Food-Web Models Predict Species Abundancesin Response to Habitat ChangeNicholas J. Gotelli1*, Aaron M. Ellison21 Department of Biology, University of Vermont Burlington, Burlington, Vermont, United States of America, 2 Harvard Forest, Harvard University, Petersham, Massachusetts,United States of AmericaPlant and animal population sizes inevitably change following habitat loss, but the mechanisms underlying thesechanges are poorly understood. We experimentally altered habitat volume and eliminated top trophic levels of thefood web of invertebrates that inhabit rain-filled leaves of the carnivorous pitcher plant Sarracenia purpurea. Pathmodels that incorporated food-web structure better predicted population sizes of food-web constituents than didsimple keystone species models, models that included only autecological responses to habitat volume, or modelsincluding both food-web structure and habitat volume. These results provide the first experimental confirmation thattrophic structure can determine species abundances in the face of habitat loss.Citation: Gotelli NJ, Ellison AM (2006) Food-web models predict species abundances in response to habitat change. PLoS Biol 4(10): e324. DOI: 10.1371/journal.pbio.0040324IntroductionThe loss of natural habitat area often is accompanied bythe disappearance of large-bodied top predators and theupper trophic levels of food webs [1–3]. However, severalpieces of evidence suggest that habitat area alone may beinsufficient to predict changes in population size. Predictionsof ecological models [4,5], patterns of food-web structure insmall versus large habitat fragments [6], and recent observa-tions of collapsing island communities [1,7] all suggest thattrophic interactions must be considered in order to predicthow abundances of populations will change in the face ofhabitat loss and alteration. However, existing tests of the roleof trophic interactions in determining species abundances ashabitats contract are correlative only. In this study, weprovide the first evidence from a controlled field experimentfor the importance of trophic structure in controllingabundances of multiple species in an aquatic food web.Moreover, we demonstrate that models of trophic structureaccount for the results better than do simpler models thatfocus only on responses of individual species to changes inhabitat size or structure, or models that include both food-web structure and habitat volume.For multitrophic assemblages, two broad classes of com-munity models predict the potential responses of populationsto habitat change. (1) Single-factor models emphasize theunique responses of individual species to variation in habitatarea. This framework includes island biogeographic models[8], as well as single-species demographic analyses [9] andassessments of extinction risk. Single-factor models alsoinclude keystone species effects, which emphasize responsesof populations to changes in the abundance of a singlekeystone species, such as a top predator [10,11]. Thisframework includ es much current research on habitatalterations by foundation species [12] and ecosystem engi-neers [13]. (2) Food-web models emphasize the shifts inabundance that result from multiple trophic interactions andthe transfer of energy and biomass through a food web. Thisframework includes top-down and bottom-up processes [14],trophic cascades [15], and more complex interactions acrossmultiple trophic levels [16].Unfortunately, published studies of the effects of habitatcontraction have relied on conventional analyses that do notexplicitly compare these alternative frameworks [17,18].Although analysis of variance and other statistical protocolscan quantify community change, they cannot be used todistinguish between simple responses of species to habitatcontraction (single-factor volume model) and more complexresponses to changes in the abundance of other species (food-web and keystone species models). A third possibility is thattrophic responses dominate the responses, even in the face ofhabitat alterations. In this study, we used realistic fieldmanipulations of habitat volume and removal of top trophiclevels of entire aquatic communities. These manipulationsinduced major alterations in habitat size and communitystructure that have been studied previously in nonexper-imental settings [1]. For the first time, we have experimentallyassessed the relative importance of autecological responses,keystone species effects, and trophic interactions in account-ing for changes in species’ abundance.ResultsThe Aquatic Food Web of SarraceniaThe macroinvertebrate community associated with thenorthern pitcher plant Sarracenia purpurea (Figure 1) is amodel system for testing mechanisms controlling abundancein the face of habitat change [19]. S. purpurea is a long-livedperennial plant that grows in peat bogs and seepage swampsthroughout southern Canada and the eastern United States[20]. The plant grows as a rosette and produces a set of six to12 new tubular leaves each year. During the growing season,Academic Editor: Robert Holt, University of Florida, United States of AmericaReceived January 22, 2006; Accepted August 1, 2006; Published September 26,2006DOI: 10.1371/journal.pbio.0040324Copyright: Ó 2006 Gotelli and Ellison. This is an open-access article distributedunder the terms of the Creative Commons Attribution License, which permitsunrestricted use, distribution, and reproduction in any medium, provided theoriginal author and source are credited.* To whom correspondence should be addressed. E-mail: [email protected] Biology | www.plosbiology.org October 2006 | Volume 4 | Issue 10 | e3240001PLoSBIOLOGYleaves open approximately every 17 d and fill with rain water;an aquatic food web quickly develops in these water-filledleaves [21]. Leaves are photosynthetically most active in theirfirst year, but persist, capture prey, and are used as macro-invertebrate habitat for 1 to 2 y [22]. The base of the food webis captured arthropod prey (predominantly ants and flies),which is shredded and partially consumed by midge (Metri-ocnemus knabi) and sarcophagid fly (Fletcherimyia fletcheri) larvae[23]. Shredded prey are then processed by a subweb ofbacteria and protozoa [24], which respectively are prey tofilter-feeding rotifers (Habrotrocha rosi) and mites (Sarraceniopusgibsonii). Larvae of the pitcher plant mosquito Wyeomyia smithiifeed on bacteria, protozoa, and rotifers [25]. Large (thirdinstar) larvae of F.