Using optical proxies for biogeochemical properties to study land coverage and terrestrial inputs of organic carbon into coastal waters from the Penobscot Watershed to the Gulf of Maine

Coastal waters are physically, biogeochemically, and therefore optically complex as a result of the commingling of waters arising from terrestrial, freshwater and marine ecosystems. Separating the influences of these three ecosystems on the optical properties of the resulting mixture is challenging, particularly given the variability within each. The longterm goals of this project are twofold: (1) to investigate the feasibility of identifying landuse changes using inwater optical observations, (2) to quantify the input of terrestrial organic carbon into the coastal marine environment. Our approach is to study the Penobscot River Watershed in Maine, from its many tributaries and subwatersheds into the main branch of the Penobscot River, Penobscot Bay and ultimately into the coastal waters of the Eastern Maine Coastal Current. The first step towards this goal is to define optical signatures associated with specific watershed land coverages measured at the source (optical proxies). The next step is to track those proxies from their respective sources to the coast, characterizing the dilution and transformative processes. Using analyses of discrete samples collected monthly from the subwatersheds, we have quantified optical proxies for both dissolved and particulate organic carbon, particulate carbon to nitrogen ratios and fluorophores unique to certain land coverages (e.g. wetlands, residence ponds, lumber mills, agriculture). We apply these proxies to in situ optical observations retrieved from sensors moored within the lower river and upper estuary to quantify the temporal patterns in the biogeochemical properties. To date we have approximately 2 years of observations, including the wettest year on record for the last century in this region. The proxies associated with the dissolved fraction are the most robust and provide signatures for both land coverage and for organic carbon concentration. Those associated with particles are robust proxies for carbon but are less informative about land coverage.