198 



Fishery Bulletin 100(2) 



Our obsei-vations are based on a larger number of tran- 

 sects than those surveyed off Oregon, but the consistency 

 of results from the two Oregon studies suggests that the 

 difference is not due to a low sample size there. Our ob- 

 servation that abundance of shortspine thornyheads in- 

 creases with depth is consistent with results from trawl 

 surveys off Alaska, Oregon, and California (Martin and 

 Clausen, 1995, Stark and Clausen, 1995, Jacobson and 

 Vetter, 1996). Off Oregon, their abundance was highest in 

 the 200-400 m depth zone, and decreased sharply between 



2.5 



1 - 



5 - 



18 



^ 



<200 



200-300 

 Depth (m) 



>300 



Figure 4 



Mean abundance of shortspine thornyhead in three depth inter- 

 vals from submersible transects in 1991 and from the 1990 and 

 1993 trawl surveys off southeast Alaska. Sample size (number of 

 transects) and standard errors of the submersible estimates arc 

 indicated in the number and vertical line with each bar 



400 m and 1400 m (Jacobson and Vetter, 1996). Off Califor- 

 nia, the highest abundance of shortspine thornyheads oc- 

 curred at 400-600 m, probably because of the warmer wa- 

 ters off California (Jacobson and Vetter, 1996). Thus, our 

 sampling depths covered the most important depth zones 

 for this species in the waters off southeast Alaska, which 

 are colder than those off Oregon. 



Bottom trawl sui-veys of fishes in the Gulf of Alaska are 

 conducted triennially, and occurred one year before ( 1990) 

 and two years after ( 1993 ) our submersible sui-vey ( Martin 

 and Clausen, 1995; Stark and Clausen, 1995). For many 

 species, trawl sui^vey results may be biased because some 

 species may be herded by the trawl doors into the path of 

 the net, resulting in overestimates of abundance when the 

 "area-swept" method is applied (Ki'ieger, 1992). Other fish 

 in the water column above the bottom may swim over the 

 net and be underestimated by the survey (Balsiger et al., 

 1985). Escape routes under the foot rope and through the 

 larger meshes in the trawl wings are also possibilities. 



The NMFS trawl survey does not cover exceptionally nig- 

 ged rocky habitats that would destroy equipment; conse- 

 quently, species that select high-relief habitats may be un- 

 dei'estimated by the survey, whereas species that select 

 low-relief soft-bottom habitats may be overestimated. Sub- 

 mersible obsei-vations provide a means to quantify the bi- 

 ases inherent in bottom trawl sui-veys. The depth catego- 

 ries we used to analyze the submersible data matched the 

 depth strata used in the NMFS triennial trawl sui-veys. 

 Mean abundance of shortspine thornyheads in submersible 

 sui^veys were several times higher that those in the 1990 

 and 1993 trawl surveys; however, the ratios of abundance in 

 the three depth zones were very similar ( Fig. 4 ). We suggest 

 that trawl sui-veys underestimate the abundance of short- 

 spine thornyheads in the Gulf of Alaska but may be 

 good indicators of relative abundance, patterns of dis- 

 tribution, and stock trends. 



Acknowledgments 



We thank Dan Ito for his encouragement and support 

 for this project. This research was supported by a 

 grant (43ABNF401902) from the U.S. Dept. of Com- 

 merce, National Marine Fisheries Ser\'ice, Auke Bay 

 Laboratory, Juneau, Alaska. 



Literature cited 



Adams, P. B. 



1980. Life history patterns in marine fishes and their 

 consequences for fisheries management. Fish. Bull. 

 78:1-12 

 Auster, P. J., L. L. Stewart, and H. Spunk. 



1989. Scientific imaging with ROVs: tools and tech- 

 niques. Mar Technol. Soc. J. 23:16-20. 

 Balsiger J. W.. D. H. Ito. D. K. Kimura. D. A. Somerton. and 

 J. M Terry. 



1985. Biological and economic assessment of Pacific 

 ocean perch (Sebastea alutus ) in waters of Alaska. U.S. 

 Dep. Commer, NOAA Tech. Memo. NMFS F/NWC-72, 



