Context. Feral pigs (Sus scrofa) are highly fecund, and populations can increase rapidly under favourable conditions. Population size can also fluctuate widely, driven largely by changes in juvenile mortality in response to food availability, but these relationships have only been explored on a limited number of sites and over short periods.
Aims. The present study aimed to investigate and quantify the numerical response of feral pig populations to changes in their food supply in north-eastern Australia.
Methods. Pig population densities were determined from aerial surveys conducted over a 21-year period on 10 regional blocks (∼2000–6000 km2) throughout the Queensland rangelands. Densities were used to calculate annual exponential rates of increase (r), which were then corrected for anthropogenic mortality (baiting and commercial harvesting). Six proxy measures of annual food supply, including rainfall, pasture biomass and pasture growth (using the AussieGRASS model), were calculated for each survey block, and assessed as predictors of corrected r. The rates of increase predicted from the first half of the data series were then applied to initial population densities to estimate successive pig densities during the second period in each bioregion.
Key results. The most parsimonious model of the numerical response had parameters common to three bioregions, with rainfall in the 12 months between surveys being the best predictor variable. Modelled densities for each bioregion were a good fit to actual, observed densities. Relationships between r and each measure of food supply at the individual block level were inconsistent.
Conclusions. Using rainfall as a measure of food supply, the numerical response relationship provides a method for predicting the dynamics of feral pig populations at the bioregional scale. Predicting population dynamics at any one site using this relationship is less precise, suggesting that differences in landscape composition affect utilisation of resources supporting population growth.
Implications. The results from the present study could be used to predict feral pig population changes at the bioregional level, supplementing or reducing the need for more frequent, expensive population surveys. This improved ability to predict fluctuations in regional feral pig populations can help guide future management actions.