When seals dive for prolonged periods, their arterial oxygen tension may decrease to values as low as 2.0 kPa. The time it takes to reach such extreme values is mainly determined by the extent of a selective peripheral vasoconstriction which distributes most of a much reduced cardiac output to the brain, while the rest of the body has to rely on local stores of oxygen and anaerobic metabolism (e.g. Blix & Folkow, 1983). In this situation brain oxygen demand would be much reduced, and hence dive duration further extended, if brain temperature was reduced. We have shown that that is indeed the case (Odden et al. 1999). In the present study we have used six harp seals (Pagophilus groenlandicus) and measured brain temperature (T_b), carotid blood temperature (T_c), muscle (m. latissimus dorsi) temperature (T_m) and rectal temperature (T_r) during experimental dives of 10 min duration. The details of the methodology followed Odden et al. (1999), the experiments being approved by the Norwegian Committee on Ethics in Animal Experimentation. We found that heart rate always fell promptly from about 100 to 8Ð10 b.p.m. upon submergence and that T_b fell significantly, in some cases as much as 3 °C, of which about 30 % took place during the first 5 min after the end of the dive, when the previously intense bradycardia was replaced by tachycardia (120Ð140 b.p.m.). T_c, which was about a degree lower than T_b before diving, fell in parallel with T_b during the dive, while T_m remained fairly stable throughout the dives, with T_r being reduced, probably due to local cooling from the hind flippers. Moreover, in some of the dives T_b started to decline even before the seal was submerged, suggesting that the decrease in T_b is not a result of passive cooling, but is under physiological control and even influenced by higher brain centres. To elucidate how this is achieved we carried out a series of anatomical studies of the vasculature of the fore flippers of three hooded seals (Cystophora cristata) after humane killing, by dissection and angiography using OEC Medical Systems, Inc., series 9600 X-ray equipment with Mixobar Colon (Astra Tech; 1 g ml^-1) as contrast medium. We found that the blood which enters the flippers is drained either through A-V shunts, which have been shown to be abundant in seal skin, to the large superficial saphenous veins, or by way of the capillary bed to smaller deep veins which split into a rete around the artery at the base of the flipper, where counter-current heat exchange occurs. We suggest that the animals achieve controlled cooling of the brain simply by regulating A-V shunt flow in the skin of the rather uninsulated fore flippers, whereby only those tissues which are perfused, notably the brain, are cooled. If we assume a similar Q₁₀-effect on cerebral metabolism in seals as in pigs, and further assume that seals can cool their brains by at least 3 °C when diving in the wild, this would result in a reduction of brain oxygen demand of 15Ð20 % (Busija & Leffler, 1987) and also provide neuroprotection against hypoxic injury.

All procedures accord with current national guidelines.