Explicit modeling of abiotic and landscape factors reveals precipitation and forests associated with aphid abundance. Issue 8 (22nd November 2016)

• Record Type: Journal Article
• Authors: Stack Whitney, Kaitlin; Meehan, Timothy D.; Kucharik, Christopher J.; Zhu, Jun; Townsend, Philip A.; Hamilton, Krista; Gratton, Claudio
• Is Part Of: Ecological applications. Volume 26:Issue 8(2016)
• Page Start: 2600
• Page End: 2610
• Publication Date: 2016-11-22
• Subjects: crop pest -- ecoinfomatics -- growing degree days -- model averaging -- precipitation -- soybean aphid -- Wisconsin, USA; Ecology -- Periodicals; Environmental protection -- Periodicals; Biology, Economic -- Periodicals; 577.05
• DOI: 10.1002/eap.1418
• Languages: English
• ISSNs: 1051-0761

Abstract

Abstract: Increases in natural or noncrop habitat surrounding agricultural fields have been shown to be correlated with declines in insect crop pests. However, these patterns are highly variable across studies suggesting other important factors, such as abiotic drivers, which are rarely included in landscape models, may also contribute to variability in insect population abundance. The objective of this study was to explicitly account for the contribution of temperature and precipitation, in addition to landscape composition, on the abundance of a widespread insect crop pest, the soybean aphid ( Aphis glycines Matsumura), in Wisconsin soybean fields. We hypothesized that higher soybean aphid abundance would be associated with higher heat accumulation (e.g., growing degree days) and increasing noncrop habitat in the surrounding landscape, due to the presence of the overwintering primary hosts of soybean aphid. To evaluate these hypotheses, we used an ecoinformatics approach that relied on a large dataset collected across Wisconsin over a 9‐year period (2003–2011), for an average of 235 sites per year ( n = 2, 110 fields total). We determined surrounding landscape composition (1.5‐km radius) using publicly available satellite‐derived land cover imagery and interpolated daily temperature and precipitation information from the National Weather Service COOP weather station network. We constructed linear mixed models for soybean aphid abundance based on abiotic and landscape explanatory variables and applied model averaging for prediction using an information theoretic framework. Over this broad spatial and temporal extent in Wisconsin, we found that variation in growing season precipitation was positively related to soybean aphid abundance, while higher precipitation during the nongrowing season had a negative effect on aphid populations. Additionally, we found that aphid populations were higher in areas with proportionally more forest but were lower in areas where minor crops, such as small grains, were more prevalent. Thus, our findings support our hypothesis that including abiotic drivers increases our understanding of crop pest abundance and distribution. Moreover, by explicitly modeling abiotic factors, we may be able to explore how variable climate in tandem with land cover patterns may affect current and future insect populations, with potentially critical implications for crop yields and agricultural food webs.