Abstract
Marginal seas represent ~15% of the ocean surface, where extensive coral reefs, diverse benthic communities, productive fisheries, and sprawling mangroves and estuaries are abundant, and all of which are vulnerable to climate change. Furthermore, human activities locally compromise these ecosystems, for example, by releasing fertilizers into rivers that drain into marginal seas, causing eutrophication that in turn creates ‘dead zones,’ while the offshore oil industry has also caused extensive damage (e.g., the 2010 Deepwater Horizon catastrophe). Marginal seas may also offer insights on ocean-driven climate shifts, since they are restricted from the ocean, having unique deep water circulation pathways and chemistry, perhaps resembling the ocean during the geologic past (e.g., the anoxic Black Sea). Furthermore, marginal seas contain paleo proxies that extend further back into the geologic record, making the study of the ancient carbon cycle and climate shifts possible. Since climate and marine chemistry are strongly influenced by the distribution of carbon (e.g., changing concentrations of atmospheric carbon dioxide), understanding its biogeochemical cycling in marginal seas will improve our understanding of climate change. In this dissertation, the contrasting environments of the Black Sea, Gulf of Mexico (GoM) and basins of the Caribbean were evaluated for biogeochemical properties and dynamics. The environmental impacts of anoxia and river input on carbon biogeochemistry in the Black Sea were investigated, utilizing data from a basin-wide transect conducted in 2013 as part of the GEOTRACES Program. The Black Sea has dissolved organic carbon (DOC) concentrations ~2.5 times higher than the open ocean and Mediterranean Sea. Previous studies have suggested that input of terrigenous DOC from rivers is responsible for the relatively high concentrations, however, anoxia may inhibit the mineralization of DOC, causing it to accumulate. The concentrations of DOC were predicted within the basin based on conservation with respect to salinity to trace riverine input; predicted values were then compared with observations to estimate net removal (i.e., deficits) and accumulation (i.e., surpluses). The cycling of DOC was also explored by examining the optical properties of dissolved organic matter (DOM), to provide insights on the composition and predominant origin of the DOM (i.e., terrigenous or marine). Net removal of DOC was identified in the subsurface waters of the Black Sea, and there was no evidence to suggest DOC accumulation under anoxia, instead suggesting that concentrations are due to input of terrigenous DOC from rivers, likely representing ~50% of the DOC in the deep basin—orders of magnitude higher than found in the deep ocean. In the anoxic water, chromophoric dissolved organic matter (CDOM) correlated well with organic matter mineralization derived from hydrogen sulfide (H2S) concentrations, suggesting that the mineralization of sinking particles in the anoxic waters. Carbon mineralization was calculated in the anoxic water based on H2S accumulation and compared with CDOM; the anoxic correlations are similar to CDOM’s known relationship with oxygen-derived mineralization in the ocean, suggesting that CDOM dynamics are fairly consistent between oxic and anoxic environments. The environmental impacts of sill depths on carbon biogeochemistry in the GoM and Caribbean (collectively referred to as the Intra-Americas Seas or IAS) were investigated, utilizing various datasets. The five major basins that comprise the IAS have dissolved inorganic carbon (DIC) concentrations similar to those found in the western North Atlantic. However, below the ~2000 m sill depths that separate the IAS basins, subtle differences in carbon concentrations exist, yet are not adequately mapped due to limited sampling coverage. Differences between the basins are apparent in oxygen concentrations from the 2013 World Ocean Atlas, showing signs of recent Upper North Atlantic DeepWater penetration, perhaps containing anthropogenic carbon. In order to investigate these differences with respect to carbon, relationships between observed DIC concentrations were established with temperature, salinity, oxygen and silicic acid measurements using multiple linear regression (MLR) models. The concentrations of DIC were calculated using MLR models and compared with available observations to determine the distribution of DIC between the deep IAS. The MLR-derived DIC concentrations represent the first coherent view of the carbon system spanning the deep IAS basins. While the IAS contain natural DIC, they may also act as reservoirs for storing anthropogenic carbon; more observations from the different basins are required to confirm their role as storage reservoirs.