An end to cane toad roulette: Using stochastic modeling to focus cane toad control efforts in Australia

Background/Question/Methods

Considered one of the world’s 100 most invasive species, the cane toad (Bufo marinus) has proven to be highly adaptable and difficult to extirpate. In Australia, invasive cane toads are particularly problematic with quantified negative effects on many native species. Deterministic population models have been used to examine potential impacts of different control strategies and have generally shown that removal of adults is likely most effective. However, these models have neglected to incorporate variability in vital rates despite available cane toad life history data from both native and invasive ranges. The effect that this variability may have on cane toad population models, on the life stage(s) that should be targeted for control, and on what management strategies should be pursued to control existing populations, is unknown. A stochastic stage-based population model with density dependence at the tadpole stage was constructed using life history data from cane toad populations in Australia in order to examine the relationship between life history parameters and two metrics of successful population control: mean adult population size and extinction risk. Because the model included both stochasticity and density dependence, we used standardized coefficients of linear and logistic regression analyses as estimates of the sensitivity of population size and extinction risk, respectively, to vital rates.

Results/Conclusions

The model yielded estimates of 200-300 adults per 100 m, which is consistent with observed population estimates from Australia and the conclusions of previous modeling efforts. Results from the linear regression analysis for mean adult population size showed that control methods targeting metamorph (<30 mm) cane toads would yield the strongest decrease in the mean number of adults. Juvenile survival, metamorph survival, adult survival, and maximum tadpole survival had the greatest effect on extinction risk, indicating that control methods that reduce survival of smaller cane toads (i.e., the use of tadpole alarm pheromones in combination with lungworm parasites and meat ants) will likely prove most effective. These results contrast those of previous studies, reflecting the importance of incorporating variability into studies of population dynamics.  However, uncertainty in our estimates of juvenile survival and the effects of proposed control methods warrant further examination. Future work will include the construction of a cane toad size-based population model to further investigate the population-level effect of control methods that alter cane toad body size.