Files in this item



application/pdf3243005.pdf (4MB)Restricted to U of Illinois
(no description provided)PDF


Title:Predicting Mobile Species Response to Landscape Structure: The Roles of Spatiotemporal Pattern and Individual Dispersal
Author(s):Stoddard, Steven T.
Doctoral Committee Chair(s):Weatherhead, Patrick J.
Department / Program:Biology
Degree Granting Institution:University of Illinois at Urbana-Champaign
Subject(s):Biology, Ecology
Abstract:How alterations of a landscape's structure, due to both natural (fire) and anthropogenic (management) sources, will differentially impact resident populations of animal species, is a critical question for the conservation management of multiple-use reserves harboring rare and endangered species. Here I examine this question with a general, spatially-explicit, stochastic, individual-based simulation of a continuous population coupled to static and dynamic fractal neutral landscape models generated by spectral synthesis. I assess three hypotheses: that populations will respond non-linearly to habitat loss and go extinct at a significant amount of remaining habitat (extinction threshold); that local, transient changes in the quality of conditions will influence overall population performance; and that differences in dispersal behavior influence species vulnerability. Existing theory predicting population response to landscape alteration characterizes the landscape as habitat or non-habitat (binary)---even though organisms likely perceive a continuum of suitability---and conceptualizes dispersal as a passive, random process---even though many species move actively in directed fashion. I show that predictions from binary landscapes are biased, inflating extinction risk and the importance of spatial structure (fragmentation). Moreover, I show that increasing the complexity of dispersal dramatically reduces predicted extinction risk and the importance of spatial structure. Together these results help explain ambiguous empirical support of extinction thresholds. Conditions within landscapes fluctuate on fine scales because of disturbances and successional processes, but whether populations will be affected is uncertain as theorists commonly expect mobile organisms to track conditions. However, I show that landscape dynamics depress population growth in proportion to their spatiotemporal autocorrelation and that extinction is probable when conditions become very spatially heterogeneous and unpredictable. With a generalized model of active dispersal, I show that the sensitivity of species depends on dispersal, and species showing more site fidelity are generally better off in dynamic landscapes. Dispersal behaviors, like sociality, perceptual abilities, and different dispersal cues all influence population success in dynamic landscapes. Overall, this research emphasizes the potential importance of spatial and temporal pattern within landscapes for population success, but that complexities inherent to real systems can significantly influence model predictions with important implications for management.
Issue Date:2006
Description:110 p.
Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 2006.
Other Identifier(s):(MiAaPQ)AAI3243005
Date Available in IDEALS:2015-09-25
Date Deposited:2006

This item appears in the following Collection(s)

Item Statistics