------------------- GENERAL INFORMATION ------------------- 1. Title of Dataset: Range-Wide North American Freshwater Mussels (Unionidae, Ambleminae) 2D and 3D Morphometric Datasets 2. Authors: Andy D. Y. Tan, Yu Kai Tan 3. Contact information: andy.tan@colorado.edu 4. Funding information: This dataset is partially funded through the bequests of Henry Monmouth Smith and Harold T. Stearns at Wesleyan University and the Kenneth Jay Boss Fellowship provided by the Smithsonian National Museum of Natural History for supporting research in USNM collections 5. Date of data collection: 2020 - 2024 -------------------------- SHARING/ACCESS INFORMATION -------------------------- 1. Licenses/restrictions placed on the data: NA 2. Links to publications that cite or use the data: In review 3 Recommended citation for the data: Tan, A.D.Y., Tan. Y. K. (2025) Range-Wide North American Freshwater Mussels (Unionidae, Ambleminae) 2D and 3D Morphometric Datasets. DOI: http://doi.org/10.25810/qs9s-8t32 --------------------- DATA & FILE OVERVIEW --------------------- 1. File and directory list: A. Filename: RENT_2025VII_2D.xlsx B. Filename: RENT_2025VII_3D.xlsx C. Filename: FigS1.png D. Filename: FigS2.png -------------------------- METHODOLOGICAL INFORMATION -------------------------- 1. Description of methods used for collecting, generating, and processing data: To explore patterns of shell variation we generated a 2D and a 3D dataset using linear measurements and structured light morphometry, respectively. Our 2D dataset contains data on 1169 animals from 10 species. We included representatives from three of the five Ambleminae tribes, with Cyprogenia stegaria, Dromus dromas, Obliquaria reflexa and Obovaria subtrotunda representing Lampsilini, Fusconaia flava, Fusconaia subrotunda and Plethobasus cooperianus in the Pleurobemini, and Pustulosa pustulosa, Theliderma metanevra and Quadrula quadrula in the Quadrulini (Appendix S1: Table S2 and Fig S1). These focal species were selected for good representation in the major natural history collections, are distributed along a wide environmental gradient, and have similar circular to ovate shell outlines. Specimens were selected from 1) Invertebrate Zoology Collection of the Smithsonian Institution’s National Museum of Natural History (USNM), Washington D.C., USA, 2) Mollusk Division of the University of Michigan Museum of Zoology (UMMZ), Ann Arbor, MI, USA, 3) Wesleyan Natural History Collections (WNHC) in the Joe Webb Peoples Museum, Wesleyan University, Middletown, CT, USA, and 4) the Academy of Natural Sciences of Drexel University (ANSP), Philadelphia, PA, USA. We verified the georeferenced coordinates of all specimens or revised them if necessary. Specimens in one lot are linked through catalogue number (Cat_No) which associates shared collecting event data (time, locality, collector etc.). Using the nhdplusTools R package, longitude and latitude for each sample was mapped to the nearest flowline feature, through association of the common identifier (ComID), to extract hydrological data (annual mean river flow rate in (ft3 s-1) from the National Hydrography Dataset Plus (NHDPlusV2; (Blodgett et al. 2024)), using hydrology, temperature, and precipitation data from 1971 to 2000 (U.S. Geological Survey National Geospatial Program 2022). Flow rates were log-transformed for downstream analyses. Given the nested, hierarchical nature of the HUC system, we extracted from HUC12s the six levels of progressively smaller hydrologic units representing region (HUC2), subregion (HUC4), basin (HUC6), subbasin (HUC8), watershed (HUC10) and subwatershed (HUC12). Our sample spans eight HUC2 water resource regions: Arkansas-White-Red Region, Great Lakes Region, Lower Mississippi Region, Missouri Region, Ohio Region, South Atlantic-Gulf Region, Tennessee Region, and Upper Mississippi Region The species are categorised into three groups based on sculpturing habits: always sculptured (AS, species that typically produce sculpture), facultatively sculptured (FS, species with highly plastic levels of sculpture), or smooth-shelled (species that are typically unsculptured). Our 3D dataset includes 53 Fusconaia subtrotunda (smooth-shelled), and 57 Pustulosa. pustulosa (sculptured). ** Linear morphometry** All linear measurements on specimens were measured to ±0.01 mm with digital callipers (2D Dataset), and include the following: length, for the maximum anterior-posterior dimension parallel to the hinge ligament; inflation, the maximum lateral dimension on the profile side across two valves, taken ~90° to width. Shell dry mass was recorded to ±0.01 g. We then calculated inflation index (ratio of inflation corrected by shell length; Fig 1b) to test for inflation change in response to river flow rate. 3D Scanning and Volumetric Morphometry 3D models of specimens were created using structured light scanning at a resolution of 0.05 mm using an Artec™ Space Spider 3D Scanner and its corresponding Artec™ Studio 15 and 16 Professional at Wesleyan University. As the scanning method is non-penetrative, both valves were scanned separately and combined into a single model. Specimens were recorded at true size, allowing linear, surface and three-dimensional measurements to be taken in the future. All digitised models were created without running the Mesh Simplification command, to retain fine sculpturing in the final model. To quantify the allocation of shell materials for both species across varying river flow regime, all specimens were scanned with two treatments: (1) both valves closely attached, producing a model of the exterior of the shell, to obtain the volume of valves plus internal cavity, henceforth known as the Filled-In Model (Vfim), (2) both valves of each specimen scanned separately and aligned to the scans from the first treatment, to obtain the volume of the shell (Vsh). Using these 3D models, volumes were measured using the built-in function in Artec Studio 16 Professional and verified with Autodesk Meshmixer 3.5. The output volume measurement is in mm3. The volume of the internal cavity enclosed by the valves (Vst, a proxy for somatic volume or soft tissue; Larson et al. 2014) was then estimated by subtracting Vfim from Vsh (V_st=V_fim-V_sh; Fig 1c). The ratio of shell volume to the soft tissue volume (S/T ratio) is calculated by V_sh/V_st to study ecophenotypic covariances in resource allocation. We consider differential volumetric ratio responses by sculptured and smooth shelled species proxies for resource allocation for either biomineralization or soft tissue growth. ----------------------------------------- DATA-SPECIFIC INFORMATION FOR: RENT_2025VII_2D.xlsx ----------------------------------------- 1. Number of variables: 29 2. Number of cases/rows: 1242 3. Variable List: See README sheet within the dataset file for variable dictionary. ----------------------------------------- DATA-SPECIFIC INFORMATION FOR: RENT_2025VII_3D.xlsx ----------------------------------------- 1. Number of variables: 23 2. Number of cases/rows: 110 3. Variable List: See README sheet within the dataset file for variable dictionary. ----------------------------------------- DATA-SPECIFIC INFORMATION FOR: FigS1.png ----------------------------------------- Figure S1: Specimen counts by species, tribe, and sculpturing habit for this dataset submission. ‘Count’ represents the total number of individuals examined, alongside the subset used for 2D and 3D morphometric datasets. Abbreviations: AS, always sculptured; FS, facultatively sculptured; UN, smooth-shelled (always unsculptured). ----------------------------------------- DATA-SPECIFIC INFORMATION FOR: FigS2.png ----------------------------------------- Figure S2: Freshwater mussel species in this dataset, categorized by sculpturing habits: (top panel left) facultatively sculptured species, (top panel right) always sculptured species, (bottom panel) smooth-shelled species, with total specimen count for dataset generation included for each species.