En poursuivant votre navigation sur ce site, vous acceptez le dépôt de cookies dans votre navigateur. (En savoir plus)

Thesis in ecology and behaviour using algorithmic tools (M/W)

This offer is available in the following languages:
- Français-- Anglais

Date Limite Candidature : lundi 5 juin 2023

Assurez-vous que votre profil candidat soit correctement renseigné avant de postuler. Les informations de votre profil complètent celles associées à chaque candidature. Afin d’augmenter votre visibilité sur notre Portail Emploi et ainsi permettre aux recruteurs de consulter votre profil candidat, vous avez la possibilité de déposer votre CV dans notre CVThèque en un clic !

Informations générales

Intitulé de l'offre : Thesis in ecology and behaviour using algorithmic tools (M/W) (H/F)
Référence : UMR7372-JULCOL-001
Nombre de Postes : 1
Lieu de travail : VILLIERS EN BOIS
Date de publication : lundi 15 mai 2023
Type de contrat : CDD Doctorant/Contrat doctoral
Durée du contrat : 36 mois
Date de début de la thèse : 2 octobre 2023
Quotité de travail : Temps complet
Rémunération : 2 135,00 € gross monthly
Section(s) CN : Biodiversity, evolution and biological adaptations: from macromolecules to communities

Description du sujet de thèse

TITLE: Historical variations in the rates of interactions with fisheries in seabird populations monitored by telemetry in the French Southern Territories.
Context:
The massive development of fisheries since the second half of the 20th century (Watson & Tidd, 2018) has many ecological consequences (Grémillet et al., 2018; Pauly et al., 2005). Among others, they cause incidental catches of highly threatened marine predators (Dias et al., 2019; Lewison et al., 2014), often when they interact directly with boats for feeding (Oro et al., 2013). The study of the consequences of fisheries is often severely limited by access to detailed, reliable and representative data on the location and total activities of all vessels (Komoroske & Lewison, 2015). Interaction risks are thus often estimated indirectly from maps of boat distributions (Komoroske & Lewison, 2015), neglecting several theoretical difficulties (Holling, 1959; Suraci et al., 2022) for example related to possible changes in predator behaviour (Barbraud et al., 2013; Weimerskirch et al., 2023) through plastic or micro-evolutionary processes.
Recently several more or less expensive and/or easy to implement methods have become available to directly document seabird-vessel interactions at sea (Collet et al., 2015; Kroodsma et al., 2018; Votier et al., 2010; Weimerskirch et al., 2020). This thesis will take advantage of these developments, mainly based on large GPS location databases, and a large dataset of long-term monitoring of seabird movements at sea in the French Southern Territories, to assess potential variations in the rate of interaction of birds with different fishing fleets and to better understand the dynamics of these interactions over time and variations in fishing activities of vessels.
Objective of the thesis :
In this context, the thesis will specifically seek to
1. explore the possibility of automating the detection of these interactions at low cost by supervised machine learning (Carneiro et al., 2022; Corbeau et al., 2019; Tremblay et al., 2014). The relatively expensive use of radar emission sensors coupled with GPS on board the birds; as well as their linkage with GPS location data of boats reported in the areas used by the birds, has made it possible to obtain for more than a hundred journeys of Great albatrosses (Diomedea exulans) an almost exhaustive annotation of all phases of interaction with boats during the journey (Corbeau et al., 2019; Weimerskirch et al., 2020; Weimerskirch H. et al., 2017). Full annotation may also be available for a second species (black-browed albatross, Thallassarche melanophris)(Collet et al., 2017a). These recent full annotation data could be used as a basis for training and validating a supervised machine learning model to try to predict interactions from trajectory (GPS) data alone.
2. to trace and compare, via the machine-learning approach and/or via the use of historical data on the fine GPS location of boats (different data sources envisaged, currently being acquired (Collet et al., 2017a; Weimerskirch et al, 2020), historical variations in the interaction rate (Collet et al., 2017a; Corbeau et al., 2021), particularly in large and eyebrowed albatrosses monitored over the long term by the CEBC (data from 2000-2023, within the framework of the IPEV 109 programme, (Weimerskirch Henri, 2018)), and to identify and compare trends. Indeed, initial results from a one-year study suggested different interaction behaviours between these two species (Collet et al., 2017a). These analyses could also benefit from age and sex data on individuals tracked by GPS (Collet et al., 2017b).
3. to link variations in interactions with historical variations in effort of various fishing fleets in the south-west Indian Ocean (Corbeau et al., 2021), which have developed more or less recently depending on the sector (Palomares & Pauly, 2011; Watson & Tidd, 2018; Weimerskirch Henri, 2018). In particular, the extent to which fishing effort predicts encounter rates and/or interaction rates will be tested (Corbeau et al., 2021; Suraci et al., 2022). These analyses could be supported by data on the wider distribution of fishing effort (Corbeau et al., 2021; Kroodsma et al., 2018).

Contexte de travail

The student will be part of the Marine Predators team of the Centre for Biological Studies of Chizé (UMR 7372, CNRS, La Rochelle University, 79360 Villiers en Bois) where he/she will be based. He/she will be attached to the Euclid Doctoral School (La Rochelle University). The student will carry out his/her thesis under the supervision of Julien Collet (La Rochelle University - CEBC). Financial support will be provided by the CNRS (MITI Project) for the salary for 3 years, as well as by a part dedicated to this thesis on an ANR- CPJ grant allocated to J. Collet (2022-27), to cover conference, defense and publication fees as well as computer equipment. The fieldwork is planned in the French Southern Territories within the framework of recurrent programmes financed and organised by the French Polar Institute and the CEBC (notably programme 109 directed by Christophe Barbraud). The subject is very well integrated into the broader research themes and/or methods of the team, and the PhD student will benefit from a stimulating environment with PhD students, researchers, teachers, engineers and post-doctoral fellows and their national or international collaborators.
Student's work programme (tasks):
(1) Development of a supervised machine learning algorithm for the analysis of annotated data: the PhD student will determine the level of fineness of the annotations to be used for the algorithm, the input variables to be considered, the distribution between training data and validation data; will choose and develop an adapted algorithm (solution considered a priori: random forest) and will interpret the results. To do this, he will be able to rely on the expertise of the supervisor, J. Collet, in the behaviour of birds and the statistical description of path data; and on the technical expertise of Marianna Chimienti (postdoc and collaborator at the CEBC), who regularly implements supervised machine learning programmes for the analysis of ecological data.
(2) Linking and analysing large databases of simultaneous spatial location data on seabirds and ships. The PhD student will ensure the implementation of algorithms for matching GPS bird location data with huge amounts of GPS location data of ships in the regions used by the birds (e.g. AIS data). Effective codes already exist but may need to be updated, improved or adapted as the volume or format of the data varies. These linked data will then be used by the PhD student to analyse the interaction rates, which will be based on solid methodological and statistical foundations developed and tested by the supervisor (Collet et al., 2017a, 2017b), and on long-standing expertise in the host team to guide the interpretation of the results (J Collet, H Weimerskirch, C Barbraud, K Delord) It is also expected that these bird-boat data matches will also be used for a second subsequent recruitment (2024-27) on more detailed questions of personality and behavioural heritability, in which the PhD student will be involved.
(3) Participation in field data collection in the framework of long-term monitoring related to the thesis topic. The PhD student will be given priority to go for at least one summer campaign in the French Southern Territories, in the framework of the IPEV 109 programme, to participate in the fitting of GPS on birds as well as in the long-term protocols (ringing control, handling and biometric measurements on large birds, etc.). This fieldwork will involve prolonged distance (simple transport to the site will take several weeks) and work in potentially harsh climatic and social conditions. It will also involve handling birds that are both large (e.g. large albatrosses of 7-12kg, with strong beaks) and strictly protected. This fieldwork will be done in a team, with at least one person already well experienced in the protocol.

(4) Scientific paper writing and communication: the PhD student will write papers based on the results obtained and communicate them to the scientific community through scientific papers in international journals and participation in conferences. He/she will also be encouraged to participate in communication actions for the general public around these themes (Science Festival, etc.).

Contraintes et risques

Fieldwork in the French Southern Territories (TAAFs) involves prolonged remoteness (with simple transport to the site requiring several weeks) and work in potentially challenging climatic and social conditions. It will also involve handling birds that are both imposing (e.g. large albatrosses weighing 7-12kg, with strong beaks) and strictly protected. This fieldwork will be carried out in a team, with at least one person who is already well experienced, within the framework of protocols and logistical, material and administrative organisation that have been mastered by the Polar Institute, the TAAFs and the CEBC staff for decades.

Informations complémentaires

The candidate should have an engineering degree and/or a master's degree in ecology (Bac +5) with strong computer and/or quantitative skills. The position requires a solid mastery of programming, particularly in R, an excellent ability to quickly mobilise and understand sophisticated statistical techniques, and good oral and written communication skills in English (level B2 or higher) for the writing of scientific articles and participation in international conferences.

A driving licence is highly recommended, as well as a good command of French, to facilitate integration in the relatively isolated CEBC site. The ideal profile is a person
- motivated by fundamental questions in ecology with prospects for applications in conservation
- who already feels sufficiently autonomous in mastering the technical tools of analysis to be able to quickly and efficiently focus the team's reflections on the biological and ecological questions that can be explored with these methods
- and with the desire and organisational skills to go on relatively long and distant missions with peace of mind.
A good understanding of the systems in question (seabirds, fisheries) will be a plus, as well as naturalist skills and in particular previous experience in handling wild vertebrates and/or setting up loggers to participate in fieldwork.