Heterotrophic bacteria are thought to be important components of aquatic ecosystems in several ways. These bacteria remineralize organic materials and convert some organic material into bacterial biomass. We examined data from 70 studies in which estimates of production of heterotrophic bacterial biomass (bacterial production) were reported for fresh-and saltwater ecosystems. In sediments, bacterial production was sigdicantly (p <0.001), positively correlated to sediment organic C content. Systems which had hlgh rates of benthic primary production (such as coral reefs) had rates of bacterial production greater than those predicted by sediment organic C content alone. In the photic zone of lakes and the ocean, bacterial production was significantly correlated with planktonic primary production, chlorophyll a, or numbers of planktonic bacteria. For all planktonic systems analysed, bacterial production ranged from 0.4 to 150 pg C 1-' d-' and averaged 20 % (median 16.5 %) of planktonic primary production. On an area1 basis for the entire water column, bacterial production ranged from 118 to 2439 mg m-2 d-' and averaged 30 % (median 27 %) of water column primary production. Heterotrophic bacterial production is, thus, a large component of total secondary production and is roughly twice as large as the production of macrozooplankton for a given level of primary production.

No abstract

Variation in bacterial production (BP) is used as an indicator of bacterial metabolism and carbon processing in the analysis of aquatic ecosystems. The allocation of carbon by bacteria to either BP or respiration (BR), however, is variable and may potentially influence the assessment of carbon cycling by bacteria in ecosystems. We studied 10 transects in the Hudson River estuary where there is a gradient in BP and BR along the flow path of the estuary. We measured BP and BR in filtered samples to derive an estimate of bacterial growth efficiency (BGE) that we could compare with independent measurements of total BP (BP T ) from unfiltered samples to evaluate the relationship of BGE to BP. We further tested the assumption that BGE derived from filtered samples is a good estimator of the ambient BGE on a subset of transects in the upriver section where bacteria dominate respiration. There was good agreement (near 1:1) between respiration measured in unfiltered samples and total BR estimated from BP T and BGE in the filtered fraction (total BR = [BP T /BGE] -BP T ). BGE averaged 0.29 but varied from 0.07 to 0.61, and did not explain the general decline in BP T along the river. Rather, BGE was strongly correlated to the residuals of the BP T and specific BP T (i.e. growth) vs. flow path (= river kilometer) relationship, indicating that shifts in bacterial carbon allocation explained local variations in bacterial metabolic activity, and these shifts were superimposed on the larger scale decline in carbon consumption and BP. The pattern in BP along the Hudson River is clearly a combination of changes in consumption as well as in BGE, to the point that the pattern in BP T or growth would be impossible to recreate from any one of these 2 components. We conclude that BGE indicates changes in carbon allocation of bacteria that reflect shifts in relative BR and BP at shorter time and space scales that are distinct from larger overall patterns in consumption and BP.

ABSTRACT. Planktonic bacterial respiration (estimated from rates of O2 consumption) was the largest fraction of total planktonic respiration at a site in the tidal freshwater Hudson River, New York, USA.No simple relationship between respiration and bacterial production was found, indicat~ng considerable variability in bacterial growth efficiencies. Mean growth efficiency was 22 % but values ranged from c 10 to >S0 %,. Non-phytoplanktonic carbon is known to be important in supporting the high rates of planktonic bacterial productivity observed in the Hudson. Experimental additions of particulate detritus derived from the most common submerged macrophyte (Vall~sneria arnericana) and wetland plant (Typha angustifolia) to Hudson water did not result in increases in bacterial productivity. In contrast, additions of dissolved organic carbon (DOC) denved from these same plants consistently yielded large increases in bacterial production to rates several-fold greater than rates measured in the field. Growth efficiencies for DOC were 37 and 63 ' V" for T angustlfolia and V americana respect~vely.These measures of respiration, production and growth efficiency allow us to estimate the amount of allochthonous carbon necessary to support bacterial productivity Required carbon inputs range from 140 to 320 pg C 1-' d-' depending on the assumed conversion efficiency. These inputs a r e 3 to 6 x net carbon fixation by phytoplankton. The bacterial assim~lat~on of dissolved organic carbon results in DOC turnover times of 2 to 3 wk which are less than the residence time of water in the mid-Hudson during most of the year. The lack of local or downstream depletion of DOC indicates that there must be new sources of DOC within this region of the Hudson. Bacterial metabolism of terrestrially-derived organic carbon has the potential to alter the quantity and composition of material supplied to the ocean.

Multiple sediment trap (MULTITRAP) arrays were deployed at 3 hydrographically and biologically distinct stations in the eastern North Pacific Ocean and used to evaluate experimentally the hypothesis that protozoan grazers are important in the diagenesis of sinking biogenic materials in the ocean. Measurements were made of adenosine triphosphate (ATP) concentrations and the abundances of bacteria and zooflagellates associated with sinking particles collected from in situ preserved and 2 sets of unpreserved (live) traps: one containing a selective eukaryote inhibitor (thiram) and the second without thiram which served as a live-control. Particulate organic carbon and nitrogen (POC and PON), NH:, NO;, NO;, dissolved organic nitrogen (DON), diatoms, and fecal pellets were quantified in live-control and live-thiram traps. In situ microbial production was estimated by measuring RNA and DNA synthesis in live-control and live-thiram traps pre-charged with 3H-adenine. Trap-collected bacteria and zooflagellates accounted for a low percentage of the total POC (2 to 11 % ) but a high percentage of the total microbial biomass (an estimated 25 to 2 100 % of total ATP). Zooflagellates were the principal protozoans collected and accounted for 2 to 25% of the total microbial biomass. Synthesis of RNA and DNA was observed at all stations and depths, indicating microbial growth in the live traps. ATP and microscopic analyses, however, independently confirmed that death rates exceeded growth rates on the sedimenting particles. Differences between the livecontrol traps (coupled microbial assemblage) and live-thiram traps (uncoupled assemblage, grazers inhibited) indicated slower decomposition and nutrient regeneration in the uncoupled system. In general, POC was higher and in the shallowest traps, NH: was lower, and DON was higher in the livethiram traps. Zooflagellates appear to strongly influence decomposition and nutrient recycling within the microbial assemblage associated with sedirnenting particles.