Hydrobiologia 489: 107–115, 2002. © 2002 Kluwer Academic Publishers. Printed in the Netherlands.

107

Seawater DMS in a perturbed coastal ecosystem Serge Despiau1 , Justine Gourdeau1 , Dominique Jamet2 , Claude Geneys2 & Jean-Louis Jamet2 1 Universit´ e

de Toulon et du Var, Laboratoire des Echanges Particulaires aux Interfaces UPRES E.A. 1723, La Garde, France 2 D´ epartement de G´enie Biologique, B.P. 132, F-83957 La Garde, France E-mail: [email protected] Received 12 February 2002; in revised form 2 October 2002; accepted 11 November 2002

Key words: DMS, coastal ecosystem, bay

Abstract DMS concentrations, chlorophyll a concentrations, abiotic parameters of water quality and quantitative samples of plankton were carried out once a month from January to December 1997 into two zones of a semi-enclosed french littoral ecosystem (Toulon Bay, NW Mediterranean sea). This bay is divided into two subecosystems by an artificial breakwater: the inner bay (polluted zone, P) is largely influenced by anthropogenic perturbations and the outer bay (less polluted, LP) is much less polluted. We found greater concentrations of DMS and chlorophyll a, of phytoplankton and zooplankton densities and biomasses in the polluted zone (P) than in the less polluted zone (LP) of the bay. The DMS concentration and phytoplankton biomass were strongly correlated, and a high degree of eutrophication may contribute, in connection with other factors, to a greater production of phytoplankton which in turn enhances the DMS production. The DMS concentration in coastal polluted zones is then greatly higher than in open sea (around three times) and this greater production should be taken into account for the global estimation, at least on a local scale, of DMS production in seawater, which is a key factor for the biogenic sulfur cycle.

Introduction Dimethyl sulfide (DMS) is an important sulfur gas produced by marine phytoplankton (Iverson et al., 1989; Belviso et al., 1990). DMS is derived from its precursor beta-dimethylsulfoniopropionate (DMSP), which is believed to act as an osmoregulator in marine algae, and is considered to be a major natural source of atmospheric sulfur as it crosses the sea/air interface (Andreae, 1990). In the atmosphere, DMS is oxidized to form aerosol particles and CCN, affecting the number and radiative properties of marine clouds (Charlson et al., 1987). There may be feedback between DMS emissions and climate as a result of the influence of temperature and insulation on the growth of phytoplankton, but this remains unclear (Bates & Quinn, 1997). Several investigations in industrialized areas, in temperate coastal areas or in estuary (Turner et al.,

1988; Leck & Rhode, 1991; Townsend & Keller, 1996; Simo et al., 1997; Cerqueira & Pio, 1999) have shown that biogenic sulfur is a significant seasonal source of atmospheric sulfur on a regional scale. But, as previously noted by Simo et al. (1997), little work has been done on the production of DMS in polluted areas and there are few reports of DMS concentrations in such systems. The Mediterranean Sea is thus an interesting experimental site, because it has a heavily developed coastline, major urban areas, a high population density, all of which are great potential causes of water pollution. This research was carried out to study, in situ, the changes and differences in DMS concentration along with the behavior of the abiotic and biological factors in two zones of the Toulon Bay (French Mediterranean coast): one highly polluted zone in the inner bay, and another, less polluted, in the outer bay.

108 The sites studied Toulon Bay (43◦ 5 N and 6◦ 0 E) is located on the french Mediterranean coast and is largely influenced by the city of Toulon and its urban area (ca. 300 000 inhabitants). The bay is separated into two parts by an artificial breakwater that limits exchange between the inner and outer water to a narrow and shallowness passage (Fig. 1). This bay has a great deal of maritime traffic due to military and commercial activities and the water has high concentrations of PAH, PCB and heavy metals, particularly in the inner part of the bay (French Phytoplankton Monitoring Network data, IFREMER, 1993). The water in the inner bay is also heavily contaminated with organic, chemical (anti-fouling paintings) and suffers from biological pollution (toxic phytoplankton) (Belin et al., 1995; Jamet et al., 2001). Previous studies (Jamet & FerecCorbel, 1996; Jamet and Bogé, 1998; Jamet et al., 2001; Richard & Jamet, 2001) showed that in the inner bay, Posidonia oceanica was absent while the chlorophyll a, phosphatase activities of organisms and the zooplankton densities were much higher than in the outer bay. On the contrary, Posidonia oceanica is abundant in the outer bay, indicating that the water quality is higher. Consequently, the two sampling zones used in this study consisted of one in the polluted area (P zone) on the west side of the breakwater and the second in the less polluted area (LP zone) on the east side, the whole constituting a nice natural laboratory for these investigations.

Material and methods This study was conducted from January to December 1997. DMS, plankton concentrations and the physicochemical parameters of water quality were measured at least once a month, on the same day, in the two zones P and LP. DMS concentrations were measured by gas chromatography with an FPD detector by the method of Nguyen et al. (1990). 0.5 l samples of surface seawater were collected in polyethylene bottles so as to leave no head-space, and were analyzed less than 2 h after collection. A sample (5–10 ml) of water was injected into a bubbler using a syringe. The bubbler was purged before each analysis with pure helium to prevent contamination with laboratory atmospheric DMS. Extraction was carried out for 10–20 min, at a flow rate of 100 ml.min−1 . The extracted gases passed through

a Teflon tube containing MgClO4 to remove water vapor and volatiles were trapped at −90 ◦ C on 60–80 mesh Tenax GC in a Teflon U-tube. The tube was then rapidly heated at +90 ◦ C and trapped gases were transferred via the helium carrier gas onto a Chromosil 310 chromatography column. The column was calibrated using standard solutions of DMS. The experimental error was 10% with a detection limit of 0.5 ng. Measurements were performed in replicates and the agreement was better than 10%. Plankton nets (10 µm and 90 µm mesh, model General Oceanic 5125, diameter 0.5 m, length 2.5 m) were used to collect phytoplankton and zooplankton, respectively. The volume of water filtered was measured by a flowmeter (model General Oceanic 2030 R) and each sample was made up of several vertical hauls at each station to avoid zooplankton aggregation. Sea water was immediately fixed with Lugol to preserve phytoplankton cells (Bourrely, 1996) and zooplankton samples were immediately stored in buffered (CaCO3) 5% formalin in sea water. Algal cells were counted with an inverted microscope (Legendre & Watt, 1972) and the phytoplankton biomass was estimated by the Lohman method (1908). All zooplankton organisms were identified and counted. The chlorophyll a concentration (µg.l−1) (Greenberg et al., 1992), orthophosphates (P–PO4 3− ) and total phosphorus (P–TP) concentrations (µM) (Eisenrich et al., 1975) were measured in each sea water surface sample. The temperature, dissolved oxygen, and salinity of seawater were measured at one-meter intervals in the water column of both sampling zones. Wind speed and air temperature data were obtained from the French meteorology network. Later on, in 1998 and 1999, bacteria and nitrate (N–NO3− ) measurements were realized in the same zones and considered in this paper assuming that they represented a characteristic of that area.

Results The DMS concentrations were 116–673 ng.l−1 over the whole study for both zones, with an annual mean value of 314 ng.l−1 (Fig. 2). As mentioned by Kettle et al. (1999), it is rather difficult to compare results obtained in different regions, seasons, and sometimes by different methods. Nevertheless, we reported in Table 1 different values obtained in the open sea or in coastal zones and we can see that our mean value is considerably higher, around three times, than those

109

Figure 1. Location of the sampling zones P and LP, separated by the breakwater (B) in the Toulon bay.

obtained in the open sea (less than 120 ng.l−1 ), higher than the values estimated by Kettle et al. (1999) for the latitude of Toulon (between 60 and 150 ng.l−1 in winter and 120 and 300 ng.l−1 in summer) and higher than the average value calculated from measurements made in other coastal zones: Ccz =280 ng.l−1 [or =185 ng.l−1 , eliminating the higher (970 ng.l−1 ) and lower (7 ng.l−1 ) values reported in Table 1]. The highest values in zone P were measured at the beginning of July (579 ng.l−1 ) and January (493 ng.l−1 ), while the average value was 365 ng.l−1 . In zone LP, the highest concentration was recorded in March (673 ng.l−1) and two other peaks (390 ng.l−1 ) were measured in January and October, as in zone P. The corres-

ponding mean value was 263 ng.l−1 , higher than the previous Ccz values. Technical problems caused the loss of the DMS concentrations for February, March and April in zone P. Nevertheless, the concentrations in zone P were significantly different from zone LP (Wilcoxon test, P = 0.0051). The average concentration in P zone, from the 10 values available, was 1.4 times higher than in zone LP and this ratio increased to 2.2 during the summer months (Table 2). The phytoplankton biomass was 1.6–130 µg.l−1 and the density 165–3500 cells.l−1 over the whole study for both zones (Fig. 2). The phytoplankton community was represented essentially by Bacillariophyceae and Dinophyceae, which accounted for ap-

110

Figure 2. From top to bottom, monthly values obtained, in zones P and LP, of: DMS concentrations; phytoplankton biomass; phytoplankton densities; zooplankton densities; and Chlorophyll a concentrations.

111 Table 1. Example of seawater DMS concentrations(ng.l−1 ) measured, in different areas and seasons, in the open sea (a), in coastal zones (b) and derived from the Kettle et al. (1999) database (c) for the latitude of Toulon. See Kettle et al. (1999), for more results (a) Open sea Bates & Cline (1985) Bates & Cline (1985) Berrescheim et al. (1989) Nguyen et al. (1990) Nguyen et al. (1990) Simo et al. (1997) Gourdeau (1999) (b): Coastal zones Turner & Liss (1985) Turner & Liss (1985) Turner et al. (1988) Turner et al. (1988) Leck & Rhode (1991) Leck & Rhode (1991) Simo et al. (1997) Cerqueira & Pio (1999) Cerqueira & Pio (1999)

Pacific Pacific Antarctic Indian Indian Mediterranean Mediterranean

118 37 118 25 87 111 95

Summer Winter Fall Winter Summer Spring March–April

North Sea Florida Great Britain Great Britain Baltic North Sea Mediterranean Portugal Portugal Ccz = 281

970 28 426 7 159 130 303 180 330

June May Summer Winter July July Spring Winter Summer

60–120 120–300

Winter Summer

(c) From Kettle et al. database Kettle et al. (1999) Latitude of Toulon Kettle et al. (1999) Latitude of Toulon

proximately 75% of the total biomass recorded. These results are in accordance with those obtained by Keller et al. (1989), Simo et al. (1997) and Kwint & Kramer (1995). In zone P, the evolution of the concentration was quite irregular with a period of high density and biomass between March and May (around 70 µg.l−1 and 2700 cells.l−1 ), one of low values from the end of July to October (less than 40 µg.l−1 and around 1000 cells.l−1 ) and a winter period characterized by a high average value, great differences from one month to the next and by the highest and the lowest values recorded, in November/January and December/February, respectively. The evolution in the LP zone was more regular than in zone P, with the highest values recorded in spring and a winter period characterized by low and quite constant values. As it was the case for the DMS concentration, the values for the zone P were much higher (except in February and December when the values were almost equivalent): the average density in zone P was 2.0

times that of zone LP and the average biomass was 2.7 times that of zone LP (Table 2). The zooplankton densities over the whole study varied from 3000 ind.m−3 in zone LP in December to 38 000 ind.m−3 in zone P in September (Fig. 2). The mean density in zone P (15 970 ind.m−3) was much greater than that in zone LP (8595 ind.m−3) (Table 2), as were the DMS concentration and phytoplankton density and biomass. The concentration of chlorophyll a in zone P varied from 0.4 to 2.5 µg.l−1 with an average concentration of 1.24 µg.l−1 . They were 0.16–1.72 µg.l−1 in zone LP, with a mean of 0.67 µg.l−1 . The concentration of chlorophyll a was always higher in zone P (except in May) as were the DMS concentration and plankton density. The maximum ratio was recorded in March (4.08), while the annual average was 1.85 (Table 2). Results of annual bacteria measurements realized later on in the same area indicated that the concentrations in zone P (3.6 105 cells.l−1 ) were always slightly

112 Table 2. Values and ratio of the different parameters measured in the zones P and LP DMS

Mean zone P Mean zone LP Ratio P / LP

ng.l −1

Phytoplankton Biomass Density µg.l −1 cells.l −1

Chlorophyll a concentration µg.l −1

Zooplankton density cells.l −1

Bacteria abundance cells.l −1

365 263 1.4

53 19.8 2.7

1.16 0.63 1.8

15971 8594 1.85

3.6 105 2.9 105 1.25

1985 985 2.0

Discussion Our results show clearly two main results.

Figure 3. Monthly variations in sea surface temperature between 1951 and 1980 and during the year 1997.

greater (1.25 time in average) than in zone LP (2.9 105cells.l−1 ). As for orthophosphate (P–PO4 3− ) and total phosphorus (P–TP) concentrations (7.3±0.3 µM), dissolved oxygen (7.1±0.6 mg.l−1 ), conductivity (47±5 mS.cm−1 ) and salinity (3.59–3.80%) values of the water column were not significantly different (