returned voluntarily by sport fishermen and com- 

 mercial troll and trawl fishermen. A $2 reward was 

 offered for the return of the tags. Several fishermen 

 were personally contacted to clarify the information 

 they provided and to seek specific information on 

 where and how they fished; all the fishermen were 

 cooperative. Assuming these fishermen were repre- 

 sentative of all those who returned tags, we believe 

 that the overall recovery information was accurate. 



The size and sex distributions of the tagged ling- 

 cod are shown in Graphs I, II, and III of Figure 2. 

 Eighty-six percent of all tagged lingcod were sexed, 

 and of this sample 87% were males. The reported 

 size ranges at maturity are 40-46 cm for males and 

 70-76 cm for females (Forrester 1969; Hart 1973). 



Operationally, we define migratory and nonmigra- 

 tory lingcod as fish recaptured at distances greater 

 than and <8.1 km (5 mi), respectively, from the tag- 

 ging site. This reference distance has been used for 

 similar purposes in previous tagging studies. Since 

 the recovery locations were usually given by the 

 name of a geographical location such as "Middle 

 Bank" or "Turn Island", there was some impreci- 

 sion in estimating the distance moved. However, the 

 fishing area associated with such named locations 

 is <8.1 km in diameter. Thus, for example, a fish 

 tagged on Middle Bank and recaptured on Middle 

 Bank was assumed to have travelled <8.1 km. 



Chi-square contingency table analysis was used for 

 comparing recapture rates by tag type and sex, for 

 comparing release-length frequency distributions of 

 migratory and nonmigratory recoveries, and for 

 comparing migrational tendencies by sex. A chi- 

 square goodness of fit test was used to test the null 

 hypothesis that the release-length distribution of all 

 recaptured lingcod was the same as that of all 

 tagged lingcod. For both of the length-frequency 

 tests, lengths were grouped into 5 mm intervals, but 

 at the tails of the distribution the intervals were 

 wider than 5 mm to follow the rule for chi-square 

 analysis that no expected cell frequency should be 

 <1.0 and that no more than 20% of expected cell 

 frequencies should be <5.0 (Zar 1974, p. 50). One- 

 way analysis of variance was used to test the null 

 hypothesis that the average time between tagging 

 and recapture was the same for fish that had 

 migrated different distances. 



Most of our tagged males and about half of our 

 tagged females were large enough to be reproduc- 

 tively mature when tagged. 



Results 



There were no significant differences among 



Length 

 at maturity 



70 

 Length cm 



120 



Figure 2.— Length-frequency distributions of tagged lingcod. I 

 - known male lingcod tagged; II - known female lingcod tagged; 

 III - all lingcod tagged; IV - release length distribution of all 

 tagged lingcod recovered less than 8.1 km from release location; 

 V - release length distribution of all tagged lingcod recovered more 

 than 8.1 km from release location. 



recovery rates by tag type (x^ = 1.90 with 2 df; 

 0.26 <P < 0.50) (Table 1). However, we suspect 

 from limited double-tagging and aquarium holding 

 of tagged fish that the large dart had better reten- 

 tion qualities for lingcod than the other two tag 

 types. 

 Through October 1985, 157 (9.3%) tagged Mngcod 



155 



