Ventilation and Gas Exchange During Shallow Hypothermia in Pigeons

There is a growing body of evidence that animals entering periods of torpor develop a respiratory acidosis. This may be important as a means of inhibiting cell metabolism (Malan, 1980, 1986). Even though such an acidosis has been demonstrated in widely different species of animals (Malan, 1980; Barnhart, 1989), this aspect of hypothermia has not been investigated in birds. In a parallel paper (Jensen and Bech, 1992), we have described the arterial acid-base balance of the domestic pigeon (Columba livia) during shallow nocturnal hypothermia and have shown that respiratory acidosis does indeed develop in this species simultaneously with a reduction in body temperature and oxygen consumption.

As the short-term regulation of systemic pH is mainly accomplished by adjusting the ratio of ventilation to CO₂ production, it was of interest to analyze the ventilatory changes that take place during regulated hypothermia. In previous papers on ventilation in torpid birds no clear-cut conclusions could be drawn regarding the acid-base changes (Withers, 1977a; Bucher and Chappell, 1989). Thus, the aim of the present study was to quantify the ventilatory changes and the changes in the composition of the air-sac gases during entry into shallow nocturnal hypothermia in the pigeon, paying special attention to the regulation of the arterial partial pressure of CO₂.

In mammals, entry into torpor is accompanied by an early and sudden decrease in the gas exchange ratio, indicating an early build-up of CO₂ and a lowering of the pH of the blood (Bickler, 1984; Nestler, 1990). In snails entering estivation, a decrease in the respiratory gas exchange ratio has been found concomitant with decreases in pH and in oxygen consumption (Rees and Hand, 1990). Furthermore, during the initial stage of arousal from hibernation, the European hamster (Cricetus cricetus) has been shown to hyperventilate, releasing CO₂ and increasing its blood pH before thermogenesis is fully activated (Malan et al. 1988).

By measuring the arterial pH and partial pressure of CO₂ in the pigeon, we were unable to demonstrate such an early build-up of CO₂. Acidosis seemed to develop gradually, i.e. there was a constant pH in vivo during entry into hypothermia. Consequently, another aim of the present study was to investigate whether the changes in the ventilatory variables were consistent with such a gradual development of acidosis.

Before the start of the experiments, domestic pigeons of both sexes were deprived of food for 1–2 weeks, with water available ad libitum. Thus, when the experiments were carried out, the pigeons were in phase II (Cherel et al. 1988) of their fasting period. They were kept in cages at room temperature (20±2°C) with a constant photoperiod, 10h:14h L:D. This photoperiod was maintained during the experiments, but the experimental ambient temperature (T_a) was 25±1°C (thermoneutral). Recordings were started in the light phase at 15:00-16:00h after a Ih stabilization period; the light was turned off at 17:00h and recording continued during the entry into hypothermia, until about 23:00h.

Two different experimental protocols were used. In six pigeons, rate of oxygen consumption ⁠, respiratory frequency (f) and inspired tidal volume (FT) were measured simultaneously. In seven other pigeons, the partial pressures of CO₂ and O₂ in the interclavicular and abdominal air sacs were measured. The mean body masses of the two groups of pigeons before the fasting period were 467 g (range 401–530 g) and 511g (range 476–552 g), respectively. During the experiments the mean body masses were 337 g (range 293–400 g) and 376 g (range 335–418 g), respectively. Body temperature (T_b) was measured in all the experiments.

and T_b were measured as described by Jensen and Bech (1992). Briefly, air was sucked through a 71 metabolic chamber and the effluent air directed into an oxygen analyzer (type 1100A, Servomex, England) for continuous recording of the O2 content. was calculated according to Withers (1977b), assuming a respiratory gas exchange ratio of 0.75. T_b was measured with a copper-constantan thermocouple (type 0.005, Finewire, California), inserted into a blind-ending tube (PP30) implanted into the peritoneal cavity.

The state during which T_b was above 40.0°C will be referred to as euthermia. When referring to hypothermia, the T_b range will be defined in each case.

Values of are expressed as mlsTPDkg^-1min^-1.

The pigeons were placed in a metabolic chamber during the experiments, so that could be measured simultaneously with the ventilatory variables. Pneumotachography was used to measure VT and f(Glass et al. 1978). Each pigeon was fitted with a mask that covered its beak and nostrils and from which a 4 cm long tube protruded (Fleisch flow transducer head). Flexible tubing attached to the head of the flow transducer was led out of the chamber and connected to a Godart pneumotachograph (Gould Inc., type 17212). Before starting each experiment, calibrations were made by injecting known volumes of air through the head of the flow transducer using a plastic syringe.

Each pigeon was fitted, under halothane anesthesia, with two polypropylene tubes (PP120), one inserted into the interclavicular air sac and the other into the right abdominal air sac. The tubes were firmly fixed to the surrounding muscles and skin with surgical thread. The tubes were sealed and the pigeons were allowed at least 2 days in which to recover before any air samples were taken.

During an experiment the pigeon was placed in the metabolic chamber for measurement of (see above). The two air-sac tubes were directed through an air-tight hole in the chamber wall and connected to a suction pump for sampling at a constant rate of 1ml min^-1. 10–15 ml of air was sucked into air-tight plastic syringes for each sample.

The partial pressures of CO₂ and O₂ in the interclavicular air-sac samples (⁠ and ⁠) and in the abdominal air-sac samples (⁠ and ) were measured, at 40.0°C, on a blood gas analyzer (Radiometer, Denmark, type BMS 3 Mk2 blood micro system). Before each experiment, the blood gas electrodes were calibrated, using two separate gas mixtures (2.96%CO₂/17.21 %O₂/79.83%N₂ and 10.00%CO₂/9.97%O₂/80.03%N₂). The partial pressure of a gas in the air sac at body temperature was calculated assuming that the air-sac gas was saturated with water vapor.

For overall mean values, the mean value for each individual pigeon was counted as a single observation. Results are expressed as means ±1S.D. When testing for differences between mean values, a two-tailed paired-sample t-test was used.

Regression lines were calculated using the method of least squares. To determine whether the slope of the regression line differed significantly from zero, a two-tailed t-test was carried out. The chosen level of significance was P<0.05.

Individual recordings of and are shown in Fig. 1; the corresponding overall mean values from two T_b ranges are given in Table 1 along with the values of f, V_T, and ⁠. declined significantly during the entry into hypothermia; this was a consequence of reductions in both f and V_T. The percentage declines in and were almost equal, so that even though some of the individual measurements showed a reduction in the ratio when T_b decreased, the overall mean values for the lower and higher T_b ranges were not significantly different. However, the ratio decreased significantly; on average by 24% from euthermia to the T_b range 37–38°C.

Fig. 2 shows the partial pressures of the interclavicular and abdominal air-sac gases. From three pigeons, we obtained air samples from both air sacs simultaneously; in another two pigeons only the interclavicular air sac was sampled, and in two more pigeons only the abdominal air sac was sampled. All the data were used to calculate the overall mean values at euthermia and hypothermia (T_b=37.0-38.0°C) given in Table 2. The changes in the partial pressures of the air-sac gases were greater in the abdominal than in the interclavicular air sac. In the latter, only the partial pressure of CO₂ changed significantly. The changes in ⁠, and were significantly different (different slopes, P<0.01).

The gas exchange ratios (RE) of the air sacs, which are shown in Fig. 3, were not affected by the changes in T_b, but RE was considerably higher in the abdominal than in the interclavicular air sac.

The respiratory frequency of birds has been found to increase as a result of using the pneumotachographic method of measurement (Bucher, 1985). Thus, the extra dead space imposed by the mask might have altered the normal respiratory variables for our pigeons. Stahel and Nicol (1988) found an increase in both/and V_T when the pneumotachographic method was used on the little penguin (Eudyptula minor) as compared to a barometric method. However, these same authors reported that the ventilation of ducks was not affected by the method used. Bech et al. (1985) measured the oxygen extraction of pigeons, in relation to the ambient temperature, using the pneumotachographic method. They concluded that oxygen extraction did not change in response to low ambient temperatures, the same conclusion as had previously been obtained using wholebody plethysmography (Barnas and Rautenberg, 1984), even though the absolute level of oxygen extraction was increased by wearing a mask. The value of measured here (0.3801 BTPS mmol^-1) does not differ significantly from that of 0.3721 BTPS mmol^-1 reported by Bech et al. (1985) for pigeons that had been fed. It should also be noted that the values measured in pigeons fitted with a mask (11.95 ±0.99 ml STPDkg^-1min^-1) were not significantly different (two-tailed two-sample r-test) from those obtained from pigeons without a mask (12.4±2.3mlsTPDkg^-1min^-1, Jensen and Bech, 1992). This indicates that the relative changes in the respiratory variables observed during the entry into hypothermia were not significantly influenced by the method used.

Two early studies of the composition of the air-sac gases in the pigeon reported widely differing values for the interclavicular air sac: 6.0 kPa and 4.1 kPa for and 11.8kPa and 14.4kPa for (Soum, 1896; Plantefol and Schamke, 1934, respectively). The divergence between the two studies was less for the abdominal air sac, for which the reported values were 2.5 kPa and 3.1 kPa and the values 16.0kPa and 15.2kPa, respectively. In the earlier study no anesthesia was used. We can offer no explanation as to why these values should differ so markedly from the results reported here (Table 2).

In the goose (Anser anser, Scheid et al. 1989) the values of (5.2 kPa) and (12.3 kPa) in the interclavicular air sac were very similar to the values reported here for the pigeon. The values for the abdominal air sac were very different, and ⁠, indicating that neopulmonic gas exchange in the goose is slower than that in the pigeon.

The higher RE values recorded for the abdominal air sac compared to those for the interclavicular air sac (Fig. 3) support the conclusion drawn from the data for the goose (Scheid et al. 1989) that the neopulmo plays a quantitatively more important role in the elimination of CO₂ than in the uptake of O₂.

The cross-current arrangement of the blood capillaries lining the parabronchi of birds makes it possible to have a higher partial pressure of CO₂ in the parabronchi than in the arterial blood. Such negative differences in the values between the arterial blood and the interclavicular air-sac gas have been reported for the duck (Powell et al. 1978), the domestic fowl (Piiper et al. 1970) and the goose (Scheid et al. 1989). The difference in the case of the goose was -0.9 kPa. In the present study of the pigeon, was 5.07kPa at euthermia compared with an arterial partial pressure of CO₂ of 3.23 kPa (Jensen and Bech, 1992), a difference between the arterial blood and the interclavicular air-sac gas of —1.84kPa. Although these values were obtained in different experimental series, they clearly indicate the effectiveness of the gas exchange mechanism of the avian lung.

Few ventilatory data exist for birds in a state of torpor. Withers (1977a) studied two species of hummingbirds (Selasphorus sasin and Calypte anna) and the poorwill (Phalaenoptilus nuttalli). For all three species f and decreased dramatically during entry into torpor, whereas a slight decrease in V_T only was observed in the poorwill. V_T decreased by 22% in the poorwill, far less than the eightfold decrease recorded in/. Our pigeons, in contrast, decreased V_T by 18% and/by 20% when entering shallow hypothermia (Table 1). During torpor, the poorwill reached a minimal T_b of 10°C associated with a tenfold reduction in ⁠. It would seem to be crucial for the poorwill to reduce f more than V_T in order to avoid increasing the relative dead space ventilation too much.

Bucher and Chappell (1989), using two different species of hummingbirds (Selasphorus rufus and S. platycercus), confirmed that the decrease in was only induced by reductions in/. Withers (1977a) found a decrease in ⁠, whereas Bucher and Chappell (1989) found an almost twofold increase when the hummingbirds changed from euthermia (T_b=40.6°C) to a state of torpor (minimal T_b=12.2°C). The same authors also reported that torpor in hummingbirds is characterised by periods of apnea lasting up to 5 min. When entering torpor, the ratio of the poorwill was constant (Withers, 1977a).

In the pigeon, the decrease in (Fig. 1) on entry into hypothermia indicated an increase in in the lungs, leading to an elevated arterial partial pressure of CO₂in vivo. This was confirmed by the results of the parallel study. Following a 4°C decrease in P_b, the total arterial CO₂ content increased by 16 %, which can be attributed to parabronchial hypoventilation and an increase in the solubility coefficient of CO₂ of 8% (calculated from Reeves, 1976). Furthermore, this is in accordance with the observed increases in both ⁠, and (Fig. 2).

The greater changes of partial pressures in the abdominal than in the interclavicular air sac may have been caused by several factors. During entry into hypothermia, a parabronchial hypoventilation was produced by decreases in both f and V_T, leading to an increased and a decreased in the residual dead space air, thus inducing a change in the composition of the air that entered the parabronchi. This could have altered the composition of the gas in the abdominal air sac more than in the interclavicular air sac, because the inspired air passes through the abdominal air sac before it enters the paleopulmo (Scheid et al. 1974) where the major part of gas exchange takes place.

Another factor that could have caused differential changes in the composition of the gas in the different air sacs is a relative redistribution of the air flow within the lung/air-sac system. Not all the air that enters the abdominal air sacs passes through the neopulmo; some of it may take a bypass route, thereby avoiding gas exchange in the neopulmo on inspiration (Scheid et al. 1989). A relative redistribution of the air passing into the neopulmo and that bypassing it to enter the abdominal air sacs directly could lead to a change in the composition of the air in this air sac without having much effect on the composition of the air in the interclavicular air sac. However, very little is known about the extent to which birds are able to regulate the air flow within their lung/air-sac system.

In mammals, most CO₂ retention leading to respiratory acidosis probably takes place at the beginning of entry into torpor, because a sudden decrease, of short duration, in the respiratory gas exchange ratio has been observed both in the deer mouse (Peromyscus maniculatus;Nestler, 1990) and in the desert ground squirrel (Spermophilus tereticaudus;Bickler, 1984). It is not clear whether this is also true for mammals entering shallow hypothermia. In the pigeon, in contrast, the respiratory acidosis apparently developed gradually with a constant pH in vivo when T_b decreased (Jensen and Bech, 1992). The present results confirm this view, i.e. the ratio remained constant and the change in the composition of the air-sac gases resulted in constant gas exchange ratios during entry into hypothermia (Fig. 3). Thus, there was no indication that the changes in P_CO2 in the air sacs could have induced any sudden decline in arterial pH that could be responsible for an initial reduction in the metabolic rate. However, the gradual development of respiratory acidosis induced by the effective parabronchial hypoventilation may have had an effect on the metabolic rate at a later stage of entry into hypothermia.