A novel field method to distinguish between cryptic carcharhinid sharks, Australian blacktip shark Carcharhinus tilstoni and common blacktip shark C. limbatus, despite the presence of hybrids
Johnson, GJ and Buckworth, RC and Lee, H and Morgan, JAT and Ovenden, JR and McMahon, CR, A novel field method to distinguish between cryptic carcharhinid sharks, Australian blacktip shark Carcharhinus tilstoni and common blacktip shark C. limbatus, despite the presence of hybrids, Journal of Fish Biology, 90, (1) pp. 39-60. ISSN 0022-1112 (2017) [Refereed Article]
Multivariate and machine-learning methods were used to develop field identification techniques for two species of cryptic blacktip shark. From 112 specimens, precaudal vertebrae (PCV) counts and molecular analysis identified 95 Australian blacktip sharks Carcharhinus tilstoni and 17 common blacktip sharks Carcharhinus limbatus. Molecular analysis also revealed 27 of the 112 were C. tilstoni × C. limbatus hybrids, of which 23 had C. tilstoni PCV counts and four had C. limbatus PCV counts. In the absence of further information about hybrid phenotypes, hybrids were assigned as either C. limbatus or C. tilstoni based on PCV counts. Discriminant analysis achieved 80% successful identification, but machine-learning models were better, achieving 100% successful identification, using six key measurements (fork length, caudal-fin peduncle height, interdorsal space, second dorsal-fin height, pelvic-fin length and pelvic-fin midpoint to first dorsal-fin insertion). Furthermore, pelvic-fin markings could be used for identification: C. limbatus has a distinct black mark >3% of the total pelvic-fin area, while C. tilstoni has markings with diffuse edges, or has smaller or no markings. Machine learning and pelvic-fin marking identification methods were field tested achieving 87 and 90% successful identification, respectively. With further refinement, the techniques developed here will form an important part of a multi-faceted approach to identification of C. tilstoni and C. limbatus and have a clear management and conservation application to these commercially important sharks. The methods developed here are broadly applicable and can be used to resolve species identities in many fisheries where cryptic species exist.
machine learning, offshore net and line fishery, shark identification