UC Berkeley

This dissertation focuses on three different aspects of the biotic and abiotic effects on pest management and community structure in agricultural systems, at a local level for a specific insect pest, at a landscape level for an insect community and in mesocosm experiments to examine the effects of predation on indirect interactions. Agricultural crops and pastures cover approximately 38% of the available land surface globally, and the intensity of agricultural production has continually increased over the past 300 years. The intensification of management practices in agricultural ecosystems, from pesticide applications to simplified rotations, have had wide-ranging consequences, from a rise in secondary pests to negative effects on native species and the efficacy of biological control from generalist natural enemy assemblages.

The walnut husk fly, Rhagoletis completa (Diptera: Tephritidae), is a secondary pest that has become more problematic in California walnut orchards with the decline in use of organophosphates, and a seasonal emergence pattern that is not well understood. To more effectively manage R. completa the factors that affect the timing of adult emergence were examined using both field data and mechanistic laboratory studies. A phenology model was developed for R. completa using a large historic data set to more accurately predict the timing of the flight period and optimize management decisions. The relationship between cumulative trap catch and degree-day accumulation for R. completa in California was modeled to predict thermal requirements for the start and mid-point of the flight period. Latitude, walnut cultivar leaf-out time, orchard age and year all contributed significantly to explain deviations from model predictions. The negative relationship between year and accumulated degree days at the start of the flight period corroborates grower observations and concerns that R. completa is emerging earlier in the season than it was 18 years ago.

A series of laboratory studies were conducted to further examine the effects of incubation temperature, pre-chill and chill durations, latitude, cultivar and size on the post-diapause development of R. completa puparia. The thermal requirement and developmental thresholds for adult emergence were estimated. Patterns based on sex were also examined in relation to puparial weight and post-diapause development. Percent adult emergence declined at both higher and lower incubation temperatures. Chill and pre-chill duration affected the thermal requirement, survival and synchronization of emergence for R. completa, with climate change implications for the future. Latitude and plant host species also affect the thermal requirement. The effects of these environmental factors on the post-diapause development of R. completa are discussed in relation to observations from other Rhagoletis species.

Insect pests and natural enemies respond to land-use intensity at varying scales. As agricultural production has intensified on the local scale over the last century, landscapes have also become increasingly homogenized. The effects of land-use intensity at both landscape and local scales on pests, generalist predators and sorghum yield were studied in Uruguay. Greater landscape intensity increased densities of pest species and provided evidence that they may interact indirectly through apparent commensalism leading to a decrease in the abundance of the generalist coccinellid community. In contrast to many field studies that have evaluated effects of local intensity on organic agriculture, in this study a multi-year rotational system was compared to continuous agriculture. Crop-grazing rotational systems resulted in lower coccinellid species richness, winter crop cultivation caused a higher proportion of non-native coccinellid individuals, and insecticide-use increased pest density. The use of path analysis provided a robust approach to examine the effects of land-use intensification on insect communities in a lesser known multi-year management system and an under-studied geographic region.

Generalist predators are increasingly recognized as important contributors to the reliability of conservation biological control. In addition, indirect interactions between prey species sharing a common predator are increasingly recognized as drivers of community dynamics. Consideration of how indirect interactions between prey species affect biological control potential is critical. Cage experiments with Acyrthosiphon pisum and Aphis gossypii were performed with varying levels of predation by adult Hippodamia convergens as well as varying levels of habitat complexity. Across predation pressure the early growth of A. gossypii appears beneficial to the net change in abundance of A. pisum over the course of the experiment, however the early growth of A. pisum provides no explanatory power for the net change in abundance of A. gossypii. Both increasing predation pressure and decreasing habitat complexity appear to change the sign and strength of indirect interactions between A. gossypii and A. pisum as predicted by ecological theory.

The negative impacts of agricultural intensification are wide-reaching and insect communities respond to the changes in their environment at multiple levels. Sustainable pest control strategies must consider long-term data sets, whenever possible, take into account environmental, local and landscape variables and evaluate patterns in the context of community ecology.