Skip to Content

EOC WS - Tracking marine top foragers (EtatJ’erre)

Workshop (Other)
EUR-OCEANS organisation / funding

This WS was selected for funding after the EOC 2012 call for proposal.
Initially scheduled for 2013 (15-17 Oct.), it had to be post-poned to 15-17 Jan. 2014.

Topic and general objectives [excerpts from proposal]

Tracking animals with electronic devices has experienced a broad and recent development in ecology, leading to the concept of movement ecology (PNAS special issue, 2008). In particular in the marine environment, specific electronic loggers, (e.g. GPS and GLS tags, (Pop-Up) archival tags, Time Depth Recorders) have been developed in the recent years allowing for the 2D or 3D tracking of large foragers (fish, mammals, seabirds). An interesting connected research avenue has emerged with the generalization of Vessel Monitoring Systems that allows for tracking movements of fishing boats. Because fishermen track a resource which is spread on large spatial areas with objectives of optimizing the catches rates, their behavior has close similarity with those of top foragers (Bertrand et al., 2007).

The spatial behavior of both marine predators and fishing boats is partially driven by the spatio-temporal distribution of their preys (e.g. small pelagic fish, squids, shrimps, or target species in general for fishermen). Once integrated over several individuals and/or several types of predators (including fishing boats) and/or over several scales, modeling their tracks opens the possibility to infer about the spatio-temporal dynamics of their preys at the oceanic basin scale, including open-sea (in contrast to acoustic survey that are limited to coastal areas and specific periods).

Each tracking device has its own strength and weaknesses that must be clarified in particular with regards to the precisions and density of the observations in space and time. If we do not intend in this project to tackle the question of position filtering (e.g. Royer et al. 2005, Nielsen et al. 2007, Tremblay et al. 2009), we address the question of integrating the uncertainties in geolocations into the analyses of the trajectories.

In parallel to data acquisitions, several types of analytical methods have been developed for tracking data (state-space models using Hidden Markov chain for behavioral states, Artificial Neural Networks, Area Restricted Search-ARS detection, etc). These methods refer to different ways of interpreting trajectories and movements and to different goals. For instance, state-space models estimate behavioral states (one amongst a fixed set of possible states, typically feeding or fishing on prey, searching for prey, and moving to another area) of an individual in time (e.g. Patterson et al. 2008). They mostly rely on discrete time markovian switch between states and require regular time step acquisition of data. ARS detections clearly target a particular type of movement within trajectories; the rest of the trajectory being left over (e.g. Knell and Kolding 2011).
Amongst the challenges that come with the new set of observations on the spatio-temporal dynamics of top predators and fishing boats is the question of the relevancy of the information provided by non random samplers of the marine ecosystem.

Expected outputs

  • Synthesis paper on existing applications and adequation of methods used to available data
  • Collective contribution to the International Statistical Ecology Conference 2014 (ISEC 2014;
  • Contribution to French grouping of research units (Groupement de Recherche or GdR)  focusing on and aiming at structuring and advancing research on Statistical Ecology / Movement Ecology ("Écologie statistique/Écologie du déplacement").

Preliminary work programme

Prior to workshop

  • literature review (N.Bez & S.Mahévas) on current modeling objectives (boats, fish, birds), available methods to analyse georeferenced observation and model behaviour - expliciting underlying assumptions (behaviour, scales), validation methods, how uncertainty is dealt with
  • inventory among all participants and partners of available data and methods used
  • exchange of data and/or methods (to produce results for the meeting)
  • partners using methods to prepare a document on
    1. objectives – assumptions – constraints
    2. input data form
    3. computer implementation (language, software)
    4. outputs
    5. publications

During the workshop

  • presentation of literature review, comparative analyses of results, proposal for structuring results, discussion
  • writing (results, discussion)
  • identification of gaps / future research avenues
  • outline of contribution to ISEC 2014
  • definition of prospective contribution to GdR on Statistical Ecology / Movement Ecology des contributions propsectives au GdR « Écologie statistique/Écologie du déplacement »


15-20 participants, upon invitation (please contact organizers if interested).


Activity leaders (see below).

Activity leader(s): 
Nicolas Bez (IRD; nicolas.bez [at] & Stéphanie Mahévas (Ifremer; stephanie.mahevas [at]
Paris, France
Financial support: 
8 k€
Start date: 
End date: 

activities | about seo