tributaries enter this reach, the lower reaches of which are non- functional from a fisheries 

 perspective (Pierce et al. 2001). Water quality is impaired (Ingman et al. 1990, Pierce et 

 al. 1997)) and riparian health declines (Marler 1997). This reach supports the lowest 

 salmonid densities for the entire Blackfoot River, with tributary spawning fish (rainbow 

 trout. WSCT and bull trout) in Very low abundance (Results Part II). Mainstem spawning 

 brown trout are in higher abundance, compared with tributary spawning fish. 



Methods 



Geomorphic assessments were completed using modified Rosgen level II channel 

 surveys (Rosgen 1996), and modified Wolman pebble counts (Wolman 1954). We 

 measured sinuosity, valley slope, channel slope, radius of curvature, meander length and 

 belt width using GIS with ADAR high resolution (one meter) imagery and USGS 7.5 

 minute quads. To calculate belt width, meander length and radius of curvature, we 

 selected a reference reach (of two fiall meanders) from the mid-portion of each reach and 

 calculated these variables with ADAR imagery using GIS. 



Modified Wolman pebble counts involved a single pebble count cross-section, 

 within wetted and bankfiill widths at a morphologically stable riffle near the mid-portion 

 of each reach, as a simple index to spawning substrate quality. This method measures 

 only the particle size on the substrate surface and likely underestimates the amount of 

 fines within the substrate and within redds. Based on sample particle size-classes, we 

 define "fine" sediment as < 0.31". This small particle closely corresponds to the particle 

 sizes that negatively influence successfiil reproduction of native salmonids (Weaver and 

 Fraley 1991, 1993; Magee 1996). We defined the suitable spawning size substrate 

 between 0.31" - 2.5" as measured at the intermediate axis, a method consistent with 

 recent bull trout spawning studies (Dunham and Reiman 2001) 



Habitat survey methods began at the upper limit of the upper reach and proceeded 

 down river through all three reaches. We measured stream channel distance using a 

 Garmin 3+ global positioning satellite receiver (GPS) unit. Measured pools were 

 randomly selected using a starting pool (pool 1-4), and then every fourth pool and the 

 preceding downstream riffle were systematically measured using a survey rod and 300' 

 tape. Measurements included: total pool length, maximum pool depth, riffle crest depth, 

 and wetted widths at the pools maximum width and the riffle crest. The difference 

 between maximum pool depth and riffle crest depth was used to calculate residual pool 

 depth. We also calculated distance between pools, and adjusted pool frequency to 

 number/ 1000'. During intensive pool surveys, we estimated pool cover based on a visual 

 estimate of percent of the pool surface area covered by large woody debris (LWD). 



Water temperatures (48-minute intervals) were monitored using Tidbit data 

 loggers in each of the three reaches (Dalton Mountain Bridge (rm 101.1), Cutoff Bridge 

 (rm 70.2), Raymond Bridge (rm 58.4)) from January 2002 through October of 2003 

 (Appendix I). We used a Mann- Whitney rank sum t-tests to test the relaUonship between 

 core winter (January and February) and core summer temperatures (July and August) 

 between the rm 101.1 site and the rm 58.4-mile site. Differences were considered 

 significant at < 0.05. 



In order to determine large woody debris (LWD) stem densities in the three 

 stratified river reaches, we counted and visually measured all large woody debris within 



57 



