Matrix Model Investigation of Invasive Species Control: Bullfrogs on Vancouver Island

Invasive species control is now a conservation priority in many parts of the world. Demographic modeling using population matrix models is a useful tool in the design of these control efforts as it identifies the life stages with the strongest influence on population dynamics. As a case in point, American bullfrogs (Rana catesbeiana) have been introduced around the world and have negative effects on native fauna. We studied demography of four populations on southern Vancouver Island, Canada, using field observations and capture-mark-recapture methods to estimate survival, growth, and fecundity. The life cycle of these introduced bullfrogs progressed in yearly increments through the following stages: eggs/small tadpoles, first-year tadpoles, second-year tadpoles, metamorphs/juvenues, and adults. Some bullfrog tadpoles were able to skip the second-year tadpole stage and metamorphosed one year after hatching. With tadpole survival estimates from the literature and field estimates of the remaining parameters we constructed a matrix population model. Prospective demographic perturbation analysis showed that bullfrog population growth rate (λ) was most influenced by the proportion of tadpoles metamorphosing early (tadpole development rate), and by early postmetamorphic survival rates. Most current control efforts for bullfrogs have focused on removing tadpoles and breeding adults, and our modeling suggests that these efforts may not be optimal. Partial removal of tadpoles may lead to higher tadpole survival and development rates and higher postmetamorphic survival due to decreased density-dependent competition. Removal of adults leads to higher survival of early metamorphic stages through reduced cannibalism. Our modeling suggests that culling of metamorphs in fall is the most effective method of decreasing bullfrog population growth rate. Our study shows how demographic information can be used to maximize the efficacy of control efforts, and our results are likely directly applicable to other invasive species with complex life cycles.