Plant Community Assembly

In graduate school in the 1970s, discussions of Hutchinson's multivariate niche (1957) and the role of competitive exclusion in structuring communities provided exciting new insights into patterns of community assembly. At the same time, however, I was immersed in the literature of plant community ecology and building my own impressions of tropical and temperate forests. For me, as for many others, it was difficult to fit plant communities into a world view of exclusive niches. While predictions of competition and niche models seemed consistent with many observations of animal interactions and patterns of abundance, interspecific interactions among plants were more subtle. The considerable overlap in requirements for a limited number of resources by tropical rain forest understory shrubs or by tall grass prairie forbs appeared to preclude any but rather diffuse competition among species. In the wet tropics, even high productivity and intense competition for resources did not seem to limit species diversity; rather there seemed no limit to the number of tree species that could coexist. New studies on patterns of tropical forest structure (e.g., Poore 1968), the role of cyclones and other disturbances in driving forest structure (e.g., Whitmore 1974) all contributed to a model of forest assembly in which chance and probability played a larger role in determining composition than did degree of niche overlap and strength of interspecific interactions. Two authors, in particular, focused my thinking around these issues: Orie Loucks's 1970 paper highlighted the malleability of plant community composition and the integral role of disturbance in the community environment. Windstorm and fire shaped the structure and composition of the forest as much as did edaphic, temperature, and moisture gradients. Margaret Brian Davis's papers (e.g., 1981, 1986, 1987) provided compelling support for a model of independent association of species articulated by H. A. Gleason (1939). From chronosequences of pollen in lake sediments, she described how tree ranges expanded independently following the retreat of the Pleistocene glaciers. Her pollen samples from different times B.P. reflected these range shifts in changing community composition. These studies indicated to me that interspecific competitive interactions leading to niche exclusivity as in animal models played a minor role in shaping communities in which tree life spans are long and environments are characterized by the high frequency of disturbance and a constantly changing climate. Our perspective that stability in the forest environment leads to an equilibrium in species composition is an artifact of the limited time in which we are able to observe change. Rather, the composition of a plant community reflects many different drivers including site history, ongoing changes in climate, the evolution and distribution of floras, dispersal and establishment capacities of species, interactions with commensal and competitive species, and the prevalence of intrinsic and extrinsic stressors such as fire, windstorm, pests, and pathogens. This view of plant community assembly is now more or less common among researchers and managers, although the search for assembly rules in plant communities continues to intrigue. Researchers find compelling questions to pursue in ecosystems subject to a variety of management and disturbance histories, as the primacy of research in habitats once thought pristine fades. Nonnative species form a growing component of regional floras, and resulting novel communities are an important part of modern landscapes. As we are challenged to manage all these ecosystems under rapidly changing climate and disturbance regimes, a probabilistic rather than deterministic vision of plant communities may prove most useful.