Bacterioplankton distribution and production in the bathypelagic ocean: Directly coupled to particulate organic carbon export?

A recently published evaluation of bacterioplankton abundance and productivity in the bathypelagic North Pacific suggests that these properties are generally coupled with particulate organic carbon (POC) fluxes. In that analysis, bacterial biomass and productivity were several-fold greater in subarctic than subtropical waters, consistent with the basin-scale distribution of POC flux and suggestive of a sinking POC -e DOC -X bacteria transformation of the carbon. To test this hypothesis, we sought to determine whether the very strong spatial and temporal gradients in POC flux in the Arabian Sea would force similar deep-ocean gradients in bacterial variables. On both a withinand between-cruise basis, there was variability in bacterial abundance and thymidine incorporation in the deep Arabian Sea, but correspondence was equivocal between these variables and several correlates to export: flux of biogenic carbon from the euphotic zone, state of the monsoon, and proximity to productive coastal upwelling zones. However, when annual mean bacterial abundance at 2,000 m was compared with annual POC flux at that depth, a strong correspondence emerged: high annual flux supported high bacterial abundance (such a correspondence was not found for bacterial productivity). This finding suggests that bathypelagic bacterial abundance responds to the long-term mean input of organic matter and less to episodic inputs. A comparative evaluation of the North Pacific revealed that although the bathypelagic bacteria there showed correspondence to deep POC flux, that variable alone would not account for the wide meridional variations in bacterial abundance that have been reported. The nutrition and sustenance of organisms in the deep sea has been a classic problem in oceanography since the discovery of life on the sea floor by the Challenger Expedition in 1872-1876 (Mills 1983). With the advent of deep-ocean sediment traps, sedimentation of particulate organic matter became recognized as the principal agent of deep-sea food supply (Moseley 1880; Tyler 1988). Subsidy with dissolved organic matter has long been invoked as a potential nutritional and energy source (J0rgensen 1976). Bacteria dominate the metabolism of deep waters below the euphotic zone (Pomeroy and Johannes 1968), but the mechanism of their nutrition is not clear. The proximal sources of organic matter for bacteria are dissolved, low-molecular-weight substances because bacterial uptake systems are restricted to transporting molecules <500 Da across cell membranes (Williams 2000). However, the ultimate sources of the substances actually transported into bacterial cells are not well characterized, even at a crude operational level. Karl et al. (1988) showed that sediment trap contents colAcknowledgments This analysis included data collected as part of the U.S. JGOFS Arabian Sea Expedition, archived at the Data Management Office (http://usjgofs.whoi.edu/research/arabian.html). H.W.D. in collaboration with F Azam generated the bacterial data as part of that program with support from OCE 9600601. POC flux data collected in the US JGOFS Arabian Sea program and used in this analysis were generated by S. Honjo and J. Dymond. Analysis of these data for this manuscript was supported by NSF OCE 0097237 (to H.W.D.) and OCE 9726091 (to D.A.H.) as part of the US JGOFS Synthesis and Modeling Program. lected from 30-600 m depths did not support bacterial growth, concluding that the organic matter lost from particles by dissolution or fragmentation was oxidized by freeliving cells. Cho and Azam (1988) observed that particleassociated bacteria were rare in the 0-1,000 m water column (<5% of total) but also found a general correspondence of vertical profiles of total bacterial biomass, bacterial production, and estimated sediment trap fluxes. These authors simultaneously advanced the hypothesis that free-living bacteria are the agents of mineralization of the organic matter transported into the deep sea in (or on) sinking particles, leading to the concept that sinking particles are the principal source of bacterial nutrition in the oceanic mesopelagic zone. These studies led to the prediction that bacterial rates and stocks might covary with particle fluxes measured by sediment traps. Ducklow (1993) compared vertical profiles of bacterial production (100-1,000 m) with the supply of sinking particulate organic carbon (POC) in the Arabian Sea and concluded that the POC flux was insufficient to meet the bacterial carbon demand. Simon et al. (1992) made the same comparison for the subarctic Pacific, estimating that bacterial metabolism accounted for 41-172% of the sinking POC flux over 80-600 m. Similarly, H.-G. Hoppe (pers. comm.) estimated from later studies in the Arabian Sea that sinking POC could support only a minor part of the bacterial carbon demand. The results of these studies implicate the large reservoir of suspended organic particles as a potential additional source of bacterial nutrition and demonstrate intense exoenzymatic hydrolysis rates by carbon-limited bacteria in the mesopelagic and deep zones. Of these studies only the last