Abstract
Salt marshes play a crucial role in coastal biogeochemical cycles and provide unique ecosystem services. Salt marsh biomass, which can strongly influence such services, varies over time in response to hydrologic conditions and other environmental drivers. We used gap-filled monthly observations of Spartina alterniflora aboveground biomass derived from Landsat 5 and Landsat 8 satellite imagery from 1984-2018 to analyze temporal patterns in biomass in comparison to air temperature, precipitation, river discharge, nutrient input, sea level, and drought index for a southeastern US salt marsh. Wavelet analysis and ensemble empirical mode decomposition identified month to multi-year periodicities in both plant biomass and environmental drivers. Wavelet coherence detected cross-correlations between annual biomass cycles and precipitation, temperature, river discharge, nutrient concentrations (NO x and PO 4 3– ) and sea level. At longer periods we detected coherence between biomass and all variables except precipitation. Through empirical dynamic modeling we showed that temperature, river discharge, drought, sea level, and river nutrient concentrations were causally connected to salt marsh biomass and exceeded the confounding effect of seasonality. This study demonstrated the insights into biomass dynamics and causal connections that can be gained through the analysis of long-term data.