sample of fish to be examined was selected, 

 the liver of each fish was removed and secured 

 in an individual plastic bag. In most instances 

 the livers were then frozen before transporting 

 them to the laboratory, although some were 

 taken to the laboratory in a fresh condition. 

 Smears of liver tissue were prepared at the lab- 

 oratory, stained by the Ziehl-Neelsen method, 

 and examined for the presence of typical acid- 

 fast bacilli. Fish were classified as tuberculous 

 when typical acid-fast bacilli were found in the 

 stained preparations within a five -minute search 

 period under oil immersion. Each smear was 

 classified according to the number of acid-fast 

 bacilli observed. The classification system 

 used was as follows: 



Negative, no bacilli found during a five- 

 minute search. 



No. 1, only 1 to 20 bacilli found during 

 a five -minute search. 



No. 2, over 20 bacilli found during a five- 

 minute search up to an average of less 

 than 1 in many fields. 



No. 3, about 1 to 24 bacilli on an aver- 

 age in each field. 



No. 4, about 25 to 150 bacilli on an aver- 

 age in each field. 



No. 5, more than 150 bacilli on an aver- 

 age in each field. 



This system is a modification, for the 

 purpose of simplification, of the rather awkward 

 Gaffky scheme of rating smears for Mycobacter- 

 ium tuberculosis outlined by Kolmer, Spaulding, 

 and Robinson (1951). Certain errors are inher- 

 ent in an arbitrary classification system such 

 as the one outlined above . The area of the liver 

 from which the smear was prepared was found, 

 in a few instances, to make a difference of as 

 many as two numbers in the system. There was 

 usually close agreement, not over one number, 

 between individuals reading a particular smear; 

 the most serious error being between negative 

 and number 1 . It is apparent with the use of 

 this system that a comparison of the degree of 

 infection between groups of fish is subject to 

 much less error than a comparison between in- 

 dividual fish. 



19 



COLUMBIA RIVER FISHERY 



General 



In order to obtain an estimate of the in- 

 cidence of tuberculosis in fish entering the 

 Columbia, it was necessary to rely on the com- 

 mercial gill-net fishery. The landings of this 

 fishery represented the only source of sufficient 

 numbers of fish for the survey. Legislation has 

 eliminated all except drift gill nets, dip nets, and 

 sport gear. By far the bulk of the Columbia River 

 catch is now taken in gill nets . 



Commercial gill -net fishing takes place 

 from the mouth of the Columbia to Bonneville 

 Dam (fig. 1), a distance of approximately 140 

 miles . Chinook, silver, blueback, and chum 

 salmon as well as steelhead trout are taken in 

 the commercial fishery. The seasons open to 

 the fishery in 1957 were as follows: 



January 29 (noon) to March 1 (noon) 

 April 30 (noon) to May 27 (6 a.m.) 

 June 20 (6 a.m.) to July 15 (6 a.m.) 

 July 29 (6 p.m.) to August 26 (6 p.m.) 

 October 29 (6 p.m.) to November 29 (6 p.m.) 



Closed weekends existed throughout the 

 open seasons as follows: 



May 1 to Aug. 1 Saturday (noon) to Sunday 



(6 p.m.) 



Aug. 2 to Aug. 25 Friday (6 p.m.) to Sunday 



(6p.m.) 



Sept. 16 to Dec.l Friday (6 p.m.) to Tuesday 



(6 p.m.) 



Whether the sampling of the landings fur - 

 nished a reasonable accurate estimate of the 

 incidence of tuberculosis in fish entering the 

 river depended, in part, on whether the migrat- 

 ing adults were adequately sampled by the gill-net 

 fishery. At the present time there are not suf- 

 ficient data available to determine whether 

 tuberculous fish are more readily taken than non- 

 tuberculous fish by the gill -net fishery. Several 

 sport fishermen, however, upon furnishing dis- 

 eased tissues of fish to our laboratory have 

 reported that these diseased fish did not "fight" 

 like normal fish. In nearly all cases the tissues 

 were heavily infected with tuberculosis . It is 



