Loss of a morph is associated with asymmetric character release in a radiation of woodland salamanders
Data files
Oct 14, 2025 version files 16.84 MB
-
Morphology.zip
5.86 KB
-
README.md
7.16 KB
-
Supplementary_Tables.zip
16.82 MB
-
Trees.zip
11.84 KB
Abstract
Color polymorphism, the occurrence of multiple discrete color morphs with co-adapted sets of traits within the same population, may provide the raw materials for rapid species formation. It has been hypothesized that fixation of a single morph can result in character release, whereby the monomorphic form evolves without the constraint of accommodating multiple adaptive peaks. However, rates of evolution between populations fixed for different morphs likely depend on the specific adaptive zones occupied by each morph. We studied the evolution of dorsal color polymorphism (striped and unstriped morphs) in woodland salamanders (Plethodon), a North American radiation in which the polymorphism can be found in even the most distantly related species (~44 Ma divergence). We estimated a phylogenomic tree of Plethodon, representing all extant taxa and including multiple samples across the range of most species. Morphometric data suggest that between-species variation exists predominantly along an axis of relative body elongation, corresponding to a terrestrial–fossorial continuum. Polymorphic species occupy an intermediate phenotypic space between the evolutionary optima of striped and unstriped species. Faster rates of body shape evolution were observed in only small-bodied unstriped species, suggesting that body elongation, which is co-adapted with the unstriped morph, could be constrained by the polymorphism. Striped species, but also large-bodied species of eastern Plethodon that lack a dorsal stripe, had slower rates of phenotypic evolution. Our results demonstrate that rates of phenotypic evolution and speciation following character release can be asymmetric and idiosyncratic depending on the alternative adaptations of each morph.
Dataset DOI: 10.5061/dryad.bvq83bknf
Description of the data and file structure
Supplementary Appendix and tables as well as phylogenomic trees and morphological data used for analyses of woodland salamanders (genus: Plethodon). README.txt files below are also included within subfolders which explain data files and their contents.
Discrete character states throughout files refer to whether species were scored as unstriped ("U"), striped ("S"), polymorphic ("S+U"), or other ("other") under a given classification scheme, which account for uncertainty in assignment and are detailed in the publication. Eight schemes used these four color morph states ("Discrete4a" through "Discrete4h"), whereas alternative schemes considered only three states ("Discrete3a" through "Discrete3d") or two states ("Discrete2a" through "Discrete2c").
Supplementary Appendix (on Zenodo)
'Plethodon_polymorphism_Evolution_SI_Appendix.pdf'
Supplementary materials & methods and results
Supplementary Tables ("./Supplementary_Tables.zip/")
'Table_S1_sample_info.xlsx'
Table S1. Sample information, including BioSample accession numbers. Some species names in the Sequence Name and Old Species Name columns reflect names prior to taxonomic revision, and updates are reflected in the New Species Name column (bold). The phylogroup column designates phylogroup membership within species recognized for this study, based on previous phylogeographic studies or on our preliminary phylogenetic analyses. Proportion missing data indicates sample-level missing data in the concatenated matrix of DNA sequences, and the MCMCTree column marks which samples were maintained as representatives for divergence dating of species or phylogroups in MCMCTree. Collection abbreviations: UF = Univeristy of Florida; MVZ = Museum of Vertebrate Zoology; UMMZ = University of Michigan Museum of Zoology; MSB = Museum of Southwestern Biology.
'Table_S2_per_locus_statistics.xlsx'
Table S2. Summary statistics for anchored hybrid enrichment data. Number of taxa includes the number of samples with any data available (i.e. did not have completely missing data).
'Table_S3_raw_morphology.xlsx'
Table S3. Raw morphological data and eight discrete color morph classifications. Costal groove counts (CG) were obtained for Fisher-Reid and Wiens (2015) except where noted in the text. Sample size (N) reflects the numbers used for all remaining morphometric variables (measured in millimeters), which represent weighted averages from means and sample sizes reported in Baken and Adams (2019) or Adams et al. (2009) except where noted in the text. SVL = snout-vent length; S.E = snout-eye length; HL = head length; BW = body width; FLL = forelimb length; HLL = hindlimb length; TL = tail length. Values with "NA" were phylogenetically imputed after continuous measurements were standardized by SVL and log-transformed.
'Table_S4_discrete_character_model_selection.xlsx'
Table S4a. Model selection tables for 12 discrete character evolution models across eight color classification schemes (4a-h). Model names are consistent with Supplementary Figure S3 in the SI Appendix.
'Table_S5_Plethodon_phylosignal.xlsx'
Table S5. Phylogenetic signal and associated p-values using Pagel's lambda and Blomberg's K. All variables except costal groove count (CG) were divided by snout-vent length and log-transformed to correct for body size. S.E = snout-eye length; HL = head length; BW = body width; FLL = forelimb length; HLL = hindlimb length.
'Table_S6_pPCA_loadings.xlsx'
Table S6. Phylogenetic PCA loadings for six morphometric traits. All traits except CG were body-size-corrected (divided by snout-vent length and log-transformed).
'Table_S7_OUwie_3cat_2cat.xlsx'
Table S7. Rates of phenotypic evolution and lineage diversification rates across color morph states using three or two color morph categories. Phenotypic optima (θ) and rates (σ2) were estimated using OUwie for the first axis of phylogenetic PCA, representing mean values and standard errors (SE) for 100 stochastic character maps repeated across eight color classification schemes. For the 3 Morph Categories optima, four simulations yielding extreme outlier optima (>100) were omitted prior to calculating means and errors. Lineage diversification rates, or tip rates, were estimated using the DR statistic (λDR). Mean λDR is the mean value for species across all eight classification schemes, while the mean range provides the lowest and highest mean values across schemes. The min λDR and max λDR represent the minimum and maximum tip rates, respectively, observed across species within each color morph state for any scheme.
Plethodon phylogenomic trees ("./Trees.zip/")
IQ-Tree and wASTRAL
'plethodon-cat.treefile'
Concatenated IQ-Tree analysis of all 198 samples
'wASTRAL-plethodon.tree'
Species tree from wASTRAL with samples assigned to species, phylogroups, or polyphyletic/paraphyletic units
Ultrametric trees
'Plethodon-MCMCtree-10Ma.tre'
Time-calibrated MCMCtree "Figtree" output tree. Includes 95% HPD. *Branch lengths are in units of 10 Ma!
Since this file is not readily usable for in most programs for comparative analyses, additional files were prepared below.
'plethodon-unpruned.tree'
Time-calibrated topology taken from 'Plethodon_MCMCtree_10Ma.tre', but with branch units converted to 1 Ma (no HPD included). Can be easily imported into R or other software for comparative analyses.
'plethodon-pruned.tree'
Same tree as 'plethodon-unpruned.tree', but intraspecific groups (e.g. phylogroups or polyphyletic samples) were pruned so that each named species is represented by a single tip.
Morphological data ("./Morphology.zip/")
'Plethodon_morphology_R_Processed.csv'
Includes morphological data used for comparative analyses.
Note that measurements represent averages at the species level from other publications (see manuscript for details), including a weighted average across studies based on sample sizes provided in those publications.
Columns
CG = costal groove count
N = sample size
SVL = snout-vent length (mm)
S.E = snout-eye length (mm)
HL = head length (mm)
BW = body width (mm)
FLL = forelimb length (mm)
HLL = hindlimb length (mm)
TL = tail length (mm; not used in this study)
Discrete2a Discrete2b Discrete2c = 2-state morph classifications (see text for description)
Discrete3a Discrete3b Discrete3c Discrete3d = 3-state morph classifications (see text for description)
Discrete4a Discrete4b Discrete4c Discrete4d Discrete4e Discrete4f Discrete4g Discrete4h = 4-state morph classifications (see text for description)
S.E.sc = snout-eye length (divided by SVL and log transformed)
HL.sc = head length (divided by SVL and log transformed)
BW.sc = body width (divided by SVL and log transformed)
FLL.sc = forelimb length (divided by SVL and log transformed)
HLL.sc = hindlimb length (divided by SVL and log transformed)
