|Home | About | Journals | Submit | Contact Us | Français|
The distribution of the phylogenetically narrow R-BT065 cluster (Betaproteobacteria) in 102 freshwater lakes, reservoirs, and various ponds located in central Europe (a total of 122 samples) was examined by using a cluster-specific fluorescence in situ hybridization probe. These habitats differ markedly in pH, conductivity, trophic status, surface area, altitude, bedrock type, and other limnological characteristics. Despite the broad ecological diversity of the habitats investigated, the cluster was detected in 96.7% of the systems, and its occurrence was not restricted to a certain habitat type. However, the relative proportions of the cluster in the total bacterioplankton were significantly lower in humic and acidified lakes than in pH-neutral or alkaline habitats. On average, the cluster accounted for 9.4% of the total bacterioplankton (range, 0 to 29%). The relative abundance and absolute abundance of these bacteria were significantly and positively related to higher pH, conductivity, and the proportion of low-molecular-weight compounds in dissolved organic carbon (DOC) and negatively related to the total DOC and dissolved aromatic carbon contents. Together, these parameters explained 55.3% of the variability in the occurrence of the cluster. Surprisingly, no clear relationship of the R-BT065 bacteria to factors indicating the trophic status of habitats (i.e., different forms of phosphorus and chlorophyll a content) was found. Based on our results and previously published data, we concluded that the R-BT065 cluster represents a ubiquitous, highly active segment of bacterioplankton in nonacidic lakes and ponds and that alga-derived substrates likely form the main pool of substrates responsible for its high growth potential and broad distribution in freshwater habitats.
Heterotrophic bacterioplankton assemblages found in a broad variety of freshwater ecosystems are frequently dominated by representatives of a few phylogenetic clusters of Betaproteobacteria and Actinobacteria (1, 7, 9, 17, 21, 23, 34, 40). One of these clusters is the phylogenetically defined R-BT065 group (34), which is currently represented by >700 environmental ribosomal sequences deposited in the GenBank database. These sequences were obtained in several nonquantitative diversity studies of many European and North American freshwater habitats (5, 10, 18, 27, 30, 42). The R-BT065 cluster is characterized by a minimum level of intragroup 16S rRNA sequence similarity of 97.3%, and investigations of recently cultivated strains demonstrated that members of the group are heterotrophic bacteria (V. Kasalický, J. Jezbera, K. Šimek, and M. W. Hahn, submitted for publication). This cluster forms a phylogenetically distinct subgroup of the so-called “Rhodoferax sp. BAL47” cluster (42). The new genus Limnohabitans was recently established for some strains affiliated with the “Rhodoferax sp. BAL47” cluster (11), and it has to be assumed that bacteria belonging to the R-BT065 cluster are also affiliated with this genus (Kasalický and coworkers, submitted for publication).
Bacteria belonging to the R-BT065 cluster in natural systems can be quantified using a homonymous fluorescence in situ hybridization (FISH) probe (34). This probe enabled intensive investigations of the ecology of R-BT065 bacteria; however, only a few freshwater systems (mainly representing larger pH-neutral lakes and reservoirs) have been quantitatively investigated to determine the presence of this bacterial group so far (2, 13, 24, 27, 30, 30). In the few habitats investigated, R-BT065 bacteria comprised 3 to 50% of the total bacteria. Importantly, these bacteria consistently had the highest net growth rates and showed high levels of substrate uptake in a freshwater reservoir (12, 13, 31, 32), as well as in an alpine lake (27). These investigations indicated that the ecology of the R-BT065 group is rather uniform and is characterized by (i) the potential for a rapid response to environmental changes, (ii) a high level of vulnerability to flagellate predation (14, 35), and (iii) a strong link with phytoplankton-derived organic material as the key growth substrate for the bacteria (24, 25, 33). However, these findings were based on quantitative investigations of only a few freshwater systems, and we can only speculate that the findings could be generalized for populations of this bacterial group in systems with different limnological characteristics. In particular, investigations of the habitat range, as well as investigations of environmental factors modulating the abundance of R-BT065 bacteria, have not been performed. Therefore, we set out to conduct a comprehensive survey of a broad variety of freshwater habitats that could support our hypothesis concerning the potential general importance of the cluster.
The specific aims of this study were (i) to quantify the R-BT065 bacteria in the surface waters of a large variety of central European freshwater ecosystems (102 habitats) spanning broad ranges of system type and size (from large and deep alpine lakes to small and shallow ponds or various fishponds), type of bedrock, altitude, trophic status, pH, and humic substance content; (ii) to determine the proportion of R-BT065 bacteria in the total bacteria and in the total Betaproteobacteria in these habitats using FISH probes; and (iii) to suggest major physical, chemical, and biological factors that influence the occurrence of the R-BT065 phylotypes in the bacterioplankton.
Altogether, 102 study sites covering a wide array of freshwater habitats in central Europe (Czech Republic and Austria) (Fig. (Fig.1),1), including alpine and prealpine lakes, humic lakes, anthropogenically acidified lakes, various types of ponds, reservoirs, large flooded sand pits, and small water holes, were sampled (see Table S1 in the supplemental material for a full list of the habitats sampled). Geographically, these habitats form four groups (Fig. (Fig.1):1): (i) ponds and reservoirs in central Bohemia, west and southwest of Prague (Czech Republic); (ii) anthropogenically acidified lakes located in southwest Bohemia in Šumava National Park (Czech Republic); (iii) fishponds, forest and humic ponds, and flooded sand pits in South Bohemia (Czech Republic); and (iv) prealpine and alpine lakes and small ponds in Austria located mostly in the Salzkammergut area (Fig. (Fig.1)1) except a few lakes located in the central part of the Austrian Alps (Niedere and Hohe Tauern).
The habitats were sampled during 2006, 2007, and 2008 from April to November. Samples were usually collected between 10:30 a.m. and 2:30 p.m. Twenty habitats were sampled twice, either during different parts of the same year (e.g., late spring after snow melt and midsummer for small mountain ponds) or at different stages of the seasonal plankton development in two consecutive years. Thus, a total of 122 samples from 102 habitats were analyzed. Water samples were usually taken from the surface layer of the epilimnion (at a depth of 0.5 m or just below the surface layer in water shallower than 0.5 m) using a 1-liter polyethylene bottle sampler attached to a retractable stick or directly with sterile, 1-liter glass Schott bottles. Samples were taken directly from the shore (habitats with a small surface area) or, in the case of larger lakes and reservoirs, from a boat, pier, or a bridge if available. Thus, samples taken from most of the habitats should be considered littoral samples. The in situ parameters measured included temperature and dissolved oxygen content (both determined with a WTW OXI196 probe [Weilheim, Germany]), as well as pH and conductivity (determined with a WTW Multi340i device [Weilheim, Germany]). For microbiological and chemical analyses, water samples were transported to the laboratory and processed immediately upon arrival (usually within 2 h after sampling).
Water samples collected in 2006 and 2007 (71 samples [for details see Table S1 in the supplemental material]) were characterized by determining the temperature, pH, concentration of oxygen, conductivity, and water absorption spectra, which were also used for estimation of the dissolved organic carbon (DOC) concentration (see below). In addition to this basic set of physical and chemical parameters, in 2008 (51 samples [see Table S1 in the supplemental material]) DOC, particulate organic carbon (POC), dissolved reactive phosphorus, total phosphorus (TP), particulate phosphorus, and chlorophyll a (Chl-a) concentrations in the samples were also determined (see below for details of the methods used). The latter parameters were also used to tentatively characterize the trophic status of the habitats studied. Since usually only one or two measurements of these parameters were available for each habitat, the characterization of the trophic status described below should be considered with caution.
Twenty-milliliter portions of samples were fixed with formaldehyde (final concentration, 2% [vol/vol]), 1- to 3-ml subsamples were filtered onto black 0.2-μm-pore-size filters (Osmonic Inc., Livermore, CA), stained with 4′,6′-diamidino-2-phenylindole (DAPI) (final concentration, 0.2% [wt/vol]), and enumerated by epifluorescence microscopy (Olympus AX 70) as described by Šimek et al. (34). At least 500 cells were enumerated in duplicate samples by using the area of at least 10 to 20 microscopic fields at a magnification of ×1,000.
Subsamples (10 to 20 ml) were fixed with a paraformaldehyde solution (final concentration, 2%) at 20°C for 6 h or at 4°C overnight. After fixation, samples were filtered onto 0.2-μm-pore-size, 47-mm-diameter Nuclepore filters (Osmonic Inc., Livermore, CA), rinsed with 1 ml of 1× phosphate-buffered saline and then with 1 ml of sterile MilliQ water, air dried, and stored at −20°C until they were processed. For in situ hybridization of bacteria on filter sections we used the catalyzed reporter deposition (CARD)-FISH protocol (26, 29). The following oligonucleotide probes (Thermo, Ulm, Germany) were employed: the BET42a probe for the Betaproteobacteria and the R-BT065 probe for the R-BT065 subcluster of the “Rhodoferax sp. BAL47” cluster of the Betaproteobacteria (42) (for the probe target see reference 34). The proportions of FISH-positive bacteria were determined directly by inspecting 600 to 1,000 cells in the replicate samples using epifluorescence microscopy (Olympus AX-70). The difference in the counts of FISH-positive bacteria on replicate filters was generally <8%.
Absorbance measurements were obtained with a Specord 210 dual-beam spectrophotometer (Analytik, Jena, Germany) at wavelengths between 250 and 440 nm using distilled water as the blank and samples filtered through glass fiber filters with a nominal pore size of 0.4 μm (GF-5; Macherey-Nagel, Düren, Germany). We used the following two indicators of the composition of dissolved organic matter (DOM) that were based on absorbance measurements. (i) Specific UV absorbance (A254/DOC) is defined as the UV absorbance of a water sample normalized for the DOC concentration and was shown to be a useful parameter for estimating the dissolved aromatic carbon content in aquatic systems (39). It was calculated by dividing the absorbance at 254 nm by the DOC concentration and was expressed in liters per milligram of carbon per meter. (ii) The ratio of absorbance at 250 nm to absorbance at 365 nm was used as a parameter that is inversely related to the molecular weight of DOM (22, 28).
DOC in the samples filtered through GF-5 filters was analyzed with a TOC 5000A analyzer (Shimadzu, Japan). Direct measurements were obtained only in 2008, whereas the DOC concentrations in the samples in 2006 and 2007 were calculated from a relationship between the DOC concentration and the absorbance at 250 nm measured in 2008 that yielded a highly significant linear correlation (r2 = 0.89, n = 51, P < 0.0001). We determined that the possible errors in the estimate of the DOC concentration obtained using the absorbance at 250 nm were generally less than 36% of the directly measured DOC values. However, since DOC concentrations spanned almost 2 orders of magnitude (Table (Table1),1), this error had a negligible effect on the overall trend of the relationship of R-BT065 bacteria to DOC concentrations shown in Fig. Fig.2.2. Particulate organic carbon (POC) on GF-5 filters was analyzed by high-temperature ignition with a TOC/SSM5000A analyzer (Shimadzu, Japan).
Dissolved reactive phosphorus was analyzed as described by Murphy and Riley (19). The total phosphorus content was determined by the molybdate method after perchloric acid digestion as described by Kopáček and Hejzlar (16). The particulate phosphorus content was determined by determining the difference between the TP content in the original samples and the TP content in the samples filtered through GF-5 filters.
Pearson's correlation analysis in the software GraphPad PRISM was used to examine direct relationships between the parameters studied. The numbers of R-BT065 bacteria were logarithmically transformed prior to the statistical analyses. Environmental factors influencing the occurrence of the R-BT065 bacteria in freshwater habitats were analyzed by multivariate analysis using the program CANOCO (38). Redundancy analysis (RDA) was carried out with centering and standardization by species. Forward selection was used to choose significant explanatory (environmental) variables. Variables were included (Fig. (Fig.2)2) when a P value of <0.05 was estimated by a Monte Carlo permutation test (999 unrestricted permutations). The results obtained with CANOCO were visualized by CanoDraw for Windows (38).
The locations of all of the habitats studied are shown in Fig. Fig.1.1. R-BT065 bacteria were detected in 98 of the 102 habitats investigated; these phylotypes were not found in only four humic lakes and ponds (Fig. (Fig.3;3; see Table S1 in the supplemental material). The proportions of the R-BT065 bacteria ranged from 0% to 29% of the total bacteria (Fig. (Fig.3).3). Notably, when the data for all 122 samples collected from the 102 habitats were averaged, the R-BT065 bacteria comprised 9.4% of the total bacteria. However, when the data for 90 samples from the pH-neutral habitats (i.e., when the data for humic ponds and lakes and acidified lakes were not included [Fig. [Fig.33 and and4])4]) were averaged, the target group accounted for 12.3% of the total bacteria.
Characteristics of the bodies of water, such as the global positioning satellite coordinates, altitude, surface area of the habitat, and basic physical and chemical background data (including pH, conductivity, temperature, water absorption spectra, phosphorus forms, and dissolved organic carbon concentration) are shown in Table S1 in the supplemental material.
The habitats sampled covered a wide range of freshwater systems in terms of surface area, altitude, pH, conductivity, trophic status, character of the bedrock, and degree of anthropogenic activities in the habitat proper and the watershed (e.g., remote alpine lakes versus heavily fertilized fishponds in South Bohemia). For instance, the altitudes of the habitats studied ranged from 264 m above sea level (Suchomasty Reservoir) to 2,309 m above sea level (Oberer Klaffersee), the depths ranged from a few centimeters (Holzteich Pond) to 171 m (Attersee), the surface areas ranged from a few square meters (puddles in the Wengen Moor raised bog) to 4,620 ha (Attersee), the pHs ranged from 3.9 to 9.1, the bacterial concentrations ranged from 0.20 × 106 to 20.1 × 106 cells ml−1, the DOC concentrations ranged from 1.7 to 38 mg liter−1, and the Chl-a concentrations ranged from 1 to 122 μg liter−1 (Table (Table1;1; see Table S1 in the supplemental material).
Geographically, the habitats form four major groups (Fig. (Fig.1;1; see Materials and Methods for details). However, a more detailed analysis of the occurrence of the R-BT065 phylotypes showed that there were substantial overlaps in some major background parameters among sites at locations that were geographically and geologically (e.g., granite versus limestone bedrock) very different. Thus, using redundancy analysis (Fig. (Fig.2),2), we related the absolute numbers of the target R-BT065 bacteria, as well as the relative proportions of these bacteria in the total bacteria and in the total Betaproteobacteria, to parameters available for all habitats. Conductivity, pH, altitude, oxygen concentration, DOC concentration, and the parameters tentatively indicating the character of the DOC (A250/A365 and A254/DOC) explained 61% of the variability in the absolute and relative numbers of the R-BT065 bacteria (Fig. (Fig.2).2). When only the parameters that are not autocorrelated were taken into account, the model still explained more than one-half (55.3%) of the variability (Table (Table2).2). Notably, the habitats examined in the analysis quite evenly covered the whole range of the core parameters detected as the driving forces shaping the relative and absolute contributions of the R-BT065 cluster to the bacterioplankton composition (Fig. (Fig.2b).2b). Surprisingly, for the 51 measurements obtained during 2008 for which parameters indicating the trophic status of habitats were also available (i.e., forms of phosphorus, POC concentration, and Chl-a concentration) no significant relationships (CANOCO analysis) to these parameters were found (Table (Table22).
Building on the statistical analyses, we grouped the habitats for which the ranges of values for the relevant parameters shown in Fig. Fig.22 were similar. Furthermore, geographic aspects, the size of the habitat, and the trophic status of the man-made habitats were also considered when the habitats were grouped. For some groups (e.g., acidified lakes in Šumava National Park) the grouping of the systems corresponded exactly to their geographical positions (Fig. (Fig.11 and Table Table1).1). In contrast, some types of habitats with similar characteristics (e.g., humic and DOC-rich systems) occurred in geologically and geographically very different areas. For instance, humic ponds surrounded by a small area of peat bog were found on limestone bedrock in the Alps in the Salzkammergut area and also at a lower altitude on the granite bedrock in South Bohemia. Considering these aspects and mainly the statistical relationships (Fig. (Fig.22 to to4),4), we propose dividing the habitats into the following six groups. (i) The humic lakes and ponds generally are habitats with small surface areas, low conductivities and pHs, high DOC contents, modest bacterial concentrations, and medium Chl-a concentrations (Table (Table1;1; see also Table S1 in the supplemental material). (ii) The acidified lakes include four anthropogenically acidified lakes of glacial origin located in Šumava National Park (southwest Bohemia) on granite bedrock with low buffering capacity; they have low pHs and conductivities, contain fairly low numbers of bacteria, and have higher proportions of low-molecular-weight compounds in the DOC pool (Table (Table1).1). (iii) The forest ponds include a group of six man-made, mostly remote ponds that are more than 3 centuries old and are located in the Novohradské Mountains (Natural Park, South Bohemia) in a densely forested area at an altitude of approximately 800 m; more than 90% of the watershed of the ponds is covered by forest dominated by Norwegian spruce (Picea abies), and the ponds contain brown water with an increased DOC concentration and a pH of about 6.5. (iv) The small shallow ponds are very small, shallow habitats located at a wide range of altitudes (Table (Table1;1; see Table S1 in the supplemental material) in a mountainous area; in some of these ponds macrophytes cover a substantial part of the bottom, but compared to the groups described above they are a more heterogeneous group of habitats in terms of pH and conductivity (Table (Table1).1). (v) The alkaline lakes comprise 36 natural, usually deep and large prealpine and alpine lakes located at a wide range of altitudes and are mostly in the Salzkammergut area of Austria (Fig. (Fig.1)1) on calcareous bedrock (alkaline; pH of most lakes, 8 to 8.5), except for three pH-neutral lakes on granite bedrock (Oberer Klaffersee, Unterer Klaffersee, and Rauhenbergsee); however, all 36 of these oligotrophic and oligomesotrophic habitats had modest numbers of bacteria and Chl-a concentrations, higher conductivities, and low DOC contents with higher proportions of low-molecular-compounds (Table (Table1).1). (vi) The fishponds and reservoirs represent a rather heterogeneous group of aquatic habitats located in the Czech Republic that share several features, including the fact that they are all man-made and are mostly medium size. They comprise dam reservoirs, two flooded sand pits in South Bohemia, and mostly fertilized fishponds in Central and South Bohemia; the trophic status of these pH-neutral systems ranges from mesotrophy (drinking water reservoirs and sand pits) to eutrophy to hypertrophy (heavily fertilized fishponds) (Table (Table1;1; see also Table S1 in the supplemental material), and the phosphorus concentrations, Chl-a concentrations, and bacterial numbers correspond well to the overall enhanced trophic status of these habitats compared to the other five groups (Table (Table11).
The redundancy analysis model (Fig. (Fig.22 and Table Table2)2) indicated that the pH had a strong influence on the relative proportions of the R-BT065 cluster in both the total bacteria and the total Betaproteobacteria. Using Pearson's correlation analysis for all data available, we examined the direct relationships between the parameters; pH by itself explained 35% and 52% of the variability in the proportions of R-BT065 bacteria in the total bacteria and in the total Betaproteobacteria, respectively (Fig. (Fig.3).3). Additionally, plotting the mean values of the R-BT065 proportions for each habitat group showed that there was an obvious trend of increasing R-BT065 proportions along the pH gradient (Fig. (Fig.4).4). The linear regression analysis of the habitat groups related to pH yielded r2 values of 0.664 and 0.676 (n = 6, P < 0.05) for the proportions of the R-BT065 cluster bacteria in the total bacteria and in the total Betaproteobacteria, respectively. Interestingly, small shallow ponds had the highest proportion of R-BT065 phylotypes in the total bacterioplankton (up to 29%; mean, 18%) (Fig. (Fig.33 and and4),4), while these phylotypes accounted for the largest proportion in the total Betaproteobacteria (up to 85%; mean, 55%) in the alkaline lakes located mostly on calcareous bedrock.
Overall, the model applied across all habitats (CANOCO, RDA method) (Fig. (Fig.2)2) showed significant positive effects of pH, conductivity, proportion of low-molecular-weight DOC (A250/A365), and oxygen concentration on the proportions of the R-BT065 cluster bacteria in the bacterioplankton. In contrast, the total amount of DOC, A254/DOC as a parameter for estimating the dissolved aromatic carbon (content of humic substances), and altitude had significant negative effects on the occurrence of the R-BT065 cluster (Fig. (Fig.22 and Table Table2).2). In addition, the model confirmed that there were strong and significant relationships to the core parameters for our suggested groups consisting of humic lakes and ponds, forest ponds, small shallow ponds, and alkaline lakes, while the relationships were less significant for the rather extreme acidified lake habitats and for the rather heterogeneous group consisting of fishponds and reservoirs, which contained man-made systems with great variability in the core parameter (Table (Table11 and Fig. Fig.22 to to4;4; see Table S1 in the supplemental material).
Our data showing the wide distribution of the R-BT065 bacteria in freshwater habitats clearly support suggestions made in previous studies, which were based on investigations of only a few habitats, concerning the ecological importance of the R-BT065 cluster in the pelagic zone of freshwater systems (2, 12, 24, 27, 32, 35). Furthermore, we established relationships between some major biotic and abiotic factors shaping the population dynamics of bacterioplankton. Thus, this study, which included a robust data set, provides novel insights into the distribution and factors influencing the occurrence of the R-BT065 cluster in a broad variety of freshwater habitats. The habitats studied are very diverse, but they are located in a relatively small geographical area, and thus not all our results can be generalized. More quantitative investigations in other regions and climatic zones are necessary in order to complete the image of the global distribution and importance of this bacterial group.
Obviously, R-BT065 bacteria prefer nonhumic habitats. Their relative proportions in humic or acidified lakes were generally less than 3% of the total bacteria (Fig. (Fig.4A);4A); moreover, in four humic habitats in Austria (pH 4.0 to 4.6) these phylotypes were not detected at all. This conclusion is clearly supported by the statistical analysis showing significant negative effects of DOM concentration, dissolved aromatic carbon content (A254/DOC), and low pH on the relative and absolute proportions of the R-BT065 bacteria in aquatic systems (Fig. (Fig.2).2). Moreover, laboratory experiments with two isolates from the R-BT065 cluster also demonstrated that these phylotypes do not grow at pH <6 (V. Kasalický, unpublished data).
Two seasonal studies showed that the relative proportions of the R-BT065 bacteria in the total bacteria could temporally vary by a factor of 2 to 4 in lake ecosystems (27, 33). In our study, 20 randomly chosen habitats were sampled twice at different times in a year to obtain a rough estimate of the seasonal variability. The relative proportions of the R-BT065 bacteria in the two samples differed by factors of 1.1 to 3.3 (mean ± standard deviation, 1.8 ± 0.6), a range quite similar to that found in previous seasonal studies (27, 33). The relative and absolute proportions of the target group can also change diurnally as a consequence of a daily cycle of primary production. To minimize the possible effect of diurnal variability, samples were generally collected between 10:30 a.m. and 2:30 p.m. Although our sampling strategy could not cover possible temporal or diurnal changes, our data clearly demonstrated that there were significant differences in the relative abundance and absolute abundance of R-BT065 bacteria between systems rich in humic substances and nonacidic systems (see Table S1 in the supplemental material).
One might ask which other key factors, aside from those suggested by the statistical analysis, determine the population size of the ecologically successful R-BT065 cluster of bacteria. With regard to the substrate preferences of these bacteria, obviously they can efficiently utilize autochthonously produced substrates, which seems to be one of the key factors supporting their high proportions in plankton environments. For instance, experiments by Peréz and Sommaruga (25) showed that the relative abundance of active R-BT065 cells exposed to alga-derived DOM was significantly higher than that in control or soil-derived DOM-amended samples. Typical low-molecular-weight components present in extracellular algal products, amino acids and monosaccharides (20, 37), were efficiently incorporated into the R-BT065 cells in natural systems (2, 12, 14, 27). Notably, lower rates of uptake of acetate, a typical fermentative metabolite, were detected in the R-BT065 bacteria (4). Last but not least, strong associations between the dynamics of the R-BT065 cluster and phytoplankton community structure and the extracellular production that could be related to the dynamics of cryptophytes have been reported for a dam reservoir (33).
Consequently, we can speculate that in highly productive systems (i.e., systems with more phytoplankton), we may expect larger relative proportions of the R-BT065 cluster bacteria. However, the results of the CANOCO analyses did not show any significant association between the target bacterial group and the Chl-a concentration or phosphorus parameters (Table (Table2),2), which are factors that are widely accepted as indicators of the trophic status. Second, high proportions of the cluster were found in all nonhumic, pH-neutral or alkaline systems practically regardless of their altitude or their trophic status (Fig. (Fig.44 and Table Table1).1). Indeed, similarly high relative proportions of these phylotypes were detected in eutrophic ponds and reservoirs and in oligotrophic prealpine lakes, as well as in ultraoligotrophic high-altitude alpine lakes, such as Oberer Klaffersee and Unterer Klaffersee (see Table S1 in the supplemental material).
The significant positive association of the proportion of the R-BT065 cluster with the A250/A365 ratio (Fig. (Fig.2)2) provides an additional indirect indication of the possible relationship between the bacteria and low-molecular-weight alga-derived substrates. Generally, the A250/A365 ratio indicates higher proportions of low-molecular-weight DOC (28) that is usually dominated in plankton by autochthonously derived substrates, such as algal exudates (8, 37). The fact that no direct relationship between the Chl-a concentration and the occurrence of the R-BT065 bacteria was found could also reflect the low temporal resolution of our study. However, we suggest that the Chl-a concentration by itself may not be a decisive factor, as high proportions of R-BT065 bacteria were found in very oligotrophic systems. Recently, Jones and coworkers (15) suggested a proxy, the ratio of water color to chlorophyll a concentration (CtCH ratio), which indicates the dominance of the allochthonous organic carbon available to bacteria (high CtCh values) over the autochthonous organic carbon available to bacteria (low values). We used this proxy and correlated it with the relative proportions of R-BT065 bacteria in the total Betaproteobacteria. This analysis yielded a significant correlation (r2 = 0.21, P < 0.001, n = 51) that further corroborates our finding of higher proportions of the target bacteria in nonhumic habitats with dominance of an autochthonous organic carbon source.
Finally, our preliminary experiments provided direct evidence of growth of isolates belonging to the R-BT065 group on alga-derived substrates in axenic cultures of Cryptomonas sp. and of green algae growing in a completely inorganic medium (V. Kasalický and K. Šimek, unpublished data). Thus, we concluded that autochthonously produced alga-derived substrates play an important role in shaping population dynamics of the R-BT065 cluster (15, 24, 33), although no direct relationship between the bulk phytoplankton biomass and the relative proportions of the cluster in the bacterioplankton can be determined from our data (Table (Table22).
DOC derived from algae is the main source of highly labile DOC for bacteria in the water column, particularly in stagnant systems where terrestrially derived DOC is assumed to be more recalcitrant (36). For a large set of data from various aquatic systems with different trophic statuses no significant relationship between the proportion of labile DOC and the total DOC concentration was found (6), while the total DOC concentration is generally much higher in humic lakes than in clear water lakes. Consequently, we hypothesize that in nonhumic oligotrophic systems the amount of labile DOC per bacterium could be sufficient and similar to the amount in systems with higher trophic status since the sizes of both phyto- and bacterioplankton standing stocks generally increase along the trophic gradient (3). Thus, the ubiquity of the target group in nonacidic systems, including oligotrophic, high-altitude alpine lakes and ponds (Fig. (Fig.33 and and4),4), does not contradict the hypothesis concerning the key role of the alga-derived labile substrates in the dynamics of the R-BT065 cluster (15, 26).
The hydraulic retention time (HRT) and pH of lakes have been suggested to be the key factors shaping the overall bacterial community composition in freshwaters (17, 23, 41). Unfortunately, HRT data are not available for most habitats investigated in our study. We hypothesize that the assumed short HRT and a sufficient labile DOC pool in our small and shallow nonhumic habitats can specifically favor members of the R-BT065 cluster based on their great growth potential and suggested life strategy as “opportunists” or “generalists” (2, 31, 32). The small shallow ponds, which are particularly rich in R-BT065 bacteria (Fig. (Fig.4),4), are habitats that are very likely more vulnerable to sudden environmental changes because of their small volumes, which result in very short HRTs in mountains with frequent and heavy precipitation. In such circumstances, competitive bacterioplankton groups should have fairly short population turnover times and life strategies that allow them to sustain or rapidly reestablish sufficient population sizes. Such ecophysiological traits have been reported for R-BT065 bacteria (2, 32). Notably, the target bacterial clade accounted for a significantly larger proportion of the total bacteria (average, ~18%) in the small shallow ponds than in the other groups of habitats (P ≤ 0.005, Mann-Whitney test), and Betaproteobacteria generally accounted for 45% of the total bacterial community (Fig. (Fig.4).4). In contrast, the relative contributions of the Betaproteobacteria to the total bacteria were on average significantly smaller (~25%; P < 0.0001, Mann-Whitney test) in the mostly large oligotrophic to oligomesotrophic alkaline lakes, which are more stable habitats, than in the small shallow ponds.
So far, the information concerning the ecophysiology of members of the R-BT065 cluster, based on in situ experimental manipulations, indicates that these phylotypes can rapidly respond to shifts in both top-down (grazer-induced) and bottom-up factors without a detectable lag phase in the growth response. Thus, they consistently displayed the fastest response and the highest net growth rate among the major FISH-detected bacterioplankton groups studied over a wide range of phosphorus availability (31, 32). Correspondingly, Alonso and coworkers (2) tentatively suggested that the members of this cluster are “generalists” among the bacterioplankton based on their typical growth patterns and substrate uptake capabilities. We hypothesize that the labile, alga-derived DOM in all photosynthetically active planktonic systems with different trophic statuses (3) could support the high metabolic and growth rates and in turn the competitiveness of the R-BT065 group in bacterioplankton (2, 12, 32).
The R-BT065 bacteria meet all of the prerequisites of a key bacterioplankton group in circumneutral habitats (Fig. (Fig.33 and and4)4) in terms of channeling a significant part of the carbon to higher trophic levels. This cluster shows remarkable uptake activity and high growth potential (12, 25, 32) counterbalanced by high grazing mortality (14, 35), and high proportions occur in communities. Although these characteristics suggest some important general trends, so far we have very limited knowledge about the diversity within this cluster. There are marked differences in major limnological parameters among the groups of pH-neutral habitats, yet these habitats are rich in R-BT065 bacteria (Table (Table11 and Fig. Fig.33 and and4).4). Thus, these phylotypes may form taxonomically divergent subunits with more diversified life strategies. To obtain deeper insights into ecophysiological traits of single representatives of the cluster, further experimental work with isolates belonging to this group originating from distinct habitats is urgently needed. Such work should provide answers whether the narrow R-BT065 cluster represents an ecologically coherent group or whether its apparent ubiquity in various freshwater environments reflects important ecological adaptation of several groups of ecotypes.
This study was largely supported by the Grant Agency of the Czech Republic under research grant 208/05/0015 awarded to K.Š. and by the Austrian Science Fund (projects P15655 and P19853 awarded to M.W.H.). Additional support was provided by projects AV0Z 60170517 and MSM 600 766 5801. KONTAKT grant MEB 060702 supported joint research activities of K.Š. and M.W.H.
We thank Petr Znachor for help with fieldwork and phytoplankton analysis, John Dolan for correction of the English, and three anonymous reviewers for their very constructive criticisms of the manuscript.
Published ahead of print on 30 November 2009.
†Supplemental material for this article may be found at http://aem.asm.org/.