Ecology of animal-mediated seed dispersal in the fragmented central hardwoods region
I studied factors affecting seed dispersal of nut-bearing deciduous trees, and examined how dispersal processes might be altered by forest fragmentation in the Midwestern U.S. The seeds of oaks (Quercus), hickories (Carya), and walnuts (Juglans) are dispersed by rodents and blue jays (Cyanocitta cristata). If forest fragmentation alters the rodent communities in habitat remnants, this could have consequences for seed dispersal and tree recruitment. My analysis of rodent-species occurrence rates throughout northern Indiana adds to evidence that eastern gray squirrels (Sciurus carolinensis) are sensitive to fragmentation. This species did not occur in small or isolated forest patches, and did not occur >15 km from large contiguous riparian forests. I conducted a 3-yr experiment to examine how seed survival would be affected in the absence of gray squirrels and fox squirrels (S. niger), 2 important seed dispersers. When Sciurus was excluded from seeds, seed survival was much lower than when these species could access seeds. However, seedling recruitment was not reduced in watersheds that lacked eastern gray squirrels. This is likely because fox squirrels, which were widespread and are tolerant of fragmentation, are able to disperse seeds in these areas. My experiments provided insight to the manner in which annual variation in seed production affects seed survival and dispersal distance. In years of low seed production, survival was extremely low for all seed types; in years of high seed production, survival was higher and varied with seed type. Rodents dispersed seeds to greater distances in seed-poor years than in seed-rich years, and dispersed more energy-rich food items to greater distances. These results provide support for escape and predator-satiation hypotheses, and challenge conventional optimal cache density models that suggest seeds should be dispersed shorter distances in low-seed years. My experiments also showed that seed removal rate data did not adequately index seed predation rates; this result has implications for conducting seed dispersal studies. Finally, I conducted feeding trials with blue jays to determine their dietary preferences for acorns of different oak species; this provided insight about jays' potential to act as long-distance dispersers of various oak species.
Swihart, Purdue University.
Off-Campus Purdue Users:
To access this dissertation, please log in to our