Fluctuations à long terme du thon rouge: validité, origines et conséquences

Describe and gain understanding on fluctuations in animal population over time and space is a central goal of population ecology. This is also crucial to ensure the sustainable management of exploited resources, particularly those with complex population dynamics such as the bluefin tuna. The goal of this work is to collect bluefin tuna catches from the ancestral Mediterranean and Atlantic trap fisheries to (i) characterise long-term fluctuations in bluefin tuna population migrating in the Mediterranean, (ii) analyse their causes et (iii) assess their implications for fisheries management. The bluefin tuna main biological and ecological characteristics are first detailed. Then, the first part presents the trap fishery, and the process of collect and validation of the historical catches. This leads to develop a data base of 54 time-series more than 20 years long, the longest ones spanning about four centuries. Mean historical catches were around 15 000 tons/year ([7 000;25 000]). Time-series are analysed in a second part. Temporal variability of trap catches may be decomposed in three components: 100-year-long periodic fluctuations, 20-year cycles and inter-annual variations. These medium- to long-term fluctuations, representing more than 50% of the variability in the time-series, were synchronous all around the western Mediterranean and adjacent North Atlantic. In contrast, short-term variability was synchronous at a local scale only. It is argued that long-term fluctuations in trap catches could be considered as a proxy for those in abundance, and a synthetic time-series is computed to depict them. Biological and ecological processes that could cause such long-term fluctuations are discussed in a third part. We tested whether long-term fluctuations in Atlantic bluefin tuna might be related to large-scale environmental changes, using long time-series of the North Atlantic Oscillation (NAO), the Length of the Day index (LOD, a proxy of the atmospheric circulation index ACI) and the temperature. Spectral analyses of trap catches, LOD and temperature displayed similar spectra with peak at low frequency, whereas those of the NAO exhibited a broad band spectrum. Regression analyses and tests of correlation did not reveal any clear relationship between trap catches on the one hand and NAO and LOD on the other hand. In contrast, long-term fluctuations in trap catches appear to be negatively and significantly related to long-term trends in temperature. Underlying processes that could explain such a relationship are discussed, with special focus on changes in migration patterns. Finally, implications of such fluctuations in term of fisheries management are discussed in a fourth part. A simulation framework, in which a simulated population dynamics model is coupled with a VPA, is built to test the pertinence of the diagnostic of stock assessment according to different scenarios on fluctuations origins and population status. In the hypothesis of fluctuations linked to variations in the recruitment, VPA appears able to reconstruct the fluctuations but the predictions may be biased if natural fluctuations are not taking into account. In the hypothesis of fluctuations linked to changes in migration pattern, VPA is not able to reconstruct the population dynamics. In a precautionary approach, quotas which follow may be unsuitable in the best case, at worst dangerous if they are defined in comparison with a reference point. The concept of cible reference points, fixe and unique over time, appears not relevant for bluefin tuna; estimate trajectory reference pertinent may prove to be more pertinent.