A conserved pattern in plant-mediated interactions between herbivores

Abstract

Plant-mediated interactions between herbivores are important determinants of community structure and plant performance in natural and agricultural systems. Current research suggests that the outcome of the interactions is determined by herbivore and plant identity, which may result in stochastic patterns that impede adaptive evolution and agricultural exploitation. However, few studies have systemically investigated specificity versus general patterns in a given plant system by varying the identity of all involved players. We investigated the influence of herbivore identity and plant genotype on the interaction between leaf-chewing and root-feeding herbivores in maize using a partial factorial design. We assessed the influence of leaf induction by oral secretions of six different chewing herbivores on the response of nine different maize genotypes and three different root feeders. Contrary to our expectations, we found a highly conserved pattern across all three dimensions of specificity: The majority of leaf herbivores elicited a negative behavioral response from the different root feeders in the large majority of tested plant genotypes. No facilitation was observed in any of the treatment combinations. However, the oral secretions of one leaf feeder and the responses of two maize genotypes did not elicit a response from a root-feeding herbivore. Together, these results suggest that plant-mediated interactions in the investigated system follow a general pattern, but that a degree of specificity is nevertheless present. Our study shows that within a given plant species, plant-mediated interactions between herbivores of the same feeding guild can be stable. This stability opens up the possibility of adaptations by associated organisms and suggests that plant-mediated interactions may contribute more strongly to evolutionary dynamics in terrestrial (agro)ecosystems than previously assumed.

Keywords: Genetic variation; herbivory; indirect interactions; induced defense; plant resistance.

Figure 1

A first instar western corn rootworm larva (Diabrotica virgifera virgifera) attacking a maize root. Picture credit: Christelle A.M. Robert.

Figure 2

Diabrotica balteata specifically avoids roots of leaf‐infested plants. Average root preferences (±SE) of dual‐choice experiments are shown. Leaves were induced by wounding or wounding and application of S. littoralis oral secretions. Stars indicate a significant preference for control plants (FDR‐corrected q < 0.025). The P‐value of an analysis of variance comparing the different choice situations is shown on the right.

Figure 3

Diabrotica balteata avoids leaf‐infested plants independently of the site of attack. Average root preferences (±SE) of dual‐choice experiments are shown. Individual leaves were induced by wounding and application of S. littoralis oral secretions. Stars indicate a significant preference for control plants (FDR‐corrected q < 0.05). The P‐value of an analysis of variance comparing the different choice situations is shown on the right.

Figure 4

Leaf herbivore identity determines D. balteata root preference. Average root preferences (±SE) of dual‐choice experiments are shown. Plants were induced by wounding and application of oral secretions of different leaf‐chewing herbivores. Stars indicate a significant preference for control plants (FDR‐corrected q < 0.03). Raw P‐values for P ≤ 0.1 are depicted. The P‐value of an analysis of variance comparing the different choice situations is shown on the right.

Figure 5

Root herbivore preference patterns do not differ between Diabrotica species. Average root preferences (±SE) of dual‐choice experiments are shown. Plants were induced by wounding and application of S. littoralis oral secretions (Star indicates an FDR‐corrected q < 0.016). Raw P‐values for P ≤ 0.1 are depicted. The P‐value of an analysis of variance comparing the different choice situations is shown on the right.

Figure 6

The plant genetic background determines D. balteata root preference. Average root preferences (±SE) of dual‐choice experiments are shown. Inbred lines with different genetic backgrounds were induced by wounding and application of S. littoralis oral secretions. Stars indicate a significant preference for control plants (FDR‐corrected q < 0.03). Raw P‐values for P ≤ 0.1 are depicted. The P‐value of an analysis of variance comparing the different choice situations is shown on the right.

Figure 7

A conserved pattern of plant‐mediated interactions between herbivores in maize. The figure summarizes the direction of the effects of leaf induction by different herbivores on the responses of different root herbivores in the different plant genetic backgrounds that were tested in this study. Note that only neutral to negative effects were observed; no combination resulted in an increase of attractiveness of the roots for the root herbivores.