Direct and indirect controls on organic matter decomposition in four coastal wetland communities along a landscape salinity gradient

Abstract Coastal wetlands store more carbon than most ecosystems globally. As sea level rises, changes in flooding and salinity will potentially impact ecological functions, such as organic matter decomposition, that influence carbon storage. However, little is known about the mechanisms that control organic matter loss in coastal wetlands at the landscape scale. As sea level rises, how will the shift from fresh to salt‐tolerant plant communities impact organic matter decomposition? Do long‐term, plant‐mediated, effects of sea‐level rise differ from direct effects of elevated salinity and flooding? We identified internal and external factors that regulated indirect and direct pathways of sea‐level rise impacts, respectively, along a landscape‐scale salinity gradient that incorporated changes in wetland type (fresh, oligohaline, mesohaline and polyhaline marshes). We found that indirect and direct impacts of sea‐level rise had opposing effects on organic matter decomposition. Salinity had an indirect effect on litter decomposition that was mediated through litter quality. Despite significant variation in environmental conditions along the landscape gradient, the best predictors of above‐ and below‐ground litter decomposition were internal drivers, initial litter nitrogen content and initial litter lignin content respectively. Litter decay constants were greatest in the oligohaline marsh and declined with increasing salinity, and the fraction of litter remaining (asymptote) was greatest in the mesohaline marsh. In contrast, direct effects of salinity and flooding were positive. External drivers, salinity and flooding, stimulated cellulytic activity, which was highest in the polyhaline marsh. Synthesis . Our results indicate that as sea level rises, initial direct effects of salinity will stimulate decay of labile carbon, but over time as plant communities shift from fresh to polyhaline marsh, litter decay will decline, yielding greater potential for long‐term carbon storage. These findings highlight the importance of quantifying carbon loss at multiple temporal scales, not only in coastal wetlands but also in other ecosystems where plant‐mediated responses to climate change will have significant impacts on carbon cycling.