intensive investigation. (We subsequently found 

 that summer is usually the period of maximum 

 mortality.) 



The method we used to estimate the total 

 numbers of dead fish was based on a mark-and- 

 recovery technique. Preparatory to making the 

 estimates, we determined the relative floating 

 qualities of tagged experimental (killed, frozen, 

 and thawed) and "natural" river-killed salmon, 

 located points where salmon lodge downstream 

 from the dam before they float, and made system- 

 atic surveys from boats and airplanes to deter- 

 mine the best sites for observing dead fish near 

 the dam. Searchers in boats below Bonneville 

 Dam recovered all the tagged and untagged 

 floating salmon used for the population estimate. 



BCF contracted with the Oregon Fish Commis- 

 sion to make this study. 



ABUNDANCE AND LOCATION OF 

 FLOATING SALMON CARCASSES 



Dead salmon in the Columbia River are most 

 evident as "floaters" (partially decomposed float- 

 ing carcasses); hereafter in this paper this term 

 refers to any floating dead salmonid. Floaters are 

 carried downstream by the current and can usually 

 be seen from a considerable distance. We searched 

 systematically from boats and aircraft in 1954 and 

 1955 and also tagged and recovered floaters to 

 learn how they disperse in the Columbia River. 

 We also introduced tagged dead chinook salmon 

 into the river at Bonneville Dam in both years 

 to provide a basis for estimating total numbers of 

 dead fish; in 1955 the observations from boats 

 were used to estimate the mortality of summer 

 chinook salmon. 



SEARCHES FROM BOATS 



1954 



In 1954 we searched for floaters downstream 

 from Bonneville Dam during three periods (table 

 1). Five boats were used; four were 4.9 to 5.5-m. 

 skiffs propelled by outboard motors, and the fifth 

 was a 7.9-m. inboard Columbia River gill-netter. 

 The skiffs were usually manned by one person 

 and the large boat by two. Efficiency of observa- 

 tions was assumed to be equal for all boats. Each 

 observer was equipped with Petersen disk tags, 

 Polaroid^ glasses, and a dip net to recover floating 



carcasses. Search time was recorded in hours and 

 minutes. Table 1 shows the numbers of floaters 

 (by species). 



Table 1. — Floaters observed during three search periods on 

 Columbia River downstream from Bonneville Dam, July 12 

 to September 12, 195 k, and numbers of chinook salmon float- 

 ers observed per hour of search 



s Trade name referred to in this publication does not imply endorsement 

 of commercial product. 



The searches were all made relatively close to 

 the dam. During the first search period (July 12 

 to August 5), observers in the five boats searched 

 midstream and shoreline sections from the dam 

 to the mouth of the Willamette River. In the 

 early part of this period an additional observer 

 was stationed on shore on a high dock at Ells- 

 worth, Wash., where he could scan the main river 

 channel with binoculars. The second search period 

 (August 9-20) included the same areas. During 

 the final search period (September 4-12), effort 

 was reduced to a roving search from two boats 

 between Ellsworth and the dam and observations 

 from the Ellsworth dock. One of the two boats 

 patrolled the 31 km. of river from Ellsworth to 

 Cape Horn, and the other the 23 km. of river from 

 Cape Horn to the dam (fig. 3). 



1955 



Searches by boat for floaters in 1955 were 

 limited mainly to the main current in midstream 

 on transects across the river channel at eight 

 locations. The boats and equipment used by 

 observers were the same as those used in 1954, 

 and search time was again recorded in hours and 

 minutes. The eight stations and their locations 

 (fig. 3) were (1) St. Helens— 93 km. below the 

 dam, and 1.6 km. upstream from St. Helens, 

 Oreg.; (2) Willamette— 70 km. below the dam, 

 immediately upstream from the confluence of 

 the Willamette and Columbia Rivers; (3) Ells- 

 worth — 53 km. below the dam at Ellsworth, 

 Wash.; (4) Reed Island — 35 km. below the dam, 

 immediately downstream from Reed Island; (5) 

 Cape Horn — 23 km. below the dam; (6) McGowan 



CHINOOK SALMON MORTALITY IN COLUMBIA RIVER NEAR BONNEVILLE DAM 



465 



