96 



Fishery Bulletin 89(1). 1991 



Discussion 



Longlines appear to be a more cost- 

 effective means of monitoring fish 

 population changes than submer- 

 sibles. However, longlines kill all 

 the collected fish whereas submer- 

 sibles do not. More data can be col- 

 lected on each caught fish at a lower 

 cost with longlines than with a sub- 

 mersible. Size, age, and sex data 

 could be collected from longline 

 catches at a cost of about $5000 per 

 day (in 1984 U.S. dollars), while 

 none of these data were available 

 from the submersible even though 

 it cost about $8000 per day to op- 

 erate. Additionally, the tilefish pop- 

 ulation based on burrow data from 

 the submersible may have been 

 overestimated because (1) "active" 

 burrows were overestimated, (2) 

 width of each burrow-count tran- 

 sect was underestimated, and (3) 

 double counting occurred when 

 transects crossed or came close to 

 crossing. The number of tilefish 

 estimated from tilefish seen was 

 about 50% larger than the estimate 

 based on longlines, and was about 

 four times less than the estimate based on burrow 

 counts. The number of burrows may be more a reflec- 

 tion of population size prior to exploitation if this area 

 was heavily fished prior to our study. 



Tilefish populations were probably underestimated 

 using longline data. But the amount of bias is unknown. 

 Capture probabilities were not constant, and this usual- 

 ly leads to underestimates (White et al. 1982). Recruit- 



