The locomotor environment and behavior of quadrupedal mammals exert functional constraints on their limbs. Therefore long bone shapes are thought to reflect at least partially the species’ locomotor ecology. Semi-aquatic species move through two media with distinct density and viscosity and their locomotor apparatus should therefore reflect a trade-off between the divergent functional constraints it faces. Adaptation to a semi-aquatic lifestyle occurred independently in otters (Lutrinae) and minks (Mustelinae). Analyzing semi-aquatic mustelids and their terrestrial relatives, we investigate long bone shape diversity, describe changes in long bone shape associated with a semi-aquatic lifestyle, and discuss the functional consequences of these shape changes. The robustness of the otter bones is highlighted and its potential ballast role discussed. Large epiphyses are observed in otters but this trend seems associated with terrestrial more than with aquatic locomotion. Thus, the most aquatic species, Enhydra lutris, presents narrow knee articulations compared to similar sized less aquatic species. Enhydra lutris presents a fore- and hind limb shape that diverge from that in other otters. Minks show bone shapes similar to each other but only Neovison vison tends to differ from its terrestrial relatives. The evolution of limb shape in this group is strongly correlated with size, locomotor mode, and phylogenetic history, leading to a morphological pattern where the roles of each of these factors are difficult to disentangle.

The challenging complexity of biological structures has led to the development of several methods for quantitative analyses of form. Bones are shaped by the interaction of historical (phylogenetic), structural, and functional constrains. Consequently, bone shape has been investigated intensively in an evolutionary context. Geometric morphometric approaches allow the description of the shape of an object in all of its biological complexity. However, when biological objects present only few anatomical landmarks, sliding semi-landmarks may provide good descriptors of shape. The sliding procedure, mandatory for sliding semi-landmarks, requires several steps that may be time-consuming. We here compare the time required by two different software packages (‘Edgewarp’ and ‘Morpho’) for the same sliding task, and investigate potential differences in the results and biological interpretation. ‘Morpho’ is much faster than ‘Edgewarp,’ notably as a result of the greater computational power of the ‘Morpho’ software routines and the complexity of the ‘Edgewarp’ workflow. Morphospaces obtained using both software packages are similar and provide a consistent description of the biological variability. The principal differences between the two software packages are observed in areas characterized by abrupt changes in the bone topography. In summary, both software packages perform equally well in terms of the description of biological structures, yet differ in the simplicity of the workflow and time needed to perform the analyses.

Identifying ecomorphological convergence examples is a central focus in evolutionary biology. In xenarthrans, slow arboreality independently arose at least three times, in the two genera of ‘tree sloths’,BradypusandCholoepus, and the silky anteater,Cyclopes. This specialized locomotor ecology is expectedly reflected by distinctive morpho-functional convergences.Cyclopes, although sharing several ecological features with ‘tree sloths’, do not fully mirror the latter in their outstandingly similar suspensory slow arboreal locomotion. We hypothesized that the morphology ofCyclopesis closer to ‘tree sloths’ than to anteaters, but yet distinct, entailing that slow arboreal xenarthrans evolved through ‘incomplete’ convergence. In a multivariate trait space, slow arboreal xenarthrans are hence expected to depart from their sister taxa evolving toward the same area, but not showing extensive phenotypical overlap, due to the distinct position ofCyclopes.Conversely, a pattern of ‘complete’ convergence (i.e., widely overlapping morphologies) is hypothesized for ‘tree sloths’. Through phylogenetic comparative methods, we quantified humeral and femoral convergence in slow arboreal xenarthrans, including a sample of extant and extinct non-slow arboreal xenarthrans. Through 3D geometric morphometrics, cross-sectional properties (CSP) and trabecular architecture, we integratively quantified external shape, diaphyseal anatomy and internal epiphyseal structure. Several traits converged in slow arboreal xenarthrans, especially those pertaining to CSP. Phylomorphospaces and quantitative convergence analyses substantiated the expected patterns of ‘incomplete’ and ‘complete’ convergence for slow arboreal xenarthrans and ‘tree sloths’, respectively. This work, highlighting previously unidentified convergence patterns, emphasizes the value of an integrative multi-pronged quantitative approach to cope with complex mechanisms underlying ecomorphological convergence.

Background: Callitrichids comprise a diverse group of platyrrhine monkeys that are present across South and Central America. Their secondarily evolved small size and pointed claws allow them to cling to vertical trunks of a large diameter. Within callitrichids, lineages with a high affinity for vertical supports often engage in trunk-to-trunk leaping. This vertical clinging and leaping (VCL) differs from horizontal leaping (HL) in terms of the functional demands imposed on the musculoskeletal system, all the more so as HL often occurs on small compliant terminal branches. We used quantified shape descriptors (3D geometric morphometrics) and phylogenetically-informed analyses to investigate the evolution of the shape and size of the humerus and femur, and how this variation reflects locomotor behavior within Callitrichidae.Results: The humerus of VCL-associated species has a narrower trochlea compared with HL species. It is hypothesized that this contributes to greater elbow mobility. The wider trochlea in HL species appears to correspondingly provide greater stability to the elbow joint. The femur in VCL species has a smaller head and laterally-oriented distal condyles, possibly to reduce stresses during clinging. Similarly, the expanded lesser trochanters visible in VCL species provide a greater lever for the leg retractors and are thus also interpreted as an adaptation to clinging. Evolutionary rate shifts to faster shape and size changes of humerus and femur occurred in theLeontocebusclade when a shift to slower rates occurred in theSaguinusclade.Conclusions: Based on the study of evolutionary rate shifts, the transition to VCL behavior within callitrichids (specifically theLeontocebusclade) appears to have been an opportunity for radiation, rather than a specialization that imposed constraints on morphological diversity. The study of the evolution of callitrichids suffers from a lack of comparative analyses of limb mechanics during trunk-to-trunk leaping, and future work in this direction would be of great interest.

Background: Biological adaptation manifests itself at the interface of different biologically relevant ‘levels’, such as ecology, performance, and morphology. Integrated studies at this interface are scarce due to practical difficulties in study design. We present a multilevel analysis, in which we combine evidence from habitat utilization, leaping performance and limb bone morphology of four species of tamarins to elucidate correlations between these ‘levels’. Results: We conducted studies of leaping behavior in the field and in a naturalistic park and found significant differences in support use and leaping performance. Leontocebus nigrifrons leaps primarily on vertical, inflexible supports, with vertical body postures, and covers greater leaping distances on average. In contrast, Saguinus midas and S. imperator use vertical and horizontal supports for leaping with a relatively similar frequency. S. mystax is similar to S. midas and S. imperator in the use of supports, but covers greater leaping distances on average, which are nevertheless shorter than those of L. nigrifrons. We assumed these differences to be reflected in the locomotor morphology, too, and compared various morphological features of the long bones of the limbs. According to our performance and habitat utilization data, we expected the long bone morphology of L. nigrifrons to reflect the largest potential for joint torque generation and stress resistance, because we assume longer leaps on vertical supports to exert larger forces on the bones. For S. mystax, based on our performance data, we expected the potential for torque generation to be intermediate between L. nigrifrons and the other two Saguinus species. Surprisingly, we found S. midas and S. imperator having relatively more robust morphological structures as well as relatively larger muscle in-levers, and thus appearing better adapted to the stresses involved in leaping than the other two. Conclusion: This study demonstrates the complex ways in which behavioral and morphological ‘levels’ map onto each other, cautioning against oversimplification of ecological profiles when using large interspecific eco-morphological studies to make adaptive evolutionary inferences.