Using Periodic Matrix Models to Simulate the Effectiveness of Alternative Reintroduction Strategies for Lizards on a Seasonal Tropical Island

ABSTRACT Conservation translocations and reintroductions are widely used to improve conservation outcomes for declining species. Reintroductions are unlikely to be successful if the threats that led to the extirpation of the focal species, such as non‐native predators, have not been ameliorated. The non‐native brown treesnake ( Boiga irregularis ) was introduced to Guam in the mid‐20th century and has devastated the native lizard fauna of this Pacific Island. Native lizard populations persist on small islands near Guam and could act as sources for reintroductions. Recently, trapping methods have been shown to greatly reduce the density of brown treesnakes, raising the possibility that native lizards could be reintroduced where brown treesnake abundance is controlled. Here we developed demographic models for four lizard species present on Guam to assess their validity as a trial for reintroducing species to meet native species recovery goals. These four species vary in their size, activity, and susceptibility to predation by brown treesnakes. We used periodic matrix models to simulate changes in lizard vital rates driven by the wet‐dry seasonal cycle found on Guam. We found that the release of 60 individuals of each species was likely to result in successful reintroduction outcomes provided adult survival in the wild is similar to rates estimated based on life history parameters. Demographic models can be used to forecast reintroduction outcomes, including predicting the probability of success and evaluating causes of failure in the event that populations do not establish.