2.2.1 Trace metals

Surface samples, i.e., SML and subsurface water (SSW: ∼ 1 m depth), were collected from a pneumatic boat deployed 500–1500 m from the research vessel in order to avoid contamination of the samples by the vessel. SML samples were collected using a glass plate sampler (Stortini et al., 2012; Tovar-Sánchez et al., 2019) that had been previously cleaned with acid overnight and rinsed thoroughly with ultrapure water (MQ water). Once at the station, the glass plate and the whole sampler were rinsed with seawater several times, and the three first dips (SML samples) were discharged. The 39×25 cm silicate glass plate had an effective sampling surface area of 1950 cm² considering both sides. In order to check for procedural contamination, we collected SML blanks at some stations on board the pneumatic boat by rinsing the glass plate with ultrapure water and collecting 0.5 L using the glass plate system. The sample signal to blank ratio was typically greater than 5 : 1 for all elements. The surface microlayer thickness (µm) was calculated using the following formula (Wurl, 2009):

where V_S is the volume sampled (mL), N the number of dips and A_p the surface area of the glass plate (cm²).

The total fraction of SML (i.e., T-SML) was collected directly from the glass plate system without filtration in acid-cleaned 0.5 L LDPE bottles, while the dissolved fraction in the SML (i.e., D-SML) was rapidly filtered on board the pneumatic boat through an acid-cleaned polypropylene cartridge filter (0.22 µm; MSI, Calyx^®). SSWs (∼1 m depth) were collected using acid-washed Teflon tubing connected to a peristaltic pump and directly filtered on the same cartridge to collect the dissolved fraction (D-SSW). All samples were acidified on board to pH < 2 with ultrapure-grade HCl in a class-100 HEPA laminar-flow hood. The metals (i.e., Cd, Co, Cu, Fe, Ni, Mo, V, Zn and Pb) were stored for at least 1 month prior to analysis. The samples were pre-concentrated using an organic extraction method (Bruland et al., 1979) and quantified by ICP-MS (Perkin Elmer ELAN DRC-e). Prior to pre-concentration and to ensure the breakdown of metal–organic complexes and the removal of organic matter (Achterberg et al., 2001; Milne et al., 2010), total fraction samples (i.e., T-SML) were digested using a UV system consisting of a UV (80 W) mercury lamp that irradiated the samples (contained in quartz bottles) over 30 min. The accuracy of the pre-concentration method and analysis for trace metals was established using Seawater Reference Material for Trace Elements (CASS 6, NRC-CNRC) and compares well with the following certified values given in brackets (in µg L⁻¹): Cd: 0.02 (0.022); Co: 0.07 (0.067); Cu: 0.44 (0.418); Fe: 1.58 (1.56); Mo: 8.1 (9.15); Ni: 0.41 (0.418); Pb:0.01 (0.011); V: 0.45 (0.49); Zn: 1.14 (1.127).

2.2.2 Ancillary parameters

The temperature and salinity of the surface seawater were measured with the underway thermosalinograph (TSG) system of the R/V Pourquoi Pas?, which was composed of a Seabird^® SBE 21 seaCAT linked to a SBE 38 thermometer situated at the seawater inlet. The seawater inlet was located 3 m under the sea surface. The wind speed at 10 m was measured with a Gill Windsonic ultrasonic anemometer from the onboard BATOS station deployed by the French meteorological agency Météo France on the vessel. Temperature, salinity and wind data were binned every 30 s by the ship's data management system TECHSAS (TECHnical Sensor Acquisition System). The average values of temperature, salinity and wind speed for a time period of 1 h around the time of SML sampling are shown in Table 1.

2.3.1 Neuston

The microorganisms inhabiting the SML are collectively referred to as the “neuston” (Engel et al., 2017). The microorganisms in the SML were sampled at the same time as the trace metal sample collection, also using a glass plate system (50×26 cm silicate glass plate with an effective sampling surface area of 2600 cm² considering both sides; Cunliffe and Wurl, 2014; Harvey, 1966). The water from the SSW was collected from around a 20 cm depth in acid-clean borosilicate bottles. The bacterial numbers were determined using flow cytometry from a 4 mL sample that was fixed with 200 mL glutaraldehyde (GDA, 1 % final concentration). The samples were stored at −20 ^∘C for 2.5 months at most until analysis and were then stained with SYBR Green I (Molecular Probes Inc.) (Marie et al., 1997). The samples were analyzed using a flow cytometer equipped with a 488 nm laser (Becton & Dickinson FACSCalibur). A plot of side scatter (SSC) vs. green fluorescence (FL1) was used to detect the unique signature of the bacterial cells. The internal standard consisted of yellow-green latex beads (Polysciences, 0.5 µm). The abundance and area of transparent exopolymer particles (TEPs) were measured microscopically following a previously described method (Engel, 2009).

2.3.2 Phytoplankton and primary production

Chl a concentration and primary production were measured in the SSW at a depth of 5 m. The primary production was measured with the ¹⁴C-uptake technique, following the methods detailed in (Marañón et al., 2000). The seawater samples, collected in Niskin bottles at dawn, were dispensed into four (three light and one dark) polystyrene bottles that were 70 mL in volume, which were amended with 15 µCi of NaH¹⁴CO₃ and incubated for 24 h inside a deck incubator refrigerated with surface seawater from the continuous water supply. The incubator was covered with a neutral density filter that provided an irradiance level of 70 % of incident PAR. After incubation, the samples were filtered, using low vacuum pressure, through 0.2 µm polycarbonate filters. The filters were then exposed to HCl fumes overnight to remove non-fixed, inorganic ¹⁴C. After adding 5 mL of liquid scintillation cocktail to the filters, the radioactivity of each sample was determined on board with a liquid scintillation counter. The dark-bottle DPM value was subtracted from the light bottle DPM value to compute the rates of carbon fixation. A value of 26 663 µgC L⁻¹ was used for the concentration of dissolved inorganic carbon (DIC), which corresponds to the mean DIC concentration at 5 m measured during the cruise. Chl a concentrations were measured by HPLC (HPLC Agilent Technologies 1200) following the method described by Ras et al. (2008).

3.3.1 Residence time of trace metals in the SML

Estimates of the residence times of metals from aerosol deposition in the SML are critical to understanding the biogeochemical processes that affect the fate and distribution of trace metals in the oceans' surface more. The equation proposed by Ebling and Landing (2017) was used to estimate the residence time (t) of particulate metals in the SML:

where [TE]_SML is the concentration of the trace element (TE) in the T-SML, d is the thickness of the SML and J_aerosol is the aerosol trace metal flux measured. J_aerosol was estimated by multiplying the metal aerosol concentrations with the deposition velocity, which is dependent on aerosol size. During the cruise, Al and Fe atmospheric concentrations were correlated at all the stations and the ratio Fe∕Al is typical of a crustal source (Fu et al., 2020). The atmospheric iron deposition fluxes are associated with mineral dust particles even during the period when the Saharan dust inputs are very low (Desboeufs et al., 2018; Guieu et al., 2010). Conversely, no correlation with Al is observed for the other metals, except during FAST 1–3. Thus, we used a mineral dust deposition velocity for Fe of 1 cm s⁻¹ and an average velocity of fine anthropogenic particles for the other metals, i.e., 0.1 cm s⁻¹ (Baker et al., 2010; Duce et al., 1991). For the calculations of the residence times throughout our different stations we used simultaneous empirical measurements of total metal concentrations in the SML (assuming that the metal in the total fraction is highly influenced by material from aerosol dust) and metal aerosol fluxes. Since Mo and Cd are not enriched in the SML they were not considered in this calculation. Although highly variable among stations, the average residence times (± SD) of Cu (5.8±6.2 h), V (2.2±1.0 h), Pb (1.8±3.4 h), Zn (1.5±1.2 h), Co (1.2±0.7 h), Ni (48±18 min) and Fe (3.6±6 min) were consistent with previous estimates in regions under low aerosol inputs (Ebling and Landing, 2017) (Table 4). The residence time calculation of Fe using the same deposition rate as rest of the metals (i.e., 0.1 cm s⁻¹) results in a residence time that is 10 times longer (i.e., average 0.6 h instead 0.06 h), even so, the residence time of Fe remains the shortest of all studied metals. The residence time of Cu (St. 3–4 and FAST 3–4), Fe (St. 3 and FAST 3) and Pb (St. 4) were an order of magnitude higher than other stations. Our results indicate that FAST 1–4 stations were affected by the dusty rain events, which increased the concentration of some metals and consequently the residence time in the T-SML (Table 4). However, the reasons for the increase at stations St. 3 and St. 4 are not evident. On average, while the highest residence times obtained for Cu and Pb are in agreement with their strong affinity to particles and therefore with a high probability of retention in the SML, other reactive elements such as Fe presented the shortest residence times. Since such a quick transfer of these metal particles to the underlying water (on the order of minutes) is unlikely (mainly due to the high content of organic matter in the SML) and the dissolution is not immediately reflected in an increase in the concentration in the dissolved fraction (i.e., D-SML), other variables (linked to physical processes, photochemistry or biological activity) probably affected the residence time of metals in the SML. Wind speed seems to not have affected the residence time of any metal in the SML (Table 3), which is probably due to the low speed registered during our campaign (9±4.99 kn) (Table 1).

Table 4Number of dips conducted to collect 500 mL of SML, thickness (µm) of the SML and calculated residence times (h) of particulate metals in the SML derived for aerosol depositions.

Download Print Version | Download XLSX

3.3.2 Neuston composition

The abundance of different microbial groups (bacteria; high-nucleic-acid-content bacteria: HNA; low-nucleic-acid-content bacteria: LNA; picophytoplankton) in the SML is shown in Table 2. Bacterial abundance in the SML ranged from 2×10⁵ to 1×10⁶ cell mL⁻¹ (average ± SD: $\mathrm{5.1}\times {\mathrm{10}}^{\mathrm{5}}\pm \mathrm{2.2}\times {\mathrm{10}}^{\mathrm{5}}$ cell mL⁻¹), which is of the same order of magnitude as the abundance measured in the SML of the coastal Mediterranean Sea (Joux et al., 2006; Obernosterer et al., 2005) and other regions (e.g., along the Peruvian Coast, with an average of $\mathrm{8.9}\times {\mathrm{10}}^{\mathrm{5}}\pm \mathrm{4.3}\times {\mathrm{10}}^{\mathrm{5}}$ cell mL⁻¹) (Agogue et al., 2005; Engel and Galgani, 2016; Joux et al., 2006; Zäncker et al., 2018). The bacterial community was dominated by low-nucleic-acid-content bacteria (LNA) with an average abundance (± SD) of $\mathrm{2.8}\times {\mathrm{10}}^{\mathrm{5}}\pm \mathrm{1.0}\times {\mathrm{10}}^{\mathrm{5}}$ cell mL⁻¹. Bacterial abundances did not differ significantly between SML and SSW (Table 2) in line with previous studies in coastal Mediterranean Sea that reported that the SML was only weakly enriched in heterotrophic bacteria (on average 1.1-fold in Joux et al., 2006, and mean 1.3 in Obernosterer et al., 2005) and comparable to the ratio in other environments, such as the Pacific (Obernosterer et al., 2008). Phytoplankton was only slightly (but significantly; t test, p=0.002, n=12) enriched in the SML with an average enrichment of 1.5 compared to the SSW; such enrichments were also observed at coastal sites of the Mediterranean Sea (Joux et al., 2006).

Microbial abundance decreases from west to east, reflecting the increasing oligotrophy (mostly due to P limitation) of the surface of Mediterranean waters (Pulido-Villena et al., 2012). Microbial abundance in the SML and reactive elements (i.e., Cu, Fe, and Zn) in the T-SML showed the same longitudinal gradients in this study, with a decreasing eastward concentration along the southern coast of the MS. In fact, bacterial abundances were significantly and positively correlated with these bioactive T-SML metals (i.e., r_s: Cu: 0.65 p<0.01; Fe: 0.53 p<0.05; and Zn: 0.49 p<0.05), suggesting that bacterioneuston could affect the concentration and fate of Cu, Fe and Zn in the SML. Bacteria could efficiently assimilate the fraction of Cu, Fe and Zn available, favoring a decrease in the D-SML fraction (Tables 1–2). No general relationship between the concentrations of metals and TEPs (high-molecular-weight polymers released by phytoplankton and bacteria, with a high metal-binding capacity; Passow, 2002) were found in the SML (Table 5). Metal assimilation by microbial communities could explain the higher residence time of Cu and Zn (on the order of hours) in the SML, although information about the metal content in seston would be necessary to corroborate this hypothesis. However, in the case of Fe with an estimated residence time of a few minutes, other processes in addition to wind speed and neuston uptake should contribute to facilitating the transfer from the SML to the underlying water. For example, photochemical reactions driven by exposure to intense solar radiation in the SML could play an important role in the dissolution processes of this metal (Boyd et al., 2010). On the other hand, Ni was strongly and negatively correlated with bacterial abundance in the D-SML ($r_{\mathrm{s}}=-\mathrm{0.93}$, p<0.01; R²=0.74, p<0.01) suggesting, in contrast to Cu, Fe and Zn, a possible inhibiting role on the microbial growth (Table 5 and Fig. 2) (see the next section for more discussion).

Figure 2Concentration of dissolved Ni plotted against bacterial abundance in both the surface microlayer (SML, black dots) and subsurface water (SSW, red open dots). The lines represent least-square linear regressions.

Download

Table 5Spearman's rank correlation coefficients for selected parameters: bacteria; HNA: high-nucleic-acid-content bacteria; LNA: low-nucleic-acid-content bacteria; picophyto: picophytoplankton; transparent exopolymer particles: TEPs; primary production: PP; and chlorophyll a: Chl a. Significant correlations at p<0.05 and p<0.01 are marked with one asterisk (italic numbers) and two asterisks (bold numbers), respectively.

Download Print Version | Download XLSX

Figure 3Salinity and concentration of dissolved trace metals in subsurface waters (SSW) plotted against longitude. The dashed lines represent least-square linear regressions.

Download

3.3.3 Subsurface water

The D-SSW concentrations of Cd, Co, Cu, Ni, Mo and Zn showed a longitudinal gradient of concentrations increasing from west to east, with significant positive correlations with longitude for Cd (r²: 0.60; p<0.05), Co (r²: 0.77; p<0.01) and Ni (r²: 0.80; p<0.01) (Fig. 3). This trend is consistent with previous studies, where the increasing eastward trend in concentration along the southern coast of the MS has been suggested to result from several factors, such as more intense Saharan deposition on the eastern MS (Guieu et al., 2002); more rapid exchange of water masses and margin inputs in the western MS (Yoon et al., 1999); or, as suggested for Co, the eastward regeneration of biogenic particulates, which yields a decrease towards the west of the dissolved Co in surface water (Dulaquais et al., 2017). Since surface salinity showed the same eastward increase and was closely correlated with those metals (r_s ranged from 0.51 p<0.05 for Mo to 0.97 p<0.01 for Ni; Table 3), the exchange with the surface Atlantic Ocean waters seems to be the main cause of this gradient of concentrations in our study. Other metals (i.e., Fe, Pb and V) did not show any clear geographical trend, and therefore the variations in their surface concentrations could be influenced by factors other than dilution or exchange, such as vertical diffusive fluxes or specific metal sources, as in the case of Fe and Pb, which have been suggested to be more affected by atmospheric inputs (Nicolas et al., 1994; Yoon et al., 1999). In fact, Pb was the only element that showed significant positive correlation with latitude (r_s: 0.88 p<0.01, Table 3), suggesting an influence of the more heavily industrialized northern region of the MS.

D-SSW concentrations of Ni were strongly and negatively correlated with microbial abundance (mainly with heterotrophic bacteria) in the underlying water (r_s: −0.91, p<0.01; R²=0.87, p<0.01) (Fig. 2 and Table 5), which could be interpreted as an indication of a potential negative role of this metal on bacteria and small phytoplankton in the top meter of the western surface of the MS, including the SML. The toxicity of Ni at concentrations of ∼50 nM has been previously demonstrated in the western MS, with inhibitions of 10 % (EC₁₀) in phycoerythrin and Chl a signals of a natural population of the picoplankton Synechococcus sp. (Debelius et al., 2011). Although that toxicity concentration (tested in picoplankton) is around 13 times higher than the average values (± SD) measured in our samples (T-SML: 4.1±0.5 nM; D-SML: 3.9±0.6 nM; and D-SSW: 3.2±0.6 nM; see Table 1), deleterious effects on the neuston and microbial communities at lower concentrations might be possible in the top meter of the sea surface. The toxicity to phytoplankton of divalent, cationic trace metals, such as Ni or Cu, is probably controlled by its free metal ion concentration (Donat et al., 1994). Although the Ni interactions with dissolved organic matter have not been studied well in seawater, they are thought to occur partly as stable organic complexes and with slow dissociation rates (Wen et al., 2011). However, intense UV radiation can alter the concentration, structure, reactivity and metal-binding capacity of the organic matter, thus increasing the proportion of free metal ions and their bioavailability and/or potential toxicity (Cheloni and Slaveykova, 2018). Even if a general decreasing trend in microbial abundance from west to east due to the increasing oligotrophy was observed, it is pertinent to mention that primary production and Chl a concentration (measured a 5 m depth) did not show any significant correlations with Ni (Table 5). It is therefore assumed that the potential toxicity of Ni would mainly affect the bacterial community in the top meter of the ocean's surface. Nickel, like other transition metals, is an essential cofactor of several enzymes; however, it becomes toxic when homeostasis fails. Multiple potential mechanisms of Ni toxicity to aquatic organisms, and in particular to bacteria, have been identified (Macomber and Hausinger, 2016). Among the different possible toxicity mechanisms (including the inhibition of Zn and Fe metalloenzymes and non-metalloenzymes) the toxicity involving reactive oxygen species (ROS) is probably the highest in surface seawater. While Ni itself is a poor generator of ROS when compared to other metals like Fe or Cu, its reactivity and ROS production can be enhanced by the displacement of redox-active iron from iron metallocenters (Macomber and Hausinger, 2016) or when chelated by oligopeptides and histidine (Brix et al., 2017), which are abundant in the SML. It appears that nickel-dependent toxicity involving ROS could be a mechanism of oxidative stress in microbial organisms of the surface of oceans.