In the second year of 2 years of study, the movements of juvenile spring Chinook salmon (Oncorhynchus tshawytscha) and juvenile summer steelhead (Oncorhynchus mykiss) through Detroit Reservoir, passing Detroit Dam, and migrating downstream to Portland, Oregon, were studied during a 1-year-long period beginning in February 2013. The primary purpose of the study was to provide empirical data to inform decisions about future alternatives for improving downstream passage of salmonids at Detroit Dam. A secondary purpose was to design and assess the performance of a system to detect juvenile salmonids implanted with acoustic transmitters migrating in the Willamette River. Inferences about fish migration were made from detections of juvenile fish of hatchery origin at least 95 millimeters in fork length surgically implanted with an acoustic transmitter and released during the spring (March–May) and fall (September–November) of 2013. Detection sites were placed throughout the reservoir, near the dam, and at two sites in the North Santiam River and at three sites in the Willamette River culminating at Portland, Oregon. We based most inferences on an analysis period up to the 90th percentile of tag life (68–78 days after release, depending on species and season), although a small number of fish passed after that period as late as April 8, 2014. Chinook salmon migrated from the tributaries of release to the reservoir in greater proportion than steelhead, particularly in the fall. The in-reservoir migration behaviors and dam passage of the two species were similar during the spring study, but during the fall study, few steelhead reached the reservoir and none passed the dam within the analysis period. Migrations in the reservoir were directed and non-random, except in the forebay. Depths of fish within 25 meters of the dam were deeper in the day than at night for Chinook salmon and similar in the day and night for steelhead; steelhead generally were at shallower depths than Chinook salmon. The primary factors affecting dam passage rates were seasonal dam operating conditions and diel period. Fish passage rates were much greater during the spring and summer than in the fall and winter, and the difference was attributed to the availability and use of the spillway near the top of the dam during the spring and summer. The flood-control purpose of the reservoir prevented spillway use during much of the fall and winter because of the low forebay elevation. Passage rates at night were greater than in the day during spring and summer (4.2 times) and during the fall and winter (14.9 times). Fish length, dam discharge, and forebay elevation also affected dam passage rates. Travel times from Detroit Dam passage to the downstream sites were shorter during the fall and winter than during the spring and summer, and were less than a median of 8.68 days to Portland. The estimated survival in the 11 kilometers (km) between Detroit Dam and the Minto Dam forebay was lower than in the remaining 241 km to the Portland site. Estimated survival per 100 km in the free-flowing reach from Minto Dam to Portland was 0.675–0.836, depending on species and season, and was similar to other free-flowing rivers in the Western United States. The high probability of fish in the reservoir reaching the dam, the chance for repeated presence near the dam, the fish depths, and the factors known to affect passage rates suggest that a properly designed surface passage route could be a viable downstream passage alternative for juvenile Chinook salmon and steelhead at Detroit Dam.

As part of the evaluations conducted at Detroit Dam, we continued to refine and improve methods for monitoring fish movements in the Willamette River. The goal was to develop stable, cost-effective, long-term monitoring arrays suitable for detection of any Juvenile Salmon Acoustic Telemetry System (JSATS)-tagged fish in the Willamette River. These data then could be used to estimate timing, migration rates, and survival of JSATS-tagged fish from various studies in the Willamette River Basin. The challenge, however, is that acoustic telemetry generally performs poorly in shallow, turbulent water, like that found in the Willamette River. We successfully designed, deployed, and maintained a series of monitoring sites near the Oregon cities of Salem, Wilsonville, and Portland. In the spring, detection probabilities at these sites ranged from 0.900 to 1.000. In the fall, the detection probabilities decreased and ranged from 0.526 to 1.000. The lower detection probabilities, particularly at the Salem site (0.526), were owing to loss of data caused by abnormally high flows as well as the 2013 Federal government shutdown, which prevented us from servicing the equipment. The monitoring sites that we installed seem to be robust and enable the efficient use of acoustic-tagged fish for studies of migration or survival in the Willamette River and similar environments.