Steve Whitlock, Rebecca Buchanan

University of Washington, 1325 Fourth Ave Suite 1515, Seattle, WA 98101

The PIT tag program in the Columbia River Basin has been an elemental component of monitoring and evaluation of vulnerable salmonid populations and their response to operational changes within the federal hydrosystem for over two decades. The monitoring system for juvenile migrants has depended on detection arrays that were installed in the juvenile bypass systems starting in the 1980s and have been present at seven of the eight lowest dams in the system since 1996. There has also been ongoing monitoring in the estuary using stationary and towed arrays since 1998. Many millions of dollars have been spent on tagging fish, installing and maintaining the system, developing the PTAGIS database, analyzing data, and communicating results, and major management decisions have rested on information gleaned from the detections of PIT-tagged salmon in the federal hydrosystem. Recent severe declines in detection counts put this important information source at risk. Here, we provide a retrospective analysis of 26 years of PIT tag release and detection data for Spring/Summer Chinook Salmon and Steelhead originating in the Snake River Basin, placing the current state of the PIT tag monitoring system in a broader historical context. Using data from hatchery and wild smolts, we estimated the joint probability of detection and survival to Lower Granite, Little Goose, Lower Monumental, Ice Harbor, McNary, John Day, and Bonneville dams and the Columbia River Estuary, characterized temporal patterns, and examined relationships to management and environmental conditions at each dam. Our analysis demonstrates that the probability of detecting juveniles at major dams has decreased across time and space in conjunction with structural and operational changes that discouraged use of powerhouse passage routes. The most precipitous drop in detections coincided with unprecedented increases in spill targets since 2020. Low detection counts within the bypass systems of many dams coincided with the addition of an impressive spillway detection system at Lower Granite Dam, which succeed in bolstering detections in the upper portion of the hydrosystem. However, these additional upstream detections have not fully compensated for the reduction in detections at downstream dams. The general decline in the frequency of detections and the lack of a compensatory increase in tag releases have reduced the quantity of data available for monitoring survival of outmigrating juveniles in the hydrosystem. The quality of survival estimates is diminished not only because fewer observations result in lower precision, but because the sparseness of some datasets can require pooling groups of tags and thus decrease the spatial or temporal resolution of the estimates and the utility of PIT tag data for guiding management decisions. The capacity to estimate survival of these focal populations in the reaches from the lower Snake River to Bonneville Dam has suffered in recent years and this issue likely extends to other populations in the Columbia basin (e.g., Mid and Upper Columbia Oncorhynchus spp). A return to the historical standard of juvenile survival monitoring while also implementing the current high spill regime will require augmenting the existing PIT tag detection system within the hydrosystem.