Microb Ecol (1995) 30:89-104
MICROBIAL ECOLOGY © 1995Springer-VerlagNewYorkInc.
Growth Limitation of Planktonic Bacteria in a Large Mesotrophic L a k e B. Schweitzer, M. Simon Limnological Institute, University of Constance, Konstanz, Germany Received: 23 September 1994; Revised: 1 December 1994
Abstract. We studied nutrient-limitation of bacterioplankton growth in Lake Constance, a mesotrophic lake, between February and August in 1992. We amended 1-txm filtrates with a single nutrient or nutrient combinations at 5 or 10 ~M final concentration, and the limiting nutrient or nutrient combination was inferred from the assay in which bacterial growth was most stimulated. The following nutrients were added individually or in combination: glucose, amino acids, peptone, and ammonium as C and N sources, and inorganic phosphate. From January until the beginning of the phytoplankton spring bloom in mid-April, C alone was growth-limiting. During the spring bloom a complex growth-limitation pattern occurred; first P was limiting, then for only 1 week C + N together, and thereafter P + C. During the clear-water phase with very low chlorophyll concentrations, P + C together limited bacterial growth again, interrupted by a period when C + N + P shortage caused a triple limitation. Later in the season, P + C were growth-limiting again. The growth efficiency (bacterial biomass prodnced/substrates used) on the basis of amino acid and carbohydrate used varied between 17 and 35%. The addition of various C and N sources indicated that the growth efficiency strongly depended on the quality of the substrates and the adaptation of the bacterial assemblages, for example, whether C and N originated from amino acids or glucose and ammonium. Introduction It is now well established that heterotrophic planktonic bacteria are important in the flux and cycling of organic and inorganic matter and energy in pelagic environments. They are the prime component of the food web in using dissolved organic matter (DOM) and are strong competitors with phytoplankton for inorganic nutrients [10, 29, 35]. An important issue in using DOM and inorganic nutrients such as phosphate and ammonium is the relation between their availability and demand
Correspondence to: B. Schweitzer, Limnological Institute, University of Constance, P.O. Box 5560 X 913, D-78434 Konstanz, Germany.
90
B. Schweitzer, M. Simon
for biosynthetic and energetic processes. The discrepancy between the relative demand and availability of the various elements C, N, P, usually controls which one limits growth unless temperature is growth-limiting [1]. The demand for the major C and N sources for biosynthetic processes is fairly constant, because the C/ N ratio in planktonic bacteria remains quite stable under various growth conditions, ranging only between 4 and 5.5 [6, 16]. The C/P ratio varies more, between 40 and 80, depending on whether P is growth limiting or not [10]. The C/N and C/P ratio in available substrates presumably is also variable, although it is difficult to estimate these ratios. Available substrates usually cannot be satisfactorily separated from recalcitrant substrates for chemical analysis [11, 41]. Another question is also in which chemical form the elements are available, for example, whether C and N are used as amino acids or as carbohydrates and ammonium. Kirchman [12] and Kirchman et al. [13] have shown that in the oligotrophic subarctic Pacific, heterotrophic bacteria prefer amino acids against glucose and ammonium. There is also evidence that the growth efficiency of planktonic bacteria depends on the chemical form of the available C source. Middelboe and SCndergaard [19] showed that the growth efficiency is substantially reduced when bacteria use high amounts of polymeric carbohydrates compared to other situations when they do not need to hydrolyze polysaccharides. It is conceivable that the growth efficiency also depends on the relative concentration and availability of carbohydrates, ammonium, and amino acids. So far it has only been shown that the growth efficiency is inversely related to the growth rate, which varies substantially during different seasons [15, 18]. Several recent studies have shown that growth of limnetic planktonic bacteria often is P limited [3, 20, 37, 39]. This is in contrast to the traditional concept that heterotrophic planktonic bacteria are usually C limited [26]. Morris and Lewis [20] even present evidence that two elements can limit bacterial growth simultaneously. We measured the growth limitation of planktonic bacteria in mesotrophic Lake Constance during different seasonal situations. We added C, N, and P individually or in various combinations to the natural bacterial assemblages filtered through 1 txm and examined their growth response. The main results show that growth of planktonic bacteria during the growing season was stimulated best by P additions and, thus, indicate that P limited growth but often in combination with C and in some situations even with C + N. We also estimated the growth efficiency and its control by availability of various nutrients and growth rates. The results show that the growth efficiency strongly depends on the chemical form of the available nutrients.
Materials and Methods The study was performed between April and August 1992 in mesotrophic Lake Constance (Bodensee), Germany. Lake Constance is a large (479 km 2) and deep (max. depth 252 m) prealpine lake in which the phosphorus load has been decreased substantially during the last 12 years, such that it became mesotrophic again [34]. Samples were collected with a clean 9-liter van Dorn bottle at 3 m at a central station in the Llberlinger See, the northwestern part of the lake (maximum depth 147 rn). Samples were kept in carefully rinsed 2-liter polyethylene bottles in a cooling box
Growth Limitation of Planktonic Bacteria
91
and transported immediately to the laboratory. Further processing of the samples started within 2 h after sampling,
Experimental Design and Sample Preparation To identify the growth-limiting nutrient, we compared growth of planktonic bacteria in 1-~m filtrates enriched with various C, N, and P sources (Table 1). We considered the most limiting nutrient or combination of nutrients to be that which stimulated bacterial growth the most. The samples were first filtered by gravity through 1-1xm Nuclepore filters. To prevent clogging, a filter was replaced when the filtration rate started to decrease substantially, usually after 200-300 ml. In total, 4-6 liters of sample water were needed for each experiment. We therefore used three 47-mm filtration units in parallel to reduce the time of filtration. Before and after filtration, the samples were kept on ice until the start of the experiment. The filtration device we used did not contaminate the samples by dissolved amino acids and removed all eukaryotic protists [28]. To minimize contamination, all filtration and further processing was done with gloves. Aliquots of 250 ml of the filtrates were subdivided into combusted Erlenmeyer flasks. To each flask, a single nutrient or a mixture of two or three nutrients was added from sterile stock solutions according to Table 1. As a C source, we used glucose; as an N source, ammonium (NH4C1); and as a P source, phosphate (NaRHPO4). In addition, in some experiments we also added as combined C and N sources dissolved free amino acids (DFAA) in a molar composition typical of the lake during the growing season [32] or peptone. Additions of all nutrients were 5 or 10 pXM each, except for phosphate, which was added at 3.5 or 7 IXM. Background concentrations in Lake Constance of total dissolved combined amino acids range between 0.8 and 2.5-p,M amino acid equivalents [28], and those of total dissolved carbohydrates range between 1.7 and 5.5-txM glucose equivalents [7]. The filtrates with the nutrients added and a control without any addition were incubated at in situ temperature in the dark for 36-48 h. Incubation usually started within 2 h after filtration. Subsamples for bacterial counts and nutrient analyses (see below) were withdrawn periodically. One subsample of 30 ml for analysis of ammonium, soluble reactive phosphorus (SRP), and carbohydrates was removed by a combusted pipette. The sample was gently filtered through a 0.2pum Nuclepore filter into an acid-rinsed 30-ml polyethylene bottle. To prevent contamination by ammonium, the bottles were filled to the top and capped immediately. A 5-ml subsample for analysis of amino acids was filtered through a 0.2-p~m Gelman Acrodisc with low protein-binding capacity (tuffrin) into a combusted glass vial. Another 5-ml subsample for enumerating bacteria was fixed with 2% formalin, final concentration. All glassware was combusted for 1 h at 550°C. Polyethylene bottles for storing subsamples were acid-washed and carefully rinsed with Milli-Q water. Samples except those for bacterial numbers were stored at -20°C until analysis within 4 weeks.
Sample Analysis Bacterial abundance was enumerated by epifluorescence microscopy after staining with 4'-6'-diamidino-2-phenylindole (DAPI) [25]. In total, between 300 and 400 cells were counted in 10 randomly chosen grids. The CV (SE/mean) varied between 0.08 and 0.15. Analysis of total dissolved amino acids as free amino acids was done after hydrolysis with 6 N HC1 (110°C for 20 h) with high-performance liquid chromatography after o-phthaldialdehyde precolumn derivatization according to Lindroth and Mopper [17] and as modified by Simon and Rosenstock [32]. Before hydrolysis, 20 pA of ascorbic acid (2 mg/ml) were added to the sample to prevent amino acid oxidation by nitrate [27]. The CV of triplicate analyses was
MICROBIAL ECOLOGY © 1995Springer-VerlagNewYorkInc.
Growth Limitation of Planktonic Bacteria in a Large Mesotrophic L a k e B. Schweitzer, M. Simon Limnological Institute, University of Constance, Konstanz, Germany Received: 23 September 1994; Revised: 1 December 1994
Abstract. We studied nutrient-limitation of bacterioplankton growth in Lake Constance, a mesotrophic lake, between February and August in 1992. We amended 1-txm filtrates with a single nutrient or nutrient combinations at 5 or 10 ~M final concentration, and the limiting nutrient or nutrient combination was inferred from the assay in which bacterial growth was most stimulated. The following nutrients were added individually or in combination: glucose, amino acids, peptone, and ammonium as C and N sources, and inorganic phosphate. From January until the beginning of the phytoplankton spring bloom in mid-April, C alone was growth-limiting. During the spring bloom a complex growth-limitation pattern occurred; first P was limiting, then for only 1 week C + N together, and thereafter P + C. During the clear-water phase with very low chlorophyll concentrations, P + C together limited bacterial growth again, interrupted by a period when C + N + P shortage caused a triple limitation. Later in the season, P + C were growth-limiting again. The growth efficiency (bacterial biomass prodnced/substrates used) on the basis of amino acid and carbohydrate used varied between 17 and 35%. The addition of various C and N sources indicated that the growth efficiency strongly depended on the quality of the substrates and the adaptation of the bacterial assemblages, for example, whether C and N originated from amino acids or glucose and ammonium. Introduction It is now well established that heterotrophic planktonic bacteria are important in the flux and cycling of organic and inorganic matter and energy in pelagic environments. They are the prime component of the food web in using dissolved organic matter (DOM) and are strong competitors with phytoplankton for inorganic nutrients [10, 29, 35]. An important issue in using DOM and inorganic nutrients such as phosphate and ammonium is the relation between their availability and demand
Correspondence to: B. Schweitzer, Limnological Institute, University of Constance, P.O. Box 5560 X 913, D-78434 Konstanz, Germany.
90
B. Schweitzer, M. Simon
for biosynthetic and energetic processes. The discrepancy between the relative demand and availability of the various elements C, N, P, usually controls which one limits growth unless temperature is growth-limiting [1]. The demand for the major C and N sources for biosynthetic processes is fairly constant, because the C/ N ratio in planktonic bacteria remains quite stable under various growth conditions, ranging only between 4 and 5.5 [6, 16]. The C/P ratio varies more, between 40 and 80, depending on whether P is growth limiting or not [10]. The C/N and C/P ratio in available substrates presumably is also variable, although it is difficult to estimate these ratios. Available substrates usually cannot be satisfactorily separated from recalcitrant substrates for chemical analysis [11, 41]. Another question is also in which chemical form the elements are available, for example, whether C and N are used as amino acids or as carbohydrates and ammonium. Kirchman [12] and Kirchman et al. [13] have shown that in the oligotrophic subarctic Pacific, heterotrophic bacteria prefer amino acids against glucose and ammonium. There is also evidence that the growth efficiency of planktonic bacteria depends on the chemical form of the available C source. Middelboe and SCndergaard [19] showed that the growth efficiency is substantially reduced when bacteria use high amounts of polymeric carbohydrates compared to other situations when they do not need to hydrolyze polysaccharides. It is conceivable that the growth efficiency also depends on the relative concentration and availability of carbohydrates, ammonium, and amino acids. So far it has only been shown that the growth efficiency is inversely related to the growth rate, which varies substantially during different seasons [15, 18]. Several recent studies have shown that growth of limnetic planktonic bacteria often is P limited [3, 20, 37, 39]. This is in contrast to the traditional concept that heterotrophic planktonic bacteria are usually C limited [26]. Morris and Lewis [20] even present evidence that two elements can limit bacterial growth simultaneously. We measured the growth limitation of planktonic bacteria in mesotrophic Lake Constance during different seasonal situations. We added C, N, and P individually or in various combinations to the natural bacterial assemblages filtered through 1 txm and examined their growth response. The main results show that growth of planktonic bacteria during the growing season was stimulated best by P additions and, thus, indicate that P limited growth but often in combination with C and in some situations even with C + N. We also estimated the growth efficiency and its control by availability of various nutrients and growth rates. The results show that the growth efficiency strongly depends on the chemical form of the available nutrients.
Materials and Methods The study was performed between April and August 1992 in mesotrophic Lake Constance (Bodensee), Germany. Lake Constance is a large (479 km 2) and deep (max. depth 252 m) prealpine lake in which the phosphorus load has been decreased substantially during the last 12 years, such that it became mesotrophic again [34]. Samples were collected with a clean 9-liter van Dorn bottle at 3 m at a central station in the Llberlinger See, the northwestern part of the lake (maximum depth 147 rn). Samples were kept in carefully rinsed 2-liter polyethylene bottles in a cooling box
Growth Limitation of Planktonic Bacteria
91
and transported immediately to the laboratory. Further processing of the samples started within 2 h after sampling,
Experimental Design and Sample Preparation To identify the growth-limiting nutrient, we compared growth of planktonic bacteria in 1-~m filtrates enriched with various C, N, and P sources (Table 1). We considered the most limiting nutrient or combination of nutrients to be that which stimulated bacterial growth the most. The samples were first filtered by gravity through 1-1xm Nuclepore filters. To prevent clogging, a filter was replaced when the filtration rate started to decrease substantially, usually after 200-300 ml. In total, 4-6 liters of sample water were needed for each experiment. We therefore used three 47-mm filtration units in parallel to reduce the time of filtration. Before and after filtration, the samples were kept on ice until the start of the experiment. The filtration device we used did not contaminate the samples by dissolved amino acids and removed all eukaryotic protists [28]. To minimize contamination, all filtration and further processing was done with gloves. Aliquots of 250 ml of the filtrates were subdivided into combusted Erlenmeyer flasks. To each flask, a single nutrient or a mixture of two or three nutrients was added from sterile stock solutions according to Table 1. As a C source, we used glucose; as an N source, ammonium (NH4C1); and as a P source, phosphate (NaRHPO4). In addition, in some experiments we also added as combined C and N sources dissolved free amino acids (DFAA) in a molar composition typical of the lake during the growing season [32] or peptone. Additions of all nutrients were 5 or 10 pXM each, except for phosphate, which was added at 3.5 or 7 IXM. Background concentrations in Lake Constance of total dissolved combined amino acids range between 0.8 and 2.5-p,M amino acid equivalents [28], and those of total dissolved carbohydrates range between 1.7 and 5.5-txM glucose equivalents [7]. The filtrates with the nutrients added and a control without any addition were incubated at in situ temperature in the dark for 36-48 h. Incubation usually started within 2 h after filtration. Subsamples for bacterial counts and nutrient analyses (see below) were withdrawn periodically. One subsample of 30 ml for analysis of ammonium, soluble reactive phosphorus (SRP), and carbohydrates was removed by a combusted pipette. The sample was gently filtered through a 0.2pum Nuclepore filter into an acid-rinsed 30-ml polyethylene bottle. To prevent contamination by ammonium, the bottles were filled to the top and capped immediately. A 5-ml subsample for analysis of amino acids was filtered through a 0.2-p~m Gelman Acrodisc with low protein-binding capacity (tuffrin) into a combusted glass vial. Another 5-ml subsample for enumerating bacteria was fixed with 2% formalin, final concentration. All glassware was combusted for 1 h at 550°C. Polyethylene bottles for storing subsamples were acid-washed and carefully rinsed with Milli-Q water. Samples except those for bacterial numbers were stored at -20°C until analysis within 4 weeks.
Sample Analysis Bacterial abundance was enumerated by epifluorescence microscopy after staining with 4'-6'-diamidino-2-phenylindole (DAPI) [25]. In total, between 300 and 400 cells were counted in 10 randomly chosen grids. The CV (SE/mean) varied between 0.08 and 0.15. Analysis of total dissolved amino acids as free amino acids was done after hydrolysis with 6 N HC1 (110°C for 20 h) with high-performance liquid chromatography after o-phthaldialdehyde precolumn derivatization according to Lindroth and Mopper [17] and as modified by Simon and Rosenstock [32]. Before hydrolysis, 20 pA of ascorbic acid (2 mg/ml) were added to the sample to prevent amino acid oxidation by nitrate [27]. The CV of triplicate analyses was
