6533b853fe1ef96bd12ad30d

RESEARCH PRODUCT

Untangling the effects of local and landscape structure on pest predation by natural enemies in an intensive crop agroecosystem

subject

description

EAGESTAD INRA; In agricultural crops, the use of biological control to reduce pest damages thanks to predation by natural enemies is an important ecosystem service. Unfortunately, the numerous mechanisms driving the efficiency of the predation function in agroecosystems are and often confounding. In this study, we investigated how local covariates and farming landscape structure affect predation rates of crop pest species in an intensive agricultural landscape in mid-west France. In 2016, predation rates of aphid and weed seeds of one species were quantified in a total of 120 fields, comprising four different types of crops as well as grasslands, selected along two landscape gradients: proportion of grassland in a 1 km radius and distance to the nearest grassland. We used an information-theoretic selection framework to test our hypotheses for several competing sets of models of increasing complexity. In addition to local effects linked to the experimental design and the focal plot, we included landscape metrics testing whether the predation rates within crops depended on the distance from the nearest grassland, which would suggest a dispersion-limited predation function, or if predation rates depended rather on the proportion of grasslands around the focal crop, following a spill-over dynamics from highly predated fields to less-predated ones. We found that predation rates varied across crop types, being maximal in grasslands and cereal crops, especially in smaller sized fields and organic farming systems, and that the predation dynamics in both aphids and weed seeds depended primarily on the distance and proportion of grasslands in the landscape.