Postcranial pneumaticity: an evaluation of soft-tissue influences on the postcranial skeleton and the reconstruction of pulmonary anatomy in archosaurs

Abstract

Postcranial pneumaticity has been reported in numerous extinct sauropsid groups including pterosaurs, birds, saurischian dinosaurs, and, most recently, both crurotarsan and basal archosauriform taxa. By comparison with extant birds, pneumatic features in fossils have formed the basis for anatomical inferences concerning pulmonary structure and function, in addition to higher-level inferences related to growth, metabolic rate, and thermoregulation. In this study, gross dissection, vascular and pulmonary injection, and serial sectioning were employed to assess the manner in which different soft tissues impart their signature on the axial skeleton in a sample of birds, crocodylians, and lizards. Results from this study indicate that only cortical foramina or communicating fossae connected with large internal chambers are reliable and consistent indicators of pneumatic invasion of bone. As both vasculature and pneumatic diverticula may produce foramina of similar sizes and shapes, cortical features alone do not necessarily indicate pneumaticity. Noncommunicating (blind) vertebral fossae prove least useful, as these structures are associated with many different soft-tissue systems. This Pneumaticity Profile (PP) was used to evaluate the major clades of extinct archosauriform taxa with purported postcranial pneumaticity. Unambiguous indicators of pneumaticity are present only in certain ornithodiran archosaurs (e.g., sauropod and theropod dinosaurs, pterosaurs). In contrast, the basal archosauriform Erythrosuchus africanus and other nonornithodiran archosaurs (e.g., parasuchians) fail to satisfy morphological criteria of the PP, namely, that internal cavities are absent within bone, even though blind fossae and/or cortical foramina are present on vertebral neural arches. An examination of regional pneumaticity in extant avians reveals remarkably consistent patterns of diverticular invasion of bone, and thus provides increased resolution for inferring specific components of the pulmonary air sac system in their nonavian theropod ancestors. By comparison with well-preserved exemplars from within Neotheropoda (e.g., Abelisauridae, Allosauroidea), the following pattern emerges: pneumaticity of cervical vertebrae and ribs suggests pneumatization by lateral vertebral diverticula of a cervical air sac system, with sacral pneumaticity indicating the presence of caudally expanding air sacs and/or diverticula. The identification of postcranial pneumaticity in extinct taxa minimally forms the basis for inferring a heterogeneous pulmonary system with distinct exchange and nonexchange (i.e., air sacs) regions. Combined with inferences supporting a rigid, dorsally fixed lung, osteological indicators of cervical and abdominal air sacs highlight the fundamental layout of a flow-through pulmonary apparatus in nonavian theropods.