Barium in twilight zone suspended matter as a potential proxy for particulate organic carbon remineralization : results for the North Pacific

Abstract

This study focuses on the fate of exported organic carbon in the twilight zone at two contrasting environments in the North Pacific: the oligotrophic ALOHA site (22°45' N 158°W; Hawaii; studied during June–July 2004) and the mesotrophic Subarctic Pacific K2 site (47°N, 161°W; studied during July-August 2005). Earlier work has shown that non-lithogenic, excess particulate Ba (Baxs) in the mesopelagic water column is a potential proxy of organic carbon remineralization. In general Baxs contents were significantly larger at K2 than at ALOHA. At ALOHA the Baxs profiles from repeated sampling (5 casts) showed remarkable consistency over a period of three weeks, suggesting that the system was close to being at steady state. In contrast, more variability was observed at K2 (6 casts sampled) reflecting the more dynamic physical and biological conditions prevailing in this environment. While for both sites Baxs concentrations increased with depth, at K2 a clear maximum was present between the base of the mixed layer at around 50m and 500m, reflecting production and release of Baxs. Larger mesopelagic Baxs contents and larger bacterial production in the twilight zone at the K2 site indicate that more material was exported from the upper mixed layer for bacterial degradation deeper, compared to the ALOHA site. Furthermore, application of a published transfer function (Dehairs et al., 1997) relating oxygen consumption to the observed Baxs data indicated that the latter were in good agreement with bacterial respiration, calculated from bacterial production. These results corroborate earlier findings highlighting the potential of Baxs as a proxy for organic carbon remineralization. The range of POC remineralization rates calculated from twilight zone excess particulate Ba contents did also compare well with the depth dependent POC flux decrease as recorded by neutrally buoyant sediment traps, except in 1 case (out of 4). This discrepancy could indicate that differences in sinking velocities cause an 3 uncoupling of the processes occurring in the fine suspended particle pool from those affecting the larger particle pool which sustains the vertical flux, thus rendering comparison between both approaches risky.