Spring molybdenum enrichment in scallop shells: a potential tracer of diatom productivity in temperate coastal environments (Brittany, NW France)

. Skeletal molybdenum/calcium ([Mo]/[Ca]) shell ratios were examined in shells of the Great Scallop


Introduction
Mollusc bivalves grow through an incremental deposition of calcium carbonate (CaCO 3 ) layers with most species exhibiting specific temporal marking in shell composition (Stecher et al., 1996;Chauvaud et al., 1998Chauvaud et al., , 2005. Bivalves can preserve, within their own exoskeleton, a chronological record of the environmental variations they have experienced during their life. Following this growth pattern, several investigations have demonstrated that variations of the historical elemental composition along the shell growth axis can be used as proxies for coastal biogeochemical processes (Dodd, 1965;Lorens and Bender, 1980;Klein et al., 1996a, b). For Published by Copernicus Publications on behalf of the European Geosciences Union.

234
A. Barats et al.: Spring Mo shell enrichment reveals diatom productivity? most of the species (Mercenaria mercenaria, Mytilus edulis, Isognomon ephippium, Ensis siliqua), trace element profiles within the shell, provide chronological records of environmental conditions experienced by the bivalves (Puente et al., 1996;Stecher et al., 1996;Giusti et al., 1999;Vander Putten et al., 2000;Richardson et al., 2001;Lazareth et al., 2003;Gillikin et al., 2005). In these studies, an approximate date of formation has often been assigned to the shell deposit to interpret historical information. Different methods such as shell sclerochronology or oxygen stable isotope composition have demonstrated that the Great Scallop Pecten maximus (L.) develops the specific feature to precipitate a distinct daily stria of CaCO 3 on its shell . This scallop has been promoted as a good candidate for environmental proxy calibration on a daily scale Lorrain et al., 2000Lorrain et al., , 2005Barats, 2006;Barats et al., 2008Barats et al., , 2009). This bivalve grows in coastal environments over a wide range of latitudes (30 • to 60 • N) and bathymetry (1 to 500 m). At temperate latitudes such as in France (Brittany), the scallop shell exhibits (i) a fast daily growth rate (maximum 350 µm per day) and (ii) an extended growth period, approximately from April to November, during the second year of growth (Chauvaud, 1998;Chauvaud et al., 2005). The shell growth rate is influenced by seawater temperature (scallop grows only if T sw >9-10 • C), and limited by massive or toxic phytoplankton blooms (Lorrain et al., 2000;Owen et al., 2002a, b). P. maximus is a non-selective filter feeder ingesting both dissolved and particulate material within the bottom waters surrounding the scallop (Chauvaud, 1998;Chauvaud et al., 1998Chauvaud et al., , 2005Lorrain et al., 2005). The variations of trace element shell concentrations provide significant and highly detailed historical information (Freitas et al., 2005(Freitas et al., , 2006Lorrain et al., 2005;Barats, 2006;Barats et al., 2007Barats et al., , 2008. For example, manganese (Mn) concentrations along the shell were recently demonstrated to be mainly governed by concentrations of dissolved Mn at the SWI (sediment water interface) being themselves controlled by freshwater inputs and benthic release (Freitas et al., 2006;Barats et al., 2008). More recently, ([Ba]/[Ca]) shell and ([Mo]/[Ca]) shell ratios were examined in Pectinidae because their variations highlighted intriguing profiles with a sharp annual increase during the spring Gillikin et al., 2008;Thébault et al., 2009).
Molybdenum (Mo) is one of the most abundant transition group metals in seawater, mainly present as the oxy anion MoO 2− 4 in oxygenated environment (Collier, 1985). Mo is generally considered to be conservative in pelagic waters with a concentration averaging 110 nmol/L in seawater, suggesting little influence of biogeochemical processes on its concentration (Collier, 1985). Coastal Mo distribution is also influenced by freshwater-seawater mixing (Dalai et al., 2005). Some studies however, highlight that Mo can also present non-conservative behaviour in coastal waters, both at the SWI (Adelson et al., 2001;Chaillou et al., 2002;Cru-sius et al., 1996;Elbaz-Poulichet et al., 2005), and in the water column (Tuit and Ravizza, 2003;Dellwig et al., 2007). Mo benthic inputs to the SWI can be induced under suboxic conditions via the reduction of sedimentary manganese oxides leading to Mo release in the overlying water as the dissolved compound, MoO 2− 4 (Crusius et al., 1996;Crusius and Thomson, 2000;Morford et al., 2001;Chaillou et al., 2002;Dalai et al., 2005;Elbaz-Poulichet et al., 2005;Morford et al., 2005). Mo released at the SWI may then diffuse back into the water column or can be authigenically reprecipitated into Mo-Fe-S forms under strict anaerobic conditions (Erickson and Heltz, 2000;Sundby et al., 2004;Tribovillard et al., 2004;Zheng et al., 2000). In the water column, Mo assimilation by diazotrophic cyanobacteria and phytoplankton is an essential catalytic factor for the majority of N 2fixing organisms and many nitrate reductase systems (Collier, 1985;Hille, 2002). Mo availability may act to limit N fixation in marine ecosystems, and consequently, may limit primary productivity (Marino et al., 1990;Cole et al., 1993). Various marine phytoplankton species were investigated for their cellular Mo contents, which were reported to be low and homogeneous (3.3 mmol Mo/mol of P, or 3.1 µmol/L of cellular volume) among the 15 phytoplankton species investigated (diatoms, green algae, coccolithophores) . This concentration obtained under identical culture conditions was apparently independent of the phytoplankton species and did not exhibit any specific assimilation related to the phylogenetic origin of the investigated species, in contrast to other micronutrients Quigg et al., 2003). Mo content in marine phytoplankton can however be enriched due to different irradiance conditions (Finkel et al., 2006). Recently, non-conservative behaviour of Mo was revealed in the water column of a coastal environment (Wadden Sea, Germany) (Dellwig et al., 2007). In summer, Mo was demonstrated to be enriched in suspended particulate matter and depleted in the dissolved phase of the seawater. The bacterial decomposition of phytoplankton was reported to promote the release of organic compounds and the formation of Mo-enriched aggregates which may thus settle to the SWI to be rapidly decomposed by microbial activity contributing to a substantial release of Mo in bottom waters (Dellwig et al., 2007).
Through the observation of ([Mo]/[Ca]) shell ratios along the daily striae of P. maximus, the objectives of this study are first, to evaluate ([Mo]/[Ca]) shell profiles as a potential record of specific biogeochemical processes occurring at the SWI, and second to provide new confirmation of the non-conservative behaviour of Mo in coastal waters. The quantitative micro-analysis of shell Mo content was previously developed using Laser Ablation -Inductively Coupled Plasma -Mass Spectrometry (LA-ICP-MS) and matrixmatched standards . Quantitative chronological profiles were defined with an accurate date and concentration assignment for each measured striae. A first evaluation of (  . Only the upper valve of the shell was considered ). Shells were cleaned by soaking in 90% acetic acid for 45-60 s to remove bio-fouling, rinsed with distilled water, and air-dried. A 45×10 mm cross section corresponding to the second year of growth (juvenile stage) was cut along the axis of maximum growth rate (Chauvaud, 1998;Chauvaud et al., 1998). A precleaning ablation of the calcite surface was carried out before LA-ICPMS analyses to avoid surface contamination. This precleaning step consisted of a quick (approximately 20 s) pre-ablation of the sample shell stria surface at a rate of 50 µm/s .

Determination of trace element concentrations in shells
Quantitative analyses of trace elements within shells were performed by LA-ICP-MS consisting of coupling a UV laser ablation unit (LSX 100 UV 266 nm, Cetac Tech.) to an ICP-MS (X7 series, Thermo Fisher). The methodological approach and analytical performances have been described in detail elsewhere (Barats et al., , 2008). Briefly, a matrix-matched external calibration was performed with lab prepared CaCO 3 pellets. The Mo calibration curve displayed good linearity with a regression coefficient above 0.99. The detection limit was approximately 27 nmol/g, and the relative standard deviations obtained for both analytical repeatability and reproducibility were below 7%. The ([Mo]/[Ca]) shell ratios were calculated dividing shell Mo concentrations by the calcium concentration in the shell (400 mg/g), and expressed in units of µmol/mol (Barats et al., , 2008). Shell Ca concentration was assumed to be constant all along the shell surface. Richard (2009) recently investigated the spatial variability of the Ca concentration in the same collection of juvenile scallop shells intercomparing two independent analytical methods. High resolution ICPMS (n > 30, mass spectrometry) and Electron micro-probe (n = 3, X-ray electronic spectroscopy) analysis demonstrated that Ca content was respectively 41.1±3.1% and 39.1±1%, within the same stria or along growth section of the shell surface. Shell analyses were performed each third stria to obtain a temporal resolution of 3 days. A date of formation was assigned to each ablated sample by backdating from the harvest date and based on the daily periodicity of stria formation in P. maximus. Chronological profiles of trace elements were then established based on the stria specific date assignment. The uncertainty of the chronological time-scale of mean shell profiles was estimated to be ±3 days. An evaluation of the shell growth rate was carried out using an image analysis technique previously described (Chauvaud, 1998).

Environmental monitoring
In  . A specific survey (2-3 days resolution) was also performed at Roscanvel in 2000 (from February to December). Bottom waters were regularly sampled (every 2-3 days) by a diver-operated Niskin sampler positioned and closed horizontally at 1 m above the SWI to avoid any disturbance the SWI and thus to preserve the characteristic of the bottom water column. After collection, the samples were filtered (<0.6 µm, Nucleopore) and acidified in 2% HNO 3 (69-70% Suprapur, Merck). Before analysis, they were diluted 50 times with Milli-Q water (Millipore). Two internal standards were also added (Y and Bi) to the diluted samples. Dissolved concentrations of Mo were then determined by ICP-MS (X7 series, Thermo Fisher) by external and internal standard calibration.

Statistical analyses
Statistical data treatment was performed to highlight environmental parameters that can vary with the change of (

Reproducibility of ([Mo]/[Ca]) shell profiles in a same scallop population from the Bay of Brest (Brittany, France)
The juvenile shell (second year of growth in 2003) of three scallops from the Bay of Brest (Roscanvel) were examined (Fig. 1a). The ([Mo]/[Ca]) shell concentration ratios showed a similar profile with average background concentrations below the detection limit (<2.7 nmol/mol), and 5 significant enrichments from May to October (Fig. 1a). A comparison of these Mo profiles among the 3 individual scallop shells reveals significant correlations (r 2 > 0.73, p < 0.05, n > 60; Table 1). This result underlines statistically a high reproducibility of ([Mo]/[Ca]) shell profiles among a same scallop population. As a consequence, an averaged ([Mo]/[Ca]) shell profile, defined as a mean of 3 shell profiles, can be established and also shows significant spring and summer enrichments ( Fig. 1b; Table 1). Analyses of a three-year-old scallop from the same population were performed from its third year of shell growth during 2003 (Fig. 1b)

Ubiquitous occurrence of ([Mo]/[Ca]) shell maxima in several scallop populations from temperate coastal environments
The comparison of (  (Table 2). These results strongly suggest that the Mo/Ca shell pattern as measured along the North West French coast indicates that the Mo uptake is a reproducible and ubiquitous phenomenon in this species and occurs irrespective of time, population or geographic location.

Recurrence of ([Mo]/[Ca]) shell maxima in scallop shells sampled from the Bay of Brest (Brittany) over a 7-year period
The inter-annual study was carried out over a 7-year period (1998 to 2004)     Apr-00 Jun-00 Aug-00 Oct-00 Dec-00

Influence of water column biogeochemistry on ([Mo]/[Ca]) shell enrichment in the Bay of Brest (1998 to 2004)
The complete environmental database was examined during the 7-year survey (1998)(1999)(2000)(2001)(2002)(2003)(2004)  ) shell maximum events usually occurred during a characteristic period of nutrient depletion (late spring). The highest depletion of silicic acid between winter and spring and its further phytoplankton uptake were always related to important maxima of ([Mo]/[Ca]) shell ratio (Fig. 4d). The comparison of ([Mo]/[Ca]) shell profiles with Pseudonitzschia spp. and silicic acid concentrations underline decreasing or minimum silicic acid concentration and the occurrence of a Pseudonitzschia spp. bloom which preceded or were concomitant with ([Mo]/[Ca]) shell maxima (Fig. 4d).
These overall observations suggest that Mo shell uptake is promoted by the significant spring pelagic productivity, which can induce changing conditions at the SWI in the environment surrounding the scallop. The 7-year survey demonstrates however that the ([Mo]/[Ca]) shell maxima cannot be directly related to specific phytoplankton species.

Evolution of the Mo partitioning between the shell and the seawater in 2000 (Bay of Brest)
A In order to better constrain the conservative or nonconservative behaviour of Mo in the bottom waters of the Bay of Brest (Roscanvel), an evaluation of the "apparent" distribution between the shell and the seawater was performed, first assuming that the particulate phase Mo is not involved. The distribution of Mo between solid carbonate and seawater is expressed by the partition coefficients:   al., 2007) which exhibits transient enrichment and depletion of Mo concentration in the particulate and dissolved phases, respectively. However, the comparison between that study and our work remains difficult to address considering the distinct experimental objectives, approaches and characteristics of the two ecosystems.
To support a dietary particulate uptake of Mo by the scallop, Mo measurements in soft tissues of Pecten maximus bivalves collected at Roscanvel in January 2003 also revealed the highest Mo concentration in the digestive gland (167±63 nmol/g), which represents 69% of the total Mo in the bivalve (soft tissues and shell included) (Barats, 2006). A study of Mo bioaccumulation in scallop tissues and organs revealed the greatest concentration of Mo in the digestive gland (Bustamante and Miramand, 2005). In addition, Mo in the digestive gland appears to be mostly bound to a soluble fraction (Bustamante and Miramand, 2005). Scallop ingestion of suspended particles would thus provide available dissolved Mo to precipitate in the calcite shell. Both shell uptake of dissolved and particulate Mo can be thus considered to induce maximum ([Mo]/[Ca]) shell levels.

Influence of benthic processes on Mo inputs to bottom waters
Benthic Mo inputs were first examined because of the extended knowledge on benthic processes that lead to the release Mo at the SWI essentially from the reductive dissolution of manganese oxides (Crusius et al., 1996;Crusius and Thomson, 2000;Morford et al., 2001;Chaillou et al., 2002;Dalai et al., 2005;Elbaz-Poulichet et al., 2005;Morford et al., 2005). Because dissolved Mn is also a good proxy for reductive benthic exchange at the SWI, variations of ([Mn]/[Ca]) shell ratio in 2000 were thus examined concomitantly with ([Mo]/[Ca]) shell ratio, dissolved Mo and Mn (Fig. 5a, b). In addition, ([Mn]/[Ca]) shell ratio were found to be correlated to both riverine and benthic inputs under reductive summer conditions in the Bay of Seine (Barats et al., 2008). In the Bay of Brest, variations of ([Mn]/[Ca]) shell ratio in 2000 were slightly decreasing in spring, steady in summer and slightly increasing in autumn (Fig. 5b). dissolved Mn in bottom seawater was not significantly higher than the rest of the year, and the water column was well mixed and oxygenated by the tidal dynamics which argues against reductive conditions (Fig. 5a- ) shell ratio also reveal a steady ratio averaging 700 µmol/mol punctuated by a maximum of 0.250 mol/mol (Fig. 5a) demonstrating distinct shell uptake routes for Mo and Mn during the spring Mo enrichment. In the Bay of Brest, benthic inputs from reductive dissolution of Mn oxides are probably not a major source of Mo for the shell. This conclusion is similar to the one suggested by Thébault et al. (2009), but does not preclude that other benthic processes can be involved.
Releases of substantial amounts of Mo from organic matter degradation during an algal decay at the SWI were previously suggested to be the best explanation for Mo enrichment in bottom waters (Kunzendorf et al., 2001;Dellwig et al., 2007). The observation in 2000 of increasing NH + 4 concentrations in seawater concomitantly with higher Mo concentrations in bottom seawater (Fig. 5) also confirm the rapid mineralization of biogenic material settling during the post bloom period (Bally et al., 2004).

Influence of a biogenic pelagic process on Mo inputs to bottom waters
The Mo input in bottom waters is assumed to be induced by pelagic biogenic processes in May 2000 as the phytoplankton concentration was increasing. Chl a and POC concentrations displayed two maximum concentrations (respectively, 2 and 15 May) with the highest maximum on 2 May ([Chl a]=3.8 µg/L and [POC]=412 µg/L) (Fig. 5d) (Sellner, 1997). Trichodesmium spp. blooms do not occur because the seawater temperature in spring is too low (<21 • C). Richelia intracellularis is a small cyanobacterium living in endosymbiotic association within some diatom genera such as Rhizosolenia and Chaetoceros, which are dominant during spring in the Bay of Brest. This association of R. intracellularis as an endosymbiont in Rhizosolenia or Chaetoceros diatoms usually occurs in warm tropical seawater (Gomez et al., 2005), and is specific to nitrate depleted ecosystems such as the Baltic Sea or the North Atlantic Ocean, and rarely observed in coastal temperate environment (Sellner, 1997;Villareal, 1992). In the Bay of Brest, no other specific pathway explaining Mo enrichment within phytoplankton can be established during the productive period of the year. This statement also agrees with the Mo concentration in phytoplankton cell data obtained from experimental culture on various genera and species which did not show Mo cell enrichment under controlled growing conditions Quigg et al., 2003). Mo cell enrichment would however, be induced by environmental and physiological conditions such as light limitation (Finkel et al., 2006) or nitrate utilization (Thébault et al., 2009). This last hypothetic process could be conclusive for the Bay of Brest exhibiting significant nitrate inputs from human activities (Ragueneau et al., 2002).

Influence of diatom productivity on Mo inputs to bottom waters
([Mo]/[Ca]) shell peak levels were usually recorded from May to July and correspond to the first major phytoplankton bloom period occurring in the water column of temperate coastal environments. This first major phytoplankton bloom is dominated by diatom species. Scallop Mo uptake and its spring enrichment may thus be related to diatom productivity. Until now, none of these diatom species have been recognized to have an effect on Mo biogeochemistry in the marine environment. Statistical data analyses were performed, using the data from the 7-year survey (1998)(1999)(2000)(2001)(2002)(2003)(2004) in the Bay of Brest, to highlight the parameters potentially related to the amplitude of ( (r 2 = 0.53, p < 0.05, n = 14) or with the maximum concentration of Chl a (r 2 < 0.32, p < 0.05, n = 14). Multiple regression analyses also underlined a relevant relationship expressing the amplitude of maximum ([Mo]/[Ca]) shell ratios (µmol/mol) according to the maximum relative abundance of Pseudonitzschia spp. (%) and the maximum concentration of silicic acid (µmol/L) in the seawater (r 2 = 0.40, p < 0.05, n = 14), these two parameters are not significantly correlated. The significance of the regression was improved by diatoms (Ragueneau et al., 2002) and that Pseudonitzschia spp. blooms usually occur during low or depleted silicic acid concentrations in seawater (Gomez et al., 2007;Pan et al., 1998;Parsons and Dortch, 2002;Prego et al., 2007). Like silicic acids, nitrates are actively taken up by diatoms (Ragueneau et al., 2002). To further establish the underlying association between diatom productivity and Mo enrichment in the scallop shell, the apparent amount of silicic acid or nitrate uptake by diatoms (i.e. "apparent" diatom spring productivity) have been evaluated. This includes subtracting the average minimum concentration measured in late spring (n = 9, May and June), compared to the average maximum one observed in winter (n = 11, December to February). As expected from our previous observations, these "apparent" spring nutrient depletions were significantly correlated to the maximum ([Mo]/[Ca]) shell ratio both for silicic acid (r 2 = 0.878, p < 0.05, n = 6, Fig. 6) and for nitrates (r 2 = 0.780, p < 0.05, n = 6). These results indicate also that the maximum ([Mo]/[Ca]) shell ratios were not directly related to instantaneous total biomass but rather to a post-bloom period. This period is characterised by silicic acid or nitrate depletion and Pseudonitzschia spp. dominance, subsequent to the major diatom production integrated over the spring growth season. Mo inputs at the SWI can thus be induced by a diatom biogenic material downward flux. Diatoms are the sole marine phytoplankton taking up Si and are characterized by large cell size and density (Sarthou et al., 2005) (Thébault et al., 2009). Our data may support these assumptions, but none of these pathways can be completely demonstrated. The processes governing Mo scavenging by biogenic particles and its further uptake by the scallop remain to be elucidated.

Conclusions
An original investigation of ([Mo]/[Ca]) shell ratio in bivalves from temperate environments is reported for the first time. The approach uses ([Mo]/[Ca]) shell profiles along the growth period of the shell and these were determined for several scallops from a same population and during a 7-year period (1998)(1999)(2000)(2001)(2002)(2003)(2004) in temperate coastal ecosystems. These profiles exhibit similar features showing a background concentration (<2.7 nmol/mol) punctuated by a transient maximum in spring (May to June). This study reveals a new evidence of the non-conservative behaviour of the Mo in coastal waters, and demonstrates the specific Mo enrichment in the surrounding shell habitat during spring-time. The extent of spring ([Mo]/[Ca]) shell enrichments were explained by the net uptake of silicate and nitrate, suggesting a connection with diatoms' spring-time productivity. The use of Mo records in Pecten maximus shells may further serve as a new proxy for biomonitoring studies in temperate coastal environments which can be extended for other marine biogenic carbonates.