Genetic Monitoring of a Lethal Control Programme for Wild Canids With Complex Mating Strategies.

Abstract

Although mammalian carnivores are ecologically important, they also drive human-wildlife conflicts. Managing carnivores using lethal control is controversial, in part because the impact of control effort is often uncertain due to limited abundance monitoring. We used an Australian metapopulation of wild dogs as a model system to investigate the feasibility of monitoring effective population size ( $N_e$ ) to detect reductions in census population size ( $N_c$ ) following control. Based on microsatellite data collected over an 11-year period, we parameterised an individual-based spatial population model for wild dogs that integrated demography, genetics, random or hierarchical mating, dispersal between subpopulations and compensatory immigration. $N_c$ and ${\hat{N}}_e$ trajectories were simulated under different proportional culling rates and genetic sampling regimes. We also used simulations without culling as null models to define 95% critical values for assessing the significance of empirical changes in ${\hat{N}}_e$ over time. We concluded there were significant reductions (39%-62%) in ${\hat{N}}_e$ in each subpopulation of the wild dog metapopulation, mostly likely due to control. In simulations assuming a hierarchical rather than random mating system, the impact of control on $N_c$ was weakened because reproduction by subordinate individuals increased as dominant individuals were removed, yet ${\hat{N}}_e$ reduced following culling. Sensitivity analyses demonstrated that ${\hat{N}}_e$ becomes an unreliable proxy of $N_c$ when compensatory immigration is strong and compensatory reproduction is weak, in which case ${\hat{N}}_e$ can increase following culling due to the immigration of new genotypes. Nonetheless, our results suggest ${\hat{N}}_e$ can provide information about wild dog $N_c$ over sufficiently short timescales to inform management.