PARENT SESSION
Symposium #20: Extreme event analysis in ecology.
Sponsored by ESA Statistical Ecology and Long Term Studies Sections
Organized by: B. Li and R. G. Balice.
Thursday, August 9, 2001. 8:00 AM to 12:10 PM. Madison Ballroom A
The role of extreme dispersal in plant migration: estimation and prediction.
McLachlan, Jason1, Clark, James1, Manos, Paul1, 1
ABSTRACT- Based on the fossil pollen record, trees in eastern North America appear to have migrated northwards at over 100 m yr-1 in the millennia following the last deglaciation. Such rapid migration rates are found across species, despite large differences in the life history attributes that might affect dispersal biology, such as net reproductive rate (R0) and seed dispersal mechanism. Long-distance seed dispersal can allow such rapid rates of spread, but parametric models using fitted fat-tailed dispersal kernels predict unrealistic continually increasing invasion speeds. We used a new model to estimate rates of spread in terms of the expected location of the furthest dispersing individual, rather than the usual expected density of individuals. We predict rates of spread that are faster than those estimated by traditional Gaussian kernels, but slower than those estimated by fat-tailed parametric models, and which are highly sensitive to R0. Migration rates based on extreme dispersal events were more consistent with maps of chloroplast DNA haplotypes we developed for red maple and American beech than established pollen-based migration maps. Our molecular data suggests that red maple, a fecund species with wind-dispersed seed, spread rapidly from ice age populations ca. 35o N, while animal-dispersed beech migrated more slowly from full-glacial refugia closer to the ice than inferred from pollen data.
KEY WORDS: migration, dispersal, chloroplast DNA, paleoecology