o 



X 



< 

 o 



2:' 

 o 

 t- 

 co 



i 



Ul 



Q: 4 



UJ 



m 

 5 



S4 



UJ 



> 

 Q: 



Ul 



lo 



§4 



w 

 q: „ 



I JULY 14,15 



-T 



200 100 



DOWNSTREAM 



MAY 6 



JUNE 14 



JUNE 27 



JULY 9 



SEPT. 13 



J 



~r 



100 



1 T 



200 300 



I UPSTREAM 



DISTANCE FROM BONNEVILLE DAM (KILOMETERS) 



400 



500 



Figure 5. — Distribution of floaters observed in 8-km. sections from the mouth of Columbia River, Astoria, 

 Oreg., to its confluence with the Snake River and up the Snake River to Lewiston, Idaho (720 km.). 

 May 6 to September 13, 1955. 



It is a simple matter to modify the standard 

 Petersen-type of estimate to fit the circumstances 

 of our experiment, as is shown below. But first, 

 it is instructive to identify our experiment as a 

 typical example of a much larger class of CIR 

 (change-in-ratio) experimental techniques (Paulik 

 and Robson, 1969). 



CIR experiments are designed to estimate 

 population characteristics, such as abundance, 

 productivity, rate of exploitation, and survival 

 and mortality rates. Any CIR estimate is based 

 on the observed differences in the relative num- 

 bers of two distinguishable types of individuals 

 at two points in time when the population is 

 observed. In our experiment the two types of 

 carcasses are those that are marked (x-type indi- 

 viduals) and those that are unmarked (y-type 

 individuals); the first observation of the popula- 

 tion is made immediately before tagging (time 1), 

 and the second observation is made at time of 

 recovery (time 2). 



Throughout the following analysis an important 

 distinction is made between carcasses (both 

 tagged and untagged) that are potentially re- 

 coverable at the time of death and carcasses 

 that are unrecoverable at the time of death. 

 These two classes of carcasses are mutually ex- 

 clusive. A potentially recoverable carcass is a 

 more or less unmutilated carcass of a fish that 

 sinks to the bottom of the river after it dies. 

 After several days, it may or may not float to 

 the surface and drift through the area in which 

 recovery crews are searching. Some recoverable 

 carcasses do not float because they are buried or 

 wedged on the bottom, eaten by scavengers, 

 stranded on the bank, or otherwise prevented 

 from floating (these carcasses are still considered 

 recoverable). An unrecoverable carcass is the 

 carcass of a fish so severely mutilated at the time 

 of death that it cannot float to the surface and be 

 recovered. Thus, a recoverable carcass has a 

 fixed chance of floating to the surface where it 



470 



U.S. FISH AND WILDLIFE SERVICE 



