Spatial scale changes the drivers of beta-diversity along land-use, environmental and geographic gradients

Abstract

Understanding the patterns of spatial change in community composition (betadiversity), and the processes that structure biological communities are central themes in ecology. While the impacts of habitat change on beta-diversity are well-studied, most studies have been restricted to a single spatial scale. As a result, the effects of changing spatial scale on beta-diversity patterns across both natural and human-modified habitats are little known. My thesis targets the effect of spatial scale on tree and bird beta-diversity in the Asia-Pacific region across gradients of latitude, elevation and land-use change. The First Chapter uses tree data (> one billion stems and > 2500 species) from 15 long-term ForestGEO plots across the Asia-Pacific region to show how spatial scale changes the relationship between beta-diversity and latitude. At small spatial scales, beta-diversity decreased with increasing latitude; but at large spatial scales, beta-diversity did not change with increasing latitude. Different relationships across spatial scales were caused by differences in species richness, which influenced β-diversity values at small spatial scales, but not at large spatial scales. The Second Chapter uses bird data from Sri Lanka to show how horizontal (geographic) and vertical (elevation) distances can influence bird beta-diversity within three different land-use types (protected rainforests, reserve buffers and intensive agriculture). I show that bird beta-diversity within all land-use types were similar across horizontal distances. However, bird betadiversity within land-use types were not similar across vertical distances; protected rainforests had higher beta-diversity than the other two habitats. The Third Chapter uses bird data from the Western Ghats – Sri Lanka biodiversity hotspot to determine the drivers of bird community assembly at three different spatial scales. The geographic barrier (the Palk Strait) is the most important driver of bird beta-diversity at large spatial scale. Land-use and environment were equally important at intermediate scales and land-use was the most important driver at small scales. In conclusion, this thesis demonstrates the importance of sampling at multiple spatial scales to better understand natural and human-influenced betadiversity. In the First Chapter I showed that spatial scale changes the relationship between beta-diversity and latitude; and improving sampling representativeness avoids the species richness dependence of beta-diversity. In the Second Chapter I demonstrated the crucial importance of conserving rainforests across the full elevation range available. In the Third Chapter I showed that considering community assembly processes at multiple spatial scales while selecting sites for biological conservation holds great promise for preventing further species loss.