Plant diversity effects on soil heterotrophic activity in experimental grassland ecosystems

The loss of plant species from terrestrial ecosystems may cause changes in soil decomposer communities and in decomposition of organic material with p otential further consequences for other ecosystem processes. This was teste d in experimental communities of 1, 2, 4, 8, 32 plant species and of 1, 2 o r 3 functional groups (grasses, legumes and non-leguminous forbs). As plant species richness was reduced from the highest species richness to monocult ures, mean aboveground plant biomass decreased by 150%, but microbial bioma ss (measured by substrate induced respiration) decreased by only 15% (P = 0 .05). Irrespective of plant species richness, the absence of legumes (acros s diversity levels) caused microbial biomass to decrease by 15% (P = 0.02). No effect of plant species richness or composition was detected on the mic robial metabolic quotient (qCO(2)) and no plant species richness effect was found on feeding activity of the mesofauna (assessed with a bait-lamina-te st). Decomposition of cellulose and birchwood sticks was also not affected by plant species richness, but when legumes were absent, cellulose samples were decomposed more slowly (16% in 1996, 27% in 1997, P = 0.006). A signif icant decrease in earthworm population density of 63% and in total earthwor m biomass by 84% was the single most prominent response to the reduction of plant species richness, largely due to a 50% reduction in biomass of the d ominant 'anecic' earthworms. Voles (Arvicola terrestris L.) also had a clea r preference for high-diversity plots. Soil moisture during the growing sea son was unaffected by plant species richness or the number of functional gr oups present. In contrast, soil temperature was 2 K higher in monocultures compared with the most diverse mixtures on a bright day at peak season. We conclude that the lower abundance and activity of decomposers with reduced plant species richness was related to altered substrate quantity, a signal which is not reflected in rates of decomposition of standard test material. The presence of nitrogen fixers seemed to be the most important component of the plant diversity manipulation for soil heterotrophs. The reduction in plant biomass due to the simulated loss of plant species had more pronounc ed effects on voles and earthworms than on microbes, suggesting that higher trophic levels are more strongly affected than lower trophic levels.