Respiration in Exercising Fowl. III. Ventilation

Both neural and humoral factors have been implicated in the control of ventilation during muscular exercise (reviews by Astrand & Rodahl, 1977; Levine, 1978). Moderate exercise in man and dog induces proportional increases in oxygen consumption and ventilation thus preserving oxygen extraction at its resting value (Flandrois, Lacour & Osman, 1971 ; Koyal et al. 1976). During severe exercise, however, ventilation rises in excess of oxygen consumption, primarily due to metabolic acidosis (Wasserman, Kessel & Burton, 1967) with the consequence that the oxygen extraction falls. In exercising birds, oxygen extraction would be expected to be influenced by environmental temperature since hyperventilation is a characteristic form of thermo-regulation in this group of animals. At environmental temperatures of 12–23 °C the extraction remained unchanged in flying crows (Bernstein, 1976) and starlings (Torre-Bueno, 1978) but suffered a marked decrease at higher temperatures in crow, budgerigar (Tucker, 1968) and raven (Hudson & Bernstein, 1981). Changes in air sac and blood gases indicate that running birds also hyperventilate the respiratory system (Kiley, Kuhlmann & Fedde, 1979; Brackenbury, Avery & Gleeson, 1981a) and the-latter authors found evidence that the degree of hyperventilation was governed by environmental temperature and by exercise intensity. In order to obtain further understanding of the way in which thermal and non-thermal factors interact in the control of breathing in exercising birds we have made direct measurements of ventilation in fowl running at different speeds at normal and elevated environmental temperatures.

Experiments were performed on four adult, female domestic fowl (av. body wt. 2·2 kg), all of which had received approximately 10 months regular training in connexion with two previous studies (Brackenbury et al. 1981a, b). Briefly, the animals were trained to run on a treadmill for periods of 10 min at speeds of 1·24–4·3 km h⁻¹ and at two sets of environmental temperatures defined as control (18 ± 2 °C) and warm (35 ± 2 °C). All four birds performed runs at both sets of temperatures. Heat-stressed birds received at least 1 h pre-exposure to the raised air temperatures before measurements were made. During exercise the birds wore a tight-fitting plastic mask (weight 13·5 g) into the walls of which were fitted two rectangular fine, stainless-steel wire screens (total surface area c. 7 cm²) through which the respired air passed. The mask was secured behind the comb and upper neck by means of two elastic bands and the end enclosing the beak was made airtight by a rubber flange. Pressure changes within the mask, brought about by airflow through the resistance screens, were detected by a cannula connected to a Grass PT5 manometer the output of which was continuously displayed on a pen-recorder. The mask was calibrated using positive and negative air flows and produced a single, linear relationship between pressure change, and airflow rate, up to 15 1 min⁻¹, which was approximately 2·5 times the maximum minute volume observed during exercise. A parallel output from the airflow channel of the recorder was led into a Grass 7P10 integrator which produced a continuous display of minute volume. Both inspired and expired air flows were summated, so that the final minute volume represented the mean of both respiratory phases. Respiratory frequency was measured directly from the airflow trace and used to estimate the tidal volume. Both volumes were expressed in BTPS conditions assuming a mean body temperature of 42 °C.

The mean environmental vapour pressure was approximately 10 Torr.

Minute volume ⁠, respiratory frequency (f) and tidal volume (V_T) all increased rapidly within the first minute of exercise in control birds then underwent more gradual changes throughout the remaining period (Figs. 1–3). At lower speeds (1·24, 2·15 and 2·9 km h⁻¹) the ventilatory response of the control birds attained a steady state within 5–6 min, following a slight overshoot during the first half of the exercise period. At the higher speeds (3·6 and 4·3 km h⁻¹) and f continued to rise and V_T continued to fall throughout the exercise period, although at a steadily decreasing rate.

Heat-stressed birds displayed a typical thermal polypnea before the start of exercise, consisting of a fourfold increase in ⁠, a sevenfold increase in f and a 40 % reduction in V_T. The panting reflex was immediately suppressed at the start of exercise and f and V_T assumed values much closer to those of the control birds (Figs. 1–4). The initial convergence of respiratory responses in control and heat-stressed birds was only temporary and soon and f began to rise at a faster rate, and V_T to fall at a faster rate, in the heat-stressed birds. The mean ventilatory responses attained in each group between the sixth and tenth minutes of exercise are shown in Fig. 5(b). These values will be referred to as ‘steady-state’ values, although strictly speaking a true steady-state respiratory response was achieved only in lightly exercising control birds. Fig. 5 (b) shows that the steady-state and V_T increased progressively with exercise intensity in control birds and heat-stressed birds although at a given exercise intensity was higher and V_T lower in the latter conditions.

At the termination of exercise, ventilation returned steadily to normal in control birds but heat-stressed birds adopted a vigorous panting response, f increasing beyond end-exercise levels and V_T falling towards pre-exercise levels (Figs. 2, 3).

The air-flow mask was designed to allow full opening of the beak since this was the attitude normally adopted by the running birds. Consequently the dead space of the mask was approximately 5 ml, which is significant compared to an estimated anatomical dead space of 7–8 ml (Kuhlmann & Fedde, 1976). However, comparison of resting ventilation in masked and unmasked birds sitting in a plethysmograph showed that there was little difference at normal temperatures. The recorded changes in f during exercise (Fig. 2) were also comparable to those measured by Brackenbury et al. (1981a). The effects of the mask only became obvious during panting; the rise in f at the end of exercise in heat-stressed birds (Fig. 2) was substantially lower than that observed in the previous study. This is not surprising since during vigorous panting at rest the tidal volume is matched much more closely than normal to the anatomical dead space and is therefore much more sensitive to alterations in the latter. The fully developed panting response of the same birds measured plethysmographically gave tidal volumes as low as 8-10 ml (unpublished observations). The effects of the airflow resistance of the mask, as opposed to the dead space, were assumed to be small ; the maximum pressure changes occurring within the mask were approximately ± 15 Pa, compared to measured changes in air-sac pressure in exercising birds of ± 500 Pa (Brackenbury & Avery, 1980).

Ventilation increases abruptly at the start of exercise in man and dogs and it was originally thought that the rise was activated by sensory feedback from proprioceptors in the muscles, joints and tendons (Dejours, 1964, 1967). However, there is also evidence that the initial ventilatory response may be triggered by an increase in cardiac output and the neurogenic hypothesis has been called into question (Wasserman, Whipp & Castagna, 1974). Steady-state ventilation in man is attained only after several minutes of exercise and is thought to be controlled by humoral factors. During moderate exercise, changes, if any, in blood and blood pH or in CSF H⁺ are insufficient to explain the relatively large ventilatory increases involved (Kao et al. 1965; Wasserman & Whipp, 1975; Wagner, Horvath & Dahmas, 1977) and the primary factor appears to be the rate at which CO₂ is delivered to the lungs (Levine, 1978).

The initial rise in in fowl, although apparent at all levels of exercise in control birds, was masked by the high resting in heat-stressed birds (Fig. 1). The rise is probably more pronounced than appears to be the case in Fig. 1 since the first measurements represent the mean values obtained over the first minute of exercise. The factors which promote the initial ventilatory response almost completely override the thermal polypnea of heat-stressed birds as shown in Figs. 2–4 and their effect is such that and f at the beginning of exercise all increase directly with external work load in both control birds and heat-stressed birds. This point is illustrated in Fig. 5 (a), which shows the minute by minute changes in f and V_T on the same graph. The dotted line in Fig. 5(a) passes approximately through the first minute measurements of each exercise and represents the relationship ml BTPS min⁻¹. Owing to the uncertainties concerning the precise timing of the initial rise in ventilation this relationship must be viewed as tentative pending more detailed experimental measurements but it is interesting that a similar direct relationship between and V_T describes the ventilatory response in man to a variety of respiratory stimuli including hypoxia, hypercapnia, acidaemia and exercise (Hey et al. 1966).

Following its rapid early rise, V_T immediately began to fall in both control and heat-stressed birds whilst f continued to rise (Figs. 2–5a). This alteration of ventilatory pattern is similar to the respiratory response of resting, hyperthermic birds and reflects the influence of rising body temperature on respiratory control. Measurements of changes in blood and blood pH in duck (Bouverot, Hildwein & LeGoff, 1974), fowl (Marder, Arad & Gafni, 1974) and flamingo (Bech, Johansen & Maloiy, 1979) confirm that the great reduction in V_T that occurs during panting in resting birds has the effect of confining most of the excess ventilation to the respiratory dead space, thereby preventing alkalosis. Fig. 5(a) shows how the respiratory pattern during exercise in all of the heat-stressed birds, and in the most actively exercising control birds, consistently tends towards the fully developed panting response of resting birds. The steady-state response (Fig. 5b) represents a compromise between thermoregulatory and metabolic requirements. The progressive rise in steady-state V_T with exercise intensity (Figs. 3, 5b) shows that the respiratory control system is less and less able to maintain a typical polypneic type of breathing in face of increasing metabolic demands, even though the mean body temperature is much higher during strenuous exercise (Brackenbury, Gleeson & Avery, 1981b).

In Fig. 6 the steady-state is plotted against the corresponding steady-state oxygen consumption measured in the same birds by Brackenbury et al. (1981a). In control birds the oxygen extraction, calculated as BTPS/0·21 BTPS, remained close to its resting value of 0·21 during moderate exercise (2–4 times resting ⁠) but fell to 0·17 at the highest work load (6·7 times resting ⁠). In contrast, in heat-stressed birds E rose from a resting value of 0·05 to 0·14 at the highest work load (7·1 times resting ⁠). Consequently, hyperventilation was less in the strenuously exercising than in the lightly exercising birds and the ventilatory equivalent (⁠BTPS/ STPD) decreased from 120 at rest to 38 at the highest work load. This almost exactly parallels the previous finding of a 2·8-fold drop in the measured fractional respiratory evaporative heat loss from 47% in resting, heat-stressed birds to 17% during the most intense exercise (Brackenbury et al. 1981b).

The increase in oxygen extraction of heat-stressed birds at higher work loads was associated with a rise in the fractional lung ventilation as shown by the progressively increasing steady-state V_T and marginally decreasing f (Fig. 5b). This is consistent with the observation that the of interclavicular air-sac gas, which reflects the composition of end-parabronchial gas, increased steadily with running speed, the absolute increase being 5 Torr at 1·24 km h⁻¹ and 14 Torr at 3·6 km h⁻¹ (Brackenbury et al. 1981a). The different ventilatory responses of hyperthermic bird at low and high work loads recalls, at least superficially, the phase I and phase II panting responses shown by resting birds and mammals (Hales, 1974). Phase II panting occurs during severe heat stress and is characterized by large increases in tidal volume and in the fractional lung ventilation, resulting in respiratory alkalosis. Severe heat therefore ultimately leads to a breakdown in the balance between thermoregulatory and respiratory demands on the respiratory system of the resting animal. With reference to exercise hyperthermia in fowl, the restriction of very large increases in ventilation to the dead space at low work loads demonstrates the effectiveness of thermal polypnea in these circumstances, but it does not necessarily follow that the large increase in parabronchial ventilation at high work loads signifies a failure in respiratory control. During heavy exercise in man, hyperventilation produces a fall in alveolar ⁠, which is known to compensate metabolic acidosis (Wasserman et al. 1967). The relationship between and at normal temperatures in running birds (Fig. 6) closely resembles that measured in human subjects, suggesting that hyper-ventilation at high work loads may be governed by metabolic acidosis in both species. The convergence of respiratory responses in fowl at the highest work loads regardless of air temperature (Figs. 5b, 6) also suggests the growing importance assumed by non-thermal factors in respiratory control.

An additional factor which may be considered in the present context concerns the prediction by Scheid (1978) that diffusion through the air capillaries of the para-bronchi might impose a limitation on gas exchange during exercise. If this were the case, lung hyperventilation might be stimulated by the need to increase the gradient between the parabronchial lumen and the periparabronchial blood capillaries. Against this possibility, however, there was no sign of hypoxaemia in either flying pigeon (Butler, West & Jones, 1977) or running duck (Kiley et al. 1979).

Finally, some data are available that permit a preliminary comparison of the ventilatory responses to exercise hyperthermia in flying and running birds. Hudson & Bernstein (1981) reported that ravens flying at 34 °C lost the equivalent of 28% of their total metabolic energy production through respiratory evaporation and slightly greater losses (35%) occurred in budgerigars flying at 36–37 °C (Tucker, 1968). underwent 3-to 4-fold increases compared to flight at 18–20 °C. The metabolic rates of these birds were 4–5 times resting and at a comparable work-load fowl exercising at 35 ± 2 °C evaporated approximately 24% of their energy production but increased by only 80% (Fig. 6). Although this comparison may at first suggest a weaker ventilatory response to hyperthermia in the running bird it is likely that even higher environmental temperatures would have elicited further increments in ventilation and indeed Taylor et al. (1971) found that at comparable work loads in the rhea, the fractional metabolic energy loss through respiratory evaporation increased from 27 % at 25 °C to 50 % at 43 °C, presumably as a result of an increase in ventilation. At the maximum work load the rhea appeared to be utilizing all the spare breathing capacity since the fractional evaporative energy loss fell below 25 % regardless of environmental temperature. Brackenbury et al. (1981b) also observed that at heavy work loads in fowl there was little difference in fractional respiratory evaporative energy losses and this is consistent with the relatively small differences in minute volume indicated in Fig. 6.

This work was supported by the Science and Agricultural Research Councils.