Date of Award

Spring 5-15-2017

Author's School

Graduate School of Arts and Sciences

Author's Department

Biology & Biomedical Sciences (Evolution, Ecology & Population Biology)

Degree Name

Doctor of Philosophy (PhD)

Degree Type



A major goal of community ecology is in understanding variation in community composition, generally termed β-diversity. This variation can result from a variety of mechanisms, including deterministic factors, wherein species sort along biotic or abiotic gradients; stochastic processes, whereby random fluctuations in population sizes cause variation in community composition; and/or dispersal limitation. Although all of these processes are likely occurring in all biological communities, a key question in community ecology research is if their relative importance may vary systematically across environmental or biogeographic gradients. In this dissertation, we combine both observational and experimental research to investigate β-diversity across a biogeographic gradient of longstanding interest in ecology and evolutionary biology, the latitudinal gradient. Diversity at the local and/or regional scale has long been known to decrease with latitude, but only relatively recently have similar trends been shown for β-diversity as well. Although this may suggest that community assembly processes that generate β-diversity may also be varying with latitude, β-diversity metrics are numerically dependent to varying degrees on different aspects of regional and local diversity. Therefore, any trends in β-diversity with latitude could simply be reflecting the well-documented trends in local and/or regional diversity, generally referred to as sampling effects. Throughout this dissertation, therefore, we employ a relatively uncommon β-diversity metric, heretofore termed β-pie, that is relatively insensitive to sampling effort (the number of individuals sampled locally) and to the shape of the regional species abundance distribution, which we believe will improve the assessment of how and why community composition may vary in space and time. In Chapter 2, we apply this metric to zooplankton communities sampled across ten latitudes in North America and three years to determine if, after accounting for the aforementioned sampling effects, there are any general trends of spatial and/or temporal turnover with latitude. Although we recovered a significant relationship between spatial β-pie and latitude in two years, these trends actually reversed from one year to the next, and there was no significant relationship in the third year. Unlike other studies documenting temporal turnover as a function of latitude, we found no relationship between temporal β-pie and latitude. These results together suggest that systematic variation in β-diversity along local and/or regional diversity gradients (such as with latitude) may simply be reflecting numerical sampling effects instead of systematic variation in community assembly processes. Chapters 3 and 4 report the results of large-scale outdoor mesocosm experiments replicated at three latitudes in North America. By using mesocosms, we attempted to limit abiotic heterogeneity and historical differences, but allow for natural variation in regional species pools to affect community assembly. Chapter 3 specifically focuses on the role of an environmental filter, drought, and asks how it affects withinsite aggregation, as well as whether its effect on β-pie varies consistently with latitude. Interestingly, we found that β-pie could either increase or decrease after the drought treatment, and although we did find regional differences in its effect, these did not vary systematically with latitude. In addition, it appears that variation in β-pie was not due to changes in local diversity (α-pie) but largely caused by changes in the regional species abundance distribution (γ-pie). Chapter 4 focused on how dispersal at different stages of assembly affects β-pie. Because we did not intentionally impose abiotic heterogeneity, this experiment focused on the interaction between dispersal, stochasticity, and species interactions in generating intraspecific aggregation during community assembly. The two dispersal treatments occurred at different stages of assembly – 1) during the initial establishment of communities, when population sizes are relatively small and demographic stochasticity might generate high variation in initial colonists, and 2) two years after communities have assembled, when population sizes are much larger and species have a greater potential to deterministically interact. Like the drought treatment, we found variable effects of dispersal on β-pie. The early dispersal treatments (high versus low) were found to increase, decrease, or have no effect on aggregation, and there was no general trend with latitude. The late dispersal treatment effects did show some interesting trends for passive dispersers, however, wherein the high dispersal treatment actually increased β-pie relative to the controls. In addition, this effect tended to decrease with latitude, suggesting that perhaps dispersal limitation plays a greater role in community assembly with decreasing latitude.


English (en)

Chair and Committee

Alan Templeton & Tiffany Knight

Committee Members

Matthew Leibold, Scott Mangan, Jonathan Myers


Permanent URL: