Survey accuracy was evaluated using the concept of sighting 

 potential. The possibility of seeing a subject that was swimming 

 or loafing in the stream was defined as the sighting potential, 

 expressed as a percentage of the length of stream surveyed. The 

 average sighting potential for land surveys was estimated at 

 64%, whereas the average sighting potential for wading and 

 floating was about 95%. 



Several stream characteristics were recorded during the 

 surveys. The number of logjams, cliff loafing sights, mid-stream 

 loafing sights, stream bed composition, bank vegetation type, 

 percent ground cover, channel type and human proximity were 

 summarized for each mile of stream (see Appendix A for 

 definitions of the above variables) . 



Flow rate, stream depth and width were measured at the mouth 

 of each creek. Flow rate was determined by timing a float 

 attached to a 5-meter length of monofilament. Steam depth was 

 measured using a calibrated rope weighted with a rock, and stream 

 width was estimated using a pair of binoculars as a range finder. 

 Volume of stream flow was then calculated using methods described 

 by Robins and Crawford (1954) . 



When harlequins were sighted, time, exact location, stream 

 habitat characteristics, and notes on behavior were recorded. 



12 



