Journal list menu

Volume 96, Issue 10 p. 2669-2678
Article

The influence of habitat fragmentation on multiple plant–animal interactions and plant reproduction

Lars A. Brudvig

Corresponding Author

Lars A. Brudvig

Michigan State University, Plant Biology Department, East Lansing, Michigan 48824 USA

E-mail: [email protected]Search for more papers by this author
Ellen I. Damschen

Ellen I. Damschen

Department of Zoology, University of Wisconsin, Madison, Wisconsin 53706 USA

Search for more papers by this author
Nick M. Haddad

Nick M. Haddad

Department of Biological Sciences, North Carolina State University, Raleigh, North Carolina 27695 USA

Search for more papers by this author
Douglas J. Levey

Douglas J. Levey

Department of Biology, University of Florida, Gainesville, Florida 32611 USA

National Science Foundation, Arlington, Virginia 22230 USA

Search for more papers by this author
Joshua J. Tewksbury

Joshua J. Tewksbury

Department of Biology, University of Washington, Seattle, Washington 98195 USA

Luc Hoffmann Institute, World Wide Fund for Nature, 1196 Gland, Switzerland

Search for more papers by this author
First published: 01 October 2015
Citations: 48

Corresponding Editor: G. A. Fox.

Abstract

Despite broad recognition that habitat loss represents the greatest threat to the world's biodiversity, a mechanistic understanding of how habitat loss and associated fragmentation affect ecological systems has proven remarkably challenging. The challenge stems from the multiple interdependent ways that landscapes change following fragmentation and the ensuing complex impacts on populations and communities of interacting species. We confronted these challenges by evaluating how fragmentation affects individual plants through interactions with animals, across five herbaceous species native to longleaf pine savannas. We created a replicated landscape experiment that provides controlled tests of three major fragmentation effects (patch isolation, patch shape [i.e., edge-to-area ratio], and distance to edge), established experimental founder populations of the five species to control for spatial distributions and densities of individual plants, and employed structural equation modeling to evaluate the effects of fragmentation on plant reproductive output and the degree to which these impacts are mediated through altered herbivory, pollination, or pre-dispersal seed predation. Across species, the most consistent response to fragmentation was a reduction in herbivory. Herbivory, however, had little impact on plant reproductive output, and thus we found little evidence for any resulting benefit to plants in fragments. In contrast, fragmentation rarely impacted pollination or pre-dispersal seed predation, but both of these interactions had strong and consistent impacts on plant reproductive output. As a result, our models robustly predicted plant reproductive output (r2 = 0.52–0.70), yet due to the weak effects of fragmentation on pollination and pre-dispersal seed predation, coupled with the weak effect of herbivory on plant reproduction, the effects of fragmentation on reproductive output were generally small in magnitude and inconsistent. This work provides mechanistic insight into landscape-scale variation in plant reproductive success, the relative importance of plant–animal interactions for structuring these dynamics, and the nuanced nature of how habitat fragmentation can affect populations and communities of interacting species.