Optical variability along a river plume gradient: Implications for management and remote sensing

Effective assessment and management of seagrass distributions require a thorough understanding of the water environment, in particular the inherent optical properties (IOPs) of the water column, and the linkage with environmental forcing such as coastal run-off. Coastal waters off the Suwannee and Steinhatchee rivers in the northeastern Gulf of Mexico where seagrass is abundant were sampled regularly in 2010 and 2011 during periods ranging from normal spring flooding to record-low flow conditions associated with widespread drought. Relationships between surface salinities, chlorophyll a concentrations (Chl-a), and IOPs were examined to characterize the optical variability of this region as well as its connection with river discharge and wind forcing. Significant spatial and temporal variability was observed for Chl-a (0.285–14.4 mg m−3), colored dissolved organic matter (CDOM) absorption at 443 nm, aCDOM(443), (0.042–7.53 m−1), and particulate backscattering at 660 nm, bbp(660), (0.002–0.067 m−1) with gradients evident between nearshore (riverine) and offshore (oceanic) waters. Overall, CDOM was the dominant optically significant constituent (OSC), contributing ∼74–75 ± 11% to total (non-water) absorption at 443 nm and 555 nm. CDOM was mainly controlled by river flow, with an inverse correlation observed between aCDOM(443) and salinity (r2 = 0.66). Chl-a showed less direct dependence on flow, exhibiting elevated concentrations (>5 mg m−3) mainly during the summer. Particulate backscattering in this region was relatively low and largely controlled by non-algal particles, as was evident by the strong relationship between bbp(660) and detrital absorption at 443 nm, ad(443) (r2 = 0.71). Compared with other coastal waters in the eastern Gulf of Mexico, aCDOM(443) was ∼3–5 times higher and bbp(660) was ∼50% lower, making the study region very dark. Given that the optical properties of coastal waters in this region are strongly influenced by CDOM derived from terrestrial discharge, remote-sensing algorithms for determining long-term Chl-a trends should focus on utilizing wavebands less influenced by CDOM.