8. THE REGULATION OF BREATHING
341
oxygen concentrations in the inspired water. De Kock (1963) has described “end buds” in the buccal cavity and on the gills of trout and has
suggested that they have a chemoreceptor function. However, control
based solely on detection of environmental oxygen would entail a precise
relationship between different concentrations of oxygen and the corresponding ventilation volumes with no feedback adjustment. Even supposing that this relationship could be modified appropriately at different
levels of exercise, it would be unlikely that such a system could give
stable control of oxygen supply under all conditions, especially with
varying gill geometry and perfusion.
It is probable that such stability can only be conferred by a control
system with receptors in the bloodstream, and the situation in mammals
argues strongly for some such system. If the receptors were located in
the vessels of the gills, control would be susceptible to local variations
in geometry and perfusion rates. Integration of activity from receptors
throughout the gills would overcome this difficulty. More satisfactorily,
the receptors could be situated downstream from the gills so as to
measure oxygen concentration in mixed arterial blood. The pseudobranch
has frequently been considered as a region where receptors might occur.
Indeed, Laurent (1967) demonstrated a sensitivity to changes in
oxygen and carbon dioxide concentrations by recording from the
nerve of an isolated and perfused tench pseudobranch. This does not
mean that all the receptors need be located in a single site. Shelton (in
Hughes and Shelton, 1962) showed that bilateral section of the IXth
cranial nerves did not abolish the ventilatory response to oxygen lack
although it was considerably modified. This experiment was repeated
with section of the Xth cranial nerves and IXth and Xth nerves together,
and in all cases the fish was able to respond to low oxygen levels. There
is thus clear evidence that oxygen receptors exists outside the areas
innervated by the IXth and Xth cranial nerves.
A suggestion that the receptor system could be located in the venous
blood system was made by Taylor et al. (1968), who based their work
on an analysis of the equilibria between environment and blood under
a variety of conditions. These workers described a digital computer
simulation of the gas exchange system in trout, using the results of
Randall and his collaborators. Equations were derived to describe oxygen
transport and storage in all parts of the system, together with three
feedback equations which controlled ventilation volume, heart output,
and gill area. All the equations could be simultaneously solved for blood
and environmental gas tensions, ventilation volume, and heart output,
by using different input driving values for metabolic rate or inspired
oxygen tension. Realistic simulations could not be achieved if arterial
341
oxygen concentrations in the inspired water. De Kock (1963) has described “end buds” in the buccal cavity and on the gills of trout and has
suggested that they have a chemoreceptor function. However, control
based solely on detection of environmental oxygen would entail a precise
relationship between different concentrations of oxygen and the corresponding ventilation volumes with no feedback adjustment. Even supposing that this relationship could be modified appropriately at different
levels of exercise, it would be unlikely that such a system could give
stable control of oxygen supply under all conditions, especially with
varying gill geometry and perfusion.
It is probable that such stability can only be conferred by a control
system with receptors in the bloodstream, and the situation in mammals
argues strongly for some such system. If the receptors were located in
the vessels of the gills, control would be susceptible to local variations
in geometry and perfusion rates. Integration of activity from receptors
throughout the gills would overcome this difficulty. More satisfactorily,
the receptors could be situated downstream from the gills so as to
measure oxygen concentration in mixed arterial blood. The pseudobranch
has frequently been considered as a region where receptors might occur.
Indeed, Laurent (1967) demonstrated a sensitivity to changes in
oxygen and carbon dioxide concentrations by recording from the
nerve of an isolated and perfused tench pseudobranch. This does not
mean that all the receptors need be located in a single site. Shelton (in
Hughes and Shelton, 1962) showed that bilateral section of the IXth
cranial nerves did not abolish the ventilatory response to oxygen lack
although it was considerably modified. This experiment was repeated
with section of the Xth cranial nerves and IXth and Xth nerves together,
and in all cases the fish was able to respond to low oxygen levels. There
is thus clear evidence that oxygen receptors exists outside the areas
innervated by the IXth and Xth cranial nerves.
A suggestion that the receptor system could be located in the venous
blood system was made by Taylor et al. (1968), who based their work
on an analysis of the equilibria between environment and blood under
a variety of conditions. These workers described a digital computer
simulation of the gas exchange system in trout, using the results of
Randall and his collaborators. Equations were derived to describe oxygen
transport and storage in all parts of the system, together with three
feedback equations which controlled ventilation volume, heart output,
and gill area. All the equations could be simultaneously solved for blood
and environmental gas tensions, ventilation volume, and heart output,
by using different input driving values for metabolic rate or inspired
oxygen tension. Realistic simulations could not be achieved if arterial