Skip to main content
Dryad logo

Differing limb functions and their potential influence upon the diversification of the mustelid hindlimb skeleton

Citation

Kilbourne, Brandon (2020), Differing limb functions and their potential influence upon the diversification of the mustelid hindlimb skeleton, Dryad, Dataset, https://doi.org/10.5061/dryad.f4qrfj6v3

Abstract

Though form-function relationships of the mammalian locomotor system have been investigated for over a century, recent models of trait evolution have hitherto been seldom used to identify likely evolutionary processes underlying the locomotor system’s morphological diversity. Using mustelids, an ecologically diverse carnivoran lineage, I investigated whether variation in hindlimb skeletal morphology functionally coincides with climbing, digging, swimming, and generalized locomotor habits by using 15 linear traits of the femur, tibia, fibula, calcaneum, and metatarsal III across 44 species in a principal components analysis. I subsequently fit different models of Brownian motion and adaptive trait diversification individually to each trait. Climbing, digging, and swimming mustelids occupy distinct regions of phenotypic space characterized by differences in bone robustness. Models of adaptive and neutral evolution are, respectively, the best fits for long bone lengths and muscle in-levers, suggesting that different kinds of traits may be associated with different evolutionary processes. However, simulations based upon models of best fit reveal low statistical power to rank the models. Though differences in mustelid hindlimb skeletal morphology appear to coincide with locomotor habits, further study, with sampling expanded beyond Mustelidae, is necessary to better understand to what degree adaptive evolution shapes morphological diversity of the locomotor system.

Methods

Data was collected using digital caliper measurements. Units of measurement: mm.

Code was written in R.

Usage Notes

Missing values are highlighted in black in Excel spreadsheet. For fitting of trait evolution models, bootstrapping of model parameters, and data simulation, the archived R code presents the OU4 model as base code. Other models (OU3_SRG, OU3_SNM, etc.) were fit by modifying this base code.

Funding

Deutsche Forschungsgemeinschaft, Award: Ki 1843-3/1

Deutsche Forschungsgemeinschaft, Award: Ki 1843-3/2