Blood gases and respiratory pattern in exercising fowl: comparison in normoxic and hypoxic conditions

Recent studies on the respiratory responses of ducks (Kiley, Kuhlmann & Fedde, 1979, 1982) and domestic fowl (Brackenbury, Gleeson & Avery, 1982; Brackenbury & Gleeson, 1983, 1986) exercised on treadmills have provided information about factors that govern changes in the rate and depth of breathing during spontaneous hyperpnea. It is clear that the respiratory rhythm generator is highly sensitive to variations in intrapulmonary and arterial Experiments on anaesthetized birds have demonstrated the importance of specifically CO₂-sensitive intrapulmonary chemoreceptors (IPCs) in the regulation of breathing pattern (Miller & Kunz, 1977; Tallman & Kunz, 1982; Tallman & Grodins, 1982) and the results from the exercise studies support this role. In chickens it has been suggested that the IPCs regulate the depth of breathing in order to match lung ventilation to CO₂ flux across the respiratory epithelium, thereby preserving normal arterial (Gleeson, Haigh, Molony & Anderson, 1985b). When, as during thermal hyperventilation, minute ventilation rises in excess of CO₂ production the development of a rapid, shallow breathing pattern, under the influence of the IPCs, ensures that most of the excess ventilation is shunted away from the parabronchi into the anatomical dead space.

When chickens are subjected to a range of exercise loads of light to moderate intensity, lung ventilation is precisely matched to increased CO₂ production and remains constant (Brackenbury & Gleeson, 1983). This happens only if rises in body temperature are kept to a minimum and in these circumstances it has been shown that increased ventilatory demand is met by coupled increased in tidal volume and respiratory rate, with the emphasis strongly on the latter variable. Exercise, however, also imposes an increased demand for oxygen as well as for the elimination of CO₂. In normal air-breathing conditions the systemic receptors do not appear to be involved in the control of ventilation during exercise hyperpnea in chickens (Brackenbury et al. 1982). However, if the birds were subjected to exercise in hypoxic conditions, the demand for lung ventilation would exceed the optimum required to maintain normal intrapulmonary In order to investigate the way in which the respiratory rhythm generator would respond to these potentially competing demands, experiments were designed to measure relationships between respiratory pattern and arterial during exercise in normal and hypoxic conditions.

The experiments were carried out on six adult, male domestic fowl (Light Sussex breed, body mass 2·7–3·5 kg), which had been trained for several months to run on a treadmill at a maximum speed of 5·0kmh⁻¹ for periods of 15–20 min. The birds were also trained to wear a loose-fitting plastic mask attached to a 1-m length of plastic tube (external diameter 0·5 cm) which was used to deliver either air or experimental gas mixtures into the mask. Before the experiments began each bird was anaesthetized with a 1:1 mixture of 30% Urethane and sodium pentobarbitone (60 mg ml⁻¹), administered intravenously, and cannulae were sewn into the clavicular air sac and a carotid artery. The free end of the arterial cannula was externalized just caudal to the outer ear and the cannula was flushed daily with heparinized saline. Experiments began 24–48 h after recovery from the anaesthetic.

Arterial blood and were measured using a Radiometer BMSMK2 blood-gas analyser thermostatted at 41·0 °C, close to resting rectal temperature (T_re).

The electrodes were calibrated with a precision gas supply and with precision buffers after each set of readings. The measured values were corrected to actual T_re using factors from Severinghaus (1965). The blood samples were drawn using a double syringe technique, one syringe being used to remove dead-space blood and saline from the cannula, the other to draw the sample. Approximately 1·5 ml of blood was drawn on each occasion and transferred immediately to the blood gas analyser. This process took less than 1 min.

Air sac pressure was monitored using a Grass PT5 manometer connected to a Grass Model 7D pen recorder. Air sac pressure was used to obtain an indication of respiratory airflow rate since there is a direct relationship between respiratory airflow and pressure changes within the lung-air sac system (Brackenbury, 1973) although this relationship is slightly non-linear. It had originally been intended to monitor airflow directly using the mask technique developed in previous studies using hens (Brackenbury et al. 1982; Brackenbury & Gleeson, 1983) but, as Gleeson, Barnas & Rautenberg (1985a) also found, the comb and wattles of the cockerel made it almost impossible to obtain an airtight seal around the mask. It was felt that the hood technique adopted by the latter authors would be unwieldy and would certainly interfere with normal respiratory movements, particularly during high-speed running. The method used to assess the rate and depth of breathing from the air sac pressure trace is shown in Fig. 1. The raw pressure trace was displayed directly and a second parallel output was led into a Grass 7P10 integrator which summated the pressure trace over both the inspiratory and the expiratory phases and automatically returned to baseline at the end of each minute. Since air sac pressure is proportional to airflow, the height of the integrator trace is proportional to the minute volume. Dividing the minute volume by the respiratory rate gave a value proportional to the tidal volume. Since airflow was not measured directly in these experiments, and in view of the slight non-linearity of the relationship between airflow and air sac pressure, no attempt was made to obtain absolute values for minute ventilation and tidal volume. However, the data can be used to obtain estimates of the relative values of these variables.

Paired sets of data were compared for significance of differences at the 5 % level using two statistical tests. First, a one-tailed i-test was used to compare blood gas/pH values in birds breathing air and 10% O₂ in N₂, respectively, since it was expected that the hypoxic birds would also become hypocapnic. Second, in order to avoid type II errors in the comparison of data from birds breathing air and 10% O₂/3% CO₂ in N₂, respectively, a two-tailed Scheffé test was employed.

These experiments were all performed at a treadmill speed of 3·2kmh⁻¹ (0·9m s⁻¹). At the beginning of each experiment the bird was fitted with the mask, the air sac and arterial cannulae were connected and the animal was placed on the stationary treadmill. The mask was continuously ventilated with a measured stream of tanked air (flow rate 10–151 min⁻¹) in order to remove end-expired gas. The bird remained standing on the treadmill until the air sac pressure trace indicated steady breathing and the first blood sample was taken. The treadmill was then started and exercise took place for 6–8 min in air-breathing conditions. A steady-state ventilatory response to exercise was usually achieved within 4–5 min. A second blood sample was taken from the running animal before replacing the airstream with one of the experimental gas mixtures delivered to the mask at the same rate. Exercise continued for a further 4–5 min until a new steady-state ventilatory response had been achieved and a final blood sample was drawn just before the treadmill was stopped. Therefore a total of 4–5 ml of blood was drawn per experiment. This represented less than 2 % of the estimated blood volume of the animals and since only one experiment was performed on each animal during any single day, it was assumed that the sampling procedure had no significant effect on haematocrit. Two experimental gas mixtures were used: 10% O2 in N₂ and a mixture of 10% O₂/3% CO₂ in N₂. Preliminary experiments had indicated that during inhalation of the hypoxic gas alone the birds became hypocapnic. The addition of 3 % CO₂ to the second gas mixture was intended to forestall this drop in and permit ventilatory measurements to be made in normocapnie hypoxic conditions. T_re was measured with a mercury-in-glass thermometer before and immediately after exercise. This process took 1–2 min and it was assumed that the handling procedure had no significant effect on the measured temperature.

In these experiments the birds breathed only air and exercise took place at a speed of 5·0kmh⁻¹ (1·4ms⁻¹). Exercise lasted 6–8min and blood was sampled before, and during the final seconds of exercise. Both series of experiments were performed at relatively low environmental temperatures (10–15 °C) in order to minimize any thermal response from the animals. Previous experiments had demonstrated that in such conditions lung ventilation was precisely matched to increased CO₂ production and arterial and intrapulmonary remained at their control values (Brackenbury & Gleeson, 1983). By measuring ventilatory characteristics in air at two exercise intensities it was therefore hoped to establish the graphical relationships between the rate and depth of breathing during isocapnic hyperpnea. This would then be used as a baseline against which to compare the ventilatory responses to hypoxia in isocapnic and hypocapnic conditions.

T_re rose significantly (P<0·05) during exercise from 41·8±0·08°C to 42·3 ± 0·06°C. However, this rise produced no apparent effect on ventilation since arterial remained unchanged from resting at both exercise intensities (Table 1), suggesting that lung ventilation and CO₂ production were matched. In addition there was no indication of the rapid, shallow panting which has been observed in chickens exercised at much higher environmental temperatures (Brackenbury & Gleeson, 1983). Increased minute ventilation during exercise was achieved mainly by increasing the respiratory rate with relatively small effects on the depth of breathing (Figs 1, 2). Representative traces of air sac pressure changes are shown in Fig. 1. The small spikes shown on the trace were synchronized with leg movements and appeared to represent small air pulsations caused by impact of the legs on the body wall enclosing the lung-air sac system.

When the running birds inhaled 10 % O₂ in N₂, minute ventilation, as indicated by the mean air sac pressure, increased by approximately 35 %. This increase was due entirely to an increased respiratory rate and the tidal volume, given by mean air sac pressure divided by respiratory rate, was reduced in value (Fig. 2). As a result of these changes the ventilatory pattern during hypocapnic hypoxia was faster and shallower than would have been the case if the same increment in ventilation had been achieved by an increase in exercise intensity in isocapnic conditions. The ventilatory characteristics in the latter instance can be estimated from Fig. 2 by projecting the isoventilatory line which passes through the hypoxic/hypocapnic exercise point, towards the line connecting the isocapnic exercise data points. The point of intersection of these two lines then gives the required rate and depth values. Arterial fell by approximately 30 Torr and the blood was slightly but significantly alkalotic compared with its control value at rest (Table 1). Inhalation of 10% O₂/3 % CO₂ in N₂ brought about a similar drop in arterial but this time arterial was held at its resting value. However, it was marginally but significantly higher than the arterial value measured in air-breathing conditions during exercise (Table 1). Ventilation increased by 80% compared with its value in air-breathing conditions and the increase was mostly due to a rise in tidal volume coupled with a small rise in respiratory rate (Fig. 2). Thus the resultant ventilatory pattern during hypoxia was deeper and slower than would have been the case if the same increment in ventilation had been brought about by isocapnic exercise. Again the ventilatory characteristics in the latter instance can be estimated from Fig. 2 by projecting the isoventilatory line, which passes through the data point for inhalation of the 10% O₂/3 % CO₂ gas mixture, towards the line connecting the isocapnic exercise data points.

The effects of exercise on the rate and depth of breathing (Fig. 2) were qualitatively similar to those previously described in running hens (Brackenbury & Gleeson, 1983, 1986) and in running cockerels (Gleeson et al. 1985a), that is ventilation is increased by rises in both tidal volume and respiratory rate with the emphasis on the latter variable. These studies used mask techniques to measure airflow directly and the agreement between them and the present study supports the use of air sac pressure as an indication of relative changes in ventilation. The relationships between minute ventilation, tidal volume and respiratory rate in exercising chickens are strongly dependent on arterial and intrapulmonary ⁠. Thermal hyperventilation can lead to a fall in which is compensated for by a shift towards a more rapid and shallow breathing pattern. In order, therefore, to establish the precise relationships between rate and depth of breathing during isocapnic hyperpnea it is necessary to minimize rises in body temperature. In the present experiments, although it was not possible to prevent a 0·5°C rise in T_re during exercise, this was not sufficient to provoke a thermoregulatory breathing response and arterial remained at its control value.

Control of respiratory pattern in ducks exercised on treadmills appears to differ somewhat from that in the domestic fowl. In ducks increased respiratory rate was responsible for the entire increase in ventilation, and tidal volume actually fell even when exercise took place in isothermic conditions (Kiley et al. 1979, 1982; Kiley, Faraci & Fedde, 1985). Gleeson et al. (1985a) have criticized the tracheal cannulation technique used in the duck studies on the grounds that it might alter the normal pattern of breathing, and it is still not certain to what extent chickens and ducks differ in their responses to exercise. However, a recent study by Woakes & Butler (1986) on swimming tufted duck, Aythya fuligula, suggests that these differences may not be great. These authors measured ventilation using a non-invasive mask technique and found that, over a range of swimming speeds, the main increases in ventilation were due to increased respiratory rate, but with significant increases in tidal volume as well at the highest work loads. Geiser, Gratz, Hiramoto & Scheid (1984) artificially increased metabolic rate and ventilatory demand in ducks by injecting them with 2,4-dinitrophenol and found large increments in respiratory rate coupled with small but significant increments in tidal volume. Some of these birds were reported to be hyperthermic, as was reflected by a 4Torr drop in arterial ⁠, and it is therefore likely that the observed increases in tidal volume were an underestimate of the isocapnic response. Gleeson (1986) also examined the effects of 2,4-dinitrophenol as well as cold exposure on metabolic rate and ventilation in unanaesthetized chickens. In these experiments arterial did not change and there was strong evidence that the alterations in respiratory pattern caused by both of these stimuli were of an identical nature to those produced by isothermic exercise.

In earlier studies on domestic fowl it was found that inhalation of hypoxic gas brought about increased ventilation almost entirely through increased respiratory rate in resting (Brackenbury & Gleeson, 1985) and exercise (Brackenbury et al. 1982) conditions. However, blood gases were not measured in these studies and it was not clear to what extent the changes in respiratory pattern during hypoxia may have been influenced by secondary alterations in ⁠. Bouverot & Sébert (1979) found that inhalation of 12% O2 in N₂ by resting, fully-conscious chickens produced a 40% increase in ventilation accompanied by a 9 Torr drop in arterial and breathing became more rapid and shallow. In the present experiments lung hyper ventilation during inhalation of 10% O₂ in N₂ evidently led to a 3–4 Torr drop in arterial and presumably a similar fall in intrapulmonary ⁠. This may well account for the development of a more rapid and shallow breathing pattern than the control pattern produced by isocapnic exercise (Fig. 2). The observed drop in arterial was not large, but even smaller levels of hypocapnia are capable of evoking very dramatic alterations in ventilatory pattern during hyperthermic exercise. It was anticipated that the addition of 3 % CO₂ to the inhalation gas would have restored the ventilatory pattern to a point on the isocapnic exercise line. It can be seen from Fig. 2 that although restoration of an arterial close to normal did result in an appropriate, adjustment in respiratory pattern, this was clearly an overcompensation. Although arterial after inhalation of the 10% O₂/3% CO₂ gas mixture was the same as its resting value in air, it was significantly higher than the value measured in exercising, air-breathing conditions (Table 1). In view of the sensitivity of the respiratory pattern generator to it is possible that even an increase of 1–2 Torr could have a large influence on the depth/rate ratio.