COS 142-6 - Are ‘big data’ necessary for detecting negative effects of landscape simplification on pest abundance and volatility?

Background/Question/Methods

Agroecologists often argue that diversifying agriculture by planting multiple crops and conserving semi-natural habitats could bolster the predators of crop pests, mitigate outbreaks, and reduce insecticide use. Yet pest control is rarely considered in landscape conservation programs. Part of the problem is that the literature is filled with examples of pests responding positively, negatively, or not at all to surrounding semi-natural habitats. This may be because few studies are conducted across enough sites to tease apart landscape effects from the natural stochasticity of pest population dynamics. Moreover, many farmers’ decisions regarding pest management are influenced by their strong risk-aversion to rare but catastrophic outbreaks that can result in total crop failure. Yet the need for long-term pest population data has largely prevented researchers from exploring landscape effects on pest population volatility and outbreaks. Here, we repurpose a government database of ~1300 olive orchards and vineyards monitored for ~13 years across Spain to study the effects of landscape simplification on three pests: olive moth, olive fly, and European grapevine moth. Specifically, we explored variation in pest abundances, population volatility, and insecticide applications on farms with varying concentrations of surrounding semi-natural habitats and crop monocultures, controlling for climate, farm management, and topography.

Results/Conclusions

We found that pest infestations, population volatility, and outbreaks increased in simplified landscapes dominated by vineyards or orchards. For example, the likelihood that grapevine moths exceeded economic thresholds for spraying insecticides rose four-fold in landscapes without nearby vineyards vs. 90% vineyard cover. As a result, insecticide sprays nearly doubled on farms surrounded by 90% vineyard cover. Similarly, we found that both the abundance and population volatility of olive flies increased in orchards surrounded by more olive groves. Critically, farmers were more than twice as likely to experience >20% losses in olive crops when comparing the most versus the least volatile populations. Finally, we found that spatiotemporal stochasticity in pest populations would have masked these strong landscape effects in smaller datasets. After reducing our dataset to typical sample sizes, only ~40% and ~25% simulations still exhibited significant landscape effects on vineyard pest outbreaks and insecticide applications. Together, our work highlights the importance of landscape management for Spanish farmers, as simplified landscapes are more likely to experience outbreaks and increased insecticide sprays. More broadly, our findings highlight the importance of collecting and analysing large pest-monitoring datasets and may explain why pest responses to landscape composition are so variable in the literature.