Factors influencing benthic bacterial abundance, biomass, and activity on the northern continental margin and deep basin of the Gulf of Mexico

Abstract

As part of a larger project on the deep benthos of the Gulf of Mexico, an extensive data set on benthic bacterial abundance (n>750), supplemented with cell-size and rate measurements, was acquired from 51 sites across a depth range of 212–3732 m on the northern continental slope and deep basin during the years 2000, 2001, and 2002. Bacterial abundance, determined by epifluorescence microscopy, was examined region-wide as a function of spatial and temporal variables, while subsets of the data were examined for sediment-based chemical or mineralogical correlates according to the availability of collaborative data sets. In the latter case, depth of oxygen penetration helped to explain bacterial depth profiles into the sediment, but only porewater DOC correlated significantly (inversely) with bacterial abundance (p<0.05, n=24). Other (positive) correlations were detected with TOC, C/N ratios, and % sand when the analysis was restricted to data from the easternmost stations (p<0.05, n=9–12). Region-wide, neither surface bacterial abundance (3.30–16.8×10⁸ bacteria cm⁻³ in 0–1 cm and 4–5 cm strata) nor depth-integrated abundance (4.84–17.5×10¹³ bacteria m⁻², 0–15 cm) could be explained by water depth, station location, sampling year, or vertical POC flux. In contrast, depth-integrated bacterial biomass, derived from measured cell sizes of 0.027–0.072 μm³, declined significantly with station depth (p<0.001, n=56). Steeper declines in biomass were observed for the cross-slope transects (when unusual topographic sites and abyssal stations were excluded). The importance of resource changes with depth was supported by the positive relationship observed between bacterial biomass and vertical POC flux, derived from measures of overlying productivity, a relationship that remained significant when depth was held constant (partial correlation analysis, p<0.05, df=50). Whole-sediment incubation experiments under simulated in situ conditions, using ³H-thymidine or ¹⁴C-amino acids, yielded low production rates (5–75 μg C m⁻² d⁻¹) and higher respiration rates (76–242 μg C m⁻² d⁻¹), with kinetics suggestive of resource limitation at abyssal depths. Compared to similarly examined deep regions of the open ocean, the semi-enclosed Gulf of Mexico (like the Arabian Sea) harbors in its abyssal sediments a greater biomass of bacteria per unit of vertically delivered POC, likely reflecting the greater input of laterally advected, often unreactive, material from its margins.

Introduction

After years of study by independent research groups using a variety of approaches, the environmental determinants of bacterial abundance in marine sediments are only coarsely resolved in the literature. Anticipated correlations with ocean depth, well known for animal populations (Rowe, 1983) before benthic bacterial numbers could be quantified reliably, were typically found to be weak or non-existent (reviewed by Deming and Baross (1993); see also Dixon and Turley, 2000) or else driven by inclusion of data from shallow coastal sites (Bak and Nieuwland, 1997). The relative constancy of benthic bacterial abundance in the upper mixed layer of pelagic sediments, described with a scaling model by Schmidt et al. (1998), was recently reinforced by a global geographic analysis of most of the available (individually small) data sets spanning a depth range of about 200–6000 m (Rex et al., 2006). Yet, instances of depth-dependent bacterial abundance do arise at the regional or local scale (e.g., Quéric et al., 2004). The enduring explanation for variability in bacterial numbers (or biomass and activity), whether or not also conforming to depth dependency, is the amount of organic material arriving at the seafloor and its physical–chemical availability in situ (Deming and Yager, 1992; Boetius et al., 2000; Turley and Dixon, 2002; and citations therein).

The collaborative Deep Gulf of Mexico Benthos (DGoMB) project and its multi-year series of sampling cruises presented an unusual opportunity to obtain benthic bacterial data (where “bacteria” and its derivatives refer to prokaryotic cells detectable microscopically by DNA-specific stains) on a much more extensive spatial scale than other deep-sea studies have afforded. The semi-enclosed nature of the Gulf of Mexico, its hydrographic complexities, and important inputs from the Mississippi River are described elsewhere (Rowe et al., 2008), but these margin-based features distinguish the Gulf of Mexico from the open ocean at comparable depths. While the Gulf of Mexico seafloor may be well known for its hydrocarbon seeps, gas hydrates, and associated microbes (e.g., MacDonald et al., 1989; Orcutt et al., 2004), the bacterial component of the pelagic sediments that accumulate on its shelf, slope, and abyssal regions are virtually unstudied. An initial snapshot from a core taken on the Sigsbee Deep in the western Gulf of Mexico (station depth of 3650 m) indicated bacterial abundances and calculated contribution to total benthic biomass that were typical of other abyssal plains (Rowe et al., 2003), raising expectations that additional data, at least for abyssal depths, might also conform to existing knowledge.

A primary objective of the bacterial component of the DGoMB project was to quantify benthic bacterial abundance in the pelagic sediments at all stations surveyed, so that the size of these communities and their biomass contributions to the total benthos could be determined and compared among stations and to other parts of the ocean (see Rowe et al., 2008). We supplemented abundance data with site-specific measurements of cell size (allowing calculation of biovolume) to improve upon the conventional approach of estimating biomass, whereby a single conversion factor is applied uniformly to counts from all stations. We also conducted experimental work shipboard at four “process” stations to evaluate production and respiration, again with the aim to place bacterial results into broader context.

Other goals included a search for spatial and temporal correlates that might help to answer general and site-specific questions about benthic bacteria. Is water depth a primary determinant of the size of their communities? Does terrestrial or riverine input lead to gradients in their abundance along depth contours or transects across the slope (shown in Fig. 1)? Do the common mesoscale features on the northern slope (basins and canyons; Fig. 1), expected to trap organic detritus from the shelf area, support larger benthic bacterial communities than surrounding environs? Does the size of the community vary inter-annually? The availability of parallel data sets generated by collaborators also allowed us to search for sediment-based chemical or mineralogical determinants of benthic bacterial abundance, including oxygen penetration, organic content, and grain size analyses (see also Morse and Beazley, 2008). Although the DGoMB project did not include measures of the supply of particulate organic carbon (POC) to the seafloor, which entails a particularly complex set of sources and processes when a basin is both semi-enclosed and river-impacted, collaborators derived estimates of vertical POC flux from measures of primary production in the overlying waters (Biggs et al., 2008). The availability of these estimates allowed us to examine whether this component of the complex supply issue could help to explain the size of underlying benthic bacterial communities. These questions and issues were approached via statistical analyses of the bacterial data and tested factors (e.g., linear regressions, two-tailed t-tests, analysis of variance, and partial correlation analyses), from all stations or from subsets representing the question at hand, and by comparing our findings to similarly obtained results in the literature.