Experiments involving supra-maximal electrical stimulation of the vagus have indicated that the stimulation of the peripheral cut ends of the branchial cardiac branches produces a more intense cardio-inhibition than the stimulation of visceral cardiac branches. It is suggested that the visceral cardiac branches may have a mainly sensory function. In no case could cardioacceleration be obtained during vagal stimulation either before or after injection of atropine, and any increases in stroke volume that occurred accompanied reductions in heart rate. This relationship was considered to be a manifestation of Starling’s Law of the heart and it has been concluded that there is no augmentary sympathetic innervation to the dogfish heart. Evidence also indicates that the Starling relationship is responsible for the increase in stroke volume which accompanies the bradycardia during hypoxia. Circulating catecholamines do not appear to be of importance in this response although they are concerned in cardio-vascular regulation during normoxia.
Elasmobranchs are of interest in that they possess two distinct pairs of cardiac branches of the vagus. One arises from the visceral branch (Marshall & Hurst, 1905) and the other from the post-branchial branch of the fourth branchial division of the vagus (Norris & Hughes, 1920). The cardio-inhibitory action of these nerves and their roles in the response to hypoxia have been described previously (Lutz, 1930; Taylor, Short & Butler, 1977).
It has recently been shown that there is an extensive cardio-regulatory sympathetic innervation of the heart in some teleosts (Govyrin & Leonteva, 1965; Otsuka & Tomisawa, 1969; Gannon & Burnstock, 1969), the adrenergic fibres presumably entering with the vagus (Gannon, 1971) as they do in amphibians (Gaskell, 1886; Langley & Orbeli, 1911). In contrast, Gannon, Campbell & Satchell (1972) illustrated a very sparse adrenergic innervation in elasmobranchs, which was confined to the vagus venosus, and they considered that this may account for the slight reduction in o-atrial conduction time on stimulation of the ducti Cuvieri, which was reported by Izquiredo (1930). The balance of evidence would suggest that the sparse adrenergic innervation of the elasmobranch heart is not involved in the control of heart rate (Bottazzi, 1902; Lutz, 1930). It therefore appears that the nervous innervation of the elasmobranch heart is entirely inhibitory and that any increase in heart rate or stroke volume will be dependent upon the release of vagal tone, or upon intrinsic or humoral regulatory mechanisms. At present, there is no clear indication of the relative roles of these regulatory mechanisms in the intact fish.
Recent evidence has suggested that the majority of the vagal activity to the dogfish heart during normoxia is exerted via the branchial cardiac branches, the visceral cardiac branches being of secondary importance in this respect (Taylor et al. 1977). In view of the importance of the two pairs of cardiac vagi for cardiac control in elasmobranchs, the present study was undertaken to determine their function using electrical stimulation techniques. We wished to see if heart rate or stroke volume could be augmented at any intensity or frequency of stimulation, either before or after injection of atropine. The relative importance of circulating catecholamines and Starling’s Law of the heart in determining heart rate and blood flow through the ventral aorta during normoxia and during the response to hypoxia has also been investigated, by using adrenergic receptor blocking agents.
The general anatomical arrangement of the innervation to the heart and gills in the dogfish (Scyliorhinus canicula L.) was described by Taylor et al. (1977). The cardiac vagi of four freshly killed fish were exposed and separated from surrounding connective tissue. They were fixed in situ using Bouin’s fluid and then removed and embedded in paraffin wax. Five µm sections were stained using the Masson’s Trichrome technique (Masson, 1929; Gurr, 1962). Sections were viewed and photographed using a photomicroscope (Leitz, Ortholux II). The number of fibres present in each cardiac nerve was counted from suitable photographs.
Eighty-four dogfish of either sex whose mass ranged between 0.52 and 1.15 kg were obtained from the aquaria of the Marine Biological Association, Plymouth, and were transferred to holding tanks in Birmingham which contained aerated, recirculated and filtered sea water at 13.5±1 °C, where they were allowed to acclimate for at least two weeks prior to experimentation.
The fish were anaesthetized by placing them in aerated sea water containing approximately 0·04 g l−1 of MS 222 (Sandoz Ltd). The actual concentration used varied between animals. The anaesthetized fish were then placed on an operating table in a constant-temperature room held at the acclimation temperature, and artificially irrigated with sea water containing MS 222. Polythene cannulae (Portex Ltd) containing heparinized dogfish saline were inserted into the first right afferent branchial artery, the caudal artery to a level just posterior to the iliac arteries in the dorsal aorta, and into the caudal vein at a level just posterior to the renal portal veins (for injection of drugs). In those fish (41) in which blood flow was measured, a cannulating electro-magnetic flow transducer (Biotronex Ltd.), 1 cm in length and with a bore of approximately 1 mm, was inserted into the ventral aorta between the 2nd and 3rd branchial arteries. This allowed the measurement of blood flow to the ftrst two pairs of afferent branchial arteries (cf. Butler & Taylor, 1975). The pulsatile flow measured by this technique has been termed ‘stroke flow’ and it is taken as an indication of stroke volume. A previous investigation has shown that these values represent approximately 37% of the actual cardiac output and stroke volume respectively, and that the same proportion of blood flows through the flow probe during the bradycardia induced by hypoxia (Taylor et al. 1977). The probe included a catheter for the measurement of the ventral aortic blood pressure, and thus an afferent branchial artery was not cannulated during this procedure.
The vagus nerves of those fish used for the electrical stimulation experiments were exposed on both sides by cutting along the lateral line from a point dorsal to the spiracle to the pectoral girdle. The muscle was incised until the anterior cardinal sinus was exposed. By paring away the connective tissue it was possible to gain access to the posterior branches of the vagus without damaging the sinus (cf. Taylor et al. 1977). The blood loss during this procedure was therefore minimal. The wounds were temporarily sutured with linen thread. In those experiments involving the cardiac chronotropic responses to electrical stimulation of the vagi, the forebrain was separated from the mid-brain and destroyed by extirpation, and the spinal cord was destroyed at a level just posterior to the vagal outflow by pithing from the posterior. This procedure immobilized the animal without affecting any of the cardiovascular responses to vagal stimulation. In six fish the anterior cardinal sinus was exposed and a polythene cannula was pushed through the connective tissue on its dorsal surface and pushed posteriorly along the sinus and down the ductus Cuvieri until its tip lay within the sinus venosus. The cannula was then secured to the muscle, and the muscle and skin were sutured. This procedure was executed with care so that the connective tissue of the anterior cardinal sinus formed a tight fit around the cannula which did not permit haemorrhage or the entry of air into the circulatory system. By this means measurement of central venous pressure could be made via a pressure transducer (S.E. Labs: S.E.M. 4·86).
The fish were then clamped into a Perspex experimental holding tank which contained approximately 8 1 of aerated, recirculated and filtered sea water maintained at 13·5 ± 0·5 °C. They were left undisturbed for 3 h prior to experimentation to allow recovery from the effects of the operation and anaesthetic (Butler & Taylor, 1971; Taylor et al. 1977). The stitches securing the wounds were removed and the muscles retracted to display the vagus, the seawater level being controlled so that it covered the spiracles but left the wounds clear. The animal was found to irrigate normally under these conditions, whilst access to the vagi was obtained.
The visceral cardiac (the cardiac branch arising from the visceral branch of the vagus) or the branchial cardiac (the cardiac branch arising from the post-branchial branch of the branchial division to the last gill arch) branches of the vagus were cleared of the surrounding connective tissue and raised onto silver hook electrodes. It was possible to pick up the visceral cardiac on to the electrodes free of any other nerves. However, the branchial cardiac was picked up at a level central to the division of the pre and post-trematic branches, since access to the branchial cardiac branch itself necessitated the undesirable procedure of opening the anterior cardinal sinus, either the branchial branch of the vagus to the fourth gill arch (the third vagal division) or that to the third gill arch were used in those experiments which involved the central stimulation of the branchial vagi. The positions adopted for the stimulation of the vagus are illustrated diagrammatically in Fig. 1. The nerves were stimulated at intensities ranging from o-i to 60 V, frequencies between 1 and 500 Hz and a pulse width of 1 ms. In some experiments the effect of vagal stimulation after injection of 0·2 mg kg−1 of atropine sulphate (Sigma) was tested. The abilities of each of the cardiac vagi to elicite a bradycardia was investigated at stimulus intensities which produced a maximal effect. It was not practical to use such a procedure during the experiments involving the simultaneous stimulation of two cardiac vagi, or those concerned with the response to changes in stimulus frequency, as a less intense bradycardia was required in order to study these responses. The central cut ends of the vagal branches were also stimulated at voltages that would produce a submaximal effect since higher stimulus voltages were found to disturb the animal. An external switch was set to trigger square wave pulses of predetermined intensity, duration and frequency from a physiological stimulator (Farnell instruments) and also to trigger an event marker on a 4-channel rectilinear pen recorder (Devices: M4). Dorsal and ventral aortic blood pressures were displayed via two pressure transducers (S.E. Labs: S.E.M. 4·86), and blood flow was measured by a Biotronex BL-610 pulsed logic electromagnetic flow meter set to an upper frequency response of 50 Hz. This equipment was calibrated at the end of an experiment with the dogfish’s blood held at the experimental temperature and stroke volume calculated by integrating the flow waveform (Butler & Taylor, 1975). In those experiments where the effects of adrenergic beta receptor blocking agents were tested, the peripheral cut ends of the cardiac vagi were electrically stimulated at a frequency of 50 Hz, a pulse width of 1 ms and at various intensities (up to 1 V) so that a reduction in heart rate of approximately 50% was obtained. This reduction in heart rate was comparable to that induced during hypoxia when the was reduced to approximately 30 mmHg.
It was found that 0·4 mg kg−1 of DL-propranolol HC1 (Sigma) was sufficient to abolish the cardio-vascular effects of 0·1 mg kg−1 of the pure adrenergic beta-receptor stimulating agent DL-isoproterenol sulphate (Sigma). Injection of 0·1 mg kg−1 of arterenol tartrate (Sigma) or epinephrine tartrate (Sigma) in animals in which 0·4 mg kg−1 of Propranolol had previously been injected caused variable circulatory changes which could be abolished by 0·4 mg kg−1 of the alpha receptor blocking agents phentolamine mesylate BP (Rogitine, Ciba) or dihydroergotamine tartrate (Sigma). These doses of alpha and beta adrenergic receptor blocking agents were, therefore, applied as standard and their potency was tested at the end of an experiment by injecting the relevant agonists.
Blood flow along the ventral aorta was monitored during the following experimental situations :
During the response to electrical stimulation of the peripheral cut end of the branchial cardiac branch of the vagus, during normoxia, both before and after injection of Propranolol (ten fish).
During the response to hypoxia both before and after injection of Propranolol (eight fish). In these experiments hypoxia was rapidly induced using a technique that has been described elsewhere (Butler & Taylor, 1971). The was reduced to approximately 30 mmHg within 1 min and then maintained at this level for 10mm.
During the responses to electrical stimulation of the peripheral cut end of the branchial cardiac branch of the vagus during hypoxia, in cardiac vagotomized fish, both before and after injection of Propranolol (four fish).
During the response to a rapidly induced hypoxia lasting appioximately 10 min after injection of both Propranolol and Rogitine (three fish).
In the present report all mean values are expressed plus or minus the standard error of the mean and the number of observations put in parentheses. Differences between means were determined by Student’s t test or, where applicable, by the paired comparisons technique (Bailey, 1959). The term ‘significant’refers to the 95 % level °f confidence (P < 0·05).
The branchial cardiac vagi contain approximately 420 myelinated fibres, whereas the visceral cardiac vagi contain approximately 300 myelinated fibres (Fig. 2). There was no appreciable variation of the number of fibres in the cardiac branches of different animals. No non-myelinated fibres could be observed and no differences between the left and right sides of the fish were apparent.
The cardio-inhibitory response to electrically stimulating the peripheral cut end of the vagus
This series of experiments was performed on 45 dogfish of mass 0·74 ±0·02 kg. During normoxia (= 147 ± 3 (5)) the mean was 88 ± 2 (5) mmHg, mean heart rate was 29 ± 2 (45) beats min−1, and mean ventral and dorsal aortic pressures were 31 ±1 (38) mmHg and 25 ±1 (35) mmHg respectively. Cardio-inhibitory responses to vagal stimulation could be obtained from fish for up to 3 days without any appreciable change in the levels of the measured variables.
The peripheral stimulation of the visceral cardiac branches of the vagus produced inhibition in only 42 % of experiments (whatever the intensity of stimulation) although ineffective nerves were often capable of producing cardio-inhibition if they were stimulated centrally, when one or both of the branchial cardiac vagi were still intact. This would, therefore, indicate that the absence of response was due to the nerve being ineffective centrifugally rather than there being a deterioration in its physiological condition. Ineffective visceral cardiac vagi were found on either the left or right side of the animal, or occasionally on both sides. The peripheral stimulation of the branchial cardiac branch of the vagus always produced a profound cardioinhibition and no differences between the left or right sides of the fish were apparent. Electrically stimulating the cardiac vagi, at voltages that were above threshold for all fibres, revealed that the branchial cardiac branches were able to cause almost complete cardiac inhibition (92 ±2 (13)% reduction in heart rate), whereas the visceral cardiac branches were only able to cause a 50 ± 19 (10) % reduction in rate (Fig. 3).
In experiments that were undertaken on 20 animals, the cardiac nerves were stimulated at frequencies ranging from 1 to 500 Hz and it was found that the degree of cardio-inhibition was dependent upon the stimulus frequency. The degree of inhibition was optimal when the cardiac nerves were stimulated at frequencies between 25 and 50 Hz (Fig. 4). There appeared to be no appreciable differences between the spectrum of frequency responses of the branchial cardiac (n = 18) and visceral (n = 6) vagi, although the mean response at 10 Hz was significantly higher for the branchial cardiac vagi (P < 0·05). In all other experiments 50 Hz was adopted as the routine frequency of stimulation, except when looking for augmentary Effects at low stimulation frequencies.
Simultaneous stimulation of the peripheral cut ends of the cardiac vagi
In 24 experiments, pairs of cardiac nerves were individually stimulated at submaximal voltages both before and after simultaneous stimulation so that it was possible to compensate for any change in the intensity of cardio-inhibition that occurred whilst the cardiac vagi were raised on to the electrodes. Fig. 5 shows a result obtained from an individual fish and suggests that the effect of stimulating the two nerves together was greater than simple summation of their individual effects. In order to obtain mean values for the effect of stimulating various paired combinations of the cardiac nerves, the effect upon heart rate of stimulating any one nerve was obtained by averaging the effects both before and after combined stimulation and compared with the effect of stimulating the two nerves together. On average, the bradycardia produced by simultaneously stimulating any two cardiac vagi was little more than a summation of their individual effects (Fig. 6).
The cardiac effects of centrally stimulating the vagal branches
The electrical stimulation of the central cut end of the branchial cardiac, visceral cardiac or branchial branches of the vagus (89 experiments on 26 animals) produced a cardio-inhibition when all other cardiac branches were intact. No differences were apparent in the capabilities of the branchial, branchial cardiac or visceral cardiac branches of the vagus, with respect to their ability to elicit a ‘reflex’bradycardia by the electrical stimulation of cut central ends.
Sequential sectioning of the cardiac efferent branches during central stimulation revealed that the cardio-inhibitory response was both ipsilateral and contralateral, and that the visceral cardiac was less capable of relaying this ‘reflex’motor activity to the heart than the branchial cardiac branches. Not only was the intensity of cardioinhibition reduced when the ‘reflex’information was conveyed to the heart via the visceral cardiac, as opposed to the branchial cardiac vagi, but the former were often found to be ineffective in producing any cardio-inhibition at all. Cardiac nerves, which did not convey efferent activity to the heart in response to central stimulation of other nerve tracts were tested to see if they were effective in producing cardioinhibition by peripheral stimulation. It has been mentioned that only 42% of the visceral cardiac vagi were effective in producing cardio-inhibition when stimulated peripherally, and in these experiments it was found that only a third of these effective visceral cardiac branches were capable of inhibiting the heart in response to central stimulation of other nerve tracts, as opposed to 30 of the 31 branchial cardiac branches that were tested.
The absence of cardio-acceleratory responses to vagal stimulation
Neither the peripheral stimulation of the cardiac branches of the vagus nor the central stimulation of the branchial or cardiac branches resulted in cardio-acceleration at any of the frequencies (1-500 Hz) or intensities (0·1-6·0 V) of stimulation investigated (105 experiments on 69 animals). All the cardio-inhibitory responses to stimulation of these nerve tracts were abolished by injection of 0·2 mg kg−1 of atropine sulphate into the caudal vein.
A bradycardia elicited by either vagal stimulation or acetylcholine injection was not followed by a post-inhibitory tachycardia in animals with normal blood pressures and heart rates, although it was often observed in animals that had deteriorated and displayed extremely low heart rates (9 beats min−1) and mean blood pressures (17 mmHg in ventral aorta).
Changes in cardiac stroke flow occurring during peripheral stimulation of the cardiac vagi
In nine animals the stroke flow to the first two pairs of afferent branchial arteries was measured during the electrical stimulation of the peripheral cut ends of the cardiac branches of the vagus. It may be seen from Fig. 7 that the reduction in heart rate is associated with an increased stroke flow. This increase was not further potentiated by stimulation at different frequencies or intensities. The increases in stroke flow that were observed were purely related to a reduction in heart rate, and there were no differences between the visceral cardiac and branchial cardiac vagi in this respect. Injection of 0·2 mg kg−1 of atropine sulphate abolished the bradycardia incited by vagal stimulation, and the associated increase in stroke volume. No further effects could be produced at any of the intensities (0·1-6·0 V) or frequencies (1-500 Hz) of stimulation investigated (five animals).
The heart rate, stroke flow and total flow changes that occur during reductions in heart rate of approximately 30% and 50%, elicited by electrical stimulation of the vagus, are presented in Table 1. A reduction in heart rate of 28·8 ± 4·2 (9)% resulted in a 33·0 ± 10·6 (9)% increase in stroke flow and as a consequence no significant reduction in total flow occurred. However, a 52·2 + 5·3 (9)% reduction in heart rate resulted in a 38·6 ± 8·8 (9)% increase in stroke flow. This change in stroke flow did not completely compensate for the induced bradycardia so that a significant reduction in the total flow occurred. In six animals the central venous pressure was measured and was found to be — 3·4 mmHg at rest and became more positive (— 2·1 mmHg) during a 50% reduction in heart rate induced by either electrical stimulation of the vagus or by hypoxia (Fig. 7).
The effects of Propranolol on cardiac stroke flow
These experiments were performed on dogfish of mass 0·75 ± 0·04 (25) kg and the mass of the dogfish was not significantly different between experimental categories. In control fish (those used for the hypoxia experiments) the mean normoxic values of ventral and dorsal aortic blood pressure were 29 ± 2 (8) and 24 ± 1 (7) mmHg respectively. The corresponding mean values for other experimental categories did not differ from these values by more than 8 mmHg (cf. Short, 1976). However, significant differences occurred in the levels of the measured variables between experimental categories, either as a result of vagotomy or injection of adrenergic receptor blocking agents (cf. Fig. 8). In general, the vagotomized animals had an increased heart rate and blood pressures and a reduced stroke flow (cf. Short, 1976). The results obtained from 17 animals in this series of experiments have indicated that there is no significant change in heart rate after injection of Propranolol. However, it is evident that this adrenergic β-receptor blocking agent severely effects the circulatory system, since significant reductions occur in stroke flow (18±4%), total flow (2215%), mean ventral aortic pressure (12 ±3%) and mean dorsal aortic blood pressure (19 + 4%). There was an increase in stroke flow during the bradycardia induced by electric stimulation of the peripheral cut end of the vagus during normoxia (Fig. 7), and this stroke flow response was not affected by injection of Propranolol (Fig. 8). However, injection of Propranolol abolished the increased stroke flow that occurred in association with the bradycardia induced by hypoxia (Fig. 8), and this response was not alleviated by injection of an a-receptor blocking agent. Propranolol also had similar action on the response to electrically stimulating the peripheral cut end of the cardiac vagus during hypoxia (in vagotomized animals).
During electrical stimulation of the peripheral cut end of the vagus, total blood flow to the first two pairs of gills was approximately 6·0 ml min−1 kg−1 during both normoxia and hypoxia, whereas the mean drop in blood pressure across the gills was, on average, 7-4 mmHg during normoxia and 3·2 mmHg during hypoxia. This indicates that the branchial blood vessels dilated during hypoxia and, using the conventional calculation for vascular resistance (see Butler & Taylor 1975), that resistance in the branchial vessels approximately halved. However, following the injection of Propranolol, hypoxia appeared to cause constriction in the branchial blood vessels, for although total flow dropped to 2 ml min−1 kg−1, the pressure difference across the gills was 7·4 mmHg giving a calculated resistance in the branchial vasculature approximately 10 x greater than during hypoxia before injection of Propranolol.
The vagal control of heart rate
Selective cardiac nerve transection techniques have previously been employed to demonstrate that the branchial cardiac vagi convey the majority of the tonic inhibitory activity to the heart during normoxia and are responsible for the greater part of the reflex bradycardia initiated by hypoxia. The visceral cardiac nerves appeared to be of secondary importance in both these respects (Taylor et al. 1977). These observations have been extended by this investigation in that the branchial cardiac vagi have been shown to have a greater capacity to inhibit the heart, when stimulated electrically, than the visceral cardiac vagi, and furthermore, the latter are often ineffective in producing a cardio-inhibition. However, no difference in the abilities of the cardiac nerves to convey afferent activity could be detected. Evidence therefore suggests that although the branchial cardiac vagi have both sensory and motor capabilities the visceral cardiac vagi may have a mainly sensory function. It is possible that a proportion of the motor fibres contained within the visceral cardiac vagi may innervate regions other than the cardiac pacemaker and this could, to some extent, explain the different cardiac chronotropic abilities of the cardiac vagi. In this connexion it is also interesting that the visceral cardiac vagi contain substantially fewer fibres than the branchial cardiac vagi. The results presented here indicate that the effects of the cardiac branches of the vagus interact at the myocardium. It is therefore possible that the visceral cardiac vagi may modulate the effects of the branchial cardiac vagi, but, since the former are often ineffective in producing cardioinhibition, their motor effects are probably only of secondary importance.
Recent investigations have shown that the afferent activity in cranial nerves IX, X, V and VII is of importance in generating the inhibitory vagal tone upon the heart (Butler, Taylor & Short 1977). In the present investigation it was evident that the electrical stimulation of the central cut ends of the branchial branches of the vagus initiated cardio-inhibition, and that this response was both ipsilateral and contralateral. In addition, it was apparent that the branchial cardiac vagi were more effective in conveying inhibitory activity to the heart when stimulated at the central cut end of the cardiac or branchial branches of the vagus.
The central stimulation of the branchial branches of the vagus always produced cardio-inhibition, whatever the stimulus intensity or frequency. These results are in direct contrast with those obtained in a number of teleosts (Kulaev, 1957, 1958) and elasmobranchs (Rodionov, 1959), where although strong stimuli caused cardioinhibition weak stimuli were reported to cause cardio-acceleration. Similar investigations on fish (Cobb & Santer, 1973) and other vertebrates (Bulbring & Burns, 1949; Marshal & Vaughan Williams, 1956; Burn & Rand, 1957; Jensen, 1958; Misu & Kirpekar, 1968) lead to the suggestion that reliable cardio-acceleratory responses to vagal stimulation or acetylcholine application can be obtained in preparations which had been allowed to deteriorate. In general, the heart was exposed and beating at a low rate or had become quiescent. This investigation has indicated that these responses do not occur in dogfish when the heart is not exposed and the heart rate and blood pressures are normal. The present investigation on intact dogfish was unable to demonstrate a tachycardia on recovery from vagal stimulation or administration of acetylcholine unless the animal was allowed to deteriorate until the heart rate was low. Therefore this phenomena is probably also dependent upon experimental conditions, and this may perhaps explain why some authors have reported a recovery tachycardia (Fange & Ostlund, 1954; Cobb & Santer, 1973) and others have noted its absence (McWilliam, 1885; McKay, 1931; Randal, 1966).
Control of cardiac stroke flow
The cardiac adrenergic innervation of teleosts is thought to reach the heart via the vagus (see introduction for references). In support of previous investigations (Young, 1931, 1933) it is thought unlikely that adrenergic fibres pass within the cardiac vagus of elasmobranchs since no non-myelinated fibres could be observed in sections of the cardiac nerves. There is physiological evidence that adrenergic nerves are absent in the vagus, since no heart rate increases can be initiated by electrical stimulation of this nerve (Bottazzi, 1902; Lutz, 1930) and such a conclusion has been borne out by the present investigation. In addition, however, it has been shown that increases in stroke flow cannot be initiated by vagal stimulation at any intensity or frequency of stimulation after administration of atropine. The only increases in stroke flow that occurred in normal circumstances were directly related to reductions in heart rate and were, therefore, considered to be a manifestation of Starling’s Law of the heart.
The results indicate that moderate reductions in heart rate do not cause reductions in cardiac output. In fact, three fish displayed a mean increase in total flow of 23 % during a reduction in heart rate of 25 % and thus, relatively small reductions in heart rate may result in an increase in cardiac output. It is possible that this mechanism may explain the increase in cardiac output that was associated with small (but nonsignificant) reductions in heart rate during the responses of S. canicula to hypoxia at 7 °C (Butler & Taylor, 1975).
The effects of Propranolol on the circulatory system may be due to the removal of the effects of catecholamines or to local anaesthetic properties of the drug (Eliash & Weinstock, 1971 ; Lee et al. 1975). However, moderately high concentrations of circulating catecholamines have been demonstrated in S. canicula (Mazeaud, 1969; Butler, Taylor, Capra & Davidson, 1978) and it seems likely that they will have important effects on the circulatory system. Recent investigations on Squalus acanthias have shown that Propranolol blocks the dilator tone of catecholamines on the gill vasculature and potentiates a vasco-constrictor response (Capra & Satchell, 1974; Capra, 1975). In addition, it has been shown that Propranolol blocks the positive inotropic responses to catecholamines in the isolated heart (Capra, 1975). The present results, obtained by injection of Propranolol, indicate that blood catecholamines are exerting a dilator tone on the branchial vasculature and are augmenting stroke volume during normoxia. The former of these is not thought to be due to side-effects of Propranolol, since large concentrations of this drug result in a decreased peripheral resistance in mammals (Lee et al. 1975). Our results also suggest that circulating catecholamines have an increased dilator effect on the branchial blood vessels during hypoxia which masks a potential vasoconstriction (cf. Satchell 1962).
An increased stroke flow is associated with the cardio-inhibition induced by electrical stimulation of the peripheral cut end of the vagus during normoxia, and it has been shown that this relationship is unaffected by Propranolol (cf. Butler & Taylor, 1975; Taylor et al. 1977). The increase in stroke volume that is associated with the bradycardia induced by hypoxia is abolished by Propranolol, thus suggesting that circulating catecholamines may be of importance (cf. Butler & Taylor, 1975). In contradiction, cardiac vagotomy abolishes both the bradycardia and the rise in stroke volume during hypoxia (Taylor et al. 1977) and no adrenergic fibres can be demonstrated in these cardiac branches. These results may seem paradoxical. However, it has been shown that hypoxia alone was not responsible for the discrepancy in the results, as stroke flow increases during the bradycardia induced by electrically stimulating the cardiac vagi during hypoxia, in vagotomized fish. This response could, however, be abolished by Propranolol and it would seem that the lack of a rise in stroke volume during hypoxia results from the combined depressant effects of hypoxia and Propranolol on myocardial contractility. Indeed, Propranolol has been demonstrated to cause direct myocardial depression in mammals (Lee et al. 1975) and may reduce myocardial blood flow (Drake, 1976). It is certainly possible that the latter may have more severe consequences during hypoxia. In addition, no significant increases in stroke volume occurred during hypoxia after both alpha and beta adren - ergic receptor blockade. It seems therefore that the lack of an increased stroke flow during hypoxia after injection of Propranolol was not due to a potentiation of alpha adrenergic receptor stimulation.
The evidence indicates that although circulating catecholamines may be of importance in maintaining cardiac activity, vascular resistance and blood pressures during normoxia, they appear to be of little importance in increasing the cardiac stroke volume during the response to hypoxia. The increase in cardiac stroke flow that occurs is a manifestation of Starling’s Law of the heart and this relationship is such that no significant reduction in cardiac output occurs during moderate reductions in heart rate.
The authors wish to thank the Science Research Council for financial support.