Metabolic Diversity of Heterotrophic Bacterioplankton Over Winter and Spring in the Coastal Arctic Ocean

Environmental Microbiology (2008) 10(4), 942–949 doi:10.1111/j.1462-2920.2007.01513.x

Metabolic diversity of heterotrophic bacterioplankton over winter and spring in the coastal Arctic Ocean

Maria Montserrat Sala,1* Ramon Terrado,2 advection, and sedimentation events that contributed Connie Lovejoy,2 Fernando Unrein1,3 and organic matter that enhanced bacterial metabolism. Carlos Pedrós-Alió1 1Departament de Biologia Marina i Oceanografia, Institut Introduction de Ciències del Mar (CSIC), Passeig Marítim de la Barceloneta 37-49, 08003 Barcelona, Spain. Arctic Ocean ecosystems are particularly sensitive to 2Dépt de Biologie and Québec Océan, Université Laval, climate changes, for example small temperature differ- Québec City, Québec G1K 7P4, Canada. ences may have large effects on the thickness and extent 3Instituto de Investigaciones Biotecnológicas-Instituto of sea ice (Smetacek and Nicol, 2005). Snow and ice cover Tecnológico de Chascomús (IIB-INTECH), Camino de are biologically important, influencing light availabity and Circunvalación Laguna Km 6 (B 7130 IWA) Chascomús, the timing of phytoplankton production in the Arctic Buenos Aires, Argentina. (Carmack et al., 2004). Primary production in oceanic areas constitutes the main organic source for bacterioplankton growth (Bird and Kalff, 1984; Cole et al., 1988), Summary mainly owing to direct bacterial use of dissolved com- Metabolic diversity of heterotrophic bacterioplankton pounds released by growing phytoplankton cells (Baines was tracked from early winter through spring with and Pace, 1991). However, recent evidence suggests that Biolog Ecoplates under the seasonally ice covered bacteria do not respond immediately to phytoplankton arctic shelf in the Canadian Arctic (Franklin Bay, production in cold waters (Bird and Karl, 1999; Pomeroy Beaufort Sea). Samples were taken every 6 days from and Wiebe, 2001; Duarte et al., 2005; Morán et al., 2006). December 2003 to May 2004 at the surface, the halo- The Arctic coastal ocean is also influenced by the input of cline where a temperature inversion occurs, and at large rivers and riverine inputs may also contribute signifi- 200 m, close to the bottom. Despite the low nutrient cant amounts of terrigenous organic matter that bacteri- levels and low chlorophyll a, suggesting oligotrophy oplankton could use for substrate (Emmerton, C.A., in the winter surface waters, the number of substrates Lesack, L.F.W. and Vincent, W.F., in press). used (NSU) was greater than in spring, when chloro- The only seasonal study to date of Arctic bacterioplank- phyll a concentrations increased. Denaturing gradient ton under the ice was over the Central Arctic Ocean, gel electrophorisis analysis also indicated that the where bacterial biomass and production in the upper winter and spring bacterial communities were phylo- water column responded strongly to phytoplankton growth genetically distinct, with several new bands appear- in summer and activity was restricted to the upper 60 m ing in spring. In spring, the bacterial community (project SHEBA, Sherr and Sherr, 2003; Sherr et al., would have access to the freshly produced organic 2003). This Central Arctic Ocean study measured the bulk carbon from the early phytoplankton bloom and the production of bacterioplankton, but it did not analyse the growth of rapidly growing specialist phenotypes possible organic substrates. Moreover, it was conducted would be favoured. In contrast, in winter bacteri- from an ice camp that drifted with the pack ice, making it oplankton consumed more complex organic matter difficult to separate changes due to seasons from those originated during the previous year’s phytoplankton due to different water masses. Despite this pioneering production. At the other depths we tested the NSU effort, winter bacterial activity under the Arctic ice remains was similar to that for the winter surface, with no largely unknown, and a detailed view into the seasonal seasonal pattern. Instead, bacterioplankton metabo- changes in bacterial utilization of carbon sources under lism seemed to be influenced by resuspension, ice cover is lacking. The CCGS Amundsen was frozen into first-year pack ice in Franklin Bay (Beaufort Sea) from December 2003 to May 2004 thus providing an excellent platform for detailed sampling throughout the winter. This Received 7 June, 2007; accepted 29 October, 2007. *For correspondence. E-mail [email protected]; Tel. (+34) 93 2309500; cruise was part of the Canadian Arctic Shelf Exchange Fax (+34) 93 2309555. Study (CASES).

Garland and Mills (1991) ﬁrst used the Biolog plates to 16 POLYMERS CARBOHYDRATES CARBOXYLIC PHEN AMINO ACIDS evaluate the metabolic potential of a whole microbial com- ACIDS 14 munity, as they allow simultaneous detection of the utilization of a large number of carbon sources. Biolog plates 12 have found application for the assessment of microbial 10 metabolic diversity in freshwater (Grover and Chrza- 8 nowski, 2000; Sinsabaugh and Foreman, 2001), estua- Surface 6 rine (Schultz and Ducklow, 2000) and, to a lesser extent, Number ofpositive results marine bacterioplankton (Hollibaugh, 1994; Jellet et al., 4 1996). The Biolog-Ecoplate contains 31 carbon sources in 2 triplicate (Insam, 1997) and has provided information, not 0 available using other techniques, on the functional diver- 16 sity of bacterioplankton in marine environments such as 14 western Antarctica (Sala et al., 2005a) and the Mediterra- nean Sea (Sala et al., 2005b; 2006b) and can also be 12 used to estimate bacterial use of dissolved organic nitro- 10

gen compounds in seawater (Sala et al., 2006a). TI 8 We used the Biolog plates as a way to estimate the 6 metabolic diversity of the heterotrophic bacterioplankton Number of positive results positive Number of through winter and at the onset of the productive spring 4 season, when phytoplankton were likely to become the 2 main source of dissolved organic matter. Our main aim 0 was to determine changes in the carbon sources poten- 16 tially used by bacterioplankton during winter and spring at 14 three different layers within the ice-covered water column of the Arctic shelf. Additionally, we also followed the com- 12 position of the bacterial community in the surface with 10 denaturing gradient gel electrophoresis (DGGE) in order 8 to assess if eventual changes in metabolic diversity were Bottom 6 related to changes in phylogenetic diversity. Number of positiveresults 4

2 Results 0 Out of the 31 substrates tested, 21 were used by communities in the Arctic water samples. Ten substrates did not -Cyclodext -Hydroxybut show any positive result: the two amines (putrescine and Glycogen α Tween 80 Tween 40 D-Mannitol D-Cellobiob N-Acetyl-D- D-Xylose Glucose-1-P Pyruvic Acid D-Galacturon γ Itaconic Acid 4-Hydroxy B 2-Hydroxy B Glycyl-L-Glu L-Threonine L-Phenylala L-Asparagine L-Serine L-Arginine phenylethylamine); three carboxylic acids (a-ketobutyric, Fig. 1. Number of positive results for each of the 31 substrates in D-malic and D-glucosaminic acid), five carbohydrates the Biolog plate along the period of the study. Substrates are (b-methyl-D-glucoside, D,L-a-glycerol phosphate, a-D- grouped into categories and results are shown for the three different layers: surface (n = 26), temperature inversion depth lactose, D-galactonic acid g-lactone and i-erythritol). Gen- (n = 24) and bottom (n = 25). erally, polymers and carbohydrates were the categories with higher number of positive results (Fig. 1). The most often used substrates (more than 25 positives out of 75 Tween 80 and N-acetyl-D-glucosamine showed more samples) were Tween 80 (a trademark for polysorbate positive results at the bottom than at other layers (not 80), glycogen, D-cellobiose, D-mannitol, N-acetyl-D- significantly different). glucosamine, and glycyl-L-glutamic acid. Chlorophyll a concentrations at the surface varied Although the differences were not significant, surface between 0.02 and 0.11 mgl-1 from December to 15 March. bacteria utilzed a greater number of substrates (number Then, as the ice started thinning, chlorophyll a increased of substrates used, NSU) than bacteria at deeper depths. rapidly, reaching values from 0.20 to 0.36 from 17 April to Likewise, most of the substrates had a higher number of the end of the study, 21 May (Fig. 2). Bacterial abundance positive results at the surface, but this was only significant did not show a clear seasonal pattern and varied between for three of them (one-way ANOVA): pyruvic acid methyl 1.5 ¥ 105 and 5.2 ¥ 105 bacteria ml-1 over the period of ester (F = 5.26, P = 0.01), D-mannitol (F = 4.12, P = 0.02) study. There was a distinct seasonal pattern in chlorophyll and D-cellobiose (F = 3.17, P = 0.03). Two substrates, concentrations. Initially, from December to the end of

(A) At the deepest layer just above the bottom, chlorophyll Winter Spring a concentrations were even lower than at the TI depth 6 10 0.4 (0.01–0.05 mgL-1) and there was no clear seasonal Bacteria Chlorophyll pattern (Fig. 4). As was the case at the TI depth, bacterial 0.3 Chlorophyll abundances had no seasonal signal. The highest values -1 over the study period and among all three layers were 5 105 0.2 from this deep region, with up to 5.6 ¥ 10 in January. a ( μ Bacteria ml Bacteria There was no pattern in the NSU, but the greatest values g l

-1 were found during a short period between February and 0.1 ) March, similar to that of the TI layer. The eight substrates with greatest number of positive 104 0 results over the period of study were chosen for a detailed (B) analysis (Fig. 5). At the surface, most of the substrates 12 Polymers were used quite regularly in winter and only occasionally Carbohydrates Carboxylic Acids in spring. Five substrates showed signiﬁcantly more 10 Amino Acids positive results in winter than in spring: pyruvic acid 8 methyl ester (P = 0.001), glycogen (P = 0.005), Tween 6 40 (P = 0.02), a-cyclodextrin (P = 0.03), and N-acetyl-D-

4 glucosamine (P = 0.04). A very different pattern was found Positive substrates Positive for both the TI and the bottom layers, with a period of 2 positive results concentrated in 5 weeks from the third 0 week of February to the second of March and only a few DDDDJJJJJFFFFFMM M AAAAAMMMM positive results for the rest of the period. This period of Fig. 2. (A) Bacterial and chlorophyll a concentration, and (B) positive results was found for most of the substrates. number of substrates used separated into substrate categories, for However, different patterns were also observed for the surface samples along the period of study. pyruvic acid methyl ester, used only during December,

March, chlorophyll a concentrations were low and started to increase during the ﬁrst 2 weeks of April. Higher chlo- (A) rophyll concentrations were present from 17 April to the Winter Spring end of the period of study, 21 May. The DGGE dendogram 106 0.4 showed that different bacterial assemblages were present Bacteria Chlorophyll during these two periods. We chose this marked change 0.3 Chlorophyll in composition to separate biologically deﬁnable winter -1 and spring periods. Number of substrates used over the 105 0.2 a

entire period showed a trend opposite to that of chloro- ( μ Bacteria ml Bacteria phyll a concentrations. In winter, the NSU varied between g l -1 1 and 12 (mean 6.1), and in spring this was between 1 0.1 ) and 3 (mean 1.9). Utilization of all categories of substrates was low in spring, but especially that of carboxylic acids 104 0 which did not show any positive results during this period. (B) 12 Chlorophyll a concentrations at the temperature inver- Polymers 10 Carbohydrates sion (TI) depth were very low, with the highest value in Carboxylic Acids Amino Acids -1 December (0.10 mgl ) decreasing slowly towards Febru- 8 ary, when concentrations varied around 0.03 mgl-1 until 6 the end of May (Fig. 3). Bacterial concentrations were greatest in January–February. Low NSU occurred during 4 Positive substrates spring, when only polymers were used. However, we also 2 found low NSU (between 1 and 2) in December and 5 0 January when bacterial abundance was around 4.5 ¥ 10 DDDDJ J J J J FFFFFMMMMMAAAAAMMMM bacteria ml-1. The largest NSU (between 7 and 9) was Fig. 3. (A) Bacterial and chlorophyll a concentration, and (B) found in February and March, when there was increased number of substrates used separated into substrate categories, for polymer and carbohydrate utilization. the temperature inversion depth samples along the period of study.

(A) a concentrations increased markedly, and maximal values Winter Spring were found much later in summer (Garneau, M.E., Roy, 106 0.4 Bacteria S., Lovejoy, C., Gratton, Y., Vincent, W.F.). Based on Chlorophyll chlorophyll a and DGGE, we divided this period into winter 0.3 Chlorophyll

-1 and spring and studied the metabolic abilities of bacterioplankton at three different layers (surface, TI depth and 105 0.2

a bottom). ( μ Bacteria ml Bacteria

g l Biolog-Ecoplates have proven to be an efficient means -1

0.1 ) of assessing microbial metabolic diversity in a variety of habitats (Insam, 1997; Sala et al., 2005). Biolog-

104 0 Ecoplates provide many simultaneous enrichment cultures with individual single carbon sources. Such an approach, (B) although limited by long incubations owing to bacterial slow 12 growth at in situ temperatures and possibility of selective Polymers 10 Carbohydrates culturing (Smalla et al., 1998) provides information on the Carboxylic Acids Amino Acids potential metabolic abilities of bacterioplankton. In the 8 Arctic, in Resolute Bay, Biolog-Ecoplates have indicated 6 higher NSU in marine than in freshwater samples (Tam

4 et al., 2003). However, these authors did not provide infor-

Positive substrates mation of the carbon sources used. Most (21 out of 31) of 2 the carbon sources in the Biolog-Ecoplate were used in our 0 study. Polymers and carbohydrates were the most fre- DDDDJ J J J J FFFFFMMMMMAAAAAMMMM quently used substrate types at all depths, similar to earlier Fig. 4. (A) Bacterial and chlorophyll a concentration, and (B) results from Mediterranean harbours (Sala et al., 2006b). number of substrates used separated into substrate categories, for the bottom samples along the period of study. The carboxylic acid pyruvic acid methyl ester and the amino acid glycyl-L-glutamic acid were also used often. In contrast to the Mediterranean samples (Sala et al., 2006b) or the Bedford Basin (Jellet et al., 1996) but similar to the and N-acetyl-D-glucosamine, preferentially used towards Antarctic environment (Sala et al., 2005a), amino acids spring. were rarely used. These results support the evidence of a Taking all depths together, most of the substrates were low amino acid uptake in samples from Resolute Bay in the used more in winter than in spring. A t-test revealed sig- Canadian Arctic (Pomeroy et al., 1990). The amino acid nificant differences for pyruvic acid methyl ester (t = 3.61, used most often, and more in winter than in spring, was P = 0.001), glycogen (t = 3.11, P = 0.003), a-cyclodextrin glycyl-L-glutamic acid, a combination of glycine and (t = 2.83, P = 0.006), glycyl-glutamic acid (t = 2.84, P = glutamic acid, two amino acids that are selectively pre- 0.006), Tween 40 (t = 2.22, P = 0.029), D-cellobiose (t = served in particulate organic matter (POM) in polar waters 2.22, P = 0.030), D-xylose (t = 2.16, P = 0.03), D-mannitol (Hubberten et al., 1995). This might explain the higher (t = 2.07, P = 0.04). The only substrate that showed utilization in winter than in spring. Utilization of glucose-1-P higher utilization in spring was Tween 80. was even lower than utilization of amino acids in surface Bacterial community composition was tracked using waters, similar to the findings by Alonso-Sáez, L., DGGE. The image and cluster analysis of a DGGE gel Sánchez, O., Gasol, J.M., Balagué, V., Pedrós-Alió, C. with surface samples showed that samples grouped into (submitted) through microautoradiography. Tan (1997) two clusters (Fig. 6). The first cluster defined the winter reported on the substrates used by oligotrophic bacteria samples and the second cluster the spring. The spring isolated from Arctic and Antarctic environments using samples had new bands that did not appear in winter. Biolog GN plates. Some substrates are common to Eco Denaturing gradient gel electrophorisis gels were also and GN plates and allow a comparison with our results. carried out for TI and bottom samples. However, cluster Similar to our observations, only some of their isolates analyses did not show a clear grouping (data not shown). metabolized amino acids. Among our most commonly used substrates, Tween 40, Tween 80, pyruvic acid and mannitol were used by all these isolated polar bacteria. Discussion There were differences to our results in that none of the The waters at the sampling station were covered by sea strains used a-cyclodextrin and glycyl-L-glutamic. It might ice during the period of this study, December to May, with be that bacteria using those substrates were unable to ice break up only at the middle of May. In April, chlorophyll grow in their enrichment cultures.

Winter Spring Fig. 5. Pattern of utilization of the main substrates in the Biolog plate along the period of study. Letters denote month of the year. A D D D D J J J J J F F F F F M M M A A A A A M M M M positive utilization of the substrate is marked as a black square whereas no utilization is marked in white. Surface Glycogen α-Cyclodextrin Tween 80 D-Cellobiose D-Mannitol N-acetyl-D-glucosamine Pyruvic Acid Methyl Ester Glycyl-L-Glutamic Acid

TI Glycogen α-Cyclodextrin Tween 80 D-Cellobiose D-Mannitol N-acetyl-D-glucosamine Pyruvic Acid Methyl Ester Glycyl-L-Glutamic Acid

Bottom Glycogen α-Cyclodextrin Tween 80 D-Cellobiose D-Mannitol N-acetyl-D-glucosamine Pyruvic Acid Methyl Ester Glycyl-L-Glutamic Acid

The majority of the carbon sources were most often used Unexpectedly, bacterial metabolic diversity was markedly at the surface layer. However, N-acetyl-D-glucosamine, lower in spring than in winter. Primary production of phy- found in chitin and bacterial peptidoglycan, was more often toplankton in spring provides new DOM to the system. used at the bottom layer. Davis and Benner (2005) experi- Davis and Benner (2005) suggested that DOM produced mentally determined that chitin was biodegraded at high in situ, with high amino acid and amino sugar content, was rates in Arctic waters, suggesting that it is not likely accu- an important source of bioreactive organic matter in the mulated in POM or dissolved organic matter (DOM). As western Arctic Ocean. This increased DOM concentration bacteria are generally three orders of magnitude more might have favoured the growth of bacterial species spe- abundant in the surface layer of sediments than in the cialized in the use of carbon sources derived from water column (Sala and Gude, 2006), resuspension events phytoplankton. Possibly, the arrival of fresh algal DOM might have enriched the bottom layer with peptidoglycan, might have selected for rapidly growing specialist and our results showing the higher utilization of N-acetyl- bacteria. Indeed, the DGGE analysis shows a clear shift in D-glucosamine in the bottom layer, could reﬂect the ability the composition of bacterial species in May, with the to degrade the peptidoglycan in bacterial cell walls. appearance of new bands not found in the DGGE lanes of There were distinct patterns in metabolic diversity over the surface winter samples. Likewise, FISH counts the period of study. Bacterioplankton production was showed an increase in the percentage of cells hybridizing assumed to rely mainly on the phytoplanktonic production with the EUB probe, suggesting a transition from a com- during spring and summer. We therefore distinguished munity of low activity in winter to a more active one in two main phases in our cycle according to chlorophyll a spring (Alonso-Sáez, L., Sánchez, O., Gasol, J.M., concentration and bacterial community composition as Balagué, V., Pedrós-Alió, C., submitted). revealed by DGGE: December to 16 April and 17 April to Through summer and fall, labile organic matter originat- 21 May. ing from the previous year primary production in the Arctic High chlorophyll a concentrations were found at the is exhausted (Cota et al., 1996; Rich et al., 1997). Organic surface in spring whereas in winter these were very low. matter remaining from previous blooms is less bioreactive

© 2008 The Authors Journal compilation © 2008 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 10, 942–949 Bacterial metabolism in the Arctic 947 A and colleagues (2005a) found that bacteria did not rely on Winter Spring phytoplanktonic production in Antarctic waters in summer. Moreover, a negative correlation between NSU and chlo- 23Dec 11Jan 29Jan 21Feb 10Mar 28Mar 12Apr 15May 28May rophyll a concentrations was found for surface NW Medi- terranean waters (Sala et al., 2006b). These results suggest that bacterial assemblages in oligotrophic systems, such as the surface layer of the Beaufort Sea in winter, have a higher number of metabolic pathways and are able to exploit a wide variety of dissolved organic carbon (DOC) molecules present at low concentrations. Bacterial NSU showed a similar pattern at the TI and bottom layers, very different from that at the surface. At both layers, chlorophyll a concentrations were close to undetectable, without a clear pattern and bacterial NSU and bacterial composition (not shown here) showed no differences between spring and winter. However, it was remarkable to find a period of 5 weeks for the TI and 3 weeks for the bottom layer, between February and March, of high metabolic diversity. As chlorophyll a concentration at those depths was very low, organic matter was not freshly produced but derived from different B sources: deepening of surface water, resuspension of sediment or advection from neighbouring water masses. Several of these events occurred over the winter. Specifi- cally: (i) a deepening of the surface isotherm (-1.8°C) occurred from 1 March to 25 March (Forest et al. 2007), indicating a movement of surface water into the TI layer. (ii) Sediment resuspension events were detected with transmissivity data (Forest et al. 2007) in early winter at the bottom layer. Sediment resuspension at our station increased both the concentration of polymeric organic matter and of bacteria attached to particles (Wells and Deming, 2006). (iii) Advection probably occurred during winter. Thus, Forest et al. (2007) observed frequent epi- sodes of increased suspended matter in all the water column between January and March, probably owing to turbulence on the shelf and slope induced by Pacific- derived eddies and convective mixing during ice growth. All these three types of events would have resulted in episodic inputs of aged organic matter and with different degrees of recalcitrance and complex composition. These Fig. 6. A. DGGE gel showing PCR-amplified bacterial 16S rRNA pulses of increased substrate utilization did not occur at genes of selected samples taken at the surface. the surface, suggesting that surface waters may have a B. Cladogram of the unweighted pair group average (UPGMA) analysis of the 16S rRNA gene DGGE carried out with surface more constant supply of organic matter from the ice algae, samples. Dates group in two different clusters with a bootstrap the flushing of ice brines, or inputs of riverine water. support of 100%, corresponding to a cluster with winter samples and a cluster with spring samples. Our study is the first to show the metabolic capacities of bacterioplankton under the sea ice of the Arctic Ocean. In principle, the oligotrophic conditions prevailing in the Arctic than freshly produced matter in the western Arctic (Davis water column during winter would suggest a scenario of and Benner, 2005) and probably more complex. The high low bacterial activity compared with spring conditions. NSU found in the surface in winter might be a response of Contrary to what was expected, bacterioplankton in the the community to expand the metabolic pathways to be Arctic winter had higher NSU than in spring at the surface, able to exploit the scarce and complex carbon sources and high NSU could also be found at deeper layers spo- available at this time. Using the Biolog approach, Sala radically. Probably, the low NSU in spring were due to the

© 2008 The Authors Journal compilation © 2008 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 10, 942–949 948 M. M. Sala et al. arrival of fresh algal DOM that selects for rapidly growing filter and the same water was then filtered through a 0.2 mm phenotypes, as indicated by the new bands found in the pore size polycarbonate filter. The 0.2 mm filters were con- DGGE gel. Sala and colleagues (2006b) reported that served in criovials with 2 ml of lysis buffer at -80°C until extraction. DNA extraction and 16S rDNA gene PCR amplifi- bacterial assemblages in the oligotrophic Mediterranean cation followed the same protocol as Pinhassi and colleagues Sea are forced to adapt to changing inputs of organic (2004). Denaturing gradient gel electrophorisis was carried matter, and therefore express a high plasticity in metabolic out using a DGGE 2401-Rev B model (CBS Scientific pathways in order to exploit the changing and scarce Company). Image analysis was carried out with the Quantity carbon sources available for growth. The present study One software (Bio-Rad). The cluster analysis was con- suggests that this pattern is applicable to other marine structed applying the unweighted pair group average (UPGMA) oligotrophic environments such as the Arctic Ocean. algorithm with the software PAST (Ryan et al., 1995).

Experimental procedure Acknowledgements

The study was carried out within the framework of the CASES, Financial support for this study was provided by grants from http://www.quebec-ocean.ulaval.ca/cases/welcome.asp. The the Generalitat de Catalunya (DURSI 2003ACES00029/ CCGS Amundsen research icebreaker overwintered at a fixed ANT), the Spanish Ministerio de Educación y Ciencia position in Franklin Bay (70°03′N, 126°18′W). Samples were (REN2002-11565-E/ANT), and the Natural Sciences and collected weekly from three depths: surface (3 m), depth of TI Engineering Research Council of Canada (NSERC) through (around 20–50 m) and bottom (220 m). Water from depths project CASES under the overall direction of L. Fortier. below 10 m was collected with a CTD connected to a rosette M.M.S. had a CSIC-I3P postdoctoral contract funded by through the moon pool. In order to avoid ship contamination, the Fondo Social Europeo. Special thanks to W.F. Vincent for surface water (3 m) was collected with Niskin bottles from a providing the opportunity to join CASES and Marie-Ève hole dug in the ice around 500 m away from the ship. Analysis Garneau for part of the bacterial counts. We thank our on of surface currents, temperature and salinity indicated that this board sampling team formed by J.M. Gasol, V. Balagué, M. sampling spot was located within the same water mass as the Bayer, O. Guadayol, D. Vaqué, M. Estrada, L. Alonso-Sáez, ship, but away from its influence. J. Felipe and M. Vidal. We sincerely thank our fellow scien- Chlorophyll a concentrations were estimated according to tists on board and the officers and crew of the CCGS Amund- Yentsch and Menzel (1963). Water samples (250–500 ml) sen for their collaboration and support. We also thank Jim were filtered onto 25 mm diameter Whatman GF/F filters and Grover and an anonymous reveiwer for constructive com- frozen at ~20°C until chlorophyll a was extracted in 90% ments and suggestions. acetone and determined fluorimetrically. Bacterial concentrations were estimated using epifluorescence microcoscopy following a modification of the method of References Porter and Feig (1980). Twenty millilitres of seawater was fixed with glutaraldehyde (1% final concentration) and then Baines, S.B., and Pace, M.L. (1991) The production of dis- filtered onto 0.2 mm black polycarbonate filters and stained solved organic-matter by phytoplankton and its importance with DAPI (4,6-diamidino-2-phenylindole) final concentration to bacteria – patterns across marine and fresh-water 5 mgml-1. Bacteria on filters were counted under an systems. Limnol Oceanogr 36: 1078–1090. Olympus-BX40-102/E epifluorescence microscope. Bird, D.F., and Kalff, J. (1984) Empirical relationships Biolog-Eco plates were used to determine patterns of sole between bacterial abundance and chlorophyll concentra- carbon source utilization. Biolog-Eco plates (Biolog) are tion in fresh and marine waters. Can J Fish Aquat Sci 41: microtiter plates in which each well contains an individual 1015–1023. carbon source, 31 carbon sources in total and a blank with Bird, D.F., and Karl, D.M. (1999) Uncoupling of bacteria and only water, all in triplicates. In addition to the carbon source, phytoplankton during the austral spring bloom in Gerlache each well includes the redox tetrazolium violet to indicate the Strait, Antarctic Peninsula. Aquat Microb Ecol 19: 13–27. oxidation of the substrates. Positive wells develop purple Carmack, E.C., Macdonald, R.W., and Jasper, S. (2004) colour. After inoculation of a 150 ml sample in each well, the Phytoplankton productivity on the Canadian Shelf of the microplates were incubated in the dark at 4°C during a period Beaufort Sea. Mar Ecol Prog Ser 277: 37–50. between 2 and 4 weeks. Several tests indicated that longer Cole, J.J., Findlay, S., and Pace, M.L. (1988) Bacterial pro- incubation time did not increase the number of positive duction in fresh and saltwater ecosystems – a cross- results. Positive wells were recorded at the end of the incu- system overview. Mar Ecol Prog Ser 43: 1–10. bation period. A substrate was considered to be positive Cota, G.F., Pomeroy, L.R., Harrison, W.G., Jones, E.P., when at least one of the three wells showed colour formation. Peters, F., Sheldon, W.M., and Weingartner, T.R. (1996) We defined the NSU as the total number of substrates Nutrients, primary production and microbial heterotrophy in utilized out of the 31 for each sample. Differences between the southeastern Chukchi Sea: Arctic summer nutrient layers or periods were tested using a one-way ANOVA and depletion and heterotrophy. Mar Ecol Prog Ser 135: 247– t-test with the program Statistica. 258. The analysis of bacterioplankton phylogenetic diversity Davis, J., and Benner, R. (2005) Seasonal trends in the was carried out by DGGE. During the DNA sample collection, abundance, composition and bioavailability of particulate water was prefiltered with a 3 mm pore size polycarbonate and dissolved organic matter in the Chukchi/Beaufort Seas

and western Canada Basin. Deep Sea Res II 52: 3396– trations of dissolved organic matter in the Central Arctic 3410. Ocean. Deep Sea Res Part II 44: 1645–1663. Duarte, C.M., Agustí, S., Vaqué, D., Agawin, N.S.R., Felipe, Ryan, P.D., Harper, D.A.T., and Whalley, J.S. (1995) J., Casamayor, E.O., and Gasol, J.M. (2005) Experimental PALSTAT, Statistics for Palaeontologists. London, UK: test of bacteria-phytoplankton coupling in the Southern Chapman & Hall. Ocean. Limnol Oceanogr 50: 1844–1854. Sala, M.M., and Gude, H. (2006) Seasonal dynamics of Emmerton, C.A., Lesack, L.F.W., and Vincent, W.F. Nutrient pelagic and benthic (littoral and profundal) bacterial and organic matter patterns across the Mackenzie River, abundances and activities in a deep prealpine lake estuaryard shelf during the seasonal recession of sea- (L. Constance). Arch Hydrobiol 167: 351–369. ice. J Mar Systems (in press). Sala, M.M., Arin, L., Balagué, V., Felipe, J., Guadayol, Ò., Forest, A., Sampei, M., Hattori, H., Makabe, R., Sasaki, H., and Vaqué, D. (2005a) Functional diversity of bacteri- Fukuchi, M., et al. (2007) Particulate organic carbon fluxes oplankton assemblages in western Antarctic seawaters on the slope of the Mackenzie Shelf (Beaufort Sea): physi- during late spring. Mar Ecol Progr Ser 292: 13–21. cal and biological forcing of shelf-basin exchanges. J Mar Sala, M.M., Balagué, V., Pedrós-Alió, C., Massana, R., Syst 63: 39–54. Felipe, J., Arin, L., et al. (2005b) Phylogenetic and func- Garland, J.L., and Mills, A.L. (1991) Classification and char- tional diversity of bacterioplankton during Alexandrium spp acterization of heterotrophic microbial communities on the blooms. FEMS Microbiol Ecol 54: 257–267. basis of patterns of community-level sole-carbon-source Sala, M.M., Pinhassi, J., and Gasol, J.M. (2006a) Estimation utilization. Appl Environ Microbiol 57: 2351–2359. of bacterial use of dissolved organic nitrogen (DON) com- Grover, J.P., and Chrzanowski, T.H. (2000) Seasonal pat- pounds in aquatic ecosystems using Biolog plates. Aquat terns of substrate utilization by bacterioplankton: case Microb Ecol 42: 1–5. studies in four temperate lakes of different latitudes. Aquat Sala, M.M., Estrada, M., and Gasol, J.M. (2006b) Seasonal Microb Ecol 23: 41–54. changes in the functional diversity of bacterioplankton Hollibaugh, J.T. (1994) Relationship between thymidine in contrasting coastal environments of the NW metabolism, bacterioplankton community metabolic capa- Mediterranean. Aquat Microb Ecol 44: 1–9. bilities, and sources of organic matter. Microb Ecol 28: Schultz, G.E., and Ducklow, H.W. (2000) Changes in bacte- 117–131. rioplankton metabolic capabilities along a salinity gradient Hubberten, U., Lara, R.J., and Kattner, G. (1995) Refractory in the York River estuary, Virginia, USA. Aquat Microb Ecol organic compounds in polar waters: relationships between 22: 163–174. humic substances and amino acids in the Arctic and Sherr, B.F., and Sherr, E.B. (2003) Community respiration/ Antarctic. J Mar Res 53: 137–149. production and bacterial activity in the upper water column Insam, H. (1997) A new set of substrates proposed for of the central Arctic Ocean. Deep-Sea Res I 50: 529–542. communiy characterization of environmental samples. Sherr, E.B., Sherr, B.F., Wheeler, P.A., and Thompson, K. In Microbial Communities. Functional Versus Structural (2003) Temporal and spatial variation in stocks of Approaches. Insam, H., and Rangger, A. (eds). Berlin autotrophic and heterotrophic microbes in the upper water Heidelberg, Germany: Springer Verlag, p. 261. column of the central Arctic Ocean. Deep Sea Res I 50: Jellet, J.F., Li, W.K.W., Dickie, P.M., Boraie, A., and Kepkay, 557–571. P.E. (1996) Metabolic activity of bacterioplankton commu- Sinsabaugh, R.L., and Foreman, C.M. (2001) Activity profiles nities assessed by flow cytometry and single carbon sub- of bacterioplankton in a eutrophic river. Freshw Biol 46: strate utilization. Mar Ecol Prog Ser 136: 213–225. 1239–1249. Morán, X.A.G., Sebastián, M., Pedrós-Alió, C., and Estrada, Smalla, K., Wachtendorf, U., Heuer, H., Liu, W.-T., and M. (2006) Response of Southern Ocean phytoplankton and Forney, L. (1998) Analysis of BIOLOG GN substrate utili- bacterioplankton production to short-term experimental zation patterns by microbial communities. Appl Environ warming. Limnol Oceanogr 51: 1791–1800. Microbiol 64: 1220–1225. Pinhassi, J., Sala, M.M., Havskum, H., Peters, F., Guadayol, Smetacek, V., and Nicol, S. (2005) Polar ocean ecosystems O., Malits, A., and Marrasé, C. (2004) Changes in Bacte- in a changing world. Nature 437: 362–368. rioplankton composition under different phytoplankton Tam, L., Kevan, P.G., and Trevors, J.T. (2003) Viable bacte- regimens. Appl Environ Microbiol 70: 6753–6766. rial biomass and functional diversity in fresh and marine Pomeroy, L.R., and Wiebe, W.L. (2001) Temperature and waters in the Canadian Arctic. Polar Biol 26: 287–294. substrates as interactive limiting factors for marine het- Tan, T.L. (1997) Biolog metabolic fingerprints for clustering erotrophic bacteria. Aquat Microb Ecol 23: 187–204. marine oligotrophic bacteria from polar regions. In Pomeroy, L.R., Macko, S.A., Ostrom, P.H., and Dunphy, J. Microbial Communities. Functional Versus Structural (1990) The microbial food web in arctic seawater – con- Approaches. Insam, H., and Rangger, A. (eds) Berlin, centration of dissolved free amino-acids and bacterial Heidelberg, Germany: Springer-Verlag, 161–170. abundance and activity in the arctic-ocean and in resolute Wells, L.E., and Deming, J.W. (2006) Significance of bacte- passage. Mar Ecol Prog Ser 61: 31–40. rivory and viral lysis in bottom waters of Franklin Bay, Porter, K., and Feig, Y.S. (1980) The use of DAPI for identi- Canadian Arctic, during winter. Aquat Microb Ecol 43: 209– fying and counting aquatic microflora. Limnol Oceanogr 25: 221. 943–948. Yentsch, C.S., and Menzel, D.W. (1963) A method for the Rich, J., Gosselin, M., Sherr, E., Sherr, B., and Kirchman, determination of phytoplankton chlorophyll and phaeophy- D.L. (1997) High bacterial production, uptake and concen- tin by fluorescence. Deep Sea Res 10: 221–231.