In 2004, the U.S. Army Corps of Engineers (Corps) completed a surface-bypass system that passes juvenile salmonids via the entrance of the old ice and trash sluiceway at the Second Powerhouse at Bonneville Dam.  This corner-collector bypass system is attracting ~15% of the migrating salmonids passing Bonneville Dam.  Consequently, the region needs an effective corner-collector PIT-tag interrogation system in order to continue to provide the PIT-tag data required to make estimates of reach survivals for juvenile salmonids and for helping to assess progress toward hydro-system performance standards.  Calculations by a subcommittee of the federal Research Monitoring and Evaluation Hydropower Work Group estimated that the corner-collector detection system will need to detect 60% of the tagged smolts using the exit flume to replace the detections that were being detected at Bonneville Dam before the operation of the corner collector.

To provide accurate estimates of reading efficiencies, we need to know precisely how many tagged fish have passed through the corner-collector system during the testing.  This will be done by releasing PIT-tagged fish within the capture velocity area in front of the entrance.  Tests done in 2006 determined that it did not matter where fish were released within the capture velocity area in front of the entrance and therefore, we will only use one release location for the 2007 tests.  Furthermore, fish-release procedures developed in 2006 proved that large numbers of fish could be released faster than what had been anticipated.  Lab and field tests conducted in 2006 showed that the SST tag was the best tag available.  Consequently, the fisheries community is making that tag model its standard tag starting in 2007.  Therefore, only one tag type will need to be included for estimating the reading efficiencies for the different salmonid species.  All of the results from the 2006 tests simplified the experimental design that will be needed for conducting the reading efficiency tests in 2007.  Although we were able to make preliminary estimates in 2006, we have learned over the years that the effectiveness of PIT-tag systems changes over time.  Because of financial constraints, Digital Angel did not make any revisions to the equipment in 2006.  And since this is a single antenna system, there is no built-in feedback and thus, the only way to effectively evaluate the performance is to release a known number of tagged fish.  

We propose to evaluate the reading efficiency of the corner-collector detection system with the main migrating salmonid populations (i.e., spring Chinook, steelhead, and fall Chinook).  We plan to release spring Chinook and steelhead over the same 2-week period in May and then conduct the tests with fall Chinook during July.  Since we will know how many fish were released, tag-reading efficiencies can be calculated by dividing the number of fish detected by the number of fish released.  Digital Angel or PSMFC will be on hand during these tests to ensure that the electronic equipment is tuned.  We plan to use SST tags to tag 2,200-2,500 smolts from each of the three main salmonid populations:  spring Chinook yearlings, steelhead yearlings, and fall Chinook salmon subyearlings.  Releasing this number of smolts will yield estimates of reading efficiencies with precision levels of around 2%.  If it is difficult to get that many river-run steelhead at the juvenile fish facility, then we will aim to release around 900-950 to get a precision level of around 3%.

Typically, ambient electromagnetic noise if high can significantly reduce the detection ability of PIT-tag systems.  The corner-collector PIT-tag system is unique because it incorporated digital-signal-processing (DSP) into its transceiver design.  In 2006, the DSP appeared to be a significant improvement as detection appeared to remain at the same level during periods of high ambient noise; however, relying on detection of ST-tagged fish made it difficult at times to confirm the performance.  We propose to monitor the noise during 2007 and use SST-tagged fish to compare detection performance during periods of high and quiet noise.