The identification of sea turtle behaviours is a prerequisite to predicting the activities and time-budget of these animals in their natural habitat over the long term. However, this is hampered by a lack of reliable methods that enable the detection and monitoring of certain key behaviours such as feeding. This study proposes a combined approach that automatically identifies the different behaviours of free-ranging sea turtles through the use of animal-borne multi-sensor recorders (accelerometer, gyroscope and time-depth recorder), validated by animal-borne video-recorder data. We show here that the combination of supervised learning algorithms and multi-signal analysis tools can provide accurate inferences of the behaviours expressed, including feeding and scratching behaviours that are of crucial ecological interest for sea turtles. Our procedure uses multi-sensor miniaturized loggers that can be deployed on free-ranging animals with minimal disturbance. It provides an easily adaptable and replicable approach for the long-term automatic identification of the different activities and determination of time-budgets in sea turtles. This approach should also be applicable to a broad range of other species and could significantly contribute to the conservation of endangered species by providing detailed knowledge of key animal activities such as feeding, travelling and resting.

Satellite remote-sensing and wildlife tracking allow researchers to record rapidly increasing volumes of information on the spatial ecology of marine megafauna in the context of global change. This field of investigation is thereby entering the realm of big data science: Information technology allows the design of completely new frameworks for acquiring, storing, sharing, analysing, visualizing, and publicizing data. This review aims at framing the importance of big data for the conservation of marine megafauna, through intimate knowledge of the spatial ecology of these threatened, charismatic animals. We first define marine megafauna and big data science, before detailing the technological breakthroughs leading to pioneering “big data” studies. We then describe the workflow from acquiring megafauna tracking data to the identification and the prediction of their critical habitats under global changes, leading to marine spatial planning and political negotiations. Finally, we outline future objectives for big data studies, which should not take the form of a blind technological race forward, but of a coordinated, worldwide approach to megafauna spatial ecology, based on regular gap analyses, with care for ethical and environmental implications. Employing big data science for the efficient conservation of marine megafauna will also require inventing new pathways from research to action.

In response to seasonality and spatial segregation of resources, sea turtles undertake long journeys between their nesting sites and foraging grounds. While satellite tracking has made it possible to outline their migration routes, we still have little knowledge of how they select their foraging grounds and adapt their migration to dynamic environmental conditions. Here, we analyzed the trajectories and diving behavior of 19 adult green turtles (Chelonia mydas) during their post-nesting migration from French Guiana and Suriname to their foraging grounds off the coast of Brazil. First Passage Time analysis was used to identify foraging areas located off Ceará state of Brazil, where the associated habitat corresponds to favorable conditions for seagrass growth, i.e. clear and shallow waters. The dispersal and diving patterns of the turtles revealed several behavioral adaptations to the strong hydrodynamic processes induced by both the North Brazil current and the Amazon River plume. All green turtles migrated south-eastward after the nesting season, confirming that they coped with the strong counter North Brazil current by using a tight corridor close to the shore. The time spent within the Amazon plume also altered the location of their feeding habitats as the longer individuals stayed within the plume, the sooner they initiated foraging. The green turtles performed deeper and shorter dives while crossing the mouth of the Amazon, a strategy which would help turtles avoid the most turbulent upper surface layers of the plume. These adjustments reveal the remarkable plasticity of this green turtle population when reducing energy costs induced by migration.

Here we propose a new machine learning algorithm for classification of human activities by means of accelerometer and gyroscope signals. Based on a novel hierarchical system of logistic regression classifiers and a relatively small set of features extracted from the filtered signals, the proposed algorithm outperformed previous work on the DaLiAc (Daily Life Activity) and mHealth datasets. The algorithm also represents a significant improvement in terms of computational costs and requires no feature selection and hyper-parameter tuning. The algorithm still showed a robust performance with only two (ankle and wrist) out of the four devices (chest, wrist, hip and ankle) placed on the body (96.8% vs. 97.3% mean accuracy for the DaLiAc dataset). The present work shows that low-complexity models can compete with heavy, inefficient models in classification of advanced activities when designed with a careful upstream inspection of the data.