344
G. SHELTON
with increased ventilation was also increased in the presence of carbon
dioxide, compared with a solely hypoxic stimulus ( Saunders, 1962). The
effect of these ventilation and utilization changes on the work of
breathing might reasonably be expected to increase the standard rate
of oxygen consumption. That they did not do so in Beamish’s experiments (1964d) could be explained by his suggestion that carbon dioxide
does not in fact stimulate the truly resting fish to hyperventilate. Certainly many fish become excitable and active if carbon dioxide is introduced into the environment and measurements on resting animals become difficult.
The significance of these responses can best be examined in the light
of some of the physical properties of carbon dioxide. In an important
contribution to the theory of aquatic gas exchange, Rahn (1966) demonstrated that because of the differences in solubilities of oxygen and carbon dioxide in water the change in carbon dioxide tension between
inspired and expired water must necessarily be small (Fig. 21). Dejours
et al. (1968) went on to show that carbon dioxide differences were even
smaller than Rahn had predicted if fish breathed in well-aerated, high
carbonate water, because of the buffering action of the carbonatebicarbonate system. For example, the carbon dioxide tension in water
expired by a goldfish at 25°C was increased by less than 1 mm Hg.
The relationships vary in different types of water, but in all cases the
tensions are low and are not substantially increased by gas exchange
at the gills. Carbon dioxide tensions in the tissues and bloodstream of
fish would also be expected to be low, and this has been repeatedly confirmed by experiment (Lenfant and Johansen, 1966; Holeton and Randall,
1967b; Stevens and Randall, 196%; Baumgarten-Schumann and Piiper,
1968).
Low tensions in the blood are also attributable to the high permeation coefficient of carbon dioxide (see chapter by Randall, “Gas Exchange in Fish,” this volume, Table I ) . The diffusion gradients between
blood and water are very much smaller for carbon dioxide than they are
for oxygen. Indeed, Kylstra et at. (1967) have derived diffusion equations
for a gill model from which they suggest that diffusion dead space for
carbon dioxide may be negligible; that is to say that persisting gradients
of tension within exchange units would be very small, assuming realistic
values for diffusion time and distance. This would not be the case for
oxygen, however, and Kylstra et al. argue that a major difference of this
type would relate the rates of oxygen uptake and carbon dioxide elimination volume in quantitatively different ways. The regulation of ventilation and oxygen uptake by reference to arterial tensions of carbon
dioxide would therefore be extremely unsatisfactory.
G. SHELTON
with increased ventilation was also increased in the presence of carbon
dioxide, compared with a solely hypoxic stimulus ( Saunders, 1962). The
effect of these ventilation and utilization changes on the work of
breathing might reasonably be expected to increase the standard rate
of oxygen consumption. That they did not do so in Beamish’s experiments (1964d) could be explained by his suggestion that carbon dioxide
does not in fact stimulate the truly resting fish to hyperventilate. Certainly many fish become excitable and active if carbon dioxide is introduced into the environment and measurements on resting animals become difficult.
The significance of these responses can best be examined in the light
of some of the physical properties of carbon dioxide. In an important
contribution to the theory of aquatic gas exchange, Rahn (1966) demonstrated that because of the differences in solubilities of oxygen and carbon dioxide in water the change in carbon dioxide tension between
inspired and expired water must necessarily be small (Fig. 21). Dejours
et al. (1968) went on to show that carbon dioxide differences were even
smaller than Rahn had predicted if fish breathed in well-aerated, high
carbonate water, because of the buffering action of the carbonatebicarbonate system. For example, the carbon dioxide tension in water
expired by a goldfish at 25°C was increased by less than 1 mm Hg.
The relationships vary in different types of water, but in all cases the
tensions are low and are not substantially increased by gas exchange
at the gills. Carbon dioxide tensions in the tissues and bloodstream of
fish would also be expected to be low, and this has been repeatedly confirmed by experiment (Lenfant and Johansen, 1966; Holeton and Randall,
1967b; Stevens and Randall, 196%; Baumgarten-Schumann and Piiper,
1968).
Low tensions in the blood are also attributable to the high permeation coefficient of carbon dioxide (see chapter by Randall, “Gas Exchange in Fish,” this volume, Table I ) . The diffusion gradients between
blood and water are very much smaller for carbon dioxide than they are
for oxygen. Indeed, Kylstra et at. (1967) have derived diffusion equations
for a gill model from which they suggest that diffusion dead space for
carbon dioxide may be negligible; that is to say that persisting gradients
of tension within exchange units would be very small, assuming realistic
values for diffusion time and distance. This would not be the case for
oxygen, however, and Kylstra et al. argue that a major difference of this
type would relate the rates of oxygen uptake and carbon dioxide elimination volume in quantitatively different ways. The regulation of ventilation and oxygen uptake by reference to arterial tensions of carbon
dioxide would therefore be extremely unsatisfactory.