Probabilistic genetic identification of wild boar hybridization to support control of invasive wild pigs (Sus scrofa)

The rapid expansion of wild pigs (Sus scrofa) throughout the United States (US) has been fueled by unlawful introductions, with invasive populations causing extensive crop losses, damaging native ecosystems, and serving as a reservoir for disease. Multiple states have passed laws prohibiting the possession or transport of wild pigs. However, genetic and phenotypic similarities between domestic pigs and invasive wild pigs – which overwhelmingly represent domestic pig-wild boar hybrids – pose a challenge for the enforcement of such regulations. We sought to exploit wild boar ancestry as a common attribute among the vast majority of invasive wild pigs as a means of genetically differentiating wild pigs from breeds of domestic pigs found within the US. We organized reference high-density single nucleotide polymorphism genotypes (1,039 samples from 33 domestic breeds and 382 samples from 16 wild boar populations) into five genetically cohesive reference groups: mixed-commercial breeds, Durocs, heritage breeds, primitive breeds, and wild boar. Building upon well-established genetic clustering approaches, we structured the test statistic to describe the difference in the likelihood of a given genotype’s ancestry vectors (sensu genetic clustering analysis) if derived strictly from the four described domestic pig reference groups versus allowing for admixture from the wild boar group. By fitting statistical distributions to test statistics of reference domestic pigs, we characterized the distribution of the null hypothesis – that a given genotype descends strictly from domestic pig reference groups. We tested the approach with simulated genotypes and empirical data from an additional 29 breeds of domestic pig represented by 435 unique genotypes; all associated test statistics for simulated and empirical domestic pig challenge sets fell within the distribution of reference domestic pigs. We then evaluated 6,566 invasive wild pigs sampled across the contiguous United States, of which 63% exceeded the maximum threshold for domestic pigs and could be statistically classified as possessing wild boar ancestry. This approach provides a scientific foundation to enforce regulations prohibiting the possession of this destructive invasive species. Further, this computationally efficient and generalizable approach could be readily adapted to quantify gene flow among ecological systems of conservation or management concern.

Methods
We assembled the Sus scrofa reference set from previously published high-resolution SNP genotypes, restricting analysis to genotypes produced with Illumina BeadArray technology (San Diego, California) across multiple commercially available arrays (Illumina PorcineSNP60, Illumina PorcineSNP60 v2, Genomic Profiler for Porcine HD, licensed exclusively to GeneSeek, a Neogen Corporation, Lansing, Michigan; Ramos et al., 2009). We augmented previously published genotypes (detailed in Smyser et al., 2020) with a subset of novel genotypes produced for this study (Appendix S1: Table S1). We restricted our analyses to loci that were available across all datasets (influenced by loci shared across arrays and the extent to which publicly available datasets were filtered by authors prior to publication) and mapped to autosomes (Sscrofa11.1 genome assembly; Warr et al., 2020). In sum, we included 33 breeds and 16 populations of European wild boar, representing a total of 1,421 reference samples (Table 1) genotyped at 28,545 biallelic loci.