Reflex, inhibitory control of heart rate, exerted by the parasympathetic vagus, is the predominant influence on cardiac output in vertebrates. However, vagal innervation of the heart is not homogeneous as some branches have greater or more complex influences on heart rate than others, and there seems to be particular heterogeneity in the extent to which different branches contribute to the generation of cardiorespiratory interactions (Taylor et al. 1999).

In the elasmobranch fish, Scyliorhinus canicula, a relatively primitive vertebrate, the sympathetic nervous system does not extend forwards to innervate the head and pharynx, so that the heart receives solely parasympathetic innervation via the vagus nerve. However, there is a dual innervation to the heart from branchial cardiac and visceral cardiac branches (Taylor et al. 1977). These may have separate roles, with efferent control exerted chiefly by activity in the branchial cardiac branches (Short et al. 1977). The activity in the branchial cardiac branches can be discriminated into two discrete types: rhythmical bursting units, which show respiration related activity, and sporadic or regular but not rhythmical units. These two types of activity have been shown to have separate origins in the brainstem (Barrett & Taylor, 1985). Electrical stimulation of the peripheral cut end of a branchial cardiac branch with continuous trains of impulses (2Ð4 ms pulses at 50 Hz) caused cardiac slowing or arrest. However, stimulation with short bursts (5Ð10 impulses) at rates between 19 and 43 bursts min^-1 caused the heart to beat at the imposed rate, even when this was faster than the intrinsic rate. This implies that the respiration-related bursting units recorded from the intact branchial cardiac branch could recruit the heart, providing a mechanism for the generation of cardiorespiratory synchrony (Taylor et al. 1999).

While peripheral stimulation of visceral cardiac branches had very little inhibitory action on the heart, central stimulation caused a much greater degree of cardiac slowing than central stimulation of a branchial cardiac branch. Thus it appears that the visceral cardiac branches have a predominantly afferent role in the reflex control of heart rate, indicating that they may innervate receptors in the heart resembling the atrial receptors of mammals.

The vagus nerve in reptiles runs to the heart, trachea, lungs and pulmonary and coronary vasculature. In snakes, peripheral electrical stimulation of the vagus results both in a bradycardia and an increase in pulmonary vascular resistance, because the vagus innervates a sphincter on the pulmonary artery as well as having an inhibitory influence on heart rate (Lillywhite & Donald, 1994). The vagus nerve in the rattlesnake, Crotalus durissus, is asymmetrical in its effects. Peripheral stimulation of the left cervical branch markedly slows the heart, while transection causes cardioacceleration. The right branch is relatively without effect on cardiac chronotropy. The functional and neuroanatomical bases of this marked asymmetry are presently under investigation.