oc D “Ur NON, D.C = ” = Ww _ & 7 LLNLILSNI NVINOSHLINS S3IYVYUGIT LIBRARIES SMITHSONIAN INSTITUTION = 2) = ote ” — ee, n 4 = os = < ao = 5 = 5 zi ZN F z g 2 g BNR 2 = 2° i Zz = S = > ~ = > Ss > 3RARIES SMITHSONIAN INSTITUTION NOILONLILSNI NVINOSHLINS S3!1YVY9I Ww S n = ae w” > = aM” = ~” ul (ep) << %, oer cog od SS m a m S Ww oy = Ww ae wn = 3RARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S31YVYsIT_| q wm : a « wn = io?) ; = = < = < = , = ar a z =e 5 z .6 < 5 g i 8 2 g g z Fe 2. a =, = eae ae A : : —NVINOSHLINS S43 [YuvVdail LIBRARI ES SMITHSONIAN INSTITUTION - 2 “ zZ Zz Ww w” 7) Y < : “ = Yin, % z ee ‘ . r sz - “GY, = t \ iis i — 5 ip > F - 5 : = 7) = | w _ & he LINLILSNI NVINOSHLINS S3IY¥VYEIT LIBRARIES. SMITHSONIAN Pa 2) al sees ” —- w” if he =e ~ z NS = Pa = oe j oO 2 aN © ANS = TH oe os If ? g 2 RR 8 : g “iy - zZ E MO 2 E z , = > = = 2 oe ; . BRARIES SMITHSONIAN NOILNLILSNI NVINOSHLINS S3Iu¥Vudl) 3 7 i 2 ce 4 ~ - ane ac. a a a oc = < 3, 2 < oA < FS oc ar oc ar ro a = 3 = 5 2 5 J . z a 2 x 3 z LALILSNI NVINOSHLINS ~°4 huvyg iT tl BRARI ES SMITHSONIAE ISTO T ION - * ) _ O ay = Oo ioe) . — w = w —_ = Ys aS a = , 2 = ES 2 : e OA : _ KX Be — ae v ash ae EY —— SJL iho N: = o = Ww i. wn SNINYINOSHLINS S3IUVUSIT_LIBRARIES, SMITHSONIAN INSTITUTION, NO!LO = ” = ee wy z= < 2) < = ‘s = < = : — i sf S = a = g z 3 z 3 = i i Z = = a 2 a ear a ee ts te ES SMITHSONIAN_INSTITUTION NOILALILSNI_NVINOSHLIWS Saluvag tT | 7p) = tah Zz ul 1B = a WX + g = 4 tx 4 snr = Ce. Wass ae + <<’ ra < aN SWS < = cc = a = FN = = rs 5 E fe ae = S 3 . 2 z2 2 SNI-NVINOSHLINS (S31¥VYGI1_ LIBRARIES SMITHSONIAN, INSTITUTION {NOU mle S) = oO Es, ae 3 = ; — = by 108) fee = \. = es = : (he a) 5 \ NUP pe ae | Yi.> -— a FE i = Of fer? 7 m SW of m m > m ‘SS Zz m z SH Zz JES SMITHSONIAN INSTITUTION NOILNLILSNI _NVINOSHLINS S| aye etTi Sees S < * = gh, = GR» = fy, 2 & oe z = , 5 z 5 SX \ 2 2 8 2 g 2% \ z me =% = z = \ _ 2 = >” = >" ea 7) i = 7) Pr ” SNI_NVINOSHLINS S3I1YVUaIT_LIBRARI ES SMITHSONIAN INSTITUTION o eZ 4 - Win, - zs o = z yf a < oa ee _ ea. c @ GY yee = | SX EX oe | oe : pee of a 5 SS ra = m” 5 = Zz ae 7 ae 2 Zr = - IES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS Saiuvudi7 LIBE 5 ~ = = Z C = is - z : 5 Al i > = > 2 > ra "4 Ad rom es) ome ad = ate = — zZ a 2 2 o — :o2) — wn s aa 1SNI pod laVa dl TLIBRARIES | SMITHSONIAN ee : z = ; < ee : = — = aera + ~«w fy = | = Zz = re, fs fe x pip ° x om LY 4 c & hs x Oo = O Yy ys = ae pe ee = - = Uj 4 s = (@ = > = Ss ” Fr 7) eae a 4. 2 LIES SMITHSONIAN INSTITUTION NOILNLILSNI_NVINOSHLIWS Sal avugi7_ LIB! uw < Lu z= Aw ul =. ? 2. Oy ‘Yb “s a < NY ac ss a /f ra < 4 Bs < pe. a Wity c oc c Na = “Ly 4 4 NX = n Wb 35 ma 5 0 = - me ad z ee | Fa LSNI~NVINOSHLIWS psaluvudi] LIBRARI ES_ SMITHSONIAN, INSTITUTION _ NOI wo \ = w — oD = a Ys se 2 bE 2 5 a \\ = = E “3 = DP NN §- 2 =e st e MADRONO A WEST AMERICAN JOURNAL OF BOTANY VOLUME XXVIII 1981 BOARD OF EDITORS Class of: 1981—DANIEL J. CRAWFORD, Ohio State University, Columbus JAMES HENRICKSON, California State University, Los Angeles 1982—DEAN W. TAYLOR, University of California, Davis RICHARD VOGL, California State University, Los Angeles 1983—-ROBERT W. CRUDEN, University of Iowa, Iowa City DUNCAN M. PORTER, Virginia Polytechnic Institute and State University, Blacksburg 1984—Mary E. BARKWOoRTH, Utah State University, Logan HARRY D. THIERS, San Francisco State University, San Francisco 1985—STERLING C. KEELEY, Whittier College, Whittier, CA ARTHUR C. GIBSON, University of California, Los Angeles Editor—CHRISTOPHER DAVIDSON Idaho Botanical Garden, P.O. Box 2140, Boise, Idaho 83701 Published quarterly by the California Botanical Society, Inc. Life Sciences Building, University of California, Berkeley 94720 Printed by Allen Press, Inc., Lawrence, Kansas 66044 iy en Z . ll For his unparalleled investigations of the origin and evolution of the vegetation of western North America, his distinguished contributions to our understanding of the role of fossil floras in unlocking the secrets of paleoclimate, his early appreciation of the importance of plate tectonics as a controlling factor in determining the distribution of plants and animals, and his contribution of a very substantial body of literature that has consistently inspired his fellow scientists, volume 28 of Madrono is dedicated to DANIEL I. AXELROD. Dan, we know you consider the herbs that we love to be a “ground cover,” unworthy of notice, but we forgive you because, largely through your own insightful analyses, together with lots of tough rock-splitting in the hot Nevada sun, you have made the communities in which they occur the best known in the world from the stand- point of their Tertiary history. Besides, you have never given us a dull moment, and we recall with enthusiasm your many fine speeches and seminars. May your years be filled with equability, punctuated with just enough climatic and edaphic stress to add variety and zest! ili TABLE OF CONTENTS ACKERMAN, JAMES D., Pollination biology of Calypso bulbosa var. occidentalis (Orchidaceae): a food deception system___________________________-_ 101 ADAM, DAVID P., ROGER BYRNE, and EDGAR LUTHER, A late Pleistocene and Holocene pollen record from Laguna de las Trancas, northern coastal Santa Cruz County, ‘Cabtornia. 33.2. 3 ea eee 255 ARMSTRONG, WAYNE P., Noteworthy collection of Wolffia punctata ___________- 3H ARMSTRONG, WAYNE P., Noteworthy collection of Wolffia columbiana _________- 187 BAKER, GAIL A., PHILIP W. RUNDEL, and DAVID J. PARSONS, Ecological rela- tionships of Quercus douglasit (Fagaceae) in ae foothill zone of Sequoia National’ Park. California 22 22-2 oe ee ee 1 BARTEL, JIM A., Noteworthy collections of Lupinus citrinus and Streptanthus JOLY IUS TOV EIUAVUL Sac cesd ee ee eee eet 184 BARTOLOME, JAMES W. and BARBARA GEMMILL, The ecological status of Stipa pulchra (Poaceae) in California _____._..-_...__.--_______-__-_-_-----_--_--_--- 172 BLECK, JOHN (see Ferren, Wayne R.) BLUESTONE, VICTOR, Strand and dune vegetation at Salinas River State Beach, SATE ORIN Aiwecsa sot eri hag ac ed eer 49 BOWERS, JANICE E., Local floras of Arizona: An annotated bibliography ______ 193 BRUNSFELD, PAMELA (see Henderson, Douglass) BRUNSFELD, STEVEN (see Henderson, Douglass) BULLOCK, STEPHEN H., Aggregation of Prunus ilicifolia (Rosaceae) during dis- persal and its effect on survival and growth ________-_--__--__-__--_-_-_-- 94 BUTTERWICK, MAry (see Parfitt, Bruce D.) BYRNE, ROGER (see Adam, David P.) CARTER, ANNETTA M. and VELVA E. RUDD, A new species of Acacia (Legu- minosae: Mimosoideae) from Baja California Sur, Mexico____________________ 220 COCHRANE, THEODORE S., Noteworthy collection of Carex deweyana subsp. LOT OV ONG cate eh aS ae Oe re a a 186 D1 Tomaso, JOSEPH M., Re-establishment of Angelica californica CU DelbieraG) ae fe eg i ee By ee ee ee eee 226 DoRN, ROBERT D. and ROBERT W. LICHVAR, A new species of Cryptantha (Bo- raginaceae) from Wyoming —___._.-_-.-_._-2__-__-__--_~-1-_--___--2_-1--_---_- 159 Dorn, ROBERT D. and ROBERT W. LICHVAR, Specific status for Trifolium hay- GEMMEVAL DOV CON Es 6 fc Isp sh er ee 188 EVANS, CHARLES J. (see Haines, Robert D.) FERREN, WAYNE R., JR., JOHN BLECK, and NANCY VIVRETTE, Malephora cro- cea (Aizoaceae) naturalized in California... ..... =. 80 GEMMILL, BARBARA (see Bartolome, James W.) HAINES, ROBERT D. and CHARLES J. EVANS, Noteworthy collections of Madia subspicata, Rafinesquia californica, Cryptantha muricata, Carex tumulicola, Epilobium minutum, Argemone munita subsp. rotundata, Apera spica-venti, Rhamnus rubra subsp. yosemitana, and Mimulus gracilipes ___-_--_-_______- 39 HALSE, RICHARD R., Taxonomy of Phacelia sect. Miltitzia (Hydrophyllaceae) 121 Hamon, Dan, Noteworthy collection of Calyptridium pulchellum ________-_------ 188 HENDERSON, DouGLass M., STEVEN BRUNSFELD, and PAMELA BRUNSFELD, Noteworthy collections of Erigeron humilis, Hymenopappus filifolius var. idahoensis, Carex rupestris, Astragalus amnis-amissi, Gentiana propinqua, POPAVCT RIMQNENSTS. 2928 nn ee se oh ee ee eee 88 HENDRIX, LYNN B., Post-eruption succession on Isla Fernandina, Galapagos_. 242 HENRICKSON, JAMES, A new subspecies of Comarostaphylis polifolia (Ericaceae) fromeCoalhiila, Mexico. ~~ Se 95 fe ee ee ee 33 HENRICKSON, JAMES, Portulaca johnstonii, a new species of Portulacaceae from the Chihtiahuan Deserta: 2.2.2 ee ee ee 78 HOLEMAN, JAMES R. (see Stebbins, John C.) iV HowaALp, ANN M. and Bruce K. Orr, Noteworthy collection of Pedicularis Crenulata ft. CONGtGO 22222. 225 sass sa sb ee aa oe JOHNSTON, MARSHALL C., Chiococca henricksonii (Rubiaceae), a new species from the Chihuahuan Desert region ________-________________-_--_----- JOHNSTON, MARSHALL C., The diandrous, hypostomatic willows (Salicaceae) of tne: © hihvalaa Desert Tee lOn eee. ies KEELER, KATHLEEN H., Cover of plants with extrafloral nectaries at four north- ern California sites _______________-__-__-_-_- eee KEELEY, JON E., Diurnal acid metabolism in vernal pool /soetes (Isoetaceae) ___ KOEHLER, DONALD L. and DALE M. SMITH, Hybridization between Cowania mexicana var. stansburiana and Purshia glandulosa (Rosaceae) ___________- LESTER, Gary S., Noteworthy collection of Cochlearia officinalis _______________- LICHVAR, ROBERT W. (see Dorn, Robert D., both entries) LITTLE, R. JOHN, Adventitious rooting in coastal sage scrub dominants _______- LUTHER, EDGAR (see Adam, David P.) MARLEY, GREGORY A., Noteworthy collections of Erigeron compactus var. con- similis, Tetradymia spinosa, Chorispora tenella, Diplotaxis muralis, Malcol- mia africana, Sclerocactus mesae-verdae, Astragalus monumentalis, Salvia microphylla var. wislizeni1, Bromus diandrus, Ranunculus testiculatus, Cer- COCOTPUS THUNLGWOUS: oo acs ee ee ee MINNICH, RICHARD A. (see O’Leary, John F.) NESOM, Guy L., Five new species of Mexican Erigeron (Asteraceae) ____________ NIXON, KEVIN C. and KELLY P. STEELE, A new species of Quercus (Fagaceae) from southenm California... 222.295. Novak, PATTI J. and KATHRYN L. STROHM, Noteworthy collection of Dedeckera CLEVE TOOTS US carts 2 ras ONS me Tey gr aa aN a RN at A gf on ee Reae OBERBAUER, THOMAS A., Noteworthy collection of Hazardia orcuttii _________- O’LEARY, JOHN F. and RICHARD A. MINNICH, Postfire recovery of creosote bush scrub vegetation in the western Colorado Desert ________-____________________- Orr, BRUCE K. (see Howald, Ann M.) PARFITT, BRUCE D. and MARY BUTTERWICK, Noteworthy collection of Oro- banche uniflora subsp. occidentalis _________-_-__------- PARSONS, DavIp J. (see Baker, Gail A.) PARSONS, DAVID J., The historical role of fire in the foothill communities of pequola. National Park oo 2-< 2s a a ee REVEAL, JAMES L., Eriogonum libertini (Polygonaceae), a new species from horthern California <.....-. 2-2 ee 2 ee eee RODSTROM, WILLIAM E. (see Lester, Gary S.) RUDD, VELVA E. (see Carter, Annetta M.) RUNDEL, PHILIP W. (see Baker, Gail A.) SMITH, DALE M. (see Koehler, Donald L.) SMITH, JAMES R. (see Stebbins, John C.) SMITH, RICHARD H., Variation in immature cone color of ponderosa pine (Pi- naceae) in northern California and southern Oregon _____-_---------- SORENG, Ros J. and RICHARD SPELLENBERG, Noteworthy collections of Ipomoea CCVOLUd ANGUSTCLILVIG MULENS. e =~ ~~ es oe To $30le = ~ ma Pe a 3 So \ ea =30- BN j . ; \ / -20- o—-—o Midday _ J ; »—e Dawn he ee wa ie -10F bd See) “S ee ed eee L l | | | ie | Bl ! | Heal S fe) N D J F M A M J J A S O N 1977 1978 BUCKEYE (W-830m) Buds »——— New leaves ——— Mature leaves ey Senescent leaves ——— eee -60 Stem elongation “——------ — -50 -40 eK eis o* \ aoe ooo -30 : p = es e \ rd -20 ‘ / \, fA = \ Ze sr 10 \ Som ner we = i \eaa ~——__°——e—° ° fee | L | | | | yoo | S O N D J F M A ™M J A S O N 1I977 1978 1981] BAKER ET AL.: BLUE OAK ECOLOGY 7 TABLE 2. SOIL CHARACTERISTICS OF COMMUNITIES WITH Quercus douglasii IN SEQUOIA NATIONAL PARK. Nutrients are reported as ppm (+s.d.). Soil particle-size classes, loss on ignition and moisture content are reported as percents (+s.d.). Mixed- evergreen Foothill woodland woodland Blue-oak Buckeye Black-oak Oak-buckeye woodland woodland woodland woodland Total N 1920 + 830 1900 + 310 2 NS O57 0 2990 + 1310 Total P 650 + 300 620 + 190 720: 2 520 1040 + 650 K A595 278i 353 + 395 S0o2 79 460 + 195 Ga 1695 + 734 1795622504: L/i36 5 422 2426 + 998 Mg 134 + 60 150 + 59 150 2928 186 + 89 N-NH, 16.3 = 8 ZOU mee <2 1337-22419 1273 2 AS N-NO, his 252 LOC 222736 Oro) 2a S 11.0 + 3.9 Soluble P 52 2 36 Sys 5 64 + 85 (Ae OO pH 62535220739 675 + 0.16 6.45 + 0.10 6.59 + 0.34 Loss on ignition 620: 25.322 6.8 + 2 So 7 1024.6 Sand 79.8 + 6.5 82.2 + 3.4 82.1 + 1.6 80.2 + 3.6 Clay O10 = 5i4 Ore 125 528) a3 Flees La 15-bar soil moisture content (is ama a) (queer eS 9.9 + 1.0 223 64) the foothill woodland, indicating lower soil fertility (Table 2). Cation contents and available forms of nitrogen and phosphorous are not significantly different. Soils in all community types are sandy loams, with a mean sand content of about 80 percent. The higher organic matter content of the oak-buckeye communities, however, gives these soils a higher moisture content at —15 bars water potential. Seasonal water-stress patterns in Quercus douglasi reflect patterns in precipitation. Although Q. douglasii is able to use water from rel- atively great depths (Lewis and Burghy, 1964), it must rely on a lim- ited ground-water supply that apparently is depleted by autumn. Quercus douglasii showed similar seasonal patterns in dawn and midday water potentials at all three study sites over 1977-1978 (Figs. 2-4). The highest stress at all sites occurred during the initial (Sep- tember, 1977) measurement at the end of a severe two-year drought. With the onset of precipitation in November, water potentials began to increase, although that increase was much slower than in evergreen shrubs at the same sites. There was a secondary peak of stress at all << Fics. 2-4. Seasonal water potentials at dawn and midday and qualitative phenology of Quercus douglasii. FIG. 2 (top). Ash Mountain site. Fic. 3 (middle). Flume site. FIG. 4 (bottom). Buckeye Campground site. 8 MADRONO [Vol. 28 250 200 a fe) PRECIPITATION (mm) fe) oO 50 rae TEMAMUIVYVASONDUFMAMJJASONDJFMAMJVJAS 1977 1978 1979 Fic.5. Monthly precipitation for Ash Mountain over the 1977-1979 period of study. Trace precipitation is indicated by a “T”. sites; in December at the Flume and Ash Mountain and during De- cember—January at Buckeye Campground, reflecting continued drought conditions following initial fall rains (Fig. 5). The highest water potentials occurred in February and March when the soil was recharged with water. Minimum water potentials occurred in July at Ash Mountain and in August at the other two sites. Extreme summer water potentials are low for all three years from 1977-1979, but there is a large difference between the dawn water potentials of the drought (1977) and mesic (1978) years (Fig. 6). In 1977 the greatest difference between dawn and midday values was —6 bars. The following year the differences ranged from —21 to —30 bars, indicating considerable overnight recovery. Intermediate values of water potential and dawn-to-midday stress differentials were gen- erally present in 1979 when precipitation was close to mean levels. Maximum 1979 water stress at the Flume site, however, exceeded that reached during the drought of 1977. Intensive studies of the water relations of Quercus douglasii trees in the coast ranges of Monterey County over a three-year period found minimum dawn water potentials of —40 bars (Griffin, 1973). This is higher than our minimum dawn water potential (below —50 bars in 1981] BAKER ET AL.: BLUE OAK ECOLOGY 9 -60 @ Dawn 0 Midday 4 [ie ‘ i t ‘ 1977 1978 1979 1977 1978 1979 1977 1978 1979 Flume Ash Mountain Buckeye =90 -40 -20 WATER POTENTIAL (bars) Fic. 6. Minimum water potentials for 1977-1979 in three populations of Quercus douglasit. both 1977 and 1979). The density of Q. douglasii on slopes with shal- low soils has been shown to be directly related to water stress (Griffin, 1973). Phenological patterns and periods of vegetative growth among sites showed variability. The greatest amount of growth and development occurred from March through May when surface soil moisture was highest, temperatures were increasing (Table 3) and water uptake by plants was highest (Figs. 2—4). Stem elongation preceded the major phenological events (Figs. 2— 4). Trees at Ash Mountain showed a pattern of gradual stem elonga- 10 MADRONO [Vol. 28 BUCKEYE LUME —— — ASH MOUNTAIN X BRANCH ELONGATION (cm) E M A M J J A S O N 1978 Fic. 7. Seasonal pattern of stem elongation in three populations of Quercus doug- lasii. Brackets enclose one standard error around the mean. tion from February through April, while growth at the other two sites occurred rapidly over a shorter period of time (Fig. 7). This rapid growth occurred at the Flume site in March and at Buckeye Camp- ground, the coolest of the sites, in May. Total mean stem elongation for the Flume, Ash Mountain, and Buckeye Campground sites was 2.0, 2.1, and 3.7 cm, respectively. The first initiation of bud swelling occurred one to two months earlier at Buckeye Campground than at the other two sites despite the cooler temperatures (Figs. 24). New leaves formed after the main TABLE 3. MEAN MONTHLY HIGH AND LOW TEMPERATURES (°C) DURING THE 1978 GROWING SEASON AT THE Quercus douglasii STUDY SITES. Data for March are means for a single week. Ash Buckeye Mountain Campground Flume High Low High Low High Low February 15,2 4.2 11.6 3:0 10.5 4.2 March 19.3 7.7 IIS S77 1252 Hae April 17.9 fe) oe ~ 13.5 6:3 May 23.6 L120 21.6 oe | 2165 10.7 June Ziie5 14.0 26.8 13.9 27.9 15.9 1981] BAKER ET AL.: BLUE OAK ECOLOGY 11 branch elongation at Ash Mountain and Flume, but at Buckeye the appearance of new leaves and branch elongation continued simulta- neously. New leaves appeared at all sites by the end of March and were mature by mid-April, with a blue-green caste and a thick cuticle. The leaves remained on the trees throughout the major portion of sum- mer and began to fall at the end of August when trees were under maxi- mum water stress. By October all trees had lost the majority of their leaves and remained leafless until the following March. Flowering in Q. douglasii is determined by conditions of the pre- vious growing season, because reproductive buds are formed at that time. No flowering was observed during the 1978 season at any site, reflecting conditions of drought in 1977. However, many catkins de- veloped after the appearance of new leaves at the Ash Mountain site in 1979. There was also somewhat greater stem elongation during this season. Average elongation was 1.5 cm on 20 March 1978 and, for the same branches a year later (16 March 1979), elongation 2.2 cm. Studies of oak community structure by Griffin (1971, 1976) and White (1966) have shown that a combination of favorable conditions must occur for successful reproduction and establishment. With their limited root systems, seedlings of Q. douglasii must endure much lower summer water potentials than mature trees (Griffin, 1973). Not only are temperature and rainfall important, but also such factors as acorn and seedling predation, grazing pressure, and fire history. Phe- nological variability from season to season has been shown by Griffin (1971) and in unpublished National Park Service data. Establishment every year is not necessary for populations of long-lived species, such as oaks, to maintain themselves. When the correct combination of favorable conditions occur, the result is the establishment of a cohort of oaks of similar ages (Griffin, 1977). Quercus douglasii stands in Sequoia National Park have a high proportion of even-aged individuals of similar size (Brooks, 1969). The greatest proportion of trees are between 60 and 100 years old and 12 to 30 cm dbh. Stand density affects diameters to some degree, so that dense stands (163 trees/ha) have trees with an average dbh of 18 cm and open stands (25 trees/ha) show a greater average dbh of 36 cm in the Coast Range of Central California (White, 1966). The factors re- sponsible for the increased density of Q. douglasii in Sequoia National Park and other areas in California are believed to be changes in land- use history and concurrent occurrence of favorable establishment con- ditions in the 1860’s and 1870’s (Vankat and Major, 1978). During this period grazing was increased and increased density of Q. douglasii was favored by removing herbaceous competition for oak seedlings and decreasing fuel levels so that fires were not as intense or frequent. In addition to direct effects on oak seedlings, fire has an indirect effect on young blue oaks by increasing vulnerability to insect damage (Law- rence, 1966). 12 MADRONO [Vol. 28 The survival of Q. douglasii in drier habitats is the result of a coordination between physiological and phenological characteristics in response to environmental conditions. Initial distribution of blue oaks and subsequent stand structure are affected by patterns of climate and physical events. ACKNOWLEDGMENTS We thank Mignonne Bivin, Mavis Hasey, Jason Greenlee, Celia Hemphill, Carolyn Wallace, and Steve DeBenedetti for help in the field work for this project. This research was supported by the National Park Service. LITERATURE CITED ALLEN, S. E., H. M. GRIMSHAW, J. A. PARKINSON, and C. QUARMBY. 1974. Chem- ical analysis of ecological materials. John Wiley and Sons, New York. Brooks, W. H. 1969. Some quantitative aspects of the grass-oak woodland in Sequoia National Park. Unpubl. report, Sequoia National Park, Three Rivers, Calif. GRIFFIN, J. R. 1971. Oak regeneration in the upper Carmel Valley, Calif. Ecology 52:862-—868. . 1973. Xylem sap tension in three woodland oaks in central California. Ecology 54:152-159. . 1976. Regeneration in Quercus lobata savannas, Santa Lucia Mountains, Cal- ifornia. Amer. Midl. Naturalist. 95:422—435. 1977. Oak Woodland. Jn: Barbour, M. G. and J. Major, eds., Terrestrial vegetation of California, p. 383—415. Wiley-Interscience, New York. JOHNSON, W., C. M. McCKELL, R. A. EVANS, and L. J. BERRY. 1959. Yield and quality of annual range forage following 2,4-D application on blue oak trees. J. Range Managem. 12:18—20. LAWRENCE, G. E. 1966. Ecology of vertebrate animals in relation to chaparral fire in the Sierra Nevada foothills. Ecology 47:278-291. LEITH, H. 1974. Phenology and seasonality modeling. Springer-Verlag, New York. LEONARD, O. A. 1956. Effect on blue oak (Quercus douglasii) of 2,4-D and 2,4,5-T concentrates applied to cuts in trunks. J. Range Managem. 9:15-19. Lewis, D. C. and R. H. BuRGHY. 1964. The relationship between oak tree roots and ground water in fractured rock as determined by tritium tracing. J. Geophys. Res. 69:2579-2588. PILLSBURY, N. H. 1978. Hardwood stand density characteristics for central coast counties in California. Unpubl. report, Central Coast Resource Conservation and Development Area, Salinas. RITCHIE, G. A. and T. M. HINCKLEY. 1975. The pressure chamber as an instrument of ecological research. Ecological Research 9:166—254. SCHOLANDER, P. F., H. T. HAMMEL, E. BRADSTREET, and E. A. HEMMINGSEN. 1965. Sap pressure in vascular plants. Science 148:339-346. VANKAT, J. L. and J. Major. 1978. Vegetation changes in Sequoia National Park, California. J. Biogeography 5:377—402. WHITE, K. L. 1966. Structure and composition of foothill woodland in central coastal California. Ecology 47:229-237. (Received 23 Jan 1980; accepted 7 Jun 1980; final revision received 14 Jul 1980.) HYBRIDIZATION BETWEEN COWANIA MEXICANA VAR. STANSBURIANA AND PURSHIA GLANDULOSA (ROSACEAE) DONALD L. KOEHLER DALE M. SMITH Department of Biological Sciences, University of California, Santa Barbara 93106 ABSTRACT The hybrid combination Cowania mexicana var. stansburiana X Purshia glandulosa has been observed at two locations in Inyo Co., California. Morphological intermediacy, reduced fertility, and the addition of species-specific flavonoids document hybridization in the narrow altitudinal zone of the species’ overlap at these localities. A unique aspect of this study is that leaf flavonoid complements of the two species and hybrids are the same whereas petal flavonoid complements are different subsets of the leaf flavonoid constituents. The discovery of this hybrid documents the third possible hybrid combi- nation among three taxa (including P. tridentata) that are currently classified within two genera and magnifies the question of the logic of this classification. Cowania mexicana var. stansburiana (Torr.) Jeps. and Purshia glandulosa Curran are long-lived shrubs or small trees of the Rosaceae. Within the southern Great Basin, they form a significant component of the vegetation from 850-2700 m (Fig. 1). These two species, with P. tridentata (Pursh) DC., have been the subjects of numerous studies because of their conspicuous abundance and importance as range plants. Ecological (Nord, 1965; Mortenson, 1970) and range manage- ment studies (USDA, 1937; Plummer et al., 1968; USDA, 1975; Blauer et al., 1975) make up the bulk of the literature. Cowania, with three or four species, and Purshia, consisting of two species, have been regarded as well-defined genera because of obvious differences in number of carpels, appearance of fruits, and other di- vergent morphological characters. In spite of this, several authors have noted interspecific and intergeneric hybridization (Brandegee, 1903; Stebbins, 1959; Stutz and Thomas, 1963; Nord, 1965; Blauer et al., 1975). This paper presents morphological, chemical, and fertility data that document hybridization between Cowania mexicana var. stansbu- viana and Purshia glandulosa. Although this hybrid combination has been suggested as possible and probable (Brandegee, 1903; Thomas, 1957; Stutz and Thomas, 1963), no documentation of either artificial or natural hybrids exists. The results of this study are of special interest in that they document the third possible hybrid combination among three taxa that are currently classified within two genera. MADRONO, Vol. 28, No. 1, pp. 13-25, 12 February 1981 3 14 MADRONO [Vol. 28 o P. glandulosa 4 C. mexicana ‘ \ var. stansburiana Q 500 km Fic. 1. Geographic distribution of P. glandulosa (dotted line) and C. mexicana var. stansburiana (solid line), and locations of populations used in this study. Brandegee (1903), describing a plant he collected in the Providence Mountains of southeastern California and that he called Cowania mexicana var. dubia, stated that “this form was also collected by Dr. C. A. Purpus on Morey Peak, Nevada in 1898, and he considered it a hybrid between Cowanzia and Purshia.” No indication was given by Brandegee of which species of Purshia Purpus considered to be pa- rental, although a Purpus specimen (6356, UC) is clearly labelled Cowania Mexicana X Purshia glandulosa. Subsequent collections and observations in the Providence Mountains indicate that P. glandulosa and C. mexicana var. stansburiana are abundant there, but P. tri- dentata does not occur that far south (Munz, 1959). Nord (1965) in- terpreted Brandegee’s C. mexicana var. dubia, from the Providence Mountains, and Purpus’ collection from Morey Peak, Nevada, as hy- 1981] KOEHLER & SMITH: COWANIA/PURSHIA HYBRIDIZATION is) brids between P. tridentata and C. mexicana var. stansburiana and proposed the hybrid be called “Purpus cliffrose (C. mexicana var. dubia Bdg.)”. Stebbins (1959), using his observations and the preliminary data of H. Stutz, P. Plummer, A. Holmgren, W. S. Boyle, and L. K. Thomas, concluded that sympatric hybridization and introgression between P. tridentata and C. mexicana var. stansburiana is widespread. In the same paper, Stebbins (1959) stated that P. tridentata and P. glan- dulosa form such extensive hybrid swarms that the identity of the parental types is completely obliterated, although no quantitative data were presented. These putative hybrid swarms occur along the eastern side of the Sierra Nevada and elsewhere in central and eastern Cali- fornia (Koehler, unpubl. data). Stutz and Thomas (1963) documented that C. mexicana var. stans- buriana and P. tridentata frequently form fertile hybrids in nature, ranging from a relatively few putative F, hybrids to situations that suggest the presence of F., segregates as well as F, and backcross derivatives, depending on the slope exposure of parental populations. These authors also suggested that P. glandulosa appears to be a sta- bilized segregate from hybrids of C. mexicana var. stansburiana and P. tridentata. Artificial hybridization by pollinating emasculated flowers of C. mexicana var. stansburiana with pollen from P. tridentata has pro- duced viable seeds and seedlings (Blauer et al., 1975). Knobloch (1972) listed 2993 reports of intergeneric hybridization in flowering plants and stated that hindrance of acceptance of hybrid- ization as a potent evolutionary force results from lack of knowledge of the extent of the process in nature. The bulk of literature dealing with homogamic intergeneric hybridization describes crosses that were artificially derived, such as Helianthus X Viguiera (Heiser, 1963), Lycopersicon X Solanum (Rick, 1951, 1960), Lychnis x Silene and Melandrium X Silene (Kruckeberg, 1962), Hordeum x Agropyron (Kruse, 1974), Tripsacum X Zea (Mangelsdorf and Reeves, 1938), and others in the Gramineae (Stebbins, 1950). Observations of natural intergeneric hybridization such as Cowania X Purshia (Stutz and Thomas, 1963) and Encelia X Geraea (Kyhos, 1967) are limited, ex- cept in the Orchidaceae where natural hybrids between genera are common (Pijl and Dodson, 1966). METHODS Two mixed stands of C. mexicana var. stansburiana and P. glan- dulosa containing putative hybrids were observed in the field and specimens were collected for study. Specimens from monotypic stands of each species were also collected for comparison (Fig. 1). The pu- tative hybrids occur on the east flank of Cerro Gordo and Waucoba 16 MADRONO [Vol. 28 TABLE 1. MORPHOLOGICAL COMPARISON OF Cowania mexicana VAR. stansburiana, Purshia glandulosa, AND PUTATIVE HYBRIDS. Numerical values represent the mean of 10 measurements on each of 5 individuals from 4 populations (20 plants) of P. glan- dulosa and of 5 individuals from 3 populations (15 plants) of C. m. var. stansburiana and of the 14 hybrids. Numbers in parenthesis are the ranges of means of 10 measure- ments/individual. Character C.m. Hybrids P.g. Pistil number 5.0 (4.2-6.4) 2.4 (2.0—3.0) 1.2 (1.0—1.3) Style length in fruit (cm) 3.8 (3.0—4.3) 1.9 (1.6—2.2) 0.7 (0.6—0.8) Style pubescence in fruit plumose short villous puberulent Achene shape oblong intermediate obovate Achene ratio, L/W 3.6 (3.0—4.3) 2.3 (2.2—2.7) 2.0 (1.9—2.2) Achene pubescence villous villous-hirsute short pubescent Hypanthium pubescence glabrous sparse tomentulose Stamen series 2 2 (see text) 1 Number of stamens >80 52.0 (43-70) <30 Petal shape obovate intermediate spatulate Mountain, located respectively near the south and north ends of Saline Valley, Inyo County, California. On Cerro Gordo, C. mexicana var. stansburiana ranges from 1600 to 2200 m and P. glandulosa ranges from 1350 to 1800 m. At Waucoba Mt. the two species range from 1850 to 2100 m and from 1700 to 2000 m, respectively. Hybrids were growing along roads and in washes paralleling roads in a narrow alti- tudinal zone of species overlap. Fourteen apparent hybrids were found at the two localities. No other mixed stands or hybrids were en- countered in these mountains or in other ranges of the region. A variety of morphological characters distinguish the parental taxa and hybrids. Table 1 shows the most striking and consistent differ- ences. A hybrid-index value for each of the 49 plants studied was derived by assigning the character states given in Table 1 a value of 2 for Cowania-like characters and a value of O for Purshia-like char- acters; intermediate states were assigned a value of 1. Voucher spec- imens are deposited at UCSB. Fertility was estimated from pollen stained for at least 24 hours in 1 percent aniline blue in lactophenol. Five hundred pollen grains were scored from each of 10 specimens of each population. Ten flowers were sampled from each of the 14 putative hybrids. Pollen grains that stained evenly were considered viable; unstained and unevenly stained grains were considered inviable. Comparison of flavonoids of flowers and leaves was performed by standard techniques (Harborne, 1967, 1968, 1973; Mabry et al., 1970). Twelve flavonoids were identified and individual plants were com- pared by two-dimensional paper chromatography using identified compounds as markers. 1981] KOEHLER & SMITH: COWANIA/PURSHIA HYBRIDIZATION 17 20 15 10 NO. OF INDIVIDUALS 5 10 15 20 HYBRID-INDEX VALUES Fic. 2. Histogram of the hybrid-index values for 49 plants. The character states given in Table 1 were assigned a value of 20 for Cowania-like characters and a value of 0 for Purshia-like characters; intermediate states were assigned a value of 10. RESULTS Morphological comparison. Several floral and mature fruit char- acteristics distinguish C. mexicana var. stansburiana, P. glandulosa, and the putative hybrids (Table 1; Figs. 2—3). Some contrasting char- acters and a typical intermediate are shown in Figs. 4 and 5. Vege- tative characters such as leaf shape, number of lobes per leaf, leaf margin revolution, and glandularity, as used in earlier studies of hy- bridization between P. tridentata and C. mexicana var. stansburiana (Thomas, 1957; Stutz and Thomas, 1963), were too variable within and overlapping between populations to be of value. Field identifi- cation of individuals in overlapping and contiguous areas was difficult when both flowers and fruits were lacking. Pollen stainability. Cowania mexicana var. stansburiana aver- aged 93.4 percent stainable pollen, ranging from 84.0 to 99.0 percent 18 MADRONO [Vol. 28 KA ew # Style Length in Fruit “if + 9 7 8 a T ST SA a) Fa SRST SST (a * aug IER SS 1 2 3 4 5 6 Average Pistil Number/Plant Fic. 3. Scatter diagram for 49 plants. The symbol in the lower right-hand corner indicates the character numbers listed in Table 1. Black circle and full-length glyph indicate Cowania-like characters. Open circle and no glyph indicate Purshia-like char- acters. Intermediate characters score as half values. for all plants sampled from three populations. Purshia glandulosa averaged 81.2 percent stainable pollen, ranging from 22.8 to 99.4 percent for all plants sampled from four populations. Stainability of hybrid pollen was comparatively low, averaging 24.4 percent and ranging from 7.8 to 55.8 percent. Six hybrids were higher than the lowest value recorded for P. glandulosa, while three individuals of P. glandulosa were below the highest value recorded for hybrids. Forty- five percent of P. glandulosa individuals had percentages below the lowest value recorded for C. mexicana var. stansburiana. Flavonoid constituents. Twelve compounds were identified from bulk leaf extracts of C. mexicana var. stansburiana and P. glandu- losa. The two species and the hybrids possess the same leaf flavonoids (Fig. 6; Table 2). A corresponding analysis of petals showed that the two species differ significantly (Table 2). Petal extracts of the two species shared five quercetin glycosides. Purshia glandulosa petal ex- tracts yielded a rhamnetin 3-O-glycoside not found in petals of C. mexicana var. stansburiana, while the latter species contained the aglycone quercetin and five glycosides based on the aglycones luteolin, gossypetin and corniculatusin not found in petal extracts of P. glandu- 1981] KOEHLER & SMITH: COWANIA/PURSHIA HYBRIDIZATION 19 Fic. 4. Fruits of P. glandulosa (G), C. mexicana var. stansburiana (M), and a hybrid (X). losa. Flavonoids extracted from petals of hybrids matched the com- bined petal flavonoid complements of P. glandulosa and C. mexicana var. stansburiana (Fig. 7). No flavonoids unique to the hybrids were found. Six hybrids lacked sufficient petals to determine their flavonoid complements. DISCUSSION Hybridization occurred in the overlap zone of populations at two locations, separated by forty miles, in the Inyo Mountains of south- apy Fic. 5. Drawing of petals of P. glandulosa (G), C. mexicana var. stansburiana (M), and a hybrid (X). 20 MADRONO [Vol. 28 - OH OH OH 0 F1G. 6. The compounds isolated in this study were based on the aglycones repre- sented above (LUTEOLIN; R = R’ = R” = H. QUERCETIN; R = OH, R’ = R” = H. RHAMNETIN; R = OH, R’ = CH;, R” = H. GOSSYPETIN; R = R” = OH, R’ = H. CORNICULATUSIN; R = OH, R’ = H, R" = OCH). eastern California. Hybrids grew in obviously disturbed areas near roadsides. Although the surrounding terrain at both sites is very rug- ged, broken, and naturally disturbed, several trips to both sites yielded no other obvious hybrids. The onset of flowering in both species is gradual from low to high elevation and the flowers are short-lived. This tends to reduce pollen flow between the species at these sites because C. mexicana var. stansburiana occurs generally at the higher elevation and P. glandulosa at the lower. This is similar to the iso- lation between C. mexicana var. stansburiana and P. tridentata (Stutz and Thomas, 1963). Pollinator specificity is not known. No sterility because of differences in chromosome number is expected be- cause both species have n = 9 (Baldwin, 1951; Blauer et al., 1975). TABLE 2. FLAVONOID COMPOUNDS OF PETAL EXTRACTS OF C. mexicana VAR. stansburiana, P. glandulosa, AND PUTATIVE HyBrIbs. The 12 compounds are all found in and are the major flavonoid constituents of the leaves of the two species and the hybrids. Compound Cowania Hybrids Purshia Luteolin 7-0-Glucoside Gossypetin 3-0-Glucogalactoside Corniculatusin 3-0-Glucoside Corniculatusin 3-0-Rutinoside Corniculatusin 3-0-Diglycoside Quercetin (as the free aglycone) Quercetin 3-0-Glucoside Quercetin 3-0-Galactoside Quercetin 3-0-Rutinoside Quercetin 3-0-Glucoxyloside Quercetin 7-0-Glucoside Rhamnetin 3-0-Glycoside x x KX KK KK KK XK X x KX KX KK KK KK KK XK x xX KX XK XK XK 1981] KOEHLER & SMITH: COWANIA/PURSHIA HYBRIDIZATION 774 15% HoAc BAW ———> Fic. 7. Diagrammatic representation of the spots revealed on paper chromatograms of the petal extracts of C. mexicana var. stansburiana X P. glandulosa hybrids. The unshaded spots are common to both putative parental species, the shaded spots are specific to C. mexicana var. stansburiana, and the speckled spot is specific to P. glan- dulosa. The numbers refer to the compounds listed in Table 2. R;0.50 indicated by cross lines on axes. No other mixed populations or hybrids were encountered in the Inyo Mountains or other mountains of this region. Mountain ranges to the south and east, such as the Providence and New York Mountains of California, and the Spring and Palmetto Mountains of Nevada, con- tain numerous populations of each species but mixed stands are rare and no hybrids were observed. Contact between the species is rela- tively rare and in turn hybridization rare in the mountains near the California-Nevada border. Pollen stainability. Average pollen fertilities of the two parental species is high. Overall fertility of the fourteen hybrids is significantly reduced. The range (7.8—55.8 percent) suggests that the hybrids are varied in their genetic make-up. Pollen fertility of P. glandulosa was low compared with C. mexicana var. stansburiana and P. tridentata, which had an average of 94.5 percent and a range of 90.0—97.6 percent (Koehler, unpubl. data). This is consistent with the hypothesis that P. glandulosa is a segregation product of C. mexicana var. stansburi- ana X P. tridentata that retains genetic variability. With P. glandu- 22 MADRONO [Vol. 28 losa retaining heterozygosity one might expect a broad fertility range in F, hybrids of C. mexicana var. stansburiana X P. glandulosa. Morphological comparison. Morphological data demonstrate the overall intermediacy of hybrids in relation to the parents. Only one character, stamen series, failed to score in a range dissimilar to either species in any of the hybrids. Stamen insertion in hybrids was highly irregular but best characterized as two series. Except for stamen series, the hybrids had a character state distinct from either species for each of the observed characters. Interplant variation in most characters indicated a high degree of morphological variability comparable with that expressed in the broad range of pollen fertility. Flavonoid constituents. Although leaf flavonoid complements of C. mexicana var. stansburiana and P. glandulosa are the same, petal flavonoid complements are distinctive. Therefore, qualitative data showing addition of petal flavonoids in morphologically intermediate plants lends strong support to the theory of their hybrid origin. It has been shown that flavonoid compounds are often inherited as simple dominant characters, involving only one or a few genes (Alston, 1964; Brehm and Ownbey, 1965, 1968; Ownbey and Brehm, 1965), and addition of flavonoids has been observed in many studies of hybrid- ization (e.g., Alston et al., 1962; Alston and Turner, 1963; Smith and Levin, 1963; Crawford, 1974). A unique aspect of the results is that the sum of the twelve petal flavonoids is found in the leaves of both species. This indicates that both parental species have the same overall genetic complement for the production of flavonoids but that different modifier genes are op- erative in the production of petal flavonoids. Genes controlling fla- vonoid synthesis in these taxa apparently fall into three classes: those controlling general production, those modifying chemical structure, and those controlling distribution within the whole plant. The differ- ence in specific petal flavonoid complements is not a relatively simple one such as the production of variants that differ only slightly in structure, but is based on the presence or absence of different com- pounds that are themselves uncommonly substituted. These data indicate a close relationship between C. mexicana var. stansburiana and P. glandulosa in their identical leaf flavonoid com- plements, yet their petal flavonoid complements demonstrate a basic difference that supports morphological and ecological differences. The chemical data also emphasize the importance of studying floral as well as vegetative flavonoids. Relationship to previous Cowania-Purshia studies. Examination of the type specimen of Cowania mexicana var. dubia, collected by T. S. Brandegee in the Providence Mountains of southeastern Cali- fornia in 1902, and of a specimen Brandegee considered to be of the same variety, collected by C. A. Purpus on Morey Peak, Nevada in 1898, indicates that they are both hybrid plants derived from P. glan- 1981] KOEHLER & SMITH: COWANIA/PURSHIA HYBRIDIZATION 23 dulosa X C. mexicana var. stansburiana. Both specimens have char- acteristics ascribed to the hybrids of this study. Several comparisons, such as pollen stainability and flavonoid analysis, could not be made but their morphological characters fall easily into the ranges of hybrids examined in this study. It was considered by one author (Nord, 1965) that these specimens were hybrids between P. tridentata and C. mex- icana var. stansburiana. This suggestion can be negated on several points. The Providence Mountains are geographically removed from the southern limit of P. tridentata by approximately 268 km, but P. glandulosa and C. mexicana var. stansburiana populations can be found within a few kilometers of each other there. Morey Peak, Ne- vada, where Purpus collected his specimen, lies within the range of all three taxa. However, the specimen lacks distinctive P. tridentata characteristics, is similar to hybrids of this study, and was labelled Cowania Mexicana X Purshia glandulosa by Purpus, suggesting that it is also a hybrid derived from C. mexicana var. stansburiana and P. glandulosa. It is noteworthy that in 1898 Purpus recognized this plant as an intergeneric hybrid. There seems little doubt that hybridization occurs between C. mex- icana var. stansburiana and P. glandulosa. The data identify a well- established syndrome of hybridity: reduced fertility, morphological in- termediacy, and the addition of species-specific flavonoids. The range of pollen fertility was suggestive of hybridization beyond the first filial generation, but petal flavonoid complements of eight hybrid individ- uals are the exact summation of parental complements. These eight individuals had pollen fertility averages below 21 percent and were the lowest of the 14 discovered hybrids in this regard. If these plants were the result of backcrossing, selfing, or intercrossing of F, hybrids, the effects of segregation on the flavonoid constituents would be ex- pected in the petal flavonoid complements. This documentation of hybridization between C. mexicana var. stansburiana and P. glandulosa closes the ring of possible hybrid com- binations in this three-taxon complex and magnifies the question of the logic of the current classification. The occurrence of hybridization among the three taxa allows a strong case for merging the taxa into one genus. The data of this study and others (Thomas, 1957; Stebbins, 1959; Stutz and Thomas, 1963; Blauer et al., 1975) indicate a close genetic relationship that is clouded by seemingly major differences in carpel characters. However, without comprehensive study that in- cludes the other taxa of Cowania, submerging taxa that long have been considered distinct and are important range plants would be premature. ACKNOWLEDGMENTS We are grateful to the curator of UC for loan of specimens used in this study, to J. B. Harborne for discussions concerning chemical identification, to B. D. Tanowitz for assistance with illustrations, and to J. C. Hickman for editorial assistance. 24 MADRONO [Vol. 28 LITERATURE CITED ALSTON, R. E. 1964. The genetics of phenolic compounds. Jn: J. B. Harborne, ed., Biochemistry of phenolic compounds, p. 171-204. Academic Press, New York. and B. L. TURNER. 1962. New techniques in analysis of complex natural hybridization. Proc. Natl. Acad. USA. 48:130—-137. , R. N. LESTER, and D. HORNE. 1962. Chromatographic validation of two morphologically similar hybrids of different origin. Science 137:1048—1049. BALDWIN, J. T. 1951. Chromosomes of Spiraea and of certain other genera of Ro- saceae. Rhodora 53:203—206. BLAUER, A. C., A. P. PLUMMER, E. D. MCARTHUR, R. STEVENS, and B. C. GIUNTA. 1975. Characteristics and hybridization of important intermountain shrubs. I. Rose family. USDA For. Serv. Res. Pap. INT-169. BRANDEGEE, T. 1903. Flora of the Providence Mountains. Zoe 5:148-153. BREHM, B. G. and M. OWNBEY. 1965. Genetic analysis of 2-dimensional chromato- graphic pattern components in Tragopogon. Amer. J. Bot. 52:647 (Abstr.). and . 1968. Inheritance of C-glycosyl flavonoids in interspecific hybrid- izations of Tragopogon. Amer. J. Bot. 55:738 (Abstr.). CRAWFORD, D. J. 1974. A morphological and chemical study of Populus acuminata Rydberg. Brittonia 26:74-89. HARBORNE, J. B. 1967. Comparative biochemistry of the flavonoids. Academic Press, London. . 1968. Gossypetin and herbacetin as taxonomic markers in higher plants. Phy- tochemistry 8:177-183. . 1973. Phytochemical methods. Chapman and Hall, London. HEISER, C. B., JR. 1963. Artificial intergeneric hybrids of Helianthus and Viguiera. Madrono 17:118—-127. KNOBLOCH, I. W. 1972. Intergeneric hybridization in flowering plants. Taxon 21:97— 103; KRUCKEBERG, A. R. 1962. Intergeneric hybrids in the Lychnideae (Caryophyllaceae). Brittonia 14:311—321. KRUSE, A. 1974. Hordeum X Agropyron hybrids. Hereditas 78:291-294. KyuHos, D. W. 1967. Natural hybridization between Encelia and Geraea (Compositae) and some related experimental investigations. Madrono 19:33-43. Mapsry, T. J., K. R. MARKHAM, and M. B. THOMAs. 1970. The systematic identi- fication of flavonoids. Springer-Verlag, New York. MANGELSDoRF, P. C. and R. G. REEVES. 1939. The origin of Indian corn and its relatives. Texas Agric. Exp. Sta. Bull. 574. MorTENSON, T. H. 1970. Ecological variation in the leaf anatomy of Fallugia End1., Cowania D. Don, Purshia DC. ex Poir., and Cercocarpus HBK. (Rosaceae). Ph.D. dissertation, Claremont Graduate School. Munz, P. A. 1959. A California flora. Univ. California Press, Berkeley. Norpb, E. C. 1965. Autecology of bitterbrush in California. Ecol. Monogr. 35:307-— 334. OwnsBEY, M. and B. G. BREHM. 1965. Inheritance of biochemical characteristics in interspecific F, and F, hybrids of Tragopogon. Science 150:381 (Abstr.). PyL, L. VAN DER and C. H. Dopson. 1966. Orchid flowers: their pollination and evolution. Univ. Miami Press, Coral Gables, FL. PLUMMER, A. P., D. R. CHRISTENSEN, and S. B. MONSEN. 1968. Restoring big-game ranges in Utah. Utah State Div. Fish and Game Publ. 68-3. Rick, C. M. 1951. Hybrids between Lycopersicon esculentum Mill. and Solanum lycopersicoides Dun. Proc. Natl. Acad. USA. 37:741-744. . 1960. Hybridization between Lycopersicon esculentum and Solanum pennellit. Proc. Natl. Acad. U.S.A. 46:78-82. SMITH, D. M. and D. A. LEVIN. 1963. A chromatographic study of reticulate evolution in the Appalachian Asplenium complex. Amer. J. Bot. 50:952—958. 1981] KOEHLER & SMITH: COWANIA/PURSHIA HYBRIDIZATION 1S STEBBINS, G. L. 1950. Variation and evolution in plants. Columbia Univ. Press, New York. . 1959. The role of hybridization in evolution. Proc. Amer. Phil. Soc. 103:231- Zole Stutz, H. C. and L. K. THOMAS. 1963. Hybridization and introgression in Cowania and Purshia. Evolution 18:183-—195. THomMaAsS, L. K., JR. 1957. Introgression in Purshia tridentata (Pursh) DC. and Cow- ania stansburiana Torr. M.S. thesis, Brigham Young Univ. USDA ForREST SERVICE. 1937. Range plant handbook. Washington, D.C. USDA FoREST SERVICE. 1975. Seeds of woody plants in the U.S. Agric. Handbook 450. (Received 17 Jul 1977; accepted 15 Nov 1977; final version received 19 Sep 1980.) ANNOUNCEMENT PAPERS IN WESTERN PLANT ECOLOGY HONORING JACK MAJOR A symposium to honor Dr. Jack Major on the occasion of his retirement will be held at the University of California, Davis, on Friday, 29 May 1981, 9 a.m. to 5 p.m. Twelve to 13 papers will be presented, each 25 minutes long. People wishing to contribute papers are invited to submit abstracts to Dr. Michael Barbour before 15 February. The symposium will be co-sponsored by the California Botanical Society; the Botany Department, UC Davis; and UC Davis Institute of Ecology. The proceedings will be published in the Institute of Ecology series. Admission will be free. An honorary dinner will be held from 7-9 p.m. For further information and dinner reservations, call or write: Dr. MICHAEL BARBOUR, Botany Department, UC Davis 95616: (916) 752-2956. COVER OF PLANTS WITH EXTRAFLORAL NECTARIES AT FOUR NORTHERN CALIFORNIA SITES KATHLEEN H. KEELER School of Life Sciences, University of Nebraska, Lincoln 68588 ABSTRACT Percent cover of plants with extrafloral nectaries was investigated in three California habitats with the same physiognomy as habitats previously studied in Nebraska (pe- rennial native grassland, riparian forest, deciduous forest). In contrast to Nebraska where cover of plants with extrafloral nectaries reached 14 percent, no plants with extrafloral nectaries were found in any California transect. Chaparral was also studied; no plants with extrafloral nectaries were found. A number of plant species with extrafloral nectaries (EFNs) have been shown to be involved in a mutualism with ants (Bentley, 1977; Deuth, 1977; Tilman, 1978; Inouye and Taylor, 1979; Koptur, 1979; O’Dowd, 1979; Pickett and Clark, 1979; Keeler, 1980b). Extrafloral nectaries are glands on a plant that secrete a nectar rich in sugars and amino acids, but are not involved in pollination. For ecological stud- ies, function rather than morphology is considered the crucial aspect of the definition of an EFN. Little is known about the distribution of plants with EFNs. Bentley (1976) and Keeler (1979) found that 0-80 percent of the plants in tropical habitats (Costa Rica and Jamaica) had EFNs, and that this correlated with ant abundance. In Nebraska, Keeler (1980a) found O-— 14 percent of the cover to have EFNs. The percent cover of plants with EFNs in Nebraska was correlated strongly with abundance of foraging ants and secondarily with plant species diversity (H’), but not with rainfall or frost-free season. This study was undertaken to test for a relationship between biome and the abundance of plants with EFNs. Three California commu- nities with plant associations similar to the Nebraska ecosystems pre- viously studied (prairie, riparian forest, and deciduous forest) were analyzed. The similar habitats are probably the result of similar rain- fall (40-80 cm/yr). Adjacent chaparral was also investigated. METHODS Vegetation analysis was carried out in natural habitats using a point- intercept method. Plants nearest to randomly chosen points along a transect were identified and scored for presence of EF Ns. At each site, 500-1000 points were recorded on two transects. Simultaneously, abundance of ants was estimated using response of ants to karo syrup MaDRONO, Vol. 28, No. 1, pp. 26-29,12 February 1981 26 1981] KEELER: EXTRAFLORAL NECTARIES uf and tuna fish baits (about 1 ml of each at 25 spots 2 m apart). Number and species of ants on baits at 15, 30, 60, 120, and 180 min. were recorded; the number of baits found reflects forager density, and num- ber of species per bait probably indicates the diversity of the ant com- munity. The California sites were as follows: 1) Native grassland. Bunch- grass prairie 6.5 km north of Stonyford, Colusa County, on the Lodoga Rd., 370 m. This was a Stipa prairie with abundant annual forbs. 2) Riparian forest understory. Two sites along the Sacramento River, off Route 45 south of Princeton, Colusa Co., were studied; opposite Road 64, and along Reservation Road, 25 m. The canopy was dominated by cottonwoods, the understory by Vitis americana, Ribes spp., and Rhus diversiloba. 3) Deciduous forest understory. The study area was in Mendocino National Forest, Colusa Co., near Deafy Glade, 1370 m. The canopy was dominated by deciduous oaks (Quercus garryana, Q. kelloggii); the understory contained a variety of forbs, especially Lupinus sp. 4) Also studied was a chaparral site at 610 m in Men- docino National Forest along road 18N01 east of the Mill Creek camp- site. All studies were carried out 10-30 May 1979. This time was chosen for maximal plant and animal activity. The season had been cool and wet through the previous week. The days on which the study was carried out were warm, clear, and windy. Annual forbs were still flowering at the prairie site and Stipa was in bud. While it is possible that some extrafloral nectaries function at other times, the availability of water and the peak animal (both ant and herbivore) activity suggest April—May as the likely time for greatest EFN activity. Air temperature was noted for all experiments. It ranged from 16°C to 33°C for 0900-1200 hr PDST, when ant baiting was carried out. RESULTS Not one plant with EFNs was found in any of the transects or observed at any of these sites (Table 1). Frequency of ants discovering baits was astonishingly low, in particular at the prairie site, where in one three-hour experiment, not one of 25 pairs of baits was found by ants. The California sites had significantly fewer ant species per bait as judged by the Wilcoxon two-sample test (n = 10,6; C = 53, p< 0.01; Sokal and Rohlf, 1969). Adding in the chaparral sites, there are still significantly fewer ants per bait at the northern California sites (n = 10,8; C = 9; p < 0.001) than at the Nebraska sites. DISCUSSION No species with extrafloral nectaries were found at any of the Cal- ifornia sites. No native plants with extrafloral nectaries are known from the California habitats studied (pers. obs.; H. G. Baker, pers. 28 MADRONO [Vol. 28 TABLE 1. PERCENT COVER OF PLANTS WITH EXTRAFLORAL NECTARIES AND BAIT VISITATION BY ANTS FOR EIGHT TRANSECTS THROUGH FOUR VEGETATION TYPES IN NORTHERN CALIFORNIA. Percent % baits cover of # points found in Mean ant Sites EFN sampled 180 min. spp/bait Prairie 0 554 12 0.12 O 865 0) 0) Riparian forest understory 0) 1064 . 33 0.43 0 1045 55 0.55 Deciduous forest understory 0) 659 87 1.61 O 542 100 Zeal Chaparral 0 420 86 1.29 0) 544 86 152 comm., 1976). Some of northern California’s introduced species do have EFNs, and have been shown to benefit by mutualism with ants (Koptur, 1979). Furthermore, Helianthella californica (Asteraceae), a forb of the Sierra Nevada, has EFNs and is found at the same latitudes as the sites studied. However, all Nebraska transects had plants with EFNs except tallgrass prairie. Even in tallgrass prairie, species with EFNs were noted outside the transect (Keeler, 1980a). This suggests that in reality, plants with EFNs are less abundant in northern California than they are in comparable habitats in Nebraska. Why this type of plant defense is not favored in these northern Cali- fornia habitats is not clear at this time. ACKNOWLEDGMENTS I thank the administrators of the Mendocino National Forest for their cooperation, Dr. and Mrs. H. G. Baker for their advice and assistance, and S. Cantwell for her help. LITERATURE CITED BENTLEY, B. L. 1976. Plants bearing extrafloral nectaries and the associated ant community: interhabitat differences in the reduction of herbivore damage. Ecology 57:815—820. . 1977. The protective function of ants visiting the extrafloral nectaries of Bixa orellana L. (Bixaceae). J. Ecol. 65:27-38. DEuTH, D. 1977. The function of extrafloral nectaries in Aphelandra deppeana Schl. & Cham. (Acanthaceae). Brenesia 10/11:135—-145. INOUYE, D. W. and O. R. TayLor. 1979. A temperate region plant-ant-seed predator system: consequences of extrafloral nectar secretion by Helianthella quinquenervis. Ecology 60:1—7. KEELER, K. H. 1979. Frequency of extrafloral nectaries and ants at two elevations in Jamaica. Biotropica 11:152—154. . 1980a. The distribution of plants with extrafloral nectaries in temperate com- munities. Amer. Midl. Naturalist 104:274-280. 1981] KEELER: EXTRAFLORAL NECTARIES 29 1980b. The extrafloral nectaries of Ipomoea leptophylla (Convolvulaceae) Amer. J. Bot. 67:216—222. Koptur, S. 1979. Facultative mutualism between weedy vetches bearing extrafloral nectaries and weedy ants in California. Amer. J. Bot. 66:1016—1020. O’DowpD, D. J. 1979. Foliar nectar production and ant activity on a neotropical tree, Ochroma pyramidale. Oecologia 43:233-248. PICKETT, C. H. and W. D. CLARK. 1979. The function of extrafloral nectaries in Opuntia acanthocarpa (Cactaceae). Amer. J. Bot. 66:618-625. SOKAL, R. and F. J. ROHLF 1969. Biometry. W. H. Freeman and Co., San Francisco. TILMAN, D. 1978. Cherries, ants, and tent caterpillars: timing of nectar production in relation to susceptibility of caterpillars to ant predation. Ecology 59:686—692. (Received 8 Apr 1980; revision received and accepted 19 Sep 1980.) CHIOCOCCA HENRICKSONII (RUBIACEAE), A NEW SPECIES FROM THE CHIHUAHUAN DESERT REGION MARSHALL C. JOHNSTON Department of Botany, University of Texas, Austin 78712 ABSTRACT Two specimens from desertic mountains of Coahuila, Mexico, are described as Chio- cocca henricksonii, distinguished by its extreme xeromorphy, microphylly, and inflo- rescences that are reduced to solitary flowers. Tetramerous flowers are reported in the genus for the first time. Chiococca henricksonii appears to have relationships to C. alba, C. pachyphylla, and C. petrina. Chiococca henricksonii M. C. Johnston, sp. nov. Frutex humilis rupestris caulibus divaricatis hispidulis, folia coria- cea elliptica hispidula laminis 4—6(—9) mm longis, flores solitarii 4—5- meri, calycibus hispidulis (Fig. 1). Low shrub; branches numerous, divaricate, short, with internodes 1—4(—8) mm long, densely hispidulous with white, erect hairs 0.1 mm long; leaves coriaceous, hispidulous; blades elliptical, 4—6(—9) mm long; petioles 1—1.5(—2) mm long; flowers solitary, axillary; pedicels 1— 1.5(—2) mm long, hispidulous; sepals 4-5, oblong, obtuse, ca. 1 mm long, hispidulous, pale green at first, persistent and darkly pigmented in fruit; corolla “dark yellow” (Johnston 8738), funnelform, ca. 7 mm long including tube ca. 4.5 mm long, gradually dilated upward, and 4—5 obtuse spreading lobes (apparently valvate in bud), caducous; stamens 4—5, ca. 6 mm long, inserted at very base of corolla-tube, paddle-shaped, with free filaments ca. 3 mm long and oblong acute anthers ca. 3 mm long, about half exserted beyond the corolla-tube and reaching about the midpoint of the corolla-lobes; ovary hispidu- lous; drupe (apparently slightly immature) laterally compressed, sub- orbicular, ca. 4 mm long and broad not including persistent sepals, white; seeds 2, brownish, narrowly ovoid. TYPE: Mexico, Coahuila, 1-2 km n. of Puerto Colorado, near crest at sw. end of Sierra de la Fragua (near 26°45’N, 102°30'W), high lime- stone ridge with forest of Pinus pinceana Gordon, shrub flat against rock, rooted in solution-holes, fruit white with black stigma, above 1750 m, 2 Sep 1941, J. M. Johnston 8738. (Holotype: LL; isotypes: GH, MICH). PARATYPE: Mexico, Coahuila, ca. 2.4 km sw. of Las Delicias on e. side of Sierra de las Delicias, at margin of scree-slope along canyon MADRONO, Vol. 28, No. 1, pp. 30-32, 12 February 1981 30 1981] JOHNSTON: CHIOCOCCA 31 Fic. 1. Chiococca henricksonii M. C. Johnston, drawn from I. M. Johnston 8738. a. Mature stem showing characteristic shoots after leaves fall. b. Mature stem with leaves of current season and fruit. c. Leaf, abaxial view. d. Flower. e. Anther, adaxial view. f. Almost mature fruit, showing subtending bracts. ca. 400 m above main spring (26°14’N, 102°49’W), with Agave lech- eguilla, Hechtia, Acacia, Dasylirion, Tecoma, Viguiera, Leucophyl- lum, Euphorbia, small shrub with snow-white fruits and one fasciated branchlet, ca. 1600 m, 15 Aug 1973, Henrickson 12471 (LL, MEXU). In Standley (1926), Johnston 8738, with its constantly tetramerous flowers, keys to the apparently closely related, monotypic, Yucatanian Asemnanthe Hooker f. But Henrickson 12471, otherwise identical to Johnston 8738, has constantly pentamerous flowers as in the rest of the genus Chiococca. I conclude that in this case the number of sepals 39 MADRONO [Vol. 28 and petals is not of taxonomic value, but that the diagnosis of Chio- cocca must be modified to take into account rare cases of tetramery. Without much more thorough study, I decline to pass judgment on the merit of retaining Asemnanthe as a distinct genus. Chiococca henricksonii is apparently related to C. alba (L.) Hitch- cock, which is widespread in warmer parts of America, and to C. pachyphylla Wernham of the Sierra Madre Oriental. It is quite distinct in habit, foliage, pubescence, and inflorescence from those two species. Chiococca henricksonii may find its closest relative in the almost equally xerophytic and microphyllous C. petrina Wiggins of Sonora and extreme western Chihuahua. The leaf-blades of C. petrina av- erage slightly larger than those of C. henricksonii and its larger flowers are borne in racemes. The two known localities for Chiococca henricksonii are only 70 km apart and lie in some of the driest, most rigorous parts of the Chihuahuan Desert Region in southwestern Coahuila, where a num- ber of other local endemic species have been discovered. Although gypseous substrates are common in the general region, the substrates at the two localities are essentially pure limestone (Henrickson, pers. comm., 1980; T. L. Wendt and E. Lott, pers. comm., 1980). I thank Jim Henrickson for permitting study of his collections, which were made with the support of the Henrickson Research Fund. LITERATURE CITED STANDLEY, P. C. 1926. Trees and shrubs of Mexico, part V. Contr. US Natl. Herb. 23:1313-1721. (Received 20 Mar 1980; accepted 28 Jul 1980; final version received 7 Aug 1980.) A NEW SUBSPECIES OF COMAROSTAPHYLIS POLIFOLIA (ERICACEAE) FROM COAHUILA, MEXICO JAMES HENRICKSON Department of Biology, California State University, Los Angeles 90032 ABSTRACT Comarostaphylis polifolia subsp. coahuilensis is described from the mountains of the central Chihuahuan Desert in Coahuila, Mexico. A discussion of fruit structure supports the generic separation of Comarostaphylis from Arctostaphylos and Arbutus. Recent field studies in connection with M. C. Johnston’s Chihua- huan Desert Flora Project have brought forth collections of a distinct taxon of Comarostaphylis from six mountain ranges in south and cen- tral Coahuila. These collections appear to lie within the general vari- ation that Standley (1924) advocated for Arctostaphylos (Comarostaphylis) polifolia H.B.K. Comarostaphylis was first segregated from Arctostaphylos by Zuc- carini (1837) on the basis of gynoecial and fruit characters. The gy- noecium of Comarostaphylis consists of 5 (rarely 4) carpels, each with 1 ovule, and the fruit is a drupe with a reddish to blackish, granular or warty exocarp, a thin mesocarp, and a solid, bony, thick-walled, spheroidal endocarp stone that contains (4—)5 seeds each with a distinct conical grayish cap on the apical end. Arctostaphylos, in contrast, may have 5—9(—10) carpels each with 1 ovule, but the mature fruit is brownish, has a smooth exocarp, mealy mesocarp, and seeds are en- closed in separate portions of the endocarp, which may then consist of 5—9(—10) separate segments. In other species of Arctostaphylos, the endocarp segments may be variously combined into groups of 2 or 3, or in one species (A. glauca Lindl.), as in Comarostaphylis and Xy- lococcus, they are combined into a single, solid, but often vertically ridged endocarp stone. Fruits of Comarostaphylis are superficially similar to those of Arbutus in that both have warty or granular exocarps. Arbutus, however, has several ovules per carpel and the seeds are enclosed in a cartilaginous to lignified endocarp wall that is open, i.e. not lignified, along the dorsal trace of each carpel. I cannot agree with the placement of Comarostaphylis within Arc- tostaphylos as has been done in Standley (1924) and Standley and Williams (1966). I support the contention (Adams, 1940) that each of these genera, as well as Xylococcus and Ornithostaphylos, is worthy of generic recognition. MADRONO, Vol. 28, No. 1, pp. 33-37, 12 February 1981 33 34 MADRONO [Vol. 28 Small (1914) recognized Comarostaphylis as a distinct genus with 22 species, six of which he described as new. Standley (1924) placed Comarostaphylis within Arctostaphylos, recognizing only 14 species and combining five of Small’s species into his Arctostaphylos polifolia H.B.K. with a comment that the key characters used by Small (1914) to distinguish these segregates were “utterly worthless”. As recognized by Standley (1924), the delineation of taxon polifolia, one of the oldest names within Comarostaphylis, has been greatly broadened to include specimens with puberulent, canescent, or glandular stems and inflo- rescences and entire, linear to oblong-lanceolate, revolute to flattened, glabrous to glaucous, puberulent, villous to somewhat glandular leaves. This entire complex of taxa is also quite variable in floral features. Ovaries can be glabrous to weakly pilose or villous and co- rolla size also varies through the range of the species. At present the actual circumscription of the species is not known as a comprehensive study has not been made of the variation encountered in the field. Our specimens from the mountains of the Chihuahuan Desert appear to fit within the morphological range of C. polifolia in the broadest sense. To name this new taxon at a rank equal to C. polifolia is not defen- sible, in my opinion. I here present the taxon as a new subspecies of C. polifolia in anticipation that further studies will recognize addi- tional, geographically distinct, minor variants of the species at the rank of subspecies (e.g., A. novoleonsis Rehder). Comarostaphylis polifolia (H.B.K.) Zucc. subsp. coahuilensis Henrickson, subsp. nov. Frutices saepe humiles nodosi vel arbores parvae ad 1.5 m altae; rami hornotini purpureo-brunnei, dense albo-puberulentes vel canes- centes. Laminae foliorum oblongo-ellipticae vel anguste oblongo-ova- tae, planae vel conduplicatae ad costum, apice saepissime mucronatae, sparse puberulentes, supra glabratae nitidae, subtus glaucae subper- sistente pubescentes. Racemi terminales puberulentes pilis crispis glandulosis; lobi calycis deltoidei, glandulosi puberulentesque ciliati roseoli; corolla ovoidae vel elliptoco-urceolatae, pallide roseolae vel roseolae, lobis orbicularibus vel ovatis reflexis (Fig. 1). Low, rounded, somewhat gnarled shrubs to miniature trees 0.2—1.5 m high; young stems maroon-brown, densely white puberulent, hir- tellous to subcanescent with trichomes 0.05—2 mm long; bark reddish brown, flaking, weathering gray in old stems. Leaves congested at stem tips; petioles 3-7 mm long, flattened to grooved adaxially, pu- berulent as young stems; blades oblong-elliptical to narrowly oblong- ovate, oblong-obovate, 2—4(—4.5) cm long, 7—11(—20) mm wide, sub- coriaceous to coriaceous, flat or conduplicately folded along midrib, obtuse to rounded or acute but mostly mucronate at apex, rounded to cuneate at base, entire, rarely slightly toothed, revolute to undulate, callose-thickened at margins, sparsely puberulent to glabrate, shining 1981] HENRICKSON: COMAROSTAPHYLIS 35 / Ly Fic. 1. Comarostaphylis polifolia subsp. coahuilensis Henrickson. A. Stem show- ing orientation of leaves and inflorescence (Johnston et al. 10842, LL). B. Flower with glandular sepals (Johnston et al. 10842, LL). C. Immature fruit on glandular-puber- ulent pedicel with bracts and rachis shown (Henrickson 13600, TEX). D—E. Extremes in leaf size from original collections. D. Subcoriaceous large leaf with slight tooth formation (Chiang et al. 9075, LL). E. Coriaceous leaf, folded along midrib (Johnston et al. 11682, LL). green above, sparsely but more persistently puberulent with trichomes O.1—0.5 mm long, glaucous, gray-green to yellow-green beneath. Ra- cemes terminal, 2.5—4.5 cm long, rachis and pedicels curly-puberulent and with stalked glands 0.2—0.5 mm long, glands red or not; pedicels 3.7—7(—10) mm long, with 3 bracts, the basal to 6 mm long, the upper bracts narrowly acuminate, to 1 mm long. Flower calyx 4-5 mm broad, lobes to 1.5 mm long, triangular, acute to acuminate, glandular and puberulent, ciliate, reddish; corollas ovoid to elliptically urceolate, pale pink to rose, 5—8(—9) mm long, 3-5 mm wide, glabrous without, pilose within, lobes erect to reflexed, orbicular to ovate, 0.8—1.0 mm long, 0.8—-1.5 mm wide, ciliate, papillate within; stamens 10, filaments 1.3—2.2 mm long, dilated, pilose at base; anthers ovoid 0.9-1.2 mm long, pink, appendages 0.4—0.6 mm long; ovary sparsely pilose, basal disk weakly 10-ribbed, sparsely pilose-ciliate; style 4-5 mm long. Drupe 4—6 mm in diameter; endocarp 3.0-3.5 mm broad, spheroidal; 36 MADRONO [Vol. 28 seeds ovoid, 1.7—2.2 mm long, the gray conical cap one-third its length. TYPE: Mexico, Coahuila, Canon de Calabazas in Sierra Mojada, s. of Esmeralda (near 27°16'N, 103°41’W), flowers pink, 6 May 1973, Johnston, Wendt, and Chiang 10881. (Holotype: LL; isotype: MEXU). PARATYPES: Mexico, Coahuila, Sierra Mojada, s. of La Esmeralda (near 27°16'N, 103°41’W), 1 Sep 1972, Chiang et al. 9075 (LL): Sierra de Parras, ca. 16 km (10 mi) w. of Parras de la Fuente (near 25°26'N, 102°16'W), 4 Nov 1972, Chiang et al. 10061 (LL); e. face of Sierra de Almagre (near 27°36'N, 103°53’W), 5 May 1973, Johnston et al. 10842 (LL); n. side Sierra de Paila, Mina El Aguirreno (near 26°06'N, 101°36’W), 5 Jul 1973, Johnston et al. 11682 (LL): Sierra de Paila, upper Canon Corazon del Toro (near 25°54'N, 101°38’W), 5 Nov 1972, Wendt et al. 10101 (LL); n. side Sierra de Organos (near 26°41'N, 103°03'W), 8 Aug 1973, Henrickson 12148 (TEX); s. part Sierra de los Organos (near 26°44'N, 103°01’W), 8 Aug 1973, Johnston et al. 12143 (LL); crest of Sierra de la Madera, above Canon de la Hacienda (near 27°03'N, 102°24’W), 27 Sep 1973, Henrickson 13600 (TEX). The new taxon is known only from limestone areas at 1500—2800 m where it occurs in open chaparral with Quercus spp., Acacia, Dasy- livion, Leucophyllum, Fraxinus greggiit, Sophora, Lindleya, Ptelea, Agave and on the southern crests of mountains in forests of Pinus strobiformis, P. arizonica, Abies coahuilensis, Pseudotsuga menziesit, Quercus greggii, and Cupressus arizonica, often in association with Arctostaphylos pungens and Philadelphus. The new taxon fits well into the C. polifolia complex as interpreted by Standley (1924) but is distinguished by the tendency to have a small shrub habit, often conduplicately folded, oblong-lanceolate to flat, ob- long-ovate leaves, closely pubescent to hirtellous stems, glandular-pu- bescent rachis and pedicels, and by its northern distribution. When comparing this new taxon with the species recognized by Small (1914) in his North American Flora treatment it would tend to key to C. hartwegiana Klotzsch from which the new taxon differs in broader leaves, non-glandular twigs, and more northern distribution. The new taxon also fits within the general description of C. caeciliana (Loes.) Small from Oaxaca but is not so tomentose. The nine collections on the new taxon are uniform except for small differences in leaf shape, those from more exposed sites tending to have smaller, more oblong-elliptical, entire, coriaceous leaves (Fig. 1E) than those of plants of presumed more protected sites (Fig. 1D). One specimen from the Sierra de la Paila (Wendt et al. 10101), lacks glandular hairs on the inflorescence while others in that collection are glandular (Fig. 1C). The new taxon is also variable in growth habit. 1981] HENRICKSON: COMAROSTAPHYLIS 37 On exposed cliffs plants only 1.5 dm tall but 1 m broad have been observed. In protected sites plants may develop into small trees. ACKNOWLEDGMENTS I thank Marshall Johnston for the Latin diagnosis, Frances Runyan for delineation of the illustration, George Diggs for commenting on the manuscript, Lynn Marshall for manuscript typing, and the Plant Resource Center at the University of Texas for use of facilities. LITERATURE CITED ADAMS, J. E. 1940. A systematic study of the genus Arctostaphylos Adans. J. Elisha Mitchell Sci. Soc. 56:1—62. REHDER, A ‘1935. Some new trees from Mexico. J. Arnold Arbor. 16:488—452. SMALL, J. x. 1914. Ericaceae. North American Flora 29:33-102. STANDLEY, P. C. 1924. Arctostaphylos. In: Trees and shrubs of Mexico. Contr. US Natl. Herb. 23:1094—1099. STANDLEY, P. C. and L. O. WILLIAMS. 1966. Ericaceae. Jn: Flora of Guatemala. Fieldiana: Bot. 24(8):88—127. ZUCCARINI, J. G. 1837. Abh. Math.-Phys. Cl. Konig]. Bayer. Akad. Wiss. 2:331. (Received 14 Feb 1980; accepted 22 Jul 1980; final version received 18 Aug 1980.) NOTEWORTHY COLLECTIONS WOLFFIA PUNCTATA Griseb. (LEMNACEAE).—USA, CA, San Diego Co., Lake Hodges, 5 km s. of Escondido, e. side of hwy I-15 (33°3'16"N, 117°3'45”W), 152 m, 14 Jun 1980, Armstrong s.n. (SD 105410). Scattered individuals, with ave. density of 10— 12 per 250 ml of water, floating at surface near shoreline among dense homogeneous population of Lemna gibba. Verified by R. F. Thorne, Jun 1980. Previous knowledge. Known from WA, OR, c. and ne. US, s. to the West Indies. A minute, free-floating rootless angiosperm, barely visible without magnification. Often associated with Lemna, Spirodela, and Azolla. The genus has likely been overlooked many times because of its small size. (Herbaria consulted: RSA, SD; published sources: Daubs, Ill. Biol. Monogr. 34. 1965; Mason, A fl. marshes Calif. 1957.) Significance. First record of Wolffia in s. CA, a se. extension for W. punctata of 975 km from Fall River Mills, Shasta Co. According to Daubs (op. cit.), two spp. of Wolffia are native to CA: W. punctata and W. columbiana. W. arrhiza (and its syn- onym, W. cylindracea) are listed, apparently incorrectly, for CA by Mason (op. cit.).— WAYNE P. ARMSTRONG, Palomar College, San Marcos, CA 92069. (Received and ac- cepted 16 Jul 1980; final version received 24 Jul 1980.) OROBANCHE UNIFLORA L. subsp. OCCIDENTALIS (Greene) Abrams ex Ferris (ORO- BANCHACEAE).—USA, AZ: Gila Co., 3-bar Wildlife Area (e. of Four Peaks), 30 Apr 38 MADRONO [Vol. 28 1958, Dick Saunders s.n. (ASU); Mohave Co., Hualapai Mts., upper Frees Wash on n. side of Dean Peak: T20N R15W S16 nw.'%, steep canyon, one plant seen, 1900 m, 16 Jun 1979, Butterwick 5149 & Parfitt (ASU, JEPS); T20N R15W S16 sw.%, one clump in moist humus of stream bed near Evrigeron and Solidago, some flowers un- derground (cleistogamous?) and producing fruits, 2070 m, 11 Aug 1979, Butterwick 5435 & Parfitt (ASU, JEPS). Verified by L. R. Heckard, 1979. Previous knowledge. The species is known from Newfoundland to Quebec and Yu- kon, s. ton. FL. TX, and s. CA. AZ is specifically mentioned (Thieret, J. Arnold Arboretum 52:425. 1971) as lacking a report of O. uniflora. Identification of the subsp. follows the treatment of Abrams and Ferris, which recognizes two subspecific taxa of w. N. Amer. in addition to the typical subsp. of e. N. Amer. According to Heckard (pers. comm., 1980) the infraspecific taxa are not sharply defined and a critical reex- amination of the species is needed. (Herbaria consulted: ARIZ, ASC, ASU, MNA; published sources: Abrams and Ferris, Illus. fl. Pac. States. 1960; Munz. A fl. S. Calif. 1974; Hitchcock et al., Vasc. pls. Pac. Northw. 5. 1959; Watson, Syst. f Orobanche sect. Gymnocaulis, M.A. thesis, Calif. State Univ. Chico. 1975; Barkley, ‘Man. fl. pls. Kans. 1968; Correll and Johnston, Man. vasc. pls. Tex. 1970; Harrington, Man. pls. Colo., ed. 2. 1964.) Significance. First records for AZ, a se. range extension of the subsp. The AZ specimens will be an important link in a comparative study of the e. and w. populations of the species (Heckard, pers. comm., 1980). Underground (cleistogamous?) flowers and fruits have not been reported previously.—BRUCE D. PARFITT, Department of Botany and Microbiology, Arizona State University, Tempe 85281 and MARY BUTTERWICK, Bureau of Land Management, 2929 W. Clarendon Ave., Phoenix, AZ 85017. (Received 28 Mar 1980; accepted 10 Jul 1980; final version received 28 Jul 1980.) HAZARDIA ORCUTTII (Gray) Greene (COMPOSITAE).—USA, CA, San Diego Co., Heri- tage Park housing development e. of El Camino Real, s. of Encinitas (near 33°03'N, 117°15'W), 75 m, 25 Aug 1979, Oberbauer 188 (SD). A vigorous population of several hundred individuals on crumbly clay soil at the interface of coastal sage scrub and chaparral. Verified by Reid Moran, Sep 1979. Previous knowledge. Known only from nw. Baja Calif. along coastal plains and hills from Colonet to about 5 km s. of the international border at La Joya (Clark, Madrono 26:105—127. 1979; R. Moran, pers. comm., 1979). Significance. First record for USA and CA, a disjunction of 60 km. Found in an area approved for a housing development. The environmental impact report for the development failed to mention this species as well as several other “rare and endan- gered” plants.—THOMAS A. OBERBAUER, Department of Planning and Land Use, County of San Diego, 1600 Pacific Coast Highway, San Diego, CA 92101. (Received 18 Aug 1980; final version received and accepted 19 Sep 1980.) 1981] NOTEWORTHY COLLECTIONS 39 Fieldwork for the following collections was supported by a grant from the Pacific Southwest Forest and Range Experiment Station, Berkeley, to the Biology Department at California State University, Fresno. Herbaria consulted for all taxa: CAS, DS, FSC, JEPS, UC; published sources for all taxa: Abrams, Illus. fl. Pac. States 1949-1960; Hitchcock, Man. grasses U.S. 1935; Hitchcock and Chase, Man. grasses U.S. 1950; Munz, A Calif. fl. 1959; Powell, Inv. rare endang. vasc. pls. Calif. 1974; Smith et al., Inv. rare endang. vasc. pls. Calif., ed. 2. 1980. All Evans, Haines, and Haines and Evans vouchers are deposited in both FSC and JEPS. MADIA SUBSPICATA Keck (ASTERACEAE).—USA, CA, Fresno Co.: Haslett Basin, USFS engineer’s camp (T11S R25E S23), 568 m, 29 May 1975, Haines and Evans C- 75373; along hwy 168, 1.6 km above Tollhouse Village (T10S R24E $31), 768 m, 8 May 1953, Quibell 1868 (FSC); 6 km w. by air from Prather, ne. corner Table Mt. (T10S R22E S820), 560 m, 15 Apr 1978, Haines 78006; Deep Creek Basin (T11S R25E $19- 20), 792 m, 24 May 1977, Haines and Evans 77036. Growing under the driplines of trees and shrubs in foothill-woodland. Common in these scattered populations; associ- ated with Quercus douglasii, Q. wislizenii, Arctostaphylos mariposa, Bromus diandrus, and Erodium cicutarium. Previous knowledge. Scattered sites from Mariposa to Butte cos. (Munz). Significance. First records s. of Mariposa Co.; a range extension of ca. 100 km. Con- sidered, but not included by Smith et al. (op. cit.). RAFINESQUIA CALIFORNICA Nutt. (ASTERACEAE).—USA, CA: Fresno Co.: Haslett Basin, USFS engineer’s camp (T11S R25E S23), 568 m, 6 Jun 1975, Haines and Evans C-75411; hwy 180, 15 kme. of Centerville (T14S R24E $14), 175 m, 2 Jun 1974, Haines 74214; Inyo Co.: Great Falls Canyon, 11.3 km n. of Trona, 806 m, 15 Apr 1940, Alexander and Kellogg 1077 (UC); Johnson Canyon, Panamint Range, 1075 m, 27 Apr 1940, Jepson 1972 (JEPS); Stanislaus Co.: Hill n. of Del Puerto Canyon, 13 Apr 1940, Hoover 4338 (JEPS); Arroyo del Puerto, Mount Hamilton Range, 461 m, 9 May 1935, Mason 8311 (UC); Calaveras Co.: Stanislaus River at crossing of Copperopolis-Sonora Road, 154 m, 22 May 1921, Tracy 5709 (JEPS). Previous knowledge. Cismontane California n. to Humboldt and Kern cos.; to UT, AZ, L. Cal. Most frequent in disturbed places. Significance. Extends dist. in Sierra Nevada n. to Calaveras Co. and e. to Inyo Co. A range extension of ca. 400 km. CRYPTANTHA MURICATA (H. & A.) Nels. & Macbr. var. MURICATA (BORAGINA- CEAE).—USA, CA: Fresno Co.: Haslett Basin, USFS Rd. 10S69, 4.5 kme. of Rd. 10S02 (T11S R25E S24), 870 m, open areas with Mimulus gracilipes, M. bolanderi, M. viscidus, and Ceanothus cuneatus, 19 May 1975, Haines and Evans C-75310; Jose Basin (T10S R23E S4), 768 m, in an open area with M. gracilipes, Eriodictyon cali- fornicum, and Plagiobothrys tenellus, 30 Apr 1974, Haines 74203; above Sampson Flats, below Delilah Lookout (T13S R26E $13), 1382 m, 24 May 1941, Simonian 849 (FSC); Madera Co.: 2.4 km se. of South Fork, e. of Mammoth Rd. (T8S R23E $17), 920 m, 5 Jun 1957, Brock 159 (FSC). Previous knowledge. Gravelly or rocky soils, many plant communities; Coast Ranges from Contra Costa Co. s., Sierra Nevada of Kern Co. to Orange Co. Significance. First Sierran records n. of Kern Co., an extension of ca. 160 km. 40 MADRONO [Vol. 28 CAREX TUMULICOLA Mkze. (CYPERACEAE).—USA, CA, Fresno Co.: Haslett Basin, USFS engineer’s camp (T11S R25E S23), 568 m, 25 Mar 1975, Haines and Evans C- 75089, common in moist areas with Plectritis californica, Bromus mollis, Plagiobothrys nothofulvus, and Erodium obtusiplicatum. Previous knowledge. Meadows and grassy slopes; Santa Cruz Id., Coast Ranges from Monterey to Del Norte and Siskiyou cos., Madera to Tuolumne cos. in the Sierra Nevada, into w. OR and WA. Significance. First record for Fresno Co., a range extension of ca. 35 km. EPILOBIUM MINUTUM Lindl. ex Hook. (ONAGRACEAE).—USA, CA, Fresno Co.: Haslett Basin, USFS Rd. 10S69, 4.5 kme. of Rd. 10S02 (T11S R25E S24), 768 m, 10 May 1975, Haines and Evans C-75237, openings in chaparral, around the bases of shrubs with Ceanothus cuneatus, Quercus wislizenii, Mimulus viscidus, and Cryptan- tha muricata. Previous knowledge. Dry open places, Coast Ranges from Santa Barbara Co. to Del Norte and Siskiyou cos.; Sierra Nevada from Madera Co. n.; to B. C., NE. Significance. First record for Fresno Co., an extension of ca. 35 km s. ARGEMONE MUNITA Dur. & Hilg. subsp. ROTUNDATA (Rydb.) G. Ownbey (PAPA- VERACEAE).—USA, CA, Fresno Co.: w. slope of Bald Mt., 4.5 km n. of Dinkey Cr. Rd. (T9S R25E 826), 2058 m, 20 Jul 1974, Haines and Evans C-74023; Tamarack Ridge, 5 kme. of hwy 168 (T9S R26E S18), 2235 m, 16 Sep 1975, Haines and Evans C-75648; Sugar Pine Hill near Rancheria Cr. (T11S R27E S25), 2012 m, 3 Oct 1975, Haines and Evans C-75714; USFS Rd. 10S70, 1.3 km n. of McKinley Grove Rd. near Tule Meadow (T11S R27E S4), 2181 m, 3 Oct 1975, Haines and Evans C-75716; along road to Mushroom Rock, 6.4 km w. of Huntington Lake Rd. (T8S R25E $17), 2335 m, 3 Oct 1975, Haines and Evans C-75708. All locations are on disturbed sites in mixed conifer forest, associated with Abies concolor, A. magnifica, and Ceanothus cordulatus. Previous knowledge. Mts. of Mojave Desert, San Bernardino and San Gabriel mts., e. slope Sierra Nevada to Shasta Co.; local populations in Lake and Colusa cos. (Munz). Significance. First records for w. Sierra Nevada. APERA SPICA-VENTI (L.) Beauv. (POACEAE).—USA, CA: Fresno Co., along USFS Rd. 10S02, 1.6 km ne. of the jct. of Rd. 10S69 in Haslett Basin (T11S R25E $14), 838 m, 6 Jun 1975, Haines and Evans C-75419, commonly distributed on open slopes with Quercus wislizenii, Bromus diandrus, B. mollis; San Luis Obispo Co.: hwy 166, 41 km w. of New Cuyama (T12N R30W 827), 427 m, 2 Nov 1975, Evans 75053, with B. diandrus and B. mollis in grassy openings. Previous knowledge. European weed, scattered locations from OR to ME (Hitch- cock and Chase). Diagnostic characters. Keys to Agrostis tenuis var. aristata in Munz (p. 1520) but is long awned from near the tip of the lemma (vs. awned from near the base of the lemma) and lemma firm at maturity (vs. membranaceous at maturity). Verified with European collections in UC. Significance. First records for CA, a range extension of ca. 1200 km from Portland, OR. RHAMNUS RUBRA Greene subsp. YOSEMITANA C. B. Wolf (RHAMNACEAE).—USA, CA: Fresno Co.: w. slope Bald Mt., 4.5 km n. of Dinkey Cr. Rd., 2058 m, 4 Aug 1975, Haines and Evans C-75611, growing in rocky areas with Quercus kelloggii, Arctostaphylos patula, and Sitanion hystrix; between Zumwalt Meadow and Bubbs Cr., Kings Canyon, 28 Jul 1940, Howell 15601 (CAS); Huntington Lake, 2150 m, 16 Jul 1917, Grant 1097 (JEPS); Florence Lake, 2243 m, 5 Jul 1952, Raven 4242 (CAS); Tulare Co.: Hockett Meadow, 2611 m, Jun 1870, Purpus 1785 (UC); Mineral King, 24 Jul 1942, Howell 17096 (CAS); Inyo Co.: Summes Creek, 2304 m, 18 Jul 1955, Raven 8313 (CAS); Calaveras Co., Big Tree Meadow, 1536 m, May 1865, Davy 1543 (UC); 1981] NOTEWORTHY COLLECTIONS 41 Sierra Co., w. side of Yuba Pass, 1536 m, 21 Aug 1951, Howell 28279 (CAS); Butte Co., 3.2 kms. of Big Bar Mt., 8 Jun 1941, Quick 41-09 (CAS). Previous knowledge. Known only from Tuolumne, Mariposa, and Mono cos. (Munz). Significance. Extends range ca. 105 km n. to Sierra Co., ca. 130 km s. to Tulare Co., and ca. 35 km e. to Inyo Co. MIMULUS GRACILIPES Rob. (SCROPHULARIACEAE).—-USA, CA, Fresno Co.: Haslett Basin, USFS engineer’s camp (T11S R25E S23), 568 m, 11 May 1975, Haines and Evans C-75245, growing with M. bolanderi, M. viscida, Ceanothus cuneatus; Haslett Basin, along USFS Rd. 10869 (T11S R25E S24), 4.5 km e. of road 10S02, 870 m, 13 Apr 1975, Haines and Evans C-75142, common in small openings in chaparral, with Plagiobothrys tenellus, Quercus wislizeni1, Cryptantha muricata, M. viscidus, M. bo- landeri; jct. of Mill Cr. Rd. and Sand Cr. Rd., 0.8 km w. of Miramonte, 950 m, 1 May 1960, Linderman s.n. (FSC); Carpenteria Botanical Area, 6.4 km w. of Lodge Rd. on hwy 168 (T10S R23E S23), 973 m, 2 May 1974, Stebbins 74123 (FSC); N. Burrough Rd. 0.8 km e. of Burrough Valley Rd., 522 m, 17 May 1964, Hardwick 144b (FSC); 7.6 km e. of Auberry Rd., 925 m: 30 May 1963, Weiler 63120 (FSC); 7 May 1964, Webb 0108 (FSC); Auberry Rd., 1137 m, 25 Apr 1941, Petersen 316 (FSC); upper Jose Basin, 1137 m, 22 May 1957, Quibell and Brock 6 (FSC); 32 km from Auberry on USFS Rd. 9807, 1230 m, 24 Apr 1962, Kirkhart 9 (FSC). Previous knowledge. Mormon Bar, Mariposa Co. (Munz). Significance. Extends the known range ca. 120 kms. These collections are the basis for the designation “rare but not endangered” in Smith et al. (op. cit.).—ROBERT D. HAINES, Tulare County Department of Agriculture, Visalia, CA 93277 and CHARLES J. EvANS, Pacific Southwest Forest and Range Experiment Station, 2081 E. Sierra Avenue, Fresno, CA 93710. (Received 3 Jun 1980; revision received and accepted 19 Sep 1980.) Herbaria consulted for all taxa: ARIZ, ASU, COLO, CS, MO, NMC, RM, UNM, UT, UTC; published sources for all taxa: Barneby, Contr. N.Y. Bot. Gard. 13. 1964; Blake, Leafl. W. Bot. 6:71. 1950; Cronquist, Brittonia 6:121—300. 1947; Cronquist et al., Intermt. fl. 1. 1972; Epling, Repert. Spec. Nov. Beih. 110:1—380. 1939; Federal Register 41(117):24523-24572. 16 Jun 1976; Fernald, Gray’s Man. Bot., ed. 8. 1950; Gould, The grasses Tex. 1975; Harrington, Man. pls. Colo. 1964; Kearney and Peebles, Ariz. fl., ed. 2. 1960; Martin, Brittonia 7:91—-111. 1950; Martin and Hutchins, A N. Mex. fl. 1980; McDougall, Seed pls. N. Ariz. 1973; Munz, A Calif. fl. 1959; Parfitt et al., Madrono 26:141—144. 1979; Strother, Brittonia 26:177—202. 1974; Wagner and Sabo, Rep. N. Mex. threat. endang. pl. spp. U.S.D.I. Fish Wildl. Serv., unpubl. reports. 1977. ERIGERON COMPACTUS Blake var. CONSIMILIS (Cronq.) (ASTERACEAE).—USA, NM, McKinley Co., Zuni Mts., Six Mile Canyon (T14N R15W S18), clay hills, 220 m, 18 Jun 1977, Wagner and Sabo 3180 (UNM). Previous knowledge. Uintah Basin, UT to ne. AZ, Navajo and Apache cos. (Cron- quist). Significance. First record for NM; a se. extension of 120 km for the species. 42 MADRONO [Vol. 28 TETRADYMIA SPINOSA Hook. & Arn. (ASTERACEAE).—USA, NM: Sandoval Co., 2 km ne. of La Ventana, 2015 m, 10 Oct 1975, Wagner 1989 (UNM); San Juan Co.: 2.1 km n. of Aztec on hwy 550, Keil 10868 (ASU); Harper NM7 (UT); 10 km s. of Bisti Trading Post, w. of hwy 371 (T23N R13W S25), edge of badlands on sandstone/siltstone interface, 1820 m: 6 Jun 1977, Marley 424 (MO); 24 May 1978, Marley 1234 (UNM); 21 km se. of Bisti, 1.4 km ne. of Black L. (T23N R12W S26/35), heavily eroded area of gray clay over soft coal, arroyo bottom, slopes with Xanthocephalum sarothrae, Atriplex confertifolia, Sarcobatus vermiculatus, Lycium pallidum, 8 Jun 1976, Spellen- berg et al. 4116 (NMC). Previous knowledge. OR to WY, s. to NM and w. to CA (Strother). Significance. First record for Sandoval Co.; 110 km se. range extension (Wagner 1989). Widely-spaced localities may suggest a marginal distribution, but may also reflect lack of floristic study in nw. NM. CHORISPORA TENELLA (Pall.) DC. (BRASSICACEAE).—USA, NM, Bernalillo Co., Al- buquerque, Univ. NM campus, e. side of Biology Bldg., disturbed ground, 25 Apr 1975, Wagner 22 (UNM). Other observations have been made from 1974-1976 along the Rio Grande flood plain in Sandoval Co., near the town of Bernalillo, and in Ber- nalillo Co. along hwy I-40 near the Rio Grande overpass. Previous knowledge. Introduced from Asia, naturalized in N.A. in scattered areas from MA, IA, and NE w. to WA and CA (McDougall). Rapidly spreading in CO (Harrington). Significance. First record for NM. Now well established in c. NM along the Rio Grande and along hwy I-40 near Albuquerque. DIPLOTAXIS MURALIS (L.) DC. (BRASSICACEAE).—USA, NM: Lincoln Co., foothills of Capitan Mts. along F.S. Rd. 56 (T8S RISE 829), grazed pinyon-juniper woodland, 2000 m, 1 Aug 1976, Wagner and Sabo 2120 (GH); Grant Co., hwy 90 at Santa Rita, road shoulders, 1880 m, 10 Aug 1977, Wagner and Sabo 3434 (UNM). Determined by R. Rollins, GH, 1980. Previous knowledge. Naturalized from Europe. Widely scattered as a weed in e. Canada, most of the U.S., including TX and AZ. Known from Tucson as early as 1913 (Kearney and Peebles, p. 337) but not collected in NM prior to Wagner and Sabo 2120 (R. Rollins, pers. comm., 1980). Significance. First records for NM. Wide separation, small population size, and lack of previous collections suggest recent establishment of D. muralis in NM. MALCOLMIA AFRICANA (L.) R.Br. (BRASSICACEAE).—USA, NM, San Juan Co.: se. of Fruitland, T29N R16W S13 nw. % se. 4%, 14 May 1979, Kramp 96 (ASU); 1.2 km n. of jet. of hwy 550 and county Rd. 52 and 524, T30N R15W, rolling hills and benches, Mancos Shale, with Atriplex corrugata, Mentzelia albicaulis , and Oenothera caespitosa, 1600 m, 17 May 1979, Marley 1810 (UNM); 18.5 km n. of Shiprock on hwy 666, w. of Blue Hill, Mancos Shale with Atriplex corrugata, A. cuneata, and Sclerocactus mesae-verdae, 1600 m, 13 May 1977, Wagner and Sabo: 2890 (UNM); 2893 (NMC). Previous knowledge. Naturalized from Mediterranean region (Harrington). In N.A., w.-c. CO to NV and Mojave Co., AZ (McDougall). Significance. First records for NM, the se.-most collections in N.A. SCLEROCACTUS MESAE-VERDAE (Boissevain ex Hill & Salisbury) Benson (CACTA- CEAE).—USA, NM, San Juan Co., e. of Chuska Mts., along hwy 666, 1.6 km s. of Sheep Springs, Menefee Fm., barren, overgrazed sandstone and siltstone hills, 1800 m, 10 Jul 1979, Marley and Heil 1990 (MO). Previous knowledge. Endemic to Mancos Shale, below 1650 m in nw. San Juan Co., NM, and se. Montezuma Co., CO; s.-most collection 8 km. s. of Shiprock along the South Hogback (Wagner and Sabo, op.cit.; Ken Heil, pers. comm., 1979). 1981] NOTEWORTHY COLLECTIONS 43 Significance. Extends the geographical range (65 km), elevational range (150 m), and documents a new substrate type. Proposed for Federal Endangered status (Fed. Reg. op. cit.), but probably should be considered only for Threatened status. ASTRAGALUS MONUMENTALIS Barneby (FABACEAE).—USA, NM, McKinley Co.: sandy slopes, vic. Peach Spring Canyon, T17N R16W S10, Point Lookout Fm., 2050 m, 25 May 1976, Manthey 853 (NY); sandy soils, w. fork Coyote Canyon, T17N R16W S2, pinyon-juniper woodland, 17 May 1977, Manthey 1692 (UNM). Determined by R. Barneby, NY, 1977. Previous knowledge. Canyons of the Colorado and lower San Juan rivers in San Juan Co., UT, and n. Navajo Co., AZ. A member of Sect. Desperati Barneby Subsect. Naturitenses Barneby along with two other highly disjunct, specialized, and recently evolved astragali: A. naturitensis Payson, known only from the Dolores R. and McE]lmo Cr. in Montrose and Montezuma cos., CO; and A. deterior (Barneby) Barneby, known only from Mesa Verde and Montezuma cos., CO. These three rare, local species are a closely related assemblage that has evolved in the Colorado Plateau region and is entirely endemic to it (Barneby). Significance. First record for NM; 140 km se. of closest sites in AZ and 220 km se. of those in UT. These collections emphasize the floristic affinities of nw. NM with the Colorado Plateau, especially with the Canyon Lands Section (Cronquist et al.), to which Astragalus monumentalis is endemic. SALVIA MICROPHYLLA Kunth var. WISLIZENII A. Gray (LAMIACEAE).—USA, NM, Hidalgo Co., Big Hatchet Mts., Thompson Canyon, T3S R15W, limestone soil, 1800 m, 7 Sep 1952, Castetter 5266 (UNM). Specimen had been misidentified. Previous knowledge. Pimaand Cochise cos., AZ, s. to San Luis Potosi and Durango, Mex. (Epling). Common in Chiricahua Mts., AZ. Synonym: S. lemmonii A. Gray. Significance. First record for NM; 80-km e. extension from Chiricahua Mts. BROMUS DIANDRUS Roth (POACEAE).—USA, NM, Bernalillo Co., Albuquerque, 1710 Gold St. SE, residential lot, disturbed ground, 1650 m, 10 Jun 1975, Wagner 1315 (UNM). Previous knowledge. Common from B.C. to ID, s. to AZ and CA (McDougall, Gould). Local in MD and D.C. and occasional n. to MA (Fernald). Has gone under the name B. rigidus Roth., a closely related Mediterranean species (Gould). Significance. First record for NM. RANUNCULUS TESTICULATUS Crantz (RANUNCULACEAE).—USA, NM, San Juan Co., 1 km e. of hwy 666, 1 kms. of NM-CO line, flat plains of Mancos Shale with Atriplex cuneata, Sporobolus airoides, and Cymopterus purpureus, 1625 m, 20 Mar 1979, Marley 1991 (UNM). Previous knowledge. Introduced recently from Mediterranean region (Harrington). In N.A. spreading WA to WY, s. to w. CO, UT, n. AZ, and OR (Parfitt et al.). Significance. First record for NM; 50-km s. extension from w. CO and 100-km ne. extension from recently reported populations in AZ (Parfitt et al.). CERCOCARPUS INTRICATUS S. Wats. (ROSACEAE).—USA, NM, San Juan Co., North Hogback, above Mine Cr., Mesa Verde Sandstone, shallow sandy soils, with Artemisia sp., Xanthocephalum sarothrae, Rhus trilobata, and Juniperus monosperma, 1750 m, 13 May 1977, Wagner and Sabo: 2928 (MQ); 2929 (UNM). Previous knowledge. s. CA, s. andc. NV, UT, and CO, and n. AZ (Martin). Significance. First record for NM, 110 kms. of the s.-most CO site and 90 km ne. of the w.-most AZ site. GREGORY A. MARLEY, 4934 Constance, Apt. C, New Orleans, LA 70115 and WARREN L. WAGNER, Missouri Botanical Garden, P. O. Box 299, St. Louis 63166. (Received 18 Jun 1980; accepted 7 Jul 1980; final version received 25 Aug 1980.) 44 MADRONO [Vol. 28 NOTES AND NEWS NECTAR-SUGARS AND POLLINATOR TYPES IN CALIFORNIA Trichostema (LABIA- TAE).—The relative proportions of glucose, fructose, and sucrose in nectar vary, with either the monosaccharides or disaccharide predominating (Percival, New Phytol. 60:235—281. 1961). The three sugars contain about an equal number of calories per gram (Stiles, Ecology 56:285-301. 1975). Energy costs in the formation of saccharide bonds may select against the use of oligosaccharides but this may be counterbalanced by pollinator preferences. For example, hummingbirds prefer sucrose-rich nectar (Stiles, Condor 78:10—26. 1976) and in a survey of hummingbird-pollinated species, Baker and Baker (Phytochem. Bull. 12:43-—45. 1979) found sucrose-rich nectar to prevail. In this note, data on nectar sugars and pollinator types are presented for five Cali- fornia species of Tvichostema. Populations were sampled in the following areas: 7. lanatum (Santa Monica Mountains); T. lanceolatum and T. ovatum (Central Valley); T. laxum (North Coast Range) and T. parishii (San Gabriel Mountains). More specific locations and a description of pollination mechanisms and nectar production are found elsewhere (Spira, M.A. thesis, Calif. State Univ., Chico. 1978; Spira, Amer. J. Bot. 67:278—284. 1980). Nectar samples were collected and analyzed using techniques de- scribed in Baker and Baker (op. cit.). The hummingbird-pollinated Trichostema lanatum has sucrose-dominant nectar (1.08 sucrose:1 glucose + fructose) while bee-pollinated T. lanceolatum (0.43:1), T. ova- tum (0.43:1) and T. laxum (0.25:1) have glucose-fructose dominant nectars. Both hum- mingbird pollination (Moldenke, Jn: Thrower and Bradbury, eds., Chile-Calif. Medit. scrub atlas. 1977) and insect pollination (Spira, 1980, op. cit.) occur in T. parishiz, which has a sucrose-dominant nectar (1.24:1). These data provide additional evidence that a tendency toward sucrose-dominant nectars is associated with hummingbird pol- lination, in spite of the increased cost in producing it. I thank Irene Baker for analyzing the nectar samples and Sigma Xi for partially funding this research. TIMOTHY P. SprrRA, Department of Botany, University of Cal- ifornia, Berkeley 94720. (Received 9 Jan 1980; returned 18 Jan 1980; revision received and accepted 7 Aug 1980.) WALNUT POLLEN IN LATE-HOLOCENE SEDIMENTS OF THE SACRAMENTO-SAN JOA- QUIN DELTA, CALIFORNIA.—Thompson (Madrono 17:1—-10. 1963) has reviewed the origin and distribution of Juglans hindsii Jepson in central California, has noted the paucity of fossil material of Pleistocene-Holocene age, and has suggested (p. 8) that “verification of Pleistocene remains, perhaps through fossil walnut pollens from the San Joaquin Delta peats, would greatly reinforce our understanding of pre-Holocene distri- bution.” While it is now known, based on radiocarbon-dated peats and peaty mucks, that the modern Delta is post-Pleistocene (Shlemon and Begg, Jn: Suggate and Cres- swell, eds., Quaternary studies. 1975), the presence of walnut pollen within Delta sed- iments still is significant, because it adds to the known areal distribution of walnut in central California a temporal dimension extending back 5000-6000 years. As part of a paleo-environmental study of the Delta supported by the California Department of Water Resources (West, report on file, Calif. Dept. Parks Rec. 1977), cores were collected from Clifton Court and Roberts Island in the southeast corner of the Delta. Both cores provided discontinuous pollen records from more than 4300 ra- diocarbon years ago. In the Clifton Court core, walnut pollen is present in small amounts (<1 percent) in four samples, one from immediately below a peaty muck dated to 2950 + 150 B.P. (GX 4221) and the others from 2, 8, and 10 cm below the lowest dated peaty muck (4340 + 150 B.P.; GX 4223). In the Roberts Island core, Juglans pollen is present only in the uppermost of the prehistoric peat layers and in each of the overlying historic age samples. The Roberts Island core samples have not been dated by radio- carbon methods but the sediments, stratigraphy, and pollen spectra are similar to those of the Clifton Court core and are assumed to be of an equivalent age. Some of the 1981] NOTES AND NEWS 45 larger-sized (45—49 um) walnut pollen grains in the historic-age sediments of the Roberts Island core may be derived from Juglans regia L. growing in orchards adjacent to the Delta (unpubl. data; Stone and Broome, Jn: Nilsson, ed., World pollen spore fl. 4. 1975). In all instances Juglans pollen is well preserved and does not appear to be redeposited from earlier sediments in the Sacramento-San Joaquin drainage systems. Accordingly, I believe that the Juglans is in primary deposits derived from trees growing within the drainage system over the last 5000 years. Thompson’s suggestion regarding the presence of walnut pollen in the Delta deposits is correct. Although the available pollen record does not preclude the dispersal of walnut in central California by prehistoric human activity, it does indicate that the trees were present in the Delta thousands of years prior to the earliest known Euro-American entry into the area.—G. JAMES WEST, Department of Anthropology, University of California, Davis 95616. (Received 2 Nov 1979; revision received and accepted 26 Sep 1980.) AGES OF INVASIVE TREES IN DANA MEADOWS, YOSEMITE NATIONAL PARK, CALI- FORNIA.—Throughout the high elevations of the Sierra Nevada, meadows are com- monly invaded by lodgepole pine (Pinus contorta). In 1978, I investigated the age structure of trees in Dana Meadows, about 2 km south of the Tioga Pass entrance station to Yosemite National Park. These meadows are bordered by monotypic stands of lodgepole pine. On both the north and south sides of the meadow, five belt transects 2 m wide were located 30 m apart. Each of the ten transects was extended perpendic- ularly from the forest-meadow ecotone into the meadow to the most distant tree en- countered; transect lengths varied from 10.4 m to 100 m. The largest number of trees in a transect was 32, and the smallest was 5. An eleventh sample, a quadrat 10 m by 2 m and containing 23 trees, was established within the meadow in an area of dense young trees. All trees rooted within the sample areas were cored at their bases with an increment borer; ages were estimated to be the number of rings on the cores plus 2. Trees too small to core were cut at their bases and the rings counted on the stumps. In addition to the trees within the samples, the four largest trees well within the meadow, which occurred in a stand of 12 individuals, were cored. In total, the ages of 149 trees were determined; these were aggregated into five-year age-classes for analysis. Invasion by large numbers of trees apparently began about 1925, although all five- year periods between 1910 and 1975 were represented by trees; a few trees germinated and survived within the meadow in still earlier years (Table 1). The oldest tree within the transects germinated about 1866 (other germination dates for trees in this stand: 1887, 1902, 1910). In general, the older trees appeared healthy, but 40 percent of the trees established after 1950 had dead leaders. In addition, 16 dead trees were encoun- tered in the 11 sample areas; judging by size-age correlations, none of these was older than 30 years when it died. The distribution of trees of certain conditions and ages along the transects produced only weak patterns. The dead trees were not restricted to particular locations along the transects, but most of the trees with dead tops were not immediately adjacent to the forest edge. The correlation between position on the transects and age was not signif- icant (r? = 0.08), although trees encountered in the outermost segment of the transects were often younger than those closer to the forest, and the oldest trees within each transect usually tended to be closer to the forest edge than to the end of the transect in the meadow. These patterns suggest that invasion into open meadow by a few individ- uals is followed by subsequent germination and survival of other trees. The invasion date of 1925 is more recent than the dates of initial tree establishment reported by either Boche (Factors affecting meadow-forest borders in Yosemite National Park, California. M.S. thesis, U.C.L.A. 1974) for meadows in the lodgepole forests of Yosemite (1898-1909), or Vankat and Major (J. Biogeogr. 5:377—402. 1978) for mead- ows in the lodgepole forests of the southern Sierra (1910). In his review of three other 46 MADRONO [Vol. 28 TABLE 1. SPRING TEMPERATURE DEVIATIONS, ANNUAL PRECIPITATION, AND NUMBERS OF TREES GERMINATING IN FIVE-YEAR PERIODS FROM 1865. Spring temperatures are deviations of five-year means from the mean monthly temperatures for March, April, and May. Precipitation values are percents of the mean for the period covered. Climatic data are from NOAA for Yosemite Valley. Spring temp. No. of trees Period (°C) Annual precip. germinating 1865— 1 1870- O 1875— 0 1880— O 1885-— 1 1890-— 0 1895— 1 1900— Z 1905— 122 0) 1910— 93 3 1915-— =, 9 84 Z 1920— 22 87 5 1925-— + O./ 89 12 1930— +0.8 76 i3 1935— =().1 119 14 1940-— +0.3 116 12 1945— +0. 1 95 5 1950— -O.2 104 12 1955— +0. 1 105 23 1960— == ).6 99 14 1965— +0 ,2 120 18 1970- +0.4 105 13 1975— =0.5 83 0 studies of such tree invasions in the southern Sierra Nevada, Wood (Holocene stratig- raphy and chronology of mountain meadows, Sierra Nevada, California. Ph.D. disser- tation, Calif. Inst. Tech. 1975) found that the initial establishment dates of meadow trees were reported as 1903, 1906, and 1924. Part of the disparity in dates probably reflects site-specific differences in environmental or historical factors, but it may also result from different criteria for identifying the “beginning” of tree invasions. If 1910 is recognized as the date of initial invasion in Dana Meadows, for example, the data presented here become more consistent with other studies. The identification of the date when the tree invasion “began” becomes especially significant when trying to correlate it with events that may have caused the unstable ecotone. Regardless of which dates are used, climatic fluctuation seems an unlikely trigger for the recent establishment of meadow trees in Dana Meadows or elsewhere in the Sierra Nevada. Warm dry weather has been suggested as a cause of tree invasion in Sierran meadows because such weather desiccates competing herbaceous vegetation and reduces suffocating soil moisture (Boche, op. cit; Wood, op. cit.). Franklin and Dyrness (U.S. For. Serv. Gen. Tech. Rept. PNW-8. 1973) found convincing the asso- ciation between warm dry weather and a period of tree establishment in subalpine meadows in the Pacific Northwest. These warm dry periods early in this century, however, were apparently not unprecedented in the recent past in the Sierra Nevada; particularly noteworthy is the warm dry episode in the mid 1800’s (Fritts, Monthly Weather Rev. 93:421-443. 1965; Bradley, Monthly Weather Rev. 104:501—512. 1976), a relatively recent period with weather presumably favorable for tree invasion when 1981] NOTES AND NEWS 47 trees did not become established in Dana Meadows or other subalpine meadows in the Sierra Nevada. The relatively cold dry conditions since 1975 may have contributed to both the lack of recent tree establishment and the high proportion of die-back among small trees. As a period, however, the years of massive tree invasion were not char- acterized by consistent and distinctive climatic conditions (Table 1). Other possible causes are environmental changes involving people. Suppression of frequent fires is sometimes considered responsible for these invasions, although the only empirical study of this agency in a Sierran lodgepole forest focuses on a wet meadow and not a dry-mesic meadow such as Dana Meadows (DeBenedetti and Parsons, J. Forest (Washington) 77:477—479. 1979). The model of Vankat and Major, which sug- gests a lag of a decade between the cessation of sheep grazing and the establishment of meadow trees, seems not to fit the data in this study; sheep were not eliminated from Dana Meadows until 1905, decades following the germination of the oldest tree (1866) and far predating the beginning of massive invasion (1925). Yet, their model fits the data better if the date of 1910 is used as the beginning of the invasion; such an inter- pretation would then also conform more closely with that of Dunwiddie (Arctic and Alpine Res. 9:393-399. 1977) who found that trees invaded a subalpine meadow in Wyoming soon after the cessation of grazing by sheep. It is difficult to isolate possible single causes of these tree invasions, however, by looking simply at the chronology of events. Particularly troublesome is the fact that sheepherders probably burned meadowlands in the Sierra regularly, and thus the elim- ination of sheep grazing involved a great reduction not only in grazing intensity but also in fire frequency. The interactions among environmental variables also detract from the attempt to find a single cause of invasion. Heavy grazing, for example, may create conditions conducive to tree invasion, but trees may not become established unless the climatic conditions are also suitable. Compounding these difficulties is the likelihood that the occasional establishment of single trees (pre-1910 or pre-1925 in Dana Meadows) is a different phenomenon (reflecting distinctive causes) trom the massive establishment of trees so common in the meadows of the Sierra Nevada (after 1925 in Dana Meadows). It is popular to suggest that forest-meadow ecotones are in “dynamic equilibrium”, in that they may fluctuate with short-term changes in environmental conditions but remain stable over longer periods of time. The high frequency of young trees that are either dead or dying supports such a view for Dana Meadows. The apparent health of older trees, however, even those well within the meadow, implies that the invasion is better interpreted as a “directional” change in the vegetation. —THOMAS R. VALE, Department of Geography, University of Wisconsin, Madison 53706. (Received 7 Mar 1980; revision received 12 Sep 1980; accepted 22 Sep 1980.) 48 MADRONO [Vol. 28 Books RECEIVED AND LITERATURE OF INTEREST Flora of the Central Wasatch Front, Utah. By Lois A. ARNOW, BEVERLY J. ALBEE, and ANN M. WYCKOFF. xiv + 663 p. Univ. Utah Printing Service, Salt Lake City. ed 2, 1980. No price listed. Will be reviewed in a subsequent issue. Inventory of Rare and Endangered Vascular Plants of California. By JAMES PAYNE SMITH, JR., R. JANE COLE, and JOHN O. SAWYER, JR. in collaboration with W. Robert Powell. viii + 115 p. California Native Plant Society Special Publication 1, Berkeley. ed. 2, 1980. $7.50. Available from CNPS, 2380 Ellsworth, Suite D, Berkeley, CA 94704. Will be reviewed in a subsequent issue. Rare, Threatened and Endangered Vascular Plants in Oregon—an Interim Report. By JEAN L. SIDDALL, KENTON L. CHAMBERS, and DAvID H. WAGNER. Oregon Nat- ural Area Preserves Advisory Committee, Salem. 1979. iv + 109 p. Available (appar- ently free) from Division of State Lands, 1445 State St., Salem, OR 97310. Threatened and Endangered Plants of Alaska. By DAvID F. MuRRAY. vii + 59 p., illus. Published cooperatively by USDA (Forest Service) and USDI (BLM). 1980. No price listed. Desert Plants. Edited by FRANK S. CROSSWHITE. Published by the Boyce Thompson Southwestern Arboretum, Box AB, Superior, AZ 85273. This new journal is oriented to the general public rather than to professional botanists and, judging from promotional literature, will cover botanical history, ethnobotany, agriculture, horticulture, flora, vegetation, and conservation of the desert regions of the southwest. Scientific Research in Sequoia and Kings Canyon National Parks: an Annotated Bibliography. By DAVID J. PARSONS and VIRGINIA A. KING. ii + 70 p. Sequoia Natural History Association, Three Rivers, CA. 1980. Available from NTIS (document PB 80- 187-313), 5285 Port Royal Road, Springfield, VA 22161. $8.00 (paper) or $3.50 (mi- crofiche). Nearly 350 references to work both published and unpublished, organized by general topic (Fire Ecology, Geology and Soils, Vegetation, Wilderness Use and Impact, etc). ANNOUNCEMENT BRUSHLAND MANAGEMENT SYMPOSIUM An international symposium on brushland management will be held in San Diego, 22-26 June 1981. The focus will be on regions having Mediterranean-type climates, including the Mediterranean Basin, southwestern United States, Africa, Chile, and Australia. Symposium topics will include the effects of brushland management on vege- tation, wildlife, soils, and hydrology; the use of prescribed burning; and new ways to make better use of brushlands. More information is available from: Chairman, Dynam- ics and Management of Mediterranean-type Ecosystems, Pacific SW For. and Range Expt. Station, 4955 Canyon Crest Drive, Riverside, CA 92507. Membership in the California Botanical Society is open to individuals ($18 per year; students $10 per year for a maximum of seven years). Members of the Society receive MADRONO free. Institutional subscriptions to MADRONO are available ($25 per year). Membership is based on a calendar year only. Applications for membership (including dues), orders for subscriptions, requests and rates for back issues, changes of address, and undelivered copies of MADRONO should be sent to the Corresponding Secretary. Renewal payments should be sent to the Treasurer. INFORMATION FOR CONTRIBUTORS Manuscripts submitted for publication in MADRONO should be sent to the editor. All authors must be members, and membership is prerequisite for review. Manuscripts and review copies of illustrations must be submitted in triplicate for all articles and short items intended for NOTES AND NEWS. Follow the format used in recent issues for the type of item submitted and allow ample margins all around. All manuscripts MUST BE DOUBLE SPACED THROUGHOUT. For articles this in- cludes title (all caps, centered), author names (all caps, centered), addresses (caps and lower case, centered), abstract, 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, and number each page. Do not use a separate cover page, “erasable” paper, or footnotes. Table captions should include all information relevant to tables. All measurements should be in metric units. Line copy illustrations should be clean and legible, proportioned (including captions) to the MADRONO page, and designed for reduction to % original size. 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 latitude and longitude references. Halftone copy should be designed for reproduction at actual size. In no case should original illustrations be sent prior to the acceptance of a manuscript. When needed they should be mounted on stiff cardboard and sent flat. No illustrations larger than 22 X 28 cm will be accepted. Presentation of nomenclatural matter (accepted names, synonyms, typification) should follow the format used for Rhus integrifolia in MADRONO 22:288. 1974. Institutional abbreviations in specimen citations should follow Holmgren and Keuken, Index Her- bariorum, 6th edition. Abbreviations of serial titles should be those in Botanico-Per- iodicum-Huntianum (Lawrence et al., 1968, Hunt Botanical Library, Pittsburgh). If the correct abbreviation cannot be determined, the full serial title should be used. Titles of books should be given in full, together with the place and date of publication, publisher, and edition, if other than the first. All members of the California Botanical Society are allotted eight pages in MADRONO over a two year period. Joint authors are each assessed the full page number. Beyond that number of pages a required editorial fee of $40.00 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 sched- ule, with equity among all members for access to its pages. Printer’s fees for illustrations and typographically difficult material @ $35.00 per page (if their sum exceeds 30 percent of the paper) and for author’s changes after typesetting @ $3.00 per line will be charged to authors. Contents, continued NOTES AND NEWS NECTAR-SUGARS AND POLLINATOR TYPES IN CALIFORNIA Trichostema (LABIATAE), Timothy P. Spira 44 WALNUT POLLEN IN LATE-HOLOCENE SEDIMENTS OF THE SACRAMENTO-SAN JOAQUIN DELTA, CALIFORNIA, G. James West 44 AGES OF INVASIVE TREES IN DANA MEADOWS, YOSEMITE NATIONAL PARK, CALIFORNIA, Thomas R. Vale 45 BooKS RECEIVED AND LITERATURE OF INTEREST 48 ANNOUNCEMENTS 25, 48 CALIFORNIA BOTANICAL SOCIETY STATEMENT OF OWNERSHIP, MANAGEMENT, AND CIRCULATION (Act of 23 October 1962; Section 4369, Title 39, United States Code) MADRONO, A West American Journal of Botany, is published quarterly at Berkeley, California. The Publisher is the California Botanicai Society, Inc., Life Sciences Building, University of California, Berkeley, California 94720. The Editor is James C. Hickman, Department of Botany, University of California, Berkeley, California 94720. The owner is the California Botanical Society, Inc., Life Sciences Building, Uni- versity of California, Berkeley, California 94720. There are no bondholders, mortga- gees, or other security holders. The average number of copies distributed of each issue during the preceding 12 months is 1000; the number of copies of the single issue closest to the filing date is 1015. I certify that the statements made by me above are correct and complete. 10 September 1980 JAMES C. HICKMAN, Editor iy M132 G0T, MADRONO VOLUME 28, NUMBER 2 APRIL 1981 WEST AMERICAN JOURNAL OF BOTANY A Contents STRAND AND DUNE VEGETATION AT SALINAS RIVER STATE BEACH, CALIFORNIA, Victor Bluestone POSTFIRE RECOVERY OF CREOSOTE BUSH SCRUB VEGETATION IN THE WESTERN COLORADO DESERT, John F. O’Leary and Richard A. Minnich SEEDLING CHARACTERISTICS AND ELEVATIONAL DISTRIBUTIONS OF PINES (PINACEAE) IN THE SIERRA NEVADA OF CENTRAL CALIFORNIA: A HYPOTHESIS, Richard I. Yeaton PORTULACA JOHNSTONII, A NEW SPECIES OF PORTULACACEAE FROM THE CHIHUAHUAN DESERT, James Henrickson MALEPHORA CROCEA (AIZOACEAE) NATURALIZED IN CALIFORNIA, Wayne R. Ferren, Jr., John Bleck, and Nancy Vivrette NOTEWORTHY COLLECTIONS COCHLEARIA OFFICINALIS, Gary S. Lester, Michael C. Vasey, and William E. Rodstrom PEDICULARIS CRENULATA f. CANDIDA, Ann M. Howald and Bruce K. Orr DEDECKERA EUREKENSIS, Patti J. Novak and Kathryn L. Strohm IPOMOEA EGREGIA, STELLARIA NITENS, Rob J. Soreng and Richard Spellenberg ERIGERON HUMILIS, HYMENOPAPPUS FILIFOLIUS var. IDAHOENSIS, CAREX RUPESTRIS, ASTRAGALUS AMNIS- AMISSI, GENTIANA PROPINQUA, PAPAVER KLUANENSIS, Douglass M. Henderson, Steven Brunsfeld, and Pamela Brunsfeld NOTES AND NEWS NOTES ON CONES AND VERTEBRATE-MEDIATED SEED DISPERSAL OF Pinus albicaulis (PINACEAE), Diana F. Tomback AGGREGATION OF Prunus ilicifolia (ROSACEAE) DURING DISPERSAL AND ITS EFFECT ON SURVIVAL AND GROWTH, Stephen H. Bullock ADVENTITIOUS ROOTING IN COASTAL SAGE SCRUB DOMINANTS, R. John Little (Continued on back cover) 49 61 67 78 80 86 86 86 87 88 91 94 96 PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL SOCIETY MADRONO (ISSN 0024-9637) is published quarterly (Jan, Apr, Jul, Oct, except following unusual delay) by the California Botanical Society, Inc., and is issued from the office of the Society, Herbarium, Life Sciences Building, University of California, Berkeley, CA 94720. Subscription rate: $25 per calendar year. Established 1916. Second-class postage paid at Berkeley, California, and additional mailing office. Return requested. Editor—CHRISTOPHER DAVIDSON Natural History Museum of Los Angeles County 900 Exposition Blvd., Los Angeles, CA 90007 (213) 744-3378 Associate Editor—JAMES C. HICKMAN Department of Botany University of California, Berkeley 94720 (415) 642-2465 Board of Editors Class of: 1981—DANIEL J. CRAWFORD, Ohio State University, Columbus JAMES HENRICKSON, California State University, Los Angeles 1982—DEAN W. TAYLOR, University of California, Davis RICHARD VOGL, California State University, Los Angeles 1983—-ROBERT W. CRUDEN, University of Iowa, Iowa City DUNCAN M. PORTER, Virginia Polytechnic Institute and State University, Blacksburg 1984—Mary E. BARKWoRTH, Utah State University, Logan HARRY D. THIERS, San Francisco State University, San Francisco 1985—-STERLING C. KEELEY, Whittier College, Whittier, CA ARTHUR C. GIBSON, University of California, Los Angeles CALIFORNIA BOTANICAL SOCIETY, INC. OFFICERS FOR 1981 President: ROBERT ORNDUFF, Department of Botany, University of California, Berkeley 94720 First Vice President: LLAURAMAY T. DEMPSTER, Jepson Herbarium, Department of Botany, University of California, Berkeley 94720 Second Vice President: CLIFTON F. SMITH, Santa Barbara Museum of Natural History, Santa Barbara, CA 93105 Recording Secretary: ROBERT W. PATTERSON, Department of Biology, San Francisco State University, San Francisco, CA 94132 Corresponding Secretary: FRANK ALMEDA, Department of Botany, California Academy of Sciences, Golden Gate Park, San Francisco 94118 Treasurer: CHERIE L. R. WETZEL, Department of Biology, City College of San Francisco, 50 Phelan Ave., San Francisco, CA 94112 The Council of the California Botanical Society consists of the officers listed above plus the immediate Past President, WAYNE SAVAGE, Department of Biology, San Jose State University, San Jose, CA 95192; the Editor of MADRONO; three elected Council Members: PAuL C. SILVA, University Herbarium, Department of Botany, University of California, Berkeley 94720; JOHN M. TUCKER, Department of Botany, University of California, Davis 95616; CHARLES F. QUIBELL, Department of Biological Sciences, Sonoma State College, Rohnert Park, CA 94928; and a Graduate Student Represen- tative, KENT HOLSINGER, Department of Biological Sciences, Stanford University, Stanford, CA 94305. STRAND AND DUNE VEGETATION AT SALINAS RIVER STATE BEACH, CALIFORNIA VICTOR BLUESTONE 7338 Mesa Drive, Aptos, CA 95003 ABSTRACT At Salinas River State Beach, Monterey Bay, California, beach and dune vegetation is relatively undisturbed. Nine zones of vegetation are distinguished along a line inland from the shore: beach, foredune, foredune hollow, mid-dune, mid-dune hollow, and reardune, the last divided into foreslope, crown, leeslope, and inland margin zones. These zones were arrived at through analysis of physiography, plant cover, density, species richness, and changing size of selected species. In general, low plants (such as Carpobrotus aequilaterus, Convolvulus soldanella, and Abronia latifolia with trailing stems, and Artemisia pycnocephala with thick pubescence) are characteristic of the foredunes to mid-dunes. Woody shrubs (such as Haplopappus ericoides, Lupinus cham- issonis, and Eriophyllum staechadifolium) along with smaller herbs and subshrubs (such as Dudleya farinosa and Eriogonum latifolium) dominate the mid-dunes through reardunes. Although several studies of the zonation of beach and dune vege- tation along the California shoreline have been undertaken (for sum- maries see Breckon and Barbour, 1974; McBride and Stone, 1976; Barbour and Johnson, 1977), such analysis within Monterey Bay has begun only recently (Bluestone, 1970; Gordon, 1979). The closest area investigated has been the Monterey Peninsula, which marks the south- ern extent of the bay (Cooper, 1967; McBride and Stone, 1976). Quantitative documentation of the beach and dune vegetation with- in Monterey Bay is particularly important at this time because in- creasing human activities are significantly altering the character of many of the beach and dune areas (Gordon, 1979). Because some of the best examples of natural beach and dune vegetation within Mon- terey Bay are found at Salinas River State Beach, the zones described in this study might serve as a paradigm for other sites in the area. METHODS Study area. Salinas River State Beach is located in the southern half of Monterey Bay, 5 km south of Moss Landing and Elkhorn Slough. Wave action and longshore currents transport sand derived from eroded marine terraces and various rivers in the northern part of the bay, and deposit it in the southern half of the bay where the dunes tracts formed are among the largest along the California Coast (Shepard and Wanless, 1971). The crests of the dune ridges tend to lie at right angles to the direction of the dominant winds, which blow onshore from a northwesterly direction. MADRONO, Vol. 28, No. 2, pp. 49-60, 3 June 1981 49 50 MADRONO [Vol. 28 A maritime climate of little seasonal temperature fluctuation pre- vails. Although rainfall is concentrated in the winter months, summer fogs moderate the hotter dry season from late spring to early fall by reducing the temperature and providing moisture along the shoreline (Moss Landing Marine Laboratory, 1971). Annual precipitation is 460 mm. On the average there is only one day of frost per year. The mean air temperature is 10°C during January and 16°C during July. Because surface water supply is limited and salt-laden winds are constant throughout the year, some have considered the beach and dune habitat essentially xeric (Purer, 1936; Martin and Clements, 1939). It is likely that soluble mineral and organic content in the soil is limited, as it is for the California coastal sand areas in general (Cooper, 1967). Ac- cumulation of organic litter and compaction of sand are greater where plant density and cover values are highest. Beginning at least 7500 years ago, the Costanoan Indians, or cul- turally similar predecessors, may have altered the beach and dune vegetation to some extent with their hunting and gathering, and lim- ited use of fire (Gordon, 1979). A Costanoan site is located near the inland margin of the transect described here. Browsing by livestock of later settlers beginning with the Spanish probably altered this vege- tation further. However, the frequency and amount of alteration was probably little enough to allow native plant growth to reestablish readily. Recent human activity on the beach and dunes has accelerated alarmingly within the last 20 years leaving visible scars, such as the three sand trails that parallel the coastline and run through the study site as shown in Fig. 1A. Furthermore, the deflation area shown ad- jacent to the transect may have been caused by human activities, and a recent fire may have occurred on the leeside of the reardune (Gordon, 1979). Other cultural imprints considered are the introduced species. These include Carpobrotus edulis from southern Africa, Ammophila arenaria from Europe, and Tetragonia tetragonioides from Australa- sla. Quantitative sampling. A total of 372 contiguous 1-m? quadrats was taken along a line inland from the shoreline at Salinas River State Beach. Along the same line during June and July of 1977, 1978, and 1980, I recorded density (number of individuals per m*) and percent cover for each species in each quadrat. Zones were defined by changes in topography or vegetation. Im- portant vegetation changes included density, cover, and species com- position. Also considered was the changing size of Haplopappus eri- coides and Eriophyllum staechadifolium, which are wide-ranging shrubs that illustrate how diameter and height change along the tran- sect. Species considered to be zone dominants were those with more than 10 percent cover; subdominants had 5-10 percent cover. Nomenclature follows Munz (1959), except for the names Carpo- ‘pasoduitiadns st yd8suel}] VY, “WO ‘Aedaquoyy ‘SUlIs0UTIs Uy - Joysag Aq uaye} ydeisojoyd [etiay (Molaq) gq “JD9SUeI} ay} BUOTe UOTIIIS SSOID “(2AOQL) Y ‘YIVI 9781S JOATY SeUTTeS ye Bale ApNjs BUT, ‘TI ‘DIY BLUESTONE: DUNE VEGETATION 0s2l 006 002 00s oo¢ OOl ay, Sso pow U ose oo¢ 0sz 002 osi 001 os a oo Ome @ 8 a = 0S Z zi— —Ov =6¢ = PA we | OA AN I si— —09 Pa ie ee errr oe cero a ota Ue ana ee ee es A ee art alee Sol S ee | | qm | anna-ain | DD | anna | I SoReee Sees eee $9N0d > oS se = 3 N n a y v 3 u MO770H MO170H HOvaa 52 MADRONO [Vol. 28 brotus aequilaterus and C. edulis, which follow Brown (1928). All taxa were compared with specimens in CAS and DS. RESULTS The quantitative data, averaged over all three years, are arrayed by zone in Table 1. Transitions between zones were often gradational, but clusters of indicator species can be identified from the table. Each zone is described briefly below. Zone lI: Beach. Occasionally beach plants such as Cakile maritima are within the area of beach inundated by the sea at high tide. The dunes along the inland portions of the beach are exposed to constant onshore winds. They are small (less than 1 m high), scattered, and seasonally transient. Cover, density, and species richness are the lowest encountered. No plant has more than 4 percent cover. Except for the grasses Elymus mollis and Ammophila arenaria that are found intermingling with other plants, each dune tends to be occupied by a single species. This is probably because the root system of a single establishing plant serves as a nucleus around which the dune is formed. Zone II; Foredunes. ‘The foredunes sharply demarcate the inland extent of the beach with an undulating ridge of sand that rises 2—4 m above the beach at an angle of 15—20° (Fig. 2). Strong onshore winds, abundant salt spray, and rapid displacement of sand further charac- terize this zone. Cover, density, and species richness are still relatively low, but cover and density contrast strikingly with the beach zone. Many plants found in the more inland zones make their first appearance here. Car- pobrotus aequilaterus is a strong dominant and Elymus mollis is sub- dominant. Many other plants grow intermingled, but no consistent patterns of association were observed. Zone IIla: Foredune hollows. This zone forms a shallow depres- sion immediately inland from the foredunes that grades onto the mid- dunes (Figs. 2, 3). Protection from wind by foredunes, accumulation of litter, and a possible freshwater table closer to the surface may help = Fics. 2-4. Beach to mid-dunes. FIG. 2 (top). Foredunes contrast sharply with beach on the left and grade into foredune hollows inland. Carpobrotus aequilaterus and other low plants dominate foredune crowns. FIG. 3 (middle). Greyish Artemisia pycnocephala of foredune hollows and mid-dunes (middle distance) contrast with the darker reardune foreslope dominated by Haplopappus ericoides (background). FIG. 4 (bottom). Vege- tation characteristic of both seaward and inland zones mixes among undulating mid- dune ridges and depressions. Sand trail at left corresponds with thin middle trail in Fig. 1B. 1981] BLUESTONE: DUNE VEGETATION 53 [Vol. 28 MADRONO 54 Ocal OC Le 0 mee Rc Oe 9 8 eat SIUOSSIMDYI SNUIGNT Zi TO GL “80 “91. TO unyofippyranys mn Yyqoiug SOI ‘90 82 ‘vO Sst ‘60 ert ‘ZT 60 ‘70 20 ‘+ DSOULAD, DKajpnq z7O ‘0 v0 ‘TO 60 ‘FO pqAnj01a] DYIUDIGKAD OO se0r uSG C1 “oS “Ci Got S71 11sD]3nNOp DOg 10z ‘SO Lrt‘ro LIz7‘so T61 ‘90 Os ‘70 2+ ‘TO sapionisa snddvgojqn H £0 “£0 61 “ZO 2° $30 pyofijn] Dlaqj14jsDD 8ST ‘60 901 ‘Z0 68 ‘80 D1 0fiqD] DIUOAQ PY POE Ome Pic ol- Le DIDYqGaIOUIK DISIMAIAP ye 11]]D]JNU SDIDBDAISP GO) 20 DIOfLYJUDAILYI DAIYIOUIO SO, + Si At 0 Pa 5 Seg 0) STUOSSIMDYI DIAISUDAT br ‘FO CLL Wt £1 0 DIJIUDPIOS SNjnAjoAUd) r'0 ‘20 8:0 ‘27'0 DUIJIADU DIAIUAP 970 ‘+ Cra» 00> “hE F60. 675 10 $1JD404]1] SNAKYIDT (ia 0 ae $aP101U08DAJ A] DIUOSDAJA I se synpa snjorqogany) co ‘+ Sap1o1ssvgD Siaasosp PE EOs FASO GOR et) SS“ er. 19A09 & YIM JOVSUeI] BUO ATUO UT pUNoO; sjueId = .4,, “T'O> Aysuep yuetd = .4,, ‘O86 pue ‘6261 ‘Z/61 Wor ssutfdures 9a1y} Jo aBCIIAV JY} JUaSaIdaI eyed ‘(,W Jad) 1aA09 UIIIId URI ST pUOdDIS ay} ‘(,WI Jad sjue[d) A}ISUIp UBIUE SI JaquINU }sIY IY], “IUI[I1OYS IY} WoT; pueTUr uaye} syeipenb ,WI-[ snonsuo0d 7/¢ WOIJ vIe VIE “LOASNVAL HOVAG ALVLS AAARY SVNITVS V ONOTY ANOZ Ad VIVG GNVIS ‘I ATAVL BLUESTONE: DUNE VEGETATION 55 1981] 0'OOT CL9 TOF 1°79 (ea vcs v'6L Cae See | JIAO) [210], SC Gx Lie Poe 9°9 O's O'?T SP cl (,tu/sjueyd) AyIsuap [e}O], 9 el 6 IT cl cl 9T 6 9 ssauyo satvads S 79 O¢ SIT Os S¢ 4 02 of (UW) 9u0z Jo YIpeag Le 20 DIUAO{IDI sNgnY 8 ‘90 D1D}1dDI DYIH) £6 ‘90 sapjnird stDYyrIvg 8°64 “7°61 IDADGADG XAADJ 6'r ‘80 2 EO Dgop1sAaaip SNYY Lc vo 8°27 ‘20 pjopiday DzLYAAK IK] * SNIADGOIS SNIOT 870 ‘TO snarvqgnf YDAD JY oe aaa 8 DIWUAO{IDI DSOY Le 80 s1ignjIags DIyIUISMP v9 ‘70 ZT ‘TO DUISSISOUDA DIIBIDY 870 ‘+ ZO ‘+ mnyofiawd wnuosouy DYyOfiursp]Yy IUKSO1YIALOD + PERRO Creere DIDIJIQUN DIUOAQY Z°0. 710 DIDP1gsnI ayIUvzZ1A0Y 7D COLO: = qt: 4 o » — U 500 meters [cae Since a Fic. 1. Location of study area near Snow Creek (Riverside Co.) in southern Cali- fornia. Stippled area denotes burned creosote bush scrub. Dotted line approximates the upper elevational extent of Larrea tridentata as does the dashed line for Opuntia echin- ocarpa. Vegetation upslope from dashed line is dominated by coastal sage scrub species admixed with some desert shrubs, and, at higher elevations, chaparral shrubs. Num- bered lines straddling fire border represent belt transects; darkened squares indicate line intercept sampling sites. Arrow shows direction in which 1973 fire burned. isted between the two sites. Species found were Euphorbia albomar- ginata (30 percent cover in unburned area, 30 percent in burned area); Erodium cicutarium (6, 10); Schismus barbatus (6, 5); Crassula erecta (3, 3); and Bromus rubens (2, 1). The following species had 0-1 percent cover in both areas: Malacothrix glabrata, Lupinus bicolor, Cryptan- tha intermedia, Lotus tomentellus, Lasthenia chrysostoma, and Ca- missonia pallida. The coefficient of community (0.82) and the per- centage similarity (84 percent) between the sites were both fairly high. DISCUSSION The portion of the California fire that extended onto the alluvial fan in the Snow Creek area was severe enough to defoliate nearly all creosote bushes, although most resprouted. Opuntia echinocarpa sur- vived by directly tolerating the fire. Living individuals bore scorched tissue five years later. No regeneration from fallen unburned joints was apparent. Tratz and Vogl (1977) found that O. acanthocarpa, a 64 MADRONO [Vol. 28 TABLE 1. FREQUENCY OF PERENNIAL PLANTS ACROSS BURN BOUNDARY IN SNOW CREEK DRAINAGE. Data (percents) are based on two belt transects. Percent foliar cover for the dominant shrubs (based on four line intercepts in each study site) is given in parentheses. Unburned Burned Species Living Dead Living Dead Larrea tridentata 24 (10) O (0) 8 (2) 3 (0.4) Opuntia echinocarpa 6 (4) 1 (0) 6 (2) 11 (4) Hymenoclea salsola 4 (5) 0 (0) 19 (10) 3 (5) Encelia farinosa 0.6 0 1.1 0 Prosopis glandulosa 5.1 0 0 0 Mirabilis tenuiloba 0 0) 0.8 0) Stillingia linearifolia 0 0 0.5 0 Salvia mellifera 0 0 0.5 0 closely related species, had 25 percent of its individuals resprouting one year after a fire on burned portions of Anza-Borrego Desert State Park, San Diego County. Opuntia acanthocarpa displayed the poorest recovery of sampled plants following that fire. Larrea tridentata and some H. salsola plants have resprouting abil- ity comparable to some chaparral plants. The abundance of H. salsola in the burned site may reflect an ability for rapid, weedy colonization. Other investigators have tentatively categorized H. salsola as a rela- tively short-lived shrub occurring in naturally disturbed areas in creo- sote bush scrub. It is also observed as a pioneer in more severely disturbed areas such as those affected by pipeline construction, power transmission lines, and off-road vehicles (Davidson and Fox, 1974; Vasek, Johnson, and Eslinger, 1975; Vasek, Johnson, and Yonkers, 1975). No fires have occurred on the study area in historic times. In 1911, a fire spread downward from the mountains to the uppermost extent of the alluvial fan, whereas in 1941 the ridge immediately west of the fan burned (U.S. Forest Service, undated). It is of interest that the lower perimeter of the 1911 fire roughly coincides with the upper elevational extent of O. echinocarpa on the alluvial fan. Lack of growth rings in Opuntia species precludes annular dating, but O. echinocarpa specimens taller than about 0.7 m may be 25 years old or more (Park Nobel, pers. comm., 1980). Its average height in our study area was about 1 m, implying that many of the individuals may have been quite old. The decreased survival rate of O. echinocarpa associated with increasingly denser vegetation found with elevational increase suggests burning intensity may set additional limits to its distribution. Data from near Tucson, Arizona, indicate that the survival fre- quency of L. tridentata after controlled burning is directly propor- 1981] O’LEARY & MINNICH: DESERT FIRE ECOLOGY 65 tional to fire intensity and season (White, 1968; Cable, 1972). Con- trolled burns in June of sparse cover of native grasses, augmented by the addition of straw, produced up to 100 percent mortality, whereas lower intensity burns on native grasses are much less damaging. Jump- ing cholla (Opuntia fulgida) and cane cholla (O. spinosior) are also susceptible to incineration (Humphrey, 1949, 1974; Reynolds and Bohning, 1956; Cable, 1967, 1972). Cattlemen have long used fire to help control cholla and mesquite. The intergradation of the creosote bush scrub into the higher ele- vation shrub communities reflects physiologic tolerance along a to- pographic moisture gradient. Beatley (1974, p. 260) suggested that the northern boundary and upper elevational limits of L. tridentata in southern Nevada “are determined primarily by rainfall in excess of a critical amount, and the rainfall regimes over probably a very long time.” She estimated that amount to be 183 mm. The upper elevational distribution of L. tridentata and O. echinocarpa also may be con- trolled locally by their susceptibility to periodic and intense fire. Dry fuel values exceeding critical levels could result in large scale mortal- ity. We suggest that in the desert scrub-coastal sage scrub ecotone, the intensity and periodicity of wildfires may be additional factors to phys- iologic tolerance in limiting the distribution and abundance of some desert perennials. ACKNOWLEDGMENTS We thank Jack Leishman for field assistance, Joan Drake for cartographic assistance, and Jonathan Sauer, Walter Westman, J. R. Griffin and an anonymous reviewer for critical comments that greatly improved the manuscript. LITERATURE CITED BARBOUR, J. G. and D. V. Diaz. 1973. Larrea plant communities on bajada and moisture gradients in the United States and Argentina. Vegetatio 28:335-352. BAUER, H. L. 1943. The statistical analysis of chaparral and other plant communities by means of transect samples. Ecology 24:45—60. BEATLEY, J. C. 1974. Effects of rainfall and temperature on the distribution and behavior of Larrea tridentata (creosotebush) in the Mojave Desert of Nevada. Ecology 55:245-261. Burk, J. H. 1977. Sonoran Desert vegetation. 7x: M. Barbour and J. Major, eds., Terrestrial vegetation of California, p. 869-889. Wiley-Interscience, New York. CABLE, D. R. 1967. Fire effects on semidesert grasses and shrubs. J. Range Managem. 20:170-176. . 1972. Fire effects in south western semidesert grass-shrub communities. Tall Timbers Fire Ecol. Conf. 12:109-127. CZEKANOWSKI, J. 1909. Zur differential Diagnose der Neandertalgruppe. Korrespon- denzbl. Deutsch. Ges. Anthropol. 40:44—47. DAVIDSON, E. and MARTHA Fox. 1974. Effects of off-road motorcycle activity on Mojave Desert vegetation and soil. Madrono 22:381-390. HUMPHREY, R. R. 1949. Fire as a means of controlling velvet mesquite, burroweed, and cholla in southern Arizona. J. Range Managem. 2(4):175-182. . 1974. Fire in the deserts and desert grassland of North America. Jn: T. T. 66 MADRONO [Vol. 28 Kozlowski and C. E. Ahlgren, eds., Fire and Ecosystems, p. 365-400. Academic Press, New York. Munz, P. A. 1974. A flora of southern California. Univ. California Press, Berkeley. REYNOLDS, H. G. and J. W. BOHNING. 1956. Effects of burning on a desert grass- shrub range in southern Arizona. Ecology 37:769-778. S@RENSEN, T. 1948. A method of establishing groups of equal amplitude in plant sociology based on similarity of species content. Kongel. Danske Vidensk.-Selsk. Biol. Skr. 5(4):1-34. TRATZ, W. M. and R. J. VOGL. 1977. Postfire vegetational recovery, productivity, and herbivore utilization of a chaparral-desert ecotone. Symposium on the Envi- ronmental Consequences of Fire and Fuel Management in Mediterranean Ecosys- tems. U.S.D.A. For. Serv. Gen. Tech. Rep. WO-3. U. S. FOREST SERVICE. undated. Fire maps for the Mt. San Jacinto District, 1911- 1973. Unpubl. reports on file, Pacific Southwest Region, San Bernardino Natl. Forest, San Bernardino, CA. VASEK, F. C., H. B. JOHNSON, and D. H. ESLINGER. 1975. Effects of pipeline con- struction on creosote bush scrub vegetation of the Mojave Desert. Madrono 23:1- £3: VASEK, F. C., H. B. JOHNSON, and T. YONKERS. 1975. Productivity, diversity and stability relationships in Mojave Desert roadside vegetation. Bull. Torrey Bot. Club. 102:106-115. WHITE, L. D. 1968. Factors affecting susceptibility of creosote bush (Larrea tridentata (D.C.) Cov.) to burning. Ph.D. dissertation, Univ. Arizona, Tucson. (Received 14 Feb 1980, revision received and accepted 6 Oct 1980.) ANNOUNCEMENT PAPERS IN WESTERN PLANT ECOLOGY A symposium in honor of the retirement of Professor Jack Major will be presented on Friday, 29 May 1981. It will be held at the University of California, Davis, campus in Memorial Union, Room II (upstairs), from 9 a.m. to 5 p.m. Approximately 12 speakers will present research or overview papers on “Western Plant Ecology.” There will be no registration fee, but participants must provide their own lunches. The symposium is being co-sponsored by the California Botanical Society; the UC Davis Institute of Ecology; and the Botany Department, UC Davis. Speakers and tentative titles are as follows: G. L. Stebbins, Floristic affinities of the high Sierra of California; D. J. Parsons, Vegetation management and research in Sequoia and Kings Canyon National Parks; J. L. Vankat, Vegetation classification and analyses in Sequoia National Park; P. W. Rundel, Nutrient cycling and chaparral in Sequoia and Kings Canyon National Parks; N. Benedict, Mountain meadows: stability and change; M. B. Morgan, Vegetation of the Rae Lakes Basin, southern Sierra Nevada; D. I. Axelrod, Age and origin of the Monterrey endemic area; J. R. Griffin, Pine seedlings, native herbs, and annual ryegrass on the Marble-Cone burn, Santa Lucia Mountains; J. A. Neilson, Distribution, ecology, and proposed revision of the Strep- tanthus morrisonii complex; R. Macdonald, Patterns of xylem sap tension and conduc- tance of foothill woodland vegetation types near Lake Berryessa; R. Gankin, Ecology and land use planning in San Mateo County; J. H. Burk, Phenology, germination, and survival of ephemerals in Deep Canyon; C. B. Davis, Seed banks and vegetation history in prairie wetlands. For further information call the UC Davis Department of Botany: (916) 752-0617. SEEDLING CHARACTERISTICS AND ELEVATIONAL DISTRIBUTIONS OF PINES (PINACEAE) IN THE SIERRA NEVADA OF CENTRAL CALIFORNIA: A HYPOTHESIS RICHARD I. YEATON Department of Biology, Pomona College, Claremont, CA 91711 ABSTRACT Seven species of the genus Pinus occupy a 3000-m elevational gradient on the western slope of the Sierra Nevada in central California. Three species of 3-needled pines replace each other abruptly over the mountainsides as do three species of 5-needled pines. The seventh species, a 2-needled pine, is found on the upper half of the mountainsides. Observed associations of pines on these slopes consist of a 2- or 3-needled species (or both) with a 5-needled species. A hypothesis is presented that interspecific competition between seedlings, ultimately for water, has led to the distributional patterns observed. This hypothesis is examined in terms of the morphological characteristics of seedlings and their needles that reflect their relative usage of water. Storer and Usinger (1963) portray seven species of the genus Pinus forming a guild over a 3000-m elevational gradient on the western slope of the Sierra Nevada in central California. The distributional pattern of these seven species as described is striking. Three species of 3-needled pines replace each other sequentially over the elevational gradient, as do three species of 5-needled pines. When one species disappears, another appears with the same number of needles per fascicle. The change with elevation is portrayed as abrupt with little or no overlap occurring between species with similar needle numbers. Also, a 2-needled species of pine is found over the upper half of the elevational gradient. Up to about 915 m there is only a 3-needled species (P. sabiniana). From 915 m to timberline several associations of pines are found. Each of these higher-elevation groups is composed of a 2-needled or a 3-needled species, or both, associated with a 5- needled pine. At no place on the mountainside are two 3-needled or two 5-needled pines described as occurring together. If the distribution of pines depicted by Storer and Usinger is accurate, the pattern of needle numbers per fascicle requires explanation. I suggest that this distributional pattern represents an example of the competitive exclu- sion principle in which interspecific competition has led to the elevation- al separation of species with similar needle number. If so, there must be ecological differences between pine species with different needle num- bers because various combinations of these species do occur. In this paper I demonstrate that the distributional patterns described by Storer and Usinger are accurate. Secondly, I present a hypothesis to explain the associations of pines that are found on the western slopes MADRONO, Vol. 28, No. 2, pp. 67—77, 3 June 1981 67 68 MADRONO [Vol. 28 of the central Sierra Nevada. Finally, I attempt to examine this hy- pothesis by a comparison of some of the morphological characteristics of species that either overlap extensively or replace each other abruptly over the elevational gradient. ELEVATIONAL DISTRIBUTIONS The elevational distribution of pines on transects of the western slopes of the Sierra Nevada in central California were determined by counting 100 individuals of the genus at stations separated from one another by 77 m altitude along State Highways 4, 88, and 108 (38°16— 34'N, 120°0-55’W; Amador, Calaveras, Alpine, and Tuolumne Coun- ties). Trees taller than 15 m were counted on 10-m-wide strips running along the contour of southeastern facing slopes at each station. Sam- pling was limited to southeastern slope aspects to reduce the variability in species abundance resulting from the different microclimates that exist at the same elevation but on different slope aspects. Only species of Pinus were counted and their relative proportions calculated at each station and plotted (Fig. 1a, b, c). In this area of the Sierra Nevada only seven species of Pinus occur: four members of the subgenus Diploxylon and three species of the subgenus Haploxylon. Pinus sabiniana Dougl. (group Macrocarpa), P. ponderosa Laws., and P. jeffreyi Grev. and Balf. (the latter two species members of the group Australis) compose the sequence of 3- needled Diploxylon pines while P. contorta Dougl. (group [nsignis) is the 2-needled Diploxylon species. Pinus lambertiana Dougl., P. monticola Doug]. (both members of the group Strobz) and P. albicau- lis Engelm. (group Cembrae) make up the 5-needled sequence of Hap- loxylon pines (taxa follow Shaw, 1914 and Mirov, 1967). For species of pines with three needles per fascicle abrupt replacements occur at approximately 840 m and 1740 m in this region. The latter replacement involves P. ponderosa and P. jeffreyi, which hybridize at the inter- faces of their distributions (Mirov, 1967). In the region studied the zone of possible hybrids is small (approximately 25 m in elevational range) and falls between sampling sites. Identifications of canopy in- dividuals of these two species are based upon gross characteristics of their barks and cones. For 5-needled species abrupt replacements oc- cur at approximately 1980 m and 2630 m (Fig. 1). For the four possible pairs of replacing species along the mountainsides, in only one instance do two species of Pinus with the same needle number occur together in the same sample. This overlap is for the high elevation, low relative abundance situation in which P. albicaulis replaces P. monticola. Thus the distributional pattern described by Storer and Usinger (1963) is verified. 1981] YEATON: SIERRA NEVADA PINE ASSOCIATIONS 69 100 - 50 1000 2000 3000 100 b u c ne > 50 0 O 32 1000 2000 3000 100 Cc 50 1000 2000 3000 Elevation (m) Fic 1. Percent occurrence of Pinus only for the seven species found on southeastern slope aspects on the western slope of the Sierra Nevada in central California. a. 3- needled pines. b. 2-needled pine. c. 5-needled pines. 70 MADRONO [Vol. 28 THE HYPOTHESIS A very precise pattern of replacements is observed among the larger individuals suggesting that their elevational distributions are deter- mined by interspecific competition between seedlings and younger age classes. Seedlings of Pinus with different needle numbers per fascicle that are associated elevationally are often found growing together in clumps along with other species of conifers. Likewise, I have found seedlings of species with the same needle number associated in clumps at elevations at which one species replaces another abruptly. Hence the possibility for intense competition exists between seedlings growing in such close proximity within these clumps. Water would seem to be the important limiting factor in this relatively open-canopied com- munity. It has been suggested that water is critical in the early life of pine seedlings (Pearson, 1930). Seedlings are dependent upon the mois- ture in the upper soil layers, where surface evaporation and transpi- ration rapidly deplete the available soil moisture (Stone and Jenkinson, 1971). For example, young ponderosa pines suffer intense intraspecific competition for available soil moisture (Weaver, 1961). Further, the smallest individuals of Pinus suffer more from decreases in moisture levels than do larger individuals (Turner, 1956; Weaver, 1961; Yeaton, 1978). Seedlings are probably most severely affected by environmental fluctuations because they are small and therefore do not have the benefit of the stabilizing effect provided by reserves accrued from growth in previous years. Hence, adaptive responses in water usage or habitat usage are most likely to be reflected in seedlings (Ledig et al., 1977). For associated species, clumps of seedlings may be located upon soil patches with varying moisture levels over the course of the growing season. Species of Pinus with adaptations to reduce the effects of moisture stress would be expected to outcompete seedlings of species without these adaptations on patches that are relatively dry over the growing season. Adaptations to reduce the effects of drought might appear as a reduction in total needle surface area over which water may be lost, a decrease in the number of stomates to reduce transpi- rational water loss, or structural differences in needle anatomy to re- duce epidermal water loss. In contrast, species of Pinus successful in competition on wetter patches may have greater total needle surface areas for increased photosynthetic surface area, increased numbers of stomates for higher rates of gas exchange during photosynthesis, and a reduced necessity for a specialized needle anatomy to reduce epi- dermal water loss. Their success on these wetter patches may be the result of an increased growth rate and their ability to store reserves to survive periods in which the characteristics of the patch are stress- ful. Coexistence of species at any one elevation may be maintained by use of patches with differing seasonal moisture regimes ranging from dry to relatively wet. 1981] YEATON: SIERRA NEVADA PINE ASSOCIATIONS 71 Species with the same needle number that replace one another abruptly elevationally are also subjected to a wide range of moisture availability. Species occurring at low elevations are faced with a hot dry summer while the species at the highest elevations are faced with the problem of a short growing season and a winter drought. Presum- ably the central species in each elevational sequence occupies the best available range of conditions for moisture and length of growing sea- son and may outcompete the species above and below it as a result of increased growth rates (Yeaton et al., 1980). NEEDLE NUMBER AND TOTAL NEEDLE LENGTH IN SEEDLINGS The hypothesis suggests that 2-needled pines should occupy drier sites than associated 3- and 5-needled species and that 3-needled pines should use drier sites than their 5-needled associates because the total needle length per fascicle will increase from 3-needled to 5-needled fascicles. It also suggests that species with the same needle number that replace each other are more similar to one another than are species that form associations. However, these relationships may be compli- cated when the whole individual is considered due to differences in needle lengths and number of fascicles. The effect of these differences can be studied by measuring the total needle lengths in all fascicles of seedlings of each species and determining if the relationship observed for fascicles holds when the whole seedling is considered. Needle width may complicate the needle number/needle length re- lationship. Haller (1962) has stated that needle width is unimportant in comparisons of species with different needle numbers and any width differences would not be great enough to affect these measurements. My analysis of needle widths agrees with Haller’s statement. The sur- face area of a fascicle is approximately 10.28rl for a 2-needled species, 12.28rl for a 3-needled species and 16.28rl for a 5-needled species (where “r” is the radius and “Il” is the length of the fascicle). These values result from viewing the fascicle as circular in cross-section. For all fascicles the outer surface area of the fascicle is 27rl. To this must be added 4rl, 6rl, or 10rl for the inner needle surfaces for 2-, 3-, and 5-needled species. Treating length of fascicle as a constant, to attain the surface area of a 5-needled species, a 3-needled associate must be 33 percent greater and a 2-needled species 58 percent greater in needle width. Similarly, a 2-needled species must be 19 percent greater in needle width to attain the same surface area as a 3-needled associate. Table 1 shows average measurements of needle width for specimens of the seven species of pines in this study. The measurements of needle width are very similar for all seven species. The pair of associated species with the greatest difference in needle width is that of the 3-needled Pinus jeffreyi with the 5-needled P. monticola (Table 2). For these species the difference is 33 percent or, in other words, the surface areas per mm of needle length of fascicles 72 MADRONO [Vol. 28 TABLE 1. ELEVATIONAL RANGE AND MEAN NEEDLE WIDTH (+ S.E.) FOR THE SEVEN SPECIES OF Pinus ON THE WESTERN SLOPES OF THE SIERRA NEVADA. N = 10 for needle-width measurements. Needles per Elevational Needle width Species fascicle range (m) (mm) Pinus sabiniana 3 100-915 0.95 + 0.03 Pinus ponderosa 3 915-1740 0.88 + 0.04 Pinus jeffreyi 3 1740-2670 0.81 + 0.04 Pinus lambertiana 5 930-1980 . 0.70 + 0.03 Pinus monticola 5 1980-2630 0.61 + 0.02 Pinus albicaulis 5 2630-3000 0.84 + 0.03 Pinus contorta 2 1930-3000 0.72 + 0.03 are equal. For all other pairs of associated species the difference in needle width is smaller than that required for this parameter to have a significant effect on the surface area of species with different num- bers of needles per fascicle. Hence, needle width is not a factor in the relationship of needle number per fascicle for associated and eleva- tionally replacing species. To study the effects of needle length, branching patterns, and spac- ing of needles along branches, the total needle lengths for seedlings of varying stem diameters were measured. Seedlings of the seven species of pines were found in areas adjacent to California State Highways 4 (through Ebbetts Pass), 88 (through Carson Pass) and 108 (through Sonora Pass). Seedlings were sampled in the centers of the distribution of each respective species. At least 25 seedlings of each species were counted. Seedlings were selected whose stems ranged from 0.5 to 30 mm in diameter as measured by calipers at a point 2.5 cm above the ground and whose diameters were fairly uniformly spaced from one another over this range. Individuals were chosen that had no apparent TABLE 2. PERCENTAGE DIFFERENCES IN NEEDLE WIDTHS FOR ELEVATIONALLY ASSOCIATED PAIRS OF PINES. Percent difference needed for needle width to be impor- tant for measurement of needle surface area for associated species of Pinus calculated as (needle width of species with fewer needles per fascicle—needle width of associated species)/(needle width of species with greater number of needles per fascicle) < 100. Negative percent indicates that needle width is greater for species with more needles per fascicle. Expected % Observed % Pairs (needles per fascicle) difference difference P. albicaulis (5)—P. contorta (2) 58 —14 P. monticola (5)—P. contorta (2) 58 18 P. jeffreyi (3}—P. contorta (2) 19 —11 P. monticola (5)—P. jeffreyi (3) 33 33 P. lambertiana (5)}—P. ponderosa (3) 46) 26 1981] YEATON: SIERRA NEVADA PINE ASSOCIATIONS 73 TABLE 3. THE RELATIONSHIP BETWEEN SEEDLING STEM DIAMETER AND TOTAL NEEDLE LENGTH FOR SEVEN SPECIES OF THE GENUS Pinus OCCURRING ON THE WESTERN SLOPES OF THE SIERRA NEVADA IN CENTRAL CALIFORNIA. Data fit a power curve, y = cx® where y = total needle length, c = a constant, x = seedling stem diameter, and E = experiment. “r’” is the coefficient of determination. Number Species sampled Constant Exponent me P. sabiniana 25 819 1.83 0.90 P. ponderosa 25 1298 1.76 0.98 P. jeffreyi 27 1187 1.69 0.90 P. lambertiana 25 1656 1.86 0.96 P. monticola 25 679 2.08 0.90 P. albicaulis 25 526 2.16 0.96 P. contorta 25 638 1.82 0.96 damage from grazing herbivores. For each seedling the diameter of each branch off the main stem was measured and all the fascicles of needles on that branch and its sub-branches counted. The length of a typical fascicle of needles on each branch was measured. This needle length was multiplied by the number of needles in the fascicle and that product multiplied by the number of fascicles counted on the branch. The total needle length for each branch on the seedling was then summed. A similar procedure was employed for fascicles growing out of the main stem between branches and these were added to the sum of the branches to give a total needle length per seedling. The relationship between total needle length per seedling and the diameter of that seedling is best fitted by a power curve. For each species of pine studied, the total needle length per seedling is equal to TABLE 4. PAIRWISE COMPARISONS OF ASSOCIATED PINE SPECIES AND ALTITUDI- NAL REPLACEMENTS. Values in parentheses are mean ratios of the square root of total needle length to stem diameter. Comparisons of the mean ratios were made using the median test. Species pairs x? P Altitudinal Replacements P. sabiniana (24.7)—P. ponderosa (28.1) 2.88 0.05 < p< 0.10 P. ponderosa (28.1)—P. jeffreyi (25.3) Zit 0.05 < p< 0.10 P. lambertiana (35.9}—P. monticola (29.4) 5.12 p= 0705 P. monticola (29.4)—P. albicaulis (27.7) 0 p = 1.00 Altitudinal Associates P. contorta (21.0)—P. albicaulis (27.7) 15.68 p < 0.001 P. contorta (21.0)}—P. jeffreyi (25.3) 7.70 p < 0.01 P. contorta (21.0)—P. monticola (29.4) L152 p < 0.001 P. jeffreyi (25.3)—P. monticola (29.4) 710 p < 0.01 P. ponderosa (28.1)—P. lambertiana (35.9) 8.00 p < 0.01 74 MADRONO [Vol. 28 a constant value times the seedling diameter approximately squared. All species fit this model well with correlation coefficients (r) of 0.95 or higher (Table 3). These data can be used to answer several questions about the species that are either associated or replacing along the western slopes of the Sierra Nevada. Are species that are associated elevationally signifi- cantly different from one another in their total needle lengths? Are associated species different such that 5-needled pines have greater total needle lengths than their 3- or 2-needled associates? What relationships exist between total needle lengths for seedlings of species that replace abruptly over the elevational gradient? Are they more similar to one another than seedlings of species that form associations? To answer these questions the ratio of the square root of total needle length to stem diameter was calculated for each seedling measured in the study. This ratio was used because the total needle length per seedling is approximately equal to the product of stem diameter squared and a species-specific constant (Table 3). Reduction of the data in this fashion reveals the sample value of this constant for each individual measured. Comparisons of these converted data were made using a Median test (Siegel, 1956) for sets of seedlings of elevationally associated or replacing pairs of species. All ratios of square root of total needle length to stem diameter for each species of 3- and 5- needled pine compared with the species replacing them higher on the mountainside have chi-square values with probabilities of occurrence greater than 0.01. Conversely, pairwise comparisons between pines that co-occur on the mountainside all have chi-square values whose probabilities of occurrence are less than 0.01 (Table 4). Furthermore, 5-needled pines attain significantly greater total needle lengths per individual than do their 2- or 3-needled counterparts. Thus, the species that replace one another elevationally have less different total needle lengths than do associated species. The pattern of needle numbers in adult trees is a reflection of the total needle surface areas of their seedlings. This pattern for seedlings, in which species similar in total needle length per seedling replace one another elevationally and in which species dissimilar in total needle length overlap elevationally is consistent with the hypothesis that water availability for seedlings determines which species can occur together. DISCUSSION The data reported here are consistent with the hypothesis that com- petition between seedlings for soil moisture has led to the distributional patterns observed for the seven species of Pinus occupying the western slopes of the Sierra Nevada. Species that are morphologically similar to one another in characters reflecting water use are displaced over 1981] YEATON: SIERRA NEVADA PINE ASSOCIATIONS 15 the gradient, while species dissimilar in these characteristics overlap extensively over the same gradient. Additional evidence from the lit- erature supports the hypothesis. Leaf anatomy is one area of possible evidence although morphological differences in their functional sense may not be clearly understood. Harlow (1947) has described the cross- sectional leaf anatomy for most species of North American pines. All Diploxylon (2- and 3-needled) species included in this study have well developed dermal regions, particularly in the hypodermal layer. The hypodermal layer is described as either biformed or multiformed with thick-walled cells. In contrast the dermal regions of the three species of Haploxylon (5-needled) pines are less well developed with no thick- ening of the hypodermal cell walls. Thickened cell walls in the dermal region of the Diploxylon species may be indicative of greater drought stresses. For example, those Haploxylon species of the group Para- cembra (Pinus monophylla and P. edulis) occupying habitats at the edge of desert regions in the southwestern United States where mois- ture stress is high are described by Harlow (1947) as having a hypo- dermal layer two to four cells thick and the cells with thickened cell walls. It may be that the thickened cell walls and more complex de- velopment of the hypodermal layers in the Dzploxylon species reflect adaptation to moisture stress to reduce epidermal water loss. A second source of information on the relative water use of asso- ciated Pinus species is data for photosynthetic and transpiration rates. Unfortunately there is very little direct evidence on rates for associated species of pines. Miscellaneous evidence does exist for two sets of associated species, P. ponderosa and P. lambertiana in the Sierra Nevada of California and P. banksiana, P. resinosa, and P. strobus in the Great Lakes region of the United States and Canada. Snow (1924), studying Pinus ponderosa (3-needled) and P. lambertiana (5- needled), reported that shade was necessary for survivorship of P. lambertiana seedlings. In contrast, shade aided but was not necessary for survivorship of P. ponderosa seedlings. He concluded that P. ponderosa was more drought resistant than P. lambertiana. A similar observation about the drought resistance of P. lambertiana and P. ponderosa was made by Pharis (1966) with a suggestion that associa- tions of these two species optimize water usage by tapping different subsurface zones of the soil. More extensive work has been done with Pinus strobus and its 2- and 3-needled associates. Belyea (1925) reported that P. strobus (5- needled) is more susceptible to wind desiccation than its associate P. vesinosa (long, 2-needled) and attributed the difference to the greater leaf surface area represented by the 5-needled species. Similar studies contrasting P. strobus with P. resinosa, P. banksiana (short, 2- needled), or P. rigida (3-needled) always place P. strobus on wetter sites than its 2- or 3-needled associates (Hutchinson, 1918; Cook et al., 1952). In a comparison of water loss in three of these species.that 76 MADRONO [Vol. 28 occur together in the vicinity of the Great Lakes, Walter and Kozlow- ski (1964) rank water loss as P. strobus > P. resinosa > P. banksiana. Brown and Curtis (1952), in a gradient analysis of the upland conifer- hardwood forests of northern Wisconsin, ranked these three species relative to the moisture-holding capacity of the soils. P. banksiana occurs on dry soils, P. resinosa on intermediate soils, and P. strobus on soils with greater moisture-holding capacity. While P. resinosa and P. banksiana are both 2-needled pines, P. resinosa has needles four times as long as P. banksiana. Other authors have reported similar patterns in water usage for other pairs of associated species differing in needle number per fascicle [Larson (1927) for P. monticola and P. contorta; Coile (1933) for P. taeda and P. echinata; Pessin (1933) for P. palustris and P. clausa; Wright (1966, 1968, 1970) for P. lambertiana and P. coulteri]. In all cases the species with the greater needle number occupied what were described as relatively wetter sites. Despite the observations for the seven species of Pinus on the west- ern slope of the central Sierra Nevada reported above, no direct evi- dence for interspecific competition between seedlings of these species exists. Both Daubenmire (1943) and Haller (1959) have suggested that water stress and/or interspecific competition are critical in determin- ing, at least in part, elevational ranges of species in western moun- tains. To test the hypothesis that interspecific competition is the or- ganizing mechanism for these seven species of Pinus, long term field studies of seedling establishment, growth, and mortality as well as transplant experiments are in progress for the two elevationally re- placing species Pinus sabiniana and P. ponderosa, and for the two elevationally associated species P. ponderosa and P. lambertiana. In addition, because most pairs of associated species comprise a Haplo- xylon and a Diploxylon species, further work on the comparative water requirements of these two subgenera is suggested, particularly with respect to differences in stomatal responses to increasing drought stress. Finally the distributional patterns observed in the central Sierra Nevada must be expanded to other sets of pines to see the generality of the patterns described here. ACKNOWLEDGMENTS I thank R. W. and L. C. Yeaton for their assistance in the field; D. Janzen, D. Guthrie, and D. Soltz for their comments on the manuscript; and V. Vance and J. and A. Waggoner for their hospitality and good wine. A grant from the American Philo- sophical Society subsidized part of the field and travel expenses and is gratefully ac- knowledged. LITERATURE CITED BELYEA, H. C. 1925. Wind and exposure as limiting factors in the establishment of forest plantations. Ecology 6:238—240. Brown, R. T. and J. T. Curtis. 1952. The upland conifer-hardwood forests of northern Wisconsin. Ecol. Monogr. 22:217—234. 1981] YEATON: SIERRA NEVADA PINE ASSOCIATIONS 77 CoILE, T. S. 1933. Soil reaction and forest types in the Duke Forest. Ecology 14:323- 335: Cook, D. B., R. H. SMITH, and E. L. STONE. 1952. The natural distribution of red pine in New York. Ecology 33:500-512. DAUBENMIRE, R. F. 1943. Soil temperature versus drought as a factor determining lower altitudinal limits of trees in the Rocky Mountains. Bot. Gaz. (Crawfordsville) 105:1-13. HALLER, J. R. 1959. Factors affecting the distribution of ponderosa and Jeffrey pines in California. Madrono 15:65—71. . 1962. Variation and hybridization in ponderosa and Jeffrey pines. Univ. Cal- ifornia Publ. Bot. 34:123-166. HarLow, W. M. 1947. The identification of the pines of the United States, native and introduced, by needle structure. Bull. New York Coll. For. Syracuse Univ. 32:1- 57. HUTCHINSON, A. H. 1918. Limiting factors in relation to specific ranges of tolerance of forest trees. Bot. Gaz. (Crawfordsville) 66:465—492. LARSON, J. A. 1927. Relation of leaf structure of conifers to light and moisture. Ecology 8:371-377. LepiGc, F. T., J. G. CLARK, and A. P. DREw. 1977. The effects of temperature treatment on photosynthesis of pitch pine from northern and southern latitudes. Bot. Gaz. (Crawfordsville) 138:7-12. Mrirov, N. T. 1967. The genus Pinus. Ronald, New York. PEARSON, G. A. 1930. Light and moisture in forestry. Ecology 11:145—160. PESSIN, L. J. 1933. Forest associations in the uplands of the lower Gulf Coastal plain (longleaf pine belt). Ecology 14:1-14. PHARIS, R. P. 1966. Comparative drought resistance of five conifers and foliage mois- ture content as a viability index. Ecology 47:211-221. SHAW, G. R. 1914. The genus Pinus. Publ. Arnold Arbor. 6:1—96. SIEGEL, S. 1956. Nonparametric statistics. McGraw-Hill, New York. SNow, S. B. 1924. Some results of experimental forest planting in northern California. Ecology 5:83—94. STONE, E. C. and J. L. JENKINSON. 1970. Influence of soil water on root growth capacity of ponderosa pine transplants. Forest Sci. 16:230-239. STORER, T. L. and R. L. USINGER. 1963. Sierra Nevada natural history: an illustrated handbook. Univ. California Press, Berkeley. TURNER, R. M. 1956. A study of some features of growth and reproduction of Pinus ponderosa in northern Idaho. Ecology 37:742-753. WEAVER, H. 1961. Ecological changes in the ponderosa pine forest of Cedar Valley in southern Washington. Ecology 42:416—420. WALTER, K. E. and T. T. KOZLOWSKI. 1964. Transpiration capacity of dormant buds of forest trees. Bot. Gaz. (Crawfordsville) 125:207-211. WRIGHT, R. D. 1966. Lower elevational limits of montane trees. I. Vegetational and environmental survey in the San Bernardino Mountains of California Bot. Gaz. (Crawfordsville) 127:184—-193. 1968. Lower elevational limits of montane trees. II. Environmental keyed responses of three conifer species. Bot. Gaz. (Crawfordsville) 129:219-226. . 1970. Seasonal course of CO, exchange in the field as related to elevational limits of pines. Amer. Midl. Naturalist 83:321-329. YEATON, R. I. 1978. Competition and spacing in plant communities: differential mor- tality of white pine (Pinus strobus L.) in a New England woodlot. Amer. Midl. Naturalist 100:285—293. YEATON, R. I., R. W. YEATON, and J. E. HORENSTEIN. 1980. The altitudinal re- placement of digger pine by ponderosa on the western slopes of the Sierra Nevada. Bull. Torrey Bot. Club 107:487-495. (Received 28 Apr 1980; revision accepted 28 Nov 1980.) PORTULACA JOHNSTONIT, A NEW SPECIES OF PORTULACACEAE FROM THE CHIHUAHUAN DESERT JAMES HENRICKSON Department of Biology, California State University, Los Angeles 90032 ABSTRACT Portulaca johnstonii, distinguished by two radial series of tapering, erect fimbriae around the seeds, is described from the Bolson de Mapimi region of the Chihuahuan Desert, Coahuila, Mexico. It is most closely related to P. retusa. While preparing a treatment of Portulacaceae for Marshall John- ston’s Chihuahuan Desert Flora, a distinctive new species of Por tulaca was encountered and is described below. Portulaca johnstonii Henrickson sp. nov. A Portulaca retusa siminibus ad marginem trichomatibus subulatis effusis biseriatis differt (Fig. 1). Glabrous, fleshy, decumbent-ascending annuals 1-2 dm wide. Leaves often subopposite, ovate-spathulate, 2-14 mm long, 1-4 mm wide (to probably larger), obtuse, rounded to truncate at tip, cuneate at base, petioles 1-2 mm long, caniculate, at margins entire, axils with fimbriate white setae 0.2—0.7 mm long. Flower 1-2 at tips of lateral branches, mostly subtended by 1-2 pairs of leaves and a pair of nar- rowly ovate, acute-acuminate scarious bracts 1.2-3 mm long, pedicels ca. 1 mm long; petals united below, 2—3.5 mm long, yellow, lobes acute; stamens 5-8, filaments united to base of corolla for 0.5-0.7 mm, free filaments 1—-1.5 mm long, puberulent above base, anthers 0.3-—0.4 mm long; style 1.5 mm long, lobes 3, ca. 0.6 mm high. Fruit 3—4 mm high, circumscissile dehiscent medially, the lid widely conical, 2-3 mm broad at base, usually constricted below tip, overtopped by paired, green sepals 3-4 mm high, these dehiscing circumscissilly with fruit, each with a medial vertically raised, crest-like keel; seeds 2-15, 1.3-1.6 mm in total diameter, body reddish-brown, 0.7—1 mm in diameter, compressed, with 3—4 concentric rows of low, radially elongated tubercules on each side, at margins with 2 rows of rust- colored, conspicuous, subulate, firm fimbriae 0.3—0.4 mm long, base of seed with a small white caruncle 0.1—0.2 mm long. TYPE: Mexico, Coahuila; Matrimonio Nuevo on road paralleling railroad between Esmeralda and Cuatro Cienegas (near 27°08'N; 103°10’W), locally common in desert flat in gravelly calcareous adobe- clay, 1075 m, 2 Sep 1972, F. Chiang C., T. L. Wendt, and M. C. Johnston 9125 (Holotype LL; isotype MEXU). This distinctive species, known only from the type collection, is MADRONO, Vol. 28, No. 2, pp. 78-79, 3 June 1981 78 1981] HENRICKSON: PORTULACA JOHNSTONII 79 Fic. 1. Portulaca johnstonii Henrickson. A. Habit. B. Leaf base showing fimbriate white setae. C. Mature circumscissile capsule showing crested conical lid, seeds, sub- tending bract, and setae. D. Seed showing radiating fimbriae and small, white, basal caruncle. E. Detail of fimbriae 0.3—0.4 mm long on seed margin, showing their orien- tation into two rows. vegetatively very similar to P. retusa Engelm. but is immediately distinguishable from this and all other species in the genus by the radiating fimbriae on the seeds. Both species are also very similar vegetatively to the widespread, weedy P. oleracea L. Portulaca johnstonii, named for Marshall C. Johnston, grows in open, clay Tobosa flats with Hilaria, Sporobolus, Ericameria, and Prosopis in the Bolson de Mapimi region of the Chihuahuan Desert. ACKNOWLEDGMENTS M. C. Johnston provided the Latin diagnosis, Bobbi Angell the illustration. I thank the University of Texas Plant Resource Center for use of facilities. (Received 9 Sep 1980; accepted 10 Oct 1980.) MALEPHORA CROCEA (AIZOACEAE) NATURALIZED IN CALIFORNIA WAYNE R. FERREN, JR. JOHN BLECK Department of Biological Sciences, University of California, Santa Barbara 93106 NANCY VIVRETTE | Santa Barbara Botanic Garden, Santa Barbara, California 93105 ABSTRACT Malephora crocea, native to Cape Province, South Africa, has been naturalized in southern California at least since 1946, but is not mentioned in California floras. Mis- identified specimens have served as the basis for incorrect reports of other species. We document the historical occurrence and current range of M. crocea in California and provide a morphological description, illustration, and comparison with other ice plants. Identification of naturalized plants often is difficult and collections are frequently misidentified. The mesembryanthemums or ice plants (Aizoaceae, sensu Melchior, 1964) are particularly difficult because of the size and complexity of the group and lack of keys to many species. The problem was amplified by the division of Mesembryanthemum sensu lato (=~Mesembryanthemaceae, Herre and Volk in Schwantes, 1947; Schwantes, 1971) into many genera. Herre (1971) distinguished 125 genera and estimated 2400 species, mostly of xeric habitats in South Africa. Hundreds of species have been cultivated in California; by 1930, Hoffman and coworkers had listed 86 for Santa Barbara. At least 22 have been used as groundcovers in California (Kimnach, 1966). Several persist after cultivation or are naturalized to some de- gree. Moran (1950) and Munz (1974) considered Carpobrotus aequt- laterus (Haw.) N. E. Br. [=Carpobrotus chilensis (Mol.) N. E. Br.] to be native. Munz listed nine species as naturalized in southern Cal- ifornia. This figure is misleading because of misidentifications. Malephora crocea (Jacq.) Schwant. (Fig. 1) is native to Cape Province, South Africa (Jacobsen, 1960, 1977). It is commonly planted in Cali- fornia in gardens and on streetsides and highway embankments. It is used for erosion control on moderate slopes and is drought tolerant (Anonymous, 1979). Some 15 herbarium specimens show it to be cul- tivated from Marin and Fresno Counties, California, to Baja California Norte, including Cedros Island, Mexico. A specimen from the James West nursery, San Rafael (West s.n., CAS) shows it was in California by 1933. According to Poindexter (1934), a purple-flowered color form was introduced at about the same time by Kate Sessions, a San Diego MADRONO, Vol. 28, No. 2, pp. 80-85, 3 June 1981 80 1981] FERREN ET AL.: MALEPHORA CROCEA 81 nurserywoman. Also, a specimen was collected in 1938 from E. O. Orpet’s Santa Barbara nursery (Hartwell s.n., SBM). Malephora cro- cea was not listed for the Santa Barbara region previously (Hoffman et al., 1930). Orpet later supplied material for freeway plantings to the State of California Division of Highways (Anonymous, 1954). Malephora crocea is naturalized in western North America mostly along the coast from northern Santa Barbara County to Baja Califor- nia Norte, Mexico. Some 26 herbarium specimens show it established at Surf, UCSB Campus, Goleta Slough, Santa Barbara, and Carpin- teria Salt Marsh, Santa Barbara County; Ventura, Ojai, Pt. Mugu, and East Anacapa Island, Ventura County; Newport Backbay, Dana Point, and Doheny Beach State Park, Orange County; Riverside, Riv- erside County; and La Mision, Baja California Norte. It occurs as a garden escape in the vicinity of dwellings and along roadsides and has become established on sea bluffs, stream banks, floodplains, coastal sage scrub habitats, and margins of estuaries. Near estuaries M. cro- cea is especially well-established and grows in several situations. It occurs in open sand and silt of disturbed areas above storm tide, usually with Atriplex patula L. subsp. hastata (L.) Hall & Clem., A. semibaccata R. Br., Carpobrotus edulis (Haw.) Schwant., Frankenia grandifolia Cham. & Schlecht., Parapholis incurva (L.) C. E. Hubbs, Salicornia subterminalis Parish, Spergularia marina (L.) Griseb., and Suaeda californica Wats. var. taxifolia (Standl.) Munz. It is locally abundant on shell middens exposed in salt marshes at Newport. This is the only habitat where we have observed numerous seedlings. It also occurs in debris at mean high and storm high tide lines where it establishes apparently from stems deposited there. Under these con- ditions it often grows with Cakile maritima Scop., Carpobrotus edulis, Osteospermum fruticosum (L.) Norl., and other naturalized and native plants dispersed by tides. Although herbarium specimens show M. crocea naturalized at least since 1946, California floras omit it (Munz, 1959, 1968, 1974; Smith, 1976). Most specimens in California herbaria have been misidentified, particularly as Disphyma crassifolium (L.) Bol. and Drosanthemum speciosum (Haw.) Schwant. Apparently it was from such misidentified specimens that Munz (1959, 1974) reported these two species in south- ern California. The same plants were included in Shetler and Skog (1978) based on Munz’s report and in Kartesz and Kartesz (1980). Although D. speciosum is cultivated in California, we have seen no evidence that it is naturalized. Howell et al. (1958) report that Dzs- phyma crassifolium occurs in San Francisco. We have seen Howell 32930 (CAS) from above Point Lobos, San Francisco, and Walther s.n. (CAS), which lacks locality data. All other specimens labeled D. crassifolium are M. crocea. Other ice plants also have been confused with M. crocea. For ex- ample, some herbarium sheets of it have been labeled Carpobrotus 82 MADRONO [Vol. 28 Fic. 1. Malephora crocea. a. Habit, showing old and new stem and rooting at nodes. b. Flower, with petals and staminodia. c. Capsule, closed. d. Capsule, open, showing valve wings, seed pockets, and bifid placental tubercles. e. Ovary, cross sec- tion. f. Leaf, cross section. g. Seed, showing rows of tubercles. aequilaterus. This error probably resulted from use of a key to species of Mesembryanthemum in Munz (1959) and to genera of the Aizoaceae in Munz (1974). Misidentified specimens of M. crocea also occur on a herbarium sheet with a pink form of Carpobrotus edulis (a possible hybrid) that is commonly cultivated and naturalized here. Lampran- thus coccineus (Haw.) N. E. Br. is listed by Munz (1974) as natural- ized in coastal environments in southern California. We have not lo- 83 FERREN ET AL.: MALEPHORA CROCEA 1981] snouosti} A[MOIIeU uae13 y}Oous ysnol (1-4) $ atnsdea pel jenba ‘¢ sutpeeids ‘ayeryjsoid 3-7 94919} uaa13 aye [1ded sapd1eqn} pue Sql yyM Yysnol al atnsdeo pei-asueio yYsIIq jenbaun ‘9—-¢ Aqqnays ‘4a.19 u01}9aS sa snouosli} snouosi4} Aun[q Ajdzeys snouost4 Apun{q SsOI9 uda13 -snoone|s jeay uveis yep 0} Uda13 an[q-snooneys IO[Oo yyoouls qYJOOUIS yJOouS goeyans Sapd1eqn} jo aoejins y}oows AyIvau qyoows SMOJ YIM Ysnol p2aS : OI-8 (OT) 6-8 sa[NIO[ ajnsdeo dyl[-Al19q atnsdes 3dA} wns A adind eyuaseu 0} yuId eyUIseUL 1o/pue a3ueio IO[OO jenboaun ‘¢ jenbaun ‘¢ jenbsun ‘9-4 sjedas IIMO[ A sutpeaids sutpeaids sutpeaids qe ‘aqyerjsoid ‘ayersoid ‘ayer}so01d ‘qeH 2°q D9 2 ‘§NaUu19909 SnYIUDACUDT = °9 “JT ‘wnso19ads wnmayjUDSOAg = ‘S ‘q ‘wnyofisspAa DUAYdsIg = ‘I ‘ ‘snaaqD]INDaD SNJOAGOGAD) = ‘D2 ‘syentut Aq poyeusisap are saieds “SLNVIg AO] YAHLO YNOY WOAA 09I049 DAOYGa]V WW HSINONILSIC LVHL SOLLSTAALOVAVH) ‘T dTav = 84 MADRONO [Vol. 28 cated any naturalized L. coccineus nor have we seen any specimens of M. crocea bearing that name. However, some specimens of M. crocea collected from plants cultivated in California were misidentified as Lampranthus spectabilis (Haw.) N. E. Br., a species commonly cultivated but not known to be naturalized in western North America. Because of the confusion of Malephora crocea with other ice plants, we include a description and illustration (Fig. 1), and provide a com- parison (Table 1) with other species. MALEPHORA CROCEA (Jacq.) Schwant., Moller’s Deutsch. Gartn.-Zei- tung 43:7. 1928.—Mesembryanthemum croceum Jacq., Fragm. Bot. 17. 1800.—Mesembryanthemum insititum Willd., Enum. Hort. Berol. 536. 1809.—Hymenocyclus croceus (Jacq.) Schwant., Moller’s Deutsch. Gartn.-Zeitung 42:27. 1927.—Crocanthus cro- ceus (Jacq.) L. Bol., Fl. Pl. S. Africa 7:255. 1927. Decumbent or prostrate shrub with pale, corky branches, occasion- ally rooting at the nodes, forming dense mats to 3 dm high, with stout, gnarled, woody stems in maturity; leaves crowded on short shoots, opposite, connate at base, erect, 2.5—6 cm long, 5-8 mm wide, bluntly trigonous, succulent, smooth, pale bluish-green and glaucous, occa- sionally reddish; flowers solitary, terminal or axillary, ebracteate, on pedicels 1-6 cm long; calyx 0.8—1.5 cm wide, the lobes 4—6, unequal, at least 2 short, acuminate, with hyaline margins; petals and stami- nodia usually orange adaxially and purple abaxially; stamens numer- ous; stigmas 8—9 (—10), plumose; ovaries obcuneiform; placentation parietal; capsule 8—9 (—10) locular, with cell lids, valve wings, adaxial seed pockets with bifid placental tubercles; seeds numerous, lenticular, 1 mm long, 0.8 mm wide, with tubercles arranged in rows (Fig. 1). Two varieties of M. crocea have been designated: var. crocea with petals orange adaxially and purple abaxially, and var. purpureocrocea (Haw.) Jacobs. and Schwant. with petals purple on both surfaces. Because there is considerable color variation among the populations examined in California, we do not distinguish color forms. Confusion over the identity of ice plants, as discovered during the investigation of M. crocea, suggests that additional taxonomic and nomenclatural problems may exist for other members of the Aizoaceae reported to be naturalized in California. We are continuing to survey the group to contribute to the correct identification of these plants. Because much of the classification of ice plants is based upon fruit characteristics, it would be helpful if collectors prepared herbarium specimens with mature fruits in addition to ample vegetative material, and flowers. LIST OF SPECIMENS More than 500 herbarium specimens were examined during this study. Approximately 40 specimens of M. crocea have been located 1981] FERREN ET AL.: MALEPHORA CROCEA 85 among these and with field observations serve as the basis for this paper. A list of the latter specimens has been distributed to herbaria cited below. Additional copies are available on request from the au- thors. ACKNOWLEDGMENTS We thank curators at CAS, CDA, DS, JEPS, LA, LAM, POM, RSA, SBBG, SBM, SD, UC, UCR, and UCSB for loans; Sherry Whitmore (UCSB) for acquisition of plant material; Bob Haller, Dale Smith, Barry Tanowitz (UCSB) and Reid Moran (SD) for comments on the manuscript; Maggie Day (UCSB) for the illustration; and, especially, Walter Wisura (RSA) for verification of our determinations and helpful insight on the ice plants. LITERATURE CITED ANONYMOUS. 1954. Colorful noon-day flowers from South Africa. Cact. Succ. J. (Los Angeles) 26:149-152. ANONYMOUS. 1969. Ice plant. Sunset Magazine. May:218-222. ANONYMOUS. 1979. Sunset Western Garden Book, ed. 4. Lane Publishing Co., Menlo Park, CA. HERRE, H. and O. K. VOLK. 1947. Mesembryanthemaceae familia nova. In: G. Schwantes. System der Mesembryanthemaceen. Sukkulentenkunde I:34—46. HERRE, H. 1971. The genera of the Mesembryanthemaceae. Tafelberg-Vitgewers Be- perk, Capetown. HOFFMAN, R., E. O. ORPET, E. WALTHER, and J. WEST. 1930. Cacti and other succulents: an annotated list of plants cultivated in the Santa Barbara region. Garden Tours Committee, Community Arts Association, Santa Barbara. HowELL, J. T., P. H. RAVEN, and P. RUBTZOFF. 1958. A flora of San Francisco. Wasmann J. Biol. 16:1—157. JACOBSEN, H. 1960. A handbook of succulent plants, vol. II]. Mesembryanthemums (Ficoidaceae). Blandford Press, London. . 1977. Lexicon of succulent plants, ed. 2. Blandford Press, Poole, England. KARTESZ, J. T. and R. KARTESZ. 1980. A synonymized checklist of the vascular flora of the United States, Canada, and Greenland. The Biota of North America, vol. 2. Univ. North Carolina Press, Chapel Hill. KIMNACH, M. 1966. Groundcover succulents of California. LASCA Leaves 16:31-48. Moran, R. 1950. Mesembryanthemum in California. Madrono 10:161—163. MELCHIOR, H. 1964. Syllabus der Pflanzenfamilien, ed. 12, vol. Il. Angiospermen. Gebruder Borntrager, Berlin-Nikolassee. Munz, P. A. 1959. A California flora. Univ. California Press, Berkeley. . 1968. Supplement to a California flora. Univ. California Press, Berkeley. . 1974. A flora of southern California. Univ. California Press, Berkeley. POINDEXTER, R. W. 1934. Some remarks on Mesembryanthemums. Cact. Succ. J. (Los Angeles) 5:533-536, 553-555, 569-571. SCHWANTES, G. 1971. The classification of the Mesembryanthemaceae. Revised and completed by H. Straka and H. D. Ihlenfeldt. 7x: H. Herre. 1971. The genera of the Mesembryanthemaceae. Tafelberg-Vitgewers, Capetown. SHETLER, S. G. and L. E. SkoG, eds. 1978. A provisional checklist of species for Flora North America. Monographs in Systematic Botany, I. Missouri Bot. Gard., St. Louis. SMITH, C. F. 1976. A flora of the Santa Barbara Region. Santa Barbara Museum of Natural History, Santa Barbara. (Received 16 May 1980; revision received and accepted 26 Nov 1980.) 86 MADRONO [Vol. 28 NOTEWORTHY COLLECTIONS COCHLEARIA OFFICINALIS L. (BRASSICACEAE).—USA, CA, Del Norte Co., Crescent City, 0.1 hectare basaltic sea stack, 1 km offshore from the intersection of Pebble Beach Dr. and Pacific Ave., 20 Jun 1980, Lester 4829 (HSC). Scattered plants on ne. slope, in crevices of rock in accumulated guano, spread over a 4 X 10-m area. Associated with Lasthenia minor subsp. maritima. Previous knowledge. Holarctic, known in Pac. Northw. from AK, B.C., WA, and OR coasts. Alaskan populations are commonly associated with seabird colonies. (Her- baria consulted: CAS, HSC, JEPS and UC (fide Alice Howard), OSC (fide Kenton Chambers), RSA (fide Robert Thorne), and WS (fide Joy Mastrogiuseppe); published sources: Munz, A Calif. fl. 1959; Munz, Suppl. Calif. fl. 1968; Hitchcock et al., Vasc. pls. Pacific Northw. 2. 1964.) Significance. First record for CA, a 55-km range extension from Cape Sebastian, OR. Locality was breeding ground for Western Gulls, Black Oystercatchers, and Pigeon Guillemots at time of collection. Should be sought on other northcoast offshore rocks that are frequented by seabirds.—GAry S. LESTER, U.S. Fish and Wildlife Service, 791 8th Street, Arcata, CA 95521, MICHAEL C. VASEY, Department of Biology, San Franscisco State University, San Francisco, CA 94132, and WILLIAM E. RODSTROM, P.O. Box 4286, Arcata, CA 95521. (Received and accepted 16 Oct 1980.) PEDICULARIS CRENULATA Benth, f. CANDIDA Macbr. (SCROPHULARIACEAE).—USA, CA, Mono Co., Sierra Nevada Aquatic Research Laboratory along n. side of Convict Cr., 1.5 km w. of hwy 395 (T4S R28E S12 se. 4%), 2160 m: 9 Aug 1978, Howald 952; 12 Aug 1978, Orr 337 (UCSB). A small colony (130 plants counted 30 Aug 1980) in moist meadow soil with Mimulus primuloides, Agoseris glauca, Penstemon oreocharis, Cirsium congdonii (C. drummondii), Stellaria longipes, Trifolium longipes, Gentiana holopetala, Parnassia palustris and Juncus macrandrus. Plants were mainly in fruit during the 1980 census. Previous knowledge. Species known from Mono Co. e. to s. WY and CO; the form always rare but scattered throughout range of species. (Herbaria consulted: UC and JEPS kindly checked by Alice Q. Howard, UCSB; published sources: Major and Bam- berg, Madrono 17:93—109. 1963); Abrams, Illus. fl. Pac. States. 3. 1951; Munz, A Calif. fl. 1959; MacBride, Contr. Gray Herb. 56:61. 1918; Sprague, Aliso 5:181—209. 1962.) Significance. Only collections of the sole CA population since 30 Jul 1933 (Peirson s.n., UC). Listed as “presumed extinct” in CA (Smith et al., Inv. rare endang. vasc. pls. Calif. CNPS Spec. Publ. 1, ed. 2. 1980). 400 km w. of the nearest population (Duck Creek, Schell Creek Range, NV). All individuals observed in 1978, 1979, and 1980 had white corollas (f. candida), although the typical corolla color is purple. Sprague noted in 1962 (but did not collect) a population of 25-30 white-flowered plants in the same area. The population lies completely within the fenced boundaries of the Sierra Nevada Aquatic Research Laboratory, a unit of the Univ. of Calif. Natural Land and Water Reserves System, and may be the only suitable habitat in the area currently protected from cattle grazing.—ANN M. Howa Lp, 419 Ellwood Beach Drive, Apt. 5, Goleta, CA 93117 and BRuCE K. Orr, Department of Biological Sciences, University of Cali- fornia, Santa Barbara 93106. (Received 29 Aug 1980; accepted 3 Sep 1980; final version received 24 Oct 1980.) DEDECKERA EUREKENSIS Reveal & Howell (POLYGONACEAE).—USA, CA, Inyo Co., White Mts., 2.4 km ne. of E. Line St., Bishop, between Silver Canyon and Poleta 1981] NOTEWORTHY COLLECTIONS 87 Canyon (T6S R33E S6 nw.'4 nw.%), 1460 m, 24 Jun 1980, Strohm 560 (UC and private collection). Extremely local on a dry gravelly n.-facing slope in Shadscale Scrub. As- sociated species include Encelia virginensis subsp. actonii, Dalea fremontii, Atri- plex confertifolia, and Petalonyx nitidus. Verified by Mary DeDecker. Previous knowledge. Known from the Last Chance Mts. and the Inyo Mts. of Inyo Co., CA. (M. DeDecker, pers. comm., 1980; Reveal and Howell, Brittonia 28:245-251. 1976). Significance. First record for the w. side of White-Inyo range and n.-most location, a disjunction of 72 km. Most of the plants were growing on the slope; however, some were found in the wash. Considered “rare and not endangered” (Smith et al., Inv. rare endang. vasc. pls. Calif., CNPS Spec. Publ. 1, ed. 2. 1980).—PatTTI J. Novak and KATHRYN L. STROHM, Inyo National Forest, Bishop, CA 93514. (Received 18 Aug 1980; accepted 9 Sep 1980.) IPOMOEA EGREGIA House (CONVOLVULACEAE).—USA, NM, Grant Co., ca. 11 km nw. of Silver City: T17S R15W center of line between S10 and S11, 2000 m, 6 Sep 1980, Spellenberg et al. 5864 (NMC). In a wnw.-draining canyon, pinyon-juniper zone, growing on pale, rather barren bedrock outcrop. Very local, ca. 50 plants. Associates included the closely related J. plummerae (5863), the rather similar, but annual, /. costellata, and the more distantly related J. hirsutula among other annuals and hemi- cryptogams; T17S R1I5W S11 w-c.%, 2000+ m, 10 Sep 1980, Fletcher 4907 (US For. Serv., Albuquerque), 3 plants. Previous knowledge. Three collections from se. AZ (two in the Huachuca Mts., sw. Cochise Co., the type locality, and one in the Santa Rita Mts., se. Pima Co.), and a fourth from Peru (cited in Macbride, Publ. Field Mus. Nat. Hist. 288. 1931). (Herbaria consulted: ARIZ, ASC, MO, NMC, UNM, Western NM Univ.; published sources: House, Torreya 6:124. 1906; Wooton and Standley, Fl. New Mex. Contr. U.S. Natl. Herb. 19. 1915; Macbride, op. cit.; Kearney and Peebles, Ariz. fl. 1951; Martin and Castetter, Checklist gymnosp. angiosp. New Mex. 1970. Significance. First record from NM, ane. range extension of 225 km. Our collection provides further evidence that J. egregia is a variant of /. plummerae, as first indicated by Macbride, who proposed the former be called J. plummerae var. cuneifolia. The two taxa are morphologically similar except in leaf form. /. plummerae, widespread in AZ, extends to sw. NM and Sonora (Kearney & Peebles, op. cit; McDougal, Seed pls. N. Ariz. 1973), and Peru (Macbride, op. cit.). At all stations known for I. egregia, including the disjunct Peruvian site, J. plummerae also occurs, but the converse does not hold. Spellenberg et al. 5863 has six specimens intact with tubers; some of these are I. plummerae, others are I. capillacea G. Don (I. muricata Cav.) as delimited in Kearney and Peebles (op. cit.). J. capillacea is said to have elongate tubers, sepals, 5— 6 mm long, and the length of the peduncle + pedicel about equal to that of the calyx. I. plummerae has globose tubers, sepals 7—9 mm long, and the length of the peduncle + pedicel up to twice the length of the calyx. In 5863 the tubers range from elongate (1.5:1, length: width) to globose (1:1), the sepals range from 6—10 mm in length, and the ratio peduncle + pedicel/calyx ranges from 1:1—1.76:1 (within-plant averages). In ad- dition, in our equally small sample of J. egregia, the five tubers ranged from nearly globose to elongate (2.5:1). George Yatskievych, who is working on the /. egregia-l. blummerae problem at ARIZ, was most helpful in researching AZ locations for this note and providing comments. Warren L. Wagner assisted us at MO. STELLARIA NITENS Nutt. (CARYOPHYLLACEAE).—Same location and date as above, Spellenberg et al. 5869 (NMC, NY). Frequency not recorded, but collection comprises 88 MADRONO [Vol. 28 21 individuals from ca. 0.5 m’. Associates included Drymaria fendleri, D. sperguloides, Bidens leptocephala, Bulbostylis funckii, Euphorbia bilobata, and Tagetes micrantha. Previous knowledge. B.C. s. to Baja Calif., e. to MT, s. through UT to c. and se. AZ. (Herbaria consulted: ARIZ, ASC, MO, NMC, UNM, Western NM Univ.; pub- lished sources; Hitchcock et al., Vasc. pls. Pacific Northw. 2. 1964; Kearney and Peebles, op. cit.; Martin and Castetter, op. cit). Significance. First record for NM, a ne. range extension of 110 km from the Chi- ricahua Mts. of AZ. This inconspicuous plant probably occurs elsewhere in NM.—RosB J. SORENG AND RICHARD SPELLENBERG, Biology Department, New Mexico State Uni- versity, Las Cruces 88003. (Received and accepted 13 Nov 1980; final version received 3 Dec 1980.) ERIGERON HUMILIS Graham (ASTERACEAE).—USA, ID: Lemhi Co., Lemhi Range, Challis N.F., moist alpine tundra on n. slope Bell Mt., 35° nw. slope on quartzite, 3400 m, 2 Aug 1978, Henderson et al. 4880 (ID, NY); Custer Co., Lost River Range, Challis N.F., moist alpine tundra on n. slope Leatherman Pass, on limestone, 3300 m, 27 Jul 1979, Brunsfeld and Brunsfeld 1235 (ID); Butte Co., s. Lost River Range, Challis N.F., moist alpine tundra at head of Elbow Canyon on limestone, 3250 m, 30 Jul 1979, Brunsfeld and Brunsfeld 1260 (ID). Three additional stations in the Lost River Range have been discovered by the authors, all in habitats similar to those described above. Plants are common in each population but often highly local. Poa alpina and P. rupicola are common associates on both limestone and quartzite substrates. Populations found on quartzite substrates also are commonly accompanied by Geum rossii var. turbina- tum. Full flower by late Jul. Verified by A. Cronquist (Henderson et al. 4880), 1978. Previous knowledge. Circumpolar but ranging s. in N.A. tos. B.C., n. WY, and nw. MT. (Herbaria consulted: CIC, ID, IDS, MONTU, NY, ORE, OSC, UTC, WS, WTU; published sources: Hitchcock et al., Vasc. pls. Pacific Northw. 5. 1955; Hitchcock and Cronquist, Fl. Pacific Northw. 1973; Dorn, Man. Vasc. pls. Wyo. 1977.) Significance. First record for ID, an extension s. of 500 km. Although not in jeop- ardy, it is listed as Rare by the Tech. Comm. on Rare and Endangered Pls., Idaho Natural Areas Council. HYMENOPAPPUS FILIFOLIUS Hook. var. IDAHOENSIS Turner (ASTERACEAE).—USA, ID: Custer Co., 12 km se. of Challis, Lost River Range, dry canyon bottom, Lime Cr. drainage, on volcanic ash, 2130 m, 18 Jun 1979, Brunsfeld and Brunsfeld 1010 (ID); Lemhi Co., steep w. slope on e. side Salmon R., 1.6 km ne. of mouth of McKim Cr. on volcanic ash, 1500 m, 14 Jun 1978, Henderson et al. 4446 (ID); Clark Co., dry, rocky soil near base of Reno Point, s. end Beaverhead Range on limestone, 1760 m, 9 Jul 1975, Henderson and Jewell 2608 (ID); Custer Co., dry, gravelly sw. slope above road along Challis Cr. at Challis N.F. boundary, 300 m e. of mouth of Pats Cr. on volcanic substrate, 1680 m, 13 Jun 1978, Henderson et al. 4413 (ID); Butte Co., Lost River Range, open ridge top, Bird Canyon Road 20 km e. of Mackay, on limestone, 2420 m, 1 Aug 1979, Brunsfeld and Brunsfeld 1301 (ID). Plants in each population generally abundant and associated with Artemisia tridentata and often with Atriplex confertifolia. Previous knowledge. This var. known only from the Salmon and Lemhi valleys of Custer and Lemhi cos., ID; a well-marked endemic of e.-c. ID. (Herbaria consulted: ID, IDF, IDS, NY, ORE, OSC, UTC, WS, WTU; published sources: Hitchcock et al., Vasc. pls. Pacific Northw. 5. 1955; Hitchcock and Cronquist, Fl. Pacific Northw. 1973.) Significance. These collections, 12 additional by the authors, and 16 by Andersen and Davies (ID), have established that this taxon is neither rare nor in jeopardy. Hab- itats in which the plants are most abundant are severely disturbed by grazing or other 1981] NOTEWORTHY COLLECTIONS 89 factors. Current land use appears to favor this var. We consider its placement in the Federal Register (1975) as a proposed threatened taxon and similar status offered by Ayensu and Defilipps (Endang. threat pls. U.S. 1978) are unwarrantable. CAREX RUPESTRIS All. (CYPERACEAE).—USA, ID, Lemhi Co., crest of Lemhi Range, Challis N.F.: dry alpine grassland on quartzite at head of Bruce Canyon, 3000 m, 30 Jun 1977, Brunsfeld and Brunsfeld 323 (ID, NY); dry rocky limestone outcrop 0.25 km s. of Trail Peak summit, 3200 m, 2 Jul 1977, Brunsfeld and Brunsfeld 375 (ID). Plants of both populations uncommon with Poa rupicola, Trisetum spicatum, Carex elynoides, and Eritrichium nanum. Verified by A. Cronquist (323), 1978. Previous knowledge. Circumboreal but extending s. in USA in Rocky Mts. to the Uinta Mts. of UT. (Herbaria consulted: BOIS, BS, ID, IDF, MONTU, NY, ORE, OSC, WS, WTU; published sources: Hitchcock et al., Vasc. pls. Pacific Northw. 1. 1969; Hitchcock and Cronquist, Fl. Pacific Northw. 1973; Cronquist et al., Interm. fl. 6. 1977; Lackschewitz, Madrono 23:362. 1976.) Significance. First records from ID. A w. extension in the USA of ca. 80 km. Although apparently in no jeopardy within ID, it is listed as Rare by the Tech. Comm. on Rare and Endangered Pls., Idaho Natural Areas Council. ASTRAGALUS AMNIS-AMISSI Barneby (FABACEAE).—USA, ID: Butte Co.: base of limestone cliffs, Middle (Bartell) Canyon, sw. end of Lemhi Range, Challis N.F., 2040 m, 18 Jul 1978, Henderson et al. 4628 (ID); base of limestone cliffs, East Canyon, sw. end of Lemhi Range, Challis N.F., 1830 m, 5 Jun 1978, Henderson et al. 4211 (ID); Custer Co.: 6.4 km ne. of Mackay, Lost River Range, base of limestone cliffs, Lower Cedar Cr. Canyon, Challis N.F., 2160 m, 14 Jun 1979, Brunsfeld and Brunsfeld 956 (ID). Plants uncommon in all locations and nearly confined to stable limestone talus at base of cliffs or ledges and cracks above talus, often in partial shade of Pseudotsuga menziesii or Cercocarpus ledifolius. Associated closely with Draba oreibata. Flowers early to mid-June, fruits Jul. An earlier collection from the type locality (Henderson 3073, ID) was verified by C. L. Hitchcock. Previous knowledge. Known only from the type locality, Pass Creek Gorge, Custer- Butte Co. line, Lost River Range. Recent examinations of this population by the authors disclose 30-40 readily-observable plants near the base of cliffs, and numerous additional plants on ledges and in cracks of the near-vertical limestone near the upper part of gorge. (Herbaria consulted: BOIS, BS, ID, IDF, MONTU, NY, ORE, OSC, WS, WTU; published sources: Hitchcock et al., Vasc. pls. Pacific Northw. 3. 1961; Hitchcock and Cronquist, Fl. Pacific Northw. 1973.) Significance. Until 1978, considered one of Idaho’s narrowest endemics. Included on the Federal list as Proposed Endangered (Federal Register, Jun 1976), and is listed as endangered by Ayensu and DeFilipps (op. cit). Collections cited above, 13 additional by the authors, and 3 by Andersen and Davies (all ID), each representing a substantial population in a different canyon in the Lost River and Lemhi ranges, establish the overall abundance of this taxon. There appear to be no immediate or proposed threats to this habitat, and the land managing agencies (USFS and BLM) have no plans for development within these areas. Although reproduction appears to be low, plants of all age classes are present in all populations examined. We believe there is no longer justification for consideration of this taxon as Endangered. GENTIANA PROPINQUA Richards (GENTIANACEAE).—USA, ID, Custer Co., Lost Riv- er Range, Challis N.F., moist alpine meadows: n. slope Leatherman Pass on quartzite, 3120 m, 27 Jul 1979, Henderson et al. 5448 (ID); lake shore at head of e. Fk. Pahsimeroi River on limestone, 2900 m, 10 Aug 1979, Brunsfeld and Brunsfeld 1353 (ID, NY). Plants rare in both populations with Deschampsia caespitosa, Poa alpina, Gentiana tenella, Carex elynoides, and Carex subnigricans. Verified by staff at NY (Brunsfeld and Brunsfeld 1353), 1979. Previous knowledge. Known from AK e. across Canada to Newfoundland and s. 90 MADRONO [Vol. 28 in Rocky Mts. to Beaverhead Co., MT. (Herbaria consulted: BS, ID, IDF, MONTU, NY, ORE, OSC, WS, WTU; published sources: Hitchcock et al., Vasc. pls. Pac. Northw. 4. 1959; Hitchcock and Cronquist, F1. Pacific Northw. 1973; Scroggan, Fl. Can. 4. 1978.) Synonym: Gentianella propinqua (Richards) Gillett. Significance. First records for ID, an extension wsw. in USA of 300 km. Listed as Rare for ID by the Tech. Comm. on Rare and Endangered Pls., Idaho Natural Areas Council. PAPAVER KLUANENSIS D. Love (PAPAVERACEAE).—USA, ID, Lemhi Co., Lemhi Range, Challis N.F., summit of Bell Mt., 3740 m, Meadow Canyon Res. Nat. Area, 2 Aug 1978, Henderson et al. 4846 (ID, NY). Extremely rare. Population of 12-15 individuals from the summit to ca. 75 m down the n. and w. faces. No individuals were seen below this elevation or on other aspects. With Arabis lemmonii and Draba lon- chocarpa in full sun among broken quartzite rocks at the summit, and with Saxifraga debilis and Poa rupicola in partial shade of near-vertical crevices on n. and w. faces. Only a few plants flowering on w. face, those of the summit and n. face still in bud. Verified by B. Ertter (NY), April, 1979. Previous knowledge. Known from Yukon Terr. s. to NM, mainly along Continental Divide. Apparently restricted to high alpine summits and ridges. Not previously known from ID. (Herbaria consulted: ID, IDS, NY, ORE, OSC, UTC, WS, WTU; published sources: Love, Brittonia 21:1—10. 1969; Dorn, Man. vasc. pls. Wyo. 1977; listed as P. radicatum Rottb. in Rydberg, Fl. Rocky Mts. and adj. plains. 1922; Weber, Rocky Mt. fl. 1976; and Scroggan, Fl. Can. 3. 1978.) Diagnostic characters. Fls single, terminal, the petals yellow; fr with brown hairs; pls scapose and densely caespitose, lvs densely hairy; scapes erect at anthesis. Singificance. First record for the Pacific Northwest and for ID, a range extension w. of ca. 480 km. Concentrated floristic studies in this region over the past seven years failed to disclose any other populations. We consider this population to be endangered: although the Bell Mt. site is isolated, there is evidence of considerable foot traffic and the actions of a single, thoughtless individual could eliminate a significant portion of this small population. Listed as endangered for ID by the Tech. Comm. on Rare and Endangered Pls., Idaho Natural Areas Council. Funding was provided in part by the Challis National Forest and by the C. R. Stillinger Trust, Univ. of Idaho.—DouGLass M. HENDERSON, STEVEN BRUNSFELD, AND PAMELA BRUNSFELD, University of Idaho Herbarium, Department of Biological Sciences, University of Idaho, Moscow 83843. (Received 15 May 1980; accepted 7 Jul 1980; final version received 29 Oct 1980.) ANNOUNCEMENT CALIFORNIA BOTANICAL SOCIETY—GRADUATE STUDENT MEETINGS The California Botanical Society Graduate Student Meetings will be held at San Francisco State Univ., 24-25 October 1981. The meeting will focus on the presenta- tion of short research papers and reports in progress by graduate students in all botanical and plant related fields. Members and non-members are invited to partici- pate. For further information please contact the Graduate Student Meetings Committee, Dept. of Biology, San Francisco State Univ., San Francisco 94132 or leave a message at (415) 469-1359. Dr. Harry D. Thiers will present a seminar Saturday evening on his recent work in Australia and the interesting fungal flora of that area. 1981] 91 NOTES AND NEWS NOTES ON CONES AND VERTEBRATE-MEDIATED SEED DISPERSAL OF Pinus albicaulis (PINACEAE).—The literature is unclear as to the events following ripening of Pinus albicaulis cones. Sudworth (For. trees Pacific slope. 1908) wrote that cones mature in late August or early September and release their seeds in September or October. “The cones dry out and open slowly. . . .” This information is contradicted by Shaw’s classification (The genus Pinus. 1914) of P. albicaulis in Cembrae, a group characterized by inde- hiscent cones. Subsequent authors acknowledged the indehiscence of P. albicaulis cones but were confused about dispersal mechansims (e.g., Bowers, Cone-bearing trees Pac. Coast. 1956; Peattie, Nat. hist. w. trees. 1953). Krugman and Jenkinson, who had access to experimental data, suggested that “seeds are dispersed when the detached cone disintegrates” (In: Schopmeyer, ed., Seeds woody pls. U.S. 1974). However, informa- tion on cone disintegration and subsequent seed dispersal comes mainly from experi- ments in which cones were protected and does not elucidate the true fate of cones and seeds. I have gathered cone-fate information from 1973 to 1979 in various locations in Inyo, Mono, Mariposa, and Tuolumne Counties in the eastern Sierra Nevada, California. When P. albicaulis cones ripen, the scales separate slightly from the core of the cone (thus, the cones are not completely indehiscent); yet, the scales hold the seeds firmly in place and are not easily dislodged. Cones I collected in 1975 still retain their seeds. The large, wingless seeds are an attractive food for many birds and rodents. Clark’s Nutcracker (Nucifraga columbiana) is sympatric with P. albicaulis wherever the pine occurs. From midsummer until the cones ripen, nutcrackers forage preferentially on unripe seeds of P. albicaulis. Nutcrackers harvest seeds by jabbing their bills repeatedly into the top or sides of cones to loosen and tear off scales. Only rarely do they first detach a cone from a tree before removing seeds. Ripe seeds are taken in quantity and stored in small “caches” consisting of 1 to 15 seeds each (median = 4) in selected “storage slopes” and throughout the forest terrain. I calculated that each nutcracker may store as many as 32,000 P. albicaulis seeds each year at subalpine elevations. These seed caches are retrived by nutcrackers in spring and early summer (Tomback, Ph.D. diss., Univ. California, Santa Barbara. 1977a; Tomback, Living Bird 16:123-161. 1977b; Tomback, Condor 82:10—19. 1980). However, many of the seeds are placed in micro- habitats favorable to germination and seedling survival. By mid-October, some seeds have been removed from many P. albicaulis cones by Clark’s Nutcrackers. These cones are partly or completely hollowed out with little of the cone intact and a shell of closed scales on the underside of the cone (Tomback, 1977a, 1977b, op. cit.). My studies of Clark’s Nutcracker suggest that this bird is an important disperser of P. albicaulis seeds. Several kinds of circumstantial evidence, such as seedlings origi- nating in nutcracker seed caches, seedling clusters from caches producing a “multi- trunked” growth form (Clausen, Evolution 19:56—68. 1965; Lueck, M.A. thesis, Oregon State Univ. 1980), the sites selected by nutcrackers for seed caches, and a consideration of alternative dispersal mechanisms, support the hypothesis that the interaction between the nutcracker and P. albicaulis is mutualistic and coevolved. A similar interaction was proposed for the nutcracker and P. edulis (Vander Wall and Balda, Ecol. Monogr. 47:89-111. 1977). More direct evidence is required to substantiate both of the proposed interactions. During late summer and early fall, Douglas squirrels (Tamiasciurus douglast) cut down numbers of P. albicaulis cones and bury them in middens. Chipmunks (Eutamias spp.) climb into trees and gnaw on the cones to extract seeds. A chipmunk-foraged cone has a distinct appearance: all but the proximal and distal scales are gone, leaving only the core. Chipmunks as well as deer mice (Peromyscus maniculatus) are known to cache germinable pine seeds (West, Ecology 49:1009-1011. 1968; Abbott and Quink, Ecology 51:271-278. 1970). Other mammals and birds take P. albicaulis seeds when cones are ripe (Tomback, 1977a, 1977b, op. cit.): e.g., golden-mantled ground squirrel (Spermophilus lateralis), 92 MADRONO [Vol. 28 Williamson’s Sapsucker (Sphyrapicus villosus), White-headed Woodpecker (Picoides albolarvatus), Mountain Chickadee (Parus gambeli), White-breasted Nuthatch (Sitta carolinensis), Cassin’s Finch (Carpodacus cassinii), Red Crossbill (Loxia curvirostra), and Pine Grosbeak (Pinicola enucleator). With such diverse foraging on P. albicaulis cones by vertebrates it is likely that few seed-bearing cones, if any, remain on trees by late fall. Vertebrate foraging has also been observed to “destroy” an entire seed crop of Pinus lambertiana (Tevis, J. Wildlife Managem. 17:128-131. 1953) and P. flexilis (Clements, Pl. succession. 1916). To test the hypothesis that most P. albicaulis seed dissemination is effected by ver- tebrates, particularly Clark’s Nutcracker, it is necessary to gather information on the condition of old and new cones encountered in the field. Here, I present the results of a preliminary study. On 9 and 10 August 1979, I surveyed P. albicaulis on the east and west slopes of Cathedral Peak, Yosemite National Park, Tuolumne and Mariposa Counties, California. On the east slope I observed the occurrence of old cones in a pure stand between 3050 m and 3250 m (25° slope). For each of the 36 trees along a line transect, I noted the number of old cones visible from my position and tree growth form (following Clausen, 1965, op. cit.). The number of old cones per tree ranged from 0 to 30 fk = 4.9 + 1.5 s.e.]. Proportions of trees encountered in each growth form category are as follows: erect trunk: 0.44; elfinwood: 0.44; and intermediate: 0.11. Fifty percent of the trees had no old cones. One old cone had been nutcracker-foraged, and the remainder were cores, such as those left by chipmunks. A linear regression analysis of number of old cones per tree vs. elevation indicated a significant negative correlation (r = —0.641, p< 0.001), which is probably the consequence of lower cone productivity of the krummholz or “elfinwood” growth form (Tranquillini, Physiol. ecol. alpine timberline. 1979). A One-sample Runs test (Siegel, Nonparametric statistics. 1956) confirmed a significantly greater occurrence of the elfinwood growth form as elevation increased (p = 0.025, one-tailed). On 8 September 1979, I surveyed the east slope of Cathedral Peak for new cones, repeating the same transect. Only 7 trees bore new cones (none of the elfinwood), and all new cones were partly or completely harvested by nutcrackers and chipmunks. The transect on the steeper (30°) west slope of Cathedral Peak zigzagged over a strip ca. 50 m wide from about 3050 m to 3200 m. Predominantly erect P. albicaulis are scattered over the lower regions of the slope, whereas dense elfinwood mats occur in talus above ca. 3200 m. At the very top of the slope where there is some shelter, trees intermediate in growth form occur. For the 44 trees on the transect, I noted the following frequencies of cone occurrence: no cones: 0.43; old cones: 0.45; new cones: 0.39; old and new cones: 0.27; vertebrate-foraged new cones: 0.23. Thus, by 10 August, verte- brates had opened 59 percent of the new cones. All the old cones observed were either cores or hollow shells. During September, 1979, I surveyed P. albicaulis in other areas to assess the con- dition of new cones. On 10 September, trees were examined in the vicinity of Budd Lake, Yosemite National Park, Tuolumne Co., California, 3050 m (Table 1). Nutcrack- ers had foraged in 75 percent of the new cones observed. Fifteen trees on each of three different slope aspects were surveyed on 15 September on the west slope of Mammoth Mountain, Mono Co., California (Table 1). Altogether, 78 percent of the new cones were completely or partly destroyed by nutcrackers and chipmunks. (From the data reported, it is apparent that the 1979 P. albicaulis cone crop was only fair to moderate in the central Sierra Nevada.) During 6 to 10 August and 1 to 3 September 1979, I examined 28 old cones found on the forest floor in both the Cathedral Peak and Kearsarge trail-Kearsarge Lakes areas, Mono and Fresno Counties, California. The cones were categorized as follows: chipmunk cones—11%, disintegrating cones (loose scales)—21% nutcracker-harvested cones— 21%, small, closed cones (ca. 5 cm diameter)—46%. Intact old cones on the ground were rare in all study areas. I pried open the closed cone scales of the small cones encountered and found only white-coated empty seeds. Thus, vertebrate-foraged cones 1981] NOTES AND NEWS 93 TABLE 1. MEAN NUMBER PER TREE AND STANDARD ERROR OF NEW AND VER- TEBRATE-FORAGED CONES. Total Vertebrate- trees New cones foraged Date Location surveyed per tree cones per tree 10 Sep 1979 Budd Lake 10 2.1 + 0.46 1.6 + 0.54 15 Sep 1979 Mammoth Mt. 45 2.0 + 0.47 1.6 = 041 accounted for 32 percent of the cones sampled, and sterile cones accounted for 46 percent. These data suggest that few if any seeds of P. albicaulis remain after vertebrates cease foraging in late fall. It is possible that any remaining cones with a full or partial complement of viable seeds disintegrate and/or abscise soon after this time. Abscission may, in part, be a weight-dependent process, which would explain why only the lighter, vertebrate-harvested old cones remain on the trees. Also, some cones seem more resis- tant to disintegration, such as those hollowed out by nutcrackers and small, sterile cones. If so few cones escape vertebrates each year, is cone disintegration the primary seed dispersal mechanism for P. albicaulis? Seed-storing vertebrates, particularly Clark’s Nutcracker, appear mainly responsible for P. albicaulis population recruitment. Not only do nutcrackers and some rodents place a large percent of seeds in sites favorable to germination and seedling survival, nutcrackers transport seeds to storage sites some distance from parent trees (Tomback, 1977a, 1977b, op. cit.). Consequently, nutcracker seed dispersal helps maintain the “pioneering” status of P. albicaulis. Seeds released by cone disintegration have lower reproductive value than those stored by nutcrackers and rodents. Because the seeds are large and wingless, many will drop near the parent tree. Thus concentrated, they may be consumed in quantity by seed predators and, because P. albicaulis is shade intolerant (Baker, Principles silviculture. 1950), they are less likely to end up in conditions favoring seedling survival. In addition, “pioneering” may be an important reproductive tactic of the species. Why are P. albicaulis cones, as well as cones of other species in Subsection Cembrae (classification of Little and Critchfield, Subdiv. genus Pinus, U.S.D.A. Misc. Publ. 1144. 1969) indehiscent? All Cembrae species are sympatric with one or more subspecies of the Eurasian Nutcracker (N. caryocatactes) (Dement’ev, Birds Soviet Union. 5. 1970) or with Clark’s Nutcracker. In a nutcracker or rodent dispersal system, natural selection should optimize the number of seeds available to dispersal agents. “Packaged” seeds should attract birds and rodents, increase their foraging efficiency, and thereby maxi- mize the number of seeds stored in favorable sites. Consequently, indehiscent cones should maximize the reproductive success of any pine with vertebrate-dispersed rather than wind-dispersed seeds. It is interesting to note that the seeds of Pinus sabiniana and P. torreyana (Subsection Sabinianae, Little and Critchfield, 1969, op. cit.) are large and bear non-functional wings, although the cones are dehiscent. The seeds are released over a period of several months (Krugman and Jenkinson, 1974, op. cit.). These pines may be evolving verte- brate seed dispersal. Either they do not have the genetic potential for indehiscent cones, or the seeds may be dispersed by a facultative tactic: seed-storing vertebrates—such as Peromyscus (McCabe and Blanchard, Three species Peromyscus. 1950) and Scrub Jays (Aphelocoma coerulescens), which cannot take seeds from the massive, closed cones— and seed fall. This also applies to P. flexilis (Subsection Strobi, Little and Critchfield, 1969, op. cit.), which has dehiscent cones. Nutcrackers and rodents, as well as seed fall, may disperse the large, wingless seeds of this species (Vander Wall and Balda, 1977, op. cit.; Tomback, Condor 82, in press. 1980). 94 MADRONO [Vol. 28 Field work was supported by a cooperative aid agreement between the U.S.D.A. Forest Service, Pacific Southwest Forest and Range Experiment Station and the Uni- versity of California at Riverside. I thank Kathryn A. Kramer for her help in the field and Jan van Wagtendonk for coordinating my work in Yosemite National Park. W. B. Critchfield, J. R. Griffin, J. C. Hickman, T. R. Plumb, and R. J. Vogl reviewed the manuscript and provided helpful comments.—DIANA F. TOMBACK, Department of Zoology and Entomology, Colorado State University, Fort Collins 80523. (Received 2 Jan 1980; revision accepted 5 Sep 1980.) AGGREGATION OF Prunus ilicifolia (ROSACEAE) DURING DISPERSAL AND ITS EFFECT ON SURVIVAL AND GROWTH.—Dispersing seeds are commonly aggregated at settlement, by vertebrates voiding cohorts of ingested seeds or storing seeds in caches, or by ants collecting and discarding seeds with elaiosomes. Aggregation may have significant con- sequences for seed and seedling survival and growth. For buried seed, aggregation may increase successful emergence (Linhart, J. Ecol. 64:375-380. 1976.). Seed predation may increase with cohort size (Wilson and Janzen, Ecology 53:954—959. 1972.) or de- crease as aggregation lowers seed density over most of the dispersal region. Likewise, later herbivory might be increased or decreased. However, competition among aggre- gated seedlings surely must be greater than among widely scattered plants. None of these effects has been widely studied, perhaps for technical reasons or for lack of data on post-dispersal seed distributions. In the California chaparral, seeds of Prunus ilicifolia are commonly dispersed by Canis latrans, with defecated and vomited cohorts containing 4—66 seeds (mean 23.5 + 2.4 s.e., n = 34) in the central Santa Monica Mountains (1975, 1977, 1978), and 3— 6 seeds at Chalone Peak, San Benito County (n = 4, 1974). This compares with seed cohorts of Washingtonia filifera dispersed by Canis latrans in eastern San Diego County, of 1-275 seeds (48.7 + 3.1, n = 252: Bullock, Principles 24:29-32. 1980.). The endo- carp/seed units of Prunus weighed 1.46 + .04 g (n = 143), and Washingtonia seeds weighed 0.10 g (n = 50). Experiments were conducted to observe survival and growth of Prunus tlicifolia, particularly with reference to aggregation of the seeds. Seeds were collected in the Santa Monica Mountains, and grown at the University of California, Los Angeles, in silty loam 25 cm deep resting on the natural substrate. The plants had only partial morning shade, and were watered only by rain. From December 1973 to November 1974, plants were grown in 1 X 2-m plots with 50 seeds each in four unreplicated conditions of inter- seed spacing: 20 cm, 10 cm, 5 cm, and O cm. Also in this year, cohorts (0 cm seed TABLE 1. SURVIVAL FROM SEED TO FIRST-YEAR SEEDLING AND MEAN ABOVE- GROUND DRY WEIGHT OF SURVIVORS (+ S.e). Aggregation n Survival Biomass (g) Seed spacing (cm) 0) 50 seeds 0.74 2.9 2035 5 50 0.64 6.1 + 1.2 10 50 0.20 62 = 1.5 20 50 0.22 142+ 0.3 Seeds per cohort 4 7 cohorts 0.79 14:9--- 279 Z fi 0.50 12a see 223 1981] NOTES AND NEWS 95 Fic. 1. Graft between roots of two three-year-old plants of Prunus ilicifolia. spacing) of 2 and of 4 seeds were grown (7 replicates each), spaced at 50 cm. From November 1974 to November 1977, plants were grown from cohorts of 1, 2, 3, 4, 5, 6, 8, 10 and 12 seeds (at least 9 replicates each), the cohorts spaced at 50 cm. Both experiments of 1973-1974 showed greatest seedling survival in the most aggre- gated conditions (Table 1). This may be attributable to decreased desiccation during the summer, due to mutual shading. The mean above-ground biomass of survivors was greatest at intermediate densities in the uniform-spacing plots, but was much greater in the cohorts than in the spaced seedlings (Table 1). Apparently the full survival value of aggregation in these conditions was attained by only 4 seeds and massive aggregation reduced growth but not early survival. From the 1974 planting, a harvest in November 1977 showed 89 percent survival from seed in cohorts of 1—8 seeds (n = 104 seeds, 32 cohorts), and 88 percent survival in cohorts of 10 and 12 seeds (n = 96 seeds, 9 cohorts). Biomass studies were not feasible but root diameters were measured on a subsample. Comparing cohorts of initial seed numbers of 1, 2—6, and 10 or 12, the total root cross-sectional areas were not significantly different between small, medium and large cohorts. However, the means for individual plants showed a sharp decrease across the three groups (F’; = 12.34, p < 0.01), and more variability among isolated plants. Grafting was commonly observed among individuals of a cohort (Fig. 1). Grafting was present in 12 of 30 cohorts with more than one survivor, and joined up to five plants. Grafting was most common near the surface and was distinct from the devel- opment of burls with multiple stems. An interesting consequence of the high survival of cohort seedlings is that any phy- siognomic shrub may contain several genets, which may have had different female parents. The colonial shrub itself may be a small breeding neighborhood, and present some variety in all aspects of its ecological behavior. Furthermore, some direct phys- iological interaction among the genets may be possible due to grafting. The fate of individual plants in dispersal cohorts merits wider and closer attention.—STEPHEN H. BULLOCK, Botany Department, San Diego State University, San Diego, CA 92182. (Received and accepted 16 Oct 1980.) 96 MADRONO [Vol. 28 ADVENTITIOUS ROOTING IN COASTAL SAGE SCRUB DOMINANTS.—Adventitious root- ing was discovered in Artemisia californica Less., Eriogonum fasciculatum Benth., Salvia mellifera Greene., and Salvia apiana Jeps. (voucher specimens in MACF) while collecting plant specimens for a flora of Starr Ranch, a 1600-ha Audubon Sanctuary 11 km east of San Juan Capistrano, Orange Co., CA. There is little information on the occurrence and degree of vegetative spread in common native species (Davis and Hey- wood, Prin. Angio. Tax. 1973). T. L. Hanes (pers. comm., 1976) indicated that ad- ventitious rooting had not previously been reported in these species of the southern California coastal sage scrub. A study of these species was initiated from February to October 1976, and during September 1980 to determine a) frequency of occurrence of. adventitious rooting; b) environmental factors necessary for root initiation; and c) possible functions of adven- titious rooting. Observations were made by examining a minimum of 50 shrubs of each species in a variety of habitats at Starr Ranch (in the coastal foothills near the southern limit of the Santa Ana Mts.) and along Black Star Road (at the northern end of the Santa Ana Mts.). The dominant species were scored for the presence or absence of adventitious roots. The results for three habitats are given in Table 1. Of the four dominants, the highest average occurrence of adventitious rooting across all habitats was E. fasiculatum (x = 81 percent), followed by S. mellifera (« = 42 percent), S. apiana (x = 41 percent), and A. californica (x = 10 percent). Both E. fasciculatum and S. mellifera produce many decumbent and vertical stems. Decumbent stems, which form the majority of adventitious roots, arise near the base of a plant and spread horizontally along the ground. Decumbent stems are often incon- spicuous when covered with soil and leaf litter. Numerous vertical shoots develop along them; when the shoots are small, they may resemble seedlings. Eriogonum fasiculatum develops numerous adventitious roots even in the most un- favorable habitats (e.g., on slopes of 80 percent gradient; on dry, south-facing slopes; and on rocky outcrops). Ten percent of the shrubs of this species sampled on level ground showed evidence of an outward circular growth pattern emanating from a single (usually dead) shrub, similar to the “clonal ring” growth of Larrea tridentata described by Vasek et al. (Madrono 23:1—13. 1975). The growth habit of S. apiana and A. californica differs from E. fasciculatum and S. mellifera in that the first two species have many erect branches and few decumbent stems. Adventitious roots develop in S. apiana and A. californica mainly where the basal portions of their vertical stems are covered with soil from erosion or silt deposits. In habitats where soils are packed, there was no evidence of adventitious rooting re- gardless of slope aspect or amount of shade. Edaphic conditions appear to be an im- portant factor in the formation of adventitious roots in these species. The environmental factors necessary for the development of adventitious roots appear to be the same as those required in horticultural applications for root development in TABLE 1. PERCENTAGES OF INDIVIDUALS OF COASTAL SAGE SCRUB DOMINANTS WITH ADVENTITIOUS Roots, ARRAYED BY HABITAT. N = total number of individuals sampled in all habitats. Habitats Species N Slopes Level Riparian Mean Eriogonum fasciculatum 59 78 ie) 100 81 Salvia mellifera 72 30 20 75 42 Salvia apiana 51 40 19 80 41 Artemisia californica 98 14 8 0 10 1981] NOTES AND NEWS 97 layering, i.e., presence of humus and moisture in close contact with the stem, and elimination of light. Adventitious rooting may serve several functions, including increasing root surface area. The rate and extent to which roots occupy soil volume is known to be critical to the survival of perennial species. In addition, avoidance of competition by stratification of root systems has been reported in a number of species and habitats (Etherington, Envir. pl. ecol. 1975). Thus, development of adventitious roots may be a means to increase water or nutrient absorption capacity. Eriogonum fasciculatum and S. melli- fera should be investigated further in this regard because their growth habits suggest that their adventitious roots function to enlarge the volume from which soil resources can be absorbed. Adventitious roots may also help to establish progeny asexually by vegetative repro- duction. The growth habits of EF. fasciculatum and S. mellifera suggest such asexual spread, which may be advantageous in harsh environments where seedling establish- ment is unlikely. However, the coastal sage dominants are all prolific seeders (Hanes, Ecol. Monogr. 41:27—52. 1971) and many seedlings are found in the wild. Additional study is needed to determine the extent of adventitious rooting and its contribution to vegetative reproduction in coastal sage dominants. Once seedlings are established, veg- etative growth may give plants a competitive advantage. Adventitious roots may also provide greater stability for a perennial shrub in a con- tinually or periodically eroding environment. Formation of adventitious roots may be an adaptive response to edaphic conditions that enables the dominants to become re- established when buried by erosion, as in areas of creep or flood. The geography of much of the coastal sage scrub community is characterized by steep terrain and rocky, sandy soils that are extremely unstable (Hanes, op. cit.). Many shrubs of the dominant species are buried under soil from slides or erosion, and adventitious roots form along the buried portions of their stems. Because many species of plants cannot survive suffocation if their roots are covered too deeply with soil or water (Daubenmire, Pls. Envir. 1974), the capacity to continue growth when buried by soil appears to be the most important function of adventitious rooting in the coastal sage dominants. I thank Drs. T. L. Hanes, C. E. Jones, S. Carlquist, and F. Lang for their comments on the manuscript.—R. JOHN LITTLE, Rancho Santa Ana Botanic Garden, Claremont, CA 91711. (Received 22 Feb 1980; revision received and accepted 16 Oct 1980.) REVIEWS Inventory of Rare and Endangered Vascular Plants of California. By JAMES PAYNE SMITH, JR., R. JANE COLE, and JOHN O. SAWYER, JR. in collaboration with W. Ros- ERT POWELL. viii + 115 p. California Native Plant Society Special Publication 1, Berkeley. ed. 2, 1980. $7.50. Available from CNPS, 2380 Ellsworth, Suite D, Berkeley, CA 94704. In response to an idea from the fertile mind of its then president, G. Ledyard Stebbins, the California Native Plant Society in 1968 launched its Rare Plant Project to develop information about the rare plants of the state. At the time, intellectual curiosity rather than legal need was the main motivation. That plants had been added to the Endan- gered Species Act in 1973 came as a surprise to the Society some time after the fact. But by then several versions of a preliminary list had already been developed and circulated to the state’s botanical community for comment, initially by Roman Gankin, and, starting in 1971, by W. Robert Powell, who contributed the enormously useful concept of the R-E-V-D coding. While federal efforts were still in developmental stages, the Society in late 1974 published the predecessor of the present edition and it imme- diately found wide applicability for federal agencies under the National Environmental 98 MADRONO [Vol. 28 Policy Act of 1969 and for various California planning agencies, both local and state- wide, especially in relation to the California Environmental Quality Act of 1970. New responsibilities of federal land-managing agencies arising from the 1973 federal act made inevitable a surge of field work on rare plants that began to produce much new information. Realizing this, as well as the existence of gaps in information contained in the 1974 inventory, CNPS’s new Rare Plant Committee under my chairmanship and working closely with Powell launched revisionary efforts in mid-1976. With the aid of many volunteers, we were able simultaneously to expand the information base by thor- ough combing of the literature in the course of preparing status reports on 247 rare plants for the U.S. Forest Service in 1977 to 1978, and to net a profit of well over ten thousand dollars from that work. This financed adding voucher information from seven additional California herbaria and reviewing these herbaria and those gleaned for the 1974 edition for information about 187 taxa that were either last-minute additions without data to the 1974 inventory or were later proposed for addition. A further 174 status reports prepared for the State Department of Fish and Game in 1979 under the direction of James P. Smith, Jr., who succeeded me as Rare Plant Committee chairman in February 1979, contributed to the wealth of new information. The revision resulting from all this activity is most welcome and is indispensable, as was its predecessor, to those with responsibilities toward rare plants. But the 1980 version is marred by numerous inconsistencies and errors, typographical and otherwise. Some, such as misspellings, are easily noticed. Others—ones in taxon codes and numbers signifying rarity, endangerment, vigor, distribution, and county and quadrangle of known occurrence—are not so easily detected. Users should be forewarned and should take advantage of the Society’s intent to issue lists of addenda (see Preface). Arrangement has been changed from the earlier version. Taxa presumed extinct are segregated into List 1, containing 44 plants. Those deemed rare throughout their ranges, whether or not endemic to California, make up the 645 plants of List 2. Taxa considered rare but not endangered throughout their ranges are in List 3 and number 447. The 236 plants of List 4 are taxa rare in California but common elsewhere. The old “main list” of edition 1 thus includes taxa now found in Lists 1, 2, and 4. I would have preferred the federal way of flagging possibly extinct plants—an asterisk signifying that information is especially desired about them—to listing them separately. Few such plants have been thoroughly searched for and, with good fortune, the situation is likely to change. At least five of the taxa in List 1 were recovered during 1980 field work; one more was recovered in 1978 but not then reported. The heading for this list is misleading in that taxa restricted to California are lumped with taxa found outside the state and not necessarily extinct there. Seven of the 44 are considered widespread outside California. Primary applicability of the inventory is to the existing legal framework, but discus- sion of this is lacking. For this reason, inclusion of the purported legal status of various taxa under federal and the new state law (Native Plant Protection Act of 1977, passed at the instigation of CNPS) is especially troublesome. With plants actively under review by both entities, such information can be of only transitory accuracy. Judging from the many queries I receive because of my activity in this area, legal status, or especially lack of it, is information all too likely to be misunderstood and misused by unknowing or unscrupulous persons in attempts to make erroneous points. Lack of legal status seldom indicates a biological judgment; rather it is likely to reflect continuing manpower and political problems. It is important to know that actual listing is not necessary under either state or federal law to bring to bear certain protective courses of action [see fuller discussions of re- quirements under state law and of the federal rare plant program in Howard, Fremontia 7(3):18. 1979, and 8(3):14—16. 1980, respectively]. For all the foregoing reasons I should have preferred no mention of legal status, thereby making more likely that inquirers would seek up-to-date information from sources also able to clarify any mistaken ideas. Ratings assigned by the Smithsonian Institution are included but have no legal sig- nificance. Smithsonian and State ratings are broken down dichotomously (“endangered” 1981] NOTES AND NEWS 99 and “threatened,” and “endangered” and “rare,” respectively) but federal ratings (also “endangered” and “threatened”) are lumped into merely “listed.” Original plans for the revision were supposed to have included listings of rare taxa by county and by quadrangle of known occurrence. Neither is present. Especially the latter would make many inquiries unnecessary and would be easily producible if the information is still computerized as was the case with the earlier inventory. Perhaps it can be issued with other addenda. The index is a useful addition. A method of footnoting the plants listed in Appendix 1 so as to distinguish among the various categories lumped therein would have been helpful.—ALIcE Q. HOWARD, University Herbarium, Department of Botany, Univer- sity of California, Berkeley 94720. A Taxonomic Study of the Ranunculus hispidus Michaux Complex in the Western Hemisphere. By THOMAS DUNCAN. Univ. California Publ. Bot. 77:1-125, 70 Figs. Univ. California Press, Berkeley. 1980. ISBN 0-520-09617-7. $7.00. Past classifications of this widespread group have been plagued by similarities and overlaps of leaf-shapes previously used as primary criteria for identification. As in most other buttercups, not only does leaf-shape in this group vary widely among individuals of the same species but even among leaves on the same plant. Duncan has sought a new classification, widening the range of characters used and applying new statistical techniques aimed at giving a more strictly objective basis for subdividing the complex. After a brief outline of data-gathering methods, the monograph provides a full dis- cussion of each of the characters to be used, some for the first time. A lucid account ensues of three different computer-assisted approaches that were tried out—cluster anal- ysis, sum-of-fractions analysis, and principal-components analysis. A logical step-by- step procedure shows how a particular “similarity value” was selected to satisfy best Duncan’s chosen species distinction (discontinuity in at least three characters). The classification that emerges from application of the first two computer-assisted ap- proaches was tested out to Duncan’s satisfaction on many thousands of herbarium specimens. Its superior discriminatory value over more traditional treatments by Benson and by Lourteig appears to be quite clear. The third computer-assisted approach, prin- cipal-components analysis, was found in this study to be much less useful than the first two. As to plant characteristics, neither leaf-shape measurements nor leaf-flavonoid compounds turned out to be useful in classifying this group, and even chromosome numbers, falling into only two classes (2n = 32 or 64), were found irrelevant taxonom- ically. A large section of the monograph is devoted to a standard descriptive revision. Each of the 20 taxa in this new treatment is also illustrated and mapped and provided with an extensive commentary that clarifies many otherwise unanswered questions. The diagnostic key attempts to avoid developmentally- and genetically-labile leaf characters, which tended to weaken previous classifications. Receptacle shape, stolons, and rhi- zomes are among new features helping discriminate taxa. An extensive appendix of background data contains 20 pages of illustrations of leaf-shape variations in each species, fully vindicating the author’s rejection of them as usable taxonomic features. Although up-to-the-minute numerical methods of data analysis have been used in this revision, its underlying assumptions seem to this reviewer to be curiously old- fashioned. It provides a vital preliminary step to a much more biosystematically com- prehensive operation yet to be carried out. Conspicuously missing are the experimental data that for half a century have been recognized as inescapable components of con- temporary systematic investigations in both plants and animals. Where are the breeding tests, transplant experiments, mass geographic samplings, and cytogenetic data crucial to evaluation of stages in the evolutionary process? But even without these other kinds of evidence, Duncan’s monograph, as a study using new methods of objective phenetic analysis, is an impressive foundation stone for 100 MADRONO [Vol. 28 more complete studies in the future. Can we dare hope that such efforts, and the funding such extensive studies require, will be sustained for building upon such excellent mor- phological preliminaries as thisp—FULTON FISHER, Department of Biological Sciences, Simon Fraser University, Burnaby, British Columbia, Canada VSA 1S6 Flora Americae Septentrionalis. By FREDERICK PURSH. xxiv + 751 p., illus. 1814 [Dec 1813]. Reprint, 1979, introduced (117 p.) and edited by JoSEPH Ewan. J. Cramer, Vaduz, Liechtenstein. Available from Lubrecht and Cramer, RFD 1, Box 227, Mon- ticello, NY 12701. ISBN 3-7682-1242-4. $60.00. In his pithy introduction, Ewan provides synopses of the place of Pursh’s Flora in botanical history, Pursh’s collecting activities, other collectors and specimen sources, conditions of publication of the Flora, and initial response of the botanical community. In addition, there is a brief chronology of Pursh’s life, a gazetteer of localities, and a very welcome annotated list of the 500 to 600 species and varieties that were first published in the Flora. The inventory is alphabetical by genus, then by species, and the annotations provide (so far as known): page in the Flora, whereabouts of type or “authentic” specimens, commentary (including references to pertinent publications), and accepted name (if original is generally relegated to synonymy). This invaluable list seems to be as com- prehensive as is practicable and represents nearly 30 years of gleanings from American and European libraries and herbaria. The Flora itself treats some 3076 species (fide Ewan). It was the “first account of North American plants to include the Pacific Northwest.” Among the “novelties” were plants then known only from the Lewis and Clark collections, including original ac- counts of Lewisia rediviva and Clarkia pulchella.—JOHN L. STROTHER, University Herbarium, Department of Botany, University of California, Berkeley 94720. TRANSITION FRANK WALTON GOULD Frank Walton Gould, Distinguished Professor Emeritus of Grass Systematics and former curator of the S. M. Tracy Herbarium at Texas A&M University died on 11 March 1981 in Austin, Texas. Gould was born in Mayville, North Dakota on 25 July 1913. He earned his bachelor’s degree from Northern Illinois University, a master’s degree from the University of Wisconsin, and Ph.D. degree in botany from the University of California. He taught biology at Dixie Junior College, St. George, Utah from 1941-1942, and at Compton Junior College, Compton, California from 1942-1944. He then worked as a botanist at the University of Arizona from 1944-1949. In 1949 Gould moved into a taxonomic position at Texas A&M University. He served as curator of the S. M. Tracy Herbarium until August 1979, when he retired. During his tenure, he built the herbarium into one of the most respected such facilities in the United States. Gould was a world renowned grass systematist and had completed teaching and research assignments in Mexico, Costa Rica, the Dominican Republic, Brazil, Puerto Rico, Sri Lanka, and England. His research projects with leading herbaria resulted in more than 80 definitive treatments of grasses that are recognized world wide. Frank W. Gould authored the books Grasses of the Southwestern United States in 1959, Grasses of the Texas Coastal Bend in 1965, Grass Systematics in 1968, The Grasses of Texas in 1975, and Common Texas Grasses in 1978. He also had completed a book on the Grasses of Baja California that will be published next year. At the time of his death he was writing a book on the grasses of Mexico.—STEPHAN L. HaTcu, Curator of the Tracy Herbarium, Texas A&M University, College Station 77843. Membership in the California Botanical Society is open to individuals ($18 per year; students $10 per year for a maximum of seven years). Members of the Society receive MADRONO free. Family memberships ($20) include one eight-page publishing allotment and one journal. Emeritus rates are available from the Corresponding Secretary. In- stitutional subscriptions to MADRONO are available ($25). Membership is based on a calendar year only. Applications for membership (including dues), orders for sub- scriptions, requests and rates for back issues, changes of address, and undelivered copies of MADRONO should be sent to the Corresponding Secretary. Renewal payments should be sent to the Treasurer. INFORMATION FOR CONTRIBUTORS Manuscripts submitted for publication in MADRONO should be sent to the editor. All authors must be members, and membership is prerequisite for review. Manuscripts and review copies of illustrations must be submitted in triplicate for all articles and short items intended for NOTES AND NEWS. Follow the format used in recent issues for the type of item submitted and allow ample margins all around. All manuscripts MUST BE DOUBLE SPACED THROUGHOUT. For articles this in- cludes title (all caps, centered), author names (all caps, centered), addresses (caps and lower case, centered), abstract, 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, and number each page. Do not use a separate cover page, “erasable” paper, or footnotes. Table captions should include all information relevant to tables. All measurements should be in metric units. Line copy illustrations should be clean and legible, proportioned (including captions) to the MADRONO page, and designed for reduction to % original size. 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 latitude and longitude references. Halftone copy should be designed for reproduction at actual size. In no case should original illustrations be sent prior to the acceptance of a manuscript. When needed they should be mounted on stiff cardboard and sent flat. No illustrations larger than 22 < 28 cm will be accepted. Presentation of nomenclatural matter (accepted names, synonyms, typification) should follow the format used for Rhus integrifolia in MADRONO 22:288. 1974. Institutional abbreviations in specimen citations should follow Holmgren and Keuken, Index Her- bariorum, 6th edition. Abbreviations of serial titles should be those in Botanico-Per- iodicum-Huntianum (Lawrence et al., 1968, Hunt Botanical Library, Pittsburgh). If the correct abbreviation cannot be determined, the full serial title should be used. Titles of books should be given in full, together with the place and date of publication, publisher, and edition, if other than the first. All members of the California Botanical Society are allotted eight pages in MADRONO over a two year period. Joint authors are each assessed the full page number. Beyond that number of pages a required editorial fee of $40.00 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 sched- ule, with equity among all members for access to its pages. Printer’s fees for illustrations and typographically difficult material @ $35.00 per page (if their sum exceeds 30 percent of the paper) and for author’s changes after typesetting @ $3.00 per line will be charged to authors. Contents, continued REVIEWS JAMES PAYNE SMITH, JR., R. JANE COLE, and JOHN O. SAWYER, JR. in collaboration with W. ROBERT POWELL, oom Tyrentory of Rare and Endangered Vascular Plants of California (Alice Q. Howard) THOMAS DUNCAN, A Taxonomic Study of the Ranunculus hispidus Michaux Complex in the Western Hemisphere (Fulton Fisher) JOSEPH EWAN, ed., FREDERICK PURSH’S Flora Americae Septentrionalis (John L. Strother) TRANSITION ANNOUNCEMENTS 97 99 100 100 66, 90 CALIFORNIA BOTANICAL SOCIETY | MIS2 Bot. MADRONO VOLUME 28, NUMBER 3 JULY 1981 Contents POLLINATION BIOLOGY OF CALYPSO BULBOSA VAR. OCCIDENTALIS (ORCHIDACEAE): A FOOD-DECEPTION SYSTEM, James D. Ackerman 101 THE HISTORICAL ROLE OF FIRE IN THE FOOTHILL COMMUNITIES OF SEQUOIA NATIONAL PARK, David J. Parsons 111 TAXONOMY OF PHACELIA SECT. MILTITZIA (HYDROPHYLLACEAE), Richard R. Halse 121 ALBERT M. VOLLMER: A MEDICAL DOCTOR WHO LOvED LILLIES, Ira L. Wiggins 133 FIVE NEW SPECIES OF MEXICAN ERIGERON (ASTERACEAE) Guy L. Nesom 136 THE DIANDROUS, HYPOSTOMATIC WILLOWS (SALICACEAE) OF THE CHIHUAHUAN DESERT REGION, Marshall C. Johnston 148 A NEw SPECIES OF CRYPTANTHA (BORAGINACEAE) FROM WYOMING, Robert D. Dorn and Robert W. Lichvar 159 ERIOGONUM LIBERTINI (POLYGONACEAE), A NEW SPECIES FROM NORTHERN CALIFORNIA, James L. Reveal 163 DIURNAL ACID METABOLISM IN VERNAL POOL ISOETES (ISOETACEAE), Jon E. Keeley 167 THE ECOLOGICAL STATUS OF STIPA PULCHRA (POACEAE) IN CALIFORNIA, - James W. Bartolome and Barbara Gemmill 172 NOTEWORTHY COLLECTIONS LUPINUS CITRINUS and STREPTANTHUS FARNSWORTHIANUS, WEST AMERICAN JOURNAL OF BOTANY Jim A. Bartel 184 THELYPODIOPSIS PURPUSII and NEMACLADUS GLANDULIFERUS VAR. ORIENTALIS, Darrell Ward and Richard Spellenberg 185 CAREX DEWEYANA subsp. DEWEYANA, Theodore S. Cochrane 186 WOLFFIA COLUMBIANA, Wayne P. Armstrong 187 CALYPTRIDIUM PULCHELLUM, Dan Hamon 188 A (Continued on back cover) 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, Herbarium, Life Sciences Building, University of California, Berkeley, CA 94720. Subscription rate: $25 per calendar year. Subscription information on inside back cover. Established 1916. Second-class postage paid at Berkeley, CA, and additional mailing offices. Return requested. POSTMASTER: Send address changes to Dr. Frank Almeda, Botany Dept., California Academy of Sciences, San Francisco, CA 94118. Editor—CHRISTOPHER DAVIDSON Natural History Museum of Los Angeles County 900 Exposition Blvd., Los Angeles, CA 90007 (213) 744-3378 Board of Editors Class of: 1981—DANIEL J. CRAWFORD, Ohio State University, Columbus JAMES HENRICKSON, California State University, Los Angeles 1982—DEAN W. TAYLOR, University of California, Davis RICHARD VOGL, California State University, Los Angeles 1983—-ROBERT W. CRUDEN, University of lowa, Iowa City DUNCAN M. PoRTER, Virginia Polytechnic Institute and State University, Blacksburg 1984—Mary E. BARKWORTH, Utah State University, Logan Harry D. THIERS, San Francisco State University, San Francisco 1985—STERLING C. KEELEY, Whittier College, Whittier, CA ARTHUR C. GIBSON, University of California, Los Angeles CALIFORNIA BOTANICAL SOCIETY, INC. OFFICERS FOR 1981 President: ROBERT ORNDUFF, Department of Botany, University of California, Berkeley 94720 First Vice President: LLAURAMAY T. DEMPSTER, Jepson Herbarium, Department of Botany, University of California, Berkeley 94720 Second Vice President: CLIFTON F. SMITH, Santa Barbara Museum of Natural History, Santa Barbara, CA 93105 Recording Secretary: ROBERT W. PATTERSON, Department of Biology, San Francisco State University, San Francisco, CA 94132 Corresponding Secretary: FRANK ALMEDA, Department of Botany, California Academy of Sciences, Golden Gate Park, San Francisco 94118 Treasurer: CHERIE L. R. WETZEL, Department of Biology, City College of San Francisco, 50 Phelan Ave., San Francisco, CA 94112 The Council of the California Botanical Society consists of the officers listed above plus the immediate Past President, WAYNE SAVAGE, Department of Biology, San Jose State University, San Jose, CA 95192; the Editor of MADRONO; three elected Council Members: PAUL C. SILVA, University Herbarium, Department of Botany, University of California, Berkeley 94720; JOHN M. TUCKER, Department of Botany, University of California, Davis 95616; CHARLES F. QUIBELL, Department of Biological Sciences, Sonoma State College, Rohnert Park, CA 94928; and a Graduate Student Represen- tative, KENT HOLSINGER, Department of Biological Sciences, Stanford University, Stanford, CA 94305. POLLINATION BIOLOGY OF CALYPSO BULBOSA VAR. OCCIDENTALIS (ORCHIDACEAE): A FOOD-DECEPTION SYSTEM JAMES D. ACKERMAN Department of Biology, University of Puerto Rico, Rio Piedras, Puerto Rico 00931 ABSTRACT Calypso bulbosa var. occidentalis is a western North American terrestrial orchid. In the northern California populations studied, these plants are allogamous and pollinated by Bombus queens and Psithyrus females. Bee visitation, however, is infrequent. Ca- lypso bulbosa var. occidentalis is a generalized food-flower mimic and must rely on exploratory visits of naive bees for pollination. Abrupt rises in the number of flowers pollinated corresponds with sudden increases in available pollinators. Fruit is set low (11-34 percent), and is comparable to that mentioned in reports of other temperate food-deceptive orchids. A large number of orchids provide no reward to their pollinators. Their visual, olfactory, and sometimes tactile cues are deceitful. Per- haps the most striking examples are the pseudocopulatory orchids, Ophrys, Trichoceros, and a variety of Australian species (Kullenberg and Bergstrom, 1976; Pijl and Dodson, 1966; Stoutamire, 1975). The majority of non-rewarding orchids, however, are pollinated by visitors apparently searching for food. Very few such systems have been stud- ied in detail. Recently, much attention has been focused on Calypso bulbosa (L.) Oakes, a strikingly beautiful circumboreal orchid, rep- resented by var. bulbosa in Eurasia, var. japonica Makino in Japan, var. americana (R. Br.) Luer in North America from the Rocky Moun- tains eastward, and var. occidentalis (Holz.) Luer in the Pacific Northwest (Luer, 1975). Mosquin (1970) and Stoutamire (1971) found that C. bulbosa var. americana is nectarless and hypothesized that pollinators are attracted to the flower because of its conspicuous col- oration. Furthermore, the flowers appear to rely on newly emerged naive bumblebees to effect pollination. Two unpublished theses treat some features of the pollination of C. bulbosa var. occidentalis. Kipping (1971) states that the flowers have nectaries, whereas Krell (1977) and my simultaneous observations in- dicate that the flowers are nectarless. These conflicting reports did not clarify whether C. bulbosa var. occidentalis has a fundamentally different pollination ecology from its eastern counterpart. The objec- tive of this study is to elucidate its pollination biology. MADRONO, Vol. 28, No. 3, pp. 101-110, 31 July 1981 102 MADRONO [Vol. 28 RA: NAN SA YW AMY NONE \ INNS . SS - 1p, 4 Bi. SA ‘ VOTED Sy we Vk 4 y N Fic. 1. Calypso bulbosa var. occidentalis and pollinator. Variation in labellum spot patterns is shown. The Bombus caliginosus queen has a pollinarium attached to its scutellum. 1981] ACKERMAN: POLLINATION OF CALYPSO BULBOSA 103 TABLE 1. PERCENT FRUIT SET OF SIX POPULATIONS OF Calypso bulbosa VAR. occidentalis. All localities are in Humboldt County, California. The number of visited flowers is the number of pollinated flowers plus the number of flowers with pollinaria removed but not pollinated. In 1978, all flowers at the first two sites listed were fortified with 15-45 wl of 10 percent sucrose solution (wt/wt). Total Number Number number of of pol- Percent of flowers _ linated fruit Locality Year flowers visited flowers set Lanphere-Christensen Dunes 1976 1273 671 141 tH Nature Conservancy, 8 km w. 1978 1161 398 82 of Arcata: dune forest. Lord Ellis Summit, Hwy 299: 1978 Cy 2 0) 0 mixed evergreen forest. Korbel, Camp Bauer at 1978 114 — 34 30 entrance: redwood forest. Kneeland, Freshwater Rd., 1978 349 —_ 56 16 0.5 km w. of jct. of Greenwood Hgts. Rd.: redwood forest. Kneeland, Maple Cr. Rd., 1978 L3/ — 47 34 2 km n. of Freshwater Rd.: mixed evergreen forest. Redwood Valley, Chezem Rd., 1978 86 a 28 33 1 km e. of hwy 299: mixed evergreen forest. FLOWER MORPHOLOGY AND METHODS Calypso bulbosa var. occidentalis grows in the thick litter of conif- erous forests. The plants are perennial and in the fall produce a single ovate leaf from an underground corm. The plants remain in this con- dition over winter. In early spring a single-flowered inflorescence emerges (Fig. 1). The sole nodding flower has rose-pink sepals and petals. The lamina of the lip is white with reddish-brown spots and is adorned with several rows of hairs. The saccate portion of the lip has numerous dark, reddish-brown stripes. The very broad, slightly arched column is also rose-pink and forms a hood over the lip. During the summer the plants become dormant: the leaf withers, the fruits dehisce, and the many thousands of seeds are dispersed. Six C. bulbosa var. occidentalis populations in Humboldt County, California were studied (Table 1). The Lanphere-Christensen Dunes Nature Conservancy population was observed during the spring sea- sons of 1976 and 1978. The other populations were studied in 1978 only. At the Lanphere Dunes, I tested Calypso for self-compatibility and autogamy. I covered ten plants with insect-exclusion cages and as 104 MADRONO [Vol. 28 these plants flowered, I self-pollinated them. Thirteen other plants were caged and left untouched. Percent seed set was estimated for all capsules developed in these experiments, and also for capsules of ten naturally pollinated flowers collected from the same population. The contents of each capsule were spread on a slide and nine fields were viewed at 40x. I considered a seed viable if it contained a well-de- veloped embryo. The flowering and fruiting phenologies of the Lanphere Dunes pop- ulation were assessed in 1976 and 1978. The parameters recorded were the number of unvisited flowers, flowers with pollinia removed, flow- ers pollinated, withered flowers, and fully developed capsules. For the other populations, I recorded the total number of flowers and polli- nations. During the 1978 season, I added 15-45 pl of 10 percent sucrose solution (wt/wt) to all flowers of two populations, Lanphere Dunes and Lord Ellis Summit (Table 1). The artificial nectar was injected into the spurs with a syringe without damage to the flowers. At least once a week flowers were fortified. All those appearing flaccid due to apparent effects of the sugar solution were removed so that only fresh flowers were present. These manipulations were designed to test the effect of food reward on foraging patterns and visitation frequencies. At approximately one-week intervals I observed the Lanphere Dunes population for pollinator activity. Bumblebees were captured and examined for pollinaria. Bees that carry orchid pollinaria are considered legitimate pollinators (Dressler, 1976). For the 1976 season, the number of bees seen per hour along a regular transect route, re- gardless of their activity, was recorded as a measure of bee abundance. These observations lasted 1-4 hours when the bees were most active (mostly 0900-1700 PST). RESULTS In these northern California sites, Calypso is one of the first-bloom- ing spring flowers. The flowering season generally occurs from March to June in Humboldt County, California (Fig. 2). Vegetative propagation is not prevalent at any of the Calypso pop- ulations in northern California with which I am familiar. Plants grow- ing in close proximity were examined for rhizomatous connections. Rarely were these connections found and not one incidence was de- tected in the exceptionally large population of 2000-3000 plants (in- cluding many seedlings) at the Lanphere Dunes. Calypso bulbosa var. occidentalis is self-compatible but not autog- amous. All ten flowers self-pollinated by hand set fruit. The mean percent seed set was 81.6 (£10.78 S.D.), which is comparable with that of the ten naturally pollinated flowers (89.4 + 9.95). All 13 caged flowers failed to set fruit. 1981] ACKERMAN: POLLINATION OF CALYPSO BULBOSA 105 Nectar was not detected in flowers in any of the Calypso populations studied nor from 30 flowers of potted plants grown in an insect-exclu- sion cage. Flowers given artificial nectar were adversely affected. Those with 15—45 wl of 10 percent sucrose solution wilted within two weeks, whereas unaltered flowers lasted an average of three. A sugar solution of 30 percent induced wilting within 24 hours. Fruit set varied from 0-34 percent (Table 1). Two flowers were visited (their pollinaria were removed) at the Lord Ellis Summit pop- ulation but no pollinations occurred. The Lanphere Dunes population had 11 percent fruit set in 1976 but dropped to 7 percent in 1978 when artificial nectar was added to flowers (G-test, Sokal and Rohlf, 1969: G = 9.94, df = 1, p < 0.005). There was a substantial decrease in the number of flowers visited (number of flowers with only their pol- linarium removed plus the number of pollinated flowers). In 1976, there were 671 flowers visited, whereas in 1978 there were only 398 (G = 33.25, df = 1, p < 0.005). Pollinator abundance was similar for the two seasons so the difference may be due to short flower life and slight lip sag induced by the artificial nectar. Insect visits to C. bulbosa var. occidentalis were infrequent; how- ever, pollinarium-laden bumblebees of three species were caught: Bombus caliginosus, B. edwardsii (queens), and Psithyrus crawfordi (female). In all cases the pollinaria were attached to the hairless region just under the scutellum (Fig. 1). For attachment to occur there, bees had to enter the flower deeply and back out with arched bodies so that the edge of the scutellum contacted the viscidium, cementing the rest of the pollinarium to the bee. Both pollinium deposition and pollinar- ium removal occur only as the bee leaves the flower. First the bee passes the stigma where it may deposit pollinia and then the pollinar- lum is removed. Pollinium deposition and pollinarium removal are not guaranteed consequences of any given visit by a pollinarium-laden bee. This is because pollinator size is variable, as is the throat of the lip. At the Lanphere Dunes population, the throat gap, measured from the tip of the column to the hairy rim of the lip, ranged from 5-10 mm. Polli- narium loads of bumblebees reflect this imprecise flower-pollinator fit. Each pollinarium has four pollinia and any number of these or none at all may be left on the stigma. At the Lanphere Dunes, approxi- mately one flower was pollinated for every four pollinaria removed. The ratio of pollinated to visited flowers was 0.27 and 0.26 for 1976 and 1978, respectively. Several bees were observed visiting C. bulbosa var. occidentalis but failed to effect pollination. On three occasions B. occidentalis queens were seen entering flowers at the Lanphere Dunes population. Each bee visited a flower without a pollinarium load and failed to pick one up when it left. Emphoropsis cf. miserabilis (Anthophoridae), one of the most common bees at Lanphere Dunes, was seen visiting 106 MADRONO [Vol. 28 1200 800 400 NO. OF FLOWERS AVE. NO. FLOWERS POLLINATED 00 BEES PER OBSERVATION HR. ms MARCH APRIL MAY 1976 Fic. 2. Phenology of Calypso bulbosa var. occidentalis and its pollinators. A. Sea- sonality of flowering. B. Frequency of pollination expressed as the average number of flowers pollinated per day per census interval. C. Availability of potential pollinators. This includes Bombus queens and Psithyrus females. Data are from the 1976 season at the Lanphere-Christensen Dunes Nature Conservancy. C. bulbosa var. occidentalis six times but was also ineffective in re- moving pollinaria. This is apparently due to its small size. These bees typically visited 2—3 flowers before leaving the site. During 1976, the number of Calypso flowers pollinated each week at the Lanphere Dunes did not always rise proportionately with an increase in the number of available flowers. A sudden drop in polli- 1981] ACKERMAN: POLLINATION OF CALYPSO BULBOSA 107 nations occurred during a week in which the population was at peak flowering (Fig. 2). However, two distinct increases in the relative abundance of bumblebee queens bracketed the peak flowering period of Calypso (Fig. 2). These two increases correspond with the two abrupt rises in the number of flowers pollinated. DISCUSSION The mechanics of pollinarium removal and deposition clearly in- dicate that C. bulbosa var. occidentalis is an outcrossing species. Autogamy did not occur, corroborating the results of Kipping (1971) and Krell (1977). Insect-mediated self-pollination is improbable be- cause bees are not inclined to reenter flowers. Furthermore, geitonog- amous pollinations are unlikely because plants produce a single flower each year and, unlike some populations of var. americana and var. bulbosa (Mosquin, 1970; Mousley, 1925; Wollin, 1975), vegetative propagation is apparently rare. The western Calypso, like the other varieties studied (Mosquin, 1970; Stoutamire, 1971; Wollin, 1975), is pollinated by several species of bumblebees. Five of the ten species caught at the Lanphere Dunes are known pollinators of C. bulbosa var. occidentalis: queen B. cal- tginosus, B. vosnesenskii, B. edwardsiu, B. mixtus, and female P. crawfordi (present data; Kipping, 1971; Krell, 1977; R. Thorp, pers. — comm., 1977). Queens of seven other species are pollinators of the western Calypso at other localities (Krell, 1977; Mosquin, 1970; Thorp, pers. comm., 1977): B. melanopygus, B. rufocinctus, B. cen- tralis, B. pleuralis, B. bifarius nearcticus, B. flavifrons, and B. fri- gidus. Two common Lanphere Dunes species, B. californicus and B. occidentalis, are thus far known to pollinate only C. bulbosa var. americana (Mosquin, 1970; Thorp, pers. comm., 1977). These and other bumblebee species that are within the appropriate size range to effect pollinarium removal and pollinium deposition might also pol- linate the western variety if they encountered it. Bumblebee workers are common during the latter half of the blooming season of C. bulbosa var. occidentalis and most are apparently too small to be effective pollinators. Because there is variability in the size of the entrance to the lip, it is possible that large workers might pollinate the western variety, as they sometimes pollinate var. americana (Mosquin, 1970; Thorp, pers. comm., 1977). The western Calypso is deceptive because it presents no reward to its pollinators. Such orchids generally have low percent fruit set (Ack- erman, 1975; Dafni and Ivri, 1979; Thien and Marcks, 1972) and populations of Calypso in North America and Europe are no exception (Kipping, 1971; Krell, 1977; Mosquin, 1970; Mousley, 1924; Wollin, pers. comm., 1979; Table 1). Food deception in the Orchidaceae may involve relatively specific mimicry systems (Nierenberg, 1972; Pijl and Dodson, 1966). There is 108 MADRONO [Vol. 28 no evidence, however, of specific mimicry involving C. bulbosa var. occidentalis and associated spring-blooming species in northern Cal- ifornia. Heinrich (1976) and Jones and Buchmann (1974) noted that bumblebees with established floral preferences may mistake at a dis- tance the flowers of one species for another. The flowers involved are of similar size, coloration, and presentation (height of inflorescences, number of flowers per inflorescence, etc.). Nevertheless, such a mis- take is short-lived because the bees veer away upon closer inspection. This behavior was not observed for any bees foraging near or passing through the populations I studied. Mimicry would have to be even more fine-tuned to lure bumblebees for successive complete visits. Those flowers visited by bumblebee queens in the vicinity of the Ca- lypso populations I studied cannot be construed on an individual species basis as models for C. bulbosa var. occidentalis. At least sev- eral aspects of associated species differed radically from the orchid (e.g., habit, flower symmetry, shape, color, size, presentation). The western Calypso may thus be regarded as a generalized food-flower mimic. The flowers possess a constellation of characteristics typical of the bumblebee-pollinated food-flower syndrome (Heinrich, 1979) and do not necessarily mimic any specific species. This mode of pollination works because bumblebees occasionally explore for new food resources (Heinrich, 1976). Pollination of C. bulbosa var. occidentalis is depen- dent on these exploratory visits of uninitiated bumblebees. Heinrich (1975a, 1975b) noted that bumblebees visited Calopogon tuberosus (L.) B. S. P., another non-rewarding orchid, an average of 5.4 times in quick succession. He suggested that the color variation of the flowers enhances the number of visits the pollinators require to learn to avoid the species. The same process may be operating in C. bulbosa var. occidentalis with its variable and irregularly patterned lip (Fig. 1). In addition, my casual checks of fragrance production indicated that some flowers are quite fragrant, whereas others in the population are scarcely scented. Variability in odor production has been noted elsewhere for both North American Calypso varieties (Bradshaw, 1919; Krell, 1977; Mousley, 1924; Stoutamire, 1971). The phenological data obtained at the Lanphere Dunes during the 1976 season supports the hypothesis of visitation by naive bees. The number of pollinations occurring each week did not correspond pro- portionately with the number of flowers open and available in the population. This anomalous situation cannot be explained on the basis of weather conditions because that season was relatively mild without any radical weather shifts (U.S. Dept. Commerce, 1976, data for near- by Eureka, California). However, the two pollination peaks did cor- respond with a jump in abundance of potential pollinators. The abrupt drop in the number of pollinations during the peak flowering period may be attibuted to the phenomenon discussed above: when the first flush of bumblebees occurred, the uninitiated bees visited Calypso in 1981] ACKERMAN: POLLINATION OF CALYPSO BULBOSA 109 search of food resources. After the first week of emergence most bees either learned that Calypso was not a food resource or readily estab- lished foraging areas and food preferences (e.g., Vaccinium ovatum Pursh and Arctostaphylos uva-ursi (L.) Spreng.) without visiting the orchid. The second flush of bumblebees then repeated this process. Despite the drop in pollinations during the peak of flowering, there is synchrony between the blooming season of the orchid and the emer- gence and availability of its pollinators. Bumblebee queens increase in abundance at the time of peak flower availability. Reliance on newly emerged bumblebee queens for pollination has been suggested for both the eastern North American and Eurasian Calypso (Luer, 1975; Mos- quin, 1970; Stoutamire, 1971; Wollin, 1975), but until now supportive data have not been published. ACKNOWLEDGMENTS I thank R. Thorp for identifying the bees; A. Montalvo and M. Mesler for aid during many stages of the study; and N. Williams, R. Dressler, J. Atwood, K. Lu, J. Howard, R. Cruden, E. Guerrant, J. Hickman, and K. Steiner for critically reading the manu- script. The illustrations were rendered by A. Montalvo. Officials of the Lanphere-Chris- tensen Dunes Nature Conservancy kindly granted permission to study plants on the reserve. Portions of this study were supported by a Sigma Xi Grant-In-Aid, a grant to N. Williams from the American Orchid Society Fund for Education and Research, and the Department of Biological Science, Florida State University. LITERATURE CITED ACKERMAN, J. D. 1975. Reproductive biology of Goodyera oblongifolia (Orchidaceae). Madrono 23:191-198. BRADSHAW, R. V. 1919. Variations in Calypso. Amer. Bot. (Binghamton) 25:152. DAFNI, A. and Y. Ivri. 1979. Pollination ecology of, and hybridization between, Orchis coriophora L. and O. collina Sol. ex Russ. (Orchidaceae) in Israel. New Phytol. 83:181-187. DRESSLER, R. L. 1976. How to study orchid pollination without any orchids. Im K. Senghas, ed., Proc. Eighth World Orchid Conf., p. 534-537. German Orchid Soc., Frankfurt. HEINRICH, B. 1975a. The role of energetics in bumblebee-flower interrelationships. In L. E. Gilbert and P. H. Raven, eds., Coevolution of animals and plants, p. 141-158. Univ. Texas Press, Austin. .1975b. Bee flowers: a hypothesis on flower variety and blooming times. Evo- lution 29:325-334. 1976. The foraging specializations of individual bumblebees. Ecol. Monogr. 46:105-128. . 1979. Bumblebee economics. Harvard Univ. Press, Cambridge, MA. Jones, C. E. and S. L. BUCHMANN. 1974. Ultraviolet floral patterns as functional orientation cues in hymenopterous pollination systems. Anim. Behav. 22:481—485. KIPPING, J. L. 1971. Pollination studies of native orchids. Unpubl. M.A. thesis, San Francisco State College. KRELL, R. 1977. The pollination ecology of Calypso bulbosa var. occidentalis (Or- chidaceae). Unpubl. M.S. thesis, Washington State Univ., Pullman. KULLENBERG, B. and G. BERGSTROM. 1976. The pollination of Ophrys orchids. Bot. Not. 129:11-20. LuER, C. A. 1975. Native orchids of the United States and Canada excluding Florida. New York Botanical Garden, New York. 110 MADRONO [Vol. 28 MosqQuIn, T. 1970. The reproductive biology of Calypso bulbosa (Orchidaceae). Can- ad. Field-Naturalist 84:291-—296. MousLey, H. 1924. Calypso. J. New York Bot. Gard. 25:25-31. . 1925. Further notes on Calypso. Torreya 25:54—-59. NIERENBERG, L. 1972. The mechanism for maintenance of species integrity in sym- patrically occurring equitant oncidiums in the Caribbean. Amer. Orchid Soc. Bull. 41:873-882. PiyL, L. VAN DER and C. H. Dopson. 1966. Orchid flowers: their pollination and evolution. Univ. Miami Press, Coral Gables. SOKAL, R. R. and F. J. ROHLF. 1969. Biometry. W. H. Freeman, San Francisco. STOUTAMIRE, W. P. 1971. Pollination in temperate American orchids. In M. J. G. Corrigan, ed., Proc. Sixth World Orchid Conf., p. 233-243. Sydney, Australia. . 1975. Pseudocopulation in Australian terrestrial orchids. Amer. Orchid Soc. Bull. 44:226-233. THIEN, L. B. and B. G. Marcks. 1972. The floral biology of Arethusa bulbosa, Calopogon tuberosus, and Pogonia ophioglossoides (Orchidaceae). Canad. J. Bot. 50:2319-2325. U.S. DEPT. COMMERCE. 1976. California. Climatological data 80 (3-6). WOLLIN, H. 1975. Kring nornans biologi. Fauna och Flora (Stockholm) 70:89—98. (Received 27 May 1980; accepted 11 Nov 1980; revision received 5 Dec 1980.) THE HISTORICAL ROLE OF FIRE IN THE FOOTHILL COMMUNITIES OF SEQUOIA NATIONAL PARK DAVID J. PARSONS National Park Service, Sequoia and Kings Canyon National Parks, Three Rivers, CA 93271 ABSTRACT The historical role of fire in the foothill-chaparral and oak-woodland communities of Sequoia National Park must be understood before a management program can be de- veloped that will assure the perpetuation of those community types. Historically, light- ning-ignited fires were supplemented by intentional ignitions by Indians and, later, by European settlers. The park fire records since 1920 portray the frequency of both man- caused and lightning fires by year, month, and elevation. Fires have been most common at higher elevations and during summer. This corresponds with the period of maximum drought stress and minimum foliage moisture content. The largest share of the area burned has been within the chaparral community. However, nearly 75 percent of the zone has not burned during this time. The role fire should be allowed to play in the low-elevation foothill communities of Sequoia National Park poses a management dilemma. Although the objective of fire management programs in national parks is generally to allow fire to play as natural a role as possible in deter- mining vegetation mosaics (Parsons, 1977), the buildup of highly flam- mable fuels following decades of fire suppression often makes this impossible. Despite evidence that periodic fire has played a vital role in the evolution and maintenance of the foothill-chaparral and oak- woodland communities of the southern Sierra Nevada (Vankat, 1977), current fuel conditions are such that naturally ignited fires cannot safely be allowed to burn. Previous years of fire suppression have resulted in extensive, highly flammable, over-mature fields of brush (Parsons, 1976). In addition to the problems encountered in controlling summer wildfires, such conditions make it difficult to implement a safe, effective program of prescribed burning. This paper reviews available information on the fire history of the foothill-chaparral and oak-woodland communities of Sequoia National Park. This information provides part of the baseline data necessary for the development of an integrated fire management program for the area. The data were derived primarily from the park’s fire atlas, which lists and maps all of the fires that have occurred since 1920. Other accounts of the early fire history of the area in and around Sequoia National Park are reviewed in Vankat (1977) and Kilgore and Taylor (1979). MapRONO, Vol. 28, No. 3, pp. 111-120, 31 July 1981 112 MADRONO [Vol. 28 It is not possible to use standard fire history techniques in brush and grassland communities. Chaparral fires typically burn all or most of the above ground biomass. Thus, while it is possible to date the last fire, there is no accurate way of dating previous ones. Similarly, fires in grass or woodland areas are commonly of low enough intensity not to leave scars on the scattered trees. While it is known that fire plays an important ecological role in chaparral and woodland com- munities (Biswell, 1974; Griffin, 1977), it is difficult to document fully the fire history of such areas. Historical accounts and personal recol- lections must be relied on heavily in such situations. THE FOOTHILL ZONE The area encompassed by this study includes more than 28,000 hectares in the southern Sierra Nevada along the western boundary of Sequoia National Park. Located within the North, Middle, Marble, East, and South Fork drainages of the Kaweah River, the foothill zone spans an elevation range between about 460 m and 1830 m. The topography is generally steep with narrow canyons and cliffs. The climate of this zone is Mediterranean-type with hot dry summers and cool moist winters. Annual precipitation averages 65 cm at the lower elevations and more than 90 cm at the upper elevations. Summer temperatures frequently exceed 38°C. Soils are primarily sandy loams of granitic origin. They are often shallow on the steeper hillsides but attain several meters in depth in more favorable locations. Following Baker et al. (1981), four major plant communities can be identified as occurring within the foothill zone: foothill woodland, cha- mise chaparral, mixed-evergreen woodland, and black-oak forest. The foothill woodland community is characterized by a grassland under- story with scattered stands of Quercus douglassit and Aesculus cali- fornica. The dominant grasses are primarily introduced annuals that became established following intensive grazing during the late 1800’s (Vankat and Major, 1978). They are thought to have replaced native annuals as well as perennial bunchgrasses (Heady, 1977). Chamise chaparral is widely distributed on the xeric slopes of the foothill zone. Adenostoma fasciculatum is the dominant and sometimes only species found in this community. The mixed-evergreen woodland is a highly variable mixture of shrubs (e.g., Cercocarpus betuloides) and trees (e.g., Aesculus californica, Quercus chrysolepis) found primarily on mesic north slopes and at higher elevations. The black-oak (Quercus kelloggiz) forest, with stands of Chamaebatia foliolosa, Arctostaphylos viscida, and A. patula, forms a transition between the foothill com- munities and the mixed-conifer forests. In earlier classification schemes, the foothill zone was generally divided into oak savanna (corresponding with foothill woodland), chaparral (including chamise chaparral and much of the mixed-evergreen woodland), hardwoods 1981] PARSONS: FOOTHILL FIRE ECOLOGY 113 (included in the mixed-evergreen woodland), and black-oak forest. Because these community types were often the basis for past fire rec- ords, it has been necessary to use them during much of this report. FIRE HISTORY Presettlement. Fires ignited by lightning are an important element in the dry summer environment of the Sierra Nevada (Komarek, 1967). Unfortunately, the lack of adequate fire dating techniques for the foothill communities has often made it impossible to quantify the frequency or size of lightning fires in presettlement times. Additional pre-European ignitions came from the local aboriginal populations. For example, it is known that prior to the first coming of Europeans to the area in the late 1850’s, the Western Mono Indians used fire to assist in hunting, to promote growth of wild food crops, and to facilitate the collection of seeds in much of the study area (Lewis, 1973; Vankat, 1977). Both Reynolds (1959) and Lewis (1973) present evidence that most Indian burning was carried out in the fall. The influence of aboriginal man on the Kaweah River region was essentially terminated by 1865, when the last Western Mono left the area (Strong, 1964). Information available for a nearby mixed-conifer forest (Kilgore and Taylor, 1979) documents a decrease in fire frequency after 1870. This corresponds with the end of Indian occupation but is before fire suppression became effective. It establishes the importance of Indian ignitions in the presettlement forest. While it is impossible to document the extent to which the same is true for the lower elevation foothill communities, it is likely that a similar pattern holds. It is thus logical to assume that the pre-aboriginal fire frequency (lightning ignitions only) was less than in the 1800’s, but greater than exists today. While it may not be possible to distinguish fully between the relative fre- quency or ecological significance of Indian versus lightning fires, it is clear that both played an important role in determining the vegetation patterns found in the foothill zone. It is the policy of the National Park Service to include both lightning and Indian ignitions as part of the natural scene. Post-settlement. During the 1860’s, European settlers first moved into the foothills of the southern Sierra in the area that is now Sequoia National Park. In succeeding years, sheepherders increasingly used the area in moving their flocks to and from the high country. They commonly set fires in the fall while coming out of the mountains in order to clear brush and provide for new, more palatable growth the following spring (Strong, 1964; Vankat, 1977). Vankat (1977) con- cluded that much of this sheepherder burning can be viewed merely as an extension of aboriginal fires. The two cultures apparently burned “for much the same reasons—to favor certain plant species and to 114 MADRONO [Vol. 28 open the forests”. Lightning-ignited fires continued to burn during this period. Following the establishment of Sequoia National Park in 1890, a policy of suppressing all fires was implemented (Vankat, 1977). Due to limited funding and manpower and poor access to many areas, suppression did not become effective until the 1920’s. A policy of fire suppression has remained in effect for the foothill region to the present time. The consequences of this suppression policy are now the cause of considerable concern. The reduction of fire frequency has resulted in an increased fuel density and an abundance of old growth, especially in the chaparral communities (Parsons, 1976; Rundel and Parsons, 1979). Increased fuel accumulations make it increasingly difficult to extinguish fires that do get started. There has also been a loss of typical age-class and community-type mosaics and, in some cases, a change in species composition. For example, Vankat and Major (1978) have documented an increase in cover and density and a decrease in diver- sity in the chaparral communities as well as an increased cover and density of oak species in the woodland types. Recent. Although fire records for the study area before the late 1920’s are incomplete, they are sufficient to document the fact that fires caused by both man and lightning were frequent occurrences. For example, early park records show at least 37 lightning fires, 6 man-caused fires, and 22 fires of unknown origin occurring within the foothills of the Kaweah River drainage (an area of 28,366 ha) between 1891 and 1919. The great majority of these fires occurred during the dry summer months when the vegetation is highly flammable. Several (both lightning and man-caused) consumed more than a thousand hectares. Since the 1920’s, nearly complete records have been maintained of all fires occurring within the park. These records provide the only available data base for recent fire history in the foothill zone. Fire frequency and size, of course, have been greatly influenced by suppres- sion activities. Records available for 1920—1929 show at least ten fires of 12 ha or larger, but of undetermined origin, burning a total of 4189 ha within the study area (the records of smaller fires have been lost). In the 49 years from 1930 to 1978 a total of 105 lightning and 107 man-caused fires were recorded for the Kaweah drainage foothill zone. Lightning fires averaged 2.14 + 2.63 (S.D.; range: 0-10) per year while 2.18 + 2.00 (range: O—8) man-caused fires were recorded per year. There were 16 years with no lightning fires, 12 with no man- caused fires, and 4 with no fires at all. Lightning fires burned a total of 592 ha while man-caused fires burned 2811 ha. Most of the fires (91 percent) burned less than 3 ha. It should be noted that the lightning fires reported here represent only those ignitions that resulted in de- tectable fires. No doubt many lightning strikes are never detected. 1981] PARSONS: FOOTHILL FIRE ECOLOGY 1S TABLE 1. TOTAL AREA AND NUMBER OF LIGHTNING FIRES BY 305 m (1000 ft) ELEVATION CONTOURS IN THE ENTIRE KAWEAH RIVER DRAINAGE, SEQUOIA NATION- AL PARK, 1930-1978. Total area No. lightning No. fires Elevation (m) (ha) fires per 1000 ha <610 627 2 32 610-914 3134 1 0.3 915-1219 5877 4 OP 1220-1524 8267 18 Did 1525-1829 10,461 81 Mil 1830-2134 12,146 99 8.2 2135-2439 12,577 139 11.0 2440-2744 12,538 118 9.4 2745-3049 9795 70 (a >3050 7758 10 1.3 Furthermore, the increased efficiency of fire suppression capabilities in recent decades has limited the total area burned to well under the area that would have burned without suppression activities. Data on the frequency of lightning fires as a function of elevation within the entire Kaweah drainage (Table 1), with the exception of the lowest elevation where the sample area is small, show a steady increase in the number of fires as well as the number of fires per unit area, up to an elevation of 2440 m. This corresponds with the findings of Komarek (1967) and Keeley (1977) for the forest and chaparral regions of California and relates in large part to an increasing fre- quency of lightning strikes with elevation. The decrease in fires at the highest elevations relates primarily to decreasing fuel supplies. Where- as lightning fires most frequently ignite in the middle elevations, under hot, dry, summer conditions they sometimes burn downslope through the highly flammable brush and grasslands. It should be emphasized that while lightning strikes are relatively rare at the lower elevations, they do occur (Griffin, 1977). It is likely that occasional lightning fires that ignite below the park would burn into the study area were they not suppressed. When ignited under the proper conditions, few igni- tions are needed to burn large areas of highly flammable chaparral and oak woodland. The distribution of lightning fires by month for the study area is presented in Fig. 1. Most of the 32 recorded July fires occurred in the latter part of the month. The late-summer peak in lightning fires cor+ responds with the onset of maximum flammability as represented by decreasing foliage moisture content (Fig. 1). This timing corresponds with the period of maximum temperatures and minimal precipitation, conditions also favoring maximum burning potential. Thus, the high incidence of lightning fires during the late summer and early fall may 116 MADRONO [Vol. 28 FOLIAGE MOISTURE CONTENT ——— NUMBER OF LIGHTNING-CAUSED FIRES FOLIAGE MOISTURE CONTENT (%) Fic. 1. Mean number of lightning-caused fires by month for Kaweah River drainage below 1830 m, 1930-1978, and mean monthly foliage moisture content for Adenostoma fasciculatum within the study area. Foliage moisture determination followed Country- man and Dean (1979). be as much a function of weather conditions and vegetation flam- mability as of lightning-strike frequency. The monthly distribution of man-caused fires within the study area corresponds closely with that of lightning fires (Fig. 2). Again, the peak occurs during the period of hot, dry weather and low foliage moisture content when the vegetation is most likely to burn. The one significant difference from the distribution. of lightning fires is the rel- atively large number of man-caused fires in June. This difference may relate to the fact that electrical storms are rare in June but the period of high visitor use has already begun (Fig. 2). The strong correlation between the timing of peak visitor use and the incidence of man-caused fires emphasizes the potential danger from unwanted wildfires during the summer months. The park fire atlas, which maps all fires that burned more than 4 ha in the study area since 1920 (unpublished data), shows that much of the foothill zone of Sequoia National Park has not burned in at least 60 years. Biswell (1974) has expressed the opinion that this is 1981] PARSONS: FOOTHILL FIRE ECOLOGY 117 —— NUMBER OF VISITORS NUMBER OF MAN-CAUSED FIRES NUMBER OF VISITORS (x103) MONTH Fic. 2. Mean number of man-caused fires by month for the Kaweah River drainage below 1830 m, 1930-1978, and monthly 1978 visitation rate for Sequoia National Park. probably a considerably longer fire-free period than most of these low- elevation communities have experienced for many centuries and per- haps for thousands of years. Together with the evidence for buildup in both live and dead fuels in stands of increasing age, this indicates a serious over-abundance of dense, over-mature, highly flammable brush. It also suggests a lack of distinct age-class boundaries that could be effective fire breaks. Lack of such boundaries can hinder the development of a safe, effective program of prescribed burning (Phil- pot, 1974). By superimposing the map of fires larger than 4 ha on a vegetation map of the park we have calculated the area burned since 1920 within each vegetation type (Table 2). The greatest area burned was in chap- arral, with the next greatest being in hardwoods. When converted to a percentage of the community type available, the oak savanna and chaparral have had the greatest proportions burned. In all, 8360 ha or slightly less than 30 percent of the foothill zone has been burned at least once by either lightning or man-caused fire since 1920. In addi- tion, 1246 ha, primarily in chaparral, have been burned two or three times during this period (Table 2). Not included in Table 2 are four prescribed burns which have been ignited as part of the park’s fire management program (Parsons, 1977). These fires, which have all occurred since 1969, have burned a total 118 MADRONO [Vol. 28 TABLE 2. AREA OF EACH FOOTHILL (<1830 m) VEGETATION TYPE BURNED BY MAN-CAUSED AND LIGHTNING FIRES IN THE KAWEAH RIVER DRAINAGE, 1920-1978. Percents are based on area of vegetation type within the drainage below 1830 m (study area) only. Percent of Vegetation No. hectares vegetation No. hectares type burned type burned reburned Oak savanna 1073 46.8 194 Chaparral 4271 38.1 673 Hardwoods 1635 20.5 251 Black oak 351 19.3 38 Conifers 1030 13.9 90 Total 8360 25.1 1246 of 1099 ha in the study area. Of this, 48 percent has been in the chaparral. Most of the area of prescribed burns (557 ha) had previously burned since 1920. MANAGEMENT IMPLICATIONS Together with data on vegetation patterns, age class boundaries, biomass accumulation, flammability, and an understanding of fire’s role in the reproduction and succession of important plant and animal species, fire history data provide a basis for predicting the immediate and long term effects of any fire and so provide an essential basis for developing a fire-management program. Such information establishes criteria for simulating the natural fire regime by re-estab- lishing the seasonal and elevational distribution of historical fires. It also helps to establish which vegetation types and even which specific areas have gone the longest without being burned. Such information is valuable in setting priorities for future management actions. The foothill zone of Sequoia National Park, like much of the rest of the foothill zone of the southern Sierra Nevada, contains extensive areas of over-mature, highly flammable brush. When unplanned fires ignite, especially at the lower elevations, it is often essential that im- mediate suppressive action be taken. Otherwise highly destructive, uncontrollable wildfires may result. This threat is of special concern due to the location of giant sequoia groves immediately uphill from the foothill communities and in the path of potential conflagrations. I conclude that through a carefully planned prescribed burning pro- gram it will be possible to reduce the accumulated fuels and at the same time increase the number of distinct age classes and vegetation boundaries. This will not only restore more natural conditions but will facilitate suppressive action on future wildfires (Philpot, 1974). While the first several prescribed burns, for reasons of control, may need to 1981] PARSONS: FOOTHILL FIRE ECOLOGY 119 be in late fall or winter, once a more diverse vegetative and age-class mosaic has been created it should be possible to burn safely at the time of year when natural fires are known to have occurred. In all cases, prescribed burns must be carefully planned to minimize risks while at the same time accomplishing desired objectives. This requires a thorough understanding of the relationship of fire behavior to mois- ture content of foliage, biomass accumulation, and various weather parameters for each vegetation type. If carefully conducted, such a program should assure the continued survival of healthy foothill wood- land and chaparral communities in the southern Sierra Nevada. ACKNOWLEDGMENTS This research was supported by the National Park Service. Rick Hedlund provided invaluable assistance in summarizing past fire records. Phil Rundel, James Parsons, Jon Keeley, and James Griffin reviewed and helped to improve the final version of the manuscript. LITERATURE CITED BAKER, G. A., P. W. RUNDEL, and D. J. PARSONS. 1981. Ecological relationships of Quercus douglasi in the foothill zone of Sequoia National Park, California. Ma- drono 28:1-12. BISWELL, H. H. 1974. Effects of fire on chaparral. 7m Fire and ecosystems, T. T. Kozlowski and C. E. Ahlgren, eds., p. 321-364. Academic Press, New York. COUNTRYMAN, C. M. and W. A. DEAN. 1979. Measuring moisture content in living chaparral: a field user’s manual. USDA For. Serv. Gen. Tech. Rep. PSW-36. GRIFFIN, J. R. 1977. Oak woodland. Jn Terrestrial vegetation of California, M. G. Barbour and J. Major, eds., p. 383-415. Wiley-Interscience, New York. Heapy, H. F. 1977. Valley grassland. /n Terrestrial vegetation of California, M. G. Barbour and J. Major, eds., p. 491-504. Wiley-Interscience, New York. KEELEY, J. E. 1977. Fire-dependent reproductive strategies in Arctostaphylos and Ceanothus. In Proceedings of the symposium on the environmental consequences of fire and fuel management in Mediterranean ecosystems, p. 391-396. USDA For. Serv. Gen. Tech. Rep. WO-3. KILGORE, B. M. and D. TayLor. 1979. Fire history of a sequoia-mixed conifer forest. Ecology 60:129-142. KOMAREK, E. V. 1967. The nature of lightning fires. Proc. Tall Timbers Fire Ecology Conf. 7:5-41. Lewis, H. T. 1973. Patterns of Indian burning in California: ecology and ethnohistory. Ballena Press, Ramona, CA. PaRSONS, D. J. 1976. The role of fire in natural communities: an example from the southern Sierra Nevada, California. Environ. Cons. 3:91—99. 1977. Preservation in fire-type ecosystems. Jn Proceedings of the symposium on the environmental consequences of fire and fuel management in Mediterranean ecosystems. p. 172-182. USDA For. Serv. Gen. Tech. Rep. WO-3 PHILPOT, C. W. 1974. The changing role of fire on chaparral lands. In Proceedings of the symposium on living with the chaparral, p. 131-150. Sierra Club, San Francisco. REYNOLDS, R. D. 1959. Effect of natural fires and aboriginal burning upon the forests of the central Sierra Nevada. M.A. thesis. Univ. California, Berkeley. RUNDEL, P. R. and D. J. PARSONS. 1979. Structural changes in chamise (Adenostoma fasciculatum) along a fire-induced age gradient. J. Range Managem. 32:462—466. STRONG, D. H. 1964. A history of Sequoia National Park. Ph.D. Dissertation. Syracuse Univ. 120 MADRONO [Vol. 28 VANKAT, J. L. 1977. Fire and man in Sequoia National Park. Ann. Assoc. Amer. Geogr. 67:17-27. VANKAT, J. L. and J. Major. 1978. Vegetation changes in Sequoia National Park, California. J. Biogeogr. 5:377—402. (Received 2 Jun 1980; accepted 12 Nov 1980; revised version received 18 Dec 1980.) TAXONOMY OF PHACELIA SECT. MILTITZIA (HY DROPHYLLACEAE) RICHARD R. HALSE Department of Botany and Plant Pathology, Oregon State University, Corvallis 97331 ABSTRACT The species of Phacelia sect. Miltitzia are found in the Great Basin physiographic province of western North America and are distinguished from other members of Pha- celia by their yellow, marcescent corollas combined with transversely corrugated seeds. An artificial key is provided to the nine species and two varieties of sect. Miltitzia recognized in the present taxonomic treatment. One new species, Phacelia monoensis, is described and Phacelia submutica is reduced to P. scopulina var. submutica. Phacelia is the largest genus in the Hydrophyllaceae, consisting of 150-200 species. The genus has a wide distribution, the greatest num- ber and diversity of species being in western North America. Many of the species-groups now included in Phacelia were recognized as distinct genera in earlier classifications (Candolle, 1845). Section Mzl- titzia is one such group. The species belonging to sect. Miltitzza are small yellow-flowered annuals found in the arid regions in and around the Great Basin (Fig. 1). These species bloom primarily in the spring and characteristically grow in clay soils having a high pH and relatively high concentrations of soluble salts. They are usually found in the sagebrush-juniper or sagebrush-rabbitbrush communities. The first species of the Miltitzia group to be described was Eutoca lutea Hooker and Arnott (1840). However, these authors questioned the plant’s affinity with Eutoca because it possessed a yellow, mar- cescent corolla; they excluded it from Emmenanthe because of its pros- trate growth habit and lack of corolla scales. Candolle (1845) subse- quently placed this species in its own new genus, Muiltitzia. Gray (1857), while describing the second known member of the group (as Emmenanthe parviflora Gray), proposed that the plants be treated as the subgenus Miltitzia within the genus Emmenanthe. Heller (1912) and Brand (1913) raised subgenus Muiltitzza to the generic rank sug- gested by Candolle. J. T. Howell (1944a), while preparing a mono- graph of Phacelia sect. Euglypta, noted the close morphological sim- ilarity between it and Miultitzia and, as a result, transferred the Miltitzia species to the genus Phacelia. Miltitzia and sect. Euglypta have several traits in common. Both have plump, transversely corrugated seeds, and the lateral attachment MADRONO, Vol. 28, No. 3, pp. 121-132, 31 July 1981 122 MADRONO [Vol. 28 =—~__ oP. monoensis : a OP. lutea var. lutea P. glaberrima *P. lutea var. calva Es salina *P. scopulina var. scopulina i nsis : z OOS . scopulina var. submutica / © P, inundata %P. tetramera oP. adenophora Fic. 1. Distribution of Phacelia sect. Miltitzia. of the ovules to fleshy placentae in Muiltitzza is characteristic of the entire genus Phacelia (Howell, 1944b). The chromosome numbers also indicate the likeness of the two groups (Constance, 1963). In Miltitzia, n = 11, 12, and 13, whereas in sect. Euglypta, n = 11, 12, 13, and 23. On the other hand the principal characteristic Miltitzia and Em- menanthe have in common is the persistent, yellow corolla. In Em- menanthe the seeds are compressed and reticulate, the pendent ovules are basally attached to the wing-like margins of membranous placen- tae, and the chromosome complement of m = 18 is unique in the Hy- drophyllaceae. Miltitzza, as Howell (1944a) concluded, is best treated as a section of the genus Phacelia. METHODS Field observations and collections were made during the spring months of 1974-1978 throughout the range of the species of sect. Miultitzia. More than 600 herbarium specimens were examined during the course of this study. Measurements of vegetative and floral parts were made on herbarium specimens from CAS, CIC, DS, GH, ID, JEPS, K, MO, NY, ORE, OSC, POM, RENO, RM, RSA, UC, US, UTC, WILLU, WS, and WTU. These, together with the field obser- vations, form the basis of the morphological and distributional data. Floral buds for chromosome number determination were collected in the field and fixed in a modified Carnoy’s solution (4 chloroform: 3 ethanol: 1 glacial acetic acid, v/v/v). Acetocarmine squashes of an- thers were obtained by using the technique of Snow (1963). CYTOLOGY Cave and Constance (1947, 1950, 1959) determined the chromosome numbers for most of the species of sect. Miltztzia. Additional counts 1981] HALSE: PHACELIA SECT. MILTITZIA 123 TABLE 1. CHROMOSOME COUNTS FOR Phacelia sEcT. Miltitzia. All collection num- bers are those of the author. Vouchers are in OSC. P. adenophora J. T. Howell; n = 12 CA, Lassen Co., Termo, 1198, 1201. NV, Washoe Co., 1195. P. lutea (Hooker & Arnott) J. T. Howell var. lutea; n = 12 ID, Owyhee Co., Sand Basin, 1014. OR: Harney Co., Stinkingwater Pass, 1018, 1143; Lake Co., Plush, 1023, 1025; Malheur Co., Succor Creek, 1153, 1154, 1295, 1302; Sheaville, 1016; Leslie Gulch, 1157; Rockville, 1017. P. lutea (Hooker & Arnott) J. T. Howell var. calva Cronquist; n = 12 ID, Owyhee Co., 43 km sw. of Marsing, 1957. from the present study are presented in Table 1; these confirm the reports by Cave and Constance for the taxa concerned. The genus Phacelia shows great diversity in chromosome numbers and contains both polyploid and aneuploid series (Constance, 1963). One such aneuploid series is found in sect. Miltitzia: in Phacelia tetramera, n = 11; in P. adenophora, P. inundata, P. inyoensis, P. lutea, P. monoensis (cited as M. scopulina, Cave and Constance, 1959), and P. scopulina, n = 12; in P. glaberrima, n = 13. TAXONOMY PHACELIA Juss. sect. MILTITZIA (DC.) J. T. Howell, Leafl. W. Bot. 4:15. 1944.—Miultitzia DC., Prodr. 9:296. 1845.—Emmenanthe Bentham subg. Muiltitzza (DC.) Gray, Proc. Amer. Acad. Arts 10:328. 1875.—TyYPE: Phacelia lutea (Hooker & Arnott) J. T. Howell. Low, diffuse, prostrate to decumbent or ascending annuals from slender taproots, usually of alkaline habitats; stems 5-30 cm long; herbage densely hirsutulous to rarely glabrous, usually more or less purplish capitate-glandular; leaves entire to toothed or pinnately lobed, 0.5—4.0 cm long, 0.2—2.5 cm wide; flowers pedicellate, in simple or branched terminal scorpioid cymes; calyx divided nearly to the base, the lobes subequal; corolla yellow to whitish, frequently pur- plish-tinged with age, marcescent; corolla scales present or lacking; stamens included, subequal to unequal, equally inserted at base of corolla tube; hypogynous disk prominent or inconspicuous; style per- sistent, 2-cleft or subentire; seeds 7-25, transversely corrugate or striate. Key to Phacelia sect. Muiltitzia Corolla tube pubescent within, at least at base; filaments pubescent. Corolla 2.0—3.5(—4) mm long; style and branches 0.5—1.5 mm long; 124 MADRONO [Vol. 28 filaments 1.5—2.5 mm long............... 1. P. monoensis Corolla (3.5—)4—8 mm long; style and branches 1.5—3.0 mm long; filaments 2.5—4.5 mm long. ............. 2. P. adenophora Corolla tube glabrous within; filaments glabrous. Seeds faintly but definitely transversely striate, 18-25 per capsule; style and branches 0.5—1.2 mm long. ....... 3. P. inundata Seeds prominently transversely corrugated. Corolla 4-10 mm long, usually longer than the calyx; style and branches 2—4 mm long. Plants densely pubescent. .......... 4a. P. lutea var. lutea Plants subglabrous, or slightly glandular in the inflores- CONCE: 4255 25 apokiea ae eciwat 4b. P. lutea var. calva Corolla 1.3—4.0 mm long, if longer usually equalling the calyx; style and branches 0.2—2.0 mm long. Plants glabrous or nearly so; corolla subrotate. ............. eee ee ee ee er ree 5. P. glaberrima Plants densely pubescent; corolla tubular to campanulate. Flowers usually 4-merous; corolla 1.3—2.0 mm long. ...... A Giiee RU eee eG ee eis Ae ae ee 6. P. tetramera Flowers 5-merous; corolla 2—4 mm long. Style and branches 1.0—2.0 mm long; capsules with 9-15 seeds. Style pubescent 4% to all of its length; capsule apic- Wlatey eee 7a. P. scopulina var. scopulina Style glabrous except at base; capsule nearly or quite WiChOlt apiculavion: {x 4546-4. cote ee eee ee i nae eae 7b. P. scopulina var. submutica Style and branches 0.5—1.0 mm long; capsules with fewer than 11 or more than 15 seeds. Seeds 18-25 per capsule, corrugations 5-8. .......... Fee Bees shoo Sas a gee etna 8. P. inyoensis Seeds 7-10 per capsule, corrugations 9-13. .......... CR eee te ee eee ee ee var e 9. P. salina 1. Phacelia monoensis Halse, sp. nov. Floribus 5-meris; segmentis calycis per anthesis 2-4 mm longis, ad maturitatem 4—6 mm longis; corolla luteola, 2-4 mm longa, extus et intus pubescenti; filamentis pubescentibus, 1.5—2.5 mm longis; stylo cum ramis 0.5—1.5 mm longo; ovulis 7-10; capsula 2.5—4.0 mm longa; seminibus 1.1—1.7 mm longis, manifeste transverse corrugatis, rugis 8-11. Annual herb; stems few or several, prostrate; herbage hirsutulous, more or less purplish capitate-glandular; leaves entire toothed or pin- nately lobed; flowers 5-merous; calyx segments in flower 2—4 mm long, in fruit 4-6 mm long; corolla tubular to campanulate, yellow, 2-4 mm 1981] HALSE: PHACELIA SECT. MILTITZIA 125 long, pubescent externally and internally; corolla scales obsolete; fil- aments pubescent, 1.5—2.5 mm long; style and branches 0.5—1.5 mm long, style pubescent to midlength; ovules 7-10; hypogynous disk prominent; capsule 2.5—4.0 mm long, apiculate, pubescent; seeds 1.1- 1.7 mm long, transversely corrugate, corrugations 8-11; n = 12. Type: USA, CA, Mono Co., in red clay, Mormon Ranch, 14 km (8.5 mi) sw. of Bodie, 30 Jun 1945, Alexander and Kellogg 4346 (Ho- lotype: UC!; isotypes: CAS! DS! NY! POM! RSA! US! UTC! WS! WTU)). Distribution. Alkaline mountain meadows of central Mono Coun- ty, California and adjacent Nevada; flowering June-July. Phacelia monoensis is probably closely related to P. adenophora because of the similarity in floral pubescence and possession of a con- spicuous hypogynous disk. The two species differ in the size of the floral parts, the presence of corolla scales and the number of ovules; they occupy quite different habitats. 2. PHACELIA ADENOPHORA J. T. Howell, Leafl. W. Bot. 4:15. 1944.— Emmenanthe glandulifera Torrey ex Watson, Bot. U.S. Geol. Explor. 40th Parallel. 257. 1871.—Miultitzia glandulifera (Watson) Heller, Muhlenbergia 8:20. 1912.—Typre: USA, NV, Virginia Mts., Jul, Watson 885 (Holotype: GH!; isotypes: NY! US!). Not Phacelia glandulifera Piper, Contr. U.S. Natl. Herb. 11:472. 1906. Miltitzia glandulifera var. californica Brand, Univ. Calif. Publ. Bot. 4:224. 1912.—-TypE: USA, CA, Lassen Co., Madeline Plains, Jun 1898, Bruce 2135 (Holotype: UC!). Stems few or several, prostrate to ascending, pubescent, capitate glands present or lacking; leaves pinnately lobed or divided, rarely entire or merely toothed, hirsutulous; flowers 5-merous; calyx segments in flower 2—5 mm long, in fruit 4-7 mm long; corolla campanulate, yellow or the lobes more or less purplish-tinged, (3.5—)4—8 mm long, pubescent externally, corolla tube sparsely to densely pubescent within; corolla scales present; filaments pubescent, unequal, 2.5—4.5 mm long; style and branches 1.5—3.0 mm long, style pubescent 4-2 its length; ovules 6-15; hypogynous disk prominent; capsule 2.5—4.5 (—6) mm long, apiculate, pubescent; seeds 1.0—1.6 mm long, transverse- ly corrugate, corrugations 8-12; m = 12. Distribution. Plains and slopes of northeastern California, western Nevada, and southeastern Oregon; flowering April—July. Phacelia adenophora is a moderately variable species but is well characterized by the conspicuous pubescence on the filaments and within the corolla tube. In central-western Nevada the plant is rather slender and delicate; to the northward it intergrades with a larger, coarser form that Brand (1912) called Miltitzia glandulifera var. cal- 126 MADRONO [Vol. 28 ifornica. However, var. californica is not sufficiently distinct to be recognized taxonomically because the variation between the two forms is continuous and the characters Brand used to distinguish the variant are found throughout the range of the species. Phacelia adenophora is closely related to P. lutea and has been treated as the same species (Jepson, 1943). Some hybridization may occur between these two taxa in northern Nevada and adjacent Or- egon. In these areas plants have been found in which the filaments and the inside of the corolla tube range from glabrous to subglabrous to pubescent. Nevertheless, it seems best to retain these taxa as sep- arate species, because in P. adenophora usually either the filaments or corolla tube are at least slightly pubescent and plants can therefore be referred readily to one or the other of the two species. 3. PHACELIA INUNDATA J. T. Howell, Leafl. W. Bot. 4:15. 1944.— Emmenanthe parviflora Gray, U.S. Pacific R.R. Reports 6:84. 1857.—Miltitzia parviflora (Gray) Brand, Das Pflanzenr. IV. 251:131. 1913.—Typre: USA, OR, Klamath Lake, Newberry s.n. (Holotype: GH!). Not Phacelia parvifiora Pursh, Fl. Amer. Sept. 1:140. 1814; nor Phacelia parviflora Phil., Anales Univ. Chile 90:226. 1895. Stems several, prostrate to ascending, pubescent, glandular; leaves subentire to pinnately lobed, hirsutulous; flowers 5-merous; calyx seg- ments in flower 3—4 mm long, in fruit 5.5—8.0(—10) mm long; corolla tubular to campanulate, yellow, 3-5 mm long, pubescent externally, glabrous internally; corolla scales present or obsolete; filaments gla- brous, 1.5—3.0 mm long; style and branches 0.5—1.2 mm long, style pubescent half to all of its length; ovules 18—25(—30); hypogynous disk prominent; capsule 4—7 mm long, apiculate, pubescent; seeds 1.1-1.8 mm long, tending to be flattened, transversely striate, striations 12— 14; = 12. Distribution. Dried edges of alkali lakes and sinks, from south central Oregon to northeast California and adjacent Nevada; flower- ing May-July. 4. PHACELIA LUTEA (Hooker & Arnott) J. T. Howell, Leafl. W. Bot. 4:15. 1944.—For synonymy and typifications see the varietal headings. Stems several, prostrate to decumbent, glabrous to hirsutulous, cap- itate-glandular to eglandular; leaves usually entire or some coarsely toothed to pinnately lobed, glabrous to pubescent; flowers 5-merous; calyx segments in flower 2.5—6.0 mm long, in fruit 4.5—10.0 mm long; corolla tubular to campanulate, yellow, (3.5—)4—10(—11) mm long, gla- brous or pubescent externally, glabrous internally; corolla scales pres- 1981] HALSE: PHACELIA SECT. MILTITZIA 27 ent or obsolete; filaments glabrous, 2.5—5.0 mm long; style and branch- es (1.5—)2—4 mm long, style glabrous or pubescent; ovules 7—15(—20); hypogynous disk prominent; capsule 3—7 mm long, apiculate; seeds 1— 2 mm long, transversely corrugate, corrugations 9-12; m = 12. 4a. PHACELIA LUTEA var. LUTEA—Eutoca lutea Hooker & Arnott, Bot. Beech. Voy. 373. 1840.—Multitzia lutea (Hooker & Arnott) DC., Prodr. 9:296. 1845.—Emmenanthe lutea (Hooker & Arnott) Gray, Proc. Amer. Acad. Arts 10:328. 1875.—Phacelia lutea var. typica, Proc. Calif. Acad. Sci. ser. 4, 25:364. 1944.—TypE: USA, “Snake Fort, Snake Country, California” [Idaho], 1837, Tolmie s.n. (Holotype: K!). Phacelia lutea var. purpurascens J. T. Howell, Proc. Calif. Acad. Sci. ser. 4, 25:365. 1944.—Typre: USA, OR, Grant Co., Hum- phrey’s, 30 Apr 1925, Henderson 5092 (Holotype: CAS!; isotypes: DS! GH! ORE!). Herbage densely hirsutulous, from eglandular to densely purplish capitate-glandular, especially in the inflorescence; corolla pubescent externally; style hairy near the base to above the middle. Distribution. Alkaline areas, usually clay and ash slopes and banks, from central Oregon east to southwestern Idaho and south to northwestern Nevada; flowering April—July. Phacelia lutea is an extremely variable species. The entire to slightly toothed leaves of typical P. lutea in northeastern Malheur County, Oregon, intergrade to the pinnately lobed leaves of P. lutea in Lake County. Style pubescence and length are also variable traits, varying as much on an individual plant as between plants. Because Howell (1944b) used these traits to distinguish var. purpurascens, it seems best to treat this taxon as part of the variable var. lutea. 4b. PHACELIA LUTEA var. CALVA Cronquist, Vasc. pls. Pacific Northw. 4:168. 1959.—TypEe: USA, ID, Owyhee Co., roadside bank alongside U.S. Hwy. 95, 6.4 km (4 mi) ne. of the Idaho— Oregon state line, 12 Jun 1946, Maguire and Holmgren 26386 (Holotype: NY!; isotypes: CAS! DS! GH! MO! UC! US! WS)). Herbage essentially wholly glabrous, or slightly purplish-glandular in the inflorescence; corolla glabrous externally; style glabrous. Distribution. Northern Owyhee County, Idaho, adjacent Malheur County, Oregon, and northwestern Nevada; flowering May—June. This variety is distinguished from var. lutea only by its glabrous condition. Mixed populations of var. lutea and var. calva have been found and in at least some of these populations there is intergradation in the amount of pubescence, plants ranging from glabrous to sub- glabrous to moderately pubescent. 128 MADRONO [Vol. 28 5. PHACELIA GLABERRIMA (Torrey ex Watson) J. T. Howell, Leafl. W. Bot. 4:15. 1944.—Emmenanthe glaberrima Torrey ex Watson, Bot. U.S. Geol. Explor. 40th Parallel. 257. 1871.—Miltitzia gla- berrima (Watson) Brand, Das Pflanzenr. IV. 251:131. 1913.— Type: USA, NV, Reese Valley, Jul 1868, Watson 886 (Lectotype: GH!; isolectotypes: NY! US!; syntype: USA, NV, Humboldt Mts., Humboldt Sink, May 1868, Watson 886 GH! NY! UC!). Stems several, decumbent, glabrous; leaves entire to subentire to rarely lobed, glabrous or with a few hairs on the margins and apex; flowers 5-merous; calyx segments in flower 2.5—3.0 mm long, in fruit 4—6 mm long, glabrous or with a few hairs at apex; corolla subrotate, glabrous, 2.5—4.0 mm long; corolla scales lacking; filaments 2.5—3.0 mm long, glabrous; style and branches 1—2 mm long, glabrous; ovules (4—)7—-10; hypogynous disk prominent; capsule 3-5 mm long, glabrous or with a few scattered hairs at apex, apiculate; seeds 1.1-1.5 mm long, transversely corrugate, corrugations 8-12; ” = 13. Distribution. Localized endemic on alkaline clay bluffs or alkaline sinks of central Nevada; flowering May-July. 6. PHACELIA TETRAMERA J. T. Howell, Leafl. W. Bot. 4:16. 1944.— Emmenanthe pusilla Gray, Proc. Amer. Acad. Arts 11:87. 1876.—Miltitzia pusilla (Gray) Brand, Das Pflanzenr. IV. 251:132. 1913.—TypE: USA, NV, Steamboat Springs, May, Watson 878, in part (Lectotype: GH!; syntype: USA, nw. Ne- vada, May 1875, Lemmon s.n. GH! UC!). Not Phacelia pusilla Buckl., Amer. J. Sci. 45:172. 1843; nor Phacelia pusilla Torrey ex Watson, Bot. U.S. Geol. Explor. 40th Parallel. 253. 1871. Miltitzia pusilla var. flagellaris Brand, Das Pflanzenr. IV. 251:132. 1913.—TyYprE: USA, OR, sterile alkaline soil of the Malheur Riv- er, 20 Jun 1898, Cusick 1946 (Lectotype: US!; isolectotypes: GH! MO! ORE! RM! UC!; syntype: USA, OR, Union Co., dry alkaline soil, 1879, Cusick 758 GH! NY! US)). Stems several, prostrate to decumbent, somewhat mat forming, pu- bescent, glands present or absent; leaves entire to shallowly pinnately lobed, pubescent; flowers 4-merous, rarely 5-merous; calyx segments in flower 1.5—3.0 mm long, in fruit 3.5—4.5 mm long; corolla cam- panulate, whitish, 1.3—2.0 mm long, pubescent externally, glabrous internally; corolla scales lacking; filaments glabrous, 1.0—1.5 mm long; style 0.2—0.4 mm long, the branches indicated only by an emargination at the apex of the style, glabrous; ovules 12—24; hypogynous disk inconspicuous; capsule 2.5—4.0 mm long, shortly apiculate, pubescent; seeds 0.7-1.0 mm long, transversely corrugate, corrugations 6—9; 2n = 22. Distribution. Alkaline flats, washes, and meadows of Nevada, adjacent California, eastern Oregon and north central Utah; flowering May-June. 1981] HALSE: PHACELIA SECT. MILTITZIA 129 This species is unique in Phacelia in being 4-merous. Howell (1944b) writes that the reduction from the usual 5-merous condition seems to be due to the suppression of one part of the adroecium, calyx, and corolla rather than the union of adjacent parts. 7. PHACELIA SCOPULINA (A. Nelson) J. T. Howell, Leafl. W. Bot. 4:16. 1944.—For synonymy and typifications see the varietal headings. Stems few to several, prostrate to ascending, hirsutulous, sparsely glandular; leaves entire to toothed to pinnately lobed, pubescent; flow- ers 5-merous; calyx segments in flower 2.5—4.0 mm long, in fruit 5—8 (—10) mm long; corolla tubular to campanulate, yellow, becoming pur- plish-tinged, 3—4(-—5) mm long, pubescent externally, glabrous inter- nally; corolla scales inconspicuous or obsolete; filaments 2-3 mm long, glabrous; style and branches 1-2 mm long, style pubescent only at base to all of its length; ovules 9-15; hypogynous disk inconspicuous; capsule 3.5—6.0 mm long, pubescent, apiculate or not; seeds 1-2 mm long, transversely corrugate, corrugations 9-12. 7a. PHACELIA SCOPULINA var. SCOPULINA—Emmenanthe scopulina A. Nelson, Bull. Torrey Bot. Club 25:380. 1898.—Multitzia sco- pulina (A. Nelson) Rydberg, Bull. Torrey Bot. Club 40:479. 1913.—Miultitzia lutea var. scopulina (A. Nelson) Brand, Das Pflanzenr. IV. 251:131. 1913.—Phacelia lutea var. scopulina (A. Nelson) Cronquist, Vasc. pls. Pacific Northw. 4:168. 1959.— Type: USA, WY, Sweetwater Co., Green River, 31 May 1897, Nelson 3056 (Lectotype: RM!; isolectotypes: GH! MO! NY! US!; syntype: USA, WY, Sweetwater Co., Green River, 30 May 1897, Nelson 3026 RM! US!). Fruiting calyx 5—7 mm long; style and branches 1—2 mm long, pu- bescent 4 to all of its length; capsule apiculate; m = 12. Distribution. Alkaline flats and slopes from southwestern Wyo- ming to central Nevada, north to southeastern Oregon and adjacent Idaho, disjunct in southwest Montana; flowering May-July. Phacelia scopulina has been considered a variety of P. lutea because in eastern Oregon, where their ranges overlap, there are plants which may appear at first to be intermediate. The two species are, however, quite distinct; individual plants can always be referred to one or the other species after examination of the critical characters. No indication of hybridization has been found. 7b. Phacelia scopulina var. submutica (J. T. Howell) Halse, stat. nov.—Phacelia submutica J. T. Howell, Proc. Calif. Acad. Sci. ser. 4, 25:370. 1944.—-TyPE: USA, CO, Mesa Co., DeBeque, 19 May 1911, Osterhout 4458 (Holotype: Accession Number 163032, RM!; isotype: RM!). 130 MADRONO [Vol. 28 Fruiting calyx 6—8(-10) mm long; style and branches 1.0—-1.5 mm long, pubescent at the base; capsule not apiculate or insignificantly so. Distribution. Local endemic on clay knolls in Mesa County, Col- orado, near DeBeque, and disjunct along the Little Colorado River near Winslow, Arizona; flowering May—June. This taxon is not well enough differentiated to deserve species rec- ognition, but it does merit varietal status. The primary characters by which it is separated from var. scopulina are the pubescence on the style and the lack of apiculation of the capsule. The amount of pu- bescence on the style is a variable character. The apiculation of the capsule is the best character distinguishing the two taxa, although some intergradation is indicated; in var. submutica the style base on the capsule may be muticous and in var. scopulina the apiculation may be very small. A collection of this variety from Arizona forms the basis for the reported occurrence in Arizona of P. glaberrima (Howell, 1944b). When the plant was collected by Newberry on the Ives’ Expedition it was misidentified as Eutoca (=Nama) aretioides. When the type description of P. glaberrima was written, Newberry’s plant was iden- tified as that species. As far as is known this is the only collection from Arizona of any species in sect. Miltitzia. 8. PHACELIA INYOENSIS (Macbride) J. T. Howell, Leafl. W. Bot. 4:16. 1944.—Miultitzia inyoensis Macbride, Contr. Gray Herb., new series, 49:41. 1917.—TyPE: USA, CA, Inyo Co., foothills w. of Bishop, 23 May 1906, Heller 8324 (Holotype: GH!; isotypes: DS! MO! NY! US!). Stems few to several, ascending to erect, pubescent, capitate-glan- dular; leaves entire to pinnately few-lobed, pubescent, glandular; flowers 5-merous; calyx segments in flower 2—3 mm long, in fruit 3.5— 4.5(—6) mm long; corolla tubular to campanulate, pale yellow, 2-3 mm long, pubescent externally, glabrous internally; corolla scales lacking; filaments glabrous, 1.5—2.5 mm long; style and branches 1 mm long, style glabrous, or with a few hairs at base; ovules 16—21(-—27); hy- pogynous disk relatively prominent; capsule 3—4 mm long, apiculate, pubescent; seeds 0.5—1.0 mm long, transversely corrugate, corruga- tions 5-8; nm = 12. Distribution. Alkaline meadows in northern Inyo and adjacent Mono County, California; flowering May-July. 9. PHACELIA SALINA (A. Nelson) J. T. Howell, Leafl. W. Bot. 4:16. 1944.—Emmenanthe foliosa M. E. Jones, Zoe 4:278. 1893.—M1zI- titzia foliosa (M. E. Jones) Brand, Das Pflanzenr. IV. 251:131. 1913.—Type: USA, UT, Tooele Co., Deep Creek Valley above Furber, 8 Jun 1891, Jones s.n. (Holotype: Accession Number 1981] HALSE: PHACELIA SECT. MILTITZIA 131 72887, POM!; isotypes: CAS! DS! MO! POM! UC! US!). Not Pha- celia foliosa Phil., Anales Mus. Nac. Chile. 53. 1891. Emmenanthe salina A. Nelson, Bull. Torrey Bot. Club 25:381. 1898.—Miultitzia salina (A. Nelson) Rydberg, Bull. Torrey Bot. Club 40:479. 1913.—TyYPE: USA, WY, Sweetwater Co., Bitter Creek, 2 Jun 1897, Nelson 3105 (Holotype: RM!; isotypes: CAS! GH! MO! NY! US!). Stems few to several, prostrate to ascending, pubescent, capitate- glandular; leaves entire to pinnately lobed, hirsutulous; flowers 5-mer- ous; calyx segments in flower 2—3(—4) mm long, in fruit 4—5(-6) mm long; corolla tubular, yellow 2—3(—4) mm long, pubescent externally, glabrous internally; corolla scales present or absent; filaments gla- brous, 1.0—1.5(—2.5) mm long; style and branches 1 mm long, style pubescent at base only or rarely pubescent to midlength; ovules 7-10; hypogynous disk prominent; capsule 3-4 mm long, apiculate, pubes- cent; seeds 1.1—1.5(—2) mm long, transversely corrugate, corrugations 9-13. Distribution. Alkaline flats and clay slopes in southern Wyoming, western Utah, and central and eastern Nevada; flowering May—June. Phacelia salina has been considered nothing more than a form of P. scopulina by Cronquist (1959) and even Howell (1944b) suggested that it might be an ecologically specialized variant of that species with smaller parts, but its characters are quite stable over its entire range and show no overlap with P. scopulina. Its nearest relative, however, is undoubtedly P. scopulina. ACKNOWLEDGMENTS I thank the curators and staffs of the herbaria listed in Methods for loans of specimens. I am especially grateful to Dr. Kenton L. Chambers and LaRea D. Johnston for their guidance and help in the completion of this study and, together with Dr. Frank Smith, for their critical comments on this manuscript. This study was part of my Ph.D. Dis- sertation research at Oregon State University, Corvallis. LITERATURE CITED BRAND, A. 1912. Die Hydrophyllaceen der Sierra Nevada. Univ. Calif. Publ. Bot. 4:209-227. . 1913. Hydrophyllaceae. In A. Engler, Das Pflanzenreich. IV. 251 (Heft 59):1- 210. Wilhelm Englemann, Leipzig. CANDOLLE, A. P. DE. 1845. Hydrophyllaceae. In Prodromus Systematis Naturalis Regni Vegetabilis 9:287-301. Fortin, Masson et Sociorum, Paris. CAVE, M.S. and L. CONSTANCE. 1947. Chromosome numbers in the Hydrophyllaceae. II. Univ. Calif. Publ. Bot. 18:449-465. . 1950. Chromosome numbers in the Hydrophyllaceae. IV. Univ. Calif. Publ. Bot. 23:363-382. . 1959. Chromosome numbers in the Hydrophyllaceae. V. Univ. Calif. Publ. Bot. 30:233-258. CONSTANCE, L. 1963. Chromosome number and classification in the Hydrophyllaceae. Brittonia 15:273-285. 132 MADRONO [Vol. 28 CRONQUIST, A. 1959. Hydrophyllaceae. In C. L. Hitchcock, A. Cronquist, M. Own- bey, and J. W. Thompson. Vascular plants of the Pacific Northwest 4:145-175. Univ. Washington Press, Seattle. Gray, A. 1857. Hydrophyllaceae. Jn U.S. Pacific R.R. Reports 6:84—-85. U.S. War Dept., Washington. HELLER, A. A. 1912. New combinations. VII. Muhlenbergia 8:20. HooKeER, W. J. and G. A. WALKER-ARNOTT. 1840. The botany of Captain Beechey’s voyage. Henry B. Bohn, London. HowELL, J. T. 1944a. A reconsideration of the genus Muiltitzia. Leafl. W. Bot. 4:12-16. 1944b. A revision of Phacelia section Muiltitzia. Proc. Calif. Acad. Sci. ser. 4, 25:357—376. JEPSON, W. L. 1943. Flora of California 3(2):281—282. Assoc. Students Store, Univ. California, Berkeley. SNow, R. 1963. Alcoholic hydrochloric acid-carmine as a stain for chromosomes in squash preparations. Stain Tech. 38:9-13. (Received 11 Apr 1980; accepted 29 Oct 1980; revised version received 18 Dec 1980.) ALBERT M. VOLLMER: A MEDICAL DOCTOR WHO LOVED LILIES IRA L. WIGGINS Department of Biological Sciences, Stanford University, Stanford, CA 94305 Albert M. Vollmer was born May 14, 1896 in San Diego, California, attended Pomona College as a premedical student during the academic year of 1916-17, served in the U.S. Naval Reserves in 1917 and 1918, and returned to Pomona for the academic years 1918 through 1920. He then entered the University of Pennsylvania School of Medicine and received the M.D. degree in 1924. He was an intern at St. Luke’s Hospital in San Francisco in 1924-25, was at Sloane Hospital for Women in 1925-26 and took the National Board examinations in 1926 and served as Visiting Obstetrician and Gynecologist at the San Fran- cisco Hospital in 1926-27. Later he was Chief Obstetrician and Gyne- cologist at St. Luke’s Hospital and was on the staff of Franklin Hos- pital in San Francisco, and Associate Clinical Professor of Obstetrics and Gynecology at the University of California. He was a member of the American Medical Association and of the Pacific Coast Society of Obstetricians and Gynecologists. By the early 1930’s “Mike”, as he was known among his friends, had developed a keen interest in the native lilies of California; and he had spent many weekends and vacation periods scouring the state and adjacent Oregon in search of stands of lilies, from which he col- lected bulbs and seeds to propagate in his San Francisco home garden. I met him through the kind offices of Dr. Philip Leighton, Professor of Chemistry at Stanford, and from 1937 onward for a bit more than a decade accompanied Mike on field trips throughout California, four lengthy trips into Baja California, and one into northern Sonora and southern Arizona. Mike corresponded with many others who held intense interests in lilies, both in the United States and abroad. Among these correspon- dents was Dr. Samuel L. Emsweller, who was Principal Horticulturist in Charge of Ornamental Plants in the experimental gardens of the U.S. Department of Agriculture at Beltsville, Maryland. Mike was instrumental in getting Emsweller to visit California in search of lily seeds and bulbs on four or five different occasions. Mike also collected additional seeds and bulbs, which he forwarded to Beltsville in sup- port of research there on propagation and culturing of these plants MADRONO, Vol. 28, No. 3, pp. 133-135, 31 July 1981 134 MADRONO [Vol. 28 and of investigations of the diseases that frequently made it very dif- ficult to grow certain species in gardens and greenhouses. Mike was very careful to avoid digging large numbers of bulbs and, whenever possible, preferred to collect seed instead of, or in addition to, a small number of bulbs. He was acquainted with Carl Purdy and obtained some material from that dealer, but disapproved of Purdy’s practice of digging large numbers of bulbs of various liliaceous plants. More than once Mike made three or four separate trips to a particular locality in order to obtain mature seeds from a threatened stand of lilies rather than dig any bulbs whatever. Furthermore, he took pic- tures in color with both still and moving picture cameras in order to preserve information about habitats of the plants. In commenting upon the work Emsweller was carrying forward, Mike wrote, “Anyone who has tried to grow the California lilies has discovered that some are difficult and fail to become established. Dr. Emsweller has had this experience, and for him this was a challenge. The first step in solving this problem was to get firsthand information, resulting in field trips, to see the lilies growing in their native habitats, listing the associated vegetation, and terrain in which they grow, rain- fall, variations in temperature, growing season, collections of samples, culturing the soil, digging bulbs, collecting seed and raising them from seed. He is the only person who has seen all the American lilies grow- ing in their native habitat”. Mike was instrumental in helping Emsweller find several rare Cal- ifornia lilies growing in the field, and in returning later to collect seed after they had found only plants in bud or with immature capsules. He carried on this field work while serving an active practice from his office on Post Street in San Francisco and attending patients in the hospital! When my appointment as Scientific Director of the Arctic Research Laboratory took me away from Stanford for several years, Mike as- sumed active support of the work of a graduate student at Stanford, Mr. Lawrence Beane, who had planned to work beyond his Master’s degree, but circumstances made this plan impracticable. However, Larry continued to hold a keen interest in lilies, and he and Mike made a number of extensive field trips to isolated localities in the state so Larry could see the plants under natural conditions and paint them from fresh material. As an outcome of these expeditions Lawrence Beane published a paper in which he and Vollmer jointly named two new species of Lilium, and Larry made three new combinations, supplied a new name for a previously published variety, and recognized a specific name proposed in 1930 by Marcus E. Jones. The last entity had been considered no more than a synonym by most California botanists (Beane, 1955). 1981] WIGGINS: ALBERT M. VOLLMER 135 Dr. Vollmer also collected herbarium specimens of most California lilies that he had originally collected for propagation, and sent much material to the California Academy of Sciences, the University of Cal- ifornia, Stanford University, and to the U.S. National Herbarium in Washington, D.C. He collected and cultivated representatives of such genera as Calochortus and Fritillaria and of several species within the Amaryllidaceae. Dr. Vollmer published several papers in the Lily Yearbook and in garden journals. In these papers he contributed substantially to knowl- edge about geographical distribution, soil requirements, and flowering periods of our native lilies (Vollmer, 1954, 1956, 1959). The 1956 paper was presented in person, along with colored moving pictures of many of California’s lilies, before the Royal Horticultural Society in London. It seems appropriate that recognition should be given to the excellent field work carried on by Dr. Vollmer, to the support he gave so gen- erously to professional botanists, and for his devotion to saving the native stands of lilies in our state. I wish that such recognition might have been provided before Mike’s death in the spring of 1977, a few weeks before his 82nd birthday. He was a staunch friend who sup- ported my field operations through a decade, and took me to several out-of-the-way lily localities in California. I thank Dr. John H. Thomas for suggesting that I prepare this tardy tribute to Albert M. Vollmer, an enthusastic student of native liliaceous plants, and a keen observer of their characteristics. LITERATURE CITED BEANE, LAWRENCE. 1955. Some undescribed lilies from the Pacific Coast and a pre- liminary revision of the Southern California species formerly associated with Lilium humboldtii. Contr. Dudley Herb. 4(8):355—366. VOLLMER, ALBERT M. 1954. Hunting lilies in California. Roy. Hort. Soc. Lily Year- book, London. 1955:53—64. . 1956. California lilies. Roy. Hort. Soc. Lily Yearbook, London. 1957:86—93. . 1959. Dr. Samuel L. Emsweller. Roy. Hort. Soc. Lily Yearbook, London. 1960:9-10. (Received 28 Dec 1980; accepted 5 Jan 1981.) FIVE NEW SPECIES OF MEXICAN ERIGERON (ASTERACEAE) Guy L. NESoM Department of Biology, Memphis State University, Memphis, TN 38152 ABSTRACT Five new species of Erigeron from northern Mexico are described—E. unguiphyllus, E. cuatrocienegensis, E. wellsii, E. stanfordii, and E. solisaltator. The first two are probably obligate gypsophiles; the second two are from the region of Pena Nevada in southeastern Nuevo Leon and west-central Tamaulipas; the last is from northeastern Chihuahua; all are narrow endemics. A new name, E. gypsoverus, is proposed for a previously described gypsophilic Erigeron. Continuing studies of Erigeron have brought to light five new species. Two of them grow in the region to be covered by the Chi- huahuan Desert Flora, currently in preparation by M. C. Johnston and J. Henrickson. Two of these (from Coahuila and from San Luis Potosi) appear to be gypsophilic and are discussed with relation to other known Erigeron gypsophiles from northern Mexico. Two others apparently are restricted to the relatively small area of Pena Nevada in west-central Tamaulipas and southeastern Nuevo Leon and join a number of other endemics known from there. The distributions of the new species, as well as of several of their possible close relatives, are mapped in Fig. 1. A new name is proposed for a previously described species that may be related to the new gypsophiles. Erigeron wellsii Nesom, sp. nov. Erigeron scaposus DC. affinis aliquantum, differt praecipue rhizo- mate crasso fibris aliquantum carnosus, foliis caulinis non-amplectens, et corollis radii ligulis latioribus non circinatis ad maturitatem (Fig. 2A). Perennials with long and thick fibrous roots, from a thick, often horizontal rhizome 0.5—4.0 cm long; caudex simple, producing 1-3 (—4) upright, monocephalous stems. Stems 15—31 cm tall, simple, mod- erately pubescent with retrorse, closely to loosely appressed or some- times spreading, extremely thin, twisted trichomes 0.8—2.2 mm long. Basal leaves in a persistent rosette, 1.8-6.5 cm long, blades 0.8-2.8 cm wide, obovate, with 2—5 pairs of crenate to crenate-serrate teeth, attenuate to petiole that is 4—-% as long as leaf, with base usually purplish; cauline leaves 6-10, alternate, sessile, not clasping, sharply MADRONO, Vol. 28, No. 3, pp. 136-147, 31 July 1981 1981] NESOM: MEXICAN ERIGERON loa . Wellsii . sStanfordii -unguiphyllus - 9ypsoverusS « pinkavii . Cuatrocienegensis . solisaltator . Sp. Fic. 1. Distribution of Evigeron species in northeastern Mexico. reduced in size upward, the lower narrowly oblong to oblanceolate, to 23 mm long and 4 mm wide, the upper linear bracts; leaves mod- erately pubescent with erect to ascending trichomes, margins eciliate. Heads 1—3(—4) on peduncles 17—85 mm long; involucres hemispheric, 138 MADRONO [Voi. 28 1981] NESOM: MEXICAN ERIGERON 139 14-19 mm wide (pressed); phyllaries in 3—4 equal to unequal series, probably reflexing after release of achenes, lanceolate to elliptic-lan- ceolate, often purplish, inner 7.2—-9.5 mm long, 0.8-1.2 mm wide, when unequal outermost 42-34 as long as inner, evenly thin or some- times with 2 lateral thickenings near base, sparsely pubescent with thin trichomes; receptacles not observed. Ray flowers 70-135 in 2-3 series, corollas white with broad lilac midstripe, drying white to com- pletely lilac, 10.5-14.0 mm long, 0.9-1.5 mm wide, not reflexing or curling with maturity. Disc corollas tubular to narrowly funnelform, constricted in lower %4-—'4, not indurated, 3.5—4.5 mm long; style branches 0.7—0.9 mm long, including the shallowly triangular to shal- lowly deltate collecting appendages 0.1-—0.3 mm long. Achenes 2.0— 2.2 mm long, 0.8 mm wide, with 2 thin, orange ribs, sparsely strigose; carpopodium 5-8 cells high; pappus of ray and disc achenes similar, of (20—)25—30 very slender bristles ca. °/6 the disc corolla height but somewhat unequal in length, simple or rarely with a few short and inconspicuous outer setae. TYPE: México, Tamaulipas, in mountains with steep cliffs, 10 km above and w. of Miquihuana, in meadows with pines present, 3110 m, 4 Aug 1941, L. S. Stanford, K. L. Retherford, and R. D. Northcraft 631 (Holotype: NY!; isotypes: GH!, MO!, OS!). PARATYPES: Nuevo Leon: rare, growing in deep moss on ne. slope of Picacho Onofre, 150 m below the summit, ca. 3300 m, small open- ing in pine woods, Cerro Pena Nevada, ca. 30 km ene. of Doctor Arroyo, 1 Aug 1977, Nesom R590 with C. Wells (LL, MEXU); in open pine forest, occasional, Pena Nevada, 42 km (26 mi) ne. of Doctor Arroyo, w. side of mt. known locally as Picacho Onofre, 3300 m, 4 Jul 1959, Beaman 2705 (MSC, US); Tamaulipas, Cerro Pena Nevada, limestone derived soils, exposed open areas, 1 Jun 1974, Patterson 1518 (LL). The isotype at MO bears the label data given above, but it is pre- sented on a “correction label”. The sheets at GH, NY, and OS have the same collection number, but the locality data are “4 km w. of Miquihuana on limestone ridges in open pine forest”. Erigeron wellsii was tentatively recognized as an undescribed species by I. M. John- ston, who distributed type sheets as “E. retherfordii sp. nov.”. How- ever, the herbarium name was never validated by publication. A num- ber of other unvalidated names on collections made by Stanford et al. and distributed by Johnston as “types” also have been found: Acacia << Fic. 2. Habit sketches of four new Erigeron species. A. Erigeron wellsii Nesom. Isotype (MO). B. Erigeron solisaltator Nesom. Holotype (LL). C. Erigeron stanfordii I. M. Johnston ex Nesom. Holotype (GH). D. Erigeron unguiphyllus Nesom. Hartman and Funk 4098 (OS). 140 MADRONO [Vol. 28 trium, Sphaeralcea oxyloba, Mentzelia retherfordii, Erigeron jimul- cans (=Erigeron commixtus Greene), Erigeron northcraftii (=Erig- eron pubescens HBK.), and Erigeron stanfordit. The name applied here to the Pena Nevada Erigeron recognizes Christopher J. Wells, who is currently a graduate student in botany at Mississippi State University. The monocephalous stems with very thin, retrorse trichomes, basal rosette with reduced cauline leaves, thin phyllaries, and simple pappus are characters that Erigeron wellsiz has in common with forms of the E. scaposus DC.-E. longipes DC. complex. However, the following features of the new species distinguish it from all plants of that com- plex to which it might appear similar: short, thick rhizome with long, thick, fibrous roots; non-clasping cauline leaves, and relatively broad rays that do not curl upon wilting or maturity. Erigeron palmeri A. Gray grows sympatrically with E. wellsit and has a similar growth habit, but the former has glabrous or glabrate leaves and stems, leaves with finely serrate or merely mucronulate, ciliate margins, and long petioles. Though plants of E. palmeri also are fibrous-rooted, they lack rhizomes. Pollen grains from several plants of Erigeron wellsii average 25 wm in diameter with a range of 22—28 um. Stainability in cotton blue averages 90 percent; micrograins are uncommon. The relatively large, stainable grains and the presence of micrograins suggest that these plants may be tetraploid. Erigeron stanfordii I. M. Johnston ex Nesom, sp. nov. Herbae perennes affinitatis obscurae, rhizomata tenuia, caules mon- ocephales, patenti-pubescentes dense, pars supera caulis fere scaposa, folia strigosa dense et paginis inferis griseo-viridis, phyllaria lineari- triangularia in 4—6 series aequilongas (Fig. 2C). Perennials from a long, thin, horizontal rhizome; upright caudex branches 0.5—9.0 cm long, produced at short intervals, bare except for dead leaves or petiole bases. Stems 18—24 cm tall, produced singly at tips of caudex branches, simple or with 1-3 short branches on lower Y;, moderately pubescent with thin, spreading trichomes 0. 1—0.6(—0.9) mm long. Leaves on lower %4—-¥% of stem, 20-45 mm long, blades 5— 9 mm wide, elliptic to elliptic-oblanceolate, entire or with 1—3 shallow, serrate teeth, attenuate to petiole 4%—% as long as leaf, not clasping, gray-green below, darker green above, moderately to densely pubes- cent with ascending trichomes, usually more densely below and often nearly pilose, margins eciliate. Heads solitary, peduncles 15-18 cm long with 1-4 linear bracts 3-14 mm long; involucres hemispheric, 10-13 mm wide (pressed); phyllaries in 4—6 imbricated series, reflexing after release of achenes, linear-lanceolate, stramineous with a dark midline, inner 6.0—8.0 mm long, 0.3—0.4 mm wide, mostly glabrous, 1981] NESOM: MEXICAN ERIGERON 141 outer densely pubescent with vitreous, spreading trichomes, densely and minutely granular—glandular; receptacles shallowly convex. Ray flowers 45—76 in 1-2 series, corollas drying white, lilac-tinged, or yel- lowish, 11.0—12.7 mm long, 1.5—1.8 mm wide, sometimes curling with maturity. Disc flowers narrowly funnelform, barely or not constricted in lower 4%, 4.0-5.0 mm long; style branches 0.8—0.9 mm long, in- cluding the shallowly triangular to shallowly deltate collecting ap- pendages 0.1 mm long. Achenes ca. 2.0 mm long and 0.4 mm wide, with 2 thin, orange ribs, sparsely strigose; carpopodium 5-8 cells high; pappus of ray and disc achenes similar, of 15—21 slender bristles 4/s the disc corolla height, with a few, inconspicuous, outer setae 0.1—0.3 mm high. Type: Mexico, Tamaulipas, in hills 19 km se. of Miquihuana on road to Palmillas in narrow, deep, and moist arroyo, 2250 m, 11 Aug 1941, L. S. Stanford, K. L. Retherford, and R. D. Northcraft 838 (Holotype: GH!; isotypes: MO!, NY!, OS!). I. M. Johnston’s name is validated here to commemorate L. S. Stanford, who led collecting trips in 1941 and 1949 to northeastern Mexico. Erigeron stanfordii is a very distinctive taxon, but the direc- tion of its affinities within the genus is not clear. It is known only from the type collection, and along with Erigeron wellsi1, is endemic to the region of Pena Nevada. The most distinctive characters of E. stanfordi are the following: slender, creeping rhizomes bearing up- right stems; clustered, undifferentiated basal and lower cauline leaves, gray-green and densely pubescent below; peduncles long, leafless, densely pubescent with thin, spreading trichomes, bearing solitary heads; phyllaries very narrowly lanceolate, in 4—6 imbricated series, inner straw-colored with a very conspicuous, narrow, greenish-brown midline. Judging from the relatively large size of the pollen grains, averaging (18—)22(—24) wm in diameter, and high stainability (98 per- cent), the plants are probably sexual tetraploids. Erigeron unguiphyllus Nesom, sp. nov. Habitu Erigeron gypsoverus Nesom, sed distinctus phyllariis glan- dulosis pubescenta patenti, radiis parvioribus numerosibus, pappo se- tarum pauciorum (Fig. 2D). Perennials with a woody taproot, highly branched at base from a relatively thick caudex region, producing a low, compact cluster of wiry, crowded-appearing, branched stems. Stems 5—9 cm tall, erect or ascending, moderately pubescent with loosely appressed to ascend- ing or mixed spreading-ascending-appressed trichomes 0.2—0.5 mm long, spreading just under the heads, abundantly but inconspicuously granular-glandular. Basal leaves absent at flowering, cauline 2.5—5.5 mm long, 0.4—0.6 mm wide, not reduced in size upwards except for occasional peduncular bracts, sessile, margins entire, eciliate, apices 142 MADRONO [Vol. 28 apiculate with a distinctive cap of indurated tissue. Heads numerous on peduncles 1-17 mm long; involucres shallowly hemispheric, 3.5-— 5.0 mm wide (pressed); phyllaries in 2—3 unequal to imbricated series, reflexing after release of achenes, lanceolate to elliptic-lanceolate, yel- lowish with a brown midregion, inner 2.1—-3.0 mm long, 0.5—0.8 mm wide, sparsely to moderately pubescent with spreading trichomes, moderately to densely punctate-glandular; receptacles shallowly to steeply convex. Ray flowers 60-110 in 2—3 series, corollas white, 2.5— 3.9 mm long, 0.2—0.5 mm wide, not curling or reflexing. Disc corollas tubular, slightly constricted in lower !/s, slightly indurated just above, 1.5—2.1 mm long; style branches 0.3—0.4 mm long, including the tri- angular to deltate collecting appendages 0.1—0.2 mm long. Achenes 0.8—1.0 mm long, 0.2—0.3 mm wide, with 2 thin ribs, sparsely strigose; carpopodium 3-7 cells high; pappus of ray and disc achenes similar, of 6-9 bristles 3/s—*/s the disc corolla height, with a conspicuous outer series of setae, squamellae, or scales 0.3—0.8 mm high. ” = 9. TYPE: Mexico, San Luis Potosi, Minas de San Rafael, Jul 1911, C. A. Purpus 5020 (Holotype: US!; isotypes: F, GH 2 sheets!, MEXU!, MO 2 sheets!, NY 2 sheets!, UC!, US!). The number for this collection in Purpus’ notebooks at UC is 5120, although the sheets were distributed as 5020. PARATYPES: San Luis Potosi: common on barren, gypseous (?) soil, Hwy 70, 9.6 km (6 mi) e. of Rio Verde, 14 Aug 1976, Hartman and Funk 4098—voucher for chromosome count, = 9 (ENCB, LL, MEXU, OS, RM); 10 km al e. de Rio Verde, sobre el camino a Bo- quilla, 1000 m, 20 Jan 1959, Rzedowski 9547 (ENCB); ca. 4 km al n. de Rio Verde, sobre el camino a Pastora, 1000 m, 21 Jan 1959, Rze- dowski 9568 (ENCB); ca. 15 km al n. de Rio Verde, sobre el camino a Pastora, 1000 m, 21 Jan 1959, Rzedowski 9588 (ENCB). Erigeron unguiphyllus is characterized by: a perennial, low, com- pact habit with wiry stems and tiny, linear leaves that are usually somewhat curved and terminated by a conspicuous, indurated api- culum or mucro; loosely ascending stem pubescence of short hairs; 2— 3 series of unequal phyllaries that are spreading-pubescent and mi- nutely punctate- or viscid-glandular; numerous and relatively short ray flowers; and a conspicuously double pappus of 6—9 fragile bristles and an outer series of short setae or scales. The diploid chromosome count of m = 9 was made from several cells at diakinesis; meiosis and tetrad formation were regular. Pollen grains are even-sized, averaging about 17.5 wm in diameter and staining 99 percent in cotton blue. According to Sousa (1969), Purpus collected several times during 1919 and 1911 at Minas de San Rafael (or Minas de San Rafael y Huascama). These mines are located in the vicinity of Rio Verde just south of Huascama—22°13’N, 100°15’W, about 96 km due east of San Luis Potosi. Erigeron unguiphyllus joins a number of other new species first collected by Purpus in this immediate vicinity. The di- 1981] NESOM: MEXICAN ERIGERON 143 Fic. 3. Habit sketch of Erigeron cuatrocienegensis Nesom. Holotype (ASU). minutive growth habit of these plants and their resemblance in habit to other known gypsophilic Erigeron species suggest that the mines are associated with gypsum outcrops. Erigeron cuatrocienegensis Nesom, sp. nov. Habitu Erigeron pinkavii Turner affinis, differt imprimis pubes- centia brevissima appressa caulium et phyllarorum, foliis caulinis in- feris eciliatis, radiis paucioribus, et pappo setarum plus numerosarum (Fig. 3). Perennials from a woody taproot, producing up to 4 stems at the base from a simple caudex, or the caudex with several very short axes, each bearing erect stems. Stems 6-21 cm high, erect or ascending, simple or usually few-branched, stems and leaves densely strigose with antrorsely appressed, short (0.1-0.6 mm), white trichomes. Basal leaves deciduous after early flowering, ca. 10-18 mm long, blades 1-— 3 mm wide, entire or with 1-2 pairs of shallow serrations, gradually narrowed to petiolar region; cauline becoming entire, linear, and ses- sile on upper 24 of stem. Heads terminal on peduncles 3—30 mm long, involucres shallowly hemispheric, 5-7 mm wide (pressed); phyllaries in 3 imbricated series, reflexing after release of achenes, innermost widely oblanceolate, 0.6—0.8 mm wide, 2.7—3.0 mm long, with narrow scarious margins, outermost narrower and ca. 2 as long as inner, moderately to densely pubescent with loosely appressed, white tri- chomes; receptacles slightly convex. Ray flowers 30—55 in 1-2 series, corollas white, sometimes drying lavender-tinged, 4.4—7.1 mm long, 0.6—0.9 mm wide, not curling or reflexing with maturity. Disc corollas tubular, constricted in lower !/s, somewhat indurated above, 1.8—2.3 mm long; style branches 0.4 mm long, including the triangular to deltate collecting appendages 0.2 mm long. Achenes 1.0—1.2 mm long, 0.3-0.4 mm wide, with 2(3) thin ribs, sparsely strigose; carpopodium 144 MADRONO [Vol. 28 1—3 cells high; pappus of ray and disc achenes similar, of 18—27 bristles 242—5/6 the disc corolla height, usually with a few, inconspicuous, outer setae up to 0.3 mm high. TYPE: Mexico, Coahuila, Poso de Anteojo (ca. 12 km wsw. of Cua- tro Cienegas), 12 Jun 1968, E. Lehto, D. J. Keil, and D. J. Pinkava 5511 (Holotype: ASU)). PARATYPES: Coahuila, desert scrub and bajada, near cave, ne.-fac- ing slope near tip of Sierra de San Marcos, 21 Mar 1972, Pinkava 10503 (ASU); nw. of Laguna Churince, (ca. 17 km sw. of Cuatro Cienegas), 13 Aug 1967, Cole et al. 3766 (ASU); Laguna Chiqueros complex of lakes and streams w. of stabilized dunes of Poso de la Becerra, ca. 16 km (10 mi) sw. of Cuatro Cienegas, 14 Aug 1967, Cole et al. 3849 (ASU). All collections of Erigeron cuatrocienegensis have been made within about a 17 km distance wsw. to ssw. of Cuatro Cienegas (Fig. 1). This narrow endemic is probably an obligate gypsophile, although the oc- currence of a gypseous substrate is not mentioned in the collection data. From this immediate locality five other narrowly endemic, gyp- sophilic Asteraceae of other genera are known, as well as an endemic Phacelia (see Atwood and Pinkava, 1977). Erigeron pinkavii, another narrowly endemic gypsophile, also grows in the Cuatro Cienegas area and has a growth habit similar to that of E. cuatrocienegensis, although this is probably a convergent adaptive complex of characters. Dissimilarities in other characters sug- gest that their closest relationships are probably not with each other. Their geographic ranges appear to be parapatric or weakly allopatric (Fig. 1); Erigeron pinkavii has been collected mostly to the n., ne., and e. of Cuatro Cienegas. The spreading stem and phyllary pubescence, ciliate leaves, 60—90 ray flowers, and the double pappus of 7-11 bris- tles and a conspicuous outer series of setae or scales are characters of E. pinkavii that mark it as clearly distinct from E. cuatrocienegensis. Judging from the small and even-sized pollen grains, averaging about 18.5 «wm in diameter, with a stainability of greater than 99 percent, plants of Erigeron cuatrocienegensis are probably diploid. Powell and Powell (1977) reported that E. pinkavii is also diploid, and their data suggest that among gypsophilic Asteraceae of the Chi- huahuan Desert, diploids are more numerous than polyploids. Besides Erigeron pinkavii and the two new gypsophiles described in this paper, still another gypsophilous, narrowly endemic Evigeron has been described from northeastern Mexico, E. gypsophilus Turner (Turner, 1975). However, E. X gypsophilus Beauverd (Beauverd, 1930) holds priority over the 1975 binomial; thus, the plants from south-central Nuevo Leon are given a new name: Erigeron gypso- verus Nesom, nom. nov.—based on Erigeron gypsophilus Turner, Wrightia 5:118. 1975. A chromosome count from a recent collection shows this species to be diploid with = 9 (Nesom R1008: LL, MEXU). 1981] NESOM: MEXICAN ERIGERON 145 All four of these gypsophilic endemics apparently are diploid, all have very restricted ranges in northern Mexico (Fig. 1); but each is strongly differentiated morphologically, and the nature of their inter- relationships is not clear. Of great interest would be a knowledge of whether they form a natural group, reflecting a gypsophilous tendency in the ancestral stock from which they radiated, as in examples from several other genera discussed by Turner and Powell (1979), or wheth- er each has independently attained its gypsophily. From northeastern Chihuahua a collection has been made (Fig. 1) of several plants (Chiang, Wendt, and Johnston 9851B, LL) that are very similar to Erigeron gypsoverus in growth habit. In addition, the label data indicate that they were growing “above ‘Los Morteros’ gyp- sum mine... (in) calcareous (and slightly gypseous?) gravel”. How- ever, these plants have more herbaceous bases and a pubescence of thicker, widely spreading or ascending below, trichomes on the stems, leaves, and phyllaries. They also have tiny heads with few rays and narrowly oblanceolate, entire leaves, and they produce abortive pol- len, indicating that they are probably polyploid. Although these plants are distinctively different from all of the other four known gypsophiles, their probable polyploidy and similarity to some forms in the variable E. modestus agamic complex bid circumspection before giving them a formal taxonomic circumscription. Further collections and obser- vations may show that they warrant recognition as a distinctive species. Erigeron solisaltator Nesom, sp. nov. A Erigeron coronarius Greene affinis, differt phyllariis erectis ad maturitatem receptaculis conicis, et corollis radii ligulis angustis non reflexis ad maturitatem (Fig. 2B). Annuals from a very slender taproot, producing 1—4 upright stems from a simple caudex. Stems 12-15 cm tall, few-branched near the middle; moderately pubescent with spreading or ascending, whitish trichomes 0.1—0.7 mm long, obscurely granular-glandular. Basal leaves mostly deciduous by flowering, 5-12 mm long, blades obovate to oblanceolate, 2-4 mm wide, entire or shallowly few-toothed, atten- uate to narrow petiole ca. 2 as long as leaf; cauline leaves oblanceolate to linear-oblanceolate, sessile to subsessile, entire, lower 6-15 mm long, 2-3 mm wide, gradually reduced in size upwards to linear bracts. Heads few, solitary, peduncles 5-35 mm long, involucres shallowly hemispheric, 5—6 mm wide (pressed); phyllaries in 3 subequal series, remaining erect after release of achenes, oblanceolate, with wide, light or scarious margins, inner 2.8—4.2 mm long, 0.6—0.8 mm wide, out- ermost 2—-'4 as long as inner, sparsely pubescent with white, spread- ing trichomes, densely granular-glandular; receptacles hemispheric to very steeply convex. Ray flowers 85—120 in 1-2 series, not reflexing with maturation, corollas white, drying with lavender tips, 3.7—5.5 146 MADRONO [Vol. 28 mm long, 0.2—0.4 mm wide. Disc corollas tubular to narrowly fun- nelform, constricted in lower '/s, indurated and inflated above, 1.8- 2.1 mm long; style branches 0.4 mm long, including the shallowly to very shallowly triangular collecting appendages 0.1—0.2 mm long. Achenes 0.9-1.0 mm long, 0.5 mm wide, with 2 thin ribs, sparsely strigose; carpopodium 5-8 cells high; pappus of ray and disc achenes similar, of 8—9 persistent bristles, with an outer laciniate corona or series of scales ca. 0.2 mm high. TYPE: Mexico, Chihuahua, zacatal, Prosopis TTDI Koeber- linia spinosa, and Hilaria mutica, ne textured, calcareous alluvium in flat (bottom of bolson), 1185 m, 0.5 km s. of Rancho El Llano, 14 Jun 1973, M. C. Johnston, T. Wendt, and F. Chiang 11317F (Holo- type: LL!). Erigeron solisaltator has many similarities with the taxa known as Achaetogeron linearifolius Watson and Achaetogeron ascendens Greenman and with Erigeron coronarius Greene (Nesom, 1980). It differs from the first two in its annual duration and in its normal complement of persistent pappus bristles, from the third in having non-reflexing phyllaries, and from all three in having hemispheric re- ceptacles and non-reflexing ligules. The location and habitat of E. solisaltator are also distinctive and different from any of its probable relatives. The single known collection was made in northeastern Chi- huahua near the Rio Grande (Fig. 1) in the shrubby vegetation of a small, undrained basin. The epithet means “sun-dancer”. ACKNOWLEDGMENTS I thank curators of the following herbaria, from which collections were borrowed or studied: ASU, GH, LL, MO, MSC, NY, OS, TEX, UC, and US. Thanks also to Tod Stuessy and John Strother for checking type collections at F and UC and to Don Pinkava for information on distributions within the Cuatro Ciénegas Basin. Ron Hartman and Vicki Funk furnished buds for the chromosome count of Erigeron unguiphyllus. The comments of the editor, John Strother, and Jim Henrickson were of essential assistance in readying the manuscript, and taxonomic advice from Bob Wilbur is appreciated. LITERATURE CITED ATwoop, N. D. and D. J. PINKAvA. 1977. A new gypsophilous species of Phacelia (Hydrophyllaceae) from Coahuila, Mexico. Madrono 24:212-214. BEAUVERD, G. 1930. Polymorphisme de quelques plantes du Massif de la Vanoise (Savoie). Bull. Soc. Bot. Geneve. Ser. II. 22:439—464. NEsSoM, G. L. 1980. A revision of the epappose species of Erigeron (Asteraceae— Astereae). Ph.D. dissertation, Univ. North Carolina, Chapel Hill. POWELL, A. M. and S. A. POWELL. 1977. Chromosome numbers of gypsophilic plant species of the Chihuahuan desert. Sida 7:80—90. SousA SANCHEZ, M. 1969. Las colleciones botanicas de C. A. Purpus en Mexico. Periodo 1898-1925. Univ. California Publ. Bot. 51:ix + 1-36. TURNER, B. L. 1975. Two new gypsophilic species of Erigeron (Asteraceae) from northern Mexico. Wrightia 5:116-119. 1981] NESOM: MEXICAN ERIGERON 147 TURNER, B. L. and A. M. PowELL. 1979. Deserts, gypsum and endemism. J J. R. Goodin and D. K. Northington, eds., Arid Land Plant Resources, p. 96-116. Proc. Int. Arid Lands Conf. Pl. Res., Texas Tech. Univ., Lubbock. (Received 28 Dec 1979; revision received 18 Dec 1980; accepted 7 Jan 1981.) THE DIANDROUS, HYPOSTOMATIC WILLOWS (SALICACEAE) OF THE CHIHUAHUAN DESERT REGION MARSHALL C. JOHNSTON Department of Botany, University of Texas, Austin 78712 ABSTRACT A key is provided for the five species of Salix in the Chihuahuan Desert Region with two-stamened flowers and leaves with stomates almost confined to the lower surfaces. One species, S. lasiolepis Bentham, was previously known and named; three new species are described from the state of Coahuila: S. pattersonii, S. riskindii, and S. wendtii. The fifth species, also from Coahuila, is discussed but remains nameless be- cause it is known only from two sterile collections. In the Chihuahuan Desert Region, delineated for floristic purposes by Johnston (1977), willows occur in a few mesic habitats, especially along the Rio Grande and its tributaries and in even more areally restricted and scattered montane canyons and high slopes. Climatic vagaries and logistical difficulties of exploration have so far prevented the gathering of comprehensive material of willows in this region and seem likely to inhibit such gathering for years to come. The challenge of future exploration is to visit each remote population at appropriate seasons in order to collect specimens representing both sexes at various stages of development with a high degree of confidence that the spec- imens are conspecific. Until such material is forthcoming, treatments of the willows of the Chihuahuan Desert Region must be considered even more preliminary and tentative than treatments of willows in other regions. Notwithstanding the incompleteness of the herbarium stores, it has been necessary recently to prepare a taxonomic treatment of Salzx for the Chihuahuan Desert Flora being compiled by Dr. James Henrick- son and me. I present here some of the results of my study. Because of space limitations I omit further mention of the arboreal, pleiandrous “black” willows, all of which have been fairly well understood by Schneider (1918), Ball (1950, 1961) and Dorn (1976, 1977) and which in this region are found at relatively low altitudes along the Rio Grande and its tributaries. I also exclude from further consideration the species of the very poorly named Salix sect. Longifoliae Andersson, namely S. taxifolia Humboldt, Bonpland & Kunth, S. exigua Nuttall and S. interior Rowlee. These three species, although diandrous, have rela- tively loosely flowered catkins and their leaf-blades have almost as many stomates on the upper as on the lower surface. They are also MaproNno, Vol. 28, No. 3, pp. 148-158, 31 July 1981 1981] JOHNSTON: CHIHUAHUAN DESERT WILLOWS 149 fairly well understood taxonomically, though far less well than are the “black” willows. With the exception of S. taxizfolia, the members of sect. Longifoliae, like the “black ” willows, are nearly restricted to the lower altitudes along rivers and creeks. The subjects of this paper are the diandrous, multistemmed shrubs or low trees with dense ascending and spreading catkins and spreading leaves with stomates almost entirely absent from their upper surfaces. In the entire region, an area almost as large as California, only a few mountain ranges provide the mesic canyons and slopes suitable for these willows: in Texas, the Davis, Chisos, and Vieja mountains; in Chihuahua, the Sierra Rica; and Coahuila, the Sierra Maderas del Carmen, the Sierra de la Madera, and, at the extreme eastern margin, the Serranias del Burro. From the entire Mexican portion of this re- gion, Johnston (1944) saw only one specimen of this species-group, a sterile specimen from the Sierra de la Madera. Since 1970, mainly through the strenuous and perceptive fieldwork of David H. Riskind, Tom Wendt, Tom Patterson, and Emily J. Lott, several collections have accumulated from northern Coahuila that represent three distinct new species. Study of a recent collection of the Sierra de la Madera population reinforces my suspicion that it represents still another new taxon, but because flowers and fruits are still not available, I refrain from proposing a name for it. TAXONOMIC TREATMENT Key to species Lower surfaces of leaves densely canescent-pubescent with antrorse, silky, white hairs, or glabrate. Aments appearing in February and March, rarely later, on leafless branches of the previous year; leaf-blades oblanceolate to lin- ear-lanceolate; stamen-filaments joined 0.2—0.7 of their length. Ovaries wholly glabrous; filaments joined only 0.2—0.3 of their Nee Se eta tes Me, eR cece nent aah eek eat te 1. S. lastolepis. Ovaries hairy at least along lines of dehiscence, sometimes all over; filaments joined more than 0.5 of their length ....... Ba aa ae eben done 2. S. riskindit. Aments appearing in May or August at end of more or less leafy ament-tipped lateral twigs; leaf-blades more or less elliptic; fil- aments free (unknown in No. 3). Aments appearing in May with the leaves, much exceeding twig- leaves; leaves entire ..2....4. 000.0005. 3. S. pattersoniz. Aments appearing in August with mature summer leaves; leaves S@uhUlALe wane eee sete te oe aes Ca 4. S. wendtit. Lower surfaces of leaves with closely appressed, scattered, rufous 102) oe a ee ee ee ec 5. Sp. nov.? 150 MADRONO [Vol. 28 1. Salix lasiolepis Bentham, Pl. Hartweg. 335. 1857.—TypeE: USA, CA, banks of the Salinas and Carmel Rivers, near Monterey, Hartweg 1955 (Holotype: K!). Salix lastolepis var. bracelinae Ball, J. Wash. Acad. Sci. 40:331. 1950.—TyYpE: USA, CA, Contra Costa Co., Antioch, Eastwood 3729 (Holotype: US)). Description of material from the Chihuahuan Desert Region: trees or shrubs, usually multi-trunked; one-year-old twigs brownish, rarely pruinose; older twigs olive-green to yellow-brown or commonly or- ange, internodes 3—8 mm long, with usually ascending-appressed but sometimes spreading, crisp, gray-white hairs 0.3-—0.5 mm long when very young, then glabrate; axillary (winter) buds prominent, often 3— 6 mm long, hairy like the youngest twigs. Leaf blades linear-lanceolate to linear-oblanceolate, (4-)5—8(-11) cm long, 6—12(-18) mm wide, usually inconspicuously gland-toothed with teeth antrorse-appressed, or nearly entire; abaxial surface with close, white, silky or crisped hairs 0.3—0.8 mm long to quickly glabrate in some specimens, epider- mis glaucous to merely pale green; adaxial surface darker green, lustrous and essentially glabrous except on midveins, when very young some leaves with appressed-antrorse silky, white hairs 0.3-0.8 mm long; petioles 2—7 mm long; stipules almost always absent, extremely small when present. Staminate aments: precocious in March or less commonly appearing in July in axils of mature leaves, ascending-ap- pressed, 9-18 mm long, 5—6(—8) mm thick; scales 1.3—1.6 mm long, obovate, blunt, appressed-ascending, abaxial surface with antrorse silky white hairs 0.7-1 mm long; adaxial pubescence similar except glabrous and glandular in lower third; filaments ca. 3-—3.5 mm long, joined 1—1.3 mm at base, free 1.7—2.7 mm above; anthers rotundly or narrowly horseshoe-shaped, 0.3—0.4 mm long and wide. Pistillate aments: precocious in March; 1—-1.5 cm long, 6-8 mm wide, ca. 50- flowered, dense; scales and flowers at first ascending, later spreading; scales obovate, blunt, 0.6—1 mm long, with antrorse, silky, white hairs 0.6—1 mm long; stipes ca. 0.5 mm long; styles 0.3—0.5 mm long, stig- mas 0.2—0.3 mm long. Fruit 2.5—-3 mm long, glabrous; seeds ca. 0.7— 0.8 mm long. In this region, S. lastolepis is known from wooded creek-canyons in igneous-rock mountains, principally in the Davis Mountains of trans-Pecos Texas, but with smaller populations in the Chisos and Vieja mountains of trans-Pecos Texas, the Sierra Rica of extreme northern Chihuahua and the Sierra Maderas del Carmen of extreme northwestern Coahuila. The localities are all above 1400 m elevation. 2. Salix riskindii M. C. Johnston, sp. nov. Frutices multicaules vel etiam arbores parvae, ramuli hornotini fo- lioli dense pubescentes; laminae foliorum oblanceolatae 3—7 cm longae 1981] JOHNSTON: CHIHUAHUAN DESERT WILLOWS Meet Fic. 1. Salix riskindii M. C. Johnston. A. Twig with inflorescences. B. Leafy twig. 9-15 mm latae persistente dense pubescentes (subtus densius) pilis antrorsis adpressis sericeis albis, petioli (1—)2—3 mm longi pubescentes. Amenta praecocia verna adscendentia staminata 15-25 mm longa 8— 12 mm crassa; filamenta 2 fere glabra partibus coalescentibus 2—3 mm longis partibus liberis 1-2 mm longis; squamae fructiferes 1-1.5 mm longae ca 1 mm latae pubescentes pilis 1-2 mm longis; ovaria ca 3 mm longa omnino pubescentes vel pubescentes non nisi secus suturas (Fig. 1). Multi-stemmed shrubs 1-3 m tall or even multi-trunked small slen- der trees to 7.5 m tall (Wendt and Lott 126); leafy twigs of the season brown-gray, densely pubescent with silky, gray-white, somewhat crisped, appressed, spreading hairs; axillary (winter) buds prominent, 5-8.5 mm long, hairy like the youngest twigs. Leaves expanding in April after flowering is completed, apparently ascending or spreading 152 MADRONO [Vol. 28 (not drooping); leaf blades (3—)6—7(—12) cm long, 9-15 (—19) mm wide, oblanceolate, entire, persistently densely pubescent (more densely so beneath) with antrorse, appressed, silky, whitish hairs; stomates pres- ent beneath, rare or absent above; petioles (1—)3—11 mm long, pubes- cent like blades. Flowers appearing in February and March on leafless branches of previous year; aments abundant on twigs but solitary at nodes, dense, sessile, ascending (staminate) or drooping at tip (pistil- late), cylindric, 15-25 mm long, 8-12 mm thick, silky-hairy, rounded at tip; winter-bud scales subtending aments buffy brown, pubescent like the leaves, 5—7 mm long. Staminate flowers (Riskind and Riskind 2052): scales oblong, 2-3 mm long, blunt, pale brown, dorsally densely pubescent with white, silky, antrorse hairs 1-2 mm long; vestigial ovary slender, conical, yellow-green, gland-like, ca. 0.8 mm long; sta- mens 2, nearly glabrous, joined part of filaments 2—3 mm long, free parts 1-2 mm long; anthers purple-brown, ca. 0.6—1 mm long, ovate- orbicular. Pistillate flowers at stage of pollen-receptivity not known, present at a later stage with young fruit in Riskind and Patterson 1944a and Wendt 126: scales brown, more or less oblong-obovate, ca. 1.5 mm long (1944a) or 1 mm long (126), 1 mm wide, pubescent with antrorse, white, silky hairs 1-2 mm long either equally densely on both surfaces (126) or more densely so on back (1944a); gland minute, oblong, scale-like, ca. 0.3 mm long; ovary stipe ca. 0.5 mm long with white, silky hairs; ovary ca. 3 mm long, hairy all over (126) or only in vertical stripes along lateral sutures (1944a); style ca. 1 mm long of which upper 0.2 mm is forked and stigmatic. Seeds (126 only) nu- merous, oblanceoloid, ca. 0.9 mm long of which lower 0.2 mm com- prises narrowed stipe; hairs numerous, 2—3 mm long attached to disk- or callus-like funicle. Type: México, Coahuila, Sierra Maderas del Carmen, Canon Car- boneras, ca. 1 km s. of El Uno, along perennial stream in pine-oak woodland, 28°59'30"N, 102°33’W, 1500-2100 m, 2 Apr 1974, Tom Wendt 126 with Emily J. Lott and David Riskind (holotype: TEX). PARATYPES: Mexico, Coahuila, Municipio de Villa Acuna, Serranias del Burro, Rancho El Bonito, Canon El] Bonito ca. 2.5 km above first dam, 29°1'30”N, 102°7'30"W, 1700 m, 11 Apr 1976, D. J. Riskind and T. F. Patterson 1944a (TEX); same locality, 20 Feb 1977, D. H. Riskind and J. Riskind 2052 (LL); same canyon, 29°0'30’N, 102°7'30"W, 1800 m, 20 Sep 1977, J. Valdés R. and A. L. Metcalf 2249 (LL). The label of Wendt 126 indicates that these willows are shrubs to small trees, sometimes exceeding 7 m in height, abundant in the “main” arroyo, extending from at least 1500 m elevation, where the plants were fully leafed on 2 April, to 2100 m where they were only in bud; the foliage is said to be “silvery light green”. The associates are said to be Quercus spp., Cupressus arizonica Greene, Pinus ari- zonica Engelmann, P. cembroides Zuccarini and Vitis sp. The label 1981] JOHNSTON: CHIHUAHUAN DESERT WILLOWS 153 of Riskind and Patterson 1944a notes that these are abundant shrubs to 3 m tall on margins of wet meadow in woodland of Quercus mueh- lenbergit Engelmann in upper reaches of Canon EI Bonito. The label of Riskind and Riskind 2052 says that these are abundant multi- stemmed shrubs to 3 m on margins of intermittent stream and cienega in deciduous woodland in upper reaches of Canon El Bonito along logging road ca. 2.5 km upstream from first dam, associated with Quercus gravesit Sudworth, Prunus mexicana Watson, Quercus muehlenbergit, and Pinus arizonica var. stormiae Martinez. The shorter stature and the slightly different ovary pubescence of the plant of the Serranias del Burro lead to the suspicion that they may deserve recognition as taxonomically distinct from the plants of the Sierra Maderas del Carmen only some 70 km farther west, but the material available is not adequate for a critical evaluation of this point. I await receipt of more adequate specimens from both areas, especially staminate plants from the Sierra Maderas del Carmen. In Schneider’s (1918) treatment of Mexican willows, S. riskindi1 keys to S. paradoxa H.B.K. but it differs from S. paradoxa in the smaller, sessile aments and much denser, more persistent pubescence of silky hairs as well as the height of joining of the stamen-filaments. Salix riskindi1 may be more closely related to S. lastolepis with which it shares the features of basal filament-joining and over-all habit. But it is quite distinct in a number of other features. 3. Salix pattersonii M. C. Johnston, sp. nov. Frutices multicaules ad 2.25 m alti, ramuli hornotini pilis multis demum glabrati. Laminae foliorum elliptico-oblanceolatae 2—6.5 cm longae 1—2.5 cm latae supra praeter costam glabrae nitidae costis al- bopubescentibus, subtus persistente dense canescento-pubescentes in- tegrae. Amenta staminata ignota; amenta pisitillata ut videtur inter mensem mali emergentia coetanea in extremitatibus ramulorum ra- mulis pedunculiformibus 1—2(—3)-nodatis 3—5-mm longis folia ramu- lorum superantia, in statu fructifera 4-5 cm longa 10-15 mm dia- metro; capsulae 6—7 mm longae dense pubescentes (Fig. 2). Multi-stemmed shrubs to 2.25 m tall; twigs of the season slender, dark brown, striate when very young with spreading to retrorse, silky, whitish hairs 0.2—0.3 mm long, older portions glabrate; internodes ca. 2-10 mm long; axillary buds pale brown, sparsely pilose, ca. 3 mm long. Leaves ascending; leaf-blades elliptic-lanceolate, 2—6.5 cm long, 1—2.5 cm wide, with acute tips and cuneate-rounded bases, entire- margined, olive-green, lustrous and essentially glabrous above when nearly mature except for white-pubescent larger veins, on abaxial sur- face waxy-papillose, glaucous and persistently densely canescent-pu- bescent with antrorse, silky, white hairs 0.3-0.8 mm long; stomates abundant beneath, absent above; petioles 2-9 mm long, with pubes- 154 MADRONO [Vol. 28 re AN, ot Ee) DI AOA NAGA E> Xe Sy SS Exe Fic. 2. Salix pattersonii M. C. Johnston. Twig with leaves and pistillate inflores- cences. cence like that of the leaves; stipules minute. Staminate aments un- known. Pistillate aments at anthesis not seen, apparently emerging in May with the leaves, terminating 1—3(—3)-noded 3—5-mm-long axillary ascending-spreading twigs, vastly exceeding twig leaves, cylindric, when in fruit 4-5 cm long, 10-15 mm thick; scales 1.5—2 mm long, lanceolate, acute, appressed, with antrorse, silky, white hairs ca. 1 mm long, less densely hairy on adaxial surface; pedicels ca. 1.6—2 mm long; gland absent; capsule 6—7 mm long, densely hairy with antrorse, silky, white hairs ca. 0.5 mm long, after dehiscence with widely re- curved valve-beaks; stigmata ca. 0.3—0.4 mm long, deltoid, about as wide as long. Seeds narrowly falcate-ovoid, compressed, nearly black, 1981] JOHNSTON: CHIHUAHUAN DESERT WILLOWS 155 at base with numerous, silky, white ascending hairs joined at central whitish easily detached, basal callus-like funicle; embryo green. Type: Mexico, Coahuila: Municipio de Ocampo, Sierra Maderas del Carmen, 102°33’N, 28°59’W, Campo 4, in cut-over, open, mixed conifer-oak woods, 27 May 1975, Riskind and Patterson 1809 (Ho- lotype: LL). The label states that the shrubs are locally common and are asso- ciated with Cupressus arizonica, Pseudotsuga menziesii (Mirb.) Fran- co, Pinus spp., Quercus sideroxyla Humboldt & Bonpland, Q. rugosa Née and Abies durangensis Martinez var. coahuilensis (I. M. John- ston) Martinez. The species is currently known only from the type collection. On the basis of over-all vegetative similarity and the shape of the bracts, this pistillate material is perhaps related to Salix oxylepis Schneider, which was known to Schneider only by material from the region near the Peak of Orizaba. Riskind and Patterson 1809 also keys to S. oxylepis in the more extensive treatment by Espinosa (1979), who reports S. oxylepis from six states and the federal district in central Mexico. As compared with S. oxylepis, however, S. patter- sonit has shorter petioles, shorter and narrower leaf blades with per- sistently pubescent abaxial surfaces, and shorter pistillate aments. Two sterile collections also from the Sierra Maderas del Carmen are similar enough vegetatively to Riskind and Patterson 1809 to be placed here tentatively: Mexico, Coahuila, Sierra Maderas del Car- men, rhyolitic s. peaks, nw. side of upper Carboneras Canyon, 28°57'N, 102°34’'W, 2400 m, uncommon small tree in mixed conifer- oak forest, 2 Apr 1974, T. Wendt 124a (TEX); along road above Catedrales Canyon in nw. part of range due w. of Campo Tres, above 2200 m, common multi-stemmed shrubs along intermittent water- courses with Pinus, Pseudotsuga, Quercus, Cupressus, Arbutus, Cor- nus, 3 Aug 1974, D. H. Riskind, B. Burleson, and J. T. Baker 1720 (LL). I think that these two collections very likely represent S. pat- tersonit. 4. Salix wendtii M. C. Johnston, sp. nov. Arbores parvissimae fruticoideae ad 3 m altae ad basem multira- mosae; ramuli hornotini primitus pilis multis demum glabrati. Lami- nae foliorum ovato-ellipticae vel ellipticae vel obovato-ellipticae (4.5—) 5—7(—8) cm longae 2—3.5 mm latae supra praeter costam glabrae nitidae costis albo-pubescentibus, subtus persistente dense canescento- pubescentes serrulatae dentibus parvis antrorsis glandularibus. Amen- ta staminata inter mensem Augusti emergentia coetanea cum follis maturis in extremitatibus ramulorum ramulis pedunculiformibus 2-3 cm longis 2—5-nodatis folia superiora breviora 20—24 mm longa 9-13 mm lata; squamae ovatae 2—2.5 mm longae acutae acuminatae pu- bescentes; stamina 2, filamenta omnino libera 5—8 mm longa (Fig. 3). 156 MADRONO [Vol. 28 Fic. 3. Salix wendtii M. C. Johnston. A. Staminate flower. B. Twig with leaves and one staminate inflorescence. “Shrubby trees” to 3 m tall, much branched at base; twigs of the season slender, dark brown, striate, when very young with numerous, spreading, silky, whitish hairs 0.5—1 mm long, later glabrate; inter- nodes ca. 1—2 cm long; axillary buds dark brown, glabrous, 4-6 mm long. Leaf blades ovate-elliptic to elliptic to obovate-elliptic, (4.5—)5— 7(—8) cm long, 2—3.5 cm wide, firm, membranous, at margins serrulate with small, antrorse gland-tipped teeth (teeth numerous on rapidly growing shoots, less so elsewhere), at tip weakly cuspidate, at base rounded, on upper surfaces olive-green, essentially glabrous and lus- trous at maturity except for white-pubescent midvein, on lower surface waxy-papillose, glaucous and persistently densely canescent-pubescent with antrorse, silky, white hairs 0.5—1 mm long; stomates abundant beneath, absent above; petioles (3—)5—8 mm long, with pubescence 1981] JOHNSTON: CHIHUAHUAN DESERT WILLOWS 157 like that of leaves; stipules on rapidly growing shoots elliptic-falcate, 4—6 mm long, highly asymmetrical, green, leaf-like in texture, serru- late, lower surface white-hairy especially near base; stipules on slower- growing shoots essentially absent. Staminate aments: appearing in August with mature leaves, terminating 2—5 noded, 2—3-cm-long twigs, ascending, shorter than uppermost twig leaves, subcylindric, 20-24 mm long, 9-13 mm wide (including stamens), with a peduncle 3-4 mm long, silky-hairy. Staminate flowers: subtending bract-scale appressed, ovate, 2—2.5 mm long, acute and acuminate, yellow-brown when dry (said to be green when fresh), pubescent with antrorse, silky, white hairs 0.8—1.3 mm long, less densely hairy above; gland ventral, truncate or even concave at tip, more or less rectangular, ca. 0.3 mm long; stamens 2, filaments wholly free, 5-8 mm long, in basal half with many antrorse silky, white hairs ca. 0.5 mm long; anthers ovoid, 0.7—0.8 mm long, yellow, each with an apical gland. Pistillate aments, flowers, and fruits unknown. Type: Mexico, Coahuila, Sierra Maderas del Carmen, 29°00'N, 102'36'W, near Campo Tres on ridge between camp and “Hell’s Kitchen” to the north, common at bases of cliffs of the ridgetop, wsw. slopes, 2600 m, 6 Aug 1974, T. Wendt and A. Adamcewicz 518 (Ho- lotype: TEX). Salix wendti1, known only from the type collection, cannot be ac- commodated in the key or descriptions of Schneider (1918). In Espi- nosa (1979) it keys to S. oxylepis but has longer petioles, shorter, narrower leaf blades, shorter staminate aments and shorter floral scales. 5. Salix sp. nov.? Two sterile specimens represent this entity: Mexico, Coahuila, Mun- icipio de Cuatro Cienegas, Sierra de la Madera, Canon del Agua, 27°3’ or 27°4'N, 102°24’W, 9 Sep 1934, C. H. Muller 3242 (GH, LL); same locality, above the canyon on n. slopes, 2450-3100 m, with Cupressus, Quercus, Pseudotsuga, Pinus, and Abies, 14 Aug 1980, T. Wendt and E. J. Lott P29 (TEX). Muller notes on his label that these are trees to 4.5 m tall with smooth, tawny trunks to 10 cm thick. Wendt and Lott (label and pers. comm., 1980) state that these are small trees 2—6(—9) m tall with 5— 100 or more stems from the base, with smooth, grayish bark. These willows are not common in the Canon del Agua at 2450-2900 m, but on the slopes above the canyon, even up to 3100 m on the Pajarito Peak, they are co-dominant with Cupressus and other conifers. The lowest, earliest leaves of the newly expanding shoots often have nar- rowly obovate or oblong blades 2—4 cm long. Somewhat later leaves have longer, narrowly obovate blades usually 4—9 cm long and 2-3 cm wide. The last-formed leaves are broadly lanceolate, 6—10 cm long, 1—2 cm wide and rather regularly serrulate with small, remote teeth. 158 MADRONO [Vol. 28 The upper surfaces are essentially glabrous. The lower surfaces are non-papillate and nearly glabrous but with closely appressed, scattered, pale rufous hairs ca. 0.4 mm long. According to Johnston (1944), Muller 3242 “appears to be referable to S. paradoxa”, but I find that it differs from that species in many features. Because Muller 3242 and Wendt & Lott P29 do not corre- spond with any previously known willows, I suspect that they rep- resent still another local endemic taxon. It is to be hoped that flowers and fruits will be made available for study. ACKNOWLEDGMENTS David Riskind, Tom Wendt, Tom Patterson, Mike Powell, Richard Worthington, and Emily J. Lott, among others, have been very helpful in making material available for study. LITERATURE CITED BALL, CARLETON R. 1950. New combinations in southwestern Salix. J. Wash. Acad. Sci. 40:324-335. . 1961. Salix. Fl. Tex. 3:369-392. Dorn, RoBERT D. 1976. A synopsis of American Salix. Canad. J. Bot. 54:2769-2789. . 1977. Willows of the Rocky Mountain states. Rhodora 79:390—429. ESPINOSA DE G. RUL, JUDITH. 1979. Salicaceae. Flora fanerogamica del Valle de Mexico. 1:95—99. JOHNSTON, I. M. 1944. Plants of Coahuila, eastern Chihuahua and adjoining Zacatecas and Durango, IV. J. Arnold Arbor. 25:431-453. JOHNSTON, M. C. 1978. Brief résumé of botanical including vegetational features of the Chihuahuan Desert Region with special emphasis on their uniqueness. J” R. H. Wauer and D. H. Riskind, eds., Trans. Symp. Biol. Resources Chihuahuan Desert Region, U.S. and Mexico, p. 335-359. U.S.D.I., Natl. Park Serv. Trans. Proc. Ser. 3. Washington, D.C. SCHNEIDER, CAMILLO. 1918. A conspectus of Mexican, West Indian, Central and South American species and varieties of Salix. Bot. Gaz. (Crawfordsville) 65: 1-41. (Received 18 Jun 1980; revised version received and accepted 19 Jan 1981) A NEW SPECIES OF CRYPTANTHA (BORAGINACEAE) FROM WYOMING ROBERT D. DORN Wyoming Department of Environmental Quality, Cheyenne 82002 ROBERT W. LICHVAR Wyoming Natural Heritage Program, The Nature Conservancy, 1603 Capitol Ave., #325, Cheyenne 82001 ABSTRACT A new species, Cryptantha subcapitata, is described from Fremont County, Wyo- ming. It is compared with C. caespitosa and C. spiculifera which it resembles most closely. In the course of field collecting in central Wyoming, we came across a mat-forming Cryptantha that resembled C. caespitosa (A. Nels.) Payson but had a different aspect. Closer examination of these plants revealed several distinct differences from C. caespitosa, including longer styles and a different type of pubescence. The longer styles suggested C. spiculifera (Piper) Payson but again the pubescence was different as were characteristics of the nutlets and inflorescence. These differences and others support recognition of this mat-forming taxon at the species level. Cryptantha subcapitata Dorn & Lichvar, sp. nov. Herba perennis caespitosa, 5—15 cm alta; folia linearia vel lineari- oblanceolata, 8-28 mm longa, 1-3 mm lata, dense strigosa; inflores- centia capitata vel subcapitata; calyx 5-7 mm longus; corolla alba, tubus 3—4 mm longus, limbus 5-6 mm latus; anthera 0.8 mm longa; stylus fructu 1.5—2 mm longior; nuculae 4, ovatae, 2-3 mm longae, dorso tuberculato-rugulosae, pagina ventrali tuberculato-rugulosae, sulco aperto triangulari (Fig. 1). Mat-forming perennial 5—15 cm high; leaves linear to linear-oblan- ceolate, 8-28 mm long, 1—3 mm wide, densely appressed strigose and with some slightly larger, spreading, pustulate hairs at least on abaxial surface and margins, the old whitish leaves persisting at base; stems greenish; inflorescence capitate or subcapitate; calyx 5—7 mm long, pubescent like the leaves; corolla white, the tube 3-4 mm long, the limb about 5—6 mm across; anthers about 0.8 mm long; nutlets 2—3 mm long, ovate in outline, dorsal surface rugose at center, mostly tuberculate near margins, the ventral surface rugose and tuberculate, MADRONO, Vol. 28, No. 3, pp. 159-162, 31 July 1981 [Vol. 28 MADRONO 160 See OES Habit: scale bar = 1 cm. Ventral Cryptantha subcapitata (from Dorn 3459). view of nutlet and four nutlets with protruding style: scale bar FIG. 1. 1.5 mm. 1981] TABLE 1. Characteristic Leaf shape Leaf pubescence Stems Inflorescence Styles Mature nutlet margins Nutlet scar C. caespitosa Obovate to oblanceolate Uniform, coarse, mostly appressed, not obviously pustulate hairs Straw colored Usually elongate Exceed nutlets by <0.5 mm Usually same as body Open C. subcapitata Linear to linear- oblanceolate Mixture of coarse, appressed, nonpustulate hairs and fewer, spreading, slightly coarser, obviously pustulate hairs Green Capitate or subcapitate Exceed nutlets by 1.5-—2 mm Same as body Open DORN AND LICHVAR: NEW CRYPTANTHA 161 SELECTED CHARACTERISTICS OF THREE TAXA OF Cryptantha. C. spiculifera Oblanceolate Mixture of fine, somewhat appressed, non- pustulate hairs and fewer, spreading, much coarser, obviously pustulate hairs Green or straw colored Usually elongate Exceed nutlets by 1.5—2 mm With narrow, smooth border prominently set off from body by smoothness and color Closed the scar open for most of length of nutlet, the opening triangular at base; style exceeding nutlets by 1.5—2 mm. Type: USA, WY, Fremont Co., T5N R6E S8, just w. of Boysen Dam, rocky calcareous ridge, 1775 m, 23 Jun 1980, Dorn 3459. (Ho- lotype: RM; isotypes: to be distributed). PARATYPES: same location and date as holotype, Lichvar 2886 (RM); USA, WY, Fremont Co., 2.4 km se. of Boysen Camp, 1550 m, 3 Jun 1964, Wight 87 (RM). Cryptantha subcapitata differs from C. caespitosa in that the inflo- rescence is capitate or subcapitate rather than normally elongate, the styles exceed the nutlets by 1.5—2 mm rather than by less than 0.5 mm, and the pubescence is different. It differs from C. spiculifera in that the inflorescence is capitate or subcapitate rather than normally elongate, the pubescence is different, and the nutlet scar is open rather than closed. These and other differences among the three species are summarized in Table 1. Cryptantha subcapitata is perhaps derived from C. caespitosa and thus can be placed in the “caespitosa group” of Higgins (1971). The 162 MADRONO [Vol. 28 pubescence and nutlets of the two are quite similar when compared to other species and the former is on the northern edge of the range of the latter. It is not likely that C. caespitosa is derived from C. subcapitata because the direction of evolution in this region appears to be from common habitats to rare, more severe habitats. Cryptantha caespitosa is found on a variety of relatively common substrates in- cluding sandy knolls, rocky slopes, and ridges. Cryptantha subcapitata is restricted to relatively uncommon calcareous substrate, a habitat in which many Wyoming endemics or near-endemics are found and in which common species are relatively few. ACKNOWLEDGMENTS We thank Arthur Cronquist and Mary Barkworth for their helpful review comments on the manuscript. LITERATURE CITED HIGGINS, L. C. 1971. A revision of Cryptantha subgenus Oreocarya. Brigham Young Univ. Sci. Bull., Biol. Ser. 13(4):1-63. (Received 11 Dec 1980; revision accepted 28 Jan 1981.) ERIOGONUM LIBERTINI (POLYGONACEAE), A NEW SPECIES FROM NORTHERN CALIFORNIA JAMES L. REVEAL Department of Botany, University of Maryland, College Park 20742 ABSTRACT Eriogonum libertini is a new species of the subgenus Oligogonum known only from serpentine outcrops in the California Coast Ranges of Shasta, Tehama and Trinity Counties. It is most closely related to E. ternatum Howell, a serpentine species of northernmost California and adjacent Oregon, differing in having a whorl of three foliaceous bracts at midlength along the stem and an inflorescence reduced to a single terminal involucre. Eriogonum libertini Reveal, sp. nov. A E. ternato Howell foliis brevioribus et semper tomentosis insuper, bracteis 3, foliaceis, inflorescentiis solitaris, floribus majoris (5-8 mm longis nec 3—5 mm longis) cum stipibus longior (1-1.5 mm nec 0.3- 0.6 mm longis) differt (Fig. 1). Low, spreading herbaceous perennial forming a rather dense mat to 4 dm across, with a densely branched, spreading, woody caudex arising from a stoutish, well-defined, woody taproot; leaves in small, well-defined, densely congested rosettes at tips of exposed caudex branches, the leaf-blades numerous, oblong to elliptic or rarely round- ed, 0.5—1(-1.5) cm long, 3—5(—7) mm wide, densely white-tomentose below, thinly tomentose and greenish above even at maturity, often drying blackish in age, the apices acute to obtuse, the bases obtuse, with plane, entire margins, on slender petioles 2—6 mm long, thinly tomentose throughout or at least near the leaf-base, otherwise ciliate marginally, the petiole-bases narrowly triangular, 0.5-3 mm long, 0.4—1 mm wide, glabrous or nearly so on both surfaces; flowering stems erect, slender, 5-15 cm long, thinly floccose, with a whorl of three foliaceous bracts about midlength; inflorescences capitate and terminal, 1—1.5 cm across; bracts lacking below involucre, restricted to middle of stem, these narrowly lanceolate to narrowly elliptic, 5—7(—10) mm long, 0.8-— 2.5 mm wide, tomentose on both surfaces although less so above, greenish-white to white; peduncles (that portion of stem between bracts and involucre) erect, slender, 2—4.5 cm long, thinly floccose; involucres solitary, turbinate-campanulate to campanulate, the involu- cral tube 4—8 mm long, 5—6(—8) mm wide, densely tomentose without, glabrous within, with 5-8 erect to slightly spreading triangular teeth 0.5—1(—1.5) mm long, the bractlets linear, 4.5—9 mm long, hirsutulous MADRONO, Vol. 28, No. 3, pp. 163-166, 31 July 1981 164 MADRONO [Vol. 28 Piz = >>>500 0 ¢€ EE g ¥£ oa > “) OO00O000 oOo o 6S Qa Qa o Oo mp) OZ22200 465 S$ PSS adass Date Fic. 3. Number of Stipa pulchra seedlings and germinable seeds in a sample of 125 6.45 cm? plots at the Hopland Field Station, Mendocino County during the 1974-1975 growing season. Species interactions. Because annual plant density was high on the study plot, as it generally is wherever Stipa is observed, further investigation of the interaction of S. pulchra with annual species was initiated. Regardless of the historical role of S. pulchra in the pristine native grassland, today a major factor in its persistence is the nature of interactions with the introduced annual grasses. The key period for observing such interactions would be during germination and seedling establishment. Poor early germination or growth would precipitate later spring mortality. We examined S. pulchra in two pot experi- ments, one growing S. pulchra alone and the second in combination with annual grasses. In both experiments, S. pulchra seeds were sown in plastic pots 14.6 cm in diameter. Pots were divided into blocks consisting of seeds gathered from three different locations (two coastal and one inland valley site). Cultures were grown at Oxford Tract at the University of California, Berkeley. The first experiment was conducted indoors in a greenhouse from 30 January to 30 March 1978. The interspecific experiment matured outdoors on benches from 30 May to 30 July 1978. In both experiments, observations were taken most frequently 1981] BARTOLOME AND GEMMILL: ECOLOGY OF STIPA PULCHRA 179 15 10 peso ere SesreamfAooonsnsaneraag/) Us Aa VIGISTURE LEVERS 5 Gg pots allowed to lose: - I 5S 15 = ~ O 10 MULCH LEVELS. seeds covered with mulch in amounts equivalent to: ® @ o 10) An--—==% 500 (560) O90 _~—«q 1000 [1120] lbs./acre, or (kg./hectare) 7 9 11:13 20 28 42 58 TIME 3 (DAYS Fic. 4. Cumulative average number of Stipa pulchra seeds germinating out of 25 under four moisture regimes and three mulch levels in pots. 180 MADRONO [Vol. 28 toward the beginning, during the initial rapid growth period, and less frequently later as growth rates leveled off. In order to follow plant growth through time, nondestructive mea- sures of growth (number of germinated seeds, length of the longest leaf, number of leaves and number of tillers) were recorded in the above experiments. A separate height—weight experiment was con- ducted to determine the relationship of these parameters to early- growth plant biomass. Length of longest leaf and number of leaves correlated best with biomass (r*? = 0.84044, p < 0.001 and r = 0.83716, p < 0.001, respectively) confirming the validity of using these measures to describe plant growth. For the first experiment, S. pulchra seeds were sown alone at den- sities of 25 seeds per pot, and were subjected to three different levels of mulch cover and four different treatments of moisture stress. Mulch treatments consisted of leached and dried annual grass straw placed on top of the pots in weights corresponding to 0, 560, and 1120 kg/ha. For the moisture treatments, an alteration of wet and dry periods was induced, because it corresponds more closely to actual field conditions than does maintaining moisture at a constant level. With a soil ten- siometer, the approximate relationship of pot weight to four soil water potentials (—3, —6, —12, and —15 bars) was established. Pots were allowed to dry down to established weights, then were rewatered. Instrumentation, and the fact that measurements were taken in rela- tively shallow pots compared to field soil depths, render only very approximate estimates of actual potential; thus the treatments should be thought of in terms of a gradation from constantly moist to infre- quently moist. Results of the above experiment showed that the final number of germinated seeds per pot differed significantly between seed sources, suggesting that there may be considerable genetic variability in natural populations. Of the 25 seeds per pot, an average of 12.75, 12.0, and 6.85 seeds germinated from Mendocino, Marin, and San Joaquin County sources, respectively. Germination patterns were much more strongly impeded by treat- ments throughout the experiment than was growth of established seed- lings. Germination was highly significantly suppressed both by the highest levels of moisture stress and the highest levels of mulch (Fig. 4). In the second experiment, constant densities of perennial grass seeds were sown in pots in combination with varying densities and species of annual grasses. Encircling five S. pulchra seeds in plastic pots 14.6 cm in diameter were annual seeds at densities of 1, 3, 9, 18, and 54 seeds per Stipa seed. These numbers of annuals correspond to field densities of 300, 900, 2700, 5100, and 16,000 seeds/m?, typical of low to moderate densities observed in the field (Bartolome, 1979). 1981] BARTOLOME AND GEMMILL: ECOLOGY OF STIPA PULCHRA 181 To reduce genetic variation in the annual grasses, we used com- mercial strains of Bromus mollis and Festuca megalura. Festuca me- galura, often mistakenly referred to as a native species (Lonard and Gould, 1974), commonly occurs in S. pulchra stands. Bromus mollis is the most widely distributed annual grass species in California (Janes, 1969). Results from pot trials support field observations reported above. Stipa pulchra seeds germinated more slowly and attained a lower density after 40 days with high densities of both Festuca megalura and Bromus mollis (Fig. 5) compared to control. Seedlings grew more slowly as represented by fewer leaves at high density (Fig. 5). Of particular interest, however, is the different effect of the two annual species. Bromus mollis appeared to have a much more detrimental effect on perennial seedlings than Festuca megalura. This effect shows up particularly well when comparisons are made between the two species averaging overall values for density versus both number of seeds germinating and leaf number per seedling. Bromus mollis showed a significant negative correlation with density and number (r = —0.7554, p < 0.01) and size (r = —0.5979, p < 0.01) of peren- nial seedlings at 40 days, whereas the relationship for Festuca, al- though present, is not significant at the 5 percent level. Values for length of the longest leaf and number of tillers show the same results as for number of leaves, and thus are not presented graphically. CONCLUSIONS Evidence presented helps clarify the ecological role of Stipa pulchra in California grasslands. Although S. pulchra has often been described as the dominant and even the equivalent of undisturbed California prairie, the results and observations above suggest a lesser role. Permanent transects at the Hopland Field Station failed to show an increase of S. pulchra over twenty years of protection from grazing by livestock. Indeed, S. pulchra decreased with protection on the Riley area, was replaced by other perennial grasses in one cluster, and by annuals in another. This lack of directional change is supported by similar Stipa stands examined at the Hastings Reservation by White (1967). The Soil—Vegetation Survey shows that S. pulchra occurs on a wide variety of sites, with broad distribution in the northern part of the state, and probably similar distribution throughout its range. S. pulchra is far from a rare or endangered species. Experimental results also support the idea that S. pulchra is op- portunistic, with few of the characteristics of typical climax species. S. pulchra germinates readily under all but severe moisture stress. It appears to establish most readily on bare ground, rather than under a cover of mulch. One might anticipate a climax species to prefer a 182 MADRONO [Vol. 28 O { A. WITH FESTUCA MEGALURA 9 pP 84 7 a DENSITY, plants / pot 7 > 5 a4 3 pm am o 3 QQ Se 2 2} c 1 14 18 23 29 40 ep) 2 5 ioe 4 53 3 fe) ~ Soe 21 o (@)) 10 9 n 8 oat oO © 6 w 5 oe) 4 = 3 — 24 = ae | 14 18 23 29 40 ep) eo Ss HB 4 o Oo a oe 2s 2 ae Ec =) 5 ooo 2 soesrscon (@)) 40 TIME , DAYS Fic. 5. Cumulative number of leaves per plant and number of germinated seeds of Stipa pulchra germinated in pots with six densities of A. Festuca megalura and B Bromus mollis. 1981] BARTOLOME AND GEMMILL: ECOLOGY OF STIPA PULCHRA 183 higher litter cover or at least be adapted to such conditions. It cannot compete successfully against a rapidly-growing, robust annual such as Bromus mollis, yet S. pulchra possesses some similar germination features (rapid germination, germination on bare sites) to those of the annuals with which it now grows. Where the annual cover is dense, S. pulchra seedlings often do not survive. Where the annual cover is reduced either by fire, grazing, or disturbance, S. pulchra seedlings thrive. Once established, a S. pulchra plant can persist under mod- erately heavy grazing. Its vigorous seeding habit and substantial quantities of viable seed place S. pulchra in a position to occupy suitable sites rapidly following disturbance of the annual cover. Stipa pulchra may have occupied a similar status in the pristine grassland, occupying areas of disturbance such as land slips and burned areas. Thus when the first ecologists observed relicts of the grassland they saw S. pulchra, not as the rem- nant climax species, but the native perennial species best able to thrive under disturbance. Far from being the dominant of the California prairie, it was a survivor because it is adapted to disturbance and does well when not grazed heavily in the spring. The dominants of the original prairie remain undetermined. The major questions remain: What was the nature of the pristine grassland, and did the new annuals from the Mediterranean replace native annuals or native perennials? What were the dominants of the pristine vegetation? These questions may be answerable through an expanded application, i.e., more native species, of the methods used in this study. Clearly, the evidence to date points to a role for S. pulchra not necessarily as the dominant pristine grass of California, but a native species adapted to disturbance. Stipa pulchra is clearly not in danger of elimination, having survived under 200 years of heavy grazing and frequent burning. ACKNOWLEDGMENTS We are grateful for the assistance of Harold Heady, Michel McClaran, Anne Ewing, Randy Rosiere, Robert Powell, Barbara Kosco, and Vincent Berg at various stages in the research and writing of this article. LITERATURE CITED BAKER, H. G. 1976. Invasion and replacement in California and neotropical grass- lands. In J. R. Wilson, ed., Plant relations in pastures. p. 368-384. CSIRO, Canberra. Barry, W. J. 1972. The Central Valley prairie. Calif. Dept. Parks and Rec., Sacra- mento. BARTOLOME, J. W. 1979. Germination and establishment of plants in California an- nual grassland. J. Ecol. 67:273-282. BEETLE, A. A. 1947. Distribution of the native grasses of California. Hilgardia 17:309- S57. BISWELL, H. H. 1956. Ecology of California grasslands. J. Range Managem. 9:19-24. BurcHAM, L. T. 1957. California rangeland. California Div. Forestry, Sacramento. 184 MADRONO [Vol. 28 CLEMENTS, F. E. 1920. Plant indicators. Publ. Carnegie Inst. Washington 290. GREEN, L. and J. R. BENTLEY. 1957. Seeding and grazing trials of Stipa on foothill ranges. U.S. For. Serv. Res. Note 128. Heapy, H. F. 1958. Vegetational changes in the California annual type. Ecology 39:402-416. . 1977. Valley grassland. In M. G. Barbour and J. Major, eds., Terrestrial vegetation of California. p. 491-514. Wiley-Interscience, New York. HULL, J. C. and C. H. MULLER. 1977. The potential for dominance by Stipa pulchra in a California grassland. Amer. Mid]. Naturalist 97:147-175. JANES, E. B. 1969. Botanical composition and productivity in the California annual grassland in relation to rainfall. M.S. thesis, Univ. California, Berkeley. KUCHLER, A. W. 1977. Map of the natural vegetation of California. Jn M. G. Barbour and J. Major, eds., Terrestrial vegetation of California. p. 909-938. Wiley- Interscience, New York. LONARD, R. I. and F. W. GouLp. 1974. The North American species of Vulpia (Gramineae). Madrono 22:217-230. Munz, P. A. and D. D. Keck. 1949. California plant communities. Aliso 2:87-105. ROBINSON, R. H. 1968. An analysis of ecological factors limiting the distribution of a group of Stipa pulchra associations within the foothill woodland of California. Ph.D. Dissertation, Oklahoma State Univ., Stillwater. SAMPSON, A. W. and E. C. McCarty. 1930. The carbohydrate metabolism of Stipa pulchra. Hilgardia 5(4):61—100. SAVELLE, G. D. 1977. Comparative structure and function in a California annual and native bunchgrass community. Ph.D. Dissertation, Univ. California, Berkeley. WHITE, K. L. 1967. Native bunchgrass (Stipa pulchra) on Hastings Reservation, California. Ecology 48:949-955. (Received 3 Jul 1980; accepted 26 Dec 1980; revision received 30 Jan 1981.) NOTEWORTHY COLLECTIONS LUPINUS CITRINUS Kell. (FABACEAE).—USA, CA, Madera Co., Indian Lakes Es- tates, Rd. 417, 3.1 kme. of jct. with Hwy 41 (T8S R21E S828 ne.%4), 685 m; 8 May 1974, Wells s.n. (CAS); 11 May 1980, Hamon 8042A and 8042B (UC); 8 Jun 1980, Bartel 1019 (UC). Open decomposing granite outcrops in digger pine/oak woodland. Associated with Calyptridium pulchellum, Cryptantha flaccida, Mimulus bicolor, M. dudleyi, Parvisedum congdonii, and Pectocarya penicillata. Previous knowledge. Reported from Fresno and Mariposa Cos. (Munz, A Calif. fl. 1959) and also as Fresno Co. endemic (Jepson, Fl. Calif. 2:277. 1936; Abrams, Ill. fl. Pacific states 2:494. 1944). Mariposa Co. reports erroneous and probably based on either incorrect county notation on label [28 May 1903, Congdon s.n. (UC)] or label data transposed during remounting [11 May 1902, Congdon s.n. (MIN)]. Significance. First record for Madera Co., a range extension wnw. of 29 km. Con- sidered rare and endangered by CNPS (Smith et al., CNPS Spec. Publ. 1, ed. 2. 1980). Under review as endangered species by the U.S. Fish & Wildlife Service (Fed. Reg. 45:82520. 1980). STREPTANTHUS FARNSWORTHIANUS J. T. Howell. (BRASSICACEAE).—USA, CA: Madera Co., Mammoth Pool Rd., 3.5 km e. of jct. with Italian Bar Rd. (T9S R23E 1981] NOTEWORTHY COLLECTIONS 185 S2 ne.4 nw.%), 1000 m: 29 May 1977, Hemphill s.n. (PUA); 8 Jun 1980, Bartel 1020 (UC), scattered on exposed slate slope in Quercus douglasii woodland, associated with Avena barbata, Lupinus benthamii, Mimulus guttatus, Pellaea mucronata, and Trifo- lium tridentata; Fresno Co., Petersen Mill Rd., 0.4 km e. of jct. with old Tollhouse Rd. (T10S R24E S17 sw.% sw.%), 1220 m, 21 Jun 1980, Bartel 1022 (UC), open granite slope in Pinus ponderosa-Quercus chrysolepis mosaic, associated with Arctostaphylos viscida, Brodiaea elegans, Bromus carinatus, Lupinus stiversit, Pellaea mucronata, and Penstemon laetus. Previous knowledge. Known from Kern, Tulare, and Fresno Cos. from metamor- phic slate substrate (Howell, Leafl. W. Bot. 10:182—183. 1965). Significance. First record for Madera Co., a range extension nw. of 28 km. First collection from granite. Recently numerous other Fresno Co. stands have been noted by the author and officers of the Sierra Natl. For. on granite in the Tollhouse area and on slate near Pine Flat Reservoir. Voucher specimens are not yet deposited. Many colonies on granite appear to be new invasions of open rock outcrops and roadcuts, suggesting a recently evolved ecotype. Considered rare but not endangered by CNPS (Smith et al., CNPS Spec. Publ. 1, ed. 2. 1980). Dropped from review as potentially endangered by the U.S. Fish and Wildlife Service (Fed. Reg. 45:82559. 1980).—Jim A. BARTEL, U.S. Fish and Wildlife Service, Endangered Species Office, 1230 “N” Street, Sacramento, CA 95814. (Received 27 Oct 1980; accepted 13 Nov 1980; revised version received 5 Jan 1981.) THELYPODIOPSIS PURPUSII (Brandegee) Rollins (BRASSICACEAE).—USA, NM, Socor- ro Co.: Sevilleta Wildlife Refuge, Los Pinos Mts., canyon directly e. of Nunn-Burris Ranch site, ca. 1825 m: 19 Apr 1975, Manthey 27 (UNM); 2 May 1980, Spellenberg and Ward 5484 (NMC); Sepultura Canyon, ca 1800 m, 19 Apr 1975, Manthey 55 (UNM); Dona Ana Co., s. end of San Andres Mts., ne. side of Black Mt. (T20S RSE S31 s.-center), 1700-1800 m, 7 May 1980, Spellenberg and Todsen 5497 (GH, NMC); Otero Co., 5.8 kme. of Hwy 70 e. of Alamogordo in Marble Canyon (T16S R10E 822 se.4%4), 1850 m, 11 May 1980, Spellenberg 5501 (GH). At each site the Thelypodiopsis was associated with Juniperus monosperma and various shrubs and perennial grasses that commonly occur with this tree. Plants were rare at all sites in 1980, possibly due to a “poor” year. Manthey’s collections are much more robust; the winter of 1974-75 was considerably wetter. In 1980, plants grew on steep slopes in shelter of rocks or at the protected bases of cliffs and gully banks in relatively inaccessible areas. The species probably is more frequent than collections indicate because spring in NM is often dry and collecting, therefore, not very rewarding. Previous knowledge. The species has been known for about 75 years from Coahuila, and within the last 20 has been found in w. TX, s. NM, and n. AZ. Its existence in NM was known from a single collection in Luna Co., Spellenberg 3002, made in 1973. (Herbaria consulted: GH, NMC, UNM; published sources: Rollins, Contr. Gray Herb. 206:1—18; Wooton and Standley, Fl. New Mex., Contr. U.S. Natl. Herb. 19. 1915. R. Rollins, pers. comm.). Significance. Besides the new county records and the indication that this species is not especially rare in NM, our 1980 collections clarify the nature of Thelypodium vernale Wooton & Standley. That species is known only from the type collection made in 1908 in “low mountains west of San Antonio”, Socorro Co., NM (Wooton & Standley, 1915). This collection apparently is lost, as noted in a revision of Thelypodium,in which T. vernale is excluded from Thelypodium and tentatively referred to Sisymbrium (Al-Sheh- baz, Contr. Gray Herb. 204:1-148. 1973.). We easily “keyed” our collections to T. vernale, and the plants match well Wooton and Standley’s description. Rollins, after viewing our material, agreed that 7. vernale and our material are apparently the same, 186 MADRONO [Vol. 28 but stated that they are properly placed in Thelypodiopsis purpusii, based on the earlier published Thelypodium purpusii Brandegee (Brandegee, Zoe 5:232, 1906). NEMACLADUS GLANDULIFERUS Jepson var. ORIENTALIS McVaugh (CAMPANULACEAE). —USA, NM, Hidalgo Co., in the s. end of the Sierra Rica, ca. 0.4 km w. of the Mexico bor- der and 3.2 kms. of the upper corner of the NM “boot-heel” (T29S R14W S36 ne. 1%), 1440 m, 14 May 1980, Spellenberg and Ward 5520 (NMC, TEX). Gravelly limestone in an arroyo, very local, only 4 plants, with Larrea tridentata, Calliandra eriophylla, and other shrubs of the Chihuahuan and Sonoran deserts. . Previous knowledge. Widespread from s. CA and sw. UT tos. AZ and nw. MEX, and known from a number of collections made during activities of the Mexican Bound- ary Survey in 1852 in the vicinity of present-day El Paso, TX (McVaugh, Amer. Mid. Naturalist 22:521-550. 1939). (Herbaria consulted: ARIZ, ASC, MO, NMC, TEX, UNM, Western NM Univ.; published sources: Correll and Johnston, Man. vasc. pl. Tex. 1970; Martin and Castetter, Checklist gymnosp. angiosp. New Mex. 1970; McVaugh, 1939; Wooton and Standley, Fl. New Mex., Contr. U.S. Natl. Herb. 19. 1915). Significance. This is the only collection of Nemacladus that unequivocally origi- nates in NM, although “stony hills near Frontera” as cited by McVaugh for Sonoran records of this species by Charles Wright in 1852 refers to a low range of hills in extreme s. Dona Ana Co., NM, and adjacent CHIH, just w. of present-day El] Paso, TX (Gray, Pl. Wrightianae, Tex.-_NM, II. 1852; Torrey, Rep. Mex. Bndy. Surv., II, Botany. 1858). Localities of collections stated to be from NM, “in the valley of the Rio Grande below Donana” (McVaugh, 1939) are also indefinite. Dona Ana is the point in NM at which the border turned west from the Rio Grande prior to the Gadsden Purchase. Our collection is the first to be made in the general region in nearly 130 years. It is ca. 165 km se. of the nearest site in AZ in Graham Co. and about that far w. of the early record from near E] Paso, TX. The plants are inconspicuous, and though probably not fre- quent, they simply might have been overlooked in the intervening years. —DARRELL WARD and RICHARD SPELLENBERG, Department of Biology, New Mexico State Uni- versity, Las Cruces, 88003. (Received 4 Dec 1980; revised version received and accepted 17 Feb 1981.) CAREX DEWEYANA Schwein. subsp. DEWEYANA (CYPERACEAE).—Mexico, Edo. de Hidalgo, Real del Monte, near Pachuca, Cupressus forest, 2850 m, 27 Aug 1944 (in fruit), E. Hernandez X.-462 (MSC). Verified by F. J. Hermann, Feb 1979. Previous knowledge. Range: Lab. and Newf. to sw. Mack. and AK, s. to PA, OH, n. IA, CO, UT and B.C. The weakly differentiated subsp. leptopoda (Mack.) Calder - & Taylor (incl. C. deweyana var. bolanderi [Olney] Boott), the characteristic phase of the cordilleran region, extends from B.C. to nw. MT, s. to s. CA, AZ and NM; also in e. Asia. (Herbaria consulted: F, MICH ex herb. F. J. Hermann, MSC, WIS; pub- lished sources: Braun, Monocotyledoneae [of Ohio]. 1967; Calder and Taylor, Canad. J. Bot. 43:1389-1391. 1965; Great Plains Fl. Assoc., Atlas fl. Great Plains. 1977; Har- rington, Man. pl. Colorado, ed. 2. 1964; Hermann, Man. Carices Rocky Mts. and Colorado Basin. 1970; Hermann, Man. Genus Carex in Mexico and Central Amer. 1974; Hitchcock et al., Vasc. pl. Pacific Northw., pt. 1. 1969; Johnston, J. Arnold Arb. 25:49-50. 1944; Mackenzie, in N. Amer. Fl. 18, pt. 3:114-117. 1931; Matuda, Las Ciperaceas Edo. Mexico. 1959; Sanchez, Fl. Valle Mexico. 1978.) Diagnostic characters. Keys to Carex bromoides Willd. in Hermann (1974), but clearly set off from that species by its thick, oblong-lanceolate, broader (1.2—1.6 versus 0.8-1.2 mm wide, ca. 3—3.5 instead of 4—5 times as long as wide) perigynia, the dorsal faces of which are nerveless or faintly nerved at base rather than conspicuously several- 1981] NOTEWORTHY COLLECTIONS 187 nerved; by its paler, usually wider (2—5 rather than 1—2.5 mm wide) leaves; by its thinner and, except for the green central zone, whitish-translucent (instead of often orangish-tinged) pistillate scales; and by its oblong-ovate, broader (1.2—1.6 mm wide, ca. 1.5 times as long as wide) achenes. In C. bromoides the achenes are 0.8—1.1 mm wide and ca. 2—2.5 times as long as wide. Significance. New to Mexico; 2030 km disjunction from nearest known populations in Las Animas and Larimer Cos., CO. One other very wide-ranging temperate species, Carex interior Bailey, has a similar range, but it extends farther s. on the Great Plains and reappears in Chihuahua and Distrito Federal, Mexico.—THEODORE S. COCHRANE, Department of Botany, University of Wisconsin, Madison 53706. (Received 10 Dec 1980; accepted 23 Feb. 1981.) WOLFFIA COLUMBIANA Karst. (LEMNACEAE).—USA, CA, San Diego Co., San Dieguito River, 1.5 km sw. of Lake Hodges Dam, s. side of Hwy S-6 (33°2'N, 117°8’W), 76 m, 26 Sep 1980, Armstrong s.n. (SD 106362). Forming dense populations at surface of quiet ponds, mixed with W. punctata with combined density of 100-150 per cm? of water surface. Associated with Lemna gibba, Azolla filiculoides, Cyperus erythrorhizos, Pluchea purpurascens, Echinodorus berteroi, Paspalum distichum, Eclipta alba, and Scirpus acutus. Previous knowledge. Known from Canada, e. US, Mexico, s. to Colombia. A minute, free-floating rootless angiosperm, barely visible without magnification. Often associated with Lemna, Spirodela, and Azolla. The genus has undoubtedly been over- looked many times because of its small size. (Herbaria consulted: RSA, SD; published sources: Daubs, Ill. Biol. Monogr. 34. 1965; Mason, A fl. marshes Calif. 1957.) Jt Fic. 1. Dense population of Wolffia from the San Dieguito River, San Diego Co., CA. A. Wolffia columbiana. B. W. punctata. Scale bar is 1 mm. 188 MADRONO [Vol. 28 Significance. First record of W. columbiana in s. CA, a se. extension of 387 km from Oso Flaco Lake, San Luis Obispo Co. This species is clearly distinguished from W. punctata by its globose frond which is minutely roughened, but not flattened, on the dorsal surface (Fig. 1).—WAYNE P. ARMSTRONG, Palomar College, San Marcos, CA 92069. (Received 26 Nov 1980; accepted 23 Feb 1981.) CALYPTRIDIUM PULCHELLUM (Eastw.) Hoov. (PORTULACACEAE).— USA, CA: Mar- iposa Co.: three small, widely separated populations on ridge e. of Ben Hur Rd. (T6S R18E $14 e.% se.% and $24 nw.4 nw.%), 592 m, 1 May 1980, Hamon 8019 (UC, FSC): this site is thought to be Pea Ridge, the type locality, found with Lupinus deflexus and a pale color form of Lupinus stiversii; granite dome 100 m w. of Mariposa Cr. and 400 m n. of Buckeye Rd. on the Jack Kirk Ranch (T6S R18E $11 se.%4 nw.%), 460 m, 11 May 1980, Hamon 8064 (UC, FSC, HSC), with Lupinus deflexus and Streptanthus diversifolius; Madera Co.: s. slope of small hill, 400 m w. of Ahwahnee (T6S R20E S36 ne. 4), 730 m, 26 May 1980, Hamon, pers. obs. (no collection made because of extremely small, impacted population), associated with Lupinus stiversii; Indian Lakes Estates, 3 km e. of SH41 on Road 417 (Picyune Rd.) two populations 400 m apart (T8S R21E 828), 610 m: 11 May 1980, Hamon 8049 (UC, FSC); 17 May 1980, Hamon 8067 (UC): these are the largest populations noted, associated with Lupinus citrinus, Mimulus layneae, Streptanthus diversifolius; Fresno Co., decomposed granite outcrop on e.-facing slope 2 km sw. of Sugarloaf Hill, Sierra Natl. For. (T9S R24E S30 sw.% sw.%), 1097 m, 20 May 1980, Hamon 8078 (UC) associated with Lupinus citrinus and Camissonia hirtella. Previous knowledge. Collected only twice; originally at “Pea Ridge” by J. W. Cong- don on 19 April 1901 (Eastwood, Bull. Torrey Bot. Club 29:79. 1902) and then by R. F. Hoover in 1938 (Hoover, Leafl. W. Bot. 2:222—225. 1940). Thought to occur only in the type locality (Hinton, Brittonia 27:197-208. 1975) and categorized as possibly extinct (CNPS Spec. Publ. 1. 1980). Diagnostic characteristics. Diminutive annual with sparsely fibrous root system. Inflorescence terminal and paniculate, stigma not sessile, inserted anthers a pale, rose- red that fades to yellow on drying. Significance. Rediscovery of species presumed extinct, with new records for Fresno and Madera Counties. Only one population found on public land (Sierra Natl. For.) with all others endangered by foothill real estate development. All populations were only a few meters in diameter with few individual plants—DAN HAMON, 2823 E. Lansing Way, Fresno, CA 93726. (Received 9 Jan 1981; accepted 18 Feb 1981.) NOTES AND NEWS SPECIFIC STATUS FOR Trifolium haydenti VAR. barnebyi (FABACEAE).—In 1947 H. Dwight Ripley and Rupert Barneby collected an unusual Trifolium in the foothills of the Wind River Mountains in Wyoming. This collection was referred by Gillett (Canad. J. Bot. 50:1975-2007. 1972) to T. gymnocarpon Nuttall, with the comment that, for- wardly directed hairs on the ovary “eliminates the possibility of its being 7. haydenii as originally identified. The leaflets, too, which are quite glabrous, fit the shape of those of T. gymnocarpon.” More recently Isely (Brittonia 32: 55-57. 1980) described T. hay- denii Porter var. barnebyi from the same collection. Both Gillett and Isely had only the original material to work with. This past summer we revisited the original locality, made extensive observations on the population, and collected more material. It was 1981] TABLE 1. Characteristic Distribution Habitat Habit Petioles Leaflets Leaflet shape Leaflet veins Peduncles Pedicels Calyx Calyx teeth Banner Ovary NOTES AND NEWS T. gymnocarpon OR, ID, & c. WY to CA, AZ, & NM Desert to low montane Cespitose Pubescent Pubescent dorsally 1—2 times as long as wide Wide spaced 5-10 pairs primary Pubescent Pubescent Pubescent About as long as tube 2.5—4 times as long as wide Pubescent 189 SELECTED CHARACTERISTICS OF THREE TAXA OF Trifolium. T. barnebyi T. haydenii c. WY Sagebrush-juniper zone Densely matted Pubescent Glabrous mostly Twice as long as wide Narrow spaced 12-18 pairs primary Pubescent Pubescent Pubescent usually 1.5—2 times as long as tube 1.5—2 times as long as wide Pubescent ID, MT, extreme nw. WY Alpine to high montane Loosely matted Glabrous Glabrous 1—1.5 times as long as wide Wide spaced 5-10 pairs primary Glabrous Glabrous Glabrous 1-2 times as long as tube 1.5—2 times as long as wide Glabrous immediately apparent that this was not 7. haydenii nor any other Trifolium known from this region. Superficially, the plants resembled 7. haydenii because of the mat- forming habit and lack of an involucre. The habitat, in the sagebrush-juniper zone, was inappropriate for T. haydenii, which is found mostly in alpine areas but descends occasionally into high montane forests. Furthermore, this location was considerably south of any known station for 7. haydenii. On closer examination the plants appeared much closer to T. gymnocarpon than to T. haydenii except for the mat-forming habit. Table 1 summarizes characteristics of the three taxa, and reveals that the “barnebyi” material is unique only in the venation pattern of the leaflets. This pattern, with veins more numerous and more closely spaced than in the other two species, is easily observed as different. There are four additional differences from 7. gymnocarpon and eight additional differences from 7. haydenii. These combined with field observations of the three taxa make it exceedingly difficult to consider the “barnebyi” material as conspecific with either of the other two species. We therefore elevate 7. haydenii var. barnebyi to specific status. Trifolium barnebyi (Isely) Dorn & Lichvar, stat. nov.—Tvifolium haydenii Porter var. barnebyi Isely, Brittonia 32:56. 1980. Mat-forming perennial to 5 cm high; petioles pubescent, 3-18 mm long; leaflets 3, 3— 11 mm long, glabrous, or rarely pubescent on dorsal midrib, mostly oblanceolate, usu- ally toothed at least above, closely veined with mostly 12-18 pair of primary veins on larger leaflets, the veins usually much less than 0.5 mm apart; peduncles and pedicels pubescent; involucre lacking; flowers 8-18, 8-13 mm long; calyx 4-6 mm long, pubes- 190 MADRONO [Vol. 28 cent at least in sinuses between teeth (rarely glabrous), the teeth mostly 1.5—2 times as long as tube; corolla whitish, drying brownish; ovary pubescent at least on dorsal suture; ovules mostly 1-4. Locally common on ledges of pale red sandstone and on sand pockets at base of sandstone outcrops at about 1950 m. elev. about 25 km sse. of Lander, Fremont Co., WY. Type: USA, WY, Fremont Co., 16 km (10 mi) s. of Perrin, 1950 m, 30 Jun 1947, Ripley and Barneby 8924 (Holotype: NY; isotypes: NY, ISC, RM!). Other collections studied: USA, WY, Fremont Co., T3IN R99W S25 sw.%, 1950 m, crevices of pale red sandstone, 27 Jun 1980, Dorn 3483 (RM); Lichvar 2955 (RM). The three species can be distinguished easily using the following key. 1. Plants clabrous throughout 2. 22.6 44-94.005 soe eee pe pe T. haydenii 1. Plants pubescent at least on peduncles, pedicels, and ovaries. 2. Plants mat-forming; leaflets glabrous except rarely on dorsal midrib. .......... ee ae gt ee ae ee Se TE ca one ar PET CEE os 2 T. barnebyi 2. Plants not mat-forming; leaflets obviously pubescent, at least dorsally. ........ eT ee eee Ee eT Te Me eee eae ee T. gymnocarpon Trifolium barnebyi is likely derived from T. gymnocarpon, even though the growth habits of the two are quite different. This view is supported by morphological similar- ities and by the dispersal and habitat. It is unlikely that 7. haydenii was ever as far south as the type locality of 7. barnebyi, for if it was, relict populations would likely have been found in the Wind River Mountains. Trifolium barnebyi is apparently ad- justed to a specialized sandstone habitat and is presumably endemic there. The few similar habitats, all within 30 km of the known population, have yet to be investigat- ed.—ROBERT D. DoRN, Box 1471, Cheyenne, WY 82001 and ROBERT W. LICHVAR, The Nature Conservancy, Wyoming Natural Heritage Program, 1603 Capitol Ave., #325, Cheyenne 82001. (Received 10 Nov 1980; accepted 30 Jan 1981; revised version received 10 Feb 1981.) Carex whitneyi OLNEY (CYPERACEAE): NOT ENDANGERED.—A survey was con- ducted from May through August 1980 to locate, document, and describe undiscovered populations of Carex whitneyi within its previously known range. The species is cur- rently on List 2 of the CNPS Inventory of Rare and Endangered Vascular Plants of California (Smith et al., 1980) and the previous known range was limited to Mariposa, Tuolumne and Fresno Cos., CA. Individual plants were identified and enumerated and culms were collected and ver- ified by J. T. Howell (Curator Emeritus, CAS). All specimens are deposited at CAS and FSC. The survey identified 35 populations (with a total of 4667 plants) in Fresno Co., 9 populations (2932 plants) in Madera Co., 10 populations (4444 plants) in Tuol- umne Co., 8 populations (432 plants) in Alpine Co., and 2 populations (42 plants) in Calaveras Co. Based on our review of available literature and herbarium and field investigations of morphological and habitat characteristics we conclude that the species should be de- scribed as: Carex whitneyi Olney.—Densely cespitose; culms 2.5—9 dm tall; blades flat or + revolute, 2-6 mm wide; spikelets 3 or 4; the terminal male, linear, 0.5—2.5 cm long, the lateral spikelets female, oblong or linear-oblong 1—3 cm long, 5-10 mm wide; female scales ovate, appressed-ascending, or + spreading at maturity, hyaline-mar- gined, 3-5 nerved; perigynia ovate, obovate, or elliptic-ovate, 3.5—-5 mm long, 1.5-—2.5 mm wide, tapering or contracted into a bidentulate or oblique beak 0.25—1.0 mm long, the beak somewhat hyaline. Habitat: dry to moist, often sandy; flat or moderate slopes; on the edge of meadows, in open to dense forests; often in disturbed soils where the surface litter has been removed; 1158 to 3658 m; Yellow Pine Forest to Subalpine Forest; Sierra Nevada from Tulare Co. to s. OR and w. NV. 1981] NOTES AND NEWS 191 We thank Peggy Smith, Leslie Zander, Mona Bourell, Bill Clark and Steve Lentz for their assistance in collecting data.—JOHN C. STEBBINS, JAMES R. SMITH, and JAMES R. HOLEMAN, Enphase, Inc. 1630 E. Menlo, Fresno, CA 93710. (Received 6 Jan 1981; accepted 23 Feb 1981.) ANNOUNCEMENT CALIFORNIA BOTANICAL SOCIETY—GRADUATE STUDENT MEETINGS The California Botanical Society Graduate Student Meetings will be held at SAN FRANCISCO STATE UNIV., 24-25 October 1981. The meeting will focus on the presen- tation of short research papers and reports in progress by graduate students in all botanical and plant related fields. Members and non-members are invited to participate. For further information please contact the Graduate Student Meetings Committee, Dept. of Biology, San Francisco State Univ., San Francisco 94132, or leave a message at (415) 469-1359. Dr. Harry D. Thiers will present a seminar Saturday evening on his recent work in Australia and the interesting fungal flora of that area. SUBSCRIPTIONS—MEMBERSHIP Membership in the California Botanical Society is open to individuals ($18 per year; students $10 per year for a maximum of seven years). Members of the Society receive MADRONO free. Family memberships ($20) include one eight-page publishing allotment and one journal. Emeritus rates are available from the Corresponding Secretary. In- stitutional subscriptions to MADRONO are available ($25). Membership is based on a calendar year only. Applications for membership (including dues), orders for sub- scriptions, requests and rates for back issues, changes of address, and undelivered copies of MADRONO should be sent to the Corresponding Secretary. Renewal payments should be sent to the Treasurer. INFORMATION FOR CONTRIBUTORS Manuscripts submitted for publication in MADRONO should be sent to the editor. All authors must be members, and membership is prerequisite for review. Manuscripts and review copies of illustrations must be submitted in triplicate for all articles and short items intended for NOTES AND NEWS. Follow the format used in recent issues for the type of item submitted and allow ample margins all around. All manuscripts MUST BE DOUBLE SPACED THROUGHOUT. For articles this in- cludes title (all caps, centered), author names (all caps, centered), addresses (caps and lower case, centered), abstract, 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, and number each page. Do not use a separate cover page, “erasable” paper, or footnotes. Table captions should include all information relevant to tables. All measurements should be in metric units. Line copy illustrations should be clean and legible, proportioned (including captions) to the MADRONO page, and designed for reduction to % original size. 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 latitude and longitude references. Halftone copy should be designed for reproduction at actual size. In no case should original illustrations be sent prior to the acceptance of a manuscript. When needed they should be mounted on stiff cardboard and sent flat. No illustrations larger than 22 < 28 cm will be accepted. Presentation of nomenclatural matter (accepted names, synonyms, typification) should follow the format used for Rhus integrifolia in MADRONO 22:288. 1974. Institutional abbreviations in specimen citations should follow Holmgren and Keuken, Index Her- bariorum, 6th edition. Abbreviations of serial titles should be those in Botanico-Per- iodicum-Huntianum (Lawrence et al., 1968, Hunt Botanical Library, Pittsburgh). If the correct abbreviation cannot be determined, the full serial title should be used. Titles of books should be given in full, together with the place and date of publication, publisher, and edition, if other than the first. All members of the California Botanical Society are allotted eight pages in MADRONO over a two year period. Joint authors are each assessed the full page number. Beyond that number of pages a required editorial fee of $40.00 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 sched- ule, with equity among all members for access to its pages. Printer’s fees for illustrations and typographically difficult material @ $35.00 per page (if their sum exceeds 30 percent of the paper) and for author’s changes after typesetting @ $3.00 per line will be charged to authors. Contents, continued NOTES AND NEWS SPECIFIC STATUS FOR Trifolium haydenii VAR. barnebyi, Robert D. Dorn and Robert W. Lichvar Carex whitneyi OLNEY (CYPERACEAE): NOT ENDANGERED, John C. Stebbins, James R. Smith, and James R. Holeman ANNOUNCEMENTS 188 190 191 CALIFORNIA BOTANICAL SOCIETY | M1S3 Bot, MADRONO VOLUME 28, NUMBER 4 OCTOBER 1981 Contents LOCAL FLORAS OF ARIZONA: AN ANNOTATED BIBLIOGRAPHY, Janice E. Bowers 193 A NEw SPECIES OF QUERCUS (FAGACEAE) FROM SOUTHERN CALIFORNIA, Kevin C. Nixon and Kelly P. Steele 210 A NEw SPECIES OF ACACIA (LEGUMINOSAE: MIMOSOIDEAE) FROM BAJA CALIFORNIA SUR, MEXICO, Annetta M. Carter and Velva E. Rudd 220 RE-ESTABLISHMENT OF ANGELICA CALIFORNICA (UMBELLIFERAE), Joseph M. DiTomaso 226 COMPOSITION OF NATIVE GRASSLANDS IN THE SAN JOAQUIN VALLEY, CALIFORNIA, Lyndon Wester 251 POST-ERUPTION SUCCESSION ON ISLA FERNANDINA, GALAPAGOS, Lynn B. Hendrix 242 A LATE PLEISTOCENE AND HOLOCENE POLLEN RECORD FROM LAGUNA DE LAS TRANCAS, NORTHERN COASTAL SANTA CRUZ COUNTY, CALIFORNIA, David P. Adam, Roger Byrne, and Edgar Luther 250 NOTES AND NEWS VARIATION IN IMMATURE CONE COLOR OF PONDEROSA PINE (PINACEAE) IN NORTHERN CALIFORNIA AND SOUTHERN OREGON, Richard H. Smith 202 INDEX TO VOLUME 28 207 - WEST AMERICAN JOURNAL OF BOTANY A 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, Herbarium, Life Sciences Building, University of California, Berkeley, CA 94720. Subscription rate: $25 per calendar year. Subscription information on inside back cover. Established 1916. Second-class postage paid at Berkeley, CA, and additional mailing offices. Return requested. POSTMASTER: Send address changes to Dr. Frank Almeda, Botany Dept., California Academy of Sciences, San Francisco, CA 94118. Editor—CHRISTOPHER DAVIDSON Idaho Botanical Garden P.O. Box 2140 Boise, Idaho 83701 Board of Editors Class of: 1981—DANIEL J. CRAWFORD, Ohio State University, Columbus JAMES HENRICKSON, California State University, Los Angeles 1982—DEAN W. TAYLOR, University of California, Davis RICHARD VOGL, California State University, Los Angeles 1983—-ROBERT W. CRUDEN, University of Iowa, Iowa City DUNCAN M. PORTER, Virginia Polytechnic Institute and State University, Blacksburg 1984—-MaAryY E. BARKWORTH, Utah State University, Logan Harry D. THIERS, San Francisco State University, San Francisco 1985—STERLING C. KEELEY, Whittier College, Whittier, CA ARTHUR C. GIBSON, University of California, Los Angeles CALIFORNIA BOTANICAL SOCIETY, INC. OFFICERS FOR 1981 President: ROBERT ORNDUFF, Department of Botany, University of California, Berkeley 94720 First Vice President: LLAURAMAY T. DEMPSTER, Jepson Herbarium, Department of Botany, University of California, Berkeley 94720 Second Vice President: CLIFTON F. SMITH, Santa Barbara Museum of Natural History, Santa Barbara, CA 93105 Recording Secretary: ROBERT W. PATTERSON, Department of Biology, San Francisco State University, San Francisco, CA 94132 Corresponding Secretary: FRANK ALMEDA, Department of Botany, California Academy of Sciences, Golden Gate Park, San Francisco 94118 Treasurer: CHERIE L. R. WETZEL, Department of Biology, City College of San Francisco, 50 Phelan Ave., San Francisco, CA 94112 The Council of the California Botanical Society consists of the officers listed above plus the immediate Past President, WAYNE SAVAGE, Department of Biology, San Jose State University, San Jose, CA 95192; the Editor of MADRONO; three elected Council Members: PAUL C. SILVA, University Herbarium, Department of Botany, University of California, Berkeley 94720; JOHN M. TUCKER, Department of Botany, University of California, Davis 95616; CHARLES F. QUIBELL, Department of Biological Sciences, Sonoma State College, Rohnert Park, CA 94928; and a Graduate Student Represen- tative, KENT HOLSINGER, Department of Biological Sciences, Stanford University, Stanford, CA 94305. LOCAL FLORAS OF ARIZONA: AN ANNOTATED BIBLIOGRAPHY JANICE E. BOWERS Office of Arid Lands Studies, University of Arizona, Tucson 85721 ABSTRACT There are more than 80 local floras for Arizona that provide information on current and past plant distributions. They are briefly summarized. Although many areas in Arizona have been intensively botanized, floras are still lacking for some important regions. Local floras, that is, relatively complete plant lists from defined areas, have value beyond simply cataloging the plants of an area. Local floras provide 1) valuable historical information, descriptions of new species, and documented range extensions; 2) data about the dis- tribution of plants, condensing locations from hundreds of herbarium specimens into a line or two of print; and 3) a baseline for following immigration and extinction of species in an area. Local floras also provide the data needed for floristic comparisons of areas (as in Raven, 1963; Schaak, 1970; Lane, 1976) and for creating and testing ecological models such as island biogeography (Harper et al., 1978), patterns of species diversity (Richerson and Lum, 1980), or species migration rates (Meyer, 1978). This bibliography is the first to concentrate on local floras of Ari- zona. Ewan (1936) lists and briefly summarizes a broad array of bo- tanical papers dating as far back as 1848. Schmutz (1978) published an extensive bibliography on native Arizona plants in which references are classified and cross-referenced according to subject, but not an- notated. Although both sources provided references for this paper, most are not cited in either bibliography. I have attempted to include all published and unpublished Arizona plant lists that appear to be reasonably complete. Incomplete lists are included when they are the only lists available for a particular area or when they are one of a series of plant lists for an area. Some very old floras, although obviously incomplete, are included because of their historical value. I also include a few plant lists from ecological studies, but omit popular guides to wildflowers or single life forms (e.g., cacti). Manuals covering large regions such as the Sonoran des- ert, Arizona or the southwest are not considered. Floristic work is continuing in Arizona. Although floras have been a frequent topic for masters’ theses (Lane, 1976; Lehto, 1970; Reeves, 1976; Schaak, 1970; Halse, 1973; and others), more and more floras MADRONO, Vol. 28, No. 4, pp. 193-209, 8 December 1981 194 MADRONO [Vol. 28 are being compiled at the direction of state and federal agencies and conservation groups. Several areas of Arizona have been neglected floristically (Fig. 1). For example, the dry mountain ranges of the lower Colorado River valley, although floristically distinct and containing a number of en- demic species, have not been treated as a single unit in any flora. Only part of this interesting region is covered by Simmons (1966). The White Mountains have been enthusiastically botanized by professional and amateur biologists for many years, but no comprehensive plant list has been compiled for this area. Floras have yet to be written for many major mountain ranges, including the Baboquivari, Pinaleno and Santa Rita Mountains. Still other areas (e.g., Grand Canyon) have been studied many times. The flora of an area is dynamic; some species immigrate and estab- lish, others disappear as their habitats are removed or changed. Any list is at best a static record of a continuously changing assemblage of plant species. NORTHERN ARIZONA 1. Grand Canyon McKee, E. D. 1934. Flora of Grand Canyon National Monument. Grand Canyon Nature Notes 9:316—321. Lists 87 taxa; classifies conspicuous plants by association; compares floras of Grand Can- yon National Park and Grand Canyon National Monument. In- complete. Patraw, P. M. 1936. Check-list of plants of Grand Canyon National Park. Grand Canyon Nat. Hist. Assoc. Bull. 6 [Grand Canyon]. Lists 635 taxa for 272,798 ha. Annotations include habitat, as- sociations, common names, local and regional distribution. Small- scale map shows collection localities. Clover, E. U. and L. Jotter. 1944. Floristic studies in the canyon of the Colorado and tributaries. Amer. Midl. Naturalist 32:591-642. Lists 490 taxa; study area was the Green and Colorado Rivers from Greenriver, Utah to Boulder Dam, Nevada. Correlates plant associations with habitat, elevation and latitude; annota- tions include collection locations, collection numbers, relative abundance, and habitat. McDougall, W. B. 1947. Plants of Grand Canyon National Park. Grand Canyon Nat. Hist. Assoc. Bull. 10 [Grand Canyon]. Lists 882 taxa for 272,798 ha; updates Patraw (1936). Annotations in- clude common names, collection locality, relative abundance, and elevational ranges. Deaver, C. F. and H. S. Haskell. 1955. Ferns and flowering plants of Havasu Canyon. Plateau 28:11—23. Lists 382 taxa for 210 ha. 1981] BOWERS: ARIZONA FLORAS 195 115° ee 109° Saw Navajo Mt | ct OF av jOGNEY 37° 1a e 4 Boysag Point lead 1 3 5 Z aN e Roaring Springs Lo ° Aenea ae, © Grandview Ge ees ARIZONA mle Se gen ANS. O 11e 012 N 13af- ed 13 13b nth Nant es. — — 17% 18° CENTRAL ARIZONA __camerc ake : = Monat = Phelps a 19, 24 \27 8 7cabin 24 20/) © 24%) 3 < 226 : oe — 236) pe IN ae RIZON 32 OUT eer 33C0 A STER | OUTHE RIZONA N Ss 30 “Ss 34 48 ® 37Q 35 320 =. aes Qe Tumamoc Hill 7 @“\Springs 36 38 115° Quitobaquito : Ec 40 8 (Daa 0 100 km el 41 43 40 109° Fic. 1. Location of local floras in Arizona. Northern Arizona: 1. Grand Canyon; la. Arizona Strip; 2. Lake Mead National Recreation Area; 3. Kaiparowits Basin; 4. Navajo National Monument; 5. Canyon de Chelly National Monument; 6. Window Rock; 7. Wupatki National Monument; 8. Sunset Crater National Monument; 9. San Francisco Mountain; 10. Bill Williams Mountain; 11. Flagstaff; 12. Walnut Canyon National Monument; 13. Oak Creek Canyon; 13a. Volunteer and Sycamore Canyons; 13b. Chevelon Canyon. Central Arizona: 14. Hualapai-Aquarius Planning Units; 15. Harcuvar Mountains; 16. Vulture Mountains; 17. Skull Valley; 18. Montezuma Castle National Monument; 19. Lake Pleasant Regional Park; 20. White Tank Mountains Regional Park; 21. McDowell Mountain Regional Park; 22. Phoenix; 23. Sierra Estrella Regional Park; 24. Three-Bar Wildlife Area; 25. Sierra Ancha; 26. Tonto National Monument; 27. White Mountains. Southwestern Arizona: 28. Cabeza Prieta Game Range; 29. Organ Pipe Cactus National Monument. Southeastern Arizona: 30. San Pedro River; 31. Aravaipa Creek; 32. Clifton; 33. Gila River; 34. Santa Catalina Moun- tains; 35. Rillito River; 36. Baboquivari Mountains; 37. Tucson Mountains; 38. Rose- mont; 39. Santa Cruz County; 40. Tumacacori Mission National Monument; 41. Syc- amore Canyon; 42. Canelo Hills; 43. Patagonia-Sonoita Creek Sanctuary; 44. Whetstone Mountains; 45. Chiricahua Mountains; 46. Huachuca Mountains; 47. Mule Mountains; 48. Hooker Cienega. 196 MADRONO [Vol. 28 Incomplete list, emphasizing spring and fall flora; discusses ge- ology and vegetation; lists crops grown by Havasupai Indians. McDougall, W. B. 1964. Grand Canyon Wildflowers. Mus. N. Ar- izona Bull. 43, Flagstaff. Lists 975 taxa for 272,798 ha; popular guide, relatively complete. About 120 species illustrated with pho- tographs; includes short species descriptions, glossary of botanical terms, keys to families and species. Schmutz, E. M., C. C. Michaels, and B. I. Judd. 1967. Boysag Point: a relict area on the North Rim of Grand Canyon in Ari- zona. J. Range Managem. 20:363—369. Ecological study; lists 88 taxa for 28 ha; discusses climate and physiography; compares floras and plant communities of Boysag Point with those of nearby mainland area. Bennett, P. S. 1968. Inventory of plants found near Hearst Tanks, South Rim, Grand Canyon. Jn Teachers Manual: Environmental Awareness, p. 77-83. Grand Canyon National Park, Grand Can- yon, Arizona 86023. Unpubl. report. Lists 185 taxa for Grand- view Natural Area; incomplete list based on one month of collec- tion; indicates introduced species. Phillips, B. G. and A. M. Phillips, III. 1974. Spring wildflowers of the Inner Gorge, Grand Canyon, Arizona. Plateau 46:149-157. Lists 186 taxa; emphasizes spring-flowering ephemerals; covers the Inner Gorge from Lee’s Ferry to Rampart Cave. Annotations include collection localities, common names, habitat, and eleva- tion. Notes new county records. Smith, E. L. 1974. The Grandview Natural Area. Jn Established Natural Areas in Arizona: a guidebook for scientists and educa- tors, p. 207-217. Office of Economic Planning and Development, Phoenix. Lists 104 taxa for 166 ha; partial checklist based on Bennett (1968); describes climate, topography, and plant com- munities. Vertebrates also listed. Phillips, A. M., III. 1975. Flora of the Rampart Cave area, lower Grand Canyon, Arizona. J. Arizona Acad. Sci. 10:148—159. Lists 270 taxa; annotations include common names, growth form, hab- itat, local distribution, and relative abundance; discusses habi- tats, lists characteristic species of each habitat. Bennett, P. S. 1978. Vascular plants, Grand Canyon National Park. Grand Canyon Computer Center. Unpubl. [computer printout] available from Western Archaeological Center, 1415 N. 6th Ave., Tucson, Arizona. Lists 1574 taxa for 493,000 ha; annotations include relative abundance, nativity, and special status (threat- ened, endangered, or sensitive); recently revised (B. G. Phillips, pers. comm.). Phillips, B. G., R. R. Johnson, A. M. Phillips, III, and J. E. Bowers. 1979. Resource values of the aquatic and riparian vegetation of Roaring Springs, Grand Canyon. Proc. Second Conf. Sci. Res. 1981] BOWERS: ARIZONA FLORAS 197 Nat. Parks, San Francisco, 1979. 4:141—-155. Natl. Techn. In- form. Serv., 5285 Port Royal Rd., Springfield, Virginia 22161. Lists 68 taxa; annotations include relative abundance, habit, and common names; discusses hydrology and plant associations; com- pares flora of Roaring Springs with that of three other springs in the Grand Canyon. la. Arizona Strip Gierisch, R. 1981. Herbarium list, Arizona Strip District. Unpub- lished list, USDI, BLM, Arizona Strip District, P.O. Box 250, St. George, Utah 84770. Lists 863 taxa for region north of Grand Canyon and south of Utah; incomplete list compiled from her- barium specimens at BLM, St. George. Plant collection in the Arizona Strip by BLM personnel is continuing. 2. Lake Mead National Recreation Area Holland, J. S., W. E. Niles, and P. J. Leary. 1979. Vascular plants of the Lake Mead National Recreation Area. Lake Mead Techn. Rep. 3. Biological Sciences, Univ. Nevada, Las Vegas. Lists 823 taxa for 607,500 ha. Discusses history of botanical collecting around Lake Mead; describes physiography and vegetation types. Annotations include habit, elevational range, local distribution, common names, collection locations, location of voucher speci- mens, relative abundance, threatened or endangered status, and synonymy. 3. Kaiparowits Basin Clute, W. N. 1919. A trip to Navajo Mountain. Amer. Bot. (Bing- hamton) 25:81-—87; . 1920. Notes on the Navago [sic] region. Amer. Bot. (Binghamton) 26:39-47. Clute (1919) described jour- ney from Flagstaff to Navajo Mountain; mentioned changes in vegetation with increasing elevation. Clute (1920) described lo- cation and habitat for collections in Nelson (1920). Nelson, A. 1920. Flora of the Navajo Reservation. Amer. Bot. (Binghamton) 26:48—56; 87-89; . 1922. Flora of the Navajo Reservation II. Amer. Bot. (Binghamton) 28:20—25. Lists 152 taxa. Incomplete list with historical interest; five new species are described. Annotations include relative abundance, habitat, species descriptions, and local distribution. Lindsay, D. A. 1959. Vascular plants collected in Glen Canyon, 1958. In Ecological Studies of the Flora and Fauna in Glen Canyon, p. 63-72. Univ. Utah Anthropological Papers 40, Glen Canyon Ser. 7. Lists 115 taxa for Glen Canyon between Hite, Utah and Lee’s Ferry, Arizona; annotations include common names, habi- tat, relative abundance, and coliection location; based on a four- week collecting trip. 198 MADRONO [Vol. 28 McDougall, W. B. 1959. Plants of the Glen Canyon area in the herbarium of the Museum of Northern Arizona. Unpublished. Museum of Northern Arizona, Flagstaff. Lists 304 taxa; compiled from collections made between 1936 and 1958; annotations in- clude collection location, number, and date, collector, relative abundance, and habitat. Gaines, X. M. 1960. An annotated catalogue of Glen Canyon plants. Mus. N. Arizona Tech. Ser. 4, Flagstaff. Lists 199 taxa. Based on a four-week collecting trip from Hite, Utah to the mouth of Kane Creek. Annotations include common names, collection lo- calities and numbers. Welsh, S. L., N. D. Atwood, and J. R. Murdock. 1978. Kaiparowits flora. Great Basin Naturalist 38:125-179. Lists 851 taxa for 1,400,000 ha; annotations include common names, collection lo- cations, collectors and collection numbers, habitat, elevational range, relative abundance, and local distribution. Lists threat- ened and endangered plants; discusses physiography. 4. Navajo National Monument Brotherson, J. D., G. Nebeker, M. Skougard, and J. Fairchild. 1978. Plants of Navajo National Monument. Great Basin Naturalist 38:19-30. Lists 293 taxa; annotations include habit, nativity, local distribution, collector, and collection date and number. Compares flora with that of Kaiparowits Basin, Arches National Park, and Uintah Basin; discusses species diversity within the Monument; describes climate, geology, and plant communities. 5. Canyon de Chelly National Monument Halse, R. R. 1973. The flora of Canyon de Chelly National Monu- ment. M.S. thesis, Univ. Arizona, Tucson. Lists 474 taxa for 33,955 ha. Includes keys to species, short species descriptions, and ethnobotanical information. Other annotations include col- lection numbers, local distribution, common names, and relative abundance. Discusses history and geology. Harlan, A. and A. E. Dennis. 1976. A preliminary plant geography of Canyon de Chelly National Monument. J. Arizona Acad. Sci. 11:69—78. Adds 44 taxa to flora; annotations include habit, rela- tive abundance, plant community, and local distribution for 112 perennials. Describes major plant communities. Schmutz, E. M., A. E. Dennis, A. Harlan, D. Hendricks, and J. Zauderer. 1976. An ecological survey of Wide Rock Butte in Canyon de Chelly National Monument. J. Arizona Acad. Sci. 11:114-125. Lists 76 taxa for 12 ha; annotations include abun- dance, frequency, and cover. Lists plant macrofossils from pack- rat midden dated at 6210 + 90 years B.P., concludes no important 1981] BOWERS: ARIZONA FLORAS 199 local changes in vegetation or climate have occurred for 6000 years. 6. Window Rock Bohrer, V. L. and M. Bergseng. 1963. An annotated catalogue of plants from Window Rock, Arizona. Navajo Tribal Museum, Window Rock. Lists 181 taxa within 16-km radius of Window Rock. Annotations include collection date, collection location, common name, habitat, elevation, plant association, and habit. 7, 8. Wupatki and Sunset Crater National Monuments McDougall, W. B. 1962. Seed plants of Wupatki and Sunset Crater National Monuments. Mus. N. Arizona Bull. 37, Flagstaff. Lists 268 taxa for 15,390 ha. Annotations include common names and short species descriptions; notes local distribution of species in the two Monuments. Includes keys to species, glossary of botanical terms. 9. San Francisco Mountain Little, E. L., Jr. 1941. Alpine flora of San Francisco Mountain, Ar- izona. Madrono 6:65—81. Lists 49 taxa for 518 ha; annotations include geographic range, habitat, relative abundance, and phe- nology. Describes plant associations and growth forms of alpine plants; discusses biogeography of some alpine species. Mentions endemics, lists 24 additional timberline species. Schaak, C. G. 1970. A flora of the arctic-alpine vascular plants of the San Francisco Mountain, Arizona. M.S. thesis, N. Arizona Univ., Flagstaff. Lists 82 taxa for 518 ha. Annotations include common names, elevational range, relative abundance, habitat, and distribution. Provides keys to species and short species de- scriptions. Discusses plant adaptations to alpine environments, krummholz, and slope effect. Compares local arctic-alpine flora with that of the Rocky Mountains; describes plant communities. Paulik, L. A. 1979. A vascular flora of the subalpine spruce-fir forest of the San Francisco Peaks, Arizona. M.S. thesis, Northern Ar- izona Univ., Flagstaff. Lists 189 taxa for 3600 ha; discusses his- tory of plant collection and ecological study on San Francisco Peaks; describes plant communities and lists characteristic species of each. Altitudinal extensions or records and plants of disturbed areas are listed separately. Distributions of two endemic species are shown on small-scale maps. 10. Bill Williams Mountain Hazen, J. M. 1978. The flora of Bill Williams Mountain. M.S. thesis, N. Arizona Univ., Flagstaff. Lists 221 taxa for 5180 ha. Discusses 200 MADRONO [Vol. 28 life zones; lists dominant plants in each zone; notes introduced plants and altitudinal records. Describes geology and climate. 11. Flagstaff Read, A. D. 1915. Flora of the Williams Division of Tusayan Na- tional Forest, Arizona. Plant World 18:112—123. Lists 284 taxa; covers the region west of Flagstaff and south of the Grand Canyon to the edge of the Colorado Plateau. Incomplete list with historical interest; notes introduced, new, or recently-described species; de- scribes vegetation of elevational zones. McDougall, W. B. 1959. Typical seed plants of the ponderosa pine zone. Mus. N. Arizona Bull. 32, Flagstaff. Lists 293 taxa for 172 ha. Includes keys to species, glossary of botanical terms, common names, and short species descriptions. Covers the area around the Museum of Northern Arizona near Flagstaff. 12. Walnut Canyon National Monument Arnberger, L. P. 1947. Flowering plants and ferns of Walnut Can- yon. Plateau 20:29-36. Lists 151 taxa for 761 ha. Incomplete list. Discusses plant communities in relation to soil moisture on north- and south-facing slopes. Spangle, P. F. 1953. A revised checklist of the flora of Walnut Can- yon National Monument. Plateau 26:86—-88. Adds 82 taxa col- lected between 1947 and 1953 to the flora of Walnut Canyon. Joyce, J. F. 1976. Vegetational analysis of Walnut Canyon, Arizona. J. Arizona Acad. Sci. 11:127—135. Ecological study; lists 93 ad- ditions to flora of Walnut Canyon. Discusses microclimates, plant communities. 13. Oak Creek Canyon Deaver, C. F. 1930. Floristic studies in Oak Creek Canyon. M.S. thesis, Univ. Arizona, Tucson. Lists 194 taxa from upper Oak Creek Canyon. Incomplete list, based on collections from three representative areas. Annotations include relative abundance and habit. Includes photographs of some plants and vegetation types. Sutton, M. 1952. A botanical reconnaissance in Oak Creek Canyon. Plateau 25:30—42. Lists 446 taxa for a 35-km stretch of Oak Creek Canyon. Road log with mileages and elevations notes interesting plants and plant associations; discusses vegetation changes with decreasing elevation. List annotated with common names. Aitchison, S. W. 1978. Oak Creek Canyon and the Red Rock Country of Arizona: a Natural History and Trail Guide. Stillwater Canyon Press, Flagstaff. Appendix lists 590 taxa for 200 ha; annotated with common names. Mammals, birds, amphibians, and reptiles are listed separately. Vegetation, fauna and geology are discussed; a road log notes features of interest. 1981] BOWERS: ARIZONA FLORAS 201 13a. Volunteer and Sycamore Canyons Schilling, M. A. 1980. A vegetational survey of the Volunteer and Sycamore Canyon region. M.S. thesis, Northern Arizona Univ., Flagstaff. Lists 376 taxa by plant community; discusses plant communities, climate, history, physiography, and geology. Plants of disturbed areas, introduced plants, and elevational extensions are listed separately. The flora is compared with the floras of Oak Creek Canyon, Walnut Canyon, Bill Williams Mountain, and the San Francisco Peaks. 13b. Chevelon Canyon Aitchison, S. W. and M. E. Theroux. 1974. A biotic inventory of Chevelon Canyon, Coconino and Navajo Counties, Arizona. Un- published report, submitted to U.S. Soil Conservation Service and USDA, Sitgreaves National Forest. For information on avail- ability contact S. W. Aitchison, Museum of Northern Arizona, Flagstaff. Lists 195 taxa; annotated with common names. Dis- cusses plant communities, slope aspect, rare or endangered species. Lists vertebrates, provides small-scale vegetation map. CENTRAL ARIZONA 14. Hualapai-Aquarius Planning Units Butterwick, M., D. Hillyard, and B. Parfitt. 1979. Annotated list of taxa of vascular plants collected in the Hualapai-Aquarius envi- ronmental impact statement area. Unpubl. report, USDI, Bur. Land Management, Phoenix District Office, 2929 W. Clarendon Ave., Phoenix. Lists 865 taxa for 698,220 ha. Annotations include common name, habitat, habit, nativity, relative abundance, vegetation type, and collectors. 15, 16, 17. Harcuvar, Vulture, and Skull Valley Planning Units Butterwick, M., P. Fischer, D. Hillyard, and D. Ducote. 1980. An- notated list of taxa of vascular plants collected in the Harcuvar, Vulture and Skull Valley Planning Units. Unpubl. report, USDI, BLM, Phoenix District Office, 2929 W. Clarendon Ave., Phoenix. Lists 559 taxa for the Harcuvar Planning Unit, 482 taxa for the Vulture Planning Unit, and 558 for the Skull Valley Planning Unit. Annotations include common names, habit, relative abun- dance, habitat, local distribution, collector and collection num- ber. 18. Montezuma Castle National Monument Spangle, P. and M. Sutton. 1949. The botany of Montezuma Well. Plateau 22:11-19. Lists 189 taxa. Incomplete list with no anno- 202 MADRONO [Vol. 28 tations. Discusses geomorphology and origin of the well (a sink- hole). Describes climate and plant communities. McDougall, W. B. and H. S. Haskell. 1960. Seed plants of Monte- zuma Castle National Monument. Mus. N. Arizona Bull. 35, Flagstaff. Lists 308 taxa for 423 ha. Includes keys to species, glossary of botanical terms, and short species descriptions. 19. Lake Pleasant Regional Park Lehto, E. 1970. A floristic study of Lake Pleasant Regional Park, Maricopa County, Arizona. M.S. thesis, Arizona State Univ., Tempe. Lists 364 taxa for 5827 ha. Discusses phenology of plants in disturbed habitats. Annotations include common names, hab- itat, collectors and collection numbers. Discusses history, geology, and climate. Lists plants for each habitat, includes photographs of vegetation types. 20. White Tank Mountains Regional Park Keil, D. J. 1973. Vegetation and flora of the White Tank Mountains Regional Park, Maricopa County, Arizona. J. Arizona Acad. Sci. 8: 35-48. Lists 332 taxa for 11,564 ha; annotations include relative abundance, habitat, and growth form. Discusses vegetation zones, includes vegetation map at 1:63,000. 21. McDowell Mountain Regional Park Lane, M. A. 1976. Vegetation and flora of McDowell Mountain Re- gional Park, Maricopa County, Arizona. M.S. thesis, Arizona State Univ., Tempe. Lists 286 taxa for 8475 ha; annotations in- clude plant community, habitat, collector and collection number, and relative abundance. Shows distribution of species among six Maricopa County regional parks. Discusses history, climate, and soils. Includes photographs and vegetation map. 22. Phoenix Hamilton, F. L. 1933. The desert garden: native plants of Phoenix and vicinity. Frances L. Hamilton, Phoenix. Deposited at Univ. Arizona Science Library, Tucson 85721. Lists 138 taxa for desert ranges near Phoenix, including Phoenix Mountain Park, Camel- back Mountain, Papago Park, and Squaw Peak. Incomplete. Pro- vides keys to families and species, line drawings of plants. An- notations include common names, habitat, short species descriptions, relative abundance, and phenology. 23. Sierra Estrella Regional Park Sundell, E. G. 1974. Vegetation and flora of the Sierra Estrella Re- gional Park, Maricopa County, Arizona. M.S. thesis, Arizona State Univ., Tempe. Lists 330 taxa for 7533 ha; annotations in- 1981] BOWERS: ARIZONA FLORAS 203 clude common names, collector and collection number, relative abundance, vegetation zone, plant community, elevation, and phenology. Describes history, geology, and climate; reviews past botanical work. Discusses plant communities and floristic differ- ences between Sierra Estrella and South Mountains. Includes photographs and a topographic map. 24. Three Bar Wildlife Area (Mazatzal Mountains) Dickerman, R. W. 1954. An ecological survey of the Three-Bar Game Management Unit located near Roosevelt, Arizona. M.S. thesis, Univ. Arizona, Tucson. Lists 286 taxa for 15,753 ha. Ecological study; incomplete list. Annotations include habitat, collection lo- cations and numbers, relative abundance, phenology, and vege- tation type. Discusses forage value for deer; describes vegetation types and effect of fire on chaparral. Includes photographs of vegetation types and map of burned areas. McCulloch, C. Y. and C. P. Pase. 1968. Checklist of plants of the Three Bar Wildlife Area. Jn Wildlife Research in Arizona 1967, p. 77-88. Arizona Game and Fish Dept., Phoenix. Lists 521 taxa for 15,753 ha; annotated with common names; notes artificially- seeded species. About 20 taxa are identified only to genus. 25. Sierra Ancha Johnson, R. R. 1960. The biota of Sierra Ancha, Gila County, Ari- zona. M.S. thesis, Univ. Arizona, Tucson. Lists 449 taxa for 5192 ha. Lists plants, birds, and mammals; plant list annotated with collection localities. Discusses life zones, analyzes distribution of plants reaching northern or southern limit in Sierra Ancha. In- cludes photographs of vegetation types. Pase, C. P. and R. R. Johnson. 1968. Flora and vegetation of the Sierra Ancha Experimental Forest, Arizona. USDA For. Serv. Res. Paper RM-41. Lists 735 taxa for 5192 ha; relatively complete list based on 35 years of botanical exploration. Discusses floristic affinities; describes vegetation types. Includes photographs and vegetation map. 26. Tonto National Monument Burgess, R. L. 1965. A checklist of the vascular flora of Tonto Na- tional Monument. J. Arizona Acad. Sci. 3:213—223. Lists 270 taxa for 454 ha; annotations include common name, life form, relative abundance, and habitat. Compares flora with that of Montezuma Castle National Monument. 27. White Mountains Judd, B. I. 1972. Vegetation zones around a small pond in the White Mountains of Arizona. Great Basin Naturalist 32:91—96. Lists 32 204 MADRONO [Vol. 28 taxa. Describes vegetation zones around Carnero Lake; lists com- mon name, percent protein, relative abundance, and vegetation zone for each species; discusses succession from pond to forest. Smith, E. L. 1974. Phelps Cabin Research Natural Area. Jn Estab- lished Natural Areas in Arizona: a guidebook for scientists and educators, p. 127-133. Office of Economic Planning and Devel- opment, Phoenix. Lists 195 taxa for 126 ha; summarizes 25 years of botanizing; annotated with common names. Includes vegeta- tion map at 1:10,500, acreage of each vegetation type. McLaughlin, S. P. 1978. Productivity of the understory community in an Arizona ponderosa pine forest. Ph.D. dissertation, Univ. Arizona, Tucson. Lists 180 taxa for 50 ha; complete list, no an- notations; an interesting area with a species-rich understory. SOUTHWESTERN ARIZONA 28. Cabeza Prieta Game Range Simmons, N. M. 1966. Flora of the Cabeza Prieta Game Range. J. Arizona Acad. Sci. 4:93—104. Lists 238 taxa for 380,700 ha; an- notations include common names, elevation, collection localities, ethnobotanical information, and use by wildlife. Illustrates plant replacement with matched photographs. Lehto, E. 1979. Plants of Cabeza Prieta Game Range. Unpubl. list. For information on availability contact the Arizona Natural Her- itage Program, 30 N. Tucson Blvd., Tucson 85719. Lists 304 taxa for 380,700 ha; no annotations. 29. Organ Pipe Cactus National Monument Phillips, J. W. 1967. A checklist of the plants of Organ Pipe Cactus National Monument. Unpubl. ms. deposited at Organ Pipe Cac- tus National Monument. Lists 456 taxa for 133,898 ha; annota- tions include local distribution, common names, and comments on identification. Adams, W. B. 1971. A checklist of the plants of Organ Pipe Cactus National Monument. Unpubl. ms. deposited at Organ Pipe Cac- tus National Monument. Lists 520 taxa for 133,898 ha; annota- tions include collection locality, common names. Adams, W. B. 1971. A flora of Quitobaquito. Unpubl. ms. deposited at Organ Pipe Cactus National Monument. Lists 93 taxa for Qui- tobaquito pond and vicinity. Phillips, J. W. 1971. Preliminary flora of Dripping Springs. Unpubl. ms. deposited at Organ Pipe Cactus National Monument. Lists 91 taxa for Dripping Springs in the Puerto Blanco Mountains. Jordan, E. H. 1975. A checklist of the plants of Organ Pipe Cactus National Monument. Unpubl. list deposited at Organ Pipe Cactus National Monument. Lists 521 taxa for 133,898 ha; compiled 1981] BOWERS: ARIZONA FLORAS 205 from herbarium specimens at the Monument and University of Arizona; annotations include common names and location of voucher specimens. Bowers, J. E. 1980. Flora of Organ Pipe Cactus National Monument. J. Arizona-Nevada Acad. Sci. 15:1-11; 33-47. Lists 522 taxa for 133,898 ha; annotations include elevational range, collection lo- cations, local distribution, relative abundance, common names, phenology, and habit. Small-scale map shows collection localities. Discusses species richness, distributional limits, paleobotany, and mesic habitats. SOUTHEASTERN ARIZONA 30. San Pedro River Gavin, T. A. 1973. An ecological study of a mesquite bosque. M.S. thesis, Univ. Arizona, Tucson. Lists 43 taxa for 137 ha; anno- tations include common names and relative abundance. Study area was along the San Pedro River near Mammoth. Birds, mam- mals, amphibians, and reptiles also listed. Discusses historic vege- tation change, examines conservation of riparian habitat. Dis- cusses seasonal aspects of flora. Includes vegetation map. 31. Aravaipa Creek Kepner, W. G. 1978. Vegetation and flora of the Aravaipa Creek Primitive Area, Graham and Pinal Counties, Arizona. Unpubl. report, USDI, BLM, Safford District Office, 425 E. 4th St., Saf- ford, Arizona. Lists 150 taxa for 3238 ha; annotated with common names. Describes vegetation types, riparian habitats; discusses hydrology, geology, and climate. Warren, P. L. and L. S. Anderson. 1980. Annotated checklist of plants of the George Whittell Wildlife Preserve. Jn T. B. John- son, ed., 1980 Progress Report for the Biological Survey of the George Whittell Wildlife Preserve, Graham and Pinal Counties, Arizona, p. 80-124. For information on availability contact the Arizona Natural Heritage Program, 30 N. Tucson Blvd., Tucson 85719. Lists 320 taxa; complete list to follow in final report. An- notations include common name, habit, elevational range, habi- tat, local distribution, relative abundance, and brief species de- scriptions. 32. Clifton Davidson, A. 1904. Flora of the Clifton district, Arizona. Bull. S. Calit@XNcad.ocl. Ss 110-111-441 8-19). 35=36,. 130-131,.5-6/_70; 6:34-36. Lists 348 taxa; annotations include collection localities, relative abundance, habitat, and phenology. Compiled from col- 206 MADRONO [Vol. 28 lections within a 40-km radius of Clifton. Incomplete list with historical interest. 33. Gila River McGill, L. A. 1979. Vascular flora. Jn Resource inventory for the Gila River Complex, eastern Arizona, p. 56-83. Unpubl. report, USDI, BLM, Safford District Office, 425 E. 4th St., Safford, Arizona. Lists 394 taxa; annotations include common names and taxonomic citations; notes established exotics. Discusses species diversity, riparian habitats. 34. Santa Catalina Mountains Whittaker, R. H. and W. A. Niering. 1964. Vegetation of the Santa Catalina Mountains, Arizona. I. Ecological classification and dis- tribution of species. J. Arizona Acad. Sci. 3:9-34; . 1968. Vegetation of the Santa Catalina Mountains. III. Species distri- bution and floristic relations on the north slope. J. Arizona Acad. Sci. 5:3-21. Lists 1210 taxa; annotations include growth form, Raunkaier life form, floristic affinity, distribution, elevational range, and relation to topographic moisture gradients. Ecological study that analyzes vegetation along moisture and elevational gra- dients. 35. Rillito River Willis, E. L. 1939. Plant associations of the Rillito floodplain in Pima County, Arizona. M.S. thesis, Univ. Arizona, Tucson. Lists 201 taxa, notes habit of each. Ecological study; discusses riparian plant communities in relation to water table; describes vegetation types. 36. Thomas Canyon (Baboquivari Mountains) Toolin, L. J. 1979. A floral survey of Thomas Canyon, Baboquivari Mountains, Pima County, Arizona. Unpubl. For information on availability contact the Arizona Natural Heritage Program, 30 N. Tucson Blvd., Tucson 85719. Lists 183 taxa. Briefly discusses major plant communities and lists noteworthy collections. 37. Tucson Mountains Thornber, J. J. 1909. Vegetation groups of the Desert Laboratory domain. In V. M. Spalding. Distribution and movements of des- ert plants, p. 103-112. Publ. Carnegie Inst. Wash. 113. Lists 442 taxa for 1036 ha. Classifies plants by habitat and growth form; lists introduced species. Discusses growth form in relation to cli- mate and habitat. Wadleigh, R. 1969. Plant list for Tucson Mountain District of Sa- guaro National Monument. Unpubl. Available from Unit Man- 1981] BOWERS: ARIZONA FLORAS 207 ager, Saguaro National Monument, Box 595, Tucson, Arizona 85704. Lists 430 taxa for 8499 ha. Notes whether plant was ob- served or collected, location of voucher specimens. Turner, R. M. 1977. Plant species list! Tumamoc Hill. Unpubl. Available from USGS Research Project Office, 301 W. Congress, Tucson, Arizona 85701. Lists 438 taxa for 1036 ha. Updates Thornber (1909). Annotations include synonymy, habitat. 38. Rosemont (Santa Rita Mountains) McLaughlin, S. P. and W. Van Asdall. 1977. Flora and vegetation of the Rosemont area. Jn An environmental inventory of the Rosemont area in southern Arizona. 1:64—98. Deposited at the Univ. Arizona Science Library, Tucson 85721. Lists 416 taxa for 6475 ha. Discusses vegetation change, floristic affinities. Analyzes plant communities, includes interesting discussion of plant com- munities on limestone. Provides photographs of vegetation types, vegetation map at 1:63,000. 39. Santa Cruz County Kaiser, J. M. 1980. Vegetation of Santa Cruz County. Unpublished manuscript. USDA Plant Protection and Quarantine Program, Nogales, Arizona. Lists 957 taxa for 322,589 ha; based on 32 years of plant collection; excludes grasses, sedges, and ferns. An- notations include species descriptions, collection localities, distri- bution in Arizona and Santa Cruz County. Profusely illustrated with color photographs. 40. Tumacacori Mission National Monument Mouat, D. A., S. J. Walker, and B. D. Treadwell. 1977. The Tu- macacori Mission National Monument floral inventory and vege- tation map project. Office of Arid Lands Studies, Univ. Arizona, Tucson. Lists 130 taxa for 4 ha: annotated with common names. Includes map of perennial species at 1:250. 41. Sycamore Canyon (Pajarito Mountains) Toolin, L., T. R. Van Devender, and J. M. Kaiser. 1980. The flora of Sycamore Canyon, Pajarito Mountains, Santa Cruz County, Arizona. J. Arizona-Nevada Acad. Sci. 14:66—74. Lists 625 taxa for 932 ha; complete list based on 35 years of collection. Discusses species with disjunct distributions, species known in U.S. only from Sycamore Canyon. Describes geographic affinities of flora. Annotations include location of voucher specimens. 42. O’Donnell Canyon (Canelo Hills) Yatskievych, G. A. 1980. Plant list for O’Donnell Canyon, Canelo Hills, Santa Cruz County, Arizona. Unpubl. For information on 208 MADRONO [Vol. 28 availability contact the Arizona Natural Heritage Program, 30 N. Tucson Blvd., Tucson 85719. Lists 251 taxa. Annotations include habitat, collector and collection dates. Notes rare or endemic species. 43. Patagonia-Sonoita Creek Sanctuary Fay, J. M. 1978. Vegetation and flora of the Patagonia-Sonoita Creek Sanctuary, Patagonia, Arizona. Unpubl. For information on availability contact the Arizona Natural Heritage Program, 30 N. Tucson Blvd., Tucson 85719. Lists 287 taxa for 125 ha; anno- tations include brief species descriptions, phenology, relative abundance, habitat, local distribution, common names, ethno- botanical information, and use of plants by wildlife. Provides keys to the species. Discusses conservation of cottonwood gallery forests. Describes local history, plant associations, floristic affin- ities, patterns of disturbance. Includes vegetation map. 44, 47. Whetstone and Mule Mountains Wentworth, T. R. 1976. The vegetation of limestone and granite soils in the mountains of southeastern Arizona. Ph.D. disserta- tion, Cornell Univ., Ithaca. Lists 515 taxa for the study area, which included parts of the Mule, Whetstone, and Huachuca Mountains. Ecological study; annotations include substrate, growth form, floristic affinity, Raunkaier life form, and distri- bution. Discusses substrate relations and species diversity in southeastern Arizona. 45. Chiricahua Mountains Clark, O. M. 1939. Chiricahua summer flora. Southwestern Monu- ments Monthly Reports (Oct):318—329; . 1940. Check list of the flora of Chiricahua National Monument. Southwestern Mon- uments Monthly Reports (Sep):201—215. Lists 601 taxa for 4308 ha. Incomplete; annotated with common names. Robinson, M. D. 1968. Summer aspect of a high coniferous forest in the Chiricahua Mountains, Arizona. M.S. thesis, Univ. Arizona, Tucson. Lists 71 taxa for the Fly’s Peak area. Ecological study; discusses relationship between slope aspect, climate, elevation, and vegetation. Describes plant communities. Reeves, T. 1976. Vegetation and flora of Chiricahua National Mon- ument, Cochise County, Arizona. M.S. thesis, Arizona State Univ., Tempe. Lists 687 taxa for 4308 ha; annotations include common names, phenology, relative abundance, habitat, eleva- tional range, and vegetation association. Describes plant com- munities; provides topographic, geologic, and vegetation maps; illustrated with photographs of vegetation types. 1981] BOWERS: ARIZONA FLORAS 209 46. Huachuca Mountains Wallmo, C. O. 1950. Vegetation mapping of Fort Huachuca Wildlife Area. In Fort Huachuca Wildlife Area surveys 1950-1951, p. 1— 43. Arizona Game and Fish Commission, Phoenix. Lists 482 taxa for 18,063 ha. Collections and species lists made by L. N. Good- ding; grasses extensively annotated. Other annotations include browse and forage value, use by wildlife, and relative abundance. Discusses vegetation types; provides vegetation map at 1:75,000. Goodding’s list was updated (to 554 taxa) in 1963. Wentworth, T. R. 1976. See 44, 47. Whetstone and Mule Mountains. Toolin, L. J. 1980. Final report on the flora of Ramsey Canyon. Unpubl. For information on availability contact the Arizona Nat- ural Heritage Program, 30 N. Tucson Blvd., Tucson 85719. Lists 343 taxa. Discusses physiography and climate; notes plants of special interest. Annotations include habit, local distribution, rel- ative abundance, phenology, common names, and habitat. 48. Hooker Cienega Yatskievych, G. and C. E. Jenkins. 1981. Fall vegetation and zo- nation of Hooker Cienega, Graham County, Arizona. J. Arizona- Nevada Acad. Sci. In press. Lists 113 taxa; study area is 8 km long. Annotations include habitat, location of voucher specimens, common names, and state and county records. Discusses vege- tational zonation and hydrology. ACKNOWLEDGMENTS I thank Steven McLaughlin, Thomas Van Devender, Charles Mason, Barbara Phil- lips and Arthur Phillips for suggesting and making available several floras and for reviewing the manuscript. Ralph Gierisch and Carl-Eric Granfelt very kindly provided plant lists, as well. I also thank Paul Mirocha for drafting the map and Ann Rosenthal for typing the manuscript. LITERATURE CITED EWAN, J. 1936. Bibliography of the botany of Arizona. Amer. Mid]. Naturalist 17:430— 454. HARPER, K. T., D. C. FREEMAN, W. K. OSTLER, and L. G. KLIKOFF. 1978. The flora of Great Basin mountain ranges: diversity, sources, and dispersal ecology. Great Basin Naturalist Memoirs 2:81—103. MEYER, S. E. 1978. Some factors governing plant distributions in the Mojave-Inter- mountain transition zone. Great Basin Naturalist Memoirs 2:197—207. RAVEN, P. H. 1963. A flora of San Clemente Island, California. Aliso 5:289-317. RICHERSON, P. J. and K. LuM. 1980. Pattern of plant species diversity in California: relation to weather and topography. Amer. Naturalist 116:504—536. (Received 5 Jun 1980; revision received 29 Jan 1981; accepted 2 Feb 1981.) A NEW SPECIES OF QUERCUS (FAGACEAE) FROM SOUTHERN CALIFORNIA KEVIN C. NIXON Department of Botany, University of Texas, Austin 78712 KELLY P. STEELE Department of Biological Sciences, University of California, Santa Barbara 93106 ABSTRACT Quercus cornelius-mulleri Nixon & Steele is described from southern California, U.S.A., and Baja California Norte, Mexico. Fused-stellate trichomes, which were pre- viously known in Quercus only in the series Virentes, are reported for the new species. The scrub oaks of the subgenus Quercus (Lepidobalanus Endl. ex Oersted) in southern California have been a source of confusion to botanists since the earliest botanical expeditions (for a complete dis- cussion of the nomenclatural history of the California scrub white oaks, see Tucker, 1952a). Trelease (1924) included all of the southern California scrub white oaks in Quercus dumosa Nutt., recognizing some varieties (such as var. turbinella (Greene) Trel.). Tucker (1952a, 1952b) established Q. turbinella Greene subsp. californica Tucker as a taxon distinct from Q. dumosa, with a range from the western edge of the Mojave Desert northward through the inner coast ranges to San Benito County. Tucker (1952a, 1953) did not include material from the desert border mountains of San Diego and Riverside Counties in his concept of Q. turbinella subsp. californica; this material, he felt, was “best referred to Q. dumosa, rather than Q. turbinella.” Recent investigations indicate that these latter populations are representative of an undescribed species, distinct in fundamental morphological char- acters, distribution, and ecology from Q. dumosa, Q. turbinella subsp. turbinella, and Q. turbinella subsp. californica. It is a pleasure to name the species in honor of Professor Emeritus Cornelius H. Muller, both for his great contributions to oak taxonomy, and for his devotion to those fortunate enough to be his students. Quercus cornelius-mulleri Nixon & Steele sp. nov. Frutices sempervirentes multiramosi 1—2(—3) m alti; rami hornotini 1—1.5 mm diametro dense tomentulosi trichomatibus sessilibus stellatis breviradiatis; gemmae 2.5—3 mm longae ovoideae glabrae. Folia crassa coriacea ovata vel oblonga vel obovata vel subrotundata ad marginem dentata dentibus saepe minus quam 2 mm longis vel integra, supra obscure viridia, trichomatibus sparsis stellatis, subtus dense tomen- MADRONO, Vol. 28, No. 4, pp. 210-219, 8 December 1981 1981] NIXON & STEELE: QUERCUS CORNELIUS-MULLERI SP. NOV. _211 tulosa trichomatibus sessilibus superpositis minutis radiis (8—)12(—16) ad basem coalescentibus usque ad 0.1 mm longis adpressis; petioli 2— 5 mm longi tomentulosi. Amenta fructifera subsessilia; cupula fructi- fera hemisphaerica vel profunde cyathiformes; glans annua saepe fu- siformis vel late conica sed variabilis. Multi-stemmed rounded shrubs 1-2 or 3 m tall, the crowns densely branched; twigs of the season 1—1.5 mm in diameter, terete, densely short-stellate tomentulose (individual rays less than 0.2 mm long), persisting tomentulose the second year, occasionally only sparsely pu- berulent and brown; buds 2.5—3 mm long, ovoid, obtuse, glabrous and dull brown, the younger scales ciliate-hairy; stipules ca. 4 mm long, subulate, strigose, caducous; leaves evergreen, rather thick and coriaceous, ovate to oblong or narrowly obovate, or subrotund, api- cally acute or rounded, basally rounded to cuneate, the margins spar- ingly toothed throughout or entire, flat or sometimes undulate, car- tilaginously thickened, the teeth sclerenchymatously tipped, the spines usually shorter than 2 mm; upper surface dull green, sparsely pubes- cent with minute stellate hairs, 0.15—0.17 mm in diameter, about 1 mm distant or much more sparse or occasionally deciduous; lower surface appearing glaucous because of a dense tomentulum of over- lapping minute stellate hairs, to 0.2 mm in diameter, closely ap- pressed, consisting usually of 12 (from 8 to 16) rays fused at their bases and forming a flattened rotate cluster; glandular trichomes absent on mature leaves; the midrib yellow against the white or ivory tomen- tulum; veins 6 or 7 on each side, slightly raised above, more prominent (even under tomentulum) beneath, irregularly branched and anasto- mosing; petioles 2-5 mm long, tomentulose similarly to the twigs or less so on the abaxial side; staminate catkins 25-55 mm long on slender sparsely pubescent rachises, the flowers distant except for the con- gested distal quarter, calyx lobes ciliate, the filaments markedly exsert- ed, anthers glabrous; pistillate catkins subsessile with clusters of 2—3 flowers or sometimes solitary; fruit annual, solitary, paired, or in clus- ters of 3, subsessile; cups hemispheric to cup-shaped, or sometimes deeply cupped, basally rounded, as much as 20 mm broad and 13 mm high, scales somewhat thickened basally, gray-tomentulose throughout or the short thin apices glabrous and brown; acorns variable, usually fusiform or broadly conical to ellipsoid, 20-30 mm long and 10-13 mm broad or smaller, glabrous except at the puberulent apices, the basal one-third enclosed in the cups. TyPE: USA, California, San Diego Co., ca. 24.7 km on McCain Valley Road n. from its junction with Interstate 8 (32°45'N, 116°20’W), 1335 m elevation, July 24, 1980, Nixon and Steele 2765 (Holotype: UCSB; isotypes: DAV, NY, RSA, SD, TEX, UC). PARATYPES: Mexico, Baja California Norte: Cantu Grade, 9.7 km e. of La Rumorosa, Moran 13147 (RSA, SD); 6 kme. of La Rumorosa, 212 MADRONO [Vol. 28 then n. 0.8 km, Steele 374C, 374E (UCSB); 6.4 kms. of La Rumorosa, Steele and Nixon 383 (UCSB); 32 kms. of La Rumorosa, Nixon and Hendon 2335A (UCSB); USA, CA, San Diego Co.: McCain’s Ranch, Manzanita Indian Reservation, Gander 8880 (SD); Montezuma Val- ley, Muller 4040 (CHM); Dubbers, Harbison and Higgins 44.129 (SD); 1.5 kme. of Burnt Rancheria Campground, Laguna Mountains, Cox s.n. (SD); 11.2 kme. on County Rd. $22 from its junction with County Rd. 82, Culp Valley, Steele and Hendon 391 (UCSB); Riverside Co.: Hidden Valley Campground, Joshua Tree Natl. Monument, Wilken 7552 (UCSB); 21 km s. of Palm Desert on CA Highway 74, Nixon and Hendon 2547 (UCSB); San Bernardino Co.: Dry Morongo Creek, Dunkle 3378 (LAM); 19 km s. on CA Highway 18 from its junction with CA Highway 247, Nixon and Hendon 2553 (UCSB, TEX). Quercus cornelius-mulleri occurs on the northeastern side of the San Bernardino Mountains, eastward to the granitic mountains of Joshua Tree National Monument, southward along the desert margin of the San Jacinto Mountains to the Laguna Mountains in San Diego County, and extends into Baja California Norte, Mexico, along the eastern escarpment of the Sierra Juarez. The southernmost known population is aproximately 40 km south of the international border, but it is likely that the species is found further to the south, possibly along the east side of the Sierra San Pedro Martir. The species usually occurs on granitic soils in association with Pinus monophylla Torr. & Frem., between elevations of 1000 m and 1800 m. Additional associates include Juniperus californica Carr., Ade- nostoma sparsifolium Torr., Rhus ovata Wats., and at the lower el- evational limits, such desert species as Larrea tridentata (Sesse & Moc. ex DC.) Cov. Morphological and distributional differences among the four taxa of scrub oak (subgenus Lepidobalanus) that are found south of the Tranverse Ranges are outlined in Table 1. The Baja California populations lie in the single-needle pinon belt (Pinus monophylla), at an elevation below those of Q. turbinella subsp. turbinella. Quercus turbinella is associated commonly with four-needle pinon (P. quadrifolia Parl. ex Sudw.) in this area. The two oak species are readily distinguished by differences in leaf spi- nation, leaf color, peduncle length, and trichome characters (see Table 1). In their zone of contact, hybrids occur sporadically, but there is little indication of introgression away from the contact area into the main populations of either species. Quercus cornelius-mulleri is not in contact with either Q. turbinella subsp. turbinella or Q. turbinella subsp. californica in San Diego and Riverside Counties. However, Q. engelmannii Greene, and to a lesser extent, OQ. dumosa, enter its range in this area. Hybridization between 1981] NIXON & STEELE: QUERCUS CORNELIUS-MULLERI SP. NOV. 213 Q. cornelius-mulleri and Q. engelmannii has been confused often with hybridization between the latter and either Q. turbinella subsp. cal- ifornica (not in the area) or Q. dumosa. The type of Q. acutidens Torr., collected by Parry (NY!) is such an intermediate and was prob- ably collected along the trail inland from San Luis Rey. In this vicinity (near Warner Springs), “pure” populations of both species come to within 8 km of each other, with intervening populations showing in- dications of hybridization. Some hybridization between Q. dumosa and Q. cornelius-mulleri occurs where coastal elements mix with des- ert elements in the areas of relatively low mountain passes. This sit- uation occurs sporadically, such as in the vicinity of Garner Valley and Santa Rosa Summit, Riverside County, and Montezuma Valley, San Diego County. From Cajon Pass (San Bernardino County) westward, Q. cornelius- mulleri is replaced by Q. turbinella subsp. californica in the interior and Q. dumosa in the coastal areas. There does not appear to be contact between Q. cornelius-mulleri and the former, although this might be expected east of Cajon Pass on the north side of the San Bernardino Mountains. Quercus cornelius-mulleri is also the common scrub oak in the granitic ranges from Morongo Valley eastward into Joshua Tree National Monument. The desert scrub oak of the New York Mountains of eastern San Bernardino County is Quercus tur- binella subsp. turbinella, which also occurs commonly to the east in the desert ranges of Arizona, New Mexico, west Texas, and northern Mexico (see Tucker, 1952a). Characters of leaf trichomes have been used to distinguish species of Quercus in previous studies (Dyal, 1936; Tucker, 1952b; Tucker and Muller, 1957). More recently, Hardin (1976, 1979) has shown the value of scanning electron microscopy (SEM) in the study of Quercus leaf trichomes. Scanning electron microscopy indicates that Q. cor- nelius-mulleri is unique among the California oaks in its possession of fused-stellate trichomes on the lower leaf surface (see Figs. 1—4). Har- din (1976) defined the fused-stellate trichome as “a ‘non-glandular’ trichome with fusion of the rays beyond the base to a maximum of two-thirds the length of the rays.” Although the foliar trichomes of Q. cornelius-mulleri are typically highly fused, the degree of fusion varies both on individual plants and within populations. No geographic pat- terns of trichome variability are apparent within the species. Hardin (1979) reported fused-stellate trichomes only for the series Virentes Trel., a small, closely related group of white oaks from the South- eastern United States, Mexico, and Central America. Since Q. cor- nelius-mulleri shares no other important diagnostic features with the Virentes (which possess connate cotyledons, pubescent anthers, and thick pubescent acorn cup-bases) there is no reason to assume any close relationship between Q. cornelius-mulleri and the latter group. [Vol. 28 MADRONO 214 wu ¢7'0-8T'0 sirey Iepnpurys ayesastun aWIOS pure SITeY 9}e]]97S SUIJA 7 94} SULINISGO JOU 9}e1apoUr 0} asreds yuadsaqnd -9yeT[a}s ATUIOJIUN = pay}00}-JuadseuIds ApIe[NsaI111 u9013-API3 saoejins yjOq ‘peiojooiun = qud1Insap APYSYs uso ‘apqeirea potusofyps ‘dsqns DIJ9uUIQAN] “O wut 9¢°0-1T7'0 sitey Je[npurys oyelestun AUBUW PUB SITeY 9}e]]9}S SUIVA ,Z 94} SULINISGO you jnq sdejANs dU} J9AO pasa}ye9s ATWAOjJIUN ‘3}e1IpOul quassaqnd -9yeJays ATUOsIuN => p2yj00} -jusosaulds Ajre[nsa1 UdaI13-ARIZ 0} UdIIB-MO][IA SIdeJINS yjoq ‘paiojootun > papunol 0} a3ep109 pDyJaurqany “dsqns DI]AU1QAN] “O wu 97'0O-8T 0 suey Je[npue[s ayelestun IUIOS pUe SITeY 94e]I[9}S SUIJA 7 94} SULINISGO jou pue qlIptul ay} eau payerjuaauos ‘asieds yuaosaqnd-9}e]]9}s Ajasreds 0} snoiqeys asoulds Ajared ‘a1tuUe 0} 938} Uap-a}eUOIONU udaI3 [[Np sdejJAns IIMO] ‘UIII3 VdeJANS Jaddn ‘paiojostun = papunol 0} anbijqo Dsounp “CO wu 97°0-9T'0 ATuo sirey 9}e][91S-pasny pue 9ze][99s SUIJA ,Z 94} BULINISGO pue adejins 1aMO] a1ITUA BULIJAOD = ‘asuep AIBA quaosaqnd-3}e][9}s AJayerapour 0} Ajasreds yuadsaulds 0} 311Ua IO pay}00} A[sutreds AIOAI 0} JUIYM JOVJINS 19MOT ‘uda13 MOT[IA 10 ABIZ soejins Jaddn ‘pa10joo1q papunois 0} anbijqo 1Ad]]NU-SNIUAOD “COC “SHVO ALIHM PNAS VINAOATTIVI NAYHLNOS AHL AO SHALOVAVH, TVNOILNGIMLSIG AGNV TVOIDOTOHdYO[T sirey a}e]JAIS Jo JajaWIeIG adA} aduadsaqng JIVJANS IMO] —a9ouadseqnd jea’T goejins 1addn —a2ouadseqnd jeaT SUISIVUL Jea'T JO[OD yea] aseq Jeo] jayoereyy ‘| aTavV]L 215 1981] NIXON & STEELE: QUERCUS CORNELIUS-MULLERI SP. NOV. pue[poom [[fq00} ‘pue,[poom Jadiunl -uourd ‘jereddeyd 10119}UI Ww OOOc—W OOe BrusosyeD “0D ovUsg ues 0} YoU (WD ‘-O7) OUIPeUIag ues) sseg uofey Wojy sasuei yseod J3UUTI uly} Ayjensn ajIssasqns 10 aqIssas ou (Z1-)8(-9) (euozliy) pue[poom Jadiunt -uould (erur0styeD efeg) ‘uoutd Jaddn ‘jereddey) jaasap wt QOOO7—W OOFT IWION elusIosIe*D eleg ‘Sexay pue euozuy 0} 4S¥9 (“S}TAT YOR MAIN) CIUIOJI[ED Ulaysea uryy 9}e[nNdunped ou (pI—-)OT(-8) pue[poom yeo uslayynos ‘jereddeyo W QOO9T—UW OF bee Le JO sasuel 4SeOd qyyiou ay} pue ‘oy euUleYyay, 0} BVIUIOsITeED efeg usayyou wos sadoys (A]}sou) [e}seoo a7e[no1aqny a[Issasqns 10 a]Issas ou (OI-)8(-9) jereddeyp) j1asap 1o/pue jjaq uoutd JIMO] IY} Ul UISIeVW JAaSap WW QOO8T—W OOOT IJION, VIUIOFITCD efeg Jo zarenf e1iaIs 94} 0} “SI, PuNse’T 94} pue ‘si OJUTDeL ues 34} Ysno1Yy} YINOos BIUIOJI[eD UIIYNOS jo ‘S}J. OUIPeUIIg ues JO SUISIBUI J1aSap 9ye[NI1IqN} 0} UI} a]Issasqns 10 ayIssas SOA (9T—)ZT(-8) syeqqey UONeAITA uoRNquysiqd safeos dnd uiooy JUIUIYIe}e UIOIY (NAS tM I[QISIA) UOISNS AVY Jaquinu Avy DIIUAO{IvI ‘dsqns DIJaULQAN] “CO DjYaUu1gany “dsqns D1]9U2GQAN} ‘CO DsSOUNp ‘CO 1A9]]NU-SNIAUAOI °C) qapeieyy ‘CHNNILNOZ) (TT ATEAV EE 216 MADRONO [Vol. 28 “a Be ik Silage om “ah ee ae ee rie Aes Fics. 1-4. Scanning electron micrographs of abaxial leaf surfaces of southern Cal- ifornia scrub white oaks. FIG. 1. Quercus dumosa Nutt. (Ventura Co., Muller 5220 (Muller Private Herbarium)). Fic. 2. Quercus turbinella Greene subsp. californica Tucker (Santa Barbara Co., Tucker 1885-4 (Muller Private Herbarium)). F1G. 3. Quer- cus turbinella Greene subsp. turbinella (Baja California Norte, Steele and Nixon 386 (UCSB)). Fic. 4. Quercus cornelius-mulleri Nixon & Steele (Baja California Norte, Moran 13147 (SD)). Scale (FIG. 1) 10 um. All figures to the same scale. Further study is needed before the relationships of Q. cornelius-mul- leri can be determined. In addition to ray-fusion in the stellate trichomes of Q. cornelius- mulleri, the high ray number, usually around twelve per trichome but 1981] NIXON & STEELE: QUERCUS CORNELIUS-MULLERI SP. NOV. 217 Fic. 5. Typical fruiting specimen of Quercus cornelius-mulleri Nixon & Steele. (San Diego Co., 3.2 km w. of Jucumba on old Highway 80 (UCSB). The bar is equal to 1 cm. Fruit and twig to same scale. often up to sixteen, amply separates it from all other California white oaks. Quercus dumosa and Q. turbinella subsp. californica possess typically eight rays per foliar trichome. The mean ray number of foliar trichomes of Q. turbinella subsp. turbinella is aproximately ten. The density of trichome cover in Q. cornelius-mulleri is also unique among the California scrub oaks. The lower leaf epidermis is typically oc- cluded by the densely packed, appressed, minute trichomes. The stel- late trichomes of the other three scrub oaks are never so dense as to obscure totally the lower leaf surface. Glandular trichomes (simple 218 MADRONO [Vol. 28 Fic. 6. Distribution of Quercus cornelius-mulleri Nixon & Steele in California and Baja California Norte, Mexico. uniseriate trichomes, as defined by Hardin, 1976) are lacking on the mature leaves of Q. cornelius-mulleri, but are consistently found on the mature leaves of the other California white oaks. These fundamental differences between Q. cornelius-mulleri and the other scrub white oaks of southern California preclude its inclusion as a subspecific taxon within any of those species. Similarly, there is no evidence that the populations which constitute Q. cornelius-mulleri were derived by hybridization from any of the extant California species. It is not intermediate morphologically or ecologically between any known species. On the contrary, it is unique among California oaks morphologically; and based upon its distribution, it appears to be more xeromorphically adapted than the other California scrub oaks. ACKNOWLEDGMENTS The scanning electron micrographs were done on an ETEC Autoscan in the Electrical Engineering Department at the University of California at Santa Barbara; we thank Don E. Zak for his assistance and advice in obtaining these micrographs. We are indebted to Professor Emeritus Cornelius H. Muller for valuable discussions and the use of his herbarium. We also thank Wayne Ferren and Dr. Dale Smith for reading a draft of this manuscript. The Latin diagnosis was prepared by Marshall C. Johnston. 1981] NIXON & STEELE: QUERCUS CORNELIUS-MULLERI SP. NOV. = 219 Herbarium curators at RSA, SD, UCSB and LAM also gave valuable assistance, es- pecially Reid Moran (SD) and Wayne Ferren (UCSB). We also give particular thanks to Emily C. Hendon and John S. McManus for encouragement and support. LITERATURE CITED DyAL, S. C. 1936. A key to the species of oaks of eastern North America based on foliage and twig characters. Rhodora 38:53-63. HARDIN, J. W. 1976. Terminology and classification of Quercus trichomes. J. Elisha Mitchell Sci. Soc. 92:151-161. 1979. Patterns of variation in foliar trichomes of Eastern North American Quercus. Amer. J. Bot. 66:576—585. TRELEASE, W. 1924. The American Oaks. Mem. Nat. Acad. Sci. 20:1—255. TUCKER, J. M. 1952a. Taxonomic interrelationships in the Quercus dumosa complex. Madrono 11:234-251. . 1952b. Evolution of the California oak Quercus alvordiana. Evolution 6:162— 180. 1953. The relationship between Quercus dumosa and Quercus turbinella. Madrono 12:49-60. TUCKER, J. M. and C. H. MuLLER. 1957. A reevaluation of the derivation of Quercus margaretta from Quercus gambellii. Evolution 12:1—17. (Received 6 Nov 1980; accepted 5 Jan 1981; revised version received 6 Feb 1981.) A NEW SPECIES OF ACACIA (LEGUMINOSAE: MIMOSOIDEAE) FROM BAJA CALIFORNIA SUR, MEXICO ANNETTA M. CARTER Department of Botany, University of California, Berkeley, 94720 VELVA E. RUDD Smithsonian Institution, Washington, D.C. 20560 California State University, Northridge 91330 ABSTRACT Acacia kelloggiana Carter & Rudd is described from Baja California Sur, Mexico. To date it is known only from Cerro Giganta at the northern end of Sierra de la Giganta and from Sierra de las Palmas, the next range to the north, a distribution from 26°08’ to 27°N. The genus Acacia is well represented in Baja California, Mexico. Eleven species are included in Wiggins’ (1980) Flora of Baja Califor- nia, four of which are considered endemic to the peninsula. We here propose another species currently known only from Cerro Giganta in the northern Sierra de la Giganta and from Sierra de las Palmas, the next range to the north. This new species, known locally as “garabatilla de espina negra’, appears most closely related to Acacia peninsularis (Britton & Rose) Standley from southern Baja California, and A. occidentalis Rose from Sonora and Sinaloa. On the basis of a sterile collection from the vicinity of Bahia Escondido (Wiggins 17529, DS), A. occidentalis was considered by Wiggins (1980) to occur in the peninsula. On exami- nation, we believe this to be a specimen of Mimosa purpurascens B. L. Robinson, “garabatilla”, a common shrub in the area. As indicated in the following key, the most conspicuous differences are in characters of leaves, spines and fruit. Key to Acacia kelloggiana and Related Species Leaves with 1—2(—3) pairs of pinnae; leaflets 3-8 pairs, spatulate-ob- ovate, often emarginate, 5-10 mm long, 2.5—6 mm wide, glabrous or puberulent (Baja California Sur, 27°N southward) ......... er cs ee A. peninsularis Leaves with 1-18 pairs of pinnae; leaflets 5—23 pairs, linear, 3-8 mm long, 0.6-1.8 mm wide, glabrous or puberulent, sometimes mi- nutely glandular-ciliate along margins. Branches pubescent (sparsely pilose); spines unguiculate, internodal, MADRONO, Vol. 28, No. 4, pp. 220-225, 8 December 1981 1981] CARTER & RUDD: NEW SPECIES OF ACACIA 2a irregularly dispersed; leaves 1.5—4 cm long with 2-4 pairs of pinnae; leaflets 5-12 pairs, puberulent to glabrescent, about 3— 5 mm long, 0.6—1.0 mm wide; flowers in heads about 1.5 cm in diameter, the calyx about 1.5 mm long, the petals 2.5 mm long; legume 2—2.5 cm wide, tortuous, chartaceous, contracted between seeds (Sonora, Sinaloa) ............ A. occidentalis Branches glabrous; spines, when present, strongly unguiculate, stip- ular, paired (or occasionally subopposite); leaves 5-15 cm long with 1—13(—18) pairs of pinnae; leaflets 9-23 pairs, essentially glabrous, minutely ciliate and glandular along margin, about 5-8 mm long, 0.6—1.8 mm wide; flowers in heads about 2 cm in diameter, the calyx 2.5—3 mm long, the petals 3—4 mm long; legume 1.5—2 cm wide, falcate, subcoriaceous, slightly con- tracted between the seeds (Baja California Sur, ca. 26°08’ to DTI Sinks, istered ecaa uc) Aa cette ak ota uke A. kelloggiana Acacia kelloggiana Carter & Rudd, sp. nov. Frutices vel arbores ca. 2—7 m alti, maximam partem spinis stipu- laribus recurvatis armati; folia 5-15 cm longa, pinnarum paribus 1— 13(—18); foliolarum paribus 9—23, foliolis linearibus, glabris, 5-8 mm longis, 0.6—0.8 mm latis, nervo medio excentrico, submarginali; inflo- rescentiae ex pedunculis axillaribus 1(—3) constantes, floribus albidis in capitulis globosis diam. 2 cm dispositis; legumina subfalcata, 8—15 cm longa, 1.5—2 cm lata, subcoriacea, glabra, seminibus 2-8 (Fig. 1). Acaciae peninsulari atque A. occidentali affinis sed foliolis mino- ribus pluribusque, leguminibus subcoriaceis haud torulosis discedit. Shrubs or small trees 2—7 m tall, crown spreading; bark glabrous, smooth except for a few low, narrow ridges, the lenticels conspicuous; spines stipular, paired or occasionally subopposite (sometimes lack- ing), broad-based laterally compressed, strongly recurved, to ca. 9 mm long, dark brown to black; stipules 4-8 mm long, narrowly linear, 1 mm wide or less, early caducous; leaves bipinnate, rarely fascicled, 5— 15 cm long; petioles minutely glandular, 1-2 cm long with a sessile cupulate gland often occurring at the mid-point, or occasionally on the lower or upper third (cupulate gland is also frequently present on the rachis at the base of the terminal pair of leaflets and occasionally between the base of the three or four subterminal pairs of leaflets); pinnae opposite or subopposite, 1—13(—18) pairs; leaflets 9-23 pairs, glabrous, but often with minute marginal glands, linear, obtuse, mu- cronulate, 5-8 mm long, 0.6—1.8 mm wide, the midrib strongly ec- centric; inflorescence of 1(—3) axillary peduncles (sometimes peduncles terminal on short axillary stems) 1—2.5 cm long, the receptacle ovoid to oblong, 1.4—3(—5) mm long, 0.8—2 mm broad; flowers in globose heads 2 cm in diameter, 25 or more per head, but only one or two developing pods, whitish, pedicellate, the inconspicuously glandular 222 MADRONO [Vol. 28 ACACLA KELLOGSLANA Carter and itudd Sp. nov. BAJA CALIFORNIA SUR, MEXICO Sierra de Ja Giyanta Fic. 1. Holotype of Acacia kelloggiana Carter & Rudd. Carter & Sousa 5152. 1981] CARTER & RUDD: NEW SPECIES OF ACACIA 223 po eo WW (TIT ryt Fic. 2. Acacia kelloggiana. Left: Unilaterally dehisced legumes. Carter 5611; Right: Strongly recurved spines. Carter 5241. pedicels 2-3 mm long; calyx glabrous (or nearly so), valvate, 2.5—3 mm long, about two thirds the petal length; petals 3—4 mm long with narrow white-membranous glandular margins; stamens exserted, ca. 8 mm long; legume 8-15 cm long, 1.5—2 cm wide, 2—8-seeded, com- pressed, somewhat falcate, sinuate-margined, slightly constricted be- tween the seeds, subcoriaceous, margin slightly thickened, tan to black in age, usually unilaterally dehiscent, the stipe ca. 5-15 mm long; seeds subovate to ovate, compressed, 7—9 mm long, 6 mm wide, 1.5— 3 mm thick, uniformly dark brown, the pleurogram inconspicuous. TYPE: Mexico, Baja California Sur, Sierra de la Giganta: vicinity of La Matancita, Arroyo Hondo, north side of Cerro Giganta, ca. 26°08'N, 111°34’W, 750 m, 13 Oct. 1966, Annetta Carter & Mario Sousa 5152 (Holotype: UC 1472700; isotypes: MEXU, US, BM). PARATYPEsS: All Arroyo Hondo collections are from Cerro Giganta (1766 m) nw. of Loreto, Baja California Sur, Mexico, the principal peak of the Sierra de la Giganta, a mountain range that extends along the Gulf of California side of the peninsula from 24°30’ to 26°30’N. Arroyo Hondo heads into the n. and ne. side of the peak and forms a huge basin at the base of high basaltic cliffs. La Matancita (750 m) is a permanent water spring high in the nw. side of Arroyo Hondo 224 MADRONO [Vol. 28 30 i) Oo Number of leaves o lr 2°°3-4- & 6 7 8 9 10 WW 12 is 14 15) 160 l¢ Je Pairs of pinnae Fic. 3. Acacia kelloggiana. Histogram representing frequency distribution of pairs of pinnae (based on 222 leaves of 14 individuals). basin. (Except as otherwise noted, all collections are at UC. Duplicates will be distributed.) Alt. 690-820 m. Arroyo Hondo: 12 Dec 1938, H. S. Gentry 4110 (DS, DES, K, UC); 24 Nov 1947, Carter, Alexander & Kellogg 2007; 14 Oct 1966, Carter & Sousa 5170, 5176; 29 Aug 1971, Carter 5622, 5623, 5623a, 5624. Vicinity of La Matancita, Ar- royo Hondo: 13 Oct 1966, Carter & Sousa 5156; 3 June 1967, Carter 5241; 28 Aug 1971, Carter 5611 (previous season’s dehisced legumes from under shrub), 56172. With scattered Nolina beldingii and legu- minous shrubs, ridge nw. of main peak (Cerro Giganta), 1400—1500 m, abundant at this elevation and some distance below, 26 Nov 1947, Carter, Alexander & Kellogg 2039. Sierra de las Palmas: La Cham- pagna, s. of Santa Rosalia, 1440-1600 m, 27-29 Apr 1952, Fox & Gentry 11805 (DES). At the height of the dry season the shrubs are leafless (e.g., Carter 5241, June); they come into leaf and bloom with the advent of the summer rains, and by November and December only the dehisced legumes are found. During the course of his 1930 field trip to Baja California Marcus Jones (1933) did some collecting in the Sierra de la Giganta. From Loreto he went by animal to Arroyo Hondo (his Arroyo “Undo”). In his field journal (p. 107) Jones says, “There is an Acacia here with white flowers and hooked spines that is very annoying and at all elevations, rarely 10 ft. high, but have seen little fruit of it.”” We have 1981] CARTER & RUDD: NEW SPECIES OF ACACIA 225 been unable to locate any Jones specimens of this Acacia in various herbaria, but he is undoubtedly referring to our A. kelloggiana. Vegetation in the basin of Arroyo Hondo is dense, with Quercus tuberculata the dominant in the several steep canyons fingering up to the basalt cliffs of the Comondu Formation (Beal, 1948, pp. 74-77). Other common species at the higher elevations (750-850 m) are Ly- stloma divaricata, Mimosa purpurascens, Erythrina flabelliformis, Croton magdalenae, Jatropha vernicosa, Karwinskia humboldtiana, Pachycormus discolor, Alvordia glomerata and Franseria arborescens. On exposed slopes and at lower elevations (600—700 m) are Lysiloma candida, Jatropha cuneata, Fouquieria diguetit and Lemaireocereus thurber1. Acacia kelloggiana is abundant in the above two associa- tions. On the lower slopes and flats of Arroyo Hondo basin Prosopis palmeri, Jatropha cinerea and Ruellia peninsularis are common. F7- cus palmeri and Pachycereus pringlei are scattered throughout. Local ranchers say that this is the only locality where they have seen “gar- abatilla de espina negra”. The Gentry and Fox collection from Sierra de las Palmas was growing in Nolina-grassland on undulating, broken terrain of a volcanic mountain top. Such a habitat and association occurs also on the crest of Cerro Giganta. The common name “garabatilla de espina negra” serves to differ- entiate Acacia kelloggiana from “garabatilla”, Mimosa purpu- rascens, a common shrub in the Sierra de la Giganta. The spines of the latter are also broad-based and strongly recurved, but they are light-colored and internodal. Acacia kelloggiana is named in memory of Louise Kellogg, with whom, in company of Annie M. Alexander, the senior author made her first trip to Baja California in 1947, as well as a number of sub- sequent memorable trips following Miss Alexander’s death in 1950. Alexander and Kellogg botanical specimens were collected in many remote parts of California and Nevada, and their collection numbers reached almost 6000. Many of their specimens serve as the bases for new taxa; duplicates have been distributed widely by UC, where the first set is deposited. LITERATURE CITED BEAL, C. 1948. Reconnaissance of the geology and oil possibilities of Baja California. Mem. Geol. Soc. Amer. 31:1—138. JONES, Marcus E. 1933. Contrib. Western Bot. 18. 158 pp. Claremont, CA. WIGGINS, I. L. 1980. Flora of Baja California. Stanford University Press, Stanford, CA. (Received 14 Oct 1980; accepted 28 Dec 1980; revision received 6 Feb 1981.) RE-ESTABLISHMENT OF ANGELICA CALIFORNICA (UMBELLIFERAE) JosEPH M. DITOMASO Biological Sciences Department, Humboldt State University, Arcata, California 95521 ABSTRACT Angelica californica Jepson emend. DiTomaso, previously included within Angelica tomentosa Watson, is re-established as a distinct species from the Sierra Nevada foothills and the Coast Ranges of northern California. Illustrations, a distribution map, and several distinguishing characters are provided. Nine species of Angelica are believed to occur in California, the eight recognized by Munz (1959) and the recently described A. callii Math. & Const. (1977). Four of these are found along the Sierra Ne- vada—Cascades axis from Shasta County to Tulare County. The re- mainder occur in the western, coastal portion of the state from Sis- kiyou and Del Norte Counties to San Diego County. Jepson (1893) described Angelica californica on the basis of a single collection from the Vaca Mountains of Solano County and noted its similarity to Angelica tomentosa Watson. In 1901, Jepson emended his treatment and demoted A. californica to a variety of A. tomentosa. Since that time, there has been a great deal of confusion about the identity of coastal foothill Angelica. As in A. arguta Nutt. ex Torrey & Gray, the ovaries of A. californica are glabrous, or nearly so, in contrast to the densely pubescent ovaries of A. tomentosa. Based on this character, specimens of A. californica sensu Jepson key to A. arguta in Munz (1959). However, most herbarium sheets of A. cali- fornica have been annotated as A. tomentosa, presumably because A. arguta is a more northern taxon. After visiting many populations and studying numerous herbarium specimens, I have found several additional differences between A. tomentosa and A. californica (Table 1). Judged on the basis of these criteria, A. californica extends as far north as Shasta County and as far east as Butte and Tehama Counties (Fig. 1). Jepson’s (1893) de- scription of A. californica was not only incomplete with respect to important morphological characters, but it also failed to indicate the range of variability within the species. The present study more ac- curately describes A. californica, defines its range, and proposes its re-establishment as a distinct species. ANGELICA CALIFORNICA Jepson emend. DiTomaso.—Angelica cali- fornica Jepson, Erythea 1:8. 1893. Angelica tomentosa S. Wats. var. californica Jepson, Fl. W. Middle Calif. 356. 1901. MADRONO, Vol. 28, No. 4, pp. 226-230, 8 December 1981 1981] TABLE 1. Character Umbel shape in mature fruit Ray orientation Fruit pubescence Leaflet color Leaflet length/width DI TOMASO: ANGELICA CALIFORNICA A. californica Flat-topped or bowl- shaped Ascending Glabrous (pubescent in Tehama Co.) Abaxial surface light green; adaxial surface green 2:1 to slightly less 227 CONTRASTING CHARACTERS OF Angelica californica AND A. tomentosa. A. tomentosa Spherical Spreading Scabrous to tomentose (glabrous in Siskiyou Co.) Both adaxial and abaxial surfaces glaucous Usually 3:1, occasionally 2:1 ratio or 5:1 Oil tubes (vittae) Interval 1-3 1 (rarely 2) Commissure 2-6 2 (occasionally 4) Total vittae 6—20 6 (occasionally 8 or 10) Sandstone, shale, or Usually serpentine volcanic Soil type Flowering time May to early July July to October Plants stout, 1—2.5 m tall, the stem and foliage glabrous to pubes- cent, strongly scented; leaves deltoid, bipinnate to three times pin- nately divided, to 12 dm in length, 8 dm in width; leaflets lanceolate to ovate or oval, (2—)4—8(—14) cm long, (1—)2—4(—8) cm broad, acute to obtuse, the larger petiolulate and with 1 or 2 narrow lobes or leaflets at base, the others sessile, length/width ratio 2:1 or less, excluding petiolule, sharply serrate, the teeth acute to acuminate, irregularly spaced, the abaxial surface glabrous to pubescent and slightly lighter in color than adaxial surface, both surfaces scabrous on veins; petiole stout, 1-6 dm long, sheathing at base; cauline leaves reduced upward, pinnate, the uppermost sheaths bladeless; inflorescence usually gla- brous, the umbels flat-topped in flower, becoming concave and bowl- shaped in fruit; involucre wanting, or rarely present; rays 15-50, 2— 13 cm long, usually glabrous, or occasionally hispidulous at base and apex, ascending or curved upward, unequal, usually webbed; invol- ucel of 1-10 inconspicuous filiform bractlets, or lacking; pedicels 1—15 mm long, spreading-ascending, usually glabrous, occasionally webbed; flowers white or rarely pinkish, the petals oval to obovate, glabrous to sparsely puberulent or rarely pubescent; styles slender, much longer than the conical stylopodium; ovaries glabrous, or rarely pubescent; fruit green to purple, oval to oblong, 6—7(—10) mm long, 4—6(—7) mm broad, the dorsal ribs low, rounded, the lateral ribs broad- er than the dorsal but narrower than to equal to the body; vittae irregular in size and variable in number (6—20), 1—3 under the inter- vals, often appearing continuous about the seed, 2—6 on commissure (Fig. 2). 228 MADRONO [Vol. OREGON DEL NORTE CALIFORNIA SISKIYOU Goose Lake \ ‘. teint Ty pe ‘ u ' LASSEN oy Tr e Lak. a e Ad R S ini L River NN SaG ; en e e \ Ne Ply J y : Z 1 Eagle Lake | ~ Rey , Shasta Lake | : 1 \ Oa : 7 Whiskey town Lake S\ Mt. Lassen gee ’ —_— ‘ SIERRA i Lake Oroville eZ a 2 ( e we \ * | (ie ee eee tes | ; ie NEVADA x \? COLUSA Stan a EOD ae j Sia ae o ed 2 a ‘go y J” PLACER Lake Tahoe o s dfn ee 4 Clear Lake ! i % a a : Ne, : — 39°N N i -_— ae < YOLO si aa" ae ys b ube Yuba River — Folsom Lake : Lake Berryessa e \ oS. | 4 rat \ oe a aad | Pd 7 AMADOR e@ e “* NAPA e SOLANO a . @ ie | ° av PACIFIC OCEAN ‘i e- L x i 4 . KC Z vA SACRAMENTO p Ne Bo @ at = ¢ wi * — TUOLUMNE cr roan aa we S aX WS \ CALAVERAS er, et: = SAN JOAQUIN » me, Jatt : oy Sues cares wg e@ CONTRA COSTA < | - p —_—_—_—__———_ \< . @ i 50 KM » ra | So Tt Ae SAN FRANCISCO ane c ae / ‘ MARIPOSA ee . pes : SANTA CLARA A MERCED ‘ \ x Fic. 1. Distribution of Angelica californica Type: USA, CA, Solano Co.: Gates Canon, Vaca Mountains, 20 June 1892, W. L. Jepson 14246 (Holotype: JEPS!. Topotype: Di Tomaso 1744, HSC). Habitat and Distribution. Dry volcanic, shale, or sandstone slopes 28 1981] DI TOMASO: ANGELICA CALIFORNICA 229 Nay ()) () 4 Wy H) ya ee YL f chy WA. Ane a Ty AY} ke) vv WARY AN) (Wy- 4 f SYA) f ff y Ny Wi); WZ TN (SWZ Fic. 2. Angelica californica. A, habit. B, basal leaf. C, mature inflorescence. D, dorsal view of entire fruit. E, transection of fruit. F, petal. All from DiTomaso 1595, 1732, and 1736. 230 MADRONO [Vol. 28 between 20 and 1600 m, from Shasta County to Butte County in the foothills of the Sierra Nevada and to Contra Costa County in the Coast Ranges. Jepson’s description of A. californica states that “the leaflets are always smaller and usually much thinner” than in A. tomentosa. In addition, he notes that A. tomentosa is “hoary-tomentose, has equal rays, and solitary depressed oil-tubes in the intervals”, as compared to “3 oil-tubes in the intervals” of A. californica. It is puzzling that Jepson neglected to consider ovary pubescence, ray orientation, and glaucousness of leaflets in his comparison of the two taxa. I have found these characters to be very effective in separating A. californica and A. tomentosa in the field. However, several of the important field characters, e.g., glaucousness, ray orientation, umbel shape, and ori- entation of the mature fruiting stem, are not always evident in her- barium material. This, and the presumed restriction of A. californica to the Vaca Mountains, may have contributed to Jepson’s later deci- sion, in 1901, and that of Mathias and Constance (1944—45) to include A. californica in A. tomentosa. ACKNOWLEDGMENTS I thank Dr. J. P. Smith Jr., Dr. J. O. Sawyer Jr., Dr. D. E. Anderson, Dr. Lincoln Constance, and especially Dr. Michael Mesler for their helpful comments and sugges- tions. I am also grateful to the directors of the following herbaria for the use of speci- mens: CAS, CDA, CHSC, DAV, DS, HSC, JEPS, PUA, UC. I would also like to thank Christina Paleno for the illustrations. LITERATURE CITED JEPpSGN, W. L. 1893. Studies in the Californian Umbelliferae. Erythea 1:8—10. . 1901. A Flora of Western Middle California. Encina Publishing Co., Berkeley. MATHIAS, M. E. and L. CONSTANCE. 1944-1945. Umbelliferae. Jn North American Flora, 28B:43-295. 1977. Two New Local Umbelliferae (Apiaceae) from California. Madrono 24:78. Munz, P. A. 1959. A California Flora. Univ. Calif. Press, Berkeley. (Received 10 Nov 1980; revised version received 3 Feb 1981; accepted 6 Feb 1981.) COMPOSITION OF NATIVE GRASSLANDS IN THE SAN JOAQUIN VALLEY, CALIFORNIA LYNDON WESTER Department of Geography, University of Hawaii, Honolulu, Hawaii 96822 ABSTRACT The native grasslands of California have undergone great change since European contact but early accounts of Spanish and Anglo-Americans provide some information about their former condition. They suggest that the dry alluvial fans of the San Joaquin Valley, which account for 60 percent by area of California grasslands, were dominated by annual species and xerophytic shrubs. Perennial bunchgrasses were common only on certain well-watered floodplains. Alterations in the grasslands of California as a result of European contact and settlement were great and began so early in the historic period that the former condition of these grasslands will always be open to question. Evidence from written documents and contemporary field observations supports the view that perennial bunchgrasses were abundant in communities now composed largely of exotic annuals. This had led to the conclusion, now widely accepted, that all native grasslands were dominated by perennial species (Munz and Keck, 1949; Clark, 1956; Oosting, 1956; Benson, 1957; Burcham, 1957; Munz, 1959; Muller and Muller, 1964; Wells, 1964; Dasmann, 1966; McCown and Williams, 1968; Crampton, 1974; Ornduff, 1974; Heady, 1977; Kuchler, 1977). Some of the interpretations of fact have been questioned and doubt has been expressed that the bunch grass- lands were as extensive as has been assumed (Biswell, 1956; Twissel- mann, 1963, 1967; Klapp, 1964; Naveh, 1967; McNaughton, 1968). The objective of this paper is to review the evidence on the nature of the prehistoric grasslands, giving emphasis to documentary informa- tion from the southern Central Valley. It was here on the dry alluvial fans that the largest tracts of native grassland occurred (Fig. 1), yet most of the evidence that has been used to reconstruct the former condition of the community is derived from much more humid sites either along the coast or at higher elevations. DOCUMENTARY EVIDENCE Perennial bunchgrasses. Toward the end of the nineteenth century the deterioration of the quality of California rangelands was investi- gated in several important surveys reviewed by Talbot and Crone- miller (1961). Of particular concern was the invasion by annuals and MADRONO, Vol. 28, No. 4, pp. 231-241, 8 December 1981 239 MADRONO [Vol. 28 SAN FRANCISCO Y Los % Banos \® SAN JOAQUIN VALLEY, CALIFORNIA Kern m VISTA @ BAKERSFIELD TEJON PASS ANNUAL PRECIPITATION CALIFORNIA PRAIRIE (Stipa spp.) of CALIFORNIA according to KUCHLER (1977) >40 Inches 20-40 Inches [] 10-20 Inches [_]<10 Inches om am om om in oe os Fic. 1. The natural setting of native grasslands in California. Upper: The San Joaquin Valley. Lower: California precipitation (left); California prairie (right), after Kuchler, 1977. 1981] WESTER: CALIFORNIA GRASSLANDS 233 replacement of perennial grasses that reduced the carrying capacity of the range. In a study of a portion of northwestern California Davy (1902) carefully documented reports of residents who observed the decline of bunchgrasses over many years. More recently Burcham (1957), in his detailed history of California rangeland, collected further early written descriptions of bunchgrass where annual grasslands stand today. Almost all of these sites are in northern coastal locations (Mendocino, Humboldt, and Monterey Counties) where mild, humid conditions prevail all year. One exception is an account from Bryant (1848, p. 309) which was made during a journey from the San Joaquin Valley to San Jose in 1847. From this plain we entered a hilly country, covered to the sum- mits of the elevations with wild oats and tufts or bunches of a species of grass, which remains green throughout the whole sea- son. The reference is unmistakably to the hills of the Coast Range where pockets of bunchgrasses may still be found. Yet this description stands in strong contrast to the scene in the San Joaquin described by the same person at another point along the way. The more arid Valley plains which stood above the river bottoms were characterized as “dry and crisp” with “large tracts of wild oats” (Bryant, 1848, p. 300). No mention was made of bunchgrasses at these locations. The absence of information about natural conditions in the San Joaquin Valley is often attributed to the lack of detail in the first accounts by the Spanish and the disruption of the ecosystem by feral herbivores before Anglo-Americans made more accurate descriptions. It is true that feral horses were present in the Valley at least by 1807 (Cook, 1960) and extremely large herds were noted after 1830, when hunting by Indians ceased as a result of decline in human pop- ulations (Bryant, 1848; Fremont, 1848; Leonard, 1904; Farquhar, 1937). However, there are instructive Spanish accounts, made even before feral livestock could have had significant effect, which leave no doubt about the scanty natural plant cover at least during certain times of the year. Zalvidea described the area around Buena Vista Lake in July 1806 in the following manner: The area covered in the morning consisted of extensive plains. In quality the land is alkaline. The shore of the lake is completely covered with a great deal of tule. Elsewhere, and in the hills bordering the plains, I saw neither pasturage nor watering places. (Cook, 1960, p. 245) Munoz, diarist on the Moraga expedition, in October of the same year judged the country he saw in present day Merced and Madera Counties to be equally barren. 234 MADRONO [Vol. 28 All the country traversed today has very poor grass and is very stony... . All the country we observed between the Tecolate [Chowchilla River] ... and the Santa Ana [Fresno River] is worse than bad. From the Santa Ana to the San Joaquin there is a little pasturage, although it is sparse and spread out widely. (Cook, 1960, p. 251) The Martinez expedition in 1816 saw the same region in May and even reported a bunchgrass growth form that almost certainly referred to Sporobolus airoides, a plant tolerant of high salinity and common in the marshes of the Valley even now. Otherwise the herbaceous vegetation was very poor. In all our trip we did not see a good tree, nor wood enough to cook a meal, nor a stone, nor even grass enough for the horses, more than bunchgrass, or what grows in the swamps. (Cook, 1960, DeZil) Similarly in June, 1824, Portilla said: “The road was flat and the land quite poor, with no grass” (Cook, 1962, p. 155). Similar observations were made by early Anglo-American travellers after 1840 and, quoted alone, they are often cited as evidence of en- vironmental degradation caused by feral cattle and horses. In fact barrenness may have been a natural condition of the landscape. Se- rious overgrazing may not have occurred until the droughts of 1861 and 1864, the first since stock had been brought to the Valley in large numbers in response to the demand for meat created by the gold rush. When Fremont passed over the land between the Kings and Kern Rivers in April 1844 he noted: To-day we made another long journey of about forty miles, through a country uninteresting and flat, with very little grass and a sandy soil. (Fremont, 1845, p. 253) Others reported similar conditions in the 1850’s. The Tulare Valley, from the mouth of the Mariposa to the Tejon pass at its head, is about one hundred and twenty miles in extent, and varies from eight to one hundred miles in width. With the exception of a strip of fertile land upon the rivers emptying into the lakes from the east, it is little better than a desert. The soil is generally dry, decomposed and incapable of cultivation, and the vegetation, consisting of artemisias and wild sage, is ex- tremely sparse. (Farquhar, 1937, p. 262) The plains between the streams are destitute of foliage, and the soil generally gravelly and poor. (Williamson, 1855, p. 13) There was but little or no vegetation, and the surface was dry and gravelly. (Blake, 1855, p. 41) 1981] WESTER: CALIFORNIA GRASSLANDS 235 After leaving the grove by the [Kern] river, we entered at once among the most desolate hills. Not a sign of herbage was seen on them—not enough to attract a bee. (Kip, 1954, p. 92) One might expect that had bunchgrasses been present at least the basal tussock would have been obvious throughout the year and might have attracted some attention. In fact, Fremont, whose descriptions are the most detailed of any explorer, does mention bunchgrasses on two occasions. Both of these were in exceptionally well-watered sites east of the delta in the Sierra Nevada foothills. Those sites receive runoff from the mountains in addition to the local precipitation. Leaving the Mo-Kel-um-ne, . . . we travelled about twenty miles through open woods of white oak, crossing in the way several stream beds—among them the Calaveras creek. These have abundant water, with good land above; and the Calaveras makes some remarkably handsome bottoms. Issuing from the woods, we rode about sixteen miles over an open prairie, partly covered with bunch grass, the timber reappearing on the rolling hills of the river Stanislaus in the usual belt of evergreen oaks (Fremont, 1848, p. 16). Emerging from the woods, we travelled in a south- easterly direction, over a prairie of rolling land, the ground be- coming somewhat more broken as we approached the To-wal- um-ne river, one of the finest tributaries of the San Joaquin. The hills were generally covered with a species of geranium, (ervodium cicutarium), a valuable plant for stock, considered very nutri- tious. With this was frequently interspersed good and green bunch grass, ... (Fremont, 1848, p. 17) Fremont makes it clear that this verdant condition was confined to the northeast portion of the San Joaquin Valley, because only a little farther south, beyond the Merced River, he noted: . . . the country had lost its character of extreme fertility, the soil having become more sandy and light... . (Fremont, 1845, p. 250) Annual herbs. If perennial bunchgrasses were not common the question arises what were the dominant herbaceous species. Other accounts from the more typical dry plains of the Valley make no men- tion of bunchgrasses but describe in some detail the annual herbs that grew abundantly, at least in wet years, and appeared to be the dom- inants in the community. For example, in the spring of 1850 a traveller making his way through the Coast Ranges at the latitude of Los Banos observed the change in aspect of vegetation as he approached the Valley. By this time we could see what had caused the mass of color so noticeable from the mountain the day before. The entire plain, 236 MADRONO [Vol. 28 as far as we could see, was covered with wild flowers. Almost all of the flowers were new to us... . As we passed below the hills the whole plain was covered with great patches of rose, yellow, scarlet, orange and blue. The colors did not seem to mix to any great extent. Each kind of flower liked a certain kind of soil best and some of the patches of one color were a mile or more across. (Mayfield, 1929, p. 9) A few years later Muir described a similar phenomenon: The Great Central Plain of California, during the months of March, April, and May, was one smooth, continuous bed of hon- ey-bloom, so marvelously rich that, in walking from one end of it to the other, a distance of more than 400 miles, your foot would press about a hundred flowers at every step. Mints, gilias, ne- mophilas, castilleias, and innumerable compositae were so crowd- ed together that, had ninety-nine per cent of them been taken away, the plain would still have seemed to any but Californians extravagantly flowery . ... Because so long a period of extreme drought succeeds the rainy season, most of the vegetation is com- posed of annuals, which spring up simultaneously, and bloom together at about the same height above the ground, the general surface being slightly ruffled by the taller phacelias, pentstemons and groups of Salvia carduacea, the king of the mints. (Muir, 1894, p. 342) Fremont also mentioned fields of wildflowers during his 1845 expe- dition even though it was early in the season (January and February). The California poppy, (Eschscholtzia Californica,) the character- istic plant of the California spring; memophila insignis [sic], one of the earliest flowers, growing in beautiful fields of a delicate blue, and erodium cicutarium, were beginning to show scattered bloom. (Fremont, 1848, p. 19) Descriptions of spring wildflower blooms have not been found in the Spanish records because most expeditions were made in summer months. However, in July, 1806, Zalvidea reported flowering of a summer growing herb, probably Hemizonia pungens, in the southern San Joaquin Valley. All this territory is covered with a species of herb which has a little stem with a yellow flower, the stalk being no more than a quarter [of a yard] high. (Cook, 1960, p. 246) The occurrence of Erodium as a component of herbaceous cover at an early date is of special interest because the common species (in- cluding E. cicutarium) are generally considered to be native of Med- iterranean Europe (Robbins, 1951; Clark, 1956; Munz, 1974). 1981] WESTER: CALIFORNIA GRASSLANDS 237 Fremont mentioned it on a number of occasions during both his 1844 and 1845 expeditions (Fremont, 1845, 1848) and leaves no doubt about its abundance in the Central Valley, and the fact that Indians made use of the plant. Instead of grass, the whole face of the country is closely covered with erodium cicutarium, here only two or three inches high. Its height and beauty varied . . . being, in many low places which we passed during the day, around streams and springs, two and three feet in height. (Fremont, 1845, p. 253) Other accounts from the first ranchers to settle a portion of the West Side plains near present day Coalinga emphasize how Erodium ap- peared to dominate the ground cover presumably in the absence of tall growing perennial grasses. This valley was covered with the finest possible stand of dry alfileria, remaining from the extremely wet winter of 1852. (Latta, 1949, p. 333) Erodium cicutarium was apparently common throughout the South- west at the time of the first scientific explorations (Torrey, 1859) and was so widely naturalized in California even early in the nineteenth century that Brewer and Watson (Calif. Geol. Survey, Bot., 1880) doubted that it was an exotic. The discovery of the species in the earliest known adobe bricks made by the Spanish suggested to Hendry (1931) that it spread into California before European settlement, a possibility that Jepson (1933) also accepted. The plant has very effec- tive dispersal mechanisms and others of the genus are native to North America and Australia. It is possible that E. cicutarium itself may have reached the New World without human assistance. If the species was either indigenous to California or spread ahead of settlement, then its presence, especially in more arid sites, cannot be used as an indi- cation of environmental degradation. Hoover (1935) felt grasses were relatively unimportant in the “prim- itive” flora of the San Joaquin Valley. References to them in the writ- ten records are rarely specific enough to allow identifications to be made with confidence. Several early references to ‘wild oats’ have been found (Bryant, 1848; Perkins, 1863; Leonard, 1904; Latta, 1949), but the name may have been applied to many annual grasses in the same way that ‘sage brush’ was used to describe any grey-green shrub and not specifically species of Artemisia. It is less likely that Avena could be mistaken for a bunchgrass that possesses a quite different life-form. RELICT ANALYSIS Observations by Davy (1902) in northwestern California strongly suggested that sites protected from grazing tended to contain more 238 MADRONO [Vol. 28 abundant native perennial bunchgrasses. In 1917 and 1918 Clements found the bunchgrass Stipa pulchra (probably including S. cernua) common in fenced railroad rights-of-way in the Central Valley and, believing them protected from grazing and burning, concluded that this drought-tolerant perennial must have dominated the grassland before grazing caused its replacement (Clements, 1934; Clements and Shelford, 1939). However, Biswell (1956) has pointed out that these sites were burned almost annually to prevent accidental fires and this Stipa, which is favored by burning, probably became established only as a result. In the southwestern San Joaquin Valley other relict sites protected from grazing, such as fenced road sides, oil fields, quarries, and ar- royos were searched but no perennial grasses were seen. Instead, xe- rophytic shrubs, particularly of Atriplex polycarpa, were often found growing more prolifically than on adjacent grazed rangeland. This saltbush is rated highly as a browse (Piemeisel and Lawson, 1937; Chatterton, 1970) and its decline under grazing has been noted (Love and McKell, 1966). It is quite possible that in the drier portion of the Central Valley, especially the West Side, this community of low shrubs was once more extensive than at present. PRESENT DISTRIBUTION PATTERNS It is often stated that Stipa pulchra and S. cernua were dominants in the grasslands and occupied more space than all the other species combined. Poa scabrella, Aristida divaricata, Koeleria macrantha, Melica imperfecta, and M. californica may have been generally wide- spread whereas Danthonia, Festuca, Deschampsia, Agrostis, and Muhlenbergia species may have had more restricted distributions (Shantz and Zon, 1924; Clements, 1934; Clements and Shelford, 1939; Beetle, 1947; Munz and Keck, 1949; Burcham, 1961; Crampton, 1974; Ornduff, 1974). Twisselmann (1963, 1967) observed Stipa to be un- common where annual precipitation falls below 245 mm and does not occur at all in places receiving less than 200 mm. The distribution map of Stipa published by Stebbins and Love (1941) supports Twis- selman’s observation by showing Stipa to be absent from the dry western San Joaquin Valley and elsewhere confined largely to riparian or foothill sites. Furthermore, this is in close agreement with an early report by Brewer to Watson when the former made collections for the first flora of the State in 1863. Stipa setigera [S. pulchra, S. cernua]. It is common on the Coast Ranges and on the foothills of the Sierra Nevada and according to Prof. Brewer, is the most common and, valuable “Bunch- grass” of the dry hills. (Calif. Geol. Survey, Bot., 1880, vol. 2, p. 286) 1981] WESTER: CALIFORNIA GRASSLANDS 239 Botanical reconnaissance made in conjunction with the railroad sur- veys of the Central Valley indicates that 75 grass species were col- lected, of which only one or perhaps two, Poa douglasiz and Elymus sp., are perennials (Durand and Hilgard, 1855). Had bunchgrasses been as abundant as supposed, one would expect them to be better represented in the collections made before significant settlement oc- curred in the region. Areas of bunchgrass identified by Crampton (1974) in the delta re- gion occupy relatively moist sites influenced by the cool, humid, mar- itime air able to penetrate to this part of the Central Valley through the San Francisco Bay gap. These conditions are not typical over the remainder of the Valley. In California, variable and unpredictable moisture, temperature, and light at the time of germination and the cool winter months cause extraordinary variation in productivity and floristic composition from year to year (Heady, 1956; Naveh, 1967). Klapp (1964) has concluded that ephemeral annuals are best adapted to this unpredictable and inconsistent climate. In its drier phases the climate is unsuitable for most perennial herbs, and the grasslands of this region are composed mainly of annual species, standing in marked contrast to communities in other temperate areas. CONCLUSIONS Evidence for the former importance of perennial bunchgrasses in the grasslands of California, and their subsequent decline as a result of grazing, exists for many places in the Coast Ranges, the Sierra foothills and in some localized, well-watered floodplains in the interior. However, the San Joaquin Valley, which contained much of Califor- nia’s native grasslands, was either wetland of fluctuating extent or dry alluvial fan. The degree of its natural aridity may not have been appreciated because earliest settlement was concentrated along the riparian oases and later large scale irrigation schemes caused almost all of the thin natural vegetation cover of the plains to be replaced by highly productive agriculture. Observations used to reconstruct the former composition of grassland communities comes from humid parts of the State that were settled first. This information cannot be extrap- olated to explain conditions in drier locations such as the San Joaquin Valley. Accounts of the San Joaquin Valley by Anglo-Americans are said to represent descriptions of an environment already degraded by the grazing of feral livestock, yet observations made by the Spanish before any significant impact of European civilization are quite consistent with those of the middle of the nineteenth century. Neither indicate the existence of perennial bunchgrasses but instead emphasize the lack of vegetation cover during the dry months and the abundance of bril- 240 MADRONO [Vol. 28 liantly flowering herbs in the spring. This leads to the conclusion that, except for riparian and wetland sites, much of the southern Central Valley supported a grassland of annual species or, in the most arid parts on some soils, a community of xerophytic shrubs with an under- story of annuals. LITERATURE CITED BEETLE, A. A. 1947. Distribution of the native grasses of California. Hilgardia 17:309- 357: BENSON, L. 1957. Plant classification. D. C. Heath, Boston. BIDWELL, J. 1866. Annual Address, Trans. Calif. State Agric. Soc. 1864—1865:202-— 213; BISWELL, H. H. 1956. Ecology of California grasslands. J. Range Managem. 9:19-24. BLAKE, W. P. 1855. Geological report. Jn Explorations and surveys for railroad route from the Mississippi River to the Pacific Ocean. Vol. 5. 33rd U.S. Congress, House Exec. Doc. 91. BRYANT, E. 1848. What I saw in California. Appleton, New York. BuRCHAM, L. T. 1957. California rangeland. Calif. Dept. Nat. Resources, Div. For- estry, Sacramento. 1961. Cattle and range forage in California: 1770-1880. Agric. Hist. 35:140— 149. CALIFORNIA GEOLOGICAL SURVEY, BOTANY. 1880. 2 vols. J. Wilson and Sons Uni- versity Press, Boston. CHATTERTON, N. J. 1970. Physiological ecology of Atriplex polycarpa: growth, salt tolerance, ion accumulation, and soil-plant-water relations. Ph.D. dissertation, Univ. California, Riverside. CLARK, A. K. 1956. The impact of exotic invasion on the remaining New World midlatitude grasslands. In W. L. Thomas, ed., Man’s role in changing the face of the earth, p. 737-762. Univ. Chicago Press, Chicago. CLEMENTS, F. E. 1934. The relict method in dynamic ecology. J. Ecol. 22:39-68. and V. E. SHELFORD. 1939. Bioecology. Wiley, London. Cook, S. F. 1960. Colonial expeditions to the interior of California, Central Valley, 1780-1820. Anthropol. Rec. 16:239—292. 1962. Expeditions to the interior of California, Central Valley, 1820-1840. Anthropol. Rec. 20:151-213. CRAMPTON, B. 1974. Grasses of California. Univ. California Press, Berkeley. DASMAN, R. F. 1966. The destruction of California. Collier-Macmillan, New York. Davy, J. B. 1902. Stock ranges of northwestern California: Notes on the grasses, forage plants and range conditions. USDA, Bur. Plant Industry, Bull. 12. DURAND, E. and T. C. HILGARD. 1855. Botanical report, Part HI. Description of plants collected upon the expedition. Jn Explorations and surveys for a railroad route from the Mississippi River to the Pacific Ocean. Vol. 5. 33rd Congress, 2nd Session, House Exec. Doc. 91. FARQUHAR, F. P. 1937. The topographic reports of Lieutenant George H. Derby. Calif. Hist. Soc. Quart. 11:247-265, 365-382. FREMONT, J. C. 1845. Report of the exploring expedition to the Rocky Mountains in the year 1842 and to Oregon and California in the years 1843-44. 28th Congress, 2nd Session, Senate Doc. 174. . 1848. Geographical memoir upon upper California. 30th Congress, 1st Session, Senate Misc. Doc. 148. HeEapy, H. F. 1956. Evaluation and measurement of the California annual type. J. Range Managem. 9:25-27. 1977. Valley Grassland. In M. G. Barbour and J. Major, eds., Terrestrial Vegetation of California, p. 491-514. Wiley-Interscience, New York. HENDRY, G. W. 1931. Adobe brick as a historical source. Agric. Hist. 5:110—127. 1981] WESTER: CALIFORNIA GRASSLANDS 241 Hoover, R. F. 1935. Character and distribution of the primitive vegetation of the San Joaquin Valley. Master’s thesis, Univ. California, Berkeley. Jepson, W. L. 1933. Yampah and filaree. Madrono 2:109-110. Kip, W. I. 1954. Early days of my episcopate. Biobooks, Oakland, CA. Kapp, E. 1964. Features of grassland theory. J. Range Managem. 17:309-322. KUCHLER, A. W. 1977. The map of the natural vegetation of California. In M. G. Barbour and J. Major, eds. Terrestrial vegetation of California, p. 909-938. Wiley- Interscience, New York. LaTTA, F. F. 1949. Black gold in the Joaquin. Caxton, Caldwell, Idaho. LEONARD, Z. 1904. Adventures of Zenas Leonard. W. F. Wagner, ed. Burrows, Cleveland. Love, R. M. and C. M. MCKELL. 1966. Proposed study, Temblor Range Research, Annual Report. 1965—1966:23. McCown, R. L. and W. A. WILLIAMS. 1968. Competition for nutrients and light between grassland species Bromus mollis and Erodium botrys. Ecology 49:981-— 990. McNAUGHTON, S. J. 1968. Structure and function in California grasslands. Ecology 49:962-972. [MAYFIELD, T. J.] 1929. San Joaquin primeval. Uncle Jeff’s story. Arranged by F. F. Latta. Tulare Times Press, Tulare, CA MulIr, J. 1894. The mountains of California. Century, New York. MULLER, C. H. and W. H. MULLER. 1964. Antibiosis as a factor in vegetation pat- terns. Science 144:889-890. Munz, P. A. 1959. A California flora. Univ. California Press, Berkeley. Munz, P. A. 1974. A flora of Southern California. Univ. California Press, Berkeley. and D. D. KEcK. 1949. California plant communities. Aliso 2:87—105. NAVEH, Z. 1967. Mediterranean ecosystems and vegetation types in California and Israel. Ecology 48:445—459. OOsTING, H. J. 1956. The study of plant communities. Freeman, San Francisco. ORNDUFF, R. 1974. An introduction to California plant life. Univ. California Press, Berkeley. PERKINS, J. E. 1863. Sheep husbandry in California. Trans. Calif. State Agric. Soc. 1863:134—-145. PIEMEISEL, R. L. and F. R. LAWSON. 1937. Types of vegetation in the San Joaquin Valley and their relation to the beet leafhopper. USDA Techn. Bull. 557:1-28. ROBBINS, W. W. 1951. Weeds of California. Calif. State Department of Agric., Sac- ramento. SHANTZ, H. L. and R. Zon. 1924. Natural vegetation. Jn Atlas of American Agri- culture. USDA, Washington, D.C. STEBBINS, G. L. and R. M. Love. 1941. An undescribed species of Stipa from Cal- ifornia. Madrono 6:137-141. TALBOT, M. W. and F. P. CRONEMILLER. 1961. Some beginnings of range manage- ment. J. Range Managem. 14:95—-102. TORREY, J. 1859. Botany of the boundary. Jn Report on the United States and Mexican boundary survey. Vol. 2, Pt. 1. 24th Congress. Ist Session, Exec. Doc., 135. TWISSELMANN, E. C. 1963. Some preliminary notes for a summary of the primitive flora of the upper San Joaquin Valley. Unpublished Ms. 1967. A flora of Kern County. Univ. San Francisco Press, San Francisco. WELLS, P. V. 1964. Antibiosis as a factor in vegetation patterns. Science 144:889. WILLIAMSON, R. W. 1855. Report of the explorations in California for railroad routes to connect with the routes near 35th and 32nd parallels of north latitude. Jn Explorations and surveys for a railroad route from the Mississippi River to the Pacific Ocean. Vol. 5. 33rd Congress, 2nd Session, House Exec. Doc. 91. (Received 19 Jun 1980; accepted 19 Nov 1980; revision received 18 Feb 1981.) POST-ERUPTION SUCCESSION ON ISLA FERNANDINA, GALAPAGOS LYNN B. HENDRIX Biology Department, Yakima Valley College, Yakima, Washington 98902 ABSTRACT In 1968 the mixed shrub forest on the western caldera rim of Isla Fernandina, Ga- lapagos, was buried by a major eruption of tephra. In June, 1971; July, 1973; and August, 1977, vegetation on the western rim was quantitatively sampled in plots along a transect extending 1.5 km from deep, barren tephra to the undestroyed original vege- tation. In the 9 years following the eruption, only a few species of weedy composites and grasses became sparsely established in gullies on the deep tephra. By contrast, where the original plants were shallowly buried, vigorous vegetative sprouting of shrubs and of rhizomatous perennial herbs resulted in nearly complete cover by 1977. Species apparently reproducing only by seed returned more slowly than those species with vegetative reproduction. Land iguana activity and precipitation appear to influence the rate and patterns of revegetation. The shrub forest is composed of weedy species able to survive volcanic disturbance and revegetate newly created open habitats. Rarely is the opportunity available to begin a study of plant succes- sion immediately following volcanic activity. The studies that have been done were usually on lava substrate, often in moist climates (Smathers and Mueller-Dombois, 1974). This paper reports six years of plant succession on tephra (fine-grained ash and coarser particles of pyroclastic origin) of arid Isla Fernandina, westernmost island of the Galapagos archipelago. These islands are one of the most active volcanic fields on earth, and their biology is closely related to their volcanic history. In May, 1968, Fernandina, a basaltic shield volcano, experienced a violent summit eruption and caldera collapse (Simkin and Howard, 1970). One feature of this activity was a ground-level surge of tephra from the caldera floor up on to the rim of the caldera and 10 km downslope to the sea. This was a cool, wet, sticky, high density, and high velocity flow (B. Nolf, pers. comm., 1977). Vegetation was oblit- erated in an area of approximately 25 km? on the western slope of the volcano and variously damaged in an area three times that. A few patches escaped in the lee of topographic shelters. Destruction was by burial under as much as 8 m of tephra toward the center of the flow (B. Nolf, pers. comm., 1977) and by uprooting, breaking, debarking, and partial burial toward the shallower edges. On the Galapagos Islands lava flows are the major alternative sub- strate to tephra. Lava flows present a most inhospitable habitat for plant colonization, and most are barren for long periods. For example, MADRONO, Vol. 28, No. 4, pp. 242-254, 8 December 1981 1981] HENDRIX: PLANT SUCCESSION, GALAPAGOS ISLANDS 243 one 1825 flow south of Punta Espinosa, Fernandina (B. Morrell, 1825, cited in Brower, 1968) had almost no plants growing on its surface after 150 years. By contrast, revegetation of the tephra on the rim of Fernandina was easily visible within nine years. The pre-eruption vegetation on the northeast caldera rim of Fer- nandina was a dense tangle of shrubs and herbs described qualitatively by Eliasson (1972) and Colinvaux (1968) as 2—3 m tall shrub forest dominated by Scalesia microcephala Robins. Similarly, on the north- west section of the rim Scalesia microcephala was a dominant shrub, but of nearly equal size and numbers were Zanthoxylum fagara (L.) Sarg., Tournefortia rufo-sericea Hook. f., and Solanum erianthum D. Don. Also present were Darwiniothamnus tenuifolius var. glandulo- sus (Harling) Crong., Lippia rosmarinifolia Anderss. var. rosmarin- ifolia, Baccharis gnidiifolia HBK., Alternanthera filifolia (Hook. f.) Howell, and a number of smaller shrubs and herbs. This shrub forest still existed apparently undisturbed around half of the rim and was, therefore, available as a source of plants for recolonization of the tephra. METHODS Reestablishment of rim vegetation on the tephra was studied 3, 5, and 9 years after the eruption (June, 1971; July, 1973; and August, 1977). A transect was established beginning on deep tephra at a per- manent metal post marking the high point (1494 m) of the west rim and running 18° east of due north. The transect extended north ap- proximately 1.2 km across deep barren tephra (Fig. 1) and continued 4 km across sparsely vegetated shallow tephra (Fig. 2) to the bound- ary of the original forest. In the barren area a circular plot of 9.1 m radius was established every 91 m and a count made of all plants encountered. In the sparsely vegetated area, circular plots of 1.5 m radius were established every 15 m for count of herbaceous species; in addition, a plot of 9.1 m radius was established at each third site (46 m intervals) for a count of shrub species. The same plots were resampled each year of the study, except that in 1977 the solitary 1.5 m radius plots were omitted. In 1973 additional sampling was done (3.1 m radius circular plots at 91 m intervals) along a 4 km line extending south from the southern edge of the barren tephra. In 1977 land iguana (Conolophus sp.) feces found in the plots were collected and the seeds they contained brought back to the laboratory. The seeds were tested for germination ability to determine if iguanas provide a possible means of seed dispersal to and across the tephra. Samples from 11 sites along the transect were put on moist paper in petri plates and placed in light. 244 MADRONO [Vol. 28 Fic. 1. View north from about meter 500 of the rim transect, 1977. Here the tephra layer was deep enough to destroy the vegetation entirely. Gullies interrupt the hard, smooth surface evident at the left foreground. Grass clumps visible at right are 0.2—0.3 m tall. RESULTS Distribution and density of dominant species encountered along the transect are shown in Fig. 3. The transect crossed three communities— barren, shrub, and the ecotone between them. These are discussed separately below. Barren area. ‘Toward the center of the tephra, where the deposit was thickest and the destruction of vegetation complete, little reveg- etation occurred in the nine years following eruption. Approximately 3 km of the rim along the west side of the caldera were still practically barren, another 3 km to the south only sparsely vegetated. The surface of the tephra was a hard crust of fine-grained material, broken in places by gullies centimeters to meters deep exposing coarser textured layers (Fig. 1). Hardening of the surface apparently occurred soon after deposit, as photos taken in successive years show essentially no change in gully pattern or size. Even the hardiest plants rarely were found growing on the hardened tephra surface. Colonization occurred almost exclusively where the surface was broken. Most plants grew in gullies, in the lines where the wall of the gully met the floor. Here seeds blown from the sur- rounding area could accumulate; the broken surface and occasional shade increased the seedlings’ chances of success. A second site of 1981] HENDRIX: PLANT SUCCESSION, GALAPAGOS ISLANDS 245 ~ Fic. 2. View north along the northern 4 km of the rim transect, 1971. Here at the tapering edge of the tephra layer, vegetation was only partially destroyed. Most vege- tation visible is 0.5—1.5 m tall. [Vol. 28 MADRONO 246 ‘(ayeos Bo0[) eY $70’ Jed sjueld jo taquinu jUssaidai sayeds [eIVIIA IU, “W006 ye ayeos Jo asuvys 9J0N “JDaSUeI} SUOTe UOT}EI0[ S}UasaIda1 sIxe [e}UOZLIOY ay], “IVad Aq sateds jueUTWOp Jo AjISUap pue UONNCINsSIq “¢ ‘OI (WYHddL MOTIVHS) HLYON (VYHd3L d33d) HLNOS Totes einen | oetoeya pees nnn glances an lee glen alee ical omen | enecetee | eesti | joememen) | eateioanae Uiceciiee |iehcures | mamas) | Seana] SaaS aes RISA Gl [ace (SaaS eS PE a | WOSrL OOVL OOEL OOZL OOLL OOOL 006 009 O0€ O WOSHL OOVL OOEL OOZL OOLL O00L 006 009 OOE O WOSHL OOHL OOEL OOZL OOLL OO0OL 006 009 o0€ OO eiebey ; | | — | | | uinj|Axoujuez | eyeydasouo1w | eIsajeos | ; eaolias-ojnd B1j4OjausNo | winyjuensa winuejos TS ENO snuosapaxy | I Sd <> ae i uod!ssadooA] : 00S i | | — So o vt wniadned uinjasiuuad ol eueodixaw OL sijsosbesg ALY: suadai ephisuy —— SG ae x snaoe1ajoO snyouos i a@NyHs ! INOLOOA IN3Y¥uva ZL6L GNY¥HS 4S3NOLOO3S ; NSYHVE EZ6L ANYHS | JNOLOOAZ » N3YYVE LZ6L 1981] HENDRIX: PLANT SUCCESSION, GALAPAGOS ISLANDS 247 colonization was at the base of rocks where morning fog condensing on the cool rock face could run to the ground. A third site was provided by scattered depressions that appeared to contain water at some time during the year. In the highlands of Santa Cruz, vegetation is most lush during a cool season from July to December (Van der Werff, 1979). No precipitation records are available for the rim of Fernan- dina; the few days observed in this study (June, July, and August) were dry. The limiting factors for plant growth on the tephra appeared to be lack of moisture and of appropriate establishment sites. Few species were able to grow at all in these conditions. Only three were found with any regularity: Sonchus oleraceus L., an annual, cosmopolitan, weedy composite; Aristida repens Trin., an annual, endemic, weedy grass; and Eragrostis mexicana (Hornem.) Link, a widespread, tropical, weedy grass. Other species found less frequently were Eragrostis ciliaris (L.) R. Br., Cyperus ligularis L., and Muhl- enbergia microsperma (DC.) Kunth, widespread tropical weedy species, and Cyperus anderssonit Boeck. and Verbena townsendii Svens, Galapagos endemics. Between 1971 and 1977 these species became less abundant at the north of the transect as shrubs grew and more abundant to the south as gradual breakup of the hard surface occurred (Fig. 3). Ecotone. This community is defined by the scarcity of any shrubs except Solanum erianthum and Baccharis gnidiifolia and the presence of three herbaceous species not found in the “barren” community: Pennisetum pauperum Steud. is a large perennial grass, endemic to Fernandina and Isabela high-elevation lava and cinder beds. Lyco- persicon cheesmanii Riley, a short-lived perennial endemic to the Ga- lapagos, and Exedeconus miersii (Hook. f.) D’Arcy, an annual com- mon in the Galapagos at sea level on sandy substrata and found also in Peru, are sprawling vines of the Solanaceae. All three are found only sparsely in the undisturbed original forest but are highly suc- cessful pioneer species on the edge of the tephra. No Exedeconus was found anywhere at the rim in 1977. The moisture-retaining surface layer of tangled vegetation and litter resulting from growth of these species provides a habitat in which other plants can grow. In 1977 the tiny moss Bryum argenteum Hedw. was found growing in extensive patches in this area. The capsules are only 8—9 mm high, and in August the entire plant appeared simply as a darker crust on the hard tephra surface. Interestingly, this species was also a post- eruption pioneer on the island of Surtsey in the North Atlantic (Fri- driksson, 1975). In 1971 much of the northern 4% km of the transect could be defined as “ecotone.” In 1973 and 1977 the ecotone community was a distinct band moving onto the tephra. This band was about 180 m wide in 1973 and 270 m wide in 1977. Figure 3 indicates the sharp demarcation 248 MADRONO [Vol. 28 of the advancing edge of vegetation that was evident in the field. This edge between “barren” and “ecotone” communities was defined by the sharp increase in both vegetative cover and number of species. The total number of species encountered in summed samples north of the barren/ecotone line (all “ecotone” plus all “shrub”) was 17-19 each year, contrasted with 8—9 species to the south (“barren” samples). This edge advanced approximately 180 m south onto the tephra between 1971 and 1973 and an additional 90 m between 1973 and 1977. Shrub area. In contrast to the still barren area at the center of the tephra, the % km at the northernmost edge experienced vigorous re- vegetation in the nine years following the eruption. Here tephra de- posits were thinner and destruction of vegetation less complete than in the km sampled across the barren tephra to the south. A sparse cover of scattered shrubs and herbs in 1971 grew by 1977 to a nearly- impenetrable tangle 2—4 m tall (Figs. 4 and 5). Densities of shrub species by year are given in Table 1. The major shrub species found in the shrub area were the same that dominated the undisturbed forest to the north, where four species were about equally dense and dominant. During revegetation, however, these four were present in strikingly different proportions. In 1977, Solanum erianthum and Tournefortia rufo-sericea were twice as nu- merous as Scalesia microcephala. Excavations showed that these two species, especially Tournefortia rufo-sericea, were sending up abun- dant vegetative shoots. Nevertheless, each apparently separate plant that arose without evident aboveground connection to another was counted as an individual. Thus much of the apparent increase in num- bers of Solanum erianthum and Tournefortia rufo-sericea was from a few genets that survived burial under tephra and sprouted vegeta- tively. In Solanum erianthum reproduction by seed also appeared to be vigorous, and it was the only shrub found as occasional pioneer individuals on the deeper tephra. Scalesia microcephala apparently reproduced by seed only, and in- dividuals of all sizes were present in 1973. By 1977, the majority of Scalesia were 2-3 m tall. Zanthoxylum fagara, common in the original forest, was practically nonexistent in the disturbed area. Nowhere were any small Zanthox- ylum seen. It appears that new plants become established with diffi- culty in this habitat but, once established at a later seral stage, persist, as Zanthoxylum was well represented in the original forest. Seed germinability. Seeds collected from iguana droppings in Au- gust, 1977, were tested for germination in December, 1977. In eight of 11 petri plates of seeds, including half of those from the most barren area, seeds germinated after two weeks (22 dicot seeds of at least two species and four grass seeds of two species). None survived long enough for identification. 1981] HENDRIX: PLANT SUCCESSION, GALAPAGOS ISLANDS 249 Fic. 4. (Upper) view north along the northern % km of the rim transect, 1973. Most shrubs are 1 to 2 m tall. Fic. 5. (Lower) view north along the northern 4% km of the rim transect, 1977. Shrubs are a solid cover 2 to 4 m tall. Compare with Figs. 2 and 4, the same view in 1971 and 1973 250 MADRONO [Vol. 28 TABLE 1. SHRUB DENSITY AVERAGED OVER NORTHERN %4 KM OF TRANSECT, AREA OF RAPID REVEGETATION (“SHRUB AREA”). Numbers are individuals per 0.1 ha. 1971 1973 1977 Solanum erianthum 3.8 107 ya Tournefortia rufo-sericea 4.9 52 130 Scalesia microcephala 1.0 41 60 Total shrub density 8.7 200 311 South of barren tephra. Vegetation to the south of the thickest tephra existed in a somewhat different pattern from that at the north- ern edge. The rim topography to the south was more varied, with more ridges and hollows, and corresponding irregularity in tephra depth and other features of the physical environment. Vegetation here grew not in a distinct advancing front as at the north edge, but in irregular patches. Some of these patches were clearly survivors from the pre-eruption forest; notable were a few sparse stands of Opuntia insularis Stewart and several stands of rim shrub forest. In other patches Solanum erianthum dominated. Over much of this area grew a sparse grass community, unlike anything to the north. Here grass cover in 1973 ranged from 10-50 percent composed of three species in the following proportions: 4 Aristida repens :2 Eragrostis mexi- cana: 1 Muhlenbergia microsperma. Also present were occasional in- dividuals of Sonchus oleraceus, Cyperus anderssonii, Verbena town- sendii, Lycopersicon cheesmanii, and Solanum erianthum. DISCUSSION Reproductive strategies. The species found growing on the tephra in the nine years since eruption may be characterized as being to some degree weedy. “Weedy” here denotes possession of some combination of the following characteristics: (1) The plant has the potential for rapid growth in perhaps temporarily favorable conditions and the tol- erance to survive harsh environmental conditions. (2) Reproductive potential by seed is high, or, alternatively, ability to grow by root or stem sprouting is well developed. (3) Ability to compete with other plants is poor. Three reproductive strategies were evident in successful colonizers of the tephra: weedy annuals, rhizomatous perennial herbs, and root- sprouting shrubs. In the harshest sites of deepest tephra, opportunistic annual herbs were the most effective colonizers. The habitats offering any possibility of success for seedlings were scattered, thus those plants producing a large number of easily dispersed seeds had the greatest chance of suc- cessful establishment. Both Sonchus oleraceus and Aristida repens, 1981] HENDRIX: PLANT SUCCESSION, GALAPAGOS ISLANDS 251 found commonly on the tephra, have fruits with hairs or awns that would facilitate their dispersal by wind across the smooth tephra. Some species spread after initial establishment by rhizome growth. Although all the grasses on the more barren tephra are cited as annuals by Reeder and Reeder in Wiggins and Porter (1971), my observations suggest that some may behave as perennials. Eragrostis mexicana in particular grew in stout-based clumps that appeared to be perennial. Several groups were dug up and found to be connected by rhizomes over a distance of a meter or more. This anomalous behavior is another example of the well-documented plasticity of weed species (Baker, 1965; Mayr, 1965). Because barren tephra is so inhospitable to even the most vigorous colonizers, extensive revegetation must await the slow breakup of the surface. Vigorous ramet production can be an important mode of growth and monopolization of resources for weedy perennials (Bunting, 1960; Baker, 1965). Solanum erianthum and Tournefortia rufo-sericea, the two shrub species most successful in early establishment on the tephra, had abundant stem and root sprouting. The apparent predominance of this form of growth in these shrubs helps explain the very rapid revegetation of the edge of the tephra. Shrubs surviving intact and roots or branches shallowly buried may all have sprouted. These shrubs show bursts of increased stem density and size resulting in rapid filling of space. Such a burst made Solanum the dominant species by 1973; by 1977 Tournefortia had attained equal density (Table 1). This rapid growth is characteristic of early successional stages (Odum, 1969) and frequently accounts for rapid revegetation after volcanic eruption (Sands, 1912; Gates, 1914; Aston, 1916; Griggs, 1918). Extensive ramet production maintains existing genotypes and pres- ent fitness. Its importance in this community supports Pickett’s (1976, p. 111) suggestion that “genetic systems favoring reduced recombi- nation are selected for in relatively severe early successional habitats.” Scalesia microcephala showed no evidence of vegetative propaga- tion and was slower to become reestablished. Seedlings were appar- ently able to compete successfully with earlier shrubs and with each other. Continual recruitment of seedlings (Table 1) suggests that Sca- lesia may eventually regain its original density. Rate of revegetation. The rate and pattern of revegetation is in- fluenced not only by differing reproductive and growth strategies but also by rainfall and by land iguanas. The Galapagos are notorious for their year-to-year fluctuations in rainfall. A 12-year record from San Cristobal (Chatham) Island, for example, shows a range in annual rainfall of 3.6—142 cm (Palmer and Pyle, 1966). It is considerably drier on Fernandina than on San Cristobal, but fluctuations are extreme here as well. Although no precipitation data are available for the rim 252 MADRONO [Vol. 28 of Fernandina, Boersma (1977), working at Punta Espinosa on the northeast corner of the island, observed no precipitation during June and August, 197i, but frequent cloud cover and precipitation more than 15 times between June and August, 1972. The weed strategies of the rim flora are those to be expected in a harsh and fluctuating environment. Climatic variation over the coming years will play a major role in determining the rate at which revegetation of the tephra will continue. | Land iguanas, found throughout the area, may facilitate revegeta- tion in two ways. They appear to play a role in seed disperal to and across the tephra, as they and their often seed-filled droppings were found on even the most barren areas. They also may influence soil development. Their droppings consists of plant material in various stages of decomposition and during the wet season may provide an excellent germination site for seeds contained within or blown to the droppings. Several workers have suggested that such addition of or- ganic matter, particularly nitrogen, to volcanic tephra is prerequisite to revegetation (Griggs, 1933; Eggler, 1963). In addition, extensive iguana tunnels help break up the solid surface of the tephra. A number of collapsed burrows were observed in the barren and ecotone areas. Plants were growing in each, usually in marked contrast to the sur- rounding unbroken and barren surface. Vulcanism and plant communities. Persistence of the shrub forest on the rim of Fernandina must be considered in the long-term context of volcanic activity. In the western Galapagos there have been at least 29 eruptions in the last 50 years, 40 eruptions in the last 150 years, and 50 eruptions since the first known eruption 183 years ago in 1797 (T. Simkin, pers. comm., 1981). Periodic disturbances must always have been a feature of Fernandina’s environment. Indeed, if “vulca- nism” is substituted for “competition”, Hutchinson’s (1951, p. 575) definition of fugitive species seems a singularly apt description of Fer- nandina’s shrub forest: “They are forever on the move, always becom- ing extinct in one locality as they succumb to competition and always surviving by reestablishing themselves in some other locality as a new niche opens. The temporary opening of a niche need not involve a full formal successional process.” One trend usually expected in succession is an increase in species number (Odum, 1969). This was not true of the first nine years of rim revegetation on Fernandina. There was, instead, rapid recolonization by most of the same few species found in the original climax forest. Whittaker (1965, p. 257) pointed out that “severe, unstable, and recent environments limit the numbers of species which have evolved to maintain themselves in these environ- ments.” Earlier workers on Fernandina tephra (Colinvaux et al., 1968; Eliasson, 1972) reported species not represented in these samples; these may well become established here in the future. 1981] HENDRIX: PLANT SUCCESSION, GALAPAGOS ISLANDS 253 Revegetation of the Fernandina tephra supports Porter’s (1976, 1979) description of the Galapagos flora as basically weedy. In this arid and periodically disturbed environment, it appears that selection has been for weedy species able to enter new terrain quickly and persist in or at the edges of mature shrub communities. Root and stem sprouting by shrubs is one means of maintaining populations in the face of periodic moderate disturbance. Classic herbaceous weed strat- egies predominate in areas of the most serious disturbance. Thus in the unstable environment of this active volcano, communities are com- posed of variously resilient species able to recover from disturbance and revegetate newly created open habitats. ACKNOWLEDGMENTS I thank J. C. Hickman for counsel and critical comments during writing of the manuscript. W. Forsythe and C. S. Hickman gave helpful comments on an earlier draft. J. C. Hickman, I. L. Wiggins, H. Robinson, and J. R. Reeder helped in iden- tification of difficult species. I thank B. Nolf, W. Eshelman, and T. Simkin for coop- eration enabling the field work, D. Porter for useful information and suggestions, and S. D. Smith for critical reading of the final draft. Contribution No. 243 of the Charles Darwin Foundation. LITERATURE CITED ASTON, B. C. 1916. The vegetation of the Tarawera Mountain, New Zealand. J. Ecol. 4:18-26. BAKER, H. G. 1965. Characteristics and modes of origin of weeds. Jn H. G. Baker and G. L. Stebbins, eds., The genetics of colonizing species, p. 147-172. Academic Press, New York. BOERSMA, P. D. 1976. An ecological and behavioral study of the Galapagos penguin. Living Bird 15:43-93. BROWER, K. 1968. Galapagos: the flow of wilderness. Sierra Club. San Francisco. BUNTING, A. H. 1960. Some reflections on the ecology of weeds. Jn J. L. Harper, ed., The biology of weeds, p. 11-26. Symposia of the British Ecological Society 1, Oxford. COLINVAUX, P. A. 1968. Eruption on Narborough. Animals 11:297—301. , E. K. SCHOFIELD, and I. L. WIGGINS. 1968. Galapagos flora: Fernandina (Narborough) caldera before recent volcanic event. Science 162:1144—-1145. EGGLER, W. A. 1963. Plant life of Paricutin Volcano, Mexico, eight years after activity ceased. Amer. Mid]. Naturalist 69:38—-68. ELIASSON, U. 1972. Studies in Galapagos plants XII: on the vegetation of Fernandina before the eruption in 1968. Bot. Notiser 125:49-61. FRIDRIKSSON, S. 1975. Surtsey. John Wiley & Sons, New York. GATES, F. C. 1914. The pioneer vegetation of Taal Volcano. Philipp. J. Sci. 9(C):391- 434. GRiGcGs, R. F. 1918. The recovery of vegetation at Kodiak. Ohio J. Sci. 19:1—57. . 1933. The colonization of the Katmai ash, a new and inorganic “soil.”” Amer. J. Bot. 20:92-113. HUTCHINSON, G. E. 1951. Copepodology for the ornithologist. Ecology 32:571—577. Mayr, E. 1965. Summary. Jn H. G. Baker and G. L. Stebbins, eds., The genetics of colonizing species, p. 553-562. Academic Press, New York. OpvuM, E. P. 1969. The strategy of ecosystem development. Science 164:262-—270. 254 MADRONO [Vol. 28 PALMER, C. E. and R. L. PYLE. 1966. The climatological setting of the Galapagos. In R. I. Bowman, ed., The Galapagos, p. 93-99. Univ. California Press, Berkeley. PICKETT, S. T. A. 1976. Succession: an evolutionary interpretation. Amer. Naturalist 110:107-119. PorTER, D. M. 1976. Geography and dispersal of Galapagos Islands vascular plants. Nature 264(5588):745—746. . 1979. Endemism and evolution in Galapagos Islands vascular plants. In David Bramwell, ed., Plants and islands, p. 225-256. Academic Press, London. SANDS, W. N. 1912. An account of the return of vegetation and the revival of agri- culture in the area devastated by the Soufriere of St. Vincent in 1902/3. West Indian Bull. 12:22—33. SIMKIN, T. and K. A. Howarp. 1970. Caldera collapse in the Galapagos Islands, 1968. Science 169:429—437. SMATHERS, G. A. and D. MUELLER-DOMBOIS. 1974. Invasion and recovery of vege- tation after a volcanic eruption in Hawaii. National Park Serv. Sci. Monog. 5. VAN DER WERFF, H. 1979. Conservation and vegetation of the Galapagos Islands. In David Bramwell, ed., Plants and islands, p. 391-404. Academic Press, London. WHITTAKER, R. H. 1965. Dominance and diversity in land plant communities. Science 147:250—260. WIGGINS, I. L. and D. M. PorTER. 1971. Flora of the Galapagos Islands. Stanford Univ. Press. Stanford, CA. (Received 7 Jul 1980; accepted 27 Dec 1980; revision received 19 Feb 1981.) A LATE PLEISTOCENE AND HOLOCENE POLLEN RECORD FROM LAGUNA DE LAS TRANCAS, NORTHERN COASTAL SANTA CRUZ COUNTY, CALIFORNIA DaviIp P. ADAM U.S. Geological Survey, Menlo Park, CA 94025 ROGER BYRNE Department of Geography, University of California, Berkeley 94720 EDGAR LUTHER Museum of Paleontology, University of California, Berkeley 94720 ABSTRACT A 2.1-m core from Laguna de las Trancas, a marsh atop a landslide in northern Santa Cruz County, California, has yielded a pollen record for the period between about 30,000 B.P. and roughly 5000 B.P. Three pollen zones are recognized. The earliest is characterized by high frequencies of pine pollen and is correlated with a mid-Wiscon- sinan interstade of the mid-continent. The middle zone contains high frequencies of both pine and fir (Abies, probably A. grandis) pollen and is correlated with the last full glacial interval (upper Wisconsinan). The upper zone is dominated by redwood (Se- quoia) pollen and represents latest Pleistocene to middle Holocene. The past few thou- sand years are not represented in the core. The pollen evidence indicates that during the full glacial period the mean annual temperature at the site was about 2°C to 3°C lower than it is today. We attribute this small difference to the stabilizing effect of marine upwelling on the temperature regime in the immediate vicinity of the coast. Precipitation may have been about 20 percent higher as a result of longer winter wet seasons. INTRODUCTION The Quaternary vegetation history of coastal California is not well understood. Several fossil floras have been published (Chaney and Mason, 1930; Mason, 1934; Warter, 1976; see Johnson, 1977 for a review), but the detailed history of vegetation change is not yet known. In large part, this uncertainty reflects the limitations of the pollen record. Only four pollen diagrams have been published (Heusser, 1960; Adam, 1975); none covers more than the past 8000 years. In this paper, we report on a pollen analysis of a 210-cm core from Laguna de las Trancas in northern Santa Cruz County (Fig. 1) that covers the period 30,000 B.P. to roughly 5000 B.P. Its pollen content indicates marked changes in vegetation. MADRONO, Vol. 28, No. 4, pp. 255-272, 8 December 1981 256 MADRONO [Vol. 28 STUDY SITE The general environmental setting of the study area has been de- scribed by Hecht and Rusmore (1973). Laguna de las Trancas lies in a small depression at the head of a landslide about 7 km southeast of Point Ano Nuevo (Fig. 1). It is situated on a marine terrace (170 m above sea level) 1 km inland from the present coastline. A radio- carbon date of 29,500 + 560 years (USGS-153) on a piece of pine wood from a depth of 312 cm near the base of the marsh deposits indicates that the marsh was formed approximately 30,000 years ago. The land- slide event may have been associated with movement along the nearby Ben Lomond or San Gregorio faults. The bedrock in the immediate vicinity of the marsh is the Santa Cruz Mudstone of Clark (1966, 1970), a siliceous organic marine mudstone of late Miocene and early Pliocene age (Greene, 1977). The topography of the coastal area is rugged, especially to the north, where the coastline intersects the San Gregorio fault, and where steep cliffs rise from the beach to an elevation of 180 m. The marsh itself is located on a narrow interfluve between Waddell and Scott Creeks, two small but perennial streams that rise in the Santa Cruz Mountains about 20 km from the coast. They occupy steep-walled valleys, the mouths of which have been drowned by the postglacial rise in sea level. The area has a Mediterranean-type climate that is characterized by winter rain and summer drought. Mean annual rainfall is about 77 cm, mostly falling between November and April (Rantz, 1971). Tem- perature extremes are rare, and seasonal averages range from 17°C in September to 9°C in January. Coastal fog is common in summer (U.S. Dept. Commerce, 1977). The present vegetation of the area forms a complex mosaic of plant communities. The general distribution of some of the more important taxa is shown in Fig. 1. Locally important along the coast is a shrub community (Type 50, Fig. 1) in which the dominant species locally are coastal sage (Artemisia californica Less.) and coyote bush (Bac- charis pilularis DC.). Farther inland the coastal shrub gives way to a coniferous woodland (Type 9, Fig. 1) in which the local dominant is Monterey pine (Pinus radiata D. Don). This species has a very restricted natural distribution and is found at only three localities along the California coast. Here at Point Ano Nuevo, Pinus radiata is at the northernmost limit of its natural range; its total range covers an area of less than 60 km? (Fowells, 1965, p. 390). There has been considerable discussion as to the causes for the very localized distribution of Monterey pine (Moulds, 1950; McDonald, 1959; Stebbins, 1965), but is is generally agreed that summer drought is an important limiting factor. Evidence of this was apparent on the eastern margins of the Ano Nuevo population during the early fall of 1981] ADAM ET AL.: PLEISTOCENE POLLEN RECORD 257 122°15' 37° Davenport Landing \ 2 Miles i) 2 Kilometers te) 10 Miles 52 ane 0 10 Kilometers Cruz Fic. 1. Map showing the locations of Laguna de las Trancas. Inset map shows the location relative to San Francisco Bay; numbered regions indicate vegetation types, and are taken from the vegetation map of California by Kuchler (1977; reproduced by permission). Vegetation types are: 2, Redwood forest (Pseudotsuga-S equoia); 9, Coastal cypress and pine forests (Cupressus, Pinus); 23, Mixed hardwood forest (Arbutus -Quer- cus); 25, Blue oak-digger pine forest (Pinus-Quercus); 29, Chaparral (Adenostoma-Arc- tostaphylos-Ceanothus); 33, Valley oak savanna (Quercus-Stipa); 36, California prairie (Stipa spp.); 37, Tule marsh (Scirpus-Typha); 38, Coastal saltmarsh (Salicornia-S par- tina); 50, Northern seashore communities (Elymus, Baccharis); and 52, Coastal prairie- scrub mosaic (Baccharis, Dantonia-Festuca). Base for large map is taken from USGS Davenport and Ano Nuevo 7.5-minute quadrangles. 1977. Several trees showed signs of stress in the form of yellow needles and premature needle fall, probably in response to the unusually se- vere drought of the two preceding years. In the main part of the stand, however, there was no evidence of drought stress; furthermore, there is no indication that the Monterey pine is a species that is doomed to an early extinction (cf. Ornduff, 1974). Reproduction is everywhere evident, and no other tree species appears to be better adapted to this particular environment. Occasional Douglas fir (Pseudotsuga menzie- sit (Mirb.) Franco) and live oaks (Quercus agrifolia Nee) are found among the pines, and in some places a mixed hardwood forest has developed (Type 23, Fig. 1); neither appears to have any consistent 258 MADRONO [Vol. 28 competitive advantage. Beneath the pines there is a discontinuous shrub layer that consists largely of California lilac (Ceanothus thyr- siflorus Esch.), California holly (Heteromeles arbutifolia M. Roem.), hazelnut (Corylus cornuta var. californica (A. DC.) Sharp), and poi- son oak (Toxicodendron diversilobum (T. & G.) Greene). Farther inland, floristic composition is largely a function of slope, aspect, and available moisture. On the more mesic sites, redwood (Sequoia sempervirens (D. Don) Endl.) is the dominant species (Type 2, Fig. 1); it is often found in association with madrone (Arbutus menziesit Pursh), tanbark oak (Lithocarpus densiflora (H. & A.) Rehd.), and the California bay (Umbellularia californica (H. & A.) Nutt.). On drier sites, Douglas fir (Pseudotsuga menziesii) and knob- cone pine (Pinus attenuata Lemmon) are locally common, as are sev- eral species of oak (Quercus agrifolia, Q. chrysolepis Liebm., Q. wis- lizenit A. DC.). Natural hybrids between Pinus radiata and P. attenuata have been reported from near Point Ano Nuevo (Fowells, 1965, p. 394). On very dry sites, chaparral species are dominant, including chamise (Adenostoma fasciculatum H. & A.), coyote bush (Baccharis pilularis) and manzanita (Arctostaphylos spp.). Chaparral species are also locally common in abandoned pastures and in areas that have recently been cleared by logging or fire. The two permanent streams that run through the area, Scott and Waddell Creeks, are fringed by a riparian woodland that includes broadleafed maple (Acer macrophyllum Pursh), California buckeye (Aesculus californica (Spach) Nutt.), red alder (Alnus oregona Nutt.), cottonwood (Populus trichocarpa T. & G.), box elder (Acer negundo L. ssp. californicum (T. & G.) Wesmael), willows (Salix spp.), and the California nutmeg (Torreya californica Torr.). Since the beginning of European settlement, the vegetation of the area has been drastically modified by human disturbance. Large areas on the marine terraces and in the valley bottoms have been cleared for agriculture. In other areas the vegetation has been variously af- fected by logging, grazing, and changes in fire frequency. METHODS A core was taken from the central part of the marsh with a 10-cm- diameter piston corer. The marsh normally consists of a floating mat of Typha and Scirpus approximately 75 cm thick. Of this, the upper half is living roots and the lower portion coarse peat. Because the mat as a whole was too loosely consolidated to be successfully cored with our equipment, we began coring at a depth of 105 cm below the water surface. Below this depth the sediments were reasonably compact, and a 210-cm core was recovered in five sections. By the fall of 1977, after 2 years of severe drought, the marsh had dried out completely, and the water table was about a meter below the ground surface. The 1981] ADAM ET AL.: PLEISTOCENE POLLEN RECORD 259 upper part of the sediments was sampled at that time, but no analyses of the top part of the section have been completed. The core shows marked changes in lithology (Fig. 2). The upper 90 cm (105 to 195 cm) consists largely of silt with an increasing proportion of coarse plant debris toward the surface. Between 195 and 210 cm there is a sand layer with some clay lenses. This material is loosely consolidated and was successfully recovered only after several coring attempts. Below the sand layer is almost a meter of brown clay and a basal 10 cm of silt. Bedrock was not encountered, but the sediments below 315 cm were too compact to be recovered with our equipment. The core was split and sampled at 5-cm intervals for pollen analysis. Constant volume samples of 2.5 cm® were taken from the undisturbed central portion of the core, and tablets containing a total of approxi- mately 25,000 Lycopodium spores were added to each sample as a control (Stockmarr, 1971). The extraction procedures followed were basically those described by Faegri and Iverson (1975); in brief, sam- ples were treated with HCl (10 percent), HNO, (10 percent) and ac- etolysis. The pollen-rich residue was then stained with 1 percent saf- ranin and mounted in silicone oil. In general, pollen preservation was good, although at certain levels, particularly the sandy levels, there was a high proportion of broken grains. A least 250 fossil grains were counted at each level. Because of the large numbers of certain pollen types present, a ratio method of count- ing was followed. For most levels, the count was made up to a total of 100 control grains. For some levels, however, the ratio of pollen grains to controls was too large to make this count feasible, in which case only 50 or, more rarely, 25 controls were counted. For very abun- dant types, such as pine, the count was stopped at 100 pine grains; the number of controls was recorded, and the count continued ex- cluding pine. When the count for a level was complete, the pine/con- trol ratio was then used to estimate the pine total. This method has the advantage of allowing for a better representation of minor taxa. POLLEN TYPES In this analysis, 26 pollen or spore types are assigned to known genera, 18 to family or subfamily, and 3 to groups of families. In addition, 15 unknown but distinctive pollen types were observed; none of them, however, accounted for more than one percent of the total count at any level. Perhaps the most frustrating feature of fossil pollen analysis in Cal- ifornia is the problem posed by the Taxaceae, Cupressaceae, and Tax- odiaceae. Their pollen grains are very similar and are often lumped together as TCT pollen (for example, see Helley, Adam, and Burke, 1972). In this study, however, we distinguish between Sequoia and other TCT pollen on the basis of the thicker exine in Sequoia grains. 260 MADRONO DEP EA IN CM LITHOLOGY POLLEN ZONE lO5 PEAT & BLACK SILT REDWOOD ZONE INCREASING SILT BLACK SILT BLACK SILT SAND & CLAY PINE-FIR LOOSE SAND ZONE BROWN SILTY CLAY DISCONTINUITY BROWN SILTY CLAY PINE FOSSILS BROWN SILTY CLAY] BROWN SILT Fic. 2. Generalized lithology of the Laguna de las Trancas core. [Vol. 28 1981] ADAM ET AL.: PLEISTOCENE POLLEN RECORD 261 Our TCT category probably includes Cupressus, Juniperus, and Tor- reya, and possibly Calocedrus, which has been found as macrofossils in deposits at Mountain View in San Mateo County dated at 21,000 to 24,000 years B.P. (Helley et al., 1972). We did not attempt to make any distinctions within the genus P7- nus. Ting (1966) has proposed that statistical analyses of morpholog- ical characters can be used to identify California pine pollen to the species level. Even when using reference material, however, we were unable to distinguish the pollen of several of the pines now growing in the vicinity of the marsh (Pinus radiata, Pinus attenuata, Pinus ponderosa Dougl. ex P. & C. Lawson), and we therefore did not at- tempt to identify fossil material below the generic level. As we indicate later, this taxonomic problem complicates the interpretation of the pollen diagram. Several different sizes of grass pollen were present, and three size classes were arbitrarily established (Gramineae A, grains <25 pum; Gramineae B, 25—40 wm; and Gramineae C, grains >40 um). Because of the wide range of possibilities no attempt is made in this paper to relate them to particular genera. The Compositae are a floristically diverse group in coastal California, and this diversity is reflected in the pollen record. Here we follow tradition (for example, Martin, 1963) in recognizing only four types: “High-spine” Compositae, Liguliflorae, “Low-spine” Compositae or Ambrosia-type, and Artemisia. Typha latifolia L. pollen is distinctive insofar as it retains the tetrad arrangement and is therefore listed as a separate type. Unfortunately, Typha angustifolia L. pollen cannot be distinguished from Spargani- um pollen or from some broken Typha latifolia tetrads, and we there- fore include all Typha-like monads in the Typha/Sparganium category. The curves in the pollen diagram represent changes in percentages rather than “absolute” values. The pollen sum includes all arboreal types but excludes herbs and aquatics. Unknown and indeterminate pollen and spores are included in the diagram under “unknowns.” RESULTS AND DISCUSSION Pollen diagrams derived from small marshes such as Laguna de las Trancas are more difficult to interpret than diagrams from lacustrine or marine environments. Marsh diagrams reflect two kinds of vege- tation change: changes in the upland vegetation and changes in the marsh itself. In order to avoid confusing local and regional effects, we excluded aquatic pollen types from the pollen sum and calculated their values as percentages of the total nonaquatic pollen. In the discussion that follows, we consider the upland record first. The diagram (Fig. 3) can be divided into the three zones shown in Fig. 2: a pine zone (levels 315-235) at the base, a pine-fir zone (levels 225-185), and a 262 MADRONO [Vol. 28 SNOZ GOOMG3Y = ANOZ UIS-3Nid J3NOZ 4Nid HET EET | a POLLEN, IN PERCENT 00 90 0 202 00 0 00m 0 0 0 0 MDM 00 HD & 0 Biri bri Siti (PI BCT Vi vhl teri 3 } x he é Sa3LIWILNID Ni ‘Hid3d 1981] ADAM ET AL.: PLEISTOCENE POLLEN RECORD 263 redwood zone (levels 175-105). We do not have radiocarbon dates for the critical zone boundaries, but we tentatively suggest the following chronology: 1) The pine zone = 30,000—24,000 B.P., corresponding to an interstade of the mid-Wisconsinan; 2) The pine-fir zone = 24,000—12,000 B.P., corresponding to the main glacial advance of the upper Wisconsinan in the mid-continent (full glacial): and 3) The red- wood zone = latest Pleistocene to Mid-Holocene (12,000 to perhaps 5000 B.P.) Because the core does not include the peat mat that at present covers the marsh, the past several thousand years are not represented in the diagram. Level 1 represents a composite surface sample taken to allow comparison of the present pollen accumulation with the fossil record. Basal pine zone. Pine pollen percentages in the lowest meter of the core are persistently high. We do not feel that it is possible to distinguish among the closed cone pines (Pinus radiata, P. attenuata, P. muricata D. Don) on pollen morphology alone; theoretically, there- fore, any combination of these species, and perhaps hybrids between them, could have been present in the area at this time, as could other species such as Pinus ponderosa, P. sabiniana Dougl., and perhaps even P. contorta Dougl. ex Loud. Fortunately, however, we can be reasonably certain that one of the pines present was the knobcone pine (Pinus attenuata). During the coring operation, an incomplete cone was recovered at 280 cm with cone scales that show the minutely spinose tip characteristic of the species (H. Schorn, oral commun., 1975). TCT pollen was encountered at all levels in this zone but never accounted for more than 5 percent of the total. As stated, several genera could be represented, including Juniperus, Cupressus, Calo- cedrus and Torreya. The only other pollen type of importance in this zone is Pseudotsuga. Several pollen types are conspicuously rare or absent, including Sequoia, Quercus, Gramineae, and Compositae. Taken as a whole, the pollen record indicates that during this time period, the upland vegetation in the vicinity of the site was coniferous forest dominated by pine and Douglas fir. The Douglas fir may have been a more important component of the vegetation than the pollen — Fic. 3. Pollen diagram for the major nonaquatic pollen types in the Laguna de las Trancas core. The horizontal scale is the same for all curves, and is for the darkly-shaded curve; the lightly-shaded curve is a 3X exaggeration of the dark curve. Depth is shown in centimeters below the water surface at the time the core was taken. The depth scale does not apply to the top sample plotted in bar-histogram form (level 1). That sample is a composite modern soil-surface sample. Frequencies less than 1 percent in level 1 are represented by a dot. The pollen sum for the modern sample excluded pollen of Plantago, which did not occur in the fossil samples. 264 MADRONO [Vol. 28 diagram indicates because its pollen is large and is commonly under- represented on pollen diagrams (Baker, 1976). Similarly, chaparral species such as Arctostaphylos manzanita Parry and Adenostoma fas- ciculatum may have been present on drier sites, but we have not observed any pollen from these insect-pollinated species in our sam- ples. Detailed paleoecological interpretation of the pine zone is precluded by the taxonomic imprecision concerning the pine and TCT pollen. In view of the limited importance of Sequoia, Quercus, Gramineae, and Compositae, we would tentatively suggest that the climate of the area at that time was cool and dry, possibly analogous to the interior valleys of the Coast Ranges of Oregon and northern California. Another possibility is that there may be no modern analog for the pine zone at Laguna de las Trancas. Pine-fir zone. At the 210-cm level, there is a marked change in pollen frequencies. Pine declines in importance, and fir and ericaceous pollen suddenly increase. The discovery of fir pollen was unexpected. Fir is not now native to the Santa Cruz Mountains, and the nearest natural stand is 100 km to the south in the Santa Lucia Mountains. We do not believe, however, that the fir pollen found at Laguna de las Trancas represents the Santa Lucia fir (Abies bracteata D. Don ex Poiteau). Comparison with modern reference materials suggest that the Laguna de las Trancas fir is more likely to be Abzes grandis, the grand fir. This species has a wide distribution in the Pacific Northwest and ranges down the California coast to the Russian River, a distance of 150 km north of Laguna de las Trancas. In northern California, the grand fir is largely restricted to the coast. It is found in association with redwood below elevations of about 600 meters, and with Bishop pine (Pinus muricata) in the immediate vicinity of the coast (Griffin and Critchfield, 1972). Unfortunately, the ericaceous pollen cannot be identified to species. In view of the similarity between the fir and Ericaceae curves, we suggest that the following are likely possibilities: Labrador tea (Ledum glandulosum Nutt. ssp. columbianum (Piper) C. L. Hitchc.), huckle- berry (Vaccinium spp.) and salal (Gaultheria shallon Pursh). All are locally common along the northern California coast, especially in sandy, low-pH environments. The Ericaceae are primarily insect pol- linated, and whatever species produced the ericaceous pollen at La- guna de las Trancas must have been growing in close proximity to the marsh. In our experience, chaparral ericads such as manzanita (Arc- tostaphylos spp.) are not well recorded in the fossil pollen record. As in the basal pine zone, Douglas fir pollen was encountered at all levels; in one sample, it accounts for 32 percent of the total tree pollen. It can be safely assumed, therefore, that Douglas fir was an important component of the vegetation. Also of interest is the increase in grass 1981] ADAM ET AL.: PLEISTOCENE POLLEN RECORD 265 SANTA CRUZ FORT BRAGG EUREKA 20 Lat 36°59’ N., Long 122°01' W. Lat 39°27’ N., Long 123°48' W. Lat 40°48’ N., Long 124°10’ W Elevation 38 m Elevation 24m Elevation 13 m 20 TEMPERATURE, IN DEGREES CELSIUS PRECIPITATION, IN CENTIMETERS Fic. 4. Comparison of climatic data for Santa Cruz, Fort Bragg, and Eureka, Calif. Data are from U.S. Department of Commerce (1964); values used are climatic normal values, with the exception of the Fort Bragg temperature data, which are for a period of 25 years. Shaded vertical bars are mean monthly precipitation, solid curve displays mean monthly temperatures, and dotted lines show mean annual temperatures. and high-spine Compositae pollen; we take this as clear evidence that the vegetation around the marsh was not completely closed forest. Taken as a whole, the pine-fir assemblage suggests that during the last full glacial interval the vegetation of the Laguna de las Trancas area was very similar to that which is found today about 2.5° farther north, along the northern coast of California. The fir pollen is the most convincing evidence of climatic change, and suggests that the full glacial climate in the vicinity of the marsh was on average at least 2° to 3°C cooler than at present. This estimate is based on a simple comparison between the mean monthly temperature curves for Fort Bragg and Santa Cruz (Fig. 4), and is only a minimum value. Fort Bragg is located directly on the coast, whereas during the last glacia- tion the Laguna de las Trancas site was several kilometers inland, and the moderating effect of the ocean upon the climate may have been less than at Fort Bragg. The lithology of the core also provides evidence of environmental change during the last full glacial interval. The high sand concentra- tion in the pine-fir zone can be interpreted in several ways. If it is fluvial in origin, it could indicate changes in rainfall in the watershed and more effective erosion and transport of sand-sized sediment. It could also be attributed to a lowering of the water level in the marsh 266 MADRONO [Vol. 28 and an increase in the transport of sand into the central area of the marsh. A third possibility is that the sand is aeolian in origin. No sedimentological studies have been carried out on the Laguna de las Trancas core, and we are not able to state definitively how the sand was deposited in the marsh. We suggest, however, that it was blown in. The watershed area of the marsh is very small (<10 ha) and it seems unlikely that, even with a significantly different precipitation regime, there would be a marked increase in the amount of surface runoff and erosion. A more plausible explanation is that during the last full glacial interval, active sand dunes were more extensive along the central California coast than they are today. During the period 30,000 B.P. to 10,000 B.P., the combined Sacramento and San Joa- quin drainage reached sea level west of the Golden Gate. The sand supply to the coast must have been considerably greater then than it is today. In this respect, the central California coast during the last full glacial interval may have been similar to the present day Oregon and Washington coast, where massive dune systems are fed by the Columbia River. At present, there is a small dune field on Ano Nuevo Point, 7 km northwest of the marsh. Conceivably, during the full glacial much of the now-submerged coastal plain was covered by ac- tive dunes, and some sand may have blown up as far as the 170-m terrace. Abies grandis currently grows on coastal dunes on the far northern California coast (Barbour and Johnson, 1977) and probably occupied the same sort of habitat near Laguna de las Trancas. Fir trees must have been growing near the marsh, because fir pollen is large and is not blown long distances. One puzzling aspect of the pine-fir zone is the absence of Sequoia pollen. It seems unlikely that the full glacial climate was severe enough to eliminate redwoods from the Santa Cruz Mountains as a whole. A more plausible explanation is that stronger winds restricted this salt- sensitive species to more sheltered, inland locations. It is generally agreed that circulation of the atmosphere was intensified during the full glacial (see, for example, Wilson and Hendy, 1971; Lamb and Woodroffe, 1970), and average wind speeds were probably higher then along the California coast than they are today, particularly in summer. The absence of redwood pollen in the pine zone may also be a reflec- tion of stronger onshore winds. Spruce (Picea) is absent from the pine-fir zone. Mason (1934) re- ported Sitka spruce (Picea sitchensis (Bong.) Carr.) needles and twigs in his Tomales flora but failed to find cones, and suggested therefore that at that location the species might have been at the southern limit of its Pleistocene range. Tomales Bay is 140 km north of Laguna de las Trancas and the flora has since been dated at 29,050 + 1100 B.P. (Berger and Libby, 1966). Spruce pollen has recently been discovered in early Holocene sediments from Bolinas Lagoon (Byrne and Berg- quist, unpublished data). 1981] ADAM ET AL.: PLEISTOCENE POLLEN RECORD 267 A present, Sitka spruce and grand fir grow together in coastal Or- egon (Fowells, 1965), but spruce does not extend as far south along the coast as fir in California (Griffin and Critchfield, 1976). A similar distributional relation may have existed during the full glacial. Considered as a whole, the pine-fir assemblage indicates a south- ward displacement of species ranges of at least 150 km during the last glaciation. This estimate is significantly less than Warter’s (1976) 320 km estimate based on paleobotanical evidence from La Brea and Car- pinteria. If both grand fir and Sitka spruce were displaced comparable distances along the coast, however, the coastal temperature change was probably not much greater than our estimate. The climatic dif- ference between Fort Bragg and Eureka is rather small (Fig. 4), and the main southern limit of Sitka spruce lies between those two sites (Griffin and Critchfield, 1976). Because Sitka spruce did not reach Laguna de las Trancas during the last full glacial, we suggest that the full-glacial climate at Laguna de las Trancas was milder than the present climate at Eureka. The implication is that average monthly temperatures were depressed by 2° to 3°C along the coast, and that precipitation was about 20 percent higher, with the increase occurring primarily in the form of longer winter rainy seasons (cf. Fig. 4). This estimate is significantly less than the CLIMAP full glacial temperature depression estimate for the California coast of 9° to 11°C (Gates, 1976). If our 2° to 3°C estimate is correct, it suggests that coastal California, unlike most other areas of North America, experienced a full glacial climate that was not very different from that of the present. Presum- ably, then as now, the Pacific Ocean had an important moderating influence. One important reason why the temperature depression may not have been great is that sea-surface temperatures close to the Cal- ifornia coast are strongly influenced by coastal upwelling, and the upwelling water undoubtedly changed temperature much less between interglacial and glacial intervals than did the normal ocean mixed surface layer farther offshore. In the same context, we emphasize that the Laguna de las Trancas pollen diagram records primarily changes in coastal climate and should not be extrapolated inland. The presence of fir at Laguna de las Tran- cas, for example, does not mean that boreal forest species or com- munity distributions migrated southward through California as a whole during the full glacial. In fact, the available evidence shows that they did not. A recently analyzed core from Clear Lake contains very little spruce or fir pollen in the levels of full-glacial age (Adam, 1979). Similarly, spruce pollen is absent from late Pleistocene sections of a core from Osgood Swamp, near Lake Tahoe (Adam, 1967). There is an interesting parallel here with the late-Pleistocene vege- tation history of eastern North America. The discovery of spruce cones and pollen in Pleistocene sediments in Louisiana was formerly thought to be evidence of a southward, en masse migration of the Boreal Forest 268 MADRONO [Vol. 28 (Deevey, 1949). More recently, it has been interpreted as the result of a more localized migration down the Mississippi Valley where cold- air drainage coming from the Laurentide Ice Sheet would have pro- vided climatically favorable conditions (Delcourt and Delcourt, 1975). In a similar but more persistent way, the cool-summer climate of the California coast allows for a southward extension of “northern” species. Redwood zone. Above 180 cm the importance of both pine and fir drops sharply and redwood increases. Also the lithology of the core indicates a shift from the high sand concentrations of the pine-fir zone to an increasing proportion of silt. Redwood clearly dominated the vegetation of the area at this time; pine, Douglas fir, and oak are rare. On the other hand, both grass and Compositae pollen reach consistently high values. We infer that red- wood was dominant on the more mesic sites and that drier sites were open grassland. Chaparral species were probably also present, but unfortunately this vegetation type is not clearly recognizable in the fossil record. The two most significant aspects of the redwood zone are the high redwood percentage and the virtual absence of pine. The redwood rise can be logically explained as simply the response to an amelioration of climate during the early Holocene. If the previous discussion re- garding the absence of redwoods during the full glacial is correct, it follows that a reduction in the strength of onshore winds would allow the redwoods to move out of the more protected locations and expand westward toward the coast. The absence of pine is less easily account- ed for. At present, Monterey pine is the dominant tree in the immediate vicinity of the site, accounting for 80 percent of the modern pollen rain (Level 1 in Fig. 3). During the early and middle Holocene, how- ever, the situation was clearly different. The low pine percentages in the redwood zone are conclusive evidence that pines were not present in the immediate vicinity of the marsh at this time. Pines produce abundant wind-dispersed pollen and are usually overrepresented in pollen diagrams. Axelrod (1967) has hypothesized that the present restricted distri- bution of the closed-cone pine forest is a result of postglacial climatic change. More specifically, he suggested that during the cooler full glacial, the closed-cone pines were widely distributed along the coast, and that as the climate became hotter and drier during the mid-post- glacial period (the xerothermic), they were restricted to the areas of their present disjunct distributions. The Laguna de las Trancas record supports this hypothesis in part, but not entirely. The main problem with the xerothermic hypothesis is that the high redwood percentages in the Holocene argue against persistent drought 1981] ADAM ET AL.: PLEISTOCENE POLLEN RECORD 269 along the coast during this period. By this we do not mean to say that the middle Holocene climate was not different from that of the present, but rather that it was not different enough to account for the disap- pearance of the pines. A more plausible explanation is that the change from glacial to postglacial climate caused the pine decline. With the exception of P7- nus attenuata, all the closed-cone pines are adapted to cool-summer climates. They are found today along the coast, where summer fogs ameliorate the effects of summer drought. During the interstadial con- ditions of the mid-Wisconsinan and the full glacial conditions of the upper Wisconsinan, a cool-summer climate probably was character- istic of most of the California coastline. During the early postglacial, however, summer droughts must have become more severe, and the closed-cone pines would have been restricted to especially favorable sites. At this time sea level was still well below its present level, and the ancestors of the Monterey pines that are now found at Laguna de las Trancas could have been 10 to 20 km to the west. At sea level rose during the Holocene, the pines could have migrated eastward to as- sume their present distribution. In other words, the pine curve in Fig. 3 is probably best explained as being a reflection of changes in climate and related changes in sea level. Aquatic sequence. The aquatic diagram (Fig. 5) shows marked changes in pollen and spore frequencies, which we interpret to be primarily a reflection of hydroseral succession. For purposes of dis- cussion, two stages can be recognized. The first stage coincides with the basal pine zone (levels 315-235) and represents open-water conditions. The pond was probably not very deep at this time but it was deep enough to prevent the estab- lishment of marsh plants such as Scirpus and Typha. The only aquat- ics consistently present were Equisetum and Myriophyllum. The li- thology of the core also indicates that the marsh had not yet formed. Except for a silty layer below 305 cm, the sediments are fine-grained and contain little organic matter. The second stage includes both the pine-fir and redwood zones and is characterized by an expanding area of marsh around the edges of the pond. This is well shown in the diagram by the Typha latifolia and Cyperaceae curves. Cyperaceae in this case almost certainly rep- resents the tules (Scirpus acutus Muhl. and possibly S. validus Vahl.). The irregular increase in Alnus also reflects the process of hydroseral succession. A similar successional trend is evident in the Azolla curve. The species represented here is Azolla filiculoides Lam., the water fern. Azolla is a floating aquatic, but needs shallow water for successful reproduction (Bonnet, 1957). The frequent changes in Azolla percent- ages probably reflect short-term changes in water depth. 270 MADRONO [Vol. 28 a é 2 es §, & ged g¢ & 2 Vw & é es & § & &§ SMM KG s 0 20 0 0 O 200 200 Level 1] 2 _ Ww z 5 2 a ZG = oN : =a 2 16 z 2g 150 A = 8 _ pe - @ -4 = aut & - | a= = Oo HT = — ON = 200- = — = -¢c G - -o ae -¥ rage a ce core : 5 x B 250 = a -Oo a - N ae _ iu - _Z2 a =a () ake) oO 0 0 20 0 0 20 0 0 20 40 60 20 40 60 80 100120140160 20 40 60 8 0 0 0 O POLLEN, IN PERCENT Fic. 5. Diagram for the aquatic types in the Laguna de las Trancas core. Depth and percentage scales are the same as for FIG. 3. The lithology of this section of the core is more complicated than that of the basal zone. As suggested, the sand layer is probably the result of increased dune activity along the coast, in which case it does not reflect any hydrological changes within the marsh itself. Above the sand layer, however, the sediments are predominantly silty and contain an increasing amount of organic material toward the top of the core. This can probably be attributed to the gradual shallowing of the pond and increased extent of the marsh. Today, the peat mat covers all but a small part of the pond. In brief, both the lithology of the core and the aquatic pollen record reflect the progressive filling of the pond and the expansion of the marsh. At the same time, this process of hydroseral succession was complicated by regional changes in climate. We recognize the danger of circular reasoning here, but suggest that at least one of the curves in Fig. 5 shows changes in frequency that are best interpreted as resulting from climatic change rather than hydroseral succession. 1981] ADAM ET AL.: PLEISTOCENE POLLEN RECORD 274 The Jsoetes curve is irregular and shows three major peaks. The peaks in themselves may or may not be meaningful, but it is probably significant that Jsoetes is restricted to the basal pine and pine-fir zones. In the eastern United States /soetes is reported as being important in lake deposits that date to the last full glacial (Frey, 1953). Unfortu- nately, the climatic tolerance of Jsoetes is not yet well understood, and we therefore cannot draw any specific conclusions from its oc- currence at Laguna de las Trancas. LITERATURE CITED ADAM, D. P. 1967. Late-Pleistocene and recent palynology in the central Sierra Ne- vada, California. In E. J. Cushing and H. E. Wright, Jr., eds., Quaternary Pa- leoecology, p. 275-301. Yale Univ. Press, New Haven. . 1975. A late Holocene pollen record from Pearson’s Pond, Weeks Creek Land- slide, San Francisco Peninsula, California. U.S.G.S. J. Res. 3(6):721—731. 1979. Raw pollen counts from core 4, Clear Lake, Lake County, California. U.S.G.S. Open-File Rept. 79-663. AXELROD, D. I. 1967. The Pleistocene Soboba Flora of Southern California. Univ. Calif. Publ. Geol. Sci. 60:1—79. BAKER, R. G. 1976. Late Quaternary vegetation history of the Yellowstone Lake Basin, Wyoming. U.S.G.S. Prof. Paper 729-E:E1—-E48. BARBOUR, M. G. and A. F. JOHNSON. 1977. Beach and dune. Jn M. G. Barbour and J. Major, eds., The terrestrial vegetation of California, p. 223-261. Wiley-Inter- science, New York. BERGER, R. and W. F. LIBBy. 1966. UCLA radiocarbon dates V. Radiocarbon 8:467— 497. BONNET, A. L. M. 1957. Contribution a l’étude des Hydropteridees, III: Recherches sur Azolla filiculoides Lamk. Rev. Cytol. Biol. Veg. 18:1-86. CHANEY, R. W. and H. L. Mason. 1930. A Pleistocene flora from Santa Cruz Island, California. Publ. Carnegie Inst. Wash. 415:1—24. CLARK, J. C. 1966. Tertiary stratigraphy of the Felton-Santa Cruz area, Santa Cruz Mountains, California. Ph.D. dissertation, Stanford Univ., Stanford, CA. . 1970. Geologic map of the southwestern Santa Cruz Mountains between Ano Nuevo Point & Davenport, California. Scale 1:24,000. U.S.G.S. Open-File Map. DEEVEY, E. S. 1949. Biogeography of the Pleistocene: Part 1, Europe and North America. Bull. Geol. Soc. Amer. 60:1315—1416. DELCOURT, H. R. and P. A. DELCOoURT. 1975. The Blufflands: Pleistocene pathway into the Tunica Hills. Amer. Midl. Naturalist 94:385—400. FAEGRI, K. and J. IVERSON. 1975. Textbook of pollen analysis. 3rd rev. ed. Hafner, New York. FOWELLS, H. A. 1965. Silvics of forest trees of the United States. U.S.D.A. Handbook No. 271. FREY, D. G. 1953. Regional aspects of late-glacial and post-glacial pollen succession of southeastern North Carolina. Ecol. Monogr. 23:289-313. GATES, W. L. 1976. Modelling the ice age climate. Science 191:1138—-1144. GREENE, H. G. 1977. Geology of the Monterey Bay region. U.S.G.S. Open-File Rept. 77-718. GRIFFIN, J. R. and W. B. CRITCHFIELD. 1972. The distribution of forest trees in California. U.S.D.A. Forest Serv. Res. Paper PSW-82/1972. HECHT, B. and B. RusmMorE. 1973. Waddell Creek: the environment around Big Basin. Univ. Calif. (Santa Cruz) Environmental Studies Program and Sempervi- rens Fund. HELLEY, E. J., D. P. ADAM, and D. BuRKE. 1972. Late Quaternary stratigraphic and 272 MADRONO [Vol. 28 paleoecological investigations in the San Francisco Bay Area. In V. A. Frizzell, ed., Guidebook for friends of the Pleistocene field trip, October 6-8, 1972. HEUSSER, C. J. 1960. Late Pleistocene environments of North Pacific North America— an elaboration of late-glacial and postglacial climatic, physiographic, and biotic changes. Amer. Geogr. Soc. Special Publ. 35. JOHNSON, D. L. 1977. Quaternary climate of coastal California. Quaternary Research 8:154-179. KUCHLER, A. W. 1977. The map of the natural vegetation of California. Jn M. G. Barbour and J. Major, eds., Terrestrial vegetation of California, p. 909-938. Wi- ley-Interscience, New York. LAMB, H. H. and A. WOODROFFE. 1970. Atmospheric circulation during the last ice age. Quaternary Research 1:29—58. MARTIN, P. S. 1963. The last 10,000 years—a fossil pollen record of the American Southwest. Univ. Arizona Press, Tucson. Mason, H. L. 1934. Pleistocene flora of the Tomales Formation. Publ. Carnegie Inst. Wash. 415:81-179. McDONALD, J. B. 1959. An ecological study of Monterey pine in Monterey County, California. M.A. thesis, Univ. Calif., Berkeley, Forest. Dept. Mou.Lps, F. R. 1950. Ecology and silviculture of Pinus radiata D. Don, in California and in southern Australia. Ph.D. dissertation, Yale Univ., New Haven. ORNDUFF, R. 1974. Introduction to California plant life. Univ. Calif. Press, Berkeley. RANTZ, S. E. 1971. Precipitation depth-duration frequency relations for the San Fran- cisco Bay Region, California. U.S.G.S. and U.S. Dept. Housing Urban Devel., S. F. Bay Region Environm. and Resources Planning Study, Basic Data Contr. No. 25, p. C237-C241. (Text also published as U.S.G.S. Prof. Paper 750-C:C237-— C241.) STEBBINS, G. L. 1965. Discussion of paper by M. H. Bannister. Jn H. G. Baker and G. L. Stebbins, eds., The genetics of colonizing species, p. 373. Academic Press, New York. STOCKMARR, J. 1971. Tablets with spores used in absolute pollen analysis. Pollen et Spores 13:615—631. TING, W. S. 1966. Determination of Pinus species by pollen statistics. Univ. Calif. Publ. Geol. Sci. 58. U.S. DEPT. COMMERCE. 1964. Climatic summary of the United States—Supplement for 1951 through 1960: California. Climatography of the United States No. 86-4. 1977. California 1977. Climatological data 81(13). Nat. Oceanic and Atmo- spheric Admin., Asheville, N.C. WARTER, J. K. 1976. Late Pleistocene plant communities—evidence from the La Brea tar pits. In J. Latting, ed., Plant communities of southern California, p. 32-39. Calif. Native Plant Soc. Special Publ. No. 2. WILsoNn, A. T. and C. H. HENDy. 1971. Past wind strength from isotope studies. Nature 234:344-345. (Received 3 Nov 1980; accepted 11 Feb 1981; revision received 2 Mar 1981.) NOTES AND NEWS VARIATION IN IMMATURE CONE COLOR OF PONDEROSA PINE (PINACEAE) IN NORTH- ERN CALIFORNIA AND SOUTHERN OREGON.—Ponderosa pine (Pinus ponderosa Dougl. ex Laws.) has been the subject of much research, possibly as much or more than any other forest tree in North America. There are now more than 3500 articles that report on some feature or relationship of ponderosa pine; possibly one third of these deal exclusively with the tree (Axelton, USDA For. Ser. Res. Pap. INT-40. 1967; Gen. Tech. Rep. INT-12. 1974; Gen. Tech. Rep. INT-33. 1978). Yet only three of these make note 1981] NOTES AND NEWS ALG of immature cone color. For the cone color of 20 trees in Idaho, Maki (J. For. 38:55— 60. 1940) reported 15 percent light (green to yellow-green), 50 percent dark (purple), and 35 percent medium (between green and purple). Krugman and Jenkinson (USDA Handb. 450. 1974) record vars. arizonica (Engelm.) Shaw and scopulorum Engelm. as green and var. ponderosa as green to yellow-green, rarely purple. Critchfield and Al- lenbaugh (Madrono 18:63—64. 1965) record only green-colored cones from the Sierra Nevada, but I observed trees with purple cones in northern California and southern Oregon. In 1976, in the South Warner Mountains of Modoc County, California, I noticed a decided increase in the frequency of purple cones with increased elevation. A good crop on ponderosa pine in northern California and southern Oregon in 1978 provided the opportunity to study immature cone color more carefully. This was un- dertaken: (1) to test for correlations of cone color with other characteristics of ponderosa that vary in this region; (2) to determine whether cone color could be a marker for some economic criterion; (3) to understand better the relationship of ponderosa to other pines. In 1978, cone color was recorded in mid-August for 33 locations (Fig. 1). Roadside positions that gave a view of a number of cone-bearing trees with front-lighting by the sun were selected. The crowns were scanned with binoculars and the cone color of each tree was classified into three categories. The color classification and nomenclature sys- tem of Kornerup and Wanscher was used (Politiken Forlag, Copenhagen. 1974) as follows—light (green to yellow-green, 28 and 29 A&B 6-8), medium (red to pale red, 8 and 9 B&C 6-8), dark (purple, 11 and 12 E&F 7-8). All cones on a tree were of one color class. Cones attacked by insects or that will abort are more nearly true pink and lack green pigment. Such cones are smaller and scattered throughout a tree among cones falling into one of the three basic classes. Any tree with cones not distinctly purple or green to yellow-green was placed in the intermediate category. A series of viewing sites was used until 100 cone-bearing trees had been classified. This usually was accomplished in less than 1.5 km. At three sites only 50—75 trees could be viewed and classified easily because of the scattered location of cone-bearing trees. Each series was considered a plot. Though ponderosa pines have irregular cone crops, most trees in many locations had maturing cones in 1978; and it was assumed that the cone-bearing trees were representative of the area. Each plot was placed into one of five groups on the basis of the frequency of trees in each of the three classes of cone color (Fig. 1). Virtually all trees in Group 1 plots had trees with only light-colored cones; less than 50 percent of the trees in Group 5 plots had light colored cones. Groups 2, 3, and 4 were gradations between 1 and 5S. The frequency of trees with medium and dark cones increased from south to northeast in the study area (Fig. 1). Group 1 and Group 2 plots were nearly all west of the Sierra Nevada-Cascade crest; Group 3, 4, and 5 plots were found only east of the crest. Elevation may be associated with the shift in cone color in parts of the region. This possibility is shown best by the three plots in the South Warner Mts. at the headwaters of the Pit River (Fig. 1:A). The Group 3 plot there was at about 1800 m; the Group 4 plot at about 2100 m; and a Group 5 plot at about 2400 m. The next three plots along the Pit River (Fig. 1:B) reinforce the relationship of cone color to elevation. A Group 4 plot was at about 1500 m; two Group 3 plots to either side were between 1200 and 1500 m. Ponderosa pine is not abundant above 1800 m in this region, but I hypothesize that higher elevation stands generally will have higher frequencies of darker cones. An elevational shift from green to purple was found by Sturgeon and Mitton (Amer. J. Bot. 67:1040—1045. 1980) for immature cones of white fir (Abies concolor [G. and G.] Lindl.) in southern Colorado. They reported only two colors, green and purple, and did not report geographical variation. They conclude that purple pigment functions in thermoregulation at high elevations. Xylem monoterpene composition is associated with shift in cone color in that it changes abruptly in the general area of the study; 3-carene increased significantly while B-pinene, limonene, and a-pinene decreased significantly from south to north and east in northern California and southern Oregon (Smith, USDA Tech. Bull. 1532. 1977). Sturgeon (Evolution 33:803—814. 1979), in an intensive study in the area, found the shift 274 MADRONO [Vol. 28 125° 120° Legend Group ' code Cone color class _ Light Medium Dark percent <1 0 >1<5 0 >10<20 >1<3 40° >50 >20<40 >3<10 <50 >30<50 >10<60 ie) 50mi Scqle eee / 100km Sierra Nevada- Cascade Crest — 2 Fic. 1. Stand classification of ponderosa pine in northern California and southern Oregon for immature cone color. A. Three sites near headwaters. B. Three sites along the Pit River. C. Localized epidemic of mountain pine beetle in 1957-1959. to occur generally at the Cascade Crest. Because data suggest that changes in resin composition are associated with cone color, three plots at the headwaters of the Pit River (Fig. 1:A) were analyzed for both characteristics simultaneously. The average monoterpene composition was essentially the same as that reported earlier (Smith, Ma- drono 21:26—32. 1971) but there was no consistent association with cone color. Trees with light-colored cones at the lowest elevation had the same approximate composition as trees with dark colored cones at the highest elevation and conversely. Resin com- position and cone color appeared to be independent. The susceptibility of ponderosa pine to the attack of the mountain pine beetle (Den- droctonus ponderosae Hopk.) appears to shift in this same general region of California and Oregon. This insect historically has had a greater tendency to be epidemic in ponderosa pine east of the crest line (Miller, USDA For. Ser. File Rep. 1920; Eaton, J. For. 39:710—713. 1941). There have been no epidemics in ponderosa pine west of the crest line. This association probably does not imply causality because many variables other than monoterpene composition can influence the severity of bark beetle epidemics. Speculations on the evolutionary relationships of ponderosa, particularly with Jeffrey pine (P. jeffreyi Grev. and Balf.) and Washoe pine (P. washoensis Mason and Stock- well), have been discussed by Haller (Madrono 16:126—132. 1961) and Wang (USDA For. Serv. Res. Pap. WO-24. 1977). Haller (1961) first concluded that Washoe pine arose through hybridization between Jeffrey pine and the Rocky Mountain variety of ponderosa pine (var. scopulorum). Later (Haller, Proc. Amer. Bot. Soc. 52:646. 1965), he suggested that Washoe pine arose from hybridization of var. ponderosa with var. scopulorum. 1981] NOTES AND NEWS IRs Both dark cone color and high proportions of 3-carene are characteristic of Washoe pine (Smith, 1971). I therefore conclude tentatively that at one time Washoe pine oc- cupied many higher-elevation sites throughout northeastern California and southern Oregon. It has been replaced slowly by ponderosa pine. The process of hybridization between the two has produced ponderosa pines with dark- and medium-colored cones and with high 3-carene resin. Color of immature cones may have more than one cause. It is possible that the hybridization of ponderosa and Washoe pine may have been the cause in northeastern California while having been selected for independently in southern Oregon. Sturgeon and Mitton’s similar cone-color results with Abzes (1980), however, suggest that dark- colored cones are somehow adaptive at higher elevations and may be selected for in- dependently in different lineages. Haller’s (1965) views of the ancient range and origin of Washoe pine are nevertheless consistent with the hybridization hypothesis.—RICH- ARD H. SMITH, Pacific Southwest Forest and Range Experiment Station, P.O. Box 245, Berkeley, CA 94701. (Received 17 Mar 1980; accepted 12 Jan 1981; final revision received 17 Feb 1981.) REVIEWERS OF MANUSCRIPTS Many people have contributed to the preparation of volume 28, in particular James C. Hickman, who served as associate editor during my initiation, and the editorial board members. Dr. Sterling Keeley, my colleague at the Natural History Museum in Los Angeles, has been particularly helpful. Reviewers of manuscripts are essential in maintaining the quality of any scientific journals, and the thoughtful, constructive crit- icisms of those listed below is gratefully acknowledged. Christiane Anderson Roger del Moral Frank Lang Loran C. Anderson Lauramay T. Dempster Merry G. Lepper Mary E. Barkworth Arthur C. Gibson Deborah Mangis Rupert C. Barneby James R. Griffin Elizabeth McClintock Spencer Barrett Edward O. Guerrant Reid Moran George L. Batchelder Sherry L. Gulmon Rodney G. Myatt Dennis Breedlove Ronald L. Hartman R. W. Pearcy Roger Byrne Harold F. Heady Duncan M. Porter Kenton Chambers Lawrence R. Heckard Robert Robichaux Tsan L. Chuang James Henrickson Carla R. Scheidlinger Curtis Clark Alice Q. Howard Leila Schultz Susan G. Conard Daniel W. Inouye John L. Strother Lincoln Constance Ann Johnson Robert F. Thorne William B. Critchfield Jon E. Keeley Richard J. Vogl Arthur Cronquist Sterling Keeley Robert D. Wright Robert W. Cruden 276 MADRONO [Vol. 28 EDITOR’S REPORT FOR VOLUME 28 Between 1 Jul 1980 and 30 Jun 1981, 66 manuscripts were received. Sixty-one percent of the submissions were articles; 23 percent were noteworthy collections; and 16 percent were notes and news. Their current statuses are as follows: in review (19 articles, 3 noteworthy collections, 4 notes and news); being revised by authors (6, 3, 1); accepted and awaiting publication (9, 4, 2); published in volume 28 (10, 11, 3). In addition, 17 articles, 4 noteworthy collections, and 6 notes received before 1 Jul 1980 were published in volume 28. Of the material appearing in volume 28, 54 percent represented articles, 33 percent noteworthy collections, and 12 percent notes and news. Only two papers were rejected. | A smooth transition between editors was possible because of the carefully organized procedures of the former editor, Dr. James C. Hickman. Madrono conventions are somewhat complicated, and because of Jim’s meticulous attention to detail and his typed guidelines, the assumption of these duties was a pleasure rather than a task. On behalf of the Society, I want to express here my deep appreciation to Jim for this help. Manuscript submissions have been slightly behind those of 1980, but time from sub- mission to publication is now about one year. This appears to be a fortuitous compro- mise between the need for quick publication and the editor’s need for a slight backlog. Most reviewers have been inordinately prompt in returning manuscripts, and authors have been very good about returning their revisions. Thus, there is a slight decrease over the 1980 season in the number of manuscripts out for revision. However my six- week absence this spring allowed a manuscript build-up so that the number of manu- scripts now in review is larger. Noteworthy collections continue to be popular items to submit, but increasing press costs are forcing a change in the format for the appearance of this sort of data. Speci- fications will appear in a future issue of Madrono. As was the case with Jim, I shall continue to welcome suggestions and criticisms from members. C.D. 8 Jul 1981 Dates of Publication of MADRONO, volume 28 No. 1, pp. 1-48: 12 February 1981 No. 2, pp. 49-100: 3 June 1981 No. 3, pp. 101-192: 31 July 1981 No. 4, pp. 193-279: 8 December 1981 1981] INDEX Di INDEX TO VOLUME 28 Classified entries: major subjects, key words, and results; botanical names and plant families (new names are in boldface); geographical areas; reviews. Incidental references to taxa (including lists and tables) are not indexed separately. Species appearing in Noteworthy Collections appear under plant family and state or country. Authors and articles are listed alphabetically in the Table of Contents. Acacia kelloggiana, new species from Baja California, 220. Adventitious rooting in Coastal Sage Scrub dominants, 96. Aizoaceae: Malephora crocea naturalized in CA, 80. Angelica californica, reestablishment of the species, 226. Arizona, bibliography of local floras, 193. New record. Orobanche uniflora subsp. occidentalis, 37. Asteraceae: Erigeron, five new species from Mexico, 136; Erigeron compactus var. consimilis, new to NM, 41; E. hu- milis, new to ID, 88; Hazardia orcuttii, new to USA & CA, 38; Hymenopappus filifolius var. idahoensis, not rare (ID), 88; Madia subspicata, range extension (CA), 39; Rafinesquia californica, range extension (CA), 39; Tetradymia spino- sa, range extension (NM), 42. Baja California, new species of Acacia, 220; Boraginaceae: Cryptantha subcapitata, new species from WY, 159; C. muricata var. muricata, range extension (CA), 39: Botanical history: Albert M. Vollmer, S25 Brassicaceae: Chorispora tenella, new to NM, 42; Cochlearia officinalis, new to CA, 86; Diplotaxis muralis, new to NM, 42; Malcolmia africana, new to NM, 42; Streptanthus farnsworthianus, range extension (CA), 184; Thelypo- diopsis purpusii, recollected (NM), 185. Cactaceae: Sclerocactus mesae-verdae, range extension (NM), 42. California: New records: Apera spica-venti, 40; Cochlearia officinalis, 86; Hazardia orcuttii, 38; Wolffia punctata, 37. New species: Eriogonum libertini, 163; Quercus cornelius-mulleri, 210. Range extensions: Argemone munita subsp. rotundata, 40; Carex tumuli- cola, 40; Cryptantha muricata var. muricata, 39; Dedeckera eurekensis, 86; Epilobium minutum, 40; Lupinus citrinus, 184; Madia subspicata, 39; Mimulus gracilipes, 41; Rafinesquia californica, 39; Rhamnus rubra subsp. yosemitana, 40; Streptanthus farn- sworthianus, 184; Wolffia columbia- na, 187. Taxa rediscovered: Caliptridium pul- chellum, 188. Calypso bulbosa, pollination biology of, 101. Campanulaceae: Nemacladus glandulifer- us var. orientalis in NM, 186. Caryophyllaceae: Stellaria nitens, new to NM, 87. Chaparral: plants lacking extrafloral nec- taries, 26. Chihuahuan Desert: new species of Chio- cocca, 30; new species of Comarosta- phylis, 33; new species of Portulaca, 78; new species of Salix, 148. Chiococca henricksonii, new species from Chihuahuan Desert, 30. Coastal Sage Scrub: Adventitious rooting in some dominant plants, 96. Colorado Desert: Postfire recovery of Creosote Bush Scrub, 61. Comarostaphylis polifolia, new species from Coahuila, 33. Community ecology: age structure of trees in Dana Meadows, Yosemite Natl. Park, 45; coastal strand and dune vege- tation, 49; creosote bush scrub, postfire recovery, 61; distribution of plants with extrafloral nectaries, 26; ecology of Quercus douglasii, 1; elevational distri- bution of pines in the Sierra Nevada, 67; fire ecology, Sierra Nevada foothills, 111; Isoetes in vernal pools, 167; Stipa pulchra in native grasslands, 172; succession after volcanic eruption, 242. Convolvulaceae: [pomoea egregia, new to NM, 87. Cowania mexicana var. stansburiana, hy- bridization with Purshia, 13. 278 MADRONO Crassulacean acid metabolism, 167. Creosote Bush Scrub, postfire recovery, 6 Cruciferae—see Brassicaceae. Cryptantha subcapitata, new species from WY, 159. Cyperaceae: Carex deweyana subsp. dew- eyana, new to Mexico, 186; C. rupes- tris, new to ID, 89; C. tumulicola, range extension (CA), 40; Carex whit- neyi, not endangered, 190. Dispersal: Seeds of Pinus albicaulis by vertebrates, 91; Prunus illicifolia, 94. Editor’s report for volume 28, 276. Endangered, threatened, or rare plants: 41, 42, 86, 88, 89, 188, 190. Ericaceae: Comarostaphylis polifolia, new species from Coahuila, 33. Erigeron, new species from Mexico, 136. Eriogonum libertini, new species from California, 163. Extrafloral nectaries, 26. Fabaceae: Acacia kelloggii, new species from Baja California, 220; Astragalus amnis-amissi, not endangered, 89; A. monumentalis, new to NM, 43; Lupi- nus citrinus, range extension (CA), 184; Trifolium barnebyi, stat. nov., 188. Fagaceae: Ecology of Quercus douglasit, 1; Quercus cornelius-mulleri, new species from CA, 210. Fire ecology: Creosote Bush Scrub, 61; Sequoia Natl. Park, foothill communi- ties, 111. Galapagos Islands, plant succession, 242. Gentianaceae: Gentiana propinqua, new to ID, 89. Gramineae—see Poaceae. Grassland communities: species lacking extrafloral nectaries, 26; species com- position in San Joaquin Valley, 231; Stipa pulchra in, 172. Hybridization, between Cowania and Purshia, 13. Hydrophyllaceae: Phacelia sect. Miltit- oh, 12 1. Idaho: New records: Carex rupestris, 89; Erig- eron humilis, 88; Gentiana propin- qua, 89; Papaver kluanensis, 90. [Vol. 28 Isoetaceae: Isoetes, physiology in vernal pools, 167. Juglandaceae: Juglans pollen from late Holocene CA, 44. Lamiaceae: Salvia microphylla var. wis- lizenit, new to NM, 43; Trichostema pollination, 44. Leguminosae—see Fabaceae. Lemnaceae: Wolffia columbiana, range extension (CA), 187; W. punctata, new to CA, 37. Malephora crocea, naturalized in CA, 80. Mexico: New records: Carex deweyana subsp. deweyana, 186. New species: Acacia kelloggii, 220; Chiococca henricksonii, 30; Com- arostaphylis polifolia, 33; Eriger- on, 136; Salix, 148. Nectaries: extrafloral nectaries, 26; nectar sugar in Trichostema, 43. New Mexico: New records: Astragalus monumentalis, 43; Bromus diandrus, 43; Cercocar- pus intricatus, 43; Chorispora tenel- la, 42; Diplotaxis muralis, 42; Erig- eron compactus var. consimilis, 41; Ipomoea egregia, 87; Malcolmia af- ricana, 42; Ranunculus testiculatus, 43; Salvia microphylla, 43; Stellaria nitens, 87. Range extensions: Sclerocactus mesae- verdae, 42; Tetradymia spinosa, 42. Rediscoveries: Nemacladus glandulifer- us var. orientalis, 186; Thelypodiop- sis purpusit, 185. Oak Woodland, plants lacking extrafloral nectaries, 26. Onagraceae: Epilobium minutum, range extension (CA), 40. Orobanchaceae: Orobanche uniflora subsp. occidentalis, new to AZ, 37. Orchidaceae: Calypso bulbosa, pollination of, 101. Paleobotany: Juglans pollen in late Ho- locene CA, 44; late Pleistocene and Ho- locene pollen record from Santa Cruz Co:, ‘CA, 255. Palynology—see under Paleobotany. 1981] Papaveraceae: Argemone munita subsp. rotundata, range extension (CA), 40; Papaver kluanensis, new to Pacific NW and ID, 90. Phacelia sect. Miltitzia, 121. Pinaceae: age structure of Pinus contorta in Dana Meadows, Yosemite Natl. Park, 45; elevational distribution of pines in the Sierra Nevada, 67; seed dispersal of Pinus albicaulis, 91; cone color variation of Pinus ponderosa, Dee Poaceae: Apera spica-venti, new to CA, 40; Bromus diandrus, new to NM, 43; Stipa pulchra, ecology of, 172. Pollination: Calypso bulbosa, 101; Trich- ostema, 43. Polygonaceae: Dedeckera eurekensis, range extension, 86; Eriogonum libertini, new species from CA, 163. Portulacaceae: Calyptridium pulchellum rediscovered (CA), 188; Portulaca johnstonii, new species from Mexico, 78. Prunus illicifolia, seed dispersal, 94. Purshia glandulosa, hybridization with Cowania, 13. Quercus: Q. douglasii, ecology, 1; Q. cor- nelius-mulleri, new species from CA, 210. Ranunculaceae: Ranunculus testiculatus, new to NM, 43. Rare species—see Endangered Species. Reviews: J. P. Smith, Jr., R. Jane Cole, and J. O. Sawyer, Inventory of rare and endangered vascular plants of Califor- nia, 97; T. Duncan, A taxonomic study of the Ranunculus hispidus Michaux complex in the Western Hemisphere, 99; F. Pursh, Flora Americae Septen- trionalis (ed. J. Ewan), 180. INDEX 279 Rhamnaceae: Rhamnus rubra subsp. yo- semitana, range extension (CA), 40. Riparian Forest: plants lacking extrafloral nectaries, 26. Rosaceae: Cercocarpus intricatus, new to NM, 43; Cowania-Purshia hybridiza- tion, 13; Prunus illicifolia, seed dis- persal, 94. Rubiaceae: Chiococca henricksenii, new species from Mexico, 30. Salicaceae: Salix, new species from Chi- huahuan Desert, 148. Sand Dunes: coastal vegetation, 49. Scrophulariaceae: Mimulus gracilipes, range extension (CA), 41; Pedicularis crenulata rediscovered (CA), 86. Sequoia National Park: ecology of Quer- cus douglasii, 1; fire ecology, 111. Sierra Nevada: ecology of Quercus doug- lasii, 1; elevational distribution of pines, 67; foothill fire ecology, 111. Stipa pulchra, ecology in native grass- lands, 172. Succession following volcanic eruption, 242. Threatened Species—see Endangered Species. Trichostema pollination, 43. Trifolium barnebyi, stat. nov., 188. Umbelliferae—see Apiaceae. Vernal pool ecology, 167. Vollmer, Albert M., 132. Wyoming: New species: Cryptantha subcapitata, 159. Yosemite National Park, 45. . a oa SUBSCRIPTIONS—MEMBERSHIP Membership in the California Botanical Society is open to individuals ($18 per year; students $10 per year for a maximum of seven years). Members of the Society receive MADRONO free. Family memberships ($20) include one eight-page publishing allotment and one journal. Emeritus rates are available from the Corresponding Secretary. In- stitutional subscriptions to MADRONO are available ($25). Membership is based on a calendar year only. Applications for membership (including dues), orders for sub- scriptions, requests and rates for back issues, changes of address, and undelivered copies of MADRONO should be sent to the Corresponding Secretary. Renewal payments should be sent to the Treasurer. INFORMATION FOR CONTRIBUTORS Manuscripts submitted for publication in MADRONO should be sent to the editor. All authors must be members, and membership is prerequisite for review. Manuscripts and review copies of illustrations must be submitted in triplicate for all articles and short items intended for NOTES AND NEWS. Follow the format used in recent issues for the type of item submitted and allow ample margins all around. All manuscripts MUST BE DOUBLE SPACED THROUGHOUT. For articles this in- cludes title (all caps, centered), author names (all caps, centered), addresses (caps and lower case, centered), abstract, 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, and number each page. Do not use a separate cover page, “erasable” paper, or footnotes. Table captions should include all information relevant to tables. All measurements should be in metric units. Line copy illustrations should be clean and legible, proportioned (including captions) to the MADRONO page, and designed for reduction to % original size. 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 latitude and longitude references. Halftone copy should be designed for reproduction at actual size. In no case should original illustrations be sent prior to the acceptance of a manuscript. When needed they should be mounted on stiff cardboard and sent flat. No illustrations larger than 22 xX 28 cm will be accepted. Presentation of nomenclatural matter (accepted names, synonyms, typification) should follow the format used for Rhus integrifolia in MADRONO 22:288. 1974. Institutional abbreviations in specimen citations should follow Holmgren and Keuken, Index Her- bariorum, 6th edition. Abbreviations of serial titles should be those in Botanico-Per- iodicum-Huntianum (Lawrence et al., 1968, Hunt Botanical Library, Pittsburgh). If the correct abbreviation cannot be determined, the full serial title should be used. Titles of books should be given in full, together with the place and date of publication, publisher, and edition, if other than the first. All members of the California Botanical Society are allotted eight pages in MADRONO over a two year period. Joint authors are each assessed the full page number. Beyond that number of pages a required editorial fee of $40.00 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 sched- ule, with equity among all members for access to its pages. Printer’s fees for illustrations and typographically difficult material @ $35.00 per page (if their sum exceeds 30 percent of the paper) and for author’s changes after typesetting @ $3.00 per line will be charged to authors. CALIFORNIA BOTANICAL SOCIETY MADRONO A WEST AMERICAN JOURNAL OF BOTANY VOLUME XXIx 1982 BOARD OF EDITORS Class of: 1982—-DEAN W. TAYLOR, University of California, Davis RICHARD VOGL, California State University, Los Angeles 1983—-ROBERT W. CRUDEN, University of Iowa, Iowa City DUNCAN M. PorRTER, Virginia Polytechnic Institute and State University, Blacksburg 1984—Mary E. BARKWORTH, Utah State University, Logan Harry D. THIERS, San Francisco State University, San Francisco 1985—STERLING C. KEELEY, Whittier College, Whittier, CA ARTHUR C. GIBSON, University of California, Los Angeles 1986—AMyY JEAN GILMARTIN, Washington State University, Pullman ROBERT A. SCHLISING, California State University, Chico Editor——CHRISTOPHER DAVIDSON Idaho Botanical Garden, P.O. Box 2140, Boise, ID 83701 Published quarterly by the California Botanical Society, Inc. Life Sciences Building, University of California, Berkeley 94720 Printed by Allen Press, Inc., Lawrence, KS 66044 TABLE OF CONTENTS ALMEDA, FRANK, Symplocos sousae, a new species of Symplocaceae from Mexico AXELROD, DANIEL I., Age and origin of the Monterey endemic area BENEDICT, NATHAN B., Mountain meadows: stability and change ______________ BENEDICT, NATHAN B. and JACK Major, A physiographic classification of sub- alpine meadows of the Sierra Nevada, California ___________________________- BOTTI, STEPHEN J. Noteworthy collection of Eriophyllum nubigenum Greene ex (ray -(ASteraceae) 4.25229 ee ee ee BOWERS, JANICE E. and STEVEN P. MCLAUGHLIN, Noteworthy collections of Hypoxis mexicana and Linum subteres _____________________--_---------------- BOWERS, JANICE E. and STEVEN P. MCLAUGHLIN, Plant species diversity in PRIZON Gye sa. 2 Woes Doe I ea eee 2 pete aes oe ce BoyvD, STEVE (see Hilu, Khidir W.) BRYANT, GENEVIEVE (see Kelley, Walt) BuRK, JACK H., Phenology, germination, and survival of desert ephemerals in Deep Canyon, Riverside County, California __________________________________ BuRKE, MAryY T., The vegetation of the Rae Lakes Basin, southern Sierra Ne- NN cam a Ie COLLINS, L. T. (see Heckard, L. R.) CONARD, S. G. and S. R. RADOSEVICH, Post-fire succession in white fir (Abies concolor) vegetation of the northern Sierra Nevada CONSTANCE, LINCOLN (see Meinke, Robert J.) Cook, STANTON A., Unilateral infertility in parapatric species of Eschscholzia (Papaveraceae): selection for isolation? _______________________________________--_ CopE, EDWARD (see Sawyer, John) Cory, JACULYN K. (see Lackschewitz, Klaus H.) Davis, W. S., Notes on the distribution of Malacothrix on the California islands DEDECKER, MARY, Noteworthy collections of Cymopterus ripleyi, Cryptantha scoparia, Astragalus argophyllus, and Eriogonum puberulum _______--_--__-- ECKENWALDER, JAMES E., Populus X inopina Hybr. nov. (Salicaceae), a natural hybrid between the native North American P. fremontii and the introduced WurasianeP 27100) 2 ee eee ELVANDER, PATRICK E., Gynodioecy in Saxifraga integrifolia (Saxifragaceae) EVERT, ERWIN F., Noteworthy collections of Carex bipartita, Carex deweyana, Carex incurviformis, Dianthus barbatus, Gentianella propinqua, Myosotis arvensis, Myosotis micrantha, Potentilla recta __________-__------------------- FELKER, PETER (see Hilu, Khidir W.) FORCELLA, F. and S. J. HARVEY, Spread of Filago arvensis L. (Compositae) in the United States 2.2: -3.22 3 2) ee GILMARTIN, AMY JEAN, Effects on Lomatium triternatum of the 1980 ash fallout from INit. St. .Elelens, 2: oe eee es ee ne ee GOODRICH, SHEREL and Mont E. LEwis, Noteworthy collections of Carex mi- croglochin, Carex parryana, Epilobium nevadense, and Kebresia simplicius- CUNO S iiss are ea en es se pte A go GRIFFIN, JAMES R., Pine seedlings, native ground cover, and Lolium multiflorum on the Marble-Cone burn, Santa Lucia Range, California ________-____-___- HARVEY, S. J. (see Forcella, F.) HAVLIK, NEIL, Noteworthy collection of Mirabilis laevis __-.-.-._-------------__-- HECKARD, L. R. and L. T. COLLINs, Taxonomy and distribution of Orobanche valida (Orobanchaceae)’ —o. -2- 28 2 ee eee HENRICKSON, JAMES, On the recognition of Trichostema mexicanum Epling (La- IMIACCAC) wer ans se ta ene Pl eh i 154 164 42 32 218 PAT 67 269 124 119 270 60 HILvu, KHIpIR W., STEVE BoyD, and PETER FELKER, Morphological diversity and taxonomy of California mesquites (Prosopis, Leguminosae) -_-_-_-----__- JERNSTEDT, JUDITH A., Floral variation in Chlorogalum angustifolium (Liliaceae) JOHNSON, DALE E., Climate diagram for the University of California Sagehen Creek iteld Stauom 22222 -2 =e ee Soe ee eo A ee eee KELLEY, WALT, GENEVIEVE BRYANT, and DIETER WILKEN, Noteworthy col- lections of Evistrum diffusum and Crypsts alopecuroides ---------------------- KRUCKEBERG, ARTHUR LEO, Noteworthy collection of Polystichum kruckebergii KuHyYoS, DONALD W. and PETER H. RAVEN, Miscellaneous chromosome numbers IM PA SUCTACCAC 2222.54. 6 ee 6 eh ee LACKSCHEWITZ, KLAUS H., PETER LESICA, ROGER ROSENTRETER, and PETER F. STICKNEY, Noteworthy collections of Antennaria monocephala, Gentiana tenella, Juncus triglumis, Koenigia islandica, Lomatium bicolor, Musineon vaginatum, Phacelia thermalis, Plantago hirtella, Ribes triste, Rorippa syl- vestris, Satureja douglasii, Saussurea densa, Veronica verna ______-_-_----_- LANG, FRANK A. and VEVA STANSELL, Noteworthy collection of Asplenium tri- chomanes 1L. (ASpleniaceae)'_ 2... ns ee LANGFORD, GAYLE (see Turner, B. L.) LESICA, PETER (see Lackschewitz, Klaus H.) LEwis, Mont E. (see Goodrich, Sherel) Major, JACK (see Benedict, Nathan B.) McCunegE, BRUCE, Noteworthy collection of Howellia aquatilis (Campanulaceae) McLAUGHLIN, STEVEN P. (see Bowers, Janice E., both entries) MCcNEAL, DALE W., JR. and MARION OWNBEY, Taxonomy of the Allium lacu- HOSUMm-COMplex, (illacede) 229k. ae ee ed eee eae ee MEINKE, ROBERT J. and LINCOLN CONSTANCE, Lomatium oreganum and L. greenmanit (Umbelliferae), two little known alpine endemics from north- CASCCEMUO ee O Te eee eee eget Psa ee ole ees tetas ole ee ee MINNICH, RICHARD A., Pseudotsuga macrocarpa in Baja California? ___________- NELSON, JANE P. (see Nelson, Thomas W.) NELSON THOMAS W. and JANE P. NELSON, Noteworthy collection of Astragalus tegetarioides M. E. Jones (Fabaceae) _..._-.--....-.._---2--_------2 2222-2. OwnBEY, MARION (see McNeal, Dale W.) PARKER, ALBERT J., Environmental and compositional ordinations of conifer forests in Yosemite National Park, California _______.___-__-_------ ee PARSONS, DAvID J., The role of plant ecological research in Sierran park man- agement: a tribute to Jack Major _--_-----------_-------_---___-__-__-__-__----._ PORTER, DUNCAN M. and Mary LINDA SMYTH, Noteworthy collection of Cen- chrus incertus M. A. Curtis (Cyperaceae) _...____._.____.____-.--.________-...- RADOSEVICH, S. R. (see Conard, S. G.) RAVEN, PETER H. (see Khyos, Donald W.) ROSENTRETER, ROGER (see Lackschewitz, Klaus H.) SAWYER, JOHN and EDWARD Cope, Noteworthy collection of Abies lasiocarpa (Hook.): Nutt, (Pinaceae) _-_-......___.--. ee SCHLESSMAN, MARK A., Taxonomy of Lomatium bicolor (Umbelliferae) _______- SMYTH, MAry LINDA (see Porter, Duncan M.) STANSELL, VEVA (see Lang, Frank A.) STEBBINS, G. LEDYARD, Floristic affinities of the high Sierra Nevada -----_-____- STICKNEY, PETER F. (see Lackschewitz, Klaus H.) STONE, R. DoucG, Noteworthy collection of Oxytheca watsonii Torrey & Gray WE Oly 2 OMAC CAG) ee tater see ee Se oe oe eh nese eee STROTHER, JOHN L. Dicoria argentea (Compositae: Ambrosiinae), a new species from Sonora, Mexico 237 87 122 274 Ag pA| 62 58 oy 123 79 i ee 58 109 220 217 218 118 189 273 TODSEN, THOMAS K., Noteworthy collections of Plummera ambigens, Eysen- hardtia polystachya, Cupheawrighti1, Aspicarpahirtella, Heuchera glomerulata eer ea EE a RN Paelee PO Cat te ht ie ea wk on 9 oN ly oS 60 (see also under Wendt, Tom) VANKAT, JOHN L., A gradient perspective on the vegetation of Sequoia National Park, California: «2222 oe 200 WENDT, TOM and THOMAS K. TODSEN, A new variety of Polygala rimulicola (Polygalaceae) from Dona Ana County, New Mexico _______-__-_______-______ 19 WERFF, HENK VAN DER, Noteworthy collections of Hieracium argutum, Cypselea humifusa, Nelumbo lutea, Lasthenia glaberrima __------------------------------ D2 WHITNEY, KENNETH D., A survey of the corticolous Myxomycetes of Califor- 0: eee ee eS ee See ee es tn ENS Poe PMNS ce 259 WILKEN, DIETER H. (see Kelley, Walt) WORTHINGTON, RICHARD D., Noteworthy collection of Salvia summa A. Nelson (aimiacéae i co 7 ee Fe ee a een ere 2 YOUNG, DAVID A., Wood anatomy of Actinocheita (Anacardiaceae) -_------------ 61 PECCATA, ERRATA, ET CONFUSIONES In addition to the phenomenally small number of typographical errors in Volume 29, most of which were of utterly no significance, the “received—accepted” data were omitted from all the articles in issue no. 1, for Jan 1982. No one has complained about this, but my apologies anyway. Benedict and Major: received 18 May 1980; revision accepted 11 Nov 1980. Meinke and Constance: received 6 Jan 1981; revision accepted 23 Mar 1981. Wendt and Todsen: received 17 Feb 1981; revision accepted 26 Mar 1981. Minnich: received 25 Feb 1981; revision accepted 1 May 1981. Cook: received 27 Mar 1980; revision accepted 29 Jul 1980. Conard and Radosevich: received 11 Jul 1980; revision accepted 27 Dec 1980. John Strother also informs me that he would have preferred the illustration in his paper (Dicoria, Fig. 1, 29(2)) to have been rotated 90° counterclockwise. He is right. It would have been better and would have taken less space. We also apologize for the collating error that resulted in a number of deletions in Madrono copies sent to the Bay area. ERIN ASONTA~ Rok. JAN 2% 1982 ° eel BA RIES MADRONO VOLUME 29, NUMBER 1 JANUARY 1982 Contents A PHYSIOGRAPHIC CLASSIFICATION OF SUBALPINE MEADOWS OF THE SIERRA NEVADA, CALIFORNIA, Nathan B. Benedict and Jack Major 1 LOMATIUM OREGANUM AND L. GREENMANII (UMBELLIFERAE), Two LITTLE KNOWN ALPINE ENDEMICS FROM NORTHEASTERN OREGON, Robert J. Meinke and Lincoln Constance 13 A NEw VARIETY OF POLYGALA RIMULICOLA (POLYGALACEAE) FROM DONA ANA COUNTY, NEW MExIco, Tom Wendt and Thomas K. Todsen 19 PSEUDOTSUGA MACROCARPA IN BAJA CALIFORNIA? Richard A. Minnich we) UNILATERAL INFERTILITY IN PARAPATRIC SPECIES OF ESCHSCHOLZIA (PAPAVERACEAE): SELECTION FOR ISOLATION? Stanton A. Cook 32 POST-FIRE SUCCESSION IN WHITE FIR (ABIES CONCOLOR) VEGETATION OF THE NORTHERN SIERRA NEVADA, S.G. Conard and S. R. Radosevich 42 NOTEWORTHY COLLECTIONS oy NOTES AND NEWS WoopD ANATOMY OF Actinocheita (ANACARDIACEAE), David A. Young 61 MISCELLANEOUS CHROMOSOME NUMBERS IN ASTERACEAE, Donald W. Kyhos and Peter H. Raven 62 REVIEWS 63 ANNOUNCEMENTS 65, 66 WEST AMERICAN JOURNAL OF BOTANY A JBLISHED 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, Herbarium, Life Sciences Building, University of California, Berkeley, CA 94720. Subscription rate: $25 per calendar year. Subscription information on inside back cover. Established 1916. Second-class postage paid at Berkeley, CA, and additional mailing offices. Return requested. POSTMASTER: Send address changes to Dr. Frank Almeda, Botany Dept., California Academy of Sciences, San Francisco, CA 94118. Editor—CHRISTOPHER DAVIDSON Idaho Botanical Garden P.O. Box 2140 Boise, Idaho 83701 Board of Editors Class of: 1981—DANIEL J. CRAWFORD, Ohio State University, Columbus JAMES HENRICKSON, California State University, Los Angeles 1982—DEAN W. TAYLOR, University of California, Davis RICHARD VOGL, California State University, Los Angeles 1983—-ROBERT W. CRUDEN, University of Iowa, Iowa City DUNCAN M. PoRTER, Virginia Polytechnic Institute and State University, Blacksburg 1984—Mary E. BARKWoRTH, Utah State University, Logan HARRY D. THIERS, San Francisco State University, San Francisco 1985—STERLING C. KEELEY, Whittier College, Whittier, CA ARTHUR C. GIBSON, University of California, Los Angeles CALIFORNIA BOTANICAL SOCIETY, INC. OFFICERS FOR 1981 President: ROBERT ORNDUFF, Department of Botany, University of California, Berkeley 94720 First Vice President: LAURAMAY T. DEMPSTER, Jepson Herbarium, Department of Botany, University of California, Berkeley 94720 Second Vice President: CLIFTON F. SMITH, Santa Barbara Museum of Natural History, Santa Barbara, CA 93105 Recording Secretary: ROBERT W. PATTERSON, Department of Biology, San Francisco State University, San Francisco, CA 94132 Corresponding Secretary: FRANK ALMEDA, Department of Botany, California Academy of Sciences, Golden Gate Park, San Francisco 94118 Treasurer: CHERIE L. R. WETZEL, Department of Biology, City College of San Francisco, 50 Phelan Ave., San Francisco, CA 94112 The Council of the California Botanical Society consists of the officers listed above plus the immediate Past President, WAYNE SAVAGE, Department of Biology, San Jose State University, San Jose, CA 95192; the Editor of MADRONO; three elected Council Members: PAUL C. SILvA, University Herbarium, Department of Botany, University of California, Berkeley 94720; JOHN M. TucKER, Department of Botany, University of California, Davis 95616; CHARLES F. QUIBELL, Department of Biological Sciences, Sonoma State College, Rohnert Park, CA 94928; and a Graduate Student Represen- tative, KENT HOLSINGER, Department of Biological Sciences, Stanford University, Stanford, CA 94305. A PHYSIOGRAPHIC CLASSIFICATION OF SUBALPINE MEADOWS OF THE SIERRA NEVADA, CALIFORNIA NATHAN B. BENEDICT Department of Biology, University of Nevada, Reno 89557 JAcK MAjor Department of Botany, University of California, Davis 95616 ABSTRACT Sierran subalpine meadows in the upper Kern River drainage are classified based on their physiographic characteristics. Two main divisions are recognized: meadows with predominantly vegetated margins (Type I), and meadows with predominantly sandy margins (Type II). Meadows of both types occur in a variety of topographic positions. Geographic distribution, sandy margins, hydrology, and geologic stability of the various meadow types are discussed. Meadows constitute an important element of the subalpine zone of the Sierra Nevada of California. Previous research on Sierran subal- pine meadows has concentrated primarily on an assessment of their condition and trend in a range management context (Armstrong 1942, Bennett 1965, Sharsmith 1959, Strand 1972, Sumner 1948). In the 1960’s an ecosystem analysis was undertaken that compiled detailed data on various components of the meadow ecosystems in the Rock Creek drainage, Sequoia National Park (Hubbard et al. 1965, Hub- bard et al. 1966, Harkin and Schultz 1966, Leonard et al. 1967, Leon- ard and Johnson 1969, Griffen et al. 1970). Without a meadow classifica- tion this information is difficult to synthesize, and cannot readily be used outside of the area where it was generated. Other data on Sierran meadows are scattered through the literature (Klickoff 1965, Pemble 1970, Beguin and Major 1975, Burke 1980, Ratliff 1979). Our research aims at understanding the dynamics of Sierran sub- alpine meadow ecosystems through plant community analysis. The objectives are: 1) description and classification of subalpine meadow vegetation; 2) correlation of vegetation types and species distribution with environmental factors; 3) study of the origins and successional sequences of meadow ecosystems; and 4) effects of meadow origins on meadow ecosystems (Benedict and Major 1979). To meet these objec- tives we have developed two meadow classifications: a physiographic classification of whole meadows, and a vegetation classification within meadows. This paper presents the physiographic classification of sub- alpine meadows. A descriptive classification of an ecosystem is an MADRONO, Vol. 29, No. 1, pp. 1-12, 15 January 1982 2 MADRONO [Vol. 29 important first step in understanding an ecosystem because it provides an organizing structure for future information. Field work was carried out during the summers of 1977, 1978, and 1979. Observations and descriptions of the physiographic character- istics of subalpine meadows seen during this period were recorded (i.e., meadow location in relation to water source, presence and lo- cation of springs and seeps, drainage patterns, glacial deposits, bed- rock outcrops, etc.). Vegetation sampling, environmental measure- ments, and soil coring were done concurrently for other aspects of this project. Field work was concentrated in the southern part of Sequoia National Park and adjacent U.S. Forest Service lands, although var- ious meadows throughout the southern Sierra were examined (Fig. 1). DESCRIPTION OF STUDY AREA The Sierra Nevada is a tilted fault block with a gradual slope to the west and a steep escarpment on the east. The crest runs in a north- south direction and includes Mt. Whitney (4418 m), the highest peak in the contiguous United States. In the southern Sierra, the Kern River runs north-south roughly parallel to and west of the main crest (Fig. 1). On the west side of the Kern River basin is the Great Western Divide with peaks to 4165 m (Midway Mountain). The river follows a fault in its upper portions (Lawson 1904, Webb 1946) forming a canyon 730 m deep opposite Whitney Creek. The meadows examined for this study occur between 2800 m and 3530 m primarily along the Whitney Creek and Rock Creek drainages. At these elevations the forest vegetation consists primarily of lodgepole pine (Pinus contorta subsp. murrayana) stands and foxtail pine (Pinus balfouriana) stands. The lodgepole pine forests are generally open with sparse understory vegetation. They occur along river bottoms, in gla- ciated valleys, and on some slopes. The foxtail pine forests are very open with little understory vegetation. They occur on steep slopes (both bedrock and morainal), and on many of the ancient erosion surfaces described by Lawson (1904) and Matthes (1950, 1962). PHYSIOGRAPHIC CLASSIFICATION In the Kern River drainage, within and just south of Sequoia Na- tional Park, there are two broad types of subalpine meadows: those with predominantly vegetated margins and continuous vegetation (Type I, Fig. 2); and those with predominantly sandy margins and sparse vegetation (Type II, Fig. 3). Type I meadows typically have dense meadow vegetation bordering directly on forest vegetation (al- though there can be distinct differences in the species composition along the forest-meadow ecotone), occur in areas glaciated relatively recently (probably late Wisconsin), and are surrounded by a forest composed mainly of Pinus contorta subsp. murrayana, at least BENEDICT AND MAJOR: SUBALPINE MEADOWS 1982] ‘yaeg jeuonenN eronbas ut says Apnys Jo uoNneI0T “Tf ‘OI JOaQUNN MOPBaWZ H221D Y290H OH Asepunog mopea-w —-— Asepunog yieg |euojyeN Bjonbags —— 814 0AND—(h) MOpBaW- 3a1}qesD saddn (€) MOPB2aWF 391}qGeID J9MO7 (2) mopeayw Apues = (1) S19}SWOIIy Z puabaq | ayy | ajeos _——$$ wy OL L n _—————n iS sayjw Ss t a Go 3 v OS ee ae = 7400 uB}aq\S o yBaq uayobs04 a —— ~~ So ys0 2 ° 3 7) Yag74 W204 3 1 a ® 2 < *\ ® Roum e] ) 0.1 m DBH) to obtain relative successional ages of the stands. Vegetation transects. The large structural differences between montane chaparral and forest sites necessitated the adaptation of sam- pling methods to vegetation structure. Sites dominated by shrubs were sampled by line intercept (Canfield 1941) to determine canopy cover- age, species composition, and vertical structure of the vegetation. Six 25-m tape transects each were established at random on sites 1, 3, 4, 5, and 10. Horizontal canopy interception and canopy height (nearest 0.2 m) were recorded by species at 0.1-m intervals. Percent cover and mean canopy height were determined for each species on each transect. Total canopy coverage was calculated as 100 percent minus the per- centage of ground surface with no canopy cover. Tree density was evaluated by counting all saplings, by species, in a 1 X 25-m belt transect adjacent to each line transect. All transects were perpendic- ular to slope contours. On sites where trees formed the dominant canopy (2, 6, 7, 8, 9, 10, 11) vegetation was sampled by the point-centered quarter (PCQ) meth- od (Cottam and Curtis 1956). Twenty points were established per site, except on site 8 where space permitted only eight points. Points (7-30 m apart depending on vegetation density) were placed along 3-5 ran- domly located transects across the slope contours. The number of transects per site was determined by the maximum transect length that was possible within each stand. Species and trunk diameter were recorded for four trees (>0.1 m DBH) at each point. Diameters were measured with a diameter tape. Determination of total density followed Morisita (1960). Density, relative density, mean basal area, total basal area, and relative basal area were determined for each species. Overstory canopy cover of each species was estimated visually at each point and mean canopy cover was determined by species for all points on each study site. Site 10 was sampled by both line transect and PCQ because trees had emerged above the brush but shrub cover was still high. Association analysis. To characterize further the vegetation and flora, releves (Braun-Blanquet 1932) were taken along each PCQ and line transect. A seven-class cover-density scale was used in which R = single individual; + = <1 percent; 1 = 1-5 percent; 2 = 6-25 per- 46 MADRONO [Vol. 29 cent; 3 = 26-50 percent; 4 = 51-75 percent; and 5 = 76-100 percent cover. Releve data were subjected to association analysis (Ceska and Roemer 1971) on a Burroughs 6700 computer. The program identified groups of species that tended to occur together. Each group of species had a corresponding group of stands, such that each of the species occurred in at least 40 percent of the stands and each stand contained at least 40 percent of the species. In addition, none of the species could occur in more than 25 percent of the stands outside the group. Each stand was identified in the association table by a three-digit number. The first two digits are the number for the study site (Table 1) and the third digit represents a stand number within the study site. Polar ordination. A polar ordination (Bray and Curtis 1957) was performed on relevé cover data. Log transformation of mean cover values for each class increased the relative importance of indicator species with low cover values. Interstand similarity was calculated using Sorensen’s community coefficient (Sorensen 1948). The maxi- mum similarity between samples in functionally identical vegetation was assumed to be 0.9 in calculations of interstand distance. For each of three axes, pole stands were selected that maximized an index of efficiency and reliability for the ordination. Stands were positioned along ordination axes using the Pythagorean method described by Beals (1960). Diameter-class distributions. Trees sampled on the PCQ transects were aggregated into 0.1-m diameter classes to generate frequency distributions of diameter classes for sites 2, 6, 7, 8, 9, 10 and 11. Fre- quency was expressed as a percentage of the total number of trees sampled on each site. Flora. A complete list of the 146 species in 35 families encountered in the field investigations is available in Conard (1980). RESULTS AND DISCUSSION Montane chaparral sites. At sites where most of the conifers were beneath the shrub canopy (1, 3, 4, 5) the average shrub cover was 81 percent (Table 2). However, low shrub cover (35 percent) and large numbers of dead and dying shrubs were noted on site 10. Sixty percent of the coniferous saplings encountered there had overtopped the shrub canopy, and A. concolor canopy coverage was 47 percent. Shrubs on site 10 were apparently being replaced by Abies concolor. Ceanothus velutinus was the dominant shrub species on sites 1, 3, 4, and 5. However, in the older stands (sites 3 and 5), cover of Arc- tostaphylos patula increased to approximately 25 percent (Table 2). Chrysolepis sempervirens was also an important species on many mon- tane chaparral sites. It was usually present in small numbers beneath the canopies of the Ceanothus velutinus and Arctostaphylos patula. 47 CONARD AND RADOSEVICH: POST-FIRE SUCCESSION 1982] € +l So cL €+ 18 0 OF 0-9 Cy Cg vl 89 sd * * * * > 50 CL * PO eee Osea 1 OOO042 70 -C +e SOD +o 1 ee 8 COG be si Se prOoOFTT €+99 TOFO0T 6 OS O0.23°0 2 S17 (uu) (%) (uu) (%) (uu) (%) 143109 I9A07 1YSI0H J9A0ZD 14310 I9A07 v ¢ ays Apnys “TVadVdVH,) ANVINOJY Ad GALVNINOCG SALIS NO SONITdVS SNOWAAINOZD AGNV SHNAHS AO SLHOITH AGNV FWOVAAAOZ) AdONV.) I + 96 ¢ + 8 e OL S'8P ST+ ST et+ el $+ 6 8 + LY 98 Se Vag SO +..t sll + 6% vO ee TO+ £0 f+ 9 TO+60 II + 8z sO';0 + €'T oo Or et T+ 7 TO PT Se ae eee 20 fea cae | 7+ (uu) (%) (uu) (%) JBI I3A07 1YyS1I0H IaA0) Ol JIAO) [e}0], JIAO) 991} [V}O], DUDIJAIQUD] SNULG DIYIUSDU *P 40]0IUOI Sa1QY Sool, J9AO) qniys BUIAT] [e}O]T, ysniq pesq CA P| MNUISSISOISIA SAQ1Y DI OfIU1IIDA SNIAINY) DIDUIBADUA SNUNA DUDIAIJNOIS XIDS SUadtaAagduas s1gajOSKAy ) Dinqwd SO]KYdDISOJIAP Snulynjaa snyiOUDa) sqniys sa1vads ‘10119 prepueys UO F pajUasaid s1e SUBITA, “JSISUNOA JU} SI p J}IS SVIIZYM JSapo 9} SI OT ANS “(I IQe) WY UMOID 4Se] IY} WIJ JWT} PI}JEUIT}SA JY} 0} SUIPIOIIV pasuKsIIe 91e So}IS APN}S “19AOD JUVdIEd [ ULY} SSI] SayBOIPUT, ‘7 ATAVL 48 MADRONO [Vol. 29 At site 10, where the tree canopy was beginning to close, C. semper- virens was the dominant shrub. These observations suggest that C. sempervirens is more shade tolerant than C. velutinus or A. patula. Other common shrub associates that had locally high cover on some sites but rarely attained a high percentage of the total vegetation cover included Salix scouleriana, Prunus emarginata, Quercus vaccinifolia, Ribes roezli1, and R. viscosissimum (Table 2). High densities of seedling and sapling Abies concolor (1100 to 11,000 st/ha) were observed on all montane chaparral sites except site 4, which had experienced the most recent burn. The high density (533 st/ha) of dead A. concolor present on that site five years after the fire indicated that it had supported an A. concolor forest in the past. The presence of several mature trees that had escaped the fire indicated that a seed source was available. The absence of A. concolor repro- duction suggests that several years may be required on some sites before conditions are favorable for A. concolor establishment. Forest sites. The mean relative density and relative basal area of Abies concolor on forest sites were 93 and 94 percent, respectively (Table 3). Total stand basal areas increased from 13 m?/ha on site 10 to 127 m?/ha on the oldest site studied (site 8). Only on site 2 were the relative densities of Abies magnifica and Pinus jeffreyi greater than 10 percent. Other occasional associates were Calocedrus decurrens and Pinus lambertiana. A clump of large Salix scouleriana occurred on site 9, although the species is more typical of montane chaparral sites. Shrub cover on forested sites was less than 5 percent except on site 10. The cover of A. concolor saplings (trees <0.1 m DBH) ranged from 1.3 percent (site 7) to 13.6 percent (site 2). Cover of A. magnifica saplings was 6.4 percent on site 6. In addition the mean basal area per tree of A. magnifica was only 28, 45, and 46 percent of that for A. concolor on sites 6, 9, and 2, respectively (Table 3). This suggests that A. magnifica is seral to A. concolor on some sites. Perhaps the cool understory microclimate at the upper elevation range of A. con- color forests favors A. magnifica reproduction. Reproduction and size-class distributions. The range in tree den- sities of the younger stands (2, 7, 9, 10, 11) was comparable with that observed by Schumacher (1926) for even-aged stands of A. concolor (>0.1 m DBH) in the Sierra Nevada. Schumacher (1926) also noted a decrease in density with increasing stand age. A negative relationship between relative age, expressed as basal area/tree (A), and tree density (d) was also observed in our study (In d = —0.073 In A + 5.04; Pr = 0.86, p < 0.01). This relationship suggests that the amount of A. con- color seedling establishment decreases with increasing stand age, even though some establishment does occur in stands of all ages. The diameter class distributions of A. concolor in stands of different 49 CONARD AND RADOSEVICH: POST-FIRE SUCCESSION 1982] (9013/01) (ey/-W) OOT 6°66 L10°0 Deel bl + S08 6£0°0 8 $9 v2 + 9891 y90°0 | ae Oe) COO 0 020°0 CeO 0 6.0 270 £t0.0 v0 LC c10°0 OFC eee Sel olac TSO0°O 9°6 L81 €20'0 b'0 + 90 v0 + 80 OT+ ST ec+ oe 610°0 6'v vac O£0°O 6°7 + £°6 s°¢ + 9°76 L100 cel SLL Ty0'O vits Ivcl £90°0 (%) (%) (9043/1) (ey/z) = (ey/rs) (9943/0) (ey/,t) (eqs) Vd P V Vd P V Vd P V SOUS [[@ JIAO Uva OT ays Apnys Z ats Apnys £90°0 v'Tol 61 + ¢Ist €1T'0 9 cIT SL £26 Of '0 1901 S = 09F v9°0 Ce $ LEO 80) 61 £¢.0 Om ST 80°0 8°0 OT 60°0 8c O¢ 890°0 BaOr v6rl €1'0 6°801 798 ce "0 v'66 ple v9°0 (901}/7U1) + (BY/,U1) (eys) (9017/01) (e8Y/,U1) (eys) (901}/,01) (ey/,W) (eys) V Vd P V Vd P V Vd P V IT aus Apnys L ays Apnys 9 ays Apnys (991}/,U1) (BY /,U1) va 8 ays Apnys 8 Ir 20 |e, ae) C0 6 OV Vd 6 ays Apnys ee Let v9 (eys) p [BIOL SS Td {d ae WV OV. sa1vads 9 + 661 661 (eqns) P [BIOL SS Td {d qo WV OV satvads ‘10119 prepuejs UO F UMOYS IIe SULIT [[PIIAG “VUDIAIINOIS XID pue ‘DUDIZAIQuD] Snulg ‘1Kadffal snuig ‘Suasdnrap SnApar0iDy ‘vayiusvu saiqgp ‘40]09U0I Sa1qP 0} AJajal SS pue “Tg ‘[g ‘dO ‘INV ‘OV ‘3S98UNOA 9U} SI OT 9}IS sevaIayM JUS JSIP[O JY} ST QB INS ‘II SL] IY} WOIJ BUT] PayeUIT}Sa VY} 0} ZUIPIOIIV pasuvlIe a1e Sd}IS APNIS ‘painsvaul 919M S9aI} OSp JO [e}0} Y “SALIS aqdaLsauoy NO Hq Wp ANO NVHL WALVANY SAAAL AO (VY) ATUL Aad VAUY TvSvgq Nva ANV ‘(Vq) Vauy Ivsvg ‘(p) ALISNAG '¢ ATAV LL 50 MADRONO [Vol. 29 100 r Site 10 Site 2 Site9 Site /l 80 60 ll, he (hme. Ue, 0.2 04 06 0.2 04 06 0.2 04 O06 Q2 04 06 fo) 80 Site 7 Site 6 Site 8 RELATIVE FREQUENCY (%) 60 40 20 02 04 06 08 02 04 06 08 1.0 02 04 06 O08 1.0 TREE DIAMETER (m) Fic. 1. Frequency distributions of diameter classes (0.1-m intervals) for trees >0.1 m DBH on sites dominated by trees. Study sites are arranged according to the estimated time from the last fire. Site 10 is the most recent whereas site 8 is the oldest. ages also indicate a decrease in seedling establishment as stands ma- ture (Fig. 1). However, both the wide range of diameters observed in old stands and the ability of A. concolor to become established under low light levels suggest that growth suppression of trees in the popu- lations is combined with a certain level of continued reproduction. Schumacher (1926) observed similar shifts in size-class distributions for A. concolor stands that he considered to be even-aged. However, as stand ages in his studies were determined only from dominant trees, an evaluation of the entire age class distribution for his stands is im- possible. In an even-aged stand, height and growth rate of individual trees should be strongly correlated. A comparison of heights and growth rates of white fir saplings on two montane chaparral sites did not support such a relationship (Conard 1980). Rather, variations in height for trees of similar growth rates encompassed nearly the entire range of values expected for a continuously reproducing population. Al- though detailed age-class information is lacking, the evidence pre- sented suggests that most A. concolor stands are not even-aged. Association analysis. Despite the wide range in age and structure of the sites selected for study, only two species groups were detected 1982] CONARD AND RADOSEVICH: POST-FIRE SUCCESSION 51 0.8 ) x \ = 5G =<. aq 0.4494/i4 56% er Bo. _\ > ®e 23 | | I | | | | | \ \ X AXIS Fic. 2. Polar ordination of stands. Stands are identified by site number. The last digit of each number identifies individual stands within study sites. Arrows indicate the direction of post-fire succession. by association analysis (Table 4). The most widespread group, which occurred in 77 percent of the stands, included Abies concolor, Cea- nothus velutinus, Arctostaphylos patula, and Chrysolepis sempervi- rens. The stands in which this group did not occur were generally those with such a high A. concolor canopy coverage that very few shrubs were present. The other species group, which occurred in 18 percent of the stands, contained Pyrola picta, Pteridium aquilinum, Chimaphila menziesii, Corallorhiza striata, and Kelloggia galioides. Other species restricted to relevés in which this group occurred were Pterospora andromedea, Sarcodes sanguinea, Hieracium albiflorum, and Pleuricospora fim- briolata. This group of nine high fidelity species includes representa- tives of five of the ten genera of Pyrolaceae in California, most of which are probably root parasites. The nearly exclusive occurrence of this group in the late phases of the seral sequence is probably in re- 52 MADRONO [Vol. 29 TABLE 4. ASSOCIATION TABLE FOR RELEVE DATA FROM INDIVIDUAL STANDS. Species-stand groupings are indicated by heavy horizontal lines. Symbols in the table are cover classes described in the text. Montane chaparral 000000000 000 00000 0 0 0 0 444 44 4 1 1121 121 213 3 3 33 3 5 5 5 1} 2, 3°45 6" V2 34°55 6) 1 42 3 45° 6 2-3 Sr?) Species name Stand number LS) _ Abies concolor I. 21-2274 Ceanothus velutinus 5 45 44 43 4 4 4 4 4 Arctostaphylos patula + 2 1 1 + 2 2 ie 2 Chrysolepis sempervirens cme) + R&S fF LY RW dbK eS - Ww me Wwhr mS DOM RoW hd bo MR HK bo Pyrola picta R Pteridium aquilinum ar Chimaphila menziesti Corallorhiza striata Kelloggia galioides 1 Pterospora andromedea Sarcodes sanguinea Pleuricospora fimbriolata Hieracium albiflorum Abies magnifica Pinus ponderosa a fs Pinus monticola Pinus lambertiana Pinus jeffreyi Calocedrus decurrens R Stephanomeria lactucina Sitanion hystrix R + + Gayophytum diffusum parviflorum R R R R + Purshia tridentata 1 Ceanothus cordulatus + R Quercus vaccintifolia 43 43 4 Ribes viscosissimum hallii 1 + + R Ribes roezlii + + + Salix scouleriana 1 t- Prunus emarginata + =F + + 1 + 1 Alnus tenuifolia Apocynum pumilum + + Symphoricarpos acutus + + 1 + 1 sponse to the low light intensities common on the forest floor (often 1 percent of full sunlight). This species group is remarkably similar to the “Pirola-Corallorhiza” union in Abies magnifica forests described by Oosting and Billings (1943), who speculated that the same union should also occur in the mixed conifer forests in the Sierra Nevada. Breaches in the canopy caused by windthrow or death of canopy trees became more common in older forests. Such openings were patchy in their distribution and lent a locally heterogeneous appear- ance to the vegetation (Bonnicksen 1975). Increased light penetration to the forest floor permitted development of some understory vegeta- tion in these areas. Stands 025, 095, 096, 097, and 074 represent local patches of low canopy coverage (Table 4). These areas are character- ized by an increased cover of Abies concolor saplings and montane chaparral shrubs, especially Chrysolepis. Where openings are large 1982] CONARD AND RADOSEVICH: POST-FIRE SUCCESSION oS TABLE 4. CONTINUED. ae Abie Transition bies concolor forest Young Mature 000110 0 1 1 ~0 2°97 91 129 9a 1 al 3.2 3 12 14 3 4~«21 ~r NNO NY DO waned Wwwo lon [on lon oe) io.<) 3 243 3 23 24 Adouey wpt> Adour) wp I> Adoue) wp [> Adouey wpt> Adour satvads Hdd Hdd Hdd Hdd Hdd Hdd I9AOD % IZA0I % I9AOD % I9AOD % I9A0I % I9A09 % Oe} acl vel ecl (aa Let jaquinu pue}s ‘O1ox JSOU PUB VINJEU ssay SEM 97] ayIG SBIIBYM JISOW }SOW PUL 2INJeUI SOUT SBM [7] OIG “JUSUIUOITAUA PU APLINU OF Zupios9e pasuesse se spuvys (71 ays Apnys) pozsa}01d ssa] auedaq adojs ay} pue pauaplM Udye} SEA JOISURI] JY} YIM Ul UOAURD oY} SE, IISIU,, SS2] SWIBIA JUSMUOATAUD BY “OY SUBS OY} Jaye sazejs [B1as juasaidai spurs [TW “MNOLNOD W-0S/1 AHL ONOTV ANAGIY 40 ANVTY AHL GNNOAV LOASNVAL W-OS7 V ONOTV GANIVLAO SFAdTAY °$ ATAWL 1982] CONARD AND RADOSEVICH: POST-FIRE SUCCESSION 5) 5 percent cover of montane chaparral shrubs (Table 4). The impor- tance of including information on canopy coverage, as well as species presence, in the ordination is illustrated by the high correlation coef- ficient for this ordination (r = 0.80) when compared with an ordina- tion based solely on species presence (r = 0.64). Stands 121 through 126 illustrate the potential influence of environ- mental conditions on the absolute time scale of the seral sequence (Table 5). These stands represent a transect from montane chaparral (125, 126) into a closed canopy Abies concolor forest (121, 122). The transect paralleled a gradient in aspect and exposure along a slope that had been burned uniformly, but where succession had proceeded more rapidly at the mesic end of the gradient. The changes in structure observed on the transect were similar to those discussed previously for other sites (Tables 2—4). A decrease in shrub canopy coverage from 98 to 1 percent between the xeric and mesic ends of the transect was associated with an increase in tree canopy coverage from 14 to 75 percent. Increased tree canopy density also corresponded with increas- ing cover by dead shrub canopies. In stand 122, where few living shrubs were observed, dead shrub canopy coverage was 80 percent. The mesic end of the gradient supported Abzes concolor-dominated forest. However, due to the xeric nature of the shrub-dominated end of this gradient, species composition of these stands was more typical of that at lower elevations. This resulted in a separation of the xeric stands (123 to 126) from other montane chaparral stands in the polar ordination (Fig. 2). Despite these large differences in species compo- sition, the transect demonstrates a directional sequence in the polar ordination that parallels the sequence observed for the other stands. ACKNOWLEDGMENTS We thank J. Major for his cooperation and insight throughout this research and for his helpful comments on the manuscript, D. Randall for assistance with vegetation analysis, and J. Greene for help with the field work. This study was a portion of the senior author’s research requirements for a Ph.D. at the University of California, Davis. LITERATURE CITED BEALS, E. 1960. Forest bird communities in the Apostle Islands of Wisconsin. Wilson Bull. 72:156-181. Bock, C. E. and J. H. Bock. 1977. Patterns of post-fire succession on the Donner Ridge burn, Sierra Nevada. USDA For. Serv. Gen. Techn. Rep. WO-3:464—469. BONNICKSEN, T. M. 1975. Spatial pattern and succession within a mixed conifer-giant sequoia forest ecosystem. M.S. thesis, Univ. California, Berkeley. BRAUN-BLANQUET, J. 1932. Plant sociology; the study of plant communities. McGraw- Hill, New York. Bray, J. R. and J. T. Curtis. 1957. An ordination of the upland forest communities of southern Wisconsin. Ecol. Monogr. 27:325-349. CANFIELD, R. 1941. Application of the line interception method in sampling range vegetation. J. Forest. (Washington) 39:388-394. 56 MADRONO [Vol. 29 CESKA, A. and H. ROEMER. 1971. A computer program for identifying species-relevé groups in vegetation studies. Vegetatio 23:255-—277. CONARD, S. G. 1980. Species interactions, succession, and environment in the montane chaparral-white fir vegetation of the northern Sierra Nevada, California. Ph.D. dissertation, Univ. California, Davis., Diss. Abstr. 41:2460 B. CoTTAM, G. and J. T. CuRTIS. 1956. The use of distance measures in phytosociological sampling. Ecology 37:45 1-460. FOWELLS, H. A., ed. 1965. Silvics of Forest Trees of the United States. Agric. Handbook 271. USDA For. Serv., Washington, DC. GorpDon, D. T. 1980. White fir. Jn F. H. Eyre, ed., Forest.cover types of the United States and Canada, p. 92—93. Society of American Foresters, Washington, D.C. GRAY, J. T. 1979. The vegetation of two California mountain slopes. Madrono 25:177-— 186. GRIFFIN, J. R. 1967. Soil moisture and vegetation patterns in northern California forests. USDA For. Serv. Res. Pap. PSW-46. GRIFFIN, J. R. and W. B. CRITCHFIELD. 1972. The distribution of forest trees in California. USDA For. Serv. Res. Pap. PSW-82. Reprinted with supplement, 1976. KILGORE, B. M. 1973. The ecological role of fire in Sierra conifer forests: its application to national park management. J. Quat. Res. 3:396-513. MorisiTa, M. 1960. A new method for the estimation of density by the spacing method applicable to non-randomly distributed populations. Physiology and Ecology 7:134-144 (1957). Transl. by M. Morris and J. P. Blaisdell from Japanese. USDA For. Serv., 20 p. Washington, D.C. OosTING, H. J. and W. D. BILLINGS. 1943. The red fir forest of the Sierra Nevada: Abietum magnificae. Ecol. Monogr. 13:259-274. RUNDEL, P. W. 1972. Habitat restriction in giant sequoia: the environmental control of grove boundaries. Amer. Mid]. Naturalist 87:81—99. RUNDEL, P. W., D. J. PARSONS, and D. T. GORDON. 1977. Montane and subalpine vegetation of the Sierra Nevada and Cascade ranges. Jn M. G. Barbour and J. Major, eds., Terrestrial vegetation of California, p. 559-600. Wiley-Interscience, New York. SCHUMACHER, F. X. 1926. Yield, stand, and volume tables for white fir in the Cali- fornia pine region. Univ. California Agric. Exp. Sta. Bull. 407. SKAU, C. M., R. O. MEEUWIG, and T. W. TOWNSEND. 1970. Ecology of eastside Sierra chaparral—a literature review. Univ. Nevada Max C. Fleischmann Coll. Agric. R71, 14 p. SORENSEN, T. 1948. A method of establishing groups of equal amplitude in plant sociology based on similarity of species content. Kongel Danske Vidensk.-Selsk. Skr. 5(4):1-34. VANKAT, J. L. and J. Major. 1978. Vegetation changes in Sequoia National Park, California. J. Biogeog. 5:377—402. WAGENER, W. W. 1961. Past fire incidence in the Sierra Nevada forests. J. Forest. (Washington) 59:734—747. WARING, R. H. 1972. An environmental grid for classifying coniferous forest ecosys- tems. Jn J. F. Franklin, L. J. Dempster, and R. H. Waring, eds., Proceedings— Research on coniferous forest ecosystems—a symposium, p. 79-91. USDA For. Serv., Portland, Oregon. WILKEN, G. C. 1967. History and fire record of a timberland brush field in the Sierra Nevada of California. Ecology 48:302—304. 1982] NOTEWORTHY COLLECTIONS SY NOTEWORTHY COLLECTIONS Readers and contributors will notice a change in the format of Noteworthy Collec- tions beginning with this issue. Increased page charges are part of the reason for this change, but also there are available many local natural history and biology journals in which such contributions are just as appropriate, if not more so, than in the pages of Madrono. Herbarium label data and extensive lists of previous localities are here no longer considered publishable forms of information within the Noteworthy Collection format. Madrono’s function will be to bring new records to the attention of floristic work- ers, who can then either write to collectors or borrow the cited specimens. Informa- tion submitted under the former heading of “previous knowledge” and complete informa- tion on “noteworthiness” will still be useful in judging the significance of the collections but will not necessarily appear in print. Arizona HYPoxXIS MEXICANA Schult. (AMARYLLIDACEAE).—Cochise Co., Garden Canyon, Huachuca Mts. (T23S R11E S1), 1905 m; 7 Aug 1979, McLaughlin and Bowers 1790 (ARIZ). Significance. First collection in AZ since the late 1800’s (Lemmon 2891). LINUM SUBTERES (Trelease) Winkler. (LINACEAE).—Coconino Co., 5.5 km sw. of Pipe Springs National Monument (T40N R6W S25), 1520 m; 12 Aug 1979 McLaughlin and Bowers 1813 (ARIZ). Significance. Second collection in AZ. Collected in the same area in 1946 (Parker 6243) but not included in Ariz. fl. (Kearney & Peebles 1960).—JANICE E. BOWERS and STEVEN P. MCLAUGHLIN, Office of Arid Lands Studies, Univ. Arizona, Tucson 85721. (Received 6 Jul 1981; accepted 26 Aug. 1981.) California ASPLENIUM TRICHOMANES L. (ASPLENIACEAE).—Del Norte Co.: ca. 36 m s. of 18 m falls on Myrtle Cr., Middle Fork of the Smith River, 3.5 km nne. of Hiouchi (T17N R1E in the se. corner of the nw.% of S33), 183 m, 30 Jun 1980, Lang 1400 (SOC). Significance. The first record for California. This collection extends the known range of the species s. and w. ca. 295 km of its nearest known location near McKenzie Pass, OR (Cole 54, OSC).—FRANK A. LANG, Dept. Biology, Southern Oregon State College, Ashland 97520 and VEVA STANSELL, Box 959, Gold Beach, OR 97444. (Re- ceived 29 May 1981; accepted 17 Aug 1981.) 58 MADRONO [Vol. 29 ASTRAGALUS TEGETARIOIDES M. E. Jones (FABACEAE).—Lassen Co., Ash Valley, along Adin-Madeline Rd., 26 km e. of Adin (T38N, R11W, Sec. 32), 1500 m, 6 Jul 1980, Nelson & Nelson 5988 (HSC, NY). (Determined by Rupert C. Barneby.) Significance. First known collection in CA, extending the range 325 km s. of its previously only known locations. This taxon is considered rare and endangered in OR and should be considered as such in CA. According to Barneby, the CA plants differ slightly from the OR ones, in that the calyx teeth are slightly longer, the banner is ca. 1 mm longer, and the corolla is purplish instead of white suffused with lilac. The habitat is also different. The CA population is growing in a sage brush steppe, whereas in OR they grow in a ponderosa pine forest.—THOMAS W. NELSON and JANE P. NELSON, Department of Biological Sciences, Humboldt State University, Arcata, CA. (Received 5 Dec 1980; accepted 23 Feb 1981.) Montana ANTENNARIA MONOCEPHALA D.C. (ASTERACEAE).—Deer Lodge Co., Anaconda- Pintlar Wilderness, E-Goat Pk. (T2N R14W SS), 2835 m, 10 Aug 1975, Lackschewitz 6266 (MONTU, WTU). Significance. First report for MT and contiguous USA. GENTIANA TENELLA Rottb. (GENTIANACEAE).—Madison Co., Gravelly Range, n. flank of Cave Mt. (T10S R1W $32), 2890 m, 7 Sep 1977, Stickney 3699, MRC; Carbon Co., Beartooth Mts., 0.5 km s. of Moon Lake (T9S R18E S820), 3040 m, 15 Aug 1979, Rosentreter 1535 (MONTU). Significance. First records for MT, a range extension of 50 km e. from Yellowstone Natl. Park. JUNCUS TRIGLUMIS L. (CYPERACEAE).—Carbon Co., Beartooth Mts., 1 km n. of WY border, 0.5 km n. of Glacier Lake (T9S R18E S29), 3000 m, 15 Sep 1976, Lack- schewitz 7033 (MONTU, MONT, COLO, NY, WTU); Hell Roaring Plateau, 3 km n. of WY border (T9S R18E $22), 3020 m, 11 Aug 1977, Lackschewitz 7782 (MONTU, COLO, NY, WTU). Significance. First records for MT, range extensions of 11 and 13 km from Park Co., WY. KOENIGIA ISLANDICA L. (POLYGONACEAE).—Carbon Co., Beartooth Mts., just along WY border off Hwy 212 (T9S R19E S32), 2960 m, 13 Aug 1977, Lackschewitz 7859 (MONTU); shallow pond, 1.5 km s. of Moon Lake (T9S R18E S829), 3040 m, 15 Aug 1979, Lackschewitz 9150 (MONTU, MONT, COLO, WTU); 0.5 kms. of Albino Lake (T9S R17E 825), 3050 m, 30 Aug 1980, Lesica 1027 (MONTU). Significance. First reports for MT, all occurrences less than 12 km from nearest known populations in Park Co., WY. LOMATIUM BICOLOR (Wats.) Coult. & Rose (APIACEAE).—Ravalli Co., Bitterroot Mts., e. slope off St. Josephs Rd. (T10N R20W 820 w.'%), 1730 m, 10 Jul 1974, Lackschewitz 5153 (MONTU, NY). (Det. by A. Conquist, NY, 1976); Selway-Bitterroot Wilderness, divide w. above Tin Cup Lake (T2N R23W S1), 1980 m, 19 Jun 1971, Lackschewitz 2728 (MONTU). Significance. First report for MT. MUSINEON VAGINATUM Rydb. (APIACEAE).—Missoula Co., ne. foothills of Bitterroot Mts., 19 km sw. of Missoula (T11N R20W S16 w.'%), summit of “McClay Mt.”, 1730 m, 29 May 1976, Lackschewitz 6477 (MONTU, WTU); 8 Jul 1976, Lackschewitz 6612 (MONTU, MONT, RM); Sapphire Mts., 9 km s. of Missoula (T12N R19W S30 nw.%), 1982] NOTEWORTHY COLLECTIONS 59 1310 m, 27 Apr 1978, Lackschewitz 79320 (MONTU, MONT, WTU); 0.66 kme. above previous location, 1460 m, Lackschewitz 7933 (MONTU); Granite Co., Garnet Range near Rattler Gulch Rd., 8 km w. of Drummond (T11N R13W S9 w.'%), 1410 m, 13 May 1977, Lackschewitz 7170 (MONTU, WTU, NY). (Collections confirmed by Ronald L. Hartman, RM, 1980.) Significance. Range extension in MT of 270 km to the nw. from previously known occurrence in Gallatin Co. PHACELIA THERMALIS Greene (HYDROPHYLLACEAE).—Phillips Co., C. M. Russell Natl. Wildlife Range, near Slippery Anne Ranger Station (T22N R23E S19), ca. 700 m, 12 Jun 1978, Lackschewitz 8125 (MONTU, MONT, NY); Garfield Co., York Island in Fort Peck Reservoir (T25N R41E S82 w.%), ca. 690 m, 28 Jun 1978, Lackschewitz 8248 (MONTU). Significance. First record for MT, ca. 800 km ne. of known range. PLANTAGO HIRTELLA H.B.K. (PLANTAGINACEAE).—Ravalli Co., Sapphire Mt. (T4N R19W S7), Sleeping Child Hot Springs Resort, 1370 m, 3 Jun 1978, Cory 1769 (MON- TU, NY). (Det. by A. Cronquist, NY, 1980.) Significance. First record for MT, more than 750 km from Grays Harbor, WA and first report of an occurrence this far inland in USA. RIBES TRISTE Pall. (GROSSULARIACEAE).—Granite Co., Anaconda-Pintlar Wilder- ness, near Falls Fork of Rock Cr. (T3N R1I5W S829 se.%), 2140 m, 25 Sep 1974, Lackschewitz 5888 (MONTU); Falls Fork Rock Cr. (T3N R15W $19), 1980 m, 20 Jul 1975, Lackschewitz 6076 (MONTU, NY, WTU). Significance. First report for MT, about 300 km e. of occurrences in the Blue Mts. OR and WA. RORIPPA SYLVESTRIS (L.) Besser (BRASSICACEAE).—Missoula Co., town of Missoula, Clark Fork island adjacent to U. of Montana campus, 970 m, 30 Jun 1979, Lackschewitz 8940 (MONTU, WTU); Konah Ranch, Clark Fork island (T13N R2W S8 se.%), 13 km w. of first collection, 940 m, 7 Aug 1979, Lackschewitz 9127 (MONTU). Significance. First report for MT, at least 280 km from nearest known occurrence. SATUREJA DOUGLASII (Benth.) Brig. (LAMIACEAE).—Sanders Co., Cabinet Gorge (T27N R34W S820), 670 m, 13 Aug 1970, Stickney 2197 (MRC, MONTU); Ravalli Co., Trapper Cr. Trail (T2N R21W S820), 1520 m, 18 Jul 1975, Cory 1494 (MONTU). Significance. First report for MT, 70 kme. of nearest known populations in Bonner and Idaho Cos., ID. SAUSSUREA DENSA (Hook) Rydb. (ASTERACEAE).—Lewis & Clark Co., Bob Marshall Wilderness, n. part of “Chinese Wall”, Cont. Divide (T24N R11W S31), 2470 m, 26 Jul 1979, Lackschewitz 9113 (MONTU, COLO, WTU). Significance. Second record for MT, 20 km w. of the occurrence in Teton Co. VERONICA VERNA L. (SCROPHULARIACEAE).—Mineral Co., near Gold Cr. (T18N R21W $31 sw.%) 3 km se. of St. Regis, 270 m, 19 May 1960, Stickney 354 (MRC); Missoula Co., 10 km w. of Missoula near Butler Cr. (T14N R20W S24), 1100 m, 8 May 1968, Stickney 1640 (MRC); Flathead Co., Hog Heaven Mining District (T25N R23W S16), 1190 m, 9 Jun 1977, Lackschewitz 7283 (MONTU, NY, WTU); Missoula Co., Waterworks Hill n. of the town of Missoula (T13N R19W S15), 1070 m, 11 May 1978, Lackschewitz 7941 (MONTU, MONT, WTU); Ravalli Co., Bitterroot Mts., Medicine Hot Springs, 1370 m, 21 May 1978, Lackschewitz 7952 (MONTU, WTU); Missoula Co., bare hills n. of Missoula (T14N R19W S35), 1220 m, 28 May 1978, 60 MADRONO [Vol. 29 Lackschewitz 7960 (MONTU, COLO, NY). (Verified by N. H. Holmgren, NY; W. A. Weber, COLO, 1980.) Significance. First report for MT and N. America.—KLAus H. LACKSCHEWITZ, PETER LESICA, ROGER ROSENTRETER, Dept. Botany, Univ. Montana, Missoula 59812, JACULYN K. Cory, SW581 Westside Rd., Hamilton, MT 59840 and PETER F. STICK- NEY, Associate Plant Ecologist, For. Sci. Lab., Drawer C, Missoula, MT 59806. (Re- ceived 4 Mar 1981; accepted 7 Aug 1981.) New Mexico PLUMMERA AMBIGENS Blake (ASTERACEAE).—Hidalgo Co.: Maverick Spring Can- yon, Peloncillo Mts. (T30S R21W S1 and S11), 1650 m, 15 Sep 1980, Todsen 800815-1 (NMC). Significance. The second locality in NM for this rare (proposed threatened) species, between the previously-known locations in NM and AZ. EYSENHARDTIA POLYSTACHYA (Ortega) Sarg. (FABACEAE).—Hidalgo Co.: Beehive Canyon, Peloncillo Mts. (T29S R21W S33), 1800 m, 4 Sep 1979, Todsen 790804-1 (NMC). Significance. Confirmation of occurrence in NM and extension n. 60 km. CUPHEA WRIGHTII Gray (LYTHRACEAE).—Hidalgo Co.: Skull Canyon, Peloncillo Mts. (T30S R21W S10), 1700 m, 3 Sep 1979, Todsen 790803-1 (NMC). Significance. First record from NM. ASPICARPA HIRTELLA Rich. sensu latu (MALPIGHIACEAE).—Hidalgo Co.: Peloncillo Mts., Skull Canyon (T30S R21W S10), 1800 m, 3 Sep 1979, Todsen 790803-2 (NMC), and Maverick Spring Canyon (T30S R21W S11), 1650 m, 6 Sep 1979, Todsen 790806-1 (NMC). Verified by William Anderson. Significance. First records for NM. HEUCHERA GLOMERULATA Rosendahl, Butters, and Lakela (SAXIFRAGACEAE).—Hi- dalgo Co.: Animas Peak (T31S R19W S28), 2100-2500 m, 28 May 1977, Todsen 770528-1 (NMC). Significance. First record for NM. A range extension of 50 km se.—THOMAS K. TODSEN, Dept. Biology, New Mexico State Univ., Las Cruces 88003. (Received 23 Mar 1981; accepted 28 May 1981.) Utah CAREX MICROGLOCHIN Wahl. (CYPERACEAE).—Duchesne Co., Ashley Natl. For., Uinta Mts., S. Fork Rock Cr. 50 km nw. of Duchesne (T2N R8W nw. USM), 2805 m, 26 Aug 1980, S. Goodrich 15061 (BRY, NY, OGDF). Significance. First record for UT, a w. range extension from c. CO. CAREX PARRYANA Dewey (CYPERACEAE).—Just below Joe’s Valley Dam (T18S R6E 24), 2190 m, s. facing slope, 25 Jun 1977, M. E. Lewis 4765 (OGDF); head of Rilda Canyon (Big East Mt.) (T16S R6E S17), 3110 m, 8 Aug 1977, M. E. Lewis 5125 (BRY, OGDEF); San Rafael Swell, Old Smiths Cabin (T21S R14E $5), 1390 m, 5 Jun 1979, J. Harris 326 (BRY). Significance. Range extension of 300 km from Cache Co. UT. EPILOBIUM NEVADENSE Munz. (ONAGRACEAE).—Miuillard Co., Fishlake Natl. For., Canyon Mts., Eightmile Cr. 11 km w. of Scipio (T18S R3W S32 sw.%), 2347 m, 12 Aug 1980, S. Goodrich 14918 (BRY, MO); above the head of John Williams Canyon, 11 km nw. of Scipio (T17S R3W S829 e.%), 2710 m, 2 Sep 1980, S. Goodrich 15144 (BRY, MO). Significance. A range extension of 230 km from s. UT. KOBRESIA SIMPLICIUSCULA (Wahl.) Mack. (CYPERACEAE).—Duchesne Co., Ashley 1982] NOTEWORTHY COLLECTIONS 61 Nat. For., Uinta Mts., S. Fork Rock Cr. 50 km nw. of Duchesne (T2N R8W S824 nw.%4 USM), 2805 m, 26 Aug 1980, S. Goodrich, 15068 (BRY, NY, OGDF). Emery Co., Scad Valley Meadow (T16S R6E S27), 2590 m, 24 Jul 1980, M. E. Lewis 6620 (OGDF). Significance. First records for UT, as. range extension of 370 km from Driggs, ID, and a w. range extension of about equal distance from CO.—SHEREL GOODRICH, Shrub Sciences Laboratory, Intermt. For. and Range Exp. Sta., USDA For. Serv., Provo, UT 84601, and Mont E. Lewis, R-4 USDA For. Serv., Ogden, UT 84401. (Received 10 Jul 1981; accepted 26 Aug 1981.) NOTES AND NEWS Woop ANATOMY OF Actinocheita (ANACARDIACEAE).—Actinocheita is a monotypic genus (A. potentillifolia (Turcz.) Bullock) of anacards occurring in the states of Guer- rero, Oaxaca and Puebla, Mexico. Although originally included in Rhus, Barkely (Ann. Missouri Bot. Gard. 24:1-10. 1937) removed the species from Rhus and placed it in its own segregate genus (Actinocheita), primarily on the basis of its axillary panicles, ovary on a gynobase, and long-villous trichomes covering the fruit. (It should be mentioned that there has been some disagreement regarding the correct specific epithet for this taxon. It appears that A. potentillifolia should be used instead of A. filicina (DC.) Barkley; see Barkley and Reed, Amer. Midl. Nat. 21:368-377. 1939; Barkley, The Biologist 28:9-23. 1945; Bullock, [Kew] Bull. Misc. Inform. 440-441. 1937; 337-339. 1939.) The purpose of this note is to present data on the wood anatomy of A. poten- tillifolia and to compare these data with those of species of Rhus and other Anacardiaceae. Wood of A. potentillifolia was collected by the author south of Oaxaca, Mexico and voucher specimens (Young 106, 107) were deposited in RSA and duplicates in ILL. Wood samples were prepared and analyzed as described elsewhere (Young, Aliso 8:133-— 146. 1974), and reference slides are available from the author. The wood of A. potentillifolia is distinctly diffuse porous, and the mean vessel mem- ber diameter was 76 wm (range = 30-115 wm, s” = 283). Vessels were mostly solitary or in short multiples, and the mean number of vessels/mm? was 33 (range = 26-40). All vessel members had simple perforation plates and were somewhat angular in outline, with alternate pitting on the lateral vessel walls (pits more or less elliptical). Mean vessel member length was 489 wm (range = 280-650 wm), and mean libriform fiber length was 942 wm (range = 570-1100 um). Tracheids were not detected. Vessel members also lacked helical sculpturing on their walls. Only uniseriate and biseriate rays were present, and the former predominated. Mean uniseriate ray height was 528 wm (range = 280-1150 pm). The rays were heterogeneous with upright, procumbent (predominated) and square cells. Crystals were present in the rays, but resin canals were not detected. Axial paren- chyma was scanty paratracheal. In its wood anatomy, A. potentillifolia is similar to Rhus as well as many other genera of tribe Rhoeae in possessing the following features: (1) simple perforation plates; (2) libriform fibers (no tracheids); (3) alternate, elliptical pitting on lateral vessel walls; and (4) heterogeneous, mostly uniseriate rays. The most distinctive feature distinguish- ing the wood of A. potentillifolia from that of species of Rhus is that the wood of Actinocheita is diffuse porous (with few vessels/mm?), whereas that of all Rhus taxa studied to date is ring porous (with many vessels/mm?’) (Young, Aliso 8:133—146. 1974). In this regard, Actinocheita is similar to Malosma, Metopium, Amphipterygium and most other genera of Rhoeae. Only four genera of Rhoeae (Cotinus, Pistacia, Rhus and 62 MADRONO [Vol. 29 Toxicodendron) have ring porous wood (Heimsch, Lilloa 8:84—198. 1942). Actinocheita also differs from Rhus in having much longer vessel members (x = 489 wm in Actino- cheita; x = 275 wm in Rhus subgen. Rhus and 208 um in R. subgen. Lobadium) and libriform fibers, and in lacking helical sculpturing of vessel walls. Again, in these fea- tures Actinocheita more closely resembles Malosma and Metopium than it does Rhus. Actinocheita differs from Malosma and Metopium in lacking resin canals in its rays. In my opinion, the data presented here support Barkley’s contention, based upon vegetative features, that A. potentillifolia should be segregated from Rhus proper. In terms of its wood anatomy, Actinocheita is as distinct from Rhus as are Malosma and Metopium, and even more so than Cotinus.—DAvID A. YOUNG, Department of Bot- any, University of Illinois, Urbana 61801. (Received 6 Feb 1981; accepted 12 Mar 1981.) MISCELLANEOUS CHROMOSOME NUMBERS IN ASTERACEAE.—Vouchers for the fol- lowing chromosome counts are deposited at MO unless otherwise indicated. Both meiot- ic and mitotic counts were made in material fixed in 3 parts absolute ethanol to 1 part glacial acetic acid and stained in acetocarmine. Anthemis cotula L. 2n = 9 bivalents. San Luis Obispo Co., CA, Raven 20160. Arnica longifolia D. C. Eat. ssp. myriadenia (Piper) Maguire. 2n = ca. 50. Mono Co., CA, Raven 20802. Chrysanthemum coronatum L. 2n = 6 bivalents + ring of 6. San Diego Co., CA, Raven 20168. Erigeron peregrinus (Pursh) Greene ssp. callianthemus (Greene) Cronq. 2n = 9 bi- valents. Tuolumne Co., CA, Raven 20803. Gnaphalium chilense Spreng. 2n = 14 bivalents. Tuolumne Co., CA, Raven 20822. Grindelia hirsutula H. & A. 2n = 12 bivalents. Napa Co., CA, Raven 20200. Leucanthemum vulgare Lam. (Chrysanthemum leucanthemum L.). 2n = 9 bivalents. Tuolumne Co., CA, Raven 20804. Santolina chamaecyparissus L. 2n = ca. 45, up to 9 bivalents, the rest univalents. Cultivated, Los Angeles Co., CA, Raven 20164. Santolina virens Mill. 2n = 9 bivalents. Cultivated, Los Angeles Co., CA, Kyhos 64-087. Sonchus asper (L.) Garsault. 2x = 9 bivalents. Napa Co., CA, Raven 20201. Tanacetum douglasii DC. 2n = 27 bivalents. Humboldt Co., CA, Munz 19869 (prog- eny, RSA 8726), Everett & Balls 18636 (progeny, RSA 8537). Tanacetum huronense Nutt. 2n = 54. Ca. 3.5 mi. sw. of Mackinaw City, Cheboygan Co., MI, Voss (MICH). Studies made by Raven of populations along the coast of northern California and in cultivation at Rancho Santa Ana Botanic Garden, Claremont, in 1961-1962, convinced him that the entities that have been called Tanacetum camphoratum Less. and T. douglasit DC. cannot be distinguished by any constant set of characteristics and should not be regarded as taxonomically distinct at any level. Populations that occur on the coastal dunes of Mendocino and Humboldt Counties are highly variable. This western entity is closely related to 7. huronense Nutt., and probably should not be regarded as specifically distinct either from it or from the circumpolar Tanacetum bipinnatum (L.) Schultz Bip., the first-named species in the complex, which also has 2n = 54. The whole entity is thus evidently hexaploid. A map of the ranges of the included “species” in North America has been given by Mickelson and IItis (Proc. Wisconsin Acad. Sci. 55:200—203. 1966; see also Hultén, Fl. Alaska, p. 892.)—DONALD W. KyHos, University of California, Davis 95616 and PETER H. RAVEN, Missouri Botanical Garden, P.O. Box 299, St. Louis 63166. (Received 23 Feb 1981; accepted 2 Mar 1981.) 1982] REVIEWS 63 REVIEWS Plants of Deep Canyon and the Central Coachella Valley, California. By JAN G. Za- BRISKE. x + 174 p. Univ. of California, Riverside. 1979. Hardbound $14.95; softbound $8.95. Deep Canyon, draining into the western Colorado Desert from the north slope of the Santa Rosa Mountains, is the site of the Philip L. Boyd Deep Canyon Research Center of the University of California, Riverside. This large drainage, from 2657-m Toro Peak to the desert floor, is diverse in topography and vegetation: it is divided here into nine habitat areas, as shown on a foldout contour map in color. After an introduction and a short chapter on the climate, the book discusses the physical setting and the vegetation for each habitat area, giving much interesting information. The appearance of the vegetation is shown by 20 clear photographs. These also help identify a few prominent plants, and individual drawings show a dozen others. An appendix gives more detailed information: a checklist of 619 plants, with altitudinal ranges shown graphically; and data from 22 surveys of perennial plants, at 122-m elevational intervals, each along a 400-m horizontal line. This attractive book will be a useful introduction to the plants for visiting biologists working at the Research Center. And because representative parts of the drainage are crossed by or easily reached from Highway 84 and the Toro Peak road, it will have a wider usefulness, helping and encouraging the botanical public to learn about this interesting broad transect of vegetation. —REID MORAN, Natural History Museum, San Diego, CA 92112. Flora Silvestre de Chile. Zona Central. By ADRIANA HOFFMANN J. 255 p. Ediciones Fundacion Claudio Gay, Santiago, Chile. 1980. (Price not given.) This paperback guide to the wildflowers, vines, shrubs, and trees of the central zone of Chile is a welcome addition to the sparse literature on the interesting flora of that country. The area covered is between Los Vilos at ca. 32° in the north and Constitucion at about 35° in the south, a distance of 400 km. This area, which includes the capital city of Santiago, is referred to as the Matorral Mediterranean Zone. Within this zone there are three floristic regions: the coastal strip (including the coastal ranges), the central valley (mostly cultivated), and an Andean area. This attractive handbook illustrates and describes approximately 450 species of woody and herbaceous plants that occur in this region. The guide is color-coded, a practice that is a familiar one in the North American popular-guide literature. The first part of the book is devoted to woody plants; subsequent sections cover herbaceous plants with white, blue to purple, red or rose, and yellow or orange flowers respectively. The brief descriptions of the species are given on even-numbered pages, and watercolor illustra- tions are grouped opposite these on odd-numbered pages. Though there are no keys, the quality of the illustrations is excellent and should make identifications of unknowns quite easy. In view of the climatic similarities between central Chile and cismontane California, it is not surprising that the two regions share a number of plant introductions from the Old World. Wild radish, scarlet pimpernel, Bermuda buttercup, mullein, cichory, hem- lock, sweet clover, and other exotics abound in both regions. California has contributed Lupinus arboreus and Eschscholzia californica to the Chilean flora; both are extremely common there. Other familiar genera are native to this central zone: there are indigenous Chilean members of Lepechinia, Ribes, Baccharis, Salix, Haplopappus, Collomia, Clarkia, Calandrinia, Phacelia, and other genera also represented in western North America. In addition to these are genera with a Latin American flavor such as Mutisia, Lapageria, Puya, Alstroemeria, Azara, Escallonia, Nolana, and Tropaeolum. The book presents a general discussion of ecological characteristics of the central 64 MADRONO [Vol. 29 zone, origins of common names, an introduction to classification and nomenclature, hints on photographing and pressing plants, a brief history of Chilean botany, and a partially illustrated glossary of terms. Interestingly, “How to use this field guide” ap- pears in English as well as Spanish, and English common names are given for the exotics. The nomenclature used seems to be up to date, though the author still recognizes Godetia and Jussiaea. This guide is thoroughly professional in all respects, and I trust it will prove to be a forerunner of others in Chile and elsewhere in South America.— ROBERT ORNDUFF, Department of Botany, University of California, Berkeley, CA 94720. The California Islands: Proceedings of a Multidisciplinary Symposium. Edited by DEN- NIS M. Power. vii + 787 p. Santa Barbara Museum of Natural History, Santa Bar- bara, CA. 1980. $20.00, paper. The task of organizing and preparing for publication papers presented in any sym- posium is tremendous; but when the symposium is multidisciplinary, as was that on the California Islands held at Santa Barbara Museum of Natural History, Santa Bar- bara, California, from 27 February to 1 March 1978, the accomplishment deserves highest praise. Dennis M. Power is to be congratulated. Over 400 people attended the symposium and 69 papers were on the program. Of these, 43 are included in this published record. The California Islands are delimited as extending from Isla Cedros and Islas San Benitos, Baja California, Mexico (ca. Lat. 28°N) on the south to the Farallon Islands, California (ca. Lat. 38°N) on the north. The papers included in the Proceedings are grouped under the following headings: Geologic History and Paleontology (3 papers); Prehispanic Man (2 papers); Vegetation Changes and the Impact of Feral Animals (5 papers); Evolution and Ecology of Land Plants (5 papers); Biogeography, Evolution, and Ecology of Marine Organisms (14 papers); Biogeography, Evolution, and Ecology of Land Animals (14 papers); Summary. Many of the papers present summaries of early work in addition to treating results of investigations carried on since the Symposium on the Biology of the California Islands organized in 1965 by the Santa Barbara Botanic Garden. Berger’s paper on early man on Santa Rosa Island indicates the presence of man on this island in association with dwarf mammoth remains, a discovery which may push the dates of man on the island, and even in the Americas, back 40,000 years. The impact of man, both prehistoric and modern, and his introductions (fire, goats, cattle, sheep, pigs) on the vegetation is an inter-relationship that is discussed not only in the second and third sections, but also in many papers of other groups. Discussions of the relatively recent plate tectonics theories are tucked into such papers as those by Wenner and Johnson and by Yanev. It would have been useful to have had a paper devoted to this subject alone. In comparison with the number of papers devoted to marine and land animals, those treating plants are disappointingly few. The Baja California islands and the northern islands received comparatively little attention in the Symposium—even less in the zoo- logical papers than in the botanical ones. The majority of the papers are concerned with the Channel Islands. An important part of the volume is the inclusion of references for each paper. Many of the bibliographies are extensive and will be most helpful to future workers. Even though quality of the contributions varies, the Proceedings will serve as a valuable reference work for those interested in the biology of the California Islands. An excellent summary by Dennis Power unifies the volume. He speaks of new re- search tools that have been brought into play in the work represented by the papers presented. He also discusses the distribution and biology of species under such subsec- tions as aboriginal man, recent landscape changes, and conservation. ANNETTA M. CAR- TER, Herbarium, Department of Botany, University of California, Berkeley, CA 94720. 1982] REVIEWS 65 The Road I Came. The Memoirs of a Russian-American Forester. By N. T. MrrRov. An autobiography. 227 p. The Limestone Press, Kingston, Ontario. 1980. ISBN 0-919642- 84-5. This book was authored by Dr. Mirov at the request of his family. It was written in such an interesting fashion that it was felt by the family that it should be shared with the botanical community. It was published in 1978 and is available through Joan Mirov, 453 Tahos Road, Orinda, CA 94563. The first printing was small and the second printing will be priced around $15.00. It is comprised of three parts: The Old World—his early years 1893-1912, college days 1912-1916, war and the revolution 1916-1922, life in China 1920-1923; The New World—dealing in how he became interested in forestry, early experiences in San Fran- cisco, Florida, the Scott Mountains and journeys to Europe and Asia; The Remaining Years—lecture tours, Canada, USSR, New Zealand, and Australia. This book is enhanced by 35 good quality pictures and two maps. It is exciting, amusing, and the reader will find himself experiencing a range of emotions as the vicissitudes of Mirov’s life are encountered. Two memorable anecdotes relate to his introduction to the concept of species while a student in Russia and the beginning of his botanical career as a naturalist on the island of Baikal.— WALTER KNIGHT, Field Associate, California Academy of Sciences, San Francisco, CA 94118. Edward Stuhl’s Wildflowers of Mount Shasta. By Edward Stuhl and Marilyn Clement Ford. Clementine Publishing Co., 123 N. Spring St., Klamath Falls, OR 97601. 1981. This book is oriented to the general public although it will be of interest to professional botanists. The introduction is written by Dr. Wm. Bridge Cooke, author of Flora of Mount Shasta. Of the 650 species of vascular plants in the area, 204 are depicted in full color paintings by Stuhl. The book’s 134 pages include beautiful black and white photographs of the mountain, an account of its plant communities, a descriptive index of the flora, the history of the area, and an account of the botanical explorations there. Hard cover $33.00, soft cover $27.00.—WALTER KNIGHT, Field Associate, California Academy of Sciences, San Francisco, CA 94118. ANNOUNCEMENT On Saturday 24 Oct 1981, Southern California Botanists and the Department of Biological Sciences, California State University, Fullerton, will co-sponsor a symposium on “Cacti and Succulents” in honor of Dr. Lyman Benson. It will be held in the Uni- versity Center Multipurpose Room, California State University, 800 N. State College Blvd., Fullerton. General admission is $10.00, students and Southern California Bot- anists members, $5.00. One unit of college credit is available through CSU Extended Education. For further information, call (714) 773-2611. Program: Ecology and evolution of the Galapagos opuntias (Dr. E. F. Anderson, Whitman College); Cactus alkaloids: chemotaxonomy (Dr. Jerry McLaughlin, Purdue Univ.); Innovative structural designs of cacti (Dr. A. C. Gibson, UCLA); Responses of cacti and succulents to water stress: shifts in metabolism from C; to CAM (Dr. I. P. Ting, UC, Riverside); Cactus chromosomes and hybridization (Dr. D. Pinkava, Arizona St. Univ.); Stonecrop family: variations on a pattern (Dr. Reid Moran, Nat. Hist. Museum, San Diego). Followed by a reception hosted by CSUF. 66 MADRONO [Vol. 29 HERBARIUM NEWS The Herbarium staff of the University of California, Santa Barbara, has initiated a publication series which serves as a vehicle for UCSB students, faculty and staff and other associated investigators. Manuscripts accepted for publication include primarily those with floristics emphasis and those for which voucher specimens are deposited at UCSB. Papers published in this series might not be available otherwise to the general scientific community. This endeavor is consistent with the goals of the UCSB Herbar- ium, summarized as follows: (1) to maintain a collections repository; (2) to provide botanical services; (3) to function as a research facility within the Department of Bio- logical Sciences; and (4) to provide educational programs. UCSB Herbarium Publication Number I, A Flora of Valentine Eastern Sierra Re- serve, is a product of floristic research conducted by UCSB students, and consists of two parts (97 pages): Part I, Valentine Camp, by Ann Howald; and Part II, Sierra Nevada Aquatic Research Station, by Bruce Orr. Valentine Eastern Sierra Reserve (VESR) is managed by the University of California, Santa Barbara, as part of the University of California’s Natural Land and Water Reserves System (NLWRS). Mrs. Edward L. Valentine donated one portion of VESR, Valentine Camp, to the University in 1973. The property consists of 136 acres immediately below the Mammoth Lakes Basin at an elevation of about 8000 ft. (2667 m). This parcel provides an unusually diverse sample of Eastern Sierra habitats on the ecotone between the sagebrush desert of the Great Basin and the coniferous forests of the higher Eastern Sierra. The Sierra Nevada Aquatic Research Station (SNARL) was originally established in 1935 by the Bureau of Sports Fisheries and Wildlife. The SNARL facilities were turned over to the University in 1973. Situated eight miles southeast of the community of Mammoth Lakes, at an elevation of about 2333 m, SNARL’s 53 acres lie just below the steep eastern slope of the Sierra Nevada. High Desert Riparian Woodland, Great Basin Sagebrush and Riparian Meadow Vegetation are plant communities found within the boundaries of SNARL. A limited number of copies of this publication have been made available through a grant from the Valentine Endowment Fund. Requests for complimentary copies should be sent to Wayne R. Ferren, Jr., Senior Museum Scientist, UCSB Herbarium, De- partment of Biological Sciences, University of California, Santa Barbara, CA 93106. THE 1981 JESSE M. GREENMAN AWARD The 1981 Jesse M. Greenman Award has been won by William R. Buck for his publication “A generic revision of the Entodontaceae” (Journ. Hattori Bot. Lab. 48:71- 159. Aug. 1980). This monographic study is based on a Ph.D. dissertation from the Department of Botany, University of Michigan, Ann Arbor. The Greenman Award, a cash prize of $250, is presented each year by the Alumni Association of the Missouri Botanical Garden. It recognizes the paper judged best in vascular plant or bryophyte systematics based on a doctoral dissertation which was published during the previous year. Papers published during 1981 are now being con- sidered for the 15th annual award, which will be presented in the summer of 1982. Reprints of such papers should be sent to: Greenman Award Committee, Department of Botany, Missouri Botanical Garden, P.O. Box 299, St. Louis, MO 63166, USA. In order to be considered for the 1982 award, reprints must be received by 1 July 1982. U.S. POSTAL SERVICE STATEMENT OF OWNERSHIP, MANAGEMENT AND CIRCULATION (Required by 39 U.S.C. 3685) 1, TITLE OF PUBLICATION PUBLICATION NO. 2. DATE OF FILING MADRONO A WEST AMERICAN JOURNAL OF BOTANY kal 613] 7] 10/28/81 3. FREQUENCY OF ISSUE A. NO, 2luts PUBLISHED/B, ANNUAL SUBSCRIPTION Auarterly ANNUAEEY TOUT ERICEYS25...00 4.LO NOWN oO F ICATION (Street, City, Coynty, St d Z[P Code) (Not print - ] Dept. of Bo tany, "Lite. Science Building Bs University ot Cee) ete re eee Alameda County, california 94720 5. LOCATION OF THE HEADQUARTERS OR GENERAL BUSINESS OFFICES OF THE PUBLISHERS (Not printers) Dept. of Botany, Life Science Building, Universit y of California, Berkeley, CA 94720 6. NAMES AND COMPLETE ADORESSES OF PUBLISHER, EDITOR, AND MANAGING EDITOR LISHER (N. d Address) i i i é . . : poeue ee! ere Calieoonta bepkeleye ca shyogT (7° Dept. of Botany, Life Science Bldg. EDITOR (Name ond Address) Christopher Davidson, Idaho Botanical Garden, P.O. Box 2140, Boise, Idaho 83701 MANAGING EDITOR (Name and Address) Same 7. OWNER (If owned by a corporation, its name and address must be stated and also immediately thereunder the names and addresses of stock- holders owning or holding 1 percent or more of total amount of stock, If not owned by a corporation, the names and addresses of the individual owners must be given. If owned by a partnership or other unincorporated firm, its name and address, as well as that of each individual must be giving. If the publication is published by a nonprofit organization, its name and address must be stated.) NAME California Botanical Society (4 corporation) ADDRESS Department of Botany, Life Science Buildin University of California, Berkeley, CA 94720 KNOWN BONDHOLDERS, MORTGAGEES, AND OTHER SECURITY HOLDERS OWNING OR HOLDING 1 PERCENT OR MORE OF TOTAL AMOUNT OF BONDS, MORTGAGES OR OTHER SECURITIES (If there are none, so state) NAME ADDRESS 9. FOR COMFLETION BY NONFSOFIT OHGANIZATIONS AUTHORIZED TO MAIL AT SPECIAL RATES (Section 132.122, PSM) The purpose, function, and nonprofit status of this organization and the exempt status for Federal income tax purposes (Check one) HAVE NOT CHANGED DURING HAVE CHANGED DURING (If changed, publisher must submit explanation of change PRECEDING 12 MONTHS PRECEDING 12 MONTHS with this statement.) AVERAGE NO. COPIES EACH ACTUAL NO. COPIES OF SINGLE 10. EXTENT AND NATURE OF CIRCULATION ISSUE DURING PRECEDING ISSUE PUBLISHED NEAREST TO 12 MONTHS FILING DATE A. TOTAL NO. COPIES PRINTED (Net Press Run) 1100 1100 B. PAID CIRCULATION 1, SALES THROUGH DEALERS AND CARRIERS, STREET -0- ae VENDORS AND COUNTER SALES 2. MAIL SUBSCRIPTIONS 1034 C. TOTAL PAID CIRCULATION (Sum of I1G3BI and 10B2) 1034 1076 D. FREE DISTRIBUTION BY MAIL, CARRIER OR OTHER MEANS SAMPLES, COMPLIMENTARY, AND OTHER FREE COPIES 9g 9 E. TOTAL DISTRIBUTION (Sum of C and D) 1043 1085 F, COPIES NOT DISTRIBUTED 1, OFFICE USE, LEFT OVER, UNACCOUNTED, SPOILED 57 15 AFTER PRINTING 2. RETURNS FROM NEWS AGENTS =0= -0 G. TOTAL (Sum of E, F1 and 2—should equal net press run shown 1100 1100 in A) Bias FGNATORE AND TITLE OF EDITOR, PUBLISHER, BUSINESS MANAGE OR OWNER 12. FOR COMPLETION BY PUBLISHERS MAILING AT THE REGULAR RATES (Section 132.121,/Postal Service Manual) 11. I certify that the statements made by me above are correct and complete. 39 U.S. C. 3626 provides In pertinent part: “‘No person who would have been entitled to mall matter under former section 4359 of this title shall mall such matter at the rates provided under this subsaction unless he filles annually with the Postal Service a written request for permission to mall matter at such rates.” In accordance with the provisions of this statute, | hereby request permission to mall the publication named In Item 1 at the phased postage rates presently authorized by 39 U.S. C. 3626. SIGNATURE AND TITLE OF EDITOR, PUBLISHER, BUSINESS MANAGER, OR OWNER ) veut r dear Sit ¢ eae bt PS porn nee 4 3526 (Page 1) (See instructions on reverse) SUBSCRIPTIONS—MEMBERSHIP Membership in the California Botanical Society is open to individuals ($18 per year; students $10 per year for a maximum of seven years). Members of the Society receive MADRONO free. Family memberships ($20) include one eight-page publishing allotment and one journal. Emeritus rates are available from the Corresponding Secretary. In- stitutional subscriptions to MADRONO are available ($25). Membership is based on a calendar year only. Applications for membership (including dues), orders for sub- scriptions, requests and rates for back issues, changes of address, and undelivered copies of MADRONO should be sent to the Corresponding Secretary. Renewal payments should be sent to the Treasurer. INFORMATION FOR CONTRIBUTORS Manuscripts submitted for publication in MADRONO should be sent to the editor. All authors must be members, and membership is prerequisite for review. Manuscripts and review copies of illustrations must be submitted in triplicate for all articles and short items intended for NOTES AND NEWS. Follow the format used in recent issues for the type of item submitted and allow ample margins all around. All manuscripts MUST BE DOUBLE SPACED THROUGHOUT. For articles this in- cludes title (all caps, centered), author names (all caps, centered), addresses (caps and lower case, centered), abstract, 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, and number each page. Do not use a separate cover page, “erasable” paper, or footnotes. Table captions should include all information relevant to tables. All measurements should be in metric units. Line copy illustrations should be clean and legible, proportioned (including captions) to the MADRONO page, and designed for reduction to *4 original size. 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 latitude and longitude references. Halftone copy should be designed for reproduction at actual size. In no case should original illustrations be sent prior to the acceptance of a manuscript. When needed they should be mounted on stiff cardboard and sent flat. No illustrations larger than 22 X 28 cm will be accepted. Presentation of nomenclatural matter (accepted names, synonyms, typification) should follow the format used for Rhus integrifolia in MADRONO 22:288. 1974. Institutional abbreviations in specimen citations should follow Holmgren and Keuken, Index Her- bariorum, 6th edition. Abbreviations of serial titles should be those in Botanico-Per- iodicum-Huntianum (Lawrence et al., 1968, Hunt Botanical Library, Pittsburgh). If the correct abbreviation cannot be determined, the full serial title should be used. Titles of books should be given in full, together with the place and date of publication, publisher, and edition, if other than the first. All members of the California Botanical Society are allotted eight pages in MADRONO over a two year period. Joint authors are each assessed the full page number. Beyond that number of pages a required editorial fee of $40.00 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 sched- ule, with equity among all members for access to its pages. Printer’s fees for illustrations and typographically difficult material @ $35.00 per page (if their sum exceeds 30 percent of the paper) and for author’s changes after typesetting (@ $3.00 per line will be charged to authors. CALIFORNIA BOTANICAL SOCIETY y i eae T SUC N 198? : ite. yw a4! BRARIES VOLUME 29, NUMBER 2 APRIL 1982 Contents POPULUS X INOPINA HYBR. NOV. (SALICACEAE), A NATURAL HYBRID BETWEEN THE NATIVE NORTH AMERICAN P. FREMONTII AND THE INTRODUCED EURASIAN P. NIGRA, James E. Eckenwalder 67 TAXONOMY OF THE ALLIUM LACUNOSUM COMPLEX (LILIACEAE), Dale W. McNeal, Jr. and Marion Ownbey 79 FLORAL VARIATION IN CHLOROGALUM ANGUSTIFOLIUM (LILIACEAE), Judith A. Jernstedt 87 TAXONOMY AND DISTRIBUTION OF OROBANCHE VALIDA (OROBANCHACEABE), L. R. Heckard and L. T. Collins 95 DICORIA ARGENTEA (COMPOSITAE: AMBROSIINAE), A NEW SPECIES FROM SONORA, MEXICO, John L. Strother 101 ON THE RECOGNITION OF TRICHOSTEMA MEXICANUM EPLING (LAMIACEAE), James Henrickson 104 ENVIRONMENTAL AND COMPOSITIONAL ORDINATIONS OF CONIFER FORESTS IN YOSEMITE NATIONAL PARK, CALIFORNIA, Albert J. Parker 109 NOTES AND NEWS s TAXONOMY OF Lomatium bicolor (UMBELLIFERAE), Mark A. Schlessman 118 SPREAD OF Filago arvensis L. (COMPOSITAE) IN THE UNITED STATES, F. Forcella and S. J. Harvey 119 CLIMATE DIAGRAM FOR THE UNIVERSITY OF CALIFORNIA SAGEHAN CREEK FIELD STATION, Dale E. Johnson 122 NOTEWORTHY COLLECTIONS WEST AMERICAN JOURNAL OF BOTANY CALIFORNIA 123 MoNnTANA 123 WYOMING . 124 <= ANNOUNCEMENTS 125 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, Herbarium, Life Sciences Building, University of California, Berkeley, CA 94720. Subscription rate: $25 per calendar year. Subscription information on inside back cover. Established 1916. Second-class postage paid at Berkeley, CA, and additional mailing offices. Return requested. POSTMASTER: Send address changes to Dr. Frank Almeda, Botany Dept., California Academy of Sciences, San Francisco, CA 94118. Editor—CHRISTOPHER DAVIDSON Idaho Botanical Garden P.O. Box 2140 Boise, Idaho 83701 Board of Editors Class of: 1982—-DEAN W. TAYLOR, University of California, Davis RICHARD VOGL, California State University, Los Angeles 1983—-ROBERT W. CRUDEN, University of Iowa, Iowa City DUNCAN M. PORTER, Virginia Polytechnic Institute and State University, Blacksburg 1984—Mary E. BARKWORTH, Utah State University, Logan HARRY D. THIERS, San Francisco State University, San Francisco 1985—STERLING C. KEELEY, Whittier College, Whittier, CA ARTHUR C. GIBSON, University of California, Los Angeles 1986—AMyY JEAN GILMARTIN, Washington State University, Pullman ROBERT A. SCHLISING, California State University, Chico CALIFORNIA BOTANICAL SOCIETY, INC. OFFICERS FOR 1982 President: WATSON M. LAETSCH, Department of Botany, University of California, Berkeley 94720 First Vice President: ROBERT ROBICHAUX, Department of Botany, University of California, Berkeley 94720 Second Vice President: VESTA HESSE, P. O. Box 181, Boulder Creek, CA 95006 Recording Secretary: ROBERT W. PATTERSON, Department of Biology, San Francisco State University, San Francisco, CA 94132 Corresponding Secretary: SUSAN COCHRANE, California Natural Diversity Data Base, Calif. Dept. of Fish & Game, 1416 9th St. Rm. 1225, Sacramento, CA 95184 Treasurer: CHERIE L. R. WETZEL, Department of Biology, City College of San Francisco, 50 Phelan Ave., San Francisco, CA 94112 The Council of the California Botanical Society consists of the officers listed above plus the immediate Past President, ROBERT ORNDUFF, Department of Botany, Uni- versity of California, Berkeley 94720; the Editor of MADRONO; three elected Council Members: LYMAN BENSON, Box 8011, The Sequoias, 501 Portola Rd., Portola Valley, CA 94025; JOHN M. TuckKER, Department of Botany, University of California, Davis 95616; CHARLES F. QUIBELL, Department of Biological Sciences, Sonoma State Col- lege, Rohnert Park, CA 94928; and a Graduate Student Representative, CHRISTINE BERN, Department of Biology, San Francisco State University, San Francisco, CA 94132. POPULUS x INOPINA HYBR. NOV. (SALICACEAE), A NATURAL HYBRID BETWEEN THE NATIVE NORTH AMERICAN P. FREMONTII S. WATSON AND THE INTRODUCED EURASIAN P. NIGRA L. JAMES E. ECKENWALDER Department of Botany, University of Toronto, Toronto, Ontario M5S 1A1, Canada ABSTRACT Populus x inopina Eckenwalder is a putative natural hybrid between P. fremontii S. Watson and P. nigra L., species of distinct subsections of sect. Azgezros Duby. It has been found in association with both parents along Coyote Creek, Santa Clara County, California. Morphological features and chromatographic profiles of leaf flavonoids of trees growing at this site support a hybrid interpretation of the type tree and suggest possible backcrossing to both parents. Apparent absence of similar hybrids at other localities at which P. nigra has been planted may reflect the unisexuality of the intro- duced trees. While most specimens of P. nigra planted in the range of P. fremontii are staminate individuals, the trees planted at Coyote Creek are pistillate, creating a situ- ation favoring hybridization. Populus fremontii S. Watson (sect. Azgeiros Duby) is a southwestern tree that frequently hybridizes with P. deltoides Marshall (sect. Az- geiros) and with P. angustifolia James and P. trichocarpa Torrey & A. Gray (both sect. Tacamahaca Spach) in their regions of sympatry (Eckenwalder 1977b, in press). No other natural hybrids of this species have previously been reported. Although P. deltoides and P. tricho- carpa are often used as parents in interspecific poplar breeding pro- grams, the small size and poor form of P. fremontii have usually excluded it from such programs (Zsuffa 1975). Stout and Schreiner (1933) reported crossing P. fremontii as female parent with P. del- toides (also the reciprocal cross), P. trichocarpa, P. nigra L. (sect. Aigeiros), and P. X berolinensis Dippel (P. nigra x P. laurifolia Led- ebour (sect. Tacamahaca)), but they apparently considered the hybrids unsuitable for cultivation, presumably discarded them, and did not describe them. Thus hybrids of P. fremontii with sympatric native species have been described, but not those with any exotic species. I found a single locality in California at which P. fremontiz has appar- ently hybridized successfully with the introduced black poplar, P. nigra. MATERIALS AND METHODS The population of cottonwoods containing the putative hybrid lines Coyote Creek in the Hellyer Unit of Coyote Creek County Parkway, San Jose, Santa Clara County, California. The site is developed for recreation but the vicinity of the creek retains its riparian woodland. Some poplar trees are planted in this area (including a row of P. nigra) MADRONO, Vol. 29, No. 2, pp. 67-78, 29 April 1982 68 MADRONO [Vol. 29 but most are apparently spontaneous (including P. fremontii and the hybrids). Presence of saplings as well as mature trees confirms the existence of conditions favoring regeneration. I first visited the site in November 1975 and sampled shoots bearing early and late leaves (Critchfield 1960, Eckenwalder 1977a) and mature winter buds. Shoot cuttings were also made, rooted in sand, and grown in a greenhouse at the University of California, Berkeley. Swelling, staminate flower buds for chromosome counts were collected in March 1976 and pre- served in 3:1 (v/v) ethanol/acetic acid. Shoots bearing male and female flowering catkins were also collected in March and fruiting specimens in April. Reproductive specimens only partly duplicated the original vegetative specimens. Vegetative and reproductive specimens were assessed for characteristics distinguishing P. fremontii from P. nigra and used to construct a hybrid index. Description of the hybrid was based on specimens from two individuals. Quantitative features of the description bracket the normal range and include outlying extremes in parentheses. Following these figures are means and coefficients of variation (mean divided by standard deviation x 100%) determined by measurement of ten randomly selected organs of each type. Chromosome counts were made from pollen mother cell squashes stained in acetocarmine and from aceto-orcein-stained root tip squash- es from plants cultivated in the greenhouse (Smith 1974). The small size of the chromosomes precluded analysis of meiotic pairing behavior in presumptive hybrids. Pollen fertility was assessed using percent stainability of 200 grains per sample by aniline blue in lactophenol, a method that may overestimate fertility (Hauser and Morrison 1964). Diameters of 10 unacetolyzed grains per sample were measured in glycerine jelly mounts. Single leaves of each individual sampled from the population were crushed and extracted in enough acidified (5% HCl) methanol to cover the fragments. Two-dimensional paper chromatograms of each extract were developed on 46 by 57 cm sheets of Whatman 3MM paper in TBA (3:1:1 (v/v) tert-butanol/acetic acid/water) and in 15% aquaeous acetic acid according to the methods of Mabry et al. (1970), which emphasize flavonoids and other polyphenolics. The resulting chro- matographic profiles were compared with reference profiles for P. fremontii and P. nigra obtained by exhaustive extraction of bulk sam- ples (Eckenwalder 1977b). Some compounds of the reference profiles were tentatively identified using chromatographic techniques and by reference to previous studies of flavonoids in the genus (Hegnauer 1973, Crawford 1974, Jones and Seigler 1975). Similarities of chro- matographic profiles to standard P. nigra, ranging from 12 to 58%, were scaled for the individuals of the local population from 0 (P. fremontit) to 100 (P. nigra) to construct a chemical index. Voucher specimens will be deposited at TRT and UC. 1982] ECKENWALDER: POPULUS x INOPINA HYBR. NOV. 69 TABLE 1. COMPONENTS OF A HYBRID INDEX SEPARATING P. fremontii FROM P. nigra USING TEN VEGETATIVE CHARACTERS. Scores of 0 for each character identify character states of P. fremontii, whereas scores of 2 identify those of P. nigra. Inter- mediate states score 1. Typical P. fremontii would total 0 on the hybrid index while typical P. nigra would total 20. States characteristic of Character P. fremontii Intermediates P. nigra Score 0 Score 1 Score 2 Terminal twig thickness >5.0 4.5-5.0 <4.5 Winter bud color greenish tan brown red brown Vesture of bud scales dense thin glabrous Terminal bud scale number 10-12 9-10 7-9 Flowering bud scale number 5-6 4 3-4 Flowering bud curvature straight hooked at tip curved Early leaf tooth number 8-14 15-19 20-25 Lamina width/length ratio 11 1.0—1.1 <1.0 Apex length/width ratio <1.0 1.0—1.6 >1.6 Lamina base shape cordate W-shaped cuneate Total index value 0 10 20 RESULTS A hybrid index based on ten vegetative characters known to sepa- rate P. fremontii from P. nigra (Table 1; Eckenwalder 1977b) clearly groups most sampled individuals with one or the other of the two species (Fig. 1). The putative F, hybrid (Eckenwalder 1132) is the only tree with an intermediate hybrid index score. This individual is inter- mediate between the presumptive parents in leaf form (Fig. 2) as well as in many other features. It has a reduced pollen fertility of 56.5% (Table 2, Fig. 1), close to the 55% average for the comparable Eu- TABLE 2. POLLEN CHARACTERISTICS OF P. fremontii AND P. X inopina FROM COYOTE CREEK AND OF SPECIMENS OF P. nigra ‘ITALICA’ CULTIVATED ON THE UNI- VERSITY OF CALIFORNIA, BERKELEY CAMPUS. Average diameter and standard devia- tion in wm of 10 grains of each sample mounted in glycerine jelly. Percent stainability of 200 grains of each sample mounted in aniline blue in lactophenol (Icb). Collection column includes Eckenwalder collection number. Taxon Collection Diameter Stainability P. nigra ‘Italica’ 1180 23.0 + 1,95 91.0 P. nigra ‘Italica’ 1182 21.2 + 1.48 96.0 P. X inopina 1172 (=1132) 31.9 + 5.69 56.5 P. fremontii subsp. fremontii 1171 (=1127) 23.022 7.31 87.5 P. fremontii subsp. fremontii 1173 2326 = 11.26 98.5 P. fremontii subsp. fremontii 1174 24.5 + 1.68 96.5 [Vol. 29 MADRONO 70 ‘(SIOMO Sp-SE YM SuTyye. / UO paseq %p¢ uoneinjyeur ansded “667[=) ozlI—l ‘egrI—t ‘eel I—y ‘OELI—B :padiu -g (%sS'°9¢ Aynsey uajod ‘¢gry ‘7ZII[=) ZEl[—}J :vuidour x ‘gd ‘TET I—e ‘(slamoy s¢e-SZ YM suTyyed g UO paseq %96 UOTJeIN}eU a[nsded ‘QOZT ‘OSTI=) S82lI—P ‘92TI—? (%S'L8 Asay uatjod ‘TZT7=) ZZII—q ‘SZ [—e :114uowmasf ‘g ‘SiaquINU UOT}IaT[OD JapfemMuUdyq “spoquiAs xas [euoT}UaAUOD Aq PeZI[VIO[ S[ENPIAIPuUI JO Sa1oIg “YaaID 3}0A0D Wor sretdod Jo (esstdsqe) d10IS XIPUI [BITWIGYD SNSI9A (9}JVUIPIO) 9109S xepUl pUqAPE ‘I ‘OI 8109S xapu| jOo|wayy) OOl O06 08 Ol O9 OS Ov O¢ Od Ol O po %96 2 28 ad 11}UOW Ad} “qG ec G “< oO a _hWS:‘9G = DuIdOU! X‘gd } Ol a rs = 7) (o) jo) OsBiu ‘g S! (5 6 Lvs | } Y Oe 71 ECKENWALDER: POPULUS x INOPINA HYBR. NOV. 1982] Wd T = Ieq IBIS ‘OZ psd ‘gd ‘7ETT :vurdour x ‘g ‘$Z—[—yea aye] ‘ZZ, [—yea] Apreo :227UOMaaf gq ‘STOQUINU UOT}IIT[OD JopfeMmMUsyIY “UOTJEUIA JOfeW YM DsSIU “gq pue ‘DUIdoUL x ‘g ‘1JUOMAAL ‘gq JO SAARI] B}e] Pue AIA Jo SBuIeIT, “7 ‘OY 4 6 DaBiu ‘d DUIG OUI LY “J Ttrotiervle sy POG S8AD9| Alida 72 MADRONO [Vol. 29 roamerican hybrid, P. X canadensis Moench (P. deltoides x P. ni- gra; Smith 1943, Eckenwalder 1977b). Its pollen grains are larger than those of either parent, but are also much more variable in size (Table 2), probably reflecting meiotic irregularities (Smith 1943). In contrast to the sharp delineation of the two species and their hybrid by the hybrid index scores, chemical index scores (Fig. 1) are more variable. Although P. X inopina is still nearly intermediate (chemical index 44), the remaining individuals are not as tightly clumped by their chemical index scores as they were by the hybrid index. The scoring of the hybrid index, however, was biased to include most variation in states assigned to the parent species and the ap- pearance of the plants suggested more intergradation than is indicated by their index scores. Morphologically, some individuals of the Coyote Creek population seem like typical representatives of P. fremontii or P. nigra, but chromatographic profiles (Table 3, Fig. 3) show that these individuals contain compounds more typical of the other species than of their own species elsewhere (e.g., compounds 1-12, 42, 43 in P. fremontii and 15-24 in P. nigra). Individuals in this population have only 21 of the 37 compounds known from the reference profiles of P. fremontii (31 compounds) and P. nigra (18 compounds; Eck- enwalder 1977b). The individual of P. X inopina has an additive chromatographic profile typical of interspecific hybrids (Alston and Turner 1962). The predominant flavones of P. fremontiz (compounds 23 and 24) are combined in P. X inopina with the flavonols that dom- inate the profile of P. nigra (compounds 27, 41-43, 51 and 53). This additive profile parallels those found in intersectional hybrids between North American species of sections Aigezvos and Tacamahaca (Craw- ford 1974, Jones and Seigler 1975, Eckenwalder, in press). Individuals identified as hybrids between P. fremontii and P. nigra at Coyote Creek are intermediate between these species in morpho- logical and chemical features. They may be described as: Populus X inopina Eckenwalder, hybr. nov. Arbor verosimiliter ex hybridatione inter P. nigra L. et P. fremontii S. Wats. orta. Inter his species intermedium est in ramulis, gemmibus hyemales colore et formis, staminibus numero, formis laminorumque. Deciduous, presumably dioecious tree of low stature (to 20 m), di- viding near base into separate trunks that ascend to a relatively nar- row, rounded crown. First year twigs coarse, the dominant twigs 4-6 mm thick (x = 5.0 mm, CV = 13%), the lateral twigs 2.5-3.5 mm thick (x = 2.9mm, CV = 11%), tan tinged with orange, turning gray- ish-tan by the third year. Winter buds brown, resinous, sparsely pu- bescent on some scales; terminal buds lance-ovate, (13—)15-17 mm long (x = 15.7 mm, CV = 8%), 4—7 mm thick (x = 5.4 mm, CV = 21%), with (7—)8(—10) bud scales (x = 8.1, CV = 11%); male flowering 13 ECKENWALDER: POPULUS x INOPINA HYBR. NOV. 1982] “MOT[IA —A ‘Buiqiosqe [An—d ‘asue1o—o ‘uveis—3 ‘useizan[q—s3q ‘an[q 14 3I[—q :eIuoUIWe YIM BuUIWIN] Jo}Je SosueYD pure jYSI] JOTOIALI[N UT SIO]OD Jo suoljeusisaq “¢ 3qey ul se paraquinu s}odg “yaed_d 2}0A40— wor DAs7Uu ‘g pue ‘purdout x ‘g ‘“WjUomaaf g Jo WeIZOJEWOIYD aISoduI0D «-¢ “DIA JOJOM / PIDD DIJA9DO / JOUDING-} (ASA) J[24°¢ dy uiBiso G2:0 OS:O Gl:0 O0:| Pe) _ 62:0 On oe (@) Q Cc (qo) °o OS:0 Cc wn (@) (@) (q>) Vey ice (@) oO <. 00°] [Vol. 29 MADRONO 74 G v I 5 aa' pouturexe sTenprAipuy a ate ate + + sprouedoidjAusyd peynueaptun 6 ‘8b ae ait ae (+) + sprouedoidjAusyd peynusptun Le ‘br ai ge + sprouedoidjAusyd paynueptun Zo 9S SS a a + IPISOIATSIP UTZadIaNb eS Ee a IPISOIATSIP [OLBJduIaeyY 1S a ate ae (+) SIpIsodA[ZouOUL [O1BJdWIIeY eb ‘7r + = ze + -—E SapisoodA[souOU UTJadIaNb It ‘22 (+) + ap + UIX9}IA vz (=) ate =i ae unuUsHo oC (+) + + + uIsAIYy9 SI (+) ah zs a8 IUOIATSE [OUDALT ST ats at a (P) Boge cept Cina DABIU ‘J DABIU ‘dq puigdoul x ‘qd 1qUoMadl *d 1quomaal ‘qd UOTIVIYIUIP!I IAILUIT, sjods IUIIIJOY IIUIIIJOY ‘AIGA S1ay}O [[e ‘uoTe[ndod ay} jo sjenprAtput [je ut Juasaid ale 6 pue ‘gp ‘17 spunoduiody ‘juanbasjut punodwiod (+) 4uasaid A[r1e[n3e1 punodwiod + ‘AYAHMASTY SAIOAdS LNAUVG OM], AHL AOA SATIAONY FONTAAITY OL AIAVdWOD AAAAD ALOAOD WOaAA DAsIU ‘gq ANV ‘DUIdoUl x ‘q ‘14UOMad{ ‘'q AO SATIAOUG OIHUVADOLVWOUHD ‘¢ AAV] 1982] ECKENWALDER: POPULUS x INOPINA HYBR. NOV. 1S buds narrowly ovate, asymmetric, spreading from the first year shoots, bent at the tip, (19-)20-—25(-27) mm long (x = 22.6 mm, CV = 12%), (4.5—)6—7(—9) mm thick (x = 6.4 mm, CV = 20%), with 4 bud scales (invariant in 25 buds); female flowering buds unknown. Leaves (Fig. 2) grayish green above, slightly whitened beneath, lacking ba- silaminar glands; early leaves deltoid-ovate with a toothless, acumi- nate, sometimes slightly falcate apex and a distinct flexure in the base in the form of an inverted W, (4—)6.5—8.5(—10) cm long (x = 7.2 cm, CV = 24%), (2.5-)6.5—9(-10) cm wide (x = 7.0cm, CV = 32%), with (12—)15-19 rounded serrations on each side (x = 16.9, CV = 14%), the largest of these in the lower third of the margin, 3-4.5(-—5) mm deep (x = 3.9 mm, CV = 18%), grading to 0.2 mm near the apex; late leaves more rounded, the apex shorter and the flexure of the base less pronounced, (2.5—)5-8.5 cm long (x = 5.9 cm, CV = 32%), (2.5-)5.5—-7.5(-—9) cm wide (x = 6.0 cm, CV = 33%), with (17—)22—26(—37) crenate teeth on each side (x = 24.0, CV = 23%), of irregular size distribution, the largest (0.4—)0.6—0.8(—1.2) mm deep (x = 0.7 mm, CV = 33%), grading down to 0.1 mm between the larg- er teeth; the petiole slightly longer than the blade. Male catkins 3-4(—5) cm long (x = 3.7 cm, CV = 19%), with 35-50 flowers (x = 39, CV = 18%); bracts obovate, laciniate, clawed, brown or tan with a brown fringe, (4-)5—7(—7.5) mm long overall (x = 6.0 mm, CV = 18%), (2.5-)3—4.5(-6) mm wide (x = 4.0 mm, CV = 26%), the claw (0.8—)1.5—2(—3) mm long (x = 1.7 mm, CV = 35%); pedicels longer at the base of the catkin than towards its apex, 0.5—1 mm long (x = 0.8 mm, CV = 33%), elongating to 1-2.5 mm after anthesis (x = 1.7 mm, CV = 36%); floral disc oblique, shallowly conical, 2—2.5(-4) mm across (x = 2.5 mm, CV = 24%); stamens inserted evenly across the disc, 25-50 (x = 37.9, CV = 24%), anthers red at anthesis; pollen spherical, psilate, inaperturate, diameter 31.9 wm + 5.69 wm. Female catkins and flowers unknown. Flowering in March. Chromosome number 2” = ca. 38 (Eckenwalder 1172). Known only from type locality; trees of this parentage produced artificially by Stout and Schreiner (1933) apparently destroyed. Type: USA, CA, Santa Clara Co.: along creek, Hellyer Unit of Coyote Creek County Parkway, San Jose, 6 Nov 1975, Eckenwalder 1132 (Holotype: UC; isotypes: TRT, and to be distributed. ) PARATYPES: Male flowering specimens from type tree: 4 Mar 1976, Eckenwalder 1172; 13 Mar 1976, Eckenwalder 1184; male flowering specimen from another tree: 13 Mar 1976, Eckenwalder 1185. The epithet reflects my surprise at finding this hybrid combination here, when, based on crossing the creek by freeway and seeing the trees from adistance at 90 km/hr, I expected to find a population of P. fremontii, P. trichocarpa, and their hybrid, P. X parryi Sargent. These hybrids are intermediate between their putative parents in many characteristics including growth habit (P. nigra is more erect 76 MADRONO [Vol. 29 than P. fremontiz); twig thickness and color (thicker than P. nigra, more orange than P. fremontiz); bud shape and color (thicker and not as red as P. nigra, without the green cast of P. fremontii); early leaf blade shape (broader, with a more cordate base than P. nigra, apex more elongate than P. fremontiz) and teeth (larger than P. nigra, more numerous than P. fremontiz); time of autumn coloration (far more advanced in early November in P. nigra than in P. fremontii); and stamen number (15-20 in P. nigra, 30-70 in P. fremontiz). If female trees are found, they should also be intermediate in carpel number (2 in P. nigra, 3-4 in P. fremontit) and disc width (2-3 mm in P. nigra, 5-8 mm in P. fremontit). DISCUSSION Natural and artificial interspecific hybridization are well known in Populus (Zsuffa 1975), whose species are all diploid with m = 19 (Smith 1943). Hybrid poplars, whether found in nature or produced artificially, are traditionally given binomia! hybrid designations (Reh- der 1940, Wagner 1970). This convention, which is also common in other woody plant genera, e.g., Quercus (Tucker 1968), is followed here for consistency and in anticipation of the discovery and possible cultivation of additional individuals of the new hybrid. The parents of the hybrid described here are both members of sect. Aigeiros, but they are assigned to distinct subsections (Bugata 1967). The American cottonwoods, P. fremontii and P. deltoides, of subsect. Americanae Bugala are easily distinguished by twig, bud, leaf, flow- er, and fruit characters from the black poplar, P. nigra, of subsect. Euroastaticae Bugata (Eckenwalder 1977a). They have probably had a separate evolutionary history for at least 10 million years, the age of the earliest known record of subsect. Americanae (Eckenwalder 1977b). There is no evidence of pre-Columbian contact between trees of the two subsections, but they were brought together following Eu- ropean settlement of North America. Spontaneous hybridization be- tween P. deltoides and P. nigra occurred in Europe during the eigh- teenth century, giving rise to the widely cultivated P. X canadensis (Boom 1957). Populus fremontit was discovered much later than P. deltoides (Eckenwalder 1977a) and has not apparently hybridized with P. nigra in Europe, where it is rarely planted (Houtzagers 1937). Populus nigra is widely planted in North America but this is the first report of local hybridization between this species and either of the native cottonwoods, P. deltoides or P. fremontii, with which it fre- quently occurs. This paucity of hybrids may be attributed to the sex of the introduced trees. Almost all black poplars planted in North America are from a single staminate clone, the Lombardy poplar (P. nigra ‘Italica’), and its pollen may not be able to compete with pollen of P. deltoides or P. fremontii in fertilizing ovules of these species in 1982] ECKENWALDER: POPULUS x INOPINA HYBR. NOV. ae natural populations (Baker 1951). At Coyote Creek, the trees of P. nigra are pistillate and, without nearby Lombardy poplars, the only seed they can set is from hybridization with P. fremontii (Eckenwalder 1977b). Hybridization would also be promoted by disturbance of the habitat at this site (Anderson 1948). Even so, difficulty in hybridization is reflected in the low percentage of capsules maturing on P. nigra here (34%; Fig. 2 j) and in the low pollen fertility of P. x inopina (56.5%; Fig. 2 f). These figures are comparable to those for male P. X canadensis (55% average pollen fertility for four trees of different cultivars; Smith 1943; Eckenwalder 1977b). All three species and both hybrids are diploids with m = 19 (Smith 1943, Eckenwalder 1977b, 1978) so barriers to crossing are not a product of differences in chro- mosome numbers (Willing and Pryor 1976). The new hybrid may be distinguished from the related P. X canadensis by its ovate, brown, less resinous winter buds (lan- ceolate, reddish, and very resinous in P. X canadensis) and by its early leaves. These are more deltoid than ovate, with coarser teeth, lacking even the sporadic single basilaminar glands of P. X canadensis, with a shorter apex, and without a distinct shoulder in the outline of the lamina base. Overall, P. X canadensis is closer to P. nigra in appearance than is P. X inopina, which more closely resembles its American parent in many respects. Collectors should search for this hybrid and others that may arise between native and introduced trees. As with all poplar studies, collections are most valuable when they are gathered from a single tree over its phenological cycle: flowering shoots in early spring, fruiting branches in late spring, and leafy shoots with early and late leaves and mature winter buds in late summer. ACKNOWLEDGMENTS This study was a portion of a Ph.D. dissertation accepted by the University of Cal- ifornia, Berkeley. I thank R. Ornduff, J. M. Tucker, and D. R. Kaplan for their counsel; S. Eckenwalder and D. E. Johnson for their companionship; S. C. H. Barrett and N. R. Morin for their insistence; anonymous planters for creating a situation con- ducive to hybridization at Coyote Creek; and builders of highway US 101 for indirectly bringing it to my attention. LITERATURE CITED ALSTON, R. E. and B. L. TURNER. 1962. New techniques in analysis of complex natural hybridization. Proc. Natl. Acad. USA 48:130-137. ANDERSON, E. 1948. Hybridization of the habitat. Evolution 2:19. BAKER, H. G. 1951. Hybridization and natural gene-flow between higher plants. Biol. Rev. Cambridge Philos. Soc. 26:302—337. Boom, B. K. 1957. Populus canadensis Moench versus P. euramericana Guinier. Acta Bot. Neerl. 6:54—-59. BuGALA, W. 1967. Systematyka Euroazjatyckich topoli z grupy Populus nigra L. Ar- bor. Kornickie 12:45-219. CRAWFORD, D. J. 1974. A morphological and chemical study of Populus acuminata. Brittonia 26:74-89. 78 MADRONO [Vol. 29 CRITCHFIELD, W. B. 1960. Leaf dimorphism in Populus trichocarpa. Amer. J. Bot. 47:699-711. ECKENWALDER, J. E. 1977a. North American cottonwoods (Populus, Salicaceae) of sections Abaso and Aigeiros. J. Arnold Arbor. 58:193—208. . 1977b. Systematics of Populus L. (Salicaceae) in southwestern North America with special reference to sect. Azgezvos Duby. Ph.D. dissertation, University of California, Berkeley. . 1978. Salicaceae, in IOPB chromosome number reports lx. Taxon 27:225. . In press. Natural intersectional hybridization between North American species of Populus L. (Salicaceae) in sections Aigezvos Duby and Tacamahaca Spach. 2. Taxonomy. Canad. J. Bot. HAUuSER, E. J. P. and J. H. MORRISON. 1964. The cytochemical reduction of nitro blue tetrazolium as an index of pollen viability. Amer. J. Bot. 51:748-752. HEGNAUER, R. 1973. Chemotaxonomie der Pflanzen. Bd. 6. Birkhauser, Basel. HOUTZAGERS, G. 1937. Het geslacht Populus in verband met zijn beteekenis voor de houtteelt. Veenman, Wageningen. JONES, A. G. and D. S. SEIGLER. 1975. Flavonoid data and populational observations in support of hybrid status for Populus acuminata. Biochem. Syst. Ecol. 2:201—206. Masry, T. J., K. R. MARKHAM, and M. B. THOMAS. 1970. Systematic identification of the flavonoids. Springer-Verlag, New York. REHDER, A. 1940. Manual of cultivated trees and shrubs, ed. 2. Macmillan, New York. SMITH, B. W. 1974. Cytological evidence. In A. E. Radford, W. C. Dickison, J. R. Massey, and C. R. Bell. Vascular plant systematics, p. 237-258. Harper and Row, New York. SMITH, E. C. 1943. A study of cytology and speciation in the genus Populus L. J. Arnold Arbor. 24:275-304. Stout, A. B. and E. J. SCHREINER. 1933. Results of a project in hybridizing poplars. J. Heredity 24:216-229. TUCKER, J. M. 1968. Identity of the oak tree at Live Oak Tank, Joshua Tree National Monument, California. Madrono 19:256-266. WAGNER, W. H. 1960. The Barnes hybrid aspen, Populus X barnesii, hybr. nov.—a nomenclatural case in point. Michigan Bot. 9:53—54. WILLING, R. R. and L. D. Pryor. 1976. Interspecific hybridization in poplar. Theoret. Appl. Genet. 47:141-151. ZSUFFA, L. 1975. A summary review of interspecific breeding in the genus Populus L. Proc. Meet. Canad. Tree Improv. Assoc. 14(2):107-123. (Received 30 Jan 1981; revised version accepted 10 May 1981.) TAXONOMY OF THE ALLIUM LACUNOSUM COMPLEX (LILIACEAE) DALE W. MCNEAL, Jr. Department of Biological Sciences, University of the Pacific, Stockton, CA 95211 MARION OWNBEY Department of Botany, Washington State University, Pullman 99163 ABSTRACT A new variety of Allium lacunosum from California, var. kernensis, is described and reduction of A. davisiae to varietal status under A. lacunosum is proposed. A key to and discussion of the four varieties of the species is presented along with a distribution map and chromosome numbers for each, all ” = 7. Frequently in western North America, species limits have been con- fused by lack of representative collections that would allow accurate appraisal of geographic variation. This situation is well illustrated by the Allium lacunosum complex. Allium lacunosum Wats. consists of four distinct geographic vari- eties extending from just north of San Francisco Bay south through the Coast Ranges and east across the Tehachapi Mountains and south- ern Sierra Nevada on to the western Mojave Desert. As described by Watson (1879) the species is restricted to serpentine soils on high ridges and peaks in the Coast Ranges from San Francisco Bay to Santa Barbara County, California. In 1908 Jones described a similar species, A. davisiae, from the western Mojave Desert. Following its original publication this latter taxon was treated as a synonym of A. lacunosum in many floristic treatments of California (Abrams 1923, Jepson 1923, Munz 1935). Ownbey (1959) resurrected A. davisiae, recognizing quantitative and distributional differences from A. lacunosum. Additional taxa with affinities to these two include material de- scribed by Eastwood (1938) as A. lacunosum var. micranthum from the inner South Coast Range and a few anomalous populations, vague- ly referred to by Ownbey (1959) and here described as A. lacunosum var. kernensis McNeal and Ownbey, occurring at the southern tip of the Sierra Nevada. As part of a revision of the Allium acuminatum alliance, of which A. lacunosum and A. davisiae are members (Saghir et al. 1966), we were able to survey a large number of collections of these taxa through the courtesy of the following herbaria: CAS, DS, GH, JEPS, MO, ND, NY, OSC, POM, RSA, UC, US, WS and WTU. Voucher spec- imens and bulbs of the putative taxa were collected over a period of several years. Bulbs were grown in Pullman, Washington and Stock- MADRONO, Vol. 29, No. 2, pp. 79-86, 29 April 1982 80 MADRONO [Vol. 29 ton, California for determination of chromosome numbers. All counts were made on pollen mother cells from fresh buds using aceto-orcein squashes. In addition bulb coats were removed, vapor-coated with silver and examined on an ETEC SEM. The data from these inves- tigations support the conclusion that A. lacunosum consists of four well-defined but closely related geographic varieties. DISCUSSION The varieties of Allium lacunosum closely resemble one another in a series of characteristics that are unique in the Allzum acuminatum alliance. We have elected to take a conservative view of these taxa because of this combination of characteristics rather than to separate them at the specific level on the basis of characters that are quanti- tative in nature. Allium lacunosum has a characteristic bulb coat reticulum with square, polygonal, or + transversely elongate meshes (Figs. 1-4). The walls of these meshes are distinctly sinuous (Fig. 5). Only one other North American species, A. acuminatum Hook., is even superficially similar to A. lacunosum in this regard, and here the walls of the meshes are much thicker and not sinuous (Fig. 6). Presumably, owing to the dry conditions under which all varieties of Allium lacunosum grow, the bulb coats do not disintegrate rapidly. They persist and accumulate, investing the bulb with a thick cover. The leaves of Allium lacunosum var. lacunosum are narrow, con- cave-convex and + falcate, a condition found only rarely in the A. acuminatum alliance. Allium lacunosum var. davisiae is distinctive in having leaves that are falcate and up to 3 mm wide. Allium lacunosum var. micranthum and var. kernensis have leaves that are narrow and straight. In all four varieties the leaves wither from the tip by anthesis; because of this, few herbarium specimens show the differences in shape that have been noted. All four varieties have portions of the ovary densely covered with minute rounded protuberances, a characteristic that is unique in the alliance. Chromosome counts show that all varieties are m = 7 (Table 1), the usual number in North American species of Allium. TAXONOMIC TREATMENT The following treatment is given to emend the circumscription of Allium lacunosum and to delimit its varietal elements. ALLIUM LACUNOSUM Wats. Proc. Amer. Acad. Sci. 14:231. 1879. For synonymy and typification see the varietal headings. Bulbs ovoid, 1—2 cm long, coats brown to yellow brown, a new coat developing annually, old coats accumulating and investing the bulb 1982] McNEAL AND OWNBEY: ALLIUM LACUNOSUM COMPLEX 81 Fics. 1-6. Scanning electron micrographs of Allium bulb coat reticulations. 1. Al- lium lacunosum var. lacunosum. 2. Var. micranthum. 3. Var. kernensis. 4. Var. da- visiae. 5. Detail of sinuous cell wall in var. lacunosum. 6. A. acuminatum. (Scale: Figs. 1-4 & 6 = 100 um; Fig. 5 = 10 pm) 82 MADRONO [Vol. 29 TABLE 1. CHROMOSOME NUMBERS FOR Allium lacunosum VARIETIES. All collec- tions are from California; vouchers are in WS unless otherwise indicated in parentheses. Our counts were all made on microsporocytes during first meiotic metaphase. * Indi- cates previously unpublished counts by Dr. Hannah C. Aase. Variety n Collection lacunosum ba Tiburon Hills, Marin Co., 22 Apr 1947, Hoffman S.N. es 0.5 mie. of Belvedere, Tiburon Peninsula, Marin Co., 13 Oct 1947, Kruckeberg s.n. micranthum 7 Above the entrance station e. side of Pinnacles Natl. Mon., San Benito Co., 20 Apr 1970, McNeal 496 (CPH). kernensis 7 Bodfish-Havilah Rd., 5.2 km se. of Hwy 178, Kern Co., 18 Apr 1970, McNeal 492 (CPH). has Road from Kernville to Glennville, e. slope of Greenhorn Mts., Kern Co., 12 May 1949, Munz 13158. davisiae 7 5.5 kms. of Mojave on Hwy 14, Kern Co., 6 Apr 1973, McNeal 1243 (CPH). ye 3.2 km w. of Palmdale, Los Angeles Co., 10 Aug 1947, Epling s.n. 7 Stoddard Wells Rd., 28.1 kms. of Barstow, San Bernardino Co., 7 Apr 1973, McNeal 1247 (CPH). with a thick cover, coats distinctly cellular-reticulate, the meshes square, polygonal or + transversely elongate with sinuous walls; leaves 2, 0.75—2 times as long as the scape, withering from the tip by anthesis; scape terete, 10-35 cm tall; bracts 2 or 3, lance-ovate to ovate, 5-15 mm long, becoming papery; umbel 5-45 (or more)-flow- ered, pedicels 1—-3.5 times as long as the flowers, each pedicel with its flower deciduous as a unit at maturity; perianth segments white or pale pink with darker mid-nerves, spreading, lance-ovate to ovate, obtuse, acute or short acuminate, 4-9 mm long and 14 mm wide; stamens 0.7—-0.8 as long as the perianth, anthers yellow, rounded at the apex; ovary 3-lobed, 3-grooved with a ridge on either side, the ridges prolonged above into a crest with 3 minute, 2-lobed central processes, ridges and crest densely covered with minute rounded pro- tuberances; stigma capitate, scarcely thickened, obscurely 3-lobed; seeds black, alveoli minutely roughened; 1 = 7. Key to the varieties of Allium lacunosum a. Umbels compact, pedicels 0.75—1.5 times as long as the flowers, scapes 10-25 cm tall 1982] McNEAL AND OWNBEY: ALLIUM LACUNOSUM COMPLEX 83 b. Leaves 1-2 times as long as the scape; scape 10—20 cm tall; flowers 7-9 mm long. Coast Ranges from Marin to Santa Barbara Co. and on Santa Rosa and Santa Cruz Islands .... PO es ee fee lake conn Maer ee hee hee ee ai 1. var. lacunosum bb. Leaves equalling the scape or shorter; scape 15-25 cm tall; flowers 6—8 (usually less than 7) mm long. East of the Green- horn Mts. in the vicinity of Lake Isabella, Kern Co. ...... OF oe a Ne en 2. var. kernensis aa. Umbels loose, pedicels spreading 1.5—3.5 times as long as the flowers; scapes 15-35 cm tall. c. Bracts 3, obtuse or acute; leaves subterete, 1 mm or less wide, straight; flowers 4-6 mm long. Inner Coast Ranges in San Benito, southern Monterey and northern San Luis Obispo BOS ate tek AO eco agiat ea 8 ae 3. var. micranthum cc. Bracts 2 acuminate; leaves flattened, up to 3 mm wide, + falcate; flowers 6-8 mm long. Carriso Plains, San Luis Obispo Co. and sporadic southward to San Bernardino and San Die- COM COSaea, ae ne ee ee eee ee es Le 4. var. davisiae 1. ALLIUM LACUNOSUM Watson var. LACUNOSUM. Proc. Amer. Acad. Sci. 14:231. 1879.—Typr: USA, CA, Santa Clara Co., summit of Mariposa Peak. 20 Jun 1862. Brewer 1284. (Holotype: GH!; isotypes: MO! UC! US)). Distribution. Tiburon Peninsula in Marin Co., south through the Coast Ranges to Santa Barbara Co. and Santa Rosa and Santa Cruz Islands, CA (Fig. 7). Apparently restricted to serpentine soil on ridges and peaks at 300-1000 m. Flowering late Apr—early Jun. This taxon has a smaller stature than any of the other varieties, its scapes varying from 10(—7) to 20(—25) cm tall. The leaves are narrow and + falcate. Umbels are compact and the ratio between flower length and pedicel length ranges from 0.75 to 1.5. The perianth seg- ments are white with deep pink or purple mid-nerves. Allium lacunosum var. lacunosum intergrades with var. davisiae in the Tehachapi Mountains, the closest point of contact between the two. 2. Allium lacunosum Watson var. kernensis McNeal and Ownbey, var. nov. Scapo 15-25 cm longo; umbella 10-45 (vel pluribus) floribus; seg- mentis perianthii 5—7(—8) mm longis, albis vel roseis, nervo medio viridi vel purpureo-viridi. Type: USA, CA, Kern Co., 1.5 km w. of Kernville Rd. on the rd. to Glennville, e. slope of the Greenhorn Mts., 12 May 1949, Munz 13158. (Holotype: POM)). 84 MADRONO [Vol. 29 o var. davisiae = var. kernensis. e var. lacunosum + var. micranthum Fic. 7. Distribution of the varieties of Allium lacunosum. Distribution. Rocky, sandy or clay soils, southern tip of the Sierra Nevada from Kernville and Pilot Knob south to Piute Mt., Kern Co., CA (Fig. 7). Flowering May—Jun. Variety kernensis is a well marked geographic variant. Specimens of this taxon are, on the average, taller than the typical variety and their leaves are narrow and straight. The flowers are somewhat small- er and usually more numerous, the ratio of flower length to pedicel length ranges from 1 to 2, and the umbels are compact. The mid- nerve of the perianth segments is commonly green in contrast to the deep pink or purple of var. lacunosum. 3. ALLIUM LACUNOSUM Watson var. MICRANTHUM Eastwood, Leafl. W. Bot. 2:101. 1938.—Type: USA, CA, San Benito Co., The Pinnacles. 3 May 1937. Eastwood & Howell 4231. (Holotype: CAS!; isotypes: GH! POM! US)). Distribution. Rocky clay and serpentine soils, Inner Coast Range of California in San Benito, southern Monterey and northern San Luis Obispo Cos. (Fig. 7). Flowering late Apr to May. 1982] McNEAL AND OWNBEY: ALLIUM LACUNOSUM COMPLEX 85 Variety micranthum differs from typical Allium lacunosum in being taller with straight, narrow leaves. The involucre commonly has three bracts, whereas two is the usual number in the other varieties. The flowers are the smallest in the complex, and the ratio of flower length to pedicel length ranges from 1.5 to 3.5. The umbels are loose. 4. Allium lacunosum Wats. var. davisiae (Jones) McNeal & Own- bey, comb. nov.—Allium davisiae Jones, Contr. W. Bot. 12:78. 1908.—TyYPE: In his protologue Jones states that this species “rows in gravelly soil and among rocks at Victor [now Victorville], California.” 8 May 1903 at 2900 ft (880 m). No specimens with data exactly conforming to this statement were found; however, specimens bearing slightly different data seem to be the collections he had in mind inasmuch as he annotated three of them as types. Two of these specimens are: California, San Bernardino Co., Vic- tor, 2600 ft (790 m). 17 May 1903 Jones s.n. (US! MO!). The third bears the same label data but is dated 18 May 1903 (US)). We assume that the differences may be accounted for as mental lapses or carelessness on Jones’ part and have concluded that the material dated 17 May and annotated as types by Jones should be interpreted as type material. Hence, we designate the US sheet #855767 as lectotype. Several herbaria have duplicates of it (CAS! DS! GH! MO! POM! WS! WTU!) which, with the exception of MO, are identified by Jones as various other species. Allium pseudobulbiferum Davidson, Bull. So. Calif. Acad. Sci. 20:49. 1921.—TyYPE: Davidson 3410, collected by Kessler from elevated ground east of the river at Victorville, San Bernardino Co., CA, 1 May 1921 (LAM!). This is an immature and fragmentary spec- imen consisting of a partial scape and an umbel. It corresponds closely with specimens of var. davisiae at the same stage of de- velopment. Distribution. Rocky, sandy, or clay soil, Mohave Desert, south to the San Jacinto Mts., Riverside Co. and northern San Diego Co., extending northwest to the Carriso Plains, San Luis Obispo Co. (Fig. 7). Flowering Apr—Jun. Variety davisiae differs from typical A. lacunosum in being, on the average, considerably taller. Specimens usually bear a larger number of somewhat smaller flowers, the ratio of flower length to pedicel length ranges from 2 to 3, and the umbels are loose. Collections from northeast Ventura Co. and from the San Jacinto Mts. appear to be intermediate between var. davisiae and the typical variety. The flow- ers in these are slightly larger than var. davisiae; however, because of the stature of the plants and the length of the pedicels, both of which considerably exceed var. lacunosum, the specimens are placed here. 86 MADRONO [Vol. 29 LIST OF EXSICCATA More than 100 herbarium specimens were examined during this study. Along with field observations these form the basis for our morphological and distributional data. Lists of these specimens are available from the senior author. ACKNOWLEDGMENTS We thank San Joaquin Delta College, Stockton, CA, for use of their scanning electron microscope in the course of this study. We also thank Dr. Robert Smutny of the Classics Department, University of the Pacific, for assistance in preparing the Latin diagnosis of Allium lacunosum var. kernensis and Dr. Hanna C. Aase for permission to use some of the chromosomal data in this paper. We appreciate critical reviews by Drs. E. F. Anderson and L. V. Mingrone. Dr. Marion Ownbey died 6 Dec 1974. Much of the research leading to this paper was done as part of my Ph.D. dissertation, for which he was the adviser and an active participant in the field work and in the taxonomic decision-making process. LITERATURE CITED ABRAMS, L. R. 1923. Illustrated Flora of the Pacific States, Vol. I. Stanford Univ. Press, Stanford, CA. EASTWOOD, A. 1938. Allium lacunosum Wats. var. micranthum. (Footnote to: Howell, J. T. A botanical visit to the Vancouver Pinnacles.) Leafl. W. Bot. 2:101. Jepson, W. L. 1923. A manual of the Flowering Plants of California. Calif. School Book Depository. San Francisco. JONES, M. E. 1908. New species and notes. Contr. W. Bot. 12:1-81. Muwz, P. A. 1935. Manual of Southern California Botany. Claremont Colleges, Clare- mont, CA. OwnsBEY, M. 1959. Allium. In P. A. Munz. A California Flora. Univ. California Press, Berkeley. SAGHIR, A. R. B., L. K. MANN, M. OwnBEY, and R. Y. BERG. 1966. Composition of volatiles in relation to taxonomy of American Alliums. Amer. J. Bot. 53:477-484. WATSON, S. 1879. Contributions to American Botany IX. Revision of the North American Lilaceae. Proc. Amer. Acad. Arts. 14:213-288. (Received 27 Jan 1981; revised version accepted 14 May 1981.) FLORAL VARIATION IN CHLOROGALUM ANGUSTIFOLIUM (LILIACEAE) JUDITH A. JERNSTEDT Botany Department University of Georgia, Athens 30602 ABSTRACT Four northern California field populations of Chlorogalum angustifolium consist of plants of two floral types: short-styled and long-styled. In short-styled plants the stigma level is below the tier of anthers; in the long-styled plants it is above the anthers. In the Butte County population studied in greatest detail, long-styled plants are twice as nu- merous as short-styled plants. The two forms do not differ significantly in anther size, number of pollen grains per anther, filament length, number of flowers per inflorescence, or height and degree of branching of the inflorescence. The two forms do differ signif- icantly in ovary and tepal lengths. Approximately 36 percent of both forms bore fruit in the field, although the average number of fruits per plant was 60 percent higher on short-styled plants (x = 1.60) than on long-styled (x = 0.99). No incompatibility exists within or between the two forms; however, seed-set was significantly lower for selfed long-styled plants than for selfed short-styled ones. No long-styled plants set seed with- out mechanical transfer of pollen, whereas up to 30 percent of flowers of short-styled plants did so. It is suggested that the presence of short-styled plants in populations of C. angustifolium insures some degree of seed production during times when pollinators are scarce. The five species and three varieties of Chlorogalum Kunth. (Lili- aceae) are remarkably uniform in most floral and vegetative charac- ters, e.g., flower size, anther color, bulb coat morphology, and inflo- rescence size (Hoover 1940). However, several populations of C. angustifolium, the second most widespread species, were found to consist of plants having two readily distinguishable floral types: short- styled and long-styled. This study investigated selected morphological characters of long- and short-styled plants of C. angustifolium Kell. from field-collected samples, including experimental pollinations to determine incompatibility between and within the two forms. Prelim- inary studies were made of degree of autogamy and floral senescence in this species. MATERIALS AND METHODS Four field populations (Palermo and Chico, Butte County, and Freeman School and Corning Rd., Tehama County) of Chlorogalum angustifolium in northern California were observed to consist of long- and short-styled plants. Bulbs collected from two of these populations (Palermo and Freeman School) were cultivated and brought to flower in the greenhouse. Artificial crosses between 288 flowers on 27 plants were made using fine forceps which were cleaned with 95 percent ethanol between pollinations. To avoid contamination of stigmatic MADRONO, Vol. 29, No. 2, pp. 87-94, 29 April 1982 88 MADRONO [Vol. 29 surfaces during emasculation, only newly opened flowers with unde- hisced anthers were used in pollinations. Within an hour after anthe- sis, masses of pollen were placed on the three minute stigmatic lobes. No distinction was made between the two whorls of stamens, which occur at the same level in both long- and short-styled forms of this species. Seeds from the crosses were harvested, counted and weighed at the time of capsule dehiscence, about 6 weeks after pollination. Flowers and buds were collected at random from the Freeman School population and preserved in 50 percent formalin-acetic acid- ethanol (FAA). Floral parts were measured using Mitutoyo dial cali- pers at 15X magnification under a dissecting microscope. Pollen pro- duction of long- and short-styled forms was compared by suspending pollen from two adjacent undehisced anthers of preserved flowers in measured volumes of 50 percent ethanol. After thorough mixing, 10 jl samples of the suspension were placed in an “Ultra-Plane Spot Lite” counting chamber and allowed to settle for 60 seconds. Sixty-five grids were counted for each of 10 long-styled and 10 short-styled sample dilutions. At the peak of the flowering season, a 50-m line transect was run through the center of the Palermo population, and adjacent inflores- cence axes were collected every 1 m along the line. These axes were measured and the number of flowers, fruits, and seeds per axis deter- mined. Differences between the two means of measurements of various parameters of the two forms (tepal length, height to first branch, num- ber of seeds per fruit, etc.) were compared using the Student’s ¢ test (Mendenhall 1971). Surface features of pollen and stigmas were examined using a Cam- bridge Stereoscan S-4 scanning electron microscope after the tissues were fixed in FAA, dehydrated through a graded ethanol series, crit- ical point dried in amyl acetate, and coated with gold. RESULTS Inter- and intrapopulation crosses using plants from two hetero- morphic populations of Chlorogalum angustifolium indicate that no incompatibility exists between long- and short-styled plants of this species. The percentage of seed-set was not significantly different in the classesS xX S,S x L,L x S,L XL, and S-selfed (Table 1). Non- emasculated flowers of the long-styled form that did not have pollen mechanically transferred to the stigmas (class L-unpollinated) failed to produce seeds, whereas 30 percent of unpollinated, non-emasculat- ed flowers of short-styled plants (S-unpollinated) set seeds. The av- erage number of seeds per fruit in cross classes S X S,S XL,L XS, L X L, S-selfed, and S-unpollinated did not differ significantly (Table 1). Only 6.3 percent of crosses of the class L-selfed resulted in seed 1982] JERNSTEDT: FLORAL VARIATION IN CHLOROGALUM 89 TABLE 1. RESULTS OF POLLINATIONS IN Chlorogalum angustifolium. Number of crosses Cross class Total resulting Seed weight Number of seeds (female parent number ineseed (mg) per fruit x male of eee parent) crosses n % X n x S n SxS 44 30 68.2 4.2 95 3.80 1.27 30 Sx i 49 34 69.4 4.6 118 3.47 1.48 34 ex 5 19 13 68.4 4.6 41 325 1.09 13 Lex 33 18 54.5 4.5 57 3.41 128 17 S-selfed 14 10 71.4 4.4 39 3.90 1410 10 L-selfed 16 1 6.3 3.6 2 2.00 — 1 S-unpollinated 60 18 30.0 4.3 69 3.83 1.04 18 L-unpollinated 53 0) 0 — 0) 0) — 0) production, with only two seeds produced in that single successful cross. Average weights per seed in the classes S XS, S XL, L XL, S-selfed, and S-unpollinated were nearly identical, and the departure of seeds from L-selfed from this value may be a result of the small number of seeds available for measurement. Of 109 plants of C. angustifolium sampled along a 50-m transect through the Palermo population, 36 were short-styled and 73, long- styled. Mean measurements of overall inflorescence height, height to first branch of the inflorescence, number of inflorescence branches, number of flowers, number of fruits, and number of seeds of the two forms are shown in Table 2. Long-styled and short-styled plants differ significantly (p < 0.05) only in average number of fruits per plant. In the field, 35.7 percent of long-styled plants and 37.1 percent of short- styled plants bore fruit at the time of sampling. Mean measurements of selected floral characters of the two forms from plants of the Freeman School population are shown in Table 3. Long-styled and short-styled plants differ significantly (p < 0.01) in ovary, style, and tepal lengths. Scanning electron microscope studies of pollen and stigmatic sur- faces of long- and short-styled plants revealed no differences between the two forms. Size and surface ornamentation of pollen grains were identical in long- and short-styled plants, and stigmatic surfaces of both forms are wrinkled at the margins and covered with a smooth layer of unornamented cuticular material along the center of the three minute lobes. DISCUSSION Four natural populations of Chlorogalum angustifolium in northern California consist of plants of two floral types: short-styled and long- [Vol. 29 MADRONO 90 OT €SO0 q802 Ol €7O qivtT Of O60T ¢°?€ O09 90 9FS O09 TSO 282 O9 @TT gqIl'8 pefAys-yz0ys OT LTT q6S'S OT 8€0O0 q80°2 OT Cor $62 09 8270 L465 09 80 +92 09 19°0 4966 pafAjs-Su0'T u s x u s x u S Xx u s x u s x u s X ULIOJ ae = So a eS ee Se eS [e10[ J (UI) (ww) uonn{Ip ajdwes (UU) (wut) (WI) yysug] 34S yysua] AIVAC I” OT Ul JayjQUe Yysua] JUDWIe]L yysug] JayjUYy yysuag eday, Jad surei3 uaTjod raquinyy [PAZ] TO" 9} 3 AazJIp q dt19sredns Aq pamo][oj suvau paiteg ‘VINAOAITVD ‘0D VNVHAL “IOOHOS NYWaaay “wnzi0fysn3un mnjDps041014) 10 NOILVINdOd OIHdMOWONALAH V WOXA SWAOT TVAOTY HLOG dO SYALAWVAVG GALOATAS JO SLNAWAANSVATL “6 ATAV £9 680 e€8T Sse 9% eO9T SE 8242 LLvS SE L7T €OF SE OFS HL? SE £06 >¥2'0S patAqs-310YS c9 «6680, 0% OL OB T «6660 «360k 8st Opse€S) «6(OL SCT TT LOH SOOL LES Ss ON'HZ-sCOOLsC'Bs«BB'9H pe]Ajs-3u0'T u S x u S x u s x u S x u S X u S X WO [BIOL J ynaj rad queld sad SIIMOPY JaquUINN sayoueiqg JaquInNN, (WI) YouRIq (UID) Spses Joquinny s}indj Jaquinny JsIY 0} JYSIOAY JYSIIy [[e1vAO [PAT SO’ 94} 3 AazFIp & ydiz9siadns Aq paMmo][oj Suva patleg “‘VINUOAITVD ‘OD ALLA ‘OWUAIV ‘wnyofiusn3uD mnjvs0L0]1Y,) 10 NOILVINdOg DIHdMOWOAALAH V WONT SWAOT TVAOTY HLOG 4O SUALAWVAVG GALOATAS AO SLNAWIANSVAW 7 ATAVE 1982] JERNSTEDT: FLORAL VARIATION IN CHLOROGALUM 91 I Fic. 1. Chlorogalum angustifolium. Long- and short-styled forms, with parts drawn to scale; see Table 3 for mean measurements. X2. styled (Fig. 1). In short-styled plants the stigma level is below the tier of anthers. In the long-styled form, it is above the height of the an- thers. No further differences exist between the two forms in terms of pollen size and shape, stamen height, stigma surface characters or pollen production. Furthermore, plants of the two forms do not differ significantly in the average overall height of the inflorescence, the degree of branching of inflorescence, number of flowers per inflores- cence, or number of seeds per fruit (Table 2). In addition to the dif- ference in style lengths between the two forms, tepal and ovary lengths are greater in long-styled flowers than in short-styled. However, while the measured differences in these two characters are statistically sig- nificant, in the author’s experience it has not been possible to distin- guish between long- and short-styled plants on the basis of tepal and/ or ovary lengths alone. The results of crossing experiments (Table 1) do not indicate the presence of incompatibility in this species within or between the two forms. The decreased seed-set in the cross class L-selfed suggests some degree of self-incompatibility in this form, which appears to be lacking in short-styled plants. Alternatively, the low seed-set for L-selfed crosses could be a result of inbreeding depression (Lloyd 1968). The S-selfed crosses do not show a similar reduction in seed-set. If inbreed- ing depression is in fact present in long-styled individuals, these results may be indirect evidence that in the field long-styled plants are more outcrossed than short-styled. The results of experiments in which long- and short-styled flowers of C. angustifolium were allowed to self spontaneously provide insight into the possible function of different style lengths in this species. Without artificial transfer of pollen from anthers to stigma, long-styled plants of C. angustifolium were never able to produce seeds, whereas 30 percent of flowers of greenhouse-grown short-styled plants set seeds (Table 1). This ability would insure some degree of seed production by short-styled plants in the event of pollinator failure, and could 92 MADRONO [Vol. 29 perpetuate the short-styled form. In light of this, however, it is some- what difficult to explain why long-styled plants predominate in the Palermo population by a 2:1 ratio. On the other hand, short-styled plants along the Palermo transect average approximately 60 percent more fruits per plant than the long-styled plants (Table 2). This dif- ference in fecundity could result from self-pollination of short-styled plants, although there is no evidence that the short-styled plants are self-pollinated in the natural habitat and no reason to suspect that pollination rates limit fruit-set in the field. Interestingly, the same percentage of long- and short-styled plants along the transect bore fruit at the time of sampling (35.7 percent and 37.1 percent, respec- tively). The mechanics of flower closing in this species suggest a means by which self-pollination may be accomplished. Floral senescence in C. angustifolium, which is similar to that reported for C. pomeridianum (Jernstedt 1980), occurs 6—8 hours following anthesis. As senescence progresses, the tepals uncurl and come together to close the flower. The filaments also collapse downward toward the ovary. This posi- tions the anther sacs directly on the stigmas of the short-styled plants, potentially resulting in self-pollination. A similar sequence of events occurs in long-styled flowers. However, because in this form the stig- ma already projects above the anther sacs, collapse of filaments of long-styled plants serves to remove anthers farther from the stigmatic surface, seemingly precluding the possibility of pollen transfer and self-pollination in this form. Results of crossing experiments suggest, however, that additional factors may be involved in the regulation of cross- and self-pollination in this species. If the collapse of anthers upon senescence were suffi- cient to cause selfing in short-styled plants of C. angustifolium, it seems likely that greater than 30 percent of S-unpollinated crosses would have resulted in seed production (Table 1). Likewise, the per- centage of seed-set in all cross classes is far from 100 percent, despite the liberal coating of stigmatic surfaces with pollen at the time of pollination. Pollen viability, style-stigma interactions, time of gamete maturation, and resource limitation could be involved, and further work is necessary to understand the control of seed-set in this species. The floral variation described here for Chlorogalum angustifolium has a superficial resemblance to the phenomenon of heterostyly, al- though a close examination shows them to differ significantly. Het- erostyly is a genetically controlled floral polymorphism in which the two or three forms of a heterostylous species produce flowers that differ reciprocally in style and stamen lengths (Baker 1964, Mulcahy 1965, Vuilleumier 1967, Ganders 1979). The floral polymorphism is usually genetically linked with a sporophytic self-incompatibility sys- tem (Ganders 1979), and the syndrome of morphological and physio- logical features is thought to be governed by a supergene (Barrett 1982] JERNSTEDT: FLORAL VARIATION IN CHLOROGALUM 93 1979). In its typical forms, heterostyly is clearly a breeding system that evolved to enforce outcrossing (Baker 1964, Mulcahy 1965, Vuilleu- mier 1967, Barrett 1977, 1978; Ganders 1979). In contrast, differences in style lengths in long- and short-styled forms of C. angustifolium are not accompanied by stamen length variation, nor is an incompatibility system found. Instead of increasing the efficiency of cross-pollination, as in classic heterostyly, the opposite seems to be the case in C. an- gustifolium, with the presence of short-styled individuals perhaps in- creasing the likelihood of selfing in the field. An understanding of the genetic basis for this floral variation must await data on seed germi- nation, seedling vigor, and floral morphology of progeny from the experimental crosses. An analogy perhaps more appropriate than that of heterostyly may be found in the development of modern tomato varieties (Rick and Dempsey 1969, Rick 1976). Although cultural conditions are known to affect style length in tomatoes (Howlett 1939), modern cultivars have come to possess very short styles, included within the anther tube, as a result of artificial selection in breeding programs. This change in stigma position has increased self-pollination and conse- quently, fruit-set, and has resulted in very low rates of outcrossing (Rick 1976, Rick et al. 1977, 1978). It may be that natural selection has resulted in a similar situation in Chlorogalum angustifolium. Fur- ther studies of the breeding sytem in this species could shed light on this process, and comparison with other species in the genus, which are not known to exhibit such floral variation, might prove instructive. ACKNOWLEDGMENTS This study represents a portion of a dissertation submitted in partial fulfillment of the requirements for the Ph.D. degree at the University of California, Davis. Support was provided in part by NSF grant DEB 77-12655. Thanks are extended to Drs. G. L. Webster, T. L. Rost, B. A. Bonner, and R. E. Wyatt for review of an earlier draft of this manuscript, to Drs. D. W. Kyhos and Curtis Clark for technical assistance and advice, and to two additional reviewers for their useful comments. LITERATURE CITED BAKER, H. G. 1964. Variation in style length in relation to outbreeding in Mirabilis (Nyctaginaceae). Evolution 18:507-512. BARRETT, S. C. H. 1977. The breeding system of Pontederia rotundifolia L., a tristy- lous species. New Phytol. 78:209-220. . 1978. Floral biology of Eichhornia azurea (Swartz) Kunth (Pontederiaceae). Aquatic Botany 5:217-228. . 1979. The evolutionary breakdown of tristyly in Eichhornia crassipes (Mart.) Solms. (Water Hyacinth). Evolution 33:499-5 10. GANDERS, F. R. 1979. The biology of heterostyly. New Zealand Journal of Botany 17:607-635. Hoover, R. F. 1940. A monograph of the genus Chlorogalum. Madrono 5:137-147. HowLeETT, F. S. 1939. The modification of flower structure by environment in varieties of Lycopersicon esculentum. J. Agric. Res. 58:79-117. 94 MADRONO [Vol. 29 JERNSTEDT, J. A. 1980. Anthesis and floral senescence in Chlorogalum pomeridianum (Liliaceae). Amer. J. Bot. 67:824—-832. LLoyp, D. G. 1968. Pollen tube growth and seed set in self-incompatible and self- compatible Leavenworthia (Cruciferae) populations. New Phytol. 67:179-195. MENDENHALL, W. 1971. Introduction to probability and statistics. 3rd edition. Wads- worth, Belmont, CA. Mutcany, D. L. 1965. Heterostyly within Nivenia (Iridaceae). Brittonia 17:349-351. Rick, C. M. 1976. Tomato, Lycopersicon esculentum (Solanaceae). In N. W. Sim- monds, ed., Evolution of crop plants, p. 268-273. Longman, London. and W. H. DEMPSEY. 1969. Position of the stigma in relation to fruit setting of the tomato. Bot. Gaz. 130:180—186. , J. F. FoBes, and M. HOLLeE. 1977. Genetic variation in Lycopersicon pim- pinellifolium: Evidence of evolutionary change in mating systems. Pl. Syst. Evol. 127:139-170. , M. HOLLE, and R. W. THORP. 1978. Rates of cross-pollination in Lycopersicon pimpinellifolium: Impact of genetic variation in floral characters. Pl. Syst. Evol. 129:31-44. VUILLEUMIER, B. S. 1967. The origin and evolutionary development of heterostyly in the angiosperms. Evolution 21:210-—226. (Received 17 Feb 1981; revised version accepted 15 Jul 1981.) TAXONOMY AND DISTRIBUTION OF OROBANCHE VALIDA (OROBANCHACEAE) L. R. HECKARD Jepson Herbarium, University of California, Berkeley 94720 L. T. COLLINS Department of Science/Technology, Evangel College, Springfield, MO 65802 ABSTRACT The name Orobanche valida subsp. howellii is proposed for populations in the central North Coast Ranges of California, some 600 km from the known range of O. valida subsp. valida. Orobanche valida subsp. valida, heretofore known from two collections of the 1920s in the San Gabriel Mountains of southern California, has been recollected from two populations in these mountains. An overlooked specimen collected in 1908 in the Topatopa Mountains of Ventura County is morphologically intermediate between the two subspecies. Orobanche valida Jepson [Section Nothaphyllon (=Sect. Myzorrhi- za)| has been known from two early collections from the San Gabriel Mountains of southern California (Munz 1959, 1974). W. A. Arm- strong collected plants of the taxon near both earlier sites in 1979. Meanwhile, orobanches of uncertain relationships were collected in north-central California by J. T. Howell in 1926, D. Hemphill in 1951, and later by others. We have since recognized their affinity to O. valida (Collins 1973, Heckard and Chuang 1975). Armstrong’s mate- rial allows us to augment earlier morphological and ecological descrip- tions and to document consistent differences from the northern plants, for which subspecific status is proposed here. OROBANCHE VALIDA Jepson, Madrono 1:255. 1929.—TyPpeE: USA, CA, Los Angeles Co., San Gabriel Mts., South Fork of Rock Creek, 6250 ft, 2 Jun 1928, Frank B. and Mabel B. Peirson 7937 (Holotype: JEPS; isotype: RSA).—Orobanche ludoviciana Nutt. var. valida (Jeps.) Munz, Bull. Torrey Bot. Club 57:621. 1930 [1931]. Pl. 39, Fig. 16. Stem fleshy, mostly underground, terminated by elongated, subspi- cate, blackish purple inflorescence, simple or with a few subordinate spikes at base of inflorescence, the entire axis (10—)15—30(—35) cm long, weakly pilose below, becoming densely so in inflorescence; scales few, narrow to broadly ovate. Inflorescence subspicate with flowers short- pedicelled (rarely to 1 cm long) below to sessile above, 4% as long as to equaling the non-floriferous portion of stem, dense throughout or with lower flowers widely spaced; bracts lance-acuminate to lanceolate MADRONO, Vol. 29, No. 2, pp. 95-100, 29 April 1982 96 MADRONO [Vol. 29 or subulate, to ca. 1 cm long, mostly shorter than the calyx, exceeding the bud before anthesis; flowers subtended by paired linear-subulate bracteoles nearly as long as the calyx. Bracts, bracteoles, and calyces blackish purple, puberulent to short-pilose with mostly glandular hairs. Calyx (4.5—)7-10(-11) mm long, deeply 5-cleft into narrow lin- ear-subulate lobes usually 1 mm broad at base, the tube 2—3 mm deep. Corolla 12—16(—18) mm long with lips 4-5 mm long, the tube strongly constricted above ovary and arching forward at constriction, sparsely to densely short-pilose or puberulent; lips short-pilose to puberulent (some hairs gland-tipped) on both upper and lower surfaces; upper lip erect to spreading and recurved, 3-5 mm long, divided 1—2 mm into 2 triangular-acute dark purple divisions; lower lip spreading, 4-5 mm long, divided ca. 3 mm into 3 triangular to triangular-lanceolate di- visions, usually exceeding upper lip ca. 0.5 mm, each lobe with whitish margins and centrally marked with a dark purplish band along mid- vein extending into the paler tube; palatal folds evident, pale yellow to whitish, puberulent or glabrous within throat of corolla. Stamens included, the upper pair ca. 5 mm long, the lower ca. 6 mm; filaments glabrous or with a few hairs at base; anthers white, almost round, ca. 1.5 mm long, each theca with apiculate base, glabrous or pilose. Ovary narrowly ovoid, the nectary not evident; placentae 2, cleft by a groove running their length; stigma peltate, crateriform, slightly bilobed with the lower lobe larger; capsule narrowly ovoid, 2-3 mm broad, 6—7 mm long. Seeds irregularly ovoid to rhombic, 0.3—-0.4 mm long, light tannish brown, favose. Key to subspecies of Orobanche valida Corolla puberulent externally, weakly so or glabrous at constriction, the trichomes ca. 0.1 mm long; palatal folds of throat glabrous; anthers glabrous; filaments glabrous at base .. 1. subsp. valida Corolla short-pilose externally, densely so at constriction, the tri- chomes 0.2—0.4 mm long; palatal folds of throat puberulent; an- thers pilose, the anther pairs held together by hairs; filaments sparsely pilose at base ................... 2. subsp. howellii 1. OROBANCHE VALIDA subsp. VALIDA. Fig. 1, F—L. Axis (10—)15—30(—35) cm long, 5-10 mm broad at mid-point, usually not enlarged basally. Bracts, bracteoles, and calyx glandular-puber- ulent, the trichome stalks 2(-3)-celled, ca. 0.1 mm long. Corolla 12-14(-16) mm long, the lips and upper tube puberulent (with tri- chomes 0.2-—0.3 mm long), becoming weakly so or glabrous at con- striction and below; palatal folds of throat glabrous. Anthers and base of filaments glabrous. m = 24 (count courtesy of T. I. Chuang, based on Armstrong s.n., 12 Jul 1979, JEPS). 1982] HECKARD AND COLLINS: OROBANCHE VALIDA ecm Fic. 1. Orobanche valida. A-E. subsp. howellii, drawn from Hemphill s.n., 14 Jul 1951. A. Habit. B. Trichomes of corolla. C. Trichomes of calyx. D. Flower. E. Corolla (open). F—L. subsp. valida, drawn from Armstrong s.n., 12 Jul 1979. F. Tri- chomes of corolla. G. Trichomes of calyx. H. Corolla. I. Corolla (open). J. Stigma. K. Ovary (XS). L. Meiotic chromosomes, n = 24. Hosts. Earlier reports on Garrya and Eriodictyon verified by Arm- strong specifically as Garrya veatchii (Armstrong s.n., 15 Aug 1979) and Eriodictyon trichocalyx var. trichocalyx (Armstrong s.n., 1 Aug 1979). 97 98 MADRONO [Vol. 29 Habitat and distribution. On slopes of loose, decomposed granite with chaparral shrubs such as Quercus chrysolepis, Q. wislizenii var. frutescens, Ceanothus leucodermis, Garrya flavescens, Cercocarpus betuloides, and Eriogonum fasciculatum subsp. polifolium. San Ga- briel Mountains, CA, 1250-2000 m. Jun—Aug. Known localities. CA: Los Angeles Co., San Gabriel Mountains: San Gabriel Divide, 1250 m, collector not specified, 5 Jul 1940 (UCSB); trail up s. fork Big Rock Creek, 2.4 km sw. of South Fork Campground, 34°23'N, 117°50'W, Armstrong s.n., 12 Jul 1979, topo- type (JEPS, RSA). San Bernardino Co.: San Gabriel Mountains, s. trail to Baldy Lookout, 1770 m, Johnston 5290 (POM); se. firebreak ridge of Lookout Mt., 1800 m, w. side of ridge near 34°14.6’'N, 117°40.3’W, Armstrong s.n., 1 Aug 1979, 15 Aug 1979 (JEPS). The specimen (Mt. Islip, Burlew, LAM) that Jepson suggested as a likely paratype of Orobanche valida is O. parishii (Jepson) Heckard. Orobanche valida subsp. howellii Heckard and Collins, subsp. nov. Fig. 1, A-E. A Orobanche valida subsp. valida bracteis, bracteolis, calycis pi- lorum stipitibus 3-cellulis ca. 0.2-0.4 mm longis obsitis, corollae tubo dense usque ad eius stricturam brevipilosa, plicis corollae interius pu- berulentis, atque antheris filamentorum basibusque pubescentibus dis- cedit. Axis (6—)10—20 cm long, ca. 3-5 mm broad at mid-point, usually enlarging basally to ca. 1 cm. Bracts, bracteoles, and calyx densely glandular short-pilose, the trichome stalks mostly 3-celled, ca. 0.2—0.4 mm long. Corolla 14—16(—18) mm long, the lips and tube densely short- pilose (with trichomes 0.4—0.7 mm long) to below constriction; palatal folds of throat puberulent. Anthers and base of filaments pilose. 2” = 48. Type: USA, CA, Mendocino Co., Impassable Rock [nw. of summit of Mt. Sanhedrin], 5600 ft, 14 Jul 1951, Donald V. Hemphill s.n. (Holotype: UC; isotype: OSC). Hosts. Reported to be parasitic on Garrya fremontit (Heckard 2331; Hemphill s.n., 21 Sep 1963) and on Quercus chrysolepis (Hemphill s.n., 14 Jul 1951). Habitat and distribution. On rocky (volcanic and ultramafic) slopes in open chaparral. Mountains, central North Coast Range, CA, 1215-1700 m. Jun—Sep. Known localities. CA: Glenn Co., Red Mt. lookout station, on ser- pentine, Hoffman 2495 (UC); Lake Co.: skyline fire trail, sw. slope Cobb Mt., Hamann s.n., Jun 1968 (CAS); Sonoma Co., head of Jacket Cr., Mt. St. Helena, Hemphill s.n., 21 Sep 1963 (CAS); Heckard 2331 1982] HECKARD AND COLLINS: OROBANCHE VALIDA 99 (JEPS); Napa Co., Mt. St. Helena, 1220 m, Howell 2203 (CAS). The plants have been sighted by Glenn Clifton in two other localities in Napa Co.: Hunting Creek (T11N, R4W, Sec 13) and near Lake Ber- ryessa (TON, R4W, Sec 21). The new subspecies is named in honor of John Thomas Howell, Curator Emeritus of the California Academy of Sciences, who has made significant contributions to our knowledge of California plants and who first collected O. valida subsp. howelliz. The most striking features distinguishing Orobanche valida subsp. valida from subsp. howellii are the differences in indument on floral parts. On the microscopic level there is a difference in trichome size on the corolla and on the bracts, bracteoles, and calyx (Fig. 1, B, C, F, G): those of subsp. howellii are about twice as long as those of subsp. valida. Available collections indicate that the two subspecies differ in other features: subsp. howelliz has a slightly larger corolla and the plants in general are smaller than subsp. valida; stems of subsp. howellii are narrow below the spike but enlarged gradually toward the base while those of subsp. valida are broader throughout. The earliest collection of O. valida, two specimens recently detected on a sheet with O. bulbosa Beck, adds complexity to the morphological distinction of the subspecies. The specimens (Dudley & McGregor 121, 4-6 Jun 1908), from the Topatopa Mountains of Ventura County in southern California, appear to be intermediate in indument of the corolla and in trichome size of the bracts and calyx. The anthers are pilose as in subsp. howellii, but the stem is similar to subsp. valida in being broad and not enlarged basally. Additional collections of O. valida in the Topatopa Mts. and other localities are needed to assess morphological constancy. It is likely that the plants will be found in other areas. Their summer flowering and the habitat in rocky chap- arral areas in remote regions have doubtlessly contributed to their infrequent collection. The relationship of Orobanche valida to other species of section Nothophyllon is difficult to assess owing to the limited number of morphological features available for comparative purposes and to lack of knowledge of evolutionary trends in these features. Seemingly the closest relatives of O. valida are the huskier, desert-dwelling plants of O. cooperi (Gray) Heller, which share such features as purplish black herbage and corolla with deep purplish markings and copious, relatively long trichomes. The two are, in our opinion, certainly spe- cifically distinct, with O. valida having narrower spikes (2—3 cm vs. ca. 4 cm broad in O. cooperi), narrower bracts (lanceolate to lance- ovate vs. lance-ovate to ovate), and smaller corollas (11-14 mm vs. 16-30 mm). Also distinguishing the two are the desert habitat of O. cooperi and its nearly complete host-restriction to the tribe Heliantheae subtribe Ambrosiinae of Compositae. Orobanche valida is morpholog- ically closer to the small-flowered race (n = 48) of O. cooperi (Heckard 100 MADRONO [Vol. 29 and Chuang 1975) than to the large-flowered race (nm = 24). This raises the possibility that O. valida subsp. valida may be ancestral, along with the large-flowered race, in an amphidiploid origin of the higher polyploid. This aspect needs further study. ACKNOWLEDGMENTS We thank Donald Hemphill and Wayne A. Armstrong for their collections for this study and Fei-Mei and T. I. Chuang for the chromosome count of Orobanche valida subsp. valida. LITERATURE CITED CoLuins, L. T. 1973. Systematics of Orobanche section Myzorrhiza. Ph.D. disserta- tion, University of Wisconsin, Milwaukee. HECKARD, L. R. and T. I. CHUANG. 1975. Chromosome numbers and polyploidy in Orobanche (Orobanchaceae). Brittonia 27:179—186. Muwz, P. A. 1959. A California flora. Univ. California Press, Berkeley. . 1974. A flora of southern California. Univ. California Press, Berkeley. (Received 8 May 1981; revised version accepted 10 Aug 1981.) DICORIA ARGENTEA (COMPOSITAE: AMBROSIINAE), A NEW SPECIES FROM SONORA, MEXICO JOHN L. STROTHER Botany Department, University of California, Berkeley 94720 ABSTRACT Dicoria argentea Strother from southern coastal Sonora is allopatric to and has larger fruits with relatively smaller bracts than its congeners. Ann Johnson noted populations of this species on beach dunes at Isla Lobos, Isla Huivulai, and the type locality, all on Sonoran coast south of Guaymas. The plants are apparently restricted geographically (reported sought but not found by Dr. Johnson, e.g., at Guaymas, Sonora, and Topolobampo, Sinaloa). According to the collector, the plants are locally dominant on the dunes and are associated with As- clepias subaphylla Woodson, Croton californicus Muell.-Arg., and Palafoxia linearis (Cav.) Lag. Dicoria argentea Strother, sp. nov. Suffrutices procumbentes ad ca. 2 dm alti et ca. 3 m diam.; folia lanceolata vel ovata 8-25 mm longa 3-12 mm lata integra sericeo- cana; bracteae flosculorum pistillatorum orbiculares vel reniformes postremo ca. 5 mm longae; flosculi pistillati 1-2; flosculi staminati 6—-18+ corollis 3-4 mm longis; achenia includentia alae 9-12 mm longa 10-13 mm lata; corpora acheniorum 6—8 mm longa 4-6 mm lata; alae acheniorum 3-4 mm latae (Fig. 1). Sprawling, procument suffrutices to ca. 2 dm high, to ca. 3 m diam.; stems appressed sericeo-canescent, less so in age; leaves sometimes opposite, mostly alternate, petioles mostly 2-10 mm long, blades lan- ceolate to lance-elliptic or ovate, 8-25 mm long, 3-12 mm wide, 3-nerved from near base, basally cuneate, apically acute, entire, both faces sericeo-canescent; heads solitary or 2—3 loosely aggregated; pe- duncles mostly 1-3 cm long; involucres roughly campanulate at an- thesis; phyllaries proper 5, free, herbaceous, ovate to lanceolate, 2-4 mm long, abaxially sericeo-canescent; “paleae” or inner phyllaries 1-2, each subtending a pistillate floret, membranous, becoming scarious, ovate to orbicular or subreniform, somewhat cupped, ca. 2 mm long, 3 mm wide at anthesis, ultimately to 5 mm long, sparsely to densely appressed-pilose abaxially; receptacles very small, slightly convex, ap- parently without paleae subtending the staminate florets; pistillate flo- rets 1-2, corollas none, style branches rather stout, ca. 1 mm long; staminate florets 6—-18+, corollas pale yellow or ochroleucous, tube 0.3-0.5 mm long, grading gradually into cylindro-campanulate throat 2.5-3.0 mm long, lobes 5, equal, deltoid, erect or incurved, 0.5-0.8 MADRONO, Vol. 29, No. 2, pp. 101-103, 29 April 1982 102 MADRONO [Vol. 29 SH yy? ML Fic. 1. Habit, immature head, and achene of Dicoria argentea. mm long, throat and lobes abaxially strigillose, stamens united by dilated filament bases, anthers free, 1.2—1.5 mm long, styles reduced, unbranched; achenes dorsi-ventrally flattened, the body black/brown, sometimes buffy-mottled on abaxial face, ovate-elliptic, 6-8 mm long, 4—6 mm wide, weakly keeled or ridged medially on one or both faces, subglabrous or sparsely puberulent and with scattered resinous glob- ules, wings stramineous, cartilaginous, 3—4 mm wide, sparsely puber- ulent, glabrescent, pectinate-erose or fimbriate; pappus none, but apex of achene may be minutely bidentate and crowned with a tuft of hairs that are soon lost. 1982] STROTHER: NEW SPECIES OF DICORIA 103 Type: Mexico, Sonora, Huatabampito (10 km s. of Huatabampo), ca. 26°38'N, 109°40’W, forming dunes just back from beach, 9 Jan 1981, Ann F. Johnson 8001. (Holotype: DAV; isotypes: to be sent to ARIZ, C, ENCB, G, K, MEXU, NY, SD, UC, US). Consistently small entire leaves plus mostly solitary heads and large fruits (9-12 mm long and 10-13 mm wide, including wings) with relatively small subtending bracts readily distinguish D. argentea from its congeners, all of which range farther north in Sonora, Baja Cali- fornia, western Arizona, southern California, southwestern Colorado, southern Nevada, northwestern New Mexico, and southern Utah (rec- ords in CAS, DS, JEPS, and UC and in standard floras of the area). Rydberg (N. Amer. FI. 33:11-13. 1922) recognized 7 species of Dz- corta; I suspect that number might be reduced to 4-6 were the genus critically reviewed. Dicoria argentea seems to be morphologically iso- lated within its genus. ACKNOWLEDGMENTS I thank Ann Johnson for bringing plants to my attention, Alan Smith for checking my Latin, and US for loaning the type of Dicoria calliptera Rose & Standl. (Received 8 Mar 1981; revised version accepted 20 Aug 1981.) ON THE RECOGNITION OF TRICHOSTEMA MEXICANUM EPLING (LAMIACEAE) JAMES HENRICKSON Department of Biology, California State University, Los Angeles 90032 ABSTRACT Additional collections of Trichostema mexicanum Epling from the Chihuahuan Desert indicate it can be consistently separated from the more wide-ranging 7. arizonicum of the mountains of Chihuahua, Sonora and adjacent Arizona and New Mexico. The former has smaller, completely blue-violet flowers and recognizable differences in in- florescence development and vestiture. Epling (1940) described Trichostema mexicanum (Fig. 1 A—E) from a single collection of C. H. Muller (3053) from Puerto de San Lazaro in west-central Coahuila, Mexico (ca. 63 km south of Monclova). The new taxon was considered most closely related to T. arizonicum Gray (Fig. 2 A—D), which occurs mainly in southeastern Arizona and ad- jacent northern Sonora with scattered collections in southwestern New Mexico, northern Chihuahua and southern Sonora, Mexico (Lewis 1945, Lewis and Rzedowski 1978). Epling noted that T. mexicanum differs from 7. arizonicum in certain reduced features, mainly vesti- ture, smaller flowers and unbranched inflorescences. Lewis (1945) sub- merged 7. mexicanum in T. arizonicum, placed in his new section Paniculatum, noting that the characteristics of Muller’s collection fell just within the range of variation in T. arizonicum. He also noted the similarity of the Muller specimen with one from the Mustang Moun- tains in Arizona collected in 1884 by Pringle. Since that time two additional collections referable to T. mexicanum, one reported by Lewis and Rzedowski (1978) from northern San Luis Potosi (Rze- dowski 6352, MEXU) and one from northern Zacatecas (113 air km northeast of Cuidad Zacatecas along highway 54 at K115; Henrickson 6675, LL), indicate a geographical and character continuity supporting the distinctness of the species. The character differences between 7. arizonicum and T. mexicanum are indicated in Table 1 and Figs. 1 and 2. The strongest, most con- sistent differences are found in corolla size as indicated from total corolla and filament lengths measured in dried specimens. Corolla color also appears to be consistently different (Table 1). According to label data, color of the four posterior corolla lobes of T. arizonicum varies within a population. The lavender color of these lobes appar- ently fades to white upon drying except when the petals dry rapidly in certain herbarium specimens. More variation is found in inflorescence development. Trichostema arizonicum typically has paired dichasia at the upper nodes but oc- MADRONO, Vol. 29, No. 2, pp. 104-108, 29 April 1982 1982] HENRICKSON: TRICHOSTEMA MEXICANUM 105 Fic. 1. Trichostema mexicanum. A. Habit showing paired flowers on long peduncles and pedicels at upper nodes. Larger leaves are commonly lobed (Henrickson 6675, LL). B. Flower, side view with slightly exserted anthers. C-D. Mature mericarp. C. Adaxial view showing conspicuous attachment scar and sessile glands, trichomes on top. D. Abaxial view showing coarse, alveolate sculpturing. E. Stem vestiture showing straight declined hairs. Magnifications as indicated. casional individuals have only two flowers per paired node. This can vary from plant to plant within a single gathering as exemplified by two collections by F. W. Gould (3695) from the Santa Catalina Moun- tains in Arizona. One sheet at LA has only racemose inflorescences with paired flowers at the nodes, whereas a sheet at ARIZ shows the 106 TABLE 1. mexicanum. MADRONO [Vol. 29 CHARACTER DIFFERENCES BETWEEN Trichostema arizonicum AND T. T. arizonicum Inflorescence Corolla length Filament length Corolla color Stem vestiture Pedicel vestiture Nutlet vestiture Nutlet ornamentation Range Habitats Flowers borne in paired, axillary 3(—5)-flowered dichasia. 9.0-13.5 mm. 15-25 mm. Four posterior lobes light lavender to white, anterior lobe blue-violet marked with white or lavender. Trichomes decurved, to 0.1(-0.2) mm long [to 0.3(—0.4) mm long in s. Sonoran populations], mixed with stipitate glands. Obscurely stipitate-glandular. Mostly few scattered, sessile to short-stipitate glands to 0.1 mm long. (Longer non- glandular trichomes in s. Sonora populations). Reticulate pattern obscure. Southern Arizona, New Mexico to s. Sonora, w. Chihuahua Oak woodland, grassland, streamsides; 1200-1800 m. T. mexicanum Flowers paired at upper nodes. 4.0-5.0 mm. 5-6 mm. Dark blue-violet throughout with lighter markings on anterior lobe. Trichomes declined (rarely some decurved), mostly 0.13 mm long, mixed with stipitate glands. Strongly stipitate-glandular. Mostly longer, straight to curved, non-glandular trichomes 0.1—0.2 mm mixed with more numerous, longer stipitate glands. Reticulate pattern bolder, more distinct. Central Coahuila, n. Zacatecas to central San Luis Potosi. Xeric habitats, Larrea grasslands to arid scrub; 1200-2300 m. characteristic paired dichasia. Variation in inflorescence development was also noted in 7. mexicanum. The Rzedowski (6352) specimen exhibits a few three-flowered dichasia. Vestiture differences between the two species hold for most speci- mens, but variation does exist. The typically decurved stem trichomes of 7. arizonicum (Fig. 1 E) are usually mixed with scattered sessile to short-stipitate, yellow-tipped glands. In some specimens pedicels have longer, coarser, decurved trichomes or more strongly developed stip- itate glands. Some specimens from southern Sonora have much longer, decurved trichomes 0.3(—0.4) mm long on stems and leaves. The may also have longer stipitate glands on the inflorescence and longer hairs on nutlets. Vestiture of 7. mexicanum also shows some variation. Stem trichomes are mostly declined (Fig. 1 E) but they may be mixed with some decurved trichomes. Stipitate glands are more abundant on ped- 1982] HENRICKSON: TRICHOSTEMA MEXICANUM 107 Fic. 2. Trichostema arizonicum. A. Habit showing paired lateral dichasia at upper nodes (Kearney & Peebles 14878, ARIZ). B. Flower, oblique lateral view showing long-exserted anthers. C. Anterior corolla lobe showing outline and pattern of white on surface. D. Stem vestiture showing characteristic decurved trichomes. Magnifications as indicated. 108 MADRONO [Vol. 29 icels in this species as well. Leaves and calyces of both species have upcurved trichomes mixed with sessile to short-stipitate glands. Of the 76 collections of 7. arizonicum studied from ARIZ, LA, LA in UC, NY, RSA-POM, TEX-LL, and UC, very few exhibited any intermediate characteristics with 7. mexicanum. None exhibited floral similarities; only two specimens of 7. arizonicum were characterized by simple racemose, instead of paniculate, inflorescences; and only a few plants were more strongly stipitate-glandular in inflorescences, a characteristic of T. mexicanum. The above noted Pringle 1884 collec- tion from the Mustang Mountains in Arizona was similar to 7. mex- icanum only in the reduced inflorescence. All other characteristics were typical of 7. arizonicum. In my opinion, the data presented above supports the recognition of 7. mexicanum as a distinct species. The following key can be used to separate the taxa. a. Corollas 4-5 mm long; dark blue-violet throughout except for white markings on the anterior lobe; filaments 5-6 mm long; flow- ers paired at nodes; stem trichomes mostly straight, declined but not decurved; pedicels usually strongly stipitate-glandular; central Coahuila to northern Zacatecas, San Luis Potosi, Mexico ...... ee ee a PES eS ORIN mr eee”: T. mexicanum aa. Corollas 9-13.5 mm long, the anterior lobe blue-violet with white markings, the other lobes white to lavender; filaments 15-25 mm long; flowers borne in paired, 3—5-flowered dichasia; stem tri- chomes decurved; peduncles mostly obscurely glandular; southern Arizona, southern New Mexico to Sonora, western Chihuahua, VEG RIGO) sey te ate oop crane, ees ae ene pee T. arizonicum ACKNOWLEDGMENTS I thank the curators of the above noted herbaria for loans, Bobbi Angell for the illustrations, and the Plant Resources Center at the University of Texas, Austin, for use of facilities. LITERATURE CITED EPLING, C. 1940. Supplementary notes of American Labiatae. Bull. Torrey Bot. Club 67:509-5 34. Lewis, H. 1945. A revision of the genus Trichostema. Brittonia 5:276-303. and J. RZEDOWSKI. 1978. The genus Trichostema (Labiatae) in Mexico. Ma- drono 25:151-154. (Received 5 May 1981; revised version accepted 29 Jul 1981.) ENVIRONMENTAL AND COMPOSITIONAL ORDINATIONS OF CONIFER FORESTS IN YOSEMITE NATIONAL PARK, CALIFORNIA ALBERT J. PARKER Department of Geography, University of Georgia, Athens 30602 ABSTRACT Quantitative phytosociological data from thirty stands sampled along an elevational gradient (1220-2440 m) in Yosemite National Park are used to generate a direct envi- ronmental ordination (elevation vs. a scalar-index integrating site topographic param- eters) and two compositional ordinations (Bray-Curtis technique and reciprocal aver- aging). In each case, elevation emerged as the principal factor controlling forest pattern. Environmental scaling revealed that lower elevation forests (1220-1900 m) in Yosemite consist of a ponderosa pine/incense-cedar type on xeric topographic settings and a white fir/incense-cedar type on mesic topographic settings. Higher elevation forest (1900-2400 m) are generally dominated by a mixture of red fir and white fir, although neither the environmental nor the compositional arrangement of stands adequately accounts for patterns of species dominance in these forests. In this study, integrated use of environ- mental and compositional techniques reinforce interpretations as well as provide insights which are not directly apparent from independent use of either approach. Both environmental and compositional ordination techniques have proven to be valuable means of analyzing patterns of plant and animal communities in “natural” settings (Whittaker 1956, Curtis 1959, Bond 1957). While Whittaker (1967) argued that the use of environmental measures (e.g., elevation, aspect, topographic position) is a more direct approach to analyzing plant-environment relationships, McIntosh (1967) and others from the “Wisconsin” school contended that it 1s preferable to avoid presupposing any necessary environmental control of vegetation pattern; rather, they defend the compositional approach to ordination because it allows the unforced emergence of significant environmental controls from vegetation data. From a biogeographic perspective, the evolution of these two ordination techniques is strong- ly tied to the character of respective study sites. Environmental or- dination techniques were developed for use in midlatitude mountain- ous regions, where elevation plays a dominant role in controlling local climate and hence vegetation patterns. Moreover, topographic con- ditions modify the effects of elevation or climate and must be incor- porated in any quantitative assessment of vegetation patterns in areas of complex relief. Compositional ordination techniques, on the other hand, were created to analyze vegetation patterns over broad geographic areas of reduced local variability, in which environmental gradients are only subtly expressed in the landscape, if at all (Curtis and McIntosh 1951). Furthermore, compositional ordination was used to impose an organizational framework on a set of geographically discontinuous sites. Although neither ordination method exhibits a MADRONO, Vol. 29, No. 2, pp. 109-118, 29 April 1982 110 MADRONO [Vol. 29 universal superiority, both approaches to vegetation study may be integrated to enhance the analysis of ecological patterns in a region (Loucks 1962). The present study provides a comparison of the results of environmental and compositional ordinations of forest stands in Yosemite National Park, California. In addition, although descriptive treatments of forest patterns in the Sierra Nevada are readily available (Klyver 1931, Munz and Keck 1949, Storer and Usinger 1963), this study represents one of the rare attempts to quantify these phytoso- ciological patterns (Rundel et al. 1977) and is the first such effort in Yosemite National Park. Thirty stands were selected for sampling in the conifer forests of the montane and lower subalpine zone (1220—2440 m) on the western flank of the Sierra Nevada in Yosemite National Park, California—twenty- five stands between South Entrance and Glacier Point, plus five stands within a 10 km radius of Crane Flat. This area possesses a relatively mild climate with a pronounced summer dry period (the period from June through August receives less than 3 percent of the total annual precipitation). Mean annual precipitation increases with increasing elevation within the sampled zone, from 850 mm at 1220 m to 1500 mm at 1830 m and above. At 1200 m (Yosemite Village), maximum/ minimum temperatures range from 30°/12°C in July to 10°/—2°C in January. Mean temperatures are presumed to decline with elevation at a rate of approximately 0.5°C per 100 m (National Oceanic and Atmospheric Administration 1978). Most stands are on weakly developed soils (Xerochrepts) derived from the Cretaceous granodiorite of which the Sierra Nevada batholith is composed. The exceptions are two stands near Glacier Point, which are on Quaternary glacial till derived from the granodiorite, and two stands near Crane Flat, which are on Paleozoic metasedimentary car- bonates flanking the batholith (Matthes 1930, Matthews and Burnett 1966). The initial material for soils on many sites appears to be derived from colluvial processes. METHODS The thirty stands sampled were stratified so that six stands occurred in each of five elevation belts, each belt 244 m in vertical extent, between 1220 and 2440 m. Each stand was judged to be homogeneous, in the sense that no environmental or vegetational discontinuities were apparent. Vegetation sampling was restricted to forests that had ap- parently been free from wholesale disturbance for extended periods. Riparian sites were excluded from consideration. Each stand was com- posed of twenty 1/100 ha circular quadrats, located by a stratified systematic unaligned technique (Berry and Baker 1968) within a 4 by 5 unit grid pattern, each grid unit being 15 m on a side. In the center of each stand topographic position (ridgetop to lower slope), cross- 1982] PARKER: ECOLOGY OF CONIFER FOREST 111 slope and down-slope configuration (convex to concave), aspect, slope steepness, and elevation were recorded. Within each quadrat, the species and basal area at breast height (1.4 m) of each individual in the tree layer were recorded. The tree layer was considered to include all stems exceeding 0.8 dm? (10 cm dbh) at breast height. Saplings (stems exceeding breast height but smaller than 0.8 cm?) and seedlings were also recorded by species. Inclusion of this information in ordi- nations, however, yielded patterns similar to those reported below, which are based on tree layer data only, and, thus are not presented. Environmental ordering of stands consisted of establishing a bivar- late space with elevation on one axis and a synthetic index of site topographic condition on the second axis. Unlike Whittaker (1956, 1967) who quantified the topographic complex gradient with a weight- ed average technique dependent on the vegetal cover (Ellenberg 1947), I have assigned positions along the topographic axis by using a mod- ification of the Topographic Potential Moisture Index (TPMI) (Parker 1980). This is a synthetic scalar index based on the sum of assigned values for the following set of slope characteristics: topographic po- sition, slope configuration, aspect, and steepness. TPMI values may range from O (driest sites) to 60 (wet sites). For each stand, be- tween O and 20 TPMI units are assigned for topographic position (ranging from 0 on ridgetops to 20 in valley bottoms) and aspect (rang- ing from O at ssw. to 20 at nne.). For each stand, between O and 10 TPMI units are assigned for steepness (ranging from 0 on slopes >30° to 10 on slopes <3°), and configuration (ranging from 10 for concave, or water collecting, slopes to 0 for convex, or water dispersing, slopes). Topographic position and slope aspect are weighted twice as heavily as the other factors because both factors were judged to have more influence on vegetation distribution, as inferred from field ob- servation and previous literature (Whittaker 1956, Hack and Goodlett 1961). The following modifications in the use of TPMI from Parker (1980) were made: soil depth was omitted as a factor in determining the TPMI because it rarely affects moisture availability patterns dur- ing the growing season in Sierran montane conifer forests (Arkley 1981), slope configuration was refined to include both cross-slope (transverse) and down-slope (longitudinal) components, and riparian sites were not sampled so as to alleviate the confounding influence of locally high groundwater. Compositional ordination of tree species was performed by two dif- ferent techniques, Bray-Curtis ordination (Bray and Curtis 1956) and reciprocal averaging (Bakuzis and Hansen 1959, Hill 1973). In each case, importance values for all tree species (after Curtis and McIntosh 1951) were used to provide a data matrix for statistical treatment. For the Bray-Curtis ordination the measure of dissimilarity used to gen- erate an ecological distance matrix among stands was the Manhattan metric [c = 1 — 2w/(a + b)]. The endpoint selection method used to 112 MADRONO [Vol. 29 generate the Bray-Curtis ordination was the regression method of Beals (pers. comm.). Reciprocal averaging, which is an iterative tech- nique that successively refines weighting values, or “scores,” for both species and stands along a single compositional gradient, was adopted for use after examining the Bray-Curtis ordination and identifying the prominence of a single environmental gradient (elevation) in the vege- tation data. The exclusion of groundcover composition from these ordinations may limit the interpretation of patterns within dominant cover types (Peet 1981), but should have little influence on the inter- pretation of environmental patterns among overstory species. RESULTS Three forest types are recognized in the direct environmental ordi- nation: ponderosa pine (Pinus ponderosa)/incense-cedar (Calocedrus decurrens) forest, white fir (Abies concolor)/incense-cedar forest, and red fir (Abies magnifica)/white fir forests (all taxonomy after Munz 1959, 1968) (Fig. 1). Most stands are readily assigned to one of these three types. However, two stands represent transitional elements be- tween types (Stands 8 and 17), a predictable result given the contin- uous variation of vegetation (Whittaker 1967), whereas three of the red fir/white fir forest stands are compositional variants that include locally important Jeffrey pine (Pinus jeffreyi; Stands 20 and 24) or lodgepole pine (Pinus contorta subsp. murrayana; Stand 28) popula- tions. The distribution of stands and forest types indicates the significance of both elevation and topographic conditions in controlling vegetation patterns (Fig. 1). The ponderosa pine/incense-cedar forest type and the white fir/incense-cedar forest type occur at elevations from 1200 to 1900 m, and are differentiated topographically. Ponderosa pine is more common at lower elevations, on south-facing slopes, and on upper slope positions, where solar radiation and potential evapotran- spiration are high and runoff is dispersed. Compensation for macro- climatic changes along the elevational gradient is reflected in the pro- gressive restriction of ponderosa pine to drier sites associated with steep, south-facing, and convex slopes above 1600 m. White fir, which is present on mesic sites in the lower elevation forests of Yosemite National Park, also displays the interplay of elevation and topographic setting. White fir is a common dominant in forests above 1600 m, but below this elevation white fir is dominant only on north-facing, con- cave slopes characterized by reduced solar radiation, reduced potential evapotranspiration, and concentrated surface runoff and soil through- flow. Stand 8, a dry mesic site at 1600 m, is considered transitional because of the equivalent importance value (25.7 percent) of both white fir and ponderosa pine. This quantification of the relative im- portance of ponderosa pine and white fir along a topographic sequence 1982] PARKER: ECOLOGY OF CONIFER FOREST 113 RED FIR/ 2200 WHITE FIR 2000 1800 ELEVATION (m) WHITE FIR/ 1600 INCENCE - CEDAR PONDEROSA PINE/ INCENCE-CEDAR 1400 10 20 30 40 TPMI Value Fic. 1. Environmental ordination of Yosemite stands. Using a scale of importance values, with 5—80.1-100.0, 4—60.1-80.0, 3—40.1-60.0, 2—-20.1—40.0, and 1—0. 1-20.0, the importance of ponderosa pine (below), white fir (right), and red fir (above) is indi- cated next to stand positions. Stand numbers are underscored to the left of plotted position. 114 MADRONO [Vol. 29 enumerated by the TPMI in lower elevation Yosemite forests corrob- orates the descriptive treatments of tree distribution patterns in the central Sierra that have been published during the last half-century (cf. Rundel et al. 1977). The zone between lower and higher elevation forests (ca. 1900 m) is under-represented in this study. I considered the occurrence of in- cense-cedar to be crucial in making this elevational distinction. All stands classified as red fir/white fir lack incense-cedar and are clearly removed from white fir/incense-cedar forests in the compositional or- dination. The compositional ordination does suggest that Stand 17 is somewhat transitional in nature, with similar importance values for red fir (26.0) and incense-cedar (24.8). Clearly, Stands 12 and 18 might also be classed as compositionally transitional. All three of these tran- sitional stands occupy a continuous portion of the mesic side of the environmental space, between 1700 and 1900 m. Red fir/white fir forests occur in a wide variety of topographic set- tings between 1900 and 2400 m in Yosemite National Park. In this elevation zone, the TPMI is not effective in differentiating habitats where red fir reaches highest dominance from those where white fir is most important (Fig. 1). There is little indication that red fir is more abundant than white fir on more mesic sites, or at higher elevations, although the regional zonal relationship of these two species suggests that red fir is better adapted to cooler, moister settings than white fir. The forests of this zone appear to be ecotonal mixtures of white fir and red fir located in the lower portion of a broad elevation zone dominated solely by red fir (Oosting and Billings 1943). The inability of the TPMI to identify habitat preferences of red and white fir on these transitional sites may be related to the confounding influence of unusual microclimatic conditions. For example, Stands 29 and 30 ap- pear on the figures to be cool, moist sites, occuring at high elevations (2200-2300 m) and on protected exposures (north-facing, lower slopes). Therefore, the dominance of white fir in these forests is puzzling until it is noticed that these stands are located immediately above Glacier Point. Because of their location on the rim of Yosemite Valley, they may experience the warming and drying influence of an upslope day- time breeze (as described for the North Rim of the Grand Canyon by Halvorson 1972), resulting in the enhancement of evapotranspiration rates. This example illustrates the need for exercising caution in in- terpreting the interaction of elevation and TPMI values, because local wind conditions can override the influence of topographic parameters in controlling water availability and demand on certain sites. The strong dominance of Stand 28 by lodgepole pine is also apparently related to wind. Here, extreme exposure on the flank of a granite outcrop induces physiological drought and locally alters winter snow- pack depth and persistence. Bray-Curtis ordination (Fig. 2) verifies the existence of the forest 1982] PARKER: ECOLOGY OF CONIFER FOREST 115 LODGEPOLE PINE PONDEROSA PINE/ _ RED FIR/ INCENSE-CEDAR WHITE FIR WHITE FIR/ INCENSE- CEDAR Fic. 2. Bray-Curtis ordination of Yosemite stands. Stands are located with respect to the first two synthetic axes of a Bray-Curtis type ordination. types recognized above. In this ordination, however, the composition of Stand 28 (dominated by lodgepole pine) is unique and warrants separate recognition from red fir/white fir forests. As was the case in the environmental space, the ordination identifies two transitional stands (8 and 17). The first axis of the compositional ordination correlates well with elevation (r = 0.847, p < 0.01). The emergence of elevation as the factor most strongly correlated 116 MADRONO [Vol. 29 with forest composition prompted the use of the first axis of a recip- rocal averaging ordination as an alternative means of arranging the stands because this technique generates an arrangement of species optima, as well as stand locations, along a single ecological gradient. Like the Bray-Curtis ordination, reciprocal averaging values for stands emphasize the elevational gradient (r = 0.814, p < 0.01). The three forest types are clearly distinguished by reciprocal averaging (ponder- osa pine/incense-cedar stand scores range from 0.0 to 7.2, white fir/ incense-cedar stand scores range from 21.0 to 37.5, and red fir/white fir stand scores (except Stand 28) range from 53.7 to 79.3). Moreover, the transitional nature of Stands 8 (18.1) and 17 (46.2) and the unique- ness of Stand 28 (100.0) are underscored. Of particular ecological in- terest is the arrangement of species optima along the elevational gra- dient, with reciprocal averaging scores increasing with increasing elevation: ponderosa pine (0.0), California black oak (Quercus kellogg- 11) (9.0), incense-cedar (14.1), sugar pine (Pinus lambertiana) (21.6), white fir (43.7), red fir (66.6), Jeffrey pine (72.4), and lodgepole pine (92.4). This quantitative arrangement of species optima by reciprocal averaging parallels the descriptive treatments of vegetation zonation published by previous workers (Rundel et al. 1977). DISCUSSION Both the environmental and compositional ordinations presented for Yosemite National Park indicate the significance of elevation in controlling forest community patterns and document quantitatively the zonal sequence of forest types as well as the associated sequence of species optima along an elevational gradient. Among the two lower elevation forest types, the environmental or- dination clearly differentiates the role played by topographic features. In this regard, the Topographic Potential Moisture Index provides an integration of slope-related hydrologic factors that is useful for seg- regating dominance patterns of ponderosa pine and white fir. By con- trast, although both of the compositional ordination techniques (Bray- Curtis and reciprocal averaging) support the recognition of two lower elevation forest types (ponderosa pine/incense-cedar and white fir/in- cense-cedar), these techniques fail to indicate the importance of mois- ture regime as influenced by topographic parameters in controlling dominance by white fir and ponderosa pine. Thus, by failing to dis- tinguish between elevation and topographic modifications of available moisture, the use of compositional ordination alone results in the loss of valuable interpretive information. Patterns of tree species occurrence in high elevation forests are not adequately explained by either environmental or compositional meth- ods. Local peculiarities in wind, snowpack, and substrate conditions obscure the expression of elevation and topographic influence on the 1982] PARKER: ECOLOGY OF CONIFER FOREST 117 distribution of forest types, rendering the TPMI ineffective in segre- gating habitats. The separation of one lodgepole pine dominated stand in both compositional ordinations underscores the need to consider factors in addition to elevation and topographic condition in examining environmental relationships among higher elevation forest types in Yosemite. If a choice between environmental and compositional ordination is necessary, direct analysis of environmental gradients may be prefer- able in situations where the complex of environmental factors influ- encing vegetation patterns is obvious, such as in landscapes of high relief. At the same time, care must be taken to avoid misinterpreting species patterns on sites modified by specific extenuating factors, such as edaphic constraints. However, in landscapes where direct environ- mental control of vegetation is not apparent, either because of subtle expression of environmental gradients or marked temporal heteroge- neity among stands, compositional ordination may help identify the most important environmental factors. I found that union of the two techniques yielded a mutual reinforcement of the interpretive pattern and provided some unique insights. Environmental, or direct gradient analysis, techniques allow direct assessment of the influence of site factors on patterns of dominance, whereas compositional ordination identifies both stand clusters and transitional stands and thus facili- tates the placement of forest types within an environmental frame- work. ACKNOWLEDGMENTS I acknowledge the financial support of the National Science Foundation and Uni- versity of Wisconsin Graduate Fellowship programs, the editorial and advisory assis- tance of Thomas R. Vale, and the field assistance of Kathleen C. Parker. Improvements suggested in review comments by Mary F. Burke and James C. Hickman were most helpful. LITERATURE CITED ARKLEY, R. J. 1981. Soil moisture use by mixed conifer forest in a summer-dry climate. Soil Sci. Soc. Am. J. 45:423-427. Bakuzis, E. V. and H. L. HANSEN. 1959. A provisional assessment of species syn- ecological requirements in Minnesota forests. Minnesota Forestry, Forestry Note 84. BEALS, E. W. Bray and Curtis ordination: a defense and an update. Manuscript sub- mitted to J. Ecol. BERRY, B. J. L. and A. M. BAKER. 1968. Geographic sampling. Jn B. J. L. Berry and D. F. Marble, eds., Spatial analysis: a reader in statistical geography, p. 91-100. Prentice-Hall, Engelwood Cliffs, NJ. Bonp, R. R. 1957. Ecological distribution of breeding birds in the upland forests of southern Wisconsin. Ecol. Monogr. 27:351-384. BRAY, J. R. and J. T. Curtis. 1957. An ordination of the upland forest communities of southern Wisconsin. Ecol. Monogr. 27:325-349. CurTIS, J. T. 1959. The vegetation of Wisconsin: an ordination of plant communities. Univ. Wisconsin Press, Madison. 118 MADRONO [Vol. 29 . and R. P. McINTosH. 1951. An upland forest continuum in the prairie-forest border region of Wisconsin. Ecology 32:476-496. ELLENBERG, H. 1948. Unkrautgesellschaften als Mass fur den Satiregrad, die Ver- dichtung und andere Eigenschaften des Ackerbodens. Ber. Landtech. 4:130—-146. HACK, J. and J. GOODLETT. 1961. Geomorphology and forest ecology of a mountain region in the central Appalachians. U.S. Geol. Surv. Prof. Paper 347. HALVORSON, W. 1972. Environmental influence on the pattern of plant communities along the North Rim of Grand Canyon. Amer. Midl. Naturalist 87:222-235. HILL, M. O. 1973. Reciprocal averaging: an eigenvector method of ordination. J. Ecol. 61:237-249. KLYVER, F. D. 1931. Major plant communities in a transect of the Sierra Nevada Mountains of California. Ecology 12:1-17. Loucks, O. L. 1962. Ordinating forest communities by means of environmental scalars and phytosociological indices. Ecol. Monogr. 32:137-166. McInTosH, R. P. 1967. The continuum concept of vegetation. Bot. Rev. 33:130-187. MATTHES, F. E. 1930. Geologic history of the Yosemite valley. U.S. Geol. Surv. Prof. Paper 160. MATTHEWS, R. A. and J. L. BURNETT. 1966. Geologic map of California. Mariposa sheet. Calif. Div. Mines Geology. Sacramento. Muwnz, P. A. 1959. A California flora. Univ. California Press, Berkeley. . 1968. Supplement to a California flora. Univ. California Press, Berkeley. . and D. D. KEck. 1949. California plant communities. Aliso 2:87—-105. NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION. 1978. Climates of the states. Vol. 1. Gale Research Co., Detroit. OosTING, H. J. and W. D. BILLINGs. 1943. The red fir forest of the Sierra Nevada: Abietum magnificae. Ecol. Monogr. 13:259-274. PARKER, A. J. 1980. Site preferences and community characteristics of Cupressus ari- zonica Greene (Cupressaceae) in southeastern Arizona. Southw. Naturalist 25:9-22. PEET, R. K. 1981. Forest vegetation of the Colorado Front Range: composition and dynamics. Vegetatio 45:3-75. RUNDEL, P. W., D. J. PARSONS, and E. T. GORDON. 1977. Montane and subalpine vegetation of the Sierra Nevada and Cascade Ranges. In M. G. Barbour and J. Major, eds., Terrestrial vegetation of California, p. 559-599. Wiley-Interscience, NY. STORER, T. I. and R. L. USINGER. 1963. Sierra Nevada natural history. Univ. Cali- fornia Press, Berkeley. WHITTAKER, R. H. 1956. Vegetation of the Great Smoky Mountains. Ecol. Monogr. 26:1-80. . 1967. Gradient analysis of vegetation. Biol. Rev. 42:207-264. NOTES AND NEWS TAXONOMY OF Lomatium bicolor (UMBELLIFERAE).—Lomatium bicolor (Watson) Coulter and Rose belongs to the group of diminutive, usually acaulescent root-tuber geophytes known as tuberous lomatiums (sect. Cous of M. E. Jones, Contr. W. Bot. 12:1-81. 1908). Recent systematic studies of this group have shown that L. bicolor consists of two morphologically and geographically intergrading varieties (Schlessman, Systematics of the tuberous species of Lomatium (Umbelliferae), Ph.D. dissertation, Univ. Washington. 1980). Variety bicolor is a regional endemic of northern Utah, southeastern Idaho and southwestern Wyoming. The range of the more common var. 1982] NOTES AND NEWS 119 leptocarpum extends from northeastern California through eastern Oregon and south- eastern Washington to Idaho, Wyoming, Colorado, and the Grand Canyon of Arizona. Representative specimens of var. bicolor tend to have irregularly thickened roots, numerous (80-250) filiform leaf segments, and relatively broad (1.5-4.5 mm), elliptic fruits with long (1-10 mm) pedicels. Variety /eptocarpum is characterized by a more or less globose-thickened root, leaves with relatively few (20-130), linear segments, and narrow (1.5-3 mm), narrowly elliptic to linear fruits with short (0.1-2 mm) pedicels. Certain specimens from central and eastern Idaho and northeastern Oregon have the numerous filiform leaf segments characteristic of var. bicolor and fruits characteristic of var. leptocarpum. Marcus E. Jones (1908) recognized Lomatium bicolor (Watson) Coulter and Rose as the closest congener of L. leptocarpum (Nuttall ex Torrey and Gray) Coulter and Rose. Mathias and Constance (North American Flora 28:161—295. 1945; Constance, pers. comm.) saw no clear discontinuity between the two taxa and treated them as a single species, L. leptocarpum. However, the epithet bicolor has priority at specific rank and recognition of two varieties requires the new combination presented here. A complete revision of the tuberous lomatiums is in preparation. Lomatium bicolor (Watson) Coulter & Rose var. bicolor, Contr. U.S. Natl. Herb. 7:237. 1900.—Peucedanum bicolor Watson, Bot. King Surv. 129. 1871.—TyYPE: USA, UT, Salt Lake Co., Parley’s Park, Watson 467 (Lectotype: US!; Isolectoptype: GH); Paratypes: GH!, US!).—Cogswellia bicolor M. E. Jones, Contr. W. Bot. 12:33. 1908. Lomatium bicolor var. leptocarpum (Nuttall ex Torrey & Gray) Schlessman, comb. nov.—Peucedanum triternatum B leptocarpum Nuttall ex Torrey & Gray, FI. N. Am. 1:626. 1840.—TyYPE: USA, Oregon (?), Wahlamet (sic) and Columbia Plains, Nuttall s.n. (Lectotype: PH!).—Peucedanum ambiguum var. leptocarpum (Nuttall ex Torrey & Gray) Coulter & Rose, Rev. N. Am. Umbell. 59. 1888.—Lomatium leptocarpum Coulter & Rose, Contr. U.S. Natl. Herb. 7:213. 1900.—Cogswellia leptocarpa M. E. Jones, Contr. W. Bot. 12:33. 1908.—Lomatium ambiguum var. leptocarpum Jepson, Madrono 1:159. 1924. Peucedanum bicolor var. gumbonis M. E. Jones, Contr. W. Bot. 10:55. 1902.—TyYPE: USA, ID, Indian Valley, 15 Jul 1889, M. E. Jones s.n. (Lectotype: US!); Monroe Creek, 20 Apr 1900, M. E. Jones s.n. (Paratypes: BM!, NY!). I thank Lois Arnow, Tom Dieffenbach, Gene Hart, and the curators of the herbaria cited in my dissertation for supplying or loaning specimens. Dan Nicolson provided valuable insight on typification. My field work was supported by a Graduate School Special Fellowship from the University of Washington and by NSF Dissertation Im- provement Grant DEB78-02482.—MArK A. SCHLESSMAN, Dept. of Biology, Box 187, Vassar College, Poughkeepsie, NY 12601. (Received 19 Mar 1981; accepted 12 Jun 1981.) SPREAD OF Filago arvensis L. (COMPOSITAE) IN THE UNITED STATES.—Filago ar- vensis 1s an annual Eurasian weed, typical of sandy pastureland and streamsides. It has been introduced into the USA and Canada, and it is spreading at an exponential rate. This may constitute some cause for alarm because the plant forms dense colonies and has been reported to be unpalatable to livestock by several stockmen who have sent specimens to MONT. Moreover, F. arvensis may be adapting to the diverse en- vironmental conditions of northwestern USA; we have found the plant growing not only in well-used pastures, but forest clear-cuts, rangelands and grainfields as well. 120 MADRONO [Vol. 29 I941-SO I9 71-80 Fic. 1. Spread of Filago arvensis by counties during the six decades from 1921-80 in Idaho, Montana, South Dakota, Washington and Wyoming. (In the 1940s F. arvensis also spread into southeastern British Columbia.) To document the spread and current distribution of F. arvensis we surveyed nine herbaria (ID, IDS, MONT, MONTU, NY, RM, UCB, WSP and WTU) for preserved specimens and identification records of the plant. Additionally, in conjunction with a general weed survey in Montana, we searched for the plant throughout the western half of that state. The migration of F. arvensis is depicted on a county basis in Fig. 1. The species was first collected in North America by A. C. McIntosh in 1926 in Lawrence Co., South Dakota, but was misidentified as Guaphalium palustre. The specimen was deposited in RM, where R. Hartman correctedly identified it in 1979. Because F. arvensis is not listed in The Vascular Plants of South Dakota (T. Van Bruggen, Iowa State Univ. Press. 1976), McIntosh’s collection also represents a state record for South Dakota. A second introduction seems to have been made in the 1930s in the northern Idaho region. 1982] NOTES AND NEWS 121 FILAGO ARVENSIS \. VY = -06744 X—-129-:30 Te OO COUNTIES INFESTED 7 DECADE fs S, I920 1940 i960 (980 TIME (YEARS) Fic. 2. Rate of spread of F. arvensis by counties in the past six decades in Idaho, Montana, South Dakota, Washington and Wyoming. Y = number of counties, X = date. This latter arrival quickly spread in all directions (including into British Columbia), but particularly east- and southeastward into Montana. By slowly spreading westward, the original South Dakota population probably gave rise to the specimens collected in north- eastern Wyoming and southeastern Montana. Evidence for a distant third introduction is the NY specimen collected in Cheboygan Co., Michigan in 1956 (A. Cronquist, in litt.). No mention of F. arvensis was made in several state and regional North American floras that we checked, thus we believe this species to be restricted in the USA to the boundaries shown in Fig. 1 for 1980 (plus northern Michigan). So far, F. arvensis has spread into at least 17 northwestern US counties. Its spread into new counties in the past six decades (1921-80) has occurred at an exponential rate (Fig. 2). From our preliminary analysis of migration rates for 250 alien weed species in the northwestern USA (Forcella and Harvey, New and exotic Weeds of Montana, I and II. Montana Dept. of Agriculture, Helena. 1981) we have concluded that exponential migration rates are characteristic of noxious weeds, whereas ruderals typically spread at approximately linear rates. Thus in the not too distant future we suspect that Filago arvensis may become a problem weed. We thank the curatorial staff of all mentioned herbaria for their aid, particularly Dr. J. H. Rumely (MONT), and appreciate the financial support of the Montana Dept. of Agriculture and APHIS, USDA.—F. ForRcELLA, Division of Plant Industry, CSIRO, Canberra, A.C.T. 2601, Australia, and S. J. HARVEY, Biology Dept., Montana State Univ., Bozeman 59715. (Received 9 Mar 1981; accepted 16 Sep 1981.) 122 MADRONO [Vol. 29 CLIMATE DIAGRAM FOR THE UNIVERSITY OF CALIFORNIA SAGEHEN CREEK FIELD STATION.—The Sagehen Creek drainage basin, e. of the Sierra crest in ne. Nevada and se. Sierra Counties, contains a mixture of forest, scrub, bog, and stream habitats and is the site of various biological research projects (see Madrono 22:115-139. 1973). The climate diagram for Sagehen Creek (Fig. 1) is drawn according to the scheme devised by Heinrich Walter (Vegetation of the earth and ecological systems of the geo-biosphere. Second edition. New York, Springer-Verlag. 1979). Monthly averages for temperature and precipitation are plotted above an abscissa representing the year from Jan (left) through Dec (right). These averages are based on data compiled by E. Alan Cranston (University of California Sagehen Creek Station weather records. Manuscript in field station library. 1970.) for temperatures collected over 16 years (Nov 1952—Mar 1954, Jan 1955—Dec 1969) and for precipitation collected over 10 years (1957, 1961-1969). Several climatological benchmarks are shown on the figure. For the period of obser- vation, mean annual temperature was 4.7°C, and mean annual precipitation was 912 mm (mostly as snow). The highest temperature recorded was 34.4°C (1 Aug 1954), and the mean daily maximum for the warmest month (Jul) was 26.7°C. The mean daily minimum for the coldest month (Jan) was —11.8°C, and the lowest temperature recorded was —33.9°C (24 Feb 1962). The mean daily minimum falls below freezing (0°C) in Sep through May (indicated by black bars below abscissa), and the actual daily minimum may fall below freezing in the remaining months Jun through Aug (diagonal stripes). The dotted area where the precipitation curve falls below the temperature curve indi- SAGEHEN CREEK (1931m) 47° 912mm [16 -10] 200 lOO 80 60 40 ZO Fic. 1. Climate diagram for Sagehen Creek Field Station, elevation 1931 m, average annual temperature 4.7°C, average annual precipitation 912 mm. (See text for expla- nation of scales and other data.) 1982] NOTES AND NEWS 123 cates the approximate period of water stress to plants (late Jun—late Sep, with the slight drought sometimes broken by summer thundershowers). Sagehen Creek drainage basin ranges in elevation from ca. 1860 m to 2672 m. The weather station itself (located ca. 120°14'30”W and 39°25’'45”N according to U.S. Geol. Surv. Truckee 15’ quadrangle. 1955) lies at 1931 m. Thus, the climate diagram is representative more of conditions in the lower part of the basin. Though but one of several methods for representing climate (see, e.g., Barbour, M. G., et al. 1980. Ter- restrial plant ecology. Menlo Park, Benjamin/Cummings Publishing Co., Inc.), this climate diagram for Sagehen Creek may be compared to more than 8000 such diagrams for stations throughout the world (Walter, Heinrich, and Helmut Lieth. 1960-1967. Klimadiagramm-Weltatlas. 3 parts. Jena, VEB Gustav Fischer Verlag).—DALE E. JOHNSON, 12283 Ranch House Rd, San Diego, CA 92128. (Received 24 Aug 1981; accepted 28 Sep 1981.) NOTEWORTHY COLLECTIONS California ERIOPHYLLUM NUBIGENUM Green ex. Gray (ASTERACEAE).—Mariposa Co., Yosem- ite Natl. Park, near the top of Chilnualna Falls, 1890 m, 2 Jun 1980, Botti 26 (YNP). Previous knowledge. Three locations in Yosemite National Park. Significance. The last recorded sighting or collection of this taxon was 4 Jun 1897 by J. W. Congdon at Chilnualna Falls in Yosemite. Its rediscovery at the same site after 83 years ended a lengthy search for this taxon, which was believed to be “possibly extinct” according to a CNPS Plant Status Report, 1977. Subsequently the site of the collection by Mrs. Dodd, 1891 (UC) in Little Yosemite Valley and three other previously unknown sites were located.—STEPHEN J. BOTTI, Resources Management Specialist, Yosemite Natl. Park, CA 95389. (Received 23 Feb 1981; accepted 12 Mar 1981.) MIRABILIS LAEVIS (Benth.) Curran (NYCTAGINACEAE).—Alameda Co., rocky slope on the nw. ridge of Mission Pk, Fremont, ca. 775 m. Colony appears to be fairly extensive. NV. Havlik 929 (UC); N. Havlik 930 (CAS). Significance. First record for Alameda Co., range extension some 200 km n.—NEIL HAVLIK, E. Bay Reg. Park Distr., 11500 Skyline Blvd, Oakland, CA 94619. (Received 3 Jul 1981; accepted 16 Dec 1981.) Montana HOWELLIA AQUATILIS Gray (CAMPANULACEAE).—Missoula Co., sw. side of Lind- bergh Lake Rd. 0.8 km nw. of Swan River crossing (NE%4 S7 T19N R16W), 1240 m, 29 Jul 1978, McCune 2287 (MONTU, MONTU at Yellow Bay, PH). (Determined by A. E. Schuyler, Jul 1978.) Significance.—First record of Howellia in MT and its easternmost record. Howellia 124 MADRONO [Vol. 29 is listed as rare and endangered in CA, ID, OR, and WA.—BrucE McCuNE, Dept. Botany, Univ. Wisconsin, Madison 53706. (Received 30 Apr 1981; accepted 20 Aug 1981.) Wyoming CAREX BIPARTITA All. (CYPERACEAE).—Park Co., Beartooth Plateau, head of Wy- oming Cr. (T58N R104W S21 SE'%), 3172 m, 21 Jul and 22 Aug 1980, Evert 2088 and 2428 (RM); Absaroka Range, at head of West Blackwater Cr. (T51N R107W S30 NW)4), 3175 m, 13 Aug 1980, Evert 2384 (RM); Absaroka Range, 0.8 km se. of Chaos Mtn. (TS5ON R108N S30 NW14), 3111 m, 28 Jul 1980, Evert 2250 (RM). Occasional on tundra in wet soil along streamlets. Previous knowledge. Wet places at high altitudes, CO, UT, and MT n. to AK, circumpolar; also in New Zealand. Significance. First reports for WY. The three WY localities are 6 km, 77 km and 85 km respectively from the nearest known population in Carbon Co., MT. CAREX DEWEYANA Schwein. (CYPERACEAE).—Park Co., 0.4 kme. of East Entrance of Yellowstone Natl. Park (T52N R109W S8 SW), 2104 m, 2 Jul 1976, Evert 1083 (RM); along Grinnell Cr., 2.4 km n. of Hwy 14 (T52N R108N S6 NW14), 2287 m, 9 Aug 1980, Evert 2343 (RM). Occasional in moist meadow and open woodland. Previous knowledge. Open woods and stream banks, Labrador to B.C. and s. and w. to PA, IA, SD, NM, AZ and CA. Significance. First reports for WY. About 247 km from the nearest known population in Rosebud Co., MT. CAREX INCURVIFORMIS Mack. (CYPERACEAE).—Park Co., Absaroka Range ne. end of Wapiti Ridge (T51N R105W S19 NE'%), 3355 m, 29 Jul 1979, Evert 1538 (RM). On tundra, among boulders near the edge of a vernal pool. Previous knowledge. Alpine ledges and turf, CA, CO and MT, n. to Alta. and AK, circumpolar; also in S. America. Significance. First report for WY. About 350 km from the nearest known population in Deer Lodge Co., MT. DIANTHUS BARBATUS L. (CARYOPHYLLACEAE).—Park Co., Aspen Cr. and Hwy 14, ca. 48 km w. of Cody (T52N R107W S23 NE'%), 1860 m, 18 Jul 1980, Evert 2129 (RM). Several colonies have been observed for eight years in woods near cabin and appear to be spreading. Previous knowledge. Native of Europe, cultivated and spreading to roadsides and groves; locally established from s. Que. to s. B.C., s. to DE and w. to ND, w. MT, w. OR, w. WA and n. CA. Significance. First report for WY. About 432 km from the nearest known population in Missoula Co., MT. GENTIANELLA PROPINQUA (Rich.) Gillett GENTIANACEAE).—Park Co., Absaroka Range, near Flora Lake (T51N R109N S25 NW), 2958 m, 20 Aug 1972 and 17 Aug 1980, Evert 178 and 2404 (RM). In moist subalpine meadow and spruce-fir groves. Previous knowledge. Meadowlands, streambanks and woods, AK s. to B.C., Alta., OR and MT, e. to Que. and Labrador. 1982] NOTEWORTHY COLLECTIONS 125 Significance. First report for WY. About 257 km from the nearest known population in Beaverhead Co., MT. This is apparently the most southerly station for this species in N. Amer. MYOSOTIS ARVENSIS (L.) Hill (BORAGINACEAE).—Park Co., along Fishhawk Cr., ca. 0.4 km s. of Hwy 14 (T52N R108W S27 SW'4), 1952 m, 5 Jul 1980, Evert 2015 (RM). In disturbed area along trail. Previous knowledge. Eurasian, established in fields and roadsides, Nfld. to Ont. and MN and s. and w. to NJ, WV and PA, local in MT, OR, WA, B.C., Man. and Sask. Significance. First report for WY. A range extension of ca. 422 km from the nearest known population in Missoula Co., MT. MyosoTIS MICRANTHA Pall. (BORAGINACEAE).—Park Co., along Hwy 14, at Fire Fighter’s Memorial ca. 51 km w. of Cody (T52N R107W S21 SW%), 1891 m, 29 Jun 1978, Evert 1258 (RM). In Moist disturbed area. (Synonym: M. stricta Link.) Previous knowledge. Road sides and old fields naturalized from Europe, s. Que. to s. B.C. and s. and w. to VA, IA, MT, ID, OR and CA. Significance. First report for WY. A range extension of about 400 km from the nearest known population in Ravalli Co., MT. POTENTILLA RECTA L. (ROSACEAE).—Park Co., 0.4 km w. of Wapiti Ranger Sta- tion, Shoshone Natl. For. (T52N R106W S21 NE'%), 1800 m, 21 Jul 1980, Evert 2134 (RM). Common in irrigated pasture. Fremont Co. (T31N R10W S24), 2165 m, 1 Sep 1977, Roth 27 (Central Wyoming Community College). Previous knowledge. Eurasian, naturalized in fields and roadsides, widespread in e. U.S. and Can., locally w. and s. to NE, SD, n. CO, TX, MT, ID and e. WA. Significance. First reports for WY. About 390 km from the nearest known population in Ravalli Co., MT.—ERWIN F. EVERT, 1476 Tyrell Ave., Park Ridge, IL 60068. (Received 2 Apr 1981; accepted 24 Aug 1981.) ANNOUNCEMENT CBS GRADUATE STUDENT AWARD, 1981 The award for the outstanding paper presented at the 1981 California Botanical Society Graduate Student Meetings, held at San Francisco State University on 24 Oct, was won by Luann Z. Wright, Department of Botany, San Diego State University. The title of her presentation was SHOOT GEOTROPISM: ITS RELATIONSHIP TO THE ACID- GROWTH THEORY. ANNOUNCEMENT The Society for Economic Botany will hold its 23rd annual meeting at the University of Alabama in University, Alabama, 14-17 June 1982. Featured will be a symposium entitled “U.S. OILSEEDS INDUSTRY-—GERMPLASM TO UTILIZATION.” Fur- ther information can be obtained from C. Earle Smith, Jr., Anthropology, Box 6135, University of Alabama, University, AL 35486. _ SUBSCRIPTIONS—MEMBERSHIP Membership in the California Botanical Society is open to individuals ($18 per year; students $10 per year for a maximum of seven years). Members of the Society receive MADRONO free. Family memberships ($20) include one eight-page publishing allotment and one journal. Emeritus rates are available from the Corresponding Secretary. In- stitutional subscriptions to MADRONO are available ($25). Membership is based on a calendar year only. Applications for membership (including dues), orders for sub- scriptions, requests and rates for back issues, changes of address, and undelivered copies of MADRONO should be sent to the Corresponding Secretary. Renewal payments should be sent to the Treasurer. INFORMATION FOR CONTRIBUTORS Manuscripts submitted for publication in MADRONO should be sent to the editor. All authors must be members, and membership is prerequisite for review. Manuscripts and review copies of illustrations must be submitted in triplicate for all articles and short items intended for NOTES AND NEWS. Follow the format used in recent issues for the type of item submitted and allow ample margins all around. All manuscripts MUST BE DOUBLE SPACED THROUGHOUT. For articles this in- cludes title (all caps, centered), author names (all caps, centered), addresses (caps and lower case, centered), abstract, 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, and number each page. Do not use a Separate cover page, “erasable” paper, or footnotes. Table captions should include all information relevant to tables. All measurements should be in metric units. Line copy illustrations should be clean and legible, proportioned (including captions) to the MADRONO page, and designed for reduction to % original size. 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 latitude and longitude references. Halftone copy should be designed for reproduction at actual size. In no case should original illustrations be sent prior to the acceptance of a manuscript. When needed they should be mounted on stiff cardboard and sent flat. No illustrations larger than 22 X 28 cm will be accepted. Presentation of nomenclatural matter (accepted names, synonyms, typification) should follow the format used for Rhus integrifolia in MADRONO 22:288. 1974. Institutional abbreviations in specimen citations should follow Holmgren and Keuken, Index Her- bariorum, 6th edition. Abbreviations of serial titles should be those in Botanico-Per- iodicum-Huntianum (Lawrence et al., 1968, Hunt Botanical Library, Pittsburgh). If the correct abbreviation cannot be determined, the full serial title should be used. Titles of books should be given in full, together with the place and date of publication, publisher, and edition, if other than the first. . All members of the California Botanical Society are allotted eight pages in MADRONO over a two year period. Joint authors are each assessed the full page number. Beyond that number of pages a required editorial fee of $40.00 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 sched- ule, with equity among all members for access to its pages. Printer’s fees for illustrations and typographically difficult material @ $35.00 per page (if their sum exceeds 30 percent of the paper) and for author’s changes after typesetting (@ $3.00 per line will be charged to authors. CALIFORNIA BOTANICAL SOCIETY MADRONO WEST AMERICAN JOURNAL OF BOTANY A PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL VOLUME 29, NUMBER 3 Contents AGE AND ORIGIN OF THE MONTEREY ENDEMIC AREA, Daniel I. Axelrod MOUNTAIN MEADOWS: STABILITY AND CHANGE, Nathan B. Benedict PHENOLOGY, GERMINATION, AND SURVIVAL OF DESERT EPHEMERALS IN DEEP CANYON, RIVERSIDE COUNTY, CALIFORNIA, Jack H. Burk THE VEGETATION OF THE RAE LAKES BASIN, SOUTHERN SIERRA NEVADA, Mary T. Burke PINE SEEDLINGS, NATIVE GROUND COVER, AND LOLIUM MULTIFLORUM ON THE MARBLE-CONE BURN, SANTA LUCIA RANGE, CALIFORNIA, James R. Griffin FLORISTIC AFFINITIES OF THE HIGH SIERRA NEVADA, G. Ledyard Stebbins A GRADIENT PERSPECTIVE ON THE VEGETATION OF SEQUOIA NATIONAL PARK, CALIFORNIA, John L. Vankat REVIEWS ANNOUNCEMENT NOTEWORTHY COLLECTIONS GALAPAGOS ISLANDS NEw MeExIcoO—TEXAS CALIFORNIA NOTES AND NEWS JULY 1982 127 148 154 164 aA 189 200 215 216 ZV 2M 218 218 SOCIETY MADRONO (ISSN 0024-9637) is published quarterly by the California Botanical Society, Inc., and is issued from the office of the Society, Herbarium, Life Sciences Building, University of California, Berkeley, CA 94720. Subscription rate: $25 per calendar year. Subscription information on inside back cover. Established 1916. Second-class postage paid at Berkeley, CA, and additional mailing offices. Return requested. POSTMASTER: Send address changes to Susan Cochrane, California Natural Diversity Data Base, Calif. Dept. of Fish & Game, 1416 9th St. Rm. 1225, Sacramento, CA 95814. Editor—CHRISTOPHER DAVIDSON Idaho Botanical Garden P.O. Box 2140 Boise, Idaho 83701 Board of Editors Class of: 1982—DEAN W. TAYLOR, University of California, Davis RICHARD VOGL, California State University, Los Angeles 1983—ROBERT W. CRUDEN, University of Iowa, Iowa City DUNCAN M. PoRTER, Virginia Polytechnic Institute and State University, Blacksburg 1984—Mary E. BARKWORTH, Utah State University, Logan Harry D. THIERS, San Francisco State University, San Francisco 1985—STERLING C. KEELEY, Whittier College, Whittier, CA ARTHUR C. GIBSON, University of California, Los Angeles 1986—AMyY JEAN GILMARTIN, Washington State University, Pullman ROBERT A. SCHLISING, California State University, Chico CALIFORNIA BOTANICAL SOCIETY, INC. OFFICERS FOR 1982 President: WATSON M. LAETSCH, Department of Botany, University of California, Berkeley 94720 First Vice President: ROBERT ROBICHAUX, Department of Botany, University of California, Berkeley 94720 Second Vice President: VESTA HESSE, P. O. Box 181, Boulder Creek, CA 95006 Recording Secretary: ROBERT W. PATTERSON, Department of Biology, San Francisco State University, San Francisco, CA 94132 Corresponding Secretary: SUSAN COCHRANE, California Natural Diversity Data Base, Calif. Dept. of Fish & Game, 1416 9th St. Rm. 1225, Sacramento, CA 95814 Treasurer: CHERIE L. R. WETZEL, Department of Biology, City College of San Francisco, 50 Phelan Ave., San Francisco, CA 94112 The Council of the California Botanical Society consists of the officers listed above plus the immediate Past President, ROBERT ORNDUFF, Department of Botany, Uni- versity of California, Berkeley 94720; the Editor of MADRONO; three elected Council Members: LYMAN BENSON, Box 8011, The Sequoias, 501 Portola Rd., Portola Valley, CA 94025; JOHN M. TUCKER, Department of Botany, University of California, Davis 95616; CHARLES F. QUIBELL, Department of Biological Sciences, Sonoma State Col- lege, Rohnert Park, CA 94928; and a Graduate Student Representative, CHRISTINE BERN, Department of Biology, San Francisco State University, San Francisco, CA 94132. AGE AND ORIGIN OF THE MONTEREY ENDEMIC AREA DANIEL I. AXELROD Department of Botany, University of California, Davis 95616 ABSTRACT The Monterey endemic area originated during late Wisconsin and Holocene times. Widespread sandy substrates afforded sites for the origination, by various modes, of new species following population movements into the region. Their sources included taxa a) from forest communities that earlier were in the lowlands under cooler, moister Pleistocene climates, b) from widespread species along the central coastal strip, and c) from interior and southern alliances that probably invaded the area during the Xero- thermic. In addition, paleoendemics of wider Quaternary distribution found a haven here under the influence of the Monterey submarine canyon on land climate; it provided an effective local foggy shield from Xerothermic conditions. The paleoendemics include two cypresses and a Pinus radiata population with relatively small cones. By contrast, the P. radiata populations at Ano Nuevo and Cambria, farther removed from a more persistent foggy shield, apparently developed larger cones in response to Xerothermic aridity. Of the diverse endemic regions in California (Stebbins and Major 1965), the Monterey area poses a unique problem. Although it is much smaller than the mountainous Santa Cruz and Santa Lucia endemic areas directly north and south, it is richer in endemics even though those areas are far more diverse in relief, climate, and rock-soil type. Furthermore, there are no serpentine exposures in the Monterey area to help account for its high endemism, as in the Santa Cruz or Mt. Tamalpais areas. In addition, whereas the larger endemic areas in the northern (Siskiyou) and southern (Diegan) parts of the state have a relatively high incidence of warm season precipitation that appears to account for the persistence of paleoendemics there (Raven and Axelrod 1978), this is not true of the Monterey area. This low-lying region not only has a much lower total rainfall than endemic areas to the north and south in the outer Coast Ranges, it receives less warm season precipitation than those areas. Furthermore, the occurrence of quartz diorite at Monterey-Carmel can scarcely explain the high endemism there. The same rock makes up the northern half of Inverness Ridge and Montara Mountain, north and south of San Francisco. Both are nearly equal in size to Monterey-Carmel endemic area yet few endem- ics occur there. Clearly, other factors must account for the occurrence in this local region of the highest number of endemics in the outer central Coast Ranges. To provide a basis for interpreting its history, some of the general features of its flora and vegetation are recalled briefly (Howitt and Howell 1963, 1973; Stebbins and Major 1965). All specific names are those in Munz (1973). MADRONO, Vol. 29, No. 3, pp. 127-147, 9 July 1982 128 MADRONO [Vol. 29 FLORA AND VEGETATION Monterey County is a meeting place for plants from the north and south. Howitt and Howell (1963, 1973) report that of the 1400-odd native species, about 150 and 160 reach their northern and southern limits there, respectively, excluding the flora of the higher Santa Lucia Mountains. In the Monterey endemic area members of the closed-cone pine forest include the notable paleoendemics Cupressus goveniana and C. macrocarpa with very restricted ranges (see Griffin and Critch- field 1972) and Pinus radiata, apparently a relict, ancestral population (see below). In addition, the area has six endemic species of Arc- tostaphylos as well as other taxa listed in the Appendix. Inland from the coast, the Salinas and Carmel Valleys are covered with coastal sage on drier slopes that alternates with a Quercus agrifolia woodland- grassland on moister flats and north-facing slopes. Farther up Carmel Valley climate rapidly becomes more continental and Quercus agrifolia woodland-grassland gives way to Quercus lobata-Q. douglasit wood- land-grassland, with Pinus sabiniana joining them farther inland. In this interior area, commencing about 20—25 km southeast of Monterey, there are numerous additional species that have a wider distribution in the inner Coast Ranges but do not extend into the coastal strip. Among these are (for others, see Howitt and Howell 1963, 1973; Lins- dale 1955): Aesclepias eriocarpa Orthocarpus attenuatus Amorpha californica Pectocarya penicillata Baccharis viminea Penstemon centranthifolius Calochortus splendens Penstemon heterophyllus Camissonia graciliflora Phacelia brachyloba Cercocarpus betuloides Phacelia douglasit Cirsium proteanum Plagiobothrys canescens Cryptantha microstachys Plagiobothrys nothofulvus Emmenanthe pendulifiora Prunus ilictfolia Ertodictyon californicum Quercus douglasit Gilia achilleifolia Quercus dumosa Gilia tenuiflora Quercus lobata Lasthenia chrysostoma Rhamnus ilictfolia Linanthus bicolor Salvia columbariae Malacothrix clevelandit Sanicula bipinnata Microseris elegans Senecio douglasit Microseris heterocarpa Trichostema lanatum Howitt and Howell (1963, p. 17) note that a number of these species, as well as those in the flora a few km southeast in Salinas Valley and adjoining interior valleys, have affinities with the Mohave Desert flora, a relationship also displayed by the floras of the inner Santa Cruz Mountains (Thomas 1961) and the Mt. Hamilton Range (Shar- 1982] AXELROD: MONTEREY ENDEMIC AREA 129 smith 1945). This interior aspect of the flora, so evident in middle and upper Carmel Valley, is reflected also by the large, relict stand of Pinus sabiniana in Pine Canyon eight miles south of Salinas. It sug- gests that the community probably had a wider occurrence on the seaward slopes of the outer Coast Ranges in the Xerothermic, of which the stand near Gorda also appears to be a relict (Axelrod 1966, p. 50). A number of taxa in the coastal Monterey region extend into the drier, inner Coast Ranges, where they are prominent members of the flora. Among these are (for others, see Howitt and Howell 1963, 1973): Aesculus californicus Lepidium nitidum Allium hickmanit Lupinus albifrons Athysanus pusillus Lupinus nanus Calochortus luteus Mahonia pinnata Chorizanthe coriacea Malacothamnus fasciculatus Clarkia cylindrica Orthocarpus purpurascens Clarkia lewtsit (=bottae) Platystemon caltfornicus Collinsia heterophylla Potentilla glandulosa Convolvulus subacaulis Ranunculus californicus Datisca glomerata Ribes malvaceum Delphinium patens Solanum umbelliferum Eriastrum denstfolium Trifolium gracilentum Eschscholzia californica Viola pendunculata This summary suggests that the Monterey endemic area recently has accumulated its unique taxa from diverse sources. Clearly, the area is a haven for conifers that had a wider distribution along the coast in the past. The Appendix shows that other endemics are allied to species now in mixed conifer forests to the north that probably invaded the lowlands of the area during the Pleistocene when rainfall was higher and temperature lower than they are now in the Monterey area. These endemics occur chiefly on sandy substrates, notably old dunes, elevated floodplains, and old coastal terraces. Also in similar sites are taxa allied to species that are widespread along the central California coast. In addition, another group has affinities with pres- ently interior and southern species that imply that a warmer, drier climate affected the area following the last major glacial-pluvial stage. ENDEMIC TAXA WITH FOREST AFFINITIES Some Monterey endemics are allied to taxa in the mixed conifer and Douglas fir forests to the north, including species of Allium, Arc- tostaphylos (4 spp.), Lupinus, Ribes, and Trifolium (Fig. 1). Although these forests occur chiefly at elevations above 750 m in the Coast Ranges to the north, they were at sea level in the San Francisco Bay region during the late Pleistocene. This is shown by the San Bruno flora (Potbury 1932), dated at 10,000 BP. It represents a Pseudotsuga 130 MADRONO [Vol. 29 forest like that now near Inverness, 60 km north in Marin County, where rainfall is 800 mm or more as compared with half that at the fossil site today. In addition, a small flora near Mountain View at the southwest corner of San Francisco Bay, dated at 20,000 to 23,000 BP, contains Calocedrus, Cupressus, Pinus, and Pseudotsuga, indicating that mixed conifer forest was then at sealevel (Helley et al. 1972), and that a much cooler, moister climate was then in the area (for estimate, see Axelrod 1981). There was a similar climate there in the late Plio- cene-early Pleistocene, as shown by the Santa Clara flora near Sara- toga, with Calocedrus, Pseudotsuga, and Pinus cf. lambertiana (Dorf 1930). Under the cooler, moister climate of the late Pliocene and early Pleistocene, mixed conifer forest ranged far to the south, occupying the lowlands of interior southern California (Axelrod 1966). The pres- ent forest patches in the central and south Coast Ranges, including the higher Santa Lucia Mountains, probably are remnants of that time. Inasmuch as the ranges were largely elevated during the late Pliocene and Pleistocene (Christensen 1966), it would appear that the mixed conifer forests that formerly blanketed the lowlands were, in essence, elevated bodily to their present moister, cooler sites where they persisted as climate became drier and warmer over the lowlands. Climatic-topographic differences over the area during the later Pleis- tocene, especially in the Xerothermic, probably account for the dis- continuous distributions of the forest taxa in the higher Coast Ranges today. Many taxa in the Monterey area that evidently originated from older, mesic forest species occur in an area of lower rainfall than that in which their presumed ancestors lived. Whereas the Monterey area receives 450-500 mm yearly, rainfall is two and three times that amount in areas where the inferred ancestral, or closely related, taxa now live. The high incidence of summer fog compensates for low rainfall to some degree. However, it appears that the usual habitat of many of the endemics on dunes and sandy areas may account for their persistence. Contrary to popular notion, dunes and old elevated ter- races are not dry. Even dunes do hold water (Bagnold 1941, p. 245- 246), the amount depending on the internal structure (bedding) of the dune, the degree of compaction, and the size of the grains, which determines porosity. The dunes in Clatsop County near Seaside in the northwest corner of Oregon are capable of storing 80 percent of the annual precipitation, which there averages 2000 mm annually (Frank 1968). In the desert of southeastern California, subsurface sands of the Eureka Dunes (Pavlik 1979, 1980), Kelso Dunes (Sharp 1966), and Algodones Dunes (Norris and Norris 1961) are moist well into summer. Pavlik showed that endemics on Eureka Dunes are in fact mesophytes as compared with nearby desert plants bordering the Eureka Dunes, situated in the lee of the Inyo-White Mountains. He also reported 1982] AXELROD: MONTEREY ENDEMIC AREA 13d Monterey Endemics ee allied to forest taxa pee | Allium hickmanii sg Arcto. edmundsii LZ Arcto. montereyensis LO Arcto. pajaroensis LES, Arcto. pumila C4 Ceanothus rigidus 6 a \ - * Chorizanthe pungens Ribes menziesii v. hystrix MONTEREY = \ \ \ \\ Fic. 1. Monterey endemics allied to forest taxa occur chiefly in the mountains to the north, under wetter and cooler climate. (pers. comm., Dec 1980) that in summer moist sand occurs in Eureka Dunes fully 150 m above the desert floor. Clearly, while the Monterey endemics derived from forest taxa are in an area of low precipitation, they persist on sandy sites presumably because they are relatively mesic. Moisture comes from winter rainfall, from drainage from bor- dering hills where precipitation is higher, and from the dense, wet fog that commonly hugs the ground during summer. In addition, non- marine terrace deposits regularly have silty or clayey lenses that act as water tables. 132 MADRONO [Vol. 29 The Monterey occurrence of numerous endemics on sandy sub- strates is not unique. Pinus torreyana, situated between Del Mar and La Jolla, is confined to the Linda Vista Terrace and Eocene Torrey Sandstone. Road cuts and the walls of gulches in these formations are wet and seep water late in the year. The population of P. torreyana on Santa Rosa Island is confined to the predominantly sandy Santa Margarita Formation which also retains ample moisture through the year. The elevated terraces between Nipomo and Lompoc have several woody endemics, including Arctostaphylos purissima, A. rudis, A. viridissima, Ceanothus impressus, Ertodictyon capitatum, and Mal- vastrum gracile. In addition, Arctostaphylos morroensis is confined to sandy hills south of Morro Bay. Other woody taxa on elevated terraces include the following that have wider distributions: Arctostaphylos cruzensis—San Luis Obispo Co. north to southern Monterey Co.; Arc- tostaphylos tomentosa var. crassifolia—Oceanside to San Diego; Bac- charis pilularis—Russian River to Pt. Sur, Monterey Co.; Ceanothus dentatus—Monterey south to San Simeon; Ceanothus gloriosus—Ma- rin Co. north to Pt. Arena; Ceanothus ramulosus—Burton Mesa, Ni- pomo Mesa, and north to Monterey Co.; Ceanothus verrucosus—San Diego Co. and southward; Eriogonum parvifoltum—Monterey to San Diego Co.; Haplopappus ericoides—Bolinas, Marin Co., south to Los Angeles Co.; Helianthemum scoparium—Mendocino Co. south to Santa Barbara Co.; Lupinus arboreus—Del Norte Co. south to Santa Barbara Co.; Lupinus chamissonis—San Francisco south to Los An- geles Co.; Lupinus variicolor—Humboldt Co. south to San Luis Obis- po Co. The preceding lists are not intended to be complete, and an- nuals and herbaceous perennials have not been considered. The data do indicate that sandy substrates, especially stabilized dunes and el- evated terraces, probably support endemic taxa because of the ade- quate moisture provided by these environments. Apart from the mois- ture factor, it is apparent that plants in these sites may be removed from competition. This may be due in part to greater moisture as compared with bordering environments, and also to the problem of moving sand and hence establishment and continued growth on a relatively unstable substrate. It is also likely that they can tolerate better the nutrient-poor sands and thereby escape competition. ENDEMIC TAXA WITH COASTAL AFFINITIES Some of the endemic taxa (Appendix) appear to be members of closely related species groups that occur along the coast, as exemplified by Chorizanthe, Eriogonum, Erysimum, Haplopappus, and Mimulus (Fig. 2). The endemics may have originated by geographic isolation, or other means, in the later Quaternary chiefly. The two varieties of Arctostaphylos tomentosa, as well as Ceanothus griseus var. horizon- talis, probably belong to this group with coastal affinities. 1982] AXELROD: MONTEREY ENDEMIC AREA 133 Monterey Endemics allied to coastal taxa Arcto. tomentosa f. trichoclada a ve f. hebeclada Ceanothus griseus v. horizontalis Eriogonum parvifolium ssp. lucidum Erysimum ammophilum Haplopappus eastwoodiae Lupinus tidestromii Malacothrix saxitilis v. arachnoidea Mimulus guttatus v. arenicola Trifolium polydon Fic. 2._ A number of Monterey endemics are related to species of primarily coastal occurrence. ENDEMIC TAXA WITH INTERIOR AND SOUTHERN AFFINITIES Some Monterey endemics (Allium hickmani1, Arctostaphylos mon- tereyensis) have isolated occurrences in the interior. Whether they are invading that area, or retreating from it, is not presently clear. A number, however, do have affinities with taxa that are in the interior or to the south in warmer regions, as Cordylanthus, Corythrogyne, Delphinium, Eriastrum, Gilia, Malacothrix, and Triteleia (Fig. 3). As with the group above, they differ from their presumed ancestral, or allied, taxa in only minor ways, often of subspecies or varietal nature. 134 MADRONO [Vol. 29 Monterey Endemics allied to interior taxa Arctostaphylos hookeri Corethrogyne_ filaginifolia v.viscidula Cordylanthus littoralis Delphinum hutchinsoniae Eriastrum virgatum Gilia tenuiflora ssp. arenaria Triteleia versicolor \ \N 3. MONTEREY < Fic. 3. Some Monterey endemics are allied to taxa that occur chiefly in the interior and to the south. They probably were derived from interior and southern taxa that in- vaded the coastal strip during the Xerothermic. The effect of this warmer drier climate of 8000 to 4000 years ago (see Axelrod 1966; 1981, p. 851) also appears to clarify the puzzling relationships dis- played by the Pinus radiata populations at ANo Nuevo and Cambria north and south of Monterey, respectively, as well as the reason for the persistence of the narrow endemics Cupressus macrocarpa and C. goveniana in the area. 1982] AXELROD: MONTEREY ENDEMIC AREA 135 PALEOENDEMICS AT MONTEREY Three conifers are restricted to the Monterey area. Cupressus go- veniana and C. macrocarpa have very limited ranges (see Griffin and Critchfield 1972) and Pinus radiata does not extend beyond the near- coastal sector around Monterey-Carmel. All of them have fossil rec- ords north and south of Monterey (Fig. 4). As outlined elsewhere (Axelrod, MS), the pine and cypresses were restricted southward by colder Wisconsin-age climate and were confined northward as drier, warmer climate spread during the Holocene. That their survival in the Monterey area probably can be ascribed to local environment is suggested by the interrelationships of the central California popula- tions of Pinus radiata. EVOLUTION OF PINUS RADIATA POPULATIONS The Pinus radiata population at Monterey has a smaller mean cone size than those at Ano Nuevo and Cambria, which are in wetter and drier areas to the north and south (Axelrod 1980, pl. 1 and fig. 2). Cones the size of the Monterey population have been recovered at Drakes Bay to the north and near Pt. Sal to the south (Axelrod 1980). In addition, smaller-coned P. radiata populations like those of the Guadalupe Island stands have been found at a number of sites of various ages (5 m.y. to 20,000 years) in the coastal region (Fig. 4). However, the larger-coned P. radiata populations at Ano Nuevo and Cambria are not now known to have a fossil record. Cone-size of the fossil Carpinteria population (radiocarbon age older than 40,000 years) is intermediate between those produced by the Monterey and Ano Nuevo stands (Axelrod 1980, fig. 2). This probably reflects the warmer climate in that very different floristic-climatic province. An earlier increase in cone size is expectable there in view of the semiarid taxa (Arctostaphylos glauca, Juniperus californica, Pinus sabiniana) that covered steep dip-slopes of Eocene and Oligocene sandstones directly east of the pine forest that inhabited the floodplain, taxa that indicate a climate warmer than that now in the region. The trend to larger-coned P. radiata populations in central Cali- fornia probably was an adaptation to increasing summer drought (Axelrod 1980), for the larger seeds would favor seedling establishment (see Baker 1972), especially under the progressively more extreme mediterranean-type climate that was emerging during the later Pleis- tocene. This suggests that the Cambria population may have origi- nated in or near its present area during the Xerothermic. It now in- habits the driest area of the three California populations, an area that presumably was warmer and drier during the Xerothermic as judged from other evidence in California (Axelrod 1966, 1981). This is implied also by the occurrence of the typically interior Pinus sabiniana in the 136 MADRONO [Vol. 29 120° O;+ €9opn 4oe Ne “ads . Oo Millerton ® 6 i tx Drakes Bay ho. — @ Spring V.Lakc 5 e MONTEREY * Little Sur—e ENDEMIC AREA 35° Ge * Pi sal Veronica Spr. ° Cc nleri Oo i arpinteria °o —Seaclifse O Willow Cr. _ + ; Rancho La Brea® 35 % NS \ @ Mi. Eden Arizona 70 Fic. 4. Fossil occurrences of paleoendemics. Pinus radiata, * Monterey population, @ Guadalupe Island population cone size; + Cupressus macrocarpa; O Cupressus go- ventana. Ages of localities are Mt. Eden, 5 m.y.; Seacliff, 1 m.y.; Veronica Springs Quarry, 1 m.y.; Drakes Bay, ? 0.75-1.0 m.y.; Carpinteria older than 40,000 BP; Rancho La Brea, 30,000 BP; Millerton, 28,000 Bp; Pt. Sal, 28,000 BP; Willow Creek, 15,000 BP; Little Sur, 10,000 BP. nearby region, and of shrubs in the adjacent South Coast Ranges that are disjunct from Los Angeles and Riverside Counties (Adenostoma sparstfoltum, Quercus dunnii), or which occur elsewhere in the coastal strip from Santa Barbara County southeastward (Ceanothus spinosus). The Ano Nuevo population may also have originated locally in response to the spread of drier, warmer Xerothermic climate. Its in- 1982] AXELROD: MONTEREY ENDEMIC AREA 137 ° Las Vegas Reno J Giant Forest Palm Springs ° ° Disneyland e ANO NUEVO MONTEREY CAMBRIA mean length mean length 9.5 cm 14.0 cm mean length 11.5 cm tS Fic. 5. Illustrating mean cone size of the California populations of Pinus radiata (from Axelrod 1980). Size of Monterey submarine canyon greatly exaggerated; see Fig. 6. Endemic areas in the outer central Coast Ranges are: 1. Pitkin-Bodega, 2. Tamalpais, 3. Santa Cruz, 4. Monterey, 5. Santa Lucia (see Stebbins and Major 1965, fig. 5). fluence may be judged from the flora of the inner Santa Cruz Moun- tains, 25 to 30 km east. Some 60-odd taxa from the inner south Coast Ranges reach their northern limit of distribution there (Thomas 1961). Since Pliocene and late Pleistocene floras from the inner Santa Cruz Mountains show that mesic, mixed conifer forest and Douglas fir forest occupied the adjacent lowlands under a climate much wetter and cool- er than that now there (Dorf 1930, Potbury 1932, Helley et al. 1972; see above), the xeric southern taxa must have entered the inner Santa Cruz Mountains following the last glacial-pluvial, that is, during the Xerothermic (Axelrod 1981). Because the Ano Nuevo grove is (and was) in a more humid region than the Cambria population, the influ- 138 MADRONO [Vol. 29 ence of Xerothermic climate was not so great there, and hence cones of the Ano Nuevo groves are not so large as those of the Cambria. However, they are larger than those produced by the Monterey pop- ulation (Fig. 5), situated only 65 km south. Immunochemical comparisons of seed proteins from the three Cal- ifornia populations of P. radiata support the idea that they diverged recently, despite the great differences in mean cone size (Murphy 1981). The immunochemical differences measured between pairs of the California populations were approximately equal, but significantly less than differences measured between California populations and those on Guadalupe and Cedros Islands. If the P. radiata populations to the north and south were affected by the Xerothermic dry climate, how did the Monterey population escape? Is this population, which had a wider late Pleistocene distribution, an ancestral, relictual pop- ulation that persisted under a favorable local climate, one that was less affected by warm, dry Xerothermic climate? Persistence under local climate. The supposition that the origin and distribution of the P. radiata populations have been controlled by local climate is consistent with the influence of topography on climate along the central coast. Between Cape Mendocino and Point Concep- tion there is only one major submarine canyon that heads in the near- shore area. This is the Monterey submarine canyon (Fig. 6), which is more than 2000 m deep and 22 km wide only 13 km from the shore at Cypress Point. It is as broad as, but deeper than, the Grand Canyon of Arizona in the area from Grand Canyon Lodge on the North Rim to Park Headquarters on the South Rim! By contrast, the continental shelf from Monterey to Cape Mendocino is fully 40 to 50 km wide out to the continental slope and it is not incised by major submarine can- yons (for detail, see U.S. Coast and Geodetic Survey, Bathymetric Maps, 1308 N-12; 1307 N-18; 1307 N-11B; 1306 N-20; scale 1:250,000, 1974). A similar broad shelf extends south to Point Conception except for the locally narrower strip where it fronts the fault scarp at the north end of the Santa Lucia Range between Nepenthe and Lopez Pt. at Lat. 36°N (see Jennings 1975; this map also shows bathymetry). The large, deep Monterey submarine canyon and its subsidiary branches to the north (Soquel Canyon) and south (Carmel Canyon) that reach close to shore are the sites of intense upwelling of deep, colder water that provides greater fog frequency in this area. Although temperature data are not available for comparison with the Ano Nue- vo and Cambria populations, maps of summer sea-surface tempera- —_ Fic. 6. The coastal strip of central California is bordered by a broad, shallow continental shelf except for the Monterey area where the Monterey submarine canyon and its tributaries (Soquel and Carmel Canyons) come in close to the shore and locally on the Big Sur coast north of Lopez Pt. 139 AXELROD: MONTEREY ENDEMIC AREA 1982] 140 MADRONO [Vol. 29 ture (Calcofi Atlas 1963) and a map showing the distribution of mean annual heavy fog (days/year) along the coast (Peace 1969, fig. 1), in- dicate a foggier summer climate at Monterey. Furthermore, in dis- cussion (Nov 1980) Dr. Larry Breaker (National Environmental Sat- ellite Service, Redwood City) stated that his several years’ study of satellite photos of the California coast shows that the present discon- tinuous groves of closed-cone pines are all at centers of high fog con- centration. The fog is then dispersed by winds along the coast ac- cording to local terrain. He also stated that the Monterey area has the highest fog frequency in this sector of the central California coast. The Monterey submarine canyon probably would have ensured a greater fog frequency for the pine groves during the Xerothermic as compared with areas to the north or south where upwelling is not as pronounced, and where the influence of drought would be greater during the Xe- rothermic. In this regard, the occurrence of fossil P. radiata (evidently the Monterey population in cone size) near the mouth of the Little Sur River is significant. The deposit, which is relatively young (about 10,000 years), is composed of species that represent a community much like that now at Monterey (Langenheim and Durham 1963), including Pinus muricata (?), Cupressus goveniana (?), Pseudotsuga menziesii, Quercus agrifolia, and a number of their usual shrubby associates, notably species of Ceanothus, Garrya, Myrica, Ribes, and Rubus. This forest probably disappeared from Little Sur, only 20 km south of Monterey, during the Xerothermic. This is consistent with the pres- ent occurrence of a few characteristically interior and southern taxa that are near their northern limit of distribution along the coast in this general area, notably Salvia mellifera and Eriodictyon tomentosum. They probably entered this coastal sector after the Little Sur forest disappeared and as coastal sage assumed dominance in this steep, well-drained terrain where fog is less frequent and persistent than at Monterey-Carmel. Pinus radiata at Monterey, with smaller cones than those at Ano Nuevo and Cambria, thus appears to be an ancestral, relict popula- tion. Its position at the head of Monterey submarine canyon during the Xerothermic probably afforded a more favorable, local climate than that in groves to the north and south, which responded to greater drought stress by developing larger cones. Local climate at Monterey probably accounts also for the persistence of the relict stands of Cu- pressus macrocarpa and C. goveniana which have even narrower dis- tributions than the pine (see Griffin and Critchfield 1972). Both cy- presses have fossil records north and south of Monterey, as does the Pinus radiata population associated with them (Fig. 4). Their restric- tion to the Monterey area during the later Pleistocene probably re- sulted from the spread of colder climate at the north and warmer, drier climate at the south, though not simultaneously. 1982] AXELROD: MONTEREY ENDEMIC AREA 141 UPWELLING AND RELICT DISTRIBUTIONS Local climate controlled by upwelling of colder, deeper water near shore may account for the discontinuous distribution and relict occur- rence of other endemics in coastal California and adjacent Baja Cal- ifornia. Pinus torreyana. This relict pine occurs in two patches on the coast- al strip on both sides of Soledad River Valley, at Del Mar to the north and Torrey Pines State Park to the south (see Griffin and Critchfield 1972). The groves are situated near the head of La Jolla submarine canyon and also lie farther west—into the ocean—than La Jolla or San Diego. According to local informants, this is the foggiest part of the coast. The relict occurrence of the pine groves may therefore reflect the highly favorable land climate in this area where precipitation totals about 250 mm annually. As noted above, their occurrence on Linda Vista Terrace and on Eocene Torrey Sandstone may also provide moisture that rises by capillarity from associated siltstones in these formations and from the underlying shales of the Del Mar Formation. That the pine is a relict is evident from its distinct morphology among pines of sect. Sabinineae and from its occurrence on Santa Rosa Island west of Santa Barbara. Pinus muricata-P. remorata. These pines occur on the coast of Baja California Norte near the mouth of Rio San Isidro, which is about 560 km (350 mi) south of their nearest mainland occurrence in the vicinity of Lompoc, California. Annual precipitation in the San Isidro area totals approximately 200-250 mm and vegetation is Diegan sage (Ax- elrod 1978). The occurrence of pines may be explained by upwelling that results in a greater fog frequency there during the warm months than elsewhere along the coast. As a result of the deep submarine terrain directly offshore (H.O., Chart BC 1206N), upwelling gives the warm months (June—September) sea-surface temperatures 3°—4° lower (i.e., 15-16°C vs. 19-20°C) in the area directly south of the San Isidro River valley where the pines occur than north of the valley where they are absent (Scripps Inst. 1962, p. 40-41). Other shrubs with relict distributions that require mild summer climate occur in this area as well as more widely to the north—but always near the coast—include Comarostaphylis diversifolia and Ribes viburnifolium. It is recalled that upwelling is also a determinant in the distribution of recent and fossil cold-water molluscan taxa on open exposed coasts, as reviewed by Valentine (1955; 1961, p. 525-527), Hubbs et al. (1962, p. 218; 1965, p. 112) and others to whom they refer. The pine groves are also near Lompoc (Tranquillon Hills, Purisima Hills) and San Luis Obispo (Pecho Hills) [see Griffin and Critchfield 1972]. These areas are characterized by foggy summer climate and also lie farther west than the adjacent coast—yjutting out into the 142 MADRONO [Vol. 29 ocean. The importance of local fog in the persistence of these relict groves was especially evident during the record hot spell of late August 1981. At that time the mid-day temperature differential 12 km from the coast was on the order of 14°C (13° as compared with 27°C). Heavy fog was also persistent near Pt. Sal midway between these areas, a site where the pines lived during the late Pleistocene (Axelrod 1980). Other taxa. Comarostaphylis diversifolia var. planifolia and Cer- cocarpus blancheae occur locally on the seaward slopes of the Santa Monica Mountains. In this area, with the Mugu, Dume, and Santa Monica submarine canyons directly off-shore, upwelling gives the lo- cale a high fog frequency in summer favorable for these taxa that are otherwise largely of insular occurrence. Coreopsis gigantea also occurs in this area close to the coast as well as on the islands, ranging north- ward to southern San Luis Obispo County in scattered, well drained, foggy coastal sites. MOoDE OF EVOLUTION Many of these endemics occur on sandy sites, including stabilized dunes, elevated terraces, and old floodplains. As noted above, these sites are mesic and may well account for the persistence of older taxa derived from alliances of more humid regions during earlier times. It is also evident that these occurrences follow the pattern of edaphic endemism discussed by Raven (1964). In it geographically marginal (especially annual) populations may invade unusual or different sub- strates during movement of a species into a region. Subsequently, they change sufficiently to be recognized as minor taxa (vars., subspp.) or full species depending on the mode of change. This may be by salta- tional speciation with catastrophic selection and consequent rapid chro- mosomal reorganization in marginal populations, by hybridization (see Arctostaphylos, in Appendix), by mutation, or by other means that have occurred in regions of diverse terrain and substrates subject to fluctuating climate during the Quaternary. These modes have been especially effective in annuals because gene flow in plant groups is very local, the effective population size is therefore very small, and the life cycle is brief. Finally, it must be reemphasized (see Raven 1976) that the nature of species is so variable that all these taxa— species, subspecies, varieties—are not of equal rank just because they have been so designated. They have had diverse modes of origin and are separated also by a time factor in which varied and unequal genetic changes have taken place in them. SUMMARY Most of the five small endemic areas in the outer, central Coast Ranges have diverse relief, climate, and substrates as well as a rela- tively high incidence of warm season rainfall, and two of them have 1982] AXELROD: MONTEREY ENDEMIC AREA 143 serpentine areas that support endemics. The Monterey endemic area, the richest one in the outer Coast Ranges, does not conform to this pattern. High endemism at Monterey owes in part to the edaphic factor, represented by the wide occurrence of endemics on dunes, terraces, and other sandy sites. These substrates apparently contain sufficient moisture to support endemics derived from forest and forest-border ancestors that were in the lowlands during moister Pleistocene phases. Other endemics are allied to taxa that evidently penetrated the Mon- terey area from the interior during the Xerothermic and some are members of species-groups distributed in coastal sites. Some evidently originated from marginal populations left behind on unusual sub- strates and then changed slightly by diverse modes. The relict occur- rence of two Cupressus species owes to their location at the head of Monterey submarine canyon, with the resultant higher fog frequency on-shore evidently providing a haven for survival during the Xero- thermic. The more persistent fog blanket may account also for the smaller-coned population of Pzmus radiata, which appears to be relic- tual and ancestral to the larger-coned populations at Ano Nuevo and Cambria that evidently reflect response to the Xerothermic climate. Evidence suggests that other relict endemics in the coastal strip, including Pinus muricata, P. remorata and P. torreyana, are confined to areas where favorable submarine topography near shore results in cold-water upwelling and higher fog frequency during summer months. The fossil record shows that a rich, floristically diverse closed-cone pine forest formed a continuous belt along the coast into the late Pleis- tocene. During colder, wetter pluvial stages mixed conifer forest taxa became associated with coastal pine forest and in the drier interglacial and postglacial times interior taxa contributed new species (subspp.., vars.) to open, sunnier sites. Thus, there is no need to invoke insular isolation to explain the evolution of its taxa, their present discontin- uous distribution, or the endemic nature of the now local floras. ACKNOWLEDGMENTS This research was supported by grants from the National Science Foundation, most recently Grant 79-05843. I am especially grateful to P. H. Raven for critical comments on the relations of endemic taxa in the Monterey area as drawn from his unparalleled, detailed knowledge of the flora. He and W. B. Critchfield have also provided valuable comments on the manuscript. Paul Leverenz, Scripps Institution of Oceanography, provided oceanographic data that have aided in interpretation of local coastal climates. Thanks are extended to the following systematists for information regarding the affin- ities of certain taxa listed in the Appendix: L. C. Anderson—Haplopappus; M. J. Gil- lett—Trifolium; L. R. Heckard—Cordylanthus; J. C. Hickman—Polygonum; R. C. Rollins—Erysimum; and P. V. Wells—Arctostaphylos. LITERATURE CITED AXELROD, D. I. 1966. The Pleistocene Soboba flora of southern California. Univ. Calif. Publ. Geol. Sci. 60:1—79. 144 MADRONO [Vol. 29 . 1977. Outline history of California vegetation. Jn M. G. Barbour and J. Major, eds., Terrestrial vegetation of California, p. 140-193. Wiley-Interscience, NY. . 1978. The origin of coastal sage vegetation, Alta and Baja California. Amer. J. Bot. 65:1117-1131. . 1980. History of the maritime closed-cone pines, Alta and Baja California. Univ. Calif. Publ. Geol. Sci. 120:1-143. . 1981. Holocene climate in relation to vegetation disjunction and speciation. Amer. Naturalist 117:847-870. BAGNOLD, R. A. 1941 (reprinted, 1956). The physics of blown sand and desert dunes. William Morrow and Co., NY. BAKER, H. G. 1972. Seed weight in relation to environmental conditions in California. Ecology 53:997-1110. CALCOFI ATLAS. 1963. California Cooperative Oceanic Fisheries Investig. Atlas 1 and 13. CASTEEL, R. W. and C. K. BEAVER. 1978. Inferred Holocene temperature changes in the North Coast Ranges of California. Northwest Sci. 52:337-342. CHRISTENSEN, M. N. 1966. Quaternary of the California Coast Ranges. Calif. Div. Mines and Geol. Bull. 190:305-314. DorF, E. 1930. Pliocene floras of California. Carnegie Inst. Wash. Publ. 412:1-112. FRANK, F. J. 1968. Availability of ground water in the Clatsop Plains, sand dune area, Clatsop County, Oregon. U.S. Geol. Surv. Open-file Report, 12 p. GRIFFIN, J. R. and W. B. CRITCHFIELD. 1972. The distribution of forest trees in California. USDA, For. Serv. Res. Paper PSW-82/1972. HELLEY, E. J., D. P. ADAM, and D. B. BURKE. 1972. Late Quaternary stratigraphy and paleoecological investigations in the San Francisco Bay area, In D. Adam et al., Progress Report on the U.S. Geological Survey Quaternary studies in San Francisco Bay area, p. 19-29. [Guidebook for Friends of the Pleistocene, Oct. 6- 8, 1972.] Howitt, B. F. and J. T. HOWELL. 1963. The vascular plants of Monterey County, California. Wassman J. Biol. 22:1—184. and . 1973. Supplement to the vascular plants of Monterey County, California. Pacific Grove Mus. Nat. Hist. Assoc. Pacific Grove, CA. Husss, C. L., G. S. Bren, and H. E. Suess. 1962. La Jolla natural radiocarbon measurements II. Radiocarbon 4:204—238. ; , and . 1965. La Jolla natural radiocarbon measurements IV. Radiocarbon 7:66—117. HYDROGRAPHIC OFFICE (DEFENSE MAPPING AGENCY). 1952. Botton contoured po- sition plotting sheet. Chart BC 1206N. 1:1,000,000. JENNINGS, C. W. 1975. Fault map of California. Geologic data Map No. 1, California Div. Mines and Geology (scale 1:750,000). Jepson, W. L. 1925. A manual of the flowering plants of California. Univ. Calif. Press, Berkeley. LANGENHEIM, J. H. and J. W. DURHAM. 1963. Quaternary closed-cone pine flora from travertine near Little Sur, California. Madrono 17:33-51. LENz, L. W. 1975. A biosystematic study of Tviteleia (Liliaceae). I. Revision of the species of section Calliphora. Aliso 8:221-258. LINSDALE, J. M. 1955. Check list of ferns and seed plants of Frances Simes Hastings Natural History Reservation. Leafl. W. Bot. 7:201—218. Munz, P. A. 1973. A California flora and Supplement. Univ. Calif. Press, Berkeley. MurpuHy, T. M. 1981. Immunochemical comparisons of seed proteins from populations of Pinus radiata (Pinaceae). Amer. J. Bot. 68:254—-259. Noss, M. A. 1963. Experimental studies on species relationships in Ceanothus. Car- negie Inst. Wash. Publ. 623. Norris, R. M. and K. S. Norris. 1961. Algodones Dunes of southeastern California. Geol. Soc. Amer. Bull. 72:605—620. 1982] AXELROD: MONTEREY ENDEMIC AREA 145 PAVLIK, B. M. 1979. A synthetic approach to the plant ecology of desert sand dunes, Eureka Valley, California. M.S. thesis, Univ. California, Davis. . 1980. Patterns of water potential and photosynthesis of desert sand dune plants, Eureka Valley, California. Oecologia 46:147-154. PEACE, R. L., JR. 1969. Heavy fog regions of the conterminous United States. Monthly Weather Review 97:116—123. PoTBURY, S. S. 1932. A Pleistocene flora from San Bruno, San Mateo County, Cali- fornia. Carnegie Inst. Wash. Publ. 415:25—-44. RAVEN, P. H. 1964. Catastrophic selection and edaphic endemism. Evolution 18:336— Rion . 1976. Systematics and plant population biology. Systematic Bot. 1:284-316. and D. I. AXELROD. 1978. Origin and relationships of the California flora. Univ. Calif. Publ. Bot. 72:1-134. SCRIPPS INSTITUTION OF OCEANOGRAPHY. 1962. Surface water temperatures at shore stations, United States West Coast and Baja California. SIO Reference 62-11. SHARP, R. P. 1966. Kelso Dunes, Mohave Desert, California. Geol. Soc. Amer. Bull. 77:1045—-1074. SHARSMITH, H. K. 1945. Flora of the Mount Hamilton Range of California. Amer. Midl. Naturalist 34:289-367. STEBBINS, G. L. and J. Major. 1965. Endemism and speciation in the California flora. Ecol. Monogr. 35:1-35. THomas, J. H. 1961. Flora of the Santa Cruz Mountains of California. Stanford Univ. Press, Stanford, CA. VALENTINE, J. 1955. Upwelling and thermally anomalous Pacific Coast Pleistocene molluscan faunas. Amer. J. Sci. 253:462-474. . 1961. Paleoecologic molluscan geography of the California Pleistocene. Univ. Calif. Publ. Geol. Sci. 34:309-442. (Received 12 May 1981; revision accepted 8 Dec 1981.) APPENDIX. ENDEMIC TAXA OF THE MONTEREY ENDEMIC AREA, THEIR GENERAL OCCURRENCE AND THEIR PROBABLE RELATIONSHIPS. (species authorities follow Munz 1973.) Pinaceae: Pinus radiata. The Monterey population has fossil records that suggest it is older than the larger-coned populations at Ano Nuevo and Cambria, which may have originated during the Xerothermic in response to warmer, drier climate (see text). Cupressaceae: Cupressus goveniana. Fossil record shows that it had a wider distribution to the north and south in the later Quaternary, and probably earlier as well. Cupressus macrocarpa. Recorded as fossil at Drakes Bay (1 m.y. old) and Rancho La Brea (30,000 years ago). No very close allies. Restricted to quartz diorite on the north and south shores of Carmel Bay. Amaryllidaceae: Triteleia X versicolor. (=Brodiaea versicolor). Known only from type collection made at Pt. Lobos State Park. Lenz (1975) considers it a natural hybrid. Asteraceae: Corethrogyne filaginifolia var. viscidula. A local endemic of the species, which is represented by 13 vars. distributed from central California into southern California, with most of them in the interior and to the south. Haplopappus eastwoodiae. Restricted to dunes about Monterey and Carmel Bays. L. C. Anderson reports (pers. comm., Feb 1981) that it is probably derived from H. ericoides, which is distributed along the coast from Marin Co. southward to Los Angeles area, chiefly on sand dunes near the coast. Malacothrix saxatilis var. arachnoidea. From lower Carmel Valley, in sandy 146 MADRONO [Vol. 29 sites. It is one of several varieties of the species that ranges from Monterey Co. southward into southern California. Brassicaceae: Erysimum ammophilum. Allied to E. menziesii, E. concinnum, and E. franciscanum, a closely related group of taxa (R. C. Rollins, letter of Feb 1981). He suggests they may represent a linearly distributed group of species adapted to coastal sands that broke up, over time, to form the present taxa. Ericaceae: Arctostaphylos edmundsii. According to P. A. Wells (letter, Feb 1981), A. edmundsii is a stabilized hybrid of A. uva-ursi X A. tomentosa rosei. A. tomen- tosa is a widespread, variable species on the central California coast. A. uva- ursi is basically a forest species from Marin Co. north, but has relict stations southward into Monterey Co. to Pt. Sur. Arctostaphylos hooker. P. A. Wells (letter, Feb 1981) states A. hookeri is derived from the A. pungens series. A. pungens is distributed from the inner south Coast Ranges into the southwestern United States and Mexico. A. hookeri occurs in the closed-cone pine forest and on dunes about Monterey. Arctostaphylos montereyensis. A unique non-sprouter, possibly derived from or re- lated to A. columbiana-A. virgata according to P. A. Wells. A. columbiana ranges from Marin Co. northward; A. virgata occurs marginal to redwood forest in Marin Co. A. montereyensis also occurs inland from Monterey for a short dis- tance. On dunes in Monterey area. Arctostaphylos pumila. Wells (letter, Feb 1981) states this is a stabilized hybrid of A. uva-ursi X A. tomentosa tomentosa. A. uva-ursi is a forest species chiefly from Marin Co. north, but has local relict occurrences south to Pt. Sur. A. tomentosa is a widely distributed, very variable species of the central California coast. The var. tomentosa is endemic to Monterey. Arctostaphylos pajaroensis. This is a stabilized hybrid of A. andersonii X A. tomentosa tomentosa according to Wells, who also states A. andersonii is en- demic to the Santa Cruz Mountains. Arctostaphylos tomentosa. The forms trichoclada and hebeclada of A. tomentosa var. tomentosa are morphs that occur with var. tomentosa only about Monterey, according to Wells (letter, Feb 1981). Fabaceae: Lupinus tidestromii. A dune to strand plant allied to and probably derived from L. littoralis of coastal strand and coastal sage from Mendocino Co. north- ward to British Columbia. Trifolium polydon. Considered a variety of T. tridentatum by Jepson (1925), a view generally concurred in by Gillett (letter, Feb 1981) though he suggests it may only be a variety of 7. variegatum. T. tridentatum is a widespread species in grassy places in cismontane California generally below 1520 m, ranging north to British Columbia. T. variegatum is found in moist grassy places in cismontane California, generally below 2440 m. Trifolium trichocalyx. Gillett (letter, Feb 1981) suggested that it may be a spo- radic hybrid between 7. microcephalum and T. variegatum, widespread species in cismontane California below 2590 m. Since only one or two plants are known, it seems best to remove it from the list of Monterey endemics. Liliaceae: Allium hickmanii. From around Monterey peninsula, but also near Jolon, in the interior. Related to a species complex of open places in forests of Coast Ranges; possibly more closely related to A. unifolium from Monterey Co. north to Del Norte Co. Polemoniaceae: Eriastrum virgatum. Sandy sites about Monterey. Apparently allied to species now in drier places, chiefly to the south and interior. Gilia tenuiflora subsp. arenaria. A segregate of the species that inhabits open places, river beds and sandy sites; distributed chiefly in the interior and south- ward to San Luis Obispo Co. Polygonaceae: Chorizanthe pungens. On dunes around Monterey; the poorly defined var. hartwegii occurs in sandy places from Santa Cruz Mts. north to San Fran- 1982] AXELROD: MONTEREY ENDEMIC AREA 147 cisco. Allied to C. cuspidata of dunes and sandy places from Santa Cruz Co. north to Sonoma Co. Eriogonum parvifolium subsp. lucidum. Pt. Lobos. A poorly defined segregate of this shrubby species that is common along coastal bluffs and dunes from Monterey south to San Diego Co. Polygonum montereyense. Known only from the type material, collected ‘on hard dry clay, Monterey.” Dr. J. C. Hickman reexamined the type material and reports (Mar 1981) it is a diseased or parasitized specimen of P. ramosissimum, a widespread species. It is therefore removed as a Monterey endemic. Ranunculacae: Delphinum hutchinsoniae. Sandy places about Monterey. Allied to D. variegatum, widespread in the inland valleys. Rhamnaceae: Ceanothus griseus var. horizontalis. Occurs at Yankee Pt., Monterey. A spreading, prostrate form of the species that ranges along the coast from Men- docino Co. to Santa Barbara Co. Ceanothus rigidus. Inhabits sandy hills and flats, Monterey peninsula. Allied to C. gloriosus distributed in closed-cone pine forest from Marin Co. north to Men- docino Co. (see Nobs 1963, fig. 39). Saxifragaceae: Ribes menziesii var. hystrix. From Monterey peninsula and border areas into the redwood forest on coastal slopes of Santa Lucia Mountains. Allied to subspecies leptosmum and senile in redwood forests northward into Sonoma Co. Scrophulariaceae: Cordylanthus littoralis. Occupies dunes back of the strand and in closed-cone pine forest at Monterey. Dr. L. Heckard, who has monographed the genus, considers C. littoralis a subspecies of C. rigidus, with which it intergrades in Upper Carmel Valley (pers. comm., Feb 1981). The species rigzdus is found chiefly in the inner south Coast Ranges. Mimulus guttatus var. arenicola. Coastal strand and dunes, Monterey area. The species is widespread in coastal plant communities in California, represented by several varieties. The var. arenicola is allied to the subsp. littoralis, which ranges along the coast from Santa Barbara Co. north to Washington. MOUNTAIN MEADOWS: STABILITY AND CHANGE NATHAN B. BENEDICT Department of Biology, University of Nevada, Reno 89557 ABSTRACT It is often assumed that mountain meadow ecosystems are fragile and temporary phenomena. An alternative hypothesis is that meadows are as stable over time as the surrounding vegetation. Evidence from palynological, stratigraphic, and successional studies on Sierran meadows supports the latter hypothesis. Three new working hy- potheses on the causes of change in meadow ecosystems are proposed. Discussions of mountain meadows often assume that meadows are “fragile,” and that they are temporary phenomena. These two as- sumptions are directly related to our understanding of meadows as seral stages in the classic hydrosere of lakes developing into forests, and imply that meadows are unstable. This paper examines these assumptions by reviewing the evidence for two hypotheses: 1) mead- ows are as stable over time as the surrounding vegetation; and 2) meadows are temporary phenomena. This will be done using data and observations from the literature including my own research over the past several years. Most of the examples come from the Sierra Nevada, California. Before reviewing the evidence, it is necessary to discuss briefly the concepts of stability and succession in relation to mountain meadows. TYPES OF MEADOW STABILITY There are many ways of classifying the different types of meadow stability. For the purpose of this discussion, stability can be divided into biologic and geologic stability (Benedict and Major 1981). Biologic stability refers to the persistence and recovery of the biologic compo- nent of an ecosystem. For example, if the species composition of a meadow does not change over time it is considered stable. Geologic stability refers to the persistence of the geologic conditions which create an environment favorable for meadow formation. In the Sierra Nevada, geologic stability is directly related to the different physio- graphic meadow types (Benedict and Major 1981). For example, a meadow that forms in a bedrock basin as a result of water accumu- lation is considered stable as long as the basin is intact and continues to collect water. This meadow type is geologically more stable than a meadow in a basin formed by a morainal dam, which is more readily eroded than a bedrock dam (Benedict and Major 1981). When the MADRONO, Vol. 29, No. 3, pp. 148-153, 9 July 1982 1982] BENEDICT: MOUNTAIN MEADOWS 149 TABLE 1. BIOLOGIC AND GEOLOGIC CONDITION IN RELATION TO VEGETATION CHANGE. Autogenic/allogenic changes, sensu Tansley 1929. Allogenic changes brought about by geologic instability may have a relatively greater effect on the vegetation than autogenic changes. Geologic condition Biologic condition Vegetation change Stable stable no change (climax stage) Stable unstable autogenic succession Unstable stable nonsensical Unstable unstable allogenic succession (autogenic succession) concepts of biologic and geologic stability are combined, they provide us with a structure to examine the successional status of meadows, and to decide if meadows are indeed temporary phenomena (Table 1). Because ecosystems are continuously changing, the division between stable and unstable conditions is partly artificial. Human recognition of change is dependent on the rate of change and the effect change has on an ecosystem. For example, in a basin meadow formed by a glacial moraine, the morainal dam is probably continuously eroding. The erosion rate may vary with season, strata in the moraine, or general longterm climatic changes. If the ecosystem can compensate for the changes in erosion rate and the height of the dam, the changes will go unnoticed and the system will be considered stable. The effect change has on an ecosystem is dependent on threshold values for both the rate and magnitude of change. For example, the decrease in height of a morainal dam may not affect the meadow until some critical height is reached. Once this height is passed, the ecosystem may undergo drastic change and be considered unstable. The recognition of change and the role of threshold values in relation to vegetation change is a complex problem and deserves further consideration. In this paper change will be recognized using plants as biometers. We can now state our hypotheses in more readily testable form: 1) meadows are both biologically and geologically as stable as the sur- rounding vegetation; and 2) meadows are temporary phenomena be- cause they are a) geologically stable but biologically unstable, and/or b) both biologically and geologically unstable. It should be noted that these hypotheses are not necessarily exclusive of each other. EVIDENCE FOR BIOLOGIC AND GEOLOGIC STABILITY There are two sources of evidence supporting the hypothesis that meadows are both biologically and geologically as stable as the sur- rounding vegetation. The first is palynological evidence. Adam (1967) presents pollen diagrams at four locations in the Sierra Nevada. The 150 MADRONO [Vol. 29 pollen record at Osgood Swamp (near South Lake Tahoe), and Soda Springs (near Tuolumne Meadows, Yosemite National Park) are per- tinent to this discussion. At the time the pollen core was taken, Osgood Swamp consisted of a small, seasonal lake surrounded by a wet mead- ow and bog complex. The Osgood Swamp pollen record begins about 10,000 BP, the approximate beginning of the Holocene. The temporal relationships are substantiated by two radiocarbon dates (Adam 1967). Soda Springs is a spring mound near Tuolumne Meadows. No radio- carbon dates are given but the record is estimated at between 7000 and 9000 BP by comparison with the Osgood Swamp pollen record (Adam 1967). From these two pollen records the following inferences can be made: 1) The geologic conditions that favored a hydric meadow environment and pollen preservation have existed at Osgood Swamp since approx- imately 10,000 BP, and at Soda Springs since approximately 7000— 9000 BP. 2) Vegetation indicative of mesic and hydric meadow envi- ronments has been present for the entire period represented by these pollen records (see Cyperaceae, Alnus, Isoetes, etc.). 3) Species indic- ative of non-meadow vegetation have varied greatly in importance over the time period represented by these pollen cores (see Abies, Acer, Artemisia, Quercus, and TCT—Taxodiaceae, Cupressaceae, Taxa- ceae. 4) Although the hydric and mesic species composition may have changed during the time represented by these pollen cores, the sur- rounding non-meadow vegetation also changed during this same time period. The second source of evidence comes from stratigraphic studies of montane meadows. Wood (1975) examined the stratigraphy of seven montane meadows in the southern Sierra Nevada. From his findings several conclusions can be drawn: 1) Five of the seven meadows have been in existence since 1200-3000 BP, and two of the meadows since 7700-9800 BP. Thus, the geologic and climatic conditions favorable for meadow formation have been present over those time periods. 2) Although no direct study of biologic change was made, the meadows have not changed to forest vegetation since 1200-3000 BP in five lo- cations, and since 7700—9800 BP in two locations. 3) No lake sediments were found in any of the seven meadows. 4) The forest vegetation on these same sites has changed to meadow vegetation over the last 10,000 radiocarbon years (Wood 1975). The above examples imply that these eight meadows and one bog have been biologically and geologically stable for various lengths of time ranging from 1200 BP to approximately 10,000 Bp. They also imply that vegetation change has occurred in both meadow and the surrounding non-meadow ecosystems over the past 10,000 radiocarbon years. In addition, the stratigraphic record suggests that meadows develop into forests only when the geologic environment passes some threshold point. 1982] BENEDICT: MOUNTAIN MEADOWS 151 EVIDENCE FOR GEOLOGIC STABILITY AND BIOLOGIC INSTABILITY Two examples of secondary, autogenic succession (Tansley 1929) in meadows support this hypothesis. An unusual lightning-ignited fire in Ellis Meadow, Kings Canyon National Park, in 1977 severely burned parts of the meadow, leaving ash 1—20 cm deep and removing 100 percent of the plant cover in some areas (DeBenedetti and Parsons 1979a, DeBenedetti 1980). One year after the fire, plant cover was 36 percent for herbaceous plants and 31 percent for moss and liverworts, with no evidence of catastrophic change in the character of the mead- ow (DeBenedetti 1980). Yearly sampling has suggested that typical meadow vegetation will eventually be reestablished (DeBenedetti, pers. comm.). This implies that the geologic conditions are stable even though the biologic component of the Ellis Meadow ecosystem is pres- ently unstable and will take an unknown number of years to stabilize. The second example is the re-establishment of meadow vegetation after removal of grazing in the Sierra Nevada. The destruction of the biologic component of meadow ecosystems in the Sierra by grazing has been well documented (DeBenedetti and Parsons 1979b). In many of these meadows, typical meadow vegetation has reestablished itself (DeBenedetti and Parsons 1979b) although the lack of baseline data makes it difficult to judge the degree of similarity between current and pristine vegetation. The degree and rate of recovery from past over- grazing may vary depending on meadow type, with xeric meadows being slowest to recover fully (Ratliff 1974). This again implies that the geologic conditions favoring meadow formation have been stable even though the biologic component has been (and may still be) re- covering from past disturbances and is thus unstable. In a few meadow ecosystems the disturbance from grazing may have been severe enough to destabilize the geologic conditions and make meadow establishment highly unlikely without manipulation of the ecosystem by man (DeBenedetti, pers. comm.). EVIDENCE FOR BIOLOGIC AND GEOLOGIC INSTABILITY In the montane meadow stratigraphy of the Sierra Nevada (Wood 1975) discussed earlier, if the entire record at each meadow is exam- ined, it is clear that over the last 10,000 radiocarbon years these mead- ows have undergone dramatic changes. As summarized by Wood (1975), the generalized stratigraphic sequence is: 1) basal layer of al- luvium deposited by pre-Holocene streams; 2) paleosol extending into basal layer dated at between 8705 and 10,185 BP, and developed under. a mesic montane forest; 3) stratified sandy deposits dated at between 8700 BP and 1200-2500 BP, and developed under a fir, yellow pine, and lodgepole pine forest; and 4) stratified deposits of sedge peat, humus-rich sandy loams, and sorted grus deposited since 2300-2500 152 MADRONO [Vol. 29 BP in a meadow environment. (As previously noted, two meadows described by Wood do not fit this generalized sequence because they have been in existence since 7700-9800 BP.) This suggests that the geologic conditions causing meadow formation and the biologic com- ponent of meadow ecosystems have been unstable over the past 10,000 radiocarbon years. It should be noted that, in a similar manner, the geologic conditions favoring forest vegetation, and the biologic com- ponent of forest ecosystems at the same locations, have also been unstable over this same time period (Wood 1975). A second source of evidence is a man-induced “experiment” at Os- good Swamp. Osgood Swamp is dammed by a Wisconsin glacial mo- raine on the downstream side (Adam 1967). This created a wet basin favorable for formation of a meadow that had persisted for approxi- mately 10,000 years. In 1963, this morainal dam was artificially breached (Adam 1967). Subsequently the water table lowered and a large number of Pinus contorta subsp. murrayana seedlings became established. This “artificial geologic instability” demonstrates the ef- fects of passing a threshold of geologic change on the biologic stability of a meadow. It suggests also that naturally occurring geologic insta- bility as a result of erosion can reach threshold values that will result in instability in the biologic component of meadows, and the subse- quent establishment of forest trees. A third source of support is the evidence that throughout the western United States various forest trees have invaded mountain meadows since approximately 1900 (Dunwiddie 1977, Franklin et al. 1971, DeBenedetti and Parsons 1979b, Vale 1981a,b). This suggests that the conditions favorable for meadow formation and maintenance have changed and are thus unstable, and that this results in biologic insta- bility as indicated by tree establishment. An alternative explanation is that biologic instability has resulted in tree establishment even though the geologic conditions are favorable for meadow ecosystems. The latter would support hypothesis 2a. The above evidence suggests that geologic and biologic instability is present in meadow ecosystems, and that forest vegetation in areas that have been occupied by meadows is also unstable. CONCLUSION Based on the above evidence, meadow ecosystems are as stable as the surrounding vegetation, and, like any vegetation type, can be viewed as temporary phenomena on a geologic time scale. Both forest and meadow ecosystems have experienced dramatic changes since the beginning of the Holocene. Some ecosystems may have changed less than others, i.e., high elevation Pinus balforiana forests, or subalpine and alpine meadows in the southern Sierra Nevada. In addition, meadow ecosystems may be more sensitive than forest ecosystems to 1982] BENEDICT: MOUNTAIN MEADOWS 153 geologic change as a result of lower thresholds of tolerance to geologic change. The evidence suggests three new working hypotheses that should continue to be examined and debated: 1) Geologic meadow instability usually results in biologic instability that may lead to the establishment of forest vegetation. 2) Biologic instability combined with geologic stability can result in changes in species composition that may lead to a meadow climax, and that do not necessarily lead to a forest climax. 3) Biologic instability and autogenic changes are not great enough to overcome ultimately the stabilizing influence of the geologic conditions that maintain meadow ecosystems. LITERATURE CITED ADAM, D. P. 1967. Late-Pleistocene and recent palynology in the Central Sierra Ne- vada, Calif. In E. J. Cushing and H. E. Wright, Jr., eds., Quaternary paleoecol- ogy, p. 275-301. Yale Univ. Press, New Haven. BENEDICT, N. B. and J. Major. 1981. A physiographic classification of subalpine meadows of the Sierra Nevada, California. Madrono 29:1—12. DEBENEDETTI, S. H. 1980. Establishment of vegetation following fire in a subalpine meadow of the southern Sierra Nevada. One year post-burn. /n Proceedings of the Conference on Scientific Research in the National Parks (2nd). Volume 10: Fire ecology, p. 325-336. NTIS, U.S. Dept. Commerce. and D. J. PARSONS. 1979a. Natural fire in subalpine meadows: a case description from the Sierra Nevada. J. Forest. 77:477—479. and . 1979b. Mountain meadow management and research in Sequoia and Kings Canyon National Parks: A review and update. Jn R. Linn, ed., Proceedings First Conference on Scientific Research in the National Parks. USDI Natl. Park Serv. Trans. Proc. 5. 2:1305-1311. DUNWIDDIE, P. W. 1977. Recent tree invasion of subalpine meadows in the Wind River Mountains, WO. Arctic Alpine Res. 9:393-399. FRANKLIN, J. F., W. H. Moir, G. W. DouGLAs, and C. WIBERG. 1971. Invasion of subalpine meadows by trees in the Cascade Range, Washington and Oregon. Arctic Alpine Res. 3:215-224. RaTLIFF, R. D. 1974. Short-hair sedge . . . its condition in the high Sierra Nevada of California. USDA Forest Serv. Pacific Southw. For. Range Exp. Sta. Res. Note PSW-293. Berkeley, CA. TANSLEY, A. G. 1929. Succession, the concept and its values. Proc. Int. Cong. Plant Sci., Ithaca 1926:677-686. VALE, T. R. 198la. Tree invasion of montane meadows in Oregon. Amer. Mid. Naturalist 105:61—-69. . 1981b. Ages of invasive trees in Dana Meadows, Yosemite National Park, California. Madrono 28:45—47. Woop, S. W. 1975. Holocene stratigraphy and chronology of mountain meadows, Sierra Nevada, CA. Ph.D. dissertation, Calif. Inst. Technology, Pasadena. (Received 17 Jul 1981; revision accepted 7 Dec 1981.) PHENOLOGY, GERMINATION, AND SURVIVAL OF DESERT EPHEMERALS IN DEEP CANYON, RIVERSIDE COUNTY, CALIFORNIA Jack H. BuRK Department of Biological Sciences, California State University, Fullerton 92634 ABSTRACT Phenology, germination, and survival in a desert ephemeral community was followed from January 1976 through January 1978. Cool season species germinated following rain in fall and/or winter. Germination of warm season Cy, ephemerals was restricted to May. Significant variations occurred in the phenology of plants germinating in re- sponse to the same rain storm. The shape of survivorship curves among cool season species dramatically varied depending on seasonality of precipitation. The relationships between the germination of desert ephemerals and temperature and rainfall were established in a general sense by the work of Went (1948, 1949), Trevis (1958a,b), and Juhren et al. (1956). Germination requirements usually restrict ephemeral activity to a spe- cific season coincident with adequate rainfall (Beatley 1974). Summer annuals usually have the C, photosynthetic pathway, assumed to be an advantage in hot, dry conditions, whereas winter annuals are typ- ically C; and seem to have few specializations other than germination requirements that adapt them to desert environments (Mulroy and Rundel 1977). However, winter ephemerals have been shown to have differential ability to survive increasing water and temperature stress (Clark and Burk 1980). This indicates that a range of physiological adaptations to stress may be present in these “mesophytic” ephemerals. This study was designed to identify species with unique germination, phenological, and survival characteristics that could be used in future research on xerophytic adaptations in desert ephemerals. METHODS The study was conducted at the Philip L. Boyd Deep Canyon Desert Research Center located on the north slope of the Santa Rosa Moun- tains about 300 m above the desert floor. The 17-year average rainfall is 109 mm, maximum temperature 47°C and minimum temperature O°C. Rainfall is highly erratic with minima of about 40 mm in 1961 and 1971 and a maximum of 300 mm during 1976 (the first year for which data are presented here). Rainfall is typically concentrated in the winter months of November through March and in the late sum- mer months of August and September. Each month has been without MADRONO, Vol. 29, No. 3, pp. 154-163, 9 July 1982 1982] BURK: DESERT EPHEMERALS 155 measurable rain at least twice during the 17-year weather records for Boyd Center (Zabriske 1979). Precipitation values used in this study were recorded at Boyd Center less than 0.5 km from the study site. The study area was located at the mouth of Coyote Canyon in a heterogeneous area consisting of sandy arroyo beds and habitats with an accumulation of boulders and/or desert pavement at the soil sur- face. Perennial vegetation consists of a wash woodland community along the arroyo margins and creosote bush scrub in the intervening areas (Burk 1977). The 1.0 ha study area was gridded into 100, 100 m? plots from which 20 sample plots were selected at random. A 1.0 m? quadrat was systematically placed in the center of each of the 20 sample sites. Each 1.0 m? plot was further divided into 100, 1.0 dm? sub-samples from which 20 were selected for counting. Permanent stakes marked each sample site and allowed precise repositioning of the quadrat at each sampling. Dividers in the 1.0 m* quadrats were of 3 mm diameter stainless steel rod. Twelve sample sites were monitored from January 1976 until January 1977 and 8 additional sites were monitored through January 1978. Visits were made at 2-4 week intervals following rain storms in February 1976, September 1976, and January 1977 until all the ephemerals died. Repeat visits were not made in August 1977 and January 1978, but data were recorded after germination was presumed complete. Individuals were counted in the 1 dm?’ sub-plots for density deter- mination, and estimates of cover for each species estimated to occupy more than 1 percent of the sub-plot were made by visual inspection. At each sample site each species was recorded as being in one or more of the following conditions: vegetative, flower buds, flowers, or seeds. Phenology of ephemeral plants not in sample frames was also record- ed. Density and cover data were used to estimate population values on m? basis by combining the results of all sample sites. Seedlings of unknown species were marked with color coded plastic toothpicks until identification was possible (nomenclature follows Munz 1974). There were a limited number of instances when seedlings did not survive and could not be identified. These are not included in any of the calculations. Because of difficulties in distinguishing the various species of Cryptantha prior to flowering, individuals of C. maritima, C. angustifolia and C. barbigera are combined for purposes of analysis. This grouping may also contain an occasional individual of the genus Pectocarya. Even though no Pectocarya individuals reached anthesis in the counted plots, isolated individuals were ob- served in the vicinity. RESULTS Seedling establishment. Three occurrences of record monthly rain- fall along with two other periods of more normal levels produced five 156 MADRONO [Vol. 29 —" 134 Precipitation (cm) 4 3 1978 2 1 JF M AM J JS A S O N_ D Fic. 1. Total weekly precipitation records at Boyd Deep Canyon Desert Research Center for the period Jan 1976—Jan 1978. * indicates rainfall stimulating germination. distinct periods of germination. Figure 1 shows the precipitation pat- tern for the period of the study. September has the highest average monthly rainfall during the year, registering a record 16.0 cm in 1976. The 17-year record for August was established in 1977 with 9.5 cm of rain; only September and December exceed August precipitation on the average. January 1978 had 11.0 cm, exceeding the long term av- erage of 10.2, and setting a record for that month. Germination also occurred in response to precipitation in January 1977 and February 1976. The only other rainfall that stimulated germination occurred in early May 1976. Table 1 shows periods of germination for each species in relationship to season. Winter refers to germination responses to December, January, and February rain; late summer, to the August 1977 rainfall; and fall to establishment following September and Oc- tober rainfall. The decision to distinguish August and September as 1982] BURK: DESERT EPHEMERALS 157 TABLE 1. PERIODICITY OF GERMINATION OF DESERT EPHEMERALS JANUARY 1976-JANUARY 1978 AT BOYD DEEP CANYON DESERT RESEARCH CENTER. Dec—Feb May Aug Sep—Oct Filago californica Langloisia schottit Chorizanthe rigida Eschscholzia minutiflora Rafinesquia neomexicana Amaranthus fimbriatus * Palafoxia arida ‘ Chorizanthe brevicornu Phacelia distans Phacelia crenulata Perityle emoryt Chaenactis carphoclinia Chaenactis fremontii Camissonia claviformis Schismus barbatus Plantago insularis Cryptantha spp. Euphorbia spp. * *¥ * * *¥ *¥ + *& *¥ *¥ ¥ ¥ KK KF KF ¥ *¥ * *¥ *¥ *¥ ¥ *¥ FF ¥ KF KF ¥ different seasons was based on the (1) very different group of ephem- erals that became established and (2) the average temperature in Au- gust, which is much more like that in July than in September. Twenty species were observed in the sample pilots with six species germinating only in response to winter precipitation. Filago califor- nica, Langloisia schottii and Eschscholzia minutiflora germinated in response to both the moderate winter rainfall of 1976 and 1977, and the very wet January of 1978. Chorizanthe rigida and Rafinesquia neomexicana were observed in the plots only in January 1978. Precipitation of 3.6 cm in May 1976 resulted in the establishment of the C, summer ephemeral Amaranthus fimbriatus and the C, an- nual/perennial Euphorbia spp. This was the only occurrence of A. fimbriatus in the plots. Storms during late spring of 1977 were rare and provided less than 4 mm of precipitation. No summer annuals were present during the summer of 1977. Euphorbia spp. included E. setiloba, an annual, and E. polycarpa, a perennial. Because of the difficulty in distinguishing the species in the pre-flowering stage, I am unsure which species germinated following the May 1976 storm since the seedlings did not survive. Palafoxia arida is a facultative pe- rennial flowering in the first year and is often active during summer and winter. This species was common in the study area but germi- nation was restricted to periods following late summer and fall rains. Eight other species germinated in the fall and winter months (1977— 1978). None of these germinated following the heavy rains of August 1977. One, Chaenactis carphoclinia, showed additional germination 158 MADRONO [Vol. 29 Schismus 2 2. Le Le Le a a a hd de de we, ae barbat “woud Cererrnreorerei«eoeftrrtrerireoroeseeri arbatus \GUDCDCDCUROURDDUODOCCRORCCURUORCCRURCDEGURDORODODOSOQOQRQUREORCRDOQGODUGBODORDQCQUOROROBOUBI (CE EY rie ets SR Sn we claviformis IOGUGGEGORRGERGOOROGERCSEESSTCSSUIGUCTODIGS lionel] Cryptantha Ze, eo 2e Le a ££ wt wd dz bdw. LLL L. LB spp “TrasestrTrTtrrrerpireséspirtitttrttrrouititst. . HODD COCO COCR DODO CR DOCOOROGCGCUODOROGDRDRCROQOUCURRGRURCRQOCOUUSORGGCGCRBORCRCRB ore] Plantago "LEILA PPL insularis TOTO OO Le Filago OLMCMI SLES LS LS ; ; wT 7 TT tr tT rnreseieuece californica OTT ; a Chaenact ag ntti a “4 4-4-4411 fremontii TOCUUOTOTTHGTTONTOEOOSONSOGOCONRCRORONGERS F leer cna ae ee el Chaenactis ; LILLE LL LLL La carphoclinia UePivtevccrercucctersceetentstatentseatstevaravertint Phacelia NN aaa LL distans TOLLE EET [aoeece ws ae as tea Phacelia LALLA crenulata ee Ti Trbuccecccccccee Ee Langloisia tet a schottii DOGUDUCUDOSSCRCRRGRRCRROREDR Ey Chorizanthe a MPAA brevicornu TO Euphorbia —— spp. Amaranthus A ae 2 ee Ge oe oe oe ee oe ee © 0 0 fimbriatus Oneupeneeen! re ee eee ee ee eee eee ee eee ee ee ee 40 60 80 100 120 140 M A M J Growing Season (days) Fic. 2. Phenological events of ephemeral plants at Boyd Deep Canyon Desert Re- search Center, spring 1976. Solid lines represent pre-reproductive periods; diagonally hatched lines, flower buds; long bars, flowering; and short bars, the period of seed production. Growing season is measured as the number of days following the rainfall that stimulated germination. in reponse to 41 mm of precipitation during April 1976. Pevityle emo- ryt and Chorizanthe brevicornu did not appear in the study plots following all fall and winter rains. The two genera that responded to rainfall during the greatest num- ber of seasons were Cryptantha and Euphorbia. Euphorbia seedlings were found at all seasons during both years of the study and Cryp- tantha in all except May 1976. Phenology. Interspecific variation in the timing of life cycle events is apparent in plants that were present during the winter and spring of 1976 (Fig. 2). Eleven species germinated in response to February rains and at the first sampling time (31 days) Schismus barbatus and Filago californica were already in the reproductive phase of the life cycle; Schismus was in flower and Filago in bud. The majority of species (82 percent) were in reproductive phase within 60 days. Chaen- actis carphoclinia formed buds within 60 days but anthesis did not occur until at least 20 days later; it continued to flower well into 1982] BURK: DESERT EPHEMERALS 159 i ec ae) cenemds he, eh, ohh th nh hn fh hnhthon dhe uhheudke barbatus [oe eg ee a dd dD TITITIU CamiSsonia exes EER claviformis Cryptantha —_—_-—: spp. ee eee ee ee ee ee es ss at TITITIIIIT ILL ITD rr Plantago EE EET] insularis remem ecm TTTIIITII II Filago a a californica ChaenactiS smmmmmmmmmmmee fremontii ~—_ Prt. ChaenactiS —___—_ carphoclinia Perityle ERASE : PIP EP PP PP LP oe Lt 2. fo LL, LL Le a emoryl ee ee ee ee = CODRCCGORDCCCROORORUGURRORGGGHORRRR008; Phacelia a distans a Th Phacelia aS ET LT IT EE a RE Ee TE SD a | crenulata Palafoxia * . Ad Al ed linearis 40 60 80 100 120 140 210 Growing Season (days) Fic. 3. Phenological events of ephemeral plants at Boyd Deep Canyon Desert Re- search Center following mid-September rainfall. See Fig. 2 for explanation of symbols. California Palafoxia linearis is now called P. arida. summer, whereas most other species ceased flowering by late April or May. Langloisia schottii and Camissonia claviformis did not begin reproductive activity until after 60 days. Camissonia claviformis flow- ered into June but Langloisia reproduction was restricted to a period of less than two weeks. Cryptantha spp. had an extensive period of reproduction lasting from mid-March through June. The next growing season was initiated by more than 13 cm of rain on 10-11 September 1976 (Fig. 3). The pre-reproductive period was longer in individuals that germinated in the fall than in those that germinated the previous spring, for all species except Cryptantha. Cryptantha, Schismus barbatus and Perityle emoryi entered the re- productive phase relatively early in the season; however, Schismus remained in the bud stage and did not flower until later. Cryptantha and Perityle began reproducing while ample soil moisture was avail- able from 55 mm of precipitation in scattered storms during late Sep- tember, October, and November. Plantago insularis and Chaenactts fremontii remained vegetative for a longer period, becoming repro- ductive prior to January rains. Camissonia claviformis and Chaenactts carphoclinia became established following fall rains but died without reproducing. (Chaenactis survived well into December.) Phacelia dis- tans and P. crenulata remained vegetative throughout the winter sea- 160 MADRONO [Vol. 29 -<-« Chaenactis 3 SMELL AES aR en = “a, wey an = = os Plantago Mu, we a, peach Cryptantha at \ | 7 oN \ Camissonia % co} \ %, A \ “ % A \ é ~ ON \ g 9 an ee . . i se =) | 2) . ’ % Sf : % ES, 1 \ } xX N N \ \ \ \ \ \ N N \ \. AN N A e 40 60 80 100 120 140 Growing Season (days) Fic. 4. Survivorship curves for selected species of desert ephemerals, spring 1976. son, initiating reproduction coincident with drying conditions during the spring. Recruitment following January precipitation was minimal for Plantago with none of the seedlings surviving to reproduce. All species set seed during February and March but only Camissonia claviformis, Cryptantha and Palafoxia arida remained alive until early April. Survivorship. Examples of contrasting survivorship curves for the spring 1976 growing season are presented in Fig. 4. Chaenacttis fre- montii and Plantago had similar high survival of seedlings followed by rapid post-reproductive mortality, whereas Cryptantha spp. and Camissonia claviformis had greater mortality in the seedling stage. Cryptantha and Camissonia showed increased survival following a midseason rain, a phenomenon not noted in Plantago and Chaenactts. Sampling methods made it impossible to construct survivorship curves for the 1976-1977 winter growing season because January re- cruits were indistinguishable from those germinating earlier. Early in the growing season, the survivorship curves of Plantago and Camis- 1982] BURK: DESERT EPHEMERALS 161 sonia did not have the concave form noted in Fig. 4. Instead, all species that germinated had very low mortality in the seedling stage and a more pronounced convex curve than during the previous spring. DISCUSSION Ephemerals are placed into three groups based on germination re- sponses noted in this study. Most species germinate during the cool season, being either restricted to midwinter (January or February) or germinating in response to rain after August. Several species (e.g., Camissonia claviformis and Chaenactis spp.) germinated in September and again in midwinter without intervening seed production. This may imply that germination requirements are dimorphic in most mem- bers of the fall to winter germinating group. Further observations are necessary to assure that microhabitat differences and seed age did not cause differential reproduction. Only during the spring of 1976 was there establishment of seedlings after the February cohort. Only Chaenactis carphoclinia germinated in response to the midspring rain in 1976. The other species may have gone into secondary dormancy, or perhaps the seed pool was depleted by midwinter germination. The former hypothesis is the more plausible one for species subjected to increasing temperature and unpredictable rainfall in late spring. The C4, summer ephemerals were represented only by Amaranthus fimbriatus and Euphorbia setiloba. The number of C4 annuals is very low compared to that in similar terrain farther to the east in the Chi- huahuan Desert of New Mexico (Syvertsen et al. 1976). The lower amount of rainfall in Deep Canyon during the summer months is the probable reason. Amaranthus was associated with the organically rich- er and more shaded habitats provided by shrubs. It may be restricted to the moderated conditions provided by the shrubs. The various species of Cryptantha form a complex of ephemerals that are potentially very long-lived and which can become established in any season from late summer through spring. The adaptations that allow plants that become established in August to remain alive and reproductive through December with less than 1 cm additional rain are the subject of further investigation in our laboratory. Ephemerals that germinate in response to the same rainfall show remarkable variations in phenological patterns. Some, like Plantago insularis, Phacelia distans, and Phacelia crenulata are capable of completing the life cycle very quickly following February germination. However, these same species have an extended period of vegetative growth following September germination and do not initiate repro- duction even when subjected to desiccation in October and November. It appears that the synergistic effects of soil water and temperature may be important in determining life cycle events in these species. Camissonia claviformis and Chaenactis carphoclinia may reproduce 162 MADRONO [Vol. 29 only during the spring months because both died before flowering, presumably because of a lack of water following fall germination. In both species new January seedlings completed the life cycle during the spring, even though significantly less rainfall occurred. Schismus barbatus is the only introduced ephemeral encountered in the sample plots. This species appears to be capable of: 1) initiating reproduction very early, whether germinating in fall or winter; 2) ad- justing the life span effectively depending upon. conditions; and 3) producing a large number of viable seeds. Schismus had a density about four times greater (288 vs. 70 m ”) than the densest native species, Filago californica. High phenotypic plasticity and successful germination of large numbers of seedlings may be related to its success as an alien species. Palafoxia arida was reported by Went and Westergaard (1949) (as P. linearis) to be unrestricted in germination. However, in that study, Palafoxia germinated only in the late summer and fall. It was ob- served flowering during all seasons in the vicinity of my plots. Comparison of survivorship curves for species that germinate in different seasons or grow under different temperature and rainfall regimes shows that determination of a characteristic form for a species is not possible. For example, Plantago insularis had a typically convex survivorship curve when germinating in February, whereas Septem- ber cohorts had a concave survivorship plot. Such variations in pop- ulation dynamics in different cohorts may be associated with consid- erable flexibility in resource allocation patterns from season to season. Detailed comparisons of resource allocation patterns in individuals from fall and winter cohorts need to be made to clarify the relation- ship. ACKNOWLEDGMENTS I thank Jan Zabriske and the staff of Deep Canyon for their help and support, M. Cruzan for help with the data, and Jerry Baskin, who commented on an earlier draft of the manuscript. LITERATURE CITED BEATLEY, J. C. 1974. Phenological events and their environmental triggers in Mojave Desert ecosystems. Ecology 55:856—-863. Burk, J. H. 1977. Sonoran Desert. In M. G. Barbour and J. Major, eds., Terrestrial vegetation of California, p. 869-889. Wiley-Interscience, NY. CLARK, D. D. and J. H. Burk. 1980. Resource allocation patterns of two California- Sonoran Desert ephemerals. Oecologia 46:86—91. JUHREN, M., F. W. WENT, and E. PHILLIPS. 1956. Ecology of desert plants, IV. Combined field and laboratory work on germination of annuals in Joshua Tree National Monument, California. Ecology 37:318—330. MuLroy, T. W. and P. W. RUNDEL. 1977. Annual plants: adaptations to desert en- vironments. Bioscience 27:109-114. Munz, P. A. 1974. A flora of southern California. Univ. California Press, Berkeley. SYVERTSEN, J. P., G. L. NICKELL, R. W. SPELLENBERG, and G. L. CUNNINGHAM. 1982] BURK: DESERT EPHEMERALS 163 1976. Carbon reduction pathways and standing crop in three Chihuahuan Desert plant communities. Southwest Natur. 21:311-320. TREVIS, L., JR. 1958a. Germination and growth of ephemerals induced by sprinkling a sandy desert. Ecology 39:681—687. . 1958b. A population of desert ephemerals germinated by less than one inch of rain. Ecology 39:688—695. WENT, F. W. 1948. Ecology of desert plants. I. Observations on germination in Joshua Tree National Monument. Ecology 29:242-253. 1949. Ecology of desert plants. II. The effect of rain and temperature on germination and growth. Ecology 30:1-13. and M. WESTERGAARD. 1949. Ecology of desert plants. III. Development of plants in the Death Valley National Monument, California. Ecology 30:26—38. ZABRISKE, J. G. 1979. Plants of Deep Canyon and the Central Coachella Valley, California. Univ. California, Riverside. (Received 9 Jul 1981; revision accepted 15 Feb 1982.) THE VEGETATION OF THE RAE LAKES BASIN, SOUTHERN SIERRA NEVADA Mary T. BURKE Department of Botany, University of California, Davis 95616 ABSTRACT Stands of subalpine-alpine vegetation from the southern Sierra Nevada were sampled (n = 97) and compared using Braun-Blanquet methodology. The study was undertaken (1) to characterize, using floristic criteria, the prinicpal vegetation types occurring within the study area; (2) to consider the influence of environmental factors on vegetational patterns; and (3) to provide a basis for making floristic and vegetational comparisons within the study area and elsewhere. The vegetational relevés (stand surveys) were organized into tables and assigned rank in a hierarchical system solely on the basis of floristic affinities. Five alliances and twelve associations were recognized and described phytosociologically and ecologically. Four major environmental features were consid- ered to order the vegetation in the landscape: moisture regime, snow cover, elevation, and substrate stability. Festucion brachyphyllae occurs in truly alpine sites within the study area. Primulion suffrutescentis is characteristic of both mid- and high-elevation sites, especially on north slopes where the snow lies late and the growing season is short. Pino-Caricion exsertae occurs on the gravelly slopes and terraces in the mid-elevation Pinus albicaulis subalpine forest. Monardello-Holodiscion microphylli is characteristic of the usually dry, stabilized talus of mid-elevations. Dodecathion redolentis represents the mid-elevation meadow of streamsides and topographic lows that remain moist to wet throughout the season. A total of 277 vascular plant taxa belonging to 38 families were found within the study area. In the Sierra Nevada of California, areas of alpine habitat are ex- tensive. The well-developed alpine flora is composed of a large en- demic (locally derived) element and a relatively small cosmopolitan arctic-alpine group (Major and Bamberg 1967). The uniqueness of the alpine flora is not duplicated in any other temperate North American mountain system; yet very few qualitative or quantitative descriptions of the vegetation exist. The flora of the Sierra Nevada alpine zone has been described by Smiley (1915), Hall and Grinnell (1919), Sharsmith (1940), and Howell (1951). The earliest very generalized vegetation descriptions of the Sierra Nevada occur in the work of Klyver (1931) and Sharsmith (1940). Munz and Keck (1949, 1950) and Munz (1959) consider the alpine region to support one broad plant community in their scheme of vegetation classification. Klikoff’s (1965) survey work in the Gaylor Lake Basin relating microenvironmental factors to vege- tation pattern was followed by Pemble’s (1970) more detailed phyto- sociological description of the Slate Creek, Dana Plateau, and Convict Creek alpine vegetation. Pemble, using 188 stand surveys with 175 species, characterized the vegetation of these areas, relating it to two MADRONO, Vol. 29, No. 3, pp. 164-176, 9 July 1982 1982] BURKE: VEGETATION OF RAE LAKES BASIN 165 sequences of ecological factors: soil parent material (granitics, non- carbonate metamorphics, and marble) and topography, especially as it relates to water availability. Chabot and Billings (1972) discussed the phytogeography and physiological ecology of Sierra Nevada alpine plants near Bishop Pass. Taylor (1976) has characterized the timber- line vegetation at Carson Pass, the northern limit for many alpine plants in the Sierra Nevada. The survey of the vegetation of the Rae Lakes Basin was made in a period of ten weeks during the summer of 1978. The aims of the survey were: 1) to characterize, using floristic criteria, the principal vegetational types occurring in the Rae Lakes Basin; 2) to provide a basis for considering the influence of environ- mental factors on spatial distribution and floristic composition of the vegetation; and 3) to allow floristic and vegetational comparisons to be made within the basin and with other areas in the Sierra Nevada and elsewhere. METHODS The field methods adopted followed standard Braun-Blanquet vege- tation sampling procedures (Birks 1973, Mueller-Dombois and Ellen- berg 1974). Uniform areas of vegetation (stands) were selected for description in the field. The selection of stands was subjective; an effort was made to record the variation of vegetation within the study area. One plot was sampled within each homogeneous stand area. The size and shape of the sample plots were not standardized, but were adjusted according to the character of the vegetation sampled. Esti- mates of percent cover of the plot area of rock, bare soil, and vege- tation (by species and strata) were made. All taxa occurring within the plots, excluding bryophytes and lichens, were recorded and cover- abundance estimated visually according to the Braun-Blanquet com- bined scale: r,+,1,2,3,4,5 (Mueller-Dombois and Ellenberg 1974, p. 60). Voucher specimens are deposited in DAV and in the Sequoia- Kings Canyon National Parks Herbarium. Nomenclature follows Munz (1959, 1968) unless otherwise noted. Environmental data for each stand, including slope, exposure, elevation, soil type, pH (Hellige pH kit), and moisture regime, were noted in the field. Estimates of moisture regime for surface and subsurface soils were assigned as fol- lows: 1) dry throughout most of the growing season; 2) moist through- out the growing season; and 3) wet, soil at field capacity or surface water visible throughout the growing season. Description of study area. The Rae Lakes Basin is located in Fresno County, California, about 30 km east of Cedar Grove and 20 km west of Independence. It lies in the southeastern Sierra Nevada, just west of the main crest (36°49’N, 118°24’W, USGS Mt. Pinchot quadrangle, 1953). The study area is an alpine lake basin that is approximately 1 km wide, tending north-south and bounded on the east, west, and 166 MADRONO [Vol. 29 south by natural watershed divides. To the north, an arbitrary line above the limits of the red fir forest defines the study area. The to- pography of the Rae Lakes Basin is extremely rugged: repeated gla- ciation of the entire region during the Pleistocene has left a legacy of steep arétes and ragged peaks, talus fields, and unstable scree slopes. This classically alpine topography, with its effects on local temperature and soil moisture, creates a myriad of microenvironments in a geo- graphically small area. Elevation is from 3046-4040 m, encompassing a wide variety of subalpine and alpine habitats. Climate may be regarded as consisting of moisture and heat regimes. Mean annual precipitation in the Rae Lakes Basin is calculated to be approximately 99 cm, based on extrapolation from Charlotte Ridge and Ellery Lake snow survey and precipitation records (Burke 1979). Although annual precipitation is high in the Sierra Nevada, it is con- centrated in the form of snow in the non-growing season. In the sum- mer, most cyclonic storms are diverted from the Sierra Nevada and, with the exception of occasional orographic thunderstorms, summer precipitation is generally low (Rae Lakes averaged 43-50 mm in sum- mer precipitation for 1972-1976). Mean monthly temperatures are cal- culated to range from —3°C (December) to 13°C (July). Using cal- culated mean monthly temperatures and precipitation figures, a Thornthwaite climatic diagram (Thornthwaite 1948, Thornthwaite and Mather 1957) was generated and evapotranspiration values de- termined (Burke 1979). Assuming 100 mm of available water stored in the soil, potential evapotranspiration (PotE) can exceed precipita- tion by 72, 91, 87, and 53 mm per month in June, July, August, and September, respectively. This results in use of soil-stored water and some drought (water deficit). In the Rae Lakes Basin, availability of and demand for water are out of phase. Four separate granitic intrusive masses occur in the Rae Lakes Ba- sin, underlying virtually all of the study area. Dark colored mafic igneous and hybrid rocks predominate in the southern end of the basin, however, and a single small outcropping of marble occurs towards the north end of the study area. No pedological studies have been carried out in the high country of the Sierra Nevada. Generally, most alpine soils are medium to coarse textured, excessively drained, have a low inherent fertility, contain large amounts of stone, cobble, and gravel, and are strongly acid. These soils are characterized by a high content of weakly decomposed organic matter in the surface horizons, weak granular structure, and silt loam textures with low (5—15 percent) clay content (Knapik et al. 1973). Due to the instability of the alpine land- scape, alpine soils are generally immature with no or weakly developed horizons. Horizons that form are often mixed by frost disturbance or mass wasting, or destroyed by erosion. Acidic soils result wherever granitic rocks are weathering; where mafic or calcium rich rocks are weathering, more alkaline soils are formed. 1982] BURKE: VEGETATION OF RAE LAKES BASIN 167 Phytosociological classification. Although relevés were sampled in as wide a range of habitat conditions as possible, all vegetational relevés are organized solely on the basis of floristic affinities, irrespec- tive of habitat. The matrix of stands and plant taxa in the association tables was initially sorted by the Ceska and Roemer computer program (1971) as modified by D. Randall and D. W. Taylor, and further rearrangements of the table were done by hand (Burke 1979). Figure 1 summarizes the final synthetic Rae Lakes Association Table (see Table 1 in Burke, 1979; 97 stands, 146 species; available at cost from author). The following list of the plants (Table 1) included in the species groups of the Rae Lakes Basin should be used in conjunction with the Rae Lakes Association Table Summary (Fig. 1). Based on a comparison of floristic relationships, relatively homogeneous relevés are grouped into associations, and each association assigned to an alliance. Syntaxa are named according to the published rules of the Code of Phytosociological Nomenclature (Barkman et al. 1976). Alli- ances are indicated by the “-ion” termination and associations by the ““etum” suffix. DESCRIPTION OF THE RAE LAKES BASIN PLANT COMMUNITIES Two levels of vegetational pattern are recognized, including five alliances and twelve associations. Detailed descriptions are provided only for the associations, the lowest ranking units in the classification. Careful use of the Association Table Summary (Fig. 1) together with the list of species groups will yield important floristic information about each association and should be considered an integral part of the following association descriptions. Ecological information about alliances may be obtained through generalization of the combined data for all associations included within a particular alliance. The letters within brackets are the abbreviations of association names used in Fig. 1. Number in parentheses is the number of stands sampled. De- tails on nomenclature type relevés is included in Table 1 in Burke (1979). Associations of the Rae Lakes Basin I. Festucion brachyphyllae. A. The Festuco-Penstemetum davidsonii [Fe.-Pe.da.] (2) occurs at high elevation (3410-3512 m) and is best developed where local to- pography is favorable for snow accumulation. The sites are well drained and surface water is never present. Relief varies from flat ridges to very steep slopes, but best development is seen on stabilized gravelly or talus slopes, or in the rock crevices and shallow sandy terraces of granitic outcroppings. Type relevé: 027. 168 MADRONO [Vol. 29 TABLE 1. RAE LAKES BASIN SPECIES GROUPS. SPECIES GROUP 1A: Penstemon davidsonii, Carex helleri, Ivesia pygmaea, Silene sargentit, Festuca brachyphylla, Calamagrostis purpurascens. SPECIES GROUP 1B: Erigeron compositus var. glabratus, Eriogonum ovalifolium var. nivale, Haplopappus macronema, Ribes cereum, Hulsea algida, Polemonium eximium, Ranunculus eschscholtzit var. Oxynotus, Oxyria digyna, Anelsonia eurycarpa (Hitch- cock et al. 1964), Arabis lemmonii vars., Draba brewer1, D. oligosperma, Poa rupicola, Androsace septentrionalis subsp. subumbellata. SPECIES GRouP 2A: Sedum rosea subsp. integrifolium, Salix orestera, Erigeron pet- iolaris, Juncus drummondi1, Potentilla breweri, Antennaria umbrinella. SPECIES GROUP 2B: Carex exserta, Lewisia sierrae, Saxifraga aprica, Penstemon heterodoxus, Calyptridium umbellatum, Rumex pauctfolius, Trisetum spicatum, Lew- isia nevadensis. SPECIES GROUP 3A: Pinus albicaulis, Juncus parryi, Selaginella watsoni, Poa ner- vosa, Eriogonum incanum, Antennaria alpina var. media, Antennaria rosea, Arabis platysperma var. howellii, Cryptogramma acrostichoides. SPECIES GROUP 3B: Penstemon newberryi, Carex rossii, Phacelia frigida, Holodis- cus microphyllus, Achillea lanulosa subsp. alpicola, Eriogonum nudum vars., Sitanion hystrix, Monardella odoratissima vars., Stipa occidentalis, Castilleja applegatei vars.., Erysimum perenne, Cirsium tioganum, Muhlenbergia richardsonis , Leptodactylon pun- gens subsp. pulchriflorum, Artemisia ludoviciana subsp. incompta, Cystopteris fragilis , Carex straminiformis, Stephanomeria tenuifolia, Ribes montigenum, Melica stricta. SPECIES GROUP 3C: Carex congdonii, Primula suffrutescens. SPECIES GROUP 3D: Senecio fremontii, Hieracium horridum, Arenaria nuttallit subsp. gracilis, Pellaea brewert. SPECIES GRrouP 4A: Cryptantha glomerifiora, Polygonum kelloggit, Gayophytum ra- mosissimum, G. racemosum, Mimulus suksdorfii, Mimulus brewert, Collinsia parvi- flora, Potentilla diverstfolia. SPECIES GROUP 5A: Epilobium angustifolium, Phyllodoce breweri, Veratrum cali- fornicum, Habenaria dilatata, Senecio triangularis, Senecio pauciflora, Lilium kel- lyanum, Veronica wormskjoldii (Hulten 1968). SPECIES GROUP 5B: Delphinium polycladon, Carex spectabilis, Agropyron trachy- caulum, Helenium bigelovii, Aconitum columbianum, Salix lemmonii, Potentilla fru- ticosa, Castilleja miniata, Thalictrum fendleri, Aquilegia formosa, Deschampsia caes- pitosa. SPECIES GROUP 5C: Muhlenbergia filiformis, Dodecatheon redolens, Senecio scor- zonella, Allium validum, Trifolium monanthum vars., Phleum alpinum, Sibbaldia pro- cumbens, Perideridia parishii, Epilobium halleanum, Carex microptera, Carex hetero- neura. SPECIES GrRouP 5D: Calamagrostis breweri, Mimulus primuloides var. pilosellus, Aster alpigenus subsp. andersonii, Carex fissuricola, Swertia perennis, Ledum gland- ulosum, Vaccinium nivictum, Danthonia intermedia, Pedicularis attollens, Eleocharis pauciflora, Gentiana newberryi, Kalmia polifolia var. microphylla, Potentilla drum- mondii, Luzula orestera, Vaccinium occidentale. SPECIES GROUP 5E: Epilobium oregonense, Calamagrostis canadensis, Gentiana holopetala, Carex rostrata. SPECIES GROUP 5F: Scirpus criniger, Pedicularis groenlandica, Viola macloskeyt, Erigeron peregrinus subsp. callianthemus, Geum macrophyllum, Luzula comosa. ASSOCIATED TAXA: Arabis lyallii var., Carex festivella, Carex multicostata, Carex phaeocephala, Carex raynoldsii, Epilobium hornemanii, Heuchera rubescens, Koeleria cristata, Pinus balfouriana, Pinus contorta subsp. murrayana, Poa incurva, Potentilla glandulosa, Senecio integerrimus var. major, Solidago multiradiata. 169 BURKE: VEGETATION OF RAE LAKES BASIN 1982] “x dnois sareads jo Jaysnjo & YIM saAaad Jo a8vyUadIed [Tews ev Jo sduaseid ay} sjuasaidai azis [[N} uvy} Sse[ Je UMEIP xOq peysed ‘UOTVBIIOSse Jo saAafat aos ut JuaseId YX Anois satdads jo J9}sN[D (3 “UOTBLIOSSB JO SQAVTeI [[B JSOWTE 1O [fe UL juasaid x dnoi3 saieds jo Jaysn[Q [J :puesaT ‘6/61 yng ‘T Iquy Ul a[qeireae eyep ajajdwoyD “TI eqey Ul pepnpur st (LY) exe} peyelvosse pue (7S 0} WI) sdnoid satsads ut exe} [[e Jo 4st] aJajduroo W ‘sasayjuared Ul SMOT[OJ UONVIDOSSe Jad s9Aafai JO SIAqUINN “UOTNAIAIISap }S$9} 9} UL pasn Wa}sAs BULIIQUINU IY} BUIMOT[OF *19}}9] pue [erauinu uewol e Aq paynuapt are pue do} ay} ssoroe payelAsiqqe are SUOT}BIIOSSe JO SIUTeN] ‘AICUIUINS 3[qGe} UOTIBIIOSSY Saye] vey 1 OY ¢-—---—----—~--------~—--- -------- --------- ------- spue}s ynoyBnoiy}, paiajjeos——-—-—----—-—---—--——--—-----——---------- iv ae eee 7 as e=3 L-u4 o) Ve) vt SdNOYD SaldadS eee eine ens ae see ee ie Se a bese jJL—--~-4 besa Eosoed mes at = & oe came ee pu Rea we S ee q VE i oa ee ees ee 5 ! ie ee a az 1 Us ia (ieee Pee | ae ye ig-m¢ ESS) Xa ey-eD] AO IQ- 94 gq 170 MADRONO [Vol. 29 B. The Festuco-Eriogonetum ovalifolii var. nivalis [Fe.-Er.ov.ni. ] (9) occurs in very high elevation, truly alpine sites (3386-3712 m, av. 3535 m). It is common in steep, loose talus slopes where soil moisture is extremely low and wind exposure is very high. The topographic location—high windy ridgetops—of some of the stands suggests that it may be able to tolerate sites that are blown free of snow in the winter. Type relevé: 043. II. Pino-Caricion exsertae. A. The Calyptridio-Caricetum exsertae [Ca.-Ca.ex.] (10) is a Carex exserta meadow type, well-developed at mid-elevations (3162—3301 m, av. 3232 m). This association is well represented throughout the Rae Lakes Basin, covering large areas of moderately dry, gently sloping terraces. Although this association is found at rather high elevations, in the Rae Lakes Basin it is better considered a dry meadow type on open gravel within the Pinus albicaulis subalpine forest. Carex exserta consistently accounts for 25—75 percent of the cover in stands sampled for this association. Type relevé: 078. B. The Erigeronto-Caricetum exsertae [Er.-Ca.ex.] (6) represents a Carex exserta meadow type occurring in relatively moist sites in the Pinus albicaulis subalpine forest. Although Pinus albicaulis is usually present, it has relatively low cover compared to the Junco-Pinetum albicaulis. Best developed at moderate to high elevations, this asso- ciation was sampled at sites from 3159-3557 m (av. 3336 m). The sites have good drainage and although the surface is often dry, subsurface moisture is usually present. These sites often receive meltwater from late-lying snowbanks. The sites are often rocky, but with more veg- etative cover than the higher elevation associations found on talus slopes. Type relevé: 052. C. A high elevation forest type, the Junco-Pinetum albicaulis [Ju.- Pi.al.] (8) was sampled at elevations from 3192-3411 m (av. 3287 m). The soil is generally dry and bare rocky spots predominate under the open canopy of this forest. Surface moisture is rare where this asso- ciation is well developed. It is characterized in part by a high Pinus albicaulis cover and the absence of a shrubby understory. Pinus con- torta subsp. murrayana and P. balfouriana are also present at Rae Lakes and are common trees throughout this subalpine forest. Slopes range from relatively flat areas on the basin floor to 30°. Type relevé: O31. Ill. Primulion suffrutescentis. A. The Primuletum suffrutescentis [Pr.su.] (6) occurs on talus slopes at the upper limits of the Pinus albicaulis subalpine forest. Elevations are from 3295-3472 m (av. 3418 m). Soils are extremely dry and rel- atively acidic. The talus is generally large and stable on slopes from 12°—35° (av. 27°). This association is characterized in part by abundant 1982] BURKE: VEGETATION OF RAE LAKES BASIN igh Primula suffrutescens, often associated with Carex congdonii and Er- togonum incanum. Type relevé: 046. IV. Monardello-Holodiscion microphyllt. A. The Stipo-Eriogonetum nudi [St.-Er.nu.] (22) is extremely com- mon throughout the study area at elevations from 3131-3460 m (av. 3295 m). This association is typical of very dry, rocky outcroppings or stable talus in the Pinus albicaulis subalpine forest. Sites are usually free of snow early in the season and the gravelly soils are moist only after a summer rain or snowfall. Slopes sampled are from 2°—52° (av. 26°). The association is best developed on slopes with western and southwestern exposures. Shrubby species, specifically Holodiscus mi- crophyllus and Ribes montigenum, are a conspicuous part of this as- sociation in many relevés. Type relevé: 034. B. The Holodisco-Mimuletum suksdorfit [Ho.-Mi.su.] (10) includes the only predominantly annual species group in the Rae Lakes vege- tation (Species Group 4A) and occurs in small pockets of damp soil trapped on otherwise dry, sunny rock outcroppings. The soil often shows evidence of frost heaving, and needle ice was observed on a number of occasions. Locally, these small annuals can be exceedingly abundant and yet, because of their minute stature, they account for very little cover. This association was sampled in stands from eleva- tions of 3192—3277 m, with an average of 3240 m. Slopes are generally steep, from 18°-48°, with an average 31°. Type relevé: 006. V. Dodecathion redolentis. A. The Potentilletum fruticosae [Po.fr.] (1) was sampled by a single relevé (relevé 104) on a marble outcropping northeast of Dollar Lake. This association occurs on wet seeps on mineralized soils at mid-ele- vations and is uncommon in the study area. Comparison of this releveé with association tables for Pemble’s (1970) study of alpine vegetation in the central Sierra Nevada and Taylor’s (1976) study of Carson Pass suggests that the Potentilletum fruticosae may be related to Pemble’s Salix anglorum antiplasta alliance and Taylor’s Dasiphora fruticosa- Potentilla breweri association. Further sampling is needed to char- acterize this vegetation unit more fully. Type releve: 104. B. The Seneciono-Caricetum spectabilis [Se.-Ca.sp.] (9) is a tall herb community at mid-elevations (3046-3228 m; av. 3125 m). It oc- curs on gentle to steep streambanks and in wet seeps. Slopes vary from 3°—45°, with an average of 18°. The soil is fairly well-drained, but retains a considerable amount of moisture throughout the field season. Where sites border seeps or slowly moving snowmelt streams, soils are generally at field capacity and are not well oxygenated. Stand- ing water is rarely present. Shrub cover values exceed those of the herbaceous layer at some sites, with Salix lemmonii forming dense thickets. Because of their density and stature, the herbaceous plants 172 MADRONO [Vol. 29 too may form tall thickets. The complexity of this meadow community suggests that it may be more than one association, but this is impos- sible to determine without further sampling. Type relevé: 088. C. The Vaccinio-Calamagrostietum breweri [Va.-Ca.br.] (10) forms very dense meadows with few rocks or bare spots. This association was sampled at elevations between 3052 m and 3283 m, with an average elevation of 3194 m. Soils of these meadows often have a thick sod layer of matted roots, are commonly poorly drained, and are at field capacity. Standing water, sometimes stagnant, may be present in shallow hollows. These meadows are frequent in topographic lows along lakeshores and in areas where the runoff of melting snowfield is channelled by macrorelief. Microrelief is generally undulating or gently sloping. Type relevé: 085. D. The Phleo-Dodecathetum redolentis [Ph.-Do.re.] (4) is another common streamside or wet lakeside community of mid-elevations. Stands sampled were from 3161-3228 m (av. 3201 m). This association occurs on moderate slopes ranging from 2°—22°, with an average 12°. The microrelief is usually convex and the sites are well-drained and generally a little drier relative to the moisture regime of the Seneciono- Caricetum spectabilis. The association is found along moderately flow- ing streams and where soils maintain a favorable moisture balance throughout the growing season. Type relevé: 040. ENVIRONMENTAL ANALYSIS Prominent discontinuous patterns characterize the vegetation of the Rae Lakes Basin. These distribution patterns can be related to gra- dients of the environmental complex. Elevation. Elevation obviously exerts a controlling influence on the large scale vertical zonation of the study area; however, even at high elevation, topographic position and substrate stability are more ac- curate predators of vegetational patterns. Certain associations, spe- cifically the Festuco-Penstemetum davidsonii and the Festuco-Eriogo- netum ovalifolii var. nivalis, are characteristic of truly high elevation habitats. Site moisture. Several of the Rae Lakes associations may be clearly separated by variation in the moisture regime of habitats (Fig. 2). At the wettest end of the spectrum, only one community, the Vaccinio- Calamagrostietum breweri, is found to be well-developed. Stream- banks and well-drained seeps also show a favorable water balance — Fic. 2. A. Moisture-pH summary. Each diagram summarizes both pH and site moisture data for moist-wet (data from 3 associations summarized), moderately dry (4) and dry associations (5) of the Rae Lakes Basin. Each box within each moisture-pH 1982] BURKE: VEGETATION OF RAE LAKES BASIN 173 A. MOISTURE-pH SUMMARY B. SLOPE / EXPOSURE moisture regime dry moist wet . esuesseceuaseusas eeas 4.0 moderately dry to moist associations (4) i | | | } mde dry associations (5) Junco- Pinetum albicaulis diagram summarizes the range of data points of all relevés within a single association. Dots represent single data points for associations with only 1 or 2 stands sampled. Each association is identified by a roman numeral and letter, following the numbering system used in the association description in the text. A general tendency from acidic to neutral soils is noted as one progresses along a moisture gradient from very dry sites to very wet sites. Detailed data for each association is available in Burke (1979). B. Slope/exposure. In each slope-exposure diagram, the concentric rings represent degree of slope in 10° increments from 0° to 50° and the lines represent the azimuth directions in 45° increments (i.e., the direction of slope). Stipo-Eriogonetum nud1 is characteristic of fairly steep slopes and is unusual on north-facing slopes; Calyptridio- Caricetum exsertae shows best development on slopes with a southwesterly aspect and stands of Junco-Pinetum albicaulis are rare on any southerly exposure. 174 MADRONO [Vol. 29 throughout the growing season and support tall herb (Seneciono-Car- icetum spectabilis) and characteristic fast-flowing streamside (Phleo- Dodecathetum redolentis) associations. Moderately dry to moist hab- itats at mid-elevations are utilized primarily by the relatively drier meadow types of the subalpine forest. The Erigeronto-Caricetum exsertae and Potentilletum fruticosae associations are best developed where the soil moisture regime is moderately moist throughout the growing season. Although surface moisture may be absent, subsurface moisture is usually available, often as meltwater seepage from late- lying snowbanks above the site. The Holodisco-Mimuletum suksdorfit represents clusters of annuals in pockets of damp gravel where melt- water or runoff is trapped in an otherwise dry, rocky habitat. These sites are subject to frost heaving early in the growing season and may be completely dry by mid-August. In a few of these sites, a continuous source of water is available throughout the dry summer season. The Calyptridio-Caricetum exsertae is the driest of the meadow types and is well-developed on moderately dry, sloping terraces. Habitats with the driest moisture regime are characterized by the Festucion brachy- phyllae and Primulion suffrutescentis alliances above tree limit, with best development seen on dry talus slopes at high elevation. At mid- elevations, dry stabilized talus supports the Stipo-Eriogonetum nudi and dry, gravelly slopes and terraces support the sparse Junco-Pine- tum albicaulis. Snow cover. In North American alpine areas, depth and duration of snow have repeatedly been found to be crucial determinants of vegetational patterns (Pemble 1970, Taylor 1976, Komarkova 1978). A complex relationship of regional climate, wind, and topography controls the distribution of snow and soil moisture, and therefore, the vegetation above treelimit. The Festuco-Eriogonetum ovalifoli var. nivalis is found on high elevation slopes that experience some degree of snow deflation during the winter. The Primuletum suffrutescentis and Festuco-Penstemetum davidsonii occur above the treelimit in sites that experience snow accumulation and a shorter growing season. In the gravelly habitats of these two associations, good drainage accounts for the lack of soil moisture during the growing season in spite of the heavy winter snowpack. At mid-elevations, the Stzpo-Eriogonetum nudi is snow-free early and very dry throughout most of the growing season. The wettest meadow association, the Vaccinio-Calamagrostte- tum breweri, is often seen in habitats where snow cover persists late into the growing season. In this association, occurring on sites where microrelief is flat or gently sloping and soil drainage is usually slow, winter snowpack correlates well with the availability of water throughout the growing season. Slope and exposure. No associations are characteristic of habitats with a specific degree of slope. Both a dry meadow type (Calyptridio- 1982] BURKE: VEGETATION OF RAE LAKES BASIN 175 Caricetum exsertae) and a wet meadow type (Vaccinio-Calamagros- tietum brewer) may be characteristic of flats and terraces with a slight degree of slope. Most plant associations in the Rae Lakes Basin are represented by stands distributed over a variety of exposures. The Calyptridio-Caricetum exsertae, however, shows best development on slopes with a southwesterly aspect. Fifty-five percent of the stands of the Stipo-Eriogonetum nudi also occur on southwestern exposures. Both of these associations reflect dry habitats with relatively long growing seasons in the subalpine forest. In contrast, the forested stands of the Junco-Pinetum albicaulis are rare on any southerly ex- posures (Fig. 2). Substrate characteristics. Because granitics underlie virtually all of the study area, few meaningful comparisons can be made between associations on the basis of parent material. However, it is noteworthy that the single relevé sampled on calcareous substrate (releve 104) has a distinct enough assemblage of taxa to be considered a separate as- sociation, the Potentilletum fruticosa. It is, however, possible to char- acterize at least the high elevation associations on the basis of substrate stability. The Festuco-Penstemetum davidsonii and Festuco-Eriogo- netum ovalifolii var. nivalis associations are both well represented at high elevation on dry soil. The Festuco-Eriogonetum ovalifolii var. nivalis is well-developed on scree and in the loose, gravelly soils of boulder fields. In contrast, the Festuco-Penstemetum davidsonii is best developed on stable substrates, from rock crevices and plateaus to trapped pockets of soil on stabilized talus slopes. The distribution of community types on the basis of soil pH is represented by Fig. 2. No obvious vegetation patterning is correlated with soil pH. A general tendency from acidic to neutral soils is noted as one progresses along a moisture gradient from very dry sites to very wet sites. Presumably, the snow meltwater may leach cations from the dry and wet sites equally in the early season. In wet sites and topographic lows, soil cations may be replaced by the cation-rich runoff from dry sites. Needles of Pinus albicaulis may also contribute to the acidity of some subalpine forest types. ACKNOWLEDGMENTS I thank Dr. Jack Major for his guidance and encouragement throughout the study. I also thank Jim Hickman for reviewing the manuscript and Bob Morgan for the illustrations. This study was funded in part by a grant from Sequoia-Kings Canyon National Parks (PX8550-8-093 1). LITERATURE CITED ATMOSPHERICS, INc. 1976. A High Sierra Precipitation Measurement Program, Final Report. Atmospherics, Inc. files, 565 E. Dalton, Fresno, CA 93727. 42 p. BARKMAN, J., J. MARAVEC, and S. RAUSHERT. 1976. Code of phytosociological no- menclature. Vegetatio 32:131-185. 176 MADRONO [Vol. 29 Birks, H. J. B. 1973. Past and present vegetation of the Isle of Skye. A paleoecological study. The Bryologist 77:277-279. BuRKE, M. T. 1979. The flora and vegetation of the Rae Lakes Basin, southern Sierra Nevada: An ecological overview. M.S. thesis, Univ. California, Davis. CreskKA, A. and H. ROEMER. 1971. A computer program for identifying species-releve groups in computer studies. Vegetatio 23:255-277. CHABOT, B. F. and W. D. BILLINGS. 1972. Origins and ecology of the Sierran alpine flora and vegetation. Ecol. Monogr. 42:163-199. Hatt, H. M. and J. GRINNELL. 1919. Life zone indicators in California. Proc. Calif. Acad. Sci. IV 9:37-67. . Hitcucock, C. L., A. CRONQUIST, M. OWNBEY, and J. W. THOMPSON. 1964. Vas- cular plants of the Pacific Northwest. Part 2: Salicaceae to Saxifragaceae. Univ. Washington Press, Seattle. HowELL, J. T. 1951. The arctic-alpine flora of three peaks in the Sierra Nevada. Leafl. W. Bot. 6:141-156. HULTEN, E. 1968. Flora of Alaska and neighboring territories: A manual of the vas- cular plants. Stanford Univ. Press, Stanford, CA. KLIKOFF, L. G. 1965. Microenvironmental influence on vegetational pattern near tim- berline in the central Sierra Nevada. Ecol. Monogr. 35:187-211. KLYVER, F. D. 1931. Major plant communities in a transect of the Sierra Nevada Mountains of California. Ecology 12:1—17. KnaPIk, L. J., G. W. SCOTTER, and W. W. PETTAPIECE. 1973. Alpine soil and plant community relationships of the Sunshine area, Banff National Park. Arctic Alpine Res. 5(3), part 2, Al61—A170. KOMARKOVA, V. 1978. Alpine vegetation of the Indian Peaks area. Jn R. Tuxen, ed., Flora et Vegetation Mundi, Band VII. Cramer, Germany. LOvE, A. 1954. Cytotaxonomic remarks on some American species of circumpolar taxa. Svensk Bot. Tidskr. 48:211-232. Major, J. and S. A. BAMBERG. 1967. Comparison of some North American and Eur- asian alpine ecosystems. In H. E. Wright and W. H. Osburn, eds., Arctic and alpine environments, p. 89-118. Indiana Univ. Press, Bloomington. MUELLER-DomBol!s, D. and H. ELLENBERG. 1974. Aims and methods of vegetation ecology. John Wiley and Sons, N.Y. Munz, P. A. 1959. A California flora. Univ. California Press, Berkeley. . 1968. Supplement to A California flora. Univ. California Press, Berkeley. and D. D. Kreck. 1949. California plant communities. Aliso 2:87—105. and . 1950. California plant communities. Aliso (suppl.) 2:199-202. PEMBLE, R. H. 1970. Alpine vegetation in the Sierra Nevada of California as litho- sequences and in relation to local site factors. Ph.D. dissertation, Univ. California, Davis. SHARSMITH, C. W. 1940. A contribution to the history of the alpine flora of the Sierra Nevada. Ph.D. dissertation, Univ. California, Berkeley. SMILEY, F. J. 1915. The alpine and subalpine vegetation of the Lake Tahoe Region. Bot. Gaz. 59:265-286. TaAyYLor, D. W. 1976. Ecology of the timberline vegetation at Carson Pass, Alpine County, California. Ph.D. dissertation, Univ. California, Davis. THORNTHWAITE, C. W. 1948. An approach toward a rational classification of climate. Geogr. Rev. 38:55—94. and J. R. MATHER. 1957. Instructions and tables for computing potential evapotranspiration and the water balance. Drexel Inst. Technol. Pub. Climatol. 10:183-311. (Received 21 May 1981; revision accepted 21 Dec 1981.) PINE SEEDLINGS, NATIVE GROUND COVER, AND LOLIUM MULTIFLORUM ON THE MARBLE-CONE BURN, SANTA LUCIA RANGE, CALIFORNIA JAMES R. GRIFFIN Hastings Reservation, University of California, Carmel Valley 93924 ABSTRACT Vegetation on 19 plots in crownfire-killed pine stands was studied. After three seasons pine seedling density ranged from 7/ha to 1685/ha in four Pinus coulteri stands. Pine seedling density was 188/ha in a P. ponderosa stand. In a stand dominated by P. lambertiana before the fire, pine seedlings numbered 3178/ha; but P. coulteri outnum- bered P. lambertiana 7:1. Shrub layers of Ceanothus spp. seedlings and hardwood tree sprouts ranged from 16 percent to 68 percent cover in these stands. Species richness was low in the P. lambertiana stand, averaging 20 species/0.1 ha plot with only one important herb species. Species richness was higher in the P. coulteri stands with one stand averaging 74 species/0.1 ha plot. By the second season Lolium multiflorum dom- inated the herb layer in all P. coulteri stands, and the native herbs had low density and biomass. Low Pinus and Ceanothus seedling survival in two P. coulteri stands may have been related to Lolium competition. The Marble-Cone fire of August 1977 covered 72,000 ha of the Santa Lucia Range in Monterey County, mainly in the Ventana Wilderness, Los Padres National Forest (Hammond 1977, Griffin 1978, Talley and Griffin 1980). Many mature pine forests on the higher ridges were heavily burned. This report summarizes observations on pine stands destroyed by crownfire on three of these ridges. Because of the burn’s wilderness status, vegetation recovery was allowed to proceed under “natural” conditions. However, an introduced annual grass, Lolium multiflorum, was sown for erosion control on much of the burn (Ham- mond 1977). This study describes the effects of this planted grass on species richness, shrub regeneration, and tree seedling survival during three postfire seasons. Mixed hardwood forests cover much of the Santa Lucia Range (Talley 1974, Sawyer et al. 1977). Three pine species are scattered in these hardwood communities, and stands dominated by each pine were studied. Within the Marble-Cone burn, Pinus lambertiana grows only on Junipero Serra Peak (JSP); P. ponderosa is more widely dis- tributed with conspicuous mature stands on Pine Ridge (PR). Pinus coultervi is a minor component of most mixed hardwood forests and adjacent chaparral, and it has dominant stands concentrated in the Chews Ridge (CR) region. Hammond (1977) summarized the large and expensive rehabilitation project that included the seeding of Lolium multiflorum. Large areas MADRONO, Vol. 29, No. 3, pp. 177-188, 9 July 1982 178 MADRONO [Vol. 29 TABLE 1. LOCATION AND HABITAT DETAILS OF PINE REGENERATION STUDY AREAS IN CROWNFIRE-KILLED FORESTS, MARBLE-CONE BURN. No. of Elev., aspect, slope, Dominant pine, most Name and location plots soil series important hardwood Junipero Serra Pk. (JSP), 4 1665-1760 m, 311-006°, Pinus lambertiana within 0.5 km of 25-45%, Junipero Quercus chrysolepis lookout loamy sand Pine Ringe (PR), 4 1360-1430 m, 171-231°, P. ponderosa 0.5—0.8 km e. of summit 20-30%, Junipero sandy loam Lithocarpus densiflorus Chews Ridge region Finch Canyon (CR-1), 3 1130-1170 m, 045-070°, P. coulteri 3 km e. of lookout 20-23%, Sheridan Q. agrifolia coarse sandy loam China Camp (CR-2), 3 1360-1450 m, 170-257°, P. coulteri 1.6 km s. of lookout 55-70%, Sur-Junipero QQ. chrysolepis complex Miller Ridge (CR-3), 3 1405-1450 m, 134-186°, P. coulteri 3.1 km sw. of 15-45%, Sheridan Q. chrysolepis lookout coarse sandy loam Summit (CR-4), 2 1490-1500 m, 075-095°, P. coulteri 0.2 km ne. of 25-28%, Sur-Junipero Q. wislizenii lookout complex of montane forest as well as chaparral were seeded. In October 1977, 460 t of seeds were aerially sown on 63,000 ha (7.8 kg/ha). Ground sampling by the seeding project suggested that seed distribution was reasonably uniform, but the final grass crop was patchy. Two regions having concentrations of “sensitive” plants, including study areas JSP and PR (Table 1), were excluded from the aerial seeding. In addition fuelbreak scars were hand seeded with a mixture of domestic grasses. This seeding included bulldozer trails near study areas CR-2,3,4 (Ta- ble 1). Although sediment loss and slope stability were studied in parts of the burn by several projects, no general report on soil erosion was compiled. No erosion data were gathered on my study areas or on comparable habitats elsewhere in the burn. The following observa- tions provide some background on how erosion affected my plots. On slopes over 60 percent at CR-2, soil loss started soon after the fire through dry creep. As Wakimoto (1979) emphasized, Lolium seeding can not control this type of gravitational flow. In late December 1977, hurricane force winds over the ridges blew ashes, soil, and some Lol- ium seeds off of the most exposed places. Then heavy rains in early 1978 removed surface soil from all plots. Rock “pedestals,” exposed rocks, and bare roots suggested that 3—8 cm of surface soil had washed 1982] GRIFFIN: POSTFIRE SUCCESSION 179 off most of the plots with slopes over 20 percent. On the steepest places rills up to 30 cm deep eroded down into the new soil surface. This wind and water erosion occurred before the Lolium formed any effective ground cover. Conrad (1979) noted that slow development of Lolium cover was common in rehabilitation projects on chaparral burns. STUDY METHODS Species lists for several regions of the Marble-Cone burn were for- tuitously included in previous work (Griffin 1975a). At PR and JSP (Griffin 1975b) I had specifically searched for plants on the sites that became study plots after the fire. Talley (1974, 1976) had prefire herb cover data relevant to the PR and JSP study areas, but his plots did not coincide with my plots. During the first postfire season I established 19 permanent plots (20 X 50 m) in pine stands killed by crownfire (Table 1). I measured diameters of all trees (>1 cm dbh). These diameters were slightly undersized due to bark loss by charring. The burned trees could be easily identified to species by bark and cone characters. In early summer of each season, during the period of maximum herb cover, I photographed every plot from permanent reference points. At the same time I counted all conifer seedlings per plot. Pinus coultert seedlings could not be readily separated from P. lambertiana seedlings the first season but were easily separated in following sea- sons. Starting in 1978 I sampled a 20 X 50 cm quadrat every 2 m along each of four transects per plot for a total of 80 quadrats. Each season I counted all shrub seedlings, recorded the presence of all rooted species, and visually estimated the percentage of bare soil in each quadrat. In 1979 visual estimates of current year Lolium cover were added to quadrat data, and in 1980 I estimated the cover of Lolium mulch. Shrub cover was sampled at a point every 2 m along four transects through each plot for a total of 100 points. All plots were visited at least one additional time per season to search for additional species, to observe pine seedling growth, and to take additional pho- tographs. The two CR-4 plots were established too late in 1978 to complete quadrat sampling, but in following years they were sampled by the same methods as the other plots. RESULTS Species richness. The vascular flora of the general region around the three study areas approached 400 species (Griffin 1975a). I found 162 species on the 19 plots over three seasons. Numbers of species on individual plots ranged from 19 at JSP to 80 at CR-2; species on individual quadrats ranged from zero on several areas to 10 at CR-2. 180 MADRONO [Vol. 29 TABLE 2. VASCULAR PLANT SPECIES RICHNESS ON THE STUDY AREAS. Species numbers represent trees (T), shrubs (S), and herbs (H) seen on 0.1 ha plots during three seasons and the average number of species per 20 X 50 cm quadrat each season. No. species all plots Average no. species 1978-1980 per quadrat Study area a's S H Total 1978 1979 1980 JSP 3 5 27 35 0.5 ibe 2.0 PR 5 7 45 57 0.6 222 201 CR-1 6 9 53 68 2.3 Def, 2.6 CR-2 5 9 86 100 2.4 4.5 4.6 CR-3 6 7 24 37 1.5 2.0 2.0 CR-4 6 10 40 56 — 225 2.4 The JSP area had unusually few species in the understory before the fire (Griffin 1975b, Talley 1976) and few species after the fire (Table 2). In 1978 only nine herb species grew on the 320 quadrats sampled at JSP. Only one of these, Lupinus cervinus, had grown on the prefire plots (Talley 1976). The only common herb was Gayophy- tum heterozygum. No grassland species were present. The richest P. coulteri area had almost three times as many species as JSP (Table 2). The CR-2 area had more species per plot than many oak woodland plots of the same size at lower elevations near Chews Ridge (Naveh and Whittaker 1979). During the first season at CR-2, 53 herb species were found on 240 quadrats. This forest borders a small grassland, and many species on the pine forest plots were typical grassland species. Native herb dominance. Only four perennial herbs were common during the first three seasons (Table 3). I had seen these species on the study areas before the fire, and individuals of these species seen after the fire were mostly sprouts not seedlings. Some Pteridium sprouts at PR were 20 cm tall 30 days after the fire, and Pteridium at PR was the most dominant perennial on any study area by 1980 (Table 3). Several perennials that were absent from the study areas before the fire germinated from dormant seeds after the fire. Lupinus abramsii at JSP was a good example (Talley and Griffin 1980), but none of these perennials became common on the study areas. The most common annuals (Table 3) were grassland species or plants that thrive in disturbed conditions. Several annuals reached a peak in coverage the second season and declined the third season— Claytonia perfoliata and Allophyllum divaricatum were good exam- ples. Allophyllum was rare in roadcuts near CR-2 before the fire. By the second postfire year it was conspicuous on and around the CR-2 plots, but it was absent from the frequency quadrats and rare else- where on the plots the third year. The best example of continued GRIFFIN: POSTFIRE SUCCESSION 181 1982] vd OT 6 Saploippu DIpD ve LG Of 6¢ Sv O¢ SN41JSIA SNARKY JOT b p I og Sb §€I UNUYINDD WNIPIA1d} J + + I> + + 02 8I Of 8i Zl eI ¢f or 2 UNITULO{YDI WNYDS) sTeIuudIag OZ OZ ¢ DIIDUDAQULAU IYJUDZ1AOY D 02 OT S SNIDAJSIUNY SNIOT IT oC 6 S Vv I SIyjOM SnmWOAg 0 oe 6 wWNnIDIIUADAIp WnycYydoy Pp Le we aa 87 te he (é Le ¢ UNIANGANGOQ]D WN1Of2A [ $2 al snupiy sang snjoT 97 «9 eT 0 0 + snupu snurgnT Z €Z 0 I> + 0 0 [> Je: 32 + Dass DYaIDY eT v + + + 2 8p 8 9€ 6 S ea Dypyofrag DIU0jKD]) + te? =e I I 0 v6 61 I> UNBAZOAIJAY UNIAYGOKDS OO. OOT 66 OOT 78 OOT 66 £=8b OOT OOT 66 ~»& ¥6 GL E> 20 + + UNAOY INU WN17OT sjenuuy 08 6L O08 6L 84 O08 64 84 08 6L $2 O8 62 SL° O8 62 81 satsads as) as (@) cao Tao Ud dSf seaie Apnys ‘syerpenb Aduanbasy uo jou ynq 407d UO sa1deds sozedIpUI ,.+,, SUOSBIS JUO jSBI] Je BULINp voIe ApN}s & UO Y%CZ jseI] 3" JO Satauanb -d1J pey satoads papnpouy “SNOSWAS AMIALSOg ATAHL ONIENG SVaAXY AGNALS NO sduapH NOWWOD dO (%) AONANOAMY ADVAGAY ‘¢ ATAVE 182 MADRONO [Vol. 29 expansion was by Gayophytum heterozygum, which I had not seen on JSP before the fire. In 1978 it was uncommon, but by 1980 it was ubiquitous over the summit. Many species were locally important after the fire in adjacent por- tions of the burn but were rare on the study areas, e.g., Argemone munita, Calystegia occidentalis, Cordylanthus rigidus, Dicentra chry- santha, Helianthemum scoparium, and Hulsea heterochroma. Pha- celia brachyloba and P. grisea are annuals that commonly respond to fire. Both were abundant in adjacent chaparral. Phacelia brachy- loba was rare on the forest plots; P. grisea became common at CR-4 (Table 3). Although Howitt and Howell (1964) reported Eriogonum sperguli- num subsp. reddingianum on JSP, I could not find this montane dis- junct on the peak prior to the fire. In 1978, however, a population less than 50 m in diameter grew on the summit near the plots. The fol- lowing season this population declined to a few plants in the midst of dense Gayophytum heterozygum. Ryegrass dominance. The JSP plots were not aerially seeded, and few seeds drifted in from adjacent seeding. In 1980 I saw no Lolium on the plots and few plants near the study area. The PR plots were not officially seeded either, but accidental seeding produced a sparse crop in 1978. This built up to 30 percent Lolium cover in 1980. Some nearby pine stands on PR had nearly 100 percent Lolium cover by 1979. All of Chews Ridge was sown with about 400 Lolium seeds/m?. This seeding produced a crop that Conrad (1979) described as gener- ally “outstanding.” Maximum height of this crop in 1978 was 120 cm, and in places I had difficulty walking through the grass. The CR-1 plots were representative of the denser grass stands. The first season 24 percent of this area remained bare, but the bare spots were mainly deep ash beds. The vegetated ground had dense Lolium about a meter tall; 237 out of 240 quadrats contained Lolium. By the second season bare ground dropped to 4 percent, and Lolium had 87 percent cover. On 34 percent of the quadrats in 1979 I could find no species except Lolium. Lolium maintained significant live cover the third season, but the plants were shorter and more spindly. By this time large piles of Lolium mulch had accumulated (Fig. 1). The CR-3,4 areas had Lolium density and cover approaching CR- 1 levels (Fig. 1). Of the areas sampled, the most vigorous growth of Lolium was at CR-4. At CR-2 Lolium density was low and cover was sparse the first year. Seventy-four percent of the soil remained bare; only half the quadrats contained any Lolium. Lolium cover at CR-2 never reached the levels of the other plots on Chews Ridge (Fig. 1). Shrub regeneration. Pine forests on the study areas have periodically 1982] GRIFFIN: POSTFIRE SUCCESSION 183 Mm conn, sMieser LIVE LOLIUM [] BARE SOIL = 100 wo C c RIM AY NT SS N rf NS . 5 LUALLLTTISSSS[SS 78 7980 787980 787980 7879 80 7980, 787960 Oe Saiz pce OF aA @ nga OO 4G nc Fic. 1. Ground cover estimates of bare soil for three postfire seasons, Lolium mul- tiflorum live cover (plus minor amounts of mulch) for second and third seasons, and accumulated Lolium mulch for third season. supported shrub layers of non-sprouting species that start after fires. The most recent fires that could have regenerated these shrub layers on the study areas were: JSP, 1901; PR, 1916; CR-1,2,4, 1928; CR-3, unknown. Shrubs resulting from those fires had nearly disappeared from the forests by 1977. The new shrub layers have hardwood tree sprouts, scattered sprouts of arborescent shrubs such as Rhamnus cal- ifornica, and abundant seedlings of three Ceanothus species. Only a small number of Arctostaphylos glandulosa sprouts and seedlings are present. The most widespread postfire shrub is C. integerrimus (Table 4). At JSP C. integerrimus reached 40 percent cover despite the loss of half the initial seedlings (Table 4). The first season these seedlings had essentially no herb competition. Seedling mortality was most evident where the seedlings were clumped together. At CR-3 not all the initial C. integerrimus seedlings could be count- ed because of the dense grass cover, and total seedling mortality was probably higher than shown in Table 4. By 1980 only a sparse cover of slow growing seedlings remained. The most vigorous C. integerrimus shrubs were at CR-4. During the winter of 1980-1981 rodents killed about 10 percent of these healthy shrubs. Thomomys bottae (pocket gophers) ate roots, and Microtus californicus (California voles) girdled stems. Ceanothus integerrimus was rare at PR, but C. papillosus, which 184 MADRONO [Vol. 29 TABLE 4. TOTAL SHRUB COVER, Ceanothus SEEDLING COVER, AND Ceanothus SEEDLING DENSITY. Total shrub cover includes shrub seedlings, shrub sprouts, and hardwood tree sprouts in shrub layer; density given in parentheses; “+” indicates seed- lings on plot but not on quadrats. Ceanothus cover % 1980 (Ceanothus density, no./m? 1978, 1980) Total cover Study area % 1980 C. integerrimus C. sorediatus C. papillosus JSP 43 40 0 (13.1, 6.7) (+, 0) PR 48 <1 2 32 (07357 0.3) (Ol 0.1) (321152) CR-1 16 5 (523505) CR-2 31 24 C3712) CR-3 27 13 (1.6, 0.6) CR-4 68 38 (e222) typically grows in chaparral, was the dominant shrub (Table 4). Arc- tostaphylos sprouts averaged 5 percent cover, the highest on any study area. At CR-1 C. integerrimus was also rare; at this lowest elevation study area C. sorediatus was important (Table 4). By 1980 this area had the lowest shrub cover and the lowest seedling density (Table 4). Pine regeneration. The pines on JSP burned on 8 August 1977. This was a relatively early date to expect high seed viability in P. lamber- tiana cones (Krugman 1966). If the seeds did not come from scorched cones on plot trees, the seeds must have dispersed up to 100 m from live trees to reach some plots. The heavier P. coulteri seeds would disperse far less, and some P. coulteri seeds probably came from pre- 1977 cones on burned trees on the plots. This species can store seeds in partially closed cones for several years (Minnich 1980). The viability of P. coulteri seeds in early August and any possible “ripening” effect in burned cones is not known. Regardless of where the seeds came from at JSP, in 1980 P. lam- bertiana had only 415 seedlings/ha to replace 651 burned trees/ha; P. coulteri had 2760 seedlings/ha to replace 98 trees/ha (Table 5). The P. coulteri seedlings were also two to three times taller than the P. lam- bertiana seedlings. In neither pine did a significant number of seeds germinate after 1978. The PR P. ponderosa trees burned a week before the JSP pines, and seeds falling from the dead pines at PR in early September had shriveled gametophytes that did not appear viable. Only 120 pine 1982] GRIFFIN: POSTFIRE SUCCESSION 185 TABLE 5. COMPARISON OF PREFIRE Pinus AND Calocedrus TREE DENSITIES IN CROWNFIRE-KILLED STANDS WITH SEEDLING DENSITIES THREE SEASONS AFTER THE FIRE. Density given in parentheses; an * indicates a few trees were present in the general area but were not on the plots. : Percent tree (seedling) density Tree (seedling) Study area density, no./ha Piso. Pate, P. po. C. de. JSP 749 13 87 0 (3178) (87) (13) (<1) PR 413 < 92 8 (200) (0) (94) (6) CR-1 857 100 (7) (100) CR-2 523 100 (310) (100) CR-3 243 100 a (90) (100) (0) CR-4 270 100 (1685) (100) seedlings/ha could be found in 1978. A larger number started in 1979, but only 40 percent survived. The seeds that produced these seedlings must have come from live trees 20-100 m away from the plots. By 1980 PR had only 188 P. ponderosa seedlings/ha to replace 379 trees/ ha (Table 5). Calocedrus decurrens seedlings were rare. The P. coulteri study areas showed great variability in pine regen- eration. On CR-1 plots 857 trees/ha were killed, and only 7 pine seed- lings/ha survived in 1980 (Table 5). Lolium was too dense for all the initial pine seedlings to be counted, but the scarcity of surviving seed- lings was clear. Since no P. coulteri trees survived within 500 m of this study area, additional seeds can not disperse into the area to supplement the rare 1978 seedlings. The CR-4 plots illustrate the other regeneration extreme. For each P. coulteri tree killed, six vigorous seedlings were growing after three years (Table 5). In 1980 some pine seedlings were 105 cm tall and already overtopping the Ceanothus shrub layer. Several dozen pine trees survived near the plots, and additional seeds might disperse onto the plots. However, almost no pine seedlings did start in 1979 or 1980. DISCUSSION Keeley et al. (1981) discussed southern California chaparral studies that show seeded Lolium growing at the expense of native herbs. In their own San Diego County survey, Keeley et al. found seeded burns with up to 28 percent Lolium cover. The dominance of Lolium on the Chews Ridge pine stands, however, greatly exceeded such chaparral 186 MADRONO [Vol. 29 results. Lolium cover, up to 87 percent at CR-1, greatly reduced the space and resources available for native herbs. Unfortunately, without any unseeded control areas available on Chews Ridge this Lolium influence could not be quantified. Lolium competition may not have prevented any native herbs from growing on the Chews Ridge plots, at least in low densities. But some species might not appear after the next fire because of unusually low dormant seed reserves returned to the soil by the few unhealthy herbs in the Lolium cover. If Lolium competition comparable to the CR- 1,3,4 areas had occurred at JSP, the tiny Eriogonum spergulinum population would have been threatened. The Chews Ridge area with the richest flora (CR-2) had the steepest slopes, the greatest erosion, and the least Lolium cover. Steep slopes increased the erosion, and the resulting soil loss reduced the supply of Lolium seeds. This soil loss also reduced the supply of native herb seeds. Yet, a variety of native species did start the first season. Prox- imity to a grassland seed source probably contributed to the higher species richness on these plots. Conrad (1979) emphasized that high shrub seedling mortality, par- ticularly in Ceanothus, was associated with Lolium cover. Data from two Chews Ridge areas support this generality. At CR-1 the 10-fold reduction in C. sorediatus seedling density must have been partly caused by Lolium competition in a relatively dry forest habitat. Low shrub seedling density and slow growth were also associated with high Lolium cover at CR-3. If Ceanothus understories are a desirable fea- ture in the development of pine stands after fire in this landscape, Lolium seeding had a negative influence on the CR-1,3 plots. In contrast, the results at CR-4 showed that Ceanothus can regen- erate successfully in Lolium under some conditions. By 1980 these plots averaged 2.7 shrubs/m?, and the shrubs were taller and more vigorous than at any other area—despite growing in dense grass. This area had a few seepage spots, and favorable soil moisture conditions may have contributed to high survival and rapid growth of Ceanothus seedlings. The only obvious threat to this developing shrub layer was heavy rodent browsing. Lolium provided a temporary grassland cover within the forest habitat. This unnatural grass cover permitted pocket gophers and voles to reach population highs. These browsers may locally thin the shrub layers until the rodents decline to more normal levels. Successful pine regeneration was generally associated with success- ful Ceanothus regeneration. An abundance of pine seedlings started at JSP in the C. integerrimus layer. But there was a significant shift from P. lambertiana dominance in the prefire forest to P. coulteri dominance in the postfire seedling crop. In terms of seedling growth P. coulteri is clearly superior. The silvicultural literature suggests that with time P. lambertiana should surpass P. coulteri in height. For 1982] GRIFFIN: POSTFIRE SUCCESSION 187 several decades, however, P. lambertiana will be scattered within P. coulteri thickets. At CR-4 there were more than enough healthy P. coulteri seedlings to replace the burned stand. At CR-2,3 not enough pine seedlings survived, and the new stands will be more open than their predeces- sors. This trend in declining P. coulteri regeneration ended with CR- 1 where the stand may not be replaced. Until the seedlings now on the area reach cone bearing age, this stand is endangered. Another burn within the next decade would eliminate the stand. By 1980 the PR plots did not have sufficient seedlings to replace the burned P. ponderosa trees. If no additional seedlings start, the new pine stand will have less than half the density of the old stand. The PR plots do have many live seed trees nearby, and there may be further seedling recruitment. Seeded Lolium seldom produces enough fuel on chaparral burns to support an early reburn. However, the growth of Lolium on the Mar- ble-Cone burn was so successful that fire became a real hazard. By July 1978 the dry grass provided continuous flash fuels that could have carried fire over significant portions of the burn. This flash fuel ac- cumulation was greater the following two summers. A summer fire on the CR-1,3 areas would have killed pine and shrub seedlings with no new seed supplies yet available. A fire in the other areas would not have been as uniformly damaging. Native herb layers at JSP and PR would not have carried a destructive surface fire during this period. Results of this pine forest study support those of many chaparral studies; significant soil eroded before Lolium cover developed; Lolium retarded growth of native herbs and increased shrub seedling mortal- ity. In this case Lolium also increased pine seedling mortality and created a high hazard for an early reburn. ACKNOWLEDGMENTS Portions of the field work were supported by the Pacific Southwest Forest and Range Experiment Station, Berkeley. I thank R. Breazeale and his staff on the Monterey Ranger District, King City for their help and interest in this project. LITERATURE CITED CONRAD, C. E. 1979. Emergency postfire seeding using annual grass. Chaps Newslet- ter, p. 5-8. Chaparral Research and Development Program. Forest Fire Lab., Riverside, CA. GRIFFIN, J. R. 1975a. Plants of the highest Santa Lucia and Diablo Range peaks, California. USDA For. Serv. Res. Pap. PSW-110. . 1975b. Ecological survey of Junipero Serra Peak candidate Botanical Area, Monterey District, Los Padres National Forest. Consulting report (P.O. 1180-PSW- 75), Pacific Southw. For. Range Exp. Sta., Berkeley, CA. . 1978. The Marble-Cone fire ten months later. Fremontia 6(2):8—14. HAMMOND, R. R. 1977. Marble Cone burn rehabilitation: accomplishment, critique, and recommendations. Monterey Ranger District, Los Padres Natl. For., King City, CA. 188 MADRONO [Vol. 29 HowltT, B. F. and J. T. HOWELL. 1964. The vascular plants of Monterey County, California. Wasmann J. Biol. 22:1—184. KEELEY, S. C., J. E. KEELEY, S. M. HUTCHINSON, and A. W. JOHNSON. 1981. Postfire succession of the herbaceous flora in southern California chaparral. Ecology 62:1608—-1621. KRUGMAN, S. L. 1966. Artificial ripening of sugar pine seeds. USDA For. Serv. Res. Pap. PSW-32. MINNICH, R. A. 1980. Wildfire and the geographic relationships between canyon live oak, Coulter pine, and bigcone Douglas-fir forests. In T. R. Plumb, techn. coord., Ecology, management, and utilization of California oaks. USDA For. Serv. Gen. Techn. Rep. PSW-44. NAVEH, Z. and R. H. WHITTAKER. 1979. Structural and floristic diversity of shrub- lands and woodlands in northern Israel and other mediterranean areas. Vegetatio 41:171-190. SAWYER, J. O., D. A. THORNBURG, and J. R. GRIFFIN. 1977. Mixed evergreen forest. In M. G. Barbour and J. Major, eds., Terrestrial vegetation of California, p. 359- 381. Wiley-Interscience, New York. TALLEY, S. N. 1974. The ecology of Santa Lucia fir (Abies bracteata), a narrow endemic of California. Ph.D. dissertation, Duke Univ., Durham, NC. . 1976. The role of fire within montane pine forests, Junipero Serra Peak, Los Padres National Forest. Consulting report (Sec. I, 000171), Lick Observatory, Univ. Calif., Santa Cruz. and J. R. GRIFFIN. 1980. Fire ecology of a montane pine forest, Junipero Serra Peak, California. Madrono 27:49-60. WAKIMOTO, R. H. 1979. Comments on seeding burned chaparral with annual ryegrass. Chaps Newsletter, p. 3-4. Chaparral Research and Development Program. Forest Fire Lab., Riverside, CA. (Received 28 May 1981; revision accepted 15 Feb 1982.) FLORISTIC AFFINITIES OF THE HIGH SIERRA NEVADA G. LEDYARD STEBBINS Department of Genetics, University of California, Davis 95616 ABSTRACT The flora of the high montane, subalpine and alpine region of the Sierra Nevada of California is believed to include four derivation categories: Old Cordilleran, Circum- boreal, Lowland Californian, and Great Basin Desert. Although these elements have mingled to the extent that species belonging to all four of them can be associated in the same community, nevertheless, they differ from each other with respect to the propor- tion of Sierran endemics that they have contributed, as well as with respect to growth habits and ecological preferences. The present review is regarded as a preliminary effort to provide a new tool for analyzing the phytogeographic, floristic, and ecological char- acteristics of the flora. With great pleasure, I dedicate this paper to Jack Major. I have obtained the familiarity with the Sierran flora that has enabled me to prepare it in large part during excursions with him and his students, accompanied by lively and absorbing discussions of the problems in- volved. His scholarly paper on a very similar topic (Major and Taylor 1977) has been a valuable source for the present survey. When an evolutionary botanist reviews the flora of a particular region, among the first questions that come to mind are: How did all the plant species get there? Did they come in from elsewhere, and if so, from how many directions? To what extent has the flora been enriched by the evolution of species in situ? What factors in the past history of the area have contributed to the present situation? Such questions can most easily be answered with respect to a region that is relatively homogenous and clearly defined. California’s Sierra Ne- vada is such a region. Four investigations of the Sierran flora have emphasized floristic affinities: those of Smiley (1921), Sharsmith (1940), Chabot and Bil- lings (1972), and Major and Taylor (1977). The last three have con- fined their attention to the alpine and subalpine zones, leaving out the high montane coniferous forests. The difficulty with this approach, as Major and Taylor recognize, is that many species highly characteristic of the subalpine and even the alpine zone descend commonly to open communities of the montane zone, so that in many parts of the Sierra the distinction between high montane, subalpine, and alpine zones is hard to make. The irregularity of timberline in these mountains, as compared with its much more regular and even occurrence in many MADRONO, Vol. 29, No. 3, pp. 189-199, 9 July 1982 190 MADRONO [Vol. 29 other mountain ranges, is discussed carefully by Major and Taylor. For these reasons, I have selected as my floristic region the parts of the Sierra designated as upper montane and subalpine by Rundel et al. (1977) plus those regarded as alpine by Major and Taylor (1977). The region included, therefore, is almost the same as that included by Smiley (1921), except that I have set its northern boundary just south of Donner Summit, north of which large interruptions exist in the subalpine flora of the Sierran crest. METHOD OF APPROACH I have made a tentative arrangement as to probable origin of all of the species of vascular plants found in the three regions: upper mon- tane, subalpine, and alpine. In the delimitation and nomenclature of species I have followed Munz (1959, 1968). The decision as to which species to include in the entire area was made for many species on the basis of personal observations. For many others, the statements of elevation in Munz (1959) were used as a guide. My experience indicates that in the northern Sierra, near Lake Tahoe, the boundary between the Lower Montane and Upper Montane Forests is about 2050 m. At the southern end of the Sierra, near Mineral King, it is about 2450 m (Smiley 1921). Consequently, I include only those species that in Munz (1959) are listed as occurring above this limit or that I have seen in the northern Sierra above 2050 m. Species that are found in the north- ern Sierra above 2600 m and that in the southern Sierra are listed as occurring above 2900 m are listed as occurring in the subalpine zone, whereas those occurring in the northern Sierra above 3150 m and in the southern Sierra above 3300 m were regarded as truly alpine. Separation of the species into four principal categories of geographic origin is more difficult but is the most logical and informative classi- fication that can be made on the basis of current knowledge. It differs in several respects from those proposed by previous authors. The ten classes recognized by Smiley (1921) are very difficult to define upon the basis of modern knowledge of the western flora and cannot be related to those recognized in the present paper. I find almost equal difficulty with the nine recognized by Sharsmith (1940). His Pacific Cordilleran plus Rocky Mountain and Sierran Cordilleran correspond roughly to Old Western Cordilleran as recognized in this paper, except that some species of the latter category are placed here in the category of “Great Basin Affinities.” His categories Boreal, Arctic-alpine and Beringian-Cordilleran correspond to Circumboreal as here recognized, whereas Widespread Mediterranean is like Lowland Californian. Great Basin is designated as such in both Sharsmith’s and the present treat- ment. The three categories recognized by Chabot and Billings (1972) are much more similar to those presented here. Both of us recognize Western Cordilleran and Widespread Arctic-alpine (Circumboreal) 1982] TABLE 1. Species Leucothoe davisiae Torr. Epilobium spp. Paeonia browni1 Dougl. Primula suffrutescens Gray Sarcodes sanguinea Torr. Horkelia tularensis (Howell) Munz Potentilla glandulosa Lindl. subsp. nevadensis (Wats.) Keck Castilleja miniata Dougl. Antennaria dimorpha (Nutt.) Tce G, Cirsium andersonii (Gray) Petrak Anelsonia eurycarpa (Gray) Macbr. & Payson Sisyrinchium idahoense Bickn. STEBBINS: FLORISTICS, SIERRA NEVADA Present classification Western Cordilleran Western Cordilleran Western Cordilleran Western Cordilleran Western Cordilleran Western Cordilleran Western Cordilleran Western Cordilleran Great Basin Great Basin Great Basin Great Basin SPECIES WHOSE GEOGRAPHIC AFFINITY IS IN DOUBT. Alternative Boreal-Holarctic Boreal-Holarctic Lowland Californian Boreal-Holarctic Boreal-Holarctic Lowland Californian Lowland Californian Lowland Californian Western Cordilleran Western Cordilleran Western Cordilleran Western Cordilleran with similar limits; their third category, Endemic Western corresponds to Lowland Californian plus Great Basin as here recognized. Major and Taylor (1977) group a few dominant species with respect to their present distribution but make no attempt to speculate about their origins. I arrived at the four categories presented below by placing a varying emphasis upon the distributional ranges of the species themselves, and those of species not found in the area under consideration that are believed to be the closest relatives of the species considered. When placing Sierran species that are endemic or have only a limited dis- tribution outside the area, emphasis was placed upon the distribution of near relatives. On the other hand, with respect to species that occur chiefly in other regions, and particularly those that enter the upper montane region from below, the distribution of the species itself was given weight equal to or greater than that of related species. The estimates of relationship are to a large extent subjective. Prin- cipal emphasis was placed upon morphological characters as outlined in taxonomic keys and descriptions. Chromosome numbers were con- sidered when available, particularly if the Sierran species was a poly- ploid having obvious diploid ancestors in other regions. The principal sources of keys and descriptions were Munz (1959), Abrams and Ferris (1923-1960), Hitchcock and Cronquist (1973), and Harrington (1954). For Carex, the standard monographs of Kukenthal (1909) and Mac- kenzie (1940) were consulted, while for Juncus, Buchenau’s (1906) monograph was essential. For information on the Circumboreal cat- egory, Hultén’s (1941-1950) Flora of Alaska was consulted, as well as those of the USSR (Komarov 1934) and Central Europe (Hegi 1936). 192 MADRONO [Vol. 29 TABLE 2. DISTRIBUTION OF SPECIES INTO CATEGORIES OF DERIVATION. Old Lowland Great Basin- Cordilleran Circumboreal California Desert Total Entire flora 342 (39%) 227 (26%) 168 (19%) 139 (16%) 876 Alpine zone only 103 (50%) 58 (28%) 15 (7%) 31 (15%) 207 In spite of all efforts, some species proved difficult to assign to one of the four categories, and their position on the lists compiled is ad- mittedly somewhat arbitrary. Some of the most puzzling are listed in Table 1. The complete lists are too long to be included in this contribution, but will be supplied upon request. RESULTS The four sources. Table 2 lists the assignment made to each of the four categories of derivation for the 876 species recognized in the high montane flora as a whole, and the 207 species found in the alpine zone, above timberline. The reasons for recognizing these four cate- gories are given below. 1. Old Cordilleran (OC). Compared with other mountain ranges in western North America, the Sierra Nevada is relatively young. Most of the uplift that elevated the range to its present height took place during the Pleistocene (Axelrod 1957, 1962; Bateman and Wahrhaftig 1966). Farther east, in the Rocky Mountain region, ranges high enough to support a subalpine and even an alpine biota have existed since the middle of the Tertiary (Axelrod 1968). Plant species adapted to these high mountains must have evolved as they were being elevated. Some species were preadapted to occupation of high montane, subalpine, and alpine zones as these zones became available following or during Sierran uplift. During the Pleistocene, when a glacial climate prevailed throughout the Sierra Nevada, the Cascade Mountains to the north must have contained much larger areas suitable to high montane, subalpine, and alpine species than they do at present. Migration south- ward along the Sierran Cascade axis is, therefore, the most probable immediate source of the Old Cordilleran species. Nevertheless, the Basin Ranges to the east most probably contributed some of these species. During the height of the Pleistocene pluvial epoch, when the valleys of the Great Basin region were filled with the waters of Lakes Bonneville, Lahontan, and others, the summit ridges of the mountain ranges that rose above the lake waters must have supported a much more mesic flora than they do now. The westernmost of these ranges, the White Mountains, is only 50 km from the Sierra, and the maximum distance between two neighboring ones of these parallel, north-south 1982] STEBBINS: FLORISTICS, SIERRA NEVADA 193 oriented ranges is only 80 km. Consequently seed transport by birds or other vectors of species having easily dispersed seeds is highly prob- able. There is no way of saying which of these two avenues of entry was the principal one. The 342 species assigned to the Old Cordilleran category either occur themselves in the Rocky Mountains or have close relatives there. The principal criterion determining their placement in this category rather than the Circumboreal group was the presence in North Amer- ica of the majority of their relatives. Examples are Hackelia, Agoseris, Antennaria, Arnica, Erigeron, some sections of Senecio, Arabis, Pol- emonium, Eriogonum subgenus Oligogonum, Lewisia, some sections of Penstemon, and Carex sect. Ovales. As expected, this is the largest of the four categories, containing 39 percent of the total 876 species, and 50 percent of the 207 species found in the alpine region (Table 2). 2. Circumboreal (C). This category includes some species that occur at low elevations in the Pacific Northwest, and a few that occur or have relatives in the eastern United States. Some of them occur or have relatives in the Rocky Mountains, but they are peripheral to and do not have centers of diversity in that region. Examples are Lonicera, Sambucus, Stellaria, various Ericaceae, Gentiana, Polygonum, Py- rola, Aconitum, Thalictrum, Amelanchier, Potentilla, Sorbus, Salix, Saxifraga, Veronica, Carex except sect. Ovales, Scirpus, Luzula, and Habenaria (sens. lat.). This is the second largest category, containing 227 species, 26 per- cent of the total, and 28 percent of the alpine species. 3. Lowland Cismontane California (LC). The 168 species that either originated in or are most closely related to species found in Lowland California constitute 19 percent of the total, but are much more poorly represented (7 percent) in the alpine zone than are any of the other categories. The great majority of them (81 percent) are found chiefly at middle or low altitudes, and enter the high montane zone, plus more rarely the subalpine zone, in warm, sunny, and dry places. They include such familiar species as Silene verecunda, Madia elegans, M. gracilis, Arctostaphylos patula, Nemophila maculata, Lotus pursht- anus, Clarkia rhomboidea, Collomia grandiflora, Linanthus ciliatus, Eriogonum nudum, Claytonia (Montia) perfoliata, Ranunculus occi- dentalis, Ribes roezlii, Orthocarpus hispidus, Solanum xantii, Viola purpurea, Brodiaea (Triteleia) hyacinthina, Danthonia californica, and Habenaria elegans. 4. Great Basin. The final category, species that have their closest affinities to the Great Basin and desert floras, contains the smallest number (139) of species, 16 percent of the total and 15 percent of those found in the alpine zone. The majority of the species occur also in the arid mid-montane zone on the east side of the Sierra, and many extend far into the pinyon-juniper belt of the Basin Ranges. Among the com- monest representatives of this group are species of Cryptantha, Arte- 194 MADRONO [Vol. 29 misia, Chrysothamnus, Machaeranthera, Descurainea, Astragalus, Phlox, Eriogonum, Ivesia, Galium, Lomatium, and annual species of Mimulus. Endemism and speciation in the High Sierra Nevada. Compared with other parts of California, the High Sierra has not been an im- portant center of speciation (Stebbins and Major 1965). Nevertheless, of the 876 species here recognized as occurring in the high montane to alpine zones, 119 or 13.5 percent are endemic to the High Sierra, while 65 percent extend beyond it only to the corresponding zones in the transverse ranges of southern California and/or the mountains of northwestern California, western Nevada and southern Oregon. These endemics are a highly heterogeneous group, belonging to 87 genera and 31 families. The great majority of them are clearly related to more widespread species that occur outside the Sierra Nevada. This condition exists for all of the 15 endemic species of Lupinus and for 9 of the 11 endemic species of Carex. The genus Carex illustrates in a striking fashion the degree to which the present flora of the High Sierra is a result of immigration rather than speciation in situ. Of the 72 species of Carex found in the High Sierra, at least 66, or 92 percent, entered the area without recognizable change, or underwent only a single event of speciation. More extensive speciation in this genus has been confined to the single large and complex section Ovales. The genus Lupinus ranks first in number of endemics. The status of described species in this genus is obscure, and several of the 15 “species” endemic to the High Sierra and neighboring ranges may deserve recognition as no higher than subspecies or varieties. Until the Sierran species of Lupinus have been studied carefully in various ways, the question remains open as to whether or not the High Sierra has been a center of speciation for this large genus. For four genera, the High Sierra appears to have been at least a minor center of speciation. Four of the six endemic species of Castilleja appear to be as closely related to each other as to species from the surrounding areas. The same could be true of the five species of Draba, although this genus needs to be studied much more carefully before the origin of its species can be deduced. With respect to two genera, Ivesia and Hackelia, the High Sierra has definitely been a center of speciation. Of the 23 species in the genus Ivesia, nine occur in the High Sierra, seven of which are en- demic. Of the remaining eight Californian species, two occur in the mountains of southern California, two in the Mohave Desert-Great Basin region to the east and south of the High Sierra, and the other four occur northward and northwest of the main Sierran axis. The High Sierra, therefore, is the center of distribution for this genus. Of considerable interest is the distribution of endemics within the 1982] STEBBINS: FLORISTICS, SIERRA NEVADA 195 TABLE 3. RELATIVE FREQUENCY OF GROWTH HABITS AMONG CATEGORIES OF DERIVATION. Great Old Circum- Lowland Basin- Cordilleran _ boreal California Desert Total Woody 21 46 | 14 92 Herbaceous perennial 312 174 104 103 693 Annual 9 7 53 22 91 342 227 168 139 876 Sierran area. The range as a whole can be divided conveniently into two parts; north and south of Yosemite National Park. The northern half has relatively few summits that are truly alpine, but its high montane and subalpine areas are more mesic, as is evident from the greater abundance of two species of conifers, Pinus monticola and especially Tsuga mertensiana. Of the 119 endemic species, 45 occur throughout the range, 20 are endemic to the northern half, and 54 are endemic to the southern half. Among those that are endemic to the Sierra plus neighboring ranges, 55 occur throughout the Sierra, 8 are confined to the northern half and 2 to the southern half. The larger number of Sierran endemics that occur only from Yo- semite southward is due partly to the greater relief and consequently greater ecological diversity of the southern area. An additional factor is, most probably, the relatively light glaciation, and in many areas complete absence of glaciation, in the high ridges and plateaus of the southern area. This permitted the survival throughout the Pleistocene of a much larger number of subalpine and alpine species than was possible in the northern area. Growth habit and derivation. With respect to growth habit, I have divided the species into three categories; woody (trees and shrubs); perennial herbs (including perennial grasses, sedges and ferns), and annuals (Table 3). Perennial herbs formed the majority (79 percent) of the total flora, whereas the percent of woody species (11 percent) is about equal to that of annuals (10 percent). Among the categories with respect to derivation, the Old Cordilleran contains the highest percentage of perennial herbs (92 percent) and the lowest of woody species and annuals. The Circumboreal category, with 20 percent, has the highest proportion of woody species, due chiefly to numerous species of Salix. The proportions of woody species among the Lowland California and Great Basin elements are similar to those in the flora as a whole. As might be expected, the species of Old Cordilleran and Circum- boreal include very few annuals (2—3 percent), whereas the Great Ba- 196 MADRONO [Vol. 29 TABLE 4. MOISTURE REQUIREMENTS AMONG CATEGORIES OF DERIVATION. Old Circumbo- Lowland Great Basin- Mesic grade Cordilleran real California Desert Total Continuously moist 63 114 21 1 199 Occasionally dry 75 71 38 9 193 Mostly dry 82 28 46 10 166 Continuously dry 127 14 63 119 318 Total 342 227 168 139 876 sin-Desert element, with 15 percent annuals and particularly the low- land California element with 32 percent annuals, have an excess. These differences are entirely to be expected. Although statistics have not been compiled, it is likely that the floras of the Lowland Cali- fornia and Great Basin Desert regions contain percentages of annuals that are not very different from those in the samples of these floras that have reached the High Sierra. Distribution of derivation categories with respect to moisture re- quirement. Among the diverse habitats found in the High Sierra, four degrees of moisture availability were recognized: (1) Moisture contin- uously available, as in wet meadows and lake and stream margins; (2) Generally moist, but with dry periods, as in the drier meadows, moist forests, and intermittent streams; (3) Generally dry, but the drought ameliorated by shade or abundant moisture during the period of snow melt; e.g., in dry forests, deep rock crevices, or rock basins; (4) Essentially xeric, such as dry rocky areas, sandy or gravelly flats, and ridge crests. The distribution into these four moisture grades of the four derivation categories is shown in Table 4. That of the species assigned to the Old Cordilleran category does not differ significantly from that of the flora as a whole. The Circumboreal species are pre- dominant in the moister habitats; those with Lowland California af- finities are more frequently adapted to the drier habitats. whereas the TABLE 5. DISTRIBUTION OF SPECIES ACCORDING TO DERIVATION IN FOUR TYPES OF COMMUNITIES IN THE VICINITY OF WRIGHT’S LAKE. OC CB LC Moist riparian forest 18 23 4 Dry forest a7 10 9 Open moraines 23 9 18 Dry rock crevices 29 8 ibe Total 97 50 48 1982] STEBBINS: FLORISTICS, SIERRA NEVADA 197 great majority of species having Great Basin affinities (86 percent) occur in the most xeric habitats. This distribution would be expected on the assumption that the great majority of the species that have entered the Sierra in relatively recent times either have not changed at all since they became part of the flora, or have undergone minimal evolution. A preliminary study of individual communities in particular habitats of the region near Wright’s Lake (2150 m) in the High Montane Zone of Eldorado County (summarized in Table 5) supports the summation presented in Table 4. Four kinds of habitats were selected: moist forests of the lake margin; dry forests; moraines in open areas having sandy or gravelly soil; and dry rock crevices. Because the number of species (6) in the Wright’s Lake area having Great Basin affinities is too small to be considered, only the other three categories of derivation were included. Table 5 presents a brief summary of the data, which will be pre- sented in full in another publication. It reinforces the data summarized in Table 4, showing that the plants of circumboreal origin are the most mesic, and those of Lowland California and Great Basin origin are the most xeric. In the two drier habitats, species assigned to all four of the categories of derivation, including one or two having Great Basin affinities, existed side by side in the same plant community. At least in the High Sierra, modern plant communities contain elements that are derived from several different sources. GENERAL DISCUSSION AND CONCLUSIONS This contribution must be regarded as a preliminary rather than a final analysis. Nevertheless, I doubt that further studies will affect the general conclusions reached here. These are as follows: (1) The largest number of species has evolved in association with mountain building and other effects (e.g., increasing climatic drought, particularly in summer) that have occurred in the western Cordillera during the Tertiary and Quaternary Periods. (2) During the Pleistocene and recent epochs, the Sierra has been in- vaded by plants from three very different sources: (a) Montane and moist forested regions to the northward; (b) Foothills and lower mon- tane regions to the west of the range; (c) Arid steppes and desert mountains to the east and southeast of the range. Some of these species of disparate origin have become closely associated with each other in the same plant community. (3) Nevertheless, most of these relatively recent immigrants have ad- justed themselves to Sierran habitats with relatively little evolutionary change. Exceptions to this rule, in the form of endemic species that are related to elements other than the Old Cordilleran, exist only in a few genera, such as Jvesia and Astragalus. 198 MADRONO [Vol. 29 Using the four derivation categories as a phytogeographic, floristic, and ecological foundation, and following more careful investigations of the species concerned, a number of problems of general interest might be attacked. Among them are: (1) In the colonization of a recent, relatively isolated mountain range, what are the relative roles of long distance dispersal and more contin- uous migration under the influence of different past climates? (2) Are recent immigrants, such as LC and GB, likely to be more abundant, less abundant or about equally as abundant as older in- habitants? Are such differences correlated with growth habit or any other characteristics of the plants themselves? (3) Are the more recent immigrants more or less randomly distributed according to habitat, or do they tend to occupy habitats that on the basis of geological evidence seem to have been formed recently? LITERATURE CITED ABRAMS, L. and R. S. FERRIS. 1923-1960. An Illustrated Flora of the Pacific States, vols. 1-4. Stanford Univ. Press, Stanford, CA. AXELROD, D. I. 1957. Late Tertiary floras and the Sierra Nevadan uplift. Bull. Geol. Soc. Amer. 68:19-45. . 1962. Post-Pliocene uplift of the Sierra Nevada, California. Bull. Geol. Soc. Amer. 73:183-198. . 1968. Tertiary floras and topographic history of the Snake River Basin, Idaho. Bull. Geol. Soc. Amer. 79:713—733. BATEMAN, P. C. and C. WAHRHAFTIG. 1966. Geology of the Sierra Nevada. In E. H. Bailey, ed., Geology of northern California, p. 2. Calif. Div. Mines Geol., Bull. 170: BUCHENAU, F. 1906. Juncaceae. Jn A. Engler, ed., Das Pflanzenreich. IV. 36. W. Engelmann, Leipzig. CHABOT, B. F. and W. D. BILLINGS. 1972. Origins and ecology of the Sierran alpine flora and vegetation. Ecol. Monogr. 42:163-199. HARRINGTON, H. D. 1954. Manual of the Plants of Colorado. Sage Books, Denver. HEGI, G. 1936. Illustrierte Flora von Mittel-Europa. C. Hanser. Munchen, Germany. Hitcucock, C. L. and A. CRONQUIST. 1973. Flora of the Pacific Northwest, Univ. Washington Press, Seattle. HULTEN, E. 1941-1950. Flora of Alaska and Yukon. Vols. 1-10. Gleerup, Lund, Sweden. KoMarovy, V. L. (ed.) 1934. Flora USSR. Akademiia Nauk, Leningrad. KUKENTHAL, G. 1909. Cyperaceae-Caricoideae. In A. Engler, ed., Das Pflanzen- reich. IV. 20 (Heft 38) W. Engelmann, Leipzig. MACKENZIE, K. K. 1940. North American Cariceae. New York Botanical Garden, New York. Major, J. and D. W. Taytor. 1977. Alpine. In M. G. Barbour and J. Major, eds., Terrestrial vegetation of California, p. 601-675. Wiley-Interscience, NY. Munz, P. A. 1959. A California Flora. Univ. California Press, Berkeley. . 1968. Supplement to A California Flora. Univ. California Press, Berkeley. RUNDEL, P. W., J. D. PARSoNs, and D. T. GorRDON. 1977. Montane and subalpine vegetation of the Sierra Nevada and Cascade Ranges. In M. G. Barbour and J. Major, eds., Terrestrial vegetation of California, p. 559-599. Wiley-Interscience, NY. 1982] STEBBINS: FLORISTICS, SIERRA NEVADA 199 SHARSMITH, C. W. 1940. A contribution to the history of the alpine flora of the Sierra Nevada. Ph.D. dissertation, Univ. California, Berkeley. SMILEY, F. J. 1921. A report on the boreal flora of the Sierra Nevada of California. Univ. Calif. Publ. Bot. 9:1-425. STEBBINS, G. L. and J. Mayor. 1965. Endemism and speciation in the California flora. Ecol. Monogr. 35:1-35. (Received 24 Aug 1981; revision accepted 7 Dec 1981.) A GRADIENT PERSPECTIVE ON THE VEGETATION OF SEQUOIA NATIONAL PARK, CALIFORNIA JOHN L. VANKAT Department of Botany, Miami University, Oxford, OH 45056 ABSTRACT The results of detrended correspondence analysis indicate that elevation is well cor- related with vegetation differences and that topographic-moisture conditions are also important. Gradients of these two environmental factors are used as axes for plotting the distribution of 15 general vegetation types. Superimposed on the resulting vegetation chart are isolines of tree density, basal area, tree cover, woody plant richness, and woody plant diversity (1/C). It is concluded that the vegetation of the park is better considered as a series of continua rather than as discrete units. This perspective produces suggestions for future research and supports expanded use of gradient analysis tech- niques in studies of the vegetation of the Sierra Nevada. Surprisingly little has been published on Sierra Nevada vegetation despite its economic and ecologic importance. Nearly all vegetation studies, as well as other ecological works, have treated the vegetation as divisible into discrete units and therefore classifiable into general categories such as those proposed by Munz (1959) for all of California. Because classification systems inherently emphasize discontinuities in vegetation, even detailed systems do not accurately represent field conditions where vegetation continua occur. Gradient analysis tech- niques can be used to illustrate vegetation continua (and discontinua). Thus, examination of vegetation data by gradient analysis can com- plement the classification approach (Mueller-Dombois and Ellenberg 1974, Whittaker 1975). Because gradient analysis methodology seldom has been applied to data on Sierran vegetation, the purpose of this study is to develop a gradient perspective on the vegetation of Sequoia National Park in the southern Sierra Nevada. METHODS I utilized two kinds of data on woody species: (1) presence-absence data compiled from surveys of 108 stands of variable size and (2) cover, density, and basal area data (with an importance percentage calculated by summing relative values and dividing by 3) obtained from 92 belt transects (each 2 X 50 m). These data were collected in 1969 as part of a study on vegetation changes within the park (Vankat 1970). The stand surveys and belt transects were located in units of MADRONO, Vol. 29, No. 3, pp. 200-214, 9 July 1982 1982] VANKAT: VEGETATION OF SEQUOIA NATIONAL PARK 201 relatively homogeneous vegetation and were scattered throughout much of the park in order to represent the variation within general vegetation types. Elevation, slope aspect and inclination, and recent land-use history data are available for the sites sampled. Vankat and Major (1978) presented other details about the field sampling meth- odology. No samples of young seral stages were used in the present study. I analyzed separately the data from the stand surveys and belt tran- sects, using the detrended correspondence analysis (DCA) program DECORANA (Hill 1979). Because this type of indirect gradient anal- ysis is not subject to some of the intersample-distance distortions that occur with other ordination methods, it is considered the best such method of ordinating vegetation data (Hill and Gauch 1980). The results of these analyses led me to arrange the stand survey and belt transect samples on graphs that had elevation and topographic-mois- ture gradients as axes. Elevation within the park is from 390 to 4419 m. I defined a 10-unit topographic-moisture gradient from moist to dry. I considered sites near streams to be at the moist extreme; sites with slopes of northeast, north, northwest, east, west, southeast, south, and southwest exposures to be at intermediate and progressively drier positions; and sites on ridgetops to be at the dry extreme. I equated arbitrarily one unit with 10° of slope. Therefore, a sample collected on an 18°, southeast-facing slope would be placed at the ninth unit of the topographic-moisture gradient. To increase the data base for my examination of the distribution of vegetation along the two gradients, I collected data from an unpub- lished vegetation map (circa 1939; available at the park headquarters). In addition to illustrating the locations of stands of various vegetation units, the map also shows contour lines and a grid of land-survey lines. I used the 12 major land-survey lines as though they were line transects and recorded line-intercept data. Specifically, for each stand of vegetation crossed by these lines, I noted the vegetation type, slope aspect and inclination, and maximum and minimum elevations along the line. This procedure provided 379 vegetation samples that I plotted on the same type of two-dimensional graph described above. The resultant graph, along with others produced with stand survey and belt transect data, provided the basis for a vegetation chart illustrating the distribution of the 15 general vegetation types listed in Table 1. Each division of the chart included the majority but not all of the samples of that vegetation type. I used belt transect data to calculate tree density (individuals >1 year of age/100 m’), basal area (m*/ha), tree cover (percent), woody plant richness (number of species/100 m?’), and diversity (1/C = the number of equally important species required to give the same Simp- son’s Index observed; after Peet 1974; C = > p,;”, where p; = impor- tance percentage of the zth species). Isolines for these parameters were 202 MADRONO [Vol. 29 TABLE 1. WoOoDy SPECIES CHARACTERISTIC OF THE 15 VEGETATION TYPES SHOWN ON THE VEGETATION CHART (FIG. 1). BLUE = Blue Oak Woodland, CHAM = Cha- mise Chaparral, MIXE = Mixed Chaparral, LOWL = Lowland Live Oak Woodland, UPLA = Upland Live Oak Woodland, BLAC = Black Oak Woodland, POND = Ponderosa Pine Forest, WHIT = White Fir Forest, JEFF = Jeffrey Pine Forest, JUNI = Juniper Woodland, REDF = Red Fir Forest, LODG = Lodgepole Pine Forest, SUBA = Subalpine Forest, MEAD = Meadow, and ALPI = Alpine. Vegetation type 2>Pma mo rPrud aM MM MOP OD S407 MO tx a elOls rman Urme My it > Species Quercus douglasit Rhamnus crocea Aesculus californica Quercus wislizenit Arctostaphylos viscida Adenostoma fasciculatum Ceanothus cuneatus Dendromecon rigida Eriodictyon californicum Ceanothus velutinus Fraxinus dipetela Fremontia californica Lonicera interrupta Rhus trilobata var. malacophylla Cercis occidentalis Cercocarpus betuloides Rhus diversiloba Quercus chrysolepis Umbellularia californica Quercus kelloggii Chamaebatia foliolosa Pinus ponderosa Pinus lambertiana Abies concolor Calocedrus decurrens Rosa spithamea Sequotadendron giganteum Arctostaphylos patula Pinus jeffreyi Artemisia tridentata Cercocarpus ledifolius Juniperus occidentalis Pinus monophylla Abies magnifica Pinus contorta Ribes montigenum Pinus albicaulis Pinus balfouriana Pinus monticola Number of stand surveys 4 9. 16. 7 SO 32 11 SONA BO 210) VOTH BOW) belt transects 14 0 0 2 0 3B TT 30, 55. 70. 105-9. (30 FO map interceptions 13 30 9 -45—- -22— 40 12 4 39 52 27 24 25 mA pA Md rd od fOr aaa KM MaMa nA ana MMMM MM mM rm mes MMM mmr MM mrs rr ms rm ~~ nanan mx a mr x 1982] VANKAT: VEGETATION OF SEQUOIA NATIONAL PARK 203 superimposed on the vegetation chart. Because some of the vegetation types were represented poorly by belt transects, I used estimates based on stand surveys and personal field experience and, in a few cases, data of other researchers (Parsons 1976, Baker et al. 1981) to orient isolines. Homologous vegetation charts with isolines of vegetation parame- ters have been produced for other mountainous regions by Whittaker (1956, 1960), Whittaker and Niering (1965), and Peet (1978). RESULTS AND DISCUSSION Detrended correspondence analysis and the vegetation chart. DCA applied separately to stand survey and belt transect data produced ordination values along first axes that correlated strongly with eleva- tion. For example, for the ordination based on importance percentage data from the belt transects (excluding Subalpine Forest transects be- cause they are unique floristically), the correlation coefficient relating ordination value and elevation was —0.96822 (p = 0.0001). DCA pro- duced little separation of samples along other axes, but where samples of different vegetation types had similar first axis ordination values they usually were from different topographic positions. Therefore, I plotted samples according to elevation and topographic-moisture con- ditions in order to produce the graphs upon which the vegetation chart is based. Figure 1 is the vegetation chart. Most of the 15 categories of vege- tation shown on the chart are widespread and nearly all are readily recognizable in the field. Table 1 lists characteristic species; greater detail has been provided by Vankat (1970), Vankat and Major (1978), Rundel et al. (1977), or other chapters of Barbour and Major (1977). Construction of the chart required simplification of the graphs of sam- ples, because different vegetation types did not occupy discrete areas on the graphs but overlapped greatly. Consequently, the chart must be considered diagrammatic. Vegetation gradients. I interpret the following as indicating that the vegetation of the park is better considered as a series of continua rather than as discrete units: first, samples of different vegetation types over- lap in the DCA ordinations as discussed above; second, many species are characteristic of more than one vegetation type as shown in Table 1 and by Vankat and Major (1978); and third, samples of the same vegetation type vary considerably as shown in field data presented by Vankat (1970). The spacing of dashes in Fig. 1 illustrates diagram- matically the breadth of continua between adjacent vegetation types. In addition, some aspects of the continua are illustrated by plots of parameter isolines on the vegetation chart (Figs. 2, 3). The chart itself and the isoline plots provide good context for discussion of the vege- tation of the park from a gradient perspective. 204 MADRONO [Vol. 29 el li 4,495 12,000 3500 AIPM er ee ae ~---— Subalpine Forest 3 30004 8 | ac 10,000 s Lodgepole Pine Forest ma ae a, OM Se a = 2500 rs: a 1s 0 2 1 38-8000 ® Zz = © 2000 O = 6000 - < Sg ff 15008 23-2357" a = lw a 1 = sieceork (MdO00 xs Va Chami " Woodland 1000 é 8 e eRe ae ae 2000 390 1280 moist dry TOPOGRAPHIC-MOISTURE GRADIENT Fic. 1. Diagrammatic chart illustrating the distribution of general vegetation types according to elevation and topographic-moisture gradients for Sequoia National Park, California. The spacing of dashes reflects the breadth of continua between adjacent vegetation types. The lowest elevational band in the park is occupied by Lowland Live Oak Woodland in riparian habitats and Blue Oak Woodland elsewhere. Increasing moisture availability is accompanied by sub- stantial and, in some cases, relatively abrupt increases in basal area, tree cover, richness, and diversity. These changes are paralleled by increasing relative importance of both Aesculus californica and Quer- cus wislizenii and decreasing importance of Q. douglasiiz. Some of these same shifts in composition and structure have also occurred during succession in Blue Oak Woodland stands on mesic sites not subject to periodic burning (Vankat and Major 1978). In general throughout the park, vegetation changes initiated by twentieth-cen- tury fire control and nineteenth-century livestock grazing add another dimension to interpreting vegetation gradients. However, as in this example, these changes frequently result in increasingly mesic condi- VANKAT: VEGETATION OF SEQUOIA NATIONAL PARK 205 1982] (183}) NOILWA313 ‘Ju9dI9d—1AIAOD 9aI} pue ‘aieyay Jod ,W—eale [eseq ‘,W OOT Jad a3e JO IeIA T< STeNnplAIpul—A}Isuap 991} aie S}IUAQ ‘WeYyd uoTe}IZ9A 9y} UO pasoduiliedns siajaueIed ainjon4}s UOT}LJIBIA JO SIUT[OS] “7 ‘DIY Y3SAOO0 33duL vauv 1VvSVa ALISNAG 3381 INJIOVWHD SYNLSIOW-DIHdVHYDOdOL ILNJIOVYS SYNLSIOW-DIHdVEYODOdOL INIIDWHDS SYNLSIOW-DIHdVEYDOdOL Aip ys!OW Aip ys!OlW Ap yS!IOW O8cl O6E S$ +006 OOS 1 00072 (si919W) NOILVAI13 206 MADRONO [Vol. 29 ELEVATION (meters) ELEVATION (feet) TOPOGRAPHIC-MOISTURE GRADIENT TOPOGRAPHIC-MOISTURE GRADIENT RICHNESS DIVERSITY (1/C) Fic. 3. Isolines of vegetation composition parameters superimposed on the vege- tation chart. Units are richness—number of woody species per 100 m? and diversity (1/ C)}—number of equally important species required to give the same Simpson’s Index observed or estimated (after Peet 1974). tions and may parallel a shift in that direction along the topographic- moisture gradient. Vankat and Major (1978) described successional changes for most of the vegetation types considered here. With increases in elevation at the moist end of the topographic- moisture gradient (i.e., in riparian habitats), there is a gradual tran- sition in the vegetation from Lowland to Upland Live Oak Woodland. The most obvious vegetation changes include decreases in richness and diversity as Quercus wislizenii and Aesculus californica are re- placed by Q. chrysolepis and as the shrub layer decreases in impor- tance. At 1200 to 1500 m, Upland Live Oak Woodland extends across the topographic-moisture gradient toward drier conditions. Basal area and tree cover decrease along this gradient, especially in sites where the steepness of canyon walls retards soil development. The transition with the Ponderosa Pine Forest appears abrupt in many areas, but no quantitative data are available. With increasing elevation in the midrange of the topographic-mois- ture gradient, Blue Oak Woodland is replaced relatively abruptly by chaparral vegetation. Although Chamise Chaparral is thought to occur under drier conditions and Mixed Chaparral under the presumably more mesic conditions of north-facing slopes, graphs of samples of these two vegetation types overlapped greatly. This may in part have resulted from inaccurate prediction of soil moisture conditions on the basis of topographic position, because chaparral vegetation has been 1982] VANKAT: VEGETATION OF SEQUOIA NATIONAL PARK 207 reported (for east of San Diego, CA) as having greater soil moisture on south-facing than on north-facing slopes (Ng 1974, see also Krause and Kummerow 1977). Regardless, gradients in vegetation structure between the two chaparral types are steep because of the lack of trees in the former and the relatively high importance of shrubby trees in the latter. Even steeper gradients exist in terms of composition. The Mixed Chaparral is characterized by some of the higher richness and diversity values of any vegetation type in the park, and Chamise Chaparral has relatively low values. Once more, a successional gra- dient is important in this portion of the vegetation chart. Parsons (1976) documented a decrease in species diversity and increases in total woody cover and dominance of Adenostoma fasciculatum in Chamise Chaparral with increased time since burning. The distribution of Montane Chaparral is not shown on the vege- tation chart, because stands are often successional in a coniferous forest sere (Hanes 1977) and because stands are interspersed through- out the chart area of all low- and mid-elevation coniferous forest types. The stands are, however, concentrated in the drier portion of the range of the Ponderosa Pine Forest. The line that separates Mixed Chaparral and Lowland Live Oak Woodland on the vegetation chart was positioned largely arbitrarily because the two vegetation types intergrade. Baker et al. (1981) com- bined the two types, along with Upland Live Oak Woodland, as a “mixed-evergreen woodland.” Although such terminology usefully unites continuously intergrading vegetation, it does not acknowledge the dominance of shrubs and shrubby trees in Mixed Chaparral as contrasted with the greater size and cover of trees in the Live Oak Woodlands, nor does it acknowledge floristic differences between the Live Oak Woodlands. Griffin (1977) recognized close floristic ties be- tween mesic chaparral and a Quercus wislizenii-dominated live oak woodland in the Sierran foothills, but also acknowledged the arbores- cent growth form of the oaks in the latter. Sawyer et al. (1977) re- viewed the mixed hardwood forests of California and described dry ridges with pure stands of Quercus chrysolepis, apparently similar to the Upland Live Oak Woodland described by Vankat and Major (1978). Therefore, I have continued to use the terms Mixed Chaparral, Lowland Live Oak Woodland, and Upland Live Oak Woodland for general vegetation types that are recognizable portions of vegetation continua. Perhaps an equivalent name for the Mixed Chaparral is Hane’s (1977) “woodland chaparral” (following Horton 1960, for the San Bernardino Mountains of southern California). At the drier end of the topographic-moisture gradient, increased elevation is accompanied by greater importance of the tree layer and by restriction of Blue Oak Woodland vegetation to the dry (ridgetop) end of the gradient. Quercus douglasii decreases in relative importance as Aesculus californica and especially Q. kelloggii increase. There are 208 MADRONO [Vol. 29 increases in richness and diversity as such species appear and, in the case of Q. kelloggii, become increasingly important. Correlated with this elevation gradient is a disturbance gradient; stands sampled at lower elevations were either burned more recently than higher stands or were grazed in the winter by National Park Service pack animals. This may reflect proximity to the park highway rather than biological or environmental differences. Higher elevation woodland stands are dominated by Quercus kel- loggii. These Black Oak Woodland stands are restricted largely to ridges; however, some ridges in this elevational range have Chamise Chaparral vegetation. The Black Oak Woodland intergrades with the Blue Oak Woodland at lower elevations and with the Quercus kelloggit dominated stands of Ponderosa Pine Forest at higher elevations, but differs from the former in the dominant species and the latter in phys- lognomy. Griffin (1977) did not recognize a Black Oak Woodland, but Baker et al. (1981) considered it a phase of foothill woodland vege- tation. The transition between Black Oak Woodland and Ponderosa Pine Forest involves some of the more abrupt shifts in structure and com- position shown by the isoline plots. The change in physiognomy from woodland to forest is reflected in increases in tree density and cover. The parameters reach very high values in the Ponderosa Pine Forest where fire protection during much of this century has resulted in a very dense understory of small trees (Vankat and Major 1978). How- ever, large trees are scattered and the increase in basal area from the Black Oak Woodland to the Ponderosa Pine Forest is less dramatic. The sparseness of large trees also may be related to the high values of richness and diversity. Whittaker (1975) indicated that environ- mental heterogeneity produced by open canopies tended to result in high plant diversities, especially in mid- to dry portions of moisture gradients. Despite relatively abrupt changes between Black Oak Woodland and Ponderosa Pine Forest, floristic gradients can be recognized. Bak- er et al. (1981) denoted a black oak forest community on north-facing slopes above 1200 m. Vankat and Major (1978) described a variety of plant communities along a xeric to mesic gradient involving both el- evation and topographic position: under xeric conditions communities were dominated by mature Quercus kelloggit with scattered mature Pinus ponderosa; more mesic sites had mature individuals of both species combined with scattered mature Calocedrus decurrens and an understory of C. decurrens and P. ponderosa; and still more mesic sites had all three species as codominants with an understory of Abies concolor and C. decurrens. Both of these understory species had in- creased in density where stands were no longer burned periodically. With increasing elevation and moisture availability, tree density, 1982] VANKAT: VEGETATION OF SEQUOIA NATIONAL PARK 209 tree cover, richness, and diversity peak in the Ponderosa Pine Forest and decline along a gradient into the White Fir Forest. Basal area increases along this gradient as average tree size becomes larger. Ac- companying floristic changes include decreases in relative importance of Quercus kelloggit, Pinus ponderosa, and Chamaebatia foliolosa, as well as shrub cover in general. Increases in importance are shown by Abies concolor, Calocedrus decurrens, and Pinus lambertiana. Sellers (1970) reported similar findings along a moisture gradient near the Kings River in Kings Canyon National Park (which is adjacent to the northern boundary of Sequoia National Park). The intergradation of the Ponderosa Pine and White Fir Forests makes it difficult to differ- entiate between them in the mid-range of the elevation and topograph- ic-moisture gradients. In fact, some authorities have combined the two as “yellow pine forest” (e.g., Munz 1959); however, at least for the park region it is better to consider them as two general forest types separated by a broad vegetation continuum. The elimination of peri- odic burning has contributed to this continuum as Abies concolor has increased in density in high-elevation Ponderosa Pine Forest stands (Vankat and Major 1978). Toward the dry end of the topographic-moisture gradient, the Pon- derosa Pine Forest is replaced at higher elevations by Juniper Wood- land that is dominated by Juniperus occidentalis with some Pinus monophylla and P. jeffreyi. Although quantitative data are lacking, I hypothesize that this elevation gradient is paralleled by decreases in the vegetation parameters illustrated. Under somewhat less xeric conditions the Ponderosa Pine Forest is replaced at higher elevations by Jeffrey Pine Forest. DCA produced an array of belt transects from the Jeffrey Pine Forest that correlated well with the topographic-moisture gradient. At approximately 2200 m, this gradient involves gradual replacement of Pinus jeffreyi by Abies concolor, Calocedrus decurrens, and, near riparian areas, Pop- ulus trichocarpa. There are accompanying increases in basal area and tree cover. The White Fir Forest occurs in the mesic half of the topographic- moisture gradient at mid-elevations. I define the forest as including the geographically scattered groves of Sequoiadendron giganteum; on the vegetation chart these are concentrated in the mid- to upper mid- elevations of the more mesic half of the area indicated as White Fir Forest. Sites of S. giganteum groves may be even more mesic than indicated by position on the topographic-moisture gradient. Rundel (1972) hypothesized that precipitation at elevations above the largest grove in the park (Giant Forest) resulted in subsurface flow of water into the grove where favorable soil moisture levels were maintained throughout dry summers. Tree density and basal area parameters reach maximum or at least high values in the stands of White Fir 210 MADRONO [Vol. 29 Forest with S. giganteuwm. In contrast, tree cover and diversity reach maximum values in stands that intergrade with the Ponderosa Pine Forest. The elevation gradient within the White Fir Forest is paralleled by an increase in basal area and decreases in density, tree cover, richness, and diversity. Floristic changes along this gradient include an increase in the relative importance of Abies concolor (except in high elevation stands where A. magnifica becomes important), a decrease in Pinus lambertiana, and the sequential losses of Quercus chrysolepis, Corylus cornuta, Cornus nuttallit, Quercus kelloggit, Calocedrus decurrens, and Sequoiadendron giganteum. Most of these trends also are appar- ent when stands with S. giganteum are analyzed separately. The transition from the White Fir Forest to the Red Fir Forest is marked by decreases in richness and diversity and a switch in the relative importances of Abies concolor and A. magnifica. The Red Fir Forest is relatively homogeneous. Except for a peak in basal area and a decrease in tree cover, there are no significant vegetation gradients correlated with increasing elevation until Pinus contorta becomes im- portant at the upper elevational limit of the forest. The presence of P. contorta results in higher richness and diversity values along a relatively narrow ecotone between the Red Fir and Lodgepole Pine Forests. Tree density is high in Lodgepole Pine Forest stands on moist sites including edges of meadows where encroachment by Pinus contorta has been occurring since early in this century (Vankat and Major 1978). This and other structure parameters show decreases as elevation increases through the Lodgepole Pine Forest to the Subalpine Forest to treeline. Richness and diversity values also decrease to treeline, except for an increase in richness values along the Lodgepole Forest- Subalpine Forest ecotone. Floristic changes with increasing elevation include the loss of Abies magnifica, followed by a shift in dominance from P. contorta to subalpine species, especially P. balfouriana and, in some sites, P. albicaulis and P. monticola. Alpine vegetation occurs above timberline. Its stands are dominated by herbs, as are stands of Meadow vegetation. The latter are found in wet habitats above 2000 m; however, many wet sites are forested and this is indicated by dashed isolines across the Meadow region of the vegetation chart. Comparison with previously published figures. Baker et al. (1981) plotted the distribution of foothill vegetation types of the park on a polar diagram of slope aspect with elevation. Their diagram may be compared to Fig. 1, although my topographic-moisture gradient is not based solely on slope aspect. The “mixed-evergreen woodland” of Bak- er et al. (1981) combines the Mixed Chaparral and both Live Oak Woodlands of Vankat and Major (1978). Its distribution corresponds closely with that of the Mixed Chaparral and Lowland Live Oak 1982] VANKAT: VEGETATION OF SEQUOIA NATIONAL PARK 2alat Woodland shown in Fig. 1; however, Baker et al. (1981) indicated a “black oak forest” at the position of Upland Live Oak Woodland in Fig. 1. Another difference is that they portrayed foothill woodland vegetation as extending farthest upslope on southeast and east aspects, i.e., the mid- to mid-dry range of my topographic-moisture gradient. My data indicate maximum elevations for Blue Oak Woodland at the dry portion of the gradient, 1.e., on south and southwest exposures and ridges (they acknowledged the ridge sites in their text). Rundel et al. (1977) presented a schematic vegetation chart based on their collective field experience. The chart had elevation and mois- ture gradients for axes and illustrated Sierran coniferous forests around 37°N latitude, which is approximately 50 km north of the park bound- ary. Their depiction of the distribution of lower elevation coniferous forests, 1.e., the Ponderosa Pine and White Fir Forests, is similar to that shown in Fig. 1, but the two vegetation charts differ greatly and increasingly in portraying the distribution of vegetation at higher el- evations. Both charts show the Red Fir Forest limited to the more mesic portion of the moisture gradient, but Rundel et al. (1977) extended its upper elevational limit to 3000 m near the middle of their moisture gradient. They also depicted the Lodgepole Pine Forest at a higher elevation than shown in Fig. 1 and as covering substantial elevational ranges toward both the mesic and xeric ends of their moisture gradient. They based this moisture gradient relationship on data collected in the Rocky Mountains, but data examined in my study do not support that extrapolation. Stands of Lodgepole Pine Forest in the park are most common in relatively mesic sites such as on plateau-like areas where they are near meadows or on gentle slopes somewhat above meadows and/or streams. Rundel et al. (1977) also indicated a band of foxtail pine (subalpine) forest decreasing regularly in elevation from an approximately 3650- 3850 m band on the mesic extreme of the moisture gradient to ap- proximately 3200-3500 m on the xeric extreme. This band falls well within the zone indicated for Alpine vegetation in all but the drier portion of the topographic-moisture gradient on my vegetation chart. Also, they indicated that the upper elevational limit of their foxtail pine forest decreased steadily with decreasing moisture conditions; however, my data indicate that it is better to represent treeline as increasing in elevation along a topographic-moisture gradient, as for example from a north-facing slope to west- and south-facing slopes. Treeline, however, does decrease in elevation in very low moisture situations such as along ridgetops. CONCLUSIONS This study is the first to apply a gradient analysis perspective to the range of vegetation found over a large area of the Sierra Nevada. The DAG MADRONO [Vol. 29 results indicate that elevation is the primary environmental factor as- sociated with vegetation differences and that topographic-moisture conditions are also important. Some relatively steep vegetation gra- dients occur in the park, but the vegetation is better considered as a series of continua, rather than as discrete units. Vegetation charts of the type produced in this study have been useful tools for portraying the environmental relations of general vegetation types (Whittaker 1975). The schematic chart for Sierran coniferous forests published by Rundel et al. (1977) is similar to that presented in this study in regard to lower elevation forests, but differs sharply for higher elevation forests. Comparison of my vegetation chart with the foothill vegetation-elevation-slope exposure diagram published by Baker et al. (1981) also shows similarities and differences. Unfortu- nately, the nature of their diagram precludes combining it with the chart of Rundel et al. (1977) so that vegetation transitions between foothill and montane zones could be compared to the transitions shown in Fig. 1. Vegetation charts are also useful for illustrating gradients in vege- tation parameters. The data presented and discussed in the previous section show that tree density is greatest in low elevation stands of the Ponderosa Pine Forest, basal area is greatest in the Red Fir Forest and White Fir Forest stands with Sequoiadendron giganteum, and tree cover is greatest in low elevation stands of both the Ponderosa Pine and White Fir Forests. Woody plant richness peaks in the Mixed Chaparral and low elevation stands of both Ponderosa Pine Forest and White Fir Forest stands with S. giganteum, and diversity reaches maximum values in low elevation stands of both the Ponderosa Pine and White Fir Forests. Of the vegetation types listed in the previous paragraph, only the White Fir Forest stands with S. giganteum have been studied exten- sively in the park or nearby areas (see references in Vankat and Major 1978, Parsons and King 1980). My findings indicate that other subjects of potentially profitable research include the Lowland Live Oak Woodland, Mixed Chaparral, Red Fir Forest, low elevation stands of both the Ponderosa Pine and White Fir Forests, and transitions be- tween these two forests and between mid-elevation forests and low- elevation woodlands and shrublands. In addition, greater application of gradient analysis techniques in investigations of Sierran vegetation is merited. The use of a gradient based on direct measurement of soil- plant moisture conditions is needed, because relationships between topographic position and moisture conditions cannot always be pre- dicted. Also, studies using a larger number of vegetation samples and covering broader geographic ranges are desirable, and attempts to complement environmental gradients with succession gradients may be rewarding. 1982] VANKAT: VEGETATION OF SEQUOIA NATIONAL PARK 213 ACKNOWLEDGMENTS During 1969 when the original field data were collected, Jack Major was my graduate program adviser, the U.S. National Park Service provided financial support and other assistance, and Sigma Xi furnished supplementary financial aid. During 1981, when this study was carried out, J. A. Gordon assisted with many of the data analyses; Miami University through Dean C. K. Williamson paid for my travel to the symposium hon- oring Jack Major; and G. A. Baker, S. G. Conrad, J. R. Griffin, and D. J. Parsons provided many helpful suggestions. I am indebted to all of the above individuals and organizations. LITERATURE CITED BAKER, G. A., P. W. RUNDEL, and D. J. PARSONS. 1981. Ecological relationships of Quercus douglasit (Fagaceae) in the foothill zone of Sequoia National Park, Cali- fornia. Madrono 28:1-12. BARBOUR, M. G. and J. Major, eds. 1977. Terrestrial vegetation of California. Wiley- Interscience, NY. GRIFFIN, J. R. 1977. Oak woodland. In M. G. Barbour and J. Major, eds., Terrestrial vegetation of California, p. 383-415. Wiley-Interscience, NY. Hanes, T. L. 1977. California chaparral. In M. G. Barbour and J. Major, eds., Terrestrial vegetation of California, p. 417-469. Wiley-Interscience, NY. HILL, M. O. 1979. DECORANA, a FORTRAN program for detrended correspon- dence analysis and reciprocal averaging. Cornell Univ., Ithaca, NY. and H. G. GAucH, JR. 1980. Detrended correspondence analysis: an improved ordination technique. Vegetatio 42:47-58. HorTON, J. S. 1960. Vegetation types of the San Bernardino Mountains. USDA For. Serv., Pacific Southw. For. Range Exp. Sta., Techn. Paper PSW-44. KRAUSE, D. and J. KUMMEROW. 1977. Xeromorphic structure and soil moisture in the chaparral. Oecol. Plant. 12:133-148. MUELLER-DoMBOISs, D. and H. ELLENBERG. 1974. Aims and methods of vegetation ecology. Wiley, NY. Munz, P. A. 1959. A California flora. Univ. California Press, Berkeley. No, E. 1974. Soil moisture relations in chaparral. M.S. thesis, San Diego State Univ., CA. Parsons, D. J. 1976. The role of fire in natural communities: an example from the southern Sierra Nevada, California. Environ. Conserv. 3:91—99. and V. A. KING. 1980. Scientific research in Sequoia and Kings Canyon Na- tional Parks: an annotated bibliography. Sequoia Nat. Hist. Assoc., Three Rivers, CA. PEET, R. K. 1974. The measurement of species diversity. Annual Rev. Ecol. Syst. 5:285-307. . 1978. Forest vegetation of the Colorado Front Range: patterns of species di- versity. Vegetatio 37:65—78. RUNDEL, P. W. 1972. Habitat restriction in giant sequoia: the environmental control of grove boundaries. Amer. Midl. Naturalist 87:81—99. , D. J. PARSONS, and D. T. GorRDON. 1977. Montane and subalpine vegetation of the Sierra Nevada and Cascade Ranges. Jn M. G. Barbour and J. Major, eds., Terrestrial vegetation of California, p. 559-599. Wiley-Interscience, NY. SAWYER, J. O., D. A. THORNBURGH, and J. R. GRIFFIN. 1977. Mixed evergreen forest. In M. G. Barbour and J. Major, eds., Terrestrial vegetation of California, p. 359-381. Wiley-Interscience, NY. SELLERS, J. A. 1970. Mixed conifer forest ecology: a qualitative study in Kings Canyon National Park, Fresno County, California. M.A. thesis, Fresno State College, CA. VANKAT, J. L. 1970. Vegetation change in Sequoia National Park, California. Ph.D. dissertation, Univ. California, Davis. 214 MADRONO [Vol. 29 and J. Major. 1978. Vegetation changes in Sequoia National Park, California. J. Biogeography 5:377-402. WHITTAKER, R. H. 1956. Vegetation of the Great Smoky Mountains. Ecol. Monogr. 26: 1-80. 1960. Vegetation of the Siskiyou Mountains, Oregon and California. Ecol. Monogr. 30:279-338. . 1975. Communities and ecosystems. 2nd ed. Macmillan, NY. and W. A. NIERING. 1965. Vegetation of the Santa Catalina Mountains, Ari- zona. (II) A gradient analysis of the south slope. Ecology 46:429-452. (Received 6 Jul 1981; revision accepted 14 Feb 1982.) 1982] REVIEWS 215 REVIEWS Flora of Chiapas. Part 1. Introduction to the Flora of Chiapas. By DENNIS E. BREED- LOVE. 35 p. California Academy of Sciences, San Francisco 94118. 1981. $2.50. Not an exchange item. Producing a flora for a tropical region is a staggering task. At present hardly any tropical countries have up-to-date floras, though several have earlier versions and many have works in progress, e.g., Costa Rica, Ecuador, Ceylon, Malesia, etc. It is gratifying to see so many tropical areas receive the attention they deserve, especidlly in view of the irony that some of the earliest floras attempted were of tropical areas, e.g., Flora Zeylanica (1747) and Aublet’s Histoire de Plantes de la Guiane Francoise (1775). The flora of Chiapas, covering an area that is technically subtropical, is a critically important project that will include plants in a transitional region, one lying on the border of the Neotropics and one that we know very little about. The area has been very active tectonically and the geology is complicated, so one might imagine that a documented flora will contribute greatly to our understanding of historical plant geography in Central and South America. Breedlove began collecting 16 years ago for this flora of an area exceeding 74,000 km”. Based on the number of collections made since then (42,000). Chiapas contains 8248 species, perhaps 1000 to 2000 species short of the estimated grand total to be expected. This may be compared to the figures for Guatemala: 109,000 km?/7749 species; and the California Floristic Province: 324,000 km?/4452 species (CFP data from Raven and Axelrod, Univ. Calif. Publ. Bot., Vol 72. 1978). The state appears to be richer floristically and more diverse geographically than Guatemala and is certainly richer and more diverse than any other Mexican state. This volume is an excellent introduction to the flora. The discussion of phytogeo- graphic regions, geological history (a bit brief), and vegetational history will provide a context for understanding the floristic data in future volumes (Part 2, Pteridophytes is reviewed in this issue). Bald catalogues of tropical species are overwhelming, so an interpretation of the results is as important as the catalogue itself. The vegetational formations described are based on Beard’s (1944) durable classifi- cation, and sometimes they seem based on little more than the number of tree layers. Tenuous as this sounds, it seems to work in a practical sense. We all know the difficulties of classifying vegetation. The problem is much greater in the tropics than it is in temperate regions. Breedlove has thus provided a listing of Chiapas forest types referred to by other authors, and where possible, he has included the Holdridge Life Zone terminology; most vegetation types are illustrated by excellent photographs. Vegetation nomenclature in some works is frequently subjective to the point that little information is conveyed by designation of “Cloud Forest” or “Lower Montane Rain- forest,” for instance; therefore, the use of synonyms overcomes individual subjectivity to some extent and allows comparison with one’s own experience in other areas. I was at first skeptical of the alleged presence of buttressed trees in “Evergreen Cloud Forest,” because buttresses are not typically found in the cloud forest of most authors. The term here applies to formations between 2000 and 2900 m, with a canopy up to 40 m. It matters less what you call it as long as you describe it. Little progress can be made toward sensible economic development of tropical regions without reliable data. The basic information in this flora, when it has been completed, will, we hope, be helpful to present and future generations of Mexicans as they wrestle with the problem of how to use land without destroying it. Because it is possible that clearing of tropical forests, as is happening in Chiapas, affects areas well outside tropical latitudes, we can all hope so.—C. DAvipson, Idaho Botanical Garden, P.O. Box 2140, Boise, ID 83701. 216 MADRONO [Vol. 29 Flora of Chiapas (D. E. Breedlove, ed.), Part 2, Pteridophytes. By ALAN R. SMITH. 370 p. California Academy of Sciences, San Francisco 94118. 1981. Price $30.00. This pteridophyte flora of Chiapas, the most southerly state of Mexico, treats 104 genera and 609 species, and notes additional species of adjacent regions that may occur in Chiapas. The work is based on about 5000 collections, most of them by Breedlove and collaborators, which provide an unusually extensive coverage for a tropical Amer- ican fern flora. The keys, nomenclature, descriptions, ecology, distribution and comments have all been carefully prepared. The fine line drawings of 90 genera and 106 species, including details as well as general habit, provide a useful aid for recognition of genera as well as illustrations of critical species. This is a scholarly treatment with obvious attention paid to detail and to accuracy. The work includes nearly all of the species of the wet tropics of Mexico and thus will have application well beyond Chiapas. The treatment of the large genus Thelypteris, with 53 species in seven subgenera, will be useful for identification and assessment of the genus over a broad area of the American tropics, and the account of Asplenium with 52 species will be equally useful. The species taxonomy has been studied in a monographic manner. There are frequent comments concerning uncertainties in the classification due to inadequate knowledge, which serve to point out areas for further work. The genera of ferns are arranged in an alphabetical sequence that avoids the difficulties of presenting a family classification, still an unsettled area of systematic pteridology. Although this is a fern flora of a relatively small part of the American tropics, its basic information provides a firm foundation for further studies of one of the major groups of tropical plants.—ROLLA TRYON, Gray Herbarium, Harvard University, Cambridge, MA 02138. ANNOUNCEMENT FLORA OF THE EASTERN MojAVE DESERT The flora of the eastern Mojave Desert, which appeared in Aliso 10(1):71—186, has been reissued as a separate by Southern California Botanists. Covering the high ranges (Kingston, Providence, New York, Clark, Ivanpah, and Mesquite Mts.) and the Kelso Dunes, the flora is the result of over 10 years of field work by the authors, R. F. Thorne, B. A. Prigge, and J. Henrickson. The plant list itself is carefully set in the context of the geography, climate, and geology of the region, and the plant communities, grouped according to Thorne’s community classification system, are each discussed in detail. Microhabitat, plant association, frequency, elevation, and Mojave distribution are given for each species. The flora is based on ca. 10,000—12 ,000 collections made over the years by many botanists in addition to the authors (notably C. B. Wolf) and represents a total of 783 species. The phytogeographical relationships of the flora and a statistical sum- mary are presented at the end. This work may be ordered from SCB Booksales, % Gardner, 777 Silver Spur Rd., Rolling Hills, CA 90274. Price $7.00 incl. tax. 1982] NOTEWORTHY COLLECTIONS 217 NOTEWORTHY COLLECTION GALAPAGOS ISLANDS CENCHRUS INCERTUS M. A. Curtis (CYPERACEAE).—Ecuador, Galapagos Ids., Wolf (Wenman) I., “Spiny-seeded Grass,” 23 Feb 1979, Nancy Jo s.n. (CAS). Previous knowledge. (See DeLisle, Iowa St. J. Sci. 37:314-316. 1963.) Significance. First record for the archipelago, the nearest known populations being in Costa Rica and Panama. It joins the closely related endemic C. platyacanthus An- derss. and the introduced C. echinatus L. as an addition to the known grass flora. Cenchrus incertus is common on the uninhabited and little-visited Isla Wolf (R. I. Bowman, pers. comm.), whence it probably was introduced by migrating birds. —DuUN- CAN M. PorRTER and MAry LINDA SMYTH, Dept. Biol., Virginia Polytechn. Inst. & State Univ., Blacksburg 24061. NOTEWORTHY COLLECTION New MEXxIcCO—TEXAS SALVIA SUMMA A. Nelson (LAMIACEAE).—New Mexico, Dona Ana Co., Organ Mts., Rattlesnake Ridge (32°14'10”N—106°31'W), 1800 m, 31 May 1980, n. slope and base of limestone outcrop, R. D. Worthington 6050 (UTEP, COLO, NMC); San Andres Mts., n. side of San Andres Pk., growing on limestone cliff, 19 Oct 1975, T. K. Todsen 510198 (NMC); TX, El Paso Co., Franklin Mts., 0.5 km wnw. from top of South Franklin Mountain (31°51'54”N—106°29'44"W), 1940 m, 10 Jul 1977 and 30 May 1981, n. slope and base of tall limestone cliff, R. D. Worthington s.n. and 7136 (UTEP, TEX). Significance. These are the first records of the species from outside the Guadalupe Mts. where reported to be endemic (Correll and Johnston, Man. Vasc. Pl. TX, 1970) and extend the known range 160 km w.—RICHARD D. WORTHINGTON, Dept. Biol. Sci., Univ. Texas, El] Paso, 79968. 218 MADRONO [Vol. 29 NOTEWORTHY COLLECTION CALIFORNIA ABIES LASIOCARPA (Hook.) Nutt. (PINACEAE).—USA, CA, Siskiyou Co.: Marble Mountain Wilderness Area, west of Wilson’s Cabin, Upper Shelley Meadows (T42N R10W S. 23) 1880 m: 18 Jul 1978, Sawyer & Cope 3240 (HSC): at junction of Sky High Lake and Shackleford trails (T33N R12W S. 26) 1910 m: 14 Jul 1971, Sawyer 2423 (HSC); slopes nw. of Deep Lake (T43N R11W S. 15) 2150 m: 16 Jul 1971, Sawyer 2431 (HSC). (The Russian Peak population (Sawyer et al. 1970. Madrono 20:413-415) was the first verified in California.) Significance. The California population can no longer be considered 80 km disjunct from those in Oregon. The new sites are n. of Russian Peak area, which still supports the largest population. The Shelley Meadows and most Russian Peak populations are associated with wet habitats. The other two are dry slope sites. This suggests that more populations are probable in California.—JOHN SAWYER and EDWARD CopPE, Dept. Biol. Sci., Humboldt State Univ., Arcata, CA 95521. NOTES AND NEWS NOTES ON THE DISTRIBUTION OF Malacothrix ON THE CALIFORNIA ISLANDS.—The purpose of this note is to update my recent paper on the distribution of Malacothrix (Asteraceae; Lactuceae) on the California Islands (p. 227-234 in The California Islands, ed. D. Power. 1980). A collection of M. similis Davis & Raven from Santa Cruz Island (Brandegee, April, 1888, UC 92228) which was inadvertently omitted from my paper has been re-examined and confirmed as that species. In addition, a collection of M. similis from San Miguel Island has been discovered (Greene s.n., September 1886, CAS 734). To the list of species on the islands should be added M. glabrata Gray on Los Islas Coronados (R. Surnurr s.n., 1920. LL). Also from Los Islas Coronados are specimens that appear to be M. saxatilis var. tenuifolia (Nutt.) Gray (Blakley 6488 & 6737, SBBG) but they are sterile and cannot be confirmed as that species with certainty. In connection with a study of M. saxatilis and related taxa I have examined specimens from Santa Catalina Island and find forms that are referrable to both M. saxatilis var. saxatilis and M. saxatilis var. tenutfolia. Representatives of a population of M. indecora Greene from Santa Cruz Island Ju- nak SC-312, SBBG) and of a population of M. squalida Greene from middle Anacapa Island (Philbrick & Hochberg B78-288, SBBG) have been grown in cultivation and the chromosome number of the former is 2” = 14 and of the latter is 2m = 28. Lastly, a recent collection of M. insularis Greene from Los Islas Coronados, south island, has come to my attention (Moran 23158, SBBG)—W. S. Davis, Dept. of Bi- ology, University of Louisville, Louisville, KY 40292. SUBSCRIPTIONS—MEMBERSHIP Membership in the California Botanical Society is open to individuals ($18 per year; students $10 per year for a maximum of seven years). Members of the Society receive MADRONO free. Family memberships ($20) include one eight-page publishing allotment and one journal. Emeritus rates are available from the Corresponding Secretary. In- stitutional subscriptions to MADRONO are available ($25). Membership is based on a calendar year only. Applications for membership (including dues), orders for sub- scriptions, requests and rates for back issues, changes of address, and undelivered copies of MADRONO should be sent to the Corresponding Secretary. Renewal payments should be sent to the Treasurer. INFORMATION FOR CONTRIBUTORS Manuscripts submitted for publication in MADRONO should be sent to the editor. All authors must be members, and membership is prerequisite for review. Manuscripts and review copies of illustrations must be submitted in triplicate for all articles and short items intended for NOTES AND NEWS. Follow the format used in recent issues for the type of item submitted and allow ample margins all around. All manuscripts MUST BE DOUBLE SPACED THROUGHOUT. For articles this in- cludes title (all caps, centered), author names (all caps, centered), addresses (caps and lower case, centered), abstract, 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, and number each page. Do not use a separate cover page, “erasable” paper, or footnotes. Table captions should include all information relevant to tables. All measurements should be in metric units. Line copy illustrations should be clean and legible, proportioned (including captions) to the MADRONO page, and designed for reduction to % original size. 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 latitude and longitude references. Halftone copy should be designed for reproduction at actual size. In no case should original illustrations be sent prior to the acceptance of a manuscript. When needed they should be mounted on stiff cardboard and sent flat. No illustrations larger than 22 X 28 cm will be accepted. Presentation of nomenclatural matter (accepted names, synonyms, typification) should follow the format used for Rhus integrifolia in MADRONO 22:288. 1974. Institutional abbreviations in specimen citations should follow Holmgren and Keuken, Index Her- bariorum, 6th edition. Abbreviations of serial titles should be those in Botanico-Per- iodicum-Huntianum (Lawrence et al., 1968, Hunt Botanical Library, Pittsburgh). If the correct abbreviation cannot be determined, the full serial title should be used. Titles of books should be given in full, together with the place and date of publication, publisher, and edition, if other than the first. All members of the California Botanical Society are allotted eight pages in MADRONO over a two year period. Joint authors are each assessed the full page number. Beyond that number of pages a required editorial fee of $40.00 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 sched- ule, with equity among all members for access to its pages. Printer’s fees for illustrations and typographically difficult material @ $35.00 per page (if their sum exceeds 30 percent of the paper) and for author’s changes after typesetting @ $3.00 per line will be charged to authors. CALIFORNIA BOTANICAL SOCIETY = ve ha — K M193 sot, (Ps k MADRONO VOLUME 29, NUMBER 4 OCTOBER 1982 WEST AMERICAN JOURNAL OF BOTANY A Contents DEDICATION TO VOLUME 29 THE ROLE OF PLANT ECOLOGICAL RESEARCH IN SIERRAN PARK MANAGEMENT: A TRIBUTE TO JAcK Major, David J. Parsons PLANT SPECIES DIVERSITY IN ARIZONA, Janice E. Bowers and Steven P. McLaughlin A NEW SPECIES OF ERICAMERIA (ASTERACEAE-ASTEREAE) FROM NORTH-CENTRAL MEXICO, B. L. Turner and Gayle Langford MORPHOLOGICAL DIVERSITY AND TAXONOMY OF CALIFORNIA MESQUITES (PROSOPIS, LEGUMINOSAE), Khidir W. Hilu, Steve Boyd, and Peter Felker SYMPLOCOS SOUSAE, A NEW SPECIES OF SYMPLOCACEAE FROM MExIco, Frank Almeda A SURVEY OF THE CORTICOLOUS MYXOMYCETES OF CALIFORNIA, Kenneth D. Whitney NOTES AND NEWS GYNODIOECY IN Saxifraga integrifolia (SAXIFRAGACEAE), Patrick E. Elvander EFFECTS ON Lomatium triternatum OF THE 1980 ASH FALLOUT FROM MT. ST. HELENS, Amy Jean Gilmartin NOTEWORTHY COLLECTIONS BRITISH COLUMBIA CALIFORNIA COLORADO REVIEW ANNOUNCEMENT INDEX TO VOLUME 29 219 220 227 234 A) Z59 259 269 270 PATON 271 274 275 236 278 IBLISHED 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, Herbarium, Life Sciences Building, University of California, Berkeley, CA 94720. Subscription rate: $25 per calendar year. Subscription information on inside back cover. Established 1916. Second-class postage paid at Berkeley, CA, and additional mailing offices. Return requested. POSTMASTER: Send address changes to Susan Cochrane, California Natural Diversity Data Base, Calif. Dept. of Fish & Game, 1416 9th St. Rm. 1225, Sacramento, CA 95814. Editor—CHRISTOPHER DAVIDSON Idaho Botanical Garden P.O. Box 2140 Boise, Idaho 83701 Board of Editors Class of: 1982—-DEAN W. TayLor, University of California, Davis RICHARD VOGL, California State University, Los Angeles 1983—-ROBERT W. CRUDEN, University of Iowa, Iowa City DUNCAN M. PoRTER, Virginia Polytechnic Institute and State University, Blacksburg 1984—-Mary E. BARKWORTH, Utah State University, Logan HARRY D. THIERS, San Francisco State University, San Francisco 1985—-STERLING C. KEELEY, Whittier College, Whittier, CA ARTHUR C. GIBSON, University of California, Los Angeles 1986—AMyY JEAN GILMARTIN, Washington State University, Pullman ROBERT A. SCHLISING, California State University, Chico CALIFORNIA BOTANICAL SOCIETY, INC. OFFICERS FOR 1982 President: WATSON M. LAETSCH, Department of Botany, University of California, Berkeley 94720 First Vice President: ROBERT ROBICHAUX, Department of Botany, University of California, Berkeley 94720 Second Vice President: VESTA HESSE, P. O. Box 181, Boulder Creek, CA 95006 Recording Secretary: ROBERT W. PATTERSON, Department of Biology, San Francisco State University, San Francisco, CA 94132 Corresponding Secretary: SUSAN COCHRANE, California Natural Diversity Data Base, Calif. Dept. of Fish & Game, 1416 9th St. Rm. 1225, Sacramento, CA 95814 Treasurer: CHERIE L. R. WETZEL, Department of Biology, City College of San Francisco, 50 Phelan Ave., San Francisco, CA 94112 The Council of the California Botanical Society consists of the officers listed above plus the immediate Past President, ROBERT ORNDUFF, Department of Botany, Uni- versity of California, Berkeley 94720; the Editor of MADRONO; three elected Council Members: LYMAN BENSON, Box 8011, The Sequoias, 501 Portola Rd., Portola Valley, CA 94025; JoHN M. TucKER, Department of Botany, University of California, Davis 95616; CHARLES F. QUIBELL, Department of Biological Sciences, Sonoma State Col- lege, Rohnert Park, CA 94928; and a Graduate Student Representative, CHRISTINE BERN, Department of Biology, San Francisco State University, San Francisco, CA 94132. DEDICATION OF VOLUME 29 There are those scientists who are known for their own research and those known for having produced a number of fine students. And there are a few who have managed to have the best of both worlds. Ecol- ogists are familiar with Jack Major’s contributions to the ecological literature, and now David Parson’s tribute to him in this issue points out his very large contribution in motivating many fine ecology stu- dents. The July Madrono (29:3) contained papers presented at a sym- posium held in Davis, California, honoring Jack Major on the occasion of his retirement, and all the presentations were made by Jack’s former students. We doubt that Jack has really retired from anything: teaching and field work are too much a part of what he is, and we take great pleasure in dedicating this volume of Madrono to him. MADRONO, Vol. 29, No. 4, p. 219, 12 October 1982 THE ROLE OF PLANT ECOLOGICAL RESEARCH IN SIERRAN PARK MANAGEMENT: A TRIBUTE TO JACK MAJOR DAVID J. PARSONS National Park Service, Sequoia and Kings Canyon National Parks, Three Rivers, CA 93271 ABSTRACT The contributions of Jack Major and his students to plant ecological research in the Sierra Nevada of California is reviewed. INTRODUCTION Management policies of the National Park Service call for the pres- ervation of naturally functioning ecosystems. In practice, Park man- agers often find these policies difficult to implement. This difficulty stems both from ambiguities in the interpretation of pertinent legis- lation (Bonnickson and Stone 1982) and the requirement that visitor use be allowed, and even encouraged. In addition to direct impacts of visitor use, such external influences as air pollution and the intro- duction of exotic species, make the preservation of natural ecosystem processes difficult, if not impossible (White and Bratton 1980). It has only been relatively recently that the National Park Service has recognized the importance of baseline ecological research upon which to base resource management programs (Leopold et al. 1963). Much of this research is now carried out by Park Service scientists stationed in the larger Parks, or through Cooperative Park Studies Units (CPSUs) that have been established on University campuses in most of the larger Western States (the most recent CPSU to be estab- lished is at the University of California, Davis). The other major source of scientific expertise and research data is that provided by university scientists. It is in this regard that Jack Major has provided significant contributions to the research input needed to manage the National Parks of the Sierra Nevada. His interest, as well as that of his students and associates, in the plant ecology of the Sierra Nevada has contrib- uted to a basic understanding of the vegetation, which in turn provides a basis for effective resources management throughout the Sierra. Jack Major first developed his interest in the mountains as a child growing up in Utah. Born and raised in the Salt Lake City area, he frequently explored the Uinta Mountains, both in the summer and winter, with his brother Ted. Jack received his first real training in botany as a high school student of A. O. Garrett. He attended the MADRONO, Vol. 29, No. 4, pp. 220-226, 12 October 1982 1982] PARSONS: ECOLOGICAL RESEARCH 224 University of Utah and Utah State Agricultural College in Logan as an undergraduate, receiving a bachelors degree in Range Management from the latter in 1942. During the next eight years he worked for the U.S. Forest Service Intermountain Forest and Range Experiment Sta- tion in Ogden as well as serving in the 10th Mountain Division during World War II (he was a ski instructor among other things). In 1953 Jack received his doctorate in Soils Science from the Uni- versity of California, Berkeley. He taught plant ecology and related courses in the botany department at the University of California, Davis from that time until his retirement in 1981. During that period he spent as much of his time as possible in the field, preferably in the subalpine and alpine reaches of the Sierra Nevada. Anyone privileged enough to accompany Jack on one of his many field trips can appre- ciate his eager enthusiasm and genuine concern both for nature and his fellow man. In addition to directing the field research of fourteen Ph.D. and five masters students, Jack found time to publish over 170 scientific articles and book reviews. Special awards and honors include a Fulbright Fellowship to study in Innsbruck, Austria and receipt in 1977 of the Distinguished Service Award from the Ecological Society of America. Throughout all of this special credit must be given to Jack’s wife Mary. She has been a partner in much of his work and the driving force behind much of his success. Together, they have made an eager and able field team under even the most trying of circumstances. Throughout his professional career Jack Major has shown a special interest in the ecology of the Sierra Nevada. Spending as much time as possible in such spots as Convict Creek, Rock Creek, and Sequoia National Park. Jack never tired of the mountains. The fact that the products of much of his research were to be of value in managing the very mountains he was so fond of made things even better. In the remainder of this paper I review selected examples of the research contributions made by Jack Major, and his students, to the basic knowledge of Sierran ecosystems and how they relate to many of the management problems faced by the National Park Service. IMPACTS OF FIRE SUPPRESSION AND GRAZING Of all the management activities that have occurred in the Sierran National Parks, fire suppression and grazing have had the greatest impact on altering natural ecosystems (Leopold et al. 1963). Both of these practices have occurred for many years and have long lasting impacts. Under the guidance of Jack Major, John Vankat became interested in the effects of these activities. His doctoral dissertation (Vankat 1970) provided a detailed historical review of both fire suppression and grazing in Sequoia National Park. This carefully doc- umented work used historical photographs, personal accounts, and 222 MADRONO [Vol. 29 vegetation transects to document presettlement conditions as well as provide the first quantitative analysis of such associated impacts as an increased threat of wildfire, shifts in successional patterns and the destruction of montane and subalpine meadows. It provided a valu- able baseline upon which fire and grazing management programs have been developed. The chaparral and coniferous forests of the Sierra Nevada evolved with periodic lightning and Indian-caused fires (Kilgore and Taylor 1979, Parsons 1981). However, during the early days of National Park management, fire, like other external influences, was considered to be bad and was to be avoided at all costs. Effective fire suppression resulted in fuel accumulations and subsequent threats of unnaturally hot and destructive wildfires. As a result of altered burning patterns shifts in species succession have occurred (Vankat 1977, Vankat and Major 1978, Parsons and DeBenedetti 1979). The documentation of the undesirable impacts of fire suppression has led to the recent es- tablishment of integrated fire management programs that include the use of naturally ignited and prescribed burns to reintroduce fire to what is thought to have been its natural role in park ecosystems (Par- sons 1980). The work of Vankat (1970, 1977) and Vankat and Major (1978) played a significant role in the successful establishment of these programs. The grazing of domestic livestock (sheep, cattle, horses, and mules) in Sierran woodlands and meadows was started long before the estab- lishment of the Sierran parks (Vankat 1970, DeBenedetti and Parsons 1979). Vankat (1970) and Vankat and Major (1978) have documented the history as well as the destructive impacts of such grazing on the plant communities of Sequoia National Park. They conclude that graz- ing has had the greatest impacts in the low elevation oak woodlands and high elevation subalpine meadows. This type of data, together with ongoing Park Service studies, provides a basis for establishing limits on grazing in backcountry meadows. Interest in the history and effects of livestock grazing led to a rec- ognition of a lack of understanding of basic ecological and geological history of Sierran meadows. Under Jack Major’s direction, Nathan Benedict became interested in these questions and pursued his disser- tation research on the vegetation and origins of Sierran meadows in the Kern River drainage of Sequoia National Park (Benedict 1981). The first published product of this research, a physiographic classifi- cation of subalpine meadows (Benedict and Major 1982), identifies two basic meadow types and discusses their formation and mainte- nance in light of meadow stability and management. Other students of Jack Major who have completed studies of low or middle elevation vegetation in the Sierra include Roman Gankin (1957; ecology of Arctostaphylos myrtifolia), Rod Myatt (1968; ecology 1982] PARSONS: ECOLOGICAL RESEARCH PR: of Eriogonum apricum) and Gary Sanford (1972; plant ecology of Sier- ran streams). The basic information provided by these studies is es- sential for effective vegetation management. SUBALPINE AND ALPINE VEGETATION STUDIES One of the most serious threats facing Sierran parks is the increasing number of backcountry users (van Wagtendonk 1981) and their as- sociated impacts on high elevation plant communities. In order to evaluate the impacts of backcountry use, park managers need detailed background information on the subalpine and alpine vegetation of the Sierra Nevada. The acquisition of such baseline descriptive data has long been of special interest to Jack Major and his students. For ex- ample, Jack’s chapter with Dean Taylor (Major and Taylor 1977) on the alpine zone of the Sierra reviews and compiles all available vege- tation information on that zone. In addition, several of Jack’s students have carried out detailed ecologic, descriptive, or taxonomic studies of many high elevation Sierran plant communities. Dean Taylor’s (1976a) doctoral study of the ecology of timberline vegetation around Carson Pass provides a thorough description of the vegetation and factors affecting its distribution and survival in the central Sierra Nevada. Although Carson Pass is not located within a National Park the plant communities and habitat requirements are similar to those found in the parks and thus Taylor’s findings are of considerable relevance to park scientists and managers. Taylor has continued to show a wide-ranging interest in Sierran plant ecology. He has provided valuable assistance and support to other investigators as well as compared Sierran floristic relationships with other areas (Taylor 1976b, 1977). Merry Lepper’s dissertation study on environmental relationships of limber pine (Pinus flexilis) in the Sierra Nevada and Rocky Mountains provides interesting insights into the factors influencing the distribu- tion, and survival and germination requirements of that high elevation species (Lepper 1974). The distributional requirements of another sub- alpine tree species, mountain hemlock (Tsuga mertensiana), was the subject of an undergraduate, NSF-sponsored study under Jack Ma- jor’s direction by this author (Parsons 1972). That experience helped solidify my interests in Sierran vegetation and provided a strong im- petus for future graduate studies in plant ecology. The contribution of nitrogen fixation by mountain alder (Alnus ten- utfolia) to the nutrient budget of a small watershed in the Lake Tahoe basin was investigated by Michael Fleschner as his dissertation study (Fleschner 1975). His findings showed alder to be a significant source of nitrogen input. Jim Neilson (1971) also analyzed the vegetation of the Tahoe Basin region and provided a basis for understanding im- 224 MADRONO [Vol. 29 pacts of existing and proposed developments. Neilson (1961) also stud- ied plant associations on glaciated granite at Sterling Lake in Nevada County. Other regional vegetation studies in the Sierra were conducted under Jack’s guidance by Rich Pemble and Mary Burke. Pemble’s disserta- tion (Pemble 1970) described the alpine vegetation of several Sierran locations as related to topography and soil parent material. The vege- tation classification he developed was based on associations of species that can be arranged to reflect a mosaic of habitat conditions. Burke (1979) also used floristic criteria to characterize subalpine and alpine vegetation types in her phytosociological study of the flora and vege- tation of the Rae Lakes Basin in Kings Canyon National Park. Burke’s study provided an understanding of vegetational patterns in terms of environmental factors and also provided a basis for making floristic and vegetational comparisons with other areas. She found that the four major environmental factors ordering the vegetation in the land- scape are moisture regime, snow cover, altitude, and substrate sta- bility (Burke 1979). Jack Major has always been particularly interested in the influence of substrate on vegetation composition and diversity. One three-year study he directed compared the flora of granitic and calcareous sites in the southern Sierra Nevada. Barbara Rice carried out much of this work, which culminated in a comprehensive checklist of plant species for the Mineral King area of what was then Sequoia National Forest and is now part of Sequoia National Park (Rice 1969). This checklist has provided a valuable basis for identifying potential threatened or endangered plants of the Mineral King area. Finally, Jack has recently co-directed Sue Conard’s study of inter- actions, succession, and environment in the montane chaparral—white fir (Abies concolor) vegetation of the northern Sierra Nevada (Conard 1980). He has also collaborated with Bob Curry (U.C. Santa Cruz) in researching and writing a recently completed survey of sites in the Sierra Nevada to be included in the Natural Landmark Program for the U.S. Department of Interior. CONCLUSIONS In addition to the valuable works produced by his students it is important that we recognize the interest and ability Jack Major himself has shown in field studies in the Sierra. He has provided valuable consultation and inspiration to countless undergraduate as well as graduate students of his colleagues both at U.C. Davis and elsewhere. His broad interest in Sierran plant ecology is further reflected in a wide variety of publications (Major 1963, Major and Bamberg 1963, 1967; Major and Taylor 1977, Stebbins and Major 1965). Jack Major has always been, regardless of the time of year or the 1982] PARSONS: ECOLOGICAL RESEARCH 225 weather conditions, eager and willing to embark on rigorous field expeditions. He can always be counted on to spend as much time as needed or possible in the field with his students and classes, or to provide consultation or advice to whoever might ask. Whether it is a downhill ski traverse of Meyers grade, a cross country ski expedition into Rock Creek to observe snow accumulation patterns on subalpine meadows, a day hike on Dana Plateau, a rigorous backpack trip through Kings Canyon to review ongoing wilderness impact and big- horn sheep research, or one of his undergraduate plant ecology classes on a trans-Sierran field trip, Jack has always been eager and willing to share his experiences and knowledge. His Munz flora in hand, and the latest European or Russian text to be reviewed in his pack, Jack Major in the field is a familiar sight to plant ecologists throughout the west. His interest in Sierran plant ecology, while seldom aimed spe- cifically at management problems, provides an excellent example of the value and need of basic research both in furthering our under- standing and improving our management of the types of ecosystems preserved in our Sierran National Parks. LITERATURE CITED BENEDICT, N. B. 1981. The vegetation and ecology of subalpine meadows of the Sierra Nevada, California. Ph.D. dissertation, Univ. California, Davis. and J. Major. 1982. A physiographic classification of subalpine meadows of the Sierra Nevada of California. Madrono 29:1-12. BONNICKSON, T. M. and E. C. STONE. 1982. Managing vegetation within U.S. National Parks: a policy analysis. Environ. Managem. 6:101-102. BurRKE, M. T. 1979. The flora and vegetation of the Rae Lakes Basin, southern Sierra Nevada: an ecological overview. M.S. thesis, Univ. California, Davis. CONARD, S. G. 1980. Species interactions, succession, and environment in the mon- tane chaparral-white fir vegetation of the northern Sierra Nevada, California. Ph.D. dissertation, Univ. California, Davis. DEBENEDETTI, S. H. and D. J. PARSONS. 1979. Mountain meadow management and research in Sequoia and Kings Canyon National Parks: a review and update. In R. Linn, ed., Proceedings of First Conference on Scientific Research in the National Parks, Vol. I]. USDI Natl. Park Serv., Trans. and Proc. Ser. 5. p. 1306— 1311. FLESCHNER, M. D. 1975. Nitrogen fixation in a small subalpine water shed. Ph.D. dissertation, Univ. California, Davis. GANKIN, R. 1957. The variation pattern and ecological restrictions of Arctostaphylos myrtifolia Parry (Ericaceae). M.A. thesis, Univ. California, Davis. KILGORE, B. M. and D. TAYLOR. 1979. Fire history of a sequoia-mixed-conifer forest. Ecology 60:129-142. LEOPOLD, A. S., S. A. CAIN, C. CoTTAM, I. N. GABRIELSON, and T. L. KIMBALL. 1963. Wildlife management in the national parks. Trans. N. Amer. Wildl. Nat. Resour. Conf. 28:28-45. LEPPER, M. G. 1974. Pinus flexilis James and its environmental relationships. Ph.D. dissertation, Univ. California, Davis. Major, J. 1963. Vegetation mapping in California. Jn R. Tuxen, ed., Bericht tiber das Internationale Symposium fur Vegetationskartierung, p. 195-218. Verlag J. Cramer. Weinheim, Germany. 226 MADRONO [Vol. 29 and S. A. BAMBERG. 1963. Some Cordilleran plants new for the Sierra Nevada of California. Madrono 17:93—109. and 1967. Some Cordilleran plants disjunct in the Sierra Nevada of California and their bearings on Pleistocene ecological conditions. Jn H. E. Wright and W. H. Osburn, eds., Arctic and alpine environments, p. 171-188. Indiana Univ. Press, Bloomington. and D. W. TayLor. 1977. Alpine. In M. G. Barbour and J. Major, eds., Terrestrial vegetation of California, p. 601-675. Wiley-Interscience, NY. MyaTT, R. G. 1968. The ecology of Eriogonum apricum Howell. M.S. thesis, Univ. California, Davis. : NEILSON, J. A. 1971. Vegetation of the Lake Tahoe region. Tahoe Regional Planning Agency, South Lake Tahoe, CA. 1961. Plant associations on glaciated granite at Sterling Lake, Nevada Coun- ty, California. M.S. thesis, Univ. California, Davis. PARSONS, D. J. 1972. The southern extensions of Tsuga mertensiana (mountain hem- lock) in the Sierra Nevada. Madrono 21:536—-539. 1980. The role of fire management in maintaining natural ecosystems. Jn Proc. Conference Fire Regimes and Ecosystem Properties, p. 469-488. USDA For. Serv. Gen. Techn. Report WO-26. 1981. The historical role of fire in the foothill communities of Sequoia Na- tional Park. Madrono 28:111-—120. and S. H. DEBENEDETTI. 1979. Impact of fire suppression on a mixed-conifer forest. Forest Ecology Managem. 2:21-33. PEMBLE, R. H. 1970. Alpine vegetation in the Sierra Nevada of California as litho- sequences and in relation to local site factors. Ph.D. dissertation, Univ. California, Davis. RicE, B. 1969. Plant checklist for Mineral King, California. Unpubl. report to Se- quoia National Forest, Porterville, CA, 27 p. SANFORD, G. R. 1972. Plant ecology of the upper reaches of streams draining two watersheds in the Sierra Nevada. Ph.D. dissertation, Univ. California, Davis. STEBBINS, G. L. and J. Major. 1965. Endemism and speciation in the California flora. Ecol. Monogr. 35:1-35. TAYLOR, D. W. 1976a. Ecology of the timberline vegetation at Carson Pass, Alpine County, California. Ph.D. dissertation, Univ. California, Davis. 1976b. Disjunction of Great Basin plants in the northern Sierra Nevada. Madrono 23(6):301-310. 1977. Floristic relationships along the Cascade-Sierran Axis. Amer. Mid. Naturalist 97:333-349. VANKAT, J. L. 1970. Vegetation change in Sequoia National Park, California. Ph.D. dissertation, Univ. California, Davis. 1977. Fire and man in Sequoia National Park. Ann. Assoc. Amer. Geogr. 67:17-27. and J. Major. 1978. Vegetation changes in Sequoia National Park, California. J. Biogeogr. 5:377—402. VAN WAGTENDONK, J. W. 1981. The effect of use limits on backcountry visitation trends in Yosemite National Park. Leisure Sciences 4:311—323. WHITE, P. S. and S. P. BRATTON. 1980. After preservation: philosophical and prac- tical problems of change. Biol. Conserv. 18:241-255. (Received 6 Jul 1981; revision accepted 7 Dec 1981.) Text originally delivered at a symposium honoring Jack Major, held at UC, Davis, on 29 May 1981. PLANT SPECIES DIVERSITY IN ARIZONA JANICE E. BOWERS and STEVEN P. MCLAUGHLIN Office of Arid Lands Studies, University of Arizona, 845 N. Park, Tucson 85719 ABSTRACT The influence of elevational range, area, and collecting time on the absolute gamma diversity of 20 local floras in Arizona is determined by multiple regression. Elevational range and collecting time together account for 77% of the variance in species number among the floras. Residuals from the prediction equation are used to derive a measure of the relative gamma diversity of each flora. The remaining variance is associated with vegetation community-types, presence of permanent water, and occurrence of major canyon environments. Vegetation community types that are rich in species are Madrean evergreen woodland and desert grassland. Sonoran desert scrub and interior chaparral are relatively poor in species. The state of Arizona, covering over 29 million hectares and ranging in elevation from near sea level to over 3800 m, supports about 3400 species of vascular plants. Although it is readily apparent that these species are not evenly distributed across the state, little work has been done to analyze objectively patterns in species diversity among various vegetation types and geographical areas in the state. The terms “diversity” and “richness” have several meanings and connotations. Whittaker (1972) distinguishes three components of di- versity. Alpha diversity is the number of species in standard size sam- ple plots, whereas beta diversity describes the change in species com- position between such plots. Gamma diversity, the number of species in a regional or local flora, can be formally defined as the product of the alpha diversity of the component communities and the degree of beta differentiation between them (Whittaker 1972). Diversity and richness usually refer to the absolute number of species in a flora. However, when used as adjectives, they imply that a flora not only possesses many species, but that it presumably has more species than most other floras. We make a distinction between absolute diversity, the number of species in a flora, and relative diversity or richness, the comparative richness or poorness in species from place to place. This paper is concerned with relative gamma diversity among regional or local Arizona floras. A few studies have examined relative alpha and beta diversity with- in a single geographical region of Arizona (Whittaker and Niering 1965, 1968; Brady 1973, Wentworth 1979). For the most part, gamma diversity of local Arizona floras has been discussed only in subjective terms (Toolin et al. 1980, Smith 1974, Lehto 1970) or by quantitative MADRONO, Vol. 29, No. 4, pp. 227-233, 12 October 1982 228 MADRONO [Vol. 29 comparison with another flora (Bowers 1980, Burgess 1965). Although numerous local and regional floras have been compiled within Ari- zona, direct comparison of absolute gamma diversity is not meaningful because these floras vary considerably in the amount of area covered, habitat diversity, and collection history. Our purpose is to provide an objective basis for comparing relative gamma diversity among local Arizona floras, identify regions and vegetation types of high and low relative diversity, and suggest other factors that appear to contribute to variation in diversity among local floras. METHODS Bowers (1981) reviewed floristic work in Arizona. From her bibli- ography we selected 20 floras for which the following data were avail- able: number of species, elevational range, areal extent, and collecting time (Table 1). We tended to avoid floras compiled for relatively small areas (<100 ha), or for discontinuous areas (e.g., Montezuma Castle National Monument, Navajo National Monument). We added to the flora of Santa Cruz County the number of Pteridophyta and Cypera- ceae reported in Kearney and Peebles (1969) and the number of Gra- mineae reported in Gould (1973) for that area. When not stated in the original flora, elevational ranges and areas were obtained from USGS topographic maps. Collecting time was defined as the number of years (to the nearest 0.5 year) invested in collecting for and compiling each flora. Where two or more floras for the same area are cited in Bowers (1981), we used the number of years between the earliest and the most recent lists as our estimate of collecting time. We examined relative diversity by comparing for each flora the observed number of species (S) with a computed estimate of the ex- pected number of species (S). Floras encompassing large areas or el- evational ranges or compiled over many years would be expected to have more species than floras covering smaller areas, less elevational range, or compiled over shorter periods. The extent to which S is related to these factors—area, elevational range, and collecting time— can be determined by multiple regression (e.g., Conner and Simberloff 1978). We used both S and the natural log of S (In S) as dependent variables; independent variables were elevational range (E), In E, area (A), In A, collecting time (T), and In T; 1.e., all linear, semilog, and log-log relationships were tested. The residuals, i.e., the differences between the observed and expected numbers of species (S — S), were then used to derive an expression for relative richness. RESULTS AND DISCUSSION The best equation for predicting S was the linear regression with elevational range and collecting time: 1982] BOWERS AND McLAUGHLIN: SPECIES DIVERSITY 229 Ser = 47 + 0.349E + 8.20T (1) with total significant (p < 0.05) R? = 0.77. In our sample, area was not as good a predictor of S as was elevational range. Elevational range is probably the better measure of habitat diversity in Arizona, where climate, soils, and vegetation change dramatically with changes in altitude (Shreve 1922, Whittaker and Niering 1965). The absolute diversity of a flora should increase with collecting effort. Furthermore we expect that the number of species encountered in an area should be a decreasing function of collecting time. There- fore, it is surprising that T provided a better fit than In T, after correcting for the variance accounted for by E. Possible explanations for the better linear fit might be: 1) over long periods of time significant changes in habitat occur that favor immigration of new species; 2) in arid regions, longer periods of time are more likely to include favorable seasons when additional species, particularly annuals, can be collect- ed; and 3) our method of quantifying collecting time might not be linearly correlated with actual effort, particularly over longer periods involving the publication of updated check-lists. It is our thesis that the remaining 23% of the variance of S, i.e., that not accounted for by E and T, is relevant to the concept of relative diversity. An examination of the residuals of equation (1) can provide insight into additional factors that influence gamma species diversity. By expressing the residuals from equation (1) as percentages, we define relative richness (R) as 100(S — Spr) SET R= The R values for each flora are given in Table 2, which ranks the floras from high to low relative richness. The interpretation of these R values is straightforward. For example, the R value for Grand Canyon National Park is 26, meaning there are 26% more species in the flora than would be expected on the basis of the elevational range and collecting time for that flora; the R value of —31 for Cabeza Prieta Game Range means there are 31% fewer species than expected. The floras in Table 2 are ranked from high to low R. Three factors that appear to be associated with relative richness are vegetation, aquatic habitats, and canyon environments. We have identified four vegetation community-types that are related to relative richness. Floras of areas dominated by Madrean evergreen woodland or desert grassland tend to have high R values, while floras from Sonoran desert communities or interior chaparral have low R values (Table 2). Montane conifer forests are somewhat equally distributed among floras with high and low R, while the small number of other 230 MADRONO [Vol. 29 TABLE 1. LOCAL ARIZONA FLORAS USED IN ANALYSIS OF RELATIVE GAMMA DI- VERSITY. References can be found in Bowers (1981). Eleva- Collect- No. tional Area ing time Flora species range (m) (ha) (yr) Aquarius Planning Unit 500 1520 135,000 220 Cabeza Prieta Game Range 304 825 380,000 13.0 Canyon de Chelly National Monument 518 630 34,000 50 Chiricahua National Monument 687 630 4300 36.0 Grand Canyon National Park 1574 2450 493,000 42.0 Hualapai Planning Unit 697 2280 247,000 2.0 Lake Pleasant Regional Park 364 265 5830 55 McDowell Mountain Regional Park 286 754 8475 2,0 Organ Pipe Cactus National Monument 522 1160 134,000 12.0 Phelps Cabin Natural Research Area 195 110 126 24.0 Rosemont Area, Santa Rita Mtns. 416 590 6500 1.5 Santa Cruz County 1148 1880 323,000 32.0 Sierra Ancha Experimental Forest 739 1270 5200 35-0 Sierra Estrella Regional Park 330 1070 7530 1.5 Sycamore Canyon Natural Area 624 427 932 45.0 Three Bar Wildlife Area 521 1750 15,750 14.0 Tonto National Monument 270 460 454 4.0 Tumamoc Hill 438 226 1036 68.0 Walnut Canyon National Monument 248 128 761 29.0 White Tank Mtns. Regional Park 632 830 11,560 1.0 vegetation types in our sample preclude drawing conclusions about their contribution to relative richness. Madrean evergreen woodlands are dominant in four of the eight floras with positive R values. Desert grassland communities occur in these four floras as well. Areas in which Madrean evergreen woodland and desert grassland contribute to high R are largely confined to south- eastern Arizona, where mild temperatures, high summer rainfall, and spring rainfall provide a favorable environment. Although richness in both community types is due largely to numerous annual and perennial herbs that proliferate during and immediately after summer rains, precipitation and temperature are also adequate to support a spring flora in these communities. Brady (1973) and Wentworth (1979) sug- gest that the scattered canopies of oaks at the upper margin of the desert grassland and the lower margin of the Madrean evergreen woodland create a heterogenous environment in which many species of herbs can flourish. Such a situation occurs at the Rosemont Area, where oak woodland and desert grassland interdigitate in shallow can- yons and rolling hills. Species with Madrean floristic affinities con- tribute significantly to the diversity of Sycamore Canyon, Chiricahua National Monument, and Rosemont, where they comprise 15, 14, and 11%, respectively, of the floras. ype | BOWERS AND McLAUGHLIN: SPECIES DIVERSITY 1982] Or v G I 0 v 0 Z v v I Z I I x (X) X X X xX x Oo x X xX = (X) (X) xX X (X) (Xx) x ner, X (xX) xX x xX >.< xX x X xX (& (XxX) X Xx (X) X ds 0d MN AN dd JI Od (X) MO (X) (X) 3d4}-A}IUNWIWOD U0T}e}990 A, ‘puvysseis ulseg yeoiy pue surejd = 5g ‘puelpoom Jajiuos ulseg yealn, = M4) ‘4S910J JajIuod vue}JUOUW I T I ¢ T I ¢ ¢ (x) D4 a ih ie D i i ve A ye (x) (x) i: / / / / OS IS WO OV O> WY UUM seIOP jo Jaquinyy O< Y UUM seiIOP jo Jaquiny Eat Oo P ao ITH 2owewny valy YoIeasoy [einyeny ulqey sdyjayd IBULY BUCS) LJIIg ezZaqey JusUINUOTY [eUOTJeNY UOAUBD NUL MA yleg [PUOIsIY LI[ISY CIIIIC yup sutuuelg redereny yu suruuelg sniuenby yleg [euoissy urleyUuNOP, [[PMOGIP Jso104 [eJUsWILIedxyY eVYyIUY PIII jusUINUOyY [PUOTIENY snjIeV_D sdig uesig yieg [euoidey ‘sui, YURL 1M eoly [einjeN UOAULD IIOWBIAS JUusWUINUOT [PUOT}BNY OJUOT, AyuNOo, ZNAId eyUeS JusUINUOTY [eUOTeNY enyeouyD yleg jeuoneN uoduey pues ‘SUIT PITY EJURS ‘vaIY JUOWIISOY JusUINUOPY [PUOTeN AT[PYD ap uoAUeD yleg [PUOIsay JURSeI[q IAT e10[ ‘(LL61) aMO'T pue UMOIG SMOTIOJ SadA}-A}IUNWIUIOD JO UOTINIYSIG (x) Aq sadAj-AVtuNUIWIOS IOUT ‘x Aq poa}JOUsp e1IOY sy} JO salIepuNog ay} UTYyWM sodA}-AyTUNUIWIOD Joe, ‘qndOs Jlasap UBIOUODS = QS ‘pur -SSPI3 JAaSapIwas = Hq ‘puv[poom uIoIBI9AN URIIPRIT = MIN ‘GN49s Wasep sARYO|T = CIN ‘HWosop ulseg way = C4 ‘[esredeyod 10l193ul = OT = AWW ‘puvyjsseis suidjeqns = 9g ‘4sa10j Jajiuod auidyeqns = 4S ‘adeospur] suneurmop (s)uokued = WoO ‘(spuod ‘saye] ‘swreass JUsUeULIEd) s}yezIqey INeNbe = OY ‘ssauyo saAnejal = Y ‘SVEOTY TVOOT 0Z HLIM GALVIOOSSY SAdA L-ALINQWWOD) NOILVLEADAA ANV ‘SHANALVAY AdVOSAGNVT ‘SSANHOIY GAILVIGY “7 ATAVE 232 MADRONO [Vol. 29 Although Whittaker and Niering (1965) found that the north-slope shrub-phase Sonoran desert community in the Santa Catalina Moun- tains is among the richest in the United States, most desert regions in our sample are relatively poor in species. Ten of the 12 floras that include Sonoran desert communities had negative R values. The north- slope shrub-phase community studied by Whittaker and Niering may have a high alpha diversity compared to other Sonoran desert com- munities because it is ecotonal between the desert and desert grassland and because it is located in canyons. Relative gamma diversity for most desert areas in our sample is low because low beta diversity across large expanses of desert masks the input from occasional com- munities with high alpha diversity. Richerson and Lum (1980) found that the desert areas of California were also relatively poor in species compared to other parts of the state. Lake Pleasant Regional Park is a Sonoran desert area notable in having the highest R value of all 20 floras included in the sample. This area is on the northern boundary of the Sonoran Desert, and its flora contains a high proportion of introduced and aquatic species. Still, these factors do not seem entirely to explain its high diversity. Areas dominated by interior chaparral also demonstrate low relative richness. Not only are interior chaparral communities composed of relatively few dominant species, but the closed canopy precludes de- velopment of a species-rich herbaceous understory (Cable 1975). In- terior chaparral is best developed in western and central Arizona, a region receiving less summer rainfall than the Madrean woodland and desert grassland areas of southeastern Arizona. Presence of aquatic habitats in otherwise arid or semi-arid regions contributes to high relative richness in a number of floras. Of the eight floras with R 210, five occur in areas with permanent streams, lakes, or ponds. Of the nine floras with R <—10, only one is for an area with permanent aquatic habitats. Aquatic habitats are particularly important at Lake Pleasant Regional Park, a Sonoran desert region with a high R value, where aquatic and wetland species comprise 8% of the flora. The flora of Organ Pipe Cactus National Monument has a higher R value than most of the other desert areas in our sample partly because of the many aquatic and semi-aquatic species found in and around sources of permanent water at the Monument. When a canyon comprises a substantial proportion of an area, the flora is often relatively rich. Six of the floras are for areas centered about major canyon features and five have positive R values. Canyon environments promote high relative richness by providing a variety of microhabitats congenial to species that would not otherwise occur in close proximity (Toolin et al. 1980). Canyon environments contribute to the high relative richness of Sycamore Canyon Natural Area, Grand Canyon National Park, Canyon de Chelly National Monument, Chi- ricahua National Monument, and Tonto National Monument. 1982] BOWERS AND McLAUGHLIN: SPECIES DIVERSITY 233 As more local floras become available, particularly for areas con- taining vegetation community-types poorly represented in our sample, this analysis can be revised and expanded. Floras of natural areas, rather than political or administrative units, would be particularly valuable. Our analysis of relative diversity does show that local floras constitute a previously underutilized data base for studying regional patterns in species distribution. LITERATURE CITED BowERs, J. E. 1980. Flora of Organ Pipe Cactus National Monument. J. Ariz.-Nev. Acad. Sci. 15:1-11, 33-47. 1981. Local floras of Arizona: an annotated bibliography. Madrono 28:193-— 209. Brapy, W. 1973. Patterns of vegetative diversity in the Huachuca Mountains. Ph.D. dissertation, Colorado State Univ., Fort Collins. Brown, D. E. and C. H. LOWE. 1977. Biotic communities of the southwest. Rocky Mtn. For. and Range Exp. Sta., USDA For. Serv. Gen. Techn. Rep. RM-41, Fort Collins, CO. BurGEss, R. L. 1965. A checklist of the vascular flora of Tonto National Monument, Arizona. J. Arizona Acad. Sci. 3:213-223. CABLE, D. R. 1975. Range management in the chaparral type and its ecological basis: the status of our knowledge. Rocky Mtn. For. and Range Exp. Sta., USDA For. Serv. Res. Paper RM-155, Fort Collins, CO. CONNOR, E. F. and D. SIMBERLOFF. 1978. Species number and compositional sim- ilarity of the Galapagos flora and avifauna. Ecol. Monogr. 48:219-248. GOULD, F. W. 1973. Grasses of the southwestern United States. Univ. Arizona Press, Tucson. KEARNEY, T. H. and R. H. PEEBLES. 1969. Arizona flora. Univ. California Press, Berkeley. LEHTO, E. 1970. A floristic study of Lake Pleasant Regional Park, Maricopa County, Arizona. M.S. thesis, Arizona State Univ., Tempe. RICHERSON, P. J. and K. LUM. 1980. Patterns of plant species diversity in California: relation to weather and topography. Amer. Nat. 116:504—536. SHREVE, F. 1922. Conditions indirectly affecting vertical distribution on desert moun- tains. Ecology 3:269-274. SMITH, E. L. 1974. Phelps Cabin Natural Research Area. Jn Established natural areas in Arizona: a guidebook for scientists and educators, p. 127-133. Planning Division, Office of Economic Planning and Development, Office of the Governor, Phoenix. TOOLIN, L., T. R. VAN DEVENDER, and J. M. KAISER. 1980. The flora of Sycamore Canyon, Pajarito Mountains, Santa Cruz County, Arizona. J. Ariz.-Nev. Acad. Sci. 14:66—74. WENTWORTH, T. R. 1976. The vegetation of limestone and granite soils in the moun- tains of southeastern Arizona. Ph.D. dissertation, Cornell Univ., Ithaca, NY. WHITTAKER, R. H. 1972. Evolution and measurement of species diversity. Taxon 21:213-—251. and W. A. NIERING. 1965. Vegetation of the Santa Catalina Mountains, Arizona: gradient analysis of the south slope. Ecology 46:429—-452. and 1968. Vegetation of the Santa Catalina Mountains, Arizona. IV. Limestone and acid soils. J. Ecol. 56:523-544. (Received 16 Mar 1981: accepted 5 Oct 1981)- A NEW SPECIES OF ERICAMERIA (ASTERACEAE-ASTEREAE) FROM NORTH-CENTRAL MEXICO B. L. TURNER and GAYLE LANGFORD Department of Botany, The University of Texas, Austin 78712 ABSTRACT A new species, Ericameria riskindii, is described from north-central Mexico, near Saltillo. It is remarkable for its remote relationship to other species in this area, its closest relative seemingly being E. juarezensis of Baja California. Recent explorations in northern Mexico by Dr. James Henrickson and colleagues has revealed a remarkable Ericameria, which we de- scribe below. Urbatsch (1978) treated the genus for north-central Mex- ico (i.e., the Chihuahuan desert species) but, not having seen material of the present collection, was unable to include the taxon. Because of its decidedly spatulate leaves, the plant will not key in his treatment. Ericameria riskindii Turner & Langford, sp. nov. Fruticuli humiles rotundati ad 25 cm alti. Folia spatulata glandu- lario-punctata conferta. Capitula solitaria terminalia sessilia vel fere sessilia. Involucrum glutinosum late turbinatum 3—4-seriatum. Rami stylorum appendicibus longis linearibus puberulis (Fig. 1). Rounded shrublets 10-25 cm high. Stems minutely hispid; inter- nodes 1-5 mm long. Leaves spatulate, 8-10 mm long, 4-5 mm wide, sparsely rough-hispid, glandular-punctuate, the very bases of the pet- ioles persisting as a series of rough scales on the old stems. Heads terminal, sessile or nearly so. Involucre broadly turbinate to nearly hemispheric, 6-7 mm high, 7-8 mm across, 3—4-seriate, the bracts glutinous, acute to narrowly obtuse. Receptacle convex, scaly-alveo- late. Ray florets ca. 13; tube ca. 2.5 mm long; ligule 4-5 mm long, 1.5-2.0 mm wide. Disk florets 30-50; corolla 5-6 mm long; tube ca. 2 mm long, gradually ampliate into a cylindrical throat 3-4 mm long, the lobes narrowly acute, ca. 1 mm long. Anther appendages narrowly acute. Style branches with linear, minutely puberulent appendages, ca. 2 mm long. Achenes 2.0—2.8 mm long, moderately pubescent throughout with stiffly appressed hairs; pappus of numerous white, persistent, bristles, 3-5 mm long. Type: Mexico, Coahuila, ca. 24 km e. of Saltillo, s. side of the Sierra de Viga, ca. 6.5 km e. of Jamé along wood cutters road. Grow- MADRONO, Vol. 29, No. 4, pp. 234-236, 12 October 1982 1982] TURNER AND LANGFORD: NEW SPECIES OF ERICAMERIA 235 Fic. 1. Sketch from holotype of Ericameria riskindii. Center, habit sketch; left, disc corolla (anthers not shown); right, ray corolla. ing in woodland with Pinus arizonica, Quercus greggii1, Pseudotsuga, Arbutus, etc., “10,000 ft.,” 15 May 1977, James Henrickson et al. 16156b (Holotype: TEX; isotypes: MEXU, RSA). PARATYPE: Nuevo Leon, Galeana Distr., Santa Rita, 2100 m, 25 Apr 1981, G. B. Hinton 18192 (TEX). Ericameria riskindii is an exceptionally distinct species. Superfi- cially it resembles Jsocoma veneta (H.B.K.) Greene, but in all of its floral characters it relates to Ericameria. Within the latter genus, it most resembles FE. juarezensis (Moran) Urbatsch, having the general habit and leaf shape of that species. Their floral features are also similar, both possessing very elongate stylar appendages; but those of E. riskindii are merely puberulent, whereas those of EL. juarezensis are decidedly hispid-pubescent. The collectors of the holotype described the plants as ““Low-rounded shrublets, aromatic, flowers yellow, in exposed limestone areas.” It is a pleasure to name the species for one of its discoverers, Mr. David Riskind, Naturalist with the Texas State Parks System and avid stu- dent of the Texas-Mexico border flora. 236 MADRONO [Vol. 29 LITERATURE CITED URBATSCH, L. E. 1978. The Chihuahuan Desert species of Ericameria (Compositae: Astereae). Sida 7:298-303. (Received 3 Apr 1981; accepted 28 Sep 1981.) ANNOUNCEMENT FIRST INTERNATIONAL CONFERENCE ON THE BUFFALO GOURD Speakers will cover a broad range of topics centered on the development of Cucurbita foetidissima as an agricultural crop plant. This will be the first opportunity for botanists and agronomists to meet and exchange research ideas on cultivation and development of this plant, which shows promise of producing a number of agricultural commodities under dry-land and minimal irrigation regimes. The conference will be held in Sydney, Australia from 16-19 Jan 1983. Further information can be obtained from Dr. Allen Gathman, Dept. of Plant Sci., Univ. of Arizona, Tucson 85721, USA. MORPHOLOGICAL DIVERSITY AND TAXONOMY OF CALIFORNIA MESQUITES (PROSOPIS, LEGUMINOSAE) KHIDIR W. HILU Department of Biology, Virginia Polytechnic Institute and State University, Blacksburg 24061 STEVE BoypD Department of Botany and Plant Sciences, University of California, Riverside 92521 PETER FELKER College of Agriculture, Texas A&I, Kingsville 78363 ABSTRACT Mesquites in California extend in distribution over 480 km from the Mexican border to about 16 km north of Death Valley, and from Parker Dam east to Bakersfield. Prosopis populations are found at —60 m to 1090 m in areas where the mean maximum July temperature is 38°C and the annual rainfall is less than 20 cm. Twenty-six popu- lations belonging to P. velutina and P. glandulosa var. torreyana were examined. Prosopis velutina is very rare in California and has been reported from a few locations in Imperial, Riverside, and Kern Counties. Its present distribution is in proximity to the Butterfield stage route that once connected San Francisco and St. Louis. The pres- ence of P. velutina in California is probably due to human introduction. Prosopis glandulosa var. torreyana is the most common species of Prosopis in California. It possesses a considerable degree of morphological variability. Principal component and cluster analyses showed that most of the variation falls in general trends that are habitat- related and sometimes form a continuum along suspected routes of dispersal. Cluster analysis revealed possible morphological intergradation among populations of P. ve- lutina and sympatric populations of P. glandulosa var. torreyana. This is probably due to hybridization. The two taxa overlap in several morphological characteristics but could be separated on the basis of a few characters, particularly leaf pubescence. Cor- relation between morphological characters was examined in terms of their importance in the systematics of the genus. The taxonomically difficult genus Prosopis L. (Mimosoideae) com- prises some 44 shrubby and tree species, 40 of which are native to the New World (Burkart 1976a). The species are distributed in the arid and semiarid regions of Southwest Asia, Africa, and America. Five to seven species are found in the Southwestern United States (Isley 1972, Burkart 1976a); among these, P. velutina Wooton, P. glandu- losa var. torreyana (L. Benson) M. C. Johnst., and P. pubescens Benth. extend in distribution to California. The North American species have been studied by Standley (1922, 1926), Britton and Rose (1928), Benson (1941), Graham (1960), Johnston (1962), Isley (1972), and Simpson (in Solbrig et al. 1977). A considerable amount of morpho- logical diversity, ecotypic differentiation, and interspecific hybridiza- MADRONO, Vol. 29, No. 4, pp. 237-254, 12 October 1982 238 MADRONO [Vol. 29 CORRELATION —0-1 0:1 0:3 0:5 0-7 0-9 l NY CLUSTER AA wD _— <<~ 2 2 3 1] 1 = v0 11 = 2 3 3 1 1 1 i = 3 11 So 2 333311 Vi 1 1 a 2 33.°3 = 3 * 33 3 O_9 a 3 3 B07 U3 3°33 3 DISCRIMINANT FUNCTION 1 Fic. 4. Centroids (symbolized by asterisks) and dispersions of clusters 1—3 of P. glandulosa var. torreyana plotted on the basis of the two discriminant functions. exhibited some overlapping with the other two. The first function was effective in separating the three clusters, whereas the second function contributed mostly to the discrimination between cluster 3 and the other two. Pinna number was by far the most important discriminating character in the first function. Other characters such as petiole and pinna length and leaflet number were fairly important (Table 3). Func- tion 1 accounted for over 77% of the variance and consequently is heavily weighted. Leaflet apex was most important in the second func- tion, followed by relative number of leaves per node, leaflet number and pinna width. Predictability of group membership was very high with a few OTUs of clusters 1 and 2 reclassified with cluster 3 (Table 4). This confirms the distinctness of the three groups. Principal component analysis was most useful in segregating the morphological variations in P. glandulosa var. torreyana. The first three factors accounted for over half of the variance and, therefore, will be emphasized in the analysis. The percent of variance accounted for by the first three factors and the loadings of the various characters are summarized in Table 2. When factors 1 and 2 were used as axes for a scatter diagram, populations of P. glandulosa var. torreyana appeared in nine mostly overlapped groups (Fig. 3). These groups are: B—southern populations from San Diego, Riverside and Yuma (Ari- zona); C—Kern County populations; D—predominantly San Bernar- dino populations; E—populations from the Salton Sea area; F—Death 248 MADRONO [Vol. 29 TABLE 3. TEN Most IMPORTANT CHARACTERS USED IN DISCRIMINATING BE- TWEEN THE THREE GROUPS OF P. glandulosa VAR. torreyana ARRANGED IN A DE- SCENDING ORDER OF THEIR CONTRIBUTION (ABSOLUTE VALUES) IN THE FIRST FUNC- TION. Percents of variance in functions 1 and 2 were 73 and 27, respectively. Characters Func- tions PNN LFL PNL PNW PBP LTL NDR LTN FRL PBL FRW LTX 1 0:76 0:74 0.35 0:34 .0.33 “0.29 0:27 “0:21 10:20 -0:09 "0:04 0.01 Z 0.22 0.47 0.29 0.34 0.58 0.23 0.56 0.39 0.05 0.48 0.30 0.68 Valley and Saline Valley; J—populations that extend along the Colo- rado River from Palo Verde to Needles then north along the Nevada border through Death Valley to Saline Valley; K—San Bernardino and Inyo County populations; and L—Riverside County populations. Although the Kern County populations appeared as a fairly distinct group (C), the first factor revealed a trend of variation from this group to San Bernardino and Inyo populations (group K). The three most important characters for this axis are leaflet length, pinna width, and the average distance between the fifth and sixth leaflets. This trend seems to be associated with decreasing xeric conditions. The second factor of the principal component revealed a trend of variation from the southern populations (group B) to Kern County (group C) and east to San Bernardino populations (group D). This factor has also shown a south—north trend along the eastern side of the state (groups J, K, L, F) where the San Bernardino mountains do not present a physical barrier. The important characters for this axis are leaflet number per pinna, leaf length, and pinna length. Because the principal component uses a character-character corre- lation matrix, a phenogram representing character correlations was generated (Fig. 5). The computed 0.87 cophenetic value shows that this phenogram well represents the matrix and has but a minimum TABLE 4. PREDICTABILITY OF GROUP MEMBERSHIP OF 97 P. glandulosa VAR. tor- reyana COLLECTIONS. The three groups represent the clusters obtained from cluster analysis (Q-correlation). Predicted group membership Actual No. of groups cases 1 2 3 1 43 39 O 4 91% O% 9% Z 25 O 23 Z O% 92% 8% S 29 Z O a7 HILU ET AL.: TAXONOMY OF PROSOPIS 249 1982] Ma 4 1dd ddd N11 NNd om lad YON M11 qi) MNd 111 id INd 141 06 ‘UOIIIVS SPOUIIJ, 9Y} UT ST SfOqUIAS Ja}IeIeYD 0} AZY “APNs SITY} UI pasn siayVIeYD ST 9Y} JO BULII}SNID Z S v O€ NOILV1TI AAO) % S l ‘COL | S 09 | , S l- 250 MADRONO [Vol. 29 amount of distortion. Leaf length is strongly correlated with pinna length and should be merged. However, in cases where there is more than one pair of pinnae, the distance between each along the rachis should be considered as a separate character. Leaf and pinna length seem to show some correlation with petiole length. Leaflet length and pinna width are highly correlated and ought to be combined because the former is indicative of the latter. These two traits appear to be correlated with the distance between leaflets. Leaflet width is inde- pendent of leaflet length. Leaflet and petiole pubescence were clustered at 0.67 correlation which indicates good correlation. Some relation (although very weak) seems to exist between number of pinnae and leaflets. Width and length of the fruit were negatively but weakly correlated and consequently the two could be used simultaneously. Graham (1960) observed negative correlation between number of pin- nae and size of leaflets, which is depicted in the phenogram at a very low correlation level (0.11). Simpson (in Solbrig et al. 1977) indicated that the longest leaflets were the widest and that leaves with greatest number of pinnae also had the largest number of leaflets. Both cor- relations are evident in the phenogram, though not especially high. DISCUSSION Prosopis velutina has its main distribution in southern Arizona and Mexico. Shreve (1951) indicated that this taxon is an important species in the Arizona Upland and the Plains of Sonora, and is the dominant or codominant tree on the flood plains and well watered level areas in the foothills of Sonora. The species has been observed in California at a few sites including Bakersfield, Temecula, 19 km east of Julian, and north of Westmoreland. This stretch of distribution is in proximity to the Butterfield Stage route that once connected San Francisco and St. Louis. Therefore, we believe that the presence of P. velutina in California is most probably due to human introduction through this route. Mesquite and screwbean are useful plants and have been an important food source for southern California Indians (Castetter and Bell 1951, Barrow 1967, Felger 1977) because of their very high yield of nutritious pods (Felker 1979). The expansion of P. velutina into California is apparently beyond its northern geographic range as im- plied by its rare occurrence in the state. Although our sample size of P. velutina is small, the morphological data do not deviate greatly from the descriptions of the species else- where (Benson 1941, Johnston 1962, Burkart 1976b). Cluster analysis and principal component techniques have shown good, but not clear- cut separation of P. velutina from P. glandulosa var. torreyana. Based on Q-correlation, P. velutina appeared at a level that could not justify a specific rank (Fig. 1). However, when the analysis was based on differences (distance method), a better separation was achieved, but 1982] HILU ET AL.: TAXONOMY OF PROSOPIS 251 not without some OTUs from P. glandulosa var. torreyana. Distance technique is known to be more effective in displaying differences be- tween close neighbors (Sneath and Sokal 1973). The discriminant func- tion analysis showed significant differences between the two taxa but discrimination was based mainly on leaf pubescence. A considerable overlap between P. velutina and P. glandulosa var. torreyana in var- ious morphological characteristics is evident (Table 1). The first two factors of the principal component segregated P. velutina from P. glandulosa var. torreyana. Factors 1 and 3 divided the northern and southern populations of P. velutina and placed them closest to those of P. glandulosa var. torreyana. The question whether P. velutina and P. glandulosa var. torreyana belong to the same or different species could not be answered satisfactorily here because of our small sample of the former. Nevertheless, the study revealed high affinities between the two. The discriminant function analysis showed that discrimina- tion between P. velutina and P. glandulosa var. torreyana is based primarily on leaf pubescence, a quantitative and rather variable char- acter. Prosopis glandulosa var. torreyana seems to bridge the mor- phological gaps between P. glandulosa and P. velutina (Table 1). These findings are supported by the reported hybridization between the two species (Peacock and McMillan 1965, Hunziker et al. 1975) and the absence of differences in flavonoid compounds (Carman 1973). Isley (1972) stated “I have difficulty regarding P. velutina as more than a desert form of P. glandulosa.” Prosopis glandulosa is abundant in valleys and dry uplands from Mexico to southern Kansas and from Louisiana to California (Isley 1972). The range of its distribution in the United States is interrupted in Arizona and Texas (Nueces Co.) by P. velutina and P. laevigata (Humb. & Bonpl. ex Willd.) M. C. Johnst. (Johnston 1962, Isley 1972). The Pecos River divides the eastern var. glandulosa from the western var. torreyana. Occasional occurrence of the varieties across the line and intergradation in morphology have been reported (Benson 1941, Isley 1972). Morphological differences between the two varieties are summarized in Table 1. Although our study showed a wider range of variation in some of the morphological traits of P. glandulosa var. torreyana, the two varieties are recognizable despite some overlap (Table 1). The various accessions of P. glandulosa var. torreyana clustered in three groups. Cluster 1 included populations primarily from the Salton Sea, Whitewater Canyon, Saline Valley, and Death Valley. The north- ern and southern populations are connected by a strip along the Col- orado River and the eastern borders of California where the mountains do not present a physical barrier to introduction. This cluster corre- sponds with groups J and F of the principal component (Fig. 3). The Saline-Death Valley and Salton Sea areas have some environmental similarities. Their annual rainfall and maximum July temperatures 252 MADRONO [Vol. 29 are comparable. Both areas are drainage sinks of their surrounding mountains, their water tables are near or above the soil surface, and fast evaporation in the two locations results in accumulation of salts. The eastern range along the Colorado River seems to represent a route of introduction from the southern to the northern areas. The characters that contributed most to the variation in this trend are leaflet number and leaf and pinna length. Cluster 2 encompassed the populations of P. velutina and those of P. glandulosa var. torreyana that are sympatric in distribution with them. Discrimination between this cluster and cluster 1 was contrib- uted by function 1 of the discriminant function while both functions contributed equally in discriminating between clusters 2 and 3 (Fig. 4). The discrimination in functions 1 and 2 was based primarily on pinna number, leaf length, leaf apex, and petiole pubescence (Table 3). These characters also are important in separating P. velutina from P. glandulosa. The affinities between P. glandulosa var. torreyana and P. velutina as depicted by the cluster analysis and principal com- ponent techniques could imply some gene flow between the two. Pop- ulations of these taxa are sympatric. Benson (1941) and Johnston (1962) have also observed morphological intergradation between P. velutina and P. glandulosa in the southwestern United States. They attributed this to possible hybridization. Cluster 3 included populations from San Bernardino and Kern Coun- ties along with a few accessions from the other two clusters. The clustering of plants from San Bernardino and Kern Counties signifies an introduction of P. glandulosa var. torreyana from the eastern sec- tion of California along a route north of the Transverse Ranges to Kern County. This trend is also observed in the principal component in which Kern County plants (group C) intergrade into the San Ber- nardino plants (group D). Factor 2 of the principal component seems to be more effective than Factor 1 in showing this trend. The present study has clearly reflected the enormous amount of variation in P. glandulosa var. torreyana (Figs. 1-3). Both cluster and principal component analyses revealed the high magnitude of variation within populations and the tendency for interpopulation grouping. Numerical analysis showed that the variation exhibits some general trends that can be correlated with habitats and suspected routes of dispersal. Variability in long-lived perennial plants, particularly those occupying harsh and unpredictable environments, has a strong selec- tive advantage. Perennials maintain their variability by outcrossing, low development of sterility barriers, and high chromosome numbers (Grant 1958). This variability buffers the population from environ- mental shocks and helps avoid decimation. This is especially true of the arid-adapted, nitrogen-fixing trees of Prosopis (Peacock and McMillan 1965, Felker and Clark 1980). These plants colonize infertile 1982] HILU ET AL.: TAXONOMY OF PROSOPIS 253 soils and often form nearly monotypic stands in parts of their ranges (ca. 30 million hectares) in the United States (Parker and Martin 1952). Prosopis achieves high genetic variability in populations by means of self-incompatability (Simpson 1977) and high chromosome numbers, 2n = 28 in diploids (Hunziker et al. 1975). The outcrossing mechanism has facilitated hybridization at the intraspecific and interspecific levels (Hunziker et al. 1975, Burkart 1976a, Felker, Boyd, and Hilu unpubl. data) and contributed to the high degree of morphological variation. Numerical techniques proved to be particularly useful in understand- ing the trends of this variation. LITERATURE CITED Barrows, D. P. 1967. Ethnobotany of the Coahuilla Indians. Malki Museum Press, Banning, CA. BENSON, L. 1941. The mesquites and screw-beans of the United States. Amer. J. Bot. 28:748—-754. BRITTON, N. L. and J. N. ROSE. 1928. Mimosaceae. North Amer. FI. 23:1-194. BuURKART, A. 1940. Materiales para monografia de género Prosopis (Leguminosae). Darwiniana 4:57-128. 1976a. A monograph of the genus Prosopis (Leguminosae subfam. Mimo- soideae). J. Arnold Arb. 57:219-249. 1976b. A monograph of the genus Prosopis (Leguminosae subfam. Mimo- soideae). J. Arnold Arb. 57:450—475. and B. B. SIMPSON. 1977. The genus Prosopis and annotated key to the species of the world. Jn B. B. Simpson, ed., Mesquite—its biology in two desert ecosystems, p. 201-215. Dowden, Hutchinson and Ross, Inc., Stroudsburg, PA. CARMAN, N. J. 1973. Systematic and ecological investigations in the genus Prosopis (Mimosaceae) emphasizing the natural products chemistry. Ph.D. dissertation, Univ. Texas, Austin. CASTETTER, E. F. and W. H. BELL. 1951. Yuman Indian agriculture. Primitive subsistence on the lower Colorado and Gila Rivers. Univ. New Mexico Press, Albuquerque. FELGER, R. S. 1977. Mesquite in Indian culture of southwestern North America. Jn B. B. Simpson, ed., Mesquite—its biology in two desert ecosystems, p. 150-176. Dowden, Hutchinson and Ross, Inc., Stroudsburg, PA. FELKER, P. 1979. Mesquite: an all-purpose leguminous arid land tree. Jn G. A. Ritchie, ed., New agricultural crops, p. 89-132. Westview Press, Boulder, CO. and P. R. CLARK. 1980. Nitrogen fixation, acetylene reductions and cross- inoculation in 12 Prosopis mesquite species. Pl. and Soil 57:177-186. GRAHAM, J. D. 1960. Morphological variation in mesquite (Prosopis, Leguminosae) in the lowlands of northeastern Mexico. Southw. Naturalist 5:187-193. GRANT, V. 1958. The regulation of recombination. /n Exchange of genetic material: mechanisms and consequences. M. Demerec, ed., Cold Spring Harbor Symp. Quant. Biol. XXIII, p. 337-363. GRAY, A. 1852. Leguminosae. Plantae Wrightianae 1:43-67. HUNZIKER, J. H., L. Poccio, C. A. NARANJO, R. A. PALACIOS, and A. B. ANDRODA. 1975. Cytogenetics of some species and natural hybrids in Prosopis (Leguminosae). Canad. J. Genet. Cytol. 17:253-262. HUTCHINSON, J. 1964. The genera of flowering plants, vol. 1. Clarendon Press, Oxford. 254 MADRONO [Vol. 29 IsLEY, D. 1972. Legumes of the U.S., VI. Calliandra, Pithecellobium, Prosopis. Madrono 21:273-298. JOHNSTON, M. C. 1962. The North American mesquites Prosopis sect. Algarobia (Leguminosae). Brittonia 14:77-89. Munz, P. A. 1973. A California flora and supplement. Univ. California Press, Berke- ley. Nig, N. H., C. H. HULL, K. STEINBRENNER, and D. H. BENT. 1970. Statistical package for the social sciences. McGraw Hill Co., NY. PARKER, K. W. and S. G. MARTIN. 1952. The mesquite problem on the southern Arizona range. USDA, Circ. 968. . PEACOCK, I. J. and C. MCMILLAN. 1965. Ecotypic differentiation in Prosopis (mes- quite). Ecology 46:35-51. ROHLF, F. J., J. L. P. KISHPAUGH, and D. KIRK. 1976. NT-SYS: numerical tax- onomy system of multivariate statistical programs. The State Univ. New York, Stony Brook. SARGENT, C. S. 1902. Prosopis juliflora. The Silva North Amer. 13 (suppl.):15—16. SHREVE, F. 1951. Vegetation and flora of the Sonoran Desert. Vol. 1. Vegetation of the Sonoran Desert. Carnegie Inst. Wash. Publ. 591:1-192. SIMPSON, B. B. 1977. Breeding systems of dominant perennial plants of two warm disjunct desert ecosystems. Oecologia 27:203—226. SNEATH, P. H. and R. R. SOKAL. 1973. Numerical taxonomy: the principles and practice of numerical taxonomy. W. H. Freeman, San Francisco. SOLBRIG, O. T., K. BAwa, N. J. CARMAN, J. H. HUNZIKER, C. A. NARANJO, R. A. PALACIOS, L. POGGIO, and B. B. SIMPSON. 1977. Pattern of variation. Jn B. B. Simpson, ed., Mesquite—its biology in two desert ecosystems, p. 44-60. Dowden, Hutchinson and Ross, Inc., Stroudsburg, PA. STANDLEY, P. C. 1922. Trees and shrubs of Mexico. Prosopis. Contr. U.S. Natl. Herb. 23:1920-1926. 1926. Trees and shrubs of Mexico. Prosopis. Contr. U.S. Natl. Herb. 23: 350-353. U.S. DEPT. COMMERCE. 1922. Climate of the world. Governm. Publ. Office, Wash- ington, D.C. (Received 7 Jul 1981; revision accepted 21 Dec 1981.) SYMPLOCOS SOUSAE, A NEW SPECIES OF SYMPLOCACEAE FROM MEXICO FRANK ALMEDA Department of Botany, California Academy of Sciences, San Francisco 94118 ABSTRACT Symplocos sousae, distinguished by solitary, axillary, 5-merous flowers, persistent floral bracts, and glabrous drupaceous fruits, is described from the Sierra Manantlan in Jalisco and the region of San Andres Chicahuaxtla in Oaxaca, Mexico. Its closest affinities are with S. coccinea Humb. & Bonpl. and S. prionophylla Hemsl., both of which are also endemic to Mexico. Ongoing exploration of montane forest vegetation in areas of west- ern and southern Mexico continues to yield collections of new woody taxa exhibiting restricted or patchy distributions. Symplocos, with ap- proximately 15 species in Mexico, is sometimes locally abundant in these upland forests, yet many of the species remain poorly under- stood. A persistent problem in clarifying the taxonomy of Symplocos, even on a regional basis, lies in the difficulty of correlating specimens collected in either flowering or fruiting condition. Although material of the new species described herein has slowly accumulated in herbaria since 1949, an understanding of its diagnostic features and probable relationships emerged only after studying the flowering and fruiting specimens recently collected by Sousa in Oaxaca. In recognition of his contribution to this study I take pleasure in naming the species for Mario Sousa Sanchez who has generously assisted many botanical colleagues in their investigations of the Mexican flora. Symplocos sousae Almeda, sp. nov. Frutex vel arbor parva (1—)3—7 m. Folia coriacea, integra vel ser- rulata, petiolata, elliptica, elliptico-ovata vel elliptico-obovata, ad ba- sem rotundata vel obtusa vel acuta, ad apicem acuta vel acuminata. Lamina 5—11 cm longa et 2.3—5.8 cm lata supra glabra, subtus strigosa vel strigillosa. Flores 5-meri sessiles vel subsessiles in foliorum supe- riorum axillis solitarii; calyx 3.5-5 mm longus, lobis ovatis, ciliatis; corolla rosea (fide collectore), campanulata, 1.2—1.6 cm longa, glabra, lobis oblongis vel obovatis, 6-10 mm latis. Stamina multiseriata. Sty- lus glaber, 7-11 mm longus. Fructus glaber, obovoideus vel ellipsoide- us, 16-24 mm longus, 12-19 mm latus, 4-locularis (Fig. 1). Shrubs reportedly 1—3 m tall or trees mostly 5—7 m. Juvenile branch- lets and vegetative buds mostly sericeous to strigose but varying to MADRONO, Vol. 29, No. 4, pp. 255-258, 12 October 1982 256 MADRONO [Vol. 29 , ‘ , ¢ f Lech O18 eyo s Z a ues ORE Ge tse tage Ce ee a 88) Fic. 1. Symplocos sousae Almeda. A, habit; B, representative leaf (abaxial surface); C, calyx lobes, style, and stigma; D, floral dissection (with two corolla lobes removed) showing floral bracts, calyx lobes, multiseriate staminal arrangement, and corolla lobes; E, cross section of mature drupe; F, mature drupaceous fruits with persistent calyx lobes. (A—D from Boutin & Brandt 2562, CAS; E from McVaugh et al. 23156, MICH; F from Sousa et al. 5163, CAS.) All scale lines equal 1 cm. glabrate. Petioles canaliculate above, 3-8 mm long and 1.5-—2.0 mm broad. Principal leaves coriaceous, entire but varying to distally ser- rulate, elliptic to elliptic-ovate or elliptic-obovate, 5-11 cm long and 2.3-5.8 cm broad, apically acute to acuminate, basally rounded but sometimes varying to obtuse or acute, glabrous and + vernicose above, densely to sparsely strigose or strigillose below with only the median nerve prominently elevated. Flowers sessile or subsessile, solitary, erect, borne in the leaf axils of distal branchlets and closely subtended by 1982] ALMEDA: NEW SYMPLOCOS FROM MEXICO 200 6—7 sessile, imbricate, persistent, glabrous bracts; lowermost bracts suborbicular to reniform, apically rounded, 1.5—2.5 mm long and 1.5— 3 mm broad, margins mostly ciliate but sometimes entire and + hya- line, the uppermost bracts broadly deltoid, apically obtuse, 3-4.5 mm long and 3.5—5 mm broad, the margins prevailingly ciliate. Calyx 5-lobed, the lobes + imbricate, ovate, 3.5-5 mm long and 3—4.5 mm wide at the base, glabrous, margins entire, ciliate, and auriculate ba- sally between sinuses. Corolla sympetalous, glabrous, + campanulate, 1.2-1.6 cm long, 5-lobed, reportedly pink but drying red; lobes con- nate basally for 2-3 mm and adnate to the filament tube for 4-6 mm basally, oblong to obovate, apically rounded, 6-10 mm wide, the mar- gins erose-ciliolate and markedly involute. Stamens multiseriate; fil- aments connate basally for 5-7 mm, free portions of the filaments ligulate, 3-6 mm long and 0.5—1.0 mm wide, often separating into clusters opposing the corolla lobes. Anthers bilocular, + globose to quadrate, mostly 1 mm long and wide, white to pale yellow. Ovary inferior, glabrous, the summit strigillose to nearly glabrous. Styles straight, glabrous, 7-11 mm long; stigma subcapitate, deeply 5-lobed. Fruits drupaceous, glabrous, obovoid to ellipsoid, 16-24 mm long, 12-19 mm broad, quadrilocular in cross section with a massive mar- ginally repand woody endocarp. TYPE: México, Jalisco, se. of El Chante and Aserradero along road near El Guisar, an abandoned lumber mill, 2743 m, 24 Nov 1968, Boutin & Brandt 2562 (Holotype: CAS!; isotypes: HNT! MEXU!). PARATYPES: Mexico, Jalisco: Sierra de Manantlan, 29 Jan 1970, Boutin & Kimnach 2979 (HNT— sheets); Sierra de Manantlan, w. of El Guisar, 30 Jan 1970, Boutin & Kimnach 2998 (HNT); Sierra de Manantlan (25-30 km se. of Autlan), along lumber roads e. of the road crossing called La Cumbre between El Chante and Cuzalapa, 20-21 Mar 1965, McVaugh et al. 23156 (MICH); hardwood-pine-fir forest in mountains of Manantlan, ca. 15 miles sse. of Autlan by way of Chante, 28 Jul 1949, R. L. & C. R. Wilbur 1936 (MICH). Oaxaca: San Andres Chicahuaxtla, Distr. de Putla, Dec 1966, MacDougall s.n. (CAS—2 sheets, MEXU); Cerro Zarzamora, San Andres Chicahuax- tla, 15 Apr 1962, MacDougall s.n. (CAS, MEXU, MICH); 1 km n. of Chicahuaxtla, municipio of Chicahuaxtla, 8 Feb 1976, Sousa et al. 5163 (CAS, MEX U— sheets). Distribution. The species is endemic to Mexico, where it reportedly occurs along streams and steep ravines on the Sierra de Manantlan of Jalisco at 2500-2750 m in forests with Abies, Cupressus, Pinus, and Quercus. It is also known from forested areas in the vicinity of San Andres Chicahuaxtla and Cerro Zarzamora in Oaxaca at 2490 m. These sites very likely represent collecting localities readily accessible from roads. Additional populations are to be expected in intervening montane regions of Michoacan and Guerrero. 258 MADRONO [Vol. 29 The salient features of S. sousae include sessile, or subsessile, sol- itary, axillary flowers; persistent, glabrous floral bracts; basally auric- ulate, ovate calyx lobes; 5-parted, glabrous corollas with involute, apically rounded lobes; and obovoid to ellipsoid fruits that are glabrous at maturity. The two known populations of this species exhibit some consistent morphological differences worthy of note. Collections from Jalisco typically have elliptic to elliptic-ovate entire leaves that are acute to acuminate apically and the summit of the ovary surrounding the style is essentially glabrous. In contrast, the Oaxaca specimens have elliptic-obovate, distally serrulate leaves that are short-acuminate apically and the ovary summit is invariably strigillose. Additional col- lections would be desirable to determine the extent of variability in these characters. I am inclined to attach little importance to these differences in view of the distance separating known populations. A peculiar feature of some collections of this species from both Jalisco and Oaxaca is the tendency of the ovary summit to become enlarged and distended beyond the persistent calyx lobes on mature fruits. The significance of these structures is unclear for they do not appear to be characteristic of all mature fruits examined. Among described congeners, S. sousae most closely resembles S. coccinea Humb. & Bonpl. and S. prionophylla Hemsl., both of which are also known only from Mexico (Brand 1901, Standley 1924). Foliar shape and the solitary, sessile flowers give S. sousae an aspect remi- niscent of S. coccinea but the latter differs in having hirtellous or hirsute distal branchlets, sericeous bracts and calyx lobes, flowers with 10-15 apically cuspidate to acuminate corolla lobes, basally pilose styles, and oblong to ellipsoid fruits that are moderately to copiously hirsute. To a lesser degree S. sousae also resembles S. prionophylla by virtue of the 5-parted corolla and elliptic or oblong-obovate leaves. Symplocos prionophylla is otherwise sharply differentiated by its uni- formly serrulate leaves, early deciduous floral bracts, densely canes- cent bracts and calyx lobes, paniculate or fascicled inflorescences of 3-5 flowers borne on short peduncles (mostly 5 mm or less), and nar- rowly cylindric, strigillose fruits. According to label information on Sousa et al. 5163 the new species is commonly known as “tunihia” or “tu-nihia” in Oaxaca, and the fruits are eaten upon turning black at maturity. ACKNOWLEDGMENTS I thank Terry Bell for preparing the line drawings and curators of the following herbaria for special loans of critical material: HNT, MEXU, MICH. LITERATURE CITED BRAND, A. 1901. Symplocaceae. Das Pflanzenreich IV. 242(Heft 6):1—100. STANDLEY, P. C. 1924. Trees and shrubs of Mexico. Contrib. U.S. Natl. Herb. 23(4): 849-1312. (Received 24 Oct 1981; accepted 21 Dec 1981.) A SURVEY OF THE CORTICOLOUS MYXOMYCETES OF CALIFORNIA KENNETH D. WHITNEY Department of Botany, University of North Carolina, Chapel Hill 27514 ABSTRACT Thirty-eight species of corticolous myxomycetes are reported from California, most of which were collected from tree bark placed in moist chambers. Included in this number are several recently described species. The substrates and California counties where each taxon was obtained are listed, along with notes on distinguishing features of each species. The corticolous myxomycetes are a diverse group inhabiting the bark surface of living woody plants. Most have minute fruiting bodies that are very difficult to detect in the field. The bark surface of living plants as a myxomycete habitat was rarely investigated until Gilbert and Martin (1933) used moist-chamber culture techniques with sam- ples of tree bark. These methods allow close observation of bark under conditions favorable to the development of myxomycete fruiting bod- ies. More recently, Keller and Brooks (1973, 1975, 1977) have made thorough collections of corticolous myxomycetes using a hand lens in the field. The myxomycetes they collected as natural fruitings follow- ing rainfall are identical to ones obtained in moist chambers, and the concept of a distinct corticolous myxomycete flora has become well recognized. The myxomycete flora of California, including several corticolous species, has been extensively studied (Kowalski 1966, 1967, 1973; Ko- walski and Curtis 1968, 1970). Most of the smaller corticolous species listed by Kowalski were obtained by collecting bark samples bearing more conspicuous corticolous myxomycetes and examining the sub- strate in the laboratory. The present study was undertaken to expand the corticolous myxomycete flora of California using moist-chamber culture techniques. Although bark samples were obtained throughout California, Butte and Lassen Counties were emphasized. A total of 38 myxomycete species were obtained in moist chambers, including seven new state records and six new species, which have been de- scribed elsewhere. Kowalski (1973) gives the number of myxomycete species reported from California as 231, about half the species recognized by Martin and Alexopoulos (1969) in their world monograph. Thus, the 38 species MADRONO, Vol. 29, No. 4, pp. 259-268, 12 October 1982 260 MADRONO [Vol. 29 reported here on the bark surface of living plants constitute a signif- icant portion of the state flora. Additionally, many areas of California have yet to be investigated; undoubtedly many other unreported or undescribed corticolous myxomycetes await discovery. MATERIALS AND METHODS In preparing moist chambers, Petri dishes were lined with filter paper. A sample of bark was placed on the paper, then the dish was flooded with distilled water. After 2—3 hours the excess water was poured off, and the moist chambers were incubated at room temper- ature or 12—-15° C. The lower temperature has been suggested as op- timum for certain corticolous species (Whitney 1980). Observations of the bark samples with a stereoscopic microscope were started within 24 hours. Myxomycete fruiting bodies can be expected to develop on the bark over a period of 4-6 weeks. Permanent collections of the fruiting bodies of larger, sturdier species were dried on the bark and mounted in small paper boxes. Smaller, more delicate specimens were picked off the substrate with fine jeweler’s forceps and mounted in lactophenol solution on a microscope slide. These slide mounts were made more or less permanent by sealing the coverslip with clear fin- gernail polish. Collections of each taxon listed are deposited in the Herbarium of the University of California, Berkeley (UC). Nomenclature follows that of Martin and Alexopoulos (1969), except where noted. LICEALES Liceaceae Licea biforis Morgan Substrates: Arbutus menziesii Pursh, Populus fremontii Wats., Quercus sp., Q. suber L., Vitis californica Benth. Distribution: Butte Co., Marin Co. Probably cosmopolitan. This species is common and often forms extensive fruitings on bark. It is easily overlooked, however, because the sporangia are dark at maturity and usually blend in well with the bark surface. The fruiting bodies of L. biforis are elongate, sessile, and open by an apical, lon- gitudinal slit. Licea castanea G. Lister Substrates: Juglans sp., Quercus sp. Distribution: Butte Co., Los Angeles Co. (Santa Catalina Island). Throughout Europe and North America. Licea castanea has chestnut-brown sporangia and is common but inconspicuous. The peridium of this species is divided into platelets by preformed lines of dehiscence. At maturity, the platelets reflex to allow spore escape. 1982] WHITNEY: CALIFORNIA MYXOMYCETES 261 Licea kleistobolus Martin Substrates: Juniperus occidentalis Hook., Platanus racemosa Nutt., Vitis californica. Distribution: Butte Co., Lassen Co. North America, Europe. This is an extremely common corticolous species with operculate sporangia. The operculum is copper-colored and has a convex center, whereas the rest of the fruiting body is dull brown to black. It can be readily obtained on grapevine bark, and often produces hundreds of sporangia per moist chamber. Licea parasitica (Zukal) Martin Substrates: Castanea dentata L., Chamaecyparis lawsoniana (A. Murr.) Parl., Cupressus macnabiana A. Murr., Pseudotsuga menziesii (Mirb.) Franco, Quercus lobata Nee, Q. suber, Umbellularia califor- nica (H. & A.) Nutt., Vitis californica. Distribution: Butte, Ma- rin, and Sonoma Cos. Throughout Europe and North America. This is perhaps the most commonly encountered Licea in California. It is operculate, with the sporangium wall and lid dull black in color. Freshly matured sporangia in moist chambers show a distinct pale line circumscribing the lid, but after drying the sporangia shrivel and the operculum becomes less distinct. Licea pedicellata (H. C. Gilbert) H. C. Gilbert Substrates: Casuarina sp., Quercus lobata. Distribution: Butte Co., Los Angeles Co. (Santa Catalina Island). Known from Europe and North America. This species has stipitate sporangia with the peridium divided into several platelets. These divisions are best seen in fresh sporangia be- cause the sporangia take on a wrinkled appearance when dry. Licea perexigua Brooks & Keller, in Keller and Brooks, Mycologia 69: 674. 1977. Substrates: Quercus douglasii H. & A., Q. lobata, Quercus sp., Umbellularia californica. Distribution: Butte, Marin, and So- noma Cos. Known from Arkansas, Kansas, Kentucky, and Missouri. Probably occurs throughout North America. This species is very small, less than 100 um in height, and is some- what difficult to detect. The sporangia are subsessile to short stipitate, and are metallic gray in color. This minute myxomycete often fruits abundantly in moist-chamber culture. Licea pusilla Schrad. Substrate: Juniperus occidentalis. Distribution: Lassen Co. Known from Europe and North America. A single collection of this species was obtained on bark, but it is probably more common than this appears to indicate. Kowalski (1966) reported this species from Butte County on decayed wood, a common substrate for L. pusilla. This species is similar to L. castanea in that 262 MADRONO [Vol. 29 both have peridia broken into platelets and have sessile sporangia of similar sizes. In L. pusilla the sporangia are purple-brown to black, and the spores are dark olive-brown. Licea castanea has pale brown sporangia with pale yellow-brown spores. Licea scyphoides Brooks & Keller, in Keller and Brooks, Mycologia 69:679. 1977. Substrates: Quercus lobata, Q. suber, Quercus sp., Salix sp., Vitis californica. Distribution: Butte, Los Angeles (Santa Catalina Is- land), Marin, and Sacramento Cos. Known from several locations in the United States. A common corticolous species in California, L. scyphoides has dull black, stipitate sporangia with a more or less equatorial line of dehis- cence. After drying the sporangia often dehisce, leaving a cup-like portion of the peridium attached to the stipe. This is a new California record. Cribrariaceae Cribraria violacea Rex Substrates: Juglans sp., Platanus racemosa, Quercus lobata, Vitis californica. Distribution: Butte, Santa Barbara, and Tehama Cos. Cosmopolitan. The long-stipitate, dark purple sporangia of this species are easily recognized and should be confused with no other corticolous myxo- mycete. It is fairly common although it did not fruit in great abun- dance during this study. Bark collected in riparian areas frequently yields this species when placed in moist chambers. ECHINOSTELIALES Echinosteliaceae Echinostelium apitectum Whitney, Mycologia 72:954. 1980. Substrates: Juniperus occidentalis, Quercus sp. Distribution: Lassen Co. Reported from Australia (D. W. Mitchell, pers. comm.). This species is common on juniper bark in Lassen County, but it has not been found elsewhere in California. It is distinguished by the presence of a minute columella borne beneath a sporelike covering at the stipe apex. Echinostelium brooksit Whitney, Mycologia 72:957. 1980. Substrates: Cupressus macnabiana, Cercocarpus ledifolius Nutt., Juniperus occidentalis, Pinus sp. Distribution: Butte Co., Las- sen Co. Known from North America and Europe. The darkly pigmented, lenticular columella and large pink spores with a thin area in the wall distinguish this species. Echinostelium brooksii is especially common in Lassen County on juniper bark. Often 1982] WHITNEY: CALIFORNIA MYXOMYCETES 263 hundreds of sporangia occur on a single piece of bark in moist-chamber culture. Echinostelium coelocephalum Brooks & Keller, in Keller and Brooks, Mycologia 68:1212. 1976. Substrates: Acer negundo L., Quercus lobata. Distribution: Butte Co. Known from several locations in the United States. The white to cream-colored sporangia of this species are usually less than 70 pm in height and are quite difficult to detect in moist cham- bers. This species has a sporelike columella and spores with distinct spore-to-spore articular surfaces. It was collected from a single location in Butte County (Pine Creek Ranch, 23 km northwest of Chico), but it probably occurs on similar substrates elsewhere in the state. Echinostelium colliculosum Whitney & Keller, Mycologia 72:641. 1980. Substrates: Juniperus occidentalis, Quercus lobata, Vitis californi- ca. Distribution: Butte, Lassen, and Sacramento Cos. Known from Europe and North America. Reported from Australia (D. W. Mitchell, pers. comm.). This species resembles EL. coelocephalum, but E. colliculosum has larger sporangia and spores, and less prominent spore articular sur- faces. Echinostelium corynophorum Whitney, Mycologia 72:963. 1980. Substrate: Juniperus occidentalis. Distribution: Lassen Co. North Carolina. Reported from Australia (D. W. Mitchell, pers. comm.). This species has been found on juniper bark in Lassen County and is only rarely encountered in California. It is distinguished by a hyaline to pale yellow-brown, hemispheric columella, and white spores bear- ing articular surfaces at points of spore-to-spore contact. Echinostelium fragile Nann.-Brem. Substrates: Artemisia tridentata Nutt., Cupressus sp., Juniperus occidentalis, Quercus lobata. Distribution: Lassen, Los Angeles (Santa Catalina Island), and Sacramento Cos. Known from Europe and North America. This is another common species on juniper bark in Lassen County, and it is occasionally found in other areas of the state. Echinostelium fragile has a fusiform, darkly pigmented columella and pink spores with a thin region in the spore wall. Echinostelium lunatum Olive & Stoianovitch, Mycologia 63:1050. 1971. Substrates: Cupressus macnabiana, Vitis californica. Distri- bution: Butte Co., Sacramento Co. North Carolina. Puerto Rico. This is one of the smallest known myxomycetes, with sporangia usu- ally less than 50 wm in height. The sporangia are easily overlooked in moist chambers, and this probably accounts for the apparent rarity of 264 MADRONO [Vol. 29 this species. It is recognized easily by its small size and the tan, cres- cent-shaped columella that supports the spores. Echinostelium minutum de Bary Substrates: Juniperus occidentalis, Quercus lobata, Pseudotsuga menziesit. Distribution: Throughout California. Cosmopolitan. Echinostelium minutum is fairly common in California, especially on the bark of Douglas fir. This species is the tallest member of the genus, with sporangia up to 500 um in height. The spore mass is white to pinkish, and a delicate branching capillitium arises from a short columella. Echinostelium paucifilum Whitney, Mycologia 72:974. 1980. Substrate: Juniperus occidentalis. Distribution: Known only from Lassen Co. This species is characterized by large, pink spores, an elongate, dark brown columella, and a sparse capillitial system. It is one of the largest Echinostelium species, with sporangia up to 350 mum in height. Clastodermataceae Alexopoulos & Brooks, Mycologia 63:926. 1971. Clastoderma pachypus Nann.-Brem., Verh. Kon. Ned. Akad. We- tensch., Afd. Natuurk., C Sect. 71:44. 1968. Substrate: Cupressus macnabiana. Distribution: Butte Co. Known from Europe and North America. This species occurs frequently on bark of Macnab cypress, a ser- pentine endemic. The chocolate brown sporangia and the capillitium forming a globose net distinguish this species. This is a new California record. TRICHIALES Dianemaceae Calomyxa metallica (Berk.) Nieuwl. Substrates: Castanea dentata, Platanus racemosa, Populus fremon- tit. Distribution: Butte Co., Sacramento Co. Cosmopolitan. This species can be found on dead wood as well as bark. The fruit- ings on bark in moist chambers usually consist of widely scattered sporangia, in contrast to collections on dead wood, which normally occur as aggregates of sporangia. Sporangia of C. metallica are sessile and iridescent purple when freshly matured; the spores are bright yellow. Trichiaceae Perichaena chrysosperma (Currey) A. Lister Substrate: Quercus sp. Distribution: Los Angeles Co. (Santa Catalina Island). Cosmopolitan. 1982] WHITNEY: CALIFORNIA MYXOMYCETES 265 Perichaena corticalis (Batsch) Rost. Substrates: Juglans sp., Quercus lobata, Vitis californi- ca. Distribution: Butte Co., Tehama Co. Cosmopolitan. These two species of Perichaena are very similar and somewhat difficult to separate. Both have reddish brown to black peridia and bright yellow spores. Perichaena chrysosperma tends to produce scat- tered plasmodiocarps, whereas P. corticalis usually forms clustered sporangia. Hemitrichia abietina (Wigand) G. Lister Substrates: Cupressus macnabiana, Pseudotsuga menzie- Sil. Distribution: Butte Co. Probably cosmopolitan. This species normally occurs on dead wood, but it is not uncommon on bark. Hemitrichia abietina has a thin, iridescent peridium that falls away in the upper portion of the sporangia at maturity, exposing the yellow spore mass and leaving behind a distinct cup. STEMONITALES Stemonitaceae Enerthenema papillatum (Pers.) Rost. Substrate: Juniperus occidentalis. Distribution: Lassen Co. Cosmopolitan. This species occurs on bark and decayed wood but is more fre- quently encountered on the latter substrate. The black stipitate spo- rangia with a shining apical disc distinguish this species. Macbrideola cornea (G. Lister) Alexop. Substrates: Common on a wide variety of woody plants. Dis- tribution: Butte, Marin, Monterey, Sacramento, Santa Barbara, and Sonoma Cos. Probably cosmopolitan. Macbrideola decapillata H. C. Gilbert Substrate: Juniperus occidentalis. Distribution: Lassen Co. Probably cosmopolitan. Macbrideola is a common genus on bark in moist chambers, occur- ring as scattered, brown, stipitate sporangia. The two listed species are similar in appearance, and can be distinguished only by micro- scopic characteristics. Macbrideola cornea has a rigid capillitium, and the tips of the capillitial threads are distinctly blunt. Macbrideola decapillata has a sparse to absent capillitial system that, when present, has sharply pointed free ends. Lamproderma arcyrionema Rost. Substrates: Cupressus macnabiana, Vitis californica. Distri- bution: Butte Co. Cosmopolitan. This is the most common species of Lamproderma on bark in Cal- ifornia. The sporangia have long, slender stalks and a persistent, iri- descent peridium. 266 MADRONO [Vol. 29 Comatricha acanthodes Alexop. Substrates: Castanea dentata, Catalpa speciosa, Quercus sp. Distribution: Butte Co., Sonoma Co. Known from Virginia and Greece. This species seems to be fairly rare in California. It can be distin- guished from other Comatricha species by the rigid, sparse capillitium and the distinctly spiny spores. This is a new California record. Comatricha fimbriata G. Lister & Cran Substrates: Common on a wide variety of tree species. Distri- bution: Throughout California. Cosmopolitan. Comatricha fimbriata is a commonly encountered corticolous myxo- mycete. The major distinguishing feature is the swollen tips of the capillitial threads. Comatricha laxa Rost. Substrates: Populus fremonti1, Yucca brevifolia Engelm. Dis- tribution: Butte Co., Riverside Co. Probably cosmopolitan. Comatricha laxa is commonly found on dead wood, but it is seen occasionally on bark in moist chambers. This species has ovoid to cylindric, reddish brown sporangia and brown spores. PHYSARALES Physaraceae Badhamiopsis ainoae (Yama.) Brooks & Keller, in Keller and Brooks, Mycologia 68:836. 1976. Substrates: Juglans sp., Quercus sp., Quercus lobata. Distri- bution: Butte Co., Sonoma Co. Known from Japan and North America. Badhamiopsis ainoae is characterized by effused plasmodiocarps and sporangia, granular lime deposits on the upper surface of the peridium, and calcareous spikes running between the upper and lower peridia. It is common in California and often forms large fruitings on bark in moist chambers. Badhamia affinis Rost. Substrate: Quercus lobata. Distribution: Butte Co., Marin Co. Cosmopolitan. This is one of the few corticolous myxomycetes easily collected in the field. Badhamia affinis often forms extensive fruitings on living trees, the large, white, stipitate fruiting bodies often covering the trunk and extending well into the branches. Badhamia bispora Whitney, Mycologia 70:672. 1978. Substrate: Juniperus occidentalis. Distribution: Known only from Lassen Co. 1982] WHITNEY: CALIFORNIA MYXOMYCETES 267 Badhamia nitens Berk. Substrates: Cupressus macnabiana, Quercus lobata. Distribu- tion: Butte Co. Probably cosmopolitan. Badhamia bispora and B. nitens are similar in most charateristics and can be distinguished best using spore morphology. Badhamia nitens has spores in clusters of four to twenty. Badhamia bispora has spores that fuse together in pairs. Physarum crateriforme Petch Substrates: Common on a wide variety of tree species. Dis- tribution: Common throughout California. Probably cosmopolitan. This species has stipitate sporangia with white, limy peridia. A large, central lime-filled columella is usually present and is distinctive. This species is fairly easy to obtain in the field due to its relatively large, stipitate sporangia. Physarum decipiens Curtis Substrate: Quercus lobata. Distribution: Butte Co. Probably cosmopolitan. This species is characterized by its yellowish to orange, calcareous peridium and its plasmodiocarpous habit. It is not common on bark in California. Didymiaceae Trabrooksia applanata Keller, Mycologia 72:396. 1980. Substrates: Juglans sp., Quercus lobata. Distribution: Butte Co., Marin Co. Known from Europe and North America. The flattened plasmodiocarps of this species are limeless and some- what iridescent. They blend in well with the bark surface and are easily overlooked. It is common in California. Diderma chondrioderma (de Bary & Rost.) G. Lister Substrates: Diospyros sp., Juglans sp., Quercus suber. Distri- bution: Butte Co. Probably cosmopolitan. Diderma chondrioderma is seen occasionally on bark in California. It can be recognized readily by the flattened white sporangia and the pinkish columella. The granular lime deposits on the peridium are often quite thick, giving the sporangia a distinctly chalky appearance. ACKNOWLEDGMENTS I would like to thank Dr. D. T. Kowalski for suggesting this project and his help throughout the course of this study. I would also like to thank Dr. Meredith Blackwell for kindly reviewing the manuscript. This study was supported in part by a grant from the National Science Foundation (SMI 76-83818). 268 MADRONO [Vol. 29 LITERATURE CITED GILBERT, H. C. and G. W. MARTIN. 1933. Myxomycetes found on the bark of living trees. Univ. Iowa Stud. Nat. Hist. 16:153-159. KELLER, H. W. and T. E. Brooks. 1973. Corticolous Myxomycetes I: Two new species of Didymium. Mycologia 65:286—294. and 1975. Corticolous Myxomycetes III: a new species of Badhamia. Mycologia 67:1218-1222. and 1977. Corticolous Myxomycetes VII: contribution toward a mono- graph of Licea, five new species. Mycologia 69:667—-684. KOWALSKI, D. T. 1966. New records of Myxomycetes from California I. Madrono 18:140—142. 1967. New records of Myxomycetes from California II. Madrono 19:43—46. ——. 1973. New records of Myxomycetes from California V. Madrono 22:97—100. — and D. H. CurTIs. 1968. New records of Myxomycetes from California ITI. Madrono 19:246-249. and 1970. New records of Myxomycetes from California IV. Madrono 20:377-381. MARTIN, G. W. and C. J. ALEXOPOULOS. 1969. The Myxomycetes. Univ. Iowa Press, Iowa City. WHITNEY, K. D. 1980. The myxomycete genus Echinostelium. Mycologia 72:950— 987. (Received 21 Jul 1981; revision accepted 5 Feb 1982.) 1982] NOTES AND NEWS 269 NOTES AND NEWS GYNODIOECY IN Saxifraga integrifolia (SAXIFRAGACEAE).—S axifraga integrifolia Hook. is a perennial herb found principally in vernally wet, mossy and grassy meadows at low elevations west of the Cascade Mountains in southern British Columbia, Washing- ton, and Oregon. In a population on a small Puget Sound island at Deception Pass, Washington, some plants were found that had only small, green to brown anthers while other plants had larger, orange anthers. Inflorescences were collected from this popu- lation, segregated according to anther type, and preserved in Carnoy’s solution. Ana- tomical studies of flowers using standard paraffin embedding, sectioning, and staining techniques showed that, in flowers with green anthers, microsporogenesis fails, leading to male-sterility, but that embryo sac development is normal. In flowers with orange anthers, microsporogenesis appears normal, abundant pollen is produced, and embryo sac development is normal. These flowers are hermaphroditic. During three field seasons, 1100 inflorescences were examined to determine the ratio of hermaphrodites to male-steriles. The ratio remained constant at 2:1. In the first year, plants were marked according to anther type and re-examined each year. Male-steriles remained male-steriles and hermaphrodites remained hermaphrodites. Male sterility in plants at Deception Pass involves a complete abortion of microsporogenesis and is not influenced by environmental factors known to affect pollen development in other plants (Jones, Amer. J. Bot. 63:657—-663. 1976). The population of S. integrifolia at Deception Pass is gynodioecious. For comparative purposes, specimens were collected from Deception Pass and from populations of S. integrifolia that did not exhibit gynodioecy. From these specimens, ovule number per carpel, percent seed set, and pollen stainability were determined. In addition, hermaphroditic plants from Deception Pass were brought into a greenhouse and were selfed in an insect-free cage to determine if they were self-compatible. The results in Table 1 show no significant differences between hermaphrodites and male-steriles at Deception Pass in ovule number per carpel or percent seed set. Plants TABLE 1. OvuLE NUMBER PER CARPEL AND POLLEN STAINABILITY FOR S. inte- grifolia AND PERCENT SEED SET OF FIELD PLANTS COMPARED WITH THAT OF Ex- PERIMENTALLY SELFED PLANTS FROM DECEPTION PASS. x is the average + standard deviation (S.D.). Numbers in parentheses refer to collections by Elvander, vouchers at WTU. No. flowers/ no. plants x SD. Number of ovules/carpel San Juan Islands/Columbia River Gorge (563, 575, 600, 607) 44/12 106 + 40.0 Deception Pass (543, 819) 27/13 15422350.) Male-steriles 14/7 154 + 24.8 Hermaphrodites 13/6 154 + 46.4 Percent pollen stainability Columbia River Gorge (563, 566, 585, 813, 814) 44/44 (jin wen Wee Deception Pass (543, 545, 588, 819) 6/6 87 + 9.4 Percent seed set/Deception Pass (543, 819) 27/13 7821107 Male-steriles 14/7 (kes o Hermaphrodites 13/6 ff mags OS: Selfed hermaphrodites 10/3 Si eae A | 270 MADRONO [Vol. 29 at Deception Pass had a higher average number of ovules per carpel and percent pollen stainability than for other populations of S. integrifolia. The differences in ovule num- ber and percent pollen stainability seem to indicate that the gynodioecious population has slightly higher potential fecundity than the other populations examined. The her- maphrodites from Deception Pass that were selfed had an average 38% seed set and thus, are at least partially self-compatible. Gynodioecy has been associated with self- compatibility as well as with effects of severe inbreeding and with maintenance of hybridity (Lewis and Crowe, Evolution 10:115—125. 1956; Jain, Genetics 46:1237—1240. 1961; Levin, Taxon 20:91—113, 1971). . Gynodioecy has not been noted previously for Saxifraga (Lewis, New Phytologist 50: 56-63. 1941; Biol Rev. Cambridge Philos. Soc. 17:46-67. 1942; Lewis and Crowe, Evolution 10:115—125. 1956) though two dioecious species have been described (Engler and Irmscher, Das Pflanzenreich. 1916; Chambers, Madrono 17:203—204. 1964). Male- sterile plants have been reported from the San Juan Islands, Washington (personal observation) and irregular meiosis has been reported in plants from southwestern British Columbia and Vancouver Island (Beamish, Canad. J. Bot. 39:567—-580. 1961). These populations have not yet been studied. The full extent and significance of gynodioecy in S. integrifolia is not yet known. I thank Melinda Denton and the Department of Botany and Plant Pathology, Oregon State University, Corvallis. This study was part of a Doctoral Dissertation at the Department of Botany, University of Washington, Seattle. A portion of this study was supported by a National Science Foundation grant.— PATRICK E. ELVANDER, Biology Board, University of California, Santa Cruz 95065. (Received 14 Aug 1981; revision accepted 20 Dec 1981.) EFFECTS ON Lomatium triternatum OF THE 1980 ASH FALLOUT FROM MT. ST. Helens.—Three populations of Lomatium triternatum (Pursh) Coult. & Rose (Umbel- liferae) that were in the path of the ash fallout had been monitored for several consec- utive years prior to the eruption of Mt. St. Helens on 18 May 1980. These populations are from Deary, Idaho; Pullman, Washington, on the corner of Stadium Way and the Moscow highway; and Magpie Hill, a kilometer north of Pullman. Annually-assembled data included population size, height, and sex ratios of tagged plants. In addition to these three populations, two others several hundred km to the south of the ash zone had been monitored. These are located 32 km west of Burns, Oregon, highway 20, and in the Challis National Forest of Idaho, 4 km from the junction of Wild Horse Creek and Fall Creek on the trail toward Moose Lake. No significant differences were found in plant size, number of leaves, or sex ratio between plants in populations within and outside of the zone of ash fallout. It is par- ticularly fortunate that both groups of populations continued to be monitored during the eruption year because of the striking difference in percent flowering individuals per population between 1980 and 1981. The same sorts of differences, however, occurred regardless of whether or not the populations were within the zone of ash fallout. The ash had no noticeable effect on plant height, number of leaves or sex ratios nor on the proportion of individuals of the populations that flowered. Flowering of this herbaceous perennial fluctuates greatly from year to year in the same population. Data are interpreted as partly resulting from the time of ashfall relative to the life cycle of this species. Lomatium triternatum had already begun to set seed at the time of maximum fallout. Had it not passed its flowering peak it might have suffered some setback, but still without local extermination, much as Mack (Science 213:537-539) noted with Veratrum and Balsamorhiza. Lomatium triternatum is not a forest dweller 1982] NOTES AND NEWS 271 and receives no protection from other plants, but its linearly dissected leaves with lax leaflets facilitated the movement of the ash through the plant without damaging leaves or stem. In view of the frequency of Cascade Range volcanism during the past several hundred millenia and the prevailing westerlies, this species from eastern Washington and Idaho may have evolved the structure that permits it to endure during ash fallouts. Support is acknowledged from the NSF, DEB 80-20937, Notice #82.—AMyY JEAN GILMARTIN, Department of Botany, Washington State University, Pullman 99164. (Received 16 Sep 1981; revision accepted 20 Dec 1981.) NOTEWORTHY COLLECTIONS BRITISH COLUMBIA POLYSTICHUM KRUCKEBERGII Wagner (POLYPODIACEAE).—Canada, B.C., Cassiar District, 1 km s. of W. Kwanika Cr. (55°33'N, 125°21’W), 1050 m, 27 Jul 1981, A. L. Kruckeberg 204 (UBC). Locally frequent on steep scree of serpentine barren. Previous knowledge. Species known from the vicinity of Lillooet, B.C., s. to CA and NM. Herbaria consulted: UBC, ORE. Significance. A range extension of ca. 580 km nnw-,; also the first record of the species from the Sub-boreal Spruce zone of British Columbia.—ARTHUR LEO KRUCKEBERG, Dept. Botany, Univ. British Columbia, Vancouver, Canada V6T 1W5S. (Received 28 Jan 1982). CALIFORNIA CYMOPTERUS RIPLEYI Barneby (APIACEAE).—Inyo Co., Coso Range, slopes of Owens Valley (T19-20S R37E); 27 Apr 1974, DeDecker 3403 (RSA and private herbarium), a few plants on a slope n. of Haiwee Reservoir (T19S R37E S34nw'4); 11 May 1978, DeDecker 4663 (UC and private herbarium), a well established population along the Cactus Flat Rd. e. of Haiwee Reservoir (T20S R37E S2nw'4); and 15 Apr 1978, Mary Ann Henry, verified by DeDecker, e. of Olancha (T19S R37E S25n'4). Previous knowledge. Known from Esmeralda, Lincoln and Nye Cos., NV. Significance. First record for CA, and most w. location, a disjunction of 160 km. Considered rare in CA, common elsewhere, by CNPS (Smith et al., CNPS Spec. Publ. 1, ed. 2. 1980). CRYPTANTHA SCOPARIA A. Nels. (BORAGINACEAE).—Inyo Co., Inyo Mts.; 20 Jul 1971, DeDecker 2776 (RSA and private herbarium), established in gully and along roadside on the e. slope of Mazourka Pk., w. of Badger Flat (T11S R35E S23sw'4); 6 Jul 1974, DeDecker 3549 (Private herbarium), same location as above; 13 Jul 1974, DeDecker 3552 (SB, CAS, RSA and private herbarium), same location as above; 20 Jun 1980, DeDecker 5047 (Private herbarium), Hines Rd. s. of Waucoba Rd. (T10S R36E S6ne\4). Previous knowledge. Known from WA, OR, ID, WY and NV. Significance. First record for CA. Although infrequent, it is probable that it is more widespread than has been noted. Considered rare in CA, common elsewhere, by CNPS (Smith et al., CNPS Spec. Publ. 1, ed. 2. 1980). ASTRAGALUS ARGOPHYLLUS Nutt. ex Torr. & Gray var. ARGOPHYLLUS (FABA- CEAE).—Mono Co., Fish Slough (T5S R32E S30sw'4); 14 May 1974, DeDecker 3481 252 MADRONO [Vol. 29 (RSA, CAS and private herbarium); 15 Jul 1976, DeDecker 4106 (Private herbarium); 2 Jun 1978, DeDecker 4716 (Private herbarium); 20 Jun 1978, DeDecker 4730 (NY and private herbarium), verified by R. C. Barneby. All of the above were from the vicinity of BLM Spring, 1280 m, where the colony is limited to about 10 plants. Another population has been noted about 1 km nw. of that spring. Previous knowledge. According to Barneby (R. C. Barneby, Atlas of North American Astragalus, Part II, 1964) the plant is known from central and nw. NV to ID, WY, MONT and central UT. Significance. First record for CA, apparently the most southerly occurrence, a dis- junct population, possibly having followed the waterways of the Pleistocene Epoch. Considered rare in CA, common elsewhere, by CNPS (Smith et al., CNPS Spec. Publ. 1, ed. 2. 1980). ERIOGONUM PUBERULUM 5S. Wats. (POLYGONACEAE).—Inyo Co., Panamint Range, Cottonwood Mts., Tin Mountain, Death Valley Natl. Mon.; 7 Jul 1978, DeDecker 4754 (CAS and private herbarium), summit Tin Mountain, 2728 m (ca. 36°02'N, 117°27'W); 24 Jul 1978, DeDecker 4760 (UC and private herbarium), summit Tin Mountain; 25 Jul 1978, observed by DeDecker and Pavlick on s. slope of Tin Mountain, 2440 m. Con- firmed by John Thomas Howell and James R. Reveal. Previous knowledge. Known from UT and NV. Significance. First record for CA, as well as the most westerly occurrence, and pos- sibly the highest in elevation, a disjunction of perhaps 250 km. Considered rare in CA, common elsewhere, by CNPS (Smith et al., CNPS Spec. Publ. 1, ed. 2. 1980). Although this plant has been known as an Eviogonum, the fact that its “involucres” are actually nodal bracts raises a question as to its generic position.—MARY DEDECKER, P.O. Box 506, Independence, CA 93526. (Received 29 Oct 1981) HIERACIUM ARGUTUM Nutt. var. PARISHII (Gray) Jepson.—San Diego Co., Corte Madera ranch, just s. of Pine Valley, 60 km e. of San Diego, 1300 m, 10 Aug 1980, van der Werff 4123 (SD, CAS); 5 Jul 1981, van der Werff 4334 (LA). Previous knowledge. Southern face of the San Gabriel and San Bernardino Mts. to Ventura Co. Mainland stations for the typical variety are Santa Barbara and San Luis Obispo Cos. Herbaria consulted: SD, RSA, POM, LA. Significance. First record for San Diego Co., a range extension of 200 km. CYPSELEA HUMIFUSA Turpin.—San Diego Co., Corte Madera ranch, just s. of Pine Valley, 60 km e. of San Diego, 1300 m, 10 Aug 1980, van der Werff 4128 (SD, CAS); 5 Jul 1981, van der Werff 4332 (LA). Not rare, on drying lake shore. Previous knowledge. A West Indian species. In California recorded only from the lower San Joaquin River, Santa Cruz, Marin, Sonoma, and Lake Cos. Herbaria con- sulted: SD, RSA, POM, LA. Significance. First record for southern CA, a range extension of 700 km. NELUMBO LUTEA (Willd.) Pers.—San Diego Co., Corte Madera ranch, just s. of Pine Valley, 60 km e. of San Diego, 1300 m, 30 Aug 1980, van der Werff 4136 (SD, CAS, LA). Well-established in a small, man-made lake. Possibly introduced with cattle from Texas. Previous knowledge. From e. Oklahoma, e. Texas to Florida, New England. Significance. First record for CA. LASTHENIA GLABERRIMA DC.—San Diego Co., Corte Madera ranch, just s. of Pine Valley, 60 km e. of San Diego, 1300 m, 3 Jul 1980, van der Werff 4017 (SD, CAS, LA; additional duplicates distributed by SD). Previous knowledge. From Washington to s. Monterey Co. Herbaria consulted: SD, RSA, POM, LA. 1982] NOTEWORTHY COLLECTIONS 213 Significance. First record for southern CA, a range extension of about 500 km.— HENK VAN DER WERFF, 4239 Arden Way, San Diego, CA 92103. (Received 29 Nov 1981) ERIOPHYLLUM CONGDONII Brandg. (ASTERACEAE).—Mariposa Co., along n. side of South Fork of Merced River from Zip Cr. to Devil Gulch 6.5 km intermittently, 620— 915 m, 23 Apr 1981, Bottz 92 (CAS); summit and nw. ridge of Iron Mtn., Sierra Natl. Forest, 1190-1890 m, 15 May 1981, Bott: 113 (CAS). Previous knowledge. Known only from Merced River canyon on ridges adjacent to Rancheria Flat at El Portal. Significance. This is the first record of E. congdonii in the South Fork of the Merced drainage and the first collection from outside the localized area surrounding Rancheria Flat. The vast number of plants observed more than doubles the known population. Also, the collection from Iron Mtn. (1890 m) is the highest known station for E. cong- donii. This may be significant in determining the relationship between E. congdonii and its closest relative E. nubigenum, which grows at a similar elevation. Due to the meagerness of known collections, E. nubigenum was at one time thought to be a dwarf alpine representative of E. congdonii, and the two were thought to grow at greatly varying elevations (Constance, L. 1937, A systematic study of the genus Eriophyllum Lag. Univ. Calif. Publ. Bot. 18:69-135). E. congdoniz is listed in the Federal Register (45FR8248082569) by the Endangered Species Office as a taxon for which enough in- formation exists to support listing as an endangered or threatened species. It is consid- ered “rare and endangered” by the California Native Plant Society (Smith et al., Inv. rare and endang. vasc. pls. Calif., CNPS Spec. Publ. 1, ed., 2, 1980). LEWISIA CONGDONII (Rydb.) J. T. Howell (PORTULACACEAE).—Mariposa Co., Sierra Natl. Forest, summit of Iron Mtn. and along nw. ridge extending to South Fork of the Merced River, 915-1890 m, 15 May 1981, Botti 117 (CAS). Approximately 7000 plants were located in areas of sparsely vegetated metamorphic rock, mostly facing north. Associated species include Allium yosemitense and Eriophyllum congdonii, both ex- tremely rare and localized. Previous knowledge. Known from El] Portal and Chowchilla Mtn. in Mariposa Co., CA, and near Yucca Point along Hwy. 180, Fresno Co., CA. Significance. This collection establishes only the fourth known site for L. congdonii. The large size of the newly discovered population is important in assessing the rarity of the taxon. In the December 15 1980 Federal Register (45FR8248082569) the Endan- gered Species Office listed L. congdonii within Category 1, which includes taxa for which enough information exists to support listing as an endangered or threatened species. Considered “rare and endangered” by the California Native Plant Society (Smith et al., Inv. rare and endang. vasc. pls. Calif., CNPS Spec. Publ. 1, ed. 2, 1980). The new population also establishes an intermediate link between the E] Portal site 8 km to the ne. and the Chowchilla Mtn. site 5 km to the s.—STEPHEN BOTTI, Resources Management Specialist, Yosemite Natl. Park, CA 95389. (Received 10 Dec 1981) OXYTHECA WATSONII Torrey & Gray (POLYGONACEAE).—Inyo Co., 1.3 km nne. of jet. State Hwy. 190 and Saline Valley Rd., Santa Rosa Wash vicinity, extreme s. end of Santa Rosa Flat (T18S R40E, S15 projected), 24 Sep 1980, Stone, Castagnoli, and de Nevers 316 (CAS, MARY, RSA). Verified by B. Ertter, 1981. Population of over 100 plants over a broad, flat, fully exposed area (1 ha or greater), 1480 m. Previous knowledge. The collection Parish & Parish 1241 from “N slope of San Bernardino Mts., near Cushenberry (sic) Springs, May 1882,” originally but erroneously referred to Oxytheca watsoni (sic), is now the type for Oxytheca parishii Parry var. goodmaniana Ertter. Oxytheca watsonii is otherwise known only from four scattered localities in c. and wc. Nevada (Ertter, Brittonia 32:80, 1980). Significance. Currently under review for listing as threatened or endangered (US Fish 274 MADRONO [Vol. 29 & Wildlife Serv., Federal Register 45:82525, 15 Dec 1980). This record is the first correct one for Oxytheca watsonii from California and represents a 225 km range extension from the nearest Nevada locality.—R. DouG STONE, Environmental Field Program, University of California, Santa Cruz 95064. (Received 19 Jan 1982) COLORADO ERIASTRUM DIFFUSUM (A. Gray) H. L. Mason (POLEMONIACEAE).—Mesa Co., nw.- facing slope of knoll, Rabbit Valley, se. from junction of Rabbit Valley Rd. and Inter- state 70 (T10S R104W S16, 30°11’N, 109°0’W), 1400 m, 23 May 1980, Kelley 80-33 (Mesa College Herb.), 16 Jun 1980, Kelley 80-92 (Mesa College Herb., CS). Previous knowledge. Widespread from s. CA, AZ, nw. MEX, sw. NM, w. TX n. to NV and UT. Herbaria consulted: BRY, COCO, COLO, CS, RM, USFS, Mesa College. Significance. This is the first record of the species and the genus in CO, representing a 200 km disjunction from nearest known localities in Garfield and San Juan Cos., UT. CRYPSIS ALOPECUROIDES (Pill. & Mitlerp.) Schrad. (POACEAE).—Jefferson Co., se. shore of Standley Lake near north Kipling St. (39°52'N, 105°7’25”W), 1520 m, 5 Sep 1980, Walter and Lormond s.n. (CS). Previous knowledge. Widespread from n. CA to s. WA and known from 1 locality each in ID and WY (Hammel and Reeder, Syst. Bot. 4:267—280. 1979). Significance. This is the first record of the species and the genus in CO, representing a 240 km disjunction from the nearest known locality in Goshen Co., WY. As noted by Hammel and Reeder (1979) this species is spreading rapidly throughout w. N. Amer. as an Old World introduction.—WALT KELLEY, Dept. of Biology, Mesa College, Grand Junction, CO 81501; GENEVIEVE BRYANT and DIETER WILKEN, Dept. of Botany and Plant Pathology, Colorado State University, Fort Collins 80523. (Received 29 Jan 1982) Correction. For the chromosome count of Tanacetum huronense Nutt., Madrono 29:62, 1982, the county should have been given as Emmet County, MI, not Cheboygan County, and, although there are other plants from the same population in the her- barium at MICH, the voucher actually grown from the plants that were counted is deposited at MO.—Peter H. Raven, Missouri Botanical Garden, P.O. Box 299, St. Louis, MO 63166. 1982] REVIEW Zt REVIEW Bromelioideae (Bromeliaceae). By LYMAN B. SMITH and ROBERT J. DOWNS. Flora Neotropica Monograph No. 14, Part 3:1493—2142, New York Botanical Garden, Bronx, NY 10458. 1979. ISBN 0-89327-210-8. $65.00! This volume is another in the growing list of welcome treatments in the Flora Neo- tropica Monograph series. Appearance of this particular title, however, gives the bo- tanical community special reason for celebration because it marks completion of the largest family of angiosperms yet to be treated in the ongoing series. Treatments for subfamilies Pitcairnioideae and Tillandsioideae appeared as Monographs No. 14, Part 1 (1974) and No. 14, Part 2 (1977) respectively. The Bromelioideae are technically distinguished from the other subfamilies by the inferior ovary, indehiscent baccate fruits, and naked seeds but the group is probably best known to both botanist and layperson through the pineapple (Ananas) and various greenhouse ornamentals such as Aechmea, Billbergia, and Cryptanthus. As interpreted by Smith and Downs this subfamily consists of 27 genera and 723 specific and infra- specific taxa with a geographic distribution extending from Mexico and the West Indies south to Chile and Argentina. By any estimate, however, the Bromelioideae are best represented in Brazil. In content and general style this volume adheres to the effective format used in previous monographs of the series. For all species and infraspecific taxa the authors present a statement of synonymy; concise descriptions; collector(s), locality and physical location of the type specimen(s); a statement of habitat, elevational range, and geo- graphical distribution supported by citation of specimens examined. Lists of hybrids and cultivars are appended to treatments for many of the genera and the volume con- cludes with a numerical list of taxa, list of exsiccatae, and a general exhaustive index. Unfortunately pages 2099-2102 are missing from the index to my review copy, so available stock of the press run should be checked for this defect. The text is enhanced by numerous distribution maps and over 200 line drawings that vary in clarity and detail. All maps depict collective ranges of species in a genus or subgenus instead of ranges for specific taxa—a prudent and justified practice because so many of the taxa are known from few collections. At a time when tropical vegetation throughout the world is facing rapid, irreversible destruction, it is distressing to learn that 195 of the taxa treated in this volume are known from type collections only. Dozens of others in the Bromelioideae are known only from the type locality or from specimens in culti- vation. After perusing this volume in some detail, I felt the desire to learn more about relationships of the species, something of their variation, and in some cases reasons for the particular taxonomic treatment chosen by the authors. Commentaries with this information are lacking in this volume and understandably reflect the fact that much remains to be learned about this curious, complex and little-collected subfamily of Bromeliaceae. Like all monographic and floristic publications dealing with tropical plants, this volume provides a useful progress report that can be reviewed and emended as new information comes to light. Smith has published voluminously on the Brome- liaceae for decades and completion of the Flora Neotropica treatment will certainly be viewed as a crowning achievement in this phase of his long and productive career. To be sure, Smith and Downs have produced a monumental treatise on the Bromeliaceae that will serve as the standard and indispensable reference for decades to come.—F RANK ALMEDA, Department of Botany, California Academy of Sciences, Golden Gate Park, San Francisco, CA 94118. 276 MADRONO [Vol. 29 REVIEWERS OF MANUSCRIPTS Perhaps the most difficult task facing an editor is the selection of appropriate re- viewers for manuscripts. He faces the dual responsibility of finding people who can provide thoughtful, constructive criticism for the author and of not overworking any given set of reviewers. Frequently those most capable of providing the kinds of advice most severely needed are the busiest people in their fields. The people listed below have generously contributed their time, over and above their own research commitments, toward maintaining the quality of Madrono, and their assistance in the preparation of Volume 29 has been greatly appreciated. David P. Adam Frank Drysdale Rexford Palmer Edward F. Anderson W. G. Ferlatte David J. Parsons G. D. Barbe Arthur C. Gibson Peter Raven Spencer Barrett James R. Griffin Jerzy Rzedowski Jerry Baskin Douglass Henderson David S. Seigler Meredith Blackwell James Henrickson Dale Smith Mary T. Burke James C. Hickman Richard Spellenberg Kenton L. Chambers David J. Keil Timothy P. Spira Curtis Clark Sterling C. Keeley John Strother Susan G. Conard William Martin Dean W. Taylor Lincoln Constance Louis V. Mingrone Ronald J. Taylor Daniel J. Crawford Reid Moran Robert F. Thorne Robert W. Cruden Erik T. Nilsen Richard J. Vogel Don G. Despain Robert Ornduff Robert Wright Robert D. Dorn 1982] REVIEW 277 EDITOR’S CONFESSIONS FOR VOLUME 29 Between 1 Jul 1981 and 30 Jun 1982, 77 manuscripts were received. This repre- sents a 17% increase over the submission rate of 1980-1981. Of this year’s manu- scripts, 45 (58%) were articles, a 12.5% increase over last year; 22 were noteworthy collections, a 46% increase; and 10 (13%) were notes, a 1% decrease. Current statuses of manuscripts are as follows: in review, 17; in revision, 22 (includes a few hold- overs from 1980-1981); accepted and awaiting publication, 23 (a 53% increase over 1980-1981); published in volume 29, 26 articles (52%), 16 noteworthy collections (32%), and 8 notes (16%). Corresponding figures for 1980-1981 were 27, 19, and 15. The 53% increase in backlog represents something of a problem because of the in- crease in publication costs. Pages published in Madrono have increased from 192 (vol. 27), to 279 (vol. 28), and ca. 280 (vol. 29). The Society faces these critical alternatives: increase dues, increase rejection rate of manuscripts, decrease the size of Madrono. A few papers were rejected this year, but Madrono’s absolute rejection rate remains low. Transition of the editorial office from Los Angeles to Idaho occurred smoothly, and I have been able to keep up with my duties. I have even been ahead of things a couple of times. Time from final acceptance to publication is now one year, an increase from 1981. I have received one compliment, one complaint, and no hate mail. Most authors appear to accept Madrono as a suitable outlet for their work and I am pleased with the continuing diversity of material received. The editorial office will continue to welcome suggestions and criticisms from members. C. D.12 Jul 1982 Dates of publication of MADRONO, volume 29 No. 1, pp. 1-66: 15 January 1982 No. 2, pp. 67-125: 29 April 1982 No. 3, pp. 127-218: 9 July 1982 No. 4, pp. 219-281: 12 October 1982 278 MADRONO [Vol. 29 INDEX TO VOLUME 29 Classified entries: major subjects, key words, and results; botanical names and plant families (new names are in boldface); geographical areas; reviews. Incidental references to taxa (including lists and tables) are not indexed separately. Species appearing in Noteworthy Collections appear under plant family, state, or country. Authors and articles are listed alphabetically in the Table of Contents. Abies concolor, post-fire succession, 42. Aizoaceae: Cypselea humifusa, range ex- tension (CA), 272. Allium lacunosum, taxonomy of species complex, 79. Alpine, two species of Lomatium in OR, 13. Amaryllidaceae: Hypoxis mexicana redis- covered (AZ), 57. See also Liliaceae. Anacardiaceae: Actinocheita, wood anat- omy, 61. Apiaceae: Cymopterus ripleyi, new to CA, 271; Lomatium bicolor, taxonomy, 118; Lomatium bicolor and Musineon vagi- natum, new to MT, 58; Lomatium or- eganum and L. greenmanii endemic to OR, 13; Lomatium triternatum, effect of ash fallout on, 270. Arizona: rediscovery of Hypoxis mexi- cana, 57; second collection of Linum subteres, 57; plant species diversity in, 227. Aspleniaceae: Asplenium trichomanes, new to CA, 57. Asteraceae: Antennaria monocephala, new to MT and US, 58; Dicoria argentea, new species from Sonora, Mexico, 101; Ericameria riskindii, new species from Mexico, 234; Eriophyllum congdonii, range extension (CA), 273; Eriophyllum nubigenum, rediscovery (CA), 123; Fi- lago arvensis, spread of, 119; Hiera- clum argutum, new to CA, 272; Las- thenia glaberrima, range extension (CA), 272; Malacothrix, on CA islands, 218; miscellaneous chromosome numbers, 62; Plummera ambigens, new to NM, 60; Saussurea densa, new to MT, 59. Baja California: Pseudotsuga macrocarpa, apparently absent, 22. Boraginaceae: Cryptantha scoparia, new to CA, 271; Myosotis arvensis, new to WY, 125; Myosotis micrantha, new to WY, 125. Brassicaceae: Rorippa sylvestris, new to MT, 59. British Columbia: Polystichum krucke- bergiz, range extension, 271. California: subalpine meadows, 1; Pro- sopis taxonomy, 237; corticolous myxo- mycetes, 259. New records: Asplenium trichomanes, 57; Astragalus tegetarioides, 58. Range extensions: Abies lasiocarpa, 218; Malacothrix, on CA islands, 218; Mirabilis laevis, 123. Taxa rediscovered: Eriophyllum nubi- genum, 123. Campanulaceae: Howellia aquatilis, new to MT, 123. Caryophyllaceae: Dianthus barbatus, new to WY, 124. Chlorogalum angustifolium, floral varia- tion in, 87. Chromosome numbers in Asteraceae, 62. Climate diagram, UC Sagehen Cr. Field Station, 122. Colorado: New Records: Crypsis alopecuroides, 271; Eriastrum diffusum, 271. Community ecology: classification of sub- alpine meadows in the Sierra Nevada, 1; conifer forest in Yosemite Natl. Park, 109; desert ephemerals, 154; dynamics of mountain meadows, 148; post-fire ecology of Lolium multiflorum, 177; post- fire succession in white-fir forest, 42; vegetation of Rae Lakes Basin, 164; vegetation of Sequoia Natl. Park, 200. Compositae—see Asteraceae. Cyperaceae: Carex bipartita, new to WY, 124; Carex deweyana, new to WY, 124; Carex incurviformis, new to WY, 124; Carex microglochin, new to UT, 60; Carex parryana, new to UT, 60; Cen- chrus incertus, new to Galapagos Isl., 217; Kobresia simpliciuscula, new to WE, 60: Desert ephemerals, phenology, germina- tion, and survival, 154. Dicoria argentea, new species from So- nora, Mexico, 101. Ecological research, tribute to Jack Ma- jor, 220. Editor’s confessions for volume 29, 277. 1982] Endangered, rare, or threatened plants: 58, LS eli ince Oe Endemism: age and origin of the Monte- rey endemic area, 127; Lomatium ore- ganum and L. greenmanii in OR, 13. Ephemerals, phenology, germination, and survival in the desert, 154. Ericameria riskindii, new species in north-central Mexico, 234. Eschscholzia, infertility in parapatric species, 32. Fabaceae: Astragalus argophyllus var. ar- gophyllus, new to CA, 271; Astragalus tegetarioides, new to CA, 58; Eysen- hardtia polystachya, confirmed in NM, 60; Prosopis, diversity and taxonomy of CA species, 237. Fallout, effect of volcanic ash on Loma- tium triternatum, 270. Filago arvensis, migration in the US, 119. Fire ecology: post-fire succession in white- fir forest, 42; pine seedlings, native ground cover, and Lolium multiflorum on the Marble-cone burn, Santa Lucia Mts., CA, 177. Floral variation in Chlorogalum angusti- folium, 87. Floristic affinities in the high Sierra Ne- vada, 189. Galapagos Islands: Cenchrus incertus, new record, 217. Gentianaceae: Gentiana tenella, new to MT, 58; Gentianella propinqua, new to WY, 125. Germination of desert ephemerals, 154. Grossulariaceae: Ribes triste, new to MT, 59. Gynodioecy in Saxifraga integrifolia, 269. Herbarium news, 66. Hybridization, Populus x inopina Eck- enwalder, a natural hybrid between P. fremontii and P. nigra, 67. Hydrophyllaceae: Phacelia thermalis, new to MT, 59. Infertility, unilateral infertility in Esch- scholzia, 32. Juncaceae: Juncus triglumis, new to MT, 58. Lamiaceae: Salvia summa, range exten- sion (NM, TX), 217; Satureja dougla- INDEX 279 sil, new to MT, 59; Trichastema mex- icanum, taxonomic recognition of, 104. Leguminosae—see Fabaceae. Liliaceae: Allium lacunosum complex, taxonomy, 79; Chlorogalum angustifo- lium, floral variation, 87. See also Amaryllidaceae. Linaceae: Linum subteres, recollected (AZ), Si Lolium multiflorum, effect on succession after the Marble-cone fire, Santa Lucia Range, CA, 177. Lomatium: Lomatium oreganum and L. greenmanii endemic to OR, 13; L. bi- color, taxonomy, 118; L. triternatum, effect of volcanic ash on, 270. Lythraceae: Cuphea wrightii, new to NM, 60. Major, Jack, volume dedication, tribute to, 220. Malacothrix, distribution on CA islands, 218. Malpighiaceae: Aspicarpa hirtella, new to NM, 60. Meadow ecology: classification of subal- pine meadows in the Sierra Nevada, CA, 1; dynamics of mountain meadows, 148. Mesquite, diversity and taxonomy of CA species, 237. Mexico: Pseudotsuga macrocarpa in Baja California?, 22; Trichostema mexica- num, taxonomic recognition of, 104. New taxa: Dicoria argentea Strother, 101; Ericameria riskindii Turner & Langford, 234; Symplocos sousae Almeda, 255. Montana: New records: Antennaria monocephala, 58; Gentiana tenella, 58; Juncus tri- glumis, 58; Koenigia islandica, 58; Lomatium bicolor, 58; Phacelia ther- malis, 59; Plantago hirtella, 59; Ribes triste, 59; Rorippa sylvestris, 59; Sat- ureja douglasii, 59; Veronica verna, 59. Range extensions: Musineon vagina- tum, 58; Saussurea densa, 59. Monterey endemic area, age and origin, ae Myxomycetes, survey of corticolous species in CA, 259. 218, Nelumbonaceae: Nelumbo lutea, new to CA, 272. 280 New Mexico: New records: Aspicarpa hirtella, 60; Cuphea wrightii, 60; Eysenhardtia polystachya, 60; Heuchera glomeru- lata, 60. Range extensions: Plummera ambigens, 60; Salvia summa, 217. New taxa: Polygala rimulicola var. mescalerorum Wendt & Todsen, 19. Nyctaginaceae: Mirabilis laevis, range ex- tension (CA), 123. Onagraceae: Epilobium nevadense, range extension (UT), 60. Oregon: Lomatium oreganum and L. greenmanil, endemic to, 13. Orobanchaceae: taxonomy and distribu- tion of Orobanche valida, 95. Papaveraceae: unilateral infertility in Eschscholzia, 32. Phenology of desert ephemerals, 154. Pinaceae: pine seedlings on the Marble- cone burn, Santa Lucia Range, CA, 177; post-fire succession in Abies concolor forest, 42. Plantaginaceae: Plantago hirtella, new to MT, 59. Poaceae: Crypsis alopecuroides, new to CO, 274; Lolium multiflorum, effect on succession after a fire, 177. Polemoniaceae: Eriastrum diffusum, new to CO, 274. Polygalaceae: Polygala rimulicola var. mescalerorum Wendt & Todsen, new species from NM, 19. Polygonaceae: Eriogonum puberulum, new to CA, 272; Koenigia islandica, new to MT, 58; Oxytheca watsonii, new to CA, 213% Polypodiaceae: Polystichum kruckebergii, range extension (British Columbia), 271. Populus: P. X inopina Eckenwalder, natural hybrid between P. fremontii and P nigra, 67. Portulacaceae: Lewisia congdonii, range extension (CA), 273. Prosopis, diversity and taxonomy of CA species, 237. Pseudotsuga macrocarpa, non-occurrence in Baja California, 22. Rae Lakes Basin, vegetation, 164. Rare species—see Endangered species. Reviews: D. E. Breedlove, Flora of Chia- pas, parts 1 and 2, 215, 216; Adriana MADRONO [Vol. 29 Hoffmann J., Flora silvestre de Chile. Zona Central, 63; N. T. Mirov, The road I came. the memoirs of a Russian- American forester, 65; D. M. Power, ed., The California islands: proceedings of a multidisciplinary symposium, 64; L. B. Smith and R. J. Downs, Bromelioideae (Bromeliaceae), Flora Neotropica monogr. 14, 275; E. Stuhl and M. C. Ford, Edward Stuhl’s wildflowers of Mount Shasta, 65; J. G. Zabriskie, Plants of Deep Canyon and the central Coachella Valley, California, 63. Rosaceae: Potentilla recta, new to WY, 125. Sagehen Cr. Field Station, climate dia- gram, 122. Salicaceae: Populus x inopina Ecken- walder, natural hybrid between P. fre- montit and P. nigra, 67. Santa Lucia Range, CA, 177. Saxifragaceae: gynodioecy in Saxifraga integrifolia, 269; Heuchera glomerula- ta, new to NM, 60. See also Grossular- laceae. Scrophulariaceae: Veronica verna, new to MT, 59. Sequoia Natl. Park, vegetation, 200. Sierra Nevada: classification of subalpine meadows, 1; floristic affinities, 189; Sierran park management, 220; succes- sion in white-fir forest, 42; vegetation of Rae Lakes Basin, 164. Succession: in white-fir forest, 42; effect of Lolium multifiorum after a fire, 177. Symplocaceae: Symplocos sousae Alme- da, new species from Mexico, 255. Texas: Salvia summa, new record, 217. Threatened species—see Endangered species. Trichostema mexicanum, taxonomic rec- ognition of, 104. Umbelliferae—see Apiaceae. Utah: New records: Carex microglochin, 60; Kobresia simpliciuscula, 60. Range extensions: Carex parryana, 60; Epilobium nevadense, 60. Vegetation: Rae Lakes Basin, 164; Se- quoia Natl. Park, 200; white-fir forest, 42. 1982] INDEX 281 Dianthus barbatus, 124; Gentianella propinqua, 124; Myosotis arvensis, 125; M. micrantha, 125; Potentilla recta, 125; Volcanic ash, effect on Lomatium triter- natum, 270. Wood anatomy, of Actinocheita (Anacar- diaceae), 61. Wyoming: New records: Carex bipartita, 124; C. deweyana, 124; C. incurviformis, 124; Yosemite Natl. Park, ecology of conifer forests, 109. iw fs SUBSCRIPTIONS—MEMBERSHIP Membership in the California Botanical Society is open to individuals ($18 per year; students $10 per year for a maximum of seven years). Members of the Society receive MADRONO free. Family memberships ($20) include one eight-page publishing allotment and one journal. Emeritus rates are available from the Corresponding Secretary. In- stitutional subscriptions to MADRONO are available ($25). Membership is based on a calendar year only. Applications for membership (including dues), orders for sub- scriptions, requests and rates for back issues, changes of address, and undelivered copies of MADRONO should be sent to the Corresponding Secretary. Renewal payments should be sent to the Treasurer. INFORMATION FOR CONTRIBUTORS Manuscripts submitted for publication in MADRONO should be sent to the editor. All authors must be members, and membership is prerequisite for review. Manuscripts and review copies of illustrations must be submitted in triplicate for all articles and short items intended for NOTES AND NEWS. Follow the format used in recent issues for the type of item submitted and allow ample margins all around. All manuscripts MUST BE DOUBLE SPACED THROUGHOUT. For articles this in- cludes title (all caps, centered), author names (all caps, centered), addresses (caps and lower case, centered), abstract, 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, and number each page. Do not use a separate cover page, “erasable” paper, or footnotes. Table captions should include all information relevant to tables. All measurements should be in metric units. Line copy illustrations should be clean and legible, proportioned (including captions) to the MADRONO page, and designed for reduction to % original size. 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 latitude and longitude references. Halftone copy should be designed for reproduction at actual size. In no case should original illustrations be sent prior to the acceptance of a manuscript. When needed they should be mounted on stiff cardboard and sent flat. No illustrations larger than 22 < 28 cm will be accepted. Presentation of nomenclatural matter (accepted names, synonyms, typification) should follow the format used for Rhus integrifolia in MADRONO 22:288. 1974. Institutional abbreviations in specimen citations should follow Holmgren and Keuken, Index Her- bariorum, 6th edition. Abbreviations of serial titles should be those in Botanico-Per- iodicum-Huntianum (Lawrence et al., 1968, Hunt Botanical Library, Pittsburgh). If the correct abbreviation cannot be determined, the full serial title should be used. Titles of books should be given in full, together with the place and date of publication, publisher, and edition, if other than the first. All members of the California Botanical Society are allotted eight pages in MADRONO over a two year period. Joint authors are each assessed the full page number. Beyond that number of pages a required editorial fee of $40.00 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 sched- ule, with equity among all members for access to its pages. Printer’s fees for illustrations and typographically difficult material @ $35.00 per page (if their sum exceeds 30 percent of the paper) and for author’s changes after typesetting @ $3.00 per line will be charged to authors. CALIFORNIA BOTANICAL SOCIETY < - “tyr th, LIBRARIES NVINO 417 OILNLILSNI NOILALILSNI ~ LIBRARIES r. NVINOSHLINS S31YVUSIT LI NOILN LIBRARIES SMITHSONIAN INSTITUTION N ‘gia LIBRARIES S3IYVUGIit TION NOILNLILSNI LIBRA NVINOSHLIWS SMITHSONIAN ait RIES NOILNLILSNI TION NOILNLILSNI INSTITUTION LIBRARIES OSHLINS S3ZIYVYgIT LI HSONIAN BRARIES_ SMITH INSTITUTION NOILNLILSNI BRARIES SMITHSONIAN NVINOSHLIWS “ SMITHSONIAN NVINOSHLIWS a SN SMITHSONIAN SMITHSONIAN S$ S31IuVYgIT LIBRARIES INSTITUTION NOILNLILSNI so aN onl SMITHS °o ia) t @) ‘y ® NOILNLILSNI SMITH: NVINOSHLINS S31} LIBRARIES LIBRARIES SJIYVYUGIT LIBRARIES SMITHSONIAN _INST INSTITUTION NOILALILSNI INSTITUTION NOILANLILSNI c Fe w “fj ow 2 Giyy, * »> Ym, , ee) , a) Oo Oo ZINSTITUTION NOILALILSNI_ NVINOSHLINS Sark < * = 1 < = = X =i 4%%.= = 4 e\yi &? : 9 SF SWS S JE bE 2 5 £6 = wy 2G fi = 27 = Ss = = = SSIUVYGIT_LIBRARIES SMITHSONIAN _INSTI m7 uw rr =~ = = ea oe Boy a a < oo = = = Oo = Oo ws aud =a INSTITUTION NOILALILSNI NVINOSHLIWS S3Ie = oO ae ‘@) oo tj = 5 2 5 FE itd = r = = Z Hs Z S3JIYVYEIT LIBRARIES INSTI Pd ; YW) a. as ys = < = 8 = zs g 2 g Ey =i = Z i > = > = ae w * at 7p) INSTITUTION NOILALILSNI_NVINOSHLIWS = S31d uJ z Sy. WwW < = BE < ce Oe —- ‘ faa) — : [oa) 5 S ne SAIYVYSIT LIBRARIES SMITHSONIAN INSTI - S ie is w al y jes) 2 2 Fh = tog aa E 4 i D z oO INSTITUTION NOILNLILSNI” NVINOSHLINS S3I1¥ = —_ & ow a) NO poms YY Aes 2 Ri f Oo WS a ily, O [3 ay. oe a WY w LIE: n (E jn Ff ly = “Sur 2 Gg = SX 2 eye TD GL, Vp = Non oc. y = “GB | = ~ > XY my s : : y —- e wn” a LIBRARI ES SMITHSONIAN INSTITUTION NOILQLILSNI NVINOSHLIWS = S3 l\yYVvua wn 2 > a WwW YY) eae WwW = “S = “ LXE 5 = SS >" = > ae Ww) cae 2 7p) ee NOIL _NWINOSHLINS S3IYVYUGIT LIBRARIES INSTITUTIC | Zz nf ae a Z ” S. = w us w | = fared = = 2 WS: : z = 2 NOS" o = a = ; o a re) as oO , = ea z J = LIBRARI ES _ SMITHSONIAN INSTITUTION NOILALILSNI_ NVINOSHLIWS S3IY¥vVud ; z o ae, 3 3 &< 2 ; EF ve Lae by = = i =) Ye. = > > EK tf, > re > = - , = - ie - a vu? é wn se 7) ; ater o z om _ 2 NOILNLILSN!I NVINOSHLINS S3IYVYUGIT LIBRARIES SMITHSONIAN ) oa w” = wee ” z = < = pate = = 5 z 4 ysl £ Te = z : aK a NE 2 5 f= Z “py = = = > = > a = os = z w za 7] s Bd ” . SMITHSONIAN INSTITUTION NOILONLILSNI NVINOSHLINS SAIYVUE : o 3 Y oe a 9 f aM a w a 4 oc. a oc = a _{ = > rs > = = 2 2 Ge = Z is Ce e Z sh LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S3IYVUE a . 2 ~STITY = ‘ z z 2 Li — A == \ v% = = FE ‘3G NO = 7 = « ay p\ Zz \, a PZ Z [KAW = Wi Ss & > J \\ ih tify, ~ 4 5 Wy BE Ww = (AR) 8 SNUQ FE 97G SAM) ET HF SMITHSONIAN INSTITUTION LIBRARIES LACE Base siege ae ei B = P a . ‘ so Fu a git . : “i . ‘ eeow dors oi padi a tate Papas WANT MES hela tee he ete pe bea 1 ra se . aS are kane wee ae Ae a She 3 Fifa gayeo pe SSE oe Pr ee aN OS ath gt aig aint: