https://doi.org/10.5061/dryad.h70rxwdqk

Linear morphometrics taken from images of dinosaur skulls, along with phylogeny analysed using variable-rate phylogenetic regression modelling in BayesTraits.

Description of the data and file structure

---- [[Data-Primary.xlsx]] ----

Primary dataset file containing raw measurements, associated annotations and sources, as well as look-up table to match taxa in data and in the tree files. The file contains the following tabs:

[Source tab]\
List of images used for data collection, their sources (specimen number, reference), taxonomy, scale info.

Description of variables

• Taxon: Linean binomial
• Specimen: specimen number or unique identifier
• Image file name: the name of the image file
• Order: taxonomic order
• Scale (mm): the scale used to calibrate the image file
• Scale type: this denotes whether the scale used is a scale bar or a morphometric (typically skull length) with a known length

[Data tab]
Data for each taxon. Empty cells represent missing data (function in for_R tab fills this with NA).

Description of variables

• L_Sk: linear length of the skull (mm)
• L_Toothrow: linear length of the biting edge length taken as the sum of the lengths along the premaxilla and maxilla (mm)
• L_Orbit: linear diameter of the orbit (mm)
• W_Skull: linear width of the skull (mm)
• L_DentBattery: linear length of the dental battery if present (mm)
• d_Bite: linear distance between the jaw joint and the posterior-most biting position on the biting edge (mm)
• Beak: 0|1 dummy variable coding for presence/absence of edentulous beaks

[match_Tree tab]
Look up table to match taxon names between morphometric dataset and tip labels in the sample of trees of Lloyd et al. (2016).

Description of variables

• Name_in_Data: Taxon name used during data collection
• Name_in_Tree: Taxon name as it appears on the Lloyd et al. (2016) tree
• Name_to_Match: This column is used to match the data and the tree. This is by default the name in the data but if there is a different corresponding name in the tree then that takes prescedence - this is determined by an IF function (e.g., =IF(B2="",A2,B2)).
• Name_to_Use: This column is ultimately the name to be used in the final analysis as it accounts for outdated names in the tree.
• USE: 0|1 binary coding to denote whether the taxon will be used in the final analysis.

[for_R]
Description of variables. Cell values are pulled from the Data tab and blank cells are filled with NAs.

• Name_in_Tree: taxon name that matches to the tip label of the Lloyd et al. (2016) trees.
• Use: 0|1 binary flag indicating which rows to use in the analyses

---- [[Toothrow linear and arc lengths.xlsx]] ----

Supplementary data file to demonstrate the relationship between linear biting edge length and the arc length of the ventral curvature of the tooth row. The sheet contains a scatterplot with a regression line fitted to demonstrate the 1:1 relationship, i.e., there is no discernible bias in taking the linear measurement over the arc length. Blank cells represent uncollected data and are left blank as the graphing function in Excel is incapable of handling NA values.

• Taxon: Linean binomial
• Specimen: specimen number or individual identifier for the source of the image
• Scale (mm): the known distance the scale was calibrated against in ImageJ.
• Scale type: type of scale used to calibrate image - scale bar (scale bar on image); skull length (where scale bar is either absent or unreliable so independent measurement of skull length was taken from literature and used to scale image)
• L_Pmx: linear length of tooth row along premaxilla (mm)
• L_Mx: linear length of tooth row along maxilla (mm)
• L_Tooth: sum of L_Pmx and L_Mx (mm)
• C_Tooth: arc length of the ventral curvature of the tooth row, taken as the sum of a series of linear distances between landmarks placed at the tips of the teeth along the tooth row (mm)
• log10_L_Tooth: log10 transformed L_Tooth
• log10_C_Tooth: log10 transformed C_Tooth --NOTE: L_Pmx, L_Mx, L_Tooth and C_Tooth were remeasured for a subset of the theropod dinosaurs (those with tooth rows instead of beaks) by MS for the sole purpose of comparing measurement discrepancies between linear biting edge length (L_Tooth) and arc lengths of the ventral curvature of the tooth row (C_Tooth). The original data used in the main analyses were measured by CK. Slight variation in measurement owe to differences between MS and CK. log10_C_Tooth was plotted against log10_L_Tooth and a trend line was fitted. The slope of 1.015 is indicative of an isometric relationshp between log10_C_Tooth and log10_L_Tooth, meaning that the two measurement of tooth row lengths scale 1:1, which the intercept of 0.0127 indicates that log10_C_Tooth is on average 1.03 times larger than log10_L_Tooth, i.e., roughly 3% over-estimate the latter. This discrepancy is minimal and within margin of error, particularly with measurements dealing with fossil organisms.

---- [[data.txt]] ----

This file is the data file that was fed into BayesTraits.

• Taxon: Linnean binomial
• Order: Grouping variable denoting one of three dinosaurian clades, Theropoda, Sauropodomorpha, and Ornithischia
• L_Sk: linear length of the skull (mm)
• L_Pmx: linear length of tooth row along maxilla (mm)
• L_Bite: linear length of the biting edge length taken as the sum of the lengths along the premaxilla and maxilla (mm)
• d_Bite: linear distance between the jaw joint and the posterior-most biting position on the biting edge (mm)
• L_Orbit: linear diameter of the orbit (mm)
• Beak: 0|1 dummy variable coding for presence/absence of edentulous beaks
• logLSk: log10 transformed L_Sk
• logLPmx: log10 transformed L_Pmx
• logLBite: log10 transformed L_Bite
• logdBite: log10 transformed d_Bite
• logLOrbit: log10 transformed L_Orbit

---- [[trees.nex]] ----

Nexus file containing 100 time-scaled trees from Lloyd et al. 2016 - this is also used as input for BayesTraits.

Sharing/Access information

Data was derived from the following sources:

• Sources are listed in Data-Primary.xlsx

Code/Software

BayesTraits V4: https://www.evolution.reading.ac.uk/BayesTraitsV4.0.1/BayesTraitsV4.0.1.html