VOLUME 57, NUMBER 2 APRIL-JUNE 2010 OX W IK 3 on EFFECTS OF FIRE ON GERMINATION OF ERICAMERIA FASCICULATA (ASTERACEAE), A RARE MARITIME CHAPARRAL SHRUB Jon ReDetka and SUSGN CC. LGD eC 6. cches sik aivscdtoucsoreats icaewben’ Te DISTRIBUTION AND COMMUNITY ASSOCIATIONS OF CAPE IVY (DELAIREA ODORATA) IN CALIFORNIA Ramona Robison and Joseph M. DiTOMASO .iccccccccccccccccccccccccccccccecceeeeeeeees 85 SPECIES BOUNDARIES IN PYRROCOMA LIATRIFORMIS AND PYRROCOMA SCABERULA (ASTERACEAE) BASED ON AFLP DATA James F: Smith, Dusty N. Perkins, Curtis R. Bj6rk, and Gina Glenne ..... 95 ONE TAXON OR Two: ARE FRASERA UMPQUAENSIS AND F. FASTIGIATA (GENTIANACEAE) DISTINCT SPECIES? Barbara L. Wilson, Valerie Hipkins, and Tom N. Kaye ...........ccccceeeeeeeeeees 106 THE EFFECTS OF LONG-TERM DROUGHT ON Host PLANT CANOPY CONDITION AND SURVIVAL OF THE ENDANGERED ASTRAGALUS JAEGERIANUS (FABACEAE) T. R. Huggins, B. A. Prigge, M. R. Sharifi, and P. W. Rundel................4+. 120 SN eee A RESURRECTION FOR SISKIYOU BELLS, PROSARTES PARVIFOLIA (LILIACEAE), A RARE SISKIYOU MOUNTAINS ENDEMIC Michael Mesler, Robin Bencie, and Bianca Hayashi .....cccccc cc cceeeeeeeeeeeeeees 129 area SEDUM VALENS (CRASSULACEAE), A NEW SPECIES FROM THE SALMON RIVER CANYON OF IDAHO CUE UL SWINGIN OIG noi eanele scant aeeeeaNe eas oN Oe Sales San as USsE Sone cacanes savledeeee oaeeeeoene 136 ABIES MAGNIFICA VAR. CRITCHFIELDII, A NEW CALIFORNIA RED FIR VARIETY FROM THE SIERRA NEVADA ROM CUGIVE HEGQIILCT mrt tereee ee reae ee eee ssh eA Sanaa. Saeed oobaticn eee at BEesa nee 14] MADRONO (ISSN 0024-9637) is published quarterly by the California Botanical Society, Inc., and is issued from the office of the Society, Herbaria, Life Sciences Building, University of California, Berkeley, CA 94720. Subscription information on inside back cover. Established 1916. Periodicals postage paid at Berkeley, CA, and additional mailing offices. Return requested. POSTMASTER: Send address changes to MADRONO, Kim Kersh, Membership Chair, Uni- versity and Jepson Herbarium, University of California, Berkeley, CA 94720-2465. kersh @berkeley.edu. Corresponding Editor—TIMOTHY LOWREY Copy Editor—RICHARD WHITKUS Museum of Southwestern Biology Department of Biology MSC03 2020 Sonoma State University University of New Mexico 1801 E. Cotati Avenue Albuquerque, NM 87131-0001 Rohnert Park, CA 94928-3609 madrono @unm.edu whitkus @ sonoma.edu Book Editor—JON E. KEELEY Noteworthy Collections Editors—DIETER WILKEN, MARGRIET WETHERWAX Board of Editors Class of: 2010—FRED Hrusa, California Department of Food and Agriculture, Sacramento, CA RICHARD OLMSTEAD, University of Washington, Seattle, WA 2011—JAMIE KNEITEL, California State University, Sacramento, CA KEVIN RIcE, University of California, Davis, CA 2012—GRETCHEN LEBUHN, San Francisco State University, CA ROBERT PATTERSON, San Francisco State University, CA 2013—ErIc ROALSON, Washington State University, WA KRISTINA SCHIERENBECK, California State University, Chico, CA CALIFORNIA BOTANICAL SOCIETY, INC. OFFICERS FOR 2010-2011 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: Rodney G. Myatt, Department of Biological Sciences, San Jose State University, San Jose, CA 95192, rgmyatt @ gmail.com Recording Secretary: Mike Vasey, Department of Biology, San Francisco State University, San Francisco, CA 94132, mvasey @sfsu.edu Corresponding Secretary: Heather Driscoll, University Herbarium, University of California, Berkeley, CA 94720, hdriscoll @berkeley.edu 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, 3060 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: Ben Carter, Department of Integrative Biology and University Herbarium, University of California, Berkeley, CA 94720, bcarter @berkeley.edu. Webmaster: Susan Bainbridge, Jepson Herbarium, University of California, Berkeley, CA 94720-2465, sjbainbridge @ berkeley.edu. This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). MADRONO, Vol. 57, No. 2, pp. 77-84, 2010 : : ‘ uct 12 2010 LIBRARIES EFFECTS OF FIRE ON GERMINATION OF ERICAMERIA FASCICULATA (ASTERACEAE), A RARE MARITIME CHAPARRAL SHRUB JON R. DETKA Division of Science & Environmental Policy, California State University Monterey Bay, 100 Campus Center, Building 53, Seaside, CA 93955-8001 jon_detka@csumb.edu SUSAN C. LAMBRECHT Department of Biological Sciences, San José State University, One Washington Sq., San José, CA 95192-0100 ABSTRACT Knowledge gaps regarding the greenhouse propagation of rare, fire-adapted plant species can impede community level conservation efforts that require fire and active revegetation as management tools. Ericameria fasciculata is a rare shrub endemic to the maritime chaparral community of the central California coast and a listed species of concern. Prescribed burning is actively used in maritime chaparral to maintain community composition and conserve several species of concern with known affinities for fire-related conditions. No study has investigated the seed viability and germination requirements for E. fasciculata. The goal of this study was to ascertain the (1) greenhouse propagation potential of E. fasciculata for planned restoration efforts and (2) to determine if fire-related conditions inhibit or promote E. fasciculata germination. Seed dissection and viability testing indicated that a large percentage of seed were empty or inviable. A greenhouse study examined the potential for fire- related germination cues from heating, light, and charate. Heating and charate had negative effects on seed germination. The combination of heating and charate treatments were particularly lethal. Exposure to light or the addition of GA; had no influence on germination rates. Results suggest that seed germination of E. fasciculata is inhibited by fire and, therefore, this species is dependent on seedling establishment between fire events. Key Words: Asteraceae, Ericameria fasciculata, fire, germination, maritime chaparral. The ability to propagate rare endemic plant species has become increasingly important with the advent of conservation goals directed at revegeta- tion and restoration of endemic plant communities (U.S. Army Corps of Engineers 1997; Padgett et al. 1999). The legal impetus for such actions in California stem from the ratification of the Endangered Species Act (ESA) and the California Endangered Species Act (CESA). In tandem with these legal requirements, a growing appreciation of native California flora and the intrinsic values associated with species diversity has prompted the need for additional information regarding the propagation of endemic plant species (Emery 1988). Much of this concern for the protection and conservation of rare species also stems from the knowledge that increased anthropogenic influ- ences (e.g., global climate change and habitat fragmentation) are moving at rates that exceed the ability of species with restricted distributions to accommodate (Davis 1989). In response to these biological crises, it has been recommended that urgency be placed on the development of conser- vation techniques that can be used to actively increase the size and distribution of rare plant populations (Primack and Miao 1992). Active restoration of rare plant populations by seed broadcasting typically fails to establish self- sustaining populations (Primack 1996). Several hypotheses have been proposed as explanations for the lack of success from introduced seed and general inability to establish self-sustaining re- generative populations. These include the need for (1) suitable habitat/community composition, (2) potentially specialized germination responses and (3) improper seasonal timing of seed collection or distribution (Primack 1996; Willson and Traveset 2000). In addition, seed broadcast- ing challenges are compounded when working with rare species as the current plant distribution may not adequately represent the conditions required for germination, required germination conditions may be unknown, and wildland seed stock may be in short supply. Therefore, restora- tionists have shifted their efforts to propagating rare species in greenhouses for reintroduction into wildlands (Gordon-Reedy and Mistretta 1997; Padgett et al. 1999). Active outplanting of greenhouse-propagated plants into suitable un- occupied habitat may increase the dispersal potential of a species with limited seed dispersal capabilities (Primack and Miao 1992). It is well established that many plant species in fire-prone Mediterranean-type plant communities have unique fire adapted seed life histories (Went et al. 1952; Sweeney 1956; Keeley and Zedler 78 MADRONO 1978; Keeley 1987). In general, these germination adaptations are responses to the drastically altered environmental conditions that are present following fire events. However, the majority of these studies investigated the role of fire in Southern California inland chaparral dominated by Adenostoma fasciculatum Hook. & Arn. (Christensen and Muller 1975a, b; Keeley et al. 1981; Moreno and Oechel 1991; Swank and Oechel 1991; Odion and Davis 2000). There are no clear trends in the degree or trend of seed germination responses among closely related taxa in the maritime chaparral plant community (Davis et al. 1989). Furthermore, most studies have investigated the role of fire on germination of common chaparral shrub species (Keeley and Zedler 1978; Keeley and Keeley 1984; Keeley 1987, 2006; Tyler 1995; Holl et al. 2000). Woody chaparral plant species with an obli- gate seeding fire life history are particularly challenging to propagate given that many are reliant on specific combinations of fire-related germination cues for their emergence (Keeley 1987; Emery 1988; Gordon-Reedy and Mistretta 1997; Padgett et al. 1999; Boyd 2007). Fire may impact seed germination by altering the micro- scale environmental conditions through heating, charate, and changes in available light (Keeley 1987; Davis et al. 1989; Baskin and Baskin 2001) During a fire, temperatures at the surface of sandy soils can exceed 600°C (Sweeney 1956). However, heating from fire dramatically decreas- es (50—200°C) with small changes in depth (1— 2 cm) and duration (<20 minutes) (Sweeney 1956; Davis et al. 1989). Germination responses of woody chaparral shrubs to dry heating at temperatures similar to near-surface burial (70— 120°C) are extremely varied, ranging from increases, decreases, and no effect on germination rates (Baskin and Baskin 2001). The increased availability of light often asso- ciated with the post-burn chaparral environment can have a significant impact on the emergence of seedlings (Sweeney 1956; Keeley 1987). Fire can dramatically reduce canopy vegetation cover and litter, increasing the amount of available light. There are few clear trends in light-facilitated germination response among plants. But, it is generally accepted that smaller seeded species have more seed residing near the surface of soils and are more dependent on light signaling than are larger seeded species (Pons 2000). Keeley (1987) noted 22 species from 15 families of woody California chaparral shrubs that exhibited a significant light-stimulated germination response. Interestingly, several species within families exhibited no consistent trend in light-related germination responses. Previous studies have demonstrated a wide variety of effects of charred wood on germination of individual species (McPherson and Muller 1969; [Vol. 57 Keeley and Pizzorno 1986; Keeley 1987; Thanos and Rundel 1995; Tyler 1996). Additionally, studies have elucidated that there are species-specific germination responses to different combinations of heat, light, and charate (Keeley and Keeley 1984; Keeley et al. 1985; Keeley 1987; Tyler 1996). Ericameria fasciculata (Eastw.) J. F. Macbr. isa stout (<5 dm tall) shrub in the Asteraceae (Hickman 1993), previously classified as Haplo- Pappus eastwoodiae H. M. Hall. E. fasciculata is listed as a species of concern (List 1B) by the California Native Plant Society (Skinner and Pavlik 1994), and is proposed for listing as an endangered species on the Federal Endangered Species List (CNDDB 2002). The most distin- guishing features of E. fasciculata are its aromatic, resinous, cylindrical leaves arranged 1n fasciculate bundles, its pale yellow radiate flower heads that bloom in July and its achenes that are attached to a dense golden-white pappus (Matthews 1997). The geographic range of E. fasciculata 1s estimated at less than 4,000 hectares (CNDDB 2002). Scattered individuals occur in coastal dune, central maritime chaparral, and coastal closed cone pine forest from 30-270 meters (MSL) elevation in Monterey Co., California, but have historically been most abundant in the central maritime chaparral plant community (Griffin 1976, 1978; Van Dyke and Holl 2003). This central maritime chaparral plant community consists of a diverse array of fire-adapted endemic sclerophyllous shrubs, residing in pre- dominately sandy soils and blanketed by the summer fog of the coastal regions (Griffin 1978). Several taxa related to E. fasciculata have demonstrated a range of post-fire responses making it difficult to infer the potential for fire stimulated seed germination (Keeley and Keeley 1984; Keeley 1987; Tueller and Payne 1987; Holl et al. 2000). Prior to this study, the germination response of E. fasciculata seed was not known. This study was prompted primarily in response to the need for information regarding (1) green- house propagation potential of E. fasciculata for planned restoration efforts and (2) to determine if burning inhibits or promotes E. fasciculata seed germination. Field observations noted low oc- currences of natural post-burn E. fasciculata seedling emergence coupled with a catastrophic mortality rate in the first year following pre- scribed burning (Detka 2007). In addition, initial attempts to propagate the species without fire- related stimuli in greenhouses yielded mediocre results (Detka personal observation). MATERIALS AND METHODS Seed Collection and Storage Mature capitula were collected from 29 plants located on the Fort Ord, Parker Flats Reserve, 2010] Monterey, CA (36°38'4.60"N, 121°46'38.78"W) during September 2005 and 2006. A voucher specimen was collected, pressed, and mounted for deposit at the Carl W. Sharsmith Herbarium, San Jose State University, CA. All collected seed was grouped by year and no cleaning or sorting was conducted. Seeds were stored loosely in brown paper bags in unlit standard refrigeration at SC— 10C. Seed Viability Testing Prior to propagation trials, three random samples of 300 seeds each were acquired from 2005 and 2006 seed stocks. Seeds were visually inspected under 10 hand lens magnification and sorted into three categories; intact, aborted, and dead. Seeds with obvious external structural deformations, such as being smaller than the mean seed length or width, were imbibed in 1 mM CaCl, solution at lab temperature for | hr and dissected under a dissection microscope to determine if an embryo was present. Those seeds with no embryos present and no signs of physical damage, predation, or fungal attack were record- ed as aborted. Seeds that appeared intact with no external deformations, physical damage, or fun- gus present were grouped as intact. Seeds with obvious physical damage from predation or fungal attack were recorded as dead. We used a 1% 2,3,5-triphenyl-tetrazolium Chloride (TZ) staining technique (Carolina Bio- logical Supply Co., Burlington, NC) to evaluate collected intact seeds for embryo viability from each seed stock for 2005 and 2006 (Lakon 1949). Prior to TZ staining we soaked intact seeds in a 1 mM CaCl, solution at laboratory temperature for 1 hour to imbibe seeds to soften the seed coat for dissection. We removed the seed coats of intact seeds under a dissection microscope and inspected for intact embryo material. Seeds containing no embryo or the presence of decayed soft tissues were recorded and pooled in the dead category. Those seeds containing intact embryos were soaked in TZ staining solution for 18 hours. Care was also taken to insure that embryos remained completely submerged in the solution with no contact to air or exposure to light. The presence of a pink to red coloration along portions of the embryo indicated viable seed. Greenhouse Germination Trials Greenhouse germination trials were conducted in the late fall following seed collection to examine the role of fire-related cues in the germination of Ericameria fasciculata. The fire- related treatments were: (1) pre-sowing heat, (2) powdered charate from Adenostoma fasciculatum wood, and (3) light. Initial germination trials had DETKA AND LAMBRECHT: FIRE AND GERMINATION OF E. FASCICULATA 19 produced poor rates of germination so gibberellic acid (GA3) treatment was applied. Seeds were sorted from remaining plant material and inspected for signs of physical damage (1.e., predation, fungal invasion) or obvious deformities. Thirty-two lots of 150 intact seeds were sorted into steel soil tins and were designated to receive orthogonally grouped treat- ments of heat (70°C—120°C) or no heat, light or dark, charate or no charate, and gibberellic acid (GA3) or no gibberellic acid (GA3). Seeds were dry heated in the open steel tins using a forced convection oven set at 70°C for 1 hr, 100°C for 5 min, and 120°C for 5 min to mimic fire conditions observed by Sweeney (1956) and recommended in Keeley (1987). A control treatment was also designated and received no heating. Immediately following heat treatment we removed seeds from the tins and placed them in 50 ml centrifuge tubes (BD Biosciences, MA). Sixteen of the 32 centrifuge tubes were designated for the GA; treatments. GA; treated seeds were imbibed with a mixture of 20 ml of 1 mM CaCl, solution and 20 ml of 100 ppm GA, solution. We designated a control treatment for the remaining 16 centrifuge tubes to receive 40 ml of CaCl, only. Seeds were soaked in solutions at laboratory temperature (22°C—25°C) for 3 hr. Seeds that were designated for dark propagation treatment were housed in centrifuge tubes wrapped in aluminum foil to prevent light exposure. Previous greenhouse trials using a sterile pre- moistened soil mixture (4 parts peat, 2 parts perlite, and 2 parts vermiculite) resulted in extremely poor seed germination response across all treatments and this prompted the adoption of Petri dish propagation techniques. Each group of 150 treated seeds were sown into 32 plastic Petri dishes (150 mm X 25 mm) containing two sheets of 150 mm #1 filter paper (Whatman Interna- tional Ltd., Maidstone, England). Filter paper was pre-moistened with 1 mM CaCl, solution and any standing solution was removed. Petri dishes were covered with their lids and sealed in re-sealable plastic food storage bags to decrease moisture loss. All Petri dishes were cold stratified for 1 month in an unlit refrigerator at 5°C—10°C. For seeds receiving charate treatment, | g of powdered charred wood was applied evenly on top of Petri dish filter paper prior to pre- moistening. Charate was made by charring fresh cut A. fasciculatum stems in a steel burn barrel with a propane torch. Once the stems appeared charred, but not completely reduced to ash, we extinguished the fire by covering the barrel with a lid. Woody charred stems (5-10 mm diameter) were removed and pulverized in a SPEX mill (SPEX CertiPrep, Metuchen, NJ) to produce a fine charate powder. 80 MADRONO Analysis of preliminary germination trials had determined that cold treatment improved mean percent germination by 6% in E. fasciculata (t = 3.530, df = 4, P = 0.024) (Detka 2007). During preliminary trials, we had noted fungal invasion in both the cold treatment and control. This prompted the testing of a potential pre-sowing disinfection treatment. Results of disinfection solution testing suggested that the solution was effective in reducing fungal invasion, but at a significant cost to seed germination (Detka 2007). Therefore, we did not use disinfection treatments in any future germination trials. Following cold stratification, Petri dishes were placed on an indoor Juliana grow rack (ACF Greenhouses, Buffalo Junction, VA) at laborato- ry temperature (22°C—25°C) and out of direct sunlight. We incubated seeds receiving dark treatment on the grow rack shelves in cardboard boxes with removable lids. Ventilation holes were placed on the backside of the boxes to allow sufficient air flow. Seeds undergoing light treat- ment were placed under 40w fluorescent bulbs (GRO-LUX Wide spectrum, Sylvania LTD., Danvers, MA) under low light (approximately 70-100 umol s'' m ’) for a 13-h photoperiod, as recommended in Comstock et al. (1989). We surveyed Petri dishes every two days to count germinated seeds and remoisten filter paper with DI water if necessary. Germination was scored based on the first observation of radicle emergence. All dark treatment dishes were surveyed under indirect green light. Monitoring continued for 30 days. We based the monitoring time period on previous growth trial observations that suggested a peak in seedling emergence approximately 10-14 days following removal from cold stratification and a rapid decline in germination thereafter. Data Analysis We used two-way ANOVA to determine if differences were evident between the observed proportions of viable seed in 2005 and 2006 seed stock. We used multi-way ANOVA to compare the proportion of seedlings emerging within and between the different propagation treatments. We used SYSTAT v. 10.0 (SYSTAT, San Jose, CA) for all statistical analyses. Levene’s test was used to test for homogeneity of variances and the assumption of normality was examined with probability plots of the residuals. RESULTS Seed Viability Testing Results of seed dissection and TZ staining indicated that approximately 10% of Ericameria fasciculata seeds were viable. In both the 2005 [Vol. 57 and 2006 seed stock, empty and dead seeds were more prevalent than viable seed (Fig. 1). There was no significant difference in the proportion of empty, viable, and dead seed condition between the 2005 and 2006 seed stock (F3.}. = 0.514, P = 0.611). Greenhouse Germination Trials The addition of GA3 CE ie = 0.269, P= 0.606) and light stimulus (F; 64 = 1.261, P = 0.266) had no significant impact on E. fasciculata germina- tion (Table 1). The use of charate had a deleterious effect on germination (F; 64 = 48.963, P < 0.001) resulting in mean germination responses less than 1% in all cases (Table 1). The interaction of charate and higher temperature treatments (>70°C) had a particularly lethal effect on E. fasciculata seed germination (F364 = 18.619, P < 0.001) (Ta- ble 1). No other significant interactions between main effects were evident. Heat treatments as a main effect had a significant effect on the germination of E. fasciculata (F364 = 23.147, P < 0.001). Post hoc tests suggested that there is a significant difference in percent germination be- tween seed experiencing lower temperature heat treatments (Control and 70°C) compared to higher temperature (100°C and 120°C) heat treatments (P < 0.001). Higher temperature treatments (>70°C) had catastrophic effects re- sulting in the near elimination (99%) of germina- tion response. The highest rates of germination occurred in seeds that received no heat treatments (control) and no charate (Table 1). Comparison of mean percent germination of seeds receiving no heating and 70°C heat treatment indicated a mean reduction in germination response of 35% with the addition of the 70°C heat treatment. DISCUSSION Ericameria fasciculata is found in fire-prone plant communities (Griffin 1978) and yet fire- related germination cues appear to have predom- inantly negative effects on seed germination. Dry heating conditions similar in temperature to those associated with near-surface burial resulted in marked decreases in germination at temperatures greater than 70°C. In addition, the presence of | charate had a particularly deleterious effect on germination. Interestingly, the presence of light had little or no impact on seed germination. This low tolerance for heating and charate and unresponsiveness to light suggests that EF. fasci- culata seed (1) existing prior to fall burning 1s largely destroyed during fall burns, (2) buried | relatively deeper in the near soil surface (>1 cm) may be able to endure exposure from low burn © intensities and germinate without exposure to light, and (3) dispersal and subsequent coloniza- 2010] 80 70 60 Seed Viability (%) & N S S eS) =) 2005 Fic. 1. Seed Stock (Year) DETKA AND LAMBRECHT: FIRE AND GERMINATION OF E. FASCICULATA 8] O Empty O Viable ® Dead 2006 Mean percent seed viability from 2005 and 2006 seed stocks. Viability percentage is based on results of seed dissection and 2,3,5-triphenyl-tetrazolium chloride (TZ) staining. Mean percentage is based on average of replicate trials (n = 3) for each seed stock. Error bars indicate SEM. tion may occur from adjacent unburned or low intensity burned sites. The observed germination responses to fire- related conditions are not uncommon in chapar- ral species known to utilize an obligate resprout- ing post-fire strategy and may indicate preferenc- es for niches in the mosaic of post-burn environmental conditions and trade-offs associ- ated with resprouting (Keeley and Zedler 1978; Keeley 1987; Baskin and Baskin 2001; Boyd 2007). For example, Keeley (1987) found that the seeds of Haplopappus squarrosus (Hook. & Arn.) Greene responded negatively to heat treatments, but charate presence and available light had no associated effect on germination. Prior to this finding, Keeley and Keeley (1984) also found that H. squarrosus was capable of vigorously resprout- TABLE 1. ing in the first-year following burns. We propose that seeds of E. fasciculata may demonstrate a similar post-fire seedling establishment strategy to H. squarrosus by occupying a niche in the low burn intensity environment. In this environment, access to light and charate would be less available due to burial depth and the increased likelihood of unburned surviving aboveground vegetation cover. In low intensity burns, mature E. fascicu- lata were more apt to vigorously crown resprout and flower (Detka 2007), increasing seed avail- ability for dispersal into areas containing little charate and more intact vegetation cover. Trends in post-burn germination of Ericameria ericoides (Less.) Jeps. may also support the observed fire-related germination responses of E. fasciculata seed. Ericameria ericoides 1s closely MEAN PERCENTAGE GERMINATION OF E. FASCICULATA IN RESPONSE TO ORTHOGONALLY GROUPED TREATMENTS OF GA3, LIGHT OR DARK, HEAT TREATMENTS, AND CHARATE. Each mean value is based on (n = 3) Petri dishes each containing 150 seeds. Temperature treatments sharing the same superscript letter were not significantly different (P > 0.05 from Bonferroni post hoc test). Standard error (SE) values are reported in parentheses. In all cases charate and non-charate treatments were significantly different. Significance values for the remaining main effects in the multi-way ANOVA were not significant (P > 0.05). Light Dark 100°C 120°C 120°C Control* 70°C 1 hr®- ~5 min” 5 min” Control? 710°C 1 he 100°C 5 min” 5 min? GA, Control 5.78 (1.11) 2.89 (0.59) a — 5.33 (1.15) 3.78 (0.44) — = Charate 0.22 (0.22) 0.44 (0.44) = —_ 0.44 (0.44) 0.22 (0.22) 0.22 (0.22) =— Control (no GA;3) Control 4.44 (1.18) 3.78 (0.44) — — 8.22 (4.26) 4.44 (1.74) — — Charate 0.22 (0.22) 0.22 (0.22) — — 0.67 (0.01) 0.44 (0.44) — —- 62 MADRONO related to E. fasciculata (Roberts and Urbatsch 2003) and resides in the same habitat range. Further experimental comparisons between the common E. ericoides and the rare E. fasciculata may serve to elucidate differences in post-fire recovery performance leading to patterns of rarity in E. fasciculata. Holl et al. (2000) observed high rates of germination in E. ericoides following surface burn treatments using fresh Adenostoma fasciculatum stems, which may initially seem contrary to trends observed in E. fasciculata. Observed germination of E. ericoides seed fol- lowing surface burn treatments may have been associated with burial depths that were deep enough to protect seed from high temperature exposure (e.g., >70°C) (Holl et al. 2000). In addition, the proposed associated toxicity of allelopathic chemicals present in A. fasciculatum charate may have been volatized as stems were reduced entirely to ash. The extremely low germination rates are apparently due to the complete absence of embryonic tissues in a large proportion of achenes. The lack of embryonic tissue in other- wise intact achenes has been frequently observed in the Asteraceae (Keeley 1987; Padgett et al. 1999; Meyer and Carlson 2001; Alkio and Grimm 2003; Ransom Seed Laboratory 2006). Previous studies have proposed that the high frequency of empty achenes is the result of increased seed abortion due to self-pollination or pollination among closely related plants (Connor and Hall 1997) or the result of variation in resource availability (Sobrevila 1989). Padgett et al. (1999) has suggested that reduced seed set reflects an adaptive mechanism to deter herbivores by hiding viable achenes among empty ones (Con- nor and Hall 1997). In this study, we did not test specific mechanisms for the observed low seed, but the implications are significant for restoring this rare species. Active restoration of E. fasciculata into areas of suitable habitat may be required to insure the conservation of the species. We recommend that active restoration efforts include wildland seed collection and viability testing in advance of prescribed burning events. Ripe capitula should be collected for site specific propagation from local populations to increase the likelihood of preserving the genetic integrity of the species and insure that seed stock is representative of healthy individuals best adapted to localized conditions (Pavlik 1996). Wildland seed collection of rare species in this fashion should also set limits on the extent of seed collection from donor plants and adopt measures to reduce the risk of decimating available wildland seed stocks from established populations (Guerrant 1996). Ericameria fasciculata propagation should in- clude preliminary tetrazolium staining assess- ments of seed stock viability. Tetrazolium stain- [Vol. 57 ing techniques can be employed quickly and inexpensively as a means of estimating germina- tion potential and the amount of seed needed to produce the projected number of seedlings for restoration efforts. In this study, estimates of germination potential using tetrazolium staining were slightly higher (9-10%) than the highest observed germination in Petri dish propagation trials (6-8%). Two explanations can account for the overestimation of seed viability using this technique. First, seeds were not rejected if they had any indication of pink to red staining along portions of the embryo. This approach increased the speed of assessment but may have reduced the accuracy by not accounting for those seeds that were experiencing the late stages of gradual tissue die-off (Lakon 1949; Grooms 2006). Secondly, bacteria and fungi can result in a surface staining of seeds. All embryos were inspected by surface scraping and sectioning to insure that staining was complete, but it is still plausible that advanced fungal invasions could have yielded a false positive reading (Lakon 1949; Gutormson 2005). Caution should be used in interpreting germination potential from Petri dish propaga- tion as several dishes experienced fungal coloni- zation that may have reduced germination. Keeley (1987) proposed that some of the fungal invasion that he observed in similar germination trials of woody chaparral shrubs may be attrib- uted to the lack of fungal resistance by empty or inviable seed. Due to the large number of potentially empty E. fasciculata seed, special care should be exercised to visually evaluate and discard seeds that appear to have signs of fungal invasion or physical defects. In addition, further propagation research should be conducted to establish if conventional vermiculite sowing techniques yield higher rates of emergence in E. fasciculata. ACKNOWLEDGMENTS Support for this work partially funded by The Arthur and Karin Nelson Foundation. The authors thank Lars Pierce, Bruce Delgado, Diane Steeck, Suzanne Worcester, Rodney Myatt, Cynthia Detka, Kali Detka, Danielle Tannourji, Leonardo Hernandez, and Ingrid Graeve. LITERATURE CITED ALKIO, M. AND E. GRIMM. 2003. Vascular connections | between the receptacle and empty achenes in sunflower (Helianthus annuus L.). Journal of Experimental Botany 54:345—348. BASKIN, C. C. AND J. M. BASKIN. 2001. Seeds: ecology, | biogeography, and evolution of dormancy and | germination, Ist ed. Academic Press, New York, | NY. Boyb, R. S. 2007. Response to fire of Ceanothus | roderikii (Rhamnaceae), a federally endangered California endemic shrub. Madronfio 54:13-21. 2010] CALIFORNIA NATURAL DIVERSITY DATABASE (CNDDB). 2002. RareFind 2 ©. California De- partment of Fish and Game, Sacramento, CA. CHRISTENSEN, N. L. AND C. H. MULLER. 1975a. Effects of fire on factors controlling plant growth in Adenostoma chaparral. Ecological Monographs 45:29-55. AND . 1975b. Relative importance of factors controlling germination and seedling sur- vival in Adenostoma chaparral. American Midland Naturalist 93:71—78. CoMSTOCK, J. P., L. DONOVAN, AND J. EHLERINGER. 1989. The role of photoperiod in determining seasonal patterns of photosynthetic activity in chaparral shrubs. Pp. 91-96 in S. Keeley (ed.), The California chaparral: paradigms reexamined. Los Angeles Museum of Natural History, Los Angeles, CA. CONNOR, D. J. AND A. J. HALL. 1997. Sunflower physiology. Pp. 113-182 in A. A. Schneiter (ed.), Sunflower technology and production. Monograph No. 35. American Society of Agronomy, Madison, WI. DAvis, F. W., M. I. BORCHERT, AND D. C. ODION. 1989. Establishment of microscale vegetation pat- tern in maritime chaparral after fire. Vegetatio 84:53-67. DAvis, M. B. 1989. Lags in vegetation response to greenhouse warming. Climate Change 15:75— 82. DETKA, J. R. 2007. Effects of fire on rare Ericameria fasciculata (Asteraceae). M.S. thesis. San José State University, San José, CA. EMERY, D. E. 1988. Seed propagation of native California plants. Santa Barbara Botanic Garden, Santa Barbara, CA. GORDON-REEDY, P. AND O. MISTRETTA. 1997. En- dangered slender-horned spineflower: germination and propagation studies reported (California). Restoration and Management Notes 15:91. GRIFFIN, J. R. 1976. Native plant reserves at Fort Ord. Fremontia 4:25-—28. . 1978. Maritime chaparral and endemic shrubs of the Monterey Bay region, California. Madrono 25:65—-112. GROoMS, L. 2006. The pros and cons of TZ testing. Seed Today 10:58—60. GUERRANT, E. O., Jr. 1996. Designing populations: demographics, genetics, and horticultural dimen- sions. Pp. 171—208 in D. A. Falk, C. I. Millar, and M. Olwell (eds.), Restoring diversity: strategies for reintroduction of endangered plants. Island Press, Washington, D.C. GUTORMSON, T. 2005. Tetrazolium, germination, dormancy and fungal interactions - much to do about molds. Seed World 143:10—12. HICKMAN, J. C. (ed.). 1993. The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. HOLL, K. D., H. N. STEELE, M. H. FUSARI, AND L. R. Fox. 2000. Seed banks of maritime chaparral and abandoned roads: potential for vegetation recov- ery. Journal of the Torrey Botanical Society 127:207-220. KEELEY, J. E. 1987. Role of fire in seed germination of woody taxa in California chaparral. Ecology 68:434-443. DETKA AND LAMBRECHT: FIRE AND GERMINATION OF E. FASCICULATA 83 . 2006. Fire severity and plant age in postfire resprouting of woody plants in sage scrub and chaparral. Madrono 53:373—379. AND S. C. KEELEY. 1984. Postfire recovery of California coastal sage scrub. American Midland Naturalist 111:105—117. ; B. A. MORTON, A. PEDROSA, AND P. TROTTER. 1985. Role of allelopathy, heat and charred wood in the germination of chaparral herbs and suffrutescents. Journal of Ecology 73:445-458. KEELEY, S. C., J. E. KEELEY, S. M. HUTCHINSON, AND A. W. JOHNSON. 1981. Postfire succession of the herbaceous flora in southern California chaparral. Ecology 62:1608—1621. AND M. PIZZORNO. 1986. Charred wood stimulated germination of two fire following herbs of the California chaparral and the role of hemicellulose. American Journal of Botany 73:1289-1297. AND P. H. ZEDLER. 1978. Reproduction of chaparral shrubs after fire: a comparison of sprouting and seeding strategies. American Mid- land Naturalist 99:142—161. LAKON, G. 1949. The topographical tetrazolium method for determining the germinating capacity of seed. Plant Physiology 3:389-394. MATTHEWS, M. A. 1997. An illustrated field key to the flowering plants of Monterey County and ferns, fern allies, and conifers. California Native Plant Society, Sacramento, CA. McPHERSON, J. K. AND C. H. MULLER. 1969. Allelopathic effects of Adenostoma_ fasciculatum, ‘““Chamise”’, in the California chaparral. Ecological Monographs 39:177—-198. MEYER, S. E. AND S. L. CARLSON. 2001. Achene mass variation in Ericameria nauseosus (Asteraceae) in relation to dispersal ability and seedling fitness. Functional Ecology 15:274-281. Moreno, J. M. AND W. C. OECHEL. 1991. Fire intensity effects on germination of shrubs and herbs in southern California chaparral. Ecology 72:1993—2004. OpIoNn, D. C. AND F. W. DAvis. 2000. Fire, soil heating, and the formation of vegetation patterns in chaparral. Ecological Monographs 70:149—169. PADGETT, P. E., L. VAZQUEZ, AND E. B. ALLEN. 1999. Seed viability and germination behavior of the desert shrub Encelia farinosa Torrey and A. Gray (Compositae). Madrono 46:126—133. PAVLIK, B. M. 1996. Defining and measuring success. Pp. 127-156 in D. A. Falk, C. I. Millar, and M. Olwell (eds.), Restoring diversity: strategies for reintroduction of endangered plants. Island Press, Washington D.C. Pons, T. L. 2000. Seed responses to light. Pp. 237-260 in M. Fenner (ed.), Seeds: the ecology of regener- ation in plant communities, 2nd ed. CAB Interna- tional, Wallingford, U.K. PRIMACK, R. B. 1996. Lessons from ecological theory: dispersal, establishment, and population structure. Pp. 202-233 in D. A. Falk, C. I. Millar, and M. Olwell (eds.), Restoring diversity: strategies for reintroduction of endangered plants. Island Press, Washington D.C. AND S. L. MIAO. 1992. Dispersal can limit local plant distribution. Conservation Biology 6:513— 519: 84 MADRONO RANSOM SEED LABORATORY, INC. 2006. Knowledge- base of native seeds — Asteraceae. Website http:// www.ransomseedlab.com/family/asteraceae.htm [accessed 22 June 2010]. ROBERTS, R. P. AND L. E. URBATSH. 2003. Molecular phylogeny of Ericameria (Asteraceae, Astereae) based on nuclear ribosomal 3’ ETS and ITS sequence data. Taxon 52:209—228. SKINNER, M. W. AND B. M. PAVLIK. 1994. California Native Plant Society’s inventory of rare and endangered vascular plants of California. Califor- nia Native Plant Society, Sacramento, CA. SOBREVILA, C. 1989. Effects of pollen donors on seed formation in Espeletia schultzii (Compositae) pop- ulations at different altitudes. Plant Systematics and Evolution 166:45—67. SWANK, S. E. AND W. C. OECHEL. 1991. Interactions among the effects of herbivory, competition, and resource limitation on chaparral herbs. Ecology 72:104—-115. SWEENEY, J. R. 1956. Responses of vegetation to fire. A study of the herbaceous vegetation following chaparral fires. University of California Publica- tions in Botany 28:143—216. THANOS, C. A. AND P. W. RUNDEL. 1995. Fire-followers in chaparral: nitrogenous compounds trigger seed germination. Journal of Ecology 83:207—216. TUELLER, P. E. AND D. E. PAYNE. 1987. The ecology and management of the genus Chrysothamnus. Pp. 1-8 in K. L. Johnson (ed.), Proceedings, 4th Utah shrub ecology workshop: the genus Chry- [Vol. 57 sothamnus; 1986 September 17-18. Utah State University, College of Natural Resources, Cedar City, UT. TYLER, C. M. 1995. Factors contributing to postfire seedling establishment in chaparral: direct and indirect effects of fire. Journal of Ecology 83:1009—1020. . 1996. Relative importance of factors contrib- uting to postfire seedling establishment in maritime chaparral. Ecology 77:2182—2195. U.S. ARMY CORPS OF ENGINEERS, SACRAMENTO DISTRICT. 1997. Installation-wide multispecies hab- itat management plan for former Fort Ord, Califor- nia, April. Sacramento, CA. Website http://www. fortordcleanup.com/adminrec/ar_pdfs/AR-BW-1787/ [accessed 21 June 2010]. VAN DYKE, E. AND K. HOLL. 2003. Mapping the distribution of maritime chaparral species in the Monterey Bay area. Report for the U.S. Fish and Wildlife Service, Ventura Fish and Wildlife Office, CA. Website http://www.elkhornsloughctp.org/ uploads/1123621594Van%20dyke%20holl%20dist %20o0f%20mar%20chap.pdf [accessed 21 June 2010]. WENT, F. W., G. JUHREN, AND M. JUHREN. 1952. Fire and biotic factors affecting germination. Ecology 33:351-364. WILLSON, M. AND A. TRAVESET. 2000. The ecology of seed dispersal. Pp. 85—110 in M. Fenner (ed.), Seeds: the ecology of regeneration in plant communities, 2nd ed. CAB International, Wallingford, U.K. MADRONO, Vol. 57, No. 2, pp. 85—94, 2010 DISTRIBUTION AND COMMUNITY ASSOCIATIONS OF CAPE IVY (DELAIREA ODORATA) IN CALIFORNIA RAMONA ROBISON ICF International, 630 K Street, Suite 400, Sacramento, CA 95814 RROBISON@ICFI.COM JOSEPH M. DIrTOMASO University of California, Davis, CA 95616 ABSTRACT Cape ivy (Delairea odorata Lem.) was found to occur throughout coastal California and southern Oregon. It was most abundant in urbanized coastal areas such as the San Francisco Bay, and Santa Cruz, Monterey, San Luis Obispo, Santa Barbara and Los Angeles counties. Field observations showed Cape ivy to occur in seven different broad community types, including both riparian and non- riparian areas. Of the two morphological forms, the exstipulate type occurred more frequently at the northern and southern ends of the distribution, and the stipulate type was more common in the middle of the distribution range, from southern Humboldt County to Los Angeles County. Only 21 locations were found that supported both stipulate and exstipulate plants, and they were most often located in urbanized coastal areas. Analysis with GIS determined the elevation, temperature and precipitation ranges that Cape ivy occupies in California. The analysis indicated that Cape ivy occurs at elevations between 0 and 891 meters, annual mean temperatures between 10.5 and 17.7°C, and in areas with annual precipitation ranging between 232 and 2270 mm. An overlay analysis of Cape ivy locations using GIS was also compared with the California Natural Diversity Database sensitive species location information to determine which species might be threatened by Cape ivy expansion. Three sensitive animals and five sensitive plants were expected to have >40% of their occurrences with a 500 m buffer to Cape ivy infestations. Key Words: German Ivy, invasive, Senecio mikanioides, South Africa, weed. INTRODUCTION Cape ivy (Delairea odorata Lem., syn. Senecio mikanioides Walp.) (Asteraceae) is native to South Africa, but has escaped cultivation to invade wildlands in Europe, Australia, New Zealand, Hawaii, and South America, as well as coastal regions of western North America (Parodi 1959; Abrams and Ferris 1960; Palhinha 1974; Zangheri 1976; Pignatti 1982; Haselwood and Motter 1983; Hirano 1983; Webb et al. 1988: Fagg 1989; Jacobi and Warshauer 1992; Scott and Delfosse 1992; Hickman 1993; Gallo 2000; DiTomaso and Healy 2007). It was first collected in California in 1892 (F. T. Bioletti s.n. UC36003), and since that time has spread to all coastal counties and many adjacent inland sites (Jepson 1951; Abrams and Ferris 1960; Thomas 1961; Hoover 1970; Howell 1970; Munz 1974; Smith 1976; Beauchamp 1986; Smith and Wheeler 1992; Hickman 1993; Best et al. 1996; Junak et al. 1995: Matthews 1997). Cape ivy’s spread into undis- turbed wildland areas of California is of great concern, particularly because effective control is difficult (Alvarez 1997; Bossard et al. 2000). Although Cape ivy has firmly established itself along California’s coast, the question remains as to whether it has occupied the full extent of its potential range. To date, only one California study has documented the community types invaded by Cape ivy within the state (Alvarez and Cushman 2002). Alvarez and Cushman (2002) compared the effect of Cape ivy on invaded and un-invaded coastal scrub, and willow and alder riparian communities. Their results showed that plots invaded by Cape ivy had a 31, 88, and 92% decrease in species diversity, abundance of native seedlings, and non-native seedlings, respectively, compared to uninvaded sites of the same community types. In this study, we provide a map of the current distribution, as well as more detailed information on the plant community types occupied by Cape ivy. In addition, we identify the distribution of the two morphological forms of Cape ivy in California, a stipulate and exstipulate type. The results provided here will help identify threatened community types or sensitive species in proximity to Cape ivy infestations. MATERIALS AND METHODS Mapping The California Exotic Pest Plant Council (now known as the California Invasive Plant Council, Cal-IPC) Cape Ivy Working Group began 86 MADRONO collecting Cape ivy distribution data in 1995. In May 1995, the distribution of Cape ivy was mapped along streams and hillsides in the coastal region of California, south of Monterey Co. (no vouchers taken). Additionally, appropriate hab- itats such as lakes, campgrounds, and parks along the coast were also surveyed. All popula- tions that were reported by Cal-IPC members, California Native Plant Society (CNPS) mem- bers, park rangers, and other concerned citizens were visited, confirmed, and described for future analysis. The boundaries of the populations were estimated and drawn on maps. The data collected were then digitized as point data on 1:100,000 topographic base maps using MapInfo Profes- sional 5.0 (LizardTech, Seattle, WA). Additional areas were surveyed in 1999, including coastal counties north of Monterey and the San Francisco Bay Area. In addition to collecting maps from field experts, data were collected using a hand-held Trimble GeoExplorer II GPS (Global Positioning System) unit (Trim- ble, Sunnyvale, CA) with an overall corrected accuracy of 1 to 3 m. A series of sites originally mapped in 1995 by Cal-IPC were re-surveyed for Cape ivy in 2000, with 95% of these locations still supporting the invasive species. Several individuals and organizations provided large Cape ivy distribution data sets that were incorporated into our database. Most notable was an extensive set of maps provided by Golden Gate National Recreation Area (GGNRA) em- ployees. These maps included data from Marin, San Francisco and San Mateo counties in the form of ArcView shapefiles. Electronic data were also provided for Pt. Reyes National Seashore by the National Park Service, Catalina Island by the Catalina Island Conservancy, and Contra Costa Co. by the Contra Costa Watershed Forum. The other data collected were on paper maps obtained from 12 sources ranging from Oregon to San Luis Obispo. These were digitized onto 1:100,000 scale topographic maps. Other mapping points were provided for a number of counties from Del Norte to San Diego. All the spatial data were brought into a GIS (Geographic Information System). In 1999, MapInfo Professional 5.0 was used to create maps, as well as store and edit the data. The GPS data was exported from Pathfinder to a MapInfo format, and ArcView shapefiles were converted to MapInfo format and included in the GIS. Some of the data provided by GGNRA were polygon or line data and these were converted to point data for the final analysis. In 2000, vegetative community types and stipulate or exstipulate morphological forms of Cape ivy were also recorded using GPS. The California Natural Diversity Data Base (CNDDB) (California Fish and Game, http:// www.dfg.ca.gov/biogeodata/cnddb/) community [Vol. 57 type which classifies vegetation using a five- number land cover code was chosen for the mapping analysis, and field data was collected using the number code. From 2001 to 2004, data were collected with a Garmin eTrex Vista GPS (Garmin Ltd., Olathe, KS) with accuracy of 15 meters alone and <3 m with the Wide Area Augmentation System (WAAS) enabled. Way- points collected with the eTrex Vista were converted to ArcView shapefiles with Waypoint+ version 1.8.03. After conversion, the data files were edited to contain attribute fields listed in Table 1. Maps presented here are in the Teale- Albers projection, geographic coordinate system NAD 1927. GIS Analysis BIOCLIM Raster Extraction. GIS analysis was performed with ArcView version 9.0 (ESRI, Redlands, CA) and the Spatial Analyst extension (version 9.0). Polygon data collected in the distribution mapping phase were converted to points and 1465 Cape ivy location points were used as the basis for GIS analysis. In order to determine the elevation and climate parameters associated with the distribution data set in California, the point data was joined with BIOCLIM raster datasets. The bioclimatic vari- ables (BIOCLIM) raster layers were derived from WorldClim interpolated climate layers (http:// www.worldclim.org/methods). The WorldClhim climate layers contain precipitation records for 47,554 locations, mean temperature from 24,542 locations, and minimum and maximum temper- ature for 14,835 locations (Hijmans et al. 2004). WorldClim altitude was obtained from the Shuttle Radar Topography Mission (SRTM) Digital Elevation Models (http://www2.jpl.nasa. gov/srtm/). Grids used in the analysis were at ! 30 seconds (1 km). A spatial join of Cape ivy © point data and BIOCLIM rasters was accom- plished with the ArcView Spatial Analyst “‘ex- tract values to points” tool. For example, when the BIOCLIM annual precipitation raster data set was spatially joined to the Cape ivy point data a column with annual precipitation was generat- ed in the attribute table. This was repeated for all | the raster layers. Excel (Microsoft Corp., Red- | mond, WA) and JMP IN (version 5.1) (SAS | Institute Inc., Cary, NC) were then used to | determine the range and mean values for the | raster layers. CNDDB Sensitive Species Overlay. Overlay | analysis was performed with the Cape ivy point | data and the California Natural Diversity Data- | base (CNDDB) sensitive species location data. | The data are available within an application | called RareFind, a Windows based program | developed by the California Department of Fish ROBISON AND DITOMASO: CALIFORNIA CAPE IVY DISTRIBUTION 87 2010] TABLE 1. ATTRIBUTES OF CAPE Ivy USED IN FINAL MAPPING SHAPEFILES. Field name Description SHAPE all points SITECODE map identification point, using county abbreviation and number COUNTY county GPS true or false VISITED date of GPS SURVEYOR surveyor or source of data ENTEREDBY person digitized by DATAFILE name of rover file for GPS data or original shapefile name SCL NAME Delairea odorata COMMENT source of data, location, directions, etc. GPS CODE waypoint code for eTrex Vista data VEGTYPE Holland (1986) numerical code used by CAGAP (Davis et al. 1998) ST_NS either stipulate (ST), exstipulate (ES) or both (STES) VIABLE viable seeds present, Yes or No LAT generated with the ‘“‘add XY” tool in ArcView 9.0 LONG generated with the “‘add XY” tool in ArcView 9.0 and Game, Sacramento, CA (http://www.dfg.ca. gov/biogeodata/cnddb/rarefind.asp) and designed to perform queries and produce reports. Rare- Find comes with GIS layers, which were used for this analysis (RareFind version 3.0.5 dated September 2, 2005). The CNDDB data consists of locations for sensitive plants, animals and natural communities as well as population data voluntarily submitted by field biologists. Sensitive species are defined as federally and state listed plants and animals, all species that are candidates for listing, all species of special concern and those species that are considered sensitive by govern- ment agencies and conservation organizations (http://www.dfg.ca.gov/whdab/pdfs/cnddbfaq. pdf). The data were then reviewed for accuracy and mapped by CNDDB personnel as “‘occur- rences”’ at various levels of precision, from specific points to non-specific buffered polygons. For the CNDDB GIS analysis, Cape ivy points were buffered out 100 m to represent the current extent of their direct or indirect influence on sensitive species locations. The “‘select by loca- tion” feature in ArcView was used to select the sensitive species occurrences, which overlapped with the 100 m buffered points. The selected polygons from the CNDDB data were then saved into a separate shapefile. Another file was created with Cape ivy points buffered out to 500 m, representing an estimate of future spread, while another shapefile with sensitive species occur- rences was generated for comparison. RESULTS AND DISCUSSION Mapping California and Oregon Cape Ivy Distribution. Cape ivy has been known to occur in California since 1892, yet many of the historic floras only mention it in passing and do not indicate it as a widespread weed (Munz 1974; Smith 1976; Beauchamp 1986; Junak et al. 1995). In the 1970’s it was noted as “‘climbing on trees, mostly willows, along coastal streams,” and “‘forming dense tangles in shaded canyons or on moist open slopes”? (Hoover 1970; Howell 1970). Floras from the 1990’s noted that it is common or invasive in coastal areas (Best et al. 1996; Matthews 1997). Surprisingly, as late as 1992 the Mendocino Flora states that it 1s “occasional but seldom collected” (Smith and Wheeler 1992). In fact, there are no voucher records of Cape ivy in Mendocino County in the Consortium of California Herbaria (http://ucjeps.berkeley.edu/consortium/) prior to 2001, despite it widespread occurrence there today. Based on the field survey, Cape ivy occurs throughout all coastal counties of California, as well as the Channel Islands (Santa Rosa, Santa Cruz and Santa Catalina) and Curry Co., Oregon (Figs. 1-3). Furthermore, it was also found in most of the major river systems along the coast. Although the vast majority of Cape ivy infesta- tions were found within a few kilometers of the coast, populations occurred 60 to 70 km inland in Contra Costa and Los Angeles counties. Interestingly, in its native range in South Africa, nearly all collections of Cape ivy have been reported to be the stipulate form (Balciunas and Smith 2006; Robinson 2006). In California, however, the exstipulate form is far more commonly encountered than in its native range (Fig. 4). Although the exstipulate form is found throughout California, it is the primary morpho- logical type in the northern extent of its range, including Curry Co., Oregon, and northern Humboldt Co., as well as the southern range of its distribution in Los Angeles and San Diego counties. The stipulate forms were most widespread throughout the center of the range of the species, from Mendocino Co. to Santa Barbara Co. 88 MADRONO [Vol. 57 Fic. |. (Fig. 4). A combination of the two morpholog- ical forms was most common in heavily populat- ed areas, particularly the San Francisco Bay region and San Luis Obispo Co. GIS Analysis The climate where Cape ivy grows in Califor- nia can be broadly described as Mediterranean. Cape ivy locations in northern California and southern Oregon. Mediterranean climates are characterized by dry summers and an average of 25 to 100 cm annual rainfall concentrated during the mild winter months (Dallman 1998). Snow is infrequent except at higher elevations, and the amount of winter rain is highly variable from year to year. BIOCLIM Raster Extraction. BLIOCLIM Ras- ter Extraction analysis indicates that Cape ivy in Fic. 2. Cape ivy locations in the San Francisco Bay Area and central California. 2010] ROBISON AND DITOMASO: CALIFORNIA CAPE IVY DISTRIBUTION 89 FIG. 3. Cape ivy locations in southern California. California occurs at elevations between O and 891 meters, annual mean temperatures between 10.5 and 17.7°C, and in areas with annual precipitation ranging between 232 and 2270 mm (Table 2). Examining the maximum temperature of the warmest month and the minimum temper- ature of the coldest month, the results suggest that Cape ivy can tolerate temperatures between 8: and 31.8°C. Vegetation community types. From the field GPS data, Cape ivy was most often observed in urban or agricultural areas (Table 3). This was expected, as Cape ivy was introduced as a horticultural plant and many of the surveys were conducted in easily accessible urban areas. Invasive populations were also common in riparian and non-native Eucalyptus forests, oak woodlands, and coastal scrub communities. In contrast, only two Cape ivy populations were observed in coniferous forests and only one occurred in a salt marsh. CNDDB sensitive species overlay. Using either a 100 or 500 m buffer, we determined the number of CNDDB sensitive species overlapping with community types known to be invaded by (Table 4). For example, 163 sensitive vascular plants were expected to overlap with predicted Cape ivy sites using a 100 m buffer around the infested location, whereas 211 sensitive species overlapped the expected Cape ivy infested areas as predicted by the 500 m buffer. Each Cape ivy infestation was predicted to overlap with a mean of 2.2 sensitive vascular plant species at a 100 m buffer, and 2.8 sensitive plants at a 500 m buffer. The number of predicted sensitive species occurrences per infestation was relatively small using the 100 and 500 m buffer areas. In all cases, except non-vascular plants, the number of overlapped occurrences and the mean number of sensitive species occurrences in Cape ivy sites increased as the buffer size increased. Although most groups only had a few predicted overlaps between sensitive species and Cape ivy infesta- tions, some species within these groups frequently overlapped in their predicted occurrences. Species that had a significant overlap in occurrences using a 100 or 500 m buffer are listed in Table 5. With the 100 m buffer, only animals overlapped with Cape ivy infestations, while the 500 m buffer overlapped both animals and plants. The percent potential overlap between each sensitive species and Cape ivy was calculated by dividing the number of predicted overlapping occurrences (100 and 500 m buffer areas) by the number of total sensitive plant occurrences. Among the sensitive vascular plant species, several showed >40% potential overlap, including the San Francisco Bay spineflower (Chorizanthe cuspidata S. Watson var. cuspidata), Franciscan thistle (Cirsium andrewsii (A. Gray) Jeps.), San Fran- cisco gumplant (Grindelia hirsutula Hook & Arn. var. maritima (Greene) M. A. Lane), perennial goldfields (Lasthenia macrantha (A. Gray) Greene ssp. macrantha) and marsh microseris (Microseris paludosa (Greene) J. T. Howell). These species are expected to be greatly impacted by the expansion of Cape ivy infestations. There was also a considerable overlap between predicted Cape ivy infestations and steelhead 90 MADRONO [Vol. 57 kK Stipulate and Exstipulate @ Exstipuate ©) Stipulate Fic. 4. Distribution of stipulate and exstipulate forms of Cape ivy, including locations where both forms co- | occur. salmon (Oncorhynchus mykiss) populations (Ta- ble 5). Using a 500 m buffer, the percentage overlap between Cape ivy and streams supporting steelhead ranged between 42 and 50%. Although no published studies have been reported on the toxicity of Cape ivy to fish, some evidence (C. | Bossard unpublished data) suggests that Cape ivy is toxic to the golden shiner (Notemigonus | TABLE 2. CAPE Ivy DISTRIBUTION ATTRIBUTES EXTRACTED FROM BIOCLIM RASTER DATA (n = 932 EXCEPT WHERE NOTED). BIOCLIM link: http://www.worldclim.org/methods. ' Quarter = three consecutive months. BIOCLIM variable Elevation (m) n =1057 Annual mean temperature (°C) Maximum temperature of warmest month (°C) Minimum temperature of coldest month (°C) Mean annual precipitation (mm) Precipitation in wettest quarter! (mm) Precipitation in driest quarter (mm) Mean + SE Minimum value Maximum value 66.7 0 891 13.3 + 0.04 10.5 17.7 23.2 + 0.08 19.5 31.8 4.6 + 0.03 1.8 eS) $26 22 11 232 2270 369 + 4.5 101 950 82203 0 72 2010] TABLE 3. codes from Holland (1986). CNDDB community ROBISON AND DITOMASO: CALIFORNIA CAPE IVY DISTRIBUTION 91 RECORDED COMMUNITY TYPES WHERE CAPE IVY WAS OBSERVED, BASED ON GPS DATA. ' Community Number of observations from field data type code! General type Specific type 11100 urban or agriculture urban or built-up land 33 11300 non-native forest Eucalyptus 1] 21310 coastal scrub northern dune scrub l 31100 coastal scrub northern coastal bluff scrub 12 32100 coastal scrub northern (Franciscan) coastal scrub 12 32200 coastal scrub central (Lucian) coastal scrub 5 32300 coastal scrub venturan coastal sage scrub 3 52120 salt marsh southern coastal salt marsh l 61110 riparian forest northern coast black cottonwood riparian 1 forest 61130 riparian forest red alder riparian forest 23 61210 riparian forest central coast cottonwood-sycamore 2 riparian forest 61220 riparian forest central coast live oak riparian forest l 61230 riparian forest central coast arroyo willow riparian forest 28 61310 riparian forest southern coast live oak riparian forest 2 61320 riparian forest southern arroyo willow riparian forest 6 62100 riparian forest sycamore alluvial woodland l 62400 riparian forest southern sycamore-alder riparian 4 woodland 63100 riparian forest northern coast riparian scrub pA 63320 riparian scrub southern willow scrub l 71160 oak woodland coast live oak woodland 13 82320 conifer forest upland redwood forest | 83120 conifer forest Bishop pine forest ] crysoleucus) and crushed Cape ivy leaves caused mortality in mosquito fish (Gambusia affinis) within three days (J. Balciunas unpublished data). However, the latter study used crushed leaves of Cape ivy which may not represent exposure typically found in nature. Because other related species (i1.e., Senecio) are known to contain pyrrolizidine alkaloids (Manske 1936; Adams and Gianturco 1956; Stelljes et al. 1991; Catalano et al. 1996), which can cause liver damage in humans, animals, and fish (Hendricks et al. 1981), the potential toxic effect of Cape ivy on steelhead populations is of concern because of its close proximity to water and its high density in many infested areas. Of the invertebrates co-occurring within pre- dicted Cape ivy populations, only one species, Monarch butterfly (Danus plexippus), showed any significant overlap, with 13 and 25% of its occurrences within the 100 and 500 m buffers, respectively (Table 5). The potential for Cape ivy alkaloids to affect the Monarch butterfly has been studied indirectly, but with no conclusions as to the potential impact. Monarch butterflies were found to have accumulated pyrrolizidine alkaloid after over-wintering in Cape ivy infested areas (Stelljes and Seiber 1990). The butterflies accumulate pyrrolizidine alkaloids after using Cape ivy as a nectar source. Although this was postulated to provide a chemical defense mech- TABLE 4. CNDDB SENSITIVE SPECIES OVERLAP WITH CAPE Ivy SUMMARIZED BY GROUP CLASSIFICATION. Mean number of species per occurrence at 100 m_~ with Cape ivy at Species overlapping with Cape ivy at Mean number of species per occurrence at 500 m Species overlapping Group classification 100 m buffer buffer = SE 500 m buffer buffer + SE Natural communities 24 2 2 03 35 242 04 Non-vascular plants 8 1? 8 = 0:3 Vascular plants 163 22.220 pala 2.8 + 0.2 Invertebrates 32 522415 oe) a ae 2 Fish ‘| 93+ 4.6 9 10.0 + 5.0 Reptiles 7 3.4 + 1.1 9 AS 1 Amphibians 4 48 + 2.8 8 6.9 + 4.6 Birds 20 2.6 + 0.6 28 3.3 2 038 Mammals Is DEEN S 18 2.8 2076 92 MADRONO [Vol. 57 TABLE 5. CNDDB SENSITIVE SPECIES LOCATIONS AND PREDICTED OVERLAP OF CAPE Ivy AND SENSITIVE SPECIES AT EITHER 100 OR 500 M BUFFERS. 'ESU = evolutionarily significant unit. Number (percent) of occurrences overlapping with Wimberor predicted Cape ivy occurrences populations tracked by Using 100 m_ Using 500 m Scientific name Common name CNDDB buffer buffer Animals Rana draytonii California red-legged frog 831 133) 39, (5) Charadrius alexandrinus nivosus western snowy plover 109 13 (12) 22420) Eucyclogobius newberryi tidewater goby M2 28 (25) 41 (37) Oncorhynchus mykiss irideus steelhead—central California 28 13 (46) 14 (50) coast ESU' Oncorhynchus mykiss irideus steelhead—south/central 2) 9 (33) 12 (44) California coast ESU Oncorhynchus mykiss irideus southern steelhead—southern 12 4 (33) 5 (42) California ESU Danaus plexippus monarch butterfly 355 43 (13) $31(25) Arborimus pomo Sonoma tree vole 208 — 10 (5) Actinemys marmorata western pond turtle 302 — 11 (4) Actinemys marmorata pallida southwestern pond turtle 308 — 13 (4) Vascular plants Campanula californica (Kellogg) swamp harebell 100 —- 10 (10) A. Heller Castilleja mendocinensis (Eastw.) Mendocino coast indian 42 — 12 (29) Pennell paintbrush Chorizanthe cuspidata S. Watson San Francisco Bay spineflower 20 — 10 (50) var. cuspidata Cirsium andrewsii (A. Gray) Franciscan thistle 27 — 11 (41) Jeps. Grindelia hirsutula Hook & San Francisco gumplant 15 —- 11 (73) Arn. var. maritima (Greene) M. A. Lane Lasthenia californica subsp. perennial goldfields 32 -- 13 (41) DC. ex Lindl. macrantha (A. Gray) R. Chan Microseris paludosa (Greene) marsh microseris 22 — 10 (46) J. T. Howell anism against potential predators, it is also possible that these alkaloids may have a direct negative affect on the butterflies. CONCLUSIONS This updated state-wide mapping of Cape ivy populations should aid in regional weed planning and in identifying areas of greatest potential invasion. Cape ivy was present in seven different broad plant community types. This is contrary to the common assumption that Cape ivy is primarily or even exclusively a riparian invasive (Hoover 1970; Smith 1976; Beauchamp 1986; Barbour and Billings 2000). State-wide trends in distribution of the two morphological forms indicate that exstipulate types occur more fre- quently at the northern and southern range of its distribution, while stipulate types are more frequent in the center of its distribution range, extending from southern Humboldt Co. to Los | Angeles Co. Only 21 locations were found that supported both stipulate and exstipulate plants | and these were most often in urbanized coastal areas. Another important aspect of this study was to | evaluate the potential threat of Cape ivy on CNDDB sensitive species known to occur in or | around invaded plant community types. Al- | though the threat to biodiversity was not. measured directly, the CNDDB dataset served | as a surrogate for native species biodiversity. This | analysis suggests that six plants of limited | distribution and nearly 50% of steelhead streams are threatened by the potential expansion of Cape ivy populations. This is of great concern to ecosystem integrity of these sensitive sites and should result in prioritization of effective Cape | ivy management programs in California and southern Oregon. 2010] ACKNOWLEDGMENTS The Cape ivy mapping work, which began in 1995 by Cal-IPC, was an inspiration as well as a starting point for this research. We thank Eva Grotkopp, Alfred Kuo and Mike Pitcairn for collecting and digitizing infor- mation, which they made available in the state-wide database. Numerous other individuals and organiza- tions, especially CNPS and Cal-IPC, provided useful data sets and Rosie Yacoub and Pat Akers from the California Department of Food and Agriculture assisted with the GPS and GIS technology. We also wish to thank Guy Kyser for providing invaluable review and advice. Finally, we thank the UC IPM Exotic Pest and Disease Program for their financial support. LITERATURE CITED ABRAMS, L. AND R. S. FERRIS. 1960. Illustrated flora of the Pacific states, Washington, Oregon and California. Stanford University Press, Stanford, CA. ADAMS, R. AND M. GIANTURCO. 1956. Senecio alkaloids: mikanoidine, the alkaloid from Senecio mikanioides. Journal of American Chemical Society 79:166—-169. ALVAREZ, M. E. 1997. Management of cape-ivy (Delairea odorata) in the Golden Gate National Recreation Area. Pp. 91-95 in M. Kelly, E. Wagner, and P. Warner (eds.), Proceedings Cali- fornia Exotic Pest Plant Council Symposium. California Exotic Pest Plant Council, Concord, CA. AND J. H. CUSHMAN. 2002. Community-level consequences of a plant invasion: effects on three habitats in coastal California. Ecological Applica- tions 12:1434-1444. BALCIUNAS, J. AND L. SMITH. 2006. Prerelease efficacy assessment, in quarantine, of a tephritid gall fly being considered as a biological control agent for Cape-ivy (Delairea odorata). Biological Control 39:516-524. BARBOUR, M. G. AND W. D. BILLINGS. 2000. North American terrestrial vegetation, 2nd ed. Cambridge University Press, New York, NY. BEAUCHAMP, R. M. 1986. A flora of San Diego County, California. Sweetwater River Press, Na- tional City, CA. BEST, C., J. T. HOWELL, W. KNIGHT, I. KNIGHT, AND M. WELLS. 1996. A flora of Sonoma County, Ist ed. California Native Plant Society, Sacramento, CA, BOSSARD, C. C., J. M. RANDALL, AND M. C. HOSHOVSKY (eds.). 2000. Invasive plants of Cali- fornia’s wildlands. University of California Press, Berkeley, CA. CATALANO, S., S. LUSCHI, G. FLAMINI, P. L. CIONI, E. M. NIERI, AND I. MORELLI. 1996. A xanthone from Senecio mikanioides leaves. Phytochemistry 42:1605—1607. DALLMAN, P. R. 1998. Plant life in the world’s Mediterranean climates: California, Chile, South Africa, Australia, and the Mediterranean Basin. California Native Plant Society, Sacramento, CA. DAvis, F. W., D. M. Stoms, A. D. HOLLANDER, K. A. THOMAS, P. A. STINE, D. ODION, M. I. BORCHERT, J. H. THORNE, M. V. GRAy, R. E. WALKER, K. WARNER, AND J. GRAAE. 1998. The California gap ROBISON AND DITOMASO: CALIFORNIA CAPE IVY DISTRIBUTION 93 analysis project — final report. University of California, Santa Barbara, CA. Website http:// www.biogeog.ucsb.edu/projects/gap/gap_rep.html [accessed 07 July 2010]. DITOMASO, J. M. AND E. A. HEALY. 2007. Weeds of California and other western states. University of California Agricultural and Natural Resources, Berkeley, CA. FAGG, P. C. 1989. Control of Delairea odorata (Cape ivy) in native forest with the herbicide clopyralid. Plant Protection Quarterly 4:107—110. GALLO, L. 2000. Floristic reports on some aliens naturalized in Croatia (Republika Hrvatska) with notes on the Italian distribution. Gortania Atti del Museo Friulano di Storia Naturale 22:77—-79. HASELWOOD, E. L. AND G. G. MOTTER. 1983. Handbook of Hawaiian weeds. University of Hawaii Press, Honolulu, HI. HENDRICKS, J. D., R. O. SINNHUBER, M. D. HENDER- SON, AND D. R. BUHLER. 1981. Liver and kidney pathology in rainbow trout (Salmo. gairdneri) exposed to dietary pyrrolizidine (Senecio jacobaea) alkaloids. Experimental and Molecular Pathology 35:170-183. HICKMAN, L. R. (ed.). 1993. The Jepson manual.: higher plants of California. University of Califor- nia Press, Berkeley, CA. HIJMANS, R. J., S. E. CAMERON, J. L. PARRA, P. G. JONES, AND A. JARVIS. 2004. The WorldClim interpolated global terrestrial climate surfaces. Version 1.3. Website http://www.worldclim.org/ [accessed 23 June 2010]. HIRANO, R. T. 1983. German ivy. Pp. 434-435 in E. L. Haselwood, G. G. Motter, and R. T. Hirano (eds.), Handbook of Hawaiian weeds, 2nd ed. University of Hawaii Press, Honolulu, HI. HOLLAND, R. F. 1986. Preliminary descriptions of the terrestrial natural communities of California. California Department of Fish and Game, Sacra- mento, CA. HOOVER, R. F. 1970. The vascular plants of San Luis Obispo County, California. University of Califor- nia Press, Berkeley, CA. HOWELL, J. T. 1970. Marin flora, 2nd ed. University of California Press, Berkeley, CA. JACOBI, J. D. AND F. R. WARSHAUER. 1992. Distribu- tion of six alien plant species in upland habitats on the island of Hawari. Pp. 155-188 in C. P. Stone, C. W. Smith, and J. T. Tunison (eds.), Alien plant invasions in native ecosystems of Hawai: man- agement and research. University of Hawaii Cooperative National Park Resources Studies Unit, Honolulu, HI. JEPSON, W. L. 1951. A manual of the flowering plants of California. University of California Press, Berkeley, CA. JUNAK, S., T. AYERS, R. ScoTT, D. WILKEN, AND D. YOUNG. 1995. A flora of Santa Cruz Island. California Native Plant Society, Sacramento, CA. MANSKE, R. H. F. 1936. The alkaloids of Senecio species. II. Some miscellaneous observations. Ca- nadian Journal Research 14B:6—11. MATTHEws, M. A. 1997. An illustrated field key to the flowering plants of Monterey County and ferns, fern allies, and conifers. California Native Plant Society, Sacramento, CA. Munz, P. A. 1974. A flora of Southern California. University of California Press, Berkeley, CA. 94 MADRONO PALHINHA, R. T. 1974. Flora de Portugal. Wheldon and Wesley, Ltd., New York, NY. PARODI, L. R. 1959. Enciclopedia Argentina de Agricultural y Jardineria. Editorial Acme S.A.C.I., Buenos Aires, Argentina. PIGNATTI, S. 1982. Flora D’Italia. Edagricole, Bologna, Italy. ROBISON, R. A. 2006. Distribution, growth analysis and reproductive biology of Cape ivy (Delairea odorata Lem., syn. Senecio mikanioides Walp.) in Califor- nia. Ph.D. dissertation, University of California, Davis, CA. ScoTT, J. K. AND E. S. DELFOSSE. 1992. Southern African plants naturalized in Australia: a review of weed status and biological control potential. Plant Protection Quarterly 7:70—-80. SMITH, C. F. 1976. A flora of the Santa Barbara region, California. Santa Barbara Museum of Natural History, Santa Barbara, CA. [Vol. 57 SMITH, G. L. AND C. R. WHEELER. 1992. Flora of the vascular plants of Mendocino County. The Uni- versity of San Francisco, San Francisco, CA. STELLJES, M. E., R. B. KELLEY, R. J. MOLYNEUX, AND J. N. SEIBER. 1991. GC-MS determination of pyrrolizidine alkaloids in four Senecio species. Journal of Natural Products 54:759—773. AND J. N. SEIBER. 1990. Pyrrolizidine alkaloids in an overwintering population of monarch butter- flies (Danus plexippus) in California. Journal of Chemical Ecology 16:1459-1470. THOMAS, J. H. 1961. Flora of the Santa Cruz Mountains of California: a manual of vascular plants. Stanford University Press, Stanford, CA. WEBB, C. J., W. R. SYKES, AND P. J. GARNOCK-JONES. 1988. Flora of New Zealand, Christchurch, New Zealand. ZANGHERI, P. 1976. Flora Italica. CEDAM, Padova, Italy. MADRONO, Vol. 57, No. 2, pp. 95—105, 2010 SPECIES BOUNDARIES IN PYRROCOMA LIATRIFORMIS AND PYRROCOMA SCABERULA (ASTERACEAE) BASED ON AFLP DATA JAMES F. SMITH AND DustTy N. PERKINS Department of Biological Sciences, MS1515, Boise State University, 1910 University Drive, Boise, ID, 83725 USA jfsmith@boisestate.edu CURTIS R. BJORK Box 131 Clearwater, B.C., VOE 1NO, Canada GINA GLENNE Western Colorado Field Office, 764 Horizon Drive, Building B, Grand Junction, CO, 81506-3946 USA ABSTRACT Previous investigations into the morphology of Pyrrocoma liatriformis sensu lato in northern Idaho and adjacent Washington have revealed two distinct morphologies that correspond to their geographical ranges. These same populations and individuals have been analyzed using AFLP data. Over 400 loci were identified among all individuals using two sets of AFLP adaptors. The data are in agreement with the morphological data and separate the populations from the Snake River Canyon/ Camas Prairie from those of the Palouse grasslands. Data clustering methodologies using both presence/absence data for all individuals and allele frequencies for each population produced similar results. We suggest the name P. scaberula be resurrected to encompass the populations from the Snake River Canyon and Camas Prairie. Key Words: AFLP, Asteraceae, Idaho, Pyrrocoma, species boundaries, Washington. Numerous species concepts have been pro- posed to unambiguously determine species boundaries (see Niklas 1997; Howard and Berlocher 1998; Wilson 1999; Coyne and Orr 2004; Sites and Marshall 2003, 2004, for more detailed summaries). Selecting a concept poses a challenge to biologists, particularly botanists, where relatively common gene flow among morphologically distinct populations seems to preclude the widespread adoption of the biolog- ical species concept (Burger 1975; Donoghue 1985; Mishler and Brandon 1987; Ellstrand et al. 1996). Over the past 20 yr, molecular methods have provided systematists an additional means of assessing species concepts beyond morphological variability (Miller and Spooner 1999; Lopez et al. 1999; Duim et al. 2001; Parsons and Shaw 2001; Dawood et al. 2002; Wiens and Penkrot 2002; Richardson et al. 2003; Martinez-Ortega et al. 2004; Sites and Marshall 2004; Whittall et al. 2004; Garzon et al. 2005; Irwin et al. 2005; Pons et al. 2006; Suatoni et al. 2006; Roe and Sperling 2007; Manoko et al. 2007; Meudt and Clarke 2007; Guo et al. 2008). Molecular data allow a means to determine if populations are genetically distinct from each other, obtain estimates of gene flow between populations, and resolve if they are mutually monophyletic. Assessing population variation to resolve taxonomic status by molecular genetic means can be done by a variety of methods including simple sequence repeats (SSRs or microsatellites, Akkaya et al. 1995), inter-simple sequence repeats (ISSRs; Smith and Bateman 2002), randomly amplified polymorphic DNA (RAPDs; Smith and Pham 1996), amplified fragment length polymorphisms (AFLPs; Vos et al. 1995) and randomly amplified fingerprinting (RAFs; Wal- dron et al. 2002). Many studies have employed AFLPs to provide evidence that a single species should be divided into multiple species where morphological data were limiting or conflicting (Kardolus et al. 1998; Lopez et al. 1999; Mueller and Wolfenbarger 1999; Duim et al. 2001; Hedrén et al. 2001; Koopman et al. 2001; Bottini et al. 2002; Parsons and Shaw 2001; Richardson et al. 2003; Martinez-Ortega et al. 2004; Whittall et al. 2004; Garzon et al. 2005; Irwin et al. 2005; Manoko et al. 2007; Meudt and Clarke 2007; Roe and Sperling 2007; Travis et al. 2008). According to Becker et al. (1995) AFLP fragments are derived from throughout the genome and bands of identical size (co-migrating) are predominantly homologous in closely related organisms (Waugh et al. 1997; Rademaker et al. 2000). Herein, we examine the AFLP variation among populations of Pyrrocoma liatriformis sensu lato. Pyrrocoma liatriformis E. Greene (syn. Haplo- pappus liatriformis (Greene) St. John) is an herbaceous perennial found in northern Idaho and adjacent Washington (Fig. 1). This taxon has 96 MADRONO OREGON Fic. 1. Map showing the distribution of populations sampled in this analysis. Abbreviations follow Table 1. | Open circles represent P. scaberula and closed circles represent P. liatriformis, the star represents P. carthamoides. | generally been viewed with a broad species concept (Hitchcock et al. 1955) that may encompass two distinct morphological species. Recent work on the morphology of P. liatriformis sensu lato has discovered that neither qualitative nor quantitative morphological characters are | uniform across P. liatriformis sensu lato (Bjork | and Darrach 2009). Instead, the morphological | variation is correlated with geographical range; | plants from the Palouse grasslands are mostly [Vol. 57 2010] heavily tomentose throughout, resinous-punctate glands are absent and flower heads tend to be smaller (10-13.8 mm) whereas those of the canyon/Camas Prairie regions are hispid, with conspicuous resinous-punctate glands and flower heads are larger (13.8—-15.1 mm). Taxonomic investigations of the type specimens of Pyrro- coma have revealed that these morphological entities have been described in the past as P. scaberula E. Greene (canyon/Camas Prairie plants) and P. /iatriformis (Palouse grasslands); the former species largely has been considered a synonym of the latter in most treatments. Bjork and Darrach (2009) studied the mor- phological variation that separates P. liatriformis from P. scaberula by examining eight continu- ously variable morphological characters for 325 plants, 201 individuals in 18 populations from the Palouse and 124 individuals in 13 populations from the canyon/Camas Prairie region. The characters showed clear non-uniformity and although there is overlap in the ranges of the characters, the means of six characters (lateral branches, number of heads, head length, head width, phyllary width, and leaf width) are statistically significant for each group of popula- tions. Furthermore, principle component analy- ses clearly separated the populations into two distinct groups. Although some populations in the Palouse grasslands share the lack of tomen- tum and strong glandularity of the canyon/ Camas Prairie plants, they were clearly attribut- able to the morphology of P. liatriformis sensu stricto based on quantitative characters. Despite the ability of the characters to separate the Palouse and canyon/Camas Prairie popula- tions into distinct clusters, there is a strong overlap among the ranges of the morphological characters. These data suggest the two species are either closely related, represent a progenitor- derivative species pair (Gottlieb 1973, 1974; Gottlieb and Pilz 1976; Crawford and Smith 1982; Ranker and Schnabel 1986: Perron et al. 2000), undergo hybridization, or are a combina- tion of these. It is the goal of this study to resolve whether the morphological species as defined by Bjork and Darrach (2009) are congruent with patterns shown by molecular data. MATERIALS AND METHODS To obtain an estimate of molecular genetic variability, we sampled 32 populations of Pyrro- coma liatriformis sensu lato from both Palouse grassland and canyon/Camas Prairie populations (Fig. 1). One additional population of P. cartha- moides was used as an outgroup for comparison. Populations, their abbreviations, and tentative Species identifications are presented in Table 1 and plotted onto a map in Fig. 1. These are the SMITH ET AL.: PYRROCOMA SPECIES BOUNDARIES 97 same populations and individuals sampled for morphological data by Bjérk and Darrach (2009). At the time sampling was conducted, exceptionally dry conditions in the region result- ed in few populations that were flowering, which minimized the number of individuals that were sampled per population (Bjo6rk personal obser- vation). Ideally 25-35 individuals were to have been sampled, but a total of 25 or more individuals was possible for only five of 33 populations (Table 1). Leaves were collected on silica gel and were the source of DNA extraction using DNeasy kits (Qiagen, Valencia, CA). One fertile stem per plant from populations of larger size was collected as a voucher. Plants from smaller populations were vouchered nondestruc- tively with photographs taken with a ruler for scale. All vouchers, including photographic ones, are deposited at the University of Idaho Stillinger Herbarium (ID). Approximately 500 ng of DNA from each individual was digested with Msel and EcoRI while simultaneously annealing the Msel and EcoRI adaptors at room temperature overnight. Annealing of the adaptors to the ends of the fragments alters the restriction site and precludes further digestion or need to perform reactions separately. To reduce the overall number of fragments amplified, and thus improve detection of homologous amplified fragments, a_ pre- selective amplification was performed using the primers EcoRI + A (5'-GACTGCGTAC- CAATTCA-3’) and MseI + C (5'-GAT- GAGTCCTGAGTAAC-3’) and 1 ul of digest- ed/ligated DNA. Pre-selective amplification was run with 20 cycles of denaturing at 94°C for 30 sec, annealing at 56°C for 1 min and extension at 72°C for 1 min. Products were diluted 1:20 with TE buffer and 3 uL of the dilution were used in selective amplification using either the A primer set, M-CAC (5’-GATGAGTCCTGAG- TAACAC-3’) and labeled (with dye for Li-Cor system) E-ACT (5’-GACTGCGTACCAATT- CACT-3’) or the T primer set, M-CTC (GAT- GAGTCCTGAGTAACTC-3’) and labeled E- ACC (5'-GACTGCGTACCAATTCACC-3’). Final selection amplification used an _ initial denaturation at 94°C for 2 min followed by 10 cycles of denaturation at 94°C for 20 sec, annealing at 66°C for 30 sec and extension at 72°C for 2 min. This was followed with 25 more cycles that differed only in reducing the annealing temperature to 56°C. Lastly there was a 30 min extension period. Final AFLP products were separated on 6.5% polyacrylamide gels and visualized on a Li-Cor LongreadIR automated sequencer (Li-Cor Bio- technology Division, Lincoln, Nebraska). Molec- ular weight size standards were run on each end of each gel. Digital images of the gels were analyzed using Gene ImagIR (Li-Cor Biotech- 98 MADRONO TABLE 1. [Vol. 57 LOCATIONS AND ABBREVIATIONS OF POPULATIONS SAMPLED IN THIS ANALYSIS, THEIR SPECIES DESIGNATION BASED ON MORPHOLOGICAL DATA, AND AMPLIFICATION SUCCESS FOR EACH SET OF AFLP ADAPTORS (A AND T). Vouchers are deposited at the University of Idaho Stillinger Herbarium (ID). Species designation . scaberula—Anatone, Asotin Co., WA . scaberula—Chesley Railroad, Lewis Co., ID scaberula—Craig Mountain, Nez Perce Co., ID . scaberula—Ferdinand Butte, Idaho Co., ID scaberula—Lawyer Canyon, Lewis Co., ID scaberula—Lime Hill, Asotin Co., WA scaberula—Redbird Road, Nez Perce Co., ID scaberula—Soldiers Meadow, Lewis Co., ID scaberula—Talmacks North, Lewis Co., ID scaberula—Upper Cold Spring Creek, Lewis Co. ID scaberula—Weissenfels Ridge, Asotin Co., WA liatriformis—American Ridge, Latah Co., ID liatriformis—Barking Dog, Whitman Co., WA liatriformis—Cedar Ridge, Latah Co., ID liatriformis—Eden Valley, Whitman Co., WA liatriformis—Genesee South, Nez Perce Co., ID liatriformis—Gross Road, Whitman Co., WA liatriformis—Joel, Latah Co., ID liatriformis—Kramer Prairie, Whitman Co., WA . liatriformis—Lenville Road, Latah Co., ID . liatriformis—Mix Road, Latah Co., ID liatriformis—Palmer Butte, Latah Co., ID liatriformis—Palouse Prairie Strip,Whitman Co. WA liatriformis—Rose Creek, Whitman Co., WA liatriformis—Armstrong Road, Whitman Co., WA liatriformis—Steptoe Butte, Whitman Co., WA liatriformis—Spaulding Road, Spokane Co., WA . liatriformis—Uniontown, Whitman Co., WA . liatriformis—Whelan Cemetery, Whitman Co., WA . liatriformis—Wawawai Grade, Whitman Co., WA . carthamoides—Smoot Hill, Whitman Co., WA VV VVDDDDDD VDDD DDDDDDDDDDDIIIIBIVBmBrvprwrs nology Division, Lincoln, NE) to determine molecular weight designations for each fragment. Gel images were edited to ensure that fragments of identical size were correctly assigned the same weights. These data were exported using Gene Profiler (Scanalytics, Inc., Fairfax, VA) and fragments were assigned an allele designation based on which set of adaptors was used (herein designated as either the A or T set of alleles) and their molecular weight. Each individual was then scored for presence or absence of each allele. Fragments greater than 600 bp and less than 49 bp were excluded from the AFLP analyses. Large fragments may be amplified with lower frequency during the process, as a result their consistency may be less reproducible and reliable. Similarly, homology of larger fragments becomes less probable since there are greater opportunities for insertions and deletions of DNA to alter fragment sizes between individuals. Smaller fragments were excluded due to potential diffi- culties in resolving fragment sizes. scaberula—Ferdinand east/Meadow Creek, Idaho Co., ID liatriformis—South end of Paradise Ridge, Latah Co., ID Number of individuals amplified: Population abbreviation (number of individuals sampled) (A/T) AN (12) 12/3 CH (25) 21/24 CM (24) 24/23 FB (25) 22/24 FE (16) 13/13 LC (12) 11/12 LH (25) 22/25 RR (25) 23/25 SO (1) 1/1 TA (12) 0/12 TS (5) 5/5 WR (8) HT AR (16) 16/15 BD (16) 16/16 CR (10) 10/10 EV (7) 6/7 GE (4) 3/3 GR (11) 10/10 JO (5) 5/5 KS (19) 19/18 LR (11) 10/11 PP (5) 5/4 PB (5) 5/5 PR (15) 15/15 PS (9) 8/8 RC (5) 5/5 RT (6) 3/2 SB (7) HA SP (3) 2/2 UN (3) 2/2 WC (25) 20/23 WW (21) 20/20 SM (23) 23/22 Data were entered into MacClade (Maddison | and Maddison 2000) and exported as a simple. table. The table was modified using AFLPDAT | (Ehrich 2006) to convert the files to formats | usable in STRUCTURE (Pritchard et al. 2000; . Falush et al. 2007) as well as to generate allele | frequency data for each population. There are many alternative approaches to_ analyze AFLP data to detect structure within the data set (Bonin et al. 2007). In general, these | break down into two analyses: 1) analysis of bands directly (presence/absence) or 2) converting | the band data into allele frequency data for each | population. Both data types can then be used in’ an array of methodologies to detect diversity and | structure within and among the sampled popula- | tions. Here we opt to make use of both band data | and allele frequency data to generate tree-based representations of the variation (Bonin et al.. 2007). We make use of two methods to analyze the data for both band and allele frequency data: 2010] Neighbor-joining (NJ) using Jaccard’s genetic similarities (bands) or Nei’s (1972) genetic distance and UPGMA. These analyses are among the most widely used methodologies for resolving species boundaries using AFLP data (Miller and Spooner 1999; Duim et al. 2001; Koopman et al. 2001; Parsons and Shaw 2001; Coulibaly et al. 2003; Jacoby et al. 2003; Richardson et al. 2003; Dehmer and Hammer 2004; Martinez-Ortega et al. 2004; Whittall et al. 2004; Garzon et al. 2005; Manoko et al. 2007; Guo et al. 2008). A data matrix of presence/absence for all bands was directly imported into PAUP* (Swofford 2002) for NJ and UPGMA analyses. Allele frequencies were used to calculate population genetic distances that were then used to generate NJ and UPGMA trees in PHY LIP (version 3.67; Felsenstein 2007). We used the Bayesian clustering method implemented in STRUCTURE 2.2 (Falush et al. 2007) to determine the optimal number of groups indicated by the data. Assuming all populations are in Hardy-Weinberg and linkage equilibria, this method assigns individuals to one of the pre-specified numbers of genetic clusters, K, using multi-locus genotypes and Markov Chain Monte Carlo sampling. We ran separate clustering simulations for all populations over a range of one to 40 clusters (1.e., K = one to 40). Individuals with data missing for either the A or T set of alleles were removed from the analyses. This reduced some populations to few samples (only three individuals for AN) and resulted in TA being removed completely. Simulations were run assuming an ancestry model that incorpo- rates admixture and correlated allele frequencies across loci for one million generations with a burn-in of 25,000 generations, sampling every 100 generations. We compared posterior probabilities of K from one to 40 clusters using the ad hoc statistic, AK (Evanno et al. 2005). This statistic has been shown to be a better estimator of structure in some data sets, especially those where homoge- neous dispersal among populations cannot be assumed (Evanno et al. 2005; Travis et al. 2008). In such cases, a common pattern is for STRUC- TURE to plateau near the true value of K, and then to continue increasing gradually. Evanno et al. (2005) showed that AK consistently returns a clear peak at the true value of K under a variety of migration models. We chose a maximum K of 40 because this exceeded the number of sampled populations. RESULTS Amplification of DNA was successful for nearly all individuals for both sets of AFLP adaptors, A and T (Table 1). A few individuals did not amplify well which is perhaps the effect of SMITH ET AL.: PYRROCOMA SPECIES BOUNDARIES 99 plant resin that inhibits the reactions. No individuals from the TA population were ampli- fied for the A alleles and only three from AN were amplified fully for the T alleles. Some individuals were not amplified for either allele. Resin was detected in the precipitation stage of the DNA extraction for some individuals of these populations. For several of the gels, bands that were distinctly different in size below 50 bp were classified as the same size by the software. Therefore the limitations of the software for resolving bands at this stage required us to eliminate these bands from the analyses. Scan- ning gels visually indicated that these fragment sizes tended to be consistent across nearly all individuals, therefore their exclusion was unlikely to affect the results whereas their inclusion may have indicated erroneous relationships among individuals and populations. For the A set of adaptors, 245 alleles were scored for 373 individuals ranging from 569 to 50 bp in size. For the T set of adaptors, 177 alleles were scored for 387 individuals ranging from 581 to 50 bp in size. The complete data matrix had 407 individuals from 33 populations and 422 alleles. The analyses that used presence/absence data for all individuals were sensitive to missing data. Individuals with missing data for either allele were either 1) clustered together regardless of population designation, or 2) in disparate parts of the tree (often included in populations that were neither morphologically or geographically simi- lar, data not shown). Therefore we removed these samples (AN1-9, AR8, CH2, 11, 12, CM11, EVS, FB9, 12, KS8, LC12, LH6,7,18, LR11, MIX], RCAS, RR4, 7, RT4, TAOI-12, WC3,6,19,20, 22,23,25, WW6, and 12) and re-ran the analyses. We also ran analyses of the A and T alleles separately to confirm the placement of the individuals excluded above in their respective population (data not shown). The analyses based on allele frequencies divided all populations into two distinct groups corresponding to Pyrrocoma scaberula and P. liatriformis based on a priori designations by Bjork and Darrach (2009; Fig. 2). Pyrrocoma carthamoides clustered within P. Jiatriformis with the UPGMA analysis (data not shown) because this analysis does not allow the outgroup to be specifically designated as such. The NJ tree in contrast (Fig. 2), results in two distinct and monophyletic groups each for P. Jiatriformis and P. scaberula. There are clear clusters within each of these groups. Populations that cluster together in both the NJ and UPGMA trees within P. scaberula are AN/FB/SO/TS, FE/WR, and LC/CH/LH/RR/ CM. Only population TA changes position between the analyses and is found close to CH/ 100 MADRONO [Vol. 57 AN FB SO TA TS Psc LC CH LH RR CM FE WR BD AR PB PS RC SB JO PP rr FR | WC Phi CR GE SP KS EV LR Ww UN oe om Pca :2 Ol FIG. 2. 0.0 Neighbor-joining based tree derived from AFLP allele frequency data. Population names follow abbreviations of Table 1. Bars to the right of the tree mark species boundaries, Psc—Pyrrocoma scaberula, Pli— | Pyrrocoma liatriformis, Pca—Pyrrocoma carthamoides. Population names are abbreviated following Table | and are designated as normal font (P. scaberula) or bold (P. liatriformis). CM/RR/LH/LC in the UPGMA analysis (data not shown). It should be noted that TA is lacking data for over half of the alleles. Within P. liatriformis there are similar clusters of popula- tions that are consistent between analyses. These are AR/PB/RC, PS/SB, JO/PP/PR, RT/WC, GR/ UN, and CR/GE/SP/KS/EV. The grouping of populations BD, LR and WW differ between the two analyses. With UPGMA, BD is close to all other populations of P. liatriformis, LR is close to RT/WC, and WW is close to CR/GE/SPKS/ EVLR/RT/WC (data not shown). The NJ analyses based on bands produced trees that were nearly identical to those based on allele frequencies (Figs. 2, 3) both in terms of clustering populations into species groups, and relationships of populations within each cluster. | The greatest differences are that 1) not all populations were recovered as a single monophy- | letic group (Fig. 3), 2) the NJ tree did not result in a monophyletic P. scaberula due to the. position of population LH, and 3) P. cartha-. moides is clustered within P. scaberula instead of | P. liatriformis with the UPGMA tree (data not. shown). | Groupings within each species are also similar | to the frequency-based methods. The major differences within P. scaberula are the position | of population LH in the NJ tree (Fig. 3) and that individual ANS was more closely associated to 2010] SMITH ET AL.: PYRROCOMA SPECIES BOUNDARIES 101 FE LH WR FE 12 FE 13 WRS & 6 NS Psc FB25 WwwW2i Pli WC WC 21 & 24 RT 2 WC 4,5,7,9 & 18 RT 3 KS SM 18 Pea 0.2 0.1 0.0 Fic. 3. Neighbor-joining based tree derived from AFLP band presence/absence. Where the majority of sampled individuals formed a single cluster only the population name abbreviation is used. In instances where individuals fell outside of their respective population cluster, they are designated with a population name and number for the individual. Population names are abbreviated following Table | and are designated as normal font (P. scaberula) or bold (P. liatriformis). Bars to the right of the tree mark species boundaries, Pse—Pyrrocoma scaberula, Pli— Pyrrocoma liatriformis, Pca—Pyrrocoma carthamoides. 102 MADRONO [Vol. 57 AK (X 1000) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 #15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 Clusters ( K) A @ &€& FY F & KRFg S$ EC HREC FoF 2 CFHE FB F aN Cy F Cu Fic. 4. Structure analysis of all populations of Pyrrocoma liatriformis sensu lato sampled in this analysis. A. Plot — showing values of AK for each value of K. B. Bayesian assignment of individuals to two clusters. The bars represent | the estimated posterior probabilities of each individual belonging to each of the two inferred clusters. WR/FE than to FB/SO/TS. The lack of unity of population AN is likely due to the fact that only approximately half of the T alleles were scorable for ANS. Population TA was excluded from these analyses due to a complete lack of A alleles. Based on analysis that utilized only the T alleles, TA was close to AN/FB/SO/TS as it is in the NJ tree based on frequencies (Fig. 2). Likewise, groupings were similar within P. liatriformis, the main exceptions being populations KS and EV which showed a closer affinity to WC/RT based on bands (Fig. 3). Plotting the actual values of K from one to 40 indicated that there was a plateau after K = 2 with small increases in probability for each subsequent value of K. With all populations except the outgroup P. carthamoides included, the greatest AK was at 2 distinct clusters (Fig. 4). These results agree with the clustering results that divide the populations into two species. Only individuals of population EV show any signifi- cant probability of being assigned to the other species (Fig. 4). DISCUSSION All analyses of AFLP data presented here, regardless of whether bands or frequencies were used, separate the populations into Pyrrocoma liatriformis and P. scaberula as determined by Bjork and Darrach (2009) using morphological - data (Figs. 2-4). The congruence of different | methodologies is largely considered a means of | overcoming potential problems of homology with | AFLP data (Koopman et al. 2001) and the results © of these analyses are congruent with previous work on morphology. | Relationships Between Species : Pyrrocoma scaberula is paraphyletic based on | NJ analysis of bands (Fig. 3). This raises the | question whether these two species may or may | not represent a progenitor-derivative pair (Gott- lieb 1973, 1974; Gottlieb and Pilz 1976). The progenitor species would be expected to be. paraphyletic since the derivative species would — have resulted from a subset of populations. | However, a second important criterion for a> progenitor-derivative species is that the derivative | species should contain a subset of the total diversity found in the progenitor. A summary _ of the presence/absence data shows that 73.2% of the alleles are shared between the two species, | 15.7% are unique to P. liatriformis and 11.1% are unique to P. scaberula. These results indicate that | a large portion of the data is shared among the > individuals and populations rather than being - unique to either the putative progenitor species — (P. scaberula) or the putative derivative (P. | 2010] liatriformis) and thus argues against a progenitor- derivative pair. Likewise, the results of the Bayesian simulations in STRUCTURE do not indicate any overlap of populations between the two species, but instead the optimal data partition is equivalent to two groups (Fig. 4). It seems more likely that the paraphyly of P. scaberula is a result of recent common ancestry with shared alleles between the populations. Relationships of Populations within Species The AFLP results clearly show population genetic structure within each species (Figs. 2, 3). Within P. scaberula, groups that consistently hold together following the tree-based methods include AN/FB/SO/TS/TA, FE/WR, and perhaps LC/LH/CH/RR/CM. Within P. liatriformis, AR/ PB/RC, SB/PS, JO/PP/PR, GR/UN, CR/GE/SP, and RT/WC are commonly recovered. Popula- tions BD, KS, WW, LR, and EV sometimes show relationships with other groups but not always. Population BD has the longest branches showing the greatest genetic distance from other popula- tions. The results from STRUCTURE indicate that the populations are best treated as two clusters, corresponding to the two species. The one exception is population EV where individuals show the greatest probability of being assigned to P. liatriformis, but also have a _ noticeable probability of being assigned to P. scaberula (Fig. 4). These data may indicate some recent hybridization within this population, be the retention of a large number of alleles that are ancestral to both species, or reflect incomplete lineage sorting. ACKNOWLEDGMENTS We would like to thank Karen Gray and Tyson Kemper who offered their time to collect and send leaf material for the AFLP analyses and two reviewers and the editors of Madrono for improving this manuscript. We would also like to thank the United States Fish and Wildlife Service and Boise State University for provid- ing funds to complete the AFLP analyses. LITERATURE CITED AKKAYA, M.S., R. C. SHOEMAKER, J. E. SPECHT, A. A. BHAGWAT, AND P. B. CREGAN. 1995. Integration of simple sequence repeat DNA markers into a soybean linkage map. Crop Science 35:1439—-1445. BECKER, J., M. KUIPER, P. Vos, F. SALAMINI, AND M. HEUN. 1995. Combined mapping of AFLP and RFLP markers in barley. Molecular and General Genetics 249:65-73. BJORK, C. R. AND M. DARRACH. 2009. An investiga- tion of morphological evidence supports the resurrection of Pyrrocoma scaberula (Asteraceae: Astereae). Journal of the Botanical Research Institute of Texas 3:231—238. BONIN, A., D. EHRICH, AND S. MANEL. 2007. Statistical analysis of amplified fragment length polymorphism data: a toolbox for molecular SMITH ET AL.: PYRROCOMA SPECIES BOUNDARIES 103 ecologists and evolutionists. Molecular Ecology 16:3737-3758. BOTTINI, M. C. J.. A. DE BusTos, N. JOUVE, AND L. PoGGIo. 2002. AFLP characterization of natural populations of Berberis (Berberidaceae) in Patago- nia, Argentina. Plant Systematics and Evolution 231:133-142. BURGER, W. C. 1975. The species concept in Quercus. Taxon 24:45—S0. COULIBALY, S., R. S. PASQUET, R. PAPA, AND P. GePpTs. 2003. AFLP analysis of the phenetic organisation and genetic diversity of Vigna un- guiculata (L.) Walp. reveals extensive gene flow between wild and domesticated types. Theoretical and Applied Genetics 104:358—366. COYNE, J. A. AND H. A. ORR. 2004. Speciation. Sinauer Associates, Sunderland, MA. CRAWFORD, D. J. AND E. B. SMITH. 1982. Allozyme variation in Coreopsis neucensoides and C. neucen- sis (Compositae), a progenitor-derivative species pair. Evolution 36:379-386. DAwoobD, A., A. CHANNING, AND J. P. BOGART. 2002. A molecular phylogeny of the frog genus Tomop- terna in southern Africa: examining species bound- aries with mitochondrial 12S rRNA sequence data. Molecular Phylogenetics and Evolution 22:407— 413. DEHMER, K. J. AND K. HAMMER. 2004. Taxonomic status and geographic provenance of germplasm acessions in Solanum nigrum L. complex: AFLP data. Genetic Research in Crop Evolution 51:551-558. DONOGHUE, M. J. 1985. A critique of the biological species concept and recommendations for a phylo- genetic alternative. The Bryologist 88:172—181. DuIM, B., P. A. R. VANDAMME, A. RIGTER, S. LAEVENS, J. R. DISKSTRA, AND J. A. WAGENAAR. 2001. Differentiation of Campylobacter species by AFLP fingerprinting. Microbiology 147:2279— DIST, EHRICH, D. 2006. AFLPDAT: a collection of R functions for convenient handling of AFLP data. Molecular Ecology Notes 6:603—604. ELLSTRAND, N. C., R. WHITKUS, AND L. C. RIESE- BERG. 1996. Distribution of spontaneous plant hybrids. Proceedings of the National Academy of Sciences USA 93:5090—5093. EVANNO, G., S. REGNAUT, AND J. GOUDET. 2005. Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study. Molecular Ecology 14:2611—2620. FALUSH, D., M. STEVENS, AND J. K. PRITCHARD. 2007. Inference of population structure using multilocus genotype data: dominant markers and null alleles. Molecular Ecology Notes 7:574—-578. FELSENSTEIN, J. 2007. PHYLIP: Phylogeny Inference Package. Software and documentation distributed by the author, Seattle. WA. Website http://evolution. genetics.washington.edu/phylip.html [accessed 28 June 2010]. GARZON, C. D., D. M. GEISER, AND G. W. MOORMAN. 2005. Diagnosis and population analysis of Pythium species using AFLP fingerprinting. Plant Disease 89:81—89. GOTTLIEB, L. D. 1973. Genetic differentiation, sym- patric speciation, and the origin of a diploid species of Stephanomeria. American Journal of Botany 60:545-553. 104 . 1974. Genetic confirmation of the origin of Clarkia lingulata. Evolution 28:244—-250. AND G. PILZ. 1976. Genetic similarity between Gaura longiflora and its obligately outcrossing derivative G. demareei. Systematic Botany 1: 181-187. Guo, Y.-P., J. SAUKEL, AND F. EHRENDORFER. 2008. AFLP trees versus scatterplots: evolution and phylogeography of the polyploid complex of Achillea millefolium agg. (Asteraceae). Taxon 57:153-169. HEDREN, M., M. F. FAy, AND M. W. CHASE. 2001. Amplified fragment length polymorphisms (AFLP) reveal details of polyploid evolution in Dactylo- rhiza (Orchidaceae). American Journal of Botany 88:1868—1880. HITCHCOCK, C. L., A. CRONQUIST, M. OWNBEY, AND J. W. THOMPSON. 1955. Vascular plants of the Northwest. University of Washington Press, Seat- tle, WA. HOWARD, D. J. AND S. H. BERLOCHER. 1998. Endless forms: species and speciation. Oxford University Press, New York, NY. IRWIN, D. E., S. BENSCH, J. H. IRWIN, AND T. D. PRICE. 2005. Speciation by distance in a ring species. Science 307:414-416. JACOBY, A., M. T. LABUSCHAGNE, AND C. D. VILJOEN. 2003. Genetic relationships between southern African Solanum retroflexum Dun. and other related species measured by morphological and DNA markers. Euphytica 32:109-113. KARDOLUS, J. P., H. J. VAN ECK, AND R. G. VAN DEN BERG. 1998. The potential of AFLPs in biosyste- matics: a first application in Solanum taxonomy (Solanaceae). Plant Systematics and Evolution 210:87-103. KOOPMAN, W. J. M., M. J. ZEVENBERGEN, AND R. G. VAN DEN BERG. 2001. Species relationships in Lactuca s. |. (Lactuceae, Asteraceae) inferred from AFLP fingerprints. American Journal of Botany 88:1881—-1887. LOPEZ, J. V., R. KERSANACH, S. A. REHNER, AND N. KNOWLTON. 1999. Molecular determination of species boundaries in corals: genetic analysis of the Montastraea annularis complex using amplified fragment length polymorphisms and a microsatel- lite marker. Biological Bulletin 196:80—93. MADDISON, D. R. AND W. P. MADDISON. 2000. MacClade version 4.0. Sinauer Associates, Inc., Sunderland, MA. MANOKO, M. L. K., R. G. VAN DEN BERG, R. M. C. FERON, G. M. VAN DER WEERDEN, AND C. MARIANI. 2007. AFLP markers support separation of Solanum nodiflorum from Solanum americanum sensu stricto (Solanaceae). Plant Systematics and Evolution 267:1—11. MARTINEZ-ORTEGA, M. M., L. DELGADO, D. C. ALBACH, J. A. ELENA-ROSSELLO, AND E. RICO. 2004. Species boundaries and phylogeographic patterns in cryptic taxa inferred from AFLP markers: Veronica subgen. Pentasepalae (Scrophu- lariaceae) in the western Mediterranean. Systematic Botany 29:965-986. MeEupT, H. M. AND A. C. CLARKE. 2007. Almost forgotten or latest practice? AFLP applications, analyses and advances. Trends in Plant Science 12:106-117. MADRONO [Vol. 57 MILLER, J. T. AND D. M. SPOONER. 1999. Collapse of species boundaries in the wild potato Solanum brevicaule complex (Solanaceae, S. sect. Petota): molecular data. Plant Systematics and Evolution 214:103—130. MISHLER, B. D. AND R. N. BRANDON. 1987. Individ- uality, pluralism, and the phylogenetic species concept. Biology and Philosophy 2:397-414. MUELLER, U. G. AND L. L. WOLFENBARGER. 1999. AFLP genotyping and fingerprinting. Trends in Ecology and Evolution 14:389—394. NEI, M. 1972. Genetic distance between populations. The American Naturalist 106:283. NIKLAS, K. J. 1997. The evolutionary biology of plants. The University of Chicago Press, Chicago, IL. PARSONS, Y. M. AND K. L. SHAW. 2001. Species boundaries and genetic diversity among Hawaiian crickets of the genus Laupala identified using amplified fragment length polymorphism. Molecu- lar Ecology 10:1765—1772. PERRON, M., D. J. PERRY, C. ANDALO, AND J. BOUSQUET. 2000. Evidence from sequence-tagged- site markers of a recent progenitor-derivative species pair in conifers. Proceedings of the National Academy of Sciences USA 97:11331-11336. Pons, J., T. G. BARRACLOUGH, J. GOMEZ-ZURITA, A. CARDOSO, D. P. DURAN, S. HAZELL, S. KAMOUN, W. D. SUMLIN, AND A. P. VOGLER. 2006. Sequence-based species delimitation for the DNA taxonomy of undescribed insects. Systematic Biol- ogy 55:595-609. PRITCHARD, J. K., M. STEPHENS, AND P. DONNELLY. 2000. Inference of population structure using multilocus genotype data. Genetics 155:945—959. RADEMAKER, J. L. W., B. HOsTE, F. J. Louws, K. KERSTERS, J. SwINGS, L. VAUTERIN, P. VAU- TERIN, AND F. J. DE BRUIJN. 2000. Comparison | of AFLP and rep-PCR genome fingerprinting with DNA-DNA homology studies: Xanthomonas as a model system. International Journal of Systematic and Evolutionary Microbiology 50:665—677. RANKER, T. A. AND A. F. SCHNABEL. 1986. Allozyme and morphological evidence for a progenitor- derivative species pair in Camassia (Liliaceae). Systematic Botany 11:433—445. RICHARDSON, J. E., M. F. FAy, Q. C. B. CRONK, AND M. W. CHASE. 2003. Species delimitation and the origin of populations in island representatives of | Phylica (Rhamnaceae). Evolution 57:816—827. Rog, A. R. AND F. A. H. SPERLING. 2007. Population | structure and species boundary delimitation of cryptic Dioryctria moths: an integrative approach. Molecular Ecology 16:3617—3633. SITES, J. W. AND J. C. MARSHALL. 2003. Delimiting — species: a Renaissance issue in systematic biology. | Trends in Ecology and Evolution 18:462—470. AND . 2004. Operational criteria for. delimiting species. Annual Review of Ecology, Evolution, and Systematics 35:199—227. SMITH, J. F. AND T. A. BATEMAN. 2002. Genetic. differentiation of rare and common varieties of. Eriogonum shockleyi (Polygonaceae) in Idaho using. ISSR variability. Western North American Natu- ralist 62:316—326. AND T. V. PHAM. 1996. Genetic diversity in. natural populations of the rare Idaho endemic Allium aaseae and its relatedness and potential | 2010] introgression with A/lium simillimum using RAPD markers. American Journal of Botany 83:717—726. SUATONI, E., S. VICARIO, S. RICE, T. SNELL, AND A. CACCONE. 2006. An analysis of species boundaries and biogeographic patterns in a cryptic species complex: the rotifer - Brachionus plicatilis. Molec- ular Phylogenetics and Evolution 41:86—98. SWOFFORD, D. L. 2002. PAUP* Phylogenetic analysis using parsimony (* and other methods). Version 10. Sinauer Associates, Inc., Sunderland, MA. TRAVIS, S. E., J. E. BAGGS, AND J. MASCHINSKI. 2008. Disentangling the role of hybridization in the evolution of the endangered Arizona cliffrose (Purshia subintegra; Rosaceae): a molecular and morphological analysis. Conservation Genetics 9:1183—-1194. Vos, P., R. HOGERS, AND M. BLEEKER. 1995. AFLP: a new technique for DNA fingerprinting. Nucleic Acids Research 23:4407-4414. WALDRON, J., C. P. PEACE, I. R. SEARLE, A. FURTADO, N. WADE, I. FINDLEY, M. W. GRA- SMITH ET AL.: PYRROCOMA SPECIES BOUNDARIES 105 HAM, AND B. J. CARROLL. 2002. Randomly amplified DNA fingerprinting: a culmination of DNA marker technologies based on arbitrarily- primed PCR amplification. Journal of Biomedicine and Biotechnology 2:141—150. WAUGH, R., N. BONAR, E. BAIRD, B. THOMAS, A. GRANER, P. HAYES, AND W. POWELL. 1997. Homology of AFLP products in three mapping populations of barley. Molecular and General Genetics 255:311—321. WHITTALL, J. B., C. B. HELLQUIST, E. L. SCHNEIDER, AND S. A. HODGES. 2004. Cryptic species in an endangered pondweed community (Potamegeton, Potamogetonaceae) revealed by AFLP markers. American Journal of Botany 91:2022—2029. WIENS, J. J. AND T. L. PENKROT. 2002. Delimiting species based on DNA and morphological varia- tion and discordant species limits in spiny lizards (Sceloporus). Systematic Biology 51:69—91. WILSON, R. A. 1999. Species. The MIT Press, Cam- bridge, MA. MADRONO, Vol. 57, No. 2, pp. 106-119, 2010 ONE TAXON OR TWO: ARE FRASERA UMPQUAENSIS AND F. FASTIGIATA (GENTIANACEAE) DISTINCT SPECIES? BARBARA L. WILSON Carex Working Group, 1377 NW Alta Vista, Corvallis, OR 97330 bwilson@peak.org VALERIE HIPKINS USDA Forest Service, NFGEL, 2480 Carson Road, Placerville, CA 95667 TOM N. KAYE Institute for Applied Ecology, 563 SW Jefferson, Corvallis, OR 97333 ABSTRACT Frasera fastigiata and F. umpquaensis are large, long-lived, perennial herbs with hollow stems, whorled leaves, large nectaries hidden by fringed hoods, and synchronized flowering. They differ in flower color and their ranges are disjunct. Some authors have treated them as conspecific due to their overall morphological similarity. The taxa can be distinguished by isozyme band patterns and by morphological traits including corolla color, relative lengths of corolla and calyx, and calyx lobe shape. Both isozyme differences and morphological differences are not completely fixed, but plants with one atypical feature can be identified by their combination of traits. The taxa should be recognized as distinct species. Key Words: Conservation, Frasera umpquaensis, isozymes, rare plants. Frasera umpquaensis M. Peck & Applegate (Peck and Applegate 1941) is a rare plant with a discontinuous range entirely west of the Cascade- Sierra axis from Lane Co., Oregon, to Trinity Co., California (Fig. 1). Young plants produce a rosette of slightly fleshy leaves, surprisingly lush in their upland habitat. After four to ten or more years, the long, thick rhizome puts up a flowering stalk that may exceed 1.7 m in height, bearing dozens of 1.2 cm-long, white to clear light green flowers that may have a purple tinge. Each of the four petal lobes bears a single large nectary surrounded and partly concealed by a fringed hood. Additional hairs arise below the nectaries, between the filaments. Flowering tends to be synchronized, with almost no plants over a large area flowering in some years and many flowering in other years. After fruiting, the plant returns to the rosette stage for four or more years. An individual plant may live for decades (Kaye 2001). Frasera umpquaensis belongs to a group of four species characterized by this unusual life history, including synchronized flowering, and by tall, hollow stems and whorled leaves (Post 1958). The other three species are F. caroliniensis Walter of eastern North America, and the western species Ff. speciosa Douglas ex Griseb. and F. fastigiata (Pursh) A. Heller. Frasera caroliniensis and F. speciosa have more open inflorescences with larger, rotate, white or greenish corollas speckled with purple or brown. Frasera speciosa is unique in this group in having two nectaries on each petal and large, fringed scales (the corona) that originate below the nectaries and partially cover them (Beattie et al. 1973). Frasera umpquaensis and F. fastigiata have denser inflorescences with > smaller, usually solid-colored flowers that do not. open flat, and their corona is represented by a row of long hairs below the nectary, originating | between the bases of the filaments (Table 1). | Frasera fastigiata grows in Idaho and southeast Washington, while F. umpquaensis lives in south-— west Oregon and northwest California (Fig. 1). Frasera umpquaensis was said to differ from F. fastigiata in the corona hairs, the length and shape of the calyx lobes, and the corolla lobe | width and apex shape (Card 1931; Peck and Applegate 1941; St. John 1941). Pringle (1990) stated that F. fastigiata can have corona hairs | and implied that earlier illustrations omitting | them (Card 1931) were in error, attributed the | supposed calyx lobe differences to diverse inter- | pretations of the words “lanceolate” and “linear” | by different botanists, and dismissed supposed differences in corolla apex shape as variable. within Frasera species. He summarized, “‘com- | parison of specimens from California identified | as .. . F. umpquaensis with specimens from the | Blue Mountains of Oregon and from Idaho. identified as ences by which the two taxa could be distin-. guished” (Pringle 1990, p. 186). ... F. fastigiata disclosed no differ- | Pringle’s (1990) rejection of F. umpquaensis species status had an air of finality, but botanists working with the plants were not satisfied. The color difference, the difference in average inflo- | rescence size, and the 500-km disjunction between their ranges suggested that they were genetically | 2010] 125°0'0"W 120°0'0"W Washington O Willamette OEIk:Meadows—— _ RY Divide-1 R-U Divide-2 Siskiyou-11 Siskiyou-2 California O Shasta-=Trinity 0 25 50 75 Mi 0 50 100 Km 125°0'0"W 120°0'0"W Fic. 1. distinct and therefore should be treated as two species. We report results of isozyme analysis and morphological studies of F. umpquaensis and F. fastigiata with the goal of clarifying their taxonomic status. We consistently use the name F. umpquaensis for the white to green-flowered plants of southwest Oregon and northern Cali- fornia, and F. fastigiata for the blue-flowered plants of Idaho and southeast Washington. MATERIALS AND METHODS Chromosomes In October 2008, the somatic chromosome number of Frasera umpquaensis was determined by observation of mitosis in root tips of plants that had been grown at the Oregon State University greenhouse from seed collected in 2007 near the Elk Camp shelter, near Sourgrass Mountain in the Willamette National Forest, Lane Co., Oregon, at 4500 ft elevation. Excised root tips were pretreated in a saturated aqueous solution of p-dichlorobenzene at 4°C for 5 hr prior to fixation in Carnoy’s fluid (95% ethyl alcohol and glacial acetic acid, 3:1 v:v). The root tips were hydrolyzed in a mixture of concentrated HCl and 95% ethanol (1:1 v:v) for 5—15 min, macerated in acetocarmine, and mounted with a small amount of Hoyer’s solution (Beeks 1955). A Zeiss Photoscope III microscope equipped with phase-contrast optics was used to determine chromosome numbers in cells undergoing mitosis. Isozymes In the summer of 1996, two to five leaves per plant were collected from 75 individuals in three WILSON ET AL.: FRASERA UMPQUAENSIS AND F. FASTIGIATA 107 115°0'0"W 110°0'0"W @ Frasera fastigiata ov Frasera umpquaensis lie Clearwater-1 @ Clearwater-2 > Umatilla 115°0'0"W Locations of populations of Frasera fastigiata and F. umpquaensis sampled for the isozyme study. populations of Frasera fastigiata and 178 indi- viduals in seven populations of F. umpquaensis (Table 2; Fig. 1). Leaf samples were shipped on ice to the National Forest Genetic Electrophore- sis Laboratory (NFGEL). For each individual, three 7 mm diam. leaf discs were placed in a Tris buffer pH 7.5 (Gottlieb 1981) and stored at —70°C. On the morning of the run, samples were thawed, macerated, and absorbed onto 3 mm wide wicks prepared from Whatman 3MM chromatography paper. Methods of electrophoresis follow the general methodology of Conkle et al. (1982) except that most enzyme stains are somewhat modified as outlined in USDA Forest Service (1995). A lithium borate electrode buffer (pH 8.3) was used with a Tris citrate gel buffer, pH 8.3 (Conkle et al. 1982), to resolve alcohol dehydrogynase (ADH), leucine aminopeptidase (LAP), phosphogluco- mutase (PGM), and phosphoglucose isomerase (PGI). A sodium borate electrode buffer (pH 8.0) was used with a Tris citrate gel buffer, pH 8.8 (Conkle et al. 1982), to resolve catalase (CAT), glutamate-oxaloacetate transaminase (GOT), triosephosphate isomerase (TPI), and uridine diphosphoglucose pyrophosphorylase (UGPP).A morpholine citrate electrode and gel buffer, pH 8.0 (USDA Forest Service 1995), was used to resolve diaphorase (DIA), florescent esterate (FEST), isocitrate dehydrogenase (IDH), and malate dehydrogenase (MDH). A Tris citrate electrode and gel buffer, pH 7.2 (USDA Forest Service 1995), was used to resolve phosphoglu- conate dehydrogenase (6PGD). All enzymes were resolved on 11% starch gels. Enzyme stain recipes follow USDA Forest Service (1995) except that GOT was stained using the recipe from Wendel and Weeden (1989). Two people independently scored each gel. When they disagreed, a third 108 SELECTED MORPHOLOGICAL TRAITS OF THE FOUR SPECIES OF FRASERA WITH HOLLOW STEMS AND WHORLED LEAVES. TABLE 1. F. speciosa F. caroliniensis F. fastigiata F. umpquaensis Trait longer than corolla lobs longer than corolla shorter than corolla lobes longer than corolla lobes Calyx lobes, at and lobes usually linear, or after anthesis Calyx lobes, shape lanceolate (sometimes to usually subulate (at least some calyx usually linear or widest above the subulate) white or light green with purple broadly subulate white or light green lobes on all plants subulate) deep blue to light purple, rarely white, base, occasionally subulate white or light green, sometimes Corolla color or brown spots flat; corolla rotate with purple spots flat; corolla rotate occasionally spotted angled; corolla nearly campanulate purple-tinged angled; corolla nearly campanulate Corolla lobes, at anthesis Corona short hairs several, few, or no long scale(s) with several several long hairs long hairs 2 per corolla lobe 6.5—8.5 mm hairs 1 per corolla lobe 2.5-3.3: mm 1 per corolla lobe 7-8 mm 1 per corolla lobe 3—4 mm Nectaries Filament length MADRCNO [Vol. 57 person resolved the conflict. For quality control, 10% of the individuals were run and scored twice. Two zones, designated 1 (faster) and 2 (slower), were resolved for each of the enzymes DIA, GOT, MDH, PGI, and TPI for a total of 18 enzyme systems. Because the limited information available (Post 1958) suggested that these species might be hexaploid and there are no crossing studies to determine the inheritance of isozymes in Frasera, we were unable to provide a genetic interpreta- tion for the complicated band patterns observed on the isozyme gels. Gels were therefore scored for (1) banding pattern, and (2) band presence/ absence. This type of data results in a phenotypic (band pattern and presence) instead of genotypic (alleles and loci) analysis. The band pattern data were used to calculate the average phenotypic identities between pairs of populations using Hedrick’s measure of phenotypic identity (Hed- rick 1971). Diversity measures within populations were calculated by several methods (after Chung et. al. 1991): (1) the number of bands found in each population, (2) percentage of the enzymes that yleld more than one band pattern among individuals in a population, (3) the average number of band patterns per stain in a popula- — tion, (4) the polymorphic index (PI), based on the > frequency of occurrence of each band, and (5) the Shannon-Weaver Diversity Index (Shannon > 1948) based on band pattern frequency. Ordination was performed in the R statistical environment (R Development Core Team 2008). | A simple pair-wise distance matrix was construct- ed from isozyme phenotypes using the function | dist.gene from the R package ‘ape’ (Paradis et al. 2004). Kruskal’s non-metric multidimensional scaling (NMS) was performed on the matrix of | pair-wise distances using the R function isoMDS. | In order to facilitate NMDS, zeros in the pair- | wise distance matrix were replaced with the. arbitrarily small value of 0.0001. Morphology We examined 27 sheets of F. umpquaensis representing 16 distinct collections and 137 sheets of F. fastigiata representing 105 distinct collec- | tions, from HSC, ID, ORE, OSC, UC, WILLU, . and WTU (Appendix 1). Selected flowers were soaked in water and opened to observe the. corona. Other traits we examined included flower color (when it could be determined from the label or dried material), number of leaves in each whorl, inflorescence width, relative length of — calyx and corolla, lengths of filaments and petals, | and length, width, and shape of calyx lobes. To | examine the relationship between the two taxa with morphometric information, we used NMS as implemented in PC-ORD (McCune and 2010] TABLE 2. umpquaensis populations are listed from north to south. NF = WILSON ET AL.: FRASERA UMPQUAENSIS AND F. FASTIGIATA 109 FRASERA UMPQUAENSIS AND F. FASTIGIATA POPULATIONS USED IN THE ISOZYME STUDY. The F. National Forest: BLM = Bureau of Land Management; RNA = Research Natural Area. TRS (township, range, and section) for Oregon and Washington are based on the Willamette Meridian; for Idaho, the Boise Meridian; and for California, the Humboldt Meridian. n = sample size. Population State Location LT R S n F. fastigiata Clearwater- | ID Clearwater NF; Giant White Pine Campground 42N 3W 2 2p Clearwater-2 ID Clearwater NF; Little Boulder Creek 39N 1W 33 "25 Umatilla WA Umatilla NF; Asotin Creek O8N 43E 28 25 F. umpquaensis Willamette OR. Willamette NF; Nevergo Creek 19S 3E a1, 25 Elk Meadows OR Eugene District, BLM; Upper Elk Meadows RNA 23S 2W 35° 325 R-U Divide-1 OR Umpqua NF; Rogue-Umpqua divide 31S JE 10. 25 R-U Divide-2 OR Rogue River NF; Rogue-Umpqua divide 31S 2E 8 27 Siskiyou- 1 OR Siskiyou NF; Bear Camp, Galice Ranger District 34S 10W 12 26 Siskiyou-2 OR Medford BLM; Hobsen Horn Gravel Pit 34S 9W 3425 Shasta-Trinity CA Shasta-Trinity NF; Fern Campground IS. 7E 36. 25 Mefford 2006). Non-metric multidimensional scaling searches iteratively for an ordination with low stress, a measure of the relationship between ranked distances in multidimensional space to the ranked distances in the reduced ordination (Peterson and McCune 2001). The following quantitative traits were used: inflorescence width, filament length, length and width of the longer calyx lobe, difference in length between two adjacent sepals, petal length, and difference in length between petal and longer sepal. In addition, the qualitative trait of flower color was scored | = blue, 0 = non-blue (white or green). We used a random seed with 250 runs of real data to ensure the ordination had low stress. Monte Carlo simulations with 250 iterations were used to assess the probability that final stress could have been obtained by chance. A stability criterion of 0.0001 was used. Student’s t-tests were performed in Microsoft Excel (Microsoft Corportation 2003) to test for significance of differences between the two taxa in sepal length and width, petal length, inflorescence width, and the difference between petal length and sepal length. RESULTS Chromosomes In 2008, counts of 78 chromosomes in each of 4 root tip cells undergoing mitosis confirmed that F. umpquaensis was polyploid and, because the base chromosome number in Frasera is 13 (Rork 1949), presumably hexaploid. Isozymes Frasera isozymes produced the variable, often complicated band patterns expected of poly- ploids. Tentative genetic interpretations could be developed only for the simpler band patterns, biasing genetic analysis against the more variable enzymes (MDH-2, PGI-1, TPI-1, and UGPP;: Appendix 2) that most clearly distinguished the two taxa. Therefore, analyzing the isozyme patterns phenotypically, as patterns and bands, was more appropriate than genetic analysis for these Frasera species (Chung et al. 1991). Most populations of Frasera fastigiata and F. umpquaensis were moderately to highly variable with 40-60% polymorphic loci (Table 3). The Willamette and Shasta-Trinity populations, iso- lated at the northern and southern ends of the F. umpquaensis range, respectively, were the least variable. In the Shasta-Trinity population all but two stains were monomorphic. Although no fixed isozyme differences distin- guished F. fastigiata from F. umpquaensis, overlap was slight in certain enzymes (e.g., MDH and PGI-1) and restricted to the Shasta- Trinity population in one enzyme (TPI-1). In general, Hedrick’s measure of phenotypic simi- larity had high values for within-taxon compar- isons and low values for between-taxon compar- isons (Table 4). The NMS ordination based on the band patterns in individual plants resulted in two clusters corresponding to the two species (Fig. 2). The geographically isolated Shasta-Trinity population of FF. umpquaensis was relatively dissimilar to other F. umpquaensis populations (Table 4). When F. uwmpquaensis and F. fastigiata differed at an enzyme for which the Shasta- Trinity population was monomorphic, the Shasta-Trinity population shared its band pattern with other F. wmpquaensis populations (e.g., for MDH, PGI-1, and PGM; Appendix 2). However, at one of its two variable enzymes (TPI-1), the Shasta-Trinity population had band patterns co) MEASURES OF PHENOTYPIC VARIATION IN ISOZYMES FOR TEN POPULATIONS OF TWO SPECIES OF FRASERA. The F. umpquaensis populations are listed from north to south. TABLE 3. Shannon-Weaver diversity Band patterns/stain Polymorphic % polymorphic Sample size per (mean) index index stain Bands stains 13.8 Population F. fastigiata 0.4403 (0.008) 0.4386 0.3709 0.5115 3.87 (0.136) 4.8932 2.5920 4.1248 2.13 (0.13) 2.06 (0.38) 53.7 (0.49) 44.4 44 (0.70) 44 38 50 24.6 (0.04) 24.4 F. fastigiata mean (SE) Clearwater- 1 1.94 (0.25) 2,39 (0,39) 55.6 24.4 Clearwater-2 Umatilla 61.1 25.0 177.6 F. umpquaensis 0.3448 (0.012) 0.2184 0.4978 0.4438 0.5415 0.2414 0.3513 O.1195 2.9207 (0.11) 1.9488 4.6252 3.8092 4.7107 1.5136 DA Lo2 1.0624 1.83 (0.12 39.7 (0.62) 27.8 41.3 (0.412) 39 47 25.4 (0.06) 25.0 F. umpquaensis mean (SE) 1.39 (0.16) 2.33 (0.48) 2.06 (0.33) 2.28 (0.42) 1.67 (0.20) 1.89 (0.34) 1.22 (0.15) Willamette 55.6 24.9 Elk Meadows 50.0 45 R-U Divide-1 55.6 47 R-U Divide-2 Siskiyou-1 38 44.4 41 29 3323 4.9 25.0 Siskiyou-2 MADRONO 11.1 Shasta-Trinity [Vol. 57 otherwise observed only in the F. fastigiata populations. At the other (CAT), 60% of the individuals had a unique band pattern that seemed attributable to a unique allele. Therefore, Hedrick’s distances indicate that the Shasta- Trinity population was as different from the northern F. umpquaensis population as from F-. fastigiata populations (Table 4). Morphology Some traits reported to distinguish FL ump- quaensis from F. fastigiata failed to separate the taxa consistently, but others were effective (Tables 5 and 6). Corona hairs were sometimes difficult to assess on herbarium specimens because chipping into the dried flowers often broke them, while soaking the flowers rendered them nearly transparent. The hairs often stuck to the fringed membrane that surrounds the nectary, and freeing them intact was difficult. Corona hairs were numerous and easy to see in all 19 F. umpquaensis flowers examined for them (Ta- ble 5). These hairs were also numerous in 18 of the 20 F. fastigiata specimens examined for them, but were often hard to see even when they were numerous. They were absent or sparse in flowers of two F. fastigiata specimens, varying among flowers in one inflorescence in one specimen (Sondenaa 327). In FL umpquaensis, calyx lobes were longer than the mature corolla lobes (Table 5); in F. fastigiata they were shorter, and the difference was statistically significant (Table 6). However, in one of the 93 F. fastigiata specimens examined for this trait, Bjork 7727, the calyx lobes were clearly longer than the corolla on many of the mature flowers. Relative calyx length was often difficult to assess because the calyx lobes were longer than the corolla in bud, in both species. After anthesis the corolla lobes withered and folded, making comparisons of length misleading unless the flower was soaked and the corolla lobes unfolded. In general, the two species were differentiated by calyx lobe shape (Table 5) and length (Ta- bles 6). In FL umpquaensis, calyx lobes were usually linear (uniform in width) at least in the proximal half or lanceolate (widest above the base), though some were subulate (widest at the base and tapering uniformly to the tip). In F. fastigiata, all calyx lobes in most inflorescences and some calyx lobes in all inflorescences were clearly subulate. In both species, the two pairs of calyx lobes were sometimes found to differ in shape and/or length. Calyx lobes were signifi- cantly longer in F. umpquaensis than in F- fastigiata and corolla lobes were signficantly longer in F. fastigiata, although the the range of variation in these traits overlapped between the two species (Table 6). 2010] TABLE 4. HEDRICK’S MEASURE OF SIMILARITY AMONG ISOZYME BAND PATTERN FREQUENCIES IN TEN POPULATIONS OF FRASERA. A value of 1.0 indicates identical variation in a population pair. Siskiyou-2 Siskiyou-1 R-U Divide-2 Willamette Elk Meadow R-U Divide-1 Population Clearwater-1 Clearwater-2 Umatilla Species Clearwater- 1 ta ta te F. fastige WILSON ET AL.: 0.894 725 0.492 Clearwater-2 Umatilla fal F. fastige 0.747 0.455 ~ 0.480 Willamette om 0.431 0.672 0.402 0.450 Elk Meadow RY Ss 0.781 0.497 0.767 0,333 O73 R-U Divide-1 0.771 0.688 0.629 0.374 0.576 0.521 0.584 0.367 0.399 0.621 R-U Divide-2 Siskiyou-1 0.680 0.774 0.702 0.766 0.657 0.673 S 0.773 0.872 O.722 0.522 0.575 Siskiyou-2 FRASERA UMPQUAENSIS AND F. FASTIGIATA tl 0.870 0.574 Or12 Ois59 0.700 0.685 0.606 0.644 Shasta-Trinity en ~a re ee he eR hh ~ e = C a ~ ~ ~ ~ é ~ = ed On a ~ ~ ~ SYS f+ f+ ff fe & ~ be in in i i | F.4 Fi Flower color was difficult to assess on herbar- ium specimens because corollas in many older specimens of both species faded to tan. Flower color was not reported on the labels of the 16 F umpquaensis specimens examined. Field workers report that the flowers are white to greenish, often lightly tinged with purple (Thomas Kaye personal observation; Jennifer Lippert, Willam- ette Natl. Forest, personal communication). Flower color was pale and greenish on the more recently collected herbarium specimens. Labels of the 21 F. fastigiata specimens that mentioned flower color reported it to be blue (including pale blue and “‘fairly deep blue’’), purple, or lavender. Corollas of the more recent dried herbarium specimens varied from deep gentian blue to light purplish blue, with few exceptions (Table 5). One Sheet (Richards 116) consisted of a shoot with blue corollas speckled with darker blue, and two shoots with pale flowers that may have been white in life. Flowers on several F. fastigiata specimens had inconspicuous darker speckles on blue corollas, and on a few sheets the speckles were relatively conspicuous (e.g., Constance 1771, Williams & Goff 16, and Wilson 241). Nonmetric multidimensional scaling based on multiple morphometric traits produced a final stress of 6.188 and usually separated Frasera fastigiata from F. umpquaensis (Fig. 3). The two F. fastigiata that overlapped the cluster of F. umpquaensis were Bjork 7727 which had sepals up to 2.3 mm longer than the petals, and the pale- flowered individual in Richards 116. Despite these anomalies, Bjork 7727 could easily be assigned to F. fastigiata because of its blue flowers, subulate calyx lobes, and two cauline leaves per whorl. The pale-flowered plants in Richards 116 had blue speckles and mostly subulate calyx lobes that were shorter than or barely longer than the corollas. Those traits, plus its occurrence in a population with blue flowered-plants, would lead to its correct identification. DISCUSSION Frasera umpquaensis and F. fastigiata are more similar to each other morphologically than they are to any other species, but they are not the same. Isozymes consistently distinguished the two taxa, although differences were not completely fixed (Fig. 2). The two taxa could be distin- guished morphologically as well (Fig. 3), but as was true for isozymes, the differences were not completely fixed. Despite these occasional incon- sistencies, all specimens could be easily identified to taxon when all traits were taken together. For example, a specimen that had an unexpected calyx length was typical of its taxon for other traits. The past confusion over the differences be- tween these taxa results in part from using MADRONO NMDS Axis 2 NMDS Axis 1 FIG. 2. [Vol. 57 © Clearwater-1 = Clearwater-2 a Umatilla OWillamette © Elk Meadows OR-U Divide -1 AR-U Divide-2 X Siskiyou-1 + Siskiyou-2 @ Shasta-Trinity Non-metric multidimentional scaling of isozyme band patterns in individual samples of Frasera fastigiata (dark symbols) and F. umpquaensis (open and shaded symbols). unreliable traits (Peck and Applegate 1941; St. John 1941; Pringle 1990). Corona hairs do help differentiate the wmpquaensis-fastigiata species pair from / speciosa, which has fringed, mem- branous corona scales, and from F. caroliniensis, which has very short hairs (Table 1), but corona hairs are variable within F. fastigiata and hard to TABLE 5. COMPARISON OF QUALITATIVE TRAITS OBSERVED IN FRASERA UMPQUAENSIS AND F. FASTIGIATA. Trait F. fastigiata F. umpquaensis Corolla color Sample size n = 62 n= 15 Blue 62 0 Pale | 15 Corolla lobes Sample size n = 96 n= 16 >sepals 5 0 = sepals 5 0 ) perpendicular to the maximum and calculating the area for an ellipse (cover = [d,/2][d>/2]z). Shrub volume was determined using an addition- al height (h) measurement and calculating the volume of an ellipsoid (volume = 4/32d,d>h). To determine the effect of canopy condition on the microclimate within host shrubs, three shrubs with canopies and three recently dead shrubs without leaves but with branches intact were selected for measurement. Light intensity (PFD) was measured using a solar monitor (Licor L1- 1776) attached to a quantum sensor (Licor LI- 190SB) placed in a horizontal position close to the interior base of the host shrub. Soil surface temperature was measured using a thermometer (Omega HH21) attached to thermocouples placed no more than | mm beneath the soil surface at the base of the shrub. Temperature and light measurement were taken on the hour from 6:00 a.m. to 8:00 p.m. Precipitation data were obtained from the remote automated weather station at Opal Mountain CA (35°09'15”E; 117°10'32"W; 988 m). This weather station is approximately 30 km from monitored Milkvetch sites at a similar elevation. The Opal Mountain data were in near perfect agreement with data collected closer to A. jaegerianus populations (Rundel et al. 2006), but has the advantage of being continuous MADRONO [Vol. 57 450 —- 1 Rent Uie eceSES | Se OE ete So 1 a er ee > oa re 400 ~ -_- | £ = < a) © = ok O wv i. ae Oo) © _— N ~ + Ww i) a Se oO io) © Oo S Oo © © oO o>) © © © © oO © o S © O - N N N N N N N N N N Fic. |. Annual precipitation (OCT-SEP) from 1999 to 2009 at the remote automated weather station at Opel Mountain, CA (35°09'15”E; 117°10'32”W; 3240 ft). Years refer to the season in which Astragalus jaegerianus 1s reproductive (for example, 1999/2000 is denoted as 2000). This weather station is approximately 18 mi from monitored A. jaegerianus sites and at a similar elevation. The dash line is mean precipitation (160.4 mm-yr _') from 1991 to 1998 at the same location. The mean precipitation during the current drought from 1999 to 2009 was 114.4 mm-yr '. Weather data archived by the Western Regional Climate Center. from 1992 to the present. Precipitation from October through September was used because it includes winter and spring rainfall that affects A. jaegerianus growth and reproduction. Thus, annual precipitation includes October through December precipitation of the previous year. Shrub data were analyzed using Statview (SAS Institute Inc., Cary, NC). Nonparametric statis- tics were used because some shrub data was not normally distributed and resistant to transforma- tion to normality (SAS 1999). Paired sign tests were used to analyze changes in shrub volume and cover. An unpaired t-test was used to compare shrub canopy condition in shrubs that supported live A. jaegerianus with shrubs in which A. jaegerianus had died since monitoring began in 1999 (Brinkman Wash) and 2003 (GCA). RESULTS The current drought began in the fall of 1998 years represent the tail end of a wet period from 1976 to 1998 (Hereford et al. 2006) that pre- sumably generated the high A. jaegerianus population numbers recorded in 1999. While the difference in precipitation between these wet and dry periods is considerable (46.1 mm-yr''), the severity of the drought and its impact on A. | jaegerianus is better appreciated by considering the years before and after 2005; the six-year © period between 1999 and 2004 had a mean precipitation of 100.5 mm-yr~', and the four year | period from 2006 to the 2009 had a mean precipitation of 61.9 mm-yr''. The year 2007 | had the lowest precipitation in the 1991 to 2009 Opel Mountain data set (22 mm). While A. jaegerianus continues to decrease in | density at our long-term study sites, its decline | has slowed and appears to be reaching a plateau | (Fig. 2). No A. jaegerianus mortality was ob- | served in Brinkman Wash populations in 2009, | and GCA populations lost only two A. jaeger- — ianus, the lowest absolute decline in seven years _ of observation. Despite these decreases in mor- | tality, A. jaegerianus numbers remain dangerous- — ly low. Of the 161 original plants at the four study — sites, only 20 remain alive, with zero recruitment | of new A. jaegerianus plants and 100% seedling | mortality since surveys began in 1999 (Brinkman (Hereford et al. 2006) and is in its eleventh year (Fig. 1). Despite an unusually wet 2005 (407 mm), this drought period has a mean precipitation of 114.4 mm-yr ' compared to the relatively wet years preceding it from 1991 to 1998, in which mean precipitation was 160.4 mm-yr |. These wet 2010] HUGGINS ET AL.: DROUGHT AND ASTRAGALUS JAEGERIANUS 123 100 (ieee Eee [a comer | ann | Se [S| ea | ee | ee ! 1 =e eee 1 Y | Brinkman Wash c 80 7 [ - | O | Z Gg 60 Oo | Some = oD) 40 a2 al Ss Ve 86.7 * e-0.192x =) Fr | R2 = 0.983 O aa T AM 1998 2000 2002 2004 2006 2008 2010 80 | Ef | oy f. Gemini Conservation Area = © 60 GE 4 50 - 40 @ ae, =o £0 =| E 20 | y=77.6 —10.93x ; 10 7 < | R* =0.950 ° O 2002 2003 2004 2005 2006 2007 2008 2009 2010 Years FIG. 2. Population declines of Astragalus jaegerianus at two study areas, Brinkman Wash (1999, 2003 through 2009) and Gemini Conservation Area (2003 through 2009). Each study area contains multiple monitored A. jJaegerianus populations. Wash) and 2003 (GCA). All monitored popula- tions have dropped to critical levels and are at risk of local extinction. Assuming that the 4. jJaegerianus mortality observed at our long-term study sites is characteristic of the species across its range, the 5723 mature A. jaegerianus plants which constituted the plants found in 2001 (Charis Professional Services Corp. 2002) would now number approximately 686 individuals. One of these populations precariously close to extinction is M2 at Brinkman Wash (Fig. 3). Since 1999, M2 has declined from 23 plants to one remaining plant. However, mortality has not been constant; population decreases were rela- tively slow between 1999 and 2003, but acceler- ated between 2003 and 2006 to a loss of 6 plants per year in 2005 and 2006 (Fig. 3). Although 2005 was an unusually wet year with more than twice mean annual precipitation, it appears that even this unusually high rainfall could not diminish the momentum of A. jaegerianus mor- tality. By 2007, population M2 had fallen to one plant that has managed to survive the last three years of intense drought. Host shrubs populations have declined simul- taneously with the decline of A. jaegerianus populations. In our shrub transects within A. jaegerianus sites, while some shrubs increased in size, total shrub cover and volume have decreased significantly by roughly 10% between 2000 and 2009 (Fig. 4; paired sign test: P < 0.001, n = 75, for both shrub cover and volume). Mortality of these long-lived shrubs has been high (48%), and the recruitment of new shrubs (5%) has been too low to maintain their populations at previous levels. Among A. jaegerianus host shrubs, shrubs 23 ALIVE 0 DEAD =a T y T “2004 7 15 ALIVE | 8 DEAD | 2006 6 ‘| . oN OO 2 © | on = [on ) e) ] . ° | @ 1 3 ALIVE ae | 20 DEAD MADRONO [Vol. 57 18 ALIVE 5 DEAD 9 ALIVE 14 DEAD 1 ALIVE 22 DEAD a ] T ] T T T T T J Ta a | an ame ] ; = Tice Fic. 3. T T i (ae a rn ine | si T T T Aerial view of Astragalus jaegerianus population M2 at the Montana Mine site (1999, 2003, 2004, 2005, 2006, 2009). Solid dots are live A. jaegerianus plants, and empty dots are dead A. jaegerianus. Population M2 decreased from 23 A. jaegerianus plants in 1999 to one plant in 2009. No recruitment has been observed at this site during this period. The position of A. jaegerianus plants was determined using each plant’s UTM coordinates. with live A. jaegerianus have more intact canopies than host shrubs that once supported A. jaegerianus, which are now dead (Fig. 5; unpaired t-test: Fy; ;;3 = 11.48; P = 0.0010). Soil surface temperature and light intensity beneath shrubs were dependent on the condition of the shrub’s canopy; shrubs with open canopies had light levels five times higher than shrubs with closed canopies, and soil surface temperature beneath shrubs with open canopies were as much as 20°C higher than shrubs with closed canopies (Fig. 6). While most shrubs do not have com- pletely open canopies, among LMMV host shrubs originally surveyed in 1999 and 2003, the average host shrub had only 45 percent of its canopy intact in 2009. DISCUSSION Studies in arid and semi-arid environments demonstrate that the shade produced by host plant canopies mitigate severe abiotic conditions 2010] ae ! _ AST P<0.001 -| 1554 + 153 7 - 157 147 75 b 145 7 1424 Shrub Volume (m3?) ——_——— 2009 P<0.001 | Shrub Cover (m2) 2000 2009 YEAR FIG. 4. Changes in shrub size between 2000 and 2009. Both shrub volume and shrub cover deceased signifi- cantly in the nine years between censuses. Shrubs were censused along a transect adjacent to A. jaegerianus population M1. A. Mean shrub cover measured as an ellipsoid (paired sign test: P < 0.001, n = 75). B. Mean shrub cover measured as an ellipse (paired sign test: P < 0.001, n = 75). by reducing air and soil temperature (Franco and Nobel 1989; Valient-Banuet et al. 1991; Paez and Marco 2000; Flores et al. 2004), and increasing soil moisture availability (Nolasco et al. 1997; Shumway 2000; Warnock et al. 2007). Facilita- tion occurs when microclimate effects such as these increase the establishment and survival of protege plants growing under host shrub cano- pies (Cody 1993). Because the facilitative effect of host plants depends on the capacity of its canopy to modify the environment beneath it, changes in canopy structure can affect the facilitative effect of the host plant (Reisman-Berman 2007). In this study we have documented the drought- induced mortality and canopy deterioration of A. jJaegerianus host plants, and have demonstrated the effect of host plant canopy foliation on soil temperature and light intensity in sub-canopy, A. jaegerianus microhabitat. We have also demon- strated a significant increase in survival of A. jJaegerianus among host plants with more intact canopies. These results support our study hy- pothesis that drought-related changes to host plant canopies affect A. jaegerianus survival, and represent an indirect negative effect of long-term HUGGINS ET AL.: DROUGHT AND ASTRAGALUS JAEGERIANUS 100 S _ 1 S 50 | P= 0.0010 : o 80 os) 70 O 2 60 = = 50 > 740 O = 30 S 20 fae} = 10 Astragalus Astragalus Dead Alive Fic. 5. Percent live host plant canopy and Astragalus jaegerianus status, GCA and Brinkman Wash site combined, 2009. Astragalus jaegerianus was found in host shrubs with more intact canopies (unpaired t-test: Fi, 1g = 11.48; P = 0.0010). Intact host shrub canopy condition was estimated as a percentage of total canopy (live plus dead canopy). 70 po + _— 60 4 + = O 50> =) r ~ © 405 " eB) ok ce 307 L eB) 20 4 - L ~®~ CLOSED CANOPY —o— OPEN CANOPY | 10 + + + } = n 4 n 6 8 10 12 14 16 18 20 1800 t + + t | 4 r a7 1600 4 B - ; Yr 1400: = ~ a 1200 T Ve E Vives | AG SB 1000 i | - © 800 4 _ = 600 + ° a) J 8 LL 400 uh IN oO 200 4 y 074 oe -200 : a ee 6 8 10 12 14 16 18 20 Time of Day Fic. 6. The effect of Astragalus jaegerianus host shrub canopy condition on shrub micro-climate in June 2009 at A. jaegerianus population M1, Montana Mine site, Brinkman Wash. A. Soil surface temperature beneath open and closed canopy shrubs (6:00 to 20:00). B. Light intensity (photon flux density, umol-m ~*-s ') beneath open and closed canopy shrubs (6:00 to 20:00). Closed circles are measurements recorded under shrubs with closed canopies. Open circles are values recorded under shrubs with open canopies. Points are means with standard errors (n = 3). 126 drought on A. jaegerianus populations. Theory suggests that positive and negative interactions should change along gradients in abiotic stress, with positive interactions dominating under harsh physical conditions where host plants ameliorate abiotic stress (Bertness and Callaway 1994). While a number of studies have demon- strated the positive, facilitative effect of host plants in stressful arid environments (Valiente- Banuet and Ezcurra 1991; Paez and Marco 2000; Pugnaire and Luque 2001; Barchuk et al. 2005; Muiriti 2006; Reisman-Berman 2007), to our knowledge, the idea that severe stress associated with long-term drought may diminish host plant facilitation through negative effects on host plant canopies has not been previously documented. The negative effects of long-term drought on Sonoran, Great Basin, and Mojave Desert peren- nial plants are well documented (Goldberg and Turner 1986; Turner 1990; Bowers 2005; Hereford et al. 2006; Miriti 2006; Hamerlynck and McAu- liffe 2008; Hamerlynck and Huxman 2009; Ralphs and Banks 2009), and are similar to drought effects described in this study for A. jaegerianus host shrubs: high shrub mortality, shrub canopy deterioration, and low recruitment. Increases and decreases in mortality associated with fluctuation in interannual precipitation have been reported in other herbaceous desert perennials (Cryptantha flava (A. Nelson) Payson, Casper 1996), and the population declines of A. jaegerianus fit this general pattern, with the exception of 2005, when adult A. jaegerianus mortality continued more or less unaffected by unusually high precipitation (e.g., M2 at Brinkman Wash experienced its highest recorded adult mortality in 2005 and 2006). Seedling establishment also responded weakly to the increase in precipitation in 2005; nine seedlings were established, went dormant through the summer of 2005, and resprouted in 2006. While this was the only observed case of seedling establishment since 1999, 2006 was again a drought year, and these resprouted, second- season plants did not achieve reproductive maturity, and failed to resprout in 2007. The reason for this insensitivity to increased precipitation in 2005 is unclear, but could be the result of the accumulated damage to host shrub canopies inflicted by long-term drought. The effects of drought on A. jaegerianus host plants may proceed rapidly because of positive feedback within the canopy/micro-climate interaction; as shrub canopies deteriorate, evapotranspiration beneath shrubs increases, which increases shrub water stress leading to further canopy deteriora- tion. This positive feedback between shrub canopy and microclimate, and the slow growing nature of desert shrubs may explain why the momentum of A. jaegerianus population declines could not be slowed by a single year of high rainfall in 2005. MADRONO [Vol. 57 Episodic recruitment associated with high precipitation has been observed or inferred from demographic analysis in a number of desert perennials (Shreve 1917; Barbour 1969; Sheps 1973; Jordan and Nobel 1979, 1982; Goldberg and Turner 1986; Turner 1990; Parker 1993: Bowers 1995; McDaniel et al. 2000; Godinez- Alvarez et al. 2003). We have previously hypoth- esized that pulses in high annual precipitation, such as those associated with ENSO events, drive A. jaegerianus recruitment, and between high recruitment years mortality occurs in a more or less constant manor (Sharifi et al. 2009). Consis- tent with this pulse model is the expectation that A. jaegerianus recruitment and mortality should be sensitive to years with high rainfall, and recruitment should increase during high rainfall years; but continued adult mortality and low recruitment through an unusually wet year like 2005 suggests that A. jaegerianus recruitment is likely to be gradual, and may occur during long- term wet periods. Long-term wet periods in the Mojave Desert occur more or less regularly, and are associated with the Pacific Decadal Oscilla- tion that causes decadal-scale variability such as prolonged dry and wet episodes (Hereford et al. 2006). Prolonged wet periods in the Mojave Desert, such as 23 yr wet period between 1976 and 1998, may positively affect A. jaegerianus population growth factors that are relatively insensitive to short-term precipitation such as the condition of slow-growing host shrub cano- pies. Similarly, dry periods result in the deterio- ration of host plant canopies, which diminishes A. jaegerianus recruitment even during years of high precipitation such as 2005. An expectation of this climate-period model is that the sensitivity of recruitment to precipitation is dependent on the climatic context in which precipitation occurs: recruitment sensitivity is high during prolonged wet periods and low during dry periods. Given these hypothetical circumstances, A. jaegerianus populations would tend to oscillate between multi-decadal, high and low population states that are determined by long-term precipitation patterns characteristic of climate-periods. Although adult A. jaegerianus mortality has occurred each year since observations began in 1999, mortality has slowed and stopped in some populations. In population M2 (Fig. 3), a single remaining LMMV has survived alone for three years despite the intense drought (mean precip- itation 50 mm-yr ', 2007-2009, Fig. 1). Astraga- lus jaegerianus it thought to be deep-rooted, and this drought-resistant plant may have access to deeper or more reliable sources of water in its fractured granite substrate, and thus better water relations, than A. jaegerianus that died earlier in the drought. This idea assumes that soil water resources are heterogeneous, and that only the most consistent water resources are able to 2010] maintain A. jaegerianus after prolonged drought. Reduced but persistent populations of A. jaeger- ianus are consistent with expectations of the climate-period model described above. Our previous studies have shown that A. jaegerianus seed density is low to extremely low in the soil seed band compared to other desert shrubs (Rundel et al. 2009; Rundel and Gibson 1996), and seed dispersal beyond host shrub canopies is rare (Rundel et al. 2009). For A. jaegerianus, these are grim ecological circum- stances: as its host shrubs deteriorate and die, and without the ability to disperse to other host shrubs, its recovery to 1999 populations levels in the immediate future is unlikely. If our climate- period model is correct, and surviving, drought resistant A. jaegerianus have access to deep, reliable water sources, drought-reduced popula- tions could persist until the current drought is over, and then expand under wetter climatic conditions. However, if drought conditions con- tinue, it is equally possible that A. jaegerianus numbers may erode further, leaving most, if not all populations in eminent danger of local extinction. Unfortunately, regional climate indi- cators suggest that the Mojave Desert may remain dry for 1 to 2 decades or longer (Breshears et al. 2005; Hereford et al. 2006). In anticipation of prolonged drought, efforts should be made to preserve the A. jaegerianus as a unique and rare component of the Mojave Desert flora. These efforts should focus on habitat preservation, experimental repopulation of en- dangered or extinct subpopulations, and further investigation into the effects of drought on facilitative interactions between A. jaegerianus and its host shrubs. ACKNOWLEDGMENTS We are grateful for assistance from: Muhammad Bari, Clarence Everly, Mark Hessing of the Environ- mental Division of the Directorate of Public Works, National Training Center, Fort Irwin; Connie Ruther- ford of the U.S. Fish and Wildlife Service for permit (TEO26656-2) to conduct this study; Dr Russell Harmon, Senior Program Manager for Terrestrial Science from the Army Research Office in administra- tion of this project; U.C.L.A. student laboratory assistants Andrew Troung and Michelle Nguyen for their keen eyesight and diligence. This study was funded by a contract with the U.S. Army (ARO proposal number: 55166-EV). LITERATURE CITED BARBOUR, M. G. 1969. Age and space distribution of the desert shrub Larrea divaricata. Ecology 50:679-685. BARCHUK, A. H., A. VALIENTE-BANUET, AND M. P. DIAZ. 2005. Effects of shrubs and _ seasonal variability of rainfall on the establishment of Aspidosperma quebracho-blanco in two edaphically contrasting environments. Austral ecology 30: 695-705. HUGGINS ET AL.: DROUGHT AND ASTRAGALUS JAEGERIANUS 127 BERTNESS, M. D. AND R. CALLAWAY. 1994. Positive interactions in communities. Trends in Ecology and Evolution 9:191—193. BRESHEARS, D. D., N. Coss, P. RICH, K. PRICE, C. ALLEN, R. BALICE, W. ROMME, J. KASTENS, M. FLOYD, J. BELNAP, J. ANDERSON, O. MYERS, AND C. MEYER. 2005. Regional vegetation die-off in response to global-change-type drought. Proceed- ings of the National Academy of Sciences USA 102:15144-15148. BoweRS, J. E., R. H. WEBB, AND R. J. RONDEAU. 1995. Longevity, recruitment and mortality of desert plants in Grand Canyon, Arizona, USA. Journal of Vegetation Science 6:551—564. . 2005. Effects of drought on shrub survival and longevity in the northern Sonoran Desert. Journal of the Torrey Botanical Society 132:421-431. CASPER, B. B. 1996. Demographic consequences of drought in the herbaceous perennial Cryptantha flava: effects of density, associations with shrubs, and plant size. Oecologia 106:144—152. CHARIS PROFESSIONAL SERVICES CORPORATION. 2002. Distributions and abundance of the Lane Moun- tain milk-vetch (Astragulas jaegerianus Munz). Unpublished report to the U.S. Army National Training Center, Fort Irwin, CA. Copy, M. L. 1993. Do cholla cacti (Opuntia spp.., Subgenus Cylindropuntia) use or need nurse plants in the Mojave Desert? Journal of Arid Environ- ments 24:139—-154. Cook, E. R., C. A. WOODHOUSE, C. M. EAKIN, D. M. MEKO, AND D. W. STAHLE. 2004. Long-term aridity changes in the Western United States. Science 306:1015—1018. FLORES, J., O. BRIONES, A. FLORES, AND S. SANCHEZ- COLON. 2004. Effect of predation and solar exposure on the emergence and survival of desert seedlings of contrasting life-forms. Journal of Arid Environments 58:1-—18. FRANCO, A. C. AND P. S. NOBEL. 1989. Effect of nurse plants on the microhabitat and growth of cacti. Journal of Ecology 77:870-886. GIBSON, A. C., M. R. SHARIFI, AND P. W. RUNDEL. 1998. Ecophysiological observations on the Lane Mountain milkvetch, Astragulas jaegerianus (Fa- baceae), a proposed endangered species of the Mojave Desert. Aliso 17:77—-82. GODINEZ-ALVARE, H., T. VALVERDE, AND P. ORTEGA- BAES. 2003. Demographic trends in the Cactaceae. Botanical Review 69:173—203. GOLDBERG, D. E. AND R. M. TURNER. 1986. Vegetation change and plant demography in permanent plots in the Sonoran Desert. Ecology 67:695—712. HAMERLYNCK, E. P. AND T. E. HUXMAN. 2009. Ecophysiology of two Sonoran Desert evergreen shrubs during extreme drought. Journal of Arid Environments 73:582—585. AND J. R. MCAULIFFE. 2008. Soil-dependent canopy die-back and plant mortality in two Mojave Desert shrubs. Journal of Arid Environments 72:1793-1802. HEREFORD, R., R. H. WEBB, AND C. I. LONGPRE. 2006. Precipitation history and ecosystem response to multidecadal precipitation variability in the Mojave Desert region, 1893-2001. Journal of Arid Environments 67:13—34. 128 HOERLING, M. AND A. KUMAR. 2003. The perfect ocean for drought. Science 299:691—694. HOLMGREN, M., M. SCHEFFER, AND M. A. HUSTON. 1997. The interplay of facilitation and competition in plant communities. Ecology 78:1966—1975. JORDAN, P. W. AND P. S. NOBEL. 1979. Infrequent establishment of seedlings of Agave deserti (Aga- vaceae) in the northwestern Sonoran Desert. American Journal of Botany 66:1079—1084. AND . 1982. Height distributions of two species of cacti in relation to rainfall, seedling establishment, and growth. Botanical Gazette 143:511-517. MCDANIEL, K. C., D. B. CARROL, AND C. R. HART. 2000. Broom snakeweed establishment following fire and herbicide treatments. Journal of Range Management 53:239—245. MuritTi, M. N. 2006. Ontogenetic shifts from facilitation to competition in a desert shrub. Journal of Ecology 94:973-979. , 9. RODRIGUEZ-BURITICA, S. J. WRIGHT, AND H. F. Howe. 2007. Episodic death across species of desert shrubs. Ecology 88:32—36. MUELLER-DOMBOIS, D. AND H. ELLENBERG. 1974. Aims and methods of vegetation ecology. John Wiley and Sons, New York, NY. NOoLaAsco, H., F. WEGAVILLASANTE, AND A. DIAZ- RONDERO. 1997. Seed germination of Stenocereus thurberi (Cactaceae) under different solar irradia- tion levels. Journal of Arid Environments 36: 123-132. PAEZ, S. A. AND D. E. MARCO. 2000. Seeding habitat structure in dry Chaco forest (Argentina). Journal of Arid Environments 46:57—68. PARKER, K. C. 1993. Climatic effects on the regener- ation trends for two columnar cacti in the northern Sonoran Desert. Annals of the Association of American Geographers 83:452-474. PRIGGE, B. A., M. R. SHARIFI, AND D. MORAFKA. 2000. Lane Mountain milkvetch surveys (progress report IT & III). Unpublished report to Dominguez Hills Foundation, Carson, CA. PUGNAIRE, F. I. AND M. T. LUQUE. 2001. Changes in plant interactions along a gradient of environmen- tal stress. Oikos 93:42-49. RALPHS, M. H. AND J. E. BANKS. 2009. Cattle grazing as a biological control for broom snakeweed: vegetation response. Rangeland Ecology and Man- agement 62:38—43. REISMAN-BERMAN, O. 2007. Age-related change in canopy traits shifts conspecific facilitation to interference in a semi-arid shrubland. Ecography 30:459-470. RUNDEL, P. W. AND A. C. GIBSON. 1996. Ecological communities and processes in a Mojave Desert ecosystem: Rock Valley, Nevada. Cambridge University Press, Cambridge, UK. ———, B. A. PRIGGE, AND M. R. SHARIFI. 2005. El Nino events as a seedling establishment and soil seed bank cue for the Lane Mountain milkvetch (Astragalus jaegerianus Munz) during 2002—2003 El MADRONO [Vol. 57 Nino conditions. Unpublished report to the U.S. Army Research Office, CA. , AND . 2006. The effect of amount and frequency of precipitation on seedling establishment and survival of the Lane Mountain milkvetch (Astragalus jaegerianus Munz). Unpub- lished report to the U.S. Army Research Labora- tory, CA. , T. R. HUGGINS, B. A. PRIGGE, AND M. R. SHARIFI. 2009. Community dynamics and soil seed bank ecology of the Lane Mountain milkvetch (Astragalus jaegerianus Munz): a model rare plant species. Unpublished report to the U.S. Army Research Office, CA. SAS. 1999, StatView reference. SAS Institute Inc., Cary, NC. SHARIFI, M. R., B. A. PRIGGE, T. R. HUGGINS, AND P. W. RUNDEL. 2009. Survival and establishment of the Lane Mountain milkvetch, Astragalus jaegerianus (Fabaceae), an endangered species, under field and controlled greenhouse conditions. In press, in CNPS 2009 conservation conference proceedings: strategies and solutions. California Native Plant Society, Sacramento, CA. SHEPS, L. O. 1973. Survival of Larrea tridentata S. & M. seedlings in Death Valley National Monument, California. Israel Journal of Botany 22:8—17. SHREVE, F. 1917. The establishment of desert perenni- als. Journal of Ecology 5:210—216. SHUMWAY, S. W. 2000. Facilitative effects of a sand dune shrub on species growing beneath the shrub canopy. Oecologia 124:138—148. TURNER, R. M. 1990. Long-term vegetation change at a fully protected Sonoran Desert site. Ecology 71:464-477. U.S. FISH AND WILDLIFE SERVICE (USFWS). 1998. Endangered and threatened wildlife and plants; determination of endangered or threatened status for five desert milk-vetch taxa from California. Federal Register 63:53596—5361. VALIENTE-BANUET, A., A. BOLONGARO-CREVENNA, O. BRIONES, E. EZCURRA, M. ROSAS, H. NUNEZ, G. BARNARD, AND E. VAZQUEZ. 1991. Spatial relationships between cacti and nurse shrubs in a semi-arid environment in central Mexico. Journal of Vegetation Science 2:15—20. association between the cactus Neobuxbaumia tetetzo and the nurse plant Mimosa luisana in the Tehuacan Valley, Mexico. Journal of Ecology 79:961-971. WALKER, G. F. AND A. E. METCALF. 2008. Genetic variation in the endangered Astragalus jaegerianus AND E. EZCURRA. 1991. Shade as a cause of the | (Fabaceae, Papilionoideae): a geographically re- | stricted species. Bulletin Southern California Acad- emy of Sciences 107:158—177. WARNOCK, A. D., M. E. WESTBROOKE, S. K. FLOR- ENTINE, AND C. P. HuRST. 2007. Does Geijera parviflora Lindl (Rutaceae) facilitate understory species in semi-arid Australia. Rangeland Journal 29:207—216. MADRONO, Vol. 57, No. 2, pp. 129-135, 2010 A RESURRECTION FOR SISKIYOU BELLS, PROSARTES PARVIFOLIA (LILIACEAE), A RARE SISKIYOU MOUNTAINS ENDEMIC MICHAEL MESLER, ROBIN BENCIE, AND BIANCA HAYASHI Department of Biological Sciences, Humboldt State University, Arcata, CA 95521 mrm!1@humboldt.edu ABSTRACT We conducted a study of Prosartes parvifolia S. Watson, a rare Siskiyou Mountains endemic, currently known from only 15 sites in Del Norte Co., California, and Curry and Josephine counties, Oregon. We found that P. parvifolia is (a) fertile, (b) probably not of hybrid origin, and (c) distinct and worthy of recognition as a species. Unlike congeners, its flowers produce ovaries with a single locule, and are pollinated by bees that buzz pollen from connivent anthers. Nectar is not produced. We provide an expanded description, illustrations, and distribution map for P. parvifolia as well as a key to the Prosartes of northwestern California and southwestern Oregon. Key Words: Liliaceae, Prosartes parvifolia, rare plant, Siskiyou Mountains. Prosartes D. Don is small genus of North American reticulate-veined Liliaceae formerly treated as part of the Asian genus Disporum (Jones 1951; Utech et al. 1995). Sereno Watson (1880) described P. parvifolia S. Watson, a species with small campanulate flowers from the Siski- you Mountains. Although Howell (1903), Jepson (1909) and Peck (1961) accepted P. parvifolia as distinct from P. hookeri Torr. and P. smithii (Hook.) Utech, Shinwari & Kawano, other authors have regarded it as either a sterile hybrid between these two more widely distributed taxa (Jones 1951; Munz 1959) or as a minor variant of P. hookeri (McNeal 1993; Utech 2002). A recent discovery of a small population near Bear Basin Butte in Del Norte Co., California, prompted a re-examination of the taxonomic status of P. parviflora. Here we provide an expanded descrip- tion of the species, and show that it is fertile, clearly distinct from the other five members of the genus, and probably quite rare. TAXONOMIC STATUS Jones (1951) regarded Prosartes parvifolia as a probable hybrid between P. hookeri and P. smithii because “it occurs in an area where these overlap, it is morphologically intermediate be- tween them, and it is sterile’. In contrast, Utech (2002) argued that the “‘the known variation in P. hookeri unquestionably encompasses the mor- phology described for P. parvifolia.” However, P. parvifolia is not intermediate between P. hookeri and P. smithii, nor does it combine the traits of these two species in the mosaic-like fashion expected of a later-generation recombinant (Figs. 1, 2; Table 1; see next section). Moreover, flowers produce well-formed, apparently viable pollen, and although the ovaries of some flowers are abortive, we have observed fruits with fully filled seeds as well as seedlings at several sites. Thus, a hybrid origin seems unlikely. Although P. parvifolia resembles P. hookeri vegetatively, the two species differ consistently for several qualitative characters (Fig. 2, Table 1). In addi- tion, they are sympatric at several sites without intergradation. FLORAL MORPHOLOGY, POLLINATION BIOLOGY, AND RELATIONSHIPS Prosartes parvifolia differs from other Pro- sartes in floral morphology and _ pollinator reward. Species in the genus form three groups based on floral plan. (1) P. hookeri, P. languinosa (Michx.) D. Don, P. maculata (Buckley) A. Gray, and P. trachycarpa S. Watson have more-or-less spreading tepals comprising a turbinate to open perianth and long filaments with exposed anthers well separated from the style and stigma (Fig. 2D). The base of each tepal is nectariferous and deeply concave. (2) The tepals of P. smithii likewise produce nectar at their concave bases, but the tepals are erect and reflexed only at the tip, forming a more-or-less cylindrical perianth with a narrow opening. The erect filaments position the anthers inside the perianth tube just below the stigmas (Fig. 2G). (3) The tepals of P. parvifolia form a campanulate perianth (Fig. 2A). In contrast to the other five species, the tepals are not strongly concave at the base, and do not produce nectar. The bases are flat or shallowly concave with a lustrous green patch contrasting sharply with the white perianth. The filaments are short and erect, and the anthers form a loose cone around the base of the style, well below the stigma (Fig. 2B). The anthers are introrse. In northern California, flowers of P. hookeri and P. smithii are pollinated by bees (mainly Bombus) which probe tepal bases for nectar and 130 Fic. 1. collect pollen either passively or actively (Mesler personal observations). Although the shiny green tepal patches of P. parvifolia resemble nectaries, the flowers offer only pollen as reward. On three separate occasions (in different populations), we witnessed bumblebees buzz pollen from the anther cone as in sympatric Ericaceae (Gaultheria shallon Pursh, G. ovatifolia A. Gray, Vaccinium ovatum Pursh) (Mesler personal observations). Prosartes parvifolia differs most strongly from other Prosartes species in gynoecial morphology. The ovary has a single locule (not three) that produces | to 4 ovules (usually 2), which are attached near the base of a parietal placenta. The ovules are held erect so that the raphe lies next to MADRONO [Vol. 57 Habit of Prosartes parvifolia. French Hill Road, Del Norte Co., California. the placenta and the micropyle faces down (hypotropous-ventral; see Simpson 2006, Fig. 11.14). In contrast the ovules of P. hookeri and P. smithii are pendent from the top of an axile placenta; the raphe faces the placenta but the micropyle points up (epitropous-ventral). The ovules of P. lanuginosa appear to be likewise epitropous-ventral. The ovules of P. trachycarpa and P. maculata have been described as horizon- tal (Jones 1951; Utech 2002), with the micropyle below the funiculus (Jones 1951). An alteration in orientation (horizontal to erect) could convert such a pleurotropous-dorsal organization to the hypotropous-ventral plan seen in P. parvifolia (Simpson 2006, Fig. 11.14). The fact that the 2010] 5mm FIG. 2. MESLER ET AL.: SISKIYOU BELLS 13] Comparison of Prosartes parvifolia (A—C), P. hookeri (D—F), and P. smithii (G and H). A, B: intact flower and flower dissection. C: glandular hair from leaf margin. D: flower dissection. E: hair from leaf margin. F: hair from upper stem. G: flower dissection. H: hair from upper stem. Hairs and flowers are shown at the same scale. ovaries of P. parvifolia are relatively small and even abortive in some flowers, coupled with routine fruit abortion, probably contributed to the conclusion of some authors (e.g., Abrams 1923; Jones 1951; Munz 1959) that the species is completely sterile. The fleshy fruits of P. parvifolia typically produce two seeds, unlike other species in genus, which (except for P. /anuginosa) usually produce at least one seed per ovary locule (Jones 1951; Utech 2002). Developing fruits are strongly asymmetrical and remain slightly so at maturity, with the stylar scar offset from the tip. This asymmetry in conjunction with parietal (vs. central or basal) placentation suggests that the unilocular condition of P. parvifolia has resulted from suppression of two of the three original locules as opposed to the loss of septa to form a common chamber. Detailed anatomical and developmental studies will be needed to verify this interpretation. The strongly divergent floral traits of P. parvifolia make it difficult to assess its relation- ships to other members of the genus, but there is currently no reason to suspect a close affinity with P. hookeri, the prior taxonomic connection between the two taxa notwithstanding. The orientation of ovules of P. maculata and P. trachycarpa and their glandular trichomes pro- vide some hint of relationship with P. parvifolia, but resolution of the issue awaits molecular phylogenetic study and determination of chro- mosome number. KEY TO PROSARTES OF NORTHWESTERN CALIFORNIA AND SOUTHWESTERN OREGON The following key allows reliable separation of Prosartes hookeri, P. parvifolia, and P. smithii in northern California and southwestern Oregon. Vestiture traits are especially useful in the field because the diagnostic glandular hairs of P. parvifolia are seen easily on both juveniles and adults, and they persist throughout the season. However, these hairs shrink and twist upon drying, making their glandular character obscure on herbarium specimens. la. Leaf margins, stems, and pedicels with glan- dular hairs; filaments <1] mm, much shorter than dehisced anthers; fruits generally with 2 seeds, style scar offset from the apex ...... aoe ae ie ere ae ee Prosartes parvifolia lb. Leaf margins, stems, and pedicels hairy or not, but lacking glandular hairs; filaments >3 mm, longer than dehisced anthers; fruits generally with >3 seeds, style scar centered at the apex 2a. Leaf margins with numerous short, sharp, forward-pointing hairs; lower blade sur- 132 TABLE 1. MADRONO McNeal (1993), and Utech (2002). Vestiture Stem Leaf surfaces Leaf margin Perianth shape Tepals Color Shape Androecium Filament orientation and anther position (post-dehiscence) Filament length Anther shape Dehiscence Gynoecium Style Ovary X.s. Locule number Ovule number Ovule orientation Pollinator reward Fruits Color x.s. shape Position of stylar scar Chromosome number Geography P. hookeri simple or branched sharp hairs (Fig. 2F) both surfaces scabrous, with numerous short, simple sharp hairs short, forward-pointing hairs (Fig. 2E) turbinate, tepals spreading from the middle, base narrowed, obtuse pale-green, yellow-green, or white oblanceolate to elliptical, lower 1/3 to 1/2 deeply folded along midvein spreading, anthers held away from style, gen exserted or +/— equal to tepals >5 mm, longer than anthers, generally unequal at dehiscence oblong to lanceolate, apex tapered with a short, blunt mucro latrorse, anther walls folded back at maturity, often twisted exserted or +/— equal to tepals, 3 minute lobes surrounding central depression at apex weakly triangular, vertices rounded 3 2/locule pendent from top of placenta, micropyle facing up nectar and pollen red to orange-red +/— terete at tip 2n = 18 widely distributed in the mountains of the Pacific Northwest to the Rockies (disjunct in Michigan), 100—2000 m COMPARISON OF PROSARTES HOOKERI, P. PARVIFOLIA, AND P. SMITHI. Based partly on Jones (1951), P. parvifolia mix of slender simple glandular hairs + shorter, eglandular clavate hairs both surfaces smooth, with simple glandular hairs slender, spreading, glandular hairs (Fig. 2C) campanulate to narrowly campanulate, tepals recurved at tip, base tapered, acute to obtuse bright white elliptical, base weakly gibbous erect, anthers included, loosely connivent around lower half of style <1 mm, much shorter than anthers, equal lanceolate, apex narrowly acute introrse, anther walls not folded back exserted or +/— equal to tepals, unlobed, obscurely cleft on one side at apex +/— terete to slightly flattened l 2 (3, 4) [total] erect from base of placenta, micropyle facing down pollen only orange-red slightly flattened offset from tip not known probably rare, limited to the Siskiyou Mountains of Del Norte Co., California, and Curry and Josephine Cos., Oregon, 600 to 1525 m [Vol. 57 P. smithii branched sharp hairs (Fig. 2H) both surfaces smooth, upper surface glabrous, lower surface glabrous or with short simple or branched hairs glabrous or with slender, spreading, simple or branched sharp hairs cylindrical, tepals closely appressed, spreading slightly at tip to form narrow opening, base truncate cream-white to white oblong-lanceolate, lower 1/5 to 1/4 deeply folded along midvein erect, anthers included, surrounding upper part of style, immediately | below stigma lobes >5 mm, longer than anthers, equal oblong, apex blunt or notched latrorse, anther walls folded back included, 3-lobed, each lobe with an obscure adaxial cleft triangular, vertices acute 3 gen >2/locule pendent from top half of placenta, micropyle facing up nectar and pollen orange to orange-red +/— terete at tip 2n = 16 common near the coast, | from the San Francisco | Bay area to British Columbia, 0—1500 m 2010] MESLER ET AL. face scabrous; stigma unlobed ........ ee ee eee Prosartes hookeri 2b. Leaf margins glabrous or with slender spreading hairs; lower blade surface smooth; stigma three-lobed . . . Prosartes smithii REVISED DESCRIPTION Prosartes parvifolia S. Watson, Botany of Cali- fornia 2:179. 1880. Disporum parvifolium (S. Watson) Torrey. 1888. Bull. Torrey Bot. Club 15: 188.—Type: USA, California, Del Norte Co., between Happy Camp and Waldo, 16 June 1879, V. Rattan s.n. (holotype: GH 30030!; isotype: DS 49627!). Plants 10—75 cm tall, sometimes clumped, from deeply buried, often vertically oriented rhizomes; flowering individuals with 1 to several aerial shoots, each with 1 to 9 spreading (shade) to strongly ascending (sun) main branches, these branched 0 to 4 times. Stems densely glandular pubescent, with slender multicellular glandular hairs and shorter, eglandular, clavate hairs; base of stem with 2-4 densely pubescent cataphyll bracts, these sometimes subtending the first or second main branch. Foliage leaves sessile; blade broadly ovate to lance-ovate, less often lance- oblong or elliptic, flat (shade) or strongly folded (sun), 1.8-5.3 cm long, 0.6—-3.4 cm wide; apex acute to acuminate, base rounded to cordate, symmetrical to slightly oblique, often +/— clasp- ing (especially when subtending major branches), both surfaces with slender, erect multicellular glandular hairs especially along veins; margins flat or undulate (sun), with slender, spreading, multicellular, glandular hairs. Inflorescences with 1-4 flowers; pedicels 4-10 mm long, densely pubescent, with multicellular glandular hairs. Flowers 8—10 mm long, 6—8 mm wide, pendent, bright white, campanulate to narrowly campan- ulate; base of perianth tapered; tepals elliptical, broadly concave, apex recurved, acute, base weakly gibbous, shallowly concave abaxially or +/— flat, with a rounded or quadrate shiny dark green patch; outer tepals 9-10 mm long, 3—5 mm wide; inner tepals 9-10 mm long, 2-4 mm wide; stamens subsessile, +/— erect, forming a loose cone around the style; filaments short and broad, 0.4-0.8 mm long, 0.4 mm wide; anthers lanceo- late, introrse, basifixed, 3.6-4.3 mm long, 0.8 mm wide, apex narrowly acute, base sagittate; pollen white; style 8-10 mm long, sparsely pubescent or glabrous at base, not centered on apex of ovary, slightly curved, exserted =1 mm or slightly included, tip very obscurely cleft on one side; ovary small (sometimes abortive), pubescent to sparsely pubescent at top, 0.9-1.0 mm long, 0.7— 1.0 mm wide, weakly angled, asymmetrical (flat or grooved on one side, convex on the other), with one locule and 2 (3) ovules; placenta parietal; ovules erect, hypotropous-ventral. Fruits : SISKIYOU BELLS [33 fleshy, orange to orange-red, slightly flattened, not expanded equally around pedicel-style axis; stylar scar displaced to one side of apex, 10— 13 mm long, 8-10 mm wide; seeds (1) 2 (3), white, 5.5—6.5 mm long when fresh. SPECIMENS EXAMINED CALIFORNIA. Del Norte Co.: E base of Hazelview Summit grade, 19 May 1929, D. Kildale 7874 (CAS); Hazelview Summit, 25 May 1929, D. Kildale 9176 (CAS); French Hill Road, 21.6 mi from intersection with Hwy 199, cut-over Douglas Fir forest, 30 June 1970, J. P. Smith and S. Silva 4254 (HSC); Near Bear Basin, herb layer of evergreen conifer forest, 41°48'21.7", 123°44’11.1", 22 June 1979, G. L. Clifton and T. Griswold 5670 (HSC); T17N, R3E, sec 25, FS road 17N05, 1.5 mi N of road 17N04, UTM 435423 E, 4632830 N (NAD 27), elev. 1005 m (3300 ft), steep roadside below Lithocarpus and Pseudotsuga, 11 June 2006, M. R. Mesler 615 (HSC); TI7N R4E, sec 32, UTM 437032E, 4629798N (NAD 27), along unlabeled logging spur of FS road 17N05, 1.0 mi from intersection with FS road 17N04, about 100 m from the main road, elev. 1067 m (4000 ft), on edge of road and under Pseudotsuga and Chrysolepis, 10 June 2007, M. R. Mesler 762 (HSC); T16N, R4E, sec 4, UTM 437109 E, 4629757 N (NAD 27), FS road 16N02, 40 paces beyond spur to Bear Basin Lookout, below road, elev. 1524 m (5000 ft), understory of Abies concolor/Abies magnifica forest, 29 June 2006, M. R. Mesler 623 (HSC); TI8N, R3E, sec 1, UTM 434606 E, 4648843 N (NAD 27), elev. 945 m (3100 ft), logging spur running S from FS 4402, 0.1 mi E of county road 316, exposed roadside, 9 September 2006, M. R. Mesler 633 (HSC); UTM 434606 E, 4648843 N (NAD 27), elev. 945 m (3100 ft), logging road running S of FS 4402, 0.1 mile E of intersection with road 316. 1 September 2006, Mes/ler 634 (HSC); TI8N, R4E, sec. 9, UTM 438620 E, 4646536 N (NAD 27), elev. 700 m (2300 ft), county road 324, 4.4 mi from its western intersection with Hwy 199, E of Hazelview Summit, 17 May 2007, Mesler 755 (HSC); 41.826°N, 123.929°W, elev. 733 m (2404 ft), French Hill Road, 7.4 mi from Hwy 199, 29 May 2008, M. Simpson 3028 (HSC). OREGON. Curry Co.: Coast Mountains, 42nd parallel, 13 June 1884, 7. J. Howell (OSU); Bear Wallow Lookout, 4 June 1932, L. Leach 3548 (OSU); T40S, R11W, sec 9, UTM 416824 E, 4660822 N (NAD 27), elev. 580 m (1900 ft), at the end of road 330, running S from FS road 1107, 0.8 mi SE of intersection with road 334, 1 September 2006, M. Mesler 636 (HSC); T40S, RIOW, sec. 24, UTM 431559 E, 4657932 N (NAD 27), elev. 1100 m (3600 ft), Buckskin Peak trail, 14 September 2007, M. Mesler 789 (HSC). Josephine 134 Kalmiopsis Wilderness | V, MADRONO [Vol. 57 Area of detail Buckskin Peak O'Brien Oregon Mtn + California Hazelview Summit A Pa Presumed type locality Bear Basin Butte ® Kilometers FIG. 3. estimated position of the type locality. Co.: Hunter’s Camp, between Chetco Ridge trail and Rough and Ready trail, 26 June 1950, A. Kruckeberg 1972 (OSU); T41S, RIOW, sec. 13, UTM 431469 E, 4650295 N (NAD 27), elev. 1070 m (3500 ft), Wimer Rd (FS 4402), 1.0 mi E of intersection of 4402 and 4402.112, 14 Septem- ber 2007, M. Mesler 790 (HSC). DISTRIBUTION AND HABITAT Prosartes parvifolia is confined almost entirely to the Smith River watershed of the Siskiyou Mountains of northwestern California and south- western Oregon (Del Norte, Curry, and Jose- Distribution of Prosartes parvifolia. Points are currently known populations; the diamond shows the phine counties; Fig. 3). Exceptions are two populations east of the Kalmiopsis Wilderness | (Coast Ranges) and one near Buckskin Peak (Illinois River watershed). A putative population, identified by G. J. Muth in 1978 (Flora of Klamath Mountains, unpublished computer-gen- erated checklist, Pacific Union College, Angwin, CA) and located near El Capitan in Siskiyou Co. (Butler 00026 [PUA]) is P. hookeri. Plants grow on various metamorphic sub- | strates (not ultramafic soils) in shaded forest | understories and forest edges as well as on) adjacent exposed roadside slopes and at logged | and burned sites, at elevations from 600 to / | | | | 2010] 1525 m. The most common tree associate 1s Pseudotsuga menziesii (Mirb.) Franco, occurring in combination with Notholithocarpus densiflorus (Hook. & Arn.) Manos, Cannon, & S. Oh at lower elevations and Abies concolor (Gordon & Glend.) Lindl. ex Hildebr. var. /owiana (Gordon & Glend.) Lemmon and 4. magnifica A. Murray at higher elevations. Other common associates are Chrysolepis chrysophylla (Douglas ex Hook.) Hjelmq., Gaultheria shallon Pursh, G. ovatifolia A. Gray, Mahonia repens (Lindl.) G. Don, Quercus sadleriana R. Br. ter, and Rhododendron macrophyllum D. Don ex G. Don. HISTORY OF COLLECTION AND RARITY Prior to our study, Prosartes parvifolia had been collected only eight times. The type collec- tion was made by Volney Rattan in 1879 along the road connecting Happy Camp, California, and Waldo, Oregon. The species was collected five years later by Thomas Howell, probably in the same general area, and then again by Lilla Leach and Doris Kildale Niles in 1929 and 1932, respectively. Each of these inveterate explorers of the Siskiyou Mountains collected the species from just a single locality or pair of closely spaced localities. The most recent collection was made near Oregon Mountain in 1998 by Veva Stansell, who reports having encountered it only once over many years of exploration (local botanist, personal communication). Prosartes parvifolia qualifies as rare, at least by virtue of its very narrow geographical distribu- tion. Currently it is known from only 15 locations spread over an area of about 525 km’; the most distant pair of sites 1s separated by only 40 km (Fig. 3). We have re-discovered all of the historical collection areas with the exception of the type locality and a site visited by Kruckeberg in 1950 (see Specimens Examined, Josephine Co., OR) on the east side of the Kalmiopsis Wilder- ness that probably lies slightly north of popula- tions we found near Buckskin Peak. Based on field reconnaissance in 2006-2009, we estimate fewer than 500 reproductive-age individuals across the 15 known sites. Our estimates may be conservative since a good deal of the roadless, rugged terrain in the Siskyou/Klamath region remains poorly explored botanically (J. Sawyer, Humboldt State Univ., personal communica- tion). Nevertheless, if such a distinctive taxon were truly abundant, we believe the many avid botanists who have worked in the area would have encountered it much more commonly. The factors responsible for the apparent rarity of P. parvifolia are unknown. The species is not a strong habitat specialist. It occurs across a wide range of elevations on a variety of relatively productive substrates in association with varying mixes of trees, in both shade and sun. The same MESLER ET AL.: SISKIYOU BELLS 135 habitat settings are common throughout the Klamath and adjoining Coast Range Mountains. Logging and road construction may have con- tributed to population declines, but the paucity of early collections suggests that the species may have been rare historically. The largest, most floriferous plants grow on otherwise bare mineral substrate along road cuts, and the largest known population occupies a recently cut and burned Douglas fir forest. Pollination deficits might be expected given small population sizes, but we have found isolated individuals with heavy fruit crops. The major threats facing P. parvifolia appear to be its limited distribution and small population sizes. ACKNOWLEDGMENTS We thank Veva Stansell, John Sawyer, Erik Jules, Jen Kalt, Jenny Nyffeneger, James P. Smith, Gibby Muth, Steve Darington, an anonymous reviewer, Lola Bell and Violet for encouragement, information, help in the field and lab, and/or comments on the manuscript. We are indebted to the curators of the following herbaria for loans: CAS, OSU, and WU. We dedicate this paper to the early plant explorers who first collected Siskiyou Bells and, especially, to our good friend, John Sawyer, whose unparalleled knowledge of the Klamath Region and passion for field work has been an inspiration. LITERATURE CITED ABRAMS, L. 1923. An illustrated flora of the Pacific States. Vol. 1. Stanford University Press, Stanford, CA. HowELL, T. J. 1897-1903. Flora of northwest America. Self-published, Portland, OR. JEPSON, W. L. 1909. A flora of California. Vol. 1. Associated Students Store. University of Califor- nia, Berkeley, CA. JONES, Q. 1951. A cytotaxonomic study of the genus Disporum in North America. Contributions of the Gray Herbarium 173:1—39. MCNEAL, D. W. 1993. Disporum. Pp. 1192 in J. C. Hickman (ed.), The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. Munz, P. 1959. A California flora. University of California Press, Berkeley, CA. Peck, M. E. 1961. A manual of the higher plants of Oregon, 2nd ed. Oregon State University Press, Corvallis, OR. SIMPSON, M. G. 2006. Plant systematics. Elsevier, Amsterdam. UrecHu, F. H. 2002. Prosartes. Pp. 142-145 in Flora of North America Editorial Committee (eds.), Flora of North America North of Mexico. Vol 26. Oxford University Press, New York, NY. , Z. K. SHINWARI, AND S. KAWANO. 1995. Biosystematic studies in Disporum (Liliaceae-As- paragoideae-Polygonateae). VI. Recognition of the North American section Prosartes as an autono- mous genus. Memoirs of the Faculty of Science Kyoto University Series of Biology 16:1-41. WATSON, S. 1880. Prosartes parvfolium. Pp. 179 in W. H. Brewer, S. Watson, and A. Gray, Botany of California. Vol. 2. Little, Brown, and Company, Boston, MA. MADRONO, Vol. 57, No. 2, pp. 136-140, 2010 SEDUM VALENS (CRASSULACEAE), A NEW SPECIES FROM THE SALMON RIVER CANYON OF IDAHO CURTIS R. BJORK Stillinger Herbarium, University of Idaho, Moscow, ID 83843 crbjork@gmail.com ABSTRACT Sedum valens (Crassulaceae) is described from the Salmon River Canyon of central Idaho. Though it shares numerous morphological traits with Sedum borschii and S. leibergii, the species differs strikingly in having myriad leaves packed into rosettes as wide as 1 dm. The leaves are ciliate, a characteristic otherwise unknown in temperate North American Sedum, except in Sedum radiatum, a highly dissimilar species. Further distinguishing characteristics are found in leaf shape, phenology, fruit characteristics and in habitat. Key Words: Crassulaceae, Idaho, Salmon River Canyon, Sedum. In the northwestern United States, the genus Sedum L. (Crassulaceae) includes 20 native taxa as circumscribed by Clausen (1975), including 5 taxa endemic to the region: S. borschii (R. T. Clausen) R. T. Clausen, S. /anceolatum Torr. var. nesioticum (Jones) Hitchce., S. /eibergii Britt., S. moranii R. T. Clausen, and S. rupicolum Jones. Clausen (1975) indentified a distinct evolutionary lineage involving S. borschii and S. leibergii, along with the California and Oregon endemic S. radiatum S. Wats. and the more widespread S. stenopetalum Pursh. This group is characterized by open, obpyramidal, cymose inflorescences of yellow flowers, widely divergent fruit follicles and observed patterns of interspecies hybrid fertility. A group of populations in the Salmon River Canyon system (hereby referred to as Sedum valens) appears to belong to this lineage, sharing its morphological distinctions while at the same time bearing consistent differences from all other species. Within this group, S. valens appears to be closest to S. borschii and S. leibergii, sharing their papillate leaves, variable numbers of flower parts, and glandular-punctate follicles. TAXONOMY Sedum valens Bjork, sp. nov. (Fig. 1).—Type: UNITED STATES, Idaho, Idaho Co., Salmon River Canyon, 16.5 air km E of Riggins, 900 m W of the junction of Elkhorn Creek and the Salmon River, elev. 609 m, on granite and granitic sand on steep canyon walls, growing with Pinus ponderosa Dougl., Pseudotsuga menziesii (Mirbel) Franco, Holodiscus discolor (Pursh) Maxim., Philadelphus lewisii Pursh, Selaginella douglasii (Hook. & Grey) Spring, Micranthes occidentalis (S. Wats.) Small, Glos- sopetalon spinescens A. Gray, Heuchera gros- sulariifolia Rydb., and Cystopteris fragilis (L.) Bernh. 45°24’N, 116°6’W, 3 December 2003, C. R. Bjork 8008 (holotype: ID, isotype: WS). Paratypes: USA. IDAHO. Idaho Co.: Salmon River Canyon, 200 m E of the mouth of French Creek, 45°25’N 116°1’W, elev. 616m, C. R. Bjérk 8007 (ID); Salmon River Canyon, 2.5 km NNW of the Salmon River on west slopes above the Wind River, 45°28’N 115°56’W, elev. 840 m, C. R. Bjérk 8006 (ID); Salmon River Canyon, 45°26'N 115°57'W, elev. 614 m, C. R. Bjérk S005 (ID); Salmon River Canyon, 45°25’N 115°59’W, elev. 624 m, C. R. Bjérk 8004 (ID). Herba biennis, folliis plurimus et ciliaris, prolificae vegitativa praecox maturescens, flor- ibus multus aureus, inflorescentia ramosa, foli- culo glandulosi divergens. Biennial, light green or yellowish green herb. Basal rosettes 3—10 cm wide, with leaves numer- ous (87-188). Rosette offshoots maturing and | detaching by the time of anthesis of the flowering rosettes. Leaves narrowly oblanceolate, (8—) 16— 38 X 24.4 mm (measurements of largest rosette leaves on dried, pressed specimens of flowering rosettes), strongly flattened dorsiventrally, the blade weakly differentiated, strongly papillate with marginal papillae on the proximal 2/3 of the leaf conspicuously lengthened, forming unicellu- | lar cilia up to 1.3 mm long. Inflorescences erect, much-branched, the peduncle 70-115 mm tall. Flowers numerous per inflorescence (36—139). Petals yellow, 3.8-6.3 * 1.2—2.2 mm (measure- | ments from dried, pressed petals). Follicles widely | divergent, 4.0-7.7 mm long, 1.5—2.8 mm tall, glandular-punctate. Flowering in April to May. No other temperate North American Sedum | taxon is known to produce cilia except S. radiatum, a highly dissimilar non-rosettiform | species, and the papillate condition is found in | only a few species (Clausen 1975). While rosette width varies greatly in S. valens, the only other | BJORK: A NEW SEDUM FROM IDAHO 137 Fic. 1. Sedum valens. Upper left: young rosettes in June; upper right: inflorescence seen from above; lower left: habit in flower, the rosette is approximately 1 dm wide; lower right: habitat in the Snake River Canyon, showing the cliffy and woodland habitats of S. valens. western North American Sedum capable of producing rosettes as large as its maximum width are S. albomarginatum R. T. Clausen, which is endemic to the Feather River Canyon of Cali- fornia (Denton 1993), and S. oregonense (S. Watson) Peck of western Oregon and northwest- ern California. Both of these species differ greatly from S. valens in reproductive and vegetative morphology. Additionally, no other North American Sedum produces rosettes bearing a number of leaves approaching that found in S. valens. Leaf and rosette characters of S. valens are the most striking morphological distinctions from S. leibergii, S. borschii and all other Sedum. Individual plants of S. valens are remarkable for their large rosettes (to 10 cm across), formed by a maximum of nearly 200 narrowly oblanceolate leaves. Rosettes of S. /eibergiii are smaller, rarely reaching 5 cm wide, and are formed by no more than 40 leaves. Leaf shape of these two species 1s unusual among North American Sedum in being both several times longer than wide, and widest near the apex. Leaves of S. valens are strongly flattened dorsiventrally and lack a distinct blade, while those of S. /eibergii are broadly elliptical to terete in blade cross-section, and are spatulate with a well defined blade. Leaves of S. va/ens are more strongly papillate, and the marginal papil- lae on the proximal 2/3 of the leaf are lengthened 138 MADRONO \\ \ Cc : 3 > aa = t ay - ~~ VY Om) yt { Washington | |! py 2 L | Oregon a = ve ” Ps Fic. 2. Map of the ranges of Sedum J/eibergii (light gray with outlines, western), Sedum borschii (medium gray with outlines, eastern), and Sedum valens (black). The arrow points to the type locality of S. valens. The dark gray areas with outlines in the range of S. borschii indicate the region of overlap in the ranges of S: leibergii and S. borschii. conspicuously, forming unicellular cilia as long as 1.3 mm. Leaves of S. Jeibergii lack cilia. Otherwise among North American Sedum, only S. radiatum has distinctly ciliate leaves (Ohba 2009), but that species differs from S. va/ens in lacking basal rosettes and non-papillate fruit follicoles. Sedum borschii produces rosettes up to only 2 cm wide, formed of no more than 15 leaves, and its leaves are ovate, elliptical or lanceolate. The papillae of S. borschii occur mostly on the leaf margins and apex, but they are inconspicuous and never lengthen into cilia. Sedum valens inflorescences are larger than those of either S. /eibergii or S. borschii. They are taller, though this is due only to branch length, not to stem length, which is roughly the same as those of S. borschii and S. leibergii (Table 1). Flower number and degrees of division in the cymes are greater than in S. /eibergii or S. borschii. The size and number of flower parts do not differ significantly. Follicle dimensions of S. va/ens are TABLE 1. [Vol. 57 greater than those of S. /eibergii, and they are longer but equally wide to those of S. borschii. Phenological differences also distinguish S. valens from S. leibergii and S. borschii. Sedum valens produces offshoot rosettes that detach around the time of flowering, producing inde- pendent clones with fully formed, surficial rosettes that do not contract in the summer. Sedum leibergii also produces offshoots prior to flowering, but they remain attached to the parent rosette well after flowering, and often through | winter. These offshoot buds do not form mature, j leafy rosettes until late winter or early spring of | the following year. Prior to that time, they remain | pale, turion-like and subsurficial in moss mats. In winter, the old, senesced rosettes and flowering © stems of the previous summer almost always bear — at least one offshoot, while old stems of S. valens — never bear attached offshoots. Sedum borschii — produces mature offshoot rosettes by the time of | flowering, but the rosettes remain permanently © attached. This gives S. borschii a suffruticose — growth form. The known population of S. va/ens occurs on | siliceous rock of the Idaho Batholith, but at the © western end of its distributional range, it extends onto the batholith margins, on contact-meta- morphics with calcareous modification. Most S. borschii populations are also on granite, while S. leibergii is thus far known only on basalt and calcareous rocks. The range of S. borschii reaches to within 10 km of S. valens, but its populations © occur at least 500 m higher in elevation. Sedum borschii grows 1n montane to subalpine woods and rock outcrops, while S. va/ens occupies drier, warmer Pinus ponderosalPseudotsuga menziesii woodlands and canyon scrub communities. Se- | dum leibergii occurs mostly northwest of the — range of S. valens, but it also occurs disjunctly © eastward in Lemhi Co., Idaho (Fig. 2). Speci- mens of S. /eibergii from Montana (MONTU, WTU) are misidentified S. borschii. Sedum — leibergii grows at similar elevations, to within | about 10 km of S. valens, but no overlap in ranges has been observed. Sedum leibergii grows in hotter, drier, usually non-forested habitats and MEANS AND RANGES OF QUANTITATIVE CHARACTERS IN SEDUM VALENS, S. LEIBERGII AND S. BORSCHI. Measurements obtained from the type, paratype specimens and specimens of the other species as cited under “‘other specimens examined”’. S. valens 27.7 (14-62) 122.4 (87-188) 88.3 (70-115) 3.2 (112-5) 72.6 (38-118) 78.7 (36-139) Character Rosette leaf length (mm) Rosette leaf number Stem length (mm) Number of cyme divisions Inflorescence width Flower number Follicle length 5.5 (4-7.7) Follicle width 2.1 (1 5—2,8) Follicle length/width ratio 2.7 (2.2-3.7) S. leibergii S. borschii 17.0 (13124) 4.9 (2.3-7.5) 10.6 (12-36) 9.0 (6-14) 83.0 (50-143) 75.3 (43-105) 1.9 (3) 1.1 (1 (2) 44.9 (18-74) 21.3 (9-44) 19.8 (4-44) 6.5 (2-15) 2.8 (2.2-3.4) 30 02-5) HO (07-12) 2311.23) 29 (0.3-3.7) 17452) 2010] almost always in moss mats on ledges and in crevices, never in forest understory. Sedum valens also often grows in moss mats, but unlike S. leibergii, it frequently occupies soils and humus amid woodland understory vegetation. ECOLOGY Sedum valens appears to be limited to lower elevations in the Salmon River Canyon and tributary canyons. About half of the observed individuals of S. valens occupy duff over granitic sand in woodland understory with Pinus ponder- osa Dougl., Pseudotsuga menziesii (Mirbel) Franco, Holodiscus discolor (Pursh) Maxim., Philadelphus lewisii Pursh, Synthyris missourica (Raf.) Pennell, Carex geyeri Boat, Poa wheeleri Vasey in Rothr., and Cystopteris fragilis (L.) Bernh. The remainder grow on mossy ledges, crevices and cliff faces with Glossopetalon spines- cens A. Gray, Heuchera grossulariifolia Rydb., Micranthes idahoensis (Piper) Brouillet & Gornall, Sedum stenopetalum Pursh, Selaginella douglasii (Hook. & Grev) Spring., and Woodsia scopulina D.C. Eaton. In either case, it grows mostly on BJORK: A NEW SEDUM FROM IDAHO 139 north- and east-facing slopes. Few individuals occur on south- or west-facing slopes, suggesting that S. valens is best adapted to relatively cool, shaded conditions. The total range of S. valens could not be elucidated due to the extremely rugged terrain and nearly impassible slopes upstream from the easternmost populations encountered. The con- tinuance of suitable habitat eastward into these impassible areas suggests that S. va/ens extends beyond the area searched. No individuals were found in apparently suitable habitat in some tributary canyons however, including French, Elkhorn, or Partridge Creeks. Sedum valens has been found no higher than 1300 m elev. So far, fewer than 10,000 individual plants have been encountered in the study area. Despite the wilderness status of the potential habitat up- stream, S. valens may be a priority for conserva- tion given its limited known range, small populations, and its proximity to a well-traveled recreation road. Since the first discovery of S. valens, large portions of the population along the road have been destroyed during a road-widening project (Karen Gray personal communication). KEY TO SEDUM OF IDAHO (EXCLUDING RHODIOLA) la. Plants rhizomatous, forming dense to loose mats often >20 cm wide; leaves alternate, bright yellow-green, 3-5 X 3-3.5 mm; growing in disturbed sites, introduced.................. 0.00222 Sedum acre 1b. Plants not or only weakly rhizomatous, not forming mats, though often clustered; leaves alternate or opposite, color various, but not bright yellow-green, if as small as S. acre, then opposite; native species, mostly in undisturbed habitats................ Da WeaVes ODPOSILE Gx. ee fig cree Gwe 4 a: Ane Goa ch Ge Do, eaves altcimate o<. o2 4a -se8 ae oe eae ee 5 i — a eee er eee a eye ear ee Sedum debile 3a. Mature follicles erect; inflorescences domed; leaves broadest at the base, lacking buds on the HOWerNe StEMS. ca. 4 ge a ees oe ee Hee es 4a. Leaves of the flowering stems 49 (rarely to 20) mm long, ovoid to elliptical, slightly flattened, curving toward the stem; rare, canyons of central Idaho.......................4. ee ee ee ee eer or eee Sedum rupicolum 4b. Leaves of the flowering stems 7-20 mm long, linear or narrowly lanceolate, terete, not or scarcely curving toward the stem; common throughout the state ................... ae ee Sedum lanceolatum var. lanceolatum 3b. Mature follicles widely spreading; inflorescences obpyramidal; leaves variously shaped, but if broadest near the base, then buds numerous in leaf axils of the flowering stems.............. 5 5a Flowering stems with sterile buds in the leaf axils; leaves keeled, the midrib persistent after the leaf withers. ......66.-2..0c00e nes ee er re Sedum stenopetalum var. stenopetalum 5b Flowering stems lacking sterile buds in the leaf axils; leaves not keeled, the midribs withering with the leaves: ..< <<... ees bbe eed 6a. Plants suffrutescent; rosette leaves 2.3—-7.5 mm long; follicle length/width ratio 1.4—2.2; growing at elevations >1200 m.... ee eee ee ee ee ee ee Sedum borschii 6b. Plants not suffrutescent; rosette leaves 13 mm long or >; follicle length/width ratio at least 2.2; mostly growing at elevations <1000 m 7a. Rosettes with 12—36 leaves, contracted and turion-like through the summer drought; leaves subterete, with a distinct blade, never ciliate; not known from granite, never in forest understory, widespread . . pat Ss fees Gag het Meee g) GMA ieee ‘a+ Gat inden, 5 Se 4 Sedum leibergii 7b. Rosettes with 87-188 leaves, growing surficially as mature, leafy rosettes through the summer drought; leaves distinctly flattened, without a distinct blade, ciliate; mostly on granite, often in forest understory, Salmon River Canyon OTHER SPECIMENS EXAMINED Sedum borschii: USA. IDAHO. Idaho Co.: Meadow Creek, above Selway Falls, 31 May Sti es as Sedum valens 1936, Rollins 1661 (WS); Seven Devils Moun- tains, 27 June 1961, Clausen 61.178.8 (ID); Patrick Butte, 22 August, 1980, Wellner 2215 (ID). Custer Co.: Camas Creek drainage, Salmon 140 River Mts., 23 July 1982, Henderson 5312 (ID). Lemhi Co.: ca. 32 air mi NW of Challis, 9 June 1982, J. Civille 286 (ID); Bighorn Crags, 1 August 1990, Moseley 1931 (ID); Warm Spring Creek, 25 July 1980, S. P. Brunsfeld 1618 (ID). Valley Co.: Salmon River area ca. 9 air mi W of Loon Creek Point, 17 June 1982, Civille 299 (ID). MONTANA. Ravalli Co.: Bitterroot Mts., W above N Kootenai Lake, 26 July 1972, Lacksche- witz 3892 (WTU); Bitterroot Mts., above Bass Creek Falls, 21 August 1976, Lackschewitz 6879 (WTU). Missoula Co.: Rattlesnake Valley, 6 km NE of Missoula, October 1942, F. Rose C42-31 (WTU). Sedum leibergii: USA. IDAHO. Idaho Co.: Snake River 0.5 mi N of Willow Creek, 19 May 1976, Henderson 2947 (ID); Hells Canyon above Wild Sheep Rapids, 23 May 1976, Henderson 3034 (ID); 3/4 mi S along SR Trail from S end of Pittsburg Landing, 13 May 1990, Loraine 2048 (ID); cliffs above Salmon River, near Lucille, 16 May 1937, Christ 7280 (ID); rocky cliff, 2 mi up Race Creek, from the mouth, W of Riggins, 29 May 1965, Baker 16784 (ID); Hells Canyon, mouth of Bernard Creek, 24 May, 1974, Wellner 131 (1D); Whitebird, Vaughn 4581 (WS). Nez Perce Co.: rocky banks along the Snake River, 4 mi E of Lewiston, 25 May 1957, Baker 14794 (ID); Lewiston, 26 May 1900, Hunter 43 (WS); S side Clearwater, 29 May 1937, Meyer 870 (WS). OREGON. Crook Co.: Ochoco NF, Grids Creek Rd., 9 June 2000, Goff 00-03 (WS). Wallowa Co.: Deep Creek, 15 May 1936, Moore, W.R. 53 (WS). WASHINGTON. MADRONO [Vol. 57 Whitman Co.: at the head of Rock Lake, 1904, | Beattie 2398 (WS); Almota, 3 June, 1976, Old s.n. (WS); Wawawai, 20 June 1901, Piper s.n. (WS); Wawawai, 2 December 2004, Bjork 8130 (ID). Garfield Co.: Ilia Grade, 17 June 1913, Darlington s.n. (WS). Klickitat Co.: Rockland, 5 May 1898, Suksdorf s.n. (WS). Yakima Co.: Rattlesnake Mts., 16 July 1902, Colton 703 (WS). Asotin Co.: i 3 mi S of Asotin, 27 May 1944, Hitchcock C.L. | 5362 (WS). ACKNOWLEDGMENTS Thanks are due to the staff of the Stillinger Herbarium, University of Idaho, and the Ownbey | Herbarium, Washington State University, to Adolf & Oluna Ceska, Ann DeBolt, Alma Hansen, Roger Rosentreter, and Sara Stark for their help with specimens and cultivation, to Jason Hollinger for his very able help generating a map, and to Trevor Goward, Terry McIntosh and the reviewers for their comments on the manuscript. LITERATURE CITED CLAUSEN, R. T. 1975. Sedum of North America north — of the Mexican Plateau. Cornell University Press, Ithaca, NY. DENTON, M. F. 1993. Sedum. Pp. 531-534 in J. C. Hickman (ed.), The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. OHBA, H. 2009. Sedum. Pp. 199-222 in Flora of North America Editorial Committee (eds.), Flora of North America North of Mexico, vol. 8. Oxford University Press, New York, NY. MADRONO, Vol. 57, No. 2, pp. 141-144, 2010 ABIES MAGNIFICA VAR. CRITCHFIELDIL, A NEW CALIFORNIA RED FIR VARIETY FROM THE SIERRA NEVADA RONALD M. LANNER' 2651 Bedford Ave., Placerville, CA 95667 pinetree30@comcast.net ABSTRACT Abies magnifica A. Murray bis var. critchfieldii var. nov. Lanner (Critchfield red fir) is described. The new variety comprises the southernmost Sierra Nevada populations of California red fir. It differs from the typical variety in having smaller cones with protruding cone bracts. Because of the protruding bracts, populations of the new variety have been assumed to be disjuncts of the bracted A. magnifica var. shastensis Lemmon (Shasta red fir), described over a century ago from Mt. Shasta and considered present in NW California and SW Oregon. However, geographic patterns of morphological variation, artificial crossing results, and recent molecular studies indicate that Shasta red fir consists of California red fir introgressed by noble fir (A. procera Rehder), and that the new variety is not hybridized with noble fir. Key Words: Abies magnifica, Abies procera, California red fir, natural hybridization, Shasta red fir. Generations of investigators have been con- fused and intrigued by a complex consisting of California red fir (Abies magnifica A. Murray bis), noble fir (A. procera Rehder), and morpho- logically intermediate populations. California red fir, which ranges south down the Sierra Nevada and noble fir, which extends north into Wash- ington are clearly differentiated by leaf, bark, and cone characters (Lamb 1912; Lanner 1999). Between their ranges, however, lies a transition ‘zone that includes the southern Cascades, Kla- math Mts., and Coast Ranges of northwestern ‘California and southwestern Oregon. In this region, trees with intermediate morphology occur that resemble California red fir but whose cones have the long protruding (exserted) bracts similar to those of noble fir, as opposed to the hidden (included) bracts of California red fir cones (Figs. 1, 2). These populations with exserted bracts, extending from about Mt. Lassen in California to Crater Lake in Oregon have long -been referred to as Shasta red fir, A. magnifica var. shastensis Lemmon or even A. shastensis (Lemmon) Lemmon, the type locality for which is Mt. Shasta, California (Sargent 1898; Little (1979). Lemmon (1890) was apparently infatuated ‘with his new variety, or species as he later discerned it, whose “‘peculiarity ... 1s connected entirely with the fact of its cone-bracts becoming long and protruded, a half to a full inch between the scales, rendering the large purple cones, thus decked out with tasseled fringes, a most beautiful object”. _ 'The author is Visiting Emeritus Scientist at the ‘Institute of Forest Genetics, USDA Forest Service, Pacific Southwest Research Station, Placerville, CA. Remarkably, protruding bracts are also found in the southernmost Sierra Nevada populations of California red fir, about 480 km. from the nearest Shasta red firs to the north. These too have, historically, often been considered to be Shasta red fir (Sargent 1898; Sudworth 1908; Chase 1911; Jepson 1923; Peattie 1953; Griffin 1993; Stuart and Sawyer 2001), despite their geographic remoteness from the northern Shasta red fir area and the absence of any such intention in Lemmon’s varietal description (Lemmon 1890). The pattern of morphological variation of trees in the northern transition zone, more noble fir- like from south to north, and from east to west within that zone (Griffin and Critchfield 1972) suggests hybridization leading to introgression. Hybridization is further suggested by the ease of artificially crossing these firs, especially when California red fir is the maternal parent but in the reciprocal cross as well (Silen et al. 1965; Critchfield 1988). Liu (1971) found this evidence compelling enough to denote Shasta red fir as A. X shastensis Lemmon. Persuasive evidence of introgression has emerged also from recent molecular studies. Oline (2008) analyzed the distribution of chloro- plast haplotypes throughout the range of Cali- fornia red fir and within the transition zone extending into southern Oregon. Sierra Nevada populations, including the southernmost bracted ones, displayed only California red fir haplo- types. But the transition zone populations, including one from the type locality of Shasta red fir, were polymorphic, with haplotypes of both species. Oline (2008) viewed these results as “supporting a broad zone of hybridization”. Oline’s results undermine the concept of a 142 MADRONO Fic. 1. Mature seed cone of California red fir with hidden (included) bracts. This morphology exemplifies the typical variety. Drawing by Taylor in Sudworth (1908). distinctive Shasta red fir variety and strongly support viewing it as a series of hybridized and introgressed California red fir and noble fir populations—in effect a geographically wide- spread mature hybrid swarm. What then of the southern “‘Shasta red fir” whose protruding bracts are “identical in their shape with those of the north” (Sargent 1898)? Ustin (1976) reported that California red fir cones from eight locations south of the Kings River watershed (Panoramic Point, Rabbit Meadow, Montecito, Little Baldy, Mineral King, Holby Meadow, Sherman Peak, and Mule Peak) had protruding bracts. I have examined cones or cone parts from ten additional locations south of the Kings (Alta Peak, Panther Peak, Tar Gap, Kaweah River, Mountain Home State Forest, Greenhorn Mts. (presumably Sunday Peak), Siretta Ridge, Bald Mountain, Mineral King Valley 1, and Mineral King Valley 2) and found all bearing protruding bracts. Sudworth (1916) reported finding in 1899 trees bearing cones with all protruding bracts, and trees with all hidden bracts at Alta Meadows, in Sequoia National Park. This location should be further investi- gated. Jeff Bisbee (personal communication) has [Vol. 57. Fic. 54.—Abies magnifica FIG. 2. Mature seed cone of California red fir with protruding (exserted) bracts. This morphology exem- plifies the new southern Sierra Nevada variety (Critch- field red fir) as well as hybrid segregates with noble fir in NW California and SW Oregon (Shasta red fir). Drawing by Taylor in Sudworth (1908). observed and photographed protruding bracts at Onion Valley and the Kearsarge Pass trail. These locations fall between 35°47’N (Sunday Peak) and 36°46’N (Onion Valley) and from 2012 m elevation (Mountain Home State Forest) to 2850 m (Sherman Peak). Ustin (1976) found that cones from twenty | Sierra Nevada locations north of the Kings had hidden bracts. Nor have bracted cones been — reported from that area in field guides or floras I have consulted, though some show illustrations of bracted cones without explanation or com- ment (Storer and Usinger 1963; Storer et al. 2004). Sargent (1898), in what was perhaps the first published mention of the bracted southern red firs, pointed out that “‘in all the central part of the range occupied by this tree its cone bracts are | acute and included’’. The only apparently incon- sistent observations on this point are those of | cones with “slightly” protruding bracts at Onion | Valley campground (Inyo National Forest) where most of the cones had protruding bracts; and at Minaret Summit and Mammoth Lakes, where | they occurred north of the Kings in an area of | hidden bracts (Bisbee personal communication). Photographs show these cones have only the free | tips of their bracts visible. This may be evidence | of interbreeding between the new variety and the | 2010] | typical variety and should be examined in more | detail. Oline’s (2008) finding of only California red fir | haplotypes in the southern Sierra Nevada popu- _ lations is not the only evidence uncoupling these populations from northern Shasta red fir. In addition, the monoterpene composition of corti- cal oleoresins has shown the southern red firs to be chemically much more similar to the typical California red fir than to Shasta red fir of the northern transition zone (Ustin 1976; Zavarin et al. 1978). For these reasons it is appropriate to provide for the southernmost Sierra Nevada populations of California red fir a new variety. A NEW VARIETY OF ABIES MAGNIFICA _ Abies magnifica var. critchfieldii Lanner, var. nov. (Critchfield red fir; Fig. 2).—Type: USA, California, Tulare Co., Mountain Home State Forest, SW 1/4 SE 1/4 Sec. 25, T19S R30E, MDB & M, in mixed conifer forest on well- drained south slope, 6600 ft. (2012 m), 7 October 1947, L. 7. Burcham 260 (holotype: UC-907558 including separately filed cone coll. no. 0335). Abies magnifica var. critchfieldii ex var. magni- fica differt in strobilus parvis (9-17 cm vs. 14— 23 cm) cum squamae bracteae in maturitas siue siccitas reflexae. California red fir is a large forest tree to over 60 m tall. Young trees are pyramidal and symmetrical, old crowns become ragged from snow breakage. Leaves linear, 6-35 mm long and flattened on lower branches (shade leaves), 7— 40 mm long and quadrangular on upper branches (sun leaves), with 2 resin ducts, crowded, bent upward, new growth silvery-glaucous turning blue-green (thus local name “‘silvertip’’), with —stomates on all surfaces, apex blunt to acute, retained to at least 12 yr. The shortest needles surround terminal buds at their base and remain to mark the annual growth increments. Twigs pubescent, turning from yellow-green to light brown to gray. Winter buds ovate with acute to _ rounded apex, 2-8 mm long, light brown, shiny, not resinous. Bark thin, silvery gray, smooth with resin blisters on young stems; thick, reddish or _ purplish brown (thus “‘‘red fir’’), deeply furrowed between broad ridges on mature trees. Seed cones oblong or cylindrical, 14-23 cm long, 6-9 cm _ wide in the typical variety, 9-17 cm long, 3-9 cm wide in var. critchfieldii, purple tinged with | brown when mature, bracts hidden in typical variety but protruding conspicuously and reflex- ing when mature, finally covering much of the cone surface in var. critchfieldii. The variety is named in honor of William B. - Critchfield (1923-1989), American forest geneti- cist, in recognition of his distinguished contribu- LANNER: A NEW CALIFORNIA RED FIR VARIETY 143 tions to the genetics, systematics, biogeography, and evolution of western North American conifers, including the California red fir beneath which he enjoyed hiking in the Sierra Nevada. A native of Fargo, N. D., he earned a bachelor’s degree in forestry (1949) and doctorate in botany and genetics (1956) at the University of Califor- nia at Berkeley. After serving as forest geneticist with the Cabot Foundation for Botanical Re- search at Harvard University (1956-1959), he was a geneticist at the Institute of Forest Genetics, a unit of the USDA Forest Service’s Pacific Southwest Research Station, at Placer- ville, CA from 1959 to his retirement in 1988. Critchfield red fir is distributed in the southern Sierra Nevada Mountains in Tulare, Inyo, and Kern (and perhaps Fresno) counties, extending into the Greenhorn Mts. in Kern Co. It is found in Kings Canyon and Sequoia National Parks and Sequoia and Inyo National Forests. It therefore comprises the southern extremity (about 1 degree of latitude) of the range of California red fir (Griffin and Critchfield 1972). Common coniferous associates are white fir, A. concolor (Gordon & Glend.) Hildebr. var. /ow- iana (Gordon & Glend.) Lemmon; Jeffrey pine, Pinus jeffreyi Balf.; western white pine, P. monticola Douglas ex D. Don; lodgepole pine, P. contorta Douglas ex Loudon subsp. murrayana (Balf.) Critchf.; whitebark pine, P. albicaulis Engelm.; and Sierra juniper, Juniperus occidenta- lis Hook. subsp. australis Vasek. The type locality, Mountain Home State Forest in Tulare County, supports white fir, sugar pine (P. lambertiana D. Douglas) and giant sequoia (Sequoiadendron giganteum (Lindl.) J. Buchholz. Critchfield red fir, as reported to date, is similar in phenotype to the typical variety except for its smaller cones with protruding bracts. However, its marginal location with respect to the species’ range may be found upon further study to harbor adaptations to a drier climate than that of the typical variety. Protruding cone bracts occur in more than twenty firs worldwide, including three North American species in addition to noble fir. They also characterize all Pseudotsuga and_ several Larix (Eckenwalder 2009). In bristlecone fir (A. bracteata [D. Don] Poit.) very long attenuated bracts characterize the species as a whole (Lanner 1999). Balsam fir (A. balsamea [L.] Mill.) has long-bracted populations termed “‘bracted bal- sam fir” (var. phanerolepis Fernald), which occur sporadically from the Appalachians of Virginia and West Virginia to the Maritimes (Hawley and DeHayes 1985). Hybridization with the long- bracted Fraser fir (A. fraseri (Pursh) Poir) has been invoked to explain this occurrence (Liu 1O7 1), It is not surprising that protruding bracts — a trait widespread in its family and common in its 144 genus—should appear in a fir with ordinarily hidden bracts. Whether there is some selective advantage to a tree that has papery objects sticking out from between the scales of its seed cones, or if we are merely observing a neutral character occasionally expressed and subject to fixation through random drift in a marginal population, cannot be judged at this time. REPRESENTATIVE COLLECTIONS CALIFORNIA. Tulare Co.: Alta Peak, Ka- weah River Basin, 1901, Ralph Hopping s.n. (UC- 400343); Panther Peak, Sequoia National Park. October 1934, P. H. Bailey & W. W. Frost s.n. (UC-525811); Tar Gap, vicinity of Mineral King, 9000 ft, 5 August 5 1904, H. M. Hall & H. D. Babcock s.n. (UC-64470); Kaweah River, ca. 1918, Ansel Hall s.n. (JEPS-46605). Kern Co.: Greenhorn Mts., 7500 ft, 31 May 1947, Lyman Benson 1618 (SDSU-01567). ACKNOWLEDGMENTS I thank Edwin Royce for specimens from _ the southern Sierra and invaluable information about conifer growth conditions there; Corie Cann, L. McGinnis, and T. McGinnis for specimens from Sequoia National Park; Andrew Doran and Michael Simpson for access to the collections at the University of California and Jepson herbaria and the San Diego State University herbarium, respectively; Michael Frankis, Keith Rushforth, and Daniel Harder for nomenclatural advice; Christine Nelson for verifying Bill Critchfield’s biographical data; and Keith Rush- forth and Rebecca Rushforth for Latin translation. LITERATURE CITED CHASE, J. S. 1911. Cone-Bearing Trees of the California Mountains. A. C. McClurg and Co., Chicago. CRITCHFIELD, W. B. 1988. Hybridization of the California firs. Forest Science 34:139—-151. ECKENWALDER, J. E. 2009. Conifers of the World, The Complete Reference. Timber Press, Portland. GRIFFIN, J. R. 1993. Pinaceae, Pine Family. Pp. 115— 121 in J. C. Hickman (ed.), The Jepson Manual, Higher Plants of California. University of Califor- nia Press, Berkeley. AND W. B. CRITCHFIELD. 1972. The Distribu- tion of Forest Trees in California. USDA Forest Service Research Paper PSW-82, Berkeley. HAWLEY, G. J. AND D. H. DEHAYES. 1985. Hybrid- ization among several North American firs. 1. MADRCNO [Vol. 57 Crossability. Canadian Journal of Forest Research 15:42-49. JEPSON, W. L. 1923. Trees of California. Associated Students Store, University of California, Berkeley. LAMB, W. H. 1912. A synopsis of the red firs. Proceedings of the Society of American Foresters 7:184-186. LANNER, R. M. 1999. Conifers of California. Cachuma Press, Los Olivos. LEMMON, J. 1890. Variety shastensis Lemmon, the Shasta red fir. California State Board of Forestry Biennial Report 3:145. LItTLe, E. L., JR. 1979. Checkhst of United States Trees (Native and Naturalized). Agriculture Hand- book No. 541. Forest Service, U.S. Department of Agriculture, Washington. Liu, T. S. 1971. A Monograph of the Genus Abies. College of Agriculture, National Taiwan Universi- ty, Taipei. OLINE, D. K. 2008. Geographic variation in chloroplast haplotypes in the California red fir-noble fir species complex and the status of Shasta red fir. Canadian Journal of Forest Research 38:2705—2710. PEATTIE, D. C. 1953. A Natural History of Western Trees. Houghton Mifflin Company, Boston. SARGENT, C. S. 1898. The Silva of North America. Vol. XII Coniferae. Houghton Mifflin Company, Boston and New York. SILEN, R. R., W. B. CRITCHFIELD, AND J. F. FRANK- LIN. 1965. Early verification of a hybrid between noble and California red firs. Forest Science 11:460-462. STORER, T. I. AND R. L. USINGER. 1963. Sierra Nevada Natural History. University of California Press, Berkeley. ; , AND D. LUKAS. 2004. Sierra Nevada Natural History, revised edition. University of California Press, Berkeley. STUART, J. D. AND J. O. SAWYER. 2001. Trees and Shrubs of California. University of California | Press, Berkeley. SUDWORTH, G. B. 1908. Forest Trees of the Pacific Slope. U.S. Department of Agriculture, Forest | Service, Washington. Rocky Mountain Region. U.S. Department of Agriculture Bulletin No. 327. Washington, DC. UsTIN, S. L. 1976. Geographic Variation in Relative Cone Bract Length, Cotyledon Number and | Monoterpene Composition of Abies magnifica in | the Southern Sierra Nevada. M.A. thesis, Califor- | nia State University, Hayward. ZAVARIN, E., W. B. CRITCHFIELD, AND K. SNAJBERK. 1978. Geographic differentiation of monoterpenes | from Abies procera and Abies magnifica. Biochem- | ical Systematics and Ecology 6:267—278. Volume 57, Number 2, pages 77-144, published 30 September 2010 . 1916. The Spruce and Balsam Fir Trees of the | | SUBSCRIPTIONS — MEMBERSHIP The California Botanical Society has several membership types (individuals ($35 per year; family $40 per year; emeritus $27 per year; students $27 per year for a maximum of 7 years). Late fees may be assessed. 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 ($70). Membership is based on a calendar year only. Life memberships are $750. Applications for membership (including dues), orders for subscriptions, and renewal payments should be sent to 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 Manuscripts submitted for publication in MADRONO should be sent to the editor preferably as Microsoft Word (.doc), Rich Text Format (.rtf), or Portable Document Format (.pdf) files. It is preferred that all authors be members of the California Botanical Society. Manuscripts by authors having outstanding page charges will not be sent for review. Manuscripts may be submitted in English or Spanish. English-language manuscripts dealing with taxa or topics of Latin America and Spanish-language manuscripts must have a Spanish RESUMEN and an English ABSTRACT. For all articles and short items (NOTES, NOTEWORTHY COLLECTIONS, POINTS OF VIEW, etc.), fol- low the format used in recent issues for the type of item submitted. Allow ample margins all around. Manuscripts MUST BE DOUBLE-SPACED THROUGHOUT. For articles this includes title (all caps, centered), author names (all caps, centered), addresses (caps and lower case, centered), abstract and resumen, five key words or phrases, text, acknowledgments, literature cited, tables (caption on same page), and figure captions (grouped as consecutive paragraphs on one page). Order parts in the sequence listed, ending with review copies of illustrations. The title page should have a running header that includes the name(s) of the author(s), and a shortened title. Avoid foot- notes except to indicate address changes. Abbreviations should be used sparingly and only standard abbreviations will be accepted. Table and figure captions should contain all information relevant to information presented. All measurements and elevations should be in metric units, except specimen citations, which may include English or metric measurements. Authors are encouraged to include the names, addresses, and e-mail addresses of two to four potential reviewers with their submitted manuscript. Authors of accepted papers are required to submit an electronic version of the manuscript. Microsoft Word 2000 or later or WordPerfect 9.0 (or later) for Windows is the preferred software. Line copy illustrations should be clean and legible, proportioned to the MADRONO page. Scales should be in- cluded in figures, as should explanation of symbols, including graph coordinates. Symbols smaller than | mm after reduction are not acceptable. Maps must include a scale and latitude and longitude or UTM references. Presentation of nomenclatural matter (accepted names, synonyms, typification) should follow the format used by Sivinski, Robert C., in MADRONO 41(4), 1994. Institutional abbreviations in specimen citations should follow Index Herbariorum. Names of authors of scientific names should be abbreviated according to Brummitt and Powell, Authors of Plant Names (1992) and, if not included in this index, spelled out in full. Titles of all periodicals, serials, and books should be given in full. Books should include the place and date of publication, publisher, and edition, if other than the first. All 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 is 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. WUNNON AN 3 9088 01558 7827 F