ABSTRACT
The gas-transporting properties of the haemolymph of Birgus latro L. were investigated in vitro. This terrestrial anomuran is restricted in distribution to the tropics, and on Christmas Island inhabits a highly stenothermal environment.
The effect of temperature on haemocyanin oxygen-affinity was pronounced (ΔH = –39 kJ mol−1) and was considered to represent the absence of any specific adaptation to environmental temperature. The Bohr effect was large for a terrestrial decapod (ϕ = –0·60), although reduced at low pH. Changes in [Ca] had a significant effect on oxygen affinity of dialysed haemolymph (ΔlogP50/ Δlog[Ca] = –0·39) whereas [Mg] had no effect. Increasing, [L-lactate] had a small effect on the oxygen affinity of dialysed haemolymph (ΔlogP50/ Δlog[lactate] =–0·013) but not whole haemolymph. Dialysis increased oxygen affinity, suggesting the presence of a dialysable component that suppresses affinity. The effect of L-lactate was inhibited in whole haemolymph.
The oxygen affinity of Birgus haemolymph was largely insensitive to effector substances, with the possible exception of Ca. In the case of lactate, at least, this was not due to a reduced sensitivity of the haemocyanin, a situation different from that in closely related species.
Carbon dioxide transport was also affected by temperature. Birgus haemolymph showed a high nonbicarbonate buffer capacity which could be correlated with a high haemocyanin concentration. It is concluded that terrestrial anomuran decapods depend on mechanical adjustments of ventilation and perfusion, rather than employing direct modulation of haemocyanin function, to optimize oxygen delivery.
INTRODUCTION
The anomuran Birgus latro is arguably one of the most terrestrial decapod species and is an obligate air-breather. The species has a wide, although contracting, distribution throughout the tropical Indo-Pacific region (Harms, 1932). The lungs of Birgus are large and have developed by extensive vascularization and evagination of the lining of the branchial chambers (Harms, 1932). Gills are present in B. latro, but are of relatively small surface area and seem unlikely to be important in gas exchange (Semper, 1878; Harms, 1932; Cameron, 1981a).
The roles of the gills and lungs in gas exchange and acid-base balance in landcrabs are poorly understood and there is little direct evidence concerning their relative contributions. In Birgus, the removal of the gills resulted in a small increase in ventilation rate and a slight respiratory acidosis (Smatresk & Cameron, 1981). This would suggest that the gills are not crucial to maintained gas exchange.
In view of this, it is probable that many of the functions carried out by gills in aquatic crustaceans may be partitioned between the two exchange organs in Birgus. Facultative partitioning of gas exchange has been demonstrated in the amphibious Holthuisana transversa in which the gills or the lungs assume ascendency depending on the respiratory medium (Greenaway et al. 1983a, 6; Greenaway, 1984; Taylor & Greenaway, 1984).
Despite a number of papers resulting from the Alpha Helix expedition to the Palau Islands (e.g. Burggren & McMahon, 1981; Cameron, 1981b; Henry & Cameron, 1981; McMahon & Burggren, 1981; Smatresk & Cameron, 1981) our understanding of the respiratory physiology of Birgus is far from complete (see also Cameron & Mecklenburg, 1973).
The role of the haemolymph in the respiration of this species is an important area requiring further study. Our knowledge of the role of modulator systems in regulating haemocyanin function in anomurans is limited to a handful of papers. These recent studies have shown that the haemolymph of terrestrial decapods is insensitive to the identified modulator/effector systems and, moreover, this insensitivity probably evolved more than once (Morris & Bridges, 1986a; Wheatly et al. 1986; Morris et al. 1987). With so few data it is impossible to conclude whether this is a general phenomenon. An important part of this study was to clarify this point, as it has fundamental implications in terms of the stress response of the oxygen uptake and delivery system. The more potent haemocyanin modulators include H+ (Bohr effect) and Mg, Ca and lactate (Mangum, 1983; Bridges & Morris, 1986; McMahon, 1986).
In addition to providing basic data on the function and evolution of haemocyanin, the study was also designed to characterize haemolymph properties essential to a comprehensive study of the respiratory physiology of Birgus. The data are discussed in relation to the ecology and behaviour of the animal with reference to the role of the haemolymph in blood gas transport in vivo (Greenaway et al. 1988).
MATERIALS AND METHODS
Animal collection and sampling
Specimens of Birgus latro (250–500 g) were collected under permit from the Australian Territory of Christmas Island (location 10°28’ S, 105 °38’ E). The crabs were individually packed and then air-freighted within 3 days of collection to the University of New South Wales, Kensington, NSW, where the investigation was carried out. Mortality during shipment was less than 1% and no animals died subsequently. The Birgus were maintained in a humidified constant-temperature room at 25°C either in terraria or, prior to an experiment, individually in plastic fish boxes. Fresh fruit and vegetables and dry, fish-based, cat food pellets were provided on a regular basis, together with ample fresh water for drinking. Both food and water were withdrawn 24 h prior to sampling.
Haemolymph samples (1 ml) were taken from the venous sinus at the base of the second walking leg. Blood was taken from 12 animals and sampling required less than 10 s. The samples were then pooled to provide a homogeneous solution allowing comparison of all data. Any clots were disrupted by forcing the blood through a 25 gauge hypodermic needle after which the blood was centrifuged at 10 000 g and the supernatant haemolymph stored for up to 4 weeks in 1 ml samples at 4°C until required. Control determinations showed no discernible change in the functional properties of the haemolymph over this period.
Initial haemolymph determinations
The concentrations of Na, K, Mg, Ca and Cu in the pooled haemolymph were determined using an atomic absorption spectrophotometer (Varian AA5). For the measurement of [Ca] all samples and standards contained LaCl3 (Sparkes & Greenaway, 1984). The concentration of chloride was measured using a chloride titrator (CMT10, Radiometer, Copenhagen) and the osmotic pressure with a vapour pressure osmometer (Wescor 5100C, Logan, USA). The concentration of L-lactate in the blood and in subsequent haemolymph preparations was determined using the Boehringer test kit (catalogue no. 139084, Boehringer Mannheim GmbH, Mannheim, FRG). This method depends on the conversion of L-lactate to pyruvate by lactate dehydrogenase and the measurement at 339 nm of the concomitant conversion of NAD+ to NADH. Pyruvate is removed from the reaction mixture by glutamate-pyruvate transaminase.
Construction of oxygen equilibrium curves
Oxygen equilibrium curves were constructed using a spectrophotometric system (Morris et al. 1987). All gas mixtures used during the investigation were supplied by Wösthoff mixing pumps (types M301a/f and SA 18, Wösthoff, Bochum, FRG). The pH of all samples of haemolymph was measured near to the P50 oxygen tension using the G299a capillary electrode in a BMS2 blood microsystem (Radiometer). The pH of the blood was normally varied by changing the CO2 content of the gas mixture. This method allowed the calculation of the Bohr coefficient (ΔlogP50/Δ pH) under the various experimental conditions employed. The Bohr coefficient and plots were in all cases calculated by regression analysis of data sets where r > 0·90. The cooperativity (n50) of haemocyanin-oxygen binding was determined by regression analysis of saturation values between 25 and 75% in accordance with the Hill equation.
The dependence of oxygen affinity on [Ca] and [Mg] was determined by constructing equilibrium curves for dialysed blood in which the concentration of Ca or Mg had been adjusted. Dialysis of Birgus haemolymph was carried out in a Ringer’s solution, based on the measured inorganic salt content of the haemolymph, with the following composition (in mmol l−1): NaCl, 316; KC1, 9·4; MgSO4, 1; MgC12, 21; CaCl2, 16·0; NaHCO3, 1. In those cases where the concentrations of Mg or Ca were altered (12 and 46 mmol l−1, 8·4 and 24·7 mmol l−1, respectively) the concentration of NaCl in the Ringer’s solution was adjusted so that [Cl] remained constant.
The specific sensitivity of the haemocyanin oxygen-affinity to CO2 was investigated using techniques described by Morris et al. (1985) in which the CO2 Bohr shift (pH altered by CO2) and the fixed-acid Bohr shifts (0·2 and 3% CO2, pH controlled with HO and NaOH) were compared. Bohr coefficients (ϕ= ΔlogP50/ ΔlogpH) were calculated by regression analysis and compared using analysis of covariance.
The effect of lactate was investigated in both dialysed and nondialysed haemolymph. Samples of haemolymph containing L-lactate at different concentrations were prepared for the construction of oxygen equilibrium curves using previously described methods (Bridges et al. 1984; Morris et al. 1985). The concentration of L-lactate was subsequently determined for all preparations and these values are provided in the appropriate figures.
Determination of CO2 equilibria
The dependence of on on the haemolymph of Birgus was determined at 15, 25 and 35 °C using published methods (Morris et al. 1985). Briefly, 100 μl samples were equilibrated in a BMS2 blood microsystem (Radiometer) with gas mixtures of varying controlled by gas-mixing pumps. The was determined according to the method of Cameron (1971). The [HCO3−] was calculated using CO2 solubility coefficients extrapolated for 80% sea water from the tables of Dejours (1981). Determinations were carried out for both oxygenated (; 1 mmHg = 133·3 Pa) and deoxygenated blood. The dependence of on was determined in triplicate and pH measured in duplicate. Unless otherwise stated all values are given as means ± 1S.D.
RESULTS
The concentrations of inorganic elements determined in the pooled sample were in mmol I −1: Na, 357 + 15; K, 9·4 ±1·2; Cl, 348 ±6; Mg, 18·6 ±2·6; Ca, 17·0 ± 1·4; and Cu, 3·7 ± 0·3. The osmotic pressure was 752 ± 48mosmolkg−1 and the concentration of L-lactate was 0·75 mmol l−1. From the measured [Cu] a theoretical maximal oxygen-carrying capacity of 1·85 mmol l−1 was calculated.
The effect of temperature on O2 affinity
Increasing temperature markedly reduced the oxygen affinity of Birgus haemolymph but did not affect the magnitude of the Bohr coefficient (analysis of covariance, P < 0·01) which had a value of –0·60 ±0·04. The Bohr coefficients calculated for each temperature are shown in Fig. 1. The effect of temperature was not discernibly different for 5 °C intervals throughout the temperature range used and the calculated values of ΔH yielded a mean value of –38·5 ± 10·9 kJmol−1. Changes in temperature had no measurable effect on cooperativity (Fig. 1), and the mean n50 was 2·84 ± 0·28.
The effect of [Ca] and [Mg]
Neither the Bohr coefficient nor the P50 of dialysed Birgus haemocyanin was altered by changes in [Mg] within the range 12 –46 mmol l−1 (Fig. 2A). The combined data could be described by a single equation with a slope of –0·60. Dialysis of the haemolymph induced a small, but significant, increase in oxygen affinity (analysis of covariance, P < 0·05) but no change in the Bohr factor.
Increases in [Ca] from 8·4 to 16 and from 16 to 24·7 mmol l−1 resulted in significant increases in oxygen affinity (analysis of covariance, P < 0·01) in the absence of a change in the Bohr factor (Fig. 2B). The effect of [Ca] on haemocyanin oxygen-affinity within the pH range used could thus be described by the equation: logP50 = 1·21 – 0·391og[Ca] (intercept ± 0·04, slope ± 0·12).
The variability of the estimate of cooperativity was greater in the dialysed blood, but the mean value of n50, 3·09 ±0·85, was not significantly altered compared with that of nondialysed haemolymph or in response to changes in [Mg] and [Ca].
The specific effect of CO2
A comparison of the CO2 and fixed-acid Bohr coefficients revealed no specific effect of CO2 when its concentration in the haemolymph was increased from 0·2 to 3·0% (Fig. 3). Over the extended pH range used the slightly curvilinear nature of ΔlogP50/ Δ pH became apparent and therefore comparisons of the Bohr coefficient (ϕ) were made over the appropriate pH ranges (Fig. 3). No dependence of n50 on CO2 concentration could be measured.
The effect of L-lactate
There was no effect of L-lactate on the oxygen affinity of whole haemolymph, but a small effect was measured for dialysed haemolymph (analysis of covariance, 0·10 > P > 0·05) as indicated in Fig. 4. The increase in haemocyanin oxygenaffinity after dialysis was again evident (P < 0·05). The small effect of L-lactate was determined at pH7·8 as ΔlogP50/Δlog[lactate] = –0·013. Again there was no discernible effect on the cooperativity of oxygen binding, n50 = 2·80 ± 0·19.
The effect of temperature on haemolymph CO2 equilibrium
The relationships between and in the haemolymph of Birgus at 15, 25 and 35 °C are shown in Fig. 5. There was a tendency for CO2 capacity to increase at lower temperatures above that dictated by the increased solubility of CO2, indicating increased Hc–CO2 association. The CO2 equilibria presented no evidence in support of a Haldane effect. Combining data from oxygenated and deoxygenated blood, the dependence of pH on temperature (ΔpH/ΔT) was calculated as –0·006 mmol l−1 pH unit−1 ( varied by 0·11 mmHg).
The [HCO3−] calculated from these data are shown in Fig. 6 which shows the effects of temperature more clearly. A temperature increase from 15 to 25 °C elicited a decrease in HCO3− capacity of >3 mmol l−1 at constant pH, without altering the nonbicarbonate buffer power of the haemolymph which averaged –15·3 mmol l−1 pH unit−1. An increase by a further 10°C to 35°C resulted in a significant increase in the buffer power to –20·3 mmol l−1 pH unit−1. It was unlikely that this result represented haemoconcentration due to evaporation of the haemolymph, as a fresh 100 μl sample was equilibrated for each determination and the gas mixtures used were humidified. Comparing deoxygenated and oxygenated blood at constant pH confirmed the absence of any significant Haldane effect.
DISCUSSION
There are few reports in the literature on the physiology of anomuran haemocyanins (e.g. Morris & Bridges, 1986a; Wheatly et al. 1986). That of Birgus was investigated, however, by Burggren & McMahon (1981) who reported P50 values (see also McMahon, 1986). Their reported value of a P50 of 21 mmHg at 30 °C (pH 7·6) is lower affinity than that reported for the same temperature in the present study (13·6 mmHg, pH 7·6). The difference could be due to variations between the Indian and Pacific Ocean populations. However, a single value of 14·5 mmHg (28°C, pH7·5) reported by Cameron & Mecklenburg (1973) for Birgus from the Marshall Islands agrees with the data presented here. These values represent, however, a moderately low oxygen affinity fairly typical of terrestrial decapods, anomurans particularly, and suggest that the pigment functions as more than a simple store for O2 (for reviews see Mangum, 1983; McMahon & Burggren, 1988; see also Greenaway et al. 1988). The calculated oxygen-carrying capacity of 1·85 mmol l−1 is close to the maximum recorded value for a decapod (Mangum, 1983). It is likely that using [Cu] to estimate oxygencarrying capacity leads, however, to an overestimate of haemocyanin concentration and O2-carrying capacity, as measured arterial was lower than that derived from [Cu] (1·32 mmol l−1) (Greenaway et al. 1988). According to the analysis of McMahon & Wilkens (1983), this high oxygen content may be correlated with a low, mass-specific, cardiac output and therefore represents an energetic saving. The calculated Bohr effect in Birgus, although relatively modest in comparison to some aquatic species, is large for a terrestrial species (for reviews see Mangum, 1983; McMahon & Burggren, 1988). Another estimate of the Bohr coefficient of –0·90 for Pacific Birgus, discussed by McMahon (1986) and McMahon & Burggren (1988), supports this conclusion.
The effect of temperature
Increased temperature significantly reduced the oxygen affinity of Birgus haemocyanin. Few higher Δ H values (–39kJmol−1) have been recorded for crustacean haemocyanin (Bridges, 1986; Burnett et al. 1988). In addition, the sensitivity was uniform over the temperature range used, a situation which differs from that reported for both the land hermit crab Coenobita clypeatus (Morris & Bridges, 1986a) and the terrestrial brachyuran Ocypode saratan (Morris & Bridges, 1986b). In contrast, the terrestrial/freshwater crab Holthuisana transversa exhibits a response similar in many ways to that of Birgus (Morris et al. 1987). In the rain forests of Christmas Island diurnal and seasonal temperature variations are only 2–3°C. Under these conditions it is difficult to imagine that there is any selective pressure for these crabs to adapt to extreme temperatures. The accumulating published data on the O2-haemocyanin temperature response now suggest that those species inhabiting stenothermal environments, or having behavioural mechanisms to avoid temperature extremes, exhibit no reduced temperature-sensitivity at temperatures near the environmental mean.
The effects of [Mg] and [Ca]
Both Mg and Ca are known to affect the oxygen affinity of most crustacean haemocyanins (see Introduction) and Ca has previously been suggested to be an effector of Birgus haemocyanin (McMahon, 1986), but until now this has been unsubstantiated. Recent investigations suggest that the haemocyanins of terrestrial crabs are insensitive to effector substances (Morris & Bridges, 1986a,b; Wheatly et al. 1986; Morris et al. 1987). Although the insensitivity to [Mg] supports this conclusion, oxygen binding by the haemocyanin of Birgus exhibits a marked Ca-sensitivity. Truchot (1975) showed that the haemocyanin of Carcinus maenas was sensitive to [Ca] (ΔlogP50/Δlog[Ca] = –0·28) and Mason et al. (1983) demonstrated a high sensitivity for Callinectes sapidus (–0·82). The measured value for Birgus (–0·39) might therefore be important. Burggren & McMahon (1981) demonstrated that 3 –4 days of dehydration could result in 14% loss in body mass and as much as 50% increase in the [Ca] in the haemolymph of Birgus. Observations on Christmas Island indicated that the crabs drank frequently from hollows in tree roots (rain water) and from ground water (3 mmol l −1 Ca) when available. Dehydration and the fluctuating [Ca] of drinking water may result in varying haemolymph [Ca] and thereby alter oxygen affinity. Whether this is the case, or indeed is of any adaptive significance, is unclear.
The effect of L-lactate
As with [Ca], physiological variations in [L-lactate] are probably not important in affecting the oxygen affinity of haemocyanin in anomuran crabs (Morris & Bridges, 1986a; Wheatly et al. 1986). A small lactate effect was measured in the dialysed haemolymph of Birgus but, for several reasons, must be considered to be of dubious physiological significance. First, a coefficient of –0·013 is relatively small (cf. Bridges & Morris, 1986) and in Birgus haemolymph at pH 7·8 an increase in [lactate] of 10·3 mmol l−1 increased O2 affinity by only 0·6 mmHg. However, Birgus produces exceptionally large amounts of L-lactate (Greenaway et al. 1988). Second, this effect of L-lactate was not apparent in whole haemolymph. Dialysis of Birgus haemolymph increased affinity and may also have removed an inhibitor of the potentiating effect of L-lactate. Recently identified modulators such as urate, which are known to accumulate in terrestrial decapods, including Birgus (Gifford, 1968; Henry & Cameron, 1981), have a secondary effect of diminishing the response to lactate (Morris et al. 1986; Morris & Bridges, 1986c). It would seem possible that a similar effect is responsible for the inhibition of the lactate effect in the haemolymph of Birgus.
The specific effect of CO2
The oxygen affinity of Birgus haemocyanin is insensitive to changes within the physiological range. Severe exercise in Birgus results in depression of haemolymph by up to 50%, usually accompanied by some increase in (Smatresk & Cameron, 1981; Greenaway et al. 1988). A potentiation of oxygen affinity by is unlikely to be beneficial under these conditions. Therefore, there would seem to be little selective pressure to promote such an effect in an airbreather, and the in vivo data indicate that it is not required (Greenaway et al. 1988).
The effect of temperature on haemolymph CO2 equilibria
Temperature increases have the predictable effect of decreasing haemolymph CO2 capacity but often to an extent beyond that due to the decreased solubility of the gas (Morris et al. 1985; Taylor et al. 1985; Morris et al. 1987). The effect of temperature change on CO2 equilibria in Birgus haemolymph between 15 and 25 °C conforms to the normal pattern. Between 25 and 35°C this is true only at high pH owing to an apparent increase in the nonbicarbonate buffer capacity of the haemolymph. The physiological importance of these latter data should be treated with some reservation until a corroborative study can be carried out.
The buffer power of the haemolymph was high at all temperatures, corresponding to a high haemocyanin concentration (140 mg ml−1 calculated from [Cu]; approx. 110 mg ml−1 calculated from assuming haemocyanin Mr = 75 000). The importance of the small amount of nonrespiratory protein normally present in crustacean haemolymph will be minimal in the presence of large amounts of haemocyanin. The measured of –15·8 mmoll− pH unit−1 was smaller than the value reported by Smatresk & Cameron (1981) (–22·7mmoll−1 pH unit−1), although these workers report resting in vivo pH values similar to those of Birgus from Christmas Island (Greenaway et al. 1988). Considering the large amount of metabolic acid released into the extracellular space, it is likely that the large amount of haemocyanin has a second, important role as a buffer substance. The value reported by Cameron (1981b), – 16·0mmol l−1 pH unit−1, was essentially the same as that measured in the present study.
There was no detectable Haldane effect. Although the haemolymph was almost completely deoxygenated and resaturated during a single circulation in exercising Birgus, this apparently played no role in the excretion of CO2. Linkage between the Bohr and Haldane effects implies that it would be impossible to evolve a significant oxygenation-dependent CO2 capacity without increasing the pH sensitivity of oxygen binding above that found here for Birgus haemolymph. Such enhanced pH sensitivity of oxygen binding may compromise oxygen loading in a terrestrial species experiencing severe haemolymph acidosis.
In conclusion, the haemocyanin of Birgus is present at high concentration in the haemolymph leading to high O2 and CO2 capacities as well as relatively elevated buffer capacity. Modulator-insensitive haemocyanins in terrestrial decapods probably evolved independently several times, leading to an increased dependence of maintained oxygen delivery on mechanical/morphological adaptations. The absence of lactate-sensitivity in Birgus haemolymph was not due only to the sensitivity of the haemocyanin but also to the inhibition of the lactate effect. Whether the Ca effect is reduced or inhibited in whole compared with dialysed haemolymph is as yet undetermined. Techniques different from those used in the present study would be required to clarify this. Temperature is the major modifying influence on haemocyanin oxygen-affinity, even though these animals occupy stenothermal habitats. A marked Bohr shift at resting pH becomes much less significant at the low pH levels seen in exercised Birgus, reducing the importance of this phenomenon. Birgus haemocyanin functionally resembles that of other terrestrial decapods but somewhat different mechanisms may be involved. Modulation of Birgus haemolymph, as with other investigated terrestrial species, appears to be unimportant, suggesting that air-breathing crabs rely primarily on ventilatory and perfusion adjustments to maintain oxygen supply to the tissues.
ACKNOWLEDGEMENTS
We thank Caroline Farrelly for her help on numerous occasions throughout this investigation and also David Hair for technical assistance. SM and BRM thank Professor T. Dawson, Dr A. M. Beal and other members of the School of Biological Science, University of NSW, for their assistance and hospitality. Financial support was provided by a NATO overseas fellowship (NERC, UK) and by an AHFMR research allowance to SM, ARGS grant A18616299 to PG and NSERC grants A5762, T7670 and IC-0265 to BRM.