Interactive effects of land-use intensity, grazing and fire on decomposition of subtropical seasonal wetlands

•The teabag index (TBI) was adopted to quantify subtropical wetland decomposition.•Land-use intensification, grazing, and fire interactively affected decomposition.•Hydrology and climate contributed to spatial–temporal variation in decomposition.•TBI is a useful indicator to inform wetland ecosystem service management. Understanding how global change drivers and their interactions affect decomposition in wetlands embedded in agricultural landscapes remains challenging, especially in tropical and subtropical biomes, which are disproportionately important for global carbon cycling yet with high uncertainties. In a long-term whole-ecosystem experiment, we used a widely-adopted indicator of decomposition, the Teabag Index (TBI), to investigate individual and interactive effects of surrounding land-use intensity, cattle grazing, and prescribed fire on seasonal wetland decomposition (i.e., decay rate k, and stabilization factor S) in an exemplar subtropical landscape in central-south Florida, USA. We ask: (1) How do land-use intensity and management practices affect the decomposition in seasonal wetlands? (2) How does wetland decomposition vary with hydrological gradient and seasonal climatic condition? (3) What are the dominant direct and indirect pathways through which land-use intensity and management practices affect wetland decomposition? Our results showed that, overall, surrounding land-use intensity exerted much stronger effects on litter decomposition compared to grazing and prescribed fire, and cattle grazing reduced decomposition rate k. There was a complex three-way interaction among land-use intensity, grazing and fire on the stabilization factor S – a proxy related to carbon sequestration potential. Grazing-fire interactive effects on S only existed in wetlands embedded in intensively-managed landscapes, where grazing reduced S in burned wetlands but did not affect S in unburned wetlands. Decay rate k showed profound temporal variation primarily driven by seasonal climatic conditions (especially precipitation), whereas S varied spatially along the hydrological gradient. Our structural equation modeling (SEM) analyses further revealed that effects on k were also indirectly manifested through changes on soil C/N, C/P, and pH, while effects on S were manifested through alterations in soil C/P ratio, pH, soil nutrients and water content. Our TBI results suggest that anthropogenic environmental changes, including land-use intensification, livestock intr