Dataset DOI: 10.5061/dryad.4qrfj6qr1

Description of the data and file structure

This data package contains the data, R scripts, and results files associated with "Clade-specific drivers of rapid vertebral evolution in squamates”. Radiographs used in the research are available at the University of Michigan’s Digital Library (https://quod.lib.umich.edu/r/rept1ic).

Files and variables

File: Evolution-vertebrae-data-files.zip

Description: 

R scripts: All scripts begin by setting up the data. A description of what each script does is given below and each script contains further annotations. 

• analysis-1-broad-summary-stats: This script runs all the statistical analyses by major grouping (snakes, elongate lizards, and non-elongate lizards). Phylogenetic regressions to test for a relationship between vertebral number and body size (SVL) and phylogenetic ANOVAs to test for a relationship between vertebral number and habitat.
• analysis-2-variance-families: This script tests whether elongate and non-elongate families differ in amount of variance in vertebral number and average vertebral length using t-tests.
• analysis-3-svl-clades: This script tests for a relationship between vertebral number and SVL in 18 squamate clades.
• analysis-4-ecology: This script tests for a relationship between vertebral number and habitat within clades.
• analysis-5-phylogenetic-signal: This script tests for phylogenetic signal of vertebral number and svl for each clade.
• analysis-6-r2-variance: This script calculates R2s for a base phylogenetic model and a best fit model (that may include body size and/or habitat) to assess how much of the variance in vertebral number is explained by phylogeny, body size, and/or habitat.
• Figure2-7 scripts: These six scripts are for plotting the six main figures in the manuscript, excluding the conceptual diagram (Figure 1). 

Other files in the folder:

• squamates-vertebrae-data.csv: csv file containing data on number of presacral vertebrae for all species in paper. The columns are:
  • tree: species name as present on the phylogeny from Title et al. (2024)
  • family: name of family
  • dataType: GenBank or genomic, this column is taken from the Title et al. (2024) data and indicates which data were used for this species in reconstructing the tree.
  • numberPresacralVert: number of presacral/precloacal vertebrae
  • source: whether the count from the previous column came from a direct count of vertebrae or count of ventral scales
  • reference: reference for vertebral count. Complete citations can be found in the Supplementary Materials. my-counts indicates that the authors counted the vertebrae.
  • completeSVL: snout-vent length, in mm. Source: Title et al. (2024)
  • elongationIndex: a metric for elongation that treats squamates as idealized cylinders and is calculated from SVL and mass. Source: Title et al. (2024)
  • type: elongation category based on elongation index (EI). Options are non-elongate lizard (if EI < 250), elongate lizard (if EI > 250), and snake (if snake).
  • eco: habitat classification for each species, source: Meiri (2024). Habitats include aquatic, arboreal, fossorial, marine (used for sea snakes and kraits), saxicolous, and terrestrial.
  • ecologyNotes: notes relevant to habitat classification, particularly noting where a species name differed between Title et al. (2024) and Meiri (2024) or if we assigned the species ecology based on the ecologies of the rest of the genus.
• tier1_ultrametric_fulltree_ddBD.tre: text file containing the Title et al. (2024) squamate phylogeny.

The /statistics folder contains all results from the analyses. The files include:

• habitat folder: contains text files with results from phylogenetic ANOVAs testing for a relationship between habitat and number of precloacal vertebrae in non-elongate lizards, elongate lizards, and snakes.
• habitat-clades folder: contains text files with results from phylogenetic ANOVAs testing for a relationship between habitat and number of precloacal vertebrae in 15 clades.
• clade-habitat-results.csv: summary of results from phylogenetic ANOVAs testing for a relationship between habitat and number of precloacal vertebrae in 15 clades. Pairwise comparisons and full results are in the habitat-clades folder. The column headings are:
  • Clade: name of clade
  • NumberSpecies: number of species
  • PValue: p-value from ANOVA
• deltaAIC-r2-results.csv: table of deltaAIC values from model fitting prior R2 decomposition. Linear regression models were fit to each clade (18 clades total) and evaluated using AIC scores.
  • Clade: name of clade.
  • Columns B-F list the different models tested for each clade, with the deltaAIC value below. NA indicates that the model was not tested for the clade (specifically, models with habitat were not tested for clades with only one habitat state like blind snakes).
  • Best Model: lists the best fitting model - the model with the lowest AIC score (or the simpler model if two models were within a deltaAIC score of 2 from each other).
• phylogenetic-signal.csv: summary of results from phylogenetic signal analyses for each clade.
  • Clade: name of clade
  • NumberSpecies: number of species
  • VertPhySignalLambda: lambda value from phylogenetic signal analysis on precloacal vertebral number
  • VertPhySignalLP: p-value from phylogenetic signal analysis on precloacal vertebral number
  • SizePhySignalLambda: lambda value from phylogenetic signal analysis on snout-vent length
  • SizePhySignalLP: p-value from phylogenetic signal analysis on precloacal snout-vent length

• r2-model-results.csv: summary of results from R2 analyses for each clade
  • Clade: name of clade
  • NumberSpecies: number of species
  • Model: name of best fit model
  • R2full: value of R2 from the best fit model
  • R2phy: value of R2 phy from analysis of the phylogeny-only model.
  • R2trait: value of R2 for the traits (body size and/or habitat depending on which variables are present in model) once phylogeny is accounted for.
  • f2phy: Cohen's F2 value for R2 phy.
  • f2trait: Cohen's F2 value for R2 trait.

• summary-vert-svl-stats.csv: summary of results from phylogenetic regressions testing for a relationship between body size and number of precloacal vertebrae in broad groups (non-elongate lizards, elongate lizards, and snakes).
  • Group: name of broad group
  • N: number of species sampled in the clade
  • t: t-value from regression
  • R2: R^2 value from regression
  • p: p-value from regression

• vert-svl-clade-stats.csv: summary of results from phylogenetic regressions testing for a relationship between body size and number of precloacal vertebrae in 18 squamate clades, using a Brownian motion model.
  • Group: name of clade
  • N: number of species sampled in the clade
  • t: t-value from regression
  • R2: R^2 value from regression
  • p: p-value from regression
• vert-svl-clade-stats-lambda-model.csv: summary of results from phylogenetic regressions testing for a relationship between body size and number of precloacal vertebrae in 18 squamate clades, using Pagel’s λ transform of the phylogenetic variance covariance matrix.
  • Group: name of clade
  • N: number of species sampled in the clade
  • t: t-value from regression
  • R2: R^2 value from regression
  • p: p-value from regression

Code/software

R version 4.3.1 was used to run all analyses and create all figures. Annotations are provided in all scripts to guide data loading, analyses, and figure creation. The following packages are required:

• ape version 5.7.1
• phylolm verson 2.6.2
• phytools version 1.9.16
• rr2 version 1.1.1