rected for continuity) showed that there were no 

 significant differences at the 0.01 level, with 1 

 degree of freedom, in return and nonreturn rates 

 between tag types for fish at liberty for 1, 2, or 3 yr 

 (Table 2). However, returns were significantly 

 better at the 0.05 level for metal tags in the third 

 year. Further, there was no significant difference 

 at the 0.01 level in the first-year return and non- 

 return rates between the two types of dart tag, 

 whether comparing each year individually or 

 comparing all years combined (Table 3). 



We also tested for differences in return and non- 

 return rates between age-groups. Fish were aged 

 from unpublished length-age tables (Rivas^). 

 Chi-square values for fish tagged at ages 1, 2, 3, 



^L. R. Rivas, Southeast Fisheries Center Miami Laboratory, 

 Natl. Mar. Fish. Serv., NOAA, 75 Virginia Beach Drive, Miami, 

 FL 33149. 



and 4 + , were not significant at the 0.01 level (Ta- 

 ble 4). The results of the chi-square test indicated 

 that tag types and ages could be combined. 



Unweighted and weighted linear regression 

 models were used to estimate immediate tag shed- 

 ding rate (1 - p) and instantaneous shedding rate 

 (L) (Table 5). The unweighted model for both tags 

 combined yielded an estimate of immediate tag 

 shedding ( 1 - p) to be 0.040 (0.042 for the weighted 

 model). The overall estimate of the instantaneous 

 rate of tag shedding (L) on an annual basis using 

 the model was 0.205 (0.186 for the weighted 

 model). (The annual rate analog for L from the 

 unweighted model is 0.19.) Therefore, the results 

 from each model were similar. We chose to use the 

 unweighted results, which give a slightly higher L 

 value. While results of the chi-square test indi- 

 cated that tag types could be combined, estimates 

 were also made for each tag type separately to 



Table 2. — Chi-square tests (df = 1) of equality of yearly return and nonreturn rates between double-tagged releases for 1971-77 

 combined, for A = 1 , 2, or 3 yr at liberty, using plastic or metal dart tags on bluefin tuna in the northwestern Atlantic Ocean. The number 

 of returns of double- tagged fish retaining both tags is n^^/^ and those retaining only one tag is n^g/i- 



•P«0.05. 



T.ABLE 3. — Chi-square tests (df = 1) of equality of return and nonreturn rates between double-tagged releases recaptured the same year 

 using plastic or metal dart tags on bluefin tuna in the northwestern Atlantic Ocean. The number of returns of double-tagged fish 

 retaining both tags during the first year after release is n^^j , and those retaining only one tag is n^jgi. 



•PsO.05 



Table 4. — Chi-square tests (df = 1) of equality of return and nonreturn rates by estimated age between double- tagged releases for all 

 years 1971-77 combined, recaptured the same year, using plastic or metal dart tags on bluefin tuna in the northwestern Atlantic Ocean. 

 The number of returns of double- tagged fish retaining both tags during the first year after release is n^j and those retaining only one 

 tagis«dsl- 



182 



