FISHERY BULLETIN: VOL. 72. NO. 1 



01 



o 



o 



o 

 £ 



O 



E 

 to 



D 



E 



LO 







E 



1/3 



a 



o 



01 ai 

 5 ^ 



J) 



Q. ^ 



E § 

 o -^ 



U1 



'5 ->^ 



a ^ 

 £ .§ 

 o — 



5" 01 



01 J<: 



a 

 £ 

 o  



CO 



01 01 



ai -1^ 



gi oi 



Oj 



D 







.? Ol 



E H 



O '- 

 lO 



gi 01 



^2 



ooomcNOO ooco 



o 

 o 

 o 



lO 00 

 -^ CO 



CN in 



CN — 



I ,— oiococoron ,-^■^■■^■00 

 oooo*— 000 inmu^o 

 *oor\ cococo TTTTTrco 



in 

 o 



o t^ 



O CO 



o 



o 



I t 1 



, GOO^COCN-— oinco-^O'— ^ 

 — OOt^00OO<3iOO00CX5 

 OOO-OOCsOO-t^cOtO^'O^ 



o o 



CO o 



t^ o 



o o 

 -- ^ 



<3 IT) 



o 

 o 

 •o 



lOlOlO^COcsCOI^COCMCO^LOOCOCM 

 TJ-^TTCO^CN^CN^CN^^-^O-OCN 



uo o 



— CO 



o o 



CO CO 

 -» CO 



m o 

 o ^ 



o 

 o 



CM 



csir^oc^j-^coocooO'^oco'^'^Trco 

 cncnCncs-^co-^ooco^co^-^"^-^^ 



O CM 



O CN 

 CO CO 



•o — 



10 U-) 



— OJ 



o 

 o 



(\coin"^r^cNLOcooooc3or^"^OcN 



CN>OOinCMCNCN'— OOOOCNlO'^r^ 

 CNCOCM^CN ^—000 CN-^COCM 



IT) 

 CO 

 CO 



O IT) 



o -a- 



■^ CM 



m IT) 

 — -"J 



UO CO 



o 

 o 



•OOOCO-^-rfTTCOO^OOO'^^OO 

 COCOCO^lOiolOOOOOOin'OoO' 



— ^^ — •^-^Tjcooo-o-^-^'^o 



o mo 



o 00 



CO CO 10 



O CN 

 CO — 



m CM 



CN CO jj, 0) CN CO 

 >- >- O .£ ^r ■- 



Q) Jl 1) 



CD ^ CO 



E ■- >- 

 Ji > > 



Q Q. 



H Q Q 2 ai Q Q 



a. 



01 ^-<N 



2 i I S 



0) ./I > .2 



l/> 



01 £ 

 E o 



"O 01 



D ■- 

 O -x: 



o;S 



CO 



1) ^- 



data on species composition. It was not practical 

 to totally weigh each catch as it came aboard or 

 to keep the fish from individual sets separated 

 for later weighing; consequently the captain's 

 catch weight estimates had to serve as our quan- 

 titative standard. The accuracy of the captain's 

 estimates was established by comparing the 

 daily total of his estimates with the daily fish 

 house landing records for the G/iIf Ranger (Table 

 2). We believed the accuracy of these estimates 

 ()•- = 0.97) justified our utilizing them for eval- 

 uating diver estimates and for quantifying 

 experimental data (Figure 3). 



RESULTS 

 Diver Estimates 



The validity of scuba-diver observations was 

 evaluated by comparing the divers' estimates of 

 the total number and species composition of fish 

 present at a structure with data obtained from 

 the purse seine catch at that structure. Numer- 

 ical estimates obtained by the divers for coastal 

 pelagic school fish were converted to weight, 

 utilizing a catch average of approximately 22 

 fish per kilogram to permit comparison with 

 purse seine catch data. 



The comparison of diver estimates with the 

 captain's estimates for the corresponding purse 

 seine catches are plotted in Figure 4 for data 

 collected 17-21 August 1970 (Table 1). Data 

 from 24 August to 27 August were not included 

 in this comparison because schools of little tunny 

 {Euthijiunm oUetteratus) began following the 

 purse seiner and were occasionally observed 

 attacking and scattering the structure-attracted 

 fish schools before the purse seine set was com- 

 pleted. A linear regression analysis of the mean 

 for each set of paired diver estimates {Y= 76.5 + 

 0.56X; ?•- = 0.68) indicates that although con- 

 siderable variation does exist, fish schools less 

 than 182 kg (400 lb) tend to be slightly overesti- 

 mated while the larger schools are increasingly 

 underestimated. A linear regression analysis 

 was also calculated for each diver's individual 

 estimates and these calculations indicated 

 that estimates made by diver 2 tend to be more 

 accurate than the more conservative estimates 

 made by divers 1 and 3. 



The purse seine catch sample data indicated 

 scuba divers were able to identify the major 



J 



184 



