A meta-population model to predict occurrence and recovery of wolves

Abstract

Wolves (Canis lupis) have been recolonizing Washington since 2008. In an effort to guide recovery and management decisions for wolves, we created a spatially explicit meta-population matrix model using vital rates based on empirical data from other states in the northwestern United States to estimate probability of occurrence, terminal extinction rates, and potential recovery time. We applied an existing habitat model for Idaho, Montana, and Wyoming to the Washington landscape to determine the extent of probable habitat. We then simulated an evenly distributed metapopulation based on average size of pack territories reported in central Idaho where average probability of occurrence exceeded 40%. Using the program RAMAS GIS, we created a female-only, stage matrix model with dispersal using population metrics from central Idaho and northwestern Montana. Model simulations that begin in 2009 suggest Washington should reach its recovery goals in approximately 12 years (2021). We used the model to project recovery timeframes and the risk of declining below recovery objectives if management scenarios are considered during recovery. This model is also intended to be a versatile and adaptive tool for managers to project potential carrying capacity and the minimum viable population in the future when locally derived empirical data become available as wolves recolonize Washington. The model framework can be easily adapted to guide management decisions of wolves in other states (Idaho, Montana, Oregon, Wyoming) or countries and it can also provide a way to identify recovery thresholds (quasi-extinction) in other areas considered for recovery where no data are currently available.