“Geometric Morphometric” Approach To Detect Body Shape Variations among Three Indian Shads
-
Kritish De [1] ; Arvind Kumar Dwivedi [2]
-
[1] Sri Sathya Sai University for Human Excellence, Navanihal
-
[2] The Assam Royal Global University, Guwahati, Assam
-
- Localización: Thalassas: An international journal of marine sciences, ISSN 0212-5919, Vol. 41, Nº. 4, 2025
- Idioma: inglés
- Texto completo no disponible (Saber más ...)
-
Resumen
Morphometric traits are essential for characterizing and classifying taxonomic groups. Traditional Morphometrics (TM) and the Truss Network System (TNS) have been employed to differentiate the sympatric Indian shad species (Tenualosa ilisha, T. toli and Hilsa kelee), though both methods have inherent limitations. In this study, digital images from 120 specimens of the three Indian shad species were analyzed using the Geometric Morphometrics (GM) approach to investigate body shape variations. Superimposition of the images successfully removed size-related effects. Procrustes ANOVA revealed a significant correlation between centroid size and shape (p < 0.0001). Canonical Variate Analysis (CVA) and Discriminant Function Analysis (DFA) demonstrated 100% differentiation among the species, which can be due to over-fitting because of limited sample size of H. kelee (6) compared to T. ilisha (86) and T. toil (28). Hence, leave-one-out cross-validation was performed in the DFA to nullify the effect of over-fitting. Mahalanobis and Procrustes distances derived from CVA were also statistically significant (p < 0.001), indicating notable differences in body shape among the species. Morphological Similarity Dendrogram (MSD) clustered the species based on morphological similarity, while wireframe analysis clearly illustrated interspecific differences in body shape. Though the findings highlight the GM approach as a rapid and effective tool for distinguishing Indian shads and offering valuable insights beyond those provided by TM and TNS, its scope is limited only to body shape variations. Therefore, an integrative approach that combines TM, TNS, GM and genetic tools is recommended to validate the findings of this study
