Hand sampling methods were adapted from Adams and Bury (2002). We used a random number generator to select 1-m transects to hand sample. Each transect was 1 m long and spanned the width of the stream (Figure 2). At each site 30% of all possible transects were sampled, which is also the approximate percentage of the area of each site sampled. A sample of 30% of all possible transect was chosen to keep the hand sampling effort consistent between sites of differing lengths. Transects that fell in an area that was too steep to sample safely or in a pool that was too deep (>60 cm) to hand sample effectively were moved to the nearest position that could be sampled (Adams and Bury 2002).
We conducted a systematic search of all surfaces within the wetted channel, by starting at the downstream end of a transect and working upstream. All easily movable objects were overturned and amphibians were either washed into 4-mm mesh wire screens or they were captured by hand (Adams and Bury 2002, Bury and Corn 1991). To minimize habitat disturbance, we left objects in place that were embedded, wedged tightly or were too heavy to be moved. Objects left in place were searched by running a hand along the object and into any spaces around the object. Only the surface layer was searched and all objects were returned to their original position. Once hand sampling was completed, sites were allowed to recover overnight prior to initiation of electroshocking. We performed two passes of removal electroshocking. All electroshocking was performed with a Smith-Root Model 12-A battery-powered backpack electrofisher. A setting of 400-500 V, 30 Hz and a 12% duty cycle were used for electroshocking. Due to the small size and low water levels of the streams sampled, we used a 15.2-cm anode ring to allow greater maneuverability. The electroshocking was performed by starting at the downstream net and working systematically upstream. Once the upper net in a section was reached, the crew turned around and electroshocked quickly as they moved back to the lower net. After the downstream sweep, the lower net was carefully inspected for any amphibians that were swept into the net. During electroshocking the entire site was surveyed, while only 30% of each site was sampled by hand.
Captured individuals were processed similarly to the hand sampling protocol, except that a different toe was clipped to indicate the pass of removal sampling on which an individual was captured. All salamanders were checked for marks to see if they had been captured during hand sampling. Pacific giant salamanders were released as close to the point of capture as possible.
A single pass of electroshocking was performed the next day as the recapture pass for mark-recapture estimates. All D. tenebrosus captures were carefully inspected for the presence of marks. All marks were recorded to determine the capture history of any individual.