194 



Fishery Bulletin 100(2) 



1 40'W 



55 30'N 



Figure 1 



Locations of stations ( • ) where transects were run off southeast Alaska. The offshore line 

 marks the 200-m isobath near the edge of the continental shelf 



harvest levels. Submersible obsei-vations with line-transect 

 methods have been used in attempts to improve estimates 

 of rockfish abundance and to understand rockfish habitat 

 associations (Richards, 1986; Pearcy et al., 1989; Krieger, 

 1992; Stein et al.. 1992; O'Connell and Carlile, 1993). 



Our goal was to assess abundance and habitat use 

 by shortspine thornyhead in the eastern Gulf of Alaska. 

 based on data from existing video records taken during 

 submersible transects. We estimated abundance of fish 

 and invertebrates and quantified substrate type. We ex- 

 plored the relationships between thornyhead abundance 

 and both physical and biotic environmental variables and 

 compared abundance estimates from submersible tran- 

 sects with those from trawl survevs in the same area. 



Materials and methods 



Sources of data were video tapes of 107 bottom transects 

 recorded at 27 stations during submarine dives in June 

 1991 on the outside coast of northern southeast Alaska 

 from Cape Ommaney to Yakatat (Fig.l). All transects were 

 conducted with the Delta submersible. This battery-pow- 

 ered two-man submersible is 4.7 m long, dives to 365 m, 

 and travels 2-6 kni/h. It is equipped with ten 150 W exter- 

 nal halogen lights, internal and external video cameras. 

 a 35-mm external camera, magnetic compass, directional 

 gyro compass, and underwater telephone and transponder 

 that allowed the submersible to be tracked from the sur- 

 face support vessel. The surface vessel recorded LORAN 

 fixes at the beginning and end of each transect. A pilot 



and observer formed the crew of the submersible: the pilot 

 attempted to maintain the submersible within 0.5 m above 

 the bottom at 3-4 km/h while the observer made obsei-va- 

 tions through a starboard porthole. 



Video recordings were made from a downward project- 

 ing external Hi-8 color video camera tilted obliquely for- 

 ward on the starboard side and included a digital read- 

 out of depth, temperature, and height above bottom. Data 

 were collected either in strip transects by using the entire 

 length of each transect, or in quadrats by freezing individ- 

 ual frames from the video. The width of the strip transect 

 was calculated from the height above bottom and field 

 of view of the camera. The field of view formed a trape- 

 zoid on the seafloor beginning almost directly beneath the 

 camera and projecting forward; its maximum width iW) 

 was estimated by calibrations performed on a subsequent 

 cruise that had the same camera and camera configura- 

 tion (Zhou and Shirley, 1997). When the submarine was 

 resting on the bottom, the camera height iH) was 0.93 m 

 and the width of the field of view (i.e. longest side of trap- 

 ezoidal field) was 1.78 m. Transect width i \V) was estimat- 

 ed by 



W= {1.78/0.93) H. 



llZhou and Shirley 1997)]. 



The height above bottom (H) was recorded at one minute 

 intervals during each transect and mean height was used 

 to estimate width ( W) of that transect. All dives were made 

 during daylight between 0600 and 1900 h. At each station 

 a series (usually 4) of parallel transects was run. and spac- 

 ing between transects was about 200 m. Transect lengths 



