Evolution, 33(3), 1979, pp. 983-997 ECTOTHERMY AND THE SUCCESS OF DINOSAURS MICHAEL J. BENTON Department of Geology, The University, Newcastle-upon-Tyne, NE1 7RU, England, U.K. Received February 28, 1978. Revised October 9, 1978 Dinosaurs were the dominant terrestrial AAAS symposium published as Olson and tetrapods of the Mesozoic and their eco- Thomas (1979) and references to papers logical roles and biological impact are gen- from this volume are, as a result, brief. erally compared with those of mammals in the Cenozoic. This has led to the as- DISCUSSION OF EVIDENCE FOR sumption that dinosaurs were endotherm- DINOSAUR ENDOTHERMY ic also (Bakker, 1971, 1979) and there Interest in the thermoregulatory physi- seems to be some evidence in support of ology of dinosaurs has revived recently this hypothesis. However, this evidence is (Bakker, 1975a; Desmond, 1975; Gould, not conclusive. Could it be that dinosaurs 1977; May, 1977; Marx, 1978; Olson and were in fact ectothermic and could their Thomas, 1979), although speculations success be explained most simply in terms have been made on this topic for at least of a normal reptilian physiology? 100 years. Audowa (1929) and Nopcsa In this paper I review the published evi- (1934) suggested that dinosaur extinction dence for and against dinosaur endother- might have been related to their ectother- my and attempt an assessment in terms of my, but other later authors have suggested the thermoregulatroy physiology of living that it was connected with their endother- animals, and I discuss the success and the my (Wieland, 1942; Russell, 1965; Cloud- extinction of dinosaurs in connection with sley-Thompson, 1971; Bakker, 1972, published data on Mesozoic paleoclima- 1973). tology. It is argued that endothermy in Broili (1941) and Schuh (1951), extrap- dinosaurs would have been energetically olating from finds in pterosaurs and syn- costly and yet there appears to have been apsids, predicted that dinosaurs would be no need for it. Ostrom (1979) complains found with "hair," necessary insulation in that critics of dinosaur endothermy merely an endotherm. However, others have re- offer alternatives and do not disprove the lated the large size and naked skin of di- arguments of their opponents. This paper nosaurs to endothermy, arguing that the attempts to show why dinosaurs (with the absence of small dinosaurs is connected possible exception of some theropods) had with the problems of excessive heat loss to be ectothermic. by small uninsulated endotherms with a Before proceeding, the four basic de- high surface/volume ratio (Bakker, 1971, scriptive terms used in thermoregulatory 1972, 1973; Ricqles, 1974). These latter physiology should be defined accurately interpretations were criticized by Thul- (Cowles, 1940, 1962; McNab, 1978). A born (1973) and Feduccia (1973) who poikilothermic animal has no control over pointed out that young dinosaurs were body temperature, which follows external small and naked and could only have conditions, while a homeothermic animal avoided cooling stress problems if they has a constant body temperature. An ec- were ectothermic. totherm derives heat from external sources The evidence most recently presented (generally solar radiation directly or indi- for dinosaur endothermy includes erect rectly) and an endotherm from internal gait, paleoclimates and distribution, small sources (generally metabolic or muscular agile dinosaurs, brain size, predator-prey heat). ratios, bone histology, and the parietal- This paper was written before the pineal complex. 983 This content downloaded from 137.222.249.97 on Fri, 29 Jan 2016 22:22:16 UTC All use subject to JSTOR Terms and Conditions 984 MICHAEL J. BENTON Erect gait. -Dinosaurs had erect stance saurs. However, this may be connected and advanced gait. Among living animals, with gait rather than thermoregulation. only endotherms have erect gait; and it Most living reptiles have a sprawling was suggested that this and the supposed stance, whereas the agile theropods had ability of dinosaurs to achieve fast speeds an erect stance. The erect posture which indicated endothermy (Schuh, 1951; Os- supported the great weight of brontosaurs trom, 1970, 1974; Bakker, 1971, 1972, allowed agile motion in smaller dinosaurs. 1973, 1974, 1975a; Dodson, 1974). How- The generally warm Mesozoic climates ever, this view has been widely criticized (see below) may have maintained their as a logical non sequitur (Thulborn, 1973; body temperatures high enough during the Bennett and Dalzell, 1973; Feduccia, day for activity and their carnivorous diet 1973, 1974; Bennett, 1974; Halstead, provided readily assimilable energy. After 1976). The fact that the only living ani- a cool night, body temperatures may have mals with erect gait are endotherms does been increased by basking or uptake of not mean that every animal with erect gait heat from rocks in the early morning as in is an endotherm-it should also be noted living reptiles. Ostrom (1978, 1979) sug- that many mammals and other endo- gests that the only possibly endothermic thermic animals (see below) have a dinosaurs were these small theropods, sprawling gait (Jenkins, 1971). In addi- particularly in view of their possible re- tion, Alexander (1976) calculated speeds lationship to birds. of only 1-3.6 ms-t from trackways of di- Brain size. -Feduccia (1973) suggested nosaur footprints, although one hadrosaur that the small lizard-like brain size of di- trackway indicates a speed of 7.5 ms- nosaurs speaks for ectothermy, but finds (Russell and Beland, 1976). Thus large of "ostrich dinosaurs" with large brains dinosaurs, at least, may have moved rel- (Russell, 1972) were interpreted as evi- atively slowly and the erect stance may be dence for endothermy (Dodson, 1974; no more than the most efficient way of Ricqles, 1974; Bakker, 1974, 1975a). supporting a heavy body (Bennett and However, the large brains of these dino- Dalzell, 1973; Feduccia, 1973; Bennett, saurs are associated with good eyesight 1974; McNab, 1978). and balance (Halstead, 1975b) and do not Paleoclimatology and distribution.- necessarily imply advanced mammal-like Finds of dinosaurs within the Cretaceous intelligence. In any case, internal ther- Arctic Circle (Russell, 1973) have been moregulation requires a smaller mass of thought to indicate endothermy also (Bak- neural tissue than that required by ecto- ker, 1975a; Desmond, 1975), but precise therms for the behavioral control of inter- data on Mesozoic climates are not given nal temperature (Jerison, 1973) and thus (Halstead, 1975b, 1976). There is no evi- brain size may not be a reliable indicator dence of glaciation in the Mesozoic of endothermy (cf. Robinson, 1971). (Schwarzbach, 1963; Bakker, 1975b) and Predator-prey ratios. -Herbivores climates at these high latitudes need not (whether endothermic or ectothermic) can have been cold. In general, Mesozoic cli- support about 5% of their biomass of en- mates were warmer than today (Dorf, dothermic predators, and for ectothermic 1970; see below). (reptile) carnivores, this predator-prey ra- Small agile dinosaurs. -Many smaller tio is apparently nearer 30-50% (Bakker, dinosaurs (especially lightly built thero- 1972, 1975a, 1975b, 1979). Bakker pods like Compsognathus, Saurorni- (1972, 1973, 1974, 1975a, 1975b) calcu- thoides and Deinonychus) were presum- lated predator--prey ratios for certain fos- ably agile and must have been able to sil populations and used the values he move rapidly (Ostrom, 1978). At first obtained as indices of ectothermy or en- sight, this is very different from many liv- dothermy. He noted a drop from 50% in ing lizards and might be thought to be the Early Permian to 10% in the Late good evidence for endothermy in dino- Permian and concluded that this repre- This content downloaded from 137.222.249.97 on Fri, 29 Jan 2016 22:22:16 UTC All use subject to JSTOR Terms and Conditions ECTOTHERMY AND THE SUCCESS OF DINOSAURS 985 sented the changeover from ectothermy to phocalcic salts, and they develop espe- endothermy in populations of fossil rep- cially in rapid growth (Ricqles, 1974). The tiles. idea has arisen that these Haversian sys- Charig (1976) listed many possible tems indicate endothermy (Enlow and sources of error in applying predator-prey Brown, 1957; Currey, 1962; Ricqles, ratios to fossil communities, such as col- 1969, 1972a, 1972b, 1974, 1976; Bakker, lector bias, incompleteness of the fossil re- 1972, 1974, 1975a; Dodson, 1974). cord, relative life spans, and interactions Large mammals, most birds, advanced of other animals in food chains. I would synapsids and dinosaurs have extensive add that some "prey" species (e.g., Bron- Haversian systems, and tunas, turtles and tosaurus) might have been unavailable for crocodiles have them less developed. They predation owing to very large size, al- are not present in small mammals, pas- though they could have been scavenged serine birds, lizards, snakes or primitive after they died, but their inclusion would reptiles (Enlow, 1969; Ricqles, 1974, greatly lower the apparent predator-prey 1976). The first two in this list are the ratio. The present-day elephants effective- "most" endothermic vertebrates (i.e., they ly have no natural predators because of have the highest rates of mass-specific their large size. Of course, the ratios can metabolism), and they might be expected only indicate whether the predator is en- to have such structures most highly de- dothermic or ectothermic: the amount of veloped (McNab, 1978). Thus, Haversian prey required would be the same whatever systems are present in most living endo- its thermoregulatory state. If ectothermic therms and some living ectotherms and tyrannosaurs ate their young or other car- tend to be associated with large size (Hal- nivores, or if they were partly scavengers, stead, 1976). Bouvier (1977) lists animals the ratios could be explained readily (Tra- with secondary Haversian systems and cy, 1976). shows that their presence is not correlated In fact, it seems that ratios of large with endothermy. The correlation seems ectothermic predators to prey closely ap- instead to be with homeothermy (McNab, proach those for endothermic predators, 1978).