Morphological diversity of the cetacean mandibular symphysis coincides with novel modes of aquatic feeding
Data files
Jul 11, 2025 version files 94.37 KB
-
Ecological_data.csv
20.19 KB
-
Ecological_data.txt
20.19 KB
-
Elongation(Complete_Estimates)_data.csv
4.73 KB
-
Elongation(Complete_Estimates)_data.txt
4.73 KB
-
Fusion_simplified_data.csv
5.15 KB
-
Fusion_simplified_data.txt
5.15 KB
-
pGLS_data.csv
6.66 KB
-
pGLS_data.txt
6.66 KB
-
README.md
10.97 KB
-
tree_pretty.tre
9.93 KB
Abstract
In whales, extreme modifications to the ancestral mammalian feeding apparatus facilitate novel modes of aquatic feeding. These modifications manifest in morphological diversity across a suite of characters, including the mandibular symphysis. Cetaceans span a range of symphyseal morphologies, with one lineage (crown mysticetes) evolving a highly mobile condition unique among mammals. Here, we use phylogenetic comparative methods to examine the evolution of symphyseal fusion and elongation across 206 extant and fossil cetacean taxa. Ancestral state reconstructions corroborate observations from the fossil record that suggest the ancestral condition for Cetacea was a fused, moderately elongated symphysis. Shifts in symphyseal morphology coincided with ocean restructuring and diversification of feeding modes. Evolutionary rates peaked in the middle-late Eocene and at the Eocene-Oligocene boundary as whales evolved shorter, unfused symphyses. During the Eocene, ankylosed mandibles became less common with the appearance of increasingly pelagic whales. Mysticetes evolved decoupled, highly mobile mandibles near the Eocene-Oligocene boundary. Several odontocete lineages underwent a trait reversal and converged on fully fused, elongated mandibles in the Miocene. Analyses evaluating the influence of ecological variables indicate strong correlations in feeding strategy, dentition, and prey type. The loss of prey-processing behavior and changes to masticatory loading regimes may explain concurrent trends in symphyseal morphology and tooth simplification. We suggest that the functional and morphological diversity of the symphysis in whales is a consequence of aquatic feeding imposing different mechanical constraints than those associated with feeding on land.
Dataset DOI: 10.5061/dryad.q83bk3jvp
Description of the data and file structure
This Dryad repository hosts the supporting data for findings presented by Strauch et al. in “Morphological diversity of the cetacean mandibular symphysis coincides with novel modes of aquatic feeding.” The associated software files (R scripts) are available on Zenodo (https://doi.org/10.5281/zenodo.15742307).
Data hosted by Dryad:
- Ecological_data.txt
- Ecological_data.csv
- Elongation(Complete_Estimates)_data.txt
- Elongation(Complete_Estimates)_data.csv
- Fusion_simplified_data.txt
- Fusion_simplified_data.csv
- pGLS_data.txt
- pGLS_data.csv
- tree_pretty.tre
Software files hosted by Zenodo:
- Ecological_Analyses.R
- Elongation_AncEst_and_Rates.R
- FusionvsElongation.R
- Modified_tree.R
See below for details.
Files
File: Ecological_data.txt
Description: Dataset used in ecological analyses. The categories and values for the five ecological variables (Feeding strategy, Habitat, Dive type, Dentition, and Prey type) follow Coombs et al. (2024). Data for taxa not included in Coombs et al. (2024) were sourced from the primary literature (https://doi.org/10.5281/zenodo.15742309, Table S3).
Variables:
- Taxa: List of extant and fossil cetacean taxa for which morphological and ecological data were collected.
- Group: Phylogenetic group to which a taxon belongs. The phylogenetic groupings are: “Mysticete”, “Odontocete”, and “Archaeocete”. All stem taxa were assigned “Archaeocete.”
- Fusion: Categorical variable. Degree of symphyseal fusion (0-3) observed for a given taxon.
- Elongation: Continuous variable. Symphyseal elongation is the proportion of the mandible that articulates at the symphysis. NA values are assigned to fossil taxa for which incomplete preservation precluded measuring or estimating elongation of the symphysis.
- Feeding strategy: Categorical variable. Taxa were assigned to one of the following modes of aquatic feeding: “Raptorial”, “Suction”, or “Filter.” NA values are assigned to taxa for which there is no data on feeding mode, and feeding mode cannot be reasonably inferred from the morphology.
- Habitat: Categorical variable. Taxa were assigned to one of the following habitat: “Riverine”, “Coastal”, “Coastal-pelagic”, and “Pelagic”. NA values are assigned to taxa for which there is little data available for habitat designation.
- Dive type: Categorical variable. Taxa were assigned to one of the following dive types: “Shallow”, “Mid”, “Deep”, or “Very deep”. NA values are assigned to all fossil taxa, as well as extant taxa for which there is no data on dive depth.
- Dentition: Categorical variable. Taxa were assigned to one of the following: “Edentulous”, “Heterodont”, “Homodont”, or “Reduced”.
- Prey type: Categorical variable. Taxa were assigned to one of the following: “Benthic invertebrates + fish”, “Cephalopods + fish”, “Fish”, “Tetrapods + fish”, or “Zooplankton + fish”. NA values are assigned to taxa for which there is no data available.
- Age: Categorical variable. Geologic age of the first appearance datum of a given taxon. NA values are assigned to taxa for which the age of the first appearance datum is not refined to the age interval.
File: Ecological_data.csv
Description: Dataset used in ecological analyses, as .csv file. Constains the same variables and values as the .txt file of the same name (see above for details).
File: Elongation(Complete_Estimates)_data.txt
Description: Dataset used in the ancestral state reconstruction for elongation.
Variables:
- Species: List of extant and fossil cetacean taxa for which symphyseal elongation was directly measured or estimated.
- Elongation: Continuous variable. Mean elongation value calculated for a taxon. Elongation is defined as the proportion of the mandible that articulates at the symphysis.
File: Elongation(Complete_Estimates)_data.csv
Description: Dataset used in the ancestral state reconstruction of elongation, as .csv file. Contains the same variables and values as the .txt file of the same name (see above for details).
File: Fusion_simplified_data.txt
Description: Dataset used in the ancestral state reconstruction of fusion.
Variables:
- OTU: List of extant and fossil taxa for which data was collected.
- Fusion: Categorical variable. Each taxon was assigned a single state (0, 1, 2, or 3) based on the mode of articulation observed at the symphysis. Taxa exhibiting polymorphism were assigned the state most frequently observed.
File: Fusion_simplified_data.csv
Description: Dataset used in the ancestral state reconstruction of fusion, as .csv file. Contains the same variables and values as the .txt file of the same name (see above for details).
File: pGLS_data.txt
Description: Dataset used to test the relationship between morphological variables (fusion and elongation).
Variables:
- OTU: List of extant and fossil taxa for which elongation could be directly measured or estimated.
- Fusion: Categorical variable. Each taxon was assigned a single state (0, 1, 2, or 3) based on the mode of articulation observed at the symphysis. Taxa exhibiting polymorphism were assigned the state most frequently observed.
- Elongation: Continuous variable. Mean elongation value calculated for a taxon. Elongation is defined as the proportion of the mandible that articulates at the symphysis.
- Group: Phylogenetic group to which a taxon belongs. The phylogenetic groupings are: “Mysticete”, “Odontocete”, and “Archaeocete”. All stem taxa were assigned “Archaeocete.”
File: pGLS_data.csv
Description: Dataset used to test the relationship between morphological variables (fusion and elongation), as .csv file. Contains the same variables and values as the .txt file of the same name (see above for details).
File: tree_pretty.tre
Description: Phylogenetic framework used in the ancestral state reconstructions. Modified version of a tree by Lloyd and Slater (2021). See below for details.
Code/software
Software
R version 4.2.3 (2023-03-15) – “Shortstop Beagle”
Copyright (C) 2023 The R Foundation for Statistical Computing
Software files hosted by Zenodo
Ecological_Analyses.R
Description: Code used to run analyses testing the relationship between morphological (fusion and elongation) and ecological variables. Chi-square tests of independence were used to test the relationship between discrete variables (fusion vs. feeding ecology). Phylogenetic groups (mysticetes, odontocetes, and archaeocetes) were tested separately to account for phylogeny. This approach was chosen because available methods of phylogenetic regression test correlation between binary characters, and all ecological traits examined in this study have more than two states. ANOVAs were used to test the relationship between continuous and discrete variables (elongation vs. feeding ecology).
The data file used in this R script is: Ecological_data.txt
Elongation_AncEst_and_Rates.R:
Description: Code used to perform ancestral state reconstructions via stochastic character mapping, for symphyseal elongation.
Elongation is binned into five character states: SY/SL = (0) 0; (1) 0.04 – 0. 21; (2) 0.21 – 0.37; (3) 0.37 – 0.54; (4) 0.54 – 0.7. This method was chosen because elongation values of crown mysticetes introduce excess zeroes into the dataset. The phytools function fitHRM is used to fit an ordered model of discrete character evolution. The resulting Q-matrix is then used to simulate stochastic character maps.
Evolutionary rate is calculated as the mean number of character transitions within equal-sized time intervals across 1,000 stochastic mapping simulations. To control for the increase in number of lineages in more recent intervals, the number of transitions is divided by the total edge length of each interval.
Data files used in this R script are: Elongation(Complete_Estimates)_data.txt and tree_pretty.tre
Fusion_AncEst_and_Rates.R
Description: Code used to perform ancestral state reconstructions via stochastic character mapping, for symphyseal fusion.
This code performs a discrete character ancestral state reconstruction for symphyseal fusion. The phytools function fitMk is used to fit the following models of discrete character evolution: (1) equal rates (ER), (2) symmetric (SYM), and (3) all rates different (ARD). The SYM model was chosen for the ancestral state reconstruction based on the Akaike information criterion (AIC) and number of assumptions.
Evolutionary rate is calculated as the mean number of character transitions within equal-sized time intervals across 1,000 stochastic mapping simulations. To control for the increase in number of lineages in more recent intervals, the number of transitions is divided by the total edge length of each interval.
Data files used in this R script are: Fusion_simplified_data.txt and tree_pretty.tre
FusionvsElongation.R
Description: Code used to test relationship between morphological variables (fusion and elongation).
This code performs a standard ANOVA to test if different states of fusion significantly differed in elongation. Phylogeny is accounted for by testing archaeocetes, mysticetes, and odontocetes separately. This method was chosen because it minimizes the impact of zero-inflated data.
Data file used in this R script is: pGLS_data.txt
Modified_tree.R
Description: Code used to modify tree by Lloyd and Slater (2021) (see below, under access information).
This code uses the phytools R package to add five extant cetaceans and 25 extinct cetaceans. For fossil taxa, position and branch length are inferred from the primary literature and Paleobiology Database (https://doi.org/10.5281/zenodo.15742309, Table S2). This code also adds six terrestrial artiodactyls, then reroots the tree around artiodactyls as the outgroup.
Supplemental files hosted by Zenodo:
Drivers_MS_Supplementary_Tables_S1-S7.xlsx
Drivers_MS_Supporting_Information.pdf- contains descriptions of supplemntary tables and supplementary figures and figure legends.
Access information
Data used in ecological analyses was derived from a dataset by Coombs et al. (2024), publicly available at: https://doi.org/10.1016/j.cub.2023.11.056
The tree used in analyses is a modified version of a tree by Lloyd and Slater (2021), publicly available at their Github repo: https://github.com/graemetlloyd/ProjectBlackFish