ABSTRACT
The fluxes of 3H2O, 22Na and 36Cl were simultaneously measured in yolk-sac larvae of cod (Gadus morhua L.) in 34 %o sea water at 4·5 °C.
The rates of turnover of all three isotopes were higher than in adult fish. Diffusional permeability coefficients, which relate ion fluxes to surface area, were however lower, indicating that larvae are less permeable than adults. Furthermore, there is close agreement between the diffusional and osmotic permeability coefficients, which supports a previous hypothesis that relatively low drinking rates in marine fish larvae are a consequence of low integumental permeability.
Estimates of the sodium and chloride concentrations derived from the equilibrium levels of 22Na, 36Cl and 3H2O indicate that yolk-sac larvae of cod regulate their body fluids hypotonic to sea water. Also, the ionic concentrations of the tissues of yolk-sac cod larvae are similar to those of adults.
Introduction
Tytler & Blaxter (1988a,b) and Mangor-Jensen & Adoff (1987) have shown that the larvae of marine teleost fish drink as part of their osmoregulation process. The rates of drinking of the larvae of cod (Gadus morhua), herring (Clupea harengus) and plaice (Pleuronectes platessa), when related to body mass, were found to be higher than for adult fish (e.g. Maetz & Skadhauge, 1968; Isaia, 1972) but substantially lower than expected from the surface to mass ratios. Tytler & Blaxter (1988b) have suggested that these differences could be explained if larvae were less permeable than adults. In fact, very low water permeability has been measured in marine fish eggs by Potts & Eddy (1973) and Mangor-Jensen (1986). The main barrier to diffusion in eggs is thought to be the vitelline membrane (Loeffler & Lovtrup, 1970). However, this structure no longer forms the integument in larvae and is therefore unlikely to offer the same protection against osmotic stress owing to unfavourable surface area to mass ratios.
The purpose of this work is, therefore, to clarify the situation by measuring the diffusional permeability coefficients of water, sodium and chloride ions in the larvae of cod, using radioisotopes, and to compare them with osmotic permeability coefficients derived from earlier measurements of drinking rates.
Materials and methods
The larvae were hatched from spring-spawned eggs of cod (Gadus morhua L.) from the west coast of Scotland. On hatching, the larvae were transferred to 2 –1 glass beakers containing 34 %o sea water (SW), which had been passed through 0 ·22 μm micropore filters and held at 4 ·5 °C in a refrigerator for up to 21 days before use.
Determination of fluxes of water, and sodium and chloride ions
3H2O, 22Na and 36Cl were measured simultaneously in 9-day-old yolk-sac stage larvae in SW at 4·5°C. 210 larvae were placed in 40 ml of filtered SW. 3H2O (185MBqml−1), 22Na (3·7–37 GBq mg−1 Na+, carrier-free) and 36Cl (110MBqg−1 Cl−), obtained from Amersham International pic, England, were added to produce a working solution containing approximately 1·5×107-3·0 × 107 counts min−1 3H2O,0·5 × 107 counts min−1 22Na and 0·6× 107 counts min−1 36C1 per ml. 22Na was assayed using a Packard Autogamma 500C instrument. A Canberra Packard 2000 CA scintillation counter was then used to determine 3H and 36Cl, after subtracting the contribution from 22Na. To measure the time course of influx of the isotopes, the larvae were sampled after 15, 30, 45, 60, 135, 150 and 330 min of exposure to the working solution. At the end of each interval, 15 larvae were removed with a wide-mouthed, fire-polished, bulb pipette into a tea strainer and then passed through three wash baths containing 400 ml of SW over a period of 6 min. The remaining larvae were transferred into 400 ml of SW to measure the effluxes. After 15, 30, 60, 90, 120 and 180min the larvae were similarly sampled. The larvae were individually removed from the final wash solutions with fine forceps and dabbed dry on a Kleenex medical wipe at the end of both influx and efflux experiments. Three subsamples, each consisting of five larvae, were then placed in 0·5 ml of Soluene (Packard) in a 7 ml plastic vial and solubilized for 15 h overnight at room temperature. Scintillation fluid (Hionic Fluor, Packard, 5 ml) was then added and the radioactivity assayed.
Determination of rate constants
Determination of permeability coefficients
Determination of water content and the concentrations of sodium and chloride
The water, sodium and chloride content of cod yolk-sac larvae were estimated from the asymptotic levels obtained by extrapolating from the exponential functions for each component. Thus, the tritiated water equilibrium level in 9-day-old cod larvae (Qeq = 5654 counts min−1, Fig. 1) obtained after exposure to a working solution with a specific activity of external bath of 17441 counts min−1 translates into a water content of 0·32 ± 0·03 mg (efflux data). The concentrations of sodium and chloride were estimated from their respective equilibrium levels and the above estimate of the water content.
The wet mass was obtained by weighing a pooled sample of 10 larvae.
Results and discussion
The 3H2O, 22Na and 36Cl influxes in 9-day-old yolk-sac larvae of cod are shown in Figs 1A, 2A and 3A. The lines describing the time courses of uptake of radioactivity are regression lines derived from equation 1.
The time courses of efflux of the three isotopes are also described by regression lines in Figs 1B, 2B and 3B which are based on equation 2. The rate constants for influx and efflux, K1 and K2, respectively, and the rates of turnover in the first hour (k), calculated using equation 3, of the isotopes are presented in Table 1.
There is good agreement between and K2 rate constants, except in the case of 3H2O where the rate constant of influx (K1 = 0·79 h−1) appears to be higher than that for efflux (K2= 0·61 h−1), but the difference was found not to be significant (t-test). The rate constants k and K2 for efflux for 22Na (0·51 and 0·45 h−1, respectively) and 36C1 (0·29 and 0·33 h−1, respectively) are also very similar, but k for 3H2O was significantly lower than K2. Rate constants for ions fluxes in adult fish have been found to be highly temperature-dependent, decreasing with decreasing temperature with Q10 of between 2·1 and 2·5 (Motais & Isaia, 1972). Since diffusional processes vary with absolute temperature, these flux changes are attributed to temperature-dependent changes in branchial blood flow. In 9-day-old yolk-sac cod larvae the gills are not yet formed and diffusion exchange is cutaneous. Rate constants in adults are also species-specific, but generally they tend to be lower than those for cod larvae (Table 1). Isaia (1972) found influx rate constants (K1) for 3H2O and 22Na to be 0·24 and 0·22 h−1, respectively, for Serranus scriba adapted to SW at 15 °C. A slightly higher 3H2O influx rate constant of 0·35 h−1 was measured by Motais & Isaia (1972) for SW-adapted eels (Anguilla anguilla) at 15 °C. In cod larvae the equivalent K1 influx rate constants for 3H2O and 22Na were 0·79 and 0·45 h− 1, respectively, which are not only higher than for adults but K1 for water is higher than that for sodium. The rate constants for 36Cl in cod larvae are lower than that for 22Na. Evans (1967) also found that the Kr influx rate constant for 36C1 was 65 % of that for 22Na in Xiphister atrapurpureus.
Diffusion permeability coefficients (Pdiff) for 3H2O of cod larvae, which relate ion flux to surface area, are much lower than in adult marine fish but higher than for eggs (Table 2). The low permeability of eggs has been attributed to the physical properties of the vitelline membrane (Potts & Eddy, 1973). It would appear that low permeability in larvae may also be attributable to the structure and composition of the integument. The relatively higher permeability in adults results from the high level of diffusion exchange which occurs across the gills. Isaia et al. (1979) measured an osmotic permeability coefficient of 30x10−6cms-1 for gills in isolated head preparations of SW-adapted trout (Salmo gairdneri), which is considerably higher than that of intact animals (Table 2). It would be of considerable interest to examine the changes in permeability during the development of larvae, when the site of diffusion moves from skin to gills. The ratio of osmotic to diffusional permeability in yolk-sac cod larvae was found to be 1:2-4 (Table 2). In view of the likely 62 % overestimate of Posm (see Materials and methods), Pdiff and Posm in cod larvae are similar. Evans (1967) and Motais et al. (1969) have also found the ratio of the permeability coefficients to be near unity in the adults of euryhaline species. The close match of the permeability coefficients larvae verifies the estimates of drinking rates made by Mangor-Jensen & Adoff (1987) and Tytler & Blaxter (1988b). It also supports the hypothesis that the relatively low drinking rates reported by Tytler & Blaxter (1988b) are a function of low skin permeability.
The water content of cod larvae was estimated, from the equilibrium level of 3H2O, to be 0·32 ± 0·03 mg, which is very close to the wet mass of 0 · 30 mg. This method is an alternative and may be a more reHable method of measuring body water content than conventional weighing techniques which rely on extrapolation to account for evaporative water loss. Pelagic larvae, which tend to be buoyant, have a high water content, for example Craik & Harvey (1984) found the water content of the tissues of cod yolk-sac larvae to 92 %. The 22Na equilibrium level (1260 counts min− 1 in a working solution of 20 · 9 counts min− 1 mmol− 1 Na+) gave a Na+ concentration in the larva of 189 mmol 1− 1. In adult marine fish the plasma Na+ concentration is generally around 180 mmol I− 1 (Rankin & Davenport, 1981), although Evans (1967) found a lower plasma Na+ concentration of 160 mmol I− 1 in Xiphister atrapurpureus, a small intertidal species. From the equifibrium radioactivity of 36Cl of cod larvae the Cl- concentration was estimated to be 148 mmol 1− 1. According to Rankin & Davenport (1981), the typical plasma chloride concentration is 150 mmol 1− 1. Evans (1967) has found a value of 156 mmol 1− 1 in X. atrapurpureus. It seems that cod larvae regulate body fluid hypo-osmotic to SW with sodium and chloride concentrations similar to that in adult fish.
ACKNOWLEDGEMENTS
We wish to thank Professor John Sargent for his generous support and invaluable advice. We are also indebted to the staff of the DAFS Marine Laboratory, Aberdeen, for the cod larvae.