Otolith shape as a valuable tool to evaluate the stock structure of swordfish Xiphias gladius in the Indian Ocean

Type Article
Date 2016
Language English
Author(s) Mahe KeligORCID1, Evano Hugues2, Mille Tiphaine1, Muths D.3, Bourjea JeromeORCID4
Affiliation(s) 1 : IFREMER, Lab Ressources Halieut, Boulogne Sur Mer, France.
2 : IFREMER, Delegat Ocean Indien, Le Port, Reunion.
3 : Univ Europeenne Bretagne, Rennes, France.
4 : IFREMER, UMR MARBEC 248, Sete, France.
Source African Journal Of Marine Science (1814-232X) (Natl Inquiry Services Centre Pty Ltd), 2016 , Vol. 38 , N. 4 , P. 457-464
DOI 10.2989/1814232X.2016.1224205
WOS© Times Cited 3
Keyword(s) connectivity, Fourier descriptors, morphometric markers, sagittal otoliths, stock discrimination
Abstract Swordfish Xiphias gladius is an oceanic-pelagic species. Its population structure in the Western Indian Ocean was studied from the shape of the sagittal otoliths of 391 individuals collected from 2009 to 2014. Normalised elliptical Fourier descriptors (EFDs) were extracted automatically using TNPC software. Principal components analysis (PCA) conducted on EFDs showed no significant effect of side (i.e. left or right otolith). Consequently, all 391 sagittal otoliths were used to identify stocks among six geographical areas: Reunion Island, Mozambique Channel, Rodrigues, South Africa, Madagascar South and Sri Lanka. To investigate the effects of sex, sampling year, sampling season, lower jaw fork length or geographical area on variations in otolith shape, redundancy analyses (RDAs) with permutation tests were conducted. The first four were non-significant (respectively, p = 0.124, p = 0.721, p = 0.197, p = 0.463), but geographical area appeared to discriminate groups significantly (p < 0.05). Furthermore, linear discriminant analysis (LDA) was performed and overall jackknife classification success reached 30%. Finally, a cluster analysis was conducted using Ward’s hierarchical algorithm, which discriminated three different groups. However, each group consisted of individual samples from all geographical areas. In conclusion, our results were unable to identify a clear geographical separation of swordfish at the Indian Ocean scale, corroborating recent genetic studies in this region.
Full Text
File Pages Size Access
8 2 MB Access on demand
Author's final draft 19 433 KB Open access
Top of the page