Abundance and production of pelagic bacteria in the southern Bay of Biscay during summer *

The first studies on bacterial abundance and production in representative areas of the shelf in the southern Bay of Biscay were carried out in coastal, shelf-break and off-shelf zones of the Central Cantabrian Sea in summer 1993 and 1994. In 1993 some coastal stations displayed the effects of an upwelling pulse. Thymidine uptake rates were minimal during upwelling conditions in coastal waters (range 0.4 4 pmol l-1 h-1). In shelf-break and off-shelf waters, thymidine uptake profiles displayed surface and subsurface peaks (range 0.3 15.5 pmol l-1 h-1). Vertical distribution of heterobacteria abundance varied from 0.28 to 2.27 x 106 cell ml-1 in the whole area studied. Cyanobacteria varied from 0.04 x 104 to 1.04 x 105 cell ml-1. Heterobacteria accounted for 13 to 46% of total particulate carbon. In this study we concluded that during summer, pelagic heterobacteria can enhance high carbon turnover rates when upwelling relaxes and primary production rates increase, allowing the accumulation of phytoplankton. In turn, bacterial production rates decay abruptly during the upwelling process. Off-shore the system is typically oligotrophic, bacterioplankton balance the primary production rates and even bacteria seem to be food-limited. Our results on abundance and production of pelagic bacteria in the Central Cantabrian Sea are within the published range for the Galician upwelling areas.


INTRODUCTION
In upwelling systems, bacterioplankton dynamics are closely related to the upwelling cycle and the dynamics of phytoplankton (Sorokin and Kogelschatz, 1979;Sorokin, 1981;Hanson et al., 1986;Bak and Nieuwland, 1993;Tenore et al., 1995).Bacterial biomass and activity are low during early phases of upwelling and become increasingly important in the cycling of carbon and nutrients as the upwelling intensity decays.
The northern and north-western Iberian coasts are characterised by upwelling, primarily wind-driven, occurring in the period from March to October (Dickson et al., 1980;Fraga, 1981;Blanton et al., 1982;Botas et al., 1990).This upwelling system shows variations of phytoplankton abundance and primary production which are similar to the succession of phases of vertical mixing during winter, early spring and autumn and thermal stratification of the water column during summer (Fernández andBode, 1991, 1994;Bode et al., 1994 a;Casas et al., 1997).Primary production maxima are often related to favourable upwelling conditions, especially during spring and summer (Fernández and Bode, 1991;Varela et al., 1991;Bode et al., 1994 a;Tenore et al., 1995).In contrast with the Galician region (NW Spain), where there are upwelling events during the spring and near the shelf-break (Varela et al., 1991;Tenore et al., 1995), the upwelling in the southern Bay of Biscay is known to occur near the coast in the summer period, where it causes the disruption of the thermally stratified surface layer (Botas et al., 1990;Fernández et al., 1991;Marañon and Fernández, 1995) and produces significant changes in planktonic community structure, biomass and primary production (Fernández and Bode, 1991;Fernández et al., 1991;Bode and Fernández, 1992 a;Fernández and Bode, 1994;Marañón and Fernández, 1995;Marañón et al. 1995).These events are particularly important in the pelagic ecosystem of the southern Bay of Biscay because the input of nutrients into the euphotic zone during summer allows for an increase in biological production when an oligotrophic situation is expected.Bode and Fernández (1992 b), analysing the biochemical composition and the size spectra of seston particles, concluded that this pelagic ecosystem would be based on a microbial food web for most of the year, but particularly during summer and in the off-shelf region where small flagellates and heterotrophic dinoflagellates dominate the microplankton community (Fernández and Bode, 1994).
There is limited information on bacterial abundance and activity during upwelling events on the Galician coast (Hanson et al., 1986;Bode et al. 1994 b;Tenore et al., 1995), despite their potential importance, bacteria were scarcely studied in the southern Bay of Biscay.The only available studies were carried out in a shallow environment in the eastern coast, (Iriberri et al., 1987(Iriberri et al., , 1990;;Barcina et al., 1992;Unanue et al., 1992).The main pattern which emerged from these studies was the marked seasonal change between the low bacterial activity and biomass values during the cold period (winterspring) and the maximum values measured during the warm period (summer-autumn).However, the place where these studies were made was not affected by the upwelling, which is restricted to the western coast of the Cantabrian Sea (Dickson et al., 1980).
The present study describes cell abundance and production of bacterioplankton in three distinct zones (coastal, shelf-break and off-shelf zone) of the Central Cantabrian Sea (southern Bay of Biscay) during summer in relation to upwelling conditions.These are the first data on bacterioplankton for this coastal region.

MATERIAL AND METHODS
Sampling was made during two cruises, ASTURIAS-0793 (July 1993) and ASTURIAS-0794 (July 1994), in the Central Cantabrian Sea (southern Bay of Biscay).Stations were arranged in a cross-shelf transect covering a bathymetric range from 30 to 900 m and some of them were sampled several times during the study (Fig. 1).We considered four types of stations according to the vertical distribution of temperature and distance from the coast: coastal stations displaying characteristics of an upwelling pulse (labelled U in the Fig. 1), coastal stations with weak-upwelling conditions (C), shelf-break stations (S), and offshelf stations (O).
Vertical profiles of temperature and salinity were obtained with a CTD SBE-25.Irradiance (400 to 700 nm) was measured with a Li-COR radiometer equipped with a 4Π sensor.Water samples for the determination of chlorophyll-a, particulate organic carbon and bacterial abundance and production were collected with acid cleaned Niskin bottles from the surface to the bottom of the euphotic zone.The water was transferred to acid cleaned carboys and transported to the laboratory within four hours of sampling in the field.Chlorophyll-a was determined in particles retained by Poretics 0.2 µm polycarbonate filters (ASTURIAS-0793 cruise) or Whatman GF/F filters (ASTURIAS-0794).Chlorophyll-a concentration was analysed by the fluorometric procedure of Yentsch and Menzel (1963) and Parsons et al. (1984).Particulate carbon was analysed in material retained by Whatman GF/F filters using either a CHN Carlo-Erba 1500N (ASTURIAS-0793) or a Perkin-Elmer 2400 analyser (ASTURIAS-0794).
Bacterial production was estimated by measuring the incorporation of [ 3 H] thymidine into the trichloroacetic acid (TCA) insoluble fraction, following the procedure of Fuhrman and Azam (1982).Three 10 ml replicates were drawn from each sample into centrifuge glass tubes, labelled with thymidine (5 nM final concentration) and incubated in the laboratory for 60 minutes.The thymidine concentration and incubation times were similar to those described by Iriberri et al. (1990), and adequate to saturate linear thymidine uptake.Temperature and irradiance conditions during the incubation were similar to those observed at the depth where samples were collected using laboratory incubators equipped with fluorescent lights.TCA-soluble pools of bacterial material were extracted on ice with TCA (5% final concentration) for 20 minutes.Particulate material was then filtered through 0.2 µm Poretics polycarbonate filters, rinsed three times with 4 ml of 5% TCA and finally washed with filtered seawater.Ethanol was not used during extraction.Blank values were obtained from thymidine labelled samples, killed with TCA and filtered immediately.
Phytoplankton carbon biomass was calculated from chlorophyll-a equivalents (chlorophyll-a + 1.51 phaeophytin) and using a C:Chl-a ratio of 50.The factor 1.51 is the ratio of molar weights of chlorophyll-a to phaeophytin.Bacterial carbon was calculated assuming 11.35 fg C cell -1 , computed for free-living bacteria from Table 1 in Iriberri et al. (1990).The conversion from thymidine uptake to bacterial carbon production was made applying the factor 1.276 x 10 11 µg C (mol thymidine) -1 in Iriberri et al. (1990).It was presumed no variation of bacterial production rates during the day (24 h).

RESULTS AND DISCUSSION
A clear upwelling pulse was detected at two coastal stations in 1993, as surface temperature increased from approximately 15 ºC near the coast (sta.U) to ca. 19 ºC at the O stations (Fig. 2).In general there were no marked thermoclines although the temperature profiles for the O stations displayed the highest gradients between 20 and 40 m.The conditions found in 1994 were similar to those of 1993, except that in coastal waters upwelling conditions were weaker, and temperature in the upper 10 m exceeded 16 ºC in all stations.Mixing phenomena in the upper 50 m of the water column are relatively frequent in this area (Botas et al., 1990) and they prevent the development of a well defined thermal stratification of the water column.Therefore, we realize that thermal distribution could be insuficient to justify our zonation into the four mentioned station types, if spatial location and chlorophyll profiles were not taken into account.
Chlorophyll-a concentration decreased, and their maxima were located deeper, from the coast towards the off-shelf, excepting the U stations in 1993 which showed the lowest chlorophyll concentrations through the water column (Fig. 3).Low phytoplankton biomass is expected in early upwelling stages (Botas et al., 1990;Varela et al., 1991).The upwelling situation observed in 1993 was described by Marañón and Fernández (1995): Cryptophyceae and other flagellates dominated through the studied transect, but diatoms increased their numbers near the coast.Upwelling dynamics caused the transport and accumulation of phytoplankton towards deep layers near the shelf-break (Fig. 3) driven by a mechanism similar to that described in the Galician upwelling (Castro et al., 1994).Bacterial thymidine uptake rates were low near the coast in 1993 because of the unfavourable upwelling conditions and ranged 0.4 -4 pmol l -1 h -1 , whereas they increased towards deeper waters with a marked subsurface maximum at the O stations (Fig. 3).In contrast, in 1994 the highest production values were at C station with 15.8 pmol l -1 h -1 , and surface maxima were found in all cases.During summer 1993 and 1994, in shelf-break and off-shelf waters thymidine uptake profiles varied between 0.3 and 15.5 pmol l -1 h -1 .After upwelling events, when water column reach a relative stability, lower turbulence and reduced advective transport of the surface layer near the coast, would favour phytoplankton accumulation and therefore bacterial growth would be enhanced, as reported in other upwelling systems (Sorokin and Kogelschatz, 1979;Sorokin, 1981;Bak and Nieuwland, 1993).Although no clear relationship was found in our study between the depth of chlorophyll-a maxima and the peaks of bacterial production, integrated stocks of phytoplankton and heterotrophic bacterial carbon, and bacterial production rates exhibited a close relation ship, as will be discussed below.
was above the chlorophyll maximum at S and O stations.Cyanobacteria increased from C to O station, and their abundances were similar to those found in some areas of the N Atlantic Ocean (Li et al., 1993) but lower than those reported for estuarine areas (Kuosa, 1990;Caron et al., 1991).Subsurface maxima of cyanobacteria occurred slightly deeper offshore, above the maxima of chlorophyll and heterotrophic bacteria, as it has been found in other areas (Olson et al., 1990).
For comparison purposes, integrated values of phytoplankton and bacterial carbon stocks, bacterial production and growth rates were computed (Table 1).Heterotrophic bacterial carbon biomass was calculated using a factor of 11.35 fg C cell -1 computed from Table 1 in Iriberri et al. (1990).Bacterial production was calculated using the same factor 1.276 x 10 11 µg C (mol thymidine) -1 as in Iriberri et al. (1990).These authors provided data for free-living marine bacteria in the south-eastern Bay of Biscay, and we decided to use their conversion factors as the best approach to our own study.We are aware that these carbon production/thymidine uptake and carbon biomass/bacterial cell factors may vary depending upon, e.g., bacteria biovolume (Bratbak, 1985) and environmental conditions (Simon and Azam, 1989;Li et al., 1993).It will be necessary to obtain empirical conversion factor in further studies in this Cantabrian region.
The largest variations occurred in the coastal zone depending on the upwelling phase: total particulate carbon and phytoplankton biomass were very low during the upwelling event in summer 1993, but peaked during weak-upwelling condition (Table 1).In the C case, we estimated that phytoplankton was 73 and 78 % of total particulate carbon in summer 1993 and 1994 respectively, while heterobacteria only accounted for 13 % in summer 1994.In contrast, progressing from the coast, heterobacterial car- bon increased up to 46 % of total particulate carbon in O station.During 1994, along with a sharp increase in bacteria numbers from coast to off-shelf, production rates seemed lower in O station suggesting a great proportion of 'dormant' bacteria (Table 1).Therefore, lower average bacterial growth rates resulted in O station compared to S and C ones.In this sense Fuhrman et al. (1980) argued that bacterioplankton may be a significant fraction of seston, even in the euphotic zone, and that bacterial growth rates are density independent.In this study, vertical profiles of carbon turnover rate of heterobacteria (not shown) were found quite similar to those of thymidine uptake rate (Fig. 3), in all cases showing two peaks: at surface and between 20 -25 m depth, above the subsurface maxima of phytoplankton.In addition, constant and pronounced maxima of thymidine uptake rates appeared at the surface, where chlorophyll-a concentration was very low.This contrasts with the correlation commonly seen between the location of maxima of phytoplankton biomass and heterobacterial rates (Fuhrman et al., 1980;Fuhrman and Azam, 1982), and suggests the presence of other food sources for heterobacteria (perhaps dissolved organic matter).
Even when no primary production measurements are available for either 1993 and 1994 cruises, we compared bacterial production rates with the published primary production values for the Cantabrian Sea, as a first approach in order to evaluate the importance of bacteria for this pelagic ecosystem.Using values given by Botas et al. (1989) and Anadón et al. (1991), and extrapolating hourly rates to daily values assuming a 12:12 h photoperiod, the average values of primary production were: 1533, 921 and 744 mg C m -2 d -1 for the coastal, shelf-break and off-shelf zones respectively.Despite the acknowledged variability of primary production rates from year to year even in the same region, these averages are representative of the pelagic system in the Central Cantabrian Sea (Bode et al., 1996).According to this, our estimations of bacterial production would represent from 54 to 124% of primary production in O stations, 84 -96% in S stations and 12 -39% in C stations.These results are indicative of an oligotrophic system off-shelf, as reported in other areas where microbial heterotrophic processes are in balance with primary production (Linley et al., 1983;Cho and Azam, 1990).Altogether, much lower turnover rates for bacteria (0.37 d -1 ) than onshelf, along with relatively low phytoplanktonic and total particulate carbon biomass were supporting this suggestion of an oligotrophic system in off-shelf waters, where heterotrophic bacteria would be well limited by food supply.
In this work we provide the first data on abundance and production of pelagic bacteria in representative areas of the continental shelf of the Bay of Biscay.Our results are within the range of values available for nearby areas in the northern and north-western Iberian Peninsula, especially with those from the Galician upwelling (Table 2).The large range of values reported for the País Vasco Coast (south-eastern Bay of Biscay) may be due to the fact that those are samples from a shallow and relatively eutrophic coastal environment which is not affected by the upwelling.In our study, average bacterial growth rates exhibited a quite narrow range, close to the lower limit of the values available for the S Bay of Biscay (Table 2), implying doubling times of the bacterial carbon from 0.8 to 2.7 days.However, these are preliminary estimates and more measurements are required to determine bacterial importance in various upwelling stages of this ecosystem.
FIG. 1. -Location of sampling stations.Dots (•) indicate stations sampled only during summer 1993.Asterisks (*) indicate stations sampled during summer 1993 and repeated in summer 1994.Labels refer to the type of hydrographic station: O, off-shelf station; S, shelf-break station; C, coastal station with weak-upwelling conditions; U, station with upwelling conditions.The station labelled as U' was found with upwelling conditions only during summer 1993, and was in weak-upwelling conditions in summer 1994.
FIG. 2. -Vertical distribution of temperature (ºC) in the upper 60 m during summer 1993 (A) and summer 1994 (B), from every station of type O, S, C and U as indicated in Fig. 1.

TABLE 1 .
-Means (± sd) of carbon stock, carbon production and growth rates of bacteria in the main zones identified.All values integrated for 0-60 m depth interval, when possible.
sta. type O sta. type S sta.type C sta. type U sta. type O sta. type S sta.type C