Experiments on Mechanisms of Tree Establishment in Old‐Field Succession: Seedling Survival and Growth

Abstract

Mechanisms influencing the relative establishment success of early successional tree species in North Carolina Piedmont old fields were investigated in multi—factor field experiments. Six trees species (Acer rubrum, Fraxinus americana, Liquidambar styraciflua, Liriodendron tulipifera, Pinus taeda, and Ulmus alata) were sown into old fields at equivalent densities. In this paper I evaluate relative seedling performance (survival and growth) with respect to two principal factors: (1) competition from old—field vegetation (weeded vs. vegetated plots) and (2) browsing by vertebrate herbivores (exclosures vs. open plots). Manipulation of the first factor tests Connell and Slatyer's (1977) alternative "models" of succession. Seedlings were monitored over a 3—yr period in each of two experiments, one initiated in a wet spring (1984) and the other in a dry spring (1985). Most seedling mortality occurred in the first growing season. In the 1984 experiment, old—field vegetation reduced survival of all species; however, competitive reductions were not large except in Liriodendron, which appeared unable to survive in herbaceous cover, and in Acer, which survived poorly in the absence of competition as well. In the other four species, survival in old—field vegetation was similar; after high mortality in the first year, seedlings exhibited continued slow mortality with time. The dry spring of 1985 resulted in greater overall mortality in all species and greater variation in species responses to vegetation cover. In contrast to survival, height growth was strongly reduced by herbaceous vegetation in all hardwood species, but not in Pinus, which grew fairly well in competition with the herbaceous community. Characterization of these responses as tolerance vs. inhibition is complex and requires consideration of long—term demographic behavior. Three species (Fraxinus, Liquidambar, and Ulmus) were damaged by vertebrate browsers, but browsing intensity varied according to site and was most severe in Ulmus, with the result that its survival and growth were selectively reduced. The expected relative prevalence of these hardwood species, based on their emergence success and seedling performance in the experiments, does not fully correspond to their apparent relative occurrence in young Piedmont hardwood communities. A significant factor that may account for the discrepancies is differential seed rain, which can numerically offset low rates of seedling survival. Patterns of seed production by the hardwood species appear consistent with the hypothesis.