ESR prepress abstract   -  DOI: https://doi.org/10.3354/esr01391

ABSTRACT: Reintroduction as a wildlife management strategy is becoming increasingly more common and has had varied levels of success. To ensure a successful reintroduction it is important to consider historical survey data, habitat preferences and species distributions in reintroduction strategies. This work presents a case study of investigations into Eastern cricket frog reintroduction site selection in New York state. Eastern cricket frogs were listed as threatened in New York State in 1983 and were subsequently listed as a state endangered species in 1999 after further declines were documented. Despite research and continued monitoring, Eastern cricket frog decline within New York remains enigmatic and little is known about the ecology and habitat use at the state level. We sought to determine abiotic and biotic drivers of Eastern cricket frog distributions in New York by fitting a generalized boosted model to identify important response variables from a suite of potential drivers, including temperature, precipitation, fragmentation, wetlands, elevation, soil characteristics and forested habitat. We used a k-fold cross-validation resampling technique to estimate the optimal number of regression trees needed for prediction. We found that the presence of more than 68 km² of wetlands, over 40 mm of snowpack, and winter soil temperatures that were above freezing were important predictors of Eastern cricket frog occupancy in New York State. We mapped Eastern cricket frog distribution and created partial dependency plots showing spatial patterns of occupancy probability relating to our response variables. Our predictions identify potential reintroduction areas to meet management objectives at the state level.