Interactive effects of plant species diversity and elevated CO2 on soil biota and nutrient cycling

摘要

Terrestrial ecosystems consist of mutually dependent producer and decomposer subsystems, but not much is known on how their interactions are modified by plant diversity and elevated atmospheric CO2 concentrations. Factorially manipulating grassland plant species diversity and atmospheric CO2 concentrations for five years, we tested whether high diversity or elevated CO2 sustain larger or more active soil communities, affect soil aggregation, water dynamics, or nutrient cycling, and whether plant diversity and elevated CO2 interact. Nitrogen (N) and phosphorus (P) pools, symbiotic N-2 fixation, plant litter quality, soil moisture, soil physical structure, soil nematode, collembola and acari communities, soil microbial biomass and micro. ora community structure (phospholipid fatty acid [PLFA] profiles), soil enzyme activities, and rates of C fluxes to soils were measured. No increases in soil Cfluxes or the biomass, number, or activity of soil organisms were detected at high plant diversity; soil H2O and aggregation remained unaltered. Elevated CO2 affected the ecosystem primarily by improving plant and soil water status by reducing leaf conductance, whereas changes in C cycling appeared to be of subordinate importance. Slowed-down soil drying cycles resulted in lower soil aggregation under elevated CO2. Collembola benefited from extra soil moisture under elevated CO2, whereas other faunal groups did not respond. Diversity effects and interactions with elevated CO2 may have been absent because soil responses were mainly driven by community-level processes such as rates of organic C input and water use; these drivers were not changed by plant diversity manipulations, possibly because our species diversity gradient did not extend below five species and because functional type composition remained unaltered. Our findings demonstrate that global change can affect soil aggregation, and we advocate that soil aggregation should be considered as a dynamic property that may respond to environmental changes and feed back on other ecosystem functions.