From landscape patterns to behavioral processes: land cover and climate effects on insects in agricultural systems

Abstract

Agricultural insect pests are influenced by factors ranging from landscape composition to plant chemical cues, yet we often manage them on a field scale. Although our concerns are often focused on pest insect abundances within one field, insect populations are interacting with the landscape and other insect populations across multiple scales during their dispersal, across life stages, and through trophic interactions. As society moves towards more sustainable agroecosystems and precision agriculture, it is important to consider insect ecology and implement a multi-scale approach for pest management. Consistent landscape effects on insect populations are difficult to identify due to large variations in crop type, natural habitat type and field configuration, as well as spatial and temporal variations in climate across studies. This dissertation investigates how surrounding landscape composition and climate shape insect communities within agriculture fields across strong landcover and precipitation gradients but within one consistent crop type while also considering the effects of bordering fields. Then, on a broader scale, how regional land cover and climate influence the spread of the invasive agricultural pest Popillia japonica Newman over time. And lastly, at a finer scale investigates a semiochemical-based mechanism for the behavioral response of P. japonica to intercropping. Together, these studies form a multi-scale viewpoint that considers multiple habitat associations across different life stages, dispersal, and chemosensory-based foraging. First, in a field study, we evaluated effects of landscape composition, precipitation, and bordering edge on insect abundance and community composition within soybean fields along a semi-natural gradient. We found that surrounding semi-natural areas affected the insect community, increasing pollinator abundance and decreasing pest abundance. We found evidence of significant interactions, where precipitation can minimize land cover’s effects on beneficial insect populations at lower levels. The interaction between land cover and precipitation suggests that low precipitation results in physiological constraints to the natural plant populations that have an indirect impact on beneficial insects through land cover in addition to the direct effect of low precipitation on the insects themselves. Second, in a regional study of the spread of the invasive Japanese beetle, Popillia japonica at the leading edge of an invasion front in the central United States, we identified habitat associations and predicted future spread, producing county scale risk maps for use in monitoring and risk assessment. Beetle occurrences were positively correlated with developed areas, corn and soybean, minimum temperature, and precipitation, and negatively correlated with grassland and pasture, and maximum temperature. There was high probability of occurrence for Japanese beetles across the entire region. Here we see the importance of habitat across various life stages as developed areas are necessary for providing P. japonica larvae with turf grass and adults can feed on corn and soybean. We also see the importance of climatic variables of temperature and precipitation which exert physiological constraints on the development and survival of P. japonica. Interestingly, we found a negative correlation with grassland and pasture despite grass roots being the preferred food source for larval P. japonica. This is likely due to regional differences in grassland root systems and a trophic interaction with grassland birds, who are natural predators of P. japonica. Third, we investigated a semiochemical-based mechanism for the behavioral response of P. japonica to intercropping soybean with sorghum. Although we found unique semiochemical profiles among our treatments, P. japonica did not show a significant preference for our treatment extracts over controls or to any of the treatments in a dual choice assay. If intercropping soybean with sorghum significantly affects the behavior of P. japonica in the field, it is likely to occur at a different scale, or in response to non-olfactory stimuli. The lack of a behavioral response to semiochemical differences in isolation, despite earlier field studies suggesting this as the mechanism for their observed differences in abundance at the field-level, indicates that P. japonica are making decisions across multiple senses and scales. Our results indicate that insect communities in agricultural fields are shaped by interacting effects of landscape composition and climate, but also that the strength and direction of these effects depend on processes operating at multiple, nested scales. Our field-scale findings demonstrate that semi-natural habitat can reduce the abundance of key soybean pests, including the Japanese beetle. Yet the observed interaction between precipitation and land cover suggests that the positive effect of natural land on the beneficial ecosystem services by natural predators and pollinators may be diminished under low precipitation conditions. Similarly, we see that on a regional-scale, temperature and precipitation are important factors in moderating the distribution and spread of P. japonica across the landscape. These regional patterns constrain expectations of P. japonica infestation at the field scale, as local management decisions are nested within regional landscape composition and climatic suitability. At the behavioral scale, the lack of response to semiochemical cues on a local scale suggests that mechanisms often assumed to drive plant-insect interaction patterns may not operate as expected in the absence of other cues. In this case, P. japonica responses to intercropping soybean and sorghum likely emerge from multi-sensory cues or from interactions within landscape context. Together, these findings highlight that field-scale management strategies are limited by regional climatic and landscape context and depend on behavioral mechanisms that may not scale linearly, underscoring the need for integrative, multi-scale approaches to sustainable pest management.