196 



Fishery Bulletin 100(2) 



ed between the shallowest depth category and both deeper 

 depth groups (Table 3, Sheffe test). Substrate type also af- 

 fected abundance (Table 3); transects with hard substrate 

 had a significantly higher density of thornyheads than tran- 

 sects on soft bottom (P=0.016, Mann- Whitney). There was 

 a significant interaction (P=0.01) of depth and substrate in 

 the ANOVA factorial analysis (Table 4) due to the sharp in- 

 crease in thornyhead abundance in depths >200 m on the 

 hard substrate than on the soft substrate (Fig. 3). 



Stepwise multiple regi'ession with all variables resulted 

 in a model that incorporated two variables: substrate type 

 and depth, with substrate type entered first. The final two- 

 variable regression model was 



A = 0.017 S + 0.00016 D - 0.033, 



(/•■-=0.2081 



where A = thornyhead abundance; 



S = proportion of bottom that is hard substrate; and 

 D = depth (m). 



Thornyhead abundance increased with depth and amount 

 of hard substrate, although there was an indication that 

 abundance may have reached maximum levels at depths 

 of200-300m(Fig. 3). 



Discussion 



Film (still or motion) and videotape recordings of tran- 

 sects have been used to assess abundance of aquatic or- 

 ganisms (Auster et al., 1989; Butler et al.'l. Potential 

 biases in such data include systematic underestimation 



Butler, J. L., W. W. Wakefield, P. B. Adams, B. H. Robison, and 

 C. H. Baxter 1991. Application of line transect methods to 

 sui'vcying demersal communities with ROVs and manned sub- 

 mersibles. Proceedings of the IEEE Oceans '91 Conference, p. 

 689-696. histitute of Electrical and Electronic Engineers, Pis- 

 cataway, NJ. 



