Files
Abstract
Generalist predators can be a useful component of biological control, but their role is complex and often context dependent. Their broader feeding allows for higher persistence in the dire conditions of an agricultural . The purpose of this research is to both evaluate new methods of investigating realized diets as well as evaluating how local and landscape factors influence the diet items of generalist predators in the field. The carabid beetle Pterostichus melanarius was used as an example of one such generalist. We evaluated shotgun metagenomics in comparison to typical metabarcoding methods to determine what prey items, pests or non-target, that P. melanarius were consuming. A general bioinformatic pipeline involving contig formation using established metagenome assemblers was also created. Shotgun metagenomics identified a broader suite of diet items, but the increased DNA requirements limited the power of statistical analysis. In addition, local landscape and management features were tested for how they impacted both the number and identity of diet items. On the same organic farms, insect populations were observed, and the influence of both local and landscape level factors were tested for predictive power. Overall community metrics and natural enemy taxa were influenced by local ground cover while brassica pests only meaningfully responded to crop size. Finally, the top-down pressure of birds on the arthropod pest and predator community was evaluated using a combination of fecal diet analysis and small-scale exclusions. We also examined whether landscape level features influenced bird predation pressure. We found that brassica pests and arthropod predators were relatively uncommon prey for birds. Accordingly, bird access yielded higher, rather than lower aphid abundances but little impact on other arthropod abundances. The broad conclusion drawn from this body of research is that generalist predators (both arthropod and avian) have a complex mix of trophic interactions that often cancel out. However, utilizing mixed techniques of broad landscape studies, molecular techniques, and on farm experimentation can form a complete picture and policy recommendations become possible