Résumé
It is commonly acknowledged that the risk of insect pest outbreaks in annual crops may be different in agroforestry systems to that in annual crops grown alone; this mainly because of: i) difference in resource availability trees and herbaceous bands at their bottom may restrain the pest population expansion; ii) micro-climate alteration radiation and therefore temperature modification due to trees may change the insect development ratio; and iii) altered prey predator equilibrium according the ‘natural enemies’ hypothesis, the enhanced vegetal diversity associated with trees may provide nutrients and protection for natural enemies of insect pests, and thus help keep them in the field, contributing to the prevention of pest outbreaks. Little data however are available to corroborate these hypotheses in temperate agroforestry systems. We conducted field experiments to test these hypotheses in a walnut winter wheat agroforestry system in southern France in two consecutive years. We monitored populations of cereal aphids, specific predators such as ladybirds (Coccinellidae), hoverflies (Syrphidae) and lacewings (Chrysopidae) and evidence of parasitism (mummified aphids), in adjacent plots of winter wheat alone and walnut wheat. Moreover, half of the agroforestry plots had lower specific richness due to weed destruction at the bottom of trees, and the other half were enriched by flower sowing in herbaceous bands. No clear differences were found in aphid or natural biocontrol agent development, neither between ‘classical’ and agroforestry plots, nor between enriched and poor agroforestry plots. The effect of temperature reduction alone on the development of cereal aphids was also tested in winter wheat by simulating tree shadow cast on the crop. Again no differences were found between ’shaded’ and ‘normal’ wheat plots, in spite of a significant difference in temperature. These results are discussed in relation to experimental context and their significance for further investigations in pest management in agroforestry systems