Using genetic species identification and environmental data to distinguish historical catches of cryptic Blackspotted Rockfish (Sebastes melanostictus) and Rougheye Rockfish (Sebastes aleutianus) in British Columbia

Visually classifying species is the most common method used in fisheries to estimate catch compositions for commercial and survey data, but catch records can be confounded when two or more morphologically similar species are classified as a single species (i.e., cryptic species)—as is the case for Blackspotted Rockfish and Rougheye Rockfish. To partition catches between the two species, I used genetic species identification data in regression models relating the proportion of Blackspotted Rockfish relative to the overall Rougheye/Blackspotted Rockfish catch to measures of set depth, location, and bottom ruggedness. The best model included a negative relationship with longitude and positive relationship with bottom ruggedness. I also used large-scale spatial predictors to estimate historical landings of each species, finding that the inclusion of trap longline commercial data after 2006 caused an increase in the relative proportion of Blackspotted Rockfish caught (out of the total Rougheye/Blackspotted Rockfish landings). Finally, I examined observer accuracy in distinguishing Blackspotted and Rougheye Rockfish and found that while 86% of fish were identified correctly, Blackspotted Rockfish were more likely to be misidentified, leading to a 55% overestimate of the actual Rougheye Rockfish catch and a 14% underestimate of the Blackspotted Rockfish catch. My results indicate that set-specific variables are most useful in estimating proportions of Blackspotted Rockfish and can be used to estimate how spatial shifts in fishing efforts will impact fishing mortality of Blackspotted Rockfish and Rougheye Rockfish.