In order to better understand the constraints on the use of barium as a coastal paleo-freshwater tracer, we surveyed the dissolved Ba distribution in Louisiana Shelf waters, including the Mississippi (MR) and Atchafalaya (AR) River plumes, during May and November 2008, and June/July 2009, which represent high, low and intermediate river discharges, respectively. Dissolved Ba was found dominantly in the <0.02 μm fraction, with no significant contribution from the 0.02 - 0.45 μm colloidal size fraction. Although apparent non-conservative surface water Ba behavior was observed during all three sampling periods, there were significant differences among the distribution patterns. River-seawater mixing experiments were supportive of substantial desorptive Ba addition only during the high discharge survey. At other times, input of Ba-enriched shelf bottom water as well as river endmember variability contributed to the apparent non-conservative behavior. During at least two of our surveys (high and intermediate river discharge), shelf bottom waters were significantly enriched in dissolved Ba relative to surface waters. While the cause of this enrichment (e.g., submarine groundwater discharge, dissolution/diffusion from the sediment, and/or an anthropogenic source such as drilling muds) could not be determined, we did observe that bottom Ba enrichment correlated with diminishing dissolved oxygen during summertime shelf bottom water hypoxia. Another interesting observation was Ba depletion in some high-salinity surface waters associated with a diatom bloom during June/July 2009. In addition, different Ba concentrations in the MR and AR appear related to inputs to the AR from the Red River as well as from the wetlands in the Atchafalaya River Basin.

Overall, our study of the Ba distribution in Louisiana Shelf waters implies that the seasonal variation of the surface water Ba-salinity relationship could lead to a considerable uncertainty in salinity prediction when using Ba as a proxy for paleo-salinity changes. Barium input to bottom waters and the extent to which this is natural or anthropogenically-affected is a particular source of uncertainty. Thus, as is the case with nearly all paleoceanographic proxies, the planktonic foraminiferal Ba/Ca ratio should be used in conjunction with other constraining proxies.