Effects of increased pCO2 and temperature on trace element (Ag, Cd and Zn) bioaccumulation in the eggs of the common cuttlefish, Sepia officinalis

35 Cephalopods play a key role in many marine trophic networks and constitute alternative 36 fisheries resources, especially given the ongoing d ecline in finfish stocks. Along the European 37 coast, the eggs of the cuttlefish Sepia officinalis are characterized by an increasing 38 permeability of the eggshell during development, wh ich leads to selective accumulation of 39 essential and non-essential elements in the embryo. Temperature and pH are two critical 40 factors that affect the metabolism of marine organi sms in the coastal shallow waters. In this 41 study, we investigated the effects of pH and temper ature through a crossed (3x2; pH 8.1 42 (pCO2, 400 ppm), 7.85 (900 ppm) and 7.6 (1400 ppm) at 16 and 19°C, respectively) 43 laboratory experiment. Seawater pH showed a strong effect on the egg weight and non44 significant impact on the weight of hatchlings at t he end of development implying an egg 45 swelling process and embryo growth disturbances. Th e lower the seawater pH, the more 46 Ag was accumulated in the tissues of hatchlings. Th e Cd concentration factor (CF) 47 decreased with decreasing pH and Zn CF reached maximal values pH 7.85, independently 48 of temperature. Our results suggest that pH and tem perature affected both the permeability 49 properties of the eggshell and embryonic metabolism . To the best of our knowledge, this is 50 one of the first studies on the consequences of oce an acidification and ocean warming on 51 metal uptake in marine organisms, and our results i ndicate the need to further evaluate the 52 likely ecotoxicological impact of the global change on the early-life stages of the cuttlefish. 53 54 55


INTRODUCTION
Atmospheric carbon dioxide (CO 2 ) concentration has increased from 280 parts per million (ppm) prior to the beginning of the industrial revolution to a current value of 380 ppm due to human activities (Solomon et al., 2007).It is now rising at a rate of ca.3.3% year -1 (Canadell et al., 2007) that will give a concentration of 700 ppm by the end of this century, according to the Intergovernmental Panel on Climate Change (IPCC) business-as-usual CO 2 emission scenario (Solomon et al., 2007).Increasing atmospheric CO 2 may have important consequences for the Earth's climate, leading to an average warming of 3°C at the Earth's surface over the course of this century (Solomon et al., 2007).Similar trends are expected for surface ocean temperature due to the warming of the surface mixed layer (Levitus et al., 2005).Surface ocean CO 2 partial pressure (pCO 2 ) is also expected to increase in proportion to the atmospheric CO 2 increase due to the oceanic uptake of anthropogenic CO 2 (Sabine et al., 2004).Increasing pCO 2 in the surface ocean causes major shifts in seawater carbonate chemistry and is likely to reduce pH by 0.2-0.4 units over the course of this century (Caldeira and Wickett, 2005).Such acidification of surface waters could affect marine organisms and in particular those having carbonate skeleton such as corals, coralline algae, foraminifera and coccolithophores for which calcification rates may decrease by 0-56% (see review by Kleypas et al., 2006).In addition to these biological effects, new data are emerging on the disturbances of physiological process such as growth, development, metabolism, ionoregulation and acidbase balance under elevated temperature and pCO 2 (e.g.Fabry et al., 2008;Widdicombe and Spicer, 2008;Pörtner et al., 2004;Pörtner, 2008).Moreover, it is widely accepted that early life stages may be the more sensitive to high pCO 2 (Pörtner and Farell, 2008) especially in invertebrates (Kurihara, 2008;Dupont and Thorndyke, 2009).Among the latter, cephalopods play a key role in many marine trophic foodwebs and constitute alternative fishery resources in the context of the ongoing decline in finfish stocks.In physiological terms, they are complex organisms with an active lifestyle and high levels of performance, e.g.high metabolic and growth rate (Pörtner et al., 1994).Recent studies have focused on the responses of these organisms to increasing temperature and pCO 2 (Melzner et al., 2007;Rosa and Seibel, 2009) and reported that their low oxygen-carrying blood protein was a target of their expected vulnerability to global warming and ocean acidification.Indeed, oxygen affinity of their haemocyanins decreased with decreasing pH (Bridges, 1994) and increasing temperature (Zielinski et al., 2001), subsequently reducing their metabolic scope.Nonetheless, data on the potential impact of both these variables on the cephalopod early life stages are relatively scarce.On the one hand, the temperature-dependence of development time in the cephalopod egg is well described (Boletzky, 1974), viz.as temperature decreases the development time increases.Moreover, temperature affects the use of the energy budget supplied by the yolk, increasing respiration of the cuttlefish embryo (Wolf et al., 1985) and reducing its growth rate (Bouchaud and Daguzan, 1989).On the other hand, D' Aniello et al. (1989) reported that squid eggs developing in acidified seawater showed reduced survival of the larvae.More globally, in the Coleoid common cuttlefish, Sepia officinalis, pH could interact with the egg development in two ways: first, the eggshell hardens once the egg is laid and becomes thicker due to pH-induced seawater polymerization of the mucopolysaccharidic components of the nidamental secretions (Gomi et al., 1986).The eggshell therefore aims at protecting the embryo against the external environment, e.g.microbial attack and predation (Boletzky, 1986) but also limits gas diffusion during the first developmental stages (Wolf et al., 1985).
Secondly, later in the development the cuttlefish embryo experiences low pH in its surrounding medium, i.e. the perivitelline fluid, because of the rising level of CO 2 as a product of the embryo respiration (Gutowska and Melzner, 2009).In this context, increasing pCO 2 in seawater could impact both the egg structure and the embryonic development.
Finally, when cuttlefish Sepia officinalis migrate during the breeding season into shallow waters to spawn (Boucaud and Boismery, 1991), the eggs laid here are thus subject to acute and/or chronic exposure to the various contaminants such as metals which are released from the human activities in the marine environment.Exposed to various dissolved trace elements, the cuttlefish eggshell is likely to act as a protective barrier that limits or hinders the incorporation of metals into the embryo during the first developmental stages, with a permeability that is element-specific (Bustamante 2002, 2004, 2006, Lacoue-Labarthe 2008a).
The subsequent incorporation of water into the perivitelline fluid as the egg swells appears to be a key process in metal penetration.Thus, we hypothesized that, in coastal shallow waters, ocean acidification and warming could affect embryonic metabolism and the shielding properties of the eggshell components, and could lead to shifts in a) the accumulation of essential element (Zn) and b) the capacity of the eggshell to protect against the penetration of non-essential or toxic elements, such as Ag and Cd, known for their contrasting uptake behaviours (Lacoue-Labarthe et al. 2008a).These three trace elements are known to be very toxic to early development stages of marine invertebrate (Calabrese et al., 1974).They are also of specific interest due to their high concentrations in polluted spawning areas of the cuttlefish along the French coasts such as the Seine Bay and the Gironde Estuary (Boutier et al., 2000;Michel et al., 2000;Roux et al, 2001).

Organisms, radiotracer and experimental procedures
Eight adult cuttlefish were collected by net-fishing off the Principality of Monaco in April and May 2008.Male and female cuttlefish were acclimated and maintained in opencircuit tanks in the IAEA-MEL premises.After mating, the fertilized eggs that were laid by each female were immediately separated to optimise their oxygenation.
The eggs (n = 300) were randomly assigned in six 5-L plastic bottles (one bottle per treatment) filled with filtered (0.45 µm) and UV sterilized Mediterranean seawater that was pumped from 30 m depth and adjacent to Monaco Bay.In each experimental bottle (closed system), seawater was constantly aerated.The light/dark cycle was 12h/12h.Eggs were maintained during the full development time in controlled conditions of temperature and pH in a crossed (2 temperature x 3 pH levels) experimental design.Seawater was renewed daily with sterilized and filtered Mediterranean seawater during the first week and then every second day to maintain good water quality.Bottles were changed and cleaned at each seawater renewal to prevent any "bottle" effect due to the development of different biomasses or to the accumulation of detritus such as fragments of external eggshell layers, or bacterial proliferation, which could affect the metabolism of eggs or the bioavailability of chemicals.
Three bottles were kept in a bath that was maintained at 16°C (ambient temperature) and three others in a bath at 19°C (elevated temperature).Temperature was controlled in each bath to within ± 0.5°C using temperature controllers connected to 300 W submersible heaters.Within each temperature condition, one bottle was maintained at ambient pH (8.10) while the two others were maintained at lowered pH (7.85 and 7.60).The values of lowered pH were consistent with those that are the most realistic modelled scenarios of ocean pH to occur by the end of this century: 7.85 (pCO 2 = 900 ppm), 7.60 (pCO 2 = 1400 ppm), as derived from various IPCC models on trajectories of carbon emissions to the year 2100 (Orr et al., 2005).
The pH was controlled in each bottle to within ± 0.05 pH unit with a continuous pH-stat system (IKS, Karlsbad) that bubbled pure CO 2 into the bottles that were continuously aerated with CO 2 -free air.The pH values of the pH-stat system were adjusted every two days from measurements of pH on the total scale.The pH was measured in each bottle using a pH meter (Metrohm, 826 pH mobile) with a glass electrode (Metrohm, electrode plus) calibrated on the total scale using Tris/HCl and 2-aminopyridine/HCl buffer solutions with a salinity of 38 and prepared according to Dickson et al. (2007).Total alkalinity (TA) shifts between two seawater renewals were assessed in a control bottle containing 45 eggs and maintained at ambient pH (ca.8.1) and at a temperature of ca.20°C.TA was measured on seawater samples filtered through 0.45 µm membranes, immediately poisoned with mercuric chloride and stored in a cool dark place pending analyses.TA was determined potentiometrically using a home-made titration system with an Orion 8103SC pH electrode calibrated on the National Bureau of Standards scale and a computer-driven Metrohm 665 Dosimat titrator.TA was calculated using a Gran function applied to pH values ranging from 3.5 to 3.0 as described by Dickson et al. (2007).The pCO 2 was determined from pH and total alkalinity using the R package seacarb (Proye and Gattuso, 2003).
Seawater of each aquarium was spiked with radioactive 110m Ag (1 kBq L -1 ), 109 Cd (1.5 kBq L - 1 ) and 65 Zn (1 kBq L -1 ) to study the bioaccumulation behaviour of the corresponding stable elements that are present in marine waters (e.g., Warnau and Bustamante, 2007).These activities corresponded to an addition of 800, 140 and 1000 × 10 -3 pmol L -1 Ag, Cd and Zn, respectively, to the natural concentrations present in the filtered Mediterranean seawater.
Although the total trace element concentrations in the aquaria were not measured, these additions of metals per spike were one to five orders of magnitude lower than the natural concentrations of metals reported in seawater (Bruland, 1983), which lead to very modest changes in metal concentrations in seawater.Radiotracers were purchased from Amersham, UK ( 110m Ag and 109 Cd) and Isotope Product Laboratory, USA ( 65 Zn): 110m Ag [as 110m AgNO 3 ; T ½ = 250 d, 109 Cd [as 109 CdCl 2 ; T ½ = 464 d] and 65 Zn [as 65 ZnCl 2 ; T ½ = 244 d].Stock solutions were prepared in 1 N nitric acid ( 110m Ag) or in 0.1 N and 0.2 N chloridric acid ( 109 Cd and 65 Zn, respectively) to obtain radioactivities that allowed the use of spikes of only a few microliters (typically 5 µL).
For each treatment, seawater and radiotracer spikes were renewed daily during the first week and then every second day to maintain constant water quality and radiotracer concentrations.Radiotracer activities in seawater were checked (i.e.counted in 150 ml of seawater) before and after each water renewal in order to determine the time-integrated radiotracer activities, i.e. the mean value of all measurements performed over the time period considered (Warnau et al., 1996).At different time intervals, the three radionuclide activities were counted in 3 dissected eggs (γ-emitters could be detected in the same sample according their γ-emissions energy) to determine the radiotracer distribution between the eggshell and vitellus or among eggshell, vitellus, embryo and peri-vitelline fluid, i.e. after 17 and 27 days at 19°C and 16°C, respectively, when the stage of development and size allowed us to both distinguish and separate the egg compartments by dissection.At hatching time, 10 eggs were weighed and 10 newly hatched cuttlefish were counted.

Radioanalyses and data treatment
Radioactivities were measured using a high-resolution γ-spectrometry system consisting of four coaxial Germanium (N-or P-type) detectors (EGNC 33-195-R, Canberra ®    and Eurysis ® ) connected to a multi-channel analyzer and a computer equipped with a spectra analysis software (Interwinner ® 6).The detectors were calibrated with an appropriate standard for each counting geometry used and measurements were corrected for background and physical decay of the radiotracers.Counting times were adapted to obtain relative propagated errors less than 5% (Rodriguez y Baena et al. 2006).They ranged from 10 to 30 min for whole eggs and from 10 min to 24h for the dissected egg compartments.Uptake of 110m Ag, 109 Cd and 65 Zn was expressed as changes in concentration factors (CF) which is the ratio between radiotracer activity in the egg, egg compartment or juvenile -Bq g -1 -and the time-integrated activity in seawater -Bq g -1 (Metian et al., 2009).This unitless term expressed the efficiency of an organism (or a biological compartment) to accumulate and concentrate an element from the seawater after a determined time of exposure.Uptake kinetics were best described by using either a linear equation (Eq.1), a saturation exponential equation (Eq.2), or a exponential equation (Eq.3): CF t = CF ss (1-e -ke t ) (Eq. 2) CF t = CF 0 e -ke t + CF ss (Eq. 3) where CF t and CF ss are the concentration factors at time t (d) and at steady-state, respectively, k e and k u are the biological depuration and uptake rate constants (d -1 ), respectively (Whicker and Schultz, 1982) and CF 0 is a constant.
Uptake of 110m Ag, 109 Cd and 65 Zn in the vitellus and the embryo during the development time were expressed as changes in total activity / concentration ratio (Load concentration ratio; LCR; g; ratio between radiotracer content in the vitellus or the embryo -Bq -and time-integrated activity in seawater -Bq g -1 ) over time (Lacoue-Labarthe et al., 2008a).Although this ratio is a rather unusual way to express metal accumulation, whole radioactivity content in the vitellus and the embryo was preferred over the concentration factor in order to overcome the problem of dramatic weight variations of these egg compartments that tends to mask the actual accumulation of metals in the whole egg.
Constants for the best fitting uptake and depuration kinetic equations (decision based on ANOVA tables for two fitted model objects) as well as their statistics of variability were estimated by iterative adjustment of the models using the nls curve-fitting routine in R freeware.The level of significance for statistical analyses was always set at α = 0.05.

Chemical speciation modelling
To ensure a consistently adequate quality all seawater was carbon-filtered prior to delivery into the tanks that were used during the acclimation period and the experimental exposure.To provide additional information on whether decreasing pH (7.60 and 7.85 from a baseline of 8.10) influenced metal(loid) bioavailability in the tested waters, the speciation of Ag, Cd and Zn was calculated using the HARPHRQ geochemical speciation code (Brown et al., 1991).The input parameters were based on measured physicochemical data (salinity, 38, temperature, 16 or 19 o C, dissolved organic carbon < 1 mg l -1 ; see Jeffree et al., 2006 for more details).

Chemical speciation
The results of the speciation modelling indicated that the decrease in pH (a 3-fold increase in H + ) from 8.10 to 7.60 had only a very minor influence on the speciation of Ag, Cd or Zn.In terms of the free metal ion concentration, generally considered to represent the bioavailable form of Ag, Cd, Zn, the concentration of Ag + increased from 0.6 to 1.0%, Cd 2+ increased from 2.7 to 2.8% and Zn 2+ increased from 46 to 56%, as the pH decreased from 8.10 to 7.60.
Consequently, any observed influence of pH on the accumulation of these three trace elements by cuttlefish eggs can be regarded as being predominantly due to responses of the exposed biological tissues and/or competition with elevated H + at the cell surface binding sites.

Biological results
Decreasing pH resulted in higher egg weight at the end of development at both temperatures (p < 0.05), with maximal values at pH 7.85 (1.60 ± 0.21 g and 1.83 ± 0.12 g at 16°C and 19°C, respectively).Increasing temperature led to an increase of the egg weight but no interactive effect of both pH and temperature was observed (p > 0.05).
Seawater pH had no significant impact on the juvenile weight at hatching time (p = 0.08) for both temperature, but hatchlings were smaller when they developed at 16°C than at 19°C (p < 0.05).
The lower the pH of the incubation seawater of the eggs, the more 110m Ag was accumulated in the tissues of hatchlings.Moreover, this effect was amplified at low temperature, i.e. 110m Ag CF was 2.5 and 1.6 fold higher at pH 7.60 than at normal pH, when eggs developed at 16 and 19°C, respectively.In contrast to Ag, the 109 Cd CF decreased with increasing pCO 2 (p < 0.05), whereas differences in temperature had no effect.Finally, 65 Zn CF showed the maximal values in the juveniles hatched at the intermediate pH 7.85, independent of temperature, and the CF at 7.60 was lower than at pH 8.10.

Uptake kinetics in the eggshell and the embryo
The uptake kinetics of 110m Ag, 109 Cd and 65 Zn in the shell of the eggs that were exposed at the three pH levels during their development are shown in Fig. 3 The 65 Zn uptake kinetics (Fig. 3) increased linearly during the first 20-22 days and 17-20 days at 16°C and 19°C respectively, independent of the rearing conditions.Then, CF slightly decreased until the end of the egg development following a linear equation (Table 2).For both temperatures, 65 Zn accumulation in the eggshell was lower at pH 7.60.
During the period of development, the radiotracer distribution was determined in the internal egg compartments, i.e. the vitellus and the embryo.In Table 3 is reported the 110m Ag, 109 Cd and 65 Zn uptake expressed in terms of metal content (load / concentration ratio; LCR; g) to take into account the vitellus reduction and the embryo growth.The radiotracers' patterns of accumulation were determined for; i), the pooled vitellus and the developing embryo, from the day 1 to 21 and from the day 1 to 14 at 16°C and 19°C and then, ii) the embryonic tissues as soon as the embryo could be separated from the vitellus (> 8 mg; stages 21-22 according to Lemaire, 1970), i.e. at day 27 and 17 at 16°C and 19°C, respectively.A significant accumulation (p < 0.05) of 65 Zn and 110m Ag in the pooled vitellus and embryo were determined at 6 and 4 days earlier (day 15 vs. 21 and 10 vs. 14 at 16°C and 19°C; Table 3) for pH 7.60 than at the other pH values, suggesting that the low pH induced an earlier change in the eggshell permeability.During the whole period of embryonic growth, 110m Ag and 65 Zn were more efficiently accumulated at pH 7.60 and 7.85, respectively, as also shown above for the hatchlings.This result implies that seawater pH had an impact on the metal accumulation capacities of the embryo during its total development.Finally, the 109 Cd distribution revealed that this element was only significantly taken up from the seawater during the last week of development, at both temperatures.

Radiotracers distribution during the egg development
The distribution of the different radiotracers between the eggshell, the perivitelline fluid, the vitellus and the embryo was also determined in this experiment (Table 4).The perivitelline fluid could be considered as an intermediate compartment between the seawater and the embryo.We therefore calculated the CF between seawater and the perivitelline fluid and between the perivitelline fluid and the embryo for Ag, Cd and Zn on the last developmental day, i.e. day 63 and 42 at 16°C and 19°C, respectively (Table 4).The results highlighted that;

DISCUSSION
In this study, one of the major results observed was that lowered pH had increased the egg weight by the end of the development, with an increase in the perivitelline fluid volume (results not shown).This suggests that the seawater pCO 2 disturbed the swelling process that occurs during the last two thirds of the whole developmental period.The water intake occurred progressively with the organogenesis and enhanced the perivitelline space that the embryo requires for its growth.The mechanistic understanding of this phenomenon is not well known in cephalopod eggs, although it was observed that the water follows an osmotic gradient that is maintained by the embryo himself (Boletzky, 1986;De Leersnyder et al., 1972).It has also been suggested that the oviducal substances from the eggshell play a key role in egg swelling, and that organic compounds cross the chorion and consequently increase the osmotic pressure of the perivitelline fluid (Gomi et al., 1986;Ikeda et al. 1993).Hence there are two possible explanations for our experimental results, viz: i) lowered pH disturbs the maintenance of the osmotic gradient in the perivitelline fluid by the embryo, and/or, ii) the components of the eggshell and its permeability were affected by the seawater pCO 2 .
For eggs reared at two different temperatures, the observed course of embryonic development was fully consistent with the previous observations reported for the common cuttlefish (Boletzky, 1986); this tends to confirm that the temperature effect was homogeneous between the three pH conditions at 16 and 19°C.It is also worth noting that reduced temperature, i.e.
16°C, decreased egg swelling compared to 19°C.Although the development time was 20 days longer at 16°C than at 19°C, the incorporation of water was still limited.This effect has also been confirmed in a subsequent experiment (unpublished data).It is known that temperature influences metabolic rate (Melzner et al. 2006), and as the temperature decreases metabolism would slow down.This suggests that the egg swelling observed depended on the embryonic metabolic level and the subsequent capacity of the embryo to maintain the osmotic gradient in the perivitelline fluid.However, our results revealed also that the egg weight increased with acidified conditions.It is also noteworthy that increasing pCO 2 in seawater leads to metabolic depression in marine organisms due to changes in their acid-base balance (e.g.Pörtner et al., 2004;Pörtner, 2008).Indeed, due to the high Bohr coefficient of their hemocyanins, cephalopods showed reduced aerobic scope under elevated pCO 2 in seawater and also showed a high sensitivity to hypercapnia (Pörtner et al., 2004;Melzner et al., 2007).It follows that the reduced metabolic rates (Rosa et al., 2009) of eggs under acidified conditions limited embryonic growth.Additionally, a recent study has demonstrated that at the end of the egg development, the cuttlefish embryo was surrounded by 10-fold higher pCO 2 values in the perivitelline fluid (i.e., ≈ pH 7.4) than those in seawater because of the embryo's respiration (Gutowska and Melzner, 2009).This result highlights that the embryo naturally experiences hypercapnia and still develops normally under such conditions.In the study reported here, the reduced size of juveniles exposed at pH 7.60 suggests that the pCO 2 of the perivitelline fluid could reach a threshold value above which the embryo does not develop normally.
Consequently, it could be valuable to determine the pH/pCO 2 levels achieved in the perivitelline fluid when eggs develop under increasing pCO 2 conditions and assess their impact on the embryo's acid-base regulation and metabolism.
During the embryonic development of cuttlefish, the greatest amounts of the metals, such as Ag, 241 Am, Cd, Co, Hg, Mn, Pb and Zn, remain associated with the eggshell (Bustamante et al., 2002(Bustamante et al., , 2004(Bustamante et al., , 2006;;Lacoue-Labarthe et al 2008a, 2009a, 2009b).Indeed, the eggshell contains a high proportion of mucin proteins that also have a high content of sulfydril-groups (Boletzky, 1986) for which Ag, Cd and Zn have a high affinity (e.g., Wedemeyer, 1968;Temara et al., 1997;Bell and Kramer, 1999).As previously described (Lacoue-Labarthe et al., 2008a), Ag and Zn accumulated linearly during the first two weeks of development, suggesting that the binding sites were not saturated during this period, in contrast to Cd.
During prolonged exposure to metals (> 20 days), the accumulation of 110m Ag, 109 Cd and 65 Zn decreased while the eggs were under exposure conditions, consistent with changes in the binding properties of the eggshell.This shift occurred at similar times (17 and 20 days at 19°C and 16°C, respectively) for both temperatures and consequently at different developmental stages, suggesting that the polymerization of the eggshell mucopolysaccharides due to seawater pH (Boletzky, 1986(Boletzky, , 1998) ) was the main factor influencing the metal bioaccumulation in the eggshell.In this way, the low pH (7.60) could affect eggshell polymerization and reduce the metals binding sites, as shown for the 110m Ag and 65 Cd (Figure 3), possibly through the competitive inhibition with H + ions.This suggests that the protective role of the eggshell in hindering the incorporation of these metals into the embryo could be adversely affected.
Concerning Cd interaction with the eggshell, two characteristics were noteworthy: firstly, Cd uptake reached a steady-state equilibrium and shifted after only 7 days at pH 7.60, strongly suggesting that the mechanisms involved in the Cd accumulation were different from the other metals.Moreover, the combined effect of pH and temperature on the maximal CF values was surprising considering that both these factors affected the chemical properties of the eggshell.Therefore, it could be proposed that the metal uptake process could be driven by the accumulation capacity of the symbiotic bacteria embedded in the eggshell nidamental layers (Bloodgood et al., 1977;Barbieri et al., 1996).As the microorganism respiration influences the oxygen diffusion through the eggshell (Cronin and Seymour, 2000), their metabolism could also affect the accumulation and the retention of metals in this egg compartment.However, to the best of our knowledge, no study has assessed the effect of temperature and pH on the metabolism of these bacteria and the consequences for their population levels in the eggshell.Finally, the metal accumulation among the internal compartments (Table 3) revealed that the permeability of the eggshell seemed to be influenced by the pH with 110m Ag and 65 Zn penetrating earlier into the pooled vitellus and embryo incubated at low pH than in those reared at normal pH.All these results highlight the role of seawater pH in reducing the shielding properties of the eggshell against the accumulation of dissolved metals.
Regarding the 110m Ag, 109 Cd and 65 Zn activities in the hatchlings, it appeared that 110m Ag and 109 Cd uptake showed a linear relationship with the increasing pH, whereas 65 Zn was optimally accumulated in the embryo at the intermediate pH.110m Ag was efficiently accumulated in the cuttlefish embryo, as previously demonstrated (Bustamante et al. 2004, Lacoue-Labarthe et al., 2008a), presumably from the time when the water permeability of the eggshell changed and the perivitelline fluid started to increase in volume.A few hours before hatching, the higher 110m Ag CF values recorded in the perivitelline fluid compared to the other elements highlighted the capacity of Ag to concentrate in the perivitelline space.Consistently, in the hypertonic perivitelline fluid, the monovalent ions such as Cl -, Na + and K + are slightly more concentrated than divalent ions such as Ca 2+ and Mg 2+ (De Leersnyder et Lemaire, 1972).
Moreover, Ag could be bound to the large molecules dissolved in the perivitelline fluid, such as the natural tranquilizer peptides (Weischer and Marthy, 1983) or the organic matter accumulated from the oviducal jelly (Gomi et al., 1986;Boletzky, 1986).It is noteworthy that Ag was more efficiently taken up with decreasing seawater pH in the juveniles.This could be explained by; i) a higher metal translocation from the eggshell to the embryo (Lacoue-Labarthe et al., 2008a) being linked with the reduced Ag retention capacity of the eggshell at lower pH and by ii) greater transfer of Ag from the perivitelline fluid to the embryo under acidified conditions in seawater.This latter mechanism may arise for two reasons: a) as mentioned above, increasing seawater pCO 2 could disturb the low pH/high pCO 2 conditions in the perivitelline fluid.This may subsequently modify the chemical speciation of the metal in the embryo's surrounding medium, thus enhancing the Ag free ionic forms which are more bioavailable; and/or b) increasing Ag uptake in the embryo could reflect disturbances of ionic regulation (Wood et al., 1999) which is highly challenged by the acid-base balance (e.g.Pörtner et al., 2004).Therefore it seems that the embryo metabolic rate controls the Ag uptake processes into its tissues.This was further confirmed by the fact that low temperature limited the Ag uptake in the hatchlings at normal pH as it decreased the respiration rate (Wolf et al., 1985).Finally, it was noteworthy that high Ag CF was correlated with a low level of egg swelling and small-sized hatchlings at the end of development at low pH.Are both these observations the results of the metabolic disturbance under acidified condition, or are these morphological impacts the first consequences of the presumably toxicity of the highly accumulated Ag?
Regarding 109 Cd and 65 Zn, the lower CF PVF/sw determined at the end of development suggested that both metal concentrations in the perivitelline fluid were close to the equilibrium with those in seawater (CF ≈ 1-4).Cadmium passed through the eggshell during the last days of development and accumulated in the embryonic tissues during the last developmental stages (Lacoue-Labarthe et al., 2008a).Considering that Cd mimics Ca (Bustamante et al. 2002;Bridges and Zalups, 2005), the decreasing Cd accumulation with increasing pCO 2 whatever the temperature was, could reflect a possible decreasing calcification rate of the embryo under hypercapnic conditions (e.g.Gazeau et al., 2007).However, it has been recently demonstrated that the cuttlebone calcification was enhanced in sub-adult cuttlefish reared at 6000 ppm CO 2 (Gutowska et al., 2008).Further studies are now warranted to determine the impact of acidified conditions on the calcification of the cuttlebone during the embryonic development.
Zn is an essential element required for the synthesis of numerous cell constituents such as proteins and enzymes (e.g., Vallee and Auld, 1990).It has been demonstrated that it is maternally transferred (Lacoue-Labarthe et al., 2008b) by incorporation in the vitellus during oogenesis and that dissolved Zn in seawater accumulated in the embryo during egg development (Bustamante et al., 2002;Lacoue-Labarthe et al., 2009b).Then, after hatching, young cuttlefish continue to bioaccumulate Zn very efficiently both from both seawater and food (Bustamante et al., 2002;Miramand et al., 2006).These facts suggest that during the embryonic development, the high embryonic requirements for Zn are not fully covered by the maternal pool.In this study, temperature had no effect on the recorded 65 Zn CF in the hatchlings, implying that a longer exposure time at 16°C did not lead to the higher metal accumulation and therefore that the Zn content in the embryo may be regulated as a function of the metabolic rate according to the developmental stages.Then, 65 Zn activities in the hatchlings and the embryos clearly showed that the metal accumulation was higher at pH 7.85 during the full developmental period and that this higher uptake was associated with a greater rate of growth of both egg and embryo.These findings give rise to the following two hypotheses: 1) the metabolic performances of the embryo increased at pH 7.85 enhancing the protein synthesis and subsequently the requirements for Zn, and/or 2) the chemical speciation of Zn in the perivitelline fluid enhanced the bioavailable ionic species for the embryo, consequently stimulating the metabolism and growth of the embryo.
In summary, this first study showed the strong and contrasting effects of pH and temperature on the bioaccumulation of several metals in the cuttlefish eggs.In the context of the ocean acidification, it appears that decreasing pH until 7.85 should lead to some possibly beneficial effects, such as a larger egg and presumably hatchling size and a better incorporation of the essential element such as Zn in the embryonic tissue.This may improve the survival the newly hatched juveniles.Moreover, the incorporation of a toxic metal such as Cd (Lacoue-Labarthe et al., submitted) in the embryonic tissue was reduced with increasing pCO 2 whereas the accumulation of Ag was strongly enhanced under acidified conditions.According to these first results, further work is now warranted to further assess the ecotoxicological consequences of combined global change effects with a greater range of anthropogenic coastal pollutants on cuttlefish egg development and the recruitment success of juveniles into their populations.
i) Ag was efficiently concentrated in the perivitelline fluid compared to the other metals (CF Ag > 100 >> CF Zn ≈ 3 > CF Cd < 2) and ii) that 110m Ag CF in the perivitelline fluid did not vary with the pH for either temperature, except at normal pH compared to the lower pHs in the 19°C-incubated group.In this experimental condition, CF values for 110m Ag, 109 Cd and 65 Zn were affected by the fact that most of the eggs sampled at this time were already hatched, leading to the loss of at least some of their perivitelline content.The 110m Ag was more effectively taken up from the perivitelline fluid with decreasing pH, with the highest CF values being attained in the embryo reared under acidified conditions.No significant effect of pH was observed on the 109 Cd CF in the perivitelline fluid whereas the Cd accumulation from the perivitelline fluid to the embryo increased with decreasing pH, at 16°C.Concerning 65 Zn, the CF PVF/sw perivitelline fluid decreased with decreasing pH, whereas the CF emb/PVF were maximal at pH 7.85 leading to the highest Zn accumulation in the hatchlings as described above.

Table 2 )
. For the eggs reared at 19°C, , reaching the lowest CF values 20 and 10 days before the time of hatching, at 16°C and 19°C, respectively.At pH 7.60, the pattern of 109 Cd accumulation changed after only 7 days of development, at both temperatures.The pH and the temperature showed a combined effect on the maximal 109 Cd CF values in the eggshell, with 109 Cd CF 7.85 > CF 8.10 > CF 7.60 and CF 8.10 > CF 7.85> CF 7.60 , at 16°C and 19°C, respectively; i.e. 930 ± 150 and 265 ± 50, respectively.
110mAg uptake in their eggshell displayed a linear pattern during the first 17 days with the uptake rate at normal pH being higher than at the lower pH values, i.e. 102 ± 3 vs.81 ± 3 and 77 ± 2 d -1 at pH 8.10 vs. 7.85 and 7.60, respectively.Following the first 17 days of development, the CFs decreased according to a single exponential equation indicating that the tracer no longer accumulated in the eggshell, but was only depurated from it.Finally, CF values reached at the end of development were 2-and 4-fold lower at lower pHs than at normal pH, i.e. 557 ± 97 and 317 ± 30 vs. 1258 ± 212 at pH 7.85 and 7.60 vs. 8.10, respectively.These results suggest that low pH limited the Ag retention in the eggshell.

Table 1 .
1. Sepia officinalis.Weight (g) (A) of the eggs at the end of development (n = 10) and Fig. 2. Sepia officinalis.Concentration factors of 110m Ag, 109 Cd and 65 Zn (CF; n = 10), in the newly hatched juvenile exposed at three different pH -pH 8.10, pH 7.85, pH 7.60 -for two temperatures, i.e. 16°C (grey) and 19°C (white).Results of the statistical analysis were reported on the Table1.

Table 1 .
Sepia officinalis.Two-way ANOVA parameters testing the effects of three pH(7.60,7.85 and 8.10) and two temperatures (16 and 19°C)on the weight of the eggs and hatchlings, and on the concentration factor (CF) of 110m Ag, 109 Cd and 65 Zn in the hatchlings at the end of the embryonic development (see Figures1 and 2).

Table 2 .
Sepia officinalis.Parameters of 110m Ag, 109 Cd and 65 Zn uptake kinetics in the eggshell of cuttlefish eggs exposed for the whole development time to radiotracers dissolved in seawater (CF; mean ± SD, n = 3) in three pCO 2 level treatments and two different temperatures (see Figure3):

Table 3 .
Sepia officinalis.Load concentration ratios (LCR; g; mean ± SD, n = 3) of 110m Ag, 109 Cd and 65 Zn, at different developmental time, in the pooled vitellus and embryo and in the separated embryo of eggs exposed to dissolved radiotracers in three pCO 2 level treatments and two different temperatures.

Table 4 .
Sepia officinalis.Uptake of 110m Ag, 109 Cd and 65 Zn expressed as CF in between the peri-vitelline fluid (PVF) and seawater and between PVF and the embryo at the end of development following three different pCO 2 levels at two different temperatures.