Abstract
The whole-body (tachymetabolic) endothermy seen in modern birds and mammals is long held to have evolved independently in each group, a reasonable assumption when it was believed that its earliest appearances in birds and mammals arose many millions of years apart. That assumption is consistent with current acceptance that the non-shivering thermogenesis (NST) component of regulatory body heat originates differently in each group: from skeletal muscle in birds and from brown adipose tissue (BAT) in mammals. However, BAT is absent in monotremes, marsupials, and many eutherians, all whole-body endotherms. Indeed, recent research implies that BAT-driven NST originated more recently and that the biochemical processes driving muscle NST in birds, many modern mammals and the ancestors of both may be similar, deriving from controlled 'slippage' of Ca(2+) from the sarcoplasmic reticulum Ca(2+) -ATPase (SERCA) in skeletal muscle, similar to a process seen in some fishes. This similarity prompted our realisation that the capacity for whole-body endothermy could even have pre-dated the divergence of Amniota into Synapsida and Sauropsida, leading us to hypothesise the homology of whole-body endothermy in birds and mammals, in contrast to the current assumption of their independent (convergent) evolution. To explore the extent of similarity between muscle NST in mammals and birds we undertook a detailed review of these processes and their control in each group. We found considerable but not complete similarity between them: in extant mammals the 'slippage' is controlled by the protein sarcolipin (SLN), in birds the SLN is slightly different structurally and its role in NST is not yet proved. However, considering the multi-millions of years since the separation of synapsids and diapsids, we consider that the similarity between NST production in birds and mammals is consistent with their whole-body endothermy being homologous. If so, we should expect to find evidence for it much earlier and more widespread among extinct amniotes than is currently recognised. Accordingly, we conducted an extensive survey of the palaeontological literature using established proxies. Fossil bone histology reveals evidence of sustained rapid growth rates indicating tachymetabolism. Large body size and erect stature indicate high systemic arterial blood pressures and four-chambered hearts, characteristic of tachymetabolism. Large nutrient foramina in long bones are indicative of high bone perfusion for rapid somatic growth and for repair of microfractures caused by intense locomotion. Obligate bipedality appeared early and only in whole-body endotherms. Isotopic profiles of fossil material indicate endothermic levels of body temperature. These proxies led us to compelling evidence for the widespread occurrence of whole-body endothermy among numerous extinct synapsids and sauropsids, and very early in each clade's family tree. These results are consistent with and support our hypothesis that tachymetabolic endothermy is plesiomorphic in Amniota. A hypothetical structure for the heart of the earliest endothermic amniotes is proposed. We conclude that there is strong evidence for whole-body endothermy being ancient and widespread among amniotes and that the similarity of biochemical processes driving muscle NST in extant birds and mammals strengthens the case for its plesiomorphy.