Morphological and cytological variation in Trillium albidum Freeman (Liliaceae)

Graduation date: 1993This work was performed on Trillium albidum Freeman (Liliaceae), to\ud clarify the pattern of morphological variation between the small-flowered\ud populations at the northern limit of its range ( T. parviflorum Soukup) and the\ud larger-flowered populations from southern Oregon and northern California. It was\ud hypothesized that only a single species of sessile-flowered Trillium occurs in the\ud range of Trillium albidum, from the Puget Sound area in Washington to Northern\ud California. Eighteen sites were visited in Washington and Oregon. Morphological\ud data taken from field-collected specimens included size of parts, ratios of length and\ud width of parts, and color of parts. Rhizome shoot scars on herbarium specimens\ud were counted to estimate relative plant age. Plant ovaries from the field were cold-treated\ud and preserved for karyotyping. It was found that plant height, petal length,\ud and leaf length positively correlated with rhizome shoot scars. Cluster analysis\ud excluding size characteristics showed a tendency of populations to form groups\ud corresponding to three geographical areas: Washington and northern Willamette\ud Valley, central Willamette Valley to the Umpqua River, and populations south of\ud the Umpqua divide. Principal components analysis revealed two groups: the\ud southern-most Oregon populations, and populations from the Umpqua River\ud northward. Karyotypes give some additional support to these patterns through the\ud limited geographical occurrence of particular chromosome types. Plants in\ud Washington showed heterozygosity of chromosomes, and a limited karyotype.\ud Chromosome BO and the heterozygous combination of E4-8 was found only in\ud plants in the Willamette Valley. One population Jackson County, Oregon has plants\ud with chromosome E6, which was not found in other populations. Though\ud population groupings were revealed by multivariate statistics, without further study\ud it was not desirable to give them formal infraspecific names. The percentage of\ud taxonomically ambiguous individuals was too high, and the amount of overlap in\ud variation too great, to permit a formal species distinction between the southern and\ud northern populations, though subspecies might be recognized after further study.\ud Based upon these data, the recognition of T. parviflorum is not justified