PHYLOGENETIC AND FOSSIL EVIDENCE FOR A COMMON BODY SIZE ATTRACTOR IN MARINE MAMMALS

Evolutionary transitions between terrestrial and aquatic habitats are rare and often have large effects on the evolutionary trajectory of the clade making the transition. Following a single transition from the marine realm to the terrestrial realm, tetrapods have subsequently re-evolved a marine lifestyle at least 30 separate times. At least six of these re-invasions of the water occurred within crown-group mammals and four [Sirenians (Sirenia), whales (Cetacea), pinnipeds (Pinnipedia), and otters (Lutrinae)] clades are extant. Although marine mammals are widely known to be larger than their terrestrial sister groups, the extent to which the body size evolution of these clades reflects common constraints of a marine lifestyle remains little studied. Here we use previously published and publicly available data on the body masses of 3832 living and 3005 fossil mammal species in addition to the most recent mammal supertree and up-to-date fossil ranges to examine the evolutionary trajectories of body size in marine mammals both through comparative phylogenetic analysis and direct interrogation of the fossil record. We find that the evolution of an aquatic lifestyle has driven three of the four aquatic mammal clades toward indistinguishable body size attractors of ~1 metric ton. In these clades, phylogenetic analysis favors an Ornstein-Uhlenbeck model of size evolution and both phylogenetic and fossil data predict comparable optimal sizes and evolutionary trajectories. Both fossil and comparative phylogenetic data indicate that the otters (Lutrinae) were the most recent clade to reinvade the water and the only clade to remain at small body size. Our results confirm that mammals living in aquatic environments have higher optimal body sizes and further suggest the existence of a body size attractor that has been discovered independently by three mammalian clades. The sustained small size of aquatic mustelids could either indicate the presence of a second attractor at small size or competitive exclusion from the ~1 ton attractor. The close agreement in statistical inference from comparative phylogenetic and fossil data both strengthens the findings described above and highlights the power of both approaches to capture the macroevolutionary dynamics of body size.