Microbes and the Dissipation of Energy and Respiration: from Cells

Home , Energy flow (ecology)

This article has been published in published been This has article or collective redistirbution other or collective or means this reposting, of machine, is by photocopy article only anypermitted of with portion the approval The O of A s e a o f M i cr o be s

> Section IV. Processes

Oceanography > Chapter 8. Energy Dissipation, Respir ation, and growth

, Volume 20, N Microbes and the Dissipation of Energy and Respiration:

umber 2, a quarterlyumber The O journal of From Cells to Ecosystems By Cr aig A . Carlson, Paul A . del Giorgio, and Gerhard J. Herndl

In the year 1974, Larry Pomeroy first chemistry is elucidated in studies of how In the contemporary aerobic ocean,

ceanography S proposed that microbes were true mov- the flow of energy through microbial the metabolic strategy for the vast ers of energy and nutrients in marine processes manifests itself as demand for majority of prokaryotes is chemohet- food webs (Pomeroy, 1974). This idea and recycling of elemental nutrients. erotrophy. Heterotrophic bacterio- ociety. Copyright 2007 by The O was later formalized as the “microbial Prokaryotes (bacteria and archaea) plankton are the major respirers (Sherr loop” (Azam et al., 1983; Pomeroy et al., are now recognized as the most abun- and Sherr, 1996; Rivkin and Legendre, this issue) in which energy and carbon dant living component of the biosphere 2001), with organic compounds serving lost from the planktonic food web in the with approximately 12 x 1028 cells found as both the electron donors as well as form of dissolved organic matter (DOM) in the oceanic water column (Whitman carbon sources, and oxygen functioning ceanography S was recovered and repackaged by hetero- et al., 1998). This prokaryotic biomass as electron acceptor. As organic matter ceanography S trophic bacterioplankton to particulate is comprised of vast phylogenetic diver- is catabolized, both organic matter and

ociety. A organic matter (POM). Their ecological sity (Venter et al., 2004; Giovannoni O2 are consumed, resulting in anabo- ociety. S role within the microbial loop is to facili- and Stingl, 2005; Moran et al., this lism (biosynthesis) and CO2 production. ll rights reserved. Permission is reserved. ll rights granted to in teaching copy this and research. for use article Republication, systemmatic reproduction, tate the transformation of DOM to POM issue; Edwards and Dinsdale, this issue, In terms of energy and carbon cycling end all correspondence to: [email protected] Th e O or (a trophic “link”) (Azam et al., 1983) or Breitbart et al., this issue) as well as within the ocean, it is these chemoor- to remineralize DOM back to its inor- metabolic diversity (King, 2005; DeLong ganotrophic organisms that are the most important physiological group. They are instrumental in the transformation of organic matter, its remineralization to Prokaryotes (bacteria and archaea) are inorganic constituents (Ducklow et al., now recognized as the most abundant Craig A. Carlson ([email protected]. ceanography S living component of the biosphere... edu) is Associate Professor, Marine Science Institute, University of California, Santa ociety, P Barbara, CA, USA. Paul A. del Giorgio ganic constituents (a respiratory “sink”) et al., 2006). Although prokaryotic pro- is Associate Professor, Département Box 1931, Rockville, MD 20849-1931, USA 1931, Rockville, O Box (Ducklow et al., 1986). These early stud- cesses and trophic interactions occur on des Sciences Biologiques, Université du ies laid the conceptual framework for the spatial scale of nanometers (Azam, Quebec à Montreal, Canada. Gerhard investigations that linked microbial pro- 1998), their sheer numbers and the J. Herndl is Department Head, Biological cesses to the flow of energy in marine rates at which they operate have major Oceanography, Royal Netherlands food webs (microbial ecology or tropho- biogeochemical implications on the Institute for Sea Research (Royal NIOZ), dynamics). The link to ocean biogeo- scale of ecosystems. Texel, The Netherlands. .

Oceanography June 2007 89 1986), and shaping of the organic and Substrate inorganic environment (Williams and I del Giorgio, 2005). These organisms and their associated processes will be the New cells focus of this article. II One of the fundamental proper- (biomass) ties that determines bacterioplankton’s ecological or biogeochemical role in the marine ecosystem is the amount of µ biomass produced per unit of organic a c1 C consumed, or the bacterial growth b efficiency (BGE) (Sherr and Sherr, c2 ATP 1996; del Giorgio and Cole, 1998). BGE c3 is a measure of the coupling between c4 d III catabolic (energy-yielding) reactions Products Cell or waste to anabolic (biosynthetic; energy- Storage requiring) reactions and is expressed by the formula, Figure 1. Substrate and energy flow within a cell. Substrate < 700 Da are actively taken up via membrane BGE = BP / (BP+BR) [1] proteins (I). As the substrate enters the cell via active uptake, it either enters into catabolic pathways (blue lines) or anabolic pathways (green lines). Monomers for anabolism can come preformed from the where BP is bacterial production and environment or as products of catabolism. The red-hashed lines represent the flow of energy to and from these metabolic pathways. Energy is conserved via substrate catabolism and ATP is produced at a rate a. BR is bacterial respiration. Here we As ATP is hydrolyzed, energy is released and utilized at rate b to drive anabolic processes such as produc- will examine how microbial energet- tion of new cells (growth; II) and production cell storage products (III). Energy is also utilized at various ics of heterotrophic bacteria partition rates to support processes that are independent of anabolism. This maintenance energy is used at rates c1 to activate uptake systems, c2 to fuel cell motility, c3 to actively eliminate waste, and c4 to repair cellular energy and carbon on a cellular level machinery. In the absence of exogenous organic substrates, the cell can yield ATP at rate d by catabolizing (Figure 1), how that partitioning affects storage material (endogenous substrates). Adapted from del Giorgio and Cole (1998) their growth efficiency, and how that growth efficiency affects their ecological and biochemical roles in the sea. et al., this issue, for details on these two through a membrane transport system Cellular Bioenergetics metabolic strategies), or the oxidation of to an electron acceptor. The generated The first law of thermodynamics states organic compounds (chemoorganotro- chemical energy is used to drive the cells’ that the total amount of energy in the phy). ATP is synthesized via one of three metabolic processes (Jones, 1983). universe is conserved and cannot be distinct mechanisms: substrate-level The second law of thermodynamics created nor destroyed. As a result, all phosphorylation (fermentation), pho- states that in all processes or reactions, organisms have evolved physiological tophosphorylation, or oxidative phos- some of the energy involved irreversibly strategies to conserve energy by col- phorylation (respiration). Except for loses its ability to do work as a system lecting, converting, and storing that obligate fermenters, all microbes carry moves from order to disorder (entropy). energy principally via the synthesis out respiration. On the cellular level, From the standpoint of energy flow of adenosine-5’-triphosphate (ATP). respiration is the key process of energy within a cell, the hydrolysis of ATP trans- Energy is harvested from light (photot- conservation in which an electrochemi- fers the energy conserved from catabolic rophy), the oxidation of inorganic com- cal potential is generated from the flow reactions to energy-requiring reactions pounds (chemolithotrophy) (see Kolber of electrons from reduced compounds that support anabolism (Figure 1).

90 Oceanography Vol. 20, No. 2 However, cells expend energy in ways polysaccharide capsule might act as Maintenance Metabolism that are independent of cell biomass pro- sorption/scavenging site for organic A more ecologically important mecha- duction via processes such as overflow molecules for subsequent cleavage by nism in natural systems is the free metabolism, futile cycles, and main- ectoenzymes and is constantly renewed energy allocated to nongrowth reac- tenance metabolism (see Russell and (Stoderegger and Herndl, 1998). Hence, tions. Known as maintenance energy, it Cook, 1995) (Figure 1). Thus, efficiency heterotrophic microbial processes might serves to keep entropy low by fixing what of energy use with regard to biomass be important sources of recalcitrant breaks within the cell; thus, it helps a cell production (growth efficiency) can vary DOM (Tanoue et al., 1995; McCarthy to maintain its molecular, cellular, and significantly depending on the physical- et al., 1998). The pool of recalcitrant or functional integrity (Hoehler, 2004). The chemical state of the environment. semi-labile DOM is biologically reactive. larger the partitioning of energy into However, its production and subsequent maintenance processes, the lower the Overflow Metabolism remineralization are uncoupled for time energetic efficiency of the cell. Studies of Laboratory cultures of bacteria grown scales of months to decades resulting in bacterial growth in pure bacterial cul- in energy and substrate-rich media accumulation (Carlson, 2002). tures demonstrate that rates of catabo- often display an uncoupling of catabolism from anabolism, resulting in nongrowth energy dissipation rather than maintenance energy expenditure (see Teixeira-de-Mattos and Neijssel, 1997; Liu, 1998, and citations therein). Under this scenario, energy generated from catabolism is in excess of that needed for anabolism; thus, the cell wastes ATP via extracellular release of DOM. It is not known to what extent energy dissipation via overflow metabolism occurs in natural systems, but several studies have demonstrated production of recalcitrant DOM in cultures of natural assemblages of bacterioplankton (Brophy and Carlson, 1989; Heissenberger and Herndl, 1994; Ogawa et al., 2001; Kawasaki and Benner, 2006). Little is known about why free-living bacterioplankton would actively release DOM extracellularly in substrate-limited systems like the open ocean. It has been shown that highly metabolically active bacterioplankton are engulfed by a polysaccharide envelope (Heissenberger et al., 1996) (Figure 2) in which a major Figure 2. TEM images of the most common types of capsules found in free-living marine bacteria. Note that the morphology of the capsules varies considerably. D and part of the cell’s ectoenzymes is embed- F (a dividing cell) were observed near marine snow particles. Scale bars: A and B— ded (Martinez and Azam, 1993). This 100 nm, C to F—200 nm. TEM pictures adapted from Heissenberger et al. (1996)

Oceanography June 2007 91 lism are independent of anabolism and appear to hold across large temperature greater catabolism of DOC in order proceed at maximal rates (Russell and ranges, the relationship is weak at any to balance elemental requirements of Cook, 1995). Maintaining maximal rates given site in which temperature range is the cell. Studies of natural assemblages of energy flow within a cell keeps cellu- < 10° C. For example, a large systematic grown on media enriched with single C lar membranes energized and the active increase in BGE (i.e., < 10% to > 35%) or N compounds generally support this transport system functional. This “ener- was observed throughout the course trend (Goldman et al., 1987). However, gized” physiological state is advanta- of a phytoplankton bloom in the Ross in culture studies where multiple C or N geous for cell growth, leaving cells poised Sea, Antarctica, despite little change in sources were available, no trend between to exploit transient but favorable envi- temperature (-2°C to +2°C) (Carlson BGE and substrate stoichiometry was ronmental conditions associated with and Hansell, 2003). In the southern observed (Goldman and Dennett, 2000). patchiness of available organic and inor- North Sea, Reinthaler and Herndl (2005) In natural systems, no discernable trend ganic nutrients (del Giorgio and Cole, observed the opposite trend in which is observed between BGE and DOM 1998). Thus, while cells sacrifice thermo- BGE increased from winter through stoichiometry largely because the actual dynamic efficiency, their BGE appears to spring as temperatures warmed. The stoichiometric ratio of the available be optimal to support maximal growth rise in BGE observed in these stud- DOM is unknown. In addition, natural rates (Westernhoff et al., 1983). ies suggests that other factors, such as assemblages of prokaryotes likely use DOM bioavailability during a phyto- multiple sources of organic C, N, and Factors That Can Affect BGE plankton bloom, had a greater effect P as well as inorganic N and P, making Abiotic Factors relative to temperature. understanding the relationship among Physical factors that hasten biochemical Additionally, abiotic factors may substrate stoichiometry, BGE, and breakdown within a cell will affect the affect bacterial energetic constraints. microbial growth extremely complex maintenance energy requirement and For example, ultraviolet irradiation, in natural systems (Vallino et al., 1996). lower BGE. In a comprehensive review exposure to toxic substances, or osmotic Other studies demonstrate a more direct of BGE, Rivkin and Legendre (2001) shock may result in large increases in relationship between nutrient limitation suggest a significant inverse relationship cell respiration associated with protec- and BGE. Tortell et al. (1999) show that between temperature and BGE, based on tion and repair, and consequent declines under iron deficiency, electron-transport observations that the fraction of respired in BGE and growth (Koch, 1997). activities associated with respiration assimilated carbon is greater at lower Although these effects have been dif- decrease, resulting in a significant reduc- latitudes compared to higher latitudes. ficult to document for marine bacterio- tion in BGE. Addition of inorganic phos- Because there are major environmental plankton in situ, there is some evidence phorus has been reported to increase gradients superimposed on this latitudi- of large declines in BGE in freshwater microbial respiration in oligotrophic nal range, it is possible that the pattern to saltwater transition zones in estuaries waters (Obernosterer et al., 2003). reported by Rivkin and Legendre (2001) (del Giorgio and Bouvier, 2002). may result from factors other than tem- Lability (Availability) of DOM perature. However, Apple et al. (2006) Nutrient Limitation Not all microbes can break down all also observed a negative relationship The stoichiometry of bacterioplank- DOM (Floodgate, 1995). The oceanic between temperature and BGE in estua- ton is relatively constant (i.e., C:N ratio DOM pool represents a continuum of rine and coastal sites, and experimentally ≈ 4–6) (Goldman et al., 1987). The C:N biological lability, ranging from refrac- confirmed the temperature effect on and C:P ratios of oceanic DOM range tory material that turns over on time BGE. The influence of temperature on from 9–18 and 180–570, respectively scales of centuries to millennia (Williams bacterial carbon metabolism is complex, (Benner, 2002). If the substrate sup- and Druffel, 1987; Bauer et al., 1992) and not all reports agree with the pat- porting bacterial growth is relatively to more available material turning terns described above. Although trends C-rich, one might expect to observe over on time scales of minutes to days

92 Oceanography Vol. 20, No. 2 (Cherrier et al., 1996; Carlson, 2002). to maintain high electrochemical poten- and Cole, 1998, and citations therein). The bulk DOM pool is often concep- tial and support transport systems for Bacteria consume a wide range of tually partitioned into broad pools of other limiting nutrients. organic compounds simultaneously, and lability, increasing in concentration Although LMW monomers are more although the nature of the organic mat- from a “labile,” to a “semi-labile,” to a easily taken up relative to polymeric ter consumed in the oceans is not well “refractory” pool (see Carlson, 2002, and material, not all LMW DOM is available known, there is evidence that very different pools are used. For example, Cherrier et al. (1999) demonstrated that at least a portion of the carbon used by oceanic bacteria is old (> 500 years), and thus Although prokaryotic processes occur on probably differs in its energetic contents the spatial scale of nanometers, their from recently fixed carbon. The relative importance of these different pools of sheer numbers and the rates at which DOM lability likely plays a major role in they operate have major implications shaping oceanic BGE patterns. on the scale of ecosystems. Viral Activity Phage infection can contribute up to 50% of bacterial mortality. Upon lytic infection, cells burst, spilling viruses and citations therein). Bacteria are limited for uptake. Indeed, the greater major- organic matter into their environment. to transporting low molecular weight ity of oceanic DOM is uncharacter- As the viral-mediated DOM release is DOM (LMW DOM < 700 Da [dalton ized LMW material that is resistant to reincorporated into bacterioplankton, = atomic mass unit]) via permeases microbial attack (Benner, 2002). Amon bacterial respiration increases by as (enzymes that transport other substances and Benner (1994, 1996) propose the much as 30% (Fuhrman, 1999). Viral across cell membranes); thus, labile com- size-continuum model in which the high infection can also decrease the BGE pounds such as dissolved free neutral molecular weight DOM is more bioreac- of noninfected bacteria as a result of sugars or dissolved free amino acids are tive and available to microbial utilization the increased energy demands associ- assimilated easily by the cell’s uptake than the majority of the LMW DOM. ated with the degradation of polymeric mechanisms (Keil and Kirchman, 1999). However, enhanced bioavailability does organic nitrogen and phosphorus lysates However, efficient DOM uptake does not necessarily equate to enhanced BGE. (Middelboe et al., 1996). not always equate to increased BGE. For Hydrolysis of large polymeric molecules example, in experiments where natural and subsequent DOM uptake can be Bacterial Diversity assemblages of oligotrophic bacterio- energetically costly and offset the nutri- Significant diversity of prokaryotic plankton were amended with labile car- tional advantage resulting in low BGE. communities exists in oceanic systems. bon in the form of glucose, Carlson et al. Vertical stratification of major prokary- (2002) observed that as much as 85% of Energetic Quality otic groups between the oceanic eupho- the glucose carbon was respired as CO2 DOM quality can also refer to the bio- tic and aphotic zones has been observed within a few days. The rapid uptake logically available energy from DOM (Giovannoni et al., 1996; Karner et al., and subsequent remineralization of oxidation. Growth efficiency and bio- 2001; Moeseneder et al., 2001; Morris this labile DOC resulted in only a small mass production of organisms grown on et al., 2002). The mechanisms respon- increase in cell production, suggesting a relatively oxidized substrate will be low sible for spatial variability of specific that most of available DOC was utilized even if the supply is high (del Giorgio prokaryotic groups or phylotypes over

Oceanography June 2007 93 depth are not well understood. However, energy for respiration, maintenance, and origin, increased overall environmental conceptual models provide compel- active growth. However, other laboratory hostility will result in an increase in the ling reasons to believe that the growth studies with the alpha-Proteobacteria proportion of the total flux of energy of specific heterotrophic microbial Candidatus Pelagibacter ubique were that is devoted to cell maintenance (EM). populations in planktonic ecosystems is unable to resolve differences in growth Associated with this increase in cell main- linked to the composition and amount response between cultures incubated in tenance, it is expected that cell-specific of DOM, as well as to the availability the light versus the dark using nutrient- respiration (SP) (standardized for differ- of inorganic nutrients. enriched cultures (Giovannoni et al., ences in cell size among systems) should Recent work in the North Sea indi- 2005). Thus, light-mediated energy fixa- also increase to fuel the increased main- cates bacterial diversity can affect the tion could potentially increase the base- tenance and repair. BGE will therefore BGE of the overall bacterial assemblage. line BGE in the illuminated layers of the decrease along this gradient of increasing Reinthaler et al. (2005) demonstrate oceans under oligotrophic conditions environmental hostility. Figure 3 is hypo- that while the function of DOM rem- by supplying additional energy to that thetical and the exact shapes of these ineralization to CO2 remains stable derived from organic substrates, but the curves are not known, but the ranges of despite shifts in bacterial assemblage factors that control photoheterotrophy BGE shown in the figure are realistic and structure, the BGE is inversely related to and its impact on the ocean carbon cycle are within what has been reported for bacterioplankton richness. This inverse are not yet fully understood. marine systems, suggesting that marine relation between BGE and bacterial bacterioplankton may experience a wide richness is largely mediated by bacte- Ecosystem range in overall environmental hostility. rial production, which is generally more and Biogeochemical On the ecosystem level, microbial variable than respiration. Implications oceanographers are interested in how As discussed above, the allocation of the hostility of the marine environment Interaction with Other carbon and energy in marine bacteria affects microbial energy partitioning Metabolic Pathways depends on many factors, and it is dif- and respiration and how the integra- The recent discovery of the widespread ficult, if not impossible, to place the tion of all these individual cell processes occurrence of rhodopsin-like pigments variation of BGE as a function of a single affects the magnitude and efficiency by in a variety of heterotrophic marine variable. In general terms, however, BGE which organic matter flows through bacterial groups (see Kolber et al., this will tend to be low in situations of either heterotrophic bacterioplankton. In issue) suggests that nonphotosynthetic insufficient supply of energy and nutri- practice, researchers have taken various light utilization may be a common fea- ents, of inappropriate stoichiometry approaches to determine the magni- ture among marine heterotrophs in of resource supply, or under increased tude of organic matter flux through the surface waters. These light-harvesting maintenance and repair requirements oceanic bacterial component, including pathways, which result in energy fixa- of the cells. These environmental fac- measurements of growth, biomass pro- tion but are not necessarily coupled to tors may be summarized as the level of duction, respiration, or a combination carbon or oxygen dynamics, could play environmental “hostility” (Figure 3). of the latter. a significant role in shaping microbial This hostility may result from extreme bioegenergetics and thus influence the lack of resources (i.e., organic matter Indirect Estimates of Bacterial patterns in BGE. Gómez-Consarnau et or nutrients), overabundance of others Carbon Demand al. (2007) demonstrate that proteorho- (i.e., pollutants, allelopathic substances), Estimates of BP via uptake of radioac- dopsin-mediated photoheterotrophy physical or chemical stresses (i.e., salin- tive tracers are by far the most common significantly stimulates production of ity and temperature gradients, UV measurements of bacterial metabolism some marine Flavobacteria under oli- radiation, extreme pH gradients), or a and are routinely carried out in marine gotrophic conditions, thus supporting combination of these. Regardless of its studies. There is thus a wealth of data

94 Oceanography Vol. 20, No. 2 Further, a recent quantitative review of published marine BGE measurements concludes that the median BGE for the surface open ocean is 8% and for the coastal ocean is 16% (Robinson, in press). Within the ocean interior, estimates of BGE are even lower. Carlson et al. (2004) estimate BGE to be 8 ± 4% for bacterioplankton in the upper mesopelagic zone after convective overturn, and Reinthaler et al. (2006) find BGE to be < 2% in the bathypelagic zone deeper than 1000 m. The finding that BGE is generally low but highly variable has both conceptual and practical consequences. From a con- Figure 3. Conceptual diagram demonstrating the relationship between environmental stressors ceptual point of view, these field obser- or environmental “hostility” and the partitioning of energy within a bacterial cell, the result- vations support the hypothesis that, in ing bacterial growth efficiency (BGE), and cell specific respiration. As environmental hostility increases, more energy is partitioned into maintenance energy (EM). Thus, bacterial growth dynamic environments under condition efficiency decreases and cell-specific respiration (SP) increases. Some combination of both of low nutrient supply, microbes have physical (temperature, pH, salinity) and chemical (toxins, substrate availability) factors contrib- evolved versatile metabolic machinery to ute to environmental hostility take advantage of multiple nutrient, carbon, and energy sources. By increasing the capacity of enzymes in the uptake/ from a wide range of oceanic regions thus total carbon consumption (BR plus assimilation systems, a larger percentage (Ducklow and Carlson, 1992; Ducklow, BP), by combining measurements of BP of energy is funneled into maintenance 1999). Cole et al. (1988) first reported with estimates of BGE, such that, processes to safeguard metabolic flex- that, in the euphotic zone, BP was ibility at the cost of energetic efficiency BCD = BP/ BGE [2] positively and significantly correlated (Teixeira-de-Mattos and Neijssel, 1997). to primary production (PP) across a One of the main findings in the past From a practical point of view, a lower wide range of aquatic ecosystems, and decade is that marine BGE is both far BGE implies a relatively larger flux of concluded that net BP averaged about more variable and much lower than carbon to support an observed BP; thus, 30% of PP. In similar analyses of seven what was traditionally assumed in mod- BCD can be a greater percentage of PP in oceanic sites, Ducklow (1999) concluded els of bacterial carbon consumption, low-productivity systems. that BP was < 20% of PP. as hypothesized by Jahnke and Craven Interestingly, this approach gener- Although the magnitude of net BP (1995) over a decade ago. For example, ally leads to estimates of BCD in the appears modest compared to PP, one there is growing evidence from field open-ocean euphotic zone that are often must consider the efficiency by which experiments that BGE generally increases similar to or greater than the estimates bacterioplankton convert DOC into along trophic gradients, from < 10% of local PP (Ducklow, 1999; del Giorgio bacterial biomass. The flux of carbon in oligotrophic oceanic water to > 20% et al., 1997) when PP values are low needed to support a given estimate of BP in the most productive environments (Figure 4). Figure 4 shows integrated is referred to as bacterial carbon demand (Ducklow and Carlson, 1992; del Giorgio euphotic zone BCD estimates in rela- (BCD). It is possible to derive BCD, and and Cole, 1998; Biddanda et al., 2001). tion to integrated PP. These data dem-

Oceanography June 2007 95 10000 the conversion factors needed to estimate BATS cell or carbon production (Ducklow Palmer LTER and Carlson, 1992; Ducklow, 2000). 1000 Equatorial Paciﬁc Measurements of BGE also have their

) Ross Sea -1

d own set of shortcomings due to techni- -2

m cal difficulties (Robinson and Williams, 100 2005). Thus, it is clear that our understanding of BGE and, consequently, BCD (mg C 10 estimates of BCD in the oceans is still very uncertain, and that direct measurements of BR and total community res- 1 piration are required to better constrain 1 10 100 1000 10000 PP (mg C m-2d-1) the flux of energy and organic matter in oceanic systems. Figure 4. Relationship between integrated bacterial carbon demand (BCD) and primary production (PP) within the euphotic zone of representative ocean sites. All data are derived from paired measurements of bacterial production and PP integrated within the euphotic Direct Measurements of zone from each study site. A common bacterial growth energy of 0.1 was used to estimate Bacterial Respiration BCD. The black line represents the 1:1 line. Data points that lie above this line indicate that bacterial carbon demand was greater than local primary production at the time of sample Bacterial respiration still represents a collection. The Bermuda Atlantic Time-series Study (BATS) data represent monthly values major technical challenge and in spite of from 1991–2003 (n = 155; see Steinberg et al., 2001 for details; data available at http://bats. bbsr.edu/). Paired BP and PP from the Equatorial Pacific (n=16) and Ross Sea, Antarctica, its obvious importance, there is a paucity (n=77) were calculated according to Ducklow (1999) (data available at http://usjgofs.whoi. of direct BR measurements. The avail- edu/jg/dir/jgofs/). All data from the Palmer Peninsula, Antarctica, (n=112) provided by able oceanic BR data set is extremely H. Ducklow and the Palmer LTER program. modest, especially compared to existing measurements of other carbon fluxes, such as bacterial and algal production. A onstrate that when PP is > 1 g m-2 d-1, (POC) flux into the interior. They find recent synthesis concluded that there are BCD was significantly less than local PP. that the gap between BCD and POC sup- fewer than 500 individual measurements However, when PP is < 1 g m-2 d-1, BCD ply lead to a deficiency of as much as for the global euphotic ocean, which can be similar to or greater than local an order of magnitude at depths greater result in a median (mean + standard estimates of PP, indicating an apparent than 1000 m. deviation) BR of 0.5 mmol C m-3 d-1 mismatch between what bacteria con- The apparent imbalance that results (1.3 + 2.3 mmol C m-3 d-1, n=105) and sume and what appears to be available from combining current measure- 3.1 mmol C m-3 d-1 (10.5 + 20.9 mmol for consumption. ments of BP and BGE and comparing C m-3 d-1, n=332) for the open ocean and Waters deeper than 200 m comprise the resulting carbon consumption with coastal regions, respectively (Robinson ≈ 75% of the volume of the global ocean, primary production or POC flux (in in press). These rates are very significant yet little is known about the microbial the case of deeper waters) may be either relative to other major carbon fluxes, processes that persist at depth. Recent a real feature of oceanic systems or the such as primary production and carbon studies demonstrate enhanced prokary- result of current uncertainties in the export, as we discuss below. otic activity and BCD within the ocean’s metabolic measurements. Routine mea- Despite the modest size of the avail- mesopelagic and bathypelagic zones surements of BP via the thymidine or able database, these direct measure- (Arístegui et al., 2005; Reinthaler et al., leucine uptake methods are only a proxy ments of BR tend to support the patterns 2006). Reinthaler et al. (2006) compare for bacterial production; accurate esti- obtained by combining BP and BGE: BCD with particulate organic carbon mates are confounded by variability in bacterial carbon consumption is high,

96 Oceanography Vol. 20, No. 2 particularly in olgotrophic systems, and complicated and potentially weakened otrophy is counter intuitive and contro- often exceeds local estimates of primary due to temporal uncoupling between versial (Williams et al., 2004). production. The existing data further respiration and primary production To test the hypothesis of net heterot- suggest that bacteria are the largest con- (Billen, 1990) or large transport of rophy within the euphotic zone of the tributors to community respiration in organic matter between ocean sectors oligotrophic gyres and address some the oceanic system. Within the eupho- (Hansell et al., 2004), and can result in criticism of temporal- and spatial- tic zone of marine systems, BR repre- an unbalance between respiration and scale problems from previous studies, sents a large fraction (≈ 50% to > 90%) local production sources. Better under- Williams et al. (2004) designed a year- of community respiration (Sherr and standing of oceanic respiration will ulti- long investigation of metabolic balance Sherr, 1996; Biddanda et al., 2001; mately help oceanographers assess the at the Hawaiian Ocean Time-series Rivkin and Legendre, 2001; Robinson magnitude of organic carbon input and (HOT) site (Station ALOHA) in the oli- and Williams, 2005). help interpret measurements of primary gotrophic North Pacific. They directly Other studies using estimates of res- production (Williams et al., 2004). measured respiration and primary pro- piration from changes in O2 in bottle Neither bacterial nor community duction by monitoring O2 changes in experiments also report that portions respiration within any ecosystem can light/dark bottles incubated for 24 hours of the surface oligotrophic open ocean exceed the supply of organic carbon. in vitro over six depths within the appear to be net heterotrophic (Duarte In circumstances where bacteria or the euphotic zone. Their integrated results and Agusti, 1998). But, what about meta- net community respire more than is indicate a metabolic deficit equivalent to bolic balance in the open sea? The appar- produced by local photosynthesis (net 40% of measured primary production. ent discrepancy is interesting and con- heterotrophy), there must be an alloch- troversial (Geider, 1997; Williams, 1998; thonous (outside the system) input of Discrepancy with Williams and Bowers, 1999; Goldman organic matter. Because all ecosystems Geochemical Estimates and Dennett, 2000), but perhaps pro- are open to exchange of organic matter, Evidence from several independent vides insight to better interpret C and local environments may not be in meta- geochemical measures all point to net energy flow in the open sea. We explore bolic balance. It is easy to envision how organic carbon production even in oli- recent findings and evaluations below. external input of organic matter from gotrophic sites (Doney, 1997). For exam-

Metabolic Balance Current evidence suggests that the total carbon consumption by hetero- Existing data...suggest that bacteria are trophic bacteria represents one of the the largest contributors to community largest components of the marine biological carbon budget, and that much respiration in the oceanic system. of this carbon is respired by bacteria. Respiration, on the ecosystem level, represents the largest sink of organic matter in the biosphere and accounts for the terrestrial runoff or advected organic ple, geochemical measurements at the total amount of organic matter oxidized, matter from coastal blooms can subsi- two most intensively studied oligotro- oxygen consumed, and CO2 produced. dize respiration in estuarine and coastal phic ocean sites at the Bermuda Atlantic On a global scale, respiration must be systems. But for open-ocean systems like Time-series Study (BATS) and HOT balanced by the input of organic mat- oligotrophic gyres that are physically iso- indicate annual patterns of O2 supersatu- ter via primary production. On the local lated from allochthonous inputs of labile ration in the euphotic zone (Steinberg scale, this relationship can become more organic matter, the notion of net heter- et al., 2001; Karl et al., 2003), O2 efflux to

Oceanography June 2007 97 the atmosphere (Emerson et al., 1997), cal estimates as evidence of net organic integrated net organic production than net export of sinking organic particles matter production in the euphotic zone, photosynthesis (Williams et al., 2004; from the surface waters (Karl et al., 1996; Williams et al. (2004) explore alternative del Giorgio and Williams, 2005). Because Steinberg et al., 2001), vertical export of possibilities to explain the discrepancy. of its great integrating properties, micro- DOC from the surface 100 m (Carlson et Rates of respiration are relatively con- bial oceanographers should include al., 1994), and long-term accumulation stant compared to the large degree of more regular measurements of respira- of DOM in the surface waters (Church primary productivity observed in aquatic tion in future studies to better constrain et al., 2002). All of these independent ecosystems (del Giorgio and Williams, the magnitude of energy flow within observations require net organic carbon 2005). Part of this constancy is due to the oceans, and to better understand production to reconcile the data. the fact that community respiration inte- the potential mechanisms that control grates all components of the food web it. Bacterioplankton are responsible for How Do We Explain the Metabolic and the various organic pools each com- a large fraction of oceanic respiration, Imbalance of Aquatic Ecosystems ponent utilizes. In contrast, autotrophic and therefore a better understanding Indicated from These Approaches? production is more episodic and can of the controls of bacterial carbon con- Granted, there are errors associated with become temporally decoupled from res- sumption, and of the partition of this the various approaches used to estimate piration (Karl et al., 2003). These pulses carbon between anabolic and catabolic community respiration, including in of energy and carbon from episodic pathways, will help to better understand vitro O2 rate measurements or microbial events will be dampened as they flow global ocean respiration. respiration from BGE and BP (Robinson through the heterotrophic component of and Williams, 2005). Nonetheless, these the food web, thus yielding a more con- Acknowledgements data are still valuable and provide sig- stant respiration signal. As a result, if the We would like to thank H. Ducklow nificant insight to our interpretation of system is undersampled through time, for contributing unpublished data metabolic balance. In their 2005 paper, the production term would be underesti- from the Palmer LTER program and del Giorgio and Williams state, “In real- mated relative to consumption, yielding for discussions that helped improve ity there is essentially no imbalance, an apparent metabolic deficit (Karl et al., this manuscript. This review was sup- because aquatic ecosystems consume 2003; Williams et al., 2004). ported by grants from the U.S. National and respire all but a small amount of Currently, the oceanographic bench- Science Foundation. organic matter available. The apparent mark for assessing the magnitude of imbalances are artifacts that result from energy and carbon that flows through References our largely insufficient understanding an oceanic ecosystem is set by measure- Amon, R.M.W., and R. Benner. 1994. Rapid cycling of high-molecular-weight dissolved organic of the magnitude and regulation of total ments of primary production. However, matter in the ocean. Nature 369:549-552. organic matter flux in aquatic ecosys- mass balance and inverse modeling Amon, R.M.W., and R. Benner. 1996. Bacterial tems.” Perhaps these apparent deficits approaches suggest that rates of organic utilization of different size classes of dissolved organic matter. Limnology and Oceanography tell us something about the approach we matter production may be underesti- 41(1):41–51. use to assess carbon and energy flow in mated by a factor of two or more for Apple, J., P.A. del Giorgio, and M. Kemp. 2006. The temperature-dependence of bacterioplankton oceans and its interpretation. ecosystems like the subtropical North carbon metabolism. Aquatic Microbial Ecology Geochemical in situ observations Pacific gyre (Williams et al., 2004) or the 43:243–254. provide better spatial and temporal global ocean (Robinson and Williams, Arìstegui, J., S. Agustì, J.J. Middelburg, and C.M. Duarte. 2005. Respiration in the mesopelagic averaging compared to in vivo incuba- 2005). Respiration is equal to the source and bathypelagic zones of the ocean. Pp. 181– tion approaches and are considered to of organic input and integrates over 205, in Respiration in Aquatic Ecosystems. P.A. have greater accuracy for estimating net greater temporal scales; thus, in an del Giorgio and P.J. leB. Williams, eds, Oxford University Press. organic carbon production (Williams undersampled system, it has the poten- Azam, F. 1998. Microbial control of oceanic carbon et al., 2004). Accepting the geochemi- tial to be a more accurate measure of flux: The plot thickens. Science 280:694–696.

98 Oceanography Vol. 20, No. 2 Azam, F., T. Fenchel, J.G. Field, J.S. Gray, L.A. production in fresh and saltwater ecosystems: A Ecology 9:3–11. Meyer-Reil, and F. Thingstad. 1983. The ecolog- cross-system overview. Marine Ecology Progress Fuhrman, J.A. 1999. Marine viruses and their ical role of water-column microbes in the sea. Series 43:1—10. biogeochemical and ecological effects. Nature Marine Ecology Progress Series 10:257–263. Church, M.J., H.W. Ducklow, and D.M. Karl. 399:541–548. Bauer, J.E., P.M. Williams. and E.R.M. Druffel. 2002. Multi-year increases in dissolved organic Geider, R.J. 1997. Photosynthesis or planktonic 1992. 14C activity of dissolved organic car- matter inventories at Station ALOHA in the respiration. Nature 388(6638):132. bon fractions in the north-central Pacific and North Pacific Subtropical Gyre. Limnology and Giovannoni, S.J., L. Bibbs, J.C. Cho, M.D. Stapels, Sargasso Sea. Nature 357:667–670. Oceanography 47(1):1–10. R. Desiderio, K.L. Vergin, M.S. Rappé, S. Laney, Benner, R.H. 2002. Chemical composition and del Giorgio, P.A., and T.C. Bouvier. 2002. Linking L.J. Wilhelm, H.J. Tripp, and others. 2005. reactivity. Pp. 59–90 in Biogeochemistry of the physiologic and phylogenetic successions in Proteorhodopsin in the ubiquitous marine bac- Marine Dissolved Organic Matter. D.A. Hansell free-living bacterial communities along an estu- terium SAR11. Nature 438(7064):82–85. and C.A. Carlson, eds, Academic Press, San arine gradient. Limnology and Oceanography Giovannoni, S.J., M.S. Rappé, K. Vergin, and N. Diego, California. 47:471–486. Adair. 1996. 16S rRNA genes reveal strati- Biddanda, B., M. Ogdahl, and J. Cotner. 2001. del Giorgio, P., and J.J. Cole. 1998. Bacterial growth fied open ocean bacterioplankton populations Dominance of bacterial metabolism in oligotro- efficiency in natural aquatic systems. Annual related to the green non-sulfur bacteria phylum. phic relative to eutrophic waters. Limnology and Review of Ecological Systems 29:503–541. Proceedings of the National Academy of Sciences Oceanography 46(3):730–739. del Giorgio, P.A., J.J. Cole, and A. Cibleris. 1997. of the United States of America 93:7,979-7,984. Billen, G. 1990. Delayed development of bacterio- Respiration rates in bacteria exceed phytoplank- Giovannoni, S.J., and U. Stingl. 2005. Molecular plankton with respect to phytoplankton: A clue ton production in unproductive aquatic sys- diversity and ecology of microbial plankton. to understanding their trophic relationship. tems. Nature 385:148–151. Nature 437(7057):343–348. Ergebnisse der Limnologie 34:191–201. del Giorgio, P.A., and P.J.leB. Williams. 2005. The Goldman, J., D.A. Caron, and M.R. Dennett. 1987. Brophy, J.E., and D.J. Carlson. 1989. Production of global significance of respiration in aquatic eco- Regulation of gross growth efficiency and biologically refractory dissolved organic car- systems: From single cells to the biosphere. Pp. ammonium regeneration in bacteria by sub- bon by natural seawater microbial populations. 267–303 in Respiration in Aquatic Ecosystems. strate C:N ratio. Limnology and Oceanography Deep-Sea Research 36(4):497–507. P.A. del Giorgio and P.J.leB. Williams. eds, 32(6):1,239–1,252. Carlson, C.A. 2002. Production and Removal Oxford University Press, New York. Goldman, J.C., and M.R. Dennett. 2000. Growth Processes. Pp. 91–151 in Biogeochemistry of DeLong, E.F., C.M. Preston, T. Mincer, V. Rich, of marine bacteria in batch and continuous Marine Dissolved Organic Matter. D.A. Hansell S.J. Hallam, N.U. Frigaard, A. Martinez, M.B. culture under carbon and nitrogen limitation. and C.A. Carlson, eds, Academic Press, San Sullivan, R. Edwards, B.R. Brito, and others. Limnology and Oceanography 45(4):789–800. Diego, CA. 2006. Community genomics among stratified Gómez-Consarnau, L., J.M. González, M. Coll- Carlson, C.A., H.W. Ducklow, and A.F. Michaels. microbial assemblages in the ocean’s interior. Lladó, P. Gourdon, T. Pascher, R. Neutze, C. 1994. Annual flux of dissolved organic carbon Science 311(5760):496–503. Pedós-Alió, and J. Pinhassi. 2007. Light stimu- from the euphotic zone in the Northwestern Doney, S.C. 1997. The ocean’s productive deserts. lates growth of proteorhodopsin-containing Sargasso Sea. Nature 371:405–408. Nature 389:1,997. marine Flavobacteria. Nature 445:210–213. Carlson, C.A., and D.A. Hansell. 2003. The Duarte, C.M., and S. Agusti. 1998. The CO2 balance Hansell, D.A., H.W. Ducklow, A.M. Macdonald, Contribution of dissolved organic carbon and of unproductive aquatic ecosystems. Science and M. O’Neil Baringer. 2004. Metabolic poise nitrogen to biogeochemistry of the Ross Sea. 281(5374):234–236. in the North Atlantic Ocean diagnosed from Pp. 123–142 in Biogeochemical Cycles in the Ducklow, H.W. 1999. The bacterial component organic matter transports. Limnology and Ross Sea. Antarctic Research Series Volume of the oceanic euphotic zone. FEMS Microbial Oceanography 49(4):1,084–1,094. 78, G. DiTullio and R. Dunbar, eds, American Ecology 30:1–10. Heissenberger, A., and G.H. Herndl. 1994. Geophysical Union Press, Washington, D.C. Ducklow, H. 2000. Bacterial production and bio- Formation of high molecular weight material Carlson, C.A., S.J. Giovannoni, D.A. Hansell, S.J. mass in the ocean. Pp. 85–120 in Microbial by free-living marine bacteria. Marine Ecology Goldberg, R. Parsons, and K. Vergin. 2004. Ecology of the Oceans. D.L. Kirchman, ed., Progress Series 111:129. Interactions between DOC, microbial pro- Wiley-Liss, Inc., New York. Heissenberger, A., G.G. Leppard, and G.J. Herndl. cesses, and community structure in the meso- Ducklow, H.W., and C.A. Carlson. 1992. Oceanic 1996. Relationship between the intracel- pelagic zone of the northwestern Sargasso Sea. bacterial production. Pp. 113–181 in Advances lular integrity and the morphology of the Limnology and Oceanography 49:1,073–1,083. in Microbial Ecology, Volume 12. K.C. Marshall, capsular envelope in attached and free-living Cherrier, J., J.E. Bauer, and E.R.M. Druffel. 1996. ed., Plenum Press, New York. marine bacteria. Applied and Environmental Utilization and turnover of labile dissolved Ducklow, H.W., D.A. Purdie, P.J.leB. Williams, and Microbiology 62:4,521–4,528. organic matter by bacterial heterotrophs in J.M. Davies. 1986. Bacterioplankton: A sink for Hoehler, T.M. 2004. Biological energy requirements eastern North Pacific surface waters. Marine carbon in a coastal marine plankton commu- as quantitative boundary conditions for life in Ecology Progress Series 139:267–279. nity. Science 232:865–867. the subsurface. Geobiology 2:205–215. Cherrier, J., J.E. Bauer, E.R.M. Druffel, R.B. Coffin, Emerson, S., P. Quay, D.M. Karl, C. Winn, L. Tupas, Jahnke, R.A., and D.B. Craven. 1995. Quantifying and J.P. Chanton. 1999. Radiocarbon in marine and M. Landry. 1997. Experimental determi- the role of heterotrophic bacteria in the carbon bacteria: Evidence for the ages of assimi- nation of the organic carbon flux from open- cycle: A need for respiration measurements. lated carbon. Limnology and Oceanography ocean surface waters. Nature 389:951–954. Limnology and Oceanography 40:436–441. 44:730–736. Floodgate, G.D. 1995. Some environmental aspects Jones, C.W. 1983. Bacterial Respiration and Cole, J.J., S. Finlay, and M.L. Pace. 1988. Bacterial of hydrocarbon bacteriology. Aquatic Microbial Photosynthesis. American Society for

Oceanography June 2007 99 Microbiology, Washington, D.C. Benner. 2001. Production of refractory dis- chemical cycling of iron in the oceans. FEMS Karl, D.M., J.R. Christian, J.E. Dore, D.V. Hebel, solved organic matter by bacteria. Science Microbial Ecology 29:1–11. R.M. Letelier, L.M. Tupas, and C.D. Winn. 1996. 292:917–920. Vallino, J.J., C.S. Hopkinson, and J.E. Hobbie. 1996. Seasonal and interannual variability in primary Pomeroy, L.R. 1974. The ocean’s food web, a chang- Modeling bacterial utilization of dissolved production and particle flux at Station ALOHA. ing paradigm. BioScience 24:499–504. organic matter: Optimization replaces Monod Deep-Sea Research Part II 43:539–568. Reinthaler, T., and G.J. Herndl. 2005. Seasonal growth kinetics. Limnology and Oceanography Karl, D.M., E.A. Laws, P. Morris, P.J.leB Williams, dynamics of bacterial growth efficiencies 41(8):1,591–1,609. and S. Emerson. 2003. Metabolic balance of the in relation to phytoplankton in the south- Venter, J.C., K. Remington, J.F. Heidelberg, A.L. open sea. Nature 426:32. ern North Sea. Aquatic Microbial Ecology Halpern, D. Rusch, J.A. Eisen, D. Wu, I. Paulsen, Karner, M.B., E.F. DeLong, and D.M. Karl. 2001. 39(1):7–16. K.W. Nelson, W. Nelson, and others. 2004. Archaeal dominance in the mesopelagic zone. Reinthaler, T., H. van Aken, C. Veth, J. Aristegui, C. Environmental genome shotgun sequencing of Nature 409:507–510. Robinson, P.J.leB. Williams, P. Lebaron, and G.J. the Sargasso Sea. Science 304:66–72. Kawasaki, N., and R. Benner. 2006. Bacterial Herndl. 2006. Prokaryotic respiration and pro- Westernhoff, H.V., K.J. Hellingwerf, and K. Van release of dissolved organic matter during duction in the meso- and bathypelagic realm of Dam. 1983. Thermodynamic efficiency of cell growth and decline: Molecular origin and the eastern and western North Atlantic basin. microbial growth is low but optimal for maxi- composition. Limnology and Oceanography Limnology and Oceanography 51(3):1,262–1,273 mal growth rate. Proceedings of the National 51(5):2,170–2,180. Reinthaler, T., C. Winter, and G.J. Herndl. 2005. Academy of Sciences of the United States of Keil, R.G., and D.L. Kirchman. 1999. Utilization of Relationship between bacterioplankton rich- America 80:305–309. dissolved protein and amino acids in the north- ness, respiration, and production in the south- Whitman, W.B., D.C. Coleman, and W.J. Wiebe. ern Sargasso Sea. Aquatic Microbial Ecology ern North Sea. Applied and Environmental 1998. Prokaryotes: The unseen majority. 18:293–300. Microbiology 71(5):2,260–2,266. Proceedings of the National Academy of Sciences King, G.M. 2005. Ecophysiology of microbial res- Rivkin, R.B., and L. Legendre. 2001. Biogenic of the United States of America 95:6,578–6,583. piration. Pp. 18–35 in Respiration in Aquatic carbon cycling in the upper ocean: Effects of Williams, P.J.leB. 1998. The balance of plankton Ecosystems, Volume 18. P.A. del Giorgio and microbial respiration. Science 291:2,398–2,400. respiration and photosynthesis in the open P.J.leB. Williams, eds, Oxford University Press, Robinson, C. In press. Heterotrophic bacterial oceans. Nature 394(6688):55–57. New York. respiration. In Marine Microbial Ecology, 2nd Williams, P.J.leB., and D.G. Bowers. 1999. Regional Koch, A.L. 1997. Microbial physiology and ecol- Edition. D. Kirchman, ed., Plenum Press, New carbon imbalances in the ocean. Science ogy of slow growth. Microbiology and Molecular York. 284:173–174. Biology Reviews 61:1,092–2,172. Robinson, C., and P.J.leB. Williams. 2005. Williams, P.J.leB., and P.A. del Giorgio. 2005. Liu, Y. 1998. Energy uncoupling in microbial Respiration and its measurement in surface Respiration in aquatic ecosystems: History growth under substrate-sufficient condi- marine waters. Pp. 147–180 in Respiration in and background. Pp. 1–17 in Respiration in tions. Applied Microbiology and Biotechnology Aquatic Ecosystems. P.A. del Giorgio and P.J.leB. Aquatic Ecosystems. P.A. del Giorgio and P.J.leB. 49:500–505 Williams, eds, Oxford University Press, New Williams, eds, Oxford University Press, New McCarthy, M.D., J.I. Hedges, and R. Benner. 1998. York. York. Major bacterial contribution to marine dis- Russell, J.B., and G.M. Cook. 1995. Energetics of Williams, P.J.leB., P.J. Morris, and D.M. Karl. solved organic nitrogen. Science 281:231–234. bacterial growth: Balance of anabolic and cata- 2004. Net community production and meta- Middelboe, M., N.O.G. Jørgensen, and N. Kroer. bolic reactions. Microbiology Reviews 59:48–62. bolic balance at the oligotrophic ocean site, 1996. Effects of viruses on nutrient turnover Sherr, E.B., and B.E. Sherr. 1996. Temporal offset in station ALOHA. Deep-Sea Research Part I and growth efficiency of noninfected marine oceanic production and respiration processes 51:1,563–1,578. bacterioplankton. Applied and Environmental implied by seasonal changes in atmospheric Williams, P.M., and E.R.M. Druffel. 1987. Microbiology 62(6):1,991–1,997. oxygen: The role of heterotrophic microbes. Radiocarbon in dissolved organic matter Moeseneder, M.M., C. Winter, and G.J. Herndl. Aquatic Microbial Ecology 11:91–100. in the central North Pacific Ocean.Nature 2001. Horizontal and vertical complexity of Steinberg, D.K., C.A. Carlson, N.R. Bates, R.J. 330:246–248. attached and free-living bacteria of the Eastern Johnson, A.F. Michaels, and A.H. Knap. 2001. Mediterranean Sea, determined by 16S rDNA Overview of the U.S. JGOFS Bermuda Atlantic and 16S rRNA fingerprints. Limnology and Time-series Study (BATS): A decade-scale look Oceanography 46(1):95–107. at ocean biology and biogeochemistry. Deep-Sea Morris, R.M., S.A. Connon, M.S. Rappé, K.L. Research Part II 48:1,405–1,447. Vergin, W.A. Siebold, C.A. Carlson, and S.J. Tanoue, E., S. Nishiyama, M. Kamo, and A. Tsugita. Giovannoni. 2002. High cellular abundance of 1995. Bacterial membranes: Possible source of a the SAR11 bacterioplankton clade in seawater. major dissolved protein in seawater. Geochimica Nature 420:806–809. et Cosmochimica Acta 59(12):2,643–2,648. Obernosterer, I., N. Kawasaki, and R. Benner. 2003. Teixeira-de-Mattos, M.J., and O.M. Neijssel. 1997. P-limitation of respiration in the Sargasso Sea Bioenergetic consequences of microbial adapta- and uncoupling of bacteria from P-regeneration to low-nutrient environments. Journal of tion in size-fractionation experiments. Aquatic Biotechnology 59:117–126. Microbial Ecology 32:229–237. Tortell, P.D., M.T. Maldonado, J. Granger, and Ogawa, H., Y. Amagai, I. Koike, K. Kaiser, and R. N.M. Price. 1999. Marine bacteria and biogeo-

100 Oceanography Vol. 20, No. 2