Data and code from: Intraspecific morphological divergence among chemosymbiotic bivalves (Lucinidae) from anchialine lakes of San Salvador Island, Bahamas
Data files
Jul 17, 2026 version files 323.65 KB
-
KokeshEtAl2026_Classifier.csv
9.98 KB
-
KokeshEtAl2026_Geochem.csv
1.70 KB
-
KokeshEtAl2026_Landmarks.TPS
290.67 KB
-
KokeshEtAl2026_Links.csv
298 B
-
KokeshEtAl2026_RCode.R
16.78 KB
-
KokeshEtAl2026_Semilandmarks.csv
356 B
-
README.md
3.86 KB
Abstract
Lakes can serve as cradles of biodiversity by facilitating adaptive radiation linked to distinct environmental conditions and ecological opportunities. On San Salvador Island, the Bahamas, karstification during Holocene transgressive flooding created anchialine lakes that have supported rapid evolutionary diversification, most notably among endemic pupfishes with specialized feeding strategies. To assess whether similar phenotypic divergence occurs in other members of these lake communities, we investigated morphological variation in shells of the chemosymbiotic infaunal bivalve Ctena sp. (Lucinidae). Live and dead Ctena specimens were sampled from muddy sediment filling karst pockets along the shorelines of three anchialine lakes, a habitat type from which chemosymbiotic bivalves have not previously been reported. Using two-dimensional geometric landmarks of shell interiors and valve margins, we found significant differences in shell shape, size, and allometric relationships among living populations from each lake. Death assemblages preserved similar among-lake patterns, suggesting that morphological divergence predates the window of time averaging represented by death assemblages. Living and dead-shell assemblages within each lake also differed significantly in shell shape, and temporal shape-change vectors from the dead to living included both shared and distinct trends among lakes.
Ongoing genomic work suggests that lake-dwelling Ctena on San Salvador represent a single endemic species. Thus, spatiotemporal patterns among lake populations likely reflect ecophenotypic responses to distinct sedimentological and geochemical conditions that influence their chemosymbiotic lifestyle. These results demonstrate the potential of combined life and death assemblage analyses to detect patterns of phenotypic divergence over geologically instantaneous timescales.
Dataset DOI: 10.5061/dryad.nk98sf88k
Description of the data and file structure
Live- and dead-collected specimens of articulated (both valves attached and intact) Ctena sp. were sampled from lakebed sediments via hand excavation and box sieving in the field. Dried and empty shells are archived at the Museum of Geology at South Dakota School of Mines and Technology (SDSM). From each of the three lakes plus two open ocean locations accessible from the shoreline (Graham's Harbour north of the Gerace Research Centre and the mouth of Pigeon Creek), porewater samples were opportunistically taken for measurements.
Files and variables
File: KokeshEtAl2026_RCode.R
Description: R code used to conduct all analyses in this study.
File: KokeshEtAl2026_Geochem.csv
Description: Spreadsheet of porewater geochemistry measurement data
Variables
- sample: Sample ID
- year: Year sample was taken
- location: Lake or open marine body form which samples were taken
- pH: The measured pH value from porewater samples, measuring the concentration of H+ ions
- pOH: The measured pOH value from porewater samples, measuring the concentration of OH- ions
- temp: The measured temperature of porewater samples in degrees Celsius
- conductivity: The measured conductivity in microSiemes per centimeter
- DO: The measured dissolved oxygen concentration in milligrams per liter
- H2S: The measured sulfide concentration in milligrams per liter
- HCO3: The measured alkalinity as bicarbonate concentration in milligrams per liter
- box_plot_order: A column used to organize locations in a desired order when generating boxplots in R
File: KokeshEtAl2026_Landmarks.TPS
Description: Digitized landmark data from imaged shell specimens, including six fixed landmarks and 35 sliding semilandmarks
File: KokeshEtAl2026_Semilandmarks.csv
Description: A helper data frame to identify semilandmarks in the analysis
Variables
- before: The first of three landmarks forming a chain
- slide: The middle of three landmarks forming a chain
- after: The last of three landmarks forming a chain
File: KokeshEtAl2026_Links.csv
Description: A helper data frame used to identify which landmarks should be connected by a line segment in data visualization
Variables
- Column_1: The first of two landmarks to draw a segment between
- Column_2: The second of two landmarks to draw a segment between
File: KokeshEtAl2026_Classifier.csv
Description: Master data frame identifying the lake and assemblage of each shell specimen
Variables
- Specimen: The specimen ID
- lake: The lake a specimen was sampled from
- state: The assemblage a specimen belongs to, being either live- or dead-collected
- group: A combined factor of lake and life state variables for data analysis
Code/software
All data analyses were conducted in version 4.5.0 of the R programming environment, which is free and openly available software. The provided code script is written to generate all statistical tests and plots presented in the manuscript. The packages 'geomorph' and 'Morpho' provide specialized functions for morphometric analyses, and the package 'vegan' aided in conducting geochemical analyses.
Access information
Other publicly accessible locations of the data:
- Physical shell specimens from which these data derive are archived at the Museum of Geology of South Dakota School of Mines and Technology (museum code SDSM).
Data was derived from the following sources:
- All data are original and were collected for the purposes of this and future studies by the authors.
