Skeletal joint morphology and mobility underlie movement, behavior and ecology in vertebrates. Joints can be categorized by their shape and articulation type, but such schemes might be unreliable for inferring function across the full diversity of vertebrates. We test hypothesized relationships between joint form and function by collecting marker-based ex vivo, cadaveric XROMM data on the shoulder and elbow joints of the tegu lizard (Salvator merianae) and Virginia opossum (Didelphis virginiana), which between them contain articulations historically classified as ball-and-socket, hemi-sellar, hinge and condylar joints. We measured 3D rotational and translational mobility at each joint and compared our experimental results against predictions based on articular morphology. Contrary to our predictions, the opossum ball-and-socket shoulder joint was less mobile – it had a smaller 3D range of motion envelope – than the tegu hemi-sellar shoulder joint and even the tegu condylar elbow joint, challenging the notion that ball-and-socket joints provide an inherent mobility advantage. However, the ball-and-socket opossum shoulder also had a less complex mobility envelope, with fewer interactions between degrees of freedom, allowing it to transition between poses more easily. Matching osteological predictions, the hinge elbow of the opossum was the least mobile. All joints exhibited coupling between rotational and translational degrees of freedom, further emphasizing the need to incorporate translational motion and soft tissue constraints for accurately modeling joint mobility. Our findings underscore the complexity of form–function relationships in vertebrate skeletal joints, and demonstrate that joint morphology alone, in the absence of soft tissues, does not provide a complete picture of joint mobility.

Author contributions

Conceptualization: L.F.-L., A.B., S.E.P.; Data curation: S.E.P.; Formal analysis: R.J.B., L.F.-L.; Funding acquisition: L.F.-L., S.E.P.; Investigation: R.J.B., L.F.-L.; Methodology: R.J.B., L.F.-L., S.E.P.; Project administration: S.E.P.; Supervision: A.B., S.E.P.; Visualization: R.J.B., L.F.-L.; Writing – original draft: R.J.B., L.F.-L., S.E.P.; Writing – review & editing: R.J.B., A.B., S.E.P.

Funding

This research was supported in part by the US National Science Foundation (grant no. DEB-1754459 to S.E.P.) and the Harvard University Robert A. Chapman Fund (to L.F.-L.).

Data and resource availability

Raw data for this study (X-ray videos, calibration files and CT-derived meshes) are available from Harvard Dataverse (https://doi.org/10.7910/DVN/OTJKEA). Interactive versions of the 3D plots in Figs 3, 4, 7 and 8 are provided as supplementary HTML files available from Harvard Dataverse (https://doi.org/10.7910/DVN/OTJKEA).

You do not currently have access to this content.