T-\Bi.E 11. — Aclital and estimated age eompositioti of 400 

 leil scales read by 6 scale readers 



Age composition (years) 



Estimated. 

 Actual 



18.4 



20. 2 41. 8 



Percent — 



38. 5 28. 2 12. 5 



2.4 

 .0 



would tend to reject the hypothesis tli;it these 

 marks are caused primarily by regeneration and 

 conclude that many of these partial marks 

 occurred naturally. 



Because the Ad-RV and Ad-LV partial marks 

 are double tin marks, we assume that their occur- 

 rence (in catch and hatchery returns) was pri- 

 marily tiie result of maxillary regeneration. 

 Hence, the data in the column Ad-EV of table 

 1-2 reflects the variability in rate of maxillary 

 regeneration. When we compare the percent 

 occurrence of Ad and KY marks with that of 

 Ad-P:V marks and the percentages of Ad and 

 E^' marks in the catches and at the hatcheries, 

 the Ad mark appears to have been caused pri- 

 marily by maxillary regeneration: and single 

 RV or LY marks appear to have resulted from 

 naturally marked fish in addition to maxillary 

 regeneration. 



In view of the above, we combined the esti- 

 mated numbers of partially marked tish with 

 their corresponding full mark for each mark type 

 except for the LV and RV marks. On this basis 



T.\BLE 12. — Recoveries of partially marked fish by region of 

 capture, type of fishery, and type of mark, 1963-66 



1 Data in table are ratios (average for all years) of estimated numbers of 

 partial marks to estimated sum of partial marks and corresponding complete 

 marks expressed in percent. 



'' EV signifies "either ventral." Marks of same general type are combined. 



Table 13. — Estimated catch and escapement of marked fall 

 Chinook salmon of Columbia River hatchery origin by area 

 of recovery, 1 963-66 



I Includes partially marked fish. 



the estimated numbers (jf marked fall cliinook 

 salmon of Columbia Ivi\er hatchery origin in the 

 catch and escapement are summarized in table 13. 



Of the r.Xil brood released from the Vl hatcher- 

 ies with Ad-RM mark, we estimated that 21,600 

 were caught by the \arious fisheries between VMVo 

 and l!)(iC>. An additional ;).40() tish escaped the 

 fisheries and returned to spawn. The catch to 

 escapement ratio for the \-l hatcheries is, there- 

 fore, about (1:1. For the four hatcheries repre- 

 sented by the other marks, this ratio was \->.:\, 

 9:1, .'^) : 1. and 7:1 (P21okoinin, Kalama. Ox- 

 bow, and Spring Creek Hatcheries, respectively). 



The estimated catch of marked fish (Ad-RM) 

 that originated from tlie 12 hatcheries relative to 

 numbers released was 21,616/5,4J:(i,4;59 or 3.97 fish 

 per 1,000 released. Considerable variation in con- 

 tribution occurred between hatcheries. The catch 

 per 1,000 tish released was 0.36, 4.50, 0.59, and 3.78 

 for Pvlokomin, Kalama, Oxbow^, and Spring Creek 

 Hatcheries, respectively. Values for P^lokomin 

 and Oxbow Hatcheries are undoubtedly too low — 

 some of the partial marks must have resulted 

 from maxillary regeneration. If w^e assume that 

 all the partial marks actually originated from 

 Elokomin and Oxbow Hatcheries, their respective 

 contribution would have been 1.53 and 1.95 jier 

 1,000 tish released. These contributions still would 

 be only one-half of those for Kalama and Spring 

 Creek Hatcheries. 



Before proceeding with the estimate of total 

 contribution (marked and unmarked) of hatch- 

 ery-reared fish to the catch, we will consider the 

 precision of the estimate of the total catch of 

 marked Hsh. 



Ecjualion (4). which provides an approxima- 

 tion of the variance of the estimated number of 

 marks in the catch, requires an estimate of the 



COXTRIBTTIOX OF COLUMBIA RIVER HATCHERIES TO FALL CHINOOK SALMON HARVEST 



379 



