95% confidence interval = N ± 1 .96 (Standard deviation). The 95% confidence 

 intervals for these estimates are found in Appendix B. 



For small stream population assessments, we also used a single pass catch-per- 

 unit effort (CPUE) method, 

 which provides an index of 

 trout abundance (Appendix A). 

 For this CPUE. we also 

 developed simple linear 

 regressions (for fish < and > 

 4.0") to help predict densities 

 from CPUE (Figure 9). These 

 regressions show a significant 

 correlation between CPUE and 

 density (P= <0.001). Small 

 stream size and highly efficient 

 electrofishing conditions in 

 study streams contributed to 

 this outcome. Although these 

 regressions demonstrate CPUE 

 to be an index to population 

 density, CPUE does not include 

 a confidence interval like the 

 actual population density 

 estimate. For this report, 

 CPUE refers to the number of 

 fish collected in a single 

 electrofishing pass and is 

 adjusted per 100' of stream (i.e. 

 CPUE of 8 means 8 fish 

 captured per 100' of sampled 

 stream). Actual population 

 estimates are referred to as 

 density/ 100'. 



CPUE/Estimate Regression 1990-2003 



y=1.3162x-H 0.5495 

 R^ = 0.8682 



Figure 9. Regressions used to predict densities from CPUE for fish < 

 4.0" (top) and for fish >4.0" (bottom). 



Fish were captured using a boat or backpack mounted electrofishing unit. In 

 small streams, we used a battery powered (Smith/Root) backpack mounted DC 

 electrofishing unit. The anode (positive electrode) was a hand-held wand equipped with a 

 1 -foot-diameter hoop; the cathode (negative electrode), a braided steel wire. On the 

 Blackfoot River and Monture Creek, we used an aluminum drift boat mounted with a 

 Coffelt Model VVP-15 rectifier and 5,000 watt generator. The hull of the boat was the 

 cathode and two fiberglass booms, each with four steel cable droppers, served as anodes. 

 We used direct current (DC) waveform with output less than 1000 watts, which is an 

 established method to significantly reduce spinal injuries in fish associated with 

 electrofishing (Fredenberg 1992). Juvenile trout were sampled in the tributaries from 

 August to November. Extra effort was used to sample stream edges and around cover to 



23 



