Assessment of playa wetland network connectivity for amphibians of the south-central Great Plains (USA) using graph-theoretical, least-cost path, and landscape resistance modelling

Context: Land-cover changes are likely affecting structural and functional connectivity for wetland-associated wildlife among the > 80,000 playa of the Great Plains (USA). Objectives: We investigated how patterns of land cover from 2008 to 2016 affected structural and functional connectivity among playas under biologically feasible scenarios and a range of vagilities for amphibians. Methods: We compared graph-theoretical, least-cost path (LCP), and landscape resistance models. Using land-cover and irrigation data and literature-based estimates of species-specific dispersal, we developed three alternative scenarios with resistance values that differed by land-cover type. Sensitivity analysis examined effects of altering the numeric range of values within models. Results: From a structural perspective, the playa network was highly fragmented, with coalescence distances greater than amphibian dispersal capabilities (> 15 km). The network was insensitive to assigning different resistance values to different land-cover types: 81% of LCP locations overlapped with at least one other modeled scenario, and 57% of LCPs overlapped among all three scenarios. Conclusions: Our findings partially support previous graph-based assessments of playa structural connectivity that indicated high levels of path redundancy within this ecological network. However, although there was extensive path overlap among scenarios, many modelled linkages were at distances well beyond the individual lifetime dispersal maxima for regional amphibians. Thus, functional connectivity would be achieved in this landscape only by spanning scales from daily movements to inter-generational dispersal.