Microb Ecol (1994) 28:255-271
Controls of the Microbial Loop: Nutrient Limitations
MICROBIAL ECOLOGYInc. © 1994Springer-Verlag New York
The Uptake of Inorganic Nutrients by Heterotrophic Bacteria D.L. Kirchman College of Marine Studies, University of Delaware, Lewes, Delaware 19958, USA
Abstract. It is now well known that heterotrophic bacteria account for a large portion of total uptake of both phosphate (60% median) and ammonium (30% median) in freshwaters and marine environments. Less clear are the factors controlling relative uptake by bacteria, and the consequences of this uptake on the plankton community and biogeochemical processes, e.g., new production. Some of the variation in reported inorganic nutrient uptake by bacteria is undoubtedly due to methodological problems, but even so, uptake would be expected to vary because of variation in several parameters, perhaps the most interesting being dissolved organic matter. Uptake of ammonium by bacteria is very low whereas uptake of dissolved free amino acids (DFAA) is high in eutrophic estuaries (the Delaware Bay and Chesapeake Bay). The concentrations and turnover of DFAA are insufficient, however, in oligotrophic oceans where bacteria turn to ammonium and nitrate, although the latter only as a last resort. I argue here that high uptake of dissolved organic carbon, which has been questioned, is necessary to balance the measured uptake of dissolved inorganic nitrogen (DIN) in seawater culture experiments. What is problematic is that this DIN uptake exceeds bacterial biomass production. One possibility is that bacteria excrete dissolved organic nitrogen (DON). A recent study offers some support for this hypothesis. Lysis by viruses would also release DON. While ammonium uptake by heterotrophic bacteria has been hypothesized to affect phytoplankton community structure, other impacts on the phytoplankton and biomass production (both total and new) are less clear and need further work. Also, even though bacteria account for a very large fraction of phosphate uptake, how this helps to structure the plankton community has not been examined. What is clear is that the interactions between bacterial and phytoplankton uptake of inorganic nutrients are more complicated than simple competition. One feature of the microbial loop that has revolutionized our view of aquatic ecosystems is the high uptake of inorganic nutrients by heterotrophic bacteria. Although we typically think that this inorganic nutrient uptake by bacteria was first observed relatively recently, uptake of at least one major inorganic nutrient was known to be important long before the term "microbial loop" was coined by Azam et al. [ 1]. Citing the work of Johannes [31 ], Pomeroy [58] pointed out that "aerobic bacteria tend to take up all available phosphorus" and went on to sketch the role of bacteria versus protozoa in remineralization, a picture that is still accepted today.
256
D.L. Kirchman
Pomeroy's review goes on to discuss the preferential use of NH4 + by phytoplankton, h o w NH4 + inhibits N O 3 - uptake, and the importance of new and regenerated production [14]. But heterotrophic use of NH4 + was not mentioned until 1977 by Eppley et al. [15]. We now know that heterotrophic bacteria can account for a large fraction of uptake of both orthophosphate (Pi) and dissolved inorganic nitrogen (DIN), principally NH4 +. I review here studies of that uptake, discuss how relative uptake by bacteria varies, and explore its importance to phytoplankton and community structure of freshwater and marine ecosystems. I do not have enough space to say much about the role of bacteria in the mineralization of organic matter to NH4 ÷ and Pi. Pomeroy's review in 1970 has it right anyway, as I already mentioned. Protists are probably more important than bacteria in the regeneration of both Pi and NH4 + , and the contribution of heterotrophic bacteria to n e t regeneration is probably small. However, Tupas and Koike [72] showed that bacteria can mineralize dissolved organic nitrogen (DON) to NH4 + even when net NH4 + uptake by these bacterial assemblages is high. This process has been shown to be important in ecosystem models (e.g., [16]), but it is not well understood.
Uptake of Phosphate and Ammonium by Bacteria Uptake of Pi by heterotrophic bacteria in freshwaters has been examined quite extensively, probably motivated by the recognition of the importance of P in limiting freshwater primary production. Several size fractionation studies have shown that heterotrophic bacteria are responsible for a large proportion of Pi uptake in lakes, usually at least 50% of total uptake (Table 1). The fraction is usually also quite high in marine systems, ranging from 5% in Peruvian coastal waters [25] to >90% in the Rhode River [17]. In addition, heterotrophic bacteria can account for a large fraction of total NH4 + assimilation in marine systems and in at least one freshwater lake (Table 2). The percentage attributable to bacteria ranges from insignificant uptake in the Delaware Bay [29] and Chesapeake Bay [21] to 78% in waters off Sapelo Island, Georgia [741. Heterotrophic bacteria appear to account for a much greater fraction of total Pi uptake than NH4 + uptake (Fig. 1), the median for Pi being 60% vs. 30% for NH4 + This difference should not be overemphasized, given the large variability in these percentages and various methodological problems. Also, only one study [65] has examined NH4 + uptake in lakes, and so any apparent differences between Pi and NH4 + may really be due to differences between marine and freshwaters; in fact, Lebo [48] found that relative Pi uptake did differ with salinity in the Delaware estuary. However, generally, Pi uptake in marine systems is just as high as in freshwaters (Table 1). Also, there are good biochemical reasons to expect Pi and NH4 + uptake to differ, as discussed below. The first step in understanding these percentages is to consider the ratio of bacterial to primary production (BACT:PRIM), which is usually expressed in carbon units. This should be instructive because nearly all of N and P assimilation has to be associated with biomass production eventually. The canonical BACT:
257
Nutrient Uptake by Bacteria Table 1.
Summary of studies examining phosphate uptake by heterotrophic bacteria
Location Freshwater Jacks Lake Jacks Lake Jacks Lake Jacks Lake Kennedy Lake Lake 227 Lake Memphremagog Lake Ontario 13 lakes Third Sister Lake Marine Delaware Estuary E. Tropical Pacific Peruvian coast Rhode River Saanich Inlet Sandsfjord Sargasso Sea Sargasso Sea S. California Median
% Total uptake by bacteria 60 60 72 >50-80 65-84 40 98 72 95 65-90 15-75 78 5 >90 >50 10-90 24-46 39 53 60% a
Comments Unenriched
Whole lake Mean at 10 h
Spatial and temporal variation Preincubation size fractionation Preincubation size fractionation Seasonal study CEPEX bags 63% in final budget Corrected for phytoplankton Corrected for phytoplankton
Reference [69] [47] [24] [3] [65] [50] [11] [68] [121 [7] [48] [25] [25] [17] [27] [70] [54] [51 ] [45]
aEach study was considered as one data point. When ranges were reported the median of the range was used
Table 2.
Summary of studies examining NH4 + uptake by heterotrophic bacteria vs. phytoplankton Location
% Total uptake by bacteria
Sapelo Is Long Is Sound Long Is Sound George Bank-NW Atlantic North Pacific Subarctic Pacific Delaware Bay Chesapeake Bay North Atlantic Kennedy Lake Median
78 5-50 20-42 15-42 50-75 45 10 10 22-39 43-88 30% a
Comments Seasonal 3 months Much chl < 1.0 Ixm 14CO2 into protein
Preincubation size fractionation Spring bloom 10 h), dark incubations with the bacterial size fraction, have proved very useful for examining DIN uptake. They have confirmed that naturally-occurring heterotrophic bacteria are capable of assimilating NO3- and urea [30, 41, 42], but few studies have examined this uptake in actual aquatic systems. Harrison and Wood [26] found a substantial amount (as much as 90% of total) of NO3- uptake by the < 1.0-p~m size fraction in coastal waters around Georges Bank, but there was also substantial chlorophyll and 14CO2 fixation, i.e., phytoplankton activity, in that size fractionation. In contrast, Kirchman et al. [44] did not have much chlorophyll in their 0.8-txm size fractions and concluded that heterotrophic bacteria accounted for 1 h) [50, 55, 66]. The biomass effect is also seen in cross-lake comparisons. Currie et al. [12] found a positive correlation between percent Pi uptake by phytoplankton (>3-1xm size fraction) and chlorophyll concentrations in 13 lakes. Data on bacterial biomass were not presented. Probably the bacteria:phytoplankton biomass ratio decreased as chlorophyll increased among these 13 lakes, a pattern suggested by other crosssystem comparisons [5, 6]. In the Rhode River estuary, Faust and Correll [17] found high correlations between Pi uptake by the 5-~m size fraction and phytoplankton biomass. Concentrations of NH4 + and Pi are also important in affecting their uptake. Several investigators have shown that addition of Pi decreased the ratio of bacterial to phytoplankton uptake, or more precisely, the ratio of uptake by small vs. large size fractions [60, 67]. Suttle et al. [67] demonstrated a similar effect with NH4 + . That is, bacteria appear to outcompete phytoplankton at low but not high concentrations of Pi and NH4 +. The competitive superiority of bacteria at least for Pi has been confirmed in chemostat experiments [10, 11], and Cotner and Wetzel [7] showed that heterotrophic bacteria have lower half-saturation constants (Km) than phytoplankton. The ability of bacteria to outcompete phytoplankton at low concentrations is usually attributed to the small size of bacteria and their high surface area to volume ratios. The trade off is that phytoplankton have higher maximum uptake velocities (Vmax) than bacteria (e.g., [70]). Consequently, bacteria better phytoplankton in low Pi concentrations, but phytoplankton prevail in high concentrations. I am not aware of any study showing this concentration effect in situ; all the studies to date have been short-term, enrichment experiments. One approach to demonstrate this concentration effect in situ would be to measure uptake across a nutrient gradient (eutrophic to oligotrophic transect) and then examine statistically the amount of variation in uptake by bacteria relative to phytoplankton explained by biomass and nutrient concentrations. The data of Harrison and Wood [26] seem to show that NH4 + uptake by heterotrophic bacteria varies in situ as we would expect. They observed that NH4 + uptake by the < 1.0-txm fraction decreased in a transect from oceanic to frontal waters with high nutrient concentrations (Fig. 2). Unfortunately, they also observed a substantial amount of chlorophyll and 14CO2fixation in the < 1.0-p~m size fraction, making it difficult to estimate the heterotrophic bacterial contribution. I think the picture suggested by Fig. 2 from Harrison and Wood [26] may be right for the wrong reasons. Another obvious factor affecting Pi and NH4 + uptake is the DOM used by bacteria. The first parameter to consider is the elemental ratios of the DOM versus
Nutrient Uptake by Bacteria
261
100 Oceanic
80
=
[
Chl
" pCO2
x pNO3" ÷ pNH4+
E =" 60
Stratified
Mixed
V
A
c G}
+
i ÷
o 40
x+
Frontal I
.13-
20 A
0
0.0
f
0.2
I
0.4 f ratio
I
0.6
, 0.8
x
Fig. 2. Fraction of chlorophyll (Chl) and uptake of 14CO~ (pCO2), NO3(pNO3-), and NH4* (pNH4 +) in the
Controls of the Microbial Loop: Nutrient Limitations
MICROBIAL ECOLOGYInc. © 1994Springer-Verlag New York
The Uptake of Inorganic Nutrients by Heterotrophic Bacteria D.L. Kirchman College of Marine Studies, University of Delaware, Lewes, Delaware 19958, USA
Abstract. It is now well known that heterotrophic bacteria account for a large portion of total uptake of both phosphate (60% median) and ammonium (30% median) in freshwaters and marine environments. Less clear are the factors controlling relative uptake by bacteria, and the consequences of this uptake on the plankton community and biogeochemical processes, e.g., new production. Some of the variation in reported inorganic nutrient uptake by bacteria is undoubtedly due to methodological problems, but even so, uptake would be expected to vary because of variation in several parameters, perhaps the most interesting being dissolved organic matter. Uptake of ammonium by bacteria is very low whereas uptake of dissolved free amino acids (DFAA) is high in eutrophic estuaries (the Delaware Bay and Chesapeake Bay). The concentrations and turnover of DFAA are insufficient, however, in oligotrophic oceans where bacteria turn to ammonium and nitrate, although the latter only as a last resort. I argue here that high uptake of dissolved organic carbon, which has been questioned, is necessary to balance the measured uptake of dissolved inorganic nitrogen (DIN) in seawater culture experiments. What is problematic is that this DIN uptake exceeds bacterial biomass production. One possibility is that bacteria excrete dissolved organic nitrogen (DON). A recent study offers some support for this hypothesis. Lysis by viruses would also release DON. While ammonium uptake by heterotrophic bacteria has been hypothesized to affect phytoplankton community structure, other impacts on the phytoplankton and biomass production (both total and new) are less clear and need further work. Also, even though bacteria account for a very large fraction of phosphate uptake, how this helps to structure the plankton community has not been examined. What is clear is that the interactions between bacterial and phytoplankton uptake of inorganic nutrients are more complicated than simple competition. One feature of the microbial loop that has revolutionized our view of aquatic ecosystems is the high uptake of inorganic nutrients by heterotrophic bacteria. Although we typically think that this inorganic nutrient uptake by bacteria was first observed relatively recently, uptake of at least one major inorganic nutrient was known to be important long before the term "microbial loop" was coined by Azam et al. [ 1]. Citing the work of Johannes [31 ], Pomeroy [58] pointed out that "aerobic bacteria tend to take up all available phosphorus" and went on to sketch the role of bacteria versus protozoa in remineralization, a picture that is still accepted today.
256
D.L. Kirchman
Pomeroy's review goes on to discuss the preferential use of NH4 + by phytoplankton, h o w NH4 + inhibits N O 3 - uptake, and the importance of new and regenerated production [14]. But heterotrophic use of NH4 + was not mentioned until 1977 by Eppley et al. [15]. We now know that heterotrophic bacteria can account for a large fraction of uptake of both orthophosphate (Pi) and dissolved inorganic nitrogen (DIN), principally NH4 +. I review here studies of that uptake, discuss how relative uptake by bacteria varies, and explore its importance to phytoplankton and community structure of freshwater and marine ecosystems. I do not have enough space to say much about the role of bacteria in the mineralization of organic matter to NH4 ÷ and Pi. Pomeroy's review in 1970 has it right anyway, as I already mentioned. Protists are probably more important than bacteria in the regeneration of both Pi and NH4 + , and the contribution of heterotrophic bacteria to n e t regeneration is probably small. However, Tupas and Koike [72] showed that bacteria can mineralize dissolved organic nitrogen (DON) to NH4 + even when net NH4 + uptake by these bacterial assemblages is high. This process has been shown to be important in ecosystem models (e.g., [16]), but it is not well understood.
Uptake of Phosphate and Ammonium by Bacteria Uptake of Pi by heterotrophic bacteria in freshwaters has been examined quite extensively, probably motivated by the recognition of the importance of P in limiting freshwater primary production. Several size fractionation studies have shown that heterotrophic bacteria are responsible for a large proportion of Pi uptake in lakes, usually at least 50% of total uptake (Table 1). The fraction is usually also quite high in marine systems, ranging from 5% in Peruvian coastal waters [25] to >90% in the Rhode River [17]. In addition, heterotrophic bacteria can account for a large fraction of total NH4 + assimilation in marine systems and in at least one freshwater lake (Table 2). The percentage attributable to bacteria ranges from insignificant uptake in the Delaware Bay [29] and Chesapeake Bay [21] to 78% in waters off Sapelo Island, Georgia [741. Heterotrophic bacteria appear to account for a much greater fraction of total Pi uptake than NH4 + uptake (Fig. 1), the median for Pi being 60% vs. 30% for NH4 + This difference should not be overemphasized, given the large variability in these percentages and various methodological problems. Also, only one study [65] has examined NH4 + uptake in lakes, and so any apparent differences between Pi and NH4 + may really be due to differences between marine and freshwaters; in fact, Lebo [48] found that relative Pi uptake did differ with salinity in the Delaware estuary. However, generally, Pi uptake in marine systems is just as high as in freshwaters (Table 1). Also, there are good biochemical reasons to expect Pi and NH4 + uptake to differ, as discussed below. The first step in understanding these percentages is to consider the ratio of bacterial to primary production (BACT:PRIM), which is usually expressed in carbon units. This should be instructive because nearly all of N and P assimilation has to be associated with biomass production eventually. The canonical BACT:
257
Nutrient Uptake by Bacteria Table 1.
Summary of studies examining phosphate uptake by heterotrophic bacteria
Location Freshwater Jacks Lake Jacks Lake Jacks Lake Jacks Lake Kennedy Lake Lake 227 Lake Memphremagog Lake Ontario 13 lakes Third Sister Lake Marine Delaware Estuary E. Tropical Pacific Peruvian coast Rhode River Saanich Inlet Sandsfjord Sargasso Sea Sargasso Sea S. California Median
% Total uptake by bacteria 60 60 72 >50-80 65-84 40 98 72 95 65-90 15-75 78 5 >90 >50 10-90 24-46 39 53 60% a
Comments Unenriched
Whole lake Mean at 10 h
Spatial and temporal variation Preincubation size fractionation Preincubation size fractionation Seasonal study CEPEX bags 63% in final budget Corrected for phytoplankton Corrected for phytoplankton
Reference [69] [47] [24] [3] [65] [50] [11] [68] [121 [7] [48] [25] [25] [17] [27] [70] [54] [51 ] [45]
aEach study was considered as one data point. When ranges were reported the median of the range was used
Table 2.
Summary of studies examining NH4 + uptake by heterotrophic bacteria vs. phytoplankton Location
% Total uptake by bacteria
Sapelo Is Long Is Sound Long Is Sound George Bank-NW Atlantic North Pacific Subarctic Pacific Delaware Bay Chesapeake Bay North Atlantic Kennedy Lake Median
78 5-50 20-42 15-42 50-75 45 10 10 22-39 43-88 30% a
Comments Seasonal 3 months Much chl < 1.0 Ixm 14CO2 into protein
Preincubation size fractionation Spring bloom 10 h), dark incubations with the bacterial size fraction, have proved very useful for examining DIN uptake. They have confirmed that naturally-occurring heterotrophic bacteria are capable of assimilating NO3- and urea [30, 41, 42], but few studies have examined this uptake in actual aquatic systems. Harrison and Wood [26] found a substantial amount (as much as 90% of total) of NO3- uptake by the < 1.0-p~m size fraction in coastal waters around Georges Bank, but there was also substantial chlorophyll and 14CO2 fixation, i.e., phytoplankton activity, in that size fractionation. In contrast, Kirchman et al. [44] did not have much chlorophyll in their 0.8-txm size fractions and concluded that heterotrophic bacteria accounted for 1 h) [50, 55, 66]. The biomass effect is also seen in cross-lake comparisons. Currie et al. [12] found a positive correlation between percent Pi uptake by phytoplankton (>3-1xm size fraction) and chlorophyll concentrations in 13 lakes. Data on bacterial biomass were not presented. Probably the bacteria:phytoplankton biomass ratio decreased as chlorophyll increased among these 13 lakes, a pattern suggested by other crosssystem comparisons [5, 6]. In the Rhode River estuary, Faust and Correll [17] found high correlations between Pi uptake by the 5-~m size fraction and phytoplankton biomass. Concentrations of NH4 + and Pi are also important in affecting their uptake. Several investigators have shown that addition of Pi decreased the ratio of bacterial to phytoplankton uptake, or more precisely, the ratio of uptake by small vs. large size fractions [60, 67]. Suttle et al. [67] demonstrated a similar effect with NH4 + . That is, bacteria appear to outcompete phytoplankton at low but not high concentrations of Pi and NH4 +. The competitive superiority of bacteria at least for Pi has been confirmed in chemostat experiments [10, 11], and Cotner and Wetzel [7] showed that heterotrophic bacteria have lower half-saturation constants (Km) than phytoplankton. The ability of bacteria to outcompete phytoplankton at low concentrations is usually attributed to the small size of bacteria and their high surface area to volume ratios. The trade off is that phytoplankton have higher maximum uptake velocities (Vmax) than bacteria (e.g., [70]). Consequently, bacteria better phytoplankton in low Pi concentrations, but phytoplankton prevail in high concentrations. I am not aware of any study showing this concentration effect in situ; all the studies to date have been short-term, enrichment experiments. One approach to demonstrate this concentration effect in situ would be to measure uptake across a nutrient gradient (eutrophic to oligotrophic transect) and then examine statistically the amount of variation in uptake by bacteria relative to phytoplankton explained by biomass and nutrient concentrations. The data of Harrison and Wood [26] seem to show that NH4 + uptake by heterotrophic bacteria varies in situ as we would expect. They observed that NH4 + uptake by the < 1.0-txm fraction decreased in a transect from oceanic to frontal waters with high nutrient concentrations (Fig. 2). Unfortunately, they also observed a substantial amount of chlorophyll and 14CO2fixation in the < 1.0-p~m size fraction, making it difficult to estimate the heterotrophic bacterial contribution. I think the picture suggested by Fig. 2 from Harrison and Wood [26] may be right for the wrong reasons. Another obvious factor affecting Pi and NH4 + uptake is the DOM used by bacteria. The first parameter to consider is the elemental ratios of the DOM versus
Nutrient Uptake by Bacteria
261
100 Oceanic
80
=
[
Chl
" pCO2
x pNO3" ÷ pNH4+
E =" 60
Stratified
Mixed
V
A
c G}
+
i ÷
o 40
x+
Frontal I
.13-
20 A
0
0.0
f
0.2
I
0.4 f ratio
I
0.6
, 0.8
x
Fig. 2. Fraction of chlorophyll (Chl) and uptake of 14CO~ (pCO2), NO3(pNO3-), and NH4* (pNH4 +) in the
