A quantitative approach to ranking management actions for an endangered pool-breeding amphibian

Management of vulnerable amphibian populations requires a better understanding of the habitat factors that will make the greatest difference in their preservation. We set out to develop a predictive model of amphibian abundance based on habitat characteristics that may influence their survival and persistence. Our study system was the Sonoma County California tiger salamander (Ambystoma californiense; SCTS), an amphibian threatened by habitat loss and fragmentation. In 2019, 298 ponds (almost every breeding pond in this federally endangered Distinct Population Segment) were surveyed to determine SCTS larval densities, a proxy for population well-being. We used model selection to identify which factors of pool morphology, the surrounding pool neighborhood, and terrestrial upland habitat best predict variation in larval densities. Our best model showed a significant positive relationship between larval SCTS densities and three variables: pool depth, number of breeding pools within 500 m of a focal pool, and amount of suitable and accessible upland habitat within 1500 m of a pool. To further refine this model, we then examined interactions and non-linearities using a generalized additive model. Based on this refined model, we simulated a variety of management scenarios to identify which management practices would most increase larval densities across the endangered Distinct Population Segment and where on the landscape employing each mitigation action would yield the greatest return. This provides the basis for a county-wide Habitat Conservation Plan for SCTS, and highlights habitat characteristics likely critical to the conservation of other pond-breeding amphibians.