792 



Fishery Bulletin 97(4), 1999 



with one time interval to either side were pooled for 

 each orientation (21 time intervals for each orienta- 

 tion) and compared by using a Watson-Williams two- 

 sample test for mean angles (Zar, 1984). There was 

 a significant difference (F=44.7, P<0.001 ) in the mean 

 angle of trawl marks detected along north-south 



E 

 S 



Z 10 



A/ =405 



Mean = 0.77 km 



SD = 0.26 



n 



rx 



Length (km) 



Figure 4 



Frequency distribution of trawl-mark length per 10-min interval. Ten 

 minute time inters'als ranged from 0.77 to 1.2 km in length (mean=0.91 ). 

 Data are from 20 randomly selected 10-min intervals from 10 different 

 sidescan tracks. 



orientated track lines (mean=357.91 compared with 

 east-west orientated track lines (mean=294.5) at 

 these intersection points. 



The mean density of trawl marks in each CDF&G 

 reporting block fitted a quadratic relationship with 

 water depth (y=-0. 0002x2 + 0.1374x -i- 2.9353, 

 7?2=o.501, P=0.64). The lowest density of trawl marks 

 was observed in the shallowest water with maximum 

 number of trawl marks observed at ~400m depth 

 (Fig. 8). Mean annual trawl hours ( 1990-94) also fit- 

 ted a quadratic model with water depth for each re- 

 porting block {y=-0.Q082x'^ + 8.7015x - 960.57, 

 i?2=o.723, P<0.001) (Fig. 8). There was a significant 

 positive correlation between the density of trawl 

 marks and the mean annual number of trawl hours 

 per reporting block (r,=0.573, P=0.048). 



The fishing blocks off Eureka can be separated into 

 three depth zones on the basis of fish assemblages 

 analyzed by DCA (Fig. 9). The upper and mid slope 

 blocks are relatively close in DCA ordination space, 

 whereas the shelf blocks showed a distinct assem- 

 blage structure. The shelf blocks nearest the coast- 

 line have effective fishing depths of <102 m (Table 

 2). From logbook data, the dominant species at these 

 depths are flatfishes (e.g. sanddabs, Petrale sole, and 

 English sole) and rockfishes. Dominant species from 

 trawling on the upper slope ( 175 to 437 m) primarily 

 consisted of Dover sole and rockfishes. Thornyheads, 

 Dover sole, and sablefish are the dominant species 

 on the mid slope (effective fishing depth from -595 

 to 837 m). 



The mean density of trawl marks was significantly 

 different (Kruskal-Wallis //=287.4, P<0.001) among 

 the three depth zones recognized by DCA 

 (Fig. lOA). Upper and mid slope depths 

 were not significantly different from one 

 another (P>0.05) but both had signifi- 

 cantly greater numbers of observable trawl 

 marks than the shelf depth zone (P<0.05). 

 Mean annual fishing effort per block also 

 showed significant differences among 

 depths (Kruskal-Wallis //=9.5, P==0.009, 

 Fig. lOB). Fishing effort was lowest in the 

 shelf habitat and highest in the upper slope, 

 although the upper slope and deep slope 

 depth zones were not significantly differ- 

 ent fi-om one another (P>0.05). 



The cumulative annual area swept by 

 commercial bottom trawls, as calculated 

 from logbook data, was compared to the 

 total area in each reporting block (Fig. 

 11). The entire study area was trawled 

 -1.5 times on an average annual basis. 

 The area trawled exceeded the total block 

 area in the majority of the blocks. This 



