Skip to Main content Skip to Navigation
Theses

From predators trajectories to preys maps: spatial estimation of purse-seiners activity and tuna in the Indian Ocean

Abstract : Nowadays spatial distribution of tropical tuna fished by the French purse-seiners in the Indian Ocean is represented with catch and effort data from logbooks. However, these data provided only one position per day and the number of fishing sets during each day: the prospected areas without any set, synonym of absence of tuna aggregations, were not known. These areas are essential to map the tuna abundances. The French purse-seiners fleet is equipped with Vessel Monitoring Systems (VMS). Trajectories obtained by the VMS were used to deduce vessels behaviour, and thus, presence (fishing) and absence (non fishing) of tuna aggregations. The objective is thus to estimate the vessel activity from its trajectory. The application of a state-space model allowed to distinguish "cruising", "tracking", "stopping", and "fishing" activities. For the majority of surveys (except those with an observer), the real activity of the vessel was not known (latent variable), which justified the modelling with a Hidden Markov model. The model was run in a Bayesian framework including prior knowledge on some parameters (speed, turning angle and transition matrix). This model was calibrated and validated thanks to observers data on 10% of the fleet. The error rate was of 10% for VMS positions data recorded every hour. After the estimation of the states, the purse-seiners activities (cruising/tracking/stop/fishing) allowed to (1) improve spatial and temporal indices of the effort of French purse-seiners, and (2) study the spatial structure of tuna presence areas (indicators variograms), and to deduce maps of tuna presence/potential presence/absence by indicators cokriging, that can be interpreted as an abundance proxy.
Complete list of metadatas

Cited literature [179 references]  Display  Hide  Download

https://pastel.archives-ouvertes.fr/pastel-00552183
Contributor : Bibliothèque Mines Paristech <>
Submitted on : Wednesday, January 5, 2011 - 4:12:11 PM
Last modification on : Wednesday, October 14, 2020 - 3:52:24 AM
Long-term archiving on: : Wednesday, April 6, 2011 - 3:13:10 AM

Identifiers

  • HAL Id : pastel-00552183, version 1

Citation

Emily Walker. From predators trajectories to preys maps: spatial estimation of purse-seiners activity and tuna in the Indian Ocean. Environmental Sciences. École Nationale Supérieure des Mines de Paris, 2010. English. ⟨NNT : 2010ENMP0013⟩. ⟨pastel-00552183⟩

Share

Metrics

Record views

819

Files downloads

1425