Invertebrate species do not only distinguish between host plant species, but also between genotypes within a plant species. Most plants interact with a diverse assemblage of invertebrate species, specialists as well as generalists, belonging to different feeding guilds and trophic groups. A key question in the ecology of plant-invertebrate interactions is whether responses of different invertebrate species to variation in plant traits are independent or correlated. This question bears directly on the issue of diffuse versus pairwise coevolution, since correlated responses preclude coevolution from operating in a pairwise manner. Positive covariation of invertebrate species across plant genotypes is not rare, suggesting a prevalence of diffuse coevolution. Studies addressing this topic have so far focused on the aboveground community of invertebrates. Most plants however also harbor a belowground invertebrate community that rivals its aboveground counterpart in diversity of species, functions and trophic complexity. Investigating the role of these belowground species in the overall covariation of invertebrates across plant genotypes would provide a first step in closing this gap in our understanding. We used six geographically and genetically distinct populations of the dune grass Ammophila arenaria to assess natural colonization by the above- and belowground invertebrate community in a field experiment.

Results/Conclusions

Most of the aboveground invertebrate species significantly reacted to plant genotype, with a clear trend of preference towards local genotypes, indicating that invertebrate diversity decreases with genetic distance of the host plant to the local genotype. Root associated species differentiated less strongly, although the effect of plant genotype on the composition of both the above- and belowground community was significant. When investigating the correlations between species across genotypes, strong positive covariation was detected both within the above- and belowground community. This suggests some generalized resistance in the plants to each group, implying diffuse selection. However, the average correlation between root and shoot associated species was significantly negative. This observed trade-off could be an ecological cost, hampering the directional selection on resistance to either group. Therefore the cost of increased resistance to aboveground invertebrates might be an increased susceptibility to root feeders. Research on other plant species is required before generalizations of these results can be made. Our study however demonstrates that including the belowground community in research of invertebrate responses to plant genotype can substantially alter our insight into the ecology of plant-invertebrate interactions.