Comparison of genetic and visual identification of cisco and lake whitefish larvae from Chaumont Bay, Lake Ontario

Abstract

Cisco (Coregonus artedi) are an important component of native food webs in the Great Lakes, and their restoration is instrumental to the recovery of lake trout (Salvelinus namaycush) and Atlantic salmon (Salmo salar). Difficulties with visual identification of larvae can confound early life history surveys, as cisco are often difficult to distinguish from lake whitefish (Coregonus clupeaformis). We compared traditional visual species identification methods with genetic identifications based on barcoding of the mitochondrial cytochrome c oxidase I gene for 726 coregonine larvae caught in Chaumont Bay, Lake Ontario. We found little agreement between the visual characteristics of cisco identified by genetic barcoding and the most widely used dichotomous key, and the considerable overlap in ranges of traditionally utilized metrics suggests that visual identification of coregonine larvae from Chaumont Bay is impractical. Coregonines are highly variable and plastic species and often display wide variations in morphometric characteristics across their broad range. This study highlights the importance of developing accurate, geographically appropriate larval identification methods to best inform cisco restoration and management efforts.