Grass invasion effects on forest soil carbon depend on landscape-level land use patterns.

Plant invasions can alter the quality and quantity of detrital and root-derived inputs entering a system, thereby influencing the activities of microbial decomposers and affecting the soil carbon cycle. The effect of these inputs on soil carbon storage is often conflicting, suggesting strong context dependency in the plant-decomposer relationship. Whether there is a generalizable pattern that explains this dependency remains relatively unexplored. Here, we (1) examine how invasion by the exotic grass Microstegium vimineum affects carbon cycling across a land use gradient, and (2) evaluate the importance of inorganic nitrogen availability and other environmental variables for explaining patterns in soil carbon. Using paired invaded and uninvaded plots, we quantified invasion effects on belowground carbon pools, extracellular enzyme activities, and native leaf litter decomposition in forests embedded in an urban, agricultural, or forested landscape matrix. Compared to the urban matrix, invasion-associated declines in total soil organic carbon in the forested and agricultural landscapes were 3.5 and 2.5 times greater, respectively. Inorganic nitrogen availability and M. vimineum biomass interacted to explain these patterns: when both nitrogen availability and M. vimineum biomass were high, invaded soils exhibited higher total organic carbon, unchanged particulate organic matter carbon, and higher mineral-associated organic matter carbon compared to adjacent uninvaded soils. Consistent with these patterns, activities of carbon-mineralizing enzymes were lower in invaded than in uninvaded soils when both nitrogen availability and M. vimineum biomass were high. By contrast,. decomposition of native leaf litter was faster when inorganic nitrogen availability and M. vimineum biomass were high. Our findings suggest that, although this invader may accelerate carbon cycling in forest soils, its effects on soil carbon storage largely depend on nitrogen availability and invader biomass, which can be altered by landscape-level patterns of land use. Additional research is needed to determine whether land use or other broad-scale processes such as atmospheric nitrogen deposition can explain context dependence in plant invasion effects on other ecosystem processes.