In addition, we assume that marked fisii beliave 

 like the unmarked hatchery fish. In particuhir we 

 must assume: 



4. Marked and unmarked hatchery fisli have 



the same survival rates and maturity 

 scliedules. 



5. Marked and unnuu-ked hatcliery llsli lia\e 



tlie same ocean distribution and are 

 etjually vulnerable to the fisheries. 

 Finally, because part of all hatchery releases bear 

 the same mark, we assume that : 



(■>. P>ither the ocean distribution and tiniinp; of 

 migration of fish from each of the hatch- 

 eries are the same or the same proportion 

 of each hatchery's production is marked. 



The appropriateness of the estimating i)roce- 

 dures described above is obviously dependent upon 

 the validity of these assumi)tions. For this reason, 

 the results must be considered as i)reliminarv until 

 sufficient data are collected to judge their validity. 

 For example, if marked and unmarked fish have 

 different sur\ival rates, then the marked to un- 

 marked ratio in the hatchery release is not ap- 

 propi'iate [in eqiurtion (6)] and some adjustment 

 should be made. To test the assumptions, additional 

 studies and data collections were incorporated 

 within the design of the marking experiment. 



In connection with the first assumption (perma- 

 nence of fin marks), marked juvenile fish were 

 held and periodically examined to determine de- 

 gree of nuirk regeneration. Also, illustrated forms 

 on which missing fins or parts of fins are shaded 

 out jn'ovide information on the samplers' in- 

 terpretation of marked fish ( fig. 5 ) . 



To test the second assuin[)tion (origin of fish 

 marked with liatchcry marks), fish were exaniiiuMl 

 for natural marks while lieing marked. In ad- 

 dition, the cooi)erating fishery agencies agreed not 

 to use the combination of marks used in this study 

 on any grou]> of ItXiri-brood chinfX)k salmon. This 

 group will be sampled as -2-, ;]-, and -t-year-old fish 

 during the I'.XiT, I'.HiS, and 1969 sampling years. 

 Any observed marks of the kind used in this stiuly 

 will be from the natural loss of fins. 



A test of scale readers was designed to test the 

 third assumption (accuracy of age assignments). 

 Scales from 400 marked fish of known age were 

 submitted to six readers from the Fisheries Re- 

 search Board of Canada. Wasliington Department 

 of Fisheries, Oregon Game Commission, Fish 



Commission of Oregon (two readers), and Bureau 

 of Connnercial Fi.sheries. Length of fish and date 

 of capture were available for each scale. 



The fourth as.sumption (equality of survival 

 rates and maturity schedules) will be examined 

 frcnn data on a different type of mark applied to 

 fish at one of the liatcheries. At Little White 

 Salmon Hatchery, a fraction of the 19()4 brood was 

 marked with TM 50 (oxytetracycline) that was 

 added to the diet. This mark was selected to iden- 

 tify a hatchery stock because it apparently does not 

 att'ect their growth or survival (Weber and Kidg- 

 way, 19f>T) . As a residt, some of the fisli are double 

 marked — with a finclip and TM 50. Returns to the 

 hatcheiy ii\ llXit;, 19(;T, 19(;s, and 1969 will be ex- 

 amined for changes since release in the proportion 

 of TM 50 marked fish which also have the finclip. 

 In addition to the TM 50 marking at this hatchery, 

 data bearing on the changes in marked to un- 

 marked ratios (i.e., finclip vs. totally unmarked) 

 Itetween release and return will be available from 

 the returns to each of the participating hatcheries. 



With the resources a^'ailable, we could not test 

 the fifth assumption (equality of ocean distribution 

 and vulnerability). Indirect evidence may be ob- 

 tained, however, by comparing the ocean distribu- 

 tion of marked fish with the distribution as 

 determined by past tagging experiments in the 

 ocean. 



The sixth assumption (e<|uality of ocean distri- 

 bution or proportion marked among hatcheries) 

 will be examined from data collected in that part 

 of the study designed to examine the variations in 

 contribution between hatcheries and, of course, 

 from the data collected from each hatchery at the 

 time of release. 



RELEASES OF FISH 



The marking phase of the hatchery contribu- 

 tion study (concluded in 1965) included marking 

 a jiortion of the fall chinook jiroduction of each 

 of the 1-! hatcheries with the same mark; marking 

 a poi'tion of the production of 4 hatcheries with 

 a unique mark; measui'ing the quality of mark- 

 ing; examining hatchery fish for naturally miss- 

 ing fins; and obtaining samples of fish to deter- 

 nune regeneration of marks aiul for measuring 

 certain physical and physiological characteristics 

 of the releases for each of the brood years. Anal- 

 yses of these latter data in terms of explaining 



COXTRIBUTIOX OF COLIMBIA RIVER HATCHERIES TO FALL CHINOOK SALMON' HARVEST 



!71 



