Aphid-infested beans divert ant attendance from the rosy apple aphid in apple-bean intercropping

Abstract

Ecological intensification of cropping systems aims at restoring multi-functionality while supporting current productivity levels. Intercropping is a form of ecological intensification involving ecological processes beneficial to farmers that do not take place in monocultures. Thus, it represents a practical approach to decrease the use of synthetic inputs such as insecticides in cultivated systems. Whereas insecticide reduction via intercropping-facilitated suppression of aphids is reported in literature, the majority of published studies focussed on herbaceous crops. Thus, the effect of intercropping on aphid populations of cultivated trees remains largely unaddressed. In this study we hypothesized that intercropping a specific companion plant within perennial crops would divert ant attendance from an aphid attacking the crop to another aphid feeding on the newly introduced plant, reducing aphid damage on the crop. We tested our hypothesis in the system of apple (Malus domestica Borkhausen), the rosy apple aphid (Dysaphis plantaginea Passerini) and the black garden ant (Lasius niger L.). Bean plants (Vicia faba) with the black bean aphid (Aphis fabae Scopoli) were intercropped within apple trees inoculated with D. plantaginea. We measured ant attendance, aphid development and survival as well as honeydew composition on both plant species through semi-field and field experiments. The majority of ants chose to attend A. fabae over D. plantaginea in the semi-field experiment with potted plants. In the orchard, a larger majority of scouts were scored on A. fabae over D. plantaginea. A higher number of D. plantaginea colonies remained active in the apple control, whilst they were almost eradicated by intercropping. Although chemical analyses of honeydew disclosed differences in the carbohydrate and amino acid profiles between aphid species, the difference in honeydew composition did not explain the preference for A. fabae. Ants did not discriminate between the two honeydew mimics both in laboratory and field bioassays. Our results showed the potential of intercropping apple trees with beans as a method to reduce ant attendance and thus colony survival. We propose that intercropping represents a bottom-up approach towards ecological intensification of perennial crops. Together with other ecosystem-based measures such as habitat management, intercropping should be considered when planning ecosystem redesign to increase biological control of pests.

Figure 1

Schematic representation of the greenhouse choice assay. (A) apple tree with a D. plantaginea colony, (B) bean plant with a A. fabae colony, (C) choice arena (D) wooden bridge connecting the pot containing the ant nest to the choice arena, (E) pot containing the L. niger nest, and (F) masonry buckets filled with water.

Figure 2

Schematic representation of the intercropping experiment in the orchard with apple trees and beans.

Figure 3

(A) Boxplot of the number of ants attending D. plantaginea and A. fabae colonies over time in greenhouse experiment and predicted values (±95% confidence intervals) of the GLMM. (B) Predicted values (±95% confidence intervals) of the GLMM representing the % of D. plantaginea and A. fabae colonies attended by L. niger over time.

Figure 4

Mean number of attending ants (±SE) per plant. Asterisk indicates statistically significant differences (GLM, χ² = 9.3, df = 1, P = 0.002).

Figure 5

(A) Proportion of surviving D. plantaginea colonies (±95% Wald confidence interval). (B) Mean number of D. plantaginea aphids (±SE) per colony. Asterisk indicates statistically significant differences (GLMM, Tukey’s test, P < 0.050).

Figure 6

Boxplot of normalized (A) sugar content and (B) amino acids content (as proportion) in honeydew collected from A. fabae (Af) and D. plantaginea (Dp). Only P-values < 0.100 are shown (Wilcoxon test). The amino acids arginine, citrulline, glutamine, leucine, phenylalanine and tyrosine were not included in the analysis because they were detected only once.

Figure 7

Boxplot of L. niger visiting water, sucrose solution and honeydew mimic of A. fabae and D. plantaginea in the (A) laboratory and (B) field experiments. Different letters indicate significant differences (GLM, P < 0.050).