3 Results
Overall, our findings in two wild pig populations suggested that the landscape predictors (e.g., wetland, linear features, and food resources) played different roles in habitat selection and contact processes. The spatial overlap of individuals’ habitat selection does not adequately represent the spatial distribution of contacts across the landscape.
3.1 Female-female pairs
In the FL site, models for female wild pig pairs showed selection for wetland, ditches, and water and avoidance of fences and roads (Figure 2). The top-selected individual-RSF model (Model 1.1 in Table S1) showed wetland areas were the most strongly selected of the resources assessed. Their contacts tended to happen less at food resources, suggesting use of food resources was subject to spatial or temporal segregation. Contacts occurred more frequently along linear features (ditches and fences) and in pastures with higher cattle density relative to their selection of these landscape features (Figure 3).
In the TX site, female wild pig pairs selected for water and bait sites and avoided wetlands and roads (Model 1.1 in Table S2; Figure 2), which resulted in strong habitat selection for riverine areas and low selection along primary roads (Figure 4). In contrast, their contact locations occurred more at bait sites and less at wetland and water areas, relative to their selection for these features.
3.2 Female-male pairs
In the FL site, individuals involved in female-male contacts selected for areas with dense tree canopy, wetlands, and locations along ditches and roads, but avoided areas with cattle supplements (Model 2.1 in Table S1; Figure 2). Prediction of the individual-RSF model across the landscape showed a high probability of selection for these individuals in wetland areas. Similar to female-female contacts, female-male contacts tended to occur along linear features such as ditches and roads. However, these contacts tended not to be cattle supplements.
In the TX site, individuals involved in female-male contacts selected areas near wetlands, water, and with dense tree canopy and bait availability, but avoided primary and secondary roads (Model 2.1 in Table S2; Figure 2), which was reflected in the prediction map (Figure 3). Contact between female-male pairs was more likely to occur in areas with high NDVI, along trails, and at bait sites.
3.3 Male-male pairs
In the FL site, male wild pigs selected wetland and high tree canopy areas and avoided cattle supplements (Model 3.1 in Table S1; Figure 2). However, their contacts happened more than expected along fences but less at places with high NDVI and cattle supplements, as shown in Figure 2 and 3. In the TX site, males selected water areas and bait piles for resources and avoided roads. For male-male contact in TX, the individual-RSF was a relatively good approximation of contact RSF, but males tended to make contact at bait sites more than habitat selection and less at roads (Figure 2 and 3).