Introduction

Designed to protect embryonic stages from predation, egg capsules also can provide severe physiological challenges to their inhabitants, as the egg wall represents a barrier to diffusion of gases. Previous work has demonstrated that oxygen diffusion coefficients \( ( {\text{K}}_{{O_{2} }} ) \) of marine animal egg capsules are typically 10–20% that of pure water (e.g., Brante 2006). In molluscs (as in all other developing embryos), oxygen consumption rates rise dramatically during development (e.g., Cronin and Seymour 2000; Brante 2006). Thus, in order to enable rising oxygen fluxes by means of diffusion, many molluscan eggs swell during development, leading to enhanced surface areas and reduced egg wall thicknesses (e.g., Kress 1972; Cronin and Seymour 2000), and consequently increased oxygen conductances (Seymour 1994). In addition, embryos inhabiting fluid filled capsules often produce convective currents that prevent the formation of pO2 gradients within the egg fluid (amphibians: Burggren 1985, fish: Rombough 1988, molluscs: Cronin and Seymour 2000). While thinning of egg capsule walls has been shown to enable consistently high pO2 values in eggs of the marine gastropod Fusitriton oregonensis (ca. 12 kPa; Brante 2006), the combination of egg swelling and convection does not prevent pO2 from consistently falling during embryonic development of the cephalopod Sepia apama, from initial values of 14 kPa down to 5–6 kPa close to hatching (Cronin and Seymour 2000). It has been hypothesized that this might eventually trigger hatching once critical pO2 values (pO2crit) are reached within the egg (pO2crit = 5–8 kPa for S. apama and S. officinalis late embryonic stages (DeWachter et al. 1988; Cronin and Seymour 2000). While low pO2 values in eggs also have been proposed to stimulate hatching in vertebrates (reptiles and birds: Vleck and Hoyt 1991, amphibians: Seymour and Bradford 1995), little attention has been devoted to the flip side of the coin: egg pCO2 and pH, especially in marine animals. In order to excrete metabolically produced CO2 at a rate that almost equals that of oxygen consumed (of course, depending on the respective respiratory quotient), a pCO2 gradient has to be maintained in the opposite direction, with high values inside the egg. To our best knowledge, there are currently no published egg fluid pCO2 values available in the literature for any marine animal. This is quite surprising, as high pCO2 values most likely go along with low pH values and potentially constitute another stressor that may significantly affect embryonic physiology. Using late embryos of the cephalopod Sepia officinalis as a model system, we aimed at characterizing in more detail the abiotic conditions within the perivitelline fluid (PVF), with an emphasis on the pCO2/pH gradient between the egg and the environment.

Materials and methods

Laboratory laid Sepia officinalis eggs were obtained from the Biological Station in Luc-sur-Mer, Université de Caen (Normandy, France) and transported to Germany at the age of 7–8 days. They were then incubated in a recirculating aquarium system (200 l volume) at the AWI Bremerhaven at 17°C (±0.2°C) for approximately four weeks. A nitrification filter (Eheim Professional 2, Eheim, Deizisau, Germany), a protein skimmer (AquaCare 2000, AquaCare, Germany) and a 36 Watt UV sterilizer (Rebie, Bielefeld, Germany), in combination with frequent water changes, aided in maintaining a high water quality within the system (pH > 8.1, pCO2 < 0.042 kPa, pO2 > 20 kPa, [NH4 +] < 0.1 mg/l, [NO2 2−] < 0.1 mg/l, [NO3 −] < 5 mg/l, S > 32 ppt). pH, pO2 and salinity were checked daily using a WTW multimeter (WTW, Weilheim, Germany), nitrogenous waste products were assessed bi-weekly using photometric test kits (Merck, Darmstadt, Germany). Carbonate system parameters were calculated from pHNBS and weekly determinations of total dissolved inorganic carbon (CT) (see below).

Eggs were placed individually on the bottom of the incubation tank until close to hatching (stages 29–30). All (n = 13) eggs of the present study were sampled on the same day. Eggs were gently lifted out of the tank and immediately sampled for PVF. All PVF samples were taken within 15 s, thus minimizing the chance of artificially increased pCO2 values caused by disturbed embryos. To measure pH and pO2, a 1 ml plastic syringe was equipped with miniature fiber optic sensors (optodes, tip diameter 140 μm, Presens GmbH, Regensburg, Germany) and filled with 200–300 μl PVF, (previously described in Melzner 2005). Stable pH values were obtained after 10 min, pO2 values after 10 s. During the measurement period, the syringe and sensors were placed in a thermostatted water bath at 17°C. The oxygen optodes were calibrated according to the manufacturer’s instructions with water vapor saturated air and a Na2SO3 solution. The pH optodes were calibrated using five seawater standards (North Sea seawater, 31 psu, 0.2 μm filtered) adjusted to pH values between 7 and 8 with 1 M HCl. A pH electrode (WTW sentix81 and pH340i pH meter, WTW, Weilheim, Germany), calibrated with Radiometer precision buffers, was used to prepare the seawater buffers. Calibration of the pH optodes with seawater buffers was found necessary as these sensors are sensitive to the ionic strength of the measurement medium.

Another 350 μl of PVF was sampled with a gas-tight glas syringe for the determination of total dissolved inorganic carbon (CT). CT was measured in triplicate (100 μl each) using a gas chromatographic method (Lenfant and Aucutt 1966, modified after Pörtner et al. 1990) on an Agilent 6890 N gas chromatograph. Carbonate system speciation (i.e., pCO2, [HCO3 −]) was calculated from CT and pHNBS using CO2SYS software (Lewis and Wallace 1998), with dissociation constants from Mehrbach et al. (1973) as refitted by Dickson and Millero (1987).

Following PVF sampling, eggs were dissected and embryo and yolk wet mass was determined using a Sartorius precision balance (Sartorius, Göttingen, Germany). Embryonic stages were determined according to Lemaire (1970).

PVF pO2, pCO2, [HCO3 −] and pH were graphed against embryo wet mass. Subsequent regression analyses were performed using GraphPad lnStat 3.0 software (GraphPad Software, San Diego, USA). A Runs test was used to test for deviations from linearity, an ANOVA was used to assess whether slopes differed from zero.

Results and discussion

The eggs of S. officinalis are particularly suited for the investigations of abiotic parameters in the PVF due to their large size. At hatching, the eggs contain >1 ml of PVF and have a diameter between 1.5 and 2 cm (see supplementary movie, S1). Schematic drawings of a freshly laid egg, as well as of late stage embryos, are nicely illustrated in Wolf et al. (1985) (Figs. 1 and 2). All of the cuttlefish eggs we investigated contained living embryos, ranging in wet mass between 134 and 310 mg (mean 238 mg, SD 44 mg). Organogenesis was completed in all embryos (stages 29–30, Lemaire 1970), and embryos had mostly absorbed their external yolk (mean wet mass of external yolk 18 mg, SD 20 mg). The larger embryos (>250 mg) were probably <1 week away from hatching, especially the two largest embryos, who had consumed all external yolk reserves (Lemaire 1970). Measured PVF pO2, pH and pCO2 values were tightly linked to embryo wet mass, increasing (pCO2) or decreasing (pO2, pH) in a linear fashion (Fig. 1a, b, c). The results of all regression analyses are depicted in Table 1.

Fig. 1
figure 1

pO2 (a), pCO2 (b), pH (c) and [HCO3−] in perivitelline fluid (PVF) of Sepia officinalis eggs (stages 29–30), displayed against embryo wet mass (excluding the external yolk sac). pCO2 and [HCO3−] were calculated from PVF CT and pH. See Table 1 for regression analyses and equations

Full size image
Table 1 Linear regression analyses. N = 13 eggs were analyzed, pCO2 and [HCO3 −] are derived parameters (calculated from pH and CT, see text), pCO2 and pO2 in kPa
Full size table

PVF Oxygen partial pressures

Sepia officinalis PVF oxygen partial pressures varied between 12.8 kPa (ca. 61% air saturation) and 4.6 kPa (ca. 22% air saturation). Wolf et al. (1985) found S. officinalis PVF pO2 values to rise with embryonic development, increasing from 12 kPa (stage 24) to 15.5 kPa (stage 29). They concluded that improved diffusion properties of the swelling and thinning egg case would aid in the maintenance of high PVF pO2. Unfortunately, Wolf et al. did not provide embryo masses, making it difficult to draw meaningful conclusions between our studies. However, the smallest of our embryos was also classified as stage 29 and its PVF pO2 value was high (>12 kPa, see Fig. 1a). It is thus not unlikely that Wolf et al. missed the final growth phase in their study, as this is when PVF pO2 rapidly declines. Support comes from the study by Cronin and Seymour (2000), who also demonstrated an inverse correlation between embryo mass and pO2 in Australian giant cuttlefish eggs (Sepia apama, 12°C). In S. apama, pO2 declined to values of about 6 kPa during the last 50 days of development. As Cronin and Seymour determined a pO2crit of 8 kPa for hatchling animals, they suggested that late embryos probably experience diffusion limitation close to hatching. pO2crit for 200–300 mg S. officinalis incubated at 18–19°C has been shown to be in the 4.5–5 kPa range (DeWachter et al. 1988), thus in the range of our lowest PVF values (4.6 kPa). These findings support the idea of oxygen diffusion limitation being one critical factor in S. officinalis late embryonic development.

PVF pCO2 and pH

As expected, PVF pCO2 values rose with increasing embryo mass, from 0.13 kPa in the smallest up to 0.41 kPa in the largest embryos (Fig 1b). Thus, cuttlefish embryos are surrounded by tenfold higher pCO2 values than those of ambient sea water (ca. 0.04 kPa) at the end of their embryonic development. High pCO2 values in the PVF also imply that blood pCO2 values must be even higher in order to maintain CO2 excretion rates across the gill/skin epithelia by means of diffusion. Typically, pCO2 values in extracellular fluids of high-power animals such as fish or cephalopods are 0.2–0.4 kPa above those of the ambient seawater (e.g., Heisler 1986; Johansen et al. 1982; Pörtner et al. 1991). Therefore, cuttlefish embryos are probably exposed to blood pCO2 values of 0.6–0.8 kPa at the end of their development. It would be quite rewarding to study blood pH regulation in embryos under such conditions, as cephalopods are known for the high pH sensitivity of their extracellular respiratory pigment hemocyanin (Melzner et al. 2007). The occurrence of special embryonic hemocyanins (Decleir et al. 1971) may be one adaptation to the high pCO2 values encountered during late embryogenesis.

The CO2 gradient between the PVF and seawater was much shallower than that of oxygen, almost by a factor of 70 (ΔpO2/ΔpCO2 = 66.6, SD 11.3). This ratio is higher than expected from Henry’s and Graham’s laws, which suggest an approximately 26-fold higher partial pressure gradient of O2 versus CO2 across the egg envelope (for seawater at 17°C, Dejours 1975). These results can only be explained if either a significant portion of the excretory CO2 is retained within the embryo to aid in the formation of the internal CaCO3 shell, or if Krogh’s diffusion coefficient for CO2 \( ( {\text{K}}_{{CO_{2} }} ) \) is much higher than \( {\text{K}}_{{O_{2} }} \) for S. officinalis egg capsules. It has been demonstrated for sea urchin larvae and corals that significant portions of inorganic carbon for calcification are derived from metabolic CO2 (>50%, Sikes et al. 1981; Furla et al. 2000). However, a more extensive follow-up study that will focus on the determination of oxygen consumption and CO2 excretion, as well as egg capsule surface area and thickness, will enable us to give estimates for both, \( {\text{K}}_{{O_{2} }} \) and \( {\text{K}}_{{CO_{2} }}. \)

As pCO2 increased, pH strongly decreased, from 7.72 to 7.23 (Fig. 1c). Again, there are no pH measurements for marine animal eggs at present that our data could be compared with. In freshwater fish (salmonids), 0.3–1.0 unit lower pH values have been recorded in egg PVF at ambient pH between 7 and 8 (Kugel and Peterson 1989). No accumulation of HCO3 − was visible in S. officinalis PVF in order to actively buffer pH (Fig. 1d), a mechanism that is used by many marine organisms to compensate for extra- and intracellular acidification (e.g., Heisler 1986; Cameron 1986). [HCO3 −] fluctuated around 2.39 mM (SD 0.08 mM) in all eggs.

It is noteworthy that cuttlefish embryos are able to form an internal CaCO3 (aragonite) shell under the low pH and high pCO2 prevailing in their egg environment. We recently demonstrated, that calcification rate in juvenile cuttlefish is not impaired at external pCO2 values of 0.4 and 0.6 kPa (Gutowska et al. 2008). This sets S. officinalis apart from other marine invertebrates studied so far, as the majority show a decrease in calcification under comparable conditions (see Fabry et al. 2008 for a review). It is tempting to propose that this capacity is causally linked to an embryo that is already adapted to cope with relatively high pCO2/low pH values.

Perspectives

As mentioned, no data on egg fluid pCO2 and pH is available for other marine organisms; however, judging from the pO2 versus pCO2 ratios obtained in our study, in comparison to pO2 values in- and around eggs or egg masses of other marine animal taxa, it seems likely that many embryos will also be surrounded by fluids of high pCO2 and low pH: For example, Diez and Davenport (1987) showed that pO2 values in shark eggs decrease from ca. 18 kPa in early embryos to ca. 10 kPa in late embryos, Fernandez et al. (2000, 2002) measured water pO2 values <5 kPa in decapod crustacean egg masses, Cohen and Strathmann (1996) found pO2 values of less than 6 kPa in egg masses of opisthobranch gastropods and those of a polychaete worm.

Delayed development of embryos in central positions of egg masses has usually been causally linked to reduced metabolic rates due to low ambient pO2 (e.g., Chaffee and Strathmann 1984). However, high pCO2/low pH may be important factors as well, as they could also elicit reductions in respiration rates: Metabolic depression, in combination with reduced growth and calcification performance, has recently been observed in mussels (Mytilus galloprovincialis) exposed to sea water of similar pCO2 (ca. 0.5 kPa, Michaelidis et al. 2005).

Quite clearly, PVF pCO2 and pH represent important abiotic factors that might influence the physiological performance of marine animal embryos to a large degree. To date, these factors have been thoroughly neglected. We hope that this study sparks some interest in studying CO2 excretion and pH homeostasis in eggs of marine animals.