CRITICAL THRESHOLDS ASSOCIATED WITH HABITAT LOSS FOR TWO VERNAL POOL‐BREEDING AMPHIBIANS

Abstract

A critical threshold exists when the relationship between the amount of suitable habitat and population density or probability of occurrence exhibits a sudden, disproportionate decline as habitat is lost. Critical thresholds are predicted by a variety of modeling approaches, but empirical support has been limited or lacking. We looked for critical thresholds in two pool‐breeding amphibians that spend most of the year in adjacent upland forest: the spotted salamander (Ambystoma maculatum) and the wood frog (Rana sylvatica). These species were selected because of their reported poor dispersal capacities and their dependency on forest habitat when not breeding. Using piecewise regression and binomial change‐point tests, we looked for a relationship between the probability of occupancy of a site and forest cover at five spatial scales, measuring forest cover in radial distances from the pond edge of suitable breeding ponds: 30 m, 100 m, 300 m, 500 m, and 1000 m. Using piecewise regression, we identified significant thresholds for spotted salamanders at the 100‐m and 300‐m spatial scale, and for wood frogs at the 300‐m scale. However, binomial change‐point tests identified thresholds at all spatial scales for both species, with the location of the threshold (percent habitat cover required) increasing with spatial scale for spotted salamanders and decreasing with spatial scale for wood frogs. Thresholds for spotted salamanders occurred at ∼30% forest cover at spatial scales of 100 m or less, with 41% cover at 500 m, and with 51% habitat cover at 1000 m. Thresholds for wood frogs ranged from 88% habitat cover at 30 m from the pond edge, declining to 44% habitat cover within 1000 m. These patterns might be explained by the different winter dispersions of these species. Knowing whether a species has a critical threshold, and at what level of cover and at what spatial scale it exists, would be essential for conservation of habitat‐sensitive species.