Ecology, 96(3), 2015, pp. 705–715 Ó 2015 by the Ecological Society of America Species associations structured by environment and land-use history promote beta-diversity in a temperate forest STEPHEN J. MURPHY, 1,6 LI ´ VIA D. AUDINO, 2 JAMES WHITACRE, 3 JENALLE L. ECK, 1 JOHN W. WENZEL, 3 SIMON A. QUEENBOROUGH, 4 AND LIZA S. COMITA 4,5 1 Department of Evolution, Ecology, and Organismal Biology, Ohio State University, 318 West 12th Avenue, Columbus, Ohio 43210-1293 USA 2 Departamento de Entomologia, Universidade Federal de Lavras, Lavras, Minas Gerais 37200 000 Brazil 3 Powdermill Nature Reserve, Carnegie Museum of Natural History, 1847 Route 381, Rector, Pennsylvania 15677 USA 4 School of Forestry and Environmental Studies, Yale University, 195 Prospect Street, New Haven, Connecticut 06511 USA 5 Smithsonian Tropical Research Institute, Box 0843-03092, Balboa, Anco ´n, Panama Abstract. Patterns of diversity and community composition in forests are controlled by a combination of environmental factors, historical events, and stochastic or neutral mechanisms. Each of these processes has been linked to forest community assembly, but their combined contributions to alpha and beta-diversity in forests has not been well explored. Here we use variance partitioning to analyze ;40 000 individual trees of 49 species, collected within 137 ha of sampling area spread across a 900-ha temperate deciduous forest reserve in Pennsylvania to ask (1) To what extent is site-to-site variation in species richness and community composition of a temperate forest explained by measured environmental gradients and by spatial descriptors (used here to estimate dispersal-assembly or unmeasured, spatially structured processes)? (2) How does the incorporation of land-use history information increase the importance attributed to deterministic community assembly? and (3) How do the distributions and abundances of individual species within the community correlate with these factors? Environmental variables (i.e., topography, soils, and distance to stream), spatial descriptors (i.e., spatial eigenvectors derived from Cartesian coordinates), and land-use history variables (i.e., land-use type and intensity, forest age, and distance to road), explained about half of the variation in both species richness and community composition. Spatial descriptors explained the most variation, followed by measured environmental variables and then by land- use history. Individual species revealed variable responses to each of these sets of predictor variables. Several species were associated with stream habitats, and others were strictly delimited across opposing north- and south-facing slopes. Several species were also associated with areas that experienced recent (i.e., ,100 years) human land-use impacts. These results indicate that deterministic factors, including environmental and land-use history variables, are important drivers of community response. The large amount of ‘‘unexplained’’ variation seen here (about 50%) is commonly observed in other such studies attempting to explain distribution and abundance patterns of plant communities. Determining whether such large fractions of unaccounted for variation are caused by a lack of sufficient data, or are an indication of stochastic features of forest communities globally, will remain an important challenge for ecologists in the future. Key words: community assembly; environmental filtering; forest diversity; habitat association; multivariate regression tree analysis; spatial dependency; spatial eigenvector; variance partitioning. INTRODUCTION Determining the underlying mechanisms that govern patterns of species diversity and community composition continues to pose a major challenge to ecologists, despite the fact that this theme has a rich history in ecological research (Whittaker 1956, Bray and Curtis 1957, Whit- taker 1960, Baldeck et al. 2013a). While such studies have made clear that topographic and edaphic gradients can drive spatial variation in species richness and turnover in community composition in plants, their relatively low explanatory power and lack of consistency among communities makes a cohesive framework for community assembly elusive (Lawton 1999, Hubbell 2001). This lack of generality is highlighted, in part, by the popular niche vs. neutral debate (i.e., deterministic vs. stochastic assembly) spurred by Hubbell’s unified neutral theory (Hubbell 2001, Adler et al. 2007, Clark 2012), as well as by the notable dearth of predictive models in community ecology (Lawton 1999, Hubbell 2001, Simberloff 2004). Researchers have, however, begun to make substan- tial progress in understanding spatial patterns of Manuscript received 12 April 2014; revised 30 July 2014; accepted 5 August 2014; final version received 26 August 2014. Corresponding Editor: T. P. Young. 6 E-mail: murphy.1132@osu.edu 705