There is a desire among land managers to use data collected as part of invasive species management to inform management efforts. For example, the Argentine black and white tegu (Salvator merianae), an invasive species in the Greater Everglades Ecosystem in Florida, is managed by multiple government agencies through removal trapping programs. Removal models can estimate abundance and removal probabilities from removal data, but they often rely on a set of problematic assumptions. While recent work has addressed some of these limitations, developing removal models with population dynamics, temporary emigration/availability, and age structure remains challenging.
Motivated by these challenges, we developed two removal models with population dynamics and temporary emigration using open-robust design to analyze multi-year tegu removal data sets in three regions. The first model is based on a random walk approach and is estimable without ancillary information. The second is a 2-age class integrated population model (IPM), which uses prior information on hatchling availability and age structured vital rates derived from expert elicitation. In each region, we estimate tegu superpopulation abundance over time, removal probabilities given trapping effort, and dynamics (e.g. birth and survival rates). Finally, we project no-removal scenarios in each location to evaluate past removal trapping programs.
Results/Conclusions
The random walk and IPM models indicated high agreement for parameters held in common, and both models fit the data well based on posterior predictive checks. The IPM yielded more precise predictions of abundance, partitioned dynamics into births, deaths, and net migration, and resolved uncertainty in growth rates. The IPM additionally estimated the total superpopulation abundances and annual capture rates corrected for bias from a cryptic hatchling stage.
Estimates of tegu superpopulation abundance over time suggest that populations are increasing in two management regions despite generally high removal rates. However, tegu abundance appears to have stabilized or may even be declining in the core region, where trapping densities is highest. This suggests it may be possible to control tegu populations, at least locally, with similar trap densities. Finally, despite positive trends in two locations, comparing estimated trends to projections from no-removal scenarios suggests significant effects of management in all locations. Importantly, we provided the first estimates of tegu abundance, capture probabilities given trapping effort, and population dynamics in these regions, which is critical for effective population management. Furthermore, our models are easily generalized to other species and are critically needed to inform invasive removal programs.