Using Occupancy Patterns to Infer Landscape Connectivity for Breeding Anurans

Landscape connectivity is critical to maintaining viable populations within patches of suitable habitat and is a function of the surrounding matrix’s resistance to an organism’s movement. As land use within the matrix varies, it can be expected that landscape connectivity will vary and as connectivity decreases, the probability of habitat occupancy by a species will decrease as well. Recognizing the difficulty in directly measuring landscape connectivity, my study attempts to model the relationship of Anuran, frog and toad, distributions to anthropogenic land use in the landscape matrix as a proxy for measuring connectivity. I use calling survey data of breeding Anurans to model occupancy probabilities in potential breeding ponds in eastern Virginia. Occupancy modeling is a likelihood-based method to estimate occupancy probabilities when the probability of detecting a species is less than one. Anthropogenic land use in the matrix around breeding sites is calculated in a Geographic Information System using a land cover map. The output of this analysis will ultimately be used to develop a spatially explicit landscape connectivity model for Anuran species in eastern Virginia. This connectivity model will enable land managers to evaluate impacts of proposed land uses on landscape connectivity and Anuran population viability. Introduction The biosphere is experiencing a rate of extinction never before witnessed in the history of life on Earth. The current rate of extinction is 10,000 times greater than the geologic background rate. 1 The majority of modern day species endangerment is attributed to human disturbance and anthropogenic habitat loss is the primary cause of global biodiversity declines. 2 In the United States urbanization is the leading cause of habitat loss. 3 Research in landscape ecology has shown that species persistence relates positively to habitat area. 4 Habitat loss not only reduces area, but fragments remaining habitat into smaller patches. Understanding the effects of fragmentation requires the evaluation of all habitat patches in the context of the surrounding landscape. 4 Fragmentation can isolate patches and thereby reduce the overall landscape connectivity of available habitat to a species. Landscape connectivity is defined as “the degree to which the landscape facilitates or impedes movement among resource patches.” 5 This definition indicates that connectivity is affected by landscape configuration and composition and is critical to maintaining viable populations. Isolated populations are vulnerable to environmental, demographic, and genetic problems, all of which can synergistically increase the probability of extinction. 6 Even the largest patches of protected habitats, if isolated, are too small for many species to persist. 7 Modern day conservation strategies emphasize preserving core habitats as reserve networks embedded in an increasingly fragmented landscape. These strategies are based on the patchmatrix landscape model, which defines patches as the habitat of interest and the matrix as everything else in the landscape or non-habitat. 8 However, this is a very simplistic view and there is an emerging argument in conservation biology that the matrix does matter and is essential to species conservation. 9 The matrix is not a homogenous landscape feature of unsuitable habitat as the patch-matrix model assumes, but is instead a complex mosaic of land cover types. 10 As landscape composition in the matrix varies, it can be assumed that landscape connectivity will vary as well. 11 Some land covers provide a high degree of connectivity where as others impede or act as complete barriers to species movement. Studies have demonstrated that many anthropogenic land covers (e.g. high urbanization) and features (e.g. high traffic roads) negatively affect landscape connectivity. 12,13 Answering questions regarding the effects of landscape connectivity on populations is inherently difficult. Connectivity studies need to be conducted at an extent measuring hundreds of kilometers in order to capture sufficient variation