UC Santa Barbara

Over the next 50 years, global food demand is forecast to double. Already it is estimated that agriculture covers about 40% of ice-free land, accounts for a third of greenhouse gas emissions, and contributes significantly to global biodiversity declines.

One means to reduce the impact of agriculture on humans and natural systems is to ensure the efficient use of pesticides. Pesticides, especially insecticides, have numerous negative externalities for human and environmental health, and their efficient use is an economic, ecological and public health priority. How land use patterns influence insect pests and insecticide demand is of special concern, because productive and efficient land use is key to meeting future food demand.

This research investigates the relationships between insecticide use and landscape configuration. It further investigates the importance of weather variability and data quality to understanding agriculture in the 21st Century. Finally, it explores ecological theory to understand how multiple natural enemies may coexist on a single resource species.

Specifically, I address the following questions: 1) is landscape simplification a consistent driver of insecticide use across time, 2) is landscape simplification a consistent driver of insecticide use across space and throughout the varied growing regions of the US, and do annual weather patterns influence insecticide use? 3) Is satellite crop data sufficiently accurate to be applied to ecological and economic questions at the sub-county level? 4) Can coexistence be driven by non-consumptive ecological interactions?

To address these questions I integrate ecological and economic theory, and apply multivariate statistical techniques to multi-year national or regional databases.

I find that, contrary to expectations from ecological theory, landscape simplification does not consistently drive insecticide use over time (Chapter 1) or space (Chapter 2). This spatio-temporal variation helps explain the ambiguous results in the literature and implies that national land use policy will have very different effects on insecticide use if regional differences are ignored. To further understand the underlying mechanisms requires fine-scale spatial information of configuration and crop type. However, leveraging satellite data for sub-county information such as spatial configuration is well suited to simplified growing regions, but highly inaccurate elsewhere (Chapter 3). Lastly, I show natural enemies and other intermediate consumers can coexist with sufficiently strong non-consumptive effects of a top predator on the dominant consumer (Chapter 4).

In 2007 US farmers applied ~70 million pounds of insecticide active ingredients. While farmers pay the purchase price, society pays for degradation of natural systems and harm to human health. To minimize the cost of insecticides to both farmers and society, now and under future climate change, we must understand what drives variation in insecticide use and what enables persistence of natural enemy diversity. My dissertation research informs these key gaps in our understanding.