Blood Flow and Pressure Changes in Exercising Octopuses (Octopus Vulgaris)

Most cephalopods are large by invertebrate standards, and some are large even in comparison with fish. All are predators, and many can move rapidly by jet propulsion. The combination of large size and high metabolic rate requires a well-developed circulatory system. However, the basic body plan of molluscs is far from ideal as a basis for the evolution of such a system for the following reasons. (1) Movement is dependent upon a hydrostatic skeleton; muscle can only be extended by pressure from antagonists, so that blood flow through the muscles cannot be helped by locomotion. (2) The gills are upstream of the heart and the high fate of oxygen uptake in the cephalopods necessitates fine vessels in the gills. This creates a second capillary bed following the main systemic peripheral resistance. (3) Cephalopod gills and hearts are inside the mantle cavity. Thus, if the mantle cavity is used to generate a jet, there will be large pressures resisting the venous return to the hearts from the head and arms. (4) Finally, most molluscs and all cephalopods use a high molecular weight blood pigment, haemocyanin, in solution though never, for reasons of viscosity, present in concentrations that result in an O₂-carrying capacity much in excess of 4vols%.

Despite these apparently considerable difficulties, the cardiovascular systems of squids such as Illex (Webber & O’Dor, 1985) and Lolliguncula (M. J. Wells, R. T. Hanlon, P. di Marco & P. G. Lee, in preparation) manage to deliver oxygen at rates in excess of 11O₂kg⁻¹ h⁻¹; rates fully comparable with those of mammals, when differences in temperature are taken into account.

Squids are, however, difficult to maintain in good physiological condition in captivity. Octopus vulgaris, although not as active metabolically, is easier to handle, comes in large sizes, has major vessels which are accessible to surgery and can be readily induced to walk (and sometimes swim) in a confined space without damaging itself; it is therefore more appropriate for an initial study of the responses of a cephalopod circulatory system to stress. The following account describes the blood pressure and flow changes that occur when an octopus is active. A parallel study of the blood oxygen and carbon dioxide contents in animals exercising under the same conditions has been reported by Houlihan et al. (1986).

Octopus vulgaris Cuvier was used in this investigation. Seven of the eight animals were mature males, between 785 and 2516g, the eighth a ripe female of 1130g. They were caught in trawls or by scuba divers. The animals were kept and fed on crabs at the Laboratoire Arago, Banyuls, France from July to September 1983. Exercise runs were made in a plastic paddling pool, l·8×l·2×0·2m deep. In most cases spontaneous movement around the pool occurred, and this was the most satisfactory way of obtaining records, since attempts to force the animals to move by touching them were liable to result in their adopting a defensive position, suckers uppermost, with intermittent cessation of ventilation, or jetting powerfully at the intruding hand, all responses that interfere with the regular heartbeat (see below). A compromise was to induce the animals to move by wading barefoot in the tank, or by presenting a white plastic bucket lid with a black ‘eye’ painted on it.

Probes to monitor pressure or flow were placed in the dorsal aorta. For pressure, Radiospares RS303-343 pressure transducers were modified for use underwater by filling the air-space below the piezoelectric element with liquid paraffin, open to the outside through a short length of capillary. The pressure side was connected to the aorta through a stainless-steel T-piece, as in the piped pressure-recording system used by Wells (1979). Back-filling with paraffin might be expected to slow the response of the transducers, but the events of interest, at around 1 Hz, are far too slow for this to be significant; the shapes of the pressure pulses obtained were the same as those reported previously using conventional SE4-82 or Statham transducers. The electrical output from the RS303-343 transducers was fed to an oscilloscope and, via an amplifier, to a modified Cambridge type 72125 cardiograph.

Flow was measured using FW series permanent magnet transducers (In vivo Metric Systems, Healdsburg, California), again connected to an oscilloscope display and cardiograph.

Calibration of the pressure transducers was done manometrically. Flow transducers were calibrated as described in Wells & Wells (1986) who found that mean pulse area × frequency showed a good correlation with measured flow over periods of a minute or so (r = 0·92 for the smaller bore of the two probes used and 0·93 for the larger). In the present series it was further noted that amplitude × frequency correlates well with the total area over such series of pulses (r = 0·98 and 0·95 for the two probes, see Fig. 1). This simpler way of estimating flow, which has the advantage that it can be applied to records made at slow speeds where it is difficult to measure area, was used in the analysis of most of the data presented here.

Two of the largest animals used had both flow and pressure monitors inserted in series into their aortas, with the pressure probe immediately downstream of the flow probe. At rest, flow and pressure both showed a rapid rise, with flow peaking before pressure, and then a two-stage fall, with the rate of decline slowing after the closure of the exit valve from the heart at the end of systole (Fig. 2A). Shadwick, Gosline & Milsom (1987) have made a more detailed analysis of this relationship in the large O. dofleini; the aorta acts as an elastic reservoir, so that flow continues between beats, ceasing only momentarily in diastole. When the animal shown in Fig. 2A was moving spontaneously over the ground at around l0cms⁻¹, walking and sometimes jetting, but not strongly, flow along the aorta averaged around 92 ml min⁻¹, with pressures of around 3 kPa in diastole and 5 kPa in systole and occasional peaks of up to 7 k Pa (Fig. 2B). At rest prior to movement, the same animal pumped just over 67 ml min⁻¹ along the aorta.

Octopus can walk or swim and often uses a combination of crawling and jet propulsion. The speed at which it moves can vary from a slow exploratory walk forwards, at around Sems⁻¹, to a jet-propelled escape, backwards at 1ms⁻¹ or more. Maldonado (1964) recorded similar speeds for jet-propelled forward attacks on crabs. The present report covers movement at the slow end of the range. Sustained jet propulsion is not seen in Octopus in aquaria, and occasional observation of animals in the sea (by M. J. Wells) suggests that they rarely, if ever, travel for more than a few metres in this manner (i.e. backwards, with the arms spread out in a flat horizontal plane, at a more leisurely pace than seen in escape or attack).

Fig. 3 shows a sequence from a slowly moving octopus, in which aortic blood pressure pulses rose from 3–4 kPa at rest to 6·5–8 kPa during the movement, returning to 3·5–5 kPa as the animal settled down again. The heartbeat frequency remained steady at 0·70 Hz throughout.

More vigorous responses to stimulation interfered with and sometimes stopped the pressure pulse. Fig. 4 illustrates two such sequences when the animal showed the ‘dymantic’ response (i.e. it flattened out, expanded the interbrachial web and paled, with dark rings around the eyes and at the edge of the web) and discharged ink through the funnel. The octopus then moved away, walking and jetting, settling down again only after removal of the stimulus. Most of these events are traceable from the pressure records, with very high pressure peaks associated with ink discharge (Fig. 4). These very high peaks probably reflect mantle pressures rather than the performance of the systemic heart. Vigorous jet production during escape responses can produce abrupt pressures of 20–40 kPa (Trueman & Packard, 1968). It is assumed that pressures of this magnitude were produced during ink ejection and were too fast for the fluid-damped RS303-343 probe to compensate fully for the ambient pressure of the mantle cavity.

Cessation of movement was followed by an increase in mean blood pressure and pulse amplitude, from 2–3 kPa at rest to 4–7 k Pa at peak. The heartbeat frequency fell from an initial 0·73 Hz to 0·63 Hz during the ‘recovery’ period, returning to 0·73 Hz 90s after the movement. It seems reasonable to suppose that the animal was paying off an oxygen debt (Wells, O’Dor, Mangold & Wells, 1983; Wells & Wells, 1984) that had been accumulated during the period of exercise when the pulse was irregular or stopped.

Table 1 summarizes the effects of exercise on pressure pulses for the five animals used in similar experiments.

Flow changes parallel pressure changes. Fig. 5 summarizes flow changes during a 20-min movement sequence, in which the octopus was continuously active. During this time the animal moved 64m at 5·3cms⁻¹, a speed that an octopus might well maintain in the sea, moving over rocks and groping for food. Flow doubled over the first 4 min of exercise, remained steady at around 100 ml min⁻¹ and then declined to ‘resting’ values over the 5 min following the end of the exercise period. This sequence closely parallels the rise and fall in oxygen uptake noted for Octopus moving in an exercise wheel (Wells et al. 1983).

In the experiment illustrated in Fig. 5 the animal moved slowly and evidently remained in oxygen balance, thus accumulating a small oxygen debt at the start of the run that was paid off within a few minutes of the end. In other cases, more vigorous movement was followed by a prolonged enhancement of flow, during which rates could transitorily exceed those achieved during exercise (Table 2).

Blood flow recordings also confirmed that vigorous jetting was incompatible with normal blood flow, particularly when powerful jets occurred in the course of a prolonged period of movement, for flow ceased on these occasions (Fig. 6).

Blood pressure, blood flow and heartbeat frequency are not, of course, independent parameters and their interactions (summarized in Fig. 7) allow us to assess other aspects of cardiovascular performance. Thus, changes in the ratio relating the pressure required to drive blood through the systemic vascular bed (in N m⁻²) to the mean stroke volume (in m³) reflect alterations in resistance (Shadwick et al. 1987). This might be expected to increase when the animal moves, for the vascular system is part of the hydrostatic skeleton. The mean aortic pressure was calculated as ⅓ (systolic + 2 × diastolic pressure); the pressure in the lateral venae cavae at around 0·15 kPa (Wells & Smith, 1987) was considered negligible for these purposes. The product of stroke volume (m³) and the pressure difference across the systemic heart [N m⁻²; which is very small at 0·2 kPa, see Wells & Smith (1987)] yields stroke work. The power output per gram (O. vulgaris hearts are close to 1·0 g kg⁻¹ body mass, Wells & Smith, 1987) can thus be derived from the stroke work and the heart rate.

There are several points of interest. Power output increases predictably, almost entirely due to the increase in stroke work, since the heartbeat rate alters so little. More surprisingly, the ratio of mean aortic pressure to stroke volume scarcely changes, which implies that resistance remains almost constant as the animal changes from rest to exercise. This is not what would be expected from an animal that is obliged to extend its muscles under pressure. Somehow vascular changes are taking place which facilitate the passage of blood during exercise. Such changes might include an increase in active pumping by the arteries and veins (many of which are muscular and innervated and show active peristalsis) and/or the active regulation of arterial capacitance.

It should be borne in mind that the changes summarized in Fig. 7 arise from experiments involving only rather gentle exercise. The three-fold increase in stroke work found here is certainly not a maximum. Smith (1985), using data from experiments involving more rapid walking in an exercise wheel (Wells et al. 1983), calculated a five-fold increase, and the animals were able to maintain this performance for at least 2h.

Previous reports on the effect of exercise in Octopus have been hampered by pressure recording systems that included a pressure pipe from the animal to a transducer. It was possible to follow the development of pressure during slow movements by the very large O. dofleini (Johansen & Martin, 1962) but for O. vulgaris the evidence that pressure rises was based on recordings made immediately after the animal had stopped moving (Wells, 1979). Records made during any rapid movement were marred by the vibration produced when towing the pressure pipe through the water.

An implanted pressure transducer, of the sort used here, solves this problem. The results were more or less what was expected. Pressure is doubled by even quite gentle exercise, remains high while the animal is active, and often for a period afterwards. Frequency, in contrast, altered little. Often the increased pulse that accompanies a rise in mean pressure was actually slower than at rest (Tables 1,2). This finding contrasts with Johansen & Martin’s (1962) results with O. dofleini, where the onset of exercise was accompanied by a 30% increase in heartbeat frequency (from 0·1 to 0·13 Hz in this very large cold-water animal).

Flow, which has not been examined before in any active cephalopod, also increases during gentle exercise. In the present experiments it doubled, which is compatible with the 2·4-fold rise in oxygen uptake observed in an exercise wheel, where the animals ran more rapidly (Wellset al. 1983). At the onset of exercise, there is a shortterm fall in arterial lasting for some tens of seconds (Smith, Duthie, Wells & Houlihan, 1985: in fig. 4 of this paper the x-axis is an order of magnitude too large) followed by a return to resting values. is very low, even at rest, so there is no scope for a significant expansion of the difference during exercise (Houlihan et al. 1986). Since heartbeat frequency remains unchanged, the extra oxygen delivery in exercise depends almost entirely upon increases in the stroke volume of the heart (Smith, 1985; Wells, 1979).

The experiments reported here included continuous activity lasting for 20 min at a time. Records of the track followed as animals were induced to move around (N = 4, including X47, summarized in Fig. 5) showed average speeds of 0·15-0·19 km h⁻¹ (2·5—3·2m min⁻¹). This is half the speed (0·34 kmh⁻¹) that can be maintained for periods of up to 2h during forced exercise in a wheel (Wells et al. 1983; Wells & Wells, 1984). This would suggest that the exercise taken in the present experiments was by no means extending the animals to their limits. However, it should be borne in mind that five out of the six octopuses used here for flow experiments were larger (in four cases considerably larger) than any of the dozen octopuses run in the wheel. The latter showed a factorial aerobic scope (active ÷ routine O₂ uptake) averaging 2·38 (size range 420—1040g). The factorial scope of the four largest animals used for the experiments reported here averaged 1-45 (size range 1735-2516 g) indicating that scope may decline with increasing size.

If the octopus needs to move at higher speeds, jet propulsion must be used Vigorous jetting can increase the animals’ speed by an order of magnitude, but the evidence available suggests that such a performance always stops the hearts, jet propulsion, at anything approaching the maximum speeds observed, can only be managed on oxygen debt. Since the possible oxygen debt seems to be very limited (about 22 ml kg⁻¹; Wells & Wells, 1984), vigorous jet propulsion is impossible as a regular means of transport.

Even at slower speeds, jet propulsion may be a very marginal operation for an octopus. At rest, the internal mantle pressure averages only 0·04—0·06 kPa as the animal ventilates (Wells & Smith, 1985). These very low pressures contrast with those measured in jetting. Eledone, which commonly swims rather than walking around its tank, showed internal mantle pressures, while jetting, in the range of 4—8kPa; a few observations with Octopus yielded similar values (M. J. Wells, unpublished observations). These peak figures for slow, jet-propelled swimming (but not escape responses, where pressures can rise to 20-40 kPa, Trueman & Packard, 1968) are well in excess of the peak pressures found in the lateral venae cavae of Octopus, which (at rest) rarely exceed 0·2kPa (Wells & Smith, 1987). Even if it is the mean rather than the peak pressures that matter (because the pulsations in the great veins are normally more frequent than the mantle contractions), it is not surprising that any resort to jetting stalls the venous return to the vessels inside the mantle supplying the branchial hearts. If the branchial hearts have insufficient blood to pump, the supply to the systemic heart, which contracts only when filled, is reduced, and the aortic pulse ceases or becomes erratic.

For Octopus this limitation on jet transport is unlikely to be important. The capacity of the octopod circulatory system is suited to the life style of a predator whose feeding forays require frequent stops to grope into crevices and among the substrate. Jetting is used to get from one patch of cover to the next or, more vigorously and perhaps comparatively rarely, for escape or a sudden jet-propelled attack on prey seen at a short distance.

In squid, which move continuously by jet propulsion, the situation must be different. Squid must somehow remain in oxygen balance during routine jet-propelled swimming. Two factors seem likely to contribute to this capacity. One is the relatively low pressures associated with routine jet propulsion in cruising squid. The other is the possible contribution of elastic recoil following compression of the collagenous tunics in the squid mantle wall. With regard to the former, Webber & O’Dor (1985) recorded peak mantle pressures of 1·5 and 3·0 and means of 0·43 at 0·99 kPa from a 550-g Illex, swimming freely in a large tank. The higher figures were recorded during a sequence when the animal was pursued by a net ‘but did not exhibit a full escape response’. These pressures are less than half those found in jetswimming octopods (see above). Webber & O’Dor’s squid were cruising at 11·5 °C at 37 and 40 jets min⁻¹, a ventilation rate some 50% higher than that of similar-sized Eledone or Octopus at 21·5°C. The squids are streamlined and evidently able to propel themselves at cruising speed with jet pressures substantially below those seeded to move the more rounded octopods.

Only one attempt to record venous pressures in a squid has so far been reported. Bourne (1982) found a pulse from 0·07 kPa in diastole to 0·63 kPa in systole in the anterior vena cava of Loligo. This would be sufficient to return the blood to the hearts during jetting. But the values are suspect, as Bourne points out, because the pulse coincided with ventilation and probably recorded mantle rather than blood pressures. Superimposed upon the larger pulses were smaller, more frequent, oscillations of around 0·2 kPa which perhaps represented the pressures generated by peristalsis of the great veins. This would be similar to the value found in Octopus, with a lower mean mantle pressure. The problem of how the venous return is achieved while the animal is jetting is evidently not as great in squids as in octopods, but the pressure available still seems inadequate to ensure the venous flow from the head and arms.

In this situation, the possibility of help from negative pressures within the mantle arising from the elastic recoil of the collagen tunics in the mantle wall becomes important. Such a possibility was inherent in Gosline & Shadwick’s (1983) demonstration that the circular muscles are not simply expanded by the radials, as was at one time believed; powerful jets are associated with contraction beyond the point found in quiet ventilation and the connective tissue tunics in the mantle wall then contribute to re-expansion.

Actual demonstration of negative pressures in the mantles of swimming squids has had to wait until quite recently. Webber & O’Dor (1986) have shown negative pressures in the region of 0·3 kPa in 490-g Illex jetting at submaximal speeds and M. J. Wells (in preparation) has obtained similar values from 12-g Lolliguncula during escape responses. Negative pressures of this magnitude would appear to solve the problem of how the blood gets back into the mantle from the head and arms. But there are two difficulties. One is that the negative phases of the jet pulses seem to be of very brief duration, lasting for only a small fraction of each jet cycle. A second is their near disappearance at normal cruising speeds. Webber & O’Dor (1986 ; figs 4, 6) show irregular values of 0·1-0·2kPa for around half of the cycles illustrated; the results with Lolliguncula imply even smaller negative pressures, often absent for many cycles on end. So the problem remains unresolved, pending the development of techniques for measuring the evidently very small pressure differentials between veins and mantle in cruising squid.