Habitat associations and predictive distribution models of commercially important rockfish species along California's Central Coast

Heather Marie Bolton, California State University, Monterey Bay

Thesis (M.S.) Division of Science and Environmental Policy

Abstract

While commercially important, the red rockfish complex, Vermilion Rockfish (Sebastes miniatus), Canary Rockfish (Sebastes pinniger) and Yelloweye Rockfish (Sebastes ruberrimus), is emblematic of our limited knowledge of the distribution and habitat associations of ecologically and economically important fishes along California’s central coast. We used videographic and photographic imagery from a remotely operated vehicle (ROV), coupled with high resolution multibeam derived maps of the seafloor to determine a) the fine scale habitat associations of red rockfishes along California’s central coast, and b) the potential distribution of small Canary Rockfish (10-40 centimeters TL) beyond surveyed areas using predictive species-specific distribution models. Across the study region, small Canary Rockfish were more frequently observed than Vermilion and Yelloweye Rockfishes, and the highest abundance of red rockfishes were observed in Bodega Bay, California. Nearly all of the Canary Rockfish observed were small, while Vermilion and Yelloweye Rockfishes were subadults and adults. At fine scales (meters), small Canary Rockfish switched their association from sand to rock as total length increased but remained close (12 to 24 meters) to rock-sand interfaces. Predictive models of small Canary Rockfish presence were 74-77% accurate, and bathymetry and distance from interface were important environmental predictor variables. The imagery-based analyses provided important ecological information about each species, while the predictive modeling allowed us to extrapolate beyond the relatively limited area transected by the ROV to the broader study region. This approach of combining methods is applicable to other species and geographies where we have to manage more than we can sample.