Microb Ecol (1995) 29:231-248
MICROBIAL ECOLOGY © 1995 Springer-Verlag New York Inc.
Attached and Free-Living Bacteria: Production and Polymer Hydrolysis During a Diatom Bloom M. Middelboe, .1 M. SCndergaard,1 Y. Letarte,1 N.H. Borch *.2 1Freshwater Biological Laboratory, University of Copenhagen, 51 HelsingCrsgade, DK-3400 Hillercd, Denmark ZRoskilde University Library, P.O. Box 258, DK-4000 Roskilde, Denmark Received: 21 July 1994; Revised: 19 September 1994
Abstract. Abundance, production and extracellular enzymatic activity of free-living and attached bacteria were measured during the development and collapse of a spring bloom in a eutrophic lake. Free-living bacteria accounted for most of the total bacterial production during the first part of the bloom. Their production had a significant positive correlation to chlorophyll (P < .01) and polysaccharide concentration (P < .02) and to potential [3-glucosidase and aminopeptidase activity (P < .05), suggesting that algal release of dissolved polymeric compounds provided an important carbon source for bacterial production. As the bloom collapsed, we observed a change in the activity and structure of the microbial community. The mean contribution of attached bacteria to total bacterial production increased from 12% during the first part of the bloom to 26% at the end. Also, the extracellular enzymatic activity of attached bacteria increased as the bloom collapsed and constituted up to 75% of the total hydrolytic activity. An estimated disparity between hydrolytic activity and the corresponding carbon demand of attached bacteria suggested a net release of dissolved organic compounds from organic particles via polymer hydrolysis by attached bacteria. Introduction The carbon flow from phytoplankton to bacteria may result from algal exudation, cell lysis, and zooplankton grazing [e.g., 2, 5, 26] and from bacterial colonization and hydrolysis of senescent algal cells and larger aggregates [20, 40, 41, 44]. The relative importance of these pathways varies with the rates of the dominating
*Present address: The Royal Veterinary and Agricultural University, Department of Ecology and Molecular Biology, Microbiology Section, Rolighedsvej 21, DK-1958 Frederiksberg C, Copenhagen, Denmark. **Present address: University of Delaware, College of Marine Studies, Lewes, Delaware 19958, USA Correspondence to: M. Middelboe
232
M. Middelboeet al.
biological processes generating bacterial substrates. Such variations in substrate supply to the bacteria may contribute to a temporal uncoupling between phytoplankton and bacterioplankton production. Several studies have investigated the coupling between the development of phytoplankton spring blooms and the resulting increase in bacterial biomass. Although an almost immediate response in bacterial abundance or production has been observed in a few studies [e.g., 3, 27], a distinct delay in bacterial biomass relative to the increase in phytoplankton biomass is often found. This delay has been observed within a time scale of 7-14 days in temperate lakes and coastal areas [e.g., 5, 9, 25, 32, 43] and up to 30 days in Antarctic waters [5]. The delay in bacterial development has been used to model bacterial growth during a phytoplankton spring bloom [5]. The model proposed that bacterial substrates were provided by a combination of algal exudates (predominantly monomers) and lysis products (predominantly macromolecules) released directly from senescent algal cells and by zooplankton. To explain the uncoupling between phytoplankton and bacterioplankton development, it was suggested that polymeric lysis products released during the breakdown of the bloom, rather than exudates, were the major source of substrate for the bacteria [5]. The model implies that hydrolysis of dissolved polymeric substrates is the major factor controlling bacterial growth during a spring bloom, given that qualitative differences between exudates and lysis products are real. Furthermore, it implies that the delay in bacterial biomass production in relation to phytoplankton development is determined by the time lag between initiation and collapse of the bloom and by bacterial affinity for the released polymeric substrates. Because the mobilization of bacterial extracellular enzymes apparently is fast (hours), a prolonged induction period does not seem to offer an explanation for the delay [8, 36]. The carbon flow via bacterial attachment and subsequent degradation of organic particles was not specified in the model [5]. Attached bacteria generally represent less than 10% of total pelagic bacteria [16, 17, 29-31, 41, 50] and are often considered of minor importance in the annual carbon budget in pelagic marine and freshwater systems [29, 41]. However, a number of observations indicate that attached bacteria may periodically occupy a dominant role in pelagic carbon turnover [4]. Highly variable bacterial production values for the >3-txm fraction in eight Danish lakes have been observed [33] without correlation with the trophic status of the lakes. These data were, however, based on a single measurement in each lake. Seasonal studies have shown the contribution of attached bacteria to increase during the breakdown of phytoplankton blooms [14, 37, 41]. In general, attached bacteria are larger and have significantly higher cell-specific uptake rates of monomeric carbohydrates and amino acids than free-living bacteria [16, 29, 30, 31, 37, 40, 41]. Indications of high cell-specific extracellular enzymatic activity of aggregated bacteria without a corresponding uptake of hydrolysates [19, 28, 44] have further emphasized the role of attached bacteria in particle degradation and, possibly, in the supply of dissolved organic compounds to the free-living bacteria [19, 20, 28, 44]. Thus, the interactions between bacteria and phytoplankton may be more complex than considered in most previous studies, especially in eutrophic systems characterized by high concentrations of particulate organic material. A
Bacterial Production and Polymer Hydrolysis
233
coherent description of the different processes that make up the overall flow of carbon from algae to bacteria is, therefore, needed. In this study, we focused on the growth and hydrolytic properties of free-living and attached bacteria and their coupling to development and breakdown of a phytoplankton spring bloom in a eutrophic lake. We followed bacterial abundance, production, and extracellular enzymatic activity in size-fractionated water samples in relation to substrate concentration and composition. Materials and Methods Sampling Water was sampled in the morning in eutrophic Frederiksborg Slotss¢, Denmark, from 25 February to 26 May 1992. Frederiksborg Slotss¢ is a shallow, dimictic lake with a mean depth of 3.1 m, and a hydraulic mean residence time of about 4 years. The annual primary production is about 400 g C m -2, and the dissolved organic carbon (DOC) is mainly of autochthonous origin [48]. All water samples were taken from the same site with a Ruttner sampler from four depths (0, 2, 4, and 6 rn) and mixed to represent the whole water column, which remained unstratified throughout the study. Subsamples for bacterial abundance, 3H-thymidine incorporation and extracellular enzymatic activity were taken from a total and from a 3.0-p~m pore size prefiltered (47-mm polycarbonate filters) water sample. To prevent cell damage and the possibility of forcing some large cells through the filter pores, fractionation was done at very low vacuum (
MICROBIAL ECOLOGY © 1995 Springer-Verlag New York Inc.
Attached and Free-Living Bacteria: Production and Polymer Hydrolysis During a Diatom Bloom M. Middelboe, .1 M. SCndergaard,1 Y. Letarte,1 N.H. Borch *.2 1Freshwater Biological Laboratory, University of Copenhagen, 51 HelsingCrsgade, DK-3400 Hillercd, Denmark ZRoskilde University Library, P.O. Box 258, DK-4000 Roskilde, Denmark Received: 21 July 1994; Revised: 19 September 1994
Abstract. Abundance, production and extracellular enzymatic activity of free-living and attached bacteria were measured during the development and collapse of a spring bloom in a eutrophic lake. Free-living bacteria accounted for most of the total bacterial production during the first part of the bloom. Their production had a significant positive correlation to chlorophyll (P < .01) and polysaccharide concentration (P < .02) and to potential [3-glucosidase and aminopeptidase activity (P < .05), suggesting that algal release of dissolved polymeric compounds provided an important carbon source for bacterial production. As the bloom collapsed, we observed a change in the activity and structure of the microbial community. The mean contribution of attached bacteria to total bacterial production increased from 12% during the first part of the bloom to 26% at the end. Also, the extracellular enzymatic activity of attached bacteria increased as the bloom collapsed and constituted up to 75% of the total hydrolytic activity. An estimated disparity between hydrolytic activity and the corresponding carbon demand of attached bacteria suggested a net release of dissolved organic compounds from organic particles via polymer hydrolysis by attached bacteria. Introduction The carbon flow from phytoplankton to bacteria may result from algal exudation, cell lysis, and zooplankton grazing [e.g., 2, 5, 26] and from bacterial colonization and hydrolysis of senescent algal cells and larger aggregates [20, 40, 41, 44]. The relative importance of these pathways varies with the rates of the dominating
*Present address: The Royal Veterinary and Agricultural University, Department of Ecology and Molecular Biology, Microbiology Section, Rolighedsvej 21, DK-1958 Frederiksberg C, Copenhagen, Denmark. **Present address: University of Delaware, College of Marine Studies, Lewes, Delaware 19958, USA Correspondence to: M. Middelboe
232
M. Middelboeet al.
biological processes generating bacterial substrates. Such variations in substrate supply to the bacteria may contribute to a temporal uncoupling between phytoplankton and bacterioplankton production. Several studies have investigated the coupling between the development of phytoplankton spring blooms and the resulting increase in bacterial biomass. Although an almost immediate response in bacterial abundance or production has been observed in a few studies [e.g., 3, 27], a distinct delay in bacterial biomass relative to the increase in phytoplankton biomass is often found. This delay has been observed within a time scale of 7-14 days in temperate lakes and coastal areas [e.g., 5, 9, 25, 32, 43] and up to 30 days in Antarctic waters [5]. The delay in bacterial development has been used to model bacterial growth during a phytoplankton spring bloom [5]. The model proposed that bacterial substrates were provided by a combination of algal exudates (predominantly monomers) and lysis products (predominantly macromolecules) released directly from senescent algal cells and by zooplankton. To explain the uncoupling between phytoplankton and bacterioplankton development, it was suggested that polymeric lysis products released during the breakdown of the bloom, rather than exudates, were the major source of substrate for the bacteria [5]. The model implies that hydrolysis of dissolved polymeric substrates is the major factor controlling bacterial growth during a spring bloom, given that qualitative differences between exudates and lysis products are real. Furthermore, it implies that the delay in bacterial biomass production in relation to phytoplankton development is determined by the time lag between initiation and collapse of the bloom and by bacterial affinity for the released polymeric substrates. Because the mobilization of bacterial extracellular enzymes apparently is fast (hours), a prolonged induction period does not seem to offer an explanation for the delay [8, 36]. The carbon flow via bacterial attachment and subsequent degradation of organic particles was not specified in the model [5]. Attached bacteria generally represent less than 10% of total pelagic bacteria [16, 17, 29-31, 41, 50] and are often considered of minor importance in the annual carbon budget in pelagic marine and freshwater systems [29, 41]. However, a number of observations indicate that attached bacteria may periodically occupy a dominant role in pelagic carbon turnover [4]. Highly variable bacterial production values for the >3-txm fraction in eight Danish lakes have been observed [33] without correlation with the trophic status of the lakes. These data were, however, based on a single measurement in each lake. Seasonal studies have shown the contribution of attached bacteria to increase during the breakdown of phytoplankton blooms [14, 37, 41]. In general, attached bacteria are larger and have significantly higher cell-specific uptake rates of monomeric carbohydrates and amino acids than free-living bacteria [16, 29, 30, 31, 37, 40, 41]. Indications of high cell-specific extracellular enzymatic activity of aggregated bacteria without a corresponding uptake of hydrolysates [19, 28, 44] have further emphasized the role of attached bacteria in particle degradation and, possibly, in the supply of dissolved organic compounds to the free-living bacteria [19, 20, 28, 44]. Thus, the interactions between bacteria and phytoplankton may be more complex than considered in most previous studies, especially in eutrophic systems characterized by high concentrations of particulate organic material. A
Bacterial Production and Polymer Hydrolysis
233
coherent description of the different processes that make up the overall flow of carbon from algae to bacteria is, therefore, needed. In this study, we focused on the growth and hydrolytic properties of free-living and attached bacteria and their coupling to development and breakdown of a phytoplankton spring bloom in a eutrophic lake. We followed bacterial abundance, production, and extracellular enzymatic activity in size-fractionated water samples in relation to substrate concentration and composition. Materials and Methods Sampling Water was sampled in the morning in eutrophic Frederiksborg Slotss¢, Denmark, from 25 February to 26 May 1992. Frederiksborg Slotss¢ is a shallow, dimictic lake with a mean depth of 3.1 m, and a hydraulic mean residence time of about 4 years. The annual primary production is about 400 g C m -2, and the dissolved organic carbon (DOC) is mainly of autochthonous origin [48]. All water samples were taken from the same site with a Ruttner sampler from four depths (0, 2, 4, and 6 rn) and mixed to represent the whole water column, which remained unstratified throughout the study. Subsamples for bacterial abundance, 3H-thymidine incorporation and extracellular enzymatic activity were taken from a total and from a 3.0-p~m pore size prefiltered (47-mm polycarbonate filters) water sample. To prevent cell damage and the possibility of forcing some large cells through the filter pores, fractionation was done at very low vacuum (
