Wed, Aug 17, 2022: 10:00 AM-10:15 AM
516D
Background/Question/MethodsAgroecosystems are a major component of terrestrial landscapes and often an interface for conflict with invasive species. Forest fragments within an agricultural matrix may conform to the principles of island biogeography theory, allowing for general predictions on invasion success. The Mississippi Alluvial Valley (MAV) is an agroecosystem with forested fragments of varying sizes that has been invaded by wild pigs (Sus scrofa). We deployed camera trapping arrays during spring 2016-2019 in 46 forested fragments within the MAV to determine wild pig presence, then used landscape structure, connectivity, and forage availability variables to determine probability of pig occupancy. Explanatory variables included: Size of forest fragment, distance to nearest fragment, corridors linking fragments, water presence within the fragment and crop type surrounding the fragment (i.e., matrix). We used model comparison to determine which variables most influenced wild pig invasion. Additionally, 20 years of remotely sensed agricultural data was used to determine how variation in matrix permeability affected the strength of relationships between other variables and pig occupancy. We predicted the metrics of island biogeography (i.e., island size and distance to mainland) would be useful predictors of pig occupancy but matrix variability associated with crop type would influence the strength of predictive power.
Results/ConclusionsAs hypothesized, wild pig occupancy was best predicted by forest fragment size and distance to next nearest forest fragment. After compiling all possible additive models, the best performing model contained variables for fragment size and distance to nearest fragment (i.e. occupancy ~ fragment size + distance to next fragment, p = 0.005). Also, some crop types (e.g., corn) may reduce resistance to movement between forest fragments. Our results indicate that island biogeography theory may serve as a useful framework to predict biological invasions in highly fragmented ecosystems. Matrix instability in fragmented terrestrial landscapes like agroecosystems where crop types vary spatiotemporally, though, may influence resistance to movement among fragments affecting the predictive power of models using island biogeography principles.
Results/ConclusionsAs hypothesized, wild pig occupancy was best predicted by forest fragment size and distance to next nearest forest fragment. After compiling all possible additive models, the best performing model contained variables for fragment size and distance to nearest fragment (i.e. occupancy ~ fragment size + distance to next fragment, p = 0.005). Also, some crop types (e.g., corn) may reduce resistance to movement between forest fragments. Our results indicate that island biogeography theory may serve as a useful framework to predict biological invasions in highly fragmented ecosystems. Matrix instability in fragmented terrestrial landscapes like agroecosystems where crop types vary spatiotemporally, though, may influence resistance to movement among fragments affecting the predictive power of models using island biogeography principles.