New perspectives on body size and shape evolution in dinosaurs.

Diversity in the body shapes and sizes of dinosaurs was foundational to their widespread success during the Mesozoic era. The ability to quantify body size and form reliably is therefore critical to the study of dinosaur biology and evolution. Body mass estimates for any given fossil animal are, in theory, most informative when derived from volumetric models that account for the three-dimensional shapes of the entire body. In addition to providing estimates of total body mass, volumetric approaches can be used to determine the inertial properties of specific body segments and the overall distribution of mass throughout the body, each of which are essential for the modelling and interpretation of form-function relationships and their associations with ecology. However, the determination of body volumes in fossil taxa is often subjective, and may be sensitive to varied artistic inference. This highlights the need for an approach to body mass estimation in which body segment volumes are systematically constrained by quantitative scaling relationships between the hard tissues that fossilise and the soft tissues only observable in extant taxa. To this end, we used recently published skeletal to soft tissue volumetric scaling factors derived from CT data of extant sauropsids to estimate body segment mass properties from skeletal models of 52 non-avian dinosaurs representing the majority of major clades and body plans. The body masses estimated by this study range from less than 200 g in the tiny avialan Yixianornis to over 60 tonnes in the giant sauropod Patagotitan, which is currently the largest dinosaur known from mostly complete skeletal remains. From our models, we infer that many previous reconstructions of soft tissue envelopes may be too small, and that many dinosaurs were therefore heavier than previous estimates. Our models generally overlap with the range of body mass estimates derived from limb bone shaft dimensions, but with considerable quantitative variability among major clades. This suggests that different taxa either differed in skeletal to soft tissue volume ratios, or that their limb bone dimensions varied relative to body mass, perhaps related to differences in locomotor dynamics and postural evolution. Our models also allowed us to investigate variation in mass distribution and body proportions across different dinosaurs from a perspective grounded in extant anatomical data, framing long-standing hypotheses about their form, function, and behaviour in a quantitative context. For example, reconstructed disparity in whole-body centres of mass reflects a broad array of postures in different dinosaur clades, while the lack of strong positive allometry in the dimensions of the weight-bearing limb segments relative to total body mass corroborates previous studies suggesting an overall decrease in dinosaur locomotor performance as body size increased.