The evolution of femoral cross‐sectional properties in sciuromorph rodents: Influence of body mass and locomotor ecology

Abstract

In several groups of mammals, adaptation to differing functional demands is reflected in long bone cross‐sectional properties (CSP), which relate to the resistance to compression and to bending loads in the craniocaudal and mediolateral directions. Members of the Sciuromorpha (“squirrel‐like” rodents) display a diversity of locomotor ecologies and span three orders of magnitude in terms of body size. The availability of robust phylogenies is rendering them a suitable group to further substantiate the relationship of long bone CSP with locomotor ecology and body mass while taking the phylogenetic non‐independence among species into account. Here, we studied 69 species of Sciuromorpha belonging to three lifestyle categories, “arboreal,” “fossorial,” and “aerial” (i.e., gliding). We hypothesized locomotor category specific loading regimes that act on femora during predominant or, in terms of gliding, critical locomotor behaviors of each category. High resolution computed tomography scans of the specimens' femora were obtained and cross‐sections in 5% increments were analyzed. Cross‐sectional area, the craniocaudal second moment of area (SMAcc), and the mediolateral second moment of area were quantified. Further, a scaling analysis was conducted for each bone cross‐section to examine how the CSP scale with body mass. Body mass accounted for variances in CSP with mainly positive allometry. The aerial sciuromorphs showed lower values of CSP compared to the arboreal and fossorial species in the distal epiphysis for all quantified parameters and along the bone for SMAcc. In contrast to previous studies on other mammalian lineages, no differences in CSP were found between the fossorial and the arboreal lifestyles.

Differing effects of size and lifestyle on bone structure in mammals

“…In contrast, the macroevolutionary analyses of bone structure, namely the distribution of bone tissue within a skeletal element (e.g. [6][7][8][9][10]), are comparatively scarce, creating a major impediment to our understanding of vertebrate evolution. Bone structure is known for its high degree of plasticity [11].…”

Weighing homoplasy against alternative scenarios with the help of macroevolutionary modeling: A case study on limb bones of fossorial sciuromorph rodents

“…The arid environment potentially contains harder soils, demanding larger forces during digging that in turn require a stronger stabilization and resistance to forces acting on the femur. However, personal qualitative observations on the smaller datasets used by Mielke et al () and Scheidt et al () indicated that all arboreal and fossorial species share overall a similar range of trait values in their trabecular and cross‐sectional robustness properties, respectively and hence, potentially, the same optima. However, their datasets need to be extended in order to enable an unbiased quantitative comparison with the traits used in our study.…”

“…Two previous studies (Mielke et al, ; Scheidt, Wölfer, & Nyakatura, ) analyzed further robustness parameters of the sciuromorph femur (trabecular parameter of the femoral head and cross‐sectional properties along the proximodistal axis, respectively). Mielke et al () found no significant differences between the arboreal and fossorial groups.…”

“…Mielke et al () found no significant differences between the arboreal and fossorial groups. Scheidt et al () demonstrated that fossorial sciuromorphs are more robust in the distal epiphyseal region than their arboreal relatives. We did not consider the femoral traits investigated by Mielke et al () and Scheidt et al (), as only four of six xerine species were sampled in both studies what we considered too few to reliably estimate trait optima.…”

“…Scheidt et al () demonstrated that fossorial sciuromorphs are more robust in the distal epiphyseal region than their arboreal relatives. We did not consider the femoral traits investigated by Mielke et al () and Scheidt et al (), as only four of six xerine species were sampled in both studies what we considered too few to reliably estimate trait optima. Moreover, the dataset of Scheidt et al () is too large (i.e., 57 variables) for the computationally heavy methods applied herein.…”