FISHERY BULLETIN: VOL 79, NO. 2 



Table 2. — Comparison of differences in recovery rates between 

 taggers for paired tagging of adult Gulf menhaden conducted in 

 1969 in the Gulf of Mexico. All chi-square values obtained from 

 2x2 contingency tables were highly significiant (P<0.001). 



Area Tagger No. tagged Actual recoveries Chi-square 



1,118 

 713 

 402 

 245 

 456 

 163 



77.05 



36.54 



120.04 



' Release totals in Table 3 contain an additional 2,500 tags released by tagger 

 A working alone. 



eries and numbers released with adjustments for 

 differential tagger induced mortality (1969 only) 

 for the adult study are given in Table 3. (The 

 number of adjusted recoveries differs from the 

 number published by Pristas et al. (1976), proba- 

 bly because slightly different methods were used 

 to adjust known recoveries and somewhat dif- 

 ferent criteria were used to judge the suitability of 

 some returns.) 



Adjustments of numbers released to account for 

 tags lost from initial mortality and shedding had 

 to be somewhat arbitrary. Kroger and Dryfoos 

 (1972) reported on a series of short-term tagging 

 mortality and shedding studies on Atlantic 

 menhaden, B. tyrannus, which tested a variety of 

 methods of insertion, fish size, chemical treatment 

 of tags, and tag size. Of this series, two were with 

 fish size and tagging methodology similar to the 

 Gulf adult study. Losses due to mortality and ini- 

 tial shedding in these two experiments were 10 

 and 249c . Since these experiments were conducted 

 under better conditions than field tagging, I 

 selected a value nearer the higher of the two esti- 

 mates, 20%, as a realistic mean rate of Type A tag 

 loss, and 10-30% as a range for testing the sen- 

 sitivity of this assumption. 



Type B errors can occur when tags are shed 

 throughout the recovery period, when tagged fish 

 have a higher rate of mortality, or when tagged 



Table 3. — Number of adult Gulf menhaden tagged in late 

 spring, by area, year, and adjusted number recovered. Gulf of 

 Mexico. 



fish emigrate from the fishing grounds. Errors of 

 this type will cause overestimation of total mortal- 

 ity and natural mortality but not necessarily rate 

 of fishing. No corrections were made for Type B 

 errors because no long-term studies of shedding or 

 mortality under this category have been con- 

 ducted with menhaden. Since tagging wounds 

 heal within a few weeks after tagging ( Kroger and 

 Dryfoos 1972) and the internal tag is stainless 

 steel with rounded, smooth edges, I would expect 

 Type B losses to be minimal. Emigration from the 

 fishing grounds is unlikely (see Recruitment sec- 

 tion above). 



Type C errors occur when the tagged individuals 

 are either more or less susceptible to capture than 

 untagged fish during the first year after release. 

 Recovery rates for later years may be representa- 

 tive, however. Adjustment for Type C error is made 

 in the estimating method, as will be shown later 



The Gulf menhaden tag-recovery data are sub- 

 ject to an additional type of error. Sometimes tags 

 from recaptured fish lodged in plant machinery 

 and were not recovered until 1 or more years after 

 entering a reduction plant (Table 4). The retention 

 of tags in plants for more than one season prior to 

 their recovery will cause estimates of total mortal- 

 ity to be too low. Since trial calculations (of simu- 

 lated data) showed that the rates of tag retention 

 noted would cause underestimation by only about 

 5.4% for fish marked with large (adult) tags and 

 3.8% for fish marked with small (juvenile) tags, 

 and attempts to adjust for this bias may introduce 

 additional error, the effects were ignored. [Tag 

 retention rates were not as large as reported by 

 Nicholson and Schaaf (1978). The most serious 

 discrepencies were for two plants during 1972, 

 where retention rates were reported as 5 and 6%. 

 It was discovered that 34 of the 59 tags reported as 

 retained for 1 yr in one plant and 92 of the 93 tags 

 reported as retained in the other plant were er- 

 roneously recorded. Corrected retention rates are 

 2 and <0.1% for the two plants.] 



Plots of log frequencies of adult recoveries for 

 each release year and area (Table 3) on years-at- 



Table 4. — Mean percentage of large ladult) and small (juve- 

 nile) test tags recovered in Gulf menhaden reduction plants, Gulf 

 of Mexico, 



'Adjusted for tagger induced mortality. 



330 



