VOLUME 59, NUMBER 3 JULY-SEPTEMBER 2012 MADRONO A WEST AMERICAN JOURNAL OF BOTANY CONTENTS POSTFIRE CHAPARRAL REGENERATION UNDER MEDITERRANEAN AND NON-MEDITERRANEAN CLIMATES Jon E. Keeley, C. J. Fotheringham, and Philip W. Rundel.................00066+ 109 MORPHOLOGICAL AND ISOENZYME VARIATION IN RHODODENDRON OCCIDENTALE (WESTERN AZALEA) (SECTION PENTANTHERA; ERICACEAE) Goode ATUSG visor Ge ncieinticvenccts Lt i AOI ick 128 MEASUREMENT OF SPATIAL AUTOCORRELATION OF VEGETATION IN MOUNTAIN MEADOWS OF THE SIERRA NEVADA, CALIFORNIA AND WESTERN NEVADA Dave A. Weixelman and Gregg M. Riegel ....ccccccccccccccccceeeeeeesttntteteeeteeeeees 143 REAPPEARANCE OF THE VANISHING WILD BUCKWHEAT: A STATUS REVIEW OF ERIOGONUM EVANIDUM (POLYGONACEAE) Naomi S. Fraga, Elizabeth Kempton, LeRoy Gross, and Duncan Bell ... 150 NEW SPECIES PTYCHOSTOMUM PACIFICUM (BRYACEAE), A NEW FEN SPECIES FROM CALIFORNIA, OREGON, AND WESTERN NEVADA, USA John R. Spence aad James RN SREVOCKs 04. .c......0.0. Riese ccececescesescceseees 156 A NEw COMBINATION IN LINANTHUS (POLEMONIACEAE) FROM IDAHO AND OREGON Joanna Le Schultz and RODerEPAICTSON occa ihoeesd ee eat eee 163 BOOK REVIEW RESEARCH & RECOVERY IN VERNAL POOL LANDSCAPES VUE IG ULI OLITIS = ah ete Or a er elias reso Ming Dees oy ie eh ran tse la lobeea iene 164 NOTEWORTHY INLINE IN re cet ears tins MAC pstscas ate ee a ep cha cine a woes B haseadaw a eae aes isloel dsantee’ 166 COLLECTIONS CFSE TIE ORIN Os Ph O80 5 htt IO ws hee nha atte esc ac Sam aaloa eons SADA TAA sna 167 PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAI end hg MADRONO (ISSN 0024-9637) is published quarterly by the California Botanical Society, Inc., and is issued from the office of the Society, Herbaria, Life Sciences Building, University of California, Berkeley, CA 94720. Subscription information on inside back cover. Established 1916. Periodicals postage paid at Berkeley, CA, and additional mailing offices. Return requested. POSTMASTER: Send address changes to MADRONO, Kim Kersh, Membership Chair, Uni- versity and Jepson Herbarium, University of California, Berkeley, CA 94720-2465. kersh @berkeley.edu. Corresponding Editor—Matt RITTER Copy Editor—RICHARD WHITKUS Biological Sciences Department Department of Biology Cal Poly, San Luis Obispo Sonoma State University | Grand Avenue 1801 E. Cotati Avenue San Luis Obispo, CA 93407 Rohnert Park, CA 94928-3609 madronoeditor @ gmail.com whitkus @ sonoma.edu Book Editor—MattT RITTER Noteworthy Collections Editor—DIETER WILKEN Board of Editors Class of: 2012—GRETCHEN LEBUMN, San Francisco State University, CA ROBERT PATTERSON, San Francisco State University, CA 2013—ERIc ROALSON, Washington State University, WA KRISTINA SCHIERENBECK, California State University, Chico, CA 2014—BRANDON PRATT, California State University, Bakersfield, CA Tom WENDT, University of Texas, Austin, TX CALIFORNIA BOTANICAL SOCIETY, INC. OFFICERS FOR 2011—2012 President: V. Thomas Parker, Department of Biology, San Francisco State University, San Francisco, CA 94132, parker @sfsu.edu First Vice President: Andrew Doran, University and Jepson Herbaria, University of California, Berkeley, CA 94720, andrewdoran @ berkeley.edu Second Vice President: Vacant. Recording Secretary: Michael Vasey, Department of Biology, San Francisco State University, San Francisco, CA 94132, mvasey @sfsu.edu Corresponding Secretary: Anna Larsen, Jepson Herbarium, University of California, Berkeley, CA 94720, secretary @calbotsoc.org Treasurer: Thomas Schweich, California Botanical Society, Jepson Herbarium, University of California, Berkeley, CA 94720, tomas @schweich.com The Council of the California Botanical Society comprises the officers listed above plus the immediate Past President, Dean Kelch, Jepson Herbarium, University of California, Berkeley, CA 94720, dkelch @berkeley.edu; the Membership Chair, Kim Kersh, University and Jepson Herbaria, University of California, Berkeley, CA 94720, kersh@berkeley.edu; the Editor of Madrono; and three elected Council Members: Chelsea Specht, Department of Plant and Microbial Biology, University of California, Berkeley, CA 94720-2465, cdspecht @berkeley.edu; Ellen Simms, Department of Intergrative Biology, 1005 Valley Life Sciences Bldg., #3140, University of California, Berkeley, CA 94720, esimms @berkeley.edu. Staci Markos, University and Jepson Herbaria, University of California, Berkeley, CA 94720, smarkos @berkeley.edu. Graduate Student Representatives: Genevieve Walden, Department of Integrative Biology and Jepson Herbarium, University of California, Berkeley, CA 94720, gkwalden@ gmail.com. Administrator: Lynn Yamashita, University of California, Berkeley, CA 94720, admin @calbotsoc.org. Webmaster: Ekaphan (Bier) Kraichak, University of California, Berkeley, CA 94720, ekraichak @ gmail.com. This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). eee EEE MADRONO, Vol. 59, No. 3, pp. 109-127, 2012 POSTFIRE CHAPARRAL REGENERATION UNDER MEDITERRANEAN AND NON-MEDITERRANEAN CLIMATES JON E. KEELEY!”, C. J. FOTHERINGHAM!”? AND PHILIP W. RUNDEL? 'U.S. Geological Survey, Western Ecological Research Center, Sequoia-Kings Canyon Field Station, Three Rivers, CA 93271 °Department of Ecology and Evolutionary Biology, University of California, Los Angeles, CA 90095 jon_keeley@usgs.gov ABSTRACT This study compares postfire regeneration and diversity patterns in fire-prone chaparral shrublands from mediterranean (California) and non-mediterranean-type climates (Arizona). Vegetation sampling was conducted in tenth hectare plots with nested subplots for the first two years after fire. Floras in the two regions were compared with Jaccard’s Index and importance of families and genera compared with dominance-diversity curves. Although there were 44 families in common between the two regions, the dominant families differed; Poaceae and Fabaceae in Arizona and Hydrophyllaceae and Rosaceae in California. Dominance diversity curves indicated in the first year a more equable distribution of families in Arizona than in California. Woody plants were much more dominant in the mediterranean climate and herbaceous plants more dominant in the bimodal rainfall climate. Species diversity was comparable in both regions at the lowest spatial scales but not at the tenth hectare scale. Due to the double growing season in the non-mediterranean region, the diversity for the first year comprised two different herbaceous floras in the fall and spring growing seasons. The Mediterranean climate in California, in contrast, had only a spring growing season and thus the total diversity for the first year was significantly greater in Arizona than in California for both annuals and herbaceous perennials. Chaparral in these two climate regimes share many dominant shrub species but the postfire communities are very different. Arizona chaparral has both a spring and fall growing season and these produce two very different postfire floras. When combined, the total annual diversity was substantially greater in Arizona chaparral. Key Words: Climate, dominance, fire, species diversity, spring and fall annuals. Chaparral is a fire-prone evergreen shrubland that is the dominant vegetation in the mediterra- nean-type climate (MTC) region of California (Keeley 2000). From a global perspective this vegetation is somewhat unique in that it not only tolerates frequent fires but many of the species have fire-dependent reproduction, similar to shrublands in other mediterranean-climate regions (Rundel 1981; Keeley and Bond 1997; Keeley et al. 2005). Chaparral shrublands, however, are not re- stricted to MTCs as this vegetation type is widely distributed in southwestern USA and disjunct to northeastern Mexico (Keeley and Keeley 1988). Arizona mirrors the MTC in the winter rains that taper off to a late spring drought, but departs from the MTC by addition of a second rainy season in the summer. However, the importance of summer rains to the dominant shrubs is a matter of some debate, as it appears that these rains play a minor role in shrub growth and reproduction (Vankat 1989). Northeastern Mex- ico has a winter drought and summer rain climate, but the physiological responses of the shrub dominants to drought are remarkably similar to Californian shrubs (Bhaskar et al. 2007). Arizona chaparral covers more than two million hectares (Schmutz and Whitham 1962; Bolander 1982) and occurs in widely disjunct patches from near Prescott in the northeast to the southeastern mountains around Tucson and east to the southwestern edge of New Mexico (Cable 1975; Pase and Brown 1982; Whittaker and Niering 1964, 1965). Arizona and California are broadly similar in the shrub dominants that are shared between these two regions (Knipe et al. 1979). Northeast Mexico chaparral 1s restricted to patches of severe substrate in the Sierra Oriental Moun- tains south of Monterey, and this vegetation shares some of the same shrub species and genera as the Arizona and California chaparral communities. The Arizona and Mexican chaparral are of interest for what they can potentially tell us about the evolution of chaparral taxa. Paleoecological studies have suggested that many chaparral shrub species originated in interior portions of the southwest (Wolfe 1964, Axelrod 1989) and contem- porary populations in Arizona and northeastern Mexico are interpreted as remnants of a Tertiary chaparral like vegetation that comprises taxa that largely originated under non-medterranean type climates (Ackerly 2009; Keeley et al. 2012). Although postfire chaparral responses have been studied in great detail in the winter rain region of California, little 1s known about 110 MADRONO Annual Precipitation (mm) S81 oo Se aN © Summer Precipitation (%) 0 Fic. 1. Patterns of (a) long-term average annual precipitation and (b) proportion falling during the ‘summer’ rainy season (defined as July, August and September) for the nearest stations to the Arizona (AZ) and California (CA) study sites used in this project. Error bars are the standard error of the mean. community responses in the bimodal rainfall region of Arizona. Postfire regeneration of Arizona chaparral has largely focused on shrub responses with relatively little attention to com- munity responses and regeneration strategies of other life forms (Pase and Pond 1964; Pase 1965; Carmichael et al. 1978). The purpose of this study was to contrast postfire recovery in the MTC California chaparral with postfire recovery in the non-MTC Arizona chaparral. We utilized data from studies of 2003 wildfires in California (same sites as in Keeley et al. 2008) and from studies of 2002 wildfires in Arizona (same sites as in Fotheringham 2009). METHODS Study Sites The Arizona sites were burned in the late spring and summer of 2002 and were distributed across six fires in southeastern Arizona and southwestern [Vol. 59 TABLE |. FIFTEEN DOMINANT PLANT FAMILIES IN ARIZONA AND CALIFORNIA POSTFIRE CHAPARRAL SITES BASED ON AERIAL COVERAGE. State/family Normalized cover Arizona Poaceae 1.00 Fabaceae 0.676 Asteraceae 0.327 Verbenaceae 0.138 Fagaceae 0.100 Molluginaceae 0.085 Convolvulaceae 0.074 Liliaceae 0.055 Malvaceae 0.054 Euphorbiaceae 0.043 Geraniaceae 0.032 Boraginaceae 0.025 Agavaceae 0.025 Rhamnaceae 0.018 Krameriaceae 0.016 California Hydrophyllaceae 1.00 Rosaceae 0.704 Cistaceae 0.345 Ericaceae 0.319 Fabaceae 0.254 Convolvulaceae 0.240 Rhamnaceae 0.237 Liliaceae 0.228 Asteraceae 0.174 Boraginaceae 0.168 Fagaceae 0.163 Poaceae 0.143 Papaveraceae 0.094 Scrophulariaceae 0.092 Fumariaceae 0.089 New Mexico (Fotheringham 2009). This study included 40 sites that were selected based on evidence of chaparral vegetation present prior to fire, fire size, range of fire severities, and accessi- bility, and were sampled in the first two postfire years. Sites were grouped by fire for analysis, except due to the small size and proximity of the Merritt and Ryan fires these were grouped together, and due to the large size of the Bullock Fire these were separated into two groups, the lower elevation Bullock and the higher elevation Upper Bullock. California sites were from five fires that burned in autumn 2003 and included 250 sites that were dominated by chaparral prior to the fires and sampled over the first two years; due to their proximity, the Grand Prix and Old fires were analyzed as a single fire. Both Arizona and California fires were distributed across a range of about 150-200 km but the former were distributed at about the same latitude in a west to east gradient and the latter along a north to south gradient (see Keeley et al. 2008 and Fotheringham 2009 for detailed maps). Chaparral sites studied in Arizona were at significantly higher elevation (AZ sites = 1620 m, CA sites 785 m). (a) KEELEY ET AL.: CALIFORNIA VS ARIZONA CHAPARRAL 111 (b) California 10 20 30 40 350 60 Plant family order 100 0 20 40 60 80 California 100 Genus order 2012] Arizona oy} WY) o) Oo) fe) © > O O 0 O 10 20> 30 .-A0 50 60 0 1.000 oar peo: ‘@)) Oo > 0.010 O 0.001 0 40 60 80 Fic. 2. ground surface cover. Field Methods In Arizona, there were two growing seasons following both the summer rains and the winter rains so sampling was first done in fall 2002 and then again in spring 2003, and this sampling regime was repeated for a second year. In the winter-rainfall chaparral of California sampling was conducted only in spring of 2004 and 2005. Each site consisted of a 20 m X 50 m sample plot, positioned parallel to the slope contour, which is considered appropriate for capturing the greatest variation in community composition (Keeley and Fotheringham 2005). Each of these tenth hectare sites were subdivided into 10 nested 100 m/? Rank order distribution of (a, b) plant families and (c, d) genera in Arizona and California. GSC is square subplots, each with a single nested | m° square quadrat in an outside corner. Cover and density were recorded for each species within the quadrats, and a list of additional species was recorded from the surrounding subplot. Cover was visually estimated and a percentage of ground surface covered was recorded for each species. Density was recorded for each species with counts where density was less than approx- imately 25 individuals per quadrat, and with estimates at higher densities. Seedlings and resprouts of the same species were counted and recorded separately. Vouchers were collected for all specimens and have been deposited in the herbarium in J. Keeley ’s laboratory. All plant [12 MADRONO [Vol. 59 TABLE 2. Top 75 NATIVE GENERA IN THE ARIZONA AND CALIFORNIA POSTFIRE SITES BASED ON AERIAL COVERAGE. Arizona Normalized cover Arizona Normalized cover Eragrostis 1.000 Amaranthus 0.029 Dalea 0.609 Portulaca 0.029 Glandularia 0.301 Bothriochloa 0.029 Bouteloua 0.258 Plagiobothryus 0.028 Bidens O22); Arctostaphylos 0.027 Quercus 0.218 Schinus 0.026 Muhlenbergia 0.217 Cryptantha 0.024 Lotus 0.200 Juniperus 0.024 Mollugo 0.187 Marina 0.022 Heterosperma 0.166 Gnaphalium 0.022 Mimosa 0.154 Boerhavia 0.022 Chamaecrista 0.150 Phaseolus 0.017 Melampodium 0.142 Chamaesyce 0.016 Urochloa 0.136 Yucca 0.015 Nolina 0.117 Cyperus 0.014 Aristida 0.105 Digitaria 0.013 Desmodium 0.104 Aeschynomene 0.013 Ipomoea 0.102 Anoda 0.013 Heliomeris 0.098 Opuntia 0.013 Panicum 0.097 Descurainia 0.013 Calliandra 0.089 Drymaria 0.012 Leptochloa 0.088 Eriogonum 0.011 Acalypha 0.073 Gymnosperma 0.010 Erodium 0.070 Commelina 0.010 Evolvulus 0.060 Dyssodia 0.010 Sida 0.056 Triticum 0.010 Sphaeralcea 0.044 Diodia 0.009 Lycurus 0.043 Gilia 0.009 Astragalus 0.043 Scleropogon 0.009 Schizachyrium 0.043 Garrya 0.009 Elionurus 0.040 Sanvitalia 0.009 Ceanothus 0.040 Linum 0.009 Crotalaria 0.038 Trachypogon 0.008 Krameria 0.036 Agave 0.008 Cathestecum 0.034 Hackelochloa 0.008 Prosopis 0.033 Erigeron 0.007 Dasylirion 0.031 Salvia 0.007 Chenopodium 0.030 California Normalized cover California Normalized cover Adenostoma 1.000 Allophyllum 0.026 Phacelia 0.379 Solanum 0.025 Calystegia 0.364 Silene 0.025 Arctostaphylos 0.354 Cercocarpus 0.023 Lotus 0.331 Malacothamnus 0.023 Ceanothus 0.280 Styrax 0.023 Helianthemum 0.261 Navarretia 0.023 Cryptantha 0.256 Hypochoeris 0.022 Quercus 0.248 Garrya 0.020 Emmenanthe 0.245 Gastridium 0.020 Calochortus 0.145 Calyptridium 0.019 Eriodictyon 0.135 Mimulus 0.018 Xylococcus 0.131 Erodium 0.018 Chlorogalum 0.100 Lepechinia 0.017 Antirrhinum 0.098 Helianthus 0.017 Marah 0.082 Nassella 0.012 Rhamnus 0.081 Penstemon 0.011 Chaenactis 0.077 Lupinus 0.010 Camissonia 0.075 Galium 0.010 Dicentra 0.068 Monardella 0.010 Dendromecon 0.061 Lomatium 0.009 Malosma 0.055 Erigeron 0.009 Salvia 0.055 Fremontodendron 0.009 2012] TABLE 2. California Normalized cover Dichelostemm 0.053 Cneoridium 0.051 Chamaebatia 0.046 Eriophyllum 0.042 Zigadenus 0.042 Pickeringia 0.041 Nemocladus 0.037 Gilia 0.037 Yucca 0.036 Pterostegia 0.036 Brassica 0.035 Hazardia 0.035 Filago 0.033 Rhus 0.031 Trichostema 0.031 Mentzelia 0.028 nomenclature follows Hickman (1993) for Cali- fornia and USDA (2009) for Arizona. Precipitation data for Arizona were obtained from http://cdo.ncde.noaa.gov/CDO/ data prod- uct (accessed May 2008) for climate stations nearest to the study sites. Precipitation data for California were obtained from the Western Regional Climate Center (http://www.wrcc.dri.edu/ summary/Climsmsca.html; accessed April 2007). Average precipitation for the sites in Arizona and California were comparable (Fig. la). Both regions have significant winter rains followed by a late spring and early summer drought. In California drought continues until late fall whereas Arizona has summer rains that begin in July and extend through September. A substantial proportion of total rain occurs during the ‘summer’ (July, August and September) in Ar- izona in contrast to California (Fig. 1b). Data Analysis Statistical comparisons and regressions were calculated and displayed graphically with Systat 11.0 (Richmond, CA, USA). Comparisons _ be- tween Arizona and California were made with a two-tailed t-test for all quantitative parameters. Compositional differences between sites within a region and between regions were evaluated using Jaccard’s similarity coefficient, which pro- vides a measure of similarity between two sets of data. This coefficient was calculated using a modified form of Jaccard’s index (see Table 10.2 in Mueller-Dombois and Ellenberg 1974), based on presence/absence as: MC No) = ae MA+ MB x 100 where MC is the number of taxa present in both regions, MA is number of taxa present only in Arizona and MB is for taxa present only in KEELEY ET AL.: CALIFORNIA VS ARIZONA CHAPARRAL 1s CONTINUED. California Normalized cover Daucus 0.008 Melica 0.007 Papaver 0.007 Cupressus 0.007 Selaginella 0.007 Elymus 0.007 Muilla 0.007 Apistrum 0.007 Ribes 0.006 Eriogonum 0.006 Lonicera 0.005 Claytonia 0.005 Chorizanthe 0.004 Pellaea 0.004 Leymus 0.004 California, and the coefficient expressed as a percentage. The value ranges from 0%, where the two data sets share no taxa, to 100% with complete overlap in taxa. This index was calcu- lated for all plant families and all genera shared between sites within a region and between regions, 1.e., for all pairwise comparisons of sites within Arizona, and within California and then for all comparisons between Arizona sites and Califor- nia sites. The non-parametric Wilcoxon signed ranks test was used to compare the Jaccard’s indices calculated within Arizona to those calcu- lated between Arizona and California sites to determine if Arizona sites were more similar to one another than they were to California. RESULTS Taxonomic Patterns Between the Arizona and California sites there were 44 plant families in common and an additional 19 families recorded just at the Arizona sites and nine just at the California sites (Appendix 1). Based on total cover over the two years of study in both Arizona and California, the top 15 families were quite different (Table 1). Although about half of the top 15 families were shared between both regions, the most dominant families were different. In Arizona the top two families were the Poaceae and Fabaceae whereas in California it was the Hydrophyllaceae and Rosaceae. Families were generally more evenly distributed in California than in Arizona, as illustrated by the observation that the top 15 families were present in sites at all fires in California, whereas in Arizona only the top 10 families were represented at all fires. A similar difference between regions is illus- trated by the pattern of equitability in rank order distribution of families (Fig. 2a, b). In Arizona 114 MADRONO [Vol. 59 TABLE 3. SPECIES FOUND IN BOTH THE ARIZONA AND CALIFORNIA STUDY SITES. This is not meant to suggest these are the only species found in chaparral of the two regions but just what was recorded from our 40 study sites in Arizona and 250 sites in California. Sambucus mexicana C. Presl. ex DC. Annuals Allophyllum gilioides (Benth). A. D. Grant & V. E. Grant Polemoniaceae Aristida adscensionis L. Poaceae Bowlesia incana Ruiz & Pav. Apiaceae Chenopodium berlandieri Mogq. Chenopodiaceae Calandrinia ciliata (Ruiz & Pav.) DC. Portulacaceae Conyza canadensis (L.) Cronquist Asteraceae Cryptantha micrantha (Torr.) I. M. Johnst. Boraginaceae Cryptantha muricata (Hook & Arn.) A. Nelson & J. F. Macbr. Boraginaceae Daucus pusillus Michx. Apiaceae Eriastrum sapphirinum (Eastw.) H. Mason Polemoniaceae Galium aparine L. Rubiaceae Gilia leptantha Parish Polemoniaceae Lepidium virginicum L. Brassicaceae Lotus humistratus Greene Fabaceae Lupinus concinnus J. Agardh Fabaceae Lupinus sparsiflorus Benth. Fabaceae Malocothrix clevelandii A. Gray Asteraceae Pectocarya setosa A. Gray Boraginaceae Phacelia distans Benth. Hydrophyllaceae Phlox gracilis (Hook.) Greene Polemoniaceae Rafinesquia californica Nutt. Asteraceae Silene antirrhina L. Caryophyllaceae Stephanomeria exigua Nutt. Asteraceae Stellaria nitens Nutt. Caryophyllaceae Stylocline gnaphalioides Nutt. Asteraceae Thysanocarpus curvipes Hook. Brassicaceae Triodanis biflora (Ruiz & Pav.) Greene Campanulaceae Triodanis perfoliata (L.) Nieuwl. Campanulaceae Vulpia microstachys (Nutt.) Munro Poaceae Vulpia octoflora (Walter) Rydb. Poaceae Yabea microcarpa (Hook. & Arn.) Koso-Pol. Apiaceae Herbaceous perennials Aristida purpurea Nutt. Poaceae Astragalus trichopodus (Nutt.) A. Gray Fabaceae Bothriochloa barbinodis (Lag.) Herter Poaceae Datura wrightii Regel Solanaceae Dichelostemma capitatum (Benth.) Alph. Wood Amaryllidaceae Gnaphalium bicolor Anderb. Asteraceae Gnaphalium canescens DC. Asteraceae Subshrubs or suffrutescents Atriplex semibaccata R. Br. Chenopodiaceae Brickellia californica (Torr. & A. Gray) A. Gray Asteraceae Encelia farinosa A. Gray ex Torr. Asteraceae Eriogonum wrightii Torr. ex Benth. Polygonaceae Gutierrezia sarothrae (Pursh) Britton & Rusby Asteraceae Porophyllum gracile Benth. Asteraceae Rhus trilobata Nutt. Anacardiaceae Senecio flaccidus Less. Asteraceae Solanum douglasii Dunal Solanaceae Yucca schidigera Roezl ex Ortgies Liliaceae Shrubs Arctostaphylos pungens Kunth Ericaceae Artemisia tridentata Nutt. Asteraceae Baccharis salicifolia (Ruiz & Pav.) Pers. Asteraceae Ceanothus greggii A. Gray Rhamnaceae Garrya wrightii Torr. Garryaceae Rhamnus crocea Nutt. Rhamnaceae Caprifoliaceae 2012] TABLE 3. Annual species alien to Arizona and California Ambrosia artemisiifolia L. Brassica nigra (L.) W. D. J. Koch Bromus madritensis L. Bromus tectorum L. Chenopodium album L. Erodium cicutarium (L.) L’>Hér. ex Aiton Lactuca serriola L. Marrubium vulgare L. Phalaris minor Retz. Poa annua L. Schismus barbatus (L.) Thell. Sisymbrium altissimum L. Sisymbrium irio L. Salsola tragus L. Sonchus tenerrimus L. Vulpia bromoides (L.) A. Gray there was a much larger difference between the top few families and the remaining families, whereas in California cover was somewhat more equally distributed among families. For example two orders of magnitude cover below the top family in Arizona comprised only about 20 families, whereas in California it was almost double that number. Unlike the family distribution, where the majority were in common between regions, many fewer genera were common between regions than were unique to one or the other region (Appendix 2). There were only 109 genera recorded from sites in both regions, but there were 148 genera recorded just in Arizona sites and 78 recorded just in California sites. Based on cover, of the top 75 genera in Arizona, only 16 were also in the top 75 in California and of the top 75 in California only 22 were in the top group in Arizona (Table 2). Genera important in both regions (defined as in the top 75) include the shrubs Arctostaphylos, Ceanothus, Garrya, and Quercus, subshrubs Baccharis, Eriogonum, Salvia, and Yucca, and suffrutescents Erigeron and Lotus (Table 2). The cover distribution for the top genera (Fig. 2c, d) followed similar curves in Arizona and California, indicating greater equi- tability in both regions than observed with families. In Arizona there were substantially more species recorded from the 40 study sites (577) than for the 250 sites in California (439) despite covering a roughly similar-sized geographical area. However, in Arizona this covered an east- west gradient and in California a north-south gradient. As a general rule none of the dominant herbaceous species in Arizona chaparral were present or well represented in California chap- arral, and vice versa. One of the most conspic- uous and widespread postfire species in Arizona KEELEY ET AL.: CALIFORNIA VS ARIZONA CHAPARRAL 1D bs. CONTINUED. Asteraceae Brassicaceae Poaceae Poaceae Chenopodiaceae Geraniaceae Asteraceae Lamiuaceae Poaceae Poaceae Poaceae Brassicaceae Brassicaceae Chenopodiaceae Asteraceae Poaceae was the fall germinating ephemeral herbaceous perennial Verbena bipinnatifida Nutt. (=Glandu- laria b.), a species not found in the California postfire chaparral. In California the most con- spicuous postfire ephemerals were Hydrophylla- ceae, most of which were absent or of very minor importance in Arizona chaparral. However, there were more than 30 minor spe- cies in common between both regions (Table 3); e.g., Allophyllum gilioides A.D. Grant & V.E. Grant, Calandrinia ciliate (Ruiz & Pav.) DC., Lupinus sparsiflorus Benth., Malacothrix cleve- landii A. Gray, and Rafinesquia californica Nutt., all of which are spring annuals. Of the her- baceous perennials the one that stands out as being very common in both regions was Diche- lostemma capitatum (Benth.) Alph. Most of the subshrubs listed were widespread but never locally common. Several shrubs were widespread in both regions, in particular Arctostaphylos pungens Kunth, Ceanothus greggii A. Gray and Rhamnus crocea Nutt. More than 15 alien species were common between both regions, and all were annuals (Table 3). The main shrub species in Arizona were seedlings of the obligate seeders Arctostaphylos pungens, Ceanothus greggii and C. fendleri A. Gray and resprouts of Quercus turbinella Greene, Rhus trilobata Nutt., and Baccharis salicifolia (Ruiz & Pav.) Pers. Postfire sites in California were dominated by resprouts and seedlings of Adenostoma fasciculatum Hook. & Arn, Arcto- staphylos spp., and Ceanothus spp. as well as resprouts of Quercus berberidifolia Liebm. Sub- shrubs and other less woody and _ shorter-lived suffrutescents were very different between these regions. In Arizona the genus Dalea was very important as well as Krameria erecta Willd. ex Schult. and species of Senecio and Solanum, but this niche was filled largely by Lotus scoparius (Nutt.) Ottley, Helianthemum scoparium Nutt. 116 MADRONO [Vol. 59 Year 1 Year 2 (a) P <0.001 i¢ 2) OQ =) 1 A ee ep) -_—" <@) Nd eB) ® 3 2 oO Ff oOo 2 Oo os ao # _. > ¢ 2B) <@) Cc ee oO & > § ne = —--—s—COC© —- YQ So Se Oo)6UCUS > a | =a 4) - 2 oO © LT a fa) =) ce Cc L Fic. 3. Plant cover in spring of year | and year 2 presented by life form (AZ = Arizona, CA = California, subshrub category includes the weakly woody suffrutescents). Error bars are the standard error of the mean. Note the scale for annuals is double that for other life forms. and Calystegia macrostegia (Greene) Brummitt in Postfire Changes in Cover and Diversity California. Herbaceous floras were very different between the two regions with Poaceae dominating Total cover was around 20% in the first in Arizona and Hydrophyllaceae in California. postfire year and not significantly different P=0.125 P<0.001 P=0.517 P<0Q.001 P<0.001 P= 0.002 2012] Year 1 80 A OW E (>) S 40 = 120 i¢ 2) ¢p) ® = N a = eras (-) Go Je oO ~ O tr ® (ox “”) N E tb Arizona Spg Spg +Fall Fic. 4. KEELEY ET AL.: CALIFORNIA VS ARIZONA CHAPARRAL Liz Year 2 P<0.001 P<0Q.001 P<0Q.001 P=0.443 Arizona CA Spg Spg Spg +Fall Species diversity (including native and non-native species) for the spring flora and total flora in Arizona and spring flora in California at three scales in both years (Arizona, CA = California, Spg = spring). Two-tailed t-test for (top P-value) Arizona spring vs. California spring and (bottom P-value) Arizona total vs. California spring. Error bars are the standard error of the mean. between Arizona and California, but in the second year cover more than doubled in Califor- nia and was significantly greater than in Arizona (P < 0.001). These two regions differed markedly in the importance of different growth forms. Shrub cover was about five times greater in California than in Arizona in the first two postfire years (Fig. 3a, b), and subshrubs also had significantly greater cover in California (Fig. 3c, d). In contrast, herbaceous species, both perennials (Fig. 3e, f) and annuals (Fig. 3g, h) had significantly more cover in Arizona than in California. By the second postfire year herba- ceous perennials, mostly grasses, had about four times more cover in Arizona and annuals had about twice as much as California. In short, postfire cover in California was more or less equally distributed among different growth forms than in Arizona. Species richness in the first spring after fire was slightly higher in California at the 1 m° scale (Fig. 4e), but not significantly different at larger scales (Fig. 4c, a). However, this does not capture the full annual diversity in Arizona due to the double growing seasons resulting from a bimodal rainfall pattern. As a consequence Arizona produced two different herbaceous floras, one in fall and one in spring. The total first year diversity (fall 2002 plus spring 2003 in Arizona vs. just the spring 2003 flora in California) was significantly higher in Arizona at all scales (Fig. 4a, c, e). 118 MADRONO [Vol. 59 Year 1 w 2 “ , ic ep) es = Oo wn Oo Oo 2 SS ee a: x= 5 ep) 7) 7) cad) = Nom O = og nO Oo ce Sn = oOo c QO o YO oOo Q ¢,) = re) ro) re) oO Pat ; aan o =e N ay = ra ce < ae < o +Fall Arizona CA Spg Spg Spg dd) P<0.001 P< 0.001 Arizona CA Spg Spg Spg +Fall Fic. 5. Species diversity at the site level (tenth ha) for the spring flora and total flora in Arizona and spring flora in California in both years presented by life form (Arizona, CA = California, Spg = spring). Two-tailed t-test for (top P-value) Arizona spring vs. California spring and (bottom P-value) Arizona total vs. California spring. Error bars are the standard error of the mean. In the second postfire spring, diversity rose at all scales in California but not in Arizona so spring floras were significantly more diverse in California at all scales (Fig. 4b, d, f). Total second year diversity was significantly greater in Arizona at the largest spatial scale (Fig. 4b). The contribution of different growth forms to diversity at the site level (tenth ha) differed between regions. In the first spring following fire, woody plant diversity was similar between both regions (Fig. 5a), but when the fall flora in Arizona was added in the total for the year was 2012] Year 1 25 20 p= 6 16 (=) © +f (ep) N O E > S&S Q. —_ (Tp) ~~ c + Oo =e —N = th CA FIG. 6. KEELEY ET AL.: CALIFORNIA VS ARIZONA CHAPARRAL 119 Year 2 Alien species diversity at three scales for the annual total (combined fall and spring for Arizona and spring for California in year 1 and year 2 (Az = Arizona, CA = California). Error bars are the standard error of the mean. Note scale remains the same in all panels. slightly greater in Arizona (Fig. 5a). Subshrub (including suffrutescents) diversity was greater in California than either the spring or spring plus fall total in Arizona (Fig. 5c). However, herba- ceous perennial diversity was greater in Arizona in the spring and even greater when fall diversity was added in (Fig. 5e). Spring annual diversity was similar in both regions (Fig. 5g) but when the fall flora in Arizona was added in (Fig. 5g) diversity of annuals was substantially greater in Arizona. These patterns remained the same in the second year (Fig. 5b, d, f, h). Alien plant cover comprised only a few percent of the total cover in the first year (measured in spring) and was not significantly different be- tween Arizona and California (P = 0.486). However, by spring of the second year, total cover of aliens in California had increased about S-fold and was significantly greater than in Arizona (P = 0.002). Species diversity of aliens was significantly higher at all spatial scales in California in both years (Fig. 6). In the second year alien species comprised 8% of the Arizona flora and 22% of the California flora. Community Similarity Jaccard’s index was used to compare the compositional similarity within regions and between regions. Since few species were important in both Arizona and California the focus was on plant families and genera. Sites were grouped by the six fires in Arizona and the four fires in California and comparisons were made on first year floras. Comparisons of Arizona sites vs. other Arizona sites, and Arizona sites vs. California sites were made for fall and spring. In California, there was only one growing season in the spring, so compar- isons of fall families or fall genera in Arizona were made against spring floras in California. 120 MADRONO [Vol. 59 TABLE 4. PERCENTAGE SIMILARITY BETWEEN BURNS USING PRESENCE/ABSENCE JACCARD’S INDEX. Arizona sites are presented from west to east and California sites from north to south. *In California, there was only one growing season in the spring, so comparisons of fall taxa in Arizona were made against spring taxa in California. AZ - AZ - Upper Bullock Bullock AZ - Merrit/ Ryan Families Fall Az - Spring CA* AZ - Oracle 73 44 53 AZ - Bullock 45 55 AZ - Upper 47 Bullock AZ - Merritt/Ryan AZ- Darnel AZ - Walnut Spring in AZ & CA AZ - Oracle 78 42 46 AZ - Bullock 48 48 AZ - Upper 49 Bullock AZ - Merritt/Ryan AZ- Darnel AZ - Walnut CA - Grand Prix/Old CA - Paradise CA - Cedar Genera Fall Az - Spring CA* AZ - Oracle 54 28 3] AZ - Bullock 31 41 AZ - Upper 26 Bullock AZ - Merritt/Ryan AZ- Darnel AZ - Walnut Spring in AZ & CA AZ - Oracle 54 33 51 AZ - Bullock 33 33 AZ - Upper 30 Bullock AZ - Merritt/Ryan AZ- Darnel AZ - Walnut CA - Grand Prix/Old CA - Paradise CA - Cedar Based on presence/absence the Jaccard’s index for families and genera in fall and spring (Table 4) generally showed that intraregional comparisons of fires (AZ fires vs. AZ fires or CA fires vs. CA fires) were more similar than comparisons be- tween regions (AZ vs. AZCA or CA vs AZCA). In Arizona the two western most sites (Oracle and Bullock) were markedly similar in families and genera but that changed with elevation (Upper Bullock) and in comparison with the eastern most sites (Table 4). In California, sites were much more similar to each other than Che = AZ Grand CA - CA- CA- Darnel Walnut Prix/Old Paradise Cedar Otay 55 56 48 40 46 45 60 61 43 39 44 40 44 42 33 33 34 36 68 60 37 32 35 37 67 40 38 43 38 39 42 44 41 56 62 55 50 58 58 58 67 57 49 59 60 44 35 36 36 37 42 a7 44 29 31 35 33 60 40 38 43 a7 40 38 43 37 70 67 76 80 79 82 37 37 14 9 13 10 31 46 11 10 10 9 3] 30 9 6 6 6 43 45 9 6 7 7 49 8 7 0. 5 10 9 10 8 28 42 17 14 L7. 1S 30 45 18 12 15 13 23 26 14 13 13 ie 30 36 8 8 8 of 42 8 6 7 5 16 14 15 14 49 49 54 63 54 61 observed within Arizona sites, despite being distributed across a similar-sized area. For both families and genera the spring flora in Arizona was much more similar to California’s spring flora than was the fall flora in Arizona. In addition, the western most sites in Arizona (Oracle and Bullock) were more similar to California than the eastern most sites (Darnel and Walnut). To summarize these patterns the average Jac- card’s index is presented for all Arizona site com- parisons, for all California site comparisons and for all comparisons of Arizona and California sites 2012] TABLE 5. COMPOSITIONAL SIMILARITY BETWEEN KEELEY ET AL.: CALIFORNIA VS ARIZONA CHAPARRAL 12] BURNED AREAS USING JACCARD’S INDEX BASED ON PRESENCE/ABSENCE FOR PLANT FAMILIES AND GENERA IN THE FALL AND SPRING, BASED ON TABLE 4. AZ = average of all pairwise comparisons of Arizona sites grouped by fire, CA = DATA IN average of all pairwise comparisons of California sites grouped by fire, AZCA = all pairwise comparisons of Arizona vs. California sites grouped by fire). *In California, there was only one growing season in the spring, so comparisons of fall taxa in Arizona were made against spring taxa in California. Average Jaccard’s percentage similarity P-value for Wilcoxon signed ranks test AZ AZCA CA AZ vs. CA AZ vs. AZCA CA vs. AZCA Families Fall* a) 39 - 0.046 <0.001 - Spring Dp, 46 WD 0.046 0.069 0.028 Genera Fall* 39 2 ~ 0.046 <0.001 - Spring 34 12 aD 0.046 <0.001 0.028 (Table 5). Based on these averages it is apparent, at both the family and genus level, California sites were significantly more similar to one another than were Arizona sites. The average for regional comparisons between Arizona and California was much lower for families and markedly lower for genera than that index calculated within each region. Aliens were not well represented in many Arizona sites and this likely contributed to the fact that at the level of both families and genera, similarity between sites was much less (Table 6) than for the flora as a whole (Table 5). This stands in contrast to the California sites where alien families and genera were quite similar between sites (Table 6). The average similarity within Arizona sites was not significantly differ- ent than the similarity index between Arizona and California sites, whereas California sites had a significantly higher index than that calculated between California and Arizona. DISCUSSION California chaparral occurs under a _ winter rain — summer drought climate in contrast to the TABLE 6. bimodal rainfall pattern characteristic of Arizona chaparral. Although both have winter rains, California sites typically have higher winter rainfall than Arizona sites (40-60% of the annual total in California vs 20-30% in Arizona). These rainfall patterns contribute to differences in fire seasons; Arizona commonly has late spring — early summer fires and the California fire season is largely in the late summer and fall (Keeley 2000), although earlier in years with dry winters (Dennison et al. 2008). In addition to occurring under a different climatic regime, Arizona chaparral tends to be distributed at higher elevations than in California sites, apparently because precipitation regimes conducive to chaparral occur at higher elevations in Arizona than in California (Mooney and Miller 1985). This likely accounts for why the ubiquitous Californian chaparral shrub Adenos- toma fasciculatum is missing from Arizona; in California it drops out of interior sites with cold winters (Keeley and Davis 2007). Arizona and California chaparral communities share many of the same dominant woody species, including species of Arctostaphylos, Baccharis, Ceanothus, Cercocarpus, Eriogonum, Garrya, ALIEN PLANT SIMILARITY BETWEEN BURNED AREAS USING JACCARD’S INDEX BASED ON PRESENCE/ ABSENCE FOR PLANT FAMILIES AND GENERA IN THE FALL AND SPRING, BASED ON SIMILAR COMPARISONS AS SHOWN IN TABLE 4. AZ = average of all pairwise comparisons of Arizona sites grouped by fire, CA = average of all pairwise comparisons of California sites grouped by fire, AZCA = all pairwise comparisons of Arizona vs. California sites grouped by fire. *In California, there was only one growing season in the spring, so comparisons of fall taxa in Arizona were made against spring taxa in California. Average Jaccard’s percentage similarity P-value for Wilcoxon signed ranks test AZ AZCA CA AZ vs. CA AZ vs. AZCA CA vs. AZCA Families Fall* Za 33 - 0.028 0.331 - Spring 26 26 75 0.028 0.950 0.028 Genera Fall* 20 6 _ 0.046 <0.001 — Spring 19 E2 59 0.028 0.022 0.028 122 P11 MADRONO Quercus, Rhamnus, Rhus, and Salvia. In both regions these dominants exhibit similar patterns of postfire recovery including resprouting and seedling recruitment from soil-stored seed banks. Following summer wildfires Arizona chaparral recovers very rapidly in concert with the summer rains that begin usually in July. All resprouting woody species initiate resprouts during this rainy season. In California, resprouting species may begin regrowth soon after fires but this appears to be dependent on soil moisture as it is commonly observed that resprouting in dry years it is delayed until the winter rainy season (Keeley 2000). The phenology of seedling recruitment in Ceanothus and Arctostaphylos shrub species 1s remarkably similar in that it occurs towards the end of the winter rainy season in both regions. Thus, the winter rainfall is one climatic charac- teristic that links these regions in terms of some functional type responses. Another is the spring postfire annual-dominated flora in both regions. However, the ephemeral postfire floras exhibit a number of differences between regions. Most noteworthy is the duel postfire floras in Arizona. Not only do fall and spring rains result in two growing seasons but different floras are produced in fall and spring. The spring flora in Arizona bears a strong systematic resemblance to California whereas the fall flora is quite distinct and has elements that have a more neotropical affinity (Fotheringham 2009). As a consequence, the total yearly diversity is substantially higher in Arizona than in California from small to large scales (Fig. 4). At the community level (1000 m7”) species richness in Arizona is comparable to some of the most species rich communities known from tem- perate latitudes (Keeley and Fotheringham 2003). Another prominent difference between these climatically different regions is the greater 1m- portance of herbaceous perennials in the Arizona chaparral. This is likely tied to the differences in summer drought between the two regions. In California the drought, on average, lasts from late spring to early fall and places a severe stress on survival of perennials, particularly herbaceous perennials. In Arizona the drought is cut short by summer rains and this works to favor surviv- al of herbaceous perennials. One of the most striking differences in the postfire floras between these two regions is the prominence of Poaceae in Arizona, in particular the very diverse and dominant C, bunchgrass flora. The importance of C, bunchgrasses is to be expected in this summer rain climate and their near total absence in California is consistent with what is known about the distribution of C4 grasses (Sage et al. 1999). CONCLUSIONS The number of similarities between California and Arizona are matched by the differences [Vol. 59 between these two regions. The most obvious similarity is that these plant communities share most of the same dominant species as well as a number of genera. The most prominent dissim- ilarity is that Arizona chaparral has both a spring growing season and a fall growing sea- son which results in two very different postfire floras. Other dissimilarities include the promi- nence of perennial grasses in Arizona, which is promoted by the summer rains and perhaps by more open shrublands. This life form is largely nonexistent in California postfire chaparral be- cause spring annuals are far better at persist- ing in landscapes dominated by a long summer drought and the closed canopy chaparral in California excludes herbaceous species to a greater degree than in Arizona. Thus, the pri- mary differences are seen in the herbaceous component of these plant communities and are largely driven by summer rains in one region and absence in another. The results from this study have implica- tions for paleoecological reconstructions. Palaeo- communities are commonly reconstructed from macrofossils and generally these are restricted to the woody component of the community because herbaceous species are seldom preserved. Based on the woody component of the contemporary California and Arizona communities one would conclude that these are similar plant communi- ties. However, the detailed community character- ization demonstrated in this paper shows that these are radically different communities. When confronted with the duel fall and spring herba- ceous communities and the major contribution of C4 perennial grasses to the Arizona postfire community one must conclude that these two regions are dominated by quite different plant communities. Thus, paleofloras reconstructed from just the woody components would be potential- ly misleading in comparisons of these types of communities. This is relevant to reconstructing the past history of chaparral as it appears that it orig- inated under summer rain conditions, apparent- ly in the southwestern portion of North America (Ackerly 2009; Keeley et al. 2012). If the Arizona chaparral is a reflection of earlier chaparral stages it strongly suggests that the primary similarity is in the woody flora. The contemporary postfire herbaceous flora in Cal- ifornia chaparral appears to be a flora, with similarities to winter rain floras from Arizona chaparral, but largely missing the fall floras of Arizona chaparral. Although lacking in diversi- ty of functional types, the Mediterranean-type climate appears to have played a role in adding to the diversification of the winter rain postfire flora. Many of the genera common in the spring floras of both regions have much greater di- versity in California than in Arizona. 2012] ACKNOWLEDGMENTS We thank Anne Pfaff for GIS assistance. Any use of trade, product, or firm names in this publication is for descriptive purposes only and does not imply endorse- ment by the U.S. government. LITERATURE CITED ACKERLY, D. D. 2009. Evolution, origin and age of lineages in the Californian and Mediterranean floras. Journal of Biogeography 36:1221—1233. AXELROD, D. I. 1989. Age and origin of chaparral. Pp. 7-19 in S. Keeley (ed.), The California chaparral: paradigms reexamined. Science Series No. 34. Natural History Museum of Los Angeles County, Los Angeles, CA. BHASKAR, R., A. VALIENTE-BENUET, AND D. D. ACKERLY. 2007. Evolution of hydraulic traits in closely related species pairs from mediterranean and non mediterranean environments of North America. New Phytologist 176:718—726. BOLANDER, D. H. 1982. Chaparral in Arizona. Pp. 60— 63 in C. E. Conrad and W. C. Oechel, (eds.), Proceedings of the symposium on dynamics and management of Mediterranean-type ecosystems. General Technical Report PSW-58. USDA Forest Service, Pacific Southwest Forest and Range Experiment Station, Albany, CA. CABLE, D. R. 1975. Range management in the chaparral type and its ecological basis: the status of our knowledge. Research Paper RM-155. USDA Forest Service, Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO. CARMICHAEL, R. S., O. D. KNIPE, C. P. PASE, AND W. W. BRADY. 1978. Arizona chaparral: plant associations and ecology. Research Paper, RM- 202. USDA Forest Service, Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO. DENNISON, P. E., M. A. MORITZ, AND R. S. TAYLOR. 2008. Evaluating predictive models of critical live fuel moisture in the Santa Monica Mountains, California. International Journal of Wildland Fire 17:18-27. FOTHERINGHAM, C. J. 2009. Postfire chaparral recov- ery in the bimodal rainfall regime in Arizona. Ph. D. dissertation, University of California, Los Angeles, CA. HICKMAN, J. C. (ed.) 1993, The Jepson manual. Uni- versity of California Press, Los Angeles, CA. KEELEY, J. E. 2000. Chaparral. Pp. 203-253 in M. G. Barbour and W. E. Billings, (eds.), North Amer- ican terrestrial vegetation. Cambridge University Press, Cambridge, U.K. AND W. J. BOND. 1997. Convergent seed germination in South African fynbos and Califor- nian chaparral. Plant Ecology 133:153—167. AND F. W. DAvis. 2007. Chaparral. Pp. 339— 366 in M. G. Barbour, T. Keeler-Wolfe, and A. A. Schoenherr, (eds.), Terrestrial vegetation of Cali- fornia. University of California Press, Los Angeles, CA. AND C. J. FOTHERINGHAM. 2003. Species-area relationships in Mediterranean-climate plant com- munities. Journal of Biogeography 30:1629—1657. AND . 2005. Plot shape effects on plant species diversity measurements. Journal of Vegeta- tion Science 16:249-256. KEELEY ET AL.: CALIFORNIA VS ARIZONA CHAPARRAL [23 AND S. C. KEELEY. 1988. Chaparral. Pp. 165— 207 in M. G. Barbour and W. D. Billings, (eds.), North American terrestrial vegetation. Cambridge University Press, Cambridge, U.K. , T. BRENNAN, AND A. H. PFAFF. 2008. Fire severity and ecosystem responses following crown fires in California shrublands. Ecological Applica- tions 18:1530—1546. , C. J. FOTHERINGHAM, AND M. B. KEELEY. 2005. Determinants of postfire recovery and succession in mediterranean-climate shrublands of California. Ecological Applications 15:1515— 1534. , W. J. BOND, R. A. BRADSTOCK, J. G. PAUSAS, AND P. W. RUNDEL. 2012. Fire in mediterranean climate ecosystems: ecology, evolution and manage- ment. University of Cambridge Press, Cambridge, U.K. KNIPE, O. D., C. P. PASE, AND R. S. CARMICHAEL. 1979. Plants of the Arizona chaparral. General Technical Report RM-64. USDA Forest Service, Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO. AND P. C. MILLER. 1985. Chaparral. Pp. 213- 231 in B. F. Chabot and H. A. Mooney, (eds.), Physiological ecology of North American plant communities. Chapman and Hall, New York, NY. MUELLER-DOMBOIS, D. AND H. ELLENBERG. 1974. Aims and methods of vegetation ecology. John Wiley & Sons, New York, NY. PASE, C. P. 1965. Shrub seedling regeneration after controlled burning and herbicidal treatment of dense pringle manzanita chaparral. Research Note RM-56. USDA Forest Service, Rocky Mountain forest and Range Experiment Station, Ft. Collins, CO. AND D. E. BROwN. 1982. Interior chaparral. Desert Plants 4:95—99. AND F. W. POND. 1964. Vegetation changes following the Mingus Mountain burn, Research Paper, RM-18. USDA Forest Service, Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO. RUNDEL, P. W. 1981. Fire as an ecological factor. Pp. 501-538 in L. O. L. Lange, P. S. Nobel, C. B. Osmond, and H. Ziegler, (eds.), Physiological plant ecology. I. Springer-Verlag, New York, NY. SAGE, R. F., D. A. WEDIN, AND M. LI. 1999. The biogeography of C4 photosynthesis: patterns and controlling factors. Pp. 313-373 in R. R. Sage and R. K. Monson, (eds.), C4 plant biology. Academic Press, San Diego, CA. SCHMUTZ, E. M. AND D. W. WHITHAM. 1962. Shrub control studies in the oak-chaparral of Arizona. Journal of Range Management 15:61—67. UNITED STATES DEPARTMENT OF AGRICULTURE (USDA). 2009, The PLANTS database. National Plant Data Center, Baton Rouge, LA.Website: http://plants.usda.gov [accessed | June 2009]. VANKAT, J. 1989. Water stress in chaparral shrubs in summer rain versus summer drought climates. Whither the Mediterranean type climate paradigm. Pp. 117-124 in S. C. Keeley (ed.), The California chaparral: paradigms reexamined. Science Series No. 34. Natural History Museum of Los Angeles County, Los Angeles, CA. WHITTAKER, R. H. AND W. A. NIERING. 1964. Vegetation of the Santa Catalina Mountains, 124 Arizona. I. Ecological classification and distribu- tion of species. Journal of the Arizona Academy of Sciences 3:9—34. AND ———. 1965. Vegetation of the Santa Catalina Mountains, Arizona. II. A gradient analysis of the south slope classification and distribution of species. Ecology 46:429-452. WOLFE, J. 1964. Miocene floras from Fingerrock Wash southwestern Nevada. Geological Survey Profes- sional Paper, Paper 454-N, 591—658. U.S. Govern- ment Printing Office, Washington, D.C. APPENDIX | Plant families recorded from just Arizona or just California postfire sites, and families recorded from sites in both regions. Only in Arizona (19) Acanthaceae Aceraceae Aizoaceae Asclepiadaceae Cactaceae Comme\linaceae Fouquieriace Juglandaceae Krameriaceae Linaceae Lythraceae Molluginaceae Pedaliaceae Pinaceae Plantaginace Polygalaceae Verbenaceae Violaceae Zygophyllacea Only in California (9) Cistaceae Crassulaceae Fumariaceae Grossulariaceae Orchidaceae Orobanchaceae Rutaceae Sterculiaceae Styracaceae In both Arizona & California (44) Agavaceae Amaranthaceae Anacardiaceae Apiaceae Asteraceae Boraginaceae Brassicaceae Campanulaceae Caprifoliaceae Caryophyllaceae Chenopodiaceae Convolvulaceae Cupressaceae Cuscutaceae Cyperaceae Ericaceae Euphorbiaceae Fabaceae MADRONO Fagaceae Fumariaceae Garryaceae Gentianaceae Geraniaceae Hydrophyllaceae Lamiaceae Liliaceae Loasaceae Malvaceae Nyctaginaceae Onagraceae Papaveraceae Poaceae Polemoniaceae Polygonaceae Portulacaceae Primulaceae Pteridaceae Ranunculaceae Rhamnaceae Rosaceae Rubiaceae Scrophulariaceae Selaginellaceae Solanaceae APPENDIX 2 [Vol. 59 Plant genera recorded from just Arizona or just California postfire sites, and genera recorded from sites in both regions. Only in Arizona (148) Abutilon Acacia Acalypha Acer Adiantum Aeschynomene Agave Agropyron Alternanthera Androsace Anemone Anisacanthus Anoda Astrolepis Bahia Baileya Bidens Boerhavia Bothriochloa Bouchea Bouteloua Brickellia Bulbostylis Calliandra Carmentia Cathestecum Chaetopappa Chamaecrista Cheilanthes Chloris Commelina Condalia Corydalis 2012] Crotalaria Crusea Cuphea Cylandropuntia Cynanchum Cynodon Cyperus Dalea Dasylirion Dasyochloa Desmanthus Desmodium Dicliptera Digitaria Diodia Ditaxis Drymaria Dyschoriste Dyssodia Echinocereus Elionurus Enneapogon Ephedra Epilobium Eragrostis Eriastrum Ericameria Euphorbia Evolvulus Fallugia Ferocactus Fouquieria Funastrum Geraea Glandularia Gomphrena Guilleminea Gymnosperma Hackelochloa Hedeoma Heliomeris Heliotropium Heteropogon Heterosperma Houstonia Hymenopappus Aymenothrix Hymenoxys Ipomoea Ipomopsis Isocoma Juniperus Krameria Laennecia Lappula Lasianthaea Leptochloa Lycurus Machaeranthe Macroptilium Macrosiphon Mammillaria Marina Melampodium Melinis Microsteris Mimosa Mitracarpus KEELEY ET AL.: CALIFORNIA VS ARIZONA CHAPARRAL Mollugo Monolepis Myosurus Nolina Notholaena Oreochrysum Panicum Pectis Pennellia Phaseolus Physalis Pinus Piptochaetium Platyopuntia Portulaca Proboscidea Prosopis Pseudognaphia Psilactis Psoralidium Sanvitalia Schistophrag Schoenocrambe Sclerocactus Scleropogon Senna Setaria Sida Sorghum Spermolepis Sphaeralcea Sporobolus Stevia Swertia Symphyotrichum Tagetes Tephrosia Tidestromia Trachypogon Trachypogon Trianthema Triticum Urochloa Verbesina Viguiera Zephyranthes Zornia Only in California (78) Achnantherum Acourtia Adenostoma Anagallis Antirrhinum Apiastrum Brachypodium Calyptridium Calystegia Camissonia Caulanthus Centaurea Centaurium Chaenactis Chamaebatia Chlorogalum Chorizanthe Clarkia Claytonia Cneoridium 126 MADRONO [Vol. 59 Cordylanthus Argemone Crassula Aristida Cupressus Artemisia Delphinium Asclepias Dendromecon Astragalus Dicentra Atriplex Dichondra Avena Dodecatheon Baccharis Emmenanthe Bowlesia Eriodictyon Brassica Eriophyllum Bromus Eschscholzia Calandrinia Eucrypta Calochortus Filago Capsella Fremontodendron Carex Gastridium Castilleja Hazardia Ceanothus Helianthemum Cerastium Hemizonia Cercocarpus Heteromeles Chamaesyce Hieracium Chenopodium Hirschfeldia Cirsium Horkelia Claytonia Hypochoeris Conium Keckiella Conyza Lepechinia Cryptantha Leymus Cuscuta Lonicera Datura Malacothamnus Daucus Malosma Descurainia Marah Dichelostemma Melica Draba Monardella Elymus Muilla Encelia Nassella Erigeron Navarretia Eriogonum Nemocladus Erodium Orobanche Galium Osmadenia Garrya Papaver Gilia Pedicularis Gutierrezia Phacelia Helianthus Pickeringia Heterotheca Piperia Hordeum Plagiobothryus Tris Polycarpon Juglans Pterostegia Lactuca Ribes Lathyrus Scrophularia Layia Scutellaria Lepidium Styrax Linanthus Tauchsia Lineria Thalictrum Linum Toxicodendron Lomatium Trifolium Lotus Uropappus Lupinus Xylococcus Malacothrix Zigadenus Marrubium In both Arizona and California (109) Mentzelia A goseris Mimulus Agrostis Mirabilis Allium Monardella Allophyllum Muhlenbergia Amaranthus Oenothera Ambrosia Pectocarya Amsinckia Pellaea Arabis Penstemon Arctostaphylos Phacelia 2012] Plagiobothrys Plantago Poa Polygala Polygonum Porophyllum Quercus Rafinesquia Rhamnus Rhus Salsola Salvia Sambucus Schismus Selaginella Senecio KEELEY ET AL.: CALIFORNIA VS ARIZONA CHAPARRAL Silene Sisymbrium Solanum Solidago Sonchus Stellaria Stephanomeria Streptanthus Stylocline Thysanocarpus Trichostema Triodanis Vicia Vulpia Yabea Yucca [27 MADRONO, Vol. 59, No. 3, pp. 128-142, 2012 MORPHOLOGICAL AND ISOENZYME VARIATION IN RHODODENDRON OCCIDENTALE (WESTERN AZALEA) (SECTION PENTANTHERA; ERICACEAE) G. F. HRUSA UC Davis Center for Plant Diversity, Department of Plant Sciences, Mail Stop 7, One Shields Avenue, Davis, CA 95616 California Dept. of Food and Agriculture Plant Pest Diagnostics Center, 3294 Meadowview Rd., Sacramento, CA 95832-1448 fhrusa@cdfa.ca.gov ABSTRACT Morphological and isoenzyme variation among populations of western azalea, Rhododendron occidentale (Torr. & A. Gray) A. Gray, were examined. Three regional parapatric groups were revealed: 1) the northern California outer North Coast Ranges; 2) the northern California and southern Oregon Klamath Ranges; and 3) the central California Sierra Nevada and southern California Peninsular Ranges. A highly variable but generally intermediate fourth group is restricted to ultrabasic substrates (serpentine) in the middle and inner North Coast Ranges of California. It is comprised of populations with recombined morphologies and alleles that were otherwise restricted to one or more of the three groups above. A revised intraspecific treatment is proposed, with the three regional groups above recognized as varieties. These are: R. occidentale (Torr. & A. Gray) A. Gray var. occidentale (outer North Coast Ranges), R. 0. var. paludosum Jeps. (Klamath Ranges), and Rhododendron. occidentale var. californicum (Torr. & A. Gray) Hrusa comb. et stat. nov. (Sierra Nevada and Peninsular Ranges). Lectotypifications of Azalea occidentalis Torr. & A. Gray isotypes at PH and GH, Azalea californica Torr. & A. Gray in Durand, and Rhododendron sonomense Greene (NDG) are also provided. Key Words: Azalea, isozyme, lectotype, morphometric, population, Rhododendron occidentale. In addition to its large and fragrant flowers Rhododendron occidentale (Torr. & A. Gray) A. Gray has long been recognized for its diversity of growth habits, floral pigmentations, and vegeta- tive forms (Kellogg 1855; Wilson and Rehder 1921; Jepson 1939, Munz and Keck 1959; Moss- man and Smith 1969; Kron 1993). The only more or less invariant characteristic is the presence of a yellow nectar guide on the upper corolla limb. It has been suggested that this variation indicates the existence of multiple species (Mossman and Smith 1969; Mossman 1977). Other observers have interpreted these polymorphisms as ran- domly distributed genotypes or as environmen- tally induced (Breakey 1960). The most recent systematic treatment of section Pentanthera (Kron 1993) did not examine the presence or patterns of intraspecific variation in R. occiden- tale, and the same can be said of the discussion in Wilson and Rehder (1921). Indeed, there has been no detailed accounting of its regional diversity. In horticultural circles its visible morphological variation has made R. occidentale a favorite of azalea breeders. Middle nineteenth- to early twentieth-century selections were prominent in these breeding programs (Mossman 1974). The results can be found for sale as the Knapp Hill and Exbury hybrid series. More recently, enthu- siasts have been seeking out novel wild forms. Especially popular are those with unusual corolla shapes and color variations (Mossman _ 1974, 1977), which are sold with names such as ‘Humboldt Picotee’, “Tatum’s Pink’, or ‘Double Dig Twelve’ (Jones et al. 2007). Accompanying the reports of these variants’ discoveries were poorly documented assertions regarding poly- ploidy; wild distributions; regional variation patterns; and edaphic, moisture, and temperature tolerances in wild populations (Breakey 1960; Mossman and Smith 1968; Mossman 1972, 1974, 1977). Ultimately, experimental evidence for any of these interpretations is lacking. Geographic Distribution Rhododendron occidentale is distributed within the Coast Ranges of California, the Klamath Ranges of northern California and southern Oregon, the Sierra Nevada, and the Peninsular Ranges of southern California. Its distribution is largely coincident to the California Floristic Province (CFP) (Raven & Axelrod 1978) with its northernmost populations only about 75 miles north of the CFP along the Umpqua River in Douglas Co., Oregon. It is found at elevations extending from sea level to near 2800 m. | Throughout its range it is restricted to sites of | permanent moisture, although these may be | subsurface sources. It is frequently found growing on ultrabasic soils, predominantly serpentine, particularly in the Klamath ranges and inner North Coast Ranges. Except for a single location at the northern tip of the Gabilan Range in 2012] Monterey Co., R. occidentale is absent from the California South Coast Ranges and Transverse Ranges. It is presumed that the warming and drying of the post-Pleistocene eliminated it from these relatively dry mountains. Claims that R. occidentale has native populations in the Puget Sound area, the Olympic Peninsula, and on Mount Rainier in Washington (Mossman 1974) require confirmation. This paper will present quantitative and qualitative analyses that describe the patterns of morphological and isoenzyme variation among wild populations of R. occidentale. A taxonomic treatment that accounts for the species’ natural variation patterns will also be proposed. MATERIALS AND METHODS Sampling Populations representing the species’ full geo- graphic range were analyzed. Thirty-six popula- tions were used for the morphological analysis, and 37 populations were used for the isoenzyme analysis (Fig. 1 and Appendix 1). All but three population samples contained a minimum of 25 individuals. The three smaller samples came from critically situated populations that did not contain that many individuals. Rhododendron occidentale is not rhizomatous, nor does it sprout from the roots. It may, however, form clones when fallen trees or branches press plants to the ground and adventitious roots develop. There was adequate visible variation among indi- viduals in both growing and dormant structures so that duplicate clone collections were readily avoid- ed. Sampling was random throughout except in the case of the smaller populations where every individual was examined. The morphological samples included both spring-collected flowering material and winter- collected dormant inflorescence buds. Thus the dormant and flowering collections used for mor- phological examination did not necessarily repre- sent the same individuals. This did not affect the analyses because the characters were defined as population summaries. Specimens for isoenzyme analysis were either fresh, spring-collected corolla tissue or mature, dormant vegetative branch tips. The sampled populations are mapped in Figure 1. Locality descriptions and sample sizes are listed in Appendix 1, and representative vouchers from those populations have been deposited at CDA and DAV. All statistical analyses were performed in JMP 5.1 (SAS Institute). Morphometric Data Traditional morphological classifications of Rhododendron have emphasized variation in floral pigmentations and vegetative trichome HRUSA: VARIATION IN RHODODENDRON OCCIDENTALE 129 types and their positions, in addition to quanti- tative physical data (Sleumer 1949; Kron 1993). The first two proved particularly useful in this study. Quantitative data were taken from specimens collected, pressed, and dried specifically for these analyses. Because of the variable bilateral form of Rhododendron occidentale corollas, special han- dling and pressing of individual flowers after their removal from the inflorescence was necessary for consistent measurement of corolla tube and limb relationships. For the quantitative characters, val- ues were defined as the mean of four measure- ment repetitions per structure per individual. As much as possible each measured structure was standardized by plant position and developmen- tal stage. Phenotypic plasticity within leaf size, shape, venation patterns, and trichome length were demonstrated in a preliminary unpublished study (Hrusa 1991). Seed and capsule features including seed wing shape and capsule sizes and shapes were highly variable within individuals and were not used. The raw measurements were taken at various scales. These were ranged to between 0 and 1 using Gower’s Transformation (Sneath and Sokal 1973). The population-based quantitative characters were defined as the proportion of the range for a given population that fell in two of three equal classes that represented the upper and lower one-third of the structure’s among-popu- lation range. Statistically, the central of the three quantitative classes for each character is always correlated to one of the outside thirds, and was excluded. The qualitative character states were defined as their frequency within each population sample. Corolla coloration data were acquired from fresh material. Trichome position and density data were taken from dried specimens. The quantita- tive and qualitative characters are listed in Table 1. Isoenzyme Data Soluble enzymes were extracted from fresh plant samples and electrophoresed on horizontal starch gels composed of 12% hydrolyzed potato starch and 3% sucrose. Gel and electrode buffers were composed of 0.0009 M L-histidine-0.0003 M citric acid and 0.065 M L-histidine-0.019 M citric acid, both at pH 5.7. Electrophoresis proceeded for 13 hours at 3.5 watts. Allelic variants were classified and identified by their relative migration distance against a stan- dard allele at each locus. This allele was one present in every population and was usually the most common. Homologies among the variant electromorphs were determined by comparing them on common gels. 130 MADRONO Fic. 1. RESULTS Morphological Variation Thirty-eight outwardly visible features (21 quantitative and 17 qualitative) were described and measured or assessed (Table 1). The among- population comparisons were performed through principal components analysis of the among- population correlation matrices. The first three principal components describ- ed 84.9% of the total variation, with four clus- ters of populations evident. These are plotted in Figure 2 and mapped in Figure 3. The Sierra Nevada and Peninsular Range populations (SNPR) grouped together. The populations of the Klamath Ranges (KR) and outer North Coast Ranges (ONCR) ordinated apart from [Vol. 59 Populations of Rhododendron occidentale used in the morphological and isoenzyme analyses. See Appendix | for geographic location and sampling details. both each other and from the SNPR populations. With two exceptions, the inner North Coast | Range Serpentine (INCRS) populations plotted | between the ONCR, KR, and SNPR popula- tions, much as they are situated geographically | (Fig. 3). The exceptions, (P-64, Kilpepper Creek, Lake Co.; and P-34, Stony Creek, Colusa Co.) | grouped among the SNPR populations. These two populations shared with those of the SNPR | the characteristics of relatively long, generally | white-colored corolla tubes, and a pubescent leaf | abaxium. | Stepwise discriminant analysis was used to | evaluate which characters were useful in distin- guishing the four groups in Figure 2. The results | indicated that individuals from the KR and ONCR populations shared combined glandular and eglandular multicellular trichomes on the leaf | 2012] TABLE l. HRUSA: VARIATION IN RHODODENDRON OCCIDENTALE 13] DEFINITIONS OF MORPHOLOGICAL FEATURES USED IN THE MULTIVARIATE ANALYSES. Asterisks (*) denote subset of nine systematically useful characters. Quantitative characters were defined as described in the text. Qualitative characters were defined as their frequency within the population sample. Quantitative 1. Ovary length: classes 1, 3 2. Anther length: classes 1, 3 summit 3. Corolla length: classes 1, 3 for vein tip 4. Corolla tube length: classes 1, 3 5. Corolla length/tube length ratio: classes 1, 3 6. Inflorescence bud length: classes 1, 3 and strigose 7. Calyx lobe length, longest: class | trichomes: glandular 8. Calyx lobe length, shortest: 18. Leaf abaxium, secondary class | vein trichomes: mixed glandular and strigose 9. Calyx lobes ratio, length of longest/shortest: class | *10. Leaf abaxium, multicellular trichomes per 25 mm‘: class 1 (<7) and 3 (>26) abaxial midvein and a lamina abaxium without unicellular hairs. There may occasionally be unicellular trichomes on and adjacent the mid- vein, but not on the abaxial surface. The ONCR populations alone were distinguished by: 1) a dense pubescence on the young twigs and dormant bud bracts; 2) winter bud bract-margin trichomes always of a glomerate-glandular type; 3) relatively large dormant flower buds; and 4) more than 25 multicellular trichomes per/25mm/? of leaf abaxial surface. The distinguishing fea- tures are marked with an asterisk in Table | and with additional morphological features summa- rized among the groups in Table 3. Klamath Ranges populations alone were dis- tinguished by: 1) mostly glabrous young twigs and either glabrous or thinly ciliate bud bracts; 2) usually ciliate, occasionally glomerate winter bud bract-margins; 3) generally a pink- to red- pigmented corolla tube; 4) a short corolla tube length in relation to the throat plus limb; and 5) the lamina abaxial surface, excluding the mid- vein, was devoid or nearly so of trichomes. Populations from the immediate coast generally were transitional to the ONCR azaleas (glome- rate bud bract-margin trichomes and less densely pubescent bud bracts and twigs). Overall the KR azaleas had the smallest corollas with the shortest floral tubes and the most pink to red pigmenta- tions in the limb and tube. Among the KR populations, only the single northernmost and most interior was morphologically unusual (P-25, Cow Creek, Douglas Co. Oregon, Fig. 2). This population had: 1) infrequently, a unicellular 11. Perianth tube: dark pink 12. Perianth veins: pink base to 13. Perianth veins: white except 14. Perianth limb: pink 15. Perianth limb: white *16. Leaf abaxium, midvein trichomes: mixed glandular 17. Leaf abaxium, midvein *19. Leaf abaxium, secondary vein trichomes: strigose Qualitative *20. Leaf abaxium, secondary vein trichomes: absent 21. Leaf margin trichomes: mixed glandular and strigose *22. Floral bud-scales, margin trichomes: glomerate *23. Floral bud-scales, margin trichomes: ciliate *24. Young twigs, vestiture: densely pubescent *25. Young twigs, vestiture: glabrous *26. Leaf abaxium, surface: single- celled pubescence 27. Floral bud bract, surface vestiture: pubescent 28. Floral bud bract, surface vestiture: glabrous pubescence on the leaf abaxium (KR azaleas are generally glabrous abaxially) and 2) midvein trichomes of a single type only (these are usually mixed strigose and glandular in the KR). The presence of these two morphological states suggests plants of the SNPR populations. How- ever, isoenzyme alleles characteristic of that group were absent. The SNPR azaleas had: 1) one type of multicellular leaf midvein trichome, either glan- dular or eglandular, not those combined as in the KR and ONCR; 2) a unicellular pubescence throughout the lamina abaxial surface; 3) second- ary vein multicellular trichomes of the same type as the midvein (glandular or eglandular); 4) absent or widely scattered multicellular trichomes on the tertiary veins (when present these were « 25/25 mm°); 5) a non-pigmented to rarely, slightly pigmented corolla tube; 6) corolla veins rarely with more than a hint of pigmentation at the distal tip. In general, SNPR corollas were either pure white or (infrequently) had a hint of pink in the tube or vein summit; and 7) the corolla tube was comparatively long in relation to the com- bined limb and throat. The SNPR plants were thus particularly distinctive morphologically. Only P-50 (Butterfly Valley, Plumas Co.) of the northern Sierra Nevada had glabrous leaf abaxia and often glabrous young twigs, both features unusual among the Sierran azaleas. Population 50 plotted between the SNPR and KR populations, the same as it is situated geographically (Fig. 3). In general, the southern California Peninsular Range azaleas had the largest and least pigmented 132 MADRONO [Vol. 59 *& INCRS A SNPR @ ONCR O KR FIG. 2. Scatter plot of the first three principal component scores for the morphological dataset. Populations 64 and 34 occur in the inner North Coast Ranges on serpentine, plotted here with Sierra Nevada populations. Labeled populations are discussed in the text. flowers with the longest floral tubes. Overall, excluding the INCRS azaleas, a cline exists from north to south in which corolla coloration diminishes and the corollas become larger overall and have relatively longer tubes. Even excluding P-34 (Stony Creek, Colusa Co.) and P-64 (Kilpepper Creek, Lake Co.) that ordinated among the SNPR, the INCRS popu- lations were weakly differentiated. They were distinguishable morphologically only by a short- est calyx lobe that was generally less than 1 mm in length, and sometimes nearly obsolete. Their multivariate scattering across the space between the KR, ONCR, and SNPR groups (Fig. 2) was the result of both quantitative intermediacy and mosaic patterns among the qualitative features. This variation was spread among the first three principal components and is suggestive of segre- gation and recombination among the analyzed features. Definitions of three readily visible and at least partially regionally distinctive features proved problematic and were not included as population- level characters in the multivariate analyses. Darkly anthocyanous new growth was a distinc- tive feature of both sun- and shade-grown INCRS azaleas. Elsewhere, azaleas growing on ultrabasic- derived soils often had at least some anthocyanin pigmentation in the young twigs, particularly in plants growing in full sun. This pigmentation (and lack of same from non-pigmented popula- tions) was maintained in greenhouse-grown seedlings, indicating it is under genetic control, at least in the INCRS (Hrusa 1991). Although the non-INCRS anthocyanin pigments were not as dark as that among individuals within the INCRS populations, this distinction could not be consistently delimited into a set of qualitative categories. Such coloration is a feature of many ultrabasic-adapted taxa (Kruckeberg 1984). Strongly bifacial leaves that were lighter on the abaxial surface characterized many popu- lations, mostly those of the SNPR. Several INCRS populations were similarly bifacial. Some of the latter were dimorphic for that char- acteristic, and in those populations the intensi- ty of the bifacial condition was particularly variable, often varying in intensity even within individuals. The timing of flower bud opening relative to leaf break is generally species specific within Rhododendron section Pentanthera (Kron 1993). Inflorescences may open concurrently to or after foliage maturation. In R. occidentale both condi- tions occur. Plants of the KR and most of the INCRS open flowers and leaf buds concurrently. Exceptions may occur in densely shaded plants where bud (and often leaf) break may be delayed. However, most followed the pattern of inflores- cence break in concert with foliar emergence. One population from the INCRS (P-64, Kilpepper Creek, Lake Co.) broke flower buds only after new growth had fully matured; yet plants in the closely related and nearby P-34 (Little Stony 2012] Wi KR => @ ONCR A SNPR %* INCRS i ' iG 3. populations are discussed in the text. Creek, Colusa Co.) opened concurrent to bud break. With one exception each, plants of the ONCR and SNPR opened their inflorescences after the foliage had fully expanded and matured. Only the ONCR P-45 (Hogback Mountain, Sonoma Co.), growing on a highly insolated and relatively dry hillside, pushed its flower buds late-concurrent to leaf break. Again with one exception, Sierra Nevada populations flowered after leaf maturity. Population 49 (Pulga Rd., Butte Co.), from an open, south-facing serpentine slope, opened its flowers with new growth expansion. Thus, while bud and leaf break patterns are mostly consistent regionally, the exceptions suggest that local environmental conditions or selection in open habitats influence timing of flower bud break. Map of the population clusters based on morphological variation as identified in Fig. 2. HRUSA: VARIATION IN RHODODENDRON OCCIDENTALE 133 Labeled Isoenzyme Variation Forty-two alleles were resolved at seven loci from four enzyme systems. The resolved variants were in malate dehydrogenase (MDH-1, MDH-2, MDH-3, 11 total alleles), phosphoglucomutase (PGM-1, PGM-2, 6 total alleles), 6-phosphoglu- conate dehydrogenase (6-PDH, | locus, 5 total alleles), and glucose-6-phosphate-isomerase (GPI-2, 20 total alleles). At each locus there was a most-frequent (primary) allele with the remaining allele(s) at lower frequencies. The presence or absence and frequency of the alternate alleles varied widely among populations. Only in P-64 (Kilpepper Creek, Lake Co.) and P-61 (Cuyamaca Peak, San Diego Co.) did the alternate alleles at PGM 134 MADRONO [Vol. 59 *INCRS 6 64 ASNPR * * re @ONCR * OKR -12 10 -8 -6 -4 -2 | 0 2 4 6 Fic. 4. Scatter plot of the first two principal component scores for the isoenzyme dataset. Labeled populations are discussed in the text. exceed slightly the frequency of the primary allele. In no case did an alternate allele replace the primary one. The first three principal components recovered 95.0% of the variation. Three population groups were evident, as opposed to four for the morphological data. The groups are plotted in Figure 4 and mapped in Figure 5. The multivariate patterns were similar to those using morphology, except for the superposition of the KR and ONCR populations. With five endemic alleles the KR group might have been expected to occupy its own isoenzyme multivar- late space as it did its own morphological space. That it does not is due to the presence of the non- private endemic alleles in only a few populations, and at relatively low frequencies. The SNPR and INCRS populations plotted over a broader multivariate space than did the adjacent KR + ONCR cluster, due to their greater allelic diversity and higher alternate allele frequencies. Moreover, those frequencies varied considerably among the populations. In terms of allele presence or absence the SNPR and INCRS were similar. However, the shared alleles were not at similar frequencies, and this is responsible for their plotting in a slightly different, but adjacent, multivariate space. A characteristic INCRS allele was MDH-3B; although this allele was also found widely scattered at generally low frequencies among the KR, SNPR, and ONCR populations, it was in every INCRS population and at higher frequencies than all but one population outside that group. Two populations of the North Coast Ranges also classified differently between the morpho- logical and isoenzyme datasets. Population 43 (Wildwood, Trinity Co.), the northernmost mor- phologically like the INCRS group (Figs. 2, 3), contained allelic complements characteristic of the KR, not those of the INCRS. Population 52 (Red Mountain, southern Hum- boldt Co.) grouped morphologically among the KR populations. However, its isoenzyme com- plement placed it between the SNPR and KR + ONCR populations, but on the opposite side of the multivariate space occupied by the likewise intermediate INCRS group (Fig. 4). Although P- 52 contains the same SNPR and KR alleles that characterized most of the populations included here in the INCRS, it was distinguished from them by a high frequency of allele MDH-2E and a low frequency of MDH-3B, the reverse of the frequencies found in the INCRS populations. Overall P-52 has mostly KR alleles. Its SNPR connection is via relatively high frequencies of alleles PGM-2B and PGM-2A as found through- out the SNPR and INCRS populations. In relation to the four geographic regions evident in the morphological analyses, four allele distribution patterns were discernible: 1) wide- spread, not ubiquitous, but occurring across geographic regions; 2) alleles common in one region, but uncommon and localized in another, HRUSA: VARIATION IN RHODODENDRON OCCIDENTALE 135 2012) ee Pree et od Hee ie ie el “SRA 10% + (%) Shared, both >10% (%) (Set um +t KD > but no endemic alleles occurred among _ the INCRS populations. Likewise, neither private nor endemic alleles occurred within the ONCR. DISCUSSION From the data, four main points may be made. First, it is clear that the azaleas of the Peninsular Ranges are allied to those of the Sierra Nevada. This is not surprising given the floristic, geologic, and climatic similarities of the forested parts of the two regions (Munz 1974; Axelrod 1976; Raven and Axelrod 1978). Azalea populations of the Sierra Nevada and Peninsular Ranges shared a distinctive allele frequency pattern at PGM where the two region’s populations all had relatively equal frequencies of three different alleles. In contrast only one of the two alternates was even a rare occurrence in either the KR or ONCR populations. Despite the presence of endemic GPI alleles in some Peninsular Range azaleas that suggests a long post-disjunction history, their morphological similarity to the azaleas of the Sierra Nevada, particularly the high southern Sierra Nevada, was marked. It would appear that these two regions were formerly part of a continuous interbreeding azalea population. Fossil azaleas of Pliocene age (pre-Sierra Nevada time) were found in the Chalk Hills flora of western Nevada (Axelrod 1962). These azaleas were associated with Sequoiaden- dron and other montane taxa that the western azalea is associated with in the Sierra today. However, Coast Range elements, including the Monterey Co. endemic Abies bracteata were also present. The leaf and capsule impressions were examined by this author, but are not assignable below section Pentanthera. SNPR KR ONCR + x x Xx x Xx xX + X xX + + x + + + + xX x xX + 11 (78.6) 8 O71) 4 (28.6) 7 (63.6) 2 (25.0) 4 (100) 6 (50) 1 (8.4) 4 (33.4) Second, while the KR and ONCR azaleas had similarly low frequencies of alternate alleles, the allele complements were different. Morphologi- cally, the azaleas of these two regions were quite distinct. It may be instructive here that Kron et. al. (1993) found no allelic differences (or variants) at enzyme systems GPI and PGM_ between Rhododendron canescens (Michx.) Sweet and R. flammeum (Michx.) Sarg., two morphologically quite distinct species. In R. occidentale, some morphological intermediacy occurred along the immediate coastline north of Fort Bragg, Men- docino Co. There the pubescent new growth and outer dormant flower bud bracts of the adjacent, more southern ONCR azalea appeared as a thin pubescence in populations that otherwise were like the more frequently glabrous KR azalea. These northern coastal azaleas did not have the multicellular leaf tertiary vein trichomes that were generally characteristic of the ONCR, and their isoenzyme complements were those of the KR. As is true for the Sierra Nevada azaleas, the ONCR and KR populations are each associated with certain distinctive forest taxa such as Sequoia sempervirens (D. Don) Endl.and Cupres- sus (Chamaecyparis) lawsoniana A. Murray bis, respectively. Based on the fossil record (Axelrod 1962, 1976; Raven and Axelrod 1978), these taxa and their associates occupied widespread geo- graphic regions long before being restricted to their current distribution in west coast cismon- tane habitats. Third, shared endemic alleles, similar within- population allele frequencies (as at PGM-2), and certain distinctive morphological features such as the superimposition of glandular and non-glan- | dular trichomes in KR + ONCR in contrast to restriction of those trichome types to different i37 HRUSA: VARIATION IN RHODODENDRON OCCIDENTALE 2012] TT ‘JUSTINIUOS-TWaS ATOIBI ‘AWLINJVUL SATRIIBIA 1SOd jusosoqnd 0} snoiqr{s LIOULO[S 10 d}eITIO ayqeliea ‘yusosoqnd Ayury} (9¢°T) UW ('E] yuaosaqnd -UIU $Z/G > 10 ‘O1azZ Ie[npur[s 10 Juasqe Iv[npur[so 10 1e[npurys (S0'O) rel YM YSTUdoIS/OVIYM AVLINJeU “BOA ysod jou ‘yeaig Jeo] ysod Ajores ‘jUaLINOUOD yusdsoqgnd 0} sno1qr{3s d1VIOWIO[S 10 9}eI]IP snoiges 0} JUSDsaqnd Ajasuap ‘a[qeiIeA (6¢°[) wu 77] snoIqes 10 jUdDsaqnd -UIU C7/$ > 10 *O1I9SZ Ie[npurys 10 Juasqe od} 9U0 Jo 10 ‘Ie[NpuULls ‘repnpuryso pure Jejnpur[s pox (900) LEI yuid owos yim Aporesr ‘ayy M yuid ‘990 ‘ayTYyM ‘Uas A}LINJEUL 9AT}e}9B9A 1SOd yuaosaqnd 9] BIDUIO[S yusossoqnd Ajasuap (p8°[) WU 6'P] snoiqey3s SW ¢Z/ST< “bary 0} Maj Je[npueyso Ajoiel ‘Ie[Npurys 10 JUasqe od} ou0 JO Ajarel ‘repnpuryso pue iejnpurys poxru (90°0) Ir] yuid *990 ‘aqyryM AT]sOUl yuid 0} a14yM JUSIINIUOS yusosaqnd ATUIY) 0} sno1qRy{s a Le snoiqei[s Ajsou (18°) WU 67] sno1qeys -WIU $Z/G >> IO O1az yuasqe odA} DUO Jo IO ‘Ie[Npurso pue re[npuel[s poxtu (O10) Srl yuid yep 0} a31yM yuid yep 0} 3314 M AJLIn}eU 10 YeaIq pnq “3d ‘SA YROIG png IdUddSaIO]JU] sousosaqnd Ie[nyjoorun Sim} Suno X SOWUOYSL)] UISIPU jORIQ “png sdUddDsaIOTU]T dININSSA “png sdUd9sa10]JU] (p's) YSus] png sduddsa10]JUT SOULOYOI] IepNy{foorun ‘QORJINS [RIxeQqe Jer] Ayisuap [erxeqge ‘SOULOYSII] IeTNT[ION [NPY SUIDA AIBI}19} ‘SOULOYSIN IelNT[ION [NPY UIDA PIU ‘SOUOYSLI] JeNTION NPY 9qn} PT[OI0D/Y3UN] P][OIOS [[BIDAO ‘ONY, QUIT] “IO[OS RTTOIOD 9qN} ‘IOTOS RI[OIOD Se a ae a ee uddNs SUONI YONO da Iajoeieyg ———e a "YISUI] B[[OIOD [210] 9Y} O} SATLJAI 9qny Jd.104S = JoquINU JOYySIH{, “UONeIADp prepurys = [eoneyyuoled “SNOIDAY IJIHdVYOOAH ONOWY FTF LINACIOIO NOUUNAGTOdOHY NIHLIAA NOILVIYVA TVOIDOTOHdYO/I ‘€ ATAVL 138 individuals in the SNPR, distinguished the azaleas of these two regions. This morphological and genetic break coincides with the floristic and paleobotanical line dividing the Klamath Ranges and the Sierra Nevada/Cascade axis (Axelrod 1962, 1976; Raven and Axelrod 1978). Again, this is evidence supporting the hypothesis that there has been a long genetic separation between the SNPR azaleas and those of the northern Califor- nia coast. Thus, although the details may be unknown, it is clear that the western azalea at present is a relict species that had a formerly more wide- spread and continuous distribution. Further, the possibility cannot be discounted that the mor- phological and isoenzyme allele groups recogniz- able in this study were already distinct at the time their primary associates occupied a much wider region of western North America than they do today (Axelrod 1962, 1976). Fourth, the isoenzyme variants present among the KR and ONCR populations combine with the alleles of the SNPR azalea to form the diverse allele complements among the INCRS popula- tions (Table 2). Moreover, a complex morpho- logical variation more or less paralleled the isoenzyme variation. These populations are discussed in more detail below. The “INCRS” Azalea The INCRS azalea populations are those growing on serpentine substrata in the hot, dry, interior parts of the North Coast Ranges of Lake, Napa, and Colusa Counties. They were repre- sented in this study by populations 2, 4, 4A, 6, 9, 34, 36, and 64 (Figs. 1, 3, 5). Population 43 shares partial morphologies with these populations but not isoenzyme alleles and is excluded from this discussion. These azalea populations occur in permanently wet habitats on high pH ultrabasic soils within open, sunny, foothill pine-leather oak—live-oak chaparral and woodland. Such an unusual habitat for azaleas has caused them to receive some research attention (Leiser 1957; Drake 1987), and the unusual ecological situation was also a factor motivating the initiation of this study. These azalea populations on the serpentine outcrops in Napa Co. east of Mount St. Helena were misinterpreted by Jepson (1925, 1939) to be Rhododendron occidentale var. sonomense Re- hder. Indeed, after his proposal of R. sonomense, Greene himself used that name for his own azalea collections from the same area. Although of a similar dwarfed size, he apparently did not recognize the distinctive morphological differenc- es between the serpentine plants east of the mountain and the more coastal ““Petaluma”’ spe- cimen on which he had based his description. As determined in this study, the azalea populations MADRONO [Vol. 59 east of Mount St. Helena contained mosaic mixtures of isoenzyme alleles otherwise found in disparate areas (Table 2). The mosaic isoenzyme allele pattern extended to the morphological variation with the complement of features in some populations like those in nearby popula- tions; in others, the complement of features resembled the plants of distant regions. More- over, there were distinctive mosaics of morpho- logical characteristics and alleles within as well as among the INCRS populations. For example, by morphology, INCRS P-34 (Stony Creek, Colusa Co.) and P-64 (Kilpepper Creek, Lake Co.) plotted among the SNPR populations (Figs. 2, 3), however, by isoenzymes they plotted among the other INCRS populations (Fig. 4). This intermediacy and among-population var- iability would best be explained via interbreeding among the KR, ONCR, and SNPR genotypes followed by recombination within and among the INCRS populations. The result is also a higher average number of alleles per locus: 2.11 for the INCRS populations; with 2.03 (SNPR), 1.97 (KR), and 1.48 (ONCR) for the other geographic groups. The INCRS populations intermediate geographic position parallels this genetic mixing, and lends support to the interpretation that these are the populational remnants of an ancient ecotone. The survival of azaleas in this region 1s apparently due to the fractured serpentine sub- stratum that resulted in the presence of permanent springs and streams in an otherwise xeric region. The history of climatic and floristic change in this region combined with the juxtaposition of sharply delimited habitats has made it a favored area for the study of plant adaptation and evolution (Major 1967; Stebbins and Hrusa 1995). The most distinctive population of the INCRS was P-36 (Gilliam Creek, Sonoma Co.). Mor- phologically this population had strongly bifacial leaves, darkly anthocyanous new growth, and corollas frequently with some pink pigmentation in the tube and limb, this latter an infrequent condition among the INCRS populations. Its isoenzyme complements included single private : alleles at both PGM-2 and GPI-2, the almost equal distribution of PGM-2 alternate alleles seen in both the INCRS and SNPR populations, and | GPI-2 alternate alleles characteristic of both the | northern SNPR populations and of the ONCR. This latter can be explained by its proximity to ONCR populations along the coast. However, its | private alleles were the only ones in the INCRS | group. The private GPI-2 allele was the highest frequency GPI variant in the population, and | among all the populations the PGM-2 private | allele was only the fourth allele seen at that locus. Overall, this population has the aspect of a | coastal form of the generally interior INCRS © azalea. The distinctive morphology and isoen- | zyme variation of P-36 suggests some uniqueness | 2012] for the azaleas on the endemic-rich serpentine habitats of the ““The Cedars” in the East Austin Creek region. Azalea genetics and morphologies characteris- tic of the Sierra Nevada in the inner North Coast Ranges are not anomalous if one accepts that glycolytic isoenzyme variation parallels or con- tributes to physiological adaptation (Gillespie 1991). Except for the azalea populations in or near The Cedars of western Sonoma Co. discussed above, most of the INCRS azaleas are in the rain shadow of several ridges and peaks including Mount Atlas, the Palisades, Goat Mountain, and Snow Mountain. These highlands block coastal air and moisture giving their shadows a hotter and drier summer and colder winter climate than that of the outer North Coast Ranges only a few miles westward. Such isoen- zyme and morphological correlation to distinc- tive habitats implies that the three geographic regions—KR, ONCR, and SNPR—support dif- ferently adapted genotypes. The discontinuous and mosaic patterns among INCRS alleles and morphologies are likely the end result of local population fragmentations, contractions, and re- expansions during the post-Pleistocene. Anacker et al. (2010) analyzed phylogenetic signals among plant taxa that occur in either serpentine or non- serpentine habitats. Their conclusion was that most serpentine-restricted taxa are younger than non-serpentine taxa in the same genus. The data presented here support that interpretation. Thus, the evidence suggests a relatively recent origin for the INCRS azaleas. However, an accounting for the presence of distant Sierran alleles and two private alleles in the far western isolated serpentine Cedars region suggests that those azaleas may have a unique history within R. occidentale. The described local recombinant patterns are in contrast to the variation within the more widespread ONCR, KR, and SNPR forms. These three groups are coherent in their regions, share morphologies and some alleles within (but less so among them) and would appear to have occupied the same ecologically distinct, if not allopatric regions for a long time. TAXONOMIC TREATMENT The three regional azalea groups determined in this study (KR, ONCR, and SNPR) were morphologically and genetically distinctive, yet vary toward each other where approaching contact, as along the northern California coast and in the northern Sierra Nevada. A treatment at varietal rank seems most appropriate as it recognizes both their distinctiveness and close relationship. As to the INCRS azaleas, further study may reveal a historical coherence worth of taxonomic recognition, but these populations are not afforded such at this time. HRUSA: VARIATION IN RHODODENDRON OCCIDENTALE 139 Rhododendron occidentale (Torr. & A. Gray) A. Gray in W. H. Brewer & S. Watson, Botany of California 1:458. 1876. Azalea occidentalis Torr. & A. Gray in Torr., Botany of the Expedition, Pac. Railr. Rep. 4:116. 1857.— Type: USA, California, Sonoma Co., Laguna de Santa Rosa, 1854, J. M. Bigelow s.n. (lectotype NY!, designated by K. A. Kron 1993; isolectotypes: GH!, PH! here designated). The lectotype of Azalea occidentalis Torr. & A. Gray (Torrey 1857) designated by Kron (1993) (NY!) is clearly part of the ONCR group, as are duplicates at GH! and PH!. Rhododendron occidentale (Torr. & A. Gray) A. Gray in W. H. Brewer & S. Watson var. occidentale. Synonyms - Rhododendron sono- mense Greene, Pittonia 2:172. Sept. 1891. Rhododendron occidentale var. sonomense Re- hder, Monograph of Azaleas; 127, 1921 (var. novus based on Rhododendron sonomense Greene). —Type: USA, California, Sonoma Co., “near Petaluma,’ May 24, 1891, Miss Carlton s.n., (lectotype: here designated, left specimen, NDG 037326!, [pre-1966 Herbarium Greeneanum 10866)). There are two individuals mounted on NDG 037326, both apparently sourced ‘‘from Peta- luma, Miss Carlton, May 24, 1891” (on ticket in pocket). The left specimen is the more complete and is here designated as lectotype. This is the only known azalea specimen seen by Greene whose gathering predates the protologue and was also purportedly collected within the species’ described distribution. Both left (lectotype) and right (non-type) specimens are considered here to be a local form of the typical variety. Rehder consistently used the term “comb. nov.” when shifting epithets or changing ranks. Although he clearly based his new variety on R. sonomense Greene, with a citation of the basionym and a partially accurate paraphrasing of Greene’s de- scription, after the intraspecific epithet Rehder added ‘“‘var. nov.”’ It is therefore interpreted here as a newly proposed variety with same type and epithet as Rhododendron sonomense Greene. Two names have been misapplied to R. occidentale var. occidentale. The first is Rhodo- dendron calendulaceum (Michx.) sensu Hook. & Arnott, Bot. Beechey Voy. 362. 1839, not sensu Torr., Fl. N. Middle United States 1:425. 1824. The other is Azalea calendulacea Michx., sensu G. Bentham, Plantae Hartwegianae 321, 1848, not sensu Michaux, Fl. Bor. Amer. 1:151. 1803. Both of these misapplications were based on specimens (!) collected near the coast by David Douglas (probably San Francisco Bay region) and T. Hartweg (“in uliginosus prope Santa Cruz’’), respectively. 140 MADRONO The typical variety, Rhododendron occidentale var. occidentale, occurs in the Outer North Coast Ranges, from northern Mendocino County south to northern Monterey and San Benito counties, California. Rhododendron occidentale (Torr. & A. Gray) A. Gray var. californicum (Torr. & A. Gray in Durand) Hrusa, comb. et stat. nov. Basionym: Azalea californica Torr. & A. Gray in Durand, Plantae Prattennianae Californicae, J. Acad. Nat. Sci. Philadelphia ns 3(2):94, June 1855, not Azalea californica (Hook.) Kuntze, Rev. Gen. 2:387. 1891. —Type: USA. “Nevada, California, Henry Pratten” s.n., s.d. (lectotype: PH!, here designated). An Azalea specimen at PH! with an annotation by Durand attributing the epithet “‘californica”’ to Gray and labeled as “Azalea californica T. & G.,” is here designated as lectotype for Azalea californica Torr. & A. Gray in Durand (Durand 1856). The protologue indicates the specimen was collected in “shady hills along Deer Creek” without mention of collection date or further locality information. While the lectotype label has written on it only ““Nevada, California” without the locality “‘Deer Creek,’ in the introduction to “Plantae Pratennianae”’ Durand equated these locations by describing Pratten’s collecting localities as “‘in the vicinity of Nevada, a place situated along Deer Creek ...”’ According to Durand (1856), Pratten’s Deer Creek collec- tions were taken in 1851. This proposed lectotype of Azalea californica Torr. & A. Gray ex Durand (PH!) is matched by the SNPR azalea. The variety occurs in the Interior North Coast Ranges, the Peninsular Ranges, and the Sierra Nevada of California. Rhododendron occidentale (Torr. & A. Gray) A. Gray var. paludosum Jeps., Man. FI. Pl. Calif., 741. 1925. , boldt Co., ““Fortuna to Eureka, filling sedgy bogs in the meadows near Loleta”’, 1902, Jepson #1916 (holotype: JEPS!). The holotype of Rhododendron occidentale var. paludosum Jeps (JEPS!) is unequivocally repre- sentative of the KR azalea. This variety occurs in the Klamath Ranges of Mendocino Co., California, to Douglas Co., Oregon. Incertae sedis: Azalea nudiflora var. ciliata Kellogg, Proc. Calif. Acad. ser. 1, 1:60. 1855 (published 1873). (7.v., probably destroyed 1906). Reported by Kellogg as “‘from the interior’, the specimen is not currently extant. KEY TO VARIETIES 1. Leaf abaxium generally with unicellular pu- bescence, not obscuring surface; multicellular [Vol. 59 trichome type on mid-vein or secondary veins of one kind, glandular or eglandular; corolla buds and open corolla white, often with a blush of pink at the lobe summits, rarely with a light pink blush in the tube, the tube otherwise white to greenish-white. Sierra Nevada, Peninsular Ranges, (inner North Coast Ranges as intro- gressants), California.......... var. californicum 1’ Leaf abaxium generally glabrous, rarely with a few unicellular hairs; multicellular trichomes on the midvein mixed eglandular and glandu- lar types; corolla buds and open corolla tube and veins usually colored dark to light pink, or less often pure white, often with a pale to dark bt pigmentation in the limb 2. Abaxial secondary veins with multicellular trichomes, these often mixed eglandular and glandular types; tertiary veins often with associated multicellular trichomes; dormant bud bracts densely pubescent, the margin trichomes glomerate; twigs of the current season densely unicellular pubescent. Outer North Coast Ranges, California var. occidentale 2' Abaxial secondary veins generally lacking multicellular trichomes, or of a single type; absent on the tertiary veins; dormant bud bracts glabrous or thinly pubescent, the margins most often ciliate, occasionally glomerate; twigs of the current season glabrous to subglabrous. Klamath Ranges of California and Oregon, from Humboldt Co., California north to Winchester Bay, OVecOn eae eiads ee owes eee var. paludosum ACKNOWLEDGMENTS This work summarizes a portion of a dissertation produced in partial fulfillment of the degree of Doctor of Philosophy in Plant Biology at the University of California, Davis. Funds were provided by a post- graduate research associateship in the section of Plant Biology, and grants from the Jastro-Shields Fund and Hardmann Foundation. A special acknowledgment to the late G. Ledyard Stebbins for his continued interest and many insightful discussions; also to the late Grady L. Webster, then Director of the J. M. Tucker Herbarium for providing an Acting Assistant Curator position for the author while this work was completed. Committee members A. T. Leiser, Lin Wu in whose laboratory the isoenzyme analyses were performed, and the late John M. Tucker are also gratefully acknowl- edged. Review and discussion by R. Bencie (HSC) and G. Wallace (RSA) of early drafts, and later by M. Ritter (OBI) and anonymous reviewers were extremely helpful. LITERATURE CITED ANACKER, B. L., J. B. WHITTALL, E. E. GOLDBERG, AND S. P. HARRISON. 2010. Origins and conse- quences of serpentine endemism in the California flora. Evolution 65:365—376. AXELROD, D. I. 1962. A Pliocene Sequoiadendron forest from western Nevada. University of California Publications in Geological Sciences 39:195—267. 1976. History of the coniferous forests, California and Nevada. University of California Publications in Botany 70:1—62. 2012] BREAKEY, E. P. 1960. Notes on our western azalea Rhododendron occidentale. Journal of the American Rhododendron Society 14:1—8. DRAKE, M. H. 1987. Nutrient uptake and growth characteristics of Rhododendron occidentale from serpentine and non-serpentine soils. M.S. Thesis, University of California, Davis, CA. DURAND, E. M. 1856. Plantae Pratennianae Californi- cae. Journal of the Academy of Natural Sciences of Philadelphia ser. 2, 3:79—104. GILLESPIE, J. 1991. The causes of molecular evolution. Oxford University Press, New York, NY. Hrusa, G. F. 1991. A morphometric analysis of variation in Rhododendron occidentale (TY. & G.) Gray among populations native to ultrabasic and non-ultrabasic substrata (Ericaceae: Rhododen- droideae). M.S. thesis, University of California, Davis, CA. JEPSON, W. L. 1925. A manual of the flowering plants of California. University of California Press, Berkeley, CA. . 1939. A flora of California. Vol. 3. University of California Press, Berkeley, CA. JONES, J. R., T. C. RANNEY, AND N. P. LYNCH. 2007. Ploidy levels and relative genome sizes of diverse species, hybrids, and cultivars of Rhododendron. Journal of the American Rhododendron Society, 221-227. KELLOGG, A. 1855. Proceedings of the California Academy of Natural Sciences, ser. 1, 1:60. Kron, K. A. 1993. A revision of Rhododendron section Pentanthera. Edinburgh Journal of Botany 50:249—-364. , L. M. GAWEN, AND M. W. CHASE. 1993. Evidence for introgression in azaleas (Rhododen- dron: Ericaceae): chloroplast DNA and morpho- logical variation in a hybrid swarm on Stone Mountain, Georgia. American Journal of Botany 80:1095—1099. KRUCKEBERG, A. R. 1984. California serpentines: flora, vegetation, geology, soils, and management problems. University of California Press, Berkeley, CA. LEISER, A. T. 1957. Rhododendron occidentale on alkaline soil. Rhododendron and Camellia Year- book. 47-51. HRUSA: VARIATION IN RHODODENDRON OCCIDENTALE 141 MAJor, J. 1967. Potential evapotranspiration and plant distribution in western states with emphasis on California. Pp. 93-126 in R. H. Shaw (ed.), Ground level climatology. American Association for the Advancement of Science, Washington DC. MOSSMAN, F. 1972. With camera, white umbrella, and tin pants in R. occidentale heartland. Journal of the American Rhododendron Society 26:4. 1974. The western azalea, Rhododendron occidentale. Journal of the American Rhododen- dron Society 26:4. . 1977. The western azalea on Stagecoach Hill. Pacific Horticulture 38:28—33. AND B. SMITH. 1968. Further trips to the Rhododendron occidentale patches. Journal of the American Rhododendron Society 22:3. AND . 1969. Rhododendron occidentale one species or many? Journal of the American Rhododendron Society 23:3. Munz, P. A. 1974. A flora of southern California. University of California Press, Berkeley, CA. AND D. D. KEck. 1959. A California flora. University of California Press, Berkeley, CA. RAVEN, P. H. AND D. I. AXELROD. 1978. Origin and relationships of the California flora. University of California Publications in Botany 72:1—134. SLEUMER, H. 1949. Ein system der gattung Rhododen- dron L. Botanische Jahrbucher fur Systematik, Pflanzengeschichte und Pflanzengeographie 74: 511-533. SNEATH, P. H. A. AND R. R. SOKAL. 1973. Numerical taxonomy: the principles and practice of numerical classification. W. H. Freeman & Co., San Fran- cisco, CA. STEBBINS, G. L. AND G. F. HRUSA. 1995. The North Coast biodiversity arena in Central California: a new scenario for research and teaching processes of evolution. Madrono 42:269-294. TORREY, J. 1857. Description of the general botanical collections. Pp. 61-167 in U.S. War Department. Reports of explorations and surveys, to ascertain the most practicable and economical route for a railroad from the Mississipi river to the pacific Ocean. Vol. HI, Part V. No. 4. Report on the botany of the expedition, Washington, DC. WILSON, E. H. AND A. REHDER. 1921. A monograph of azaleas. Harvard University, Cambridge, MA. [Vol. 59 ~ MADRONO 142 yooIg so1oopy Joddn ‘jsol0-+{ 93819 sepesogd se] DAOID vSOdIIvI,, “YIeg [PUONRNY dIWIOSO XK sos] Addey ‘y1eg [euOneN o}IUWIOSO XK DAOID JUBID ‘yIeg [eUONeN vrlonbas DAOIQH POOMPdY 199k py ssuridg yopieg pure yoo1g Joddod[ry jo uonsassojuy Cpc AMP SuUOTR ‘YIIOD [YA oN * sureyunoyy ouloese ues ‘uodkueyD yeq Jo peoy ‘dures inod¢g [ALD sprerp vsezy “ureyunoyy oyUIDeS URS yieg 81S OyoURY voORUIRAND ‘UdI[D PvIezZV yieg 21e1S UlvjUNO|Y IeWOjeg ‘AaT[eA sueOdg JAsddy IOI ‘S'n jo q wy | xoidde py suey ‘TITH yorossseys DAIOSIY BLIS KIeZY ‘dPIAAD[UTYPIPY punoisdurey Ayuiy Jo q ¢ AMP Jo qs uy ¢ “xoidde ‘yaorD uotds09¢g ‘YIIQ swMool_ “ye 1g ulseg 3sIg INSYJOH JO S WY OQ] ~ ‘AITVA SSVID ‘SYIID O1apRosag ‘asuvyYy uRpIqey yooulg uoWNpY ‘Wsa104 [eJUsUITIDdxY Nospolg ON Yoo UOTIDIg JOOYSS Jo SIaJBMpPRdY “UIRJUNOPY Poy jNoOYyooyT sury ezueuog Jo N Wy 6~ ‘ospry sury PZUBUOY Opis NI “OAT AULT, YO iseq siayempevoy ‘AleINqL poweuuy 30q PIU IUILIeG ‘vary [eoluRjog AgeA ALsaINg ‘jso1O4 [eUOMeNY seuny[d Pa es[Ndg JO AA MW 8/T “OL AMY oul] “OD vurouos/sedenN AA Apo VIPSUILUT ‘py [OoYysS AayeA ZuUPI- Yo suse ensy jo siajyeMproy * ulejyuNOPy yorqsopH ospry su0Yydaya Ll “YOotD WS YO} A “POOMPITIA JO MA UD 6~ “OE AMF 211 | AMP tuvonsdn ury / “OAT OLIBARN suryied durey yorhoog jo q ury | ‘xoidde yoorn yorhoog ‘siedjeue | juno DJAIOSIY 9181S POOMPSY SUOIISULIY JO NI ATOVeIPOUIUUT BdIY UOI]RIIDIY eIS YOUD unsny ‘aye] pooMmpay Moyoq UoAUPd ‘YIID WIRTH JSOIO-] [RUOT]BNY OUTIDOpUs|I ‘YIID JozZuaI pure yoIQ Auo.S sWT soUaNTUOoS ‘py ureyUNOPY Iepay SNUIWID] PY IYOLIYIH saoqge Ulry C°] “YIuUID uosuvyeN Ppl OL ‘S'N JO AN GY p~ “Pa ITH Youos YO “PPra JO A MW OT “1TE AMH UO ospiiq Jo weansdn wy Z/] ~ “YaI1D MOD 661 ‘SN JO M WY L~ “py ureyunoyy Jeyod 14s udLig.O JO MS WY C'¢ ‘py (OWI AM) UTeJUNOPY DUO (AdssouUdxY OYeT) WieG UUOD 1e BZ] AM UMOJIPPIY pure AagyeA odog usdmjaq ‘uOUkD sng 8c] AM JO uonounl jo N wy oz ‘pay AqypeA sary 87] AMH JO uonounl jo N wy 7 “py AIBA Softy uonount py AopyeA sopyD-pra ABA odog Jo q Wy ¢'¢ “yao1D odod uondtosaq OOS OOr I O0€ I 000C 0091 SCS 009 I 008 I OS9l OOS I SSI SL OS6 Ore Orv OScl OOOT Ors I SOT OOS SLC OSt OOTT St O8t cot SLV O8C OOS OO O06t OS 08 OOC SVC OVC SLI uOneAIA | XIGNdddVY NM.OE.VC.CTI N.0C,9¢.611 M.0C.¢¢.611 M.S 1.8S.811 MOC, ee .0C1 M.S 1,0€.cC1 N.00.S0.9 11 NM.OV.eV9T 1 M.00,9¢.9T 1 NM.00,¢S.9 TI N.00,90.0C1 M.S L,V0.vC1 NLL. Croc N.00,¢ 1.1 M.SC.SCoICl N.OV. 8.01 N.00,6¢.¢Cl N.00,6€.CC1 NM.tl,6$.0C1 M.S TLC M.SV.LEOCCI NM.00,9C.CC1 M.00,L0.¢C1 M.00,00-¢C1 Mut L,9t.ccl MCI, 100€C1 M.0E EE 0CCI M.00,9C.CC1 M.0C,68.¢Cl NM.0¢,08.€C1 M.S L,OPoecl N.SV.VV-ECI Mi8lL.ecoccl M.LV.9COCTI M.00,CC.CCI NM.0¢,1C.CCI M.CV.ACoCCI “NOVEL 8E "NUS OELE ‘NOOO. PPLE “N00,SP.9€ “NOE, £0.6€ *N00,01.6€ “NUSPLYOEE “NUSVLVEE “NUS 1,89 C€ ‘N.00,070€€ ‘N00, ¢ 1.1? *N.00,SS 00 ‘NLIL,9061P ‘N.00,01 LE ‘NO, CPo9E “NOES o8€ “NLOS PS 06 “NUS TL LOoTv “N.8¢,00.0P “NLOELV.6E “N00,S€68€ “NOE ,OT.8E ‘N00, 0C.0P "NLS L016 “NUSTPS OLE “NUIT, VEo8E ‘NOE. LT.6€ "N100,0C.8€ ‘N.OE, 60.19 "NWOT, SSCP “N00, 1-TP “NUS L.€05CP “NOT, 67-8€ “NWO, TPo8€ "N00, €£.8€ “NUSPTEo8E "NOP. 8E 08 apnyisuosapnyneq] eden esodieyy evsodiuvyyy aIeIN IO0e[q ae IPISIOATY OPISIIATY ossIq urs OsaIq ues eels +PIOQuny AVWULL], ZnId vues oulusg ues ePE IOP tal 1 PC AWULLT seun[d seuuntd BUIOUOS BUIOUOS AYU oulso pus ULIe|y BULOUOS esnjoy eUIOUOS dION [9d UO ‘sevejsnoq UO ‘ourydssor WO ‘ourydasor eden aye Tseden eden eden eden A\uno, ICO Vo9 tv “AVC pure VC) 1® pasnoy O18 SIOYONOA IANKLIUDSIIdIY “payloads doYyM Jd99x9 VIUIOJI[LD Ul 3B SOUNOD [PY “SI9JOUI UL UDAIS 91K SUONRADY “] BINS UL ISOY} 0} pUOdsaIIOS sioquInu UOHnK[Ndog SNOILV1NdOd FTIFINAGCIOIO NOYUTNAGOGOHY AO LSIT GAONA adda eO*L) MADRONO, Vol. 59, No. 3, pp. 143-149, 2012 MEASUREMENT OF SPATIAL AUTOCORRELATION OF VEGETATION IN MOUNTAIN MEADOWS OF THE SIERRA NEVADA, CALIFORNIA AND WESTERN NEVADA DAVE A. WEIXELMAN USDA Forest Service, Range Ecologist, 631 Coyote Street, Nevada City, CA 95959 dweixelman@fs.fed.us GREGG M. RIEGEL USDA Forest Service, Pacific Northwest Region, Area Ecology Program, 63095 Deschutes Market Road, Bend, OR 97701 ABSTRACT The presence of autocorrelation invalidates all standard statistical tests unless special corrections are made. Because of this, it is important to know the degree of spatial autocorrelation in order to know how to sample. Mountain meadows were sampled to determine spatial autocorrelation of vegetation at the plant community level. A total of 40 meadows were sampled in the eastern Sierra Nevada, California. At each meadow a dominant plant community was selected for sampling. Sampling consisted of placing 10 * 10 cm quadrats at 1-m intervals on a 20-m transect and recording the presence for all vascular plant species rooted in the quadrats. Sites varied in plant species composition and number of species present. For each plot, ordination analysis in the form of reciprocal averaging was used to derive positions for each quadrat on axis |. The scores from axis 1 were analyzed by semivariance to obtain the spatial dependence of the quadrats. Overall, three semivariance patterns were seen; A) plant communities that were autocorrelated at distances of less than one meter: B) communities that were autocorrelated between | m and 15 m; C) communities that were autocorrelated at distances greater than 20 m. Results indicate that for semivariogram type B, on average, sites were autocorrelated to a distance of 3.6 m, meaning that quadrats separated by greater than 3.6 m were independent. Beta diversity was significantly (P < 0.05) lower for semivariance type C than for either semivariance types A or B. These results are useful for determining spacing of sample points in mountain meadows to ensure spatial and statistical independence for presence/absence data. Key Words: Beta diversity, meadow, sampling, spatial autocorrelation, vegetation. One of the general patterns in ecology is that, on average, sites that are closer will be more similar (Fortin 1999). This is known as positive spatial autocorrelation (Mistral et al. 2000). The presence of autocorrelation invalidates all stan- dard statistical tests unless special corrections are made (e.g., Dale et al. 1991). Because of this, it is important to know the degree of spatial autocor- relation in order to know how to sample. In this study, mountain meadows in the eastern Sierra Nevada, California were sampled to estimate the amount of autocorrelation within plant commu- nity types. Methods for analyzing spatial autocorrelation in ecology have commonly been for univariate data (Fortin 1999). Such methods have been used, with single-variate indices such as Moran’s I or Geary’s c, on individual variates such as fruit production (Koenig and Knops 1998), plant height and diameter (Kuuluvainen et al. 1998), flower and vegetative characters (Chung and Noguchi 1998) and gene frequencies (Sokal et al. 1998). However, communities comprise many Species, and are hence multi-variate. To examine the spatial pattern of gradients, workers such as Palmer (1988), Jonsson and Moen (1998), Ohlson and Okland (1998), Meisel and Turner (1998), and Wagner (2003) summarized the whole community into ordination scores, and analyzed them by semivariance. This approach examines spatial variation in the major gradients, using only that fraction of the variation in species composition that is captured in the ordination. Our aim is to document the spatial pattern for Sierra Nevada mountain meadow communities. We sampled with a fixed quadrat size (10 x 10 cm) using quadrats at Il-m intervals along a transect of 20 m length. The size of the quadrat from which species associations are calculated, as well as the length of transects may influence the detected association pattern (e.g., Ver Hoef et al. 1989). With very long transects, covering multiple habitat types and communities, environmental gradients will be captured (Rydgren et al. 2003). Such gradients are likely to occur as a result of environmental variation rather than as a cause of species interactions within specific plant commu- nities. At the opposite end, with very short transects, stochastic effects, due to few individu- als, may restrict the ability to detect significant spatial patterns among species (Jonsson and Moen 1998) within communities. The focus of 144 MADRONO [Vol. 59 Fic. 1. this paper is at intermediate scales, e.g., at the plant community level, where patterns in species associations and beta-diversity relate to the intrinsic patch sizes present within the plant community being sampled. METHODS Study Sites Forty meadow sites were sampled in the eastern Sierra Nevada of California and Nevada at latitudes between 38° and 40° (see Fig. 1). All sites were located on National Forest lands. Livestock grazing had occurred on all sites since the 1860’s and the sites are generally representa- tive of the history of livestock grazing use in the Sierra Nevada. Sampling occurred between June of 1994 and August of 1996. Elevation of sites ranged from 7100 to 9600 feet. Depth to water table varied from O (at the surface) to 50 cm in mid-summer. All sites were classified as wet or moist meadow types using the USDA Forest Location of study sites in Sierra Nevada mountain range, California and eastern Nevada. Service classification for Sierra meadows (Weixel- man et al. 1996) and the dominant soil taxon was Typic Cryaquoll (Soil Survey Staff 1998). Species composition generally consisted of sedges, rushes, and forbs. Dominant species included Nebraska sedge (Carex nebrascensis Dewey), blister sedge (Carex vesicaria L.), western aster (Aster occi- dentalis Nutt.), Kentucky bluegrass (Poa praten- sis L.), and yarrow (Achillea millefolium L.). At each site, a transect line 20 m in length was randomly located within a homogenous plant community and 10 < 10 cm quadrats were placed at 1 m intervals along the transect for a total of 20 quadrats on each transect. Previous studies using presence absence methods in mountain meadows have used a 10 * 10cm quadrat size for sampling (Mistral et al. 2000; Moseley et al. 1986, 1989; Weixelman et al. 1996; USDA Forest Service 2008). The vegetation data consisted of presence/absence for each vascular plant species rooted in the 10 X 10 cm quadrat. Plant nomenclature used in this paper conforms to Hickman (1993). 2012] Semivariance “Sill? “Nugget” ee “Range” Distance Fic. 2. Theoretical interpretation of a semivariogram with an asymptotical model showing the proportion of variance found at increasing distances of paired samples. The ‘sill’ is the variance around the average value of the variable. The ‘range’ is the maximum distance at which samples show spatial dependence, the ‘nugget’ is the variance found at a scale finer than the smallest sampling scale. Statistical Analysis Species presence/absence data for each quadrat on a transect line was analyzed using reciprocal averaging (RA), an ordination program. RA is an indirect ordination technique which extracts gra- dients present in the species composition data assuming a unimodal relationship between the species abundance and the gradients (Hill 1973). To obtain a regionalized variable reflecting species composition, we used the ordination scores from reciprocal averaging (RA). RA then calculated scores on axis | for each quadrat. If the species composition in neighboring quadrats 1s similar, the ordination should place these quadrats close to each other in ordination space. However, if the species composition is unrelated to the spatial location of the quadrats, samples close to each other in ordination space may be considered randomly located in geographical space. Each site was analyzed for spatial independence using the data from the 20 quadrats. The ordination analyses were performed with the package PC-ORD version 3.0 (McCune and Mefford 1997). Scores for each quadrat on axis 1 of RA were then analyzed by semivariogram (Robertson 1987) to examine the distance at which quadrats were autocorrelated within transects. Semivario- grams are plots of the spatial dissimilarity (measured by semivariance) between points separated by known distances, plotted against those distances. Normally, points in close prox- imity are more similar than points farther apart, so that semivariance among points increases with distance until a maximum semivariance, called the sill, is reached (see Fig. 2). The distance at which the semivariance stops increasing is called the range, and the point where the semivariance begins (distance equals zero) is called the nugget. WEIXELMAN AND RIEGEL: AUTOCORRELATION OF VEGETATION IN MEADOWS 145 Samples separated by distances closer than the range are statistically dependent, while those separated by distances greater than the range, are not, because at distances greater than the range the semivariance equals the sample vari- ance, implying zero spatial correlation (Trangmar et al. 1985). Using the geostastistical package GS+, we calculated semivariograms of the quadrats using the axis | scores from ordination. Semivariances were calculated up to within 10 pairs of the maximum distance between all points (1.e., 20 m). For transects that exhibited spatial autocorrela- tion, semivariogram models for range, nugget, and sill were fit using a non-linear least squares technique (Robertson 1987). These models includ- ed linear with a sill, spherical, exponential, and Gaussian curves. We chose the best fitting of these four curves based on the best fit of the residuals about the curve, particularly at the sill and nugget ends of the curve. If a transect exhibited zero autocorrelation, the sample variance was used for sill and nugget variances and zero was used for the range. If a transect exhibited spatial autocorrela- tion with no sill it was considered to have a nonstationary mean (Trangmar et al. 1985). In this case, the samples were dependent out to the maximum distance of the transects, in this case 20 m. Because mountain meadows are made up of a patchwork of a number of plant communities, the size of each community is sometimes less than 20 m and sometimes greater than 20 m. Based on the author’s experience, at distances much greater than 20 to 30 m, changes in environment, including changes in hydrology, become significant and changes in plant composition are more likely due to environmental gradients. Beta Diversity Robert H. Whittaker (1960) defined beta diversity as the variation in species composition among sites in a geographic area. In our case, this is the variation in species composition among quadrats along the transect line. Whittaker (1960) established a straightforward measure of beta diversity, which will here be called fy: By, =(s/a)—1 where f,, = beta diversity, s = total number of species occurring on the transect, and aq = average number of species occurring in the quadrats. The measure f},, 1s easy to calculate and explicitly relates the components of diversity a and fp, to overall diversity, s. Statistical Tests All statistical tests were performed using SPSS version 9.0 (SPSS 1998). Tests of significance 146 MADRONO [Vol. 59 Y oe) 5 E as & Ao) n 0.0 4.5 9.1 13.6 18.2 Separation Distance (m) iP) oO c Bo FE D> n 0.0 4.5 9.1 1st 18.2 Separation Distance (m) dyne< 9) Y & = E 9) Yn 0.0 4.5 91 13.6 18.2 Separation Distance (m) Fic. 3. Semivariograms of data taken from three individual sites to illustrate three types of spatial pattern. The dashed lines denote the semivariance around the average value. Type A: communities that showed no clear spatial pattern, 1.e., autocorrelation of less than 1 m; type B: meadow communities with a clear range and sill; and type C: meadow communities with an increasing semivariance beyond 20 m separation, i.e., autocorrelation to distances beyond 20 m. Also shown are stylized diagrams illustrating species turnover (curved lines on the right) along a transect. 2012] TABLE 1. MEAN AND STANDARD DEVIATION OF BETA DIVERSITY AND AVERAGE NUMBER OF SPECIES BY TYPE OF SEMIVARIOGRAM. Superscripts that are different within the same column represent a statistically significant difference (P < 0.05). Average Average Type of Beta number elevation semivariogram diversity of species (m) A (n = 20) Mean 220° LI? 26/97 SD 1.28 6 328 B (n = 13) Mean Oo lee 2686" SD 0.88 5 208 C (n = 7) Mean 133° q 2668* SD 92 2 360 were used to determine if differences existed in beta diversity and number of species on a site for classes of differing spatial pattern. RESULTS Results of ordinating quadrat data followed by semivariance analysis showed three general spa- tial patterns of vegetation (see Fig. 3): Type A, B, and C. The semivariogram for type A communi- ties was generally flat, with no range evident. Therefore, plant species in this type were auto- correlated at distances of less than 1 m, which was the smallest scale measured in this study. These sites were higher in beta diversity and contained many species. The semivariogram for type B communities exhibited a distinct range and sill between 1 m and 20 m. Sites in type B were positively autocorrelated to a distance of 3.6 m on average. The semivariogram for type C communities showed a continual rise with no sill or range visible. Quadrats in this type were positively autocorrelated at distances greater than 20 m. In type A communities, species turnover as indicated by the f-diversity index was high (see Table 1) as was the average number of vascular plant species on the sites. In type B communities, B-diversity was also high. Sites in this type typically had high numbers of species with rapid turnover in composition along transects as indicated by the high f-diversity index for these sites. The semivariogram for type B showed a range at distances ranging from 3 m to 15 m (Fig. 3, type B). The average distance at which quadrats were autocorrelated in type B commu- nities was 3.6 m. The P-diversity index averaged nearly the same as for type A (Table 1). The number of species for type B was also similar to type A. Type C plots had the lowest B-diversity of the three semivariance types (Table 1). In addi- WEIXELMAN AND RIEGEL: AUTOCORRELATION OF VEGETATION IN MEADOWS 147 x< 0) TC = i= oO _ 0) a O < je) aa) A B C Spatial Pattern Type Fic. 4. Comparison of beta diversity among the three types of spatial pattern seen (A, B, and C), see text for explanation of types of spatial pattern. Error bars represent 95% confidence interval. tion, type C had the lowest average number of species on each site. Significant differences (P < 0.05) in B-diversity were found between type A and B as compared to type C communities. DISCUSSION Spatial patterns of herbaceous species in meadows vary with environment, species ecolog- ical characteristics, species interactions, grazing, and ground disturbing activities. In this study, sites were located within a plant community in order to limit variation based on environmental factors. Therefore, results of this study are most applicable in situations where one is interested in spatial variation within meadow plant communi- ties that typically occur in the Sierra Nevada. The three types of plant community spatial patterns were different in their pattern of spatial autocorrelation (Fig. 3). In this study, plant community spatial patterns were positively auto- correlated according to the following rank distanc- es: type A < type B < type C. Type C communities were typically homogeneous stands dominated by clonal graminoid species including sedges and/or rushes. These stands had fewer species than either type A or B. In contrast, type A and B communities were composed of many small forbs including species indicating disturbance as well as clonal graminoid species. Most species present were either obligate wetland (OBL), facultative wetland plant species (FACW) or facultative wetland species (FAC) using the U.S. Fish and Wildlife wetland rating system for plants (U.S. Fish and Wildlife Service 1988). Robert H. Whittaker (1960) defined beta diversity as the variation in species composition among sites in a geographic area. In our case, this is the variation in species composition among 148 MADRONO quadrats along the transect line. Using Whi- taker’s equation provides an index of the variation in number of species present in each quadrat along the transect line, in this case 20 m. For homogenous plant communities, such as clonal patches of sedges, the beta diversity was low. The 95% C.I. for beta diversity in type C ranged from 0.55 to 2.2. Plant communities in this group were typically clonal, rhizomatous sedge species and included Nebraska sedge (Carex nebrascensis), Blister sedge (Carex vesi- caria), and analogue sedge (Carex simulate Mack.). Beta diversity in type A and B were higher due to more species and higher density of individual plants and ranged from 1.9 to 3.4 in type B and 2.2 to 3.1 in type A (95% C.1). Type A and B communities were not different in beta diversity even though semivariance diagrams were different between the two groups. Type A communities did show a higher beta diversity for some plots but generally overlapped type B when the 95% C.l. was plotted (see Fig. 4). Type A communities were autocorrelated at distances of less than one meter, while type B communities were autocorrelated at an average distance of 3.6 m. Type and A and B communities were composed of early successional forbs and a mix of graminoid species. Typical plant species present in type A and B communities included Kentucky bluegrass (Pod pratensis), western yarrow (Achillea millefolium), and western aster (Aster occidentalis). These types were representa- tive of disturbance communities and meadow types with larger seasonal fluctuations in water table. Significant differences (P < 0.05) in B-diversity were found between type A and B as compared to type C communities. Thus, knowing the fB- diversity of a plant community using the methods described here could potentially be used to determine which type of spatial pattern exists in a meadow community. Spatial pattern will depend both on the size of the quadrat being used and the distance separating the quadrats. Spatial pattern may also vary depending on soil moisture types, drier meadows would be expected to be different than wet meadows based on wider spacing of plants in drier meadows. For practical considerations and sampling in the field, beta diversity can be used as a rough indicator of spatial autocorrelation in plant communities. When determining rooted frequen- cy using the 10 < 10 cm frame as in this study and | meter spacing along transect lines, a beta diversity value using the Whitaker index of less than about 2.2 would indicate, with high proba- bility, a spatial autocorrelation distance greater than 20 m. While a beta diversity index of greater than 2.2 would indicate a spatial autocorrelation distance closer to that expected for semivariance types A and B. It is reasonable to expect the spatial [Vol. 59 relationships of the plant community to vary in different environments, and some care must be taken when trying to extrapolate the results of this work. ACKNOWLEDGMENTS The authors would like to thank Desi Zamudio for help with soil identification and Karen Zamudio for help with vegetation work. LITERATURE CITED CHUNG, M. G. AND J. NOGUCHI. 1998. Geographic spatial autocorrelation of morphological characters of the Hemerocallis middendorffii complex (Lilia- ceae). Annales Botanici Fennici 35:183—189. DALE, M. R. T., D. J. BLUNDON, D. A. MACISAAC, AND A. G. THOMAS. 1991. Multiple species effects and spatial autocorrelatiojn in detecting species associations. Journal of Vegetation Science 2: 635-642. FORTIN, M.-J. 1999. Spatial statistics in landscape ecology. Pp. 253-279 in J. M. Klopatek and R. H. Cardner, (eds.), Landscape ecological analysis: issues and applications. Springer-Verlag, New York, NY. HICKMAN, J. C. (ed.) 1993, The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. HILL, M. O. 1973. Reciprocal averaging: an eigenvector method of ordination. Journal of Ecology 61: 237-251. JONSSON, B. G. AND J. MOEN. 1998. Patterns in species associations in plant communities: the importance of scale. Journal of Vegetation Science 9:327—332. KONIG, W. D. AND J. M. H. KNops. 1998. Testing for spatial autocorrelation in ecological studies. Eco- graphy 21:423—429. KUULUVAINEN, T., E. JARVINEN, T. J. HOKKANEN, S. ROUVINEN, AND K. HEIKKINEN. 1998. Structural heterogeneity and spatial autocorrelation in a natural mature Pinus sylvestris dominated forest. Ecography 21:159—-174. McCune, B. AND M. J. MEFFORD. 1997. PC-ORD. Multivariate analysis of ecological data, Version 3. MjM Software Design, Gleneden Beach, OR. MEISEL, J. E. AND M. G. TURNER. 1998. Scale detection in real and artificial landscape using semivariance analysis. Landscape Ecology 13: 347-362. MISTRAL, M., O. BUCK, D. MEIER-BEHRMANN, D. A. BURNETT, T. E. BARNFIELD, A. J. ScoTT, B. J. ANDERSON, AND B. J. WILSON. 2000. Direct measure- ment of spatial autocorrelation at the community level in four plant communities. Journal of Vege- tation Science 11:911—916. MOSELEY, J. C., S. C. BUNTING, AND M. HIRONAKA. 1986. Determining range condition from frequency data in mountain meadows of central Idaho. Journal of Range Management 39:561—565. ; AND 1989. Quadrat and sample sizes for frequency sampling mountain meadow vegetation. Great Basin Naturalist 49: 241-248. OHLSON, M. AND R. H. OKLAND. 1998. Spatial variation in rates of carbon and nitrogen accumu- lation in a boreal bog. Ecology 79:2745—2758. 2012] PALMER, M. W. 1988. Fractal geometry: a tool for describing spatial patterns of plant communities. Vegetatio 75:91—102. ROBERTSON, G. P. 1987. Geostatistics in ecology: interpo- lating with known variance. Ecology 68:744-748. RYDGREN, K., R. H. OKLAND, AND T. OKLAND. 2003. Species response curves along environmental gra- dients. a case study from SE Norwegian swamp forests. Journal of Vegetation Science 14:869—880. SOIL SURVEY STAFF. 1998, Keys to soil taxonomy, 8th ed. USDA, Washington, DC. SOKAL, R. R., N. L. ODEN, AND B. A. THOMSON. 1998. Local spatial autocorrelation in biological variables. Biological Journal of the Linnean Society 65:41—62. SPSS Inc. 1998, SPSS base 9.0 for Windows user’s guide. SPSS Inc., Chicago, IL. TRANGMAR, B. B., R. S. YOST, AND G. UEHARA. 1985. Application of geostatistics to spatial studies of soil properties. Advances in Agronomy 38:45—94. U.S. FISH AND WILDLIFE SERVICE. 1988, National list of vascular plant species that occur in wetlands. U.S. Fish & Wildlife Service, National Conservation WEIXELMAN AND RIEGEL: AUTOCORRELATION OF VEGETATION IN MEADOWS 149 Training Center, Shepherdstown, WV, Website: http://library.fws.gov/WetlandPublications.html [ac- cessed 31 March 2012]. USDA FoREST SERVICE. 2008, Natural Resources Information Service (NRIS). Ecosystem Manage- ment Coordination (EMC), Washington, DC, Web- site: http://www.fs.fed.us/emc/ [accessed 31 March 2012]. VER HOEF, J. M., D. C. GLENN-LEWIN, AND M. J. A. WERGER. 1989. Relationship between horizontal pattern and vertical structure in a chalk grassland. Vegetatio 83:147-155. WAGNER, H. H. 2003. Spatial covariance in plant communities: integrating ordination, geostatistics, and variance testing. Ecology 84:1045—1057. WEIXELMAN, D. A., D. C. ZAMUDIO, AND K. A. ZAMUDIO. 1996. Central Nevada riparian field guide. USDA, Forest Service, Intermountain Re- gion, Ogden, UT. WHITTAKER, R. H. 1960. Vegetation of the Siskiyou mountains, Oregon and California. Ecological Monographs 30:279-338. MADRONO, Vol. 59, No. 3, pp. 150—155, 2012 REAPPEARANCE OF THE VANISHING WILD BUCKWHEAT: A STATUS REVIEW OF ERIOGONUM EVANIDUM (POLYGONACEAE) NAOMI S. FRAGA, ELIZABETH KEMPTON, LEROY GROSS AND DUNCAN BELL Rancho Santa Ana Botanic Garden, 1500 North College Avenue, Claremont, CA 91711 nfraga@rsabg.org ABSTRACT Eriogonum evanidum Reveal is an annual herb in the buckwheat family that is endemic to southern California, USA and northern Baja California, Mexico. It was described by James Reveal in 2004 and was determined to be extinct due to the lack of recent observations and accessioned specimen collections. During field surveys conducted in 2007 and 2008 this species was rediscovered across its known range in southern California. Three of 10 historic occurrences that were presumed extirpated were located. One new occurrence was documented in the vicinity of Holcomb Valley, San Bernardino Mountains, CA. Location and habitat information are provided and the current conservation status of this species is discussed. Key Words: California, conservation, Eriogonum, endemic, extinct, Polygonaceae, rare, rediscovery. Eriogonum evanidum Reveal (vanishing wild buckwheat) is an annual herb in the buckwheat family (Polygonaceae) that was recently described (Reveal 2004; Fig. 1). This species was given the common name ‘vanishing wild buckwheat’ be- cause it was presumed extinct due to the lack of recent observations and herbarium specimen records (Reveal 2005; CNPS 2011). Focused surveys were conducted by botanists at Rancho Santa Ana Botanic Garden (RSABG) in late summer and fall of 2007 and 2008 to locate E. evanidum across its known range in southern California, USA. Eriogonum evanidum was found to be extant across its historic range within the United States, although the status of this species in Baja California, Mexico is unknown. Although several occurrences were located, it is likely that some occurrences have been extirpated (Fig. 2). Details of the 2007 and 2008 field surveys are discussed here, including additional information on distribution, habitat, and conservation status. TAXONOMIC AND COLLECTION HISTORY Eriogonum evanidum was discovered during herbarium studies by Reveal (2004). The new species was described from specimens that were previously identified as Eriogonum foliosum S. Watson. In his description, Reveal (2004) stated that “this distinctive species has been hidden quietly under Eriogonum foliosum S. Watson awaiting rediscovery of an extant population so that it might be more precisely characterized than possible from dried material.” Reveal designated a holotype that was collected in 1902 (Abrams 2894) and paratypes collected from as early as 1893 (Alderson 399) to 1967 (Ziegler s.n.). Based on accessioned herbarium specimen records examined by Reveal (2004), it appeared that most of the documentation for E. evanidum took place between 1920 and 1940, primarily from the vicinity of Big Bear Valley in the San Bernardino Mountains of southern California (Reveal 2004; Table 1); an area that has experienced substantial development over the last century. Eriogonum evanidum was “‘presumed extinct” in the Flora of North America (Reveal 2005) and considered “possibly extirpated” in the second edition of the Jepson Manual (Baldwin et al. 2012), although Reveal has reported seeing specimens of the E. evanidum that were collected in the late 1990’s (Costea and Reveal 2011). Two specimens that were not cited in the original description, and therefore not likely viewed prior to its descrip- tion, are housed at RSA and were collected in 1976 (Davidson 4471) and 1994 (Hirshberg s.n.). FIELD SURVEY METHODS Surveys for E. evanidum were conducted by botanists at RSABG using a focused, or intuitive- controlled, field survey method (USDA FS 2005), which targets habitats with the highest potential for locating target species at the appropriate time for proper identification. Herbarium specimen records (RSA, UCR), databases (CalFlora, Con- sortium of California Herbaria, and California Natural Diversity Database), and literature reports were used to identify historic populations of E. evanidum and these were targeted for field surveys. Surveys by botanists at RSABG were conducted during August 2007, and August and September of 2008. U.S. Department of Agricul- ture (USDA) Forest Service Element Occurrence forms were used to document all populations of E. evanidum that were encountered (e.g., exact location, population status, existing or potential threats or disturbances, habitat, associated 2012] ah & ¢ * * * “S < : P a. 5 a Fic. 1. Eriogonum evanidum in flower with a fisher space pen (9.6 cm X 0.8 cm) for scale. species). Herbarium specimens were collected and deposited at the RSA herbarium and _ photo- graphs of plants were contributed to the CalPho- tos database (calphotos.berkeley.edu). USDA Forest Service Element Occurrence forms, along with U.S. Geologic Survey (USGS) maps of surveyed populations, were submitted to the California Department of Fish and Game (CDFG), and USDA Forest Service. Plant identifications were made using taxonomic keys and descriptions in The Jepson manual: higher plants of California (Hickman 1993), A flora of southern California (Munz 1974), and Flora of North America (Reveal 2005). Identifications were verified through comparison with annotated specimens in the RSA herbarium. DISTRIBUTION In the United States, E. evanidum is restricted to southern California and has been documented in the San Bernardino Mountains in San Bernardino County, San Jacinto Mountains in Riverside County, and the Laguna Mountains in San Diego County (Fig. 2). Eriogonum evanidum has also been reported from northern Baja California, Mexico (Costea and Reveal 2011). FRAGA ET AL.: STATUS REVIEW OF ERIOGONUM EVANIDUM 151 Occurrences previously documented in the vicin- ity of Big Bear Lake and Baldwin Lake in the San Bernardino Mountains were not found in 2007 or 2008. It is possible that these occurrences have been extirpated due to development in the region (lable 1): There are three occurrences in southern Cali- fornia that are doubtful and not supported by herbarium specimens (Table 1). These include reports from Valencia in Los Angeles County, Warner Springs in Riverside County, and Viejas Mountain in San Diego County (CNPS 2011: CNDDB 2011). The only source for the reported occurrence at Valencia in Los Angeles County is a report prepared by Dudeck and Associates (unpublished) to the CDFG and Newhall Land and Farming Company (CalFlora 2011; CDFG 2011). The population at Warner Springs was referenced by Reveal (1989) in his treatment of EF. foliosum. The source for the occurrences at Viejas Mountain is from Craig Reiser’s 1994 account in Rare Plants of San Diego County (CNDDB 2011). None of these occurrences has been verified in the field, and none are documented by herbarium specimens. In addition, all three are reported at lower elevations than vouchered occurrences and therefore they may lack suitable habitat (Table 1). Future surveys should be conducted to verify if E. evanidum occurs at these locations. HABITAT In previous floristic treatments, habitat for E. evanidum was described as sandy to gravelly flats and slopes, in sagebrush communities, oak woodland and montane conifer woodlands at 1150-2300 meters in elevation (Reveal 2005; CNPS 2011; Costea and Reveal 2011; Fig. 3). The described habitat requirements were sub- stantiated while conducting surveys, except that plants were not found in oak woodland. Oak woodland vegetation was presumably included in the habitat description because of the previous reference of this species at Warner Springs in Riverside County. Occupied habitat for E. evanidum included sandy soils derived from decomposed granite, in primarily full sun with little to no leaf litter, on a flat aspect in sagebrush scrub dominated by Artemisia tridentata Nutt. The sagebrush community was often surrounded by coniferous forest dominated by Pinus jeffreyi Balf. or P. ponderosa P. Lawson & C. Lawson. Eriogonum evanidum was also observed to occur in dry meadows dominated by Artemisia triden- tata in Holcomb Valley, San Bernardino Moun- tains, CA. The following is a list of species associated within E. evanidum compiled from several loca- tions throughout its range (* denotes not native species): Artemisia tridentata, Astragalus douglasti 118°0'0"W 117°0'0"W ‘oY NT: ; pies io ie: # Los Angeles Riversic oc ee m4 Beach COS SS Ke : : > & ORANG! = . : : oe, COUNTYSY, RIVERSIDE ? Legend Eriogonum evanidum @ Historical Occurrences Z\ 2007-2008 Survey Occurrences 118°0'O"W 117°0'0"W FIG. 2. MADRONO Distribution of Eriogonum evanidum based on vouchered occurrences. [Vol. 59 116°0'0"W NEVADA CALIFORNIA Las Vegas poplin Los Angeles LJ ARIZONA Phe San Die BAJA CALIFORNIA IMPERIAL COUNTY 116°0'0"W Historic occurrences are represented by a black circle and occurrences located in 2007 and 2008 are represented by a white triangle. See Table | for detailed information (source herbaria, collection date, locality, etc.) for each occurrence. (Torr. & A. Gray) A. Gray var. parishii (A. Gray) M. E. Jones, *Bromus tectorum L., Castilleja cinerea A. Gray, Chenopodium leptophyllum (Moq.) Nutt. ex S. Watson, Chrysothamnus viscidiflorus (Hook.) Nutt., Cryptantha micrantha (Torr.) I. M. Johnst., Gutierrezia sarothrae (Pursh) Britton & Rusby, Ericameria pinifolia (A. Gray) H. M. Hall, Eriogonum baileyi S. Watson, E. davidsonii Greene, E. wrightii Torr. ex Benth. var. subscaposum S. Watson, Eriastrum sapphirinum (Eastw.) H. Mason, Lessingia glan- dulifera A. Gray, Nicotiana attenuata Steud., Penstemon centranthifolius (Benth.) Benth., Pinus Jeffreyvi, P. ponderosa, *Sisymbrium altisimum L., Stephanomeria exigua Nutt., and Trichostema micranthum A. Gray. IDENTIFICATION Eriogonum evanidum co-occurs with other annual taxa in the genus Eriogonum that it may be confused with including: E. baileyi, E. davidsonii, and E. gracile Benth. Eriogonum evanidum can be distinguished from these species by its relatively small flowers (0.8—1.2 mm), outer perianth lobes that are more or less hastate in fruit, and stems that are tomentose. In contrast E. baileyi has flowers that are 1.5—3 mm, the outer perianth lobes oblong to oblong-obovate, gener- ally constricted near middle, and stems are glabrous or tomentose. Eriogonum davidsonii has flowers that are 1.5—-2 mm, the lobes are oblong-obovate, and stems are glabrous. Eriogo- num gracile has flowers that are 1.5—3 mm, the lobes are lanceolate to oblong, and stems are generally tomentose, or sometimes glabrous (Baldwin et al. 2012). Although not co-occurring within the United States, there are two additional species (E. foliosum and E. hastatum Wiggins; both endemic to Mexico) that may be confused with E. evanidum in Baja California, Mexico. Eriogonum evanidum can be distinguished from both of these species on the basis of several morphological characters—both E. foliosum and E. hastatum have sprawling habits, foliaceous inflorescence bracts, and elliptic basal leaves while E. evanidum has an erect habit, scalelike inflorescence bracts, and broadly ovate to orbicular or reniform basal leaves. CONSERVATION STATUS Eriogonum evanidum is not listed by the State of California or Federal government as threat- ened, or endangered, but is considered by the California Native Plant Society as “seriously endangered in California” (California Rare Plant Rank 1B.1; CNPS 2011). Eriogonum evanidum 1s also on the Sensitive species list for the Cleveland and San Bernardino National Forests. Presum- ably several populations have been extirpated in FRAGA ET AL.: STATUS REVIEW OF ERIOGONUM EVANIDUM 153 2012] _oOoO PoYLIoA JON BIOUgIeA ‘UIs epeU OSE VI 9007-das- | OAdO oyey uIMpleg G€dGdAN) Ut jou “gq pure oye] Ivog sig useMjoq WAS €LOT-8S0T dads 1¢61-das-L SOOSTIMON yuv}xo A[qissod ING “BOOT Ul pareooyar 6L6S9VSU 10U “GQCAND Ur 10N ‘oyeyT urlmpreg ‘Was €LOT dds 6C61-SNW-€T “‘6€TEGLINOd “ILLTLINOd ‘punoisdured AaeA quiosjoH] 9SOSELVSU ‘SOPSELVSU jo {J ‘MOpeoU YaaID Noquey WS Q9ETT-STCT ads 80-8NV-07 “SOOSELVSU ‘O99SELVSU ‘ge ABMYSIF] WO Tu ¢ INOqY ‘yIoID noqiuedg suojly ‘uoAuey uasnd uvA Jo uoToas 1addq WAS TCT adds LO-3NV-rI CTIEGEVSU “EPOETLVSU ‘(SsulpueT yusuIdojaasq “A sApun]) aye] Ivog sig aAoqe 1eLy ‘WS OOIZ ads I€6[-sny-¢ EISTOSON “ETSLSSdaLr “AQTIVA vag ‘qn[D pur[poom yuswIdojaasq “A ued Jo}ag JO YNOS UvaNs SUCTY WAS PeIT das 6761-190-S TOPE6LINOd £66SS INOd juawidojaaaq “q ‘AQIVA Ivog ‘surejuno|w gs OOIT dads O61 -8NW-¢ “PEETINOd “699T6EVSU (P661 JostIoy) POYLIOA JON ouldiy SPs ds ANON OsaIq URS Jo sjUL[g Wey TL AeMYsIpy Ol 9SO]O pue ‘JO opIs N ‘AQTLA JOUIeH surejuNoyy O}UIOKLS URS TOVI-S9ET AIT 8o-das-9] TIOSELVSU “P79061 VSU POLLISA JON ssulidg JOY Ioure Ay CL6 ds ANON POOT [P2ATY (1107 HOD) yuouIdoyaAap suIsSnoy LOLL6SVSU 0} onp payedinxs sem ‘TSOPIICS “99€C6EVSU 99U9I1INIIO SIYy} JO UonAod VW ‘ATTRA WIg JO pud N sureyunoyy vunsey] O9LT ds 80-8sNV-SZ ‘ISTIPLVSU ‘1997T7AS SOION, UOI]BVIOT (UI) “ATA ree) P2}S9][OD Ise] 91eq dDINOS a a ee eee ee ‘poyedinxg Al[qissog = Y ‘sureyuno|] OUIpsIeuslog ueg = WAS ‘suoNneIAsIqqe J9yIO ‘OUIpIeUlog UPS = GS ‘OpIsioATyY = ATY ‘OSsIq, ueg = Cg ‘suoneiAsiqqy AluNOD ‘auIKeH pure ysi{ Jo Juounredad eviuslojyeg = O4dD ‘asaTJoD euowo0g = WOd ‘Ag[ayYIIG PIUIOJ][VD JO AYSIOAIUF, = DP ‘uosdar = Sgaf ‘UspIeH o1uejog euy eluRs OYysUR Yy = VSU “Whneasnyy A1OIsTF{ [RINJVN OSIIG| URS = CF ‘sWIAUOIOR UMTIeQIdF{ “SALVLS CaLIN() dHL NI WAGINVAT WANODOIYY AOA SNOILVOOT GALYOdAY ‘[ ATIVE 154 se se st af wer Pe ats ‘ Fic. 3. the vicinity of Big Bear Lake and _ possibly Baldwin Lake due to development in the region. In addition, one population in Pine Valley has been extirpated due to development; Hirshberg revisited a population she previously documented in 1994 and states that ‘“‘there is now a house there, and the population has apparently been extirpated”’ (CCH 2011). The primary threats to this species include development, dispersed rec- reation (vehicle use off designated roads, hiking, equestrian use, etc.), and non-native plant estab- lishment (CNPS 2011). SURVEY RESULTS There are ten documented locations of E. evanidum (Fig. 2); four of these are known to be extant (Table 1). Several occurrences in the vicinity of Big Bear Lake and Baldwin Lake were not located and are possibly extirpated (Table 1). One historic occurrence in the San Jacinto Mountains in Garner Valley (Hemet Valley), one historic occurrence in the Laguna Mountains in Pine Valley, and one occurrence in the San Bernardino Mountains along Van Dusen Canyon Road were located. An occurrence in the vicinity Holcomb Valley in the San Bernardino Mountains was newly documented as a result of this study (Table 1). There are several report- ed locations for which there are no voucher MADRONO Typical habitat for Eriogonum evanidum. Photo from Pine Valley, San Diego County, CA. specimens (Valencia, Warner Springs, and A\I- pine, Table 1), therefore the existence of these occurrences is suspect until verified. In 2008 a survey was conducted at Warner Springs and no suitable habitat found. DISCUSSION There are several factors that may have influenced the lack of recent documentation for this species and previous conclusion that this species was extinct. First, E. evanidum has a highly limited distribution, only occurring in localized microsites within its distribution, and is generally not locally common. Second, this species com- monly occurs in the general vicinity of similar looking species (e.g., E. baileyi, E. davidsonii). Third, E. evanidum has exceedingly small flowers and therefore diagnostic characteristics can be difficult to detect in the field. Lastly, E. evanidum flowers from August to September, a time of year when few botanists make collections. Other summer to fall-blooming plants, such as Deinan- dra mohavensis (D. D. Keck) B. G. Baldwin have been mistakenly considered extinct (Sanders et al. 1998). This brings to bear the importance of collecting plant specimens late in the growing season in the summer and fall months. There are undoubtedly additional occurrences of EL evanidum that have not been documented ( i 1 | 2012] and we recommend additional surveys are conducted to further assess the status of this species. The status of CNPS list 1B.1 is ap- propriate given the current information. All extant occurrences are known from National Forest lands, and are therefore currently protected from development; however, this species may also be present in undocumented locations on private property. While E. evani- dum is extant throughout its current range with- in the U.S., this species has been impacted by anthropogenic disturbances (e.g., dispersed rec- reation, OHV use, development). Specifically plants in Pine Valley in San Diego County have been impacted by development. This was noted by Hirshberg (CCH 2011) who was unable to locate an occurrence she previously document- ed because it has been extirpated by develop- ment (CCH 2011). The conservation status of E. evanidum was brought to light because it was presumed extinct in recent floristic treatments; if not for these recent publications, this species could have remained undetected and hidden in herbaria. ACKNOWLEDGMENTS This project was supported by the USDA Forest Service (San Bernardino and Cleveland National Forests). We wish to thank Scott Eliason, Melody Lardner, and Lisa Young for logistical support. LITERATURE CITED BALDWIN, B. G., D. H. GOLDMAN, D. J. KEIL, R. PATTERSON, T. J. ROSATTI, AND D. H. WILKEN, (eds.) 2012. The Jepson manual: vascular plants of California, 2nd ed. University of California Press, Berkeley, CA. CALIFORNIA DEPARTMENT OF FISH AND GAME (CDFG). 2011, Biological resources technical report for the Entrada Site, Los Angeles, County, California. Website: http://nrm.dfg.ca.gov/FileHandler. ashx?DocumentVersionID=21408 [accessed December 1, 2011]. CALIFORNIA NATIVE PLANT SOCIETY (CNPS). 2011, Inventory of rare and endangered plants (online edition, v7-10c). California Native Plant Society, FRAGA ET AL.: STATUS REVIEW OF ERIOGONUM EVANIDUM [55 Sacramento, CA. Website: http://www.cnps.org/ cnps/rareplants/inventory/ [accessed November 11, 2011]. CALIFORNIA NATURAL’ DIVERSITY DATABASE (CNDDB). 2011, RareFind 4.0. California Depart- ment of Fish and Game, Sacramento, CA. Website: https://nrmsecure.dfg.ca.gov/cnddb/view/query. aspx [accessed November 11, 2011]. CALFLORA: INFORMATION ON CALIFORNIA PLANTS FOR EDUCATION, RESEARCH AND CONSERVATION [WEB APPLICATION]. 2011, The Calflora Database [a non-profit organization], Berkeley, CA. Website: http://www.calflora.org/ [accessed: November 11, 2011]. CONSORTIUM OF CALIFORNIA HERBARIA (CCH). 2011, Consortium of California Herbaria Search Page. Consortium of California Herbaria, Berkeley, CA. Website: http://ucjeps.berkeley.edu/consortium/ [accessed November 11, 2011]. COSTEA, M. AND J. L. REVEAL. 2011. Eriogonum. In The Jepson eFlora. Website: http://ucjeps.berkeley. edu/IJM.html [accessed November 11, 2011]. HICKMAN, J. C. (ed.). 1993. The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. Munz, P. A. 1974. A flora of southern California. University of California Press, Berkeley and Los Angeles, CA. REVEAL, J. L. 1989. The eriogonoid flora of California (Polygonaceae: Eriogonoideae). Phytologia 66:295— 414. —. 2004. New entities in Eriogonum (Polygona- ceae: Eriogonoideae). Phytologia 86:121—159. . 2005. Eriogonum. Pp. 221-430 in Flora of North America Editorial Committee (eds.), Flora of North America North of Mexico, Vol. 5: Magnoliophyta: Caryophyllidae, part 2. Oxford University Press, New York, NY. SANDERS, A. C., D. BANKS, AND S. Boyb. 1998. Rediscovery of Hemizonia mohavensis Keck (Aster- aceae) and addition of two new localities. Madrono 44:197—200. UNITED STATES DEPARTMENT OF AGRICULTURE, FOREST SERVICE (USDA FS). 2005, Threatened, endangered and sensitive plants survey: field guide. Washington, DC. Website: http://fs.fed.us/r6/ sfpnw/issssp/documents/inventories/inv-sp-tesp- survey-field-guide-2005-03.doc [accessed December 1, 2011]. MADRONO, Vol. 59, No. 3, pp. 156-162, 2012 PTYCHOSTOMUM PACIFICUM (BRYACEAE), A NEW FEN SPECIES FROM CALIFORNIA, OREGON, AND WESTERN NEVADA, USA JOHN R. SPENCE Glen Canyon National Recreation Area, P.O. Box 1507, Page, AZ 86040-1507 JAMES R. SHEVOCK California Academy of Sciences, Department of Botany, 55 Music Concourse Dr., Golden Gate Park, San Francisco, CA 94118-4503 jshevock@calacademy.org ABSTRACT Ptychostomum pacificum J. R. Spence & Shevock, a new and highly distinctive species restricted to fen habitats within coniferous forests in California, Oregon, and extreme western Nevada is described and illustrated. This species appears to be related to P. turbinatum (Hedw.) J. R. Spence but is easily distinguished by a combination of features including large size, percurrent yellowish-brown costa with numerous incrassate cells at the leaf tip, strongly recurved leaf margins, and elongate cylindrical to narrowly pyriform capsules. Key Words: Bryum, cascades, mosses, Sierra Nevada. The genus Bryum sensu lato with nearly 450 currently recognized species (Crosby et al. 2000) remains a complex and taxonomically challeng- ing group of mosses. However, this polyphyletic genus has recently been divided into several segregate genera (Spence 2005, 2007; Spence and Ramsay 2005; Holyoak and Pedersen 2007). The Bryaceae are a major component of the bryoflora along the Pacific slope of North America, and California, with 60 taxa, leads all other states and provinces of North America in the number of species in this family (Norris and Shevock 2004; Malcolm et al. 2009; Spence unpublished data). Historically, bryologists generally avoid collecting Bryum if sporophytes are lacking since most keys have relied heavily on sporophytic characters. This collecting approach has greatly hindered work on this group of mosses since many useful gametophytic characters are currently recognized to make identification of sterile collections possible. During the course of developing the treatment of the Bryaceae for the Flora of North America Project (FNA Vol. 28 in prep.), the first author determined that several additional taxa remain undescribed, and many of these new taxa reside in California. This paper is an effort to reduce this backlog of species awaiting formal publication. This new species is described in the genus Ptychostomum Hornsch., which has been shown to be distinct from Bryum sensu stricto (B. argenteum Hedw. and its allies) by both molec- ular and morphological studies (Pedersen et al. 2003; Spence 2005; Holyoak and Pedersen 2007). Most of our familiar boreal-temperate species of Bryum actually belong in Ptychostomum, and have been transferred elsewhere for the Bryophyte Flora of North America (Spence 2005, 2007). Based on the herbarium record examined to date (CAS, UC), ‘“P. pacificum”’ was first collected by the second author in 1975. This collection was sent to Dan Norris (UC) who at the time at Humboldt State University was working toward a California moss flora. The specimen was returned as “Bryum sp. possibly undescribed.’ Many collections of this fen species were subsequently obtained by the second author but sporophytes remained unknown. In 1999, a population of “‘P. pacificum”’ was discovered by David Toren (CAS) in Lake County, California that contained a few sporophytes, and a couple years later the second author found sporophytes in a Sierra Nevada occurrence. Of the nearly 75 occurrences documented by herbarium vouchers, sporophytes have only been documented seven times. Interestingly, Andrews (1935) describes “sterile forms” of Ptychostomum turbinatum (Hedw.) J. R. Spence from the west (without specific localities) with stem lengths reaching 10 cm or more, which suggests that at least some prior collections of ‘‘P. pacificum”’ may exist in other herbaria and are likely to be labeled as P. turbinatum, a much smaller species. TAXONOMIC TREATMENT Ptychostomum pacificum J. R. Spence & Shevock, sp. nov. (Figs. 1, 2). —TYPE: USA, Califor- nia, Fresno Co., Sierra National Forest, Highway 168 above Huntington Lake, 37°12'59.7"N, 119°11'30.4"W, 7375 ft., 2 Sep 2002, Shevock & Ertter 22887 (holotype: CAS; isotypes H, KRAM, MO, NY). 2012] SPENCE AND SHEVOCK: A NEW SPECIES OF PTYCHOSTOMUM 157 Lee { Po Caclo=GD CP AVR? Ge aa =< Gay Se =e BZ 5 = ae | = = = eS RST Be SEZ a — Z4 CES CO ) 2= PIE pearance PEPSI TE eas ae C2 25% (>) cae INS < t= C5) C= : a e \ Ny SS VAG AAS Saka Se ers La CFS CS ) Fic. 2. Ptychostomum pacificum J. R. Spence & Shevock. 1. Male plant. 2-6. Perigonial bracts, sequentially from outermost to innermost. 7—13. Perichaetial leaves, sequentially from outermost to innermost. 14. Female plant with mature capsule, wet. 15-16. Operculate capsules, dry. 17. Deoperculate capsule, wet. 18. Exothecial cells at base of urn and stomata. 19. Mid-urn exothecial cells. 20. Exothecial cells at orifice, portion of peristome and spores. (1-6 | from Shevock 14233; 7-20 from Toren 9586 B; both in KRAM). Scale bars: a — 0.5 cm (2-13); b — 0.5 cm (14); ¢ — 0.5 cm (1); d — 1 mm (15-17); e — 100 um (18-20). 2012] Plantae robustae, caulibus usque 12 cm; mar- gines folii proximales recurvae, limbidio pervaldo, ex 2-4— stratis composito, partim bistratoso, ad apicem fol cellulis distalibus laminalibus incras- satis, saepe rubro-brunneis, costa apicem folii non attingenti vel percurrenti; capsula elongata, cylin- drica vel clavata, 3-5 mm, brunnea, symmetrica. Plants medium to robust, in dense or open turfs, green or yellow-green. Stems 2-12 cm, red-brown, lacking central strand, evenly foliate to somewhat crowded distally, often strongly radiculose. Leaves narrowly to broadly ovate- lanceolate to ovate, 3-5 mm, somewhat enlarg- ed towards stem apex, weakly concave distally, leaves below becoming strongly concave and sometimes cucullate, weakly to moderately keeled, yellow-green to bright green, becoming blackish to red-brown when old, strongly con- torted when dry, distally spirally twisted, erect- spreading to spreading wet, not or weakly decurrent; apex acute to acuminate, often colored golden-brown, margins revolute to mid-leaf, smooth to weakly serrulate distally, costa brown to red-brown, strong, not reaching apex to percurrent, rarely very short excurrent in stout serrulate awn, limbidium extremely strong, 2—4 rows, yellowish, partially bistratose below, distal and median laminal cells rhomboidal to hexago- nal, 15—25 um wide, 30—70 um long, mostly 2—3:1, thin walled except colored distal cells which are strongly incrassate, proximal laminal cells grad- ually rectangular, about the same width or somewhat narrower than cells above, some- what longer than cells above, 15—20 (25) um wide, 40-90 um long, often reddish, alar cells not differentiated. Specialized asexual reproduction absent. Sexuality dioecious; outer perichaetial leaves similar to vegetative leaves, inner leaves smaller, more triangular, with more acuminate apex; perigonium distinctly enlarged and conspic- uous, Outer leaves similar to vegetative leaves, inner leaves broadly ovate and abruptly acute. Seta 2-4 cm, straight, red-brown to yellow-brown, slender. Capsule elongate cylindric to clavate, 3— 5 mm, brown, symmetric, mouth brown to red- brown, somewhat constricted when mature below mouth, exothecial cells irregularly elongate, 15— 25 um wide, 30—60 um long, mostly 4:1, shorter in 2—3 rows at mouth; peristome well developed, exostome teeth yellow, lanceolate, SOO—-650 um long, strongly trabeculate; endostome membrane high, '% height of exostome, pale hyaline to yellowish, processes well developed, perforations oval, smooth, cilia somewhat variable, 2—3, some- times short, mostly appendiculate; operculum low convex, apiculate. Spores pale yellow or pale brown, finely papillose, 12-16 um. Paratypes: USA. CALIFORNIA. Alpine Co.: Stanislaus National Forest, McCormick Creek drainage, S base of The Dardanelles, 7830 ft, 6 Aug 2007, Willits 500 (CAS). El Dorado Co.: SPENCE AND SHEVOCK: A NEW SPECIES OF PTYCHOSTOMUM Lo 1.5 mi N of Grass Valley near Tahoe Rim Trail, Nichols 221 (USFS-LTBMU). Fresno Co.: Kings Canyon National Park, Pacific Crest Trail, Middle Fork Kings River, Le Conte Canyon between Ranger Station and Grouse Meadow, 8400 ft, 8 Sep 1999, Shevock & Haultain 18634 (CAS, NY); Sierra National Forest, Kaiser Wilderness, tributary of Rancheria Creek, 9240 ft, 16 Jul 2001, Shevock 20994 (CAS, H, KRAM): E of Idaho Lake, 8875 ft, 16 Jul 2001, Shevock 21004 (CAS, UBC); forest road 10S66 above Bear Creek Rd., 8300 ft, 27 Jul 1996, Shevock & Bourell 13994 (CAS, KRAM): West Snow Corral Meadow, 7100 ft., 27 Jul 1996, Shevock & Bourell 14022 (CAS, DUKE, F, NY); between House and Ahart Meadows, 7200 ft, 28 Jul 1996, Shevock & Bourell 14040 (CAS, KRAM, UC); John Muir Wilderness, Wet Meadows, between Rancheria and Crown Valley, 8550 ft, 16 Aug 1996, Shevock & York 14117 (CAS, DUKE, MO, NY, OSC) & /4//8 (CAS, H, KRAM, UBC, US); near Spanish Lake, 8600 ft, 17 Aug 1996, Shevock 14165 (CAS, COLO, H, UC); Hoffman Meadow, 6800 ft, 30 Jul 2000, Shevock & Norris 19856 (CAS, US). Humboldt Co.: Shasta-Trinity National Forest, Headwaters of Kerlin Creek about 5.4 air mi WSW of Hyampom, 5460 ft, 23 Jul 2009, Lenz 226 (CAS). Glenn Co.: Mendocino National Forest, Telephone Camp W of Plaskett Mead- ows, 6900 ft, 18 Jul 1998, Toren 7224 (CAS, MO, NY). Lake Co.: Mendocino National Forest, Mt. Sanhedrin, headwaters of Mill Creek, 5450 ft, 17 Oct 1999, Toren, Dearing & Heise 7633 (CAS); Cirque of N slope of Hull Mountain, 6300 ft, 6 Sep 1999, Toren 7620 (CAS, NY); Snow Moun- tain Wilderness, Milk Ranch Meadow, 6300 ft, 25 Jun 2002, Toren 9151 (CAS) and East Peak, 6400 ft, 26 Jun 2002, Toren 9167 (CAS). Lassen Co.: N slope of Dyer Mtn., 3.4 air mi SSW of Westwood, 5900 ft, 22 Jun 2005, Lenz 1835 & 1837 (CAS). Madera Co.: Devils Postpile Na- tional Monument, meadow N of Rainbow Falls, 7450 ft, 25 Sep 2001, Shevock & Dulen 21310 (CAS); Sierra National Forest, 1.2 mi NW of Beasore Road at Cold Saddle, 7250 ft, 24 Sep 2001, Shevock 21231 (CAS, H, KRAM): near Chipmunk Meadow, 6850 ft, 11 Jul 2003, Shevock 24166 (BOL, CAS, KRAM, NY): W of Grey Mountain, 6800 ft, 19 Jun 2002, Laeger 1466 (CAS); NE of Fresno Dome, 7450 ft, 6 Jul 2000, Shevock & Kellman 19622 (CAS, DUKE): Central Camp Rd near Gaggs Camp, 5875 ft, 11 Jul 2003, Shevock 24156 (CAS, DUKE, F); Poison Meadow, 6850 ft, 18 Jun 2002, Shevock, Norris, & Clines 22420 (CAS, UC); China Meadow off of FS road 8S70, 6000 ft, Shevock & Norris 20241 (CAS); Yosemite National Park, between Grouse and Crescent Lake, 8300 ft, 29 Jul 2009, Shevock & Smith 33206 (CAS, MO, NY, UC). Mariposa Co.: Yosemite National 160 MADRONO Park, Turner Meadow, 7365 ft, 28 Jul 2009, Shevock, Taylor, Smith, & Colwell 33181 (CAS, MO, UC); McGurk Meadow, 6900 ft, 24 Sep 2001, Shevock 21232 (CAS, YM); Mono Mead- ows, 6940 ft, 24 Sep 2001, Shevock 21239 (CAS, MO, YM) and 11 Jul 2009, Shevock & Hutten 33174 (CAS). Modoc Co.: Modoc National Forest, Headwaters of Lassen Creek, 5.8 air mi SE of Sugar Hill, 6550 ft, 23 Oct 2007, Lenz 4097 (CAS). Nevada Co.: Tahoe National Forest, Pat Yore Flat, 6165 ft, 22 Jul 2010, Wishner 1029] (CAS, UC). Placer Co.: Tahoe National Forest, Duncan Fen, 6710 ft, 17 Oct 2008, Wishner 9721 (CAS, UC); Gates Fen, 5435 ft, 17 Oct 2008, Wishner 9731 (CAS, UC); Tadpole Creek Fen, 6260 ft, 17 Oct 2008, Wishner 9711 (CAS, UC). Plumas Co.: Plumas National Forest, Daly Cow Camp Fen, head of Big Pine Ravine, 1 km NNE of Camel Peak, 1591 m, 8 Nov 2006, Toren & Janeway 9505 (CAS, MO, UC) and Janeway & Toren 8952 (CHSC); Vaca Fen near head of S Branch Middle Fork Feather River, 1743 m, 8 Nov 2006, Toren & Janeway 9507 (CAS, NY, UBC, UC) and 10 Jun 2009 Dillingham & Toren 2572 (CAS); above Black Rock Campground W of Little Grass Valley Reservoir, 5150 ft, 15 Jun 2007, Toren 9586B (CAS, KRAM, UC); S Branch of Ward Creek, 6670 ft, 7 July 2005, Dillingham & Toren 2057 (CAS, UC); Bucks Summit, 5360 ft, 14 Nov 2003, Dillingham et al. 1123 (CAS); 3 mi W of Bucks Lake, 5255 ft, 13 Nov 2003, Dillingham 1048 (CAS, UC); China Gulch Fen, 5410 ft, 10 Oct 2003, Dillingham & Toren 1036 (CAS, UC); SE of Red Mountain, 5200 ft, 13 Nov 2003, Dillingham & Toren 1187 (CAS); E of Tamarack Flat, 5400 ft, 13 Sep 2001, Dillingham & Norris 404 (CAS); 7.5 mi NE of Quincy, 6285 ft, 23 Aug 2004, Dillingham 1654 (CAS, UC); tributary to Rabbit Creek, 5300 ft., 16 Jun 2005, Dillingham 2042 (CAS). San Bernardino Co.: San Bernardino National Forest, Champion Lodgepole Meadow about 0.5 mi. from Siberia Creek trailhead, 7475 ft., 26 Sep 2011, Eliason & Williams SEO9F&G (CAS). Shasta Co.: Shasta-Trinity National Forest, Clark Creek ca. 2 mi E of Red Mountain, Norris, Lenz, & Hillyard 108204 (CAS, UC); Lassen Volcanic National Park, Dersch Meadows, 6590 ft, 22 Aug 2008, Shevock & Showers 31901 (CAS, NY). Sierra Co.: Tahoe National Forest: Sierra Buttes Fen, 7250 ft, 17 Aug 2009, Wishner 12013 (CAS, UC). Tehama Co.: Lassen Volcanic National Park, N base of Mt. Conard, 7175 ft, 22 Aug 2008, Shevock & Showers 31925 (CAS, MO, NY, UC, US). Trinity Co.: Shasta-Trinity Na- tional Forest, Headwaters of West Branch Crow Creek, 0.5 air mi SSW of Mumbo Lake, 6260 ft, 21 Jul 2009, Lenz 4201 (CAS). Tulare Co.: Sequoia National Forest, Headwaters of Free- man Creek, 7000 ft, 19 Jul 1975, Shevock 4599 (CAS) and /0638 (CAS); Freeman Creek Trail, [Vol. 59 7000 ft, 24 Aug 1996, Shevock 14233 (CAS, KRAM, UC); Quaking Aspen Campground, 7000 ft, 15 Jun 2001, Shevock 20954 (CAS); Cold Spring below Portuguese Pass, 7200 ft, 2 Sep 1996, Shevock 14314 (CAS, KRAM); Clicks Creek, 7900 ft, 24 Aug 1996, Shevock 14252 (CAS, MO, NY); Golden Trout Wilderness, Jacobsen Meadow, 8400 ft, 23 Oct 2000, Laeger 394 (CAS), between Redwood Crossing and Long Meadow, 6800 ft, 27 Jul 1983, Shevock 10604 (CAS, UC), and 9 Nov 1997, Shevock 16713 (CAS, MO, NY, UC); Sirretta Meadows, Ernest C. Twisselmann Botanical Area, 8800 ft, 16 Aug 1998, Shevock 17522 (CAS) and stringer of Sirretta Meadows, 9000 ft, 26 Aug 2006, Laeger & Cone 3586 (CAS); Machine Creek E of Round Meadow, 9060 ft, 29 Sep 2001, Shevock, et al. 21380 (CAS, MO, NY, UBC, US); Round Meadow, 9000 ft, 2 Sep 1996, Shevock 14320 (CAS, UC); Mosquito Meadow, 8800 ft,10 Jul 1999, Shevock 18416 (CAS, CONC, MO); Woodcock Meadow near Buena Vista, 7300 ft, 25 Oct 1997, Shevock 16659 (CAS, H, MO, NY); Weston Meadow, 6800 ft, 28 Jun 1996, Shevock & York 13650 (CAS, KRAM); Jennie Lakes Wil- derness, between Marvin Pass and Mitchell Peak, 9300 ft, 11 Oct 1996, Shevock 14553 (CAS, MO, NY) and Rowell Meadows, 8850 ft, 2 Aug 2002, Laeger & Hayden 1611 (CAS); Sequoia National Park, Quinn Snow Survey Cabin NW of Soda Butte, 8200 ft, 24 Oct 2000, Laeger 430 (CAS); Log Meadow, Giant Forest, 6800 ft, 25 Oct 1997, Shevock 16679 (CAS); NE of Bald Dome, 8595 ft, 23 Jul 2008, Jones 315D (COLO, KRAM). Tuolumne Co.: Stanislaus National Forest, Sapps Meadow, 6725 ft, 4 Jun 2009, Willits 501 (CAS); Yosemite National Park, ridge SE of Kibbie Lake, 7365 ft, 12 Sep 2008, Colwell et al. 08-596a (CAS, YM); E of Knapp, Colwell et al. 09-494 (CAS, YM). NEVADA. Washoe Co.: Sierra Nevada, Humboldt-Toiyabe National Forest, Tahoe Meadows, 8740 ft, 13 Sep 2009, Wishner 9512 (CAS, UC). OREGON. Douglas Co.: wetland in basin below jct. of forest road 250 and 251, 5475 ft, 6 Aug 2007, Wagner m2308 (CAS). Lane Co.: Willamette National Forest, E side of Little Groundhog Mountain at jct. with forest road 452, 6.5 mi ESE of south end of Hills Creek Reservoir, 5240 ft, 23 Jul 2004, Wagner m1390 (CAS, UC); Three Sisters Wilderness, Quaking Aspen Swamp, 8 mi E of McKenzie Bridge, 4465 ft, 8 Aug 1999 & 1 Oct 2009, Wagner m0730a, m2556a & m2556b (CAS). Etymology: the species is named for its distribution along the Pacific Coast of the United States in California, Nevada, and Oregon. TAXONOMIC RELATIONSHIPS Ptychostomum pacificum has previously gone un-noticed due to its morphological similarity to 2012] several related species in western North America. The new species appears to be closest to P. schleicheri (Schwagr.) J. R. Spence and P. turbinatum (Hedw.) J. R. Spence, but also mimics large specimens of P. pseudotriquetrum (Hedw.) J. R. Spence & Ramsay. All these species occur on wet sites on soil, organic muck, and sometimes wet rocks, and all can occur in various kinds of wetlands. Until sporophytes were discovered it was generally thought to be a robust form of P. turbinatum. However, that species has broadly turbinate capsules, a plane leaf margin (occasion- ally revolute proximally), usually short-excurrent costa, and is a much smaller plant with smaller leaves (to 4 mm) and shorter stems (to 3—4 cm). Ptychostomum schleicheri 1s also robust, but is characterized by its broad, uncontorted, yellowish leaves, very broad laminal cells, and turbinate capsules. Ptychostomum pseudotriquetrum differs from P. pacificum by its strongly decurrent leaves, distinctly short-excurrent costa, shorter proximal lamina cells, and unistratose limbidium. The most distinctive features of P. pacificum include the large size, colored leaf tip with incrassate cells, mostly percurrent or shorter costa, proximally recurved leaf margins, very strong limbidium, large conspicuous perigonia, and elongate capsule. HABITAT AND ECOLOGY Ptychostomum pacificum is restricted to peren- nially wet fen habitats within coniferous forests primarily dominated by Pinus contorta Lounon, Abies concolor (Gordon & Glend.) Lind., Abies magnifica A. Murray, or a combination of these species. Few flowering plants occur among populations of Ptychostomum pacificum. Taxa listed on herbarium labels for multiple occurrences of P. pacificum include Camassia quamash (Pursh) Greene, Dodecatheon jeffreyi Van Houtte, Dro- sera rotundifolia L., Eriophorum crinigerum (A. Gray) Beetle, Kalmia polifolia Wangenh., Ledum glandulosum Nutt., Pedicularis groenlandica Retz., Phalacroseris bolanderi A. Gray, Rhododendron occidentale (Torr. & A. Gray) A. Gray, Saxifraga oregana Howell, Salix spp., Vaccinium uliginosum L. ssp. occidentale (A. Gray) Hultén, and Vera- trum californicum Durand. Bryophytes generally associated with Ptychostomum pacificum include Aulacomnium palustre (Hedw.) Schwagr., Drepa- nocladus aduncus (Hedw.) Warnst., Philonotis americana (Dism.) Dism., P. tomentosa Mol. in Lor., Ptychostomum weigelii (Spreng.) J. R. Spence, Sphagnum spp., and occasionally, Meesia triquetra Angstr. Fen habitats throughout the range of Ptychostomum pacificum are acidic in pH and all populations contain one or more members of the Ericaceae, usually Vaccinium, Kalmia, or Ledum. Ptychostomum turbinatum is reported to be mildly calciphilous, P. schleicheri is reported to SPENCE AND SHEVOCK: A NEW SPECIES OF PTYCHOSTOMUM 161 be acidophilous, and P. pseudotriquetrum appears to be tolerant of a relatively broad range of pH conditions. At first glance, P. pacificum colonies (especially male plants) are reminiscent of Rhizomnium pseudopunctatum (Bruch & Schimp.) T. Kop. in stature and mat growth-form, although P. pacificum 1s considerably more yellow-green in color. However, with a hand-lens inspection it is clear that this plant is actually a member of the Bryaceae and not the Mniaceae. DISTRIBUTION Populations of Ptychostomum pacificum range from the Cascades of central Oregon in the Three Sisters Wilderness, Willamette National Forest, Lane Co. southward to the southern portion of the Sierra Nevada of California on the Sequoia National Forest, Tulare Co. with a disjunct occurrence recently discovered in the San Bernar- dino Mountains, San Bernardino National Forest. In northern California this species also extends west and south from the Klamath Mountains into the Northern Coast Ranges, Mendocino National Forest and just east of Lake Tahoe in Nevada, Humboldt-Toiyabe National Forest. Although P. pacificum occurs across a wide geographical area, the actual habitat of perennially wet fens within montane to subalpine coniferous forests is con- siderably restricted. Populations range in the north from 4465 feet to over 9300 feet in the southern portion of its range. Based on the herbarium record, this species is more frequently encountered in the Sierra Nevada of California, especially from Yosemite National Park south- ward. However, this may be an artifact of the more systematic bryophyte inventory work that has occurred in this portion of the species’ range. CONSERVATION IMPLICATIONS Fens throughout the Pacific Slope are very fragile habitats and are ecologically diverse and species-rich environments. Within the coniferous zone these fens comprise less than two percent of the landscape. Some of these fens have been adversely impacted by intensive grazing activities, or have been drained, channelized, or the water flow through them altered by road construction or erosion by headcutting. However, these riparian systems today are viewed as important habitat types for biodiversity. Nearly all reported populations of Ptychostomum pacificum are on public lands either administered by the USDA, Forest Service or the USDI, National Park Service. Since fen habitats are increasingly likely to receive a wide variety of protective measures by these land-management agencies, the long-term conservation of this narrowly distributed species seems secure. 162 MADRONO ACKNOWLEDGMENTS We thank Halina Bednarek-Ochyra for the wonder- ful illustration plate. Patricia Eckel provided the Latin diagnosis for which we are most appreciative. We also thank the various botanical personnel at several national forests and national parks for approving collecting permits to develop baseline inventory infor- mation for California, Nevada, and Oregon bryophytes. We are also appreciative of colleagues Dave Wagner, David Toren, Colin Dillingham, Carl Wishner, and Martin Lenz who shared with us specimens and their field observations regarding Ptychostomum pacificum. Comments provided by William Buck and an anony- mous reviewer enhanced the final version. LITERATURE CITED ANDREWS, A. L. 1935. Family Bryaceae. Pp. 184-210 in A. J. Grout (ed.), Moss flora of North America north of Mexico, Vol. 2. Published by the author, Newfane, VT. CrosBy, M., R. E. MAGILL, B. ALLEN, AND S. HE. 2000. A checklist of the mosses. Missouri Botanical Garden, St. Louis, MO. [Vol. 59 HOLYOAK, D. T. AND N. PEDERSEN. 2007. Conflicting molecular and morphological evidence of evolution within the Bryaceae (Bryopsida) and its implications for generic taxonomy. Journal of Bryology 29: 111-124. MALCOLM, B., N. MALCOLM, J. R. SHEVOCK, AND D. H. Norris. 2009. California mosses. Micro- Optics Press, Nelson, New Zealand. Norris, D. H. AND J. R. SHEVOCK. 2004. Contributions toward a bryoflora of California I. A specimen- based catalogue of mosses. Madrono 51:1—131. PEDERSEN, N., C. J. Cox, AND L. HEDENAsS. 2003. Phylogeny of the moss family Bryaceae inferred from chloroplast DNA sequences and morphology. Systematic Botany 28:471-482. SPENCE, J. R. 2005. New genera and combinations in Bryaceae (Bryales, Musci) for North America. Phytologia 87:15—28. . 2007. Nomenclatural changes in the Bryaceae (Bryopsida) for North America II. Phytologia 89: 110-114. AND H. P. RAMSAY. 2005. New genera and combinations in the Bryaceae (Bryales, Musci) for Australia. Phytologia 87:61—71. MADRONO, Vol. 59, No. 3, p. 163, 2012 A NEW COMBINATION IN LINANTHUS (POLEMONIACEAE) FROM IDAHO AND OREGON JOANNA L. SCHULTZ 1060 Driscoll Ridge Road, Troy, ID 83871 ROBERT PATTERSON Department of Biology, San Francisco State University, San Francisco, CA 94132 patters@sfsu.edu ABSTRACT In preparing the treatment for Polemoniaceae for the Flora North America North of Mexico it is necessary to propose a new combination in Linanthus (Polemoniaceae). Key Words: Endemism, Leptodactylon, Linanthus, Polemoniaceae. As many as 10 infraspecific taxa have been recognized for Leptodactylon pungens (Torr.) Nutt. (IPNI 2008), and while floristic treatments have differed in use of infraspecific taxa, we consider Leptodactylon pungens subsp. hazeliae (Peck) Meinke distinctive and warranting recog- nition. When Porter and Johnson (2000) revised the taxonomy of Polemoniaceae, conforming to the goal of recognizing only monophyletic genera, they transferred all seven species formerly recognized as Leptodactylon to Linanthus; how- ever, they did not consider infraspecific taxa that had previously been recognized in Leptodactylon pungens. In preparation for the treatment of Linanthus for the Flora North America north of Mexico project, a new combination is required: Linanthus pungens subsp. hazeliae. TAXONOMIC TREATMENT Linanthus pungens (Torr.) J. M. Porter & L. A. Johnson subsp. hazeliae (Peck) J. L. Schultz & R.Patt., comb. nov. Basionym: Leptodactylon hazelae Peck, Proc. Biol. Soc. Wash. 49:111. 1936; L. pungens subsp. hookeri (Dougl. ex Hook.) Wherry f. hazelae (Peck) Wherry, Amer. Mid. Naturalist 34:383; L. pungens subsp. hazeliae (Peck) Meinke, Madrono 35(2):107. 1988. — Type: USA, Oregon, Wallowa Co., dry rocky slope, Snake River Canyon near mouth of Battle Creek, 13 April 1934, Barton s.n. (holotype: WILLU 18415). Linanthus pungens subsp. hazeliae is a rare and very narrow endemic to the Snake River Canyon region that forms part of the Idaho-Oregon border. It occurs within the overall range of Linanthus pungens subsp. pungens, but the two subspecies are not sympatric. Linanthus pungens subsp. hazeliae has opposite, soft-filiform distal leaf lobes, while subsp. pungens has alternate, acicular, sharp-tipped distal leaf lobes. Meinke (1988) and Moseley (1989) provide careful and complete discussions of the distinctive features and habitat of the subspecies, including comment on its rarity. Descriptions and discussions of infraspecific taxa of Linanthus pungens have been inconsistent and often lacking in detail; however, analytical approaches have improved since these taxa were described, and there are more botanists today who can contribute valuable field information. Therefore, it seems an appropriate time for a thorough systematic study of Linanthus pungens throughout its range. LITERATURE CITED THE INTERNATIONAL PLANT NAMES INDEX. 2008. Published on the internet. Website http://www. ipni.org/index.html [accessed 17 June 2011]. MEINKE, R. J. 1988. Leptodactylon pungens subsp. hazeliae (Polemoniaceae), a new combination for a Snake River endemic. Madrono 35:105—111. MOSELEY, R. K. 1989. Field investigations of Lepto- dactylon pungens ssp. hazeliae (Hazel’s prickly phlox) and Mirabilis macfarlanei (Macfarlane’s four-o-clock), region 4 sensitive species, on the Payette National Forest, with notes on Astragalus vallaris (Snake Canyon milkvetch) and Rubus bartonianus (bartonberry). Technical Report, Idaho Department of Fish and Game, Boise, ID.Website https://fishandgame.idaho.gov/ifwis/idnhp/cdc_pdf/ moser89k.pdf [accessed 28 March 2012]. Peck, M. E. 1936. Six new plants from Oregon. Proceedings of the Biological Society of Washing- ton 49:111. PORTER, J. M. AND L. A. JOHNSON. 2000. A phylogenetic classification of Polemoniaceae. Aliso 19:55-91. MADRONO, Vol. 59, No. 3, pp. 164-165, 2012 REVIEWS Research & Recovery in Vernal Pool Landscapes. Edited by D. G. Alexander and R. A. Schlising. 2011. Studies From The Herbarium, No. 16. California State University. Chico, CA. 175 pp. ISBN 978-0-9761774-3-2. Price: $12.00, paper- back. Vernal pools are seasonally ephemeral wet- land ecosystems, generally of regions with a Mediterranean climate. Vernal pool basins be- come inundated during the winter, support a colorful procession of highly specialized plants and animals through the spring, and then become completely dry by summer. These charismatic ecosystems have intrigued scientists from a varie- ty of disciplines: evolutionary biology, ecology, genetics, taxonomy, geology, hydrology, and (in- creasingly) conservation biology. This broad spectrum of workers has produced a large and diverse body of literature pertaining to vernal pools, much of which has been first presented to the scientific community in a series of vernal pool symposia and subsequent proceedings (Jain 1976b; Jain and Moyle 1984; Ikeda and Schlising 1990; Witham et al. 1998; Schlising and Alexander 2007). In the introduction to the proceedings from the first vernal pool symposium at the University of California at Davis, editor Subodh Jain ac- knowledged that he and other symposium orga- nizers wondered if “‘there is enough known about vernal pools to hold such a meeting” (Jain 1976a). Many dozens of academic papers, six vernal pool conferences (counting the first), and 35 years later, scientists who study vernal pool ecosystems have again contributed original papers toward a con- ference proceedings: Research & Recovery in Vernal Pool Landscapes (Alexander and Schlising 2011). As with previous vernal pool conference proceedings, this most recent volume presents timely research on this important and imperiled California Floristic Province ecosystem. The 14 contributed papers in this volume are arranged into five sections: Plants, Animals, Geo- logy and Soils, Management, and Preservation History and Recovery Plans. The Plants section begins with a paper by keynote address speaker Ellen Bauder, a plant ecologist who has studied the vernal pools of southern California for over 20 years. Drawing upon this experience, Bauder provides guidance on the topic of appropriate experimental design in vernal pool ecosystems, which can be highly variable spatially and temporally. The three other papers in this section include a population ge- netics analysis of Butte County meadowfoam (Limnanthes floccosa ssp. californica) by Sloop; a far-reaching study of the ecology and evolution of vernal pool Lasthenia species by Emery et al.; and an ecological study by Leong examining the relationship between a vernal pool endemic plant species (Blennosperma bakeri) and its pollinators in naturally occurring and created vernal pools. The Animals section comprises three research articles: the first is a methodologically innovative study of the migration distance of terrestrial- phase California tiger salamanders (Ambystoma californiense) by Searcy and Shaffer; the second 1s a contribution by Bogiatto et al. quantifying the usage of inundated vernal pools by geese and swans; and the third is an editors’ summary of the presentation given by Rogers on macroinverte- brate bioassessment. The Vernal Pools Geography and Soils sec- tion 1s composed of two papers. The first is a continuation of the important work of Holland to document the distribution and areal extent of vernal pool landscapes in the Central Valley. In this Herculean study, Holland prepared maps of vernal pool landscapes by manually photointer- preting high-resolution (l-m) National Agricul- tural Imaging Program (NAIP) imagery in a Geographic Information System (GIS). Using the resulting GIS layer, Holland calculated vernal pool extent as of 2005, overall loss since baseline conditions were evaluated (1976-1995), loss by county, and loss to various land-use types. Three color figures nicely illustrate the method em- ployed, and a fold-out, full-color plate showcases the resulting map of vernal pool landscapes of the Central Valley. Also in this section is a paper prepared by the editors that summarizes an oral presentation by Conlin on substrates encountered during the Butte Area Natural Resources Con- servation Service Soil Survey. Although brief, this paper is nicely illustrated and informative. The Management and Preservation History and Recovery Plans sections together comprise five papers. Bauder and Bohonak describe a pro- mising new method for vernal pool functional assessment in southern California; Schohr high- lights the growing coalition of individuals and groups that support managed grazing of Cali- | fornia landscapes (including vernal pool land- scapes); and the editors provide a summary of a | talk by Witham on the planning process associ- | ated with the construction of UC Merced. These | sections conclude with two papers on the topic of vernal pool conservation planning in California | and adjacent Oregon: one by a long-time water | rights activist and vernal pool conservationist, and the other by a biologist from the U.S. Fish 2012] and Wildlife Service. These contributions provide an interesting insight into the process, and yes, politics, of conservation planning. This volume would be a solid addition to the bookshelves of anyone interested in the natural history of western North America, as many of the contributed papers are of general biological interest. For those who study vernal pools, how- ever, this book is the latest in a series of in- dispensable vernal pool conference proceedings that together constitute a substantial portion of the research on these fascinating ecosystems. The volume is not perfect; for example, some might question the editors’ choice to include summaries of oral presentations where the original presenter elected to not contribute a paper. But consider- ing the strength of the contributed papers, the very low price ($12.00 + tax, shipping), and the fact that any income from sales beyond the cost of printing benefits the non-profit organization Studies From the Herbarium, any minor critiques seem immaterial. Additional information on this volume can be found at: http://www.csuchico.edu/herbarium/index.shtml http://www.csuchico.edu/herbarium/studies/ detailed-book-list.shtml —C. MATT GUILLIAMS, UC/JEPS Her- baria & Department of Integrative Biology, BOOK REVIEW 165 University of California, Berkeley CA 94720- 2465; matt_g@berkeley.edu. LITERATURE CITED ALEXANDER D. G. AND R. A. SCHLISING (eds.). (2011). Research & recovery in vernal pool landscapes. Studies from the Herbarium No. 16. California State University, Chico, CA. IKEDA D. H. AND R. A. SCHLISING (eds.). (1990). Vernal pool plants, their habitat and_ biology. Studies from the Herbarium No. 8. California State University, Chico, CA. JAIN S. 1976a. Foreword in: Vernal pools: their ecology and conservation. Institute of Ecology Publication No. 9. University of California, Davis, CA. (ed.). (1976b). Vernal pools: their ecology and conservation. Institute of Ecology Publication No. 9. University of California, Davis, CA. JAIN S. AND P. MOYLE (eds.). (1984). Vernal pools and intermittent streams. Institute of Ecology Publica- tion No. 28. University of California, Davis, CA. SCHLISING R. A. AND D. G. ALEXANDER (eds.). (2007). Vernal pool landscapes. Studies From the Herba- rium No. 14. California State University, Chico, CA. WITHAM C. W., E. T. BAUDER, D. BELK, W. R. FERREN JR., AND R. ORNDUFF (eds.). (1998). Ecology, conservation, and management of vernal pool ecosystems. Proceedings from a 1996 Confer- ence. California Native Plant Society, Sacramento, CA. MADRONO, Vol. 59, No. 3, p. 166, 2012 NOTEWORTHY COLLECTION Montana MIMULUS CLIVICOLA Greenm. (PHRYMACEAE). —Sanders Co., Lolo National Forest, Clear Creek drainage, ca. 9.6 km WNW of Thompson Falls, 47.6160N, —115.4785W (WGS84), in upland montane forest with Pinus ponderosa, Pseudotsuga menziesii, Pseudoroegneria spicata, Collinsia parviflora, Hypericum perforatum, Centaurea stoebe ssp. micranthos, Carex geyeri, and Spiraea betulifolia, in grassy 0.02-hectare opening among mature conifers, well-drained gravelly silt loam, SSE aspect, 55% slope, elev. 1168 m, approx. 900 plants primarily in bare soil patches, 22 July 2010, C. Odegard 48 (MONTU). The collection location is one of eight sites in the Clear Creek drainage where M. clivicola was found in July 2010 and June—July 2011. Collectively, these sites consist of a single ‘“‘element occurrence” (NatureServe 2012a) that contains about 11,000 M. clivicola plants. All of the sites are on partially forested, southerly, 55— 70% slopes; site elevations vary from 840 to 1250 m. Previous knowledge. Mimulus clivicola (bank mon- keyflower, hill monkeyflower, North Idaho monkey- flower) is endemic to the inland Pacific Northwest, from northern and west-central Idaho to northeastern Or- egon. A historic (1889) collection from ‘“‘Washington”’ has not been confirmed; no extant populations are known from the state (Washington DNR 2012). In northern Idaho M. clivicola is found in open conifer stands on southerly slopes at elevations below 1250 m, while in west-central Idaho and northeastern Oregon it often occurs in unforested habitats at elevations up to 1675 m (Lorain 1993). Throughout its range M. clivicola 1s typically found in open pockets of vernally moist, exposed mineral soil (Lorain 1993; Consortium of Pacific Northwest Herbaria 2012). Mimulus clivicola is documented within 13 km of the Montana border in the Coeur d’ Alene, St. Joe, and North Fork Clearwater river basins of northern Idaho (IF WIS 2011). At the headwaters of the North Fork St. Joe River south of Lookout Pass, it occurs less than 2 km from the state line (IFWIS 2011). Significance. The collection is the first report of M. clivicola in Montana. The Clear Creek sites are ca. 24 km east of the nearest documented M. clivicola site in Idaho (IFWIS 2011). Considering the species’ close proximity to the Montana border south of Lookout Pass, M. clivicola should also be searched for in western Mineral County, Montana. In the early 1990’s M. clivicola was a candidate for federal listing as a threatened or endangered spe- cies (USDI, FWS 1995), and it was managed as a For- est Service sensitive species on national forest lands throughout its range in the 1990’s. It is now considered globally secure (G4), but vulnerable (S3) in both Idaho and Oregon (NatureServe 2012b). It is provisionally ranked S1S3 in Montana while its conservation status is reviewed (MTNHP 2012). Due to the species’ rarity in Montana and potential impacts to the Clear Creek sites from proposed forest management activities, in 2011 the Forest Service Northern Region designated M. clivicola as a sensitive species in Montana. —CRAIG ODEGARD, Botanist, Plains Ranger Station, Lolo National Forest, 408 Clayton St, Plains, MT 59859. codegard@fs.fed.us. LITERATURE CITED CONSORTIUM OF PACIFIC NORTHWEST HERBARIA. 2012. Specimen data for Mimulus clivicola. Univer- sity of Washington Herbarium, The Burke Museum of Natural History and Culture, Seattle, WA. Website: http://www.pnwherbaria.org [accessed 20 February 2012]. IDAHO FISH AND WILDLIFE INFORMATION SYSTEM (IFWIS). 2011. IFWIS Mimulus clivicola locations provided by Jim Strickland. Botany Data Coordi- nator, Idaho Dept. of Fish and Game, Boise, ID. Contact through website: https://fishandgame. idaho. gov/ifwis/portal/ [accessed 30 November 201 1]. LORAIN, C. C. 1993. Conservation assessment for Mimulus clivicola (bank monkeyflower). Prepared for USDA Forest Service, Northern Region, Missoula, MT. Website: https://fishandgame.idaho. gov/ifwis/iidnhp/cdc_pdf/lorac93a.pdf [accessed 8 January 2012]. MONTANA NATURAL HERITAGE PROGRAM (MTNHP). 2012. Montana field guide. Montana Natural Heri- tage Program, Helena, MT. Website: http://fieldguide. mt.gov/detail_PDSCR1BOS0O.aspx [accessed 17 Janu- ary 2012]. NATURESERVE. 2012a. Habitat-based plant element occurrence delimitation guidance. NatureServe, Arlington, VA. Website: http://www.natureserve. org/explorer/decision_tree.htm [accessed 7 January 2012); 2012b. NatureServe Explorer: an online encyclopedia of life. NatureServe, Arlington, VA. Website: http://www.natureserve.org/explorer [ac- cessed 7 January 2012]. U.S. DEPARTMENT OF INTERIOR, FISH AND WILDLIFE SERVICE (USDI, FWS). 1995. Endangered and threatened wildlife and plants; 12-month finding on a petition to list Mimulus clivicola (bank monkey- flower). Federal Register 60 (187): 49818-49819. Website: http://ecos.fws.gov/docs/federal_register/ fr2896.pdf [accessed 17 January 2012]. WASHINGTON DEPARTMENT OF NATURAL RESOURCES (DNR). 2012. Field guide to selected rare plants of Washington. Washington Dept. of Natural Re- sources, Washington Natural Heritage Program, Olympia, WA. Website: http://www1.dnr.wa.gov/ nhp/refdesk/fguide/pdf/mimcli.pdf [accessed 11 January 2012]. MADRONO, Vol. 59, No. 3, p. 167, 2012 NOTEWORTHY COLLECTION CALIFORNIA TRIFOLIUM TRICHOCALYX A. A. Heller (FABACEAE). —Mendocino Co., The Conservation Fund, Big River Forest, 9 mi. E of Russian Gulch State Park, 1.6 mi. SW of McGuire Hill (742 ft.), East Branch Little North Fork Big River, 214 m (702 ft.), Mathison Peak 7.5’ quadrangle, 39°20.102'N, 123°38.176'W, [NAD83], May 27, 2011, Hulse-Stephens 1952 (UC); same site different location along road, 39 °20.093'N, 123 °38.195' W, [NAD83], June 2, 2011, Heise 2916 (UC). Observed at two locations, within 0.25 miles of each other, one with approx. 5000 plants the other with 50 plants, on road 21020, north-facing slope within redwood/Douglas fir/tanoak forest. Plants were grow- ing in shaded, moist soil of a seasonal logging road that was last graded 5 years prior. Associated species include: Trifolium dubium, T. variegatum, Hypochaeris radicata, Acmispon parviflorus, Gamochaeta ustulata, Deschampsia elongata, Bromus carinatus var. carinatus, Juncus patens, Rubus ursinus, R. leucodermis, Iris douglasiana, Cirsium vulgare, Cyperus eragrostis, Equi- setum telemateia subsp. braunii. Previous knowledge. Trifolium trichocalyx (Monterey clover) is an herbaceous annual endemic to California and is both state and federally endangered (CNDDB 2012). Previously the species was only known to occur in a 206-acre (83 ha) area in the central portion of the Monterey Peninsula within naturally occurring Mon- terey pine forest. This area is completely surrounded by residential and recreational development so there is little habitat available for the species to expand into. In its closed-cone forest habitat 7. trichocalyx is very responsive to fire and other disturbances, and without it becomes scarce. Up to 1000 plants were observed following a 1987 fire, thereafter none in 1992 and only 22 in 1995 (USFWS 2009). We know of no recent reports. Significance. This northern population extends the range of the species approximately 200 mi (322 km) north of the Monterey Peninsula. The habitat here is markedly different in terms of geology, soils, and forest composition thus adding significantly to our knowledge of the ecology and distribution of this species. In Mendocino County 7. trichocalyx is a species of old logging roads situated on mesic north-facing slopes of redwood/Douglas fir forest and in lieu of fire appears to be dependent on grading, which reduces shade and competition. At least at the pre-flowering stage (R. Morgan, CNPS fellow, personal communication) the two forms are identical in all respects, except that the Mendocino form differs in having a faint chevron on the leaflets. According to Morgan ‘the chevron difference could indicate that the Mendocino occurrence may represent a long-standing native population rather than a recent introduction”. No apparent differences were seen between the Mendocino and Monterey forms after examining herbarium sheets of 7: trichocalyx from the Pacific Grove Museum of Natural History (V. Yadon, Director Emeritus, Pacific Grove Museum of Natural History, personal communication). Total DNA was prepared from a dried leaf sample of the Mendocino County 7. trichocalyx using the DNeasy Plant Mini Kit (QIAGEN, Germany) and following the manufacturer’s recommended protocol, except that the elution buffer was pre-heated to 65°C. The nuclear internal transcribed spacer region of 18S- 26S rDNA and the chloroplast trnL intron region were amplified and sequenced as_ previously described (Ellison et al. 2006). DNA homology searches of the GenBank DNA sequence database were conducted using BLAST (http://blast.ncbi.nlm.nih.gov/Blast.cgi) to establish sequence identities. Both sequences were found to be identical to those reported previously for the Monterey population of this species (Ellison et al. 2006). ACKNOWLEDGMENTS Special thanks to Randall Morgan and Vern Yadon who confirmed our identification of JT. trichocalyx, provided timely and relevant information regarding the Monterey Peninsula occurrences, and offered useful suggestions for this article. We also acknowledge Scott Kelly of The Conservation Fund and Tony LaBanca of the California Department of Fish and Game for their support in the long-term conservation of this rare taxon. —KERRY HEISE, 453 Mendocino Dr., Ukiah, CA 95482; GERI HULSE-STEPHENS, 915 East Hill Road, Willits, CA 95490; NICHOLAS ELLISON, Grasslands Research Centre, Tennent Drive, Private Bag 11008, Palmerston North 4442, NZ. kheise@copper.net. LITERATURE CITED CALIFORNIA NATURAL DIVERSITY DATABASE (CNDDB). 2012. Trifolium trichocalyx. California Department of Fish and Game, Sacramento, CA. Website: http://www.dfg.ca.gov/biogeodata/cnddb/ [accessed 02 March 2012]. ELLISON, N. W., A. LISTON, J. J. STEINER, W. M. WILLIAMS, AND N. L. TAYLOR. 2006. Molecular phylogenetics of the clover genus (Trifolium — Leguminosae). Molecular Phylogenetics and Evo- lution 39:688—705. UNITED STATE FISH AND WILDLIFE SERVICE (USFWS). 2009. Trifolium trichocalyx, (Monterey Clover), 5-year review: Summary and evaluation. U.S. Fish and Wildlife Service, Ventura Fish and Wildlife Office, Ventura, CA. Volume 59, Number 3, pages 109-168, published 27 August 2012 SUBSCRIPTIONS — MEMBERSHIP The California Botanical Society has several membership types (individuals ($40 per year; family $45 per year; emeritus $32 per year; students $27 per year for a maximum of 7 years). Late fees may be assessed. Beginning in 2011, rates will increase by $5 for all membership types except life memberships, for which rates will increase by $100, and student memberships, which will not show a rate increase. Members of the Society receive MADRONO free. Institutional subscriptions to MADRONO are available ($75). Membership is based on a calendar year only. Life memberships are $850. Applications for membership (including dues), orders for subscriptions, and renewal payments should be sent to the Membership Chair. Requests and rates for back issues, changes of address, and undelivered copies of MADRONO should be sent to the Corresponding Secretary. INFORMATION FOR CONTRIBUTORS Submit manuscripts to http://www.edmegr.com/madrono/. If you are submitting a manuscript for the first time to the MADRONO PeerTrack website, you must first register by clicking "register now" and following the instructions. Manuscripts by authors having outstanding page charges will not be sent for review. Manuscripts may be submitted in English or Spanish. English-language manuscripts dealing with taxa or topics of Latin America and Spanish-language manuscripts must have a Spanish RESUMEN and an English ABSTRACT. For all articles and short items (NOTES, NOTEWORTHY COLLECTIONS, POINTS OF VIEW, etc.), follow the format used in recent issues for the type of item submitted. Detailed information for contributors, including instructions for manuscript preparation, is available at the California Botanical Society website, www.calbotsoc. org; follow the link for MADRONO. Authors are encouraged to include the names, addresses, and e-mail addresses of two to four potential reviewers with their submitted manuscript. All California Botanical Society members current in the volume year that their contributions are published are allowed five free pages per volume year. Additional pages will be charged at the rate of $40 per page. Joint authors may apply all or a portion of their respective five-page allotments to a jointly-published article. Partial pages will be charged as full. The purpose of this fee is not to pay directly for the costs of publishing any particular paper, but rather to allow the Society to continue publishing MADRONO on a reasonable schedule, with equity among all members for access to its pages. Printer’s fees for color plates and other complex matter (including illustrations, charts, maps, photographs) will be charged at cost. Author’s changes after typesetting @ $4.50 per line will be charged to authors. Page charges are important in maintaining Madrono as a viable publication, and timely payment of charges 1s appreciated. At the time of submission, authors must provide information describing the extent to which data in the manuscript have been used in other papers that are published, in press, submitted, or soon to be submitted elsewhere. SMITHSONIAN INSTITUTION LIBRARIES WMI 088 01700 3708