Ecological life history of Polymnia canadensis, a monocarpic species of the North American Temperate Deciduous Forest: Demography

A demographic investigation was conducted to assess variation in life histo ry of Polymnia canadensis (Asteraceae), a geographically-widespread, herbac eous species of deciduous forests in eastern North America. During 1985-199 4, 23,063 seedlings of P. canadensis were monitored at five central Kentuck y study sites. Numbers at the end were: biennials, 554; triennials, 142; wi nter annuals, 16; monocarpic perennials, 2; tricarpic perennials (three yea rs), 3; and dicarpic perennials (some skipped years) that matured in the fi rst year of life, 23; in the second year, 60; and in the third year, 9. Wee kly cohorts of P. canadensis generally exhibited Deevey Type III survivorsh ip with highest seedling mortality in summer associated with low soil moist ure. Wide spatial variation in life history was displayed by the fact that fall germination cohorts at dry sites generally had greater germination and survivorship than at mesic sites during seedling establishment, while the reverse was true for spring cohorts. This led to more reproductive individu als in fall germination cohorts than spring cohorts at dry sites and genera lly the opposite case at mesic sites. Forest shade in mesic sites caused sl ower growth and a greater frequency of longer-lived reproductive individual s than in open, dry sites, but it also resulted in higher survivorship than at dry sites during moderate drought. Annual population growth rate averag ed across four years was not significantly different between a dry site and mesic site, indicating that despite forest shade, P. canadensis persisted in the mesic site as well as it did in the dry site. Population structure v aried among years, seasons, and study sites; at two study plots, a 2-year f lowering cycle of mass seeding and senescence persisted for 4 years. At a s maller scale, there was little difference in survivorship between study plo ts within sites or between quadrats within study plots, while in a few inst ances there were large differences in the number of reproductive individual s. Weekly cohorts that germinated early within seasonal cohorts had greater number of reproductive individuals than later weekly cohorts, but not cons istently greater survivorship. Several droughts induced temporal variation that was as important as spatial variation. During these droughts, the popu lation size of all cohorts that germinated prior to the droughts declined t o zero at all sites, and biennials were the longest lived type of reproduct ive individuals. During drought, Deevey Type I survivorship was prevalent, and lack of seed rain led to dependence on persistent seed banks for recrui tment in some seasonal cohorts.