Variation in long-term threshold responses to habitat availability

Background/Question/Methods

The minimum amount of habitat needed in a landscape for a species to persist has been a widely recognized conservation target. Several studies have identified habitat thresholds below which the probability of persistence declines rapidly. However, most of these studies have aimed at one or a few focal species and were set in a particular region. Little is known of the extent to which thresholds vary spatially and between species, making it difficult to extrapolate and interpret previous results because characteristics specific to the study region (e.g. level of fragmentation) and species (e.g. level of specialization) are determining factors of species responses. Based on remote sensed land cover data and repeated state-wide breeding bird atlases of the states of Vermont and New York, we derived logistic and segmented regression models of long-term responses to habitat availability (in % cover) for 25 forest associated breeding bird species. We used an information-theoretic approach to select the best model to describe each species persistence response in both states. The segmented regression models corresponded to threshold responses and provided threshold estimates of minimum habitat cover required for persistence.

 Results/Conclusions

Out of all 25 species, less species showed threshold responses to habitat availability in Vermont (11 species) than New York (21 species). For those species showing threshold responses in both states, the average of the percentage of forest cover at which thresholds of persistence occurred was significantly lower in Vermont (42.04%, S.E.=8.32) than in New York (62.97%, S.E.=4.18). Persistence thresholds ranged from 34.25% (Pileated woodpecker) to 95.83% (Winter wren) in New York and from 17.08% (Nashville warbler) to 88.55% (Scarlet Tanager) in Vermont. Ranking species from low to high thresholds in both states showed that different species responded differently in both states (i.e. a shift in species ranks), indicating that differences in species responses were not merely systematic. This complexity highlights once more the difficulty of generalizing minimum habitat requirements across species and regions. Yet, our study also indicates how we can go about modeling species responses for many species and in many regions using large scale data sets in a fairly quick and systematic manner. This study is the first step in a long-term project that aims at investigation of the factors governing minimum habitat requirements and providing critical information to conservation management and policy.