Bacterial Remineralization of Particles
21 , as observed during SERIES. Trapping efficiency is thought to be determined by the characteristics of sinking particles, trap design and upper-ocean hydrodynamics 19. The location of the 'In' traps relative to the patch was monitored using a logging fluorometer 10 m subsurface on the surface-tethered array. Traps remained near the patch centre; on day 23 the array was at the periphery for 36 h before recovery and therefore underestimated vertical export. 'Out' traps were deployed at least 20 km northeast of the patch. Limited replicates were available from trap cups, and the standard error (n ¼ 3) for fluxes was ^1.3 mmol Si m 22 d 21 and ^2.5 mmol C m 22 d 21 for the 75 and 100 m 'In' traps on day 14. We had no information on bacterial growth efficiency, which was required to calculate bacterial carbon demand 29 : demand ¼ bacterial production £ (1/growth efficiency), published values display a range from 0.1 to 0.7 (ref. 29). Therefore, we employed indirect approaches (changes in DIC concentrations, selective preservation of opal and ammonium accumulation) to estimate particle remineralization. analysis of sinking rates of natural copepod and euphausiid fecal pellets. Intense hydrolytic enzyme activity on marine aggregates and implications for rapid particle dissolution. Sinking particles as a possible source of nutrition for the large calanoid copepod Neocalanus cristatus in the sub-arctic Pacific Ocean. A 3-dimensional time-dependent approach to calibrating sediment trap fluxes. Vertical diffusion and nutrient supply to the surface mixed layer of the Antarctic Circumpolar Current. eddy diffusion, particle sinking, and the interpretation of sediment trap data. Acknowledgements We thank the officers and crews and scientists on board the John P. Tully, El Puma and Kaiyo Maru. We are grateful to S. Toews for shoreside logistical support. The manuscript was improved by comments from K.