Global Ecology and Biogeography · 2016 · 281 citations · 40 references
Natural ResourcesEcologyLandscape EcologyHabitat FragmentationLandscape MetricsPatch DynamicsFragmentation MetricsHabitat Fragmentation EffectsLandscape FragmentationPatch FragmentationWildlife EcologyWildlife-habitat RelationshipsBiodiversity Responses To FragmentationFragmentation Effects On BiodiversityFragmentation Effects On DiversityHabitat Fragmentation ModelingLandscape Fragmentation EffectsSpecies Sensitivity AnalysisSpecies-habitat Relationships
Abstract Aim Elucidating patterns in species responses to habitat fragmentation is an important focus of ecology and conservation, but studies are often geographically restricted, taxonomically narrow or use indirect measures of species vulnerability. We investigated predictors of species presence after fragmentation using data from studies around the world that included all four terrestrial vertebrate classes, thus allowing direct inter‐taxonomic comparison. Location World‐wide. Methods We used generalized linear mixed‐effect models in an information theoretic framework to assess the factors that explained species presence in remnant habitat patches (3342 patches; 1559 species, mostly birds; and 65,695 records of patch‐specific presence–absence). We developed a novel metric of fragmentation sensitivity, defined as the maximum rate of change in probability of presence with changing patch size (‘Peak Change’), to distinguish between general rarity on the landscape and sensitivity to fragmentation per se. Results Size of remnant habitat patches was the most important driver of species presence. Across all classes, habitat specialists, carnivores and larger species had a lower probability of presence, and those effects were substantially modified by interactions. Sensitivity to fragmentation (measured by Peak Change) was influenced primarily by habitat type and specialization, but also by fecundity, life span and body mass. Reptiles were more sensitive than other classes. Grassland species had a lower probability of presence, though sample size was relatively small, but forest and shrubland species were more sensitive. Main conclusions Habitat relationships were more important than life‐history characteristics in predicting the effects of fragmentation. Habitat specialization increased sensitivity to fragmentation and interacted with class and habitat type; forest specialists and habitat‐specific reptiles were particularly sensitive to fragmentation. Our results suggest that when conservationists are faced with disturbances that could fragment habitat they should pay particular attention to specialists, particularly reptiles. Further, our results highlight that the probability of presence in fragmented landscapes and true sensitivity to fragmentation are predicted by different factors.
40
Fitting Linear Mixed-Effects Models Using <b>lme4</b>
Douglas M. Bates, Martin Mächler, Benjamin M. Bolker et al. · Journal of Statistical Software · 2015 · 85.6K citations · Full text
Kenneth P. Burnham, David R. Anderson · Sociological Methods & Research · 2004 · 11.6K citations
Phylogenies and the Comparative Method
Joseph Felsenstein · The American Naturalist · 1985 · 10.1K citations
The comparative method in evolutionary biology
Choice Reviews Online · 1992 · 5K citations