Spatial ecology of humpback whales in breeding areas : habitats, distribution and movements in the South Pacific

• Main Author: Derville, Solène
• Other Authors: Ecologie marine tropicale des océans Pacifique et Indien (ENTROPIE Nouvelle-Calédonie ), Institut de Recherche pour le Développement (IRD Nouvelle-Calédonie )-Ifremer - Nouvelle-Calédonie, Institut Français de Recherche pour l'Exploitation de la Mer (IFREMER)-Institut Français de Recherche pour l'Exploitation de la Mer (IFREMER)-Université de la Nouvelle-Calédonie (UNC), Sorbonne Université, Claire Garrigue
• Format: Doctoral or Postdoctoral Thesis
• Language: English
• Published: HAL CCSD 2018
• Subjects: Habitat mapping; Spatial management; Movements; Humpback whale; Species Distribution Models; Baleine à bosse; Modèles de distribution d’espèces; Déplacements; Cartographie d’habitats; Gestion spatiale; [SDE.BE]Environmental Sciences/Biodiversity and Ecology; [SDU.STU.OC]Sciences of the Universe [physics]/Earth Sciences/Oceanography; Baleine; High Island; Pacific; Humpback Whale; Low Island; Megaptera novaeangliae
• Online Access: https://tel.archives-ouvertes.fr/tel-02865352; https://tel.archives-ouvertes.fr/tel-02865352v2/document; https://tel.archives-ouvertes.fr/tel-02865352v2/file/these_DERVILLE_Sonia_2018.pdf

Understanding the social and environmental drivers of the distribution and movements of marine megafauna is essential to their conservation. Cetaceans are elusive and mobile species, whose management requires an improved understanding of habitat use patterns. This thesis is aimed at investigating the spatial ecology of an endangered population of humpback whales (Megaptera novaeangliae) in the South Pacific Ocean. Using a multidisciplinary dataset collected between 1995 and 2018 in New Caledonia and Oceania, this thesis addressed three objectives, 1) investigate best practices to cetacean species distribution modeling, 2) acquire fundamental knowledge on the distribution, habitats and movements of humpback whales in Oceania breeding grounds, and 3) predict priority conservation areas and potential threats to humpback whales. Modeling the distribution of a migratory large whale from non-systematic visual survey and citizen science data provided valuable space-use predictions when uneven survey effort and statistical overfitting were specifically addressed. Generalized Additive Models were favored for their complexity trade-off, ecological interpretability and transferability. Models of habitat use revealed a preference for a diversity of shallow habitats (low island and atoll lagoons, barrier reef and high island slopes, banks and seamounts) spread over a relatively large thermal range over Oceania. Shallow seamounts and banks were identified as major breeding and nursing habitats and play a key role in the connectivity within and between populations. This unique and unexpected use of pelagic waters has important consequences for the spatial management of humpback whales. The predictions of present and future suitable humpback whale breeding habitats at multiples scales provide science-based evidence for priority conservation areas, and enable mitigation of threats from anthropogenic activities and climate change in the South Pacific. Comprendre les facteurs sociaux et environnementaux de la distribution et des ...