Phylogeny and foraging behaviour shape modular morphological variation in bat humeri

Abstract

Bats show a remarkable ecological diversity that is reflected both in dietary and foraging guilds (FGs). Cranial ecomorphological adaptations linked to diet have been widely studied in bats, using a variety of anatomical, computational and mathematical approaches. However, foraging-related ecomorphological adaptations and the concordance between cranial and postcranial morphological adaptations remain unexamined in bats and limited to the interpretation of traditional aerodynamic properties of the wing (e.g. wing loading [WL] and aspect ratio [AR]). For this reason, the postcranial ecomorphological diversity in bats and its drivers remain understudied. Using 3D virtual modelling and geometric morphometrics (GMM), we explored the phylogenetic, ecological and biological drivers of humeral morphology in bats, evaluating the presence and magnitude of modularity and integration. To explore decoupled patterns of variation across the bone, we analysed whole-bone shape, diaphyseal and epiphyseal shape. We also tested whether traditional aerodynamic wing traits correlate with humeral shape. By studying 37 species from 20 families (covering all FGs and 85% of dietary guilds), we found similar patterns of variation in whole-bone and diaphyseal shape and unique variation patterns in epiphyseal shape. Phylogeny, diet and FG significantly correlated with shape variation at all levels, whereas size only had a significant effect on epiphyseal morphology. We found a significant phylogenetic signal in all levels of humeral shape. Epiphyseal shape significantly correlated with wing AR. Statistical support for a diaphyseal-epiphyseal modular partition of the humerus suggests a functional partition of shape variability. Our study is the first to show within-structure modular morphological variation in the appendicular skeleton of any living tetrapod. Our results suggest that diaphyseal shape correlates more with phylogeny, whereas epiphyseal shape correlates with diet and FG.

Keywords: Chiroptera; foraging ecology; functional morphology; geometric morphometrics; humerus; modularity.

FIGURE 1

Phylogenetic relationships between sampled taxa based on Shi and Rabosky’s (2015) phylogeny. Branch colours represent foraging guild categories (C= carnivores, P= phytophagous, G = gleaners, H = hawkers, T = trawlers and TL = terrestrial locomotion). Families represented in our sample: Craseonycteridae (Cr), Emballonuridae (Em), Furipteridae (Fu), Hipposideridae (Hp), Megadermatidae (Mg), Miniopteridae (Mi), Molossidae (Ml), Mormoopidae (Mo), Mystacinidae (Mt), Myzopodidae (Mz), Natalidae (Na), Noctilionidae (No), Nycteridae (Ny), Phyllostomidae (Ph), Pteropodidae (Pt), Rhinolophidae (Rl), Rhinonycteridae (Rn), Rhinopomatidae (Rp), Thyropteridae (Th) and Vespertilionidae (Ve). 3D models of humeri illustrate humeral diversity in sampled taxa. Represented taxa clockwise from bottom left to bottom right: Desmodus rotundus, Furipterus horrens, Nycteris grandis, Macroglossus minimus, Myotis daubentonii and Molossus molossus

FIGURE 2

Landmarking protocol used to quantify humeral morphology. From left to right humeri are presented in anterior (far left), medial (centre left), posterior (centre right) and lateral (far right) views. Proximal (top right) and distal (bottom right) epiphyses are also presented. Homologous landmarks are represented by numbers 0‐30 and curves used to place semi‐landmarks are represented by C0‐C5

FIGURE 3

Humeral shape disparity of whole‐bone (left), diaphyseal (centre) and epiphyseal (right) morphology. Shape disparity was decomposed based on foraging guild categories: C = carnivores, P = phytophagous, G = gleaners, H = hawkers, T = trawlers and TL = terrestrial locomotion

FIGURE 4

Morphospace (PCA, a and c) and phylogenetically corrected morphospace (pPCA, b and d) based on whole‐bone shape data. Dot colours represent foraging guild categories (C = carnivores, P = phytophagous, G = gleaners, H = hawkers, T = trawlers and TL = terrestrial locomotion), and polygon colours suborder (purple = Yangochiroptera, yellow = Yinpterochiroptera). Landmark heatmaps of shape change represent magnitude of shape variation across each PC by comparing the minimum and maximum of each component. Humeri 3D models represent position of landmark heatmaps; red colours representing greater variation and yellow colours lower variation

FIGURE 5

Diaphyseal (left) and epiphyseal (right) morphospaces of humeral morphology based on PCAs of shape data. Dot colours represent foraging guild categories (C = carnivores, P = phytophagous, G = gleaners, H = hawkers, T = trawlers and TL = terrestrial locomotion), and polygon colours suborder (purple = Yangochiroptera, yellow = Yinpterochiroptera)

FIGURE 6

Diaphyseal (A and C) and epiphyseal (B and D) phylogenetically corrected morphospaces of humeral morphology based on pPCAs of shape data. Dot colours represent foraging guild categories (C = carnivores, P = phytophagous, G = gleaners, H =hawkers, T = trawlers and TL = terrestrial locomotion), and polygon colours suborder (purple=Yangochiroptera, yellow = Yinpterochiroptera)

FIGURE 7

PLS biplot of first two axes of diaphyseal and epiphyseal shape covariation. Dot colours represent foraging guild categories (C = carnivores, P = phytophagous, G = gleaners, H = hawkers, T = trawlers and TL = terrestrial locomotion)