Results 



A summation of random sample 

 examinations of adult livers for the years 1958 

 and 1959 (tables 1 and 2) showed that of the fish 

 known or suspected to be of hatchery origin, 

 only those from the Eagle Creek National Fish 

 Hatchery were free from the infection. This 

 hatchery has never fed raw salmon. It has been 

 postulated that the disease is probably main- 

 tained in hatchery stocks by the feeding of 

 infected salmon products to juvenile fish. The 

 apparent lack of infection in fish returning to 

 Eagle Creek supports this . 



No acid-fast bacilli were seen in pre- 

 sumably wild sockeye salmon ( Oncorhynchus 

 nerka) from the Okanogan River in 1958. This 

 is in contrast to the findings of the previous 

 year when 36 percent of the sample fish were 

 infected. It is possible that in 1957 a number 

 of hatchery strays may have entered this system. 



The incidence of disease in marked fish 

 returning to the Little White Salmon and Spring 

 Creek National Fish Hatcheries apparently was 

 high (tables 3 and 4) . Among 36 marked 

 populations available for examination, all but 

 6 were positive to some degree. The numbers 

 of fish representing the majority of groups are 

 small, and the observed infection rates are 

 necessarily subject to limited interpretation. 

 There appeared to be a tendency toward higher 

 infection rates in fish marked prior to hatchery 

 release than in random samples. In 1958, 

 marked fall chinook returning to the Little White 

 Salmon Hatchery were 64 percent infected in 

 contrast to 56 percent for a random sample of 

 the same species. In 1959, 52 percent of the 

 marked fall chinook salmon returning to the 

 Spring Creek Hatchery were infected, compared 

 with only 24 percent of the random sample. 

 Similar results were obtained in 1957 when 55 

 percent of the marked fall chinooks returning 

 to Little White Salmon were infected, while 

 acid-fast bacilli were found in only 12 percent 

 of the random sample . 



Two factors, singly or in combination, 

 may play a role in these differences. First, 

 marked fish may be reared in the hatchery for 

 longer periods of time than unmarked fish, 



thereby increasing the risk of becoming diseased 

 through infected food. Secondly, the random 

 sample may become diluted with wild fish, 

 thereby lowering the infection rate . In view of 

 the rather large differences between marked 

 and unmarked samples returning in 1957 and 

 1959, it would appear that the first theory is 

 more reasonable. 



Juvenile fish were sampled only in 1958 . 

 Infections were relatively rare and were found 

 only at the Carson and Coleman National Fish 

 Hatcheries (table 5). It is strongly suspected 

 that the observed absence of infection in young 

 fish is more apparent than real and that 

 bacilli may be present in such small numbers 

 as to make detection virtually impossible by 

 examination of smears. A comparison of the 

 incidence of infection in adult random samples 

 for the years 1957-58-59 (table 6) revealed 

 striking differences in several instances. The 

 most notable changes occurred in the years 

 1957 and 1958, when the infection rate increased 

 at Little White Salmon Hatchery from 12 per- 

 cent to 56 percent, and the Spring Creek 

 station infection rate rose from 2 percent to 

 35 percent. However, no definite trend could 

 be established towards increased numbers of 

 infected fish in the areas studied for 3 years . 

 It is suspected that the pronounced fluctuations 

 were influenced by the relative amounts of 

 infected food fed during each of the 3 brood 

 years . 



That elimination of raw salmon 

 products as food may drastically reduce or 

 eliminate mycobacterial infections in fishes 

 was strongly suggested at a State -operated 

 steelhead trout hatchery. Prior to the 1960-61 

 spawning season, adults returning from salt 

 water showed a high percentage of infection. 

 Fish returning during the 1960-61 and 1961-62 

 spawning seasons were from brood years in 

 which no raw salmon products had been fed. 

 The author was unable to find acid-fast 

 bacilli in fish examined during these two 

 spawning seasons. 



However, other factors may be in- 

 volved. Harlan E. Johnson noted (personal 



