Key points are not available for this paper at this time.
Invasive wild pigs Sus scrofa are globally recognised as one of the most damaging invasive species: they harm ecosystems, agriculture, cultural heritage and transmit significant diseases. However, in Australia effective management of wild pigs is hindered by outdated and non-quantitative estimates of broadscale wild pig distribution. This is partly due to fragmented and disparate data held by many stakeholders across jurisdictions and institutions. Here we produced a contemporary map of wild pig relative abundance for Australia using a continental-scale integrated species distribution model fit to 51,444 presence-only, 11,176 presence-absence, and 34,981 abundance-absence records from 30 data sources. We used five ecological covariates and a spatial random effect to estimate wild pig distribution across Australia, as well as two spatial layers to correct for search effort bias in presence-only observations. Our model showed wild pigs to be more widespread across northern, eastern and southwestern mainland Australia than previously appreciated, and largely absent from the arid interior. The full model had high predictive performance when tested against an independent presence-absence dataset (>0.85 AUC ROC ) everywhere except for northern Australia, which was also the only region in which predictive performance did not improve with the addition of spatial random effects. Our study demonstrates that integrated species distribution models can be successfully applied to non-equilibrium species over broad spatial scales using fragmented data sources. However, in addition to computational challenges, data integration at this scale requires complex stakeholder coordination across different jurisdictions and institutions to ensure fair and accurate distribution modelling.
Rees et al. (Tue,) studied this question.