UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Taxonomic study of Trillium ovatum forma hibbersonii O’Neill, Darlene, M. 1995

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

Item Metadata


831-ubc_1996-0076.pdf [ 8.96MB ]
JSON: 831-1.0087047.json
JSON-LD: 831-1.0087047-ld.json
RDF/XML (Pretty): 831-1.0087047-rdf.xml
RDF/JSON: 831-1.0087047-rdf.json
Turtle: 831-1.0087047-turtle.txt
N-Triples: 831-1.0087047-rdf-ntriples.txt
Original Record: 831-1.0087047-source.json
Full Text

Full Text

TAXONOMIC STUDY OF Trillium ovatum  forma hibbersonii by  Darlene M . O'Neill B.Sc, The University of British Columbia, 1983 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE  in  THE FACULTY OF GRADUATE STUDIES Department of Plant Science  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA 1995 © Darlene M . O'Neill, 1995  In  presenting  this  degree at the  thesis  in partial  University of  freely available for reference copying  of  department  this or  publication of  fulfilment  British Columbia, and study.  thesis for scholarly purposes by  his  or  her  PuA*J"f  DE-6 (2/88)  requirements  I agree  that the  representatives.  may be It  this thesis for financial gain shall not  S c ^ M C 6  The University of British Columbia Vancouver, Canada  Date  the  I further agree  permission.  Department of  of  is  that  an  advanced  Library shall make it  permission for extensive  granted  by the  understood be  for  that  allowed without  head  of  my  copying  or  my written  11  Abstract The entity known as Trillium ovatum forma hibbersonii Taylor and Szczawinski was first discovered on the west coast of Vancouver Island in 1938 by J A . Hibberson (Holotype UBC 73131). Leonard Wiley invalidly published the taxon as Trillium hibbersonii in 1968. A review of the taxonomic status by T.M.C. Taylor and A.F. Szczawinski resulted in the publication of the taxon as Trillium ovatum forma hibbersonii (1975). The diagnosis stated this dwarf variant was part of the continuous series of intergradations of Trillium ovatum Pursh (Taylor and Szczawinski, 1975). Taylor and Szczawinski did not cite any small or intermediate samples that supported their conclusions. The present study was initiated to provide a comprehensive re-evaluation of the taxonomic status: separate species from Trillium ovatum; subspecies, variety or forma of Trillium ovatum']  This study considers morphological differences, flavonoid analysis and Random Amplified Polymorphic DNA (RAPD), from natural populations and cultivated plants. I conclude that compelling evidence supports the raising of Trillium ovatum forma hibbersonii  to specific status. Characteristics which support this conclusion are:  1) the limited geographical distribution of Trillium ovatum forma hibbersonii 2) different ecological niche and strong morphological divergence from Trillium ovatum 3) distinct dormancy/germination requirements 4) quantitative differences in the major flavonoid constituents of T. ovatum and T. ovatum forma hibbersonii.  Ill  Table of Contents Abstract  ii  Acknowledgements  *  List of Tables  vi  List of Figures  vii  1. Introduction  1  2. Distribution and Ecology  9  3. Materials and Methods  17  3.1 Cytology  17  3.2 Pollen Staining  17  3.3 Crossing Methods  18  3.4 Morphological and Development Data  18  3.5 Flavonoid Materials and Methods  19  3.5.1 Isolation and Purification Methods  23  3.5.2 Identification of Compounds  25  3.6 DNA Materials and Methods  26  3.6.1 Plant Genomic DNA purification  26  3.6.2 Agarose Gel Electrophoresis  28  3.6.3 Random Amplified Polymorphic DNA  29  4. Results and Literature Reviews  30  4.1.1 Cytology Literature Review  30  4.1.2 Cytology Results  31  IV  4.2.1 Pollen Literature Review  31  4.2.2 Pollen Results  33  4.3.1 Pollination and Seed Dispersal Literature Review  34  4.3.2 Pollination Results  35  4.4.1 Germination and Interspecific Crossing Literature Review  36  4.4.2 Germination and Interspecific Crossing Results  37  4.5 Morphology: Description  40  4.5.1 Trillium subgenus Trillium  40  4.5.2 Trillium ovatum Pursh ssp. ovatum  40  4.5.3 Trillium ovatum forma hibbersonii  44  4.6.1 Morphology: Comparison  49  4.6.2 Key to Western Trillium subgenus Trillium  57  4.6.3 Morphological Analysis  59  4.7.1 Flavonoid Literature Review 4.7.2 Flavonoid Results  65 .  68  4.8.1 Random amplified polymorphic DNA (RAPD) Literature Review  81  4.8.2 Random amplified polymorphic DNA (RAPD) Results  82  5. Discussion  87  5.1 Taxonomic Considerations  87  5.2 Phytogeny  90  6. Conclusions  93  7. Literature Cited  vi List of Tables Table 1. Collection locale of Trillium ovatum specimens examined.  10  Table 2. Source of leaf material for 2D chromatogram profiles.  20  Table 3. Source of leaf material forflavonoidanalysis.  21  Table 4. Voucher specimens.  22  Table 5. Source of leaf material for DNA analysis.  27  Table 6. Pollen stainability.  33  Table 7. Crossing data Trillium "hibbersonii" and T. ovatum.  37  Table 8. Comparison of quantitative characters of Trillium ovatum and Trillium "hibbersonii".  50  Table 9. Comparison of qualitative morphological characters of Trillium ovatum and Trillium "hibbersonii".  51  Table 10. Summary of Trillium ovatum and Trillium "hibbersonii"staX\stics.  52  Table 11. Canonical loadings of the variables used in the discriminant analysis of Trillium "hibbersonii" and T. ovatum.  64  Table 12. Flavonoids identified in Trillium "hibbersonii".  11  Table 13. Ultraviolet-visible absorption spectral data.  78  Table 14. Shift reagent spectral data.  79  Table 15. List of primers used and the respective oligonucleotide sequence. Trillium species which showed any amplification products are listed. H - T. "hibbersonii", O - T. ovatum, G - T. grandiflorum, St - T. stylosum, K-T. kamtschaticum.  83  Vll  List of Figures Figure 1 . Distribution of pedicellate Trillium in western North America.  3  Figure 2. Range of Trillium ovatum and T. ovatum ssp. oettingeri based on herbarium collections seen.  5  Figure 3. Range of Trillium "hibbersonii".  7  Figure 4: Typical site of Trillium ovatum.  11  Figure 5. Hesquiat Lake site.  15  Figure 6. Trillium "hibbersonii" at Hesquiat Lake site.  15  Figure 7. Hesquiat Lake site, Trillium "hibbersonii" growing in rock cleft.  16  Figure 8. Chromosome set from root tip squash of Trillium "hibbersonii".  32  Figure 9. Chromosome set from root tip squash of Trillium "hibbersonii".  32  Figure 10. Sketch of development of Trillium "hibbersonii" seed, radicle and rhizome.  39  Figure 11. Trillium ovatum in flower, Native Garden, UBC Botanical Garden.  41  Figure 12. Aged flower of Trillium ovatum Native Garden, UBC Botanical Garden.  41  Figure 13. Illustration of corolla shapes in Trillium ovatum, T. "hibbersonii", T. grandiflorum  and T. nivale.  43  Figure 14. Trillium "hibbersonii" in the North American Garden, UBC Botanical Garden.  45  Figure 15. Trillium "hibbersonii".  45  Figure 16. Aged flower of Trillium "hibbersonii" is dark pink.  46  v'ui  Figure 17. Anthers in Trillium "hibbersonii" in which the pollen sacs did not develop.  46  Figure 18. Sketch of abortive anthers of Trillium "hibbersonii".  47  Figure 19. Trillium "hibbersonii" grown in cold frame, UBC Botanical Garden.  53  Figure 20. Dark pink petals of agedflowerof Trillium "hibbersonii".  53  Figure 21. Dehiscent anthers of Trillium ovatum.  55  Figure 22. Trillium ovatum in April, North American Garden, UBC Botanical Garden.  56  Figure 23. July photograph of same group of Trillium ovatum as in Fig. 22.  56  Figure 24. Trillium "hibbersonii" mature capsule and cross-section of immature capsule.  58  Figure 25. Trillium ovatum mature capsule and cross-section of immature capsule.  58  Figure 26a. Boxplot comparisons of height, pedicel length, petal and sepal measurements from Trillium "hibbersonii" and T. ovatum.  60  Figure 26b. Boxplot comparisons of leaf measurements and leaf ratios from Trillium "hibbersonii"  and T. ovatum.  61  Figure 26c. Notched boxplots are used for comparison of the medians of the ten measurements and two ratios for T. "hibbersonii" and  T. ovatum.  Figure 26d. Scatterplot of distance scores obtained from discriminant analysis.  62  64  Figure 27. Flavonoid biosynthetic pathway andflavonoidwith carbon numbering system.  66  ix  Figure 28. 2-dimensional profile of Trillium "hibbersonii" vacuolar flavonoids.  70  Figure 29. 2-dimensional profile of Trillium ovatum vacuolar  70  flavonoids.  Figure 30. 2-dimensional profile of Trillium rivale vacuolar flavonoids.  71  Figure 31. 2-dimensional profile of Trillium grandiflorum vacuolar flavonoids.  72  Figure 32. 2-dimensional profile of Trillium nivale vacuolar  72  Figure 33. 2-dimensional profile of Trillium stylosum vacuolar Figure 34. 2-dimensional profile of Trillium sessile vacuolar  flavonoids. flavonoids.  73  flavonoids.  73  Figure 35. Trillium "hibbersonii" fractions from Sephadex LH-20 Column 3.  75  Figure 36. Trillium ovatum fractions from Sephadex LH-20 Column 2.  76  Figure 37. Amplification of Trillium "hibbersonii" and T. ovatum DNA using primer #6.  86  Figure 38. Amplification of Trillium "hibbersonii" and T. ovatum DNA using primers #6, #100, #70, #60. Figure 39. Hypothesis for the phylogeny of Trillium based on Berg (1958).  86 91  X  Acknowledgements I would especially like to thank Dr. Gerald Straley for suggesting this project, believing that I could do it then patiently waiting. Thanks to the Flora of North America Project and the Henry Eddie Foundation for funding and the UBC Botanical Garden for the collecting trip to the type locality. Peter Buckland provided unforgettable hospitality at Boat Basin. I am grateful to Dr. Brian Ellis for lab space, materials and all his knowledgeable graduate students and technicians who answered all my questions. Dr. Bruce Bohm also supplied lab space, materials and wonderful grad students. I would like to thank the Curators of the following herbaria for use of their facilities and specimen loans: UBC, V, MO, RSA, UCB, UW. I would like to thank all the people at the Botanical Gardens and in the Department of Plant Science who helped just by being themselves. The close-up photography by David Wilson (DW) was greatly appreciated/Thanks to Lynn Noble for allowing me to use her collection in this study and for her art and enthusiasm. Karen helped with proof reading and knows more about trillium than any CA. Finally, thanks to Michael for putting up with everything.  1  1. Introduction In  Species Plantarum o f 1753 Linnaeus first recognized and published the genus Trillium  and described three North American species, T. cernuum, T. erectum and T. sessile. The characteristic features o f the genus are the whorl o f three leaves (also referred to as leafy bracts (Freeman, 1975; Murrell, 1969)) subtending a terminal, solitary, trimerous flower which is either sessile or pedicellate. Trillium is generally placed in the family Liliaceae (Cronquist, 1981; Takhtajan, 1969), more specifically in the tribe Parideae sensu Krause (1930) along with Paris, Medeola and Scoliopus. The genus Trillium is sometimes placed in its own family Trilliaceae and distinguished from Liliaceae by the whorled rather than alternate leaves. The family Trilliaceae was first proposed by John Lindley in 1836. Takhtajan (1986) revised his earlier conclusions and Hutchinson (1959, 1973) placed Trillium in its own family based on chemical evidence. Trillium has also been placed in another segregate family, Convallariaceae (Abrams, 1923, Thomas 1961). The genus Trillium is composed o f approximately 30 to 50 species (depending on taxonomic interpretation) which occur in two regions o f the temperate, northern hemisphere; North America and eastern Asia. North America is further subdivided into east and west regions as no Trillium species are found in central North America. None o f the species are native to more than one region, either Asia and North America or east and west N o r t h America. The genus Trillium is divided into two subgenera, the pedicellate-flowered (Trillium) and the sessile-flowered (Phyllantherum R a f ) . Members of the subgenus Phyllantherum occur only in North America with representatives in both east and west regions. In total, there are approximately eight species in western North America, two pedicellate species (Figure 1) and  . 2 five of the subgenus Phyllanlhemm.  In eastern North America there are approximately 35  species in total, 13 pedicellate-flowered and 22 sessile-flowered (Freeman, 1975). Major taxonomic works that contribute to our knowledge of the pedicellate Trillium species are by Gates (1917), Gleason (1906) and Rendle (1901). A chemotaxonomic study of the sessile flowered Trillium species (Murrell, 1969) provides a biochemical profile of the phenolics with reference to Freeman's (1969) treatment which is based on morphology and distribution. A revision and description of the sessile-flowered species was published by Freeman (1975). Trillium ovatum  Pursh belongs to the subgenus Trillium and has a range from southern British  Columbia east to Montana and south to Colorado and California (Figure 2). The type locality for T. ovatum is on the Columbia River at the foot of the Cascade Mountains. The species was named by Frederik Pursh, from material collected on April 10, 1806 by Meriwether Lewis during the Lewis and Clark expedition (Pursh, 1814). Considering the extent of the range and the diverse environments that it encompasses it is not unexpected that T. ovatum is a polymorphic species. Synonyms for Trillium ovatum are as follows: T. obovalum  Hook. Fl. Bor. Am. 2:180. 1840  T. californicum  Kellogg, Proc. Cal. Acad. 2:50. 1860  T. crassifolium  Piper, Erythea 7:104. 1899  T. scouleri  Rydberg, Bull. Torrey Club 33:394. 1906  T. ovatum  var. stenosepalum Gates, Ann. Mo. Bot. Gard. 4:61.1917  T. venosum T. ovatum  Gates, Ann. Mo. Bot. Gard. 4:66. 1917  ssp. oettingeri Munz & Thorne, Aliso 8(1): 15. 1973  3  C^AN AL) A  Range of western Trillium species  Figure 1. Distribution of pedicellate Trilliums in western North America. Enlarged area shows approximate ranges of Trillium "hibbersonii", T. ovatum, T. ovatum ssp. oettingeri and T. rivale.  4 Gates (1917) noted that a continuous series of intermediate forms existed between T. ovatum, the variety stenosepalum and T. venosum where any of the three overlapped in geographical range. The variety stenosepalum was described by Gates (1917) as a "transitional" variety differing from var. ovatum in the narrowness of the sepals (3.5-6.0 mm wide) and the inconspicuous sepal veins. The range was outlined as western Montana and southern Washington to central California. In 1948 S.J. Smith examined the type specimen of stenosepalum  (MO 1825559) and determined that no segregation of this variety was  warranted. In the Flora of California (Munz, 1959; Hickman, 1993) the status of variety was removed from stenosepalum and placed in the list of synonyms. The 1949 Smith annotation on the T. venosum type specimen (MOl825569) indicates that it is within the range of variation of T. ovatum rather than a separate species. The five species listed as synonyms for 7! ovatum, as described, do not affect this study other than providing examples of the variation within the species. None are possible descriptions of Trillium ovatum  Pursh forma hibbersonii Taylor & Szczawinski.  In 1973 a new taxon was collected and described from the mountains of northern California, T. ovatum  ssp. oettingeri (Munz & Thorne, 1973). Subspecies oettingeri differs from  subspecies ovatum principally by the distinctly petiolate smaller leaves and its limited geographical range (Figure 2). It also differs in the flower attitude, which is more or less nodding, the shorter and narrower petals, shorter peduncle, and overall smaller stature (Hickman, 1993; Munz & Thorne, 1973). The stem height ranges from 10 to 25 centimetres, flower colour is white at anthesis and leaves are distinctly petiolate. The type specimen (RSA 234713) was examined by the author and found not to affect this study.  Figure 2. Range of Trillium ovatum and T. ovatum ssp. oettingeri based on herbarium  collections seen. Shaded areas indicate county or locale of specimens included in this study.  6  The entity now known as Trillium ovatum forma hibbersonii was first discovered on the west coast of Vancouver Island, on April 25, 1938 by Jack Arthur Hibberson (Holotype UBC 73131). The type locality for this form is near Boat Basin, Hesquiat Harbour (Figure 3). Although forma hibbersonii  has gained popularity as a garden plant it has received very little  attention in the scientific community. Gardeners who have grown the plant have always believed that it is a "good" species. Leonard Wiley considered this dwarf to be a distinct species and named it Trillium hibbersonii in Rare Wild Flowers of North America (Wiley, 1968). This is a nomen nudum as Wiley failed to follow the International Code of Botanical Nomenclature (ICBN) requirements for publication of a new name. Specifically, he did not validly or effectively publish the name with a Latin description in a journal normally used by taxonomists. Also, a nomenclatural type was not cited. In 1974 T.M.C. Taylor and A.F. Szczawinski reviewed the taxonomic status of the dwarf variant. The opinion of Taylor and Szczawinski was that this dwarf variant, although more frequent on western Vancouver Island, was simply part of the continuous series of intergradations of T. ovatum (Taylor & Szczawinski, 1975). Although not in favour of attaching names to forms, Taylor and Szczawinski subsequently published the new name, Trillium ovatum forma  hibbersonii,  according to ICBN requirements. They felt it was justified in this case as the form was commonly exhibited at garden shows incorrectly labelled as T. hibbersonii. Taylor and Szczawinski did not cite any of the small or similar samples that they based their judgement upon. There is no documentation of a visit by Taylor and Szczawinski to the type locality prior to their determination. In 1958 they collected this form on the banks of the Clennyuck River, Kyuquot Sound (UBC Herbarium 76310). Two collecting trips to Boat Basin were  7  Figure 3. Range of Trillium "hibbersonii". The type locality and collection sites of all known herbarium specimens are shown. This distribution indicates a narrow endemic entity restricted to Vancouver Island, British Columbia.  8  reported by T.C. Brayshaw (pers. comm.) in 'Notes on some threatened or endangered plant species of British Columbia, from records at the Royal British Columbia Museum', however he also noted that there were no specimens of T. ovatum f.hibbersonii in the provincial herbarium. A visit to the Royal B.C. Museum Herbarium confirmed the lack of specimens. According to the herbarium curator, J. Pinder-Moss, many specimens collected by A. Ceska, a co-author of the Hesquiat Lake Report No. 358, have not been accessioned to this date although Ceska (pers. comm.) stated that no herbarium specimens were collected. Wiley (1968) reported that Szczawinski and T.M.C. Taylor and others re-collected T. ovatum f. hibbersonii in 1956 at an unspecified location on the west coast of Vancouver Island.  For brevity and clarity Trillium ovatum forma hibbersonii is referred to as T. "hibbersonii" throughout this text. The obvious morphological differences that characterize T. "hibbersonii" are the dwarfing of all its parts, flower colour pink in bud through anthesis, position of the stigmatic surface above the anthers, presence of pigmentation on the anther connective, narrow lanceolate leaves, and the flower attitude due to the bend of the pedicel. The present study was initiated to provide a comprehensive re-evaluation of the taxonomic status: separate species from Trillium ovatum; subspecies, variety or forma of T. ovatum! This study considers morphological differences, flavonoid analysis and Random Amplified Polymorphic DNA (RAPD), from natural populations and cultivated plants.  9 2. Distribution and Ecology The approximate range of T. ovatum extends from 50° to 36° latitude, southern British Columbia through Washington and Oregon to Monterey County (Santa Cruz Mountains), California, east to the Rocky Mountains of British Columbia and south through Idaho and western Montana into Wyoming and Colorado. The only population found to occur in Alberta is located in the southeast corner of the province in Waterton Lakes National Park (Kuijt, 1982). Populations of the Pacific Coast and the Rocky Mountains are separated by the Columbia River Plateau of the Harney Basin, centred in Washington and Oregon, and the Great Basin lying to the south. (See Table 1 & Figure 2). The habitat of 71 ovatum generally consists of moist, shady sites under open or dense forest with acidic soils (Figure 4). Colonization of open meadow or bare mountain slopes is limited by soil moisture and mild daytime temperature requirements in the spring growing season (Wiley, 1968). Pacific coast populations of T. ovatum are found in Coastal Western Hemlock (Tsuga heterophylla)  and Douglas Fir (Pseudotsnga menziesii var. menziesii) biogeoclimatic  zones (BCMF, 1988) and Redwood (Sequoia sempervirens) forests in California. Rocky Mountain populations inhabit Interior Cedar-Hemlock (Thujaplicata,  Tsuga  heterophylla)  and Interior Douglas Fir (Pseudotsnga menziesii var. glauca) zones. Under more arid conditions Trillium populations are also found in moist zones along streams associated with Ponderosa Pine (Pinus ponder osd) and Lodgepole Pine (P. con tor ta) (Fukuda and Channel, 1975). Trillium ovatum is used as an indicator plant in coastal British Columbia for fresh (young) and very moist nitrogen-rich soils with Moder and Mull humus forms (Klinka et al., 1989). These soil types are associated with friable forest floors.  10  Table 1. Collection locale of Trillium British Columbia # Specimens 246 incl. West 67 Van. Is. 2 Gulf Is. 28 East California  Del Norte Humboldt Marin Mendocino Monterey San Mateo Santa Clara Santa Cruz Siskiyou Sonoma Trinity Unknown  8 18 10 15 3 30 1 10 41 1 2 7  Colorado  Jackson Routt  1 5  Idaho  Adams Boise Bonner Clearwater Idaho Latah Nez Perce Valley Washington  6 4 13 11 9 57 3 8 2  Montana  Beaverhead Flathead Gallatin Jefferson Lake Lewis & Clark Mineral Missoula Unknown  1 6 8 1 1 2 6 52 1  ovatum  specimens examined. Oregon # Specime Baker 3 Benton 4 6 Clackamas 1 Clatsop Coos 1 Hood River 5 Jackson 3 14 Josephine Lane 3 Linn 2 Marion 2 Multnomah 2 2 Wallowa Washington 2 Yamhill 1 Washington Asotin' 14 Chelan 7 Clellam. 2 Cowlitz 1 3 Grays Harbor Jefferson 3 King 34 3 Kitsap Kittitas 3 4 Klickitat 1 Lewis Mason 4 10 Pierce Skamania 1 Snohomish 11 6 Spokane Thurston 10 Whatcom 1 Whitman 3 3 Yakima Wyoming Carbon. 7  11  12  Herbarium collections indicate only two areas where T. "hibbersonii" populations occur. Both are on the extreme west coast of Vancouver Island within the zone mapped as Coastal Western Hemlock zones (BCMF, 1988) (Figure 3). In Figure 3 the shaded area indicates areas of similar terrain and zoning, all with difficult access where T. "hibbersonii" may potentially occur. In the B.C. Vascular Inventory (Taylor and MacBryde, 1977) and in Rare Vascular Plants of B.C. (Straley el al., 1985) T. "hibbersonii" is listed as endemic. The holotype (UBC 73131) of Trillium "hibbersonii" was collected at 610 metres elevation within Ecological Reserve #358, Hesquiat Harbour, Clayoquat Sound (Figure 3). The reserve report includes a vegetation and site survey . The area is comprised of three vegetation communities: I.  Danthonia intermedia - Cladonia spp  II. Pinus contorta/Tsuga heterophylla - Vaccinium ovatum - Hylocomium splendens III. Tsuga heterophylla/Thuja plicata - Vaccinium alaskaense/Gaultheria shallon - Blechnum spicant. The Danthonia intermedia - Cladonia spp. community, which includes Trillium  "hibbersonii', is found on rock outcrops with a very shallow folisol (organic layer) and some seepage early in the season. Similar vegetation communities have not been surveyed (Ceska et al, 1982). The second area where T. "hibbersonii" is documented is in Kyuquot Sound where a population occurs along the Clennyuck River (UBC Herbarium 76310) (Figure 3). Within Ecological Reserve #358 at Hesquiat Harbour another population was located at less than six metres elevation and sampled by the author. There was no soil development on the rocky  13 collection site. Moisture seepage zones were apparent although probably seasonal. The site aspect was south to southwest and partly exposed to exposed (Figures 5, 6, & 7). Saxifraga ferruginea,  Cryptogramma acrostichoides (C. crispd), Huperzia  haleakalae  (Lycopodium selagd)  and Vaccinium ovatum were, collected at the Hesquiat Harbor sampling  site along with 71 "hibbersonii". None of these species are indicators of the environment favoured by 71 ovatum as outlined by Klinka et al. (1989). The environment indicated by these species is that of moderately dry to very dry nitrogen-poor fresh soils, Mor humus forms (typical of acidic and compacted forest floors) and shallow to very shallow soils over exposed coarse fragments and bedrock (Klinka et al., 1989). No specimens of 71 ovatum were located on the portion of the reserve surveyed by the author. A third population of 71 "hibbersonii" is located on a series of cliffs overlooking Rae Lake, three kilometres southwest of the Hesquiat Lake reserve. This population was located by Hans Roemer (pers. comm.) by identifying steep cliff zones on a topographic map prior to going out into the field. Garden cultivated 71 "hibbersonii" used in this analysis were from different sources and grown under various conditions: pots in cold frames, shady garden and exposed garden conditions. These differing sites and conditions did not alter the growth habit from the habit of the wild population located at Hesquiat Harbor. Notable especially in the UBC Botanical garden collections is the clump forming habit. Seedlings form a portion of the clump as do vegetative offshoots. Trillium ovatum  is protected by the Dogwood, Rhododendron and Trillium Protection Act  (1960). No plants were destroyed as part of this research. Samples consisted of one leaf or  14 less and any larger samples were collected from garden cultivated plants. Thefiveherbarium samples collected for documentation did not include the rhizomes.  15  Figure 5. Hesquiat Lake site. Steep slopes with thin organic layer overlying rocks. Pinus contona is in the foreground and Pseudotsuga menziesii in the midground.  Figure 6. Trillium "hibbersonii" at Hesquiat Lake site. Folding or curling of the leaves about the midline is common. Moss and lichens are abundant. An organic layer covers the underlying bedrock and rubble. Moisture due to seepage.  16  17 3. Materials and Methods 3.1 Cytology  Chromosome squashes were prepared from root tips collected from immature plants of T. "hibbersonii",  at the three-leaf stage. Plants were refrigerated for 3 days prior to removal of 5  mm of new root growth. The root tips were fixed in 6:3:1 solution of 70% ethanol: glacial acetic acid: chloroform for a minimum of one week. The standard aceto-carmine stain technique was used. Approximately 1 mm of each root tip was placed on a slide with one to two drops of aceto-carmine. A rusty nail, which acts as a mordant, was used to macerate the tissue until the colour changed from bright red to a purple-red. Hoyer's medium (Beeks, 1955) was added on a 1:1 ratio to aceto-carmine and mixed into the macerated material. A cover slip was set in place and pressure applied on the slide to spread the chromosomes. A period of approximately one week was required for the cytoplasm to clear. The chromosomes were observed at late prophase to early metaphase at 400X magnification . 3.2 Pollen Staining  The pollen samples of T. "hibbersonii" were collected from the UBC Botanical Garden collections and from herbarium specimens. Pollen samples from live and herbarium specimens were also collected from T. ovatum. Stamens of live specimens were removed from the plant and the pollen grains transferred to the slide from the dehiscing anther. Pollen from herbarium specimens was collected by inverting the specimen over a slide and gently tapping the anther with forceps. Two drops of aniline (cotton) blue dissolved in lactophenol (Maneval, 1936) were placed on the slide and covered with a glass slip. Approximately 24 hours was allowed for the stain to permeate the tissues prior to examination.  18 3.3 Crossing Methods A limited number of inter-taxa cross-pollinations were made. Plants were emasculated and bagged when sepals were beginning to recurve and prior to petals unfurling. Dehiscing anthers were collected and the pollen was dispersed onto the stigmatic surface of the mother plant. Pollen not used immediately was stored in a desiccator. In both taxa the anthers were not dehiscent for the first two or three days after the flower opened although the stigma appeared to be receptive. 3.4 Morphological and Development Data Flower and vegetative character measurements were made on both fresh and herbarium specimens. Shrinkage of specimens under proper drying conditions was found to be uniform, with length/width ratios and base to tip/base to centre ratios essentially equal for dried and fresh specimens (Duncan and Calhoun, 1962). Ten measurements were made and indicators of leaf shape (length/width ratio and base to tip ratio/ base to widest) were calculated from this data. Length and width measurements were made on T. "hibbersonii" seeds. Qualitative assessment was used for rating placement of the receptive portion of the stigma in relation to the anther (either above or below the level of the anther), ovary shape, pedicel attitude, fusion or non fusion of sepal base. Flower colour was rated in bud and at anthesis using the Royal Horticultural Society Colour Scheme. The colours of aged flowers, especially those of T. ovatum, varied widely and were assigned a colour range. Location of anthocyanin accumulation in sepals, pedicel, leaves and scape was also noted. Two or more sets of measurements were taken from selected individuals under cultivated conditions. The measurements were taken after emergence and prior to anthesis, at anthesis  19 and one month after anthesis or at petal withering. These measurements were used to determine if leaf shape remained constant (length/ratio, base to widest/base to tip) during seasonal development and to indicate the amount of leaf, sepal and petal expansion and height changes. For statistical analysis the data were analyzed using SYSTAT (Wilkinson, 1990). Notched boxplots for median and range comparison and scatterplots for discriminant analysis results were plotted using SYGRAPH (Wilkinson, 1990). 3.5 Flavonoid Materials and Methods Leaf material for the 2D profiles in the overview was collected from the UBC Botanical Garden (Table 2). These specimens were selected for the profiles because all except Trillium rivale  were growing under similar soil conditions and site aspect within a ten square meter  area. By using samples grown under similar conditions the potential effects of environment on flavonoid profile variation are minimized. Although no specimens of 7! rivale are in the UBC Botanical Garden collection it is important to include this species for comparison for two reasons: it is a western North American pedicellate species whose range overlaps that of T. ovatum,  and it shows a putative relationship with T. ovatum (Gates, 1917; Berg, 1958).  Leaf material for the individual, population and juvenile profiles was collected from wild and garden cultivated populations as outlined in Table 3. Voucher specimens have been deposited in the UBC Herbarium (Table 4).  20 Table 2 . Source of leaf material for 2D chromatogram profiles University of British Columbia Botanical Garden, North American Section, Alpine Garden. Lat: 49° 16*N. long: 123° 14'W. Collection Date: 06 May 1993 Taxon  Accession No  T. ovatum T. "hibbersonii" T. grandiflorum  014331-0156-74 010698-0156-74 006043-0055-73 or 013379-0327-76 013380-0327-76 003049-0117-71 014181-0156-74 .  T. stylosum T. nivale T. sessile T. sessile var. rubrum T. rivale  Leaf Weight grams 0.1782 0.1886 0.1867  014983-0273-77 Josephine Co. Oregon  Collection Number 296 297 298  0.1097 0.1861 0.1120  300 299 302  0.1490  301 6448, 6441 G.B. Straley  21 Table 3. Source of leaf material for flavonoid analysis  Date  Location  Number Individ.  2d of Individ.  2d of Pop.  2d of Juvenile  LH-20 Column  T. ovatum : Wild Populations  06/16/91 Chilliwack 06/16/91 Chilliwack 07/22/92 Chilliwack 04/30/91 Langley 04/30/91 Langley 07/22/92 Langley 04/30/91 Lynn Valley 04/30/91 Lynn Valley 05/02/91 Cathedral Gr 05/02/91 Cathedral Gr 05/19/91 Hamilton Sw 04/13/92 Hamilton Sw  3 36  *  2 8 28 2 8 17 1 19 21 25  * *  * *  * * *  * * *  * * *  *  2 2 2 2 2 2 2 2 2 2 2 1  T. ovatum : U B C Botanical Garden Plants  05/23/91 Asian Garden 08/02/91 Asian Garden 07/02/92 Asian Garden 06/14/91 B C & N Am  5  *  3  *  5 9  *  2 2 2 2  T. "hibbersonii": Wild Populations and Garden Plants  05/16/91 UBC Cold Fr. 5 * * * 3 06/14/91 UBC N Am 8 * * * 3 03/28/92 Noble Garden 2 * 3 05/21/92 UBC Cold Fr. 2 3 06/14/91 UBC N Am 3 3 05/11/92 Hesquiat (wild) 18 * * 3 UBC Botanical Garden includes: Asian Garden, B.C. Native Garden, North American section of the AlpineGarden and the cold frames. Cathedral Grove, Van. Is. Hamilton Swamp, Van. Is Hesquiat Harbor, Van. Is. Chilliwack Lake Langley Lynn Valley Noble Garden, West Vancouver  22  Table 4. Voucher specimens Specimens are located in the U.B.C. Herbarium Location  Latitude  Longitude  Date  Herbarium number  Trillium "hibbersonii"  Hesquiat Hbr  49°30'N  126°23'W  04/11/92  205459  T. ovatum  Hamilton Swamp Lynn Valley Langley Chilliwack River  49° 16'N 49°22'N 49°06'N 49°16'N  124°29'W 123°03'W 122 34'W 124°29'W  05/19/91 05/11/93 05/19/91 07/08/93  206841 207595 206841 in process  0  23 3.5.1 Isolation and Purification Methods  Isolation and purification methods used for flavonoid constituents were those described by Wilkins and Bohm (1976) and modified by Gornall and Bohm (1980), with only minor modifications. Initial extraction of fresh plant material was in 100% methanol and 80% methanol for dried plant material. Leaf material was soaked overnight or until no further colour was extracted. Solvent was changed several times as required. The total extraction of chlorophyll was used as an indication of complete flavonoid extraction. Small samples were spotted on Polyamide DC-6.6 plates for 2 dimensional thin-layer chromatography (TLC) for initial comparisons of individual and population profiles (intrataxon variation) as well as for inter-taxa variation. Individual samples and populations of T. ovatum  and T. 'hibbersonii' were bulked separately when no intra-taxon variation was found  in the 2D profiles. The methanolic extract was evaporated to dryness at 40° C using a rotary evaporator. The evaporation of methanol leaves a green viscous residue which was then extracted with boiling water and filtered using a Buchner funnel under vacuum, filter paper and Celite 504. The water-soluble substances, including flavonoids, are in the filtrate. A separatory funnel was utilized for the repeated extraction of the aqueous filtrate with water saturated n-butanol. Ammonium sulphate [(NH ) S0 ] was also added to assist in separation 4  2  4  by changing the ionic balance of the solution. The n-butanolic extracts were combined and evaporated. The phenolic fraction remaining as residue was then taken up in approximately 5 mis of 100% methanol for storage in vials.  24 Resolution of the phenolic fraction into constituent flavonoid glycoside classes was accomplished using Sephadex LH-20 column chromatography. Three Sephadex LH-20 columns were run using bulk samples as outlined in Table 3. The LH-20 gel (45 grams) was allowed to swell in 30% methanol in the column prior to addition of the phenolic fraction. The methanol concentration of the phenolic fraction was adjusted to 30%, equivalent to the concentration of the primary column elution solvent. Excess solvent was eluted from the column prior to the addition of the phenolic fraction. A pipette was used to place the phenolic fraction in a narrow band on top of the LH-20 gel. Elution was started using a 30% methanolwater solvent and proceeded with methanol-water mixtures increasing by increments of 10%. Ultraviolet (UV) light (336 mn) was used to monitor the fractionation. Collection of the fractions began when the first fluorescent band (blue) of material began to elute. Fractions collected were evaporated to dryness in vacuo before storing in 1 to 2 mis of 100% methanol. The final step after elution with 100% methanol was to wash the column with 2 aliquots of acetone. One to five drops of each fraction collected from column chromatography was spotted on Polyamide DC-6.6 plates. One-dimensional chromatography of the fraction sets were run separately in the aqueous solvent (water-butanol-acetone-dioxane 70:15:10:5) and the organic solvent (dichloroethane-methanol-methylethyl ketone-water 50:25:21:4). The aqueous solvent system separates the aglycones, monoglycosides, diglycosides and triglycosides. The organic solvent system separates individual compounds within each glycosylation class, ie. kaempferol, quercetin and myricetin glycosides of the same glycoside class. Standards were also run with the fraction sample sets.  25 Once the compounds were resolved, the individual compounds in each fraction were separated and purified using preparative thin-layer chromatography (TLC) and the solvents described above. Related fractions with the same chromatographic behaviour were combined prior to the purification procedure to minimize loss of material during purification. 3.5.2 Identification of Compounds  The identification of flavonoid compounds was based on chromatographic behaviour on TLC plates under UV light, both with and without ammonia vapour, and the colour reaction with diphenylboric acid ethanolamine complex spray (Mabry et al, 1970; Markham, 1982; Wilkins & Bohm, 1976). Standards were also used for comparison and identification of chromatographic behaviour under the conditions described. UV absorption spectral data from a methanol solution of the flavonoid and the use of shift reagents permitted identification of the aglycone and provided information on the position of hydroxyl groups and sugar(s) by comparison with reference spectral data (Mabry et al., 1970; Markham, 1982). Complete hydrolysis of all samples was carried out, since the limited amount of sample did not permit subsampling. The purified fraction (in 100% methanol) was diluted with distilled water and 5 drops 1M trifluoroacefic acid (TFA). The solution was placed in an 80-100° C. waterbath for 60 minutes. After 20 minutes, an additional 5 drops TFA was added. Distilled water was added throughout the hydrolysis process to maintain liquid volume. The hydrolysed sample was partitioned four times using water-saturated ethyl acetate. The sugars remain in the water-methanol phase while the aglycone phenol is soluble in the ethyl acetate. The sugarcontaining aqueous phase was evaporated to dryness using a rotary-evaporator. The sugar was then taken up with a minimum amount (1-2 drops) of 1:1 methanol-water.  26 The sugar solution was spotted on Polygram Sil G commercial sheets along with standards containing equal amounts of rhamnose, arabinose, galactose, xylose and glucose. The chromatograms were developed with CHCl -MeOH-H 0 (16:9:2) for approximately two 3  2  hours in a small tank. The chromatogram was dried before spraying with thymol reagent (0.9 g thymol, 19 ml 95% ethanol, 1 ml cone. H S0 ) and again allowed to dry thoroughly. The 2  4  chromatogram was then placed silica gel side down on a warm hot plate, in a fume hood, to develop the colours and locate the sugars. A glass plate placed on top of the chromatogram ensured more even heat distribution. The aglycones were evaporated to dryness and taken up in minimal amounts of 100% methanol for storage. The aglycones were spotted on a Polyamide DC-6.6 plate along with standards (quercetin, rutin, kaempherol-3-O-glucose, kaempherol-3-O-galactose and hydrolysed rutin) and run in the aqueous solvent. This procedure is used to confirm that hydrolysis was complete and allows for confirmation of compound identification by comparison of spot colour to standards as described above. 3.6 DNA Materials and Methods 3.6.1 Plant Genomic DNA purification The DNA extraction method of Rogers and Bendich (1988) is based on a CTAB (cetyltrimethylammonium bromide) procedure for use with milligram amounts of fresh, herbarium and mummified plant tissues. This procedure, tested by the authors on more than 60 different tissue types from more than 30 species, does not require cesium chloride density gradient techniques, and thus significantly reduces time and expense, although DNA size may  27  '  not be maximized (Escote-Carlson, 1991). The sources of plant material are outlined in Table 5. Milligram portions of leaf material that had been stored at -20 C were ground in 0  Table 5. Source of leaf material for D N A analysis.  University of British Columbia Botanical Garden Lat: 49° 16'N. long: 123° 14'W. Location Leaf Weight Taxon or Accession # mg 1184 B.C. Native T. ovatum 964 N.American T. grandiflorum T. stylosum 013380-0327-76 3.1 003049-0117-71 3.1 T. kamtschaticum 3.1 014181-0156-74 T. sessile T. "hibbersonii"  92.0 Hesquiat Hbr. 49°30'N 126°23'W  Sample Label O G St  K S  H  microfuge tubes using plastic pestles. Samples of T. grandiflorum and 7'. ovatum tissue were ground in a mortar as they did not break down quickly enough in the microfuge tubes. Liquid nitrogen was used in place of dry ice during the grinding procedure. Hot (65 ° C) 2x CTAB (2% w/v CTAB, 100 mM Tris pH 8.0, 20 mM EDTA pH 8.0, 1.4 M NaCl, 1% PVP (polyvinylpyrrolidone MW 40000) was added to the ground material. Yeast tRNA, which acts as a carrier, was added to all samples smaller than 70 mg fresh weight. An emulsion was formed by adding one volume of chlorofonn/isoamyl alcohol (24:1) and mixing thoroughly. This was followed by 30 seconds centrifugation at 11,000 g. The supernatant solution was transferred to a new microfuge tube, while the chloroform phase (lower) and cellular debris were discarded. One tenth volume 10% CTAB (10% w/v CTAB, 0.7 M NaCl) at 65° C was added and another chloroform/isoamyl alcohol extraction performed as above. An equal volume of CTAB precipitation buffer (1% w/v CTAB, 50 mM Tris pH 8.0, 10 mM EDTA pH  28 8.0) was added to the supernatant in a clean microfuge tube and gently mixed. Nucleic acid precipitation was not observed immediately, so the precipitation was allowed to proceed for 15 minutes on ice in a 6 C refrigerator. All samples were centrifuged for 30 seconds 0  (11,000 g), except G which required 60 seconds for pellet formation. After removal of the supernatant, high-salt TE buffer (10 mM Tris pH 8.0, 1 mM EDTA pH 8.0, 11 M NaCl) was used to rehydrate each sample. Two volumes cold 95% ethanol was added to each sample and mixed gently before incubation at -20 C for 10 minutes. DNA was pelleted by centrifugation 0  for 10 minutes at 11,000 g. The supernatant was discarded and one volume cold 80% ethanol was added. Samples were centrifuged for another 5minutes. The DNA samples were dried in a desiccator for 30 minutes before rehydration in 0.2x TE buffer (1.0 mM Tris pH 8.0, 0.1 mM EDTA pH 8.0). RNase A at 100 ug/ml was added before incubation at 37° C for 1 hour to eliminate RNA. For all samples, the amount of DNA extracted was quantified using the GeneQuant RNA/DNA Calculator by Pharmacia. Select samples of Trillium ovatum and T. "hibbersonii" were gel quantified as a secondary check.  3.6.2 Agarose Gel Electrophoresis Minigels of 1.4% agarose were prepared by melting 0.56 g of agarose in 40 ml lx TAE buffer (0.04 M Tris-acetate, 0.001 M EDTA) and adding 2 ul EtBr (10 ug/ml) prior to pouring into the gel forms. 1% TAE buffer was used during electrophoresis. lOx gel loading buffer (0.25% bromophenol blue, 0.25% xylene cyanol, 30% glycerol in H 0, Maniatus Loading Buffer III 2  (Sambrook, 1989) was added to each DNA reaction sample (2 pi /15 pi). Samples were run  29 at a current of 80 V for 40 minutes to 1 hour. Gels were viewed and photographed under ultraviolet light.  3.6.3 Random Amplified Polymorphic DNA Prior to DNA extraction, the individual plants were identified to species and source. Most samples were collected from the UBC Botanical Garden Collections except for Trillium "hibbersonii" which was sampled from the wild population at Hesquiat Harbour (Table 5). The DNA preparations were used in RAPD-PCR reactions as described by Williams et al (1990). The lOmer primers for RAPD analysis were obtained from the UBC Nucleic Acid Protein Service Unit. For each amplification reaction approximately 25 ng DNA, 1 unit Taq buffer (10 mM TrisHCl pH 9.0, 50 mM KCL, 1.5 mM MgCl 0.1 % Triton X100, 0.2 mg/ml 2  gelatin, Appligene), 0.2 uM primer, 0.1 uM of each dNTP and 2 ul Taq DNA polymerase (5000 units per ml, Appligene) were used. Reactions were performed in a Bio-Oven III Thermocycler by Biotherm. An initial denaturation step of 5 minutes at 94° C was followed by 44 cycles consisting of 1 min. at 94° C (2.0 C sec" ), 1 min. at 35° C (0.1° C sec ), 2 0  1  -1  minutes at 72° C (2.0 C sec" ). The final step is 1 cycle of 4 minutes at 72° C. Ramp times 0  1  are given in brackets. Different concentrations of MgCl , (from 1.9 mM to 3.0 mM) were tested in order to optimize 2  reaction conditions. A concentration of 2.1 mM MgCl was used in the final reactions. 2  30  4. Results and Literature Reviews 4.1.1 Cytology Literature Review The chromosome base number for all Trillium species is five. All of the sessile-flowered trilliums are diploid with a chromosome number of 2n= 10+0-7 beta chromosomes (Freeman, 1975). Most of the North American pedicellate-flowered Trilliums also have a chromosome number of 2n=10. Chromosome counts of both 2n=10 and 2n=15 are published for two eastern North American species, T. erectum (Sparrow et al, 1952; Samejima, 1987) and T. flexipes  (Gotoh, 1937; Samejima, 1987). The Asiatic Trillium species include a diploid,  triploids, amphidiploids, tetraploids and allohexaploids (Kurabuyashi, 1958). Trillium ovatum is a diploid (2n=10; Warmke, 1937) consisting of two metacentric, six submetacentric and two acrocentric chromosomes (Fukuda & Channell, 1975). A chromosome base number offivewas reported for T. "hibbersonii" (Taylor & MacBryde, 1977).  Cytological work on the genus has been extensive since the 1930's due to the large size of the chromosomes and ease of working with them. The cytology of the Pacific Coast trilliums was investigated by Warmke (1937). The basic chromosome complements of Trillium ovatum, T. chloropetalum, T. rivale and T. petiolatiim were found to be very similar consisting of a large V, intermediate V, small V, a J and a knobbed chromosome. Warmke noted small differences in ratios of arm length, size and satellites that could be used to help distinguish between the species. Karyotype analysis in the early studies used the differential reaction of constitutive heterochromatin to cold treatment and differential staining techniques developed by Darlington and LaCour (1940). Darlington and Shaw (1959) and Bailey (1954, 1958)  31 c o n d u c t e d k a r y o t y p e studies i n v o l v i n g 71 ovaii/m, with an emphasis o n descriptions o f patterns that c o u l d potentially be used as t a x o n o m i c indicators. Pattern differentiation i n 71 ovatum related to population size and breeding structure was analyzed by K u r a b a y a s h i (1963). C o n t i n u i n g from Kurabayashi's study, F u k u d a and Channell (1975) analyzed p o p u l a t i o n variation i n 71 ovatum based on c h r o m o s o m e pattern variation in conjunction w i t h m o r p h o l o g i c a l data (leaf shape). A correlation o f c h r o m o s o m a l pattern variation and leaf shape v a r i a t i o n w a s related to distribution along a cline, between east and west regions and w a s also related to evolutionary history.  4.1.2 Cytology Results T h e c h r o m o s o m e number o f the garden cultivated 71 "hibbersonii" used in this study is 2n=10. T h e root-tip squashes illustrated in Figures 8 & 9 are at late prophase to early metaphase. T h e c h r o m o s o m e number confirms that reported by T a y l o r & M a c B r y d e (1977) and is based o n four counts from t w o plants.  4.2.1 Pollen Literature Review T h e m o r p h o l o g y o f 71 ovatum pollen was elucidated by M . Takahashi (1982) u s i n g scanning and transmission electron m i c r o s c o p y . A study i n v o l v i n g seventeen N o r t h A m e r i c a n species found that the pollen is characteristically spherical and inaperturate, 2 0 - 3 0 urn in diameter w i t h variation o c c u r r i n g in the exine structure (Takahashi, 1982). Differences in exine ornamentation were used to form five groups where members were indicated to be closely related and independent o f the other groups.  Trillium ovatum is in the g r o u p characterized b y  spinulate exine ornamentation as are 71 grandiflorum  and 71 nivale. F o u r specimens o f  32  Chromosome Counts  33 Trillium ovatum "hibbersonii"  were studied and no infraspecific variation was noted. No samples of T.  or T. rivale were examined in the Takahashi study.  4.2.2 Pojlen Results  Both the pollen from Trillium ovatum and T. "hibbersonii" are similar in size and in gross morphological appearance. The percentage of viable pollen, indicated by stainability (Hauser and Morrison, 1964) also lie within similar ranges (Table 6). Table 6. Pollen stainability. Trillium "hibbersonii"  % Via  UBC Botanical Garden 2a  . 97  UBC Botanical Garden 2b  94  UBC Herbarium 73131, Type  '71  UBC Herbarium 25476 , Boat Basin  82  Pot A #3 1993, used for crossing  92  §ptf,.  mmllwmrOyatHHi  UBC Native Garden la  87  UBC Native Garden lb  62  UBC Native Garden lc  86  UBC Herbarium 5904  98  UBC Herbarium 78984  94  UBC Herbarium 82316  99  UBC Herbarium 113005  94  Native Gdn, Bog 1993, used for crossing  100  hk  34  4.3.1 Pollination and Seed Dispersal Literature Review The study of pollination within the genus is incomplete. Berg (1958) listed two types of insect pollination within the genus, flowers with pleasant fragrance attracting bee pollinators and malodorousflowersattracting fly pollinators. A study concerning five sessile-flowered Trillium  species of central Tennessee found that the most effective pollinators were beetles  (Coleoptera),  followed byflies(Dip/era), wasps (Hymenoptera) and true bugs (Hemiptera)  (Patrick in Adams, 1975). Ihara (1973) concurs that the main cross-pollinator in the genus Trillium is probably Coleoptera. (Thomisxdae Misuminops sp.)  Adams (1975) noted pollen grains on the leg of a spider  found on T. lute urn while Patrick (in Adams, 1975) concluded  that spiders were possible pollinators as representatives of two spider families were noted in many instances. A study of the biology of T. nivale (Nesom, 1985) noted frequent visitation by honeybees (Apis mellifera) at one site and in approximately 30% of flowers at two other sites a beetle (Nitidulidae Meligelhes simplipes) was observed. The breeding system of T. nivale  is described as facultatively xenogamous with a brief pistillate phase (protogynous).  Dyer's (1963) claim that Trillium species are self-sterile was disputed by Ihara (1973), who did, however, find that self-pollination can also be inhibited due to partial protandry or protogamy rather than a system of S-allele incompatibility. In a study on mode of speciation, Fukuda (1987) noted that all American species are diploid, with self-pollination the dominant mode of breeding, such that even in naturally co-existing populations hybrids are not common. Among the lapanese members of the genus there are many hybrids and out-breeding is dominant. The polyploid system has also developed in the Asiatic species.  35  In contrast to Fukuda's theories Ohara (1991) concluded from a breeding study of four Japanese Trillium species, that inbreeding is the predominant breeding system although the ability to outbreed is still present with varying levels of xenogamous characteristics between the species. In this same study the most common insect visitors to T. kamtschaticum and T. tschonoskii smallii,  belonged to Diptera and Coleoptera. Insect visitation to T. apetalon and T.  which lack or partly lack petals, was rare (Ohara, 1991). In a study on a mixed  population of the Japanese species T. apetalon and T. smallii no interspecific hybrids were found, despite being experimentally compatible (Uchino, 1989). The pollination study confirmed that self-pollination was predominant and possibly the isolating factor for both species even though flowering seasons are coincident. Seed dispersal experiments conducted by Berg (1958) demonstrated that effective dispersal of Trillium ovatum, T. rivale, T. chloropelalum, T. peliolalum and T. erectum  is accomplished  by ants. Myrmecochory has also been reported for T. recurvatum (Robertson, 1896), T. grandiflorum (Gunther and Lanza, 1989; Gates, 1940, 1941) and T. nivale (Smith el al.,  1989; Nesom and La Duke, 1985). In 1983, Mesler and Lu studied the effectiveness of ant dispersal and pattern of distribution of Trillium ovatum seeds in coastal redwood forests (Mesler & Lu, 1983). 4.3.2 Pollination Results Both Trillium ovatum and T. "hibbersonii" exhibit protogyny, with the stigma appearing receptive for two or three days prior to anther dehiscence. A flower beetle (Chtysomelidae) was collected and identified from T. ovatum in the UBC  36  Botanical Garden BC Native Garden Collection, North American Garden and the Lynn Valley population. A sap beetle (Nilu/idae) was also collected from 71 ovatum in the BC Native Garden. A beetle in the Chrysomelidae was also collected from 71 "hibbersonii" in the North American Garden. None of the beetles were observed travelling from plant to plant. Bees were not observed visiting either 71 "hibbersonii" or 71 ovatum although bees frequently  visited the 71 nivale located nearby. 4.4.1 Germination and Interspecific Crossing Literature Review  Seed from North American species of Trillium generally requires two years to germinate  (Lighty, 1993; Case, 1988; Dyer, 1963). Seed dormancy occurs in both pedicellate and sessile-flowered species (Samejima, 1962). Berg (1958) noted that dormancy or after-ripening requirements were common in the genus and that the Trillium embryo is incomplete (in the sense of Goebel, in Berg, 1958) being less than one tenth the length of the endosperm. Apomixis in Trilliium  species was reported by Jeffery and Haertl (1939a&b) with the  endosperm as the source of diploid embryos originating as adventitious buds. Christian (1980a) reported that fertile seeds were produced by crossing 71 ovatum x T. rivale and 71 rivale x 71 ovatum, but it was unknown whether the hybrids were fertile as none had  reached flowering stage at the time of the article. From seed germination to flowering stage averages six or seven years (Case, 1988).  37  Criteria for rating the viability of seeds were outlined by Channel & Haga (1982) where seeds from a compatible cross are very firm due to normal development of the endosperm and the normal brown pigmentation of the seed coat is present. Seeds from incompatible crosses may be normal in size with brown seed coat pigmentation but are not firm due to abnormal endosperm development. Slight finger pressure causes the seed to split, extruding a liquid, which probably represents the abnormally developed endosperm material. Barriers to cross-pollination thus appear to be of the post-fertilization type, which become manifest during seed endosperm development (Jeffrey and Haertl, 1939b). One of the factors that contributes to the inter-sterility of Trillium that Dyer (1963) observed is failure of the endosperm to develop properly. 4.4.2 Germination and Interspecific Crossing Results Table 7. Crossing data for Trillium "hibbersonii" and T. ovatum.  ? Parent T.  "hibbersonii"  <f Parent T. ovatum  Number 1  x  Results 1 fruit, 41 seeds, none viable  T.  "hibbersonii"  T. ovatum  1  unknown material stolen  T. ovatum  T. "hibbersonii"  12  11 fruit undeveloped 1 fruit, seeds inviable  Fruit did not develop in 11 out of 12 cases where the maternal parent was T. ovatum. In one instance, this may be due to the plastic protector not being removed soon enough. In the case  38  where a fruit developed the seeds were creamy to light yellow in colour, eliasomes were small but present, and the endosperm was abnormal, being composed of liquid. Trillium "hibbersonii" "hibbersonii"  seeds that resulted from a manual cross-pollination with T.  were immediately prepared for cold-moist stratification. The seeds were pre-  soaked in water for 12 hours, added to a pre-moistened 50:50 peat perlite mix and placed in a glass container with a ventilated lid. The duration of the cold period was six weeks at a temperature of 4°C. The cold treatment was followed by warm-moist stratification, also 6 weeks in duration at a temperature of 18°C. Twelve seeds collected from manually pollinated T. ovatum  (1431-156-77) were also prepared in the same manner. Trillium "hibbersonii"  exhibited 90% breaking of dormancy (radicle emergence and shoot emergence) after only one cold period (Figure 10). The free germination of fresh T. "hibbersonii" seed without exhibition of double dormancy was also noted by F.F. Hunt (1991) and A. Guppy (1992). In contrast the twelve seeds of T. ovatum did not show signs of radicle emergence even after two cold/warm stratification cycles. All seeds were transferred to a peat pot with the same media mix used during stratification and placed in a sheltered site outdoors in early spring 1992. Four seeds of T. ovatum produced shoots in spring of 1993. Longer stratification periods for T. grandiflorum increase  the germination rate (Lee, 1995) and may be a factor in the slow and  low percentage germination of T. ovatum in this trial. Trillium ovatum  typically requires two cold periods for germination (Christian, 1980b); after  the first period of cold and the slow warming of the soil, the radicle emerges. The cotyledon remains dormant until the end of the second cold period when it finally emerges. Hypocotyl  39  dormancy is indicated by the failure of the cotyledon to elongate during the first year of germination.  Figure 10. Sketch of development of Trillium "hibbersonii" seed, radicle and rhizome (eliasomes removed). Seed collected in July 1994 from UBC (140-90?) and placed under cold-moist stratification conditions. Shoot appears near the end of the six week cycle.  40 4.5 Morphology: Description 4.5.1 Trillium subgenus Trillium  Herbaceous perennial with a short, thick rhizome; erect stem with scarious basal sheath, 1 or more per rhizome. Leaves, simple, glabrous, 3 in a whorl, subtending solitary, terminal, pedicellate, bisexualflower.Sepals 3, green, persistent, alternate with petals; petals 3, withering rather than deciduous; stamens 6, straight, basifixed linear anthers, 2-celled, introrse dehiscence; stigmas 3, sessile, inner surface stigmatic, linear with curled tip, curling more as flower ages; ovary with 3 carpels, placentation axial; maturing into a non-pulpy, berry-like capsule; seeds numerous, reddish-brown, ovoid, red-brown to light brown with whitish to yellow, soft eliasome attached on the concave side adjacent to the funiculus. Eliasome as large as or larger than seed. Flowering time March-May. 4.5.2 Trillium ovatum Pursh ssp. ovatum  White or Western Trillium, Western Wakerobin Figures 11, 12 & 13. Scape erect, (5.0) 17.2-32.2 (59.9) cm, red colouration common. Leaves sessile to subsessile, widely rhombic to rhombic-ovate to deltoid-obovate, acute to abruptly acuminate, cuneate base. Blade 5.6-12.6 cm long and 3.6-10.2 cm wide, length to width ratio 1.1-1.7. Three to five prominent veins; parallel and net veins conspicuously sunken giving a rugose texture. Petals ovate to oblong, 2.5-4.3 cm long, 0.9-1.9 cm wide, white from bud till after pollen dehiscence then turning to shades of pink to deep wine purple, longer and broader than sepals, withering not deciduous. Sepals narrowly oblong to lanceolate, 1.7-3.3 cm long and 0.3 to 1.1 cm wide, persistent, red colouration minimal, most common on margins, late in  Figure 11. Trillium ovatum in flower, Native Garden, UBC Botanical Gardens. White bags cover flowers that were emasculated and crossed using T. "hibbersonii" pollen.  Figure 12. Agedflowerof Trillium ovatum, Native Garden, UBC Botanical Garden. Colour of aged flower ranges from deep pink to wine red. Note the deeply veined leaf texture and slender, erect pedicel.  42 development, beginning as petals are colouring and then withering. Sepals fused 1 to 2+ mm at the base. Peduncle slender, 2.0-5.4 cm, reddish colouration, erect or declining at fruit maturation. Corolla spreading from base, not funnel shaped, petals spreading 0 to 45 to 0  receptacle (Figure 13). Stamens 6, filaments creamy-white, shorter than anthers; anthers yellow, not extending beyond creamy-white connective. Stamens approximately 1/3 length of petals. Stigma position less than or equal to top of anther. Ovary vase-shaped to subglobose, 6-angled to distinctly winged, yellow-green to yellow-brown. Fruit a berry-like capsule with irregular to regular dehiscence along 3 longitudinal sutures. Seed length 2.5 - 4.0 mm, width 1.5 -2.5 mm. Aberrations are common with sepals and stamens forming petaloid structures or the leaves and flower parts occurring in two's, four's or five's (Goodspeed, 1917; Gates, 1917). Abberant colour forms (green on petals, white on sepals) or unusual flower colour forms; pink, yellowgreen or green (UBC Botanical Gdn #3049-117-71) also occur. Stamens were noted which did not produce pollen but were normally developed except for creamy-white nonproductive anther sacs. Named clones: 'Kenmore' - double flowers, pale pink at anthesis? 'Edith' - double flowers Tillicum' - double flowers  43  Petals 0to45«  Sepals Pedicel Leaves  A  > 4 5  Scape Spreading from base  Funnel-form  Trillium ovatum Pursh  Trillium grandiflorum (Michaux) Salisbury  Trillium "hibbersonii"  Trillium Hivale Riddell  Figure 13. Illustration of corolla shapes in Trillium ovatum, T. "hibbersonii", T.grandiflorum and T. nivale.  44  4.5.3 Trillium ovatum forma hibbersonii Taylor & Szczawinski  Hibberson's Trillium, Rock Trillium Trillium hibbersonii Wiley nomen nudum Trillium X hibbersonii (Wiley, 1968) Trillium ovatum X T rivale ( Mitchell, 1969)  Figures 14, 15, 16 & 17. Scape erect ( 0.6) 1.0 - 8.2 (17) cm, red colouration common. Leaves sessile, occasionally subsessile-petiolate, elliptic to narrowly elliptic, acute to short acuminate, cuneate base, margins reddish, undulate. Blade 3.3-5.3 cm long, 1.3-2.5 cm wide, length to width ratio 1.92.7. Three to five prominent veins, smooth leaf texture. Leaf often not held flat but halves held at angle to the longitudinal axis. Petals ovate to elliptic, 1.6-2.4 cm long and 0.7-1.1 cm wide, slightly undulate, pink (RHS red purple group 65 A in Bud, 65 C when flower completely open), become darker pink with age, especially along venation, withering after anthesis, but not detaching. Sepals lanceolate, shorter than the petals, 1.1-1.7 cm long and 0.4-0.6 cm wide, persistent, red colouration present on margins and venation at anthesis. Sepals not fused at the base. Peduncle relatively stout, 0.6-1.6 cm, turned toward horizontal below the flower (to approximately 45 °), red colouration especially marked along striations. Corolla not funnelshaped but ascending, spreading from base 0 to 45 (Figure 13). Base of perianth angular. 0  Stamens 6, straight, about 1/2 petal length, filaments shorter than anthers, anther connective purple on dorsal side, pollen sacs yellow and not extending beyond the connective. Stigma position at or above top of anther. Developed ovary sub-globose to nearly spherical (Taylor et al, 1975), 3 carpels obvious in the shape of the fruit, ridged but not prominently winged,  45  Figure 14. Trillium "hibbersonii" in the North American Garden, UBC Botanical Garden. Clumping suggests vegetative increase but seedlings are present and easily distinguished, (black marker label 8.5 cm.) Photo GBS.  Figure 15. Trillium "hibbersonii". The stigmatic surface is held above the introrsely dehiscent anthers. The sepals and leaf margins are rimmed with red. Note the smooth leaf texture. Photo D.W.  46  Figure 16. Trillium "hibbersonii" grown in cold frame, U B C Botanical Garden. A t anthesis the petal colour is pink. Leaf and sepal margins are red, sepals are not fused at the base. Photo D . W .  Figure 17. Anthers in Trillium "hibbersonii" in which the pollen sacs did not develop. U B C Botanical Gardens cold frame. Improper development was observed in two other populations. Photo D . W .  47  3  mrn  Figure 18. Sketch of abortive anthers of Trillium "hibbersonii". The filaments are creamy-white and the connective tissue is purple. Undeveloped 'pollen sac' tissue lacks pigmentation.  48  yellow-green to slight red-brown. Fruit a berry-like capsule with irregular dehiscence. Seed length (2.0) 2.6 (3.3) mm, width (1.0) 1.3 (2.0) mm. Eliasome, yellowish up to 3/4 size of seed. A photograph of the plants discovered by J.A. Hibberson and transplanted to his garden in Victoria, B.C. doesn't show the leaves folded about the midline (Clark, 1976). This may be a populational characteristic or a stress response, possibly to high light levels. Tetramerous forms were described by Wiley (1972) and Taylor et al. (1975). A tetramerous individual was observed in 1995 in the UBC Botanical Garden collection (cold frame Pot B: 140-90?). This is an unstable form, as it was not observed in 1994. The fourth petal was incompletely formed from a stamen. Other teratological forms were observed by the author in the UBC Botanical Garden collection (028973-01-90) and in the Noble collection. Several individuals appeared to be normally developed except for the anthers (Figure 17 & 18). The abortive stamens are light purple-white, thickened and incurved at the distal end, extending only slightly beyond the ovary. Flowers have either all fertile, developed stamens or all abortive stamens. The ovaries appeared normal and when pollinated set seed. The appearance of undeveloped anthers appears to be due to environmental factors rather than the result of inbreeding. Pot B of the cold frame collection (140-90?) was repotted in July 1994. Several individuals had undeveloped anthers at this point(1994), yet in the following season (1995) all these individuals had normally developed anthers. However, individual specimens in the Alpine Garden Collection (010698 -0156-74) had undeveloped anthers again in 1995. Another UBC variant specimen had leaf and flower parts occurring in two's, two stigmas, two carpels, and two leaves. One leaf was a developmental fusion of two as evidenced by the two tips.  49  4.6.1 Morphology: Comparison  The impact of environment on phenotypic expression, especially on vegetative characters, is an important consideration. Herbarium specimen data collected in this study encompasses the entire range of Trillium ovatum and represents the continuous range of variation present in nature due to both environmental and genetic effect. Fresh specimen sampling was concentrated in the north western limit of the range of Trillium ovatum where it is sympatric with T "hibbersonii". Sampling was limited to areas of relatively easy access except for the area in the ecological reserve containing the type locality of T. "hibbersonii". The data collected from the UBC Botanical Garden places both taxa under similar environmental conditions as in a transplantation study. Both taxa of the UBC accessions in the North American Garden have been in place for 19 years and many seedlings have reached the reproductive phase over the years. The ability to distinguish between the taxa within the UBC Botanical Garden collection emphasizes the genotypic source of the variation. Once the reproductive stage has been attained, T. "hibbersonii" can consistently be distinguished from T. ovatum in bud or at the flowering phase using qualitative and/or quantitative morphological characters. The quantitative and qualitative character comparisons are summarized in Table 8 and Table 9.respectively. Ten morphological measurements and two ratios were scored and the results are summarized in Table 10. Trillium "hibbersonii" "hibbersonii"  and T. ovatum are easily distinguished in bud and flower. Trillium  petals are pink in bud and anthesis (Figure 16). Trifoliate juveniles can generally  be distinguished based on overall size. Both taxa flower soon after the shoot emerges from the  50 Table 8.  Comparison of quantitative characters of Trillium ovatum and Trillium "hibbersonii".  Character  Trillium ovatum  Trillium "hibbersonii"  Huwci Colnui  Royal Hort. Soc. white group  R.H.S. Red puiple group  anllicM.s  155 D  65 C  hud  155D  65 D  aged  pink-red-vvine  dark pink/purple  Height (In lea\c.i) u n  (5.0) 16.1-31.9 (59.9)  (0.6) 1.0-8,2 (17.0)  Pedicel Length cm  (0.8) 2.1 - 5.3 (9.1)  (0.5) 0.6 - 1.6 (2.8)  Loaf Length cm  (2.0) 5.6- 12.6 (20.6)  (2.2) 3.3 - 5.3 (7.0)  Leaf Width cm  (1.3) 3.7 - 10.3 (20.9)  (0.8) 1.3 -2.5 (3.3)  Leal Ra.se ID Widest cm  (0.7) 2.1 -5.3 (10.2)  (0.6) 0.9- 1.9 (3.3)  Leal" Widest to Tip cm  (1.5) 3.5-7.5 (13.3)  (1.3) 2.1 - 3.5 (4.7)  Leaf Shape (length/vvidili i  (0.9) 1.2-1.6 (2.6)  (1.4) 1.9-2.7 (3.6)  Leaf Length Ratio (bT/bo  (1.3) 2.2-3.0 (5.1)  (1.9) 2.6-3.6 (3.1)  Petal Length cm  (1.3) 2.5-4.3 (7.0)  (1.0) 1.6-2.4 (2.9)  Petal Width cm  (0.3) 0.9- 1.9 (3.7)  (0.4) 0.7-1.1 (1.4)  Sepal Length cm  (0.6) 1.7-3.3 (5.2)  (1.0) 1.1 - 1.7 (1.9)  Sepal Width cm  (0.2) 0.4- 1.0 (2.5)  (0.3) 0.4 - 0.6 (0.6)  Seed Length mm  2.5-4.0  Seed Width mm  1.5-2.5  -  •  (2.0)  2.6  (3.3)  (1.0)  1.3  (2.0)n=43  Reported as one standard deviation around the mean, (minimum) & (maximum) values.  Length = b - T Width = Li - d Widest to Tip = c - T Base to Widest = b - c Leaf Shape = Length / Width Length Ratio = b - T / b - c  CO' fD  a  P  fD 3 »—•  oo O o" o ,  •I  §  o c 3 3  ro »— r<'  a. o'  r  COJ 3 I 'ic c 1—tfD — GO  OO  2.  <  C  •a  EL  ro  o  P  T3  CD  p o •a o  CO  fa  < fD 3 P  r  fD P •-») H  3  O  3  -fD  fD P  P  s/i  fD  fD P  <-*>  00 3P T3 fD  H » cr  w •  5T  H  \o  n o  O  s  3  C 05  "73  »  3  o  fD  O  3  s  3  s  O  P  fD fD  o •I  fD P  v>  3-'  I era - •  CJQ "I  •n fD 3 fD  I-  S3.  O  3 05  fD  3  O 05 2! o era fD 3 era _, a.  fD 3  a.  00  -1 fD fD 3  fD era fD fD  <  fD  CL  03  O  o_ o  VI  303  c  "a fD CL  3  p CL fD O  fD  3  P_  fD A o T3 3 fD —i EL P " ci3  3  3 fD l  cr P  fD  3-  05  fD  P 3  fD  3"  5'  3  CL  O  K)  C/5 < fD  3  3  a.  P  fD  fD  c o c  o  c 3  ofD  CL  c  vi  C  o cr o VI JTD Cu  a  3  P  cr  jr  fD  2  3  r-t-  P  5' o  <  3 © i •a  3"  o  era.  "O C  c  P~  CT  a.  o o  VI  IA  c ro  3 O  P CFQ  c  oo I-a 00 3  1»  o  o  c  3  fD  3 P  3  fD  3  era fD  O  3 cr o"  CT ©_  3*  O  3  CT O  < JTD  a.  5 S  2L 9. CL.  6 cr o  2  <  a.  rt re  P  fD O  P  < 3 ft 3 fD fD ~ 52. O  o -a  fD O 3  a. o VI  P  o  2. 2. EL  era  fD  a. oP  •"I  T3  CL o  It  fD 3 O g oT S- era fD =• A g to  I era  fD  S fD  3  f  P ^> P 3 3"  fD  fD D-  3  P  5'  fD  CL  o  cT c  !-|  0 3. 3 3  C  cr oo o o  1 o 3  P  fD  3  00  3  p  V  p  3 3"  fD  O  CL  fD  00 ,  0  ?  5  r< cr n> 3  CL  VI  fD  CL  u  O fD 3  o  3  O O  3 3  fD 3  fD  a.  c  fD O  O P O  c  O  VI r+  fD  3"  O*  cr  CL  VI  -a  2-  3"  cr 2o  fD  3.  o" p  33 O fD  CL  era  CL  cr  c 3  O  TS  s  3 P  >-|  o  Q.  2 2  1^  <-+>  Cu g  fD  3  fD P  3*  O  3, 2P  o C 3 3  fD  o  ft"  a a  M T3  n  oo  -o 'TO o -a o •a  2 ft  s»  "O  oo H  a &  Tp  OJ —  J-  o o  —o o  O—  -J KJ O KJ  p > — O OJ O ON  P P ON OJ  OJ OJ  p - o KJ O Ui  P P L/i KJ  OJ O OJ Ov  o —o KJ O J>  o o J> NJ  KJ  O —O - O to  O O KJ '—  O O »—• <0 OJ  p P  O — CD b b -  O O — o  p  o o  p  H  c  cr  3  <  o c a  a  p — p ki b u i  ifc. io ^  Tp  00 H  a >—' —  2  O O  3 ft <  NJ O bo Ui  O  00  T3 w  OJ !—•  P P ON OJ  o o J>  OJ —  p —p KJ b L/i  p p L/i KJ  b vo  i—' ON  — KJ b\  bv  S p Lo bo 5  ;„  KJ  ~J — tVO ^  — — OJ  OJ  —  -j — ca ^ * ^  KJ — — b\ bv  — KJ b\ bs  O —O VO vo  O p VO vo  —J O  KJ O -S3 ^ vo ^  o — o Ol  P P OM OJ  KJ — w ^  o — o bv bo  o o bo bv  O —O J— OJ  p p  KJ O *-  00 r 00 ^  KJ  OJ  o —o - b *.  b\  OJ —  -~J  ^  KJ —  —— ' ' OJ K> i„  —  —O J> J^  o  t-  vO O  1  00  3  ~J KJ 50 OJ  ON 4^  -J -J  OJ  —  co  H  OJ  i.j o  OJ ^—  £  vl  U  OJ  KJ KJ P b  ~ J vo  KJ —  v)  O  LJ  Tj  1  v]  C«  [T  & b ^ -j  OJ  S  ^  CO  — OJ  U1  OJ T3  OJ  O  on — hfl  ~J  OJ  3  VO  5 * ^ 00  KJ  b\  KJ O l/i KJ  •P* KJ OO  g  t-  J> O  00  8 ^ ^ - J O-i r  Ji. — KJ — b  KJ  o — o — KJ  o o• KJ —  KJ O ON VO  —J Ov - jo OO  o o KJ —  O O  O-i — OJ  co OS KJ ON H  J>.  b —  KJ KJ  oo  U 1  H  00  o  o  29  5«  53  Figure 19. Aged flower of Trillium "hibbersonii" is dark pink. The sepals are clearly not joined at the base. Striation in the pedicel is prominent. Flower attitude due to the bend in the pedicel. Photo D.W.  Figure 20. Dark pink petals of aged flower of Trillium "hibbersonii". The anther connective was purple prior to anthesis. The anthers extend beyond the ovary but below the stigmatic surface. Photo D.W.  54 ground and before full height and leaf expansion is achieved. Specimens collected at this phase could be improperly classified as intermediates if height is the only criterion used. Examination of the stigma/anther relationship can be used to separate the taxa even prior to the flower opening. The stigma of Trillium "hibbersonii" extends beyond the anther tips in mature flowers (Figure 20) and is visible and approximately equal to the anthers in the bud phase. In T. ovatum  the anthers extend beyond stigmatic surface at the matureflowerstage (Figure 21)  while at the bud phase the stigma is positioned well below the anthers and is hidden by the anthers. An additional character that is useful to separate problematic specimens is that of leaf shape in conjunction with texture. Leaves and sepals of T. ovatum continue to expand during the season, increasing significantly during anthesis (Berg, 1958; Renner, 1980) (Figures 22 & 23). Expansion also occurs in T. "hibbersonii". This continued growth increases the assimilation potential of the plant as needed for resource allocation to seed development and replenishment of the rhizome. Starting at anthesis, measurements of height, pedicel length, leaf length and width, sepal length and width were taken for intervals ranging from one to seven weeks. All plant parts measured increased in size. The greatest growth rate during the test period was observed at anthesis in both taxa. The length/width ratio, an indicator of leaf shape, remained consistent over time for both taxa. The length/width ratio for 71 "hibbersonii"  of 2.3 reflects a narrower leaf than the 1.4 ratio of 71 ovatum. Distance from  the widest point of the leaf to the tip averaged 2.8 cm for T. "hibbersonii" and 5.5 cm for T. ovatum.  Fukuda and Channel (1975) published length/width ratios for 71 ovatum ranging  from 1.31 to 1.43 in five American states (WA, OR, CA, MT, ID) with values for distance  55  Figure 21. Dehiscent anthers of Trillium ovatum. Dehiscence is introrse and relative height of stigma surface is less than or equal to the top o f the anthers. Anther connective is creamy-white and at this developmental stage the petal colour is white.  Figure 22. Trillium ovatum in April, North American Garden, UBC Botanical Garden. Blooms open shortly after plant emerges from ground and before leaves have fully expanded. Petals are ovate, no red colouration on sepals or leaves. (Ruler length 15 cm).  Figure 23. July photograph of same group of T. ovatum, as in Fig. 23, from a different angle. Plant height and pedicel length has increased along with obvious leaf expansion. Petals, anthers and stigma have withered. Note prominant wings on capsules. (Marker label 11 cm wide).  57  from the widest point to the leaf tip ranging from 3.8 to 4.8 cm. Length/width ratios were calculated from Renner's data on 71 ovatum (1980) with results ranging from 1.21 to 1.46. Renner had divided the range of 71 ovatum into 5 regions: Central California, CaliforniaOregon, Corvallis-Columbia River, Washington and the Rockies. The results from both the Fukuda and Renner studies concur with the average length/width ratio of 1.4 collected in this study for Trillium ovatum. Floral parts of Trillium are persistent due to the absence of abscission tissue (Berg, 1958). Petals and stamens wither while sepals remain green and functional beyond flowering. Petal measurements from herbarium specimens can be problematic as signs of withering are masked, to some extent, by drying. Sepals wither at a later stage which can result in the same problem. At the postfloweringstage the capsule of 71 ovatum is easily distinguished by the prominent wings on the maturing fruit (Figure 24). The maturing fruit of T. "hibbersonii"is sub-globose to nearly spherical (Figure 25) but an undeveloped capsule is indented (Figure 20) and although it lacks wings or ribs it may be misleading. 4.6.2 Key to Western Trillium subgenus Trillium 1 Flowers sessile..... 1 Flowers pedicellate  subgenus Phyllantherum subgenus Trillium 2  2 Stigma club-shaped and leaves petiolate 2 Stigma slender, more or less recurved, lvs sessile to petiolate 3 Anthers below exserted stigma, mature ovary not winged 3 Anthers extend beyond stigma , ovary 6 angled, winged 4 Leaves sessile to subsessile, flower erect 4 Leaves petiolate, flower nodding  Trillium  rivale  3 71 "hibbersonii"  4  71 ovatum ssp. ovatum 71 ovatum ssp. oettingeri  58  I  1  »  7 mm  1  H mm  Figure 24. Trillium "hibbersonii" mature capsule and cross-section of immature capsule. Capsule is sub-globose to almost spherical. The stigma extends beyond the anthers.  I  <  1  I  .  \  Figure 25. Trillium ovatum. mature capsule and cross-section of immature capsule. The capsule is prominently winged or ribbed. The anthers extend beyond the stigma at all stages of development.  59 4.6.3 Morphological Analysis  Frequency distribution plots showed that many of the measured variables are not normally distributed. This affects the usefulness of the statistics summarized in Table 10. Better descriptors of the data than the mean and standard deviation are found in distribution-free statistics. Median and range, distribution-free statistics, summarize the data without transformation. All ten measurements and two ratios on Trillium "hibbersonii" (Taxon Ff) and T. ovatum  (Taxon O) were analyzed and notched box plots produced (Figure 26a & 26b A-L). A box plot compares the median and range of the taxa and identifies the outliers. The 95% confidence intervals are visualized by a notch around the median (Figure 26c). If the notched area of the taxa medians do not over lap there is 95% confidence that the population medians are different (McGill etal., 1978). Notched box plots (Figure 26a, 26b) were constructed from the data. These measurements cover height, pedicel length, petal length and width, sepal length and width, leaf length and width, leaf length from the base to the widest point plus the length from the widest point to the tip. In all of the cases there is a 95 percent certainty that the medians of the two populations are different (Figure 26c). The ratios calculated to indicate leaf shape and to factor out differences between herbarium and fresh samples (Figure 26b K & L ) also have different medians at the 95 percent confidence level. Overlap in the ranges of the two taxa does occur in all measurements such that no character can be reliably used alone to distinguish the taxa. Multivariate statistic analysis uses linear equations to summarize differences using a number of variables. In discriminant analysis a grouping variable such as taxa can be used, a priori..  60 Figure 26a. Boxplot comparisons of height, pedicel length, petal and sepal measurements from Trillium "hibbersonii" (H) and T. ovatum ( 0 ) .  D.  0  TAXON  H  61 Figure 26b. Boxplot comparisons of leaf measurements and leaf ratios from Trillium "hibbersonii" (H) and T. ovatum (O).  62  Figure 26c. Notched box plots are used for comparison of the medians of the ten measurements and two ratios for Trillium "hibbersonii" (Taxon H ) and T. ovatum (Taxon O) in Figure 26a & 26b. When the intervals around two medians do not overlap the two population medians are different at the 95% level of confidence. Hspread is the absolute value of the difference between the medians of the upper and lower halves. Upper and lower range limits are 1.5 Hspreads of the medians of the upper and lower halves. Outliers indicated by * are outside the 1.5 Hspreads of the range and o indicate outliers outside 3 Hspreads.  *  —  >1.5 Hspreads of the hinges  Upper limit of range  95% Confidence Interval values  Lower limit of range  -L  H  o TAXON  63 Discriminant functions and classification probabilities are calculated on the observations divided into groups to highlight the differences between those groups. The classification probabilities are computed from the Mahalanobis distances where probability of group membership for each case depends on proximity to a group's location in the discriminant space. A two-dimensional plot graphically illustrates grouping and separation between groups. The assumption of multivariate normality is not met in this study and the results of the discriminant analysis only model the original data set (Johnson & Wichern, 1982). In SYST AT (Wilkinson, 1990) any case with missing values is excluded from the analysis. For Trillium "hibbersonii" this drastically reduced the number of cases. The significance tests are considered suspect when more than one-fifth of the fitted cells have a frequency of less than five. The scatterplot of distances calculated from the discriminant analysis of T. "hibbersonii" and T. ovatum plant measurements (Figure 26d) shows that the cases group by taxa with only a  slight amount of overlap. Leaf length/width ratios and petal width are important in the canonical loadings (Table 11) of Distance 1 and Distance 2. Of the 353 cases analyzed only sixteen incorrect group predictions were made. All of the Trillium "hibbersonii" were correctly classified. The sixteen cases of T. ovatum that had incorrect group predictions were probably not misidentified or mis-grouped. Quantitative and qualitative characters outlined previously and summarized in Tables 8 and 9 but not used in the discriminant analysis would likely verify the original grouping.  64 Table 11. Canonical loadings of the variables used in the discriminant analysis of Trillium "hibbersonii" and T. ovatum.  Canonical Loadings Distance 1  Variable  „ Canonical Loadings Distance 2  1.3  0.9  29.5  42.8  leaf centre to tip  1.5  1.6  petal length  1.3  -1.4  petal width  5.5  11.2  sepal length  0.7  -1.3  sepal width  -1.3  2.6  pedicel leaf length/width ratio  1  CO  4  1  0  -  -  -  -  -  . o"  OO  "0  oo '  0  0 o  °o  1  1  5  10  15  DISTANCE(I) o Trillium "hibbersonii"  ' Trillium ovatum  Figure 26d. Scatterplot of distance scores obtained from discriminant analysis of morphological measurements from Trillium "hibbersonii" and T. ovatum.  65 4.7.1 Flavonoid Literature Review The pattern o f flavonoid occurrences, termed the "biochemical profile" by Turner and Alston (1959), provides another set o f characteristics which may be used in a taxonomic treatment. Flavonoid compounds are based on a C - C - C nucleus (Figure 27) and are divided into five 6  3  6  main classes. Within the groups, the characteristics of individual compounds can vary greatly due to the type and location o f side groups. The compounds are physiologically stable and the extraction and separation processes are simple. Identification o f the basic compounds is also easy. There is wide distribution of flavonoids in plants with some flavonoid classes widespread and some rare. The use o f flavonoid compounds for chemotaxonomic purposes is extensive (Harborneefa/., 1985; Bohm, 1987). Previous work on the flavonoid chemistry of Trillium has been limited. A n unpublished doctoral thesis completed by J.T. Murrell, Ir. Trillium (subgenus Phyllantherum)  (1969) found that among sessile-flowered  the phenolic compounds (including flavonoids) were  sufficiently variable to be used as an assessment o f systematic treatments based on morphology and distribution. Murrell used biochemical profiles to separate and group the species and did not identify the compounds. Phylogenetic relationships were suggested by the degree o f similarity o f profiles as an indication of the degree o f relatedness. Quercetin and kaempferol glycosides and aglycones appear to be the dominant flavonoids in the Murrell study. Murrell found that in his " M Y " compounds (flavonol glycosides) minor quantitative variation between populations occurred but in certain of his species groupings major quantitative differences were highly characteristic.  66  Biosynthetic Pathway  phenylpropauoids  chalcones  flavanones  flavones R=H  dihydroflavonols I anthocyanins  flavonols  R=OH  Basic Flavonoid  Figure 27. Flavonoid biosynthetic pathway and flavonoid with carbon numbering system.  67 Three members of the subgenus Phyllantherum, Trillium sessile, T. cuneatum and T. stamineum were reported to contain two cyanidin glycosides (Asbury, 1973). Uniform Re-  values for the cyanidin compounds reported for species of this group led Asbury to conclude that it was likely that all cyanic trilliums contain the same two compounds. Corroboration of the presence of two cyanidin glycosides in Trillium cuneatum was furnished by Adams in 1975, who also found these compounds in T. luteum (subgenus  Phyllantherum).  Adams (1975) reported quercetin and kaempferol glycosides that contained arabinose as a prevalent sugar. A wide variety of mono- to tetraglycoside forms of flavonols were observed in other species of Trillium (Adams, 1975). A flavonol glycoside was isolated from aerial organs of Trillium tschonoskii (subgenus Trillium) by Nakano et al. (1983) and identified as 3-0-(2"'-O-acetyl-a-L-arabinopyranosyl-  (1 -6)- p -D-galactopyranosyl) kaempferol. Five quercetin flavonols and two cyanidin compounds were detected in a heterocyanic population of Trillium sessile (subgenus Phyllantherum), with variation in occurrence and quantity among the three phenotypes. A compound tentatively identified as quercetin 3,7-0tetraglycoside was found to occur in the pink and yellow phenotypes (Les et al, 1989); Arabinose glycosylation was again found to occur widely. Four flavonol glycosides were identified from leaf samples of Trillium tschonoskii, acetylated kaempferol 3-0-arabinosylgalactoside, kaempferol 3-0-arabinosylgalactoside, acetylated quercetin 3-0-arabinosylagalactoside and quercetin 3-0-arabinosylgalactoside (Yoshitama et al, 1992). Intraspecific variation of flavonoids indicated that two or three chemotypes could be present in T. tschonoskii and their occurrence correlated to geographical distribution and  68 differences in karyotype analysis (Yoshitama et al., 1992). Further work by Yoshitama et al. (1994) identified four flavonol glycosides in the leaves of T. tschnosokii; kaempferol/quercetin 3-0-2"'-0-acetyl-a-L-arabinopyranosyl-(l-6)-P-D-galactopyranoside (TAK, TAQ) and kaempferol/quercetin 3-0-a-L-arabinopyranosyl-(l-6)-P-D-galactopyranoside (TK, TQ). In the leaves of T kamtschaticum the major flavonoids are TAQ and TQ while in the leaves of T. apetalon TAQ, TAK occur with the same glycosides of isorhamnetin (Yoshitama et ai,  1994). A comparison of flavonoids in different plant tissues of T. kamtschaticum found quercetin dominant in the leaves, kaempferol dominant in the petals and both quercetin and kaempferol in the sepals (Yoshitama et al., 1994). Cyanidin 3-t9-rhamnosylglucoside was identified as the dominant anthocyanin in the sepals, stamens and fruit of Trillium apetalon (Yoshitama et al., 1994). 4.7.2 Flavonoid Results  No qualitativeflavonoidpolymorphism was found within the northern range of T. ovatum sampled in this study, either between individuals, populations or sampling over successive years. Sample collection ranged from April until August encompassing earlyfloweringto fruit maturation stages, without qualitative changes in flavonoid profile. Both the cultivated and the wild collected T. "hibbersonii" also showed no intra-taxon variation in the flavonoid pattern. Profiles of juvenile plants, either at the one leaf or three-leaved, non-flowering stage, of T. ovatum and T. "hibbersonii" were qualitatively identical to the respective mature flowering  plant profiles. A series of two dimensional chromatograms were run to provide an overview of the vacuolar flavonoid patterns of Trillium species. Seven Trillium taxa were used, six were collected from  69 similar site aspect and growing conditions within a ten square meter area (Table 2) in the U B C Botanical Gardens. The 2-D chromatograms (Figures 28-34) illustrate that flavonoid analysis provides characters that can be used to separate the taxa into species groups. Trillium  nivale  is considered to be closely related to T. grandiflorum and this is reflected in the similar profiles. (T nivale and T. grandiflorum are placed in the same group with T. ovatum based on pollen ornamentation (Takahashi, 1982).) Trillium ovatum and T. "hibbersonii" would be separated into another group. Trillium rivale, T. sessile and T. stylosum are dissimilar from each other or the grouped (T. ovatum and T. grandiflorum) taxa although T. rivale is closer to the T. ovatum group except for the lack o f the quercetin monoglycoside. Certain patterns are apparent: all major flavonoid compounds are flavonols, specifically quercetin and kaempferol glycosides, based on interpretation of spot colour (Markham, 1982) and comparison with standards; diglycosides are the predominant compounds and the triglycoside (Figures 28-34 Spot 6) is common to all taxa. The 2-dimensional chromatogram profiles of T. ovatum and T. "hibbersonii" are qualitatively identical with seven constituents (Figures 28 & 29).The most striking characteristic is the quantitative variation between the taxa. Quantitative variation is indicated by the greater intensity and spot size o f T. "hibbersonii" compared to T. ovatum especially in the diglycoside compounds. In the preliminary 2D survey, the initial T. ovatum profiles lacked the monoglycoside and displayed only one quercetin diglycoside spot. Heavier loading o f the chromatograms did confirm the presence o f the monoglycoside, and allowed detection o f other diglycosides. A quantitative difference was indicated, as larger quantities o f material were extracted and then concentrated prior to spotting.  70 Trillium "hibbersonii" Vacuolar Flavonoid Chromatogram  -  m  •  2  R  A •  •  >> Org 1-7 reference numbersforflavonoidsidentified Table 12 Dint before spray iBUl Dark before rpnry  Oreo, after spray Yclkrw/oraruje after spray |fff|l) Dark before spray ' < ' Dashed circle mhcatee VtlP' Yelow after spray ' Umer wncenttalion R - Rutin (Quercetin 3-0>^minnglifrisifki) • -Origin (  Figure 28. 2-dimensional profile of Trillium "hibbersonii" vacuolar flavonoids.  Trillium ovatum Vacuolar Flavonoid Chromatogram  l i i p  R  1  t  •  • Org 1-7 reference numbers for flavonoids identified Table 12 Daft before spray D»k before spray  (\) Vjy  Green after apray Yelcrw/crangc after spray Dai before spray ,' *i Dadwl ciick indicates Yelow after spray lower corOTtration R - Rutin (Quercetin 3-O-riamnoglucoside) • -Origin  Figure 29. 2-dimensional profile of Trillium ovatum vacuolar flavonoids.  71 Trillium rivale Vacuolar Flavonoid Chromatogram  -  € P  i 1 Nil!/ R  t .  Dare before spray  • Jill \if§§i/  Green after spray Dark before spray Yellow after spray  jHll ^gpF ,' '  Dark before spray Ydlow/ondge ate spray Dashed circle indicates 1"™" concentration  R - Rutin (Quercetia 3-O-mamnoglvicoside) • -Origin ^^^^^^^^^^^^^^^^^^^^^^^  Figure 30. 2-dimensional profile of Trillium rivale vacuolar flavonoids.  72 Trillium grandiflorum Vacuolar Flavonoid Chromatogram  • R  •  •  Dare before spray  Irlllllfjj  Dark before spray  Green after spray j Dare before spray - Yellow after spray  'ttipjl'  Ydlow/orasuje after apray Dashed circle fndioates lo^wr concentration  R - Rutin (Quercetin 3-Q-rbftmnnglirnsiAff) • - Origin  Figure 31. 2-dimensional profile of Trillium grandiflorum vacuolar flavonoids.  Trillium nivale Vacuolar Flavonoid Chromatogram  i l l  ^  m  R  t• 1  .  ©  Dark before spray Qreai after spray  Dark before spray Ydlow/oraoce after spray  Dark bdbre spray Yelow after spray  Dashed circle indicates lower concentration  R - Rutin (Quercetin 3 -O-rfiatnno^ucoude) • -Origin  Figure 32. 2-dimensional profile of Trillium nivale vacuolar flavonoids.  73 Trillium stylosum Vacuolar Flavonoid Chromatogram  • (  A  ^illi/  Dark before spray  jMfflh  Green after spray Dark before spray Yellow after spray  V$ggj' Yellow/orange sfter spray \ '  Dark before spray Daa^chdernificases lower concentration  R - Rutin (Quercetin 3-0-rnarnnogluco6ide) • - Origin  Figure 33. 2-dimensional profile of Trillium stylosum vacuolar flavonoids.  Trillium sessile Vacuolar Flavonoid Chromatogram  • R  •ntS&QhB,  mm  A,  A •  •  Dark before spray  rtlfili  Dark before spray  Green after spray Dark before spray Yellow after spray  VGgW  Yellow/orange after spray Dashed circle indicates lower concentration  , i v_ J  R • Rutin (Quercetin 3-O^rramnoglucoside) • -Origin  Figure 34. 2-dimensional profile of Trillium sessile vacuolar flavonoids.  74  The amount of sample obtained from column 1 (Trillium ovatum, dry weight 10 gm) was insufficient for purification. Subsequent partitioning of the phenolic fractions of Column 2 (71 ovatum, dry weight > 30 gm) and Column 3 (71. "hibbersonii", dry weight 4.5 gm) showed that both profiles were based on sevenflavonolderivatives (Figures 35, 36). The triglycoside was present in T. "hibbersonii" (Fractions 1 & 2) and in T. ovatum (Fractions 7  & 8)  in trace  amounts. A compound detected and purified from T. ovatum Fraction 14/15 was not detected in T.  "hibbersonii".  The flavonoid profiles of T. "hibbersonii" and T. ovatum are qualitatively identical with respect to the major constituents. The seven major compounds, purified from the 71 "hibbersonii"  fractions, were at least partially identified (Table 12) using the spot colour,  comparison with standards, ultra-violet absorption spectral data (Table 13) and shift reagent spectral data (Table 14). The major constituents are flavonols; quercetin and kaempferol -3-0monoglycosides and -3-O-diglycosides. Three different sugars were isolated and identified; glucose, galactose and rhamnose. The order of sugar linkage on the diglycosides was not determined due to the limited amount of sample available for hydrolysis. In T. ovatum Fraction 14/15 the compound produced the uv profile of a flavanone or dihydroflavonol. Further identification was not possible due to the small amount present. This compound was not detected in the  71 "hibbersonii"  fractions even in trace amounts. Flavanone and  dihydroflavonol are precursors toflavonolsand anthocyanins (Figure 27). The buildup of these precursors to detectable levels in step in the reaction compared to  71 ovatum  71 "hibbersonii"  may indicate lower enzyme activity at this where the reaction is weighted more  towards the end products resulting in greater concentration of flavonols and/or anthocyanins.  75  O  •  •  ^  • in «  H t ai  •  fl0  *m  0 «  la  fa  a M  2  * Si *  «N H  ri  •  i  |  i  OO • ©c© • ii  ©  «  E  0 1- 0\ &  u  E  so  •  © ©  3 0  15 M  ^^'n'li'i''iii'l  t  ?!  II 0i*  05 2"  V  CO) •  O  •  • 8 5 -a 1 lilili'ili'  •  ••<©  •oQ  snooi  III! 00  Q>-  76  . ,- ,  ^1  •  r-^  • III 00 vo •  1 «•-  Oo  5  •  ip  a «  •  .2 *  IHIK  *  A  •  »  A  E M o  •  m  «  S  0 h  »  »  1  11  U  s S s  IIP  0  »  U 0  |S 1  »  •  !! II ! i! |  11  I « »  »  OIUBSJQ  •o  77 Table 12. Flavonoids identified in Trillium "hibbersonii"  Refer to Figures 28 & 35. Mono-glycosides  Fraction  Quercetin-3 -O-galactoside Kaempherol-3 -O-glucoside  9-12 YB 9-12 GB  Di-glycosides  2D reference *  1  (order of sugar attachment not determined)  Quercetin-3 -O-rhamnoglucoside Quercetin-3-O-rhamnose/galactose Quercetin-3 -O-diglucoside Kaempherol-3-O-diglucoside Kaempherol-3-0-  5/6 L M Y 5/6 B Y 5/6 U M Y 5/6 TG 4 TG *  2 2? 3? 5? 4?  1/2 *  6  Tri-glucoside  Quercetiri-3-O-rhamnose-diglucoside  insufficient material for purification and/or hydrolysis. Tri-glucoside flavonoid in Trillium sessile.  Refer to Figure 34. (order of sugar attachment not determined) Quercetin-3 -O-rhamnose-diglucoside  6  Colour and position of spot in each profile fraction is denoted below: YB - yellow spot on bottom GB - green spot on bottom L M Y - lower middle yellow spot B Y - bottom yellow spot U M Y - upper middle yellow spot TG - top green spot *2D reference refers to spot # in Figure 28.  78  Table 13. Ultraviolet-visible absorption spectral data.  Trillium  "hibbersonii"  Methanol A1C1 II I II I 3  FRACTION 4 BY 4 TG 5/6 B Y 5/6 L M Y 5/6 U M Y 5/6 TG 9/10 B Y 9/10 TG Trillium  252, 356 265, 353 258, 360 262, 360 266, 356 262, 351 256, 354 262, 351  261, 362 267, 353 273, 433 267, 368 268, 406  266, 396 271, 394 270, 402 270, 401 271, 398  NaAcetate BoricAcid NaOCH II I t=0 min. t=5 min. II I II I II I 270, 384 263, 376 275, 413 275,412 272, 362 268, 354 275, 404 274, 403 270, 397 265, 391 281, 426 280, 426 270, 389 268, 388 274, 411 274, 408 271, 374 266, 360 274, 410 274, 409 2  ovatum  Methanol AICI3 II I II I FRACTION 10 B Y 10 M Y 10 TG 11 Y B 11 TG 14/15 17/18 B Y 17/18 TG Trillium  HC1 II I  258, 360 268, 392 262, 336 262, 336 265, 346 267, 348 258, 360 267, 393 270, 322 285,sh322 . 254, 334 266, 326  NaAcetate BoricAcid NaOCH II I t=0 min. t=5 min. II I II I II I 268, 384 268, 382 263, 379 276,417 278,417 272, 351 277, 348 273, 364 266, 350 274, 402 273, 402 268, 401 265, 377 264, 283 278, 424 278, 426  HC1 II I  2  sessile  NaAcetate BoricAcid NaOCH II I t=0 min. t=5 min. II I II I II I FRACTION 258, 358 269, 428 268, 399 263, 375 263, 376 273, 425 273, 425 Top YO Methanol AICI3 II I II I  HC1 II I  2  o o  T3  T3  O  o c  o <u -a o a  O  o  o  O  aj  e  S  O OJ T3  X O  22 •* o  •a  o  CN .3  o  C  C  r~ in +  o in +  vo  m +  m +  X o  X 0  b oo e  o  X 0 1  o  •3  c5 oo ° os -cc  cn +  x o  o  1  oo oo +  + CQ  * X  x  x  e  d  x  O oo a •c  U  t  a  •  3.CQ ¥  3  < CQ +  OS  o  f-i 00 00  |9  X  +  u •if  5 X  X  <S 9 - o .  2 o  00  6  e  oo e  CQ  + r - + c^ + r~- + r^ + r^  a x  X 0  cN -J.  + CQ  § X § X S X § X 5 X  -o  X o  c  —• 'G oo _ i C N i in + CQ + CQ +  c ° g 00 g O '5 ^ CN J , —' + CQ +  in + X  a  c  VO  m +  1  '•3  ON  i-  CN  +  + in + in + in + m + m  in +  3  CN +  m  in cn  m vo CN  <u  O  o  vO co CN VO CN  vO  in cn  in cn  vo VO  VO m CN  CN  VO cn  CO CN VO CN  oin o"  CO  CN  5n= >C *Q  O  CQ  in  D  in g cn CN S VO CN  < o o > CQ H H o o *S in ON Os  oR  J»  ^*  CQ o  •3• O oo  CQ  + m  + >n  CN  o  VO cn 00 in CN  o  59  C/5  vO in co  x  a X  * X  i  ii ?  u  cn cn +  "S <u a,  cs H  X O  VO cn  in" vo  CN vO cn cn CN VO CN  O H o  o  VO  m oo" m  CN  CQ  • 3 - vo m CN CN cn w m cn  O m >* vo oo m vo  CN CN CN CN  > O CQ H O in oo oo —  ^?  oo m m cn oo" O >-  o  80  The triglycoside, spot 6 on the 2-D profiles and faintly visible in fractions of both taxa, was not in quantities sufficient for purification. During the overview of taxa it was found that a triglycoside exhibiting the same chromatographic behaviour was a constituent in all the profiles (Figures 28-34 Spot 6). The concentration in 71 sessile was sufficient to purify and identify as quercetin-3-O-rhamnose/diglucoside (Table 12-14). It was impossible to work with equivalent amounts of plant material from the two taxa, due to the size and relative rarity of 71 "hibbersonii". Approximately 4.5 grams of T. "hibbersonii" leaf material was used for LH-20 Column 3 in comparison to an excess of 30 grams of T. ovatum leaf material for Column 2. A quantitative difference between the taxa was indicated during the purification process, especially for the diglycosides, based on intensity of the band and number of plates spotted. This led to the decision to use  71 "hibbersonii"  for the  compound identifications and sugar analysis despite the smaller initial quantities of material. Fractions of 71 ovatum which could be adequately purified were also used for corroboration of compound identification. Analysis using high performance liquid chromatography would be required to quantify the indicated concentration differences. Impact of the environment and/or developmental stages onflavonoidexpression would also have to be considered if using a sensitive quantitative technique.  81 4.8.1 Random amplified polymorphic DNA (RAPD) Literature Review  No prior information on the target genome is required for the RAPD technique (Williams et al., 1990) which is based on the polymerase chain reaction (PCR) technique originally developed by K.B. Mullis (1986). The assay involves the use of a thermostable D N A polymerase, oligonucleotide primers, nucleotide triphosphates (dNTP), reaction buffer and the target D N A template. For conventional PCR specific nucleotide sequence information is required for constructing primers which are complementary to the 3' ends flanking the D N A sequence to be amplified. In the RAPD technique, RAPDs are generated using random 10base oligonucleotide primers with a high GC content (50% or greater) which facilitates binding of primer to DNA template for detectable levels of amplification products (Williams et al., 1990). Polymorphisms are visualized and scored directly from ethidium bromide stained agarose gels. Resolution of the gel is a limiting factor for band identification. The sensitivity of RAPD primers to single-base changes in the primer target sites (Williams et al, 1993) has applications for molecular taxonomy particularly for phylogenetic analysis of individuals that are closely related. Wolff et al. (1993) used RAPDs for identification of cultivars of Dendranthema grandiflora but found that genetic variability among Dendranthema  species was too high to study genetic distances.  The Stylosanthes giiianensis assess genetic  species complex was investigated by Kazan et al. (1993) to  relationships and variation using RAPDs. Polymorphisms between 45  accessions provided information for calculating genetic distance from which a phenogram was constructed to illustrate genetic relationships among the taxa investigated. The differentiation at the D N A level supported previous classifications of the taxa as distinct species and taxa  82  relationships based on morphological-agronomic characters, seed protein patterns, rhizobial affinities, crossability and pollen stainability of the hybrids was reflected in the phenogram (Kazan etal, 1993). The objective of another study of genome relationships using RAPDs was to distinguish four diploid Lotus species and to provide information on the origin of the tetraploid Lotus corniculatus (Campos et t//., 1994). Although this was an introductory study of a large genus,  the polymorphisms among Lotus species indicating relationships were reported to closely parallel results from meiotic chromosome analyses, crossing behaviour and isoenzyme studies (Campos et al, 1994). RAPDs are also useful for discriminating polymorphisms within the species (Williams et ai, 1993; Campos et al 1994) and can be used to discriminate between geographical origins (Russell etal, 1993). A combination of RAPDs and standard PCR was used by Arnold et al. (1991) for the identification of random'and specific genetic markers for testing an evolutionary genetic hypothesis concerning the hybrid origin of Iris nelsonii. 4.8.2 Random amplified polymorphic DNA (RAPD) Results Trillium species (T. grandiflorum,  T. sessile, T. stylosum, T. kamtschaticum  Table 15) were  included for positive control and comparison during optimization experiments. Poplar genomic DNA {Populus trichocarpa x deltoides) provided an unrelated positive control. Negative controls were also included in each set. Primers from the 10-Mers Set 100/1 were obtained from the University of British Columbia Nucleic Acid - Protein Service Unit. Of the 19 primers tested (UBC 1-10, 20, 30, 40, 50, 60, 70, 80, 90, 100) only 7 showed  83  Table 15. List of primers used and the respective oligonucleotide sequence. Trillium species  which showed any amplification products are listed. H - T. "hibbersonii", 0 - T. ovatum, G - T. grandiflorum, St - T. stylosum, K-T. Seqilnce  %'S pIlies^feplififelM  1  CCT G G G CTT C  G, St  2  CCT G G G CTT G  3  CCT G G G C T T A  4  CCT G G G C T G G  5 .  CCT G G G T T C C  6  CCT G G G CCT A  7  CCT G G G GGT T  8  C C T G G C GGT A  Prime,  9  .  K  H, O, G, St  CCT G C G C T T A  10  GGG GGG ATT A  20  TCC GGG TTT G  30  CCG GCC TTA G  40  T T A CCT G G G C  50  TTC CCC G C G C  60 v  T T G GCC G A G C  H, O  . 70  GGG CAC GCG A  H, O  80  GTG CTC T A G A  90 '  G G G GGT T A G G  100  ATC GGG TCC G  H,0  H, O  kamtschaticum.  84  amplification products for any of the Trillium species tested (Table 15). Of the primers which produced amplification products in T. "hibbersonii" and T. ovatum UBC 6, UBC 70 and UBC 100 produced discernable polymorphic bands while primers UBC 30 and UBC 60 should be retested (Figures 37 & 38). The objective of this section was an introductory investigation of the feasibility of using RAPDs as a tool to study and characterize the relationship between Trillium  "hibbersonii"  and T. ovatum at the genome level. Optimization of reaction conditions, especially MgCl and concentration of DNA, proved to 2  be a lengthy process which used up a large part of the limited supplies of DNA. The ramp time of 0.1° C s" for the transition to the annealing portion of the cycle (35° C to 72° C) is 1  very long (>6 min.) compared to 2 minutes (0.3° C s") which was also tried. Williams et al. 1  (1993) and Penner et al. (1993) use the fastest available transitions between temperatures. The faster transition failed to produce any amplification products/This lack of results could also be attributed to other variables (DNA concentration, appropriate primer for the species) but as the longer ramp time had produced results it was used in all subsequent amplification cycles. For reproducibility of results the same conditions must be used. The type of thermocycler and the overall temperature profiles are the most important variables to affect reproducibility in R A P D s (Penner et al, 1993). The amount of DNA used in a reaction is also important to reproducibility of results and a strong signal (Williams et al, 1993). Gel quantification of the extracted DNA samples differed from the geneQuant by a factor of two or greater which may have contributed to lack of definable bands or smears.  85  No conclusions can be based on these polymorphism because it is a very introductory investigation and only compares individuals. However, it is a good starting point for further investigation. A working protocol is in place and three primers have been identified which amplify DNA from both of the key taxa in this study and Primer UBC 6 successfully amplifies DNA tracts from other taxa in the genus. Polymorphism between the taxa has been demonstrated in the present work. Further investigation would have to consider intra and inter-populational polymorphisms. Geographic origin has been discriminated using RAPDs (Russell et al, 1993). Since Trillium ovatum has a large range, and geographic origin can be discerned based on karyotype (Fukuda and Channel, 1975), RAPDs may be useful and lead to information on the taxonomic relationship between T. ovatum and T. "hibbersonii".  86  Figure 37. Amplification of T. "hibbersonii" and T. ovatum using primer #6. Polymorphisms are visible in the 975 bp and 750 bp range.  6 100 H 0 MW 0 H  70 60 O H O H MW  n  1 —  *—  4072 3054 = l«5 —1011 5165 /06 2(a6  O - T.  "hibbersonii"  H - T. ovatum  M\V - molecular weight std. 1 kb  Figure 38. Amplification of T. "hibbersonii" and T. ovatum using primers #6, #100, #70, #60. Polymorphisms are visible with primer #6 (repeat), #100 and #70 all in the 600 to 1018 bp range.  87  5. Discussion 5.1 Taxonomic Considerations The subdivisions of Species are Subspecies, Varietas, Subvarietas, Formae and Subformae, in descending order. Despite the ordered ranking of infraspecific taxa under the rules of the International Code of Botanical Nomenclature the distinction of the ranks of subspecies and varietas is not clear and conventional usage differs. In fact, subspecies and varietas are used interchangeably. At present two infraspecific ranks have been used in Trillium ovatum, that of subspecies and forma. Trillium "hibbersonii" is presently placed at the rank of forma based on the diagnosis of Taylor and Szczawinski (1975). According to Cronquist (1988) distinctive phenotypes that occur within a 'typical' population and do not have persistent populational significance are denoted by the rank of formae. Formae are also described as differences of a minor nature caused by environmental factors (Porter, 1967). T. "hibbersonii" does not occur as an occasional abberation amongst populations of T. ovatum. In the native habitat T. "hibbersonii" exists in populations, which interbreed and produce offspring of the same type. Therefore the rank forma is clearly not a suitable designation. Subspecies occur within a definable geographical range and may be ecologically or morphologically separated. Subspecies do form populations and are not generally genetically isolated so interbreeding and continuous intergradation with other subspecies of the same species occurs when their ranges meet (Griffiths and Ganders, 1983).  88 A species, such as T. ovatum, which occupies extensive geographical area is composed of ecological races as outlined by Turesson (in Porter, 1967), where physiological fitness and selection pressures affect survival not morphological characters, as a result the ecotypes have often been interpreted as species. Fukuda and Channell (1975) illustrated the diversity of karyotypes in T. ovatum, and related these characters to morphological characters and evolutionary pressures. The different environmental conditions on the Pacific coast and in the Rocky Mountains and the karyotype variations that have developed in these now isolated regions may be on the path of divergence resulting in two distinct species (Fukuda & Channell, 1975). Trillium ovatum Pursh ssp. oettingeri Munz & Thorne has a small, definite range at altitude in the Salmon Mountains of Siskiyou County, California. The plant is described as smaller in stature, with a narrower petal, smaller flower, shorter peduncle and leaves that are smaller and more distinctly petiolate than T. ovatum ssp. ovatum. No information on interbreeding, intermediates and germination is available. Trillium "hibbersonii" occupies a limited geographic range on the extreme west coast of Vancouver Island. Trillium "hibbersonii" is morphologically distinct, forms populations and occupies a different ecological niche than T. ovatum. The criteria listed in the previous paragraph would allow for the treatment of T. "hibbersonii" as a subspecies. The following characteristics of T. "hibbersonii" indicate that the status of species is more appropriate: i)  differences in germination and dormancy requirements  ii)  stigma/stamen position indicating the promotion of outbreeding  89  iii) lack of morphological intermediates in the U B C Botanical Garden Plants under favourable conditions: a) presence of seedlings of different ages b) spatial proximity for 19 years c) concurrent or overlapping flowering d) sharing common pollinators iv) no intermediates were noted in examination of herbarium material which covered the full range of T. ovatum v) no intermediates were sited by Taylor and Szczawinski despite their claim that T. "hibbersonii" could be found throughout the range of T. ovatum vi) attempts to cross-pollinate failed to produce viable seed set. vii) flower pink in bud and throughout anthesis to withering. The application of the specific rank to T. "hibbersonii" fulfils the most basic criteria of species according to Cronquist (1988), that a species must be distinguished by ordinary means. The notched box plots illustrate that morphological measurements can be used to differentiate between the taxa especially if combined with stigma/anther relation, or colour of petals especially in bud or at anthesis. Colour variation is well documented within the genus and many colour forms have been described and named so this character cannot be used alone. Freeman (1975) and Murrell (1969) noted variation in every sessile-flowered species in the eastern United States but found the colour of floral parts in Trillium is most important in a population sense, especially paired with a well-defined geographic range (Freeman, 1975).  90 Limited geographical distribution, different ecological niche in combination with strong morphological divergence, different dormancy/germination requirements, differences in minor flavonoids and quantitative differences in the major flavonoid constituents, indications o f D N A polymorphism are compelling reasons for raising the status o f T. "hibbersonii" to the rank o f species. The raising o f T. "hibbersonii" to the specific status would also result in the designation o f T. "hibbersonii" as an endangered species. 5.2 Phylogeny Berg (1958) postulated one ancestral species which possessed an erect stem and peduncle, white petals, a berry and endozoochory (Figure 39). Trillium grandiflorum is a representative o f Berg's stage 2. The white petals and pulpy berry o f the ancestral species are retained but the fruit may be erect or declining and myrmecochory is established. F r o m Trillium grandiflorum to T. ovatum (Berg's stage 5) is marked by the loss o f endozoochory and the pulpy berry while myrmecochory is further established as the method o f seed dispersal. The pulpy berry is replaced by a non-pulpy capsule with regular to irregular dehiscence. The flower is erect and the fruit declines. These changes were probably accompanied, or followed, by separation due to glaciation. B e r g considers another western North American pedicellate species, Trillium rivale (stage 6), to be a derivation o f the T. ovatum type (stage 5) that is adapted for a deep shade habitat. The development o f petioles places the leaves in a better position for light and energy assimilation. Gates (1917) also indicated a relationship between T. ovatum and T. rivale, with T. rivale putatively derived from T. ovatum as a dwarf mutant.  91 1 Ancestral species erect stem and peduncle whiteflower,probably nectar berry and endozoochory  Establishment of myrmecochory Trillium grandiflorum white, erect flower berry declining or erect endozoochory myrmecochory Dwarf habit Neutral to calcareous soils Development of petioles  Loss of endozoochory Loss of berry (pulp) Separation (by glaciation)  Trillium ovatum white, erect flower capsule declining myrmecochory  Trillium nivale white flower, pink on dorsal flower sub-erect fruit declining myrmecochory  Habitat in deep shade Development of petioles  Trillium ovatum subspecies oettingeri white, nodding flower capsule declining  Dwarf habit Habitat more exposed, undeveloped soil Change in dormancy requirements Increased outcrossing Increased ultra-violet protection  Dwarf habit Seasonal moisture Change in dormancy requirement  Trillium rivale whiteflowerwith pink/purple spots flower erect then recurved capsule recurved myrmecochory  Figure 39. Hypothesis for the phylogeny  Trillium "hibbersonii" pinkflower,bend in peduncle capsule declining stigma extends beyond anthers  of Trillium based on Berg (1958).  92 Trillium ovatum ssp. oettingeri occupies a shadier habitat than T. ovatum and has also developed short petiolate leaves. The relationship between T. ovatum and T. "hibbersonii" has been variously described: •  T. ovatum forma hibbersonii (Taylor & Szczawinski, 1975)  •  A dwarf and colour mutant of T. ovatum (Griffiths & Ganders, 1983)  •  Derived by mutation from T. ovatum (Beamish, annotated herbarium label U B C #73131)  •  A natural hybrid with one of the putative parents T. ovatum (Wiley, 1968)  •  A hybrid of T. rivale and T. ovatum (Mitchell, 1969)  The habitat occupied by T. "hibbersonii" is more exposed, drier and with less soil development than the wooded habitat of T. ovatum. The increase in solar light and potential damage to the plants due to U V radiation can be offset by changes in the protective mechanisms of the plant. Folding of the leaf about the midline to minimize surface area during times of high exposure and the apparent increase in leaf thickness of T. "hibbersonii" over T. ovatum (qualitative observation) are both possible protective mechanisms that have developed in T "hibbersonii". Both flavonols and anthocyanin have been found to provide protection from ultra-violet radiation (Li etal.,, 1993; Takahashi, 1991). The protective role of these compounds may be utilised in T. "hibbersonii" and reflected in the qualitative increase in several of the flavonol compounds isolated from the leaves compared with the qualitative amounts present in T. ovatum. The increased presence of anthocyanin in T. "hibbersonii" was denoted by red margins on the leaves and sepals as well as the presence in the stem, pedicel, petals and anther connective. Even under similar growth and light conditions the deposition  93 pattern of anthocyanin in 71 ovatum differed from 71 "hibbersonii". Red leaf margins were not present while sepal and petal colouration correlates to aging and senescence. Trillium rivale and T. "hibbersonii" are geographically separate but both ranges lie within the range of T. ovatum. Trillium rivale and T. "hibbersonii" occupy different ecological niches although both taxa appear to depend on the availability of seasonal moisture in the form of seepage. These two taxa also share two features; both are without the dormancy requirements of T ovatum and the ovaries are 3-lobed, sub-globose to globose and without the prominent wings of T. ovatum. The anthers of T. rivale are positioned higher than the stigma, as in T. ovatum, but the stigmas are short and club-shaped not recurving as in both T. ovatum and T. "hibbersonii". Fertile seeds were produced from the cross of T. ovatum x T. rivale and the reverse cross (Christian, 1980a) but none had reached the flowering stage at the time of the article. 6. Conclusions Trillium "hibbersonii" is endemic to the west coast of Vancouver Island with a small geographic range. Herbarium specimens have been collected from two locations. The Danthonia intermedia - Cladonia species community which includes T. "hibbersonii" indicate a different ecological niche from the environment indicated by T. ovatum (Klinka et al, 1989) in terms soil type and development, moisture and amount of light exposure. Morphological divergence has progressed to the point that T. "hibbersonii" and T. ovatum can be separated using morphological characters, both qualitative and quantitative. Discriminant analysis confirmed the grouping by taxa although a larger data set for T. "hibbersonii" and inclusion of qualitative characters such as stigma/anther relation would  94 improve the soundness of this method. The exsertion of the stigma beyond the anthers indicates the promotion of outcrossing in T. "hibbersonii". Different dormancy/germination requirements result in the germination of T. "hibbersonii" after one cold period without the double dormancy found in T. ovatum. The quantitative differences in the major flavonoid constituents between T. "hibbersonii" and T. ovatum and the presence of a minor flavonoid constituent in the 71 ovatum profile indicate a close relationship of the taxa but a discernable one using quantitative techniques. Introductory work using random amplified polymorphic D N A (RAPD) indicates polymorphisms are present. Trillium "hibbersonii" fulfils the criteria for specific status. No morphological intermediates were noted during examination of herbarium specimens or in the U B C Botanical Garden and none were cited by Taylor and Szczawinski in their diagnosis of T. "hibbersonii" as a form (1975). The recognition of Trillium "hibbersonii" as a species requires publication in a recognized journal according to the International Code of Botanical Nomenclature regulations. The type specimen will be maintained (UBC # 73131) and the new description of the species Trillium hibbersonii (Taylor & Szczawinski) OTSfeill and Straley will be applied.  95 7. Literature Cited Abrams, L . 1923. Illustrated Flora of the Pacific States, Vol. 1. Stanford Univ. Press. Stanford, California. Adams, A.B. 1975. A study of Trillium  cuneatum and T. luteum. M.Sc. Thesis, University of  Tennessee, Knoxville. Arnold, M L . , C M . Buckner, and J.J. Robinson. 1991. Pollen-mediated introgression and hybrid speciation in Louisiana irises. Proc. Natl. Acd. Sci. 88:1398-1402. Asbury, S.S. 1973. The anthocyanin pigments of sessile-flowered Thesis, George Peabody College for Teachers, Nashville.  Trillium (Liliaceae). M . A .  Province of B.C. Ministry of Forests Research Branch. 1988. Biogeoclimatic Zones of B . C . Bailey, P C . 1958. Differential chromosome segments in eight species of Trillium. Bull. TorreyBot. Club 85:201-214. Bailey, P.C. 1954. Differential reactivity in chromosomes as an indicator of species in  Trillium. Bo. Gazette. 115:241-248. Beeks, R . M . 1955. Improvements in the squash technique for plant chromosomes. Aliso 3:131-133. Berg, R.Y. 1958. Seed dispersal, morphology, and phytogeny of Trillium. Skr. Norske Vidensk.-Akad. Oslo, Mat.-Naturvidensk. Kl. 1958:1-36. Bohm, B.A. 1987. Intraspecific flavonoid variation. Botanical Review. 53:197-279. Campos, L.P., J.V. Raelson and W.F. Grant. 1994. Genome relationships among Lotus species based on random amplified polymorphic D N A (RAPD). Theor. Appl. Genet. 88:417-422. I;  Case, Frederick W. Jr. 1988. Trillium propagation and conservation. Fine Gardening. 2:36-37. Ceska, A., O. Ceska, R. Ogilvie, and Ff. Roemer. 1982. Report #358 Hesquiat Lake. Ecological Reserves Unit, B.C. Min. of Lands, Parks & Housing. 12 pp. Christian, Paul. 1980a. Trilliums, the Western American species. Part 2. Bull. Alpine Gard. Soc. 48(2) 139-143.  96 Christian, Paul. 1980b. Trilliums, propagation and cultivation. Part 4. Bull. Alpine Gard. Soc. 50(2)339-342. Clark, L.J. 1973. Wild Flowers of British Columbia. Gray's Publishing Limited, Sidney, B.C. Cronquist, Arthur. 1981. A n Integrated System of Classification of Flowering Plants. The New York Botanical Garden. Columbia University Press, New York. 1262 p. Cronquist, Arthur. 1988. Evolution of the Flowering Plants. The New York Botanical Garden. Bronx, New York. 2nd ed. 555 pp. Dahlgren, H . , T. Clifford, and P.F. Yeo. 1985. The Families of the Monocotyledons. Springer-Verlag, Berlin. Darlington, C D . , and L.F. LaCour. 1940. Nucleic acid starvation of chromosomes of  Trillium.!. Genet. 40:185-213. Darlington, C D . , and G.W. Shaw. 1959. Parallel polymorphism in the heterochromatin of Trillium species. Heredity l3(l):89-121. Dogwood, Rhododendron and Trillium Protection Act. 1960. British Columbia Laws, Statutes, etc. 119:1209-1210. . Duncan, W.H., and U.T. Calhoun. 1962. Data on sessile-flowered Trillium (Liliaceae). Castanea 27:165-173. Dyer, A.F.1963. Endosperm development after controlled pollination within and between species of Trillium and Paris. Chromosoma 14:549-567. Escote-Carlson, L.J. 1991. Plant Molecular Biology Laboratory Course manual. Biotech Lab. U B C , Vancouver. 127 pp. Freeman, J.D. 1969. A revisionary study of sessile-flowered Trillium L . (Liliaceae). Ph.D. Thesis. Vanderbilt University, Nashville, Tennessee. 523 pp. Freeman, J.D. 1975. Revision of Trillium subgenus Phyllantherum (Liliaceae). Brittonia 27:1-62. Fukuda, I., and R.B. Channel. 1975. Distribution and evolutionary significance of chromosome variation in Trillium ovatum. Evolution 29(2):257-266. Fukuda, I., and W.F. Grant. 1980. Chromosome variation and evolution in Trillium grandiflorum. Can. J. Genet. Cytol. 22:81-91.  97 Fukuda, I. 1987. Mode of speciation in Trillium - Chromosme evolution and differentiation of the reproductive system. Am. J. Bot. 74:641-642. Gates, B.N. 1940. Dissemination by ants of the seeds of Trillium grandiflorum. Rhodora, 42:194-196. Gates, B.N. 1941. Observations in 1940 on the dissemination by ants of the seeds of Trillium grandiflorum. Rhodora, 43:206-207. Gates, R.R. 1917. A systematic study of the North American genus Trillium, its variability, and its relation to Paris mdMedeola. Ann. Mo. Bot. Gard. 4:43-87. Gleason, H A . 1906. The pedunculate species of Trillium. Torrey Bot. Club 33: 387-396. Goodspeed, T.H., and R.P. Brandt. 1916. Notes on the Californian species of Trillium. Univ. Cal. Publ. Bot. 7:1-38. Gornall, R.J., and B A . Bohm. 1980. The use of flavonoids in the taxonomy ofBoykinia and allies (Saxifragaceae). Can. J. Bot. 58:1768-1779. Gornall, R.J., a n d B A . Bohm. 1978 Angiosperm Flavonoid Evolution: A Reappraisal. Sys. Bot. 3(4) 353-368. Gotoh, K. 1937. Jap. J. Genetics 13:209. Griffiths, A.J.F., and F.R. Ganders. 1983. Wildflower Genetics. Flight Press. Vancouver, B.C. Gunther, R.W., and J. Lanza. 1989. Variation in attractiveness of Trillium diaspores to a seed-dispersing ant. Am. Midland Naturalist. 122(2):321-328. Guppy, Art. 1992. Two Trilliums. Alpine Gdn Club of B C Bull. 35(3):52-53. Harborne, J.B., M . Boardley, H.P. Linder. 1985. Variations in Flavonoid Patterns within the Genus Chondropetalum (Restionaceae). Phytochemistry 24(2):273-278. Hauser, E . , and J.H. Morrison. 1964. The cytochemical reduction of nitro blue tetrazolium as an index of pollen viability. Am. J. Bot. 51:748-752. Hickman, J.C. ed. 1993.The Jepson Manual. Higher Plants of California. Univ. of California Press. Berkeley and Los Angeles, Cal. Hunt, F F . 1991. Q. Bull. Alpine Gard. Soc. 59(4) 386.  98  Hutchinson, J. 1973. The Families of Flowering Plants Arranged According to a New System Based on Their Probable Phylogeny. 3rd Ed. Clarendan Press, Oxford. Hutchinson, J. 1959. The Families of Flowering Plants Vol. TJ 'Monocotyledons'. Clarendon, Oxford. Ihara, M . 1973. A facet of the breeding system in American species of Trillium. Unpubl. M.Sc. Thesis, Vanderbilt Univ. Nashville, Tenn. 221 pp. Jeffrey, E . C , and E.J. Haertl. 1939a. Apomix in Trillium. Cellule 48:79-88. Jeffrey, E . C , and E.J. Haertl. 1939b. The production of unfertilized seeds in Trillium. Science (Wash. D.C.) 90:81-82. Kazan, K., J.M. Manners and D.F. Camerson. 1993. Genetic relationships and variation in the Stylosanthes guianensis species complex assessed by random amplified polymorphic DNA. Genome 36:43-49. Klinka, K., V.J. Krajina, A. Ceska and A.M. Scagel. 1989. Indicator Plants of Coastal British Columbia. UBC Press. Vancouver, B.C. Krause, K. 1930. In Die naturllichen Pflanzenfamilien (Engler, A. and Prantl, K., eds) Ed. 2, Bd 15a Wilhelm Engelmann Leipzig. Kuijt, Job. 1982. A Flora of Waterton Lakes National Park. Univ. of Alberta Press. Edmonton, AB. Kurabuyashi, M . 1958. Evolution and variation in Japanese species of Trillium. Evolution 12:286-310. Kurabuyashi, M . 1963. Karyotype differentiation in Trillium sessile and T. ovatum in the western Unites States. Evolution 17:296-306. Lee, Penny. 1995. Trilliumsfromseed. Wildflower. Winter 1995 36-37. Les, Donald H., R. Whitkus, F.A. Bryan and L.E. Tyrell. 1989. Biochemical basis of floral color polymorphism in a heterocyanic population of Trillium sessile (liliaceae). Amer. J. Bot. 76(1)): 23-29. Li, Jiayang, Tsai-Mei Ou-Lee, R. Raba, R. G. Amundson and R.L. Last. 1993. Arabidopsis Flavonoid Mutants Are Hypersensitive to UV-B Irradiation. The Plant Cell 5:171-179.  99  Lighty, R.W. 1993. Garden use and species conservation need not conflict with research. Public Garden 8(4):32-33. Mabry, T.J., K.R. Markham and M.B. Thomas. 1970. The Systematic Identification of Flavonoids. Springer-Verlag. New York. 354 pp. McGill, R., J. W. Tukey, and W.A. Larsen. 1978. Variations of box plots. Amer. Statistician, 32:12-16. Maneval, W.E. 1936. Lactophenol preparations. Stain Tech. 11:9. Markham, K.R. 1982. Techniques of Flavonoid Identification Biol. Tech. Series. Academic Press. London. 113 pp. Mesler, JR., and K.L. Lu. 1983. Seed dispersal of Trillium ovatum in second growth redwood forests. Am. J. Bot. 70:1460-1467. Mitchell, R.J. 1969. The Genus Trillium. J. Sottish Rock Gard. Club. 11:271-283 Mitchell, R.J. 1970. Genus Trillium - II. J. Scottish Rock Gard. Club. 12(46): 16-20. Mitchell, R.J. 1989. Trillium, part 2 - Western North America. The Plantsman, 2(2):67-79. Mullis, K B . 1990. The unusual origin of the polymerase chain reaction. Sci. Am. April. 5665. Munz, P. A. 1959. Flora of California. Univ. Calif. Press. 1681 pp. Munz, P.A., andR.F. Thorne. 1973. A new northern California Trillium. Aliso 8(1): 15-17. Murrell, J.T. Jr. 1969. A Chemotaxonomic Study of Sessile-flowered Trillium (Liliaceae). Unpubl. Ph.D. Thesis, Vanderbilt University, Nashville, Tennessee. 303 pp. Nakano, K., A. Maruhashi, T. Nohara, T. Tomimatsu, N . Imamura and T. Kawasaki. 1983. A flavonol glycoside and a sesquiterpene cellobiosidefromTrillium tschonoskii. Phytochemistry. 22(5): 1249-1251. Nesom, G.L., and J.C. La Duke. 1985. Biology of Trillium nivale (Liliaceae). Can. J. Bot. Vol. 63:7-13. Ohara, Masashi, K. Okazaki, and S Kawano. 1991. Predominant inbreeding and resource limitation in four Japanese Trillium species. Miyabea. 1: 21-30.  100 Penner, G.A., A. Bush, R. Wise, W. Kim, L. Domier, K. Kasha, A. Laroshe, G. Scoles, S.J. Molnar and G. Fedak. 1993. Reproducibility of random amplified polymorphic DNA (RAPD) analysis among laboratories. PCR Meth. & Appl. 2:341-345. Porter, C.L. 1967. Taxonomy of Flowering Plants. 2nd Ed. W.H. Freeman & Co. New York.  Pursh, F. 1814. Flora Americae Septentrionalis i. p. 245. Rendle, A.B. 1901. Notes on Trillium. J. Bot. 39: 321-335. Renner, M A . 1980. Studies in Trillium (Liliaceae) the pedicellate Trillium of the Pacific Northwest. Unpubl. M A . Thesis, Humboldt State University, California. Richards, E. 1987. Preparation of genomic DNA from plant tissue. In Current Protocols in Molecular Biology. Wiley Interscience. pp 231-233. Robertson, C. H. 1896. Flowers and insects. XVI. Bot. Gaz., 221:266-274. Rogers, S.O., and A.J. Bendich. 1988. Extraction of DNA from plant tissues. In Plant Mol. Biol. Manual A6: 1-10. Kluver Acad. Pub. Dordrecht. 5:69-76. Russell, JR., F. Hosein, E. Johnson, R. Waugh and W. Powell. 1993. Genetic differentiation of cocoa (Theobroma cacao L.) populations revealed by RAPD analysis. Mol. Ecol. 2:89-97. Sambrook, J., E.F. Fritsch, and T. Maniatis. 1989. Molecular Cloning, A Laboratory Manual. Cold Spring Harbor Lab. press, Cold Spring Harbor, NY. Samejima, J., and K. Samejima. 1962. Studies on the eastern Asiatic Trillium (Liliaceae). Acta HortiGothob. 25:157-257. Samejima, J., and K. Samejima. 1987. Trillium Genus Illustrated. Hokkaido University Press. Smith, B., P.D. Forman and A.E. Boyd. 1989. Spatial patterns of seed dispersal and predation of two myrmecochorous forest herbs. Ecology 70(6): 1649-1656. Sparrow, A . H . , U. Pond, and R.C. Sparrow. 1952. Amer. Nat. 86:277. Stapleton, Ann E. and V. Walbot. 1994. Flavonoids can protect Maize DNA from the Induction of Ultraviolet Radiation Damage. Plant Physiol. 105: 881-889.  101  Straley, G.B., R.L. Taylor, and G.W. Douglas. 1985. The Rare Vascular Plants of B.C.. National Museums of Canada. Syllogeus No. 59. Takahasi, Akihisa, K. Takeda and T. Ohnishi. 1991. Light-Induced Anthocyanin Reduces the Extent of Damage to DNA in UV-Irradiated Centauria cyanus Cells in Culture. Plant Cell Physiol. 32(4): 541-547. Takahashi, M . 1982. Pollen morphology in North American species of Trillium. Amer. J. Bot. 69(7):1185-1195. Takhtajan, A.L. 1969. F l o w e r i n g Plants. O r i g i n a n d Dispersal. Oliver and Boyd, Edinburgh. Takhtajan, A.L. 1980. Outline of the classification of the flowering plants. Bot. Review 46:225-359. Takhtajan, A.L. 1986. Floristic Regions of the W o r l d , (translated by Theodore J. Crovello) Univ. of Cal. Press, Berkeley Cal. 522 pp. Taylor, R.L., and B. MacBryde. 1977. Vascular Plants of B . C . : Descriptive Resource Inventory.  Tech. Bull. 4. The Botanical Garden, UBC. Vancouver.  Taylor, R.L., and S. Taylor. 1977. Trillum ovatum Pursh. Davidsonia. 8(1):9-14. Taylor, T.M.C. and A.F. Szczawinski. 1975. Trillium ovatum forma hibbersonii Taylor et Szczawinski f. nov. . Syesis 7: 250. Thomas, J.H. 1961. F l o r a of the Santa C r u z M o u n t a i n s of C a l i f o r n i a . Stanford Univ. Press. Stanford, California. Turner, B.L., and R. Alston. 1959. Segregation and recombination of chemical constituents in a hybrid swarm of Baptisia laevicaulis X B. viridis and their taxonomic implications. Am. J. Bot. 46.678-686. Uchino, A., H. Kanazawa and H. Watanabe. 1989. Structure and dynamics of natural populations of polyploid trilliums: V. Predominant self-pollination in Trillium apetalon and Trillium smallii association. Plant Species Biol. 2(1-2): 101-108. Warmke, H E . 1937. Cytology of the Pacific Coast Trilliums. Amer. J. Bot. 24:376-383. Wiley, L. 1968. R a r e W i l d Flowers of N o r t h A m e r i c a . Leonard Wiley, Portland, Oregon.  102  Wilkins, C.K., and B.A. Bohm. 1976. Flavonoids ofHeuchera micrantha var. diversifolia. Can. J. Bot. 54:2133-2140. Wilkinson, L. 1990. SYGRAPH: The system for Graphics. Evanston, IL: SYSTAT, Inc. Wilkinson, L. 1990. SYSTAT. Evanston, IL: SYSTAT, Inc. Williams, C A . 1975. Biosystematics of the Monocotyledoneae - Flavonoid patterns in leaves of the Liliaceae. Biochem. Syst. Ecol. 3:220-244. Williams, J.G.K., A. Kubelik, K J . Livak, J. A. Rafalski and S.V. Tingey. 1990. DNA polymorphisms amplified by arbitrry primers are useful as genetic markers. Nucleic Acids Res. 18(22): 6531-6535. Williams, J.G.K., M . Hanafey, JA. Rafalski and S.V. Tingey. 1993. in "Methods in Enzymology" p. 704. Academic Press. Wolff, K., and J. Peters-van Rijn. 1993. Rapid detection of genetic variability in chrysanthemum (Dendranthema grandiflora Tzvelev) using random primers. Heredity 71(4):335-41. Yoshitama, K., N . Ishikura, T. Maso, S. Oka, M . Kai and A. Uchino. 1992. The flavonoid chemistry of the genus Trillium. I. Flavonol glycosides in the leaves of Trillium tschonoskii Maxim, and their geographical variations. Bot. Mag. Tokyo 105: 555-563. Yoshitama, K., Maki Kai, Yuko Shida and Toru Oyamada. 1994. Flavonoid glycosides of Trillium species (Liliaceae). Absts of the XVII Int. Conf. Groupe Polyphenols, Polyphenols Actualites 11.  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items