Intraspecific Variation in Sapling Mortality and Growth Predicts Geographic Variation in Forest Composition

With a view toward understanding variation in species composition among different forest communities, I examined species—specific growth and mortality of juvenile trees (2.3—78 mm diameter at 10 cm above the ground) at three contrasting sites. Two sites differing in soil mineralogy and elevation (schist/gneiss uplands vs. calcareous bedrock valley) were situated in northwestern Connecticut, USA. To examine variation over a more extensive geographic scale, I located the third site in central—western Michigan, USA. Among the three sites, the deciduous species (American beech, white ash, and sugar maple) showed little intraspecific variation in models of relative radial growth rate as a function of light availability. Strikingly, a substantial component of the variation in radial growth at the Michigan site could be explained by sapling growth models originally calibrated for the Connecticut sites. In contrast to the deciduous species, the evergreen species (white pine and eastern hemlock) exhibited significant intraspecific differences in growth models between the two Connecticut sites. Intraspecific variation in growth models among sites was characterized by significantly different estimates of growth at low light. This result suggests that light and other resources can be simultaneously limiting, and challenges the application of Liebig's law of the minimum to tree sapling growth. For the deciduous species, mortality processes exhibited more variation among the sites than did growth. I found significant site differences in mortality as a function of recent growth for both sugar maple and white ash on the calcareous soils in comparison to the schist/gneiss soils in Connecticut. In general, the mortality functions at the Michigan and the upland Connecticut site were similar, and both differed from the Connecticut calcareous site. Community dynamics at each of the three sites were simulated with a model of forest dynamics called SORTIE, which incorporated observed differences in sapling mortality and growth functions among the sites. Using SORTIE, among—site differences in only juvenile survivorship and growth were sufficient to predict dominant species in the adult canopy, demonstrating the critical importance of sapling stages to community dynamics. If a species performs differently at a particular site, its interactions with other species at the site will also change, potentially contributing to further differences in community species composition. Changes in species' performance in response to different site conditions and the species' subsequent interactions are critical to understanding differences among the dynamics and compositions of these forest communities.