The contrasted impacts of grasshoppers on soil microbial activities in function of ecosystem productivity and herbivore diet
Herbivory can have contrasted impacts on soil microbes and nutrient cycling, which has stimulated the development of conceptual frameworks exploring the links between below- and aboveground processes. The "productivity model" predicts that herbivores stimulate microbial activities and accelerate nutrient mineralization in productive ecosystems, while they have an opposite effect in less productive ecosystems. In parallel, the "diet model" predicts that herbivores feeding on conservative plants accelerate nutrient cycling while those feeding on exploitative plants decelerate nutrient cycling, due to changes in litter inputs. Since these two frameworks can lead to conflicting predictions in some cases, experimental evidence combining herbivore diet and plant productivity is required. During two consecutive years, we conducted an experiment controlling the presence of three grasshopper species consuming either grasses, forbs or both in twelve natural and managed alpine grasslands with contrasted productivities. In order to assess the effects of herbivory on soil microbes, we measured their extracellular enzymatic activities, biomass and potential nitrogen mineralization (PNM). Soil and vegetation were also characterized to test how much they modulated the effects of herbivory on microbes. Contrary to the predictions of the diet model, the effects of herbivory on microbial characteristics did not depend on the herbivores diet, but were influenced by primary production, though in a way that differed from the productivity model. The most productive sites were constituted by exploitative plant species which depleted N resources in the soil, and by microbes producing relatively few extracellular enzymes, leading to a lower PNM. Herbivory increased microbial biomass and decreased the production of extracellular enzymes in those sites, possibly through the stimulation of root exudates produced by exploitative species. The least productive sites were characterized by conservative plants, high soil C content, and by microbes having a resource acquisition strategy (more extracellular enzymes, higher PNM). Herbivory decreased microbial biomass and increased the production of extracellular enzymes in those sites. This pattern can be explained by the loss of carbon associated with insect respiration, which increases the resource requirements of microbes and by a lower production of root exudates by conservative species. Therefore, the effects of two years of herbivory on soil microbes were at odds with the productivity model, which focuses instead on longer term effects corresponding to herbivory-induced changes in plant species composition. This highlights the multidimensional feature of the impacts of herbivory on ecosystem functioning, both in space and time.