UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A biosystematic study of the Lasthenia californica complex (Asteraceae) Desrochers, Andrée M. 1992

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1993_spring_phd_desrochers_andree.pdf [ 6.36MB ]
Metadata
JSON: 831-1.0086380.json
JSON-LD: 831-1.0086380-ld.json
RDF/XML (Pretty): 831-1.0086380-rdf.xml
RDF/JSON: 831-1.0086380-rdf.json
Turtle: 831-1.0086380-turtle.txt
N-Triples: 831-1.0086380-rdf-ntriples.txt
Original Record: 831-1.0086380-source.json
Full Text
831-1.0086380-fulltext.txt
Citation
831-1.0086380.ris

Full Text

A BIOSYSTEMATIC STUDY OF THELasthenia californica COMPLEX (ASTERACEAE)byANDREE MANON DESROCHERSB.Sc., McGill University, 1985M.Sc., Macdonald College ofMcGill University, 1987A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIESDepartment of Botany, University of British ColumbiaWe accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIADecember, 1992© Andree Manon DesrochersIn presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)Department of 7 oTA ki yThe University of British ColumbiaVancouver, CanadaDateDE-6 (2/88)ABSTRACTLasthenia californica DC. ex Lindley is an annual,presumably obligate outcrossing species and is the mostwidespread taxon of the 17 species in this mostlyCalifornian genus. A biosystematic study of this complex hasshown that this species is polymorphic in its flavonoidchemistry, electrophoretic banding patterns, and morphologyat the intra- and interpopulation levels. A detailedanalysis of a population located in the Biological Preserveof Stanford University (Jasper Ridge) showed that there is astrong correlation among flavonoid profiles, pappusmorphology and isozyme patterns. A long term study at JasperRidge suggests that the spatial distribution of pigmenttypes along several transects has remained essentiallyconstant over a period of 10 years. Indices of geneticdifferentiation in populations and preliminary findings ofbreeding experiments also suggest population subdivision andlow levels of crossability between plants with differentflavonoid profiles. It appears that these forms flower atdifferent times. The pattern of variation observed at JasperRidge suggests that the two genotypes represent a northernrace and a southern race. It is premature to concludewhether the population structure at Jasper Ridge is uniqueto this site.Variation in morphology, flavonoid chemistry, andisozymes failed to group the populations of L. californica into recognizable taxonomic categories. However, thecharacter states observed (i.e., in morphology, flavonoids,and isozymes) at Jasper Ridge show a polarized distribution(north-south) in the other 34 populations examined acrossthe species' range and represent geographical races. Thereis also a positive correlation among the data sets in somepopulations although this correlation is different than theone observed at Jasper Ridge.Ecological races - one coastal and one inland - werealso observed. The formation of various races in the L.californica complex is believed to represent a step towardspeciation and it seems that the geographical speciationmodel best reflects the patterns of variation observed.Diploid, tetraploid, and hexaploid populations are foundthroughout the range of the species and areindistinguishable geographically, morphologically,chemically or by their allozymes. The isozyme study suggeststhat the tetraploid populations are of an autopolyploidorigin. The only hexaploid population sampled was notanalysed for its isozyme variation.iiiTABLE OF CONTENTSAbstract^ iiList of Tables^List of Figures viiAcknowledgement^ ixIntroduction 1Literature review on L. californica. ^ 10Geographical and ecological races 10Cytological races^ 10Chemical races 13Evolution^ 14Purpose of this research^ 17Materials and methods 20Population sampling^ 20Sampling at Jasper Ridge 20Morphological analysis 24Pollen viability and size^ 27Statistical analysis 28Chromosome counts^ 30Flavonoid analysis 30Isozyme analysis 31Crossing experiments 36Results^ 38Results obtained for the Jasper Ridge population^ 38Results obtained for all populations^ 47Flavonoids^ 47Chromosome counts^ 52Pollen 52Morphology 55Isozymes^ 64Crossing experiments^ 76Discussion 82General patterns of variation^ 82Differentiation in the Jasper Ridge population^ 84Population differentiation 87Nature and origin of polyploid populations^ 90Evolution^ 93Summary and conclusion^ 96References cited 98ivLIST OF TABLES1- Locations of the Lasthenia californica populationscollected in 1988, 1989, and 1990^ 212- Analyses performed on each population of Lasthenia californica^ 233- Morphological characters measured on the populationsof Lasthenia californica^ 264- Flavonoid profiles occurring in the counties surveyedusing herbarium specimens of Lasthenia californica....325- Flavonoid profiles occurring in each population ofLasthenia californica^ 396- Genetic variation at isozyme loci for 33 populationsof Lasthenia californica 457- Genetic diversity statistics for each group^ 468- Distribution of flavonoids in the Lasthenia californica pigment profile types^ 489- Means for pollen viability and pollen grain size foreach population examined^ 5310- One-wy nested analyses of variance for pollendiameter and pollen stainability from three polyploidlevels where eight and 17 populations, respectively,were included^ 5411- Results obtained for the qualitative charactersrecorded on population samples ofLasthenia californica^ 5612- Means and F-values from analysis of variance for the11 quantitative characters included in themorphometric analyses of Lasthenia californica^ 5713- Allelic frequencies for each population of Lasthenia californica^ 6514- Number of alleles observed and shared by the cytotypesof Lasthenia californica^ 6615- Gene diversity statistics for each of the sevenpolymorphic loci over all populations^ 7416- Mean genetic identities and ranges of identities forpair-wise comparisons of populations for Lasthenia californica^ 77vi17- Genetic identities between the groups^ 7818- Results obtained from the reciprocal crossing andselfing experiments^ 80LIST OF FIGURES1- Geographical distribution of Lasthenia californica.... 32- Diagrams showing some of the morphological variationamong three races of Lasthenia californica describedby Ornduff^ 113- Distribution of the 36 populations included inthe study of Lasthenia californica^ 224- Map of transects in the Jasper Ridge population^ 255- Flavonoid and pappus types along the four transects inJasper Ridge for 1989^ 406- Flavonoid and pappus types along the four transects inJasper Ridge for 1990 417- Photographs of starch gels showing electrophoreticpatterns of plants exhibiting different flavonoidprofiles in the Jasper Ridge population^ 438- Geographical distribution of flavonoid profiletypes for Lasthenia californica^ 519- Distribution of the canonical means against thefirst two axes for 34 populations based on 20morphological characters^ 5810- Distribution of the canonical means against thefirst two axes for 34 populations based on 9qualitative characters^ 5911- Geographical distribution of pappus types forLasthenia californica 6212- Diagram showing pappus variation in Lastheniacalifornica^ 6313- Photographs of starch gels showing electrophoreticpatterns of diploid and tetraploid plants inLasthenia californica for Pgi-2 and Acon-2^ 6814- Geographical distribution of allelic frequencies atNadhdh for Lasthenia californica^ 6915- Geographical distribution of allelic frequencies at6Pgd-1 for Lasthenia californica 7016- Photographs of starch gels representing intra- andinterpopulation variation in Lasthenia californica forLAP, SKDH, and PGM^ 72viiviii17- Cluster analysis (UPGMA) of 33 populations ofLasthenia californica, including the Jasper Ridgepopulation^ 79ACKNOWLEDGEMENTI would, first of all, like to give heartfelt thanks toDr. B.A. Bohm whose advice and assistance were invaluableduring the course of this study. I thank him for his patientguidance and constructive comments provided during thepreparation of this manuscript. His skill as a supervisor,and his financial support facilitated my time and work atthe University of British Columbia, and made the conclusionof this research project possible.I would also like to express my gratitude to the restof my committee, Drs G.E. Bradfield, F.R. Ganders. and G.B.Straley, whose sound advice and interest shown in my workwere greatly appreciated. I am especially grateful to Dr.F.R. Ganders who generously provided space for the isozymeanalysis in his already busy and crowded laboratory.The administration and staff of Jasper Ridge BiologicalPreserve deserve special thanks for making the facilities atthe Preserve available for this research. As well, the useof specimens from the following herbaria was muchappreciated: JEPS, UC, CAS, OSUF, and LA.Sincere thanks to Terry T. McIntosh whose supportduring the course of this study and his field assistancemade this research considerably easier.I also wish to extend my thanks to several of my fellowstudents, P. Ann Eastman, Meg Stookey, Alan R. Reid, GarthBrown, and Jeff Glaubitz for their good humor, helpfulness,and interest in theoretical discussion.Finally, I sincerely thank my parents, George andRgjeanne Desrochers, for their continued understanding, loveand support throughout my academic career.ixINTRODUCTIONThe Flora of California is well known for its speciesrichness and its wealth of endemics. Its diversity is theresult of the presence of a relatively large number ofancient, relic species and of the active speciation whichhas occurred in recent time (Stebbins and Major, 1965).Raven (1977) proposed that it is the newly evolved speciesthat mostly contributed to the size of the flora and theoverall proportion of endemism in California. Both therelic and newly evolved species are important parts of theflora of California. They represent, however, two extremesituations that can easily be distinguished based on theecological, geological, and historical characteristics ofthe areas where these taxa are found. Many species endemicto California and adjacent areas vary in a patternassociated with the complex physiography and geologicalhistory of the region and do not fall in either of thesecategories, i.e., as relic or new taxa. Such species ofteninclude populations or races that might either be distinctfor one or a few characters or are intermediate between wideranging entities that differ by a suite of characters.Lasthenia californica DC. ex Lindley represents such anexample, a species endemic to the California FloristicProvince. This region includes southwestern Oregon andportions of northern Baja California but excludes some ofthe southeastern desert regions of California as well aspart of the state east of the Sierra-Cascade axis (Howell,11957). The geographical distribution of L. californica covers most of the California Floristic Province, portionsof the southeastern deserts of California in addition tosouthern Arizona (Fig. 1). The climatic regime of most ofthe area occupied by the species is Mediterranean with long,dry summers and short, mild, wet winters. The regioncontains a large number of annual species. The level ofendemism at the species level, in California, isapproximately 30% (Raven and Axelrod, 1978). In addition,the area is characterized by a substantial topographic andedaphic diversity (Oakeshott, 1971).Lasthenia californica belongs to a relatively smallgenus that includes 17 species, 16 of which are restrictedto the Pacific coast of North America and one of which isknown only from Central Chile (Ornduff, 1966; Vasey, 1985).The majority of the species are endemic to the CaliforniaFloristic Province (Ornduff, 1966). Lasthenia is a memberof the subtribe Baerinae of the Heliantheae and is unique inthe subtribe owing to the presence of radial thickenings onendothecial cells (Robinson, 1981).The present circumscription of the genus includesBaeria, Crockeria, and Lasthenia s.s. (Hall, 1914, 1915;Jepson, 1925; Keck, 1959; Ferris, 1960) which, according toOrnduff (1966), are best treated as congeneric taxa. In hismonograph on the genus, Ornduff (1966) accommodated thelarge degree of interspecific variation in Lasthenia s.l. bysubdividing the genus into six sections. Most members of2Figure 1- Geographical distribution of Lasthenia californica. n=haploid chromosome number. (AfterOrnduff (1966).)3the genus are easily distinguished by small but constantmorphological characters. However, the large degree ofintraspecific variation found in some species (e.g., L.californica) or groups of species (e.g., L. fremontii, L.burkei and L. conjugens) renders the delimitation of thosetaxa more difficult.Interspecific variation in the genus is seen inmorphology, habitat or ecological tolerances, chromosomenumbers and breeding systems. Morphological variation isobserved in characters such as plant size, leaf shape,nature of pubescence, and floral and fruit features.Chromosome numbers in the genus include a wide array wheren=4, 5, 6, 7, 8, 12, 16, and 24, and are believed to be theresult of aneuploid reductions and polyploid increases(Ornduff, 1966). Most species in the genus are isolated bybarriers to crossing, by strong interspecific sterility orboth. Only four species in the genus are known to be self-compatible and largely self-pollinated, the others areobligate outbreeders. Ecological tolerances vary. Somespecies are restricted to extreme habitats such as seabird-breeding sites (e.g., L. maritima) (Vasey, 1985) or tohighly saline soils of low valley areas (e.g., L.chrysantha). In contrast, L. californica occurs over a widerange of habitats including coastal bluffs, inland valleys,and desert areas.Another well-studied aspect of the genus is itsflavonoid constitution. More than 20 flavonoid glycosides4are known to occur within Lasthenia (Bohm et al., 1974,Saleh et al., 1971). Although most species can becharacterized by their flavonoid constitution, both inter-and intrapopulation variation occur (Bohm and Banek, 1987;Ornduff et al., 1974). Studies of artificial hybridsindicate that inheritance pattern among classes offlavonoids may vary in several ways (Ornduff et al., 1973).Hybrids may either be lacking compounds of one parent or mayshow additive patterns or produce novel compounds. Ingeneral, the results obtained in these studies support thetaxonomic treatment proposed by Ornduff (1966) and in somecases even helped to clarify the position of problematictaxa within the genus (Bohm et al., 1974).Ornduff's monograph and the chemical studies helped toelucidate taxa relationships within Lasthenia. Otherstudies of the genus have concentrated on groups of closelyrelated species. The taxonomic status of the progenitor-derivative species, L. minor (DC) Ornduff and L. maritima (Gray) M. Vasey, was reviewed by Vasey (1985) who usedmorphological evidence while Crawford et al. (1985) examinedallozyme variation. Particular attention was given to asecond group, L. fremontii (Torr. ex Gray) Greene, L.burkeii (Greene) Greene, and L. conjugens Greene, sincethese species are readily crossed and produce a largeproportion of vigorous and fertile hybrids (Ornduff, 1969and Ornduff, 1976). Ornduff used this group to speculate onthe possible mechanisms of evolution in Lasthenia. Ornduff5(1976) suggested that the large phenotypic variation amongspecies is the result of relatively few genetic differencesand is largely the result of catastrophic selection. Laterinvestigations by Crawford et al. (1985), and Crawford andOrnduff (1989) using allozyme variation supported theprogenitor-derivative hypothesis for L. minor and L.maritima, and agree that L. fremontii, L. burkei, and L.conjugens are closely related.Although much seems to be known about the genus, L.californica remains a 'problem' species. The nature andprocesses responsible for the variation found in L.californica are not well understood. The species shows ahigh degree of morphological, ecological, cytological andbiochemical diversity and is probably the most variable ofthe 17 species included in the genus. Plants in thisspecies vary from the large-headed, succulent, but shortplants, found on the coastal regions to tall, delicatespecimens with smaller heads in inland regions. Populationsof this species occur on coastal bluffs, in open grasslands,oak woodlands, alkali flats, chaparral, and in the desert,and may be found growing on a wide variety of soil types.There are also intraspecific differences in plant habit,nature of pubescence, achene and pappus morphology,flavonoid chemistry and in chromosome number. The magnitudeof these differences is so large that the species was placedunder three genera and at one point was divided into sevenspecies, or 11 taxa when the varieties are included.6The taxonomic history of the species spans over 150years. In 1836, Fischer and Meyer described specimens foundin the Bodega Bay area as Baeria chrysostoma and laterduring that year de Candolle erected the genus Burriela toaccommodate plants similar to Baeria chrysostoma butdiffering in the possession of pappus. By 1841, fivespecies describing what is now known as L. californica wererecognized under the genus Burriela: two by de Candolle(1836) and three by Nuttall (1841). Nuttall's treatmentlisted four species occurring in the same area near SantaBarbara, of which three were found in a single population.In 1842, Torrey and Gray described an additional species,Burriela chrysostoma T.& G., and mentioned the similaritiesbetween the two genera Baeria and Burriela but found itinadvisable to increase the synonymy by employing the nameBaeria for the whole genus. It was not until 1874 that thespecies described under Burriela and Baeria were treated ascongeneric and placed under Baeria by Gray. By 1886, Grayhad described and recognized seven species, or 11 taxaincluding the varieties, all of which referred to L.californica s.1.. These taxa were separated by acombination of characters such as nature of pubescence,shape of achene, and shape and colour of the pappus. Itseems that what is now understood as intra- andinterpopulational variation was then accommodated bydescribing new taxa both at the specific and infraspecificlevels.7In 1894, Greene simplified the synonymy of the entiregenus by treating the genera Lasthenia, Baeria, andCrockeria as congeneric and merging the species underLasthenia. As a result, three species, L. chrysostoma Greene, L. ctracilis Greene, and L. hirsutula Greene wererecognized for the L. californica complex. Unfortunately,Greene did not explain the reasoning behind his viewpointand the congeneric recognition of the taxa was soonabandoned. Between 1897 and 1903, Greene describedadditional species for the complex under the genus Baeria.Four years later, Hall acknowledged the close relationshipamong the species of this complex by recognizing only onespecies, Baeria chrysostoma, and accommodated themorphological variation by describing eight forms under B.chrysostoma var. gracilis. In 1942, Howell relegated B.palmeri Gray to the varietal level as B. chrysostoma F. & M.var. palmeri (Gray) Howell comb. nov., based on specimensfrom Guadalupe Island.In their taxonomic treatments, both Ferris (1958 and1960) and Keck (1959) agreed that the variation found inthis assemblage of populations was best accommodated byrecognizing three infraspecific taxa: Baeria chrysostoma ssp. chrysostoma, B. chrysostoma ssp. hirsutula, and B.chrysostoma ssp. aracilis. Their treatments did not bringnew information on the biology of the species but representthe first agreement on the delineation of taxa in thecomplex. It took another six years for new information on8the biology of the species to be available when Ornduff(1966) reviewed the taxonomy of the entire genus Lasthenia.His monograph dealt with various aspects of the morphology,ecology, breeding systems, and genetic and cytologicalrelationships which he used to clarify the taxonomic statusof the species and to speculate on patterns of evolution inLasthenia.Ornduff (1966) adopted Greene's earlier view intreating Lasthenia, Baeria, and Crockeria as congeneric.However, it was not until 1978 that L. californica wasrecognized as the legitimate name for this species whenJohnson and Ornduff (1978) had the opportunity to examinetype specimens that were previously unavailable. They foundthat the holotype of L. californica and the holotype of L.chrysostoma were the same species, and because publicationof the epithet californica predated that of chrysostoma, thespecies name was changed to L. californica.During his investigation of the genus, Ornduff (1966)examined a large number of specimens of L. californica whichled him to defer the recognition of the three infraspecifictaxa previously described. Ornduff found that a largenumber of populations could be distinguished based onmorphological differences used in previous treatments, buthe also realized that several populations includedindividuals that could not be assigned to any of thesubspecies traditionally recognized. He referred to thisgroup as a "complicated aggregation of races" needing9further attention. Investigations subsequent to Ornduff'smonograph serve to further demonstrate the variability foundwithin this species (e.g., Bohm et al., 1974 and 1989).Literature reviewGeographical and ecological racesThe geographical races referred to by Ornduff (1966)are the coastal and inland races, the latter being furtherdivided into inland races from desert or non-desert areas.Figure 2 illustrates some of the morphological differencesamong the three races. According to Ornduff's description,plants of the coastal race are characterized by branchesthat are stout, decumbent, with internodes usually shorterthan those of inland races. These plants usually live for alonger period of time, germinate earlier and require alonger time to flower than the inland races. Plants of theinland race, from desert areas, are shorter than the otherinland race, are highly branched, the branches originatingfrom the lower leaf axils. Plants from the inland race fromnon-desert areas are erect, relatively unbranched, but ifso, the branches arise from middle and upper leaf axils.Differences between these races are apparently retained inprogenies when grown in the greenhouse suggesting that thisintraspecific variation is genetically controlled (Ornduff,1966).Cytological racesUntil the present work, only diploid (n=8) andtetraploid (n=16) populations were known to occur in1011Figure 2- Diagrams showing some of the morphologicalvariation among three races of Lasthenia californica described by Ornduff. a) inland nondesert race, b) coastal race, and c) inland desertrace. After Ornduff (1966).L. californica. Most of the 80 populations surveyed byOrnduff (1966) for chromosome counts were diploids. InCalifornia and northern Baja California tetraploid cytotypesare widely distributed but not common, and only tetraploidpopulations are found in southern Oregon (Fig. 1). There isno report of diploid and tetraploid individuals occurring inthe same population. Ornduff (1966) found no morphologicaldifferences (including pollen grain size) or ecologicalpreferences associated with the cytotypes.Ornduff (1966) conducted a hybridization program andobtained the following results. Hybrids between diploid andtetraploid plants produce pollen with low percentagestainability with most values ranging between 0-19%.Hybrids between tetraploid individuals of differentpopulations generally produced high percentage of stainablepollen grains (80 to 100%) but hybrids between diploidindividuals from different populations showed a wide rangeof pollen stainability (0 to 100%). Meiosis was regular inhybrids with good pollen as well as in some partly sterilediploid hybrids. In other hybrids, meiosis was irregularwhere cell bridges and a variable number of univalents wereobserved at metaphase I.Most of the crosses performed by Ornduff (includingdiploids and tetraploids) produced less than 30% seed set.No positive relationship was observed between the percentageof seed produced by a cross and the fertility of hybrids.In addition, there was no correlation between the percentage12of seed set and the morphological, ecological orgeographical characteristics of the parental strains.Ornduff (1966) believed that there existed severalchromosome races with various degrees of distinctiveness atthe diploid level and that the tetraploid races haveprobably arisen by hybridization and chromosome doublingbetween various diploid races.Chemical racesStudies on the flavonoid constitution of L. californicareveal that there exists inter- and intrapopulationvariation in these pigments (Bohm et al., 1974, 1989).Interpopulation variation was observed in a survey of sixpopulations, including both diploid and tetraploidpopulations where bulk leaf material was used for theanalyses (Bohm et al., 1974). Three chemical races wereobserved and the only tetraploid population examinedproduced the richest array of compounds. No clearcorrelation was found between these results and the variousraces described in the species, except that the populationsfrom inland localities lacked some of the compounds (Ornduffet al., 1974). However, considering the extensivedistribution of the species and the low number ofpopulations surveyed, more populations needed to be examinedto verify such a correlation.Intrapopulation variation was observed following anextensive survey of one population of L. californica in theJasper Ridge Biological Preserve of Stanford University13(Bohm et al., 1989). This survey provided six years of datawhere flavonoid pigments of individual plants along severaltransects were analysed. Four flavonoid pigment types (A,B, C, and D) were observed, with three (A, B, and C)occurring in high frequencies (Bohm et al., 1989).Intrapopulation variation is not unique to L. californica but the distribution of the flavonoid patterns along thetransects is unusual. The distribution of the flavonoidprofiles along the transects did not vary over a period ofsix years. Over the six year period the frequency ofoccurrence did not change significantly for profile type A,but some fluctuation in the frequencies of occurrence wasobserved for profile types B and C. Progeny from a maternalplant exhibiting flavonoid type A exhibited the same patternwhile progeny from a maternal plant exhibiting flavonoidtype B or C displayed either profile B or C. During thecourse of this investigation three additional populationswere examined: one contained all three types of flavonoidpigments, one contained only type A and one contained type Cplants only (Bohm et al., 1989).EvolutionClearly, L. californica is a highly variable speciesexhibiting inter- and intrapopulation variation in severalcharacteristics. Interpopulation variability is expressedin the formation of numerous races which seem to beseparated spatially, by barriers to crossing, and by hybridsterility at the diploid level (Ornduff, 1966). Ornduff14suggested that genetic changes may have accumulated inisolated populations as a result of long spatial isolation.That genetic changes have occurred in this species issupported by the narrow range of tolerance of someecological races and by the retention of morphologicalcharacters in individuals grown under uniform conditions.According to Ornduff (1966) some of the races may showenough variation to deserve taxonomic recognition. Whetherthe recognition of taxa within this group is justifieddepends on the degree of divergence that has occurred amongraces. Ornduff (1966) emphasized the large degree ofvariability in this species but did not attempt to quantifyany of this information. His observations on the morphologyand ecology of the species were based mostly on theexamination of herbarium specimens and collection ofmaterial throughout the range of the species.The presence of races, local and geographical, alsosuggests that the L. californica complex may be an exampleof gradual speciation. This process, as summarized in Grant(1981), is described as occurring in steps where, after theformation of local races, further selection may lead to theformation of geographical races and then species that differin morphology, cytology and ecology. The examination ofthose features (i.e., morphology, cytology, genetics)provides evidence used to test hypotheses about the mode ofspeciation in taxa. The information also provides insightson the possible forces responsible for the maintenance of15variation within the species. Such forces could beselection for adaptation to local environments, and/or ofgenetic drift as a result of isolation.The consistent occurrence of specific flavonoidpatterns across a sharp boundary further complicates thepattern of variation in L. californica. Similar patternshave been reported where plants showed phenotypic and/orallozyme differences across a boundary (Jain and Bradshaw,1966; Hamrick and Holden, 1979). Several of theseinvestigations showed that the sharp boundary between the'races' was the result of strong selective pressure whereunfit genotypes were eliminated on one side of the boundary(e.g., Watson, 1969) or where gene flow across the boundaryis reduced through the development of self-fertility (e.g.,Antonovics, 1969). Little seems to be known about thebiology of L. californica at the Jasper Ridge Preserve.Bohm et al. (1989) already emphasized that further work onthe reproductive biology of this population is needed. Theysuggested that the pattern observed could be the result ofmaternal inheritance or a manifestation of restricted geneflow. If it is the manifestation of restricted gene flowthen one would expect to find some differences along thetransects that would act to reduce gene flow. Differencescould be in edaphic conditions such as in pH, moisture oreven changes in wind exposure, and perhaps these differencesare concordent with variation in the morphological and/orgenetic characteristics of the plants along the transect.16Purpose of this researchThe information available through Ornduff's (1966) andBohm et al.'s (1989) investigations have indicated a highlevel of variability in L. californica. However, little isknown about the processes or mechanisms responsible for thediversity found within the species. Previous work on thespecies also suggested that this complex may includeentities deserving taxonomic recognition either at theinfraspecific or specific level. A first step in trying tosolve questions about the classification and processes ofevolution in a taxon involves gathering information thatwill allow comparison between different aspects of thebiology of the taxon. For classification purposes one hasto be able to define and describe the organism andpractically this means finding a morphological character ora set of characters that distinguish the taxa. Ideally, theuse of diagnostic characters should also reflect the amountof genetic differentiation. It is often difficult to relatemorphological differentiation (either external morphology orchemical) with genetic differentiation found within a taxonsince such characters often have a complex genetic basis.The use of isozyme data can provide a useful estimate ofgenetic variation. Several reviews (e.g., Gottlieb, 1981,Crawford, 1985) have shown the ease with which isozyme datacan be gathered and interpreted. The large number ofstudies for which isozyme data have been used haveestablished a basis against which new information can be17compared. Gottlieb (1977, 1981) and Crawford (1983, 1985)have shown how these data can be used in the classificationof taxa and how they can provide information about processesof evolution such as speciation, or hybridization.Additionally, Hamrick et al. (1979), Loveless and Hamrick(1984), and recently Hamrick and Godt (1989), demonstratedhow isozyme data may relate to the life historycharacteristics of a taxon at the population and specieslevels.The present work involves a biosystematicinvestigation of the L. californica complex. The mainpurpose of this study was to examine how variation inmorphology, flavonoid chemistry, and isozymes is apportionedwithin the species complex. Both morphological andflavonoid variation have previously been examined for arelatively small number of specimens or populations. Thepresent investigation extends the data set to a largernumber of populations and individuals per populationallowing comparisons to be made between the levels ofdiversity found within and among populations. This studymostly concentrated on levels of allozyme variation withinand among populations of the L. californica complex sincethis data set may provide a more powerful estimate ofgenetic variation than is obtained with morphological andflavonoid data. Particular attention was devoted to theJasper Ridge population for several reasons: 1) it was ofinterest to determine whether the known variation patterns18in flavonoid chemistry were also represented inmorphological and allozyme data sets, 2) if patterns ofdifferentiation at this site are not unique, this populationcan be representive of population subdivisions for thespecies and can serve as a model, and finally, 3) thelocation of the site allows one to establish long termstudies without fear of destruction or disturbance throughhuman activity.This investigation of L. californica was undertaken inthe hope of answering specific questions. These are:1) Are the individuals in the Jasper Ridge populationvariable only for their flavonoid chemistry?2) If not, is there a correlation between flavonoidchemistry and other data sets?3) Is the variation found in one or some of the datasets in Jasper Ridge representative of the variationfor the same data sets throughout the range of thespecies?4) Is there any evidence that L. californica asdelineated now represents an aggregation of taxa, ateither the specific or subspecific level?5) What are the forces or mechanisms most likelyresponsible for the differentiation found within thespecies?This investigation of L. californica was undertakenwith the belief that the information could be used tosuggest evolutionary mechanisms and delineation of taxa, andto formulate hypotheses for future work.19MATERIALS AND METHODSPopulation SamplingThe locations of the 36 populations studied arepresented in Table 1. The populations were selected in anattempt to cover the geographical range and the differenthabitats where the species is known to occur in Oregon,California, Arizona and Baja California. Figure 3 shows thedistribution of the populations examined. During the threeyears of sampling (1988, 1989 and 1990) only one population(437) representing a coastal habitat was found south of SanFrancisco County. For each population, material wascollected for morphological, flavonoid and isozyme analyses.In most cases different individuals were used for each ofthese analyses. For some populations (327, 428, 435 and452) one or two of these analyses were omitted due to thelack of material available at the time of sampling. Thematerial used for the isozyme analysis of 12 populations wascollected in different years due to the lack of matureachenes during the first year of collection. Table 2enumerates the analyses performed for each population.Voucher specimens were also collected from each populationand are deposited at the University of British ColumbiaHerbarium (UBC).Sampling at Jasper RidgeIn 1989 and 1990, four transects were examined at theJasper Ridge site. Transects I, II, and III were 25 metersapart. Transect IV sampled in 1989 was 50 meters east of2021TABLE 1- Locations of the Lasthenia californica populationscollected in 1988, 1989, and 1990.COUNTIES^LOCATION^POPULATION PLOIDYDESIGNATIONARIZONA^YavapaiYavapaiPimaBAJACALIFORNIABagdadHillsideTanque VerdeMission (near)Cantamar433^2x434 2x435^2x436^2x437 2xCALIFORNIA San Benito^Panoche (near)^323^2xMonterey Priest Valley (near)325 4xKern^Maricopa^327^4xKern Tehachapi 328 4xSan Diego^Cuyamaca lake^330^4xVentura Ventucopa (South of)331 4xMarin^Mt Tamalpais^333^2xSonoma Salt Point State Pk 335 2xTehama Red Bluff 425^2xTehama^Paskenta^428 2xRiverside^Lake Matthew^438^4xSanta Barbara Los Olivos 439 6xSan Luis Obispo Santa Margarita^440^2xSan Benito^Hollister^441 2xSanta Clara^Hwy 152,East of^442^4xSanta NellaTuolumne^Jamestown^443^2xAmador Jackson 444 2xSacramento^Clay 445^2xSan Mateo Jasper Ridge^447-462^2xMarin^Point Reyes 449^2xMarin Dillon Beach^450 2xSonoma Fort Ross 451^2xMendocino^Fort Bragg^452 2xShasta Millville 460^2xFresno^Hwy 198OREGONSan DiegoRiversideLos AngelesFresnoJacksonJackson(near Coalinga)Anza BorregoMurrietaPalmdaleDunlapGold HillTable Rock464^4x468 2x471^2x473 4x475^2x453^4x454 4x452450449335451333445••444447/462 •443An 044244111.• 323 •475On,325• 4400327 0328033104730438• 473• 46803309437• 433• 4345354• 460mik•425W42822180°+380436240km0435 Figure 3- Distribution of the 36 populations includedin the study of Lasthenia californica. Numbersbeside symbols represent population designation.• n=8, on=16, • n=24.TABLE 2- Analyses performed onLasthenia californica.each population ofPOP MORPH FLAV ISOZYME POLLEN1988 1989 1990 VIAB DIAM323 * * *325 * * * *327 *328-432 * * * * * *330 * * *331 * * *333 * * *335 * * *425 * * * *428 *433 * * *434 * * * *435 * * * * *436 * * * * *437 * * * * *438 * * * * * *439 * * * * *440 * * * * *441 * * *442 * * * * *443 * * *444 * * * * *445 * * *447-462 * * * *449 * * * * * *450 * * * *451 * * *452 * * * *453 * * * * *454 * * * *460 * * * *464 * * *468 * * * *471 * * * *473 * * * *475 * * *Pop=population designationMorph=morphological analysisFlav=flavonoid analysisViab=pollen viabilityDiam=pollen diameter23transect I, and transect IV sampled in 1990 was 25 meterseast of transect I. Figure 4 shows the study site. Thetransects ranged from about 32 to about 62 m. in length.The terrain slopes gently from Fire road, where samplingbegan, to the boundary of an oak woodland habitat, wheresampling ended. Plants were collected at every meter alongeach of these transects. In 1989, two plants were collectedat every meter. Flowers from one plant were used forflavonoid analysis, and the achenes from that individualwere germinated and the seedlings or mature plants were usedfor isozyme analysis; the second plant was used for themorphological analysis. Details concerning these analysesare presented in the following sections.^In 1990, only oneplant was collected at every meter. The flower heads ofthis plant were used for the flavonoid analysis, to scorethe shape of the pappus, and the achenes were germinated andthe seedlings were used for the isozyme analysis.Morphological AnalysisThirty-four populations were examined for morphologicalvariation: two from Oregon, two from Arizona, two from BajaCalifornia, and 28 from California (Table 2). Wholeindividuals, including roots, were collected from eachpopulation and pressed for later examination. Twenty plantsper population, with flowers past anthesis, were selectedfor the morphometric analysis. Each individual was scoredfor 24 characters: 11 quantitative and 13 qualitative (Table3). To allow a comparison between the results of this and24I^50m.^1if Oak-GrasslandBoundary1 ---"K—II1111111i1I11I25Figure 4- Map of transects in the Jasper Ridge population.TABLE 3- Morphological characters measured on thepopulations of Lasthenia californica.26QUANTITATIVE CHARACTERS:Plant heightHead diameterLeaf lengthLeaf widthRay flower lengthRay flower widthPhyllary numberPhyllary lengthPhyllary widthAchene lengthPappus lengthQUALITATIVE CHARACTERS:Branching patternDensity of pubescence:leaf surfacephyllary tipphyllary surfacestemLeaf marginRay flowers, colourAchene shape in lengthAchene surface (vestiture)Pappus shapePappusPappus colour% achene/headPLH cmHDD cmLFLE cmLFWI cmLE cmWI cmNOPLE cmPWI cmALE cmPL cmBR 0=upper 1/2, 1=lower 1/2,2=rosette, 3=none1DL 0= very low,PT 1=moderate,PS^2=high,ST 3=noneMA* 0=toothed, 1=smoothCO 1=yellow, 2=yellow and WhiteSH 1=linear, 2=clavateSU 0=glabrous, 1=pubescent,2=papillaeP^0=absent, 1=subulate2=lanceolate, 3=linearPA* 0=absent, 1=presentPC* 0=absent, 1=white, 2=brownPER* 0=60-100%, 1=30-60%, 2=0-30%* Characters not included in statistical analysesprevious studies, some of the characters included here havealso been used to separate the taxa in previous taxonomictreatments. Some of the characters listed in Table 3require more explicit definitions. For head diameter (HDD),the widest portion of the head including the outermost rayflowers was measured. Leaf measurements (LFLE and LFWI)were taken from the first (when intact) or second set ofleaves below the most mature flower head. Ray flower width(WI) and phyllary width (PWI) were measured in their widestportion. All measurements for PLE, PWI, PT and PS werescored from the same, outermost phyllary.Pollen Viability and SizeSeventeen populations were examined for percentagepollen stainability and eight populations were examined forpollen grain diameter. At least one population representingeach cytotype was included in the data sets. Two hundredand fifty-five individuals were examined for percentagepollen stainability by leaving the pollen of individualplants to stain for 48 hours in lactophenol and anilineblue. Pollen grains were classified as normal andpresumably viable when they appeared fully stained, and asaborted when unstained or lightly stained and/or shrunken.Counts were done under a light microscope at 40X.Pollen grain diameter was measured for 116 individuals.Three to four pollen grains were measured for eachindividual and, when possible, 15 individuals per population27were scored. Measurements were taken using a lightmicroscope at 40X with an ocular micrometer.Statistical AnalysisStatistical analyses were performed at the Universityof British Columbia on the MTS system using the subroutinesGLM and CANDISC available in SAS (SAS Institute 1985a,1985b). All cytotypes were included in these analyses.One-way analyses of variance (ANOVA) were performed on eachof the 11 quantitative characters. Prior to conducting anANOVA, a Bartlett test was performed on each of thequantitative characters to test the homogeneity of variancesamong populations. The results showed significantdifferences for seven of the 11 characters. However, theANOVA and canonical analyses were still performed since theaim of this study was to explore the possible relationshipsamong the populations. The examination of population meansalso indicates variability among the morphologicalcharacters that I consider biologically different. Toexplore the possible relationships among populations twocanonical discriminant analyses were performed: one whereboth quantitative and qualitative characters were used; onewhere only the qualitative characters were included in orderto account for the potential increase in the overall size ofplant parts often associated with polyploidy. Four of thequalitative characters listed in Table 3 were not includedin the canonical analyses. Leaf margin (MA) was scoredbecause it was reported to be variable in previous studies.28No variation was recorded for the individuals collected inthe field but several individuals from different populationshad marginal foliar teeth when grown in a controlledenvironment. Presence or absence of pappus (PA) and pappuscolour (PC) were not included because they wereinterdependent with other variables; when pappus (PA) wasscored as absent, pappus colour (PC) was also scored asabsent, and pappus length (PLE) was scored as zero.Percentage of achenes per head (PER) was scored as anindication of the growth stage of the individuals examined.In addition to the graphs showing the distribution ofpopulation means along the first two axes for each analysis,the relative contribution of the characters to the canonicalaxes were shown by a set of character vectors. Thepositions of the end points of the vectors were obtainedfrom the canonical loadings for each character on therespective axis. These were standardized by dividing eachby the square root of the corresponding element of thediagonal of the inverse within-group covariance matrix(Reyment et al., 1984).Measurements for percentage pollen stainability andpollen tetrad diameter were analysed by nested one-wayanalysis of variance using the GLM subroutine available inSAS. An arcsine square-root transformation was performed onthe percent pollen stained per population in order tofulfill the requirements of normality. Multiple range29tests, Student-Newman-Keuls (SNK) and Duncan tests, wereperformed on each of these data sets.Chromosome CountsThe chromosome number for each population wasdetermined from meiotic chromosome squashes for one to threeindividuals per population. Flower buds collected fromnatural populations and from plants grown in environmentalcabinets were fixed in Carnoy's solution (3 chloroform:6alcohol:1 glacial acetic acid), and transferred to 70%ethanol for storage after 24 hours. Anthers were hydrolysedin 1 N HC1 for 20 minutes at room temperature and stainedwith haematoxylin.Flavonoid AnalysisBohm et al. (1989) showed that in L. californica thefull array of flavonoid compounds can be observed usingfloral tissue. Thirty-three different populations wereexamined: two from Oregon, three from Arizona, two from BajaCalifornia and 26 from California (Table 2). Flower headsof 25 to 40 individuals were collected and kept in separatelabeled envelopes for each population. The basis for thepopulation screening followed the method used by Bohm et al.(1989) where the major flavonoid variants were cleanlyresolved using one-dimensional thin-layer chromatography.The solvent system employed consisted of an aqueous-basedmixture (70 water:15 n-butano1:10 acetone:5:dioxane).Flavonoid compounds were obtained by placing flower heads ofeach plant in small vials to which a few drops of methanol30were added. Extraction of the floral flavonoids wascomplete after 30 minutes and samples for chromatographywere taken directly from the vials by means of capillaryspotting tubes. Extracts of about 25 individuals werechromatographed on a 20 X 20 cm homemade plate. Thechromatographic medium used was Polyamid 6.6 spread to adepth of ca. 0.3mm. Each individual was scored as havingone of the flavonoid pigment types (A, B, C and D) followingthe designation used in Bohm et al. (1989).In addition to the population study, flavonoid profileswere examined using herbarium specimens borrowed from thefollowing herbaria: JEPS, UC, CAS, OSUF, and LA. Thepurpose of this survey was to provide a better estimate ofthe pattern of variation found in flavonoid profilesthroughout the range of the species. Herbarium sheets wereselected from those with individuals from which it waspossible to remove a few florets without damaging thespecimens. One to four individuals were examined per sheet.A total of 563 individuals from 398 herbarium sheets werescored for their flavonoid profiles. The counties and thenumber of individuals examined in each county are presentedin Table 4.Isozyme AnalysisAchenes were collected from 34 populations; two fromOregon, two from Arizona, two from Baja California, and 28from California (Table 2). Seeds were germinated and grownin controlled environmental chambers (10 hours daylight at3132TABLE 4- Flavonoid profiles occurringsurveyed using herbariumcalifornica.in thespecimenscountiesof LastheniaSTATEFlavonoid Profiles No. ofHerbariumSheetsCOUNTY A B C DArizona Cochise 1 1Gila 12 5Graham 5 2Maricopa 15 9Pima 19 13Pinal 19 10Yavapai 5 2California Alameda 4 4 6Amador 2 2Butte 2 2Calaveras 9 2 9Colusa 2 1Contra Costa 4 10 12Del Norte 5 4Eldorado 4 2 4Fresno 7 1 1 7Humboldt 7 5 10Kern 21 6 15Kings 3 2Lake 9 3 11Los Angeles 13 2 11Madera 3 1 1 4Marin 10 4 10Mariposa 3 4 5Mendocino 7 2 6Merced 4 1 5Monterey 14 3 11Napa 9 1 9Orange 6 1 5Placer 2 1 3Riverside 11 1 10San Benito 15 8San Bernardino 16 2 12San Diego 14 11San Francisco 8 5San Joaquin 3 1 3San LuisObispo 19 13San Mateo 11 8 12Santa Barbara 26 1 20Santa Clara 9 4 8Santa Cruz 12 4 12Shasta 5 3Solano 2 10 8Sonoma 1 3 4Stanislaus 5 3 1 6Sutter 3 3Tehama 2 4 3Tulare 4 6 9Tuolumne 8 1 7Ventura 19 1 13Oregon Curry 1 1Douglas 1 1Jackson 13 10BajaCalifornia 22 2 2015°C, 14 hours darkness at 12 °C). For populations 434, 449,450, 452, and 453 difficulties arose in obtaining matureachenes due to differences in the time of floweringthroughout the range of the species when sampled; for thosepopulations leaf material was collected in the field andkept on ice to be analyzed for allozyme variationimmediately on arrival at the University of BritishColumbia.A total of 1390 plants were assayed by horizontal gelelectrophoresis; 456 individuals from 11 tetraploidpopulations and 894 individuals from 22 diploid populations.The average number of individuals examined in eachpopulation is presented in Table 6. Forty individuals froman hexaploid population were scored for their isozymepatterns but those results were not included in the analysesdue to difficulties in interpreting the banding patterns.Enzymes were extracted from leaves of either seedlingsor mature plants; both gave the same banding patterns, andactivity and resolution were similar for given plants. Thetris-HC1 grinding buffer-PVP solution of Soltis et al.(1983) was employed with 10% PVP. Electrophoresis wasconducted on 10% to 11% starch gels made with differentcombinations of gel and electrode buffers according to theparticular enzymes assayed. A citric acid-morpholine geland electrode buffer (Odrzykoski and Gottlieb, 1984),adjusted to pH 6.4, was used to resolve aconitase (ACON),glyceraldehyde 3-phosphate dehydrogenase (G3PDH),33phosphoglucomutase (PGM), phosphogluconate dehydrogenase(6PGD), and shikimate dehydrogenase (SKDH). A modifiedsystem 8 (Soltis and Rieseberg, 1986) was used for leucineaminopeptidase (LAP), malic enzyme (ME),phosphoglucoisomerase (PGI), and nicotinamide adeninedinucleotide dehydrogenase (NADHDH). Enzymatic assaysfollowed Soltis et al. (1983), except for NADHDH whichfollowed Cheliak and Pitel's (1984) assay. Menadionereductase and isocitrate dehydrogenase expressed activitybut were not scored because they did not exhibit consistentactivity.Since the number of progeny obtained from crossingexperiments was small, the number of loci and alleles foreach enzyme were inferred for each population. Severallines of evidence were used to interpret the bandingpatterns. Subunit structure and locus numbers commonlyobserved in diploid plant species (Gottlieb 1981, 1982;Weeden and Wendel, 1989) and in other species of Lasthenia (Crawford et al., 1985, and Crawford and Ornduff, 1989) wereassumed. The examination of segregation patterns in progenyfrom individual plants collected in nature as well asrecording the relative intensity of the bands forpolymorphic loci provided valuable information. Since nogenetic analyses were performed to confirm the hypotheticalinheritance patterns, the phenotypes were interpretedconservatively. This approach invoked using the lowestnumbers of loci and alleles.34When more than one locus was observed for an enzymesystem the one with the most anodal position was designated1, the next 2, and so forth. For each locus, the allelesspecifying the allozymes were called a, b, c, etc, theclosest to the front being designated as a.The following genetic parameters were calculated at thepopulation level: the proportion of polymorphic loci (P),the proportion of heterozygous loci (P'), observedheterozygosity (Ho) (direct count), and the mean number ofalleles per locus using all loci (A), and, for diploidpopulations, expected heterozygosity (He) (or genediversity). At the population level those parametersrepresent population means. Gene diversity statistics (Ht,Hs, Dst, and Gst) and standard genetic identities werecalculated utilizing the methods of Nei (1972, 1973)implemented by the GENESTAT-PC program (Version 2.1, by PaulLewis and Richard Whitkus; Whitkus, 1988). A UPGMA(unweighted pair-group method) phenogram was generated usingNTSYS-pc (version 1.5 of Rohlf, 1988) based on Nei'sunbiased genetic identities matrix obtained from GENESTAT.Estimates of gene flow were calculated from the geneticdata. The number of migrants per generation (Nm) wasestimated using Wright's formula (Wright, 1951) where Nm=(1- Fst)/4Fst. The Fst value was replaced by the Gst, anequivalent value.35Crossing ExperimentsA limited number of reciprocal crosses between plantsof flavonoid types A and C were performed; 10 (A X A), 8 (CX C), 10 (A X C) and 10 (C X A). For each cross, plantsfrom the same population were utilized. Twenty crosses usedindividuals from Jasper Ridge, four crosses included diploidindividuals from three different populations (461, 468, and479) and four crosses used tetraploid individuals from twodifferent populations (473 and 453). In addition, 22 plantswere left to self (10 A type, 12 B/C type); 12 individualswere from Jasper Ridge, two were individuals from diploidpopulations (449 and 475), and eight were tetraploidindividuals (328, 453, 454, and 473).Application of pollen was accomplished by gentlyrubbing the heads of the tagged plants together daily untilthe flower heads had withered. The flower heads of theparent plant were collected when they reached maturity andwere kept in separate labeled envelopes. The percentage ofachenes produced was estimated and fell in three categories:more than 30%, less than 10%, and zero. No percentagesbetween 10 and 30% were observed. Difficulties arose indefining successful versus unsuccessful crosses for thefollowing reasons: the size and the number ofcapitulescences per plant, and the flowering time variedamong individuals, therefore the low percentage of achenesproduced for a given cross might have been the result of thelow amount of pollen received by the maternal plant. On the36other hand, the occasional presence of fungus gnats in thegrowth cabinet might have been responsible for thepollination of some plants from an unknown donor. Giventhose uncontrollable events, a maternal plant with aproduction of fruit above 30% was considered as a successfulcross and a production below 10% was recorded asunsuccessful. I believe that by using this approach thispreliminary experiment provides information on crossabilitybetween plants exhibiting different flavonoid types.37RESULTSThe results obtained from the analyses are presented intwo sections: the first one concerns the Jasper Ridgepopulation only; the second one examines all populations,including the Jasper Ridge population. The Jasper Ridgepopulation is considered first since it will explain whysome groups were defined a priori. Details concerning eachanalysis are provided in the second section.Results obtained for the Jasper Ridge PopulationExamination of the Jasper Ridge population of Lasthenia californica in 1989 and 1990 showed the occurrence of threeflavonoid profiles, A, B, and C, where A and C were the mostfrequent (Table 5). The distribution of these patternsalong the four transects for 1989 and 1990 closelycorresponded to the distribution previously reported by Bohmet al. (1989). Each transect included at least twoflavonoid profiles. Figures 5 and 6 show the spatialdistribution of these patterns in Jasper Ridge for 1989 and1990.Each individual collected along the transects in 1989was scored for 22 morphological characters. One character,pappus morphology, separated the individuals in two majorgroups; one group included 75 individuals with linear pappusand a second group included 103 individuals with lanceolatepappus and four with subulate pappus. The distribution ofthe pappus shape was mapped along the four transects andshowed a strong correlation with flavonoid type (Figure 5).38TABLE 5- Flavonoid profiles occurring in each population ofLasthenia californicaPOPULATIONSTotal NoofA B C D Individuals323 40 40325 36 36328 65 65330 33 1 1 35331 35 35333 54 6 60335 25 5 30425 33 33433 32 32434 25 1 2 28435 20 10 30436 19 6 25437 30 30438 26 4 30439 29 1 30440 30 30441 27 2 1 30442 30 30443 30 30444 22 8 30445 22 1 4 27447** 54 5 123 182449 30* 30450 30 30451 30* 30453 30 30454 30 30460 33 33462** 69 18 98 185464 30 30468 30 30471 30 30473 32 32475 11 24 35*modified profile**Jasper Ridge population3940Figure 5- Flavonoid and pappus types along the four transects inJasper Ridge for 1989.TRANSECT IO 14^2012^60 Meters1^1^147CCCCCCCCCCCCCBCCCCCC^ B/C...CCCBCCCCCCAAAA...AAA* **********TRANSECT IIO 12^38^44^50 Meters1^1 1^1^1CCCCCCCCCCCCCCACCCCC...CCBBCAAACAAAAAA* ***^*********TRANSECT IIIO 8^12^23^32 Meters1^1^1CCCCCCCACCCACCCCCCCCCC 1^1AAAAAAAAAA* *^************TRANSECT IV32^40 Meters1 1AAAACCCCCCCAACAAAC.ACCCAAACAAAACCACAACCA***^* ***** *************************** Indicates the meter at which linear pappus was scored. Except forfour plants all others had lanceolate pappus.0 12 15 241 1 1 1Figure 6- Flavonoid and pappus types along the four transects inJasper Ridge for 1990.TRANSECT IO 10^30^43^48^62 Meters1^1 1^1^1CCCCCCBBCCBCC...BBCCCCCBCBCBCCC..BABBCAA^A*^***********TRANSECT IIO 10^33^43^50 Meters1^1 1^1 1CCCCCCCCBCBCCCCCACCBCC...BCC....CAAAAAAAA* ********TRANSECT IIIO 10^26^33 Meters1^1 1 1CCCBCCCCBCCCCCCCBCCCCCCCCAAAAAAAA* ********TRANSECT IVO 10^40 Meters1^1 1AAAAACCCAAAAAAAAAA...AAAAA*****^******************* Indicates the meter at which linear pappus was scored. All otherplants had lanceolate pappus.41In 1989 different individuals were used for those twoanalyses. Fifty-four sampled coordinates had individualsexhibiting flavonoid pattern A, and linear pappus was scoredat 52 of those coordinates. Flavonoid profile type B/C wasscored for 126 individuals and linear pappus was scored atonly 23 of those coordinates. The remaining 103 sitesincluded four plants with subulate pappus and 99 withlanceolate pappus.^In 1990 the three analyses wereperformed on the same individual plants. Pappus shape wasthe only character scored for the individuals collected thatyear. The distribution of the pappus shape along thetransects is strongly correlated with flavonoid type (Figure6). All plants with flavonoid pattern A were scored ashaving linear pappus. Except for one individual, plantsexhibiting flavonoid profile type B/C had lanceolate pappus.This individual, found along transect III, exhibitedflavonoid pattern B but had a linear pappus which ischaracteristic of a plant with flavonoid pattern A.For the 1990 collection, 77 plants were examined forisozyme variation in the Jasper Ridge population. Twoisozymes, Nadhdh and 6Pgd-1, separated the population intotwo groups. One group was characterized by the presence ofthe b allele at Nadhdh and the faster alleles (a, b, and c)at 6Pgd-1; the second group is characterized by the presenceof the a allele at Nadhdh and the slower alleles (e, f, andg) at 6Pgd-1 (Fig. 7). Similar results were obtained fromthe analysis of the 1989 plants from Jasper Ridge.42Figure 7- Photographs of starch gels showing electrophoreticpatterns of plants exhibiting different flavonoidprofiles in the Jasper Ridge population. Numberson the right of the photograph designate isozymes.Letters on the gels represent the flavonoidprofile of maternal plant. a) banding patternsat Nadhdh, b) banding patterns at 6Pgd-1.There is a strong correlation between the three datasets obtained for the Jasper Ridge population in 1990. Allplants with flavonoid pattern A have linear pappus, the ballele for Nadhdh and the faster alleles for 6Pgd-1 (Figs 5,6 and 7). Most of the plants with flavonoid patterns B or Chave lanceolate pappus, the a allele for Nadhdh and theslower alleles for 6Pgd-1. In 1990, only 5 individuals didnot show this correlation and these individuals exhibitedflavonoid profile types B or C. Three out of fourindividuals were heterozygous at 6Pgd-1 with one of thealleles that characterize plants with flavonoid profile A.One plant with flavonoid pattern C was homozygous b atNadhdh, the allele characteristic of a plant with profile A.A fifth plant with flavonoid profile C had a linear pappusbut had the allozymes expected given its flavonoid profile.Parameters representing allelic richness and thefixation index (F) for the Jasper Ridge population are shownin Table 6. Nei's genetic diversity statistics and estimateof gene flow (Nm) for the population are presented in Table7. Values of P (percentage of polymorphic loci), A (averagenumber of alleles per locus), and Ho (observedheterozygosity) are high for the Jasper Ridge population andare comparable between the two years of sampling (i.e., JR89and JR90). Those values are also high for the a priori groups (i.e., A89, B89, A90, and B90) but are lower than thevalues obtained when the population subdivisions are pooled(JR89 and JR90). The high Gst estimate for the entire444 5TABLE 6- Genetic variation at isozyme loci for 33 populations of Lastheniacalifornica.Population^n^P^P'^A^Ho*^He^Fdesignati n Tetraploidpopulations:325 75 .75 .75 2.63 .278327 12 .5 .4 1.75 .180328 65 .75 .75 2.38 .259330 23 .75 .63 2.13 .197331 63 .88 .75 2.88 .249438 32 .63 .63 1.88 .269442 30 .75 .75 2.63 .224453 44 .75 .63 2.0 .286454 17 .75 .50 1.88 .148464 10 .33 .33 1.3 .167473 23 .75 .63 1.88 .223Mean: .69 .61 2.12 .265Diploidpopulations:323 37 .75 .75 2.25 .186 .228 .184333 65 .75 .63 2.63 .257 .299 .140335 20 .50 .50 1.38 .164 .145 -.131425 27 .63 .50 2.75 .267 .290 .079434 20 .38 .38 0.88 .140 .146 .041435 20 .75 .63 2.5 .275 .276 .004436 11 .63 .63 1.75 .130 .148 .122437 30 .50 .50 1.63 .125 .123 -.016440 33 .75 .75 2.38 .229 .264 .133441 24 .75 .63 2.0 .225 .227 .009443 31 .75 .75 2.25 .154 .186 .172444 39 .63 .63 2.13 .269 .231 -.165445 33 .50 .50 1.63 .186 .193 .036449 61 .75 .63 1.88 .112 .181 .381450 40 .88 .63 2.63 .173 .224 .228451 32 .63 .63 1.88 .198 .244 .189460 13 .50 .50 1.38 .225 .197 -.142468 25 .63 .63 2.0 .173 .200 .135471 19 .67 .50 1.83 .113 .124 .089475 20 .50 .50 1.88 .179 .204 .123481 30 .75 .75 2.0 .133 .250 .468A89 27 .63 .63 2.0 .163 .172 .052B89 68 .75 .75 3.13 .180 .225 .200A90 41 .63 .63 2.38 .174 .201 .134B90 33 .75 .75 3.13 .193 .264 .269Mean: .65 .61 2.09 .201 .210 .118JR89** 95 .75 .75 3.13 .172 .288 .402JR90** 74 .75 .75 3.25 .184 .343 .464* No significant differences between Ho and He based on Chi square test.** Jasper Ridge population, including all transects for 1989 and 1990.n=mean number of individuals examined per populationP=Proportion of polymorphic lociP'=Proportion of heterozygous lociA=Mean number of alleles per locusHo=Observed heterozygosityHe=Expected heterozygosityF=Wright's fixation indexTABLE 7- Genetic diversity statistics for each group.Total genetic diversity (Ht), mean diversity withinpopulations (Hs), gene diversity among populations(Dst), and proportion of gene diversity amongpopulations (Gst). Wright's estimate of migrationrate (Nm). JR=Jasper Ridge.Hs Ht Dst Gst NmPolyploid 0.230 0.344 0.115 0.333Diploid 0.217 0.344 0.127 0.370JR 0.219 0.374 0.156 0.417 0.350Species 0.221 0.347 0.130 0.363 0.43946population indicates population subdivision (Table 7). Inaddition, both F and Nm values suggest some restriction ofgene flow and/or population differentiation at this site.The F-values are slightly lower for the subdivisions thanthey are when the subdivisions are pooled. Estimates ofallelic richness (P, A, and H) tend to be higher for thesubdivisions with plants exhibiting flavonoid patterns B/Cthan the estimates obtained for plants exhibiting flavonoidpattern A. Those results suggest that the A-type plantshave lower genetic diversity than the B/C-type plants. Itis interesting to find that the F-values are lower for the Aplants than they are for the B/C plants. Those estimatesare all above zero and suggest that mating is not completelyrandom in the population subdivisions, especially for theB/C plants.Results obtained for all PopulationsFlavonoidsThe examination of a large number of individuals fromthe entire range of the species demonstrated that L.californica showed the presence of the same four flavonoidprofiles reported by Bohm et al. (1989). The flavonoidconstituents for each profile, A, B, C, and D, are presentedin Table 8. Profile C consists of aurone and chalconemonoglycosides and a set of flavone and flavonolglucuronides. Profile B adds luteolin 7-glucoside. ProfileA has the fundamental C profile plus sulphated kaempferoland quercetin diglycosides plus prominant eriodictyol47TABLE 8 - Distribution of flavonoids in the Lastheniacalifornica pigment profile types. (After Bohmet al.,^1989.)Profile typesPigments A^B^C^DGlucuronides +^+ + +Anthochlorsa +^+ + +Sulfated diglycosides +Nonsulfated diglycosidesb +Eriodictyol 7-glucoside + +Luteolin 7-glucoside +aAurone and chalcone glycosides.bBased upon limited sample.48glycosides. Profile D, the rarest of the four, is similarto the A profile except that the flavonol diglycosides lackthe sulphate group.The first survey for flavonoid variation consisted ofexamining individual flower heads taken from the entirerange of the species by using herbarium specimens. Theresults from this survey are presented in Table 4. In all,563 individuals were examined (398 herbarium sheets)providing information for plants from seven counties inArizona, 44 counties in California and three counties inOregon plus collections from Baja California. Flavonoidprofile type A is the most common throughout the range ofthe species. All specimens (76) from Arizona exhibitedprofile A pigments and all specimens (15) from Oregonexhibited profile type C. Most specimens from Californiaexhibited profile A (321) and C (124) pigments; profiles Band D were observed for only three specimens in this state.In Baja California 22 specimens showed profile type A andtwo exhibited profile type B.Thirty-three populations were also examined for theirflavonoid profiles. Three of these populations were fromArizona, two from Baja California, two from Oregon and theremaining populations were from throughout the species'range in California. Except for two populations (449 and451) the profiles were clearly discernable as A, B, C or D(Table 5). The two unusual profiles were modified Bprofiles that require additional detailed chemical study and49represent the only significant deviation from the originalfour profiles previously reported.The distribution of profile types in the populations ispresented in Table 5. Profile type A is the most commonthroughout the range of the species. Thirteen populationsexhibited profile type A only; five populations exhibitedprofile type C only. Most of the mixed populations, both Aand B/C types present, are dominated by profile type A.Population 475 and Jasper Ridge (447 and 462) are the onlypopulations where profile type C occurred in higherfrequency than profile type A.The geographical distribution of the pigment types ispresented in Figure 8. This figure shows that thedistribution of the profile types is not random within therange of the species. The two Oregon populations and thetwo most northerly populations from California exhibit theB/C profile. The populations from central California areeither A and/or B/C types while the populations in southernCalifornia, here defined as south of San Francisco Bay, arepredominantly A type; 86 to 100% of the individualsexhibited profile type A in these populations. Populations440 and 475 are the only two exceptions for this region.One Arizona and one Baja California population sampled arealso exclusively A types. The other three populations fromthose two areas are mixed types, having mostly A profileplants.504. 47/462 442• 323 e75• 325• 4644441.?4, 530460042551335450449454451333^,ikn445glar44441443..,_180.04400439 •328• 331 4.n1111473.438• 47146830436240km...437 . a• 4330434435Figure 8- Geographical distribution of flavonoid profiletypes for Lasthenia californica. The percentageof flavonoid profile type A in a population isdisplayed by increased darkening, with solidcircles=100%. 0 =modified flavonoid profile B.Chromosome CountsThe basic chromosome number in L. californica is n=8(Ornduff, 1966). Chromosome counts revealed the presence ofthree ploidy levels in this complex: 24 populations arediploid, 11 are tetraploid and one is hexaploid (Table 1).Diploid and tetraploid populations were known to occur inthis species, however the discovery of an hexaploidpopulation is new to this complex. This population (439) islocated in Santa Barbara county and was the only oneencountered with this ploidy level. The two Oregonpopulations consisted of tetraploid individuals, the twoBaja California and the three Arizona populations includeddiploid individuals only. All coastal populations examinedwere diploid populations. The other diploid and tetraploidpopulations were found throughout California (Fig. 3).PollenMean pollen grain size for the populations examinedranged from 16.49 to 20.57pm and mean pollen stainabilityper population varied from 55.8 (population 436) to 90.9%(population 449) (Table 9). A one-way nested analysis ofvariance indicated significant variation in pollen diameterand in pollen stainability among populations and cytotypes(Table 10). For each data set Student-Newman-Keuls (SNK)and Duncan multiple range tests resulted in the detection ofgroups including populations with different ploidy levels,suggesting that variation in pollen diameter and in pollen52TABLE 9- Means for pollen viability and pollen grain sizefor each population examined. Pollen viability ispresented as the proportion of stained pollengrains. Pollen size is given in pm.POPULATION^POLLENNUMBER^VIABILITYPOLLENSIZEPLOIDYLEVEL435 0.763 16.49 2x436 0.558 2x437 0.720 2x440 0.854 17.25 2x444 0.805 17.10 2x449 0.909 2x452 0.796 18.72 2x460 0.857 2x468 0.706 2x471 0.878 2x328 0.894 18.68 4x438 0.682 20.48 4x442 0.895 19.52 4x453 0.708 4x463 0.903 4x473 0.884 4x439 0.810 20.57 6x53TABLE 10- One-way nested analyses of variance for pollendiameter and pollen stainability from threepolyploidy levels where eight and 17 populations,respectively, were included. *=P<0.05; **=P<0.001.Pollen diameter Pollen stainabilityd.f.^s.s F d.f. s.s FPolyploidy 2^637.5433 133.95** 2 0.2390 3.04*Population 5^220.5107 18.53** 14 3.7755 6.86**Individual 24^17.2633 0.30 - _Error 374^890.0477 - 239 9.3894Total 405 1765.3651 255 13.4039d.f.=degree of freedoms.s.=sum of square54stainability were largely the result of populationdifferences irrespective of the ploidy level.MorphologyThe results obtained for the qualitative andquantitative measurements are listed in Tables 11 and 12,respectively. The Jasper Ridge population was included inthe morphometric analysis and was labeled as I, II, III andIV. Each number represents the transect along which theindividuals were collected in 1989 and each transect wastreated as a separate population. Except for onecharacter, width of ray flower (WI), the F-values for theother 10 quantitative characters indicated among-populationdifferences for each character (p <0.001) (Table 11).The graphs of the canonical means for the populationsplotted against the first two axes for each character setare shown in Figures 9 and 10. The first and second axesaccounted for 32 and 18% of the variation, respectively,when all characters were included in the analysis (Fig. 9);when qualitative characters only were used in the analysis,the first and second axes accounted for 34 and 25% of thevariation, respectively (Fig. 10).The distribution of population canonical means wassimilar for both data sets. Nine populations, including thefive coastal populations sampled north of San Francisco,were distinctly separated from the others when either of thedata sets was used. The highest loadings obtained for thefirst and second canonical axes indicated that four55•1311 DL PT PS ST CO SH SU P PA PC PER2 3 2 3 0 1 3 0 1^2 3 0 1 2 1 2 1 2 2 0^1 2 3 0 1 0 1 1I 0.10 0.65 0.08 0.17 0.82 0.15 0.03 0.03 0.97 0.17 0.83 0.05 0.95 1.0 0.77 0.23 0.02 0.98 0.05 0.73 0.22 1.0 1.0 0.70 0.20 0.10II 0.14 0.78 0.08 0.36 0.64 0.02 0.98 0.12 0.88 1.0 1.0 0.74 0.26 1.0 0.74 0.26 1.0 1.0 0.62 0.26 0.12III 0.06 0.72 0.16 0.06 0.19 0.81 1.0 0.25 0.75 0.06 0.94 1.0 0.53 0.47 0.06 0.94 0.53 0.47 1.0 1.0 0.88 0.06 0.06IV 0.05 0.65 0.20 0.10 0.12 0.88 1.0 0.23 0.77 1.0 1.0 0.10 0.90 1.0 0.02 0.13 0.85 1.0 1.0 0.55 0.15 0.30323 1.0 0.95 0.05 1.0 1.0 0.95 0.05 0.30 0.70 1.0 1.0 0.10 0.90 0.10 0.90 0.10 0.90 0.05 0.50 0.45325 0.05 0.40 0.55 1.0 1.0 1.0 0.05 0.95 0.20 0.80 0.90 0.10 1.0 0.10 0.90 0.10 0.90 0.10 0.90 0.15 0.50 0.35328 0.65 0.25 0.10 0.35 0.65 1.0 1.0 0.05 0.95 0.20 0.80 0.10 0.90 0.20 0.70 0.10 1.0 1.0 1.0 0.65 0.05 0.30330 0.20 0.60 0.10 0.10 1.0 1.0 1.0 1.0 0.05 0.95 1.0 0.95 0.05 0.10 0.90 0.10 0.90 0.10 0.90 0.05 0.50 0.45331 0.80 0.20 1.0 1.0 1.0 0.10 0.90 1.0 0.85 0.15 0.05 0.90 0.05 0.10 0.90 0.10 0.90 0.10 0.90 0.40 0.45 0.15333 0.75 0.25 0.05 0.80 0.15 1.0 1.0 0.05 0.80 0.15 0.15 0.85 0.30 0.70 0.15 0.60 0.25 0.05 0.40 0.55 1.0 1.0 0.10 0.30 0.60335 0.10 0.55 0.35 0.85 0.15 1.0 1.0 0.85 0.15 1.0 1.0 1.0 1.0 1.0 1.0 0.05 0.05 0.90425 0.10 0.55 0.35 1.0 1.0 1.0 0.05 0.95 1.0 0.15 0.85 0.05 0.35 0.60 0.15 0.85 1.0 1.0 0.10 0.45 0.45428 0.20 0.80 0.05 0.95 1.0 1.0 1.0 1.0 1.0 0.10 0.85 0.05 1.0 1.0 1.0 0.20 0.80432 0.10 0.45 0.25 0.20 0.05 0.95 1.0 0.400.60 1.0 1.0 1.0 0.20 0.80 1.0 1.0 1.0 0.10 0.20 0.70433 0.60 0.25 0.15 0.10 0.90 1.0 1.0 1.0 1.0 0.95 0.05 1.0 0.75 .25 1.0 1.0 0.30 0.15 0.55434 0.60 0.25 0.15 0.05 0.95 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 0.30 0.70436 0.15 0.50 0.20 0.15 0.45 0.10 0.45 1.0 0.70 0.05 0.25 1.0 1.0 1.0 0.95 0.05 0.80 0.20 1.0 1.0 0.20 0.40 0.404374380.050.250.500.750.45 0.450.200.500.800.05 1.01.00.400.150.500.850.10 1.01.01.01.00.700.950.300.050.951.00.05 0.750.100.850.250.05 0.101.00.90 0.101.00.900.800.400.100.500.100.10439 0.40 0.50 0.05 0.05 1.0 0.05 0.95 1.0 1.0 1.0 0.75 0.25 0.05 0.70 0.25 0.200.50 0.30 0.15 0.85 0.20 0.80 0.10 0.70 0.20440 0.10 0.70 0.20 0.05 0.95 1.0 0.20 0.80 - 1.0 1.0 1.0 1.0 1.0 1.0 1.0 0.70 0.10 0.20441 0.05 0.50 0.45 0.05 0.95 1.0 1.0 1.0 1.0 0.35 0.65 1.0 0.95 0.05 1.0 1.0 0.60 0.30 0.104424430.14 0.710.600.100.200.050.20 0.950.400.050.600.051.00.95 0.051.00.95 0.951.00.051.01.0 0.751.00.250.100.100.050.900.85 0.200.900.800.100.20 0.801.00.20 0.801.00.900.700.050.250.050.05444 0.20 0.50 0.15 0.15 0.40 0.60 0.25 0.75 0.05 0.95 1.0 1.0 0.05 0.95 0.15 0.85 0.20 0.80 1.0 1.0 0.25 0.15 0.60445 0.10 0.45 0.05 0.40 0.20 0.60 0.20 1.0 1.0 0.15 0.85 1.0 1.0 0.10 0.05 0.05 0.70 0.30 1.0 1.0 0.15 0.30 0.55449 0.05 0.65 0.10 0.20 1.0 1.0 1.0 1.0 0.95 0.05 1.0 1.0 0.05 0.95 0.05 0.95 0.05 0.95 0.40 0.40 0.20450 0.05 0.70 0.25 0.55 0.45 1.0 0.55 0.45 1.0 1.0 1.0 0.80 0.10 0.10 1.0 1.0 1.0 0.45 0.55451 0.45 0.25 0.30 0.45 0.40 0.15 1.0 0.10 0.850.05 0.05 0.95 0.95 0.05 1.0 0.95 0.05 1.0 1.0 1.0 0.05 0.10 0.85452 0.05 0.05 0.05 0.85 0.15 0.85 0.10 0.10 0.80 0.10 0.90 1.0 1.0 1.0 0.30 0.55 0.15 1.0 1.0 1.0 1.0453 0.25 0.05 0.70 0.10 0.80 0.10 1.0 1.0 1.0 0.10 0.90 1.0 0.15 0.30 0.55 0.10 0.90 1.0 1.0 0.05 0.20 0.75454 0.10 0.90 0.40 0.60' 1.0 1.0 1.0 0.10 0.90 1.0 0.15 0.50 0.35 0.20 0.80 1.0 1.0 0.15 0.25 0.60460 0.10 0.45 0.45 0.35 0.65 1.0 1.0 1.0 1.0 0.40 0.60 0.15 0.85 1.0 1.0 1.0 0.15 0.65 0.20464 0.20 0.65 0.15 1.0 1.0 1.0 1.0 1.0 0.70 0.30 0.05 0.95 0.150.05 0.75 0.05 0.15 0.85 0.15 0.85 0.75 0.25468 1.0 0.20 0.80 0.85 0.15 0.55 0.45 1.0 1.0 0.40 0.60 1.0 0.050.50 0.45 0.05 0.95 0.05 0.95 0.30 0.10 0.60471 0.05 0.80 0.05 0.10 0.75 0.25 1.0 1.0 1.0 1.0 0.95 0.05 1.0 0.20 0.80 1.0 1.0 0.40 0.50 0.10473 0.05 0.95 0.05 0.95 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 0.85 0.10 0.05475 0.05 0.85 0.05 0.05 1.0 1.0 1.0 1.0 1.0 0.60 0.40 1.0 0.50 0.50 1.0 1.0 0.60 0.15 0.25TABLE 11- Results obtained for the qualitative characters recorded onpopulations of L. californica. The results are presented asfrequencies. *See Table 3 for definitions of these characters.TABLE 12- Means and F-values from analysis of variance for the11 quantitative characters included in themorphometric analyses of Lasthenia californica.Results for the Bartlett test show no significantdifferences for LFWI, PWI, ALE, and PL.Pop.samples(n=39) *PLH HDD LFLE LFWI LE WI^NO PLE PWI ALE PLI 12.19 1.15 1.40 0.08 0.49 0.60 7.80 0.48 0.20 0.21 0.22II 12.44 1.25 1.38 0.08 0.52 0.26 8.14 0.49 0.19 0.20 0.23III 13.14 1.37 1.57 0.09 0.56 0.28 8.34 0.50 0.21 0.20 0.22IV 11.48 1.31 1.60 0.12 0.56 0.31 7.73 0.49 0.23 0.18 0.22323 7.05 0.69 1.23 0.10 0.32 0.15 6.70 0.37 0.12 0.21 0.15325 12.93 1.52 1.82 0.12 0.50 0.24 11.25 0.49 0.19 0.25 0.18328 12.34 1.64 1.57 0.19 0.57 0.24 10.45 0.64 0.27 0.26 0330 10.92 1.56 1.94 0.22 0.55 0.24 11.00 0.50 0.24 0.20 0.18331 10.27 1.38 1.51 0.16 0.51 0.24 10.10 0.47 0.19 0.21 0.16333 9.50 1.36 1.96 0.14 0.61 0.28 9.25 0.52 0.22 0.23 0.29335 5.83 1.26 1.83 0.13 0.47 0.27 9.30 0.45 0.26 0.20 0425 11.87 1.35 0.99 0.08 0.69 0.33 6.50 0.51 0.25 0.21 0.25428 11.71 1.24 1.12 0.07 0.54 0.28 9.90 0.48 0.17 0.19 0432 9.03 1.46 1.48 0.13 0.60 0.31 8.90 0.60 0.23 0.28 0433 7.63 1.82 1.18 0.09 0.49 0.21 10.60 0.46 0.18 0.21 0.16434 6.55 1.15 0.66 0.06 0.45 0.18 10.65 0.37 0.12 0.20 0.17436 7.63 0.90 0.73 0.03 0.32 0.18 7.70 0.34 0.16 0.16 0.17437 13.14 1.75 1.25 0.17 0.73 0.38 11.40 0.55 0.31 0.19 0.21438 15.70 1.30 1.26 0.07 0.46 0.21 9.35 0.46 0.19 0.21 0.16439 13.70 1.07 1.20 0.08 0.47 0.18 9.25 0.48 0.19 0.21 0.15440 11.49 1.31 1.60 0.07 0.57 0.24 9.50 0.48 0.19 0.19 0441 20.86 1.54 1.64 0.11 0.62 0.31 10.30 0.55 0.20 0.20 0.21442 20.96 1.77 2.18 0.11 0.72 0.33 9.10 0.62 0.28 0.23 0.23443 13.61 1.24 1.76 0.07 0.49 0.27 7.60 0.46 0.18 0.21 0.20444 14.52 1.24 2.25 0.10 0.54 0.27 8.55 0.50 0.24 0.22 0.24445 10.37 1.12 1.43 0.08 0.50 0.23 7.30 0.46 0.21 0.19 0.24449 5.77 1.66 0.82 0.16 0.64 0.36 10.50 0.58 0.26 0.24 0.19450 5.88 1.69 1.66 0.13 0.70 0.40 10.05 0.57 0.25 0.21 0451 3.77 1.34 1.31 0.11 0.57 0.28 8.95 0.50 0.20 0.20 0452 4.15 1.31 0.69 0.14 0.48 0.34 8.45 0.43 0.23 0.20 0.18453 12.11 1.37 1.58 0.10 0.62 0.27 6.85 0.56 0.22 0.24 0.29454 7.60 1.07 1.33 0.10 0.52 0.24 5.95 0.49 0.20 0.25 0.27460 11.39 1.44 1.33 0.10 0.68 0.32 7.45 0.59 0.25 0.22 0.28464 9.45 1.08 1.05 0.08 0.42 0.20 8.05 0.44 0.19 0.20 0.16468 7.66 1.31 1.19 0.11 0.48 0.21 7.95 0.44 0.20 0.20 0.18471 9.85 1.20 1.07 0.09 0.48 0.23 8.30 0.39 0.16 0.17 0.16473 14.39 1.61 1.51 0.08 0.63 0.26 8.50 0.62 0.25 0.28 0.24475 12.52 1.10 0.92 0.07 0.54 0.21 8.00 0.44 0.17 0.20 0.19Means 10.97 1.32 1.43 0.10 0.54 0.29 8.73 0.49 0.21 0.21 0.17F 40.95 7.55 10.1 12.7 19.7 0.47ns 15.67 14.8 23.1 22.3 102Note: ns, not significant; all others significant at p<0.001level.*See Table 3 for definitions of these characters.573.5-PT58PL452•PLHSTCO449.7.5-5.5-473•335•450•451432•443•.445x.. 425•444323• 4.37404 A71..11-• - Oa325• .433 330•^434•^.475436•^•442•4394684138..441454•^•453333• 460-0.5--2.5-440..428 •331328- 4.5- 61.5^-41.5^-25^- 015^1.15^3.5CAN 1Figure 9- Distribution of the canonical means against thefirst two axes for 34 populations based on 20morphological characters. Character vectorsshowing the relative contribution of the first sixcharacters to the two axes are included. Table 3defines those characters. I, •11, III, and IV,represent the transects examined in 1989 in JasperRidge.6.5- •452 PT59 4.5-COST•335^ •449.443•4542.5-•473•46005-•U-1.5--•450^ .445Try-•451333• it 411.444323•^425 V0325 43• •I475 011• 0433412 4F.439441•^434 438330• •436432••428440.•468-3.5-331.•328-5^ 1^3CAN 1Figure 10- Distribution of the canonical means against thefirst two axes for 34 populations based on 9qualitative characters. Character vectorsshowing the relative contribution of the firstfive characters to the two axes are included.Table 3 defines those characters. I, II, III,and IV represent the transects examined in 1990in Jasper Ridge.-5.5characters were responsible for the separation of thesepopulations as seen in Figures 9 and 10. In order ofimportance, these were presence/absence of pappus (P andPL), colour of ray flower (CO), type of vestiture on achenes(SU), and plant height (PLH). The five coastal populations(335, 449, 450, 451, and 452) contained individuals that arecharacterized by the colour of the ray flower (entirelyyellow) and the individuals are shorter. The mean plantheight for these populations ranged from 3.77 to 5.88 cm.The mean for all populations was 10.97 cm and ranged from3.77 to 20.96 cm (Table 12). The other characters (PL/P,and SU) contributed to the separation of seven populationsincluding three of the coastal populations (335, 450, and451). In those three coastal populations and in an inlandpopulation (440) all individuals lacked pappus and 80 to100% of the achenes were glabrous. All the individuals inpopulations 328, 428, and 432 had epappose achenes.Populations 328 and 432 were collected at the same locationbut in different years (1989 and 1990). They were the onlytetraploid populations that were separated from the otherpopulations.The canonical analysis performed using the qualitativecharacters only resulted in the separation of fouradditional populations (331, 443, 468, and 473) along thesecond axis. Density of pubescence on the stem (ST) and onthe phyllary tip (PT) accounted for the separation of thosepopulations. The density of pubescence on the stem was low60for the individuals in populations 473 and 443 but high inpopulations 328 and 331. Two populations were characterizedby low density (468) or lack of pubescence (452) on thephyllary tip.Individuals with entirely yellow ray flowers and/orlacking a pappus are not unique to the above populations.There were populations throughout the species' range inwhich one to six individuals were epappose and/or had yellowray flowers. However, the only populations with 95 to 100%of the individuals characterized by entirely yellow rayflowers were found in coastal habitats.The geographical distribution of pappus shape in thepopulations is shown in Figure 11. There were fourcharacter states for this morphological feature (Figure 12).Several populations (23) included two or sometimes threecharacter states. The geographical distribution of thischaracter is not random. All populations (8) that includeindividuals with linear pappus are located in the northernpart of the species' range. Those populations included 453and 454 from Oregon, 460, the most northerly population inCalifornia, and populations 333, 442-467 (Jasper Ridge),444, 445, and 449 (a coastal population). Populations forwhich the majority of individuals have subulate pappus arefound in the southern part of the range, including Arizonaand Baja California. Other character states (i.e., pappuslacking and lanceolate) are found throughout the range ofthe species.6141530454442/4674420323 c4750325-G0464©4400439 ©328033141•473180062452335 0 •N451460• 425428450 0 333 445449■, 1444443437043847140 4680330043304342 40 kM36Figure 11- Geographical distribution of pappus types forLasthenia californica. • =linear pappus,C)=subulate pappus,^=lanceolate pappus,c)=all individuals epappose, and open section ofthe circle represents the percentage of epapposeindividuals in the population.63A^B^C^DFigure 12- Diagram showing pappus variation in Lasthenia californica. a) lanceolate, b) subulate, c)epappose, and d) linear. (After Ornduff, 1996)IsozymesEnzyme electrophoresis resulted in the clear resolutionof nine enzyme systems. Three of these enzymes, PGM, SKDH,and LAP were not included in the analyses presented heresince the highly variable banding patterns could neither beinterpreted nor confirmed by progeny analysis. Theremaining six systems included isozymes specified by nineloci, of which eight were well resolved and are discussed;those are Pgi-2, 6Pgd-1, 6Pgd-2, Acon-1, Acon-2, Nadhdh, Me,G3pdh. Pgi-1 was detected but could not be clearlyresolved. This difficulty has been encountered in otherspecies of Lasthenia by Crawford et al. (1985), and Crawfordand Ornduff (1989) who reported that this isozyme is locatedin the chloroplast.Allelic frequencies obtained for each population arepresented in Table 13. The number of alleles shared betweencytotypes for each enzyme system examined is presented inTable 14. G3pdh was monomorphic for the same allele in allindividuals examined. Eight populations, including bothcytotypes, possessed Me-a and Me-b in heterozygotes.Frequencies of the uncommon allele, Me-b, ranged from 0.01to 0.09. Of the 36 alleles detected for the eight loci, 30were shared by the diploid and tetraploid cytotypes. Sixalleles were found only in diploid populations and none wereunique to the tetraploid populations (Table 14). Ingeneral, all populations shared the same common allele atevery locus. Acon-2e and 6-Pgd-la were unique to64Loci :Alleles:ANadhdh Acon-1 Acon-2 6-Pgd-1 6-Pgd-2 Pgi-2 Me-1B A B^C A B C D E A B C D E F G A B C D^E F A B C D E^F G H I J A B325 0.97 0.03 0.02 0.91^0.07 0.08 0.57 0.29 0.06 0.10 0.88 0.02 0.08 0.07 0.62 0.120.05 0.06 0.14 0.74 0.12 1.0327 1.0 0.88^0.12 0.04 0.29 0.61 0.06 1.0 0.58 0.26 0.080.04 0.04 0.14 0.82 0.04 1.0328 0.93 0.07 0.95 0.05 0.08 0.41 0.50 0.01 0.88 0.12 0.06 0.0B 0.51 0.270.07 0.01 0.28 0.69 0.03 1.0330 1.0 0.96^0.04 0.07 0.35 0.58 0.02 0.98 0.05 0.02 0.86 0.05 0.02 0.10 0.85 0.05 0.91 0.09331 0.93 0.07 0.06 0.92^0.02 0.145 0.44 0.41 0.005 0.03 0.94 0.03 0.09 0.06 0.82 0.010.02 0.25 0.01 0.72 0.02 0.99 0.01438 1.0 0.98^0.02 0.11 0.70 0.19 0.01 0.78 0.21 0.06 0.72 0.22 0.17 0.67 0.14 0.02 1.0442 0.02 0.98 0.99^0.01 0.08 0.72 0.19 0.01 0.82 0.16 0.02 0.20 0.02 0.64 0.020.10 0.02 0.05 0.03 0.79 0.13 1.0453 0.01 0.99 0.74^0.26 0.23 0.42 0.34 0.01 0.63 0.37 0.50 0.50 0.02 0.03 0.92 0.03 1.0454 1.0 0.50 0.50 0.75 0.25 0.85 0.15 0.15 0.05 0.75 0.05 0.17 0.78 0.05 0.95 0.05464 1.0 n.a. n.a. 1.0 0.10 0.15 0.60 0.15 0.15 0.05 0.75 0.05 1.0473 1.0 0.90^0.10 0.10 0.82 0.08 0.75 0.25 0.02 0.87 0.11 0.19 0.77 0.04 0.98 0.02323 0.84 0.16 0.85^0.15 0.13 0.75 0.12 0.07 0.93 0.17 0.30 0.46 0.030.03 0.01 0.04 0.95 0.01 1.0333 1.0 0.06 0.86^0.08 0.04 0.54 0.39 0.03 0.40 0.50 0.08 0.02 0.35 0.63 0.02 0.09 0.01 0.71 0.18 0.01 0.98 0.02335 1.0 0.97^0.03 0.08 0.37 0.42 0.13 1.0 0.19 0.81 0.02 0.93 0.05 1.0425 1.0 0.94^0.06 0.14 0.38 0.04 0.42 0.02 0.29 0.42 0.19 0.06 0.04 0.15 0.06 0.70 0.09 0.06 0.04 0.02 0.78 0.02 0.08 1.0434 1.0 1.0 0.40 0.60 1.0 0.75 0.25 0.02 0.11 0.82 0.05 1.0435 1.0 0.78^0.22 0.055 0.50 0.39 0.055 0.04 0.94 0.02 0.14 0.19 0.63 0.02 0.02 0.02 0.28 0.63 0.07 0.95 0.05436 1.0 0.95^0.05 0.08 0.67 0.17 0.08 0.96 0.04 0.09 0.86 0.05 0.09 0.86 0.05 1.0437 1.0 0.98^0.02 0.10 0.76 0.14 1.0 0.01 0.86 0.13 0.12 0.02 0.82 0.02 0.02 1.0440 0.94 0.06 0.75^0.25 0.13 0.6 0.26 0.01 0.35 0.19 0.38 0.08 0.10 0.03 0.84 0.03 0.03 0.02 0.95 1.0441 0.95 0.05 0.96^0.04 0.06 0.67 0.27 0.07 0.93 0.27 0.21 0.52 0.18 0.02 0.73 0.07 1.0443 0.98 0.02 0.01 0.85^0.14 0.01 0.17 0.74 0.08 0.03 0.85 0.12 0.10 0.88 0.02 0.13 0.84 0.03 1.0444 1.0 0.02 0.79^0.19 0.17 0.53 0.25 0.05 0.89 0.11 0.25 0.01 0.69 0.05 0.01 0.11 0.01 0.87 1.0445 1.0 1.0 0.23 0.32 0.35 0.10 0.81 0.17 0.02 0.24 0.03 0.73 0.04 0.89 0.07 1.0449 0.02 0.98 0.93^0.07 0.08 0.84 0.08 0.44 0.56 0.15 0.84 0.01 0.08 0.05 0.87 1.0450 0.02 0.98 0.75^0.25 0.08 0.55 0.28 0.09 0.09 0.91 0.21 0.79 0.04 0.04 0.01 0.86 0.03 0.01 0.01 0.99 0.01451 1.0 0.02 0.76^0.22 0.12 0.52 0.29 0.07 0.24 0.70 0.06 0.26 0.74 0.03 0.94 0.03 1.0460 1.0 1.0 0.50 0.50 0.69 0.22 0.06 0.03 0.44 0.53 0.03 0.96 0.04 1.0468 1.0 0.02 0.80^0.18 0.08 0.28 0.62 0.02 0.03 0.05 0.83 0.09 0.05 0.93 0.02 0.19 0.81 1.0471 1.0 n.a. n.a. 0.87 0.13 0.05 0.87 0.03 0.05 0.08 0.90 0.02 0.95 0.05475 1.0 1.0 0.18 0.77 0.05 0.17 0.81 0.02 0.06 0.83 0.11 0.33 0.02 0.48 0.04 0.02 0.11 1.0481 0.07 0.93 0.85^0.15 0.35 0.58 0.05 0.02 0.55 0.04 0.06 0.35 0.19 0.81 0.11 0.89 1.0Al 1.0 0.92^0.08 0.15 0.68 0.15 0.02 0.90 0.10 0.08 0.04 0.78 0.10 0.02 0.06 0.02 0.90 1.081 0.82 0.18 0.01 0.95^0.04 0.20 0.73 0.06 0.01 0.05 0.94 0.01 0.09 0.03 0.83 0.04 0.01 0.02 0.30 0.05 0.01 0.59 0.01 0.01 0.01 1.0A2 1.0 0.85^0.15 0.04 0.78 0.16 0.02 0.03 0.84 0.13 0.18 0.02 0.68 0.010.11 0.05 0.01 0.88 0.05 0.01 1.082 0.94 0.06 0.02 0.86^0.12 0.11 0.85 0.03 0.01 0.03 0.02 0.02 0.84 0.09 0.06 0.37 0.48 0.020.02 0.05 0.41 0.01 0.51 0.06 0.01 1.0TABLE 13- Allelic frequencies for each population of Lasthenia californica.G3pdh was monomorphic for the same allele in all populations. Al, A2,Bl, and B2 represent A priori groups from the Jasper Ridgepopulation. See text for definitions.TABLE 14- Number of alleles observed and shared by thecytotypes of Lasthenia californica.Total No.of allelesNo. of allelesobserved in No. ofsharedalleles2x 4xAcon-1 3 3 3 3Acon-2 5 5 4 4G3pd 1 1 1 1Me 2 2 2 2Nadhdh 2 2 2 26-Pgd-1 7 7 5 56-Pgd-2 6 6 6 6Pgi-2 10 10 7 7Total 36 36 30 3066populations 425 and A2, respectively. All of the otheralleles that occurred in low frequencies were found in morethan one population and were widespread throughout thegeographical range of the species. Except for Pgi-2j and6Pgd-le, alleles unique to the diploid cytotype wereobserved in low frequency (<3%). Unbalanced heterozygositywas commonly observed for individual tetraploids at Pgi-2and Acon-2 (Figs 13a and b). Three-banded patterns at themonomeric, Acon-2, were also detected for tetraploidindividuals (Fig. 13b).Allelic frequency for Nadhdh and 6Pgd-1 separated thepopulations into two groups with distinct geographicaldistributions (Figs 14 and 15). Most populations containedone of the two alleles at Nadhdh, a or b, in high frequency(>93%) (Table 13). A similar pattern was observed for 6Pgd-1 but it involves sets of alleles. The populations aredifferentiated according to a set of slow or fast alleles.The frequency of the faster alleles (i.e., a, b, and c) at6Pgd-1 was above 81% for 11 populations. Populations 425and 481 had lower frequencies of 71 and 59% for thesealleles, respectively. The remaining populations have highfrequencies for the slower alleles (d, e, f, and g).The geographical distribution of the allele frequenciesfor Nadhdh and 6Pgd-1 is presented in Figures 14 and 15,respectively. All populations north of San Francisco weredominated by the b allele at Nadhdh and the faster alleles(a, b, and c) at 6Pgd-1. Except for populations 442, all67AmoimilmeinianowFigure 13 - Photographs of starch gels showingelectrophoretic patterns of diploid andtetraploid plants of Lasthenia californica forPgi-2 and Acon-2. Numbers of the right of thephotograph designate isozymes. a) Pgi-2. Lanes1, 2 and 3 are examples of unbalancedheterozygosity in tetraploid individuals. Lane4, normal heterozygosity in a diploid individual.b) Acon-2. Lanes 1 and 2 represent unbalancedheterozygosity in tetraploid individuals. Lanes3 and 4 represent three-banded patterns intetraploid individuals.8Figure 14- Geographical distribution of allelic frequenciesat Nadhdh for Lasthenia californica. Thefrequency of allele b in a population isdisplayed by increased darkening, with solidcircle=100%.7nFigure 15- Geographical distribution of allelic frequenciesat 6Pgd-1 for Lasthenia californica. Thefrequency of the faster alleles (a, b, and c) ina population is displayed by increased darkening,with solid circle=100%.populations in the southern part of the range were dominatedby the a allele at Nadhdh and the slower alleles at 6Pgd-1.Measures of genetic structure for the populations arepresented for three groups in Tables 6 and 7. One groupincluded all polyploid populations, a second group includedall diploid populations, and a third group represented theJasper Ridge population where a priori groupings wereestablished based on flavonoid profile types and the year ofcollection. The Jasper Ridge population included fourgroups: Al and A2 represented groups with individualsexhibiting flavonoid profile type A collected in 1989 and1990, respectively, and B1 and B2 represented flavonoid typeB/C individuals collected in 1989 and 1990, respectively.Lasthenia californica has relatively high amounts ofgenetic variation; only one locus (G3pdh) was monomorphic(Table 13). Although LAP, SKDH, and PGM were not includedin this analysis, the zymograms for theses enzyme systemssuggested high levels of polymorphism at the intra- andinter-population levels (Figs 16a, b, and c)). It is likelythat the interpretation and the inclusion of these datawould have resulted in estimates of genetic variationcomparable to what was obtained. The values obtained for P(proportion of polymorphic loci), P' (proportion ofheterozygous loci), and A (mean number of alleles per locus)at the population level were similar among the populationsand between cytotypes (Table 6). The mean number of allelesper locus per population ranged from 1.30 to 3.25 (excluding71•ii^01^.cqt,irc'h^ Tyl^rit:n 1populations 464 and 434 due to the low number of individualsand loci sampled). The mean numbers of alleles per locuswere 2.12 and 2.09 for the tetraploid and diploidpopulations, respectively, with an overall mean of 2.17 forthe species. Values of the proportion of polymorphic locifor the populations ranged from 0.50 to 0.88 with means of0.69 and 0.65 for tetraploid and diploid populations,respectively (Table 6). Most of the polymorphism occurringat a locus was found in an heterozygous form as seen by thesimilar values obtained for P and P'. Higher values for Pand A were obtained when no a priori groups were assumed forthe Jasper Ridge population. Within a population, the locishowing the greatest diversity were Acon-2, 6Pgd-2 and Pgi-2with Hs values of 0.516, 0.416, and 0.320, respectively(Table 15). Me and Nadhdh had the lowest diversity valuesfor Hs.Values of observed (Ho) and expected (He)heterozygosities for each diploid population are presentedin Table 6. Expected heterozygosities were calculated forthe diploid populations only since the Hardy-Weinberg modelassumes diploidy and sexual reproduction. The mean Ho forall populations is 0.186, and ranges from 0.112 to 0.286.The mean Ho values for the diploid and tetraploidpopulations were 0.201 and 0.265, respectively. The chisquare test indicated that the observed heterozygosityvalues for the diploid populations were not significantlydifferent from the expected heterozygosities. However, He73TABLE 15- Gene diversity statistics for each of the sevenpolymorphic loci over all populations. Hs=genediversity within populations, Ht=total gene diversity withinthe species, Dst=gene diversity among populations,Gst=coefficient of gene differentiation.Loci Hs Ht Dst GstNadhdh .043 .504 .461 .915Aconl .189 .206 .018 .085Acon2 .516 .596 .080 .1346-pgd-1 .268 .632 .365 .5776-pgd-2 .416 .479 .063 .132Pgi-2 .320 .341 .021 .062Me .016 .017 .0004 .02574was generally higher than Ho, especially when no a priori groups were used in the analysis of the Jasper Ridgepopulation. Fixation index (F) for diploid populations arepositive, except for four populations, and range from 0.004to 0.468 (Table 6). The relatively high F values obtainedfor the populations suggest inbreeding or populationsubdivision in these populations. When a priori groups areused in the analysis of the Jasper Ridge population the F-values vary from 0.052 to 0.269. When no a priori groupsare considered, the values obtained for the collections from1989 and 1990 are quite high (0.402 and 0.464) againindicating that some functional inbreeding or populationsubdivision occurs in this population.Genetic diversity statistics are presented in Table 7.Both the genetic diversity within populations (Hs) and thecoefficient of genetic differentiation (Gst) were high,indicating a high level of genetic diversity within andamong populations. The examination of the genetic diversitystatistics for each locus indicate an extremely high Gstvalue (0.915) at Nadhdh and a relatively high Gst value(0.577) at 6Pgd-1 (Table 15). The high Gst values at thosetwo loci probably contributed to the high Gst value obtainedfor the species, especially since only eight loci wereincluded in the analyses. The high Gst values for Nadhdhand 6Pgd-1 also reflect the among population differentiationin the distribution of the alleles at those two loci.^Thealleles at those loci showed a polarized (north-south)75distribution through the geographical range of L.californica (Figs 14 and 15).Mean genetic identity for pairwise comparisons amongall populations of the species was 0.837 and ranged from0.611 to 1.0 (Table 16). Similar values and ranges wereobtained within each of the three groups(i.e., polyploid,diploid, and JR). Mean genetic identities between the threegroups were similar and were all above 0.963 (Table 17).The cluster analysis revealed two groups with a geneticidentity of 0.72 (Fig. 17). The high cophenetic coefficient(0.96) indicates that the phenogram satisfactorily portraysthe relationships in the original genetic identity matrix(Sneath and Sokal, 1973). The formation of the two groupswas again influenced by the allelic frequencies at theNadhdh and 6Pgd-1 loci. All populations in the lower partof the graph were found in the northern part of the species'range and includes the Jasper Ridge individuals exhibitingflavonoid type A. The tetraploid population 442 was theonly population from the south to cluster with the northernpopulations. The second group included southern populationsonly and the individuals from Jasper Ridge exhibitingflavonoid pattern B or C.Crossing ExperimentsThe results obtained from the crossing experiment arepresented in Table 18. Some of the plants died before theyhad the opportunity to produce fruits, therefore, the totalnumber of maternal plants in Table 18 is lower than76TABLE 16- Mean genetic identities and ranges ofidentitiespopulationsRidge.for pair-wise comparisons offor Lasthenia californica. JR=JasperGroupsMeanIdentityRange ofIdentitiesPolyploid 0.850 0.620-0.998Diploid 0.836 0.643-1.000JR 0.801 0.707-0.975Species 0.837 0.611-1.00077TABLE 17- Genetic identities between thegroups. JR=Jasper Ridge.Polyploids^DiploidsPolyploidsDiploids^0.963JR^0.975^0.98078790.70^0.75^0.80^0.85^aso^0.95^1.00Genetic IdentityFigure 17- Cluster analysis (UPGMA) of 33 populations ofLasthenia californica, including the Jasper Ridgepopulation (Al, A2, Bl, and B2). Geneticidentity between the two groups=0.72. Copheneticcorrelation=0.96. S=Southern population,N=Northern population.TABLE 18- Results obtained fromselfing experiments.the reciprocal crossing andA* B/C*% of seedsproduced:^0^<10% >30% 0 <10% >30%Number ofreciprocalcrosses:10 (A X A)^2^2 15 /19**8^(C X C) 14 /1410^(A X C)^7^1 2 /10 3 5 1 /9Number ofplants selfed:10^A^0/1012 B/C 3 /12*Flavonoid profile type of the maternal plant.**The number after slash bar (/) indicates the number ofmaternal plants that survived during the experiments.80expected. When plants exhibiting the same flavonoid patternwere crossed all the C type produced fruits, and 15 out of19 plants with flavonoid type A had a fruit set above 30%.The remaining four plants were from the Jasper Ridgepopulation: two did not produce any fruit and two had fruitsets below 10%. Ten reciprocal crosses were performed usingparent plants with different flavonoid profiles. Three ofthose crosses resulted in the production of seed sets above30% where two of the maternal plants were A type and one wasa C type.Three of the 22 individuals left to self produced largeseed sets; one diploid individual was from population 475and two were tetraploid individuals from population 454 inOregon.81DISCUSSIONGeneral Patterns of VariationLasthenia californica shows relatively high levels ofvariation in morphology, flavonoid chemistry and isozymes.The morphometric analysis of the species failed to separatethe populations into coherent groups based on the charactersexamined in this study. The nine populations that separatedfrom the remaining 27 included five coastal populations andfour inland populations. Ray flower colour (i.e., entirelyyellow) and plant height (i.e., short) are the charactersthat contributed to the separation of the coastalpopulations. However, those two character states are notunique to those populations. The individuals in the fourinland populations share one trait only, absence of pappus,but again this character state was found in otherpopulations throughout the range of the species.Morphological variation in L. californica was alsoobserved between two years of collection. Populations 328and 432 were collected at the same locality in the Tehachapimountains but in different years (1988 and 1989,respectively). The position of these two populations inFigures 9 and 10 suggests that they are more closely relatedto other populations than they are to each other.Differences in the qualitative characters scored wereobserved, especially for density of pubescence on theleaves, phyllaries, and on the stem (Table 11). Thepopulations means for the quantitative characters indicate82that the values are higher for the 1988 sample (328) thanthey are for the 1989 sample (432) (Table 12). Thedifferences between the two years might be the result offluctuation in rainfall and temperatures, or of the samplingof younger, therefore smaller, individuals in 1989.The extensive flavonoid survey of L. californicademonstrated the presence of the four pigment profilespreviously observed as well as a new flavonoid profile intwo populations along the coast north of San Francisco.Minor variants of the basic flavonoid types have also beennoted. The examination of individual flower heads takenfrom herbarium specimens covering the entire range of thespecies indicates that the four flavonoid profiles are foundthroughout the range of the species. The examination of 33additional populations also supports the heterogeneity inthe flavonoid chemistry of the species and indicates thatthe two known flavonoid profiles A and C predominate in L.californica.The isozyme analysis supports the morphological andchemical analyses in that the level of allozyme diversity inL. californica is relatively high. High levels ofheterogeneity are expected for a species with the ecologicaland historical characteristics of L. californica. Severalreviews (e.g., Hamrick et al., 1979, Hamrick, 1982, Lovelessand Hamrick, 1984, and Hamrick and Godt, 1989) havedemonstrated that there can be a strong correlation betweenlevels of allozyme diversity and the ecological and83historical traits (e.g., mating systems, geographical range,successional status) of plant species. Those reviewscorrelated estimates of allelic richness and thedistribution of the genetic diversity with particularbiological aspects of a species. In a recent paper,reviewing over 450 species Hamrick and Godt (1989) concludedthat geographic range combined with the plant breedingsystem are the best predictors of allozyme variation. Forexample, Hamrick and Godt (1989) presented mean values atthe population level of P (percentage of polymorphic loci),A (average number of alleles per locus), and He (genediversity) for 85 widespread species of 43%, 1.72, and0.159, respectively. Values of P=65%, A=2.09, and He=0.21were obtained at the population level for L. californica(Table 6). Although estimates of allozyme diversity areslightly high for L. californica, in general, they arecomparable to the mean values reported by Hamrick and Godt(1989) for taxa with similar ecological and historicalcharacteristics (i.e., geographic range and breedingsystem).Differentiation in the Jasper Ridge PopulationThe examination of the Jasper Ridge populationindicates subpopulation division at this site. Thedistribution of flavonoid patterns presented by Bohm et al.(1989) combined with the results of this study suggest thatthe pattern of occurrence at this site has remained constantover a period of 10 years. The analyses performed in this84study showed that there is a strong correlation betweenflavonoid type, pappus shape and allele frequencies atNadhdh and 6Pgd-1, suggesting differentiation within thispopulation. The population forms two coherent groups: onegroup is characterized by linear pappus, flavonoid patternA, the b allele at Nadhdh and the faster set of alleles at6Pgd-1, whereas the other group is characterized bylanceolate pappus, flavonoid pattern B or C, the a allele atNadhdh and the slower set of alleles at 6Pgd-1.Parameters that estimate genetic differentiation inpopulations further support population subdivision at theJasper Ridge site. The high Gst (0.417) approaches the meanvalue estimated for selfing species (Gst=0.510). Theestimated Gst value indicates that 42% of the total geneticvariation at the Jasper Ridge site exists as differencesbetween subdivisions. The calculated Nm value of 0.350 isless than 1.0, which is commonly taken as the breakpointbelow which genetic drift can play a major role indetermining the distribution of genetic variation amongpopulation subdivisions (Wright, 1951). A priori groupswere defined based on the flavonoid chemistry of thematernal plants for the Jasper Ridge population. The geneticidentity between those groups and the fixation index (F) foreach group (i.e, A89, A90, B89, and B90) and when the groupswere pooled (i.e., JR89 and JR90) were estimated. Geneticidentities (I) at Jasper Ridge range from 0.707 to 0.975(Table 16) and suggest population differentiation at this85site. The range of genetic identities falls below the meanvalue estimated by Crawford (1983) (1=0.789) for pairs ofcongeneric species. The higher F values estimated for 1989and 1990 samples (0.402 and 0.464, respectively) whencompared to the estimates for each group defined a priori suggest population subdivision, inbreeding, and/or theaction of selection (Table 6). Estimates that express thedistribution of genotype frequencies in the Jasper Ridgepopulation indicate population subdivision at this site anda reduced level of gene flow within this popultation.The preliminary breeding experiment performed usingindividuals from different subdivisions indicates a lowlevel of crossability between the subdivisions in the JasperRidge population. Results from this experiment furthersupport the existence of population subdivision as suggestedby the measures presented in the previous paragraph (i.e.,Gst, Nm, I and F values). The factors responsible forreduced gene flow are unknown but asynchronous flowering mayplay a role: during the three years this population wasvisited, the flower of the plants with flavonoid pattern Bor C had produced their fruits when the plants exhibitingflavonoid pattern A were beginning to flower.The examination of the Jasper Ridge populationindicates that this population is representative of theheterogeneity found within the species. The four pappusshapes found throughout the species' range were also foundin the Jasper Ridge population. Three pappus shapes were86scored from the field collections (i.e., linear, subulateand lanceolate pappus) and epappose achenes were found inthe progeny resulting from one cross between plantsexhibiting flavonoid pattern A and C. The populations alsoinclude the four flavonoid patterns detected throughout thespecies. Each subdivision in the Jasper Ridge populationhas a level of allozyme diversity (P, A and He) comparableto the other populations of the species and those values arehigher when the subdivisions are pooled (i.e., for JR89 andJR 90) (Table 6).Population DifferentiationAlthough the morphological, flavonoid, and isozymeanalyses indicate that L. californica has a relatively highlevel of variation, differentiation among populations wasalso detected. Pappus shape, flavonoid chemistry, andisozymes are usually correlated geographically, but notnecessarily in all individuals in all populations. Thegeographic distribution of pappus shape shows that linearpappus occurs only in northern populations and that subulatepappus is found mostly in the southern populations (Fig.11). Simple flavonoid profiles (B and C), the b allele atNadhdh, and the faster set of alleles at 6Pgd-1 occur in thenorthern part of California and in Oregon while the morecomplex profile (A), the a allele at Nadhdh and the slowerset of alleles at 6Pgd-1 appear in southern California,Arizona and Baja California (Figs 8, 14, and 15).87Measures indicating genetic differentiation amongpopulations reinforce the view of interpopulation variation.The mean genetic identity for L. californica is 0.837 andranges from 0.611 to 1.0 among populations. The range ofgenetic identities (I) extends below the mean valuespresented by Crawford (1983) (1=0.789) for pairs ofcongeneric species. The mean genetic identities for pair-wise comparisons of populations between the three groups(diploid, tetraploid, and Jasper Ridge) are quite high(1>0.962) indicating that genetic differentiation among allpopulations is probably not the result of polyploidy. TheGst value (0.363) for the species implies that 36% of thegenetic diversity is maintained as a result ofinterpopulation differentiation. The indirect estimate ofgene flow (Nm=0.439) is relatively low reinforcing the ideathat pollen and seed movement among populations of L.californica is reduced.Interpopulation differentiation as suggested by thegenetic indices presented above reflects the formation oftwo groups of populations in L. californica. As well, thecluster analysis (UPGMA) separated the populations into twoassemblages based on the allelic frequencies at Nadhdh and6Pgd-1. Those two loci have high Gst values; Gst=0.915 forNadhdh and Gst=0.577 for 6Pgd-1. Given the high Gst valuesat these two loci it is not surprising to find low geneticidentities among some populations, a relatively high Gst88value for the species, and reduced level of gene flow amongpopulations of L. californica.One of the purposes of this study was to find outwhether L. californica represents an aggregation of taxa ateither the specific or subspecific level. Although there isconsiderable differentiation among the populations forpappus shape, flavonoid chemistry, and isozymes, there is noindication that the populations form coherent taxonomicgroups as suggested in the early taxonomic treatments.Previous to Ornduff's (1966) taxonomic treatment, species orsubspecies in this complex were recognized based oncharacters such as nature of pubescence, shape of pappuspaleae and in one case, shape of the leaves. Mymorphometric analyses of L. californica showed that thosecharacters are too variable for grouping populations intotaxonomic categories. None of the characters thatcontributed to the separation of the nine populations (fivecoastal and four inland) from the remaining 27 populationsin the canonical analyses were unique to these samples.Flavonoid chemistry and allelic frequency separate thepopulations into two groups but those groups cannot bedefined morphologically.The polarized distribution (north-south) of theflavonoid patterns, allozymes and pappus shape indicates theformation of geographical races in L. californica. None ofthose races (chemical, allozymic, and morphological)correspond to Ornduff's races. Ornduff (1966) described89three races (coastal, inland non-desert, and inland desert)based on growth habit. In this study variation in growthhabit was observed between the coastal populations and theinland populations. In addition, the coastal populationswere the only ones for which 95 to 100% of the individualsexhibited entirely yellow ray flowers, and these plants werealso characterized as having leathery leaves. However, thelatter character was not included in this analysis since thedifference between leathery and non-leathery leaves wasoften difficult to detect on pressed specimens. Thevariation observed in ray flower colour, growth habit, andtexture of the leaves supports the existence of two of thethree ecological races described by Ornduff (1966). Theseare the coastal and inland races. However, theintrapopulation variation observed in growth habit for theinland populations in this study does not support therecognition of desert and non-desert races.Nature and Origin of Polyploid PopulationsTetraploid populations of L. californica did not showany differences in their flavonoid chemistry or morphology(including pollen grain size and stainability) when comparedwith diploid populations. The absence of differentiationmakes it impossible to use these features in assessing thenature and origin of polyploid populations. The isozymedata were more useful in assessing the nature of thepolyploid populations.90Soltis and Rieseberg (1986) demonstrated that enzymeelectrophoretic analyses provide useful genetic insights indetecting autopolyploid speciation based on theirinvestigation of Tolmiea menziesii. Their study wasespecially important since naturally occurringautopolyploids had been largely uninvestigated. Theypredicted that attributes of naturally occurringautopolyploids should include: 1) extremely high allozymicsimilarity to diploid progenitors, 2) tetrasomicinheritance, 3) higher heterozygosity than the diploidprogenitor (but fixed heterozygosity would not be expected),and 4) maintenance of three or four alleles at a singlelocus in a single plant. Although speculative at thispoint, there are several indications that the polyploidpopulations in L. californica arose via autopolyploidy.There generally is a slight increase in observedheterozygosity in the tetraploid populations compared to thediploid populations. Although the total number of allelesdetected for the tetraploid populations is lower than forthe diploids, the mean number of alleles per locus wascomparable between the two cytotypes (Table 6). As well, nonew alleles were detected in the tetraploid populations: thealleles present in this cytotype comprise a subset of thealleles detected in the diploid populations (Table 14). Thecontribution of a genome other than that from the diploidpopulations of L. californica could result in the presenceof additional, new alleles in tetraploid populations. A few91tetraploid plants exhibited three alleles in unbalancedheterozygosity at the monomeric locus, Acon-1. Unbalancedheterozygosity was also observed for several individuals atboth Acon-2 and Pgi-2, reflecting possible tetrasomicinheritance. However, formal genetic analyses are necessaryto determine mode of inheritance. These observations takentogether strongly suggest an autopolyploid origin for thetetraploid populations of L. californica.The isozyme data presented here are not sufficient toconfirm the origin of the tetraploid populations. However,the artificial hybridization program conducted by Ornduff(1966) provides some insights regarding their origin.Ornduff (1966) reported an absence of multivalents innaturally occurring tetraploids. He also demonstrated thatcytologically distinct diploid races occur in the speciesbased on evidence from cytological behavior and othersources (i.e., morphological and ecological). Based onthese observations, Ornduff (1966) believes that tetraploidpopulations of L. californica have probably been formedindependently several times.The lack of multivalent formation combined with theisozyme data lend support to Ornduff's speculation. Normalbivalent formation in polyploid plants was traditionallybelieved to be an indication of allopolyploidy. However,several studies have reported lack of multivalent formationin naturally occurring autopolyploids (e.g., Coreopsis grandiflora var. longipes (Crawford and Smith, 1984),92Tolmiea menziesii, (Soltis and Rieseberg, 1986), Vacciniumcorymbosum (Krebs and Hancock, 1989), Heuchera micrantha,(Ness et al., 1989). Autopolyploidy may arise fromhybridization between genetically distinct individuals(Stebbins, 1980). Autopolyploidy derived from a singleplant would possess two different alleles at each locus,whereas autotetraploids derived via crossing betweengenetically different plants could potentially exhibit up tofour different alleles at each locus (Ness et al., 1989).The alleles in the three banded patterns observed at Acon-2in tetraploid individuals are shared by both cytotypes in L.californica. This suggests that hybridization in the senseof Stebbins (1980) has been involved in the formation of atleast some polyploid populations, or alternatively, thatpolyploids that originated independently have subsequentlycrossed with each other. Either of these two processesmight explain the lack of mutivalent formation reported byOrnduff (1966).EvolutionIt was one of the purposes of this study to infer themode of speciation in L. californica. I believe that thepatterns of variation observed best fit the model ofgeographical speciation. This model, as summarized in Grant(1981), is a gradual and conservative process. Differentselection pressures on allopatric populations, perhapscombined with drift, are believed to cause gradualdivergence in morphological, chemical, cytological, and93ecological features resulting in the formation of localand/or geographical races. Because this process is assumedto occur over a long period of time, it is usually expectedto find greater divergence at gene loci specifying isozymeswhen compared with other modes of speciation such as quantumor rapid speciation (Gottlieb, 1981, Crawford, 1983, 1985).As well, there might be some positive correlation betweenthe various features that characterize the races. Severalcharacteristics of the L. californica complex suggest thatthe species represent a stage in the process of geographicalspeciation. Those are; the relatively old age of thespecies, low levels of crossability between populations, andthe formation of geographical, ecological, and cytologicalraces.Ornduff (1966) believes that the species is relativelyold based on the following observations: L. californica hasnot become specialized in the kind of habitats it occupies,populations of the species have the basic chromosome numberfor the genus (n=8), and most populations have retainedseveral morphological characters believed to be primitive inthe genus (e.g., leaves entire, herbage pubescent, pappusmonomorphic, achenes pappose). As well, Ornduff (1966)found a low level of crossability between populations withthe same chromosome number.This study has demonstrated that the species hasdiverged at two gene loci; several populations are fixed foralternate alleles at Nadhdh and 6Pgd-1, and the mean genetic94identity between the isozyme races is low (1=0.72). Inaddition, several populations differ in flavonoid patterntypes and pappus shape along a north-south gradient. Theexamination of figures 8, 11, 14, and 15 indicates thatthere is a positive correlation among some of the featuresfor which the species has diverged. There is also aformation of ecological races. One race occupies coastalhabitats and the other occupies a variety of inland habitatsincluding desert areas. Those two races differ inmorphological characteristics such as ray flower colour,plant height, and leaf texture. Finally, the speciesincludes three cytotypes (n=8, 16, and 24) but there is noevidence that the cytotypes have differentiatedmorphologically or chemically.The formation of various races in the L. californicacomplex should not be surprising. The species covers arelatively large geographic range of heterogeneous habitatsand several of its populations are isolated by long distanceand/or unsuitable habitat. Lasthenia californica is anannual and presumably an obligate outcrossing species madeup of several large populations, therefore, given enoughtime, one might expect the populations to diverge as aresult of selective pressure and/or drift. An unusualaspect of the species is to find a low level of crossabilitybetween the races in the Jasper Ridge population. It islikely that two colonists with different genotypes -one anorthern type and the other a southern type- have95established the Jasper Ridge population. However, anobligate outcrossing species is expected to interbreed,recombining the genotypes. This has not happened at theJasper Ridge site, presumably because: 1) flowering timediffers between the two races, 2) there is a low level offertility between the two races (perhaps they represent twocolonists coming from populations that were somewhat low incrossability), and 3) colonization of differentmicrohabitats so that crossing between the two colonists isless frequent or hybrids are less fit to the microhabitats.Differences in flowering time and low levels of crossabilitybetween the two genotypes were observed at the Jasper Ridgesite. However, there is no information available on whetherthe different genotypes occupy different microhabitats . Aswell, whether the maintenance of two genotypes is the resultof the establishment of two colonists coming frompopulations with low level of crossability or the result ofselection for different microhabitats is unknown. A closeexamination of the ecological characteristics at this sitemight help to elucidate the factors responsible for thedistribution pattern observed in the Jasper Ridgepopulation.Summary and ConclusionThis biosystematic study of L. californica hasdemonstrated intra- and interpopulation variation inmorphology, flavonoid chemistry, and isozymes. The patternsof variation observed lead to the following conclusions:961) There is a strong correlation between the occurrenceof flavonoid patterns, pappus shape and allelefrequencies at Nadhdh and 6Pgd-1 in the Jasper Ridgepopulation,2) The two races at Jasper Ridge not only maintaintheir identities but differ in flowering time, andhave a low level of crossability,3) There is some evidence that the two races at JasperRidge represent the establishment of two coloniststhat might have evolved different adaptations intheir original habitats,4) Flavonoid pattern types, pappus shape, and allozymesare geographically correlated but not in allindividuals in all populations,5) In addition to the geographical races, there hasbeen a formation of ecological races,6) The races are not correlated with ploidy levels,7) It is impractical to distinguish infraspecific orspecific taxa,8) There is evidence that the tetraploid cytotypesrepresent autotetraploid populations and that thosepopulations have formed independently several times,9) The patterns of variation observed in flavonoidchemistry, morphology, and isozyme suggest that L.californica represents a stage in the process ofgeographical speciation.97REFERENCES CITEDAntonovics, J. 1969. Evolution in closely adjacent plantpopulations. V. Evolution of self-fertility. Heredity23:219-238.Bohm, B.A. and H.M. Banek. 1987. Flavonoid variation inLasthenia burkei. Biochem. Syst. Ecol. 15:57-59.Bohm, B.A., A. Herring, K.W. Nicholls, L.R. Bohm, and R.Ornduff. 1989. A six-year study of flavonoiddistribution in a population of Lasthenia californica(Asteraceae). Amer. J. Bot. 76(2):157-163.Bohm, B.A., N.A.M. Saleh, and R. Ornduff. 1974. Theflavonoids of Lasthenia (Compositae). Amer. J. Bot.61:551-561.Candolle, A.P. de. 1836. Burriela. Prodr. 5:663-664.Cheliak, W.M. and J.A. Pitel. 1984. Techniques for starchgel electrophoresis of enzymes from forest treespecies. Information Report PI-X-42, Petawawa NationalForestry Institute, Can. For. Serv., 49pp.Crawford, D.J. 1983. Phylogenetic and systematic inferencesfrom electrophoretic studies. Pp. 257-287. In: Isozymesin Plant Genetics and Breeding. Part A. S.D. Tanksley,and T.J. Orton (eds). Elsevier Science Publishers B.V.,Amsterdam.Crawford, D.J. 1985. Electrophoretic data and plantspeciation. Syst. Bot. 10(4):405-416.Crawford, D.J. and R. Ornduff. 1989. Enzyme electrophoresisand evolutionary relationships among three species ofLasthenia (Asteraceae: Heliantheae). Amer. J. Bot.76(2):289-296.Crawford, D.J. and E.B. Smith. 1984. Allozyme divergence andintraspecific variation in Coreopsis grandiflora (Compositae). Syst. Bot. 9:219-225.Crawford, D.J., R. Ornduff, and M.C. Vasey. 1985. Allozymevariation within and between Lasthenia minor and itsderivative species L. maritima (Asteraceae). Amer. J.Bot. 72(8):1177-1184Ferris, R.S. 1958. Taxonomic notes on Western plants. Contr.Dudley Herb. 5:99-108.Ferris, R.S. 1960. Baeria. Pp.212-219. In: IllustratedFlora of the Pacific States. Vol. 4. L. Abrams and R.S.Ferris (eds). Stanford University Press, Stanford.98Fischer, F.E.L. and C.A. Meyer. 1836. Baeria chrysostoma.Ind. Sem. Hort. Petrop. 2:29.Gottlieb, L.D. 1977. Evidence for duplication and divergenceof the structural gene for phosphoglucoisomerase indiploid species of Clarkia. Genetics 86:289-307.Gottlieb, L.D. 1981. Electrophoretic evidence and plantpopulations. Prog. Phytochem. 7:1-46.Gottlieb, L.D. 1982. Conservation and duplication ofisozymes in plants. Science 216(23):373-380.Grant, V. 1981. Plant Speciation. 2nd ed.. ColumbiaUniversity Press, New York.Gray, A. 1874. Baeria. Proc. Am. Acad. 9:196-197.Gray, A. 1886. Baeria. Synoptical flora of North America.2nd ed.. Vol. I, Part 11:452.Greene, E.L. 1894. Lasthenia. Manual of the Botany of theRegion of San Francisco Bay:203-206.Greene, E.L. 1897. Baeria. Flora Franciscana:437-440.Greene, E.L. 1903. Baeria. West. Amer. Plants 2:8.Hall, H.M. 1907. Compositae of southern California. Univ.Calif. Publ. Bot. 3:1-302.Hall, H.M. 1914. Baeria, Lasthenia. Pp. 76-81. In: NorthAmerican Flora, Vol 34, Part 2. New York BotanicalGarden, New York.Hall, H.M. 1915. Notes on Baeria and Lasthenia. Torrey Bot.Club Bull. 42:111-116.Hamrick, J.L. 1982. Plant population genetics and evolution.Amer. J. Bot. 69(10):1685-1693.Hamrick, J.L. and M.J.W. Godt. 1989. Allozyme diversity inplant species. Chapt. 3. Pp. 43-63. In: PlantPopulation Genetics, Breeding and Genetic Resources.A.H.D. Brown, M.T. Clegg, A.L. Kahler, and B.S. Weir(eds). Sinauer Associates, Inc., Sunderland,Massachusetts.Hamrick, J.L. and L.R. Holden. 1979. Influence ofmicrohabitat heterogeneity on gene frequencydistribution and gametic phase disequilibrium in Avenabarbata. Evolution 33:521-533.99Hamrick, J.L., Y.B. Linhart, and J.B. Mitton. 1979.Relationships between life history characteristics andelectrophoretically detectable genetic variation inplants. Ann. Rev. Ecol. Syst. 10:173-200.Howell, J.T. 1942. A list of vascular plants from GuadalupeIsland, Lower California. Leafl. West. Bot. 3:145-155.Howell, J.T. 1957. The California flora and its province.Leafl. West. Bot. 8:133-138.Jain, S.K. and A.D. Bradshaw. 1966. Evolutionary divergenceamong adjacent plant populations. I. The evidence andits theoretical analysis. Heredity 21:407-441.Jepson, W.L. 1925. A manual of the flowering plants ofCalifornia. University of California Press, Berkeleyand Los Angeles.Johnson, D.E. and R. Ornduff. 1978. Lasthenia californica(Compositae), another name for a common goldfield.Madrono 25(4):227.Keck, D.D. 1959. Baeria, Crockeria, and Lasthenia. Pp. 1140-1144. In: A California Flora. P.A. Munz (ed).University of California Press, Berkeley and LosAngeles.Krebs, S.L. and J.F. Hancock. 1989. Tetrasomic inheritanceof isoenzyme markers in the highbush blueberry,Vaccinium corymbosum L. Heredity 63:11-18.Loveless, M.D. and J.L. Hamrick. 1984. Ecologicaldeterminants of genetic structure in plant populations.Ann. Rev. Ecol. Syst. 15:65-95.Nei, M. 1972. Genetic distance between populations. Amer.Natur. 106:283-292.Nei, M. 1973. Analysis of gene diversity in subdividedpopulations. Proc. Nat. Acad. Sci. USA. 70(12):3321-3323.Ness, B.D., D.E. Soltis, and P.S. Soltis. 1989.Autopolyploidy in Heuchera micrantha (Saxifragaceae).Amer. J. Bot. 76(4):614-626.Nuttall, T. 1841. Burriela. Trans. Am. Phil. Soc. II, 7:380-381.Oakeshott, G.B. 1971. California's changing landscapes. NewYork.100Odrzykoski, I.J. and L.D. Gottlieb. 1984. Duplications ofgenes coding 6-Phosphogluconate dehydrogenase inClarkia (Onagraceae) and their phylogeneticimplications. Syst. Bot. 9(4):479-489.Ornduff, R. 1966. A biosystematic survey of the goldfieldgenus Lasthenia (Compositae:Helenieae). University ofCalifornia Pub. 40:1-92.Ornduff, R. 1969. The origin and relationships of Lastheniaburkei (Compositae). Amer. - J. Bot. 56(9):1042-1047.Ornduff, R. 1976. Speciation and oligogenic differentiationin Lasthenia (Compositae). Syst. Bot. 1:91-96.Ornduff, R., B.A. Bohm, and N.A.M. Saleh. 1973. Flavonoidsof artificial interspecific hybrids in Lasthenia.Biochem. Syst. 1:147-151.Ornduff, R., B.A. Bohm, and N.A.M. Saleh. 1974. Flavonoidraces in Lasthenia (Compositae). Brittonia 26(4):411-420.Raven, P.H. 1977. The California flora. Pp. 109-137. In:Terrestrial Vegetation of California. M.G. Barbour andJ. Major (eds). John Wiley & Sons Inc. New York.Raven, P.H. and D.A. Axelrod. 1978. Origin and relationshipsof the California flora. Univ. Calif. Pub. Bot. 72:1-134.Reyment, R.A., R.E. Blackith, and N.A. Campbell. 1984.Multivariate Morphometric, Academic Press Inc., London.Robinson, H. 1981. A revision of the tribal and subtriballimits of the Heliantheae (Asteraceae). SmithsonianContr. Bot. 51:1-102.Rohlf, F.J. 1988. NTSYS-pc. Numerical taxonomy andmultivariate analysis system. Exeter Publishing Ltd,New York.Saleh, N.A.M., B.A. Bohm, and R. Ornduff. 1971. Flavonoidsof Lasthenia conjugens and Lasthenia fremontii.Phytochemistry 10:611-614.SAS Institute. 1985a. SAS User's Guide: Basics. 5'th ed. SASInstitute Inc., Cary, N.C.SAS Institute. 1985b. SAS User's Guide: Statistics. 5'th ed.SAS Institute Inc., Cary, N.C.Sneath, P.H.A. and R.R. Sokal. 1973. Numerical taxonomy. SanFrancicso. W.H. Freeman and Co.101Soltis, D.E, C.H. Haufler, D.C. Darrow, and G.J. Gastony.1983. Starch gel electrophoresis of ferns: Acompilation of grinding buffers, gel and electrodebuffers, and staining schedules. Amer. Fern J. 73(1):9-27.Soltis, D.E. and L.H. Rieseberg. 1986. Autopolyploidy inTolmiea menziesii (Saxifragaceae): genetic insightsfrom enzyme electrophoresis. Amer. J. Bot. 73(2):310-318.Stebbins, G.L. 1980. Polyploidy in plants: unsolved problemsand prospects. Pp. 495-520. In: Polyploidy: BiologicalRelevance. W.H. Lewis (ed.). Plenum Press, New York,London.Stebbins, G.L. and J. Major. 1965. Endemism and speciationin the California flora. Ecol. Monog. 35(1):1-35.Torrey, J. and A. Gray. 1842. Burriela. Flora N. Amer.2:378-379Vasey, M.C. 1985. The specific status of Lasthenia maritima (Asteraceae), an endemic of seabird-breeding habitats.Madrono 32(3):131-142.Watson, P.J. 1969. Evolution in closely adjacent plantpopulations. VI. An entomophilous species, Potentillaerecta, in two contrasting habitats. Heredity 24:407-422.Weeden, N.F. and J.F. Wendel. 1989. Genetics of plantisozymes, Chapt. 2. Pp. 47-72. In: Isozymes in PlantBiology. D.E. Soltis and P.S. Soltis (eds). DiscoridesPress, Portland, Oregon.Whitkus, R. 1988. Modified version of GENESTAT: a programfor computing genetic statistics from allelic frequencydata. P1. Genetics Newslett. 4:10.Wright, S. 1951. The genetical structure of populations.Ann. Eugen. 15:323-354.102

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0086380/manifest

Comment

Related Items