Published August 21, 2023 | Version v1
Dataset Open

A geometric morphometric approach to identify uncomplete snake vertebrae from raptor bird feeding remains

  • 1. Department of Systematic Zoology and Ecology, Eötvös Loránd University
  • 2. Department of Ecology, University of Debrecen

Description

The Hungarian meadow viper (Vipera ursinii rakosiensis) is an endangered subspecies of Vipera ursinii, which faces high predation pressure, partially due to avian species. To create a systematic method for estimating the measure of predation pressure, we developed a geometric morphometric approach to identify both undamaged and damaged vertebrae of snake species found in Hungarian meadow viper habitats from raptor feeding remains. We used linear discriminant analysis with a reference material of vertebrae from identified snake species as training data. We also tested its efficiency by predicting the identification results of different simulation levels based on vertebra completeness. We practiced this method on vertebrae of unknown species of snakes obtained from nests and pellets of short-toed snake eagles (Circaetus gallicus, n=9), common buzzards (Buteo buteo, n=14) and Montagu’s harriers (Circus pygargus, n=3). The identification approach showed high accuracy, even in the case of missing landmarks to some extent. We identified vertebrae remnants of Natrix natrix (n=172, 83.9%), Coronella austriaca (n=10, 4.9%) and V. u. rakosiensis (n=23, 11.2%). Both, the reptile specialist C. gallicus and the generalist B. buteo proved to be preying on V. u. rakosiensis, while samples of C. pygargus did not contain any snake remains despite of previous observations of V. ursinii predations. Our approach is applicable for other studies and taxa as well, therefore can be a practical tool for classification of incomplete vertebrae, which is otherwise hardly identifiable. Furthermore, it could be applied to help estimate predation pressure on endangered snake species.

Files

known_vertebrae_landmarks.csv

Files (787.3 kB)

Name Size Download all
md5:a6668ba3cf21473b19d6fbb55f73c450
659.1 kB Preview Download
md5:a28467e3194ddb09c585271c12b55abc
121.3 kB Preview Download
md5:27554419bdc90d841684f3a8cb17b40d
6.9 kB Download