Urine Production Rate and Water Balance in the Terrestrial Crabs Gecarcinus Lateralis and Cardisoma Guanhumi

Land crabs of the family Gecarcinidae occur widely in the tropical and subtropical Americas and also in the Indo-Pacific region. The genus Gecarcinus is considered to be the most terrestrial of this family and is found in dry burrows, sometimes far inland (Bliss, 1963, 1968). Cardisoma generally lives in low-lying swampy areas and burrows down to ground water (Gifford, 1962; Feliciano, 1962; Herreid & Gifford, 1965).

Bliss (1968) has discussed some aspects of salt and water balance in her review of the adaptations of terrestrial decapods. In adult land crabs, water uptake from moist substrates is thought to occur via the abdominal setae to pericardial sacs, branchial chamber and thence to the haemolymph. Water uptake has been shown to be under hormonal control (Bliss, Wang & Martinez, 1966). Mantel (1968) has shown that the gut of Gecarcinus is permeable to water and salts, which may be taken up from food. Behavioural adaptations are important to ensure that water loss is kept to a minimum.

Because the haemolymph and urine of land crabs are isosmotic, it has been suggested that the antennal gland plays no significant role in the conservation of water (Gross, 1963; Bliss, 1968). This suggestion is based on little independently derived evidence such as that which might be provided by studies of the excretion of filtration-marker substances. For example, it has been shown in Potamon edulis, a semi-terrestrial freshwater crab, that U/B ratios of the marker substance inulin are greater than 1, even though haemolymph and urine are isosmotic. It is suggested that reabsorption of water with some sodium and chloride uptake could account for this (Harris & Micallef, 1971; Harris, 1975). Flemister (1958) found that injected inulin was cleared from the haemolymph of Gecarcinus lateralis kept on sand, indicating that urine was formed during exposure to dry conditions. Flemister was unable to collect urine from the nephropore of his crabs and he assumed that the inulin was retained within the antennal gland. This suggested that complete reabsorption of the primary urine occurred in these conditions. However, in the semi-terrestrial crab Uca mordax, urine inulin concentrations did not exceed those of the haemolymph even under conditions where water reabsorption from the urine would be expected (Schmidt-Nielsen, Gertz & Davis, 1968).

In the present study, the rate of clearance of inulin and inulin U/B ratios of Car disoma guanhumi and Gecarcinus lateralis were measured. This was in order to determine the rate of urine production under different experimental conditions and to assess the importance of urinary reabsorption in the water economy of the land crabs.

Adult G. lateralis and C. guanhumi were collected by hand from burrows near Bellairs Research Institute, Barbados (Chace & Hobbs, 1969). They were kept in the laboratory at 22 °C on sand moistened with sea water into which they burrowed. They were fed on coconut flesh.

Rates of inulin clearance were determined using (hydroxy-[¹⁴C]methyl) inulin (Radiochemical Centre, Amersham). Gecarcinus Ringer solution, 0·1 ml (Skinner, Marsh & Cook, 1965), containing 0·8 μCi of radioisotope was injected into each weighed animal by Agla microsyringe. The injection site was a predrilled hole, covered by a rubber membrane to prevent leakage, in the dorsal carapace above the heart. One and a half hours after injection, 0·5 ml of haemolymph was removed from each animal and mixed with 0·5 ml of distilled water on a planchette. The samples were air-dried and counted, using a Mullard MX 168 G-M counter and Panax scaler. Samples of urine were prepared similarly for counting, o-i ml of urine was collected directly from the nephropore by temporarily displacing the 3rd maxilliped, which almost completely covered the opening. Standards of known dilutions, with distilled water, of the original injection medium and known volumes of inactive haemolymph and urine were used to prepare quench curves. Sample counts were corrected to 100% efficiency using these curves. The degree of initial dilution after injection, and therefore the inulin space, was calculated, and knowing individual space values, the rate of clearance, C (in ml day⁻¹), was determined from the slope of a semi-logarithmic plot of haemolymph [¹⁴C] activity against time. The urine production rate, V (ml day⁻¹), was calculated from the relationship C × U/B. V, where U/B is the final ratio of inulin activity of urine haemolymph. V was also expressed as % body weight day⁻¹. All animals were weighed on a Mettler balance accurate to 0·1 g. A standard, timed weighing procedure was used which included the removal of sand grains adhering to mouthparts, chelae and paraeopods. Regression lines were fitted to inulin clearance data by the least squares method using a Wang 720 C programmable calculator. The statistical significance of the difference between means was calculated using Student’s t test.

To investigate the possibility that water was reabsorbed from the urine of G. lateralis to an extent significant in the water economy of the animal, released urine was diverted laterally by means of rubber strips bonded to the carapace by Eastman 610 adhesive (Fig. 4b). The effectiveness of the seal between carapace and rubber strip was tested. Animals were injected with Gecarcinus Ringer containing indigo-carmine which is cleared by the antennal gland and colours the urine. Observations showed that the animals released urine as discrete droplets which were retained on the rubber strips and were dispersed laterally from the region surrounding the branchial exhalant opening. On S.w.-moist sand, animals operated on in this way maintained a constant body weight. Groups of control and experimental animals with strips in position were injected with [¹⁴C]inulin and transferred to dry sand (R.H. 55%; T = 22 °C). Haemolymph and urine [¹⁴C] inulin activities and osmolalities were measured. The weight of each animal was recorded initially and subsequent weighings made and corrected for weight loss due to haemolymph removal. Osmolality was determined by cryoscopy using an Advanced Osmometer 3W. 0·25 ml samples of haemolymph and urine, diluted 1 :1 with distilled water, were prepared. Standards were treated similarly.

Semi-logarithmic plots of [¹⁴C]inulin activity against time in G. lateralis and C. guanhumi are shown in Fig. 1. Rates of inulin clearance expressed as ml day⁻¹ and % body weight day⁻¹ are shown in Table 1. From these values and mean final U/B ratios the rates of urine production have been calculated (Table 1).

The mean rate of urine production in G. lateralis on s.w.-moist sand was 10’47 ± 1 ·53 (S.E.M.) % body weight day⁻¹. The rate of C. guanhumi was 1·89±0-34% day⁻¹ in animals kept in similar conditions.

In dry conditions G. lateralis showed a marked decrease in the rate of inulin clearance (Fig. 1). The animals were maintained on dry sand in glass containers (R.H. 55%; T = 22 °C) and allowed no access to water. During the experiments the animals lost weight. The mean clearance was 2·15 ±0·68% body weight day⁻¹. The rate of urine production was not calculated since a final inulin U/B ratio could not be determined before the animal died.

There appeared to be a reduction in the rate of urine filtration in G. lateralis maintained in dry conditions. Rates of inulin clearance in C. guanhumi on dry sand were not measured.

The U/B ratio of injected inulin increased in G. lateralis to a mean value of 0·98 ± 0·02 (S.E.M.) after 48 h and was maintained at this level during the course of the experiment (166 h) (Fig. 2). The relationship between [¹⁴C]inulin activity in the haemolymph and the urine (Fig. 3) shows that there was no significant difference between them after 48 h.

Inulin U/B ratios close to unity were also recorded in C. guanhumi on s.w.-moist sand (0·93 ±0·07) and the time course of the changes in U/B ratio were similar. In G. lateralis on dry sand it became difficult to obtain urine samples large enough to determine [¹⁴C] activity after 24 h (Fig. 2).

In marine decapods, inulin U/B ratios generally exceed 1 (Riegel, 1972), suggesting an ability to withdraw water from the primary urine. In these land crabs there was no appreciable withdrawal of water and even a slight dilution of the primary urine seems to have taken place, particularly in C. guanhumi. However, it is more likely that experimental error would account for the ratios deviating slightly from 1.

In G. lateralis the nephropore is almost completely covered by the 3rd maxilliped. Beneath the nephropore lies a pad of setae bordering the groove running ventro-laterally to the margin of the 3rd maxilliped (Fig. 4a). Carson (1974) has suggested that in Gecarcinus ruricola the copious growth of micro-organisms found in the interstices of the setal pad may be important in removing nitrogenous compounds from the urine before fluid is reabsorbed into the haemocoel at the base of the carapace groove that supports the pad. In G. lateralis the arrangement of the nephropore and setal pad is similar. The released urine was diverted to investigate reabsorption of urine (Methods and Fig. 4b). The rates of [¹⁴C] inulin clearance of control and experimental animals were not significantly different (P>0·10) (Table 1).

It had been found in preliminary experiments that diversion of urine in animals kept on s.w.-moist sand did not affect the animals’ ability to maintain water balance. The possibility that reabsorption or urinary water might be important in drier conditions was investigated.

On dry sand G. lateralis showed a decrease in body weight (Fig. 5). This was assumed to be solely due to water loss and to be linear in the first 50 h. The rate of the loss was expressed as a percentage of the initial body weight per day corrected for differences in surface area for different animal weights by the two-thirds rule’ (Table 2). There were no significant differences in the rates of water loss between control and experimental animals (P>0· 10).

On s.w.-moist sand the haemolymph osmolality of G. lateralis was maintained relatively constant. In animals on dry sand it increased sharply and continued to rise until death (Fig. 6). After 70 h the difference between haemolymph osmolalities in control and experimental groups of G. lateralis on dry sand was not significant (P>0·10). An unexpected longer survival time (between 100– 120 h) was recorded in the experimental group. This may be due to the higher mean body weight of these animals, although Bliss (1968) noted that size did not influence survival in this species. She found a mean survival time of 89 h at, however, a higher temperature (30 °C) and relative humidity (78 %).

Determination of both haemolymph and urine osmolalities during desiccation in G. lateralis showed that the U/B osmolality ratio remained at 1·01 ± 0·05 (S.E.M.) up to 24 h after transfer to dry conditions (Fig. 7). Urine samples could not be collected after this time. This was not significantly different from the mean ratio of animals kept on s.w.-moist sand (1·04 ±0·02) (P>0·10) (Fig. 7).

Although the possibility remains that] the experimental procedure was not completely effective in preventing water reabsorption from the released urine, it appears that this process is not an important water conservation mechanism in G. lateralis in dry conditions.

Gecarcinus lateralis living in s.w.-moist sand is apparently in water balance and produces urine at a relatively high rate (10·47% body weight day⁻¹). Rates recorded in many freshwater and estuarine decapods are similar, e.g. Astacus fluviatilis 8·22 %, Bryan (1960); Eriocheir sinensis 3·6– 18·7%, Scholles (1933), De Leersnyder (1967); Carcinus maenas 4·4 % Binns (1969). These rates can be contrasted with those recorded in potamonids, freshwater semi-terrestrial crabs with an Old World distribution, e.g. Potamon nHoticus 0·05– 0·6% (Shaw, 1959); Potamon edulis 0·58% (Harris 1975). The rate found in C. guanhumi was slightly faster than the latter values at 1·89 % body weight day⁻¹ on s.w.-moist sand. This was considerably slower than that of G. lateralis in similar conditions. This is surprising since the former species is usually considered to be less well adapted to terrestrial conditions (Bliss, 1968; Gifford, 1968). The semi-terrestrial ghost crab Ocypode albicans has a rate not dissimilar to that of C. guanhumi of 2·3 % body weight day⁻¹ (Flemister, 1958). No values have been previously reported for the rate of urine production in land crabs, although Flemister (1958) measured inulin clearance rates in G. lateralis.

On s.w.-moist sand the maintenance of a constant body weight implies that the relatively high rates of urine production in G. lateralis, together with water loss by other routes, are balanced by water and salt uptake. Bliss (1963) found that G. lateralis was able to take up water from damp substrates and suggested that water was transported from abdominal setae in contact with the substrate to haemolymph circulating through the gills via the pericardial sacs. Spaargaren (1975) suggested that water was taken up from food in G. ruricola.

The tendency to dismiss the antennal gland as an organ important in osmoregulation in land crabs ignores the possibility of isosmotic withdrawal of water and salts from the primary urine. Production of copious urine isosmotic with the haemolymph represents a potential drain on the body salt content. However, judging from the present results, there appears to be no net water or salt reabsorption from the primary urine in either G. lateralis or C. guanhumi.

The evidence presented here suggests that the urinary water recycling mechanism suggested by Carson (1974) is not an important component of the total water loss in dry conditions. The setal pad situated beneath the nephropore also guards the entrance to the exhalant branchial canal and may prevent the entry of sand grains during burrowing. The presence of micro-organisms in the interstices of the setae may be regarded as a neutral association with the crab. Gifford (1968) has suggested that in Cardisoma urine drains mainly into the branchial exhalant canal also. However, comparison of the position of the nephropore in C. guanhumi with that of G. lateralis shows that the arrangement is quite different in the former. Here the nephropore opens directly onto the carapace and is not covered by the 3rd maxilliped. Urine drains away from the oral region onto laterally situated setal pads.

In dry conditions G. lateralis is capable of decreasing the rate of clearance of inulin from the haemolymph to just over a quarter of the rate on s.w.-moist sand. Flemister (1968) also found that inulin was cleared more slowly in dry conditions. Rates of clearance calculated from his data are slower than those reported here on s.w.-moist sand but faster than the dry sand values. It is assumed that the decrease in clearance rate represents a decrease in the rate of filtration into the coelomic end-sac. Two possibilities can be considered here. Firstly, that there is a decrease in the permeability of the basal lamina to water and permeant solutes such as inulin (Schmidt-Nielsen et al. 1968). Secondly, that permeability remains unchanged but the haemolymph pressure in the antennal artery or adjacent haemocoel is reduced. If the hydrostatic pressure in the coelomic end-sac remains constant and the colloid osmotic pressure of haemolymph is a small component, the net filtration pressure will be reduced (Belman, 1976). It would be interesting to investigate these possibilities in G. lateralis.

The effect on the water balance of the animal of a reduction of the urinary water loss in dry conditions can be assessed by comparing urine production rates and water loss rates. Assuming that inulin clearance represents urine production rate, then urinary water loss is nearly 40 % of the total on dry sand. Maintenance of the urine production rates found on s.w.-moist sand would almost double the water loss of animals in dry conditions. This decrease in rate may be an adaptation to reduce water loss when the animal exhibits nocturnal running in low humidities (Bliss, 1968).

This work was supported in part by the Browne Research Fund of the Royal Society, and by the Research Fund of the University of Leicester. Radioisotope counting facilities were kindly provided by the Department of Physics, University of the West Indies, Cave Hill, Barbados. I would like to thank the Director, Dr Finn Sander, of the Bellairs Institute of McGill University, St James, Barbados, for his hospitality, and Dr A. P. M. Lockwood, of the University of Southampton, for criticism of the manuscript.