Howe I propose to discover whether seed-eating birds and rodents and foliage-eating mammals influence plant diversity and community composition as agricultural land is converted to native tallgrass vegetation. Birds and mammals select seeds by size, and most herbivorous mammals select seeds by size, and most herbivorous mammals prefer to eat foliage of some plant species over others. However it is not known whether preferences evident in the lab actually alter composition and diversity of tallgrass prairie communities, either in conserved remnants or in synthetic assemblages created by restoration. Understanding such basic processes as seedbank depletion and intensive mammalian herbivore will aid in land reclamation, prairie restoration, and conservation of hundreds of native species which now occupy peripheral oldfield or woodland habitats, or increasingly isolated prairie remnants now (( 1% of pre-European extent on rich soils suitable for intensive agriculture. An exclosure experiment in southwestern Wisconsin will determine whether consumer choices by vertebrates regulate vegetative composition during establishment of replicated restorations of prairie grass and forbs. I will seed 45 representative species of fallow agricultural land that has been prepared for two years. Different levels of vertebrate access (no access, bird access, rodent access, total access including deer) should produce different species composition in replicated plots that initially receive the same number an composition of seeds. Cricetid and microtine rodents should depress large-seeded species (( 3.0 mg/seed), sparrows should depress species with seeds of intermediate size (0.3-2.0 mg/seed), while populations of plants with seeds to small to be of interest to vertebrates (( 0.2 mg/seed) should be most diverse and abundant when granviores are present. With all vertebrates absent, large-seeded species should be disproportionately abundant. Rodent and deer access after emergence should favor grasses because of selective herbivore on forbs. It is remarkable that foraging preferences considered axiomatic in foraging ecology of seed selection and herbivore are generally untested in tallgrass prairie communities, and therefore cannot be applied to conserving or restoring a rich flora. Recent experience has shown that when restoration is done as a science, it is very revealing. This study will augment scientific understanding needed by federal, state, and local governments, private interests, and environmental groups that show an increasing commitment to "restoration from scratch" of species assemblages that are vanishingly rare on rich agricultural soils.
