Combining genetic markers with stable isotopes in otoliths reveals complexity in the stock structure of Atlantic bluefin tuna (Thunnus thynnus)

Deirdre Brophy,Naiara Rodríguez-Ezpeleta,Haritz Arrizabalaga,Igaratza Fraile

doi:10.1038/s41598-020-71355-6

Deirdre Brophy, Naiara Rodríguez-Ezpeleta + Show 2 more

Open Access

https://doi.org/10.1038/s41598-020-71355-6

Copy DOI

Abstract

Atlantic bluefin tuna (Thunnus thynnus) from the two main spawning populations in the Mediterranean and Gulf of Mexico occur together in the western, central and eastern Atlantic. Stock composition of catches from mixing areas is uncertain, presenting a major challenge to the sustainable management of the fisheries. This study combines genetic and chemical markers to develop an integrated method of population assignment. Stable isotope signatures (δ13C and δ18O) in the otolith core of adults from the two main spawning populations (adult baselines) showed less overlap than those of yearlings (12–18 months old) from western and eastern nursery areas suggesting that some exchange occurs towards the end of the yearling phase. The integrated model combined δ18O with four genetic markers (SNPs) to distinguish the adult baselines with greater accuracy than chemical or genetic markers alone. When used to assign individuals from the mixing areas to their population of origin, the integrated model resolved some (but not all) discrepancies between the chemistry and genetic methods. Some individuals in the mixing area had otolith δ18O values and genetic profiles which when taken together, were not representative of either population. These fish may originate from another Atlantic spawning area or may represent population contingents that move away from the main spawning areas during the first year of life. This complexity in stock structure is not captured by the current two-stock model.

Highlights

Increasing global food demands, climate change and habitat loss place unprecedented pressure on the world’s fish populations[1,2,3]
When the assignment threshold was set to 80%, individuals were assigned to a population if over 80% of the trees predicted that it belonged to that population and were unassigned if the votes for both populations were < 80%
Rates of classification accuracy were lower for the yearling baseline; 81.7% were assigned to the correct population with a probability threshold of 50%

Summary

Introduction

Increasing global food demands, climate change and habitat loss place unprecedented pressure on the world’s fish populations[1,2,3]. A fundamental step in the effective management of a fishery is to identify the appropriate management unit (stock)[6] This is wrought with difficulties; fish populations rarely maintain geographically discrete distributions throughout their life cycle and many fisheries exploit mixed aggregations[7]. A flexible approach to stock identification is needed, which recognises the broad spectrum of stock structure scenarios that e xist[7,15] Both genotypic and phenotypic traits are being used in combination to characterise fish populations[16,17,18,19]. The limitations of the current management approach are recognised and alternative assessment models are being tested within a management strategy evaluation framework that incorporates different scenarios of population mixing[41,42]. The potential contribution of ABFT from other spawning areas needs to be addressed through development and refinement of stock identification approaches

Objectives

Methods

Results

Conclusion