Geometric morphometric analysis of ...
Type de document :
Compte-rendu et recension critique d'ouvrage
DOI :
Titre :
Geometric morphometric analysis of Pleuronectiformes vertebrae: A new tool to identify archaeological fish remains?
Auteur(s) :
Dierickx, Katrien [Auteur]
Oueslati, Tarek [Auteur]
Histoire, Archéologie et Littérature des Mondes Anciens - UMR 8164 [HALMA]
Profico, Antonio [Auteur]
Oueslati, Tarek [Auteur]

Histoire, Archéologie et Littérature des Mondes Anciens - UMR 8164 [HALMA]
Profico, Antonio [Auteur]
Titre de la revue :
Journal of Anatomy
Pagination :
982-996
Éditeur :
Wiley
Date de publication :
2023-07-26
ISSN :
0021-8782
Mot(s)-clé(s) en anglais :
Geometric morphometric, Pleuronectiformes, vertebrae, archaeology, fish
Discipline(s) HAL :
Sciences de l'Homme et Société/Archéologie et Préhistoire
Résumé en anglais : [en]
Abstract Flatfish (Pleuronectiformes) vertebrae are difficult to identify to species due to the lack of diagnostic features. This has resulted in a lack of understanding of the species abundances across archaeological ...
Lire la suite >Abstract Flatfish (Pleuronectiformes) vertebrae are difficult to identify to species due to the lack of diagnostic features. This has resulted in a lack of understanding of the species abundances across archaeological sites, hindering interpretations of historical fisheries in the North Sea area. We use a new approach, utilising a combined 2D landmark‐based geometric morphometric analysis as an objective and non‐destructive method for species identification of flatfish vertebrae from the North Sea area. Modern specimens were used as a reference to describe the morphological variation between taxa using principal component analysis (PCA) and to trial an automated classification using linear discriminant analysis. Although there is limited distinction between taxa using PCAs, the classification shows high accuracies, indicating that flatfish species identifications using geometric morphometrics are possible. Bone samples ( n = 105) from two archaeological sites in the United Kingdom and France were analysed using this approach and their identifications were verified using collagen peptide mass fingerprinting. The success rate of species identification was usually less than 50%, indicating that this technique has limited applicability due to preservation/fragmentation of archaeological fish bone. Nonetheless, this could prove a valuable tool for modern and non‐fragmented samples. Furthermore, the technique applied in this study can be easily adapted to work on other landmark datasets.Lire moins >
Lire la suite >Abstract Flatfish (Pleuronectiformes) vertebrae are difficult to identify to species due to the lack of diagnostic features. This has resulted in a lack of understanding of the species abundances across archaeological sites, hindering interpretations of historical fisheries in the North Sea area. We use a new approach, utilising a combined 2D landmark‐based geometric morphometric analysis as an objective and non‐destructive method for species identification of flatfish vertebrae from the North Sea area. Modern specimens were used as a reference to describe the morphological variation between taxa using principal component analysis (PCA) and to trial an automated classification using linear discriminant analysis. Although there is limited distinction between taxa using PCAs, the classification shows high accuracies, indicating that flatfish species identifications using geometric morphometrics are possible. Bone samples ( n = 105) from two archaeological sites in the United Kingdom and France were analysed using this approach and their identifications were verified using collagen peptide mass fingerprinting. The success rate of species identification was usually less than 50%, indicating that this technique has limited applicability due to preservation/fragmentation of archaeological fish bone. Nonetheless, this could prove a valuable tool for modern and non‐fragmented samples. Furthermore, the technique applied in this study can be easily adapted to work on other landmark datasets.Lire moins >
Langue :
Anglais
Vulgarisation :
Non
Source :