580 



FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 



the stream with a boxlike trap at the upstream end. 

 The net, 60 feet long and 3^ feet high, was made 

 of two layers of netting for added strength. One 

 layer of netting was ^-inch mesh, bar measure, the 

 other was 1% inches, bar measure. A trap box 3 

 feet square, constructed of K-inch galvanized hard- 

 ware cloth on a wooden frame, had a conical en- 

 trance on the downstream face with a 1-inch open- 

 ing at the apex. Rectangular portable traps (36 

 X 18X18 inches), were constructed of %-inch gal- 

 vanized hardware cloth over a wooden frame. 

 Conical entrances at each end were 16 inches deep 

 and had 2-inch openings at the apex. A remov- 

 able top was held in place with dowels and cleats. 



A Petersen tag consisting of two 



-inch- 



diameter plastic discs, one red and one white, was 

 attached to the first dorsal fin by means of a brass 

 pin. One disc was numbered so that identification 

 of individual specimens was possible. Tagged 

 specimens were released as follows: 257 in section 

 E; 417 in section A; and 286 one and one-half miles 

 downstream from section A. Date and number 

 of lampreys tagged and captured are included in 

 table 18. 



Schaefer's method for estimating the total num- 

 ber of specimens in a migratory population was 

 employed. This method is well adapted for esti- 

 mating the numbers of adult lampreys on their 

 spawning run because it takes into consideration 

 the changing abundance of lampreys in the tribu- 

 tary stream during the several weekly periods of 

 tagging and recovery. Even though a variable 

 number of tagged or marked lampreys was re- 

 leased in the stream during the marking periods, 

 this number was directly proportional to the total 

 catch of unmarked or untagged lampreys. In 



table 18 is a summary, by weekly periods, of the 

 number of lampreys tagged, number of tagged 

 lampreys recaptured, and total number of lam- 

 preys captured. 



An estimate of the number of sea lampreys on 

 the 1951 spawning run (table 19) was computed 

 from the formula: 



n ni =m„ 



m„ rn.' 



where, n ai = the estimated number of lampreys 

 based on the a" 1 tagging period and the i'" recovery 

 period; m 0l =the number of lampreys tagged 

 during the a' h period of tagging and recovered 

 during the i' h period of recovery; T =the number 

 of lampreys tagged during the a ,h tagging pe- 

 riod; m a = the total number of tagged lampreys 

 recovered during each a'* recovery period; C,= the 

 total number of lampreys recovered during the i th 

 recovery period; m., = the total number of tagged 

 lampreys recovered during each i' h recovery pe- 

 riod. The summation of n ai values gave a popu- 

 lation estimate (N) of 9,390. 



Fiducial limits at the 95-percent level were cal- 

 culated from the formula: 



V N—n pq 

 N-l"n 



P=V±\ 



where, P=the population estimate at the 95-per- 

 cent confidence level ; -/V= the population estimate ; 

 p = the total number of tagged lampreys divided 

 by the population estimate N; <7=1 — p; n = the 

 total number of lampreys captured; X=1.96 for 

 the 95-percent confidence limits. Upper and 



