|Author(s)||Gloaguen Pierre1, Mahevas Stephanie1, Rivot Etienne2, Woillez Mathieu2, 3, Guitton Jerome2, Vermard Youen4, Etienne Marie-Pierre5|
|Affiliation(s)||1 : IFREMER, F-44311 Nantes 03, France.
2 : AGROCAMPUS OUEST, UMR Ecol & St Ecosyst 985, CS84215, F-35042 Rennes, France.
3 : IFREMER, F-29280 Plouzane, France.
4 : IFREMER, Channel & North Sea Fisheries Dept, F-62321 Boulogne Sur Mer, France.
5 : AGROPARISTECH, UMR MIA 518, F-75231 Paris, France.
|Source||Environmetrics (1180-4009) (Wiley-blackwell), 2015-02 , Vol. 26 , N. 1 , P. 17-28|
|WOS© Times Cited||22|
|Keyword(s)||hidden Markov model, vessels dynamics, RECOPESCA, autoregressive process, Baum-Welch algorithm|
|Abstract||The understanding of the dynamics of fishing vessels is of great interest to characterize the spatial distribution of the fishing effort and to define sustainable fishing strategies. It is also a prerequisite for anticipating changes in fishermen's activity in reaction to management rules, economic context, or evolution of exploited resources. Analyzing the trajectories of individual vessels offers promising perspectives to describe the activity during fishing trips. A hidden Markov model with two behavioral states (steaming and fishing) is developed to infer the sequence of non-observed fishing vessel behavior along the vessel trajectory based on Global Positioning System (GPS) records. Conditionally to the behavior, vessel velocity is modeled with an autoregressive process. The model parameters and the sequence of hidden behavioral states are estimated using an expectation–maximization algorithm, coupled with the Viterbi algorithm that captures the most credible joint sequence of hidden states. A simulation approach was performed to assess the influence of contrast between the model parameters and of the path length on the estimation performances. The model was then fitted to four original GPS tracks recorded with a time step of 15 min derived from volunteer fishing vessels operating in the Channel within the IFREMER RECOPESCA project. Results showed that the fishing activity performed influenced the estimates of the velocity process parameters. Results also suggested future inclusion of variables such as tide currents within the ecosystem approach of fisheries.|
Gloaguen Pierre, Mahevas Stephanie, Rivot Etienne, Woillez Mathieu, Guitton Jerome, Vermard Youen, Etienne Marie-Pierre (2015). An autoregressive model to describe fishing vessel movement and activity. Environmetrics, 26(1), 17-28. Publisher's official version : https://doi.org/10.1002/env.2319 , Open Access version : https://archimer.ifremer.fr/doc/00179/29049/