VOLUME 57, NUMBER | INVASIVE HOLLIES (JLEX, AQUIFOLIACEAE) AND THEIR DISPERSERS IN THE PACIFIC NORTHWEST AAD) Gs OAS (4 OO ae Or Oe EIT EDEN PERE» UTES oe en ee OT DISTURBANCE, RESOURCES, AND EXOTIC PLANT INVASION: GAP SIZE EFFECTS IN A REDWOOD FOREST Brent C. Blair, Deborah K. Letourneau, Sara G. Bothwell, and Ga Caos CE [OR EA a ae rene eer ee on Nore ee MORPHOLOGICALLY CRYPTIC SPECIES WITHIN DOWNINGIA YINA (CAMPANULACEAE) LASG VM. SCHUUCIS 2. icc oscne PN Acts LO eB ee bcesscasissceaistvees POLLINATION AND REPRODUCTION IN NATURAL AND MITIGATION POPULATIONS OF THE MANY-STEMMED DUDLEYA, DUDLEYA MULTICAULIS (CRASSULACEAE) C. Eugene Jones, Frances M. Shropshire, Robert L. Allen, and YoussepGe Atlan ..,..ccec Gay OR ES ceed ov Wee de Aggy estaiaatsces chebectesans A NEw SPECIES OF DISTICHLIS (POACEAE, CHLORIDOIDEAE) FROM BAJA CALIFORNIA, MEXICO Flesteiale, Dell cg 7h MAL Po cccsei AGH op eacae walehee ne csscvcvsssoaseens CHENOPODIUM LITTOREUM (CHENOPODIACEAE), A NEW GOOSEFOOT FROM DUNES OF SOUTH-CENTRAL COASTAL CALIFORNIA Nuri Benet-Pierce’and Michael Gy SUNDSON .........5cccdecievesecccscsccesccccsnecsens DESERT WISDOM/AGAVES AND Cacti: CO,, WATER, CLIMATE CHANGE DOIG) COME SIF CLI ceccusavantianidiakers + Wore vssandtainens kc Wi erapumd aa entaaed acme JAX (ZA OLS WN aI ale OR EU a oO Oot od eae A ERE |) 2) RYT EY GAT) [ ee ie Sane tn) 2 > ENIAC Tye eo OS ora EO UO OIE CF JANUARY-MARCH 2010 MAbrRONO (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 MSCO03 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—FReED Hrusa, California Department of Food and Agriculture, Sacramento, CA RICHARD OLMSTEAD, University of Washington, Seattle, WA 2011—JAMIE KNEITEL, California State University, Sacramento, CA KEVIN Rice, University of California, Davis, CA 2012—GRETCHEN LEBUHN, San Francisco State University, CA ROBERT PATTERSON, San Francisco State University, CA 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. 1, pp. 1-10, 2010 oS MS Tm “CIBRARIES INVASIVE HOLLIES ULEX, AQUIFOLIACEAE) AND THEIR DISPERSERS IN THE PACIFIC NORTHWEST to ZUTL PETER F. ZIKA University of Washington Herbarium, Box 355325, Seattle, WA 98195-5325 Zikap@comcast.net ABSTRACT Naturalized //ex aquifolium L. (English holly) was first collected in the Pacific Northwest in 1953, based on herbarium records. Field surveys showed it is now commonly naturalized from northwestern California to coastal British Columbia. //ex crenata Thunb. and /. opaca Aiton were also found growing outside of cultivation, but rarely. A key and seed illustrations are provided to distinguish these three //ex species. Between 2003 and 2006 twice-weekly visits to naturalized and cultivated hollies in Seattle revealed seven species of birds disseminating seeds by eating the fruits. American robins, Turdus migratorius, accounted for 96% of 2796 frugivory observations on J. aquifolium, followed by European starlings, Sturnus vulgaris (3.2%). Ilex aquifolium fruits ripened in October and persisted for six months, yet 99% of all fruit was consumed between November and February. A study of I. aquifolium seed fate found pre-dispersal diurnal seed predation was rarely observed. Bird- regurgitated seed was more frequently attacked by nocturnal rodents in a sheltered forested setting in Clark Co., Washington (39% losses), compared to an exposed urban setting in Seattle (2% losses). The percentage of viable seed surviving rodent attack was higher in the urban sample (66%) than in the forest sample (24%). Commercial and ornamental use of /. aquifolium is extensive in the coastal region and less-invasive alternatives should be considered, to provide food and cover for urban avians without degrading natural areas. Key Words: American robin, English holly, //ex aquifolium, invasive plants, seed dispersal, seed predators, Turdus migratorius. Holly, the genus //ex, is the largest genus of woody dioecious plants, with more than 500 species worldwide (Cuénoud et al. 2000; Loizeau and Spichiger 2004). More than 30 holly species are cultivated in gardens in western North America, as well as a large number of named hybrids (Omar 1994; Galle 1997; Jacobson 2006). No native //ex species are found on the Pacific coast of North America. Ilex species are recently escaped (a non-native growing outside of cultivation, without human intervention) or naturalized (a non-native growing and reproducing outside of cultivation) in western North America. //ex ovaries ripen into a drupe, usually containing 3-4 nutlets (pyrenes). For convenience, I refer to these ecological dispersal units (diaspores) as fruits and seeds. Little is known about the interactions between //ex species and their seed dispersers and seed predators of the region, although these data can be important for dealing with invasive species. Therefore, in addition to investigating the collection history and distribution of escaped or naturalized //ex species in the region, prelim- inary studies on holly dispersal biology are reported here: 1.e., feeding behavior of frugivo- rous birds, and seed fate and viability after bird dispersal. At least eight English birds, including six thrush species, are known to disperse the seeds of English holly, Mex aquifolium L., in its native range (Snow and Snow 1988). Olmsted (2006) reported some unexpected species consuming holly fruit in the Pacific Northwest, such as (American) blackbirds and chickadees. I attempt- ed to reproduce her findings by systematically observing concentrations of fruiting holly species (naturalized and cultivated) in or near Seattle’s Washington Park Arboretum over three years, to resolve which birds were responsible for the most frugivory. In the settled landscape of southern England one study found frequent interactions between avian predators and their prey, flocks of fruit-eating birds, which affected fruit-gathering behavior (Snow and Snow 1986). So I recorded the behavior of urban American robin flocks when gathering fruit. Seed viability and the fate of seeds handled by birds were examined for possible effects of seed predators in two settings: an urban area and ina typical rural forest. I focused on the most widespread and invasive holly in western North America, //ex aquifolium, and asked what species ate the seeds by day, how frequently, and what percentage of seeds were destroyed by seed predators after they were transported by birds. Tlex aquifolium seed is protected by a thick bony exocarp (Fig. 1) and germination is delayed 18— 36 mo in Europe (Beckett and Beckett 1979; Arrieta and Suarez 2004). For comparison a three-year outdoor seed germination test was conducted in Seattle. N MADRONO [Vol. 57 2mm Fic. 1. Seeds of escaped hollies in the Pacific Northwest. //ex aquifolium, (a) lateral view, (b) proximal view. J/ex opaca, (c) lateral view, (d) proximal view. //ex crenata, (e) lateral view, (f) proximal view. METHODS Distribution Holly distribution data were compiled from the literature and specimens at the following herbar- ia: A, BM, CDA, CHSC, COCO, DAV, DAVFP, DBG, DECV, ELRG, FTU, GH, HSC, JEPS, KHD, LINN, MALA, NEBC, NLSN, NY, ORE, OSC, POM, RSA, SCCBC, SFUV, SOC, UBC, UC, UCR, UVIC, V, WILLU, WLK, WS, WTU, and WTUH (acronyms from Holmgren et al. 1990). Additional collections consulted includ- ed: the Shasta-Trinity National Forest, Redding, California; Reed College, Portland, Oregon; Olympic National Forest in Olympia, Washing- ton; The Evergreen State College in Olympia; and Fort Clatsop National Memorial, near Astoria, Oregon. The study area was broadly defined as the lowlands west of the Cascade Range in southwest British Columbia, western Washing- ton, and western Oregon. Populations were considered naturalized and mapped if they were obviously not planted and reproducing outside of cultivation, or if herbarium labels indicated they were not cultivated. Field surveys for naturalized holly were conducted on 50 d between 2000 and 2006. Herbarium vouchers from representative naturalized holly populations were deposited at WTU. Frugivory Studies The 21 holly taxa in Table 1 were studied at the edge of second-growth forest in the former holly plantings of the Washington Park Arboretum, part of the University of Washington Botanic Gardens in Seattle, King Co., Washington (Omar 1994), or areas within two km of the arboretum, including the University of Washington campus, and the adjacent Montlake neighborhood (AI- berti et al. 2001). Frugivory observations were made two times a week during daylight hours between December 2003 and March 2006, while walking to and through the grounds of the arboretum looking for bird activity. All observa- tions of animals eating fruits or seeds were recorded. Individual bird observations began when the first fruit was swallowed and ended when the bird stopped feeding and left the fruit source. It was soon evident that American robins were the most frequent frugivore to visit natu- ralized //ex, although this aspect of their natural history was not recorded in ornithological literature, so I compiled detailed notes of their feeding behavior. To estimate the transport of fruits and seeds, a count of total English holly fruits swallowed in one feeding bout was made for 25 American robins in Seattle. Ten large robin flocks were also timed (in minutes) when feeding on fruit, starting with the first bird perching on a 2010] TABLE 1. ZIKA: INVASIVE HOLLIES WW NUMBER OF OBSERVATIONS OF BIRDS SWALLOWING /LEX FRUITS IN THE PACIFIC NORTHWEST, 2004— 2006. Avians are American robin, European starling (ES), hermit thrush (HT), cedar waxwing (CW), American crow (AC), varied thrush (VT), and northern flicker (NF). Nomenclature follows Andrews (1997). Hex Robin ES x altaclerensis (Loudon) Dallim. 858 87 aquifolium L. 2690 90 aquifolium X cornuta P| < attenuata Ashe 168 x beanii Rehder 48 ciliospinosa Loes. 20 cornuta Lindl. & Paxton 43 cornuta X latifolia X pernyi oS cornuta X pernyi 5 crenata Thunb. 40 decidua Walter 146 dipyrena Wall. hybrid 10 integra Thunb. 14 < koehneana Loes. 18 latifolia Thunb. 3 maximowicziana Loes. l opaca Aiton 338 pernyi Franch. a7 serrata Thunb. 2 verticillata (L.) A. Gray 163 yunnanensis Franch. 6 Total observations 4732 7a Jo 95.12 3.56 fruiting branch and swallowing fruit, ending when the last individual departed. Most frugivory observations were made at close range or with Zeiss 7 X 42 binoculars. Fresh samples of ten fruits were gathered in Seattle and measured for each cultivated species in Table 1 (100 fruits of naturalized I. aquifolium), then the seeds were manually extracted, cleaned, counted and mea- sured, to determine the range of fruit and seed sizes and the average number of seeds per fruit. Seed Predation Seed predation was detected in several ways. Preliminary study showed birds usually swal- lowed holly fruits whole and departed, but seed predation was obvious when a bird lingered on the fruiting branch, mashed the fruit in its bill, slowly separating and dropping pulp while extracting, manipulating, and crushing seeds. Squirrels also sat on a fruiting branch, discarding fruit pulp and cracking seeds with their teeth, which was audible from 5 m. Seed predation by birds and squirrels was diurnal, producing small amounts of shredded fruit pulp where they attacked seeds. In contrast, evidence of nocturnal seed predation was indirect. The best evidence came from small gnawed holes in bird-regurgi- tated holly seeds on the ground, with no adjacent shredded fruit pulp. This was assumed to be (nocturnal) rodents feeding on seed contents; their preference for seeds over fruit flesh shown Avian HT CW AC VT NF _ Total Jo 4 949 19.08 i 11 2 | | 2796 56.20 a7 0.54 168 3.38 2 80 Lo! ] 21 0.42 | 44 0.88 2 97 1.95 5 0.10 3 43 0.86 | l 148 2.98 10 0.20 14 0.28 18 0.36 3 0.06 | 0.02 1 3 342 6.87 Ss] 0.74 2 0.04 1 164 3.30 6 0.12 42 19 2 2 l 4975 0.84 0.38 0.04 0.04 0.02 by untouched freshly fallen fruits within a few cm. Several times in Seattle I saw indications of nocturnal rodents (perhaps a rat sp.) climbing shrubs and feeding on the contents of seeds of Cotoneaster franchetii Bois, leaving large amounts of fruit flesh and broken seed husks below the shrub, with many shredded and seedless fruits remaining on the branches. Fruit- ing hollies were checked for evidence of similar arboreal seed predation by rodents throughout the study, in daylight hours; no direct nocturnal observations of rodents were attempted. Seed Viability Seed viability for lex aquifolium was deter- mined from freshly regurgitated seeds at sites where American robin frugivory was observed along sidewalks and lawn edges in Montlake, Seattle, as well as from second-growth Pseudo- tsuga menziesii (Mirb.) Franco forest near the high school in Camas, Clark Co., Washington. A sample of 500 regurgitated seeds was gathered in Montlake in January 2004, planted in one cm of soil in unirrigated pots left outdoors, and monitored for 3.5 yr to record length of time to germination (Barnea et al. 1991). Additional seeds from the same sites were scored for rodent damage, consisting of a gnawed exocarp and missing embryo. Undamaged seeds were halved with a razor and examined with a dissecting microscope. Grey firm embryos were scored as 4 MADRONO Fic. 2. Distribution map of naturalized [lex aquifo- lium in western North America, based on herbarium specimens. A few records extend beyond the map boundaries, to the north tip of Vancouver Island in British Columbia (50°35'N, 126°56'W; Zika 22740 V), [Vol. 57 viable seeds. Liquid, discolored, blackened, or absent embryos were scored as inviable seeds. RESULTS AND DISCUSSION Distribution Literature, herbarium records, and field obser- vations showed three holly taxa escaped from cultivation in northwestern North America: J/ex aquifolium (English holly), £ crenata Thunb. (Japanese holly), and Z opaca Aiton (American holly) (Zika and Jacobson 2005). Their seeds are illustrated in Fig. 1. A key is provided to separate them. Key to //ex Growing Outside of Cultivation in the Pacific Northwest 1. Leaves less than 30 mm long, less than 15 mm wide, minutely dentate, never spiny; fruit black, 4.8-6.5 mm _ diam.; seeds nearly smooth I. crenata 1’ Leaves more than 40 mm long, more than 20 mm wide, entire or spiny-margined; fruit red, 7-13 mm diam.; seeds grooved and strongly ridged 2. Fresh leaves scarcely shiny or dull above; pistillate flowers solitary and scattered on thE CWS «508 oP as, ches ee ee I. opaca 2' Fresh leaves glossy above, pistillate flow- ers clustered on short spurs, in subumbels of 1-8 I. aquifolium Only /lex aquifolium was abundant enough to represent a conservation concern in the Pacific Northwest; the other hollies were documented as escapes at just one location each. I. opaca was vouchered from a single escaped sapling in King Co., Washington (Zika 20447 WTU). Ilex crenata was restricted to two small shrubs on a brushy pondshore in Snohomish Co., Wash- ington (Zika 20423 & Jacobson WTU). Ilex X attenuata Ashe (I. cassine L. X opaca) was collected as an escape once in 1977 in Sacramento Co., California (Hrusa et al. 2002), but was not recorded escaped in the study area, even though it fruits in cultivation in Seattle. In the Pacific Northwest I found //ex aquifo- lium was thoroughly naturalized at low elevations west of the Cascade Range (Fig. 2). I found it reproducing outside of cultivation at hundreds of locations, including forests of all age classes, dominated by Picea sitchensis (Bong.) Carriére, Pseudotsuga menziesii, Acer macrophyllum Pursh, Alnus rubra Bong., or Populus balsamifera L. subsp. trichocarpa (Torr. & A. Gray) Brayshaw. English holly varied from infrequent to common and south to Monterey Co., California (36°36’N, 121°54’W; Zika 23683 RSA). 2010] ZIKA: INVASIVE HOLLIES 5 TABLE 2. MONTHLY FRUGIVORY OBSERVATIONS FOR JLEX AQUIFOLIUM, 2004—2006. Avian Oct Nov Dec Jan Feb Mar Apr—Sept Total % American robin 7 314 632 872 862 2 ] 2690 96.21 European starling 23 51 11 > 90 322 Hermit thrush 1 l 0.04 Cedar waxwing 5 6 1] 0.39 American crow 2 2 0.07 Varied thrush | l 0.04 Northern flicker l 0.04 Total observations a 337 683 884 876 2 7 2796 % 0.25 12.05 24.43 31562 31.33 0.07 025 in fencerows, thickets, roadsides, lakeshores, and floodplains. The majority of naturalized plants were found in rural, suburban, or urban wood- lots, fencelines, and hedges, where nearby pistil- late cultivated plants provided a seed source. Plant density was highest in some urban green- belts, with young stands of Pseudotsuga and an understory dominated by naturalized I. aquifo- lium rather than native shrubs. Ilex aquifolium Collection History English holly was introduced to the Pacific Northwest as an ornamental by 1869 (Ticknor 1986). Fruiting boughs were popular yuletide decorations, so by 1891 the species was estab- lished in commercial orchards. A regional indus- try continues to this day, providing an estimated 85% of the world’s crop of cut branches, which totaled 300 tons in 1963 (Ticknor 1986). lex aquifolium was first noted naturalized in the Pacific Northwest by Brayshaw (1960) and Taylor and MacBryde (1977). Plants were appar- ently uncommon at first and the species was not included in local and regional floras of the time (e.g., Hitchcock and Cronquist 1961, 1973; Szczawinski and Harrison 1972; Creso 1984). The oldest herbarium collection is dated 1953 (Vancouver Is., M. C. Melburn s.n. V). Prior collections, such as a 1931 sheet from the Columbia River Gorge (Yuncker & Welch 3703 NY) presumably represent cultivated plants as their labels do not specifically state they are escapes. English holly was mentioned as a locally frequent garden escape in British Columbia ‘“‘on south Vancouver Island, [and] less frequent on the lower mainland” (Douglas et al. 1989). Within a decade it was reported as “‘frequent in southwestern British Columbia” (Douglas et al. 1998), indicating it was spreading rapidly. In California, 1 aquifolium was absent from state floras (e.g., Munz and Keck 1965; Munz 1968) until recorded from the northern coast by McClintock (1993). A naturalized plant was first collected in 1976 in Humboldt Co. (Barker 1594 HSC). In treatments of Oregon plants, Peck (1961) and Thilenius (1968) did not include the species. The first Oregon record was collected in 1986 (Zika 9818 OSC). More recently Gray (2005) noted Z. aquifolium was naturalized in both disturbed stands and old growth forests at low elevations west of the Cascade Range. My field surveys disclosed I. aquifolium was natural- ized in every urban area in western Washington, although the first herbarium gathering was only in 1987 (Carnevali 203 ELRG). Ilex aquifolium was also reported naturalized in Hawai‘l (Wagner et al. 1999), New Zealand (Williams and Karl 1996), and Australia (Gleadow and Ashton 1981). Olmsted (2006) reported the species naturalized on the coast of New England, but there are no vouchers at NEBC (R. Angelo, New England Botanical Club herbarium, personal communication), and the report is dismissed here as a mistake for native populations of I opaca. Frugivore Studies The hollies studied (see Table 1) have fruits 5— 13 mm diam. and seeds 2—5 mm diam. Appar- ently none were too large to be swallowed by the local frugivorous birds; seven species were observed swallowing the fruits of the 21 J/lex taxa, including cultivated hybrids (Table 1). Native birds were the primary consumers of fruit, but 4% of the feeding observations repre- sent introduced European starlings (Sturnus vulgaris). Indigenous birds swallowing //ex fruits, in order of frequency, include: American robin (Turdus migratorius), hermit thrush (Catharus guttatus), cedar waxwing (Bombycilla cedorum), American crow (Corvus brachyrhynchos), varied thrush (/xoreus naevius), and northern flicker (Colaptes auratus). Robins, often flocking in winter, consumed 95% of the fruit of all combined J//ex taxa. Olmsted (2006) reported /. aquifolium fruits in the Pacific Northwest provide food for, among others, *‘...blackbirds, mourning doves, finches, chickadees and non-native house sparrow,” but did not provide supporting data, and I was unable to confirm her reports in this study. Those birds were common and seen near or in holly during the three years of field observations, but they ignored //ex fruits. American robins (Table 2) were responsible for 96% of fruit consumption observations for Ilex 6 MADRONO aquifolium (n = 2690), and accounted for 99% of the frugivory observed on /. opaca (n = 338), and 93% of the frugivory observed on /. crenata (n = 40). Robins were common year-round residents in all habitats with naturalized //ex (Sallabanks and James 1999). Published literature documents robins eating the fruits of 1. opaca, I. verticillata (L.) A. Gray, and /. decidua Walter (which are important wildlife food in eastern North Amer- ica, see Martin et al. 1951), but not the other //ex species in Table 1. These results suggest that pest control pro- grams for non-native birds like rock pigeons (Columba livia) and starlings would have a negligible effect on the dispersal of naturalized holly. On the other hand, the data present a strong argument that urban populations of American robins eat a great deal of //ex in their winter diet, resulting in considerable dispersal of seed into urban thickets and woodlots. American Robin Feeding Behavior The local movement of J/ex aquifolium seed was easily observed when American robins foraged in the study area between November and February (Table 2), before the onset of spring breeding and a shift in diet to consumption of more invertebrates (Wheelwright 1986). Rob- ins typically foraged in loose flocks of 5—75 birds. Part of the flock advanced towards a fruit tree, in stages, finally arriving and feeding rapidly. Returning to one or several prominent arboreal perches to process the fruit, they maintained a predator watch as a group (Howe 1979; Snow and Snow 1986, 1988; Fleming 1988) before returning to the fruit source. I refer to these lookout points as “relay trees.’’ Flock members repeatedly advanced from the relay tree(s) to feed on fruit. Holly berries were taken while perched, or occasionally snatched in flight. A few fallen fruits were consumed on the ground, or snapped with the bill by leaping from the ground to a low branch. Occasionally fruit was carried away in the bill, and either swallowed or dropped from a new perch. On Jlex aquifolium, feeding bouts for individ- ual American robins in flocks averaged 44 sec, with a range of 10-115 sec. A sample of 100 English holly fruits gave an average of 3.9 seeds per fruit. My observations of 25 robins feeding on I. aquifolium gave an average of 5.2 fruits swallowed, or an estimated 20.3 seeds (3.9 seeds/fruit < 5.2 fruits = 20.3 seeds) per feeding bout. //ex aquifolium has relatively large seeds, 5— 8 mm. Presumably most were regurgitated within ca. 15 min and very few seeds were defecated (Murray et al. 1993). Flocks of foraging robins were observed swallowing large numbers of fruits and seeds. In one observation, as many as 157 robins fed [Vol. 57 | undisturbed over a 30 min period on Ilex | aquifolium, resulting in removal of an estimated | 3187 seeds (20.3 seeds/bird X 157 birds = 3187 | seeds). For the flock, this observation represents a | potential removal rate of 106 seeds/minute (3187 | seeds + 30 min = 106 seeds/minute). In another | observation, 122 robins fed on I. aquifolium over | 20 min before scattering at the approach of their major avian predator, a Cooper’s hawk (Accipiter | cooperi). A similar calculation showed they | transported an estimated 2476 seeds, removing | approximately 124 seeds/min. | American robin flocks commonly used relay — trees 10-50 m from the fruit source. Flock | members moved holly seeds to many locations, © as some birds varied their approach and depar- — ture vectors, or fed on more than one fruit species | (Kwit et al. 2004). Winter soils were usually | unfrozen in the study area, so some flock members occasionally interspersed frugivory with foraging for invertebrates along brushy edges and in lawns, transferring seeds to additional sites. — When a predator alarm was given the birds fled, — resulting in some robins carrying seeds 500 m | before regurgitation. These observations are consistent with those of Holthuijzen and Sharik (1985), who found flock-feeding birds such as American robins, European starlings, and cedar waxwings facilitated long-distance dispersal of large quantities of seed when present. My observations suggest the variable feeding behav- ior of American robin flocks, with the use of different relay trees, make them effective dispers- ers for Ilex aquifolium (Schupp 1993; Jordano and Schupp 2000). Seed Germination, Predation and Viability Germination of [lex aquifolium seed is delayed 18—36 mo in Europe (Beckett and Beckett 1979; Arrieta and Suarez 2004). Regurgitated seeds I gathered and planted January 2004 germinated 29 mo later. Thousands of regurgitated J. aquifolium seeds were found under relay trees in Seattle during the study, and it was common to see American robins regurgitate the seeds after feeding on holly fruits. Seedlings were frequent in these sites. I found regurgitated seeds showed no physical differences from seeds extracted from fresh fruits, as did Meyer and Witmer (1998). Diurnal seed predation of J/ex species was rarely observed in the three years of the study, and is apparently insignificant before dispersal. The seeds of J. yunnanensis Franch. were taken © from fresh fruits once by a spotted towhee (Pipilo maculatus). Similarly, introduced eastern gray squirrels (Sciurus carolinensis) fed on Ilex fruits in Seattle, loudly cracking open the seeds while discarded pulp accumulated below the tree. Squirrels were seen and heard eating the seeds — of I. X altaclerensis (Loudon) Dallim. (n = 15), | 2010] use a gb: “erent sr apesagaamamaas . ee Beat Tae ce ‘aie 2 2 tid Ww he YS ya Es, ie J Rodent seed predation of J//lex aquifolium, showing gnawed holes and missing embryo, with mm scale. I. aquifolium (n = 5), I. cornuta X pernyi (n = 3), I. decidua (n = 2), and I. opaca (n = 1). In Spain, Obeso (1998) found evidence noc- turnal rodents were climbing trees and taking seeds from fresh //ex aquifolium fruits. I sought similar evidence from nocturnal visitors, such as abundant discarded fruit pulp and gnawed seed cases on the ground directly below numerous slashed and damaged seedless fruits still attached to pedicels on the branches. Diurnal seed predators observed in the study (squirrels and birds) never produced similar displays, they always picked the fruit before removing the seeds. I was able to observe nocturnal rodent damage a few times on fruits of cultivated Cotoneaster franchetii in Seattle, but never on holly, although I examined thousands of fruiting holly branches during daylight hours. I did not attempt direct nocturnal observations of rodents interacting with //ex fruits. In forested settings bird-regurgitated seeds were easiest to find under naturalized pistillate holly trees, as in Europe (Alcantara et al. 2000; Obeso and Fernandez-Calvo 2003). Post-dispers- al seed predation, evidenced by small gnawed holes and a missing embryo (Fig. 3), was assigned to small nocturnal rodents such as mice (Jones and Wheelwright 1987; Garcia et al. 2005). This type of seed damage differed from diurnal seed ZIKA: INVASIVE HOLLIES 7 TABLE 3. NUMBER OF VIABLE SEEDS OF J/LEX AQUIFOLIUM AFTER RODENT PREDATION, IN URBAN (SEATTLE, KING CO.) AND FORESTED (CAMAS, CLARK Co.) SITES, DETERMINED BY SECTIONING REGURGITATED SEED SAMPLES. Forest Urban Seed type N I N I Rodent damage 300 39 43 2 Viable 184 24 1403 66 Inviable 285 37 680 32 Total 769 2126 predation as practiced by squirrels, which left accumulations of discarded pulp. Squirrels ex- tracted seeds from fresh fruit picked and held in the forepaws, and their damage also differed in that they seemed to crush or crack open J//ex aquifolium seeds instead of gnawing small holes in them to remove the embryo. Although they may occasionally do it, I never saw squirrels gather or eat scattered regurgitated holly seeds on the ground. So I scored the damage shown in Fig. 3 as nocturnal, not diurnal, rodent seed predation. My examination of regurgitated //ex aquifo- lium seeds in woodland settings and edges invariably showed significant nocturnal rodent predation, as noted in Europe (Smal and Fairley 1982; Obeso 1998; Kollmann and Buschor 2002). Nocturnal rodents damaged 39% of regurgitated I. aquifolium seed sampled on the ground near pistillate 1 aquifolium trees in sheltered forest and forest edge settings in Clark Co. (Table 3). These results are qualitatively similar to studies of /. opaca in the eastern United States (Kwit et al. 2004) and LI aquifolium in Spain (Garcia et al. 2005; Arrieta and Suarez 2005). In contrast, Seattle’s urban walks, lawns, and hedges near cultivated //ex trees had a substantial seed rain that was largely ignored by seed predators, with only 2% post-dispersal seed predation by noctur- nal rodents. The general lack of cover and suppressed seed predation together suggest a powerful nocturnal predator influence, possibly urban cats (Crooks and Soulé 1999; Haskell et al. 2001). A secondary effect on seed viability may also result from the differences in post-dispersal seed predation in forested and urban sites. In the forested sample (n = 769), 76% of the regurgi- tated seed was either damaged by nocturnal rodents or was inviable (Table 3). In the urban sample (n = 2126), only 34% of the seed was either damaged by nocturnal rodents or was inviable. Said differently, 24% of surviving seed was viable in the forest, compared to 66% in the urban sample. Nocturnal rodents may be able to detect and ignore inviable seed in the forest, and apparently are unable or unwilling to attack viable seed in exposed situations in urban settings. Although these are small samples, the 8 MADRONO seed studies suggest post-dispersal seed predation is negligible for bird-disseminated holly seed in cities, and may provide a partial explanation for the relative success of J. aquifolium in urban and residential areas (Kollmann 2000). Conservation and Horticultural Implications Invasive woody plants in North America raise numerous conservation concerns, altering plant communities and displacing the native biota (Catling 1997; Pimental et al. 2000; Friedman et al. 2005; Reinhart et al. 2006). I/ex aquifolium is dispersed by the ubiquitous American robin and colonizes forests, edges, and settlements. It represents a long-term management problem in natural areas (Mack et al. 2000; Reichard and White 2001; Dlugosch 2005). As Temple (1990) and Low (2002) discussed, attempts to control or restrict sale of invasive but popular ornamentals like I aquifolium are not always welcomed by gardeners or distributors. Improved public out- reach and education are needed, as are ecologi- cally benign substitutes. Table 1 suggests winter- fruiting //ex alternatives exist in the garden trade. They are attractive ornamentals offering cover and food for winter bird flocks, but are apparently non-invasive, as measured by the lack of herbarium records of plants collected outside of cultivation, and an absence of seedlings around irrigated pistillate plants in gardens or arboreta. These all are in contrast to 1. aquifo- lium, with many adventive herbarium vouchers and which produces numerous seedlings in the immediate area of pistillate plants. However, potential hybrid replacements for 7. aquifolium in Table 1 should be tested for seed viability (perhaps a proxy for invasiveness), and moni- tored for their capacity to reseed in our climate over a longer period than this study. Nonetheless, some hollies seem to show promise as horticul- tural options preferable to L aquifolium. These include red-fruited deciduous shrubby species like I. decidua and I. verticillata, popular with birds in Seattle and in England (Ridley 1930). Ilex xX meserveae 8S. Y. Huis a little-known low-growing I. aquifolium hybrid (. aquifolium X rugosa F. Schmidt) but its fruits were outnumbered and ignored by birds in the study area. Landscapers might instead favor evergreen trees with the form as well as the color of English holly, like 1 x koehneana Loes. (I. aquifolium X_ latifolia Thunb.), 7 x beanii Rehder (1. aquifolium X dipyrena Wall.), and especially 1. X altaclerensis (Ul. aquifolium X perado Aiton). The latter accounted for 19% of all //ex frugivory observa- tions, and has the dense growth, bright fruit color, and dark shiny foliage most similar to /. aquifolium. From an ornamental, ornithological, and invasive standpoint, /. < altaclerensis may be the best available replacement, based on my initial results. However, any holly must be rigorously tested for invasiveness, and should be | commercially available, before promotion as an | alternative to 1. aquifolium in the Pacific North- | west. ACKNOWLEDGMENTS [Vol. 57 | I thank Linda Brooking for the seed drawings, and | Ben Legler for the map. Susyn Andrews, Dennis Paulson, Scott Pearson, Sarah Reichard, Rex Salla- | banks, and Robert Sundstrom were generous with their time in discussions about avians and J//ex. I am grateful to the staff of the institutions cited for providing access to living material, herbarium specimens or loans, especially David Giblin, Wendy DesCamp and Randall Hitchin. Ed Alverson, Elizabeth Gould, Art Jacobson, Frank Lang, and Fred and Ann Weinmann provided | invaluable aid with the field work. LITERATURE CITED ALBERTI, M., E. BOTSFORD, AND A. COHEN. 2001. | Quantifying the urban gradient: Linking urban planning and ecology. Pp. 89-115 in J. H. Marzluff, R. Bowman, and R. Donnelly (eds), Avian ecology — and conservation in an urbanizing world. Kluwer — Academic Publishers, Boston, MA. ALCANTARA, J. M., P. J. REY, F. VALERA, AND A. M. SANCHEZ-LAFUENTE. 2000. Factors shaping the seedfall pattern of a bird-dispersed plant. Ecology — 81:1937-1950. ANDREWS, S. 1997. //ex Linnaeus. Pp. 164-173 in J. Cullen, J. C. M. Alexander, C. D. Brickell, J. R. Edmondson, P. S. Green, V. H. Heywood, P.-M. Jorgensen, S. L. Jury, S. G. Knees, V. A. Mat-— thews, H. S. Maxwell, D. M. Miller, E. C. Nelson, N. K. B. Robson, S. M. Walters, and P. F. Yeo, (eds.), The European garden flora, Vol. V, dicot- | yledons (Part III). Cambridge University Press, Cambridge, UK. ARRIETA, S. AND F. SUAREZ. 2004. Germination and seed bank depletion of holly U/ex aquifolium L.) in — four microhabitat types. Seed Science Research 14:305—313. AND . 2005. Spatial dynamics of //ex aquifolium populations seed dispersal and_ seed bank: understanding the first steps of regeneration. Plant Ecology 177:237—248. BARNEA, A., Y. YOM-TOv, AND J. FRIEDMAN. 1991. Does ingestion by birds affect seed germination? | Functional Ecology 5:394—402. BECKETT, K. AND G. BECKETT. 1979. Planting native trees and shrubs. Jarrold & Sons, Ltd., Norwich, UK. BRAYSHAW, T. C. 1960. Key to the native trees of | Canada. Bulletin Forestry, Ottawa, ON. 125, Canada Department of | CATLING, P. M. 1997. The problem of invading alien | trees and shrubs: some observations in Ontario and a Canadian checklist. Canadian Field-Naturalist 111:338—342. CRESO, I. 1984. Vascular plants of western Washington. Johnson Cox Printers, Tacoma, WA. CROOKS, K. R. AND M. E. SOULE. 1999. Mesopredator release and avifaunal extinctions in a fragmented — system. Nature 400(6744):563—S66. 2010] CUENOUD, P., M. A. DEL PERO MARTINEZ, P.-A. LOIZEAU, R. SPICHIGER, S. ANDREWS, AND J.-F. MANEN. 2000. Molecular phylogeny and biogeog- raphy of the genus //ex L. (Aquifoliaceae). Annals of Botany 85:111—122. DLuGOosCcH, K. M. 2005. Understory community changes associated with English ivy invasions in Seattle’s urban parks. Northwest Science 79:53—60. DOUGLAS, G. W., G. B. STRALEY, AND D. MEIDINGER. 1989. The vascular plants of British Columbia, Part 1. BC Ministry of Forests, Crown Publications, Victoria, BC. , : , and J. POJAR (eds.). 1998. Illustrated flora of British Columbia, Vol. 1. BC Ministry of Environment, Lands and Parks, Ministry of Forests, Victoria, BC. FLEMING, T. H. 1988. The short-tailed fruit bat. A study in plant-animal interactions. University of Chicago Press, Chicago, IL. FRIEDMAN, J. M., G. T. AUBLE, P. B. SHAFROTH, M. L. ScoTT, M. F. MERIGLIANO, M. D. FREEHLING, AND E. R. GRIFFIN. 2005. Dominance of non- native riparian trees in western USA. Biological Invasions 7:747—751. GALLE, F. C. 1997. Hollies, the genus //ex. Timber Press, Portland, OR. GARCIA, D., J. R. OBESO, AND I. MARTINEZ. 2005. Rodent seed predation promotes differential re- cruitment among bird-dispersed trees in temperate secondary forests. Oecologia 144:435—446. GLEADOW, R. M. AND D. H. ASHTON. 1981. Invasion by Pittosporum undulatum of the forests of Central Victoria. I. Invasion patterns and plant morphol- ogy. Australian Journal of Botany 29:705—720. GRAY, A. N. 2005. Eight nonnative plants in western Oregon forests: Associations with environment and management. Environmental Monitoring and As- sessment 100:109—127. HASKELL, D. G., A. M. KNUPP, AND M. C. SCHNEI- DER. 2001. Nest predator abundance and urbani- zation. Pp. 243-258 in J. H. Marzluff, R. Bowman, and R. Donnelly (eds.), Avian ecology and conservation in an urbanizing world. Kluwer Academic Publishers, Boston, MA. HITCHCOCK, C. L. AND A. CRONQUIST. 1961. Vascular plants of the Pacific Northwest, Part 3. University of Washington Press, Seattle, WA. AND ———. 1973. Flora of the Pacific Northwest. University of Washington Press, Seat- tle, WA. HOLMGREN, P., N. H. HOLMGREN, AND L. C. BAR- NETT. 1990. Index herbariorum, Part 1: The herbaria of the world, 8th ed. New York Botanical Garden, Bronx, NY. HOLTHUIJZEN, A. M. A. AND T. L. SHARIK. 1985. The avian seed dispersal system of eastern red cedar (Juniperus virginiana). Canadian Journal of Botany 63:1508—-1515. Howe, H. F. 1979. Fear and frugivory. American Naturalist 114:925—931. HRUuSA, F., B. ERTTER, A. SANDERS, G. LEPPIG, AND E. DEAN. 2002. Catalogue of non-native vascular plants occurring spontaneously in California be- yond those addressed in The Jepson Manual — part 1. Madrono 49:61—-98. JACOBSON, A. L. 2006. Trees of Seattle, 2nd ed. Published by the author, Seattle, WA. ZIKA: INVASIVE HOLLIES 9 JONES, E. AND N. T. WHEELWRIGHT. 1987. Seasonal changes in the fruits of Viburnum opulus, a fleshy- fruited temperate zone shrub. Canadian Journal of Botany 65:2291—2296. JORDANO, P. AND E. SCHUppP. 2000. Seed disperser effectiveness: the quantity component and patterns of seed rain for Prunus mahaleb. Ecological Monographs 70:591—615. KOLLMANN, J. 2000. Dispersal of fleshy-fruited species: a matter of spatial scale? Perspectives in Plant Ecology, Evolution and Systematics 3:29—51. AND M. BUSCHOR. 2002. Edge effects on seed predation in deciduous forests of northern Swit- zerland. Plant Ecology 164:249—261. KwitT, C., D. J. LEVEY, AND C. H. GREENBERG. 2004. Contagious seed dispersal beneath heterospecific fruiting trees and its consequences. Oikos 107: 303-308. LOIZEAU, P.-A. AND R. SPICHIGER. 2004. Pp. 26—27 in N. Smith, S. A. Mori, A. Henderson, D. W. Stevenson, and S. V. Heald (eds.), Flowering plants of the Neotropics. The New York Botanical Garden and Princeton University Press, Princeton, NJ. Low, T. 2002. Feral future: the untold story of Australia’s exotic invaders, 2nd ed. The University of Chicago Press, Chicago, IL. MAcK, R. N., D. SIMBERLOFF, W. M. LONSDALE, H. EVANS, M. CLOUT, AND F. A. BAZZAz. 2000. Biotic invasions: causes, epidemiology, global consequences, and control. Ecological Applications 10:689—7 10. MARTIN, A. C., H. S. ZIM, AND A. L. NELSON. 1951. American wildlife & plants: a guide to wildlife food habits. McGraw-Hill Book Company, New York, NY. McCLINTOCK, E. 1993. Aquifoliaceae, Holly Family. Pp. 169 in J. C. Hickman (ed.), The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. MEYER, G. A. AND M. C. WITMER. 1998. Influence of seed processing by frugivorous birds on germina- tion success of three North American shrubs. American Midland Naturalist 140:129—139. Munz, P. A. 1968. Supplement to a California flora. The University of California Press, Berkeley, CA. AND D. D. KECK. 1965. A California flora. The University of California Press, Berkeley, CA. MURRAY, K. G., K. WINNETT-MURRAY, E. A. CRO- MIE, M. MINOR, AND E. MEYERS. 1993. The influence of seed packaging and fruit color on feeding preferences of American robins. Vegetatio 107/108:217—226. OBESO, J. R. 1998. Patterns of variation in J//ex aquifolium fruit traits related to fruit consumption by birds and seeds predation by rodents. Ecoscience 5:463-469. AND I. C. FERNANDEZ-CALVO. 2003. Fruit removal, pyrene dispersal, post-dispersal predation and seedling establishment of a bird-dispersed tree. Plant Ecology 165:223—233. OLMSTED, D. 2006. English holly, //ex aquifolium. Pp. 42-43 in P. D. Boersma, S. H. Reichard, and A. N. Van Buren (eds.), Invasive species in the Pacific Northwest. University of Washington Press, Seattle, WA. OMAR, T. 1994. The woody plant collection in the Washington Park Arboretum. Serial Publication 1. Center for Urban Horticulture, University of Washington, Seattle, WA. Peck, M. E. 1961. A manual of the higher plants of Oregon, 2nd ed. Binfords and Mort, Portland, OR. PIMENTEL, D., L. LACH, R. ZUNIGA, AND D. MorRRISON. 2000. Environmental and economic costs of nonindigenous species in the United States. BioScience 50:53—65. REICHARD, S. H. AND P. WHITE. 2001. Horticulture as a pathway of invasive plant introductions in the United States. BioScience 51:103—113. REINHART, K. O., F. T. MAESTRE, AND R. M. CALL- AWAY. 2006. Facilitation and inhibition of seed- lings of an invasive tree (Acer platanoides) by different tree species in a mountain ecosystem. Biological Invasions 8:231—240. RIDLEY, H. N. 1930. The dispersal of plants throughout the world. L. Reeve and Co., Ashford, UK. SALLABANKS, R. AND F. C. JAMES. 1999. Turdus migratorius, American robin. The Birds of North America 462:1—27. ScHupp, E. W. 1993. Quantity, quality and the effectiveness of seed dispersal by animals. Vegetatio 107/108:15—29. SMAL, C. M. AND J. S. FAIRLEY. 1982. Food of wood mice (Apodemus sylvaticus) and bank voles (Cle- thrionomys glareolus) in oak and yew woods at Killarny, Ireland. Journal of Zoology 191:413-418. SNow, B. AND D. SNOw. 1988. Birds and berries: A study of ecological interaction. T & AD Poyser, Carlton, U.K. SNow, D. W. AND B. K. SNOw. 1986. Some aspects of avian frugivory in a north temperate area relevant to tropical forest. Pp. 159-164 in A. Estrada and T. H. Fleming (eds.), Frugivores and seed dispers- MADRONO [Vol. 57 al. Dr. W. Junk Publishers, Netherlands. Dordrecht, The SZCZAWINSKI, A. F. AND A. S. HARRISON. 1972. Flora | of the Saanich Peninsula. Annotated list of | vascular plants. Occasional Paper 16, BC Provin- cial Museum, Victoria, BC. TAYLOR, R. L. AND B. MACBRYDE. 1977. Vascular plants of British Columbia: a descriptive resource inventory. Technical Bulletin 4, Botanical Garden, University of British Columbia, The University of | British Columbia Press, Vancouver, BC. TEMPLE, S. A. 1990. The nasty necessity: eradicating exotics. Conservation Biology 4:113-115. THILENIUS, J. F. 1968. The Quercus garryana forests of | the Willamette Valley, Oregon. Ecology 49: 1124-1133. TICKNOR, R. L. production of cut English holly (lex aquifolium) in the Pacific Northwest. Ornamentals Northwest | Newsletter 10:12—15. WAGNER, W. L., D. R. HERBST, AND S. H. SOHMER. 1999. Manual of the flowering plants of Hawa1’l, | revised ed., Vol. 1. Special Publication 97, Bishop Museum, University of Hawai’i Press, Honolulu, | HI. WHEELWRIGHT, N. T. records. The Auk 103:710—725. WILLIAMS, P. A. AND B. J. KARL. 1996. Fleshy fruits of indigenous and adventive plants in the diet of birds | in forest remnants, Nelson, New Zealand. New © Zealand Journal of Ecology 20:127—145. ZIKA, P. F. AND A. L. JACOBSON. 2005. Noteworthy | collections Madrono 52: 209-212. from Washington. 1986. Overview of commercial | 1986. The diet of American robins: an analysis of U.S. Biological Survey © MADRONO, Vol. 57, No. 1, pp. 11-19, 2010 DISTURBANCE, RESOURCES, AND EXOTIC PLANT INVASION: GAP SIZE EFFECTS IN A REDWOOD FOREST BRENT C. BLAIR!, DEBORAH K. LETOURNEAU, AND SARA G. BOTHWELL Department of Environmental Studies, University of California, Santa Cruz, CA 95064 blairb@xavier.edu GREY F. HAYES Elkhorn Slough National Estaurine Research Reserve, Watsonville, CA 95076 ABSTRACT Fluctuations in plant resource availability are hypothesized to promote exotic plant invasion by allowing propagules already present in an area a chance to successfully compete for unused resources. To examine the relationship between resource enrichment and exotic species invasion, we used selective logging canopy gaps over a range of sizes (56 m* to >1500 m°’) in a redwood forest (Santa Cruz County, CA) as a surrogate for disturbance intensity and level of pulsed resource enrichment. Measurements of abiotic conditions in gaps ca. 10 yr after logging suggest light is the primary difference in current resource availability, though a pulse of light and nutrients likely occurred at the time of gap formation. Exotic species richness and relative cover increased significantly as gap size increased. In a separate manipulative experiment, we compared understory plant composition between artificially shaded and unshaded plots in 2.5-year-old logging gaps. Shaded plots had a lower proportion of exotic species than did adjacent, unshaded plots, showing that light is a critical resource for exotic species in redwood forest habitats. Taken together, these results support the view that both physical disturbance and increased availability of scarce resources contribute to a community’s susceptibility to invasion and suggest a linear relationship between the size of logging gaps and the magnitude of exotic species invasion. Key Words: Canopy gap, disturbance, redwood forest, selective logging, Sequoia sempervirens, understory. The probability of invasion by non-native plant species is determined by the supply of introduced propagules, the capacity of these species to establish, and the susceptibility of the environment to invasion (Lonsdale 1999). Sus- ceptibility, or invasibility, of the environment is determined by bottom-up forces (such as light and nutrients), top-down forces (such as herbi- vores and pathogens) and lateral forces (facilita- tive and competitive interactions among plants) (Davis et al. 2000). Theories on invasibility often focus on the dynamics of bottom-up forces and Suggest that increases in resource availability (e.g., light, moisture, nutrients) promote invasi- bility of plant communities. For example, in- creased water supply in drought-prone areas often promotes invasion (Li and Wilson 1998; Davis et al. 1999; Dukes and Mooney 1999) as does the addition of limiting nutrients in North American grasslands (Stohlgren et al. 1999). Alternative theory suggests that physical distur- bance acts by disrupting existing species interac- tions, diminishing the competitive intensity for resources within plant communities, and thus allows foreign invaders to take a foothold (Rejmanek 1989; Hobbs and Huenneke 1992). "Present address: Department of Biology, Xavier University, Cincinnati, OH 45207. Disturbance may also increase unused resources in a community by disrupting resource uptake. Davis et al. (2000) suggest that it 1s the presence of unused resources rather than total amount of resources that is critical to invasive species SUCCESS. While studies confirm that both physical disturbance and changes in resource availability promote exotic species invasion (Li and Wilson 1998; Stohlgren et al. 1999; Rodgers and Parker 2003; Glasgow and Matlack 2007), the magnitude of change attributable to each of these factors is rarely studied. However, we know that plant competitive intensity declines as the magnitude of unused resources increases (Davis et al. 1998). Therefore, a large increase in resource availability should boost the success of exotic species invasions. Canopy gaps caused either by natural treefalls or logging events are one type of disturbance that increases local resource availability. Canopy gap formation causes an immediate resource pulse at the forest floor. The quantity and quality of light increase proportionate to the amount the over- story shade is diminished (Collins et al. 1985). Higher precipitation throughfall and lower tran- spiration may cause soil moisture to increase (Collins et al. 1985), but this trend may be mediated in areas with coastal fog (Dawson 1 MADRONO 1998). Soil disturbance caused by fallen or extracted trees can also create a mineral soil seedbed critical for plant germination (Battles et al. 2001). Decaying plant debris from fallen or extracted trees, in combination with reduced plant uptake, increase nutrient availability (Mat- son and Vitousek 1981; Vitousek 1985a, b; Frazer et al. 1990). Eventually these abiotic resources return to base levels, but return time varies among these resource classes and could be quite long in areas where soil mineralization rates are low. Surplus nutrients that are not absorbed by rapidly colonizing plants are lost, in the relatively short term, through erosion and leaching (Uhl et al. 1982; Vitousek 1985b). Similarly, excess soil moisture and newly formed mineral soil seedbeds will decline as plants and their roots re-colonize empty space above and belowground. However, light levels decline slowly and canopy closure may take years or decades to complete (Moore and Vankat 1986). Thus, canopy gaps cause a resource pulse whose components re-equilibrate at different rates. Canopy gaps cause measurable changes in herbaceous species composition in forest ecosys- tems (Davison and Forman 1982; Moore and Vankat 1986; Glasgow and Matlack 2007). Moore and Vankat (1986), for example, found that while total species richness remained un- changed in canopy gaps, species composition changed substantially with early spring annuals declining and late spring and summer species becoming more abundant. California’s coastal redwood forest communities tend to be composed of native species, with low light levels in the understory providing a potential barrier to colonization by the many exotic plants that thrive in disturbed sites in the region. Selective logging events are different than natural treefalls as they remove large merchantable tree boles while un- merchantable stumps, branches and leaf litter remain. Experiments using selective tree removal have found that changes in understory plant composition are similar to those in natural treefall gaps of similar size (Collins et al. 1985; Collins and Pickett 1988a). We conducted two complementary field studies using canopy gaps formed during. selective logging to examine the effects of physical disturbance (tree removal) and resource pulses on exotic species invasion in a coast redwood (Sequoia sempervirens (D.Don) Endl.) forest. In the first study we used forest canopy gaps of different sizes that were created in the 1990’s by selective logging operations to examine the effects of logging disturbance magnitude on invasibility in the understory plant community (referred to as the gap size study). Gap sizes in this study encompassed a range of over an order of magnitude in area (56 m°* to 1612 m’) and were of similar size to natural treefalls found in [Vol. 57 redwood forests (160 m?* to 1770 m’) (Sugihara 1992; Busing and Fugimori 2002) as well as other temperate forests (8 to 1320 m’) (Barden 1981; Romme and Martin 1982; Collins and Pickett 1988b). In the second study we tested for a direct effect of light as a pulsed resource after logging by using paired, artificially shaded (using shade | cloth) and unshaded plots in newly created | logging gaps in the same forest (referred to as the Jight effect experiment). In these studies, we tested the hypothesis that | exotic plant species survival and dominance are > positively influenced by physical disturbance (tree removal) and resource pulses (light and nutrients) ~ created by gap formation during selective log- ging. We predicted that exotic species richness and cover would increase concomitantly with the © size of canopy gap, in the gap size study. In the | light effect experiment, we expected that sections | of canopy gaps covered by shade cloth would experience a reduced influx of exotic species after | gap formation when compared to unshaded | regions of the same canopy opening. METHODS Study Site For both experiments, we used selective logging sites in a redwood forest located in the Santa Cruz Mountains at Swanton Pacific | Ranch, ca. 21 km north of Santa Cruz, CA (37°04'’N. 122°14'W), a 3200 acre property owned and managed by the California Poly- technic University, San Luis Obispo. The region receives approximately 700 mm of rainfall annually, mostly between November and May, and has a mean temperature of | 13C. The forest, which was clear-cut in the early 1900’s, is dominated by coast redwood (Sequoia sempervirens) and Douglas fir (Pseu- dotsuga menziesii (Mirb.) Franco). In 1991 and 1995, California Polytechnic designated forest sections to be selectively logged. They removed | individual and small stands of trees, leaving canopy gaps of various sizes scattered within the forest. For the gap size comparison, we estab- lished 16 understory plant census plots (8 < 8 m) in clearings under canopy gaps. All gaps were | located in an area of ca. 0.5 km?* (Fig. 1). Gaps were identified through a comprehensive search > of the logged area and all gaps over 100 m° were | used. The shade cloth experiment was established in two sections of the same forest, after a third | selective logging operation completed in Novem- ber 2004. These sections were not affected by the 1991 or 1995 logging events. For this second experiment, we established paired plots (shaded and unshaded) in 14 clearings under logging gaps. 2010] 250 Meters Fic. 1. The location of canopy gaps at Swanton Pacific Ranch (n = BLAIR ET AL.: GAP SIZE EFFECTS IN A REDWOOD FOREST 13 16) used in gap size comparison are represented by black circles showing relative sizes and proximities. Gap Size Effects Plot evaluation. For the gap size comparison, gaps were initially identified as clearings of >100 m? with no mature tree stems. Canopy gap openings were measured in each of eight compass directions as the distance from plot center to points directly below the edge of surrounding canopy foliage (Brokaw 1982). Hemispherical photographs were taken with a Nikon 6006 camera (Nikon, Melville, NY, USA) and a Peleng 8-mm fisheye lens (Peleng, Minsk, Belarus) using Kodak Elite-Chrome film (200 ASA, Eastman Kodak, Rochester, NY, USA). The camera was mounted on a tripod, pointed skyward, and positioned so that the top of the photograph corresponded to due north. After leveling the camera, two photographs were taken at each plot. The photograph at each site with the best contrast was used for analysis. Slides were digitized using a Polaroid Sprint 35 mm scanner (Boston, MA, USA). Images were then analyzed using the computer program Gap Light Analyzer 2.0 (Frazer et al. 1999) to determine percent transmitted global photosynthetically active ra- diation (PAR). Percent transmitted PAR repre- sents the amount of above-canopy direct and diffuse PAR incident beneath the canopy (Can- ham 1988). To determine temperature and humidity we placed data loggers and probes (Campbell, Logan, UT, USA) at the center of six plots that represented the range of gap sizes present in this study. We measured temperature and relative humidity in five plots and temperature in the sixth between July 18 and September 17, 2003. This period represents the warmest and driest weeks of the year (Mediterranean climate) when temperature and humidity differences between small and large gaps should be most pronounced. Data loggers calculated minimum and maximum daily temperature and relative humidity. We conducted linear regression analysis (SPSS 1999) to determine the effects of gap size on temperature and humidity. Soil properties. To determine if soil moisture (surface and root zone) varied with gap size in the wet and dry season we measured water content in each plot, using two 5 cm-diameter soil cores from a 0-5 cm and 5—20 cm depth in June (dry season) and November (wet season), 2003. Fresh 10 g sub-samples from each core (4 per plot) were oven dried at 90C to constant weight and subsequently weighed to calculate water content. The levels of essential plant nutrients (N, P, K, Mg and Ca) in the plots were determined from the remaining soil in the 0—S cm cores. These 14 MADRONO samples were air dried and passed through a 2 mm sieve. The two dried 0-5 cm soil samples from each plot were pooled and mixed thoroughly before analysis for available N (nitrate and ammonium), P, K, Mg and Ca. Soil N, Mg and Ca were determined using soil sub-samples extracted with a sodium acetate solution and P was extracted with Bray’s solution (Page et al. 1982). Soil chemical analyses were carried out at Perry Laboratories (Watsonville, CA, USA). Gap size effects on soil properties were examined using regression analysis (SPSS 1999). Vegetation purveys. We surveyed each of the sixteen 8 m X 8 m plots in April of 2003, during the peak flowering season for understory herbs (February—May), to determine the total number of plant species and the number of exotic plant species present. We estimated plant cover for each species in the center 6 m X 6 m area of each plot, to reduce potential edge effects. Within this area, we randomly chose nine of the possible thirty-six 1 m° subplots. For each of the 9 subplots, a 50 cm X 50 cm grid with twenty-five 10cm X 10cm cells was held over the vegetation at approximately | m height. The number of cells (0-25) that contained a particular species served as a measure of relative cover for that species. We quantified plant species richness and diversity by using the number of species per unit area in the whole plot censuses (S;), and a measure of evenness (J). The index J is defined as follows: Si = SS Pyne; Ja —_i=! In S; where P; is the relative frequency of occurrence of every species in each plot’s nine point-count subplots, and S, is the total number of species in each 64 m*° plot. Exotic (S.) and native (S,,) species richness were determined using the whole plot (8 m X 8 m) census data. To determine exotic species relative cover (C,) we used the quadrat cell count data to obtain percentages for each plot. The index C, is defined as follows, Se ee al Ce = Se Ss ej + Nj =] i= 1 i where e; 1s the number of occurrences of exotic species 7 in the nine 25 cell grids, n; is the number of occurrences of native species 7 in the nine 25 cell grids, S. and S,, are the total number of exotic and native species found in the plot. To evaluate treatment effects on species diversity and com- position we conducted regression analyses (SPSS [Vol. 57 1999). The effects of gap size and percent canopy cover were tested on the number of exotic (S.) and native (S,,) species, exotic relative cover (C,), and evenness (J). The index C. was square-root transformed prior to analysis to meet assump- tions of normality. Light Effects — Experimental Manipulations and Vegetation Survey In April of 2004, 12 clearings (>100 m7) under closed canopy were identified by their proximity to trees to be removed during the upcoming logging operation. We surveyed each of the clearings to determine the total number of native and exotic plant species present. After tree extraction in October of 2004, two additional clearings were added and one of the previous sites was discarded due to lack of increased light penetration at the forest floor after logging. The remaining 11 previously surveyed and 2 new clearings corresponded to areas of increased light penetration (canopy openings). Within each of the 13 clearings, we demarcated a pair of 5m X 5 m plots, and randomly assigned the shade treatment to one plot in each pair. The remaining plot was used as an open (unshaded) control. In January 2005, shade cloth (80% shade) was suspended from a PVC frame 1.5 m in height, with a 2 m T-bar in the center to elevate the center of the shade canopy and reduce litter accumulation on the structure. Shade cloth was draped over the edges of the structure but left 1 m above the ground uncovered to allow access by arthropods and ensure airflow. Each pair of plots (shaded and open) was established within 10 m of each other within a single clearing. When on a slope, paired plots were positioned to have the same slope aspect. In July 2007, all pairs of shaded and open plots were censused for the total number of native and exotic plant species. The relative number of plant species in shade and open plots was compared using a paired t-test for equal variances on the normally distributed differences between light and dark plots for total plant species, native plant species, and exotic plant species per plot (SPSS 1999). RESULTS Gap Size Effects Abiotic factors. Percent transmitted global PAR increased significantly with gap size (Ta- ble 1). Average maximum daily temperatures also increased with gap size (Table 1). However, minimum daily temperatures were similar across all gap sizes. In smaller gaps, we measured higher minimum humidity levels than in larger gaps, but maximum humidity levels were similar (Table 1). 2010] BLAIR ET AL.: GAP SIZE EFFECTS IN A REDWOOD FOREST 15 TABLE 1. EFFECT OF CANOPY GAP SIZE ON GAP PROPERTIES. Significant P values are indicated in bold. Property Coefficient R° F (N) P-value Transmitted global PAR (%) 0.010 0.61 14.91 (16) 0.002 Native species richness 0.004 0.24 4.43 (16) 0.054 Exotic species richness 0.002 0.71 33.68 (16) <0.001 Exotic relative cover 0.0002 0.41 2.9 f (116) 0.007 Temperature (C°) Minimum —(0.002 0.32 1.89 (6) 0.241 Maximum 0.006 0.83 19.40 (6) 0.012 Humidity (%) Minimum —0.040 0.82 13.935) 0.034 Maximum —0.003 0.54 3.47 (5) 0.159 Soil moisture was significantly greater during the wet season than the dry season (t = 8.11, df = 15, P < 0.001). However, sampling within each period showed no significant differences in moisture levels between gap sizes at the 0—S cm or 5—20 cm depths. No gap size-dependent trends were found for nutrient availability. Vegetation. The understory vegetation in selectively logged redwood forest gaps was typical of that found in natural forest. Common native species such as Oxalis oregana Nutt. (redwood sorrel), Polystichum munitum (Kaulf.) Presl. (western sword fern), Rubus ursinus Cham. & Schlecht (California blackberry), Stachys bullata Benth. (California hedge nettle), and Trillium ovatum Pursh. (western wake-robin) were present. The five largest canopy gaps (832 m*-1612 m’) contained between 18 to 27 understory plant species, medium gaps (212 m*— 624 m’) had 15 to 26 species, and the smallest gaps (<150 m/’) had between 11 and 21 species. The total number of native understory plant species in canopy gaps showed no significant relationship with gap size (Table 1) or light availability, though native species richness tended to increase with gap size (coefficient = 0.004, r° = 0.24, P = 0.06). In contrast, the number and percent cover of exotic plant species increased significantly with gap size and light availability (Table 1, Fig. 2). The exotic plant species found in understory plots were: Cirsium vulgare (Savi) Ten. (bull thistle), Cortaderia jubata (Lem.) Stapf (jubata grass), Erechtites minima (Poir.) DC. (Australian fire- weed), Myosotis latifolia Poir. (forget-me-not), Rubus discolor Weihe & Ness (Himalayan black- berry), Vulpia sp. (annual fescue), and Torilis sp. (hedge-parsley). The relationship between gap size and estimated light availability was strong (1 = 0.61) and both were good predictors for exotic species number (r* = 0.71 (gap size) vs. r> = 0.82 (light availability)). However, light availability was a far better predictor of exotic species relative cover (tr? = 0.61) than gap size (r* = 0.41) (Table 1, Fig. 2). The exotic and native species evenness was unrelated to either gap size (coefficient = 0.0001, r7 = 0.15, P = 0.134) or to light availability (coefficient = 0.005, r> = 0.14, P = 0.147). 5 y= -0.47 + 0.15x r?=0.82P <0.001 Exotic Spcies Richness (S,) y = -18.09 + 1.75x r*=0.61P<0.001 Exotic Species Cover (%) 5 10 15 20 25 30 35 Global Transmitted PAR (%) FIG. 2. Relationship between exotic species richness (A), and exotic species relative cover (B) with global transmitted PAR (%). Dotted lines represent 95% confidence intervals. 16 MADRONO Number of Species Open FIG. 3. and open plots (n = 14 pairs). Light Effects on Vegetation The number of plant species occurring in pre- harvest understory in April, 2004 was half the plant species richness occurring within the forest gaps comparison (above) surveyed the previous year (average of 10.3 species in closed canopy before logging versus 20 species, on average, in all gaps). Plant species richness did not increase strongly after 2.5 yr, with a mean richness of 11.9 (+£0.5 SE) per 5m X 5 m light plot (n = 14). However, in the pre-harvest samples, all the species were natives, compared to a mean of 2.9 (+0.3 SE) exotic species per 5m X 5 m in post- logging open plots. Neither did the addition of shade cloth after logging have a significant effect on plant understory species richness or the number of native species present, on average, in the plot. However, there was twice the number of exotic species present in open plots than in shaded plots (t = —4.2, df = 13, P = 0.001) (Fig. 3). DISCUSSION The increasing richness and cover of exotic plant species across selectively logged forest gaps of increasing size supports our prediction that the magnitude of disturbance positively affects exotic species invasion, directly and/or through a resulting pulse in available plant resources. The removal of individuals or clusters of timber trees resulted in direct disturbance of existing vegeta- tion. After disturbance, the amount of light reaching the understory immediately increased and soil likely had a short-term nutrient enrich- ment as nutrients were released from decaying vegetation and fewer plant roots were present for [Vol. 57 Mmm Native Exotic Shaded Mean native and exotic species richness (+SE) in gaps 32 mo after selective logging in artificially shaded nutrient uptake (Matson and Vitousek 1981; Vitousek 1985a; Frazer et al. 1990; Frazer et al. 1999). The coupling of decreased resource use by native species with increased total resource availability during the initial disturbance period would have made more resources available to exotic species, which were otherwise suppressed by understory conditions. Over a decade after logging events took place, light availability still varies significantly and directly with the size of the gap created. Invasion of these logging gaps by exotic species and the increase in exotic species richness in gaps of increasing size was likely due to this increase in light availability, and possibly other plant resources whose initial increases are no longer detectable. The persistence of these exotic species is enabled by the length of time required for canopy gaps to close, returning PAR to pre-disturbance levels. Our shade-cloth study suggests that distur- bance and light play complementary roles. There was a doubling of the exotic species richness in unshaded plots after logging compared to the number of species found in shaded plots within the same logging gaps. Further, we found an increase in exotic species even within the exper- imentally shaded plots compared to the unde- tectable level of exotic species in our pre-harvest vegetation surveys. Previous studies have found that physical disturbance has the greatest impact on a site’s invasibility if it is coupled with increased resource availability (Burke and Grime 1996; White et al. 1997; Leishman and Thomson 2005). Burke and Grime (1996), for example, showed in a manipulative field experiment that, while both physical disturbance and fertilization increased the invasibility of limestone grassland, exotic species were most successful in displacing | 2010] their native counterparts when both disturbance and fertilization were present. However, the conditions under which the exotic species in this study typically occur, and the habits of invasive exotic plants more gener- ally, indicate that light availability may be relatively more important than the influence of disturbance itself, through the physical disruption of established vegetation. Exotic plants, whether intentionally introduced for agricultural or orna- mental use (diCastri 1989) or unintentionally introduced (1.e., agricultural weeds) (Heywood 1989), tend to originate from high-light environ- ments. It is thus not surprising that exotics tend to be light-demanding species (Fine 2002) that are shade intolerant (Mack 1996). These characteris- tics suggest that many exotic species may be successful invaders only after disturbances that increase light availability. Supporting this idea, research in western Oregon found greater num- bers of exotic species in the understory of old- growth Douglas-fir forests than in un-thinned second growth forests with lower light availability at the forest floor (Bailey et al. 1998). Exotic species such as Cirsium vulgare, Erechtites mini- ma, and Rubus discolor, found in our study, are shade intolerant and, when found in undisturbed forests, are gradually out-competed by shade tolerant understory plants (Amor 1974; Mulda- vin et al. 1981; McDonald and Tappeiner 1986). The importance of treefall gaps in forest ecology is well known. However the impacts of tree harvest on plant community structure and diversity are not clearly understood in Mediter- ranean forests. In late successional forests, selective logging is often one of the preferred forest management methods because it more closely emulates natural disturbance patterns in uneven-aged forests and maintains mature forest structure (Webster and Lorimer 2002). Studies find selective logging to be superior to other more disruptive management systems (e.g., clearcut and shelterwood logging) in minimizing exotic species colonization (Battles et al. 2001). Unfor- tunately, several studies suggest that selective logging is disruptive in subtle and indirect ways. For instance, regeneration of certain species 1s greater with natural gaps rather than logging gaps (Nagaike et al. 1999). Other critics cite the obvious problems and damage that occurs through the tree extraction process (Vasiliauskas 2001) and creation of logging roads and trails (Kreutzweiser and Capell 2001). Old logging roads and trails at our site may serve as pathways for propagules of exotic species into and through the redwood forest habitat (Costa and Magnus- son 2002), which can then establish when even low-impact logging techniques are applied. Perhaps a more pertinent framework is to consider how the forest as a whole responds to artificial gaps in the long term. Even a decade BLAIR ET AL.: GAP SIZE EFFECTS IN A REDWOOD FOREST 17 after gap formation, we found obvious vegetative differences among gaps of different sizes. Log- ging-induced changes in understory species com- position are sometimes long-lived (Duffy and Meier 1992; Meier et al. 1995) and forests often revert slowly to their original structure over the course of decades or longer (Alaback and Her- man 1988; Halpern and Spies 1995; He and Barclay 2000). In an unlogged forest, succession between periods of natural disturbances (wind, fire) would bring the forest back to its original vegetative composition as shade-tolerant species gradually outcompete the light demanding ones that came in after disturbance. However, because exotic propagules are likely ubiquitous in remain- ing redwood forests, logging gaps may well be more likely to experience long-term shifts to- wards exotic composition than natural, pre- invasion treefall gaps. Exotic plant species are a ubiquitous compo- nent of terrestrial ecosystems today, and one that often negatively influences natural habitats (Vi- tousek et al. 1997). A range of impacts has been documented to occur in terrestrial systems (see Levine et al. 2003 for review). The impact of exotic species in the redwood forest understory 1s unknown, but their spread in logging gaps may change hydrology, mycorrhizal composition, and interrupt regeneration of disturbance-dependent native species, possibly leading to their extinction. Although exotic species will likely decline locally as gap closure occurs, at a larger spatial scale exotic species are a permanent component of this forest ecosystem. Exotic plants can be expected to take advantage of logging-induced canopy gaps within the forest, highlighting the importance of research on how exotic species impact forest community function. The next step, should their effects be detrimental, would be to explore ways to reduce exotic species spread. Washing logging equipment to limit the spread of exotic seeds, for example, may be a cost effective method to reduce propagule pressure during logging opera- tions (Brooks 2007). While it is inevitable that some form of logging occurs in most publicly and privately held redwood forests, improved meth- odology could reduce the impact of forest management both by minimizing disturbance and diminishing propagule pressure on forested lands. ACKNOWLEDGMENTS We wish to thank W. Mark, B. Dietterick, and S. Auten for logistical support and permission to use the Swanton Pacific Ranch forest site, California State Parks botanist T. Hyland for initial identification of understory plants and help with determining their status as native or exotic species, and J. G. Armstrong, D. Howes, N. Jacuzzi, E.P. Kress, C. MacDonald, A.W. Malisos, T. R. F. Roubison, and R. Welch for field assistance. The manuscript was improved by comments from, J. Hagen, D. Plante, A. Racelis and E. Zavaleta. 18 MADRONO LITERATURE CITED ALABACK, P. B. AND F. R. HERMAN. 1988. Long-term response of understory vegetation to stand density in Picea and Tsuga forests. Canadian Journal of Forest Research 18:1522—1530. AMOR, R. L. 1974. Ecology and control of blackberry (Rubus fruticosus L. agg.): I. reproduction. Weed Research 14:231—238. BAILEY, J. D., C. MAYRSOHN, P. S. DOESCHER, E. ST PIERRE, AND J. C. TAPPEINER. 1998. Understory vegetation in old and young Douglas-fir forests of western Oregon. Forest Ecology and Management 112:289-302. BARDEN, L. S. 1981. Forest development in canopy gaps of a diverse hard wood forest of the Southern Appalachian Mountains USA. Oikos 37:205—209. BATTLES, J. J., A. J. SHLISKY, R. H. BARRETT, R. C. HEALD, AND B. H. ALLEN-DIAZ. 2001. The effects of forest management on plant species diversity in a Sierran conifer forest. Forest Ecology and Man- agement 146:211—222. BROKAW, N. V. L. 1982. The definition of treefall gap and its effect on measures of forest dynamics. Biotropica 14:158—160. Brooks, M. L. 2007. Effects of land management practices on plant invasions in wildland areas. Pp. 147-162 in W. Nentwig (ed.), Biological invasions: ecological studies 193. Springer, Heidel- berg, Germany. BURKE, M. J. W. AND J. P. GRIME. 1996. An experimental study of plant community invasibil- ity. Ecology 77:776—790. BUSING, R. T. AND T. FUGIMORI. 2002. Dynamics of composition and structure in an old Sequoia sempervirens forest. Journal of Vegetation Science 13:785—-792. CANHAM, C. D. 1988. An index for understory light levels in and around canopy gaps. Ecology 69:1634-1638. COLLINS, B. S., K. P. DUNNE, AND S. T. A. PICKETT. 1985. Responses of forest herbs to canopy gaps. Pp. 217-234 in S. T. A. Pickett and P. S. White (eds.), The ecology of natural disturbance and patch dynamics. Academic Press, Orlando, FL. AND S. T. A. PICKETT. 1988a. Demographic responses of herb layer species to experimental canopy gaps in a northern hardwoods forest. Journal of Ecology 76:437-450. AND . 1988b. Response of herb layer cover to experimental canopy gaps. American Midland Naturalist 119:282—290. CosTA, F. AND W. MAGNUSSON. 2002. Selective logging effects on abundance, diversity, and composition of tropical understory herbs. Ecolog- ical Applications 12:807-819. DAvis, M. A., J. P. GRIME, AND K. THOMPSON. 2000. Fluctuating resources in plant communities: a general theory of invasibility. Journal of Ecology 88:528—534. K. J. WRAGE, AND P. B. REICH. 1998. Competition between tree seedlings and herbaceous vegetation: support for a theory of resource supply and demand. Journal of Ecology 86:652—661. ——., : , M. G. TJOELKER, T. SCHAEF- FER, AND C. MUERMANN. 1999. Survival, growth, and photosynthesis of tree seedlings competing [Vol. 57 with herbaceous vegetation along a _ water-light- nitrogen gradient. Plant Ecology 145:341—350. DAVISON, S. E. AND R. T. T. FORMAN. 1982. Herb and shrub dynamics in a mature oak forest, a 30 yr study. Bulletin of the Torrey Botanical Club 109:64—73. Dawson, T. E. 1998. Fog in the California redwood forest: ecosystem inputs and use by plants. Oecologia 117:476-485. DICASTRI, F. 1989. History of biological invasions with special emphasis on the old world. Pp. 1—30 in J. A. Drake, H. A. Mooney, F. diCastri, R. H. Groves, F. J. Krunger, M. Rejmanek, and M. Williams (eds.), Biological invasions: a global perspective. John Wiley and Sons, Chichester, U.K. DuFFy, D.C. AND A. J. MEIER. 1992. Do appalachian herbaceous understories ever recover from clear- cutting? Conservation Biology 6:196—201. DUKES, J. S. AND H. A. MOONEY. 1999. Does global change increase the success of biological invaders? Trends in Ecology & Evolution 14:135—139. FINE, P. V. A. 2002. The invasibility of tropical forests by exotic plants. Journal of Tropical Ecology 18:687—705. FRAZER, D. W., J. G. MCCOLL, AND R. F. POWERS. 1990. Soil nitrogen mineralization in a clearcutting chronosequence in a Northern California conifer forest. Soil Science Society of America Journal 54:1145. FRAZER, G. W., C. D. CANHAM, AND K. P. LERTZ- MAN. 1999. Gap light analyzer (GLA): imaging software to extract canopy structure and gap light transmission indicies from true-colour fisheye photographs, users manual and program docu- mentation. Simon Fraser University, Burnaby, B. C., and the Institute of Ecosystem Studies, Mill- brook, NY. GLAsGow, L. S. AND G. R. MATLACK. 2007. The effects of prescribed burning and canopy openness on establishment of two non-native plant species in a deciduous forest, southeast Ohio, USA. Forest Ecology and Management 238:319—329. HALPERN, C. B. AND T. A. SPIES. 1995. Plant species diversity in natural and managed forests of the Pacific Northwest. Ecological Applications 5:913—934. , F. AND H. J. BARCLAY. 2000. Long-term response of understory plant species to thinning and fertilization in a Douglas-fir plantation on southern Vancouver Island, British Columbia. Canadian Journal of Forest Research 30:566—572. HEYwoop, V. H. 1989. Patterns, extents and modes of invasions by terrestrial plants. Pp. 31-60 in J. A. Drake, H. A. Mooney, F. diCastri, R. H. Groves, F. J. Krunger, M. Rejmanek, and M. Williams (eds.), Biological invasions: a global perspective. John Wiley and Sons, Chichester, U.K. Hoss, R. J. AND L. F. HUENNEKE. 1992. Disturbance diversity and invasion implications for conserva- tion. Conservation Biology 6:324—337. KREUTZWEISER, D. P. AND S. S. CAPELL. 2001. Fine sediment deposition in streams after selective forest harvesting without riparian buffers. Canadian Journal of Forest Research 31:2134—-2142. LEISHMAN, M. R. AND V. P. THOMSON. 2005. Experimental evidence for the effects of additional water, nutrients and physical disturbance on invasive plants in low fertility Hawkesbury Sand- 2010] stone soils, Sydney, Australia. Journal of Ecology 93:38—49. LEVINE, J. M., M. VILA, C. M. D’ANTONIO, J. S. DUKES, K. GRIGULIS, AND S. LAVOREL. 2003. Mechanisms underlying the impacts of exotic plant invasions. Proceedings of the Royal Society of London - Series B: Biological Sciences 270: 775-781. Li, X. AND S. D. WILSON. 1998. Facilitation among woody plants establishing in an old field. Ecology 79:2694-2705. LONSDALE, W. M. 1999. Global patterns of plant invasions and the concept of invasibility. Ecology 80:1522—-1536. Mack, R. N. 1996. Predicting the identity and fate of plant invaders: emergent and emerging approaches. Biological Conservation 78:107—121. Matson, P. A. AND P. M. VITOUSEK. 1981. Nitrifica- tion potentials following clearcutting in the Hoo- sier National Forest, Indiana. Forest Science 27:781-791. McDONALD, P. M. AND J. C. TAPPEINER. 1986. Weeds: the cycles suggest controls. Journal of Forestry 84:33—37. MEIER, A. J.,S. P. BRATTON, AND D. C. DUFFY. 1995. Possible ecological mechanisms for loss of vernal- herb diversity in logged eastern deciduous forests. Ecological Applications 5:935—946. Moore, M.R. AND J. L. VANKAT. 1986. Responses of the herb layer to the gap dynamics of a mature beech-maple forest. American Midland Naturalist 115:336—-347. MULDAVIN, E. H., J. M. LENIHAN, W. S. LENNOX, AND S. D. VEIRS. 1981. Vegetation succession in the first ten years following logging of coast redwood forests. Technical Report 6, Redwood National Park, Arcata, CA. NAGAIKE, T., T. KAMITANI, AND T. NAKASHIZUKA. 1999. The effect of shelterwood logging on the diversity of plant species in a beech (Fagus crenata) forest in Japan. Forest Ecology and Management 118:161-171. PAGE, A., R. MILLER, AND D. KEENEY (eds.). 1982. Methods of soil analysis, Part 2: chemical and microbial properties. American Society of Agron- omy, Madison, WI. REJMANEK, M. 1989. Invasibility of plant communities. Pp. 369-388 in J. A. Drake, H. A. Mooney, F. diCastri, R. H. Groves, F. J. Krunger, M. Rejmanek, and M. Williamson (eds.), Biological invasions: a global perspective. Wiley, Chichester, U.K. RODGERS, J.C. AND K. C. PARKER. 2003. Distribution of alien plant species in relation to human BLAIR ET AL.: GAP SIZE EFFECTS IN A REDWOOD FOREST 19 disturbance on the Georgia Sea Islands. Diversity and Distributions 9:385—398. ROMME, W. H. AND W. H. MARTIN. 1982. Natural disturbance by tree-falls in old-growth mixed mesophytic forest: Lilley Cornett Woods, Ken- tucky. Pp. 367—383 in R. N. Muller (ed.), Proceed- ings of the 4th Central Hardwood Forest Confer- ence. University of Kentucky, Lexington, KY. SPSS Inc. 1999. SPSS for Windows, rel. 10.0.0. Chicago, IL. STOHLGREN, T. J.. D. BINKLEY, G. W. CHONG, M. A. KALKHAN, L. D. SCHELL, K. A. BULL, Y. OTSUKI, G. NEWMAN, M. BASHKIN, AND Y. SON. 1999. Exotic plant species invade hot spots of native plant diversity. Ecological Monographs 69:25—46. SUGIHARA, N. G. 1992. The role of treefall gaps and fallen trees in the dynamics of old growth coast redwood (Sequoia sempervirens (D.Don) Endl.) forests. Ph.D. dissertation. University of Califor- nia, Berkeley, CA. UHL, C., C. JORDAN, K. CLARK, H. CLARK, AND R. HERRERA. 1982. Ecosystem recovery in Amazon Caatinga forest after cutting, cutting and burning and bulldozer clearing treatments. Oikos 38: 313-320. VASILIAUSKAS, R. 2001. Damage to trees due to forestry operations and its pathological significance in temperate forests: a literature review. Forestry 74:319—-336. VITOUSEK, P. M. 1985a. Community turnover and ecosystem nutrient dynamics. Pp. 325-334 in S. T. A. Pickett and P. S. White (eds.), The ecology of natural disturbance and patch dynamics. Academic Press, Orlando, FL. . 1985b. Community turnover and ecosystem nutrient dynamics. Pp. 325-334 in S. T. A. Pickett and P. S. White (eds.), The ecology of natural disturbance and patch dynamics. Academic Press, Inc., Orlando, FL. , C. M. D’ANTONIO, L. L. LOOPE, M. REJMA- NEK, AND R. WESTBROOKS. 1997. Introduced species: a significant component of human-caused global change. New Zealand Journal of Ecology 21:1-16. WEBSTER, C. R. AND C. G. LORIMER. 2002. Single-tree versus group selection in hemlock-hardwood for- ests: are smaller openings less productive? Canadi- an Journal of Forest Research 32:591—604. WHITE, T. A., B. D. CAMPBELL, AND P. D. KEMP. 1997. Invasion of temperate grassland by a subtropical annual grass across an experimental matrix of water stress and disturbance. Journal of Vegetation Science 8:847—854. MADRONO, Vol. 57, No. 1, pp. 20-41, 2010 MORPHOLOGICALLY CRYPTIC SPECIES WITHIN DOWNINGIA YINA (CAMPANULACEAB) LisA M. SCHULTHEIS Foothill College, 12345 El Monte Road, Los Altos Hills, CA 94022 schultheislisa@foothill.edu ABSTRACT The Downingia yina species complex (Campanulaceae), centered in northern California and southern Oregon, currently contains three morphologically distinguished species: D. yina, D. elegans, and D. bacigalupii. This complex of species is notable for high levels of morphological and cytological variation, with chromosome counts of nm = 6, 8, 10, and 12. Molecular evidence suggests three main clades within this complex, corresponding more with cytological variation than with morphological variation. Additionally, the molecular evidence suggests a phylogeographic pattern associated with the Cascade Ranges, where members of the clade characterized by chromosome counts of n = 6, 8, and 10 are distributed primarily to the west of the Cascades while members of the clade characterized by chromosome counts of 7 = 12 are distributed primarily to the east. A third clade characterized by n = 10 is localized in the Lake of the Woods region of southern Oregon. Evidence from morphological, cytological, interfertility, and molecular data was used to re-examine the delimitation of species within this complex. Downingia elegans and D. bacigalupii are maintained, while D. yina is split into three morphologically cryptic species (D. yina, D. willamettensis, D. pulcherrima) that do not form a clade. Key Words: Campanulaceae, chromosome races, cryptic species, Downingia, phylogeography. The Downingia yina species complex is a monophyletic group (Schultheis 2001) comprising D. yina Applegate, D. bacigalupii Weiler, and D. elegans (Lindl.) Torr. The species complex represents a cytologically and morphologically variable group centered in northern California and southern Oregon. Chromosome numbers within the complex include n = 12 in D. bacigalupii, n = 10 in D. elegans, and races of n = 6, 8, 10 and 12 in D. yina (Weiler 1962; Foster 1972; Lammers 1993). Morphologically, both D. bacigalupii and D. elegans are distinguished from D. yina by an exserted staminal column with a sharp bend between the anthers and filament, and by the concave oval-shaped lower corolla lip with relatively parallel corolla lobes. Downingia baci- galupii can be distinguished from D. elegans by the corolla’s lighter shade of purple and by the yellow pigmentation in the corolla throat, a feature also found in D. yina. Morphological variation within D. yina has led some workers to recognize additional species or infraspecific taxa. Downingia yina was described by Applegate (1929) from a localized region of the southern Cascade Ranges in Klamath Co., Oregon. Shortly thereafter, Peck (1934, 1937) described two additional larger flowered species: D. willamettensis Peck from the Willamette Valley of Oregon, and D. pulcherrima Peck from eastern Oregon. In the first monograph of the genus, McVaugh (1941) recognized D. yina and D. willamettensis, including D. pulcherrima in the latter. McVaugh noted (1941), however, that D. yina and D. willamettensis were not readily distinguishable, and ultimately treated them as varieties within D. yina, var. yina and var. major McVaugh, respectively (McVaugh 1943). He distinguished the two varieties based on fruit characteristics (fusiform with hyaline lines in var. yina; subulate without hyaline lines in var. major), plant stature (larger and more erect in var. major), and geographic location of the popula- tions. Weiler (1962) found that the differences described between D. yina and D. willamettensis were not maintained under greenhouse condi- tions. He accordingly recognized only D. yina with no infraspecific taxa, although noting that fresh material of D. pulcherrima was not exam- ined. Weiler (1962) also noted that individuals of D. yina sensu lato tended to be decumbent to the west of the Cascade Ranges, and erect to the east. Foster (1972) was unable to find consistent morphological differences to correspond with cytological races in D. yina, but did note an ecological trend. She observed that D. yina chromosome races n = 6, 8, and 10 are found in habitats characterized by Kuchler (1964) as Oregon-oak woodland or cedar-hemlock-Doug- las fir mosaic while the D. yina chromosome race n = 12 is found in California mixed evergreen forest and juniper-steppe woodland, as charac- terized by Kichler (1964). Both Foster (1972) and Ayers (1993) followed Weiler (1962) in recognizing only D. yina. The present study emerged largely from a systematic investigation of the genus Downingia, in which molecular data unexpectedly suggested the existence of morphologically cryptic lineages within D. yina (Schultheis 2001), corresponding in part to infraspecific taxa previously recog- SCHULTHEIS: CRYPTIC SPECIES IN DOWNINGIA YINA 2 2010] ) o Cascade Range aan ae e \. eo Sa 2 aD. elegans = S ‘ (Clade |) : —" \ Fa ‘ SS Clade |<—'—- Clade Ill |_ | es : > > : ro ' vs — 8 | ° 42 i“ lade Ill extending 7 west of Cascades 119 59° W Fic. 1. Map of the northwestern USA showing localities of samples included in this study. Each symbol may represent one or multiple samples from the vicinity. The dashed line roughly corresponds to the geographic barrier created by the Cascade Range. Triangles = Downingia elegans. Squares = D. bacigalupii. Circles = samples previously included in D. yina. Filled circles = now assigned to D. willamettensis. Open circles = now assigned to D. pulcherrima. Circles with line through center = now assigned to D. yina sensu strictu. Clade I, Clade I] and Clade III refer to clades identified in phylogenetic analyses. nized. The situation was further complicated (Schultheis 2001) by the apparent para- or polyphyly of D. yina with respect to D. elegans and D. bacigalupii. The D. yina species complex thus represents a mixture of morphologically cryptic and morphologically distinctive lineages that may not correspond to the species currently recognized (Ayers 1993). The aim of this study was to further investigate the relationships and circumscriptions of D. bacigalupii, D. elegans and D. yina using morphological data, additional nuclear and chloroplast molecular sequence data to supplement Schultheis (2001), and available cytological and interfertility data (Weiler 1962; Foster 1972). METHODS Taxon Sampling Collections were made from throughout the range of D. elegans, D. bacigalupii, and D. yina (Appendix 1; Fig. 1). Herbarium collections provided important supplemental material. Downingia bicornuta A.Gray, D. concolor E. Greene, D. cuspidata (E. Greene) Rattan, D. N N TABLE |. MADRONO [Vol. 57 CHARACTERS USED IN MORPHOLOGICAL ANALYSES OF THE DOWINGIA YINA COMPLEX. Characters 1— 10 are quantitative and measured in millimeters, characters 11—14 are qualitative, and characters 15—19 are ratios. ‘Characters used in cladistic analyses, with character states noted in brackets. Character number Character l. sepal dorsal sepal, length 2: back slit dorsal surface 3. side slit lateral surface 4 upper lobe 5 lower lobe 6. filament! 7. anther 8 anther angle! 9 lower angle 0 horns' Character definition and how assessed corolla base to dorsal slit, length; equivalent to height of corolla tube along corolla base to lateral slit, length; equivalent to height of corolla tube along upper corolla lobes, length lower corolla lip, length filament tube, length (<6 mm [0], >6 mm [1]) anther tube, length angle between anther and filament tubes (<50 [0], >70 [1]) angle of divergence between lobes of the lower corolla lip anther horns, length (<0.62 mm [0], >0.62 mm [1]); refers to triangular projections on each of the two smaller anthers it. anther back 12. upper lobe orientation 13. yellow! 14. lower lobe shape 15. filament/anther 16. sideslit/backslit! ee filament/backslit! 18. upper/lower lobe 19: backslit/upper lobe montana (E. Greene) Rattan, and D. orndatissima E. Greene were chosen as outgroup taxa based on previous phylogenetic analyses within the genus (Schultheis 2001). Molecular Generation of sequence data. Extraction of total DNA from 24 samples first reported in Schultheis (2001) and 9 new samples (Appendix 1) involved use of either the CTAB protocol of Doyle and Doyle (1987) or Hillis et al. (1996) with minor modifications (Schultheis 2001), or use of Qiagen DNeasy Plant mini kits following manufacturer’s instructions. Most plant tissue samples were stored in a cooler while in the field and transferred to a —80C freezer within one week of collection. Voucher specimens were either the same plant from which tissue for DNA extraction was taken, or were other plants from the same site. Sequence data were generated from the nuclear 18S—26S rDNA internal transcribed spacer (ITS) and the chloroplast 3’¢rnK intron. Amplification and sequencing methods changed during the course of the project, as new techniques became available. Single-stranded DNAs of ITS 1 and ITS 2 were generated, purified, and manually sequenced following Baldwin (1992). Double-stranded DNAs of ITS 1, ITS 2 and the 3’¢rnK intron were generated, purified, and sequenced using automat- ed sequencing technology following Schultheis (2001). Sequences are deposited in Genbank. trichomes on anther dorsal surface: abundant (0), few (1), none (2) upper corolla lobes, orientation: parallel (0), intermediate (1), divergent (2) yellow on lower corolla lobe: present (0), absent (1) lower corolla lobe, shape: acute (0), intermediate (1), mucronate (2) filament length/Anther length length of lateral slit/Length of dorsal slit (=0.6 [0], <0.6 [1]) filament length/Length of dorsal slit (<1 [0], 1—2 [1], >2 [2]) upper lobes length/Lower lip length length of dorsal slit/Upper lobes length Sequence alignment. All alignments were visual. Sites coded with ‘*?” or with an IUPAC-IUB ambiguity code represent basepairs where se- quence produced with neither primer produced a sufficiently strong or clear signal for confident basepair assignment. Indels, coded as “‘-’’, were treated as missing data. Two regions were excluded from the ITS dataset due to ambiguous sequence alignment (positions 132—138 and 284— 291 of the aligned ITS data set). Evaluation of sequence data. Separate and combined analyses using a parsimony criterion were conducted for ITS and 3’trnK intron data. All analyses employed heuristic searches with 10,000 replicates of random sequence addition and tree-bisection-reconnection (TBR) branch swapping. Conservative estimates of clade sup- port were assessed using 10,000 replicates of the ‘fast’? bootstrap option in PAUP 4.0b5. Decay analyses (Donoghue et al. 1992; Bremer 1994) using Autodecay (Eriksson 1998) were conducted for the 3’trnK intron and the combined molec- ular analyses. Morphology Fresh and/or herbarium material was exam- ined from 80 localities (Appendix 1, Fig. 1) and 450 flowers. Nineteen characters were included for phenetic analyses, including 10 quantitative, 4 qualitative, and 5 ratio characters (Table 1). All 2010] characters are floral, because vegetative charac- ters are not generally useful for distinguishing species of Downingia. Characters were observed or measured against a ruler under a dissecting scope, except for anther horn length which was measured with an ocular micrometer. Morphometric analyses. Analyses of variance were conducted to identify characters differing significantly among the three currently recognized species, and Tukey tests were used to identify which species differed. The same was done within D. yina for the three groups identified by molecular analyses (see results). For multivariate analyses, a data matrix was created containing the average value for each character from each collection locality. Multivariate analyses included cluster analysis using Euclidean distances and single linkage, discriminant function analysis, and Principal Components Analysis (PCA), the latter using a matrix standardized so that each character had a mean of zero and a standard deviation of one. All statistical analyses were performed with SYSTAT 5.2.1. Cladistic analyses. One qualitative and five quantitative characters (indicated in Table 1) were used in a cladistic analysis of the 26 populations for which molecular data were also available, plus one population per outgroup taxon. Phylogenetic analyses using a parsimony criterion were con- ducted with PAUP 3.1.1 (Swofford 1993) or PAUP *4.0b5. The analysis employed a heuristic search with 100 replicates of random taxon addition and TBR branch-swapping. Qualitative characters excluded from the analysis were poly- morphic within most populations. Character states for the quantitative characters were deter- mined by searching for gaps within the character distribution among specimens that were greater than 2 times the average population standard deviation (Archie 1985). Most quantitative char- acters were excluded from the cladistic analysis because no character states could be defined. The character “‘locule’’, referring to the number of locules in the ovary, separates the D. yina complex from the outgroup taxa. The morphological data matrix is provided in Table 2. Cytology Chromosome counts were obtained from unpublished theses (Weiler 1962; Foster 1972) and from numerous specimens deposited at the UC and JEPS herbaria as chromosome vouchers (Appendix 1). Chromosome number was treated as an ordered character. All known chromosome counts for D. bacigalupii, D. elegans, D. bicor- nuta, D. cuspidata, D. ornatissima and_ D. montana report a single number for each of the species (Wood 1961; Foster 1972; Weiler 1962; Lammers 1993). All samples of these taxa were SCHULTHEIS: CRYPTIC SPECIES IN DOWNINGIA YINA 23 scored based on chromosome counts reported for the species, regardless of whether a count was obtained from the population sampled here. The only exception is D. bacigalupii sample 585-99, which was scored as unknown since the popula- tion is at the limits of the species range, and no chromosome counts were available from the vicinity. Chromosome counts for D. concolor are n = 8 and n = 9 (Weiler 1962; Lammers 1993). The samples of D. concolor included here fall within the known geographic range of n = 9 reports for D. concolor (Weiler 1962), and were scored as such. Populations of D. yina were scored based on the geographic proximity of the population to a population with a documented chromosome number (indicated in Table 2; Weiler 1962; Foster 1972). MacClade version 3.0 (Maddison and Maddison 1992) was used to reconstruct the most parsimonious chromosome numbers characterizing each node on trees produced from the combined analysis of the ITS and 3’trnK datasets. Analyses of Combined Molecular, Morphological, and Cytological Data A partition-homogeneity test (Farris et al. 1995) performed in PAUP *4.0 (Swofford 2001) confirmed combinability of the ITS plus 3’trnK data (P = 0.247; 1000 replicates, heuristic searches with random addition and TBR branch swapping), and of the molecular data with the morphological and cytological data (P = 0.094). Morphological and cytological data were com- bined as a single partition for the test since cytological data consisted of only one character (chromosome number). A branch and bound search of the combined data was conducted under a parsimony criterion. Clade support was assessed using 10,000 replicates of the “‘fast” bootstrap option. The morphological data came from the same or neighboring populations as the sequence data (Table 2). The cytological data consisted of chromosome numbers and did not include information regarding meiotic configura- tions of chromosomes in hybrid plants. Interfertility Information regarding interfertility and cross- ability among members of the D. yina complex comes from Weiler’s unpublished thesis (1962), in which he documented the results of numerous interspecific crosses within Downingia. His data include qualitative assessments of seed set, germination, and hybrid condition (e.g., flower- ing, green, chlorotic, dying in seedling stage), some quantitative assessments of pollen stain- ability, and analysis of meiotic configurations. Information regarding interfertility and cross- ability within D. yina comes from Foster’s 24 MADRONO TABLE 2. [Vol. 57 MORPHOLOGICAL DATA MATRIX USED FOR CLADISTIC ANALYSES. Sample numbers correspond to Appendix 1, with the following prefixes: B = Downingia bacigalupii, E = D. elegans, Y = D. yina, M = D. montana, = D. concolor, O = D. ornatissima, BI = D. bicornuta, CU = D. cuspidata. Characters and states are listed in Table 1. The “locule” character refers to the number of locules in the ovary [bilocular (0), unilocular (1)]. For D. yina the ““chromosome” character refers to the chromosome number based on reports or vouchers from the same or a neighboring population, indicated in parentheses. This character was included in the analyses of all data combined, but was not included in the analysis of morphological data alone. For some samples, the morphological data were combined with the molecular data from a neighboring population, indicated in parentheses. Sample 10 B Schultheis 585-99 1/0 B Schultheis 240-95 1 1/0 B Schultheis 237-95 0 B Schultheis 231-95 1/0 B Schultheis 251-95 E Schultheis 243-95 E Schultheis 242-95 E Schultheis 320-96 E Weiler 60138 (Foster 70-15-4) Y Oswald & Ahart 3943 Y Schultheis 247-95 Y Tracy 3217 Y. 7. Obrien s.n. Y Schultheis 236-95 Y D. Barbe 348 Y Schultheis 241-95 Y Schultheis 584-99 Y Schultheis 245-95 Y Weiler 61449 Y Schultheis 581-99 Y R. Bacigalupi 7978 Y Cook 962 Y Peck 16291 (Foster 68-210) Y Schultheis 319-95 Y Weiler 61333 (Foster 68-51) Y R. Bacigalupi 7894 BI Schultheis 100-95 C Schultheis 195-95 M Schultheis 235-95 CU Schultheis 179-95 (197-95) O Schultheis 180-95 — — SO 2 © oo: co oO Oo ooo Oo © © Oo 0 'o 0 oO oo] SS S| "CN unpublished thesis (1972). She documents meiotic configurations and pollen stainability for crosses between individuals of the same and different chromosome races. I assigned each of Foster’s parent populations to a molecular clade, based either on sequence data from her voucher specimens, or on close proximity of the vouch- ered population to a population with sequence data. I applied an ANOVA to Foster’s raw pollen stainability data to examine whether there were significant decreases in stainability in hybrids between versus within chromosome races, and between versus within molecular clades. RESULTS Cladistic Analyses Levels of divergence for the ITS dataset ranged from 0.0 to 0.017, excluding outgroups. Analysis Character 13 16 17 Locule Chromosome 0 2 0 12 0 l 12 0 [2 1/0 l 12 10 1 10 1/0 10 10 1/0 7 10° (Foster 70-96-11) 10 (Foster 70-96-11) 9 12 (Weiler 60207) 12 (Foster, Siskiyou) 12 (Foster, Harney) 12 (Foster 70-43-15) 12 (Foster 70-43-15) 10 (Weiler 61200) 10 (Weiler 61451) 10 (Weiler 61200) 8 6 (Foster 68-210) 12 (Weiler 61383) 10 12 (Foster,Wasco) 11 rook ooococososcoos congo coor ors sje Seoeoecoo.seoecoooeoooseocoeo =] So 2s cn) — OooooceococeoCcocococmclcmCcOClcC COCO OO Or SSE HK Ooo oO of ITS data resulted in 104 minimum-length trees based on 43 parsimony-informative characters (length = 100; CI = 0.90, 0.83 without uninfor- mative characters; RI = 0.91). Levels of diver- gence for the 3’trnK dataset ranged from 0.0 to 0.027, excluding outgroups. Analysis of the 3'trnK dataset resulted in 42 trees based on 19 parsimony-informative characters (length = 67; CI = 0.94, 0.83 without uninformative charac- ters; RI = 0.89). Combined molecular analyses produced 2 trees based on 43 parsimony-infor- mative characters (length = 158; CI = 0.91, 0.77 without uninformative characters; RI = 0.89). Combined molecular, morphological and cyto- logical analyses produced 72 trees based on 50 parsimony-informative characters (length = 176; C.I. = 0.87, 0.72 without uninformative charac- ters; RI = 0.86). All analyses (ITS dataset; 3’trnK dataset; combined molecular datasets; combined molecu- 2010] 98 83 an | 9 Fic. 2. SCHULTHEIS: CRYPTIC SPECIES IN DOWNINGIA YINA N N Grade III n=l2_ PD. yina D. Barbe 348 D. yina Oswald& Ahart 3 n=? n=|2 w= 12 1). bacigalupii 230-95 Wl? 1). bacigalupii 240-95 wl? PD. bacigalupii 251-95 W=12_ 1). bacigalupii 237-95 W127). yina 319-95 n=l2_ 7). yina Foster 70-43-15 w=l2_ 1). yina Foster 71-1-14 tl Ey BS yina_ T. Obrien sn wl2_ 1). yina 241-95 n=l2_ 7 yind R. Bacigalupi 7894 943 D. yina 236-95 62 n=10 D. yina Foster 70-84 We10 D. yina 581-99 Clade II D. yina R. Bacigalupi 797 D. bacigalupii 585-99 52 D. elegans Foster 70-15-4 D. yina Foster 71-13 D. elegans 242-95 D. elegans 243-95 Clade I D. elegans 320-96 D. yina 247-95 D. yina Foster 68-210 D. yina Foster 68-51 D. yina Foster 70-96-11 100 wll PD). montana 235-95 w= 1). montana 250-95 99 w=) 1). concolor 287-95 n= 1). concolor 195-95 n=l? D. ornatissima 180-95 n=l 7). bicornuta 100-95 WI) 1). cuspidata 197-95 Outgroups The strict consensus of 104 minimum-length ITS parsimony trees (length = 100; C.I. = 0.90, 0.83 w/o uninformative characters; R.I. = 0.91) produced from a heuristic search with 10,000 replicates of random taxon addition and TBR branch swapping. Numbers above the branches indicate bootstrap values generated from 10,000 replicates of the “‘fast’” bootstrap method. Sample collection numbers correspond to Appendix 1. If only a number is indicated, the collection was by Schultheis. Chromosome counts are uniform for all species except Downingia yina. Sources for D. yina chromosome counts are indicated in Table 2. lar, morphological and cytological datasets) except that of morphological data alone resulted in three main clades or grades (Figs. 2—5). Clade I comprised D. elegans and D. yina pro parte. Clade II comprised D. yina pro parte. Clade III comprised D. bacigalupii and D. yina pro parte. Primary differences among the trees produced from different analyses were the following: (1) There was a sister relationship between Clades I and II in trees resulting from analyses of all datasets but the ITS dataset, in which Clade II is aligned with grade III (Fig. 2). (2) Downingia elegans sample Foster 70-15-4 was resolved as part of Clade I in all trees except those resulting from analysis of the 3’trnK dataset, in which it fell in an unresolved position between Clades I and II (Fig. 3). This sample is from Snow Mountain, in Lake Co., California, at the southern limit of the species range (Fig. 1). (3) Downingia bacigalupii sample 585-99 is aligned with Clade I in ITS trees (Fig. 2), but is sister to other members of Clade III in all other trees. Sample 585-99 is from Josephine Co., Oregon, at the western periphery of the species range 26 MADRONO 62 89 FIG. 3. [Vol. 57 nl? D. bacigalupii 240-53 nl DD). bacigalupii 251-95 wD. bacigalupii 237-95 om : ea wale D. yina 319-96 ee 64 = —_ . Es ee . cuspidata 197-95 Fic. 4. The strict consensus of two minimum-length trees (length = 158; C.I. = 0.91, 0.77 w/o uninformative characters; R.I. = 0.89) from a heuristic search of combined ITS and 3’¢rnK intron data, with 10,000 replicates of random taxon addition and TBR branch swapping. Numbers above the branches indicate bootstrap values generated from 10,000 replicates of the “fast”? bootstrap method. Numbers below the branches are decay indices. Sample collection numbers correspond to Appendix |. If only a number is indicated, the collection was by Schultheis. Chromosome counts are uniform for all species except Downingia yina. Sources for D. yina chromosome counts are indicated in Table 2. divergence between lobes of the lower corolla lip in particular distinguished D. elegans and D. bacigalupii from D. yina. The former two taxa had more sharply bent anthers and less divergent lower corolla lobes than D. yina. The filament of D. bacigalupii was longer on average than that of D. elegans and D. yina. Additionally, D. elegans could be distinguished by the lack of yellow pigmentation on the lower corolla lip. PCA analyses of all samples using all data showed clear separation among D. elegans, D. bacigalupii, and D. yina, particularly when principal components I and III were plotted (Fig. 6). This separation was also clear when only ratio characters were used or when ratio charac- ters were excluded. Characters of particular importance in the PCA analyses were the anther angle, the filament/back slit ratio, and the angle of divergence between the lower corolla lobes. The percent of total variance explained by components I, IH, and III was 45.2%, 17.5%, and 12.0%, respectively. 73 MADRONO D. [Vol. 57 bacigalupii 240-95 ; D. bacigalupii 251-95 . D. bacigalupii 237-95 jam n=12 Dias . = 70 - YING 241-95 w Includes D. bacigalupii and as) ' 1 n=12 ; D. pulcherrima (formerly D. yina D. yina 319-96 eee Dorm) os Oo n=12 . 2 D. yina 236-95 i ? n= jt D. yina T. O'brien (Siskiyou n=? : us 96 D. bacigalupii 585- 5 n= 10 = D. yina 581-99 a 97 y — Includes D. yina s.s. 4 =10 , - D. yina R. Bacigalupi 7978 | UO n= 10 D. elegans 320-96 87 62 = 10 3 = D. elegans 242-95 1 n= 10 — D. elegans 243-95 -¥ Includes D. elegans and n= 10 5 S LD. willamettensis (formerly D. yina) l 66 D. elegans Foster 70-15-4 s aa ! n= 10 D. yina 247-95 n= 10 . D. yina Foster 68-51 =i - D. montana 235-95 = N : Leet §=1). concolor 195-95 a n=12 ee © D. ornatissima 180-95 | 5p _ n= 11 ; = os D. bicornuta 100-95 © 9 11 - D. cuspidata 197-95 FiG. 5. The strict consensus of 72 minimum-length trees (length = 176; C.I. = 0.87, 0.72 w/o uninformative characters; R.I. = 0.86) produced from a branch-and-bound search of combined data. The data matrix included the ITS, 3’trnK, morphological, and cytological data. Numbers above the branches indicate bootstrap values generated from 10,000 replicates of the “‘fast’’ bootstrap method. Numbers below the branches are decay indices. Sample collection numbers correspond to Appendix |. If only a number is indicated, the collection was by Schultheis. Chromosome counts are uniform for all species except Downingia yina. Sources for D. yina chromosome counts are indicated in Table 2. Cluster analysis (not shown) produced two main groups, one with D. yina samples and the other with a mixture of D. elegans and D. bacigalupii samples. One sample of D. elegans (Ehlers & Erlanson 39) and one sample of D. bacigalupii (582-99) together joined at the base of the D. yina cluster. Variation within Downingia yina. Univariate analyses revealed that significant character dif- ferences were evident between D. yina samples from the three main molecular clades, but with overlapping ranges (Table 4). No qualitative characters could be used to uniquely identify the three groups. In general, Clade II samples tended to be smaller for most quantitative characters measured (Table 4). Samples from Clade I tended to have a less sharply bent anther and a wider angle of divergence between the lobes of the lower corolla lip than did samples from Clades II and III. PCA analyses of only D. yina samples using all data did not clearly distinguish between samples from Clades I, I], and HI (not shown). Discrim- inant function analysis of D. yina samples showed better separation of the three groups, but with areas of overlap (Fig. 7). Characters of particular importance in the discriminant function analyses were anther length, the angle of divergence between the lobes of the lower corolla lip, and trichome density on the dorsal anther surface. Cluster analysis (not shown) grouped all of the D. yina samples together, but did not resolve groups corresponding to Clade I, II, and HI samples. 2010] SCHULTHEIS: CRYPTIC SPECIES IN DOWNINGIA YINA 29 TABLE 3. UNIVARIATE STATISTICS FOR THE DOWNINGIA YINA COMPLEX. Means, standard deviations and ranges (in parentheses) are provided for each character within each species. Superscripts indicate groups that are significantly different from one another using Tukey multiple comparison tests following ANOVA. Groups with no superscript or that share a superscript are not significantly different. D. elegans D. yina D. bacigalupii Character (n = 62) (i= 317) (n = 69) Sepal (mm) 5.31 + 1.30% (3.0-8.0) 4.75 + 1.28" (0.50—10.0) 5.81 + 1.80% (3.0—-10.0) Side slit (mm) 2.33 + 0.45% (1.5-3.0) 4.10 + 0.718 (1.75-6.0) 2.96 + 0.54© (2.04.25) Back slit (mm) 4.48 + 0.75“ (3.0-6.0) 4.80 + 0.898 (2.25-—7.5) 3.87 + 0.91° (2.25-6.0) Upper lobe (mm) 4.04 + 1.16* (2.0-7.0) 3.71 + 0.93% (2.0-6.75) 6.75 + 1.35® (4.0-11.0) Lower lobe (mm) 6.53 + 1.82“ (3.5—-12.0) 6.18 + 1.30% (3.0-9.5) 8.40 + 1.80" (5.0-14.0) Filament (mm) 5.27 + 1.48% (2.5—7.5) 3.17 + 0.927 (1.25—5.75) 7.48 + 1.50© (3.25-9.75) Anther (mm) 2.68 + 0.46% (1.5-3.5) 2.18 + 0.368 (1.25—3.0) 2.89 + 0.42° (1.25-3.75) Anther angle (degrees) 84.07 + 14.40% (28.0-90.0) 22.18 + 9.11? (0.0—-51.0) 88.80 + 6.75% (38.0—90.0) Lower angle (degrees) 9.16 + 11.924 (0.0—-50.0) 53.69 + 12.798 (20.0—-90.0) 12.31 + 9.88% (0.0—30.0) Horns (mm) 0.44 + 0.08“ (0.26-0.75) 0.41 + 0.114 (0.13-0.79) 0.60 + 0.108 (0.32—0.86) Filament/anther 1.95 + 0.34% (1.2—2.55) 1.45 + 0.30" (0.625—2.375) 2.58 = 0.35% (1.63.6) Side slit/back slit 0.53 + 0.10% (0.35—1.0) 0.86 + 0.09% (0.47-1.3) 0.79 + 0.13° (0.5-1.11) Filament/back slit 1.18 + 0.23% (0.56-1.75) 0.66 + 0.18" (0.38-2.0) 1.97 + 0.36 (1.42-3.56) Upper lobe/lower lobe 0.64 + 0.15“ (0.25-1.0) 0.61 + 0.14% (0.31-1.28) 0.81 + 0.14% (0.54-1.14) Back slit/upper lobe 1.19 + 0.35“ (0.6—2.22) 1.37 + 0.418 (0.5-3.11) 0.59 + 0.18° (0.35—1.14) Cytology = 6 and nm = 8 were documented from Marion and Lane counties in Oregon, respectively (Weiler Chromosome numbers within the Downingia yina complex appear to correspond to the molecular clades identified with ITS and 3’trnK sequences (Figs. 2-4, Appendix 1). All samples in molecular Clade I for which chromosome counts were available were n = 10 in D. elegans and n = 6,8 or 10 in D. yina. Downingia yina counts of n a 2 fH Z zZ ] 1 Oy = S 'S) = = 0 iS) E a A, -] -2 -3 -2 -1 O i 2 S 4 PRINCIPAL COMPONENT (3) Fic. 6. Plot of principal components one and three using the characters listed in Table 1. Filled symbols _ represent Clade I. Symbols with a line through the center represent Clade II. Open symbols represent Clade HI. Triangles = Downingia elegans. Squares = _ D. bacigalupii. Circles = D. yina. 1962; Foster 1972). All samples in molecular Clade II were D. yina with n = 10. All samples in molecular Clade III were n = 12 in either D. yina or D. bacigalupii. Character state reconstruction suggests an ancestral state of m = 10 in Clades I and II, and an ancestral state of m = 12 in Clade III. The ancestral state for the entire D. yina complex is equivocal. Interfertility Foster’s results (1972) show that within D. yina, crosses between populations with different chromosome numbers showed a significant re- duction in pollen stainability relative to crosses between populations with the same chromosome numbers (Table 5; Foster 1972). Similarly, pollen stainability was significantly reduced in crosses between populations presumed to be from different molecular clades relative to those presumed to be from the same molecular clades (Table 5; Foster 1972). Weiler’s results (1962) from interspecific recip- rocal crosses (Table 6) reflect Foster’s results within D. yina in that pollen stainability and meiotic irregularities seemed to be affected more by differences in chromosome number than by species identification (D. elegans, D. bacigalupii, or D. yina). Crosses between D. elegans (n = 10) and n = 10 populations of D. yina, for example, produced 10 bivalents and no significant reduc- tions in pollen stainability (Table 6), in contrast to the reduction in pollen stainability for crosses within D. yina but between populations of different chromosome number (Table 5; Foster 1972). 30 MADRONO [Vol. 57 TABLE 4. UNIVARIATE STATISTICS WITHIN DOWNINGIA YINA. Means, standard deviations, and ranges (in parentheses) are provided for each character within inferred molecular clades. Superscripts indicate groups that are significantly different from one another using Tukey multiple comparison tests following ANOVA. Groups with no superscript or that share a superscript are not significantly different. Clade I (n = 100) 1.0348 (3.0-7.75) 0.63 (2.75—6.0) 0.784(3.25-7.0) 0.718 (2.0-5.75) 0.96© (3.25-8.5) 0.62° (2.0-4.75) 0.33° (1.25-2.75) 8.108 (2.0-40.0) 12.24® (35.0—90.0) 0.08 (0.19-0.61) 0.23 (1.0—2.0) 0.08% (0.62—1.07) 0.08® (0.5—0.9) 0.12° (0.33—0.9) 0.448 (0.78-3.11) Character Sepal (mm) Side slit (mm) Back slit (mm) Upper lobe (mm) Lower lobe (mm) Filament (mm) Anther (mm) Anther angle (degrees) Lower angle (degrees) Horns (mm) Filament/anther Side slit/back slit Filament/back slit Upper lobe/lower lobe Back slit/upper lobe oo ~ NS) te Nae lee a dae de de tae ae Uae it Tae DISCUSSION The Downingia yina species complex currently comprises three species that are readily distin- guished from one another on the basis of morphological characteristics (Weiler 1962; Ayers 1993; Fig. 6, Table 3). Downingia elegans and D. bacigalupii differ from D. yina in that the anthers form a sharp angle relative to the filaments, and the lower corolla lobes are relatively parallel versus divergent in D. yina. The chromosome numbers and the yellow patches on the lower corolla lobes readily distinguish D.bacigalupii 4 © 3 fe) n | 2 fe) N ° ° [e) : ae Ee T- $8 3 4 a re) 000 {o) 6 (e) e ee ry ie) é 4 | J sf 3. = ie) @ 8 ed ° ® S -]> ye. a o a S ° ° -7 i= e e =a 3 4 . e 4 | | | | =A ae O 2 4 6 CANONICAL FACTOR 1 Fic. 7. Plot of canonical factors one and two from discriminant function analysis of Downingia yina samples using the characters listed in Table 1. Filled circles = Clade I. Circles with a line through the center = Clade I. Open circles = Clade III. Clade II (n = 30) 4.22 + 0.96" (0.5—6.0) 3.40 + 0.49 (2.75-4.5) 3.74 + 0.538 (3.0-5.0) 3.44 + 0.81% (2.05.0) 5.09 + 0.99" (3.25-6.75) 2.33 = 0.45" (1.75-3:5) 1750033" (125-2725) 22.50 + 6.114? (7.0-33.0) 46.27 + 12.40% (20.0-68.0) 0.34 + 0.04? (0.26—-0.42) 1.37 = 0.30 (0.875—2.2) 0.91 + 0.09? (0.77—-1.25) 0.62 + 0.084 (0.47—-0.75) 0.68 + 0.138 (0.45—0.92) 1.17 + 0.414 (0.60—2.22) Clade HII (n = 187) 4.82 + 1.424 (2.0-10.0) 4.17 + 0.734 (1.75-6.0) 4.94 + 0.884 (2.25-7.5) 3.90 + 0.994 (2.0-6.75) 6.46 + 1.394 (3.0-9.5) 3.40 + 1.014 (1.25-5.75) 2.28 + 0.324 (1.25-3.0) 23.96 + 9.524 (0.0-51.0) 52.06 + 12.08% (22.0-78.0) 0.42 + 0.134 (0.13-0.77) 1.48 + 0.34 (0.625~2.375) 0.85 + 0.10% (0.47-1.31) 0.69 + 0.22 (0.38-2.0) 0.62 + 0.15“ (0.31-1.3) 1.34 + 0.38 (0.50—2.5) from D. elegans. As outlined in the introduction, previous workers (Peck 1934, 1937; McVaugh 1941, 1943) recognized that D. yina may represent multiple taxa, which were variously named: D. yina Applegate, D. yina Applegate var. major McVaugh, D. willamettensis Peck, D. pulcherrima Peck. Recent molecular analyses lent merit to these interpretations, but sampling within the D. yina complex was very limited (Schultheis 2001). The additional molecular data presented here substantiates these patterns, and demonstrates that samples of D. yina fall into three separate molecular clades, with D. elegans and D. baciga- lupii nested within two of these three clades (Figs. 2-4). Taken independently, this paraphy- letic or polyphyletic pattern with respect to the sequence data from either the nuclear or chloro- plast genomes (Figs. 2-4) might only represent gene rather than organismal phylogenies (Doyle 1992; Knox 1998). High resolution molecular data are expected to reveal patterns in which paraphyletic progenitor species (with respect to the molecular data) give rise to monophyletic derivative species (Rieseberg and Brouillet 1994; Graybeal 1995; Olmstead 1995), in this case D. TABLE 5. MEAN CROSSES BETWEEN PERCENT STAINABLE POLLEN IN | POPULATIONS OF DOWNINGIA | yYINA. Significant differences occur for populations | with the same versus different chromosome numbers » and for populations from the same versus different molecular clades. (Raw data taken from Foster (1972) | and reanalyzed). Cross type Mean SE n P value Chromosome numbers same 93.4 84 5 0.001 Chromosome numbers differ 49.9 5.6 II Same molecular clade 83.0 8.3 7 0.007 Different molecular clades AS3 1:3 .9 | j 2010] TABLE 6. MEAN PERCENT STAINABILITY AND MEIOTIC CONFIGURATIONS IN INTERSPECIFIC CROSSES WITHIN THE DOWNINGIA YINA COMPLEX. Data taken from Weiler (1962). D. elegans D. bacigalupii D. elegans >95% n= 10 D. bacigalupii 30.8—78.3% = 93570 n= 12 Ich3 + 9II + II D. yina 51.3-79.5% >95% n= 12 1201 D. yina >95% 50-70% n = 10 101 Ich3 + 9II + II yina independently giving rise to D. elegans and D. bacigalupii. If D. yina populations were integrated through gene flow with one another, but to the exclusion of D. elegans and D. bacigalupii, D. yina would eventually proceed to monophyly with respect to the molecular data, and the currently recognized species would be appropriate, or could be accommodated with terms indicating their unresolved or transitional status (““metaspecies”, Donoghue 1985; “‘ferre- species’, Graybeal 1995; “‘plesiospecies”’, Olm- stead 1995). What is compelling in this example is the correspondence of gene geneologies from more than one gene with geographic, cytological, and interbreeding data; a correspondence that makes a case for multiple organismal lineages (Avise 1994), and thus multiple species (de Queiroz 1998, 1999) within D. yina. Geography In the D. yina complex, cytological races and molecular clades appear to be roughly segregated along the Cascade Ranges (Fig. 1). When a concordant pattern emerges between phylogenet- ic and geographic subdivisions of a group, this often indicates little to no gene flow among subdivisions. This point has been emphasized in phylogeographic studies (Avise et al. 1987) and has received confirmation from population ge- netic models (Slatkin 1989). The correspondence between molecular clades within the D. yina complex and the distribution of these clades to either the west or east of the Cascade Ranges is striking (Figs. 1-5), and suggests that the moun- tain range serves as a geographic barrier to gene flow. The Cascade Ranges have been recognized as a geographic barrier in other contexts, clearly affecting differences in climate (Peck 1941; Orr and Orr 1996), and floristic composition (Peck 1941) to the west versus the east. The Klamath- Siskiyou region at the California-Oregon border is where the striking segregation of D. yina ‘molecular clades to the east and west of the Cascade Ranges is much less evident (Fig. 1). SCHULTHEIS: CRYPTIC SPECIES IN DOWNINGIA YINA 3] Clade I is found to the west of the Cascade Ranges, except that D. e/egans extends eastward into eastern Washington and Idaho. Clade II is localized to a region in the Cascade Range of southern Oregon (Fig. 1), in the vicinity of Lake of the Woods and Upper Klamath Lake, Oregon, and cannot readily be designated as “east” or “west”. Molecular clade HI is primarily east of the Cascades, but extends west into the Klamath- Siskiyou region. It is possible that the Klamath- Siskiyou region was the source from which the D. yina complex dispersed northward to the east and west of the Cascades, a scenario similar to hypotheses of post-glaciation dispersal presented by Whittaker (1961) and Soltis et al. (1997). Cytology Cytological variation within the D. yina complex mirrors the molecular phylogeny and the geography for the group, with n = 12 samples primarily east of the Cascade Ranges, and n = 10 samples primarily to the west. Populations of D. yina within Clade I have n = 6 or 8 in the northwestern reaches of the species range, an observation which prompted Foster (1972) to suggest a trend of decreasing chromosome numbers as one progressed from the southeast to the northwest of D. yina’s range. Foster’s (1972) proposed explanation for this trend, based on meiotic configurations in numerous hybrids between the different chromosome races of D. yina, was that the races arose through Robertso- nian translocations producing either a dysploid series of reductions from a starting point of n = 12, or a series of reductions from n = 11 with an increase to nm = 12. Foster’s work (1972) unfortunately did not include D. e/egans and D. bacigalupii, perhaps because the potential deriva- tion of these taxa from within D. yina was not reflected in the taxonomy. If D. elegans and D. bacigalupii arose from within D. yina, as suggest- ed by the molecular data, the simplest explana- tion 1s that they arose from n = 10 and n = 12 populations of D. yina, respectively. The homol- ogy of D. elegans and n = 10 D. yina genomes, and of D. bacigalupii and n = 12 genomes is supported by interfertility data discussed below. Interfertility If D. yina contains the multiple divergent lineages suggested by the molecular data, one might expect levels of interfertility to correspond with the molecular clades. Indeed, levels of interfertility appear to correspond more with the molecular clades and the chromosome numbers of the populations examined than with species identification. For example, individuals of D. yina from Clade I show greater interfertility with D. elegans than with individuals of D. yina 32 from Clade III (Tables 6 and 7). Similarly, individuals of D. yina from Clade III show greater interfertility with D. bacigalupii than with individuals of D. yina from Clades I or I. In sum, patterns of interfertility do not appear to correspond to the species currently recognized, but do appear to correspond to chromosome races and molecular data, both of which corre- spond to geography. While levels of fertility may be reduced in crosses between chromosome races or molecular clades, reproductive barriers are not complete. Nor are reproductive barriers complete among the three species currently recognized. Popula- tions exist with hybrids between D. bacigalupii and D. yina, and between D. elegans and D. yina (Weiler 1962; Schultheis personal observation). These populations may either resemble a hybrid swarm, with a wide variety of hybrid forms, or may contain readily distinguishable parental forms and only a few hybrids (Weiler 1962; Schultheis personal observation). Regardless of whether reproductive barriers are complete or incomplete, the currently recognized species of the D. yina complex do not correspond to patterns of interfertility within the group. Hypothesized Organismal Lineages Within the Downingia yina Complex In sum, there appear to be three main lineages within the D. yina species complex. Members of the first lineage (Clade I) are characterized by either a ““D. yina” or ““D. elegans” morphology, and are distributed primarily west of the Cas- cades, with D. elegans extending eastward into eastern Washington and Idaho. “‘D. yina’’ indi- viduals are n = 6, 8, or 10. “D. elegans” individuals are n = 10. Within this lineage, the “D. elegans’ members form a clade, excluding sample Foster 70-15-4, from the southern periph- ery of the “D.elegans” range. The scant support for the “D. elegans” clade comes from morpho- logical characters, some of which are polymor- phic within populations (Table 2). The second hypothesized lineage (Clade II), localized in the Lake of the Woods region of the Cascades in southern Oregon, is characterized by a “D. yina” morphology and n = 10. Support for this clade comes entirely from molecular charac- ters. Members of the third hypothesized lineage (Clade III) are characterized by either a “‘D. yina”’ or “D. bacigalupii’ morphology, n = 12, and a distribution primarily to the east of the Cascades, into southwestern Idaho and western Nevada, and extending westward into the Klamath/ Siskiyou region of southern Oregon and northern California. Within this lineage, the ““D. bacigalu- pii’ samples form a clade to the exclusion of sample 585-99, from the western periphery of the MADRONO [Vol. 57 range. The ““D. bacigalupii”’ clade is supported only by morphological characters (Table 2). Morphological analyses presented here were unable to clearly distinguish among D. yina samples falling into different molecular clades (Fig. 7; Table 4), which largely correspond to variation in D. yina chromosome numbers. Similarly, Foster (1972) was unable to find morphological differences corresponding to the chromosome races within D. yina. The chromo- some races and the molecular clades within D. yina are morphologically cryptic. Further exam- ination of morphology may reveal differences missed thus far, but even in the absence of such differences, it is desirable to recognize what are hypothesized to be organismal lineages. Based on the information currently available, I choose to recognize five species, with names assigned based on nomenclatural priority and the phylogenetic placement of the type specimens: D. elegans (Lindley) Torrey, D. bacigalupii Weiler, D. yina Applegate, D. willamettensis Peck, and D. pulcherrima Peck. Ideally taxon names, including species names, should only be assigned to clades (Mishler and Donoghue 1982; Misher and Theriot 2000). This strict application of a phylogenetic species concept only applies full species status to D. elegans, D. bacigalupii, and D. yina sensu stricto. Downingia willamettensis and D. pulcherrima comprise the “‘D. yina’’ samples from Clades I and III respectively. These samples are not resolved as clades, but may still be named as metaspecies (Donoghue 1985), plesiospecies (Olmstead 1995) or ferrespecies (Graybeal 1995). An alternative to recognizing five species is to recognize a single species, D. elegans (based on nomenclatural priority), and five varieties. While both alternatives recognize the same taxa, differ- ing only in the rank applied (species or variety), the recognition of five species more clearly emphasizes the molecular, cytological and fertility diversity | within this complex group. In this case, names are also available at the species rank whereas new | names or combinations would be needed if the taxa were recognized at the varietal rank. Features of the five taxa are summarized in Table 7. It is unfortunate that the three species — previously referred to D. yina (D. yina s.s., D. willamettensis, and D. pulcherrima) are morpho- logically indistinguishable given current informa- | tion. Even those features of most importance in| discriminate function analysis (included in Ta- | ble 7) show such overlap as to be of minimal use for field identification. Weiler (1962) did note. that D. yina tended to be decumbent in the west | and erect in the east (which would correspond to | D. willamettensis and D. pulcherrima respective- | ly), but this can be difficult to detect on herbarium sheets. This feature, as well as corolla coloration (particularly useful for distinguishing | D. elegans and D. bacigalupii), is worth noting in (oa) laa) SCHULTHEIS: CRYPTIC SPECIES IN DOWNINGIA YINA 2010] ,(JuepUNnqeR 0] dUOU 3q UPd) Maj A[TRIOUNS 400 + UU CC OT + wu p¢ juosoid x(T] H Sd0IdOp TS) JUISIDAIP (C6 + Sd0IS9P (PZ) JUNq ATdseys jou uOSdIIQ UIOYINOS pure eIusojI[ed usoyyiou jo UOIBI1 NOATYSIC-Y PUL] OUI PIBMSOM SUIPUS}X9 ‘BIUJOJI[ED pue ‘UOSIIO ‘UOISUTYSe AA Ul SAPRVdSeD Jo F W QOOT> AT[es19ue3 Tl =U puid ‘q puldsayojnd -q ,(JUBpUNQe 0} sU0U 3q uvd) DUMAIaYyojnd ‘g pure puid ‘gq uevy} Juepunqe a1oul AT[e19uas ao Q + WU ['Z 90 + WU Oe’ juasoid x(T] = Sooisop 6S) JUISIDAIP (1'8 = sooisop /°8T) yuoq Ajdieys jou BINION (e) UIO]sSoOMYy}IOU ul sosuRy JSVOD YON JO AA ‘UOBSIIO pue uoiSsuUryseM UI SOpBdseD JO M VO “OD 94PT UI UW QS9 -W OST> y(JuUepUuNnqeR 0] 9U0U 9q UBD) puisiayojind “q pue sisuajJaUD]JIM ‘q uey) JURpUNge sso] AT[e10ues «xe OQ + WU C/T CO + WU ¢'? jyuosoid a(Cl = SddIZ9P Op) JUIBADAIP (19 = saoisap STZ) Jusq Aj[dievys jou SPOOM 24} JO dxVT pue oye] yewmryy JaddyQ ussdjsomy}i0u udaMjJoq ‘UOSIIO uray Nos JO SOPBISBD 0} POZI][RIO] Ww OTSI—00cI Ol =u put ‘Gg (juepunqge 0} ouUOU 3q UD) Mo} AT[eIOUNS VO + WU 67 (sorsods 19y}0 ueYy dSRIDAB UO JOSUOT) GS] = WWW Cy yuosoid (66 % SddIsOp CT) [oT[esed Ajivou (L°9 = SadIsap 9°9g) UAaq A[dseys BPPADN UI9}SOM pure oyep] UId}soOMYINOS SUOBIIO uIOyINOS puP KIUIOJITED UsINYIIOU JO UOISSI NOATYSIS-YCUIL | OJUL PIBMISIM SUIPUD}X9 [UOSIIOC pue vIUulojed Ul sapeosend Jo q a QOOC> Tl =u udnjpsiovq ‘q (uepuNqe oO} sU0U 9q UBD) MOJ AT[RIOUNS 50 = EME ? CT + wu ¢’¢ juosqe (6 1] += saeisap 76) [orfered Ayresu (prt + SdaIsap ['pg) yUNq Aydieys oyep] pue uojsuIYyseM UIO]SVO OUI PIBM]SLd SUIPUD}XO SBIUIOJI[ED UI sosURY ISVOD YON JO M ‘UO SUTYSeAA pue UOSIIO UI sapRdseED JO MA Ww 000T> Ol =u SuDBIJA ‘G SISUIJJOUID] JIM "J puid ‘gq udnjps1ovq ‘q SUDBI]2 “CT doRjAns [esiop Jayjue uO SOWOYSI | ysud] 1oyVUuy aqn} JUsUIPII "jeOIY} B]JOIOD UL MOTIAA Sogo] B][OIOD IOMO'T aqn} JUSUIR]Ly 0} SATII g[sue IdyUYy ASooyd1op TOUTICER STP o1ydeiso0ay UONBAIA ASO[OVD UONBOTISSL[D IOUIO] ‘({ xtpusddy) Apnjs sty} url sodures wiorjy syUdWIOINSROW UO paseg ‘UOT]eIADP plepuR\s | Uv ‘VUId ‘G UTYIIM SISA[VUR UOTOUNJ JUVUILULIOSIP UI POTJNUSP!I SoINeoJ SOJBOIPUT x ‘VNIA (d AONV IdQTVDIOVE “d ‘SNVYOITA VIONINMOCG SV GAIAISSVTIQ ATYAWNYOA SAIOddS FAI AHL ONIHSINONILSICGQ: SAYNLVAY AHL AO AUVNWWNAS ‘L ATAVL$e 34 MADRONO STATEHWY 138 ys ZA £2 423N, 123.3 W Medford Cc Om A Scale 1:2,000,000 FIG. 8. Lake of the Woods ( [Vol. 57 Legend © Downingia pulchernma va Downingia yina Oregon Ecoregions SS Cascades SS) Eastern Cascades Slopes and Foothills ea) Klamath Mountains Klamath Falls Map illustrating the distribution of Downingia yina sensu strictu relative to adjacent D. pulcherrima populations in the southern Cascade Range of Oregon. The map does not illustrate D. bacigalupii, which is also found in the pictured region. Triangles = D. yina s.s. (n = 10, Clade II). Circles = D. pulcherrima (n = 12, Clade III). Downingia yina s.s. samples are located within the southern tip of the Cascades ecoregion of Oregon (following Thorson et al. 2003). future collections. Unless reliable features are identified, we must rely on geographic location for field identification, ideally with confirmation from molecular and/or cytological data. At present I recommend that specimens collected west of the Cascades in Oregon and Washington, and west of the North Coast Ranges in California are best assigned to D. willamettensis. Specimens collected east of the Cascades in Oregon or Washington are best assigned to D. pulcherrima. Downingia pulcherrima is also located in the Klamath and Siskiyou regions of northern California (documented in this study as far west as Coffee Creek, just west of Clair Eagle Lake, Trinity Co.) and southern Oregon (documented in this study as far west as Medford, Jackson Co.). Downingia pulcherrima and D. willametten- sis are generally above and below elevations of 250 m respectively. Downingia yina sensu strictu 1s localized to the southern tip of the Cascade Range in Oregon. This study documents popula- tions from the northwestern edge of Upper Klamath Lake to Lake of the Woods (Klamath Co.). Based on my current understanding of the distribution for D. yina, | recommend assigning to this taxon any collections found in the Cascades ecoregion of southern Oregon (eco- region as delimited in Thorson et al. 2003), while assigning those found in neighboring areas outside of this ecoregion to D. pulcherrima. Figure 8 provides a map delimiting the distribu- tion of D. yina relative to D. pulcherrima. Priorities for refining our current understand- ing of this species complex include obtaining molecular data from additional populations (particularly at the limits of species ranges, including Washington state) additional sampling of cytological variation, and exploration of morphological or ecological features to distin- guish D. yina sensu strictu, D. willamettensis, and D. pulcherrima. Key to Taxa of the Downingia yina Species Complex la. Anthers abruptly bent, >70° to filaments; lower corolla lip lobes + parallel. 2a. Corolla 3-colored (blue, white, yellow); lower corolla lobes obtuse, mucronate Steal ie Gas oo cS gS Boi IN SN femee Nee D. bacigalupii 2b. Corolla 2-colored (blue, white); lower corolla 1Obés acute. 240 ie uace es D. elegans lb. Anthers not or + bent, <45° to filaments; lower corolla lip lobes divergent, not parallel. 3a. Plants generally east of Cascades, extend- ing into Klamath Ranges in southern Oregon and northern California; generally >250 m (but <250 m along Columbia River, Washington). 4a. Localized to southern Oregon Cas- cades, between northwestern Upper Klamath Lake and Lake of the Woods, plants at 1200-1510 m... D. yina 4b. East of Cascades, extending into Klamath Ranges in southern Oregon and northern California, plants gener- ally at: 2000 me 2a. 23 D. pulcherrima 2010] 3b. Plants generally west of Cascades in Oregon and Washington, and west of North Coast Ranges in California; generally <250 m (but 650 m on Snow Mountain, Lake Co., Califormiay =... Sew ees eek D. willamettensis ACKNOWLEDGMENTS This work represents partial fulfillment of the requirements for obtaining a Ph.D. from the University of California at Berkeley. Thank you to my committee members Bruce Baldwin, Brent Mishler and Tom Bruns, for commenting on a much earlier draft of this work, and to Staci Markos for commenting on a recent draft. A special thanks to Sue Bainbridge and K. Allison Lenkeit-Meezan for assistance with preparing maps. Nancy Morin and an anonymous reviewer provided constructive and appreciated comments. Thank you to the JEPS and UC herbaria for graciously providing me with space to work over an extended period of time, and to Ellen Dean at DAV for access to unmounted specimens. The Lawrence Heckard Endow- ment Fund of the Jepson Herbarium provided research and publication support. LITERATURE CITED APPLEGATE, E. I. 1929. Two new Downingias from Oregon. Contributions from the Dudley Herbari- um of Stanford University 1:97—98. ARCHIE, J. W. 1985. Methods for coding variable morphological features for numerical taxonomic analysis. Systematic Zoology 34:326—-345. AVISE, J. C. 1994. Molecular markers, natural history and evolution. Chapman & Hall, New York, NY. , J. ARNOLD, R. M. BALL, JR., E. BERMING- HAM, T. LAMB, J. E. NEIGEL, C. A. REEB, AND N. C. SAUNDERS. 1987. Intraspecific phylogeogra- phy: the mitochrondrial DNA bridge between population genetics and systematics. Annual Re- view of Ecology and Systematics 18:89—522. AYERS, T. 1993. Downingia. Pp. 460-462 in J. C. Hickman (ed.), The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. BALDWIN, B. G. 1992. Phylogenetic utility of the internal transcribed spacers of nuclear ribosomal DNA in plants: an example from the Compositae. Molecular Phylogenetics and Evolution 1:3—16. BAUDER, E. 1992. Ecological monitoring of Downingia concolor ssp. brevior (Cuyamaca Lake downingia) and Limnanthes gracilis ssp. parishii (Parish’s slender meadowfoam). Report numbers 844-06- 328, FG 7414, FG 9443. California Department of Parks and Recreation Southern Division, and California Department of Fish and Game Natural Heritage Division, Sacramento, CA. BREMER, K. 1994. Branch support and tree stability. Cladistics 10:295—304. DE QUEIROZ, K. 1998. The general lineage concept of species, species criteria, and the process of specia- tion: a conceptual unification and terminological recommendation. Pp. 57—78 in D. J. Howard and S. H. Berlocher (eds.), Endless forms: species and speciation. Oxford University Press, New York, NY. . 1999. The general lineage concept of species and the defining properties of the species category. Pp. 49-89 in R. A. Wilson (ed.), Species: new SCHULTHEIS: CRYPTIC SPECIES IN DOWNINGIA YINA 35 interdisciplinary essays. The MIT Press, Cam- bridge, MA. DONOGHUE, M. J. 1985. A critique of the biological species concept and recommendations for a phylo- genetic alternative. Bryologist 88:172—-181. , R. G. OLMSTEAD, J. F. SMITH, AND J. D. PALMER. 1992. Phylogenetic relationships of Dip- sacales based on rbcL sequences. Annals of the Missouri Botanical Garden 79:249—265. DoyLe, J. J. 1992. Gene trees and species trees: molecular systematics as one-character taxonomy. Systematic Botany 17:144—-163. AND J. L. DOYLE. 1987. A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochemical Bulletin 19:11—15. ERIKSSON, T. 1998. AutoDecay, ver. 4.0 (program distributed by author). Department of Botany, Stockholm University, Stockholm, Sweden. FARRIS, J. S., M. KALLERSJO, A. G. KLUGE, AND C. BULT. 1995. Testing significance of incongruence. Cladistics 10:315—319. FOSTER, R. I. 1972. Explosive chromosome evolution in Downingia yina. Ph.D. dissertation. University of California, Davis, CA. GRAYBEAL, A. 1995. Naming. species. Biology 44:237-250. HILLIs, D. M., B. K. MABLE, A. LARSON, S. K. DAVIS, AND E. A. ZIMMER. 1996. Nucleic acids IV: sequencing and cloning. Pp. 321-381 in D. M. Hillis, C. Moritz, and B. K. Mable (eds.), Molecular systematics, 2nd ed. Sinauer Associates, Inc., Sunderland, MA. KNOX, E. 1998. The use of hierarchies as organizational models in systematics. Biological Journal of the Linnean Society 63:1—49. KUCHLER, A. W. 1964. Potential natural vegetation of the conterminous United States. American Geo- graphical Society Special Publication no. 36. Amer- ican Geographical Society, New York, NY. LAMMERS, T. G. 1993. Chromosome numbers of Campanulaceae. III. Review and integration of data for subfamily Lobelioideae. American Journal of Botany 80:660—675. MADDISON, W. P. AND D. R. MADDISON. 1992. MacClade: analysis of phylogeny and character evolution, ver. 3.0. Sinauer, Sunderland, MA. McCVAUGH, R. 1941. A monograph of the genus Downingia. Memoirs of the Torrey Botanical Club 19:1—57. . 1943. Campanulaceae (Lobelioideae). North American Flora 32A:1—134. MISHLER, B. D. AND M. J. DONOGHUE. 1982. Species concepts: a case for pluralism. Systematic Zoology 31:491—S03. AND E. THERIOT. 2000. The phylogenetic species concept sensu Mishler and Theriot: mono- phyly, apomorphy, and phylogenetic species con- cepts. Pp. 44-54 in Q. D. Wheeler and R. Meier (eds.), Species concepts and _ phylogenetic theory: a debate. Columbia University Press, New York, NY. OLMSTEAD, R. G. 1995. Species concepts and plesio- morphic species. Systematic Botany 20:623—630. OrR, E. L. AND W. N. ORR. 1996. Geology of the Pacific Northwest. The McGraw-Hill Companies, Inc., New York, NY. Peck, M. E. 1934. New Oregon plants. Proceedings of the Biological Society of Washington 47:185—188. Systematic 36 MADRONO the Biological Society of Washington 50:93—94. . 1941. A manual of the higher plants of Oregon. Binford & Mort, Publishers, Portland, OR. RIESEBERG, L. H. AND L. BROUILLET. 1994. Are many plant species paraphyletic? Taxon 43:21-—32. SCHULTHEIS, L. M. 2001. Systematics of Downingia (Campanulaceae) based on molecular sequence data: implications for floral and chromosome evolution. Systematic Botany 26:603—621. SLATKIN, M. 1989. Detecting small amounts of gene flow from phylogenies of alleles. Genetics 121:609— 612. SOLTIS, D. E.. M. A. GITZENDANNER, D. D. STRENGE, AND P. S. SOLTIS. 1997. Chloroplast DNA intraspecific phylogeography of plants from the Pacific Northwest of North America. Plant Sys- tematics and Evolution 206:353—373. SWOFFORD, D. L. 1993. PAUP: phylogenetic analysis using parsimony, version 3.1. Computer program . 1937. New plants from Oregon. Proceedings of [Vol. 57 distributed by the Illinois Natural History Survey, Champaign, IL. . 2001. PAUP*: phylogenetic analysis using parsimony. Version 4. Sunderland: Sinauer Asso- ciates, Inc. Publishers, Sunderland, MA. THORSON, T. D., S. A. BRYCE, D. A. LAMMERS, A. J. Woops, J. M. OMERNIK, J. KAGAN, D. E. PATER, AND J. A. COMSTOCK. 2003. Ecoregions of Oregon (color poster with map, descriptive text, summary tables, and photographs): map scale 1:1,500,000, U.S. Geological Survey, Reston, VA. WEILER, J. H., JR. 1962. A biosystematic study of the genus Downingia. Ph.D. dissertation. University of California, Berkeley, CA. WHITTAKER, R. H. 1961. Vegetation history of the Pacific Coast states and the “central” significance of the Klamath Region. Madrono 16:5—23. Woop, C. W., JR. 1961. A study of hybridization in Downingia (Campanulaceae). Journal of the Ar- nold Arboretum 2:219-262. SCHULTHEIS: CRYPTIC SPECIES IN DOWNINGIA YINA TT (ON) eddiany Jo AQ TI ¢ “OD JaeMIRATD ‘¢zOZ IIUDISUO) x. (ON) AMO pudad syvyJ JO aspq 0D IQUUOG “6F UOSUDIAT WY SAY] x (ON) Pasual Jo § MU gy VOD yemousg ‘QQ0E 421 « OHVdI (ON) AaeA MeYysudOH 20D Away ‘corZ/ ADDAL (ON) LOT AM Jo | ‘SUT JO SHUI] § OD OUTDopuay ‘9STO9 APIOM x COUN MOUS,, SB TL6] CS T6CCAV J9}SOJ UI PSD) OL = u ‘IPL6CCAV “OPI6TTAV (AVG) Us ulqed Al[>y Ieou ‘ureyUNO| MOUS 10D oyeT P-LT-OL 49ISO] (T961 JOPIOM UL Pad) OL = u (ON) ded oiqnd Yae1D swag jo N Tw og ‘ureyunopy MOUS “OD WPT 8ET09 49/19 x (ON) AaTeA seqeueT -oD ipjoquny *zo¢e7 ov4] * VINYOATTVO SUuDSaJa DISUIUMOG (OM) Ilo AYoY JO MS tu ¢ ‘uoNrIg 1osuey UBSTPSd “OD YWUePY CCP 49/10 6rl6ccCAV ‘LEL6CCAV “9ET6TTAV (SdHL) OUDY JO AM TU ¢ “99 IY VOD yeMEPy ‘¢6-zEz SPOUTIAUOS & (ON) YHE.O JON MW L'€ “661 AMY VOD ourydasoe ‘6 7¢79 42/1OM (ON) OPTEM JO SS PJO JO MA MU p'9 “OD ourydasor ‘eze7o ged bis ae IS l6ccAV (Sdaf) uonouns “E€L6CCAV “TELOCTAV SOAR) JO § “YeuD Apeoy puv ysnoy ye yieg opiske [eoruvjog “OD surydasor “66-C9¢ S1ayIINYIS » (Sdaf) ‘SPY puepapy pure yoy yqvy Jo 19U10d FN “psojpoy JON “OD UOSAICLF “66-CEEF SIOYIINYIS’ » 8889L1TIV “E8EC9TAV “T8ECOIAV (Sddf) Aaa JON “S6E YM UONoUN! Jo N ru | ‘97 AMP UO VOD Asse “C6-OFT SAYIN) » NOOAAO 8L89LITV 0069LIAV (Sddf) eurdyed 0} Jyo uy Jo N Ww /"0 “EZ-W peo! ‘AaTTeA BLOTS “OD BIIIIS “C6-1O7 SIOYIINYIS) » (Sddf) €C-W PY pur ourdjed 07 peo Jo 19UI109 MS ‘Aap BLOTS “OD RBIISIS “C6-COT SIOYIINYI » (SdAf) eur] AJuNOS NoArysig-viseYys Jo § IW ZE “68 TY VOD RIseYS “SO-PET SLOYIINYIS 5 (Sddf) SUlssoio proses pure $6 Py 41D 3 “MoYooT Jo N v.09 SOPOW “S6-TET SIOYIINYIS x LLECOITAV “OLEEITAV (SdH) “Id JepaDd Jo § TW €'0 “b6 PU AID UO ‘NoYooT Jo N “0D sopopy “SO-OEE S1ayINYIS (Sddf) 9b VA UO GET TA JON twp] ‘Aqued Jo N 20D sopoypy “u's ANSSOL 'S (Sdaf) 199lo1g dLDIJBO.IPAH Aa|[eA YONI O} oue.ua ye 667 AMP “IOQloqnyy JO MS :OD Uasser’T *C6-FE 7 SlaYyINYIS x. VINYOATTVO ndnppsi9vg visuuMog Joquinu swosowoliy) souonbes yuRgquay Ioyono A ey mn ee (ON) 10 Sdaf 3B ore SUOTdaT]OD pouOlssazqR “AW Ie pousodap are 19}s0-4 Aq suoNsaT[09 pojunowuy ‘ejJep sUIOsOWOIY9 J99]]OD OF pasn susutiseds soy Jaquinu swWOsoWOIYS prlojdey pue SISATBUR Ie[NOV[OU 9Y} UT pasn suauTIOdds OJ UdAIT ome (S)Ioquinu UOIsssd0R addUaNbas yuR_GuaH “YSH9}se ue Aq payeorpul oie siskjeue [eosojoydsow oY} Ul posn susWIs9dg “ALVLS GNV SHIOddS YWAGNGQ GadnNouyH “ACNLS LNAYUNNAD NI Gas SYAHONOA NAWIOddS | XIGNdaddV [Vol. 57 ~ MADRONO eee (ON) puelpoom svou ‘AVA OUSUIBIOCS SOD O[OX ‘OFT 2992L x (ON) AW2D sajoD suoye 1] BIg O—em=> “GHNNILNO, “[ XIGNddd VY 40 4] 6L89LIAV “O9EE9TAV (SdAL) ‘PU 99-1QPID JO JJO “PY Nouleg Jo apis § ‘ayxPT YoTING JO N “OD snvysturys “C6-O87 s1ay1NyoS » VINYOATTVO DULISSIJDUAO DISUIUMOG (Sddf) punorsduie>) pos9LIAV “l6ec9ltVv SoIOT "PVN SMmOpKo|Y SIN JO | MeO “TE VA JO JfO “Pa Ipjoquiny 70D sng *¢6-O¢T slaysnyos OLSOLIAV “6LEC9TAV “8LEC9IAV (Sdaf) ou] AyUNOD NOATYsIC/eISeYS JO § TW TE “68 VA “OD KIseYS “C6-CET SIOYIINYIS' » VINUYOATTVO DuDjUuol DISUIUMOG (Sdaf) ‘Py PuoutoT ysoT O689LIAV ‘P9CEIIAV YUM UOTIasIa}UT JO N snl “CLT VY JO Opis M “MOPKa|] ‘PUOUIOT YIOT 20D ay] “66-Z6T s1ayynyrs (Sdaf) ‘PU VOUT pue ‘97 AMP] ‘py uUosiIng Jo suonounl je ‘MOAIasaY ByOuRUIRD JO YS VOD SPBIDAPTRD “66-62 SIAYIINYIS » VINYOATTVO vyppidsnd visumuMog (Sdaf) (T66T F9pneg 998) O] SCHULTHEIS: CRYPTIC SPECIES IN DOWNINGIA YINA L6OEL9IAV “96EL9OTAV OIS oyeT vovurvdAny 18 Jopneg ‘q Aq pdydd[[OD pads WIOIJ UMOID 10D O8dIq US ‘66-ZO7 SlayIjNYyIS EL89LIAV C9ECOlIAV (Sdaf) ‘Pu ABA sdog-softyD pue py ssory AoyeA odog Jo uonounr oD eden “¢6-c6/ s1ayINYS » VINdOATTV OD AOJOIUOD VISUIUMOG LO89LIAV ‘OPEC9IAV “6EfC9lAV (Sdaf) 9-V Te UO ‘NT PIM JO AN “ACT Soe “OD ePuURY IL “C6-OOT SIOYIINYIS' x VINYOATTVO DINUAOIIC DISUIUMOG Jaquinu swosowolyy souanbes yuegquay Iayono A, “GHNNILNO,) ([T XIGNddd VY 2010] MADRONO, Vol. 57, No. 1, pp. 42—53, 2010 POLLINATION AND REPRODUCTION IN NATURAL AND MITIGATION POPULATIONS OF THE MANY-STEMMED DUDLEYA, DUDLEYA MULTICAULIS (CRASSULACEAE) C. EUGENE JONES, FRANCES M. SHROPSHIRE, ROBERT L. ALLEN, AND YOUSSEF C. ATALLAH Department of Biological Science, California State University, Fullerton, CA 92834-6850 ceyones@fullerton.edu ABSTRACT We investigated the reproductive biology of the rare and endangered plant, Dudleya multicaulis at five separate sites, three natural and two mitigation sites. We employed dawn to dusk observations to determine the spectrum of pollinators visiting D. multicaulis, took pollen samples from visitors to determine floral constancy, sampled nectar to determine volume produced per flower, examined the number of flowers per inflorescence, the number of those flowers that produced seed, and total seed set to determine reproductive output, completed seed germination tests to determine viability, and transplanted germinated seedlings from Petri dishes to soil to determine how well seedlings survive transplanting. Dudleya multicaulis was visited by flower beetles, native and European honey bees, flies, and a variety of other insects. Nectar production per flower averaged 0.12 ul. Bees averaged 99% floral constancy to D. multicaulis. Reproductive output measured by flower production and fruit/seed set were not significantly different among sites. Among all populations, the average fruit set ranged from 86.9 to 94.4%. The large fruit set coupled with the diversity of floral visitors suggests that D. multicaaulis is not pollinator limited. Data suggest that D. multicaulis is capable of self-pollination in absence of vectors. Seed germination and transplanted seedling survival did not differ significantly among sites. Results suggest that sowing seed may be better for plant establishment rather than transplanting when mitigation is necessitated. Key Words: Auto-fertility, Dudleya multicaulis, pollination, reproductive output, seedling survival, transplanted. Information on the reproductive biology of rare plants can provide some assistance in understanding why some plants are rare and others are common (Kearns et al. 1998). Of special importance are cases where rare plants, which are to be extirpated as a result of development, are physically transplanted to new sites or seeds from existing populations are sown in new locations intended to serve as mitigation sites. Data relative to the reproductive biology of such species should play a significant role in decision-making regarding the management, sal- vaging, and moving of such rare plants as part of a mitigation process. Information of this type may indeed prove critical to the success or failure of the establishment of salvaged plants or seeds in mitigation areas. Dudleya multicaulis (Rose) Moran (Crassula- ceae), the many-stemmed Dudleya, is recognized as a rare and endangered plant in California and elsewhere (List 1B.2) by the California Native Plant Society (CNPS 2005). As part of the mitigation process necessitated by the Final Project Environmental Impact Report for the Santiago Hills If Planned Community and certified by the City of Orange in 2000 (Hom- righausen unpublished), this sensitive species was transplanted or seeded to new areas as part of a pilot study for future mitigation. Mitigation sites were selected based on “their similarity to the existing population sites in terms of vegetation composition and cover, apparent soil type, and depth, slope, and aspect”’ (Homrighausen unpub- lished). A patchily-distributed geophyte, D. mutlticaulis is typically associated with the coastal sage scrub plant community of southern California (Dodero 1995; Marchant et al. 1998). Little is known about its reproductive biology (RCIP 2003), although several possible bee, fly and flower beetle pollinators are projected to be involved (Dodero 1995). To provide information relative to the repro- ductive biology of this rare species, we observed the developmental sequence of flowering and investigated the pollination biology of this species during the peak flowering period in May of 2005 at the Santiago Hills site (Jones, Shropshire, and Allen unpublished), which is within the Santiago Hills Hf! Planned Community and the East Orange development projects (Homrighausen unpub- lished). Subsequently, we examined the repro- ductive output, seed germination, and seedling | survival and reproductive effort for natural and mitigation plant material. Specifically, we ad- dressed the following questions: |) What visits D. multicaulis diurnally? 2) Might the plant self | without a vector? If so, what is the mechanism of | 2010] this selfing? 3) How constant are the visitors to D. multicaulis? 4) How much nectar is produced per flower in D. multicaulis? 5) What is the repro- ductive output in the natural and mitigation populations of D. multicaulis? 6) How viable are the seeds produced by plants in the natural versus the mitigation populations? 7) Do transplanted natural and mitigation population seedlings survive and reproduce during the first year? MATERIALS AND METHODS Dudleya multicaulis is a member of the succulent family Crassulaceae (the stonecrops). Detailed descriptions of the family, genus, and this specific species can be found on line (http:// ucjeps.berkeley.edu/cgi-bin/get_JM_treatment. pl?3284,3295,3324). Dudleya multicaulis is an herbaceous perennial that comes up each year from over wintering underground corm-like tuberous caudices. Dudleya multicaulis occurs on heavy clay and rocky soils in barren areas among coastal sage scrub and chaparral com- munities (Munz 1974) and was originally found from coastal Los Angeles County south to Camp Pendleton and inland to Riverside and San Bernardino Counties, in California. In D. multicaulis, the flowering stalk is often multiple-branched and bears lemon yellow flow- ers. According to Munz (1974), the many- stemmed Dudleya flowers in May—June; howev- er, both BLM (2005) and Marchant et al. (1998) give the blooming season as April—July, which is more consistent with our observations. Nascent inflorescences of D. multicaulis start to appear in March and April, each beginning as a_ pink- stemmed stalk produced near the center of the plant. Each primary stalk usually forked at least once, producing two secondary stalks. Some secondary stalks fork again, producing tertiary stalks. A single flower appears at the first fork and is the first to open. From there, blooming continues up the stalk in succession (Fig. 1). Flower “‘1”’ opens first, reaches peak bloom, if pollinated tending to develop a reddish tinge on the petals, and begins to form fruit. Relative ages of each inflorescence can be estimated by examining the condition of their flowers. Young inflorescences have their lowest flowers open and none in fruit. Intermediate-aged stalks have open flowers mid-way along the inflorescence branches with the lowest in fruit. Older inflorescences are in flower at the tips (‘‘n’’ flowers) and in fruit below. In late summer or fall, follicles dehisce and fall off of the plant. Seeds are about 0.8 mm long. Caesares and Koopowitz (unpublished) report that the average flower, with its five follicles, produces about 12 seeds, of which approximately 52% were viable when germinated under nursery conditions. All aboveground parts senesce in JONES ET AL.: POLLINATION BIOLOGY OF DUDLEYA MULTICAULIS 43 First flower to open FiG. 1. Blooming sequence. Generalized flowering pattern of Dudleya multicaulis. 1 = first flower to open, 2 = second flowers to open, 3 = third flowers to open, 4 = fourth flowers to open, 5 = fifth flowers to open, n = last flowers to open. Flowers 6 though (n—1) were intentionally left unlabeled. summer, leaving only the dried inflorescence in place. The small stature and growth habit of this species make it difficult to see and, as a result, it is easily overlooked by botanical surveyors. Study Sites Primary study site. The site where the pollina- tion studies were conducted from 13 to 15 May 2005 is located in the Santiago Hills just east, south east of Irvine Regional Park, Orange Co., California. Here a rather large population of D. multicaulis occurs on a northwestern facing slope near an abandoned stretch of the old Santiago Canyon Road. A series of four D. multicaulis subpopulations were initially delimited for study. Beginning lower on the slope and proceeding to the top of the hill, the four subpopulations were identified as follows: subpopulation C — the control site that was used for the collection of D. multicaulis floral visitors (insects captured at this site were used for identification and for pollen constancy studies). Subpopulations |, 2 and 3a were dedicated for use in the dawn-to-dusk pollinator observation studies. On the second day of the dawn-to-dusk studies, subpopulation 3a was replaced by a nearby subpopulation (3b), which contained a larger number of D. multi- caulis plants in flower. Site 3b was located at the top of the hill adjacent to the fence line separating 44 MADRONO TABLE 1. GPS COORDINATES FOR THE SUBPOPULA- TIONS STUDIED AT SANTIAGO HILLS, ORANGE COUNTY, CALIFORNIA. Subpopulation Latitude Longitude C 33°47.242'N 117°44.802’W l 33°47.240'N 117°44.792'W 2 33°47.226'N 117°44.782'’W 3a 33 47.217 N 117°44.778'’W 3b 33 4/2135'N 117°44.772'W the overall study site from a Toll Road (SR-261). GPS coordinates for these sites are presented in Table 1. Ancillary study sites. The mitigation sites in Weir Canyon (GPS coordinates 33°48.784'N, 117°44.767'W) and Limestone Canyon (GPS coordinates 33°43.522’N, 117°39.721'W) were examined on 15 May 2005 and 27 May 2005, respectively, to determine how many of the mitigation plants were flowering. These plants were counted and later (on 9 July 2005 at Weir Canyon and on 14 July 2005 at Limestone Canyon) examined to determine how many of the flowers on these plants produced one or more follicles and whether these fruits contained one or more fully formed seeds. Fully formed seeds were assumed to be viable and were later utilized in the germination studies. Pollination Pollinators/visitors—Dawn-to-dusk observa- tions. To determine pollinator behavior, diversity, and the relative importance of each of the major pollinator groups, a series of dawn-to-dusk surveys was conducted during the peak D. multicaulis bloom at the Santiago Hills study site from 13 through 15 May 2005. Peak bloom is herein defined as the time when greater than 50% of the plants were in flower. Pollinators visiting D. multicaulis plants were observed during at least 10 min out of each hour beginning on the hour after sun up and continuing throughout the day until 50 min after the hour before sun down. This survey involved three consecutive days of observation. At the study site, each of the three subpopu- lations (1, 2, and 3) was selected on the basis of the ease with which field assistants could observe a sizeable number of plants. Two observers were employed to conduct simultaneous observations at each subpopulation during the three days of study. Each person observed and recorded the visitors to D. multicaulis plants and the number of flowers each visited in the initial subpopulation (e.g., 1) during the first 10 min of each hour. The observers then had 10 min to move to the second subpopulation (2) where visitors and visits were observed and recorded from 20 min after the [Vol. 57 hour until half past the hour. Finally, these same observers rotated to the third subpopulation (3) and repeated the process from 40 min after the hour until 50 min after the hour. Each day the starting subpopulation was rotated so that, during the three-day period, each of the three study plots or subpopulations was the first to be sampled at the start of the observations for that day. A visitor was defined as any organism that actually landed on and came into contact with the anther(s) and/or the stigma(s) of the flower. Visits were defined as the number of times that a particular visitor landed on one or more flowers of D. multicaulis and probed that flower for nectar and/or pollen. Data were subsequently analyzed in terms of number of visitors and visits. Pollinator/visitor collection and _ identification. Representative samples of visitors were collected from 13 to 15 May between 9:00 and 18:00 at subpopulation C. Organisms seen visiting three or more flowers were captured in an insect net or by using a blowing aspirator and placed in killing jars charged with ethyl acetate. Each specimen was returned to the laboratory, pinned and prepared for identification and pollen sampling. Hymenopteran samples were taken to Roy Snelling at the Natural History Museum of Los Angeles County for identification. All other visitors were identified, at least to order, by the investigators. Pollen analysis. Each captured visitor was examined under a Bausch and Lomb dissecting microscope to determine if pollen was present on the visitor and, if so, where it was located. A 3 cm piece of double-sided Scotch® tape with one end cut to a point and that end was used to pick up any available pollen from the visitors under the dissecting scope. Once the pollen had been transferred from the visitor to the double-sided tape, the tape was placed on a 7.62 cm X 2.54cm < 1 mm glass microscope slide. One or two drops of cotton blue (1% aniline blue in lactophenol) were added to stain the pollen grains and the slide allowed to sit for at least 24 hrs for the stain to take effect. Slides were then viewed under a Leitz compound microscope and any pollen grains present were identified as either D. multicaulis pollen (no other species of Dudleya were in flower in the local area) or foreign pollen (using pollen reference slides). The number of plant species and pollen grains found on each individual visitor was used to determine which pollinators carried the pollen of D. multicaulis and how constant they were to D. multicaulis. A minimum of 100 pollen grains were examined for each specimen, except in the case of two of the flower beetles, where only 10 and 23 total pollen grains were located and indentified. Pollinator constancy was defined on a percentage basis. The higher the percentage 2010] of one pollen species in a sample, the more specific that pollinator was to that particular plant species. A pollinator was considered to be “constant” when that pollinator visited a given species at least 95% of the time during a single foraging flight. Nectar samples. Near subpopulation 3b, five plants that were in bud but had no open flowers were entirely covered with light colored knee high nylon stockings on 13 May 2005. These stockings served as pollinator exclusion bags and were secured with a twist-tie to create a seal between the bag and the stem of the D. mut/ticaulis plant to ensure that no pollinators visited the flowers. After approximately five days, these five plants were brought back to the laboratory where the presence of nectar was subsequently sampled using | ul Drummond “microcaps” disposable micro-pipettes (Drummond Scientific Company, Broomall, PA). On 18 May 2005, at least 3 newly- opened flowers on each of the five plants were probed with the micro-pipettes to determine if nectar was being produced and, if so, how much was being secreted per flower. Reproduction Reproductive output. Between 9 July and 14 July 2005, plants at the Santiago Hills study site, as well as the Weir Canyon and Limestone Canyon mitigation sites, were examined to determine the number of flowers produced per inflorescence and how many of those flowers contained one or more follicles. This was done to determine if there were differences in flower production per inflorescence among the sampled sites and to determine the percentage of fruit set per flowers produced. Also, while examining the fruit, mature flowers with fruit were harvested from the control (C) subpopulation at the Santiago Hills study site, from the natural population at the Weir Canyon site, and from the mitigation plants at Weir and Limestone Canyons. Two sub-samples were examined at the Weir Canyon natural population. The first sample (identified as population 1) was taken from the lower portion of the natural population on the north west facing slope and the second sample (identified as population 2) was removed from plants that co-occurred with the mitigation plants at the top of the same natural population. A total of 10 or 11 flowers were harvested at each site, one each from 10 or 11 different plants, except for the Weir and Limestone Canyon mitigation sites where more than one flower was harvested per plant to achieve a sample of 10 flowers. The number of fully formed seeds per fruit and per flower was determined. Twenty-five inflorescences, one each from separate plants, were sampled from each of the JONES ET AL.: POLLINATION BIOLOGY OF DUDLEYA MULTICAULIS 45 subpopulations (C, 1, 2, 3a, and 3b) studied at the Santiago Hills site. At the Weir Canyon site, one inflorescence each from 50 naturally occurring separate plants found on the north west facing slope were examined and one inflorescence each from four of the eight plants that had been spotted and marked with flags on 15 May 2005 were examined; the other four marked plants could not be located. One inflorescence each from seven of the eight plants that had been identified and marked with flags on 21 May 2005 at the Limestone Canyon Site were also examined. The eighth flagged plant at this location could not be located. The seeds harvested from these samples were then submitted to germination tests. Seed germination tests. A total of 208 seeds from the Santiago Hills site, 101 seeds from the Limestone Canyon site, 137 seeds from the Weir Canyon natural occurring plants, and 12 seeds from the Weir Canyon mitigation site, were harvested from the flowers produced by the plants in each of these four sites. Of these, a subsample of 100 seeds (except for the Weir Canyon mitigation site where all seeds recovered were utilized) were placed on moistened 38 Ib. 8.9 cm circles of regular seed germination paper (Anchor Paper Company, St. Paul, MN) in 100 x 15 mm Fisherbrand disposable sterile petri dishes (Fisher Scientific, Los Angeles Office, Tustin, CA). A total of 18 petri dishes were utilized as follows: five petri dishes with 20 seeds per dish or 100 seeds per each were prepared for the Santiago Hills, Limestone Canyon, and the Weir Canyon natural sites. Since there were so few inflorescences produced by the mitigation planting at the Weir Canyon site, there were fewer seeds available so only 3 petri dishes with 4 seeds per each or a total of 12 seeds were prepared for the germination tests. The petri dishes were watered with 5 ml of deionized water and placed in individual Ziploc® one quart storage bags (A product of S. C. Johnson & Sons, Inc., Racine, WI), labeled with an identi- fication code, and then randomly placed in one of two Percival Model E-30B growth chambers (Percival Scientific, Inc., Perry, IA). Each growth chamber was then set on 11 hr of daylight with 15°C daytime temperature and 10°C nighttime temperature. Germination was monitored daily from 3 October 2005 through 28 November 2005. Transplanted seedling survival tests. A sample of the germinated seedlings from each site was transplanted into 5.5 * 5.5 KX 8.5 cm (W X D X H) black plastic pots filled with potting soil on 5 January 2006 and followed through the growing season of 2006. The potting soil was a mix of an organic fraction (50%) that included peat moss (6 parts by volume) and forest humus (9 parts by volume) and of an inorganic fraction (50%) that included washed plaster sand (6 parts by volume) A6 MADRONO [Vol. 57 Total Visits - Dawn to Dusk 7 '¢) come oo co — ———— 60 | = -« Flower Beetles = = Honeybees | = @ Other Bees | 50 = Flies | = 7 All Others | s = | “ — 2 40 = | 8 = = | he = — F = = | £ 30 = = a = 4 J = = = | z = = = = 20 | = = = = ! > Zt = = ~ = = y : = /, = tp = Mae Ge: EYES = f= A= B= See — — = = —- nae 0 — = = VMa= Vat a 14:00 15:00 16:00 17:00 18:00 Time FIG. 2. Total visits by all visitors by time of day for all study plots (1, 2, 3a, and 3b) for 13—15 May 2005 combined. and pumice (9 parts by volume). Time-released fertilizer was added at the rate of 40z/10 gallons soil mix and dolomite (Ca and MgCo3) at 5oz/ 10 gallons of soil mix. The time released fertilizer used was Sierrablen 18N:7P:10K + Fe. The potted plants were placed outdoors and watered daily or as needed following rains. They were monitored for survival and reproduction at the end of the growing season on 23 June 2006. Statistical Analyses Sites were compared using one-way analysis of variance (ANOVA) or a Kruskal-Wallis Rank Sum Test when required for flowers produced per plant, fruit-set per flower, seeds per flower, seed germination, and survival of transplanted seed- lings derived from the germination tests. Tests were done using Excel. RESULTS Pollination Pollinators/visitors. Hymenopterans included the European honey bees (Apidae, Apis mellifera L.), two bee species in the family Halictidae (Halictus tripartitus Ckll. and Lasioglossum [Dia- lictus|] sp.), one bee species in the family Mega- chilidae (Hoplitis grinnelli [Ckll]), and possibly two separate species of ants (all Order Hyme- noptera). Other visitors included soft-winged flower beetles (Melyridae, Dasytinae, possibly Lystrus sp.) and weevils (both Order Coleoptera), several flies including those in the family Syrphi- dae as well as other families (Order Diptera), with only a few individuals each of true bugs (Order Hemiptera), leafhoppers (Order Homoptera), and flower mites (Order Acari). Of these visitors, the most frequent and/or most important (judg- ing by behavior within the flowers that indicated a high probability of successful pollination) included the flower beetles, honey bees, and bees in the families Halictidae and Megachilidae. Dawn-to-dusk observations. The results of the total dawn-to-dusk observations are summarized on a diurnal basis (Fig. 2). It is interesting to note that flower beetles were found visiting the flowers during the entire daily observational period. Non-native European honey bees tended to be more common in the afternoon hours, whereas the native solitary bees tended to frequent the flowers earlier in the day. Flies seemed to visit the 2010] flowers early in the morning. All other visitors appeared to show a bimodal distribution arriving in the morning and then again in the afternoon. The frequency of visits by the various groups of potential pollinators (visitors) for all study plots and all three days of observation was combined. There was considerable variation among the three subpopulations regarding the frequency of visits by the various groups, but the total frequency distribution provides a good representation of the overall visits to Dudleya multicaulis. Of the total visits to D. multicaulis, flower beetles accounted for 31% of the visits and they represented 56% of the visitors. European honey bees and other bees, in contrast, made 27% and 19% of the visits respectively, but were repre- sented by only 8% and 8% of all visitors, indicating that the bees were typically visiting more than one flower per foraging bout on D. multicaulis. Flies and all others accounted for 9% and 14% of all visits to D. multicaulis, but had 10% and 18% of all visitors respectively. As in the case of the visits at each of the subpopulation study sites, there was considerable variation in the relative frequency of the various groups of visitors at each study site. Considering each subpopulation individually, 78% of the visits observed in subpopulation | were from flower beetles, whereas none of the other groups contributed more than 8% of the visits. Further, 84% of all D. multicaulis floral visitors were flower beetles. Visits to flowers of D. multicaulis at subpop- ulation 2 were more equally distributed among the various pollinating groups with 32% of all visits being by other bees, 28% by flower beetles, and 36% by all others. Flies and European honey bees accounted for only 3% and 1% of the visits respectively. In terms of visitors, flower beetles represented 32% of the visitors, whereas other bees and all others accounted for 31% and 31% of the visitors respectively. Flies and European honey bees accounted for 3% of the visitors each. Since we used two separate plots for subpop- ulation 3, two distinct patterns were observed for the visits and visitors to subpopulations 3a and 3b. Subpopulation 3a, which was utilized only on 13 May 2005, showed 44% of all visits were by the all others group, 31% by flower beetles, 17% by flies, 6% by other bees and 2% by European honey bees. An examination of the visitors for the Same subpopulation shows that 40% of the visitors were in the all others group, 36% were flower beetles, 15% were flies, 6% were other bees, and 3% were European honey bees. In contrast, visits to D. multicaulis flowers in subpopulation 3b, which was observed 14—-15 May 2005 showed that 39% of the visits were by European honey bees, 32% by other bees, 12% by JONES ET AL.: POLLINATION BIOLOGY OF DUDLEYA MULTICAULIS 47 flower beetles, 11% by flies, and only 6% by all others. This represents quite a contrast with the visits observed at subpopulation 3a and may simply reflect the consequence of a much larger floral display present at subpopulation 3b in comparison to 3a. Data for visitors of the various groups at this subpopulation (3b) show that members of some of the groups made multiple visits per foraging bout (e.g., honey bees and other bees with only 14% and 25% of the visitors), whereas individuals of other groups of visitors usually visited only a single flower per foraging bout. Pollen analysis. Pollen taken from the sampled visitors was identified. The three bee species (European honey bee, n = 6 and halictid bee species, n = 5, exhibited an average floral constancy of 98.7 and 99.7% with standard deviations of 2.61 and 0.67 respectively. The same was also true for the soft-winged flower beetle (Melyridae, n = 4), which had an average constancy of 74.8, but the standard deviation was much higher at 26.6. Nectar analysis. Dudleya multicaulis plants produced an average of 0.12 ul per flower. Average nectar production per the five sampled plants varied from 0.08 ul to 0.17 ul per flower. Nectar production per flower was minimal. Reproduction Flower and fruit production. Data regarding the number of flowers produced per inflorescence and the percentage fruit-set for sampled plants at the various study sites/subpopulations are presented in Table 2. Although there were no significant differences among sites (Fg9 = 0.94, P > 0.52), the subpopulations at Santiago Hills generally produced a few more flowers per inflorescence than either of the mitigation populations at Weir Canyon or Limestone Canyon. Of the latter two, the Weir Canyon mitigation site, which was located within approx- imately 30 meters of a natural population of D. multicaulis, produced a few more flowers per inflorescence than those at Limestone Canyon, a population which was separated from a natural population of D. multicaulis by well over 2 km (Table 2). There were also no significant differences for average fruit-set among sites (Kruskal-Wallis Rank Sum Test value = 4.34, P > 0.82). However, the average fruit-set was always greater than 85%. The range in fruit-set varied among and within the D. multicaulis populations from a low of 60% in the Limestone Canyon mitigation population to 100%, a high value that was found in all studied populations including the Lime- stone Canyon mitigation population. AS MADRONO TABLE 2. [Vol. 57 NUMBER OF FLOWERS PRODUCED PER INFLORESCENCE AND PERCENTAGE FRUIT-SET FOR PLANTS IN THE VARIOUS STUDY POPULATIONS AND SUBPOPULATIONS. n = the number of inflorescences sampled. Ave. fl. = average flower number, SD fl = standard deviation for that average, and R fl = range of number of flowers produced per inflorescence. Ave. % fr. = average percentage fruit set, SD fr = standard deviation for that average, and R fr = range of fruit set per flowers produced on inflorescences. Study site/subpopulation n Ave. fl. Santiago Hills — subpop. C 25 22 Santiago Hills subpop. 1 25 14.6 Santiago Hills subpop. 2 25 a2 Santiago Hills subpop. 3a pee 20.0 Santiago Hills subpop. 3b 25 34.5 Weir Canyon natural pop. | 50 jis Bs Weir Canyon natural pop. 2 21 11.2 Weir Canyon mitigation plants 4 55 Limestone Canyon mitigation plants a 5.6 Seed production. No significant differences in average seed production per site was found among populations (F;.6 = 3.65, P > 0.10). The average seed production per flower varied by population from a low of only 0.3 in the Weir Canyon mitigation population to a high of 5.4 in the Santiago Hills subpopulation C. The Santiago Hills population produced the highest number of fully-formed seeds per flower, (each flower having 5 separate fruits [follicles]), followed by the natural population at Weir Canyon. The mitigation plants at the Limestone Canyon site produced the next highest number of seeds per flower, whereas the Weir Canyon site produced the fewest number of seeds per flower. In fact, only one plant of the four plants sampled from this latter site contained any seeds. Seed germination. The percentage of seeds germinating by site was not significantly different from one another (Kruskal-Wallis Rank Sum Test value = 0.17, P > 0.98). An examination by site showed that at least 25% of the seeds had germinated after the first 48 hr of the tests. Percent germination at each site was quite good with all sites ranging from 62% at the Limestone Canyon Mitigation Site, to 65% at the Santiago Hill Subpopulation C, to 83.3% at the Weir Canyon Mitigation Site, to a high of 85% at the Weir Canyon Natural Population. It is interesting to note that the two Weir Canyon sites had the highest germination percentages. This may be important to the ultimate survival of the popu- lation at the Weir Canyon mitigation site since so few seeds were produced by the meager number of surviving mitigation plants at that site. Transplanted seedling survival and reproduction. Transplanted seedling survival to successful reproduction did not differ significantly by site (F34 = 0.29, P > 0.83). However, of all transplanted seedlings from all the study sites, a minimum of 25% of them survived to flowering and fruit production. The lowest survival was found in the Limestone mitigation site (at 25%) SD fl R fl Ave. % fr SD fr % R fr 5.43 6-31 91.6 7.14 78.6—100 6.33 5—33 94.4 6.20 80.0—100 5.01 6—27 86.9 9.16 64.3—100 7.54 10-37 92.2 5.89 81.1—100 DigA2 11-94 92.5 5.99 80.0—100 Dat 2-35 89.7 11.10 60.0—100 G27 2—29 94.2 7.69 71.4100 1.71 527 923 9.0 83.3—100 1.90 2-8 87.0 14-7 60.0—100 and the highest was in the Weir Canyon mitigation plants (37.5%). Conversely, between 62.5% and 75% of the transplanted seedlings died prior to maturity, indicating a relatively minimal transplantation survival rate even under the nearly ideal conditions used during this study. DISCUSSION Pollination by biotic agents is a mutualism that has the potential to control important aspects of plant reproduction and can play a critical role in the survival and management of rare species (Schemske et al. 1994; Kearns and Inouye 1997; Bernardello et al. 1999; Kaye 1999; Timmerman- Erskine and Boyd 1999; Spira 2001). Therefore, a knowledge of the pollination biology of any rare species takes on greater importance given the potential effect of such interactions can have on the continued existence of the rare species. Pollinator Activity and Floral Constancy Observations of pollinator activity were only made during the peak time of flowering. Future studies should examine pollinator activity during | early, mid- and late flowering periods to deter- mine the total spectrum of visitors (potential pollinators) and how it may or may not vary from the beginning to the end of the blooming | period. The observations of pollinators within the current time frame revealed that the primary pollinators as judged by their behavior at the | flowers (which included contacting the anthers | and/or stigmas during a floral visit) were European honey bees and bees in the families Halictidae and Megachilidae, although flower | beetles were usually the most abundant visitor at most of the plots. Six specimens of flower beetles | were examined to determine if they carried | Dudleya multicaulis pollen and this pollen of D. multicaulis was found on four of those individ- uals. Given the observed behavior of flower beetles within the flowers, the most likely role 2010] they play in the pollination process of D. multi- caulis is in selfing within a flower. Our data support the suggestion that D. multicaulis has adopted a generalist pollination strategy (see Waser et al. 1996; Gomez and Zamora 1999, for a more detailed overview of this strategy). Plants living in fluctuating envi- ronments such as the southern California Med- iterranean climate have to deal often with substantial annual variation in rainfall. Such variability in rainfall in seasonally dry environ- ments can have a substantial effect on the number of plants that emerge from dormancy, grow to maturity, successfully flower and set fruit (Beat- ley 1974). The generalist pollination strategy then provides a mechanism to ensure some successful reproduction even in years in which plant population levels, flowering resources, and pos- sibly pollinator numbers and diversity are re- duced by lack of rainfall (Waser et al. 1996; Aigner 2001, 2003, 2005; Gomez and Zamora 2006). One of the potential consequences of a reduction in the diversity of potential pollinators in dry years is the potential loss of pollinator species that are more likely to effect outcrossing between plants. This occurs because flowers of species whose population levels fall low enough to reduce the floral rewards to levels that do not meet the energetic needs of the pollinators that are likely to facilitate outcrossing, such as many species of bees (Sih and Baltus 1987; Jennersen and Nilsson 1993; Conner and Neumeier 1995). As a result, selfing is more likely since remaining pollinators are ones (like flower beetles) that require fewer resources to meet their energetic needs. Fruit Set The total number of visitors seen visiting the flowers of D. multicaulis during our study was relatively small. Although fruit set varied among the subpopulations investigated, the differences were not significant. When we harvested inflo- rescences to determine fruit set, we found that nearly every flower had five fully developed follicles, indicating that reproduction did not seem to be pollinator limited. Fruit set was so high (in every case over 85%) that we suspected D. multicaulis might be at least partially self compatible (Sutherland 1986). Sutherland (1986) reviewed the fruit/flower ratios of many plant species and determined that high ratios, certainly those above 33%, were found in plants that were at least partially self-compatible. In view of the small number of visitors observed during this study and the high fruit set, we suspected that D. multicaaulis may not require a pollinator to effect fruit production (hence self fertile, see Harding et al. 1974; Lloyd and Schoen 1992). JONES ET AL.: POLLINATION BIOLOGY OF DUDLEYA MULTICAULIS 49 Nectar Production We found that nectar production per flower was low in comparison to species of D. reported for the subgenus Dudleya (Levin and Mulroy 1985). This reduced nectar production is a characteristic of species that do not to rely on pollinators to effect successful reproduction (Levin and Mulroy 1985). Self Fertility We closely examined the flowers of D. multi- caulis and found some interesting features that may contribute to the high fruit-set in_ this species. Selfing without a vector within a single flower may occur. Each flower has 10 stamens, five alternate and five oppoite the petals. The five pistils begin to fold back into the groove of the V- shaped petals and their styles begin to elongate. During this process, the stigma becomes receptive to pollen deposition. If the receptive stigma does not receive pollen via normal pollinator facilitat- ed transfer, the virgin stigma can pick up pollen as the style elongates and pushes the stigma past the anther on the stamen opposite the petal. If pollen remains on these anthers opposite the petals, selfing without a vector can occur if the pollen remains viable. The observed floral morphology suggests that D. multicaulis may not require a pollinator to effect fruit production and may be able to get pollen into contact with receptive stigmas without the involvement of biotic agents. We emphasize that this is a tentative conclusion and requires further data from additional experimental proce- dures before it can be confirmed. Specifically, bagging or exclusion experiments are required to determine the breeding system of D. multicaulis. If seed is produced by selfing without a vector, then germination and seedling fitness tests should be completed. Further, any seeds produced in the bagging experiments that result from selfing with a vector (transfer of pollen from a flower on a plant to another flower on the same plant) or outcrossing should also be tested for germination and seedling survival. Levin and Mulroy (1985) found that significantly more seed was produced by outcrossing in species in the genus Dudleya subgenus Dudleya than by selfing with or without a vector and that seedlings from outcrossed seeds also survived better than those produced by either mode of selfing. Reproductive Output Reproductive output, as judged by seed production was not significantly different among sites and was reasonable at all sites except the mitigation plants at Weir Canyon. Of the four plants sampled from that group, only one 50 MADRONO produced any fully formed seeds. This finding suggests that selfing does not always occur. This group of plants bears watching and may not survive with such low reproduction. Average seed production per flower also did not vary signifi- cantly among our study populations (ranging from a low of 0.3 to a high of 5.4 seeds per flower) and were much lower than the approxi- mately 12 seeds produced per flower found by Casares and Koopowitz (unpublished). Seed Germination and Seedling Transplantation Tests were completed on the seeds produced by D. multicaulis to see if they will germinate and result in successful offspring. The seed extracted from plants from each of the four study sites germinated quite well and the percent germina- tion was not significantly different among sites. Germination ranged from 62% for the Limestone Canyon mitigation site to 85% for the Weir Canyon natural site. Our germination results are higher than those found by Caesares and Koopowitz (unpublished) who recorded a germi- nation rate of about 52% under nursery condi- tions. It would appear that seeds produced by all plants demonstrate sufficient viability to ensure successful seed reproduction. Further, when the germinated seedlings from these seeds were transplanted into pots and placed out-of-doors under relatively normal conditions, except for regular watering, between 25% (Limestone Can- yon mitigation site) and 37.5% of the plants (Weir Canyon mitigation site) survived and successfully produced one or more seeds by the end of the first year. It should be noted, however, that between 62.5% and 75% of all transplanted seedlings died during this first year when they were grown under nearly ideal conditions. Trans- planting of seedlings or adult plants to new locations would not seem to be a viable alternative to sowing harvested seed as a mitiga- tion measure for this species. It should be noted that the two mitigation sites were dissimilar in that the Weir Canyon mitigation site was within approximately 30 m of an existing natural population, whereas the Limestone Canyon mitigation site was quite remote from any existing natural population of D. multicaulis (ca. 2 km). Reproductive Strategies Wilken (unpublished) investigated the repro- ductive strategies of D. nesiotica (Moran) Moran, another member of the subgenus Hasseanthus and concluded that it is self-compatible but requires a vector to facilitate reproduction. Levin and Mulroy (1985) studied the pollination biology of several species in the genus Dudleya subgenus Dudleya and found that two of the [Vol. 57 three major groups of species in this subgenus demonstrated a significant degree of self-fertility. They attributed this to unreliable pollinators and/ or environmental unpredictability. By unreliable pollinators, they meant pollinators that varied considerably in abundance both temporarily and spatially (Levin and Mulroy (1985). In our study, pollinator abundance appeared to be minimal. It could be that the past few drought years have had a negative effect on insect populations. It may take a few wet years for insect populations to return to normal. Environmental variability, and thus unpredict- ability of resources and pollinators, has certainly been a factor in the development of southern California ecosystems as rainfall varies consider- ably in both amount and pattern from year to year. Therefore, if self-fertility is found to be a significant mode of reproduction in D. multi- caulis, then it may represent an adaptation that increases overall reproductive success in habitats like the coastal sage scrub community and for species like D. multicaulis (Moeller 2006). How- ever, it again needs to be emphasized that in Dudleya subgenus Dudleya, selfing with a vector and outcrossing both resulted in more seed production and, in the case of outcrossed seed, better fitness of the seedlings (Levin and Mulroy 1985). Similar seed set results were also found for D. nesiotica in that it produced about the same fruit set when manually selfed (22.1 seeds per flower) or when outcrossed (20.3 seeds per flower). However, if emasculated and unpolli- nated, no fruit set occurred (Wilken unpub- lished). Wilken (unpublished) provides no data relative to the possibility that self-fertility can occur within flowers in time, if vector facilitated pollination does not occur before the senescence of the flower. Self pollination is also prevalent in habitats with short growing seasons (Runions and Geber 2000; Mazer et al. 2004). D. multicaulis occupies such a habitat, one characterized by extreme annual variation in rainfall, which tends to favor small flowers (Strauss and Whittall 2006). Small- er flowers like those found in D. multicaulis increase the likelihood of selfing because of the close proximity of the anthers and stigmas (Snell and Aarssen 2005). This association of small flower size and variable water availability has been shown to increase selfing in several annual plants genera (Guerrant 1989). The breeding biology of a rare species 1s a very important issue that requires careful consider- ation by decision makers when movement of plants is required for mitigation purposes. For example, if a rare plant requires no pollinator and still sets abundant seed, and assuming such seed germinates and the progeny survive, then, at least in the short term the sowing of this seed may increase the probability of successful mitigation 2010] in cases where plants must be removed from a site. This appears to be the best option for D. multicaulis. However, selfing can have more long-term consequences that include increased inbreeding depression and increased homozygosity in the interbreeding population and, thus, decreased genetic variation at the colony scale. One of the goals of many conservation programs is to maintain genetic diversity in species that are rare, threatened, or have small population size like Dudleya multicaulis (Frankel and Soule 1981; Simberloff 1988). For this reason, genetic studies of rare plants should be completed whenever possible. Information from these studies can establish much about the species that will assist in its successful management (Ellstrand and Elam 1993). Genetic Structure and Selfing In a previous study of the genetic structure of D. multicaulis by Marchant et al. (1998), they concluded that there is little evidence for signif- icant gene flow between populations and that local populations tended to show heterozygote deficit. They also indicated that reduced genetic variability within populations of D. multicaulis might be a consequence of founder effects and subsequent mating among relatives. We would add that selfing should also be considered. In this regard, Marchant et al. (1998) did note that D. multicaulis can self, but indicated that they had not investigated if selfing in D. multicaulis lowered the fitness of the progeny. Data from Levin and Mulroy’s study (1985) of Dudleya subgenus Dudleya suggest that lowered fitness may indeed be the case. Marchant et al. (1998) additionally state that variation among D. multicaulis populations tended to be significant, further indicating that gene flow by either pollen transport or seed dispersal was limited. How far apart, then, must D. multicaulis populations be for genetic isolation to be significant? That remains to be determined for D. multicaulis, but an interesting recent study by Boose et al. (2005), examined genetic variation in Navarretia leucocephala, and concluded that distances of 1100 to 1800 m were often sufficient to result in significant genetic differentiation between populations. Therefore, for species like D. multicaulis with limited pollen and seed dispersal capabilities, it is quite probable that significant interpopulational variation in genetic structure could occur at these distances or less. However, it may be that selfing is a key component in the survival of this species. A recent paper by Morgan et al. (2005) demon- strated, using models, that plants with population densities that vary annually with environmental conditions (like D. mul/ticaulis) may avoid extinc- JONES ET AL.: POLLINATION BIOLOGY OF DUDLEYA MULTICAULIS 5 tion by increased reliance on autogamy, especial- ly when they are pollinated by generalist pollina- tors (as is the case with D. mu/ticaulis). Further, their models also showed that delayed selfing is always favored. At least selfing without a vector appears to occur in D. mutlticaulis only if pollinator services are not forthcoming since D. multicaulis is protandrous. If our model of selfing without a vector is shown to be a functional mode of reproduction in D. mul/ticaulis, it may mean that newly established mitigation popula- tions may be able to persist without going extinct because of their ability to self without a pollination vector. In fact, they may be able to persist long enough to develop sufficiently large plant populations to attract the generalist polli- nators required to facilitate outcrossing and increase genetic diversity (Jarne and Charles- worth 1993). CONCLUSIONS There is much more research to be done to elucidate the reproductive biology of Dudleya multicaulis to provide the background data required to increase the probability of the successful preservation of this species. However, we suggest that transplantation of plants to new sites may not be as good a mitigation measure as seeding the new sites with seeds derived from those plants. It should be noted that our germination tests were completed under con- trolled conditions suggesting that artificial water- ing following seed inoculation of a new location may be necessary to ensure adequate germination and survival. ACKNOWLEDGMENTS We thank Angela Knips for her assistance through- out the project, Frank Wegscheider for his field assistance, Senta Breden and Katie Levensailor for all the hours they spent observing visitors to Dm, and Kelly Argue for all the hours related to the laboratory studies of germination and the greenhouse study of seedling survival and reproduction following trans- planting. Unless otherwise noted, all photographs and line drawings in this report were done by Robert L. Allen. Special thanks to Roy Snelling of the Natural History Museum of Los Angeles for the hymenopteran identifications. We gratefully acknowledge the contract grant support provided to C. E. J. by the Irvine Company through a subcontract from LSA Associates, Inc., Irvine, California. LITERATURE CITED AIGNER, P. A. 2001. Optimality modeling and fitness trade-offs: when should plants become pollinator specialists? Oikos 95:177—184. 2003. The evolution of specialized floral phenotypes in a heterogeneous pollinaiton envi- ronment: fitness trade-offs in Dudleya (Crassula- ceae) species complex. Ph.D. dissertation. Univer- sity of California, Riverside, CA. 52 MADRONO . 2005. Variation in pollination performance gradients in a Dudleya species complex: can gerneralization promote floral divergence? Func- tional Ecology 19:681—689. BEATLEY, J. C. 1974. Phenological events and their environmental triggers in Mojave Desert ecosys- tems. Ecology 55:856—863. BERNARDELLO, G., G. J. ANDERSON, L. S. PATRICIO, M. A. CLELAND, T. F. STUESSY, AND D. J. CRAWFORD. 1999. Reproductive biology of Lac- toris fernandeziana (Lactoridaceae). American Journal of Botany 86:829—840. BoosgE, D. L., B. R. HARMELING, AND R. T. TUR- COTTE. 2005. Genetic variation in eastern Wash- ington populations of MNavarretia leucocephala (Polemoniaceae): a vernal pool endemic. Madrono 52:99-106. BUREAU OF LAND MANAGEMENT (BLM). 2009. Special status plants. Many-stemmed Dudleya (Dudleya multicaulis), Crassulaceae. Bureau of Land Management, California Office, Sacramento, CA, Website http://www.ca.blm.gov/pa/ssp/plants/ dudleya_multicaulis.html [accessed 21 June 2010]. CALIFORNIA NATIVE PLANT SOCIETY (CNPS). 2010. Inventory of rare and endangered plants of California (Online Edition, v7-10b). California Native Plant Society, Sacramento, CA, Website http://cnps.site.aplus.net/cgi-bin/inv/inventory.cgi [accessed 21 June 2010]. CONNER, J. K. AND R. NEUMEIER. 1995. Effects of black mustard population size on the taxonomic composition of pollinators. Oecologia 104:218—224. DODERO, M. W. 1995. Phylogenetic analysis of Dudleya subgenus Hasseanthus (Crassulaceae) using mor- phological and allozyme data. M.S. thesis. San Diego State University, San Diego, CA. ELLSTRAND, N. C. AND D. R. ELAM. 1993. Population genetic consequences of small population size: implications for plant conservation. Annual Re- view of Ecology and Systematics 24:217—242. FRANKEL, O. H. AND M. E. SOULE. 1981. Conservation and evolution. Cambridge University Press, Cam- bridge, UK. GOMEZ, J. M. AND R. ZAMORA. 1999. Generalization vs. specialization in the pollination system of Hormathophylla_ spinosa (Cruciferae). Ecology 80:796—805. GUERRANT, E. O. 1989. Early maturity, small flowers and autogamy: a developmental connection? Pp. 61—84 in J. H. Bock and Y. B. Linhart (eds.), The evolutionary ecology of plants. Westview Press, Boulder, CO. HARDING, J., C. B. MANKINEN, AND M. H. ELLIOTT. 1974. Genetics of Lupinus VII. Outcrossing, auto- fertility, and variability in natural populations of the Nanus group. Taxon 23:729—738. JARNE, P. AND D. CHARLESWORTH. 1993. The evolu- tion of the selfing rate in functionally hermaphro- dite plants and animals. Annual Review of Ecology and Systematics 24:441—466. JENNERSTEN, O. AND S. G. NILSSON. 1993. Insect flower visitation frequency and seed production in relation to patch size of Viscaria vulgaris (Caryo- phyllaceae). Oikos 68:283—292. KAYE, T. N. 1999. From flowering to dispersal: reproductive ecology of an endemic plant, Astra- galus australis var. olympicus (Fabaceae). American Journal of Botany 86:1248—1256. [Vol. 57 KEARNS, C. A. AND D. W. INOUYE. 1997. Pollinators, flowering plants, and conservation biology. Bio- Science 47:297—307. : , AND N. M. WASER. 1998. Endangered mutualisms: the conservation of plant-pollinator interactions. Annual Review of Ecology and Systematics 29:83—112. LEVIN, G. A. AND T. W. MULROy. 1985. Floral morphology, nectar production, and breeding systems in Dudleya subgenus Dudleya (Crassula- ceae). Transactions of the San Diego Society of Natural History 21:57—70. LLoyb, D. G. AND D. J. SCHOEN. 1992. Self- and cross- fertilization in plants. I. Functional dimensions. International Journal of Plant Science 153: 358-369. MARCHANT, T. A., R. ALARCON, J. A. SIMONSEN, AND H. KoopowitTz. 1998. Population ecology of Dudleya multicaulis (Crassulaceae); a rare narrow endemic. Madrono 45:215—220. MAZER, S. J., H. PAZ, AND M. D. BELL. 2004. Life history, floral development and mating system in Clarkia xantiana (Onagraceae): do floral and whole-plant rates of development evolve indepen- dently? American Journal of Botany 91:2041—2050. MOELLER, D. A. 2006. Geographic structure of pollinator communities, reproductive assurance, and the evolution of self-pollination. Ecology 87: 1510-1522. MorGANn, M. T., W. G. WILSON, AND T. M. KNIGHT. 2005. Plant population dynamics, pollinator forag- ing, and the selection of self-fertilization. American Naturalist 166:169—183. Munz, P. A. 1974. A flora of southern California. University of California Press, Berkeley, CA. RIVERSIDE COUNTY INTEGRATED PROJECT (RCIP). 2003. Western Riverside County multiple species habitat conservation plan (MSHCP), Vol 2. The MSHCP reference document, County of Riverside, CA. RUNIONS, C. J. AND M. A. GEBER. 2000. Evolution of the self-pollinating flower in Clarkia xantiana (Onagraceae). I. Size and development of floral organs. American Journal of Botany 87:1439—1451. SCHEMSKE, D. W., B. C. HUSBAND, M. H. RUCKELS- HAUS, C. GOODWILLIE, I. M. PARKER, AND J. G. BisHop. 1994. Evaluating approaches to the conservation of rare and endangered plants. Ecology 75:584—606. SIH, A. AND M.S. BALTUus. 1987. Patch size, pollinator behavior, and pollinator limitation in catnip. Ecology 68:1679-1690. SIMBERLOFF, D. 1988. The contribution of population and community biology to conservation science. Annual Review of Ecological Systematics 19:473— 512. SNELL, R. AND L. AARSSEN. 2005. Life history traits in selfing versus outcrossing annuals: exploring the ‘time limitation” hypothesis for the fitness benefit of self-pollination. BMC Ecology 5:1—14. SpPiRA, T. P. 2001. Plant-pollinator interactions: a threatened mutualism with implications for the ecology and management of rare plants. Natural Areas Journal 21:78—88. STRAUSS, S. Y. AND J. B. WHITTALL. 2006. Non- pollinator agents of selection on floral traits. Pp. 120-138 in L. D. Harder and S. C. H. Barrett 2010] JONES ET AL.: POLLINATION BIOLOGY OF DUDLEYA MULTICAULIS os) (eds.), Ecology and evolution of flowers. Oxford Clematis socialis (Ranunculaceae). Journal of the University Press, Oxford, UK. Torrey Botanical Society 126:107—116. SUTHERLAND, S. 1986. Patterns of fruit-set: what controls WASER, N. M., L. CHITKA, M. V. PRICE, N. M. fruit-flower ratios in plants? Evolution 40:117—128. WILLIAMS, AND J. OLLERSON. 1996. Generaliza- TIMMERMAN-ERSKINE, M. AND R. S. BOoyp. 1999. tion in pollination systems, and why it matters. Reproductive biology of the endangered plant Ecology 77:1043—1060. MADRONO, Vol. 57, No. 1, pp. 54-63, 2010 A NEW SPECIES OF DISTICHLIS (POACEAE, CHLORIDOIDEAE) FROM BAJA CALIFORNIA, MEXICO HESTER L. BELL Rancho Santa Ana Botanic Garden, 1500 N. College Ave, Claremont, CA 91711 hester.bell@cgu.edu ABSTRACT Based upon a specimen first collected by S. N. Stephenson, a new grass species, Distichlis bajaensis H. L. Bell, is described. Stephenson hypothesized that this specimen was a hybrid between D. Jittoralis and D. spicata. Analyses of sequences of nuclear internal transcribed spacer (ITS) and chloroplast ndhF and trnL—-trnF and an examination of gross morphology, blade and lemma micromorphology, and blade transectional anatomy demonstrate that this grass is a new species that may be sister to the remaining Distichlis. The blades of D. bajaensis are yellow-green; those of D. /ittoralis and D. spicata are blue-green. Distichlis bajaensis can be distinguished from D. Jittoralis by its exserted inflorescences with glumes present and from D. spicata by its short (0.8—-1.5 cm) blades with a bend toward the adaxial side. At and distal to the bend, there are antrorse hairs along the medial vascular bundle. Distichlis bajaensis is known from a single large population growing along alkaline seeps in Arroyo Rosarito in Baja California, Mexico. RESUMEN Se describe como especie nueva de las gramineas a Distichlis bajaensis H.L. Bell, basada en un espécimen colectado por la primera vez por S. N. Stephenson. Stephenson postulo que este espécimen era un hibrido de D. Jittoralis y D. spicata. Los analisis de secuencias de ADN nuclear (ITS) y del cloroplasto (ndhF y trnL-trnF), asi como los estudios de morfologia general, micromorfologia (lema y lamina) y anatomia (hoja), demuestran que esta graminea es una especie nueva y que puede ser hermana a las especies restantes de Distichlis. Las hojas de D. bajanensis son amarillentos verdes pero estas de D. Jittoralis and D. spicata son azulinos verdes. Es posible diferenciar D. bajaensis de D. littoralis por las inflorescencias exsertas con glumas y de D. spicata por las hojas cortas (0.8—1.5 cm) con una curva hacia la cara abaxial. Hay pelos antrorsos a lo largo del nervio central antes del medio. Distichlis bajaensis se conoce de una sola poblacion grande que crece a lo largo de filtrars alcalinas en la localidad de Arroyo Rosarito, Baja California, México. Key Words: Baja California, Chloridoideae, Distichlis, halophytic grass. A putative hybrid between Monanthochloé littoralis Engelm. and Distichlis spicata (L.) Greene from Baja California, Mexico was re- ported by Stephenson (1971). The putative hybrid resembled M. Jittoralis in vegetative morphology and D. spicata in inflorescence structure. Mo- nanthochloé littoralis is distributed in coastal regions of subtropical Mexico and USA with one inland population known from Coahuila, Mexico. Distichlis spicata has a much broader distribution in coastal and inland North and South America. Recent work has placed Monanthochloé into synonymy with Distichlis based upon anatomical, morphological, and molecular evidence (Bell and Columbus 2008). Thus, ©. /ittorialis is hereafter referred to as D. Jittoralis and the putative hybrid is considered interspecific. The present study was undertaken to determine if the population from Stephenson (1971) was still extant and to test Whether the plants belonging to this population are hybrids, as Stephenson hypothesized. A hypothesis of hybrid origin predicts that incongruence may be observed between phylog- enies derived from nuclear (biparentally inherit- ed) and chloroplast (uniparentally inherited) DNA sequences (McDade 1992; Rieseberg et al. 1996 and refs. therein; Blattner 2004; Jakob and Blattner 2006). To test this hypothesis, I present new sequence data derived from both nuclear and chloroplast genomes. These data are added to existing matrices from Bell (2007) and Bell and Columbus (2008). In addition, whole plant morphology, micromorphology of the abaxial surfaces of blades and lemmas, and _ blade transectional anatomy of the putative hybrid plants were studied and compared to other species of Distichiis. METHODS During the springs of 2008 and 2009, extensive _ searches for historical populations of the putative | hybrid were conducted near El Nuevo Rosarito | in Baja California, Mexico. Observations were made of the growth habit and site conditions. Leaf material was dried in silica gel for DNA | analysis, blades, culms, and spikelets were | 2010] preserved in FPA (formalin:propionic acid:etha- nol, 1:1:18) for anatomical investigations, and pressed, dried herbarium specimens were pre- pared for morphological studies. For the remain- der of this report, the putative hybrid will be referred to as Baja grass. Genomic DNA was extracted from leaf tissue from Stephenson 68-304a (MSC 216526) and freshly collected material (Bell 458, RSA 754084) using DNeasy Plant Mini Kits (Qiagen, Valencia, CA). Sequences of the nuclear ribo- somal internal transcribed spacer (ITS) as well as chloroplast ndhF and trnL—trnF were amplified using primers and protocols described in Bell and Columbus (2008). In order to detect possible allelic variation in ITS, the amplification product was cloned using a TOPO TA kit (Invitrogen, Carlsbad, CA) following the manufacturer’s instructions. Ten colonies per sample were screened. Cycle sequencing was conducted at Rancho Santa Ana Botanic Garden on an ABI 3130xl Genetic Analyzer (Applied Biosystems, Foster City, CA) following the protocols of Bell and Columbus (2008). Sequences were assembled, edited, and incor- porated into existing alignments from Bell (2007) and Bell and Columbus (2008). The same out- group taxa used in Bell and Columbus (2008) (Allolepis texana (Vasey) Soderstr. & H. F. Decker; Bouteloua dactyloides (Nutt.) Columbus; Eragrostis obtusiflora (E. Fourn.) Scribn.; Jouvea pilosa (J. Presl) Scribn.) were employed in this study. Maximum parsimony (MP) and Bayesian inference (BI) analyses were conducted on the ITS and combined chloroplast (ndhF + trnL— trnF) data sets using PAUP* (Swofford 2002) and MrBayes vers. 3.0b4 (Huelsenbeck and Ronquist 2001) using the search parameters outlined in Bell and Columbus (2008). In addition, ITS 1 and ITS 2 were analyzed separately (Yokota et al. 1989; Liu and Schardl 1994; Mai and Coleman 1997). Branch support was assessed via posterior probabilities (PP), parsimony bootstrap (BS), and Bremer Support Values (BSV) (Bremer 1988) following Bell and Columbus (2008). Sequences generated during this study were submitted to GenBank and accession numbers are given in Appendix 1. Abaxial surfaces of leaf blades and lemmas were observed following the methods of Bell and Columbus (2008). Transectional anatomy of leaf blades was examined following Columbus (1999). Descriptive terminology follows Ellis (1976) for anatomy and Ellis (1979) for morphology. RESULTS Collection Site As described by Stephenson (1971), I found Baja grass growing along alkaline seeps in BELL: DISTCHLIS BAJAENSIS a3 Arroyo Rosarito adjacent to Mexico Hwy 1, southwest of El Nuevo Rosarito, approximately 100 km north of the border with Baja California Sur. Coordinates of the collection site are 28°43'36"N 114°43'17"W. Baja grass was one of the dominant species at the site and one of the few grasses present although some D. spicata was noted also. No D. littoralis was observed. Stephenson found fragments of both male and female plants at the heavily grazed site in 1968; only male plants were located during my extensive searches. Burros and cattle were observed in the area, but the population was not heavily grazed during the time of my collections in 2008 and 2009. Morphological features of the earlier collection (Stephenson 68-304a, MSC 216526), e.g., exserted inflorescences, spikelets with glumes, suggested affinities to D. spicata. However, in the field, its growth habit resembles that of D. /ittoralis; thus, it is clear why Stephenson would have considered D. littoralis to be a possible parent. Like D. littoralis, Baja grass possesses stolons and fre- quently grows up through (as on a trellis) adjacent plants such as species of Juncus and Lycium. The leaves of both D. Jittoralis and D. spicata are usually dark blue-green; those of Baja grass are yellowish-green. DNA Sequence Analysis Nine of ten cloned ITS sequences from Baja grass (Stephenson 6&-304a) were identical; the tenth sequence differed by a single base pair. Sequences from recently collected material (Be// 458) of ITS (to the group of nine) and ndhF were identical to those generated from Stephenson 66- 304a; only sequences from Stephenson 68-304a (including trnL—trnF) were used in the analyses. Descriptive statistics for the MP analyses are given in Table |. Both specimens of Baja grass (Stephenson 68-304a and Bell 458) showed a single unique indel, a three base pair repeat in ITS. The trees with the highest log-likelihood value are shown in Figure 1 (ITS) and Figure 2 (combined chloroplast). In the ITS analyses, Baja grass is supported as sister to all other Distichlis (BS = 91%, PP = 1.00, BSV = 4). In the combined chloroplast tree, Baja grass is retrieved in a polytomy with all other Distichlis species with good support for the clade (BS = 99%, PP = 1.00, BSV = 7). When ITS 1 and 2 were analyzed separately, the topological position of Baja grass changes (data not shown). With ITS 1 (in the MP strict consensus tree), Baja grass resolves as sister to D. laxiflora and D. scoparia; with ITS 2, it resolves as sister to the D. spicata clade. However, neither of these positions was supported. In addition, when an ITS sequence from the Baja grass was aligned and analyzed with a dataset 56 MADRONO TABLE 1. [Vol. 57 DESCRIPTIVE STATISTICS FOR MAXIMUM PARSIMONY ANALYSES. MP = maximum parsimony, PIC = parsimony informative characters, CI = consistency index, RI = retention index. J missing Region Aligned length data ITS 641 0 ndhF + trnL—-trnF 2111 + 1040 0.1 98/4] Distichlis humilis JU ARG | oy *! Distichlis humilis BOL */139 Distichlis eludens SLP MEX Distichlis eludens SLPc MEX | x Distichlis laxiflora BA ARG ee | |’ Distichlis laxiflora CO ARG Distichlis scopatia RN ARG / 96N 96/3 B Distichlis scoparia VA CHI 96/5 * {41 Distichlis australis RN ARG * * I Distichlis australis SC ARG */7 4 Distichlis acerosa LR ARG */44| * © Distichlis acerosa CA ARG * /7¢ Distichlis littoralis TX USA * L Distichlis littoralis coCA USA 73/3~ Distichlis spicata inCA USA 107 Distichlis spicata PERU 92/2 F Distichlis spicata TX USA *1 b Distichlis spicafa coCA USA Distichlis spicata BC CAN Distichlis distichophylia SA AUS Distichlis distichophylila VIC AUS * {15 | Distichlis palmeri SOc MEX * I Distichlis paimeri SOf MEX Distichlis spicata CO MEX 6Ri37 ee cas VRUSA * Distichlis spicata CH ARG Distichlis spicata VA CHI Baja Grass Eragrostis obtusifiora 96/3 | * 91/4 */9 * * 87/2] | * Jouvea pilosa Allolepis texana Bouteloua dactyloides Fic. 1. Tree with the highest log-likelihood score from Bayesian analysis of ITS. Bootstrap values followed by Bremer support are given above the branches, and below are posterior probabilities. An asterisk indicates 100% bootstrap or 1.00 posterior probability. Branches marked with an arrow collapse in the strict consensus from parsimony analysis. Geographical abbreviations are as follows: ARG = Argentina (BA = Buenos Aires, CA = Catamarca, CH = Chubut, CO = Cordoba, JU = Jujuy, LR = La Rioja, RN = Rio Negro, SC = Santa Cruz); AUS = Australia (SA = South Australia, VI = Victoria); BOL = Bolivia; BC CAN = British Columbia, Canada; CHI = Chile (AN = Antofagasta, VA = Valparaiso); MEX = Mexico (CO = Coahuila, SLP = San Luis Potosi, SO = Sonora); USA (CA = California, TX = Texas, VR = Virginia). + of MP MP tree trees length PIC Cl RI 26 500 148 0.76 0.82 209 275 84 0.83 0.86 Distichlis spicata TX USA Distichlis spicata BC CAN 87/2 Distichlis spicafa cocA USA i Distichlis spicata CO MEX Na] = Distichlis spicata VR USA Distichlis spicata PERU shes Distichlis spicata CH ARG 86/1 Distichlis spicata VA CHI Distichlis distichophyliia SA AUS Distichlis distichophylia VIC AUS Distichlis spicata inCA USA Distichlis paimeri SOc MEX Distichlis palmeri SOf MEX Distichlis scoparia RN ARG Distichlis scoparia VA CHI Distichlis laxiflora CO ARG ‘ Distichlis laxiflora BA ARG Distichlis humilis JU ARG Distichlis humilis BOL Baja Grass 99/7 98/4 Distichlis acerosa LR ARG * 99/51 * ! Distichlis acerosa CA ARG * 962 nistichiis littoralis TX USA * L Distichlis littoralis coCA USA 98/4 Distichlis australis RN ARG * | *! Distichlis australis SC ARG * {7 Distichlis eludens SLP MEX . Distichlis eludens SLPc MEX \j——————_ Bouteloua dactyloides Eragrostis obtusifiora 85/2 Allolepis texana Jouvea pilosa Fic. 2. Tree with the highest log-likelihood score from Bayesian analysis of combined chloroplast data set (ndhF + trnL—trnF). Bootstrap values followed by Bremer support are given above the branches, and below are posterior probabilities. An asterisk indicates 100% bootstrap or 1.00 posterior probability. Branches marked with an arrow collapse in the strict consensus from parsimony analysis. Geographical abbreviations are the same as in Fig. 1. Fic. 3. Comparisons of abaxial blade surfaces; A. Baja grass (Bell 458), B. Distichlis spicata (Bell 231), C. Distichlis littoralis (Bell 260). a = papilla, b = clustered papillae, c = short cell, d = long cell, e = microhair, cz = costal zone, 1z = intercostal zone. Scale bar applies to A, B, and C. derived from 84 chloridoid genera (Bell 2007), it resolved as sister to Distichlis (data not shown). Micromorphology Abaxial surfaces of leaf blades of Baja grass are highly papillate making it difficult to observe features such as long and short cells, microhairs and stomates (Fig. 3A). In the costal zones of blades of Baja grass and D. spicata, there are regular pairs of large and small papillae (Fig. 3A, B). In intercostal zones of Baja grass and D. littoralis, papillae form complexes associated with microhairs (Fig. 3A, C). In Baja grass and species of Distichlis, stomates occur in two files along each edge of the intercostal zone; stomates are frequently obscured by complexes of papillae making them difficult to observe from a surface BELL: DISTCHLIS BAJAENSIS 7 Fic. 4. Comparisons of abaxial surfaces of lemmas; A. Baja grass (Bell 458), B. Distichlis spicata (Bell 277), C. Distichlis littoralis (Bell 260). a = papilla, b = clustered papillae, c = short cell, d = long cell, e = stomate. Scale bar applies to A, B, and C. view. Stomates of Baja grass and Distichlis have dome shaped subsidiary cells. Abaxial surfaces of lemmas of Baja grass have many papillae that obscure features such as microhairs and stomates (Fig. 4). There are many complexes of papillae similar to those found on species of Distichlis. Microhairs and stomates appeared to be more sparse on lemmas of Baja grass than in Distichlis but they may be hidden by papillae. Anatomy Blade transectional anatomy of Baja grass is similar to that of Distichlis species (Fig. 5). The outline of the blade transection is broadly U- shaped. There are adaxial furrows between all vascular bundles to a depth of about half of the blade thickness. Furrows are absent or shallow on the abaxial side. Blades possessed about 14 total vascular bundles, three of which were Ist order. Examination of species of Distichlis found from 18—24 (7-9 Ist order) vascular bundles in blades of D. spicata and 9 (3 Ist order) in D. 58 MADRONO rmvb AEDS sd FIG. 5. Comparison of blade anatomy; A. Baja grass (Bell 458), B. Distichlis spicata (Bell 375), C. Distichlis littoralis (Bell 260). a = outer bundle sheath, b = inner bundle sheath, c = xylem, d = phloem, e = mesophyll, f = sclerenchyma, g = microhair, h = papillae, 1 = colorless cells. 0.02 mm § ; FS x Bi i -_ “4 > on 2 3 e . a ~ gee Pit ee Ps i tS aa oe « FIG. 6. [Vol. 57 littoralis (Bell and Columbus 2008). Second order vascular bundles form a regular arrangement between the Ist order bundles; a single 3rd order bundle is found at each margin. Sheath cells in all vascular bundles are elliptical in shape. The outline of 3rd order bundles is round and that of Ist and 2nd order bundles are elliptical. First order vascular bundles have a continuous double sheath that is not interrupted and lacks exten- sions. Phloem is directly adjacent to the inner sheath, and metaxylem is narrow. Walls of the inner sheath are thickened. Chloroplasts are centripetally arranged in the outer sheath cells. Very narrow strands of sclerenchyma are found on both adaxial and abaxial sides of most vascular bundles, and a small sclerenchyma cap occurs at the margins. Mesophyll forms a single layer of radially arranged cells. Colorless cells form uni- to multiseriate columns between all vascular bundles. Bulliform cells are associated with colorless cells at the base of furrows. Other epidermal cells are small and have numerous papillae on both surfaces. First order vascular bundles of Baja grass show Kranz anatomy of the type that predicts NAD-ME C, photosynthesis (Prendergast and Hattersley 1987). Bicellular microhairs of Baja grass are dumb- bell or flask shaped, with a portion of the basal cell sunken below the epidermis into mesophyll or colorless cells (Fig. 6). DISCUSSION Analyses of molecular data do not support the hypothesis that Baja grass is a hybrid between D. littoralis and D. spicata (McDade 1992; Rieseberg et al. 1996). In both ITS and combined chloro- plast (ndhF + trnL—trnF) analyses, Baja grass does not group with any other species but is supported as sister to or a member of Distichlis (Figs. 1 and 2). However, three South American endemics, D. humilis, D. laxiflora, and D. scoparia, are resolved Bicellular microhairs, A. Baja grass (Bell 458), B. Distichlis spicata (Bell 231), C. Distichlis littoralis (Bell 260). a = distal cell, b = basal cell. Scale bar applies to A, B, and C. 2010] ITs i 11 21 31 | Allolepis TCGTGACCCTGACCAAAACAGACTGTGAATGTGTCATCC Bouteloua@ —s nweeeeevee Dececcves Avie Cr meio GA see eeree EragroStiS 0. cesecesesecseees Tesccee CG eevers GS shokeleusiexe te G Jouvea oeAnecAvece Nee, i6) aioe: ey sx ii6l 6 ai fe! 6:56 CTs weew awe D. Spicata —=— sewers accacnascccssces Cosees Ciscsioieteiiecrs) Baja gQrasSS =—=— na eeeeeseees Tease rGesnne Ciaveuerete CA ayexe soe A D. Littoralis wee. STisiceiyeiie-(et/eye:e eee apecsvene.s Gieue e ciiee etove Guevee ies ndhF 1882 1891 1901 1911 | | | Allolepis AATTTTAAAAATTCCCTTGTAAAAGGGAATCCCAAAABAGTT Bouteloua —=— ww eeneeevvccee Tow aTecCevncsese Mg Te leleawere AA EFAGrOSELS Os © see ie 66 ole Sie se Weleie) oa ee 66 680 0 vase pci sa 8 ee, AA JOUVECEA ———— ww ww wc cw reece cern r ere reece seeenes Decevccce D. SPiCAtA — ns oeeerereverccecscssenees TT ievg:tecsriecetie: sisi sie! eves Baja grass (Sve veeus couiene er eyerevereve (eles tovvel svaneneyevelraricr ferenenevershe:sceie AA D. Littoralis wcvecucencscvscses Tate G:scetere vest el o7eieie! a ececaie 6 Fic. 7. Patterns of variation in two sections from this study’s molecular data sets. Top, the beginning of nuclear ITS; bottom, a relatively variable region of ndhF. The first four taxa are the outgroup for this study; they are followed by D. spicata (Bell 231), Baja grass (Stephenson 68-304a), and D. littoralis (Bell 260). in conflicting positions by nuclear and chloro- plast markers demonstrating that there is ade- quate signal in these datasets to detect potential reticulation. A visual comparison of sequence segments from ITS and ndhF (Fig. 7) does not reveal the additive pattern that would be predicted if Baja grass were a hybrid between D. Jittoralis and D. spicata. If Baja grass were a relatively recent hybrid I would expect to see polymorphisms that were compatible with derivation from D. Jittoralis or D. spicata. If the hybridization event occurred in the distant past so that homogenization of ITS alleles had taken place (as is indicated by the finding of nine identical clones), then I would expect that sequences of Baja grass would resemble one or the other of the putative parents. If Baja grass were a hybrid, I would expect that chloroplast sequences would be the same or highly similar to one of the putative parents. As seen in Fig. 7, these are not the patterns that are observed. Variation in sequences from Baja grass does not suggest derivation from either D. littoralis or D. spicata. Baja grass has the same blade organization as species of Distichlis (Fig. 4). Blades are U- shaped, with vascular bundles separated by furrows and columns of colorless cells. There are few Ist order vascular bundles with narrow xylem elements. There is some variation in the amount of sclerenchyma but its distribution is similar. Dumb bell or flask shaped microhairs are found in Baja grass and all species of Distichlis as well as Eragrostis obtusiflora and a few other more distantly related halophytic chloridoids. There is evidence that these microhairs are the site of salt secretion in halophytic chloridoids (Oross and Thomson 1982; Amarasinghe and Watson 1988; Warren and Brockelman 1989; BELL: DISTCHLIS BAJAENSIS 59 Ramadan 2001; Bell and O’Leary 2003). Salt crystals have been observed on the surface of Baja grass blades. Habitat preferences and anatomical and mor- phological similarities of the Baja grass to other Distichlis species support its inclusion as a new species within the genus. All Distichlis species occur in saline or alkaline habitats, are dioecious, have multi-nerved lemmas, Kranz anatomy that predicts NAD-ME type C4, photosynthesis, nu- merous papillae on blade surfaces, bulbous bicellular microhairs, columns of colorless cells between vascular bundles, narrow metaxylem elements in Ist order vascular bundles, and relatively few Ist order vascular bundles per blade. Based upon these shared characters and the molecular evidence, this grass 1s described as a new species of Distichiis. TAXONOMY Distichlis bajaensis H. L. Bell. sp. nov. (Fig. 8). — Type: MEXICO, Baja California, Municipio de Ensenada, salt marsh in arroyo | km SW of Rosarito, area dominated by juncus and salt grasses, heavily grazed by burros and goats, October 1968, Stephenson 68-304a (holotype: MSC 216526! [not MSC 216528 or 289874)]). Paratype: MEXICO, Baja California, Munici- pio de Ensenada, southwest of El Nuevo Rosarito, 28°36'40"N, 114°03'03"W, 100 m elevation, broad, dry arroyo with alkaline seeps, growing with Distichlis spicata (L.) Greene, Juncus acutus L., Allenrolfea sp., Lycium sp., and Salicornia sp., 2 April 2008, Bell 458, (BCMEX, MEXU, MO, RSA, UC, US). Gramen perenne decumbens rhizomatosum stoloniferum ramis intravaginalibus secus_ sto- lones, 8-12 cm altum, ligulae pilis linea minuta dispositis, laminis 8-15 mm longis ad collo patentibus, in apicem pungentem sensim decres- centibus, ad faciem adaxialem parum flexis, pilis antrorsis secus margines et faciem abaxialem fascis vascularis medi ad et supra flexuram. Sprawling, decumbent perennial with rhizomes and stolons, 8—12 cm tall, intravaginal branching along stolons, culms | mm in diameter, glabrous, sheaths open, glabrous, with tiny hairs along margins, ligules a minute line of hairs, blades 8— 15 mm long, spreading at collar, narrowing gradually to pungent tip, with slight bend toward adaxial side, antrorse hairs along margins and along the abaxial side of the median vascular bundle at and above bend, male inflorescences a small panicle of racemes, inflorescences exserted above blade tips on peduncles of up to 1 cm, flattened pedicels 3—5 mm with toothed margins, 2-5 spikelets per inflorescence, 2—4 florets per spikelet, 1st glume 3 mm, 2nd glume 5 mm, both hyaline with a single nerve, lemmas 7—9 mm with 60 MADRONO [Vol. 57 FiG. 8. Distichlis bajaensis. a. Female plant habit; b. Rhizome; c. Male plant habit; d. Detail of blades; e. Male spikelet; f. First glume from male spikelet; g. Second glume from male spikelet; h. Lemma from male spikelet; 1. Palea from male spikelet. a and c from Stephenson 68-304a; b, d —i from Bel/ 458. Illustration by Amanda Labadie. 2010] TABLE 2. LITTORALIS, AND D. SPICATA. D. bajaensis 0.8—1.5 narrow gradually to pungent tip divaricate slight bend toward adaxial side Character Blade length (cm) Blade tips Blade angle from culm Blade curve or bend Glumes present Male inflorescence yes exserted Plant color yellowish green 7-11 indistinct nerves, hyaline, palea_ slightly shorter than lemma, enclosed within lemma, anthers 2.5—3.5 mm, straw colored (some with purple tinge). No fresh female inflorescences or caryopses were examined. Stephenson (1971) observed extensive grazing in the collection area and noted “only fragmentary grass specimens could be obtained”’. He was not able to collect caryopses but provided observations of ovaries and stigmas. Table 2 gives characters that can be used to distinguish between D. bajaensis, D. littoralis, and D. spicata. A distinctive field character is a small bend near the middle of the leaf blade (Fig. 8d). Generally, at and above the bend, short, antrorse hairs occur along the median vascular bundle on the abaxial surface. Future studies of D. bajaensis will focus on the total distribution of this species and the relation- ship of this species to the rest of Distichlis. At present, the species is known from a single large population that appeared to be all or predomi- nately male. It is crucial to learn if other populations exist, the proportions of sexes in those populations, and their proximity to the Arroyo Rosarito population. Although Stephen- son found male and female plants during his 1968 collection, John and Charlotte Reeder observed only male plants in 1979 (R. Felger, University of Arizona, personal communication). Distichlis species are capable of extensive vegetative repro- duction via rhizomes and stolons and _ highly skewed sex ratios have been observed in many populations (e.g., D. distichophylla, Connor and Jacobs 1991; D. spicata, Freeman et al. 1976; Eppley et al. 1998). Even though vegetative reproduction occurs, the conservation status of D. bajaensis may well be extremely fragile with few or no female plants in existence. The ITS phylogeny places D. bajaensis as sister to the remaining Distichlis (Fig. 1). If this is corroborated by future work, D. bajaensis will hold a phylogenetic position that is critical to investigating character development and evolu- tion in Distichlis by enabling researchers to better <0.8 narrow abruptly to divaricate straight or slight curve absent bluish green BELL: DISTCHLIS BAJAENSIS 61 COMPARISON OF CHARACTERS THAT CAN BE USED TO DISTINGUISH DISTICHLIS BAJAENSIS, D. D. littoralis D. spicata 22.0 narrow gradually to blunt tip (rarely pungent) appressed or divaricate generally straight blunt tip toward abaxial side present yes bluish green understand pleisomorphies vs. apomorphies in the genus. ACKNOWLEDGMENTS I thank Carol Annable and Gwyneth Pett for assistance with collecting. Alan Prather, curator of MSC, kindly facilitated a loan of Stephenson 68-304a and granted permission for destructive sampling for DNA analysis. Thanks to Amanda Labadie for the illustration, to Mark Garland for the Latin diagnosis, and to Cristina Martinez-Habibe and Gilberto Ocampo for the translation of the Resumen. J. Travis Columbus photographed the blade transection and microhair (Figs. SA and 6A). Kelly Allred, Lynn Clark, Lucinda McDade, Linda Prince, and Erin Tripp read earlier drafts and provided many helpful sugges- tions that improved the manuscript. Lon E. Bell gave support and encouragement at every stage of this project. LITERATURE CITED AMARASINGHE, V. AND L. WATSON. 1988. Compara- tive ultrastructure of microhairs in grasses. Botan- ical Journal of the Linnean Society 98:303—319. BELL, H. L. 2007. Phylogenetic relationships within Chloridoideae (Poaceae) with emphasis on subtribe Monanthochloinae. Ph.D. dissertation. Claremont Graduate University, Claremont, CA. AND J. T. COLUMBUS. 2008. Proposal for an expanded Distichlis (Poaceae, Chloridoideae): sup- port from molecular, morphological, and anatom- ical characters. Systematic Botany 33:536—S51. AND J. W. O'LEARY. 2003. Effects of salinity on growth and cation accumulation of Sporobolus virginicus (Poaceae). American Journal of Botany 90:1416—1424. BLATTNER, F. R. 2004. Phylogenetic analysis of Hordeum (Poaceae) as inferred by nuclear rDNA ITS sequences. Molecular Phylogenetics and Evo- lution 33:289-299. BREMER, K. 1988. The limits of amino acid sequence data in angiosperm phylogenetic reconstruction. Evolution 42:795—803. CoLuMBuUs, J. T. 1999. Morphology and leaf blade anatomy suggest a close relationship between Bouteloua aristidoides and B. (Chondrosium) erio- poda (Gramineae: Chloridoideae). Systematic Bot- any 23:467-478. 62 MADRONO CONNOR, H. E. AND S. W. L. JAcoss. 1991. Sex ratios in dioecious Australian grasses: a preliminary assessment. Cunninghamia 2:385—390. ELuis, E. P. 1976. A procedure for standardizing comparative leaf anatomy in the Poaceae: I. The leaf-bade as viewed in transverse section. Bothalia 12:65—-109. . 1979. A procedure for standardizing compara- tive leaf anatomy in the Poaceae: II. The epidermis as seen in surface view. Bothalia 12:641—671. EppPLey, S. M., M. L. STANTON, AND R. K. GROSBERG. 1998. Intrapopultion sex ration variation in the salt grass Distichlis spicata. The American Naturalist 152:659-670. FREEMAN, D. C., L. G. KLIKOFF, AND K. T. HARPER. 1976. Differential resource utilization by the sexes of dioecious plants. Science 193:597—598. HUELSENBECK, J. P. AND F. RONQUIST. 2001. MRBAYES: Bayesian inference of phylogenetic trees. Bioinformatics 17:754-755. JAKOB, S.S. AND F. R. BLATTNER. 2006. A chloroplast genealogy of Hordeum (Poaceae): long-term per- sisting haplotypes, incomplete lineage sorting, regional extinction, and the consequences for phylogenetic inference. Molecular Biology and Evolution 23:1602—1612. J-S. AND C. L. SCHARDL. 1994. A conserved sequence in internal transcribed spacer | of plant nuclear rRNA genes. Plant Molecular Biology 26:775—778. MAI, J. C. AND A. W. COLEMAN. 1997. The Internal Transcribed Spacer 2 exhibits a common secondary structure in green algae and flowering plants. Journal of Molecular Evolution 44:258—271. McDApgE, L. A. 1992. Hybrids and _ phylogenetic systematics II. The impact of hybrids on cladistic analysis. Evolution 46:1329—1346. Oross, J. W. AND W. W. THOMSON. 1982. The ultrastructure of the salt glands of Cynodon and Distichlis (Poaceae). American Journal of Botany 63:939—-949. PRENDERGAST, H. D. V. AND P. W. HATTERSLEY. 1987. Australian Cy grasses (Poaceae): leaf blade anatomical features in relation to C4 acid decar- boxylation types. Australian Journal of Botany 35:355-382. RAMADAN, T. 2001. Dynamics of salt secretion by Sporobolus spicatus (Vahl.) Kunth from sites of differing salinity. Annals of Botany 87:259—266. RIESEBERG, L. H., J. WHITTON, AND C. R. LINDER. 1996. Molecular marker incongruence in plant hybrid zones and phylogenetic trees. Acta Botanica Neerlandica 45:243—262. STEPHENSON, S. N. 1971. A’ putative Distichlis xX Monanthochloé (Poaceae) hybrid from Baja Cali- fornia, Mexico. Madrono 21:125—127. SWOFFORD, D. L. 2002. PAUP*: phylogenetic analysis using parsimony (* and other methods), v. 4.0 beta 10. Sinauer Associates, Sunderland, MA. WARREN, R. S. AND P. M. BROCKELMAN. 1989. Photosynthesis, respiration, and salt gland activity of Distichlis spicata in relation to soil salinity. Botanical Gazette 150:346—350. YOKOTA, Y., T. KAWATA, Y. HDA, A. KATO, AND S. TANIFUJI. 1989. Nucleotide sequences of the 5.8S rRNA gene and internal transcribed spacer regions in carrot and broad bean ribosomal DNA. Journal of Molecular Evolution 29:294—301. LIU, [Vol. 57 APPENDIX | LIST OF TAXA SAMPLED Taxa used as sources of DNA for the molecular phylogeny. See Bell and Columbus (2008) for additional details about Distichlis morphology and anatomy. Biogeographical abbreviations used in Figs. 1 and 2 are underlined. GenBank accession numbers (starting with EF or GU) appear in this order: ITS, trnL-trnF, ndhF. For a few specimens, the trnL-trnF sequence was not available and this is designated as ‘NA’. Allolepis texana (Vasey) Soderstr. & H. F. Decker. USA. TEXAS: Be// 240 (RSA), EF153021, EF156670, EF561646. Bouteloua dactyloides (Nutt.) Columbus. MEXICO. QUERETARO: Columbus 2329 (RSA), EF153026, EF156675, EF561647. Eragrostis obtusiflora (E. Fourn.) Scribn. MEXICO. MICHOACAN: Bell 314 (RSA), EF196874, EF196902, EF561648. Jouvea pilosa (J. Presl) Scribn. MEXICO. JALISCO: Bell 247 (RSA), EF153057, EF156706, EF561649. Distichlis acerosa (Speg.) H.L. Bell & Columbus (= Monantho- chloé acerosa (Griseb.) Speg.). ARGENTINA. LA RIOJA: Bell 389 (RSA), LR ARG, EF196897, EF196924, EF561671. CATAMARCA: Bell 392 (RSA), CA ARG, EF196898, EF196925, EF561672. Distichlis australis (Speg.) Villamil. ARGENTINA. RIO NEGRO: Bell 330 (RSA), RN ARG, EF196875, EF196903, EF561650. SANTA CRUZ: Bell 357 (RSA), SC ARG, EF196876, EF196904, EF561651. Distichlis bajaensis H.L. Bell. MEXICO. BAJA CALIFORNIA: Bell 458 (RSA), GU562862, NA, GU562863; Stephen- son 68-304a (MSC), ITS Clones 1, 2, 3, 5, 6, 7, 8, 9, 10 GU562864, ITS Clone 4 GU562865, GU562867, GU562866. Distichlis distichophylla (Labill.) Fassett. AUSTRALIA. VICTORIA: Cochrane 1198 (MEL), VIC AUS, EF196877, EF196905, EF561652. SOUTH AUSTRALIA: 12 October 2003, Walsh s. n. (12 October 2003), SA AUS, EF196878, EF196906, EF561653. Distichlis eludens (Soderstr. & H.F. Decker) H.L. Bell & Columbus, (=Reederochloa eludens So- derstr. & H.F. Decker). MEXICO. SAN LUIS POTOSI: Bell 250 (RSA), SLP MEX, EF153077, EF156726; Columbus 4133 (RSA), SLPc MEX, EF196901, EF196928, EF561676. Distichlis humilis Phil. ARGENTINA. JUJUY: Bell 405 (RSA), JU ARG, EF196879, EF196907, EF561654. BOLIVIA. DEPARTAMENTO ORURO: Peterson 12833 (US), BOL, EF196880, NA, EF196908. Distichlis laxiflora Hack. ARGENTINA. BUENOS AIRES: Bell 367 (RSA), BA ARG, EF196881, EF196909, EF561656. CORDOBA: Bell 381 (RSA), CO ARG, EF196882, EF196910, EF561657. Distichlis littoralis (Englem.) H.L. Bell & Columbus (=Monanthochloé littoralis Engelm.). USA. TEXAS: Bell 236 (RSA), coTX USA, EF153065, EF156714, EF561673. CALIFORNIA: Bell 260 (RSA), coCA USA, EF196900, EF196927, EF561674. Distichlis palmeri (Vasey) Fassett ex I. M. Johnst. MEXICO. SONORA: Columbus 3586 (RSA), SOc MEX, EF196883, EF196911, EF561658; Felger 91- 39 (RSA), SOf MEX, EF196884, EF196912, EF561659. Distichlis scoparia (Nees ex Kunth) Arechav. ARGEN- TINA. RIO NEGRO: Bell 328 (RSA), RN ARG, EF196885, EF196913, EF561660. CHILE. VALPA- RAISO: Bell 374 (RSA), VA CHI, EF196886, EF196914. EF561661. Distichlis spicata (L.) Greene. USA. CALIFORNIA: Bell 231 (RSA), inCA USA, 2010] EF153040, EF156689, EF561662; Be//l 259 (RSA), coCa USA, EF196890, EF196918, EF561665. TEXAS: Bell 237 (RSA), TX USA, EF196887, EF196915, EF561663. VIRGINIA: Bell 290, VR USA, (RSA), EF196892. EF196920, EF561667. CANADA. BRITISH COLUM- BIA: Bell 277 (RSA), BC CAN, EF196891, EF196919, EF561666. MEXICO. COAHUILA: Bell 245 (RSA), BELL: DISTCHLIS BAJAENSIS 63 CO MEX, EF196888, EF196916, EF561664. ARGEN- TINA. CHUBUT: Bell 340 (RSA), CH ARG, EF196893, EF196921, EF561668. CHILE. VALPA- RAISO: Bell 375 (RSA), VA CHI, EF196895, EF196922, EF561669. PERU. REGION LAMBA YE- QUE: Columbus 3432, (RSA), PERU, EF196896, EF196923, EF561670. MADRONO, Vol. 57, No. 1, pp. 64-72, 2010 CHENOPODIUM LITTOREUM (CHENOPODIACEAB), A NEW GOOSEFOOT FROM DUNES OF SOUTH-CENTRAL COASTAL CALIFORNIA NURI BENET-PIERCE AND MICHAEL G. SIMPSON Department of Biology, San Diego State University, San Diego, CA 92182, USA nuribpierce@gmail.com ABSTRACT Chenopodium littoreum is described as new. It had been incorrectly cited in the past as C. carnosulum Mog. var. patagonicum (Phil.) Wahl, a variety of the South American C. carnosulum. However, C. littoreum differs from the C. carnosulum complex in having narrowly elliptic to lanceolate and mostly unlobed leaves, consistently five stamens per flower, and seeds that are invariably horizontal. Chenopodium littoreum is similar to another South American taxon, C. patagonicum Phil. (=C. philippianum Aellen), but the latter differs in having basally lobed leaves, sepals fused above the middle, and generally one or two (rarely five) stamens. Chenopodium littoreum has a range currently known only from coastal dunes of San Luis Obispo Co. and Santa Barbara Co. of the Central Coast of California, plus a single historic collection from Los Angeles Co. of the South Coast of California. Key Words: Chenopodium, C. Chenopodiaceae, dune flora, coastal goosefoot. Chenopodium (Chenopodiaceae; Amarantha- ceae sensu APG III 2009) is a large genus of approximately 100 species of mostly temperate plants, with a worldwide distribution. It is segregated from the related genus Dysphania (ca. 32 species) in recent treatments (Clemants and Mosyakin 2003a, b). Although many species of Chenopodium are weeds, some are economi- cally important, such as the pseudo-grain C. quinoa of South America (Mabberley 2008). The preparation of the Chenopodium treatment (Clemants and Benet-Pierce in preparation) for the second edition of The Jepson Manual necessitated the resolution of issues left pending by the untimely death of Dr. Steve Clemants of the Brooklyn Botanic Garden. One major issue was the taxon Chenopodium carnosulum Mog. var. patagonicum (Phil.) Wahl, several specimens of which had been cited as occurring (and presumably naturalized) in San Luis Obispo and Santa Barbara counties, California (Wilken 1993). After reviewing the literature and observ- ing numerous specimens and specimen images, we are convinced that the California taxon in question does not correspond to Chenopodium carnosulum Mogq., nor to C. patagonicum Phil. (C. philippianum Aellen; see below), and therefore has been an ongoing case of misidentification. We propose here that what was _ previously identified as Chenopodium carnosulum var. pata- gonicum is actually an undescribed, new species. We presume it to be native and endemic to California, as specimens of this taxon have not been found elsewhere. Chenopodium littoreum Benet-Pierce & M. G. Simpson, sp. nov. (Fig. 1).—Type: USA, California, San Luis Obispo Co., road along carnosulum var. patagonicum, C. patagonicum, C. philippianum, Jack Lake, ca. 9 km south of Arroyo Grande, ca. 16 m, 35.03858°N, 120.60378°W, 15 May 1966, R. F. Hoover 9856 (holotype: OBI 17235; isotypes: CAS 473439, 473440, 473441). Paratypes (see Fig. 1F, G for locality map): USA. CALIFORNIA. Los Angeles Co.: Playa del Rey, 33.96184°N, 118.4468°W, 14 May 1904, G. C. Grant s.n. (DS 91772). San Luis Obispo Co.: Oceano, 35.0946°N, 120.622327°W, 30 April 1910, G. F. Condit s.n. (UC 455220); Oceano Dunes, 35.09456°N, 120.622327°W, 30 May 1931, R. Hoffman 420 (CAS 189558); Oso Flaco Lake, 35.02941°N, 120.62756°W, 13 May 1950, L. S. Rose 50116 (CAS 367246, RSA 63058, UC 942915); Morro Bay, 35.37257°N, 120.863926°W, 9 June 1967, R. F. Hoover 10629 (OBI 17236); Morro Bay, 35.37257°N, 120.863926°W, 29 June 1969, J. R. Potter 51 (OBI 4176); Little Coreopsis Hill, 35.03433°N, 120.615°W, 25 May 1980, A. P. Griffiths s.n. (OBI 56356); Black Lake, Highway 1, 35.05885°N, 120.609709°W, 25 April 1985, D. Keil 18563 (OBI); Los Osos, 35.31548°N, 120.86648°W, 9 June 1985, D. Keil 18790 (OBD). Santa Barbara Co.: SBC Vandenberg Air Force Base, 34.79311°N, 120.621247°W, 23 August 1996, D. Keil 25849 (OBI 67573); North Base, 34.74747°N, 120.62801°W, 23 August 1996, D. Keil 25947 (OBI 67553). Chenopodium littoreum differt a C. carnosulum Mog. foliis integerrimis anguste ellipticis lanceo- latis vel late lanceolatis plerumque non-lobis basi cuneatis, apice mucronulatis, 5 stamenibus, et semenibus complanatis; differt a C. patagonicum Phil. et C. philippianum Aellen foliis integerrimis anguste ellipticis lanceolatis vel late lanceolatis plerumque non-lobis, calycis ulterioribus separa- tis, et 5 stamenibus. 2010] Ss Luis Obispo County Z CALIFORNIA é | > Hoover No. > | . AAN DILGO STATE RIVERATY HERBARIUM (SDSU? RF. me Fron remov ol for triosomdc study for Ut Jepscer Mircuzt San Luis Obispo Co. Santa Barbara Co. F Fic. 1. BENET-PIERCE AND SIMPSON: CHENOPODIUM LITTOREUM SP. NOV. 65 “4 5 eos San-tuis . “=3% - Obispo: mes 4 Edns “T= ters pre , fA fae 4 -“t al ... | San LuiszObispo Co./ San Lus “Pismo Sees gtES Obispo Bay pee ‘Minis s at 5 a fg eg Grande. esther Huasra yn wt Niger GN se re 121° W 35°N | } 4 | a { Dallender \ Hey ye } ipomo ns oy ¢ tOAS 241 { bY © ar Santa Maria car Medco Betteravia Lie j ti a, Orcutt | * Devon e” Brokoen / Santa Barbara Co. } r '® Vandenberg + 9 AFR yd yf -bosal \ - + \ elie RAL F ‘2, Vandenberg mn) ) Village } Uy “Ty Slit Chenopodium littoreum. A. Herbarium specimen (OBI 17235, holotype). B. Specimen (OBI 67553, paratype). Close-up of leaves, showing narrowly elliptic to lanceolate shape. C. Specimen (CAS 473441, isotype). Single leaf close-up; note farinose surface. Scale bar = | mm. D, E. Specimen (OBI 17235, holotype). D. Fruit, showing calyx lobes distinct almost to base (arrow). E. Flower, removed, showing five stamen filaments. Scale bar = 1mm. F. Distribution map of known collections. G. Close-up of specimen localities in Santa Barbara and San Luis Obispo counties. Chenopodium littoreum differs from C. carno- sulum Mogq. by having entire, narrowly elliptic, lanceolate, or widely lanceolate, mostly non- lobed, basally cuneate leaves, apex mucronulate, 5 stamens, and horizontal seeds; it differs from C. patagonicum Phil. and C. philippianum Aellen in having entire, narrowly elliptic, lanceolate, or widely lanceolate, mostly non-lobed leaves, with calyx lobes distinct to near base, and 5 stamens. Annual prostrate herb, branched from _ base, forming mats to ca 4 dm in diameter. Leaves alternate; petioles 5—9 mm long; blades narrowly elliptic, lanceolate, or broadly lanceolate, rarely basally lobed, 6—15 (20) mm long, 3—8 mm wide, 66 MADRONO light green; base cuneate, apex acute, obtuse, or rounded, often mucronulate, farinose adaxially, densely farinose abaxially. Inflorescence of glom- erules up to 7 mm wide, in axillary and terminal spikes and panicles, 1-15 cm long; bracts leaf- like. Flowers perfect, radial, approximately 1 mm in diameter; perianth uniseriate; calyx synsepal- ous, with five lobes, distinct to near base, lobes apically obtuse, densely farinose abaxially. Sta- mens five, distinct, whorled, antisepalous; fila- ments terete, yellow, with laterally dehiscent, dithecal, subbasifixed anthers. Gynoecium syn- carpous, hypogynous; ovary superior, with two stigmas. Placentation basal with one curved ovule. Fruit an achene, horizontal, dark brown, lenticular, margin rounded, approximately 1 mm in diameter; fruit wall minutely tuberculate to smooth, attached to the seed, but becoming loose at maturity. Seeds 0.9-1 mm in diameter, peri- spermous; seed coat smooth, black-brown to red. Distribution and habitat: Chenopodium littore- um is currently known from dunes of a narrow coastal strip of the Central Coast of California (San Luis Obispo and Santa Barbara counties), and one collection from the South Coast of California (Los Angeles Co.; Fig. 1F, G). Phenology: Chenopodium littoreum appears to flower and fruit from late April to as late as August. Etymology: The specific epithet, /ittoreum, Latin (pronounced li-TOR-e-um), translates as ““of the seashore,” in reference to the coastal distribution of this species. Suggested common name: Coastal Goosefoot. DISCUSSION California collections of Chenopodium littore- um, described here, have mostly been identified as Chenopodium carnosulum Mog. var. patagonicum (Phil.) Wahl (basionym C. patagonicum Phil.), purportedly a Californian variety of an otherwise mostly South American species. However, the species C. carnosulum 1s markedly different in a number of features from C. /ittoreum. Christian Horace Bénédict Alfred Moquin- Tandon described Chenopodium carnosulum in 1849. It is mostly found in the southernmost tip of South America, in Chile and Patagonia in Argentina, but specimens have been cited from Peru and Mexico. Examination of an on-line image of the holotype of C. carnosulum Mog. (K 583167, Port Gregory, Patagonia, Argentina; Fig. 2A) shows a plant with leaves that are relatively small, rhombic-deltoid, and strongly lobed; this is in contrast to the elliptic or lanceolate, mostly unlobed leaves of C. littoreum (Fig. 1[A—C). Physical examination of other specimens of C. carnosulum (UC 559383; GH 25/655, 257651, 257652; and GH (Mexia 7960, not accessioned; Fig. 2 B—E) and of the infra- [Vol. 57 species C. carnosulum Mog. var. scabricaule (Speg.) Aellen & Just (GH 257656) all show similar features. The leaves of all of these specimens are small, rhombic-deltoid and strong- ly lobed (elliptic to lanceolate or widely lanceo- late and mostly unlobed in C. /ittoreum); the flower has only one stamen or occasionally 2 (consistently 5 in C. /ittoreum); many of the seeds are vertical or oblique (consistently horizontal in C. littoreum); and the fruit wall is often mottled (mottling absent in C. /ittoreum). In addition, the description of Chenopodium carnosulum Mogq. from the protologue (Moquin-Tandon 1849) states: ““Folia 3-4 lin. [=6.3—8.4 mm] longa (incl. petiolo 1/2-1 lin. [=1-2.1 mm]), 1 1/2—2 lin. [=3.2-4.2 mm] lata, subcarnosa; superiora rhom- beo-deltoidea ....”" This description of the leaves as rhombic-deltoid with a length:width ratio of approximately two substantiates our observa- tions of images and specimens of this taxon. In summary, the significant disparities between C. carnosulum Mogq. and the taxon described here definitively rules out any possible identity be- tween the two. Given that the basionym for Wahl’s taxon Is C. patagonicum Phil., we investigated the features of that taxon in comparison to C. Jittoreum. The original description by Philippi (1895) of C. patagonicum reads: “‘foliis ... integerrimis, ovatis seu oblongo-triangularibus, basi sub truncates vel trapezoideis, interdum basi utrinque unidenta- tis...,” translated as ‘“‘the leaf is entire, ovate or oblong-triangular with base subtruncate, or [leaf] trapezoidal, sometimes basally one-toothed from both sides.’’ These characters are different from the narrowly elliptic to widely lanceolate (base cuneate) leaves of C. littoreum, which cannot be described as trapezoidal or subtruncate. The accompanying description in Spanish by Philippi just below the Latin one, says “‘su lamina 21 milimetros de lonjitud 1 [sic] 15 milimetros de anchura, pero la mayor parte de las hojas tienen la mitad de ese tamano ...” (“its blade 21 mm long by 15 mm wide but the majority of the leaves are closer to half of this size’’). The measurements of 21mm by 15 mm are inconsistent with the leaf length of C. carnosulum (ca. 6 to 8 mm) and are not those of an elliptic to widely lanceolate leaf either, as in the Californian C. /ittoreum. In the original description, the leaves of C. patagonicum (Philippi 1895) resemble those of C. carnosulum in shape, but are apparently larger in size. Additional evidence of the distinctiveness of C. littoreum comes from synonomy. Aellen (1929) and Aellen and Just (1943) combined three previously described Argentinian taxa - C. fuegianum Speg. (1896), C. patagonicum Phil. (1895), and C. scabricaule Speg. (1902) (the last having three varieties) with C. carnosulum Mog. (1849), which has nomenclatural priority. Thus, these authors considered C. patagonicum Phil. to 2010] BENET-PIERCE AND SIMPSON: CHENOPODIUM LITTOREUM SP. NOV. 67 ROYAL ROTANIC GARDFNG KEW MM KOODSE3166 “61° RE 'NYTIONINNDD ‘OH Wd TH au Cheep, Carano) whine ROTEL BOTANIC GARDENS KEW Vag en AION | 2 Pome, nto 1/889) K000593167 A Tyre « KEW NEGATIVE No 6104 ~ 8 JUL 1963 ! A Ch _ Canim Ma | FIG. 2. Chenopodium carnosulum Mog. A. Holotype (K 583167). Note relatively short, rhomboid to deltate, basally lobed leaves. BE. Specimen, Mexia 7960 (GH, s.n.). B. Close-up of shoot, showing similar, rhomboid leaves. C. Single leaf, showing rhomboid shape with two lateral lobes. Scale bar = Imm. D, E. Close-up of flower remains, showing calyx lobes distinct nearly to base and only two stamens (arrows). Scale bars = 1 mm. F. C. carnosulum (Chenopodium parryi Standl.) specimen C. Parry 780 (central Mexico, 1878, MO 46467, isotype). Note identical, rhomboid-trapezoid leaves. 68 MADRONO be the same taxon as C. carnosulum, which, as we already have shown, is quite distinct from C. littoreum. These authors presumably thought that any variation between these three species, in a genus well known for its lack of definite and stable leaf characters, was insufficient to warrant separate species status from C. carnosulum. Aellen, having revised the genus in the American continent, pointed out that the original collection of C. carnosulum did not come from California, as Moquin-Tandon had noted, but from Port Gregory, Patagonia. Aellen (1929) annotated the type specimen collected by O. Cunningham, the same that Moquin-Tandon had identified as the holotype of C. carnosulum Mog. (K 583167). Aellen was clear in his opinion of this: ““Moquin made a mistake when he stated, in the ‘Prodomus’, California as the native country of the original plant. The exemplary originates from Patagonia (Port Gregory). This lead to the fate of the species being sealed in the South American literature. North American botanists were certainly mystified by Ch. carnosulum Mog., as it couldn’t be found in California. S. Watson (I.c.) treated it as a ‘doubtful species.’ Standley (1.c.) mentioned it from Mount Orizaba, Mexico; yet the identification is not certain.” (Aellen 1929, translation by D. Pierce-Knies, personal commu- nication). In order to ascertain the presence of the South American C. carnosulum in North America, we studied other species that have been associated with C. carnosulum. One of them, C. parryi Standl., was for a time an accepted taxon. The type specimen from Mexico (MO 46467, C. Parry 780, central Mexico, 1878; Fig. 2F) shows a species with a trilobed leaf much lke C. carnosulum, described by Standley as “... leaf- blades triangular or triangular-rhombic in out- line, 3—5 mm. long, 3—4 mm. broad, 3-lobed, ...”’ (Standley 1916). Wahl (1965) also considered this species, stating ““The type (no other collection has been referred to it) fits in geographically with the other two Mexican records even if these were difficult to place with any ... C. Parryi Standley seems to be the same as C. carnosulum Mog. vat. carnosulum” (Wahl 1965). And we concur, as the type from MO (Fig. 2F) shows the same rhom- boid, basally lobed leaf as in C. carnosulum, evidently different from that of C. Jittoreum. Thus we confirmed the presence of C. carnosulum in North America, but not in the United States. H. A. Wahl, who revised the genus Chenopo- dium in North America (Wahl 1954, 1965) had recognized the California taxon as puzzling, citing several specimens from CAS that ‘‘when I examined them in 1955, could not be placed with any known North American species. These were from sand dunes or similar habitats along or near the coast in San Luis Obispo and Santa Barbara counties, California’? (Wahl 1965, p. 137). Wahl [Vol. 57 (1965) believed that the California specimens in question were C. patagonicum, which he then reduced in rank to C. carnosulum var. patagoni- cum. Wahl based his opinion solely on what he described as a photograph of the type of C. patagonicum, which he said “is such an exact match for the California plants as to leave no doubt as to their inclusion with this species” (Wahl, 1965, p. 138). As representatives of this taxon, Wahl cites one Chilean specimen (Bauch- tien s.n., in part, Feb. 1903, US; this specimen not listed on the US database); two Mexican specimens (Seaton 184, 6 Aug. 1891, GH; this specimen not listed on the Harvard University Herbaria database; Balis B5503, 22 Sept. 1938, UC), and several California specimens (Eastwood 789, 2 July 1906, CAS; Hoffmann s.n., 29 March 1939, CAS; Condit s.n., 30 April 1910, UC; Hoffmann 420, 30 May 1931, CAS; and L. S. Rose 50116, 13 May 1950, CAS, UC). However, his conclusions are puzzling, given the disparity in leaf morphology (let alone stamen number) between C. /ittoreum and C. carnosulum. We have not seen the specific Chilean specimens he mentioned, but we have examined other speci- mens of C. carnosulum. Having seen all of the same specimens of California collections, we firmly believe they do not correspond to C. carnosulum. Wahl, however, treated the Califor- nia taxon as a variety of C. carnosulum, presumably on account of the differences he observed and because C. patagonicum had already been treated as a synonym of the former by Aellen (1929) and Aellen and Just (1943). We have been unable to physically examine specimens of C. patagonicum Phil., but we have now seen an image of the type (SGO 38811; Fig. 3). The type specimen does look similar to C. littoreum in leaf morphology in that some leaves are narrowly elliptic to widely lanceolate. How- ever, most leaves, in particular the mature ones, are “‘trullate’’ in appearance, 1.e., rhombic with a more elongate upper half, with two, small lobes near the base, and a mostly rounded apex (Fig. 3B). Thus, leaf morphology of C. patagoni- cum 1s somewhat different from that of C. littoreum, and intermediate to that of a typical C. carnosulum (Fig. 2). It 1s plausible that it was the picture of this plant, identified as C. patagonicum Phil., that convinced Wahl that the Californian plants were equivalent, introduced from South America. From the SGO 38811 image of the C. patagonicum type, we noted that this specimen had been annotated as C. philippianum (A. Marticorena, annotated 2000; Fig. 3C). In addi- tion, C. patagonicum has been treated as a synonym of C. philippianum in at least one recent treatment (Marticorena 2008). If indeed these two taxa are equivalent, we do not understand why C. patagonicum Phil. (1895) would not have MUSEO NAOONAL DE HS TORE NATURAL CRE $G0000001638 o o a] g rad eo a” 2 PS > c Pa 2 3 3 NAL ; hearer Pe RBARISCND Qt 038814 ii] J mS CHILE E Sd TOR eZ ETIQUETA DE IDENTIFICACION 4 ? ty } isd = , , enapep tian fie deg end euiaas G L Memapped tien ftbigen ope Vee BENET-PIERCE AND SIMPSON: CHENOPODIUM LITTOREUM SP. NOV. 69 mature leaves trullate, basally lobed (Glee) % : * y indice pe i is ous. Le; Ye CLE PL fet a lel de : a Li : : AebecopeoD teens A ae Wa zl bi Wes Li lean ce sone Ley HF Stl c HERBARIO DEL MUSEO NACIONAL DE MISTORIA NATURAL, CHILE. ETIQUETA DE yp. TIFICACION (banopodiuu. as i pitas Hue, Det. Uarticortnra 2000 _ ETIQUETA DE CORRECCION. faery Ly /. peep “4 eee Latha tat =e / / . - C (: penpfptiaure UA a porictcrr~ th Mate ial prontada bajo pe QUSPIEIOS UE oh Ofte mea ne Coordinador de Aa f / é Asuntos [nieramericanos, Ministerfa de Agritultura y det, C. Munoz P. K-/PVF Compania Manufacturera de Papeles y Cartones. FIG. 3. Chenopodium patagonicum Phil. Type specimen (SGO 38811). A. Whole herbarium sheet, B. Close up of larger plant (at left on sheet). Note leaves varying from narrowly elliptic to widely trullate, with two, small lobes near base. Scale bar = 1 cm. C. Close-up of herbarium labels. Note original designation as C. patagonicum Phil., annotated as Chenopodium philippianum Aellen by A. Marticorena (2000). nomenclatural priority over C. philippianum Aellen (1929). This discrepancy we hope to _ address in a later study in conjunction with our _ Chilean colleagues at SGO. Because C. philippianum looks superficially similar to C. Jittoreum, and indications are it may be equivalent to C. patagonicum, it was particularly important to thoroughly investigate the former from specimens. We have physically examined C. philippianum (GH 257649; Fig. 4A— C), the same specimen Wahl had also examined and which he had determined to be different from the California collections. We found the leaves to resemble C. patagonicum, being generally rhom- boid and lobed, although they are much larger and with lobes much less pronounced than C. 70 MADRONO NLPED STATES DSPARTMRNT OF AGRICULTURE UN URAY HEADARIUM OF HARVARD UNIVESSITY NEW VOKK SOTANICAL QAKDRN EXPLONATIONS IM BOUTH AMERIOA eps}. i ead A 6 fhe ven parla ae exrucbeaiuans Hey Ausoltion Label | ‘ Wrearavth sea! craw, Anton ‘ < aga HN MT ‘ es | : FIG. 4. [Vol. 57 Chenopodium philippianum Aellen. A-C. Specimen GH 257649. A. Herbarium sheet. B, C. Leaves, showing somewhat trullate to widely lanceolate shape, with slight lobbing near base. Scale bar = 1 mm. D, E. Specimen GH 21730. Scale bars = 1 mm. D. Fruit, showing calyx fused (arrow) more than halfway to apex. E. Fruit, showing remains of two stamens (arrows). carnosulum. We have been able to ascertain that the leaf apices are rounded to obtuse and generally not mucronulate, which is often the case in C. /ittoreum. In the two type specimens of C. philippianum (K 583181 and K 58382, both images available on line), the leaves are even more strongly lobed than the specimen we physically examined, but they probably represent more mature plants. C. philippianum also has a variable number of stamens (mostly 2—3, occa- sionally 5) (GH 21730; Fig. 4E). In addition and perhaps more significantly, the sepals of C. philippianum are fused to half or more than half of their length (Fig. 4D), whereas in C. /ittoreum the calyx is fused well less than half its length (Fig. 1D), calyx fusion being somewhat useful diagnostically in Chenopodium. Thus, we can rule out this species being the same as the Californian taxon on the basis of the leaf shape and apex, calyx fusion, and stamen number (Fig. 4). In general, though, this species does show stronger similarities with C. Jittoreum than do any taxa of the C. carnosulum complex, and future molecular work could better elucidate their relationship. To further explore the C. patagonicum type, we asked the curator of SGO in Santiago, Chile, to examine the type of C. patagonicum (SGO 38811; Fig. 3). Dr. M. Munoz reported the 2010] specimen having 2 and 5 stamens and a calyx fused to around the middle (personal communi- cation). These findings would support the con- sideration that C. patagonicum Phil. and C. philippianum Aellen are the same species. We also reviewed the diagnosis of C. philippianum by Aellen (1929). Aellen had problems identifying the material from which he diagnosed this species: “The labeling of the Philippianum material is extremely difficult. To approximate the species is only indirectly possible. The Washington original material of Cordillera de Talca is a very incomplete, small specimen, which can’t be accurately identified; the one from Berlin is a little more complete, but does not feature any fully developed seeds Philippi, seemingly, never published his Ch. Andinum ...” (translation by D. Pierce-Knies, personal communication). It is plausible that Aellen (1929) described C. philippianum as a new species (even given the poor material he had seen), unaware that it was equivalent to C. patagonicum. In the past, he had incorrectly accepted C. patagonicum to be a synonym of C. carnosulum even if he had done this while issuing a warning that the synonomy of C. carnosulum could be in doubt: ““Assumedly, it [C. carnosulum] was newly characterized by Phlilippi or Spegazzini; it still needs to be established with certainty whether it is the same as Ch. patagonicum Phil. or Ch. fuegianum Speg. or Ch. Scabricaule Speg.” (Aellen 1929, transla- tion by D. Pierce-Knies, personal communica- tion). We have recently seen an image of C. fuegianum (SGO 59002), which is now identified as C. carnosulum var carnosulum, C. carnosulum having priority over C. fuegianum. Aellen’s concerns also give further credence that these two species, C. philippianum and C. patagonicum, could be the same. On the other hand, when Wahl examined the California collections, specimens that had been sent to Wahl by R.F. Hoover from San Luis Obispo, the notion that C. /ittoreum could be a new species did occur to him. He wrote (Wahl 1965): ““The possibility of these representing an undescribed species was considered but the known occurrence on the west coast of varieties of species native in the drier and colder parts of southern and western South America [C. macro- —spermum Hook. f. var. farinosum (Wats.) J. T. Howell, C. chenopodioides (L.) Aellen var. Degenianum (Aellen) Aellen and var. Lengyelia- num (Aellen) Aellen] suggested a possible similar _ relationship for these relatively restricted plants.”’ Wahl never confirmed this relationship. We have _ been able to determine that the above naturalized | | _ Chenopodium species for the most part have vertical seeds, and probably are not comparable at all to C. littoreum; they were presumably cited as an analogy, indicating that because other South American species have become established BENET-PIERCE AND SIMPSON: CHENOPODIUM LITTOREUM SP. NOV. 71 in California, what we are calling C. /ittoreum could have been as well. Thus, although it was presumably a picture of the type of C. patagonicum that convinced Wahl of its equivalence to what we are describing as C. littoreum, we can only rely on the facts: 1) that C. patagonicum 1s described as having “ovate or oblong-triangular with base subtruncate, or [leaf] trapezoidal, sometimes basally one-toothed from both sides, 21 mm long by 15 mm wide” in the protologue (Philippi 1895), agreeing more with the leaf shape of C. carnosulum and C. philippia- num but not with C. /ittoreum; 2) that the type of C. patagonicum shows differences in leaf mor- phology from C. /ittoreum in the former being trullate in shape with basal lobes; 3) that C. patagonicum has been considered a synonym of C. carnosulum by some authors (Aellen 1929; Aellen and Just 1943), a taxon quite different from C. littoreum; and 4) that C. patagonicum 1s apparently equivalent to C. philippianum, a taxon that we have been able to show differs from C. littoreum in having stamen number 2-3 or occasionally 5, a more extensive sepal fusion, and differences in leaf morphology. Therefore, we do not believe that C. patagonicum Phil., nor by extension C. carnosulum Mog. var. patagonicum (Phil.) Wahl, nor C. philippianum Aellen are the same taxon as C. /ittoreum. No other South American taxa that we know have been associ- ated at any point with these species’ characteris- tics. We have thoroughly reviewed every species at one time associated with C. carnosulum and C. patagonicum. We have reviewed Chilean (Marti- corena, 2008; Reiche, 1911) and Argentinean (Toloaba, 2006) keys to Chenopodium and have found no other species that would fit the description of C. /ittoreum. In particular, it is the highly restricted range of C. /ittoreum, in the absence of any other likely candidate in Cheno- podium keys for South America, Baja California (Wiggins, 1980), or neighboring North American states (Clemants and Mosyakin 2003a), plus its differences with the above-mentioned species, that supports the conclusion that it 1s endemic, particularly in a region well known for dune endemic vegetation (D. Keil California Polytech- nic State Univ., personal communication). In conclusion, the Californian Chenopodium littoreum described here does not conform to any of the South American taxa that have been associated with it nor to any other we have separately considered, and its narrow range makes it unlikely that it should be. Chenopodium littoreum 1s also unlike any other North Ameri- can species in the genus. Although it shares some characters with other Chenopodium species found here, with the usual horizontal seed and five perianth parts, none of these taxa is prostrate. Other Chenopodium species in North America that are either prostrate or somewhat decumbent 72 MADRONO have vertical or vertical and horizontal seeds and have usually one or two stamens, or other differing vegetative or floral characters. We end with this quote from Wahl (1954): ““No group of plants of comparable size and wide distribution known to the writer has suffered the lack of understanding of the taxa involved as has the genus Chenopodium ... The reasons for this lie in (1) the ecological variability characteristic of weedy annuals, (2) the fact that important diagnostic characters are present in the seeds, which are of small size and often lacking from collected material, (3) the repetition of similar variations in habit and leaf shape in distinct species and (4) the lack of pubescence characters in most species.”’ The convergence of these factors probably contributed to the confusion that has surrounded C. /ittoreum, this new Californian species, to this day. We are hopeful that future molecular work will clarify some of the confusion in this complex and lead to further elucidation of the relationships among South and North American taxa. ACKNOWLEDGMENTS Our sincere thanks to Drs. D. Keil, A. Marticorena, M. Munoz for their expert advice, to Drs. D. Trock and E. Zacharias for their generous assistance with locating needed specimens, and the following herbaria for allowing us to examine material or images CAS-DS, GH, JEPS, OBI, RSA-POM, SGO, and UC. Thanks to E. N. Genovese, Professor Emeritus of Classics and Humanities at SDSU, for his help on the name of the name of this new species and to Lee M. Simpson for his help with the images. We specially want to thank our reviewers for their help and willingness to work under a very tight deadline. LITERATURE CITED AELLEN, P. 1929. Beitrag zur Systematik der Chenopo- dium —Arten-Amerikas, vorwiegend auf Grund der Sammlung des United States National Museums in Washington, D.C. I, Hl. Feddes Repertorium Specierum Novarum Regni Vegatabilis 26:31—64, 119-160. AND T. JUST. 1943. Key and synopsis of the American species of the genus Chenopodium L. American Midland Naturalist 30:47—76. APG III. 2009. An update of the Angiosperm Phylogeny Group classification for the orders and [Vol. 57 families of flowering plants: APG III. Botanical Journal of the Linnean Society 161:105—121. CLEMANTS, S. E. AND S. L. MOSYAKIN. 2003a. Chenopodium. Pp. 273-299 in Flora of North America Editorial Committee (eds.), Flora of North America north of Mexico, Vol. 4. Oxford University Press, NY. AND . 2003b. Dysphania. Pp. 267-273 in Flora of North America Editorial Committee (eds.), Flora of North America North of Mexico, Vol. 4. Oxford University Press, New York, NY. MABBERLEY, D. J. 2008. Mabberley’s plant-book: a portable dictionary of the higher plants, their classification and uses, 3rd ed. Cambridge Univer- sity Press, Cambridge, U.K. MARTICORENA, A. 2008. Clave Para la Identificaci6n de las Especies de Chenopodium en Chile. Website http://www.chlorischile.cl/chenopodium/ chenopodium.htm [accessed 10 June 2010]. MOQUIN-TANDON, C. H. B. A. 1849. Chenopodium carnosulum. In Augustin Pyramus de Candolle. Prodromus Systematis Naturalis Regni Vegetabilis 13(2):64. PHILIPPI, R. A. 1895. Plantas Nuevas Chilenas. Anales de la Universidad de Chile 91:419. REICHE, K. 1911. Estudios criticos de la Flora de Chile. Anales de Universidad de Chile 6:148—159. Avail- able at: http://www.biodiversitylibrary.org/item/ 10736#1 [accessed 15 June 2010]. STANDLEY, P. C. 1916. Chenopodium. Pp. 1-93 in North American Flora, Vol. 21. The New York Botanical Garden, NY. TOLOABA, J. A. 2006. Chenopodiaceae Vent. Aportes Botanicos de Salta 7:18, Herbario del Museo de Ciencias Naturales de Salta (MCNS), Buenos Aires, Argentina. THE FLORA OF BAJA CALIFORNIA. 2009, onwards. San Diego Natural History Museum, San Diego, CA. Website http://bajaflora.org [accessed 10 June 2010]. WAHL, H. A. 1954. A preliminary study of the genus Chenopodium in North America. Bartonia 27: 1-46. 1965. Chenopodium carnosulum and some related taxa in North and South America. Leaflets of Western Botany 10:137—156. WIGGINS, I. L. 1980. Flora of Baja California. Stanford University Press, Stanford, CA. WILKEN, D. H. 1993. Chenopodium. Pp. 506-511, 513 in J. C. Hickman (ed.), The Jepson manual: higher plants of California. University of California Press, Berkeley, CA. MADRONO, Vol. 57, No. 1, p. 73, 2010 REVIEW Desert WisdomlAgaves and Cacti: CO>, Water, Climate Change. By PARK S. NOBEL. 2010. iUniverse, Inc., New York, NY and Blooming- ton, IN. 182 pp. ISBN 978-1-4401-9151-0 $16.95, soft cover. ISBN 978-1-4401-9152-7 $6.00, eBook. Park Nobel is well known to plant biologists interested in plant-environment interactions. From 1979 through 2009, the ISI Web of Science lists 245 peer-reviewed articles he has authored or coauthored and almost all involve the physiology of agaves and cacti. In addition, Nobel is author or editor of four books dealing with agaves and or cacti and is also the author of a unique textbook on environmental plant physiology (Nobel, 2009). So why this new book, Desert Wisdom? This book is similar to Nobel’s other books in that it draws on a wide range of research that has been conducted on plants in general and agaves and cacti in particular. Desert Wisdom differs from its predecessors in that it is written for a broader audience and it takes a position of advocacy for planting agaves and cacti in locations around the globe that are predicted to become hotter and drier. The writing style in this book is less formal than is typical for Nobel but he does not abandon the quantitative perspective that attracts many readers to his work. Desert Wisdom contains seven chapters. The first chapter stands out from the others in that it focuses on commercial uses of agaves and cacti rather than their environmental physiology. This is interesting reading, particularly if you know little about the historical use of agaves and cacti or their present-day economic importance. Most species of agaves and cacti possess the unusual photosynthetic pathway known as Crassulacean Acid Metabolism (CAM). Chapter two describes CAM’s biochemical features and how the CAM pathway improves water conservation. Chapter three explains how well many agaves and cacti tolerate drought and temperature extremes. Chapter four, titled “‘Issues of Global Climate Change’, reviews and defines the problems plants as well as humans will be facing in the future. In the first 16 pages of chapter four, Nobel describes the basis and scope of the problem. He summa- rizes the main conclusions that can be derived from global change models, particularly with regards to plants. The remainder of chapter four describes how plants and particularly agaves and cacti should be able to adapt to changing climate. Nobel uses a systematic, direct approach to analyze what global change models can tell us. Besides being important to the thesis of this book, I think many readers will find this concise, non alarmist presentation a very useful Overview of climate change. In chapter five Nobel explains an Environmental Productivity Index (EPI) that he has developed and which can be used to predict the effect of different environmental factors on net CO, uptake. Separate indices for plant response to light, temperature, water availability, nutrient availability, and CO, concentration are determined and then these indices are multiplied together. The resulting number or EPI is used to interpret the growth of agaves and cacti in the field. Chapter six is an exploration of plant productivity based on the EPI. The overall ideas in chapters five and six are straight forward to grasp but derivation of the individual indices is not so clear. It took a review of some of the original research citations for this reviewer to understand what must be measured to determine the water availability index, for example. The final chapter summarizes how agaves and cacti should be important players in man’s response to climate change. Nobel argues the high productiv- ity of agaves and cacti in hot and dry conditions makes these plants ideal for combating desertifi- cation. Agaves and cacti may serve a more direct economic role since they can serve as carbon sinks and provide carbon credits or the related carbon offsets. Agaves and cacti could also be utilized as fodder for livestock or as stocks for biofuels. Nobel adds clear definitions in the text to minimize jargon and includes a_ glossary of important terms. I did find the organization of references into separate topics cumbersome. The separate lists are meant to aid those that want to read further about a specific topic. However, checking citations while reading the text is confusing since many references could fit into more than one grouping. I note Desert Wisdom is a bargain, listing for $16.95 for the bound copy and $6.00 for an eBook version (see Nobel’s website for information: www.eeb.ucla.edu/nobel). The interesting material, logical arguments, and direct writing style make this book an interesting read. Interested non-scientists as well as scientists with wide ranging backgrounds should enjoy and find something new in Desert Wisdom. —DaAvip J. LONGSTRETH, Department of Biological Sciences, Louisiana State University, 202 Life Sciences Building, Baton Rouge, LA 70803. btlong@lsu.edu. LITERATURE CITED NOBEL P. S. 2009. Physiochemical and environmental plant physiology, 4th ed. Academic Press, Oxford, UK. MADRONO, Vol. 57, No. 1, p. 74, 2010 NOTEWORTHY COLLECTION ARIZONA PUNICA GRANATUM L. (LYTHRACEAE).—Graham Co., river sand dune of Gila River, N of Deadman Canyon, 32.897717°N, —109.467783°W, riparian, a few local shrubs with red flowers, associated species include Tamarix ramosissima, Hymenoclea monogyra, Lappula occidentalis, Baccharis salicifolia, Mentzelia veatchiana, Prosopis velutina, Baccharis sarothroides, Tripterocalyx wootonii, Eriogonum trichopes, Allionia incarnata, Sola- num elaeagnifolium, Senecio flaccidus, Mentzelia multi- flora, Calycoseris wrightii, Hordeum murinum glaucum, Stephanomeria exigua, Ipomopsis longiflora, 5 May 2004, Wendy Hodgson 17923 and Dixie Damrel (DES). Pinal Co., near Dudleyville, floodplain east side of San Pedro River, ~350 meters from river channel under large Salix, 32.926611°N, —110.733278°W, elev. 649 me- ters, mixed Populus fremontii, Salix gooddingii, Ta- marix, Prosopis velutina river terrace community, other associated species include Chloracantha — spinosa, Sonchus asper, Sisymbrium irio, Bromus diandrus, Bromus madritensis ssp. rubens, B. catharticus, Clematis drummondii, Conyza canadensis, Hordeum murinum ssp. glaucum, Matelea producta, Hedosyne ambrosiifolia, Nicotiana glauca, Silybum marianum, Rumex sp., two plants seen, one 1.5 meters tall, the other 0.5 meters tall, flowering and fruiting. 19 Jun 2008, Michael Denslow 2587 and Elizabeth Ray (BOON, ASU). Previous knowledge. Pomegranate is native to western Asia and has been cultivated since antiquity (Davidson 1999). It is reported as introduced in six states in the southern United States including California and Utah (USDA, NRCS 2009). The plant has not previously been reported outside cultivation from Arizona (Shreve and Wiggins 1964; Kearney and Peebles 1969; Lehr 1978; Anonymous 2009; USDA, NRCS 2009). It was likely first introduced into Arizona as a fruit crop though non-fruiting cultivars are also available. Early settlers planted pomegranates near springs (e.g., Quito- baquito in Organ Pipe Cactus National Monument, Bowers 1980). It is widely cultivated and is a common component of many cultivated landscapes today. Significance. The specimens cited here are from riparian areas and do not appear to be individuals that are persisting from cultivation. The sites were not within human settlements and showed no signs of anthropogenic disturbance. The plants were found to be flowering and fruiting. These are the first reports of this shrub established outside cultivation in the flora of Arizona. Based on these records this species should be sought elsewhere in riparian habitats in Arizona. —MICHAEL W. DENSLOW, BOON Herbarium, P.O. Box 32027, Appalachian State University, Boone, NC 28608-2027; GABRIELLE KATZ, Department of Geog- raphy and Planning, P.O. Box 32066, Appalachian State University, Boone, NC 28608-2027; and WENDY HODGSON, Desert Botanical Garden, 1201 N. Galvin Parkway, Phoenix, AZ 85008. md68135@appstate.edu. LITERATURE CITED ANONYMOUS. 2009. Flora of Arizona: compilation of vascular plants of Arizona Project, Flora of North America, and information found Arizona collection herbaria. Southwest Environmental Information Network (SEINet). Website http:// swbiodiversity.org/seinet/checklists/checklist.php?cl =] [accessed 21 December 2009]. Bowers, J. E. 1980. Flora of Organ Pipe Cactus National Monument. Journal of the Arizona- Nevada Academy of Science 15:33-47. DAVIDSON, A. 1999. The Oxford companion to food. Oxford University Press, Oxford, UK. KEARNEY, T. H. AND R. H. PEEBLES. 1969. Arizona flora: 2nd edition with supplement. University of California Press, Berkeley, CA. LEHR, J. H. 1978. A catalogue of the flora of Arizona. Desert Botanical Garden, Phoenix, AZ. SHREVE, F. AND I. L. WIGGINS. 1964. Vegetation and flora of the Sonoran Desert. Stanford University Press, Stanford, CA. USDA, NRCS. 2009. The PLANTS Database, Na- tional Plant Data Center, Baton Rouge, LA. Website http://plants.usda.gov [accessed 21 Decem- ber 2009]. MADRONO, Vol. 57, No. I, p. 75, 2010 ERRATUM In the Note by Malcom and Radke (2008), the incorrect family is provided for Lilaeopsis schaff- neriana var. recurva. The correct family is the Apiaceae (e.g., USDA, NRCS 2010). LITERATURE CITED MALcoM, J. W. AND W. R. RADKE. 2008. Livestock trampling and Lilaeopsis schaffneriana var. recurva (Brassicaceae). Madrono 55:81. USDA, NRCS. 2010. The PLANTS Database National Plant Data Center, Baton Rouge, LA. Website http://plants.usda.gov/ [accessed 24 June 2010]. Volume 57, Number 1, pages 1—76, published 31 August 2010 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. SMITHSONIAN INSTITUTION LIBRARIES AQUI 558 1317