Microb Ecol (1993) 26:79-99
MICROBIAL ECOLOGYInc. © 1993 Sprhtger-VedagNew York
Factors Controlling Bacterial Production in Marine and Freshwater Sediments Bettina C. Sander and Jacob Kalff McGill University, Department of Biology, 1205 Dr. Penfield Ave., Montreal, Quebec, Canada H3A 1B1
Received: November 2, 1992; Revised: April 26, 1993
Abstract.
We collected benthic bacterial production data measured by 3H thymidine incorporation (TTI) (25 studies), frequency of dividing cells (FDC) (3 studies), dark-CO2 assimilation (1 study) and 3H-adenine uptake (2 studies) from the literature, which included 18 marine, 6 river, and 2 lake studies. In all of the studies that used the TTI method, 3H-DNA was isolated and incubations were carried out at in s i t u temperatures. Most of the researchers also determined 3H-DNA extraction efficiencies and isotope dilution, thus interpretable estimates of bacterial production were used in the analysis. In marine sediments, bacterial production rates were linked to bacterial biomass, bacterial abundance, sediment organic matter, temperature, and sediment chlorophyll a, with these variables explaining between 40% and 68% of the variation in production rates. Simple relationships between production and bacterial biomass or bacterial abundance, or between production and sediment organic matter, were improved by also including temperature in the analysis of marine sediments. Sediment organic matter explained an appreciable fraction (58%) of the observed production in freshwater sediments. Temperature was the most powerful predictor of the observed variability in specific growth rates (r 2 = 0.48 and r 2 = 0.58) in marine and freshwater sediments, respectively. Thus, bacterial production and specific growth rates are most closely linked to substrate supply and temperature in marine and freshwater sediments.
Introduction The notion that bacterial production is important in aquatic food webs has provided the basis for much research [35, 60]. Bacterial production rates have been estimated in a wide variety of marine and freshwater planktonic environments (for review see [68]). Much less is known about marine and freshwater sediments, but bacterial abundances have been reported to be 2-1000 times higher in sediments than in the
Offprint requests to: B. C. Sander.
80
B.C. Sander and J. Kalff
overlying water column of marine [12, 41, 49] and freshwater [15, 59, 67] ecosystems. If bacterial production were to be roughly proportional to bacterial abundance, bacterial production rates in sediments would be expected to be much higher than in the water column. The postulated high rates of bacterial production in sediments affect the chemistry of the overlying water through bacterial consumption of hypolimnetic oxygen [29], reduction of nitrate, methane, and sulphate [18, 52, 67], and the release of phosphorus from sediments [7, 66]. Following the development of the 3H thymidine incorporation (TTI) method [21], sufficient data have gradually become available to allow an assessment of the relation between site-specific bacterial production rates and various environmental factors. This was first done by Cole et al. [13], who reported significant relationships between benthic bacterial production, sediment organic matter, and benthic bacterial biomass, albeit for a small data set. The marine and freshwater plankton has been more extensively studied where bacterial abundance, chlorophyll a concentrations, temperature, and primary production have been linked to both bacterial production and specific growth rates [13, 68]. Despite the relationship reported by Cole et al. [13] between benthic bacterial production and organic matter, the interpretation presented is not unambiguous. Their model expresses both organic matter and bacterial production per gram dry weight of sediment, thereby introducing a possibly important statistical bias [6, 7, 59]. Many correlations between bacteria and organic matter turn out to be significant only as a result of the per gram dry weight standardization of both the bacterial and the sediment attributes. If the bias is removed, the correlations are weakened. This is acknowledged by Cole et al. [13], who report less of the variability in bacterial production to be explained by organic matter when both variables were expressed per square meter rather than per gram dry weight of sediment. The present study uses benthic bacterial production rates collected from the literature: (1) to search for possible relationships between benthic bacterial production and selected environmental variables; (2) to examine the predictive power of models of sediment bacterial production reported by Cole et al. [ 13] with a new and larger data set; (3) to test if marine and freshwater differ in their benthic bacterial productivity; and (4) to examine if factors controlling bacterial production in the water column differ from those in the sediments.
Materials and Methods
Data Collection Bacterial production rates were collected from the literature along with the following biological and physico-chemical variables, where available: bacterial abundance, bacterial biomass, bacterial volume, specific growth rate, pool size of exogenous and endogenous thymidine (TdR), sediment chlorophyll a, sediment organic carbon, temperature, sediment water content, Eh, pH, mean depth, and Secchi depth. Variables were selected because: (1) they have been correlated with bacterial production in previous work; or (2) they represent potential substrate sources (i.e., organic carbon or chlorophyll a) to bacteria, thereby possibly influencing production rates. The data included bacterial production rates measured in marine, freshwater, estuarine, and fiver sites (Table 1). Marine sites included salt marshes, tropical mangrove sediments, sandy beaches, reef flats, seagrass flats, mud flats, coral reefs, silt-clay sediments, intertidal flats, bay sites, sites near river
Bacterial Production in Sediments
81
Table 1. Sources of data for the analysis. Bacterial production rates estimated by 3H-thymidine incorporation (e), frequency of dividing ceils (A), dark CO 2 assimilation (+), and 3H-adenine uptake (*) Marine sediments
[1] (e) [2] (e) [31 (e) [8] (e) [10] (e) [14] (*) [171 (e,A) [241 (e) [331 (e) [34] (e) [421 (e) [43] (e) 144] (e) [481 (e) [50] (*) [53] (e) [61] (+) [64] (e)
Lake sediments
River sediments
[5] (e) [16] (e,A)
[4] (e) [19J (e) [201 (e) Hudson and Roff, in press (e)
[30] (e) [39] (e)
deltas, and saline ponds. The much more limited river data included samples from blackwater rivers, a forested river, an open stream, and a mountain stream. To date, only two studies have reported lake sediment production rates, one oligotrophic and the other hypereutrophic. The limited freshwater data are, therefore, biased towards flowing waters. To obtain an unbiased data set, studies were chosen based on the following criteria: (1) both benthic bacterial production rates and benthic bacterial abundances were reported because bacterial abundance has been shown to be a good predictor of bacterial production in pelagic systems [68]; (2) the methods of analysis were mentioned and explained; and (3) numbers were either given in tables or could be extracted from figures using an image analyzer. Bacterial production was measured by: (1) the rate of 3H-TdR incorporation (TTI) (n = 25); (2) frequency of dividing cells (FDC) (n = 3); (3) dark-CO2 assimilation (n = 1); or (4) 3H adenine uptake (n = 2). Only 3H TdR uptake rates were recorded in one study, but not converted to bacterial production rates [16]. To make the conversion, a conversion factor of 10 9 cells produced (nmol TdR incorporated) -1 [57] was used to convert rate of 3H TdR uptake to number of cells produced. This conversion factor was used because it represents the mean value of conversion factors determined in batch cultures of coastal marine bacteria grown under different nutrient conditions and at different temperatures. All other TTI studies examined used conversion factors obtained from the plankton literature (1-2 × 109 cells produced (nmol TdR incorporated) -1) [21, 57]. To calculate bacterial production rates as grams of carbon produced from number of cells produced, a carbon per cell conversion factor of 2.2 × 10 -13 gC txm-3 [9] and a mean cell volume of 0. l /xm3 were used. The particular conversion factor lies in the middle of the range of conversion factors (0.5-5.0 × 10 -13 gC Ixm-3) reported in the literature [55]. Bacterial abundances were obtained by epifluorescence direct counts, employing either acridine orange or DAPI, except in the study by Sorokin [61] where bacterial cells were stained with erythrosin and fuchsin. While all environmental variables were not reported in every study, some could be derived by simple calculations. Specific growth rates (day-~) were estimated by dividing bacterial production (mgC m -2 day -I) by bacterial biomass (mgC m-2). Alternatively, specific growth rates (day -1) could be determined from turnover time (day) or from generation time (day).
82
B.C. Sander and J. Kalff
Biological and physico-chemical variables were not always expressed in the same units. We extrapolated hourly rates to daily values. Secondly, to remove any statistical bias [6], variables expressed per gram dry weight of sediment were converted to aerial units (per square meter) by taking into account the depth of sediment over which the particular factor was measured. As the depth of sediment analyzed by Fallon et al. [17] was not given, we assumed a depth of 1 cm on the basis of a previous study of the same site by Newell and Fallon [48]. Conversion to aerial units was easy if sediment density was provided; otherwise density was estimated from the reported sediment water content [58] as: sediment density (gDW m1-1) = - 0 . 0 2 x water content (%, g/gDW) + 1.72 r 2 = 0.95, n = 190, p < 0 . 0 0 0
(1)
Finally, sediment organic matter was converted to sediment organic carbon, assuming that organic carbon comprises half of the organic matter present [23].
Data Analysis Data were analyzed by ordinary least squares regression using SYSTAT [70]. Data were fit to the model y = a + b(x) for simple linear regression and y = a + bl(x l) + bz(X2) q- " " " + bk(Xk) for multiple regressions. Data were logarithm transformed to their normal base 10, arcsine, or square root, transformed when necessary to equalize the variance over the range of observations and to meet the normality requirements of least squares regression [73]. A correction factor [CF = exp (1.152 x SEE2), where SEE = standard error of the estimate] is required for the conversion to the arithmetic from the logarithmic scale [62]. We also present the Model II slope as a better measure of the true relationship between x and y variables when there is, as expected, uncertainty in the measurement of x [56]. To compare the predictive power of univariate and multivariate models, adjusted coefficients of correlation (r 2) that account for the differences in the number of parameters in the models were computed. The following abbreviations are used in the text: TdR = thymidine, TTI = 3H-thymidine incorporation method, PROD = benthic bacterial production (mgC m -2 day-1), POOL = pool size of exogenous and endogenous TdR (nM), ABUND = bacterial abundance (cells m 2), BIOM = bacterial biomass (gC m-2), VOL = cell volume (p, m3), SGR = specific growth rate (day-t), CHLA = sediment chlorophyll a (rag m-2), ORG-C = sediment organic carbon (gC m - Z ) , O M = sediment organic carbon (%, g/gDW), TEMP = temperature (°C).
Results
Mean, minimum, and maximum values of the variables analyzed are summarized in Table 2. In marine sediments, highest bacterial production rates (43200 mgC m -z day-1) [48] and lowest bacterial abundances (0.05 x 1013 cells m-2), as well as specific growth rates (0.005 day -a) [17], were measured in nearshore sandy, marine sediments of Sapelo Island, Georgia. Lowest rates (1 mgC m -z day -1) were reported from the Dutch Wadden Sea at -1.3°C [64]. Highest bacterial abundances (328.7 x 1013 cells m -z) and specific growth rates (5.5 day -1) were reported in sediments of mangrove estuaries by Alongi [1]. Surprisingly, higher bacterial production rates (348 mgC m -2 day -~) and specific growth rates (2.5 day -1) were estimated in the fine sediments of an oligotrophic lake [16] than in sediments of a hypereutrophic lake [5], even though more bacterial cells were present in the latter (21.9 x 1013 cells m -z) than in the former (0.3 × 1013 cells m-2). Both the highest (20726 mgC m - 2 day -1) and the lowest (0.2 mgC m -2 day- 1) bacterial production rates were reported in sandy sediments of two blackwater rivers studied by Findlay et al. [20] and Meyer [39]. Lower bacterial numbers
Bacterial Production in Sediments
83
Table 2. Sample size, mean, minimum, maximum values of bacterial production (PROD; mgC m -2 day-1), pool size of exogenous and endogenous thymidine (POOLSIZE; nr~ TdR), bacterial abundance (ABUND; 1013 cells m-2), bacterial cell volume (VOL; txm3), bacterial biomass (BIOMASS; gC m-2), specific growth rate (SGR; day-l), sediment organic carbon (ORG-C; gC m-2), sediment chlorophyll a (CHLA; mg m-e), temperature (TEMP; °C), and redox potential (Eh; mV) Variable Marine PROD POOLSIZE ABUND VOL BIOMASS SGR ORG-C CHLA TEMP Eh
Lake PROD POOLSIZE ABUND VOL BIOMASS SGR ORG-C CHLA TEMP Eh
Sample size
Mean
Minimum
Maximum
221 5 233 4 211 203 43 57 56 40
959.9 0.09 19.4 0.2 3.8 0.5 3.9 39.9 19.2 219.8
1 0.07 0.05 0.09 0.001 0.005 0.1 1.5 - 1.3 7
43200 0.1 328.7 0.3 55.9 5.5 18 138.2 30.5 370
13 5 13 5 11 16 5 . 13
86.5 4.3 6.9 0.1 2.1 0.8 7.4 . 17
2 1.5 0.3 0.08 0.07 0.004 0.2
348 6.5 21.6 0.1 7.6 2.5 10.9
.
River PROD POOLSIZE ABUND VOL BIOMASS SGR ORG-C CHLA TEMP Eh
. .
2
.
98 10 96 12 96 17 61 2 75 .
.
1235 2780 10.3 0.2 0.4 1.3 9.4 89.5 20.6 .
30
.
0.2 226 0.01 0.05 0.003 0.02 0.05 64.3 8 .
20726 7103 15.4 0.3 2.9 5.9 55.8 114.6 32
.
were again o b s e r v e d in sediments harboring higher bacterial production rates. H u d s o n and R o f f (in press in C J F A S ) reported highest specific growth rates (5.9 d a y - 1 ) in sandy sediments o f an Ontario stream, whereas lowest specific growth rates (0.02 d a y - a ) were m e a s u r e d in the H u d s o n R i v e r [4].
Bacterial
Production
Bacterial p r o d u c t i o n was significantly related to bacterial b i o m a s s only in m a r i n e sediments (Fig. l a ) , where bacterial b i o m a s s explained 48% o f the variability in
84
B.C. Sander and J. Kalff
100000
A
1o000 E (5
• "
•
:
q: ° " • [] I
IOOO
E
•
[]
ea a
~ ~
[]
.ram
marine
d~o,.n:t'~baE-i:n~i~ a"~_
m
~
" I~aD•
a
e-
.o +5
100
[]
a e ~L~J.CrV~ ~ q2~l~l
'~- ' F rCH i~ - "r-
,A
"0
.o C3.
+o
•
°
.
=
MICROBIAL ECOLOGYInc. © 1993 Sprhtger-VedagNew York
Factors Controlling Bacterial Production in Marine and Freshwater Sediments Bettina C. Sander and Jacob Kalff McGill University, Department of Biology, 1205 Dr. Penfield Ave., Montreal, Quebec, Canada H3A 1B1
Received: November 2, 1992; Revised: April 26, 1993
Abstract.
We collected benthic bacterial production data measured by 3H thymidine incorporation (TTI) (25 studies), frequency of dividing cells (FDC) (3 studies), dark-CO2 assimilation (1 study) and 3H-adenine uptake (2 studies) from the literature, which included 18 marine, 6 river, and 2 lake studies. In all of the studies that used the TTI method, 3H-DNA was isolated and incubations were carried out at in s i t u temperatures. Most of the researchers also determined 3H-DNA extraction efficiencies and isotope dilution, thus interpretable estimates of bacterial production were used in the analysis. In marine sediments, bacterial production rates were linked to bacterial biomass, bacterial abundance, sediment organic matter, temperature, and sediment chlorophyll a, with these variables explaining between 40% and 68% of the variation in production rates. Simple relationships between production and bacterial biomass or bacterial abundance, or between production and sediment organic matter, were improved by also including temperature in the analysis of marine sediments. Sediment organic matter explained an appreciable fraction (58%) of the observed production in freshwater sediments. Temperature was the most powerful predictor of the observed variability in specific growth rates (r 2 = 0.48 and r 2 = 0.58) in marine and freshwater sediments, respectively. Thus, bacterial production and specific growth rates are most closely linked to substrate supply and temperature in marine and freshwater sediments.
Introduction The notion that bacterial production is important in aquatic food webs has provided the basis for much research [35, 60]. Bacterial production rates have been estimated in a wide variety of marine and freshwater planktonic environments (for review see [68]). Much less is known about marine and freshwater sediments, but bacterial abundances have been reported to be 2-1000 times higher in sediments than in the
Offprint requests to: B. C. Sander.
80
B.C. Sander and J. Kalff
overlying water column of marine [12, 41, 49] and freshwater [15, 59, 67] ecosystems. If bacterial production were to be roughly proportional to bacterial abundance, bacterial production rates in sediments would be expected to be much higher than in the water column. The postulated high rates of bacterial production in sediments affect the chemistry of the overlying water through bacterial consumption of hypolimnetic oxygen [29], reduction of nitrate, methane, and sulphate [18, 52, 67], and the release of phosphorus from sediments [7, 66]. Following the development of the 3H thymidine incorporation (TTI) method [21], sufficient data have gradually become available to allow an assessment of the relation between site-specific bacterial production rates and various environmental factors. This was first done by Cole et al. [13], who reported significant relationships between benthic bacterial production, sediment organic matter, and benthic bacterial biomass, albeit for a small data set. The marine and freshwater plankton has been more extensively studied where bacterial abundance, chlorophyll a concentrations, temperature, and primary production have been linked to both bacterial production and specific growth rates [13, 68]. Despite the relationship reported by Cole et al. [13] between benthic bacterial production and organic matter, the interpretation presented is not unambiguous. Their model expresses both organic matter and bacterial production per gram dry weight of sediment, thereby introducing a possibly important statistical bias [6, 7, 59]. Many correlations between bacteria and organic matter turn out to be significant only as a result of the per gram dry weight standardization of both the bacterial and the sediment attributes. If the bias is removed, the correlations are weakened. This is acknowledged by Cole et al. [13], who report less of the variability in bacterial production to be explained by organic matter when both variables were expressed per square meter rather than per gram dry weight of sediment. The present study uses benthic bacterial production rates collected from the literature: (1) to search for possible relationships between benthic bacterial production and selected environmental variables; (2) to examine the predictive power of models of sediment bacterial production reported by Cole et al. [ 13] with a new and larger data set; (3) to test if marine and freshwater differ in their benthic bacterial productivity; and (4) to examine if factors controlling bacterial production in the water column differ from those in the sediments.
Materials and Methods
Data Collection Bacterial production rates were collected from the literature along with the following biological and physico-chemical variables, where available: bacterial abundance, bacterial biomass, bacterial volume, specific growth rate, pool size of exogenous and endogenous thymidine (TdR), sediment chlorophyll a, sediment organic carbon, temperature, sediment water content, Eh, pH, mean depth, and Secchi depth. Variables were selected because: (1) they have been correlated with bacterial production in previous work; or (2) they represent potential substrate sources (i.e., organic carbon or chlorophyll a) to bacteria, thereby possibly influencing production rates. The data included bacterial production rates measured in marine, freshwater, estuarine, and fiver sites (Table 1). Marine sites included salt marshes, tropical mangrove sediments, sandy beaches, reef flats, seagrass flats, mud flats, coral reefs, silt-clay sediments, intertidal flats, bay sites, sites near river
Bacterial Production in Sediments
81
Table 1. Sources of data for the analysis. Bacterial production rates estimated by 3H-thymidine incorporation (e), frequency of dividing ceils (A), dark CO 2 assimilation (+), and 3H-adenine uptake (*) Marine sediments
[1] (e) [2] (e) [31 (e) [8] (e) [10] (e) [14] (*) [171 (e,A) [241 (e) [331 (e) [34] (e) [421 (e) [43] (e) 144] (e) [481 (e) [50] (*) [53] (e) [61] (+) [64] (e)
Lake sediments
River sediments
[5] (e) [16] (e,A)
[4] (e) [19J (e) [201 (e) Hudson and Roff, in press (e)
[30] (e) [39] (e)
deltas, and saline ponds. The much more limited river data included samples from blackwater rivers, a forested river, an open stream, and a mountain stream. To date, only two studies have reported lake sediment production rates, one oligotrophic and the other hypereutrophic. The limited freshwater data are, therefore, biased towards flowing waters. To obtain an unbiased data set, studies were chosen based on the following criteria: (1) both benthic bacterial production rates and benthic bacterial abundances were reported because bacterial abundance has been shown to be a good predictor of bacterial production in pelagic systems [68]; (2) the methods of analysis were mentioned and explained; and (3) numbers were either given in tables or could be extracted from figures using an image analyzer. Bacterial production was measured by: (1) the rate of 3H-TdR incorporation (TTI) (n = 25); (2) frequency of dividing cells (FDC) (n = 3); (3) dark-CO2 assimilation (n = 1); or (4) 3H adenine uptake (n = 2). Only 3H TdR uptake rates were recorded in one study, but not converted to bacterial production rates [16]. To make the conversion, a conversion factor of 10 9 cells produced (nmol TdR incorporated) -1 [57] was used to convert rate of 3H TdR uptake to number of cells produced. This conversion factor was used because it represents the mean value of conversion factors determined in batch cultures of coastal marine bacteria grown under different nutrient conditions and at different temperatures. All other TTI studies examined used conversion factors obtained from the plankton literature (1-2 × 109 cells produced (nmol TdR incorporated) -1) [21, 57]. To calculate bacterial production rates as grams of carbon produced from number of cells produced, a carbon per cell conversion factor of 2.2 × 10 -13 gC txm-3 [9] and a mean cell volume of 0. l /xm3 were used. The particular conversion factor lies in the middle of the range of conversion factors (0.5-5.0 × 10 -13 gC Ixm-3) reported in the literature [55]. Bacterial abundances were obtained by epifluorescence direct counts, employing either acridine orange or DAPI, except in the study by Sorokin [61] where bacterial cells were stained with erythrosin and fuchsin. While all environmental variables were not reported in every study, some could be derived by simple calculations. Specific growth rates (day-~) were estimated by dividing bacterial production (mgC m -2 day -I) by bacterial biomass (mgC m-2). Alternatively, specific growth rates (day -1) could be determined from turnover time (day) or from generation time (day).
82
B.C. Sander and J. Kalff
Biological and physico-chemical variables were not always expressed in the same units. We extrapolated hourly rates to daily values. Secondly, to remove any statistical bias [6], variables expressed per gram dry weight of sediment were converted to aerial units (per square meter) by taking into account the depth of sediment over which the particular factor was measured. As the depth of sediment analyzed by Fallon et al. [17] was not given, we assumed a depth of 1 cm on the basis of a previous study of the same site by Newell and Fallon [48]. Conversion to aerial units was easy if sediment density was provided; otherwise density was estimated from the reported sediment water content [58] as: sediment density (gDW m1-1) = - 0 . 0 2 x water content (%, g/gDW) + 1.72 r 2 = 0.95, n = 190, p < 0 . 0 0 0
(1)
Finally, sediment organic matter was converted to sediment organic carbon, assuming that organic carbon comprises half of the organic matter present [23].
Data Analysis Data were analyzed by ordinary least squares regression using SYSTAT [70]. Data were fit to the model y = a + b(x) for simple linear regression and y = a + bl(x l) + bz(X2) q- " " " + bk(Xk) for multiple regressions. Data were logarithm transformed to their normal base 10, arcsine, or square root, transformed when necessary to equalize the variance over the range of observations and to meet the normality requirements of least squares regression [73]. A correction factor [CF = exp (1.152 x SEE2), where SEE = standard error of the estimate] is required for the conversion to the arithmetic from the logarithmic scale [62]. We also present the Model II slope as a better measure of the true relationship between x and y variables when there is, as expected, uncertainty in the measurement of x [56]. To compare the predictive power of univariate and multivariate models, adjusted coefficients of correlation (r 2) that account for the differences in the number of parameters in the models were computed. The following abbreviations are used in the text: TdR = thymidine, TTI = 3H-thymidine incorporation method, PROD = benthic bacterial production (mgC m -2 day-1), POOL = pool size of exogenous and endogenous TdR (nM), ABUND = bacterial abundance (cells m 2), BIOM = bacterial biomass (gC m-2), VOL = cell volume (p, m3), SGR = specific growth rate (day-t), CHLA = sediment chlorophyll a (rag m-2), ORG-C = sediment organic carbon (gC m - Z ) , O M = sediment organic carbon (%, g/gDW), TEMP = temperature (°C).
Results
Mean, minimum, and maximum values of the variables analyzed are summarized in Table 2. In marine sediments, highest bacterial production rates (43200 mgC m -z day-1) [48] and lowest bacterial abundances (0.05 x 1013 cells m-2), as well as specific growth rates (0.005 day -a) [17], were measured in nearshore sandy, marine sediments of Sapelo Island, Georgia. Lowest rates (1 mgC m -z day -1) were reported from the Dutch Wadden Sea at -1.3°C [64]. Highest bacterial abundances (328.7 x 1013 cells m -z) and specific growth rates (5.5 day -1) were reported in sediments of mangrove estuaries by Alongi [1]. Surprisingly, higher bacterial production rates (348 mgC m -2 day -~) and specific growth rates (2.5 day -1) were estimated in the fine sediments of an oligotrophic lake [16] than in sediments of a hypereutrophic lake [5], even though more bacterial cells were present in the latter (21.9 x 1013 cells m -z) than in the former (0.3 × 1013 cells m-2). Both the highest (20726 mgC m - 2 day -1) and the lowest (0.2 mgC m -2 day- 1) bacterial production rates were reported in sandy sediments of two blackwater rivers studied by Findlay et al. [20] and Meyer [39]. Lower bacterial numbers
Bacterial Production in Sediments
83
Table 2. Sample size, mean, minimum, maximum values of bacterial production (PROD; mgC m -2 day-1), pool size of exogenous and endogenous thymidine (POOLSIZE; nr~ TdR), bacterial abundance (ABUND; 1013 cells m-2), bacterial cell volume (VOL; txm3), bacterial biomass (BIOMASS; gC m-2), specific growth rate (SGR; day-l), sediment organic carbon (ORG-C; gC m-2), sediment chlorophyll a (CHLA; mg m-e), temperature (TEMP; °C), and redox potential (Eh; mV) Variable Marine PROD POOLSIZE ABUND VOL BIOMASS SGR ORG-C CHLA TEMP Eh
Lake PROD POOLSIZE ABUND VOL BIOMASS SGR ORG-C CHLA TEMP Eh
Sample size
Mean
Minimum
Maximum
221 5 233 4 211 203 43 57 56 40
959.9 0.09 19.4 0.2 3.8 0.5 3.9 39.9 19.2 219.8
1 0.07 0.05 0.09 0.001 0.005 0.1 1.5 - 1.3 7
43200 0.1 328.7 0.3 55.9 5.5 18 138.2 30.5 370
13 5 13 5 11 16 5 . 13
86.5 4.3 6.9 0.1 2.1 0.8 7.4 . 17
2 1.5 0.3 0.08 0.07 0.004 0.2
348 6.5 21.6 0.1 7.6 2.5 10.9
.
River PROD POOLSIZE ABUND VOL BIOMASS SGR ORG-C CHLA TEMP Eh
. .
2
.
98 10 96 12 96 17 61 2 75 .
.
1235 2780 10.3 0.2 0.4 1.3 9.4 89.5 20.6 .
30
.
0.2 226 0.01 0.05 0.003 0.02 0.05 64.3 8 .
20726 7103 15.4 0.3 2.9 5.9 55.8 114.6 32
.
were again o b s e r v e d in sediments harboring higher bacterial production rates. H u d s o n and R o f f (in press in C J F A S ) reported highest specific growth rates (5.9 d a y - 1 ) in sandy sediments o f an Ontario stream, whereas lowest specific growth rates (0.02 d a y - a ) were m e a s u r e d in the H u d s o n R i v e r [4].
Bacterial
Production
Bacterial p r o d u c t i o n was significantly related to bacterial b i o m a s s only in m a r i n e sediments (Fig. l a ) , where bacterial b i o m a s s explained 48% o f the variability in
84
B.C. Sander and J. Kalff
100000
A
1o000 E (5
• "
•
:
q: ° " • [] I
IOOO
E
•
[]
ea a
~ ~
[]
.ram
marine
d~o,.n:t'~baE-i:n~i~ a"~_
m
~
" I~aD•
a
e-
.o +5
100
[]
a e ~L~J.CrV~ ~ q2~l~l
'~- ' F rCH i~ - "r-
,A
"0
.o C3.
+o
•
°
.
=
