Search for: All records

Abstract Locomotor evolution in synapsids involved numerous functional shifts associated with the transition from sprawled to erect limb postures on the line to therian mammals. Given that bone structure frequently reflects functional requirements, this study investigated evolutionary changes in synapsid humerus and femur proportions as a lens to evaluate functional shifts through time. A total of 936 bones were measured, representing 330 species across the full 320+ million years of synapsid history. This dataset was used to test whether transformations in stylopod proportions are consistent with inferred changes in bone loading mechanics, alignment of joint and muscle forces, muscular control of the shoulder and hip, and differential support of body weight by the fore‐ and hindlimbs. As variation in bone dimensions may also correlate with bone or body size, this study first developed a novel approach for calculating species‐specific, size‐corrected measures of bone proportions. By disentangling the effect of body size from functional signals recorded in bone geometry, this then enabled a node‐to‐node appraisal of how bone allometry itself evolved through time. Ancestral state reconstruction of size‐corrected stylopod proportions reveals trends that broadly support many hypothesized shifts in locomotor biomechanics along the therian stem lineage. However, patterns of transformation are frequently complex, suggesting functional mosaicism, and stylopod proportions that typify therians as a whole are often not achieved until crown Theria itself. Several instances of temporary trend reversal are also inferred, particularly within non‐mammalian cynodonts, indicating greater functional or ecological diversification in this group. 

Synopsis Size has an impact on various aspects of an animal's biology, including physiology, biomechanics, and ecology. Accurately and precisely estimating size, in particular body mass, is therefore a core objective of paleobiologists. Two approaches for estimating body mass are common: whole-body volumetric models and individual element-scaling (e.g., bones, teeth). The latter has been argued to be more accurate, while the former more precise. Here, we use minimum convex hulls (MCHs) to generate a predictive volumetric model for estimating body mass across a broad taxonomic and size range (127 g – 2735 kg). We compare our MCH model to stylopodial-scaling, incorporating data from the literature, and find that MCH body mass estimation is both more accurate and more precise than stylopodial estimation. An explanation for the difference between methods is that reptile and mammal stylopod circumference and length dimensions scale differentially (slope 1.179 ± 0.102 vs. 1.038 ± 0.031, respectively), such that reptiles have more robust bones for a given size. Consequently, a mammalian-weighted stylopodial-scaling sample overestimates the body mass of larger reptiles, and this error increases with size. We apply both estimation equations to a sample of 12 Permo-Triassic tetrapods and find that stylopodial-scaling consistently estimates a higher body mass than MCH estimation, due to even more robust bones in extinct species (slope = 1.319 ± 0.213). Finally, we take advantage of our MCH models to explore constraints regarding the position of the center of mass (CoM) and find that relative body proportions (i.e., skull:tail ratio) influence CoM position differently in mammals, crocodylians, and Permo-Triassic tetrapods. Further, we find that clade-specific body segment expansion factors do not affect group comparisons but may be important for individual specimens with rather disproportionate bodies (e.g., the small-headed and large-tailed Edaphosaurus). Our findings suggest that the whole-body volumetric approach is better suited for estimating body mass than element-scaling when anatomies are beyond the scope of the sample used to generate the scaling equations and provides added benefits such as the ability to measure inertial properties. 

The ‘sprawling–parasagittal’ transition was a major postural shift that occurred in the ancestors of mammals, [PSE1] underpinned by musculoskeletal reorganization of the limbs. However, ‘when’ and ‘how’ this important postural shift occurred is unknown. While the anatomical changes characterizing this transition can be traced through the fossil record, how these relate to functional changes, and the acquisition of parasagittal posture, remains poorly understood. Here, we produced three-dimensional musculoskeletal models of the forelimbs of extant (n=3) and fossil (n=8) taxa that phylogenetically and functionally span the sprawling–parasagittal transition. We calculated joint range of motion (ROM) to determine a 3D pose space, using the novel APSE algorithm (Accelerated Pose Searching with Electrostatics). We then estimated muscle moment arms (MMAs) across the entire pose space for all muscles crossing the shoulder and elbow joints. Models of extant species were validated against empirical measures of ROM and MMA derived from ex vivo XROMM (X-ray reconstruction of moving morphology). Among extant species, our parasagittal taxon occupied a distinct region of pose-space, with more retracted and depressed shoulder joint angles. MMA data show increased emphasis on shoulder elevation associated with a parasagittal posture, but greater shoulder depression in sprawlers. Results from the fossil species show complex, non-linear patterns of forelimb transformation, demonstrating that the ‘sprawling-parasagittal’ transition is characterized by homoplasy and postural variation within the mammalian lineage. 

The ‘sprawling-parasagittal’ transition was a major postural shift in the ancestors of mammals, resulting in musculoskeletal reorganization of the forelimbs that underpins modern mammal locomotor diversity. However, ‘when’ and ‘how’ this important postural shift occurred is unknown. While the anatomical changes characterizing this transition can be traced through the fossil record, how these relate to functional changes, and the acquisition of parasagittal posture, remains poorly understood. We produced three-dimensional musculoskeletal models of the forelimbs of extant (n=3) and fossil (n=8) taxa that phylogenetically and functionally span the sprawling–parasagittal transition. We calculated joint range of motion (ROM) to determine a 3D pose-space, using the novel APSE algorithm (Accelerated Pose Searching with Electrostatics). We then estimated muscle moment arms (MMAs) across the entire pose space for all muscles crossing the shoulder and elbow joints. Models of extant species were validated against empirical measures of ROM and MMA derived from ex vivo XROMM (X-ray reconstruction of moving morphology). Among extant species, in both models and experiments, our parasagittal taxon occupied a distinct region of pose-space, with more retracted and depressed shoulder joint angles. MMA data show increased emphasis on shoulder elevation associated with a parasagittal posture, but greater shoulder depression in sprawlers. We hypothesised the fossil taxa would follow trends in these postural variables – e.g., increasing shoulder retraction and elevation MMAs through time – but they instead showed complex, non-linear patterns of forelimb transformation. We demonstrate that the ‘sprawling-parasagittal’ transition is characterized by considerable homoplasy and continuous postural variation throughout mammalian evolution.