lower San Francisco volcanic field, Coconino Co., 1. Noteworthy collection: Dactyloctenium radulans, 88. Arundo donax, role of floods and bulldozers in break-up and dispersal, 216. Asclepiadaceae (see Apocynaceae) Asclepias schaffneri, resurrection of, 69. Asteraceae: Arctotis, notes on two southern African spp. growing in CA, 244; Eriophyllum lanatum, hybrid zone in OR, 269. Noteworthy collection: Senecio linearifolius var. linear- ifolius, from CA, 312. Atriplex amnicola, A. glauca, A. muelleri, A. stipitata, A. undulata, noteworthy collections from CA, 306. Bestia longipes, identification, distribution and family placement, 291. Boraginaceae (see Cryptantha) Brassicaceae (see Lilaeopsis) C, photosynthesis without Krantz anatomy, 143. California: Alpine flora of White Mts, CA, 202; Arctotis, notes on two southern African spp. growing in CA, 244; Arundo donax, role of floods and bulldozers in break-up and dispersal, 216; Bestia longipes, identifi- cation, distribution and family placement, 291; conifer seedling survival under closed-canopy and manzanita patches in the Sierra Nevada, CA, 191; Clarkia calientensis and C. tembloriensis, taxonomic reassess- ment, 297; Cryptantha crinita, habitat and distribution, 76; Cumathamnion sympodophyllum, notes on early collections and distribution, 248; Dissanthelium califor- nicum, rediscovery and status on Santa Catalina Id., CA, 60; Eleocharis macrostachya and Orcuttia tenuis, spatial and temporal investigation, 257; Euphorbia terracina morphological traits and invasive potential in coastal southern CA, 52; invasive plants not in Jepson, 93; late Holocene wetland vegegation and climate, Shasta Co., 15; Oenothera wolfii and O. glazioviana hybridization, 132; Salsola tragus complex, 113; subalpine fellfield community ecology and eco- physiology, Mt. Pinos, CA, 41; vegetation change in central Sierra Nevada, CA, 223; vegetation mapping and ecotone analysis, southern Coast Range, 26; Yucca brevifolia, glucose-6-phospate isomerase variation and genetic structure, 285. New taxa: Arctostaphylos ohloneana, 238; Clarkia tembloriensis subsp. longistyla, CC. tembloriensis subsp. tembloriensis, 301; Eriastrum harwoodii and E. signatum, 82; Rosa gymnocarpa var. serpentine and R. pisocarpa subsp. ahartii, new taxa from CA & OR, 170; Salsola ryanii, 126. Noteworthy collections: Atriplex amnicola, A. glauca, A. muelleri, A. stipitata, A. undulata, 306; Calocedrus decurrens, Carex longii, 89; Eleusine coracana subsp. africana, 310; Rytidosperma caespitosum, 90; Senecio linearifolius var. linearifolius, 312; Viburnum edule, 306. Calocedrus decurrens, noteworthy collection from CA, 89. Campanulaceae (see Jasione) Caprifoliaceae (see Viburnum) Carex lenticularis var. dolia, noteworthy collection from WY, 179; C. longii, noteworthy collection from CA, 89. Castanopsis (see Notholithocarpus) Castilleja levisecta, effects of host plants and implications for reintroduction, 151. Chenopodiaceae: Sa/sola tragus complex in CA, 113. New taxa: Neokochia, new genus from North America, 251; N. americana and N. californica, new combs., 255; Salsola ryanii, new sp. from CA, 126. Noteworthy collections: Atriplex amnicola, A. glauca, A. muelleri, A. stipitata, A. undulata from CA, 306. Chromosome counts: Eriophyllum lanatum and hybrids, 278. Chrysolepis (see Notholithocarpus) Clarkia calientensis and C. tembloriensis, reassessment, 297. Colorado (see Subalpine forest) Compositae (see Asteraceae) Conifer seedling survival under closed-canopy and manzanita patches in the Sierra Nevada, CA, 191. Cruciferae (see Brassicaceae) Cryptantha crinita, habitat and distribution, 76. Cumathamnion sympodophyllum, notes on early collec- tions and distribution, 248. Cupressaceae (see Calocedrus) Cyperaceae (see Carex and Eleocharis) taxonomic Dactyloctenium radulans, noteworthy collection from AZ, 88. Dedication of Vol. 55 to John O. Sawyer, Jr., 320 Delesseriaceae (see Cumathamnion) Dissanthelium californicum, rediscovery and status on Santa Catalina Id., CA, 60. Drought reconstruction (see Paleoecology) Ecotone (see Vegetation mapping) Editor’s report, 315. Eleocharis macrostachya, spatial and temporal investiga- tion, 257. Eleusine coracana subsp. africana, noteworthy collection from CA, 310. Eriastrum: E. sparsiflorum reassessment, 82. New taxa: E. harwoodii and E. signatum, 82. 318 Ericaceae (see Arctostaphylos) Eriophyllum lanatum, hybrid zone in OR, 269. Euphorbia terracina morphological traits and invasive potential in coastal southern CA, 52. Euphorbiaceae (see Euphorbia) Fagaceae (see Notholithocarpus) Festuca viviparoidea ssp. krajinae, noteworthy collection from WY, 179. Fire history (see Paleoecology) Floras: Alpine flora of White Mts, CA, 202; vascular flora of lower San Francisco volcanic field, Coconino CocsAZa. Genetic structure (see Yucca) Genetic swamping (see Oenothera) Gentianaceae (see Gentianella) Gentianella calanchoides, G. ericothamna, G. gilioides, G. herrerae, G. polyantha, noteworthy collections from Peru, 91. Geraniaceae (see Geranium) Geranium lucidum, noteworthy collection from WA, 178. Glucose-6-phospate isomerase (see Yucca) Gramineae (see Poaceae) Hemiparasitic plants (see Castilleja) Hybridization: Eriophyllum lanatum, hybrid zone in OR, 269; lochroma, homoploid hybridization, 280; O6e- nothera wolfii and O. glazioviana hybridization, 132. Invasive plants: Arctotis, notes on two southern African spp. growing in CA, 244; Arundo donax, role of floods and bulldozers in break-up and dispersal, 216; CA catalog of nonnative plants not in Jepson Manual, 93; Euphorbia terracina in coastal southern CA, 52. Iochroma, homoploid hybridization, 280. Jasione montana, noteworthy collection from WA, 178. Juncaceae (see Luzula) Keys: Annual Eriastrum with stamens exterted less than 1/2 corolla lobe length, 87; Arctostaphylos in the central and southern Santa Cruz Mts., 242; Rosa pisocarpa subsp., 174; R. gymnocarpa vars., 176; separation of Bestia from Isothecium, 292. Kochia (see Neokochia) Krantz anatomy (see Orcuttieae) Lembophyllaceae (see Bestia) Lilaeopsis schaffneriana var. recurva, livestock trampling effects on, 81. Lithocarpus (see Notholithocarpus) Luzula campestris, noteworthy collection from WA, 178. Lycurus (see Muhlenbergia) Marine algae (see Cumathamnion) Massenerhebung or mass-elevation effect (see Subalpine fellfield) MEXICO (see Arceuthobium and Asclepias) Modoc Plateau, CA (see Paleoecology) Mojave Desert (see Yucca) Montana (see Ranunculus) Moss (see Bestia) Mt. Pinos, CA (see Subalpine fellfield) Muhlenbergia alopecuroides, new comb. from Argentina, 159. MADRONO [Vol. 55 Neokochia, new genus from North America, 251; N. Americana and N. californica, new comb., 255. Notholithocarpus, new genus from western North Amer- ica, 181. New taxa: N. densiflorus var. densiflorus, N. densiflorus var. echinoides, 188. Oenothera wolfii and O. glazioviana hybridization, 132. Onagraceae (see Clarkia and Oenothera) Orcuttia tenuis, spatial and temporal investigation, 257. Orcuttieae (Poaceae), C4 photosynthesis without Krantz anatomy, 143. Oregon: Eriophyllum lanatum, hybrid zone, 269; Oc6c- nothera wolfii and O. glazioviana hybridization, 132; Rosa gymnocarpa var. serpentine and R. pisocarpa subsp. ahartii, new taxa from CA & OR, 170. Paleoecology: Late Holocene wetland vegegation and climate, Shasta Co.,:CA, 15. Parasitic plants (see Arceuthobium) Pasania (see Notholithocarpus) Paulownia tomentosa, noteworthy collection from WA, 178. PERU (see Gentianella) Picea englemannii (see Subalpine forest) Pinaceae: Conifer seedling survival under closed-canopy and manzanita patches in the Sierra Nevada, CA, 191; subalpine forest, origina of differences at a CO location, 303. Pinyon-juniper woodland (see Floras) Pinus aristata (see Subalpine forest) Pinus lambertiana (see Conifer seedling survival) Poaceae: Arundo donax, role of floods and bulldozers in break-up and dispersal, 216; Dissanthelium californi- cum, rediscovery and status on Santa Catalina Id., CA, 60; Orcuttia tenuis, spatial and temporal investigation, 257; Orcuttieae, C, photosynthesis without Krantz anatomy, 143. New taxon: Muhlenbergia alopecuroides, new comb. | from Argentina, 159. | Noteworthy collections: Dactyloctenium radulans from AZ, 88; Eleusine coracana subsp. africana, collection | from CA, 310; Festuca viviparoidea ssp. krajinae, | from WY, 179; Rytidosperma caespitosum from CA, 90. Polemoniaceae (see Eriastrum) President’s report, 314. Prunus padus, noteworthy collection from WA, 178. Ranunculaceae (see Ranunculus) Ranunculus jovis, noteworthy collection from MT, 91. Reviews: California Native Plants for the Garden by Carol Bornstein, David Fross and Bart O’Brien, 169; Native | Treasures: Gardening with the Plants of California by | Michael Nevin Smith, 180. Rhodophyta (see Cumathamnion) Rocky Mts. (see Subalpine forest) Rosa gymnocarpa var. serpentine and R. pisocarpa subsp. | ahartii, new taxa from CA & OR, 170. Rosaceae: New taxa: Rosa gymnocarpa var. serpentine | and R. pisocarpa subsp. ahartii from CA & OR, 170. | Noteworthy collections: Prunus padus, Sorbus hybrida, | from WA, 178. | Rytidosperma caespitosum, noteworthy collection from CA, 90. | Salsola: S. tragus complex in CA, 113; S. ryanii, new sp. — from CA, 126. | 2008] San Francisco volcanic field, AZ, vacular flora of, 1. Sawyer, John O., Jr., Dedication of Vol. 55 to, 320 Scrophulariaceae (see Castilleja and Paulownia) Senecio linearifolius var. linearifolius, noteworthy collec- tion from CA, 312. Sierra Nevada, CA: Conifer seedling survival under closed-canopy and manzanita patches in the Sierra Nevada, CA, 191; vegetation change, 223. Solanaceae (see Jochroma) Sorbus hybrida, noteworthy collection from WA, 179. Subalpine fellfield community ecology and ecophysiolo- gy, Mt. Pinos, CA, 41. Subalpine forest, origin of differences at a CO location, 303. Vegetation change in central Sierra Nevada, CA, 223. Vegetation mapping and ecotone analysis, southern Coast Range, CA, 26. Vegetation Type Mapping (VIM) project, 223. Venidium (see Arctotis) INDEX TO VOLUME 55 319 Vernal pool plants: Eleocharis macrostachya and Orcuttia tenuis, spatial and temporal investigation, 257; Orcut- tieae (Poaceae), C4, photosynthesis without Krantz anatomy, 143. Viburnum edule, noteworthy collection from CA, 306. Viscaceae (see Arceuthobium) Washington: Noteworthy collections: Geranium lucidum, Jasione montana, Luzula campestris, Paulownia tomentosa, Prunus padus, Sorbus hybrida, 178. White Mts, CA, alpine flora, 202. Wyoming: Noteworthy collections: Carex lenticularis var. dolia, Festuca viviparoidea ssp. krajinae, 179. Yucca brevifolia, glucose-6-phospate isomerase variation and genetic structure, 285. MADRONO, Vol. 55, No. 4, pp. 320-321, 2008 DEDICATION JOHN O. SAWYER, JR. The California Botanical Society dedicates this volume of Madrono to one of California’s most influential plant ecologists, John O. Sawyer, Jr., in recognition of his distinguished accomplishments in research, in teaching, and in conservation. John was born in Chico, California in 1939. As an undergraduate he attended California Polytechnic State University, San Luis Obispo and California State University, Chico (as they are now known), where he earned his bachelor’s degree in 1961. He received his M. S. in ecology (1963) and Ph. D. in plant ecology (1966) from Purdue University. While pursuing his graduate degrees, John was an instructor at Chico, a graduate assistant at Purdue, and a field ecologist for a consulting firm with projects in Costa Rica and Thailand. His work in those two countries resulted in productive collaborations with A. A. Lindsey and A. W. Kichler and two major publications. In 1966, John joined the faculty of Humboldt State University where he remained until his retirement in 2001. John has published more than forty books and papers, with more in preparation. While there were excursions into such diverse topics as the biological formations of the eastern United States, the vegetation life zones of Costa Rica, and even a flora of a region in northern Thailand, John’s passion has been the vegetation of California, particularly of the northwestern part of the state. Having gone into the field with him for almost forty years, I can attest to his boundless enthusiasm and incredible knowledge. You can see this so clearly in his latest book, Northwest California: a Natural History. He is the co-author, along with Todd Keeler-Wolf, of A Manual of California Vegetation. It is recognized as the standard reference on the subject. A new edition is in the final stages of preparation. John has also contributed treatments of the montane and subalpine vegetation of the Klamath Mountains, the forests of northwestern California, and alpine vegetation in Terrestrial Vegeta- tion of California, edited by Michael Barbour, et al. John, as his students would say, has this thing about woody plants, especially conifers. His Trees and Shrubs of California, co-authored by John Stuart, also at Hum- boldt State, has received excellent reviews. Sawyer’s research was recognized by his faculty colleagues when he was selected as Scholar of the Year at Humboldt State University in 1997. Some plant ecologists have little interest in or knowledge of the flora itself. Not so with Sawyer. He has been involved in documenting the range extension of subalpine fir into California; authored a paper on the serpentine flora of the Lassics; authored or co-authored taxonomic innovations in Rhamnus and Frangula, four family treatments in The Jepson Manual of Higher Plants of California, and the treatment of Rhamnaceae that will appear in the Flora of North America North of Mexico. Five editions of our Keys to the Families and Genera and Vascular Plants in Northwestern California have been published; twenty-two editions our checklist of the plants of that same region have appeared. John is also accomplished with the clippers, field press, and plastic bag, the result being about 10,000 collection numbers. While a faculty member at Humboldt State, John taught general ecology and developed upper division courses in plant ecology, arctic and alpine ecology, and a general education class in California natural history. He was also a prime mover in a very popular spring offering, the desert field trip. Students always commented that his courses were both intellectually and physically demand- ing. One of John’s former students, Andrea Pickart, recalls that he was both demonstrating “critical think- ing’ and demanding it of his students long before it became an academic cliché. By physically demanding, I mean that John is not known for leading leisurely strolls across the countryside. Of course you were going to go to the top of that mountain; there might be interesting manzanitas there. The lack of a well-defined trail is a trivial impediment. These days, John’s field trips for the California Native Plant Society and his Jepson Work- shops are very popular, with a number of return participants. John served as the major professor for about 50 graduate students and was a committee member for a number of others. What is remarkable is how many of them remain close friends and colleagues. They now have important faculty, federal and state agency positions, and are successful consultants around the country. They constitute an extended family. “Once a Sawyer stu-— dent...”’ Photo by Michael Kauffmann, October 2008 2008] John has also made significant contributions to the study of our state’s rare and endangered plants and its forests. He was an editor of the second edition of the California Native Plant Society’s Inventory of Rare and Endangered Plants of California. He has held several important positions in CNPS, including the presidency of the society. It was under his leadership that CNPS began its transformation into a truly statewide organization. John remains active in the society, especially with the North Coast Chapter. In 2006, he became a Councilor for the Save the Redwoods League. DEDICATION SrA Had John been involved in drafting this dedication, he would have wanted to acknowledge the encouragement that he has received from Dale Thornburgh, his long- time friend and colleague, and Jane Cole, his wife and intrepid companion in the field. Allow me to do that on his behalf. JAMES P. SMITH, JR. Professor of Botany, Emeritus Department of Biological Sciences Humboldt State University MADRONO A WEST AMERICAN JOURNAL OF BOTANY VOLUME LV 2008 BOARD OF EDITORS Class of: 2008—ELLEN DEAN, University of California, Davis, CA ROBERT E. PRESTON, Jones & Stokes, Sacramento, CA 2009—DONOVAN BAILEY, New Mexico State University, Las Cruces, NM MARK BORCHERT, USFS, Ojai, CA 2010—-FRED HRUSA, California Department of Food and Agriculture, Sacramento, CA RICHARD OLMSTEAD, University of Washington, Seattle, WA 2011—JAMIE KNEITEL, California State University, Sacramento, CA KEVIN RICE, University of California, Davis, CA Editors—JOHN C. CALLAWAY Department of Environmental Science University of San Francisco San Francisco, CA 94117 callaway@usfca.edu AND JOHN HUNTER c/o Center for Plant Diversity University of California Davis, CA 95616 johnhunterca@sbcglobal.net | Published quarterly by the | California Botanical Society, Inc. Life Sciences Building, University of California, Berkeley 94720 Printed by Allen Press, Inc., Lawrence, KS 66044 MADRONO, Vol. 55, No. 4, pp. 1i-1v, 2008 MADRONO VOLUME 55 TABLE OF CONTENTS Alvarado-Cardenas, Leonardo O. (see Fishbein, Mark) Anderson, R. Scott, et al., A late Holocene record of vegetation and climate from a small wetland in Shasta County, California Arguello, Leonel (see DeWoody, Jennifer, et al. ) Baer-Keeley, Melanie, Review of Native Treasures: Gardening with the Plants of California by Michael Nevin Baer-Keeley, Melanie, Review of California Native Plants for the Garden by Carol Bornstein, David Fross and Bart O’Brien : gaia econ ieia ak Gage ate pute oaeaocetees seas Baum, David A. (see Smith, Stacey DeWitt) Boland, John M., The roles of floods and bulldozers in the break-up and dispersal of Arundo donax (giant POCO) ye 8s ore a tc os eon se nee cee ae ot Lent Ne essen acias Mnetoec sete Boyd, Steve (see Riefner, Richard E., J. Travis Columbus and Steve Boyd) Boykin, Laura M., William T. Pockman and Timothy K. Lowrey, Leaf anatomy of Orcuttieae (Poaceae: Chloridoideae): More evidence of C4 photosynthesis without Kranz anatomy Brigham, Christy (see Riordan, Erin C.) Cannon, Charles H. (see Manos, Paul S.) Christie, Kyle, Vascular flora of the lower San Francisco volcanic field, Coconino County, Arizona Chu, Ge-Lin, and S. C. Sanderson, The genus Kochia (Chenopodiaceae) in North America __ : es Clark, Michelle, et al., A spatial and temporal investigation of Eleocharis macrostachya and Orcuttia tenuis Columbus, J. Travis (see Peterson, Paul M.) Columbus, J. Travis (see Riefner, Richard E., J. Travis Columbus and Steve Boyd) Columbus, J. Travis (see Roberts, Fred M.) Curto, Michael (see Steers, Robert J.) Dawson, Todd E. (see Plamboeck, Agneta H.) Dean, Ellen, et al., Catalogue of nonnative vascular plants occurring spontaneously in California beyond those addressed in the Jepson Manual—Part Uo DeWoody, Jennifer, et al. Genetic evidence of hybridization between Oecnothera wolfii (Wolfs evening primrose) and O. glazioviana, a garden escape _____. Elizondo, Martha Gonzalez (see Mathiasen, Robert L., et al. ae Elizondo, M. Socorro Gonzalez (see Mathiasen, Robert L., et al.) Elliott, Brian A., and Samantha S. D. Mackey, Habitat and distribution of Cryptantha crinita Greene (Boraginaceae) Enriquez, I. Lorena Lopez (see Mathiasen, Robert ioe ‘et al. ) Ertter, Barbara (see Dean, Ellen, et al.) Errter, Barbara and Walter H. Lewis, New Rosa (Rosaceae) in California and Oregon Fairbanks, Dean (see Clark, Michelle, et al.) Fishbein, Mark, Veronica Juarez-Jaimes and Leonardo O. Alvardo-Cardenas, Resurrection of Asclepias schaffneri (Apocynaceae, Asclepiadoideae), a rare Mexican milkweed __ Flores, Jorge A. Tena (see Mathiasen, Robert L., et al.) Gaskin, J. F. (see Hrusa, G. F.) Giblin, David (see Legler, Ben) Gibson, Arthur C. (see Rundel, Philip W., Arthur C. Gibson and M. Rasoul Sharifi) Gibson, Arthur C., Philip W. Rundel and M. Rasoul Sharifi, Ecology and ce of a subalpine fellfield community on Mount Pinos, southern California —__ Chae Otte Goforth, Brett, and Richard A. Minnich, Noteworthy collection from California a Goldman, Douglas, Noteworthy collection from California sesteeaeucste Gowen, David, New taxa following a reassessment of Eriastrum sparsiflor um (Polemoniaceae) ~ erate Hanson, Marilyn F. (see Wiens, John F.) Hemingway, Carroll (see Wiens, John F.) Hipkins, Valerie D. (see DeWoody, Jennifer, et al.) Holland, V. L. (see Steers, Robert J.) Howell, Brian E. (see Mathiasen, Robert L., et al.) Hrusa, Fred (see Dean, Ellen, et al.) Hrusa, Fred (see Riefner, Richard E., et al.) Hrusa, G. F., and J. F. Gaskin, The Salsola tragus complex in California (Chenopodiaceae): Characterization and status of Sa/sola australis and the autochthonous allopolyploid Salsola ryanii, sp. nov. Imper, Davis (see DeWoody, Jennifer, et al.) Jass, Renata B. (see Anderson, R. Scott) Jennings, Steven A., A reexamination of the origin of forest differences at a subalpine location in Colorado Juarez-Jaimes, Veronica (see Fishbein, Mark) 93 1324 76 | 113, 303. 2008] TABLE OF CONTENTS Kaye, Thomas N. (see Lawrence, Beth A.) Kelch, Dean G., President’s report for Volume 55 Kennedy, Jeffrey A. (see Thorne, James H.) Knapp, Denise A. (see McCune, Jenny L.) Kolberg, Vannesa J. (see Smith, Stacey DeWitt) Lawrence, Beth A., and Thomas N. Kaye. Direct and indirect effects of host plants: Implications for reintroduction of an endangered hemiparasitic plant (Castilleja levisecta) Legler, Ben, David Giblin and Peter F. Zika, Noteworthy collections from Washington Lewis, Walter H. (see Errter, Barbara) Linstrand, Len, HII, and Julie Kierstead Nelson, Noteworthy collection from California Lis, Richard (see Clark, Michelle, et al.) Leppig, Gordon (see Dean, Ellen) Lepschi, Brendan (see Riefner, Richard E., et al.) Lowrey, Timothy K. (see Boykin, Laura M.) Lyman, Jennifer, and Clayton McCracken, Noteworthy collection from Montana Mackey, Samantha S. D. (see Elliott, Brian A.) Mahoney, Alison M., and Robert J. McKenzie, Notes on two southern African Arctotis species (Arctotideae: Asteraceae) growing in California Malcolm, Jacob W., and William R. Radke, Livestock trampling and Lilaeopsis schaffneriana var. recurva (Brassicaceae) eee Mallinson, David (see Riefner, Richard E., et al.) Manos, Paul S., Charles H. Cannon and Sang-Hun Oh, Phylogenetic relationships and taxonomic status of the paleoendemic Fagaceae of western North America: Recognition of a new genus, Notholithocarpus Massatti, Rob, and Aaron Wells, Noteworthy collections from Wyoming Mathiasen, Robert L., et al. Distribution of dwarf mistletoes (Arceuthobium spp., Viscaceae) in Durango, WHERIGO: io og cee lene cence lace wees ceeeer noedec dadeccnsewcedeceeweneuess McCracken, Clayton (see Lyman, Jennifer) McCune, Jenny L., and Denise A. Knapp, The rediscovery and status of Dissanthelium californicum (Poaceae) on Santa Catalina Island, California McKenzie, Robert J. (see Mahoney, Alison M.) Merritt, Robert (see Wimmer, Amy T. Toulson) Minnich, Richard A. (see Goforth, Brett R.) Mooring, John S., An Eriophyllum lanatum (Asteraceae) hybrid zone in Oregon Morgan, Brian J. (see Thorne, James H.) Nelson, Julie Kierstead (see Linstrand, Len, IIT) Norris, Daniel H. (see Shevock, James R.) North, Macolm (see Plamboeck, Agneta H.) Oh, Sang-Hun (see Manos, Paul S.) Parker, V. Thomas (see Vasey, Michael C.) Peterson, Paul M., and J. Travis Columbus, Muhlenbergia alopecuroides (Poaceae: Muhlenbergiinae), a new combination —_ Saree eee ee Plamboeck, Agneta H., Malcolm North and Todd E. Dawson, Conifer seedling survival under closed- canopy and manzanita patches in the Sierra Nevada Pockman, William T. (see Boykin, Laura M.) Primrose, Brant (see White, Scott D.) Pringle, James S., Noteworthy collections from Peru Radke, William R. (see Malcolm, Jacob W.) Riefner, Richard E., J. Travis Columbus and Steve Boyd, Noteworthy collection from California Riefner, Richard E., et al., Noteworth collections from California Riordan, Erin C., et al., Morphological traits and invasive potential of the alien Euphorbia terracina (Euphorbiaceae) 1n coastal southern California Roberts, Fred M., and J. Travis Columbus, Noteworthy collection from California Rundel, Philip W. (see Gibson, Arthur C., Philip W. Rundel and M. Rasoul Sharifi) Rundel, Philip W. (see Riordan, Erin C.) Rundel, Philip W., Arthur C. Gibson and M. Rasoul Sharifi, The alpine flora of the White Mountains, California eect Sanders, Andrew (see Dean, Ellen) Sanderson, S. C. (see Chu, Ge-Lin) Schierenbeck, Kristina (see Clark, Michelle, et al.) Scott, Jared (see Mathiasen, Robert L., et al.) Sharifi, M. Rasoul (see Gibson, Arthur C., Philip W. Rundel and M. Rasoul Sharifi) Sharifi, M. Rasoul (see Rundel, Philip W., Arthur C. Gibson and M. Rasoul Sharifi) Shaw, A. Jonathan (see Shevock, James R..,) Shevock, James R., Daniel H. Norris and A. Jonathan Shaw, Identification, distribution and family placement of the pleurocarpous moss Bestia longipes Smith, James P., Jr., Dedication of Volume 55 to John O. Sawyer, Jr. Smith, Stacey DeWitt, Vanessa J. Kolberg and David A. Baum, Morphological and cytological evidence for homoploid hybridization in Jochroma (Solanaceae) il 314 [>] 178 306 91 244 81 181 179 161 202 291 320 280 iv MADRONO [Vol. 55 Smith, Susan J. (see Anderson, R. Scott) Spaulding, W. Geoffrey (see Anderson, R. Scott) Steers, Robert J.. Michael Curto and V. L. Holland, Local scale vegetation mapping and ecotone analysis in the southern Coast Range, Caltformia: 22.22.2222. ese eee ne ee Thorne, James H., Brian J. Morgan and Jeffrey A. Kennedy, Vegetation change over sixty years in the central sierra’ Nevada, Catitotnia, SA. Wem ee eee Tiszler, John (see Riordan, Erin C.) Vasek, Frank C., A taxonomic reassessment of Clarkia calientensis and Clarkia tembloriensis Vasey, Michael C., and V. Thomas Parker, A newly discovered species of Arctostaphylos (Ericaceae) from the central California coast Wells, Aaron (see Massatti, Rob) Westfall, Robert D. (see DeWoody, Jennifer, et al.) White, Scott D., and Brant Primrose, Noteworthy collection from California = Wiens, John F., Marilyn F. Hanson and Carroll Hemingway, Noteworthy collection from Arizona Wimmer, Amy T. Toulson, and Robert Merritt, Glucose-6-phosphate isomerase variation and genetic structive: Fuceq:-brevijolia (Ae avacede): oe eng a eee eee Wynne, Michael J., Notes on early collections and the distribution of the red alga Cumathamnion sympodophyllum _______. Zika, Peter F. (see Legler, Ben DATES OF PUBLICATION OF MADRONO, VOLUME 55 Number 1, pages 1—92, published 16 April 2008 Number 2, pages 93-180, published 15 December 2008 Number 3, pages 181—250, published 30 January 2009 Number 4, pages 251-321, published 31 July 2009 26 223 297 238 312 88 285 248 SUBSCRIPTIONS — MEMBERSHIP Membership in the California Botanical Society is open to individuals ($35 per year; family $40 per year; emeritus $27 per year; students $27 per year for a maximum of 7 years). Late fees may be assessed. Members of the Society receive MADRONO free. Institutional subscriptions to MADRONO are available ($70). Membership is based on a calendar year only. Life memberships are $750. Applications for membership (including dues), orders for subscriptions, and renewal payments should be sent to CBS c/o Jepson Herbarium, University of California, Berkeley, CA 94720-2465. Requests and rates for back issues, changes of address, and undelivered copies of Ma- DRONO 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, five 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 abbreviations 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 are encouraged to include the names, addresses, and e-mail addresses of two to four potential reviewers with their submitted manuscript. 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 included 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 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 MADRONO. Joint au- thors 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 manuscript have been used in other papers that are published, in press, submitted, or soon to be submitted elsewhere. SMITHSONIAN INSTITUTION LIBRARIES HINTON T IMUM 3 9088 01498 5 Pee e eee o ere rere eres er eres e eee EEE EEE EOE SESE EEE ETES EEE SEES EES EEE ESTEE HEED EL EL EEE OEEE EL Oe w DEDICATION; JOHN'O. SAWYER, JRe2i0i2.. dG ReGe. ...ssscsccncsccccssesesssesseccccsnconsseeuta TABLE OF CONTENTS FOR VOLUME: $5 /.05265....cccsssscccvossssccccoosssececsoacsseccconesedsceo sail VOLUME 56, NUMBER | JANUARY-MARCH 2009 MADRONO A WEST AMERICAN JOURNAL OF BOTANY “ yw ot uy bose | AN by | CONTENTS EFFECT OF TERRESTRIAL MOLLUSC HERBIVORY ON HOLOCARPHA MACRADENIA (ASTERACEAE) SEEDLINGS IN CALIFORNIA COASTAL PRAIRIE UNDER DIFFERENT CLIPPING REGIMES TD OATS IVES VIG Coots secach us eeueaen cise s chee eee cuat ran uncon ramen ee | A COMPARISON OF THE SHORT-TERM EFFECTS OF TWO FUEL TREATMENTS ON CHAPARRAL COMMUNITIES IN SOUTHWEST OREGON Kendra G. Sikes and Patricia S.. Mutt occ 252g vcceecc0 PEER Gi Rescsvessouvescnee 8 REPRODUCTIVE BIOLOGY OF THE SAN FERNANDO VALLEY SPINEFLOWER, CHORIZANTHE PARRYI VAR. FERNANDINA (POLYGONACEAE) C. Eugene Jones, Frances M. Shropshire, Laura L. Taylor-Taft, Sean E. Walker, Leo C. Song, Jr., Youssef C. Atallah, Robert L. Allen, Darren R. Sandquist, Jim Luttrell and Jack He Burk oe aI LA savvvevvvveveveneees 23 ON THE RELATIONSHIP OF STREPTANTHUS VERNALIS AND STREPTANTHUS BARBIGER (BRASSICACEAE) Richard: O: Donne lca PU gad eo Woe be PES aes Paha sa svc cacnnonnoeanseees 43 To CALIFORNIA WITH JEPSON’S “PHYTO-JOGS” IN 1913 RichGrd-GeBetd emia i). ie Tae ease vv vv ve Uy bnaaas Sie MA ERSGI NY cote van vonnveseees 49 A COMMENT ON THE ZONAL, INTRAZONAL, AND AZONAL CONCEPTS AND SERPENTINE SOILS Earl B. AlexaQndeye dy occa cece MR occcc onc MERCER RUA sos veneassseediceoosens a7 EVIDENCE OF EXTREME ROOT PROLIFERATION IN RESPONSE TO THE PRESENCE OF A NUTRIENT RICH RESOURCE PATCH BY ERIOGONUM PARVIFOLIUM IVE GTN. IVIVEVOVICH x eccacch acne cso Ml Ueto a6 cas fa cated cee: Oikene en ieamerneutenassemeeetceee 58 PLANT INVASIONS: HUMAN PERCEPTION, ECOLOGICAL IMPACTS AND MANAGEMENT. EDITED BY B. TOKARSKA-GUZIK, J. H. BROCK, G. BRUNDU, L. CHILD, C. C. DAEHLER, AND P. PYSEK WVIGHCON IRC INANE lGssanatttaitmahannrau-siuavadoas oasbwustione uae ses esaen sus ventdentingsieeseeneeene 60 THE CALIFORNIA DESERTS: AN ECOLOGICAL REDISCOVERY. BY BRUCE M. PAVLIK NYO eG SOM ase hess ans rier as Ch te ect aay Oh, oes seatsa te aeeete eee 60 CS AISI ORINDA Beas OOM hxc aie teat lei nen ce healt Ole nmmese Ane eta ca net ce lan date ee ee 63 (SGT eB Gaver Meenas ince NT Po hc wens PURI PET A a Fare vO ee ERR SER IGRUR TRARY tenn OP 66 ILO) INST NUN er a tetas tes nar 1. Oir satcinh ire ke ram aa cn tarsuc Nam ae Nes Seine aly aaiennc aco aa nate 67 PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL SOCIETY MabrONO (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 MADRONO, Kim Kersh, Membership Chair, Uni- versity and Jepson Herbarium, University of California, Berkeley, CA 94720-2465. kersh @berkeley.edu. Corresponding Editor—TIMOTHY LOWREY Copy Editor—RICHARD WHITKUS Museum of Southwestern Biology Department of Biology MSC03 2020 Sonoma State University University of New Mexico 1801 E. Cotati Avenue Albuquerque, NM 87131-0001 Rohnert Park, CA 94928-3609 madrono @unm.edu whitkus @sonoma.edu Book Editor—JON E. KEELEY Noteworthy Collections Editors—DIETER WILKEN, MARGRIET WETHERWAX Board of Editors Class of: 2009—DOoNOVAN BAILEY, New Mexico State University, Las Cruces, NM MARK BorcHerrt, USFS, Ojai, CA 2010—FRED Hrusa, California Department of Food and Agriculture, Sacramento, CA RICHARD OLMSTEAD, University of Washington, Seattle, WA 2011—JAMIE KNEITEL, California State University, Sacramento, CA KEVIN Rice, University of California, Davis, CA 2012—GRETCHEN LEBUHN, San Francisco State University, CA ROBERT PATTERSON, San Francisco State University, CA CALIFORNIA BOTANICAL SOCIETY, INC. OFFICERS FOR 2009-2010 President: Dean Kelch, Jepson Herbarium, University of California, Berkeley, CA 94720, dkelch @berkeley.edu First Vice President. Andrew Doran, University and Jepson Herbaria, University of California, Berkeley, CA 94720, andrewdoran @ berkeley.edu Recording Secretary: Nishanta Rajakaruna, Department of Biological Sciences, San Jose State University, One Washington Square, San Jose, CA 95192, Nishanta.Rajakaruna@sjsu.edu Corresponding Secretary: Heather Driscoll, University Herbarium, University of California, Berkeley, CA 94720- 2465, heather.driscoll @ nature. berkeley.edu Treasurer: Thomas Schweich, California Botanical Society, Jepson Herbarium, University of California, Berkley, CA 94720, tomas @ schweich.com The Council of the California Botanical Society comprises the officers listed above plus the immediate Past President, Michael Vasey, Department of Biology, San Francisco State University, 1600 Holloway Avenue, San Francisco, CA 94132, mvasey @sfsu.edu; the Editor of Madrono; three elected Council Members: James Shevock, Department of Botany, California Academy of Sciences, 55 Music Concourse Drive, San Francisco, CA 94118, jshevock@ nature.berkeley.edu; Roy Buck, Jepson Herbarium, University of California, Berkeley, CA 94720, roybuck@ email.msn.com; Chelsea Specht, Department of Plant and Microbial Biology, University of California, Berkeley, CA 94720-2465, cdspecht@nature.berkeley.edu. Graduate Student Representatives: Ben Carter, Department of Integrative Biology and University Herbarium, University of California, Berkeley, CA 94720-2465, bcarter@ berkeley.edu. Webmaster: Susan Bainbridge, Jepson Herbarium, University of California, Berkeley, CA 94720- 2465, sjbainbridge @ berkeley.edu. This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). MADRONO, Vol. 56, No. 1, pp. 1—7, 2009 EFFECT OF TERRESTRIAL MOLLUSC HERBIVORY ON HOLOCARPHA MACRADENIA (ASTERACEAE) SEEDLINGS IN CALIFORNIA COASTAL PRAIRIE UNDER DIFFERENT CLIPPING REGIMES DOMINIC M. MAZE Department of Botany and Plant Pathology, Oregon State University, Corvallis, OR 97331-2911 dom_maze@yahoo.com ABSTRACT Herbivory is generally overlooked as a factor influencing the viability of endangered plant populations in western North American grasslands. I conducted experiments to determine the effects of terrestrial mollusc herbivory for survival of a federally endangered, annual forb, Holocarpha macradenia (Santa Cruz tarplant), in California’s coastal prairie. In addition, I utilized litter removal and clipping regimes from a continuing study of site-specific responses to evaluate the relationship between herbivory rates from terrestrial molluscs and disturbance regimes. I planted seedlings of H. macradenia into copper bordered quadrats that excluded molluscs and into non-copper bordered controls at two sites in California’s coastal prairie. Three treatment regimes of vegetation clipping and litter removal were utilized. Herbivory damage was assessed using an ordinal scale every two weeks throughout the rainy season. Molluscs were also trapped and identified. Herbivory amounts were significantly less when molluscs were excluded from plots for all disturbance regimes; however, there was no. significant difference in herbivory levels between the disturbance regimes. A significant amount of invertebrate herbivory on H. macradenia was by the non-native slug, Deroceras reticulatum (grey garden slug), and this herbivory significantly affected the survivability of H. maradenia seedlings in California coastal prairie. Key Words: Coastal prairie, grasslands, herbivory, Holocarpha macradenia, Santa Cruz tarplant, terrestrial mollusc. Many recent studies concerning California’s coastal prairies have focused on management strategies for maintaining populations of native plants (Corbin et al. 2004). Strategies for maintaining native plant populations include prescribed fire (Belsky 1992; Davidson and Kindscher 1999; Hatch et al. 1999), soil distur- bance (Belsky 1992) mowing (Davidson and Kindscher 1999; Hayes and Holl 2003a, b), and limited grazing (Belsky 1992; Menke 1992; Davidson and Kindscher 1999; Hatch et al. 1999; Hayes and Holl 2003a, b). Primary concern about conservation of native grasslands has focused on the influence of habitat destruction (Hayes 1998) and competition from exotic grasses (Hayes and Holl 2003b; Seabloom et al. 2003), with little regard to the effects of nonnative, mollusc herbivores on native plants. The exper- iment reported here examines the effects of herbivory by non-native, terrestrial molluscs on seedling survival of Holocarpha macradenia Greene, a federally listed, endangered, annual forb. Anecdotal observations have indicated that a large component of herbivory of H. macradenia is at the seedling stage and by molluscs (G. Hayes, pers. obs.). With H. macradenia populations requiring management and with most remaining populations experiencing decline or interannual variability in numbers, my study evaluates the effect of mollusc herbivory on the viability of H. macradenia seedlings, a component not examined in previous studies. Temperate grasslands have a high abundance of terrestrial molluscs (Crawley 1983; Hulme 1996). In California’s coastal prairie, D. reticula- tum Miller, a nonnative species from Europe, is particularly abundant and widespread (B. Roth, American Malacological Society, personal com- munication). Where endemic, studies examining the effects of herbivory by D. reticulatum have been performed (Rodriguez and Brown 1998; Scheidel and Bruelheide 1999a, b; Hanley et al. 1995a, b; Frank 1997; Hatto and Harper 1969). For many species of plants, these studies found that mollusc herbivory is an important factor affecting survivorship and abundance. The amount of damage from herbivory to seedlings in grasslands may be more important than other forces affecting seedling survival (Dirzo and Harper 1980), with mollusc herbivory often comprising a significant amount of this damage (Wilby and Brown 2001; Hitchmough 2003). In addition, the accumulation of plant litter can be a major factor 1n species competition (Hayes and Holl 2003a, b). The accumulation of plant litter in grasslands may limit seedling survival by not allowing favorable conditions for growth (Milchunas et al. 1995), increasing susceptibility to fungal pathogens by extending 2 MADRONO the duration of free water on leaf surfaces for fungal spore adherence and germination (Bradley et al. 2003), and also by providing a moist, protected habitat for terrestrial molluscs (Maze, pers. obs; Hitchmough 2003). The relatively recent introduction of D. reticulatum, which probably took place when the Spanish began grazing cattle in California (B. Roth, American Malacological Society, personal communication), may be affecting species composition in Califor- nia’s coastal grasslands as it affect species’ composition in its native range (Cottam 1986; Edwards and Gillman 1987; Hanley et al. 1995a). Because mollusc prefer plants at the seedling stage (Duthoit 1964; Byers and Bierlein 1982; Hanley et al. 1995b), have preference-based feeding (Dirzo 1980; Hanley et al. 1995b), and can be a major factor affecting plant population size and viability (Hanley et al. 1996; Scheidel and Bruelheide 1999b; Fenner et al. 1999; Fritz et al. 2001), this herbivory may have implications for H. macradenia survivorship and persistence of its now fragmented and threatened populations. Competition is likely intense in California’s coastal prairie (Stromberg et al. 2002), and any competitive advantage may have major implica- tions for a species’ population viability (Hulme 1996). This competition, coupled with herbivory, can have significant effects on the survivability of plants (van der Wal et al. 2000). These factors, in addition to evidence that slug herbivory is a potent force affecting community structure and plant performance (Hill and Silvertown 1997; Scheidel and Bruelheide 1999a, b, 2001; Fritz et al. 2001; Buschmann et al. 2005), raise the question of to what extent does slug herbivory, affect H. macradenia survivability? Therefore, my research goals were to answer the following questions: 1) What is the extent and the effect of mollusc herbivory on H. macradenia seedlings? 2) Which mollusc species are herbi- vores of H. macradenia and are they native or introduced species? 3) Do grassland management strategies employing the clipping and removal of vegetation influence mollusc herbivory? MATERIALS AND METHODS Study Species Holocarpha macradenia is a federally endan- gered, glandular, annual plant of California’s coastal prairie, with a growth habit that ranges from prostrate and branching to erect and monopodial. Small plants can produce a single flower head, while larger plants have a rigid main stem and lateral branches that grow to the height of the main stem (1—5 dm) and produce many flower heads. The leaves are larger and linear at the base of the plant (up to 12 cm), and are reduced up the stem. Holocarpha macradenia [Vol. 56 produces more central disk flowers than any related species (California Department of Fish and Game 2002). Holocarpha macradenia is known from grass- lands below 150 m in elevation. Historically, it occurred from northern Monterey County, north to Marin County. Currently known Santa Cruz tarplant populations are frequently associated with non-native grasses (e.g., in the genera Avena, Bromus, Hordeum, Briza, and Vulpia). Native associates include species of related Asteraceae, and of Juncus and Danthonia (CDFG 2002). Current H. macradenia populations are in flux and survival of the species 1s dependent on management strategies employed (Hayes 1998), although many management efforts to increase its numbers have met with limited success. With only 13 natural populations, effective manage- ment strategies for both protecting the remaining populations and for facilitating the reintroduc- tion of H. macradenia are needed. Experimental Design I conducted research between December 2003 and April 2004, from the first substantial rains to the end of the rainy season, at two coastal prairie sites within 4 km of the ocean, and less than 150 m in elevation. The sites were separated by 25 km, north to south. I conducted my experi- ment in 52 m X 52 m enclosures that have been undergoing clipping and vegetation removal manipulations since 1998 (Hayes and Holl 2003a, b). Study sites were enclosed with fencing in the fall of 1998 to keep out cattle and feral pigs. Within these enclosures, Hayes and Holl randomly allocated 30, 7 m X 7 m plots with ten treatments. For my study, I utilized three of the ten treatments (described below). The two sites were: “Elkhorn”, 36°52'4.3"N, 121°44'23.8"W (near the Elkhorn Slough reserve in Monterey Co.) and “UCSC”, 36°59'5.5"N, 122°3'0.9"W (on the University of California at Santa Cruz campus). The Elkhorn site is about 200 m from one of the largest known natural populations of H. macradenia while the UCSC site has no known historical population. At both sites, I utilized plots with three treatments of clipping and removal of cut vegetation: clipped 2 times a year, clipped 6 times a year, and untreated “‘control” plots (Hayes and Holl 2003b). Treatments were accomplished with | a weed-whip and removal of cut vegetation with a | rake after each clipping. Control plots were not | manipulated. Clipping and raking regimes greatly | reduced both vegetation cover and biomass, | although these were not measured during my | study. For each clipping regime at each site I placeda | treated quadrat, bordered with copper flashing, and a control quadrat delineated with string. All | 2009] were | m?* and contained 16 4-week old seedlings, evenly dispersed within the plot. Each plant was 2326 cm from any neighboring seedling or the plot edge. I placed my quadrats in the southeast- erly corners of the Hayes and Holl 7 < 7 m plots, 1 m from the corners. Each of the quadrats for this study was assigned separate plots from the Hayes and Holl study. To ensure accurate herbivory measurements, clipping and raking were suspend- ed for each quadrat for the duration of my study. Whereas previous experiments used the broad application of molluscicides as an abundance- and density-reducing tool (Hanley et al. 1995a; Rees and Brown 1992; Frank 1997), both the sensitive nature of these grasslands and the possible presence of unknown, native molluscs warranted another approach. I used 8.5 cm copper flashing firmly adhered to the soil layer using galvanized 16-penny nails to create a barrier around these quadrats. Terrestrial mol- luscs cannot cross copper (Grewal et al. 2003) and the flashing employed was manufactured for this purpose (U.S. Patent # 4,471,562). In addition to the copper flashing, slug traps were placed in each southeasterly corner of the treated (copper-bordered) quadrats. The slug traps were a pitfall type consisting of a buried 12 once aluminum can a third filled with a Pilsner type beer. The traps further decreased the density and abundance of existing and emerging molluscs within the manipulated plots and provided specimens for identification. Slug traps were originally also placed outside of the fenced enclosures to further examine the mollusc species present; however, these traps were unsustainable due to damage from cattle. Experimental Procedure Seed of AH. macradenia was collected from stock plants and sown on November 18, 2003 at the University of California Santa Cruz, Thimann greenhouse. Seedlings were moved outside of the greenhouse 2 wk after sowing. The seedlings were outplanted on December 18th and 19th into the quadrats, as the local rains were becoming sub- stantial. Measurements of the extent of herbivory on H. macradenia seedlings were taken every two weeks from January 3, 2004 to April 6, 2004, with each site being measured within 2 d of each other. Herbivory damage was assessed visually on an ordinal scale of 0 to 4, corresponding to 4 = 100%, 3 = 75%, 2 = 50%, 1 = 25%, 0 = 0% of the plant damaged. Mollusc herbivory was easily identified: signs included characteristic feces, mucus ‘“‘trails’’, a hand lens revealed rasped margins of damage indicating radular-type herbivory, and often, especially after precipitation events, I observed slugs actively consuming the seedlings. I measured vegetation height in each plot by dropping a paper plate onto the vegetation and MAZE: MOLLUSC HERBIVORY OF HOLOCARPHA MACRADENIA 3 measuring from the plate’s center to the soil surface (Davis and Sherman 1992). I averaged vegetation depth and litter for all plots within each clipping regime at the start of the study: clipping 2 times a year, 5.75 cm; clipping 6 times a year, 4.0 cm; “‘controls”’, 16.3 cm. In “‘control” plots, vegetation litter was so thick from the five years of the Hayes and Holl study that no change in the height of the vegetation was observed during the duration of the study. Statistics Measurements were compiled for each plot by taking the mean levels of herbivory for all seedlings for each date, respective of manipula- tion or control for each clipping regime. SAS version 8.01 (SAS Institute, Cary, NC) was employed for the univariate repeated measures ANOVA with sampling date, slug exclusion, and clipping regimes as the independent variables. A separate univariate repeated measures ANOVA was also employed to test for differences in survivability of H. macradenia with slug exclusion as the independent variable. Seedling mortality was examined by computing nonparametric estimates of survivor functions, stratified by site, after recording the fate of each seedling. For this analysis of survivability, SAS version 9.1.3 (SAS Institute, Cary, NC) was employed with LIFETEST procedures, which allowed for seedlings that experienced mortality due to non-mollusc factors or that were alive at the end of the study, to be right-censored. For these survivor functions, a log-rank test for homogeneity between treatments was performed. RESULTS Both the Elkhorn and UCSC sites showed significantly less herbivore damage in the copper- bordered plots as opposed to the control plots throughout the study (F,.7 = 8.44; P = 0.0198, Fig. 1), with the rate of herbivory increasing over time as the study progressed (F549 = 3.42; P = 0.0283). Differences in herbivory among clipping re- gimes were not significant with regards to levels of herbivory (Fs ,,; = 1.79; P = 0.2285). Analysis of seedling mortality indicated that terrestrial molluscs have significant effects on the survivorship of H. macradenia, (log-rank, Chi- square statistic = 53.31, P < 0.0001, Fig. 2). Analysis by the same statistical method but without stratification by site resulted in similar values (log-rank, Chi-square statistic = 55.15, P = 0.0001). At the end of the study (107 d), the survival estimate for control seedlings at both sites was 13.5%, while the mollusc-excluded seedlings experienced 55.8% survivorship. The probability of survival decreased at a greater rate 4 MADRONO 100 Percent Herbivory [Vol. 56 m Cu slug exclusions OO Controls 80 70 60 50 40 30 20 10 14 30 46 70 89 107 Day Sampled Fic. 1. Mean percent herbivory for all H. macradenia seedlings with regard to presence or absence of slug exclusions and without regard to clipping regimes. Error bars indicate standard error. in control plots than in copper-bordered plots at both sites (Fig. 2), with estimated mean survival times reflecting this trend difference (Table 1). DISCUSSION The hypothesis that terrestrial molluscs signif- icantly affect H. macradenia seedlings via herbiv- Probability of Survival Days BiGe2: homogeneity, stratified by location. ory is strongly supported by the results of this experiment. Exclusion and trapping of molluscs led to decreased rates of herbivory over the duration of the experiment with one notable exception, which was one copper-bordered plot at UCSC that underwent twice-yearly clipping. Initial rates of herbivory in this plot were statistically identical to that of the untreated —R Copper —— Control Survival probability curves for mollusc-excluded (copper) and control seedlings. From a log-rank test for 2009] TABLE 1, MEAN SURVIVAL TIME ESTIMATES (IN DAYS) AND STANDARD ERRORS FOR H. MAZE: MOLLUSC HERBIVORY OF HOLOCARPHA MACRADENIA 5 MACRADENIA SEEDLINGS AT BOTH SITES FOR MOLLUSC-EXCLUDED (COPPER) AND CONTROL SEEDLINGS. From a log-rank test for homogeneity. UCSC control 76.8 3.03 Mean survival time (days) Standard error controls due to a breach in the copper stripping of about 13 cm, which led to a large initial standard error for the combined means of the manipulated plots. The breach was due to cows which entered the fenced enclosure and damaged the copper barrier sometime between herbivory assessments. The copper barrier was breached for no longer than 13 d. The fact that the herbivory rates in the breached plot were so similar to those in the corresponding control plots is a testament to both the effectiveness of the copper (in light of the herbivory rates in the non-breached plots) and to the tenacity of the slugs themselves. After the repair, seedlings slowly recovered while the control seedlings of the corresponding plot were effectively decimated. Seedlings performed differently at the two sites. The Elkhorn seedlings appeared more vigorous than their counterparts at the UCSC site, although statistical analyses showed that differ- ences in herbivory and mortality were not significant. The Elkhorn site is next to a natural population of H. macradenia from which the seed for this experiment was collected. Two weeks after the last herbivory assessments were made most of the remaining UCSC seedlings had died, although seemingly not from invertebrate herbiv- ory. Seedlings may have died because the site is not suitable for H. macradenia. Not all coastal grasslands are suitable habitat for H. macradenia, suggesting other factors such as soil type and microclimate should be considered for any reintroduction efforts. Six of the 134 molluscs trapped over the duration of the study were not D. reticulatum. These were all one species of native snail and all were collected at the Elkhorn site in February. The snails were members of the family Succinei- dae, with the genus being either Succinea or Catinella which is associated with seasonally wet conditions (B. Roth, American Malacological Society, pers. comm.). These native snails are probably less effective herbivores than D. reticu- latum based on the small number trapped and their diminutive nature (0.5 cm in diameter) relative to D. reticulatum (up to 8 cm in length in this experiment). The impact of mollusc-induced mortality was perhaps the most striking result of this study. The survivorship curves illustrate the significant effect that mollusc herbivory had on H. macradenia in this study (Fig. 2). The fact that almost all Elkhorn control 88.6 2.08 UCSC copper 99.0 2.14 Elkhorn copper 101.0 1.98 herbivory was mollusc induced before day 70, and that one introduced mollusc was responsible for most (if not all) of this herbivory, suggests that D. reticulatum might be adversely affecting annual forb persistence in not only coastal but also interior grasslands (Severns 2006). Mollusc herbivory continued throughout the duration of the study with increased emergence of arthropod herbivores resulting in similar survival proba- bility slopes between treatments after day 70 (Fig. 2); however, this was not statistically analyzed. Mollusc herbivory was fairly constant with varying vegetation depth and clipping regimes until the last data set, where herbivory in the heavier thatched, non-clipped plots increased at UCSC (data not shown). This is perhaps due to the increasingly warmer and drier habitat as the rainy season ended. The fact that herbivory did not differ among clipping and vegetation removal regimes or controls earlier in the season might be due to the small size of the quadrats, unknown distance of movements of D. reticulatum in California coastal prairie, or ability of D. reticulatum to tolerate different amounts of cover and vegetation. This does not necessarily imply that in an unmanipulated setting, differences in disturbance history would not be important. Disturbance could affect the surrounding vege- tation, which could then affect factors including germination time, seed mortality and predation, and seed dispersal. As for movements of D. reticulatum, in European rapeseed crops complete losses occurred within 2m of slug habitat, although another species of slug was also present (Frank 1997). Given that nonnative slug herbivory induced low survival for H. macradenia seedlings and that previous studies such as Bevill et al. (1999) illustrated that short-term protection of juvenile rare plants from invertebrate herbivores can improve the probability of long-term persistence of populations, future management of H. macra- denia should examine possible ways to limit slug herbivory. Obviously, the cost, labor, and main- tenance of copper flashing make this method highly impractical. More feasible methods in- clude 5% metaldehyde pellets (a molluscicide), which is undesirable because of adverse impacts on native molluscs in the remaining remnants of coastal prairie in California. It was observed that a few (4 total) seedlings in heavily grazed control 6 MADRONO plots were never or only slightly grazed over the duration of the experiment, suggesting there might be a phenotype that resists mollusc herbivory as shown with Asarum caudatum (wild ginger) (Cates 1975). Results of this study suggest that D. reticulatum herbivory should be consid- ered when devising management or reintroduc- tion strategies for H. macradenia. ACKNOWLEDGMENTS This research was made possible by Dr. Karen Holl, my advisor at the University of California, Santa Cruz, and Dr. Grey Hayes for his suggestions to start paying attention to H. macradenia and herbivory by slugs, his novel approaches to slug exclusion, and for his help and revisions. Also, thanks to Kim Hayes of the Elkhorn Slough Foundation for permission to conduct this study at the Foundation’s Porter Ranch site and to the helpful comments of three anonymous reviewers, Dr. Patricia Muir and Dr. John C. Hunter. Finally, thank you to Dr. Barry Roth of the California Malacozoo- logical Society for kindly taking his valuable time to identify molluscs and for providing much needed historical information regarding D. reticulatum. LITERATURE CITED BELSKy, A. J. 1992. Effects of grazing, competition, disturbance and fire on species composition and diversity in grassland communities. Journal of Vegetation Science 3:187—200. BEVILL, R. L., S. M. LOUDA, AND L. M. STANFORTH. 1999. Protection from natural enemies in managing rare plant species. Conservation Biology 13:1323— 1331, BRADLEY, D. J., G. S. GILBERT, AND I. M. PARKER. 2003. Susceptibility of clover species to fungal infection: the interaction of leaf surface traits and environment. American Journal of Botany 90:857—864. BUSCHMANN, H., M. KELLER, N. PORRET, H. DIETZ, AND P. J. EDWARDS. 2005. The effect of slug grazing on vegetation development and plant species diversity in an experimental grassland. Functional Ecology 19:291—298. BYERS, R. A. AND D. L. BIERLEIN. 1982. Feeding preferences of three slug species in the laboratory. Melscheimer Entomological Series 32:5—11. CALIFORNIA DEPARTMENT OF FISH AND GAME (CDFG). 2002. Santa Cruz Tarplant. California Department of Fish and Game, http://www. elkhornsloughctp.org/uploads/1108408439CDF Greport.pdf (accessed April 3, 2009). CATES, R. G. 1975. The interface between slugs and wild ginger: some evolutionary aspects. Ecology 56:391—400. CoRBIN, J. D., C. M. D’ANTONIO, AND S. J. BAIN- BRIDGE. 2004. Tipping the balance in the restora- tion of native plants: experimental approaches to changing the exotic:native ratio in California grassland. Pp. 154-179 in M. Gordon and L. Bartol (eds.), Experimental approaches to conser- vation biology. University of California Press, Berkeley, CA. CoTTAM, D. A. 1986. The effects of slug grazing on Trifolium repens and Dactylis glomerata in mono- culture and mixed swards. Oikos 47:275—279. [Vol. 56 CRAWLEY, M. J. 1983. Herbivory: the dynamics of animal plant interactions. Blackwell Scientific Publications, Oxford, U.K. DAVIDSON, C. AND K. KINDSCHER. 1999. Tools for diversity: fire, grazing and mowing tallgrass prairie. Ecological Restoration 17:36. DAvis, L. H. AND R. J. SHERMAN. 1992. Ecological study of the rare Chorizanthe valida (Polygonaceae) at Point Reyes National Seashore, California. Madrono 39:271—280. DirRzo, R. 1980. Experimental studies on plant-slug interactions I. The acceptability of thirty plant species to the slug Agrilimax caruanae. Journal of Ecology 68:981—998. AND J. L. HARPER. 1980. Experimental studies on slug-plant interactions II. The effect of grazing by slugs on high-density monocultures of Capsella bursa-pastoris and Poa annua. Journal of Ecology 68:999-1011. DuTnHolIT, C. M. G. 1964. Slugs and food preferences. Plant Pathology 13:73-78. EDWARDS, P. J. AND M. P. GILLMAN. 1987. Herbivores and plant succession. Pp. 295-314 in A. J. Gray, M. J. Crawley, and P. J. Edwards (eds.), Coloni- sation, succession and stability. Blackwell Scientific Publications, Oxford, U.K. FENNER, M., M. HANLEY, AND R. LAWRENCE. 1999. Comparison of seedling and adult palatability in annual and perennial plants. Functional Ecology 13:546-551. FRANK, T. 1997. Slug damage and the number of slug pests, Arion lusitanicus and Deroceras reticulatum, in oilseed rape grown beside sown wildflower strips. Agriculture, Ecosystems and Environment 67:67-78. FRITZ, R. S., C. HOCHWENDER, D. A. LEWKIEWICZ, S. BOTHWELL, AND C. M. ORIANS. 2001. Seedling herbivory by slugs in a willow hybrid system: developmental changes in damage, chemical de- fense, and plant performance. Oecolgia 129:87—97. GREWAL, S. K., P. S. GREWAL, AND R. B. HAMMOND. 2003. Susceptibility of North American native and non-native slugs (Mollusca: Gastopoda) to Phas- marhabtitis hermaphrodita (Nematoda: Rhabditi- dae). Biocontrol Science and Technology 13:119- | 125. | HANLEY, M., M. FENNER, AND P. EDWARDS. 1995a. | An experimental field-study of the effects of | mollusc grazing on seedling recruitment and | survival in grasslands. Journal of Ecology 83:621— 627. : , AND 1995b. The effect of | molluse grazing on seedling recruitment in artifi- cially created grassland gaps. Oecoloica 106:240— | 246, ' , AND . 1996. Molluse grazing | and seedling survivorship of four common grass- | land plant species: the role of gap size, species and season. Acta Oecologica 17:331—341. | Hatcu, D. A., J. W. BARTOLOME, J. S. FEHMI, AND | D. S. HILLYARD. 1999. Effects of burning and | grazing on a coastal California grassland. Resto-_ ration Ecology 7:376—381. | HATTO, J. AND J. L. HARPER. 1969. The control of slugs and snails in British cropping systems, especially grassland. International Copper Re- | search Association Project 15:1—25. 2009] HAYES, G. F. 1998. The Saga of the Santa Cruz Tarplant. The Four Seasons 10:18—21. AND K. D. HOLL. 2003a. Cattle grazing impacts on annual forbs and vegetation composition of mesic grasslands in California. Conservation Biol- ogy 17:1694—1702. AND . 2003b. Site-specific responses of native and exotic species to disturbances in a mesic grassland community. Applied Vegetation Science 6:235-244. HILL, B. AND J. SILVERTOWN. 1997. Higher-order interaction between molluscs and sheep affecting seedling numbers in grassland. Acta Oecologica 18:587—S96. HITCHMOUGH, J. D. 2003. Effects of sward height, gap size, and slug grazing on emergence and establish- ment of Trollius europaeus (Globeflower). Restora- tion Ecology 11:20-28. HULME, P. E. 1996. Herbivores and the performance of grassland plants: a comparison of arthropod, mollusc, and rodent herbivory. Journal of Ecology 84:43-51. MENKE, J. W. 1992. Grazing and fire management for native perennial grass restoration in California grasslands. Fremontia 20:2. MILCHUNAS, D. G., A. S. VARNAMKHASTI, W. K. LAUENROTH, AND H. GOETZ. 1995. Forage quality in relation to long-term grazing history, current-year defoliation, and water. Oecologica 101:366—374. REES, M. AND V. BROWN. 1992. Interactions between invertebrate herbivores and plant competition. Journal of Ecology 80:353—360. RODRIGUEZ, M. AND V. BROWN. 1998. Plant compe- tition and slug herbivory: effects on the yield and MAZE: MOLLUSC HERBIVORY OF HOLOCARPHA MACRADENIA q biomass allocation pattern of Poa annua L. Acta Oecologica 19:37-46. SCHEIDEL, U. AND H. BRUELHEIDE. 1999a. Selective slug grazing on montane meadow plants. Journal of Ecology 87:828-838. AND . 1999b. Selective slug grazing on montane meadow plants. Journal of Ecology 87:839-848. AND . 2001. Altitudinal differences in herbivory on montane compositae species. Oeco- logia 129:75—86. SEABLOOM, E. W., E. T. BORER, V. L. BOUCHER, R. S. BURTON, K. L. COTTINGHAM, L. GOLDWASSER, W. K. GRAM, B. E. KENDALL, AND F. MICHELI. 2003. Competition, seed limitation, disturbance, and reestablishment of California native annual forbs. Ecological Applications 13:575—592. SEVERNS, P. M. 2006. Seasonality, habitat preference and life history of some Willamette Valley wet prairie terrestrial molluscs in western Oregon, USA. The Veliger 48:220—227. STROMBERG, M. R., P. KEPHART, AND V. YADON. 2002. Composition, invasability, and diversity in coastal California grasslands. Madrono 48:236—252. VAN DER WAL, R., M. EGAS, A. VAN DEL VEEN, AND J. BAKER. 2000. Effects of resource competition and herbivory on plant performance along a natural productivity gradient. Journal of Ecology 88:303—306. WILBy, A. AND V. BROWN. 2001. Herbivory, litter and soil disturbance as determinants of vegetation. Oecolgia 127:259-26. MADRONO, Vol. 56, No. 1, pp. 8-22, 2009 A COMPARISON OF THE SHORT-TERM EFFECTS OF TWO FUEL TREATMENTS ON CHAPARRAL COMMUNITIES IN SOUTHWEST OREGON KENDRA G. SIKES! AND PATRICIA S. MUIR Department of Botany and Plant Pathology, Cordley Hall 2082, Oregon State University, Corvallis, OR 97331-2902 ABSTRACT Fuel treatments to reduce fire danger are being applied to public lands throughout the western United States, affecting significant acreage at considerable expense. This study compares the short term effects of two fuel treatment methods, shrub mastication and “hand cut, pile, and burn” (HPB), on chaparral communities in southwestern Oregon. Ceanothus cuneatus dominated the study sites where permanent paired plots were established on either side of treatment-control boundaries. Over a two year period, we recorded all vascular plant species within each treatment or control plot, along with an abundance class for each species. The effects of treatment on species composition and abundance of forbs and graminoids, overall as well as by plant trait group, were surprisingly small. Time since treatment, | yr or 2 yr, had a stronger effect on species composition than did treatment method. Species abundance and richness were greater in the first year after treatments than in the second year or in controls. In the second year, after both types of treatments, species abundance and richness were reduced, while after mastication treatment, these measures were lower than in control areas. The HPB treatment had a greater effect on plant communities than did mastication, due, at least in part, to the presence of fire rings from burned piles. Compared to their surrounding treated plots, fire rings had greater proportions of both annuals (95% versus 71%) and introduced weeds (35% versus 21%) in the second year after treatment. Ceanothus germination was stimulated in fire rings but also occurred in most plots, including controls. Both types of treated areas had more Ceanothus seedlings than their controls. Short term evidence suggests that the HPB treatment may lead to an increase in weedy and exotic species and the mastication treatment may reduce species diversity. The HPB treatment may also increase native species diversity by allowing fire-cued species to establish. Additional monitoring over time is needed to assess longer term treatment consequences for these northern chaparral communities. Key Words: Burn piles, Ceanothus cuneatus, chaparral, fire management, fuel reduction, shrub mastication. Fuel treatments have become an increasingly important aspect of public lands management in the western United States. Decades of fire suppression have increased fuel loads in some forests and shrublands, potentially increasing spread and severity of fires when they occur. Recent large-scale wildfires and the expansion of housing in proximity to wildlands have resulted in loss of property and increased motivation to prevent further losses (Keeley 2002). The primary purpose of fuel treatment is to remove, reduce, or alter combustible plant material that can act as fuel for a fire, thus reducing wildfire risks. Other potential benefits include rejuvenating senescent shrubs, increasing forage for wildlife (Lillywhite 1977; Rogers et al. 2004), and improving conditions for forbs and grasses. Risks associated with fuel treatments include invasion or expan- sion of exotic plant species (e.g., Merriam et al. 2006; Perchemlides et al. 2008) and loss or reduction of native species caused by depleted ‘Present address: California Native Plant Society, 2707 K Street, Suite 1, Sacramento, CA 95816. E-mail: sikesk@lifetime.oregonstate.edu. or treatment-damaged seed banks, mycorrhizal communities, or resprouting success (e.g., Smith et al. 2004; Busse et al. 2005; LeFer and Parker 2005; Knapp et al. 2007). These potentially negative effects depend on the ecosystem being treated, fuel loads, historical fire regimes, recent — fire-free intervals, and method of treatment (e.g., Keeley and Fotheringham 2001la; Veblen 2003; Knapp et al. 2007). Mechanical methods of fuel treatment have become more common because of risks associat- ed with prescribed fire and the effects of smoke , on air quality in populated areas. Mechanical | treatments are being used in fire-adapted plant communities, such as chaparral, though conse- quences for the ecosystems are not well under- stood (Knapp et al. 2007). Chaparral most commonly burns in stand-replacing fires that do not leave records such as fire scars, resulting in uncertainty about past fire regimes (Keeley and Fotheringham 2001b). Estimates of a fire return interval for chaparral in the pre-industrial United | States range from 20-40 yr (Leenhouts 1998) to | 50-80 yr (Zedler 1995). A range of regeneration | strategies coexists in this diverse plant communi- | ty, making chaparral resilient to fire and other | 2009] disturbances (Lavorel 1999; Keeley et al. 2005b). While some chaparral plant species are long-lived (Keeley et al. 2005a), many have relatively short life spans, and are dependent on fire for reproduction or the creation of sites for estab- lishment. Replacing fire with mechanical fuel treatment is likely to affect the regeneration of some species and the community in general. Chaparral is found in regions with the hot dry summers and cool wet winters of a Mediterra- nean climate, and is often dominated by a single shrub species with small, evergreen leaves. Ceanothus cuneatus (Hook.) Nutt. (Rhamnaceae) dominates the chaparral of central and northern California and Oregon’s Rogue Valley, “the northernmost outpost of typical North American chaparral” (Detling 1961, p. 354). Chaparral is the most xeric vegetation type in southwestern Oregon, with C. cuneatus being the most drought and heat tolerant of the area’s woody dominants (Detling 1961). In the Rogue Valley, Ceanothus chaparral is usually found adjacent to the more mesic habitat of Quercus garryana Dougl. ex Hook. (Fagaceae) oak woodlands, while Arcto- staphylos viscida Parry (Ericaceae) is found overlapping the two adjacent zones (Detling 1961). Ceanothus stands are often interspersed with grass openings. The Medford District of the Bureau of Land Management (BLM) in southwestern Oregon manages a checkerboard of public lands inter- mixed with private, inhabited lands. In recent years, the BLM has allotted substantial funds to treating fuels in both chaparral and oak wood- lands using two methods: mastication and hand cut, pile, and burn (HPB). Mastication uses heavy machinery on caterpillar treads that is equipped with a rotating brush cutting disk on a moveable arm. The machine can shred woody material up to 30 cm in diameter. The resulting chips and coarsely shredded debris are left on site. In the HPB treatment, fuel removal is accom- plished using chainsaws and the cut material is piled for later burning. Burn piles are generally about 3 m in diameter and up to 2 m high; a few sheets of plastic are incorporated into them to keep their centers relatively dry for burning during the rainy season. Both treatment methods open the canopy, increasing light availability on the ground, but neither removes the litter layer or approximates the soil surface conditions found after fire (Kauffman 2004). Mastication does not remove fuels, but alters their structure, leaving woody debris on the soil surface. Little woody debris remains after HPB treatment. Small patches of the HPB treatment area are subjected to high intensity fire from burned brush piles. There are concerns about the potential conse- quences of the management techniques being used on the plant communities in southwestern SIKES AND MUIR: EFFECTS OF FUEL TREATMENTS ON NORTHERN CHAPARRAL ? Oregon. While fuel management has been applied to thousands of hectares in this region since the mid-1990’s (USDI 1999), effects of treatments on fuel conditions, subsequent fire behavior, or plant or animal communities have received little attention (but see Perchemlides et al. 2008). The overall objective of this study was to determine the short term effect of fuel treatments on plant communities in which Ceanothus cuneatus is the dominant shrub. Specifically, we addressed the following questions. Do fuel treatments enhance or reduce opportunities for native plant estab- lishment? Do they increase the spread of non- native weeds? Do the two types of fuel treatments affect the plant community differently? How does the presence of fire rings from burned piles affect species composition? We analyzed differences in plant community response and Ceanothus regen- eration one and two years after both types of treatments. METHODS Study Area This research was conducted on public lands in the Butte Falls Resource Area (42°32’'N, 122°37'W) of the Medford District Bureau of Land Management (BLM). Study areas were located within a 10 km radius in the foothills of the Cascade Range in the Rogue River water- shed, northeast of Medford, Jackson County, in southwestern Oregon (Fig. 1). The Rogue Valley is bounded by the Siskiyou Mountains to the west and south, and by the Cascade Mountains to the east. The valley floor, up to about 750 m in elevation, has been categorized as the Interior Valley Zone which includes oak woodlands, coniferous forests, grassland, and chaparral (Franklin and Dyrness 1973). At higher eleva- tion, mixed conifer forest begins to dominate (Franklin and Dyrness 1973). Chaparral occurs in the driest areas up to 1100 m (Detling 1961). Elevation at our study sites ranges from 500 to 920 m. Precipitation varies with elevation, aver- aging 47 cm per yr at 395 m in Medford, and 92 cm in Butte Falls at 762 m (NOAA 2004; Fig. 1). Only 20 percent of the annual precipita- tion falls between April and September (Johnson 1993). The normal average July temperature 1s 20.7°C at 482 m (Lost Creek Dam), and 22.6°C at Medford. The average January temperature is 3.2°C at 482 m, and is similar throughout the elevation range of our study sites (NOAA 2004). Study sites have moderate (4%~—31%) slopes, predominantly south-facing. Soils are Carney clay, McMullin-Rock outcrop complex, and Medco-McMullin complex (Johnson 1993). Both the Carney clay and Medco soils have very slow permeability. The Carney soil has a high clay content throughout and therefore can cause areas 10 MADRONO OCEAN S = a @) , 3 GOLD BEAGK., O LL O < QO Fic. 1. [Vol. 56 | 43 ‘_ : | | | | | | | | * BUTTE, FALLS ! ! 25 Kilometers Map of Jackson County, Oregon (in white), including major highway routes, significant population centers, and the Rogue River. Plot pairs are shown as dots within the grey oval in the vicinity of Butte Falls. of standing water. The clay subsoil of the Medco soil keeps the water table high in winter and limits the effective rooting depth. McMullin soils are moderately permeable, but shallow, support- ing mainly shrubs, grasses and forbs (Johnson 1993). Ceanothus cuneatus was the dominant woody plant at all of the study sites. Quercus garryana (Oregon white oak) was common, with some oak canopy occurring in about half of the plots sampled. Herbaceous vegetation was dominated by annuals, especially non-native annual grasses. All sites have a history of livestock grazing. No evidence of recent fire or clearing was discovered from a search of aerial photos on file at the BLM (taken approximately once each decade since 1966). Treatment Prescriptions In both HPB and mastication treatments, approximately 75% of the shrub cover was removed (M. Wineteer, BLM, personal commu- nication). Prescriptions varied slightly by man- agement unit, but, in general, all shrubs under or within 3 m of tree crowns were removed. Shrubs in the open were thinned to produce a space between clumps of approximately 7.6 m (Fig. 2). Large healthy trees (1.e., conifers >18 cm diam. at breast height (DBH) and hardwoods >25 cm DBH) were not removed. Smaller trees were retained with a 7.6 m spacing. Management units were contiguous areas with a single treatment prescription carried out by a single entity. Units varied in size from 1.2 to 76 ha. Treatment conditions often varied from place to place within a unit because treatments could take days or months to be completed. All burning of hand piles for the HPB treatment occurred in November or December of 2002 (2 yrs before sampling) or 2003 (1 yr before sampling). Cutting and piling, however, began as early as November of the previous year. Mastication treatments were completed in April, May, or June of 2002 and in November or | December of 2003. Field Methods Plot establishment. Pairs of permanent plots (n | = 26 plots) were established in summer 2003 near | the boundaries of fuels reduction units slated to be treated the following fall or winter. We > selected areas where the plant community and | environment appeared similar on both sides of | the boundary. Plot pairs were randomly located on either side of such boundaries. Random — locations were rejected and redetermined if they | resulted in conditions that clearly differed in | community or aspect within a pair. Each plot was 50 m X 1 m (50 m?’) and was at least 15 m from | the marked treatment boundary. Nine of the © pairs established pre-treatment were subsequently | treated, and re-sampled in 2004. One of the nine 2009] Mts FIG. 2. pairs, an HPB treatment and control, was only partially treated, having been cut and piled, while the piles remained unburned. Additional plots were established in 2004 near the boundaries of fuel management units that had already been treated. Some were in areas that had been treated in 2003 at similar times as the pre- treatment plots set up the previous year. These nine retrospective plots were added to compen- sate for pairs established in 2003 whose scheduled treatments had not occurred. One plot was matched with an existing control, resulting in five additional pairs sampled 1 yr since treat- ment. Other plots (n = 25) established in 2004 were in areas that had been treated in 2002. In locating retrospective plots, we reduced the required distance between plot and boundary to 7.5m since we no longer needed to allow for potential differences between the boundary flagged for treatment and the actual treatment boundary. Species abundance data reported here were collected between May 5 and July 28, 2004 in 52 plots, including six pairs for each treatment type and year and four additional unpaired plots that SIKES AND MUIR: EFFECTS OF FUEL TREATMENTS ON NORTHERN CHAPARRAL | Photos of a management area before and after mastication treatment. Pre-treatment photo (top) was taken in July 2003, post-treatment (bottom) in May 2004. were sampled | yr post-treatment. Pair members were sampled on or near the same day. Plots were distributed across 13 land survey sections (each 2.59 sq km in area) and 21 management units. Some units were adjacent to each other, so that some pairs in different units were closer together than pairs in the same unit. The distances between pairs ranged from approximately 100 m to 20 km (Fig. 1), with the average being 870 m. Sampling method. Physical data recorded for each 50 m? plot included cover estimates for bare soil (including gravel-sized rock), rock, litter, woody debris, and standing dead C. cuneatus shrubs (rooted in plot and >30 cm in height), with the following cover classes: 0 = none; 1 = < 10% cover; 2 = 10-25% cover; 3 = 25—50% cover; 4 = > 50% cover. We recorded numbers of Ceanothus in each of four life stages: seedlings (plants that germinated that year with little or no woody tissue); immatures (plants with some woodiness =30 cm tall); living stumps (plants that had been cut but still had green foliage); and matures (uncut 12 mature plants >30 cm tall). Living stumps did not show vigorous resprouting as observed in other woody species; it was unclear whether these plants would survive. The abundance of each vascular plant taxon rooted in the plot was recorded. We used the following broad abundance classes, adapted from the Forest Health Monitoring protocol for lichens (USDA/FS 2002): 0 = Absent (not present in plot), 1 = Rare (1-3 individuals in the plot), 2 = Uncommon (4-10 individuals present), 3 = Common (more than 10 individuals but less than 50% cover), 4 = Abundant (greater than 50% cover). Most taxa were recorded as species; however, some that were not reliably distinguished in the field were combined with other species in their genus. Samples of unknown or uncertain species were collected for later identification or verification. See Sikes (2005) for a full species list and a list of vouchers accessioned by OSC (Sikes 45A-232). Because the species composition of fire rings resulting from burned piles appeared different from the rest of an HPB treated plot, and because the proportion of the plot covered by fire rings was small, we adopted an additional sampling procedure to capture their species composition. A circular plot with an area of 0.25 m° was placed at the visually estimated center of each of three fire rings that most overlapped the HPB treated plot, or were closest to it. Species abundance within the circular plot was recorded using the same abundance codes given above. Percent cover of rock, litter, and woody debris was also estimated using the substrate cover codes de- scribed above. Forty-two fire rings were sampled in June or July 2004, three for each of 14 HPB treatment plots. All had been burned in the months of November or December, half of them in 2003, the other half in 2002. Data Analysis Multivariate analyses were performed using PC-ORD for Windows, Version 4.25 (McCune and Mefford 1999). Other statistical tests were performed using S-PLUS 6.2 for Windows. Data included abundance codes for all taxa in four categories of post-treatment plots, distinguished by treatment type and time since treatment, and their paired controls. Number of taxa present in each plot was included as a plot attribute. Before analyses, we deleted species that occurred in fewer than four plots, to reduce heterogeneity in the data in order that the more rare species did not overly influence our results and to strengthen overall species composition patterns. We also deleted all woody plants from the primary dataset since we were interested in the treatment effects on other components of the plant communities. D MADRONO [Vol. 56 We used blocked multi-response permutation procedures with Euclidean distances (McCune and Grace 2002) to compare species composition between treatment and control groups. Multi- response permutation procedure (MRPP) is a nonparametric test for multivariate difference between two or more pre-defined groups (Zim- merman et al. 1985). The test provides a P-value based on randomized group reassignments and A-values that represent chance-corrected within- group agreement. A = 1 indicates complete agreement where all items are identical within groups, and A = O when heterogeneity within groups is the same as expected by chance (McCune and Grace 2002). Blocking by pair focused the analysis on differences within the pair, while accounting for variation among pairs. We tested for differences between treatment and controls in all matched pairs, HPB treatments (including one partially treated pair), mastication treatments, plots treated the year prior to sampling, plots treated 2 yr prior to sampling, and the four subgroups made up of each treatment type by each treatment year. We also created a data matrix to represent the treatment effect on each pair. For each matched pair, species abundance values for the control were subtracted from those for the treatment. The full set of 229 taxa, including rare and woody species, was used. We used this difference matrix to calculate a matrix of changes in species traits for assessing treatment effects on life forms (forb, graminoid, shrub, tree, annual, or perennial), geographic origin (native or non-native), and weediness of community constituents. Traits were determined from Hickman (1993) and the Plants database (USDA, NRCS 2004). Weedy plants (characteristic of disturbed habitats) were also categorized as either native or introduced weeds. Plants accorded special status by the Medford District BLM (M. Wineteer, personal communi- cation) were also indicated. The change in species trait by pair matrix was created by multiplying the difference matrix described above by a matrix of traits by species. Within-pair differences in each trait were tested for differences from zero using one-sample t-tests. Two-sample t-tests tested whether changes in traits differed between treatment types and between time-since-treatment groups. We also | tested for changes in species richness and canopy | cover between these groups. T-tests were chosen © because a normal distribution was approximated; — however, P-values should be treated with caution | because multiple comparisons were made and > geographic proximity between some pairs prob- | ably compromised their independence. To compare how many species of each trait | had changed in treated versus control plots, a second change of species trait by pair matrix was © created by converting each change in species | 2009] TABLE 1. TESTS FOR DIFFERENCES IN COMMUNITY COMPOSITION SIKES AND MUIR: EFFECTS OF FUEL TREATMENTS ON NORTHERN CHAPARRAL 13 BETWEEN FUEL-TREATED PLOTS AND MATCHED CONTROLS, MRPP BLOCKED BY PAIR (WOODY SPECIES DELETED). The larger the A-value the less likely the similarity within the pre-defined groups can be expected by chance. Within-group agreement Group Number of plots (A) Probability (P) All matched pairs 48 0.007 0.024 HPB treatment 24 0.012 0.036 Mastication treatment 24 0.004 0.234 1 yr since treatment 24 0.023 0.002 2 yr since treatment 24 0.005 O12 1 yr since HPB treatment 12 0.029 0.038 2 yr since HPB treatment [2 0.009 0.254 l yr since mastication 12 0.029 0.038 2 yr since mastication Be 0.002 0.428 abundance by pair in the difference matrix to | or —l (increase or decrease) such that the subse- quent multiplication resulted in a matrix contain- ing the number of species changed rather than the quantitative changes in abundance values. We then used the full 229 taxa dataset to make a presence-absence matrix for control plots and calculate the number of species per trait in these plots. Percent of species changed by trait was calculated using the average number of species differing between treatment and control for a given trait divided by the average number of species with that trait in the control plots. We also tested differences in treatment effects on species composition between the following groups using MRPP: plots established before treatment in 2003 versus post-treatment in 2004, time since treatment (1 yr or 2), and mastication versus HPB. These analyses used the matrix of treatment — control differences in species abun- dances, from which we deleted all woody taxa, resulting in 208 taxa for 25 pairs of plots. The relatively even distribution of data in this matrix made it unnecessary to delete rare species to reduce skewness or coefficients of variation. The matrix contained negative numbers, so MRPPs were based on Euclidean distances (McCune and Grace 2002). Effect of treatment on Ceanothus. We tested for differences in abundances of Ceanothus life stages collectively between the two treatments and between years since treatment using MRPP on treated plots only, and for differences in abundance of each individual stage between treatments and controls using a two-sample t- test. We compared within-pair differences in abundance of each Ceanothus stage (treated plots minus paired controls) between years since treatment and between treatment types using two-sample t-tests. Fire rings of HPB treatment. We tested for differences in the species composition of fire rings between the first year and the second year after burning, and between the two ages of HPB treated plots in their entirety using MRPP (Sorenson distance; McCune and Grace 2002). All species were retained to better represent species richness. We also compared the average proportions of species by trait found in the two ages of fire rings and their associated treatment plots. We used Indicator Species Analysis (ISA; Dufréne and Legendre 1997) to look at species differences between the two ages of fire rings, and the two ages of HPB plots. ISA produces indicator values that combine relative abundance and relative frequency of each species by group to represent faithfulness and exclusivity to that group. A Monte Carlo test of 1000 randomiza- tions (McCune and Grace 2002) was used to test the significance of species indicator values. RESULTS Pair members established pre-treatment in 2003 did not differ in species composition before treatment (blocked MRPP: P = 0.175; Sikes 2005). In 2004, after treatments had occurred, the average species richness per plot (a/pha diversity) was 71.8, and beta diversity (the total number of species in all plots/a/pha) was 3.19. Species richness was greater by an average of 4.2 species in treatment plots than in control plots (P = 0.039, one-sample t-test on differences between matched treatment and control plots in 2004; 95% confidence interval from 0.2 to 8.2 species). In addition, treated plots had greater mean species abundance compared to control plots, averaging an increase in abundance for 5.6 species (P = 0.017, one sample t-test; 95% confidence interval from 1.1 to 10.2 species). There was significantly more cover of woody debris in both types of treatment plots than in control plots (P = 0.001, one-sample t-test on treatment minus control); the mean cover class for control plots was 0.96 (almost all with <10% cover woody debris) and 1.36 for treated plots (with about one third recorded as 10-25% cover of woody debris). No other abiotic plot attributes differed significantly after treatment. 14 MADRONO [Vol. 56 MHPB 1YR @HPB 2YR Number of Species Trait Group FIG. 3. Change in species traits for “‘hand cut, pile, and burn” (HPB) treatment and control pairs | and 2 yr since treatment. Bars show average number of species that increased or decreased in abundance between treatment and control plots by trait, with standard errors. See Table 2 for means across years and the percentage change by trait group. Treatment Effect on Species Composition After treatments were applied, herbaceous plant communities differed (P < 0.05) between treatments and controls when all treatments were compared to all controls (blocked MRPP on 48 plots; Table 1). The species composition of the HPB-treated plots also differed from their controls, whereas the mastication treatment plots and controls did not differ (Table 1). This difference in results between treatment types indicates that the HPB treatment, in the short term, had a greater effect on species composition than did mastication. In all cases, however, the effect sizes were modest, as judged by small measures of within-group agreement (A). When treatments and controls were grouped by time since treatment, communities in 1-yr- since-treatment plots differed from their controls, while the 2-yr-since-treatment plots did not (Table 1). Within each treatment type, commu- nities also differed between treatment and con- trols | yr after treatment, while neither differed 2 yr post-treatment (Table 1). These results suggest that treatment effects on community composition were greatest immediately after treatment. Differences in species abundance between matched treatment and control plots in 2004 provided a measure of community composition change presumably due to treatment. Pairs established in 2003 versus 2004 did not differ in overall species composition changes (MRPP on matrix of treatment-control differences: A = 0.002, P = 0.206), indicating that results were not affected by whether plot establishment occurred before or after treatment. Species composition changes also did not differ between pairs grouped by treatment type (MRPP: A = 0.002, P = 0.223). There were, however, signif- icant differences in composition changes between pairs sampled | yr versus 2 yr post-treatment (MRPP: A = 0.008, P = 0.015). The HPB treatment resulted in larger increases in species richness and abundance than did the mastication treatment (Figs. 3, 4; Table 2). Spe- cies abundance in all trait groups except trees was greater in HPB treated plots across years than in untreated plots (Fig. 3, Table 2). Compared to mastication, the HPB treatment caused signifi- cantly greater increases in abundance of native weeds and shrubs (P = 0.018, df = 23, two- sample t-test). An increase in shrub abundance after treatment that removed shrubs was possible due to seedling recruitment post-treatment. Both treatments tended to favor annuals more than perennials. Differences between responses of non- natives and natives were not as pronounced. When differences were separated by time since treatment, the | yr since treatment pairs generally 2009] SIKES AND MUIR: EFFECTS OF FUEL TREATMENTS ON NORTHERN CHAPARRAL 15 G Mastication 1YR @Mastication 2YR g gin —iddCY : if fc 7 mn eee ee E oe acne £ > | = We ee ee S 0 % S x) S % 2 PSP SS SE EE SES wo ra S © & ~ wr? w? 3? oe NS Trait Group FIG. 4. Change in species traits for mastication treatment and control pairs 1 and 2 yr since treatment. Bars show average number of species that increased or decreased in abundance between treatment and control plots by trait, with standard errors. See Table 2 for means across years and the percentage change by trait group. showed greater increases in species abundance than did the 2 yr since treatment pairs (Figs. 3, 4; Table 2). More graminoids, trees, perennials, and natives had increased abundance in treated relative to untreated plots | yr after treatment than did those groups in plots sampled 2 yr post- treatment (P < 0.05 by two-sample t-test). Of these trait groups, all but graminoids were less TABLE 2. abundant in treated than in untreated plots 2 yr after treatment. While species richness increased in the first year after fuel treatments, there were fewer species in treated than in control plots after 2 yr. For first year plots, the average species richness was 74.1 species in treated plots and 62.5 species in the controls. In second year plots, the CHANGES IN NUMBERS OF SPECIES PER TRAIT GROUP BY TREATMENT TYPE, ACROSS YEARS SINCE TREATMENT, EXPRESSED AS THE MEAN NUMBER OF SPECIES THAT INCREASED OR DECREASED (TREATMENT MINUS CONTROL; AS SHOWN IN FIGs. 3, 4). The percentage change in the trait group is calculated as the mean change in species divided by the average number of species for that trait in the control plots. Hand, pile, and burn Mastication Species Percent change in trait Species Percent change in trait Trait group differed group differed group Forbs 5.5 10.9 13 25 Graminoids 3.5 22.5 1.6 Led Shrubs 0.6 23.8 =1.0 —32.4 Trees —0.4 mck = =25.6 Annuals 6.8 14.7 ot =| Perennials Le 9.9 0.0 0.0 Non-natives 2.0 12.6 0.3 2.1 Natives He) 13.6 1.3 2.6 Weedy plants 4.5 17.6 0.3 0.9 Native weeds 20D 258 —0.4 = 344 Non-native weeds 1.8 12.4 0.6 4.0 Special status 0.0 0.0 0.4 100.0 16 MADRONO TABLE 3. [Vol. 56 INDICATOR VALUES (IV) FOR SPECIES IN PERCENTAGE OF PERFECT INDICATION FOR SECOND YEAR FIRE RINGS. Perfect indication (100%) occurs when the species is always present in that group and never occurs in other groups. Only statistically significant indicator species are listed (P < 0.05). Boldface indicates non- native species. Annuals, not noted as weedy IV Agoseris heterophylla (Nutt.) Greene (Asteraceae) Cardamine oligosperma Nutt. (Brassicaceae) Clarkia purpurea ssp. quadrivulnera (Dougl. ex Lindl.) H.F. & M.E. Lewis (Onagraceae) Cryptantha torreyana (Gray) Greene (Boraginaceae) Madia exigua (Sm.) Gray (Asteraceae) Linanthus bicolor (Nutt.) Greene (Polemoniaceae) Madia Molina spp. (Asteraceae) Phlox gracilis (Hook.) Greene (Polemoniaceae) average was 73.6 species per plot in treated plots and 76.0 species for controls. This difference in species richness between treatment and control of the time-since-treatment groups was statistically significant (P = 0.004, df = 23, two-sample t- test). The mean change in species richness was 10.5 species greater in the first year than in the second (95% confidence interval from 3.7 to 17.3 species). When the year-since-treatment groups were further divided into their treatment groups, we found that species richness was more reduced in the second year after mastication than in the corresponding year after HPB treatment. In fact, 2 yr after treatment, the HPB treated plots supported more species than their control plots. The same trends are visible in the number of species per trait that changed in abundance (Figs. 3, 4; Table 2). Effect of Treatment on Ceanothus When all treatments were compared to all controls, seedlings were significantly more abun- dant in treatments (P = 0.024, two-sample t-test). However, the abundance of immature Ceanothus did not differ between treatments and controls (P = 0.19, two-sample t-test across treatment types and years). Comparing within-pair differences, uncut mature Ceanothus were substantially less abundant after treatment (see Fig. 2), as was cover of standing dead Ceanothus (both P < 0.001, one-sample t-test). The combined life stages of Ceanothus were less abundant after treatment (P = 0.030, one-sample t-test). Abundances of Ceanothus life stages in treated plots differed between years since treatment (MRPP: A = 0.068, P = 0.004), but not between the treatment types (MRPP: A = 0.021, P = 0.129). The abundance of the species across Weedy annuals IV Aira caryophyllea L. (Poaceae) 48.9 Bromus japonicus Thunb. ex Murr. 38.1 (Poaceae) Bromus tectorum L. (Poaceae) 33.3 Cerastium glomeratum Thuill. 33.3 (Caryophyllaceae) Galium L. spp. (Rubiaceae) 76.2 Epilobium L. spp. (Onagraceae) 85.7 Gastridium ventricosum (Gouan) Schinz & 42.9 Thellung (Poaceae) Lactuca serriola L. (Asteraceae) 28.6 Lotus humistratus Greene (Fabaceae) 33.3 Myosotis discolor Pers. (Boraginaceae) 28.6 Veronica L. spp. (Scrophulariaceae) 29.6 Vulpia myuros (L.) K.C. Gmel. (Poaceae) 47.6 stages was more reduced by mastication than by the HPB treatment (P = 0.050, two-sample t-test comparing within-pair differences). The within- pair difference in the abundance of immature Ceanothus was greater in the second year than in the first year after treatment (P = 0.003, two- sample t-test); abundance of this life stage was reduced in year one, but in year two, immatures were more abundant in treated than in untreated plots. Other life stages did not differ significantly in their within-pair difference between years- since-treatment nor treatment types. Fire Rings of HPB Treatment Eighty-nine species were recorded in fire rings, despite their relatively small area, which was about half of the 184 species found in the associated HPB treatment plots. Species compo- sition in fire rings differed between | yr and 2 yr post-treatment groups (MRPP: A = 0.108, P < 0.001). Only one species was a significant indicator for first year fire rings: Brodiaea elegans Hoover (Liliaceae), a native perennial geophyte (ISA: IV = 50.0). Twenty species were significant indicators for second year fire rings. All of these were annuals, most were weedy, and about half were non-native (Table 3). Species composition and abundance in the HPB treatment plots in entirety also differed between the | yr and 2 yr since treatment groups (MRPP: A = 0.082, P = 0.005). Twelve species were significant indicators for one of the two groups (Table 4). Only five were annuals, and only two of these were considered weedy (Myosotis discolor Pers., Boraginaceae; Trifolium wildenowii Spreng., Fabaceae). The non-native Myosotis was the only species that indicated both fire rings and HPB treatment plots. 2009] SIKES AND MUIR: EFFECTS OF FUEL TREATMENTS ON NORTHERN CHAPARRAL 17 TABLE 4. INDICATOR VALUES (IV) FOR SPECIES IN PERCENTAGE OF PERFECT INDICATION FOR “SHAND CUT, PILE, AND BURN” (HPB) TREATMENT PLOTS BY YEAR SINCE TREATMENT. Perfect indication (100%) occurs when the species is always present in that group and never occurs in other groups. Only statistically significant indicator species are listed (P < 0.05). Boldface indicates non-native species. l yr since treatment IV Clarkia gracilis (Piper) A. Nels. & J.F. Macbr. 76.9 (Onagraceae) Elymus elymoides (Raf.) Swezey (Poaceae) 68.0 Lomatium utriculatum (Nutt. ex Torr. & Gray) 82.6 Coult. & Rose (Apiaceae) Myosotis discolor Pers. (Boraginaceae) 76.5 Plagiobothrys cognatus (Greene) I.M. Johnston 71.4 (Boraginaceae) Poa secunda J. Presi. (Poaceae) 75.0 Trifolium willdenowii Spreng. (Fabaceae) 80.0 The age of the fire rings affected the average number of species present. First year fire rings averaged 3.2 species per 0.25 m° plot, while second year rings averaged 14.2 species. The average abundance value per species was 1.1 in the first year, indicating that almost all species were represented by only 1-3 individuals. The average abundance class was 1.5 in second year fire rings, which corresponds to half of the species having 1-3 individuals and half having 4-10. The HPB treatment plots in entirety, with 200 times the area of the fire ring plots, supported an average of 77.6 species per plot, with an average abundance value of 2.5; most species were represented by more than 10 individuals. First year fire rings had significantly fewer species in all trait groups except shrubs and perennials than did second year fire rings (P < 0.001, two-sample t-test). In contrast, native weeds was the only trait group for which species numbers differed significantly between years in HPB treatment plots overall (P = 0.013, two- sample t-test). There was an average of 3.7 more native weed species in | yr than in 2 yr since treatment plots (95% confidence interval between 1.0 and 6.5 species). Since the proportions of species representing various trait groups did not differ appreciably between years since treatment, we combined ages for HPB treatment plots overall for comparisons with fire rings. In HPB treatment plots overall, the herbaceous species were mostly annuals (70.6%). First year fire rings, however, had slightly more perennials than annuals (56% versus 44%); most appeared to have resprouted after the fire. Very little colonization of the newly open area had occurred by year one. In the second year fire rings, 94.8% of the species were annuals. The higher propor- tion of annual species and greater species richness in year two indicate that more colonization had occurred after two years. 2 yr since treatment IV Agoseris grandiflora (Nutt.) Greene 100.0 (Asteraceae) Calochortus tolmiei Hook. & Arn. 81.8 (Liliaceae) Galium porrigens Dempster (Rubiaceae) 100.0 Hesperolinon micranthum (Gray) Small 61.2 (Linaceae) Horkelia daucifolia (Greene) Rydb. 85.7 (Rosaceae) The proportion of native weeds, compared to the sum of non-native weeds and non-weedy plants, was similar in the two ages of fire rings and HPB treatment plots overall, ranging be- tween 15.2 and 17.3% of the species present. The proportions of non-native weeds, however, varied strikingly among these plot types. There were comparatively few non-native weeds in the first year fire rings (6.1%), but non-native weed species increased to twice the number of native weeds in the second year, accounting for 35.3% of the species present. The HPB treatment plots showed an intermediate proportion of non-native weeds (20.8% of the species present). Several species occurred in only one of three situations: fire rings, HPB treatments, or HPB controls. Even species that appeared to be stimulated by fire, however, such as Ceanothus seedlings, were present in control plots as well as in treated areas. Three species occurred in more than one HPB plot but not in any control plots, and all of these grew in association with burn piles. One of them, Lactuca serriola L. (Aster- aceae), an annual non-native weed, was an indicator species for the second year fire rings (Table 3). The other two species are native annuals, Stephanomeria virgata Benth. (Astera- ceae) and Gnaphalium palustre Nutt. (Astera- ceae). Though only Gnaphalium was categorized as weedy, Stephanomeria is a composite with wind-dispersed seeds. DISCUSSION The pre-treatment and control plots did not differ significantly in species composition (Sikes 2005), thus post-treatment differences between treatment and control can most probably be interpreted as treatment effects. The short term treatment effects on the herbaceous community that we detected were generally small. Other factors, such as presence of oak canopy, had a 18 MADRONO stronger influence on species composition than did treatment (Sikes 2005). Both oaks and Ceanothus provided important habitat for natives and perennials. While open areas were dominated by non-native annual grasses such as Tae- niatherum caput-medusae (L.) Nevski (Poaceae), they also supported several native annuals that are of special interest to the BLM including Navarettia subuligera Greene (Polemoniaceae) and Plagiobothrys greenei (Gray) I.M. Johnston (Boraginaceae). While treatment-induced differences in herba- ceous plant communities were smaller than expected, given the dramatic reductions in shrub cover, communities in treated areas did differ from those in controls. Treatment effects were larger for HPB than for mastication treatments, in general, and also tended to be stronger in the first year after treatment than in the second year. First year communities of both treatments differed from controls, while communities did not differ in the second post-treatment year of either treatment. The lack of mastication treat- ment effects across years probably occurred because the general increase of species abundance in the first year after mastication was counter- balanced by a general decrease in the second year. In general, the effect of time since treatment was stronger than the effect of treatment type. Taking all species composition differences be- tween matched treatment and control plots into account, we found that the two types of fuel treatment did not differ, consistent with commu- nity responses 4 to 7 yr after treatment in a nearby study area (Perchemlides et al. 2008), while the year-since-treatment groups did differ. These findings parallel those for comparisons between trait group abundances, in which twice as many trait groups differed by time since treatment as by treatment type. There was a general increase in plant species abundance and species richness with treatment, with the exception that both were lower in treated areas than in controls 2 yr after mastication. Overall shrub abundance did not differ between treatments and controls across treatment types, even though shrubs were targeted for reduction. While Ceanothus abundance across all life stages was significantly decreased after treatment, the effect was smaller than anticipated because of the survival of some cut stems and an increase in Ceanothus seedlings with treatment. Fuel treat- ments encouraged regeneration, even in the absence of fire, and it appears that the reduction of standing fuel will be short-lived. Ceanothus species can revert to closed crowns within five to seven years after mechanical brush clearing (Green 1977), and even young chaparral can burn readily under some wildfire conditions (e.g., Fried et al. 2004; Keeley and Fotheringham 2001b; Moritz et al. 2004). Overall shrub [Vol. 56 abundance was also affected by the occurrence of species other than Ceanothus. Other shrub species either were not removed because of low density or tended to resprout vigorously. Woody species that appeared to resprout without fail after cutting included Toxicodendron diversilobum (Torr. & Gray) Greene (Anacardiaceae), Quercus garryana, Amelanchier alnifolia (Nutt.) Nutt. ex M. Roemer (Rosaceae), and Prunus subcordata Benth. (Rosaceae). Rapid restoration of a shrub canopy following treatment, while perhaps unde- sirable from a fuel management perspective, can significantly decrease the likelihood that exotic species presence and abundance will increase post-treatment (Keeley et al. 2005a, b). The HPB treatment caused a general increase in shrub abundance, while mastication decreased them. Abundance of Ceanothus, in particular, was apparently more reduced by mastication than by the HPB treatment. The fire rings of the HPB treatment were responsible for this differ- ence between treatment types. Though Ceanothus seedlings occurred in most plots, whether control or treatment, their affinity for burned areas was evident, and resulted in greater differences in seedling abundance between treatment and con- trol in HPB than in mastication treatments. Previous work has emphasized that C. cuneatus is an obligate seeder that generally requires fire for seedling establishment and will not resprout after fire (e.g., Keeley 1992a). Therefore, it was unexpected to find so many Ceanothus seedlings outside of the fire rings and in control plots where disturbance was minimal. Seed dormancy in C. cuneatus 1s due to a hard impermeable seed coat that may be cracked by the heat of fire or by scarification (mechanical breakage; Keeley 1991). The seed coat may also deteriorate with time to allow germination (Quick and Quick 1961). In addition, some fraction of seed produced by fire- recruiting species, including Ceanothus species, often lacks dormancy (Keeley 1991). These alternative situations, which allow germination of Ceanothus in the absence of fire, appeared to be in operation throughout our study area, consistent with reports of C. cuneatus seedlings and immatures being found 4 to 7 yr after HPB © or mastication treatments in a nearby study area | (Perchemlides et al. 2008). Because Ceanothus seedlings were generally | more abundant in mastication treatments than in | controls, it appears that the treatment may have | increased germination, perhaps by scarifying seed | or improving microhabitats. A congener, C. | greggii, Showed increased germination a year _ following the clipping of standing chaparral at about 10cm from the ground (Moreno and | Oechel 1991). Though often described as having | no substantial germination in the absence of fire, in a previously disturbed site that was quite open | and invaded by exotic annual grasses, Keeley 2009] (1992b) found both seedlings and uneven-aged shrubs of C. cuneatus. The other species trait group that was signif- icantly more abundant in HPB treatments than in mastication treatments was native weeds. An average of 26% of native weed species increased in HPB treatments relative to controls, while the corresponding change was a 3% decrease in mastication plots. Native weeds was the only trait group for which significantly more species were present in HPB plots in the first year after treatment than the second, even though only one of seven indicator species for first year HPB treatments (versus second year) was a native weed, Trifolium willdenowii. The proportion of native weeds was not greater in fire rings than in the larger treatment plot, so fire rings were not responsible for the difference between treatments. In fact, only 2 of the 20 indicator species of second year fire rings were native weeds, Epilo- bium L. spp. (Onagraceae) and Lotus humistratus Greene (Fabaceae). Weedy plants would be expected to increase after either type of treatment, especially in the first year, because of the availability of newly exposed sites for establishment. Although treat- ment types did not differ in overall cover of bare ground or woody debris (Sikes 2005), treatment- specific differences in the distribution of woody debris could help explain the difference in native weed abundance between treatment types. Mas- tication leaves pieces of debris distributed fairly evenly over the area, perhaps inhibiting seed germination or seedling establishment, while HPB leaves debris massed at the edge of burn piles. Another difference between the two treat- ments that might affect native weed success is degree of soil disturbance. Both mastication and HPB treatments cause disturbance, but mastica- tion causes some areas to be compacted by the treads of heavy equipment and leaves other areas of soil undisturbed. HPB treatment lightly compacts and disturbs most areas of the soil with foot traffic. The larger increases in species abundance for treated plots compared to controls in the first year versus the second year after treatment may reflect an initial pulse of resource availability. In the first year, more light and resources were newly available, so the ground layer responded accordingly. Disproportionately large increases might be expected for weedy plants, since they specialize in colonizing new and disturbed habitats. However, the weedy trait groups were not significantly more abundant in the first year than the second year, and their percentage changes were not large compared to those for other trait groups. These findings contrast with other studies, perhaps due to the already abundant weed presence in our sites prior to treatment. For example, weedy and exotic plants SIKES AND MUIR: EFFECTS OF FUEL TREATMENTS ON NORTHERN CHAPARRAL 19 were more abundant on fuel breaks than on adjacent untreated areas throughout California (Merriam et al. 2006). Indeed, four to seven years after treatment, cover by exotic annual grasses was nearly twice as high on masticated or HPB treated sites compared to controls in a chaparral study area within the same BLM district (Perch- emlides 2006; Perchemlides et al. 2008). The fire rings that result from HPB treatment deserve individual consideration. The footprints of burn piles provide sites for invasive or weedy plants to establish (Korb et al. 2004). Though previous research has focused on larger slash piles that result from forest thinning operations, the principles invoked in those cases also apply in our situation. Human-condensed fuel piles can burn at higher temperatures or over a longer period than most naturally occurring fuel loads, result- ing in increased soil heating and greater damage to biotic and abiotic soil properties. Long duration soil heating causes more damage than shorter duration heating (DeBano et al. 1979). Prolonged heating is also associated with the burning of deep accumulations of masticated wood residues (Busse et al. 2005), either in prescribed burns or with unintended wildfire. Increased damage may take the form of greater water repellency (e.g., MacDonald and Huffman 2004), altered soil chemistry and structure (Shea 1993), seed mortality, and mycorrhizal steriliza- tion (Korb et al. 2004). Natural fires usually occur under dry condi- tions rather than in the moist conditions that are purposely chosen for burning piles to reduce fire danger. Burn season has complex influences on soils and biota (e.g., Knapp et al. 2007). For example, while moist soil can reduce many impacts of heating during fire, damage to soil micro-organisms can be increased in moist soils (DeBano et al. 1979; Busse et al. 2005; but see Smith et al. 2004) and seed germination of some chaparral species is decreased under moist as compared to dry heat (LeFer and Parker 2005). These factors, coupled with seasonal differences in availability of native and exotic species seed, indicate that differences in timing of pile burning could affect post-treatment responses of the plant communities. Some factors relevant to fire effects are unique to the chaparral ecosystem. Waxy-leaved shrub- lands are especially susceptible to post-fire water repellency of soil, because of hydrophobic substances produced by chaparral plants (Beschta et al. 2004). Chaparral has a thinner litter layer than forests and therefore soils are less insulated from heat (DeBano et al. 1979). In addition, shrubland wildfires often produce higher soil temperatures than forest fires because of their low, single stratum stature (Christensen 1985). Thereiore. burn, pile tenaperatures in chaparral may be closer to naturally occurring 20 MADRONO temperatures than would be the case in a forest system, although, even in chaparral, it is not likely that natural conditions would produce such concentrated fuel loads and resultant heat, or heating of as long duration, as occur in burn piles. In standing chaparral, seed recruitment after an autumn fire tends to be concentrated in areas that were gaps in the pre-burn vegetation, reflecting the high temperatures that occur in relatively dense areas of vegetation (Odion and Davis 2000). Though Ceanothus germination was promoted in the fire rings, soil temperatures may have been higher than those occurring during wildfires. Occasional Ceanothus (C. cuneatus var. fascicu- laris) and Arctostaphylos seedlings occurred in the highest fire intensity areas in a chaparral system, whereas no other taxa germinated or resprouted at those temperatures (Odion and Davis 2000). Ceanothus greggii showed increased germination when fuel loads were moderately increased in a chaparral stand, and its germination was similar to that occurring under normal fuel load when fuels were greatly enhanced (Moreno and Oechel 1991). Reactions to changes in fuel load varied by species, but the majority of chaparral species present showed decreased seedling production with increased fire intensity (Moreno and Oechel 1991). Without the diversity of fire intensity inherent in a fire through standing chaparral, in which soil temperatures range from. slightly heated to severely heated over the landscape, species diversity is likely to be reduced. The indicator species of the second year fire rings were mostly weedy and half of the species were non-native. The fact that most of the indicators for the associated HPB treatment plots were native perennials demonstrates the radical difference between the environments of the fire rings and the surrounding treatment area. Many seeds in the seed bank were likely killed by high soil temperatures under the burning piles. The plants that occur in the fire rings tend to be species whose underground tissues can withstand high temperatures and prolonged heating or colonizers that are efficient dispersers. It is apparent that many non-native species succeed as colonizers of these open sites. Several species that have been noted as having fire cue-stimulat- ed germination were present in the vicinity of fire rings, including Arctostaphylos viscida (Fried et al. 2004), Toxicodendron diversilobum, C. cunea- tus, Gilia capitata Sims (Polemoniaceae), Stepha- nomeria virgata, Clarkia purpurea (Curt.) A. Nelson & J.F. Macbr. (Onagraceae; Keeley 1991), Trifolium microcephalum Pursh (Faba- ceae), and Juncus bufonius L. (Juncaceae; Odion 2000). For Ceanothus and Stephanomeria, the association with fire rings was particularly evident, as discussed previously. Other genera present in the vicinity of fire rings include species [Vol. 56 that have been noted as post-fire recruiters, including Calystegia, Lotus, Cryptantha, Gnapha- lium, Galium, Collinsia (Keeley 1991), and Na- varretia (Odion 2000). The changes in proportion of species by trait across years in fire rings give an interesting snapshot of fire ring colonization, showing increased domination by annuals and introduced weeds over the two years. The proportions of species attributes in the larger associated treat- ment plot lie between the two extremes of the first year and second year fire rings’ distribution of species attributes. Longer term studies of succes- sion in fire rings are needed to determine whether they provide a locus for enhanced invasion by such species into the surrounding area or they eventually return to a similar species composition as their environs. CONCLUSION This study did not show a large effect of fuel treatments on herbaceous plant communities in the short term, though vegetation structure was altered by removing or redistributing woody biomass. Fuel treatments may have caused relatively little community alteration because of a history of disturbance and the already extensive occurrence of introduced species in our study site. Alternatively, our relatively coarse abundance data may have been insufficiently detailed to detect changes that occurred. Both fuel treatments increased species richness initially, probably because both caused distur- bance and increased resource availability. The greatest effect was detected in the first year after treatment. By the second year species abundance was generally lower in the mastication treatment plots than in associated controls. In the HPB treatment, species abundance was lower in the second year than the first year, but was still higher than in their controls. The effects of treatment on overall species composition were stronger for the HPB treatment than the mastication treatment. The primary factor responsible for the difference between the two treatments appeared to be the fire rings that remain after piles are burned in the HPB treatment. Though soil heating is likely greater or longer lasting under the burn piles than it | would be during a chaparral fire, this treatment | introduces the element of fire into a community | that is adapted to it, and it may allow fire- | adapted species in the community to persist. It is likely that factors beyond the presence of fire rings contributed to the differential effects of | the two treatments. Levels of soil disturbance and | distribution of woody debris are markedly | different in the two treatments, and these factors | should influence species composition. However, | 2009] the importance of these factors cannot be assessed based on the data that we collected. With the evidence at hand, it appears that neither fuels reduction treatment is a definite detriment to the plant community over the one to two year post-treatment period we studied. Short term data suggest that the HPB treatment may lead to an increase in weedy and non-native species. At the same time, however, it may increase native diversity by promoting species with fire-cued germination. In contrast, the mastication treatment appears to reduce species diversity. Both treatments tended to promote Ceanothus germination, and rapid recovery of a shrub canopy is likely to reduce abundance of weedy and non-native species over time (Keeley et al. 2005a). With patchy application, both treatments will increase the heterogeneity of the overall northern chaparral community in the absence of wildfire. Our results do not lead to a clear recommen- dation concerning future management of chap- arral fuels in southwestern Oregon. Further monitoring will increase our understanding of the longer term effects of these fuel treatments on plant communities. In the meantime, managers should be aware of the potential negative effects that are linked to either treatment. ACKNOWLEDGMENTS The authors would like to thank the Medford Oregon District of the Bureau of Land Management that funded this research through a Challenge Cost Share contract with Oregon State University, as well as M. Wineteer and L. Kayes for logistical assistance. B. McCune, P. E. Hosten, D. A. Pyke, D. L. Luoma, E. A. Holt, and K. A. Perchemlides provided helpful com- ments on the research. The comments of John Hunter and two anonymous reviewers greatly improved the manuscript. LITERATURE CITED BESCHTA, R. L., J. J. RHODES, J. B. KAUFFMAN, R. E. GRESSWELL, G. W. MINSHALL, J. R. KARR, D. A. PERRY, F. R. HAUER, AND C. A. FRISSELL. 2004. Postfire management on forested public lands of the western United States. Conservation Biology 18:957-967. Busse, M. D., K. R. HUBBERT, G. O. FIDDLER, C. J. SHESTAK, AND R. F. Powers. 2005. Lethal soil temperatures during burning of masticated forest residues. International Journal of Wildland Fire 14:267-276. CHRISTENSEN, N. L. 1985. Shrubland fire regimes and their evolutionary consequences. Pp. 85-100 in S. T. A. Pickett and P. S. White (eds.), The ecology of natural disturbance and patch dynamics. Aca- demic Press, Orlando, FL. DEBANO, L. F., R. M. RICE, AND C. E. CONRAD. 1979. Soil heating in chaparral fires: effects on soil properties, plant nutrients, erosion, and runoff. Research Paper PSW-145, Pacific Southwest Forest and Range Experiment Station, Forest Service, SIKES AND MUIR: EFFECTS OF FUEL TREATMENTS ON NORTHERN CHAPARRAL 21 United States Department of Agriculture, Berke- ley, CA. DETLING, L. E. 1961. The chaparral formation of southwestern Oregon, with considerations of its postglacial history. Ecology 42:348—357. DUFRENE, M. AND P. LEGENDRE. 1997. Species assemblages and indicator species: the need for a flexible asymmetrical approach. Ecological Mono- graphs 67:345—366. FRANKLIN, J. F. AND C. T. DYRNEss. 1973. Natural vegetation of Oregon and Washington. General Technical Report PNW-8, U.S. Department of Agriculture, Forest Service, Pacific Northwest Forest and Range Experiment Station, Portland, OR. FRIED, J. S., C. L. BOLSINGER, AND D. BEARDSLEY. 2004. Chaparral in southern and central coastal California in the mid-1990’s: area, ownership, condition, and change. PNW-RB-240. U.S. De- partment of Agriculture, Forest Service, Pacific Northwest Research Station. Portland, OR. GREEN, L. R. 1977. Fuel reduction without fire — current technology and ecosystem impact. Pp. 163— 171 in H. A. Mooney and C. E. Conrad (eds.), Proceedings of the symposium on the environmen- tal consequences of fire and fuel management in Mediterranean ecosystems, August I—5, 1977, Palo Alto, California. Forest Service, U.S. Department of Agriculture, Washington, D.C. HICKMAN, J. C. (ed.) 1993. The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. JOHNSON, D. R. 1993. Soil survey of Jackson County area, Oregon. United States Department of Agri- culture, Soil Conservation Service. Available at: ftp://ftp-fc.sc.egov.usda.gov/MO1/text_pdf/oregon/ or632_text.pdf. Accessed May 15, 2009. KAUFFMAN, J. B. 2004. Death rides the forest: perceptions of fire, land use, and ecological restoration of western forests. Conservation Biol- ogy 18:878—-882. KEELEY, J. E. 1991. Seed germination and life history syndromes in the California chaparral. The Botan- ical Review 57:81—116. 1992a. Demographic structure of California chaparral in the long-term absence of fire. Journal of Vegetation Science 3:79—90. . 1992b. Recruitment of seedlings and vegetative sprouts in unburned chaparral. Ecology 73:1194— 1208. . 2002. Fire management of California shrub- land landscapes. Environmental Management 29:395—408. AND C. J. FOTHERINGHAM. 2001a. History and management of crown-fire ecosystems: a summary and response. Conservation Biology 15:1561—1567. AND . 2001b. Historic fire regime in southern California shrublands. Conservation Bi- ology 15:1536—-1548. , M. BAER-KEELEY, AND C. J. FOTHERINGHAM. 2005a. Alien plant dynamics following fire in Mediterranean-climate California shrublands. Eco- logical Applications 15:2109—2125. : , AND . 2005b. Determinants of postfire recovery and succession in Mediterranean- climate shrublands of California. Ecological Ap- plications 15:1515—1534. 29 MADRONO KNAPP, E. E., D. W. SCHWIk, J. M. KANE, AND J. E. KEELEY. 2007. Role of burning season on initial understory vegetation response to prescribed fire in a mixed conifer forest. Canadian Journal of Forest Research 37:11—22. Kors, J. E., N. C. JOHNSON, AND W. W. COVINGTON. 2004. Slash pile burning effects on soil biotic and chemical properties and plant establishment: rec- ommendations for amelioration. Restoration Ecol- ogy 12:52-62. LAVOREL, S. 1999. Ecological diversity and resilience of Mediterranean vegetation to disturbance. Diversity and Distributions 5:3-13. LEENHOUTS, B. 1998. Assessment of biomass burning in the conterminous United States. Conservation Ecology [online] 2(1). Available at: http://www. ecologyandsociety.org/vol2/issl/artl/. Accessed May 15, 2009. LEFER, D. AND V. T. PARKER. 2005. The effect of seasonality of burn on seed germination in chaparral: the role of soil moisture. Madrono 52:166—174. LILLYWHITE, H. B. 1977. Animal responses to fire and fuel management in chaparral. Pp. 368—373 in H. A. Mooney and C. E. Conrad (eds.), Proceedings of the symposium on the environmental consequences of fire and fuel management in Mediterranean ecosystems, August 1-5, 1977, Palo Alto, Califor- nia. Forest Service, U.S. Department of Agricul- ture, Washington, D.C. MACDONALD, L. H. AND E. L. HUFFMAN. 2004. Post- fire soil water repellency: persistence and _ soil moisture thresholds. Soil Science Society of Amer- ica Journal 68:1729—1734. McCUNE, B. AND J. B. GRACE. 2002. Analysis of ecological communities. MjM Software Design, Gleneden Beach, OR. AND M. J. MEFFORD. 1999. Multivariate analysis of ecological data version 4.25. MjM Software Design, Gleneden Beach, OR. MERRIAM, K. E., J. E. KEELEY, AND J. L. BEYERS. 2006. Fuel breaks affect nonnative species abun- dance in Californian plant communities. Ecological Applications 16:515—527. MoRENO, J. M. AND W. C. OECHEL. 1991. Fire intensity effects on germination of shrubs and herbs in southern California chaparral. Ecology 72:1993—2004. Moritz, M. A., J. E. KEELEY, E. A. JOHNSON, AND A. A. SCHAFFNER. 2004. Testing a basic assump- tion of shrubland fire management: how important is fuel age? Frontiers in Ecology and Environment 2:67-72. NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRA- TION (NOAA). 2004. Climatological data annual summary Oregon 2003. 109(13), 1—36. ODION, D. C. 2000. Seed banks of long-unburned stands of maritime chaparral: composition, germi- nation behavior, and survival with fire. Madrono 47:195—203. AND F. W. DAVIS. 2000. Fire, soil heating, and the formation of vegetation patterns in chaparral. Ecological Monographs 70:149—169. [Vol. 56 PERCHEMLIDES, K. A. 2006. Impacts of fuel reduction thinning treatments on oak and chaparral commu- nities of southwestern Oregon. M.S. thesis. Oregon State University, Corvallis, OR. , P. S. MuIR, AND P. E. HOSTEN. 2008. Responses of chaparral and oak woodland plant communities to fuel-reduction thinning in south- western Oregon. Rangeland Ecology and Manage- ment 61:98—109. Quick, C. R. AND A. S. QUICK. 1961. Germination of Ceanothus seeds. Madrono 16:23—30. ROGERS, J. O., T. E. FULBRIGHT, AND D. C. RUTHVEN, III. 2004. Vegetation and deer response to mechan- ical shrub clearing and burning. Journal of Range Management 57:41-48. SHEA, R. W. 1993. Effects of prescribed fire and silvicultural activities on fuel mass and nitrogen redistribution in Pinus ponderosa ecosystems of central Oregon. M.S. thesis. Oregon State Univer- sity, Corvallis, OR. SIKES, K. G. 2005. The effects of two fuel reduction treatments on chaparral communities in southwest Oregon. M.S. thesis. Oregon State University, Corvallis, OR. SMITH, J. E., D. MCKAy, C. G. NIWA, W. G. THIES, G. BRENNER, AND J. W. SPATAFORA. 2004. Short- term effects of seasonal prescribed burning on the ectomycorrhizal fungal community and fine root biomass in ponderosa pine stands in the Blue Mountains of Oregon. Canadian Journal of Forest Research 34:2477-2491. UNITED STATES DEPARTMENT OF AGRICULTURE (USDA), FOREST SERVICE. 2002. Phase 3 Field Guide. Section 10. Lichen Communities, Version 4.0. October 2007. USDA, Forest Service: Forest Inventory and Analysis Program. Available at: http:// www.fia.fs.fed.us/library/field-guides-methods-proc/ docs/2007/p3_4-0_sec10_10_2007.pdf. Accessed May 26, 2009. USDA, NATIONAL RESOURCES CONSERVATION SER- VICE (NRCS). 2004. The PLANTS Database, Version 3.5. National Plant Data Center, Baton Rouge, LA. Available at: http://plants.usda.gov. UNITED STATES DEPARTMENT OF THE INTERIOR (USDI). 1999. Ecosystem restoration in the Ash- land Resource Area. USDI Bureau of Land Management, Medford District Office, Medford, OR. VEBLEN, T. T. 2003. Key issues in fire regime research for fuels management and ecological restoration. USDA Forest Service Proceedings RMRS-P-29. ZEDLER, P. H. 1995. Fire frequency in southern California shrublands: biological effects and management options. Pp. 101—102 in J. E. Keeley and T. A. Scott (eds.), Brushfires in California wildlands: ecology and resource management. International Association of Wildland Fire, Fair- field, WA. ZIMMERMAN, G. M., H. GOETZ, AND P. W. MIELKE, JR. 1985. Use of an improved statistical method for group comparisons to study effects of prairie fire. Ecology 66:606—611. MADRONO, Vol. 56, No. 1, pp. 23-42, 2009 REPRODUCTIVE BIOLOGY OF THE SAN FERNANDO VALLEY SPINEFLOWER, CHORIZANTHE PARRYI VAR. FERNANDINA (POLYGONACEAE) C. EUGENE JONES', FRANCES M. SHROPSHIRE, LAURA L. TAYLOR-TAFT’, SEAN E. WALKER, LEO C. SONG, JR., YOUSSEF C. ATALLAH, ROBERT L. ALLEN, DARREN R. SANDQUIST, JIM LUTTRELL, AND JACK H. BURK Department of Biological Science, California State University, Fullerton, California 92834-6850 ABSTRACT In response to conservation concerns, the reproductive biology of the San Fernando Valley Spineflower was investigated, focusing on pollination interactions and seed germination. Pollination by a variety of aerial visitors, as well as autogamy (a facultative selfer, showing about 25% selfing), appear to contribute significantly to fruit/seed set. There was a significant correlation between the numerous different floral visitors (many went uncollected) and the invertebrate fauna in the immediately surrounding coastal sage scrub community indicating that this taxon is visited by a substantial variety of potential pollinators and is probably not pollinator-limited. Although there were many potential pollinators, only six species, including three species of ants, made up the vast majority of visits to the flowers at the two study sites. Many of the invertebrate visitors to the flowers of the San Fernando Valley Spineflower exhibited a high rate of constancy. An overall generalist strategy is suggested. Seed set was high and a germination rate of over 70% occurred without pre-treatment. Key Words: Ants, Chorizanthe parryi var. fernandina, floral constancy, generalist pollination strategy, mixed mating, pollination biology, Polygonaceae, San Fernando Valley Spineflower, selfing. The present study investigated a variety of the factors associated with the reproductive biology of Chorizanthe parryi S. Watson var. fernandina (S. Watson) Jepson, the San Fernando Valley Spineflower (SFVS), an herbaceous low-growing annual thought to be extinct (Hickman 1993) until its 1999 rediscovery on the Ahmanson Ranch in Ventura County, California. Subse- quently, it was found on the Newhall Ranch, 17 mi northeast of the Ahmanson Ranch, in Los Angeles County (California Natural Diversity Database 2001). Although the Ahmanson Ranch and the Newhall Ranch both support large populations of the SFVS, fewer than 20 acres of habitat at the Ahmanson Ranch (Meyer 2003) and no more than 25 acres at the Newhall Ranch are known to support this species (Mary Meyer, personal communication, California Department of Fish and Game). Therefore, a total of not more than 45 acres of habitat currently exist (at least have been discovered) where this species still can be found. Historically, this taxon 1s reported to have had a much larger range extending from Lake Elizabeth in Los Angeles County to near Del Mar in San Diego County (Munz and Keck 1959; Glenn Lukos Associates, Inc. 1999; Jones et al. ' Author of correspondence, cejones@fullerton.edu °Current address: School of Life Sciences, Arizona State University, P.O. Box 874501, Tempe, AZ 85287- 4501 2002). Currently, it is designated as a List 1B.1 plant (Rare, Threatened, or Endangered in California or Elsewhere; seriously endangered in California) by the California Native Plant Society and is State-listed Endangered (CNPS 2001) and a Federal candidate for similar listing (CNPS 2005). Knowledge of the reproductive biology of a rare plant is often critical to any management plan developed to ensure the long-term survival of that species (Kearns and Inouye 1997). Such studies involve a detailed analysis of all aspects of plant reproductive biology, including the breed- ing system, pollination interactions, fruit/seed set, dispersal and germination, growth and survival, etc. (Kearns and Inouye 1993). Following the rediscovery of the SFVS on the Ahmanson Ranch, a series of surveys and directed research activities were undertaken to determine the size and extent of the on-site populations, any off-site occurrences, and factors important to its survival. These initial studies were reported in Sapphos Environmental, Inc. (2001), which also includes a summary of the known information regarding the pollination biology of this plant. On the basis of an apparently abundant seed set and a brief field observation in 1999, C. E. Jones suggested pollinators were probably not limiting the repro- duction of this species. The current study was then developed to address systematically the reproductive factors of pollination interactions and germination success. 24 MADRONO Fic. 1. San Fernando Valley Spineflower at the Ahmanson Ranch, Ventura Co., California. Photo by Bob Allen. Scope of Study Given the importance of reproductive biology to the survival of endangered plant species (Harper 1979), this study was designed to initiate an information database addressing the repro- ductive biology of the SFVS. The specific hypotheses examined are as follows: 1) The SFVS is pollinated by a variety of invertebrate pollinators. 2) Seed-set in the SFVS is not pollinator limited. 3) The SFVS has visitors that demonstrate a high degree of constancy. 4) The SFVS 1s a facultative selfer. 5) SFVS seed germinates readily without special treatment. MATERIALS AND METHODS Plant Species The SFVS occurs primarily in dry, sandy places within coastal scrub communities at elevations below 350 m (Munz and Keck 1959). Reveal (1989) has described the SFVS as occurring between about 215 and 335m, a somewhat narrower range than that listed for C. p. var. parryi (90-350 m). Stems of the SFVS spread more or less horizontally from the base to form a low, flat- topped, grayish plant 0.2—0.8 (1) dm high and 0.5—4 (6) dm across (Jepson 1925; Reveal 1989; Fig. 1). Leaves are mostly basal, 0.5—2.5 (4) cm long, oblanceolate to oblong, and canescent (Jepson 1925; Reveal 1989; Hickman 1993). The predominantly sessile, single-flowered involucres are more or less openly distributed in small clusters (Munz and Keck 1959) at branchlet ends [Vol. 56 (Jepson 1925) and are urn-shaped, 1.5—-2 mm long, grayish pubescent, bearing six bracts and three awns (Reveal 1989). In the SFVS, these involucral awns are straight rather than hooked, a trait that distinguishes C. p. var. fernandina from the more widely distributed C. p. var. parryi (Reveal 1989). The sessile flowers are 2.5—3 mm long with a greenish-white tube and 6 white, sparsely hairy lobes, occurring in two series of 3 (Reveal 1989; Hickman 1993). Filaments and anthers of the nine stamens are white (Reveal 1989), whereas the pollen varies from white to pink. The pollen is heteromorphic (1.e., pollen wall sculpturing with- in a single pollen grain varies), a characteristic found in only eight taxa within the genus (Russell 2003). The significance of such sculpturing is unknown. The ovary is glabrous (Jepson 1925) and bears three styles with dry stigmas (Reveal 1989). Nectar is present around the base of the ovary and between the filaments. The flowers are protandrous (Taylor-Taft 2003) and are pro- duced in late spring, April-June (Munz and Keck 1959). The fruit is a brown achene, 2.5—-3 mm long with a 3-angled beak (Reveal 1989). Seeds of the genus Chorizanthe are reported to contain a straight embryo and abundant endosperm (Re- veal 1989). Voucher SFVS specimens were deposited in the Fay A. MacFadden Herbarium (MACF) at California State University, Full- erton, California. Study Sites Investigations were first carried out in 2001 at the Ahmanson Ranch, the initial rediscovery site, located in the southeastern corner of Ventura County, California (Fig. 2). SFVS populations are found primarily on the slopes of Laskey Mesa and in isolated areas on the Mesa itself. Plants usually occur within open areas free of a significant number of competing species and are generally associated with San Andreas soils. These soils are composed of fine particles, soft sandstone, and loose sandy gravelly deposits (Glenn Lukos Associates Inc. 1999). The specific study locations (GPS coordinates: 34°10.360'N, 118°40.839'W to 34°10.473'N, 118°40.277'W; Fig. 2) were selected because they contained abundant SFVS. Each site is charac- terized by having mostly barren soil, surrounded by coastal sage scrub communities, which are often substantially invaded by annual Mediterra- nean grasses and ruderals. Of the five sites chosen, four occur on southwest-facing slopes and the fifth occurs nearby on the mesa proper. Subsequent investigations were completed in 2004 at the Newhall Ranch located in Los Angeles County, California (Fig. 3). The specific study locations (34°24.743’N, 118°37.786'W to 34°25.975'N, 118°35.044’W; Fig. 3) were selected 2009] 118°41'W FIG. 2. JONES ET AL.: REPRODUCTION IN SAN FERNANDO VALLEY SPINEFLOWER 25 Approximate locations of study sites on Laskey Mesa, Ahmanson Ranch. Aerial map for this figure provided by Terraserver/Globexplorer, (copyright 2005; used with permission). because they had sufficient numbers of SFVS for observational requirements. Again, each site is characterized by supporting sparsely vegetated areas containing some bare ground and light litter, surrounded by coastal sage scrub commu- nities, again with a substantial non-native annual grass component. Of the areas chosen for study, site 1 at the Grapevine Mesa is located on a west facing slope, site 2 at Airport Mesa is on a southwest facing slope, and site 3 at the Magic Mountain area is on a southeastern facing slope (Fig. 3). Dawn-to-Dusk Observations Ahmanson Ranch. Limited pollinator availabil- ity has proven to be a problem in some arid zone rare plants. To determine if pollinators are limiting reproductive success in the SFVS, we examined pollinator behavior, diversity, and the relative importance of each of the major pollina- tor groups by employing a series of dawn-to-dusk surveys that were conducted during the early (from 20 April through 22 April 2001), mid- (4 May through 6 May 2001) and late (18 May through 20 May 2001) bloom of the SFVS at three separate study sites (1, 2, and 4; Fig. 2) on the slopes of Laskey Mesa, Ahmanson Ranch. A total of 126 hr of dawn to dusk observations were completed during three observation periods (42 hr each during the early, mid and late seasons). For the purposes of this study, early bloom is defined as the time when approximately 25% of the SFVS plants were in flower, mid-bloom as the time when at least 50% of the SFVS plants were in flower, and late bloom as the time when approximately 75% of the SFVS plants had completed flowering. Dawn-to-dusk means that the possible pollinators visiting SFVS plants were observed during at least 10 min out of each hour beginning on the hour after sun up and continuing throughout the day until 50 min after the hour before sun down. Each survey involved three consecutive days of observation. At each of the three study sites, three subpopulations (e.g., 1A, 1B, and IC) were selected on the basis of the ease with which one person could observe a sizeable number of plants simultaneously. One observer was assigned to each of the three separate study sites at the Ahmanson Ranch and, later, to the three New- hall Ranch study locations as well. That person observed and recorded the visitors to the SFVS plants in the initial subpopulation (e.g., 1A) during the first 10 min of each hour. That same observer then had 10 min to move to the second subpopulation (1B) where visitors were observed and recorded from 20 min after the hour until half past the hour. Finally, that same observer then rotated to the third subpopulation (1C) and repeated the process from 40 min after the hour until 50 min after the hour. 26 MADRONO [Vol. 56 118°41°Ws118°40'Wsd118°39'W_s-118°38'Ws118°37'Ws118°36'W_s118°35'W Newhall Ranch - San Fernando Valley Spineflower Conservation Plan 2004 Pollinator Study 3000 Feet 914.4 Meters Henry Mayo Drive/ Saugus Ventura Road, State Route 126 Golden State Freeway |5 4 Magic Airport t Mountain- Mesa Parkway Grapevine Mesa @ Magic Mountain Valencia Parkway 118°41;Ws118°40'Wsid118°39'Ws118°38'Wsi118°37'W_s118°36'W_s118°35' W Fic. 3. Specific locations of sites investigated on the Newhall Ranch in Los Angeles County. Names for the Study Sites are: 1—Grapevine Mesa Site, 2—Airport Mesa South Site, and 3—-Magic Mountain Site. For the purposes of this study, a “‘visitor” was defined as any organism that actually landed on and came into contact with the anther(s) and/or the stigma(s) of the flower. “‘Visits” were defined as the number of times that a particular visitor landed on a SFVS flower and probed that flower for nectar and/or pollen. Data were subsequently analyzed in terms of number of visits and visitors. Diurnal temperatures were measured with a digital, indoor/outdoor thermometer (Digital Thermo-Clock, available from Oregon Instru- ments, P.O. Box 1190, Cannon Beach, OR 97110). Newhall Ranch. Subsequently, similar dawn-to- dusk surveys were completed twice during the blooming period (from 23 April through 25 April 2004 and 7 May through 9 May 2004) of the SFVS at three separate study sites at the Newhall Ranch. These latter observations were conducted only during the mid- and late-bloom periods of the SFVS because 2004 was a very dry year (only about 19.1 cm of rainfall versus 54.6 cm in 2001), whereas the 95 yr average for this area is 47.3 cm) and many fewer plants were available for study and because those that were available flowered earlier and for a shorter period of time than had the plants at the Ahmanson Ranch in 2001. Based on the results of SFVS investigations at the Ahmanson Ranch, dawn-to-dusk observa- tions were conducted at the Newhall Ranch between the hours of 9:00 a.m. and 7:00 p.m. (Jones et al. 2002). A total of 90 hr of dawn to dusk observations were completed during three observation periods (30 hr each during the early, mid and late seasons). All other details related to how the surveys were conducted were identical to those at the Ahmanson Ranch. Pollinator Collection and Identification Ahmanson Ranch. Representative samples of visitors were collected on 20-22 April and 4-6 and 18-20 May between 10:00 and 17:00. Sampling was primarily conducted at a different location (Site 5; see Fig. 2) than those used for the dawn-to-dusk observations in order to eliminate the possibility of decreasing pollinator visitations through collection. Organisms seen visiting three or more flowers were captured in an insect net or by using a blowing aspirator and placed in killing jars charged with ethyl acetate. Each individual was then placed in a vial with 70% ethyl alcohol. 2009] Newhall Ranch. To determine if there was a positive association between the invertebrate community in the vicinity of the SFVS and actual visitors to the SF VS, thus indicating the extent to which the SFVS was being visited by the potential vectors available in the specific area where the plants were located, we set up a time based sampling method to capture these potential pollinators. Individual insects that were on or in the area of the SFVS were collected using aspirators and nets. Samples were collected for a total of 30 person minutes at each site and each captured individual was placed into a glassine envelope. Collections were primarily conducted at a location (subpopulation) near, but not within, the dawn-to-dusk study subpopulations at each of the three sites. Again, this was done in order to eliminate the possibility of decreasing pollinator visitations as a result of collection. One sample was collected on April 23rd, three on May 7th, and three on May 8th, 2004. For analysis, these were pooled for a single site for a single day. For each insect collected, we noted whether it was found on or near the SFVS. We also employed pitfall traps to sample ground dwelling arthropods. Each trap consisted of a single 16 oz plastic cup filled with approx- imately 4 0z of propylene glycol to act as a preservative. Three traps were placed at each site, each covered with hardware mesh to prevent the capture of vertebrates. A single pitfall sample consisted of approximately 48 h of continuous trapping (from Friday afternoon until Sunday afternoon). Pitfall traps were open from the 23rd until the 25th of April and from the 7th until the 9th of May 2004. All captured arthropods were identified to order, morphospecies, (this being essentially a recognizable taxonomic unit, see Oliver and Beattie 1993, 1996), or to species, if possible. Pollen Analysis Ahmanson Ranch. Visitors collected for identi- fication were returned to the laboratory and the vials shaken to remove pollen from the body surface. The cap of each vial was removed and the ethyl alcohol was allowed to evaporate down to about a single drop. That drop was then placed on a slide and, following evaporation, cotton blue (1% aniline blue in lactophenol) was added to stain the pollen grains. Slides were viewed under a Leitz compound microscope where an a priori set minimum of 200 pollen grains were identified using reference slides. The number of plant species and pollen grains found on each individ- ual visitor was used to determine which pollina- tors carried the pollen of SFVS and how constant they were to the SFVS. Pollinator constancy was defined on a percentage basis. The higher the percentage of one pollen species in a sample, the JONES ET AL.: REPRODUCTION IN SAN FERNANDO VALLEY SPINEFLOWER 2g more specific that pollinator was to that partic- ular plant species. For the purpose of this study, a pollinator was considered to be “‘constant”? when that pollinator visited a given species at least 95% of the time during a single foraging flight. All captured visitors were examined and a determi- nation was made of the pollen they were carrying. Newhall Ranch. Each visitor captured while visiting at least three flowers was examined under a Bausch and Lomb dissecting micro- scope to see if pollen was present on the visitor and, if so, where it was located. A 3 cm piece of double sided Scotch tape with one end cut to a point. The pointed end was used to pick up any available pollen from the visitors under the dissecting scope. Once the pollen had been transferred from the visitor to the double-sided tape, the tape was placed on a 7.62 cm X 2.54 cm X | mm glass microscope slide. One or two drops of cotton blue were then added to stain the pollen grains and the slide allowed to sit for at least 24 hrs before examination. Slides were viewed under a Leitz compound micro- scope and were identified using reference slides (prepared with known SFVS pollen using the identical staining technique). Types and num- bers of pollen grains recovered from each individual were used to determine which polli- nators carry the pollen of SFVS and how constant they were to the SFVS. Exclusion Experiments Exclusion or bagging experiments were at conducted at Ahmanson Ranch at Site 3 (on the Mesa) to determine whether the SFVS requires a vector to facilitate the pollination process and to determine the relative importance of ants and other crawling insects as pollinators. The bagging experiments were set up on 7 April 2001 along an old dirt road selected to minimize possible destruction of SFVS plants since the experimental design required the use of well- separated plants. Prior to blooming, 30 plants were selected haphazardly to serve as controls (Control 1). Another 30 plants were selected and surrounded by Tangle-Trap® (Insect Trap Coat- ing Brand, the Tanglefoot Company, Grand Rapids, Michigan 49504) to prevent any terres- trial pollinators from reaching the flowers. This series, termed ‘‘Experimental” (ant free), served to determine the role of flying versus crawling insects (primarily ants) as pollinators. A third set of 30 plants was covered with a wire cage screened with nylon and secured to the substrate with u-shaped sections of wire. The base of these cages was also surrounded with Tangle-Trap®. This latter series, which excluded both crawling and flying insects, was used to determine if the SFVS is self-pollinated and was termed “‘Self”’. 28 MADRONO These last two sets of 30 plants (Experimental and Self) were initially also surrounded by concentric rings of cinnamon (suggested as an environmentally friendly barrier) as well as Tangle-Trap®, with cinnamon on the inner ring. This practice was abandoned approximately half way through the experiments since the cinnamon appeared to be toxic to plant leaves coming into contact with it. A final group (fourth set) of 30 representative plants was selected haphazardly to serve as an additional (second) set of controls (Control 2). This set became necessary when it was noted that the originally selected control plants were some- what smaller than the two series of experimental plants. All plants chosen were evenly distributed along a plot approximately 115 m long by 3 m wide. Individual plants located more than 8 cm from the closest SFVS neighbor were chosen for the self-pollination treatment and the experimen- tal treatment to allow room for the Tangle-Trap® layer without having to sacrifice surrounding plants. Tangle-Trap® was renewed weekly. After the plants set fruit, they were harvested, placed in paper bags, and returned to the laboratory. Fruits were then removed from their involucres and a minimum of 200 involucres was sampled from each plant to determine fruit set for each treatment and the controls. Data were analyzed using ANOVA (Tukey’s multiple com- parison procedure) in Minitab version 13.31 (Minitab, Inc., State College, PA 16801-3008). To test if the exposure to cinnamon affected fruit set, each of the plants exposed to cinnamon was divided into two portions. Fruit set on the inner half (older portions of the plant, exposed to cinnamon) was compared to fruit set on the outer half (younger part of the plant, not exposed to cinnamon) using a paired t-test. Nectar Availability In order to provide information on the nectar rewards being presented to visitors at Ahmanson Ranch, one ul microcapillary tubes were carefully inserted into 20 flowers on each of 23 separate plants at Site 3 on 14 May 2001. The flowers to be sampled on each of the selected plants were carefully enclosed within nylon bags the previous day to ensure that no nectar was removed by visitors prior to the sampling procedure. Only newly opened flowers were sampled. Microcapil- lary tubes were placed in 4 dram glass screw-cap vials and transported to the laboratory for measurement. The amount of nectar in each microcapillary tube was measured under a Bausch and Lomb dissecting microscope and the amount of nectar produced per flower was calculated. To determine any possible diurnal pattern of SFVS nectar production, mature plants grown in [Vol. 56 1 gallon pots at the California State University, Fullerton, greenhouse complex were sampled. These plants were in the later phase of flowering and open flowers had been exposed to pollen vectors prior to this study. Sixty hours before observation, several branch portions on each plant were enclosed in nylon bags to exclude pollinators. On 1 July 2001, six plants were moved into the greenhouse, placed in an exclu- sion room, and the bags partially removed to allow access to the newly opened flowers. Nectar sampling was conducted at 9:00, 13:00, and 17:00. At the initial sampling time, nectar was collected from 20 flowers, 10 each on two separate plants. At 13:00, flowers on these two plants were re- sampled and 20 flowers on two new plants were also sampled. Finally, at 17:00, all flowers were re-sampled and 20 more flowers on two addi- tional plants were sampled. Nectar was sampled with one ul microcapillary tubes from open flowers only. Seed Germination Representative SFVS seeds were collected on 10 July 2001 from plants at Ahmanson Ranch Study Site 5. Fruits were sampled from the distal portions of flowering branches. All seeds were cleaned and removed from the surrounding floral remains to ensure that only whole seeds were being tested. Except where noted, all seed treatments were placed in an unheated greenhouse under ambient light in Petri dishes with one sheet of Whatman No. | filter paper. Dishes were watered with reverse osmosis (RO) water as needed. Germination was defined as the emergence of the embryonic root or radicle. Samples of 50 seeds each (except for Treatment 2 where 51 seeds were used) were tested in the following ways: 1) Control. 2) Leach 24 hr: seeds leached 24 hr under running tap water, then placed in a Petri dish. 3) Leach and Stratify: seeds leached 24 hr, placed in a Petri dish, then stratified for 2 wk in a refrigerator at 3-4°C. 4) Stratify Only: seeds sown directly in a Petri dish, then stratified for 2 wk in a refriger- ator at 3-4°C. 5) Direct Planting: seeds sown in a layer of sand on top of greenhouse soil mix in plastic flower pots, watered until soaked, then placed on outside benches. Pots were watered as necessary to keep the sand from drying out. This treatment served as a supplementary control and was used to determine the basic non-treatment viability of the seeds. The time interval for these germination tests was approximately six weeks for seeds that were 2009] 1200 0 Ants 1000 & Honey bees W@ Other bees Flies & Beetles 800 {) Others 2 ‘a 600 > 400 200 0 _ ~ Mra Ik | LALE =e & NY) s JONES ET AL.: REPRODUCTION IN SAN FERNANDO VALLEY SPINEFLOWER 29 Z ap Ly UOUAUUNUTAVEGTE UGH OEUUEUEUUCUUCUUAUOTUUUUEUERTTRAOUTRTTGUTUUUETDTAOAERTEEAUG TU U UOT UOUAUEETUU UOT AOEAUUCUUE TT OTE ENETTTT ET Time FIG. 4. Total dawn-to-dusk visits observed, by hour, during all observation periods for the major visitor groups at all study sites at the Ahmanson Ranch during the 2001 SFVS flowering season. subjected to the stratification treatments and three weeks for all other treatments. RESULTS Dawn-to-Dusk Observations Ahmanson Ranch. The results of these dawn-to- dusk observations are summarized in Fig. 4. It 1s apparent that this herbaceous plant received numerous visits (total of 9816) from a wide range of invertebrate organisms; however, among the invertebrate groups, five species accounted for nearly 75% of all visits made to the SFVS. These included: two species of ants, Dorymyrmex insanus Roger (37.8%) and Solenopsis xylonii McCook (2.6%) (Hymenoptera: Formicidae); two species of beetles, Zabrotes sp. (1.9%) (Coleoptera: Bruchidae: Amblycerinae), and Em- menotarsis quadricollis LeConte (2.4%) (Coleop- tera: Melyridae: Dasytinae [identification tenta- tive]); and the European honey bee, Apis mellifera (30.3%) (Hymenoptera: Apidae). Other taxa captured on the flowers of the SFVS at the Ahmanson Ranch included Diptera in the families Bombylidae, Syrphidae, Calliphoridae, Sarcophagidae, and Tachinidae, Hymenoptera in the families Ichneumonidae, Chrysididae, Sphe- cidae, Halictidae, Andrenidae, Megachilidae, Pompilidae, Vespidae, and Formicidae, and Lepidoptera in the families Riodinidae and Hesperiidae. Early season.—We performed detailed analy- ses on each of the early, mid- and late seasons and for all three seasons combined. The total number of individual visitors (hereafter referred to as visitors) and the total number of flower visits (hereafter referred to as visits) by each were analyzed separately (Figs. 5 and 6). In terms of visits, an analysis of the early (20— 22 April 2001) series of observations (Fig. 5) showed that of the 1662 visits were made by Visits 100% 90% 80% 70% g 60% nee re ees = 50% = Flies 8 lll Beetles a 40% BAnts 30% 20% 10% 0% Early Middle Late Total FiG. 5. Summary of the early (n = 1662), middle (n = 5184), late (n = 2984), and total season (n = 9830) SFVS visits recorded during dawn-to-dusk observations at the Ahmanson Ranch in 2001. 30 Visitors 100% 80% 2 60% W Others j= = Bees t Flies 9 ll Beetles a 40% + WAnts a 20% 0% Early Middle Late Total Fic. 6. Summary of the early (n = 179), middle (n = 804), late (n = 710), and total season (n = 1693) SFVS visitors recorded during dawn-to-dusk observations at the Ahmanson Ranch in 2001. various invertebrates of which >54% were made by ants and 40% were by bees nearly all of which were made by Apis mellifera. Further, of the 179 individual visitors observed during the early bloom (see Fig. 6), ants were dominant (84%). Among the ants, the dominant species was Dorymyrmex insanus, which account- ed for 74% of the visitors. Apis mellifera accounted for 7% of the total visitors during this early portion of the blooming season. Mid-season.—Regarding visits, an analysis of the middle (4-6 May 2001) series of observations (Fig. 5) showed that of the 5184 visits by various invertebrates, of which 49% were by bees, the vast majority of those were made by Apis mellifera. Additionally, of the 804 individual visitors observed during the mid-blooming period (Fig. 6), 21% were ants, 46% were beetles, 19% were flies, and 9% were bees, of which 6% were Apis mellifera. Among the ants, significant visitors were Dorymyrmex insanus (9.3%) and Solenopsis xylonii (8.1%). Late season.—Again, in terms of visits, an analysis of the late series of observations (Fig. 5) revealed that of the 2984 visits by various invertebrates, dominated by ants (71%). Of the 710 individual invertebrate visitors (Fig. 6), 77% were made by ants. Among the ants, the dominant species was Dorymyrmex insanus, Which accounted for 71% of all visitors during the late SFVS blooming season. Entire flowering season.—If all visit data from early, mid-, and late season sampling periods are combined for the three study sites (Fig. 5), the total number recorded is 9830. Visits by ants accounted for 37% of this total, while 8% were made by beetles, and 55% were made by all other invertebrates. Within this latter group (55%), MADRONO [Vol. 56 more than 38% were made by flies and more than 59% were made by bees. Of the bee visits, 90% were made by Apis mellifera. Of the 1693 visitors noted throughout the season (Fig. 6), 51% were ants, 27% were beetles, 13% were flies, and 6% were bees. It is interesting to note that the greatest number of visits (5184) and visitors (804) was recorded during the mid-blooming period (4-6 May 2001) when the vast majority of the SFVS plants were in full bloom. This compares to 1662 visits and 179 visitors during the early bloom (20— 22 April 2001) and 2984 visits and 710 visitors during the late bloom (18—20 May 2001). Newhall Ranch. Visitors to the flowers of the SFVS at the Newhall Ranch were dominated by one species of ant, Forelius mccooki, and flower beetles in the family Melyridae (unknown Dasy- tinae). Together these taxa made up nearly 50% of all floral visitors to the SFVS at the Newhall Ranch. Besides these dominant taxa, other species captured on the flowers of the SFVS at the Newhall Ranch included representatives from the Coleoptera in the family Bruchidae, Diptera in the families Bombylidae, Syrphidae, Calliphor- idae, Sarcophagidae, and Tachinidae, Hymenop- tera in the families Chrysididae, Sphecidae, Halictidae, Megachilidae, Pompilidae, Vespidae, and Formicidae, and Lepidoptera in the family Riodinidae. Mid-season.—An examination of the visits made by each of the visitor groups observed during the mid-season period (23—25 April 2004) shows that flies (79%) greatly outnumbered beetles (16%) in terms of the number of flowers — actually visited by individual visitors at Site 1. | Flies (67%) also dominated the visits at Site 2, — followed by beetles (11.5%) and ants (11%). Ants | (61%) made the most numerous visits at Site 3 © followed by beetles (32%). Total visits varied | among the three sites from 2021 at Site 1, 633 at © Site 2, and 2488 at Site 3. Overall, 5142 visits were made to SFVS flowers during the mid-season flowering period. During that time, flies (40%) | dominated the floral visits, followed by ants _ (33%) and beetles (23%; Fig. 7). Visitors to the flowers of the SFVS varied | substantially among the three study sites during the mid-season. Flies (67%) and beetles (27%) | t dominated the visitors at Site 1 and at Site 2° (58.5% for flies and 21.5% for beetles), whereas | flies were replaced by ants (43%) as dominant visitors along with beetles (42%) at Site 3. Total | visitors also varied considerably among the sites, — with 722 visitors at Site 1, 130 at Site 2, and 483 | at Site 3. Overall, 1335 visitors were observed on | the flowers of the SFVS during the mid-season | flowering period. During this period, flies | (42.5%) beetles (32%), and ants (20%) were the | dominant SFVS floral visitors (Fig. 8). Observa- | 2009] Visits 100% > 90% + 80% - x 70% - 8 @ Others re) 4 fa 60% # Bees = 50% + G Flies 9 li Beetles 5 40% - rt @ Ants 30% 20% - 10% ~ 0% + = Middle Late Total Fic. 7. Summary of mid-season (n = 5142), late season (n = 2864), and total (n = 8006) visits recorded during dawn-to-dusk observations at the Newhall Ranch in 2004. tions of interest include the total lack of honey bee visitors and the variation in ant species present at each location. Late season.—An examination of the number of flowers visited during the late season observa- tions (7-9 May 2004) by each of the visitor groups shows that flies (90%) greatly outnum- bered beetles (6.5%) at Site 1. Flies (31%), beetles (28%), ants (25%), and bees (15%) almost equally dominated the visits at Site 2 (Fig. 7). Ants (78%) made the most numerous visits at Site 3, followed by bees (11%) and beetles (9%). Total visits varied among the three sites with 1483 at Site 1, 372 at Site 2, and 1009 at Site 3. Overall 2864 visits, or about half the number of visits seen during the mid-season, were made by the visitors to the flowers of the SFVS during the late season flowering period. During this late season, flies (51%) dominated the floral visits, followed by ants (32%) and beetles (10%; Fig. 7). Visitors to the flowers of the SFVS also varied substantially from site to site during our late- season observations. Flies (83%) dominated the visitors at Site 1, followed by beetles (12%). At Site 2 there was a more equal distribution of visitors with beetles (31%), ants (28%), flies (25.5%) being the dominant visitors, whereas ants (70%) were by far the dominant visitors at Site 3. Total visitors also varied considerably among the sites, with 429 visitors at Site 1, 133 at Site 2, and 171 at Site 3. Overall, 733 visitors, or about half the number of visitors seen during the mid-season, were observed on SFVS flowers. During this late season, flies (54%) dominated the floral visitors, followed by ants (24%) and beetles (16%; Fig. 8). Entire flowering season.—If all the data from the mid- and late season sampling periods are combined for the three study sites (Fig. 7), the JONES ET AL.: REPRODUCTION IN SAN FERNANDO VALLEY SPINEFLOWER a Visitors 100% ——— —— 90% -_ 80% _ 70% 8 | @ Others e 60% #8 Bees € 50% + Flies S 40% ; ili Beetles a 30% meee 20% I ; =e 10% 0% +— - Middle Late Total Fic. 8. Summary of mid-season (n = 1335), late season (n = 733), and total (n = 2068) SFVS visitors recorded during dawn-to-dusk observations at the Newhall Ranch in 2004. total number of recorded visits is 8006, which is reasonably close to the total number of visits (8168) recorded during the mid- and late flower- ing periods of study at the Ahmanson site, even though the 2004 season was a much drier year that produced many fewer individual plants. Flies dominated the floral visits (45.5%), followed by ants (32.5%) and beetles (16%). Bees, especially honey bees, were not well represented among the visitors to the flowers of the SFVS during the entire blooming period. Total visitors also varied considerably among the sites. Overall, 2068 visitors were observed on the flowers of the SFVS during the mid-season flowering period (Fig. 8). During this period, flies (42.5%) dominated the floral visitors, followed by beetles (32%) and ants (20%). Pollinator Collection and Identification Ahmanson Ranch. Identified visitors to the SFVS, collected primarily at Study Site 5, included 5 species of beetles, 20 species of flies, 3 species of butterflies, and 27 species of bees, ants, wasps, and their relatives. For a complete list of visitors, contact CEJ via email. These were all captured on the flowers of the SFVS. Newhall Ranch. In total, we captured 4223 individuals on the SFVS flowers and the adjacent coastal sage scrub community. This sample consisted of 12 different insect orders, two arachnid orders (Acarina and Araneae), one myriapod (Chilopoda), one crustacean (Isopoda), and a large number of Collembola (Fig. 9). Non- insect taxa made up a large portion of the sample (2267 individuals). In particular, 45% of the sample was made up of Collembola. Taxa representing 7 different insect orders were captured on or in close proximity to the SFVS [Vol. 56 32 MADRONO 100 ; iC = 10 + Led Y U £ a = 3 1 2 < o = hed o oO 01 ; I | 0.01 4 . ql xe << <4 - ~Y xe 2 x a & & » & "8 3 et ee 3 a ax ee ; NS os ra ee G ¥ x a N2 1°) ¢ ® Q we FIG. 9. Relative abundance of different invertebrate taxa captured at Newhall Ranch. These taxa were collected on the flowers and, also in the surrounding coastal sage scrub community. flowers (Fig. 10). These same 7 orders were also found in the surrounding coastal sage scrub community (a total of 1912 individuals, Fig. 10). The relative abundance of the insects captured on the flowers (44 individuals) is largely reflective of their relative abundance in the surrounding coastal sage scrub community (R* = 0.9337; Fig. 11). In terms of the insect species diversity observed at the Newhall Ranch, we identified 101 different morphospecies. Hymenopterans (bees and ants) were the most diverse order of insects, followed by Dipterans (flies), Coleopterans (beetles), and Hemipterans (true bugs) (Table 1). The most abundant orders were generally the most diverse. @ Near Flower On Flower 60 S = 50 @ vu <= 40 uc S 30 a P- e20 = 10 ac ft) MS oe SE oe eo ” & < EN SOS UN Fic. 10. Insect community composition (relative abundance) of different insect orders captured on flowers and near flowers. The near flowers sample includes data from pitfall traps and hand collected individuals. Non-Pollinator (NP) orders includes groups that were never observed on SFVS (Arche- ognatha, Dermaptera, Phasmatodea, Orthoptera, and Psocoptera). This group makes up only 3% of the entire sample. Of the 101 different morphospecies, 49% are represented by a single specimen and 16% by two specimens. Only in 7 morphospecies did we collect 50 or more individuals (Fig. 12). These data indicate that there are a number of infrequent species present and very few abundant species. These data were only recorded at New- hall Ranch and were not taken at Ahmanson Ranch. The taxa of fauna collected varied with the sampling method. Pitfall traps primarily capture active ground-dwelling arthropods and tend to underestimate the abundance of inactive or non- ground dwelling species (e.g., Adis 1979; Spence and Niemela 1994; Work et al. 2002). Using pitfall traps, we captured 1665 individuals from — 69 different morphospecies and from hand- collecting we captured 291 individuals from 51 | 60 R2 = 0.9337 50 40 30 20 10 % of Community Sample 0 10 20 30 40 50 60 % of Sample on Flower Fic. 11. Relationship between the relative abundance | of taxa captured on the spine flower and their relative abundance in the community (R* = 0.9663, P < 0.05). | 2009] TABLE 1. DISTRIBUTION OF MORPHOSPECIES COL- LECTED ACROSS INSECT ORDERS. Order Number of morphospecies Hymenoptera 31 Diptera 18 Coleoptera Bs) Hemiptera 14 Homoptera 7 Orthoptera 6 Lepidoptera 3 Archeognatha 2 Psocoptera | Phasmatodea l Dermaptera l different morphospecies. Results of the two sampling methods share only 21 species in common (Jaccard Coefficient = 0.21). In addi- tion, based on species accumulation curves (Colwell and Coddington 1994), we have sampled the species captured with pitfall traps much more thoroughly than the species we collected by hand (Fig. 13). This is evident since the species accumulation curve for the hand collection is essentially linear whereas the pitfall species accumulation curve is hyperbolic and the rate of species accumulation has decreased. Since pollinator diversity was estimated using only hand-collected individuals, it is possible that we have underestimated the number of pollinator species. Pollen Analysis and Constancy Ahmanson Ranch. Of all insect floral visitor specimens captured, only those of Apis mellifera were captured in sufficient numbers and had enough pollen lodged on their bodies (an a priori set number of 200 pollen grains was established as an adequate subsample to examine from each specimen) to carry out a complete analysis of the degree of their constancy to the SFVS. Data on 50 45 40 35 30 25 1 2 3 4 5 6 7 JONES ET AL.: REPRODUCTION IN SAN FERNANDO VALLEY SPINEFLOWER 33 60 Observed 50 - — «= Singletons 40 3g Vv A 30 wn re) o 20 a E ped = 0 a) Number of Samples 80 70 Observed - - - - Singletons 60 W v 3 50 a « 40 Oo © 30 fe 3 20 2 10 0 0 S 10 LS 20 Number of Samples FIG. 13. Species accumulation curves for pitfall trapped (upper) and hand collected (lower) samples. Singletons are species represented by a single specimen. the purity of pollen loads was collected from 10 different individuals of Apis mellifera. In all but one case, these honeybees were very constant to the SFVS with constancy levels ranging between 96-99%. The one exception carried only 59% SFVS pollen. Overall, the pollen loads on Apis mellifera averaged 94% (SD = 12.4) SFVS pollen and an average of 4.5% (SD = 2.1) other pollen species per sample. 8 9 10 iito 2ito 31ito 41to >50 20 30 40 50 Number of Specimens Fic. 12. Number of individuals collected of each species. 34 MADRONO The remaining taxa visiting the flowers of the SFVS either did not carry at least 200 pollen grains or their sample size was less than three, an examination of the pollen loads of these visitors was completed to determine their specificity to the SFVS flowers. Of the non-honeybee visitors captured while visiting the SFVS flowers, 56 carried one or more pollen grains in general (see Jones et al. 2004). Of those 56, 48 (86%) carried one or more pollen grains of the SFVS. Of the 48 that carried pollen of the SFVS, 28 (58%) carried only SFVS pollen. The average constancy for all 48 specimens that carried at least one pollen grain of the SFVS was 87%, with the range varying from 1.6 to 100%. The constancy for the most abundant visitor recorded on the SFVS flowers at the Ahmanson Ranch, Dorymyrmex insanus, was 100%; howev- er, we had only a sample of two individuals. These two carried 10 and 9 pollen grains of the SFVS respectively. Five individuals of Solenopsis xylonii, another prominent ant visitor to the flowers of the SFVS, also proved to be carrying SFVS pollen (between 8 and 37 pollen grains) and exhibited an average constancy of 98%. The average constancy exhibited by all ants was 98% and ranged from 89 to 100%. Newhall Ranch. The 43 insect floral visitors caught while visiting the SFVS represented at least 14 different species of potential pollinators. Of those 43, 58% carried pollen loads of one or more pollen grains. The 25 floral visitors that carried pollen loads represented at least 10 different taxa of potential pollinators. Of these, 72% carried only SFVS pollen, whereas the remainder carried mixed pollen loads, but all included some pollen from the SFVS (for a complete list of the individual species, contact CEJ via email). Of the 17 individuals of the small red ant species Forelius mccooki, caught on the SFVS flowers and sampled for pollen, 76.5% carried one or more SFVS pollen grains. Of the 13 that carried pollen, 69% carried only SFVS pollen. The remainder carried mixed loads, but all included some pollen of the SFVS. Exclusion Experiments Regarding the two sets of plants exposed to cinnamon, no significant difference was found between the number of fruits produced on the inner half of the plants and the outer portion (t- test Control I, t = 0.2554, P > 0.8; Control II, t = 0.2713, P > 0.8; Experimental (without ants), t = 0.307, P = >0.8; and Self, t = 0.2413, P > 0.8; in all cases n = 5, df = 4), thus indicating that the cinnamon layer did not affect fruit set and, therefore, that there was no difference in efficien- cy of pollination of the plants throughout the duration of the experiment. [Vol. 56 Fruit Set in SFVS Fruit set % Control 1 Control2 Experimental Self Treatment Fic. 14. Comparison of SFVS fruit set between the two sets of controls (Control 1, plants selected prior to the beginning of plant bloom, and Control 2, plants selected at the end of the experimental period. Treatment ‘Experimental’ excludes crawling insects, especially ants, and treatment ‘Self’ excludes all visitors. Percent fruit set is determined by dividing the total number of fruits set by the total number of flowers examined. Error bars indicate one standard deviation. Those marked ‘‘a” are not significantly different from one another but are significantly different from the Self, which is marked “‘b’”’. ANOVA (f = 48.02, n = 119, P< 0.0001) followed by a Tukey’s Test (q = 3.69, n = 119, df = 118, P < 0.01). An analysis (ANOVA) of the fruit set in the two sets of controls and the two exclusion treatments (Fig. 14) showed a significant differ- ence in fruit set among the two sets of Controls, the Experimental, in which crawling insects (such as ants) were prohibited from visiting the SFVS | flowers, and the Self plants (f = 48.02, n = 119, P | < 0.0001). Specifically, there was no significant difference between the two sets of Controls or | between the Controls and the Experimental | plants, but there was a significantly lower number | of fruits produced by the selfing treatment » (Tukey’s multiple comparison, q = 3.69, n = 119, df = 118, P < 0.01). Nectar Analysis In the sampling of 460 bagged SFVS flowers (20 each on 23 separate plants), the average nectar production per flower was 0.0034 ul with a range from 0.0 to 0.014 ul per flower. Bee flowers | normally have an average of 2.5 + 1.1 wl per flower (Cruden et al. 1983). Insufficient nectar _ was produced for refractometer determinations — of sugar content. Nectar sampling to determine any possible. diurnal pattern of nectar production is presented | in Table 2. No measurable nectar was detected at. the 9:00 sampling period. However, measurable nectar was collected in all afternoon samples. 2009] TABLE 2. ON | JULY 2001. N = 20 flowers per plant sample. 900 hrs—ul/fl Plants | and 2 0.0000 Plants 3 and 4 =. Plants 5 and 6 = Average 0.0000 Calculated nectar values ranged from 0.00234 to 0.00938 ul per flower. Seed Germination The results of the seed germination tests are presented in Table 3. Germination in the Control seeds after 6 wk was 74% whereas germination in all other treatments was noticeably lower. In particular, stratification, either alone or in combination, decreased percent germination. DISCUSSION The reproductive biology of rare and endan- gered plants has been of great interest to biologists charged with developing successful management strategies for these species (Purdy et al. 1994; Schemske et al. 1994; Luiyten et al. 1996; Bernardello et al. 1999; Kaye 1999; Timmerman-Erskine and Boyd 1999). This aspect of the conservation or reintroduction of rare species involves not only understanding the factors affecting seed production, but also factors affecting long-term successful propagation (Gi- blin and Hamilton 1999). In order for a population to remain stable, the plants must both flower and receive sufficient pollinators in order to produce viable seeds. Those seeds must then receive enough water and nutrients, avoid predation, and grow to mature flowering plants capable of producing the next generation. Inter- ference with any of these steps will inhibit reproduction and, if consistent over time, may result in reduced populations (Kaye 1999). Results from this investigation establish an information base regarding the initial steps in the reproduction process of the SFVS that should TABLE 3. SEFVS SEED GERMINATION RESULTS AS OF 4 FEBRUARY 2002. Sample size Number % Treatment (N) germinated germination Control 50 3] 74 Leach 24 hr 51 22 44 Leach and stratify 50 9 18 Stratify only 50 5 10 Direct planting 50 2D 50 JONES ET AL.: REPRODUCTION IN SAN FERNANDO VALLEY SPINEFLOWER 35 DIURNAL NECTAR PRODUCTION IN THE SFVS, AS MEASURED IN THE CSUF GREENHOUSE COMPLEX 1300 hrs—nl/fl 1700 hrs—l/fl 0.00391 0.00344 0.00938 0.00234 — 0.00313 0.00664 0.00297 prove useful to conservation biologists responsi- ble for protecting this endangered taxon. Pollinator Diversity—Generalist Strategy Data from the dawn-to-dusk visitor surveys at both the Ahmanson and Newhall Ranches are consistent with a generalist (both within and between years) rather than a specialist pollination vector strategy. We documented a wide variety of taxa, terrestrial as well as aerial (including a number of different species of ants, bees, beetles, and flies) that visit the SFVS flowers and would appear to be capable of effecting pollination with subsequent seed set. With the exception of one of the three species of ants that are frequent visitors to the flowers of the SFVS, we did not examine in any detail the efficiency of the various possible vectors at facilitating fruit set. Given that the ant visitors were among the most frequent to the flowers of the SFVS, we did examine the efficiency of the major ant visitor at Ahmanson Ranch (Dorymyrmex insanus) as a pollinator of the SFVS and this work is reported elsewhere (Jones et al. unpublished). Both the Northern Hemisphere (annuals) and South American (mostly perennials) members of the genus Chorizanthe occupy xerophytic habitats in which rainfall is quite variable from year to year (Goodman 1934; Reveal and Hardham 1989). Such variation may favor the generalist pollination strategy that seems to be the rule for species in this genus. The generalist strategy has been found in all members of this genus that have received any attention regarding the pollinators that frequent their flowers (Reveal and Hardham 1989; Bauder 2000; U.S. Fish and Wildlife Service 2001; Murphy 2003). Specialization of pollinators has been a key concept of plant-pollinator coevolutionary mod- els (Thompson 1994). This assumption is based on the “most effective pollinator principle”, which states that natural selection should proceed toward floral phenotypes that attract a limited spectrum of potential pollinators and result in an increase in the effectiveness of fruit set in the plant (Stebbins 1970). This process by which the flowers are molded by a small group of related and effective pollinators is referred to as “‘adap- tive specialization” (Thompson 1994; Herrera 1996; Johnson and Steiner 2000). 36 Although this principle has often been used as the underlining support for the idea that evolu- tion should proceed toward more specialized mutualistic associations between specific flowers and their pollinators, Stebbins (1970) emphases that this refers to the “predominate and most effective vector’ and does not mean that the plant is “‘pollinated exclusively by this vector,” and as such, supports the concept of pollinator syndromes. In other words, by this “‘principle’’, a plant species should have one or only a few related primary pollination vectors (e.g., two or more bee species with similar morphologies and behaviors would fall into the bee pollination syndrome—melittophily of Faegri and van der Pijl 1971), but certainly can have a few to several other pollinators as well. In fact, as pointed out by Futuyma (2001), there is a continuum between species that are exclusive specialists pollinated by only a single vector species to species that are pollinated by a vast array of vector species. Therefore, each species has to be individually evaluated regarding how specialized or general- ized its pollination system actually is. So the question becomes, why do some plant species adopt more inclusive guilds of potential pollinators that include species of dissimilar sizes and behaviors representing different taxonomic groups? When this occurs, we refer to the plant as having adopted a generalist pollination strategy (see Heithaus 1974; Waser et al. 1996; Gomez and Zamora 2006; Olesen et al. 2007; Ollerton et al. 2007 for a more complete overview of the generalist strategy). From the perspective of an annual plant, a generalist pollination strategy tends to increase the likelihood of at least some successful fruit/ seed set in fluctuating environments that are the norm in the variable southern California Medi- terranean climate (Waser et al. 1996; Gomez and Zamora 2006). However, there may be genetic consequences related to the different sets of insect vectors and their behavior, ranging from visits of many flowers on a single plant before moving to the next plant (favoring selfing), to moving quickly among flowers on separate plants (favor- ing outcrossing) (Stebbins 1950; Schmitt 1983). In our study, rainfall varied considerably between the times of the study at the Ahmanson and Newhall Ranch sites (2001 and 2004 respectively) and resulted in many fewer SFVS plants being produced at both sites during the 2004 season than had been produced in 2001. Fewer plants mean fewer resources are available to attract and hold the services of potential pollinators, espe- cially those with high energy requirements like the European honey bee (Johnson and Hubbell 1975; Schaffer et al. 1979; Sih and Baltus 1987; Jennersen and Nilsson 1993; Conner and Neu- meier 1995). However, a decrease in resources should not be as important to potential pollina- MADRONO [Vol. 56 tors with minimal energetic requirements such as ants and other small-bodied pollinators. In our study, 2001 was a good year and produced sufficient SFVS plants that the bloom even attracted the European honey bee, Apis mellifera, which was one of the top five visitors to the plants that season (Jones et al. 2002). However, the 2004 year was much drier and produced many fewer SFVS plants at both sites (Ahmanson and Newhall Ranches) and resulted in no honey bees being attracted to these plants, even though they were observed visiting other taxa in the vicinity (Jones et al. 2004). During both seasons, ants were common visitors to the SFVS plants. As a result of the way they forage, ants would tend to facilitate selfing more frequently than outcrossing (Jones et al. 2002, 2004; unpublished data). In contrast, honey bee behavior on these flowers would tend to increase the probability of outcrossing over selfing. Although honey bees collected both nectar and pollen when visiting the SFVS flowers, the minimal nectar rewards produced per flower, forced them to visit many flowers to achieve a full nectar crop prior to returning to the hive. An intriguing aspect to consider is how such annual differences in pollinators might affect the genetic structure of the offspring in any given season. One would expect that the progeny produced during drier years would reflect more selfing, whereas that produced in wetter years would reflect more outcrossing. It would be quite interesting to collect seed produced under such environmental conditions and examine the genet- ic variation of progeny produced to determine if pollinators have significant effects on the genetic | structure of the progeny produced during such | divergent rainfall years (Barrett 2002). Local Invertebrate Community in the Coastal Sage Scrub Plant Community and Potential Pollinators for the SFVS The insect community at the Newhall Ranch . was quite diverse and both sampling methods captured a number of infrequent species as | evidenced by the high percentage (~50%) of | species represented by only a single specimen (singletons). Given this and the fact that the species accumulation curves did not reach an | upper asymptote, it is very likely that we have | underestimated the insect diversity in this com- | munity. However, these results are not unlike other studies of arthropod communities in coastal « sage scrub (CSS). Burger et al. (2003) found that | approximately 51% of their 169 morphospecies | were represented by a single specimen. Their | results and ours suggest that the terrestrial | arthropod community in the surrounding com- | munity is very diverse and that adequately | sampling the terrestrial arthropod diversity will 2009] require a much greater effort than that repre- sented in this study. We found that 14 of 101 insect morphospecies were captured on the flowers of the SFVS and might be potential pollinators. Of these 14 species, 10 carried grains of SFVS _ pollen. Interestingly, our sample of potential pollinators contained 8 singletons, which also suggests that there is a substantial number of species with very low abundance in this community and, again suggests that we have probably underestimated the insect diversity in this community. Regardless of these issues, it is clear that the most abundant orders of insects in our sample (e.g., Hymenop- tera, Coleoptera) also represent the largest number of potential pollinators. Seed Set and Pollinator Limitation The effectiveness of the SFVS visitors as pollinators is demonstrated, at least partially, by the high seed set (50-60%) registered in both sets of controls associated with the exclusion exper- iments completed at the Ahmanson Ranch study site. Additionally, the data suggest that the contribution of ants and flying insects to seed set 1s equivalent to that of aerial visitors alone. This would indicate that ants are not normally required for full seed set and supports the idea that this species has adopted a generalist pollina- tion strategy (Waser et al. 1996; Gomez and Zamora 2006). However, ants may be important pollinators when other vectors are scarce (Jones, et al., unpublished data) The actual pollinating species may depend primarily upon vector availability, the diversity of which varies with seasonal and annual environmental fluctuations. In any case, pollinator limitation would appear not to be a problem for the SFVS. Pollinator Constancy Harper (1979) has noted that most rare plant taxa rely on insect pollination and that survival of many rare plants depends on the maintenance of sufficient pollinator populations. As defined here, constancy is the condition that exists when a floral visitor frequents only a single species on a given foraging bout. Many pollinators such as honeybees, bumblebees and lepidopterans have demonstrated constancy to specific species (Free 1963; Lewis 1989; Goulson et al. 1997). Such flower constancy increases the likelihood of a plant receiving pollen from a member of the same species and, in turn, increases the likelihood of successful fertilization. This also generally de- creases the flower handling time for the pollinator (Waser 1986; Chittka et al. 1999). In addition, when pollinators visit a single species they are less likely to transfer a pollen grain of another species that could possibly clog the stigma with incom- JONES ET AL.: REPRODUCTION IN SAN FERNANDO VALLEY SPINEFLOWER 37 patible pollen (Waser 1986; Chittka et al. 1999). Frequently rare species exhibit reduced seed set (Baskin and Baskin 1998), one possible symptom of low constancy. However, seed set and con- stancy are both high in the SFVS. Constancy, as determined by captured speci- mens from both the Ahmanson and Newhall Ranches (details for the latter are given in table 4 in Jones et al. 2004), was high among most of the floral visitors. The significant number of visitors, both terrestrial and aerial, demonstrating con- stancy to the SFVS may also be reinforced by the patchy distribution pattern of this taxon, which facilitates both nectar collection and constancy (Chittka et al. 1999). Although we examined a relatively small sample of ants captured on SFVS flowers at the Ahmanson Ranch, those that were had pollen loads that were 98% specific to the flowers of the SFVS, indicating that these ant species were purposefully visiting these plants, using them as a food source, and in the process picking up pollen and probably facilitating the successful reproduc- tion of the SFVS (Jones et al. 2002). Nectar Availability Nectar production has an energetic cost for plants. Plants of arid environments with profuse flower production per plant and subjected to fluctuating rainfall from year to year are likely to produce smaller individual flowers and less nectar per flower when under stress. Consequently, their normal nectar production per flower may very well be relatively small due to this evolutionary constraint (Southwick 1984; De la Barrera and Nobel 2004). The SFVS occupies a seasonally dry habitat that varies considerably in the amount of rainfall received prior to and during the growing season of the plant. Quite likely as a result of such constraints, the nectar production per flower in the SFVS is minimal. In such cases there is still a possibility that through rapid visits to a large number of flowers pollinators could be able to collect sufficient nectar to satisfy their energetic needs (especially taxa with higher energetic demands, Spira 1980). In comparison to the average nectar per flower found in bee flowers (Cruden et al. 1983), the low nectar production per flower observed in the SFVS forces larger bodied floral visitors (like honey bees) to visit many flowers during a single foraging bout, helping to ensure the pollination of many flowers and, perhaps, facilitating in- creased outcrossing among plants in the popula- tion (Proctor et al. 1996). It is interesting to note that diurnal nectar production was highest during the early after- noon and remained fairly constant throughout the remainder of the day. Floral visitation was also highest during this period, particularly 38 MADRONO among the ants, which are smaller bodied potential pollinators that have smaller energetic requirements per individual (Peters 1983; Degen et al. 1986). Breeding System—Selfing Our data from the bagging experiments completed at the Ahmanson Ranch study site indicate that the SFVS is at least partially self- fertile. In fact, nearly 30% of the flowers set fruit with viable seeds in our controlled experiments where the plants were bagged preventing any pollination vectors from having access to the flowers; thus, these flowers set fruit without the services of a vector. However, our data also indicate that more flowers set fruit when the flowers are exposed to potential pollination vectors. It would appear that at least some of this fruit set is due to cross pollination suggesting that the SFVS can probably set fruit by selfing when pollinators are limiting or when pollinators are small and tend to frequent many flowers on the same plant (ants), as well as, by outcrossing when pollinators are abundant or when larger polli- nators such as honey bees are among the floral visitors. Stebbins (1957) indicated that geographically restricted plants, such as the SFVS, are likely to be self-compatible. He postulated that in species whose populations fluctuate frequently, selection would favor self-compatibility during times of smaller populations. He also suggested that self-fertilization is common in annual plants of California and other Mediterranean regions due to dramatic climate variability. During particularly dry years, conditions fa- vorable for cross-pollination may be absent or only present for a short time. Furthermore, Hagerup (1932) suggested that dry climates lead to lower populations of pollinators and Roubik (2001) has shown annual variability in pollina- tor populations. Karron (1991) and Barrett (2002) have also noted that rare plant taxa may be more likely to be self-compatible than more widespread species. Numerous rare plant species have been shown to be self-fertile (Kunin and Shmida 1997; Bosch et al. 1998; Anderson et al. 2001). These self- pollinating flowers generally have smaller, less showy flowers (Proctor et al. 1996; Barrett 2002). Such small flower size and short relative distance from anther to stigma are also often associated with self-compatibility (Kunin and Shmida 1997; Anderson et al. 2001). The SFVS demonstrates these characteristics. Situations in which some fraction of the fruit is set by selfing and some fraction is set by outcrossing creates a mixed mating strategy (Vogler and Kalisz 2001). Such a strategy appears [Vol. 56 to be present in plants found in seasonally variable habitats with unpredictable rainfall regimes, resulting in large variations in plant population sizes and unpredictable pollination vector populations. Under these conditions, a mixed mating strategy seems to provide some assurance of reproductive output each and every year (Barrett 2002, 2003). A more thorough discussion of the potential roles of ants versus honey bees in facilitating selfing or outcrossing and a discussion of the potential importance of each to the continued survival of the SFVS will be presented in a separate paper (Jones et al. unpublished data). Seed Germination Plants require more than successful pollination and fertilization to complete sexual reproduction. First, the seeds produced must also be viable. Secondly, many species have complex germina- tion patterns (Bewley and Black 1994) that may contribute to differential survival and recruitment (Burdon et al. 1983), an important consideration in the biology of rare species. Therefore, an examination of the germination potential of any seeds produced is necessary in order to form an accurate picture of the reproductive capacity ofa particular species. In general, SFVS seeds seem to germinate easily, without the need for any special treatment. Even the seeds that did not germinate may remain viable and might actually represent seeds that exhibit delayed germination and become part of a seed bank (Baskin and Baskin 1998). Although SFVS seeds that were subjected to special treatments seemed to germinate at a lower rate than control seeds, this may be due simply to the short timeline of these experiments. Stratifi- cation, in particular, appeared to slow the » germination process but this may result from | the slowing effects of lower temperatures per se | on physiological processes, rather than on special © seed dormancy requirements. Obviously, this | aspect of the reproductive process requires » further study. An accompanying study of possible fire-related | germination cues, such as heat shock, smoke, and | charred wood showed that none of these treat- | ments significantly stimulated germination in the | SFVS (Sandquist 2003). The general ease of germination allows for the | rapid establishment of SFVS populations in open spaces, much like a weedy strategist (Baskin and | Baskin 1998), and supports the establishment of the dense patches that provide the floral display and nectar resources required to attract flying | pollen vectors. Clearly, maintenance of these | dense patches in the face of drought, disturbance | and alien species should be a matter of concern for managers. 2009] Conservation Concerns—Effects of Population Size Variability Low population sizes, resulting in decreased floral display and nectar resources, would be expected to lead to decreases in the number of flying visitors. Our data indicate that during harsh, dry, growing seasons, as was the case for our Newhall Ranch study, the SFVS may survive by producing a significant number of progeny via selfing without a vector. A significant decrease in the number of floral visitors or the production of a significant number of progeny via selfing without a vector would have important genetic implications in terms of interpopulation gene flow (Barrett 2002). Certainly, an additional study that should be undertaken is a detailed analysis of the population genetics of this species throughout its extant range in order to determine its genetic health and to establish management strategies to maintain or enhance that health. Summary The reproductive biology of the SFVS, at least in the aspects studied here, appears to be characterized by great flexibility. No special treatment was required for seed germination, allowing for the rapid establishment of the dense SFVS patches observed in the field. Pollination interactions involved a substantial variety of insects, both terrestrial and aerial, and included a mixed mating system characterized by being able to set fruit without the need of a pollination vector. However, far more fruit were set when the flowers were exposed to potential pollination vectors. Five species at the Ahmanson Ranch were responsible for 75% of the total visits to SFVS flowers: Apis mellifera; two species of native ants, Dorymyrmex insanus and Solenopsis xylonii; and two species of beetles, Zabrotes sp. and Emmenotarsis quadricollis. The small size of the latter four native species (all under 5 mm in length) allows easy entry into the SFVS flowers, whereas the larger, introduced, Apis mellifera accesses the nectar rewards with its long probos- cis. At the Newhall Ranch, by far the most common insect captured on the flowers of the SFVS (17 of 43 or 39.5%) was the small red ant, Forelius mccooki. Size and mobility differences among these six prominent visitors may have implications for their respective roles in the SFVS pollen flow. Flight distance and, therefore, pollen dispersal range is much greater for Apis mellifera than for the five native species. The five native species tend to visit flowers on the same plant (the SFVS proved to be a facultative selfer) or on nearby neighbors. It would seem that Apis mellifera would foster longer distance pollen dispersal than the other four species and would have a greater JONES ET AL.: REPRODUCTION IN SAN FERNANDO VALLEY SPINEFLOWER 39 likelihood of facilitating outcrossing. Native bees may normally play this role but few were observed during this study. Therefore, the role of the smaller species might be one of facilitating reproductive success through within-plant or within-patch pollination, whereas the larger, more mobile pollinators, like Apis mellifera, should provide the inter-patch or inter-subpopu- lation pollen flow important for the maintenance of the SFVS genetic diversity. The presence of both guilds of pollinators should facilitate overall reproductive success. Clearly, this generalized pollination strategy would be highly advanta- geous In an environment with large spatiotempo- ral variability such as that found in the southern California climate zone, where seasonal and annual variation in pollinator assemblages ap- pears to be substantial. Although visits to flowers of the SFVS do not necessarily correspond to successful pollination events, because of the diversity and frequency of the observed visitors and the fact that the data were only taken on visitors that were actually visiting the stamens and/or came into contact with the stigmatic surfaces, we conclude that the SFVS is not likely to be pollinator limited. Further, it is neither limited by lack of seed production nor seed germination. We conclude with the suggestion that this species may be rare due to the destruction of suitable habitat. ACKNOWLEDGMENTS The authors thank Senta Breden, Mina deLeest, Kelly Garron, Dawn Hendricks, Fern Hoffman, Katie Levensailor, Jamie Miner, Greg Pongetti, Nelida Rojas, Romeo Sison, and Matthew Weekes for field assistance, the staffs at the Ahmanson and Newhall Ranches, Dudek and Associates, Inc., Sapphos Environmental, Inc., and Lenora O. Kirby of the Las Virgenes Institute for their assistance, as well as Roy Snelling, Charles L. Bellamy, Eric Fisher and Steve Gaimari for arthropod identifications. Contract grant support from Newhall Land and Farming Corp., Newhall, California, via Dudek and Associates, Inc., Encinitas, California and from Sapphos Environmental Inc., Pasadena, Califor- nia is gratefully acknowledged. LITERATURE CITED Apis, J. 1979. Problems interpreting arthropod sam- pling with pitfall traps. Zoologischer Anzeiger 202:177-184. ANDERSON, G. J., G. BERNARDELLO, T. F. STUESSY, AND D. J. CRAWFORD. 2001. Breeding system and pollination of selected plants endemic to Juan Fernandez Islands. American Journal of Botany 88:220-233. BARRETT, S. C. H. 2002. The evolution of plant sexual diversity. Nature Reviews Genetics 3:274—-284. . 2003. Mating strategies in flowering plants: the outcrossing-selfing paradigm and beyond. Philo- sophical Transactions of the Royal Society of London. Series B, Biological Sciences 358: 991-1004. 40 BASKIN, C. C. AND J. M. BASKIN. 1998. Seeds: ecology, biogeography, and evolution of dormancy and germination. Academic Press, San Diego, CA. BAUDER, E. T. 2000. Recovery and management of Orcutt’s spineflower (Chorizanthe orcuttiana). Final report. Prepared for the California Department of Fish and Game, South Coast Region. San Diego, CA. BERNARDELLO, G., G. J. ANDERSON, L. S. PATRICIO, M. A. CLELAND, T. F. STUESSY, AND D. J. CRAWFORD. 1999. Reproductive biology of Lac- toris fernandeziana (Lactoridaceae). American Journal of Botany 86:829—840. BEWLEY, J. D. AND M. BLACK. 1994. Seeds: physiology of development and germination, 2nd ed. Plenum Press, New York, NY. BoscH, M., J. SIMON, J. MOLERO, AND C. BLANCHE. 1998. Reproductive biology, genetic variation and conservation of the rare endemic dysploid Delphin- ium bolosii (Ranunculaceae). Biological Conserva- tion 86:57—66. BURDON, J. J., D. R. MARSHALL, AND A. H. D. BROwN. 1983. Demographic and genetic changes in populations of Echium plantagineum. Journal of Ecology 71:667—680. BURGER, J. C., R. A. REDAK, E. B. ALLEN, J. T. ROTENBERRY, AND M. F. ALLEN. 2003. Restoring arthropod communities in coastal sage scrub. Conservation Biology 17:460—467. CALIFORNIA NATURAL DIVERSITY DATA’ BASE (CNDDB). 2001. Chorizanthe parryi var. fernan- dina. California Department of Fish and Game. Sacramento, CA. Unpublished files. CALIFORNIA NATIVE PLANT SOCIETY (CNPS). 2001. Inventory of rare and endangered plants of California, 6th ed. Rare plant scientific advisory committee, David P. Tibor, Convening Editor. California Native Plant Society. Sacramento, CA. . 2005. Inventory of rare and endangered plants (online edition, v6-05a). California Native Plant Society, Sacramento, CA. Accessed from http:// www.cnps.org/inventory CHITTKA, L., J. D. THOMSON, AND N. M. WASER. 1999. Flower constancy, insect psychology, and plant evolution. Naturwissenschaften 86:361—377. COLWELL, R. K. AND J. A. CODDINGTON. 1994. Estimating terrestrial biodiversity through extrap- olation. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 345:101—-118. CONNER, J. K. AND R. NEUMEIER. 1995. Effects of black mustard population size on the taxonomic composition of pollinators. Oecologia 104:218— 224. CRUDEN, R. W., S. M. HERMANN, AND S. PETERSON. 1983. Patterns of nectar production and _ plant- pollinator coevolution. Pp. 80-125 in B. Bentley and T. Elias (eds.), The biology of nectaries. Columbia University Press, New York, NY. DEGEN, A. A., M. GERSANI, Y. AVIVI, AND N. WEISBROT. 1986. Honeydew intake of the Weaver Ant, Polyrhachis simplex (Hymenoptera: Formici- dae) attending the aphid Chaitophorous populialbae (Homoptera: Aphididae). Insectes Sociaux, Paris 33:211-215. DE LA BARRERA, E. AND P. S. NOBEL. 2004. Nectar: properties, floral aspects, and speculations on origin. Trends in Plant Science 9:65—69. MADRONO [Vol. 56 FAEGRI, K. AND L. VAN DER PIJL. 1971. The principles of pollination ecology, 2nd revised ed. Pergamon Press, Oxford, U.K. FREE, J. B. 1963. The flower constancy of honeybees. Journal of Animal Ecology 32:119—-131. FUTUYMA, D. 2001. Ecological specialization and generalization. Pp. 177-192 in C. W. Fox, D. A. Roff, and D. J. Fairbairn (eds.), Evolutionary ecology. Oxford University Press, Oxford, U.K. GIBLIN, D. E. AND C. W. HAMILTON. 1999. The relationship of reproductive biology to the rarity of endemic Aster curtus (Asteraceae). Canadian Jour- nal of Botany 77:140-149. GLENN LUKOS ASSOCIATES, INC. 1999. Report: biology of the San Fernando Valley Spineflower, Ahman- son Ranch, Ventura County, California. Glenn Lukos Associates, Inc., Laguna Hills, CA. GOMEZ, J. M. AND R. ZAMORA. 2006. Ecological factors that promote the evolution of generaliza- tion in pollination systems. Chapter 7, Pp. 145-166 in N. M. Waser and J. Ollerton (eds.), Plant- pollinator interactions from specialization to gen- eralization. The University of Chicago Press, Chicago, IL. GOODMAN, G. J. 1934. A revision of the North American species of the genus Chorizanthe. Annual of the Missouri Botanical Garden 21:1—102. GOULSON, D., J. OLLERTON, AND C. SLUMAN. 1997. Foraging strategies in the small skipper butterfly, Thymelicus flavus: when to switch? Animal Behav- ior 53:1009-1016. HAGERUP, O. 1932. On pollination in the extremely hot air at Timbuctu. Dansk Botanisk Arkiv 81:1—20. HARPER, K. T. 1979. Some reproductive and _ life history characteristics of rare plants. and indica- tions for management. Great Basin Naturalist Memoirs 3:129—137. HEITHAUS, E. R. 1974. The role of plant-pollinator interactions in determining community structure. Annals of the Missouri Botanical Garden 61:675—691. HERRERA, C. M. 1996. Floral traits and plant adaptation to insect pollinators: a devil’s advocate approach. Pp. 65—87 in D. G. Lloyd and S. C. H. Barrett (eds.), Floral biology. Chapman and Hall, New York, NY. HICKMAN, J. C. (ed.). 1993. The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. JENNERSTEN, O. AND S. G. NILSSON. 1993. Insect flower visitation frequency and seed production in relation to patch size of Viscaria vulgaris (Caryo- phyllaceae). Oikos 68:283—292. JEPSON, W. L. 1925. A manual of the flowering plants of California. University of California Press, Berkeley, CA. JOHNSON, S. D. AND K. E. STEINER. 2000. Generaliza- tion versus specialization in plant pollination systems. Trends in Ecology and Evolution 15:140-143. JOHNSON, L. K. AND S. P. HUBBELL. 1975. Contrasting foraging strategies and coexistence of two bee | species on a single resource. Ecology 56:1398—1406. JONES, C. E., J. BURK, F. SHROPSHIRE, L. TAFT, Y. ATALLAH, R. ALLEN, AND L. SONG. 2002. The | pollination biology of the San Fernando Valley Spineflower, Chorizanthe parryi var. fernandina (S. — Watson) Jepson. Final Report. Prepared for the ee 2009] Ahmanson Land Company, Calabasas, California, under contract with Sapphos Environmental, Inc., Pasadena, CA. , S. E. WALKER, F. M. SHROPSHIRE, R. L. ALLEN, D. R. SANDQUIST, AND J. C. LUTTRELL. 2004. Newhall Ranch investigation of the San Fernando Valley Spineflower Chorizanthe parryi S. Watson var. fernandina (S. Watson) Jepson. Final report, prepared for the Newhall Ranch and Farming Company, Valencia, California in consul- tation with Dudek and Associates, Inc. Encinitas, CA. KARRON, J. D. 1991. Patterns of genetic variation and breeding systems in rare plant species. Pp. 87—96 in D. A. Falk and K. E. Holsinger (eds.), Genetics and conservation of rare plants. Oxford University Press, Oxford, U.K. KAYE, T. N. 1999. From flowering to dispersal: reproductive ecology of an endemic plant, Astrag- alus australis var. olympicus (Fabaceae). American Journal of Botany 86:1248—1256. KEARNS, C. A. AND D. W. INOUYE. 1993. Techniques for pollination biologists. University Press of Colorado, Niwot, CO. AND . 1997. Pollinators, flowering plants, and conservation biology. Bioscience 47:297—307. KUNIN, W. E. AND A. SHMIDA. 1997. Plant reproduc- tive traits as a function of local, regional, and global abundance. Conservation Biology 11:183— 12: Lewis, A. C. 1989. Flower visit consistency in Pieris rapae, the Cabbage Butterfly. Journal of Animal Ecology 58:1—13. LUITEN, S. H., J. G. B. OOSTERMEIJER, N. C. VAN LEEUWEN, AND H. C. M. DENNIS. 1996. Repro- ductive success and clonal genetic structure of the rare Arnica montana (Compositae) in the Nether- lands. Plant Systematics and Evolution 201:15—30. MEYER, M. 2003. Back from the brink of extinction. Outdoor California 64:4—7. Munz, P. A. AND D. D. Keck. 1959. A California flora. University of California Press, Berkeley, CA. MuRPHY, A. 2003. Endangered plant conservation: do pollinators matter to the robust spineflower? M.A. thesis. Department of Environmental Studies, San Jose State University, San Jose, CA. OLESEN, J. M., Y. L. DUPONT, B. K. EHLERS, AND D. M. HANSEN. 2007. The openness of a flower and its number of flower-visitor species. Taxon 56:729-736. OLIVER, I. AND A. J. BEATTIE. 1993. A possible method for the rapid assessment of biodiversity. Conserva- tion Biology 7:562—568. AND . 1996. Designing a cost-effective invertebrate survey: a test of methods for rapid assessment of biodiversity. Ecological Applications 6:594-607. OLLERTON, J., A. KILLICK, E. LAMBORN, S. WATTS, AND M. WHISTON. 2007. Multiple meanings and modes: on the many ways to be a generalist flower. Taxon 56:717—728. PETERS, R. H. 1983. The ecological implications of body size. Cambridge University Press, New York, NY. Proctor, M., P. YEO, AND A. LACK. 1996. The natural history of pollination. Timber Press, Incorporated, Portland, OR. JONES ET AL.: REPRODUCTION IN SAN FERNANDO VALLEY SPINEFLOWER 41 PURDY, B. G., J. B. RANDALL, AND S. E. MACDONALD. 1994. Genetic variation, breeding system evolution, and conservation of the narrow sand dune endemic Stellaria arenicola and the widespread S. longipes (Caryophyllaceae). American Journal of Botany 81:904-911. REVEAL, J. L. 1989. The eriogonoid flora of California (Polygonaceae: Eriogonoideae). Phytologia 66:295— 414. AND C. B. HARDHAM. 1989. A revision of the annual species of Chorizanthe (Polygonaceae: Eriogonoideae). Phytologia 66:98—198. ROUBIK, D. W. 2001. Ups and downs in pollinator populations: when is there a decline? Conservation Ecology 5(1), 2. [online] URL: http://www. consecol.org/vol5/iss 1/art2/ RUSSELL, A. D. 2003. Phylogenetic analysis and morphological study of the subfamily Eriogonoi- deae (Polygonaceae) with an emphasis on the genus Chorizanthe. M.S. thesis. Biology Department, San Diego State University, San Diego, CA. SANDQUIST, D. 2003. Investigation of possible fire- related germination cues for San Fernando Valley Spineflower, Chorizanthe parryi var. fernandina (S. Watson) Jepson. Technical report. Prepared for Sapphos Environmental Inc. Pasadena, CA. SAPPHOS ENVIRONMENTAL INC. 2001. An investigation of the San Fernando Valley Spineflower for the Ahmanson Land Company. Sapphos Environmen- tal, Inc., Pasadena, CA. SCHAFFER, W. M., D. B. JENSON, D. E. Hosss, J. GUREVITCH, J. R. TODD, AND M. V. SCHAFFER. 1979. Competition, foraging energetics, and the cost of sociality in three species of bees. Ecology 60:976—987. SCHEMSKE, D. W., B. C. HUSBAND, M. H. RUCKELS- HAUS, C. GOODWILLIE, I. M. PARKER, AND J. G. BISHOP. 1994. Evaluating approaches to the conser- vation of rare and endangered plants. Ecology 75:584—606. SCHMITT, J. 1983. Density-dependent pollinator forag- ing, flowering phenology, and temporal pollen dispersal patterns in Linanthus bicolor. Evolution 37:1247-1257. SIH, A. AND M.S. BALTUS. 1987. Patch size, pollinator behavior, and pollinator limitation in catnip. Ecology 68:1679—1690. SOUTHWICK, E. E. 1984. Photosynthate allocation to floral nectar: a neglected energy investment. Ecology 65:1775—1779. SPENCE, J.R. AND J. K. NIEMELA. 1994. Sampling carabid assemblages with pitfall traps: the madness and the method. Canadian Entomologist 126:881—894. SPIRA, T. P. 1980. Floral parameters, breeding system and pollinator type in Trichostema (Labiatae). American Journal of Botany 67:278—284. STEBBINS, G. L. 1950. Variation and evolution in plants. Columbia University Press, New York, NY. . 1957. Self fertilization and population variabil- ity in the higher plants. American Naturalist 91:337-354. 1970. Adaptive radiation of reproductive characteristics in angiosperms, I: Pollination mech- anisms. Annual Review of Ecology and Systemat- ics 1:307—326. TAYLOR-TAFT, L. 2003. The reproductive biology of the San Fernando Valley Spineflower (Chorizanthe parryi var. fernandina (S. Watson) Jepson (Poly- n%) MADRONO gonaceae). M.S. thesis. Department of Biological Science, California State University, Fullerton, CA. THOMPSON, J. D. 1994. The coevolutionary process. University of Chicago Press, Chicago, IL. TIMMERMAN-ERSKINE, M. AND R. S. BoyD. 1999. Reproductive biology of the endangered plant Clematis socialis (Ranunculaceae). Journal of the Torrey Botanical Society 126:107—116. UNITED STATES FISH AND WILDLIFE SERVICE. 2001. Endangered and threatened Wildlife and plants; reopening of public comment period and notice of availability of draft economic analysis of proposed critical habitat determination for the Chorizanthe robusta var. hartwegii (Scotts Valley spineflower). Federal Register 66:48227—-48228. [Vol. 56 VOGLER, D. W. AND S. KALISZ. 2001. Sex among the flowers: the distribution of plant mating systems. Evolution 55:202—204. WASER, N. M. 1986. Flower constancy: Definition, cause, and measurement. American Naturalist 127:593-603. , L. CHITTKA, M. V. PRICE, N. M. WILLIAMS, AND J. OLLERTON. 1996. Generalization in polli- nation systems, and why it matters. Ecology 77: 1043-1060. WorRK, T. J., C. M. BUDDLE, L. M. KORINUS, AND J. R. SPENCE. 2002. Pitfall trap size and capture of three taxa of litter dwelling arthropods: implica- tions for biodiversity studies. Environmental En- tomology 31:438-448. MADRONO, Vol. 56, No. 1, pp. 43-48, 2009 ON THE RELATIONSHIP OF STREPTANTHUS VERNALIS AND STREPTANTHUS BARBIGER (BRASSICACEAE) RICHARD O’DONNELL 1317 Cornell Ave., Berkeley, CA 94702 dickodonnell@earthlink.net ABSTRACT The relationship between Streptanthus vernalis and Streptanthus barbiger (including the taxonomic synonym Mesoreanthus fallax) is examined. Morphological and allozyme comparisons show that the species are unrelated. Key Words: E. L. Greene, Mesoreanthus, Streptanthus vernalis, allozyme analysis. Streptanthus vernalis O’Donnell and Dolan is a newly described species from Three Peaks in Lake County, CA (O’Donnell and Dolan 2005). In the paper describing this species, O’Donnell and Dolan (2005) compared S. vernalis morphol- ogy to that of sympatric Streptanthus species and to S. batrachopus J. Morrison, a species known only from Marin County, CA (S. batrachopus was included because some observers of S. vernalis thought that it was S. batrachopus). We also performed allozyme analysis to measure the genetic distance between S. vernalis and other Streptanthus species. The possibility that S. vernalis is Mesoreanthus fallax E. Greene has been raised (Dr. Dean Taylor personal communication 2006). Mesor- eathus fallax is a taxon segregated from 5S. barbiger E. Greene by Greene because of perceived morphological differences, examined below. Streptanthus barbiger was not included in the comparisons in O’Donnell and Dolan (2005) because we did not consider it to be related to S. vernalis, and there was no population of S. barbiger within 15 kilometers of the only known S. vernalis population. However, to address the possibility that S. vernalis is M. fallax, in this paper I compare S. vernalis to S. barbiger. STREPTANTHUS BARBIGER Streptanthus barbiger, a widely distributed but uncommon serpentine endemic of the inner Coast Range of northern California, was first described by Greene (Greene 1888) based on a specimen collected in June 1888 by A. B. Simonds from Highland Springs, Lake County, CA. The specific epithet barbiger (bearded) presumably referred to the “short bristly white pubescence”’ which was perceived on the tips of the sepals. Greene supplemented his description of S. barbiger (Greene 1894), adding that the entire calyx, not just the sepal tips, was ‘“‘commonly bristly-hairy, but often glabrous’. He noted elsewhere that it was common in Napa Co., specifically at Muiravalle, an estate west of the town of St. Helena. Later, Greene split S. barbiger into three species under the generic name Mesoreanthus (Greene 1904, 1903-1906): (1) M. barbhiger, the “type species’, with hirsute sepals; (2) M. fallax, a glabrous, glaucous species collected in July 1891 in the hills above St. Helena in the Napa Valley; and, (3) M. vimineus, also glabrous and glaucous, which Greene described as a showier plant with large white blossoms. Mesoreanthus vimineus was collected by C.F. Baker in early May 1903 near Lakeport. The specimens of M. fallax that Greene used to describe that new species were fruiting and had few blossoms and no leaves; thus, his description 1s sketchy and incomplete. Perhaps it was their lack of leaves that prompted Greene to apply the specific epithet fallax to the poor specimens he found: fallax means “unclear, deceptive’. A topotype of M. fallax in the California Academy of Sciences herbarium (Greene 16323, CAS 296902) appears to be typical S. barbiger: the basal leaves are oblanceolate, entire, with a purple cast, and the cauline leaves are linear. The calyx is almost glabrous. However, all four of the white petals are marked with purple where usually only the two lower are so marked. Only one other M. fallax accession known to the author is Greene’s of 1891 in the Harvard Herbarium (Barcode 40579); one specimen in the University of Califor- nia Herbarium of S. barbiger (UC10858) has been annotated M. fallax, annotator unknown. Streptanthus barbiger is described by Buck et al. (1993) in The Jepson Manual as a variable species, a characterization confirmed by the significant variation in descriptions of S. barbiger since it was first described, as indicated on Table 1. Buck et al. (1993) may have found additional justification for their description be- cause they lumped unidentified plants from Tehama and Lake Counties under S. barbiger. A coauthor of Buck et al. (1993) says he “lumped several things under Streptanthus barbiger E. ee [Vol. 56 (ysnTyM Apo1e.1) ajdind Ayyje1sues JOMO] SYSHIYM —/+ ulsAprur ofdind yim Jied I9Moy ‘ojTYM ired taddn ‘podstio oo earl jo 19}uU99 Woy A[IeoU SuIUUNI MPD WOLT pueq odind yyWM oy ied I9MmoyT SouIyM ired taddn ‘Ajao1e9os 10 podsiio jou ‘popunol pospo ou1yM wu Z jaddn ‘wut g[-¢ jou ‘suoT, Wd | Noe quuly ‘suoy soul] ¢ ‘pol-ylep “[[eus ov1yM ‘poxapor A[SUOMS “OTYM s[ei1og sousdsoqnd SuO] WU /—9 WYN} ayyM ApStiq jeorde ou jnq sirey JOYS B YIM snoiqe[s —/+ JJOS YIM SoUIToUIOS sjejod Surpsdoxd JO =F poop sd ayy Ajjetoues ‘(ystdind) ‘sdij poamnoor pue sd} poammosol Ayjenba ‘poaimool UVM O1TyM MO]TJOA YSTUdoIs SUISIeU UO oUTTeAY SnOUvIqUISUT 10 pue surpeoids YsIusois ‘oyITe sd pieMmo} Ayje1oues ‘UU Q—¢ ‘ystjdind 0} uso13 YsoIyM YUM YSsTusoIs ‘suo] AroA sdy puev [enbo Ajivou MO]T[OA “Ud0183 “UU /—-9 jedas BuO] = — — [Jews ‘aTIssosqns Soul] p ‘aTISsasqns SNOIgRIS ‘[eIpeIIg xATea suldselo —/+ SOUITJIUIOS ‘ITISSAs Ayjei1oues ‘ommua —/+ poddi-asuevio ® Ayjetouas ‘1eouly, —/+ ajoned y10ys B 0} ‘QTISSOS “SITUS S1BOUT] Z ‘UD [[> SdAed] UMOP UMPIP JO ‘aTISsas Iaaou jnq 2 sulnes 1oddn pue ‘ayeNOLINe [eV 1B JOU ayeTosouR] A[MOIIeU a a[pprut ‘poroned IO ‘aseq 78 d}R[NILINe soulynes ‘poddn < ‘ayejUap A[asIvOS ‘a7e]UNP IdMOT 94} -a8URIO 99} [voIde 2 Ajjetouas ‘9}eA0gO poonpel IO ‘dIIVUd ‘d} e[OSNUP] ‘tut | sored “wd p-¢ AJAPIM 0} d}vTODNULTqGO ‘molieu AOA Joddn -IROUI] Pd}eSUO]I O} ‘yugfnoons ‘MOTaq o[dind ‘UD (> SdapRIq IaMO] ‘SUO] WD Q[—9 IOMOT Ivaul] Sopeyq oy} ‘suoy| ‘OA0GR UdIdIB *9]VUdIO ‘pajjasol Jou —/+ [eseq ‘o1NUd ‘jOo1O ‘RoUTTqns (WD $Z’SI-S'Z) ‘Ul 9-] umouyun ainus ‘resuly = Ayyearde “repnsiqso :;eseq jeryy xA[BO jdao0x9 noysno1y} SUO] os A[poysreu snoiqes wu ¢—Z 10919 sjaoIpod ud}JO ‘puNndas ‘asoo| xe] pue ‘oy e[norued puosas ‘SUO] UD G7T—C DUIS XP] apy eB Apso saura0e1 suo] soyids Sunmay -dSOUTDNBI ATISOO] SOUITJOUOS ‘ASOUMIIOVI IOUDOSOIOTJU]T aseq aaoqe wo SuIpoURIq ynoysnosy} poyouriq ‘49019 aseq dy} dsAOQgK WO AJaa1j ‘Snooneys MO]Oq WO, poyoueig poyouriq ATpe1ouas ‘snooneys ‘snoiqge[s soyoueig }O019 YIM UW9}s ‘SnoIge[s “1opusys Iapugys ‘19019 Io adums Ayyensn 1qeH wp $-1 wp lt (up L-E~) We ST CTI (up 9-¢€~) UTI (WIS Cg) “Ul TI Wd QT-T —- WYSIOY Jue jenuue Ajng—Avyy ‘Tenuue Ajng—oung ‘penuue Ajng—j, ‘Tenuue jenuue ARI -Yolepy “Penuue potsad SULIOMO]-] (C661) ‘Te 19 Yong (6S61) ZUNI 9€61 Uosdor pO6T SUseIH (8881) 9US9IDH (SOOT) UPTO 7% [TIUU0d.O 4as81gAbg SnyJUvIda4jg —- dabgabg snyjuvi dag 4d81GADG SNYIUDIdIAS xDyo{ snyjuvasosapy sasigapg snyjuvidals sypusaa SNYJUDIAIALS ‘MADIGU VA SQHINVLdANLS AGNV SITFNYFA SQHINV.LddaALS AO SHALOVUVHOD TVOIDOTOHdYOW AHL AO NOSTYVdNO) “T FTaV EL 44 O’DONNELL: RELATIONSHIP OF STREPTANTHUS 45 2009] pud duo 18 IO Q ‘SUIM SP99S UI9M19q poMmolIeUu —/+ “PpdaIno —/+ wo /-z ‘sulpeoids poonpor aired 1sddn ‘s1oyyuRe Spasnj tied 1addn :syuoweyly (£661) ‘Te 19 Yong AdB1GADG SNYJUDIAIALS posulm JOU opis WU ¢'] jnoqe “sUuOT WD / —¢ dISSOs ‘poaAamnsos yeyYMOWOS ‘Sulpeoids 2]eUU09 Jaddn :sjusweyly (6S61) Zun| Aasigdpg snyjuvj dads SSoTUIS1eUW ‘passoiduios posuIm jou IVI “[BAo — posulm ‘osurso pos sjeorpod surpeoids yAOYS AIDA uO SpIeMUMOp PoaIns ‘SUO| dSO[NIO} (UID Qg°¢) “Ul CT AJaINISqO !paAino ‘asO[N10} APYSI]S Anysts ‘surpeaids ‘possaiduios poamnodal dSO[NIO} POXd[Jo1 10 poxajol ‘MOLIVU AISA ‘Ieouly A[MOLIVU ‘Wd ¢—p 0} dn ‘yoOI9 onbiIsg HIYM “9181118 eS -IBdUT] :S.1dq]Ue ‘VULNS 9} 0} [jews ysoulye poyun povlosul AoA aired 1addn Jred ysoutoddn “Id1J (JSOMOT SO[PPIul 0} sToyUe YIUUUUINS 0} oy} ‘ojdind posny pdajiosul ‘ay;pprut Aemy[ey poywun ared ssddn ylep :syuoWeyly {POAINIII “paytosxa powun 19MoyT :do} 78 ‘sjuaUIe]Ly ‘spedos ‘sired penboun ‘ayeuUuods A[peoig poyun saddn :sjuowepy 0} yenbo sted 1zsddn AIQA 9014} Ul ysouttoddn :sited sary} SUSUIRIS 9€6] Uosdos PO6T USED (8881) SUs2IH (SOOT) URTOd 7 Teuu0d.O 4asiqgdpg Snyjuvj dass XDD snyluvasosaPy dasigipg snyjuvj dass sypudaa snylUudjdaas ‘GHNNILNOD ‘| ATAVE 46 MADRONO Greene” in his treatment of the species (Dr. Dean Taylor personal communication). Thus, S. barbi- ger as described in Buck et al. (1993) is probably not a single entity but rather a collection of entities. Table | compares S. vernalis to S. barbiger (including the synonymous M. fallax) as it was described by Greene and subsequent authors. Table 1 retains the language from the descrip- tions of S. barbiger to preserve the intentions of each author as to diagnostic characters. This results in some cases in different measurements of the organs (e.g., centimeters, inches, lines) and different characters being featured. Notwith- standing the diversity of descriptive choices that the authors from Greene to Buck et al. (1993) have made, the morphological differences be- tween S. vernalis and S. barbiger are clear. COMPARISON OF MORPHOLOGY The treatments of S. barbiger shown in Table 1 indicate a wide range of variation in important characters: height, calyx vestiture, and the purple veining on the lower petals. The siliques are variously described as recurved/reflexed and spreading, and spreading only; the torulose character of the silique has remained faint. The branching pattern varies widely, from branching from the base, to branching above, to generally branching; there appears to be no consensus on the branching. As for the leaves, the authors focus on the narrow, linear cauline leaves. Jepson (1936) and Buck et al. (1993) mention the lower cauline leaves which are long, wide and crenate (or coarsely dentate); none of the authors indicate that these leaves are often blotched with grayish irregular marks. Specimens observed by the author near the Highland Springs Road in 2006 possessed bristly calyces, white upper petals, the lower pair marked with purple veining, and recurved siliques. Some of the individuals possessed long, crenate basal leaves, and all possessed linear, almost filiform cauline leaves. In May 2007, a drought year, the author found an early flowering population of S. barbiger near the Highland Springs Reservoir. This population was generally shorter than other populations observed in nearby locations in previous years (<13 cm) and its other organs proportionately smaller than those found a few miles up the Highland Springs Road in 2006. The basal leaves were erect, clustered at the base and were entire or coarsely dentate; the cauline leaves were linear, filiform. The flowers, however, were typical of S. barbiger; 1.e., upper petals white, lower petals white with purple veining. The author also observed a population of S. barbiger on Spring Mt. Road west of St. Helena in July 2007. The plants of this population had glabrous calyces, [Vol. 56 lax racemes, recurved siliques and, except for the basal leaves, which were no longer present, were otherwise identical to those found along the Highland Springs Road in 2006. The author cannot confirm that this Spring Mt. Road population was that from which Greene sampled, or the source of the specimen in the California Academy of Sciences herbarium. Serpentine habitat in the area, including the former Mir- avalle estate, west of St. Helena is limited to a narrow band surrounded by volcanic substrate. Since Greene’s time, the area has been planted with vineyards. It is possible that vineyards or other construction have replaced the S. barbiger populations that Greene observed. The comparisons in Table 1 show that S. vernalis does not resemble S. barbiger. S. vernalis has no purple color in its petals or sepals, and the silique is erect and distinctly torulose. The seeds are winged. The leaf morphology of specimens of S. barbiger from Napa, Lake and Sonoma Counties in the Jepson Herbarium and at the California Academy of Sciences herbarium is very distinct from that of S. vernalis. The basal leaves of the former are usually oblanceolate and deeply crenate (sometimes entire) and the cauline leaves are entire, linear and sometimes filiform. This is true for the cauline leaves of specimens observed by the author in the hills west of St. Helena (Spring Mountain Road), the type locality of M. fallax, and for the specimens from Highland Springs, the type locality of S. barbiger. The leaf characters of these specimens are entirely different from those of S. vernalis. Streptanthus vernalis does not resemble the > topotype of M. fallax in the herbarium of the California Academy of Sciences or M. fallax as_ Greene described it. S. vernalis does not have the very long spreading sepals reported for M. fallax. | Where M. fallax is described as having dark-red | lower petals edged with white, the lower petals of | S. vernalis are white, suffused with a yellow tint and the upper petals are pure white (pers. obs.). Greene describes the fruiting racemes of M. fallax: as “long and lax”; the fruiting branches of S. vernalis are short and retain their ascending and erect stature (pers. obs.) TAXONOMIC STATUS OF MESOREANTHUS FALLAX’ After Greene segregated the glabrous species. M. fallax from the “‘type species” M. barbiger, M. — fallax was not used. In the Flora of Middle Central California (Jepson 1911) Jepson lists S.. barbiger but does not refer to M. fallax. Much. later in Vol. 2 of the Flora of California (Jepson 1936), Jepson refers to both M. fallax and M. vimineus as synonyms of S. barbiger as does. Abrams (1944). It appears that the glabrous calyx that had distinguished M. fallax from M. barbiger was assimilated into S. barbiger. Only 2009] TABLE 2. NEIV’S UNBIASED GENETIC IDENTITY (NEI O’DONNELL: RELATIONSHIP OF STREPTANTHUS 47 1978) VALUES FOR PAIRWISE COMPARISONS OF STREPTANTHUS VERNALIS, S. BREWERI SSP. BREWERI, AND COLLECTIONS FROM THREE DIFFERENT LOCATIONS FOR S. BARBIGER. S. vernalis S. breweri S. barbiger 1 S. barbiger 2 S. vernalis * * x x S. breweri 0.600 * ‘i * S. barbiger 1 0.298 0.526 * 53 S. barbiger 2 0.365 0.593 0.992 : S. barbiger 3 0.204 0.408 0.891 0.842 Munz and Keck (1959) and Abrams (1944) mention the occasional hairy calyx in S. barbiger. Jepson (1936) remarked: Greene (Fr. Fl.) describes the calyx as ‘‘bristly-hairy’’, an apparently unusual char- acter for S. barbiger as we now know it. The calyces sometimes exhibit parallel hairlike lines in relief which imitate appressed bristles. However, the entire breweri-barbiger-niger series 1S unusually productive of variables, even for Streptanthi No Streptanthus species with the characters of M. fallax appears to have taken its place. The only specimens in herbaria attributed to M. fallax in the Missouri Botanical Garden’s TROPICOS database is Greene’s of 1904/05. It appears that systematists since Greene have not recognized M. fallax as a separate species. While the name may be conserved, it cannot be applied to S. vernalis; morphological and allozyme evidence (see below) against such application is unambiguous: S. vernalis is not M. fallax. In fact a M. fallax was recently (Al-Shebaz and Taylor 2008) relegated to the synonymy of a new species S. vimineus A\I- Shabaz & D. W. Taylor, and as no claim is made that S. vinimeus is the same as S. vernalis the question is moot. To complete the reassignment of the genus Mesoreanthus I use this opportunity to consign the name M. barbiger to the synonymy of S. barbiger. MOLECULAR ANALYSIS Allozyme analysis was conducted by Dr. Rebecca Dolan (written comm. July 2007) on samples from three S. barbiger populations (Barbiger 1—Highland Springs Reservoir; Barbi- ger 2—Highland Springs Road about 2 mi west of Highland Springs Reservoir; and Barbiger 3— Spring Mt. Road), one S. breweri ssp. breweri population, and the sole population of S. vernalis following the procedures of Dolan (1995). Pair- wise values of Nei’s genetic identity (Nei 1978) in this analysis show S. vernalis to be most similar to _S. breweri and very distant from the samples of S. barbiger. The three S. barbiger samples are very similar, with an average identity value of 0.91. However, the S. barbiger populations in the vicinity of Highland Springs are more closely allied to each other than either is to the Spring Mountain Road population. The data in Table 2 supplement the data in the original paper (O’Donnell and Dolan 2005). Notwithstanding the similarity to S. breweri ssp. breweri shown in Table 2, S. vernalis remains most closely allied to S. morrisonii, a sympatric species, as shown 1n the original paper. CONCLUSION Streptanthus barbiger and Streptanthus vernalis are separate species. Morphology alone is suffi- cient to distinguish the two species. Molecular analysis supports the distinction, and points to different origins for the two species. ACKNOWLEDGMENTS The author thanks Dr. Rebecca Dolan for perform- ing the molecular analysis used in this paper and for tabulating the results and for her comments on several early drafts. He also thanks Dr. Steve Edwards for his review and comments on the final draft. LITERATURE CITED ABRAMS, L. 1944. Hlustrated flora of the Pacific states, Vol Il. Stanford University Press, Stanford, CA. AL-SHEBAZ, A. AND M. S. MAYER. 2008. New or noteworthy Streptanthus (Brassicaceae) for the Flora of North America. Novon 18:279—282. BUCK, R. E., D. W. TAYLOR, AND A. R. KRUCKE- BERG. 1993. Streptanthus. Pp. 439-444 in J. Hickman (ed.), The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. DOLAN, R. W. 1995. The rare, serpentine endemic Streptanthus morrisonii (Brassicaceae) species com- plex revisited using isozyme analysis. Systematic Botany 20:338—346. GREENE, E. L. 1888. New or noteworthy species. Pittonia 1:217-218. . 1894. Manual of the botany of the region of the San Francisco Bay. Cubert & Co., San Francisco, CA. . 1903-1906. Leaflets of botanical observation and criticism, Vol |. Washington D.C. . 1904. West American Cruciferae. Leaflets of Western Botany 1:81—91. JEPSON, W. 1911. A flora of western middle California. Cunningham, Curtiss, and Welch, San Francisco, CA. 48 MADRONO [Vol. 56 . 1936. A flora of California, Vol 2. Califor- NEI, M. 1978. Estimations of average heterozygosity nia School Book Depository, San Francisco, and genetic distance from a small number of CA. individuals. Genetics 89:583—590. Muwnz, P. AND D. KECK. 1959, 1968. A California flora) O’DONNELL, R. AND R. DOLAN. 2005. A new species and supplement. University of California Press, of Streptanthus (Brassicaceae) from Three Peaks in Berkeley, CA. Lake County, CA. Madrono 52:202—206. MADRONO, Vol. 56, No. 1, pp. 49-56, 2009 TO CALIFORNIA WITH JEPSON’S “PHYTO-JOGS” IN 1913 RICHARD G. BEIDLEMAN University and Jepson Herbaria, University of California, Berkeley, CA 94720-2465 ABSTRACT The Second International Phytogeographical Excursion of 1913 was initiated by University of Chicago’s Henry Cowles and represented the first international party of plant geographers to visit the United States. Among the twelve Europeans who participated were Cambridge University’s Alfred Tansley, first president of the British Ecological Society (and later to coin the word “‘ecosystem’’), and Adolf Engler, taxonomist and biogeographer of Berlin’s Royal Botanical Gardens. Arriving in New York City at the end of July, the excursionists (nicknamed the “‘Phyto-jogs” by Willis Jepson) traveled by train to the West Coast, with stops in a variety of ecosystems. After arriving at San Francisco Bay from the Pacific Northwest in early September, the Phyto-jogs, guided by University of California’s Jepson, proceeded to Yosemite National Park for field trips and a famous group photograph among Mariposa Grove’s Giant Sequoias. On September 12 in Oakland, several of the European scientists presented major lectures at the California Botanical Society’s first annual banquet. The following days included visits to sites of local interest, including Burbank’s experimental garden, Mount Tamalpais and Muir Woods with Alice Eastwood, and the Carnegie Institution’s Coastal Laboratory at Carmel. The last two days covered the California desert, including the Salton Sea, and Tucson’s Carnegie Desert Laboratory within the Sonoran Desert. A year later Professor Tansley, reflecting on the phytogeographical diversity of America and the varied researches of its many energetic ecologists, wrote: “In that vast field of ecology America has secured a commanding position.” Key Words: California Botanical Society, Alice Eastwood, Adolf Engler, International Phytogeo- graphical Excursion of 1913, Willis Jepson, Mt. Tamalpais, Alfred Tansley, Yosemite. It was indeed a memorable congregation of men and women gathered for a photograph at the base of the General Sheridan Tree, a Giant Sequoia (Sequoiadendron gigantum) in Yosemite National Park’s Mariposa Grove. The year was 1913, the date September 9. The special occasion was a photo opportunity for, as Professor Jepson nicknamed them, “The Phyto-jogs” (Jepson 1913a, p. 201), members of the first international plant geography tour to visit North America, including California: ““The Second International Phytogeographical Excursion’’. This photo- graphic memorabilia first appeared in print just a tad over ninety years ago, on page 12 of the first issue (1916) of Madrono, journal of the newly formed California Botanical Society (Fig. 1). It is easy to pick out three of the women in the group photograph, wearing jaunty white hats, two with white blouses and long skirts. Edith Tansley on the far left side of the picture was the wife of eminent plant ecologist Alfred G. Tansley from England’s Cambridge University Botany School, seated below her with hands clasped. This very year Professor Tansley became presi- dent of the world’s first ecological society (British Ecological Society), and two decades later (1935) he would introduce an important new word and concept into ecology: “‘ecosystem’’. Next to Mrs. Tansley is Dr. Marie Brockmann-Jerosch, whose botanist husband Henryk (above and to her left, wearing a hat) and she were from Zurich, Switzerland. Dr. Edith S. Clements is on the opposite side of the group, below her husband Professor Frederic E. Clements (he with black tie, black hat band, and book in hand), botanist from the University of Minnesota who was gaining fame for his array of new ecological ideas and terms associated with plant communities and succession. Edith Clements herself would later lay her claim to West Coast fame, writing a lengthy article for National Geographic (May issue, 1927) on “Wild Flowers of the West’, profusely illustrated with her beautiful color paintings, to be embellished into a book in 1928, Flowers of the Coast and Sierra. The distinguished looking, rotund gentleman on Edith Clements’ right, with white waistcoat over his paunch, is Dr. Adolf Engler, plant classifier, editor, and biogeographer, from the Royal Botanical Gardens and Herbarium in Berlin. His system for taxonomically arranging herbaria was used worldwide for years, as was the monumental Vegetation der Erde that he edited. Beside Engler is Professor Carl Schroter of the Swiss Institute of Technology’s Botanical Muse- um in Zurich, with long gray beard and mustache, his fedora’s brim pushed up, looking more like an old mountain man than a botany professor. Between Schroter and Mrs. Brock- mann-Jerosch, with arms folded, is University of Chicago’s plant ecologist Dr. Henry C. Cowles, the editor of the Botanical Gazette, who like Clements was developing ideas on plant succes- sion, especially based on his work with Lake Michigan sand dunes. 50 MADRONO Fic. 1. Park, California. Among others were Professor Carl von Tubeuf of Munich’s Forestry Research Institute (kneel- ing on Professor Tansley’s right, with a drooping handle-bar mustache to match Schroter’s); tall Dr. Carl Skottsberg of the University of Uppsala, Sweden, on Mrs. Clements’ left, with forearm raised, and possibly above her University of Amsterdam’s Dr. Theodoor J. Stomps. Sitting two to the right of Dr. Tansley in the photograph is Dr. George D. Fuller (University of Chicago), with Dr. Eduard Rubel of Zurich behind him, and seated to Fuller’s left Dr. Ove Paulsen of Copenhagen’s Botanical Museum, Denmark. Behind Edith Tansley is moss authority Dr. George Nichols of Yale University; and mostly obscured behind Skottesberg, Dr. Harvey Hall, co-author with his wife Carlotta of A Yosemite Flora (1912). Elsewhere is Professor Jean Massart of Belgium’s Botanical Museum. Meanwhile, front and almost center, sitting on the ground to Tansley’s left, in boots, suit coat, white shirt and black tie, holding his western hat by one fist, is U.C. Berkeley’s botany professor Willis Jep- son, “Chairman of the California Committee of Arrangements” and first president of the new California Botanical Society The First International Phytogeographical Excursion, organized by the British Vegetation Committee and held in Great Britain in August, 1911, proved to be a great success. One partic- ipant, American botanist Cowles, was inspired to [Vol. 56 Sa Group Photograph of Phyto-jogs at The General Sheridan Tree, Mariposa Grove, Yosemite National organize a similar international excursion to introduce European botanists to the American flora and become familiar with some of the New World plant ecologists and their research. This 1913 excursion would start on the East Coast and cross the country by train to the West Coast, with a number of appropriate stops, and returning through the Southwest (Tansley 1913, 1914; Nichols 1914). As it turned out, there would be nearly two hundred participants, including the dozen foreign visitors, the American phytogeo- graphical hosts, and researchers at the various stops. Almost a year ahead of time Cowles had contacted professors Willis Jepson at the Univer- sity of California (Berkeley) and LeRoy Abrams at Stanford University about the California visit. Soon a committee was formed to put together a California itinerary that would appeal to bota- nists interested in the developing fields of plant ecology and biogeography. The premier visit would be to Yosemite National Park and then to one of the coast Redwood areas. With respect to the latter, the Stanford botanists—George James Pierce, William Cooper, and especially Abrams—favored Big Basin (California Red- wood Park) and, understandably, the Stanford campus, while the Berkeley and California Academy of Sciences savants favored Muir Woods and Mt. Tamalpais. There were sugges- tions for a bayside brine pond tour by Professor 2009] Pierce, and a marine algal field trip led by Professor Setchell of the University of California at Berkeley. James Barr, manager of the forth- coming (1915) Panama-Pacific International Ex- position at San Francisco, optimistically hoped a presentation could be made for the “‘distin- guished visitors”, extolling the proposed attrac- tions of the forthcoming exposition, to drum up interest in international attendance. However, this presentation never materialized. After San Francisco, the excursion was to proceed to the Carnegie Institution’s new Coastal Laboratory at Carmel, with a return east by way of the Carnegie Institution’s Desert Laboratory in Tucson, Ar- izona. Now at last, in New York City during the closing days of July, 1913, many members of the European traveling party convened and were armed with six detailed sectional programs describing the impending continental tour. Short- ly they were whisked off to see an “edaphic prairie’ on Long Island, the pine barrens in New Jersey, stop-offs at the botanical garden in Brooklyn and the more famous one in the Bronx. Then the participants were on their way to Chicago for a welcome by University of Chica- go’s botanist John M. Coulter, field trips to remnant low and high prairie, a beech-maple woodland, and exploration of the Lake Michigan sand dunes where Cowles had pursued his research. Early the second week of August the excursion arrived in Lincoln, Nebraska. Greeted by Pro- fessor Charles Bessey and a temperature of 108 degrees in the shade, there was the Commer- cial Club’s afternoon tour of the Lincoln Prairie, accompanied by Nebraska’s governor, followed by Dr. Raymond Pool’s lantern-slide lecture on Nebraska sand hills vegetation. In mid-August several days were spent around a short-grass, high-plains agricultural experiment station near Akron in southeast Colorado, hosted by Colo- rado ecologist Dr. Homer Shantz and joined there by Frederic Clements and his wife. Next was the Pikes Peak region, first with a day in the unique ponderosa pine “Black Forest”, which extends eastward on an elevated sedimentary plateau surrounded by plains grassland. The following week was with the Clements at Minnehaha-on-Ruxton summer resort, their mountain ecology research retreat (the Alpine Botanical Laboratory), alongside the famous Pikes Peak Cog Railway several miles up Engelmann Canyon (named after famous bota- nist George Engelmann of St. Louis). Here during the week the Phyto-jogs were joined by Dr. Engler from Berlin. Next, aboard the Denver and Rio Grande train it was across the continen- tal divide at Tennessee Pass, through the Great Basin shadscale desert, and up Price River Canyon to Salt Lake City. There, after The BEIDLEMAN: JEPSON’S PHYTO-JOGS 5] Commercial Club’s luncheon, a visit through salt- flat brushland ended with a quick saline dip at Saltair Beach alongside the Great Salt Lake. Sunday, August 24, was spent to the south in the Tooele Valley, where Department of Agriculture botanists had carried out an intensive floral study just a year earlier. The next week involved railroading through sagebrush and forest lands to the state of Washington, especially around Mount Rainier, then into Oregon by train on September 2. The three days at Medford, hosted by the Commercial and University clubs, included a caravan of seven private autos on a junket into coniferous forest surrounding volcanic Crater Lake, and a tour of Medford’s orchards, famous for their pears. On September 6 the Southern Pacific Rail- road’s “Shasta Limited”, bound for California, departed south into the Siskiyou Mountains, through the Siskiyou Tunnel; and while dusk fell, “caught our last glimpse of the magnificent snow- covered summit of Mount Shasta” (Tansley 1914, p. 271). At dawn the Phyto-jogs were to look out the train’s western windows onto the shoreline of San Francisco Bay. On Sunday morning, September 7, 1913, the party of European and American natural scien- tists were greeted by Professor Jepson at the 16th Street train station and headed on for breakfast at the pier. For each person, a California excursion program had been prepared, including the travel schedule as well as detailed botanical and geological information. The Southern Pacific Company had put together a special scenic folio, “T.P.E. in America”, individually embossed with each participant’s name, containing twenty large, beautiful photographs “showing some of the more remarkable features of the State’, including Mt. Shasta, Muir Woods, Yosemite’s El Capitan, and Palm Canyon in the Colorado Desert foothills (Jepson 1913b, no. 527). The original trip itinerary had been so specific that it listed five-minute arrival and departure times for some stops. Naturally the fine-tuned schedule soon became obsolete. Monday, the first full excursion day in California, was to be spent touring the Mount Tamalpais region. But an eleventh-hour change of plans “‘threw out of joint” the program Jepson had meticulously prepared for the printer. Instead, immediately after Sunday breakfast the entourage departed by train for El Portal, gateway to Yosemite National Park (Jepson 1913a, 1913-1914). In the hot, dry San Joaquin Valley there was a lunch stop at Tracy. Transferring to the Yosemite Valley Railroad at Merced, enthusiasm height- ened with the entry into Merced River Canyon and the changing foothills panorama of blue oak (Quercus douglasii), gray pine (Pinus sabiniana), California black oak (Quercus kelloggii), and a patchwork of buckbrush (Ceanothus cuneatus), 52 MADRONO manzanita (Arctostaphylos) and chamise (Ade- nostoma fasciculatum). Next morning at the El Portal Hotel the Yosemite Transportation Com- pany’s largest (seventeen-passenger) horse-drawn stage and a smaller one picked up the excursion members, and three hours were spent, guided by Willis Jepson and Harvey Hall, touring Yosemite Valley, with a lunch stop at historic old Sentinel Hotel near Yosemite Falls. Then it was into two small stages of the Big Tree Line, heading south with time for a photo opportunity of the valley from Artist’s Point, on past Chinquapin Junction as the sun set, and to the Wawona Hotel for the night, pausing “‘occasionally for a few minutes at points of interest.” Tuesday was the all-day trip southeast to the lower and upper stands of giant sequoias in Mariposa Big Tree Grove. At first “sight of the great giants”, Dr. Schroter “‘took off his hat and waved it wildly!” (Jepson 1913-1914, p. 2). Here four leisurely hours were spent in the impressive forest. Jepson, anticipating that the Phyto-jogs would cherish a diminutive sequoia cone, un- reachable from the high branches, had arranged with “Mr. Zeus, a Greek citizen, and coworker with Mr. Franklin and Mr. Farraday, to strike gently the top” of the Indiana Tree, to provide cones for inspection and collection (Jepson 1916, p. 14). At the General Sheridan sequoia near Big Tree Cabin, an 1885 replica of grove-discover Galen Clark’s old homestead, official Mariposa Grove photographer Baxter took the memorable group picture. In the lower grove, eleven Phyto- jogs clambered atop the base of The Fallen Monarch for another picture, matching the one in their Southern Pacific Company’s scenic folio portraying the U.S. Army’s Sixth Cavalry lined up on and alongside the Monarch with many of their horses. Late afternoon came the return to Wawona Hotel and a welcome night’s sleep. The following day, September 10, the troupe was off to the Yosemite high country at Glacier Point, through beautiful stands of tall California red firs (Abies magnifica) with an array of late- season Sierran flora which aroused “‘the some- what jaded spirits of the party” (Jepson 1913a, p. 193). After a luncheon along the Glacier Point roadside near Peregoy Meadow, there was a little time for the botanists, who had received special permission from the national park to make plant collections. Jepson was amused to watch the participants at their respective pursuits: Schroter and Engler were zealous collectors; Clements would pick up “‘scraps,”’ ask Jepson or Hall to identify the plants, then throw the scraps away. Administrator Cowles “‘does not collect at all” (Jepson 193l1c, p. 193). Reaching Glacier Point in mid-afternoon, select stalwarts scaled the nearby granite summit of Sentinel Dome with its interesting xerophytic plants, Engler climbing “‘nearly to the top but not [Vol. 56 quite.”’ Choicest discovery was “the rat-tail, Stellariopsis” (now Ivesia santolinoides, its slen- der, silvery, densely haired leaves worthy of today’s common name, mousetail Ivesia). Schr6- ter declared ecstatically: “‘that species alone was worth the trip” (Jepson 1913-1914, p. 5). On September 11, after over-nighting at the Glacier Point Hotel, many of the party com- menced the twelve-mile hike from Glacier Point down the I[lilouette Trail past [lilouette Falls, Nevada Falls, Vernal Falls, and finally onto the valley floor and east towards Camp Curry. Robust Dr. Engler was exhausted by the long trekking at high altitude, but none the less he refused to ride the horse awaiting him below Vernal Falls, and trudged on to Camp Curry. There awaiting him, of all things, was his valuable pocket watch, accidentally left behind at the Glacier Point Hotel but delivered by courier down the literally vertical Four-Mile- Trail from Glacier Point. Incidentally, during the scenic descent along the Hlilouette Trail into the incomparable Yosemite Valley, Jepson observed the Europeans’ “‘wonder grew and grew”’ (Jepson 1913-1914, p. 5). Originally the schedule called for spending the night at Camp Curry, the main valley hostelry. Indeed, David Curry had planned for the eminent group “‘the biggest fire-fall (smoldering Califor- nia red fir embers dumped over the cliff from Glacier Point) and the biggest camp fire’, and around the flickering light of the fire a sympo- sium on the formation of Yosemite Valley (Jepson 1913b, no. 395). But alas, there was only time at Camp Curry for a late lunch, then by stage back to El Portal for the night, and onto the train next morning for “‘the long tedious hot dusty” trip to Berkeley. The exhausted excursion members were re- turned to Berkeley’s Hotel Shattuck on Friday at 4:30 p.m.; but the day was by no means over as they were scheduled for the first banquet of the newly formed California Botanical Society in downtown Oakland that evening! Somehow hastily groomed, the excursionists filed into a private dining room at Hotel Oakland promptly at 7 p.m. on the 12th of September 1913 (Jepson 1916, pp. 14-18). At the banquet Professor Jepson, first presi- dent of the new society, greeted the assemblage of | more than a hundred, noting that the Phyto-jogs had just returned from Yosemite, “‘laden with botanical spoils, and covered impartially with the dust of the San Joaquin.’ Even Jepson himself | had managed to garner 46 specimens. He went on to emphasize that this group of international natural scientists represented a new school, the plant ecologists, ““who are leading us back to the field and woods ... making important use of the observations of the old-time naturalists.” Dr. Setchell from Berkeley then welcomed the visitors 2009] on behalf of the University of California, wishing them “‘the greatest success in their further studies in this state.” At this point President Jepson introduced Dr. C. Hart Merriam, eminent for his distributional work with both plants and animals, resulting in the life zone concept. Merriam noted the “‘special interest that California has for the naturalist from the great diversity of its soil and climate conditions.’ His brief comments were followed by Jepson’s presentation of Dr. David Barrows, acting president of the University of California, who early in his professional career had researched the ethno-botany of California’s Cahuilla Indians. Jepson then turned to Professor Tubeuf of the University of Munich. Dr. Tubeuf, apologizing for speaking in German, expressed thanks for the princely hospitality of the gentlemen, and espe- cially, the ladies present; the latter “‘fair as your skies, rosy as your wine’’. Indeed they were the only group of American women the Excursion had socialized with in the New World. Dr. Tubeuf reflected that as a lad he had read “of your high mountains, wonderful trees, and fields of glorious bloom.” And it “‘was the dream of my youth to see this paradise.”” For Tubeuf the high point of the California visit so far had been the Mariposa Grove. Standing among the immense conifers, he thought he fleetingly glimpsed a butterfly. “It alighted; and behold it was no butterfly but a bird, a hummingbird. How most remarkable, at the same moment to see the smallest of birds and the greatest of trees.’ In closing, Tubeut praised “‘the freedom, the aban- don, the largeness, the youth of your Western life.” And he expressed the hope that the new “California Botanical Society will live to the age and dignity of your mighty Sequoias.”’ At last Jepson had ‘“‘the especial honor of presenting Dr. Engler, Director of Berlin’s Royal Botanical Gardens, who, apologizing for his limited English, made a well received presenta- tion, after which the entire audience rose and drank a toast to the eminent scientist.” Finally, the crowd moved to the hotel ballroom where Switzerland’s Dr. Schroter presented a lecture on the flora of the Alps, illustrated with beautiful lantern slides, a talk delivered in English ‘“‘which was excellent for a man who had just given only a few months to the language” (Jepson 1913a, p. 198). Incidentally, the original anticipation was to have Dr. Schroter’s lecture paid for by the University of California or by Stanford, but it turned out that neither institution, as Stanford’s Professor Pierce observed, would be ‘exactly lavish in its official recognition of the German Imperial Botanist or any of the others.”’ World War I, of course, commenced the next year (Jepson 1913b, no. 627). At evening’s end Professor Schréter expressed to Jepson that “it was the finest dinner and BEIDLEMAN: JEPSON’S PHYTO-JOGS Se) grandest occasion the party had experienced since landing in America” (California Botanical Society Collection 1913, letter, Jepson to Mrs. D. Van de Veer, Sept. 15, 1913). Overall, Jepson concluded that “‘the dinner was excellent, the speeches were satisfactory and our guests seemed thoroughly to enjoy themselves” (Jepson 1913a, p. 196). But undoubtedly the salient memory of this banquet which would linger most indelibly for many was the sudden appearance of botanist Sarah Lemmon at the podium (Jepson 1913b, no. 622). As Jepson stood up to introduce President Barrows, she materialized from the audience (though she had not been invited), “interrupted the programme, seized the floor and sailed grandly on’’, narrating the manifold but unappreciated past achievements of forester-husband John Lemmon (who had died in 1908), as well as those of Sarah herself, who had by 1903 successfully championed the California poppy (Eschscholzia californica) as the state flower. Jepson later wrote southern California botanist Samuel Parish that “Thus are pro- grammes smashed in the interests of free speech ... Iam sure you will be heartily pleased to know that Mrs. Lemmon’s brilliance and safeguards against interruption when talking have not been dimin- ished by years or the ravages of time” (Jepson 1913b, No622), Next morning, Saturday, the Phyto-jogs were dispersing in various directions around San Francisco, some to the University of California campus, to Stanford University, to salt marshes along the bay near Redwood City, and to nearby artificial brine ponds to view “‘organisms living under fatal conditions” (Jepson 1913b, no. 627), still others to see more redwoods and chaparral. Professor Jepson took Schroter and Stomps up to Santa Rosa to visit Luther Burbank, with whom Jepson had made an appointment (Jepson 1913a, pp. 198-201). After Burbank warned them that he restricted his tours to five minutes, Burbank ushered the botanists into his garden, pointing out exotics he had procured from various places. These were of little interest to the special visitors, who preferred to see some of Burbank’s experi- mental plants. Schréter and Stomps were partic- ularly impressed by the unique prickleless black- berries (Santa Rosa and Sebastopol Rubus) that Burbank demonstrated by leaning over and rubbing his cheek along the branches. When Burbank noticed a crowd of tourists peeking into the garden at the little group of savants, he half- snarled to Jepson “‘See there, I can’t come out here but people begin to gather. There was a regular riot one day and I had to send for the police!” Sunday’s excursion for the Phyto-jogs was to Mount Tamalpais and Muir Woods (Fig. 2), with Jepson turning the guiding over to Alice East- wood of the California Academy of Sciences (Jepson 1913b, no. 444-445). It was a well organized day. Coming across by ferry from the 54 MADRONO [Vol. 56 FiG2. boots perched on a wood footrest. East Bay to Sausalito, the people ate breakfast on the boat. Met by Alice Eastwood, the party entrained for Mill Valley and was ready for the 8:15 Mt. Tamalapais and Muir Woods Train. As time permitted, there would be hikes which featured the flora of the Redwood forest and the coastal mountain chaparral. On the Rock Spring Trail the plant ecologists could see the peculiar vegetation which grows on serpentine outcrops, and from the scenic train the early July fireburn was visible, with ‘“‘vigorous” new growth. Alice had obtained free passes on the electric railway (but forgot to get one for herself), and she had prevailed on her Botany Club to underwrite a free luncheon at West Point Inn, the terminus of the train line near the mountaintop. Most impressively, Alice managed to secure complimentary beer for the European visitors during the heat of the September afternoon. The day on “The Mountain,” as she put it, indubi- tably proved to be the “‘only free entertainment of the Excursion!” (Wilson 1955, pp. 132-133). Some forty years later, Miss Eastwood would have appreciative reason to recollect Dr. Skotts- berg of Sweden with whom she became acquaint- ed on the Mt. Tamalpais junket. It was upon his recommendation that she was invited to be honorary president of the Seventh International Botanical Congress in Uppsala (Wilson 1955, p. 132; Bonta 1992, p. 102). At Tamalpais, however, Eastwood had devoted much of her attention to Dr. Engler, taking him along a Group Photograph at Mt. Tamalpais, California. Alice Eastwood is seated on Dr. Engler’s left, with her special trail where she pointed out the different species of manzanita (Arctostaphylos) and two silk-tassels (Garrya elliptica and G. fremontii), new to him (Wilson 1955, pp. 132-133). Engler, as a botanical garden man himself, told East- wood that the San Francisco Bay region would be a wonderful site for a botanical garden, and she confided her hopes that one would eventually be developed in San Francisco’s Golden Gate Park. Of course, the Conservatory of Flowers had been initiated there about twenty-five years earlier; and in 1940 the Strybing Arboretum and Botanical Gardens would be officially opened. From San Francisco, with Jepson and East- wood now freed of hosting responsibilities, most members of the International Phytogeographical Excursion on September 15 traveled south by train to Carmel and the Carnegie Institution’s Coastal Laboratory. Here they spent two days under the guidance of Daniel T. MacDougal, director of the Carnegie Desert Laboratory founded at Tucson, Arizona, late in 1902. Aiding MacDougal as “‘his willing lieutenants” were W. A. Cannon, his former New York Botanical Garden colleague, now Resident Investigator at the Desert Laboratory, Godfrey Sykes, the laboratory’s manager, and Forrest Shreve, on his way to becoming one of the Southwest deserts’ foremost plant ecologists and in 1915 a founder of the Ecological Society of America. It wasn’t just torrid desert heat in Arizona which had encouraged the Carnegie Institution to 2009] set up summer headquarters for the Desert Laboratory at Carmel (Howard 1945, pp. 433— 446; McGinnies 1981, pp. 10-14). Becoming interested in Burbank’s plant breeding experi- ments, the Institution in 1905 initiated a five-year, $50,000 grant to assist Burbank’s work, and sent out botanist Dr. George Shull to cooperate with Burbank in his research at Santa Rosa. By 1907, with Burbank and Shull at an impossible scientific and personality impasse, Cannon from Tucson replaced Shull. During that summer Cannon became familiar with officers of the Carmel Development Company on the Monterey Penin- sula, resulting in an enthusiastic offer from the company to build an adjunct laboratory for the Carnegie Institution at Carmel-by-the-Sea near the cool California seashore. Two years later, by which time Burbank had become persona non grata as far as Carnegie scientists were concerned, the laboratory became the summer headquarters for the Desert Laboratory. By 1913 when the Phytogeographical party visited, the so-called Coastal Laboratory was an integral part of the Carnegie’s Department of Botanical Research. Here on the Monterey Peninsula, the “‘center of one of the most remarkable communities of endemic plants in existence,” (Nichols 1919, p. 63) the Phyto-jogs hastily reveled in the natural bounty, from Monterey pine (Pinus radiata), Monterey cypress (Cupressus macro- carpa), and oak woodlands to chaparral and sand dune, ably tutored by Stanford’s Abrams and Cooper. Meanwhile, Berkeley’s Setchell focused on the peninsula’s panoply of coastal seaweeds. During the Phyto-jogs brief stay, MacDougal and his colleagues, with a generous special grant from the Carnegie Institution, lavished ‘“‘to the needs, both physical and intellectual, of the visitors.”” As ecologist Tansley later reflected, “‘this portion of the tour formed a most brilliant close to a brilliantly successful excursion’ (Tansley 1914, p. 325). On Wednesday morning, September 17, the remaining excursionists boarded the Southern Pacific train for Los Angeles (Tansley 1914, pp. 326-333). How appropriate for the Europe- ans and Easterners aboard this “Wild West” train that near San Luis Obispo there should be an exciting “hold-up”. A fire in a railroad tunnel forced the train passengers off the train. With luggage in hand they had to hike for a mile over a mountain path through smoldering chaparral, the night illuminated only by railroad men’s lamps and torches. A rescue train beyond the tunnel took the party on to an unscheduled stop at Santa Barbara, with a morning visit to sub- tropical gardens and the old Spanish mission. Then there was dinner in Los Angeles and all aboard the night-train east to the desert. The last day in California was spent with a morning tour from Mecca to the Salton Sea BEIDLEMAN: JEPSON’S PHYTO-JOGS JD (below sea level), formed by accidental diversion of the Colorado River into a playa basin, but since 1907 slowly evaporating. Here the Phyto- jogs took note of the changing plant distribution along the advancing shoreline. By the next year MacDougal would publish his own Salton Sea succession observations in The New Phytologist. During the afternoon a stop was made at a flourishing date palm plantation near Mecca, where it was noted that the temperature that day only reached 115 degrees in the shade. During the evening of September 19, the Southern Pacific train carried the international party over the Colorado River about eight miles north of Mexico and out of the Golden State into the newly formed State of Arizona. The botanists awoke to a Sonoran Desert vista of Giant saguaro (Cereus giganteus) cacti as they sped by train for Tucson and the Desert Laboratory for a several-day stay. There they spent a day touring the laboratory’s site on Tumamoc Hill, had another group photograph, and the next day explored by motor the botanical diversity of the Santa Catalina Mountains, backdrop of Tucson. Then they divided into two parties, one of which visited the Grand Canyon while the other made a circuit up the north slope of Mt. Lemmon, named in honor of botanist Sarah Lemmon, who in 1881 was the first white woman to reach its summit. Yes— none other than the unexpected disrupter of the grand California Botanical Society banquet! The West Coast excursion of the Phyto-jogs had essentially reached its conclusion. The general response of the European scientists to the California exposure was well summed up by Professor Tubeuf, writing Professor Jepson from Munich, Germany, on November 23, 1913 (Jep- son 1913b, no. 710): “I shall never forget the beautiful days in California and am now working with the greatest pleasure on my notations, photographs and collections ...”’ The retrospective observation from the Excursion by Dr. Brock- mann-Jerosch took a slightly different twist: ““Ach, Gott, dies ist kein Urwald. Sehen Sie nur die Stumpfen!”’ (“Oh, God, this is no virgin forest. You see only the stumps!” (Sears 1969, p. 131). Inevitably it would be phytogeographer Alfred Tansley who put into print in The New Phytol- ogist the revelation of the Phyto-jogs’ exhaustive excursion across the varied ecosystems of the United States, and the numerous phytogeogra- phers enroute who were already busy studying the diversified ecological bounty: **... the most vivid impression I personally obtained was of the earnestness and single-mindedness of American science. In that vast field of ecology America has secured a commanding position and from the energy and spirit with which the subject is being pursued by very numerous workers and in its most varied aspects, there can be little doubt that 56 MADRONO her present pre-eminence in this branch of biology—one of the most promising of all modern developments—will be maintained” (Tansley 1914, p. 333). None of these distinguished foreign scientists would ever return to California and become residents. But Professor Adolf Engler’s Reprint Collection, 183 linear feet including some 25,000 items, purchased from Engler’s widow by Her- man Knoche and presented to the Dudley Herbarium in 1945, now resides at the California Academy of Sciences (Daniel personal commu- nication). ACKNOWLEDGMENTS The author appreciates the very helpful editorial suggestions provided by Madrono editors John Hunter and Richard Whitkus, as well as the reviewer of the original manuscript, and my wife Linda who reviewed drafts. Thanks also to Dr. Thomas Daniel, The California Academy of Sciences, for information on the Engler reprint collection at the Academy and for verifying the identification of Alice Eastwood in the Mt. Tamalpais group photograph. The primary assemblage of The Phytogeographic Excursion photographs is in the Clements Collection, American Heritage Center, University of Wyoming, Laramie, WY. At the Univer- sity of California and Jepson Herbaria Archives there are 57 photographs, mostly taken in California, from which the selections in this narrative have been made. The unpublished correspondence, field books, and other memorabilia are also in these herbaria archives. LITERATURE CITED BONTA, M. M. 1992. Women in the field. Texas A & M University Press, College Station, TX. [Vol. 56 CALIFORNIA BOTANICAL SOCIETY COLLECTION. 1913. Correspondence for 1913. University and Jepson Herbaria Archives, University of California, Ber- keley, CA. HOWARD, W. L. 1945. Luther Burbank - a victim of hero worship. Chronica Botanica 9(5—6): 433-446. The Chronica Botanica Co., Waltham, MA. Published as a book. Burbank and the Carnegie Institution. JEPSON, W. L. 1913a. Field Book, Vol. 27:155,178,190— 201. University and Jepson Herbaria Archives, University of California, Berkeley, CA. 1913b. Correspondence, Bound Volume 8. University and Jepson Herbaria Archives, Univer- sity of California, Berkeley, CA. . 1913-1914. Field Book, Vol. 28:1—23. Univer- sity and Jepson Herbaria Archive, University of California, Berkeley, CA. 1916. The International Phytogeographic Excursion in California. Madrono 1:12-18. MCGINNIES, W. G. 1981. Discovering the desert: the legacy of the Carnegie Desert Botanical Laborato- ry. University of Arizona Press, Tucson, AZ. NICHOLS, G. E. 1914. The International Phytogeo- graphic Excursion in America. Torreya 14:55—64. SEARS, P. B. 1969. Plant ecology. Pp. 130—131 in Joseph Ewan (ed.), A short history of botany in the United States. Hafner Publishing Company, New York, NY. TANSLEY, A. G. 1913. International Phytogeographic Excursion (I.P.E.) in America, 1913. The New Phytologist 12:322—336. . 1914. International Phytogeographic Excursion in America, 1913 (cont.) The New Phytologist 13:30-41, 83-92, 263-275, 325-333. WILSON, C. G. 1955. Alice Eastwood’s wonderland: the adventures of a botanist. California Academy of Sciences, San Francisco, CA. MADRONO, Vol. 56, No. 1, p. 57, 2009 A COMMENT ON THE ZONAL, INTRAZONAL, AND AZONAL CONCEPTS AND SERPENTINE SOILS EARL B. ALEXANDER Soils and Geoecology, 1714 Kasba Street, Concord, CA 94518 alexandereb@att.net Key Words: Serpentine, soils, zonal. Some prominent ecologists have stated that serpentine soils (soils with ultramafic parent materials) are azonal. Actually, only a small percentage of the serpentine soils are azonal. In order to rectify the inappropriate characteriza- tion of serpentine soils as azonal, it is helpful to review the origin and application of the zonal, intrazonal, and azonal terms in soil classification. A soil classification system based on the zonal, intrazonal, and azonal concepts was first devel- oped by Sibirtzev in Russia (Glinka 1914). These concepts were subsequently utilized in the USA (Marbut 1927). Then they were abandoned in the precursor to the present Soil Taxonomy of the USDA (Soil Survey Staff 1960), because the zonal concepts are too “‘arbitrary” to apply objectively and consistently (Kellogg 1963). Zonal soils are those that reflect the influence of the regional climate. Very shallow soils and soils in recent alluvium, volcanic ash, eolian sand, or loess that lack significant pedological devel- opment are azonal soils. They are those rudi- mentary soils that in similar deposits are similar in all climatic regions (or zones), ignoring vegetative differences. Intrazonal soils are those that reflect the influence of local conditions such as poor soil drainage that make the soils different from the “‘normal” zonal soils. A major problem with these concepts 1s deciding what is normal. Is it a 10,000 yr old soil in glacial drift or is it a 100,000 yr old soil on basalt, granite, or serpentinized peridotite? The decision might depend on the common lithology in a region, but the soils in a region with predominantly one of these parent materials (for example, granite) might be quite different from the soils in a climatically similar region with different parent materials (for example, basalt). In fact, well drained and developed serpentine soils (those other than Entisols) might be considered zonal soils in a region where the lithology is predominantly ultramafic rocks. Most serpentine soils in the botanical California Region are Alfisols, Mollisols, and Inceptisols (Alexander et al. 2007). Only Entisols might definitely be called azonal soils, and Entisols occupy only about 0.8% of the serpentine area in the Region (Alexander 2004). Therefore most of the serpentine soils are either zonal or intrazonal, depending upon how one applies these concepts. Those who consider ultramafic rocks to be unusual components of the continental crust would assume that most serpentine soils are intrazonal. The zonal-intrazonal-azonal concepts are very useful, even though they are insufficiently defin- itive to use in a modern system of soil classifica- tion. The zonal and intrazonal concepts can be applied to plant associations as well as to soils. Zonal plant associations may be common on zonal soils and intrazonal plant associations may be common on intrazonal soils. The azonal concept, however, is not easily applied to plant communities. Perhaps the plant associations that conform most closely to the azonal concept are those in early stages of succession. Some plant species, such as Bryum argenteum Hedwig and Achillea millefolium L., have the azonal charac- teristics of being widespread across many climatic zones, but the azonal concept may not be applicable to plant associations. LITERATURE CITED ALEXANDER, E. B. 2004. Varieties of ultramafic soil formation, plant cover, and productivity. Pp. 9-17 inR.S. Boyd, A. J. M. Baker, and J. Proctor (eds.), Ultramafic rocks: their soils, vegetation, and fauna. Science Reviews, St. Albans, Herts, UK. , R. G. COLEMAN, T. KEELER-WOLF, AND S. P. HARRISON. 2007. Serpentine geoecology of West- ern North America. Oxford University Press, New York, NY. GLINKA, K. 1914. Die typen de bodenbildung, ihre klassifikation und geographische verbreitung. Ver- lagsbuchhandlung Gebriider Borntraeger, Berlin. KELLOGG, C. E. 1963. A new system of soil classifica- tion. Soil Science 96:1—5S. MARBUT, C. F. 1927. A scheme for soil classification. First International Congress of Soil Science, Proceedings 4:1—31. SOIL SURVEY STAFF. 1960, Soil classification, a comprehensive system, 7th approximation. United States Department of Agriculture, Soil Conserva- tion Service, Washington DC. MADRONO, Vol. 56, No. 1, pp. 58-59, 2009 EVIDENCE OF EXTREME ROOT PROLIFERATION IN RESPONSE TO THE PRESENCE OF A NUTRIENT RICH RESOURCE PATCH BY ERIOGONUM PARVIFOLIUM MILAN J. MITROVICH' Department of Biology, San Diego State University, 5500 Campanile Drive, San Diego, CA 92182 milan.mitrovich@gmail.com ABSTRACT I discovered evidence of extreme root proliferation by Eriogonum parvifolium following the recovery of skeletal remains of a large, radio-tracked snake buried in the sand dune environment at the Tijuana Estuary in San Diego, California. Located beneath an isolated E. parvifolium, the recovered remains were grossly entangled by a large number of roots generally absent from the adjacent soil. The apparent massive localized proliferation of roots suggests decomposition of the animal created a favorable soil mosaic readily utilized by the plant in an otherwise low nutrient environment. Key Words: Root proliferation, Eriogonum parvifolium, nutrient patch, root growth response, resource heterogeneity. A number of plant species show plastic growth responses when associated with the patchy distribution of resources (Robinson 1994; Hodge 2004). One type of growth response expressed when a plant encounters a nutrient rich patch of resources is the proliferation of lateral root branching (Hodge 2004). This can be massive in a region well supplied with a nutrient when the remainder of the root system is nutrient deprived (Robinson 1994). Eriogonum parvifolium Smith (Polygonaceae) is a woody perennial shrub common to stabilized sand dune formations in the coastal area of central and southern Califor- nia (Purer 1936). In dune environments individ- uals of this species can attain 0.3 m to 1.0 m in height and 0.5m to 2.0m in spread with the taproot extending to a depth of 1.2 m and lateral roots spreading out two to three times as far as the parts above ground (Purer 1936; Hickman 1993). Sand dunes are low nutrient environments that show considerable resource heterogeneity (Hesp 1991). Living in the low resource environ- ment associated with sand dune formations, FE. parvifolium may show a significant root growth response following an encounter with a large nutrient resource. Here, I provide evidence of extreme root proliferation by E. parvifolium in response to encountering a rich resource patch, specifically a >500 g buried animal corpse. In 2002, as part of a multi-year study on the space-use behavior of the Coachwhip Snake (Masticophis flagellum), 1 tracked 12 Coachwhips surgically implanted with radio-transmitters at the Tijuana River National Estuarine Research Reserve in San Diego County, California 'Present Address, Green Shield Ecology, 15360 Barranca Parkway, Irvine, CA 92618. flagellum. However, this individual showed no (32°34'’N Latitude, 117°07’W Longitude, 0-6 m elevation; see Mitrovich et al. in press). The 1024 ha research reserve, bordered by the cities of Tijuana (Mexico) and Imperial Beach (USA), is a | complex landscape with tidal marsh, alkali flats, | fallow agricultural fields, riparian woodland, © upland sage scrub, and sand dune habitats. The > dune formations are limited to a narrow strip | bordering the ocean and characterized by short | dunes rarely reaching further than 125 m inland. | The dunes lack trees and are largely vegetated by — low-lying perennial plant species including Abro- | nia umbellata, A. maritima, Ambrosia chamissonis, | Atriplex leucophylla, Calystegia soldanella, Ca- missonia cheiranthifolia, E. parvifolium, and Rhus | integrifolia. On August 10, 2002 one of the radio-tracked | snakes, a large, 1.13 m (snout-to-vent length), 560 g female, who had been active in the dune. environment throughout the spring and summer seasons, was recorded using a rodent burrow located in a small, isolated dune partially covered | by E. parvifolium. The use of a small mammal burrow as a retreat site is not uncommon for M. movement away from the burrow the rest of the year through the spring of 2003. As a result of detecting no movement the following spring when all other tracked snakes were emerging from winter hibernation, I concluded the radio- tagged M. flagellum had died the previous fall or winter. This was confirmed when I dug-up the remains of the animal with the still functioning transmit- ter on May 28, 2003, 290 days after the snake was first recorded as entering the burrow. I recoverec the remains of the snake approximately 0.6 m tc 0.7 m below the surface of the sand dune. The 2009] remains were located directly beneath a single, large E. parvifolium. The burrow had collapsed and the transmitter, although intact, was sepa- rated from the remains by several centimeters. Only skeletal remains were left. No soft tissue or skin of the animal was present. The snake’s vertebral column was not complete as it was broken in several sections and not all sections were recovered. Although the root structures of E. parvifolium were largely absent from the excavated soil removed during the digging process, a large mass of fine roots from the plant were entangled with the recovered segments of the vertebral column. The roots were so completely entwined with the vertebrae it was not possible to collect segments of the vertebral column without also harvesting roots. It appeared as though the roots entwining the skeletal remains were still connected to the root system of the plant. The way the roots had grown so tightly around the skeleton suggests the soil was nutrient-rich in the immediate vicinity of the animal’s remains. As microbes in the soil broke down the animal tissue during the preced- ing months, the subsequent release of nutrients (e.g., N, P, Ca) was likely to be high and highly desirable by the plant, explaining a realized benefit behind the proliferation. Decaying verte- brate tissue is known to enrich adjacent soil horizons (Ad] 2003). Collectively the evidence suggests E. parvifo- lium responded to the resource-rich patch by increasing occupation of the immediate area surrounding the body. This evidence for signifi- cant root proliferation in E. parvifolium is not surprising given the environment (i.e., nutrient poor soils) where the response occurred. It is specifically in these nutrient poor soils where plasticity of root growth is expected to be most extreme following an encounter with a large -resource patch (Robinson 1994). Clearly, the 500+ g animal corpse created a large organic ‘input to the otherwise resource poor environment of the sand dunes. Plant-animal interactions are MITROVICH: EXTREME ROOT PROLIFERATION BY ERIOGONUM PARVIFOLIUM a9 often categorized as mutualistic (e.g., pollination, seed dispersal) or antagonistic (e.g., herbivory; Gurevitch et al. 2006). Animal droppings are well known to contribute to soil quality mosaics that favor some plants’ growth (Harper 1977). Crea- tion of a favorable soil mosaic during decompo- sition is an additional benefit, albeit inadvertent. ACKNOWLEDGMENTS Research was funded by the National Oceanic and Atmospheric Administration—National Estuarine Re- search Reserve System dissertation fellowship (F#NAI17OR1185) to MJM, San Diego State University, and U.S. Geological Survey. I thank S. N. Handel for his helpful comments on an earlier version of the manuscript. The California Department of Parks and Recreation and U.S. Fish and Wildlife Service granted access to land, which is also greatly appreciated. LITERATURE CITED ADL, S. M. 2003. The ecology of soil decomposition. CABI Publishing, Cambridge, MA. GUREVITCH, J.,S. M. SCHEINER, AND G. A. Fox. 2006. The ecology of plants, 2nd ed. Sinauer Associates, Sunderland, MA. HARPER, J. L. 1977. Population biology of plants. Academic Press, New York, NY. Hesp, P. A. 1991. Ecological processes and plant adaptations on coastal dunes. Journal of Arid Environments 21:165—191. HICKMAN, J. C. (ed.). 1993. The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. HopcGE, A. 2004. The plastic plant: root responses to heterogeneous supplies of nutrients. New Phytolo- gist 162:9-24. MITROVICH, M. J., J. E. DIFFENDORFER, AND R. N. FISHER. In Press. Behavioral response of the Coachwhip (Masticophis flagellum) to habitat fragment size and isolation in an urban landscape. Journal of Herpetology. PURER, E. A. 1936. Studies of certain coastal sand dune plants of southern California. Ecological Mono- graphs 6:1—87. ROBINSON, D. 1994. The responses of plants to non- uniform supplies of nutrients. New Phytologist 127:635-674. MADRONO, Vol. 56, No. 1, pp. 60-62, 2009 REVIEWS Plant Invasions: Human Perception, Ecological Impacts and Management. Edited by B. To- KARSKA-GUZIK, J. H. BROCK, G. BRUNDU, L. CHILD, C. C. DAEHLER AND P. PYSEK. 2008. Backhuys Publishers, Leiden. xvii + 427 pp., 103 figures, 50 tables. Paperback, Euro 126.00, ISBN 978-3-8236-1528-6. Over the last two decades, plant invasions have become a subject of an increasing number of national and international meetings. This volume presents key contributions from the 8th Interna- tional Conferences on the Ecology and Manage- ment of Alien Plant Invasions (EMAP1), held at the University of Silesia, Katowice, Poland, in 2005. In total, 27 chapters were written by 71 authors from 18 countries and four continents. The volume is divided into four sections: 1— Human perception and role in biological inva- sions (four chapters), 2—Biology, ecology and distribution of invasive species (seven), 3— Invasibility of habitats and impacts of invasive species (12), 4—Control and management (four). As in the previous EMAPi volumes, standardized international terminology is used (Richardson et al. 2000, PySek et al. 2004). First, browsing through the Contents, we may be somewhat disappointed that only two contri- butions are from the USA (C. C. Daehler: Invasive plant problems in the Hawaiian Islands; J. H. Brock: Ecology and management of A/hagi maurorum in Arizona). Nevertheless, because plant invasions are a global problem, we can learn important lessons from studies conducted in Australia, Europe, or Africa. Moreover, several contributions in this volume are of general importance, addressing very basic questions of invasion biology. For example, Daehler’s analysis of major motivations for plant introductions in Hawaii (nostalgia, neophilia, economics, im- provement of ecosystem services) could serve as a model for similar studies in other countries. Similarly, approaches used in the chapter by Philip Hulme and co-authors on multiscale analyses of plant invasions to Mediterranean islands could be applied in different parts of the world. Twenty contributions from different parts of Europe serve as examples of a very well- organized research on plant invasions over this continent. Recent publication of the European catalogue of all known invasive species (DAISIE 2009) is an impressive result of cooperation across Europe. The book is very well edited and packed with interesting data. However, as in the previous volume that I reviewed for Madrono (Rejmanek 2003), one chronic weakness of plant invasion biology still remains: a lack of rigorous evidence for assumed harmful impacts of invasive taxa in natural and seminatural areas. The phrase “ecological impacts” is in the title of this new volume, but only a very few contributors are dealing with this topic. Moreover, if they do, their conclusions are based on descriptions of pairs of invaded and non-invaded plots (space for time substitution). The only exception is experimental study on competition between Acacia longifolia and native woody species in Portugal (Christiane Werner et al.). Although, unfortunately, rather expensive for a paperback ($161), this volume is definitely worth attention of all botanists interested in plant invasions. —MARCEL REJMANEK, Section of Evolution and Ecology, University of California, Davis, CA 95616; mreymanek@ucdavis.edu. LITERATURE CITED DAISIE. 2009. Handbook of alien species in Europe. Springer, Dordrecht. PyseEk, P., D. M. RICHARDSON, M. REJMANEK, G. L. WEBSTER, M. WILLAMSON, AND J. KIRSCHNER. 2004. Alien plants in checklists and floras: toward better communication between taxonomists and | ecologists. Taxon 53:131—143. REJMANEK, M. 2003. Review of L. Child et al. (eds.) Plant invasions: ecological threats and manage- | ment solutions. Madrono 51:392-393. RICHARDSON, D. M., P. PYSEK, M. REJMANEK, M. G. | BARBOUR, F. D. PANETTA, AND C. J. WEST. 2000. Naturalization and invasion of alien plants: con- | cepts and definitions. Diversity and Distributions | 6:93-107. The California Deserts: An Ecological Rediscov- | ery. By BRUCE M. PAVLIK. 2008. University of © California Press, Berkeley and Los Angeles, CA _ and London, UK. 384 pp. ISBN 0-520-25140-7 _ $60.00, hardcover. ISBN 0-520-25145-8 $27.50, paper. People around the world have an unquench- | able fascination with deserts. The California Deserts: An Ecological Rediscovery by Bruce _ Pavlik will help to fill the niche in any library | for deserticolous or desertophilic readers thirsty” for interesting knowledge about the three deserts | 2009] of California. This is not the first book penned on the subject, and will not be the last. But Pavlik has made an attempt to combine and weave together, in a fairly readable and colorfully illustrated fashion, a broad range of topics on ‘the California desert bioregion.” To appreciate these wonderful natural resources, ultimately one needs to appreciate that our desert bioregion has been an ever-changing landscape. Derivative desert life forms have evolved in a cumulative manner from lineages inhabiting western North America during wetter epochs. The compositions of the present communities have been so highly modified during fluctuations in climatic and geologic conditions, and human activities, that none of us would probably recognize the original formation. Now we have the privilege and responsibility to understand how the ecosystems function, and in the pursuit of that goal to reconstruct the whys. Pavlik’s point, proposed by his subtitle (an ecological rediscovery), is that biologists were not the first to discover the nature of California’s deserts, but that humans living here for thousands of years knew many secrets and essences of desert life, therefore scientists have often rediscovered what the original inhab- itants already experienced. Most readers will likely appreciate the author’s introduction to folk knowledge of desert via a fictionalized narrative, of how an ancient seed- carrying basket was left within a small cave in Eureka Valley, found the next century by famous desert naturalist Mary DeDecker, and how that basket led to rediscovery of ancient knowledge. Next we read a condensed historical account of how westerners successively visited California’s deserts and learned from the original inhabitants, leading eventually to scientific curiosity and our modern knowledge. Following are chapters on the complex physical origins and abiotic charac- teristics of deserts; on the creatures, present and past, inhabiting southeastern California; on the adaptations of selected, extant desert species to deal with extremes in the environment; and on the ecology of desert communities. The text ends with an overview of recent changes inflicted on our deserts and what has been and is being done to preserve and restore extant communities. No glossary is provided. In general, the book is designed to be a bridge of knowledge between the scientific and management literature and the learned reader, professional scientist or not, who may know little about desert biology. Toward that goal, the amount of professional jargon is greatly re- duced, metric units are not used (with several _ exceptions), literature citations are mostly absent from paragraphs, memorable factoids and short essays are included to capture interest, and excellent many contributors, are used to accompany the colored images, accumulated from BOOK REVIEWS 61 text. This is an attractive book with a nice, clean layout. Approximately 85% of the text pages have photographic images, graphs, line drawings, or tables. A research scientist will, on the other hand, find many parts somewhat uncomfortable to read, lacking precision in how observations are present- ed—selectively and with too much poetic license— avoiding the appropriate terms (e.g., seeds that actually are fruits or fruitlets), and presenting many findings and offering conclusions that would not be admissible in a professional journal. So it 1s that books like this may bend too far to make a subject understandable to amateurs, and therein plant seeds—or fruits—in minds of the general public that scientists will have to push against to overturn because speculation morphs into “common knowledge.”’ My greatest disappointment was with the long chapters on “Operations and Origins” and ““Remarkable Biota.’ These together equal the core for any treatment on adaptations of desert organisms. Basic principles are not used effec- tively to give the reader a clear route to understanding desert adaptations, and some common desert features are never discussed. Much interesting information is presented, and a person who knows the literature will recognize where the data were originally published. But numerous statements are at best misleading, often lost in translation from the vast scientific literature on ecophysiology and structure-func- tion relationships. The reader does not have an opportunity, without literature cited, to deter- mine what is correct and what is not. An explanation of types of photosynthesis is terribly flawed, including diagrams that misrepresent carbon fixation in both Cy and CAM (initially cytosolic and not within chloroplasts), origin and diffusion of oxygen, and basic characteristics of each pathway, along with other errors in text and a table. Rather than taking a modern approach on resistance to stress, which any amateur can understand, the author adopted a totally artificial treatment of strategies for species as_ being ‘“drought-susceptible,” ““drought-avoiding,” and “drought-tolerant,’ and thereby takes an ap- proach popular before the onset of modern ecophysiology and now largely out of date. For the Madrofio audience, the book has some taxonomic shortcomings. It is already antiquated in that the author did not always adopt scientific names of Flora of North America (volumes available) and those combinations to be used in the upcoming edition of The Jepson Manual. Here is an opportunity to pass along the best available names for vascular plant taxa, to begin the process of breaking down public resistance to changes in scientific binomials, but instead wastes efforts of scores of biologists trying to sort out the phylogeny of organisms. Second, on the majority of pages common names of organisms 62 MADRONO are used, with a paucity of scientific names, this to reduce jargon and text length, but seldom are binomials repeated, so a reader must flip back to an earlier entry, often pages away, to make the match. Some of that problem would have been eliminated by always including the scientific name in figure legends, and including them also with common name and a former scientific name in the index. Third, as expected by all of us, some latinized names are misspelled (e.g., see back cover), because great proofreading and double checking is required. Finally, there is the peren- nial problem of having some common names that authors insist on being capitalized while others, especially plant names, are not. No modern molecular data are discussed to define paleoen- demics versus neoendemics, and, in fact, Bursera microphylla is classified as a paleoendemic even [Vol. 56 though that native is not endemic to the California desert bioregion. The California Deserts was written not as a review with scientific precision, but rather citing relatively few references (pp. 305-327). That accepted, there are many seminal books and articles that should have been referenced to allow readers to pursue their interests and network information. Honestly, there is plenty of interesting coverage to enjoy in this reasonably priced book. Readers will look forward to reading the final chapter, in which Pavlik passes along some really important knowledge about saving our deserts. —ARTHUR C. GIBSON, Department of Ecology and Evolutionary Biology, UCLA, Los Angeles, CA 90095-1606; agibson@biology.ucla.edu. MADRONO, Vol. 56, No. 1, pp. 63-69, 2009 NOTEWORTHY COLLECTIONS CALIFORNIA LEPECHINIA ROSSII S. BOYD & MISTRETTA (LA- MIACEAE).—Ventura Co., Topatopa Mountains, Pine Canyon, tributary of the lower Sespe Canyon drainage along the east side of Santa Paula Peak ridge; near 34.44833°N, 118.96633°W [NAD 83] 1600 ft (488 m); 6 Jun 2007, L. Gross, 2791 (RSA; MO, SBBG, UC, UCR). Topatopa Mountains, Coldwater Canyon, tributary of the lower Sespe Canyon drainage, north- east of Santa Paula Peak ridge (Coldwater Canyon is the next drainage upstream [north] from Pine Canyon); near 34.46239°N, 118.96312°W [NAD 83] 1550 ft (472 m), 13 Jun 2007, L. Gross, V. Arvizu & S. Boyd 2792 (RSA; CAS, SBBG, US). Previous knowledge. Lepechinia rossii (Ross’ pitcher sage) is a member of section Calycinae, which includes four additional species endemic to California and adjacent Baja California, Mexico—L. calycina (Benth.) Epling, L. cardiophylla Epling, L. fragrans (E. Greene) Epling, and L. ganderi Epling—and is most readily distinguished from these taxa by its geniculate inflores- cence axes, bent at 60°—90° angles relative to the subtending stems, and large, foliaceous inflorescence bracts which are generally equaling or exceeding their adjacent flowers in length, and little reduced distally (Boyd & Mistretta 2006, Madrono 53: 77-84). Lepechi- nia rossii 1s endemic to the western Transverse Ranges of southern California, being known from two some- what disjunct populations, one in the Liebre Mountains (northwestern Los Angeles Co.) and one in _ the Topatopa Mountains (southeastern Ventura Co.) (Boyd & Mistretta, loc. cit.). Within the Liebre Mountains, L. rossii is known only along the summit and northern flank of Red Mountain, within the watershed of Ruby Canyon. In the Topatopa Moun- tains, L. rossii is well documented from the lower portions of Tar Creek, a major tributary to the lower portion of the extensive Sespe Canyon watershed (Boyd & Miuistretta, loc. cit.). Efforts to locate additional populations of L. rossii within the Liebre Mountains region, e.g., in adjacent portions of Elizabeth Lake Canyon, Warm Springs Canyon, and on Warm Springs Mountain, have been unsuccessful. Within the Topa- topa Mountains, two historic specimens (Evermann s.n., 24 Mar 1917 [CAS #25345] and Hoffmann s.n., 21 Mar 1927 [SBBG #6403]), both non-flowering and with somewhat ambiguous locality data, suggest L. rossii may be more widely distributed within the lower Sespe Canyon watershed of the Topatopa Mountains region (Boyd & Mistretta loc. cit.). Evermann’s specimen is from “‘Pine Creek, near Sespe,’’ presumably this being Pine Canyon, a major tributary of Sespe Canyon whose confluence is downstream and to the west of Tar Creek. The locality on Hoffmann’s specimen is more vague, “Sespe Canyon.” This may refer to the old town of Sespe, near the confluence of Sespe Creek and the Santa Clara River, just west of Fillmore, or perhaps Sespe Hot Springs, or most anywhere along the entire 88 km main trunk of Sespe Creek and the many tributaries throughout its 690 square km? watershed. Significance. The Pine Canyon population of L. rossii _ documented by Gross 2791 was in full anthesis at the time of collection, providing vouchers with well developed and diagnostic inflorescences and flowers. This population is likely the same from which Ever- mann collected his substerile specimen in 1917, and validates the determination of that otherwise ambigu- ous voucher. Our specimen from Coldwater Canyon (Gross et al. 2792) appears to be the first record of L. rossii for that drainage. Together, these two collections help refine our understanding of the overall scope of the Topatopa Mountains metapopulation of this narrow endemic. Lepechinia rossii, as with other Californian species of Lepechinia in general, is essentially a pioneer species within chaparral habitats. Plants tend to occur in small, often widely scattered stands. They are _ usually associated with some sort of physical disturbance of vegetation and soil. Not uncommonly, this disturbance is anthropogenic in origin, e.g., cut or fill slopes along roads, and along cleared fuel breaks across ridgelines. This certainly characterizes known occurrences of L. rossii in the Tar Creek area of the Topatopa Mountains, and in the Liebre Mountains, where most plants are known from very near current or old roads and fuelbreaks. Unlike these previously known L. rossii populations, however, both the Pine and Cold- water canyon populations of L. rossii are strictly associated with “natural” disturbance, a fact that may provide a_ useful model in predicting additional occurrences within the Sespe Canyon/Topatopa Moun- tains region, as well as the extensive area between these populations and those of the Liebre Mountains to the east. Both the Pine and Coldwater canyon populations of L. rossii were small in size (ca. 5 plants in Pine Canyon, ca. 10 in Coldwater Canyon), and limited in areal extent. Each population is situated towards the base of steep, north-northeasterly slopes within mid-elevation portions of their respective drainages, and essentially restricted to areas disturbed by relatively recent landslides. These slides, perhaps 2 to 10 years old, are associated with steeply-bedded sedimentary substrates in areas where chaparral vegetation burned 2 to 5 years earlier. The presence of recent fire combined with physical soil disturbance and relatively mesic exposure appears to be key factors controlling distribution of this narrow endemic. Based on this information, we recommend future surveys for L. rossii to include the entire Santa Paula Peak ridge, especially its eastern end, as well as the northern slopes of San Cayetano Mountain. Other likely areas include the West Fork of Sespe Canyon south of Topatopa Peak, Stone Corral Creek north of Pigeon Flat, and other adjacent drainages along the Sespe Canyon watershed. Similarly, the northern facing flanks of Devils Heart Peak and Sulphur Peak, especially areas between ca. 450 m to 900 m (ca. 1500 ft to 3000 ft), warrant surveys for possible new localities of this species. —LEROY GROSS, VALENTIN ARVIZU, and STEVE BoybD, Rancho Santa Ana Botanic Garden, 1500 N. College Avenue, Claremont, CA 91711. leroy.gross@cgu. edu. 64 MADRONO CALIFORNIA SCHIZYMENIA DUBYI (Chauvin ex Duby) J. Agardh 1851: 171 (SCHIZYMENIACEAE).—Monterey Co., attached to floating docks Monterey Marina, Monterey Harbor, Monterey. 36°36'07''N, 121°53'25'’W. Thalli cystocarpic and sterile (13 September 2006 (UC 1934344), 15 July 2007 (UC 1934344), 8 November 2007 (UC 1934345), 6 March 2008 (UC 1934346), J. R. Hughey). Previous knowledge. Native to Europe, Africa, Asia, New Zealand, and Australia (M. D. Guiry and G. M. Guiry, AlgaeBase. World-wide electronic publication, National University of Ireland, Galway. http://www. algaebase.org; searched on 24 May 2008). Type locality: Cherbourg, Manche, France (P. C. Silva, P. W. Basson & R. L. Moe, Catalogue of the Benthic Marine Algae of the Indian Ocean, 1996: 323). First reported from Pacific North America based on material from Esqui- malt, British Columbia (W. H. Harvey 1862, Journal of the Linnean Society of London: Botany 6: 174). Citing Harvey, W. A. Setchell and N. L. Gardner (1903, University of California Publications Botany 1:356) included S. dubyi in their list of algae from northwestern America. They admitted however “we know nothing of this plant, but suspect that we may have included it under Sarcophyllis californica.” [Dilsea californica (J. Agardh) Kuntze 1891: 892]. Y. Yamada (1928, Scien- tific Reports of the Tohoku Imperial University, Biology 3:497-534) reported S. dubyi from Japan, but it was later removed from the flora by I. A. Abbott (1967, Bulletin of the Southern California Academy of Sciences 66:162) who treated Japanese specimens as conspecific with Schizymenia pacifica (Kylin) Kylin 1932: 10. Based on a recent phylogenetic analysis of the large subunit of ribulose-1, 5-biphosphate carboxylase/ oxygenase gene (rbcL), Gavio et al. (2005, Gulf of Mexico Science 83:38—57) demonstrated that S. dubyi from Japan is conspecific with S. dubyi from the Atlantic ocean, but different from S. pacifica from the northeastern Pacific. However, the two earlier reports of S. dubyi from the eastern Pacific have gone without comment by later workers. Significance. Previously misidentified specimens (as S. pacifica) include: 26 May 1978, R. Setzer (AHF 84371 in UC); 14 July 1977, L. A. Abbott (GMS 13296); 24 May 1974, I. A. Abbott (GMS 13187); 17 June 1972, L A. Abbott (GMS 11395); 8 May 1970, L. A. Midon & J. N. Norris (AHF 80259 in UC); 24 June 1966, N. L. Nicholson (GMS 2121 & 2122). Schizymenia dubyi also occurs in Ventura Co. just east of Carpinteria (13 January 1957, E. Y. Dawson, AHF 63782 in UC) and in Los Angeles Co. in the upper mid-littoral at Royal Palms (26 June 1972, R. Setzer (AHF 78717 in UC); 8 March 1972, R. Setzer (AHF 77975 in UC)). Mature thalli of S. dubyi can usually be distinguished from S. pacifica by their darker color (maroon rather than brown-pink), ruffled margins, longer stipe (3 mm vs. 1 mm or lacking), and ostioles that are approximately twice as large (45-60 um) as those found in S. pacifica (25-40 um). Our identification was confirmed by analyzing rbcL gene sequences following the methods described by J. R. Hughey et al. (2008, Phycologia 47:124-155). Two specimens were analyzed and found to be identical in sequence (GenBank FJ013041 & FJ013042), but differed from S. dubyi from Brittany, France (A Y294389) and Iwateken, Japan (AY 294388) by 5 nucleotides. In comparison, an rbcL sequence of S. [Vol. 56 pacifica from Washington (A Y294393) differed from S. dubyi by 69 nucleotides. These data confirm the presence of S. dubyi in California, but highlight the need for further investigations into the life history and taxonomy of species of Schizymenia in the northeast Pacific. —JEFFERY R. HUGHEY, Division of Science and Mathematics, Hartnell College, Salinas, California 93901. jhughey@hartnell.edu; and KATHY ANN MILL- ER, Herbarium, University of California, Berkeley, California 94720-2465. CALIFORNIA COTONEASTER HORIZONTALIS. Decne. (ROSA- CEAE).—Del Norte Co., pasture edge, 3 air km N of Lake Earl, elev. 8 m, 18 Jul 2007, Zika 23166 (CAS, DAO, GH, HSC, OSC, RSA, UC, WTU). Previous knowledge. Wall cotoneaster is recorded as a garden escape on the coast of Oregon, Washington and British Columbia, so wild plants on the north coast of California were expected. It is a low shrub readily distinguished from all other wild deciduous Cotoneaster in California by the small leaves, less than 2 cm long and less than 1 cm wide. It is recognizable at a distance by the strongly planar branching in a_ herringbone pattern. Significance. First collection for California as an escape from cultivation. COTONEASTER FRIGIDUS Wall. ex Lindl. (ROSA- CEAE).—Alameda Co., below Engineers Road, Straw- berry Canyon, Berkeley, 27 Sep 1934, C. M. Belshaw 199 (UC). Previous knowledge. Tree cotoneaster is uncommon in cultivation in North America, but is a common escape from gardens in the British Isles (Stace, in New Flora of the British Isles, 2nd edition, Cambridge University Press, Cambridge, UK. 1997). Hrusa et al. (Madrono 49:61—-98. 2002) were presumably including this databased specimen, which was labeled Cotoneaster lacteus W. W. Sm., when they said C. Jacteus was: “noted elsewhere in the east San Francisco Bay Area.” — The two can be separated by their foliage in the following key. la. Leaves thin, deciduous, often 10—15 cm long; veins Supericial 22.5 Js. es 2 nee C. frigidus | 1b. Leaves thick, evergreen, less than 10 cm long; Veins SUNKEN ac eatitioee oe e ee C. lacteus | Significance. First collection for California as an | escape from cultivation. PRUNUS SPECIOSA (Koidz.) Nakai (ROSACEAE).— Alameda Co., Strawberry Canyon E of UC-Berkeley © campus, wooded area at base of fire trail behind UC — Botanical Garden, 15 Apr 1999, B. Ertter 16502 (UC); ff shady riparian of Strawberry Creek, near Hamilton — Gulch, Strawberry Canyon, elev. 230 m, 19 Mar 2007, — Zika 22924 (DAO, NY, RSA, UC, UWEC, WTU). ; Previous knowledge. Prunus speciosa (syn. Cerasus i speciosa (Koidz.) H. Ohba, Prunus lannesiana (Carriere) | E. H. Wilson forma albida (Makino) E. H. Wilson) is) known as Oshima cherry, and is endemic to Japan (Chang et al., in Botanical Journal of the Linnean Society 154:35—54. 2007). It is cultivated on the Pacific 2009] Coast as an ornamental, but much less commonly than garden hybrids derived from it. This population was reported as Prunus serrulata Lindl. of China (syn. Cerasus serrulata (Lindl.) Loudon) in Hrusa et al. (loc. cit.). Prunus speciosa has leaf teeth with glandular tips and glandular-serrate sepals, while P. serru/ata has leaf teeth without glands and entire sepal margins. Prunus speciosa is also known as an escape in Orange Co., North Carolina (Weakley s.n. NCU). Significance. First report as an escape from cultiva- tion in California. PRUNUS X YEDOENSIS Matsum. (ROSACEAE).— San Francisco Co., city of San Francisco, adventive tree 4 m tall, Golden Gate Park, elev. 75 m, 22 Mar 2007, Zika 22931 (NY, UC, UWEC, WTU). Previous knowledge. Prunus Xyedoensis, yoshino cherry, is traditionally considered to be a_ hybrid between P. subhirtella Mig. and P. speciosa derived and grown in Japan in the 1800’s (Rehder, Manual of Cultivated Trees and Shrubs Hardy in North America, MacMillan Co., New York, 1927), and the protologue suggests the type was cultivated (Japan: in hortis Tokyoensibus ample culta). Recently classified at the rank of species, with no mention of hybridity or parentage, by Ohba (in Iwatsuki et al., eds., Flora of Japan, Volume IIb, Kodansha Ltd., Tokyo, 2001) and also by Chaoluan and Bartholomew (Flora of China 9:404-420. 2003). Those authors treat it in the genus Cerasus, as C. yedoensis (Matsum.) A. N. Vassiljeva, but we prefer to treat Cerasus as a subgenus of Prunus (Jacobson and Zika, Madrono 54:74-85. 2007). Prunus xX yedoensis is a commonly planted ornamental in North America. Chaoluan and Barthol- omew (loc. cit.) say it is native to Japan and Korea (Cheju Island), but do not provide supporting details, and possibly they are referring to naturalized plants spread from cultivation. Garden plants produce viable seed, and yoshino cherry has been collected escaped from cultivation in Washington. Significance. First report as an escape from cultiva- tion in California. —ARTHUR L. JACOBSON, 2215 E. Howe St., Seattle, WA 98112. arthurleej@earthlink.net; and PETER F. ZIKA, WTU Herbarium, Box 355325, University of Washington, Seattle, WA 98195-5325. CALIFORNIA JUNCUS FALCATUS E. MEY. SUBSP. SITCHENSIS (BUCHENAU) HULTEN (JUNCACEAE).—Del Norte Co., near lagoon, Crescent City, 1899, W. R. Dudley s.n. (DS); Humboldt Co., 0.5 mi SW of Samoa; 16 Jun 1936, H. S. Yates 5654 (RSA, UC); wet places near the shore, Stone Lagoon, 3 Aug 1924, J. P. Tracy & H. E. Parks 6746 (UC); Big Lagoon, 18 Oct 1925, J. P. Tracy 7290 (UC); low flats in sand dunes, ocean beach at N end of Humboldt Bay, 13 Oct 1930, J. P. Tracy 9218 (UC); sand dunes, Samoa Peninsula, 7 Aug 1965, R. F. Thorne 35223 & P. Everett (BM, CAS, RSA). Previous knowledge. Mapped in Alaska, British Columbia and Oregon by Brooks (Juncus subg. Graminifolii, in Flora of North America 22: 228. 2000). Kirschner et al. (Juncus subg. Juncus sect. Graminifolii, in Species Plantarum: Flora of the World 7:49—50. 2002) cited specimens as far south as Coos Co., Oregon, while NOTEWORTHY COLLECTIONS 65 noting reports from Asia or Japan refer to the related Asian species J. prominens (Buchenau) Miyabe & Kudo. All Juncus falcatus records in California were assigned to subsp. falcatus (as var. falcatus) by Swab (Juncus, pp. 1157-1165, in J. C. Hickman (ed.), The Jepson Manual, University of California Press, Berkeley, CA. 1993). Juncus falcatus subsp. falcatus is native to Australia as well as California, where it is restricted to the south coast (SCo) and central coast (CCo) geographic subdivisions. California plants can be divided into two subspecies using the following key. la. Inner tepals (petals) blunt or hooded, usually less than 4.5mm long; anthers less than 1.7 mm long; fruit apex notched, conspicuous and roughly equaling the tepals, often globose to broadly elliptic; NCo ....... subsp. sitchensis lb. Inner tepals acuminate, usually more than 5 mm long; anthers usually more than 1.7 mm long; fruit apex usually acute to truncate, inconspicuous and much shorter than the tepals, usually elliptic to oblong; CCo, SCG ee eo ay eo Aa te a ee wie oe eare subsp. falcatus Significance. First report for California. The plants are variable in size, and usually sort well into the two subspecies, but a few central California specimens show some transitions, and the rank of subspecies seems more appropriate than species for the two taxa. The relationship of Californian and Australian representa- tives deserves investigation; southern hemisphere spec- imens are traditionally called subsp. falcatus but may more closely approach subsp. sitchensis in some morphological characters. JUNCUS INTERIOR WIEGAND (JUNCACEAE).—San Bernardino Co., granitic sand, Fourth of July Canyon, W New York Mountains, 1845 m, 4 Jun 1973, J. Henrickson 10551 (RSA); same canyon, 1829 m, 30 Aug 1973, J. Henrickson 12703 (RSA). Previous knowledge. Interior rush was mapped from Ohio west to Saskatchewan, Wyoming and New Mexico by Brooks (Juncus subg. Poiophylli, in Flora of North America 22:228. 2000). Kirschner et al. (Juncus subg. Agathyron sect. Steirochloa, in Species Plantarum: Flora of the World 8:17—57. 2002) expanded the range west to include Utah and Arizona. Kartesz (Synthesis of the North American Flora, Version 2.0, CD Rom. 2003) shows herbarium records as far west as Mohave Co., Arizona, adjacent to San Bernardino Co., California. Juncus interior often has a pinkish base to the rather tall stems, but otherwise closely resembles J. tenuis Willd., which has shorter stems only rarely pinkish instead of green at the base. Juncus tenuis is a common species on damp disturbed ground on the Pacific Coast. The two can be distinguished by the following key. la. Auricles of early season shoots acuminate (rarely acute), uniformly pale or translucent, 1-8 mm long; stem with 0-1 strong longitudi- nal ridges visible on one _ side; bracteoles subtending flowers acute to blunt...... J. tenuis 1b. Auricles of early season shoots rounded (rarely acute), usually opaque and with the marginal (outer) half thinner textured than the basal (inner) portion, 0.2-0.6 mm long in CA collections; stem with 4-6 strong longitudinal ridges visible on one side; bracteoles subtend- ing flowers acuminate, usually aristate . J. interior 66 MADRONO Significance. First report for California, and the species should be sought in additional washes in desert mountains within the eastern Mojave Desert of California. JUNCUS NEVADENSIS S. WATS. VAR. INVENTUS (L. F. HEND.) C. L. HitcHc. (JUNCACEAE).—Humboldt Co., frequent, boggy places, Big Lagoon, 6 m, 18 Oct 1925,.J. P. Tracy 7293 (UC). Previous knowledge. Nevada rush is a_ variable species, and authors disagree on its taxonomy. Follow- ing Hitchcock and Cronquist (Flora of the Pacific Northwest. University of Washington Press, Seattle, WA. 1973), the 2008 Oregon Plant Atlas (available at: http://oregonflora.org/atlas.php) and Kartesz (loc. cit.) map var. inventus as an Oregon endemic along every county of the Pacific Coast, from Clatsop Co. south to Curry Co. Kirschner et al. (Juncus subg. Juncus sect. zophyllum, in Species Plantarum: Flora of the World 7:151-270. 2002) followed Clemants (Juncus subg. Septati, in Flora of North America 22:240—255. 2000) and did not recognize any varieties of J. nevadensis. I agree with Cronquist (Juncaceae, in Intermountain Flora: Vascular Plants of the Intermountain West 6: 47-64. 1977) that the plants of the interior are not readily divided into geographic and morphological varieties, but more study is needed of the patterns of variation. However, the plants of coastal sand dunes are disjunct, with much more regularly and_ strongly flattened leaves, compared to the slightly flattened to tubular leaves of inland populations. Coastal seeds tend to be ovate and slightly plumper than the usually elliptic seeds from the interior. The number of heads and tepal color seem to fluctuate without correlation. The best discriminator is the stamen, noted in the following key. la. Anthers usually much longer than the fila- ments, inland and montane...... var. nevadensis 1b. Anthers shorter than to equaling the filaments, Coastal. co 5% sexes Ole ooo we ee var. inventus Significance. First report for California, and to be sought in interdunal swales along the northern coast. Plants of the Willamette Valley in northern Oregon need study, and may represent another distinct popu- lation. Some plants with very fine foliage from the Sierra Nevada approach J. mertensianus Bong., and may prove to be separable. —PETER F. ZIKA, WTU Herbarium, Box 355325, University of Washington, Seattle, WA 98195-5325. Zikap@comcast.net. COLORADO CAREX CONOIDEA WILLDENOW (CY PERACEAE).— Jefferson Co., Meyer Ranch Park, Jefferson County Open Space, along US Hwy 285, about 0.8 km (0.5 mi) east of Aspen Park and 25.7 km (16 mi) west of Denver, 2403 m (7885 ft), 7.5’ Conifer quad, UTM NAD83 Zone 13S 76683E “°77372N; 29 June 2008, Steve Popovich 8508, with Pamela F. Smith, Anton A. Reznicek, Loraine Yeatts, and Leo Bruederle (MICH, KHD, COLO, CS, RM). Approximately 50 plants in wet sedge meadow along south side of South Turkey Creek, with Carex brevior, C. microptera, C. buxbaumii, [Vol. 56 Juncus mertensianus, Equisetum arvense, Cirsium cana- densis, _Deschampsia_ caespitosa, Hierochloe _ hirta, Phleum pratense, Iris missouriensis, Crunocallis chamis- soi, Allium geyeri, Bistorta bistortoides, and Neolepia campestris. Previous knowledge. Distributed throughout much of northeastern North America, from Manitoba south to central Missouri and east to northwestern North Carolina and Newfoundland. Uncommon, with the exception of New England; occupies open meadows, wet prairies, and shores of lakes, ponds, and rivers. Reported from one site in New Mexico (J. D. Coop 2003, The New Mexico Botanist 25:7) and another in Arizona, where it is presumed to have been introduced, but not naturalized (R. F. C. Naczi 1992, Systematics of Carex Section Griseae (Cyperaceae), Ph.D. dissertation, University of Michigan, Ann Arbor, MI.; R. F. C. Naczi and C. T. Bryson 2002, in Flora of North America Editorial Committee, Flora of North America North of Mexico, vol. 23. Oxford University Press, New York, NY). Similar taxa in the region (Wyoming, Colorado) include Carex crawei and C. blanda, from which C. conoidea is differentiated on the basis of its impressed nerves (most visible in living material). In addition, C. crawei is colonial from long-creeping rhizomes and has lower pistillate scales awnless or, occasionally, with a *+smooth awn; pistillate spike bracts shorter than to +equaling staminate spike; and peduncles of the pistillate spikes mostly smooth. Carex conoidea 1s cespitose, forming small clumps from short rhizomes, and has lower pistillate scales with a conspicuous scabrous awn; longer pistillate spike bracts, usually exceeding the staminate spike; and peduncles of the pistillate spikes scabrous. Carex blanda differs from both species in having perigynium nerves ca. 25—30 (vs. ca. 12—25); perigynia cuneately or even concavely tapered to the base when dry (vs. rounded); apex of perigynium abruptly bent, the orifice pointing to the side (vs. nearly straight); and culms sharply trigonous to +winged, soft and easily compressed (vs. firm and wiry, not easily compressed). Significance. First documented occurrence of C. conoidea from Colorado. Eastern sites are over 1250 km distant. Meyer Ranch Park has had a long and varied history of land use, and this new occurrence adds complexity to the question of nativity for the New Mexico and Arizona sites. ACKNOWLEDGMENTS We acknowledge the Colorado Native Plant Society, which sponsored the Carex workshop and field trip leading to this new record for Colorado. —PAMELA F. SMITH, 4824 Overhill Dr., Fort Collins, CO 80526; and ANTON A. REZNICEK, University Herbarium, 3600 Varsity Dr., University of Michigan, Ann Arbor, MI 48108-2228; and LORAINE YEATTS, Denver Botanic Gardens, Kathryn Kalmbach Herbar- ium, 909 York St., Denver, CO 80206; and STEVE J. PoPOVICH, Arapaho and Roosevelt National Forests and Pawnee National Grassland, 2150 Centre Avenue, | Building E, Fort Collins, CO 80526-8119; and LEO P. BRUEDERLE, Department of Integrative Biology, CB 171, University of Colorado Denver, POB 173364, | Denver, CO 80217-3364. Leo.Bruederle@ucdenver.edu. | 2009] MONTANA ACHILLEA FILIPENDULINA LAM. (ASTERACEAE).— Missoula Co., w side of Duncan Drive | km nw of Rattlesnake School in the Rattlesnake Valley on the outskirts of Missoula. One large plant on the road edge with Bromus inermis and Centaurea maculosa, 1065 m, T13N RI9OW SWI1/4 82, 22 September 2007, P. Lesica 9885 (MONTU). Significance. First report of this common introduced ornamental escaping in MT (Dorn, R. D. 1984, Vascular plants of Montana, Mountain West Publish- ing, Cheyenne, WY; United States Department of Agriculture PLANTS database, www.plants.usda.gov). AJUGA REPTANS L. (LAMIACEAE).—Flathead Co., NE end of Lake McDonald, one large colony at the edge of a dirt road in a _ hemlock forest with Symphoricarpos albus and Ranunculus repens, 975 m. 29 June 2003, P. Lesica 8667 (MONTU), verified by K. Chambers (OSU). Significance. First report of this exotic for MT (Dorn 1984, loc. cit.; PLANTS database, loc. cit.). ALLIARIA PETIOLATA (BIEBERSTEIN) CAVARA & GRANDE (BRASSICACEAE).—Missoula Co., Mis- soula, ne corner of the Natural Sciences Building, University of Montana campus, | plant near the bicycle rack with Poa pratensis and Chenopodium berlandieri, 945 m, 4 May 2007, P. Lesica 9720 (MONTU). Significance. First report of this prolific weed for MT (Dorn 1984, loc. cit.; PLANTS database, loc. cit.). MIMULUS HYMENOPHILUS MEINKE (PHRYMA- CEAE).—Lake Co., Mission Mtns., along the trail to Mission Falls, ca. 75 plants on a moist, limestone shelf protected by an overhang on a s-facing slope, 1370 m, TI8N RI8W S12, 25 July 1983, P. Lesica 2758 (MONTU, NY). Determined by M. Carlson (UAAH). Significance. First report for MT (Dorn 1984, loc. cit.; PLANTS database, loc. cit.); otherwise known only from a few side canyons of the Snake River 280 km sw in se WA. ORTHOCARPUS TOLMEI HOOKER & ARNOTT (ORO- BANCHACEAE).—Gallatin Co., Madison Range, s of Ernest Miller Ridge, common in tall herb meadows with Helianthella uniflora and Balsamorhiza sagittata, 2710 m, TIOS R5E 829, 10 July 2007, P. Lesica 9815 with P. Kittelson (MONTU, NY). Verified by N. Holmgren (NY). Significance. First report for MT (Dorn 1984, loc. cit.; PLANTS database, loc. cit.); previously adjacent mcton Co,, WY, PAPAVER CROCEUM LEDEB. (PAPAVERACEAE).— Stillwater Co., Beartooth Range, near the old mine just west of Chrome Lake, locally common along the road with Potentilla fruticosa and Astragalus australis. 2560 m, T5S RI6E S31, 23 Aug 1999, P. Lesica 7963 (MONTU) determined by D. Murray (ALA). Significance. First report for the continental U.S (PLANTS database, loc. cit.). Cultivated, escaped and very well established, in arctic North Atlantic areas. Likely more widespread but lurking in collections under P. nudicaule, which has been the collective name for the common introduced poppy in North America—but distinct from P. nudicaule in Beringia, which are native and part of a circumpolar. PEDICULARIS CRENULATA BENTH. (SCROPHULAR- IACEAE).—Beaverhead Co., along a backwater slough NOTEWORTHY COLLECTIONS 67 of the Beaverhead River 1 km SW of Dalys, scattered clumps in a moist meadow with Juncus balticus and Thermopsis montana, 1630 m, TIS R1IOW 82. 6 July 2003, P. Lesica S688 (MONTU); along sloughs of the Beaverhead River just below Clark Canyon Dam, locally abundant in moist meadows with Juncus balticus and Aster occidentalis, 1660 m, TIS RIOW S832. 10 July 2003, P. Lesica 8693 (MONTU, RM), verified by R.D. Dorn (RM). Significance. First report for MT (Dorn 1984, loc. cit.; PLANTS database, loc. cit.), 400 km nw of the nearest location in w WY. PLAGIOBOTHRYS SALSUS (BRANDEGEE) JOHNSTON (BORAGINACEAE).—Beaverhead Co., Centennia Valley, shallow reservoir in an alkaline wetland 5 km sw of Antelope Peak, common in drying mud with Hordeum jubatum and Chenopodium glaucum, 2040 m, T13S R4W S27, 26 Aug 2004 P. Lesica 8984 (MONTU, BYU). Verified by L. Higgins (BRY). Significance. First report for MT (Dorn 1984, loc. cit.; PLANTS database, loc. cit.), 500 km nw of the nearest station in ne NV. Possibly introduced by livestock. SYMPHYOTRICHUM MOLLE (RYDB.) NESOM (AS- TERACEAE).—Big Horn Co., Bighorn Range, near head of Line Creek immediately n of the WY state line and Bighorn N.F. Road 11, 5.5 km n of Sheep Mtn. and 2.5 km se of Windy Point Lookout, ungrazed Deschampsia cespitosalElymus trachycaulus meadow on slightly east-dipping slope of limey clay-loam with scattered surface rock of whitish limestone, 2740 km, TIS R31E SW1/4SW1/4 SE1/4 S34, 28 July 2001. W. Fertig 19787 (MONTU, RM). Significance. First report of this Bighorn Mtn. endemic for MT (Dorn 1984, loc. cit.; PLANTS database, loc. cit.). —PETER LESICA, Herbarium, Division of Biological Sciences, University of Montana, Missoula, MT 59812, peter.lesica@mso.umt.edu; and WALT FERTIG, 1117 W. Grand Canyon, Kanab, UT 84741. OREGON LIMBELLA FRYEI (R. W. WILLIAMS) OCHYRA (AM- BLYSTEGIACEAE, BRYOPHYTA).—Curry Co., Floras Lake, north shore, 42°54’'11"N, 124°30'14’W, elevation 3 m, on organic debris and basal stems of vascular plants over dune sand, in dense vegetation of three types, 1) weedy shoreline dominated by Lotus corniculatus, Carex obnupta, and Potentilla anserina ssp. pacifica, with a scattering of Salix hookeriana seedlings and associated bryophytes: Fontinalis antipyretica, Kindbergia praelonga, Chiloscyphus pallescens, Riccardia chamaedryfolia, and Calliergonella cuspidata, 2) a swampy area dominated by Comarum palustre and Carex obnupta with Calliergon cordifolium and 3) a Salix hookerianal Vaccinium uliginosum/Carex obnupta swamp with Sphagnum squarrosum, 17 June 2008, D.H. Wagner m2402, m2409a, m2410 (OSC, UC, NY). Previous knowledge. Endemic to Oregon, long known from a single locality at Sutton Lake, Lane Co. Presumed extirpated from the original, type locality in Coos Co. (J. A. Christy 2000. Limbella. Bryophyte Flora of North America, Provisional Publication, 68 MADRONO Missouri Botanical Garden. http://www.mobot.org/ plantscience/BFNA/v2/AmbILimbella.htm). Significance. Second locality for this extremely rare moss that has been searched for extensively, ca. 140 km south of the other extant population. The habitats in which this was observed are different from the Sutton Lake locality and provide new search images for future exploration. HAPLOMITRIUM HOOKERI (SMITH) NEES’ (HAP- LOMITRIACEAE, MARCHANTIOPHYTA).—Lane County, Sutton Beach, on moist sand next to Sutton Creek, solitary shoots, both male and female with sporophytes, mixed most often in colonies of Phaeo- ceros carolinianus but also with Anthoceros fusiformis, Blasia pusilla, Aneura pinguis, Cephalozia bicuspidata, Cephaloziella hampeana, Jungermannia rubra, and Pohlia annotina, 44°04'05"N, 124°07'25”W, elevation 1m, 18 June 2008, D. H. Wagner m2414d (OSC, UC). Previous knowledge. In Oregon known from a single site in the Cascade Mountains, Three Sisters Wilder- ness, Lane Co. and in Washington State known from a single site in the northern Cascades, otherwise circum- boreal at high latitudes, everywhere rare. Significance. Second collection in Oregon, 188 km west of previously known site; first record from sea level in North America (S. Bartholomew-Began 2001. Haplomitriaceae. Bryophyte Flora of North America, Provisional Publication, Missouri Botanical Garden. http://www.mobot.org/plantscience/BFN A/v3/ Hap]Haplomitriaceae.htm). —DAVID H. WAGNER, Northwest Botanical Insti- tute, P.O. Box 30064, Eugene, OR 97403-1064. davidwagner@mac.com. WASHINGTON DRYOPTERIS CRISTATA (L.) Gray (DR YOPTERIDA- CEAE).—Pend Oreille Co., on logs or marshy margin, over limestone, 0.5 air km S of Upper Lead King Lake, 48°56.4'N, 117°21.3'W, 770m, 28 Aug 2007, Zika 23515 (WTU); swampy small cove, Lake Leo, 48.6°N, 117.4°W, 1000 m, 6 Sep 2007, Arnett 07-132 & Giblin (WTU). Previous knowledge. This circumboreal species ranges in North America from Montana to Newfoundland, south to North Carolina. The Flora of North America does not include Washington within the range for this species (J.D. Montgomery and W.H. Wagner Jr., Dryopteris, Flora of North America, Volume 2, page 285). The first report of D. cristata for Washington was 1965, and the Washington Natural Heritage Program currently tracks 20 populations of this species in Stevens and Pend Oreille Counties (http://wwwl.dnr.wa.gov/ nhp/refdesk/lists/plantrnk.html). Its restricted occur- rence in the northeast portion of the state is the primary factor responsible for its assigned conservation status of S2 (Imperiled—6 to 20 occurrences, very vulnerable to extirpation) in Washington. Dryopteris cristata also has been collected in northern Idaho (FD. Johnson, s.n., Chase Lake, Bonner County (ID); S.J. Brunsfeld 2013, Perkins Lake, Boundary County (ID)), occurrences similarly omitted by J.D. Montgomery and W.H. Wagner Jr. Significance. These collections confirm the presence of this species in Washington, and suggest that [Vol. 56 additional populations are likely to be found through additional survey work in this undercollected area of Washington. Future floristic treatments for D. cristata in North America should include Washington and Idaho as part of the range of this species. JUNCUS BREVICAUDATUS (Engelmann) Fernald (JUNCACEAE).—Pend Oreille Co., Lake Leo boat ramp, 48.649°N 117.496°W, 1000 m, 6 Sep 2007, Giblin 1164 & Arnett (WS, WTU); Frater Lake, 48.654°N, 117.483°W, 977 m, 7 Sep 2007, Arnett 07-150 & Giblin (WS, WTUV); Stevens Co., Little Twin Lakes, 48.574°N, 117.644°W, 1127 m, 4 Sep 2007, Arnett 07-100 & Giblin (WS, WTU). Previous knowledge. The species has been reported from Minnesota to Newfoundland, south to Pennsyl- vania, with a narrow band extending south along the Appalachian Mountains to Tennessee. Disjunct popu- lations are known from Arizona, Colorado, Utah, Wyoming, Alberta, and coastal areas of Oregon and British Columbia. Significance. These are the first records from the interior Pacific Northwest. Each of our sites are in the Colville National Forest, and specimen identifications were confirmed by Peter Zika (WTU). The previous Oregon and British Columbia collections were made in coastal areas under cranberry cultivation, suggesting that this species was introduced by farmers (P. F. Zika, 2003, Journal of the Torrey Botanical Society. 130: 43— 46). The collections here raise the possibility that this species may be more widespread throughout western North America than previously reported. It is unclear whether the current distribution pattern of J. brevicau- datus from the interior West results from _ true disjunction, is the artifact of undercollection, or is due to specimen misidentification. —DAviID E. GIBLIN, University of Washington Herbarium, Box 355325, Seattle, WA 98195-5325; and JOSEPH L. ARNETT, Washington Natural Heritage Program, P.O. Box 47014, Olympia, WA 98504-7014. dgiblin@u.washington.edu. WASHINGTON PELTANDRA VIRGINICA (L.) SCHOTT (ARACEAE).— Whatcom Co., about 8 km west of Ferndale on the west shore of Lake Terrell, 48°51.748’N 122°41.943'W, elevation 65 m. Growing on shore in shallow water to about 20 cm deep, associated with Typha latifolia L., with dense Phalaris arundinacea L. higher on the shore. 12 July 2007, A. Fullerton and S. Boothe (WTU). Previous knowledge. Peltandra virginica is native to the eastern United States and southeast Canada and is most common on the Atlantic Coastal Plain. Its range has expanded to the Midwestern states over the last 30 yr. Introduced populations are recorded in Califor- nia and Oregon as far north as the south shore of the Columbia River in Clatsop County (S. A. Thompson, Araceae, pp. 128-142 in Flora of North America Editorial Committee, Flora of North America North of Mexico, vol. 22, 2000; WTU Herbarium on-line database www. washington.edu/burkemuseum/collections/ herbarium/index.php accessed 6 February 2008). Significance. This is the first record for Washington State. The small population is located on the undevel- 2009] oped shore of Terrell Lake, a Wildlife Area managed by the Washington Department of Fish and Wildlife. Peltandra virginica is available in the horticulture trade, however due to the remote nature of the site we feel this population more likely grew from seed transported by waterfowl from locations where the plant persists in Oregon or California. HIBISCUS MOSCHEUTOS L. (MALVACEAE).—Doug- las Co., 1 km northwest of the town of Rock Island on the northeast shore of Big Bow Lake, 47°23.08'N 120°9.29'W 19, elevation 201 m. Several plants growing in shallow water and just above the waterline along the east end of the lake; associated with Schoenoplectus acutus (Muhl. ex Bigelow) A. & D. Love, Typha latifolia L., and Phalaris arundinacea L. September 2005, J. Parsons (WTUV). NOTEWORTHY COLLECTIONS 69 Previous knowledge. The native range of Hibiscus moscheutos includes Ontario, Canada, the eastern United States east of the Mississippi River and across the southern states to New Mexico (United States Department of Agriculture PLANTS database, ac- cessed February 7, 2008 www.plants.usda.gov). Significance. This is the first record for Washington State. Hibiscus moscheutos is widely available in the horticulture trade. This population 1s likely an escaped garden ornamental. The population has been observed for four years and appears to be persisting and expanding slowly among other native an introduced shoreline species. —JENIFER PARSONS and ARLINE FULLERTON, Washington Department of Ecology, 15 W. Yakima Ave, Suite 200, Yakima, WA 98902. jenp461(@ecy.wa.gov. Volume 56, Number 1, pages 1-69, published 31 August 2009 SUBSCRIPTIONS — MEMBERSHIP Membership in the California Botanical Society is open to individuals ($35 per year; family $40 per year; emeritus $27 per year; students $27 per year for a maximum of 7 years). Late fees may be assessed. Members of the Society receive MADRONO free. Institutional subscriptions to MADRONO are available ($70). Membership is based on a calendar year only. Life memberships are $750. Applications for membership (including dues), orders for subscriptions, and renewal payments should be sent to CBS c/o Jepson Herbarium, University of California, Berkeley, CA 94720-2465. Requests and rates for back issues, changes of address, and undelivered copies of MA- DRONO should be sent to the Corresponding Secretary. INFORMATION FOR CONTRIBUTORS Manuscripts submitted for publication in MADRONO should be sent to the editor preferably as Microsoft Word (.doc), Rich Text Format (.rtf), or Portable Document Format (.pdf) files. 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. For all articles and short items (NOTES, NOTEWORTHY COLLECTIONS, POINTS OF VIEW, etc.), fol- low 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, five 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 review copies of illustrations. The title page should have a running header that includes the name(s) of the author(s), and a shortened title. Avoid foot- notes except to indicate address changes. Abbreviations should be used sparingly and only standard abbreviations 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 are encouraged to include the names, addresses, and e-mail addresses of two to four potential reviewers with their submitted manuscript. Authors of accepted papers are required to submit an electronic version of the manuscript. Microsoft Word 2000 or later or WordPerfect 9.0 (or later) 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. 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 Index Herbariorum. 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 au- thors 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 manuscript have been used in other papers that are published, in press, submitted, or soon to be submitted elsewhere. SMITHSONIAN INSTITUTION LIBRARIES CeCe eo eee eee eee eH ESET E EER EE HEE EO SHEE EEO SHES ETE EEE EEE EH EH EE ESET EEE EEET ESE SETH ESES ESE ET ES EEE SES ESS EO ESE ESO LEDS ASHINGTON stn bihial decescceccscosesccpetageaihee shaspiteUNO’s wesc sisiceleoslassisise sins cslisiesiisa@ets siete eee CONTENTS IEW SPECIES VOLUME 56, NUMBER 2 APRIL-JUNE 2009 MADRONO A WEST AMERICAN JOURNAL OF BOTANY THE EFFECTS OF REVETMENT ON STREAMSIDE VEGETATION IN SEQUOIA SEMPER VIRENS (TAXODIACEAE) FORESTS Will Russell Gnd SAVQKG Terdda ..éc.scsseicasenB0y es xetesisss¢eiePessllasiessssanscDeveesseees 71 Dust DEPOSITION EFFECTS ON GROWTH AND PHYSIOLOGY OF THE ENDANGERED ASTRAGALUS JAEGERIANUS (FABACEAE) Upekala C. Wijayratne, Sara J. Scoles-Sciulla and Lesley A. Defalco .... 81 EFFECTS OF FIRE AND GROUNDWATER EXTRACTION ON ALKALI MEADOW HABITAT IN OWENS VALLEY, CALIFORNIA Daniel W. Pritchett And Sara J. MANNING .cccccccccccccccccccccsccseccccccccceeseeeeseees 89 AARCEUTHOBIUM RUBRUM (VISCACEAE) IN MEXICO Robert L. Mathiasen, Carolyn M. Daugherty, and Brian P. Reif ............. 99 NOTES ON CALIFORNIA MALVACEAE INCLUDING NOMENCLATURAL CHANGES AND ADDITIONS TO THE FLORA STEVEN TILT. 5.000.000 eG Se Os Eid SE oc cesceccoceveceevecees 104 HISTORICAL, NOMENCLATURAL, AND DISTRIBUTIONAL NOTES ON Two PACIFIC CoasT KELPS: LESSONIOPSIS LITTORALIS AND PLEUROPHYCUS GARDNERI (PHAEOPHYCEAE, LAMINARIALES, ALARIACEAE) PaAUC? Silva. iE hoc Wh cca cee MUU SUA ccc eceseooevenes 12 ARCEUTHOBIUM ABIETINUM SUBSPECIES WIENSII, A NEW SUBSPECIES OF FrR DWARF MISTLETOE (VISCACEAE) FROM NORTHERN CALIFORNIA AND SOUTHERN OREGON Robert L. Mathiasen and Carolyn M. DaughePty......ccccccccccccccccceeeeeetttees 118 A NEW SPECIES OF STREPTANTHUS (BRASSICACEAE) FROM TRINITY COUNTY, CALIFORNIA Thomas W. Nelson and Jane P. NeISONn...cicccccccccccceseccccseeccccusescccuecccceueeeess 127 SE NIETORINIDS caer oc Bu oeeet gata sae tans aataie a abuse ecRcaksMeacttuia ate cuccuie unt nee etn ates ies eeieoe tances 130 MD AION ete Sere seca notac anenede ate ces snen sc cenuine fe tase cates aatsa gut vonnaes wentuceneaa rh. tareeuanceesSeas 130 ORE GOIN tcc rem acre rseeie cess So ccreed Navargeconees wametesat aes suee Selauuset aun vt aeess saesoe tone eeaicncsase Nesaee 131 INN ANAC) NE LIN ag eee Flag Ms sae ata ogre ok Sau tinea Si meaisieateal ec Betzo ato tnopbamtaned ue aeeannaaeae ene aaaee: 134 INTE OC CO) Fem sn edna os acto soat ez os sschsoe oe Sasa wna cic eee Me nas lactate aaa ee Seana NNVSEiA cestenouecaeds cares 135 PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL SOCIETY MADRONO (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 MADRONO, Kim Kersh, Membership Chair, Uni- versity and Jepson Herbarium, University of California, Berkeley, CA 94720-2465. kersh@berkeley.edu. Corresponding Editor—TIMOTHY LOWREY Copy Editor—RICHARD WHITKUS Museum of Southwestern Biology Department of Biology MSC03 2020 Sonoma State University University of New Mexico 1801 E. Cotati Avenue Albuquerque, NM 87131-0001 Rohnert Park, CA 94928-3609 madrono@unm.edu whitkus @ sonoma.edu Book Editor—JOn E. KEELEY Noteworthy Collections Editors—DIETER WILKEN, MARGRIET WETHERWAX Board of Editors Class of: 2009—DOoNoOvVAN BAILEY, New Mexico State University, Las Cruces, NM MARK BorcHerrt, USFS, Ojai, CA 2010—FReED Hrusa, California Department of Food and Agriculture, Sacramento, CA RICHARD OLMSTEAD, University of Washington, Seattle, WA 2011—JAMIE KNEITEL, California State University, Sacramento, CA KEVIN RIcE, University of California, Davis, CA 2012—GRETCHEN LEBUHMN, San Francisco State University, CA ROBERT PATTERSON, San Francisco State University, CA CALIFORNIA BOTANICAL SOCIETY, INC. OFFICERS FOR 2009-2010 President: Dean Kelch, Jepson Herbarium, University of California, Berkeley, CA 94720, dkelch @berkeley.edu First Vice President: Andrew Doran, University and Jepson Herbaria, University of California, Berkeley, CA 94720, andrewdoran @ berkeley.edu Recording Secretary: Nishanta Rajakaruna, Department of Biological Sciences, San Jose State University, One Washington Square, San Jose, CA 95192, Nishanta.Rajakaruna@sjsu.edu Corresponding Secretary: Heather Driscoll, University Herbarium, University of California, Berkeley, CA 94720- 2465, heather.driscoll @nature.berkeley.edu Treasurer: Thomas Schweich, California Botanical Society, Jepson Herbarium, University of California, Berkley, CA 94720, tomas @schweich.com The Council of the California Botanical Society comprises the officers listed above plus the immediate Past President, Michael Vasey, Department of Biology, San Francisco State University, 1600 Holloway Avenue, San Francisco, CA 94132, mvasey @sfsu.edu; the Editor of Madrono; three elected Council Members: James Shevock, Department of Botany, California Academy of Sciences, 55 Music Concourse Drive, San Francisco, CA 94118, jshevock@ nature.berkeley.edu; Roy Buck, Jepson Herbarium, University of California, Berkeley, CA 94720, roybuck@ email.msn.com; Chelsea Specht, Department of Plant and Microbial Biology, University of California, Berkeley, CA 94720-2465, cdspecht@nature.berkeley.edu. Graduate Student Representatives: Ben Carter, Department of Integrative Biology and University Herbarium, University of California, Berkeley, CA 94720-2465, bcarter@ berkeley.edu. Webmaster: Susan Bainbridge, Jepson Herbarium, University of California, Berkeley, CA 94720- 2465, sjbainbridge @ berkeley.edu. This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). | , | | | DE VEGETATION IN SEQUOIA SEMPERVIRENS (TAXODIACEAE) FORESTS WILL RUSSELL AND SAYAKA TERADA Department of Environmental Studies, San Jose State University, San Jose, CA 95192-0115 will.russell@sjsu.edu ABSTRACT Stream-bank stabilization structures, or revetment, can impact riparian systems by confining watercourses to the point where natural functions are limited. Removal of existing revetment may have considerable impacts to stream-bank stability however. The effects of revetment and revetment removal on vegetation and stream-bank morphology were measured on three streams in forests dominated by Sequoia sempervirens (Coast Redwood) in northern California. Data were collected using randomly located transects on three treatments; “‘no-revetment,” “‘revetment-intact,” and “revetment removed.” Results were compared between treatments using ANOVA at a 0.05 level of significance. On all three sites, species richness, vegetation cover, and tree seedling density were found to be highest where no revetment existed compared to where revetment was intact. Stream depth and stream-bank slope were highest where revetment was intact. Recovery of vegetation following removal of revetment was more site specific being most pronounced where recovery time was greatest and stream-bank restoration efforts were highest. Key Words: Coast redwood, revetment, Sequoia sempervirens, stream restoration. Stream-bank stabilization structures, known as revetment, are common in areas where roads and trails are in close proximity to a watercourse. These structures are designed to support the stream-bank, to protect roads and trails, and to reduce erosion from the stream-bank into the stream. In recent years land managers have become concerned with the ecological ramifica- tions of these protective measures, and have considered the removal of revetment on a limited scale. One of the sites where revetment removal is currently being considered is Redwood Creek at Muir Woods National Monument. There is concern, however, that the process of removing revetment may pose risks to streamside vegeta- tion. This study is an attempt to provide basic information on the relative impacts of revetment installation and removal on riparian vegetation. This study is limited to streams within the northern California Sequoia sempervirens (coast redwood) forest type in order to provide a reasonable comparison with Redwood Creek. Revetment is constructed out of a variety of materials, including rock structures often referred to as riprap, rock filled wire structures known as Gabion-baskets, structures made of logs known as cribbing or bank-armor, or stacks of cement bags. Installation of these structures on stream- banks tends to slow erosion, reducing the amount of fine sediment input into adjacent stream channels (Harvey and Watson 1989; Engber 2002). In some cases revetment is installed where stream-banks have started to fail, while in others, revetment is installed as a preemptive measure. While revetment protects stream-banks, it also confines stream channels to the extent where natural functions may be limited. Watercourses, by nature, are dynamic systems (Kauffman et al. 1997). Reinforcing a stream-bank confines the movement of the stream, and consequently limits natural fluvial processes such as the development of flood plains, pools, and point bars (Knudsen et al. 1987; Smith et al. 1993). In addition, confining a stream tends to deepen its channel over time, which lowers the water table and eliminates the deposition of nutrient rich silt in the riparian zone. Revetment also reduces the input of gravel into a stream, which is important for the formation of spawning beds for salmonid species (Flosi and Reynolds 1994), and also negatively impacts other wildlife species (Chapman and Knudsen 1980; Hortle and Lake 1983; Knudsen et al. 1987; Crispin and Roberts 1993; Quinn and Peterson 1996). Alterations to natural stream processes have negative effects on plants as well. Fluvial processes are known to affect the recruitment and succession of riparian plant communities and other ecological processes in a number of ways (McBride and Strahan 1984a, b; Buer et al. 1989; Lisle 1989; Minore and Weatherly 1994; Maho- ney and Rood 1993). Point bars, which can be negatively effected by the construction of revet- ment, provide favorable conditions for the recruitment of riparian trees such as alder and willow (Russell and McBride 2001; Russell et al. 2002). In addition, periodic silt deposition may give S. sempervirens a competitive edge over other coniferous species (Stone and Vasey 1968). Removal of revetment is intended to return a stream to a more natural state and to promote inherent fluvial processes. However, effects of 72 MADRONO revetment, and revetment removal on riparian vegetation are not well understood. In particular, there is concern that increased levels of stream- bank erosion following revetment removal could endanger existing trees, and limit the recovery of riparian vegetation. Because of the direct impacts that revetment installation has on stream-bank morphology and likely repercussions for stream-bank vegetation, this study has been designed to test the hypoth- eses that the diversity, abundance, and recruit- ment of vegetation would be highest where revetment has not been installed, moderate where it has been removed, and lowest where it is intact; and that stream-bank slope, and stream depth would be lowest where revetment has not been installed, moderate where it has been removed, and highest where it is intact. METHODS The diversity, abundance, and recruitment of streamside vegetation, along with stream channel characteristics, were analyzed on three streams in S. sempervires forests in northern California. Transects were randomly located within three treatments, where revetment was in place “‘revet- ment-intact,”’ where no revetment was present ‘“‘no-revetment,”’ and where revetment had been removed “‘revetment-removed.” Data was ana- lyzed within and between study sites. Study Sites Three study sites were sampled for comparti- son. All of the sites were located in northwest California in the S. sempervires forest on federal and state lands. The number of possible sites was limited, as very few streams in this vegetation type have had revetment removed. The first site, Redwood Creek, is a perennial stream in Marin County, California, 20-kilometers north of San Francisco. Approximately 1.5-kilometers of the total 5.8-kilometer length of Redwood Creek is within the boundaries of Muir Woods National Monument. A dense forest composed of S. sempervirens and Umbellularia californica (Cali- fornia bay-laurel) shades the stream within the monument, with the exception of the extreme down stream portion where A/nus rubra (red alder) dominates. The Civilian Conservation Corp installed stone revetment on 26 sections of the stream within Muir Woods during the 1930s. The revetment has survived mostly intact to the present time, with only a few sections having fallen into the streambed. Two treatments, ‘“‘revetment-intact’”’ and “‘no-revetment,’’ were designated on this site for sampling. The “‘revet- ment-removed” treatment was not sampled on this site, as no-revetment removal projects had been implemented at the time of this study. [Vol. 56 The second site, Little River, is a perennial stream located in Mendocino County, California, approximately 0.5-kilometers north of the town of Little River. The entire length of the stream is located within Van Damme State Park. The vegetation is dominated by Sequoia sempervires with an understory of Tsuga heterophylla (west- ern hemlock) and A/nus rubra (red alder) common near the mouth of the stream. Older sections of riprap that were constructed in the 1930’s are visible along the stream from three to six kilometers inland from the coast. Newer sections of revetment, installed in the 1960’s and constructed primarily of gabion baskets, are in place adjacent to road crossings closer to the coast. A single large section of these gabion basket structures, 60-meters in length, was removed on the north side of the stream in 1996. Significant effort was taken to facilitate vegetation recovery following revetment removal including the installation of large woody debris to inhibit bank erosion, and planting of stabilizing vegetation (Lowe et al. 1996). The third site, Fife Creek, passes for approx- imately 2.5-kilometers through Armstrong Red- woods State Reserve in Sonoma County, Cali- fornia. It is a perennial stream, but drops below the surface within the park boundaries in some areas during the dry summer months. The section of stream used for this study was the farthest inland of the three sites. The canopy is dominated by Sequoia sempervires, Umbellularia californica, and Lithocarpus densiflorus (tan-oak). A series of check dams flanked by sections of revetment were installed along the course of the stream within park boundaries in the 1960s. Riprap, cribbing, and cement bags were used adjacent to road crossings, and on unstable portions of stream- bank. Approximately 600-meters of check dams, and adjacent revetment, were removed in 1998. Some effort was taken to protect stream-banks. Vortex weirs and woody debris were installed following revetment removal. However, efforts to facilitate vegetation recovery and protect stream- banks were modest compared to the Little River revetment removal project. Significant increased erosion and head cutting has taken place since removal. Sampling Methods A total of 120 meter-wide transects were used to collect data, with 15 randomly placed transects | within each treatment on each site. Transects | were separated by a minimum distance of 5- | meters. The Fife Creek site included three | treatments (revetment-removed, revetment-in-_ tact, and no-revetment) for a total of 45 transects | (Fig. 1). The Little River site included the same | three treatments for a total of 45 transects | (Fig. 2). The Redwood Creek site included two | 2009] RUSSELL AND TERADA: EFFECTS OF REVETMENT ON STREAMSIDE VEGETATION Ei Fife Creek Sampling Location | 2003 x No-revetment & Revetment-intact @ Revetment-removed 1 rr docino | BA < ence 2 Area Tn, Enlarged ss, = ey. oh . .*Guernevilic # +} ; he | by ca, . 4 : = 1 © t Mult Woods 4 & Sea, Francisco ‘ Fic. 1. Location of sample points on Fife Creek in Armstrong Redwoods State Park, California. *Saae. sewer” | Little River Sampling i cotion 2003 - - ed treatments (revetment-intact, and no-revetment) for a total of 30 transects (Fig. 3). Transects originated at the edge of the stream-bank at the level of summer base flow (defined as the lowest annual stream level) and continued, perpendicu- lar to the stream, to the height of the second terrace (approximately 1.5-meters above the elevation of the height of summer base flow; Fig. 4). The overall length of each transect varied between 2 to 10-meters depending on the slope of the streambank. Three sampling approaches were employed along each transect. Data were collected at the point of origin of the transect, within the 1-meter width of the transect, and along the length of the transect using the line intercept method (Kaiser 1983). Data collected at the point of origin included stream depth, stream-bank slope (mea- sured with an Abney level), bank full width, channel slope, and tree canopy cover (estimated using a spherical densiometer). Within the 1- meter width of each transect, the occurrence of all vascular plant species was recorded. The percent cover of each understory species was determined using ocular estimates. The basal area of each tree was calculated from measurement of diam. taken 1.4-meters from the base the up-hill side of the tree. The number and species of all tree seedlings (trees less than 1-meter in height) were recorded. The line intercept method was used to record cover type (e.g. herbaceous plants, shrubs, trees, litter, bare ground) at 10-centimeter inter- vals along the center of each transect. In addition, elevation of the stream-bank was measured at the Aan doc itic 7 Area | Enlarged A < samevitla 4 4, i ™ Mult Wioodt@ one ~* -- “« -.- 4 Revetment-intact e Revetment-removed P Fic. 2. | | | x No-revetment | | | | | | | 490 Meter Base Map. LSGS Topo Map, 1978 Location of sample points on Little River in Van Damme State Park, California. 74 MADRONO | Mendocino = \ ae te qovemville ‘ Area Enlarged near rer eer wsserresesaametireaierarenee peetostarenaanri Teta reeeT EEA | Mui Woods ¢ : a | (Sar Francisco | aes i | i | SRE SURE isc woe Santa eon ew mR anc, cS OE FIG. 3. same 10-centimeter intervals with reference to the point of origin. Analysis of Variance (ANOVA) was used to examine variables between treatments. Fisher’s LSD was conducted as a post-hoc test to determine significant differences between treat- ments at a significance level of 95 percent. RESULTS Species Richness The diversity of species was found to be significantly different between treatments. For all sites combined, a total of fifty-one species were Summer base flow [Vol. 56 a eee ay Redwood Creek | Sampling Location, | 2003 x No-revetment 4 Revetment-intact fy 200 4ao Melor Fareernereeemeeereneecnnn iets NiSSe-SO SPSS SSO JSGS Topo Map, 1998 Location of sample points on Redwood Creek in Muir woods National Monument, California. found “‘revetment-intact”’ transects, fifty-eight on ‘“‘no-revetment”’ transects, and sixty species on ‘‘revetment-removed”’ transects (Appendix 1). On Little River a significantly lower species richness was found on “revetment-intact” than “‘no- revetment” and “‘revetment-removed” (Fig. 5). Species richness was not statistically different between “‘no-revetment’’? and ‘“‘revetment-re- moved” (P < 0.01, P < 0.01, P = 0.40). In contrast, data from Fife Creek indicate that “‘no- revetment” had higher species richness than both “‘revetment-removed”” and “revetment-intact.”’ Species richness for “‘revetment-removed”’ and ‘‘revetment-intact”’ were not statistically different (P < 0.01, P < 0.01, P = 0.34). On Redwood Transect Point of origin Fic. 4. Cross section of stream channel with transect origin at the level of summer base flow, terminating at the second terrace (approximately 1.5-meters above summer base flow level). 2009] Little River revetment intact peels scons removed no revetment 2 ca 6 8 10 12 Oo Fife Creek revetment intact Ce, a 3 removed |! no revetment oO NO BSS Redwood Creek revetment | intact 0 2 4 6 8 species/plot FIG. 5. Richness of plant species (mean number of species + standard error) per transect on three streams in the northern California Sequoia sempervirens forest. Treatments sharing the same lower-case letter within the same chart were not significantly different (P > 0.05). Creek, where only two treatments occurred, species richness was significantly higher for “‘no- revetment” than “‘revetment-intact’’ (P = 0.03). Vegetation Cover The total cover of vegetation measured with the line intercept technique was higher where no revetment had been installed for all three sites (Fig. 6). On Little River the percent cover of vegetation was significantly higher for ‘“‘no- revetment” and “‘revetment-removed” compared ito “revetment-intact” (P < 0.01, P < 0.01, P < 0.71). On Fife Creek, percent cover was signifi- cantly higher for “‘no-revetment” than “‘revet- -ment-removed” and ‘‘revetment-intact.’’ No sig- nificant difference was found between “revetment-removed”’ and “‘revetment-intact” (P =< 0.01, P < 0.01, P = 0.50). On Redwood Creek, cover was significantly higher for “‘no-revetment”’ / than ‘“‘revetment-intact” (P < 0.01). RUSSELL AND TERADA: EFFECTS OF REVETMENT ON STREAMSIDE VEGETATION 75 Little River ve ot a 2 intact ve, i ° removed no revetment Fife Creek revetment a intact La revetment removed no revetnhent Pa .- b - Redwood Creek revetment intact a - no revetment ee b 20 40 % cover FIG. 6. Percent cover of vegetation (+standard error) on three streams in the northern California Sequoia sempervirens forest. Treatments sharing the same lower-case letter within the same chart were not significantly different (P > 0.05). Tree Density and Canopy Cover The greatest number of tree seedlings was found where no revetment had been installed on all sites (Fig. 7). The average density of tree seedlings was statistically different between “‘re- vetment-intact”” and both “‘revetment-removed” and “‘no-revetment”’ on Little River. No signifi- cant difference was found between “‘revetment- removed” and ‘“‘no-revetment” (P < 0.01, P < 0.01, P = 0.11). On Fife Creek, tree seedling density was statistically higher for ‘‘no-revet- ment’? compared with “‘revetment-intact” and ‘“‘revetment-removed.” There was no significant difference between “‘revetment-intact” and “‘re- vetment-removed (P = 0.43, P < 0.01, P < 0.01). There was also no significant difference between treatments on the Redwood Creek site (P = 0.09). Mixed results were found for the canopy cover of trees on the three sites (Fig. 8). Little River 76 MADRONO “Little River revetme 2 ea piarllicts iat b removed | no revetment i — > | Fife Creek revetment | intact 4 revetment | removed } ) revetment intact | ino revetment 0.5 ‘Seedlings/plot Fic. 7. Mean density of seedlings per sample (+stan- dard error) on three streams in the northern California Sequoia sempervirens forest. Treatments sharing the same lower-case letter within the same chart were not significantly different (P > 0.05). was not significantly different between “‘revet- ment-intact” and “‘no-revetment.” “‘Revetment- removed” was significantly lower than both of the other treatments, however (P = 0.04, P = 0.18, P = 0.02). On Fife Creek ‘‘no-revetment”’ had significantly higher tree canopy cover than both ‘“‘revetment-intact’’ and ‘‘revetment-re- moved.” No significant difference was found between “‘revetment-intact”’ and “‘revetment-re- moved” (P = 0.16, P = 0.02, P = 0.02). Tree canopy cover was not significantly different between treatments on Redwood Creek (P = 0.32). Stream-bank Morphology Stream-bank slope was higher for “‘revetment- intact’ compared to “‘no-revetment”’ on all sites (Fig. 9). On Little River, “‘revetment-removed” was not significantly different than “‘no-revet- ment,’ both of which were significantly lower [Vol. 56 ; revetment | intact ! : revetment removed ‘no revetment rr 80 100 ; Fife Creek revetment | intact yeeros —- removed no revetment | : ~ Redwood Creek oct intact e ino revetment % cover Fic. 8. Percent cover of trees (+standard error) on three streams in the northern California Sequoia sempervirens forest. Treatments sharing the same lower-case letter within the same chart were not significantly different (P > 0.05). than “‘revetment-intact” (P < 0.01, P< 0.01, P=. 0.44). On Fife Creek, stream-bank slope was not | statistically different between “‘revetment-re- | moved” and “‘revetment-intact,’’ which were both significantly higher than “‘no-revetment” (P = 0.62, P < 0.01, P = 0.03). On Redwood Creek, the stream-bank slope was significantly higher on. ‘“‘revetment-intact’’ compared to ‘“‘no-revetment”’ (P=0:01); | Stream depth was higher for “revetment-, intact’’ compared to “‘no-revetment”’ for all sites (Fig. 10). On Little River “‘no-revetment” was. significantly higher than “‘revetment-intact.”’ No significant difference was found between other treatments (P = 0.23, P < 0.01, P = 0.09). Stream. depth was significantly higher for “‘revetment- intact” and ‘‘revetment-removed”’ compared to ‘“‘no-revetment”’ on Fife Creek. No difference was found between “‘revetment-intact’” and “‘revet- ment-removed”’ (P = 0.36, P < 0.01, P < 0.01). On Redwood Creek stream depth was signifi- 2009] RUSSELL AND TERADA: EFFECTS OF REVETMENT ON STREAMSIDE VEGETATION ae Little River revetment- intact revetment- | removed | ino-revetment revetment- | intact revetment- removed ‘no-revetment Redwood Creek a a revetment- intact 'no-revetment FIG. 9. Stream-bank slope (+standard error) on three streams in the northern California Sequoia sempervirens forest. Treatments sharing the same lower-case letter within the same chart were not significantly different (P => ().05). ee cantly higher for “‘revetment-intact” than “‘no- frevetment’ (P = 0.03). DISCUSSION The results of this study support the hypoth- eses that revetment negatively impacts both vegetation and stream bank morphology. Species richness, vegetation cover, and tree recruitment were highest where there was no revetment compared to where revetment was intact on all three study sites. Stream-bank slope and stream depth were also higher where revetment was intact compared to where revetment had not been Installed. The effect of removal of revetment on recovery of vegetation was not as clear however, as significant differences were apparent between Sites. On Little River, vigorous vegetation recovery ~Was apparent where revetment had been re- Little River revetment- intact og ee a removed 10 20 30 j=) Fife Creek revetment- intact a oe TS — 2 removed |! a — no-revetment () 10 20 30 Redwood Creek revetment- intact no-revetment Ree ae 0 5 10 15 10. FIG. streams in the northern California Sequoia sempervirens forest. Treatments sharing the same lower-case letter within the same chart were not significantly different (P = 09): Stream depth (+standard error) on three moved. In contrast, little evidence of vegetation recovery was found on Fife Creek following revetment removal. The difference in recovery between Fife Creek and Little River was likely due in part to differences in recovery time, pre- treatment site conditions, and post-treatment restoration effort. Revetment on Little River was removed two years earlier than it was on Fife Creek allowing for more development of vegeta- tion cover. Stream bank conditions were also superior on Little River prior to revetment removal. Habitat alteration was much more extensive on: Fite’ Creek, where-a.serics. of check dams and road crossings were installed along the entire course of the stream (Jackson and Marcus 2004). The recovery of vegetation on Little River was also superior due to a more intensive restoration efforts following revetment removal. Stream-bank slope was lower on Little River, where revetment had been removed, compared to 78 Fife Creek where there was little difference between the restored stream bank slope and the stream bank slope where revetment was still intact. The post-revetment removal stream bank grading that took place on Little River reduced erosion allowed for vegetation to take root (Lowe et al. 1996). In addition, superior post-revetment recovery of vegetation may have been related to better site conditions on Little River where vegetation cover and species diversity were higher overall. Management of riparian areas has significant consequences for vegetation and the wildlife species that are dependent on that vegetation (Russell 2009; Kauffman et al. 1997). Perhaps the most profound of these management impacts comes from alterations of the stream channel itself. Confining streams through the construc- tion of revetments limits natural fluvial processes that create the diversity of habitat necessary to support all of the species inherent in riparian systems. Restoring stream channels that have been disturbed by revetment is, therefore, clearly advantageous, as long as the restoration process does not cause greater disturbance than the initial installation. Removal of revetment, by necessity, exposes stream banks with little or no vegetative cover. Without adequate treatment of freshly exposed stream-banks, vegetation may not only be slowed in recovery, but existing vegetation could be damaged through erosion. The compar- ison of these case studies of revetment removal in the coast redwood forest type suggest that post removal restoration should be planned carefully in the context of existing site conditions in order to facilitate the return of vegetation and stream channel processes to a more natural condition. ACKNOWLEDGMENTS Funding for this project was received from the NPS Golden Gate National Recreation Area and the USGS/ BRD Western Ecological Research Station. Logistical support was provided by Rene Pasquanele of the California Department of Parks and Recreation. LITERATURE CITED BUER, K., D. FORWALTER, M. KISSEL, AND B. STOHLERT. 1989. The middle Sacramento River: human impacts on physical and ecological process- es along a meandering river. Pp. 22-32 in D. L. Abell (ed.), Proceedings of the California riparian systems conference: protection, management, and restoration for the 1990s. General Technical Report PSW-110. U.S. Department of Agriculture, Pacific Southwest Forest and Range Experiment Station, Berkeley, CA. CHAPMAN, D. W. AND E. KNUDSEN. 1980. Channel- ization and livestock impacts on salmonid habitat and biomass in western Washington. Transactions of the American Fisheries Society 109:357—363. CRISPIN, V., R. HOUSE, AND D. ROBERTS. 1993. Changes in instream habitat, large woody debris, and salmon habitat after the restructuring of a MADRONO [Vol. 56 coastal Oregon stream. North American Journal of Fisheries Management 13:96—102. ENGBER, E. 2002. Rebuilding the riparian using bioengineering: Some basic techniques for restor- ing, stabilizing and revegetating riparian areas. Pp. 29-30 in H. Vaughn (ed.), Proceedings of the twentieth annual salmonid restoration federation conference, Ukiah, CA. FLOSI, G. AND F. L. REYNOLDs. 1994. California salmonid stream restoration manual. California Department of Fish and Game, Sacramento, CA. HARVEY, M. D. AND C. C. WATSON. 1989. Effects of bank revetment on Sacramento River, California. Pp. 47—-S0 in D. L. Abell (ed.), Proceedings of the California riparian systems conference: protection, management, and restoration for the 1990s. Gen- eral Technical Report PSW-110. U.S. Department of Agriculture, Pacific Southwest Forest and Range Experiment Station, Berkeley, CA. HORTLE, K. G. AND P. S. LAKE. 1983. Fish of channelized and unchannelized sections of the Bunyip River, Victoria. Australian Journal of Marine and Freshwater Research 34:441—450. JACKSON, D. AND L. MARCus. 2004. Fife Creek in Armstrong Woods State Park monitoring program results and cross-sections. Appendix 4 in L. Marcus (ed.), Russian River creek stewardship: monitoring and assessment summary report. Laurel Marcus and Associates, Oakland, CA. KAISER, L. 1983. Unbiased estimation in line-intercept sampling. Biometrics 39:965—976. KAUFFMAN, J. B., R. L. BESCHTA, N. OTTING, AND D. LYTJEN. 1997. An ecological perspective of riparian and stream restoration in the western United States. Fisheries 22:12—24. KNUDSEN, E. AND S. J. DILLY. 1987. Effects of riprap bank reinforcement on juvenile salmonids in four western Washington streams. North American Journal of Fisheries Management 7:351—356. Liste, T. E. 1989. Channel-dynamic control on the establishment of riparian trees after large floods in northwestern California. Pp. 9-13 in D. L. Abell (ed.), Proceedings of the California riparian sys- tems conference: protection, management, and restoration for the 1990s. Report PSW-110. U.S. Department of Agriculture, General Technical. Pacific Southwest Forest and Range Experiment. Station, Berkeley, CA. Lowe, P., S. ZEMBACH, AND W. VANDIVERE. 1996. | Gabion Removal and Channel Restoration for. Little River, Van Damme State Park. Report to the | California Department of Parks and Recreation, , Mendocino District. Parks and Recreation, Mendocino, CA. MAHONEY, J. M. AND S. B. Roop. 1993. A model foul assessing the effects of altered river flows on the recruitment of riparian cottonwoods. Pp. 32-46 in , B. Tellman, H. J. Cortner, M. G. Wallace, L. F.: DeBano, and R. H. Hamre, (eds.), Riparian management: common threads and shared inter- | ests. General Technical Report RM-226. U.S. Department of Agriculture, Forest Service, Rocky. Mountain Forest and Range Experiment Station, | Fort Collins, CO. McBRIDE, J. R. AND J. STRAHAN. 1984a. Establish- California Department off ment and survival of riparian species on gravel bars » of an intermittent stream. American Midland Naturalist: 112:235-—245: : 2009] RUSSELL AND TERADA: EFFECTS OF REVETMENT ON STREAMSIDE VEGETATION 72 AND 1984b. Fluvial processes and woodland succession along Dry Creek, Sonoma County, California. Pp. 110-119 in R. E. Warner and K. M. Hendrix, (eds.), California Riparian Systems: ecology, conservation, and productive management. University of California Press, Berke- ley. GA: MINORE, D. AND H. G. WEATHERLY. 1994. Riparian trees, shrubs, and forest regeneration in the coastal mountains of Oregon. New Forests 8:249—263. QUINN, T. P. AND N. P. PETERSON. 1996. The influence of habitat complexity and fish size on over-winter survival and growth of individually marked juvenile coho salmon (Oncorhynchus kisutch) in Big Beef Creek, Washington. Canadi- an Journal of Fisheries and Aquatic Sciences 53:1555—1564. the Coastal Redwood region. Forest Ecology and Management 257:1427—-1433. AND J. R. MCBRIDE. 2001. The relative importance of fire and watercourse proximity in determining stand composition in mixed conifer riparian forests. Forest Ecology and Management 150:259-265. , AND K. CARNELL. 2003. Influence of environmental factors on the regeneration of hardwood species on three streams in the Sierra Nevada. Madrono 50:21—27. SMITH, I. dD:.R;. Ci SIDLE,. P2 E. PORTER, AND:J, (R; NOEL. 1993. Effects of experimental removal of woody debris on the channel morphology of a forest, gravel-bed stream. Journal of Hydrology 152:153-178. STONE, E. C. AND R. B. VASEY. 1968. Preservation of RUSSELL, W. 2009. The influence of timber harvest on coast redwood on_ alluvial flats. Science the structure and composition of riparian forests in 159:157-161. APPENDIX | The presence of plant species under three treatments on streams in the northern California Sequoia sempervirens forest. The three sites are abbreviated in the chart as Little River (LR), Fife Creek (FC), and Redwood Creek (RC). No-revetment Acer macrocarpa LR Actae rubra Adeocaulon bicolor RC Adiantum aleuticam RC Adiantum pedatum Alnus rubra LR Anagallis arvernsus Anthyrium flix-femina LR, RC Aralia californica RC Asuram caudatum Boykinia elata LR Bromus vulgaris LR Carex sp. LR, FC, RC Cardamine oligosperma Claytonia perfoliata Claytonia siberica LR, RC Claytonia sp. RC Clintonia andrewsiana Conium maculatum LR Corylus cornutus LR, FC, RC Cystopteris fragilis LR, FC, RC Dactylis glomerata Dicentra farmosa LK Disporum hookerii LR, RC Dryopteris sp. FC Epilobium brachycarpum Epuisetum sp. LR, RC Gallium sp. LR, FC, RC Hedera helix FC Heracleum lanatum LR Hieracium albiflorum — Holocus lanatus Huchera pilosissima FC Huchera sp. LR, RC Hydrophyllum occidentales LR Juncus sp. _ Lithocarpus densiflorus POSKC Lolium sp. Marah oreganus LR Mitella caulescens Revetment-intact Revetment-removed RC FC RC RC LR LR LR FC LR, RC LR LR LR LR, RC LR LR LR LR, RC LR LR FC FC LR LRORC LR RC RC LR, RC LR LR LR RC LR LR FC LR. ORC LR, FC PG RC LR LR LR LR LR LR LR LR, RC LR LR LR PC. .RC FC LR LR LR LR, PC 80 Myosotis latifolia Oenanthe sermentosa Oxalis oregana Petasites palmatus Polypodium californicum Polypodium gycyrrhiza Polysticum munitum Pseudotsuga menziesii Pteridium aquilinum Pyrola picta Ranunculus repens Ribes sanguineum Rosa californica Rosa gymnocarpa Rubus discolor Rubus parviflora Rubus ursinus Rubus spectabillis Rumex crispus Rumex ovtusifolius Salix sp. Sambucus callicarpa Sambucus racemosa Scrophularia californica Sequoia sempervirens Smilacina racemosa Stacys sp. Tellima grandiflora Toxicodendron diversilobum Trillium sp. Tsuga heterophylla Umbellularia californica Urtica dioica Vaccinium ovatum Vancouvaria hexandra Vancouvaria plantipetala Veratrum fimbriatum Veronica americana Viola glabella Viols sempervirens Whipplea modesta Woodwardia fimbriata MADRONO APPENDIX 1. CONTINUED No-revetment LR, BC, RC FC, RC Revetment-intact [Vol. 56 Revetment-removed MADRONO, Vol. 56, No. 2, pp. 81-88, 2009 DUST DEPOSITION EFFECTS ON GROWTH AND PHYSIOLOGY OF THE ENDANGERED ASTRAGALUS JAEGERIANUS (FABACEAE) UPEKALA C. WIJAYRATNE!, SARA J. SCOLES-SCIULLA AND LESLEY A. DEFALCO U.S. Geological Survey, Western Ecological Research Center, 160 N. Stephanie St., Henderson, NV 89074 uwijayratne@usgs. gov ABSTRACT Human expansion into the Mojave Desert is a significant threat to rare desert plants. While immediate habitat loss is often the greatest concern, rare plants situated near areas where soil surfaces experience frequent disturbance may be indirectly impacted when fine particulate dust accumulates on leaf surfaces. Remaining populations of the federally listed Astragalus jaegerianus (Lane Mountain milkvetch) occur on land open to expanding military activities and on adjacent public land with increasing recreational use. This study was initiated to determine whether dust accumulation could decrease the vigor and fitness of A. jaegerianus through reduced growth. Beginning in early May 2004, plants located on Bureau of Land Management (BLM) land were dusted bimonthly at canopy-level dust concentrations ranging from 0 to 32 g/m’, and physiology and growth were monitored until late June when plants senesced. The maximum experimental dust level simulates dust concentrations of Mojave Desert perennials neighboring military activities at a nearby army training center. Average shoot growth declined with increasing dust accumulation, but seasonal net photosynthesis increased. Further investigation of plants grown in a greenhouse supported similar trends. This pattern of greater net photosynthesis with increasing dust accumulation may be explained by higher leaf temperatures of dusted individuals. Ambient dust deposition measured in traps near field plants (May 2004—July 2004) ranged from 0.04—-0.17 g/m?/ d, which was well below the lowest level of dust on experimental plants (3.95 g/m?’/d). With this low level of ambient deposition, we expect that A. jaegerianus plants in this population were not greatly affected by the dust they receive at the level of recreational use during the study. Key Words: Anthropogenic disturbance, Astragalus jaegerianus, endangered species, eolian dust deposition, plant physiology. Many factors threaten the survival of Mojave Desert plant species. One of the most prevalent and far-reaching is anthropogenic development, which not only impacts species directly through plant injury and habitat loss but generates many indirect threats as well (Lovich and Bainbridge 1999; Belnap and Warren 2002). Atmospheric dust generated by soil surface disturbance is an indirect threat particularly relevant for the federally endangered perennial Astragalus jaeger- ianus Munz (Lane Mountain milkvetch) because the remaining populations occur on lands subject to increased vehicular traffic. Individuals of A. jaegerianus are found in four geographically distinct populations in a limited area of the western Mojave Desert (U.S. Fish and Wildlife Service 2004). Two of the four popula- tions (Brinkman Wash-Montana Mine and Par- adise Wash) are located almost entirely within lands withdrawn for the planned expansion of the Department of the Army’s National Training Center (NTC) at Ft. Irwin, California. A third population (Goldstone) is on existing NTC lands ‘Current address: U.S. Geological Survey, Forest and Rangeland Ecosystem Science Center, 3200 S.W. Jefferson Way, Corvallis, OR 97331 and the fourth population (Coolgardie Mesa) is found primarily on Bureau of Land Management (BLM) lands just south of the NTC. The USS. Fish and Wildlife Service listed A jaegerianus as endangered due to its endemism and its sensitivity to military activities and other vehicular traffic (U.S. Fish and Wildlife Service 1992, 1998). A better understanding of the impact from multiple land uses will be helpful in evaluating prospects for long-term conservation and recovery of the species. Protective fencing may prevent direct damage to plants by military vehicles (U.S. Fish and Wildlife Service 1998) or recreation, but concern remains that habitat quality within these protect- ed areas will be adversely affected by airborne dust raised by vehicles (U.S. Fish and Wildlife Service 2004). Long term exposure to a degraded habitat could counteract the effectiveness of fenced areas as a primary means of protection. Studying the effects of dust deposition on the physiology, growth and reproduction of A. jaegerianus is necessary for assessing the feasibil- ity of protecting populations that occur near high activity vehicle routes. In dry regions like the American Southwest, eolian dust is a natural occurrence. Dust storms 82 MADRONO in the western part of the Mojave Desert are generated primarily by cyclonic activity during winter and spring (Brazel and Nickling 1987). Dry lakebeds (playas) occur within the region where A. jaegerianus is found, providing a natural source of fine particulate dust. Human activity on arid soils such as military transports, training maneuvers, off-highway vehicles (OHV), mining activities, and intensive grazing greatly increases fugitive dust by denuding the land of vegetation and exposing areas from which dust can be raised (Adams et al. 1982; Grantz et al. 1998; Padgett et al. 2007). Human-induced changes in playa hydrology may also increase nutrient fluxes and influence plant communities that are downwind of the disturbed area (Blank et al. 1999). Fugitive dust associated with human distur- bances may impact the health of desert plants and is of particular concern for rare species with limited ranges. Sharifi et al. (1997) demonstrated that visibly dusty Mojave Desert perennials proximate to military activities at the NTC exhibit a 21-58% reduction in photosynthesis and a decrease in total shoot length. If leaf surface dust similarly reduces photosynthesis in A. jaegerianus, decreased growth and reproduc- tion over time could negatively impact popula- tion viability of this rare species. The goal of this study was to determine Whether surface dust accumulation could de- crease the vigor of A. jaegerianus individuals. We hypothesized that photosynthesis would decrease with accumulation of dust on the leaves, with a resultant decrease in growth. In addition, we examined the site characteristics that influence the amount of airborne dust intercepted by the plant canopy. METHODS Field Studies In April 2004, we selected A. jaegerianus plants from the previously surveyed Coolgardie Mesa population, located on land administered by California BLM (Barstow Field Office). The plants were then sorted into four size classes based on the dimensions of their canopies. We randomly chose five plants from each size class and randomly assigned leaf-level target dust treatments: 0, 8, 16, 24 and 32 g/m’, yielding four replicate plants per concentration (total n = 20). The maximum dust treatment was similar to heavily dusted Atriplex canescens (Pursh) Nutt. (Chenopodiaceae) growing near trails created by military personnel vehicles at the NTC (Sharifi et al. 1997). The dust concentrations were chosen to reflect potential future conditions for the two A. jJaegerianus populations (Brinkman Wash-Mon- tana Mine and Paradise Wash) which are located in the planned expansion area of the NTC. [Vol. 56 Currently, the Coolgardie Mesa population experiences soil surface disturbances created by a nearby OHV recreation site. Off-highway vehicle use is evident throughout the population (personal observation) and may increase in intensity. The quantity of dust needed to achieve target dust treatments was based on the ellipsoid area of each A. jaegerianus canopy. Experimental dust was generated by passing soil collected from nearby roadbeds through a series of soil sieves. This procedure produced dust particles small enough to fall through the L-shaped trichomes and onto the surface of the leaf, which was verified by microscopic examination. We distrib- uted a pre-measured quantity of dust onto A. jJaegerianus individuals through a 45 um sieve to ensure uniformity of application on May 07, May 25, June 08, June 22, July 06, and July 20. We chose three shoots of approximately equal size on each plant for monitoring growth. Shoot length and leaf number were recorded bimonthly prior to and throughout the dusting period (April 28—July 20, 2004). Shoot growth was calculated as the difference in maximum shoot length and initial shoot length, averaged over the three shoots for each plant. From May 7 to June 22, photosynthesis and water potential were measured on shoots separate than those for growth. Prior to each dust application, one new fully-expanded leaf on each plant was randomly selected. Following the accumulating bimonthly dust applications, we measured leaf-level net photosynthesis (Pye) for each plant with an open compensating photo- synthesis system (Li-6400, Li-Cor, Inc., Lincoln, NE) and mid-day water potentials (\) using a Scholander-type pressure chamber (PMS Instru- ment Co., Corvallis, OR). To ensure that dust accumulation did not impair the infrared gas analyzer measurements, we cleaned the sample > cell using ethanol when sample automatic gain control (AGC) values reached 3000 mV. In an. effort to allow plant responses to acclimate, | photosynthesis and water potential was not. measured until the morning after an evening | dusting. Photosynthesis was measured between | 0900 and 1200 hr and mid-day water potential | (\y) was measured immediately between 1200 and | 1300 hr on the same leaves. Leaf net photosyn- . thesis was expressed on a leaf-area basis by | recording an image of each leaf with light- | sensitive diazo paper and measuring the area. using a leaf-area meter (Li-3000A, Li-Cor Inc.). | Actual leaf-level dust concentration was mea- | sured both before and after bimonthly dusting. | Prior to dust application, a randomly selected leaf: adjacent to the physiological leaf was rinsed with | 20 ml of de-ionized water to determine the pre-' application concentration of dust. Immediately | following dust application, another randomly 2009] selected adjacent leaf was rinsed as described above to measure the post-application concen- tration. The resulting residue was dried to a constant mass in a convection oven at 50°C and weighed. All leaves used for dust concentration measurements were marked to avoid using them for future dust-concentration measurements. Dust concentration was expressed on a leaf-area basis by recording an image of each leaf with light-sensitive diazo paper and measuring the area using a leaf-area meter (L1-3000A, Li-Cor Inc.). To determine the levels of ambient dust deposition, or accumulation rates, for plants in the Coolgardie population, we constructed dust traps based on a design by Reheis and Kuhl (1995). Traps were made of bundt cake pans with hardware cloth placed approximately five centi- meters below the top rim. Clear marbles were placed on top of the hardware cloth, one layer thick. The rough surface simulated by the marbles caused dust particles to fall out of the flow of air into the traps. The marbles also prevented subsequent passing winds from picking up and lifting away the trapped dust. Traps were installed 10-60 m and 50-170 m from one- and two-track roads, respectively. Fifteen traps were placed within the Coolgardie population, with four traps placed among the study plants and 11 associated with additional milk-vetch plants. The traps were positioned on fence posts at shrub height, approximately one meter above the soil surface. Traps were rinsed monthly using de- ionized water, and the residue dried and weighed. Dust concentration was expressed based on the inside area of the bundt cake pan. Greenhouse Studies We established A. jaegerianus individuals in the greenhouse in April 2004 using seeds collected from the Goldstone population in 2003. The seeds were scarified with sandpaper (Sharifi and Prigge personal communication) and sown in pots modified from PVC pipe measuring 30 cm in length (10 cm diam.) with wire mesh on the bottom to allow for drainage. Shoots were supported with trellises fashioned from plastic chicken wire and later transplanted to five-gallon pots to discourage constriction of the root system, which might influence photosynthetic _ responses. To clarify results obtained from the field experiment, we dusted plants in the greenhouse ' from December 16, 2004 to January 27, 2005. Five replicate plants were randomly assigned either a dusted treatment or non-dusted control (1 = 10). Two shoots of equivalent size on each plant were chosen to monitor growth. Dust was created and applied weekly using the field _ protocol. The target dust concentration level for WIJAYRATNE ET AL.: DUST EFFECTS ON 4. JAEGERIANUS 83 the dusted treatment was the same as the target maximum level applied in the field (32 g/m”). The quantity of dust applied was based on the area of the study plant canopy as in the field experiment. The treated plants were visibly dusty compared to the untreated control plants and retained dust on the shoots and leaves between applications. At the end of each month, we washed three leaves of each target plant with 20 ml of de-ionized water to determine average cumulative dust concentra- tion. Shoot length and leaf number were recorded weekly for the growth shoots. Leaf-level net photosynthesis was also measured weekly on one “new” leaf and one “old” leaf from each plant. New leaves were characterized as the most recent, fully formed leaf on a shoot. Old leaves were initially selected by tagging the fifth leaf from the growing tip on a randomly chosen shoot. Once this initial leaf showed signs of senescence, such as yellowing or desiccation, we switched to the next younger leaf on the same shoot. Analysis Maintaining a fixed leaf-level dust concentra- tion on plants in the field was challenging because of the difficulty of dusting whole canopies of disparate branching architecture, as well as varied wind direction and velocity between sampling periods (Fig. 1). Though we were unable to maintain target dust concentrations, we did apply a constant pressure of dust throughout the study period and achieved a dust concentration gradi- ent. To express this constant pressure, we integrat- ed the fluctuating dust concentration over the study period, using the trapezoidal rule generated in SigmaPlot (version 9.01, Richmond, CA) and 50 . id @ a > Pa is) = 404 an E Dm 30 | : ; i : a 5 20+ as c & ® e e Pa = % e 8 104 7° o a g eae Qa @ 0 ® : T : T T T 0 10 20 30 40 50 Dust concentration (g/m?) - May 7 FiG. 1. Leaf dust concentrations measured on May 25 before dust application did not reflect dust concentra- tions applied 2 wk earlier on May 7 in the field (Fj.17 = 0.26, P = 0.61, R* = 0.02), indicating low retention of leaf dust between applications. Line indicates the expected 1:1 ratio if all applied dust was retained between sampling periods. 84 MADRONO divided by the length of the study period to obtain the average daily dust concentration that each individual plant experienced over the duration of growth and physiology measure- ments (g/m*). Photosynthesis and mid-day water potential were similarly expressed as an integrat- ed average using SigmaPlot. Physiology data from the fourth sampling period were omitted because plants began senescing by this time; therefore average daily dust concentration for these measurements were adjusted accordingly. All data were analyzed using SAS statistical software (version 9.1.3, Cary, NC). Patterns in plant physiology and growth associated with dust application in the field were analyzed using simple linear regression. Shoot growth and leaf number data from the field were heteroscedastic, and leaf number data included several zeros. Therefore, shoot growth was log-transformed before analysis, and leaf number was analyzed using Poisson regression with overdispersion included in the model (GLM procedure). Growth responses to dust treatments applied to green- house plants were analyzed with a nonparametric Wilcoxon rank sum test. Photosynthetic respons- es were analyzed using a split-plot ANOVA, with dust treatment as the whole plot and leaf age as the split-plot. A multiple linear regression was used to determine whether dust trap characteris- tics (elevation and distance from single- and dual- track vehicle routes), mean maximum daily air temperature and rainfall were associated with monthly deposition of dust in traps (Fernandez 2003). RESULTS Field Studies Phenological measurements indicated a steady increase in shoot length across all plants until most plants reached peak growth in mid June. As dust concentration increased, shoot growth de- creased (for log-transformed data, P = 0.01, F; 18 = 7.67, R* = 0.37) and there was a trend towards lower leaf production (P = 0.06, y°; 7s = 3.43; Fig. 2). Net photosynthesis increased as the concentration of dust on leaves increased (P = 0.02, Fi 13 = 6.47, R* = 0.26), but no difference in water stress across the dust concentration gradi- ent was detected (P = 0.95, F; 17 = 0.00, R? = 0.00; Fig. 3).We found a positive correlation between dust level and leaf temperature (Pearson correlation coefficient r = 0.49, P = 0.03) and between leaf temperature and leaf-level photo- synthesis rate (Pearson correlation coefficient r = 0.63, P < 0.01; Fig. 4). Ambient dust deposition rates in the traps for May 2004—July 2004 ranged from 0.04—0.17 g/m7?/ day. The study area is intersected by one unimproved single-track route and several dual- [Vol. 56 Log transformed shoot growth (mm) No. leaves produced/shoot 0 5 10 15 20 25 30 35 Average daily dust concentration (g/m?) FIG. 2. Seasonal shoot growth (a) (i.e., change in shoot length) declined with increasing leaf dust concentration (Log[shoot growth] = 2.62 — 0.20*[dust concentration]). Leaf production (b) declined slightly as leaf dust concentration increased. track vehicular routes. Rainfall, mean maximum daily temperature and distance to the single-track route were significant factors explaining daily dust | concentration (overall model, P < 0.01, Fo, 116 = 39.96, R* = 0.41). High dust concentrations were | associated with low monthly rainfall, high mean © maximum daily temperature (Fig. 5), and prox- imity to the single track road. Wind direction and | speed likely varied throughout the area as well but | were not recorded during this study. Greenhouse Studies Cumulative leaf-level dust concentration on. treated plants at the end of the experiment ranged _ from 20-40 g/m’. Neither shoot growth (P = 0.30, z = 0.52) nor leaf production (P = 0.46, z = | 0.11) differed between dusted plants and control — plants (Fig. 6). Dusted plants appear to have : higher seasonal photosynthesis than controls but | this difference was not statistically significant (P | = (0.33, F,.3 = 1.06; Fig. 7). As we expected, new | leaves had significantly higher rates of photosyn- | thesis than old leaves (P = 0.04, F,.3: = 5.763) Fig. 7). While greenhouse results were largely. i) =) S & fe2) o a o (umols CO2 m-2 s-1) net Average daily P (MPa) mid Average daily y 0 2 4 6 8 10 12 14 16 18 20 22 Average daily dust concentration (g/m’) Fic. 3. Average leaf-level net photosynthesis (a) increased with increasing leaf dust concentration. Average midday water potential (b) did not change significantly with increasing leaf dust concentration. non-significant, the trends are consistent with field results. DISCUSSION Studies of dust deposition on desert plants suggest that photosynthesis and growth are negatively impacted by the dust that falls on the surfaces of leaves (Beatley 1965; Sharifi et al. 1997, 1999). Though not all of the dust experi- mentally applied to A. jaegerianus plants was retained on the leaf surfaces in our field study, the net accumulation of dust after repeated applica- tions reduced shoot growth. Astragalus jaeger- ianus individuals are rarely observed growing outside of the canopy of desert shrubs (Prigge et al. 2000), which they use as a trellis for support (Gibson et al. 1998). At the same time, A. _jaegerianus requires at least two-thirds full sunlight to reach maximum net photosynthetic rates (Gibson et al. 1998). Consequently, we _ expect that A. jaegerianus individuals experienc- ing prolonged exposure to dust next to heavily _used vehicle routes may not grow to the height needed to acquire adequate sunlight and achieve -maximum net photosynthetic rates. Flower and fruit production during the study was poor irrespective of dusting treatment and were not WIJAYRATNE ET AL.: DUST EFFECTS ON 4. JAEGERIANUS 85 26.0 a) 25.5 25.0 23.5 @e Average daily leaf temperature (C°) oo pee! [ae 0 2 4 6 8 10 12 14 16 18 20 22 Average daily dust concentration (g/m?) (umols CO2 m-2 s-1) net Average daily P 3° 22.5 23.0 23.5 24.0 24.5 25.0 25.5 26.0 Average daily leaf temperature (C°) FIG. 4. Relationships between (a) dust concentration and leaf temperature (leaf temperature = 23.84 + 0.09* [dust concentration]) and (b) leaf temperature and leaf- level photosynthetic rate (photosynthesis = —149.78 + 7.37 *[leaf temperature]). analyzed (1.e., only 28% of the shoots measured had flowers or fruits), thus it remains to be seen whether reproductive effort over the life of the plant is compromised by diminished growth. Dusted A. jaegerianus individuals in the field had reduced shoot growth compared to undusted plants. Contrary to our original hypothesis, however, average net photosynthesis increased with leaf dust concentration and is counter- intuitive because of the decrease in shoot growth. We considered that dust accumulation could stimulate compensatory leaf production or shift internal plant resources within the photosynthetic mechanism to enhance photosynthesis. Conse- quently, we expected greater leaf production and/ or higher photosynthetic rates in new leaves on dusted plants than those on undusted plants. We measured photosynthesis on both new and old leaves on dusted and control plants in the greenhouse. Similar leaf production and photo- synthetic rates between dusted and undusted leaves do not support this explanation. Alterna- tively, higher dust loads on leaves are known to increase absorption of incident radiation, raising leaf temperature and shifting the temperature- 86 MADRONO (“~~ Dust deposition —@— Precipitation —tt— Max daily temperature 0.14 0.12 0.10 0.08 0.06 0.04 Dust deposition (g m2 d-1) 0.02 0.00 Fic. 5. Sept Oct [Vol. 56 70 50 60 40 = mca s/ ® re) — — < 40 ©. 30 a @ | 3 {e) @O 305 20 3 =| = 203 o 10 © 10 = 0 0 Nov Dec Jan April hem © = — 2 o LL Mean ambient dust deposition (+standard deviation) collected in 15 traps over the study area was highest during the late spring months when 4A. jaegerianus plants were active. No precipitation fell during May, June and July. Dust concentrations were averaged over February and March because trap contents were not emptied at the end of February. response curve (Hirano et al. 1995). Leaves of plants in the field that had higher dust concen- trations also had higher leaf temperatures. Dust- induced increases in leaf temperatures and subse- quent photosynthetic rates during early spring would extend the activity period that milkvetch could maintain positive net photosynthetic rates, but as spring temperatures increased, leaf tem- peratures of dusted plants likely lowered net photosynthetic rates, thus reducing shoot growth. Dusted plants in the greenhouse tended to have lower shoot growth and greater net photosynthe- sis, though the results were not statistically significant. While it is possible that we were 25 7 a) b) 20 5 15 4 Average twig growth/shoot (cm) jyooys/paonpoid saaea] # BbeuaAy 10 - + 3 + 2 54 + 4 0 T T 0 Control Dust Control Dust FIG. 6. Average shoot growth (a) and leaf production (b) for greenhouse grown plants (n = 5 for each treatment; +standard deviation). Neither shoot growth nor leaf production were significantly different between dusted and control plants. unable to detect a significant treatment difference due to the low number of plants available for the experiment, enhanced water status of greenhouse plants may also have offset the negative effects of dust. Sharifi et al. (1999) showed that naturally dusted Larrea tridentata (Sess¢ & Moc. ex DC.) Coville individuals that received irrigation had higher shoot water potentials and more growth than non-irrigated individuals. The greenhouse A. jaegerianus plants were watered regularly, which may have been sufficient to counteract | any effect of dust accumulation. Ambient dust deposition for the Coolgardie | population during the study period was on the 35 Ga New ES Old 30 + ee ie hy = N 20+ e) e) 2 15 4 ° = = 10 Fs a 5 ea 0 Control Dust FIG. 7. Net photosynthesis was not different between | dusted and control plants in the greenhouse. However, | new leaves did exhibit significantly higher rates of net photosynthesis than old leaves. 2009] upper end of yearly dust accumulation rates reported for the Mojave (Reheis 2006) over a 16- year period. The dust traps used in this study create local turbulence causing any dust remain- ing on the top of the trap to be carried away, and thus may be underestimate the true rate of accumulation on vegetation (Reheis and Kuhl 1995). Integrating trap dust fluxes over May and June, which roughly corresponds to the field dusting experiment, enabled us to compare the result to control (non-dusted) study plants. The cumulative daily dust concentration in the traps ranged from 1.14—2.82 g/m’, while cumulative dust concentration on control plants was 4.15— 9.10 g/m°. With this low level of ambient cumulative deposition, we expect that A. jaeger- ianus plants in this population were not greatly affected by the dust they received from unim- proved vehicle routes by the end of the study. In addition, all of our study plants recovered from experimental dusting after heavy winter rains and put out new growth for the 2005 season. Future protection of this population will depend on minimizing the proliferation of single-track routes, especially during periods of low rainfall when dust evolution is greatest. Experimental dust accumulation on study plants simulated potential scenarios for the Brinkman Wash-Montana Mine and Paradise Wash populations, located in the expansion area of the NTC, where fugitive dust 1s expected to increase due to military training. Proposed conservation areas (U.S. Fish and Wildlife 2005) will be more beneficial if there is an adequate buffer zone between the dust source and the plants. Buffer zones have been an effective means of preventing other airborne contaminants such as pesticides that drift onto non-target plants or areas (de Snoo 1999; Robinson et al. 2000; Burn 2003) and may help prevent excess dust from drifting onto A. Jaegerianus populations. Even though direct loss of habitat is the greatest threat for A. jaegerianus, dust generated by recreation and expanded military training near A. jaegerianus habitat is a concern for the continued persistence of plants, whether protect- ed in conservation areas or in areas fenced to prevent direct losses of individuals. Low ambient dust levels for the Coolgardie population did not pose a hazard during this study; however, monitoring of dust generation should continue because increased recreational use of the area is a concern. Dust impacts within the NTC expansion area can potentially be reduced by providing sufficient distance, or a buffer, between the source of dust and A. jaegerianus populations. Designation of buffer zones should also take into consideration the dust concentration threshold at which A. Jaegerianus reproduction is reduced to below the WIJAYRATNE ET AL.: DUST EFFECTS ON 4. JAEGERIANUS 87 rate required to maintain minimum viable populations. Monitoring of dust deposition, as well as measurements of A. jaegerianus physiol- ogy, growth and reproduction should continue after the expansion occurs to assess efficacy of the buffer zones in maintaining A. jaegerianus populations. ACKNOWLEDGMENTS We acknowledge the support of Connie Rutherford and Ray Bransfield at the USFWS (Ventura, CA office) during all phases of this study. In addition, we appreciate the help of Charles Sullivan at CA-BLM in Barstow, CA for obtaining approval to work on BLM land. We thank Marith Reheis for advice on dust trap design. Assistance from Student Conservation Associ- ation Resource Interns Anna Braswell, Crystal Ed- wards, Lisa Jones and Stacy Smith is appreciated as well as help from our USGS coworkers Adam Nilsson, Karissa Smith, Dustin Haines, Todd Esque, Ken Nussear, J.R. Machett and Bridget Lair. We thank Julie Yee for her review of the statistics and suggestions that improved the final manuscript. Funding for this project was generously provided by USFWS, under FWS permit ##TE-022630-1. Any use of trade, product, or firm names in this publication is for descriptive purposes only and does not imply endorsement by the U.S. government. LITERATURE CITED ADAMS, J. A., A. S. ENDo, L. H. STOLZy, P. G. ROWLANDS, AND H. B. JOHNSON. 1982. Controlled experiments on soil compaction produced by off- road vehicles in Mojave Desert, California. Journal of Applied Ecology 19:167—176. BEATLEY, J. C. 1965. Effects of radioactive and non- radioactive dust upon Larrea divaricata Cav., Nevada Test Site. Health Physics 11:1621—1625. BELNAP, J. AND S. WARREN. 2002. Patton’s tracks in the Mojave Desert, USA: an ecological legacy. Arid Land Research and Management 16:245—258. BLANK, R. R., J. A. YOUNG, AND F. L. ALLEN. 1999. Aeolian dust in a saline playa environment, Nevada, U.S.A. Journal of Arid Environments 41:365-381. BRAZEL, A. J. AND W. G. NICKLING. 1987. Dust storms and their relation to moisture in the Sonoran-Mojave Desert region of the South- western United States. Journal of Environmental Management 24:279-291. BURN, A. 2003. Pesticide buffer zones for the protection of wildlife. Pest Management Science 59:583—590. DE SNOO, G. R. 1999. Unsprayed field margins: effects on environment, biodiversity and agricultural practice. Landscape and Urban Planning 46:151—160. FERNANDEZ, G. 2003. Data mining using SAS appli- cations. Chapman and Hall/CRC, Boca Raton, Fle, GIBSON, A. C., M. R. SHARIFI, AND P. W. RUNDEL. 1998. Ecophysiological observations of Lane Mountain Milkvetch, Astragalus jaegerianus (Fa- baceae), a proposed endangered species of the Mojave Desert. Aliso 17:77-82. GRANTZ, D. A., D. L. VAUGHN, R. FARBER, B. KIM, M. ZELDIN, T. VAN CUREN, AND R. CAMPBELL. 88 MADRONO 1998. Seeding native plants to restore desert farmland and mitigate fugitive dust and PM). Journal of Environmental Quality 27:1209-1218. HIRANO, T., M. KIYOTA, AND I. AIGA. 1995. Physical effects of dust on leaf physiology of cucumber and kidney bean plants. Environmental Pollution 89:255-261. LovIcH, J. 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. PADGETT, P. E., W. M. DOBROWOLSKI, M. J. ARBAUGH, AND S. A. ELIASON. 2007. Patterns of carbonate dust deposition: implications for four federally endan- gered plant species. Madrono 54:275—285. PRIGGE, B. A., M. R. SHARIFI, AND D. MORAFKA. 2000. Lane Mountain Milk-vetch Surveys. Progress Report III. Prepared for the National Training Center, US Department of the Army, Contract # DSCA05-99-P-0759. Dominguez Hills Foundation, Carson, California. REHEIS, M. C. 2006. A 16-year record of eolian dust in Southern Nevada and California, USA: Controls on dust generation and accumulation. Journal of Arid Environments 67:487—520. AND R. KIHL. 1995. Dust deposition in southern Nevada and California, 1984-1989: rela- tions to climate, source area, and source lithology. Journal of Geophysical Research 100:8893-8918. ROBINSON, R. C., R. G. PARSONS, G. BARBE, P. T. PATEL, AND S. MURPHY. 2000. Drift control and [Vol. 56 buffer zones for helicopter spraying of bracken (Pteridium aquilinum). Agriculture Ecosystems & Environment 79:215—231. SHARIFI, M. R., A. C. GIBSON, AND P. W. RUNDEL. 1997. Surface dust impacts on gas exchange in Mojave Desert shrubs. Journal of Applied Ecology 34:837-846. : , AND . 1999. Phenological and physiological responses of heavily dusted creosote bush (Larrea tridentata) to summer irrigation in the Mojave Desert. Flora 194:369-378. UNITED STATES (U.S.) FISH AND WILDLIFE SERVICE. 1992. Endangered and threatened wildlife and plants; proposed rule for seven desert milk-vetch taxa from California and Nevada. 57 Federal Register 90 (8 May 1992):19844-19851. . 1998. Endangered and threatened wildlife and plants: determination of endangered or threatened status for five desert milk-vetch taxa from Califor- nia. 63 Federal Register 193 (6 October 1998): 53596-53615. . 2004. Endangered and threatened wildlife and plants; proposed designation of critical habitat for Astragalus jaegerianus (Lane Mountain milk- vetch). 69 Federal Register 66 (6 April 2004): 18018-18035. . 2005. Endangered and threatened wildlife and plants; designation of critical habitat for Astragalus jaegerianus (Lane Mountain milk-vetch); Final Rule. 70 Federal Register 67 (8 April 2005): 18220-18241. MADRONO, Vol. 56, No. 2, pp. 89-98, 2009 EFFECTS OF FIRE AND GROUNDWATER EXTRACTION ON ALKALI MEADOW HABITAT IN OWENS VALLEY, CALIFORNIA DANIEL W. PRITCHETT 401 East Yaney St., Bishop, CA 93514 skypilots@telis.org SARA J. MANNING! Inyo County Water Department, P. O. Box 337, Independence, CA 93526 smanning@telis.org ABSTRACT Alkali meadow habitat—a groundwater dependent ecosystem—is rare in California, and its response to fire has not been documented. We sampled vegetation in this habitat across a pumping- induced depth-to-water (DTW) gradient immediately before, then eight weeks after, a summer wildfire. After the fire in the burned area, we documented vigorous re-growth of native perennial grasses in areas of shallow groundwater but no re-growth of shrubs. The dominant grass Sporobolus airoides flowered in the shallow end of the DTW gradient but never advanced beyond vegetative phenology (leaves) in the drawn-down end. DTW explained 77% of the grass cover variance in the post-fire burned area, and 87% and 94% of the grass cover variance in the pre-fire and post-fire unburned (control) areas, respectively. This suggests that post-fire re-growth was re-establishing a cover-DTW relationship already present before the fire. The principal short-term fire effect was the elimination of shrub cover due to apparent shrub mortality. Our study shows fire may be an effective management tool for regenerating alkali meadow in areas of shallow groundwater. However, in areas subject to long-term water table drawdowns we found negligible grass re-growth and increased vulnerability to erosion, suggesting fire may accelerate the process of type-conversion from meadow to xeric shrubland. Key Words: Distichlis spicata, fire, groundwater-dependent, groundwater pumping, habitat loss, Sporobolus airoides, vegetation change, water table. On July 6, 2007, lightning strikes along the base of the eastern escarpment of the Sierra Nevada in Owens Valley, California, started several fires. They became known as the Inyo Complex fire as they coalesced and burned >14,000 ha, almost destroying the towns of Independence and Big Pine. The fires started in sagebrush/bitterbrush shrubland high on alluvial fans and were blown by strong winds downslope through Mojave Desert shrubland to burn alkali meadow habitat at the fan toes. Re-growth after the fire varied considerably in the alkali meadow. In some areas we observed native perennial Cy grasses, Sporobolus airoides (alkali sacaton) and Distichlis spicata (saltgrass, nomenclature follows Baldwin et al. 2002), re- sprouting within days after the fire. In other areas we saw no re-growth. After several weeks, grass re-growth was so robust in some areas it was difficult to tell that a fire had occurred. Alkali meadow habitat is rare in California (Sawyer and Keeler-Wolf 1995), and there are few quantitative data regarding its response to fire. Anecdotal observations made after other recent Owens Valley fires suggest S. airoides and D. 'Present address: Big Pine Paiute Tribe Environmen- tal Department, P.O. Box 700, Big Pine, CA 93513. spicata recover rapidly in areas of shallow groundwater, but we had not had the opportunity to closely monitor recovery after fires in areas subject to groundwater drawdowns. Understand- ing the response of alkali meadow to fire in areas of lowered water table is important because drawn-down water tables underlie much of this habitat in Owens Valley. The Inyo Complex fire burned a portion of alkali meadow habitat which covered a large gradient in depth-to-water table (DTW) at the time of the fire. It thus offered an unplanned opportunity to observe alkali meadow response to fire in areas of both shallow and deeply drawn-down groundwater. Our primary objectives were to: 1) document species composition, cover, frequency, and phe- nology of the observed post-fire growth through- out the burned portion of the study area; 2) compare characteristics of pre- and post-fire vegetation; and 3) determine the extent to which DTW could account for vegetation patterns in the study area. METHODS Study Site Owens Valley lies on the western edge of the Great Basin in eastern California at an elevation 90 MADRONO of approximately 1200 m above mean sea level. The Sierra Nevada borders the valley on the west, and the White and Inyo Mountains flank its eastern side. Owens Valley summers are hot and dry; winters are cold, and more than _three- quarters of average annual precipitation falls during the October to March cold period. Although the climate is arid, with average annual precipitation of 130 mm, abundant water from Sierran snowmelt flows into Owens Valley and annually recharges shallow groundwater aquifers which underlie thousands of hectares of the valley floor (Hollett et al. 1991). As a result, valley floor vegetation is dominated by phreatophytic plant species assembled in alkali meadow and _ sink habitats (City of Los Angeles and County of Inyo 1991; Howald 2000). The growing season for alkali meadow species in Owens Valley com- mences in late March (first of spring), reaches a peak in late June when leaves and stems are fully grown, then enters winter dormancy in early October (Sorenson et al. 1991). Alkali meadow vegetation is characterized as groundwater dependent because its estimated actual evapotranspiration exceeds annual precip- itation (City of Los Angeles and County of Inyo 1991). Under the classification of groundwater- dependent ecosystems by Eamus et al. (2006), alkali meadow habitat falls under the “‘phreato- phytic” category, because groundwater is not regularly expressed on the ground surface. In Owens Valley, large expanses of alkali meadow occur at the toes of Sierran alluvial fans located several miles from open water or Owens River, and the study site is in such a setting. Most alkali meadow habitat in Owens Valley is on land owned by the City of Los Angeles and is managed for grazing and water export to Los Angeles by the Los Angeles Department of Water and Power. In the mid 1980's, “‘parcels”’ of relatively homogeneous vegetation in Los Angeles’ holdings were delimited, mapped, and inventoried for species composition and cover. Groundwater levels and vegetation cover and composition on selected parcels have been regularly monitored since 1991. The study site (Fig. 1), located approximately 12 km north of the town of Independence, consists of two adjoining parcels classified in the 1980’s as alkali meadow, a classification consistent with Lee’s (1912) ‘“‘grassland”’ classification made over 70 yr before. Both study site parcels were dominated before the fire by two native phreatophytic grasses S. airoides and D. spicata, and, in places, by the encroaching shrubs Artemisia tridentata, Atriplex lentiformis subsp. torreyi, and/or Chrysothamnus nauseosus. Histor- ically, livestock grazing had occurred throughout the study area (typically from October through May), including during spring 2007. After the July 2007 fire, livestock were excluded from the [Vol. 56 study area for the remainder of the 2007 growing season. Substrate of the study site is a combination of ancient beach, bar, or river channel sediments. These deposits lie between alluvial fan deposits upslope to the west and fluvio/lacustrine deposits of the valley floor downslope to the east (Hollett et al. 1991). Soils are moderately- to poorly- drained loams (USDA Natural Resources Con- servation Service 2002). Vegetation Sampling Inyo County Water Department (ICWD) monitors vegetation composition and cover each summer along randomly-located 50-m transects in selected parcels. ICWD uses the point-inter- cept technique (Bonham 1989) to determine species composition of the top canopy layer at 0.5-m intervals along the transect. The endpoints of these transects are not permanently marked, but the starting location 1s recorded with a GPS unit, and its randomly-generated compass bear- ing 1s recorded. Two weeks before the Inyo Complex fire, ICWD read point-intercept tran- sects at 48 random locations throughout the two parcels comprising the study area (Fig. 1). We conducted visual reconnaissance of the burned area several times, starting one week after the fire. Eight weeks after the fire, August 30— September 6, 2007, we used GPS and compass bearings to approximately relocate then re- sample the 48 pre-fire random transect locations in the study area. Twenty were in burned areas and 24 in un-burnt areas. Four of the 48 transects were not sampled because they crossed the burn boundary. To increase the size and spatial extent of post-fire sampling, we randomly selected 22 additional transects throughout the study site. Altogether, we obtained data from 37 transects in the burned area and 29 from the unburned “control” area after the fire (Fig. 1). We em- ployed the same point-intercept method used before the fire to measure species composition and estimate top layer canopy cover. We also noted the most advanced phenology of each species sampled along each transect in the burned area. Phenology was not recorded along pre-fire transects nor along transects in the unburned control area. Precipitation and Water Table Data ICWD maintains a precipitation gauge in the study area, and its precipitation totaled 31 mm for the period October 1, 2006—April 30, 2007. No precipitation was recorded from April 17, 2007, until August 28-30, when 3 mm _ were recorded. Average precipitation for the ICWD gauge period of record (1993-2006) was 160 mm. 2009] 389000 4osso00f 4084000 4: PRITCHETT AND MANNING: EFFECTS OF FIRE ON ALKALI MEADOWS a1 390000 1 Kilometers 389000 Fic. 1. 390000 Study area and sample sites. The 204-ha study area is outlined in black. The burned portion 1s indicated by stippling and covers 76 ha. Open circles represent start points of transects read before the fire; solid triangles represent start points of transects read after the fire in the burned area; open squares represent start points of transects in the unburned (control) area. Map projection: UTM Zone 11, NAD 27. To show the historic DTW gradient in the study area we digitized the 8 ft (2.4 m) and 4 ft (1.2 m) contours from Lee’s (1912) map and included them in Fig. 2. Lee (1912) estimated DTW in southern Owens Valley during the period 1908-1911 based on readings of 169 observation wells, seven of which were in or adjacent to the study area. Lee gathered his data before large scale groundwater pumping had been initiated in Owens Valley. To show the 2007 DTW gradient, we ob- tained DTW data collected on or near April 1, 2007, from 14 observation wells in or immedi- ately adjacent to the study area (Fig. 2). We chose April 1 DTW data because early spring groundwater levels typically represent the an- nual high stand before decline due to the onset of seasonal evapotranspiration (Lee 1912), and because they are used by ICWD and other researchers for tracking annual changes in groundwater depth as well as for investigations of DTW-cover relationships (Elmore et al. 2006). Next, we applied ordinary kriging, as implemented in ArcGIS 9.2, Spatial Analyst extension software (circular variogram, 10-m grid cells, and default values for other param- eters), to the April 2007 DTW data to create a DTW grid covering the study area. We chose kriging as opposed to other interpolation techniques because it is recommended in cases where sample sites are irregularly spaced (Le- gendre and Legendre 1998). We derived DTW 92 MADRONO 389000 40€3006 389000 FIG. 2. [Vol. 56 390000 390000 Historic and 2007 water table gradient. Solid black line outlines study area. Dotted lines represent April 2007 water table contours (labeled in meters from ground surface). Bold-faced dashed lines represent the —8 ft (—2.4 m) and —4 ft (—1.2 m) water table contours mapped by Lee (1912). Crossed circles represent monitoring wells used for kriging, and adjacent numbers represent April 2007 DTW values at the wells. Scale and map projection as in Fig. 1. contours from the grid and included them in Fig. 2. The current DTW gradient in the study area— with drawdowns of up to 5 m and alterations of subsurface flow relative to conditions mapped by Lee (1912)—1s controlled by large scale ground- water extraction which began in 1970 (Fig. 3) (City of Los Angeles and County of Inyo 1991; Steinwand and Harrington 2003a, b; City of Los Angeles Department of Water and Power and Inyo County 2007). To estimate DTW under each vegetation transect, we overlaid the 2007 transects on the kriged 2007 DTW grid and calculated the DTW as the mean of all grid cells intersected by the transect. Analysis We grouped transect data into three sets for analysis: 1) pre-fire data covering the entire study area; 2) post-fire data from the burned portion of the study area; and 3) post-fire data from the unburned control portion of the study area. To characterize vegetation we calculated each spe- cies’ mean cover and frequency based on the transects sampled in each dataset. We used a bootstrap technique to estimate standard errors because of the skewed distributions of most species (Elzinga et al. 1998). We calculated the Mann-Whitney U statistic to determine if the mean cover of S. airoides, the dominant species in the post-fire burned area, differed from its cover 2009] 12000 10000 8000 6000 4000 2000 Total Pumping (hectare-meters) PRITCHETT AND MANNING: EFFECTS OF FIRE ON ALKALI MEADOWS 93 Water Year (to yr shown) FIG. 3. water-year cumulative totals, 1931—2005. in the pre-fire and post-fire control datasets. We calculated the chi square statistic to find out if frequencies of the three most common species in the post-fire burned area differed significantly from their frequencies in the pre-fire and post-fire control datasets. We calculated the Mann-Whit- ney U statistic to determine if there was a significant difference in mean DTW between post-fire burned transects with flowering S. airoides versus those with vegetative-only indi- viduals. To allow visual comparison of cover and frequency of each species among the three datasets we juxtaposed data in a table. We used ordinary least-squares regression (as implemented in Sigma Plot 10.0 software) to attempt to explain variance in grass cover (summed cover of S. airoides and D. spicata at each transect) as a function of estimated April 2007 DTW at each transect in each dataset. Each transect’s grass cover and DTW represented a single data point in these regression analyses. We performed exploratory linear regressions on raw data, then transformed grass cover data by adding one to each value and taking the natural log of the sum. We then regressed the trans- formed data against DTW using a sigmoid curve. We chose the sigmoid curve rather than the straight line for both practical and theoretical reasons: it explained more variance, and species abundance can be better related to environmental gradients by unimodal curves (ter Braak 1996), of which sigmoid curves are special cases (ter Braak and Looman 1995). RESULTS Fire Effects At our first site reconnaissance within days after the fire, virtually all above-ground biomass had been reduced to ash or charred wood. By eight weeks after the fire, seven species showed measur- able re-growth in the burned area (Table 1). Total Groundwater pumping from wellfield in which study site is located. Amounts (in ha-m) are shown as five transect cover ranged from zero to 38%. Sporobolius airoides and D. spicata were the dominant graminoid species, dominant overall, and showed great spatial variation in cover. Glycyrrhiza lepidota was the third most abundant herb. Sporobolius airoides and Heliotropium curassavicum sampled in burned-area post-fire transects were in both vegetative and flowering states, while we sampled D. spicata only in vegetative state. Average cover of S. airoides, the dominant species in the burned area after the fire, was significantly lower than S. airoides measured in pre-fire and post-fire control datasets (P < 0.05) (Table 1). However, post-fire frequencies of the three most common species in the burned area did not differ significantly from frequencies in the pre-fire and post-fire un-burned control area datasets (Table 1). Although most graminoid species recorded before the fire were recorded after the fire, few other herbaceous and woody species were sampled on post-fire transects in the burned area. A single hit on Salix exigua was the only documentation of post-fire shrub re-growth (Table 1). DTW Effects Grass cover was significantly correlated with DTW in all three datatsets (P < 0.01). DTW explained 87% of the variance in pre-fire grass cover, 77% of the variance in the post-fire burned area, and 94% of the variance in the post-fire control area (Fig. 4). The mean DTW for transects with flowering S. airoides was 2.6 m (SE = 0.20) and was shallower than the mean for transects in which S. airoides was in vegetative state only, 3.6 m (SE = 0.32), and the difference was significant (P < 0.05). Heliotropium curassavicum was also observed flowering along transects after the fire, but the sample size was too low for statistical analysis. ~ [Vol. 56 LI 8 Ss fe) Z e) a ja : 9 6 ve 6v “AOD “XP t00 I¢0 vc 0 990 69°0 ‘body “SUBLL a Ss HG. ca | b 61 Oc ‘sueI] “UINN SJOISUBI] 67 LOOT IsNSNY “‘pouInqun dI1J-1isog MIVA IIAOD ITf-1SOd soVUSISOp (,.) YSUOSY “JOosUvI] IJSUIS B UO SaIddds UDAIS B JO IOAOD JSOYSIY JY} SI boi ‘suery,, ‘Wy sem saioods UdAIS B YSTYM UO s}ddsUBI] JO JAQUINU Sd} vOIpUI :d1B SWIOJ YIMOIDH “(SUONLIO] SuT[dures 1OJ | “SI 99S) “SAFTAUNS NOILVLAOAA AUIA-LSOd ANV -dud NI AONANOAAY ANV ‘AFAOD ‘NOLLISOUNOD Salads Aq payenoyes sem 94 “AOD “XP t0'0 c0'0 80°0 vc 0 £00 £90 99°0 "Dody “SUBLL 6 I VC SC ‘sued “UINN spoosuely L¢ LO> OS Ce x19 “AOD Dd LOOT SNSNY ‘pound d11J-1s0Oq ve | LC Sv “AOD “XP vr'0 cS 0 OO COO cO'0 cO'0 vO'O vO'O vO'O vO'0 6c 0 vO'O Se 0 SLO LLO ‘boy “SUBIL New ae coo _ NN — OMmMNTANNANNA Gere RNN bc | ny Le ‘sued “UN NY SIOISUP. Bp 8 0 80 i) LOOT FUNG ‘dTIS dIIQUD dITJ-I1g > CN Se ae ee ee qcoocoeoaqooaaq ee ea numboiut ppapns SNIDINIMMAIA SNJDGOIADS DNBIXI XIJVDS snluapAjod snuiupylO10sg SUIISOUDI Xa]d1AL py ‘ds pluauouvnydals pIDJUap1A] ‘dsqns pypJuapll] VISNUIJAP 1A AAO] ‘dsqns siusofijua, Xajdialp snsoasndU SnNULUDYIOSAAYD DSOUADI DAIYJUDLIDYIDI DSOWDIDA DULOIOAAA SLAD]JIXD DAT SLUDINAISD{ SPIdajISp syplUuapigI0 YpIYdoAjIN’ psoida] DIJIAJDIV WNIIADSSPANI wUnidoAOl aH pyofidossdAy vissvg plopida] DZIYAIAIAT) ‘ds xaupy snoujpg snoune pIpII1ds syYysig sapio.v snjogo1odsy so1sods jueld NNnNNANAMN S-qns OoOonmMmoOnseeaeo os WI0-] (SOO > d) 1891 Q AouTTYAA-UUPYY Aq pouTULIajap se sanyRA [O.WUOD a1J-jsod pue daitj-did UvY} IOMOT ATJUBOTJTUSIS ce 6 AOS “XPI[,, °, SUBI] “WINN, Aq s}oosURI] [210] OY] SUIPIAIp ‘sueI] “WINNY,, “(S) qniys pue ‘(Y) Snoadeqsay JayIO “(S) PlourlureIs ‘Tl ATIVE y = 3.78/(1+exp(-(x- -4.59)/0.83)) R? =.87 O o 8% Boe Ln (pre-fire grass cover) -4 DTW (m) y= 2.79/(1+exp(-(x- -4.19)/0.49)) R2 =.77 Ln (post-fire grass cover) -8 -4 1] DTW (m) > Ty= 3.74/(1+exp(-(x- -4.19)/0.74)) ] S |R7=.94 sy oO 44 oO 2 ae ed ia a 0 5 a ° = | rs Qo 24 | 6 & / S s a core oh Cc | Pree ce . -8 -4 2 0 DTW (m) Fic. 4. Grass cover as a function of DTW in meters below ground surface. Dotted lines represent 95% confidence limits: (A) Pre-fire dataset, n = 48; (B) Post-fire burned area dataset, n = 37; (C) Post-fire un- burned control dataset, n = 29. All coefficients were significant at P < 0.01. DISCUSSION Fire Effects The fire’s timing in summer, the noted thoroughness of combustion of above-ground biomass, and the meager precipitation in 2007 PRITCHETT AND MANNING: EFFECTS OF FIRE ON ALKALI MEADOWS 1) (both before the fire and in the eight weeks after the fire), led us to expect little re-growth before the 2008 growing season (following winter precipitation). We sampled vegetation just eight weeks after the fire because, contrary to our expectations, we had observed vigorous re- growth in places, and we desired to document the species showing re-growth and investigate the spatial variation in that re-growth. We had no expectation that re-growth was complete and, therefore, expected mean cover over the burned area to be lower than mean cover of the pre-burn and post-burn control areas (Table 1). A note- worthy finding was that short-term re-growth of two grasses (and the herb G. /epidota) did attain pre-fire and post-fire control levels in terms of frequency (Table 1), suggesting the fire caused little mortality among the dominant herbaceous species. This finding is consistent with the results of Smith and Kadlec (1985), who found no substantial below-ground mortality of saltgrass after burning in a Utah marsh. The four graminoid and forb species which attained greater than trace cover values after the fire in the burned area (Table 1) show anatom- ical/physiological traits which facilitate survival after fire. Perennating buds occur below ground on rhizomes for D. spicata and at the root crown for the bunchgrass, S. airoides. Glycyrrhiza lepidota 1s rhizomatous while H. curassavicum has underground rootstocks able to give rise to new stems. For these reasons, in addition to the dry, hot, post-fire conditions which made seed germination and seedling survival extremely unlikely, we treated live cover sampled in the post-fire burned area as re-growth from plants already established before the fire, rather than new recruits. In contrast to the re-growth of herbaceous species, shrubs may have suffered high mortality. Artemisia tridentata is known to be killed by fire and re-establish by seed (Tirmenstein 1999a). The absence of C. nauseosus after the fire (Table 1) suggests mortality because it has been reported by several researchers to show “‘rapid”’ post-fire recovery through re-sprouting as well as by seed germination (Tirmenstein 1999b), and we have observed it re-sprouting within weeks after other Owens Valley fires. Fire effects on A. lentiformis subsp. forreyi have not been systematically studied, though some data suggest post-fire recovery 1s by seed germination rather than re- sprouting (Meyer 2005). Observations in Owens Valley show high rates of A. lentiformis subsp. torreyi germination in spring following winters with abundant precipitation (ICWD data on file). We found no data regarding fire effects on Machaeranthera carnosa. Overall, the absence of most shrub species in our observations and sampling data eight weeks after the fire suggests shrub mortality. 96 MADRONO DTW Effects One of the principal reasons we sampled post-fire vegetation was to investigate the spatial variation in re-growth observed in reconnaissance site visits. DTW explained most of the variance (Fig. 4b) in grass cover in our post-fire sample of the burned area. The significant difference in DTW between flowering and vegetative S. airoides samples is further evidence for the importance of DTW—as opposed to properties of the fire itself—in controlling patterns of re-growth. The fact that DTW also explained most of the variance in grass cover in both pre-fire and post-fire control datasets (Figs. 4a, 4c) suggests the post-fire recovery in the burned area is re-establishing cover—DTW relationships which existed before the fire. These patterns had previously been obscured from view by encroaching shrubs. These relationships between grass cover, phe- nology, and DTW may be understood as effects of differential water availability across the DTW gradient in both burned and un-burned portions of the study area. Cox et al. (1990) asserted that a “waving sea” of alkali sacaton could not be maintained where annual precipitation ranges from 150-400 mm, while revegetation experi- ments elsewhere in Owens Valley have docu- mented water requirements for D. spicata to range from 200-800 mm/yr (Dickey et al. 2005). However, with average annual precipitation of 160 mm and only 34 mm total precipitation prior to and during the 2007 growing season, a beautiful, ““waving sea” of alkali sacaton and saltgrass developed in the burned area at the shallow end of the DTW gradient eight weeks after the fire, but not at the drawn-down end. The simplest explanation for the observed patterns is differential groundwater availability. Livestock grazing is another important factor potentially controlling patterns of cover and phenology, but because livestock were not permitted in the study area after the fire, low cover in parts of the burned area was not a direct result of grazing. The pattern of un-grazed post- fire grass re-growth was similar to the pattern of grazed pre-fire grass distribution. While grazing undoubtedly affects the site, our results argue against differential grazing as a direct explanation for the observed variation in grass cover. We are not the first to infer the importance of DTW as a primary factor in controlling vegetation patterns in arid lands. Los Angeles Department of Water and Power eneimeer: C. HH. Lee. (1912) documented the “very striking” relationship be- tween vegetation and DTW in this portion of Owens Valley nearly a century ago, and ground- water availability is used to explain vegetation patterns on valley floors throughout the Great Basin (Coville 1893; Odion et al. 1992; Castelli et al. 2000; West and Young 2000; Cooper et al. 2006). [Vol. 56 The 2007 DTW gradient, which explains a high percentage of current variation in grass cover, is controlled by groundwater extraction (previously noted) and is quite different from the gradient Lee mapped in 1912 (Fig 2). The 2007 distribu- tion in grass cover, therefore, is best interpreted as a response to this altered hydrologic regime. Elmore et al. (2006) explained changes in perennial cover of Owens Valley alkali meadow habitat over 20 yr as a function of changes in DTW and precipitation. Cooper et al. (2006) described a similar case in Colorado. In our study we substituted space for time by examining grass cover before and after fire in a single growing season and obtained similar results. Management Implications At the shallow end of the DTW gradient (Fig. 2) the robust recovery of native grass provided dramatic evidence that fire can serve as a force for meadow regeneration. This finding is not surprising: burning to remove shrubs and promote grasses has been practiced in California since pre Euro-American settlement (Anderson 2006). More relevant to our research, Wright and Chambers (2002) conducted and studied results of controlled burns in Great Basin riparian meadow habitat located in both shallow and drawn-down water table sites. They concluded that burning is an effective tool for restoring meadows with shallow groundwater. At the drawn-down end of the DTW gradient, fire may be an agent of community change rather than a force for meadow regeneration. We expect vegetation there will cease to meet criteria for classification as groundwater dependent meadow and, instead, will come to resemble that of nearby precipitation-dependent desert shrublands. The expected multi-year period for shrub re-coloniza- tion, combined with negligible grass cover, will leave the drawn-down end of the DTW gradient highly vulnerable to erosion. In Owens Valley, — wind poses an especially great erosional force, and a major highway (US 395) was closed on at least one occasion after the fire due to the density of airborne meadow soil where the highway crossed the study area. Other studies predict vegetation change as a response to hydrologic | alterations under phreatophytic communities | (Odion et al. 1992; Manning 1999; Wright and | Chambers 2002; Naumburg et al. 2005; Patten et | al. 2008). Our data suggest fire may accelerate the | rate of the predicted vegetation change. Conclusion We conclude that if the 2007 DTW gradient is | maintained in the future, vegetation at opposite | ends of the gradient will follow different trajec- | 2009] tories. At the shallow end we expect grass cover will reach or exceed pre-fire levels in summer 2008 (preliminary data suggest this outcome) and alkali meadow habitat will be maintained. At the drawn-down end the fire will facilitate the conversion of groundwater-dependent alkali meadow to a more xeric community of annuals and, eventually, shrubs. Rather than acting as an independent force shaping patterns of vegetation, our short-term response data suggest the effect of the fire over the entire DTW gradient will be the long-term amplification of differences already present before the fire. These differences are associated with differential water availability. Our examination of fire effects led to documen- tation of groundwater dependence and the effects of groundwater limitation. ACKNOWLEDGMENTS We thank Jerry Zatorski and Sondra Grimm of the Inyo County Water Department for conducting the field work. We also thank the University of California White Mountain Research Station, John Callaway, and an anonymous reviewer. LITERATURE CITED ANDERSON, M. K. 2006. The use of fire by Native Americans in California. Pp. 417-430 in N. G. Sugihara, J. W. VanWagtendonk, K. E. Shaffer, J. Fites-Kaufman, and A. E. Thode, (eds.), Fire in California’s ecosystems. University of California Press, Berkeley, CA. BALDWIN, B. G., S. Boyb, B. J. ERTTER, R. W. PATTERSON, T. J. ROSATTI, AND D. H. WILKEN, (eds.). 2002. The Jepson desert manual: vascular plants of southeastern California. University of California Press, Berkeley, CA. BONHAM, C. D. 1989. Measurements for terrestrial vegetation. Wiley, New York, NY. CASTELLI, R. M., J. C. CHAMBERS, AND R. J. TAUSCH. 2000. Soil-plant relations along a soil-water gradi- ent in Great Basin riparian meadows. Wetlands 20:251—266. City OF Los ANGELES AND COUNTY OF INyo. 1991. Water from the Owens Valley to supply the second Los Angeles Aqueduct: 1970 to 1990, and 1990 onward, pursuant to a long term groundwater management plan, Final Environmental Impact Report. State Clearing House no. 89080705, De- partment of Water and Power, City of Los Angeles, CA. CiTy OF Los ANGELES DEPARTMENT OF WATER AND POWER (LADWP) AND INYO COUNTY. 2007. Agreement between the County of Inyo and the City of Los Angeles Department of Water and Power regarding an interim management plan for groundwater pumping in the Owens Valley. Exhibit A. Inyo County Water Department, Independence, CA. http://www.inyowater.org/Water_Resources/ InterimManagementPolicy/IMP_Final_1Mar07. pdf [accessed 2009, July 05]. Cooper, D. J., J. S. SANDERSON, D. I. STANNARD, AND D. P. GROENEVELD. 2006. Effects of long term water table drawdown on evapotranspiration PRITCHETT AND MANNING: EFFECTS OF FIRE ON ALKALI MEADOWS 97 and vegetation in an arid region phreatophyte community. Journal of Hydrology 325:21—34. COVILLE, F. V. 1893. Botany of the Death Valley expedition. U.S. Department of Agriculture, Con- tributions from the U.S. National Herbarium, Vol. 4. U.S. Government Printing Office, Wash- ington, DC. Cox, J. R., A. K. DOBRENZ, AND B. MCGUIRE. 1990. Evaluation of some alkali sacaton ecotypes collect- ed in Mexico. Applied Agricultural Research 5:164-168. DICKEY, J.. M. HALL, M. MADISON, J. SMESRUD, M. GRISWOLD, Q. COTTEN, M. HEILMANN, G. Ro- LAND, J. JORDAHL, R. HARASICK, W. BAMOssy, R. COLES, L. WHEELER, P. BROWN, K. BURTON, R. FORNELLI, I. ANDERSON, M. RIEDEL-LEHRKE, R. TILLER, AND J. RICHARDS. 2005. Stabilizing Owens dry lake surface with irrigated saltgrass, Part 2: the managed vegetation project. Ecesis 15:2. EAMUS, D., R. FROEND, R. LOOMES, G. HOSE, AND B. MURRAY. 2006. A functional methodology for determining the groundwater regime needed to maintain the health of groundwater-dependent vegetation. Australian Journal of Botany 54:97-114. ELMORE, A. J., S. J. MANNING, J. F. MUSTARD, AND J. M. CRAINE. 2006. Decline in alkali meadow vegetation cover in California: the effects of groundwater extraction and drought. Journal of Applied Ecology 43:770-779. ELZINGA, C. L., D. W. SALZER, AND J. W. WIL- LOUGHBY. 1998. Measuring and monitoring plant populations. USDI Bureau of Land Management Technical Reference 1730-1. BLM National Busi- ness Center, Denver, CO. HOLLETT, K. J.,. W. R. DANSKIN, W. F. MCCAFFREY, AND C. L. WALTI. 1991. Geology and water resources of Owens Valley, California. U.S. Geo- logical Survey water- supply paper 2370-B. U.S. Government Printing Office, Washington, DC. HOWALD, A. 2000. Plant communities. Pp. 94-207 in G. Smith (ed.), Sierra East: edge of the Great Basin. University of California Press, Los Angeles, CA. LEE, C. H. 1912. An intensive study of the water resources of a part of Owens Valley, California. U.S. Geological Survey water-supply paper 294. U.S. Government Printing Office, Washington, DC. LEGENDRE, P. AND L. LEGENDRE. 1998. Numerical ecology: second English edition. Elsevier Science, Amsterdam, The Netherlands. MANNING, S. J. 1999. The effects of water table decline on groundwater-dependent Great Basin plant communities in the Owens Valley, California. Pp. 231-237 in E. D. McArthur, W. K. Ostler, and C. L. Wambolt (eds.), Proceedings: Shrubland ecotones. 1998 August 12-14; Ephraim UT. Proceedings RMRS-P-11. USDA Forest Service, Rocky Mountain Research Station, Ogden, UT. MEYER, R. 2005. Atriplex lentiformis. In Fire Effects Information System, [online]. USDA Forest Ser- vice, Rocky Mountain Research Station, Fire Sciences Laboratory (Producer). http://www.fs.fed. us/database/feis [accessed 2009, July 05]. NAUMBURG, E., R. MATA-GONZALEZ, R. G. HUNTER, T. MCLENDON, AND D. W. MARTIN. 2005. Phreatophytic vegetation and groundwater fluctu- ations: a review of current research and application of ecosystem response modeling with an emphasis 98 MADRONO on Great Basin vegetation. Environmental Man- agement 35:726—740. OpDION, D. C., R. M. CALLAWAY, W. R. FERREN, JR., AND F. W. DAvis. 1992. Vegetation of Fish Slough, an Owens Valley wetland ecosystem. Pp. 171-197 in C. A. Hall, V. Doyle-Jones, and B. Widawski (eds.), The history of water: Eastern Sierra Nevada, Owens Valley, White-Inyo Moun- tains. White Mountain Research Station Sympo- sium Vol. 4. White Mountain Research Station, Bishop, CA. PATTEN, D., L. ROUSE, AND J. C. STROMBERG. 2008. Isolated spring wetlands in the Great Basin and Mojave Deserts, USA: Potential response of vegetation to groundwater withdrawal. Environ- mental Management 41:398-413. SAWYER, J. O. AND T. KEELER-WOLF. 1995. A manual of California vegetation. California Native Plant Society, Sacramento, CA. SMITH, L. M. AND J. A. KADLEC. 1985. Comparison of burning and cutting of Utah USA marsh plants. Great Basin Naturalist 45:462—-466. SORENSON, S. K., P. D. DILEANIS, AND F. A. BRANSON. 1991. Soil water and vegetation respons- es to precipitation and changes in depth to ground water in Owens Valley, California. U.S. Geological Survey water-supply paper 2370-G. U.S. Govern- ment Printing Office, Washington, DC. STEINWAND, A. AND R. HARRINGTON. 2003a. Simula- tion of water table fluctuations at permanent monitoring sites to evaluate groundwater pumping. Inyo County Water Department, Independence, CA. http://www.inyowater.org/ICWD_Reports/ Water_Table_Simulation.pdf [accessed 2009, July O35); AND ————. 2003b. Water table fluctuations resulting from management under the Drought Recovery Policy and the Green Book, 1989 to 2000. Inyo County Water Department, Indepen- [Vol. 56 dence, CA. http://www.inyowater.org/ICWD_ Reports/Water_Table_Fluctuations_GB.pdf [acc- essed 2009, July 05]. TER BRAAK, C. J. F. 1996. Unimodal models to relate species to environment. DLO-Agricultural Mathe- matics Group, AC Wageningen, The Netherlands. AND C. W. N. LOOMAN. 1995. Regression. Pp. 29-66 in R. H. G. Jongman, C. J. F. ter Braak, and O. F. R. Van Tongeren (eds.), Data analysis in community and landscape ecology. Cambridge University Press, Cambridge, U.K. TIRMENSTEIN, D. 1999a. Artemisia tridentata ssp. tridentata. In Fire Effects Information System, [online]. USDA Forest Service, Rocky Mountain Research Station, Fire Sciences Laboratory (Pro- ducer). http://www.fs.fed.us/database/feis. [accessed 2009, July O05]. 1999b. Chrysothamnus nauseosus. In Fire Effects Information System, [online]. USDA Forest Service, Rocky Mountain Research Station, Fire Sciences Laboratory (Producer). http://www.fs.fed. us/database/feis. [accessed 2009, July 05]. UNITED STATES DEPARTMENT OF AGRICULTURE (USDA), NATURAL RESOURCES CONSERVATION SERVICE. 2002. Soil survey of Benton-Owens Valley and parts of Inyo and Mono Counties, CA. http:// soildatamart.nres.usda.gov/Manuscripts/CA802/0/ Benton_OwensValley_CA.pdf [accessed 2009, July OS]. WRIGHT, J. M. AND J. C. CHAMBERS. 2002. Restoring riparian meadows currently dominated by Artemi- sia using alternative state concepts: above-ground vegetation response. Applied Vegetation Science 5:237-246. WEsT, N. E. AND J. A. YOUNG. 2000. Intermountain valleys, and lower mountain slopes. Pp. 256—284 in M. G. Barbour and W. D. Billings (eds.), North American terrestrial vegetation. 2nd ed. Cam- bridge, New York, NY. MADRONO, Vol. 56, No. 2, pp. 99-103, 2009 ARCEUTHOBIUM RUBRUM (VISCACEAE) IN MEXICO ROBERT L. MATHIASEN School of Forestry, Northern Arizona University, Flagstaff, AZ 86011 Robert.Mathiasen@nau.edu CAROLYN M. DAUGHERTY Department of Geography, Planning, and Recreation, Northern Arizona University, Flagstaff, AZ 86011 BRIAN P. REIF School of Forestry, Northern Arizona University, Flagstaff, AZ 86011 ABSTRACT Arceuthobium rubrum (Viscaceae) is a distinctive species of dwarf mistletoe having red to reddish- brown plants and red, shiny fruits. It is primarily distributed in the Sierra Madre Occidental of Durango, Mexico, but in 1972 a population of A. rubrum was reported from Oaxaca, Mexico, more than 1000 km south of the nearest population in southern Durango. Initially, this population was classified as a disjunct population of A. rubrum, but in 1989 it was described as a new species: A. oaxacanum. However, our morphological measurements and observations of the phenology for plants from Durango and Oaxaca indicate these populations are morphologically similar and flower and disperse seed at approximately the same time, supporting the results of recent molecular analyses indicating that the Oaxacan populations represent disjunct populations of A. rubrum. RESUMEN Arceuthobium rubrum (Viscaceae) es una especie distintiva de muérdago enano de color rojo a café rojizo y con frutos rojo brillantes. Se distribuye principalmente en la Sierra Madre Occidental de Durango, Mexico, pero en 1972 se reporto una poblacion para Oaxaca, México, a mas de 1000 km al sur de la poblacion mas cercana al sur de Durango. La poblaci6n de Oaxaca fue clasificada iniclalmente como una poblacion disyunta de 4. rubrum, pero en 1989 fue descrita como una nueva especie: A. oaxacanum. Sin embargo, nuestras mediciones morfologicas y observaciones de la fenologia de las plantas de Durango y Oaxaca indican que esas poblaciones son morfologicamente similares y florecen y dispersan las semillas aproximadamente al mismo tiempo, lo cual apoya los resultados de recientes analisis moleculares que indican que las poblaciones de Oaxaca representan poblaciones disyuntas de 4. rubrum. Key Words: Arceuthobium oxacanum, Arceuthobium rubrum, México, morphology, phenology. Arceuthobium rubrum Hawksw. & Wiens (Ru- by dwarf mistletoe, Viscaceae) is a distinct species with reddish to brownish-red plants and shiny fruits and is a common parasite of several pines (Pinus spp., Pinaceae, Table 1) in northwestern Mexico (Hawksworth and Wiens, 1965, 1972, 1996; Mathiasen et al. 2008). When a dwarf mistletoe with brownish red plants and shiny, red fruits was discovered in western Oaxaca on Pinus pseudostrobus Lindley by Dr. Roger Peterson in 1972, Hawksworth and Wiens (1977) classified it as a disjunct population of A. rubrum. However, based on further studies of two populations of A. rubrum in Oaxaca, Hawksworth and Wiens concluded that these populations were sufficient- ly distinct from those in Durango to describe the Oaxaca populations as a new species: A. oax- acanum Hawksw. & Wiens (Hawksworth and Wiens 1989). They separated A. oaxacanum from A. rubrum based primarily on differences in plant size, plant color, length of pistillate spikes, branching angle of staminate and pistillate spikes, and the tendency to form systemic infections or not (Hawksworth and Wiens 1989). Host range was also mentioned as a possible distinction between these taxa, although the hosts of A. oaxacanum do not occur in the range of A. rubrum (Table 1). Comparisons of A. rubrum and A. oaxacanum have been difficult because Hawksworth and Wiens (1989) did not include a description of the flowers of A. oaxacanum when they described it and only included information on the mean length of fruits (3.5 mm) in their Latin descrip- tion of this species. Furthermore, in their revised monograph of Arceuthobium, Hawksworth and Wiens (1996) did not include any information on the characteristics of the flowers or fruits of A. oaxacanum. More recently, Nickrent et al. (2004) reported that A. rubrum and A. oaxacanum could not be distinguished using internal transcribed spacer DNA and chloroplast trnl DNA sequenc- 100 TABLE 1. HOSTS OF ARCEUTHOBIUM RUBRUM IN DURANGO AND OAXACA FROM HAWKSWORTH AND WIENS (1989, 1996). 'Host susceptibility class is defined in Hawksworth and Wiens (1996). *Hosts of Arceuthobium oaxacanum from Hawksworth and Wiens (1989, 1996). *Does not occur in Oaxaca (Perry, 1991; Farjon and Styles, 1997). *Does not occur in Durango (Perry, 1991; Farjon and Styles, 1997). °Distributed in both Durango and Oaxaca, but did not occur in stands infested by A. rubrum in Oaxaca. Host susceptibility class! Durango Oaxaca’ Principal Pinus cooper? Pinus lawsonii* P. durangensis’ — P. michoacana* P. engelmannii’? — P. pseudostrobus* P. herrerav’ P. teocote’ Occasional P. oaxacana* Durango MADRONO [Vol. 56 es. Therefore, in 2007, we began studies of the Durango and Oaxaca populations of A. rubrum in order to gather additional morphological data for their flowers, fruits, and seeds. This allowed the flower and fruit characteristics of A. rubrum and A. oaxacanum to be compared for the first time. We also made additional observations of the phenology of both species starting in 1999. METHODS Morphological Measurements Specimens from two of the three currently known populations of A. oaxacanum were sampled in 2007 and five populations of A. rubrum were sampled in 2008 (Fig. 1). These populations included the type localities for A. rubrum and A. oaxacanum (Hawksworth and Gulf of Mexico Fic. 1. Approximate locations of populations of A. rubrum sampled in Durango and Oaxaca, Mexico. Sites are | listed under specimens examined; site 1 is the type locality for A. rubrum and site 6 the type locality for — A. oaxacanum. 2009] Wiens 1965, 1989; Fig. 1). For each population, 20-30 non-systemically infected branches with mature plants (10-15 males and 10—15 females) were collected. The largest dwarf mistletoe plant from each infection was measured; plant mea- surements were made within 24 hrs of collection. Measurements for flowers and fruits/seeds were completed in July or August when flowers were open and fruits were mature. The following morphological characters were measured: 1) plant height, basal diameter, third internode length and width, and color; 2) mature fruit length, width, and color; 3) seed length, width and color; 4) staminate flower diameter for flowers with 3 perianth lobes; 5) number, color of the adaxial surface, and length and width of staminate perianth lobes; 6) anther distance from the perianth lobe tip; and 7) anther diameter. A total of 50 measurements were made for each of these characters. Furthermore, we only measured plants from non-systemic infections for both species, because plants from systemic infections tend to be smaller than those from non-systemic infections (Hawksworth and Wiens 1996). A one- way analysis of variance (ANOVA, P = 0.05) was used to determine if there were statistical differences between the means of the morpholog- ical characters measured. Specimens examined. MEXICO. DURANGO. 47 km E of El Salto on Mexico Rte. 40, on Pinus teocote, 2 Aug 2008, Mathiasen 0808 (ASC) (site 1, Fig. 1, type locality of A. rubrum); 6.5 km S of Mexico Rte. 40 on rd to Regocio, on Pinus teocote, 31 Jul 2008, Mathiasen 0813 (ASC) (site 2, Fig. 1); 12.6 km S of Mexico Rte. 40 on rd. to Regocijo, on Pinus tecote, 31 Jul 2008, Mathiasen 0816 (ASC) (site 3, Fig. 1); 51 km S of Durango on rd. to La Flor, on Pinus teocote, 30 Jul 2008, Mathiasen O&11 (ASC) (site 4, Fig. 1); 5 km W of El Salto on Mexico Rte. 40, on Pinus teocote, 29 Jul 2008, Mathiasen 0809 (ASC) (site 5, Fig. 1). OAXACA. 13 km S of Miahuatlan on Mexico Rte. 175, on Pinus lawsonii, 22 Jul 2007, Mathiasen 0727 (ASC), and 24 Jul 2007, Mathia- sen 0729 (ASC) (site 6, Fig.1, type locality of A. oaxacanum); 7 km W of Santo Tomas Tamazu- lapan, on Pinus lawsonii, 22 Jul 2007, Mathiasen 0726 (ASC) (site 7, Fig. 1). Phenology Observations of flowering and seed dispersal in Durango were made in December 1999, March 2003, March 2004, July 2005, March 2007, September and October 2007, and July 2008. Observations in Oaxaca were made in December 2000, September 2004, September 2006, and July 2007. MATHIASEN ET AL.: ARCEUTHOBIUM RUBRUM IN MEXICO 101 RESULTS Morphological Measurements The means and ranges for plant heights, basal diameters, and third internode dimensions of the populations we sampled in Oaxaca and Durango were nearly identical (Table 2). Male plants were slightly larger than female plants from both regions, but approximately the same for each sex. Only the length of the third internode of male plants from Oaxaca was significantly different than the length of this internode for male plants from Durango. Hawksworth and Wiens (1989, 1996) reported that the mean height of A. rubrum plants (male and female combined) from Dur- ango was approximately 10 cm and those from Oaxaca were approximately 12 cm. When we combined male and female plant height data, the mean heights of plants from both areas were approximately 12 cm (Table 2). The ranges in plant heights Hawksworth and Wiens (1989) reported were approximately the same; 8—18 and 8-20 cm for plants from Durango and Oaxaca, respectively. The plant height ranges we recorded were only slightly different than those reported by Hawksworth and Wiens, but we also found that the range in plant heights was approximately the same in both regions. The mean widths for staminate spikes were also similar between the populations in Oaxaca and Durango, but the length of staminate spikes was significantly different (Table 3). Although the means for the flower characters measured were slightly larger for the flowers from male plants in Oaxaca, the means were not significant- ly different. The flower diameters for 3-merous flowers were smaller than most other species of TABLE 2. PLANT HEIGHTS, BASAL DIAMETERS, AND THIRD INTERNODE MEASUREMENTS FOR ARCEUTHOBIUM RUBRUM IN DURANGO AND OAXACA. Measurements are expressed as mean (standard deviation) values. Mean values between the two locations are not significantly different except for the third internode length in males (ANOVA, P = 0.05). Character Durango Oaxaca Plant height (cm) Male 14.2 (3.9) 13.9 (327) Female 10:3°(2.5). .10:0'@.7) All plants (cm) P2.3'(3:)) 194327) Basal diameter (mm) Male 3,7 (1.0) 3:2 (0.0) Female 4.0 (1.0) 3.8.(1.0) Third internode length (mm) Male 10-9°(3 5) 9.8 (3.2) Female Yel (2,3) Ouse (2.5) Third internode width (mm) Male 2.8 (0.6) 2.7 (0.6) Female 3.0 (0.6) 2.2 (0.6) 102 Arceuthobium (Hawksworth and Wiens 1996), averaging only 1.6 and 1.7 mm in Durango and Oaxaca, respectively. Another unusual character- istic of the flowers of A. rubrum reported by Hawksworth and Wiens (1965, 1996) is that it had flowers that were barely open, even during their peak of anthesis. We also observed this characteristic in Durango, but some flowers did open to the extent that they could be easily measured. Our observations in Oaxaca indicated the flowers there also do not open completely during anthesis. Fruit and seed dimensions demonstrated the greatest difference between the mistletoe popula- tions in Oaxaca and Durango, but differences were only 0.1—0.2 mm larger for the populations in Oaxaca and were not significantly different (Table 3). Although Hawksworth and Wiens (1989, 1996) reported that the mean fruit length in Oaxaca and Durango was about 3.5 mm, none of the mature fruits we measured in either state were that small. The mean lengths of mature fruits we measured in Oaxaca and Durango were nearly one mm larger than reported by Hawks- worth and Wiens (4.4 and 4.3 mm, respectively). We measured mature fruit as large as 5 mm in Durango and 5.2 mm in Oaxaca. Plant color for male and female plants was similar for populations from Oaxaca and Dur- ango also. The predominant color of male and female plants in Durango and Oaxaca was reddish-brown (76%, and 70%, respectively). Other plants were either red or brown, but a few female plants in Durango were very dark; nearly black in color. Plants from both regions that were red when fresh dried to a dull brown. Another distinction between the Oaxaca and Durango populations noted by Hawksworth and Wiens (1989, 1996) was that the dwarf mistletoe typically induced systemic infections on its pine hosts in Oaxaca, while those in Durango typically TABLE 3. STAMINATE SPIKE, FLOWER, FRUIT, AND SEED MEASUREMENTS FOR ARCEUTHOBIUM RUBRUM IN DURANGO AND OAXACA. Measurements are expressed as mean (standard deviation) values. Mean values between the two locations are not significantly different except for staminate spike leneth (ANOVA, P = 0.05). Character Durango Oaxaca Staminate spike length (mm) 11.4 (2.4) 10.8 (3.4) Staminate spike width (mm) 1.5 (0.1) 1.6 (0.2) Staminate flower diameter (mm) 1.6 (0.2) 1.7 (0.2) Perianth lobe length (mm) 0.7 (0.06) 0.8 (0.06) Perianth lobe width (mm) 0.7 (0.10) 0.8 (0.14) Anther diameter (mm) 0.3 (0.08) 0.4 ( 0.08) Mean fruit length (mm) 4.3 (0.3) 4.4 (0.4) Mean fruit width (mm) 2.8 (0.2) 3.0 (0.3) Mean seed length (mm) 2.4 (0.2) 2.5 (0.3) Mean seed width (mm) 1.2 (0.09) 1.3 (0.13) MADRONO [Vol. 56 did not. However, our observations in Oaxaca indicated that the mistletoe also commonly formed non-systemic infections there. In Dur- ango, we observed that A. rubrum frequently induced systemic infections, particularly on Pinus teocote Schiede ex Schltdl. & Cham. and Pinus durangensis Martinez. Therefore, the infrequency of non-systemic infections in Oaxaca versus systemic infections in Durango does not appear to be a consistent difference between these populations. Furthermore, we did not observe any distinctive differences in the size of the witches’ brooms caused by these dwarf mistletoes on their pine hosts in Oaxaca versus Durango, although Hawksworth and Wiens (1989, 1996) reported that the mistletoe in Oaxaca induced larger brooms on its hosts. Hawksworth and Wiens (1989) also justified the classification of the populations in Oaxaca as a distinct species because plants there had spikes that branched at nearly right angles (90°) from the main axis of plants while the spikes of plants in Durango usually branched at angles of about | 45°. We observed that younger plants in both areas tended to have spikes that branched at about 45° and older plants had more spikes that | branched at nearly right angles to the main axis | of the plant. In addition, the young staminate | spikes near the top of older male plants from | both regions tended to branch at about 45° and older spikes commonly branched at right angles. However, these branching patterns varied widely between different populations. Therefore, the branching angle of spikes was not a consistent — difference between the populations in Durango | and Oaxaca. The A. are much larger than those in other areas of Durango (Hawksworth and Wiens 1996). We. examined two populations in July 2008 that | Hawksworth and Wiens (1996) classified as A. | rubrum near Altares and classified these as Arceuthobium vaginatum (Willd.) Presl subsp. | vaginatum. Male and female plants of these populations were very large, dark brown to) nearly black, and had much larger basal diame- | rubrum plants we. examined elsewhere in Durango. Although we also noted that some of the fruits of these | populations were shiny, they were not dispersing | we | ters (>1 cm) than the A. seed as were populations of A. rubrum examined in Durango at that time. Furthermore, | male plants of the Altares populations were not, flowering in July 2008. They appeared to have | already flowered earlier in the year. Therefore, | the phenology of the Altares populations also | supports their classification as A. subsp. vaginatum because this taxon flowers in March-April and disperses seed in August through September (Hawksworth and Wiens vaginatum rubrum populations near Altares, | Durango have been reported to have plants that | 2009] 1996). Although the Altares populations Hawks- worth and Wiens’ classified as A. rubrum were parasitizing Pinus arizonica Engelm., they never classified this pine as a host of A. rubrum (Hawksworth and Wiens 1996, see pages 49 and 364). Phenology Male plants in Durango began flowering in early July and flowering peaked from mid to late July, but some plants continued to flower though August into early September. Male plants in Oaxaca began flowering at the same time, but some plants were also observed flowering in mid September. Therefore, the time of anthesis is nearly the same for these dwarf mistletoe populations, but extends slightly longer in Oaxaca. Seed dispersal started in mid July in Durango and continued into September, with its peak in late August to early September. Seed dispersal in Oaxaca occurred slightly later based on observations in September 2007. Female plants in Oaxaca were just starting to disperse seed in mid September that year, so seed dispersal peaked in late September and continued into October. The slightly later periods for flowering and seed dispersal for the Oaxaca populations may be due to their more southern distribution or that they occurred at lower elevations (1760— 2100 m) than the populations in Durango (2300— 2600 m). DISCUSSION Our morphological data and observations of the phenology of the A. rubrum populations from Oaxaca and Durango supported the classification of these widely separated populations as one species, as was concluded earlier by Hawksworth and Wiens (1977) and supported with molecular analyses by Nickrent et al. (2004). Therefore, we recommend that the dwarf mistletoe populations with reddish brown plants in Oaxaca be treated as extremely disjunct populations of A. rubrum. Because of the similarity between the male and female plants, flowers, and fruits of the Oaxaca and Durango populations of A. rubrum, the classification of these populations as different species is not warranted at this time. The only host of A. rubrum that occurs in both Durango and Oaxaca is Pinus teocote (Perry 1991; Farjon and Styles 1997), but this pine does not occur in the A. rubrum populations we sampled in Oaxaca (Table 1). Therefore, host susceptibility based on natural infection of pines in Oaxaca and Durango could not provide information on the taxonomic relationship of MATHIASEN ET AL.: ARCEUTHOBIUM RUBRUM IN MEXICO 103 these populations (Hawksworth and Wiens 1989). Although the pine hosts of A. rubrum in Oaxaca are not closely related to its pine hosts in Durango (Farjon and Styles 1997), this is insufficient evidence to support the classification of the Oaxaca populations as a different species. The large geographic separation of the A. rubrum populations in Oaxaca from those in Durango (ca. 1200 km) certainly suggests that these populations could have developed charac- ters that support separate taxonomic status. Although some of the characters we measured were slightly larger for the populations of A. rubrum we sampled in Oaxaca than those in Durango, we do not feel the differences were large enough to support the recognition of the Oaxaca populations as a different species or subspecies. If additional populations of A. rubrum were to be found in the large geographic gap between Oaxaca and Durango, it would lend further support to our interpretation of the widely separated populations of A. rubrum being conspecific. ACKNOWLEDGMENTS The assistance of Dr. M. Socorro Gonzalez Elizondo with an early review of the manuscript and the Spanish translation for the Resumen is greatly appreciated. LITERATURE CITED FARJON, A. AND B. STYLES. 1997. Pinus (Pinaceae). Flora Neotropica, Monograph 75, New York Botanical Gardens, New York, NY. HAWKSWORTH, F. G. AND D. WIENS. 1965. Arceutho- bium in Mexico. Brittonia 17:213—238. AND . 1972. Biology and taxonomy of dwarf mistletoes (Arceuthobium). Agricultural Handbook 401, USDA Forest Service, Washing- ton, DC. AND . 1977. Arceuthobium in Mexico: additions and range extensions. Brittonia 29:411-418. AND . 1989. Two new species, nomen- clatural changes, and range extensions in Mexican Arceuthobium (Viscaceae). Phytologia 66:3—11. AND . 1996. Dwarf Mistletoes: Biology, pathology, and systematics. Agriculture Handbook 709, USDA Forest Service, Washington, DC. MATHIASEN, R. L., M. S. GONZALEZ, M. GONZALEZ, B. E. HOWELL, I. L. LOPEZ, J. M. SCOTT, AND J. A. TENA. 2008. Distribution of dwarf mistletoes (Arceuthobium spp., Viscaceae) in Durango, Mex- ico. Madrono 55:161—160. NICKRENT, D. L., M. A. GARCIA, M. P. MARTIN, AND R. L. MATHIASEN. 2004. A phylogeny of all species of Arceuthobium (Viscaceae) using nuclear and chloroplast DNA sequences. American Journal of Botany 91:125—138. PERRY, J. P. 1991. The pines of Mexico and Central America. Timber Press, Portland, OR. MADRONO, Vol. 56, No. 2, pp. 104-111, 2009 NOTES ON CALIFORNIA MALVACEAE INCLUDING NOMENCLATURAL CHANGES AND ADDITIONS TO THE FLORA STEVEN R. HILL Illinois Natural History Survey, 1816 S. Oak Street, Champaign, IL 61820 srhill@inhs.uiuc.edu ABSTRACT The writing of revised treatments for selected California Malvaceae for the upcoming second edition of the Jepson Manual and the Flora of North America series (volume 6) has made several nomenclatural changes and explanations necessary. New combinations are made here for taxa in Sidalcea, including Sidalcea asprella subsp. nana, Sidalcea calycosa subsp. rhizomata, Sidalcea celata, and Sidalcea sparsifolia. Several taxa previously included within Sidalcea malviflora have been removed from that species and re-interpreted, resulting in the resurrection and acceptance of the names Sidalcea asprella Greene and Sidalcea elegans Greene. Comments are presented here on the status of Hibiscus lasiocarpos and Lavatera vs. Malva in the California flora. At least one native species has been added to the flora, namely, //iamna rivularis, though it may no longer occur in the state. Four species of Malvaceae have become naturalized or have been found as waifs in recent years and are added to the flora, namely, Anoda pentaschista, Lagunaria patersonia, Lavatera olbia, and Lavatera trimestris. Key Words: California, Hibiscus, Lagunaria, Lavatera, Malva, Malvaceae, North America, Sidalcea. The preparation of revised treatments of several genera of California Malvaceae for the upcoming revision of the Jepson Manual— Higher Plants of California (TJM1=Hickman 1993) and for the new Flora of North America, Volume 6 (in preparation) led to the need for several nomen- clatural changes as well as an explanation for some of the changes as compared to previous treatments (e.g., Hill 1993). In addition, several taxa have been noted that were not included in TJM1 for the flora of California. Newly described taxa in Sidalcea have been or shall be published elsewhere (Hill 2008; Clifton, Buck and Hill unpublished). NEW COMBINATIONS AND INTERPRETATIONS IN SIDALCEA Sidalcea A.Gray is the most species-rich of the genera of Malvaceae in California and it is a near-endemic there. It is also one of the most perplexing of the genera taxonomically, and while several attempts have been made to better define the taxa (Roush 1931; Hitchcock 1957; Dimling 1991; Hill 1993; Andreasen and Baldwin 2001, 2003a, b; Andreasen 2005), some remain difficult to delineate. The treatment by Hitchcock (1957) attempted a synthesis using phytogeographic, morphological, and chromosomal data, and he utilized four ranks: genus, species, subspecies, and variety in an attempt to sharply define the variants. An examination of his treatment revealed that most species, subspecies, and varieties that he described were said to have transitional individuals to other taxa, and in some groups of species it was nearly impossible to identify many of the individuals conclusively. His hand-written notes on specimens in some herbar- ia also revealed his frustration with these plants (e.g., on Blankinship s.n., JEPS 2856, the type of Sidalcea malviflora (DC.) A. Gray var. celata Jeps., C. L. Hitchcock wrote: “S. malvaeflora ssp. celata—unless I find cause for changing my opinion’, and on JEPS 2855: “S. malvaeflora ssp. celata—I believe I shall call this’’). A previous treatment of the genus in California (Hill 1993) attempted to make some sense of the species, but it tended to err on the side of combining variants rather than recognizing them to reduce the number of names. After working several more years with these species as well as with many more both new and old collections, I have attempted to clarify some of the problems created by combining the variants. Admittedly, the changes still have not resulted in a ‘perfect’ treatment by any means, but my first goal has been to re-interpret some of the variants. The | changes in interpretation are supported by both | morphological and geographic consistency aftera — reexamination of type and additional material. Second, the work of Andreasen and Baldwin (2001, 2003a, b) and Andreasen (2005) utilizing . new molecular phylogenetic data has helped to clarify some of the relationships within the genus ~ since the 1993 treatment, and a goal was to . reposition and rename some taxa to incorporate | some of the major changes suggested by the molecular work. Among the hypotheses support- | ed by the new data are the following: 1) the basal | perennials are Sidalcea hickmanii Greene, S. - malachroides (Hook. & Arn.) A. Gray, and S. | stipularis J. T. Howell & True, 2) four lineages, or 2009] clades, of the remaining perennial species appear to be well-supported, the ‘malviflora clade’, the ‘oregana Clade’, the ‘glaucescens clade’, and the ‘asprella clade’, 3) within the ma/viflora clade, the primarily coastal subspecies of Sidalcea malvi- flora (subsp. malviflora, subsp. laciniata C. L. Hitche., subsp. patula C. L. Hitche., subsp. purpurea C. L. Hitche., and subsp. rostrata (Eastw.) Wiggins) form a very coherent and closely related group, whereas the somewhat more interior subsp. sparsifolia C. L. Hitchce. and subsp. californica (Torr. & A. Gray) C. L. Hitche. (and subsp. dolosa C. L. Hitche. ?) are somewhat divergent from those, and 4) the plants treated as S. malviflora subsp. asprella (Greene) C. L. Hitche. in the 1993 revision are not in the same lineage, or clade, as the other subsp. of S: malviflora, nor are they all necessarily very close to each other, but, instead, are more closely related to the mountain species S. glaucescens Greene and the foothill species S. robusta Heller ex Roush. While several additional working hypotheses can be derived from the molecular work, these four, especially the latter three, would seem to affect the classification and nomenclature of the California perennials the most. The inland, mostly mountain plants that had been tossed into the ‘dust bin’ of Sidalcea malviflora subsp. asprella had to be reassessed, and this has been the emphasis in the recent studies. Hitchcock (1957) considered Sidalcea malvi- flora [‘malvaeflora’| to be a single widespread species ranging from Baja California, Mexico, north to the Willamette Valley of Oregon, and he divided it into 12 rather geographically coherent subspecies, some of which were subdivided into varieties. Eleven of the subspecies were recog- nized in California, subsp. virgata (Howell) C. L. Hitche. of Oregon being the only exception. Hill (1993) reduced the number of California subspe- cies to eight, combining Hitchcock’s subsp. celata, elegans, and nana into the single Sierran subsp. asprella partly because of the numerous comments on transitional individuals in Hitch- cock’s 1957 revision. Over the years since, and after the examination of many more collections, it was decided that this subspecies circumscrip- tion has become far too broad to be useful, and refinement has been attempted. This group of difficult variants resides primar- ily in Andreasen and Baldwin’s (2003a, b) ‘asprella clade’. Regarding this ‘asprella clade’, _ Hitchcock’s treatment and keys were generally unusable. Plants of very different appearance from distant geographical areas and_ habitats would often key to the same subspecies. Andrea- _ sen and Baldwin (2003a, b) demonstrated that the Hill (1993) concept of subsp. aspre/la was actually _ polyphyletic, and their different samples of that subspecies did not cluster together in the final analysis. HILL: NOTES ON CALIFORNIA MALVACEAE 105 Inheriting this problem, I decided to start over and reexamine the type specimens in the group, keeping this new molecular data result always in mind. Within this group, a new and undescribed species of Sidalcea had also been brought to my attention (Clifton, Buck and Hill unpublished), and studies of this as well as the other entities within the ‘asprella’ and ‘glaucescens’ clades have helped to resolve the problems to a certain extent. I decided to recognize and describe as best I could the morphologically and geographically distinct taxa that sorted out with the new data. Therefore, I now propose the following nomenclatural and taxonomic changes within Sidalcea. SIDALCEA ASPRELLA Greene subsp. ASPRELLA, Bulletin of the California Academy of Sciences 1:78.1885.—Type: USA, California, Yuba Co., near Camptonville, 1 Jul 1884, E. L. Greene s.n. (lectotype, here designated: CAS 1121!). Synonym: Sidalcea malviflora |‘malvaeflora’| (DC.) A. Gray subsp. aspre/la (Greene) C. L. Hitche., University of Washington Publica- tions in Biology 18:25. 1957. Edward L. Greene, in describing this species in 1885, cited two specimens, and some of his other remarks (p. 78) bear repeating: “‘On bushy hillsides of the lower Sierras, just below the habitat of Chamaebatia; apparently not collected before last season; found by Mrs. Curran in El Dorado County, and by the writer on Mr. John Ramm’s ranch, near Camptonville, in Yuba County. Peculiar, at least among the perennial species, in having the leaves all of precisely the same shape, the lowest and the uppermost differing only in point of size. The rough pubescence is likewise very characteristic.”’ The lectotype shows these features well, as do many other specimens from the Sierras. However, Hitchcock (1957) changed the circumscription of this species to include many plants with hair, habit, and leaf features that did not match the type or Greene’s conception. Over the years since, numerous specimens that vary considerably from the original concept have been determined to be this species, and the name has become a ‘dustbin’ for difficult Sierra plants. While some variation certainly appears to be present, the resurrection and more precise application of Greene’s original name, and to a greater extent, his concept of the species, should prove useful for current and future studies. SIDALCEA ASPRELLA Greene subsp. NANA (Jeps.) S. R. Hill, comb. nov.—Sidalcea reptans var. nana Jeps., Flora of California, 2:489. 1936.— Type: USA, California, Trinity Co., Soldier’s Ridge, SE Trinity Co. (Yollo Bolly Moun- tains), 24 Jul 1897, W. L. Jepson 14061 (holotype: JEPS 2856!; isotype: JEPS 2858!). Synonym: Sidalcea malviflora [‘malvaeflora | 106 MADRONO subsp. nana (Jeps.) C. L. Hitche., University of Washington Publications in Biology 18:29. 195d. Willis L. Jepson, in describing this as a variety in 1936, considered it to be a close relative of Sidalcea reptans Greene because of its very long thin rhizomes. The specimens on the type sheet are dwarfed, and all of the inflorescences are <10 cm long. Not all specimens are as small as the type—instead, robust individuals can reach a height of as much as 40 cm, yet they still share the other important diagnostic characters. The morphological features match those of S. asprella subsp. asprella well, but the subspecies is distinctive in its long slender rooting rhizomes and its fewer-flowered, often short inflorescences and occasionally few-leaved (1—3) stems that are decumbent-based. Otherwise the hairs through- out and toothed leaf lobes are a close match to S’ asprella subsp. asprella. Studies by Andreasen and Baldwin (2003a, b) show this plant to be, perhaps, closer to S. glaucescens than S. asprella, but more samples are needed to test this as the morphology does not support this placement (e.g., S. glaucescens and the very similar S. multifida Greene do not have elongated rooting rhizomes of any kind). Certainly this subspecies, as well as the typical subspecies, can no longer be included within the more coastal S. malviflora based upon the molecular data, and so a new name was needed. SIDALCEA CALYCOSA subsp. RHIZOMATA (Jeps.) Munz ex S. R. Hill, comb. nov.—Sidalcea rhizomata Jeps., Manual of Flowering Plants of California 629. 1925.—Type: USA, Cali- fornia, Marin Co., Point Reyes Peninsula, marsh near Russell’s Creamery, 16 Sep 1900, W. L. Jepson 1174 (holotype: JEPS 2861); isotypes: JEPS 2859! MO!—not yet acces- sioned with a number at the time of its inspection). There are two sheets of the type collection at JEPS; the sheet accessioned as JEPS 2861 has Jepson’s handwritten designation as ‘type’ on it, and includes the fertile material; the duplicate sheet, JEPS 2859, consists of the sterile, creeping, rooting rhizomes with scattered leaves also characteristic of this subspecies. Munz (Munz and Keck 1959 p. 132; Munz 1968 p. 12) used the name “‘Sidalcea calycosa M. E. Jones subsp. rhizomata (Jeps.) Munz,” but did not validly publish this new combination there or elsewhere. He, perhaps, did not realize that post-1952 new combinations require direct references to the basionyms. Its inclusion here serves to validate the combination. SIDALCEA CELATA (Jeps.) S. R. Hill, comb. nov. —Sidalcea malviflora [‘malvaeflora’| var. celata [Vol. 56 Jeps., Flora of California 2:493. 1936.—Type: USA, California, Shasta Co., Olinda, 11 May 1911, J. W. Blankinship s.n. (holotype: JEPS 2856!; isotype: JEPS 2855!, possibly WIS!—but dated 16 Apr 1911). The two JEPS specimens are quite different at first glance — JEPS 2856 is a single stem with nicely spread leaves, and JEPS 2855 has 2 stems and a good caudex, but the leaves are badly wrinkled. The two together supply a good series of characters to define the species, however JEPS 2856 bears a label indicating it is the type, and JEPS 2855 bears a label indicating it is an isotype. Both bear several annotation labels by the experts, including C. L. Hitchcock whose com- ments have been included above. While the lower leaves of the holotype have some resemblance to those of the type of S. asprella, other characters do not fit that species; some of the contrasting features include the lack of rhizomes, the presence of stiff reflexed bristle hairs at the base of the stem (a primary character for S. celata), and the upper leaves have narrow, often entire lobes, whereas S. asprella, as here defined, generally has some short rhizomes, coarse stellate hairs at the stem base, and upper leaves that are somewhat similar to those below, with wider lobes that are generally toothed. Upon using these characters on additional specimens, I discovered that Sidalcea celata is a species that is rather narrowly distributed in dry open oak woodlands mostly associated with serpentine in — Shasta and adjacent Tehama Cos., whereas S. asprella appears to be widely distributed in the central and northern Sierra Nevada range and to | the northwest, at the margin of more mesic | coniferous woodlands, either associated with | serpentine and serpentine-like minerals, or not. SIDALCEA ELEGANS Greene, Cybele Columbiana | 1:35. 1914.—Type: USA, Oregon, Josephine | Co., Eight Dollar Mountain, 12 Jun 1904, C. V. Piper 6171 (holotype: US 527772!; photo- graph of holotype at MO 940080!). Synonym: | Sidalcea malviflora |[‘malvaeflora| (DC.) A. Gray subsp. elegans (Greene) C. L. Hitchce., © University of Washington Publication in Biol- — opy 16:27,-1957: Sidalcea elegans is rather easily distinguished | from the other members of the ‘aspre/la clade’ by means of the relatively long, soft, simple hairs at the base of the stem, sometimes so sparse as to be | nearly lacking. The stems are characteristically | brittle and easily snapped when fresh, a character | not mentioned for other taxa in the genus (but | not especially useful on herbarium specimens!). | The upper stems are sometimes glaucous, and because of that feature as well as the one-sided | inflorescences that are often slightly curved 2009] between the flowers, the long acuminate calyx lobes, and the decumbent stems the species has sometimes been reported as Sidalcea glaucescens, a species without the long rooting rhizomes of S. elegans and that is not known in Oregon. Sidalcea elegans appears to be restricted to serpentine, and it is found in the Klamath Mountains of California and Oregon. Roush (1931) treated this taxon as a synonym of S. aspre/la and Hitchcock (1957) stated that “If asprella were to be treated as a species, ssp. elegans would best be considered thereunder.” In contrast, Dimling (1991) stated “Since this subspecies [S. malviflora [‘malvae- flora’| subsp. e/egans] 1s so clearly distinct from S-. malvaeflora ssp. asprella, its taxonomic identity will not be discussed further’’. Hill (1993) treated it as a synonym of S. malviflora (‘malvaeflora ) ssp. asprella and left it within that variable complex. In an attempt to clarify its position and nomenclature, the name Sidalcea elegans Greene is here resurrected, because the taxon 1s not a part of S. malviflora based on molecular evidence, and because it appears to have several features that separate it easily from S. asprella. SIDALCEA SPARSIFOLIA (C. L. Hitche.) S. R. Hill, comb. nov.—Sidalcea malviflora [‘malvaeflora | subsp. sparsifolia C. L. Hitche., University of Washington Publications in Biology 18:32. 1957.—Type: USA, California, Kern Co., 1 mile south of Ft. Tejon, 29 May 1952, C. L. Hitchcock 19546 (holotype: WTU; isotypes: UTC 88184! DS 368036 at CAS!). Andreasen and Baldwin (2003b) included this plant in their molecular studies of Sidalcea, and found that it grouped generally with the coastal subspecies of S. malviflora as well as with S. covillei Greene, S. pedata A. Gray, and S: neomexicana A. Gray. They stated: ““The position of S. malviflora subsp. sparsifolia basally to the clade (jk 76%) consisting of the other subspecies of S. malviflora plus S. pedata and S. neomex- icana, provides evidence for the paraphyly of S. malviflora and may justify treatment of S. malviflora subsp. sparsifolia as a separate spe- cies’. I agree with this, not only because of the molecular data but because of its more inland range and semi-desert habitats, as well as a series of morphological differences. It is rather similar to the other species in the clade particularly in the morphology of the fruits and of the pubescence of its various parts. The reduced stem leaves and shortened rhizomes may be adaptations to its transitional hot and dry desert environment, and it is the southernmost species of the genus in southern California and Baja California, along with its desert wetland-adapted relative S. neo- mexicana. Extreme variation remains problematic in this species despite its removal from S. malviflora. HILL: NOTES ON CALIFORNIA MALVACEAE 107 Hitchcock (1957) divided his Sidalcea malviflora subsp. sparsifolia further into four morphologi- cally defined varieties that also have some geographic coherence. These, upon further study, both morphological and molecular, perhaps could be defined as subspecies of this newly circumscribed species. I have not yet focused on this aspect in the current study, but some of the extremes are not only quite different in appear- ance, but they are also somewhat difficult to separate from S. malviflora subsp. californica— another mostly inland taxon that needs addition- al study. SIDALCEA DIPLOSCYPHA VS. SIDALCEA KECKII Sidalcea keckii Wiggins has been of special interest in California because it was once thought to have been extirpated (Hill 1993) then, upon being rediscovered at a later date, it was proposed and accepted for inclusion in the Federal Register as a Federally Endangered plant species (United States Fish and Wildlife Service 2000). It was thought to be restricted to the White River region of southern Tulare County. In the years since it was rediscovered, it has been sought out there and elsewhere. It was known since its original description to be very closely related and similar to Sidalcea diploscypha (Torr. & A.Gray) A.Gray in its annual habit, its leaf morphology, its flowers and fruits, and especially regarding its pubescence—as only these two annual sidalceas have numerous long fine perpendicular hairs along the stem. The molecular work of Andrea- sen and Baldwin (2001, 2003b) and especially Andreasen (2005) utilizing new molecular phylo- genetic data demonstrated convincingly that the two species are distinct. It has been proposed by some that S. keckii is quite recognizable because on the inside of the calyx there are five reddish spots thought not to be present in S. diploscypha. A re-investigation of the two species from both old and new herbarium specimens revealed that both species can have these red spots (sometimes reduced to narrow red lines) on the internal calyx surface. Specimens sorted out well using other characters, to the point that it appears that Sidalcea keckii is more wide-ranging than previ- ously thought. Several morphological features can be used to distinguish the two similar species. Regarding the hairs on the stem—one group of specimens (first identified as S. diploscypha) had, mixed with the characteristic long perpendicular hairs, a consid- erable number of glandular hairs and odd multicellular trichomes with green pigment (that resembled short algal filaments) on both the upper stems and on calyces and these were found also on specimens known to be S. keckii. This same group of specimens had one or a very few tiny bristles on the upper portion of the fruit 108 MADRONO where many sidalceas have a small cusp or mucro (sometimes called a ‘beak’) whereas S. diploscy- pha has no such bristles on its fruit. The upper leaves of S. keckii are not only lobed, typical of both species, but the lobe tips have three equal teeth on the widened lobe apex whereas the lobes of the upper leaves of S. diploscypha are usually narrow throughout, and either entire or the lateral teeth are positioned far below the central elongated tooth. The primary difference between the two species is the presence of long, multi- divided bracts and stipules at the base of the flowers and upper leaves, respectively, of S. diploscypha vs. the smaller undivided bracts and stipules in the same positions in S. keckii. Upon examining a large number of specimens, this feature did hold up well—but there were always a few that did not ‘quite fit-—and so a very few specimens with divided bracts were called S. keckii. As a result of these character observations, several specimens of S. diploscypha from Colusa, Fresno, Merced, Napa, Solano, and Yolo coun- ties were re-annotated as S. keckii. Most of these had some features of S. diploscypha, and it appeared that introgression might be playing a role. This geographic distribution suggested that specimens of S. keckii might also be found in Butte and Lake counties but no specimens examined from those counties had the definitive assemblage of characters of that species and so all were annotated as S. diploscypha. Sidalcea diploscypha appears to be a species that prefers serpentine, whereas S. keckii is not so restricted. An examination of habitat and sub- strate preferences in Napa Co. where both substrates and species have been found nearly side-by-side, showed that those on serpentine sorted out nicely to S. diploscypha and those in the adjacent sandstone-derived soils were S. keckii though individuals were often only a few meters distant from one another (B. Ertter, [UC/ JEPS], J. Ruygt, personal communication). More molecular work on these newly interpreted populations from Solano to Colusa counties may help to further unravel the relationship between these two taxa, but, as interpreted now, the Federally Endangered S. keckii, while still exceedingly uncommon, is now reported for seven counties, rather than just one or two, as previously thought. The number of populations currently extant is still unknown. The following key is offered to distinguish the two species: la. Upper paired stipules (at petiole bases) and bracts (at pedicel bases) each divided to base into 2 or more linear lobes nearly equal to or longer than calyx; length of central tooth of middle leaf lobe on upper stem leaves much longer than lateral teeth, so lobe has a single apical tooth, or lobes entire; inflorescence, calyx generally not densely glandular .... S. diploscypha [Vol. 56 1b. Upper paired stipules and bracts each simple, linear, undivided (a few divided in robust plants) generally shorter than calyx; length of 3 apical teeth of widened middle leaf lobe on upper stem leaves essentially equal; inflores- cence, calyx generally with many minute glandular, multicellular simple hairs... . S. keckii HIBISCUS LASIOCARPOS The California Hibiscus was treated by Hill (1993) as part of the widespread Hibiscus lasiocarpos Cav. in accordance with the opinions of Fryxell (1988) and others. The California populations (primarily in the Sacramento Valley) remain quite scarce and isolated from any other populations of this species, the closest of which are in northwestern Chihuahua, Mexico, and in Dona Ana Co., New Mexico. Its scarcity causes it to be of conservation concern in California (List 2: Plants Rare, Threatened, or Endangered in California, But More Common Elsewhere). I have reconsidered this restricted California plant, and I have decided to recognize it as Hibiscus lasiocarpos Cav. subsp. lasiocarpos var. occiden- talis (Torr.) A.Gray. Its nomenclatural history follows. HIBISCUS LASIOCARPOS Cav. var. OCCIDENTALIS (Torr.) A.Gray, Proceedings of the American Academy of Arts and Sciences 22:303. 1887 [4 Mar 1887] (as “‘/asiocarpus var. occidentalis’). —RHibiscus moscheutos L. var. occidentalis Torr., United States Exploring Expedition, Phanerogams. Pacific North America 17(2):256. 1874.—Type: USA, California, Sacramento Co., Sacramento Valley, s.d., Wilkes Expedition 1364 (holotype: NY). Hibiscus lasiocarpos var. californicus (Kellogg) L. H. Bailey, The Standard Cyclopedia of Horticulture 1486. 1915.—Afibiscus californicus Kellogg, Proceedings of the California Acad- emy of Sciences 4:292. 1873.—Type: USA, California, San Joaquin Co., on island near Middle River bridge, San Joaquin River (Byron-Stockton Hwy) Alexander & Kellogg 3526 (neotype: CAS; isoneotypes: UC, US)— designated by P. A. Fryxell, Systematic Botany Monographs 25:211 (1988). This variety continues to be included here ~ within Hibiscus lasiocarpos Cav. That species, | including the California plants, has recently been — treated by Blanchard (2008) as H. moscheutos L. | subsp. /asiocarpos (Cav.) O. J. Blanchard. At , least one flora (Gleason and Cronquist 1991) has equated Torrey’s California variety with all of © H. lasiocarpos, calling it H. moscheutos var. | occidentalis Torr. That concept is not accepted | here because Torrey’s type is from California and clearly represents only the isolated Cali- | fornia population here included, following Gray’s | 2009] example, within H. Jasiocarpos. While there is some variation in pubescence in these two species, treatments such as Godfrey and Wooton (1981) and Mohlenbrock (1986) use this feature in particular to distinguish these species. Other treatments of the Malvaceae (e.g., Hull 1982) and most world floras use some or many hair characters to distinguish mallow species from one another, as this family has hair types varying from simple to bilateral several-rayed hairs, multi-rayed stellate hairs, glandular hairs, as well as stellate-lepidote hairs in varying combinations and densities that usually remain surprisingly consistent within a taxon. I continue to accept Hibiscus moscheutos L. and Hibiscus lasiocarpos Cav. as distinct species based primarily on the following characters: la. Upper leaf surface glabrous or sparsely pubescent, darker than the densely felty- pubescent lower surface especially when dry; capsules glabrous or sparsely pubescent on sutures; involucral bracts uniformly, minutely CANECSCEN Eo ig Skene Bh Shek Hibiscus moscheutos L. lb. Upper leaf surface densely soft-pubescent like the lower surface, and usually similar in color; capsules densely pubescent through- out; involucral bracts densely coarsely pu- bescent with both short-stellate hairs and longer spreading simple hairs especially near the: Mares. 2G 24.424 Hibiscus lasiocarpos Cav. According to this species concept, the Califor- nia plants fit within H. Jasiocarpos. This California variety was distinguished by Bailey as having more uniformly cordate leaves and a less hairy capsule than the typical variety. Furthermore, individuals of this taxon charac- teristically produce long starchy rhizomes from which they often propagate themselves in their native habitat, marshes and deltaic areas subject to unreliable water levels (e.g., Sacramento, 28 Sep 1989, C. M. Richard 098928 [OAKL: 4 sheets]). This variety often exceeds 2 m in height, and the capsules are globose and 2.5—3 cm, whereas the typical variety normally grows to 2m or less (not infrequently <1 m tall) and has subglobose or short-cylindic capsules 2—2.5 cm long. LAVATERA VS. MALVA The genera Lavatera L. and Malva L. have undergone a significant revision since the publi- cation of TJM1. Studies by Ray (1994, 1995) on nuclear rDNA Internal Transcribed Spacer (ITS) sequence data as well as morphological features led to the conclusion that the species of Malva and Lavatera are all closely related, and that a significant number of species assigned to both genera were more closely related to one another than previously thought. In particular, several species in Lavatera were found to be more closely HILL: NOTES ON CALIFORNIA MALVACEAE 109 related to Malva sylvestris L., the type species of Malva, than they were to Lavatera trimestris L., the type species of Lavatera, by means of both sets of data. Both genera are still accepted, but circumscriptions have changed, and the closely related taxa could no longer be maintained within two separate genera. Therefore, the realignment of species within Lavatera had to be formalized. Nomenclature for the species of Malva included in TJM1 remain the same. Ray (1998) chose to maintain both genera with the types as stated above, and he defined them not only by their ITS characters, but also by a series of mericarp features. Ma/va was distin- guished from Lavatera primarily by its mericarps that 1) are rounded in only the axial direction on the abaxial side, 2) have lateral angles or edges, 3) completely or nearly completely enclose the seed, 4) do not separate readily from the seed, and 5) act as a dispersal unit. This group contains not only cosmopolitan weedy species formerly in- cluded in both genera, but also several unusual disjunct taxa in Australia, Baja California, Mexico, and California, USA, that had formerly been treated within the genera Lavatera or Saviniona Webb & Berthelot. Ray (1998) proposed new combinations and new names for several of the taxa formerly placed within Lavatera. Consequently, all three Lavatera species in TJM1 are now considered to belong in Malva. Ray (1998) proposed the name Malva dendromorpha M. F. Ray as a substitute for Lavatera arborea L., as he thought that Malva arborea was a name already taken and unavail- able. However, the name that he cited, ““Malva arborea St.-Hil.”, was never published (a Sphalma typographicum (misprint) in Index Kewensis), and so the next available name, Malva arborea (L.) Webb. & Berthelot (1836. Histoire Naturelle des Iles Canaries, pt. 2. Phytographia Canariensis 1:30.) based on the Linnaean name is the correct name in Malva and Malva dendromorpha M. F. Ray becomes superfluous. Ray (1998) also proposed the name Malva linnaei M. F. Ray to replace the name Lavatera cretica L., as Malva cretica Cav. had already been used for a different plant. However, it was brought to my attention (Hinsley 2009) that the name Malva pseudolava- tera Webb & Berthelot (1836. Histoire Naturelle des Iles Canaries, pt. 2. Phytographia Canariensis 1:29.) had been proposed as a substitute name for Lavatera cretica long before Ray’s substitute name, and can be considered to be the correct name for the plant, making Ray’s name super- fluous. For the third California Lavatera, Ray (1998) proposed the new combination Malva assurgentiflora (Kellogg) M. F. Ray for the indigenous species formerly called Lavatera assurgentiflora Kellogg, and this is now its name in Malva. Further study may indicate that there are two distinct subspecies within this coastal 110 California species as suggested by Philbrick (1980). Lavatera itself as currently defined (Ray 1998) is only rarely found as an introduced plant in North America and in California in particular. Both L. olbia L. and L. trimestris L. have been found as waifs in the state (see below). A third species, Lavatera thuringiaca L., has been found rarely as an escape in more northern parts of North America. ADDITIONAL MALVACEAE IN THE FLORA Tliamna_ rivularis (Dougl.) Greene was not included in the flora of California in 1993 (Hickman 1993). Two specimens collected by Joseph P. Tracy on August 14, 1939, in Humboldt Co. recently came to light. They had been overlooked for many years and there are no other known records of this native species in California, though it is much more common north of the state. The vouchers for this species are: CALIFORNIA. Humboldt Co.: Willow Creek Canyon, along Trinity Highway, in woods near stream, altitude 2500 feet, 14 Aug 1939, J. P. Tracy 16104 (MO 1191877!, MO 1191878!). It is doubtful that it still exists in the state, but there is always the chance it persists. It should also be sought in Lassen or Modoc counties where there is suitable habitat. [/iamna_ latibracteata Wiggins is well-known from redwood forest regions in Humboldt County, and it differs from J. rivularis by its wider bractlets (ca. 1 cm wide and long vs. 2 mm wide X 4.5—6 mm long in I. rivularis) and its dense pubescence on the undersides of the leaves (hairs sparse in J. rivularis). Iliamna bakeri (Jeps.) Wiggins is found in more inland chaparral sites in northern California, and it has more shallowly lobed leaves and shorter (2-5 cm vs. >5 cm), stouter petioles than the other two. Several species of Malvaceae have become naturalized or have been found as waifs in recent years and are added to the flora. ANODA PENTASCHISTA A. Gray. CALIFORNIA: Imperial Co.: Collins and Flood, Bard, weed in citrus, two trees involved in 40 acre grove, 9 Sep 1983, L. Pineda & R.A.Flock s.n. (CDA 4902, CDA 4903, RSA 327698). LAGUNARIA PATERSONIA (Andr.) G. Don. CAL- IFORNIA. San Diego Co.: Camp Pendleton, south of Santa Margarita River, 200 ft west of Stuart Mesa Road, and north of old sewage treatment ponds, elevation 3 m, 21 Jul 2007, C. Martius 401 (SD 179434!). LAVATERA OLBIA L. CALIFORNIA. Orange Co.: Laguna Canyon, 17 Jun 1994, O.F. Clarke s.n. (UCR 120561). San Francisco Co.: shrubs to 8 ft. tall, commonly naturalized on non- irrigated waste ground of formerly cultivated MADRONO [Vol. 56 garden, Victor Reiter’s garden, 1195 Stanyon St., San Francisco, 4 Aug 1970, T7.C. Fuller s.n. (CDA 5008). LAVATERA TRIMESTRIS L. CALIFORNIA. Santa Barbara Co.: edge of water, Lauro Canyon Reservoir near San Roque Rd., Santa Barbara, 25 Jun 1975, C. F. Smith 10902 (CDA 5007; RSA 535067). Cited in Smith (1976). ACKNOWLEDGMENTS I thank Katarina Andreasen and Bruce Baldwin for sharing their data on these plants and for stimulating correspondence concerning these taxa in recent years. Others have also been generous with their time and experience with these plants, including Barbara Ertter, Richard Halse, Stewart Robert Hinsley, Fred Hrusa, Lawrence Janeway, Kanchi Gandhi, Jake Ruygt, and Debra Trock, to mention a few. I would especially like to thank Bruce G. Baldwin and the Lawrence R. Heckard Endowment Fund of the Jepson Herbarium for the funding to visit the UC/JEPS and CAS herbaria to study herbarium specimens in support of my work on several Malvaceae genera for the upcoming revision of the Jepson Manual: Higher Plants of California. The UC/JEPS and CAS staff were especially hospitable during my visit and very generous with their time. A continuing grant from the Illinois Department of Transportation is gratefully acknowledged for financial support towards research and publication of these studies LITERATURE CITED ANDREASEN, K. 2005. Implications of molecular systematic analyses on the conservation of rare and threatened taxa: contrasting examples from Malvaceae. Conservation Genetics 6:399-412. AND B. G. BALDWIN. 2001. Unequal evolution- ary rates between annual and perennial lineages of checker mallows (Sidalcea, Malvaceae): evidence from 18S-26S rDNA internal and external tran- scribed spacers. Molecular Biology and Evolution 18:936—-944. AND . 2003a. Nuclear ribosomal DNA sequence polymorphism and hybridization in checker mallows (Sidalcea, Malvaceae). Molecular Phylogenetics and Evolution 29:563-581. AND . 2003b. Reexamination of relation- checker mallows (Sidalcea, Malvaceae) based on molecular phylogenetic data. American Journal of | Botany 90:436—-444. BLANCHARD, O. J. 2008. Innovations in Hibiscus and Kosteletzkya (Malvaceae, Hibisceae). Novon 18:4-8. DIMLING, J. 1991. Comments on Sidalcea (Malvaceae) ships, habital evolution, and phylogeography of of the Klamath Mountains of Oregon and Califor- _ nia. Madrono 38:249—267. FRYXELL, P. A. 1988. Malvaceae of Mexico. Systematic Botany Monographs Vol. 25. The American Society of Plant Taxonomists. University of Michigan Herbarium, Ann Arbor, MI. GLEASON, H. A. AND A. CRONQUIST. 1991. Manual of | the vascular plants of northeastern United States and adjacent Canada, 2nd ed. The New York Botanical Garden Press, Bronx, NY. ; : 2009] GODFREY, R. K. AND J. W. WOOTEN. 1981. Aquatic and wetland plants of southeastern United States. Dicotyledons. The University of Georgia Press, Athens, GA. HICKMAN, J. C. (ed.). 1993. The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. HILL, S. R. 1982. A monograph of the genus Malvastrum (Malvaceae: Malveae). Rhodora 84:1-83, 159-264, 317-409. . 1993. Sidalcea. Pp. 755-760 in J. C. Hickman (ed.), The Jepson manual: higher plants of Califor- nia. University of California Press, Berkeley, CA. . 2008. Two new subspecies of Sidalcea hickmanii (Malvaceae) in California. Journal of the Botanical Research Institute of Texas 2:783—791. HINSLEY, S. R. 2009. Malvaceae info. http://www. malvaceae.info/index.html. [accessed 2009, July 23]. HitcHcock, C. L. 1957. A study of the perennial species of Sidalcea. Part I. Taxonomy. University of Washington Publications in Biology 18:1—79. MOHLENBROCK, R. H. 1986. Guide to the vascular flora of Illinois, revised edition. Southern Illinois University Press, Carbondale, IL. Munz, P. A. 1968. A California flora: supplement. University of Californai Press, Berkeley, CA. HILL: NOTES ON CALIFORNIA MALVACEAE 11] AND D. D. Keck. 1959. A California flora. University of California Press, Berkeley, CA. PHILBRICK, R. 1980. Distribution and evolution of endemic plants of the California islands. Pp. 173— 187 in D. M. Power (ed.), The California Islands: proceedings of a multidisciplinary symposium. Santa Barbara Botanic Garden, Santa Barbara, CA. RAY, M. F. 1994. A contribution to the systematics of Lavatera and Malva (Malvaceae) and related genera. Ph.D. dissertation. University of Califor- nia, Berkeley, CA. . 1995. Systematics of Lavatera and Malva—a new perspective. Plant Systematics and Evolution 198:29_-53. . 1998. New combinations in Malva (Malvaceae: Malveae). Novon 8:288—295. RousH, E. M. F. 1931. A monograph of the genus Sidalcea. Annals of the Missouri Botanical Garden 18:117-244. SMITH, C. F. 1976. A flora of the Santa Barbara region, California. Santa Barbara Museum of Natural History, Santa Barbara, CA. UNITED STATES FISH AND WILDLIFE SERVICE. 2000. Endangered status for Sidalcea keckii. Federal Register 65:7758—7759. MADRONO, Vol. 56, No. 2, pp. 112-117, 2009 HISTORICAL, NOMENCLATURAL, AND DISTRIBUTIONAL NOTES ON TWO PACIFIC COAST KELPS: LESSONIOPSIS LITTORALIS AND PLEUROPHYCUS GARDNERIT (PHAEOPHYCEAE, LAMINARIALES, ALARIACEAE) PAUL C. SILVA University Herbarium, University of California, Berkeley, CA 94705-2465 psilva@berkeley.edu ABSTRACT The names of two eastern North Pacific kelps (Phaeophyceae, Laminariales, Alariaceae) were inadvertently validated by Josephine Tilden on labels to specimens distributed in the exsiccata ‘“American Algae’’. Lectotypes for Lessonia littoralis and Pleurophycus gardneri, which concomitantly are the types of the names of the monospecific genera Lessoniopsis and Pleurophycus, respectively, are herein designated from among specimens housed in the Herbarium of the University of Minnesota, Tilden’s home institution. Lessoniopsis ranges from Kodiak Island, Alaska, to the Big Sur coast of Monterey County, California, whereas Pleurophycus ranges from southeastern Alaska to Pt. Piedras Blancas, San Luis Obispo County, California. Molecular data suggest that both Lessoniopsis and Pleurophycus belong to the Alariaceae, a departure from their previous familial placements in the Lessoniaceae and Laminariaceae, respectively. Key Words: W. G. Farlow, Laminariales, Lessonia, Lessoniopsis, Pleurophycus, DeAlton Saunders, W. A. Setchell, Josephine Tilden. LESSONIOPSIS LITTORALIS As recounted by Setchell and Gardner (1903, p. 267), Lessoniopsis littoralis Farlow et Setchell ex Tilden was first brought to the attention of science by Elihu Hall of Oregon, who sent a specimen to William Gilson Farlow at Harvard. Farlow (1875, p. 355) assigned it tentatively to Lessonia fuscescens Bory (syntype localities: Concepcion, Chile, and the Falkland Islands). Later, Farlow (1876, p. 707) assigned the Hall specimen to L. nigrescens Bory (type locality: Cape Horn). After having seen a living specimen at Monterey in 1885, Farlow decided that Hall’s specimen represented an undescribed species of Lessonia for which he proposed the manuscript name L. /ittoralis. This name can be found on labels to specimens in the Farlow Herbarium (FH) or sent by Farlow to other herbaria. The species was still undescribed when William Albert Setchell, who had been appointed Professor of Botany at the University of California in 1895, first saw it in the field — at Cypress Point, Monterey Peninsula, in July 1896. In his note- book, under no. /4/7, Setchell wrote ““Lessonia litoralis [using an alternative spelling of the epithet] Farlow mscr.”’, but on the labels to his specimens he wrote ““Lessonia litoralis F. & S.”. The use of “F. & S.” for the authorship appears to follow Josephine Tilden’s exsiccata, “‘Ameri- can Algae” (fourth century, dated February 20, 1900), in which a description of “‘Lessonia littoralis Farlow and Setchell. Mss.” is given on the label to no. 342. The basis for this accredi- tation is unclear as there is no evidence that Farlow or Setchell had provided Tilden with pertinent information. In fact, Tilden had previ- ously severed relations with Setchell (letter to Setchell dated November 21, 1898) because of the ‘unpleasant spirit” and “‘ungentlemanly tone” of his criticism of the first fascicle of her exsiccata. On the label to no. 342, Tilden gave credit to DeAlton Saunders for the determination. Saun- ders had recently returned from serving as botanist on the Harriman Alaska Expedition, which gave him the opportunity to observe and collect this kelp. An explanation of how Tilden came to describe the species can be found in a letter from Saunders to Setchell, dated March 1, 1900: ‘“‘T was somewhat surprised on receiving Miss Tilden [’s] set [American Algae Century IV] that I had determined so many species. While I was working Christmas at Minneapolis she often showed me things that bothered her & I suggested that they look like so & so—That evidently constituted a determination in her mind. Thinking to save her a great deal of unnecessary blundering & perhaps the chagrin of redescribing your Lessonia I told her my understanding of it & suggested her writing you... I believe her work for pure superficiality cannot be equalled.” Farlow was scarcely less critical. In a letter to Setchell dated March 12, 1896, with regard to the first fascicle of ““American Algae’, Farlow wrote: ‘‘Miss Tilden [’s] specimens are worse than any I ever saw except Kitzing’s... A good part of 2009] FIG. 1. SILVA: PACIFIC COAST KELPS HERBARIUM OF THE UNIVERSITY OF MINNESOTA. hessonia littoralis Farlow Am. Alg. 342. a i eal ieee are SS ee Lectotype of Lessonia littoralis Farlow et Setchell ex Tilden (MIN 3607). 3607 Pacey Sh TROUT RE RS | 114 SECRETS HES wee SEBS TAR \ aa a Ses FIGs 2; MADRONO copyright reserved JOSEPHINE E. TILDEN. American Algz, 340. Pleurophycus gardneri SercHELL and Scan Mss. Frond 1-2 meters in length, rising from a branched holdfast, forming a rather flattened elongated stipe which continues as a wide midrib area through the lamina; lamina wrinkled and fluted on either side of midrib portion; sporangia forming elongated patches on upper portion of mid- rib, Attached to white fir log. Washed up on beach, North bay, San Juan island, Washington. J. E. T. 28 My 1898. Det. De Alton Saunders Lectotype of Pleurophycus gardneri Setchell et Saunders ex Tilden (MIN 3613). EEE Ee 2009] the other numbers which I have examined are incorrectly named. I felt like writing a savage notice of the century but I concluded that it was better to say nothing. Whatever criticism emanates from Harvard is believed in a good part of the country to be merely ill natured jealousy of progressive Western botanists and any criticism, however just, excites sympathy for the person criticized and has no weight as criticism.” Whatever her shortcomings as a taxonomist, Tilden made significant contributions to phycol- ogy. She was a pioneer in the study of thermal algae in the United States. She founded the University of Minnesota Seaside Station at Port Renfrew on the west coast of Vancouver Island. She compiled and published a card index to phycological literature, “Index Algarum Univer- salis’’, popularly called the ““Tilden Index’’, which remains useful despite being largely unknown or ignored by present-day taxonomists. She also published a somewhat idiosyncratic but thought- provoking text book, The algae and their life relations (Tilden 1935), a pioneer work in which classification was addressed with reference to pigmentation, food reserves, and evolutionary environmental changes. Because no other collections were cited on the label, the specimens distributed as no. 342 in Tilden’s ““American Algae” constitute the type collection of Lessonia littoralis. This collection was made by Tilden on August 3, 1898, at Baird Point on the Strait of Juan de Fuca, Vancouver Island, British Columbia, Canada. Three speci- mens of this distribution are in the Herbarium of the University of Minnesota (MIN) while two are in the Herbarium of the University of California (UC), all so fragmentary as to validate Farlow’s criticism of the quality of Tilden’s collections. One specimen (MIN 3607), which lacks the exsiccata label but is otherwise labeled “‘Lessonia littoralis Farlow. Am. Alg. 342.”, is somewhat more informative than the other four specimens seen in this study and is herein designated lectotype (Fig. 1). All other distributions of no. 342 are isotypes (isolectotypes). The name Lessonia littor- alis must be accredited either to Tilden (when using an abbreviated format) or to Farlow et Setchell ex Tilden (when using an extended format), but not to Farlow et Setchell because Tilden wrote the diagnosis. The type locality has been incorrectly indicated as Cypress Point, Monterey Peninsula, by Smith (1944, p. 145) and Nicholson (in Abbott and Hollenberg 1976, p. 246). Lessonia littoralis was made the type of a new genus, Lessoniopsis by Reinke (1903), thus restricting Lessonia to the southern hemisphere. Lessoniopsis littoralis (Tilden) Reinke is a cuma- phyte, growing only in the most exposed rocky sites, usually just below Postelsia palmaeformis SILVA: PACIFIC COAST KELPS KS Rupr. in the lowermost intertidal and upper subtidal zones. It ranges from Kodiak Island, Alaska (Druehl 1970) southward to Kasler Point, Monterey County, California (Si/va 802 in UC). PLEUROPHYCUS GARDNERI Setchell and Saunders, independently from one another, recognized Pleurophycus gardneri Setch- ell et Saunders ex Tilden as an undescribed species representative of an undescribed genus of kelps. The first specimens available to Setchell were those collected in 1898 by N. L. Gardner, a high school teacher in Oak Harbor on the west coast of Whidbey Island, Washington. The alga that Setchell immediately recognized as unde- scribed is similar to Laminaria, with a single blade borne on a sturdy stipe, but unlike the latter genus, its blade has a midrib. For this alga Setchell proposed the manuscript name Pleur- ophycus gardneri. On June 26, 1899, Saunders collected the same kelp at Yakutat Bay, Alaska. In a letter to Setchell dated September 1, 1899, Saunders wrote: “It is a good thing for both of us that I stayed a week in Pudget Sound & collected with Mr Gardner. Otherwise I should have done you an injustice—TI should have had the description of your Pleurophycus published under a different name.” Setchell then proposed joint authorship and Saunders hesitantly accepted the proposal. In subsequent correspondence, the place of publica- tion was discussed, Saunders favoring the forth- coming results of the Harriman Expedition while Setchell favored a note intended for the journal Zoe. While Setchell and Saunders were dithering, Tilden published a valid description of this new genus and species under the name “ Pleurophycus gardneri Setchell and Saunders Mss.” (label to no. 346, American Algae, fourth fascicle, Febru- ary 20, 1900). As with Lessonia littoralis, Tilden came to publish the first description of this kelp with the inadvertent assistance of Saunders during his Minneapolis visit at Christmas time in 1899. Setchell (1901) published a more complete description in the journal Zoe, which appeared early in 1901 (between February 15, when the editor, Mrs. Brandegee, sent proof to Setchell, and March 26, when Mrs. Weeks acknowledged receipt of a reprint from Setchell). The account of the algae of the Harriman Alaska Expedition (Saunders 1901) was published November 15, 1901. Plate LII in that account is a drawing of ‘an almost perfect specimen ... collected by Miss J. E. Tilden in Puget Sound’’. This specimen has not been located. The specimens distributed by Tilden under no. 346 in both sets of ““American Algae’? housed 116 at UC and the set at MIN are predictably fragmentary. The specimen at Minnesota (MIN 3613) comprises a horizontal swath of a blade, a small piece of a stipe, and the base of a blade minus the holdfast. The specimens were collected by Tilden on May 28, 1898, at North Bay, San Juan Island, Washington, and determined by Saunders. Because no other collections are cited on the label, these specimens must be considered as constituting the type collection. No specimen in the Herbarium of the University of Minnesota was indicated by Tilden as the holotype. There- fore, no. 346 of ““American Algae’’ (MIN 3613) is herein designated lectotype of Pleurophycus gardneri (Fig. 2). All other distributions of this number are isotypes (isolectotypes). The generic and specific names must be accredited either to Tilden (when using an abbreviated format) or to Setchell et Saunders ex Tilden (when using an extended format), but not to Setchell et Saunders (as has always been done in the past), because Tilden wrote the diagnosis. The type locality is San Juan Island and not Whidbey Island (as has been erroneously assumed by previous authors). At the time when the known southern limit of Pleurophycus gardneri was Coos Bay, Oregon, Setchell and Gardner (1925, p. 607) speculated that “It possibly extends many miles further south, even to the coast of California.’ The first collection from California seems to be a plant obtained in the drift in Mendocino Bay on August 21, 1965, by E. K. Daniels (UC 1318218). Kjeldsen (1972) reported the finding of a large population on rocks at extreme low tide level at Fort Bragg, Mendocino County, and an isolated specimen from the drift at Salt Point, Sonoma County. VanBlaricom et al. (1986) studied subtidal populations of Pleurophycus off Pt. Sierra Nevada and Pt. Piedras Blancas in San Luis Obispo County, thus greatly extending the range southward. They also retrieved two thalli of Pleurophycus from the anchor upon departure from a site near Pt. Sur. Pleurophycus was later found growing on an unnamed bank near Pt. Sur (subsequently named Schmieder Bank) by a Cordell Expedition in 1989 (UC 1575256, 1575257) and was reported as forming a zone at depths of 30-45m all along the coast of Monterey County from Carmel to Pt. Sur (Spalding et al. 2003). Although the progression of records suggests a southward movement of this kelp, there is no reason to believe that the distribution in Califor- nia has changed in recorded history. CLASSIFICATION OF LESSONIOPSIS AND PLEUROPHYCUS The numerous genera of kelps are held together in a single order, Laminariales, distin- guished by a unique structural plan and a nearly MADRONO [Vol. 56 unique life history that involves separate game- tophytic and sporophytic generations. Gameto- phytes are very small filamentous thalli_ that undergo oogamous sexual reproduction. The thallus of the sporophyte is parenchymatous and comprises a holdfast, stipe, and blade, the latter two structures joined by a transitional meristematic zone. Sporangial sori are borne either on a central blade or on special lateral sporophylls. Whereas gametophytes show rela- tively little variation, there is an astonishing number of variations of the sporophyte, leading to the recognition of a large number of mono- typic genera. In a long-standing but simplistic classification, longitudinal splitting of the meri- stem and the production of lateral sporophylls by the stipe immediately below the transition zone were cardinal points in the recognition of three families of Pacific coast kelps: Lamiunariaceae, without splitting or lateral sporophylls; Alaria- ceae, without splitting but with lateral sporo- phylls; and Lessoniaceae, with splitting but without lateral sporophylls. Lessoniopsis has a complex morphology that makes classification difficult. The meristem splits repeatedly to produce a very large number of narrow, ribbed, sterile blades, each blade subtended by 1-3 pairs of ribless sporophylls. This straddling of two families (Lessoniaceae and Alariaceae) was ac- commodated by Setchell and Gardner (1925) by placing the genus in its own tribe within the Lesssoniaceae, deciding on this family rather than the Alariaceae because the thallus of Lessoniopsis more closely resembles that of Lessonia than that of Alaria. Prior to the recognition of the Alariaceae and Lessoniaceae by Setchell and Gardner (1925), the Lessoniopseae had been established by Setchell (1912) as a tribe of Laminariaceae. Pleurophycus, by contrast, has a simple form that apparently is not very different from Laminaria. Germann (1986), however, has shown that at the end of the growing season the blade abscisses, in contrast to other members of the Laminariaceae in which the old blade disappears progressively in response to senescence and abrasion by wave action in winter storms. After the blade has abscissed, the abscission layer is protected by an outgrowth of cortical cells, forming a scar that remains at the distal end of newly formed blades. Another unusual feature is the formation of sporangial sori on the midrib as well as on the wings of the blade. Despite the distinct morphology of such genera as Postelsia, Macrocystis, and Nereocystis, the Laminariales show relatively little molecular diversity. Nucleotide sequence comparisons of various parts of the nuclear, chloroplast, and mitochondrial genomes of 42 species reveal unexpected groupings of genera within the Alariaceae-Laminariaceae-Lessoniaceae complex 2009] (Lane et al. 2006). In the new alignment, both Lessoniopsis and Pleurophycus are included in the Alariaceae along with the traditionally placed members Alaria and Pterygophora. The presence of lateral sporophylls in Lessoniopsis is a character in agreement with A/aria and Pterygo- phora, but it is difficult to imagine the evolution- ary transformation of Pleurophycus from an alarioid ancestor to its present-day laminarioid form. The molecular-based Laminariaceae has an even more surprising circumscription. In addition to genera traditionally assigned to this family, it includes Nereocystis, Macrocystis, Pelagophycus, and Postelsia. The inclusion in the Laminariaceae of these four genera, which have traditionally been assigned to the Lessoniaceae, greatly nar- rows the breadth of the latter family, but molecular data supports the transfer of Egregia and Eisenia from their traditional placement in the Alariaceae to the Lessoniaceae, where they join the southern hemisphere genera Lessonia, Ecklonia, and Eckloniopsis. On the basis of molecular data, Lane et al. (2006) established a fourth family, the Costariaceae, comprising Agarum, Costaria, Dictyoneurum, and Thalassio- phyllum. These familial realignments imply that longitudinal splitting in the transition region and the production of sporophylls are characters that have arisen more than once, thus increasing the difficulty of writing diagnoses at the family level. ACKNOWLEDGMENTS I thank Anita F. Cholewa, Curator of the Herbarium of the University of Minnesota, for arranging a loan of specimens. I am also indebted to my colleague Kathy Ann Miller, whose help in preparing the manuscript was indispensable. I appreciate the helpful suggestions offered by Tim Lowrey (Corresponding Editor) and two reviewers. LITERATURE CITED ABBOTT, I. A. AND G. J. HOLLENBERG. 1976. Marine algae of California. Stanford University Press, Stanford, CA. DRUEHL, L. D. 1970. The pattern of Laminariales distribution in the northeast Pacific. Phycologia 9:237-247. SILVA: PACIFIC COAST KELPS ib ie FARLOw, W. G. 1875. List of the marine algae of the United States, with notes of new and imperfectly known species. Proceedings of the American Academy of Arts and Sciences 10:351—380. . 1876. List of the marine algae of the United States. Report of the U.S. Commissioner of Fisheries 1873—1875:691—718. GERMANN, I. 1986. Growth phenology of Pleurophycus gardneri (Phaeophyceae, Laminariales), a decidu- ous kelp of the northeast Pacific. Canadian Journal of Botany 64:2538—2547. KJELDSEN, C. K. 1972. Pleurophycus gardneri Setchell & Saunders, a new alga for northern California. Madrono 21:416. LANE, C. E., C. MAYES, L. D. DRUEHL, AND G. W. SAUNDERS. 2006. A multi-gene molecular investi- gation of the kelp (Laminariales, Phaeophyceae) supports substantial taxonomic re-organization. Journal of Phycology 42:493—512. REINKE, J. 1903. Studien zur vergleichenden Entwick- lungsgeschichte der Laminariaceen. Schmidt & Klaunig, Kiel. SAUNDERS, D. 1901. Papers from the Harriman Alaska Expedition. XXV. The algae. Proceedings of the Washington Academy of Sciences 3:391—486. SETCHELL, W. A. 1901. Notes on algae. I. 5:121-129. 1912. The kelps of the United States and Alaska. Pp. 130-178 in F. K. Cameron, A preliminary report on the fertilizer resources of the United States. Washington DC. AND N. L. GARDNER. 1903. Algae of north- western America. University of California Publica- tions in Botany 1:165—-418. AND . 1925. The marine algae of the Pacific coast of North America. Part HI. Melano- phyceae. University of California Publications in Botany 8:383-898. SMITH, G. M. 1944. Marine algae of the Monterey Peninsula, California. Stanford University Press, Stanford, CA. SPALDING, H., M.S. FOSTER, AND J. N. HEINE. 2003. Composition, distribution, and abundance of deep- water (>30 m) macroalgae in central California. Journal of Phycology 39:273-284. TILDEN, J. E. 1935. The algae and their life relations. University of Minnesota Press, Minneapolis, MN. VANBLARICOM, G. R., D. C. REED, C. HARROLD, AND J. L. BopIN. 1986. A sublittoral population of Pleurophycus gardneri Setchell and Saunders 1900 (Phaeophyceae: Laminariaceae) in central Califor- nia. Bulletin of the Southern California Academy of Sciences 85:120—122. Zoe MADRONO, Vol. 56, No. 2, pp. 118-126, 2009 ARCEUTHOBIUM ABIETINUM SUBSPECIES WIENSHIT, A NEW SUBSPECIES OF FIR DWARF MISTLETOE (VISCACEAE) FROM NORTHERN CALIFORNIA AND SOUTHERN OREGON ROBERT L. MATHIASEN School of Forestry, Northern Arizona University, Flagstaff, AZ 86011 USA Robert.Mathiasen@nau.edu CAROLYN M. DAUGHERTY Department of Geography, Planning, and Recreation, Northern Arizona University, Flagstaff, AZ 86011 USA ABSTRACT We describe Arceuthobium abietinum subspecies wiensii (Wiens’ dwarf mistletoe, Viscaceae), a dwarf mistletoe that severely parasitizes red fir and Brewer spruce in northwestern California and southwestern Oregon. This classification is based on morphological and host range differences between white fir dwarf mistletoe (Arceuthobium abietinum f. sp. concoloris), red fir dwarf mistletoe (Arceuthobium abietinum f. sp. magnificae), and Wien’s dwarf mistletoe. Male and female plants of Wiens’ dwarf mistletoe were consistently smaller than those of both white fir and red fir dwarf mistletoes, whose plants were about the same size. The shoot color of Wiens’ dwarf mistletoe was frequently green-brown or red-brown, while the shoot color of white fir and red fir dwarf mistletoes was typically yellow-green or yellow. Differences in host specificity also distinguish Wiens’ dwarf mistletoe from white and red fir dwarf mistletoes. RESUMEN Se describe Arceuthobium abietinum subespecie wiensii (muérdago enano de Wiens, Viscaceae), un muérdago enano que parasita severamente a Abies magnifica y a Picea breweriana en el noroeste de California y suroeste de Oregon. Esta clasificacion se basa en diferencias morfoldgicas y de hospederos entre el muérdago enano del oyamel blanco (Arceuthobium abietinum f. sp. concoloris), el del oyamel rojo (Arceuthobium abietinum f. sp. magnificae), y el de Wiens. Las plantas masculinas y femeninas del muérdago enano de Wiens son consistentemente mas pequenhas que las del muérdago enano del oyamel blanco y las del oyamel rojo, las cuales son de tamano similar entre si. El color de las ramas del muérdago enano de Wiens fue frecuentemente café verdoso a café rojizo, mientras que el de los mueérdagos del oyamel blanco y del oyamel rojo tipicamente fue verde amarillo o amarillo. Adicionalmente, el muérdago enano de Wiens se distingue de los otros dos por diferencias en la especificidad de sus hospederos. Key Words: Abies concolor, Abies magnifica, Arceuthobium, dwarf mistletoe, Picea breweriana. Fir dwarf mistletoe (Arceuthobium abietinum Engelm. ex Munz) is a common parasite of white fir (Abies concolor (Gordon & Glend.) Hilde- brand; including var. /owiana (Gordon) A. Murray), and red fir (Abies magnifica A. Murray; including var. shastensis Lemmon) in the Sierra Nevada Mountains, Cascade Range, and the Klamath and Siskiyou Mountains of northwest- ern California and southwestern Oregon (Hawks- worth and Wiens 1972, 1993, 1996). Fir dwarf mistletoe populations that parasitize white fir in the Sierra Nevada Mountains do not infect red fir, while the fir dwarf mistletoe populations infecting red fir do not infect white fir, yet they are all morphologically similar. Recognizing this distinction, Hawksworth and Wiens (1972, 1996) named two special forms of A. abietinum: A. abietinum Engelm. ex Munz formae speciales concoloris Hawksworth & Wiens, which parasit- izes white fir, and Arceuthobium abietinum Engelm. ex Munz f. sp. magnificae Hawksworth & Wiens, which parasitizes red fir. Parmeter and Scharpf (1963) were the first to report the extreme host specificity of fir dwarf mistletoe populations based on cross inoculation studies and their field observations in mixed red and white fir stands in the Sierra Nevada Mountains. Although Hawksworth and Wiens (1972, 1996) did not find morphological or phenological differences between the fir dwarf mistletoe populations parasitizing white and red fir, their field observations in the Sierra Nevada Mountains confirmed the host specificity report- ed by Parmeter and Scharpf. Due to the economic impact that A. abietinum has on true firs in California, Hawksworth and Wiens (1972, 1996) argued that since the host affinities of the two fir dwarf mistletoes were so distinct, they deserved taxonomic recognition and designated them as formae speciales in accordance with \ 2009] recommendation 4B of the International Code of Botanical Nomenclature. Although Nickrent et al. (2004) suggested a taxonomic classification that treats A. abietinum under A. campylopodum Engelm. based primarily on molecular data, we reject this alternative classification and adopt the taxonomic treatment for Arceuthobium of Hawksworth and Wiens (1972, 1996) which distinguishes A. abietinum from A. campylopodum using morphological and host range differences between these species. We also agree with Hawksworth and Wiens’ (1996) classification of the dwarf mistletoe populations that parasitize Pacific silver fir (Abies amabilis Douglas ex J. Forbes) and noble fir (A. procera Rehder) in Oregon and Washington as hemlock dwarf mistletoe (Arceuthobium tsugense (Ro- send.) G. N. Jones) although we have determined that two subspecies of hemlock dwarf mistletoe parasitize these true firs to different degrees (Mathiasen and Daugherty 2005, 2007). The subspecies of hemlock dwarf mistletoe that occur on Pacific silver and noble firs can be distin- guished from fir dwarf mistletoe by their differences in shoot color and their inability to parasitize red fir (Hawksworth and Wiens 1996; Mathiasen and Daugherty 2005, 2007). Further- more, white fir dwarf mistletoe does not infect western hemlock (Tsuga heterophylla (Rafael) Sargent) although this hemlock is frequently associated with infected white firs in Oregon and grand firs (Abies grandis (Douglas ex D. Don) Lindley) in Washington (Hawksworth and Wiens 1996). In 1996 we were shown a population of dwarf mistletoe severely parasitizing red fir and Brewer spruce (Picea breweriana Watson) by Dr. Gregg DeNitto, USDA Forest Service, which occurred near Baldy Mountain northwest of Happy Camp, California (Fig. 1, site 11). Although the dwarf mistletoe plants there resembled fir dwarf mistle- toe, plants on both hosts were green-brown or red-brown instead of the typical yellow-green to yellow of fir dwarf mistletoe in the Sierra Nevada Mountains (Hawksworth and Wiens 1996). Furthermore, the dwarf mistletoe occasionally infected white fir indicating it was not Arceutho- bium abietinum f. sp. magnificae (sensu stricto). Our examination of another population of dwarf mistletoe parasitizing white fir and Brewer spruce near Flat Top Mountain west of Grants Pass, Oregon (Fig. 1, site 6) also showed the plants at this location were green-brown to red-brown, but otherwise resembled fir dwarf mistletoe. The Flat Top Mountain population had been previously Classified as A. abietinum f. sp. concoloris because white fir was also parasitized there (Hawksworth et al. 1967; Hawksworth and Wiens 1972, 1996). Since the fir dwarf mistletoe on Baldy and Flat Top Mountains infected both red and white firs, as well as Brewer spruce, we began intensive MATHIASEN AND DAUGHERTY: ARCEUTHOBIUM ABIETINUM SUBSP. WIENSII be, morphological, phenological, and host suscepti- bility studies of the populations of dwarf mistletoes parasitizing true firs and Brewer spruce in northern California and southwestern Oregon in 1998. Our morphometric analyses and data on host susceptibility differences support the de- scription of a new subspecies of fir dwarf mistletoe. Arceuthobium abietinum Engelm. ex Munz subsp. wiensii Mathiasen & C. Daugherty, subsp. nov. Wiens’ dwarf mistletoe.—Type: USA, Califor- nia, Siskiyou Co., 41°49’21”"N, 123°28'05”"W. Elev. 1820 m. Klamath National Forest, Baldy Mountain, 18.2 km west of Indian Creek on forest rd. 17N11 (Doolittle Creek rd.), parasitic on Picea breweriana, 8 Aug., 2008. R. L. Mathiasen and C.M. Daugherty O&23 (holo- type: ASC; isotypes: JEPS, UNM, US). Plantae 4-16 (9) cm altae; surculi prinicipales basi 2—6 (3) mm diam.; internodis tertiis 6—27 (14) mm longis, 2 mm latis; fructus maturi 4.2 mm longi, 3.0 mm latis; anthesis mense Julio—Au- gusto; fructus maturitas Septembri—Octobri. In Abies magnifica et Picea breweriana parasiticae. Plants 3.8—16.1 cm in height (mean ca. 9 cm); basal diameter of dominant plants 1.8—5.8 mm (mean 3.1 mm); third internode length 5.6— 27mm (mean 14.0 mm) and 1.9mm wide (Figs. 2 and 3); staminate plants primarily green-brown, but many red-brown; pistillate plants primarily green-brown, but some red- brown or, rarely, yellow-brown; staminate flow- ers 3 or 4-partite, flower diameter 2.0—3.8 mm (mean 2.8 mm); mature fruit length 3.1—5.0 mm (mean 4.2 mm) and 2.2-3.5 mm wide (mean 3.0 mm)(Fig. 4). Seeds 1.9-2.9 mm long (mean 2.4 mm) and 0.9—-1.5 mm wide (mean 1.1 mm). Phenology: Anthesis from early July through late August with peaks in late July to early August; seed dispersal from early September to mid October with peaks in late September to early October. Habit: Principally parasitic on Abies magnifica and Picea breweriana. Occasionally parasitic on Abies concolor and Abies lasiocarpa (Hook.) Nutt. Rarely parasitic on Pinus monticola Dou- glas ex D. Don. The host susceptibility classi- fication used here and below is based on the system described in Hawksworth and Wiens (1996). Distribution: Wiens’ dwarf mistletoe occurs in northwestern California from South Fork Moun- tain (Trinity Co.) north through the Siskiyou and Klamath Mountains into southwestern Oregon on Flat Top Mountain (Josephine Co.)(Fig. 1). Elevational range is from ca 1500 m on South Fork Mountain to as high as 2000 m on Baldy Mountain northwest of Happy Camp, California. Representative specimens: Arceuthobium abie- tinum subsp. wiensii—Paratypes: [all citations 120 MADRONO [Vol. 56 O - wiensii Ol - magnificae @ - concoloris Fic. 1. Approximate locations of populations sampled for Arceuthobium abietinum subsp. wiensii, A. abietinum f. sp. magnificae and A. abietinum f. sp. concoloris. Open circles indicate populations of subsp. wiensii, open squares — represent f. sp. magnificae and closed circles represent f. sp. concoloris. Locations: 1—Big Tree rd N of Trout Lake; 2—east side of Suttle Lk.; 3—17 km S of Sisters on rd 16; 4—south boundary of Crater Lake Nat. Park on Rte. 62; — 5—1 km E of entrance to Joseph Stewart St. Park; 6—Flat Top Mtn.; 7—Steve Fork Cr.; 8—Althouse Mtn.; 9— Bolan Mtn.; 10—Jct. of Greyback rd and rd to Kelly Lk.; 11—Baldy Mtn. 18 km W of Indian Creek on Doolittle; Cr. rd; 12—Rock Creek Butte 41 km N of Orleans; 13—Yellow Jacket Ridge on trail to Chimney Rock Lk.; 14— Etna Summit on rd to Sawyers Bar; 15—3 km N of Eaton Lk. in Russian Peak Wilderness; 16—South Fork Mtn. | 20 km N of Rte. 36; 17—South Fork Mtn. 8 km N of Rte. 36; 18—-10 km W of Stewart Hot springs on Park Creek rd; 19—1 km W of Black Butte trailhead; 20—10 km W of McCloud on Rte. 89; 21—11 km W of entrance to. Lassen Nat. Park on Rte. 44; 2216 km S of north entrance to Lassen Nat. Park on Rte. 89; 23—17 km SW of: 2009] FIG. 2. based on Abies magnifica except as noted] USA. OREGON. Josephine Co.: Flat Top Mountain, on Picea breweriana, 2003, Mathiasen 0357 (&) (ASC); same site, 2007, Mathiasen 0741 (0 & 9) (ASC); same site, on Abies lowiana, Mathiasen 0742 (& & 9) (ASC) (Fig. 1, site 6); Steve Creek, on Picea breweriana, 2003, Mathiasen 0360 (7 & 2) (ASC) (Fig. 1, site 7); Althouse Mountain, 2003, Mathiasen 0336 (& & 9) (ASC) (Fig. 1, site 8). CALIFORNIA. Siskiyou Co.: Bolan Moun- tain, 2007, Mathiasen 0743 (& & 9) (ASC); same site, 2008, Mathiasen 0835 (9) (ASC) (Fig. 1, site 9); Baldy Mountain, 1998, Mathiasen 9884 (&’) (ASC) (Fig. 1, site 11); same site, on Abies lowiana, Mathiasen 9883 (&) (ASC); same site, on Picea breweriana, Mathiasen 9885 (&) (ASC); same site, 2003, Mathiasen 0334 (°) (ASC); same site, on Picea breweriana, Mathiasen 0332 (&) << MATHIASEN AND DAUGHERTY: ARCEUTHOBIUM ABIETINUM SUBSP. WIENSH 12] (*) hi ve Zy KP a fia : SUEY 1 cm Male plants of Arceuthobium abietinum subsp. wiensii in September. (ASC); same site, on Abies lowiana, Mathiasen 0333 (&°) (ASC); same site, 2007, Mathiasen 0746 (0) (ASC); same site, on Picea breweriana, Mathiasen 0745 (9) (ASC); same site, 2008, on Picea breweriana, Mathiasen & Daugherty 0823 (o & 9) (ASC) and Mathiasen & Daugherty 0833 (¢) (ASC); Etna Summit, 2008, Mathiasen 0830 (9 & 9) (ASC) (Fig. 1, site 14); Trinity Co.: South Fork Mountain, 2008, Mathiasen OS65 (o & 9) (ASC) (Fig. 1, site 16). A total of 7 populations of Arceuthobium abietinum subsp. wiensii were sampled from within its geographic range (Fig. 1). From each population, 20 to 40 infections were collected and the dominant shoot from each infection was used for morphological measurements. For each pop- ulation, 10 or 20 male and female infections were collected. Old Station on Rte. 44; 24—13 km SE of rd A21 on Rte. 44; 25—6 km W of Meadow Valley on road to Bucks Lake; 26—Grizzly Summit W of Bucks Lake; 27—1 km E of Lyons Cr. Trailhead; 28—-Echo Summit on U.S. 50; _ 29—9 km W of Crane Flat on Tioga Pass rd, Yosemite Nat. Park; 30—16 km W of Crane Flat on Tioga Pass rd, Yosemite Nat. Park; 31—10 km E of Fish Camp on rd 6807; 32—11 km E of Fish Camp on rd 6S07. FIG. 3. MADRONO \) Zw 2 } PA eau, seth yin nA MOI. SU \Y \ Ww eee WS NS AS 4 sy Negscimr a mT Wa Ht ae =< >. — : i HI TTT : = . RE S “Le ZUR EES j ‘ ae. le ? that will be mature the following year. FIG. 4. Ss Le Qld sc mea NAG ! AN 1 cm Mature fruits of Arceuthobium abietinum subsp. wiensii in October. [Vol. 56 Female plants of Arceuthobium abietinum subsp. wiensii in November. These plants have developing fruits 2009] In order to make a comparison with the morphological characters of Arceuthobium abie- tinum f. sp. magnificae a total of 6 populations of f. sp. magnificae were sampled from the central and northern Sierra Nevada Mountains (Fig. 1). From each population, 20 to 40 infections were collected and the dominant shoot from each infection was used for morphological measure- ments. For each population, 10 or 20 male and female infections were collected. In addition, 20 populations of Arceuthobium abietinum f._ sp. concoloris were sampled throughout most of its geographic range (Fig. 1). From each population, 10 male and 10 female infections were collected and the dominant shoot from each infection was used for morphological measurements. The dwarf mistletoe plant characters measured were those used by Hawksworth and Wiens (1996) for taxonomic classification. The following morphological characters were measured: height, basal diameter, third internode length and width, and color of the tallest male and female shoot from each infection collected; mature fruit length, width, and color; seed length, width and color; staminate flower diameter; number, length and width of staminate perianth lobes; and anther diameter and anther distance from the perianth lobe tip. Plants were measured within 24 hr after collection and were measured using a digital caliper, a dissecting microscope with a microme- ter, or with a Bausch and Lomb 7 hand lens equipped with a micrometer. A minimum of 100 measurements or observations of color were made for each character above. A one-way analysis of variance (ANOVA, P = 0.05) was used to determine if there were statistical differences between the means of the morpholog- ical characters measured. Plants of Arceuthobium abietinum f. sp. con- coloris and f. sp. magnificae were morphologically similar as reported by Hawksworth and Wiens (1972, 1996). Although several morphological characters were similar between the special forms of A. abietinum and A. abietinum subsp. wiensii, there were several consistent morphological differences (Table 1). Male and female plants of subsp. wiensii were consistently smaller than those of both f. sp. concoloris and magnificae, whose plants were about the same size. Differ- ences in plant height were not statistically different between the special forms of A. abieti- num, but the plant heights of subsp. wiensii were different than the plant heights of the special forms for both male and female plants. As noted earlier, the color of plants of subsp. wiensii were frequently green-brown or red-brown, while the color of plants of f. sp. concoloris and f. sp. magnificae were typically yellow-green, or yellow -as described by Hawksworth and Wiens (1972, 1996). However, occasionally the female plants, -and rarely the male plants of f. sp. magnificae MATHIASEN AND DAUGHERTY: ARCEUTHOBIUM ABIETINUM SUBSP. WIENSII 123 were green-brown, particularly at the northern end of its geographic range, but none were red- brown. The reddish-brown to nearly orange color of the plants of subsp. wiensii is a distinctive character of this dwarf mistletoe. Staminate flowers of subsp. wiensii were smaller in size (both 3 and 4-merous flowers) than those of f. sp. concoloris and f. sp. magnificae (Table 1). The mean diameters of both 3- and 4-merous flowers for subsp. wiensii were statistically differ- ent than those of the special forms. Hawksworth and Wiens (1972, 1996) reported that staminate flower diameters of A. abietinum (for both special forms) were 2.5 mm, so this must have been for 3- merous flowers because our measurements of 3- merous flower diameters averaged 2.6 and 2.7 mm for A. abietinum f. sp. concoloris and f. sp. magnificae, respectively. Hawksworth and Wiens did not report flower diameters for 4- merous flowers, but we found that the average diameter of 4-merous flowers was smaller for subsp. wiensii (3.2 mm) than for the special forms (3.5 mm). Another character which typically is not informative, but was significantly different between subsp. wiensii and the special forms was the distance of the anther from the tip of the perianth lobe. This distance averaged 0.4 mm (range 0.2—0.7 mm) for the special forms, but averaged 0.6 mm for subsp. wiensii and some anthers were as much as one mm from the tip of the perianth lobe. This character could often be observed when comparing the anther location on perianth lobes between the special forms and subsp. wiensii using a 10 hand lens. Fruits of Arceuthobium abietinum subsp. wien- sii were smaller on average than those for both special forms (Table 1), but the range in fruit length and width overlapped for all three taxa. However, the mean length of the fruits of subsp. wiensii was Statistically different than the mean lengths of the special forms, but not the mean width. Furthermore, the color of fruits for subsp. wiensii was green-brown to green while the color of fruits for both special forms was consistently green. Seeds of subsp. wiensii were also smaller on average than those of the special forms, but the differences were not statistically significant. The principal hosts of Arceuthobium abietinum subsp. wiensii were red fir and Brewer spruce, but it is most common on red fir. We presently know of only three locations where it occurs on Brewer spruce; Flat Top Mountain and Steve Creek, Josephine Co., Oregon and Baldy Mountain, Siskiyou Co., California. However, it severely parasitizes Brewer spruce at these locations, but only a few spruces are infected near Steve Creek and the 2002 Biscuit Fire killed most of the severely infected spruce on Flat Top Mountain. However, our observations on Flat Top Moun- tain and those of Hawksworth et al. (1967) before the Biscuit Fire indicated that many Brewer 124 MADRONO TABLE 1. [Vol. 56 MORPHOLOGICAL MEASUREMENT RESULTS FOR WHITE FIR DWARF MISTLETOE, RED FIR DWARF MISTLETOE, AND WIENS’ DWARF MISTLETOE. Data are presented as means with range, and (sample size) below. Means followed by different letters in the same row were significantly different (ANOVA, P = 0.05). Plant heights in cm; all other measurements in mm. 'Distance of anther from the tip of the perianth lobe. Character mistletoe mistletoe mistletoe Female plants Height I2Z0A 1L8A 9.5°B 7.0-19.6 (200) 5.6-18.4 (100) 3.8-16.1 (130) Basal diameter 3.7 A 3.6A 3.2 B 2.3—6.4 (200) 2.4-6.1 (100) 1.8-5.8 (130) Third internode 16.0 A IS.8 A 14.7 B length 9.3-26.8 (200) 8.1—24.7 (100) 5.6-27.0 (130) Third internode 22A pa 1.9 B width 1.5—3.1 (200) 1.43.0 (100) 1.3—2.9 (130) Color yellow-green/yellow yellow-green/yellow/green- green-brown/red-brown brown Male plants Height 122A 116A 8.9 B 6.7—19.4 (200) 7.2-19.2 (120) 3.5-15.7 (160) Basal diameter 3.6.4 3.5 A 3445 2.2—6.1 (200) 1.9-4.8 (120) 1.9-5.2 (160) Third internode IS9 A IS.4 A 13.5 B length 7.0—23.8 (200) 9.9-24.4 (120) 5.4-23.0 (160) Third internode 2.3 A 2.2A 1.9 B width 1.6—3.6 (200) 1.6—3.1 (120) 1.42.9 (160) Color yellow-green/yellow yellow-green/yellow/green- green-brown/red-brown brown Staminate flowers Diameter—3-merous 2.6A 2.7 A 2.4B 2.3—3.0 (150) 2.43.1 (130) 2.0—2.9 (140) Diameter—4-merous 3.5 A SSA 3.2 B 3.1-4.0 (150) 2.64.1 (130) 2.6-3.8 (140) Perianth lobe length 1L3A L3A 112A 1.1—1.7 (300) 1.0-1.9 (260) 0.8—2.0 (280) Perianth lobe width LIA LIA LOA 0.8—1.4 (300) 0.9-1.5 (260) 0.7—1.4 (280) Anther diameter 0.5 A O.5A OSA 0.2—0.7 (300) 0.3—0.7 (260) 0.40.7 (280) Anther distance! 0.4 A 0.4 A 0.6 B 0.2—0.6 (300) 0.3-0.7 (260) 0.4—-1.0 (280) Mature fruits Length 4.7A 46A 4.2B 3.35.6 (160) 3.7—5.7 (120) 3.1—5.0 (150) Width 3.1A SIA 3.0 A 2.3-4.0 (160) 2.4-3.9 (120) 2,2-3.5 (150) Color green, slightly glaucous green, slightly glaucous green to green-brown, slightly glaucous Mature seed Length 2.5 A 205 A 24 A 1.9-3.3 (160) 2.0—3.3 (120) 1.9-2.9 (150) Width LIA LIA LIA 0.8—1.6 (160) 0.9-1.6 (120) 0.9-1.5 (150) Color dark green dark green dark green White fir dwarf Red fir dwarf Wiens’ dwarf spruces were severely infected there. Many severely infected Brewer spruces are present on the east slopes of Baldy Mountain as well as many dead spruces killed by the mistletoe. Wiens’ dwarf mistletoe occasionally infects white fir at Flat Top Mountain, Baldy Mountain, Etna Summit, and South Fork Mountain (Fig. 1, sites 6, 11, 14, and 16). It rarely infects western | white pine at Flat Top Mountain, but we have not found it parasitizing western white pine at. other locations where this tree is associated with | severely infected red fir or Brewer spruce. Our observations in the Sierra Nevada Mountains | fully support the earlier reports of the host. 2009] TABLE 2. MATHIASEN AND DAUGHERTY: ARCEUTHOBIUM ABIETINUM SUBSP. WIENSII 125 PRINCIPAL MORPHOLOGICAL AND HOST DIFFERENCES BETWEEN WHITE FIR DWARF MISTLETOE, RED FIR DWARF MISTLETOE, AND WIENS’ DWARF MISTLETOE. Mean plant heights in cm, all other means in mm. 'Host susceptibility system follows Hawksworth and Wiens (1972, 1996). Host classifications for Wiens’ dwarf mistletoe are based on data from field observations. Host classifications for white fir dwarf mistletoe are based on field observations and data presented in Hawksworth and Wiens (1996). Host classifications for red fir dwarf mistletoe are based on data presented in Hawksworth and Wiens (1996). Character White fir dwarf mistletoe Mean plant height Male 122 Female 12.0 Mean basal diameter Male 3.6 Female 3.7 Mean length of third internode Male 15.9 Female 16.0 Mean flower diameter 3-merous 26 4-merous 355 Mean fruit length 4.7 Anther distance from tip 0.4 Plant color: Male Female yellow-green/yellow yellow-green/yellow Host susceptibility’ Red fir dwarf mistletoe yellow-green/yellow/green-brown yellow-green/yellow/green-brown Wiens’ dwarf mistletoe i 8.9 8 Se) a2 3.1 3.6 3.2 15.4 [345 1538 14.7 2a 2.4 35 ee 4.6 4.2 0.4 0.6 green-brown/red-brown green-brown/red-brown White fir principal host immune occasional host Red fir immune principal host principal host Brewer spruce immune unknown principal host Western white pine rare host unknown rare host specificity of the special forms of A. abietinum there. Because subsp. wiensii infects both red fir and white fir, it is physiologically and genetically distinct from both of the special forms. Further- more, at three locations where f. sp. concoloris was severely infecting white fir, it did not infect associated Brewer spruce (Fig. 1, sites 10, 12, and 15). The absence of infection of Brewer spruce by f. sp. concoloris is further support that it 1s genetically distinct from subsp. wiensii. The geographic range of subsp. wiensii does not overlap with f. sp. magnificae and there is a distinct gap in the distribution of these dwarf mistletoes (Fig. 1). We have been unable to locate any infection of red fir near Mt. Shasta or west of there around Mt. Eddy, although red fir is common in these areas. Hawksworth and Wiens (1996) illustrated this geographic gap in their distribution map for red fir dwarf mistletoe (see their fig. 16.6). However, f. sp. concoloris is common in these areas (Fig. 1, sites 18, 19, and 20 and see Hawksworth and Wiens 1996, fig. 16.3). Our map (Fig. 1) only includes the locations of populations we sampled during this study and is not meant to replace the distribution map for f. sp. magnificae in Hawksworth and Wiens (1996) which illustrates the location of several other populations of this mistletoe. Although the distributions of A. abietinum subsp. wiensii and f. sp. concoloris overlap, we have not observed any areas where they are sympatric. Incidence of infection of white fir was consistently low (36-47%) in mixed red fir/white fir stands infested with subsp. wiensii. In contrast, in stands with f. sp. concoloris, incidence of infection of white fir was common (>91%). In addition, f. sp. concoloris plants can easily be distinguished from subsp. wiensii by plant size and color. Differences in plant size, geographic distribu- tion, and host range have been the principal characters used to separate subspecies of Ar- ceuthobium (Hawksworth and Wiens 1972, 1996; Hawksworth et al. 1992; Wass and Mathiasen 2003; Mathiasen 2007, 2008; Mathiasen and Daugherty 2007) and these are the principal characteristics that distinguish Arceuthobium abietinum subsp. wiensii from the special forms of A. abietinum (Table 2). Because we also found that the morphological similarities between the special forms of A. abietinum reported by Hawksworth and Wiens (1972, 1996) are consis- tent, particularly in the Sierra Nevada Mountains where the special forms are often sympatric, we concluded that describing these populations as subspecies of A. abietinum was not appropriate. 126 Arceuthobium abietinum subsp. wiensii 1s named in honor of Delbert Wiens, ardent student of Arceuthobium and one of the principal contemporary architects of the genus. His col- laboration with the late Frank Hawksworth spanned over 25 yr during which they described most of the currently recognized species of Arceuthobium in Mexico and Central America, as well as several taxa found in California that many botanists and foresters recognize as distinct species (Hawksworth and Wiens 1993; Mathiasen and Marshall 1999, Hansen and Lewis 2000, Geils et al. 2002). ACKNOWLEDGMENTS We thank Dr. M. Socorro Gonzalez Elizondo for the Spanish translation for the Resumen and Victor Leshyk for help with the figures. LITERATURE CITED GEILS, B. W., J. CIBRIAN TOVAR, AND B. Moopy. 2002. Mistletoes of North American conifers. USDA Forest Service General Technical Report RMRS-GTR-98, Rocky Mountain Research Sta- tion, Ft. Collins, CO. HANSEN, E. AND K. J. LEwIs. 1997. Compendium of conifer diseases. APS Press, St. Paul, MN. HAWKSWORTH, F. G. AND D. WIENS. 1972. Biology and classification of dwarf mistletoes (Arceutho- bium). Agriculture Handbook 401, USDA Forest Service, Washington, D. C. AND . 1993. Viscaceae. Pp. 1092-1097 in J. C. Hickman (ed.), The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. MADRONO [Vol. 56 AND . 1996. Dwarf Mistletoes: Biology, pathology, and systematics. Agriculture Handbook 709, USDA Forest Service, Washington, D. C. , AND D. P. GRAHAM. 1967. Dwarfmis- tletoe on Brewer spruce in Oregon. Northwest Science 41:42—44. , AND D. L. NICKRENT. 1992. New western North American taxa of Arceuthobium (Viscaceae). Novon 2:204-211. MATHIASEN, R. L. 2007. A new combination for Hawksworth’s dwarf mistletoe (Viscaceae). Novon T7222, . 2008. New combinations for Arceuthobium aureum (Viscaceae) in Mexico and Central Amer- ica. Novon 18:571—579. AND C. M. DAUGHERTY. 2005. Susceptibility of conifers to western hemlock dwarf mistletoe in the Cascade Range of Washington and Oregon. Western Journal of Applied Forestry 20:94—100. AND 2007. Arceuthobium tsugense subsp. amabilae, a new subspecies of hemlock dwarf mistletoe (Viscaceae) from Oregon. Novon VE222-227, AND K. MARSHALL. 1999. Dwarf mistletoes in the Siskiyou-Klamath Mountain Region. Natural Areas Journal 19:379—385. NICKRENT, D. L., M. A. GARICA, M. P. MARTIN, AND R. L. MATHIASEN. 2004. A phylogeny of all species of Arceuthobium (Viscaceae) using nuclear and chloroplast DNA sequences. American Journal of Botany 91:125—138. PARMETER, J. R. AND R. F. SCHARPF. 1963. Dwarf- mistletoe on red fir and white fir in California. Journal of Forestry 61:371—-374. WASS, E. F. AND R. L. MATHIASEN. 2003. A new subspecies of Arceuthobium tsugense (Viscaceae) from British Columbia and Washington. Novon 13:268—276. MADRONO, Vol. 56, No. 2, pp. 127-129, 2009 A NEW SPECIES OF STREPTANTHUS (BRASSICACEAE) FROM TRINITY COUNTY, CALIFORNIA THOMAS W. NELSON AND JANE P. NELSON 4244 Fairway Drive, Eureka, CA 95503-6413 janetomnel@humboldtl.com ABSTRACT Streptanthus oblanceolatus T. W. Nelson and J. P. Nelson is here described and illustrated. It is endemic to steep metavolcanic bluffs along the gorge of the Trinity River above its confluence with the New River, where three voucher specimens have been obtained. Key Words: Brassicaceae, California, metavolcanic, new species, Streptanthus, Trinity County. This novelty was first collected, but not recognized as new, by Richard W. Spellenberg in 1965 on a steep, metavolcanic bluff in the gorge of the Trinity River at its confluence with the New River, Trinity County. We were alerted to it by a note in The Jepson Manual (Buck et al. 1993) appending the treatment of Streptanthus tortuosus Kellogg, that stated the existence of an undescribed species near Burnt Ranch in Trinity County. Streptanthus oblanceolatus is most close- ly related to the S. tortuosus complex based on its bracteate racemes and expanded receptacle (AI- Shehbaz, Missouri Botanical Garden, personal communication). TAXONOMY Streptanthus oblanceolatus T. W. Nelson and J. P. Nelson, sp. nov. (Fig. 1)—Type: USA, Cali- fornia, Trinity Co., steep bluff along State Route 299, 100 m E of Shasta-Trinity National Forest boundary, ca. 1.69 km (1 mile) W of Burnt Ranch, T 5N, R 6E, sec. 3, 418 m (1350 ft) elevation, 2 June 2004, 7. W. Nelson and J. P. Nelson 9217 (holotype, HSC; isotypes, MO, CAS). Herba biennis, caulibus ligneis 50-100 cm longis, bracteis lanceolatus, petalis luteis, non nisi pari supero filamentorum arcte adhaerentis, semi- nibus 3/4 circumferentiae alatis, siliquis patentibus parum recurvatis, et substrates non serpentines. Plant biennial; stems woody, erect, glabrous, glaucous, 50-100 cm long, arising from woody taproot system; inflorescence with many primary and secondary branches and many small lance- Olate bracts; leaves oblanceolate, dark green adaxially, gray and glaucous abaxially, 4.5-7 cm long, 4-6 mm wide, reduced in size upward; calyx _biradial; sepals yellow, inverse boat-shaped, recurved at tips, 8-9 mm long, enlarged at base, partially enclosing lower portions of limb of petals; petals yellow, bilaterally symmetrical, Tecurved at tips, long exerted, upper pair 13— 16 mm long, lower pair 12—13 mm long; stamens: adaxial pair of filaments 13-16 mm long, exsert- ed, nearly equaling petals in length, connate on lower 2/3rd, anthers sterile, lateral pair 7.5— 9.0 mm long, abaxial 9-11 mm long, free; anthers fertile; pistil green, 6-7 mm long, stigma entire; siliques torulose, spreading, slightly curved: seeds oblong, ca. 2 mm long, ca. 1 mm wide, brown, rugose on one surface, ribbed on the other, winged 3/4 of circumference; seedling: stem 22— 25 cm tall, leaves subequal, oblanceolate with 2 small teeth on each side in the upper half (Fig. 1). Paratypes. USA. CALIFORNIA. Trinity Co.: along NW slope of New River from mouth to 1/4 mile upstream on steep, rocky banks, 17 June 1965, R. Spellenberg 1153 (HSC); growing near Grays Falls campground, 70—100 ft. NE of swinging bridge beside foot trail, 3 April 1972, D. Santana 699 (HSC). DISTRIBUTION AND HABITAT Streptanthus oblanceolatus is likely to be ex- tremely rare as it is known only from the three collections cited herein, all collected within five miles of one another in Trinity County. It may be more widespread on cliffs on the opposite side of the Trinity River gorge and in the New River gorge where there are many steep, rocky bluffs. Howev- er, we were unable to search this roadless area as the old swinging bridge that once provided access across the Trinity River gorge has been removed, and attempts to gain access from an old trail above the New River gorge were unsuccessful. Streptanthus oblanceolatus grows in fissures and soil pockets on an exposed, nearly vertical rock face above State highway 299 and in gravels below the highway, within an open Quercus chrysolepis—Pseudotsuga menziesii forest on east to north-facing aspects. Its few associates include Selaginella wallacei, Eriogonum nudum, Keckiella corymbosa, Sedum spathulatum, Pentagramma triangularis subsp. triangularis, and Polystichum imbricans subsp. imbricans. Overstory and under- 128 MADRONO [Vol. 56 SS SS Wee > Fic. 1. I[llustration of Streptanthus oblanceolatus. A. Mature plant at anthesis; B. Flower in lateral view; C. Front | view with sepals removed; D. Top view with sepals removed; E. Seedling; F. Seed side view. 2009] TABLE l. MORPHOLOGICAL COMPARISON OF STREPTANTHUS OBLANCEOLATUS AND THE. S. NELSON AND NELSON: A NEW SPECIES OF STREPTANTHUS 129 TORTUOSUS COMPLEX. Characteristics of S. tortuosus from Buck et al. 1993. Character S. oblanceolatus Longevity Biennial Adaxial stamen pair Connate below middle, with sterile anthers Basal leaves Oblanceolate Cauline leaves Inflorescence bracts Numerous Stigma Entire Sepals Yellow Siliques Spreading, straight to slightly curved story cover are lacking and shrub and herb cover are sparse (<2%). A few plants were found on a roadside mound of rockslide scrapings in associ- ation with Bromus hordeaceous, Briza maxima and Avena fatuda. RELATIONSHIPS Streptanthus has traditionally been divided into three subgenera based primarily on filament fusion. Subgenus Streptanthus does not occur in California. The two generally Californian sub- genera are Pleiocardia, in which filaments are free to the base, and Euclisia, in which filaments are partially to fully connate (Hoffman 1952). Under this model, S. oblanceolatus would belong in subgenus Euclisia. However, we are following the advice of Dr. Al-Shehbaz (personal communica- tion) who, in the absence of molecular data, discourages recognition of infrageneric categories in medium-sized and larger genera of the Brassicaceae, due to considerable homoplasy in morphological features. Filament connation has also evolved in the closely related Caulanthus and Thelypodium. Molecular studies in conjunction with morphology will be necessary to determine relationships and character evolution within Streptanthus and related genera. Streptanthus oblanceolatus was noted as unde- scribed in an annotation to the treatment for S: tortuosus in The Jepson Manual (Buck et al. 1993). Streptanthus oblanceolatus and S. tortuosus are both characterized by an expanded receptacle and racemes that are bracteate below or between the flowers. Based on these characters, S. oblanceolatus appears to be most closely related to the S. fortuosus complex (Al-Shehbaz personal communication). Table 1 compares morphologi- cal differences between the two taxa. ACKNOWLEDGMENTS Thomas W. Nelson passed away unexpectedly in October of 2006 as he was about to submit this Oblanceolate, reduced upward S. tortuosus Annual to perennial Free to the base, with fertile anthers Oblong to widely ovate Round to oblong, upper often larger Generally | Weakly lobed Purple (yellowish in var. flavescens) Generally reflexed, curved manuscript to Madrono. The manuscript is his 12th species description and represents his final contribution to the botanical world and our understanding of the California flora. It was through getting to know Dr. Doris Niles while in his 30s that Tom discovered his passion for plants. He changed his major from Physics to Botany, started exploring the local mountains, wrote the Flora of the Lassics mountain range for his Master’s thesis, and worked for several years as herbarium curator at HSC. Tom spent many years searching different environments for new plants, although he had a special interest in serpentine plants. When he found something unfamiliar he left no stone unturned to determine whether it was an undescribed taxon. He was always happy to share his knowledge with others, and in turn was always seeking to learn from other botanical experts. I am very grateful for the many special years we spent together following his passion. We thank Dr. Guy L. Nesom and Dr. Ihsan [. Al- Shehbaz for the Latin description and Christina Paleno for providing the fine illustration. We thank Dr. Al-Shehbaz, Dr. John Hunter and an anonymous reviewer for their constructive comments on the manuscript. We are especially grateful to Dr. Al- Shehbaz for his generous help with generic relation- ships. We also thank Dr. Ronald L. Hartman for his review of the original manuscript, and Dr. James P. Smith, Jr. and Robin Bencie for access to the excellent collection of northwest California plants at HSC. We are grateful to Susan Erwin and Sydney Carothers for assistance in the field and to Sydney for revising the original manuscript. We appreciate that Dr. Dean W. Taylor brought this very rare Streptanthus to our attention. LITERATURE CITED BUCK, R. E., D. W. TAYLOR, AND A. R. KRUCKEBERG. 1993. Streptanthus. Pp. 439-444 in J. C. Hickman (ed.), The Jepson manual: higher plants of Califor- nia. University of California Press, Berkeley, CA. HOFFMAN, F. W. 1952. Studies in Streptanthus. A new Streptanthus complex in California. Madrono 11:221—233. MADRONO, Vol. 56, No. 2, pp. 130—135, 2009 NOTEWORTHY COLLECTIONS CALIFORNIA GERANIUM YEOI Aedo & Munoz Garm. (GERA- NIACEAE).—Del Norte Co., common, grassy road- banks, Ocean View Drive near North Indian Road, N of Smith River mouth, 25 m, 19 Oct 2008, Zika 24223 (DAO, DAV, MA, NY, UC, US, WTU); common in ditches, Ocean View Drive near Eagle Crest develop- ment, 30 m, 3 Nov 2008, Zika 24271 (CAS, CHSC, GH, HSC, MA, MO, RSA, WTU); shady roadside thickets, Ocean View Drive 0.3 km N of Lopez Road, 10 m, 3 Nov 2008, Zika 24272 (MA, WTUV). Previous knowledge. Yeo’s geranium is endemic to central Maderia in Macaronesia, where it is frequent in woodlands, on sunny slopes and along irrigation channels (Press et al. 1994. Flora of Madiera. The Natural History Museum, London). It appears in garden catalogs and older floras as G. rubescens Yeo (not G. rubescens Andrews). It undoubtedly was first introduced in northern California as a garden ornamental, although it is biennial and not perennial. A useful illustration and key can be found in Yeo (1997. The European Garden Flora 5:26—-50). In its foliage and habit the species resembles a husky and large-flowered G. robertianum L., with petals 17-21 mm long. Duplicates were kindly verified by Geranium expert Carlos Aedo in Madrid, who writes that it is most similar to another Madeiran endemic that has escaped from cultivation in California, G. palmatum Cav. Dr. Aedo notes the two are most easily separated by the glandular indument, which is purple in G. palmatum and colorless with red tips in G. yeoi. They can also be separated by habit and stamen length. Significance. This is the first report of Geranium yeoi as a wild plant in the flora of California and North America, according to Carlos Aedo (personal communication, MA Herbarium, Real Jardin Botanico). The species was present intermittently along several miles of roadsides, and should be sought elsewhere in the northern North Coast region of California, as well as in adjacent Oregon. JUNCUS ANTHELATUS (Wiegand) R. E. Brooks (JUNCACEAE).—Sonoma Co., Pitkin Marsh, near Forestville, lower part of the Upper Marsh, in the open on marshy ground, 14 Jul 1951, P. Rubtzoff 472 (RSA); same site, 28 Jul 1951, P. Rubtzoff 564 (CAS, DS). Previous knowledge. Kentucky rush is native to eastern North America (Brooks. 2000. Juncus subg. Poiophylli, in Flora of North America Editorial Committee, eds. Flora of North America North of Mexico. Vol. 22), and adventive in British Columbia. Wetland acreage has been severely reduced in this part of the Northern Outer Coast Ranges, but hosts a number of rare species with conservation issues. These wetland remnants should be surveyed for Juncus anthelatus, which may have been overlooked among the similar J/. tenuis Willd., a California native that shares with it a tufted habit, green to reddish tepals, and 1-8 mm long acute or acuminate semi-translucent auricles at the base of the leaf blades (at least on spring and summer shoot growth). The two can be separated by the following key: la. Fruit <2.5 mm, <3/4 length of perianth; stem with 2—6 prominent wide pale ridges visible per side; flowers usually < internodes and scat- Juncus anthelatus lb. Fruit >2.5 mm, >3/4 length of perianth; stem with 0—1 prominent wide pale ridges visible per side; flowers often > internodes and tending to be more clustered towards branch tipsJuncus tenuis Significance. First report for California. JUNCUS PLANIFOLIUS R. Br. (JUNCACEAE).— Humboldt Co., common in damp sunny ditches, Route 101 near Exit 728 for Trinidad Beach, 110 m, 19 Oct 2008, Zika 24229 (CAS, CHSC, GH, HSC, MO, PRA, RSA, UC, UCR, WS, WTU). Previous knowledge. New Zealand rush is indigenous to New Zealand, Australia, Chile, and the Juan Fernandez Islands, and is reported naturalized in Hawai, Ireland, and Oregon (Kirschner et al. 2002. Juncus subg. Juncus sect. Graminifolii, in Species Plan- tarum: Flora of the World, Vol. 7). In my experience it is a common weed in disturbed ground and on freshwater shores along the coast in southern Oregon. Naturalized populations in Oregon and California are unusually variable in habit. Plants a mere | cm tall can flower as an annual. Most plants appear to be cespitose perennials 20— 50 cm tall, but careful excavation of some will reveal elongated rhizomes. Significance. First report for California, and likely to be found elsewhere in the North Coast region. —PETER F. ZIKA, WTU Herbarium, Box 355325, University of Washington, Seattle, WA 98195-5325. Zikap@comcast.net. IDAHO ALICIELLA (Gilia) TRIODON (Eastw.) Brand (PO- LEMONIACEAE).—Owyhee Co., southeast of Oreana about 9 km west of where Oreana cutoff road crosses | Birch Creek and about 1 km southwest from there, in | ashy barrens with Eriogonum ochrocephalum, Tetradymia glabrata, Eriogonum salicornioides, Artemisia spinescens, and Atriplex confertifolia, 42°52.110'N 117°17.147'W, 1130 m, 11 June 2006, D. Mansfield 06-42 (CIC). Previous knowledge. Three-tooth gilia ranges from |, southern California across southern Nevada, Utah, and | Arizona, to Colorado (J. M Porter. 1998. Aliso 17:230- 46; A. Day. 1993. Gilia, in J.C. Hickman, ed. Jepson | Manual. University of California Press, Berkeley). Curtis Bjérk reported collecting this same species in : this vicinity in 2005 (personal communication). Significance. First report for Idaho. PRENANTHELLA EXIGUA (A. Gray) Torr. (ASTER- | ACEAE).—Owyhee Co., east of Oreana in gravelly soil with lacustrine subsoil with Artemisia spinescens and | Atriplex, T5S R1E S11, 800m, 21 June 1978, Rag Rosentreter s.n. (CIC, UI; det. C. R. Bjérk, 2002); | southeast of Oreana about 1/2 mi west of where Oreana | cutoff road crosses Birch Creek, in ashy barrens with | a ae Eriogonum ochrocephalum, Tetradymia glabrata, Erio- | gonum salicornioides, Artemisia spinescens and Atriplex | confertifolia, 42°58.843"N_ 116°18.978"W, NAD 83, | 2009] 880 m, 11 June 2006, D. Mansfield 06-41 (CIC); about 10 mi west-southwest of Grandview on Oreana cutoff road., in shadscale barrens with Tetradymia glabrata, Chrysothamnus viscidiflorus ssp. puberula, Cryptantha spiculifera, Oryzopsis hymenoides, and Stanleya pinnata. 42°57.03’”N 116°17.88”"W, NAD 83, 940 m, 2 July 2005, D. Mansfield 05-386 (CIC). Previous knowledge. Desert prenanthella (syn. Lygo- desmia exigua (A. Gray) A. Gray) is a southwest desert species known previously to extend to Washoe and Humboldt Counties in Nevada (A. Cronquist. 1994. Asteraceae, in A. Cronquist et al., eds. Intermountain Flora. Vol. 5. New York Botanical Garden, Bronx. NY). These collections are the basis for the USDA Plants Database citation in Idaho (http://plants.usda.gov). Significance. First report for Idaho. POGOGYNE FLORIBUNDA Jokerst (LAMIACEAE).— Ada Co., south of Boise, about one mile south of Tenmile Creek and 3/4 mi west of Cole Road, near the head of a rocky watercourse tributary to the main watercourse in the area, growing between large basalt cobbles paving the bottom of an ephemeral watercourse, with Artemisia ludoviciana, Elymus cinereus, Tae- niatherum caput-medusae, Rumex crispus, Veronica per- egrina, Deschampsia danthonoides, Navarettia sp., and Epilobium paniculatum, area outside the watercourse has burned and is dominated by a Bromus tectorum/ Sysimbrium altissimum community. 23 May 2000, H. Swartz OOHSO03 (CIC, verified by D. Mansfield). Previous knowledge. Profuseflower mesamint is a Modoc Plateau species extending from northeastern California to Klamath and Lake Counties in Oregon (Oregon Plant Atlas; http://cladonia.nacse.org/platlas/ jcelass/OPAJava20.htm). It was recently found in Mal- heur County, Oregon. Significance. First report for Idaho. UTRICULARIA GIBBA L. (LENTIBULARIACEAE).— Custer Co., in Mays Creek bog, 23.5 road km south of Stanley and 2 km south on Forest Service Rd. 210, just west of Hwy 75, in standing water of Sphagnum bog with Carex spp., Drosera anglica, Juncus spp., and Mimulus primuloides, 44°01.170"N 114°51.538"W, NAD 83, 2090 m, 13 July 2007, M. Markin 0767 (CIC, verified by D. Mansfield); and 6 July 2008, M@. Markin 0801 (CIC). Previous knowledge. Humped bladderwort is distrib- uted widely in eastern North America and west of the Cascade/Sierra/Coast Range crest from California to British Columbia (USDA Plants Database, /oc. cit.). Significance. First report for Idaho and possibly for the Rocky Mountains. ERIOGONUM PALMERIANUM Reveal (POLYGONA- _ CEAE).—Owyhee Co., Bruneau Canyon downstream _ from the mouth of the East Fork, in sedimentary deposits _ of silt, sand, and clay along the river bank and scattered | basalt and rhyolite with Juniperus scopulorum and Celtis _ reticulate, T10S R7E S4 SE of SW, 970 m, 12 October 1980. L.C. Smithman with J.E. Smithman, B. Pierce, and . G. Pierce, LS-580 (CIC, verified by J. Reveal). Previous knowledge. Palmer’s buckwheat is distribut- ed throughout the deserts of southwestern United _ North America _ University Press, New York, NY). States extending into the northern Great Basin in _ Nevada. (J. L. Reveal. 2005. Eriogonum, in Flora of North America Editorial Committee, eds. Flora of North of Mexico. Vol. 5. Oxford Significance. First record in Idaho. This early | collection is the basis for Idaho being cited in Flora NOTEWORTHY COLLECTIONS 13] of North America (J. Reveal. 2005. loc. cit.). The collection, originally identified as E. nidularium, was attributed to this taxon during a review of herbarium materials in preparation of the treatment of this genus for the Flora of North America. —DONALD H. MANSFIELD and MELINDA MARKIN, Harold M. Tucker Herbarium (CIC), Department of Biology, The College of Idaho, Caldwell, ID 83605. dmansfield@collegeofidaho.edu. OREGON PRENANTHELLA EXIGUA (A. Gray) Torr. (ASTER- ACEAE).—Malheur Co., fewer than 10 individuals on west-facing, tan, chalky ash outcrop east of Owyhee River north of Lambert Rocks, one mile south of Bogus Creek with Caulanthus crassicaulis, Cryptantha propria, Stanleya pinnata, and Eriogonum novonudum, T29S R41E SS NW % NE %4, 43°05.13'N 117°42.65'W, NAD 83, 950 m, 7 June 2002, D. Mansfield 02-220 (CIC, OSC; verified K. Chambers). Previous knowledge. Desert prenanthella (syn. Lygo- desmia exigua (A. Gray) A. Gray) is a southwest desert species known previously to extend to Washoe and Humboldt Counties in Nevada (A. Cronquist. 1994. Asteraceae, in A. Cronquist et al. Intermountain Flora, Vol. 5. New York Botanical Garden, Bronx. NY). No plants were observed on a subsequent visit to this location in 2005, though there was no precipitation on the site until 11 March of that year, so this was an unusual drought year in which we may not have expected to see the population. Significance. First report for Oregon. This collection is the basis for Oregon being cited in Flora of North America (K. L. Chambers. 2006. Prenanthella, in Flora of North America Editorial Committee, eds. Flora of North America North of Mexico. Vol. 19. Oxford University Press, New York, NY). PHYSARIA COBRENSIS (Rollins & E. A. Shaw) N. H. Holmgren (BRASSICACEAE).—Malheur Co., on al- kaline clay flats, 1.5 mi east of Anderson Crossing, Guadalupe Meadows Quad, 42°7.113'N 117°16.434'W, 24 May 2002, H. Nielsen 2002383 (CIC); very few plants in one pocket of crunchy, cracking clay loam with Artemisia tridentata var. wyomingensis, Agropyron spicatum, Poa secunda, Elymus elymoides, Eriogonum ovalifolium, and Delphinium on the southwest flank of Black Butte, west of the road, Guadalupe Meadows Quad, T39S R47E S19 SW %4SW %4, 5680 ft, 16 June 2005, Jean Findley JF1199 (CIC); ona playa just north of Black Butte and northeast of Anderson Crossing of West Little Owyhee River with Wyoming sage, Poa secunda, Phacelia, Allium spp., and Eriogonum ovalifolium, 42°11.672'N 117°12.264'W, 5540 ft, 12 June 2005, H. Nielsen 1041 (CIC, OSC); northeast of Anderson Crossing with Poa sandbergii and Eriogonum caespitosum, T39S R41 E S22 SW %4, 5500 ft, 7 June 1996, Jean Findley 691 (CIC). Previous knowledge. This bladderpod (syn. Lequerella kingii S. Watson var. cobrensis Rollins and Shaw) is known previously from northern Washoe and Hum- boldt Counties in northwestern Nevada and in eastern Nevada and southeastern Idaho (N. H. Holmgren. 2005. Brassicaceae, in N. H. Holmgren, et al., eds. Intermountain Flora. Vol. 2b. New York Botanical Garden, Bronx. NY). Significance. First reports for Oregon. 132 MADRONO CAULANTHUS CRASSICAULIS (Torr.) S. Watson var. GLABER M. E. Jones (BRASSICACEAE).—Malheur Co., few plants on whitish, rocky, shallow outcrop south of main road west of McDermitt, NV, with Artemisia tridentata var. wyomingensis, winterfat, and Castilleja, Boghole Spring Quad, T41S R41E S12 SW Ys SW 4, 5060 ft, 15 June 2005, Jean Findley JF1189 (CIC, OSC); few plants on whitish, rocky, shallow outcrop south of main road west of McDermitt, NV, with Artemisia tridentata var. wyomingensis, winterfat, and Castilleja, Boghole Spring Quad, T41S R41E S12 SW % SW %, 5060 ft, 7 June 2005, Jean Findley JF 1137 (CIC,OSC, verified by K. Chambers); few plants on cracking clay loam, NV, with Artemisia tridentata var. wyomingensis, Pseudoroegneria spicata, Astragalus curvicarpus, near Anderson Crossing, 16 June 2005, Jean Findley JF 1198 (OSC, verified by K. Chambers). Previous knowledge. This variety 1s known previously from central Nevada and southwest into Utah and Arizona (N. H. Holmgren. 2005 /oc. cit.). Significance. First reports for Oregon. ERIOGONUM HOOKER! Torr. & A. Gray (POLY GO- NACEAE).—Malheur Co., on white shaly outcrop with Atriplex confertifolia and Oryzopsis hymenoides about 3.5 mi upriver from Birch Creek on ash slopes and cliffs north of Greeley Bar campsite on north side of river, flowers creamy yellow, 43°12.615’N_ 117°32.620'W, NAD 83, 865 m, 8 July 2004, D. Mansfield 04-329 with R. Stacy (CIC, OSC); on the Owyhee River above Greeley Bar about three miles upriver from Birch Creek, on tan ash above Hot Springs camp with other annual Eriogonum spp. and Cheanactis spp., The Hole in the Ground Quad, 43°12.6'’N 117°32.6'W, NAD 83, 914 m, 10 June 2002, D. Mansfield 02-358 with H. Kugler, H. Nielsen, and J. Loehrke, (CIC, verified by J. Reveal); in Leshe Gulch north of Runaway Gulch Road on road cut, T26S R45E S17 NE “% NW %, 28 August 2007, Jean Findley JF1240 (CIC). Both populations have been seen regularly in subsequent years. Previous knowledge. Hooker’s wild buckwheat is known previously from throughout Utah and Nevada. These collections are the basis for Oregon being cited in Flora of North America (J. L. Reveal. 2005. Eriogonum, in Flora of North America Editorial Committee, eds. Flora of North America North of Mexico. Vol. 5. Oxford University Press, New York, NY). Significance. First reports for Oregon. ASTRAGALUS TENELLUS Pursh (FABACEAE).—Mal- heur, Co., on open flat gravels with xeric shrubs at entry to mine site south of Spur Road into Opalite Mine, 14 mi NW of McDermitt, Nevada, T40S R40E S33 SW Ya SW %, 42°02.968'’N 118°02.211'W, 5200 ft, 12 June 2007, Jean Findley 1239 (CIC). Previous knowledge. Pulse milkvetch is widespread from Minnesota to Utah, southeastern Idaho, the Bonneville basin in Utah, and into Nevada, with a disjunct population in northwestern Washoe County, Nevada, having a wide ecological amplitude with both high elevation and valley populations. (R. Barneby. 1989. Fabaceae, in A. Cronquist et al. Intermountain Flora, Vol. 3b. New York Botanical Garden, Bronx. NY). Significance. First report for Oregon. ALICIELLA (Gilia) TRIODON (Eastw.) Brand (PO- LEMONIACEAE).—Malheur Co., on hills north of Rome on road to Crooked Creek, in barren, sandy desert scrub with abundant weedy, annual flora and [Vol. 56 Gilia leptomeria, Chaenactis stevioides, and Kochia americana, 42°51.075'N_ 117°41.151’W, NAD 83, 1070 m, 31 May 2005, D. Mansfield 0523 (CIC, verified by L. Johnson); about 50 km north of Rome between Bogus Creek and Chalk Basin on the east side of the Owyhee River on white, shaly, barren outcrop with Stanleya pinnata, shadscale, Eriogonum novonudum, Chaenactis macrantha, Hymenopappus, Caulanthus cras- sicaulis, and Penstemon — miser, 43°05.235'N 117°42.432'W, NAD 83, 1000 m, 13 June 2005, D. Mansfield 05195 (CIC); about 10 km north of Arock between Jordan Valley and Rome, on white, shaly, barren outcrop with Stanleya viridiflora, Eriophyllum, Hesperostipa comata, Salvia dorii, and Astragalus tetrapterus, T30S R42E S9 NW % SW ss W%, 42°57.786'N 117°35.244’'W, NAD 82, 1200 m, 13 June 2005, D. Mansfield 05193 (CIC, verified by L. Johnson); Owyhee uplands, at Dry Creek '% mile downstream of confluence with Hurley Spring Creek, 24 air miles south of Harper, common in gravelly swale on point bar in shrub-steppe with Artemisia tridentata, Achnatherum hymenoides, and Nama densum var. parviflorum, T23S R42E S34, 43°31.194’N 117°33.340’, 1100 m, 22 May 2003, N. Otting 524 (CIC, verified by L. Johnson); about 6 km south of Rome on a bluff southeast of Owyhee Springs near Owyhee River Cliffs on barren soils with shadscale, Gilia inconspicua, Salvia dorii, Dalea, and Eriogonum ovalifolium var. purpurescens, 42°47.508'N_ 117°35.973', NAD 83, 1200 m, 2 June 2005, D. Mansfield 0566 (CIC, verified by L. Johnson); on hills north of Rome on road to Crooked Creek in barren, sandy desert scrub with abundant weedy annual flora and Gilia leptomeria, Chaenactis stevioides, and Kochia americana, 42°51.075'N 117°41.151'W, NAD 83, 1050 m, 31 May 2005, D. Mansfield 0523 (CIC). Previous knowledge. Three-tooth gilia ranges from southern California across southern Nevada, Utah, and | Arizona, to Colorado (J. M. Porter. 1998. Aliso 17:230— | 46; A. Day. 1993. Gilia, in J. C. Hickman, ed. Jepson | Manual. University of California Press, Berkeley). The | populations collected in 2005 have not been seen in | these locations since 2005. That year Curtis Bjork | reported collecting this same species in southern Idaho | (personal communication). Significance. First reports for Oregon. DODECATHEON PULCHELLUM (Raf.) Méerr. ssp. | SHOSHONENSE (A. Nelson) Reveal (PRIMULA- © CEAE).—Malheur Co., about 4 km above Three Forks | of the Owyhee River by cliffs on the east bank of the river, © rooted in moss on a seeping cliff face with Viola | nephrophylla, Epilobium ciliatum, Agrostis stolonifera, | and Berula erecta, T35S R45ES10SE “% NW 1%, 42°32'N | 117°11', NAD 83, 1150 m, 12 July 2001, D. Mansfield 01- | 266 (CIC, verified by J. Reveal); at Three Forks of the Owyhee River, T34S R45E 835, 4000 ft, 27 April 1982, | Roger Rosentreter 2625 (CIC, verified by J. Reveal); north — of Hwy 95 at Crooked Creek Crossing between Burns | Junction and Rome on BLM land in alkaline swale | meadow with Juncus balticus, Crepis runcinata, Poa | nevadensis, and Potentilla gracilis, T32S R4l1E S6, 42°48.3576'N 117°44.2938'W, NAD 83, 1050 m, 3 June | 2004, D. Mansfield 04015 (CIC, verified by J. Reveal). Previous knowledge. This large, alkaline, warm | springs subspecies of the common beautiful shooting © star is known from southern Idaho. These collections | are the basis for Oregon being cited in Flora of North America (J. L. Reveal. In press. Dodecatheon, in Flora | of North America Editorial Committee, eds. Flora of 2009] North America North of Mexico. Oxford University Press, New York, NY). Significance. First reports for Oregon. MENTZELIA CONGESTA (Nutt.) Torr. & A. Gray (LOASACEAE).—Malheur Co., 21 km southeast of Rome, south of Burns Junction where Highway 95 crosses Crooked Creek, on alkaline, sandy flat just south of Crooked Creek with Sarcobatus vermiculatus, rabbitbrush, Thelypodium sp., Camissonia boothii ssp. alyssoides, Amaranthus spp., and Abronia_ turbinata, 42°40.594'N_ 117°51.888’W, NAD 83, 1900 m, | June 2005, D. Mansfield 05-43 (CIC). Previous knowledge. The Ventana blazing star is distributed in southwestern California in the Transverse Ranges, through the east side of the Sierra Nevada, and into western Nevada (B. Prigge. 1993. Loasaceae, in J. C. Hickman, ed. Jepson Manual University of Califor- nia Press, Berkeley). Significance. Though this is cited as “rare in se. Oregon and on the Snake River Plain of s. Idaho” (N. H. Holmgren, et al. 2005. Loasaceae, in N. H. Holmgren et al., eds. Intermountain Flora. Vol. 2b. New York Botanical Garden, Bronx. NY), I have been unable to find any records. This has apparently been collected earlier in the 20th century by Morton Peck, but there is no record in either the Oregon Plant Atlas of the Oregon Flora Project or in Peck’s Manual of Higher Plants of Oregon (Binfords and Mort, 1941). This species has not been seen at the site of this collection in the three years since 2005, despite intensive searches. POGOGYNE FLORIBUNDA Jokerst (LAMIACEAE).— Malheur Co., at a silver sage playa north of Bull Flat between Toppin Creek and Owyhee River, Beaver Charlie Breaks Quad, growing in playa bottom with Artemisia cana, Sarcobatus vermiculatus, Epilobium torreyi, Plagiobothrys cognatus, and Phacelia thermalis, T37S R48E S21] SW, 1560m, 30 June 1999, D. Mansfield 99-76 (CIC, verified by D. E. Boufford); dry silver sage playa 17 plus km northeast of Anderson Crossing, on grazed playa with Artemisia cana, Agoseris heterophylla, Epilobium brachycarpum, and Iva axillaris, T37S R48E S21 SW, 42°20.493'N 117°06.084’W, NAD 83, 1560 m, 14 June 2003, D. Mansfield with W. Harvey, C. Davis, and E. Mansfield 03-220 (CIC). Previous knowledge. Profuseflower mesamint is a Modoc Plateau species extending from northeastern California to Klamath and Lake Counties in Oregon (Oregon Plant Atlas, http://cladonia.nacse.org/platlas/ jclass/OPAJava20.htm; USDA PLANTS Database, http://plants.usda.gov). These collections are the basis for the USDA Plants Database citation in Oregon (http://plants.usda.gov). Significance. This is a considerable range extension from Lake County, OR. Other playas nearby have been _ searched for this species during the past 9 yr to without _ finding additional populations. COLLOMIA RENACTA Joyal (POLEMONIACEAE).— _ Malheur Co., immediately south of Anderson Crossing | on the east side of the West Little Owyhee in dark _ gravelly/sandy soils just above riverbank with Pseudor- _ oegneria _ Meadows Quad, T40S R46E S3 SE Y% NW % NW spicata and Basin Big Sage, Guadalupe Za, 42°07.778'N 117°19.031'W, 5400 ft, 23 May 2002, _ H. Nielsen 2002373 (CIC, verified by L. Johnson): about 0.5 mi west of Anderson Crossing on a gravelly southwest-facing hillside with Basin Big Sage and Pseudoroegneria spicata, T40S R46E S4 NW % NE NOTEWORTHY COLLECTIONS 133 Ya NE %4, 42°07.921'N 117°19.489'W, 5520 ft, 16 May 2002, H. Nielsen 2002331 (CIC, verified by L. Johnson); dozens of plants in clay with fine gravels and some rocks, east and adjacent to the road to Lucky Seven Cow Camp, north of main road to Anderson Crossing, with Artemisia longiloba, Sitanion hystrix, Phlox hoodii, Lomatium spp., Eriogonum caespitosum, and many other annual/perennial forbs, Starvation Spring Quad, T38S R44E 824 SW 4 SE 1%, 5540 ft, 27 June 2005, Jean Findley JF1207 (CIC, verified by K. Chambers); east of road to Lucky Seven Cow Camp and north of main road to Anderson Crossing on slightly north-facing slope in clay loams with fine gravels on surface, dozens of plants, with Artemisia longiloba, Sitanion hystrix, Phlox hoodit, Lomatium spp., Eriogonum caespitosum, and many other annual forbs, Chipmunk Basin Quad, T38S R44E S25 NE 4 NE '4, 5600 ft, 28 June 2005, Jean Findley, JF1209 (CIC, verified by K. Chambers). Previous knowledge. Barren Valley collomia is en- demic to southeastern Oregon. It was described from populations in the Barren Valley of eastern Malheur and western Harney counties. The range extends into Elko Co. NV (L. Johnson and R. L. Johnson. Systematic Botany 31:349—360). Significance. The collections reported here are range extension from other known Barren Valley populations, and serve as the basis for Malheur County citations in the Oregon Plant Atlas of the Oregon Flora Project (Joc. cit.). The site of the largest population of this species (Nielsen 2002373) has been quite variable during the past five years with 50—60 individuals in 2004, 700 in 2005, 10 in 2006, 0 in 2007, and 282 in 2008. CRYPTANTHA GRACILIS Osterh. (BORAGINA- CEAE).—Malheur Co., about five kilometers up the Owyhee River above Warm Springs, at Three Forks southwest of Jordan Valley, on green Leslie Gulch ash tuff outcrop with rabbitbrush, Chenactis douglasii, Eriogonum vimineum, and some blue bunch wheatgrass, 42°30.063'N 117°12.548'W, 1200 m, 27 May 2007, D. Mansfield 07089 (CIC); on barren hillsides northeast of the Owyhee River Ford north of the Birch Creek Campground below takeout about 1.5 km north of Birch Creek Ranch, on ashy north-facing slopes at both the top of the ridge and down the steep slope, 43°14.072’N 117°29.247’'W, 990 m, 24 May 2007, D. Mansfield 07071b (CIC). Previous knowledge. Slender cryptantha spreads from southeastern California, across Nevada into southern Utah and Arizona. It has been collected in Malheur County, but only rarely so, and has been misidentified as C. flaccida on occasion resulting in confusion. Significance. The populations along the Owyhee River are noteworthy in confirming the presence of this taxon in Oregon. Though the range is not extended by these, the species is sufficiently rare that these are noteworthy. See Oregon Plant Atlas of the Oregon Flora Project (/oc. cit.). TRIFOLIUM LEIBERGH A. Nelson & J. F. Macbr. (FABACEAE).—Malheur Co., about 34 km south of Jordan Valley and 16 km north of North Fork Camp- ground on road to North Fork Owyhee Campground and Three Forks two miles north of crossing back to Idaho, on barren outcrops with A//ium parvum, Perider- idia bolanderi, Microseris nutans, Eriophyllum lanatum, surrounded by mesic sage scrub with Idaho fescue, Poa sandbergii, and squirreltail, 42°42.537'N 117°02.703'W, 1500 m, 9 June 2005, D. Mansfield 05102 (CIC). Previous knowledge. Leiberg’s clover is known from a few populations in the Independence Mountains and 134 Jarbidge Mountains of northern Nevada with disjuncts near Drewsey in Harney County, Oregon. (R. Barneby. 1989. loc cit.) Significance. This collection near the Oregon-Idaho border represents a connection between the disjunct populations near Drewsey, OR and those in the Independence Mountains in Nevada, separated by about 200 km from each. ELATINE BRACHYSPERMA A. Gray (ELATINA- CEAE).—Malheur Co., in Antelope Creek drainage 6.5 km northeast to turnoff to Anderson Crossing en route to Peacock Lake, about 32 km northeast of McDermitt, NV, in muds of shore of small reservoir with Downingia laeta, Veronica peregrina, Psilocarphus brevissimus, Plagiobothrys hispidulus, and Juncus bufo- nius, 42°19.631’N 117°36.099'W, 1650 m, 27 June 2006, D. Mansfield 06-113 (CIC). Previous knowledge. This plant is irregularly distrib- uted throughout much of the western U. S. Significance. This collection is a range extension from other known populations. HETERANTHERA DUBIA (Jacq.) MacMill. (PONTE- DERIACEAE).—Malheur Co., at confluence of West Little Owyhee River and Owyhee River, growing in river about 0.5 km above Five Bar, Drummond Basin Quad, T36S R47E S15 SW 4, 42°26.4'N 117°11.0'W, 1340 m, 10 July 2002, D. Mansfield 02-828 with H Kugler, H. Nielsen, and J. Loehrke (CIC); along Owyhee River about 1.5 km upriver from Three Forks and just below Warm Springs, in a pond created by high water river channel with Potamogeton strictifolius, P. nodosus, P. alpinus, P. foliosus, Scirpus acutus, and Typhus latifolia, 42°31.84'N 117°11.06’W, 1200 m, July 19 2006, D. Mansfield 06-222 (CIC). Previous knowledge. Widely distributed but with few records from the Intermountain West. Significance. These collections in the Owyhee River drainage are range extensions from Lake County Oregon (Oregon Plant Atlas, /oc. cit.) and the Modoc Plateau in California (USDA Plants Database, /oc cit. ). —DONALD H. MANSFIELD and MELINDA MARKIN, Harold M. Tucker Herbarium (CIC), Department of Biology, The College of Idaho, Caldwell, ID 8360S. dmansfield@collegeofidaho.edu. WYOMING AMPHICARPAEA BRACTEATA (L.) Fernald (FABA- CEAE).—Crook Co., headwaters of Middle Fork of Hay Creek, Bear Lodge Mts., TS4N R62W S7 and 17, 1298-1329 m. Growing on creek banks, terraces, and a seasonally-flooded channel in a deep wooded valley, with Pteridium aquilinum, Heracleum sphondylium, Corylus cornuta, Betula papyrifera. 25 Jun 2008, J. Larson 11253 (RM); same site, 30 Jul 2008, J. Larson 11336 (RM). Previous knowledge. Widespread in eastern North America, previously known from the southern Black Hills, Pennington Co., South Dakota (Great Plains Flora Association. 1977. Atlas of the Flora of the Great Plains. Iowa State University Press. Ames, IA). Significance. First record for Wyoming, a second record for the Black Hills region. ASTRAGALUS DIVERSIFOLIUS A. Gray (FABA- CEAE).—Sweetwater Co., Chain Lakes, Circle Bar Lake and Mud Lake, Great Divide Basin, T23N R93W S4 and MADRONO [Vol. 56 5; T23N R91W S7 and 18; T23N R92 W S12 and 13; T24N R95W S5 and 6; T25N R95W S833; 1981—2018 m. Growing on alkaline meadows around playa lakes, with Pyrrocoma lanceolata, Triglochin maritima, Spartina gracilis, Sporobolus airoides, Carex praegracilis. 30 Jun 2007, B. Heidel 2935 (RM); 1 Jul 2007, B. Heidel 2939 (RM); 4 Jul 2008, B. Heidel 3160, 3161 (RM, BYU), confirmed by S. Welsh; same site, 24 Jul 2008, B. Heidel and J. Larson 3181 (RM, BYU, NY), confirmed by S. Welsh; same site, 26 Aug 2008, B. Heidel 3237 (RM). Previous knowledge. First collected by Thomas Nuttall on the 1834 Wyeth Expedition from ‘‘Sandy plains of the Colorado of the West, near the sources of the Platte”, generally referring to the Green River Basin, Wyoming, Nuttall s.n. (holotype at BM, isotypes at GH, K, NY, PH). Also known from _ widely- separated areas in east-central Idaho, southwestern edge of the Salt Lake Desert in eastern Juab and western Tooele Cos., Utah, and Spring Valley in southern White Pine Co., Nevada (R. Barneby. 1964. Atlas of North American Astragalus. Memoirs of the New York Botanic Garden, Vol. 13, NYBG Press, Bronx, NY; S. Welsh. 2007. North American Species of Astragalus Linnaeus: a taxonomic revision. Brigham Young University Press. Provo, UT; A. Tiehm. 1984. Madrono 31:123—127; J. Morefield. 2001. Nevada Rare Plant Atlas, http://heritage.nv.gov/atlas/atlas.html; Ida- ho Conservation Data Center. 2009. Idaho Special Status Plants, http://fishandgame.idaho.gov/cms/tech/ CDC/plants/; Utah Native Plant Society. 2009. Utah Rare Plant Guide, http://www.utahrareplants.org/). Significance. First recent records for Wyoming, over 420 km from the nearest historic station in Bingham Co., Idaho and over 450 km from Juab Siding in Juab Co., Utah: CAREX FOENEA Willd. (syn. CAREX AENEA Fernald) (CYPERACEAE ).—Crook Co., headwaters of Middle Fork of Hay Creek, Bear Lodge Mts., T54N R62W S17, 1262 m. Growing in a spring-fed, marshy opening in deep, wooded valley, with Carex pellita, Salix bebbiana. 25 Jun 2008, B. Heidel and J. Larson 3141 (RM, MICH, SDC), confirmed by A. A. Reznicek. Previous knowledge. Widespread in northern North | America (J. Mastroguiseppe et al. 2002. Carex section | Ovales in Flora of North American Editorial Commit- | tee, eds. Flora of North America North of Mexico. Vol. 23. Oxford University Press, New York, NY), in | northeastern South Dakota and more recently recog- © South | 527 (SDC) | annotated by G. Larson. Previous reports of Carex | foenea for Wyoming (R. D. Dorn. 2001. Plants of Wyoming, 3rd ed., Mountain West Publish- | nized from the Black Hills in Custer Co., Dakota, 23 Jun 1998, J.R. Johnson ing. Cheyenne, WY) were based on C. siccata Dewey. Significance. First record for Wyoming, a second | record for the Black Hills region. CAREX INTUMESCENS Rudge (CYPERACEAE).— Crook Co., headwaters of Middle Fork of Hay Creek, Bear Lodge Mts., T54N R62W S17, 1256 m. Growing | at a spring and spring-fed wetland in a deep, wooded | valley, with Carex disperma, C. interior, C. vulpinoidea, 30 Jul 2008, B. Heidel 3202 (RM). Previous knowledge. Widespread in eastern North | Vascular | i America, previously known from streams feeding out of | | the crystalline core and metamorphic portions of the central Black Hills in Custer and Pennington Cos., South Dakota (G. E. Larson and J. R. Johnson. 1999. ) q 2009] Plants of the Black Hills and Bear Lodge Mountains. South Dakota Agricultural Station, B732. Brookings, SD). Previous unpublished reports of C. intwmescens in Wyoming were based on a misidentification. Significance. First record for Wyoming, a minor range extension in the Black Hills region. CAREX SCOPARIA Schkuhr ex Willd. (CYPERA- CEAE).—Crook Co., headwaters of Middle Fork of Hay Creek, Bear Lodge Mts., T54N R62W S17, 1280 m. Growing along an open sandy creek bank of wet meadow vegetation in a deep, wooded valley, with Agrostis stolonifera, Glyceria grandis, Scirpus micro- carpus, 30 Jul 2008, J. Larson 11356 (RM, SDC), confirmed by G. Larson. Previous knowledge. Widespread in North America, previously known from the Black Hills in Lawrence Co. (Great Plains Flora Association. 1977. loc. cit.), Custer and Pennington Cos. (SDC), South Dakota. Significance. First record for Wyoming, a minor range extension in the Black Hills region. FIMBRISTYLIS PUBERULA (Michx.) Vahl var. INTERI- OR (Britt.) Kral (CYPERACEAE).—Niobrara Co., Niobrara River, T31IN R60W S7 and S18, 1445 m. Growing in a broad, sub-irrigated wet meadow in open plains, with Panicum virgatum, Equisetum laevigatum, Muhlenbergia richardsonis, Elymus trachycaulus, Pedi- cularis pulchella, 5 Aug 2006, B. Heidel 2875 (RM, SDC), confirmed by G. Larson. Previous knowledge. Known from southwestern, central and southern Great Plains states including Nebraska and Utah, previously known from Sioux Co., Nebraska (R. B. Kaul et al. 2006. The Flora of Nebraska. University of Nebraska Press. Lincoln, NE). Significance. First record for Wyoming, a minor range extension from Sioux Co., Nebraska. SCIRPUS PENDULUS Muhl. (CY PERACEAE).—Con- verse Co., Duck Creek, a tributary of the Cheyenne River, T37N R73W S14, 1527 m. Growing in sub- irrigated wet meadow between scattered cottonwood groves in open plains, with Agrostis stolonifera, Schoenoplectus pungens, Equisetum laevigatum, Elymus trachycaulus, Juncus longistylis, 15 Aug 2006, B. Heidel 2882 (RM, SDC), confirmed by G. Larson. Previous knowledge. Widespread in eastern North America, eastern Great Plains, southern states, Color- ado, California and Oregon, previously known from Boulder Co., Colorado Weber 50/1 (CU), where possibly a nursery escape (W. A. Weber and R. C. Wittmann. 2001. Colorado Flora: Eastern Slope, 3rd ed., University Press of Colorado. Boulder, CO). Significance. First record for Wyoming. —BONNIE HEIDEL, Wyoming Natural Diversity Database, Dept. 3381, University of Wyoming, 1000 E. University Ave., Laramie, WY 82071; and JILL LARSON, Black Hills National Forest-Northern Hills Ranger District, 2014 N. Main St. Spearfish, SD, 57783: bheidel@uwyo.edu. MEXICO SALVIA BRANDEGEEI MUNZ (LAMIACEAE).—Baja California, the bay of San Quintin, saddle and gentle NOTEWORTHY COLLECTIONS 135 slopes on eastern side of volcano ‘Riveroll’—the most northern of the cluster of volcanoes near N30.48910, W-116.01684, 399 ft. 28 March 2008, Sula Vanderplank, Jorge Ochoa, Dylan Hannon and Duncan Bell 080328-14 (RSA); on sandy east- to northeast-facing hillside above and west of the top of a prominent, steep, small rocky vernal waterfall, on the east (inland) slope of the cone, adjacent to the area where large-scale mining is taking place near the summit. Growing with Hazardia berberidis, Marah macrocarpa, Ephedra californica, Artemisia californica, Lycium andersonii, Aesulus parryi, Lasthenia_ californica, Oenothera wigginsii, Crassula connata, Encelia californica, Helianthus niveus, Nema- caulis denudata, Euphorbia misera, Dichlostemma_ pul- chellum, Eriogonum fasciculatum, Amblyopappus pusil- lus, Dudleya attenuata ssp. orcuttii, Mesembryanthemum crystallinum, and Lastarriaea coriacea. This population has approximately 130 plants and has probably already been impacted directly by local mining activities. Previous knowledge. The core range of S. brandegeei is in coastal northwestern Baja California. Records indicate that it occurs from the area near El Retiro (between Punta Banda and Santo Tomas) in the north. to Punta Colonet (Herbarium specimens: SD; RSA: UCR; BCMEX; HCIB; P.A. Munz 1974, A Flora of Southern California, University of California Press, Berkeley). A disjunct northern population is known from Santa Rosa Island (Averett, D. E., and K. R. Neisess, in J. C. Hickman [ed.] 1993, The Jepson Manual: Higher Plants of California, University of California Press, Berkeley; P.A. Munz 1974, Joc. cit.). Significance. The collection from the bay of San Quintin represents a southern range extension of ca 100 km from the nearest known population in Colonet. Given the small number of plants and populations this may represent a significant increase in the global numbers for this species. Sa/via brandegeei is considered Rare, Threatened or Endangered in California and elsewhere (list 1b.2) and has a global rank of G2 (California Native Plant Society (CNPS). 2008. Inven- tory of Rare and Endangered Plants (online edition, v7- O8b). California Native Plant Society. Sacramento, CA. Accessed on Jun. 7, 2008 from http://www.cnps.org/ inventory). Many of the plants are >2 m across and appear senescent, with often a high proportion of dead branches; few young plants were observed. Native sand bees (Anthophora sp.) were seen visiting the flowers repeatedly. In addition to herbarium vouchers, live material was collected as cuttings, one cutting from each of approximately 50 individuals, to be grown at Rancho Santa Ana Botanic Garden, and seed collection is pending. It is hoped that, should this population disappear from mining or other activity, some repre- sentation of it will exist ex sitw into the future. —SULA VANDERPLANK (sula.vanderplank@cgu.edu), DUNCAN BELL, Rancho Santa Ana Botanic Garden, 1500 N. College Avenue, Claremont, CA 91711; DYLAN HANNON, The Huntington Library, Art Collections, and Botanical Gardens, 1151 Oxford Road, San Marino, CA 91108; and JORGE OCHOA, Long Beach City College, Pacific Coast Campus, 1305 E. Pacific Coast Highway, Long Beach, CA 90806 & City of Los Angeles Dept. of Recreation and Parks, Forestry Division, 3900 West Chevy Chase Drive, Los Angeles, CA 90039. Volume 56, Number 2, pages 71-135, published 14 October 2009 SUBSCRIPTIONS — MEMBERSHIP Membership in the California Botanical Society is open to individuals ($35 per year; family $40 per year; emeritus $27 per year; students $27 per year for a maximum of 7 years). Late fees may be assessed. Members of the Society receive MADRONO free. Institutional subscriptions to MADRONO are available ($70). Membership is based on a calendar year only. Life memberships are $750. Applications for membership (including dues), orders for subscriptions, and renewal payments should be sent to CBS c/o Jepson Herbarium, University of California, Berkeley, CA 94720-2465. Requests and rates for back issues, changes of address, and undelivered copies of Ma- DRONO should be sent to the Corresponding Secretary. INFORMATION FOR CONTRIBUTORS Manuscripts submitted for publication in MADRONO should be sent to the editor preferably as Microsoft Word (.doc), Rich Text Format (.rtf), or Portable Document Format (.pdf) files. 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. For all articles and short items (NOTES, NOTEWORTHY COLLECTIONS, POINTS OF VIEW, etc.), fol- low 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, five 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 review copies of illustrations. The title page should have a running header that includes the name(s) of the author(s), and a shortened title. Avoid foot- notes except to indicate address changes. Abbreviations should be used sparingly and only standard abbreviations 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 are encouraged to include the names, addresses, and e-mail addresses of two to four potential reviewers with their submitted manuscript. Authors of accepted papers are required to submit an electronic version of the manuscript. Microsoft Word 2000 or later or WordPerfect 9.0 (or later) 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. 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 Index Herbariorum. 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 MADRONO. Joint au- thors 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 manuscript have been used in other papers that are published, in press, submitted, or soon to be submitted elsewhere. iiMN SMITHSONIAN INSTITUTION LIB UUM 3 9088 01512 6147 VOLUME 56, NUMBER 3 JULY-SEPTEMBER 2009 "i ee ‘ahi git i ng 4 oe ig x 4 A WEST AMERICAN JOURNAL OF BOTANY RIK " ISS bot ‘(CONTENTS Ear Ly Post-FirE PLANT ESTABLISHMENT ON A MOJAVE DESERT BURN Scott R. Abella, E. Cayenne Engel, Christina L. Lund, and ES SIC “ae SD CHOEH sancti en re pene een eerste eae enc ocamiees aie) INTERACTIVE EFFECTS OF SIMULATED HERBIVORY AND INTERSPECIFIC COMPETITION ON SURVIVAL OF BLACKBRUSH (COLEOGYNE RAMOSISSIMA) SEEDLINGS SUNON A. LEL siccseccccsescacsscavsvccavscccnscnvvense steeds ovcvecvceo Sle Gh ecocsvercossseuces 149 DIVERSITY, REPRODUCTION, AND POTENTIAL FOR INVASIVENESS OF EUCALYPTUS IN CALIFORNIA Matt Ritter and Jenn YOSt....00hilid ii, ccccc0eeeteeGoln eth okt Eas edlbaaileccoessevavecoesnsees 155 MOLECULAR PHYLOGENY OF THE PUNGENTES SUBSECTION OF CHORIZANTHE (POLYGONACEAE: ERIOGONOIDEAE) WITH EMPHASIS ON THE PHYLOGEOGRAPHY OF THE C. PUNGENS-C. ROBUSTA COMPLEX Chris Brinegarnadnd SQNArG BATON |. coo echo hess ages 000 Sendefbave sober J68 PLANT COMMUNITIES AND FLORISTIC DIVERSITY OF THE EMERALD LAKE BASIN, SEQUOIA NATIONAL PARK, CALIFORNIA Philip W. Rundel, Michael Neuman, and Peter Rabenold...............00000000+ 184 HISTORICAL RANGE EXTENSIONS FOR JUNIPERUS CALIFORNICA (CUPRESSACEAE) AND JUGLANS CALIFORNICA (JUGLANDACEAE) AT RANCHO MUSCUPIABE, SAN BERNARDINO COUNTY, CALIFORNIA Brett R. Goforth and Richard A. MInnichy.i..ccccccccccccecceeeeee cece 199 IEW SPECIES ELEVATION OF PHACELIA CICUTARIA VAR. HUBBYI (BORAGINACEAE) TO SPECIES STATUS Laura M. Garrison and Robert Patterson ........c.ccccceccceccneccuccnsccuccusceucensces 205 A NEw COMBINATION IN 7RIFOLIUM VARIEGATUM (FABACEAE) IVIFCHGACUZL NV INMGCIIL eee see eae ssa ee ee ee re 208 (OK REVIEWS SYSTEMATICS, EVOLUTION, AND BIOGEOGRAPHY OF COMPOSITAE. EDITED BY V. A. FUNK, A. SUSANNA, T. F. STEUSSY, AND R. J. BAYER TS OD SLELCTS OF Rectan etre menace eae rate NS Tae ee ee ee 209 YTEWORTHY ORIN cea aR aete teeth esas eee eas cs cna Manteca sa om Boron eae ce neasnete 211 )LLECTIONS (S58) VEL) 0 1 ea ne nc TY TIS ee 212 PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL SOCIETY MADRONO (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 MADRONO, Kim Kersh, Membership Chair, Uni- versity and Jepson Herbarium, University of California, Berkeley, CA 94720-2465. kersh@berkeley.edu. Corresponding Editor—TIMOTHY LOWREY Copy Editor—RICHARD WHITKUS Museum of Southwestern Biology Department of Biology MSC03 2020 Sonoma State University University of New Mexico 1801 E. Cotati Avenue Albuquerque, NM 87131-0001 Rohnert Park, CA 94928-3609 madrono@unm.edu whitkus @ sonoma.edu Book Editor—JON E. KEELEY Noteworthy Collections Editors—DIETER WILKEN, MARGRIET WETHERWAX Board of Editors Class of: 2009—DONOVAN BAILEY, New Mexico State University, Las Cruces, NM MARK BorcuHe_ert, USFS, Ojai, CA 2010—FReED Hrusa, California Department of Food and Agriculture, Sacramento, CA RICHARD OLMSTEAD, University of Washington, Seattle, WA 2011—JAMIE KNEITEL, California State University, Sacramento, CA KEVIN Rice, University of California, Davis, CA 2012—-GRETCHEN LEBUHN, San Francisco State University, CA ROBERT PATTERSON, San Francisco State University, CA CALIFORNIA BOTANICAL SOCIETY, INC. OFFICERS FOR 2009—2010 President: Dean Kelch, Jepson Herbarium, University of California, Berkeley, CA 94720, dkelch @ berkeley.edu First Vice President: Andrew Doran, University and Jepson Herbaria, University of California, Berkeley, CA 94720, andrewdoran @ berkeley.edu Recording Secretary: Nishanta Rajakaruna, Department of Biological Sciences, San Jose State University, One Washington Square, San Jose, CA 95192, Nishanta.Rajakaruna@sjsu.edu Corresponding Secretary: Heather Driscoll, University Herbarium, University of California, Berkeley, CA 94720- 2465, heather.driscoll @ nature. berkeley.edu Treasurer: Thomas Schweich, California Botanical Society, Jepson Herbarium, University of California, Berkley, CA 94720, tomas @schweich.com The Council of the California Botanical Society comprises the officers listed above plus the immediate Past President, Michael Vasey, Department of Biology, San Francisco State University, 1600 Holloway Avenue, San Francisco, CA 94132, mvasey @sfsu.edu; the Editor of Madrofo; three elected Council Members: James Shevock, Department of Botany, California Academy of Sciences, 55 Music Concourse Drive, San Francisco, CA 94118, jshevock@ nature.berkeley.edu; Roy Buck, Jepson Herbarium, University of California, Berkeley, CA 94720, roybuck@ email.msn.com; Chelsea Specht, Department of Plant and Microbial Biology, University of California, Berkeley, CA 94720-2465, cdspecht @nature.berkeley.edu. Graduate Student Representatives: Ben Carter, Department of Integrative Biology and University Herbarium, University of California, Berkeley, CA 94720-2465, bcarter@ berkeley.edu. Webmaster: Susan Bainbridge, Jepson Herbarium, University of California, Berkeley, CA 94720- 2465, sjbainbridge @ berkeley.edu. This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). MADRONO, Vol. 56, No. 3, pp. 137-148, 2009 EARLY POST-FIRE PLANT ESTABLISHMENT ON A MOJAVE DESERT BURN SCOTT R. ABELLA School of Environmental and Public Affairs, University of Nevada Las Vegas, 4505 S. Maryland Parkway, Las Vegas, NV 89154-4030 scott.abella@unlv.edu E. CAYENNE ENGEL School of Environment and Public Affairs, University of Nevada Las Vegas, 4505 S. Maryland Parkway, Las Vegas, NV 89154-4030 CHRISTINA L. LUND' Bureau of Land Management, Las Vegas Field Office, 4701 N. Torrey Pines Drive, Las Vegas, NV 89130 JESSICA E. SPENCER? Public Lands Institute, University of Nevada Las Vegas, 4505 S. Maryland Parkway, Las Vegas, NV 89154-2040 ABSTRACT Fire has become more extensive in recent decades in southwestern United States arid lands. Burned areas pose management challenges and opportunities, and increasing our understanding of post-fire plant colonization may assist management decision-making. We examined plant communities, soils, and soil seed banks two years after the 2005 Loop Fire, located in a creosote-blackbrush community in Red Rock Canyon National Conservation Area in southern Nevada’s Mojave Desert. Based on a spring sampling of 20, 0.01-ha plots, live + dead cover of the exotic annual Bromus rubens averaged nine times lower on the burn than on a paired unburned area. Perennial species composition shifted from dominance by late-successional native shrubs (e.g., Coleogyne ramosissima) on the unburned area, to dominance by native perennial forbs (e.g., Sphaeralcea ambigua, Baileya multiradiata) on the burn. Species richness of live plants averaged 26% (100 m° scale) and 239% (1 m? scale) greater on the burn compared to the unburned area. Only 5% of Larrea tridentata individuals resprouted, compared to 64% of Yucca schidigera and baccata. Fire and microsite (interspace, below L. tridentata, or below Yucca) interacted to affect several O-5 cm soil properties, with higher pH, conductivity, and total P and K on burned Yucca microsites. Bromus rubens density in 0-5 cm soil seed banks was four times lower on the burn, and its distribution among microsites reversed. Below-shrub microsites contained the most B. rubens seeds on the unburned area, but the least on the burned area. Intense fire below shrubs may have increased seed mortality, an idea supported by >3-fold decreases we found in emergence density after heating seed bank samples to 100°C. Our study occurred after a post-fire period of below-average precipitation, underscoring a need for longer term monitoring that characterizes moister years. Key Words: Colonization window, exotic species, grass-fire cycle, soil seed bank, wildfire. Wildfire has become widespread in southwestern _USA deserts. In a record 2005 fire season in the Mojave Desert, for example, more than 385,000 hectares burned (Brooks and Matchett 2006). This area represents approximately 3% of the entire Mojave Desert. Fueled in large part by exotic annual grasses, these fires burned desert shrub- lands thought to have only burned infrequently historically (Brooks 1999; Esque and Schwalbe 2002). Burns now occupy significant portions of desert landscapes, posing prominent management ‘Present address: Bureau of Land Management, California State Office, 2800 Cottage Way, Sacramento, @A 95825. * Present address: U.S. Army Corps of Engineers, 701 San Marco Blvd., Jacksonville, FL 32207. challenges. In addition to such concerns as post- fire soil erosion, burns pose challenges and opportunities for meeting mandates of land management agencies to manage for native species and communities in areas such as national parks and conservation areas. Fires in arid lands have been cited to initiate a “‘grass-fire cycle,” where fire promotes resurgence of the exotic grasses that fueled the fire to create a frequent-fire regime (D’Antonio and Vitousek 1992). Such a fire cycle threatens the sustainability of many fire-suscepti- ble, infrequently regenerating native species, such as Coleogyne ramosissina Torr. (Rosaceae) or Yucca brevifolia Engelm. (Agavaceae). However, just as in more mesic regions, deserts contain early successional native species that might benefit from fire (Cave and Patten 1984). 138 MADRONO Improving our understanding of plant coloni- zation on desert burns is important for evaluating future fire hazard, whether natural colonization will meet management objectives for burned sites, or for planning active revegetation if this becomes a management goal. It is possible that opportu- nities exist for intervention in the grass-fire cycle immediately following a fire if exotic grass competition 1s temporarily reduced while avail- able nutrients liberated by the fire increase. This time period may serve as a colonization window, which is often a key feature of plant succession in other regions. For example, periodic droughts provided windows for colonization by underrep- resented species during a 40-year permanent plot study of succession in New Jersey (Bartha et al. 2003). The post-fire environment, together with the frequent dry years characterizing desert climates, may permit colonization by species not dominant prior to fires (Lei 2001; Brooks 2002). In addition to precipitation, two of the many factors that may influence plant colonization are soil seed banks and post-fire soil properties. Seed banks can provide propagules of both native and exotic species if emergence requirements are met following disturbance (Warr et al. 1993). Seed bank contributions to plant colonization after wildfires hinge upon the density of seeds surviv- ing fire, their germinability, and post-fire seed deposition and dispersal (Lei 2001). Soil pH often increases after fire through combustion of organic acids or release of base cations. Correspondingly, availabilities of some soil nutrients also frequently increase initially post-fire (Raison 1979). On one hand, pulses of nutrient availability, following precipitation events, are considered important to the functioning of native desert plant communities (Noy-Meir 1973). On the other hand, exotic species often respond with vigorous growth to nutrient additions (James et al. 2006). We measured plant communities, soil seed banks, and soil properties during the second year after the 2005 Loop Fire in the eastern Mojave Desert. This burn occurred within a National Conservation Area and encompasses a popular tourist and recreation site. The fire burned within and spread across a scenic loop road that receives 900,000 visitors annually (Bureau of Land Management, Las Vegas Field Office, Las Vegas, NV). The area is regarded as a symbol of the “natural” desert landscape surrounding metro- politan Las Vegas, Nevada. Resource managers are concerned about post-fire succession in relation to aesthetics, native community recovery, future fuel loads, and if or how active revegeta- tion should be a management goal. Our objec- tives were to (1) compare plant species richness, composition, and soil properties between the burn and a matched unburned area, (2) assess resprouting of burned dominant native shrubs, and (3) assay soil seed bank composition and [Vol. 56 experimentally test responses of seed bank samples to the fire cues of heat and smoke. METHODS Study Site This study occurred inside the Loop Drive area of Red Rock Canyon National Conservation Area, managed by the Bureau of Land Manage- ment and located 15 km west of the western Las Vegas suburbs in Clark Co., southern Nevada. Precipitation at the Spring Mountain Ranch State Park weather station, 7 km south of the study site, has averaged 30 cm/yr (range = 6-61 cm/yr, coefficient of variation [CV] = 45%) during a 1977-2008 period of record (Western Regional Climate Center, Reno, NV). January daily mini- mum temperatures have averaged —1°C (range = —6 to 2°C, CV = 12%), and July maximum temperatures have averaged 36°C (range = 33-— 38°C, CV = 2%). We studied the 348-ha Loop Fire, which started from a lightning ignition on July 22, 2005. Maximum temperature on this day was 38°C, and erratic winds gusted up to 80 km/hr during the fire. Typical flame heights were =2 m and flame lengths were =6 m (Troy Phelps, Bureau of Land Management, personal commu- nication). Fire managers also noted that the fuel provided by exotic annual grasses facilitated ignition of native shrubs and allowed the fire to burn continuously across the landscape. The fire left fewer than five unburned islands, all <1 ha in size, within the burn perimeter. We matched an unburned area immediately adjacent to the west of the burned area that was similar in climate, soils, and vegetation. Based on plot sampling described below, mean elevation of the burned area (1274 m) differed by 54 m from the mean elevation of the unburned area (1220 m). Soils on the burned and unburned areas were both mapped in the county soil survey as >90% of the 731 (Purob-Irongold association) and 732. (Purob extremely gravelly loam) types (Lato 2006). These soils are classified as Typic and Calcic — Petrocalcids, have alluvium parent material de- | rived from limestone, and occur on fan piedmont | landforms. Based on identifying resprouts and _ burned shrub skeletons on the burned area, both | the burned and unburned areas were similarly dominated by Coleogyne ramosissima, Yucca | schidigera Roezl ex Ortgies (Agavaceae), Yucca | baccata Torr. (Agavaceae), and Larrea tridentata | (DC.) Coville (Zygophyllaceae). Plot Sampling We overlaid a square grid of 315 ha encom-— passing 91% of the irregularly shaped burn and randomly located ten 10 m X 10 m (0.01 ha) plots | within the grid using randomly selected coordi- | 2009] nates in a geographic information system. We located 10 plots in the adjacent unburned area using the same method but in a square grid of 130 ha that the site would accommodate. Each plot contained six, 1 m X | m subplots located at the four plot corners and centered at 5 m along the southern and northern plot boundaries. We visually categorized areal percent cover of each live perennial species and of live and dead annual species rooted in each subplot using Peet et al.’s (1998) cover classes: | = trace (assigned 0.1%), 2 = 0-1%, 3 = 1-2%, 4 = 2-5%, 5 = 5-10%, 6 = 10-25%, 7 = 25-50%, 8 = 50-75%, 9 = 75-95%, and 10 = 95—-100%. We measured both live and dead annuals to provide a measure of the recent annual communities on plots. Dead annuals (especially the exotic grasses) also provide the fuel that allows fire to spread between perennial plants. We quantified measurement reproducibil- ity in subplot data by remeasuring a total of three randomly selected subplots on three different plots. Species richness/m° was identical in original and repeated measurements. The maximum difference in cover class between original and repeated measurements was one class. We inven- toried whole plots for species not already occurring in subplots and categorized cover of these species on a whole-plot basis. Plants not able to be readily identified to species in the field were collected, pressed, and identified to species when possible. Six vegetative specimens, which occurred on only one plot each, could not be identified to species or genus and were deleted from the data set. Nomenclature, classification of species life forms (e.g., shrub, forb), longevity (e.g., annual, perennial), and native/exotic status in North America followed the PLANTS data- base (USDA 2007). We counted the number of individuals of each shrub species on each plot. On a 10 m X 40 m (0.04 ha) plot originating at each 10 m X 10 m plot, we further counted the number of Larrea tridentata, Yucca baccata, and Yucca schidigera individuals and classified them into one of four categories: unburned alive, burned but crown _ survived, resprout, or dead (Rogers and Steele 1980). _ We collected soil and soil seed bank samples _ from three interspaces (>1 m from any shrub) _and from below the canopies of three each of the largest Larrea tridentata and Yucca spp. (Y. _ baccata or Y. schidigera) adjacent to each plot. In _ the shrub microsites, we collected a 200-cm® soil and a 200-cm* seed bank sample (cores 7 cm in diameter) halfway between the main central stem and the outer canopy edge on both the north and south sides of each shrub. Samples were collected _ from a 0-5 cm depth, which could include a litter | layer (only the below-shrub microsites contained _ appreciable litter) only for the seed bank samples. We included litter as part of the 0-5 cm seed bank ABELLA ET AL.: POST-FIRE PLANT ESTABLISHMENT 139 sampling depth because litter can trap seeds (Warr et al. 1993). We chose this depth for both soil and seed bank sampling because upper soils become hottest during fire and may more strongly reflect post-fire differences than deeper layers (Raison 1979; Patten and Cave 1984). This depth also includes the highest density of viable seeds in desert seed banks and is a maximum depth from which most seeds can emerge or reach emergence depth (Guo et al. 1998). We compos- ited soil and seed bank samples on a plot basis for each of the three microsite types. This resulted in a soil volume of 1200 cm* each for soil and seed banks for each microsite type on each plot. We also collected separate 200-cm* samples from each microsite for measuring soil bulk density (2- mm sieving followed by drying at 105°C for 24 hr). Sampling occurred in 2007 (two years post-fire) in spring (April-May), the time of peak annual plant community growth in the Mojave Desert (Beatley 1966). We did not collect soil seed bank samples during what is considered the optimal time (fall or early winter just prior to germination of winter annuals) in this region. However, as only 6 cm of precipitation fell (as opposed to the long-term average of 23 cm) from September 2006 through March 2007 prior to sampling, emergence of annual plants was limited. There- fore, depletion of soil seed banks through field germination should not have been a major factor affecting the detection of seeds in the greenhouse soil seed bank assay. Soil Laboratory Analysis The air dried, <2-mm fraction of soil samples was analyzed for texture (hydrometer method), electrical conductivity and pH (1:1 soil:water), total C and N (Leco C/N analyzer), and extractable P and K (Olsen NaHCO; method) following methods in Amacher et al. (2003). Measurement error, based on analyzing a dupli- cate sample every 10 samples and comparing original and duplicate means, ranged from 0.2% (electrical conductivity) to 6.8% (total N). Soil Seed Bank Experiment We assessed responses of seed bank samples to simulated fire exposure in a factorial experiment consisting of heat (two levels: presence or absence of 100°C heating) and liquid smoke (two levels: presence or absence of 10% smoke). Treatments were performed on 240 cm? soil volumes extract- ed from microsite/plot composite samples. We conducted heating treatments by placing samples in tins in an electric oven. Samples remained in the oven until the soil reached 100°C for one minute, which required about 10 min, at which time samples were removed from the oven. We 140 monitored temperature using a probe accurate to 0.1°C and placed directly in the soil. The soil cooled to 65°C within five minutes of being removed from the oven and to 45°C within 10 min. We chose the 100°C treatment to be within the temperature range previously reported to occur in upper soil layers during burns in the Sonoran and Mojave Deserts (Patten and Cave 1984; Brooks 2002). For smoke only samples and after heating for heat xX smoke samples, we applied 120 ml (diluted to 10% smoke by volume using tap water) of commercially available liquid smoke (Wright’s Brand, Roseland, NJ) to each sample. Liquid smoke contains the same butanol compound contained in air smoke that can affect germination in some species (Flematti et al. 2004). The non-smoke samples received 120 ml of tap water not containing liquid smoke. After applying treatments, we placed each sample in a 2 cm thick layer on top of 300 cm? of sterile potting soil in 700-cm* square plastic pots. We randomly arranged these pots, including six pots containing only potting soil to check for seed contamination (none was detected), on a bench in a greenhouse. This greenhouse did not receive supplemental lighting and was maintained at approximately 28°C during the first three months (summer) of the emergence period and 17°C during the last three months (fall). Samples were watered by an automated misting system that delivered approximately 1.5 cm of water/day. We conducted periodic (at least monthly) inven- tories during a six-month emergence period; 93% of the total number of emerging seedlings emerged within the first 10 d. We also checked samples every 1—2 d to monitor for potential mass mortality between inventories, which we did not detect. Data Analysis Statistical inference is limited to the particular fire that we studied (van Mantgem et al. 2001). We compared live, dead, and total mean cover of Bromus rubens L. (Poaceae), a fuel-producing exotic annual grass, and community species richness (per m* and per 100 m°’) between the burned and unburned area using two-tailed t- tests. Findings can change based on the scale of analysis for species richness, which is why we examined effects of fire at two different scales. Based on species importance values for live vegetation for each plot, we ordinated species composition using non-metric multidimensional scaling (Sorensen distance, thorough mode) in PC-ORD (McCune and Mefford 1999). We calculated importance values as the average of relative frequency (based on six subplots per plot) and relative cover (with percent cover calculated as the midpoint of a cover class). We tested the hypothesis of no difference in live plant species MADRONO [Vol. 56 composition between burned and unburned areas using multi-response permutation procedures (Zimmerman et al. 1985). We based this analysis on Serensen distance and performed it in PC- ORD. To examine the fidelity of live species with burn status, we performed indicator species analysis (Dufréne and Legendre 1997) using species importance values and a Monte Carlo test of significance (1000 permutations). Indicator species analysis combines the relative abundance and frequency of a species within a group to produce an indicator value that ranges from zero (no fidelity) to 100 (maximum fidelity). We analyzed univariate soils data as a_ split-plot analysis of variance (ANOVA) with burn status as the whole-plot factor and microsite (inter- space, below Larrea, or below Yucca) as the subplot factor using JMP (SAS Institute 2004). Tukey’s test was used for mean separation. Bromus rubens comprised 92% of seeds in the seed bank experiment, so we focused the statis- tical analysis on this species. Logl0 transformed B. rubens seed counts were modeled using a mixed-model ANOVA (SAS PROC MIXED; REML) with burn status, microsite, heat, smoke, and all two-, three-, and four-way interactions as fixed effects (SAS Institute 1999). This was a partially nested design, so plot nested within burn status and its two- and three-way interactions with microsite, heat, and smoke were treated as random effects. RESULTS Relative to longer lived species, live annual cover was low in the unburned area and both live and dead annual cover was low in the burned area. Live annuals averaged only 0.05% (SD = 0.09%) cover in the unburned area (0.4% of the 15% total mean live plant cover), and 0.6% (SD = 0.6%) in the burned area (27% of the 2% total mean live plant cover). Total mean cover of Bromus rubens was nine times lower in the burned — than in the unburned area, a difference driven by | significantly greater (t = —5.32, P < 0.001) dead | cover in the unburned area (Fig. 1). Although | live cover was sharply lower than dead cover in > both areas, live B. rubens cover was significantly | greater (t = 3.09, P = 0.006) in the burned area | relative to the unburned area. Species richness of live plants was 3.4 times greater at a 1-m° scale and 26% greater at a 100- m° scale in the burned area compared to the unburned area (Fig. 2). Exotic annuals in the burn comprised 50% of richness/m*, but this | percentage declined to 18% of richness/100m°. | Only 7% of richness at both scales in both the burned and unburned areas consisted of native annuals. Native perennial forbs constituted 34— | 35% of richness at the two scales on the burn, | compared to 2-13% on the unburned area. |} 2009] t = -5.18, P < 0.001, df = 18 5 3 O Burned Fic. 1. ABELLA ET AL.: POST-FIRE PLANT ESTABLISHMENT 141 @ Dead OAlive Unburned Areal cover of Bromus rubens on a burned and unburned area, Red Rock Canyon National Conservation Area, Mojave Desert. Error bars are one standard deviation for total cover. Total mean cover was compared between the burned and unburned areas using a two-tailed t-test. Conversely, native perennial shrubs composed 76-79% of richness in the unburned area and only 5—30% in the burned area. Species composition differed sharply between the burned and unburned areas (Fig. 3). Importance of Coleogyne ramosissima and Yucca baccata was positively correlated with unburned plots in the ordination. In contrast, species associated with burned plots included the native perennial forbs Sphaeralcea ambigua A. Gray (Malvaceae), Erio- gonum inflatum Torr. & Frém. (Polygonaceae), and Baileya multiradiata Harv. & A. Gray ex A. Gray (Asteraceae), the perennial grass Dasyochloa pulchella (Kunth) Willd. ex Rydb. (Poaceae), and the exotic annuals Erodium cicutarium (L.) L’Hér. ex Aiton (Geraniaceae) and Bromus rubens. Consistent with the ordination results, average Serensen similarities among burned (53%) and ‘unburned plots (47%) were about four times greater than the average similarity between burned and unburned plots (12%). This difference in species composition between the burned and unburned areas was significant based on multi- ‘response permutation procedures (T statistic = _—12.1, chance-corrected within-group agreement |A statistic = 0.27, P < 0.001). Distributions of individual species also reflect- ed differences between burned and unburned species composition (Fig. 4). Twelve species constituted 71-78% of all relative cover. Based on indicator species analysis, six of these species were significant (P < 0.05) indicators for the burn and three for the unburned area. Resprouting varied between measured shrub species, with only 5% of burned Larrea tridentata resprouting compared to 64% of burned Yucca (baccata and schidigera; Fig. 5). No measured L. tridentata and only 3% of Yucca in the burned area had crowns that survived the fire. Although we did not measure dead Co/eogyne ramosissima in the burned area, we found live individuals (three mature individuals presumably avoided by the fire) in only one plot on the burned area and no resprouts. In comparison, live C. ramosissima density averaged 4,560/ha (SD = 3,041) in the unburned area. Several soil properties differed significantly among microsites and between the burned and unburned areas (Table 1). There was a burn xX microsite interaction for four properties (pH, conductivity, P, and K) because the Yucca microsite contained higher levels in the burned than in the unburned area compared to the other microsites. Total C varied both with burn status and microsite, being greater on the burn. Total N was higher in Yucca microsites than in interspaces but did not differ overall between burned and unburned areas. Soil texture did not differ significantly among microsites or between burned and unburned areas. In the seed bank experiment, seed density of Bromus rubens reversed among microsites with burn status, resulting in a significant burn xX microsite interaction (Fs 34 = 47.6, P < 0.001). Seed density was greatest below shrubs on the unburned area, but interspaces contained the most seeds on the burn. Heat also was a significant effect (F; 13 = 101, P < 0.001) that sharply reduced the density of emerging seeds. Therefore, results for the three-way interaction of 142 MADRONO (a) 1 m?: t = 4.55, P < 0.001, df =18 ” 2] oO c £ Oo B8_8_8_8_8_8_8 8888.88 ” o Oo ® a. ” Y VM (b) 100 m?: t = 2.81, P = 0.012, df =18 2] ” ® c = Bad 2] o Oo ® a. ”) es WOE Burned FIG. Plant species richness on a burned and unburned area, Red Rock Canyon National Conservation Area, [Vol. 56 @ Native perennial shrub i Native perennial grass Ej Native perennial forb Mi Native annual grass Native annual forb OC Exotic annual grass Exotic annual forb AAANAAAARRAAAR aa OR eo Unburned eee Desert. Error bars are one standard deviation for total mean richness. Total mean richness was compared between the burned and unburned areas using a two-tailed t-test. burn status X microsite X heat (Fs 36 = 11.5, P< 0.001) are presented in Fig. 6. DISCUSSION Study Assumption and Context An assumption of this study is that differences between burned and unburned areas result primarily from the fire and not from pre-existing vegetation or environmental differences. The adjacent burned and unburned areas demonstrat- ed a priori similarity in mapped soil type (Lato 2006) and dominant vegetation type. Plot sam- { pling corroborated this similarity, with mean’ elevation differing by only 54 m, soil texture nearly identical differing by only 1—2% in sand, silt, or clay (Table 1), density was probably higher in the burned area (Fig. 5). Because of this, we did not statistically compare post-fire L. tridentata density between burned and unburned areas, as instead we focused on the proportion of burned individuals that resprouted. Annual precipitation was 134% of average in 2004 preceding the fire and 169% of average in the fire year of 2005. These high-precipitation and similar dominant shrub composition. Pre-existing Larrea tridentata 2009] EROCIC € yucEC & yuck SPHAMB ERIINF ABELLA ET AL.: POST-FIRE PLANT ESTABLISHMENT 143 Axis 1 FiG. 3. Non-metric multidimensional scaling ordination of plant species composition on a burned and unburned area, Red Rock Canyon National Conservation Area, Mojave Desert. The lengths of vectors are proportional to correlations with ordination axes, and only vectors with r? values = 0.3 are shown. Vectors for species are abbreviated as the first three letters of the genus and species: BAIMUL = Baileya multiradiata, BRORUB = Bromus rubens, COLRAM = Coleogyne ramosissima, DASPUL = Dasyochloa pulchella, EPHNEV = Ephedra nevadensis, ERIINF = Eriogonum inflatum, EROCIC = Erodium cicutarium, KRAERE = Krameria erecta, SPHAMB = Sphaeralcea ambigua, and YUCBAC = Yucca baccata. Vectors for soils are shown as microsite (int = interspace, It = Larrea tridentata, yuc = conductivity, K = potassium, and P = phosphorus). years resulted in accumulation of annual plant biomass flammable after senescence (Brown and Minnich 1986). Precipitation has been below average since the fire, averaging 65% of average in 2006 and only 14% of the January-May mean in 2007 up to the sampling time for this study (Western Regional Climate Center, Reno, NV). Annual plant production has been negligible since the fire, and our results may be different had different weather patterns occurred after the fire. This observation underscores a need for longer term monitoring, particularly given the high interannual variability characterizing Mo- Jave Desert climates (Beatley 1966). Bromus rubens Total mean cover (live + dead) of Bromus _rubens was sharply lower on the burned than the unburned area (Fig. 1). Furthermore, B. rubens density in the soil seed bank of the burn was only 28% of the density on the unburned area (Fig. 6). Yucca spp.) and variable (C = total carbon, EC = electrical Seed density also reversed among microsites, with greater density in interspaces in the burn but greater density below shrubs in the unburned area. Below-shrub microsites, where B. rubens is most abundant aboveground in unburned areas, likely burned more intensively due to greater fuel accumulation than in interspaces (Brooks 2002). This intense burning probably killed seeds, an idea supported by results of our experimental 100°C heating that sharply reduced emergence from samples. It is important, however, to determine temperature thresholds for seed mor- tality of B. rubens in comparison with other species. If B. rubens has a higher or lower threshold, this could influence its ability to exclude other species after fire. We detected 12 species other than B. rubens in seed bank samples, but the low abundance of these species (only 8% of the total seeds detected) precluded making these comparisons. Our finding that Bromus rubens was reduced after fire concurs with Brooks (2002), who found 144 MADRONO Relative cover (%) Fic. 4. [Vol. 56 @ Burned 0 Unburned Species Mean relative cover and indicator values (top of bars and only given for species that were significant indicators at P < 0.05) of the 12 most dominant species on a burned and unburned area, Red Rock Canyon National Conservation Area, Mojave Desert. ASTSPP = Astragalus spp., BAIMUL = Baileya multiradiata, BRORUB = Bromus rubens, COLRAM = Coleogyne ramosissina, DASPUL = Dasyochloa pulchella, ENCVIR = Encelia_ virginensis, EPHNEV = Ephedra nevadensis, ERIINF = Eriogonum inflatum, EROCIC = Erodium cicutarium, SPHAMB = Sphaeralcea ambigua, YUCBAC = Yucca baccata, and YUCSCH = Yucca schidigera. that B. rubens biomass declined after prescribed fires at three Mojave Desert sites. In that study, B. rubens was significantly decreased for four years after fire in shrub microsites and for two years in interspaces. Other studies have reported varying patterns of B. rubens abundance after fire. For example, no clear trend in cover of this species was apparent in a 1—37 yr time-since-fire chronosequence in the northeastern Mojave Desert (Callison et al. 1985). Similarly, Cave and Patten (1984) found that post-fire B. rubens abundance varied between wildfire and_ pre- scribed fire in the upper Sonoran Desert. As Brooks and Matchett (2003) note, longer term patterns of post-fire B. rubens abundance can hinge upon climate, disturbance (e.g., grazing), 400 eH Burned 300 - 250 200 No. individuals/ha Larrea Yucca Fic. 5. and other factors. It is unclear how much time is needed after wildfire for B. rubens to return to pre-fire levels, or which combinations of factors. (e.g., native community composition, soil prop- | erties) could result in higher or lower post-fire B. rubens abundance. Native Perennial Species Composition Although resprouts of perennial shrubs cur- | rently contribute little to plant cover on the burn, this contribution may increase over time as existing resprouts grow, particularly for Yucca baccata and schidigera. Our finding that these’ two Yucca species vigorously resprout concurs | with Minnich’s (1995) findings in Joshua Tree | Unburned O Unburned-alive M Burned-crown survived Resprout @ Dead Larrea Yucca Density of Larrea tridentata and Yucca spp. (schidigera and baccata) on a burned and unburned area, Red Rock Canyon National Conservation Area, Mojave Desert. Error bars are one standard deviation for total mean density. ABELLA ET AL.: POST-FIRE PLANT ESTABLISHMENT 145 2009] pleo 0 COV9'0 tVc9 0 61e8 0 ISvs'0 19ZS 0 Ice. 0 1000°0> 1000 0> [000'0> $900°0 70000 8c00'0 COIS 0 Scor 0 16st 0 1000 0> [000° 0> 1000 0> 1000°0> 1000 0> 1000 0> IPSOIOIPA] SPoffo 1OJOB ISlO'0 cOl0'0 VECCO CIvs 0 9ILE 0 L8ct 0 L100°0 LO00'0 1000 0> 1000'0> 61cL 0 (OT) 4¥L9°0 (C1) TI (07) €€ (TI) ss (Ch) PLT0 ($9) SLE (TE) IPTE (OL) 99€ (€7) 9L°0 (I) QL‘L (8) 990L°0 (L) qe79°0 (81) CI (C0) Ce (91) 96 (Se) GPST'0 (6p) 9P'€ (O€) 94LIr (17) 9417 (LI) 9L°0 (1) 99°L pouinquy (L) 8680 (IT) ®82°0 Gia aa (OT) TE (ET) 96 (17) 990°0 (IL) PO'T (Le) PTSI (LE) 99T (LZ) 90 (1) 99°L (8) 94€L°0 (8) 979°0 (C1) vi (Ol) re (8) €¢ (ST) B8T°0 (ET) 86'S (17) BrSL (17) B76 (97) P6'T (¢) PEL (% UONPLIVA JO JUDIOIJJOOS) UBdT (11) 999°0 (C1) 48S'0 (OT) €1 (6) €€ (S) o¢ (po) GeSTO (pI) Gets (CT) 99TH (17) 997E (eS) qo"] (CT) GS°L pouing -(.U19/3) dd (uu9/3) (IM %) KeID (41M %) YIS (1M 2%) pues (1M %) N (IM %)O (Sy/slU) > (Sy/8W) d (W/SP) DA Hd aqqeiey, ‘popnypour jou SPM UOJ WU T< SY} JO SUINIOA 1vY} 1dd9x9 D9AOQK SB ORS DY} Poe[Ndyed ‘AjIsuSp YIN, “UOTIRIF WU T< SY} JO SLUNJOA POpNySUT YOIYM ‘OUUINJTOA UMOUY JO 3109 B UI poulejUuos (WUT T>) [IOS JO ssvUT AIP-UDAO dU} SB Po}e[NdTRd ‘AyIsUap YIN, “AWANINPUOD [RILINSI[A_ “SUBIU [BUONIPpe IY} JO UOISNPOUT dy} AQ JATIVAIOSUOD IIOUT IPLUT SEM 1S9} JOY-1SOd dy} VY} SSpPo[MOUYIV OM YSNOY}L “SOURSTIUBIS [BIISO]OI dALY ABU SOZIS JOdJJa JY} ASN’IOG SUIT dU} JO [[e JUdSaId SAA “UOTIOVIOJUT JUBOTFIUSIS ATTRINSNLIS B SABY JOU PIP SITQRLIVA JY} JO SUIOS YSNOY} UdAD ‘poyUasaid dIv S[OAI] BISOAOIU x UNG [[e 1OJ suvaut sy_L, “Adoued (Y1asIpIyIS 10 YIDIIDG) VINX MOPG = DONA pue ‘AdOuevd DIDJUAPIA] VIAADT MOTIQ = LT ‘ooedsidjul = [NI ‘SUONeIAIIQGe IISOIOIPY “(QSAL S$ Ad SWAON YW AbM-2U0) SQ’) > d 18 JAJIP Sla}19] pareys NoYIM suvspy “LYASAd JAVLOW “VAYY NOILVAYASNOD) TVNOILVN NOANVD MOOY Gay “VAYV GANYNANA) GNV GANUN V NO SALISOWIPL ONOWY SAILYadOUd MOS “] ATIVE 146 Burned Seeds/m? FIG. 6. MADRONO [Vol. 56 Unburned 0 Control @ Heat aT: YUC Microsite Emergence density of Bromus rubens among treatments performed on 0—5 cm soil seed bank samples from a burned and unburned area, Red Rock Canyon National Conservation Area, Mojave Desert. Error bars are one standard deviation. Means without shared letters differ at P < 0.05 and were compared using Tukey’s test. National Park in the southwestern Mojave Desert, where 80-84% of Yucca schidigera resprouted after fire. The relatively low (5%) proportion of resprouting for Larrea tridentata that we recorded is lower than Muinnich (1995: 17%) and two studies in the Sonoran Desert (Rogers and Steele 1980: 23%; McLaughlin and Bowers 1982: 37%). This proportion is similar, however, to Brown and Miunnich’s (1986) 3% resprout frequency at another site in the Sonoran Desert. Differences in resprouting frequency among studies could be related to differences in fire intensity, site factors (e.g., soils, prevalence of herbivory), or climate (O’Leary and Minnich 1981; Gibson et al. 2004). Our study site also was situated near the upper elevational limit for L. tridentata, possibly limiting resprouting com- pared to other studies. Relative to the unburned area, fire converted a late-successional shrub community dominated by high cover of Coleogyne ramosissima to a community with the perennial cover component dominated by forbs. Dominant native perennials colonizing the burn, such as Baileya multiradiata, Sphaeralcea ambigua, Eriogonum inflatum, and Dasyochloa pulchella, have previously been re- ported as colonizers after fire or other distur- bances (e.g., road abandonment, land clearing). For example, B. mu/tiradiata exhibited a density 1.7 times greater than the next highest species (among 35 species) one year after fire in the upper Sonoran Desert (Wilson et al. 1995). Similarly, S. ambigua, the most abundant perennial on the burn in our study, had the third highest density among all burn colonizers in Wilson et al.’s (1995) study. Brooks and Matchett (2003) found that these two species also had greater cover on burned than unburned areas in the Mojave Desert. Fifty to seventy years after agricultural field abandonment at 20 sites in the eastern Mojave Desert, Carpenter et al. (1986) found that cover of S. ambigua was three times greater on abandoned fields than off field. Dasyochloa pulchella and E. inflatum also had greater cover on disturbed than undisturbed areas in Mojave Desert studies of succession on a cleared borrow pit (Vasek 1983) and a pipeline corridor (Abella et al. 2007). These observations suggest that the major native perennial colonizers of the burn in our study are early colonizers of several distur- bance types. Colonization Windows and Revegetation Approaches Brooks (2002) proposed that reduced Bromus — rubens immediately following fire may offer a_ colonization window for actively revegetating native species on desert burns. Although the effectiveness of this window may depend on> precipitation, we found that several native | perennials colonized the burn even during the | period of below-average precipitation character- | izing our study. Despite this colonization, total | live native plant cover remained more than seven | times lower on the burned than the unburned | area. Although active revegetation (e.g., seeding) | can be expensive and prone to failure (Lovich and | Bainbridge 1999), managers may wish to attempt | revegetation of desert burns for several reasons. | Revegetation may improve aesthetics, reduce — fugitive dust, increase animal forage, and provide > competition with exotic plants (DeFalco et al. | 2007). Based on several studies, natural reestab- | lishment of late-successional species such as Coleogyne ramosissima or Larrea tridentata will | require long time periods, often longer than 30 yr | and possibly longer than 100 yr (Callison et al. | 1985; Lei 1999; Lovich and Bainbridge 1999). A- potential concern about reestablishing these late- | 2009] successional shrubs is that B. rubens is most abundant below their canopies, facilitating fuel accumulation (Brooks 2002). Therefore, reestab- lishing these shrubs may need to occur in combination with B. rubens control treatments. A complementary approach to test, in addition to revegetating with late-successional species, could be augmenting establishment of early colonizing native species (e.g., Baileya mutltiradiata and Sphaeralcea ambigua). In a Mojave Desert seeding study, B. multiradiata established at a density of 3 plants/m°’, second most among 12 seeded species (Walker and Powell 1999). If native species could be established in a post-fire colonization window, this early successional vegetation type can persist for more than 30 yr (Carpenter et al. 1986; Abella et al. 2007). Although these species are not thought to form the fertile islands most conducive to B. rubens establishment, it is unclear whether this vegeta- tion type can depress B. rubens resurgence. In combination with experiments, longer term mon- itoring of natural post-fire colonization patterns may provide insight into these types of practical questions. However, short-term data on initial colonization patterns such as provided by this study also are crucial, as managers often must plan revegetation activities soon after a fire. ACKNOWLEDGMENTS We thank the Bureau of Land Management (South- ern Nevada District) for funding the study through a cooperative agreement with the University of Nevada Las Vegas (UNLV); Jill Craig (UNLV) for help with fieldwork; the UNLV research greenhouse for provid- ing space for the seed bank experiment; Cheryl Vanier (UNLV) for performing the statistical analysis of the seed bank experiment; and Sharon Altman (UNLV) and two anonymous reviewers for providing helpful comments on the manuscript. LITERATURE CITED ABELLA, S. R., A. C. NEWTON, AND D. N. BANGLE. 2007. Plant succession in the eastern Mojave Desert: an example from Lake Mead National Recreation Area, southern Nevada. Crossosoma 33:45—54. AMACHER, M. C., K. P. O’NEILL, R. DRESBACH, AND C. PALMER. 2003. Forest inventory and analysis manual of soil analysis methods. U.S. Department of Agriculture, Forest Service. Website http://nrs. fs.fed.us/fia/topics/soils/documents/FIA.P3.Soils. Lab.Manual.2003.pdf. [accessed 23 November 2009]. BARTHA, S., S. J. MEINERS, S. T. A. PICKETT, AND M. L. CADENASSO. 2003. Plant colonization windows in a mesic old field succession. Applied Vegetation Science 6:205—212. BEATLEY, J. C. 1966. Ecological status of introduced brome grasses (Bromus spp.) in desert vegetation of southern Nevada. Ecology 47:548—554. Brooks, M. L. 1999. Alien annual grasses and fire in the Mojave Desert. Madrono 46:13—19. ABELLA ET AL.: POST-FIRE PLANT ESTABLISHMENT 147 . 2002. Peak fire temperatures and effects on annual plants in the Mojave Desert. Ecological Applications 12:1088—1102. AND J. R. MATCHETT. 2003. Plant community patterns in unburned and burned blackbrush (Coleogyne ramosissina Torr.) shrublands in the Mojave Desert. Western North American Natural- ist 63:283—298. AND 2006. Spatial and temporal patterns of wildfires in the Mojave Desert, 1980— 2004. Journal of Arid Environments 67:148—164. BROWN, D. E. AND R. A. MINNICH. 1986. Fire and changes in creosote bush scrub of the western Sonoran Desert, California. American Midland Naturalist 116:411—422. CALLISON, J., J. D. BROTHERSON, AND J. E. BOWNS. 1985. The effects of fire on the blackbrush (Coleogyne ramosissima) community of southwest- ern Utah. Journal of Range Management 38:535-538. CARPENTER, D. E., M. G. BARBOUR, AND C. J. BAHRE. 1986. Old field succession in Mojave Desert scrub. Madrono 33:111—122. CAVE, G. H. AND D. T. PATTEN. 1984. Short-term vegetation responses to fire in the upper Sonoran Desert. Journal of Range Management 37:491—496. D’ANTONIO, C. M. AND P. M. VITOUSEK. 1992. Biological invasions by exotic grasses, the grass/ fire cycle, and global change. Annual Review of Ecology and Systematics 3:63—87. DEFALCcOoO, L. A., G. C. J. FERNANDEZ, AND R. S. NOWAK. 2007. Variation in the establishment of a non-native annual grass influences competitive interactions with Mojave Desert perennials. Bio- logical Invasions 9:293—307. DUFRENE, M. AND P. LEGENDRE. 1997. Species assemblages and indicator species: the need for a flexible asymmetrical approach. Ecological Mono- graphs 67:345—366. ESQUE, T. C. AND C. R. SCHWALBE. 2002. Alien annual grasses and their relationships to fire and biotic change in Sonoran desertscrub. Pp. 165—194 in B. Tellman (ed.), Invasive exotic species in the Sonoran region. University of Arizona Press and the Arizona-Sonora Desert Museum, Tucson, AZ. FLEMATTI, G. R., E. L. GHISALBERTI, K. W. DIXON, AND R. D. TRENGOVE. 2004. A compound from smoke that promotes seed germination. Science 3057977. GIBSON, A. C., M. R. SHARIFI, AND P. W. RUNDEL. 2004. Resprout characteristics of creosote bush (Larrea tridentata) when subjected to repeated vehicle damage. Journal of Arid Environments 57:41 1-429. Guo, Q., P. W. RUNDEL, AND D. W. GOODALL. 1998. Horizontal and vertical distribution of desert seed banks: patterns, causes, and implications. Journal of Arid Environments 38:465—478. JAMES, J. J.. M. A. CAIRD, R. E. DRENOVSKY, AND R. L. SHELEY. 2006. Influence of resource pulses and perennial neighbors on the establishment of an invasive annual grass in the Mojave Desert. Journal of Arid Environments 67:528—534. LATO, L. J. 2006. Soil survey of Clark County area, Nevada. U.S. Department of Agriculture, Natural Resources Conservation Service. U.S. Government Printing Office, Washington, D.C. 148 MADRONO LEI, S. A. 1999. Postfire woody vegetation recovery and soil properties in blackbrush (Coleogyne ramosis- sima) shrubland ecotones. Journal of the Arizona- Nevada Academy of Science 32:105—115. . 2001. Postfire seed bank and soil conditions in a blackbrush (Coleogyne ramosissima Torr.) shrub- land. Bulletin of the Southern California Academy of Sciences 100:100—108. LovicH, J. E. AND D. BAINBRIDGE. 1999. Anthropo- genic degradation of the southern California desert ecosystem and prospects for natural recovery and restoration. Environmental Management 24:309-326. McCuneg, B. AND M. J. MEFFORD. 1999. PC-ORD: multivariate analysis of ecological data. Version 4. User’s guide. MjM Software Design, Gleneden Beach, OR. MCLAUGHLIN, S. P. AND J. E. BOWERS. 1982. Effects of wildfire on a Sonoran Desert plant community. Ecology 63:246—248. MINNICH, R. A. 1995. Wildland fire and early postfire succession in Joshua tree woodland and blackbrush scrub of the Mojave Desert of California. San Bernardino County Museum Association Quarter- ly 42:99-106. Noy-MeirR, I. 1973. Desert ecosystems: environment and producers. Annual Review of Ecology and Systematics 4:25—51. O’LEARY, J. F. AND R. A. MINNICH. 1981. Postfire recovery of the creosote bush scrub vegetation in the western Colorado Desert. Madrono 28:61—66. PATTEN, D. T. AND G. H. CAVE. 1984. Fire temper- atures and physical characteristics of a controlled burn in the upper Sonoran Desert. Journal of Range Management 37:277—280. PEET, R. K., T. R. WENTWORTH, AND P. S. WHITE. 1998. A flexible, multipurpose method for record- ing vegetation composition and structure. Castanea 63:262-274. RAISON, R. J. 1979. Modification of the soil environ- ment by vegetation fires, with particular reference to nitrogen transformations: a review. Plant and Soil 51:73-108. ROGERS, G. F. AND J. STEELE. 1980. Sonoran Desert fire ecology. Pp. 15-19 in M. A. Stokes and J. H. [Vol. 56 Dieterich (tech. coords.), Proceedings of the fire history workshop. General Technical Report RM- 81. U.S. Department of Agriculture, Rocky Moun- tain Forest and Range Experiment Station, Fort Collins, CO. SAS INSTITUTE. 1999. SAS/STAT user’s guide. SAS Institute, Inc., Cary, NC. SAS INSTITUTE. 2004, JMP user’s guide. SAS Institute, Inc., Gaty, NC: UNITED STATES DEPARTMENT OF AGRICULTURE (USDA), NATURAL RESOURCES CONSERVATION SERVICE. 2007. The PLANTS database. National Plant Data Center, Baton Rouge, LA. Website http:// plants.usda.gov. [accessed 23 November 2009]. VAN MANTGEM, P., M. SCHWARTZ, AND M. KEIFER. 2001. Monitoring fire effects for managed burns and wildfires: coming to terms with pseudoreplica- tion. Natural Areas Journal 21:266—273. VASEK, F. C. 1983. Plant succession in the Mojave Desert. Crossosoma 9:]—23. WALKER, L. R. AND E. A. POWELL. 1999. Effects of seeding on road revegetation in the Mojave Desert, southern Nevada. Ecological Restoration 17:150—-155. WARR, S. J., K. THOMPSON, AND M. KENT. 1993. Seed banks as a neglected area of biogeographic research: a review of literature and sampling techniques. Progress in Physical Geography 17:329-347. WILSON, R. C., M. G. NAROG, A. L. KOONCE, AND AND B. M. CORCORAN. 1995. Postfire regeneration in Arizona’s giant saguaro shrub community. Pp. 424-431 in L. H. DeBano, P. H. Ffolliott, A. Ortega-Rubio, G. J. Gottfried, R. H. Hamre, and C. B. Edminster (tech. coords.), Biodiversity and management of the madrean archipelago: the sky islands of southwestern United States and northwestern Mexico. General Technical Report RM-GTR-264. U.S. Department of Agriculture, Forest Service, Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO. ZIMMERMAN, G. M., H. GOETZ, AND P. W. MIELKE. 1985. Use of an improved statistical method for group comparisons to study effects of prairie fire. Ecology 66:606—611. MADRONO, Vol. 56, No. 3, pp. 149-154, 2009 INTERACTIVE EFFECTS OF SIMULATED HERBIVORY AND INTERSPECIFIC COMPETITION ON SURVIVAL OF BLACKBRUSH (COLEOGYNE RAMOSISSIMA) SEEDLINGS SIMON A. LEI Department of Educational Psychology, University of Nevada, Las Vegas, 4505 Maryland Parkway, Las Vegas, NV 89154-3003 leis2@unlv.nevada.edu ABSTRACT I investigated the individual and joint effects of simulated herbivory and interspecific competition on survival of Coleogyne ramosissima Torr. (blackbrush) seedlings. Seeds of C. ramosissima and Bromus rubens L. (red brome grass) were collected at mid-elevations (1220 to 1770 m) of the Spring Mountains in southern Nevada. A pot trial experiment was conducted for six months (27 wk) in a controlled environmental greenhouse. This trial experiment, consisting of a 2 < 2 factorial design with simulated rodent herbivory and interspecific competition with B. rubens as the main effects, resulted in four treatments. Herbivory on C. ramosissima by heteromyid and non-heteromyid species of rodents was simulated by clipping the top 3 cm of young shoots. Significant interaction was detected between herbivory and competition for C. ramosissima seedling survival. When herbivory and competition were examined independently, both factors had significant adverse effects on C. ramosissima seedling survival, with the former having greater negative effect than the latter. Results of this study suggest that simulated rodent herbivory and B. rubens competition limited survival independently, and that a combination of herbivory and competition caused synergistic reductions in C. ramosissima seedling survival. Understanding the role of herbivory and plant competition in reducing survivorship of seedlings is crucial to the management and regeneration of C. ramosissima shrublands in the Mojave Desert. Key Words: Bromus rubens, Coleogyne ramosissima, competition, herbivory, seedling survival, southern Nevada, Spring Mountains. Herbivory and competition from plant neigh- bors are two major biotic factors that determine the growth, survival, and reproduction of plant individuals, and subsequently the abundance of plant populations (Harper 1977; Crawley 1983; Gurevitch et al. 2000; Hamback and Beckerman 2003). Field and greenhouse studies further suggest that manipulation of herbivore and plant neighbor densities frequently have synergistic effects on plant performance (Hamback and Beckerman 2003). Plants growing at high density can be water and nutrient stressed, and therefore easily succumbs to herbivores (Mattson and Haack 1987; Karban et al. 1989; Maron 2001). As a result, plants stressed by overcrowding that are attacked by herbivores often suffer greater reduction in survival compared to plants dam- aged by herbivores growing free from competi- tors (Parker and Salzman 1985; Maron 2001). Seedlings incur considerably higher mortality than juveniles or adults in most plant popula- tions, with both herbivory and plant competition proposed as being important mechanisms causing seedling death (Maron 1997). Previous research studies have suggested that single biotic factors are usually inadequate and unlikely to fully explain community structure (Pickett et al. 1987; Myster and McCarthy 1989). Herbivory can alter competitive interactions, while compe- tition can alter the ability of seedlings to respond to herbivory (Hjalten et al., 1993; Maron 2001). Once plant tissue has been lost to herbivory, seedlings may suffer from reduced competitive abilities (Hendrix 1988). This effect of herbivory may subsequently lead to increased seedling mortality or decreased seedling growth (Meiners and Handle 2000). Herbivory and competition may function in synergistic ways to affect plant survival, with seedlings more likely than adults to suffer this negative interaction (Crawley 1983; Meiners and Handel 2000). Coleogyne ramosissima Torr. (blackbrush) ts a shrub from the Rosaceae plant family that grows in the transition zone between warm and cold deserts of southern California, southern Utah, southern Nevada, northwestern Arizona, and southwestern Colorado of the United States (Meyer and Pendleton 2005). It is found at 760 to 1980 m in elevations, ranging from 0.3 to 1.2 m in height (Lei 1995). Coleogyne ramosissima often forms dense, nearly monospecific stands at mid- elevations (1200 to 1500 m) in the Mojave Desert of southern Nevada, bounded by Larrea triden- tata-Ambrosia dumosa (creosote bush-white bur- sage) shrublands below and by Pinus monophylla- Junipers osteosperma (pinyon pine-Utah juniper) woodlands above on desert mountain slopes (Le! 1995: Lei and Walker 1997). 150 In southern Utah and Nevada, Coleogyne ramosissima seedling mortality due to herbivory most often involves rodents, consisting of both heteromyid and non-heteromyid species (Meyer and Pendleton 2005). Herbivory from rodents was primarily at the cotyledon and young seedling stages. In a field experiment conducted in southern Utah, rodent herbivory, including Dipodomys spp. (kangaroo rats) has an adverse impact on C. ramosissima seedlings (Meyer and Pendleton 2005). Dipodomys microps (chisel- toothed kangaroo rat), D. merriami (Merriam’s kangaroo rat), and D. deserti (desert kangaroo rat) were commonly found at C. ramosissima shrublands of southern Nevada (Beatley 1976). Establishment success of C. ramosissima seedlings was a consequence of the interplay of rodent herbivory and abiotic factors (Meyer and Pen- dleton 2005). Herbivory is a factor, but drought was the primary factor, resulting in a combina- tion of low precipitation, low soil moisture, as well as high air and soil temperatures (Meyer and Pendleton 2005). Interspecific competition occurs within shrub- lands between shrubs and annual grasses that differ in life history traits, including life cycles and growth patterns (Florentine and Fox 2003). Annual grasses can use soil moisture and mineral nutrients faster than perennial shrubs (Jackson and Roy 1986). The intensity of interspecific competition with Bromus rubens L. (red brome grass) at different densities on survival of C. ramosissima seedlings has been observed using greenhouse experiments (Lei 2006). In general, C. ramosissima shrub seedlings do not compete well with adjacent B. rubens when seedlings emerged simultaneously with B. rubens (Lei 2006). In mixed cultures of two species involving C. ramosissima and B. rubens plants, the survivor- ship and growth characteristics of shrub species are heavily dependent on the density and time of emergence of adjacent herbaceous species (Lei 2006). Bromus rubens was introduced into the western United States from southern Europe during the mid-nineteenths century, but did not spread into the Mojave Desert until the early twentieth century (Burcham 1975; Hunter 1991; Lei 2006). Most likely the introduction of B. rubens with low forage value was unintentional (Newman 1991). Bromus rubens invaded springs, roadsides, and disturbed areas of the Mojave Desert during 1920's; it was not common, however, until after 1950 (Hunter 1991; Lei 2006). Bromus is common in some locations around 1575 to 1970 m in elevation at the Nevada Test Site of southern Nevada (Hunter 1991; Lei 2006). Bromus rubens is often a codominant or subdominant herba- ceous species in C. ramosissima communities of southern Nevada and southern California (Beat- ley 1966; Newman 1991). MADRONO [Vol. 56 The individual effects of rodent herbivory and B. rubens competition on C. ramosissima survival have been documented (Meyer and Pendleton 2005; Lei 2006, 2008). Nevertheless, the interac- tive effects of herbivory and competition on survival of young C. ramosissima seedlings remain poorly understood. Do herbivory and competition have independent or synergistic effect on C. ramosissima seedling survival? In this study, I report results of a greenhouse experiment designed to assess the extent to which simulated rodent herbivory through clipping and interspecific competition with B. rubens individ- ually and jointly affected C. ramosissima seedling survival. METHODS Seed Collection and Germination Coleogyne ramosissina and B. rubens seeds were collected during July and August 2007 in the Spring Mountains (36°N 09’W, II5°N 43’W; elevation 1220 to 1770 m) of southern Nevada. From casual observations, flattened, wrinkled, punctured, or empty (potentially inviable) seeds of both species were discarded in the field. A total of 500 C. ramosissima and 1000 B. rubens seeds, along with their corresponding field soils from the upper 10 cm, were collected initially from C. ramosissima-dominated shrublands. Approxi- mately half of all seeds were randomly collected from the soil surface, while the remaining half was randomly collected from 312 C. ramosissima — and 181 B. rubens plants. Potentially viable seeds were transported to a biology laboratory at Nevada State College (NSC) for storage. Seeds were placed at room temperature (22°C) for 4-5 mo prior to conducting the pot trial | experiment. Coleogyne ramosissima and B. rubens seeds were stored at 4°C for two and six weeks, respectively, in the dark (dry-chilling) to obtain | | maximum germination (Newman 1991; Pendle- | ton and Meyer 2004; Lei 2006). During the germination period, 20 seeds of the same species were placed between two layers of germination blotters moistened with tap water inside a 10-cm | diameter Petri dish. Petri dishes were placed in | transparent zip-loc bags in a cool chamber, and water was added as necessary to maintain » moisture of blotters during incubation. Coleogyne | ramosissima seeds were incubated at 4°C, while B. rubens seeds were incubated at 22°C in the dark. Germination occurred between 2-3 wk for C. ramosissima, and occurred within two weeks for B. rubens. Experimental Design and Statistical Analysis One greenhouse experiment, consisting of a 2 X 2 factorial design with simulated rodent | a 2009] herbivory through clipping and competition with B. rubens as the main effects, resulted in four treatment groups: 1) herbivory, 2) compe- tition, 3) presence of herbivory and competition, and 4) absence of herbivory and competition (control). This experiment was designed to assess the extent to which simulated rodent herbivory and interspecific competition with B. rubens influenced survival of young C. ramosissima seedlings. The clipping and competitive treat- ments were designed to simulate biotic stresses confronting C. ramosissima seedlings in nature. The density of B. rubens (4 individuals) in the pots was moderate, and was within the range of natural variation in the field. On the average, there were fewer than three C. ramosissima seedlings per square meter in the field (Lei personal observation). Clipping, in some instanc- es, may not fully duplicate natural herbivory (Fowler and Rausher 1985), but is probably a good simulation of the feeding damage caused by Dipodymis spp. In the field, Dipodymis spp. commonly consume the aboveground shoots, leaving only a small piece of the main stem projecting above ground. The appearance of a defoliated and a clipped seedling were therefore similar (Fowler and Rausher 1985). The timing and severity of the clipping treatment corre- sponded to the timing and severity of defoliation in nature (Lei personal observation). Five-week old C. ramosissima seedlings and three-week old B. rubens seedlings were planted into 65 mm diameter, 350 mm tall, cone-shaped containers (pots) in the greenhouse of southern Nevada. There were 120 individuals of C. ramosissima in each of the four treatment groups for a total of 480 seedlings. Each pot contained only one C. ramosissima seedling for a total of 480 pots. Each pot also consisted of one-third of perlite and two-thirds of natural field soil, thoroughly mixed, without adding fertilizers in order to maintain a low soil fertility level. Perlite was used to improve aeration and drainage (Lei 2004, 2006). Prior to planting, C. ramosissima seedlings were allocated by size to various experimental treatments, so that no single treatment had disproportionally large or small seedlings. Pots were very lightly moistened with tap water for 27 wk (six months), simulating natural changes in soil moisture in the field from late fall through late spring season. Initially, seedlings were watered every 4-5 d for the first three weeks, and thereafter watered every 7-9 d until the end of the pot trial experiment. During the course of study, light and air temperature regime resembled winter and spring weather conditions in C. ramosissima shrublands at mid-elevations of southern Nevada. For instance, mean monthly minimum and maximum air temperatures in winter were —2.1°C and 12.4°C, respectively, LEI: HERBIVORY, COMPETITION AND COLEOGYNE SEEDLING SURVIVAL IS and in spring were 7.9°C and 22.7°C, respectively. Mean photoperiods were approximately 10:55 h light: 13:05 h dark in winter and approximately 12:50 h light: 11:10 h dark in spring (Lei personal observations). The experiment commenced in early December 2006. Initially, 480 C. ramosissima seedlings were planted in the center of pots, with only one seedling per pot. At this time, half of all seedlings (240 pots) were randomly selected to experience competition with B. rubens, while the remaining half had single C. ramosissima seedlings only. In each of the 240 pots selected for the competitive treatment, four individuals of B. rubens were planted on the periphery with a single C. ramosissima in the center. Twelve weeks later (early March 2007), exactly half of 240 pots (120 pots) containing B. rubens were randomly select- ed to simulate herbivory through clipping (her- bivory-competition interaction group), while the remaining half had C. ramosissima_ seedlings continually grown together with B. rubens (competitive treatment group). Also, in early March 2007, exactly half of the 240 pots (120 pots) containing single C. ramo- sissima seedlings were randomly selected to simulate herbivory (herbivory treatment group). The remaining 120 pots containing C. ramosis- sima seedlings without exposure to herbivory and competition served as a control population (control group), and were used to compare individual and interactive effects of herbivory and competition on seedling survival. To determine whether there were interactive effects of herbivory and competition on C. ramosissima seedling survival, the top 3 cm of young shoots were clipped only once. Young shoots were removed with scissors. Scissors were sterilized in 70% isopropyl alcohol between clippings in order to minimize pathogen spread among C. ramosissima seedlings. Survivorship of clipped seedlings was then observed and recorded 15 wk later. The entire greenhouse experiment was six months in duration, and I terminated this experiment in early June 2007. Survivorship of C. ramosissima seedlings was assessed based on the presence of green leaves and growing shoots. The status of all experimen- tal treatments was classified into two categories: alive or dead. Surviving seedlings exhibited net growth, while dead seedlings exhibited no growth with brittle, dark-brown leaves several weeks after planting into a pot. The proportion of C. ramosissima seedlings that survived six months after simulated rodent herbivory and B. rubens competition was com- puted. Log-linear analysis (Analytical Software 2007) was performed to assess individual and interactive effects of herbivory and competition on seedling survival. Statistical significance was determined at P = 0.05. 152 MADRONO [Vol. 56 TABLE 1. SURVIVORSHIP OF YOUNG C. RAMOSISSIMA SEEDLINGS, ALONG WITH RESULTS OF LOG-LINEAR ANALYSIS (CHI-SQUARE) SHOWING SIMULATED RODENT HERBIVORY THROUGH CLIPPING AND INTERSPECIFIC COMPETITION WITH B. RUBENS USING A GREENHOUSE EXPERIMENT (N = 120 PER TREATMENT GROUP IN EACH VARIABLE). Survivorship was based on six month (27 wk) for four treatments. Statistical significance is determined at P= 0105, Variable Survival proportion (%) Chi-square P-value Control 130 Competition treatment 65.0 4.62 0.0317 Herbivory treatment fae. 3137 <(Q.0001 Herbivory by competition interaction 21.6 45.88 <0.0001 RESULTS Mortality of C. ramosissima seedlings occurred in the absence of simulated rodent herbivory through clipping and interspecific competition with B. rubens (control group) (Table 1). How- ever, C. ramosissima seedlings in the control group had a substantially higher survivorship than those that grew with B. rubens and/or those that experienced simulated herbivory (Table 1). When herbivory and competition were examined independently, C. ramosissima seedling survivor- ship was significantly reduced in herbivory pots compared to herbivory-free pots (Table 1). Sim- ilarly, seedling survivorship was. significantly reduced in competition pots compared to com- petition-free pots (Table 1). Herbivory had a larger adverse impact on C. ramosissima seedling survival than did interspecific competition, de- spite the fact that seedlings were tall and robust before experiencing simulated herbivory. There was a significant herbivory-by-competi- tion interaction for C. ramosissima_ seedling survival, with the lowest survivorship occurring in the presence of herbivory and competition and the highest in the control group (Table 1). The presence of B. rubens significantly decreased the ability of C. rubens seedlings to survive following simulated herbivory (Table 1). Seedling survivor- ship was only 21.6% after 15 wk of clipping (Table 1). Regardless of treatments, all seedlings without true photosynthetic leaves died, while the majority of resprouting seedlings and seedlings with true leaves survived. DISCUSSION This study demonstrated that simulated rodent herbivory and interspecific competition with B. rubens independently decreased the probability of C. ramosissima seedling survival, with the former having a larger adverse effect than the latter. The interactive effect of herbivory and competition on seedling survival was best described as synergis- tic. Synergistic effects on C. ramosissima seedling survival were more important as they showed effects beyond simple addition of herbivory and competition. Growth and survival often decline tremen- dously as multiple stresses are imposed on plants, especially seedlings (Saunders and Puettman 1999). The ability of herbivory and competition to determine seedling survival is well-established (Meiners and Handel 2000). Seedling recovery from herbivory is hypothesized to decrease with increasing plant competition (Saunders and Puettmann 1999). Field and greenhouse experi- ments further suggest that manipulation of herbivore and plant neighbor densities frequently have synergistic effects on plant growth and survival (Hanback and Beckerman 2003). In this study, survivorship of C. ramosissima seedlings showed an interaction between the effects of simulated herbivory and B. rubens competition. The effects of a given level of herbivory are expected to depend on the competitive environ- ment of the plants experiencing the herbivory (Fowler and Rausher 1985). Herbivory largely defines C. ramosissima seed- _ ling survival. Herbivory determines if plant competition can occur because it acts as a filter for future competitive interactions (Myster and © McCarthy 1989). Competition may subsequent © affect seedling growth if surviving an initial — herbivore attack (Myster and McCarthy1989). | In this study, a number of C. ramosissima | seedlings did not survive simulated rodent herbivory; those seedlings that survived generally | suffered from reduced competitive abilities. Nev- | ertheless, some C. ramosissima seedlings over- | came the long-term effects of herbivory by | resprouting in a field experiment conducted in | southern Utah (Meyer and Pendleton 2005). | Results of this study depart from the tradi- | tional ‘‘single mechanism” view of community | dynamics and succession (Connell and Slatyer | 1977), and are consistent with a multi-model view | of herbivory and plant competition acting in | concert (Menge and Sutherland 1976; Myster and | McCarthy 1989). Seedlings are often the demo- | graphic stage that is most susceptible to mortality | from biotic stress compared to juvenile and adult | stages (Maron 1997; Meiners and Handel 2000). | In this study, the effect of herbivore attack on survivorship depended heavily on whether seed- | lings were experiencing interspecific competition, | 2009] and the effects of competition on survivorship also depended heavily on whether seedlings were experiencing herbivory attack. Thus, my results caution against attempts to assign a primary role to either herbivory or plant competition alone. Research studies that focus on herbivory or plant competition in isolation can result in incorrect prediction of community or population trajecto- ries (Meiners and Handel 2000). Such studies may grossly underestimate the importance of herbivory and competition in structuring plant communities, thus leading to inappropriate man- agement recommendations (Meiners and Handel 2000). This greenhouse study largely reflected the moderate levels of rodent herbivory and inter- specific competition of C. ramosissima seedlings with B. rubens in the field. A strong relationship exists between experimental set-up and _ field conditions because rodents typically browsed the top few centimeters of young C. ramosissima shoots during early first spring season (Meyer and Pendleton 2005), leaving only a small piece of the main stem projecting above ground. Conse- quently, there was a substantial drop in seedling number from mid-March through mid-April due to rodent predation events (Meyer and Pendleton 2005). Coleogyne ramosissima seedlings were simultaneously competing with B. rubens for limited resources. Previous research studies have shown that B. rubens substantially limited C. ramosissima seed germination and seedling survi- vorship (Meyer and Pendleton 2005; Lei 2006). In the absence of B. rubens competition, C. ramo- sissima seedling survivorship increases substan- tially (Lei 2006). Therefore, the greenhouse experiment conducted in this study likely cap- tured important herbivory and competition mechanisms prevalent 1n the field. In addition, from a perspective of C. ramosis- sima shrubland management and regeneration, the invasiveness of B. rubens is relatively new to the Mojave Desert. Coleogyne ramosissima re- production has always been episodic. Mast fruiting is an evolutionary strategy to reduce the impact of seed predation and herbivory. The new pressures of B. rubens competition, along with the concurrent dramatic increase in fire frequency present a new threat to the persistence of the C. ramosissima shrubland. This experiment demon- strates that B. rubens competition may further reduce the frequency of episodic recruitment of C. ramosissima in the Mojave Desert. ACKNOWLEDGMENTS I gratefully acknowledge Steven Lei and David Valenzuela for collecting seeds in the field and for setting up one experiment with four treatments in the greenhouse at the College of Southern Nevada (CSN). Prior to conducting experiments, seeds were stored and incubated at the Nevada State College (NSC). Steven LEI: HERBIVORY, COMPETITION AND COLEOGYNE SEEDLING SURVIVAL 153 Lei and two anonymous reviewers provided helpful comments on previous versions of this manuscript. LITERATURE CITED ANALYTICAL SOFTWARE. 2007. Statistix 8.0, an inter- active statistical program for microcomputers. Analytical Software, Tallahassee, FL. BEATLEY, J. C. 1966. Ecological status of introduced brome grass (Bromus spp.) in desert vegetation of southern Nevada. Ecology 47:548—554. —. 1976. Environments of kangaroo rats (Dipod- omys) and effects of environmental change on populations in southern Nevada. Journal of Mammalogy 57:67—93. BURCHAM, L. 1975. California rangeland. Department of Natural Resources, Sacramento, CA. CONNELL, J. H. AND R. O. SLATYER. 1977. Mecha- nisms of succession in natural communities and their role in community stability and organization. American Naturalist 11:1119—1144. CRAWLEY, M. J. 1983. Herbivory: the dynamics of animal-plant interactions. University of California Press, Berkeley, CA. FLORENTINE, S. L. AND J. E. Fox. 2003. Competition between Eucalyptus victrix seedlings and grass species. Ecological Research 18:25—39. FOWLER, N. L. AND M. D. RAUSHER. 1985. Joint effects of competitors and herbivores on growth and reproduction in Aristolochia reticulata. Ecolo- gy 66:1580—1587. GUREVITCH, J., J. A. MORRISON, AND L. V. HEDGES. 2000. The interaction between competition and predation: a meta-analysis of field experiments. American Naturalist 155:435—453. HAMBACK, P. A. AND A. P. BECKERMAN. 2003. Herbivory and plant resource competition: a review of two interacting interactions. Oikos 101:26—37. HARPER, J. L. 1977. Population biology of plants. Academic Press, San Diego, CA. HENDRIX, S. D. 1988. Herbivory and its impact on plant reproduction. Pp. 246—263 in J. L. Doust and L. L. Doust, (eds.), Plant reproductive ecology. Oxford University Press, New York, NY. HJALTEN, J., K. DANELL, AND L. ERICSON. 1993. Effects of simulated herbivory and _ intraspecific competition on the compensatory ability of birches. Ecology 74:1136—1142. HUNTER, R. 1991. Bromus invasions on the Nevada Test Site: Present status of B. rubens and B. tectorum with notes on their relationship to disturbance and altitude. Great Basin Naturalist 51:176—182. JACKSON, E. L. AND J. ROy. 1986. Growth patterns of Mediterranean annual and perennial grasses under simulated rainfall regimes of southern France and California. Oecologia Planatarium 7:191—212. KARBAN, R., A. K. BRODU, AND W. C. SCHNATHORST. 1989. Crowding and a plant’s ability to defend itself against herbivores and diseases. American Natu- ralist 134:749—760. S. A. 1995. A gradient analysis of Coleogyne communities in southern Nevada. Master’s Thesis, University of Nevada, Las Vegas, NV. ——. 2004. Intraspecific competition among black- brush (Coleogyne ramosissima) seedlings in an environmentally controlled glasshouse. Journal of the Arizona-Nevada Academy of Science 37:100—-104. LEI, 154 MADRONO . 2006. Interspecific competition between Coleo- gyne ramosissima seedlings and Bromus rubens. Southern California Academy of Sciences 105:76—-84. . 2008. Impacts of cotyledon removal on survival of blackbrush (Coleogyne ramosissima: Rosaceae) seedlings. Southern California Academy of Scienc- es 107:178-182. AND L. R. WALKER. 1997. Biotic and abiotic factors influencing the distribution of Coleogyne communities in southern Nevada. Great Basin Naturalist 57:163-171. MARON, J. L. 1997. Interspecific competition and insect herbivory reduce bush lupine (Lupinus arboreus) seedling survival. Oecologia 110:284—290. . 2001. Intraspecific competition and subterra- nean herbivory: individual and interactive effects on bush lupine. Oikos 92:178—186. MATTSON, W. J. AND R. A. HACCK. 1987. The role of drought in the outbreak of plant-eating insects. Bioscience 37:110—118. MEINERS, S. J. AND S. N. HANDEL. 2000. Additive and nonadditive effects of herbivory and competition on tree seedling mortality, growth, and allocation. American Journal of Botany 87:1821—1826. MENGER, B. A. AND J. P. SUTHERLAND. 1976. Species diversity gradients: Synthesis of the role of predation, competition, and temporal heterogene- ity. American Naturalist 110:351—369. [Vol. 56 MEYER, S. E. AND B. K. PENDLETON. 2005. Factors affecting seed germination and seedling establish- ment of a long-lived desert shrub (Coleogyne ramosissima: Rosaceae). Plant Ecology 178:171-187. MYSTER, R. W. AND B. C. MCCARTHY. 1989. Effects of herbivory and competition on survival of Carya tomentosa (Juglandaceae) seedlings. Oikos 56:145-148. NEWMAN, D. 1991. Element stewardship abstract for Bromus rubens. The Nature Conservancy, Arling- ton, VA. PARKER, M. A. AND A. G. SALMAN. 1985. Herbivore enclosure and competitor removal: effects on juvenile survivorship and growth in the shrub Gutierrezia microcephala. Journal of Ecology 73:903-913. PENDLETON, B. K. AND S. E. MEYER. 2004. Habitat- correlated variation in blackbrush (Coleogyne ramosissima: Rosaceae) seed germination response. Journal of Arid Environments 59:229—-243. PICKETT, S. T. A., S. L. COLLINS, AND J. J. ARMESTO. 1987. Model, mechanisms and pathways of succes- sion. Botanical Review 53:335-371. SAUNDERS, M. R. AND K. J. PUETTMANN. 1999. Effects of overstory and understory competition and simulated herbivory on growth and survival of white pine seedlings. Canadian Journal of Forestry Resources 29:536—546. MADRONO, Vol. 56, No. 3, pp. 155-167, 2009 DIVERSITY, REPRODUCTION, AND POTENTIAL FOR INVASIVENESS OF EUCALYPTUS IN CALIFORNIA MATT RITTER AND JENN YOST Biological Sciences Department, Cal Poly, 1 Grand Ave., San Luis Obispo, CA 93407 mritter@calpoly.edu ABSTRACT In the 150 years since their introduction to the state, species in the genus Eucalyptus have become the most common non-native trees in California. A clearer understanding of the ability of different species to reproduce in the state is important for how we monitor the ecological impact of these abundant non-native trees and for predicting possible future invasions. Here we present current data on the diversity of Eucalyptus in California, which species are spontaneously reproducing, or have the potential to do so, where they can be found, how they can be identified, and our analysis, based on herbarium and field observations, of the potential ecological impacts of various species in the locations where they have been introduced. We also present a new dichotomous identification key, and botanical drawings of all naturalized species. We discuss the degree to which factors such as life history traits, commonness of planting, and native range influence reproductive behaviors of different species. Key Words: Australia, California, Eucalyptus, invasive, key, naturalization, weed. The genus Eucalyptus L’Hér. (Myrtaceae) includes some of the most important solid timber and paper pulp forestry trees in the world (Doughty and Places 2000). They have also become the most abundant, ecologically success- ful, and controversial exotic trees in California. Species in the genus were first imported into the state as early as the mid 1850s, to be grown initially as horticultural oddities for the nursery trade, then later as a promising fiber source and possible savior of a forecasted timber drought (Butterfield 1935; Santos 2006). By 1880, a number of species, but primarily E. globulus Labill. (Fig. 1A), were being extensively planted for lumber, pilings and posts, fuel wood, medic- inal products, tannin, oil, windbreaks, and as street and park trees (Groenendaal 1983). As the California forestry and fuel economy evolved many of the thousands of hectares of Eucalyptus plantations remained uncut, and parts of the state are now bearing the ecological legacy of this vast unharvested crop. Since the time of its initial introduction into California, the genus was promoted by various agents including private landholders, commercial firms, and state and federal agencies. The planting craze that took place around the turn of the last century, the enthusiasm of certain individuals in high-profile state public office, and biomass fuel experimentation by the U.S. De- partment of Energy and the California Depart- ment of Forestry following the 1973 Arab oil embargo have all significantly helped propagate the genus in the state. The prevalence of eucalypts in California is more the result of large scale intentional plantings than it is the result of extensive naturalization. Needless to say, these species, mostly E. globulus, have become contro- versial in the regions where they are now conspicuous features of the landscape. There are many popular articles containing the most emotive writing, much of which is based loosely at best on scientific observation, regarding the various impacts of eucalypts on California’s landscapes and wildlife. They are admired as erosion control, wildlife habitat, and aesthetically valuable landscape and heritage trees or demon- ized as America’s largest, most fire prone, most dangerous, bird killing, weeds (Bulman 1988; Williams 2002). The genus, which contains more than 700 species, according to the most recent formal classification (Brooker 2000), is almost entirely endemic to the Australian continent with a small number of species occurring natively in the southern Philippines, New Guinea and parts of Indonesia (Williams and Woinarski 1997). Euca- lypts exhibit a great range of adaptation to different moisture conditions (Boland and Hall 1984), and they rival other large tree genera (1.e., Ficus, Pinus, Quercus) in having great diversity in mature tree size. Species range from small multi- stemmed shrubs (mallees), to some of the tallest and largest forest trees on earth. In fact, the putative tallest angiosperm in North and South America is a 75.05 m (246.2 ft) E. globulus off the coast of California, on Santa Cruz Island (Steve Sillette personal communication, Humboldt State University). A great number of species, representing the gamut of diversity in the genus, have been introduced into cultivation in California over the last 150 yr. The correct identification of Eucalyptus species in cultivation is often difficult, Pic. -£, Eucalyptus camaldulensis (Sanders 21982, CAS). C. Eucalyptus fastigata (Ritter 345, OBI). D. Eucalyptus citriodora (Eastwood s.n., CAS 45201). All size bars = 1 cm. but is essential for studying the potential of different species to become invasive. The bark, leaves, and reproductive structures are greatly varied and at times all need to be examined for accurate identification (Pryor 1976; Brooker and Kleinig 1996). Many species retain the dead bark year after year, giving rise to a trunk covered ina hard, weathered, outer layer (e.g., E. sideroxylon A. Cunn. ex Woolls), while others annually decorticate, resulting in a completely smooth trunk (e.g., E. citriodora Hook. = Corymbia citriodora (Hook.) K. D. Hill & L. A. S. Johnson). Eucalypts are definitively heterophyl- lous, with juvenile leaves that differ from adult MADRONO [Vol. 56 Hlustrations of fruits and buds of: A. Eucalyptus globulus subsp. globulus (Howell 32582, CAS). B. leaves in phyllotaxis, shape, petiolation, and glaucousness (Jacobs 1955). Juvenile leaves are commonly sessile, decussate, glaucous, oriented horizontally, discolorous (dorsiventral) and often cordate, orbicular, or ovate in shape, whereas adult leaves tend to be petiolate, alternate, glabrous, pendulous, lanceolate, and concolorous (isobilateral) (Coppen 2002). As in most genera, the defining aspects of | Eucalyptus are in the reproductive structures. The flowers of only a small number of species develop singly in leaf axils (e.g., E. globulus), while more commonly they develop in 3-, 7-, 9-, Il-, etc. flowered umbels (intact inflorescences always — FIG. 2. RITTER AND YOST: EUCALYPTUS IN CALIFORNIA Illustrations of fruits and buds of: A. Eucalyptus polyanthemos (Twisselmann 18559, CAS). B. Eucalyptus kitsoniana (Ritter 263, OBI). C. Eucalyptus conferruminata (McClintock s.n., CAS 994288). D. Eucalyptus pulchella (Kawahara S800, CAS). All size bars = | cm. have odd numbers of flowers) or heads. Individ- ual umbels may develop singly or paired in leaf axils (e.g., E. camaldulensis Dehnh., Fig. 1B and E. fastigata H. Deane & Maiden, Fig. 1C) or in branched axillary or terminal panicles (e.g., E. citriodora, Fig. 1D and E. polyanthemos Schauer, Fig. 2A). Individual flowers, which are often small, white, and inconspicuous in the tree crown, have either one or two bud caps (opercula) derived from the fused petals and/or sepals. In a number of species (mostly subgenus Symphyo- myrtus (Schauer) Brooker) the outer bud cap, derived from united sepals, sheds early in the development of the flower, leaving a diagnostic ring-like scar around the middle of the inner bud cap. The inner bud cap is shed at anthesis, exposing numerous spreading stamens. The inferior ovary is sunken in and fused to the hypanthium (invaginated pedicle) wall. After fertilization, when the stamens and the style fall from the flower, the ovary develops into a woody capsule with valves dehiscing at the top, allowing tiny, wind-dispersed seeds to be shed (Slee et al. 2006). There have been a number of past treatments of cultivated and naturalized eucalypts in Cali- fornia. In Eric Walther’s 1928 key to the species grown in California, he included 99 distinct species, known to be growing in the state at the time (not necessarily naturalized however), and made mention of over 100 others (Walther 1928). Three species, E. polyanthemos, E. globulus, and E. tereticornis Sm., were treated in Munz’s 1959 flora of California, and a number of other commonly planted species including E. siderox- ylon, E. viminalis Labill., and E. camaldulensis have been listed as naturalized in the floras of various regions and counties (Howell 1958; Munz and Keck 1959; Howell 1970; Beauchamp 1986; Thomas 1991; Smith and Wheeler 1992; Junak et al. 1995; Moe et al. 1995; Matthews 1997; Best et al. 2000). The two most commonly planted 158 eucalypts in California, FE. globulus and E. camaldulensis, are treated in the California Invasive Plant Inventory Database (California Invasive Plant Council 2006—2009) with invasive ratings of moderate and limited respectively. In the most recently published state flora, The Jepson Manual: Higher Plants of California, nine species were included as naturalized (McClintock 1993). Since this 1993 volume, new observations of eucalypt naturalization have been made and those discoveries are reported here. In order to further elucidate the status of the genus in California, we have generated a database of all, or nearly all, past and current collections of Eucalyptus in the state’s many arboreta, botanical gardens, experimental forestry sites, and other public and private plantings. We report here our results and interpretations from many hours of field observations, herbarium study, and plant collection throughout the state. We present current data on the diversity of Eucalyptus in California, which species are spontaneously reproducing, or have the potential to do so, where they can be found, how they can be identified, and our analysis of the potential ecological impacts of the various species where they have been introduced. When monitoring new plant invasions and potential invasions, correct species identification is paramount. For this reason we have included a new key to species, notes on identifying morphological characters, and approximate distributions of naturalized species in California. It is our hope that this paper can act as a guide to the most commonly found naturalized eucalypts for field botanists, land managers, landscape architects, horticultur- alists and silviculturalists, as well as anyone wishing to learn more about the genus in our State. METHODS From 2003 to 2008, the authors visited and studied the Eucalyptus collections in herbaria throughout the state including, the University of California and Jepson Herbarium (UC, JEPS), the California Academy of Sciences (CAS), the Chico State Herbarium (CHSC), the U.C. Riverside Herbarium (UCR), the San Diego Natural History Museum (SD), the Hoover Herbarium of California Polytechnic State Uni- versity, San Luis Obispo (OBI), the Cheadle Center for Biodiversity and Ecological Restora- tion (UCSB), and the Santa Barbara Botanical Garden Herbarium (SBBG). In addition, field observations and collections of Eucalyptus were made from stands and small plantings of multiple species in 31 of California’s 58 counties, including all coastal counties with the exception of Del Norte County. Morphological data and planted ranges for the species notes and key were based MADRONO [Vol. 56 on field observations of living specimens and construction of the key was accomplished by conventional means. For all sites observed, the level of reproduction was gauged based on the number, if any, of spontaneously occurring new individual trees (from seed, not stump sprouting). In order to determine naturalization the number of young trees (those that were clearly not planted) in the area of adult trees were counted, distances from original introduction sites were approximated, sapling juvenile leaves were visually inspected for proper identification, and saplings were pulled up, with the root included, for herbarium vouchers. Species were considered naturalized if new propagules met the criteria defined by Richardson et al. 2000 (species establishes new self-perpetuating populations, undergoes dispers- al, and becomes incorporated into resident flora). Where it was relevant, sapling age was deter- mined by main stem growth ring analysis by cutting the main trunk and counting the growth rings in the cross section. We also noted any observations made on herbarium vouchers of collections being made from apparently repro- ducing or clearly reproducing stands. We also visited living Eucalyptus collections at many of California’s botanical gardens, arboreta, publics parks, university campuses, and private collections including the Huntington Botanical Gardens (San Marino), the University of Cali- fornia Botanical Garden (Berkeley), the Stanford University campus (Palo Alto), Golden Gate Park and Strybing Arboretum (San Francisco), Balboa Park (San Diego), the Los Angeles County Arboretum (Arcadia), U.C. Davis Arbo- retum (Davis), U.C. Riverside Botanical Garden (Riverside), Fullerton Arboretum (Fullerton), Vasona Lake Eucalyptus Grove (Los Gatos), Quail Botanical Garden (Encinitas), the Ruth Bancroft Garden (Walnut Creek), U.C. Santa Barbara campus (Goleta), Palomar College Arboretum (San Marcos), Orpet and Franceschi Parks (Santa Barbara), Cal Poly campus (San Luis Obispo), and the U.C. Santa Cruz Arbore- tum (Santa Cruz). Records for all accessioned Eucalyptus were compiled from the above collec- tions as well as any published records of species planted at some time in the past in California. A database of species names and locations was created based on any species designated in published records, identified in living collections, or in visited herbaria. The classification system for Eucalyptus followed in this paper is found in Brooker 2000. RESULTS After thorough examination of herbaria, living collections, and introduction records, we found evidence for the introduction, or attempted 2009] introduction, of 374 distinct species of Eucalyptus to California beginning in 1853. Of these 374 taxa, we were able to confirm that as of December 2008, 202 species are represented by one or more mature living trees in the state. Forty eight attempted introductions are apparently now no longer extant in California. We were not able to locate or confirm the existence, or non- existence, of the remaining 124 species. Of the 202 extant species in California, only 38 are widely planted (represented by 10 or more trees in 15 or more different locations), and more than 150 are represented by fewer than 5 mature individuals (approximately 20 of these being represented by a single surviving mature speci- men). Six of the seven subgenera (as recognized by Brooker in 2000) are represented in California. Most of the Eucalyptus diversity in California can be found in only a few collections. The U.C. Santa Cruz Arboretum, the Los Angeles County Arboretum, and the Huntington Botanical Gar- den each have more than 75 extant species and records of having attempted many more. Other collections at the U.C. Davis Arboretum, the Stanford campus, the Cal Poly campus in San Luis Obispo, and the U.C. Berkeley Botanical Garden have 25 or more different extant species. There are many other collections in California with fewer than 25 species. In the horticultural collections and forestry plantations in California we found evidence of regular and widespread spontaneous reproduction from seed of 18 of the 202 extant species (Table 1). Nine of these 18 species were included in the treatment of naturalized Eucalyptus in the 1993 edition of the Jepson Manual (McClintock 1993). Contrary to this treatment, we could find no evidence for the naturalization, or spontaneous reproduction of FE. pulverulenta Sims, the very commonly planted species used for stem cuttings by the cut flower industry. We observed nine previously unrecorded occurrences of eucalypt naturalization. Seven of these nine newly ob- served naturalized species were found in the Max Watson Grove at the Arboretum at U.C. Santa Cruz (36°58'49.11"N, 122°3'29.17”"W), where ap- proximately 80 different species were planted in 1964. In this grove we observed extensive reproduction, with trees ranging in age from | to 30 yr (based on main stem growth ring analysis), of both commonly planted species such as E. camaldulensis (Fig. 1B) and species rare in California such as E. kitsoniana Maiden (Fig. 2B). One interesting case of recently recognized eucalypt naturalization in Southwestern Califor- nia is that of E. conferruminata D. J. Carr & S. G. M.Carr (Fig. 2C). This species has been sold and planted widely in California, under the misap- plied name E. l/ehmannii (Schauer) Benth., for more than half a century for use primarily as a RITTER AND YOST: EUCALYPTUS IN CALIFORNIA [59 TABLE 1. COMMON EUCALYPTUS SPECIES IN CALIFORNIA. A. Taxa naturalized in California. B. Commonly planted taxa that would be expected to reproduce if planted more frequently, based on taxonomic similarity to reproducing species and reports from other areas with similar climates. C. Commonly planted Eucalyptus species for which there is no evidence of reproduction. Expected No evidence of Naturalized naturalization naturalization (A) (B) (C) E. camaldulensis E. amygdalina E. calophylla E. citriodora E. blakelyi E. cornuta E. cladocalyx E. botryoides E. erythrocorys E. conferruminata E. dalrympleana _ E. ficifolia E. fastigata E. dives E. leucoxylon E. globulus E. gunnii E. macranda E. grandis E. maculata E. melliodora E. kitsoniana E. neglecta E. nicholii E. macarthurii E. nicholii E. pauciflora E. mannifera E. paniculata E. preissiana E. ovata E. radiata E. pulverulenta E. parvula E. regnans E. punctata E. polyanthemos E. resinifera E. rudis E. pulchella E. rubida E. spathulata E. robusta E. rudis E. torquata E. sideroxylon E. saligna E. diversicolor E. tereticornis E. megacornuta E. viminalis dense screen along roadways, houses, and agri- cultural fields. We have not found E. /ehmannii, as recognized by Carr and Carr (1980), growing in California (Carr and Carr 1980). Eucalyptus lehmannii has a bud cap 9-15 times as long as wide, a thin peduncle that is 5—9 cm long, and a palpably round apical bud, where the closely related and very commonly grown E. conferru- minata has a bud cap that is 44.5 times as long as wide, a thick, strap-like peduncle that 1s 2-4 cm long, and an apical bud that is palpably trigonous when rolled between the fingers. Eucalyptus lehmannii is also capable of regeneration after fire by sprouting from an underground lignotuber, where E. conferruminata is an obligate seeder (Nicolle 2006). The first observation of reproduc- tion of E. conferruminata was made in San Diego County in 2006 (Jon Rebman personal commu- nication, San Diego Natural History Museum). We have confirmed collections of spontaneously reproducing FE. conferruminata from San Diego County (Rebman 163346, OBI, 33°7'15.49"N, 117°16'20.04"W), Santa Barbara County (Ritter & Yost 379, OBI, 34°26'0.76"N, 119°52'19.55”"W), and San Luis Obispo County (Ritter & Yost 380, OBI, 35°14'21.39"N, 120°38'26.12”W). Repro- duction was moderate in the areas described above, with fewer than 50 seedlings at each site. Collections have also been made from an area where reproduction was extensive, with hundreds of new plants regenerating from seed, in Gaviota, Santa Barbara County (Ritter & Yost 381, 160 MADRONO Number of taxa no overlap overlapping in native range a3 E. polyanthemos E. conferruminata E. cladocalyx FIG. 3. [Vol. 56 C_] Arid [J Wet Summer, Dry Winter G8 Uniform A A MM Dry Summer, Wet Winter (Mediterranean) ¥ Native ranges and seasonal rainfall patterns of Eucalyptus reproducing in California. A. The approximate native ranges of the 18 taxa spontaneously reproducing in California. Only species which have a portion of their range not overlapping with other species are labeled. The FE. camaldulensis native range (throughout Australia except for the southwest) 1s excluded from the figure. B. The seasonal rainfall patterns within the native ranges of reproducing taxa. Based on Williams (1997) page 94. 34°28'24.92"N, 120°12'48.86"W). This reproduc- tion began in earnest after the 2004 Gaviota fire, which burned to the sea through an area densely planted with E. conferruminata. In areas where E. conferruminata reproduction is taking place, all observed saplings were less than 10 yr old, based on main stem growth ring analysis. DISCUSSION The over 200 species of Eucalyptus living in California represent a unique example of a significant and purposeful introduction of a genus of trees into cultivation outside their native range. Eucalyptus species are not only the most prevalent non-native trees in California; there is no other genus of introduced trees represented by more species in the state (McMinn and Maino 1935). This diversity of Eucalyptus represents an opportunity to observe the process of naturaliza- tion, or lack of naturalization, in different and closely related species. It became apparent during the course of this study that different eucalypts are spontaneously reproducing at different rates, sometimes regardless of how frequently they are planted. In some groves, such as the Max Watson Grove at the Arboretum at U.C. Santa Cruz, where over 70 different mature species remain, we observed copious reproduction from seed of some species and could find no evidence of reproduc- tion of other species, planted just a short distance away. The degree of spontaneous reproduction in different species may correlate with the taxonom- ic subgenus and section to which they belong, and therefore future invasions could possibly be predicted for closely related yet uncommonly planted species (Table 1). Of the eighteen taxa that have become naturalized in the state, all but 3 are in the subgenus Symphyomyrtus (Pryor and Johnson 1971). The exceptions are E. fastigata (Fig. 1C) and E. pulchella Desf. (Fig. 2D), both in the subgenus Eucalyptus (Monocalyptus of Pryor and Johnson 1971), and E. citriodora in the subgenus Corymbia (see species notes for a brief discussion of the recent elevation of the subgenus Corymbia to the genus level). Within the subge- nus Symphyomyrtus, 7 of the 18 naturalized taxa are in the large section Maidenaria L. D. Pryor & L. A. S. Johnson ex Brooker, which has a total of 80 taxa. It is possible that other species in this section, if planted more commonly, would be predicted to naturalize more readily in California than species in other sections (Table 1). In a contrasting situation, Symphyomyrtus sections Dumaria (23 taxa) and Bisectae (50 taxa), which are represented by over 50 living taxa in California, have only one species (E. conferrumi- nata) that is apparently naturalized. Eucalyptus, although almost entirely endemic to Australia, has a broad native range (~1.6 X 10’ km/’) including species that have evolved in myriad climate types, including temperate rain- forests, deserts, humid subtropical coastal areas, and Mediterranean climate regions. The 18 taxa naturalized in California are native primarily to southeast and east Australia, with the exception of E. conferruminata and E. cladocalyx F. Muell., which have coastal distributions in southern West Australia and southern South Australia (Fig. 3A). Eucalyptus camaldulensis has a wide- spread but largely inland distribution (occurring Fic. 4. RITTER AND YOST: EUCALYPTUS IN CALIFORNIA 161 Illustrations of fruits and buds of: A. Eucalyptus cladocalyx (R. Philbrick B65-44, CAS). B. Eucalyptus mannifera (Ritter 226, OBI). C. Eucalyptus grandis (Broder 1472, CAS). D. Eucalyptus parvula (McClintock s.n., CAS 474279). All size bars = 1 cm. in every mainland Australian state) and is usually found near permanent or seasonal watercourses. It is interesting to note that most of the naturalized taxa in California are native to areas with different seasonal rainfall patterns than those found in Mediterranean climate areas of cismontane California. Eastern Australia, and parts of southeastern Australia, have either uniform year-round or summer maximum pre- cipitation (Fig. 3B). Of the 18 naturalized taxa, only E. conferru- minata and E. cladocalyx (Fig. 4A) have native ranges that fall entirely in winter maximum rainfall Mediterranean type climate areas of Australia. At present, there are 74 taxa alive in California with a native range entirely in Mediterranean areas of southern and southwest- ern Australia. Of these 74, 13 are planted widely (represented by 10 or more trees in 15 or more different locations). There is no evidence of spontaneous reproduction from seed or natural- ization of any of these species, with the exception of E. conferruminata and E. cladocalyx. It is noteworthy that some eucalypt species reproduce frequently in California while others, which are closely related, planted as frequently and in the same locations, do not apparently reproduce. Two examples of this phenomenon are E. camaldulensis and its closest relative E. rudis Endl., and E. sideroxylon and two of its closest relatives E. leucoxylon F.Muell. and E&. melliodora A. Cunn. ex Schauer. Eucalyptus camaldulensis 1s widely naturalized in central and southern California. Its closest relative, E. rudis, which is endemic to southwestern West Australia, is commonly planted in California, Fic. 5. MADRONO [Vol. 56 Illustrations of fruits and buds of: A. Eucalyptus macarthurii (Ritter 232, OBI). B. Eucalyptus ovata (Huber 1203, CAS). C. Eucalyptus robusta (Pollard s.n., CAS 556588). D. Eucalyptus sideroxylon (McClintock s.n., CAS 863986). All size bars = 1 cm. especially along roadways in the central western and southern parts of the state, yet there is no evidence of even occasional reproduction from seed of this species. Similarly, E. /eucoxylon and E. melliodora, both horticulturally important species that are planted frequently throughout the state, apparently do not reproduce, whereas E. sideroxylon saplings can be found in many areas where this species is grown. Conversely, E. nicholii Maiden & Blakely, which is planted very commonly throughout the state, apparently rarely or never reproduces although its close relatives E. mannifera (Fig 4B), E. macarthurii, and FE. parvula do so extensively. Why this disparity in reproduction of closely related species exists in California is a question requiring further study. The native ranges of the closely related species, mycorrhizal fungal associations (especially during seed germination), soil types, and a number of other factors that might affect germination and recruitment of new trees could all play a role in explaining this observed variation in reproduction. We observed widespread reproduction of a number of species in the Max Watson Grove at the Arboretum at U.C. Santa Cruz, such as E. fastigata (Fig. 1C), E. grandis W. Hill ex Maiden (Fig. 4C), E. kitsoniana (Fig. 2B), E. parvula L. A. S. Johnson & K. D. Hill (Fig. 4D), and E. macarthurii H. Deane & Maiden (Fig. 5A), which are not planted frequently elsewhere in Califor- nia. Eucalyptus kitsoniana, E. parvula, and E. macarthurii are rarely grown elsewhere outside Australia (Jacobs 1981) and are very uncommon | in California. It is tempting to speculate that if | other fast growing timber species such as EF. | fastigata and E. grandis were planted widely in California, instead of FE. globulus, then there could be as much or more reproduction and expansion of these plantations. Had E. diversico- lor F. Muell., a fine timber species from Southwest Australia that apparently does not reproduce where it is grown in California, been promoted instead of E. globulus, would there be no issue with expanding groves? Not surprisingly, | a correlation exists between the commonness of | 2009] which a species is planted in California and the number of observations (and possibly the actual degree) of spontaneous reproduction and natu- ralization. The recently recognized and locally extensive reproduction of E. conferruminata (Fig. 2C) in parts of California is an interesting case of possible naturalization after a long period of latency. There have been a number of observa- tions made of non-native species that only become invasive after a long lag time subsequent to their initial introduction (Ellstrand and Schierenbeck 2000). This species, which is planted commonly along a number of roads, highways, and freeways in western Central and Southern California (and has been for over 50 yr), is considered a serious threat to wildlands in the Mediterranean climate Cape Province of South Africa where it has been planted as a sand-binder and windbreak. Euca- lyptus conferruminata is listed as a Category 1 plant as defined in the South African Conserva- tion of Agricultural Resources Act 43 (1983) (its planting, propagation, and importation is pro- hibited) in the Western Cape, due to the fact that it forms thickets in areas of coastal fynbos (Richardson et al. 1996; Henderson 2001; Le Maitre et al. 2002; Forsyth et al. 2004). Why after RITTER AND YOST: EUCALYPTUS IN CALIFORNIA 163 being planted so commonly in California for more than 50 yr is this species now just beginning to become naturalized? Eucalyptus conferruminata, which reproduces in the wild only after fire, may only become naturalized after repeated fires in cultivated areas (Nicolle et al. 2008). Another possibility is that a different genotype of this species was introduced into South Africa; a genotype that does not require fire for profuse reproduction. Genetic fingerprinting or chloro- plast haplotype analysis would be useful in elucidating any genetic component to these vastly different levels of naturalization of the same species in similar Mediterranean climates (McKinnon et al. 2001; Freeman et al. 2001). Guide To Using The Key This key includes the 18 species known to the authors to spontaneously reproduce in Califor- nia. There are many commonly planted species for which there is no evidence of any reproduc- tion (see above) and these have been omitted from the key. However, the user of the key is encouraged to use the species notes (Appendix 1) as there are lists of non-reproducing species that are Closely related to those found in the key. Key To Eucalyptus Species Spontaneously Reproducing In California 1. Inflorescences of terminal or axillary panicles of umbels 2. Leaves lanceolate, lemon-scented; bark smooth ene Hak es Meio ee eu HL, oe nee eh ated E. citriodora 2’ Leaves ovate, elliptic, or orbicular, not lemon-scented; bark variable (often rough [subsp. vestita L. A. S. Johnson & K. D. Hill], or occasionally smooth, [subsp. po/yvanthemos L. A. S. Johnson & K. D. ANN) je eaeiseh eds ores, AP ee oe hes keene cath as iqeee eA eh eet epee aye een heen E. polyanthemos 1’ Inflorescences of unbranched umbels or heads borne in leaf axils or flowers borne singly in leaf axils 3. Flowers borne singly in leaf axils, + sessile; fruit > 1.5 cm wide..................... E. globulus 3’ Flowers in 3-to 15 flowered axillary umbels or heads; fruit =1 cm wide 4. Leaves lighter abaxially (discolorous) 5. Bark rough, persistent on trunk and large branches, thick, fibrous; fruit valve tips remaining fused after dehiscence E. robusta 5’ Bark smooth, shedding from trunk and large branches, occasionally rough up to ~1 m on trunk; fruit valve tips distinct after dehiscence 6. Fruit prominently ribbed, barrel-shaped; valves of mature fruit not exserted (sunken (eaSiCle Ma ANCL): at sect ie ho aol © oe Sa ee ee Ge Sea Pe oe E. cladocalyx 6’ Fruit smooth, obconical; valves of mature fruit exserted and incurved......... E. grandis 4’ Leaves same color on both sides (concolorous) 7. Bark rough, persistent on trunk and large branches, brown to black 8. Bark deeply furrowed, hard, black; outer stamens without anthers (staminodes); TAMEDES FEC OF WOME so 6c oi- 5 oe ES oe eee se eee ee E. sideroxylon 8’ Bark fibrous, brown; stamens all fertile; filaments always white 9. ‘Umibels often paited in leaf axils;*bud cap scat absent « ....% 64.6644 c0% E. fastigata 9’ Umbels always singe in leaf axils; bud cap scar present.............. E. macarthurii 7’ Bark smooth, shedding from trunk and large branches, sometimes rough up to ~2 M on trunk, gray, white, or tan 10. Flowers and fruit fused at the base into a dense, spherical head, >3 cm in diameter... Dh ied, Gr Pa ah PNM 2d, ed aS POE IR a, Wend ye US Bt ts nets Gcaees ane een Ne ce E. conferruminata 10’ Flowers and fruits free at the base ll tntloréscences mostly 3-tlowered umbels. ..4 3 4ad Pe ei ee ee eee oo E. viminalis 11’ Inflorescences 5-to 15-flowered umbels or heads 12. Valves of mature fruit not exserted (sunken below level of hypanthium rim) 13. Flowers and fruit stalked; leaves linear, generally =0.5 cm wide; bud cap scar ADSCINE Wats ache ais Bae cain a Sh ss) Re ee eae ee a eae ae ales E. pulchella 13’ Flowers and fruit sessile; leaves lanceolate, elliptic, or ovate, generally =1 cm wide; bud cap scar present 164 MADRONO [Vol. 56 14. Mature crown retaining large numbers of opposite, sessile, juvenile leaves; adult leaves == sem wide. 2 2 5. een ce eee eee a Gee E. parvula 14’ Mature crown with only alternate, stalked, adult leaves; adult leaves > 2cm wide........ oe ee ee eae ae ee ee oe E. kitsoniana 12’ Valves of mature fruit level with or exserted beyond rim of hypanthium 15. Bark smooth to ground level, powdery to the touch, mottled, shedding in plates; leaves narrow-lanceolate, dull bluish green ............. E. mannifera 15’ Bark often rough up to ~1m on trunk, not powdery to the touch, shedding in short strips; leaves lanceolate to broad-lanceolate, glossy green 16. Fruit obconic;: valves slightly exserted or occasionally level with hy pantiium rim: DUG: CapxCOMICss415 per umbel; bud cap_ scar absent. Rarely planted in California, extensive | reproduction in Joaquin Miller Park, Alameda Co. | (37°48'49.40"N, 122°10'57.43"W). Related species: £. | amyegdalina, E. dives, E. radiata. | Eucalyptus robusta Sm. (Fig. 5C). Pollard s.n. March. 1970 (CAS 556588); Sanders 13849 (UC 1792828). Bark 2009] thick, fibrous, spongy, reddish brown, easily torn from trunk; leaves discolorous, glossy; fruit valves remain joined at tip after dehiscence. Occasional landscaping plant throughout western California. Reproduces occa- sionally. Related species: E. botryoides, E. resinifera. Eucalyptus sideroxylon A. Cunn. ex Woolls (Fig. 5D). McClintock s.n. August 1989 (CAS 863986); Gross 387 (UC 1870241). Bark rough, hard, gray-black; buds 7 per umbel; staminodes present. Very common landscaping plant throughout western Cali- fornia. Reproduces occasionally. Related species: E. leucoxylon, E. melliodora. Eucalyptus tereticornis Sm. (Fig. 6A). Boyd s.n. February 1981 (CAS 931360); Ertter 17604 (UC RITTER AND YOST: EUCALYPTUS IN CALIFORNIA 167 1789216). Similar in appearance to E. camaldulensis; bud cap horned not beaked; seeds dark brown not yellow. Grown commonly in California. Occasionally form hybrids with E. camaldulensis in California. Related species: E. blakelyi, E. dealbata, E. dwyeri. Eucalyptus viminalis Labill. (Fig. 6B). Keil 25866 (OBI 542293); Ertter 10146 (UC 1607671). Bark gray, rough to 1 m above ground level then smooth above, shed in long ribbons; buds 3 per umbel; fruit cup- shaped; fruit valves exerted beyond capsule rim. Very common landscaping plant throughout western Cali- fornia. Extensively naturalized in parts of San Fran- cisco Bay Area; reproduces regularly elsewhere in the state. Related species: E. dalrympleana, E. rubida. MADRONO, Vol. 56, No. 3, pp. 168-183, 2009 MOLECULAR PHYLOGENY OF THE PUNGENTES SUBSECTION OF CHORIZANTHE (POLYGONACEAE: ERIOGONOIDEAE) WITH EMPHASIS ON THE PHYLOGEOGRAPHY OF THE C. PUNGENS-C. ROBUSTA COMPLEX CHRIS BRINEGAR' Conservation Genetics Laboratory, Department of Biological Sciences, San Jose State University, San Jose, CA 95192 acbrinegar@hotmail.com SANDRA BARON 119 Rancho Road, Watsonville, CA 95076 ABSTRACT The nuclear ribosomal internal transcribed spacer and chloroplast rbcL-accD intergenic spacer regions of species in the Pungentes subsection of Chorizanthe (spineflower) were analyzed. The goals of the study were to ascertain the phylogenetic relationships of taxa in the C. pungens-C. robusta complex to each other and to the other Pungentes species. The ITS phylogenies showed that C. diffusa was most distant from the other species and that C. cuspidata, C. pungens and C. robusta were not monophyletic. The four varieties of the C. pungens-C. robusta complex resolved into montane and coastal groups but the groupings were not in agreement with current classification. Statistical parsimony analysis divided 26 chloroplast DNA haplotypes into four groups with the 18 sampled populations of the C. pungens-C. robusta complex displaying a high degree of phylogeographic structure despite some incomplete lineage sorting. The haplotype analysis was not discordant with the ITS phylogeny. Possible taxonomic revisions and implications for conservation and management are discussed. Key Words: Chloroplast DNA haplotypes, Chorizanthe, internal transcribed spacer, phylogeography, Pungentes, spineflower. Chorizanthe R. Br. ex Benth. (spineflower) is a New World genus composed of approximately 10 perennial and 41 annual species, with most of the annuals found in western North America and principally in California (Reveal and Hardham 1989). The most recent revision of Chorizanthe annuals (Reveal and Hardham 1989) placed all but two species in the subg. Amphietes Reveal & Hardham which was further divided into four sections. Within the sect. Ptelosepala Nutt. (including Herbaceae Benth. ex Goodman), seven species endemic to the California central coast were assigned to the subsect. Pungentes Good- man. Four of these species are currently under federal protection due to the extensive loss of coastal dune and sandy inland habitat: C. howellii Goodman and C. valida S. Watson each have only one known population in Mendocino and Marin counties, respectively, while C. pungens Benth. and C. robusta Parry are confined to fragmented populations in the Monterey Bay region. Of the more widespread species, C. cuspidata S. Watson var. villosa (Eastw.) Munz is found from the Point Reyes Peninsula to Bodega Bay, with var. cuspidata restricted to isolated populations in San Francisco and 'Present address: 1500 Middle Road, New Portland, ME 04961. northern San Mateo counties. The remaining | two species, C. diffusa Benth. and C. angustifolia © Nutt., are distributed from Santa Barbara to | Santa Cruz counties. Classification of species in Pungentes has been | confounded by intraspecific variation and, in | areas of overlapping ranges, interspecific similar- | ities. Until the revision of Goodman (1934), | many of the currently recognized species in the | subsection were often considered varieties of | Chorizanthe pungens (Parry 1884; Jepson 1914). — The C. pungens-C. robusta complex has been an | especially difficult taxonomic group due to. historical confusion over type specimens, coastal | vs. inland expressions, overlapping ranges, diffi- | culty in locating rare populations, and disagree- | ment over taxonomic validity of varieties (Reveal - and Hardham 1989; Reveal and Morgan 1989; | Ertter 1996). | Chorizanthe pungens var. pungens (Monterey spineflower) is a generally prostrate, wiry herb | with white to pink involucral margins found in: coastal Monterey Co. with some fragmented ° populations extending into the Salinas Valley. | The montane variety, C. pungens var. hartwegiana | Reveal & Hardham (Ben Lomond spineflower), | is more erect with purple to pink involucral. margins and is confined to the sandhills of the Santa Cruz Mountains. Chorizanthe robusta var. robusta (robust spineflower) is generally taller 2009] and more erect than C. pungens with larger and whiter involucral margins. A shorter expression of C. robusta var. robusta has been observed at the southern end of the range in the backdune habitat of coastal Monterey Co. where it overlaps the northern range of C. pungens var. pungens. The Santa Cruz Mountains variety, C. robusta var. hartwegii (Benth.) Reveal & R. Morgan (Scotts Valley spineflower) is smaller with purple to pink involucral margins. Despite these assign- ments, the definition of C. robusta is still not satisfactorily settled (Reveal and Hardham 1989). This taxonomic uncertainty was reflected in the decision by the U.S. Fish and Wildlife Service to list the entire complex in order to protect the existing threatened and endangered populations, regardless of their taxonomic status (Ertter 1996). Phylogeographic analysis of both nuclear and chloroplast loci has proven to be a useful tool in reconstructing the evolutionary histories of re- cently divergent plant taxa (Wojchiechowski et al. 1999; Comes and Abbott 2001; Kimball et al. 2003; Chiang et al. 2004; Mort et al. 2004; Cornman and Arnold 2007; Yamanaka et al. 2008; Wang et al. 2009). However, based on population genetic theory, it is unreasonable to expect recently divergent taxa to have completely resolved gene tree phylogenies (Hudson and Coyne 2002; Rosenberg 2002). Non-coalescence of ancestral polymorphisms can sometimes lead to discordance between nuclear and plastid gene trees (Soltis and Kuzoff 1995; Comes and Abbott 2001; Yamanaka et al. 2008) and frustrate attempts at species delimitation based on exclu- Sivity criteria, such as reciprocal monophyly. Despite these difficulties, models suggest that useful phylogenetic information can be derived in the presence of reticulation and incomplete lineage sorting (Lyons-Weiler and Milinkovitch 1997; Madison and Knowles 2006; Knowles and Carstens 2007), and several empirical studies have shown that to be the case (Wolfe and Elisens 1996; Levin 2000; Cubas et al. 2006; Jakob and Blattner 2006). Limited molecular systematic studies have been performed in Polygonaceae (Lamb Frye and Kron 2003; Sanchez and Kron 2008), but none have been undertaken at the subgenus or species complex level where recent divergence and rapid diversification pose greater challenges. In this investigation, the nuclear ribosomal ITS and chloroplast rhcL-accD intergenic spacer regions were analyzed to help clarify the systematics of the Pungentes subsection of Chorizanthe. Partic- ular emphasis was given to the phylogeographic assessment of several populations of the C. pungens-C. robusta complex to ascertain its evolutionary relationship to the rest of the subsection, provide data for possible taxonomic revision, and identify genetically rare populations for conservation and management purposes. BRINEGAR AND BARON: CHORIZANTHE SUBSECT. PUNGENTES PHY LOGENETICS 169 METHODS Plant Collection Two to four leaf samples were taken from 10— 20 plants of each Chorizanthe population de- scribed below (more plants were sampled from the larger Sunset State Beach populations). Leaf tissue was either frozen on dry ice in the field and later lyophilized or placed directly in a coin envelope and dried in the presence of desiccant. Voucher specimens were archived in the Carl W. Sharsmith Herbarium (CSH), San Jose State University (accessions 15177-15202). Varieties of Chorizanthe pungens and C. robusta were collected from several sites in the Monterey Bay region representing the majority of known populations of the complex (Fig. 1, inset). For C. robusta var. robusta, two small but morphologically distinct populations in Pogonip Park were sampled. Isolated inland populations of var. robusta were sampled at Branciforte, Rodeo Gulch, Freedom, and Buena Vista. The Manresa and Sunset State Beach populations are coastal expressions with the Sunset State Beach backdune population representing the southern- most extent of var. robusta. Adjacent to this latter population of var. robusta is the northernmost population of C. pungens var. pungens in fore- dune habitat. Three var. pungens inland popula- tions were sampled at the former Fort Ord site near the southern extreme of the variety’s range, and three mid-range foredune populations were sampled at Zmudowski, Salinas River and Marina State beaches. The two largest popula- tions of C. pungens var. hartwegiana were sampled at the Bonny Doon Ecological Preserve and Quail Hollow Park sites in the Santa Cruz Mountains along with the largest C. robusta var. hartwegii population at Glenwood Preserve in Scotts Valley. The approximate locations of collection sites of the other taxa in Pungentes are shown on the map in Fig. 1. Chorizanthe howellii was sampled from the only known population in low dune habitat at MacKerricher State Park near Fort Bragg (Mendocino Co.), and C. valida was taken from its only extant population at G Ranch in Point Reyes National Seashore (Marin Co.). Three populations of C. cuspidata var. villosa were sampled from Bull Point Trailhead, Abbotts Lagoon and South Kehoe Creek at Point Reyes National Seashore. The var. cuspidata was sampled at Baker Beach in Golden Gate National Recreation Area in San Francisco. The northern expression of C. diffusa (sometimes referred to as var. diffusa) was collected from two close populations at Corona Ridge east of Point Lobos State Reserve in Monterey Co. while C. an- gustifolia was sampled from the south end of Morro Bay and the nearby Montana de Oro 170 BONNY DOON ECOLOGICAL PRESERVE Z [_] C. pungens var. hartwegiana SUNSET 44 (_] C. robusta var. hartwegii STATE BEACH = (_| C. robusta var. robusta ZMUDOWSKI STATE BEACH | Bic. pungens var. pungens C. cuspidata var. villos C. valida : ; : SALINAS RIVER eCastroville C. cuspidata var. cuspidat. J STATE BEACH é Monterey Bay end MARINA be STATE BEACH C. diffusa ‘ a 5 miles C. angustifolia Q Fic. 1. MADRONO San Lorenzo R. POGONIP(4 N [Vol. 56 QUAIL HOLLOW PARK [_] GLENWOOD PRESERVE Felton @ Scotts Valley BRANCEL [ |]RODEOGULCH FORTE [_] FREEDOM [_] BUENA VISTA STATE BEACH eWVatsonville Location of Chorizanthe (subsect. Pungentes) taxa used in this study. Inset: locations of populations analyzed from the C. pungens-C. robusta complex in the Monterey Bay region. At Sunset State Beach, C. pungens var. pungens and C. robusta var. robusta occurred in foredune and backdune habitat, respectively. State Park in San Luis Obispo Co. Chorizanthe douglasii Benth. of the sister subsect. Legnota Reveal & Hardham was collected at the former Fort Ord site in Monterey Co., and Eriogonum nudum Benth. var. nudum was collected at the Buena Vista site in Santa Cruz Co. The latter variety was used as the outgroup for phylogenetic analysis. DNA Extraction, PCR Amplification and Sequencing DNA was extracted and amplified from dried leaf tissue (2 mm’) using the REDExtract-N- Amp'™ combination plant extraction and PCR kit (Sigma-Aldrich, St. Louis, MO). The univer- sal plant nuclear rDNA internal transcribed spacer (ITS) PCR primers A and B as described by Blattner (1999) were used at 0.5 uM each to amplify an approximately 750 bp product con- taining the entire ITS-1, 5.8S rDNA and ITS-2 regions. DNA extracts (4 uL) were used in a final reaction volume of 20 uL. PCR conditions were 94°C (2 min) followed by 30 cycles of 94°C (45 sec), 55°C (1 min), 72°C (2 min) and a final extension step of 72°C (6 min). Polygonum cuspidatum-based PCR primers (Inamura et al. 2000) were initially used to amplify part of the | chloroplast rbcL and accD genes along with their | intergenic spacer. However, in Chorizanthe the PCR product sequence variability for the two | coding regions was low, so the new forward | primer RBCL (5’-CCGCTTGTGAAGTATG | GAAAG-3’) and reverse primer ACCD (5’- AGGATCAAGACTGCCCATTG-3’) were de- | signed based on Chorizanthe sequences to amplify an approximately 770 bp product containing all | of the spacer region but very little of each - flanking gene sequence. PCR conditions were | the same as for the ITS region except that 35 | cycles of amplification were used. Amplified DNA was purified using DNA | Clean and Concentrator’™-5 spin columns | (Zymo Research, Orange, CA). Sequencing for | both loci was performed by Geneway Research © (Hayward, CA) using Big Dye™ v.3.1 sequencing | mix (Applied Biosystems, Foster City, CA) and | the forward PCR primers on an ABI 3730XL | DNA Analyzer (Applied Biosystems). Reverse primers were used for sequencing only in the rare | case when forward priming gave low quality or | incomplete reads. Higher quality sequence was © obtained for the ITS region when GC-rich buffer (Applied Biosystems) was used. ITS sequences | 2009] TABLE l. BRINEGAR AND BARON: CHORIZANTHE SUBSECT. PUNGENTES PHY LOGENETICS 171 GENBANK ACCESSION NUMBERS OF NUCLEAR RIBOSOMAL DNA INTERNAL TRANSCRIBED SPACER REGION. See Methods for descriptions of populations. Taxon C. howelli C. valida C. cuspidata var. villosa cuspidata var. cuspidata angustifolia diffusa robusta var. hartwegii iY oh Oo robusta var. robusta C. pungens var. hartwegiana C. pungens var. pungens C. douglasii E. nudum var. nudum Population MacKerricher G Ranch Abbotts Lagoon Bull Point Kehoe Creek Baker Beach Morro Bay Montana de Oro Pt. Lobos Pop1 Pt. Lobos Pop2 Glenwood Pogonip Pop2 Rodeo Gulch Buena Vista Manresa Sunset backdune Bonny Doon Sunset foredune Zmudowski Salinas River Fort Ord Pop! Fort Ord Pop2 Fort Ord Pop3 Fort Ord Buena Vista Specimen (GenBank accession number) MK-9 (EU753776), MK-10 (EU753777), MK-17 (EU753778), MK-18 (EU753780), MK-19 (EU753779) GR-1| (EU753781), GR-3 (EU753783), GR-18 (EU753782), GR-19 (EU753785), GR-20 (EU753784) AL-2 (EU753741), AL-3 (EU753744) BP-1 (EU753740), BP-3 (EU753743) KC-1 (EU753742) BB-1 (EU753756), BB-3 (EU753757) MB-1 (EU753738), MB-3 (EU753739) MO-18 (EU753735), MO-19 (EU753737), MO-20 (EU753736) PL1-6 (EU753730), PL1-8 (EU753731) PL2-16(BU793732), PL2-19 (BU753733), PL2-20 (EU 793734) GL-1 (EU753745), GL-5 (EU753746), GL-7 (EU753747) GL-18 (EU753748) PO2-3 (EU753753) RG-1 (EU753773), RG-2 (EU753770) BV-1 (EU753769) MA-7 (EU753754), MA-10 (EU753764), MA-16 (EU753765), MA-17 (EU753755) SB-1 (EU753763), SB-3 (EU753767), SB-4 (EU753768) BD-1 (EU753751), BD-3 (EU753749), BD-12 (EU753750) SF-5 (EU753762), SF-7 (EU753760), SF-9 (EU753766), SF-10 (EU753752), SF-18 (EU753772) ZM-12 (EU753758) SR-2 (EU753759) FO1-1 (EU753774) FO2-1 (EU753761) FO3-1 (EU753771) CD-1 (EU753775) EN-1 (EU753786) were assigned GenBank accession numbers EU753730—EU753786. Chloroplast haplotype sequences were assigned accession numbers EU794965—EU795005. At the request of the GenBank administrator, any haplotype sequence shared by different taxa was assigned a separate accession number for each taxon. Sequence Alignments and Phylogenetic Analysis ITS region sequences were aligned using Clus- talX (Thompson et al. 1997). Where there was polymorphism in individual sequences, the IU- PAC ambiguity codes were used. Neighbor- joining phylogenetic analysis of ITS sequences using the Kimura two-parameter model (Kimura 1980) and maximum parsimony analysis using the close-neighbor-interchange option (search level 1) were performed with MEGA4 software (Tamura et al. 2007). In both analyses, gaps were treated as missing data and bootstrap consensus trees were inferred from 1000 replicates (Felsenstein 1985). The ITS region of E. nudum var. nudum was used as the outgroup sequence for both trees. Chloroplast rbcL-accD intergenic spacer re- gion (cpDNA) haplotypes were aligned manually. Statistical parsimony analysis to estimate haplo- _ type genealogy was conducted using the program TCS (Clement et al. 2000). Following the procedure of Jakob and Blattner (1996), gaps were treated as a fifth state, each indel was treated as a single mutational event, and the single mononucleotide (T/A) repeat was omitted from the analysis. Maximum parsimony phylo- genetic analysis was performed as with the ITS sequence. Pairwise distances for data sets of both loci were calculated in MEGA4 using a Kimura two- parameter model. RESULTS ITS Sequence Variation GenBank accession numbers of the ITS sequences for 56 Chorizanthe specimens and the Eriogonum nudum outgroup are provided in Table 1. The combined ITS-1, 5.8S and ITS-2 regions resulted in an alignment of 683 nucleo- tides. In the Chorizanthe sequences there were a total of 68 (10.0%) variable sites, 36 of which were unique to C. diffusa. In the highly conserved 5.8S sequence, C. diffusa sequences showed a single base insertion and a polymorphic site not seen in the other Chorizanthe taxa. Four other single-base indels were unique to C. diffusa (one 85 96 58 98 95 87 94 50 65 99 FIG. 2. MADRONO [Vol. 56 C. howellii (MK-18) C. howellii (MK-19) C. valida (GR-1) C. valida (GR-3) CladeA C. douglasii(CD-1) C. cuspidaia var. villosa (BP-1) C. cuspidata var. villosa(KC-1) C. angustifolia (MB-1) C. angustifolia (MO-18) C. cuspidata var. cuspidata (BB-1) C. cuspidata var. cuspidata (BB-3) C. robustavar. robusta (SB-1) C. robustavar. robusia(BV-1) C. robusta var. robusta (MA-7) Clade B C. robusta var. robusta (PO2-3) C. pungens var. pungens (ZM-12) C. pungens var. pungens (FQO1-1) C. pungens var. pungens (SF-5) C. pungens var. pungens (SR-2) C. robustavatr. hartwegii(GL-1) C. pungens var. hartwegiana(BD-1) C. robusta var. hartwegii(GL-5) C. pungens-C. robusta complex C. pungens var. hartwegiana(BD-3) C. diffusa(PL1-6) C. diffusa (PL2-19) E. nudum (EN-1) Maximum parsimony bootstrap consensus tree of the complete ITS-1/5.8S/ITS-2 sequence. Numbers at ' branches are bootstrap values based on 1000 replicates with values below 50% omitted. Population and specimen number are in parentheses (see Table 1). in ITS-1, three in ITS-2). Site polymorphism in individual sequences was common throughout subsect. Pungentes. Pairwise distances between Chorizanthe taxa were 0.0—-6.5% with an average of 1.8%. Exclud- ing C. diffusa, the pairwise distance range was 0.0—-1.5% with an average of 0.9%. Within the C. pungens-C. robusta complex, the average pairwise distance was only 0.2% with some individuals of C. pungens var. pungens and C. robusta var. robusta having identical sequences. ITS Phylogeny The maximum parsimony (Fig. 2) and neigh- bor-joining (Fig. 3) consensus trees were con- structed using ITS sequences from four individ- | uals each of Chorizanthe pungens var. pungens | and C. robusta var. robusta, two from each of the & remaining Pungentes taxa, and one from C. douglasii. Increasing the number of individuals had no significant effect on the topology of either» tree: Similar topologies were observed in both trees. with all members of subsect. Pungentes (except | Chorizanthe diffusa) falling within Clades A and B. In both analyses there was very strong. bootstrap support for an early divergence of C. diffusa. In the neighbor-joining tree the long terminal branches in individuals from the two C. | diffusa populations were caused by eight base substitutions. 2009] BRINEGAR AND BARON: CHORIZANTHE SUBSECT. PUNGENTES PHY LOGENETICS go| C. howellii(MK-18) C. howellii (MK-19) C. valida (GR-1) 39! ©. valida (GR-3) 20 C. douglasii(CD-1) gg| ©. cuspidata var. villosa(BP-1) C. cuspidatavar. villosa(KC-1) 58 91{ C. angustifolia (MB-1) C. angustifolia (MO-18) g6| C. cuspidata var. cuspidaia(BB-1) C. cuspidata vat. cuspidata (BB-3) C. robusta vat. robusia(PO2-3) 30 C. pungens var. pungens (FO1-1) C. robusta var. robusta(BV-1) C. robusta var. robusta(SB-1) 58 C robusta var. robusta (MA-7) C. pungens var. pungens (SR-2) a C. pungens var. pungens (ZM-12) C. pungens var. pungens (SF-5) | ~ 3 CladeA Clade B 0.01 FIG. 3. C. robusta var. hartwegii(GL-1) C. pungens var. hartwegiana(BD-3) C. pungens-C. robusta complex C. robusia var. hartwegii(GL-5) C. pungens var. hartwegiana(BD-1) C. diffusa(PL1-6) set C. diffusa (PL2-19) E. nudum (EN-1) Neighbor-joining bootstrap consensus tree of the complete ITS-1/5.8S/ITS-2 sequence. Numbers at branches are bootstrap values based on 1000 replicates with values below 50% omitted. Population and specimen number are in parentheses (see Table 1). Clade A consists of Chorizanthe howellii and C. valida, two of the most northerly members of subsect. Pungentes. In Clade B of both phylog- enies C. douglasii, C. cuspidata var. villosa and C. angustifolia diverged earlier than C. cuspidata var. cuspidata and the C. pungens-C. robusta complex. All Clade B taxa have ranges south of the Clade A species except for C. cuspidata var. villosa at Point Reyes which is adjacent to the only extant population of C. valida. In both the maximum parsimony and neigh- _bor-joining analyses, the Santa Cruz Mountain varieties of the Chorizanthe pungens-C. robusta complex (C. pungens var. hartwegiana and C. robusta var. hartwegii) fall within a separate group from the coastal/coastal terrace members of the complex (C. pungens var. pungens and C. robusta var. robusta). The maximum parsimony analysis suggests some resolution between C. pungens var. pungens and C. robusta var. robusta, but bootstrap support for those branches was extremely weak (<10%). Sequence similarity (and often sequence identity) at the ITS locus makes it impossible to distinguish between some individ- 174 uals of C. pungens var. hartwegiana and C. robusta var. hartwegii and also between some individuals of C. pungens var. pungens and C. robusta var. robusta. It should be noted that the specimens of C. pungens var. pungens and C. robusta var. robusta used in these phylogenies were chosen to be representative of the north to south range of the sampled populations. As with the two varieties of C. cuspidata, neither the two varieties of C. pungens nor the two varieties of C. robusta, as currently recognized, were found to be monophyletic by these analyses. There is weak bootstrap support (<50%) for some interior branches of Clade B in the maximum parsimony phylogeny where only 16 parsimony informative sites were found in Clade B member sequences. Despite the fact that genetic distances separating all Clade B taxa are small (<1%), similar branches are better supported in the neighbor-joining phylogeny, although the support is still only weak to moderate. Within Clade B, removal of Chorizanthe douglasii from the neighbor-joining analysis (not shown) has the largest positive effect on bootstrap support, raising the bootstrap value at the initial branch of Clade B from 58% to 77%. Separate phylogenetic analyses of the ITS-1 and ITS-2 sequences (not shown) provided some insight into the generally weaker bootstrap support for interior branches of Clade B. Neighbor-joining analysis of the ITS-1 region (which contains 70% of all variable sites in the entire ITS-1, 5.8S and ITS-2 sequence) produced a topology very similar to Fig. 3. However, a tree based solely on the ITS-2 region was significantly different with less bootstrap support. An ITS-2 neighbor-joining tree grouped Chorizanthe an- gustifolia, C. pungens var. hartwegiana, C. robusta var. hartwegii, C. howellii and C. valida in a clade with weak bootstrap support (26%). Also, the ITS-2 tree grouped C. cuspidata var. villosa with C. pungens var. pungens, C. robusta var. robusta and C. cuspidata var. cuspidata with 56% bootstrap support whereas the ITS-1 tree pro- vided 87% bootstrap support for the grouping of the entire C. pungens-C. robusta complex with C. cuspidata var. cuspidata. In the ITS-2 tree, C. douglasii diverged after C. diffusa and prior to the remaining members of Pungentes. Chloroplast Haplotype Sequence Variation GenBank accession numbers of the 26 cpDNA haplotype sequences (A—Z) identified in 445 individuals from 27 populations of Chorizanthe in subsect. Pungentes are provided in Table 2. Individual sequences ranged in length from 671—687 nucleotides. Manual alignment resulted in a 715 nucleotide sequence which included nine indels ranging from 1—15 bases, nine substitu- tions, and a variable mononucleotide (T/A) MADRONO [Vol. 56 TABLE 2. GENBANK ACCESSION NUMBERS OF CHLOROPLAST RBCL-ACCD INTERGENIC SPACER HAPLOTYPES. ‘Haplotypes shared by different taxa were submitted to GenBank under each taxon’s name. -Outgroup (Eriogonum nudum var. nudum) haplotype. GenBank accession numbers! EU794965, EU794983 EU794966 EU794967, EU794973, EU794974 EU794970 EU794975 EU794976 EU794977, EU794978, EU794979, EU794980, EU794989 EU794971, EU794968, EU794982, EU794993 EU794994 EU794995 EU794998 EU794999 EU795000 BU795001 EU795002 EU795003 EU795004 EU795005 Haplotype EU794969, EU794972 EU794984, EU794996 EU794985 EU794986 EU794987 EU794988 EU794981, EU794990 EU794991, EU794997 EU794992 KASS CHV AOVOZZ UAT TOMO, TN Z repeat ranging from 8—14 bases. Four indels were | unique to single haplotypes (K, M, P, W), two haplotypes (T, Z) contained two unique indels | each, and two haplotypes (Q, R) shared the same | deletion. Haplotype A was the consensus se- | quence for all haplotypes. Pairwise distances | between haplotypes ranged from 0.0—0.8% with | an overall average of 0.3%. Statistical Parsimony Network of Haplotypes A 95% parsimony connection limit of 11 steps | was calculated by TCS using a 665 nucleotide | alignment of cpDNA haplotypes where all indels | had been reduced to single mutations. The | mononucleotide repeat was excluded, but its . inclusion did not affect the final genealogy. Due | to extremely minor differences between many of | the haplotypes, this collapsed the number of | unique sequences from 26 to 13. The resulting | haplotype network is shown in Fig. 4. All of the | original haplotypes that are separated by commas | in Fig. 4 were collapsed into a single haplotype | by TCS. For example, Haplotypes B and C_ differed only by 1 and 2 additional bases, | respectively, at the mononucleotide repeat rela-_ tive to Haplotype A, so elimination of this non-| parsimony informative region from the alignment | 2009] BRINEGAR AND BARON: CHORIZANTHE SUBSECT. PUNGENTES PHY LOGENETICS 175 Fic. 4. TCS network of 26 Chorizanthe cpDNA haplotypes. Highly related haplotypes (separated in the figure by commas) were collapsed into unique haplotypes after statistical parsimony analysis. Lines represent a single parsimony-informative mutation. Numbers 1-4 were given to each of the four haplotype groups. collapsed these sequences together. Lines joining haplotypes represent a difference of a single parsimony-informative mutation. There were no missing intermediate haplotypes uncovered by the analysis. Three of the haplotype groups (Groups 2-4) branch off from the remaining core group (Group 1) represented by Haplotypes A, B and C. Haplotype A was selected by TCS as having the highest outgroup probability. A maximum parsimony phylogenetic analysis (not shown) was consistent with the statistical parsimony network. Group 4 haplotypes diverged early in the phylogeny and were separated from Groups 1-3 with moderate bootstrap support (70%). The latter haplotype groups resolved into two weakly resolved clades (<50% bootstrap support). One clade contained the Group | haplotypes while its sister clade was split into Group 2 and Group 3 subclades which had 64% and 65% bootstrap support, respectively. Haplotype Frequencies and Distributions Frequencies of the 26 cpDNA _ haplotypes found in 27 populations of 11 Chorizanthe taxa are shown in Table 3 (expressed as percentages). Within the C. pungens-C. robusta complex, several highly fragmented and very small popu- lations (<2000 individuals) of C. robusta var. robusta were fixed at one haplotype (both Pogonip Park populations, Branciforte, Rodeo Gulch, and Freedom). The only C. pungens var. pungens population with a fixed haplotype was at Zmudowski State Beach. However, the highest degree of haplotype diversity occurred in popu- lations just north and south of Zmudowski State Beach. The foredune (C. pungens var. pungens) -and backdune (C. robusta var. robusta) popula- tions of Sunset State Beach had six and seven haplotypes, respectively, while Salinas River State | Beach (C. pungens var. pungens) had seven. Fixation was also observed in other subsect. Pungentes taxa: C. howellii, C. valida, both populations of C. diffusa, both populations of C. angustifolia, and one population of C. cuspidata var. villosa. The C. cuspidata var. cuspidata population at Baker Beach was nearly fixed at one haplotype. Fourteen of the 26 haplotypes were found in only one population each. Of those, Haplotypes B, F, H, M, Q, S, Y and Z were extremely rare, collectively representing only 3.6% of all samples tested. The most widespread haplotypes were D and J, each occurring in eight populations. Overall, Haplotype D was the most common (19.1%), particularly in Chorizanthe robusta var. robusta (48.6%) and the single C. cuspidata var. cuspidata population (94.1%). Chorizanthe robus- ta var. robusta and C. pungens var. pungens displayed the greatest haplotype diversity (11 and 13 haplotypes, respectively). Chorizanthe howellii, C. valida, C. angustifolia and C. diffusa each had a single unique haplotype. Chorizanthe cuspidata var. villosa had three unique haplotypes. Haplotype sharing was common in all varieties of the Chorizanthe pungens-C. robusta complex and in C. cuspidata var. cuspidata. Chorizanthe pungens var. pungens and C. robusta var. robusta shared seven haplotypes (D, I, J, K, L, N and P), five of which are found in the Sunset State Beach populations of each variety. Haplotype A was shared by C. pungens var. pungens and C. pungens var. hartwegiana. Chorizanthe robusta var. robus- ta, C. robusta var. hartwegii, and C. pungens var. hartwegiana shared Haplotype C while C. pun- gens var pungens, C. pungens var. hartwegiana, and C. cuspidata var. cuspidata shared Haplotype O. Haplotypes D and N also occurred in C. cuspidata var. cuspidata and C. robusta var. hartwegii, respectively. [Vol. 56 ~ MADRONO 176 = = = —= a= vane ass — — — —- — — uooseyT snoqqy = (0z DsopjiA “IVA DJDpidsnd “Dy re = == —_ ie _— a 3s — — _— _- — jUIoOg [ING OZ DSO]IA “1eA DIDpidsnd “Dd O1O [ep vuRJUO[, = =e = = = —_ ao — — — a _- — wy Avg OLIOPL = QT PYOfISNBUY “Dy cdod a = — — a es — — = — _ -- — {dog sogoT 1ulog—s 0 Psnffip “Dd = = = mae = —_ — = = — — ~~. --- youey DTI ppyva “Dd —_ = == == — — = — _ _: — — = IQYOMIYOVYY 1jjamoy “Dy = = ee = — =— ae — — L'?6 — — — yoeog 1oyeq Ss L| pippidsnd “red vyppidsna-D = —_ — 0'0r 2= = = — — O'OE — — — ¢dod PlO 1.104 OT suasund “Ie Ssuasund ‘> = —_ = OC] —_ = = — — — a — TPs qdog plo MOY 6 suasund “Ie Ssuasund ‘Dy _ —_ — 0'Sb oe oe a — — — — — OO! [dog PIO MOL OZ suasund “ied suasund -D = — —_ Oc = = = — — — — — OS eULIeI, =—s QZ suasund “IRA suasund *y Gc ULp — Sc Ac — — — a =o — — 6S ISAT SBUTTeS LI suasund “1eA Suasund “yD = — — _ OOI a = — — — — — = IysMOpnUuy—s 6 | suasund “1eA Suasund “y — hes be & i tL 6 1c = == — — _—. —- — — ouNps1oy JOSUNS EZ suasund “1eA Suasund -y _ ire 9°8h eo BT == = — — hg — — — sunpyoeq josung = Cg DISNGOA “IRA DISNGOA “Dd —_ an 0°0¢ 23 ae = —_ a 00S — — — ~- esoluRrI, =—s_« QT DISNGOA “IRA VISNGOA “D ae — = O'0¢ _ 001 Orcs — _— 0'SZ _- — — eISIA PUDONG (QT DISNGOA “IBA DISNGOA “D = a = = —_ = — = — OO —— — — wopsslt (QZ DISNGOA “IRA DISNGOL “Dd = — ce —_ — = —_ _ —_ OOT == — — yong oapoy Il DISNGOA “IVA DISNGOA “Dd —_ —— =— —_ ame = = = — OO! -- — — a10jlouvig CT DISNGOA “IBA DISNGOL “Dd _ as — — = = = — — _- OO! — — qdog diuos0g al DISNGOA “IVA VISNGOA “Dd ——. _ = = = — = — —— OO! = — — {dog dtuos0g_—s J DISNGOA “IBA DISNGOA “D = = = = _— = = O'S — ~- 0'S9 — —= poomugyy (7 NBIMIADY “IRA DISNGOA D DUDIBIMIADY = zat a —_ = —_ = — — -- 9°89 QR] ara | MOT[OH [lend 9 “IeA suasund “yD DUDIBIMJADY —_ _ = — — = — — — _- 9°87 — £99 uoog Auuog ie “Ie@A suasund ~y W ‘I al f I H D | d d J d V uonrndog u pera AH ‘odAjojdey suo AJUO pomoys puv AT[enba poydures a1aM sarsads asoy} Jo suonvyndod yog, ‘sodvjucoiod se possoidxo oie 7—-Y sodAjojdeyH jo saiuonbel4y “NOILVINdOg A VXV I, SFTLNIONAd NI SAdALOTAVE WNQCdO AO AONANOAY “€ ATIVE$ Ly. BRINEGAR AND BARON: CHORIZANTHE SUBSECT. PUNGENTES PHYLOGENETICS 2009] OST. OSL =~: = OOT =a asi — OOT A x NM === —— 6S —_ i O'S OST OOI — ty OSC = O'S ag 008 OS ae - = 455 =F ——e age By ces Bal = Ls — =. => VC —— ——— QT — O d O N ‘GHNNILNOD ‘¢€ ATEV EL uooser’yT] s}0qqyv WUurOd [IN O1O [Ep vURIUO|[ 2 ARG OLIOPY Tdod 27 [dod soqo7T] 110d Yvuea IOYOIAIISPOV I] yorog Ioyeg ¢dod plo M04 cdod pio 04 [dod plo 1104 BULIv IAI IDA SPUTTeS TySMOpNUWZ suNpos1oy yasuNS sunpyorq jasuns BSOIUP I BISIA vUONg WOPdo1-] yon oapoxy do sIouRIg qdog diuos0g [dog diuosog poomusy MOTOH [!eNO uoog Auuog uonendog DSO]IA “IRA VIDpIdsno - DSO]JIA “IRA DIDpIdsno - PIOf{LUSNBUD * psnffip DplDA 1JAMOY * pippidsnd “red pppidsn. suasund “ea Ssuasund * suasund “1eA Suasund ° suasund ‘1eA Suasund * suasund “tea Suasund ° suasund “tea Suasund * suasund “ea Suasund * suasund “eA Suasund * DISNGOA “YEA DISNGOA * DISNGOA “YEA DISNGOL * DISNGOA “A DISNGOA * DISNGOA “YEA DISNGOA * DISNGOA “IA DISNGOA * DISNGOA “XA DISNGOA * DISNGOA “IA DISNGOA * DISNGOA “IBA DISNGOA ° 1BIMJADY “IVA DISNGOA ° DUDIBAMIADY ‘IBA SUasUuNnd DUDIBIMIADY “IVA SUasUNd UOXeL >: WOOD OUUUDUOUUDUDODUDUOOU.S e, ro) 178 MADRONO BONNY DOON ECOLOGICAL PRESERVE C. howellii am C. cuspidata var. villosayy C. validal4 K C. cuspidata var. cuspidata C. angustifolia Q } S Fic. 5. POGONIP PARK 2 [=] Group 2 haplotypes [_] Group 3 haplotypes [_] Group 4 haplotypes Monterey Bay 5 miles [Vol. 56 QUAIL HOLLOW PARK 2d GLENWOOD PRESERVE BRANCE- FORTE MANRESA STATE BEACH Foredune SUNSET STATE BEACH ZMUDOWSKI CD STATE BEAC SALINAS RIVER /“ | STATE BEACH } MARINA y ; } STATE BEACH S Frequency of cpDNA haplotype groups in populations of Chorizanthe subsect. Pungentes taxa. Haplotypes were assigned to Groups |I-4 by statistical parsimony analysis (Fig. 4). Group frequencies were calculated from Table 3 data. Pogonip Park and Fort Ord populations are numbered as in Fig. 1. Haplotypes were grouped according to the TCS parsimony analysis (shown in Fig. 4). The haplotype group frequencies were then calculated for each population and are displayed graphically in Fig. 5. Taxa outside of the Chorizanthe pungens-C. robusta complex show no phylogeo- graphic correlation with haplotype groups. How- ever, within the C. pungens-C. robusta complex (Fig. 5 inset), Group | haplotypes are much more prevalent in the Santa Cruz Mountain varieties (C. pungens var. hartwegiana and C. robusta var. hartwegii). Group | extends into one of the most northerly C. robusta var. robusta populations (Pogonip Park Population 2). The remaining inland populations of C. robusta var. robusta are dominated by Group 2 haplotypes. The two coastal populations at Manresa State Beach and Sunset State Beach (backdune) are at this variety's southern extreme and are the only populations of C. robusta var. robusta with high frequencies of Group 3 haplotypes (50 and 77%, respectively). The northernmost population of C. pungens var. pungens begins in the foredune of Sunset State Beach and consists entirely of Group 3 haplotypes. With the exception of Fort Ord Population 2 (which is 84% Group 1), the remaining C. pungens var. pungens populations are mostly or entirely composed of Group 3 | haplotypes. Group 4 haplotypes are relatively | rare except in populations of C. howelli and C. diffusa. DISCUSSION Phylogeny of Subsect. Pungentes Reveal and Hardham (1989) considered Chor- izanthe diffusa to be the primary unit of | Pungentes based on its glabrous flower and distinct involucral margin. Indeed, both the | maximum parsimony and neighbor-joining phy- | logenetic analyses of the ITS region strongly | support a more ancient divergence of C. diffusa | from the rest of the subsection. The single | cpDNA haplotype (V) found in the two sampled | populations of C. diffusa is unique to that species | and belongs to the less common Group 4 | haplotypes which are found in C. howellii and, | to a lesser degree, C. pungens var. pungens and C. | robusta var. robusta. | The other members of Pungentes separate into two clades in the ITS phylogenies. An early | north-south split is suggested with one lineage | leading to Chorizanthe howellii and C. valida 2009] (Clade A). These two species are typically distinguished from the others in the subsection by their involucral morphology. Reveal and Hardham (1989) postulated a hybrid origin of C. howellii based on its apparent tetraploid chromosome number (Hardham 1989) and spec- ulated that one of the parents was C. cuspidata var. villosa or perhaps C. valida. The ITS phylogeny is more supportive of C. valida as a prospective parent. The cpDNA data do not shed light on this matter since all three of these entities have haplotypes which fall into different groups. Clade B has lower bootstrap support in the ITS phylogenies than Clade A, especially in the maximum parsimony analysis. However, the maximum parsimony tree provides a similar Clade B topology as the more strongly supported neighbor-joining tree. Most notable in Clade B is the lack of monophyly of varieties in three of the species: Chorizanthe cuspidata, C. pungens and C. robusta. Instead of grouping with C. cuspidata var. villosa, C. cuspidata var. cuspidata groups with the C. pungens-C. robusta complex and also shares greatest cpDNA haplotype similarity with C. robusta var. robusta. It was noted in the last revision (Reveal and Hardham 1989) that spec- imens from extirpated populations of C. cuspi- data in Alameda Co. may have been misidentified as C. robusta. Also, earlier revisions had classified C. cuspidata var. cuspidata as C. pungens var. cuspidata (Parry 1884; Jepson 1914). Prior to its reduction to varietal status (Munz 1958), C. cuspidata var. villosa was recognized as the separate species C. villosa (Eastwood 1903; Goodman 1934). Even so, all specimens of C. cuspidata var. villosa analyzed have Group 2 cpDNA haplotypes as do most C. cuspidata var. cuspidata and C. robusta var. robusta individuals. Analysis of more loci will be necessary to resolve the true phylogenetic relationship between the C. cuspidata varieties and the C. pungens-C. robusta complex. Current classification of the Chorizanthe pun- gens-C. robusta complex is inconsistent with the ITS phylogeny which shows both species to be non-monophyletic. The phylogeny, however, shows a biogeographically consistent pattern in that the two coastal/coastal terrace varieties of the complex (C. pungens var. pungens and C. robusta var. robusta) group together while the two Santa Cruz Mountains varieties (C. pungens var. hartwegiana and C. robusta var. hartwegii) form a separate, but closely related group. Genetic distances between the two groups are low (<0.4%), suggesting a recent divergence. The ITS phylogeny of the C. pungens-C. robusta complex is supported by cpDNA haplotype distributions as discussed in a later section. Further investigation into the cause of weaker bootstrap support for interior branches of Clade BRINEGAR AND BARON: CHORIZANTHE SUBSECT. PUNGENTES PHY LOGENETICS 179 B in the neighbor-joining ITS trees showed that the ITS-2 region produced a phylogeny different than ITS-1 especially in the placements of Chorizanthe douglasii, C. angustifolia and C. cuspidata var. villosa. The phylogeny derived from the entire ITS region (Fig. 3) reflects the greater influence of the more rapidly evolving ITS-1 sequence and provides a better supported and more phylogeographically consistent tree than the ITS-2 region. Soltis and Kuzoff (1995) also found different evolutionary rates for ITS-1 and ITS-2 regions in the Heuchera group which affected placement of taxa when analyzed sepa- rately. Haplotype Analysis and Phylogeography of the Chorizanthe pungens-C. robusta Complex In our study, phylogenetic resolution at the chloroplast locus was limited because of a low mutation rate at the cpDNA rbcL-accD inter- genic spacer and incomplete lineage sorting (42% of cpDNA haplotypes were not species-specific). Ancestral haplotypes can be maintained through multiple speciation events and therefore be inherited through multiple lineages (Madison and Knowles 2006). Such a mechanism has been invoked to explain cpDNA/ITS gene tree dis- crepancies in Coreocarpus (Kimball et al. 2003) and Senecio (Comes and Abbott 2001). In the latter study, ancestral haplotypes may have been maintained for 0.4—-1.0 million years and, as a result, only a small percentage of the haplotypes were diagnostic for species. Traditional methods of phylogenetic recon- struction rely on assumptions that do not necessarily hold at the population level where sequence divergence 1s often low and ancestral haplotype frequencies are high (Crandall and Templeton 1993; Huelsenbeck and Hillis 1993). Network representations have been developed to provide genealogical reconstructions which take into account such population level phenomena (Templeton et al. 1992; Excoffier and Smouse 1994). TCS statistical parsimony analysis (Clem- ent et al. 2000) has been used in several phylogeographical studies to create genealogical networks of cpDNA haplotypes as an alternative to the construction of traditional phylogenetic trees (Grivet and Petit 2002; Honjo et al. 2004; Jakob and Blattner 2006; Gonzales et al. 2008; Tan et al. 2008; Yamanaka et al. 2008; Fehlberg and Ranker 2009; Wang et al. 2009). By grouping highly similar cpDNA haplotypes together, TCS analysis uncovered significant phylogeographic structure in the Chorizanthe pungens-C. robusta complex despite some incom- plete lineage sorting or reticulation, most notably among Group | haplotypes which appear in all four varieties of the complex. The limited number of mutational events connecting all haplotypes in 180 the four haplotype groups, and the fact that there are no missing intermediate haplotypes, reinforc- es how recently divergent the Pungentes taxa are. Coalescent theory suggests that ancestral haplotypes should have interior positions in the genealogical network and occur at high frequen- cies (Crandall and Templeton 1993). Although the Group 2, 3 and 4 haplotypes radiate from Group 1, the frequency of Group | haplotypes is high only in the Santa Cruz Mountains varieties and in one Fort Ord population of the Chor- izanthe pungens-C. robusta complex. The most widely spread haplotypes in Pungentes species are in Group 4. Like Group | haplotypes they do not appear at a high frequency across the subsection, but Group 4 haplotypes diverge early in a maximum parsimony phylogenetic tree. Consid- ering the sequence similarity between all haplo- types (and the possibility of homoplasic muta- tions) we are hesitant to speculate as to the identity of ancestral haplotypes. The geological history of Central California is too complex to easily track the migration of populations in the Chorizanthe pungens-C. robus- ta complex, but there is some indication in the cpDNA data that plants moved south to north, as more haplotype diversity is found in the southern populations. Some Group | haplotypes are shared between southern Monterey Bay and Santa Cruz Mountains populations that are not found in the intermediary coastal C. robusta populations, so plants may have radiated both up the coast from the Monterey area as well as from an inland approach into the Santa Cruz Moun- tains, perhaps via the Salinas Valley where an ancient dune field was connected to the still preserved relict dunes of Fort Ord (Smith et al. 2005). This will have to remain an open question unless C. pungens var. pungens from rare popu- lations in the Salinas Valley can be found. However, many of these populations might be extirpated, leaving only very old herbarium specimens for future analysis. The cpDNA analysis, while not providing independent verification, does not contradict the ITS phylogeny which suggests that the Santa Cruz Mountain varieties and the coastal/coastal terrace varieties of the complex diverged from a common ancestor. The Group | haplotype in the Pogonip Park 2 population of Chorizanthe robusta var. robusta near its northern extreme could point to an ancestral biogeographical connection with the montane varieties which are dominated by the Group | haplotypes. It is not obvious from the data whether this C. robusta var. robusta haplotype, unique in the Pogonip Park 2 population, is an artifact of incomplete lineage sorting (as 1s probable with the Group | haplotypes in southern C. pungens var. pungens populations) or is due to hybridization with a montane population. The latter explanation is MADRONO [Vol. 56 supported by the fact that this population has a high percentage of pinkish flowers which is common in the montane varieties but less so in the general population of C. robusta var. robusta. Group 2 haplotypes were most likely fixed in the mid-range populations of Chorizanthe robusta var. robusta through genetic drift. The clinal transition from Group 2 to predominately Group 3 haplotypes in the southern C. robusta var. robusta populations argues for an extremely close relationship to C. pungens var. pungens where Group 3 haplotypes predominate, especially at the northern extreme of its range where it overlaps with the southernmost population of C. robusta var. robusta. In this area of overlap there is also less morphological distinction between the two. One exception is C. robusta var. robusta from Manresa State Beach (and from one area at the north end of Sunset State Beach). These plants are consistently so large they could never be confused with any of the other known populations. A possibility that needs to be examined further, especially in light of the genetic similarity of these two varieties, is that the larger Chorizanthe robusta var. robusta may have arisen from genome duplication in C. pungens var. pungens (Reveal, Cornell University, personal communication). Polyploidy would thereby act as a reproductive barrier between the two varieties despite their close proximity in foredune and backdune habitat at Sunset State Beach. Karyotyping has been performed on many Chorizanthe species (Hard- ham 1989) but not C. robusta var. robusta. Possible Taxonomic Revisions As alluded to earlier, there is a long history of disagreement regarding the classification of the \ | Chorizanthe species and varieties that are now | included in subsect. Pungentes (see Reveal and | Hardham 1989; Reveal and Morgan 1989; Ertter | 1996). The subsection was named by Goodman (1934) who assigned to it C. pungens, C. howellii, C. valida, C. douglasii, C. stellulata, C. villosa, C. angustifolia, C. cuspidata, C. diffusa and C. robusta. His revision was a marked departure | from that of Jepson (1914) who suggested that the latter three species be reduced to varieties under C. pungens. Subsequent revisions reduced C. villosa to a variety of C. cuspidata (Munz 1958), removed C. douglasii and C. stellulata to | the new sister subsection Legnota (Reveal and | Hardham 1989), and added the hartwegiana varietal to C. pungens (Reveal and Hardham | 1989) and the hartwegii varietal to C. robusta | (Reveal and Morgan 1989). The ITS sequence | phylogeny does not fully support any of these | classification schemes. Based solely on the ITS data, one could | reasonably collapse all of the current Chorizanthe | 2009] pungens-C. robusta complex into varieties of C. pungens, the species having taxonomic precedence (Bentham 1836). Such a revision would not be a departure outside the limits of previous classifi- cations. The cpDNA data, despite evidence of incomplete lineage sorting, show many shared haplotypes and clinal haplotype distributions between these varieties which are indicative of a close evolutionary relationship and generally supportive of the ITS phylogeny. Final resolution of the classifications of the two varieties of C. cuspidata will require more analysis, but in light of the current data and the historical treatment of these varieties the validity of C. cuspidata as a species designation should be reexamined. Although we focused on subsect. Pungentes, two observations suggest that a phylogenetic reevaluation of sect. Pte/osepala 1s in order. First, because of its genetic distance from the rest of the subsection, it is possible that Chorizanthe diffusa belongs in another subsection. Second, analysis of the single C. douglasii specimen (currently in subsect. Legnota) places it within the main group of Pungentes species. An analysis of C. ste/lulata will need to be performed to determine whether one or both of these members of subsect. Legnota should be returned to Pungentes. During the course of this study, we confirmed by ITS sequencing that an erect form of Chorizanthe (mistakenly identified as C. robusta var. robusta) at Point Reyes National Seashore was, in fact, C. cuspidata var. villosa which is endemic to the area. The confusion most likely arose from microhabitat differences as these plants were in sheltered areas off the coast. Also, C. cuspidata var. villosa has Group 2 cpDNA haplotypes (X, Y and Z) which are very similar to those found in the Santa Cruz Co. C. robusta var. robusta and the Baker Beach C. cuspidata var. cuspidata. This may explain morphological sim- ilarities especially when combined with variations in habitat. Implications for Conservation and Management In a recent recovery plan (U.S. Fish and Wildlife Service 2004), Chorizanthe rebusta var. robusta populations were grouped into four geographic recovery units (Point Reyes, Northern Santa Cruz, Aptos and Southern Santa Cruz). Based on our cpDNA haplotype data, the Aptos Recovery Unit (which includes the Freedom population in this study) could be combined with the genetically similar Northern Santa Cruz Recovery Unit (which includes the Pogonip and Branciforte populations). The Point Reyes Re- ‘covery Unit should be omitted entirely based “upon our analysis of the ITS region from several putative Point Reyes C. robusta var. robusta specimens, all of which had the same sequence as _C. cuspidata var. villosa that is endemic to the BRINEGAR AND BARON: CHORIZANTHE SUBSECT. PUNGENTES PHY LOGENETICS 181 area. Removal of the Point Reyes Recovery Unit will have a significant impact on the estimate of C. robusta var. robusta plants. Classifying populations of recently evolved taxa for the purpose of protection and manage- ment can be difficult if not impossible (Ertter 1997), as was chronicled by Ertter (1996) who described the process which led to the listing of the entire Chorizanthe pungens-C. robusta com- plex (U.S. Fish and Wildlife Service 1994) despite the inability of experts to come to a consensus on classification. This decision has provided a high degree of protection to a group that has long been an important component of coastal ecosys- tems and has conserved populations that have radiated into new habitats or persisted as habitats changed. Although the current study has shed light on the evolutionary relationships of species in subsect. Pungentes, it has raised questions about others. However, our results support and high- light that subsect. Pungentes is indeed a group with a high degree of evolutionary adaptation and recent change. There are a number of populations that are in geographically close, yet ecologically different, habitats. These plants exhibit sometimes minor morphological and genetic differences—attributes helpful in adapt- ing to changing environments. Consequently, our results support the current practice of protecting multiple small and often genetically diverse populations. ACKNOWLEDGMENTS This project was funded primarily by U.S. Fish and Wildlife Service (USFWS) contracts #101813Q101 and #-801017M276. Additional funding was provided by Point Reyes National Seashore contract ##R 8539030259. We are indebted to Connie Rutherford of the Ventura USFWS office for her support in initiating this project. For collecting permission, advice or other assistance we would like to thank the City of Scotts Valley, the City of Santa Cruz, Jane Rodgers (Point Reyes National Seashore), Melanie Gogol-Prokurat and Jeannine DeWald (California Department of Fish and Game), Renee Pasquinelli, Vince Cicero, George Gray and Kenneth Gray (California Depart- ment of Parks and Recreation), Vern Yadon, David Styer and Doreen Smith (California Native Plant Society), Randy Morgan, David Imper (USFWS), Dr. James Reveal (Cornell University), Dr. Barbara Ertter (Jepson Herbarium, U.C.-Berkeley), Bruce Delgado (Bureau of Land Management) and Peter Brastow (National Park Service). We thank Maliheh Movassat for providing expert technical assistance for some of the genetic analyses. LITERATURE CITED BENTHAM, G. 1836. On the Eriogoneae, a tribe of the order Polygonaceae. Transactions of the Linnaean Society of London 17:401—420. BLATTNER, F. 1999. Direct amplification of the entire ITS region from poorly preserved plant material 182 using recombinant PCR. BioTechniques 27:1180-1185. CHIANG, T. Y., K. H. HUNG, T. W. HSU, AND W. L. Wu. 2004. Lineage sorting and phylogeography in Lithocarpus formosanus and L. dodonaeifolius (Fagaceae) from Taiwan. Annals of the Missouri Botanical Garden 91:207—222. CLEMENT, M., D. POSADA, AND K. A. CRANDALL. 2000. TCS: a computer program to estimate gene genealogies. Molecular Ecology 9:1657—1659. CoMES, H. P. AND R. J. ABBOTT. 2001. Molecular phylogeography, reticulation, and lineage sorting in Mediterranean Senecio Sect. Senecio (Astera- ceae). Evolution 55:1943—1962. CORNMAN, R. S. AND M. L. ARNOLD. 2007. Phyloge- ography of Jris missouriensis (Iridaceae) based on nuclear and chloroplast markers. Molecular Ecol- ogy 16:4585-4598. CRANDALL, K. A. AND A. R. TEMPLETON. 1993. Empirical tests of some predictions from coalescent theory with applications to intraspecific phylogeny reconstruction. Genetics 134:959—969. CuBAS, P., C. PARDO, AND H. TAHIRI. 2006. Morpho- logical convergence or lineage sorting? The case of Cytisus purgans auct. (Leguminosae). Taxon 55:695—704. EAsSTwoob, A. 1903. New species of western plants. Bulletin of the Torrey Botanical Club 30:483—502. ERTTER, B. 1996. Saga of the Santa Cruz spineflowers. Fremontia 24:8—11. . 1997. Taxonomic grays vs. black and white expectations: implications for conservation man- agement of diversity. Pp. 11-13 in T. N. Kaye, A. Liston, R. M. Love, D. L. Luoma, R. J. Meinke, and M. V. Wilson, (eds.), Conservation and management of native plants and fungi. Native Plant Society of Oregon, Corvallis, OR. EXCOFFIER, L. AND P. E. SMOUSE. 1994. Using allele frequencies and geographic subdivision to recon- struct gene trees within a species: molecular variance parsimony. Genetics 136:343—359. FEHLBERG, S. D. AND T. A. RANKER. 2009. Evolu- tionary history and phylogeography of Encelia farinosa (Asteraceae) from the Sonoran, Mojave and Peninsular Deserts. Molecular Phylogenetics and Evolution 50:326—335. FELSENSTEIN, J. 1985. Confidence limits on phyloge- nies: an approach using the bootstrap. Evolution 39:783-791. GONZALES, E., J. L. HAMRICK, AND S. M. CHANG. 2008. Identification of glacial refugia in south- eastern North America by phylogeographic analy- ses of a forest understorey plant, Tri/limm cuneatum. Journal of Biogeography 35:844-852. GOODMAN, G. J. 1934. A revision of the North American species of the genus Chorizanthe. Annals of the Missouri Botanical Garden 21:1—102. GRIVET, D. AND R. J. PETIT. 2002. Phylogeography of the common ivy (Hedera sp.) in Europe: genetic differentiation through space and time. Molecular Ecology 11:1351—1362. HARDHAM, C. B. 1989. Chromosome numbers of some annual species of Chorizanthe and related genera (Polygonaceae:Eriogonoideae). Phytologia 66:89— 94. Honjo, M., S. UENO, Y. TSUMURA, I. WASHITANI, AND R. OHSAWA. 2004. Phylogeographic study based on intraspecific sequence variation of chlo- MADRONO [Vol. 56 roplast DNA for the conservation of genetic diversity in the Japanese endangered species Primula _ sieboldii. Biological Conservation 120:211—220. Hupbson, R. R. AND J. A. COYNE. 2002. Mathematical consequences of the genealogical species concept. Evolution 56:1557—1565. HUELSENBECK, J. P. AND D. M. HILLIS. 1993. Success of phylogenetic methods in the four-taxon case. Systematic Biology 42:247—-264. INAMURA, A., Y. OHASHI, E. SATO, Y. YODA, T. MASUZAWA, M. ITO, AND K. YOSHINAGA. 2000. Intraspecific sequence variation of chloroplast DNA reflecting variety and geographical distribu- tion of Polygonum cuspidatum (Polygonaceae) in Japan. Journal of Plant Research 113:419-426. JAKOB, S.S. AND F. R. BLATTNER. 2006. A chloroplast genealogy of Hordeum (Poaceae): long-term per- sisting haplotypes, incomplete lineage sorting, regional extinction, and the consequences for phylogenetic inference. Molecular Biology and Evolution 23:1602—1612. JEPSON, W. L. 1914. Polygonaceae. Pp. 376-428 in A flora of California, volume 1, part 4. Jepson Herbarium and Library, University of California, Berkeley, CA. KIMBALL, R. T., D. J. CRAWFORD, AND E. B. SMITH. 2003. Evolutionary processes in the genus Coreo- carpus: insights from molecular phylogenetics. Evolution 57:52-61. KIMURA, M. 1980. A simple method for estimating evolutionary rates of base substitutions through comparative studies of nucleotide sequences. Jour- nal of Molecular Evolution 10:111—120. KNOWLES, L. L. AND B. C. CARSTENS. 2007. Delim-. iting species without monophyletic gene trees. Systematic Biology 56:887—895. LAMB FRYE, A. S. AND K. A. KRON. 2003. rbcL | phylogeny and character evolution in Polygona- ceae. Systematic Botany 28:326—332. LEVIN, R. A. 2000. Phylogenetic relationships within Nyctaginaceae tribe Nyctagineae: evidence from | nuclear and chloroplast genomes. Systematic Bot- | any 25:738—750. LYONS-WEILER, J. AND M. C. MILINKOVITCH. 1997. A | phylogenetic approach to the problem of differen- | tial lineage sorting. Molecular Biology and Evolu- tion 14:968—975. MADISON, W. P. AND L. L. KNOWLES. 2006. Inferring | phylogeny despite incomplete lineage sorting. Systematic Biology 55:21—30. Mort, M. E., D. J. CRAWFORD, AND K. N. FAIR- | FIELD. 2004. Phylogeny and character evolution in California Coreopsis (Asteraceae): insights from. morphology and from sequences of the nuclear and. plastid genomes. Systematic Botany 29:782-—789. Munz, P. A. 1958. California miscellany. IV. Aliso | 4:87—100. PARRY, C. C. 1884. Chorizanthe R. Brown. A revision of the genus, and rearrangement of the annual) species — with one exception, all North American. Proceedings of the Davenport Academy of Natural Science 4:45—63. REVEAL, J. L. AND C. B. HARDHAM. 1989. A revision of the annual species of Chorizanthe (Polygonaceae: Eriogonoideae). Phytologia 66:98—-198. AND R. MORGAN. 1989. A new combination in, Chorizanthe robusta C. Parry (Polygonaceae: Erio- 2009] gonoideae) from California. Phytologia 67:357— 360. ROSENBERG, R. R. 2002. The probability of topological concordance of gene trees and species trees. Theoretical Population Biology 61:225—247. SANCHEZ, A. AND K. A. KRON. 2008. Phylogenetics of Polygonaceae with an emphasis on the evolution of Eriogonoideae. Systematic Botany 33:87—96. SoLTIS, D. E. AND R. K. KUZOFF. 1995. Discordance between nuclear and chloroplast phylogenies in the Heuchera group (Saxifragaceae). Evolution 49:727-742. SMITH, D. P., K. GREF, A. HOFMANN, AND L. TURRINI-SMITH. 2005. Are “‘stable shorelines”’ and ‘“‘broad beaches” mutually exclusive manage- ment goals along Southern Monterey Bay? Report No. WI-2005-09, The Watershed Institute, Califor- nia State University Monterey Bay, Seaside, CA. TAMURA, K., J. DUDLEY, M. NEI, AND S. KUMAR. 2007. MEGA4: Molecular Evolutionary Genetics Analysis (MEGA) software version 4.0. Molecular Biology and Evolution 24:1596—1599. TAN, B., K. Lru, X. L. YUE, F. Liu, J. M. CHEN, AND Q. F. WANG. 2008. Chloroplast DNA variation and phylogeographic patterns in the Chinese endemic marsh herb Sagittaria potamogetifolia. Aquatic Biology 89:372—378. TEMPLETON, A. R., K. A. CRANDALL, AND C. F. SING. 1992. A cladistic analysis of phenotypic associa- tions with haplotypes inferred from restriction endonuclease mapping and DNA sequence data. III. Cladogram estimation. Genetics 132:619—633. THOMPSON, J. D., T. J. GIBSON, F. PLEWNIAK, F. JEANMOUGIN, AND D. G. HIGGINS. 1997. The BRINEGAR AND BARON: CHORIZANTHE SUBSECT. PUNGENTES PHYLOGENETICS 183 ClustalX windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Research 24:4876-4882. U.S. FISH AND WILDLIFE SERVICE. 1994. Endangered and threatened wildlife and plants: endangered status for three plants and threatened status for one plant from sandy and sedimentary soils of central and coastal California. Federal Register 59:5499-S511. —., 2004. Recovery plan for Chorizanthe robusta var. robusta (robust spineflower). U.S. Fish and Wildlife Service, Portland, Oregon, vii + 71 pages. WANG, L., R. J. ABBOTT, W. ZHENG, P. CHEN, Y. WANG, AND J. Liu. 2009. History and evolution of alpine plants endemic to the Qinghai-Tibetan Plateau: Aconitum gymnandrum (Ranunculaceae). Molecular Ecology 18:709—721. WOJCHIECHOWSKI, M. F., M. J. SANDERSON, AND J. M. Hu. 1999. Evidence on the monophyly of Astrag- alus (Fabaceae) and its major subgroups based on nuclear ribosomal DNA ITS and chloroplast DNA trnL intron data. Systematic Botany 24:409—437. WOLFE, A. D. AND W. J. ELISENS. 1996. Evidence of chloroplast capture and pollen-mediated gene flow in Penstemon section Peltanthera (Scrophularia- ceae). Systematic Botany 20:395—412. YAMANAKA, M., S. KOBAYASHI, AND H. SETOGUCHI. 2008. Distinct geographical structure across species units evidenced by chloroplast DNA haplotypes and nuclear ribosomal ITS genotypes of Corylopsis (Hamamelidaceae) in the Japanese islands. Botan- ical Journal of the Linnean Society 157:501—518. MADRONO, Vol. 56, No. 3, pp. 184-198, 2009 PLANT COMMUNITIES AND FLORISTIC DIVERSITY OF THE EMERALD LAKE BASIN, SEQUOIA NATIONAL PARK, CALIFORNIA PHILIP W. RUNDEL, MICHAEL NEUMAN, AND PETER RABENOLD Department of Ecology and Evolutionary Biology, University of California, Los Angeles CA 90095 rundel@biology.ucla.edu ABSTRACT The Emerald Lake Basin forms a subalpine watershed in the upper drainage of the Marble Fork of the Kaweah River in Sequoia National Park. The basin is 120 ha in area and covers an elevational range from 2804 m at Emerald Lake to 3416 m at the summit of Alta Peak. The flora of the basin includes 202 vascular plant species, distributed into 132 genera and 41 families, with the Asteraceae (25 species) and Poaceae (23 species) as the largest families. Herbaceous perennials make up about three-fourths of the flora, but unlike alpine habitats many of these are relatively tall upright species, particularly in wet habitats. The woody plant flora includes five tree species, all conifers, with Pinus monticola as the dominant. There are 19 species of woody shrubs present, including six evergreen and 13 winter deciduous species. Eight plant habitat types were delineated on the basis of geomorphic position related to soil conditions and water availability that influence species composition (Billings 1974). These were subalpine conifer forest, willow thicket, wet meadows, moist rock crevices, dry meadows, dry rock crevices, colluvium, and fellfields. Key Words: Fellfield, Pinus jeffreyvi, Sierra Nevada, subalpine flora, subalpine vegetation. Subalpine and alpine basins of the Sierra Nevada provide important ecosystems not only for their biodiversity but also for monitoring environmental impacts of global change resulting from higher temperatures, altered patterns of snowmelt, and the deposition of anthropogenic nitrogen and phosphorus. These high mountain systems are sensitive to small changes in growing season conditions of temperature and water availability (Bowman and Saestedt 2001; Malan- son et al. 2007). Additionally, there is little potential for nitrogen uptake by the limited vegetation cover growing on shallow granitic soils and the short growing season. Low levels of nitrogen sequestration and deposition of anthro- pogenic nitrogen onto the winter snowpack produces pulses of nitrogen associated with snowmelt into aquatic systems, and thus an early warning indicator of nitrogen saturation for downstream forested basins (Sickman et al. 2003b; Williams et al. 1995). The Emerald Lake Basin in Sequoia National Park provides an excellent case study for ecosystem stability in a small subalpine watershed. The Emerald Lake Watershed Study (1984-1989) was organized by the California Air Resources Board as a means of better understanding the influence of atmospheric inputs of nutrients to basin processes and ecosystem structure (Tonnessen 1991). These studies have provided detailed and in many cases long-term continuing databases to understand hydrologic flow and water balance (Kattelmann and Elder 1991), nutrient enrichment (Sickman et al. 2003a, b), solute chemistry of snowmelt and runoff (Williams and Melack 1991), nitrogen fluxes and transformations (Williams et al. 1995), and long-term models of hydrochemical respons- es (Wolford et al. 1996; Meixner et al. 2004). Our objective in this paper is to describe the floristic diversity of the Emerald Lake Basin and to characterize the mesotopographic distribution of plant communities within the basin. MATERIALS AND METHODS Study Site The Emerald Lake watershed comprises a rugged basin about 120 ha in area in the upper drainage of the Marble Fork of the Kaweah River above Tokopah Falls in Sequoia National Park (36°35'49"N lat., 118°40’30’W long.). The | basin is roughly triangular in shape trending | northwest to southeast, with elevations ranging | from 2804 m at Emerald Lake at the lower end of the basin to the apex of the triangle at the summit | of Alta Peak at 3416 m (Figure 1). Two spurs | emanating from Alta Peak form ridges along the | northeastern boundaries of the » drainage, while the lower drainage boundaries are | southern and open to the northwest. The Emerald Lake Basin is a glacial cirque, | carved from granitic parent material. Bedrock ; exposed by glacial scouring and frost action | covers nearly half of the basin surface, with the | remaining half covered by talus and thin soils in | approximately equal proportions (Sisson and Moore 1984; Tonnessen 1991). Exposed rock faces in the basin contain many smaller fracture joints, providing cracks where residual soils can | 2009] RUNDEL ET AL.: EMERALD LAKE BASIN VEGETATION AND FLORA 185 y = Met Station California / Emerald Lake Inflow Streams 42 40 38 36 34 32 | D 425-120 -415 as) \ Alta Peak \ 3416 m \ N 1 m x Contour interval 25 m Fic. 1. The Emerald Lake Basin, Sequoia National Park, California. Emerald Lake at the lower margin of the basin lies at 2804 m elevation while Alta peak at the southeast corner is the high point at 3416 m. Map by T. Meixler and used with permission. collect. The basin is typical of many subalpine and alpine lake basins in the Sierra Nevada with its weakly buffered calcium bicarbonate surface waters (Williams and Melack 1991). Median slope angles in the basin are about 30°, but many of the upper slopes are considerably steeper. A major fracture in the granite forms a large joint that extends across the lower eleva- tions of the basin in a southwest to northeast direction. Emerald Lake itself, 2.7 ha in area, is a tarn formed by glacial quarrying of fractured rock along this joint. There is a smaller lake at 2963 m elevation above Emerald Lake near the geographical center of the basin. The climate of the Emerald Lake Basin 1s typical of the Mediterranean-type regime of the southern Sierra Nevada, with 75-90% the annual precipitation falling as snow in the winter months (Stephenson 1990). For Emerald Lake, rainfall comprises only about 10% of annual precipita- tion and occurs predominantly in the autumn. There are limited long-term data on precipitation levels present in the basin. Mean annual precip- itation is about 1600 mm, but amounts are highly variable between years (Sickman et al. 2003a). Emerald Lake is typically covered by ice from November to June or July. Snowmelt typically begins in April with peak discharge typically occurring in June (Sickman et al. 2003b). Summers are generally dry except for an occa- sional convective storm associated with mon- soonal air masses from the east. Field Studies The field studies collecting the data presented in this paper were carried out between June and September in 1984, 1985, and 1987. Over this period all parts of the basin were repeatedly visited to provide as complete an assessment as possible of the vascular plant flora. These field studies identified, characterized and mapped a series of plant communities determined by a combination of their physiographic position within the watershed, soil accumulation, and seasonal water availability. These communities broadly resemble the mesotopographic gradient of alpine vegetation described by Billings (1974). Each of these communities had distinctive floristic and plant life-form dominance. No attempt was made to assign these communities to previous classifications of subalpine and alpine vegetation alliances and associations for the Sierra Nevada (see Sawyer and Keeler-Wolf 2007) as these applied relatively poorly to the Emerald Lake Basin because of broad ecotonal gradients between communities. Representative biomass samples were collected from each community and a complete census and diameter measurement of all trees in the basin was completed. The scientific names used in this paper follow those of The Jepson Manual (Hickman 1993), with the exception of the familial classification where the Liliaceae and Amaryllidaceae are treated as separate families here. RESULTS Flora The flora of the Emerald Lake Basin includes 202 vascular plant species, distributed into 130 genera and 42 families (Table 1). The largest 186 MADRONO [Vol. 56 TABLE 1. SUMMARY OF THE FLORA OF THE EMERALD LAKE BASIN, SEQUOIA NATIONAL PARK, SIERRA NEVADA, CALIFORNIA. Division Families | Genera Species Pteridophyta + 10 fa Coniferophyta 1 2 5 Dicotyledonae 32 93 139 Monocotyledonae 5 22 47 Total 42 130 (2 family present is the Asteraceae with 25 species, followed in order by the Poaceae (23 species), Scrophulariaceae (16 species), Cyperaceae (12 species), and Rosaceae (11 species). These five families comprise 44% of the total flora. Notable is the absence of any species of Fabaceae, as other alpine floras typically have this family well represented (Rundel et al. 2008). At the generic level, Carex forms the largest group with 10 species. Other genera with four or more species include Mimulus (6 species), Epilobium (4 spe- cies), and Saxifraga (4 species). The Pteridophyta are represented by 11 species in the basin, all herbaceous perennials (Table 1, Appendix 1). The ferns are most important in wet meadows and mesic rock crevices where Crypto- gamma acrostichoides, Athyrium alpestre, Cystop- teris fragilis, and Woodsia scopulorum are all common. Other pteridopytes are more typical of dry meadows and dry rock crevices, as with Aspidotis densa, Pellaea bridgesii, and Selaginella watsonii. Ferns also are common in colluvial habitats, as described below. Five species of coniferous trees are present in the basin (Table 1, Appendix 1), with four of these pines. More than 70% of these trees are Pinus monticola, with Pinus contorta subsp. murrayana and P. balfouriana ssp. austrina comprising 17% and 9.5% of the trees respective- ly. Only a few individuals of Pinus jeffreyi and Abies magnifica var. shastensis are present. Angiosperms form more than 92% of the basin flora, with a total of 186 species (Table 1). Dicots include three-fourths of the angiosperm total (139 species), while monocots provide the other quarter (47 species). Among the latter, graminoid life forms of grasses, sedges, and rushes form the dominant element. Life Forms Woody plants make up only a moderate proportion of the flora of the Emerald Lake Basin. As described above, there are five species of evergreen trees present, all conifers. There are 19 species of woody shrubs present, 9% of the flora, with six of these evergreen shrubs and the other 13 species as winter deciduous shrubs (Appendix 1). The evergreen shrubs include five species of Ericaceae and Fagaceae—Arctostaph- ylos nevadensis, Chrysolepis sempervirens, Kalmia polifolia, Ledum glandulosum, and Phyllodoce breweri. The first two of these are largely restricted to growing within the open pine stands of the basin. Kalmia polifolia is found in willow thickets and wet meadows, while L. glandulosum and P. breweri can be found in a variety of communities. The sixth species of woody shrub, barely qualifying beyond being a subshrub, is Eriogonum wrightii (Polygonaceae), a low-grow- ing dry meadow species. The most prominent deciduous shrub in the basin 1s Salix orestera, which forms extensive willow thickets. A number of other deciduous shrubs are less abundant and generally exhibit a lower stature. These include five species of Rosaceae—Amelanchier utahensis, Holodiscus mi- crophyllus, Prunus emarginata, Sorbus californica, and Spiraea densiflora. Also present are Lonicera conjugialis, L. involucrata, and Sambucus race- mosa (Caprifoliaceae), Ribes cereum and R. montigenum (Grossulariaceae), Jamesia america- na (Philadelphaceae), and Acer glabrum (Acer- aceae). An additional eight species (4% of the flora) in the basin can be classified as suffrutescent subshrubs. These are Aster breweri and Erica- meria discoidea (Asteraceae), Epilobium canum (Onagraceae), Leptodactylon pungens, and Phlox diffusa (Polemoniaceae), Primula suffrutescens (Primulaceae), and Penstemon newberryi (Scro-— phulariaceae). These subshrubs are most charac- | teristic of more xeric habitats such as pine forests, | dry meadows, dry rock crevices and colluvial | sites. | The overwhelming life forms for floristic | dominance in the basin are herbaceous perenni- | als, which comprise three-fourths of the total | flora. These are species that characteristically die | back to ground level at the end of each growing | season. Within this broadly defined category are > rosettes, broad-leaved tussocks, perennial grami- | noids, geophytes, mats and cushions, and bien- | nials. The 156 species of herbaceous perennials | include 11 pteridophytes, seven monocot geo- phytes, 39 graminoids, and 98 herbaceous peren- | nial dicots. At the subalpine elevations of the. Emerald Lake Basin, many perennial dicot species are upright and relatively tall, in contrast | to low-growing species that characterize the. alpine environments. | The seven species of geophytes present are in. the Amaryllidaceae and Liliaceae (A//ium cf. | campanulatum, A. obtusum, Fritillara pinetorum, | Lilium kellevanum, Smilacina racemosa, Veratrum californicum, Zygadenus venenosus). Six species of | herbaceous perennials are hemi-parasites. These | are Castilleja applegatei, C. miniata, C. nana, | Pedicularis atollens, and P. semibarbata (Scro-. phulariaceae), and Orobanche uniflora (Oroban-. 2009] chaceae). Pterospora andromedea (Ericaceae) is a ‘“‘saprophyte’’, epiparasitic through fungal myce- lium on vascular plant roots. Annual plants are infrequent in the basin, with 14 species comprising 7% of the flora. Annuals present at Emerald Lake are Gnaphalium palustre (Asteraceae), Cryptantha sp. (Boraginaceae), Cuscuta californica (Cuscutaceae), Phacelia eise- nii (Hydrophyllaceae), Gavophyton humile (Ona- graceae), Gilia capillaris and Linanthus. ciliatus (Polemoniaceae), Collinsia torreyi, Mimulus bre- wert, M. laciniatus, and M. whitneyi (Scrophular- iaceae), Galium bifolium (Rubiaceae), Saxifraga bryophora (Saxifragaceae), and Muhlenbergia filiformis (Poaceae). Cuscuta californica is an annual parasitic vine. The annual flora is most apparent in dry meadow habitats, although there are certainly conspicuous species present in both wet meadows and wet rock crevices. Vascular Plant Communities Subalpine conifer forest. Although scattered conifers, particularly pines, are widespread in the Emerald Lake Basin, an open community of subalpine conifer forest is well developed only in a small area in the extreme northeast corner of the basin near Emerald Lake. The four conifers species present in this community are Pinus Jeffreyi. Abies magnifica, Pinus monticola, and a few individuals of Pinus contorta subsp. mur- rayana. The thin soils of the pine sites are sandy in texture and very poor in organic matter, allowing these soils to dry quickly in the summer when snowmelt ceases. This condition may account for the relatively small number of species, 30 in total, encountered in this commu- nity. Beyond this small area of conifer forest, scattered individuals of Pinus monticola are widespread on the mesic benches of the south- west-facing slopes of the basin. In addition to these southwest-facing slopes, P. monticola also occurs as scattered trees on relatively flat dry meadows below the cirque of Alta Peak. There are several areas of the basin where P. monticola appears to be slowly expanding its range by colonizing dry meadows and dry crevice habitats. A fifth conifer, Pinus balfouriana subsp. austrina, is restricted to a relatively small area on the ridge running along the south- and southwest-facing margin of the basin. The understory of the conifer forest stands is typically dominated by Chrysolepis sempervirens, which locally exhibits ground cover up to 80— 100%. In more open forest stands, there is a low cover of dry meadow herbs, most notably Jvesia santalinoides and Achillea millefolium. On the lower slopes of the basin, the stands of C. sempervirens mix with Phyllodoce breweri at the margins of dry meadow habitats. Also present in RUNDEL ET AL.: EMERALD LAKE BASIN VEGETATION AND FLORA 187 these stands are Arctostaphylos nevadensis, Ame- lanchier utahensis, Prunus emarginata, and Sorbus californica. Although Lonicera conjugialis may be present, it is more common growing along wet rock crevices. None of the conifer forest commu- nities has a unique herbaceous flora, with those species present also typical of dry meadow habitats. Willow thicket. Willow thickets are widespread across the Emerald Lake Basin and dominate significant areas of the basin. Along with Salix orestera, the indicator species, there are a characteristic set of understory shrubs and herbs, which typically cover 100% of the ground surface. Characteristic species include ericaceous shrubs such as Ledum glandulosum, Kalmia polifolia, Phyllodoce breweri, and Vaccinium nivictum. Tall herbs present include Senecio triangularis, Epilobium angustifolium, Lilium kel- leyanum, Ligusticum grayi, and Calamagrostis canadensis. The high productivity of leaf litter in this habitat has lead to the development of thick and hydric organic soils. The willow thicket community typically intergrades to wet meadow communities around its margins. During our study we observed that willows appeared to be colonizing wet meadow and wet rock crevice habitats in several areas. With a dominance of woody plant cover and associated tall herbaceous perennials, the species richness of willow thickets is relatively low. Only 28 species were encoun- tered. Wet meadows and moist rock crevices. Wet meadow and moist rock crevice communities form diverse habitats in the basin, and exhibit a number of distinctive associations. These associ- ations are typically distributed in relatively flat areas on west-facing slopes above Emerald Lake where soils accumulate to moderate depths and remain moist for much or all of the growing season. Variations in soil texture, soil organic matter content, and soil moisture dynamics appear to be the primary physical factors separating different wet meadow associations. The largest areas of wet meadow habitat are located high in the basin, adjacent to and above Parson’s Pond, where snowmelt keeps the soils saturated well into the summer. Other large areas of wet meadows occur on a bench running southwest from the major area of Pinus mon- ticola, and adjacent to willow thickets along the main drainage of the basin. The high species richness of wet meadows and moist rock crevices, with 105 species present, makes for a large number of indicator species. The most important of these for wet meadows are Senecio triangularis, Calamagrostis canadensis, Carex spectabilis, C. nigricans, Vaccinium nivic- tum, Aster alpigenus, Dodecatheon jeffreyi, D. subalpinum, Eriophorum criniger, Juncus merten- 188 MADRONO sianus, and the moss Polytrichum juniperinum. Areas with saturated soils in early summer often support stands of tall herbs such as Veratrum californicum, Mertensia ciliata, Sphenosciadium capitellatum, and Helenium bigelovii. Although there is a significant overlap of wet meadow species with those of dry meadows, a condition not surprising in view of community gradients of moisture availability and soil depth, the high cover of wet meadow associations is often but not invariably a distinguishing feature. In some areas, wet meadows grade gradually into willow thickets, while other areas exhibit a transitional association between wet and dry meadows formed by a heather turf occurring in the granitic joint along the east face of the basin. This latter community is dominated by Phyllo- doce breweri, with Vaccinium nivictum, Juncus parryi and several dry meadow herbs as associ- ated species. While P. breweri is widespread across many habitats in the basin, only in this area does it form the dominant groundcover. Moist rock crevice communities are concen- trated on the east face of the basin, on benches along the major joint area, and in various pockets of rocky talus. These are areas below wet meadows that provide sufficient moisture inputs to support the growth of mesic shrubs and herbs. While plants cover the total surface along these moist crevices, the crevices themselves cover less than 5% of the rock surface where they occur. Characteristic species in these habitats include Sambucus racemosa, Lonicera involucrata, Mer- tensia ciliata, Aquilegia pubescens, and Thalictrum fendleri. Although plant cover over the crevices themselves is virtually 100%, only about 5% of the rock faces provide crevice habitats. A related habitat includes a group of species that grow among large boulders 1—3 m in diameter in slide areas which maintain moist soils throughout the summer. In addition to the species above Lonicera conjugialis, Aquilegia formosa and Helenium bigelovii are commonly present, while smaller herbaceous perennials are largely absent. Dry meadows and dry rock crevices. Dry meadow associations occupy a relatively large area of the Emerald Lake Basin, most notably in the relatively flat plateau and benches below the northeast ridge and boundary of the basin. In these sites, soils are formed of shallow layers of decomposed granite with low organic matter, and typically dry early in the growing season as indicated by the early senescence of herbaceous species. There are also significant areas of dry meadow northeast of Parson’s Pond, adjacent to the stand of Pinus monticola in the northeast corner of the basin, in the northeast fault area, and in small, scattered areas of relatively flat topography on south-facing slopes. [Vol. 56 The overall diversity of dry meadow and dry crevice associations is surprisingly high, with 84 species encountered. Indicator species for the dry meadow habitats include Jvesia_ santalinoides, Juncus parryi, Dicentra nevadensis, Eriogonum incanum, Calyptridium umbellatum, Sedum obtu- satum, and grass species such as Achnantherum occidentale and Muhlenbergia filiformis. The density and cover of dry meadow associations is highly variable, with moisture relations a critical factor (Klikoff 1965; Burke 1982; Benedict 1983). Much of the area of dry meadows has only 20— 25% plant cover, although the dry meadow northeast of Parsons Pond has 50% or more cover. Small stands of Dicentra nevadensis high on the east slope of the basin approach 100% groundcover. Dry rock crevices are widespread all across the fractured areas of granite faces in the Emerald Lake Basin, but highly diffuse in coverage and account for only a very small total area. These habitats, usually relatively narrow cracks no more than 50 cm wide, support a number of specialist species as well as herbaceous perennials common to dry meadows. While these crevices appear dry at the surface, the presence of suffrutescent perennials suggests good water availability deep in the crevices. The characteris- tic dry crevice species include Penstemon new- berryi, Holodiscus microphyllus, Spiraea densi- flora, and Sedum obtusatum. Eriogonum nudum, Ivesia pygmaea and a variety of dry meadow herbs may also be present but are eclipsed in> coverage by the larger shrubs. As with the wet | rock crevices, plant cover is often close to 100% | along these cracks, but these crevices cover no> more than 5% of the rock surface. | Colluvium. The colluvium association occurs in | a few steeply sloped areas along the fault trace on | the southwestern boundary of the basin and in | two smaller areas to the south where large | boulders change to smaller granitic boulders | 0.3—-1.0 m along their major axis. The 26 species | present occurred in areas with shallow soil | development between the boulders. Most of the highly heterogeneous plant cover, which averaged | about 25% of ground surface, was composed of. small herbaceous perennials up to 20 cm in) height. Species which dominated this habitat both | in abundance and cover were Senecio fremontii, Eriogonum incanum, Carex lanuginosa, Phlox | diffusa, and Erysimum capitatum. The first of these is particularly important, making up a’ major part of the cover. Two pteridophytes, | Cryptogramma acrostichoides and Selaginella watsonii, grow along the margins of the boulders | and provide significant cover. Two species, | Primula suffrutescens and Anaphalis margarita- | cea, appear to be restricted to the colluvium habitat. | 2009] Fellfields. Relatively little area of alpine fell- field is present in the Emerald Lake Basin. This community is found at high elevation on a plateau below Alta Peak, a site with Pinus balfouriana on a north-facing slope overlooking Pear Lake. Compared to the colluvium habitat, the fellfield has a ground surface scattered with smaller granite rocks and a large area of exposed soil. Unlike the colluvium habitat where plants have colonized virtually all of the available areas of bare soil, this fellfield is very sparsely vegetated and low in diversity. Nowhere is plant cover greater than about 10%, with a typical cover closer to 5%. Seven species were encountered here—Chaenactis alpigena, Carex helleri, Ranun- culus escscholtzti, Minuartia nuttallii, Eriogonum incanum, Saxifraga tolmiei and Trisetum spica- tum. The first two of these were encountered nowhere else in the basin. DISCUSSION The relatively large species diversity of the Emerald Lake Basin is consistent with similar levels of diversity reported for other high Sierra Nevada basins. Burke (1982) reported the pres- ence of 277 plant species in the Rae Lakes Basin which covers approximately 6 km?’ with eleva- tional ranges from 3046-4040 m. Cheng (2004) provided summary descriptions of U.S. Forest Service Research Natural Areas in the central Sierra Nevada that included subalpine meadow and alpine talus and scree slope communities. Three of these, Clark Fork (1869-3826 m eleva- tion, 874 ha), Highland (2650-2815 m elevation, 178 ha), and Snow Canyon (2499-3001 m eleva- tion, 285 ha) contained 227, 200, and 223 taxa, respectively. These levels of biodiversity can be explained only partially by the elevational gradi- ents included within the basins. More important is the diversity of habitat conditions present despite a low level of plant cover and biomass. The relative abundance of woody species in the Emerald Lake Basin provides evidence of the less stressful conditions of this subalpine area com- pared with true alpine communities where shrub species are rare (Rundel et al. 2008). Neverthe- less, herbaceous perennials do form the largest life form in the flora as they do in alpine habitats. In both subalpine and alpine habitats, herba- ceous perennials have the characteristic of maintaining large proportions of total biomass belowground where they play an important role in carbohydrate storage over the winter months (Mooney and Billings 1960; Billings 1974; Rundel et al. 2005). Interesting among the Poaceae in the Emerald Lake Basin is the presence of three species of Muhlenbergia, grasses with C4 metabolism. These include two perennial and one annual species. Although Cy, metabolism in grasses is normally RUNDEL ET AL.: EMERALD LAKE BASIN VEGETATION AND FLORA 189 characteristic of low, subtropical habitats, Muh- lenbergia forms an exception with species reach- ing to alpine environments (Sage and Sage 2002). The 7% of the flora represented by annual plant species at Emerald Lake 1s similar to typical levels of about 8% reported for the summer dry alpine areas of the Sierra Nevada and White Mountains (Jackson 1985; Jackson and Bliss 1982; Rundel et al. 2008). Annual species are rare in typical circumboreal arctic-alpine floras of the Northern Hemisphere where they commonly comprise only 1—2% of the flora (Billings 2000). Ecological and biogeographic data on high- mountain and related alpine meadows in the Sierra Nevada have recently been summarized (Fites-Kaufman et al. 2007; Sawyer and Keeler- Wolf 2007), as has the general state of knowledge on Sierra Nevadan meadows in an earlier report (Ratliffe 1985). The diverse associations of subalpine meadow communities in the Emerald Lake Basin illustrate the complexity of adapting simple systems of community classification as proposed in previous studies for specific basins (Pemble 1970; Burke 1982; Benedict 1983; Taylor 1984). Clearly soil depth and the amount and seasonality of soil moisture availability are critical components of species distribution. A detailed study of the floristics and physiographic classification of seven subalpine meadows in Sequoia National Park found that these factors as well as soil frost action and conditions of water chemistry, temperature, and depth of flooding explained the major part of floristic variation (Benedict and Major 1982; Benedict 1983). Subalpine dry meadow associations have been described for other areas of the Sierra Nevada (Khkoff 1965; Rathff 1979, 1982; Taylor 1976; Burke 1982), but with only limited understanding of controlling factors for species distribution. The plant communities of the Emerald Lake Basin would fall within a variety of associations as described in the classification system of Sawyer and Keeler-Wolf (1995). Lower areas of the basin would be classified as Mixed Subalpine Forest, while associated ericaceous shrublands and wil- low thickets would be termed the Montane Heather—Bilberry, Montane Wetland Shrub, or Holodiscus Series. The majority of the basin would fall under “‘habitat” classifications of this system as Subalpine Upland Shrub habitat, Subalpine Wetland Shrub habitat, Subalpine Meadow habitat and Alpine habitat. These categories are much too crude, however, to appropriately separate the plant associations present in the Emerald Lake Basin. With continuing concerns about global change, nutrient enrichment, and acid deposition in high mountain basins such as that of Emerald Lake, it is important to have baseline data on floristic diversity and ecological communities. Continuing 190 studies in the Emerald Lake Basin can provide a means of identifying and assessing future envi- ronmental changes. ACKNOWLEDGMENTS We thank the California Air Resources Board for supporting this research. Gail Baker, Evan Edinger, Carmen Crivellone, and Karen Poulin provided major assistance in carrying out the field work for this project. The assistance of Kathy Tonnessen, Dave Parsons, and the staff of Sequoia and Kings Canyon National Parks is acknowledged with thanks. LITERATURE CITED BENEDICT, N. B. 1983. Plant associations of subalpine meadows, Sequoia National Park, California. Arctic and Alpine Research 15:383—396. AND J. MAJoR. 1982. A physiographic classi- fication subalpine meadows of the Sierra Nevada, California. Madrono 29:1—12. BILLINGS, W. D. 1974. Adaptations and origins of alpine plants. Arctic and Alpine Research 6:129-142. . 2000. Alpine vegetation. Pp. 536-572 in M. G. Barbour and W. D. Billings (eds.), North Amer- ican terrestrial vegetation. 2nd ed. Cambridge University Press, Cambridge, UK. BOWMAN, W. D. and T. R. SEASTEDT (eds.). 2001. Structure and function of an alpine ecosystem: Niwot Ridge, Colorado. Oxford University Press, Oxford, UK. BURKE, M. T. 1982. The vegetation of the Rae Lakes Basin, southern Sierra Nevada. Madrono 29:164—176. CHENG, S. (ed.). 2004. Forest Service Research Natural Areas in California. General Technical Report PSW-GTR-188. U.S. Department of Agriculture, Forest Service, Pacific Southwest Research Station, Albany, CA. FITES-KAUFMAN, J. A., P. RUNDEL, N. L. STEPHEN- SON, AND D. A. WEIXELMAN. 2007. Montane and subalpine vegetation of the Sierra Nevada and Cascade ranges. Pp. 456-501 in M. G. Barbour, T. Keeler-Wolf, and A. A. Schoenherr (eds.), Terres- trial vegetation of California, 3rd ed. University of California Press, Berkeley, CA. HICKMAN, J. (ed.) 1993. The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. JACKSON, J. L. 1985. Floristic analysis of the distribu- tion of ephemeral plants in treeline areas of the western USA. Arctic and Alpine Research 17:251—260. AND L. C. BLIss. 1982. Distribution of ephemeral herbaceous plants near treeline in the Sierra Nevada, California, USA. Arctic and Alpine Research 14:33-44. KATTELMANN, R. AND K. ELDER. 1991. Hydrologic characteristics and water balance of an alpine basin in the Sierra Nevada. Water Resources Research 27:1553-1562. KLIKOFF, L. G. 1965. Microenvironmental influence on vegetational pattern near timberline in the central Sierra Nevada. Ecological Monographs 35: 187-211. MALANSON, G. P., D. R. BUTLER, D. B. FAGRE, S. J. WALSH, D. F. TOMBACK, L. D. DANIELS, L. M. MADRONO [Vol. 56 RESLER, W. K. SMITH, D. J. WEIss, D. L. PETERSON, A. G. BUNN, C. A. HIEMSTRA, D. LIPTZIN, P. S. BOURGERON, Z. SHEN, AND C. I. MILLAR. 2007. Alpine treeline of western North America: linking organism-to-landscape dynamics. Physical Geography. 28:378—396. MEIXNER, T., C. GUTMANN, R. BALES, A. LEyY- DECKER, J. SICKMAN, J. MELACK, J., AND J. MCCONNELL. 2004. Multidecadal hydrochemical response of a Sierra Nevada watershed: sensitivity to weathering rate and changes in deposition. Journal of Hydrology 285:272—285. Mooney, H. A. AND W. D. BILLINGS. 1960. The annual carbohydrate cycle of alpine plants as related to growth. American Journal of Botany 47:594—-5S98. PEMBLE, R. H. 1970. Alpine vegetation in the Sierra Nevada of California as lithosequences and in relation to local site factors. Ph.D. dissertation. University of California, Davis, CA. RATLIFFE, R. D. 1979. Meadow sites of the Sierra Nevada, California: classification and species relationships. Ph.D. Dissertation, New Mexico State University, Las Cruces, NM. 1982. A meadow site classification for the Sierra Nevada, California. USDA Forest Service, General Technical Report, PSW-60. 1985. Meadows in the Sierra Nevada of California: state of knowledge. General Technical Report, PSW-84. U.S. Department of Agriculture, Forest Service, Pacific Southwest Research Station, Albany, CA. RUNDEL, P. W., A. C. GIBSON, AND M. R. SHARIFI. 2005. Plant functional groups in alpine fellfield habitats of the White Mountains, California. Arctic, Antarctic and Alpine Research 37:358—365. ; , AND . 2008. The alpine flora of the White Mountains, California. Madrono 55: 204-217. SAGE, R. F. AND T. L. SAGE. 2002. Microsite characteristics of Muhlenbergia richardsonis (Trin.) Rydb., an alpine Cy grass from the White Mountains, California. Oecologia 132:501—508. SAWYER, J. O. AND T. KEELER-WOLF. 2007. Alpine vegetation. Pp. 539-573 in M. G. Barbour, T. | Keeler-Wolf, and A. A. Schoenherr (eds.), Terres- . trial vegetation of California, 3rd ed. University of | California Press, Berkeley, CA. SICKMAN, J. O., J. M. MELACK, AND D. W. CLOw. 2003a. Evidence for nutrient enrichment of high- elevation lakes in the Sierra Nevada, California. Limnology and Oceanography 48:1885—1892. SCHIMEL. 2003b. Mechanisms underlying export of N from high-elevation catchments during seasonal | transitions. Biogeochemistry 64:1—24. SISSON, T. W. AND J. G. Moore. 1984. Geology of Giant Forest-Lodgepole Area, Sequoia National | Park, California. Geologic Investigation Report | OF 84-0254, U.S. Geological Survey, Denver, CO. | STEPHENSON, N. L. 1990. Climatic control of vegeta- | tion distribution: the role of the water balance. | American Naturalist 135:649—670. TAYLOR, D. W. 1976. Ecology of the timberline vegetation at Carson Pass, Alpine County, Cali- | fornia. Ph.D. Dissertation. University of Califor- | nia, Davis, CA. , A. LEYDECKER, C. C. Y. CHANG, C. KEN- | DALL, J. M. MELACK, D. M. LUCERO, AND J. | 2009] . 1984. Vegetation of the Harvey Monroe Hall Research Natural Area, Inyo National Forest, California. Unpublished report on file, U.S. Department of Agriculture, Forest Service, Pacific Southwest Research Station, Albany, CA. TONNESSEN, K. A. 1991. The Emerald Lake watershed study: introduction and site description. Water Resources Research 27:1537—1539. WILLIAMS, M. W., R. C. BALES, A. D. BROWN, AND J. M. MELACK. 1995. Fluxes and transformations RUNDEL ET AL.: EMERALD LAKE BASIN VEGETATION AND FLORA 19] of nitrogen in a high-elevation catchment, Sierra Nevada. Biogeochemistry 28:1—31. AND J. M. MELACK. 1991. Solute chemistry of snowmelt and runoff in an alpine basin, Sierra Nevada. Water Resources Research 27:1575—1588. WOLFORD, R. A., R. C. BALES, AND S. SOROOSHIAN. 1996. Development of a hydrochemical model for seasonally snow-covered alpine watersheds: Applica- tion to Emerald Lake Watershed, Sierra Nevada, California. Water Resources Research 32:1061—1074. [Vol. 56 MADRONO g Juniperus californica A: “Plat of the Rancho Muscupiabe finally confirmed to Miguel White, surveyed under instructions from the U.S. Surveyor General by Henry Hancock, Dep. Sur., November 1867.” B: Relocated survey stations on the boundary of Rancho Muscupiabe that record prominent native vegetation. RESULTS AND DISCUSSION Only prominent native vegetation was record- ed on the 1867 map of Rancho Muscupiabe (Fig. 2). Platanus racemosa is recorded as ‘‘Syc- amore” on the Rancho boundary at Nealeys Corner, Devore Heights, Tokay Hill, Cable Canyon, Badger Canyon, and Little Sand Can- yon. The ““Manzanita bush”’ recorded for another boundary survey station at Devore Heights is most likely Arctostaphylos glauca Lindl. A station recorded as ““Granite Rock in juniper bushes Sta. 32." documents Juniperus californica. This site was relocated at Lat. 34°9'36.51”N, Long. 117°14'46.61"W, on the Harrison Mountain quadrangle, adjacent to section 24, TIN, R4W, San Bernardino Base Meridian, approximately 900 m NW of Little Sand Creek at Del Rosa. Juglans californica was recorded at two locations on the boundary survey. East of Lytle Creek, “Walnut Sta. 25” is in Devil Canyon at Lat. 34°12'56.66"N, Long. 117°19'45.67"W, adjacent to Section 32, TZN, R4W, on the San Bernardino North quadrangle. West of Lytle Creek, ““Walnut Sta. 42” is at the SE base of San Sevaine Ridge at Lat. 34°10'51.23"N, Long. 117°26'40.68"W, ad- jacent to section 7, TIN, RSW, approximately 10 km NE of Etiwanda, on the Devore quadran- 202 MADRONO Ras eye | tt ee Bo Rehign bs Juniperus californica sk 1867 @ Extant FIG. 3. gle. It is unlikely that these identifications were made in error because the species are morpho- logically distinct from other native taxa. Juniperus californica was not known to have occurred in alluvial fan scrub at Del Rosa (Minnich and Everett 2001). Widely disjunct stands occur to the west on alluvial fans below the San Gabriel Mountains at Azusa, Upland, Cucamonga Canyon, and Lytle Creek (Fig. 3). East of Lytle Creek along the base of the San Bernardino Mountains, scattered stands occur on terraces in Cajon Creek, and in the Santa Ana River wash between Redlands and Mentone. Nearest-neighbor distances are 7 to 10 km SE to scattered individuals along alluvial terraces of the Santa Ana River Wash between Redlands and Mentone (Vasek and Thorne 1988, p. 819-821; Clarke et al. 2007), 13 km W to stands in alluvial fan scrub along Cajon Creek (Minnich and Everett 2001), ca. 15 km SE at toa reported locality at San Timoteo Canyon (Clarke et al. 2007), 21 kmStoan occurrence at the Box Springs Mountains, and 27 km NW to extensive populations near Cajon Junction (Minnich and Everett 2001). The cismontane region of San Bernardino County is the easternmost range of Juglans californica (Jepson 1908, 1917; Munz 1974). East of Juglans californica woodlands in the Chino Hills, widely scattered extant localities occur on alluvial fans at the southern base of the San Gabriel Mountains, and eastward along the foothills of the San Bernardino Mountains from Cajon Pass to Banning Pass (Fig. 4). A collection of Juglans californica by L. C. Wheeler in 1942 (RSA610986) for a locality reported as “San Gabriel Mountains region, foot of San Gabriel Mountains: 1/2 mile SW mouth of Lytle Creek,” matches with the 1867 Rancho Mucupiabe bound- [Vol. 56 Extant distribution of Juniperus californica in the vicinity of Rancho Muscupiabe. ary survey location of ““Walnut Sta. 42” near the SE end of San Sevine Ridge (Fig. 4). East of Lytle creek, the CCH botanical collection database mapped extant occurrences of Juglans californica in foothills of the San Bernardino Mountains at Devil Canyon, and eastward at Badger Canyon (3 km SE), Waterman Canyon (5 km E), near Seven Oaks Dam (25 km SE), Mentone (26 km SE), Oak Glen Creek (31 km SE), Mill Creek Canyon (37 km E), Yucaipa (38 km SE), and Millard Canyon near Banning Pass (56 km SE). Mapped occurrences of Juniperus californica and Juglans californica on the Rancho Muscu- piabe boundary survey of 1867 predate local botanical collections of these species first recorded in the late nineteenth century (Table 1). The CCH database lists the earliest documented botanical collection of Juniperus californica in the cismon- tane region of San Bernardino County by S. B. and W. F. Parish (JEPS48231) on 3 Jan 1882 for the locality of ““San Bernardino Valley.”’ The earliest collection of Juglans californica in San Bernardino County was by Marcus E. Jones (POM 123336) on 28 Apr 1882 for the locality, ““Coastal Plains and Basin; San Bernardino Basin region Colton.””! 'The CCH database reports an erroneous collection date of 19 Aug 1824 for a different botanical specimen of Juglans californica (POM 122548), since the collector, Marcus E. Jones, was born in 1852 (see Lenz 1986). POM 122548 was most likely collected 19 Aug 1924, when Marcus E. Jones was working at Claremont College, nearby the locality of ““Cajon Pass’’ where POM 122548 is reported to have been collected. The Rancho Santa Ana Botanic Garden Herbarium holdings include two other collections by him on 19 Aug 1924 at Cajon Pass (POM 122540 & POM70013). The specimen in question lists his personal collection number as s.7., as with his two other collections from Cajon Pass that day. 2009] = ae 655, b Fianna ina AO ine la Lb y nda ra TPN iis Mey cori 28 GOFORTH AND MINNICH: RANGE EXTENSIONS AT RANCHO MUSCUPIABE Le Biz Nee LC Te i cep a ad Soin a ey ae Bl i) i) QW L/ aR. % SS aaa Po ‘ert eA SKA a od sae Se mien TTS af 7 iT Nb ‘ : j F Sey é JANA ta Tot Ss a ee te ¢ Kan CON gran PEA plant ; vain >. wel 4 BAN ane mR OINO POs i ei . ‘RIVERS! ais Ve baw 2 fm ar, AN Juglans californica 31867 @ Extant Fic. 4. Extant distribution of Juglans californica in the vicinity of Rancho Muscupiabe. Note the location of Etiwanda circled on the map. Historical Ecologists Grove and Rackham (2001, 18) state, “landscape history is best arrived at from the records of identifiable sites, which can be traced down the centuries in the archives and compared with what is there now.” The Rancho Muscupiabe boundary survey in 1867 mapped site-specific localities of Juniperus californica and Juglans californica that can be more precisely relocated than historical botanical collections of these species which reported localities in general reference to the San Bernar- TABLE 1. CISMONTANE REGION OF SAN BERNARDINO Co. dino region (Table 1). Comparison of extant distributions to the historical survey map indicate that subsequent land use extirpated Juniperus californica from the Del Rosa locality near Little Sand Creek, while Jug/ans californica persisted at Devil Canyon where land use has been less intensive due to rugged topography. The distribution data compiled in this study build upon comprehensive atlases of California vegetation, which did not map the easternmost distribution of Juglans californica in the San EARLY BOTANICAL COLLECTIONS OF JUNIPERUS CALIFORNICA AND JUGLANS CALIFORNICA IN THE , ““Coastal Plains and Basin; San Species Accession no. Collector Juglans californica POM 123336 Marcus E. Jones Juglans californica JEPS58690 S. B. and W. F. Parish Juglans californica UC847334 Marcus E. Jones Juglans californica UCS56653 H. M. Hall Juglans californica JEPS58695 W. L. Jepson Juniperus californica JEPS48231 S. B. and W. F. Parish Juniperus californica UC87 S. B. Parish Juniperus californica JEPS48334 S. B. Parish Juniperus californica | JEPS48333 S. B. Parish Juniperus californica JEPS48324 W. L. Jepson Juniperus californica POM9881 I. M. Johnston Locality Date 28 Apr 1882 Bernardino Basin region Colton” “foothills San Bernardino Mts.” Apr 1888 “Colton” 28 Apr 1892 ‘between San Bernardino and Apr 1902 Cajon Pass” “Cajon Pass”’ 27 May 1914 “San Bernardino Valley” 3 Jan 1882 “San Bernardino Valley” I Feb: 1889 “San Bernardino Southern 1 Feb 1893 California, San Bernardino” “San Bernardino Southern 1 Jan 1894 California, San Bernardino” “Cajon Pass”’ 28 May 1914 “Transverse Ranges; San Gabriel 20 Nov 1916 Mts region Foothills above Upland. Dry wash” 204 Bernardino Mountains. The vegetation type map (VIM) survey of California (Weislander 1935) was conducted in the Rancho Muscupiabe area from 1929-1930, and did not map Juglans californica on either the San Bernardino 15 min quadrangle (164C) nor the Redlands 15 min quadrangle (164D). Griffin and Critchfield (1972) used VIM data to map the distribution of trees in California, and their maps do not show Juglans californica in the foothills of the San Bernardino Mountains. Instead, they suggest that ‘one apparently natural disjunct colony occurs to the east of the main population in San Bernar- dino County (Chino Hills) near Etiwanda.”’ Etiwanda is located on a cismontane alluvial fan below Cucamonga Peak at the base of the eastern San Gabriel Mountains (Fig. 4). A synthesis of data provided by the Rancho Muscupiabe boundary survey in 1867, early botanical collections (Table 1), historical taxo- nomic studies (Jepson 1908, 1917), and mapped extant occurrences (Fig. 4) provide a fuller picture, showing that the eastern distributional range of Juglans californica extends along the southern base of the San Bernardino Mountains to Millard Canyon near Banning Pass. LITERATURE CITED BURKE, J. H., C. E. JONES, W. A. RYAN, AND J. A. WHEELER. 2007. Floodplain vegetation and soils along the upper Santa Ana River, San Bernardino County, California. Madrono 54:126—137. CLARKE, O. F., D. SVEHLA, G. BALLMER, AND A. MONTALVO. 2007. Flora of the Santa Ana River and environs. Heyday Books, Berkeley, CA. GROVE, A. T. AND O. RACKHAM. 2001. The nature of Mediterranean Europe: an ecological history. Yale University Press, New Haven, CT. GRIFFIN, J. R. AND W. B. CRITCHFIELD. 1972. The distribution of forest trees in California. U.S. Department of Agriculture, Forest Service Re- search Paper PSW-82. Pacific Southwest Forest and Range Experiment Station, Berkeley, CA. HANCOCK, H. 1867. Plat of the Rancho Muscupiabe. On file, Tomas Rivera Library Special Collections, map case, second drawer, file A2-F1l: Historical maps of Riverside and San Bernardino, University of California, Riverside, CA. HANES, T. L., R. D. FRIESEN, AND K. KEANE. 1989. Alluvial scrub vegetation in coastal southern California. Pp. 187-193 in D. L. Abell (ed.), Proceedings of the California riparian systems conference: protection, management, and restora- tion for the 1990s; 1988 September 22—24; Davis, CA. U.S. Department of Agriculture, Forest Service General Technical Report PSW-GTR-110. MADRONO [Vol. 56 Pacific Southwest Forest and Range Experiment Station. Berkeley, CA. JEPSON, W. L. 1908. The distribution of Juglans california Wats. Bulletin of the Southern California Academy of Sciences 7:23—24. 1917. The native walnuts of California. Madrono 1:55—57. KEELEY, J. E. 1990. Demographic structure of Califor- nia black walnut (Juglans Californica; Juglanda- ceae) woodlands in southern California. Madrono 37:237-248. LENZ, L. W. 1986. Marcus E. Jones: western geologist, mining engineer & botanist. Rancho Santa Ana Botanic Garden, Claremont, CA. LESKINEN, C. A. 1972. Juglans californica: local patterns in southern California. M.S. Thesis, University of California, Los Angeles, CA. MINNICH, R. A. AND R. G. EVERETT. 2001. Conifer tree distributions in southern California. Madrono 48:177-197. MULLALLY, D. P. 1992. Distribution and environmen- tal relations of California black walnut (Juglans californica) in the eastern Santa Susana mountains, Los Angeles County. Crossosoma 18:1—17. Munz, P. A. 1974. A flora of southern California. University of California Press, Berkeley, CA. QUINN, R. D. 1990. The status of walnut forests and woodlands (Juglans californica) in southern Cali- fornia. Pp. 42-54 in A. A. Schoenherr (ed.), Endangered plant communities of southern Cali- fornia: proceedings of the 15th annual symposium, October 28, 1989. Southern California Botanists Association, California State University, Fullerton, CA. RAup, H. F. 1940. San Bernardino, California: settlement and growth of a pass-site city. Univer- sity of California Publications in Geography 8:1—64. ROBERTS, F. M., S. D. WHITE, A. C. SANDERS, D. E. BRAMLET, AND S. BoyD. 2004. The vascular plants of western Riverside county, California: an anno- tated checklist. FM Roberts Publications, San Luis Rey, CA; RUNDEL, P. 2007. Sage scrub. Pp. 208-228 in M. G. Barbour, T. Keeler-Wolf, and A. A. Schoenherr (eds.), Terrestrial vegetation of Califor- nia, 3rd ed. University of California Press, Berke- ley, CA. SMITH, R. L. 1980. Alluvial scrub vegetation of the San Gabriel River Floodplain, California. Madrono 27:126-138. VASEK, F. C. AND R. F. THORNE. 1988. Transmontane coniferous vegetation. Pp. 797-832 in M. G. Barbour and J. Major, (eds.), Terrestrial vegetation of California, 2nd ed. California Native Plant Society, Sacramento, CA. WEISLANDER, A. 1935. A vegetation type map of California. Madrono 3:140—144. MADRONO, Vol. 56, No. 3, pp. 205—207, 2009 ELEVATION OF PHACELIA CICUTARIA VAR. HUBBYI (BORAGINACEAE) TO SPECIES STATUS LAURA M. GARRISON!” AND ROBERT PATTERSON Department of Biology, San Francisco State University, 1600 Holloway Avenue, San Francisco, CA 94132 ABSTRACT Phacelia cicutaria Greene var. hubbyi (J. F. Macbr.) J. T. Howell warrants elevation from varietal to species status based on morphological evidence. It is distinguished morphologically from P. cicutaria var. cicutaria and P. cicutaria var. hispida (A. Gray) J. T. Howell by its more robust habit, more densely congested inflorescences, lack of mottled markings on the corolla, longer stamens and style, thick shaggy hairs on the stems and leaves, and calyx lobes that closely invest the fruit. Key Words: Boraginaceae, Hydrophyllaceae, Hydrophylloideae, Phacelia. The genus Phacelia (Boraginaceae subf. Hy- drophylloideae sensu Angiosperm Phylogeny Group [Stevens 2001 onwards]) is taxonomically diverse with approximately 200 species (Wilken et al. 1993). Species in the genus are often distin- guished by seemingly slight morphological differ- ences; however, these differences are generally stable within species. More recently, molecular sequence data have supported integrity of species that have been described based on morphology (Gilbert et al. 2005; Hansen 2005; Garrison 2007; Walden et al. 2008). In addition to the large number of species in the genus, a substantial number of subspecies, varieties, and formae have been described. While many of these are slight morphological variants of the species, occasionally exceptions appear where a named variety differs markedly from the other infraspecific taxa. During a detailed mo- lecular systematic study of the P. crenulata complex based on nrITS sequence data and morphological observations/data (Garrison 2007) it became clear that the taxon traditionally known as P. cicutaria var. hubbyi was appreciably distinct from the other two varieties, P. cicutaria var. cicutaria and P. cicutaria var. hispida. Phacelia hubbyi was described originally as a variety of P. hispida (which itself is now recognized as P. cicutaria var. hispida) by Macbride (1917) based on “‘... hirsute rather than hispid inflorescence with suberect pedicels’”” and “*... much softer hairiness.” All three taxa are native to California, and while their ranges overlap somewhat (Fig. 1), they are fundamen- tally discrete biogeographically. While Macbride did not mention other characters, Garrison’s (2007) analysis added a more robust habit, denser ‘Author for correspondence, email: Igarrison@ calacademy.org. *Present address: Department of Botany, California Academy of Sciences, San Francisco, CA 94118. inflorescences, absence of mottled marks on the corollas, and longer stamens and styles, thereby further distinguishing P. hubbyi from other purportedly related taxa. Garrison’s (2007) molecular phylogenetic re- search removes P. hubbyi from an alliance with the other varieties of P. cicutaria in favor of a closer relationship with P. distans Benth., P. tanacetifolia Benth., P. ramosissima Lehm., and P. umbrosa Greene (Garrison 2007). While a more complete phylogenetic study of Phacelia is underway (Garrison et al. unpublished), we present this taxonomic note in order for the new name to be included in the forthcoming second edition of The Jepson Manual of Califor- nia Plants. TAXONOMIC TREATMENT Phacelia hubbyi (J. F. Macbr.) L. M. Garrison, comb. et stat. nov. Phacelia hispida var. hubbyi J. F. Macbr. Contr. Gray Herb. 49:29. 1917.— Type: USA, California, Ventura County, Ojai Valley, 20 May 1896, Frank W. Hubby 31 (holotype, GH). Phacelia cicutaria Greene var. hubbyi (J. F. Macbr.) J. T. Howell. Leafl. W. Bot. 3:120. 1942. Phacelia tanacetifolia Benth. var. hubbyi (J. F. Macbr.) Jeps. & Hoover in Jepson. Fl. Calif. 3:258. 1943 DISCUSSION Phacelia hubbyi, traditionally referred to as P. cicutaria var. hubbyi, is morphologically distinct from other varieties of P. cicutaria. The most noticeable difference is in the pubescence of the inflorescence axes and calyces, which in P. hubbyi is wavy-hairy (shaggy-hirsute fide Jepson 1943) in contrast to the stiff-hairy pubescence in P. cicutaria var. cicutaria and P. cicutaria var. hispida. This feature alone is in line with typical 206 0 100mi_ ss on 0 100 km FIG. 1. MADRONO EQ a P. cicutaria var. hispida [Vol. 56 ea W P. hubbyi @ P. cicutaria var. cicutaria Baa TE EES CET ECLCP EEE EEE OT Distribution of Phacelia hubbyi, P. cicutaria var. cicutaria, and P. cicutaria var. hispida based on data from the Consortium of California Herbaria (http://ucjeps.berkeley.edu/consortium/about.html). Symbols represent collections from beyond the primary range of the taxon. differences distinguishing species in the genus (e.g., Constance 1951). Based on morphological evidence presented here and supporting molecular genetic evidence (Garrison 2007), P. hubbyi warrants elevation from varietal to species status. Additional mor- phological features distinguish P. hubbyi. Ra- cemes are much denser (i.e., flowers crowded more closely together) and remain so in fruit in P. hubbyi, while in P. cicutaria var. cicutaria and P. cicutaria var. hispida internodes between cap- sules, especially those at the proximal end of the inflorescence, tend to elongate as the inflores- cence matures. While this character is difficult to quantify, the differences are readily apparent in live and herbarium material. In fruit the calyx lobes in P. cicutaria var. cicutaria and P. cicutaria var. hispida spread away from the base of the capsule, while in P. hubbyi the calyx closely invests the capsule, as it does in P. tanacetifolia (Fig. 2). Hoover (1943) noted that a spreading calyx occurs also in P. cryptantha, while the calyx closely invests the capsule in P. ramosissima, P. tanacetifolia, and P. distans; this pattern is in accord with Garrison’s (2007) molecular phylogeny. Phacelia hubbyi has a particularly robust habit and stout stems as compared with P. cicutaria var. cicutaria and P. cicutaria var. hispida, as has been noted by other authors that have addressed fruit and calyx and calyx Phacelia cicutaria Phacelia tanacetifolia Fic. 2. Posture of calyx lobes in fruit of Phacelia hubbyi and P. tanacetifolia. Reprinted from The Jepson manual, J. Hickman, ed., 1993, with permission from the Jepson Herbarium. © Regents of the University of California. 2009] the taxon (“plant stout” [Voss 1935], “‘stem very stout” [Jepson 1943]). ACKNOWLEDGMENTS We thank Art Gibson and Andy Sanders for fruitful discussions, the Jepson Herbarium for permission to use illustrations from The Jepson Manual, and Gary Hannan and Diane Ferguson for helpful comments. LITERATURE CITED CONSTANCE, L. 1951. Hydrophyllaceae. Pp. 476-532 in L. Abrams (ed.), Hlustrated flora of the Pacific states, Vol. 3. Stanford University Press, CA. GARRISON, L. M. 2007. Phylogenetic relationships in Phacelia (Boraginaceae) inferred from nrITS se- quence data. Unpublished M.S. thesis. San Fran- cisco State Univ., CA. GILBERT, C., J. DEMPCyY, C. GANONG, R. PATTERSON, AND G. SPICER. 2005. Phylogenetic relationships within Phacelia subgenus Phacelia (Hydrophylla- ceae) inferred from Nuclear rDNA ITS sequence data. Systematic Botany 30:627—634. HANSEN, D. R. 2005. Phylogenetic relationships between and within Phacelia (Boraginaceae) Sec- tions Whitlavia and Gymnobythus. Unpublished M.S. Thesis. San Francisco State Univ., CA. GARRISON AND PATTERSON: PHACELIA CICUTARIA VAR. HUBBYI 207 HOoveER, R. F. 1943. Taxonomic note. Pp. 257 in W. L. Jepson (ed.), A flora of California, Vol. 3. Jepson Herbarium and Library. University of California, Berkeley, CA. JEPSON, W. L. 1943. A flora of California, Vol. 3. Jepson Herbarium and Library. University of California, Berkeley, CA. MACBRIDE, J. F. 1917. Notes on the Hydrophylla- ceae. Contributions from the Gray Herbarium 49:29. STEVENS, P. F. 2001 onwards. Angiosperm Phylogeny Website. Version 9, June 2008. Website http:// www.mobot.org/MOBOT/research/APweb _ [ac- cessed 30 August 2009]. Voss, J. W. 1935. A revisional study of the Phacelia hispida group. Bulletin, Southern California Acad- emy of Sciences 33:169-178. WALDEN, G. K., L. M. GARRISON, AND R. PATTER- SON. 2008. Phylogenetic analysis of Phacelia (Boraginaceae) inferred from chloroplast ndhF and nuclear ribosomal internal transcribed spacer (nrITS) region sequences (abstract). Botany 2008, Vancouver, B.C. WILKEN, D. H., R. R. HALSE, AND R. PATTERSON. 1993. Phacelia. Pp. 691—706 in J. C. Hickman (ed.), The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. MADRONO, Vol. 56, No. 3, p. 208, 2009 A NEW COMBINATION IN TRIFOLIUM VARIEGATUM (FABACEAE) MICHAEL A. VINCENT W.S. Turrell Herbarium (MU), Department of Botany, Miami University, Oxford, OH 45056 USA vincenma@muohio.edu ABSTRACT A new combination, Trifolium variegatum var. geminiflorum is proposed, and a key is provided for varieties of 7. variegatum. Key Words: California, Idaho, Oregon, Trifolium, Washington. In preparing the treatment of 7rifo/ium for the new Jepson Manual, it became necessary to make a new combination for the following, since no name for this taxon exists at the rank of variety. Trifolium variegatum Nutt. var. geminiflorum (Greene) Vincent, comb. nov. —T. gemini- florum Greene, Pittonia 3(17B): 216-217. 1897. —Lectotype (herein designated): USA, Califor- nia, Amador Co., New York Falls, April 1892, G. Hansen 1, (NDG [not seen]; isolectotype K!). Many names have been published, at the rank of species and variety, for taxa that fall within the range of morphological variation of what is now called Trifolium variegatum Nutt. The differences among the entities are slight and intergrade, except at the extremes of the range of morphol- ogy. In fact, even within a population, nearly the full range of variation can be seen, depending upon available water and light. In spite of this, some entities do seem to merit recognition, though not at the level of species. The varieties being recognized are var. variegatum, var. gemini- florum comb nov., and var. major Lojac., which may be distinguished as in the following key. 1. Inflorescence 1—5-flowered; corolla 3.5— STAM A ane Bs ae Se ae Bee we eed var. geminiflorum 1’. Inflorescence 5S—many-flowered; corolla 6—16mm 2. Inflorescence 1-1.5 cm, 5—10-flowered, corolla 6-10 mm.......... var. variegatum 2’. Inflorescence 1.5-3 cm, 10—many-flow- ered, corolla 9-lk/ mm. ....4:..4. var. major This variety, which ranges through much of California and north into Washington and Idaho, is equivalent to Isely’s 7. variegatum “phase 5”’, as treated in the current edition of the Jepson Manual (Isely 1993), and is equivalent to what was called “7. pauciflorum” by Lojacono (1883), though 7. pauciflorum Nutt. is a synonym of T. variegatum var. variegatum. A full discussion of the taxonomy of this group will be given in a forth-coming revision of Trifolium sect. Involucrarium Hook. (sensu Zoh- ary and Heller 1984). LITERATURE CITED ISELY, D. 1993. Trifolium (Fabaceae). Pp. 646—654 in J. C. Hickman (ed.), The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. LOJACONO, M. 1883. Revisione dei Trifogli dell’ Amer- ica settentrionale. Nuovo Giornale Botanico Ita- liano 15(2):113—198, plates HI-V. ZOHARY, M. AND D. HELLER. 1984. The genus Trifolium. Israel Academy of Sciences and Human- ities, Jerusalem, Israel. MADRONO, Vol. 56, No. 3, pp. 209-211, 2009 REVIEWS Systematics, Evolution, Compositae. Edited by V. A. FUNK, A. SUSANNA, T. F. STEUSSY, AND R. J. BAYER. 2009. Interna- tional Association for Plant Taxonomy, Univer- sity of Vienna, Rennweg 14, 1030, Vienna, Austria. xix + 965 pp (cloth). ISBN 978-3- 9501754-3-1. $110 (USA and Canada, $120 elsewhere). I tell students in my plant taxonomy course that if they are serious about being a taxonomist they must learn the Compositae. They cannot be daunted by the size and intricacy of the family— it is just too important a group to pass by. The editors of Systematics, Evolution, and Biogeogra- phy of Compositae state this more eloquently and topically: *“To understand biodiversity across our planet’s landscape requires understanding of this massive, diverse, and fascinating family.”’ The Compositae Book (as it is called on The Interna- tional Compositae Alliance (TICA) web site) is the most recent summary of what we know about this family. There have been earlier summarizing volumes (e.g., Heywood et al. 1977; Bremer 1994; Hind 1996), but this is the first to include results of analysis of extensive phylogenetic information based on molecular sequence data. The phylogenetic information presented is refer- enced largely to the (shall I say awesome?) Compositae supertree produced initially by Funk et al. (2005) and continually refined since. There is also a strong emphasis on biogeography of tribes and clades throughout the book. At first glance, the book is stunning—to look at and to lift (it weighs in at 3.785 kilos). At around 1000 pages it is loaded with information that is relatively easy to retrieve. Despite there being around 80 authors across 44 chapters plus five appendices, the book reads well. There is even the occasional light spirited inclusion, for example the remark from a famous synantherol- ogist that “‘speculating on base chromosome numbers offers, perhaps, the finest of all vehicles for intellectual auto-stimulation.”’ The book is richly illustrated with generally high quality photos and diagrams—the production of the photos on the pages is commendable. In particular there is a rich selection of photos of comps that are unfamiliar to many of us, exotic looking genera from around the world. As I perused the photos I could not help wondering which of these genera would survive in my yard on the central California coast. Seeing these images will impress readers who still believe that all members of the family are “doggone yellow comps.”’ and Biogeography of The inside back cover is a summary of the supertree (you can also download an 8 foot by 4 foot detailed version of the supertree at the TICA we site), the branches of which are differentially colored to reflect biogeographical affinity of higher taxa. On the inside front cover there is pocketed a removable color code piece to facilitate interpreting the biogeographic informa- tion on the summary tree on the inside back cover. Nice touch. Very little was left out in the making of this book. Section I, the introduction, contains chapters on the history of comp taxonomy, an interesting read that combines the history of taxonomic thought on the Compositae with personal, humanizing details of the early workers in the family (such as Vaillant being a popular botany teacher as evidenced by his lectures being given at six in the morning and attended by hundreds of listeners). There is also a fascinating chapter by Simpson on the economic importance of the family, despite the relative dearth of economically useful comps, given the size of the family. Section II contains chapters on character evolution in the family: chromosome numbers, secondary chemistry, microcharacters, pollen. The pollen chapter in particular 1s augmented by a list of characters measured, a data matrix of pollen characters used for supertree taxa, a summary of pollen features for each tribe, and an enormous bibliography containing over 1200 citations on composite pollen. Jeffrey’s chapter on evolution of Compositae flowers summarizes current thoughts on the topic. Many interpreta- tions that were held only a few decades ago have been radically overturned upon the appearance of molecular data. A final chapter in the section on oceanic island comps, Baker’s Law, polyploidy is sure to stimulate the interest of future researchers intrigued by island biogeography. Section III is really the meat of the book, beginning with a chapter on relationships of the family with other Asterales, followed by a chapter on classification of Compositae—this is a summary chapter for quick descriptions of tribes. The next thirty-two chapters cover, in supertree sequence, each tribe in the family, each written by active researchers specializing in that tribe. Chapters are organized for the most part in uniform pattern: historical overview; phylogeny; subtribal taxonomy; morphology; anatomy; pol- len; embryology; chromosome numbers; chemis- try; biogeography; evolution; economic uses. This systematic organization makes it easy to compare tribes across chapters. 210 Californian/western botanists (Madrono read- ership in large part) will find this book especially useful, considering the large diversity of comps in western North America. I found myself becoming immersed in the details of tribes with which I have a working familiarity. Even the staunchest synantherophobes will likely be fascinated by what we now know about relationships within the Heliantheae alliance, neatly summarized by Baldwin in Chapter 41. Likewise, the Cichorieae, generally easy to recognize as a tribe but beyond that often regarded by non-experts as a tangle of difficult genera, seem to be less intimidating after reading Chapter 24. The chapters that drew me the most were those covering the little-known and (still) incom- pletely studied tribes. Indeed, some of these tribes were erected precisely because modern data isolated them from groups with which they had been traditionally allied. For example, treatments of Gymnarrheneae, Pertyeae, Oldenburgieae, Corymbieae, and our own western Hecastoclei- deae not only fill in gaps in our knowledge of the family, they also highlight the numerous taxo- nomic side branches that have evolved. Similarly, the early chapter (Ch. 12) on the tribes at the basal end of the supertree gives a perspective on ancestral Compositae. With these phylogenetic revelations, we can get down to the business of studying character evolution within the family. At the end of the book there is a summary chapter that explains how the supertree was constructed, a brief description of all tribes, and a discussion of the age of the family and its area of origin and subsequent radiations. Here, and throughout the book, we are reminded that conclusions in the book are only the 2009 version, that there remains a lot to do, and that numerous labs are continuing to refine their studies. There are five appendices: an_ illustrated glossary of Compositae; the aforementioned pollen literature; original figure legends for plates in Chapter 1; new names and combinations; and a complete list of literature cited (74 pages). The illustrated glossary is overall useful, and nicely illustrated, but sometimes strange and inconsis- tent. With a subject as intricate as the Compo- sitae that has been studied by thousands of students, it is inevitable that an enormous terminology has evolved. It is equally inevitable that definitions vary across different authors. While this might not trouble experts in the family, novices are quite likely to become confused at the apparent lability of definitions. Take, for example, the term ligulate. Under the entry for Ligulate Floret the term describes a corolla shape. The following entry is Ligulate Head, at which readers are referred to Liguli- florous Head. The next entry, Ligule, describes the strap-shaped portion of a ligulate floret, MADRONO [Vol. 56 followed by the caveat that the term is “‘used in some references for the lamina of a ray floret.” While the entries mentioned go some distance toward differentiating ligulate, liguliflorous, and ligule, even in the glossary itself there are inconsistencies. Under the entry for Synoecious, for example, we read about ligulate heads. I would have liked to have read a_ direct statement to the effect that “‘ligulate’’ describes corollas (flowers), ‘“‘liguliflorous” describes heads, etc. Other oddities include the absence from the glossary of some specialized terms. Syncalathia and homo-/heterocalathia are not in the glossary but are used (yet, not defined) in Jeffrey’s chapter (Ch. 8) on evolution of composite flowers. On the other hand I was perplexed by the inclusion of terms so commonly used and understood (e.g., androecium, ovary) and that do not carry a special meaning in the context of the family. Curiously, the entry for Pollen Grains is largely a description of the plunger-pollination system in composite flowers. Sure, there are the very occasional blots, such as poorly type-set lines (cf. p 695, column 1, next to last line), inconsistent dates for literature in text and in citation, a misspelling or two, and the occasional quotation mark inside the period, but in reality these are likely to be noticed only by reviewers looking for mistakes. Overall, for a book this size with as many authors as it has, it is well-edited and presented. To purchase the book visit the TICA web site (http://www.compositae.org/) and download the order form. It’ll run you $110 (includes shipping!) in the U.S. and Canada, which my colleagues and students unanimously consider a bargain (re- member, 3.785 kilos of Compositae!). The cost is $120 for other areas. Visit YouTube for an impressive advert for the book (http://www. youtube.com/watch?v=mrlVVfW86gA). So who should/will buy this book? Taxono- mists, teachers, students, lay botanists, and other professionals will enjoy owning it. It stands to be a major reference for a long time, and a major inspiration for the next generation of composite systematists. If you’re new to the family, or if you have been ducking the Compositae for a while, this volume will change the way you look at this captivating family of plants. —BoB PATTERSON, Department of Biology, San Fran- cisco State University, 1600 Holloway Avenue, San Francisco, CA 94132. patters@sfsu.edu. LITERATURE CITED BREMER K. 1994. Asteraceae: cladistics and classifica- tion. Timber Press, Portland, OR FUNK V. A., R. J. BAYER, S. KEELEY, R. CHAN, L. WATSON, B. GEMEINHOLZER, E. SCHILLING, J. L. PANERO, B. G. BALDWIN, N. GARCIA-JACAS, A. SUSANNA, AND R. K. JANSEN. 2005. Everywhere 2009] but Antarctica: using a supertree to understand the diversity and distribution of the Compositae. Biologiske Skrifter 55:343—373. HeEywoop V. H., J. B. HARBORNE, AND B. L. TURNER (eds.). 1977. The biology and chemis- MADRONO, Vol. 56, No. 3, p. 211, 2009 BOOK REVIEWS 214 try of Compositae. Academic Press, London, U.K. HIND D. J. N. 1996. Proceedings of the International Compositae Conference, Kew 1994. Royal Botan- ical Gardens, Kew, U.K. NOTEWORTHY COLLECTION COLORADO COTONEASTER LUCIDUS Schltdl. (ROSACEAE).— Denver Co., yard weed, W 3lst Avenue, Denver, 1670 m, 3 Jul 2003, Zika 18517 (COLO, WTU); Boulder Co., Long Canyon, Boulder Mountain Park, 2130 m, 18 Aug 1991, Hogan 1514 (COLO); El Paso Co., forest edge, Cascade, 2300 m, 30 Aug 2006, Zika 22827 (COLO, DAO, NY, RM, WTU); El Paso Co., thickets, Manitou Springs, 1940 m, 31 Aug 2006, Zika 22873 (CS, WTU); Gilpin Co., aspen grove, Golden Gate Canyon State Park, 17 Jun 2000, Smookler & Senser 639 (KHD). Previous knowledge. Hedge or shiny cotoneaster is a cold-hardy native of Siberia and Mongolia. The fall foliage is attractive, and it is widely planted in northern North America, where often misidentified as C. acutifolius Turcz. Duplicates of Zika collections have been confirmed by Jeanette Fryer of Froxfield, Eng- land. Reports of C. franchetii Boiss. from Fort Collins, Larimer Co. (Ells, J. 2006. Rocky Mountain Flora. The Colorado Mountain Club Press, Golden.) are based on photos of immature fruits of C. lucidus. Significance. First records escaped from cultivation in Colorado. Dispersal of the species from gardens is presumably by avians. When C. lucidus was collect- ed in Cascade black-headed grosbeaks (Pheucticus melanocephalus) were observed opening fruits and eating the contents of the seeds, and American robins (Turdus migratorius) were observed swallowing whole fruits and disseminating the seeds. SORBUS AUCUPARIA L. (ROSACEAE).— EI Paso Co., cultivated in area and escaping to riparian zone, Ruxton Creek, near Fairview Ave., Manitou Springs, 1990 m, 4 Sep 2006, Zika 22875 (CS, MO, WTU): El Paso Co., Ruxton Creek riparian, Engelman Canyon, Manitou Springs, 2070 m, 4 Sep 2006, Zika 22876 (COLO, MTMG, WTU). Previous knowledge. Rowan, or European mountain ash, is native to Eurasia and commonly grown as an ornamental. It has been recorded as an adventive across northern North America. In Colorado it can be found growing with Sorbus scopulina Greene, a native species with sticky and shiny winter buds (varying from densely hairy to glabrous) and shiny upper leaflet surfaces. Sorbus aucuparia has dull leaflets and densely pubescent winter buds that are not shiny or sticky. Significance. First records escaped from cultivation in Colorado. —PETER F. ZIKA, WTU Herbarium, Box 355325, University of Washington, Seattle, WA 98195-5325. Zikap@comcast.net. MADRONO, Vol. 56, No. 3, p. 212, 2009 NOTEWORTHY COLLECTIONS CALIFORNIA HESPEREVAX ACAULIS (Kellogg) Greene var. 4MBUS- TICOLA Morefield (ASTERACEAE).—Los Angeles Co., San Gabriel River floodplain, Irwindale, Santa Fe Dam County Regional Park, along both sides of southwestern end of self-guided nature trail, ca. 0.3 km southwest of Interpretive Center, dwarf herbs in open, vernally-mesic flat of cryptogamic crust soils, on alluvial terrace of San Gabriel River, associated with Pectocarya_ penicillata, Crassula connata, Stylocline gnaphaloides, Schismus barbatus, Camissonia_ bistorta, Cryptantha intermedia, Navarretia hamata, Eriogonum gracile, Salvia columbariae, Centaurium venustum, and Erodium cicutarium, surrounding habitat alluvial scrub of Rhus integrifolia, Malosma laurina, Opuntia xvaseyi and Ribes aureum, 34°6'58.13"N, 117°56'43.63"W, 149 m, 14 March 2009, M. C. Long 890 (RSA). Previous knowledge. Fire evax is known from northern and central California but from very few sites in southern California (J. C. Hickman, ed. 1993. Jepson Manual, University of California Press, Berkeley; James Morefield personal communication). Recorded in southern California from Santa Barbara Co., Santa Ynez Mtns., San Marcos Pass, C. F. Smith 4436, Apr 25 1956 (JEPS, SBBG), western Riverside Co., Santa Rosa Plateau (E. W. Lathrop and R. F. Thorne. 1985. Flora of the Santa Rosa Plateau, Southern California Botanists Special Publication No. 1), northern San Diego Co., Aqua Tibia Mtns., Cutca Valley, Banks & Steinmann 1915, 13 April 1997 (RSA), and San Diego Co., Barber Mountain, 2854 ft. S. Bell 143, May 17, 1995 (SD). (Consortium of California Herbaria, ac- cessed April 2009). Significance. First collection for Los Angeles Co., CA, one of only a few collections from southern California, and fills an approximately 260 km gap in range between Santa Barbara Co. and western River- side Co. and northern San Diego Co. (James Morefield, personal communication). The plants were first found and identified by Jane Strong on 12 March 2009, and so far appear to be limited to a small, local population, 5 to 7m in extent along both sides of the nature trail, with an estimated 100-200 plants. Plants flowering and withering early with no plants located on repeat visit April 11, 2009. —MICHAEL C. LONG, Los Angeles County Depart- ment of Parks and Recreation, Natural Areas Division, 1750 N. Altadena Dr., Pasadena, CA 91107. mlong@ parks.lacounty.gov. ARCTOSTAPHYLOS RAINBOWENSIS Keeley — and Massihi (ERICACEAE).—San Diego Co., southern end of the Merriam Mountains, W of Interstate 15, N of the City of Escondido, population on east- facing slope, ca. 5 meters of well-established dirt path/road, portions of property in the process of being cleared and graded for residential development, 33,171333°N; 117115611" W., 1247 feet (380 meters), 10 August 2007, C. M. Guilliams & J. Dillane 417 & 418 (SDSU). Previous knowledge. Arctostaphylos rainbowensis is an narrowly-distributed, burl-sprouting shrub previously thought to be restricted to northern San Diego Co. and southwestern Riverside Co. Recently described, the protologue indicates that the species occurs N of the San Luis Rey River in San Diego Co., and S of Pauba Valley in adjacent Riverside Co. (Keeley and Massihi. 1994. Madrono 41:1—12). Due to its limited distribution and the likely extirpation of many populations during the agricultural and residential development of the region, Keeley and Massihi (/oc. cit.) suggest that the species requires legal protection. The species is currently listed by the California Native Plant Society as a List 1B.1, and is proposed for coverage under two regional habitat conservation plans, the Riverside County MSHCP and the San Diego County North County MSCP. It is not protected under the California Endangered Species Act, or its federal counterpart. Significance. The recent collection of A. rainbowensis from the Merriam Mountains extends the range of the species approximately 20 kilometers to the south. Although seemingly insignificant, this range extension suggests that the species is likely present in small numbers in the chaparral-covered hills between the San Luis Rey River and the City of Escondido. This region, situated between the towns of Rainbow and De Luz in the north and the greater San Diego metropolitan area in the south, is the target of much residential and commercial development. Therefore, it is imperative that biological surveys in this region target A. rain- bowensis as a potentially occurring, sensitive taxon, and that rare plant surveys of chaparral in this region are performed at a level that would detect rare plants present in low-numbers. Chaparral is notoriously difficult to traverse, and as such, plant surveys targeting species that might potentially occur in chaparral vegetation are often “‘windshield” surveys along roads or binocular surveys from ostensibly suitable vantage points. Indeed, this population of A. rainbowensis from the Merriam Mountains would likely remain undetect- ed if it were not located near a dirt road. This population was originally observed by Mr. Craig Reiser, a noted botanist of San Diego, and brought to my attention by Mr. James Dillane, an avid botanist and schoolteacher. Although this population of A. rainbowensis had been documented in_ the California Natural Diversity Database, the locality information was imprecise and no voucher collection had been made. This collection provides a voucher and confirms the species identification. —C. MATT GUILLIAMS, Department of Integrative Biology, University of California Berkeley, 3060 Valley Life Sciences Building, Berkeley, CA 94720. matt_g@ berkeley.edu. Volume 56, Number 3, pages 137-212, published 2 March 2010 SUBSCRIPTIONS — MEMBERSHIP Membership in the California Botanical Society is open to individuals ($35 per year; family $40 per year; emeritus $27 per year; students $27 per year for a maximum of 7 years). Late fees may be assessed. Members of the Society receive MADRONO free. Institutional subscriptions to MADRONO are available ($70). Membership is based on a calendar year only. Life memberships are $750. Applications for membership (including dues), orders for subscriptions, and renewal payments should be sent to CBS c/o Jepson Herbarium, University of California, Berkeley, CA 94720-2465. Requests and rates for back issues, changes of address, and undelivered copies of Ma- DRONO should be sent to the Corresponding Secretary. INFORMATION FOR CONTRIBUTORS Manuscripts submitted for publication in MADRONO should be sent to the editor preferably as Microsoft Word (.doc), Rich Text Format (.rtf), or Portable Document Format (.pdf) files. 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. For all articles and short items (NOTES, NOTEWORTHY COLLECTIONS, POINTS OF VIEW, etc.), fol- low 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, five 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 review copies of illustrations. The title page should have a running header that includes the name(s) of the author(s), and a shortened title. Avoid foot- notes except to indicate address changes. Abbreviations should be used sparingly and only standard abbreviations 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 are encouraged to include the names, addresses, and e-mail addresses of two to four potential reviewers with their submitted manuscript. Authors of accepted papers are required to submit an electronic version of the manuscript. Microsoft Word 2000 or later or WordPerfect 9.0 (or later) 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. 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 Index Herbariorum. 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 MADRONO. Joint au- thors 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 manuscript have been used in other papers that are published, in press, submitted, or soon to be submitted elsewhere. SMITHSONIAN INSTITUTION LIBRARIES we 01534 6414 VOLUME 56, NUMBER 4 OCTOBER-—DECEMBER 2009 MADRONO A WEST AMERICAN JOURNAL OF BOTANY Qk | miss Od li \i\ i 4 i gh CONTENTS PLANT COMMUNITY WATER USE AND INVASIBILITY OF SEMI-ARID SHRUBLANDS BY Woopy SPECIES IN SOUTHERN CALIFORNIA Anna L. Jacobsen, R. Brandon Pratt, L. Maynard Moe, and OFLU TE VOCTS eee re earache ores art one Ee eee Rise a ts Poe MNO 213 IN VITRO PROPAGATION, CRYOPRESERVATION, AND GENETIC ANALYSIS OF THE ENDANGERED HEDEOMA TODSENII (LAMIACEAE) Valerie C. Pence, G. Douglas Winget, Kristine M. Lindsey, Bernadette L. Plair, and Susan M. Chaar tsi.) cs sco... 00. cock focbch oaeccnvccneccoeeee DDN SOIL AND COMMUNITY CHARACTERISTICS ASSOCIATED WITH HAZARDIA ORCUTTII (ASTERACEAE) George L. Vourlitis, Julie Miller, and Kari Colei.....cccccccccccccscsesssesssssseeees pp Ae) ECOLOGY AND GROWTH OF WHITELEAF MANZANITA WITHIN A PONDEROSA PINE PLANTATION IN SOUTHWEST OREGON T. J. Hanson.and Michael Newt0n!. vce ee. ee ee 238 VEGETATION AND FLORA OF A BIODIVERSITY HOTSPOT: PINE HILL, EL DoRADO COUNTY, CALIFORNIA, USA James L. Wilson, Debra R. Ayres, Scott Steinmaus, and Michael Baad... 246 THE TYPE OF CAREX SCIRPOIDEA VAR. GIGAS (CYPERACEAE) | | Kathryn Mauzii2 afoot HUE AN concen MEHR SINE oa cccccensensoene 279 | THE STATUS OF JUNCUS MARGINATUS (JUNCACEAE) IN CALIFORNIA | Peter Fo Lika... Prcsosscossssnnse MUM isi Noe brevsesneoree NEUNES Uoooneorenscnssorsnsces 283 JEW SPECIES A NEW SIDALCEA (MALVACEAE) FROM NORTHEASTERN CALIFORNIA Glenn L. Clifton, Roy E. Buck, and Steven R. Hill ..ccccccccccccccccccceceeseseeees 285 PTEWORTHY (TETHORING re ee eae ecole andere eee eect seizes event ee eee 293 JLLECTIONS JiS(V 2 a5 1h01 1 <1 61) sete ere RO or OR eure een se 295 PRESIDENT S REPORT FOR VOLUME 50 sccccpovac6s.dasssecamscosees09sscsesssnenoeseqnesessnaceeame 296 IS DIFORS: (REPORTIROR VOLUME 50 ce ccececasnsc es oo0eesncseosee ossvcstesaneeossvisessaseeseageanes 297 FE VIEWERS OR VIA NIUS@RIPUS orc. scores eases sone cre. s-5sess sures ueeciesceaseaesveeteesesseoccae 298 HIS NOVO ISU INGE 5 Ol oss acta teen shes eee sae aa aca cess wou egaestns ceteencneeeaeecads anteazen: 299 MP OG AUN = ecto sneer sar tasen ce Sleeves aies sateen So ed aa ters Taal amen saonteaarocau ange a sawaeetesteenes 302 (EABEE OF (CONTENTS FOR VOLUME? 50) 21ig sats oaceiness eee aces ste ace ceca caceentees eseeuseec ll PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL SOCIETY MADRONO (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 MADRONO, Kim Kersh, Membership Chair, Uni- versity and Jepson Herbarium, University of California, Berkeley, CA 94720-2465. kersh @berkeley.edu. Corresponding Editor—TIMOTHY LOWREY Copy Editor—RICHARD WHITKUS Museum of Southwestern Biology Department of Biology MSC03 2020 Sonoma State University University of New Mexico 1801 E. Cotati Avenue Albuquerque, NM 87131-0001 Rohnert Park, CA 94928-3609 madrono @unm.edu whitkus @ sonoma.edu Book Editor—JON E. KEELEY Noteworthy Collections Editors—DIETER WILKEN, MARGRIET WETHERWAX Board of Editors Class of: 2009—DoNOVAN BAILEY, New Mexico State University, Las Cruces, NM MARK BorcHeRT, USFS, Ojai, CA 2010—FReED Hrusa, California Department of Food and Agriculture, Sacramento, CA RICHARD OLMSTEAD, University of Washington, Seattle, WA 2011—JAMIE KNEITEL, California State University, Sacramento, CA KEVIN Rice, University of California, Davis, CA 2012—GRETCHEN LEBUMN, San Francisco State University, CA ROBERT PATTERSON, San Francisco State University, CA CALIFORNIA BOTANICAL SOCIETY, INC. OFFICERS FOR 2009-2010 President: Dean Kelch, Jepson Herbarium, University of California, Berkeley, CA 94720, dkelch @berkeley.edu First Vice President: Andrew Doran, University and Jepson Herbaria, University of California, Berkeley, CA 94720, andrewdoran @berkeley.edu Recording Secretary: Nishanta Rajakaruna, Department of Biological Sciences, San Jose State University, One Washington Square, San Jose, CA 95192, Nishanta.Rajakaruna@sjsu.edu Corresponding Secretary: Heather Driscoll, University Herbarium, University of California, Berkeley, CA 94720- 2465, heather.driscoll @ nature.berkeley.edu Treasurer: Thomas Schweich, California Botanical Society, Jepson Herbarium, University of California, Berkley, CA 94720, tomas @schweich.com The Council of the California Botanical Society comprises the officers listed above plus the immediate Past President, Michael Vasey, Department of Biology, San Francisco State University, 1600 Holloway Avenue, San Francisco, CA 94132, mvasey @sfsu.edu; the Editor of Madrono; three elected Council Members: James Shevock, Department of Botany, California Academy of Sciences, 55 Music Concourse Drive, San Francisco, CA 94118, jshevock@ nature.berkeley.edu; Roy Buck, Jepson Herbarium, University of California, Berkeley, CA 94720, roybuck@ email.msn.com; Chelsea Specht, Department of Plant and Microbial Biology, University of California, Berkeley, CA 94720-2465, cdspecht@nature.berkeley.edu. Graduate Student Representatives: Ben Carter, Department of Integrative Biology and University Herbarium, University of California, Berkeley, CA 94720-2465, bcarter@ berkeley.edu. Webmaster: Susan Bainbridge, Jepson Herbarium, University of California, Berkeley, CA 94720- 2465, sjbainbridge @ berkeley.edu. This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). MADRONO, Vol. 56, No. 4, pp. 213-220, 2009 JUL 75 2010 LIBRARIES PLANT COMMUNITY WATER USE AND INVASIBILITY OF SEMI-ARID SHRUBLANDS BY WOODY SPECIES IN SOUTHERN CALIFORNIA ANNA L. JACOBSEN, R. BRANDON PRATT, AND L. MAYNARD MOE Department of Biology, California State University, Bakersfield, CA 93311 USA ajacobsen@csub.edu FRANK W. EWERS Biological Sciences Department, California State Polytechnic University, Pomona, CA 91768 USA ABSTRACT Soil moisture is a key limiting resource in arid and semi-arid environments for woody shrub species. We assessed if three arid communities differed in their level of dry season soil moisture content and if the dominant species in these communities differed in their ability to use soil moisture. Water potential of all of the dominant woody plant species occurring in chaparral, coastal sage, and Mojave Desert communities and soil moisture content of these sites were measured seasonally. Species occurring in the Mojave Desert exhibited the most negative water potentials while the coastal sage community displayed the least negative water potentials. Dry season volumetric soil moisture content of the Mojave Desert site was lowest (7%), the chaparral site was intermediate (10%), and the coastal sage scrub site had the moistest dry season soil (20%). These moisture differences developed even though the coastal sage and chaparral communities both received the same annual rainfall and had similar soil characteristics. Of the three communities, the coastal sage community may be particularly susceptible to invasion by woody shrub species because its soil moisture content would allow for germination and persistence of a wider range of potential invaders. Current differences among sites in numbers of non- native woody species are consistent with predicted differences in susceptibility to non-native species based on community water use and dry seasonal soil moisture. Key Words: Chaparral, coastal sage, invasive species, Mediterranean-type ecosystems, Mojave Desert, R*, soil moisture, water potential. Much of undeveloped southern California is dominated by arid and semi-arid woody shrub- lands, including the shrub communities of the chaparral, coastal sage scrub, and Mojave Desert (Fig. 1). Water is a limiting resource in these communities, particularly for the dominant ever- green woody shrub species, which remain active year-round. Indeed, drought-induced mortality of species in these communities at both the seedling and adult stages has been reported (Frazer and Davis 1988; Williams et al. 1997; Davis et al. 2002; Paddock III 2006), suggesting the importance of water availability in shaping these communities. The use of the same resource by species should lead to the exclusion of species that are compet- itively inferior at acquiring and using this resource. This principle was described by Gause and is given the name “‘competitive exclusion” (Hardin 1960). While species persistence depends on many variables, Leibig’s law of the minimum _ States that in any given environment there is one _Tesource that is most limiting to species persis- _ tence. Where species compete for a single limiting resource, species resource use can be used to predict the outcome of competitive interactions (R* rule; Tilman 1982). This pattern of resource competition has also been termed exploitation _ competition (Levine 1976; Vance 1985). The R* rule predicts the outcome of compet- itive interactions based on differential depen- dence of species on a given resource level for positive population growth (Fig. 2A; Tilman 1982). The growth rate of a population will decline as the availability of the limiting resource declines. The level of resource availability at which the population growth rate equals zero is the R*. If two species have different R* then the species with the lower R* will be able to competitively exclude the species with the higher R*. This is because the species with lower R* can draw resource availability down to a level that results in a decline in the population of the species with the higher R* (Fig. 2A). This same model may also be applied to interactions among plant communities (Fig. 2B). Natural communities are shaped by many factors and models that examine only one or a few factors are undoubtedly simplifications; however, such simplified models have been shown to have utility in predicting outcomes of competitive interactions. The R* rule and similar models have been used to predict the outcome of competition in controlled experiments (Titman 1976; Ciros-Pérez et al. 2001; Fox 2002), patterns of species abundance within communities (KOiv and Kangro 2005; Harpole and Tilman 2006), The semi-arid shrub communities from the PIG. 1, winter rainfall region of southern California: chaparral (A), coastal sage scrub (B), and Mojave Desert scrub (C). These photos were taken of the sites that were sampled in the present study in Spring 2006. Photos by A. Jacobsen. and patterns of species succession (Herron et al. 2001). Recently, these models have also been applied to species invasions and determination of community invasibility and native versus non- native competitive interactions (Herron et al. 2001; Booth et al. 2003; Fargoine et al. 2003; Krueger-Mangold et al. 2006; Funk and Vitousek 2007). While many of these studies have focused MADRONO [Vol. 56 L 4) _ S O = o>) e 2 ® = 2. O OW Low a Rp* High > oO Cc o S oO . © Community B Ee 1) = ® i) ad Low Rp* Rp” High Resource availability Fic. 2. Solid curves show the dependence of popula- tion growth rate on resource availability (A; based on Tilman 1982) and the dependence of species communi- ties on resource availability (B) if a single resource is limiting. In panel A, the dashed line indicates the point at which the mortality rate in a population equals the replacement rate. The R* indicates the amount of resource resulting in zero population growth. Species B requires a higher amount of resource than Species A for the population to increase (i.e., Rg* > Ra*); therefore, the long term outcome of competition between Species A and Species B would be the competitive exclusion of Species B. This results because Species A is able to reduce resources below the level at which Species B can maintain positive population growth. In panel B, the dashed line indicates the point at which individuals are | so infrequent in the community that the community cannot be sustained. In this example, Community A would be able to competitively exclude Community B. on nutrient availability (for plants) or food - availability (for animals), moisture availability and plant water use have been shown to predict | competitive outcomes in some plant communi- ties, particularly in arid regions (Cleverly et al. 1997; Booth et al. 2003). j YU \ We examined differences in plant water poten- | tial and soil moisture among three semi-arid plant communities of southern California to assess whether these communities differed in their ability to utilize soil water resources. 2009] TABLE 1. PLANT COMMUNITIES. Vegetation type and location Chaparral—Cold Creek Canyon Preserve, Santa Monica Mountains, CA, USA (34.5N 118.4W) Coastal Sage Scrub—Pepperdine University, Malibu, CA, USA (34.2N 118.4W) JACOBSEN ET AL.: COMMUNITY WATER USE AND INVASIVE SPECIES THE DOMINANT Woopby SPECIES PRESENT IN AND SAMPLED IN EACH OF THREE SEMI-ARID Mojave Desert Scrub—Red Rock Canyon State Park, CA, USA (35.2N 117.6W) Species Family Adenostoma fasciculatum Hook. & Arn. Rosaceae Adenostoma sparsifolium Torrey Rosaceae Arctostaphylos glandulosa Eastw. Ericaceae Ceanothus cuneatus (Hook.) Nutt. Rhamnaceae Ceanothus megacarpus Nutt. Rhamnaceae Ceanothus oliganthus Nutt. Rhamnaceae Ceanothus spinosus Nutt. Rhamnaceae Malosma laurina (Nutt.) Abrams Anacardiaceae Quercus berberidifolia Liebm. Fagaceae Rhus ovata S. Watson Anacardiaceae Artemisia californica Less. Asteraceae Encelia californica Nutt. Asteraceae Eriogonum cinereum Benth. Polygonaceae Hazardia squarrosa (Hook. & Arn.) Greene Asteraceae Lotus scoparius (Nutt.) Ottley Fabaceae Malacothamnus fasciculatus (Torrey & A. Gray) Greene Malvaceae Malosma laurina (Nutt.) Abrams Anacardiaceae Salvia leucophylla Greene Lamiaceae Salvia mellifera Greene Lamiaceae Ambrosia dumosa (A. Gray) Payne Asteraceae Atriplex canescens (Pursh) Nutt. Chenopodiaceae Atriplex polycarpa (Torrey) S. Watson Chenopodiaceae Coleogyne ramosissima Torrey Rosaceae Hymenoclea salsola A. Gray Asteraceae Isomeris arborea Nutt. Capparaceae Larrea tridentata (DC.) Coville Zygophyllaceae Lepidospartum squamatum (A. Gray) A. Gray Asteraceae Lycium andersonii A. Gray Solanaceae Predawn water potential of woody plant branch- lets or leaves estimates availability of soil moisture. More negative values indicate less available soil moisture reserves. Plants that can tolerate more negative water potentials can continue to extract soil water when plants that are less tolerant are no longer able to obtain enough soil moisture. Thus, species and commu- nities that tolerate more negative predawn and midday water potentials and in which water is the most limiting resource will have a lower R*. Differences in soil moisture and species water potential may also be related to the invasibility of these communities by non-native species. Arid and semi-arid communities with a lower R* would be predicted to be vulnerable to invasion by fewer species (1.e., only the limited number of potential invaders that are most water stress tolerant) (Pratt and Black 2006). Results are discussed as they apply to the invasion potential of communities. However, it should be noted that such results apply only to woody shrub species because predictions of competitive outcomes based on limiting resources appear most useful among similar functional groups (Krueger-Mangold et al. 2006). MATERIALS AND METHODS Three diverse semi-arid sites were selected based on their abundance of woody shrub species. These sites represented chaparral, coastal sage, and Mojave Desert plant communities occurring in the winter rain-fall area of southern California (Fig. 2; see Jacobsen et al. 2007, 2008 for site descriptions). The chaparral site was dominated by 10 woody plant species, the coastal sage site was dominated by 9 woody plant species, and the Mojave Desert scrub community was dominated by 10 woody plant species (Table 1). These species accounted for 89, 99, and 73% of the woody plants in each community, respectively. The average annual precipitation for these sites is approximately 440 mm for both the chaparral and coastal sage communities and 180 mm for the Mojave Desert site (average annual precipitation from 1999-2007 based on the July to June rain year). At each site, woody shrub species were sampled using a modified point-quarter sampling method to determine basic stand structural characteristics (Cox 1985). Fifty-four to 60 points were sampled at each site and plant height, basal diameter, crown diameter, and interplant basal and crown distance were all determined. Differ- ences in traits among communities were tested using ANOVAs (Statview v. 5.0.1, SAS Institute Inc., Cary, NC, USA). Percentage volumetric soil moisture content (m°*/m*) was measured monthly from late August 2006 through April 2007 at each of the three sites. 216 MADRONO [Vol. 56 TABLE 2. MEANS AND STANDARD ERRORS OF CANOPY HEIGHT, CROWN DIAMETER, BASAL DIAMETER, AND DISTANCE BETWEEN INDIVIDUAL PLANT BASES AND CROWNS IN THREE PLANT COMMUNITIES: CHAPARRAL, COASTAL SAGE, AND MOJAVE DESERT SCRUB. Different letters in each column indicate significant differences among communities. ' From outside edge of base to outside edge of base. > From edge of crown to edge of crown. Distance between Distance between Crown diameter Basal diameter Vegetation type Height (m) (m) (m) bases! (m) crowns? (m) Chaparral 2.57 2 OLOT-a 1.87 + 0.08 a O17 = O00l a 0.642 0.054 —LIis210:06-4 Coastal Sage Scrub 0.83 += 0:03 b. 0.99 = 0.05 b 0.10 + 0.01 b 0.61 + 0.03 a —0.47 + 0.04 b Mojave Desert Scrub 0.75 = 0:03 ¢ 1.03 + 0.04 b 18 + 0.00 a 1.67 + 0.08 b 0.70 + 0.08 c Volumetric soil moisture content of the upper soil RESULTS layers (upper 30 cm) was measured at the base of the same individuals at each sampling time (n = 14-26) using a TDR probe (CS615, Campbell Scientific, Inc., Logan, UT, USA) attached to a datalogger (CR23X Micrologger, Campbell Sci- entific, Inc., Logan, UT, USA). Within each site, values for sample locations were compared using repeated measures ANOVAs. Minimum seasonal soil moisture of each site was determined as a community wide mean from the month with the lowest soil moisture values. Minimum volumetric soil moisture across communities was compared using ANOVA followed by a Bonferroni/Dunn post-hoc analysis (Statview v. 5.0.1, SAS Institute Inc., Cary, NC, USA). Monthly water potential was measured at midday on all dominant woody shrub species at each site from February 2006 and continuing through April 2007 (see Jacobsen et al. 2008 for Methods) using a pressure chamber (Model 2000 Pressure Chamber Instrument, PMS Instruments, Corvalis, OR, USA). Predawn water potentials were sampled during the dry season in August and September 2006. Water potential was mea- sured on leaves or small branchlets (for species with very small leaves) from six individuals per species at each sampling period. Mean water potentials were calculated for each species at each sampling time. These values were pooled across species and the frequency of water potentials were calculated for each community in order to determine the community wide range in soil moisture resource use. The invasibility of communities was compared to current non-native and invasive species pres- ence in communities by determining the number of woody species occurring in each community, including the percentage of these woody species that are native to the region, not native, and which are both non-native and also invasive. Species were included as woody species if they were reported by Hickman (1996) to be trees, shrubs, subshrubs, or woody perennials. Species presence in communities was determined based on species geographic ranges and ecology report- ed in Hickman (1996), The Jepson Manual Online (University of California and Jepson Herbaria 2001) and regional floras (McAuley 1996; Dale 2000; Faull 2005). Plants within these three communities differed in height, crown diameter, and basal diameter (Table 2). Overall, chaparral species were taller and had greater crown diameters relative to the other two communities (Table 2). The distance between plants was not different among the chaparral and coastal sage communities although crown overlap was greater in the chaparral due to their significantly wider crown diameters. Indi- viduals occurring at the Mojave Desert site were further apart than in the other two communities and this, combined with their smaller canopy sizes, resulted in an open canopy (Table 2). Dry season volumetric soil moisture of the shallow soil layers significantly differed among the sites and with season (Figs. 3 and 4C). These sites do not differ in soil texture at the measured soil depth (Jacobsen et al. 2007) so these differences should equate to differences in soil water potential of the shallow soil layers of these sites; however, we did not measure soil water. potential versus volumetric soil moisture for these — sites. Dry season soil moisture values for the fall of: 2006 followed an average rain year and thus. should be representative for soil moisture condi- | tions in average years during the dry season. The three communities significantly differed in their | dry season volumetric soil moisture content. (Fig. 4C, P < 0.001). Minimum volumetric soil moisture content of the Mojave Desert site was - the driest (7.1 + 0.1%, P = 0.001 compared to the | chaparral and coastal sage scrub), the chaparral | site was intermediate (10.3 + 0.6%, P < 0.001 | compared to the coastal sage scrub), and the. coastal sage scrub site had the most moist dry | All of the sites. season soil (20.3 + 1.0%). exhibited significantly higher soil moisture values — during the wet season from December to March | (Fig. 2); however, rainfall during the winter. 2006-2007 wet season was much less than average (approximately 100-130 mm across all sites), sO wet season soil moisture values may. deviate from the typical pattern in an average , rainfall year. Midday water potentials were strongly corre-. lated with predawn water potentials in August and September (P < 0.001, r° = 0.771; Jacobsen 2009] JACOBSEN ET AL.: COMMUNITY WATER USE AND INVASIVE SPECIES oa Chaparral Coastal Sage Scrub Mojave Desert Scrub G 05 6 G 04 : o 5 0.3 , ab ab, O02 S aaaa a (e) Y 0.0 © © © aN © © © A © © © A v © S v Vv © S Vv v © S Vv Date (m/d/yy) Fic. 3. Volumetric soil moisture content of the upper soil layers (upper 30 cm) measured from late August 2006 through April 2007 at each of the three arid shrub communities (chaparral, coastal sage, Mojave Desert). Within each panel, different letters indicate soil moisture values that are significantly different. et al. 2007; Pratt et al. 2008). Midday water potentials were approximately one MPa lower than predawn water potentials measured on the same plant (range of 0.88 to 1.17 MPa difference between predawn and midday). The three communities examined in the present study differed in the minimum midday water potentials that could be tolerated by species within each community (Fig. 4A and B). Species within the Mojave Desert community reached the lowest water potentials (—9.2 MPa), spe- cies within the chaparral were intermediate (—8.7 MPa), and woody species from the coastal sage scrub community experienced the least negative water potentials (—5.4 MPa). All of the communities had some species that main- tained relatively high water potentials even in the dry season (Fig. 4A and B). These minimum water potential values are consistent with site differences in dry seasonal soil moisture content (Fig. 4C). At present, the coastal sage contains more non- native and invasive woody species than the chaparral and Mojave Desert communities (Ta- ble 3). In the coastal sage, 39% of the woody species are non-native, compared to only 21.1% for the chaparral, and 7.1% for Desert Scrub. Of woody species that are categorized as invasive or noxious, the coastal sage contains 4.4% invasive/ noxious species compared to only 2.5% for the chaparral and 0% for the Mojave Desert scrub. DISCUSSION Species of woody shrubs in the chaparral, coastal sage, and Mojave Desert communities are able to tolerate relatively low water potentials (less than —7 MPa in all communities compared to a limit of —2 MPa for most crop plants; Tyree and Zimmerman 2002). Among these three communities, the Mojave Desert community contains the species that experience the lowest water potentials seasonally and therefore are able =) Chaparral MB Coastal Sage {_] Mojave Desert o —_— 0.001 Relative Frequency (=) ie —— Chaparral —— Coastal Sage —--- Mojave Desert 0.1 Relative Frequency oO = 0.001 0.0001 -14 -12 -10 -8 -6 -4 -2 0 Water Potential (MPa) Chaparral b C Coastal Sage C Mojave Desert 1a 0 5 10 15 20 25 30 Volumetric Soil Moisture Content (%) FIG. 4. The frequency of water potential occurrence in three arid plant communities, the chaparral (medium gray fill, solid line), coastal sage scrub (black fill and dashed line) and Mojave Desert scrub (light grey fill, dashed-dotted line) (A and B) and the minimum seasonal volumetric soil water content for these same three communities (C). Different letters after bars in panel C indicated soil moisture values that are significantly different. 218 MADRONO [Vol. 56 TABLE 3. THE NUMBER OF WOODY SPECIES OCCURRING IN THREE PLANT COMMUNITIES OF SOUTHERN CALIFORNIA, INCLUDING THE PERCENTAGE THAT ARE NATIVE TO THE REGION, NOT NATIVE, AND ARE BOTH NON-NATIVE AND INVASIVE. | Includes only chaparral and coastal sage species that are distributed in the western Transverse Ranges of southern California. This includes species that have distributions that are more extensive than this region but which include it, such as species that have ranges of the south west of California or the California Floristic Province. Community Number of woody species Chaparral 203 Coastal Sage Scrub 182 Mojave Desert Scrub 154 to utilize more limited shallow soil moisture reserves than species in the other two communi- ties. This suggests that among these communities, the Mojave Desert community has the lowest R* value compared to the other communities, assuming that soil moisture availability is the most limiting resource. Differences in soil mois- ture availability between the Mojave Desert community and the chaparral and coastal sage communities are likely related to differences in precipitation as well as differential water use by the species present in these communities (Jacob- sen et al. 2008). Among sites that receive the same amount of precipitation, differences in woody plant water use significantly affect soil moisture availability. In the present study, the chaparral and coastal sage sites experienced similar annual rainfall (approximately 440 mm) and had similar soil texture (Jacobsen et al. 2007), yet they have different dry season soil moisture levels. This 1s presumably because of differences in water use by species in these communities (Jacobsen et al. 2008). Species in the chaparral community are able to attain and survive lower water potentials than species within the coastal sage community allowing them to draw shallow soil moisture down to lower levels. Such biotically mediated differences in water resource availability are predicted to occur (Huxman et al. 2005) and have been reported in other communities (re- viewed in Levine et al. 2003). Among the chaparral and coastal sage communities, this difference in water use may be related to the drought evading strategy of coastal sage species, many of which have seasonally dimorphic leaves or reduce their canopies during the dry season, compared to the evergreen chaparral species which maintain a full canopy and utilize a drought resisting strategy (Mooney 1989). Community water use of deeper soil layers may be very different than the patterns found in the present study, which examined only shallow soil layers (<30 cm soil depth). Indeed, different species within these communities tap different soil layers (Hellmers et al. 1955) and_ shift between shallow and deeper soil layers seasonal- ly. Additionally, plants in different communities display differential water use patterns (Jacobsen Native (%) Non-native (%) = Invasive/noxious (%) 78.8 22 225 61.0 39.0 4.4 92.9 | 0.0 et al. 2008). However, differences in the soil moisture of shallow soil layers may be particu- larly important in the establishment of woody plant seedlings and in determination of the invasibility of these communities. Dry season soil moisture has been found to be linked to seedling survival for five woody species from the Mediterranean Basin (Spain) (Padilla and Pugnaire 2007). Since most of the invasive or naturalized species already present in these California communities are native to other Mediterranean-type climate regions (Knops 1995; Salo 2004; Keeley 2006), the results of this recent study may be particularly applicable to these southern Californian communities. At dry season soil moistures over 18% nearly all of the — studied Mediterranean seedlings survived, where- as there was very little survival at <12% soil : moisture, and no survival predicted for soil that ' had <8% soil moisture (Padilla and Pugnaire » 2007). The amount of available water for a given > soil moisture content varies depending on soil | texture, which differed between the study by. Padilla and Pugnaire (2007) and the sites in the | present study. Therefore, at the same percentage | soil moisture there would be slightly more water | available to plants in the southern California» communities (with sandy and loamy sand soils; | 2007) compared to the site in) Spain where plants were grown in silt soil. Yet, | the results suggest that while the chaparral and: Jacobsen et al. Mojave Desert communities (with 10% and 7% dry season soil moisture content, respectively) | would be able to resist invasion (in an average : year), nearly all seedlings of these Mediterranean | species would be able to persist in the coastal sage + community through the dry season (20% dry season soil moisture content). While dry season data in the present stud : were collected during an average rainfall year, it |] should be noted that the invasion process 1s likely dependent on stochastic abiotic events including disturbances such as land use changes or fire, and’ altered weather patterns that result in greater than average levels of precipitation (Davis et al. 2000; Mazia et al. 2001; Lloret et al. 2005). These ; ; ; factors also suggest that, at present, the coastal _ sage community may be the most susceptible to invasion relative to the chaparral and Mojave. 2009] Desert communities. When sites experiencing similar rainfall are examined, coastal sage shrub species are less able to lower the level of available soil water resources relative to chaparral species. Additionally, the coastal sage community is at least partially disturbance dependent and likely experiences more disturbance than the other two communities at present. The predicted differences in susceptibility to non-native and invasive species among these communities (predicted by water resource limita- tion of these communities and R*) are consistent with current numbers of non-native woody species in these communities (Table 3). The coastal sage contains more non-native and invasive species than the chaparral and Mojave Desert communities. In both the coastal sage and chaparral communities woody invaders are pre- dominantly native to other Mediterranean-type climate regions and include, Acacia spp., Nicoti- ana glauca, Cytisus spp. and Spartium spp. (Knops et al. 1995; Keeley 2006). Arid and semi-arid sites that exhibit low productivity appear to be less broadly susceptible to invasion than more moist and productive sites (Stohlgren et al. 2002; Otto et al. 2006); however, semi-arid and arid plant communities in southern California have still experienced significant inva- sion by non-native species. Most of these invasive species are annual herbs and grasses (Knops et al. 1995; Sax 2002; Salo 2004; Keeley 2006; Lambri- nos 2006). While invasion of these communities by annuals appears to be at least partially dependent on increased nitrogen availability (Brooks 2003; however see Padgett and Allen 1999), the timing and availability of soil moisture is also important (Salo 2004). Indeed, the shallow moisture available in the coastal sage site examined in the present study would be readily utilized by shallow rooted annuals and herbaceous species and we observed a greater density of non-native annuals and herba- ceous species at this site compared to the chaparral and Mojave Desert sites. The R* rule and similar models have been useful for predicting invasibility of some commu- nities. Using this model to predict susceptibility to invasion by woody shrub species in these three woody shrub communities of southern California suggests that the coastal sage community may be most susceptible to woody species invasion if soil moisture is a limiting resource. This suggests that ‘monitoring and control of woody naturalized shrubs may be most efficient if efforts are focused on this southern Californian plant community, as ‘the chaparral and Mojave Desert communities “appear to be more impervious to shrub invasion. ! ACKNOWLEDGMENTS | This work was supported by an NSF Graduate Research Fellowship and Doctoral Dissertation Im- provement and by a Michigan State University | JACOBSEN ET AL.: COMMUNITY WATER USE AND INVASIVE SPECIES Z19 Department of Plant Biology Grant supporting A. L. J. We thank L. Alan Prather, Douglas W. Schemske, Frank W. Telewski, and Stephen D. Davis for valuable advice and discussions. A. L. J. and R. B. P. thank the Andrew Mellon Foundation and NSF Grant IOS- 0845125 to R. B. P. for support. LITERATURE CITED BooTH, M. S., M. M. CALDWELL, AND J. M. STARK. 2003. Overlapping resource use in three Great Basin species: implications for community invisi- bility and vegetation dynamics. Journal of Ecology 91:36-48. BROOKS, M. L. 2003. Effects of increased soil nitrogen on the dominance of alien annual plants in the Mojave Desert. Journal of Applied Ecology 40:344-353. CIROS-PEREZ, J.. M. J. CARMONA, AND M. SERRA. 2001. Resource competition between sympatric sibling rotifer species. Limnology and Oceanogra- phy 46:1511—1523. CLEVERLY, J. R., S. D. SMITH, A. SALA, AND D. A. DEvITT. 1997. Invasive capacity of Tamarix ramosissima in a Mojave Desert floodplain: the role of drought. Oecologia 111:12-18. Cox, G. W. 1985. Laboratory manual of general ecology, Sth ed. WCB/McGraw-Hill, New York, NY. DALE, N. 2000. Flowering plants: the Santa Monica Mountains, coastal and chaparral regions of southern California. California Native Plant Soci- ety, Sacramento, CA. DAvIs, M. A., J. P. GRIME, AND K. THOMPSON. 2000. Fluctuating resources in plant communities: a general theory of invasibility. Journal of Ecology 88:528—534. DAvis, S. D., F. W. Ewers, J. S. SPERRY, K. A. PORTWOOD, M. C. CROCKER, AND G. C. ADAMS. 2002. Shoot dieback during prolonged drought in Ceanothus chaparral of California: a possible case of hydraulic failure. American Journal of Botany 89:820-838. FARGOINE, J., C. S. BROWN, AND D. TILMAN. 2003. Community assembly and invasion: an experimen- tal test of neutral versus niche processes. Proceed- ings of the National Academy of Sciences USA 100:8916—-8920. FAULL, M. 2005. Toward a flora of Red Rock Canyon State Park: a preliminary checklist. California Native Plant Society, Kern County, CA. Website http://www.kerncnps.org/PlantLists/redrockcyn/ rrcflora.pdf [accessed 27 January 2010]. Fox, J. W. 2002. Testing a simple rule for dominance in resource competition. The American Naturalist 159:305-319. FRAZER, J. M. AND S. D. DAvis. 1988. Differential survival of chaparral seedlings during the first summer drought after wildfire. Ocecologia 76:215—221. FUNK, J. L. AND P. M. VITOUSEK. 2007. Resource-use efficiency and plant invasion in low-resource systems. Nature 446:1079—1081. HARDIN, G. 1960. The competitive exclusion principle. science 131:1292—1297. HARPOLE, W. S. AND D. TILMAN. 2006. Non-neutral patterns of species abundance in grassland com- munities. Ecology Letters 9:15—23. 220 MADRONO HELLMERS, H., J. S. HORTON, G. JUHREN, AND J. O’KEEFE. 1955. Root systems of some chaparral plants in southern California. Ecology 36:667—678. HERRON, G. J., R. L. SHELEY, B. D. MAXWELL, AND J. S. JACOBSON. 2001. Influence of nutrient availability on the interaction between spotted knapweed and bluebunch wheatgrass. Restoration Ecology 9:326-331. HICKMAN, J. C. (ed.). 1993. The Jepson manual. University of California Press, Berkeley, CA. HUXMAN, T. E., B. P. WILCOx, D. D. BRESHEARS, R. L. Scott, K. A. SNYDER, E. E. SMALL, K. HULTINE, W. T. POCKMAN, AND R. B. JACKSON. 2005. Ecohydrological implications of woody plant encroachment. Ecology 86:308—319. JACOBSEN, A. L., R. B. PRATT, S. D. DAVIS, AND F. W. EWERS. 2007. Cavitation resistance and seasonal hydraulics differ among three arid Californian plant communities. Plant, Cell and Environment 30:1599-1609. : : , AND . 2008. Compar- ative community physiology: non-convergence in water relations among three semi-arid shrub communities. New Phytologist 180:100—113. KEELEY, J. E. 2006. Fire management impacts on invasive plants in the western United States. Conservation Biology 20:375—384. KNops, J. M. H., J. R. GRIFFIN, AND A. C. ROYALTY. 1995. Introduced and native plants of the Hastings reservation, central coastal California: a compari- son. Biological Conservation 71:115—123. Korv, T. AND K. KANGRO. 2005. Resource ratios and phytoplankton species composition in a strongly stratified lake. Hydrobiologia 547:123—135. KRUEGER-MANGOLD, J., R. SHELEY, AND R. ENGEL. 2006. Can R*s_ predict invasion in semi-arid grasslands? Biological Invasions 8:1343—1354. LAMBRINOS, J. G. 2006. Spatially variable propagule pressure and herbivory influence invasion of chaparral shrubland by an exotic grass. Oecologia 147:327-334. LEVINE, L. M., M. VILA, C. M. D’ANTONIO, J. S. DUKES, K. GRIGULIS, AND S. LAVOREL. 2003. Mechanisms underlying the impacts of exotic plant invasions. Proceedings of the Royal Society of London B 270:775—781. LEVINE, S. H. 1976. Competitive interactions in ecosystems. The American Naturalist 110:903—910. LLORET, F., F. MEDAIL, G. BRUNDU, I. CAMARDA, E. MORAGUES, J. RITA, P. LAMBDON, AND P. E. HULME. 2005. Species attributes and invasion success by alien plants on Mediterranean islands Journal of Ecology 93:512—520. MAZIA, C. N., E. J. CHANETON, C. M. GHERSA, AND R. J. C. LEON. 2001. Limits to tree species invasion in pampean grassland and forest plant communi- ties. Oecologia 128:594—602. McAULEY, M. 1996. Wildflowers of the Santa Monica Mountains. Canyon Publishing Company, Canoga Park, CA. Mooney, H. A. 1989. Chaparral physiological ecolo- gy—paradigms revisited. Pp. 85—90 in S. C. Keeley (ed.), The California chaparral: paradigms reexam- [Vol. 56 ined. The Natural History Museum of Los Angeles County, Science Series No. 34. OTTo, R., B. O. KrusI, C. A. BURGA, AND J. M. FERNANDEZ-PALACIOS. 2006. Old-field succession along a precipitation gradient in the semi-arid coastal region of Tenerife. Journal of Arid Environments 65:156—178. PADDOCK, III, W. A. 2006. Adult mortality of chaparral shrubs following severe drought. M.Sc. Thesis, Michigan State University, East Lansing, MI. PADGETT, P. E. AND E. B. ALLEN. 1999. Differential responses to nitrogen fertilization in native shrubs and exotic annuals common to Mediterranean coastal sage scrub of California. Plant Ecology 144:93-101. PADILLA, F. M. AND F. I. PUGNAIRE. 2007. Rooting depth and soil moisture control of Mediterranean woody seedling survival during drought. Function- al Ecology 21:489—495. PRATT, R. B. AND R. A. BLACK. 2006. Do invasive trees have a hydraulic advantage over native trees? Biological Invasions 8:1331—1341. , A. L. JACOBSEN, R. MOHLA, F. W. EWERS, AND S. D. DAvIs. 2008. Linkage between water stress tolerance and life history in seedlings of nine chaparral species (Rhamnaceae). Journal of Ecol- ogy 96:1252—1265. SALO, L. F. 2004. Population dynamics of red brome (Bromus madritensis subsp. rubens): times for concern, opportunities for management. Journal of Arid Environments 57:291—296. SAX, D. F. 2002. Native and naturalized plant diversity are positively correlated in scrub communities of California and Chile. Diversity and Distributions 8:193-210. STOHLGREN, T. J.,G. W. CHONG, L. D. SCHELL, K. A. RIMAR, Y. OTSUKI, M. LEE, M. A. KALKHAN, | AND C. A. VILLA. 2002. Assessing vulnerability to | invasion by nonnative plant species at multiple | spatial scales. Environmental Management 29:566—-577. | TILMAN, D. 1982. Resource competition and commu- | nity structure. Monographs in Population Biology 17. Princeton University Press, Princeton, NJ. TITMAN, D. 1976. Ecological competition between | algae: experimental confirmation of resource-based | competition theory. Science 192:463—465. TYREE, M. T. AND M. H. ZIMMERMANN. 2002. Xylem : Structure and the Ascent of Sap. Springer-Verlag, Berlin, Germany. UNIVERSITY OF CALIFORNIA AND JEPSON HERBARIA. | 2001. Online interchange for California floristics. The Regents of the University of California, | Berkeley, CA. Website http://ucjeps.berkeley.edu/ interchange.html [accessed 01 January 2010]. VANCE, R. R. 1985. The stable coexistence of two) competitors for one resource. The American, Naturalist 126:72—86. WILLIAMS, J. E., S. D. DAVIS, AND K. PORTWOOD. | 1997. Xylem embolism in seedlings and resprouts | of Adenostoma fasciculatum after fire. Australian Journal of Botany 45:291—300. —_—_ eee MADRONO, Vol. 56, No. 4, pp. 221—228, 2009 IN VITRO PROPAGATION, CRYOPRESERVATION, AND GENETIC ANALYSIS OF THE ENDANGERED HEDEOMA TODSENITI (LAMIACEAE) VALERIE C. PENCE, G. DOUGLAS WINGET, KRISTINE M. LINDSEY, BERNADETTE L. PLAIR, AND SUSAN M. CHARLS Center for Conservation and Research of Endangered Wildlife, Cincinnati Zoo & Botanical Garden, 3400 Vine St., Cincinnati, OH 45220 valerie.pence@cincinnatizoo.org ABSTRACT Todsen’s pennyroyal (Hedeoma todsenii R. S. Irving, Lamiaceae) is a federally endangered species from the mountains of south central New Mexico that rarely produces seed. In vitro propagation methods were developed to provide material for cryostorage and for reintroduction, if that becomes necessary. Cultures were initiated from shoot tips taken from the ex situ collection at The Arboretum at Flagstaff, resulting in 12 genetic lines that were maintained on MS medium with 0.1 mg/L BAP and 0.01 mg/L NAA. Tests with other media commonly in use in CREW’s Endangered Plant Propagation Program indicated that MS medium with 0.5 mg/L BAP increased shoot production and MS medium with 0.5 mg/L IBA increased root production over the maintenance medium. Other concentrations of IBA tested did not improve rooting, and a pulse of IBA followed by culture on charcoal-containing medium did not increase rooting significantly above the control. Approximately half of the plants moved to soil survived acclimatization, regardless of previous treatments. Survival through cryopreservation averaged 35% with no significant difference between the encapsulation dehydration and encapsulation vitrification procedures, and shoot tips from all 12 lines have been banked in liquid nitrogen for long-term storage. RAPD analyses indicated that there was less diversity among plants that exist in close proximity in situ than among genotypes that are separated by more distance. These propagation, cryopreservation, and genetic analysis protocols are all methods that can be used as tools to provide support for the long-term conservation of this species. Key Words: Cryopreservation, endangered plant, Hedeoma todsenii, in vitro, RAPD. Hedeoma todsenii R.S. Irving (Todsen’s pen- nyroyal, Lamiaceae) (Irving 1979) is a federally endangered species found in three population clusters in the Sacramento and San Andres Mountains of south central New Mexico. Be- cause of its remote locations, most of the threats to this species are natural, including fire, climate change, limited natural habitat, and limited sexual reproduction. However, changes in land management, illegal grazing, and military activ- ities may also pose threats (U.S. Fish and Wildlife Service 2001). Seeds of this species are rarely available for propagation or preservation. H. todsenii has been shown to be capable of out-crossing, but it appears to have a low percentage of flowers producing seed, a very low number of seeds per flower, and a low germination rate in the laboratory (Huenneke 1993). This study was undertaken to develop in vitro propagation and cryopreservation methods for H. todsenii to supplement traditional ex situ conservation methods and in situ conservation efforts. In vitro methods can provide a variety of tools _ for conservation. These include in vitro collecting (IVC), a method for initiating tissue cultures in the field (Pence et al. 2002). IVC can be used when seeds are not available and when conditions for the transport of cuttings may not be optimal. In vitro propagation methods can then be used to increase the number of plants available for reintroduction, research, education, etc. When seeds are not available for germplasm storage, shoot tips from in vitro cultures can be cryopre- served, and this can be an alternative method for long-term germplasm conservation. The growth of H. todsenii, as large mats, and its low seed production and germination rate have led to speculation that the plants are primarily reproducing asexually in the wild. Comparing DNA sequence information could show whether they do in fact reproduce clonally (Sydes and Peakall 1998). Random amplified polymorphic DNA (RAPD) has been used to reveal DNA sequence differences between species and even those more subtle differences associated with hybrids, ecotypes, or individual members of a population (Krasnyanski et al. 1998; Liu et al. 2006). Analysis of the number of identical DNA bands between individuals can be used to group plants (Adams et al. 2003). Comparison of these groupings to the geographical locations of the plants can determine the apparent genetic diver- sity, gene flow and genetic interaction within and between separated populations (Morden and Loeffler 1999; Skoula et al. 1999; Khanuyja et al. 2000). This information can prove invaluable in cases where an endangered species is being 2919) prepared for reintroduction (Maki and Horie 1999: Mattner et al. 2002). In the case of AZ. todsenii, which appears to rely heavily on vegetative reproduction, it would be of interest to know the extent and distribution of genetic variation within the species. This study describes the application of in vitro methods to the propagation and preservation of H. todsenii. In addition, it describes the results of RAPD studies on the in vitro lines established and the genetic relationships of the populations from which they originated. METHODS Establishment of Cultures Shoots of Hedeoma todsenii were obtained from the ex situ collection at The Arboretum at Flagstaff, and cultures were initiated twice using two different methods. Method I. Multiple shoots were collected from each of four plants and sent overnight to the Center for Conservation and Research of En- dangered Wildlife (CREW). These were then surface sterilized with a 1:20 dilution of commer- cial bleach with 0.05% Tween 20 for 10 min, followed by 2 rinses in sterile, pure water. The shoots were then placed onto medium in 25 X 150 mm borosilicate culture tubes with colorless polypropylene closures, 15 ml of medium/tube. The medium consisted of Murashige and Skoog (1962) salts and minimal organics (Linsmaier and Skoog 1965) (PhytoTechnology Laboratories, Shawnee Mission, KS), 3% sucrose, 0.22% gelrite (Kelco, San Diego, CA) (=MS basal medium), with 0.5 mg/L benzylaminopurine (BAP), 0.05 mg/L naphthaleneacetic acid (NAA), and 100 mg/L active benomyl (methyl 1-(butylcarba- moyl)-2-benzimidazolecarbamate, 95%) (Sigma- Aldrich, St. Louis, MO). This medium was similar, but not identical, to the medium reported as effective in propagating a related species, H. multiflorum (Koroch et al. 1997). Multiple shoots were cultured from each of the four genetic lines received. Contamination was monitored visually, and when it was found in some of the original isolates, the tissues were cut to remove contam- inated areas, re-sterilized and placed on the same medium with the addition of 500 mg/L carben- icillin and 200 mg/L cefotaxime; both were filter sterilized and added to the autoclaved medium (all antibiotics were from Sigma-Aldrich). Even though there was contamination in some of the cultures, some material from each genetic line survived to establish lines HT-1 to HT-4. Method 2. This method used the technique of IVC (in vitro collecting) (Pence 2005). One shoot tip was collected from each of 55 documented genotypes. These were surface sterilized in the MADRONO [Vol. 56 field by wiping with 70% ethanol and were placed into 7 ml borosilicate scintillation vials. Each vial contained 2.5 ml of the same medium used previously, except that antibiotics were added as a drop (0.05 ml) of a freshly made, filter sterilized solution (of 0.25 mg/ml vancomycin and 5 mg/ml cefotaxime) into each vial, after the tissue was added. The vials were then sent overnight to CREW where they were evaluated for browning and contamination. The 36 shoots that remained clean and green after a week were transferred to fresh medium of the same formulation, but without fungicide or antibiotics. Nine (25%) of these shoots initiated growth, and these were subcultured onto GM medium (Table 1) and maintained as lines HT-5 to HT-13. One line was lost in subsequent culture, and lines HT-1 to HT- 12 were used for the growth experiments de- scribed here. The survival of only 9 genotypes from the original 55 collected resulted from the fact that only one sample of each genotype was collected. Other work in this laboratory has demonstrated that collecting multiple samples of a particular genotype generally ensures survival of at least one sample from which cultures can be established (Pence 2005). At the time of the H. todsenii collections, the development of propaga- tion protocols was the goal, rather than the initiation of multiple genotypes. In subsequent collections, multiple samples have been cultured from each genotype, when possible. Shoot Propagation and Rooting Propagation and further experiments with these cultures centered on the use of three media that have been in standard use in the Endangered Plant Propagation Program at CREW. All were MS basal medium with 0.8% agar replacing gelrite, with no antibiotics or fungicide. The media included a low hormone, general mainte- — nance medium (GM medium); a second propa- gation medium with a higher level of cytokinin (P medium); and a rooting medium (R medium) (Table 1). These were tested with H. todsenii to quantify their effects on growth and rooting. Stock maintained on GM medium was used for all experiments. Cultures in tubes, as described above, were maintained at 26°C, under Cool- White fluorescent lights, 16:8 hrs light:dark cycle, | at approximately 40-60 umol/m2/s photosynthet- _ ically active radiation (PAR). All genotypes were labeled and maintained separately. In Experiment 1, shoot and root production were compared on GM, P, and R media. Shoots | of all 12 lines were cultured on all three media, 20 | shoots of each line on GM and R media and 10 | shoots of each line on P medium. The number of shoots produced per culture, the number of | shoots with roots, and the number of roots per shoot were recorded after two months. This | 2009] TABLE 1. GROWTH REGULATOR CONCENTRATIONS IN MEDIA FOR PROPAGATING H. TODSENI TESTED IN Two EXPERIMENTS. 1) A comparison of three media commonly used in this laboratory; and 2) a comparison of four concentrations of IBA. All media contained MS basal medium with 0.8% agar. Growth regulator (mg/L) Test Medium BAP NAA IBA Experiment | GM 0.1 0.01 oe P 0.5 — _ R a _ 0.5 Experiment 2 0 IBA —- — 0 0.5 IBA _- — 0.5 2 IBA — ~ 2 10 IBA _ _— 10 experiment was done twice with all 12 genetic lines. The shoot multiplication rate for stem pieces with nodes vs. shoot tips was also determined using GM medium alone, and the number of shoots resulting after two months of culture was recorded. This experiment was also done with all 12 genetic lines. In Experiment 2, media with MS basal medium with agar and either 0, 0.5, 2 or 10 mg/L IBA were tested. Although this was done specifically to enhance rooting, the shoot multiplication rate, the number of shoots with roots, and the number of roots per shoot were all recorded after two months. This experiment was done twice with all 12 genetic lines, 10 shoots of each line on each of the four media per replicate. An alternative method for rooting was also tested, using a 7- day pulse on MS basal medium plus agar and 0 or 50 mg/L IBA followed by culture on Woody Plant (WP) basal medium (Lloyd and McCown 1980) plus 0.8% agar and 0.05% activated charcoal (Sigma-Aldrich). After two months on charcoal, the presence or absence of roots was scored, and this experiment was done twice with each of the 12 genetic lines of H. todsenii. Acclimatization Rooted shoots were carefully removed from culture, rinsed to remove medium from the roots, and planted in a moist 4:1:1 mix of commercial play sand: soil mix (a coarse, custom soil mix, Ammon Wholesale Nursery, Burlington, KY): clay (from a wooded area in southwestern Ohio). Because of its availability, an autoclave was used to sterilize the soil. Soil was approximately 3 cm deep, in 10 X 10 X 11.5 cm polystyrene culture _ boxes (Phytotech) with lids. These were incubated in the laboratory under fluorescent lights (a mix of Gro-Lux and CoolWhite), 16:8 hr light:dark cycle, 45 + 10 umol/m?/s PAR, at 21—23°C and ambient humidity outside the boxes. Each container had four, 1 cm diam. drainage holes PENCE ET AL.: IN VITRO METHODS FOR HEDEOMA TODSENII 225 in the bottom and one, 2 cm diam. Sun Biofilter membrane (Sigma-Aldrich) over a hole approx- imately 1 cm in diam. in the lid. The plants were monitored carefully and the boxes watered with purified (reverse osmosis) water when needed to avoid the chlorine and calcium in the local water source. The plants were carefully watched, and if they remained green and unwilted, after 5 wk the lids were loosened and opened slightly. Lids were opened more every 2-3 d until the plants appeared stable and acclimatized, at which time the lids were removed completely. Survival of plants taken from an IBA pulse experiment and from the IBA concentration experiment was compared after 7 wk. Cryopreservation of Shoot Tips Shoot tips, approximately 1 mm in length, were isolated for cryopreservation from cultures 34 to 51 d in age that had been grown on MS basal medium with 0.5 or 1.0 mg/L BAP. After isolation, shoot tips were placed onto GM medium plus 0.3 M mannitol and 10 uM abscisic acid (ABA) with 0.8% agar in 60 * 15 disposable petri plates, 15 ml of medium per plate, for a preculture of 48 hr. Shoot tips were then prepared for liquid nitrogen (LN) exposure using either the encapsulation dehydration procedure (10 trials) (Fabre and Dereuddre 1990) or the encapsulation vitrification procedure (24 trials) (Hirai et al. 1998). With the encapsulation dehydration pro- cedure, 5—10 tips each were recovered as controls after the preculture, encapsulation, and pretreat- ment steps, while 10—20 tips each were tested after the drying and LN exposure steps. With the encapsulation vitrification procedure, 5—10 tips each were recovered as controls after the preculture and encapsulation steps, while 10—15 tips each were recovered after the PVS2 and the LN exposure steps. The percent of moisture remaining in samples of some beads that were air dried during the encapsulation dehydration procedure was deter- mined gravimetrically by weighing the dried beads, placing them in an oven at 95°C overnight and then reweighing the beads. The percent moisture was calculated on a wet weight basis. After LN exposure for at least 30 min, shoot tips from the encapsulation dehydration proce- dure were thawed at ambient temperature for 20 min and transferred to recovery plates of GM medium. Those from the encapsulation vitrifica- tion procedure were thawed 1n a 38°C water bath, rinsed with a solution of 1.2 M_ sucrose, and transferred to plates of GM medium. For recovery growth, shoot tips were incubated under the same temperature and light conditions used for culture maintenance. Survival was measured as the number of shoot tips remaining green and showing growth at 2 wk. 224 MADRONO Data Analysis For growth experiments, data for the 12 genotypes were combined to evaluate the effects of treatment, while data for treatments were combined to evaluate the effects of genotype. Data on shoot and root production and survival through cryopreservation were analyzed using StatView 5.0.1 (SAS Institute Inc.). One-way analysis of variance (ANOVA) was calculated and the significance of differences was deter- mined by using the Tukey-Kremer post hoc test. Genetic Analysis Total genomic DNA was extracted from plants that were micropropagated. Micropropagated plants were initiated from material held at the Arboretum at Flagstaff, originally collected from the more northern of the two population clusters in the Sacramento Mountains. According to records, the collection sites were at or near the following areas of separated populations. HT4 was collected on a north-facing slope. HT3, 11 and 12 were collected about 2200 ft ESE of the collection site for HT4. Samples HT2, 5, 6, 7, 8, 9 and 10 occurred along a slope extending north- east to southwest about 6000 ft ESE of the collection site for HT4. HT1 was collected from a population about 7700 ft south of the collection site for HT4. All plants were collected at elevations above 6200 ft. Weighed samples (ca. 100 mg) were frozen in LN, ground to a fine powder, and extracted by a modification of the CTAB procedure (Stewart and Via 1993). Yields of DNA of at least 30 ug were typical. Concentrations of DNA _ were determined using a Pharmacia Ultrospec Plus spectrophotometer (Uppsala, Sweden). The aver- age ratio of absorbance at 260 nm/280 nm was 1.6. The apparent DNA concentration was adjusted to 50 ug/ml with TE buffer and used without further cleanup. Polymerase chain reactions were run in an Idaho Technology (Salt Lake City, UT) Rapid- cycler using borosilicate glass capillary tubes (part no. 1706). Samples were prepared and run in duplicate or triplicate. Each sample tube of 10 ul contained 50 mM Tris (pH 8), 4 mM MgCl, 20 mM KCl, 0.005 mg BSA, 0.1 mM dNTPs, 0.6 units TAQ (TaKaRa), 0.4 uM ten- base oligomer (ten-mer) and 10 ng genomic DNA. The amplification protocol was | min at 92°C followed by 2 cycles of 7 secs at 42°C, 70 secs at 72°C and 60 secs at 92°C, then 38 cycles of 7 secs at 42°C. 7/0 secs.at 72°C, lsec-at 92 -C with a final 4 min at 72°C. Ten-mers were obtained from Operon Technologies (now www.GeneLink. com) and their numbering system was used. Gels were prepared by microwaving Seachem GTG agarose (BioWhittaker Molecular Applica- [Vol. 56 tions, Rockland, Maine.) at 1.6% in TBA buffer (0.0005% ethidium bromide was added just before the 50°C agarose was poured). Samples from the capillary tubes (10 ul) were mixed with a 2 ul solution of 50% glycerol containing bromo- phenol blue and xylene cyanole as marker dyes. Comparison to a 100 bp ladder was used to estimate fragment size. Electrophoresis was carried out at 80 volts on a 5 X 8 cm gel. The separated fragments were visualized using a UV- transilluminator and photographed in black and white with a digital camera (Sony Mavica FD83). Photos were digitally enhanced to improve contrast using ArcSoft Photo Studio 2.0 (Arc- Soft, Inc., Fremont, CA) software and printed on an HP Deskjet 890C. Scoring of RAPD bands was done directly on the printed output. From an initial group of ten different ten-mers, four were selected based on the reproducible and discrete banding patterns in the RAPD assay: Operon numbers All (CAATCGCCGT), ABOS5 (CCCGAAGCGA), G17 (ACGACCGACA) and X03 (TGGCGCAGTG). The presence or ab- sence of a total of 49 individual bands was scored as a character when a band was present in at least one of the twelve lines of micropropagated H. todsenii. Each band was considered a separate character, and information on the presence or absence of a band was analyzed using the Phylip package of programs (Nei’s Distance Measure, RESTDIST, and the Neighbor-Joining program, NEIGHBOR). The resulting tree indicated the clustering of individual plants into groups based on number of bands in common (Felsenstein 1983, 1985; Backeljau et al. 1995). RESULTS Shoot Propagation An average of over 5 shoots per culture | developed after two months on GM medium, © but propagation rates were increased significantly on P medium (Table 2). Significantly less shoot propagation occurred on R medium, which was used to initiate rooting. There was no difference in the rate of shoot multiplication at 0, 0.5 or | 2 mg/L IBA, but there was a small but significant | increase in shoots on 10 mg/L IBA. Propagation on GM was compared in stem pieces with and without apices. While there was propagation from both explants, there were » significantly more shoots from apical pieces — (6.37 + 0.25) than from nodes of decapitated | shoots (5.68 += 0.25) (P < 0.05). When shoot multiplication rates were com- pared between lines in these three experiments, | there was only one consistent difference between | lines. In each experiment, HT-9 was among the | top three lines in the number of shoots produced, while HT-3 was always among the lowest three. | | 2009] TABLE 2. PENCE ET AL.: IN VITRO METHODS FOR HEDEOMA TODSENII 229 AVERAGE NUMBER OF SHOOTS, ROOTS PRODUCED PER SHOOT, AND PERCENT OF SHOOTS ROOTING IN Two EXPERIMENTS. 1) A comparison of three media commonly used in this laboratory; and 2) a comparison of four concentrations of IBA. All media contained MS basal medium with 0.8% agar. Lower case letters indicate statistical differences for the parameter measured within the experiment (P < 0.01). Test Medium Shoots/culture Experiment 1 GM 5.53 + 0.16a P 8.10 0. 35b R 4.21 + 0.14c Experiment 2 0 IBA 2,80 = 10. 13a 0.5 IBA 3.25 + 0.14a 2 IBA 3.39 + 0.17a 10 IBA 4.23 + 0.21b There was also a significant difference between the number of shoots produced from these two lines in all three experiments (P < 0.05). Rooting When rooting was compared on GM, P, and R media, root initiation occurred on all three media, but there were significantly more shoots producing roots as well as significantly more roots per shoot on the IBA-containing R medium (Table 2). When higher levels of IBA were tested, they did not increase the rate of rooting, and a small decrease in the number of roots per shoot was seen on 10 mg/L compared with 0.5 and 2.0 mg/L IBA. Roots produced on medium with 10 mg/L IBA were generally longer than on other media, but they were more likely to be aerial and produced from nodes above the medium (data not shown). When genetic lines were compared in this experiment, lines HT-2, HT-4, and HT-5 had significantly more rooting than other lines (P < 0.05) in both experiments. Rooting was also obtained from shoots that were cultured for 7 d on 0 or 50 mg/L IBA followed by two months on charcoal medium with no hormones. Rooting from the control (53.2 + 7.5%) was not statistically different from the IBA treated shoots (65.3 = 5.6%) Acclimatization Of 225 plants acclimatized in two rooting experiments, 54% were surviving at one month. There was no statistical difference in the survival rates of plants rooted with the IBA pulse and IBA concentration experiments and no differenc- es in the survival of rooted plants from any of the individual treatments in those experiments or from different propagation media. There also appeared to be no effect of genetic line on survival through acclimatization in these trials. Cryopreservation Survival of shoot tips through cryopreserva- tion ranged from 10-93%, with an average of Roots/shoot Rooting % 0.9 = Ola 21.43 = 65a li33: 2205224 29.85 + 8.0a 3.88 = 0.26b 62,08: 8.50 2.82 + 0.20ac 66.88 + 3.59a 3.61 + 0.23a 69.95 + 5.00a 3.25°2210,.234 70.53 + 4.06a D209 £0196 50.65 6.334 35%, but there was no significant difference in the average rates of survival using encapsulation dehydration (29.9 + 4.5%) and encapsulation vitrification (37.4 + 4.0%). Moisture levels were determined in five of the encapsulation dehydra- tion experiments and ranged from 14.1% to 24.3%, with an average of 20.0%. In this range of moisture, survival ranged from 20% (24.3% moisture) to 53% (22.8% moisture) with no correlation between moisture and survival. Analysis of survival data for the steps in the encapsulation dehydration procedure indicated that there was no significant difference in survival through the first three steps, but that survival decreased significantly with LN exposure (P < 0.01) (Fig. 1). Similarly, with the encapsulation vitrification procedure, there was a significant decrease in survival with LN exposure (P < 0.01). Genetic Analysis In spite of the small number of samples available from each location there was a remark- ably consistent correlation of the within-popula- tion diversity with the collection location (Fig. 2). For example, HT1 and HT4 were collected from sites | and 4 that were separated by the greatest distance, and these had fewer characters in common. They also differed from HT3, HT11, and HT12, which were collected from several plants in the same locale but on a different mountain slope (site 2). HT2 and HTS5-10 were collected from yet another slope (site 3) and, while they showed some differences, HTS, HT8, HT9 and HT10 were remarkably similar. These preliminary results suggest a possible association of low genetic diversity within populations and greater diversity between populations. DISCUSSION This work demonstrates that in vitro methods and biotechnology have the potential for playing a role in the conservation of H. todsenii. This species is found in three small isolated clusters of populations in the mountains of New Mexico and is threatened by its limited numbers, drought, 100 fa i) ide 80 \X \ 40 Percent Survival \ Yy / J 7 7 \ 20 Yi Precult Encap MADRONO Pretr [Vol. 56 Mi Survival Evit Survival EDeh \ oe PVS2 Drying Treatment Step Fic. 1. Survival of shoot tips of H. todsenii through two cryopreservation protocols: encapsulation dehydration (EDeh) and encapsulation vitrification (EVit). Survival was tested after preculture (precult), encapsulation (encap), pretreatment (pretr) (EDeh only), PVS2 (EVit only), drying (EDeh only), and liquid nitrogen exposure (LN). For each protocol, different letters indicate significant differences (P < 0.01). climate change, and its apparent limited ability for sexual reproduction. The methods described here can supplement traditional methods for propaga- tion and germplasm analysis and storage, in order to provide a back-up for the natural population and to assist in its effective conservation. The establishment of the in vitro cultures used in these experiments was done both by IVC and by processing cuttings sent to the lab by overnight shipment, demonstrating that either method can be utilized with this species. The media and methods used for propagating H. todsenii shoots in vitro are similar but not identical to those used elsewhere to propagate H. multiflorum (Koroch et al. 1997). With the latter, half-strength MS medium with a higher level of BAP was used for propagation, while indoleacetic acid (IAA) was used for rooting. With AH. todsenii, higher concentrations of BAP resulted in hyperhydric shoots, and thus the low concentration of BAP in the GM medium was used. In the case of acclimatization of H. todsenii, the method used for propagation or rooting did not appear to affect survival ex vitro. Similarly, Koroch et al. (1997) observed no effect on survival of different sucrose concentrations used in preconditioning H. multiflorum for acclimatization. This work was done as part of the Endangered Plant Propagation Program at CREW, where multiple endangered species are targeted for study at any one time. In dealing with a new © species, a few media that have worked well with © other species for propagation and rooting are © tested first and then modified as needed. Since | mass propagation on a commercial scale is not the goal, demonstrating the technique and . producing a few hundred plants may be done | without optimization of the media. In these studies with H. todsenii, three media that are used frequently in this lab were at the core of these studies, and both shoot and root initiation were measured on all of these. The protocols | established have since been used to initiate | cultures from shoots of new genotypes sent to | CREW as shoots from both an ex situ collection » and a wild population. Based on the results of : these studies, the higher level of cytokinin (P | medium) is used to initiate and build up the number of shoots in cultures, even though these cultures may tend to become hyperhydric if | maintained at length on this medium. Then the lower hormone medium (GM) is used to main- tain and normalize the cultures. Rooting is done on the R medium, as these studies indicated no benefit of increased IBA concentrations. Shoot tips of H. todsenii survived cryopreser- vation protocols and were banked in liquid nitrogen for long-term storage. Successful cryo- preservation of shoot tips has also been reported for other Lamiaceae species, including several 2009] HT7 ** HT2** HT3 *** HT11 *** HT 12 *** HT4 * HT6 ** HTS ** HT8** HT10** HT9** Tere Bre cree cee ena a RAPD analyses of 12 lines of H. todsenii. The four collection sites are indicated by number of asterisks. Note the correlation of the similarity of RAPD characters with the geographical location of the plant. species of Mentha and Solemostemon rotundifo- lius (Towill 1990; Hirai and Sakai 1999: Sakai et al. 2000; Niino et al. 2000) using encapsulation dehydration and vitrification procedures. The effects of cold-acclimation and of the vitrifying solution on Mentha shoot tips have also been studied (Volk et al. 2006; Senula et al. 2007). Ina comparison of three cryopreservation methods using Mentha spp., controlled cooling gave the highest survival, but good survival was also observed using either encapsulation dehydration or vitrification, with some variation depending on the genotype (Uchendu and Reed 2008). In our studies, encapsulation dehydration and encapsu- lation vitrification gave comparable survival with 1. todsenii. _ The results of genetic analysis using RAPDs shows that intra-population diversity was less than that seen between populations, but inter- ‘population diversity in this species was also low. However, only 12 lines were available for analysis, and these all came from only one of ‘the three population clusters of this species: the northern cluster in the Sacramento Mountains. Future studies are planned to obtain more PENCE ET AL.: IN VITRO METHODS FOR HEDEOMA TODSENII 22) samples from the less accessible southern Sacra- mento cluster, as well as from the cluster in the San Andres Mountains, which is approximately 45 mi west of the two Sacramento Mountain clusters. Analysis of genotypes from these other areas will provide a more complete picture of the inter-population diversity of H. todsenii. It should also help direct ex situ conservation efforts by identifying areas with the most diversity within this species. Those areas could then be targets for more intense collecting and storage ex situ. Obtaining additional genotypes will also pro- vide further material for germplasm storage. Since seed production in this species is low and the populations are in remote areas, seeds are rarely available and seed banking is not an adequate option for long-term ex situ conserva- tion. As a back-up to the maintenance of an actively growing ex situ collection at the Arbo- retum at Flagstaff, shoot tips from multiple genotypes of H. todsenii are being maintained at CREW in LN for long-term storage. Although more labor intensive than traditional seed bank- ing, cryostorage of shoot tips may provide the only method currently available for ex situ banking of germplasm from species producing few or no seeds. As an isolated mountain species, H. todsenii is among the species most vulnerable to the effects of climate change, while its small numbers make it vulnerable to drought, fire, or other cata- strophic environmental events. In vitro methods provide a variety of tools that can aid in the conservation of the endangered H. todsenii. In vitro culture, which can be initiated by IVC, if needed, can readily propagate shoots that can be rooted in vitro, providing plants that can be acclimatized to soil. In vitro cultures provide shoot tips that can be cryopreserved as a substitute for seed banking. They also provide material that can be analyzed readily by RAPDs for genetic diversity, which, in turn, can provide direction for targeting further collections in areas of greatest diversity. These ex situ activities can, in turn, provide a back-up to conservation efforts to protect the species in situ and ensure the maintenance of H. todsenii into the future. ACKNOWLEDGMENTS The authors gratefully acknowledge the assistance of Dr. Joyce Maschinski and Sheila Murray at The Arboretum at Flagstaff for making the original collec- tions and sending material from the ex situ collection at TAF for these studies, for providing information on the species, and for helpful comments; Dr. John R. Clark for preliminary studies, and Jennifer Rieger, Lisa Cleveland, Rachel Kennedy, Sean Carr, and Jodi Omnitz for technical assistance. The development of the propagation and cryopreservation protocols was supported, in part, by grants from the Institute of Museum and Library Services, received in collaboration with the Center for Plant Conservation (St. Louis, MO). pH) MADRONO LITERATURE CITED ADAMS, R. P., A. E. SCHWARZBACH, AND R. N. PANDEY. 2003. The concordance of terpenoid, ISSR and RAPD markers, and ITS sequence data sets among genotypes: an example from Juniperus. Biochemical Systematics and Ecology 31:375—387. BACKELJAU, T., L. DE BRUYN, H. DE Wop, K. JORDAEM, S. VAN DONGEN, R. VERHAGEN, AND B. WINNEPENNINCKX. 1995. Random amplified polymorphic DNA (RAPD) parsimony methods. Cladistics 11:119—130. FABRE, J. AND J. DEREUDDRE. 1990. Encapsulation dehydration: a new approach to cryopreservation of Solanum shoot tips. CryoLetters 11:413—426. FELSENSTEIN, J. 1983. Parsimony in systematics: biological and statistical issues. Annual Review of Ecology and Systematics 14:313—333. 1985. Confidence limits in phylogenies: an approach using the bootstrap. Evolution 39:783— 791: HIRAI, D., K. SHIRAI, S. SHIRAI, AND A. SAKAI. 1998. Cryopreservation of in vitro-grown meristems of strawberry (Fragaria x ananassa Duch.) by encap- sulation-vitrification. Euphytica 101:109—115. AND A. SAKAI. 1999. Cryopreservation of in vitro-grown axillary shoot-tip meristems of mint (Mentha spicata L.) by encapsulation vitrification. Plant Cell Reports 19:150—155. HUENNEKE, L. F. 1993. Interaction of breeding system and genetic structure in Hedeoma todsenii (Lamia- ceae), a rare mint of New Mexico, Final report to Center for Plant Conservation and New Mexico Forestry and Resources Conservation Division, Santa Fe, NM. IRVING, R.S. 1979. Hedeoma todsenii (Labiatae), a new and rare species from New Mexico. Madrono 26:184-187. KHANUJA, S. P. S., A. K. SHASANY, A. SRIVASTAVA, AND S. KUMAR. 2000. Assessment of genetic relationships in Mentha species. Euphytica 111: 121-125. Korocn, A. R., H. R. JULIANI, JR., H. R. JULIANI, AND V. S. Trippi. 1997. Micropropagation and acclimatization of Hedeoma multiflorum. Plant Cell Tissue and Organ Culture 48:213—217. KRASNYANSKI, S., T. M. BALL, AND K. C. SINK. 1998. Somatic hybridization in mint: identification and characterization of Mentha piperita (x) M. spicata hybrid plants. Theoretical and Applied Genetics 96:683—687. LINSMAIER, E. M. AND F. SkooG. 1965. Organic growth factor requirements of tobacco tissue cultures. Physiologia Plantarum 18:100—127. Liu, J.. L. WANG, Y. GENG, Q. WANG, L. LUO, AND Y. ZHONG. 2006. Genetic diversity and population structure of Lamiophlomis rotata (Lamiaceae), an endemic species of Qinghai—Tibet Plateau. Genet- ica 128:385—394. LLoyp, G. AND B. McCown. 1980. Commercially feasible micropropagation of mountain laurel, Kalmia_ latifolia, by use of shoot-tip cultures. Proceedings of the International Plant Propaga- tors’ Society 30:421—427. MAKI, M. AND S. Horie. 1999. Random amplified polymorphic DNA (RAPD) markers reveal less genetic variation in the endangered plant Cerastium fischerianum var. molle than in the widespread [Vol. 56 conspecific C. fischerianum var. fischerianum (Car- yophyllaceae). Molecular Ecology 8:145—150. MATTNER, J., G. ZAWKO, M. ROSSETTO, S. L. KRAUSS, K. W. DIXON, AND K. SIVASITHAMPARAM. 2002. Conservation genetics and implications for resto- ration of Hemigenia exilis (Lamiaceae), a serpen- tine endemic from Western Australia. Biological Conservation 107:37—-45. MORDEN, C. W. AND W. LOEFFLER. 1999. Fragmen- tation and genetic differentiation among subpopu- lations of the endangered Hawaiian mint Haplos- tachys haplostachya (Lamiaceae). Molecular Ecology 8:617—625. MURASHIGE, T. AND F. SKOOG. 1962. A revised medium for rapid growth and bioassays with tobacco tissue cultures. Physiologia Plantarum 15:473—497. NINO, T., A. HETTIARACHCHI, J. TAKASASHI, AND P. K. SAMARAJEEWA. 2000. Cryopreservation of lateral buds of in vitro-grown innala_ plants (Solemostemon rotundifolius) by vitrification. Cryo- Letters 21:349-356. PENCE, V. C. 2005. In vitro collecting (IVC) I. The effect of media and collection method on contam- ination in temperate and tropical collections. In vitro Cellular and Developmental Biology-Plant 41:324—332. , J. A. SANDOVAL, V. M. VILLALOBOS A., AND F. ENGELMANN (eds.). 2002. In vitro collecting | techniques for germplasm conservation. IPGRI Technical Bulletin 7. International Plant Genetic Resources Institute, Rome, Italy. SAKAI, A., T. MATSUMOTO, D. HIRAI, AND T. NIINO. 2000. Newly developed encapsulation-dehydration protocol for plant cryopreservation. CryoLetters | 21:53-62. | SENULA, A., E. R. J. KELLER, AND T. SANDUIJAV. | 2007. Cryopreservation of cold-acclimated mint (Mentha spp.) shoot tips using a simple vitrification | protocol. CryoLetters 28:1—12. SKOULA, M., I. EL HILALI, AND A. M. MAKRIS. 1999. | Evaluation of the genetic diversity of Salvia | fruticosa Mill. clones using RAPD markers and | comparison with the essential oil profiles. Bio- | chemical Systematics and Ecology 27:559—568. STEWART, C. N., JR. AND L. E. VIA. 1993. A rapid CTAB DNA isolation technique useful for RAPD | fingerprinting and other PCR applications. Bio- | Techniques 14:748—751. ) SyYDES, M. AND R. PEAKALL. 1998. Extensive clonality © in the endangered shrub Haloragodendron lucasii — (Haloragaceae) revealed by allozymes and RAPDs. Molecular Ecology 7:87—93. TOWILL, L. E. 1990. Cryopreservation of isolated mint | shoot tips by vitrification. Plant Cell Reports | 9:178-180. UCHENDU, E. E. AND B. M. REED. 2008. A compar- | ative study of three cryopreservation protocols for) effective storage of in vitro-grown mint (Mentha: spp). CryoLetters 29:181—188. U.S. FISH AND WILDLIFE SERVICE. 2001, Todsen’s; pennyroyal (Hedeoma_ todsenii) revised recovery, plan. U.S. Fish and Wildlife Service, Albuquerque, _ NM. VOLK, G. M., J. L. HARRIS, AND K. E. ROTINDO. 2006. | Survival of mint shoot tips after exposure to’ cryoprotectant solution components. Cryobiology 52:305—308. MADRONO, Vol. 56, No. 4, pp. 229-237, 2009 SOIL AND COMMUNITY CHARACTERISTICS ASSOCIATED WITH HAZARDIA ORCUTTITI (ASTERACEAE) GEORGE L. VOURLITIS, JULIE MILLER, AND KARI COLER Department of Biological Sciences, California State University, San Marcos, CA 92096 georgev@csusm.edu ABSTRACT Hazardia orcuttii (A. Gray) Greene is a 5-10 dm tall perennial shrub that is native to coastal sage scrub communities of southern California and northern Baja California. This species was listed as threatened by the California Department of Fish and Game in 2002 and is a federal candidate species, and the only known population in the U.S. is on a 1.6 ha mesa located in Encinitas, California. Very little is known about the general ecology of this species, thus, the goal of this research was to characterize the basic soil physical and chemical properties and plant community characteristics associated with this species. Research was conducted between January 2004 and July 2005 in 12.56 m? randomly-located plots that either contained or lacked H. orcuttii. Soil in plots containing H. orcuttii had significantly higher clay, soil organic matter, total N, and soil moisture content than plots lacking H. orcuttii, while plots lacking H. orcuttii had significantly more surface litter content. Significant differences were also observed in plant species abundance between plots containing and lacking H. orcuttii, indicating fundamental differences in plant community composition associated with patches of H. orcuttii. Our data support the notion that H. orcuttii is a soil endemic; however, it is unclear whether H. orcuttii prefers soil richer in clay or is restricted to these soils because of other factors. Given the restricted nature of H. orcuttii, and the proximity of the extant population to residential areas, habitat protection from human degradation and fire should be a high priority. RESUMEN Hazardia orcuttii (A.Gray) Greene es un arbusto perenne que mide 5—10 dm de alto que es nativo a la comunidad perteneciente de la salvia chaparro costeno en el sur de California y ella parte fronterizo de Baja California Norte. Esta especie fue enumerada como una que eata bajo de amenaza por el Departamento de California pescaderia y casa en el 2002, y la unica poblacion en los Estados Unidos, esta en una mesa de 1.6 ha ubicado en Encinitas, Calif. Poco se sabe sobre la ecologia general de esta especie, asi, el fin de esta inverrstigacion era de caracterisar las propiedades basicos fisicos y quimicos de la tierra y las carecteristicas planta associadas con esta comunidad de esta especie. Invistigaciones fueron condujidos entre Enero 2004 y Julio 2005 en 12.65 m2 en parcelas establecidas al azar unos conteniendo y otros careciendo H. orcuttii. Tierra en parcelas conteniendo H. orcuttii tenian significativamente alto niveles de barro, mater organica del suelo, nitrogeno total, y contenido de humedad del suelo que las parcelas careciendo H. orcuttii tenian significatimante mas contenido de revoltura al superficie. Diferencias significativas fueron observados en la abundancia de especie de plantas entre parcelas conteniendo y careciendo de H. orcuttii, indicando diferencias esenciales en el compuesto de la comunidad de las plantas asociado con parches de H. orcuttii. Nuestros datos apoyan la nocion que H. orcuttii es endemica de la tierra sin embargo, no es claro si H. orcuttii prefiere tierra rico en barro o esta limitado a esta tierra por otros elementos. Dado por la naturaleza limitado de H. orcuttii, y la cercania de la proximidad de una poblaciones que existe en areas residenciales, proteccion de los habitos degradantes causados por la humanidad y el fuego derian ser de alta prioridad. Key Words: Asteraceae, biodiversity, chaparral, coastal sage scrub, human impacts, soil, threatened plants species. Habitat loss reduced the extent of coastal sage scrub by 72% from 1970-1990 (Pryde 1992). causing some plant and animal species to become threatened or endangered. One such species, Hazardia orcuttii (A. Gray) Greene (Orcutt’s Hazardia), is a 5-10 dm tall resinous evergreen shrub in the Asteraceae family that is native to Naritime sage scrub-chaparral communities of southern California and northern Baja California ~Hickman 1993) and was listed by the California Department of Fish and Game as threatened in August 2002 (Gogol-Prokurat and Osborne 2002). Specimens collected between 1920 and 1985 indicate that the distribution of H. orcuttii ranged from Encinitas, California to Punta Colonet. Baja California, Mexico. The current distribution is uncertain (Gogol-Prokurat and Osborne 2002) and only two of the 13 previously documented Mexican populations have been located as of 2004. This plant naturally occurs in only one documented location in the United States, on a 1.6 ha mesa (elev. 90-120 m) approximately 5 km from the Pacific coast near Lux Canyon in the Manchester Conservation Area in Encinitas. 230 California. Moreover, plants are distributed in patches that occupy an approximately 0.15 ha area located in the SW corner of the mesa. The number of plants in this population has been estimated to be 50-700 (Oberbauer 1981; Gogol- Prokurat and Osborne 2002; Vourlitis et al. 2006). The population appears to be long established (based on field observations of plant size and woodiness), and voucher specimens from 1979 indicate that the population has been established at Lux Canyon for at least 30 yr. Hazardia orcuttii occurs in a sage scrub-chaparral habitat along with other perennial species such as Rhus integrifolia (Nutt.) W. H. Brewer & S. Watson., Adenostoma fasciculatum Hook. & Arn., and Artemisia californica Less. (nomencla- ture according to Hickman 1993). Little is known about the basic ecology, including population structure and habitat re- quirements of AH. orcuttii; however, previous research suggests that H. orcuttii is restricted to soils with higher clay content (Oberbauer 1981). Research designed to determine the basic ecolog- ical requirements of H. orcuttii is needed to conserve this species (Gogol-Prokurat and Os- borne 2002). Given the current status of this species the main objectives of this research were to characterize the soil physical and chemical properties and the plant community associated with the extant H. orcuttii population at Lux Canyon. MATERIALS AND METHODS Site Description Field measurements were conducted from January 2004—-July 2005 at Lux Canyon (33°1'48"N, 117°15'6"W) in the Manchester Conservation Area in Encinitas, CA. Lux Canyon is approximately 5 km east of the ocean at an elevation that ranges from 10 m above sea level in the valley bottom to 100 m on the mesa top (Center for Natural Lands Management (CNLM) 2005). Vegetation consists of Diegan sage scrub and southern maritime chaparral (CNLM 2005), and the main soil types consist of Altamont clay (Typic chromoxerert) on the mesa top and a loamy, alluvial Huerhuero complex (Typic natrixeralf) on the eroded slopes and valley bottoms (Bowman 1973). Climate data obtained since 1998 from the National Oceanic and Atmospheric Administration for Palomar Airport in Carlsbad, California located approximately 13 km north of Lux Canyon with similar coastal exposure indicates a maritime Mediterranean-type climate with warm-dry sum- mers and cool-wet winters. Average annual rainfall is approximately 200 mm (7.9 in.) and average annual maximum and minimum tem- perature is 19.8 and 12.7°C, respectively. The MADRONO [Vol. 56 wettest month is February with 64 mm of rainfall and the driest month is August with 0.4 mm rainfall. Field Sampling and Data Collection Field plots consisting of 12.56 m* permanent circular quadrats were randomly established in sub-sites containing H. orcuttii (n = 13 plots; hereafter referred to as “‘H. orcuttii plots’’) and in adjacent sub-sites lacking H. orcuttii (n = 10 plots; hereafter referred to as “non-H. orcuttii plots”). As mentioned above, sub-populations of HI. orcuttii are restricted to a 0.15 ha portion of the mesa top. Within this area circular plots were established within patches containing HZ. orcuttii and patches lacking H. orcuttii using a random coordinate system. We attempted to pair each plot containing H. orcuttii with a plot lacking H. orcuttii, but the spatial distribution of paths and shape of the vegetation fragments precluded an adequate paired-design resulting in unbalanced replication for H. orcuttii and non-H. orcuttii plots. Measurements of plant species abundance were conducted over a total of 4 sample campaigns (January and July of 2004 and 2005) in the H. orcuttii plots and non-H. orcuttii plots described above. All individuals rooted within each plot (Chapman 1976; Barbour et al. 1999) were counted and measured for width along 2-axes (the maximum width and the axis perpendicular to the maximum width) and height from the ground surface to the top of the shrub (Bonham — 1989). Soil and surface organic matter (litter, which is dead plant matter >1 mm in diameter) was collected in April 2004 to coincide with the | spring growing season and the main period of seed germination. Samples were obtained from | the non-H. orcuttii plots (n = 10) and a subset of the H. orcuttii plots (n = 10) to preserve as | much as possible the paired-sampling design | between H. orcuttii and non-H. orcuttii plots. Surface litter was collected within a 312.5 cm? rectangular quadrat that was centered on a randomly chosen point in each plot. After litter removal, soil samples were obtained from surface (0-10 cm) and subsurface (30-40 cm) soil layers using a 173.5 cm* bucket auger. Soil. samples were transferred from the core samplers | to polyethylene sample bags and immediately returned to the lab and stored at 4°C until) analysis. | Sample Processing and Data Analysis Plant species density was quantified as the number of individuals per species per unit plot. area and cover was quantified as the area of each shrub species per plot divided by the area of the. f yo 2009] plot. The area (A) of each individual shrub was calculated as nD’/4, where D is the average diameter of each individual calculated from the measurements of maximum and _ perpendicular width. Frequency of occurrence was calculated as the number of plots that a particular species was encountered. Indices of relative abundance were calculated from the estimates of absolute abun- dance by dividing a given absolute abundance for a particular species by the total abundance of all species. These relative indices were combined to yield an estimate of the index of relative importance (JRJ), which was calculated as the sum of the individual relative density, cover, and frequency of occurrence indices (Chapman 1976; Barbour et al. 1999). Soil samples were sieved to remove rocks and organic matter =2 mm in size prior to laboratory analyses. Litter samples were passed through a | mm sieve to remove mineral debris, dried at 70°C for 1 wk, weighed, and ground to pass through a 40 mesh sieve. Total N and P content of soil and litter was quantified using micro-Kjeldahl methods (Bremner 1996). Percent gravimetric soil water was calculated as [(M; Ma) M,]\*100 where Mr was the fresh mass of soil and M, was the mass of soil after drying at 105°C (Robertson et al. 1999). Percent soil organic matter was quantified by combusting soil at 700°C for 1 h in a muffle furnace (Nelson and Sommers 1996). Soil bulk density was calculated as the mass of dry soil per unit volume (Robertson et al. 1999). Soil particle size distribution was measured using the Bouyoucos hydrometer method (Gee and Bauder 1986). Cluster analysis was used to assess the similar- ity in community composition between the habitats containing and lacking H. orcuttii. For this analysis, the plot values of the /R/ of each species were used to determine the degree of similarity between discrete plots, and the “Euclidian Distance’” method was used to determine the relative distance between each plot (Hintze 2005). Soil physical and chemical data were analyzed using a 2-way analysis of variance (ANOVA) with site (H. orcuttii vs. non-H. orcuttii plots) and depth (0-10 and 30— 40 cm) treated as fixed effects. Data were tested for normality and heteroscedasticity using the Anderson-Darling and Levene’s tests, respec- tively. Data violating assumptions of normality and heteroscedasticity were LN-transformed to fulfill the assumptions of ANOVA. Cluster analysis and ANOVA were performed using NCSS (version 2004, Kaysville, Utah, USA). Differences in litter pool biomass and N and P content between H. orcuttii and non-H. orcuttii _ plots were analyzed with a randomized-t-test (Sokal and Rohlf 1995) using MS-Excel (Chris- tie 2004). VOURLITIS ET AL.: HABITAT CHARACTERISTICS OF HAZARDIA ORCUTTII 25) $216.79" D: 1.09 SxD: 0.19 0-10 30-40 Mmmm Present _ Absent 5 a 0 200 400 600 800 a Sand content (mg/g) a) Shao hy4e wale D: 3.68 SxD: 0.02 0-10 30-40 @mumm@ Present (7) Absent 0 100 200 300 400 Clay content (mg/g) Fic. |. The mean (+SE) sand (A) and clay (B) content of soil in plots where Hazardia orcuttii was present (closed bars; n = 10 plots) or absent (open bars; n = 10 plots) at different soil depths. Also shown are the results of a 2-way ANOVA (F-statistic) where sub- site (habitat) and depth were fixed effects (df = 1,36 for sub-site (S), depth (D), and the sub-site X depth interaction (S*D)), ***P = 000i: **P — 0.0L. =P < 0.05. RESULTS Soil Physical and Chemical Properties Sand content of surface (0-10 cm) and sub- surface (30—40 cm) soil in areas lacking A. orcuttii was significantly higher than in areas containing H. orcuttii (Fig. 1A; F,36 = 16.79; P < 0.001). Clay content of soil in areas with H. orcuttii was significantly higher than in areas lacking H. orcuttn (hie EB. diya6, = 13.25, P= 0:005). Differences 1n particle size distribution were large enough that plots containing H. orcuttii were characterized as having a sandy clay loam soil while plots lacking H. orcuttii were characterized as having a sandy loam soil. Plots containing H. orcuttii also had significantly higher soil water content (Fig. 2A), especially in the sub-surface, and a significantly higher soil organic matter (SOM) content (Fig. 2B). Total soil N content (Fig. 3A) was significantly higher in plots con- taining AH. orcuttii, especially in the surface soil 252 S: 7.36* Di2(.(2"* SxD: 0.39 0-10 30-40 Mmm Present Absent € 0 50 100 150 200 250 = Soil water content (mg/g) ros oO fa S: 10.88** D: 1.68 SxD: 2.11 0-10 30-40 Gm Present Absent 0 10 20 30 40 50 60 Soil organic matter content (mg/g) FIG. 2. The mean (+SE) water (A) and organic matter (B) content of soil in plots where Hazardia orcuttii was present (closed bars; n = 10 plots) or absent (open bars; n = 10 plots) at different soil depths. Also shown are the results of a 2-way ANOVA (F-statistic) where sub- site (habitat) and depth were fixed effects (df = 1,36 for sub-site (S), depth (D), and the sub-site X depth interaction (S*D)): ***P =< 0.001; **P < 0.01: *P = 0.05. layer. In contrast, total soil phosphorus (P) content was similar for plots with and without H. orcuttii (Fig. 3B). The N and P concentration of surface litter was not statistically different between AH. orcuttii and non-H. orcuttii habitats; however, because sur- face litter biomass was nearly 3-times higher in non-H. orcuttii plots (P = 0.003; Fig. 4A), litter N and P pools sizes were significantly higher in non-H. orcuttii plots (Fig. 4B, C). Thus, while H. orcuttii plots had significantly higher SOM and total soil N, plots lacking H. orcuttii had a significantly higher surface litter pool and litter N content. Community Composition Cluster analysis of the sample plots indicated two discrete vegetation assemblages (Fig. 5) based on the importance values (JR/) of the plant species. The first group consisted of all of the MADRONO [Vol. 56 S: 8.68** D: 46.99*** SxD: 0.08 Gm Present (7) Absent 5 o 20 40 60 80 100 € Total soil N (gN/m?) a 0-10 30-40 Gam Present C——) Absent 0 2 4 6 8 10 A2 14 16 18 Total soil P (gP/m?) Fic. 3. Mean (+SE) total soil N (A) and P (B) content in plots where Hazardia orcuttii was present (closed bars; n = 10 plots) or absent (open bars; n = 10 plots) at different soil depths. Also shown are the results of a 2-way ANOVA (F-statistic) where sub-site (habitat) and depth were fixed effects (df = 1,36 for sub-site (S), depth (D), and the sub-site X depth interaction (S*D)). ee POLO P= 0 Ole PaO, plots lacking H. orcuttii (“‘B”’ plots; Fig. 5), while the second major grouping consisted of the plots where H. orcuttii was present (““A”’ plots; Fig. 5). This clustering described the maximum variation in the sample plots (r = 0.89), and indicated two discrete sub-communities at Lux Canyon. Some species were common to both sub- communities, including Adenostoma fasciculatum, Artemisia californica, Eriogonum fasciculatum Benth., Quercus dumosa Nutt., Rhus integrifolia, | and Deinandra fasciculata (DC.) Greene (Ta- ble 1). However, A. fasciculatum, Q. dumosa, and | R. integrifolia had higher mean values of /RJ over the 2-year field study in non-H. orcuttii plots, while FE. fasciculatum, Dudleya edulis (Nutt.) Moran, and D. fasciculata were more abundant | in H. orcuttii plots (Table 1). Other species including Mimulus aurantiacus Curtis, Xylococcus bicolor Nutt., Yucca schidigera Ortgies, and Y. | whipplei Torr. were conspicuously lacking in H. orcuttii plots, while Lotus scoparius (Torr. & A.: Gray) Ottley, Ferocactus viridescens (Torr. & A. 2009] T T p = 0.003 | 2000 Litterpool biomass g/m? (gN/m?) Litterpool N content Litterpool P content (gP/m?) Absent Hazardia orcuttii Present Fic. 4. Mean (+SE) litter biomass (A) and total litter N (B) and P (C) pool sizes in plots where Hazardia orcuttii was present or absent (n = 10 plots per sub- site). Also shown are the results of a randomized t-test, where the probability of committing a type-I error (P- value) was calculated over 1000 iterations. 0.5 0.4 Dissimilarity » FIG. 5. VOURLITIS ET AL.: HABITAT CHARACTERISTICS OF HAZARDIA ORCUTTII 253 Gray) Britton & Rose, and Baccharis pilularis DC. were observed only in H. orcuttii plots (Table 1). Significant differences in total density and cover were also apparent between H. orcuttii and non-H. orcuttii plots (Fig. 6). Plots lacking H. orcuttii had significantly higher cover in 2004 (Fig. 6A), which explains in part the significantly higher surface litter pool (Fig. 4A). Plots lacking H. orcuttii also had significantly lower density than A. orcuttii plots during each measurement campaign (Fig. 6B) suggesting that, on average, non-H. orcuttii plots were dominated by fewer but larger shrubs. One of the most abundant shrubs in plots lacking H. orcuttii was A. fasciculatum (Table 1), which 1s a chaparral shrub that can reach heights of 2 m and an area of 3- 4 m° (Munz 1974; Riggan et al. 1988). Temporal variations 1n stand cover and density were relatively higher in H. orcuttii plots than in non-H. orcuttii plots. For example, H. orcuttii plots experienced a nearly 4-fold increase in stand cover (Fig. 6A) and a more than 5-fold increase in stand density (Fig. 6B) between July 2004 and January 2005. Over the same period plots lacking H. orcuttii experienced no increase in stand cover (Fig. 6A) and a 2-fold increase in stand density (Fig. 6B). Thus, H. orcuttii plots experienced larger temporal variation in overall plant species abundance than non-H. orcuttii plots, which may have implications for H. orcuttii recruitment and survival. Temporal variation in the index of relative importance (JR/) of the six dominant shrub species (excluding H. orcuttii) present in both Non-H. orcuttii plots A6 H. orcuttii plots Cluster analysis of the index of relative importance (JR/J) values of plant species observed in plots containing (A1-13) and lacking (B1-10) Hazardia orcuttii. 234 TABLE lI. MADRONO [Vol. 56 MEAN (+ISD) INDEX OF RELATIVE IMPORTANCE (JR/J) FOR ALL SPECIES OBSERVED IN 12.56 mM? PLOTS CONTAINING (PRESENT; N = 13) AND LACKING (ABSENT; N = 10) HAZARDIA ORCUTTIT AT LUX CANYON, ENCINITAS, CALIFORNIA. Data were collected over 4 sampling campaigns conducted in January and July of 2004 and 2005. 7RI was multiplied by 100; — not observed. Nomenclature and habitat data according to Hickman (1993). Species Adenostoma fasciculatum Hook. & Arn. Artemisia californica Less. Dudleya edulis (Nutt.) Moran Eriogonum fasciculatum Benth. Hazardia orcuttii (A. Gray) Greene Ferocactus viridescens (Torr. & A. Gray) Britton & Rose Mimulus aurantiacus Curtis Opuntia littoralis (Engelm.) Cockerell Quercus dumosa Nutt. Rhus integrifolia (Nutt.) W. H. Brewer & S. Watson Xylococcus bicolor Nutt. Yucca schidigera Ortgies Yucca whipplei Torr. Lotus scoparius (Torr. & A. Gray) Ottley Deinandra fasciculata (DC.) Greene Marah macrocarpus (Greene) Greene Dichondra occidentalis House Centaurium venustum (A. Gray) B. L. Rob. Anagallis arvensis L. Cholorgalum parviflorum Wats. Zigadenus fremontii (Torr.) S. Watson Dichelostemma pulchella (Salisb.) Heller Baccharis pilularis DC. Eriophyllum confertiflorum (DC.) A. Gray Gnaphalium californicum DC. Gnaphalium sp. Cryptantha sp. Stephanomeria sp. Thistle Unknown annual Unknown herbaceous perennial Unknown grass : : plot types revealed substantial changes in shrub species composition over 2004—2005, especially in the AH. orcuttii plots (Fig. 7). In A. orcuttii plots, A. fasciculatum and D. fasciculata exhibited 3—4 fold increases in /R/J during the study period, while A. californica and E. fasciculatum experi- enced a 2~—3-fold decline in /R/J (Fig. 7). In contrast, plots lacking H. orcuttii experienced substantially less temporal variation in relative abundance; however, R. integrifolia was an obvious exception (Fig. 7). These data indicate rapid and dynamic species turnover in H. orcuttii plots and more stable community dynamics in non-H. orcuttii plots. DISCUSSION Soil Physical and Chemical Properties Some soil physical and chemical properties at Lux Canyon were significantly different between plots containing and lacking AH. orcuttii. One HAazardia orcuttii Present Absent 23,0) 23,3 66.8 + 9.8 64.6 = 36.4 49.9 + 8.9 | Be ete WA ae Ie =the WO AD = D2? 1 0 eee 32.4: 155 — 132.0 —— — 37.6 = 4,1 | es eee | 55+ 0.5 212 el 1720 S45 134022 7.6 47-6 = 187 — 2322007 — 8.62 4.1 — 23° 1.9 8.2 + 3.4 — 55.6 = 79.2 99+ 12.5 — 39+ 54 — | aus J — le 34 03 = 0.6 0.6 = 13 S002 eS 10.9 + 21.8 0.4 + 0.9 — O09 21-7 0.5 + 1.0 0.4 + 0.8 — — 0:67 2 152 0.4 + 0.7 L832" 347 0.6 + 1.3 O62 1.2 — 0:5 22 11 O5'=09 — 8.6 + 16.0 526 2 720 0.9 + 1.7 L214 — 0.6 + 1.2 10.7. 218-7 28 a 8 a3 potentially key difference was soil texture, where H. orcuttii plots had significantly higher clay content and lower sand content than non-H. orcuttii plots (Fig. 1). This observation is sup- ported by previous research and is consistent with the notion that H. orcuttii may be a soil endemic (Oberbauer 1993). Gravimetric soil water content was significantly higher in H. orcuttii plots (Fig. 2A), which presumably reflects the higher | clay content of the soil. Soil texture controls a | variety of processes that control plant species distribution, including soil water holding capac- | ity, nutrient retention, organic matter stabiliza- | tion, seed germination, and seedling recruitment (Baskin and Baskin 1990; Oberbauer 1993; Schimel et al. 1985; Kluse and Doak 1999; Walck | et al. 1999: Hook and Burke 2000). However, it is | unclear why H. orcuttii at Lux Canyon is) restricted to soil with higher clay content. For | example, percent germination of H. orcuttii seeds | was actually higher in soil types that had lower clay content (Miller 2008); thus, H. orcuttii is 1 ) 2009] apparently not restricted to soil with higher clay content because of seed germination. Haczardia orcuttii plots had higher soil organic matter (SOM) content than non-A orcuttii plots (Fig. 2B), which presumably indicates differences in plant species composition and/or rooting depth between habitats (Jobbagy and Jackson 2000). The increase in SOM was apparently not due to an increase in aboveground litter input because the surface litter pool was nearly 3-fold lower in A. orcuttii plots (Fig. 4A). Rather, H. orcuttii plots had a higher abundance of shrubs associated with coastal sage scrub (Table 1), which typically have a shallower, more horizon- tally-distributed root system than species charac- teristic of evergreen chaparral (Hellmers et al. 1955). Hazardia orcuttii plots also had higher total soil N, suggesting higher overall fertility compared to non-H orcuttii plots (Marion and Black 1988). Given that SOM represents a large pool of N in terrestrial soils (Hook and Burke 2000), these results presumably reflect the signif- icantly higher SOM content observed in H. orcuttii plots. Surface litter was more than 3-times higher in non-H. orcuttii plots, which has important implications for the germination and recruitment of H. orcuttii. For example, the germination and seedling recruitment of Chorizanthe pungens Benth. var. hartwegiana Reveal and Hardham (Polygonaceae) is reportedly inhibited by chap- arral vegetation, possibly as a result of allelopa- thy and/or the development of a larger surface litter pool that alters the microclimate and reduces light availability at the soil surface (Kluse and Doak 1999). Results from germination experiments indicate that percent germination of HI. orcuttii seeds was significantly lower in complete darkness, which may simulate light conditions under a deep surface litter layer (Miller 2008). While there are substantial seasonal variations in the production of above- ground litter and the size of the surface litter pool in coastal sage scrub and chaparral (Vourlitis et al. 2009), the difference in the litter pool between plots containing and lacking //. orcuttii observed in April coincides with the most active time for seed germination for chaparral and coastal sage shrubs. The relatively larger surface litter pool observed in non-H. orcuttii plots may inhibit seed germination and/or seedling recruitment, which may be one impor- tant mechanism causing H. orcuttii to be restricted to more clay-rich soils. Community Composition Our results indicate fundamental differences in species composition between H. orcuttii and non- H. orcuttii plots (Fig. 5). These differences are presumably due in part to spatial variations in VOURLITIS ET AL.: HABITAT CHARACTERISTICS OF HAZARDIA ORCUTTII —@— Present 3.0 O- Absent 20 Total cover (m? plant area/m? ground area) © on © ros) = aS Total density (individuals/m? ground area) 0 Jan04 Jul04 Jan05 Jul05 Measurement period FIG. 6. Average (+SE) total shrub cover (A) and density (B) in plots containing Hazardia orcuttii (closed- circles, solid-lines; n = 13 plots) and plots lacking orcuttii (open-circles, dotted-lines; n = 10 plots). Differences in mean cover and density between habitat types were assessed using a two-sample t-test. * P < 0.052."* P= 0.01. soil texture, which exerts a strong influence on soil water availability and plant species distribu- tion (Westman 1981). Hazardia orcuttii plots had significantly lower cover and higher density (Fig. 6) and more rapid and dynamic species turnover (Fig. 7) than plots lacking H. orcuttii, and it is possible that these interspecific dynamics affect H. orcuttii recruitment, survival, and fecundity. For example, the higher temporal variation in species composition in H. orcuttii plots implies higher variation in the intensity of competitive interactions, availability of “‘safe sites’ important for Hl. orcuttii recruitment, and/or available resources (Menges 1990; Watson et al. 1994; Kluse and Doak 1999; Walck et al. 1999). Similar interspecific controls on plant growth, fecundity, and recruitment have been observed for other Asteraceae including Coreop- sis lanceolata L. (Folgate and Scheiner 1992), Ratibida columnifera (Nutt.) Wooton & Standl. (Vargas-Mendoza and Fowler 1998), Solidago shortii Torr. & A. Gray (Walck et al. 1999) and Deinandra conjugens (D. D. Keck) B. G. Baldwin (Bauder et al. 2002). Presumably similar process- es may be important in limiting the local distribution of H. orcuttii. N WwW oy 0.9 K++ | 0.0 | 4 1 ; 0.8 -Ef —@®- Present] 98/Ri —@— Present : [ O- Absent } Mes O-- Absent 0.4 1 04+ 1 O Index of relative importance (/R/) 0.0 | t t t 0.0 aieeeeetn af | Df Ou 1.51 0.20 + | t 5 1.0 + O 0.10 0.5 ¢ 1 0.05 j | 4 pom | 0.0 | O_O 1} 9.90 Jan04 Jul04 Jan05 Jul05 Jan04 Jul04 Jan05 Jul05 FIG. 7. Temporal variation in the index of relative importance (JR/) for the six dominant plant species Adenostoma fasciculatum (Af), Artemisia californica (Ac), Eriogonum fasciculatum (Ef), Rhus integrifolia (Ri), Deinandra fasciculata (Df), and Quercus dumosa (Qd) in plots containing AH. orcuttii (closed-circles, solid-lines) and plots lacking H. orcuttii (open-circles, dotted-lines). Together, individuals of these species made up 70% of all shrubs and perennials observed in both plot types. Note the difference in scale for Df and Qd. Conclusions and Conservation Eecommendations Rabinowitz et al. (1986) define a rare species as one that has a restricted geographic range, narrow habitat tolerance, and/or low local abundance. Hazardia orcuttii appears to possess all three of these traits. For example, the only known extant U.S. population of AH. orcuttii is restricted to a 1.6 ha mesa near Lux Canyon, Encinitas, CA. Within this narrow geographic range, H. orcuttii individuals appear to be restricted to patches of soil with high clay content, but it is unclear whether H. orcuttii is restricted to clay soils because of intrinsic (1.e., seed germination and/or recruitment) or extrinsic (interspecific competition) factors. Knowledge of why H. orcuttii is restricted to more clay-rich soils will undoubtedly inform and improve the success of future conservation efforts. Estimates from 2002 indicated approximately 600 plants at Lux Canyon (Gogol-Prokurat and Osborne 2002), while more recent estimates (Vourlitis et al. 2006) indicated a population size of 510 individuals, suggesting that the population has declined since 2002. The low local abundance, restricted geo- MADRONO [Vol. 56 graphic range, and narrow habitat tolerance suggest that H. orcuttii is rare by all criteria. Protection of Lux Canyon from human degradation and fire should be a high priority. The Lux Canyon site is used for recreation purposes, and human activities lead to the creation of paths, trampling and damage of vegetation, accumulation of waste, and urban runoff. These threats increase the potential for catastrophic fire, which can either damage or completely eliminate that only known U. S. population of H. orcuttii. ACKNOWLEDGMENTS This research was supported through a grant provided by the California Department of Fish and Game. Assistance with field and laboratory research provided by James Coler, Kristen Faulkner, Ron Kirker, Courtney Nance, Markus Spiegelberg (Center for Natural Lands Management), Jaime Willson, Mark Wolcott and Lauren Zuniga is gratefully appreciated. LITERATURE CITED BARBOUR, M. G., J. H. BURK, W. D. PITTS, F. S. GILLIAM, AND M. W. SCHWARTZ. 1999. Terrestrial plant ecology, 3rd ed. Benjamin Cummings, Menlo Park, CA. BASKIN, J. M. AND C. C. BASKIN. 1990. Seed germination ecology of poison hemlock, Conium maculatum. Canadian Journal of Botany 68:2018— 2024. BAUDER, E. T., J. SNAPP-COOK, AND J. SAKRISON. 2002. Ecology and management of Deinandra conjugens (D. D. Keck) B.G. Baldwin (Otay Tarplant). Final Report. California Department of Fish and Game, Region 5, San Diego, CA. BONHAM, C. D. 1989. Measurements of terrestrial vegetation. John Wiley & Sons, Inc., New York, NY. BOWMAN, R. 1973. Soil survey of San Diego County, California. Soil Conservation Service and United States Department of Agriculture, Washington, DC. BREMNER, J. M. 1996. Nitrogen-total. Pp. 1085—1122 in D. L. Sparks, A. L. Page, P. A. Helmke, R. H. Loeppert, P. N. Soltanpour, A. Tabatabai, C. T. Johnson, and M. E. Sumner (eds.), Methods of soil analysis: Part 3. Chemical Methods. Soil Science Society of America Book Series No. 5, Soil Science | Society of America, Inc., Madison, WI. CENTER FOR NATURAL’ LANDS (CNML). Website http://www.cnlm.org/cms/ [ac- cessed 27 January 2010]. CHAPMAN, S. B. 1976. Methods in plant ecology. Halsted Press, Inc., New York, NY. MANAGEMENT | CHRISTIE, D. 2004. Resampling with Excel. Teaching | Statistics 26:9—14. FOLGATE, L. A. AND S. M. SCHEINER. 1992. Distribu- | tion of a restricted locally abundant species: effects of competition and nutrients on Coreopsis lanceo- — lata. American Midland Naturalist 128:254—269. GEE, G. W. AND J. W. BAUDER. 1986. Particle-size analysis. Pp. 383-412 in A. Klute (ed.), Methods of soil analysis: part | physical and mineralogical methods, 2nd ed. Soil Science Society of America . | 2009] Book Series No. 5, Soil Science Society of America, Inc., Madison, WI. GOGOL-PROKURAT, M. AND M. OSBORNE. 2002. Report to the Fish and Game Commission on the status of Orcutt’s Hazardia (Hazardia orcuttii). Status Report 2002-01. California Department of Fish and Game, Habitat Conservation Planning Branch, Sacramento, CA. HELLMERS, H., J. S. HORTON, G. JUHREN, AND J. O’KEEFE. 1955. Root systems of some chaparral plants in southern California. Ecology 36:667—678. HICKMAN, J. C. (ed.). 1993. The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. HINTZE, J. 2005. NCSS and PASS. Number Cruncher Statistical Systems, Kaysville, UT. Hook, P. B. AND I. C. BURKE. 2000. Biogeochemistry in a shortgrass landscape: control by topography, soil texture, and microclimate. Ecology 81: 2686-2703. JOBBAGY, E. G. AND R. B. JACKSON. 2000. The vertical distribution of soil organic carbon and its relation to climate and vegetation. Ecological Applications 10:423—436. KLUSE, J. AND D. F. DOAK. 1999. Demographic performance of a rare California endemic, Chor- izanthe pungens var.hartwegiana (Polygonaceae). American Midland Naturalist 142:244—256. MARION, G. M. AND C. H. BLACK. 1988. Potentially available nitrogen and phosphorus along a chap- arral fire cycle chronosequence. Soil Science Society of America Journal 52:1155—1162. MENGES, E. S. 1990. Population viability analysis for an endangered plant. Conservation Biology 4:52-62. MILLER, J. 2008. Reproductive biology and germina- tion requirements of the California state threatened species Hazardia orcuttii (Asteraceae). M.S. thesis. California State University, San Marcos, CA. Munz, P. A. 1974. A flora of southern California. University of California Press, Berkeley, CA. NELSON, D. W. AND L. E. SOMMERS. 1996. Total carbon, organic carbon, and organic matter. Pp. 961-1010 in D. L. Sparks, A. L. Page, P. A. Helmke, R. H. Loeppert, P. N. Soltanpour, A. Tabatabai, C. T. Johnson, and M. E. Sumner (eds.), Methods of soil analysis: part 3. Chemical methods. Soil Science Society of America Book Series No. 5, Soil Science Society of America, Inc., Madison, WI. OBERBAUER, T. 1981. Noteworthy collections: Hazar- dia orcuttii (Gray) Greene (Compositae). Madrono 28:38. VOURLITIS ET AL.: HABITAT CHARACTERISTICS OF HAZARDIA ORCUTTII 23) . 1993. Soils and plants of limited distribution in the Peninsular Ranges. Fremontia 21:3—7. PRYDE, P. R. 1992. San Diego: an introduction to the region. Kendall-Hunt Publishing Co., Dubuque, IA. RABINOWITZ, D. H., S. CAIRNS, AND T. DILLON. 1986. Seven forms of rarity and their frequency in the flora of the British Isles. Pp. 182-204 in M. E. Soule (ed.), Conservation biology: the science of scarcity and diversity. Sinauer Associates, Inc., Sunderland, MA. RIGGAN, P. J., S. GOODE, P. M. JACKS, AND R. N. LOCKWOOD. 1988. Interaction of fire and commu- nity development in chaparral of southern Califor- nia. Ecological Monographs 58:155—176. ROBERTSON, G. P., D. C. COLEMAN, C. S. BLEDSOE, AND P. SOLLINS (eds.). 1999. Standard soil methods for long-term ecological research. Ox- ford University Press, New York, NY. SCHIMEL, D., M. A. STILLWELL, AND G. R. Woop- MANSEE. 1985. Biogeochemistry of C, N, and Pina soil catena of the shortgrass steppe. Ecology 66:276—282. SOKAL, R. R. AND F. J. ROHLF. 1995. Biometry. 3rd ed. W. H. Freeman and Company, San Francisco, CA. VARGAS-MENDOZA, M. C. AND N. L. FOWLER. 1998. Resource-based models of competitive interactions. I. Intraspecific competition in Ratibida columnifera (Asteraceae). American Journal of Botany 85:932— O30; VOURLITIS, G. L., J. KIRKER, AND K. COLER. 2006. Research for the management and conservation of Orcutt’s Hazardia (Hazardia orcuttii). Final Re- port. California Department of Fish and Game, Habitat Conservation Planning Branch, Sacra- mento, CA. , S. C. PASQUINI, AND R. MUSTARD. 2009. Effects of dry-season N input on the productivity and N storage of Mediterranean-type shrublands. Ecosystems 12:473-488. WALCK, J. L., J. M. BASKIN, AND C. E. BASKIN. 1999. Effects of competition from introduced plants on establishment, survival, growth and reproduction of the rare plant Solidago shortii (Asteraceae). Biological Conservation 88:213—219. WATSON, L. E., G. E. UNO, N. A. MCCARTY, AND A. B. KORNKVEN. 1994. Conservation biology of a rare plant species Eriocaulon kornickianum (Erio- caulaceae). American Journal of Botany 81:980— 986. WESTMAN, W. E. 1981. Factors influencing the distribution of species of Californian coastal sage scrub. Ecology 62:439-455. MADRONO, Vol. 56, No. 4, pp. 238-245, 2009 ECOLOGY AND GROWTH OF WHITELEAF MANZANITA WITHIN A PONDEROSA PINE PLANTATION IN SOUTHWEST OREGON T. J. HANSON Forest Ecologist, FEC Consulting, LLC, Beaverton, OR 97075-2184 MICHAEL NEWTON! Department of Forest Science, Oregon State University, Corvallis, OR 97331-5704 Mike.newton@oregonstate.edu ABSTRACT Sclerophyll shrubs such as Arctostaphyllos viscida Parry (Ericaceae; whiteleaf manzanita) are often considered obstacles to afforestation. The establishment of conifer plantations in southwest Oregon presents challenging problems for initial seedling survival and subsequent growth. This region is characterized by hot dry summers, cool moist winters, and rocky, shallow soils at low elevations. The growth of competing hardwoods within conifer plantations creates the traditional problem of undesirable competition for the silviculturist. Accumulation of manzanita biomass also contributes heavily to fire hazard while at the same time providing biomass of potential value as biofuel. This study describes productivity of whiteleaf manzanita stands of varying density growing within the confines of Pinus ponderosa Dougl. ex C. Lawson var. ponderosa (Pinaceae; ponderosa pine) plantations of the same age on poor sites. We explore the ability of this shrub to generate biomass as a potential energy source. Growth prediction equations are for height, basal diameter, biomass per individual, and biomass per hectare by stand density. The accumulated biomass after 14 yr of growth ranges around 43 metric tons/ha, and represents a strong negative influence on the supposed productivity of interplanted pines of the same age. Energy content of biomass amounts to an accumulation rate of 1.6 * 10° megajoules/ha/yr on such sites, indicating a large biofuel potential by age 14 or more. Harvesting the manzanita will also provide a potential benefit to productivity of the pines. Key Words: Arctostaphylos viscida, biomass, competition, growth habit, manzanita, stand density, yield. Growth equations for sclerophyllous shrubs elevations in much of California west of the are rare, as are reports of their value as resources. The search for energy sources in non-commercial forest species may have potential for exploitation both for fuel, and also for benefiting commer- cially valuable conifers. We examine growth patterns of whiteleaf manzanita on hot, dry sites in southwest Oregon as a function of stand density and age following spacing and weed control. These equations are one part of a multi- phased project demonstrating the impact that indigenous hardwood competition has on the supposed productivity of conifer plantations in this region. Current interest in biomass suggests further interpretation of these data toward potential utilization. summit of the Sierra Nevada Mountains. It may be considered a weed species when mixed with conifer plantations (Newton and Cole 2008), but we here evaluate its role as a potential resource. Toward this end, we quantify the growth of | whiteleaf manzanita within young ponderosa pine plantations and document its potential role as a biomass source. The whiteleaf manzanita data presented here represent the first long-term stand-based dataset | for this species while under management. It describes the growth response of whiteleaf | manzanita growing at densities of 1700 to | 27,000 shrubs per hectare within the confines of | a ponderosa pine plantation in southwest Ore- — gon. We summarize the history of the study ~ below, relying on allometrics for individual plants from other studies of whiteleaf manzanita in the | area that adapt growth data to biomass (Latt © 1985; Hughes et al. 1987; Minore et al. 1988). Whiteleaf manzanita is a common sclerophyll shrub indigenous to southwest Oregon. It regen- erates prolifically from a seed bank after a major disturbance such as mechanical site preparation prior to planting conifer plantations. This species is abundant in interior sites at low elevations through much of southwest Oregon and mid- MATERIALS AND METHODS The research areas for this study are nested | ' Present address and correspondence: Department of | Within three mixed conifer plantations growing | Forest Engineering, Resources and Management, Ore- gon State University, Corvallis, OR 97331. on mechanically cleared shrub lands in the> Applegate Valley of southwest Oregon (for — . 2009] HANSON AND NEWTON: GROWTH HABITS AND YIELD OF WHITELEAF MANZANITA 239 detailed site descriptions, see White and Newton 1989). The original design of these plantations was to provide an opportunity to study compe- tition and growth of Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco var. menziesii) and ponderosa pine seedlings under various competi- tion levels of whiteleaf manzanita and herbs. The original conifers were thinned to 14 ponderosa pine and 6 Douglas-fir per treatment, a density that offered negligible competition during the first five years, but an increasing amount later. Most of the planted Douglas-fir succumbed to moisture deficiency in their seventh and eighth year as a result of unusually high summer temperatures and the superior moisture extrac- tion and stress tolerance of whiteleaf manzanita. Our study uses a randomized block design with three replications in each of which are six 0.04-ha square plots containing six different densities of shrubs in a randomly located geometric progres- sion of shrubs per hectare, separated in density by a factor of 2 from highest density to lowest. Treatment designation is based on spacing between individual manzanitas plants, in feet, square spacing equivalent. Treatment 2, for example reflects spacing of 2 * 2 feet (60 xX 60 cm; 27,000 shrubs/ha) and treatment 8 represents 8 X 8-foot spacing (240 x 240 cm; 1700 shrubs/ha). Within each plot, a subsample of 25 shrubs in clusters of 5 was selected for repeated measure- ment in a systematic sample with location at a random starting point. All shrubs were measured at ages 3, 4, 5, 7, 8 and 14 yr, the age span in which they had full crowns. Measurements in year 21 indicated severe suppression and crown dieback under influence of ponderosa pines, and are excluded from our analysis. All shrubs and pines were established the same year. All replications are within 4 km of each other, centered at approximately 42°45’N, 123°02’W, and 640-750 m elevation (Delorme Mapping Company 1991). All sites are in southern Jackson Co., Oregon, and are on southwest or southeast aspects with approximately 10—20 percent slope. The soils at two replications vary between 50 and 80 cm deep, with local deep spots up to 150 cm. Most are clay loam of the Vannoy soil series. The soils at the third replication appear more productive and are of the Caris-Offenba- cher soil series (Stearns-Smith and Hann 1986). _ These soils are gravely loams with the same depth as the Vannoy series. The geologic parent material for all three replications is of metamor- phic-sedimentary origin. Original estimated coni- _ fer growth was Site V or poorer for Douglas-fir (height at 50 yr <17 m, per McArdle et al. 1961) and Site IV or poorer for ponderosa pine (height ~<13 m at 50 yr, per Meyer 1938). Height and breast height age measurements of local domi- nant trees places the site index between 12 and 27 m at 50 yr. Adjacent stands of whiteleaf manzanita roughly 70 yr old range between 3-4 m tall, with some stems reaching 180 mm diam. at the base, indicating productive sites in terms of adaptation of this manzanita species. Our stands originally had a component of Ceanothus cunea- tus (Hook.) Nutt. (common buckbrush.), ordi- narily found on less productive sites. In addition to whiteleaf manzanita, other vegetation includes Arbutus menziesii Pursh (Pacific madrone), Rhus diversiloba Torr. & Gray (poison oak.), Bromus tectorum L. (downy brome), Galium aparine L. (bedstraw), Caucalis microcarpa Hook. & Arn. (California hedge parsley), Madia elegans D. Don ex Lindl. (tar- weed), Verbascum thapsus L. (common mullein), Centaurea solstitialis L (yellow star-thistle.), Quercus kelloggii Newberry (California black oak), Fragaria vesca L (woods strawberry), and Epilobium minutum Lindl. ex Lehm. (willow- weed). These species were removed in the study plots except that herbs were retained at one density of manzanita (13,500/ha) after an initial One-year treatment with hexazinone to permit establishment of the ponderosa pine seedlings. Annual precipitation varies between 60 and 150 centimeters throughout the region (Stearns- Smith and Hann 1986), these sites are 1n the driest subregion of southwest Oregon. Approximately 65 to 75 percent of this rainfall occurs between November and February (Hobbs et al. 1992). The sites of all replications had burned about 1930, resulting in large areas of even-aged white- leaf manzanita. Scattered ponderosa pine and Douglas-fir occurred in cove sites (draws) nearby. All study sites were cleared in preparation for this research in 1980, following domination by near- pure stands of manzanita for 50 yr. Site preparation for planting consisted of scarification by crawler tractors equipped with brush blades. Brush was pushed into windrows and burned. The tractor scarification created an ideal germination seedbed for the persistent and abundant seeds from whiteleaf manzanita. Pro- fuse germination resulted in a patchy distribution with some areas producing 500,000 seedlings per hectare (White 1989). After burning the wind- rows, the area was ripped down to a 45 cm depth with a crawler tractor along the contours at approximately two-meter intervals. Afforestation planting occurred in the spring of 1981 with equal numbers of two-year-old (2 + 0) bareroot Douglas-fir and ponderosa pine stock. Spacing was about 1600 seedlings per hectare. After planting, one site was sprayed with 1.7 kg/ha hexazinone for herbaceous weed control. The other two sites were mulched with 90 * 90 cm laminated Kraft paper impregnated with asphalt. The herbicide treatment had no observable effect on the manzanita or Ceanothus germinants, or on the vigor of oak and madrone sprouts. 240 MADRONO [Vol. 56 TABLE 1. SUMMARY OF WHITELEAF MANZANITA (ARVI) TREATMENT INSTALLATIONS. There are three replications of each treatment. 'Completely randomized within blocks. Treatment ID Treatment description! 2 27,000 AR VI/hectare 28 13,500 ARVI/hectare 28h 13,500 ARVI/hectare w/ herbs 4 6720 ARVI/hectare 56 3360 ARVI/hectare 8 1700 ARVI/hectare Installation of the original study occurred in early spring 1983. Control of manzanita density for the various treatments was with chemical thinning or interplanting natural seedlings. The treatments ranged from zero to 27,000 shrubs per hectare in regular square spacings, using 20— meter-square plots (0.0405 ha; 0.1 acre). Table 1| presents a description of the treatments and densities of manzanita remaining after age 3. At the time of installation, the conifer density ranged from 1050 to 1760 seedlings per hectare depend- ing upon mortality during the first two years after planting. Researchers removed all shrubs other than designated whiteleaf manzanita. Planted manzanita seedlings were wild stock carefully dug up on-site from outside plot boundaries and planted the same day wherever gaps existed in the planned grid spacings. All plots had comparable initial densities of manzanita germinants, thus avoiding the confounding between the probabil- ity of establishment and site conditions. Treatment establishment and subsequent her- baceous weed control involved several chemical applications over a period of years. Initially, researchers covered the desired manzanita and conifer seedlings with soft-drink cups prior to broadcast spraying a mixture of 3.3 kg/ha simazine, 2.8 kg/ha glyphosate, and 3.8 kg/ha 2,4-D. The total volume of this application was 121 liters per ha (White 1989). In early 1984, an additional broadcast spray of 4.4 kg/ha of simazine controlled herbaceous weeds. Oak and madrone sprouts were killed with directed sprays of glyphosate or 2,4-D with care taken to avoid manzanitas. Whiteleaf manzanita individual-shrub param- eters were recorded on permanent samples of 25 shrubs per plot with numbered tags and periodic measurements. Although the measurements did not occur every year after 1985, measurements did occur in 1987, 1989, 1994, and 2002. A small reduction in the number of tagged shrubs over time resulted from natural mortality under intraspecific competitive stress. Owing to wide- spread dieback of crowns observed in the 2002 measurements that compromised estimates of gross growth, our estimates of biomass here will be limited to the first 14 yr. Three studies (Latt 1985; Hughes et al. 1987; Minore et al. 1988) developed leaf area and ARVI per ha Sq. m per ARVI 26,910 0.3716 13,455 0.7432 13,455 0.7432 6727 1.4864 3364 29129 1682 5.9458 biomass equations for whiteleaf manzanita. The studies by Minore et al. (1988) and Latt (1985) used datasets collected from open-grown white- leaf manzanita. We used the equations developed by Hughes et al. (1987) as having been derived from shrubs in varying densities of stands rather than individual shrubs. We did not verify with complete biomass measurements of our own, but did consult with Hughes in choice of equations and proximity of his data sources to our study (southwest Oregon). The variation between growing conditions for the data used to produce the biomass equation (Hughes et al. 1987), and the growing conditions on the treatments in this study may represent a source of error. However, this was the only biomass equation available that approximated the growing conditions found in young conifer plantations, and was selected for suitability to the range of sizes in our study. Analysis is by non-linear regression, reflecting exponential and logarithmic parameters. Above- ground yield prediction equations are given for whiteleaf manzanita growing at varying densities for 14 yr within a conifer plantation ecosystem. These prediction equations use two independent variables. The first is the number of residual whiteleaf manzanita stems per hectare. The second is the plantation age, which is also the age of the manzanita from seed. Even though these shrubs were measured at age 21, influence of interplanted ponderosa pine confounded yields by causing death to large portions of crowns, an | observation that led to rejection of that data set in our yield equations. For the regressions in this study, the natural log of the manzanita growth parameter and of | age was taken before running the regression procedure. Square root transformations helped to | emphasize the rapid drop in growth when shrubs | are subjected to a small amount of competition. | Also, regressions were fit against individual replications rather than treatment means. Resid- | ual plots showed independence of data and, — therefore, a correct transformation. Each growth | regression presented is a single regression of a_ growth parameter on age and shrub density. The | methodology was set forth by the Quantitative | Sciences Group of the Department of Forest Science (Sabin and Stafford 1990). With this, we | 2009] HANSON AND NEWTON: GROWTH HABITS AND YIELD OF WHITELEAF MANZANITA 241 Fic. 1. Height/age relationship as influenced by stand density for whiteleaf manzanita (ARVI). Accelerating height trend with reduced density is highly significant, as is reduced height growth age. organized the original regressions into a number of terms starting with the main effects and ending with cubed interaction terms. Regressions were then run using the REG procedure within the SAS™ statistical software. This software pro- vides a P-value level for each term in the regression. Terms not achieving a P-value of at least five percent were backed out of the regression and the procedure was run again. After several iterations, all remaining terms had a P-value less than or equal to the five percent level, achieving a good fit to the data. RESULTS The growth form of manzanita shrubs on the lowest-density treatments remained open for the first eight years. While open grown, the size of whiteleaf manzanita is predictable for individual shrubs, as described by Muinore et al. (1988). Mortality and decreased height growth of indi- vidual shrubs became apparent as the treatments reached crown closure, but total growth per hectare continued to increase. The continued rise in growth rate per hectare per year indicates that crowns were coalescing toward full site occupan- cy beyond age 14 despite mortality of some manzanita. Manzanita height growth approaches maxi- mum increments within the first 10 yr (Fig. 1). Height increment continues somewhat more rapidly at wide spacings than at the closest spacing in a pattern that suggests even greater _ heights at wider spacing than shown in our data. This figure shows height is increasing at a decreasing rate, yet adjacent 60+ year shrubs _ were three to four meters tall in the absence of _ conifers, presumably having reached their maxi- mum height potential. 120.0 —e— Trimt 2 —a-= Trtmt 28 100.0 —e— Trtmt 28h Trtmt 4 —w=—= Trtmt 56 Trtmt 8 ie x = 80.0 ie d — ® Bi £ 60.0 = . & Q SA rc) % 40.0 eae = . fa) —_ 20.0 eZ: 3 4 5 6 7 8 9 10 11 12 13 14 Age (years) Fic. 2. Basal diameter of whiteleaf manzanita (ARVI) growing at varying densities within ponderosa pine plantations in southwest Oregon, including treatment 28 with herbs. Raw means of all replications. Increasing divergence of height curves with age and differences in shrub density is highly significant. Diameter growth, hence basal area, is very responsive to spacing (Fig. 2). Treatment 8 (1700 ARVI/ha; 2.4 * 2.4 m spacing), with the least dense manzanita, has the highest basal diameter growth while treatment 2 (27,000 ARVI/ha; 60 x 60-cm spacing), the densest manzanita, has the lowest basal diameter growth but the highest biomass growth. The effect of herbaceous vege- tation is evident in treatment 28 with 13,500 ARVI/ha (85 X 85-cm spacing). This treatment 1s consistently larger in diameter than treatment 28h (85 X 85 cm, with herbs after year 1) that has the same density of manzanita without herba- ceous control. Figure 2 shows the negative influence of herbs by being out of sequence in the array of diameter growth curves for the array of stand densities. This aberration does not appear in the response surface of diameter on age and spacing that illustrates interaction of age x density (Fig. 3). Figure 4 shows the average aboveground oven- dry stem volume biomass per manzanita shrub computed from equations developed by Hughes et al. (1987). Density control strongly influences the net biomass of individual manzanita shrubs. High-density plots produce smallest individuals, but the greatest total biomass (Fig. 5). In view of the large sizes of widely spaced shrubs, one might expect that as self-thinning occurs on dense stands, yields of various densities will merge. However, at age 14 the yields are continuing to diverge in favor of dense stands despite the onset of mortality. The density of manzanita on all treatments provides a sufficient amount of intraspecific Basal Diameter (mm) Fic. 3. Basal diameter growth for whiteleaf manzani- ta (ARVI), by age and initial density. The increase in diameter with decrease in shrub density and increasing age is highly significant. competition to cause mortality by age 14. Table 2 describes the loss in stems per hectare at several points in the lives of these stands due to mortality at the different treatment densities. The data from Applegate plots are not of sufficient quantity (25 shrubs per plot) to allow useful regression for changes in shrub popula- tions. Mortality varied sporadically by age depending upon the starting number of manza- nita stems. By age 14, the percent mortality roughly aligns with treatment density, ranging from eight percent for the least dense treatment to 21 percent for the most dense. The only treatment with herbs, 28h, had the highest mortality with 24 percent. Although the manza- nita have now shaded out all herbs, they still Biomass (kg/ ARV!) FIG. 4. Oven dry stem biomass for individual white- leaf manzanita (ARVI), by age and initial density. Increase in biomass with decreasing density of shrubs is significantly greater with increasing age. lower densities and MADRONO [Vol. 56 eae Co = fe) = ” ” © 5 s 14 o 2 ~~ © PIG? 5: manzanita (ARVI), by age and initial density. Yields are adjusted for mortality. Increase in yield with increasing density of shrubs is highly significant and increases with age. show residual effects of herb competition in their first several years of growth. The various growth parameters are influenced differently by density and its interaction with age. Height growth, (Fig. 1), does not show a large variation between the low and high density treatments although heights are greater with wider spacing. However, basal area, being a squared function of diameter and the largest contributor to biomass, declines sharply with just a small amount of intraspecific competition. Figure 4, oven-dry stem biomass per shrub, demonstrates that a small amount of intraspecific competition triggers a substantial decrease in biomass. When biomass is converted to a per hectare basis, the highest biomass is found in the treatments that have the largest number of stems. This relationship of biomass to density persists through age 14; seen in Fig. 5 where mortality is taken into consideration. The largest increment in total biomass yield with increasing density from the widest spacing 1s in the difference between 1682 and 3364 shrubs per ha. Further increments in yield with increas- ing density are smaller, and are potentially | compromised by the apparent relation between | density and mortality. Equations of the response surfaces shown in | Figs. 1, 3, 4, and 5 are available from the | corresponding author. DISCUSSION The whiteleaf manzanita in this study represent | a view of potential opportunities for capturing | biomass from plants that may otherwise have | neutral or negative value, and/or which may be | able to produce significant yields on sites too harsh for production of higher-value forest Oven dry stem biomass per ha for whiteleaf 7 i 2009] HANSON AND NEWTON: GROWTH HABITS AND YIELD OF WHITELEAF MANZANITA 243 TABLE 2. MORTALITY SUMMARY FOR WHITELEAF MANZANITA (ARVI) AT THE END OF THE 2ND, 7TH, AND 13TH GROWING SEASON AFTER PLANTATION ESTABLISHMENT. 'Treatment descriptions are given in Table 1. ARVI stems per hectare (percent mortality) Treatment ID! Initial Age 2 ; 26.910 26.910 (0) 28 13,455 13.455 (0) 28h 13.455 13.455 (0) 4 6727 6727 (0) 56 3364 3364 (0) 8 1682 1682 (0) species. The manzanita on our plots are growing at densities representative of conditions in both fire-originated stands and in managed conifer plantations within this region. Although manza- nita’s competitive influence on conifers is well documented (White 1989; Ortiz-Funez 1989; Newton and Cole 2008), the value of manzanita as a bio-fuel is generally unrecognized either as a salvage material or as a crop. Several studies have described similar yields of shrubs in unregulated densities (Zavitkovski and Newton 1968; Hughes et al. 1987; Minore et al. 1988) as measured either in total yield or on terms of individual shrub growth. Zavitkovski and Newton (1968) displayed with Ceanothus veluti- nus Douglas ex Hook. var. velutinus a pattern of upper asymptote in yield at age 14, more or less when not suppressed by conifers. Revisitation of those stands in 2003 finds most of the shrubs suppressed or dead, usually under influence of conifers (Newton personal observation). Our data here originate in uniform stands with a continuum of measurements that terminate at age 14 when yields were increasing rapidly. They are also in full view of older stands with much higher biomass and larger sizes of individual shrubs. These observations suggest the possibility of a potentially much larger estimate of asymptotic maxima. Our stands were mixed with ponderosa pines at a low stocking. The ponderosa pines did not cause high mortality of shrubs until after age 14, but the last yields measured probably reflect a degree of suppression despite showing a positive flexion in growth curves at that point. Pure stands of these shrubs at the wider spacings we used would undoubtedly produce yields of large shrubs with totals greater than any measured in this study. We therefore regard these data as a beginning in understanding shrubs as a crop. While it is not unlikely that growth would continue to increase in the absence of planted conifers, the initial purpose of the study included retention of pines as the dominant crop, thus extension of production cycles beyond 14 yr may not be appropriate. _ The biomass growth of whiteleaf manzanita is done at the expense of sequestering water Age 7 Age 14 26,192(2.7) 2d LOO 4213s) 12.737°,3) 11,482 (14.7) i122 10,226 (24.0) 6189 (8.0) 5920 (12.0) 3274 (2.7) 3050 (9.3) 1592 (5.4) 1547 (8.0) resources needed by the conifer component (White and Newton 1989; Newton and Cole 2008), perhaps with less effect on the manzanita than on the conifers. The prediction equations (available from the corresponding author) dem- onstrate manzanita’s potential to utilize the site resources, of which the most limiting is water. When water resources are diminished, the com- petitiveness of intermingled conifers is apparently reduced more because of differences in ability to extract water from bedrock (Zwieniecki and Newton 1995). The effect of manzanita compe- tition on conifers is relatively insensitive to manzanita density (Newton and Cole 2008); hence if one is to grow mixed species, high density manzanita will provide maximum growth of this species so long as pine spacing is wide enough for shrubs to develop with adequate sunlight for a decade or more. Harvest of manzanita is likely to release the already domi- nant ponderosa pine (Newton and Cole 2008). The precise age for optimum harvest will need definition by experiments that evaluate harvest at different ages for different densities of pine and manzanita. Within the manzanita stand, height growth responds to manzanita density. The best height growth occurs where manzanita density is the lowest. The least favorable height growth occurs at high densities and where herbs are not controlled. For optimum development of a whiteleaf manzanita/ponderosa mixed culture, wide spacing of pines and medium spacing of manzanita appear to have merit, especially if herbs are suppressed. Herbaceous cover is apparently able to deplete enough surface water (down to 75-cm soil depth) to impact whiteleaf manzanita, a species capable of extracting soil and rock moisture when these shrubs were 5—7 yr old at xylem pressures of —7.0 MPa. Suppression of herbs favored both pines and manzanitas. Mortality of both Douglas-fir and manzanita may have contributed to thrift of residual shrubs. Competition from manzanita eliminated the Douglas-fir on all treatments except those with the lowest manzanita density. The ability to secure soil moisture provides manzanita with a decided survival advantage over coniferous com- 244 petitors in southwest Oregon. Thus, Douglas-fir apparently had little effect on manzanita growth by depletion of soil water, but the dense foliage of Douglas-fir would, in any case, lead to suppres- sion of this shade-intolerant shrub had water stress not interfered. Figures 4 and 5 demonstrate the power in numbers of small shrubs. Even though individual manzanitas in dense stands are small, the high stem density produces the highest biomass per hectare as long as heavy shade is not a factor. At age 14 (1994 growing season) treatment 2 contained almost 61 percent more aboveground biomass than the next highest manzanita density. Although there was some compensatory growth on individual shrubs with each increment of spacing, biomass increment per hectare at age 14 increased with each doubling of density. As stands age and crowding sets in at wider spacings, growth differences between various spacings may merge. By age 14, the biomass of treatment 28 without herb control (13,500 stems/ha) and the next treatment density (treatment 4, with 6720 stems/ha) were virtually equal. Early herb control for four years literally doubled the biomass per shrub. Major differences are apparent in biomass accumulation resulting from varying levels of shrub density at different ages. As in any plant population, resources per plant decrease with increases in density. The average size of an individual manzanita on treatment 2 is 2.25 kg while the average size of an individual manzanita on treatment 8 is 6.14 kg, a difference by a factor of 2.7 resulting from area of occupation 16 times larger. The aboveground biomass of individual manzanita plants relates to the crown cover. The observed differences in water use and competitive influence was remarkable in early years (White and Newton 1989) but as crowns of shrubs coalesced, differences in interaction of shrubs and pines decreased (Newton and Cole 2008). Interaction with overstory ponderosa pine likely compromises long-term projection of man- zanita growth. We observed evidence of pine release with removal of manzanita at age 21 (Newton and Cole 2008) and also observed deteriorating condition of manzanita as conifers gained in dominance after age 20. Yet shrub stands in the vicinity now over 70 yr old were dominated by very large individual manzanita shrubs even though within the range of ponder- osa pine seedfall. These sites were clearly totally occupied for most of their age span. We postulate that where pine and manzanita are to be managed as mixed stands, harvest or other removal of the shrubs roughly 15 yr after establishment will allow the conifers to develop into a productive overstory. Conversely, if manzanita is managed in pure stands, it may be feasible to grow the shrubs to much _ higher MADRONO [Vol. 56 volumes in a longer rotation in order to increase yield per hectare per year and minimize harvest cost per unit of yield. The growth form of whiteleaf manzanita lends itself to relatively efficient utilization of this shrub for biomass. The single stems at the base would permit severance near the ground to allow transfer into a processor of some description with little residue on the ground. Absolute conversion efficiency into usable biomass will depend on equipment development. However, the conversion of biomass to energy, based on dry weight of biomass of all above-ground manzani- ta, would be roughly 15-50 metric tons per hectare at age 14, depending on density of the stand, with an energy content of 0.56 to 1.86 megajoules per ha per yr of growth when grown to this age. The data on manzanita/conifer mixtures pro- vide valuable information for an understanding of conifer growth during the first 25 yr after plantation establishment. Manzanita 1s superior at commandeering moisture held at high tension in soil and rock and it can maintain an expansion in biomass extending to an age when mixed hardy conifers begin to reach commercial size. Whereas the impact of manzanita on conifer growth during these years is substantial, this study demonstrates an opportunity to harvest a crop of manzanita biomass early in a rotation while leaving dominant and released ponderosa pine to produce a commercial harvest later. This experiment indicates that biomass accu- mulation by manzanita is greater than that of conifers that co-exist with the shrubs for over a decade. A shrub that can accumulate 40+ tons per ha in 14 yr on sites of marginal productivity for conifers represents a significant source of fuel, both for commercial exploitation as a biofuel or extractive source, or for hazard of wildfire. Fire- risk shrub stands are presently being treated at substantial cost with massive machinery to reduce risk of catastrophic fires near homes and highways. Equipment similar to that used for reduction of fuel in this way is evidence that similar equipment could harvest the manzanita in forms useful for power generation or even commercial heating facilities. Whiteleaf manza- nita does not sprout; harvest of the shrubs before the mixed conifers succumb offers an opportunity to facilitate long-term conifer production while making an asset of a major competing species. CONCLUSIONS Sclerophyllous shrubs in southwestern Oregon and California represent a large source of biomass of potential importance as: a) a source of fuel or fiber, b) a severe fire hazard, or c) an obstacle to conifer regeneration on moderately productive sites. Whiteleaf manzanita, as a ) ; 2009] HANSON AND NEWTON: GROWTH HABITS AND YIELD OF WHITELEAF MANZANITA representative of this community, will accumulate several tons per hectare per year which, if accumulated for decades, represents a fuel load capable of intense heat if not utilized or controlled. This paper presents the first yield equations for whiteleaf manzanita growing under variable densities in a plantation environment in south- west Oregon. Yields per hectare per year are greater than those of juvenile conifers even when conifers dominate the site in the first decade. The lost early volume growth of conifers attributable to competition is offset during the first decade- plus, more or less, by increasing manzanita production. We expect plantation dynamics to shift toward the conifer component with me- chanical removal of the shrubs as per Newton and Cole (2008). This will vary in time and intensity depending on the initial manzanita density. The conifer component does not appear to have the ability to overcome this initial growth loss during the first few decades, and pinpointing optimum timing for harvest of shrubs with release of conifers needs refinement. As biofuels gain in value, potential yields of shrub communities, mixed stands and associated tree species will gain in importance. Developing an ability to accurately predict shrub and conifer productivity, growing in pure stands and mix- tures on such harsh sites represents a new area in forest mensuration that has yet to be addressed. ACKNOWLEDGMENTS Diane White, Liz Cole, Laural Witherspoon-Joos, Rob Pabst, Tom Hughes, Mary O’Dea and Atilio Ortiz-Funez participated in various elements of stand establishment, data collection and analysis in early years. Maciej Zwienecki and Ruth Willis provided assistance in data recording and shrub measurements. The USDI Bureau of Land Management provided land, site preparation and administrative support for this work. LITERATURE CITED -DELORME MAPPING COMPANY. 1991, Oregon atlas and gazetteer: topo maps of the entire state. Delorme Publishing Co., Freeport, ME. | ‘Hosss, S. D., S. D. TEscH, P. W. OwsTon, R. E. STEWART, J. C. TAPPEINER, AND G. E. WELLS. | 1992. Reforestation practices in southwestern Oregon and northern California. Forest Research ’ 245 Laboratory, Oregon State University, Corvallis, OR. HUGHES, T. F., C. R. LATT, J. C. TAPPEINER, AND M. NEWTON. 1987. Biomass and leaf-area estimates for varnishleaf ceanothus, deerbrush, and whiteleaf manzanita. Western Journal of Applied Forestry 2:124-128. LATT, C. R. 1985. The development of leaf area and biomass in the whiteleaf manzanita (Arctostaphylos viscida Parry) brushfields of southwest Oregon. M.S. thesis, Oregon State University, Corvallis, OR. MCARDLE, R. E., W. H. MEYER, AND D. BRUCE. 1961. The yield of Douglas-fir in the Pacific Northwest (revised). USDA Technical Bulletin 201. Washing- ton, DC: MEYER, W. H. 1938. Yield of even-aged stands of ponderosa pine. USDA Technical Bulletin 630. Washington, D.C. MINORE, D., H. G. WEATHERLY, AND J. E. MEANS. 1988. Growth of whiteleaf manzanita (Arctostaph- ylos viscida Parry). Forest Science 34:1094—1100. NEWTON, M. AND E. C. COLE. 2008. Twenty-six-year response of Douglas-fir and ponderosa pine plantations to woody competitor density in treated stands of madrone and whiteleaf manzanita. Forest Ecology and Management 256:410—420. ORTIZ-FUNEZ, A. 1989. Survival and growth of Douglas-fir and ponderosa pine during eight years of whiteleaf manzanita and herb competition in southwest Oregon. M.S. thesis, Oregon State University, Corvallis, OR. SABIN, T. E. AND S. G. STAFFORD. 1990. Assessing the need for transformations of response variables. Special Publication 20, Forest Research Laborato- ry, Oregon State University, Corvallis, OR. STEARNS-SMITH, S. C. AND D. W. HANN. 1986. Forest soil associations of southwest Oregon. Stock Number 1435, Forest Research Laboratory, Ore- gon State University, Corvallis, OR. WHITE, D. E. 1989. Competitive interactions between Douglas-fir or Ponderosa pine and whiteleaf manzanita. Ph.D. thesis, Oregon State University, Corvallis, OR. AND M. NEWTON. 1989. Competitive interac- tions of whiteleaf manzanita, herbs, Douglas-fir and Ponderosa pine in southwest Oregon. Canadi- an Journal of Forest Research 19:232—238. ZAVITKOVSKI, J. AND M. NEWTON. 1968. Ecology of snowbrush, (Ceanothus velutinus); its role in forest regeneration in the Oregon Cascades. Ecology 49:1134-1145. ZWIENIECKI, M. A. AND M. NEwTON. 1995. Roots growing in rock fissures: their morphological adaptation. Plant and Soil 172:181—187. MADRONO, Vol. 56, No. 4, pp. 246—278, 2009 VEGETATION AND FLORA OF A BIODIVERSITY HOTSPOT: PINE HILL, EL DORADO COUNTY, CALIFORNIA, USA JAMES L. WILSON Department of Biological Sciences, Sierra College, 5000 Rocklin Rd., Rocklin, CA 95677 DEBRA R. AYRES! Department of Evolution and Ecology, University of California, One Shields Ave., Davis, CA 95616 drayres@ucdavis.edu SCOTT STEINMAUS Department of Biological Sciences, California State Polytechnic University, 1 Grand Avenue, San Luis Obispo, CA 93407 MICHAEL BAAD Department of Biological Sciences, California State University, Sacramento, 6000 J Street, Sacramento, CA 95819 ABSTRACT Pine Hill lies near the center of a gabbrodiorite intrusion in the foothills of the Sierra Nevada mountain range in El Dorado County, CA, USA. We assembled an extensive flora, examined the distribution and associations of vascular plant taxa, and specifically focused on associations of six rare plant taxa. The influence of environmental variables on plant distribution was investigated using a stratified random plot sampling technique and applying canonical correspondence analyses. The site contained over 10% (741 plants) of the flora of the entire state of California including seven rare species. Species segregated into chaparral, oak woodland, and grassland communities. In chaparral and woodland, and on serpentine sites, over 75% of the flora was comprised of native species. The non-serpentine grassland community was dominated by exotic species (64% exotic) and contained no rare species. Shrub and tree cover were the most important biotic factors associated with plant species distribution; serpentine substrate, soil texture, elevation, and degree of disturbance were the most important abiotic factors. Five rare species were restricted to gabbro soils. Consideration of beta diversity contributed little to our understanding of vegetation patterns. Our analyses identified two types of chaparral which we termed ‘Xeric Seeding” and “‘Mesic Resprouting” to reflect the environmental conditions and the fire regeneration strategy of the vegetation. Each chaparral type contained different rare species whose regeneration strategies were concordant with chaparral regeneration type. Key Words: CANOCO, canonical correspondence analysis, chaparral, gabbro, obligate resprouter, obligate seeder, rare plants, TWINSPAN. Mediterranean-climate regions are known for the high diversity of their flora, collectively containing almost 20% of the world’s vascular plant species while comprising an area less than 5% of the earth’s surface (Cowling et al. 1996). This is due to a combination of factors acting at local to regional scales such as plant growth-form diversity and differential responses to distur- bance, plant assemblages composed of habitat specialists and geographical vicariants, and spa- tial variation in resources due to topographic diversity and edaphic complexity (Cowling et al. 1996). In California shrublands, edaphic special- ists, and patches in which varied seral stages occur following fire add to floristic richness. ' Author for correspondence. Located near the center of a gabbro soil: formation in the Sierran foothills 48 km east of; Sacramento, CA, Pine Hill stands as one of; California’s remarkable ‘‘ecological islands’’. (Stebbins 1978), possessing a rich floristic diver-, sity and a high concentration of rare and. endangered plants (Fig. 1). The vegetation con-' sists of open grassland, oak woodlands, and: chaparral. The Pine Hill complex forms a 104 km?; gabbrodiorite volcanic intrusion of Mesozoic origin (approximately 175 million years in age),§ that is surrounded by metamorphic rocks, with some granitic intrusions, and serpentine rock) lands (Springer 1968). Serpentine occurs as rocky/} outcrops or as ridges which extend in a north- south direction. At the time this study was begun} in the mid-1980s, at least six rare and endangered } plant taxa were considered to exist only on Pine Hill or in the immediately surrounding areas) 2009] FIG. |. Green Valley Rd WILSON ET AL.: BIODIVERSITY OF PINE HILL 247 Salmon South Falls Rd Fork f American River Rescue - Gabbro Soil Intrusion Map showing the location of the gabbro soil intrusion which extends from S U.S. highway 50 to the South Fork of the American River, encompassing the towns of Cameron Park and Rescue. The center of the gabbro soil intrusion is at approximately 38°43’ north latitude and 120°59’ west longitude. (Howard 1978; El Dorado County 2007; Baad personal obervation). Since these species were only known from gabbro soils at this locality, it appeared as if the rare plants were restricted to soils derived from gabbro parent materials. Serpentine areas serve as an important edaphic comparison with gabbro. Serpentine is classed as an ultrabasic or ultramafic, cold intrusive rock. It is high in ferro-magnesium silicates and is especially noted for its low calcium and high magnesium levels (Whittaker et al. 1954; Kunz 1985). High concentrations of heavy metals like chromium and nickel are also generally common in this rock type. The high proportion of endemic species associated with serpentine soils has generated much study ranging from the evolu- | tionary ecology of plant tolerance to the structure of plant communities found on serpentine (see ‘Brady et al. 2005 for a review). The gabbro soils ‘are considered to be edaphically similar to ‘serpentine because of their mineral composition ‘and because they appear to influence plant ‘distributions. However, gabbro-derived soils in ‘El Dorado County have a higher Ca/Mg ratio ‘(Goldhaber et al. 2009), and lower concentrations of chromium and nickel (Morrison et al. 2009) han are characteristic of serpentine soils. ' Changes in topography strongly affect the distribution of plants by providing micro-cli- 4 | mates significant to species survival (Mason 1946; Spurr and Barnes 1973; Mooney et al. 1974; Ricklefs 1976; Hocker 1979). In California chaparral, topographically-governed moisture and insolation levels may be reflected in patterns of shrub species distribution due to their affect on germination and seedling survival (Meetemeyer et al. 2001); hot, exposed sites tend to contain species with seeds cued to germinate after fire and seedlings that have high tolerance to drought, while sheltered slopes contain resprouting species with seeds that depend on cool, moist conditions for germination and subsequent growth. The topography of the Pine Hill complex is rich in its variety of slope and aspect varying from small flat valleys with gently rolling hills to steep river canyons and prominent peaks (though only a few of these extend above 600 meters in elevation) and thus, topography may play an important role in the diversity of the area’s flora and in the distribu- tion of the rare plants on the Pine Hill complex. The overall climate is relatively consistent over the entire region and is characteristic of California’s Mediterranean climate with warm, dry summers and cool, wet winters. The average annual precipitation, recorded nearby at Fol- som Lake, is 65 cm and occurs mostly in the form of rain in the winter months (USBR 1981). 248 In addition to being noted for its unique plant life. the Pine Hill region of El Dorado County was considered a desirable area for residential development. Easy freeway access from the city of Sacramento encouraged rapid and extensive development with much of the land being cleared for commercial and residential uses. By 1996 several plant species were federally listed under the U.S. Endangered Species Act as endangered, threatened, or of special concern (U.S. Fish and Wildlife Service 1996); others were listed as rare by the California Native Plant Society (Table 1) due to urbanization, habitat fragmentation, road construction, herbicide spraying, change in fire frequency, off-road vehicle use, unauthorized dumping, overgrazing, mining, and competition from invasive alien vegetation. Preserves to protect the rare species have been established throughout the Pine Hill area (for a brief history see Brink 2010). Of the 2,024 ha (5,001 acres) that are within the target recovery area’s boundaries, at least 325 ha (803 acres) have been lost due to development while 1,309 ha (3,234 acres) within the recovery boundary are protected within formal preserves (DeLacy, American River Con- servancy: Hinshaw, Bureau of Land Manage- ment, personal communications). The federal listing of five species has been effective in providing protection for large areas of a unique chaparral (‘Northern Gabbroic”’, Holland 1986) and has provided collateral protection for seven rare, but unlisted plant species (Pavlik 2003). Our goals in this study were to compile a flora for the Pine Hill region, classify the plant communities using Two-way Indicator Species Analysis (TWINSPAN), and to investigate the distributions of plant communities in relation to environmental factors using canonical correspon- dence analysis and permutation tests in CA- NOCO. We considered both biotic factors (vegetation cover, cover by exotic, native or rare species, vegetation height, etc.) and abiotic factors (slope, aspect, rock type, soil chemistry, disturbance, etc.). Further, we wished to specif- ically determine the community and plant asso- ciations, and environmental correlates of the rare and endangered plants. MATERIALS AND METHODS Study Area In order to evaluate the influence of gabbro soil on plant distribution, we extended the boundaries of the study area beyond the imme- diate Pine Hill area to include other soil types. We located 148 plots between the elevations of 120 and 670 meters and approximately between north latitude 38°38’ and 38°57’. Pine Hill, at an elevation of 628 meters (USGS 1973) is located near the center of the study area (at approxi- MADRONO [Vol. 56 mately 38°43’ north latitude and 120°59’ west longitude). Approximately 60% of the plots were on gabbro soil. The Floristic Study of Pine Hill and Vicinity Plant identification and taxonomy used in this work conform to the nomenclature of Hickman (1993). Existing specimens from the California State University, Sacramento herbarium (CSUSH) were used to confirm identifications. Plant specimens were collected between 1981 and 1985 during all seasons and placed in CSUSH. Whenever rare plant species were observed during explorations of the study area, their locations were mapped onto USGS 7.5 min quad maps and any unusual circumstances noted. Map locations were converted to UTM coordinates in 2008. Selected sites previously recorded by others were visited to confirm the presence of rare species, but the primary emphasis of this study was to find new rare plant locations. New locations were reported to the California Natural — Diversity Database (2008). Stratified Random Plot Study Aerial photographs (USGS 1979) were used to map the overall distribution of the basic vegeta- tion types and the fraction covered by each vegetation type was estimated using graph paper. Ground truthing verified photo interpretation. | The vegetation map served as a guide to locate | the stratified, random sample plots as well as a_ means of calculating coverage area for vegetation | types as they occurred upon the Pine Hill gabbro | formation in 1979. From these calculations, | chaparral was the most widespread vegetation | with a cover of 44.8%, followed by woodland at | 28.3% and grassland at 26.9%. The number of sample plots per vegetation | type was assigned in proportion to the relative | aerial coverage of each type. Since a comparison | was to be made between vegetation on gabbro - soil and that on non-gabbro soil, the number of) plots ‘‘on” and “off” the gabbro needed to be} relatively consistent within the percentages of | each vegetation type found on the gabbro: formation. Approximately 40% of the plots: assigned to each vegetation category were located on non-gabbro soil. Appropriate plot sizes were determined exper-. imentally using a nested plot technique and standard species area curve calculations for greater than 90% coverage (Mueller-Dombois. and Ellenberg 1974). This technique to determine plot size was used to insure that the samples. taken from each vegetation type would be comparable in species diversity. The actual plot sizes used for each vegetation category were as follows: for chaparral, 42 m? (3.25 m X 13 m); for 2009] TABLE 1. WILSON ET AL.: BIODIVERSITY OF PINE HILL 249 THE EIGHT RARE VASCULAR TAXA OF THE PINE HILL GABBRO COMPLEX, THEIR LISTING STATUS, PERCENT OF PLOTS WHERE FOUND, THE SOIL TYPES WHERE THEY GREW, AND THEIR FIRE REGENERATION STRATEGIES (F = FACULTATIVE SEEDER/RESPROUTER; R = OBLIGATE RESPROUTER; S = OBLIGATE SEEDER). 'Known from other soil types outside the Pine Hill area. ° Not found during this study, but reported to be present (California Department of Fish and Game, 1978; Aparicio 1978); the legitimacy of H. suffrutescens as a distinct taxon is controversial. Percent -of Fire Common plots where regeneration Taxon Federal status name found Soil type strategy Calystegia stebbinsii Brummitt endangered Stebbins’ O.7 gabbro' S (possibly F) morning- glory Ceanothus roderickti W. Knight endangered Pine Hill 6.5 gabbro 5 ceanothus Chlorogalum grandiflorum Hoover not listed Red Hills LO gabbro! R soaproot Fremontodendron californicum endangered Pine Hill 1.4 gabbro F (Torr.) Coville ssp. decumbens flannelbush (R. Lloyd) Munz Galium californicum Hook. & Arn. endangered El Dorado 5.0 gabbro R ssp. sierrae Dempster & Stebbins bedstraw Helianthemum suffrutescens Schreib. not listed Bisbee Peak 0.0 not found? 5? rush-rose Packera layneae (Greene) W.A. threatened Layne’s 4.3 gabbro, R Weber & A. Léve butterweed serp, meta Wyethia reticulata Greene species of El Dorado oe 2 gabbro R concern County mule ears woodland, 100 m* (5 m X 20 m); and for indicating minimum (SW = 1, S = 2, W = 3, grassland, 25 m* (2.5 m X 10 m). Rectangular SE = 4, NW = 5, E = 6, N = 7, NE = 8). plots were used as they yield more representative data than plots of other shapes (Mueller-Dom- bois and Ellenberg 1974). A total of 148 sample plots was established throughout the study area between July 1984 to February 1985; vegetation and floristic data were taken during spring and summer 1985. At the end of the study period, only 139 of these plots remained. Nine plots were lost due to road building or development. Specific plot locations were assigned using a stratified random sampling method. This method allowed the sampling of Specific areas, in between anthropogenically disturbed places, while retaining the advantages of random sampling. A random numbers chart was used to determine direction of travel, distance taken to reach a specific point, and to determine plot orientation. Specific study areas were chosen on the basis of observed environ- mental variation in the interest of including significant gradients for data analysis. Environmental data recorded at each plot location were slope, aspect, elevation, soil texture and rock types, soil calcium and magnesium, disturbance, available water, and vegetation cover. Specific slope and aspect measurements were determined using a pocket transit. To reflect the sun exposure, aspects were assigned numer- ical values on a gradient from 1| to 8 with 1 (SW) indicating maximum exposure, and 8 (NE) | Surface estimates of soil texture were made by rating the proportions of rock to clay and a numerical scale was constructed to indicate a gradient from extreme rocky outcrop (value of 1) to soils of mostly fine silt and clay (value of 4). Elevations were estimated at each plot location using topographic maps. Geology substrate maps and field identification of the rocks within each plot were used to determine the parent material of the soil. The U.S. Department of Agriculture’s soil surveys (Rogers 1974; USDA 1980) were used to check field observations on rock and soil parameters. The levels of calcium and magnesium in the soil were determined using the Model 14855 Soil Calcium and Magnesium Test Kit available from Hach Co., Loveland, Colorado. Note was taken of any evidence of disturbance due either to human activities, such as grazing or clearing, or natural events, such as fire. Distur- bances were recorded with regard to (1) the extent to which they affected the plants within the plot and (2) recentness of their occurrence. These two factors were rated. Ratings on recentness (time) were scaled with end points from 1 (long ago - little or no evidence remaining) to 7 (recent - within the last 2 yr). Extent of the disturbance was rated from 1 (disturbance area and type minimal) to 4 (major disturbance, all plants destroyed). The two factors were multiplied by each other to obtain a value for each _ plot. 250 Observable surface water was estimated using a scale as follows: 1 = always dry, no water nearby; 3 = near seasonal water supply, mostly dry; 5 = near a permanent source of water, stream or lake; and 7 = water within plot most of the year. Differences in cover were estimated on the basis of the total amount of plant cover present in the three structural levels of trees, shrubs and groundcover (herbs and grasses). The method used for estimating cover was a modification of methods described in the literature (Daubenmire 1974; Mueller-Dombois and Ellenberg 1974). The cover values used in this study were: 8 = 95.1 to 100% cover; 7 = 75.1 to 95% cover; 6 = 50.1 to 75% cover; 5 = 25.1 to 50% cover; 4 = 10.1 to 25% cover; 3 = 5.1 to 10% cover; 2 = 1.0 to 5% cover; | = <1% cover. In addition to the measurable data gathered for each plot, other factors were included. “Latitude” values for each plot were assigned as the distance in miles north from the southern- most plot location in the study. We noted the number of rare species found within each sample plot. The soil survey for El Dorado Co. (Rogers 1974) rates the suitability of various sites for general farming using the Storie Index rating which takes into account soil profile, texture, slope and other conditions such as drainage. High ratings imply few restrictions to agricultural plants while lower ratings indicate increased limitations to farming. Since the Storie Index is a calculation indicating a soil and plant growth relationship, it was included in the analysis. Depth to bedrock was also noted from the soil survey (Rogers 1974). Table 2 lists the physical, descriptive, and vegetation variables considered in the study. The Shannon diversity index (H’) was com- puted for each sample and used as a measure of alpha diversity or the species diversity within samples (Krebs 1999). The Shannon H evenness index (evenness = H’/log(N)) was used as a measure of how equitable and homogeneous species diversity was among samples. Equitability assumes a value between 0 and 1 with | being complete evenness. Diversity and evenness were compared for each rock formation and vegeta- tion category. Data Analyses Two-way Indicator Species Analysis (TWIN- SPAN) (Hill 1979) is a classification program which organizes plot samples into community groups on the basis of species composition (identity and cover) using a divisive clustering algorithm. Plots with similar associations are grouped together by TWINSPAN and _ the program organizes species on the basis of their affinities for these groups into plant associations. We analyzed our data using a FORTRAN MADRONO [Vol. 56 version of TWINSPAN and that ran on a main-frame computer (Alcor) at the University of California, Davis in 1985. Canonical Correspondence Analysis (CCA here after) is a constrained ordination technique that finds axes of the greatest variability in community composition for a set of samples (ter Braak 1986; ter Braak and Smilauer 2002). Community composition is defined by the number, identity, and abundance of species. CCA uses weighted averaging to search for the best explanatory variables where species abun- dances are the weights. Assuming the species have unimodal responses for the explanatory variables, weighted averaging is the simplest way to find the species optima (1.e., species scores) for those variables. A preliminary detrended corre- spondence analysis (DCA) by segments was used to assess segment length of gradients using CANOCO for Windows (Hajek et al. 2002; ter Braak and Smilauer 2002). The DCA showed that gradients were 5.20 standard deviations long and thus were conducive to unimodal methods such as CCA (Leps and Smilauer 2003). As well, data diagnostics were performed to access the assumption of unimodal response of the species data to the explanatory variables. The CCA program CANOCO (Leps and Smilauer 2003) was used to arrange all plant species along the measured environmental gradi- ents. The quantitative and nominal environmen- tal variables we used are listed in Table 2. Species cover class values were backtransformed to percent cover using the midpoint value of the cover class and then were log transformed (plus a constant of one) because of the many zero values in order to remedy the positive skew in frequency distribution of species cover. Species with low overall cover were downweighted in the analysis to reduce the undue influence of these rarer species on the CCA (Fig. 2). This influence occurred because many of the low cover species co-occurred in samples with a few more common species (ter Braak and Smilauer 2002). All measured and computed (e.g., Shannon H)- environmental variables were subjected to Monte | Carlo permutation tests in CCA to provide p-_ values to assess the marginal significance of each | variable individually. The conditional effect of’ each variable was assessed as each was added to a_ model during forward selection to explain total | variation in community structure. During this: process, multicolinearity was detected among. several of the variables causing a slight arch) effect in the CCA biplots. A correlation matrix. was generated and sorted using the CORR procedure in SAS software for all environmental variables in order to identify redundant environ- ' mental variables (SAS Institute Inc. 2004). Any} pairwise correlation exceeding 0.60 resulted in the’ selection of the most objective and ecologically . 2009] TABLE 2. WILSON ET AL.: BIODIVERSITY OF PINE HILL 251 VARIABLES SELECTED BY FORWARD SELECTION AND TESTED BY MONTE CARLO PERMUTATION. The variable codes were used in the CCA biplots. The marginal effect (A) for each variable 1s a measure of the variance each explains when that particular variable is the only environmental variable used. The variables were categorized as abiotic or biotic for variance partitioning (see text for details). ' Designated as nominal variables, all others are quantitative. * Variables not selected by MonteCarlo simulation. Variables Code Ay Definition or how measured: ABIOTIC Aspect Aspt 0.13 measured with a Brunton pocket transit Bedrock Bdrk 0.11 depth to bedrock Ca/Mg? Ca/Mg chemically tested soil values in situ Disturbance Dist O25 numerical assessment of degree and recency of disturbance Elevation Elev 0.13 estimated from 7.5 min. topographic maps Latitude Lati 0.11 distance north from southernmost plot in miles Gabbro' Gabb 0.12 nominal variable designates gabbro rock formation sites Serpentine! Serp 0.17 nominal variable designates serpentine rock formation sites Granite! Gran 0.10 nominal variable designates granite rock formation sites Metamorphic’ Meta 0.09 nominal variable designates metamorphic rock formation sites Slope Slpe 0.17 measured with a Brunton pocket transit Soil Ca CA 0.09 chemically tested soil Ca in situ Storie index Stor 0.30 index of agricultural suitability Surface Text 0.20 soil texture field estimate Water H20 0.17 availability of surface water in or near plot BIOTIC Cover Covr 0.19 percent of plot area covered by all plants estimated visually Diversity Dive 0.28 #families/#species Evenness Even 0.16 calculated as H’/In(Exot + NatS) Groundcover GrCov 0.31 percent of plot covered by forbs estimated visually Height? Height estimate of overall plant height Exotic species Exot 0.40 number of introduced species Native species NatS 0.38 number of native species Aree cover TrCov 0.46 percent of plot covered by trees estimated visually Rare species Rare 0.20 number of rare species Shade?’ Shade estimate of coverage at 5 dm height Shannon H’ Shan 0.23 calculated as H’ (Krebs 1999) Shrub cover ShCov 0.58 percent of plot covered by shrubs estimated visually Unique species Uniq 0.13 species found in only a single plot Chaparral! Chap 0.48 nominal variable designates chaparral sites Grasslands' Gras 0.51 nominal variable designates grassland sites Woodland! Wood 0.45 nominal variable designates woodland sites meaningful variable of the pair, and elimination of the other correlated variable with the exception of two pairs of important explanatory variables that had correlations exceeding 0.8: ground cover was correlated with grasslands, a nominal site variable, and tree cover was correlated with ““Woodland’’, also a nominal site variable (both correlations >0.80). The remaining explanatory variables were subjected to another forward selection and Monte Carlo permutation to remove those variables that did not explain significant portions of the overall variance singly without the influence of any other variable. These _ variables were highly unlikely to contribute to an overall explanatory model of species variability among the sites. Multicollinearity was not detected in subsequent CANOCO analyses with _ the final set of environmental variables. In a final CCA analysis, significant variables were identi- fied and their conditional P-values estimated by Monte Carlo permuation. oD) £ = D o s < O O Species rank (based on % cover) FIG. 2. Down-weighting scheme used for CCA, where a weight of 1 means the species carries its original influence on the ordination and lower weights reduce less frequently occurring species undue influence on the analysis (see ter Braak and Smilauer 2002). Species were arranged on the X-axis from most frequent on the left to least frequent on the right. 252 The final CCA diagram of species scores with biplot scaling, and biplot scores of the quantita- tive variables and centroid scores of the nominal variables were interpreted for community struc- ture (Leps and Smilauer 2003). Multiple CCAs were run to partition the total variance into separate ‘biotic’ (B) factors (plant cover, species numbers, etc.) and ‘abiotic’ (A) factors (soil calcium, soil type, water availability, etc.; Ta- ble 2) (Legendre 2007). We did this to see how much of community composition was determined by site characteristics such as resource availability (A), by plant-plant interactions (B) and how much was shared between these two categories (C). We estimated the A, B, and C fractions using five partial constrained ordinations. From these five analyses we were able to decompose the total variance in the species data set into abiotic, biotic, and shared sources of explained variance. Variance decomposition was performed where the two spatially explicit variables, longitude and latitude from UTM data were partitioned from the remaining environmental variables (Legendre et al. 2005). This decomposition was done to assess for differences in spatial (beta) diversity. RESULTS Floristic Content of the Study Area Over one thousand plant specimens were collected on numerous trips to the region. The final list of plants from the entire study area, on and off gabbro, is a composite of species identified by various individuals working in the area (Appendix 1). The list includes 741 distinct taxa (including 91 subspecies or variaties, 8 species of ferns, and 3 species of mosses) in 376 genera, representing 91 families. The families with the most taxa were Asteraceae (108 species and subspecies), Poaceae (71), and Fabaceae (58). During the plot study, 342 species and varieties were identified within the plot borders (Appendix 1). The taxa found in the plots belonged to 216 genera that occurred within 66 vascular plant families; 267 (approximately 78%) were Califor- nia natives. The mean number of plants found in each plot was 24, and the mean percentage of California native taxa occurring throughout all plots was 64.1% (Table 3). We found 219 species in ““Woodland” areas, of which 76% were native species. One ‘“‘Woodland”’ plot, 100 m?° in size, contained 61 species of plants. The chaparral contained 190 species, of which 76% were native. Within the ““Woodland” areas, serpentine and gabbro had the highest levels of natives at 96% and 81%, respectively. On the other hand, only 36% of the 149 species found in grassland were native species according to Hickman (1993). Serpentine grasslands, however, had a greater proportion of native species (66%) than non- MADRONO [Vol. 56 TABLE 3. NUMERICAL DISTRIBUTIONS OF THE STUDY AREA’S FLORA AS SAMPLED BY THE STRATIFIED RANDOM PLOT STUDY. Number in parentheses is the number of plots in the category. Categories Values A. Overall taxa distribution Number of taxa All plots (139) 342 Gabbro plots (80) 253 Serpentine plots (17) 141 Metamorphic and granite 225 plots (42) Grassland plots (38) 149 Woodland plots (38) 219 Chaparral plots (63) 190 B. Gabbro soils only Number of taxa Grassland gabbro plots (22) 85 Woodland gabbro plots (22) 145 Chaparral gabbro plots (36) 150 C. Species densities Mean taxa per plot All plots 24 Chaparral plots 21 Grassland plots 20 Woodland plots 35 Gabbro plots 26 Metamorphic plots 23 Granite plots 20 Serpentine plots 24 D. Percent native taxa Mean percent per plot All plots 64.1 Chaparral plots eee Grassland plots 36.1 Woodland plots 76.3 Gabbro plots 64.0 Metamorphic plots 62.5 Granite plots 372 Serpentine plots 76.4 serpentine grasslands. The 100 most frequently encountered species in the plots, which included rare species Calystegia stebbinsii Brummitt, Cea- nothus roderickii W. Knight, Chlorogalum grand- iflorum Hoover, and Wyethia reticulata Greene along with the three other listed species (Fremon- todendron californicum (Torr.) Coville ssp. de- | cumbens (R. Lloyd) Munz, Galium californicum | Hook. & Arn. ssp. sierrae Dempster & Stebbins, | and Packera layneae (Greene) W.A. Weber and. A. Léve) are listed in Table 4 with their 4-letter | species codes. The low-growing native herb Galium porrigens Dempster was the most common species found (Table 4) and grew in over 80% of the ““Wood-_ land” and shrub plots, but was never found in grasslands (Table 5) while the exotic grasses Aira caryophyllea L. and Bromus madritensis L. occurred in about 80 plots and were found in. all three community types (Tables 4 and 5). The. shrubs with the highest frequency and cover were native species Arctostaphylos viscida Parry and 2009] Adenostoma fasciculatum Hook. & Arn. (Table 4) which were found in ca. 75% of the “‘tall, closed-canopy chaparral’ and ‘‘Woodland”’ plots, and on all soil types, but were never found in grassland (Table 5). The tree with the highest frequency and cover was the native oak Quercus wislizenii A. DC., a dominant species of ““Woodland”’. It was frequently found in shrub plots and was also found in a few grassland plots. Classification Based on their floristic composition, the 138 plots were classified by TWINSPAN into three main communities: ““Woodland”’, Shrub, and “Grassland”. Table 5 lists the classification of the 100 most common species although the analysis was run using all 347 species and varieties. ““Woodland”’-type communities were generally found on non-serpentine soils. Within the ““Woodland” community types, TWINSPAN further delimited ““Blue Oak Savanna’, a com- munity dominated by Quercus douglasii Hook. & Arn. and mostly-native forbs; ““Woodland”’, a native-species rich community characterized by high diversity of trees, including the oaks Quercus wislizenii and Q. kelloggii Newb. and Ponderosa pine (Pinus ponderosa C. Lawson), vines includ- ing native honeysuckles (Lonicera spp.) and abundant poison-oak (Toxicodendron diversiloba (Torr. & A. Gray) Greene), and native grasses, forbs, and bulbs; and a “Chaparral-Woodland” transitional community type that was character- ized by the shrubs toyon (Heteromeles arbutifolia (Lindl.) M. Roem.), redbud (Cercis occidentalis Torr.) and coffee berry (Rhamnus tomentella Benth. ssp. crassifolia (Jeps.) J. O. Sawyer), and Foothill pine (Pinus sabiniana Douglas) — species which also grew on serpentine soils. Wyethia reticulata, a species of concern, was included in the main ““Woodland” group, near the ““Chapar- ral-Woodland” transition group. The Shrub-dominated communities were found on all soil types including serpentine. Shrub communities were divided into ‘‘Short- Chaparral” dominated by native low-growing shrubs, forbs, and grasses - a high proportion of which were found growing on serpentine soils; “Tall, Closed-Canopy Chaparral” dominated by the shrubs Adenostoma fasciculatum Hook. & Arn. (chamise), Arctostaphylos viscida Parry (manzanita), the low growing Salvia sonomensis Greene, and the rare species Ceanothus roderickii and Chlorogalum grandiflorum.; and openings in chaparral, ““Open Chaparral, where the exotic grasses Vulpia myuros (L.) C. C. Gmel. and Aira caryophyllea L. were commonly found. Both of these grasses had high occurrence in all three main community types. WILSON ET AL.: BIODIVERSITY OF PINE HILL 203 In the “Grassland” community type, 80% of the most frequently encountered species were exotic. ““Grasslands’” were dominated by exotic annual grasses, especially the brome grasses (Bromus spp.), oats (Avena spp.), and exotic forbs, especially Hypochoeris spp. and Erodium spp. Results of CCA of the Pine Hill Vegetation Shrub and Tree Cover (quantitative variables) and community classifications (nominal vari- ables) explained the highest amount of variance in the CCA when we evaluated the marginal significance of each variable individually (Ta- ble 2). Serpentine was the only soil type that explained much variation (Table 2). The condi- tional effect of each variable was assessed as each was added to a model during forward selection to explain total variation in community structure (Table 6). The final model that resulted from forward selection found the Shrub and Tree cover variables to have the highest conditional effects (Aq = 0.58 and 0.45, respectively) and thus were the first variables to be included in the multivar- late model (Table 6). The Serpentine variable was the only abiotic variable (and only rock forma- tion) found to have a moderately high condition- al (Aaj = 0.15) effect relative to the biotic variables, followed by elevation, surface texture, and degree of disturbance (A, = 0.09, 0.08, 0.08, respectively). The first two axis of the CCA biplot depicted three main clusters around variables that gener- ally describe communities dominated by grass- land, chaparral, and woodland species (Fig. 3). There was a smaller cluster of species scores situated between the ““Woodland” and chaparral clusters. The tree, shrub and exotic species variables had the longest arrows in the CCA biplot, and were therefore most strongly related to community structure. The first CCA axis (x- axis) was dominated by information contained in exotic species numbers to the right (TExosp -ccAt = 0.73) and shrub cover to the left (tTspeoy-ccal = —0.91) (Fig. 3), and separated the open grass- lands and blue oak and valley oak savannas from shrub and tree dominated woodlands and shrub- lands. The shrub species were most often native shrub species (1% sncov-Nats = 0.74). The second CCA axis (y-axis) was dominated by tree cover (tT+Cov-CCA2 = —90.91) and ‘“‘Woodland”’ sites (tTWood-CCA2 = ~—90.89) in one direction, and chaparral sites (tChap-ccA2 = 9.63) in the other direction, and separated the chaparral from “Woodland”. The proximity of species in the CCA biplot was indicative of their co-occurrence in the samples and aggregations of species were sorted into communities (Table 7). The species with the highest cover observed in this study, Adenostoma fasciculatum (ADFA) is most closely associated 254 MADRONO [Vol. 56 TABLE 4. STUDY CHARACTERISTICS OF THE ONE HUNDRED MOST FREQUENT TAXA IN THE STUDY, WHICH INCLUDED RARE TAXA CERO, CHGR, AND WYRE, PLUS FOUR OF THE RARE TAXA, CAST, FRCA, GACA, AND PALA. Taxa are listed by their four-letter codes. Rare species are denoted with an asterisk. Taxon Number of Average code Taxon Family plots cover (%) ACMI — Achillea millefolium L. Asteraceae ADFA ~— Adenostoma fasciculatum Hook. & Arn. Rosaceae 60 16.8 AECA — Aesculus californica (Spach) Nutt. Hippocastanaceae AETR — Aegilops triuncialis L. Poaceae AICA Aira caryophyllea L. Poaceae 80 | es) ARVI Arctostaphylos viscida Parry Ericaceae a2 13.6 AVBA ~— Avena barbata Link.. Poaceae 43 2.4 AVFA — Avena fatua L. Poaceae BAPI Baccharis pilularis DC. ssp. consanguinea (DC.) C.B. Asteraceae Wolf BRDI Bromus diandrus Roth Poaceae oi) Sal BRDS — Brachypodium distachyon (L.) P. Beauv. Poaceae BREL ~ Brodiaea elegans Hoover Liliaceae BRHO ~— Bromus hordeaceus L. Poaceae 56 8.7 BRLA — Bromus laevipes Shear Poaceae BRMA — Bromus madritensis L. Poaceae a ee) BRMI ~— Briza minor L. Poaceae BRST Bromus sterilis L. Poaceae CAAL ~— Briza minor L. Liliaceae 51 0.3 CABR — Carex brainerdii Mack. Cyperaceae CAOL = Cardamine oligosperma Torr. & A.Gray Brassicaceae *CAST Calystegia stebbinsii Brummit Convolvulaceae CECU Ceanothus cuneatus (Hook.) Nutt. Rhamnaceae CEGL ~~ Cerastium glomeratum Thuill. Caryophyllaceae CELE Ceanothus lemmonii Parry Rhamnaceae CEOC — Cercis occidentalis Torr. Fabaceae CEPA Ceanothus palmeri Trel. Rhamnaceae *CERO Cenothus roderickti W. Knight Rhamnaceae *CHGR_ Chlorogalum grandiflorum Hoover Liliaceae CHPO Chlorogalum pomeridianum (DC.) Kunth Liliaceae CLLA Clematis lasiantha Nutt. Ranunculaceae CLPE Claytonia perfoliata Donn ex Willd. Portulacaceae CYEC = Cynosurus echinatus L. Poaceae DICA Dichelostemma capitatum Alph. Wood Liliaceae DIMU — Dichelostemma multilflorum (Benth.) A. A. Heller Liliaceae DIVO Dichelostemma volubile (Kellogg) A. A.Heller Liliaceae ELGL — Elymus glaucus Buckley ssp. jepsonii (Burtt Davey) Poaceae Gould ELMU = Elymus multisetus (J.G. Smith) Burtt Davy Poaceae ERCA — Eriodictyon californicum (Hook. & Arn.) Torr. Hydrophyllaceae ERCI Erodium cicutarium (L.) L’>Her. ex Aiton Geraniaceae ERLA — Eriophyllum lanatum (Pursh) Forbes var. grandiflorum Asteraceae (A. Gray) Jeps. ERBR ~— Erodium brachycarpum (Godr.) Thell. Geraniaceae FIGA Filago californica Nutt. Asteraceae *FRCA = Fremontodendron californicum (Torr. Coville) ssp. Sterculiaceae decumbens (R. Lloyd) Munz GAAP — Galium aparine L. Rubiaceae 48 12 *GACA_ ) Galium californicum Hook. & Arn. ssp. sierrae Rubiaceae Dempster & Stebbins GADI Galium divaricatum Pourr. ex Lam. Rubiaceae GAPO — Galium porrigens Dempster Rubiaceae 82 0.8 GAVE © Gastridium ventricosum (Gouan) Schinz & Thell. Poaceae GEDI Geranium dissectum L. Geraniaceae GEMO © Geranium molle L. Geraniaceae HEAR = Heteromeles arbutifolia (Lindl.) M. Roem. Rosaceae 66 6.4 HEMI — Hesperolinon micranthum (A. Gray) Small Linaceae HOMU = Hordeum murinum L. ssp. leporinum (Link) Arcang. | Poaceae HOVI Holocarpha virgata (A. Gray) D.D. Keck Asteraceae HYGL = Aypochaeris glabra L. Asteraceae 41 0.5 HYRA = Hypochaeris radicata L. Asteraceae 2009] WILSON ET AL.: BIODIVERSITY OF PINE HILL 255 TABLE 4. CONTINUED. Taxon Number of Average code Taxon Family plots cover (%) IRMA — Iris macrosiphon Torr. Iridaceae LASU ~— Lathyrus sulphureus A. Gray Fabaceae LECA ~— Lepechinia calycina (Benth.) Epling ex Munz Lamiaceae LOHI Lonicera hispidula (Lind|.) Dougl. ex Torr. & Caprifoliaceae A. Gray var. vacillans A. Gray LOIN Lonicera interrupta Benth. Caprifoliaceae LOMI ~— Lotus micranthus Benth. Fabaceae LOMU © Lolium multiflorum L. Poaceae LUBI Lupinus bicolor Lindl. Fabaceae LUCO — Luzula comosa E. Mey. Juncaceae MAEX Madia exigua (Sm.) A. Gray Asteraceae MAGR_ Madia gracilis (Sm.) D. D. Keck Asteraceae MECA Melica californica Scribn. Poaceae METO~ Melica torreyana Scribn. Poaceae MICA —= Micropus californicus Fisch. & C.A. Mey. Asteraceae MOVI = Monardella villosa Benth.ssp. villosa Lamiaceae *PALA = Packera layneae (Greene) W.A. Weber & A. Léve Asteraceae PETR Pentagramma triangularis (Kaulf.) Yatsk., Pteridaceae Windham & e. Wollenw. PIPO Pinus ponderosa C. Lawson Pinaceae PISA Pinus sabiniana Douglas Pinaceae PLER Plantago erecta Morris Plantaginaceae POCO — Polygala cornuta Kellogg Polygalaceae QUCH = Quercus chrysolepis Leibm. Fagaceae QUDM Quercus dumosa Nutt. Fagaceae QUDO = Quercus douglasii Hook. & Arn. Fagaceae QUKE = Quercus kelloggii Newb. Fagaceae QUWI = Quercus wislizenii A. DC. Fagaceae 40 9.9 RAOC — Ranunculus occidentalis Nutt. var. eisenii (Kellogg) Ranunculaceae A. Gray RHIL Rhamunus ilicifolia Kellogg Rhamnaceae 38 0.7 RHTO —= Rhamnus tomentella Benth. ssp. crassifolia (Jeps.) Rhamnaceae J.O. Sawyer SABI Sanicula bipinnata Hook. & Arn. Apiaceae 46 0.6 SACR = Sanicula crassicaulis Poepp. ex DC. Apiaceae 40 0.7 SASO Salvia sonomensis Greene Lamiuaceae 40 6.8 SIMA Sidalcea malvaeflora (DC.) A. Gray ex. Benth. ssp. Malvaceae asprella (Greene) C.L. Hitchce. SEAR — Senecio aronicoides DC. Asteraceae STME — Stellaria media (L.) Vill. Caryophyllaceae TACA — Taeniatherum caput-medusae (L.) Nevski Poaceae TOAR ~ Torilis arvensis (Huds.) Link Apiaceae 52 3.8 TODI Toxicodendron diversilobum (Torr. & A. Gray) Greene Anacardiaceae 53 4.7 TRDU — Trifolium dubium Sibth. Fabaceae TRMI Trifolium microcephalum Pursh. Fabaceae TRPR ~~ Trifolium pratense L. Fabaceae TRWI Trifolium willdenovii Spreng. Fabaceae VINI Vicia sativa L. ssp. nigra L. Fabaceae VIVA Vicia villosa Roth ssp. varia (Host) Corb. Fabaceae VUHI Vulpia myuros (L.) C.C. Gmel. var. hirsute Hack. Poaceae VUMY Vulpia myuros (L.) C.C. Gmel. Poaceae 72 2.4 VUPA Vulpia microstachys (Nutt.) Munro var. pauciflora Poaceae (Scribn. ex. Beal) Lonard & Gould *WYRE Wryethia reticulata Greene Asteraceae with the cluster “Chaparral 1”’ (Yellow group in Figure 3; Table 7). Additional shrub species in “Chaparral 1” include Arctostaphylos viscida (ARVI), Ceanothus lemmonii Parry (CELE), and Quercus dumosa Nutt. (QUDM) and low- growing Salvia sonomensis (SASO). Four rare species were most abundant in the ““Chaparral 1” cluster and closely associated with each other; Calystegia stebbinsii (CAST), Ceanothus roder- ickiti (CERO), Chlorogalum grandiflora (CHGR), 256 MADRONO [Vol. 56 TABLE 5. SUMMARY OF TWINSPAN CLASSIFICATIONS THE 100 MOST FREQUENTLY ENCOUNTERED PLANT TAXA. The listing is arranged into the three main TWINSPAN community types (WOODLAND, SHRUB, GRASSLAND). The three right-hand columns contain the frequency (percentage of plots) of taxa found in shrub- dominated, tree-dominated and open, grassland type plots; frequencies in bold text are plants characteristic of the main community type. Within the main types, TWINSPAN community sub-types are delimited. Species abbreviations as in Table 4. Underlined species are those that occurred in ca. 10% or more plots in each main community type. Plants with an asterisk are rare species among the top 100 species. Plant names followed by ‘““Serp”’ were found in serpentine plots. Plant taxa Native/ Introduced Life form Shrub plots Woodland plots Grassland plots WOODLAND Blue Oak Savanna: MAGR Serp N forb 1.6 31.6 5.3 TRWI Serp N forb Io 15.8 7.9 QUDO Serp N tree 3.2 55.3 PAA | GAAP Serp I forb fie $4.2 18.4 CLPE Serp N forb 32 31.6 0:5 RAOC N forb 34.2 2.6 CAOL Serp N forb 4.8 23.7 270 CYEC Serp I grass 4.8 78.9 10.5 SACR Serp N forb 11.1 78.9 as) TOAR Serp I forb [27 92.1 18.4 Woodland AECA N tree 1.6 26.3 CLLA Serp N vine 16 15.8 QUCH Serp N tree 19 QUWI Serp N tree 19.0 71.1 2.6 DIVO Serp N bulb ee 57.9 2.6 BRLA N grass 52.6 ELGL N grass 68.4 2,0 LASU N forb 2.6 Che, LOHI N vine 39.5 LOIN Serp N vine a2 36.8 LUCO N forb 68.4 SIMA N forb 1.6 36.8 ACMI N forb 39:5 IRMA Serp N forb Be 44.7 PIPO N ice ZSit QUKE N tree 52.6 TODI Serp N vine 23.8 92.1 MECA Serp N grass I2ee 47.4 * WYRE N forb 4.8 18.4 CABR N forb 6.3 15.8 MOVI N forb 19 28.9 Chaparral-Woodland Transition CAAL Serp N bulb 47.6 Tet CEOC N tree oD PRE | HEAR Serp N shrub 44.4 71.1 RHTO Serp N shrub 23:8 31.6 BRDS I grass 19 10.5 PISA Serp N tree 23:8 26.3 PETR Serp N fern 19.0 28.9 SHRUB Short Chaparral BRMA Serp I grass 69.8 65.8 13.2 SABI Serp N forb 39.7 36.8 [S38 BAPI N shrub 12.7 152s 26 VUPA Serp N grass 19.0 13.2 210 CEPA Serp N shrub 4.8 2.6 CHPO Serp N bulb 15.9 26.3 GAPO Serp N forb 81.0 84.2 METO Serp N grass oD 18.4 RHIL Serp N shrub 30.2 44.7 ELMU Serp N grass 17.5 ioe2 ERLA Serp N forb 20.6 15.8 2009] WILSON ET AL.: BIODIVERSITY OF PINE HILL 257 TABLE 5. CONTINUED. Plant taxa Native/ Introduced Life form Shrub plots Woodland plots Grassland plots POCO Serp N shrub 23.8 28.9 SEAR N forb 9.5 18.4 Tall Closed-Canopy Chaparral CECU Serp N shrub 7.9 5:3 ADFA Serp N shrub 76.2 18.4 ARVI Serp N shrub 74.6 35:3 FIGA Serp I forb 14.3 26 GAVE Serp I grass 54.0 2.6 LECA N shrub 27.0 fi) MAEX Serp N forb 30.2 2.6 QUDM Serp N tree 15.9 vo DIMU Serp N bulb 23.8 525 = Pe: CELE N shrub 25.4 10.5 *CERO N shrub 12.7 2.6 * CHGR N bulb Dhue ERCA N shrub 20.6 HEMI Serp N forb 46.0 SASO N forb 58.7 53 Open Chaparral VUMY Serp I grass 66.7 34.2 44.7 GADI I forb 19.0 = Pe) AICA Serp I grass 77.8 50.0 2.9 MICA Serp N forb 14.3 26 PLER Serp N forb 12.7 19 GRASSLAND LOMI Serp N forb pe = Fe 18.4 DICA Serp N bulb 20.6 hO:5 31.6 HYGL Serp I forb 25.4 25 50.0 HYRA Serp I forb Pia 72) Va | VINI I forb 10.5 42.1 BRHO Serp I grass 13.9 2a. 7 97.4 ERCI | forb 1.6 39.5 HOMU Serp I grass 26.3 LUBI Serp N forb 50.0 TRDU Serp I forb 1.6 44.7 TRPR Serp I forb 16 me) molt AETR Serp I grass 1.6 2.6 44.7 AVEA Serp I grass SZ 52.6 BRDI I grass 263 68.4 ERBR Serp I forb 1.6 $1.6 GEDI Serp I forb Ope. 53,5 LUMU Serp I grass ons 39.5 BREL N bulb 4.8 18.4 60.5 TACA I grass 2.6 50.0 VUHI Serp I grass 6.3 2:0 13.2 HOVI N forb 18.4 VIVA I forb 2.6 18.4 BRMI I grass 126 18.4 34.2 CEGL I forb a2 28.9 34.2 GEMO I forb 16 2h 31.6 AVBA Serp I grass 22a 29 36.8 BRST Serp I grass 316 21.1 STME I forb ea 28.9 TRMI N forb oS 2a 23:1 and Fremontodendron californicum ssp decumbens and CA-poor (“CA”) but moderately deep (FRCA) (Table 8). “Chaparral 1” was found on = (““Bdrk’’). southerly facing slopes (““Aspt’’) and was associ- A second high diversity shrub-dominated ated with soils derived from serpentine and cluster — “Chaparral 2” (Blue group Fig. 3; gabbro that were rocky (“Text”), dry (““H20’), Table 7) — was located in CCA space between 255 MADRONO TABLE 6. RESULTS OF FORWARD SELECTION IN ORDER OF VARIABLE INCLUSION INTO THE FINAL MODEL. The conditional effects (Aa) are the additional variance explained by that variable upon its inclusion into the model. All variables contributed significantly to the model (P-value < 0.05). Variable codes follow Table 2. Variable A P-value F ratio ShCov 0.58 0.002 14.00 TrCov 0.45 0.002 11.48 NatS 0.19 0.002 5.06 Serp 0.15 0.002 Ao? Chap 0.14 0.002 3.69 Uniq 0.12 0.002 3.42 Elev 0.09 0.002 2.66 Text 0.08 0.002 226 Dist 0.08 0.002 2.29 Lati 0.07 0.002 2.16 Dive 0.07 0.002 212 Rare 0.07 0.002 2.10 Stor 0.07 0.002 2.00 GrCov 0.06 0.002 1.96 Gabb 0.06 0.002 1.83 Bdrk 0.06 0.002 1.75 Exot 0.05 0.002 1.74 Wood 0.06 0.004 L772 Slpe 0.05 0.004 1.66 Cov 0.05 0.004 1.67 CA 0.05 0.002 1.60 Shan 0.05 0.026 1333 Even 0.04 0.018 1.40 Gran 0.05 0.018 1.62 Aspt GACA | | \ geet oe [Vol. 56 Adenostoma-Arctostaphylos ‘‘Chaparral 1” and the ““Woodland”’ communities. Species that char- acterize “Chaparral 2”’ were shrubs Heteromeles arbutifolia (HEAR), Cercis occidentalis (CEOC), and Rhamnus tomentella ssp. crassifolia (RHTO); the sedge Carex brainerdii Mackensie (CABR), and two rare species, Packera layneae (PALA) and Wyethia reticulata (WYRE). Foothill Pine, Pinus sabiniana (PISA) was placed between “Chaparral 2” and “Woodland” (Fig. 3). Envi- ronmental variables associated with ‘‘Chaparral 2” included steeper slopes than “‘Chaparral 1”’, but with more moderate (non-southerly) aspect and higher water availability. This group was strongly associated with other native species, and negative- ly associated with disturbance, exotic species, and the ““Grassland”’ cluster. Both shrub clusters were associated with higher numbers of families per species (“Dive”) than the ‘‘Grassland”’ cluster which was dominated by species in Poaceae. The ““Woodland” cluster (Red group in Fig. 3; Table 7) was associated with north facing slopes, the presence of water, and shallow, metamorphic- or granite-derived soils with high calcium and few surface rocks. Not surprisingly, 1t was associated with high tree cover (““TrCov’”), and high total cover (“Covr’’). This cluster was associated with high species diversity (“Shan’’), especially native species (““NatS’’), and species diversity was homo- geneous among plots (““Even’’). ““Woodland” was negatively associated with disturbance (“Dist”’). AV, VIV. __Gras Al \ oUDO A! GAAP - TOAR SQULO -1.0 FIG. 3. 1.0 First two canonical correspondence analysis (CCA) axes depicting biplot scores of the 50 most abundant species (hollow triangles), quantitative (arrows), and nominal (filled triangles) environmental variables. The four clusters of species associations are those corresponding to sites with many introduced species (green), to woodland sites (red), to chaparral type 1 sites (yellow), and to chaparral type 2 sites (blue). See Table 4 for species abbreviations list and Table 2 for list of factors and their abbreviations. 2009] WILSON ET AL.: BIODIVERSITY OF PINE HILL 259 TABLE 7. THE FOUR CLUSTERS OF SPECIES ASSOCIATIONS BASED ON CHI SQUARE DISTANCES FROM THE CCA ON 104 TAXA (COLORS AS IN FIG. 3). Taxa abbreviations as in Table 4. The rare taxa are in bold text. The lower case letters following taxon abbreviations in Chaparral 1 and 2 refer to fire regeneration mechanisms: f = facultative seeder, r = obligate resprouter, r? = potential to resprout suggested by underground perennating structures, s = obligate seeder. (Anderson 1991; Keeley 1991; Hickman 1993; Franklin et al. 2004; personal observation). Cluster 1 grassland Cluster 2 Cluster 3 chaparral 1— xeric Cluster 4 chaparral 2 — (green) woodland (red) seeders (yellow) mesic resprouters (blue) AETR HYRA ACMI LUCO ADFA-f CHGR-r HEMI CAAL-r AVBA LOMI AECA MAGR AICA CHPO-r LECA-s CABR-r? AVFA LOMU BRLA PEAZ ARVI-s DIMU-r MAEX CAOL-r? BRDI LUBI BRST PIPO BAPI-r ELMU MICA CEOC-r&s BREL SIMA CLPE PISA BRDS ERCA-r POCO CLLA BRHO TACA CYEC QUCH BRMA ERLA QUDM-r HEAR-r BRMI TRDU CYGR QUDO CAST-s FIGA RHIL-r MECA-r? CEGL TRMI DIVO QUKE CECU-s FRCA-f SABI METO DICA TRPR BLGL QUWI CELE-s GADI SASO-f MOVI ERCI TRWI GAAP RAOC CEPA GAPO VUMY PALA-r GEDI VIHI GACA SACR CERO-s GAVE VUPA PETR-r? GEMO VINI IRMA SIMA RHTO-r HOMU VIVA LASU TOAR SEAR-r? HOVI VUHI LOHI TODI WYRE-r HYGL Plants found in this cluster included trees such as the oaks (Quercus sp; QUWI, QUKE, QUCH, QUDO, QULO), Aesculus californica (Spach.) Nutt. (AECA), and Pinus ponderosa (PIPO), vines such as Toxicodendron diversiloba, and low growing forbs such as Galium spp. including the rare G. californicum ssp. sierrae (GACA). Many of the exotic species such as annual grasses Avena fatua L., Bromus diandrus Roth, B. hordeaceous L., Lolium multiflorum L., and Taeniatherum caput-medusae (L.) Nevski (AVFA, BRDI, BRHO, LOMU, and TACA) and forbs Trifolium dubium Sibth., TJ. pretense L., and Erodium brachycarpum (Godr.) Thell. (TRDU, TRPR, and ERBR) occurred in the “‘Grassland”’ cluster and were top ranked along the distur- bance arrow (Green group Fig. 3; Table 8). ““Grassland”’ was associated with granitic soils on generally level sites, and was highest rated for agriculture according to the Storie Index. ““Grassland’’ was negatively associated with shrub cover, plant family diversity, and rare species and strongly associated with Exotic Species (““ExoS’’), Since one of the initial goals of this study was to investigate the existence of plant communities that included rare and endangered plants living upon relatively unique soils, special attention was given to plots that included rare species. Within the plot study, only 19 plots possessed rare taxa; all of those were located in either chaparral or woodland areas of gabbro soils. None of the rare plant species was found in the ‘‘Grassland”’ cluster (Table 7). Of the rare taxa, Calystegia stebbinsii (CAST), Ceanothus roderickii (CERO), Chloro- galum grandiflorum (CHGR), and Fremontoden- _ dron californica ssp decumbens (FRCA) were most abundant in the ““Chaparral 1” cluster and closely associated with each other. Galium californicum ssp sierrae (GACA) was found in the ‘““Wood- land” cluster adjacent to “Chaparral 2’. Packera layneae (PALA) and Wyethia reticulata (WYRE) were more abundant in the ““Chaparral 2” cluster. Variance partitioning of biotic sources of vari- ance from abiotic sources revealed that 12.5% of the total species variation was explained by purely abiotic factors and 18.6% by biotic factors (Tables 8 and 9). According to permutation tests, both of these sources of variation were significant (P = 0.002) and were of equal weight in explaining variance (at the 5% level). The two categories of variables shared 14.8% of the total species variance. Partitioning the explanatory variables into spatially explicit (longitude and latitude) and the remaining environmental variables suggested that there may be a small amount of beta diversity among the sites. A linear model of spatial variables explained about 1.3% of the total species variation. An additional 1.1% of the variation was explained jointly by spatial ar- rangement and the remaining environmental variables. Whereas the full model (P = 0.002) explained a significant portion of species varia- tion according to permutation tests, the purely spatial sources (P = 0.054) explained only a marginally significant portion. Species diversity in terms of the Shannon diversity index (H’) tended to be highest on gabbro soils and lowest on metamorphic soils, and was highest in ““Woodland”’ and lowest in “Chaparral” plots (Fig. 4). Species evenness among sites was similar within rock formation groups; **“Wood- land” and “Grassland” plots were more homoge- neous than the chaparral plots. 260 TABLE 8. MADRONO [Vol. 56 VARIANCE EXPLAINED BY ABIOTIC AND BIOTIC FACTORS IN THE PINE HILL FLORA. The trace is the sum of all canonical variables in the analysis. The F ratio and P-values were generated by Monte Carlo permutation tests (see text for details). Source Trace Abiotic ignoring biotic yg I Biotic ignoring abiotic 2.104 Both 2.891 Abiotic adjusted for biotic 0.787 Biotic adjusted for abiotic 1173 Total inertia 6.290 DISCUSSION The Pine Hill area stands out as an ecological island of considerable interest due to its diverse flora, vegetation types, rare plant species, and uncommon geology. The 731 species of vascular plants found there and on its borders account for more than 10% of the plant species found in the entire state of California (6,885 species, Hickman 1993) while encompassing less than 0.05% of the area of the entire State. Within this small area we found a diversity of plant forms (ferns, grasses, forbs, shrubs, vines, and trees) within three main community types, many native species including edaphic endemic species, a rich non-native flora, geological and topographic complexity that created numerous habitats, and natural and human-caused disturbances that created tempo- ral diversity. Any or all of these factors interacted to produce an area about 200-fold more diverse on average than the State as a whole. The distributions of species were related equally to biotic (cover, native species diversity, etc.) and abiotic variables (serpentine soil, soil texture, etc.). Variance in species distributions due to spatial constraints or correlations was small (<2% of variation), which suggests that dispersal limitations have not played a role in community structure at the spatial scale of the Pine Hill gabbro intrusion although dispersal limitations may have played a role at both larger and smaller spatial scales (Bell 2005). TWIN- SPAN and CCA analysis were in agreement in identifying three basic vegetation types within the study area. The first and most common of these was chaparral shrublands. Overall the chaparral of the study area was rich in terms of native species diversity and had relatively few exotic species. Much of the chaparral was composed of extremely thick stands of Adenostoma fascicula- tum (chamise) and/or Arctostaphylos viscida (whiteleaf manzanita). This type of chaparral occurred on south and southwest facing slopes on gabbro or serpentine soils. A second type of chaparral, denoted by the presence of evergreen shrubs Heteromeles arbutifolia and Rhamnus tomentella, and deciduous shrub/tree Cercis occidentalis, occurred on sites with moderate F ratio P value % variance 3.326 0.002 213 4.452 0.002 33.4 3.340 0.002 46.0 1.818 0.002 b225 2.712 0.002 18.6 100.0 exposure and was intermediate in our analysis between “Woodland” and “‘Chaparral. The two main strategies by which chaparral plants regenerate after fire are vegetative re- sprouting and recruitment from seeds whose germination is cued by fire. Shrubs such as Arctostaphylos viscida and Ceanothus cuneatus (Hook.) Nutt. are referred to as obligate seeders as the plants are killed by fire and the species must regenerate from long-lived seed stored in the soil seed bank (Keeley 1987, 1991). While the seedlings are able to exploit the high light, nutrient, and water availability of the post-fire environment in the spring following fire, they are then subject to severe moisture stress during the summer drought. As a consequence, these species have evolved higher tolerance to drought than the seedlings of obligate resprouters (Keeley 1998). Obligate resprouters, such as Heteromeles arbu- tifolia and Rhamnus tomentella, are not killed by fire but resprout from underground structures such as lignotubers, roots, and/or rhizomes following fire. They do not depend on fire to cue the germination of their seeds; indeed seeds may be short-lived or killed by fire’s heat. However, some resprouters, such as Wyethia reticulata (Ayres in press), may not flower until the shrub canopy is removed and thus are indirectly dependant on fire for sexual repro- duction. In general, seedlings of resprouting species are less drought tolerant than the seedlings of seeders (Davis et al. 1998; Keeley 1998) and may require shaded, mesic sites for seedling survival, such as under the shrub canopy. Some species, such as Adenostoma fasciculatum, are termed “‘facultative seeders” as they employ both strategies; the plants and seeds both survive fire and thus these species can both resprout and germinate following fire. Based on species response models Meentemeyer et al. (2001) have suggested that limitations on seed germination and seedling survival affect land- scape patterns of shrub establishment with fire- dependant seeding species occurring on xeric, exposed slopes, while resprouting species are more common on protected, mesic sites. This interpretation is consistent with the chaparral communities we found. 2009] TABLE 9. VARIANCE DECOMPOSITION OF THE EFFECT OF ABIOTIC AND BIOTIC FACTORS ON GABBRO ASSOCIATED VEGETATION. Computations are based on CCA analyses presented in Table 9 and the components correspond to those depicted in Figure 3. Component Source Variance Percentage A Pure abiotic WSs [25 B Shared 0.930 14.8 C Pure biotic 1.173 18.7 D Residual 5.399 54.0 Our study suggests that there are two distinct chaparral types in what has been previously identified as one community, ““Northern Gab- broic Chaparral” (Holland 1986), and more recently as the (Arctostaphylos viscida — Adenos- toma fasciculatum) | Salvia sonomensis Associa- tion (Klein et al. 2007). ““Chaparral 1”, dominat- ed by chamise (ADFA) and manzanita (ARVI) was associated with a harsh set of environmental conditions in the CCA and contained a distinct set of plant species many of which respond to fire by facultative or obligate seeding (Fig. 3, Ta- ble 7). We termed this community “‘Xeric Seed- ing’ to denote the harsh environment and dominant method of fire regeneration. As well, this type of chaparral was identified and classified using TWINSPAN as “Tall Chaparral” (Ta- ble 5). “Chaparral 2”’, identified as a ““Chaparral- Woodland” transitional type in TWINSPAN, was characterized in the CCA by more moderate environmental conditions and species that em- ploy a resprouting strategy to survive fire, e.g., evergreen shrub species Heteromeles arbutifolia (HEAR), and Rhamnus tomentella (RHTO), and deciduous Cercis occidentalis which both re- sprouts following fire (Anderson 1991) and has long-lived seed that survives fire. We termed this type of chaparral “‘Mesic Reprouting”’. “Woodland”, the second main woody vegeta- tion type, appeared where the chaparral-covered slopes came together to form a pattern of drainage gullies and stream courses, and extended into the lower and narrower riparian canyons of the region. Woodland vegetation, with occasional elements from higher elevation forest (e.g., Pinus ponderosa), followed the pattern of drainage courses and streambeds. In addition to serving as riparian tree cover, woodland vegetation covered the north-facing slopes of the steeper hills and ridges as well. A rich variety of native plant taxa occurred in the ‘““Woodland”’ and this community had the highest Shannon’s H’ diver- sity index (Fig. 4). In many wooded areas, three structural layers or strata were found: a canopy _of overstory trees, an understory layer of shrubs -and smaller trees, and an herbaceous ground cover. Like the chaparral, the woodland vegeta- ‘tion varied in density. Some areas were extremely thick and almost impenetrable; these were WILSON ET AL.: BIODIVERSITY OF PINE HILL 261 0.8 0.7 Gabbro e @ Metamorphic 4 Serpentine @ Granite © Chaparral Q Savanna 4 Woodland 0.4 : ait Shannon's evenness (H/log(N)) oO o Shannon's diversity (H) Fic. 4. Shannon’s diversity (X-axis) and evenness (Y- axis) for species associations based on soil (solid shapes) and vegetation type (open shapes). identified using TWINSPAN as species-rich ““Woodland”’’. The upper layer of this vegetation type was usually quite closed, providing cooler micro-climates beneath the canopy of live oaks and vines. Other ‘““Woodland” types were open, park-like meadows of native and exotic forbs with scattered Blue Oaks (QUDO) (*“Blue-Oak Savanna’, Table 5). Intermediate between “Woodland” and “Shrubland” was a community that contained shrubs typical of ““Chaparral 2” and included Foothill Pine (PISA) (“Chaparral- Woodland” Transition Table 5). In the wider, open valleys of the region, the chaparral and woodlands gave way to the third basic vegetation type, the grasslands. Most of the species were common exotic annuals (e.g., Avena spp. Bromus spp., Erodium spp., Lotus spp., Trifolium spp., Tables 5 and 7) that germinated in the fall and early spring, set seed, and were dead by early summer. This species composition was typical of what has been observed in the California foothill grasslands for at least several decades (Bentley and Talbot 1948) with the exception of more recent arrivals, Aegilops triuncialis L. and Taeniatherum caput-medusae. In the Pine Hill area, this vegetation was strongly associated with high numbers of exotic species, high levels of disturbance, granitic soils, little slope and a high Storie Index. They appeared as open sunny meadows with occasional scattered oaks (Quercus douglasii, QO. wislizenii and occa- sionally Q. lobata Nee) and California buckeye (Aesculus californica) that provided disconnected patches of shade. Past and current grazing practices may maintain this vegetation type (Bentley and Talbot 1948). Rare Taxa of Pine Hill Area No single location or vegetation type was found to contain all of the rare plant species. Of 262 the three basic vegetation types in the Pine Hill area, only the exotic-dominated ‘“‘Grassland” lacked rare plant species. Calystegia stebbinsii, Ceanothus roderickii, Galium californicum ssp. sierrae, Fremontodendron californicum ssp decum- bens, and Wyethia reticulata were only found on gabbro soils, although C. stebbinsii is known to occur on serpentine soils in Nevada County (CNDDB 2008), and Packera layneae occurred on three soil types (Table 1). It is not obvious from our analyses why five rare species should be restricted to gabbro-derived soil in El Dorado Co. In fact, serpentine substrate played a larger role in community structuring than gabbro in our CCA analysis. Stringent environmental condi- tions were associated with both rare (FRCA, CERO, CAST) and widespread (ADFA, ARVI) species; less stringent conditions were similarly associated with both rare (GACA, PALA, WYRE) and widespread species (HEAR, RHTO, TODI). Dispersal limitation may play a role restricting species distributions at the scale of single habitat patches and over broader regional scales where seed movement is infrequent (Bell 2005), but it apparently did not play a large role at the spatial scale of our study. In short, we did not find an explanation for the limited distribu- tions of the rare species. The rare species have been observed recovering after controlled burns as well as wildfires. Studies of recovery after fires of both types in the Pine Hill area indicated that Ceanothus roderickii, Fremontodendron californicum ssp. decumbens, and Calystegia stebbinsii recover from fire through seeds in the soil whose germination is promoted by fire (Boyd 1987, 2007; Nosal 1997) (Table 1). Calystegia stebbinsii, a short-lived twining vine with a woody caudex and rhizomes, may also be able to resprout after short-interval fires as has been observed for C. macrostegia (Greene) Brummitt, a congener with similar growth traits, in southern California chaparral (Keeley et al. 2006). Wyethia reticulata (Boyd 1987; Ayres and Ryan 1997), Chlorogalum grand- iflorum (personal observation), Fremontodendron californicum ssp. decumbens (Boyd 1987) and Packera layneae (personal observation) can re- sprout from underground roots, bulbs, or rhi- zomes after fire. Significantly, each chaparral type contained a different assemblage of rare species; ““Chaparral 1” contained four rare species (CAST, CERO, CHGR, FRCA) while ““Chaparral 2” contained two rare species (PALA, WYRE). Galium cali- fornicum ssp. sierrae (GACA) was located in CCA space in the ““Woodland” community near the border with ““Chaparral 2”. While our results were based on only 19 plots containing rare species, recently Gogol-Prokurat analyzed 79 chaparral relevés containing One or more rare plants from the Pine Hill area (Gogol-Prokurat MADRONO [Vol. 56 2009). She found that relevés where ‘“‘Chaparral 2” plants (e.g., CABR, CEOC, HEAR, RHTO) were present at cumulative cover values of 3% or higher had more occurrences of resprouting species WYRE, PALA, and CHGR, and facul- tative seeder FRCA than plots that did not contain these mesic chaparral species. CERO and CAST, obligate seeders were found predominant- ly in xeric Chaparral type | relevés. Thus, the modes of regeneration of the rare species are tied to environmental harshness and the regeneration strategies of diagnostic com- mon shrub species. This association is impor- tant for the preservation of these rare plants for the following reasons: 1) both types of chapar- ral should be targeted for preservation as each potentially contains a different sub-set of rare species; 2) the search for new populations of a particular rare species, especially those species present only in the seed bank, may be facilitated by looking for diagnostic shrub species; 3) while the regeneration of populations of one or possibly two species (CERO and _ possibly CAST) requires fire, the regeneration of others (WYRE, PALA, CHGR) may be possible with mechanical removal of the shrub canopy to promote flowering, and/or planting seed into the thick litter of established stands (FRCA, see Boyd and Serrafini 1992); and, 4) if artificial populations are deemed necessary, the selection of the appropriate type of chaparral for each species may promote the success of those efforts. Galium californicum ssp sierrae (GACA) was the only rare species not found in chaparral. Much of its biology, including its mode of regeneration following fire, is unknown. Thought to be an oak woodland species, GACA was placed within the Quercus kelloggii | Arctostaph- ylos viscida Provisional Association by Klein et al. (2007), an association that included several of the “Chaparral 2” shrubs identified here (e.g., HEAR, CEOC, and RHTO) and rare perennial Wyethia reticulata (Fig. 3). Of note, after a 2007 fire G. californicum ssp sierra was observed resprouting near fire-killed trunks of Q. kelloggii trees in a community that contained resprouting Packera layneae, W. reticulata, Heteromeles arbutifolia, and reseeding Cercis occidentalis — plants of or in close association to ‘““Chaparral 2” vegetation. This occurrence suggests that the native community of this tiny plant may be more like ““Chaparral 2” than oak woodland. ACKNOWLEDGMENTS We have greatly benefited from the assistance of Dr. M. Josephine Van Ess (deceased), including plant identification, field work, inclusion of her extensive — plant list (see Appendix 1) and helpful suggestions and encouragement. We also wish to thank Dr. Mary A. Reihman and Dr. Marda West (deceased) for their assistance, encouragement and critical reading of early — 2009] versions of this document, and Dr. Melanie Gogol- Prokurat and two anonymous reviewers for critical readings of the current versions. The Central Valley Project Conservation Program and Central Valley Project Improvement Act Habitat Restoration Program provided funding (Grant # 06FG204164 for Gabbro Soil Endemic Plant Research). LITERATURE CITED ANDERSON, M. K. 1991. California Indian horticulture: management and use of redbud by the Southern Sierra Miwok. Journal of Ethnobiology 11:145— L7. APARICIO, J. 1978. The plants of Ione. Fremontia 6:14—-16. AYRES, D. R. in press. Effects of fire on the demography of three rare chaparral plants (Calys- tegia stebbinsii, Ceanothus roderickii, and Wyethia reticulata). Proceedings of the California Native Plant Society, Sacramento 2009. AND F. J. RYAN. 1997. The clonal and population structure of a rare endemic plant, Wyethia reticulata (Asteraceae): allozyme and RAPD analysis. Molecular Ecology 6:761—772. BELL, G. 2005. The co-distribution of species in relation to the neutral theory of community ecology. Ecology 86:1757—1770. BENTLEY, J. R. AND M. W. TALBOT. 1948. Annual- plant vegetation of the California foothills as related to range management. Ecology 29:72-79. BRINK, C. 2010. History of the Pine Hill Preserve. Website: http://www.pinehillpreserve.org/about/ history/index.htm [accessed 05 March 2010]. BoybD, B. 1987. The effects of controlled burning on three rare plants. Pp. 513-517 in T. S. Elias (ed.), Conservation and management of rare and endan- gered plants. California Native Plant Society, Sacramento, CA. Boyb, R. S. 2007. Response to fire of Ceanothus roderickii (Rhamnaceae), a federally endangered California endemic shrub. Madrono 54:13-21. AND L. L. SERAFINI. 1992. Reproductive attrition in the rare chaparral shrub Fremontoden- dron decumbens Lloyd (Sterculiaceae). American Journal of Botany 79:1264—1272. BRADY, K. U., A. R. KRUCKEBERG, AND H. D. BRADSHAW. 2005. Evolutionary ecology of plant adaptation to serpentine soils. Annual Review of Ecology, Evolution and Systematics 36:243—266. CALIFORNIA NATURAL DIVERSITY DATABASE (CNNDB). 2008. Clarksville, Shingle Springs, and Placerville Quads. California Department of Fish and Game, Sacramento, CA. Website: http://www. dfg.ca.gov/biogeodata/cnddb/mapsanddata.asp [accessed 29 March 2010]. ~COWLING, R. M., P. W. RUNDEL, B. B. LAMONT, M. K. | ARROYO, AND M. ARIANOUTSOU. 1996. Plant diversity in Mediterranean-climate regions. Trends in Ecology and Evolution 11:362—366. DAUBENMIRE, R. F. 1974. Plants and environment, a textbook of plant autecology. John Wiley & Sons, | New York, NY. Davis, S. D., K. J. KOLB, AND K. P. BARTON. 1998. Ecophysiological processes and demographic pat- terns in the structuring of California chaparral. Pp. 297-312 in P. W. Rundel, G. Montenegro, and F. M. Jaksic (eds.), Landscape disturbance and WILSON ET AL.: BIODIVERSITY OF PINE HILL 263 biodiversity in Mediterranean-type ecosystems. Springer-Verlag, Berlin, Germany. EL DORADO COUNTY. 2007. Pine Hill Preserve — a brief history and issue paper. El] Dorado County Board of Supervisors, Placerville, CA. Website: http:// www.co.el-dorado.ca.us/bos/wwwroot/Attachments/ eadf4302-3c24-473a-bee4-9cc2f2673a84.doc [ac- cessed 08 April 2010]. FRANKLIN, J., C. L. COULTER, AND S. J. REy. 2004. Change in a southern California chaparral com- munity over 70 yr related to fire history. Journal of Vegetation Science 15:701—710. GOLDHABER, M. B., J. M. MORRISON, J. M. HOLLO- WAY, R. B. WANTy, D. R. HELSEL, AND D. B. SMITH. 2009. A regional soil and sediment geo- chemical study in northern California. Applied Geochemistry 24:1482—1499. GOGOL-PROKURAT, M. 2009. Ecology and demogra- phy of four rare endemic plants on gabbro soils: implications for habitat suitability modeling and conservation. Ph.D. dissertation, University of California, Davis, CA. HAJEK, M., P. HEKERA, AND P. HAJKOVA. 2002. Spring fen vegetation and water chemistry in the Western Carpthian flysch zone. Folio Geobotanico 37:205—224. HICKMAN, J. C. 1993. The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. HILL, M. O. 1979. TWINSPAN ~— a Fortran program for arranging multivariate data in an ordered two- way table by classification of the individuals and attributes. Cornell University, Ithaca, NY. Hocker, H. W. 1979. Introduction to forest biology. John Wiley & Sons, New York. HOLLAND, R. F. 1986. Preliminary descriptions of terrestrial natural communities of California. Unpublished report. California Department of Fish and Game, Sacremento, CA. Website: http:// www.cal-ipc.org/ip/inventory/pdf/HollandReport. pdf [accessed 05 March 2010]. HOWARD, A. Q. 1978. Pine Hill — a case in point. Fremontia 5:3—S. KEELEY, J. E. 1987. Role of fire in seed germination of woody taxa in California chaparral. Ecology 68:434-443. 1991. Seed germination and life history syndromes in the Californian chaparral. Botanical Review 57:81—116. . 1998. Coupling demography, physiology and evolution in chaparral shrubs. Pp. 257-264 in P. W. Rundel, G. Montenegro, and F. M. Jaksic (eds.), Landscape disturbance and biodiversity in Medi- terranean-type ecosystems. Springer-Verlag, Berlin, Germany. , C. J. FOTHERINGHAM, AND M. BAER-KEELEY. 2006. Demographic patterns of postfire regenera- tion in Mediterranean-climate shrublands of Cali- fornia. Ecological Monographs 76:235—255. KLEIN, A., J. CRAWFORD, J. EVENS, T. KEELER-WOLF, AND D. HICKSON. 2007. Classification of the vegetation alliances and associations of the north- ern Sierra Nevada Foothills, California. Report prepared for California Department of Fish and Game. California Native Plant Society, Sacramento, CA. Website http://www.dfg.ca.gov/biogeodata/ vegcamp/pdfs/VegMappingRpt_Vol2_northern_ Sierra_Foothills.pdf [accessed 05 March 2010]. 264 Kress, C. J. 1999. Ecological methodology, 2nd ed. Benjamin/Cummings Publishing, Menlo Park, CA. KUNZ, M. E. 1985. Influence of serpentine derived soil on vegetational pattern at mid elevations in the northern Sierra Nevada of California. M.A. thesis, California State University, Sacramento, CA. LEGENDRE, P. 2007. Studying beta diversity: ecological variation partitioning by multiple regression and canonical analysis. Journal of Plant Ecology 1:3-8. , D. BORCARD, AND P. R. PERES-NETO. 2005. Analyzing beta diversity: partitioning the spatial variation of community composition data. Ecolog- ical Monographs 75:435—450. Leps, J. AND P. SMILAUER. 2003. Multivariate analysis of ecological data using CANOCO. Cambridge University Press, New York. MASON, H. L. 1946. The edaphic factor in narrow endemism. I. the nature of environmental influenc- es. Madrono 8:209—226. MEENTEMEYER, R. K., A. MOODY, AND J. FRANKLIN. 2001. Landscape-scale patterns of shrub-species abundance in California chaparral: the role of topographically mediated resource gradients. Plant Ecology 156:19-41. Mooney, H. A., S. L. GULMON, D. J. PARSONS, AND A. T. HARRISON. 1974. Morphological changes within the chaparral vegetation type as related to elevational gradients. Madrono 22:281—285. MorRISON, J. M., M. B. GOLDHABER, L. LEE, J. M. HOLLOWAY, R. B. WANTY, R. E. WOLF, AND J. F. RANVILLE. 2009. A regional-scale study of chro- mium and nickel in soils of northern California, USA. Applied Geochemistry 24:1500—1511. MUELLER-DOMBOIS, D. AND H. ELLENBERG. 1974. Aims and methods of yegetation ecology. John Wiley & Sons, New York. NEWBERRY, D. W. (ed.). 1972. The plants and animals of Folsom Lake State Recreation Area. California Department of Parks and Recreation, Sacremento, CA. NOosAL, T. 1997. A demographic study of Stebbins’ morning-glory (Calystegia stebbinsii Brummit, Convolvulaceae), a California state-listed and federal listed endangered plant species. M.A. thesis. California State University, Sacramento, CA. PAvilk, B. 2003. Plants that protect ecosystems: a survey from California. Biodiversity and Conser- vation 12:717—729. RICKLEFS, R. E. 1976. The economy of nature. Chiron Press, Portland, OR. MADRONO [Vol. 56 ROGERS, J. H. 1974. Soil survey of El Dorado area, California. United States Department of Agricul- ture, Washington, D.C. SAS INSTITUTE INC. 2004. SAS. OnlineDoc® 9.1.3. Cary, NC. SPRINGER, R. K. 1968. The Pine Hill intrusive complex, El Dorado County, California. M. A. thesis. University of California, Davis, CA. SPURR, S. H. AND B. V. BARNES. 1973. Forest ecology. Ronald Press, New York. STEBBINS, G. L. 1978. Why are there so many rare plants in California? I. environmental factors. Fremontia 5:6—10. AND D. SMITH. 1960. Plant list for Pine Hill State Nature Reserve, El Dorado County. Califor- nia Department of Fish and Game, Sacremento, CA. TER BRAAK, C. J. F. 1986. Canonical correspondence analysis: a new eigenvector technique for multivar- late direct gradient analysis. Ecology 67:1167—1179. AND P. SMILAUER. 2002. CANOCO reference manual and CanoDraw for Windows user’s guide: software for canonical community ordination (Version 4.5). Microcomputer Power, Ithaca, NY. UNITED STATES BUREAU OF RECLAMATION (USBR). 1981. Fact sheet, Folsom Dam - Folsom power- plant - Folsom Lake, Central Valley project, California. U.S. Department of the Interior, Washington, D.C. UNITED STATES DEPARTMENT OF AGRICULTURE (USDA). 1980. Soil survey of Placer County, California, western part. Washington, D.C. UNITED STATES (U.S.) FISH AND WILDLIFE SERVICE. 1996. Endangered and threatened wildlife and plants; determination of endangered status for four plants and threatened status for one plant from the central sierran foothills of California. Federal Register 61(203): 54346-54358. UNITED STATES GEOLOGICAL SURVEY (USGS). 1973. California topographic 7.5 min series quadrangle map: Shingle Springs, CA. U.S. Department of the Interior, Washington, D.C. . 1979. N.A.S.A. aircraft standard photo-series, color infrared print scene id. no. 5790027775243, frames 5243-5244. Eros data center, U.S. Depart- ment of the Interior, Washington, D.C. WHITTAKER, R. H., R. B. WALKER, AND A. R. KRUCKEBERG. 1954. The ecology of serpentine soils. Ecology 5:258—288. 2009] FLORA OF PINE HILL, EL DORApo COUNTY, CALIFORNIA. WILSON ET AL.: BIODIVERSITY OF PINE HILL APPENDIX 1 Determination of taxa in the flora comes from the following sources: N = Newberry (1972), R = Stebbins and Smith (1960), S = Stebbins (1978), V = Van Ess (unpublished plant list), and W = Wilson (this paper). Determination of native (N) or introduced (I) status of plants found in the plot study is from Hickman (1993). Occurrence of listed plants on specific substrate (Rock) is as follows: G = found on gabbro related and possibly other soils, NG = found on non-gabbro soils only, and — = insufficient information, substrate unknown. Family Aceraceae Aizoaceae Aizoaceae Alismataceae Amaranthaceae Anacardiaceae Anacardiaceae Apiaceae Apiaceae Apiaceae Apiaceae Apiaceae Apiaceae Apiaceae Apiaceae Apiaceae Apiaceae Apiaceae Apiaceae Apiaceae Apiaceae Apiaceae Apiaceae Apiaceae Apiaceae Apiaceae Apiaceae Apiaceae Apiaceae Apocynaceae Apocynaceae Aristolochiaceae Aristolochiaceae Asclepiadaceae Asclepiadaceae Asteraceae Asteraceae Asteraceae _ Asteraceae Asteraceae | Asteraceae _ Asteraceae _ Asteraceae _ Asteraceae Asteraceae _ Asteraceae _ Asteraceae Asteraceae Asteraceae _ Asteraceae Asteraceae Taxon Acer macrophyllum Pursh Cypselea humifusa Turp. Mollugo verticillata L. Alisma plantago-aquatica L. Amaranthus californicus (Moq.) S.Watson Rhus trilobata Nutt. Toxicodendron diversilobum (Torr. & A.Gray) Greene Anthriscus caucalis M.Bieb. Apiastrum angustifolium Nutt. Daucus carota L. Daucus pusillus Michx. Eryngium vaseyi J.M.Coult. & Rose var. vallicola (Jeps.) Munz Foeniculum vulgare Mall. Lomatium macrocarpum (Nutt. ex Torr. & A.Gray) J.M.Coult. & Rose Lomatium marginatum (Benth.) J.M.Coult. & Rose Lomatium utriculatum (Nutt. ex Torr. & A.Gray) J.M.Coult. & Rose Osmorhiza chilensis Hook. & Arn. Perideridia gairdneri (Hook. & Arn.) Mathias Perideridia kelloggii (A.Gray) Mathias Perideridia parishii (J.M.Coult. & Rose) A.Nelson & J.F.Macbr. Sanicula bipinnata Hook. & Arn. Sanicula bipinnatifida Douglas ex Hook. Sanicula crassicaulis Poepp. ex DC. Sanicula tuberosa Torr. Scandix pectin-veneris L. Tauschia hartwegii (A.Gray) J.F.Macbr. Torilis arvensis (Huds.) Link Torilis nodosa (L.) Gaertn. Yabea microcarpa (Hook. & Arn.) Koso-Pol. Apocynum cannabinum L. Vinca major L. Aristolochia californica Torr. Asarum hartwegii S.Watson Asclepias cordifolia (Benth.) Jeps. Asclepias fascicularis Decne. Achillea millefolium L. Achyrachaena mollis Schauer Agoseris grandiflora (Nutt.) Greene Agoseris heterophylla (Nutt.) Greene Agoseris retrorsa (Benth.) Greene Ambrosia psilostachya DC. Anaphalis margaritacea (L.) Benth. & Hook.f. Anthemis cotula L. Artemisia douglasiana Besser Aster chilensis Nees Aster radulinus A.Gray Baccharis pilularis DC. ssp. consanguinea (DC.) C.B.Wolf Balsamorhiza deltoidea Nutt. Balsamorhiza macrolepis Sharp Bidens frondosa L. Brickellia californica (Torr. & A.Gray) A.Gray Native or introduced N VT rt Zt 2 Zi ZZ, 27, Lid. ~2. - L712, «Z Ly «hi Pie ZL, Source Rock WNS < ZO) q) < < Zz DN v2) ZQAAQAG AgAaAANH Q Q Q QQ) } leer esha aeal| “suey aaa MONAT EN Aes eo el, (Ge aNGNevene ie: QQqaz aA aga 266 Family Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae MADRONO APPENDIX 1. CONTINUED. Taxon Calycadenia multiglandulosa DC. Calycadenia truncata DC. Carduus pycnocephatlus L. Centaurea melitensis L. Centaurea solstitialis L. Chaenactis glabriuscula DC. Chamomilla suaveolens (Pursh) Rydb. Chondrilla juncea L. Cichorium intybus L. Cirsium andersonii (A.Gray) Petr. Cirsium occidentale (Nutt.) Jeps. var. californicum (A.Gray) Keil & C.Turner Cirsium occidentale (Nutt.) Jeps. var. occidentale Cirsium vulgare (Savi) Ten. Conyza canadensis (L.) Cronquist Ericameria arborescens (A.Gray) Greene Erigeron foliosus Nutt. Erigeron inornatus (A.Gray) A.Gray Erigeron philadelphicus L. Eriophyllum lanatum (Pursh) Forbes var. grandiflorum (A.Gray) Jeps. Filago californica Nutt. Filago gallica L. Gnaphalium californicum DC. Gnaphalium canescens DC. ssp. beneolens (Davidson) Stebbins & Keil Gnaphalium canescens DC. ssp. microcephalum (Nutt.) Stebbins & Keil Gnaphalium luteoalbum L. Gnaphalium palustre Nutt. Gnaphalium purpureum L. Grindelia camporum Greene Grindelia procera Greene Helenium puberulum DC. Helianthus californicus DC. var. nevadensis (Greene) Jeps. Helianthus annuus L. ssp. lenticularis (Douglas ex Lindl.) Cockerell Helianthus californicus DC. Hemizonia fitchii A.Gray Hesperevax acaulis (Kellogg) Greene Hesperevax sparsiflora (A.Gray) Greene Heterotheca grandiflora Nutt. Holocarpha virgata (A.Gray) D.D.Keck Holozonia filipes (Hook. & Arn.) Greene Hypochaeris glabra L. Hypochaeris radicata L. Lactuca saligna L. Lactuca serriola L. Lagophylla glandulosa A.Gray Lagophylla ramosissima Nutt. Lasthenia californica DC. ex Lindl. Layia fremontii (Torr. & A.Gray) A.Gray Layia pentachaeta A.Gray Leontodon taraxacoides (Vill.) Mérat Lessingia leptoclada A.Gray Lessingia nemaclada Greene Lessingia virgata A.Gray Madia elegans D.Don ex Lindl. Madia elegans D.Don ex Lindl. ssp. densifolia (Greene) D.D.Keck Madia elegans D. Don ex Lindl. ssp. vernalis D.D.Keck Madia exigua (Sm.) A.Gray Native or introduced Let 22 ZZ, ZA, ZZ, Source WVNR [Vol. 56 Q Z QQ0QQ000Z000 Q QQQQQQQ Q ANAGA AARDAADAO Q Q QQ a Ay ZAAAAAAAAAAADADAA 2009] Family Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Asteraceae Berberidaceae Betulaceae Blechnaceae Boraginaceae Boraginaceae Boraginaceae Boraginaceae Boraginaceae - Boraginaceae - Boraginaceae | Boraginaceae _ Boraginaceae . Boraginaceae _ Boraginaceae _ Boraginaceae _ Boraginaceae _ Boraginaceae _ Boraginaceae Brassicaceae _ Brassicaceae Brassicaceae WILSON ET AL.: BIODIVERSITY OF PINE HILL APPENDIX 1. CONTINUED. Taxon Madia gracilis (Sm.) D.D.Keck & J.C.Clausen ex Applegate Madia minima (A.Gray) D.D.Keck Madia rammii Greene Madia subspicata D.D.Keck Micropus californicus Fisch. & C.A.Mey. Microseris acuminata Greene Microseris sylvatica (Benth.) A.Gray . Packera layneae (Greene) W.A.Weber and A.Loéve Pseudobahia heermannii (Durand) Rydb. Psilocarphus brevissimus Nutt. Psilocarphus tenellus Nutt. Rafinesquia californica Nutt. Rigiopappus leptocladus A.Gray Senecio aronicoides DC. Senecio flaccidus Less. var. douglasii (DC.) B.L.Turner & T.M.Barkley Senecio vulgaris L. Silybum marianum (L.) Gaertn. Solidago californica Nutt. Solidago canadensis L. ssp. elongata (Nutt.) D.D.Keck Solidago occidentalis Nutt. Soliva sessilis Ruiz & Pav. Sonchus asper (L.) Hill Sonchus oleraceus L. Stebbinsoseris heterocarpa (Nutt.) K.L.Chambers Stephanomeria virgata Benth. Stylocline filaginea (A.Gray) A.Gray Stylocline gnaphalioides Nutt. Taraxacum officinale F.H.Wigg. Tragopogon dubius Scop. Tragopogon pratensis L. Wyethia angustifolia (DC.) Nutt. Wyethia bolanderi (A.Gray) W.A.Weber Wyethia helenioides (DC.) Nutt. Wyethia reticulata Greene Xanthium strumarium L. Berberis aquifolium Pursh var. dictyota (Jeps.) Jeps. Alnus rhombifolia Nutt. Woodwardia fimbriata Sm. Amsinckia menziesii (Lehm.) A.Nelson & J.F.Macbr. Amsinckia menziesii (Lehm.) A.Nelson & J.F.Macbr. var. intermedia (Fisch. & C.A.Mey.) Ganders Cryptantha flaccida (Douglas ex Lehm.) Greene Cryptantha intermedia (A.Gray) Greene Cryptantha micrantha (Torr.) I.M.Jonst. Cryptantha muricata (Hook. & Arn.) A.Nelson & J.F.Macbr. Cryptantha muricata (Hook. & Arn.) A.Nelson & J.F.Macbr. var. denticulata (Greene) I.M.Johnst. Cynoglossum grande Douglas ex Lehm. Myosotis discolor Pers. Pectocarya pusilla (A.DC.) A.Gray Plagiobothrys canescens Benth. Plagiobothrys fulvus (Hook. & Arn.) I.M.Johnst. var. campestris (Greene) I.M.Johnst. Plagiobothrys nothofulvus (A.Gray) A.Gray Plagiobothrys stipitatus (Greene) I.M.Johnst. var. micranthus (Piper) I.M.Johnst. Plagiobothrys tenellus (Nutt. ex Hook.) A.Gray Arabidopsis thaliana (L.) Heynh. Arabis sparsiflora Nutt. Athysanus pusillus (Hook.) Greene Native or introduced N Leake ZZ, Z. iG e Lz ZZ ZZ Z Source VW Rock G eo Q Q) seer QQOAQMQQAQQQAQAQAAZAQGQAAAAAAAAAD ANAAAZARAADAAA q) Q Ao QO QO2000 0 Z 268 Family Brassicaceae Brassicaceae Brassicaceae Brassicaceae Brassicaceae Brassicaceae Brassicaceae Brassicaceae Brassicaceae Brassicaceae Brassicaceae Brassicaceae Brassicaceae Brassicaceae Brassicaceae Brassicaceae Brassicaceae Brassicaceae Brassicaceae Brassicaceae Brassicaceae Brassicaceae Callitrichaceae Campanulaceae Campanulaceae Campanulaceae Campanulaceae Caprifoliaceae Caprifoliaceae Caprifoliaceae Caprifoliaceae Caprifoliaceae Caprifoliaceae Caryophyllaceae Caryophyllaceae Caryophyllaceae Caryophyllaceae Caryophyllaceae Caryophyllaceae Caryophyllaceae Caryophyllaceae Caryophyllaceae Caryophyllaceae Caryophyllaceae Caryophyllaceae Caryophyllaceae Caryophyllaceae Caryophyllaceae Caryophyllaceae Chenopodiaceae Cistaceae Cistaceae Convolvulaceae Convolvulaceae Convolvulaceae Convulvulaceae Cornaceae Crassulaceae MADRONO APPENDIX 1. CONTINUED. Taxon Barbarea verna (Mill.) Asch. Brassica rapa L. Capsella bursa-pasturis (L.) Medik. Cardamine oligosperma Nutt. Draba verna L. Erysimum capitatum (Douglas ex Hook.) Greene Hirshfeldia incana (L.) Lagr.-Foss. Lepidium nitidum Nutt. Lepidium oblongum Small Lepidium strictum (S.Watson) Rattan Raphanus raphanistrum L. Raphanus sativus L. Rorippa curvisiliqua (Hook.) Besser ex Britton Rorippa nasturtium-aquaticum (L.) Hayek Rorippa palustris (L.) Besser var. occidentalis (S.Watson) Rollins Sisymbrium altissimum L. Sisymbrium irio L. Streptanthus polygaloides A.Gray Thysanocarpus curvipes Hook. Thysanocarpus curvipes Hook. var. elegans (Fisch. & C.A.Mey.) B.Rob. Thysanocarpus radians Benth. Tropidocarpum gracile Hook. Callitriche verna L. Githopsis pulchella Vatke Githopsis specularioides Nutt. Heterocodon rariflorum Nutt. Triodanis biflora (Ruiz & Pav.) Greene Lonicera hispidula (Lind1.) Douglas ex Torr. & A.Gray var. vacillans A.Gray Lonicera interrupta Benth. Sambucus mexicana C.Presl ex DC. Symphoricarpos albus (L.) S.F.Blake var. /aevigatus (Fernald) S.F.Blake Symphoricarpos mollis Nutt. Virburnum ellipticum Hook. Cerastium glomeratum Thuill. Lychnis coronaria (L.) Desr. Minuartia californica (A.Gray) Mattt. Minuartia douglasii (Fenzl ex Torr. & A.Gray) Mattf. Petrorhagia dubia (Raf.) G.Lopez & Romo Sagina apetala L. var. barbata Fenzl. Saponaria officinalis L. Scleranthus annuus L. Silene antirrhina L. Silene californica Durand Silene gallica L. Spergula arvensis L. Spergula rubra (L.) J.Presl & C.Presl Stellaria media (L.) Vill. Stellaria nitens Nutt. Velezia rigida L. Chenopodium ambrosioides L. Helianthemum scoparium Nutt. Helianthemum suffrutescens Schreib. Calystegia occidentalis (A.Gray) Brummitt Calystegia purpurata (Greene) Brummitt ssp. saxicola (Eastw.) Brummitt Calystegia stebbinsii Brummitt Convolvulus arvensis L. Cornus glabrata Benth. Crassula connata (Ruiz & Pay.) A.Berger Native or introduced I I N —_ ZZ PZ 2 LZ OL, Le eA. LiL ZZ ZZ, ZZ 24 2 Source WVNS [Vol. 56 v2) e) O nn ZN Ge a ANE? q) GG Gs 9) Q QQQ0 OO ee eae A QQQ0 AAQDAAAAO 2009] Family Crassulaceae Crassulaceae Crassulaceae Crassulaceae Crassulaceae Cucurbitaceae Cucurbitaceae Cupressaceae Cuscutaceae Cuscutaceae Cyperaceae Cyperaceae Cyperaceae Cyperaceae Cyperaceae Cyperaceae Cyperaceae Cyperaceae Cyperaceae Cyperaceae Cyperaceae Cyperaceae Cyperaceae Cyperaceae Cyperaceae Cyperaceae Cyperaceae Cyperaceae Cyperaceae Datiscaceae Dipsacaceae Dryopteridaceae Equisetaceae Equisetaceae Equisetaceae Ericaceae Ericaceae Ericaceae Ericaceae Euphorbiaceae Euphorbiaceae Euphorbiaceae Euphorbiaceae Euphorbiaceae Euphorbiaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae WILSON ET AL.: BIODIVERSITY OF PINE HILL APPENDIX 1. CONTINUED. Native or introduced Taxon Crassula tillaea Lester-Garland N Dudleya cymosa (Lem.) Britton & Rose N Parvisedum congdonii (Eastw.) R.T.Clausen Parvisedum pumilum (Benth.) R.T.Clausen Sedum spathulifolium Hook. Marah fabaceus (Naudin) Naudin ex Greene var. agrestis (Greene) Stocking Marah watsonii (Cogn.) Greene Calocedrus decurrens (Torr.) Florin Cuscuta californica Hook. & Arn. Cuscuta californica Hook. & Arn. var. breviflora Engelm. Carex athrostachya Olney Carex barbarae Dewey Carex brainerdii Mack. N Carex densa (L.H.Bailey) L.H.Bailey Carex dudleyi Mack. Carex gracilior Mack. Carex nebrascensis Dewey Carex nudata W.Boott Carex praegracilis W.Boott Carex rossii Boott Carex senta Boott Carex subbracteata Mack. Cyperus eragrostis Lam. Cyperus rotundus L. Cyperus squarrosus L. Eleocharis acicularis (L.) Roem. & Schult. Eleocharis pachycarpa Desv. Lipocarpha micrantha (Vahl.) G.Tucker var. minor (Schrad.) Friedl. Scirpus acutus Muhl. ex Bigelow var. occidentalis (S.Watson) Beetle Datisca glomerata (C.Presl) Baill. Dipsacus fullonum L. Dryopteris arguta (Kaulf.) Watt N Equisetum arvense L. Equisetum hyemale L. ssp. affine (Engelm.) A.A.Eaton Equisetum laevigatum A.Braun Arbutus menziesii Pursh Arctostaphylos manzanita Parry Arctostaphylos viscida Parry Rhododendron occidentale (Torr. & A.Gray) A.Gray Chamaesyce maculata (L.) Small Chamaesyce ocellata (Durand & Hilg.) Small Chamaesyce serpyllifolia (Pers.) Small Eremocarpus setigerus (Hook.) Benth. Euphorbia crenulata Engelm. Euphorbia spathulata Lam. Astragalus gambelianus Sheldon Cercis occidentalis Torr. ex A.Gray Cytisus scoparius (L.) Link Hoita macrostachya (DC.) Rydb. Hoita orbicularis (Lind|.) Rydb. Lathyrus jepsonii Greene var. californicus (S.Watson) C.L.Hitche. Lathyrus latifolius L. Lathyrus nevadensis S.Watson Lathyrus sulphureus W.H.Brewer ex A.Gray Lotus grandiflorus (Benth.) Greene Lotus humistratus Greene Lotus micranthus Benth. Lotus purshianus (Benth.) Clem. & E.G.Clem. Z ZL -L Pile LiL LL, Source WVS WVNS = < 2) —4 444 2 ane Oa on Q) ee a ae Oe Gare rare) VOOl eG 2ZozzZ aumae Q) QAQAAMDMQAADAAZAAD QAQAQAAD| 269 Family Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fabaceae Fagaceae Fagaceae Fagaceae Fagaceae Fagaceae Fagaceae Fagaceae Fagaceae Garryaceae Garryaceae Gentinaceae Gentinaceae Gentinaceae MADRONO APPENDIX 1. CONTINUED. Taxon Lotus scoparius (Nutt.) Ottley Lotus wrangelianus Fisch. & C.A.Mey. Lupinus albifrons Benth. Lupinus benthamii A.Heller Lupinus bicolor Lindl. Lupinus bicolor Lindl. ssp. microphyllus (S.Watson) D.Dunn Lupinus bicolor Lindl. ssp. pipersmithii (A.Heller) D.Dunn Lupinus latifolius Lindl. ex J.Agardh Lupinus latifolius Lindl. ex J.Agardh var. columbianus (A.Heller) C.P.Sm. Lupinus microcarpus Sims var. densiflorus (Benth.) Jeps. Lupinus microcarpus Sims Lupinus nanus Douglas ex Benth. Lupinus polyphyllus Lind. Medicago polymorpha L. Melilotus indica (L.) All. Melilotus officinalis (L.) Lam. Pickeringia montana Nutt. Robinea pseudoacacia L. Rupertia physoides (Douglas ex Hook.) Grimes Trifolium albopurpureum Torr. & A.Gray Trifolium albopurpureum Torr. & A.Gray var. olivaceum (Greene) Isely Trifolium barbigerum Torr. Trifolium bifidum A.Gray var. decipiens Greene Trifolium bifidum A.Gray Trifolium ciliolatum Benth. Trifolium depauperatum Desv. Trifolium dubium Sibth. Trifolium glomeratum L. Trifolium gracilentum Torr. & A.Gray Trifolium hirtum All. Trifolium incarnatum L. Trifolium microcephalum Pursh Trifolium microdon Hook. & Arn. Trifolium pratense L. Trifolium subterraneum L. Trifolium variegatum Nutt. Trifolium wildenovii Spreng. Vicia americana Muhl. ex Willd. Vicia benghalensis L. Vicia hirsuta (L.) Gray Vicia sativa L. Vicia sativa L. ssp. nigra (L.) Ehrh. Vicia villosa Roth Vicia villosa Roth ssp. varia (Host) Corb. Quercus chrysolepis Liebm. Quercus douglasii Hook. & Arn. Quercus dumosa Nutt. Quercus durata Jeps. Quercus kelloggii Newberry Quercus lobata Née Quercus wislizenti A.DC. Quercus X moreha Kellogg Garrya congdonii Eastw. Garrya fremonti Torr. Centaurium muehlenbergii (Griseb.) W.Wight ex Piper Centuarium venustum (A.Gray) Rob Swertia albicaulis (Douglas ex Griseb.) Kuntze var. nitida (Benth.) Jeps. Native or introduced N N “ZZ Z Z LZ. LA, A ZA, 77 A Source Be ee eee ee ee ee Oe OO OO ee Ze Ge ae Qa Qa. aqgggna/% nn WVNS ay FR N x Se Bana £83 [Vol. 56 2009] Family Geraniaceae Geraniaceae Geraniaceae Geraniaceae Geraniaceae Geraniaceae Geraniaceae Hippocastanaceae Hydrophyllaceae Hydrophyllaceae Hydrophyllaceae Hydrophyllaceae Hydrophyllaceae Hydrophyllaceae Hydrophyllaceae Hypericaceae Hypericaceae Hypericaceae Iridaceae Iridaceae Iridaceae Juglandaceae Juncaceae Juncaceae Juncaceae Juncaceae Juncaceae Juncaceae Juncaceae Lamiuaceae Lamiaceae Lamiaceae Lamiaceae Lamiaceae Lamiaceae Lamiaceae Lamiuaceae Lamiaceae Lamiaceae Lamiaceae Lamiaceae Lamiaceae Lamiaceae Lamiaceae Lamiaceae Lamiaceae Lamiuaceae Lamiaceae _ Lamiaceae _ Lamiuaceae _ Lamiaceae Lauraceae Liliaceae Liliaceae Liliaceae Liliaceae Liliaceae Liliaceae Liliaceae Liliaceae _ Liliaceae _ Lilaceae WILSON ET AL.: BIODIVERSITY OF PINE HILL APPENDIX 1. CONTINUED. Taxon Erodium botrys (Cav.) Bertol. Erodium brachycarpum (Godr.) Thell. Erodium cicutarium (L.) L’Her. ex Aiton Erodium moschatum (L.) L’ Her. ex Aiton Geranium carolinianum L. Geranium dissectum L. Geranium molle L. Aesculus californica (Spach) Nutt. Emmenanthe penduliflora Benth. Eriodictyon californicum (Hook. & Arn.) Torr. Nemophila heterophylla Fisch. & C.A.Mey. Nemophila maculata Benth. ex Lindl. Nemophila menziesii Hook. & Arn. Phacelia cicutaria Greene Phacelia imbricata Greene Hypericum concinnum Benth. Hypericum mutilum L. Hypericum perforatum L. Tris hartwegii Baker Tris macrosiphon Torr. Sisyrinchium bellum S.Watson Juglans californica S.Watson var. hindsii Jeps. Juncus balticus Willd. Juncus bufonius L. Juncus effusus L. var. pacificus Fernald & Wiegand Juncus nevadensis S.Watson Juncus oxymeris Engelm. Juncus tenuis Willd. Luzula comosa E.Mey. Lamium amplexicaule L. Lamium purpureum L. Lepechinia calycina (Benth.) Epling ex Munz Lycopus americanus Muhl ex W.Bartram Marrubium vulgare L. Mentha aquatica L. Mentha arvensis L. var. villosa (Benth.) S.R.Stewart Mentha piperita L. Mentha pulegium L. Mentha spicata L. Monardella villosa Benth. ssp. villosa Monardella viridis Jeps. Pogogyne serpylloides (Torr.) A.Gray Prunella vulgaris L. var. lanceolata (W.Bartram) Fernald Pycnanthemum californicum Torr. Salvia sonomensis Greene Satureja douglasii (Benth.) Briq. Scutellaria californica A.Gray Scutellaria siphocampyloides Vatke Scutellaria tuberosa Benth. Stachys stricta Greene Trichostema lanceolatum Benth. Umbellularia californica (Hook. & Arn.) Nutt. Allium hyalinum Curran Allium peninsulare Lemmon ex Greene Allium sanbornii Alph.Wood Allium serra McNeal & Ownbey Bloomeria crocea (TYorr.) Coville Brodiaea elegans Hoover Brodiaea purdyi Eastw. Calochortus albus Douglas ex Benth. Calochortus luteus Douglas ex Lindl. Calochortus monophyllus (Lindl.) Lem. Native or introduced Ly 2a Ly Lepr Lz ZZ, Source Rock QQ) 0000) 02000 Be OGG) GEG) Oh A ag ad aN PoE ay QAAAQQAAQAAAAAAAAAQ QAAANAA| Family Liliaceae Liliaceae Liliaceae Liliaceae Liliaceae Liliaceae Liliaceae Liliaceae Liliaceae Liliaceae Liliaceae Liliaceae Liliaceae Liliaceae Liliaceae Liliaceae Liliaceae Liliaceae Liliaceae Liliaceae Liliaceae Limnanthaceae Limnanthaceae Limnanthaceae Linaceae Linaceae Linaceae Loasaceae Lythraceae Lythraceae Malvaceae Malvaceae Malvaceae Marsileaceae Oleaceae Oleaceae Onagraceae Onagraceae Onagraceae Onagraceae Onagraceae Onagraceae Onagraceae Onagraceae Onagraceae Onagraceae Onagraceae Onagraceae Onagraceae Onagraceae Onagraceae Orchidaceae Orchidaceae Orchidaceae Orchidaceae MADRONO APPENDIX 1. CONTINUED. Taxon Calochortus superbus Purdy ex J.T.Howell Calochortus venustus Douglas ex Benth. Chlorogalum angustifolium Kellogg Chlorogalum grandiflorum Hoover Chlorogalum pomeridianum (DC.) Kunth Dichelostemma capitatum Alph. Wood Dichelostemma congestum (Sm.) Kunth Dichelostemma multilflorum (Benth.) A.Heller Dichelostemma volubile (Kellogg) A.Heller Erythronium multiscapoideum (Kellogg) A.Nelson & P.B.Kenn. Fritillaria micrantha A.Heller Lilium humboldtii Roezl & Leichtlin ex Duch. Lilium pardalinum Kellogg Odontostomum hartwegii Torr. Trillium chloropetalum (TYorr.) Howell Triteleia bridgesii (S.Watson) Greene Triteleia hyacinthina (Lindl.) Greene Triteleia ixioides (S.Watson) Greene Triteleia ixioides (S.Watson) Greene ssp. scabra Greene Triteleia laxa Benth. Zigadenus venenosus S.Watson Limnanthes alba Benth. Limnanthes douglasii R.Br. var. rosea (Hartw. ex Benth.) C.T.Mason Limnanthes striata Jeps. Hesperolinon micranthum (A.Gray) Small Linum bienne Mill. Linum usitatissimum L. Mentzelia laevicaulis (Douglas ex Hook.) Torr. & A.Gray Lythrum hyssopifolia L. Rotala ramosior (L.) Koehne Sidalcea calycosa M.E.Jones Sidalcea hartwegti A.Gray Sidalcea malvaeflora (Sesse & Mocino ex DC.) A.Gray ex Benth. ssp. asprella (Greene) C.L.Hitche. Marsilea vestita Hook. & Grev. Fraxinus dipetala Hook. & Arn. Fraxinus latifolia Benth. Camissonia micrantha (Hornem. ex Spreng.) P.H.Raven Clarkia biloba (Durand.) A.Nelson & J.F.Macbr. Clarkia gracilis (Piper) A.Nelson & J.F.Macbr. Clarkia purpurea (Curtis) A.Nelson & J.F.Macbr. Clarkia purpurea (Curtis) A.Nelson & J.F.Macbr.ssp. quadrivulnera (Douglas ex Lindl.) F.H. Lewis & M.E. Lewis Clarkia rhomboidea Douglas ex Hook. Clarkia unguiculata Lindl. Epilobium brachycarpum C.Presl Epilobium canum (Greene) P.H.Raven ssp. /atifolia (Hook.) P.H.Raven Epilobium ciliatum Raf. Epilobium cleistogama (Curran) P.Hoch & P.H.Raven Epilobium densiflorum (Lindl.) Hoch. & P.H.Raven Epilobium minutum Lindl. ex Lehm. Epilobium torreyi (S.Watson) Hoch. & P.H.Raven Ludwigia peploides (Kunth) P.H.Raven Epipactis gigantea Douglas ex Hook. Piperia elegans (Lindl.) Rydb. Piperia unalascensis (Spreng.) Rydb. Spiranthes porrifolia Lindl. Native or introduced ZL, OZ LL LL Ae LLL Le. Ly ZZ Zi La ZZ, Source S 27°32 <3 a2 2 332 “233 83 2 25 ve) 7) | QQQQ00 Q000 VNS | | a QAO Z EN EGC) GGG Ae eee! QQQQ000 | QQQQ AAAMAZ AAOZA ZQQQ a) [Vol. 56 Rock qQ) Q Q 2009] WILSON ET AL.: BIODIVERSITY OF PINE HILL eI) APPENDIX 1. CONTINUED. Native or Family Taxon introduced Source Rock Orobanchaceae Orobanche bulbosa (A.Gray) G.Beck N WVR G Orobanchaceae Orobanche fasciculata Nutt. R G Orobanchaceae Orobanche uniflora L. N WR G Orobanchaceae Orobanche uniflora L. var. sedii (Suksd.) Achey V NG Papaveraceae Eschscholzia caespitosa Benth. N WVNS G Papaveraceae Eschscholzia californica Cham. VNR G Papaveraceae Eschscholzia lobbii Greene N WVN G Papaveraceae Meconella californica Torr. VN G Papaveraceae Platystemon californicus Benth. WVN NG Pinaceae Pinus ponderosa C.Lawson N WVNSR G Pinaceae Pinus sabiniana Douglas ex Douglas N WVNSR_ G Pinaceae Pseudotsuga menziesii (Mirb.) Franco var. menziesii WNS G Plantaginaceae Plantago erecta Morris N WVNSR_ G Plantaginaceae Plantago lanceolata L. WVNSR G Plantaginaceae Plantago major L. VN G Poaceae Achnatherum lemmonii (Vasey) Barkworth N WVSR G Poaceae Aegilops triuncialis L. I WVR G Poaceae Agrostis exarata Trin. Vv G Poaceae Aira caryophyllea L. I WVSR G Poaceae Andropogon virginicus L. S G Poaceae Avena barbata Pott ex Link I WVSR G Poaceae Avena fatua L. I WVR G Poaceae Brachypodium distachyon (L.) P.Beauv. I WVSR G Poaceae Briza minor L. I WVSR G Poaceae Bromus arenarius Labill. S NG Poaceae Bromus carinatus Hook. & Arn. N WVR G Poaceae Bromus diandrus Roth I WVSR G Poaceae Bromus hordeaceus L. I WVS G Poaceae Bromus laevipes Shear N WVSR G Poaceae Bromus madritensis L. I WVSR G Poaceae Bromus madritensis L. ssp. rubens (L.) Duvin I WVSR G Poaceae Bromus sterilis L. I WVR G Poaceae Bromus tectorum L. Vv G Poaceae Crypsis schoenoides (L.) Lam. S G Poaceae Cynodon dactylon (L.) Pers. S G Poaceae Cynosurus echinatus L. I WVSR G Poaceae Danthonia californica Boland var. americana (Scribn.) R G Hitche. Poaceae Danthonia unispicata (Thurb.) Munro ex Macoun N WV G Poaceae Deschampsia danthonioides (Trin.) Munro R G Poaceae Digitaria sanguinalis (L.) Scop. Vv G Poaceae Echinochloa crusgalli (L.) P.Beauv. Vv G Poaceae Elymus elymoides (Rat.) Swezey V G Poaceae Elymus glaucus Buckley ssp. jepsonii Burtt Davy N Vv G Poaceae Elymus multisetus (J.G.Sm.) Burtt Davy N WVSR G Poaceae Eragrostis hypnoides (Lam.) Britton, Sterns & Vv NG Poggenb. Poaceae Gastridium ventricosum (Gouan) Schinz & Thell. I W G Poaceae Holcus lanatus L. S G | Poaceae Hordeum depressum (Scribn. & J.G.Sm.) Rydb. N WV G _ Poaceae Hordeum marinum Huds. ssp. gussoneanum (Parl.) I WVR G | Thell. | Poaceae Hordeum murinum L. ssp. leporinum (Link) Arcang. I WVR NG _ Poaceae Hordeum vulgare L. W G | Poaceae Koeleria macrantha (Ledeb.) Schult. R G | Poaceae Leersia oryzoides (L.) Sw. VS G Poaceae Lolium multiflorum Lam. I WV G | Poaceae Lolium perenne L. I WV G | Poaceae Lolium temulentum L. I W G Poaceae Melica californica Scribn. N WVSR G Poaceae Melica torreyana Scribn. N WVSR G - Poaceae Muhlenbergia rigens (Benth.) Hitche. VS G Poaceae Nassella cernua (Stebbins & R.M.Love) Barkworth N WVSR G Family Poaceae Poaceae Poaceae Poaceae Poaceae Poaceae Poaceae Poaceae Poaceae Poaceae Poaceae Poaceae Poaceae Poaceae Poaceae Poaceae Poaceae Poaceae Poaceae Poaceae Poaceae Poaceae Poaceae Poaceae Poaceae Poaceae Polemoniaceae Polemoniaceae Polemoniaceae Polemoniaceae Polemoniaceae Polemoniaceae Polemoniaceae Polemoniaceae Polemoniaceae Polemoniaceae Polemoniaceae Polemoniaceae Polemoniaceae Polemoniaceae Polemoniaceae Polemoniaceae Polemoniaceae Polemoniaceae Polemoniaceae Polemoniaceae Polemoniaceae Polemoniaceae Polygalaceae Polygonaceae Polygonaceae Polygonaceae Polygonaceae Polygonaceae Polygonaceae Polygonaceae MADRONO APPENDIX 1. CONTINUED. Taxon Nassella pulchra (Hitche.) Barkworth Panicum acuminatum Sw. var. acuminatum Panicum capillare L. Phataris aquatica L. Phalaris lemmonii Vasey Phalaris minor Retz. Piptatherum miliaceum (L.) Coss. Poa annua L. Poa bulbosa L. Poa compressa L. Poa pratensis L. Poa secunda J.Presl ssp. secunda Poa tenerrima Scribn. Polypogon maritimus Willd. Polypogon monspeliensis (L.) Desf. Scribneria bolanderi (Thurb.) Hack. Setaria pumila (Poir.) Roem. & Schult. Sorghum halepense (L.) Pers. Taeniatherum caput-medusae (L.) Nevski Vulpia bromoides (L.) Gray Vulpia microstachys (Nutt.) Munro var. ciliata (Beal) Lonard & Gould Vulpia microstachys (Nutt.) Munro var. confusa (Piper) Lonard & Gould Vulpia microstachys (Nutt.) Munro var. pauciflora (Scribn. ex Beal) Lonard & Gould Vulpia myuros (L.) C.C.Gmel. Vulpia myuros (L.) C.C.Gmel. var. hirsuta Hack. Vulpia octoflora (Walter) Rydb. var. hirtella (Piper) Henr. Allophyllum divaricatum (Nutt.) A.D.Grant & V.E.Grant Allophyllum gilioides (Benth.) A.D.Grant & V.E.Grant Collomia heterophylla Hook. Gilia capitata Sims Gilia capitata Sims ssp. pedemontana V.E.Grant Gilia tricolor Benth. Gilia tricolor Benth. ssp. diffusa (Congd.) H.Mason & A.D.Grant Linanthus androsaceus (Benth.) Greene Linanthus bicolor (Nutt.) Greene Linanthus ciliatus (Benth.) Greene Linanthus dichotomus Benth. Linanthus filipes (Benth.) Greene Linanthus montanus (Greene) Greene Linanthus parviflorus (Benth.) Greene Linanthus pygmaeus (Brand) J.T.Howell Navarretia eriocephala H.Mason Navarretia filicaulis (Torr. ex A.Gray) Greene Navarretia intertexta (Benth.) Hook. Navarretia pubescens (Benth.) Hook. & Arn. Navarretia viscidula Benth. Navarretia viscidula Benth. ssp. purpurea (Greene) H.Mason Phlox gracilis Greene Polygala cornuta Kellogg Chorizanthe membranacea Benth. Chorizanthe polygonoides Torr. & A.Gray Chorizanthe staticoides Benth. Eriogonum nudum Douglas ex Benth. Eriogonum umbellatum Torr. Eriogonum vimineum Douglas ex Benth. Polygonum arenastrum Jord. ex Boreau Native or introduced N I N ZS eZ, ZZ Z i Ley SP Sees LD Source SR QQQQ00| AQAADAAOAADO0A0M S Q. ZZQQOO © A200 0 are) | 0 | ZQAQAAAG AAQARDAAAAZ OQ | [Vol. 56 Q 2009] Family Polygonaceae Polygonaceae Polygonaceae Polygonaceae Polygonaceae Polygonaceae Polygonaceae Polygonaceae Polygonaceae Polygonaceae Polypodiaceae Portulacaceae Portulacaceae Portulacaceae Portulacaceae Portulacaceae Portulacaceae Primulaceae Primulaceae Primulaceae Primulaceae Pteridaceae Pteridaceae Pteridaceae Pteridaceae Pteridaceae Pteridaceae Pteridaceae Pteridaceae Ranunculaceae Ranunculaceae Ranunculaceae Ranunculaceae Ranunculaceae Ranunculaceae Ranunculaceae Ranunculaceae Ranunculaceae Ranunculaceae Ranunculaceae Ranunculaceae Ranunculaceae - Ranunculaceae ~ Ranunculaceae _ Rhamnaceae Rhamnaceae . Rhamnaceae _ Rhamnaceae - Rhamnaceae Rhamnaceae _ Rhamnaceae ~ Rhamnaceae Rhamnaceae _ Rosaceae _ Rosaceae - Rosaceae _ Rosaceae - Rosaceae WILSON ET AL.: BIODIVERSITY OF PINE HILL APPENDIX 1. CONTINUED. Taxon Polygonum californicum Meisn. Polygonum convolvulus L. Polygonum punctatum Elliot Pterostegia drymarioides Fisch. & C.A.Mey. Rumex acetocella L. Rumex conglomeratus Murray Rumex crispus L. Rumex obtusifolius L. Rumex pulcher L. Rumex salicifolius Weinm. var. denticulatus Torr. Polypodium californicum Kaulf. Calandrinia ciliata (Ruiz & Pav.) DC. var. menziesii (Hook.) J.F.Macbr. Claytonia exigua Torr. & A.Gray Claytonia parviflora Douglas ex Hook. Claytonia perfoliata Donn ex Willd. Montia fontana L. Portulaca oleracea L. Anagallis arvensis L. Centunculus minimus L. Dodecatheon hendersonti A.Gray Trientalis latifolia Hook. Adiantum jordanii C.H.Mull. Aspidotis californica (Hook.) Nutt. ex Copel. Cheilanthes intertexta (Maxon) Maxon Pellaea andromedaefolia (Kaulf.) Fée Pellaea mucronata (D.C.Eaton) D.C.Eaton Pentagramma pallida (Weath.) Yatsk., Windham & E.Wollenw. Pentagramma triangularis (Kaulf.) Maxon Pteridium aquilinum (L.) Kuhn var. pubescens Underw. Aquilegia formosa Fisch. ex DC. Clematis lasiantha Nutt. Delphinium gracilentum Greene Delphinium hansenii (Greene) Greene Delphinium hesperium A.Gray Delphinium patens Benth. Isopyrum occidentale Hook. & Arn. Ranunculus aquatilis L. var. hispidulus E.Drew Ranunculus arvensis L. Ranunculus californicus Benth. Ranunculus hebecarpus Hook. & Arn. Ranunculus hystriculus A.Gray Ranunculus muricatus L. Ranunculus occidentalis Nutt. var. eisenii (Kellogg) A.Gray Thalictrum fendleri Engelm. ex A.Gray var. polycarpum Tort. Ceanothus cuneatus (Hook.) Nutt. Ceanothus integerrimus Hook. & Arn. Ceanothus lemmonii Parry Ceanothus leucodermis Greene Ceanothus palmeri Trel. Ceanothus roderickii Knight Rhamnus californica Eschsch. Rhamunus ilicifolia Kellogg Rhamnus tomentella Benth. ssp. crassifolia (Jeps.) J.O.Sawyer Adenostoma fasciculatum Hook. & Arn. Amelanchier utahensis Koehne Aphanes occidentalis (Nutt.) Rydb. Cercocarpus betuloides Nutt. Chamaebatia foliolosa Benth. Native or introduced N ZZ os ZZ Zz LUZ Zt LZ 2,27, Z Ta Source WV | QQQAQ AAAAH000OO| a]F = Q CQ) SGN AZ Gre) Geer er aeke) QQQQQQ2Z202Z0 QO | 8 q) Q) oletimommaoueiotalma tats 276 MADRONO [Vol. 56 APPENDIX 1. CONTINUED. Native or Family Taxon introduced Source Rock Rosaceae Fragaria vesca L. ssp. californica (Cham. & Schltdl.) N — Staudt Rosaceae Heteromeles arbutifolia (Lind|.) Roem. N WVNSR G Rosaceae Horkelia californica Cham. & Schltdl. ssp. dissita Vv G (Crum) Ertter Rosaceae Horkelia fusca Lindl. ssp. parviflora (Nutt. ex Torr. & Vv G A.Gray) D.D.Keck Rosaceae Malus sylvestris Mall. N ~~ Rosaceae Oemleria cerasiformis (Hook. & Arn.) J.W.Landon N W NG Rosaceae Potentilla glandulosa Lindl. N WN G Rosaceae Potentilla glandulosa Lindl. ssp. reflexa (Greene) VS G D.D.Keck Rosaceae Prunus ilicifolia (Nutt. ex Hook. & Arn.) D.Dietr. N = Rosaceae Rosa californica Cham. & Schltdl. N W NG Rosaceae Rosa eglanteria L. Vv G Rosaceae Rubus armeniacus Focke WVS G Rosaceae Rubus leucodermis Douglas ex Torr. & A.Gray N WN G Rosaceae Rubus ursinus Cham. & Schltdl. N WVNS G Rosaceae Sanguisorba minor Scop. ssp. muricata (Spach ex NV G Bonnier & Layens) Nordborg Rubiaceae Cephalanthus occidentalis L. var. californicus Benth. VNS G Rubiaceae Galium aparine L. N WVNSR G Rubiaceae Galium bolanderi A.Gray N WVSR G Rubiaceae Galium californicum Hook. & Arn. ssp. sierrae N WVR G Dempster & Stebbins Rubiaceae Galium divaricatum Lam. I WV G Rubiaceae Galium murale (L.) All. V NG Rubiaceae Galium parisiense L. I WVNS G Rubiaceae Galium porrigens Dempster N WVSR G Rubiaceae Sherardia arvensis L. I WV G Rutaceae Ptelea crenulata Greene N WVS G Salicaceae Populus fremontii S.Watson WVS G Salicaceae Salix exigua Nutt. WVN G Salicaceae Salix gooddingii C.R.Ball N WVN G Salicaceae Salix laevigata Bebb S G Salicaceae Salix lasiolepis Benth. WVN G Salicaceae Salix lucida Muhl. ssp. lasiandra (Benth.) E.Murray Vv G Salicaceae Salix melanopsis Nutt. VN G Santalaceae Comandra umbellata (L.) Nutt. ssp. californica (Eastw. VS G ex Rydb.) M.Piehl Saxifragaceae Boykenia occidentalis Torr. & A.Gray WVN G Saxifragaceae Darmera peltata (Torr. ex Benth.) Voss V G Saxifragaceae Lithophragma affine A.Gray VN NG Saxifragaceae Lithophragma bolanderi A.Gray VS G Saxifragaceae Lithophragma heterophyllum (Hook. & Arn.) Torr. & N VN G A.Gray Saxifragaceae Lithophragma parviflorum (Hook.) Torr. & A.Gray Vv G Saxifragaceae Philadelphus lewisii Pursh ssp. californica (Benth.) N WVNS G Munz Saxifragaceae Saxifraga californica Greene N WVNS G Scrophulariaceae = Antirrhinum cornutum Benth. N WV G Scrophulariaceae = Antirrhinum vexillocalyculatum Kellogg ssp. breweri Vv NG (A.Gray) D.Thomp. Scrophulariaceae = Castilleja applegatei Fernald Vv G Scrophulariaceae = Castilleja attenuata (A.Gray) T.1.Chuang & Heckard N WVSR G Scrophulariaceae = Castilleja exerta (A.Heller) T.1.Chuang & Heckard VR G Scrophulariaceae = Castilleja foliolosa Hook. & Arn. N WVNSR G Scrophulariaceae = Castilleja lacera (Benth.) T.I.Chuang & Heckard Vv G Scrophulariaceae = Castilleja lineariloba (Benth.) T.1.Chuang & Heckard WV G Scrophulariaceae = Castilleja rubicundula (Jeps.) T.1.Chuang & Heckard N WVN G ssp. lithospermoides (Benth.) T.1.Chuang & Heckard Scrophulariaceae = Castilleja subinclusa Greene VN G Scrophulariaceae = Collinsia heterophylla Buist ex Graham WVNSR G 2009] Family Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Scrophulariaceae Selaginellaceae Selaginellaceae Selaginellaceae Simarubaceae Solanaceae Solanaceae Solanaceae Solanaceae Solanaceae Solanaceae Solanaceae Solanaceae Sterculiaceae Styracaceae Tamaricaceae Typhaceae Typhaceae Urticaceae Valerianaceae Valerianaceae Valerianaceae Verbenaceae Verbenaceae Verbenaceae WILSON ET AL.: BIODIVERSITY OF PINE HILL APPENDIX 1. CONTINUED. Taxon Collinsia sparsiflora Fisch. & C.A.Mey. var. bruceae (M.E.Jones) Newsom Collinsia sparsiflora Fisch. & C.A.Mey. var. collina (Jeps.) Newsom Collinsia tinctoria Hartw. ex Benth. Cordylanthus pilosus A.Gray ssp. hansenii (Ferris) T.I.Chuang & Heckard Grateola ebracteata Benth. Keckiella breviflora (Lindl.) Straw Keckiella lemmonii (A.Gray) Straw Kickxia elatine (L.) Dumort. Linaria canadensis (L.) Dum.Cours. var. texana (Scheele) Pennell Lindernia dubia (L.) Pennell var. anagallidea (Michx.) Cooperr. Mimulus aurantiacus W.Curtis Mimulus cardinalis Benth. Mimulus douglasii (Douglas ex Benth.) A.Gray Mimulus guttatus DC. Mimulus kelloggii (Curran ex Greene) Curran ex A.Gray Mimutlus layneae (Greene) Jeps. Mimutlus pilosus (Benth.) S.Watson Mimulus tricolor Hartw. ex Lindl. Pedicularis densiflora Benth. ex Hook. Penstemon azureus Benth. Penstemon heterophyllus Lindl. Scrophularia californica Cham. & Schltdl. Scrophularia californica Cham. & Schltdl. ssp. floribunda (Greene) R.J.Shaw Triphysaria eriantha (Benth.) T.I.Chuang & Heckard Triphysaria pusilla (Benth.) T.I.Chuang & Heckard Verbascum blattaria L. Verbascum thapsus L. Veronica arvensis L. Veronica peregrina L. ssp. xalapensis (Kunth) Pennell Veronica persica Por. Selaginella douglasii (Hook. & Grev.) Spring Selaginella hanseni Hieron. Selaginella wallacei Hieron. Ailanthus altissima (Mall.) Swingle Datura stramonium L. var. tatula (L.) Torr. Datura wrightii Regel Nicotiana acuminata (Graham) Hook. var. mu/tiflora (Phil.) Reiche Nicotiana attenuata Torr. ex S.Watson Nicotiana glauca Graham Nicotiana quadrivalvis Pursh Solanum americanum Mill. Solanum xantii A.Gray Fremontodendron californicum (Torr.) Coville ssp. decumbens (R. Lloyd) Munz Styvrax officinalis L. var. redivivus (Torr.) Howard Tamarix parviflora DC. Typha domingensis Pers. Typha latifolia L. Urtica dioica L. ssp. holosericea (Nutt.) Thorne Plectritis ciliosa (Greene) Jeps. Plectritis macrocera Torr. & A.Gray Valerianella locusta (L.) Laterr. Phyla nodiflora (L.) Greene var. nodiflora Verbena bonariensis (A.DC.) A.Gray Verbena hastata L. Native or introduced N Source WVN V WVN WVNS gzZredd FIG. 1. [Vol. 56 Sidalcea gigantea. A. Plant base, habit, with lower and mid-stem leaves. Scale 1 cm. B. Stem base, enlarged. Scale | cm. C. Plant apex, inflorescence, habit with reduced leaves. Scale 2 cm. D. Pulvinus at apex of petiole and attachment to blade. Scale 2 cm. E. Stipules and attachment. Scale 5 mm. F. Mericarp, 3 views: 1) lateral, 2) back, 3) top. Scale 1 mm. G. Fruiting calyx, side view. Scale 1 cm. H. Calyx lobe enlarged. Scale 5 mm. I. Petal. Note: veins are actually pale — pattern is indicated. Scale 1 cm. J. Stamen column, styles of pistillate flower. Scale 5 mm. A, B, E drawn from Janeway & Castro 7696; C, D, I, J drawn from Janeway & Castro 7698; F, G, H drawn from D. W. Taylor 15370. Mlustration by Steven R. Hill. scattered (not dense) simple or 2-4-rayed (6- rayed on lower surface) stellate hairs to 0.5 mm long, gradually reduced in size towards stem apex, in the inflorescence greatly reduced and irregularly 2—3 lobed; inflorescence compound or simple, each branch an interrupted spike-like raceme generally 10-25 cm long, axis glaucous, with minute tufted stellate hairs, bracts single, tardily deciduous, narrowly lanceolate, often bifid, 2.5—-3.0 * 0.7 mm, little longer than the pedicel, much shorter than calyx, canescent, attached with pedicel to raised pad on stem; species gynodioecious, flowers pistillate with aborted anthers, or bisexual; flowers on all plants generally 10-20 per branch, well-separated (about 5 mm apart) on stellate canescent pedicels 2-3 (-5) mm long, these longest below; mature calyx (5—) 6-8 mm long, fused at the base, lobes 5, 3-veined, (3.5—) 4-5 mm long, densely stellate on outside (surface + obscured), the inner surface glabrous except for the similarly pubescent inner lobe margins; petals 5, 14-20 mm long < 7-8 mm wide in bisexual flowers, 7-9 mm X 4—5 mm in pistillate flowers, the claw densely fringed with hairs; the staminal column stellate-puberulent, 5— 6 mm long in pistillate flowers, to 8 mm in bisexual flowers; stigmas 6-8, in pistillate flowers extended 4 mm beyond aborted anthers, not. extended beyond anthers in bisexual flowers, | sometimes poorly developed; mericarps 6-8, 3 mm long X 3.0-4.0 mm tall * 2-3 mm thick | (nearly equally sided), surface conspicuously | 2009] CLIFTON ET AL.: SIDALCEA GIGANTEA ZS Fic. 2. Sidalcea gigantea colony at the head of Little Grizzly Creek, Plumas Co., CA, showing colonial habit and stature of plants. Scale divisions in decimeters. Plants toward rear of photo are upslope; rod shows stature of immediately adjacent plants. Voucher: Janeway & Castro 7698 (CHSC, VT), 11 August 2002. Photograph by Lawrence Janeway. reticulate-alveolate dorsally and on outer margins of sides, less so towards axis, margins bulging, back with obvious groove, a few minute glandu- lar and/or stellate hairs on top surface, apical cusp about 1.0 mm long with a few minute simple hairs at tip; seed smooth, glabrous, dark brown, about 1.5 mm X 1.3 mm X 2 mm tall, often with a few minute hairs near hilum. The species produces an extensive rhizome system and often occurs in dense colonies of hundreds or thousands of stems 1n an area as small as 40 m°. Stems are more scattered where the Species occurs at forest margins or on drier roadbanks (where there 1s still significant moisture nearby). Several populations have been discovered that extend to 10-20 acres (D. W. Taylor, botanical consultant, personal communication). The chromosome number is unknown. This species normally flowers from mid-July to September (rarely to October), with peak flower- ing approximately 25 July—l10 August. Representative specimens: USA. CALIFOR- NIA. Butte Co.: Flea Valley, T23N R4E S24, 3 Jul 1999, G. Clifton 36019 (JEPS); ca. 4.8 mi S of Inskip at water supply pond for Sterling City T24N R4E S21, 31 Jul 1980, R. A. Schlising & R. Banchero 3940 (CHSC); 1.1 mi S of Inskip along Skyway, 31 Jul 1980, R. A. Schlising & R. Banchero 3941 (CHSC); NE of Coyote Gap on Four Trees Rd T23N R2E S36, 11 Jul 1981, R. A. Schlising et al. 4140 (CHSC); near N end of Concow Reservoir T22N R4E S10, 8 Aug 1981, R. A. Schlising & C. A. Lawler 4190 (CHSC); N and E of De Sabla and Lovelock at “Doon MADRONO [Vol. 56 Pigs: Yuba Co., CA. Photograph by Roy E. Buck. Camp” just N of Little West Fork of Feather River T24N R4E S31, 26 Aug 1982, R. A. Schlising 4374 (CHSC); Road 28N35 at a branch of Wildcat Creek ca. 1 mi N of Four Trees- Coyote Gap Rd, 11 Aug 1998, V. Oswald & L. Ahart 9466 (UC ); along Fall River, W of the bridge across Fall River, ca. 6 (air) mi NE of Feather Falls T2IN R7E S827, 22 Jul 1993, L. Ahart 7083 (UC); West Branch of Wildcat Creek, about 2 mi NE of Coyote Gap, T23N R6E S30, 3 Oct 1993, L. Ahart 7269 (CHSC, JEPS); Four Trees Rd 0.6 km N of Coyote Gap, T23N R5E $35, 11 Aug 2002, L. P. Janeway & B. Castro 7696 (CHSC, JEPS, VT); along Road 24N13 crossing an uppermost small branch of Locker- man Creek 0.6 km SSW of Logue Meadows T24N RSE S833, 4 Sep 2005, L. P. Janeway 8534 (CHSC); at Concow Rd crossing of uppermost Keyser Creek 0.8 km ENE of Bald Mountain summit, T24N R4E S12, 12 Aug 2006, L. P. Janeway 8838 (CHSC); ca. 2 air mi NE of Stirling City, Plumas Natl. Forest, 6 Aug 1995, D. W. Taylor 15379 (JEPS, UC); ca. 3 air mi NE of Stirling City, near Jackass Flat, Plumas Natl. Forest, 6 Aug 1995, D. W. Taylor 15380 (JEPS, UC); ca. 6 air mi E of Stirling City, Keyser Creek, ca. 1.5 mi S of Bald Mountain summit, Lassen Natl. Forest, 6 Aug 1995, D. W. Taylor & R. Fallscheer 15381 (JEPS ); ca. 2 air mi N of Stirling City, Reston Rd, ca. 1.5 mi SW of Bald Mountain summit, Lassen Natl. Forest, 6 Aug 1995, D. W. Taylor & R. Fallscheer 15389 (JEPS, UC ); near Concow Reservoir (ca. 6 air mi E of Portion of inflorescence of Sidalcea gigantea showing open pistillate flowers. Hampshire Creek watershed, Paradise) along Hoffman Rd 0.3 mi W of Concow Rd, Lassen Natl. Forest, 6 Aug 1995, D. W. Taylor & R. Fallscheer 15397 (ILLS); Four Trees Rd (Forest Road 23N00) ca. 0.3 mi N of Coyote Gap, 22 Jul 2004, D. W. Taylor et al. 19207 (JEPS); Skyway 1.4 mi S of Inskip [‘Ins- keep’], 12 Jul 1984, V. Oswald 1569 (CHSC ); N side of Oroville-Quincy Hwy ca. | mi S of Butte — Plumas Co. line, 21 Jul 1978, M. S. Taylor 1834 (CHSC, MO); road to Bald Mt. ca. % mi SW of Bald Mt., ca. 3 mi NE of Stirling City, 16 Aug 1982, M. S. Taylor 4996 (CHSC, MO). Plumas Co.: N side of road to dam across Slate Creek ca. 100 yds N of Scales Rd ca. 5.5 mi due NE of. Strawberry Valley, 23 Oct 2006, L. Ahart 13461 (CHSC); along the Oroville-Quincy Highway | paralleling Little Grizzly Creek west of Bucks | Lake, 11 Aug 1998, V. Oswald 9482 (CHSC); Grizzly Forebay at Bucks Lake T24N R6E S34, | 31 Jul 1975, F. T. Griggs & A. Pass 273 (CHSC); | Oroville-Quincy Highway 1.1 mi W of Little | Grizzly Creek, T23N R7E S7, 15 Aug 1996, L. Po | Janeway 5069 (CHSC, UC); Forest Service Rd _ 23N18 at Little North Fork Middle Fork Feather | River, 20 Aug 1996, L. P. Janeway 5075 (CHSC, | UC); Oroville-Quincy Highway (and into adja- | cent meadow) 0.5 km W of Road 23N73Y at head of Little Grizzly Creek, 12 Aug 2002, L. P. | Janeway & B. Castro 7698 (CHSC, VT); between Road 24N13 and a NE ttributary to upper Lockerman Creek ca. 1.3 km S of Logue. Meadows T23N RSE S3, 4 Sep 2005, L. P.. Janeway 8532 (CHSC); along Road 24N13 along | 2009] the E side of Lynch Meadows T24N RSE S27, 4 Sep 2005, L. P. Janeway 8535 (CHSC); Oroville- Quincy Highway 0.5 mi W of Forest Service Rd 23N73Y, 16. Aug: 1999; Bs G. Baldwin 1089 (JEPS); Onion Creek headwaters ca. 0.5 air mi SW of American House, along Forest Rd 2INOLY ca. 0.3 mi W of junction Forest Rd 21N99, 2 Aug 2001, D. W. Taylor 17878 (JEPS). Shasta Co.: Goose Valley, ca. 4 air mi NW of Burney, Shasta-Trinity Natl. Forest, W margin of valley, 2 Jul 1995, D. W. Taylor 15210 (JEPS, UC ); Goose Creek at its debouchment into Goose Valley (ca. 5 mi NW of Burney), 30 Jul 1995, D. W. Taylor 15370 (ALLS); Highway 299 about 3 mi W of Hatchet Mountain Pass, tributary to Hatchet Creek 1 mi W of Carberry Flat, 30 Jul 1995, D. W. Taylor 15372 (ILLS); Hwy 299 2.1 km W of Hatchet Mountain Summit, 8 Sep 1998, V. Oswald & L. Ahart 9568 (CHSC, JEPS): Hatchet Mountain Pass Hwy 299 T35N R2E S30, 3 Jul 1999, G. Clifton 36018 (JEPS); South Fork Digger Creek, 0.8 km W of Heart Lake, 24 Jul 1997, D. W. Taylor 16151 (UC). Sierra Co.:, bog on the N side of Forest Road 20N35 ca. 0.5 mi E of intersection with County Rd 690 (about | air mi NE of Union Hill), 2 Aug 2001, D. W. Taylor 17889 (JEPS ). Yuba Co.: meadow along Forest Rd 35 about 2 road mi SW from Union Hill, 2 Aug 2001, D. W. Taylor 17880 (JEPS); ca. 1.7 air mi S of Sly Creek Reservoir and 5 mi E of Woodleaf along La Porte Rd 0.1 mi E of Travis Saddle [topotype], 30 Jul 1998, B. Castro & L. Gehrung &18 (CHSC). DISTRIBUTION AND HABITAT Sidalcea gigantea is found on the western slope of the northern Sierra Nevada in extreme northeastern Yuba, western Sierra, eastern Butte and western Plumas Counties, north to the extreme southern Cascades in east-central Shasta Co., California. The southernmost known local- ity is near Hampshire Creek east of Clipper Mills (Butte Co.) in Yuba Co. The northernmost known locality is in Goose Valley in east-central _ Shasta Co. The north-south extent of its range is approximately 160 km. Although it is notably unknown in Tehama Co. in the middle of its range, much of that area is remote and rugged and additional populations may eventually be _ found in the intervening areas. Sidalcea gigantea occurs at elevations between _ approximately 670-1950 m. The species does not _ appear to have very specific substrate preferences, as it occurs on a variety of sedimentary, igneous, and metamorphic substrates, including Paleozoic marine sedimentary rocks, Mesozoic granitic rocks, Mesozoic ultrabasic intrusive rocks, Pleis- _ tocene volcanic andesite, Pliocene volcanic basalt, _ Paleozoic metavolcanics, Jurassic-Triassic meta- volcanics, Pre-Cretaceous metavolcanics, and CLIFTON ET AL.: SIDALCEA GIGANTEA 209 Quaternary glacial deposits (Lydon et al. 1960; Burnett and Jennings 1962). Sidalcea gigantea is found at sites that are relatively mesic but that are usually adjacent to wetter areas, and it can grow in the open or in moderate shade. Habitats on labels have included locations near seeps or springs, sphagnum fens, streams, wet meadows, and shaded banks in middle to upper montane coniferous forest (especially Mixed Conifer Forest), as well as in drier edges of meadows. It has been collected most often in moderately moist meadows and around the margins of wet meadows dominated by a variety of herbaceous species, and in ecotonal areas between meadows and montane coniferous forests dominated by Pinus ponderosa C. Lawson, Calocedrus decurrens (Yorr.) Florin, Pseudotsuga menziesii (Mirb.) Franco, Abies concolor (Gordon & Glend.) Lindl. ex Hildebr., and Quercus kelloggii Newb. Occasionally it occurs in Openings in and around the margins of red fir forest dominated by Abies magnifica A. Murray, aspen forest dominated by Populus tremuloides Michx., or in riparian forest domi- nated by species such as Salix lasiolepis Benth. and Acer macrophyllum Pursh. It sometimes occurs along stream banks and it sometimes grows in disturbed areas, including roadside ditches, road banks, and graded areas, though with a stream or meadow nearby. IDENTIFICATION AND TAXONOMIC RELATIONSHIPS Sidalcea gigantea when well-developed is easily distinguished from the other species of Sidalcea. It is the one of the largest in stature of all of the species, with stems to 2.5 m tall. It is also distinctive in producing an extensive rhizome system that sometimes gives rise to dense, concentrated colonies of hundreds or even thousands of stems (Fig. 2), although stems are sometimes more scattered or limited. Additional characters, in combination, also serve to distin- guish S. gigantea from the other taxa of Sidalcea, including: 1) long (to 0.6 m), thick (to 1.0 cm in diameter) rhizomes that are bristly-pubescent when young; 2) stem bases with many long, retrorse, bristly hairs; 3) thick (to 1.4 cm in diameter), hollow stems; and 4) fruit segments that are almost equally three-sided. This new species has been confused most frequently with Sidalcea malviflora subsp. celata (Jeps.) C. L. Hitche. (=S. celata (Jeps.) S. R. Hill) and S. malviflora subsp. asprella (Greene) C. L. Hitche. (=S. asprella Greene), both of which were included within S. ma/viflora subsp. asprella by Hill (1993) and are now treated as distinct species (Hill 2009). Several Sidalcea species can reach heights of more than | m, and a few tend to reach 1.5 m and above (Hitchcock 1957), 290 including S. campestris Greene (Willamette Val- ley, OR), S. candida A. Gray (CO, NM, UT, WY), S. cusickii Piper (western OR), S. mala- chroides (Hook. & Arn.) A. Gray (coastal CA and OR), S. neomexicana A. Gray (ID and eastern OR to CA, CO, south to northern Mexico), S. oregana subsp. spicata (Regel) C. L. Hitche. (CA, OR, NV), and S. oregana subsp. valida (Greene) C. L. Hitche. (Sonoma Co., CA). Some of these have stems that can infrequently attain 1 cm in diameter and are somewhat hollow at this size, but their other characters do not match those of the new species. Schlising (1987) treated specimens now identified as S. gigantea as S. malviflora subsp. celata, but Oswald (2002) recognized them as an unnamed taxon. As interpreted by Oswald (2002) and as currently recognized, S. celata (=S. malviflora subsp. celata) is restricted to a small area of southwest- ern Shasta Co. and adjacent northern Tehama Co. within the Inner North Coast Ranges. Its range does not overlap that of S. gigantea. Sidalcea asprella, as recognized by Hill (2009), is more widespread, ranging at least from the central Sierra Nevada in Mariposa Co. north- ward to southwestern Oregon (including portions of the northern Sierra Nevada, southern Cascade Ranges, and Klamath Ranges), a range that either includes or is adjacent to that of S. gigantea. Within the area of overlapping range, S. asprella can be distinguished from S. gigantea by its shorter stature (normally <1 m), its ascending-supported, often decumbent-based stems, generally by its less free-rooting short rootstocks or rhizomes generally <10 cm (to 11— 30 cm) long, and by a stem base with minute or larger stellate (not simple) hairs. An early monograph by Roush (1931) would place this new species within Sidalcea Subgenus Eusidalcea Section Perennes. A revised treatment of Sidalcea in North America (north of Mexico) is being produced for the Flora of North America series by one of the co-authors of this article (Hill) and a discussion of species alignments will MADRONO [Vol. 56 be presented there along with a revised nomen- clature. Most recently, S. gigantea has been grouped with putatively related taxa within an informal ‘asprella or ‘celata’ group (Andreasen and Baldwin 2001, 2003a, b). According to Katarina Andreasen (Uppsala University, per- sonal communication), who along with Bruce Baldwin has studied the DNA of most of the species of Sidalcea (Andreasen and Baldwin 2001), ““There are essentially no differences between the [molecular] sequences in this ‘ce/ata’ group’. Sidalcea gigantea was discussed again in subsequent molecular phylogenetic studies of Sidalcea that utilized internal and external transcribed spacers (ITS and ETS) of 18S—26S nuclear ribosomal DNA (Andreasen and Bald- win 2003a, b). Among the specimens studied by Andreasen and Baldwin were four samples (Baldwin 1089, Clifton 36018, Clifton 36019, all at JEPS, and Oswald 9466 at UC) of the then unnamed S. gigantea, and they could not be separated significantly from samples of other species in this grouping, including S. asprella, based on their DNA sequences (Andreasen and Baldwin 2003a). This study placed S. gigantea in a well-supported “‘aspre/lla clade” that also included S. asprella and S. celata (cited in the paper as ‘asprella 4’), as well as what they called S. maxima M. Peck, S. hirtipes C.L. Hitche., S. campestris Greene, and a putative hybrid, S. asprella X S. oregana (Torr. & A. Gray) A. Gray subsp. oregana. These studies also placed the “asprella clade” within a larger “glaucescens clade” that contained, in addition to the above taxa, S. glaucescens Greene, S. multifida Greene, S. robusta A. Heller, and another sample of S. asprella. The following key serves to distinguish Sidal- cea gigantea and the other perennial Sidalcea taxa found within or near its range in northeastern California (including the following counties: Butte, Nevada, Plumas, Shasta, Sierra, Tehama, and Yuba). Names used are according to Hill (2009). 1. Mericarp mucro 0; calyx bractlets 3; leaves generally evenly arrayed on stem both early and late in season (not mostly on lower stem or basal), similar in shape and unlobed [Nevada Co. only]............. ee aE a ee eee S. stipularis J. T. Howell & True 1’ Mericarp mucro | (occasionally minute); calyx bractlets 0; some leaves located on lower stem or stem base at least early in season (or not), occasionally rosette-like, basal leaves crenate, leaves above generally deeply lobed = 2. Plants (1—)1.5—2(—2.5) m tall, with rhizomes to (0.5—)1 cm in diameter, 4-6 dm long, bristle-hairs reflexed; stem base erect, hollow, with dense reflexed bristles; mericarp + equilateral.................... N dak pets ake S. gigantea G. Clifton, R. E. Buck & S. R. Hill | Plants generally less than 1(—1.5) m tall, with or without rhizomes, if present these generally less than 1 cm in diameter; stems solid (hollow), hairs variable, occasionally bristle-like or lacking; mericarp equilateral or not 3. Mericarp narrowly wing-margined, generally glabrous; stem erect, glabrous, glaucous above; [Butte Co. only] ant Med ene Ee ee S. robusta Heller ex E. M. Roush | 2009] CLIFTON ET AL.: SIDALCEA GIGANTEA 3’ Mericarp not wing-margined, generally pubescent; stem erect or not, usually pubescent (sometimes sparsely so), glaucous or not Stems long-prostrate, free-rooting, base with bristle-hairs 2—3 mm long, 0 stellate; leaf hairs simple; mericarps densely bristly stellate-puberulent on top, back, and/or mucro; [Nevada Co. UTE CSO Ul Us| Mare eae a Psat Ma eek ARCA, Ai Sea, eee nia hte Wie, Guan ne: ENN ART ot By aot Mocs S. reptans Greene Stems not long-prostrate nor very elongate and free-rooting, more compact, base with shorter bristles or hairs stellate; leaf hairs various; mericarps sparsely puberulent to nearly glabrous 5. Stems ascending to sprawling, not strictly erect from base; plant height usually 0.5 m or less 4. 4’ 6. Stems not rooting at base; plants very glaucous; stem base glabrous to sparsely stellate-pubescent; basal leaves few, generally 5-lobed; mericarps glandular-pub- PSUGLUN (co) OL UR eR eR een ae Oe ee oer ee eC es era S. glaucescens Greene Stems commonly rooting at base; plants scarcely if at all glaucous; stem base stellate- pubescent; most leaves basal, generally 7-lobed; mericarps with few minute bristles RMI Te Tieton ties aegis Mate eae AoE S. asprella Greene subsp. nana (Jeps.) S. R. Hill 5° Stems usually erect from base; plant height 0.8 m or more ve Stems several, clustered, erect and not rooting at base; flowers relatively small; inflorescence not secund, often congested; petals of bisexual flowers 1.0—1.5 cm long 8. Open flowers and fruits often not overlapping on axis; calyx hairs uniformly stellate; bud bracts equal to or shorter than buds. ................... S. oregana (Torr. & A. Gray) A. Gray subsp. oregana 8’ Open flowers and fruits usually overlapping on axis; calyx hairs uniformly stellate or often with some bristles; bud bracts usually longer than buds ........... eae a S. oregana subsp. spicata (Regel) C. L. Hitche. 7 Stems few or one, erect or ascending, rooting at base or not; flowers larger; inflorescence often secund, not congested; petals of bisexual flowers 2—2.5 cm long 9. Plant with caudex, stem bases not rooting, rhizome 0; stem generally erect, free- standing, base with stiff reflexed bristle- (coarse few-rayed stellate) hairs; leaves lobed, differing in shape, uppermost with linear entire or semi-entire lobes; [Shasta, Tehama cos. only] ee Ore ree any eee ee S. celata (Jeps.) S. R. Hill 9° Plant with caudex or not, gen with short rooting stem bases or rhizomes <10 cm (to 30 cm); stems often weak, supported by other plants, base with minute or larger stellate hairs; leaves generally lobed and all similar in shape, lobes generally’ toothed; [more widespread, Humboldt to Mariposa cos.]................. RARITY AND CONSERVATION STATUS Sidalcea gigantea is endemic to California. Colonies are generally local, and although more than forty collections are known, many are from the same few colonies. Sidalcea gigantea also appears to be a species with ecologically special- ized microhabitat preferences that are probably largely determined by a specific soil moisture regime. Observations also suggest that the species increases or becomes more robust after fire (D. W. Taylor, personal communication). Fires may be, at least in part, responsible for its occurrence in localized populations. Suitable habitat is widely scattered but limited in extent within the species’ range, but there appears to be additional apparently suitable habitat within its range from which the species is absent. Most of the known localities of Sidalcea gigantea are relatively undisturbed. The species appears to tolerate limited surface disturbances that do not substantially damage the rhizome system. It can colonize some roadside areas, but _ these plants are somewhat depauperate and there is no evidence that the species is tolerant of extensive or prolonged disturbance that results in significant soil disruption. It is likely to appear after fires from the soil seed bank. Logging _ Operations and associated road building or road maintenance and improvement near moist drain- S. asprella Greene subsp. asprella ages, especially where plants are near road margins, could threaten some populations. Much additional study is needed. Given the known distribution of Sidalcea gigantea, the remoteness of many known locali- ties, and the nature and level of potential threats to the species, this species should not be considered endangered at present. It is, however, uncommon and local enough that its status should be monitored. We recommend that S. gigantea be considered for inclusion on List 4 (Plants of Limited Distribution—A Watch List) of the California Native Plant Society’s (CNPS) Inventory of Rare and Endangered Vascular Plants of California (CNPS 2001, 2008). ACKNOWLEDGMENTS We would like to thank the curators of the herbaria that generously shared their specimens with us, including (but not limited to) CAS, CHSC, JEPS, MO, and UC. We would also like to thank Linnea Hanson, Plumas National Forest botanist, and William Davilla of EcoSystems West for their support. We also thank two anonymous reviewers for comments which greatly improved the manuscript. Special thanks go to Dean Wm. Taylor for sharing his information on this new species as well as his many specimens of it. We also especially thank Lawrence Janeway for sharing his knowledge of, and specimens of, this new species. We have enjoyed his enthusiasm and encouragement to 292 work on this and other checker mallows on which he has also conducted considerable field and herbarium work. A portion of this research was supported by a grant from the Lawrence R. Heckard Endowment Fund of the Jepson Herbarium (to Hill) which allowed travel to and research at the herbaria of the University of California, Berkeley, (JEPS, UC) and the California Academy of Sciences (CAS) in March 2008. A continuing grant from the Illinois Department of Transportation is gratefully acknowledged (by Hill) for financial support towards research for these studies. LITERATURE CITED ANDREASEN, K. AND B. G. BALDWIN. 2001. Unequal evolutionary rates between annual and perennial lineages of checker mallows (Sidalcea, Malvaceae): evidence from 18S—26S rDNA internal and external transcribed spacers. Molecular Biology and Evolu- tion 18:936—-944. AND . 2003a. Reexamination of relation- ships, habital evolution, and phylogeography of checker mallows (Sidalcea; Malvaceae) based on molecular phylogenetic data. American Journal of Botany 90:436—-444. AND . 2003b. Nuclear ribosomal DNA sequence polymorphism and hybridization in checker mallows (Sidalcea, Malvaceae). Molecular Phylogenetics and Evolution 29:563—581. BURNETT, J. L. AND C. W. JENNINGS. 1962. Geologic map of California: Chico sheet. Scale 1:250,000. California Division of Mines and Geology, Sacramento, CA. CALIFORNIA NATIVE PLANT SOCIETY. 2001, Inventory of rare and endangered plants of California, 6th ed. MADRONO [Vol. 56 Rare Plant Scientific Advisory Committee, David P. Tibor, Convening Editor. California Native Plant Society, Sacramento, CA. . 2008. California Native Plant Society inventory of rare and endangered plants. Online edition. Version 7-08b, 2 April 2008. Website: http://cnps. web.aplus.net/cgi-bin/inv/inventory.cgi [accessed 12 March 2010]. HILL; S, R. 1993..Sidalcea’ Pp. 755-700. in J. 3G: Hickman (ed.), The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. . 2009. Notes on California Malvaceae including nomenclatural changes and new additions to the flora. Madrono 56:104—-111. HITCHCOCK, C. L. 1957. A study of the perennial species of Sidalcea. Part I. Taxonomy. University of Washington Publications in Biology 18:1—79. LypDon, P. A., T. E. GAY, AND C. W. JENNINGS. 1960. Geologic map of California: Westwood sheet. Scale 1:250,000. California Division of Mines and Geology, Sacramento, CA. OSWALD, V. H. 2002. Selected plants of Northern California and adjacent Nevada. Studies from the Herbarium, Number 11. California State Universi- ty, Chico, CA. RousH, E. M. F. 1931. A monograph of the genus Sidalcea. Annals of the Missouri Botanical Garden 18:117—244. SCHLISING, R. A. 1987. Malvaceae of Butte County, California. Studies from the Herbarium, California State University, Number 5. California State University, Chico, CA. MADRONO, Vol. 56, No. 4, pp. 293-295, 2009 NOTEWORTHY COLLECTIONS CALIFORNIA DASYA_ SESSILIS Yamada (DASYACEAE).—San Diego Co., attached to floating docks at 4000 Coronado Bay Rd., Coronado Cays, Coronado Bay, San Diego, 32°37'50.75’N, 117°08'00.75"W, _ thalli tetrasporic, 28 December 2008, J. R. Hughey s. n. (UC 1944736, UC 1944737); same location, thalli cystocarpic, 28 November 2009, J. R. Hughey s. n. (UC 1944757); Orange Co., attached to vertical faces of concrete pilings from 0 to 3 m and on floating docks, Huntington Harbor, Huntington Beach, 33°43'20. 748"N, 118°03'21.8874’W, thalli tetrasporic, 23 August 2006, A. Lyman s. n. (UC 1944739); Los Angeles Co., attached to vertical sides of concrete pilings and wood on floating docks, Long Beach, 33°46'36.012’N, 118°12'37.8354"W, thalli tetrasporic, 26 February 1976, R. Setzer s. n. (AHFH 81053 in UC [UC 1846061]) as D. pedicellata var. stanfordiana (Farlow) Dawson); same location, 24 August 2006, A. Lyman s. n. (UC 1944738). Previous knowledge. Native to Japan (Y. Yamada 1928, Scientific reports of the Tohoku Imperial University 4:497—-534), type locality: Asamushi, Fu- tago-Jima, Oshima, Benten-Jiwa, Pacific. Also reported from Korea (Y. Lee and S. Kang 2001, A catalogue of the seaweeds in Korea, pp. 1-662) and the Philippines (P. C. Silva et al. 1987, Smithsonian Contributions to Marine Sciences 27:1—179). Introduced to France via oyster importation (M. Verlaque 2002, Phycologia 41:612-618), Spain via mussel cultivation (V. Pena and I. Barbara 2006, Anales del Jardin Botanico de Madrid 63:13—26), and Portugal, by an unknown vector (R. Araujo et al. 2009, Botanica Marina 52:24-46). Fertile tetrasporophytes of the invasive D. sessilis were collected in abundance in France from June to December and in Spain year round. Sexual plants from France and Spain were reported from March to October. Significance. First report of D. sessilis from the northeastern Pacific. Specimens from San _ Diego, Huntington Harbor, and Long Beach were in good morphological and anatomical agreement with illustra- tions of this species (Yamada, Joc. cit.; Verlaque, Joc. cit.; Pena and Barbara, Joc. cit.). The tetrasporophytic plants from San Diego were conspicuous, growing up to 15 cm high, forming bushy thalli with fuzzy axes and laterals. Dasya sessilis is distinguishable from the native species, D. sinicola var californica (Gardner) Dawson, by having wider axes (2+ mm vs. 0.5—1.4 mm). Identification was confirmed by analysis of part of the cytochrome oxidase subunit 2 and 5S RNA genes, and the cox2—cox3 intergenic spacer of the mitochondrial genome on a specimen from San Diego (GenBank GU473263). The resulting 198 bp (base pair) fragment was identical in sequence to the invasive Mediterranean specimen of D. sessilis, but differed from D. baillouviana by 12 bp. A sequence from a Dasya specimen from Coyote Point, San Francisco Bay (GenBank ~GU473264) differed from D. sessilis by 6 bp. Additional confirmation was obtained from the ribulose-1l, 5- biphosphate carboxylase/oxygenase (rbcL) gene from the Huntington Beach and Long Beach specimens which was identical to D. sessilis from San Diego (GenBank GU473265—GU473267). GRATELOUPIA LANCEOLATA (Okamura) Kawaguchi (HALYMENIACEAE).—San Diego Co., attached to floating docks at 4000 Coronado Bay Rd., Coronado Cays, Coronado Bay, San Diego, 32°37'50.75"N, 117°08'00.75"W, thalli tetrasporic and _ sterile, 28 December 2008, J. R. Hughey s. n. (UC 1944744): same location, 27 July 2009, J. R. Hughey s. n. (UC 1944743, UC 1944742); Ventura Co., on docks, Port Hueneme, 34°08'59.964"N, 119°12'36.0354”"W, thallus tetrasporic, 26 July 2006, A. Lyman s. n. (UC 1944741); San Francisco Co., attached to floating docks at the San Francisco Marina, San Francisco, 37°47'14.37’"N, 122°26'18.90"W, thalli tetrasporic, procarpic, and sterile, 15 September 2009, K. A. Miller s. n. (UC 1944746-UC 1944748). Previous knowledge. Native to Japan (K. Okamura 1934, Icones of Japanese algae 7:42), type locality: Ku Province, Enoshima, Tateyama, Kazusa, Pacific. Re- ported from the Mediterranean, Thau Lagoon, France (M. Verlaque 2001, Oceanologica Acta 24:29-49) where it was likely introduced in the 1970s with the importation of the Japanese oyster (M. Verlaque et al. 2005, Phycologia 44:477—496). Specimens of G. lanceo- lata from France were collected from March to December, and reported to have established large, reproductive populations (Verlaque et al., Joc. cit.). Recently, fertile material of G. lanceolata was discov- ered in southern California at Santa Catalina Island in the spring of 2003 and 2008, and in central California at the mouth of Elkhorn Slough in Moss Landing in May, June, and July of 2008 (K. A. Miller et al. 2009, Phycological Research 57:238—241). Miller et al. (/oc. cit.) speculated that mariculture of oysters played a role in the introduction of G. lanceolata, which might be growing cryptically at other sites in California, Oregon, and Washington. Significance. Discovery of G. lanceolata at three new localities in California. Thalli of G. lanceolata exhibited considerable morphological variation depending on age and collection locality. The mature specimens collected in San Diego in December were approximately 20 cm high and 3 cm wide, with uniformly dark red blades that forked once or twice above, and lacked epiphytes. Juvenile specimens from San Diego collected in July matched the description above, however mature thalli were pigmented brown to black with fully to partially eroded blades, and covered with diatoms and bryozo- ans. The fronds from Port Hueneme were dark red and expanded to 20 cm high and 5 cm wide, linear to lobate in shape with a few marginal proliferations, and covered with epiphytic bryozoans. Mature thalli collected from San Francisco were 100 cm high and 40 cm wide and uniformly dark red brown, bullate and with proliferous margins, and often found covered with non-native tunicates. Grateloupia lanceolata is distinguishable in harbors and bays from the native species in California, G. californica Kylin, by having thicker blades at maturity (600 um-l1 mm vs. 400-700 um), a darker color (dark red vs. greenish-brown), and medullary filaments that run anticlinally and periclinally in 294 direction vs. predominantly periclinally in G. califor- nica. Identification was confirmed using Internal Transcribed Spacer-1 (ITS-1) sequences. The ITS-1 sequence (GenBank GU339499) obtained from G. lanceolata from San Diego differed by one nucleotide from eight identical sequences of G. lanceolata from other localities: Port Hueneme (GenBank GU339500), San Francisco (GenBank GU339501—GU339503), Thau Lagoon, France (GenBank AF412010 and AF412011), Santa Catalina Island (GenBank FJ013039), and Moss Landing (GenBank FJ013040). Although oyster spat was implicated in past introduc- tions of G. lanceolata, it is unknown at this time if mariculture, international shipping via ballast water, or hull fouling by coastal shipping vessels are the responsible vectors. These collections support the speculation of Miller et al. (Joc. cit.) that G. lanceolata was introduced to harbors from southern to northern California, and is growing undetected among native Californian species. GRATELOUPIA TURUTURU Yamada (HALY MENIA- CEAE).—Santa Barbara Co., attached to docks in the Santa Barbara boat harbor, 34°24'27.90’N, 119°41'32.40"W, thalli tetrasporic and cystocarpic/ spermatangial, 1 August 2009, J. R. Hughey s. n. (UC 1944749, UC 1944750). Previous knowledge. Native to Japan and Korea, type locality: Muroran, Hokkaido, Pacific (Verlaque et al., loc. cit.). Grateloupia turuturu was first reported outside of its native range 1n 1973 in Portsmouth, England (W. F. Farnham and L. M. Irvine 1973, British Phycological Journal 8:208—209). Since then, it has invaded the northwest Atlantic, France, The Netherlands, Portugal, Spain, the Canary Islands, the west coast of Africa from Mauritania to Namibia, Tasmania, New Zealand, Russia, and much of Asia (M. D. Guiry and G. M. Guiry, AlgaeBase. World-wide electronic publication, National University of Ireland, Galway. http://www. algaebase.org; [accessed 27 October 2009]). Earlier reports of this species in the Atlantic were incorrectly identified as G. doryphora (Montagne) M. A. Howe by Farnham and Irvine (loc. cit.) and M. Villalard- Bohnsack and M. H. Harlin (1997, Phycologia 36:324-328). B. Gavio and S. Fredericq (2002, Europe- an Journal of Phycology 17:349—-359) were the first to show that G. turuturu was the correct name to apply to the species invading the Atlantic. Grateloupia turuturu spreads by spores contained in the ballast water of ship hulls (Villalard-Bohnsack and Harlin, /oc cit.). Once the ballast water is discharged, the spores germinate and thrive in varied environments, including waters that are nutrient enriched and those that fluctuate in salinity and temperature. Based on the aggressiveness of this marine alga, as well as its presence in major shipping ports, G. furuturu was predicted to spread throughout North America and the rest of the world (C. Simon et al. 1999, Botanica Marina 42:437—-440). Grateloupia turuturu 1s characterized as having simple blades that are pinkish to maroon in color, lubricous in texture, ruffled on the margins, and growing in clumps of 5—12 (Verlaque et al., loc. cit.). Significance. First report of G. turuturu in the northeastern Pacific. Specimens from Santa Barbara representing both tetrasporangial and gametangial thalli were in good morphological and anatomical agreement with illustrations of this species (Gavio and Fredericq, Joc. cit.; Verlaque et al., loc. cit.). Sporo- MADRONO [Vol. 5 phytic thalli are linear in shape, while gametophytes are suborbicular in outline. In contrast to G. californica and G. lanceolata (refer to anatomical comparison above), G. turuturu has relatively thin blades (200 um—350 um), loosely interwoven, predominantly anticlinally arranged medullary filaments, and a phenology where growth and spore production take place in the fall. Identifica- tion was confirmed on two of the Santa Barbara specimens using ITS-1 (GenBank GQ499329 and GQ499330) and rbcL gene sequences (GenBank GQ499331 and GQ499332). DNA sequences from specimens from Santa Barbara were identical to previously published sequences of G. turuturu from Japan and Korea, and to introduced specimens from New Zealand, North America, Britain, and France. In Santa Barbara, G. turuturu was found attached growing on the docks in a harbor with 1,133 slips that accommodates resident house and pleasure boats, and traveling sail boats and yachts. Since G. turuturu was growing in this rather small harbor, it is likely that this seaweed was introduced by traveling boaters, rather than by commercial shipping vessels. If this is the case, Santa Barbara is not likely the site of primary introduction. It is more probable that G. turuturu was first established in a larger shipping port, then secondarily introduced via sail boat by ballast water or by attachment to the hull, although an analysis of the latter mode of introduction by recreational yachts was found unlikely (F. Mineur, M. P. Johnson and C. A. Maggs 2008, Environmental Management 42:667—676). Thorough surveys in the nearby marinas of San Francisco, Newport, and San Diego may support this hypothesis. Based on how rapidly G. turuturu has spread globally and this report of its newly recognized presence in California, this species should be considered the most invasive red seaweed on the planet. NEOSIPHONIA HARVEYI (J. W. Bailey) M.-S. Kim, H.-G. Choi, Guiry and G. W. Saunders (RHODOME- LACEAE).—Humboldt Co., attached to the underside of floating wood docks and on the vertical sides of concrete pilings 0 to 6 m in depth, Humboldt Bay, 40°48'25.344"N, 124°09'59.8674’W, thallus tetrasporic, 8 August 2006, A. Lyman s. n. (UC 1944754). Previous knowledge. Native to Japan; introduced to Norway, British Isles, Atlantic Europe, New Zealand, east coast of North America from Newfoundland to North Carolina (L. MeclIvor et al. 2001, Molecular Ecology 10:911—919) and Florida (D. S. Littler et al. 2008, Submersed plants of the Indian River Lagoon, 286 pp.); lectotype locality: Stonington, Connecticut, USA (C. A. Maggs and M. H. Hommersand 1993. Seaweeds of the British Isles. Volume 1. Rhodophyta. Part 3A. Ceramiales, 444 pp.). In California, McIvor et al. (Joc. cit.) identified a specimen collected in 1994 from Monterey Bay as Polysiphonia acuminata Gardner. This © specimen shared the same rbcL haplotype (i.e., haplo- | type F) as that from invasive specimens of Neosiphonia | harveyi from New Zealand and North Carolina, and is | closely related to a haplotype from Honshu, Japan. | Significance. Second report from northeast Pacific | Ocean and California, approximately 650 km north of | original range. This inconspicuous seaweed is easily | overlooked or mistaken for native species. The thallus is | distinguished from other taxa by the following charac- | teristics: up to 5 cm tall, dark red, 4 pericentral cells, | trichoblasts, and tetraspores that occur in spiral series. Identification of the Humboldt sample was confirmed | 009] by analyzing a portion of the rbcL gene. The nucleotide rbcL sequence (GenBank GU339504) generated from the Humboldt specimen was identical to a specimen of N. harveyi from Akkeshi, Hokkaido, Japan (GenBank AF342901). Both the Humboldt and Akkeshi sequences contain a genetic marker (A > G at position 231) that Mclvor et al. (/oc. cit.) defined as diagnostic for haplogroup B. These data support a second, unrelated introduction of N. harveyi to California. Oyster culture is a likely vector for seaweed introductions (F. Mineur et al. 2007, Biological Conservation 137:237—247). The history of oyster farming in California has _ been reviewed by E. M. Barrett (1963, Fish Bulletin 123. The California Oyster Industry. UC San Diego: Scripps Institution of Oceanography Library. Website http:// escholarship.org/uc/item/1870g57m). In 1896, adult eastern oysters, Crassostrea virginica, were imported from New York populations experimentally and raised in Humboldt Bay, but failed by 1912. They were again imported from the east coast from 1935 until the early 1940s. Since 1902, the Pacific oyster, Crassostrea gigas, was imported from various sites in Japan to oyster farms in Puget Sound, Washington. In 1928, the first experimental planting of Pacific oysters in California was made by the California Department of Fish and Game in Tomales Bay, with several other experimental plantings following in the early 1930s. According to P. S. Galtsoff (1930, Oyster industry of the Pacific coast of the U.S. Report U. S. Commissioner Fisheries, pp. 367— 400), oysters from Akkeshi Bay were deemed the best adapted for transplanting to North America. However, the decision to exclude Pacific oysters from Humboldt Bay, the largest California bay available for oyster culture, delayed the state’s development of the industry. Importation of the Pacific oyster from Japan to Humboldt Bay was initiated in 1953, and in 1957-58, Pacific oyster spat were imported from Willapa Bay, Washington to Humboldt Bay. Large scale production of Pacific oysters in Humboldt Bay began in 1955 and has continued to be an important industry. Although it is clear from our work that Neosiphonia harveyi in Humboldt Bay originated in Japan, the trajectory of the introduction is unknown. It may be primary (directly NOTEWORTHY COLLECTIONS 295 from Japan to California), secondary (from Japan to the east coast of the U.S. or to Washington, thence to California) or even tertiary, since oysters from San Francisco Bay were exported to Humboldt Bay (Barrett, /oc. cit.). Specimens collected by A. Lyman were gathered as part of a California Department of Fish and Game/ Office of Spill Prevention and Response project funded through the Introduced Species Study (Grant Number P0875029). —JEFFERY R. HUGHEY, Department of Science and Engineering, Hartnell College, Salinas, CA 93901; KATHY ANN MILLER, Herbarium, University of California, Berkeley, CA 94720-2465; and ASHLEIGH LYMAN, Marine Pollution Studies Laboratory, Moss Landing Marine Laboratories, Moss Landing, CA 95039; jhughey@hartnell.edu. NEW MEXICO BOERHAVIA PTEROCARPA S. Watson (NYTAGINA- CEAE).—Luna Co., disturbed ground near an isolated house at 5165 Veranda Rd. SE, near Deming, NM, about 1.5 miles S and 3 miles E of the Luna County Courthouse, 32°11.558’N, 107°26.091'W, elev. 1314 m, 26 Aug 2009, Jercinovic 917 (NMC, UNM). Previous knowledge. Boerhavia_ pterocarpa occurs sporadically from southern Arizona (Cochise, Pima, and Yuma Counties) into northeastern Sonora in Mexico. Significance. First U.S. report of this taxon outside of Arizona. The owner of the residence mentioned above has attended Master Gardener conferences in Sierra Vista, AZ, and purchased plants there. It is quite possible that the arrival of this Boerhavia in New Mexico is related to these purchases. —EUGENE M. JERCINOVIC, 6285 Algodo6n Rd. SW, Deming NM 88030. gjercinovic@earthlink.net. MADRONO, Vol. 56, No. 4, p. 296, 2009 PRESIDENT’S REPORT FOR VOLUME 56 Dear CBS member, I am honored to provide my first President’s Report for the Society. Dean Kelch has transitioned to the Past President’s position. CBS Recording Secretary Nishanta Rajakaruna has accepted a new position in Maine. When he leaves, Mike Vasey, a former President, will step in to take his place as Recording Secretary. Although he will be residing in Maine most of the year Dr. Rajakaruna will continue his research on Lasthenia and other California plants. Andrew Doran (Ist Vice President), Rod Myatt (2nd Vice President), Heather Driscoll (Corresponding Secretary), Kim Kersh (Membership Chair) and Tom Schweich (Treasurer) continue as officers of the Society. Other members of the Council included Staci Markos, Ellen Simms, Chelsea Specht and the Graduate Student Representative Ben Carter. Thank you Susan Bainbridge for your work as Webmaster this past year. The Editors of Madrono, Rich Whitkus and Tim Lowrey, make up the remainder of the Board. This is the first volume of Madrono edited by Drs. Whitkus and Lowrey, and they are quickly bringing the journal back under control and on time after a transitional year. They have brought a commitment to quality and professionalism that gives all of us confidence in the future of the journal. The past year was an active time for the Board of the California Botanical Society. The Board this year is moving to bringing the Society totally online through the website. We hope to provide members with electronic access to the journal in upcoming months. Submissions of manuscripts and reviews have already converted to online submission to some extent, but may move entirely to electronic processing in the near future, along with membership renewals and other functions. We’re excited by this process but are moving carefully to ensure sustainability of these improvements. Meanwhile, we also are working on getting all past issues of Madrono available through an electronic journal access website. So many classic articles are hiding in the older issues, from floras and species treatments, to insightful papers on ecology and evolution; electronic access to these articles would raise the status of the journal as well as providing us all great insight into our history as botanists of western North America. As I write this, the Annual Banquet and Graduate Student Meeting at San Jose State University have already occurred. These were both quite successful with a fantastic talk by Doug Schemske from Michigan State University on studies of adaptation and speciation. We congratulate Susie Woolhouse (SJSU graduate student) and Rod Myatt for their exceptional work organizing this successful meeting. In fact, the graduate student talks and posters were so excellent at this meeting the judges could not agree which were the best talks for awards, so we ended up with multiple awards. Winners were Brian Anacker (UC Davis), Naomi Fraga (Rancho Santa Ana), Jennifer Goropse (San Jose State), Dena Grossenbacher (UC Davis), Matt Guilliams (UC Berkeley), Kristen Hasenstab-Lehman (Rancho Santa Ana), Stephanie Porter (UC Davis) and Jenn Yost (UC Santa Cruz). Congratulations to you all. Our membership base is the foundation of the Society and your support allows us to promote botanical research and education. Increasing our membership is thus a priority, so please continue to encourage your colleagues to join us and to publish in Madrono. Please consider providing a sponsoring membership or subscription to a foreign scientist or scientific institution to support botan- ical research in economically depressed, developing countries. For more information on making such a gift, please contact Corresponding Secretary Heather Driscoll (hdriscoll@berkeley.edu). The Society also welcomes gifts or other contributions to our endowment. Finally, I note that the Society reaches its centennial year in 2015, which is not that far away. The Board has already begun considering ideas about how to celebrate that event. We’re open to ideas from the membership, so as you come up with them, please pass them along to board members. V. Thomas Parker December 2009 MADRONO, Vol. 56, No. 4, p. 297, 2009 EDITORS’ REPORT FOR VOLUME 56 We are pleased to report the publication of this volume of Madrono by the California Botanical Society (CBS) in 2009. The journal is continuing to work towards a timelier rate of publication with turnaround time in the past year having averaged six months. As always, we are extremely grateful to all the individuals who serve are reviewers and contribute to the quality of the articles. This year we received 36 new manuscripts and 35 were accepted for publication. Several manuscripts were also carried over from the previous year. The current volume includes 28 articles (including Notes), five new taxa, 18 Noteworthy Collections, and three Book Reviews. There was a mix of systematic and ecological manuscripts submitted. Many of the systematic manuscripts incorpo- rated current molecular techniques and data as well as cutting edge data analysis methods. It is notable that manuscripts reporting taxa new to science are being submitted on a regular basis. As Editors, we have enjoyed our interactions with contributors and reviewers this past year and look forward to another year of continuing the long and distinguished tradition in botanical publication represented by Madrono. Tim Lowrey Richard Whitkus December 2009 MADRONO, Vol. 56, No. 4, p. 298, 2009 Ihsan Al-Shebaz Kelly Allred Tina Ayers Jeffrey Bacon James Bartolome David Bates Orland Blanchard David Bramlet Gregory Brown Leo Bruederle Kenton Chambers Lynn Clark Daniel Crawford Ellen Dean Mary Dellavalle Kristy Duran Carolyn Ferguson Gregory Filip Paul Fryxell John Guretsky Gary Hannan Ronald Hartman Max Hommersand Richard Jensen Erik Jules Michael Kane Sarah Kimball Gretchen LeBuhn Sandra Lindstrom REVIEWERS OF MADRONO MANUSCRIPTS 2009 Craig Lorimer Diane Marshall Nancy Morin Esteban Muldavin Robert Naczi Dean Nicolle Cathy Offord Beth Painter Robert Patterson Paul Peterson Burton Pendleton James Pringle Joanna Redfern James Reveal Philip Rundel Jake Ruygt Jeff Saarela Adriana Sanchez Louis Santiago David Shaw James Shevock Kelly Steele Nelli Sugi Michael Williams Hugh Wilson Tamara Zelikova Jennifer Zettler Wendy Zomlefer MADRONO, Vol. 56, No. 4, pp. 299-301, 2009 INDEX TO VOLUME 56 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. Achillea filipendulina, noteworthy collection from MT, 67. Ajuga reptans, noteworthy collection from MT, 67. Aliciella triodon, noteworthy collection from ID, 130, and OR, 132. Alkali meadow, effects of fire and groundwater extraction on habitat, 89. Alliaria petiolata, noteworthy collection from MT, 67. Amphicarpaea_ bracteata, noteworthy collection from WY, 134. Anoda pentaschista, naturalized in CA, 110. Araceae (see Peltandra) Arceuthobium abietinum subsp. wiensii, new subsp., 118; A. oaxacanum and A. rubrum in Mexico, 99. Arctostaphylos: A. viscida, ecology and growth in ponderosa pine forest in southwest OR, 238. Noteworthy collection: A. rainbowensis from CA, 212. Asteraceae: Hazardia orcuttii, soil community character- istics, 237; Hedeoma todsenii, in vitro propagation, cryopreservation and genetic analysis, 221; Holocarpha macradenia, effect of terrestrial mollusc herbivory in CA coastal prairie, 1. Noteworthy collections: From CA: Hesperevax acaulis var. ambusticola, 212 from ID: Prenanthella exigua, 130; from MT: Achillea filipendulina, Symphyotri- chum molle, 67. Astragalus: A. jaegerianus, effects of dust deposition on growth and physiology, 81. Noteworthy collections: A. diversifolius from WY, 134; A. tenellus from OR, 132. Boerhavia pterocarpa, noteworthy collection from NM, 295: Boraginaceae: New taxon: Phacelia hubbyi, 205. Noteworthy collections: Cryptantha gracilis from OR, 133; Plagiobothrys salsus from MT, 67. Brassicaceae: Streptanthus barbiger and_ S. relationship, 43. New taxon: Streptanthus oblanceolatus, 127. Noteworthy collections: From MT: Alliaria petiolata, 67; from OR: Caulanthus crassicaulis, 132, Physaria cobrensis, 131. Bromus rubens (see Coleogyne) Bryophyta: Amblystegiaceae (see Limbella) vernalis California: Alkali meadow, effects of fire and groundwa- ter extraction on habitat, 89; Astragalus jaegerianus, effects of dust deposition on growth and physiology, 81; Chorizanthe parryi var. fernandina, reproductive biology, 23; C. subsection pungentes molecular phylog- eny with emphasis on the C. pungens-C. robusta complex phylogeography, 168; Emerald Lake Basin, Sequoia Nat. Park, plant communities and floristic diversity, 184; Eriogonum parvifolium, extreme root proliferation, 58; Eucalyptus, diversity, reproduction and potential invasiveness, 155; Hazardia orcuttii, soil community characteristics, 237; Holocarpha macrade- nia, effect of terrestrial mollusc herbivory in coastal prairie, 1; Jepson, Willis, and his ‘““Phyto-jogs”, 49; Juglans californica and Juniperus californica, historical range extensions Rancho Muscupiabe, San Bernardino Co.,199; Juncus marginatus, status in CA, 283; Mal- vaceae, notes on taxa, including nomenclatural changes and additions to the flora, 104; Pine Hill, El Dorado Co., vegetation and flora of a biodiversity hot spot, 246; plant community water use and invasibility of semi-arid shrublands by woody species, 213; streamside revetment effect in Sequoia sempervirens forests, 71; Streptanthus barbiger and S. vernalis relationship, 43. New Taxa: Phacelia hubbyi,205; Sidalcea asprella subsp. nana, S. calycosa subsp. rhizomata, S. celata, S. sparsifolia, new combinations, 104; S. gigantea, 285; Streptanthus oblanceolatus, 127. Noteworthy collections: Arctostaphylos rainbowensis, 212; Cotoneaster frigidus, C. horizontalis, 64; Dasya sessilis, Grateloupia lanceolata, G. turuturu, 293; Hesperevax acaulis var. ambusticola, 212; Juncus anthelatus, 130; J. falcatus subsp. sitchensis, J. interior, 65; J. nevadensis var. inventus, 66; J. Planifolius, 130; Geranium yeoi, 130; Lepechinia rossii, 63; Neosiphonia harveyi, 294; Prunus speciosa, P. X yedoensis, Schizymenia dubyi, 64. California Botanical Society (see Jepson, Willis) Carex: C. scirpoidea var. gigas, type determination, 279. Noteworthy collections: From CO: C. conoidea, 66; from WY: C. foenea, C. intumescens, 134; C. scoparia, 135. Caulanthus crassicaulis, noteworthy collection from OR, 132; Chaparral: Comparison of short-term effects of two fuel treatments, 8; Hazardia orcuttii, soil community characteristics, 237; plant community water use and invasibility of semi-arid shrublands by woody species in southern CA, 213. Chorizanthe: C. parryi var. fernandina, reproductive biology, 23; C. subsection pungentes molecular phylog- eny with emphasis on the C. pungens-C. robusta complex phylogeography, 168. Coastal prairie (see Holocarpha) Coastal sage: Hazardia orcuttii, soil community charac- teristics, 237; plant community water use and invasi- bility of semi-arid shrublands by woody species in southern CA, 213. Coleogyne ramosissima, interactive effects of simulated herbivory and interspecific competition on seedling survival, 149. Collomia renacta, noteworthy collection from OR, 133. Colorado: Noteworthy collections: Carex conoidea, 66; Cotoneaster lucidus, Sorbus aucuparia, 211. Competition, interspecific (see Coleogyne) Compositae (see Asteraceae) Cotoneaster: Noteworthy collections: From CA: C. frigidus, C. horizontalis, 64; from CO: C. lucidus, 211. Cruciferae (see Brassicaceae) Cryopreservation (see Hedeoma) Cryptantha gracilis, noteworthy collection from OR, 133. Cupressaceae (see Juniperus) 300 Cyperaceae: Carex scirpoidea var. gigas, type determina- tion, 279. Noteworthy collections from WY, Carex foenea, C. intumescens, 134; C. scoparia, Fimbristylus puberula var. interior, Scirpus pendulus, 135. Dasya sessilis, noteworthy collection from CA, 293. Dasyaceae (see Dasya) Distichlis spicata (see Alkali meadow) Dodecatheon pulchellum, noteworthy collection from OR, 132. Dryopterideaceae (see Dryopteris) Dryopteris cristata, noteworthy collection from WA, 68. Editors’ Report for Vol. 56, 297. Elatinaceae (see Elatine) Elatine brachysperma, noteworthy collection from OR, 134. Ericaceae (see Arctostaphylos and Vaccinium) Eriogonum parvifolium, extreme root proliferation, 58. Noteworthy collections: E. hookeri from OR, 132; E. palmerianum from ID, 131. Eucalyptus, diversity, reproduction and potential inva- siveness in CA, 155. Fabaceae: Astgragalus jaegerianus, effects of dust depo- sition on growth and physiology, 81. Noteworthy collections: Amphicarpaea bracteata from WY, 134; Astragalus diversifolius from WY, 134; A. tenellus from OR, 132; Trifolium leibergii from OR, 133. Farlow, W. G., 112. Fimbristylus puberula var. interior, noteworthy collection from WY, 135. Fire: Effects of fire and groundwater extraction on alkali meadow, 89; comparison of two fuel treatments on chaparral communities, 8; early post-fire establishment on a Mojave Desert burn, 137. Floras: Emerald Lake Basin, Sequoia Nat. Park, CA, 184; Pine Hill, El Dorado Co., CA, 265. Gabbro soil (see Pine Hill) Geranlaceae (see Geranium) Geranium yeoi, noteworthy collection from CA, 130. Gilia (see Aliciella) Gramineae (see Poaceae) Grasslands (see Holocarpha) Grateloupia lanceolata, G. turuturu, noteworthy collec- tions from CA, 293. Groundwater extraction (see Alkali meadow) Halymeniaceae (see Grateloupia) Haplomitrium hookeri, noteworthy collection from OR, 68. Hazardia orcuttii, soil community characteristics, 237. Hedeoma todsenii, in vitro propagation, cryopreservation and genetic analysis, 221. Herbivory (see Coleogyne and Holocarpha) Hesperevax acaulis var. ambusticola, noteworthy collec- tion from CA, 212. Heteranthera dubia, noteworthy collection from OR, 134. Hibiscus: H. lasiocarpos, nomenclatural discussion, 104; H. moscheutos, noteworthy collection from WA, 69. Holocarpha_ macradenia, effect of terrestrial mollusc herbivory in CA coastal prairie, 1. Hydrophyllaceae (see Boraginaceae) MADRONO [Vol. 56 Idaho: Noteworthy collections: Aliciella triodon, 130; Eriogonum palmerianum, 131; Prenanthella exigua, 130; Pogogyne floribunda, Utricularia gibba, 131. Invasive plants: Eucalyptus, diversity, reproduction and potential invasiveness in CA, 155; plant community water use and invasibility of semi-arid shrublands by woody species in southern CA, 213. Jepson, Willis, and his ‘‘Phyto-jogs’’, 49. Juglandaceae (see Juglans) Juglans californica, historical Bernardino Co., CA, 199. Juncaceae (see Juncus) Juncus: J. marginatus, status in CA, 283. Noteworthy collections from CA: J. anthelatus, 130; J. falcatus subsp. sitchensis, J. interior, 65; J. nevadensis var. inventus, 66; J. planifolius, 130. Juniperus californica, historical range extension, San Bernardino Co., CA, 199. range extension, San Keys: Hibiscus lasiocarpos and H. moscheutos, 108; Sidalcea perennial spp in CA, 290; Trifolium variegatum varieties, 20. Labiatae (see Lamiaceae) Lagunaria pattersonia, naturalized in CA, 110. Lamiaceae: Noteworthy collections: Ajuga reptans from MT, 67; Lepechinia rossii from CA, 63; Pogogyne floribunda from ID, 131, and OR, 133; Salvia brandgegeei from Mexico, 135. Laminariales (see Lessoniopsis and Pleurophycus) Lavatera olbia, L. trimestris, naturalized in CA, 110; Lavatera vs. Malva, 109. Leguminosae (see Fabaceae) Lentibulariaceae (see Utricularia) Lepechinia rossii, noteworthy collection from CA, 63. Lessonia (see Lessoniopsis) Lessoniopsis littoralis, historical, distributional notes, 112. Limbella fryei, noteworthy collection from OR, 67. Loasaceae (see Mentzelia) nomenclatural and Malva (see Lavatera) Malvaceae: Anoda pentaschista, Lagunaria pattersonia, Lavatera olbia, L. trimestris, naturalized in CA, 110; notes on California taxa, including nomenclatural changes and additions to the flora, 104. New taxa: Sidalcea asprella subsp. nana, S. calycosa subsp. rhizomata, S. celata, S. sparsifolia, combinations, 104; S. gigantea, new species, 285. Noteworthy collection: Hibiscus moscheutos, from WA, 69. Marchantiophyta: Haplomitriaceae (see Haplomitrium) Mentzelia congesta, noteworthy collection from OR, 133. Mesoreanthus fallax, taxonomic status, 46. MEXICO: Arceuthobium oaxacanum and A. rubrum, 99; | Salvia brandgegeei, noteworthy collection, 135. Mimulus hymenophilus, noteworthy collection from MT, | 67. Mojave Desert: Astragalus jaegerianus, effects of dust new deposition on growth and physiology, 81; Coleogyne | ramosissima, interactive effects of simulated herbivory and interspecific competition on seedling survival, 149; early post-fire plant establishment, 137; plant commu- nity water use and invasibility of semi-arid shrublands | by woody species in southern CA, 213. Mollusc, terrestrial (see Holocarpha) 2009] Montana: Noteworthy collections: Achillea filipendulina, Ajuga reptans, Alliaria petiolata, Mimulus hymenophi- lus, Orthocarpus tolmei, Papaver croceum, Plagio- bothrys salsus, Symphyotrichum molle, 67. Myrtaceae (see Eucalyptus) Neosiphonia harveyi, noteworthy collection from CA, 294. Nevada (see Coleogyne) New Mexico: Hedeoma todsenii, 1n vitro propagation, cryopreservation and genetic analysis, 221. Noteworthy collection: Boerhavia pterocarpa, 295. Nyctaginaceae (see Boerhavia) Oregon: Arctostaphylos viscida, ecology and growth in ponderosa pine forest, 238; comparison of two fuel treatments on chaparral communities, 8; Juncus mar- ginatus, status, 283. Noteworthy collections: Aliciella triodon, Astragalus tenellus, 132; Collomia renacta, Cryptantha gracilis, 133: Dodecatheon pulchellum, 132; Elatine brachy- sperma, 134; Eriogonum hookeri, 132; Haplomitrium hookeri, 68; Heteranthera dubia, 134; Limbella fryei, 67; Mentzelia congesta, 133; Physaria cobrensis, 131; Pogogyne floribunda, 133; Prenanthella exigua, 131; Trifolium leibergii, 133. Orobanchaceae (see Orthocarpus) Orthocarpus tolmei, noteworthy collection from MT, 67. Owens Valley, CA (see Alkali meadow) Papaver croceum, noteworthy collection from MT, 67. Papaveraceae (see Papaver) Peltandra virginica, noteworthy collection from WA, 68. Phacelia hubbyi, elevation from P. cicutaria var. hubbyi, 205: Pine Hill, El Dorado Co., CA, vegetation and flora of a biodiversity hot spot, 246. Physaria cobrensis, noteworthy collection from OR, 131. Plagiobothrys salsus, noteworthy collection from MT, 67. Pleurophycus. gardneri, historical, nomenclatural and distributional notes, 112. Poaceae (see Bromus) Pogogyne floribunda, noteworthy collections from ID, 131, and OR, 133. Polemoniaceae (see Aliciella and Collomia) Pollination biology (see Chorizanthe) Polygonaceae (see Chorizanthe and Eriogonum) Pontederiaceae (see Heteranthera) Prenanthella exigua, noteworthy collections from ID, 130, and OR, 131. President’s Report for Vol. 56, 296. Primulaceae (see Dodecatheon) Propagation, in vitro (see Hedeoma) Prunus speciosa and P. X yedoensis, noteworthy collec- tions from CA, 64. Reviews: Plant Invasions: Human Perception, Ecological Impacts and Management eds. B. Tokarska-Guzik, et al., 60; Systematics, Evolution and Biogeography of Compositae by V. A. Funk, et al., 209; The California INDEX TO VOLUME 56 301 Deserts: An Ecological Rediscovery by Bruce M. Pavlik, 60. Rhodomelaceae (see Neosiphonia) Riparian systems (see Sequoia sempervirens) Root proliferation (see Eriogonum) Rosaceae: Coleogyne ramosissima, interactive effects of simulated herbivory and interspecific competition on seedling survival, 149. Noteworthy collections: From CA: Cotoneaster frigi- dus, C. horizontalis, Prunus speciosa, P. * yvedoensis 64. From CO: Cotoneaster lucidus, Sorbus aucuparia, 211. Salvia brandgegeei, noteworthy collection from MEXI- CO; 135: Saunders, DeAlton, 112. Schizymenia dubyi, noteworthy collection from CA, 64. Schizmeniaceae (see Schizymenia) Scirpus pendulus, noteworthy collection from WY, 135. Scrophulariaceae (see Mimulus) Sequoia Nat. Park, CA, Emerald Lake Basin plant communities and floristic diversity, 184. Sequoia sempervirens forests, effect of streamside revet- ment, 71. Serpentine soils, comment on zonal, intrazonal and azonal concepts, 57. Sidalcea: New combinations from CA, S. asprella subsp. nana, S. calycosa subsp. rhizomata, S. celata, S. sparsifolia, 104; new species, S. gigantea from CA, 285. Sierra Nevada (see Pine Hill and Sequoia Nat. Park) Sorbus aucuparia, noteworthy collection from CO, 211. Sporobolus aeroides (see Alkali meadow) Stream restoration (see Sequoia sempervirens) Streptanthus: S. barbiger and S. vernalis relationship, 43: S. oblanceolatus, new sp. from CA, 127. Strother, John L., dedication of Vol. 56 to, 302. Subalpine flora and vegetation (see Sequoia Nat. Park) Symphyotrichum molle, noteworthy collection from MT, OF, Taxodiaceae (see Sequoia sempervirens) Tilden, Josephine, 112. Trifolium: T. leibergii, noteworthy collection from OR, 133; 7. variegatum var. geminiflorum, new combination, 208. Utricularia gibba, noteworthy collection from ID, 131. Vaccinium macrocarpon, association of Juncus marginatus with, 283. Viscaceae (see Arceuthobium) Washington: Noteworthy collections: Drvopteris cristata, 68; Hibiscus moscheutos, 69; Peltandra virginica, 68. Wyoming: Noteworthy collections: Amphicarpaea brac- teata, Astragalus diversifolius, Carex foenea, C. intu- mescens, 134; C. scoparia, Fimbristylus puberula var. interior, Scirpus pendulus, 135. MADRONO, Vol. 56, No. 4, pp. 302-303, 2009 DEDICATION JOHN L. STROTHER This volume of Madrono is dedicated to Dr. John L. Strother. Born in Conroe, Texas, 1941, he earned a B.S. at Sam Houston State College, a Master’s at Washington University in St. Louis, and his Ph.D. in 1967 at the University of Texas, Austin, under the direction of Dr. Bilhe Turner. His dissertation was a systematic study of the tagetaean genus Dyssodia, and he remainsR an avid synantherologist to this day. Upon receiving his doctor- ate, John accepted a staff position at the University of California Herbarium, the place he still calls home. He has served as its ““‘Deputy Director’ on various occasions. John’s research focuses on composite taxonomy, with an occasional dalliance into other families. During his professional career he has authored nearly 90 scientific articles, over 140 generic treatments of Compositae for Flora North America North of Mexico, and an assortment of other contributions to taxonomic resourc- es, including the Flora of Chiapas, The Jepson Manual, and Tarweeds & Silverswords. He loves his research, especially that relating to arcane nomenclatural details that most systematists abhor. He possesses a_ broad overview of what good monography ought to be, especially that of a classical nature. John is widely known for being a lover of the English language and having a penchant for grammatical accuracy, word use, and, what else, the use of commas, colons, and periods (he is likely to edit this dedication). His cleverness with words extends into botanical nomenclature, coining an array of generic names that can best be described as entertaining: Complaya (the beach composite), Elaphandra (named for Dr. Ron Hartman—elaphos is a Greek word for a stag), Wamal- chitamia (coined after the Tzeltal name for Dr. Dennis Breedlove), Jefea (named for his mentor, Billie Turner), Oblivia (purportedly an anagram of Bolivia, the source of the type material), and probably his best known coining, Damnxanthodium (it should be obvious that this is a genus of yellow flowered composites). And there is Zyzyxia, a name “arbitrarily formed.” John’s editorial skills were put to good use when he assumed co-editorship of Madrono, along with his UC Herbarium colleague Alan Smith. John later served as editor of Systematic Botany, and he continues to serve as an editor for the Flora North America project. His former mentor, Professor Turner, writes about John: “Honest, direct, forceful, hardly a tactful bone in his sternum.”’ I (Patterson) recall reading a Strother review where he commented on the author’s written abstract: “This is merely a description of contents. I should like to see an abstract.”’ Yet, as an editor that rare tactful bone would appear, as he often returned a review of a marginally acceptable manuscript with the comment to “give the author an opportunity to make the needed corrections.”’ John always treated authors with respect. John has also been, at times, a taxonomic activist. His publications include insightful critiques and commentar- les on issues of nomenclature and classification, and invitations to his fellow plant taxonomists to speak out and be heard, or in today’s parlance, to represent. John’s position at UC did not require him to teach classes, but he would offer the occasional seminar in nomenclature and give almost yearly lectures in courses on plant systematics. He is a natural teacher, articulate and precise in explanations, witty, and encouraging. He has co-taught many Jepson Herbarium workshops on the Sunflower family with his colleague, Bruce Baldwin. For many years John taught the Compositae to the plant taxonomy class at SFSU (eventually he considered Patterson adequately competent to give the lecture and save himself a trip). Some time in the late 1980’s John developed a passion outside of botany, for fountain pens. He approached this passion as a taxonomist would, cataloguing them, studying their history and their morphology. This fascination (and perhaps this is not a strong enough word—maan, he likes them!) led him to amass thousands of pens, to frequent pen shows around the country, buying, selling, trading, even repairing. I (Patterson) have made two long road trips with John to botanical meetings, and both were punctuated with numerous stops at antique malls to assess their fountain pen stock. John is known for his generosity with his time. When asked about matters of nomenclature (he is an astute student of the International Code of Botanical Nomenclature) or to assist with identification of a plant, he is usually willing to put on hold what he is doing to help. More often than not he ends up ferreting 2009] DEDICATION out more details than the original questioner had in mind. John is known by many for his to-the-point delivery and economy of verbiage. If asked how he is doing, he will likely answer, ““Well.”” His business card for years reads: STROTHER—BERKELEY. John is also quite an epicure, an aficionado of artisanal brews, and a guy who prefers to pour his own—woe betide the waitperson who gets this wrong. So let us end this dedication in Strotherian style: Blue Skies, John. Bob Patterson Professor of Biology San Francisco State University Alan Smith University Herbarium University of California, Berkeley with comments from Dr. Billie L. Turner Professor Emeritus University of Texas ae ’ MADRONO A WEST AMERICAN JOURNAL OF BOTANY VOLUME LVI 2009 BOARD OF EDITORS Class of: 2009—DONOVAN BAILEY, New Mexico State University, Las Cruces, NM MARK BORCHERT, USFS, Ojai, CA 2010—FRED HRUSA, California Department of Food and Agriculture, Sacramento, CA RICHARD OLMSTEAD, University of Washington, Seattle, WA 2011—JAMIE KNEITEL, California State University, Sacramento, CA KEVIN RICE, University of California, Davis, CA 2012—GRETCHEN LEBUMN, San Francisco State University, CA ROBERT PATTERSON, San Francisco State University, CA Corresponding Editor—TIMOTHY LOWREY Museum of Southwestern Biology MSC03 2020 University of New Mexico Albuquerque, NM 87131-0001 madrono@unm.edu AND Copy Editor—RICHARD WHITKUS Department of Biology Sonoma State University 1801 E. Cotati Avenue Rohnert Park, CA 94928-3609 whitkus@sonoma.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 MADRONO, Vol. 56, No. 4, pp. u—iv, 2009 MADRONO VOLUME 56 TABLE OF CONTENTS Abella, Scott R., et al., Early post-fire plant establishment on a Mojave Desert burn Alexander, Earl B., A comment on the zonal, intrazonal, and azonal concepts and serpentine soils Allen, Robert L. (see Jones, C. Eugene, et al.) Arnett, Joseph L. (see Giblin, David E.) Arvizu, Valentin (see Gross, LeRoy) Atallah, Youssef C. (see Jones, C. Eugene, et al.) Ayers, Debra R. (see Wilson, James L.) Baad, Michael (see Wilson, James L.) Baron, Sandra (see Brinegar, Chris) Beidleman, Richard G., To California with Jepson’s ““Phyto-jogs” in 1913 _ Bell, Duncan (see Vanderplank, Sula) Boyd, Steve (see Gross, LeRoy) Brinegar, Chris, and Sandra Baron, Molecular phylogeny of the pungens subsection of Chorizanthe (Polygonaceae: Eriogonoideae) with emphasis on the pore ity of the C. pungens-C. robusta COMMDICK 5.606 5.0 eed eee eee Bruederle, Leo P. (see Smith, Pamela EF. et al.) Buck, Roy E. (see Clifton, Glenn L.) Burk, Jack H. (see Jones, C. Eugene, et al.) Charls, Susan M. (see Pence, Valerie C.) Clifton, Glenn L., Roy E. Buck and Steven R. Hill, A new Sidalcea (Malvaceae) from northeastern AEA ere cee eek le re ie eee eat oe eee eee eae Coler, Kari (see Vourlitis, ‘George L. _ Daugherty, Carolyn M. (see Mathiasen, Robert L., and Carolyn M. Daugherty) Daugherty, Carolyn M. (see Mathiasen, Robert L., Carolyn M. Daugherty and Brian P. Reif) Defalco, Lesley A. (see Wiayratne, Upekala C.) Engel, E. Cayenne (see Abella, Scott R.) Ewers, Frank W. (see Jacobson, Anna L.) Fertig, Walt (see Lesica, Peter) Fullerton, Arline (see Parsons, Jenifer) Garrison, Laura M., and Robert Patterson, Elevation of Phacelia cicutaria var. hubbyi (Boraginaceae) to SPCCICS StAUUS inten cect bree ue dese ee gh a, e pen ga dees Ea eee een tee teen et Giblin, David E., and Joseph L. Arnett, Noteworthy collection from W asSHmM@ton: Sa... 2.2.0000c ee Gibson, Arthur C., Review of The California Deserts: An Ecological Rediscovery by Bruce M. Pavlik ___ Goforth, Brett R., and Richard A. Minnich, Historical range extensions for Juniperus californica (Cupressaceae) and Juglans californica (Juglandaceae) at Rancho Muscupiabe, San Bernardino