10.5061/DRYAD.3J9KD51FT
Figueirido Castillo, Francisco Borja
0000-0003-2542-3977
Universidad de Malaga
Martín-Serra, Alberto
Universidad de Malaga
Pérez-Ramos, Alejandro
0000-0003-1417-4338
Universidad de Malaga
Velasco, David
Universidad de Malaga
Pastor, Francisco
Universidad de Valladolid
Benson, Roger
University of Oxford
Serial disparity in the carnivoran backbone unveil a complex adaptive role
in metameric evolution
Dryad
dataset
2020
disparity
metameric structures
constrain
ecomorphology
Vertebral column
Ministerio de Ciencia e Innovación
https://ror.org/05r0vyz12
CGL-2015-68300P
Ministerio de Ciencia e Innovación
https://ror.org/05r0vyz12
CGL-2017-92166-EXP
European Union’s Horizon 2020 research and innovation program*
European Research Council
https://ror.org/0472cxd90
TEMPO-677774
European Union’s Horizon 2020 research and innovation program
2020-08-10T00:00:00Z
2020-08-10T00:00:00Z
en
967935 bytes
1
CC0 1.0 Universal (CC0 1.0) Public Domain Dedication
Multi-element systems such as the vertebral column of vertebrates
represent a major challenge to phenotypic quantification and
macroevolutionary analyses. The vertebral column is a metameric structure,
composed of serially repeated subunits, and much of what is known so far
has been inferred from sparse anatomical samples, providing little insight
into local-scale (i.e. vertebra-to-vertebra) variation and its
macroevolutionary importance. This limits understanding of how
evolutionary constraints and functional adaptation interact during the
evolution of multi-element phenotypes. Here, we quantify morphological
disparity across all subunits (vertebrae) of the pre-sacral column in the
mammalian order Carnivora. We address how vertebral morphology varies
among elements, and the extent to which these patterns have been
structured by constraints and/or evolutionary adaptation to locomotory
capabilities, using 3D geometric morphometrics and multivariate analyses
for high-dimensional phenotypes. We find that lumbars and posterior
thoracics exhibit high individual disparity but low serial
differentiation. These vertebrae are pervasively recruited into locomotory
functions, exhibiting high-dimensional ecomorphological signals and
patterns of evolution indicative of relaxed constraints. Cervical and
anterior thoracic vertebrae have low individual disparity and greater
serial differentiation. Individual vertebrae in these regions unexpectedly
also show signals of locomotory adaptation that were not generally
recognized by previous studies. These are characterized by low-dimensional
ecomorphological signals and overall constrained patterns of evolution.
Our findings support the hypothesis that the lumbar region is a key
innovation that increases ecological versatility of mammalian locomotion.
Nevertheless, locomotory adaptation is more widely distributed along the
mammalian axial skeleton. This has been masked by local-scale variation
and low phenotypic variability in comparison with other skeletal
structures such as the skull or limbs. Our analyses demonstrate that the
strength of ecomorphological signal does not have a predictable influence
on macroevolutionary outcomes even within the same structure, and
undermine the traditional view that highly constrained skeletal units are
strongly limited in their potential to adapt to new ecological avenues.
Our findings emphasize the importance of quantifying local-scale variation
in functionally versatile, multi-element phenotypes such as the vertebral
column, or indeed, the vertebrate skeleton as a whole.
All presacral vertebraewere scanned in 3D with either micro-computed
tomography (CT) scanning or a NextEngine® surface scanner. CT scans were
segmented in Avizo. This resulted in 3D models of 1097 vertebrae. The 3D
models were imported into Meshlab (Cignoni et al. 2008) to reduce the size
of the models. In addition, we used the software Netfabb to edit the 3D
models and to separate each vertebrae of a given vertebral column. To
capture the morphology of the vertebrae, we digitized 34 homologous
landmarks on the cervical vertebrae (C03-C07), 32 homologous landmarks on
the thoracic vertebrae (T01-T14), and 36 homologous landmarks on the
lumbar vertebrae (L1-L7) (see Fig. 2; Table S2). The landmarks were
digitized with the software Landmark from IDAV (Wiley et al. 2005) and the
x,y,z coordinates of each landmark were exported as a Text file. Raw
landmark data were imported into the R package geomorph version 3.1.0
(Adams et al. 2019). Each region of the vertebral column (cervical,
thoracic, and lumbar) was separated into different datasets, and
therefore, all subsequent analyses were performed independently.