Résumé: Terrestrial plants seeds, spores and pollen are often dispersed by wind. Likewise, most eggs and larvae of marine organisms are dispersed by oceanic currents. It was historically believed that the spatial scale at which dispersal occurs was orders of magnitude smaller for plants than for fish. However, recent empirical estimates of seed and larval dispersal suggest that these dispersal scales are more alike than previously thought. The modeling approaches used to simulate aerial and aquatic dispersal are also converging. Similar biophysical models are developed, in which outputs of Eulerian models simulating the main physical forcing mechanism (wind or currents) are used as inputs to Lagrangian models that include biological components (such as seed terminal velocity or larval vertical migration). These biophysical models are then used to simulate trajectories of the biological entities (seeds, larvae) in three dimensions. We reflect on these converging trends by first putting them into an historical perspective, and then by comparing the physical and biological processes represented in marine larva vs. terrestrial seed dispersal models, the data used for the models output corroboration, and the tools available to perform simulations. We conclude that this convergence offers the opportunity to bridge the gap between two scientific communities which are currently largely disconnected. More broadly, we also see our comparison across systems as a useful way to strengthen the links between aquatic and terrestrial ecology by sharing knowledge, methods, tools, and concepts.

Résumé: Understanding larval connectivity patterns is critical for marine spatial planning, particularly for designing marine protected areas and managing fisheries. Patterns of larval dispersal and connectivity can be inferred from numerical transport models at large spatial and temporal scales. We assess model-based connectivity patterns between seamounts of the Southwestern Indian Ocean (SWIO) and the coastal ecosystems of Mauritius, La Reunion, Madagascar, Mozambique and South Africa, with emphasis on three shallow seamounts (La Pemuse [LP], MAD-Ridge [MR] and Walters Shoal [WS]). Using drifter trajectory and a Lagrangian model of ichthyo-plankton dispersal, we show that larvae can undertake very long dispersion, with larval distances increasing with pelagic larval duration (PLD). There are three groups of greater connectivity: the region between the eastern coast of Madagascar, Mauritius and La Reunion islands; the seamounts of the South West Indian Ridge; and the pair formed by WS and a nearby un-named seamount. Connectivity between these three groups is evident only for the longest PLD examined (360 d). Connectivity from seamounts to coastal ecosystems is weak, with a maximum of 2% of larvae originating from seamounts reaching coastal ecosystems. Local retention at the three focal seamounts (LP, MR and WS) peaks at about 11% for the shortest PLD considered (15 d) at the most retentive seamount (WS) and decreases sharply with increasing PLD. Information on PLD and age of larvae collected at MR and LP are used to assess their putative origin. These larvae are likely self-recruits but it is also plausible that they immigrate from nearby coastal sites, i.e. the southern coast of Madagascar for MR and the islands of La Reunion and Mauritius for LP.

Résumé: Yellowfin tuna, Thunnus albacares, is one of the most important seafood commodities in the world. Despite its great biological and economic importance, conflicting evidence arises from classical genetic and tagging studies concerning the yellowfin tuna population structure at local and global oceanic scales. Access to more powerful and cost effective genetic tools would represent the first step towards resolving the population structure of yellowfin tuna across its distribution range. Using a panel of 939 neutral Single Nucleotide Polymorphisms (SNPs), and the most comprehensive data set of yellowfin samples available so far, we found genetic differentiation among the Atlantic, Indian and Pacific oceans. The genetic stock structure analysis carried out with 33 outlier SNPs, putatively under selection, identified discrete populations within the Pacific Ocean and, for the first time, also within the Atlantic Ocean. Stock assessment approaches that consider genetic differences at neutral and adaptive genomic loci should be routinely implemented to check the status of the yellowfin tuna, prevent illegal trade, and develop more sustainable management measures.