Long-term ecological research can resolve effects of disturbance on ecosystem dynamics by capturing the scale of disturbance and interactions with environmental changes. To quantify how disturbances interact with long-term directional changes (sea-level rise, freshwater restoration), we studied 17 yr of monthly dissolved organic carbon (DOC), total nitrogen (TN), and phosphorus (TP) concentrations and bacterioplankton productivity across freshwater-tomarine estuary gradients exposed to multiple disturbance events (e.g., droughts, fire, hurricanes, and low-temperature anomalies) and long-term increases in water levels. By studying two neighboring drainages that differ in hydrologic connectivity, we additionally tested how disturbance legacies are shaped by hydrologic connectivity. We predicted that disturbance events would interact with long-term increases in water levels in freshwater and marine ecosystems to increase spatiotemporal similarity (i.e., synchrony) of organic matter, nutrients, and microbial activities. Wetlands along the larger, deeper, and tidally influenced Shark River Slough (SRS) drainage had higher and more variable DOC, TN, and TP concentrations than wetlands along the smaller, shallower, tidally restricted Taylor River Slough/Panhandle (TS/Ph) drainage. Along SRS, DOC concentrations declined with proximity to coast, and increased in magnitude and variability following drought and flooding in 2015 and a hurricane in 2017. Along TS/Ph, DOC concentrations varied by site (higher in marine than freshwater wetlands) but not year. In both drainages, increases in TN from upstream freshwater marshes occurred following fire in 2008 and droughts in 2010 and 2015, whereas downstream increases in TP occurred with coastal storm surge from hurricanes in 2005 and 2017. Decreases in DOC:TN and DOC:TP were explained by increased TN and TP. Increases in bacterioplankton productivity occurred throughout both drainages following lowtemperature events (2010 and 2011) and a hurricane (2017). Long-term TN and TP concentrations and bacterioplankton productivity were correlated (r > 0.5) across a range of sampling distances (1-50 km), indicating spatiotemporal synchrony. DOC concentrations were not synchronized across space or time. Our study advances disturbance ecology theory by illustrating how disturbance events interact with long-term environmental changes and hydrologic connectivity to determine the magnitude and extent of disturbance legacies. Understanding disturbance legacies will enhance prediction and enable more effective management of rapidly changing ecosystems.

Net ecosystem carbon balance is a comprehensive assessment of ecosystem function that can test restoration effectiveness. Coastal peatlands are globally important carbon sinks that are vulnerable to carbon loss with saltwater intrusion. It is uncertain how wetland carbon stocks and fluxes change during freshwater restoration following exposure to saltwater and elevated nutrients. We restored freshwater to sawgrass (Cladium jamaicense) peat monoliths from freshwater marshes of the Everglades (Florida, U.S.A.) that had previously been exposed to elevated salinity (approximately9 ppt) and phosphorus (P) loading (1 g P m À2 year À1 ) in wetland mesocosms. We quantified changes in water and soil physicochemistry, plant and soil carbon and nutrient standing stocks, and net ecosystem productivity during restoration. Added freshwater immediately reduced porewater salinity from >8 to approximately 2 ppt, but elevated porewater dissolved organic carbon persisted. Above-and belowground biomass, leaf P concentrations, and instantaneous rates of gross ecosystem productivity (GEP) and ecosystem respiration (ER) remained elevated from prior added P. Modeled monthly GEP and ER were higher in marshes with saltwater and P legacies, resulting in negative net ecosystem productivities that were up to 12Â lower than controls. Leaf litter breakdown rates and litter P concentrations were 2Â higher in marshes with legacies of added saltwater and P. Legacies of saltwater and P on carbon loss persisted despite freshwater restoration, but recovery was greatest for freshwater marshes exposed to saltwater alone. Our results suggest that restoration in nutrient-limited freshwater wetlands exposed to saltwater intrusion and nutrient enrichment is a slow process.