MEPS 452:253-267 (2012)  -  DOI: https://doi.org/10.3354/meps09618

Comparative analysis of methods for inferring successful foraging areas from Argos and GPS tracking data

Anne-Cécile Dragon1,2,3,*, Avner Bar-Hen2, Pascal Monestiez4, Christophe Guinet1

1CEBC-CNRS, 79360 Villiers en Bois, France
2MAP5-CNRS, 45 rue des Saints-Pères, 75270 Paris Cedex 06, France
3LOCEAN-UPMC, 4 place Jussieu, boite 100, 75252 Paris Cedex 05, France
4INRA, UR546 Biostatistique et Processus Spatiaux (BioSP), 84914 Avignon, France

ABSTRACT: Identifying animals’ successful foraging areas is a major challenge, but  such comprehensive knowledge is needed for the management and conservation of wild populations. In recent decades, numerous specific analytic methods have been developed to handle tracking data and to identify preferred foraging areas. In this study, we assessed the efficiency of different track-based methods on Argos and GPS predators’ tracks. We investigated (1) the consistency in the detection of foraging areas between track-based methods applied to 2 tracking data resolutions and (2) the similarity of foraging behaviour identification between track-based methods and an independent index of foraging success. We focused on methods that are commonly used in the literature: empirical descriptors of foraging effort, Hidden Markov Models (HMMs) and first passage time analysis. We applied these methods to satellite tracking data collected on 6 long-ranging elephant seals equipped with both Argos and GPS tags. Seals were also equipped with time depth recorder loggers from which we estimated an independent index, based on the drift rate and the changes in the seals’ body condition, as a proxy for foraging success along the tracks. Favourable foraging zones identified by track-based methods were compared to locations where the body condition of the seals significantly increased. With or without an environmental covariate, HMMs were the most reliable for identifying successful foraging areas on both high (GPS) and low (Argos) resolution data. Areas identified by HMMs as intensively used were congruent with the locations where seals significantly increased their body condition given a 4 d metabolisation lag.


KEY WORDS: Area-restricted-search · Drift dive · First bottom time · Mirounga leonine · Movement analysis · State-space modelling


Full text in pdf format 
Supplementary material 
Cite this article as: Dragon AC, Bar-Hen A, Monestiez P, Guinet C (2012) Comparative analysis of methods for inferring successful foraging areas from Argos and GPS tracking data. Mar Ecol Prog Ser 452:253-267. https://doi.org/10.3354/meps09618

Export citation
Mail this link - Contents Mailing Lists - RSS
- -