Response of soil food-web structure to defoliation of different plant species combinations in an experimental grassland community

We established a greenhouse experiment based on replicated mini-ecosystems to evaluate the effects of defoliation of different plant species combinations on soil food-web structure in grasslands. Plant communities, composed of white clover (Trifolium repens), perennial ryegrass (Lolium perenne) and plantain (Plantago lanceolata), were subjected to the following defoliation treatments: no defoliation of any species (control) and selective trimming of all possible one-, two- and three-way combinations of the species either to 27 cm height (weak defoliation) or to 15 cm height (strong defoliation) above the soil surface three times over a 10-week period. Successive defoliations removed the largest amounts of shoot mass from systems in which T. repens was included among the defoliated species because T. repens dominated aboveground plant biomass. At the final harvest shoot mass was lowest in treatments that included defoliation of T. repens, while total root mass was on average lower in strongly than in weakly defoliated systems and did not differ between the control and defoliation treatments. Total shoot production was not affected by defoliation. Microbial basal respiration and soil NO3-N concentration differed between the combinations of defoliated species; e.g. microbial respiration was on average 32% lower in systems in which only L. perenne was defoliated than in systems in which only T. repens was defoliated. Microbial biomass and soil NH4–N concentration were not significantly affected by defoliation treatments. Enchytraeid abundance differed significantly between the combinations of defoliated species: in systems in which only L. perenne was defoliated enchytraeid abundance was on average 88% lower than in systems in which all species or only T. repens were defoliated. Enchytraeid abundance was also positively associated with total defoliated shoot mass. Abundances of both bacterial-feeding and fungal-feeding nematodes were affected by the combination of defoliated species; e.g. the abundance of bacterial feeders was on average 52% lower in systems in which only T. repens was defoliated than in systems in which both P. lanceolata and T. repens were defoliated. Fungal-feeding nematodes were also more numerous in strongly than in weakly defoliated systems and positively associated with total defoliated shoot mass. Herbivorous nematode abundance was not significantly affected by defoliation treatments. The results show that the response of soil food webs to defoliation can be affected by which combination of species in a plant community is defoliated. Further, it seems that the role of the combination of species that are defoliated may for some components of the soil biota (e.g. fungal-feeding nematodes) be explicable simply in terms of the total mass of foliage removed. However, for other components of the soil biota (e.g. bacterial-feeding nematodes and enchytraeids) species-specific properties of different plant species in the combination of defoliated species are also clearly important, over and above simple mass removal effects of defoliation.