• McMcracken, Karen E.
  University of Maine
  

I experimentally examined spatial and temporal patterns of seed predation on three tree species in a temperate oak-pine forest in southern Maine, USA. Rodents were the principal seed predators. Rates of seed loss varied with tree species and exposure to three different suites of potential seed predators (all vertebrate seed predators, medium-sized rodents [primarily squirrels] and small rodents [primarily mice and voles]).   Acer rubrum seeds were removed more slowly (mean of 17.7% within 2 days across all habitats and treatments in 1991) than seeds of either Quercus rubra or pinus strobus (e.g. > 99% removed within 2 days across all habitats and treatments for each species in 1991). Levels of final removal (cumulative removal final census) varied with year but not with microhabitat ; i.e. seed predation in four types of forest gap was not significantly different than in intact forest. Both field data and experimental feeding trials with captive Peromyscus leucopus and Clethrionomys gapperi suggest that a significant  proportion of white pine seeds are eaten at time of detection, while red oak acorns are more likely to be cached.  Captive Peromyscus leucopus cached and ate significantly more red oak acorns than Clethrionomys gapperi.  These results and those of Chapter 2 suggest that red oak mast may be more important in overwintering success in  Peromyscus than in Clethrionomys. Nevertheless, both species did consume red oak acorns, suggesting that in southern Maine these species are not substantially inhibited by high tannin levels in red oak acorns, as has been suggested by researchers elsewhere.