The Effects of Forced and Voluntary Diving on Ventilation, Blood Gases and Ph in the Aquatic Amphibian, Xenopus Laevis

Most cardiorespiratory studies on diving in anuran amphibians have involved either submersion of restrained animals or closing off access to the surface for freely moving individuals. Periods of breath holding under these circumstances result in a well developed diving bradycardia, together with a rapid decline of the blood oxygen store and a combined respiratory and metabolic acidosis (Lenfant & Johansen, 1967; Jones, 1972; Emilio, 1974; Lillo, 1978; Emilio & Shelton, 1980). This has been taken to suggest that anaerobic pathways make a significant contribution to the overall energy metabolism of the submerged animal. Anaerobiosis is indeed a major constituent of the activity metabolism of many amphibians as shown by the increase of lactate in the blood and tissues during exercise (Bennett, 1978). After exercise, the restoration of the lactate concentrations and HCO₃⁻/CO₂ equilibria in the blood takes several hours (Hutchison & Miller, 1979a,b; McDonald, Boutilier & Toews, 1980; Boutilier, McDonald & Toews, 1980). Recovery from enforced anaerobic dives in Xenopus appears to be equally time-consuming (Emilio & Shelton, 1980), and this must have some impact on the frequency and duration of subsequent dives.

There is little evidence, however, to indicate that anaerobic metabolism is of major importance during periods of unrestricted voluntary diving. For example, voluntary dives in both Amphiuma and Xenopus are terminated before oxygen stores become fully depleted (Toews, Shelton & Randall, 1971; Emilio & Shelton, 1974; Boutilier, 1984). In addition, although these animals will voluntarily dive for up to an hour or more, they do not show a post-dive hyperpnoea. There is reason to doubt, therefore, that anaerobiosis plays a significant role in the diving metabolism of unrestrained amphibians during periods of voluntary submergence. In the present paper, we examine this hypothesis by comparing the blood acid–base and oxygenation characteristics together with pre-and post-dive breathing patterns in voluntarily and forcibly submerged Xenopus laevis.

Xenopus laevis (95–153 g) were obtained from a commercial supplier and held in the laboratory at 25°C for several months before the experiments. Each animal was anaesthetized by immersion in a 0·06% solution of tricaine methane sulphonate (MS-222, buffered to pH 7·0), and a cannula was chronically implanted into the femoral artery as described by Boutilier (1984).

Two series of experiments were carried out, both of which continuously monitored ventilation using a pneumotachograph apparatus (see Boutilier, 1984). Briefly, air flow out of a small cylindrical breathing chamber (positioned at the water surface) was passed through a pneumotachograph screen. Pressures on either side of the screen were detected by a differential pressure transducer (Hewlett-Packard Model 270) whose output was stored on an instrumentation tape recorder (Racal Store 4) and recorded on a Medelec For-4 oscillograph. Air flow rates were later integrated to give ventilation volumes. The output from the transducer was linearly related to flow over the range of flow rates encountered. Below the breathing hole the animal was enclosed in an aerated volume of water (25 ±0·5°C) large enough to permit free movements for voluntary diving and surfacing. All animals were given at least 24 h to recover from the surgery and get used to the experimental chamber. The entire apparatus was shielded with a one-way blind to enable observations to be made and to minimize disturbance.

In the first series of experiments (N = 8 animals), a blood sample was withdrawn near the end of a 2-to 3-min breathing spell. As soon as the animal voluntarily submerged, a Perspex lid was inserted between the aerial and aquatic compartments so as to prevent any further lung ventilations for 30 min. Just before the lid was removed, a second blood sample was taken (time zero). As the animal then resumed its voluntary diving-emergence behaviour, samples were taken at +1, +2, +4, +8 and +22 h following the dive. The timing of these samples was not made to coincide with any particular part of the breathing-nonbreathing cycles. The second series of experiments (N = 12 animals) was designed to examine voluntary dives of similar length to the enforced 30-min submergence. When the animals were observed to exhibit their characteristic ‘burst-breathing’ behaviour (bursts of lung ventilations separated in time by long dives; Boutilier, 1984), blood samples were taken every 10–15 min. Owing to the unpredictable length of a voluntary dive and the fact that dives of 30 min duration are less frequent than shorter periods of submergence (Boutilier, 1984), data were collected for numerous dives of 15 min and less. Some of these data have been reported elsewhere (Boutilier, 1984).

Voluntarily diving Xenopus exhibit a wide variety of breathing patterns (cf. Boutilier, 1984), one of which (‘burst breathing’) is characterized by long periods of diving punctuated with comparatively short outbursts of lung ventilation. A breathing burst following a voluntary dive of 30 min or more was always essentially the same as that which immediately preceded the period of submergence (Fig. 1A,B), whereas that following a 30-min enforced dive was always markedly different (Fig. 1C,D). Analysis of the periods of breathing at the surface, both before and after a 30-min enforced dive, is shown in Table 1. The average inspiratory volume of a breath following an enforced dive was not significantly different from that seen prior to the dive. As the mean frequency of breathing was also unchanged, the inspiratory flow (minute ventilation) was the same after as before the period of enforced submergence (Table 1). The large increase in ventilation following enforced dives was therefore due to the increased amount of time spent at the surface (Table 1; Fig. 1C,D).

An analysis of the number of breaths and their associated volumes of inspired gas is shown in Fig. 2 for a single animal contributing to the mean data in Table 1. It is evident that upon emergence from an enforced dive, the frequency of breathing and the associated inspiratory flow volumes were markedly increased ; over the first 2 min of the surface period following the enforced dive, the ventilation frequency was 9·5 min⁻¹ whilst inspiratory flow amounted to more than 27 ml min⁻¹ 100 g⁻¹. These values are far in excess of the mean data shown in Table 1. An initial hyperpnoea following an enforced dive was more evident in some animals (Fig. 1C) than in others (Fig. ID). As is the case for animals diving voluntarily (Boutilier, 1984), changes in the amount of gas inspired were mainly brought about by altering the amount of time spent ventilating at the surface rather than through changes in the depth of breathing (Table 1), though the latter did at times show an appreciable variation (e.g. Fig. 1C). It seems likely that the differences in breathing patterns seen after voluntary and forced dives are caused by the relatively greater anaerobic production of energy in the latter. Furthermore, the variations in the breathing patterns during recovery from enforced dives (e.g. Fig. 1C versus ID) are probably the result of different levels of activity during the periods of submergence.

The respiratory acid-base variables of arterial blood samples, taken either during or shortly after a breathing burst in a freely diving animal (levels shown at B in Fig. 3), fall within the range of values previously reported for this species at 25 °C (Emilio & Shelton, 1980; Boutilier, 1984). As is the case for other chronically catheterized amphibians (McDonald et al. 1980; Boutilier et al. 1980; Boutilier & Toews, 1981), the levels of arterial blood lactate were uniformly low (0·73 ± 0·10 mmol l⁻¹). The very much higher levels of lactate (5–10 mmol l⁻¹) and lower pH values (7·4–7·6) reported for blood obtained from Xenopus by cardiac puncture (Putnam, 1979; Jokumsen & Weber, 1980) are uncharacteristic of blood samples drawn from indwelling catheters in freely moving animals. Animals in the present experiments were isolated from visual and mechanical disturbances by enclosing the experimental chamber behind a blind and cushioning the entire apparatus on foam padding. Blood sampling under these conditions caused no apparent disturbance to the animals.

Although each animal submerged voluntarily, observations after the lid was in place showed that the animals became variously engaged in active attempts to reach the surface before the 30-min time period had elapsed. In a few instances, these bouts of activity occurred early on in the ‘dive’ with the animal eventually returning to the bottom of the tank. Presumably, the activity associated with attempts to emerge marked what would have been the end of the voluntary period of submergence. As voluntary dives vary considerably in their length (Boutilier, 1984), it is not surprising that the onset and duration of such activity was also quite variable.

Enforced dives of this sort led to marked changes in the acid-base status of the arterial blood (Fig. 3). By the end of a 30-min dive (time 0, Fig. 3), a 10Torr increase in arterial blood and a 4·5 mmol l⁻¹ rise in blood lactate had caused the pH of the plasma to decline by 0·24 pH units. The net result of these respiratory and metabolic acid additions to the blood was an overall reduction in the true plasma bicarbonate concentration from 29·2 to 26·4 mmol l⁻¹ (time B to 0, Fig. 3). By 1 h following the dive, all variables had begun to change in direction, with sampling over subsequent hours revealing a gradual return to levels characteristic of the voluntarily breathing animal. Complete recovery of the acid-base status following such dives took more than 4h (Fig. 3).

Similar patterns of change and subsequent recovery were also observed for simultaneously measured levels of arterial blood ⁠, O₂ content and haematocrit (Fig. 4). During the 30-min enforced dive, arterial declined from 80·8 to 20·5 Torr and O₂ content fell to 2·8 vols % from a pre-dive level of 8·5vols%. Over the same time period, haematocrit levels increased considerably from a pre-dive mean of 26% to a value of nearly 40% following the dive. Between 1 and 2h following the dive, all of these variables returned to the levels seen prior to the enforced period of submergence (Fig. 4).

Initial experiments clearly showed that upon emergence from prolonged 30-to 40-min voluntary dives, the blood acid-base status and oxygenation were usually restored to normal pre-dive levels within the first breathing period (Fig. 5 ; Shelton & Boutilier, 1982). Arterial blood samples taken from six Xenopus diving freely for periods of about 30 min remained low in lactate (Fig. 6), suggesting that anaerobiosis was not important during such dives. Arterial rose and pH fell, rapidly at first, slowing later throughout the dive (Fig. 6). The acidosis was mainly respiratory in origin with little or no metabolic component. Mean data for pH, and lactate concentrations of the Xenopus subjected to forced dives of about the same duration as the voluntary dives also appear in Fig. 6, showing the much greater changes for forced than voluntary dives.

In both forced and voluntary 30-min dives, arterial fell on average by approximately 60Torr (Fig. 6). Voluntary dives appeared to be terminated when levels of arterial reached approximately 20 Torr. Evidently, the metabolic acidosis associated with enforced dives was not the result of complete extraction of the blood O₂ store, since the levels of arterial in these animals were as high as those of the voluntarily diving animals after 30min of submergence (Fig. 6).

There are certain problems inherent in making measurements of diving behaviour and of the changes this behaviour produces in gas exchange, cardiovascular performance and metabolic activity. Most observational and experimental methods cause the behaviour to change in some way (Butler & Jones, 1982; Butler & Woakes, 1979; Kooyman et al. 1980). For example, enforced dives on restrained animals, which have been used in the study of bradycardia, selective peripheral vasoconstriction and anaerobiosis of the classical diving response cause many changes that are not typical of animals making voluntary dives. Even when animals are unrestrained, the cannulation and sampling techniques, and even the presence of the experimenter, can affect diving behaviour. In Xenopus, for example, activity levels are higher and dives much shorter than in the totally undisturbed animal (Emilio & Shelton, 1974, 1980). In the free dives described in this paper, every precaution was taken to ensure that the animals were undisturbed, so that a valid comparison could be made with the forced dive experiments.

The evidence presented here leads us to conclude that anaerobiosis is unimportant in free dives, even when they are of considerable duration. There is no accumulation of lactic acid and the fall in blood O₂, the increase in CO₂ and the associated respiratory acidosis are all corrected within the first few breaths, involving 1 or 2 min at the surface (Figs 5, 6).

In view of these conclusions, an assessment of the body’s capacity to store respiratory gases, particularly O2, and of the gas exchange through the skin, is important. Previous analyses, based on experiments on restrained animals (Jones, 1972) or on measurements of the depletion rates in lung and blood gases of active animals (Emilio & Shelton, 1974), have claimed that O₂ stores are quickly used during a dive. The O₂ requirements of undisturbed Xenopus at 20–25 °C, moving only to come to the surface to breathe, is 4·2ml O₂100⁻¹ h⁻¹ or less, of which at least 0·6 ml 0·2 100g⁻¹ h⁻¹ is supplied through the skin (Emilio & Shelton, 1974). In more disturbed animals, O₂ consumption is 8·6 ml O₂ 100 g⁻¹ h⁻¹ and skin exchange is 2·5 ml O₂100 g⁻¹ h⁻¹ (Emilio & Shelton, 1980). The extent of the O₂ stores may be equally variable. The degree of lung inflation can vary by several millilitres, but it seems unlikely that an animal could dive effectively with more than 7–8 ml of air in the lung. A fall in lung from 120 to 30 Torr during a dive (Shelton & Boutilier, 1982) would yield 0·8 to 0·9ml O₂. The blood volume of Xenopus is 13·4 ml 100 g⁻¹ (Emilio & Shelton, 1980) and O₂ capacity is 8·5 vols % (Boutilier & Shelton, 1986). Assuming that 25 % of the blood is arterial and fully saturated and that the remainder is venous and contains 6·3 vols %, the total blood store would be 0·9 ml O₂, of which 0·6–0·7ml would be usable as blood ₂ declined to 20 Torr. Usable tissue stores would be small, at less than 0 ·1 ml. The total usable store of 1·4–1·7 ml O₂ would allow an undisturbed 100-g Xenopus to remain submerged for 23–28 min, whereas a disturbed and more active animal would need to breathe after 14–16 min.

These estimates agree remarkably well with the present and previously reported observations (Boutilier, 1984) and substantiate the hypothesis that free dives are aerobic. Still longer aerobic dives would be possible if the estimates were rather conservative, as they probably are. There is evidence that the skin O₂ uptake increases during a dive (Emilio & Shelton, 1974). The Po gradient across the skin increases as the dive progresses and it is also likely that capillary recruitment occurs in the skin under these conditions (Poczopko, 1957, 1959; Burggren & Moalli, 1984). More O₂ may be taken from the lung than the change suggests, because lung volume declines during a dive. Finally, the estimates of tissue O₂ requirements during a dive may be too high, as they are based on averaged figures over several breathing-diving cycles. Muscle O₂ requirements must increase as the toad swims to the surface and ventilates the lungs, but these demands will be met by O₂ uptake during the breathing period itself. The tissue O₂ requirements in a negatively buoyant, completely inactive toad will therefore be less than the average figures suggest.

The accumulation and removal of CO₂ during a voluntary dive seem to present few problems. The increases more rapidly in the first 10 min of such a dive and causes a mild respiratory acidosis (Figs 5, 6). Towards the end of a 30-min dive, and pH are changing little, probably because of increased CO₂ losses through the skin. The values reached at the end of 30-min voluntary dives are less than those seen in forced dives because of the absence of protons of metabolic origin. In the forced dives, these products of anaerobiosis give rise to CO₂ by dehydration of bicarbonate ions.

The claim that O₂ stores are renewed and excess CO₂ removed during the first few breaths after a dive perhaps needs some qualification. Only the lung (Shelton & Boutilier, 1982) and the arterial blood components have been examined in detail. Further work on the renewal of the venous and tissue stores is in progress. However, it is known that pulmonary blood flow increases substantially when the animal is breathing and that venous blood is preferentially conveyed to the pulmocutaneous vessels in the partially divided heart (Shelton, 1976). Both characteristics are obviously of major benefit in restoring blood and tissue stores in an animal in which the lung is repeatedly ventilated during the time at the surface (Shelton, 1985).

At this stage, we can make no firm conclusions about the relevance of the changes seen in forced dives to voluntary diving behaviour. It is clear from the fall in pH and bicarbonate concentrations that metabolic protons are present in the blood of animals after a 30-min forced dive. The accumulation of lactate is also evidence of extensive anaerobiosis, as is the prolonged recovery period (Fig. 3), which in Xenopus takes about 8 h to complete. However, these dives are invariably associated with increased levels of activity and the overall patterns of acid-base changes in the blood are very similar to those seen in studies of exercise, either in air or with free access to air (McDonald et al. 1980; Boutilier et al. 1980; Boutilier, Emilio & Shelton, 1986). It could be argued that O2 stores are slowly exploited in inactive dives but that sudden bursts of mechanical power and of energy consumption in the voluntary musculature are served very largely by’anaerobic pathways, even though the stores may still contain some O₂. Xenopus, in common with many other amphibians and reptiles (Shelton & Boutilier, 1982), can survive prolonged submergence with marked acidoses and lactate accumulations, well beyond the limits of the 30-min dives described in this paper. The ability to tolerate high concentrations of anaerobic endproducts may be an adaptation to high activity levels (Bennett & Licht, 1973, 1974; Hutchison & Miller, 1979a,b), prolonged diving, or both. Unfortunately the present experiments do not show whether free dives are ever extended to the point where available O₂ stores are completely depleted and anaerobiosis begins.

The range of both internal and external conditions under which extended dives may begin and end are largely unknown. They are worth investigation not only to test whether anaerobiosis can be an important component in a prolonged dive but also to give more insight into the complexities of breathing and diving behaviour. The present experiments suggest that prolonged voluntary dives end and breathing begins when O₂ stores are reduced to certain levels. The relative constancy of CO2 and pH indicates that these factors are of lesser importance. Yet, they cannot be totally ignored because breathing is stimulated to begin at relatively high levels of if is also high (Shelton & Boutilier, 1982). Similarly, it could be argued that a breathing period ends when the O₂ store has been renewed. However, lung ventilations continue for some time after alveolar and arterial are restored to high levels (Shelton & Boutilier, 1982) and so it is also necessary to suppose that the tissue and venous compartments of the store are monitored (Shelton, 1985). In addition other factors, such as the acid-base balance in blood and tissues, must be of importance in ending a breathing period because ventilation at the end of a forced dive seems to be too prolonged simply to be involved in renewing the O₂ stores. Until the full repertoire of these intermittently breathing animals has been studied, the precise nature of the control processes will continue to be elusive.

Funding from the Commonwealth Scholarship Commission is gratefully acknowledged.