European hake (Merluccius merluccius) stock structure in the Mediterranean as assessed by otolith shape and microchemistry

• Published in: Fisheries Research
• Main Authors: Morales-nin, Beatriz, Pérez-mayol, Sílvia, Mackenzie, Kirsteen, Catalán, Ignacio A., Palmer, Miquel, Kersaudy, Thibaut, Mahé, Kelig
• Format: Article in Journal/Newspaper
• Language: English
• Published: Elsevier BV 2022
• Subjects: Merluccius merluccius; Population; Sagittal otolith; Shape; Geotags; Mediterranean Sea; Hake; North East Atlantic
• Online Access: https://archimer.ifremer.fr/doc/00783/89472/94966.pdf; https://archimer.ifremer.fr/doc/00783/89472/94967.docx; https://doi.org/10.1016/j.fishres.2022.106419; https://archimer.ifremer.fr/doc/00783/89472/

The European hake Merluccius merluccius is the third most valuable species for the North-East Atlantic and the Mediterranean fisheries. European hake has been rated as overexploited in the Mediterranean, thus careful management is advisable. Mediterranean hake is well-differentiated from Atlantic hake, but sub-population structure within the Mediterranean, and how this structure could be translated into stocks (operative management units), is still an elusive topic. Otolith shape and chemistry (concentration of trace elements) have been systematically used to distinguish fish stocks. Our aim was therefore to assess the discrimination capability (assigning fish to the correct geographical unit) of otolith shape and microchemistry at two geographical scales within the Mediterranean: (1) the official geographical subareas (GSAs), and (2) three larger units previously suggested by genetic markers (i.e., Western Mediterranean, Adriatic with Central Mediterranean, and Eastern Mediterranean). Two complementary analyses were completed because shape is more easily analyzed than chemistry. First, a large sample of juvenile hake (n = 1656) from 40 Mediterranean GSAs subunits was used for shape analysis. Second, a subsample of those fish (n = 154) from 10 GSAs was analyzed for both otolith shape and microchemistry. Irrespective of the type of data (shape and/or chemistry) and geographical scale (GSAs versus the 3-units), between-unit differences were always statistically significant. However, according to the large within-unit variability, discrimination capability was always poor but better at the GSA scale, and even better when both shape and microchemistry were combined. Moreover, unsupervised clustering methods (the number and limits of the units are data-driven and not a priori defined as above) failed to find an optimal structure. Overall, these results are fully compatible with the hypothesis of a continuous gradient, within which discrete spatial units cannot be safely recognized.