of fish was carried out at U. S. Fish and Wild- 

 life Service hatcheries in Ennis and Bozeman, 

 Montana, during 1954 and 1955. All marking 

 was done by clipping various combinations of 

 fins . 



It was assumed at the beginning of this 

 investigation that among adult fish (above eight 

 inches) the ratio of unmarked fish to marked 

 fish in the fishermen' s catch would be the same 

 as the ratio of unmarked to marked fish in the 

 plantings. This assumption was proved invalid 

 from examination of scale samples collected by 

 creel census clerks. The scales of planted 

 hatchery fish over eight inches can be easily 

 distinguished from scales of wild fish of the 

 same size group. Growth under hatchery con- 

 ditions is more regular and no clear annuli are 

 formed (figure 2). The ratios of marked to un- 

 marked hatchery fish in the creel in 1954 was 

 1:1 . 10, while the ratio of marked to unmarked 

 hatchery fish in the plantings was 1:2.6. In 

 1955 the ratio of marked hatchery fish to un- 

 marked hatchery fish in the catch was 1:1.57, 

 while the ratio in the planting was 1:3 . 125. 

 Therefore, we decided to distinguish hatchery 

 fish from wild fish by the scale method only 

 for final analyses of recovery rates . 



All fish were marked from one to six 

 weeks prior to actual planting. A check was 

 made of regeneration of fins which might occur, 

 especially in the fingerling fish. On the finger- 

 ling plants, there were differences in the amount 

 of regeneration among individual lots of fish. 

 The anals and ventrals showed marked regener- 

 ation in many cases . The pectorals showed the 

 least regeneration, and the dorsals showed a 

 moderate amount of regeneration. Creel census 

 clerks were instructed to carefully check all fins 

 on the fish examined and it was usually possible 

 to discern the original mark on a regenerated 

 fin since regenerated fin rays grow differently 

 from normal fin rays . Therefore, regenera- 

 tion would not upset the accuracy of the finger- 

 ling recovery estimate data since only returns 

 checked by clerks were used in the calculations. 



There is a possibility that unmarked 

 fingerlings might survive better than marked 

 fingerlings. Cooper (1953a) found essentially 

 the same survival rates in unmarked and marked 

 brook and brown trout. Cope and Slater (1957) 



summarized several papers on possible differ- 

 ent survival rates between unmarked and marked 

 fingerling salmon and considered this factor as 

 a possible error in their recovery rate estimates. 

 In the present study, the trout fingerlings were 

 retained in the hatchery for several weeks after 

 marking and any definite mortality trend would 

 have been observed before planting. As shown 

 later, the results of the study would not have 

 been significantly changed, even if the mortality 

 rates of the marked fingerlings were 100 percent 

 greater than those of the unmarked fingerlings . 



Age and growth 



Most of the scale samples (1,273) used 

 in determining age and growth of brown and rain - 

 bow trout and whitefish were obtained from fish- 

 ermen's creels in conjunction with the creel 

 census from 1953 to 1957. A few (36) scale 

 samples from small browns and rainbows were 

 collected in 1957 with an electric shocker. The 

 clerks measured the total length of each fish to 

 the nearest five millimeters or to the nearest 

 one -tenth of an inch. Scales were collected 

 from, the fish on the body area below the dorsal 

 fin and midway between the lateral line and the 

 mid-dorsal line of the fish. Scales were mounted 

 in sodium silicate and read on a standard scale 

 microprojector . 



The samples of scales collected did not 

 contain enough small fish for determination of 

 accurate body-scale relationships. Regressions 

 of body length to anterior scale radius were 

 plotted and calculated by solution of simultaneous 

 equations . Our plotted data showed that linear 

 relationships were sufficiently accurate for 

 practical purposes. Back calculations were 

 made with a straight -Une nomograph, using the 

 Y intercept determined from the regression 

 analyses. 



Fish populations 



Studies on fish populations were made 

 from September 3 through September 13, 1957, 

 on the Firehole and Madison Rivers with elec- 

 trical shocking equipment . No population data 

 were collected from the Gibbon River . 



On the upper Firehole River (Stations A 

 and B on figure 1), a 220 -volt A.C. electric 



