Abstract
In most ecosystems, plant roots from different species decompose in mixtures and in the presence of living roots; however much root decomposition research has focused on how roots of individual species or artificial mixtures decompose in the absence of living plants. We thus examined two poorly studied components of root litter decomposition: 1) whether decomposition of root mixtures can be predicted from the sum of the decomposition rates of each component species and 2) how living plants influence rates of root decomposition. Decomposition rates of roots from three perennial herbaceous Mediterranean species grown in monocultures and in two- and three-species mixtures were determined after a one-year incubation period under their living community and in non-vegetated soil (bare soil). Soil respiration in the presence of glucose (substrate induced respiration, SIR) was measured in each plant community and in bare soil. Decomposition rates of root mixtures cannot be predicted from decomposition rates of the component species, both additive and non-additive effects were observed; the presence of low quality roots of Carex humilis in mixtures strongly negatively influenced root decomposition. The presence of living plants stimulated root decomposition in monocultures and two-species communities, likely through an enhanced microbial activity (SIR) under plant communities. This study highlights that root decomposition cannot be predicted from decomposition rates of the component species and is more influenced by endogenous factors or root litter functional composition than by plant community composition.