Possible drivers and implications of changes in inter-spawning area connectivity
The observed levels of genetic differentiation, hardly compatible with the high level of gene flow suggested by our results in constant equilibrium, suggest that connectivity patterns between ABFT spawning grounds could be subjected to temporal changes and that an increase in migration from the Mediterranean Sea towards the known western Atlantic spawning areas could have a genetic homogenizing effect. Such a homogenizing effect is expected to be correlated with migration rates and the effective population size (Ne) of the recipient population (Lowe and Allendorf 2010, Gagnaire et al. 2015), which in turn relates to the number of adult individuals among other factors (Waples 2022) and would consequently be affected by fluctuations in the population’s abundance. The abundance of ABFT stocks have undergone strong changes during the last ~60 years. After both the western and the eastern Atlantic stocks reached a critical status, including the collapse of several fisheries around the 1960-1970s (Fromentin 2009, Porch et al. 2019), the western Atlantic stock has not recovered as rapidly as the eastern Atlantic stock (of Mediterranean origin), whose estimated abundance has been one order of magnitude larger for several decades (ICCAT 2017), despite decades of conservation efforts. This slow recovery could result from a regime shift over the last decades, due to the combination of oceanographic changes in the equatorial Atlantic and overfishing (Fromentin et al. 2014b) possibly affecting both migration rates and effective population sizes. Fluctuations in the abundance and distribution of eastern Atlantic bluefin tuna during the last century were largely explained by the Atlantic Multidecadal Oscillation (AMO) (Faillettaz et al. 2019) and long-term trends in temperature (Ravier and Fromentin 2004). Moreover, coinciding with the last negative AMO period starting in the 1960s, ABFT had disappeared from several North-East Atlantic areas where it is reappearing during the current positive AMO phase starting in the mid-1990s (Horton et al. 2020, Nøttestad et al. 2020, Aarestrup et al. 2022). Furthermore, increasing catches of ABFT in Greenland waters show that the northern limit of ABFT distribution was expanded northwards during the last decade by mostly individuals of Mediterranean genetic origin (Jansen et al. 2021). Based on electronic tagging data, the proportion of individuals of eastern origin present in the western Atlantic has also increased during the last two decades (Aalto et al. 2021). Interestingly, AMO positive or warm phases as well as current global warming involve an increase in habitat suitability in most of these northern areas (Fromentin et al. 2014b, Faillettaz et al. 2019). In summary, population size, distribution, and migratory behavior of ABFT has been undergoing changes during the last decades, probably due to changes in environmental conditions, fishing pressure, conservation efforts or combined effects of these. These changes could explain migration intensification from the recently expanded eastern stock towards the more slowly recovering western Atlantic stock. Our analyses support that the recent short-term genetic effects of immigration on the variance in ancestries are much stronger in the Slope Sea than in the Gulf of Mexico. This could be due to behavioral preferences or favorable conditions which would make it easier for Mediterranean individuals to reach and reproduce in the Slope Sea, or due to a smaller effective population size compared to the Gulf of Mexico. The homogenizing effect of this unidirectional migration would ultimately lead to the genetic swamping of the GOM-like genetic component. It is thus possible that genetic differentiation is detected in the existing samples because gene flow is relatively recent relative to mean generation times estimated to average 9.6 years for the Western stock (Collette et al. 2011), and that this genetic divergence will be attenuated across generations in the future. While genetic connectivity does not necessarily equate with demographic dependence, genetic erosion would not necessarily imply demographic decline either, but could have more unpredictable demographic consequences.