Estimates of salmon and steelhead smolt to adult survival and return are used in fisheries management and the assessment of hatcheries, habitat improvement projects, hydroelectric facility operations, and other activities. The estimates are also used to forecast future abundance and understand environmental processes affecting salmon. This report reviews the estimation of smolt-to-adult survival (SAS) and smolt-to-adult return (SAR) for Columbia River Basin salmon, with a focus on terminology, methodology, data inputs, and other attributes affecting their use. SAS and SAR are related but different metrics to represent salmon smolt survival to adulthood and return for spawning. SAS is the estimated proportion of smolts leaving some specified location that survive to adulthood and are either taken in ocean or freshwater fisheries, stray, or return to a designated location in the river system (e.g., a hatchery or stream). SAR is the estimated proportion of smolts leaving a specified location that return to that or another designated location on their return as adults. The distinction between SAR and SAS, and the terms return and survival, should be clearly defined when used. Notably, salmon taken in ocean fisheries are considered to have survived for the purposes of estimating SAS, but they have not returned to their designated location in the river and thus do not count towards the SAR estimate. The use of SAR to estimate survival relies on the assumption of little or no ocean fishery interceptions.
The overall goal of this report is to promote the clear and consistent use of SAR and SAS. To that end, the ISAB: 1) reviews how SAR and SAS are commonly estimated and used for Columbia River Basin salmon, 2) discusses some of the key assumptions and limitations in their use, 3) presents some of the complexities associated with the apparently simple terms “smolt, adult, return, survival” and 4) makes recommendations to help practitioners and readers best use and understand these metrics. The ISAB’s goal is neither to criticize past studies nor to discourage the use of these metrics. Rather, the ISAB seeks to heighten awareness of the variation underlying their estimation and the pitfalls related to their inconsistent or unclear application.
Estimating SAR and SAS requires designated locations where smolt and adult abundances are estimated. These locations are often different for the two estimators for a given stock and for different stocks, and the methods for estimating abundance depend on the mark and recapture methodology and sampling techniques. Consequently, for a single cohort of smolts, SAR and SAS estimates may be similar or very different, and the choice of which estimate to report, or to report both, can affect our understanding of a population’s trend or the effects of management. Moreover, directly comparing SAR or SAS estimates that are produced with different tagging methods (e.g., coded wire tags (CWTs), passive integrated transponder (PIT)-tags, and parentage-based tagging), from different locations, representing stocks of different origins and life histories, or different definitions of smolts and adults can introduce unintended biases and can lead to erroneous conclusions. These metrics are available for public use, so comparisons are commonly made within and beyond the Columbia River Basin. Consequently, well-defined terms, transparent methods, and consistent application are important for sound science and use in the management process.
This report does not advocate specific ways of estimating and interpreting salmon return and survival, but the ISAB provides examples of the use of SAR and SAS in the region that illustrate some of the issues the ISAB has identified. The ISAB also considers some of the complexities in estimating mortality from fishing and natural causes at sea. Apportioning mortality to different years and life stages has frustrated fisheries scientists for decades. Mortality during the early marine phase may determine the success of the cohort, but it is very difficult to estimate with any confidence. In the Columbia River Basin, this early marine phase has special importance because it is more plausibly linked to the delayed effects of smolt exposure to the hydrosystem during seaward migration than is mortality that occurs years later.
SAR and SAS estimates are essential to salmon conservation and management efforts in the Columbia River Basin and the broader science and management communities. Those who generate and use the estimates and manage data archives all contribute to their usability. The ISAB makes the following summary recommendations to improve the estimation, interpretation, documentation, and usability of SAR and SAS metrics in the Columbia River Basin:
Provide clear study objectives and describe the application for studies using SAR and SAS. Clearly define and consistently use the terms SAR, SAS, smolt, adult, return, and survival. Describe how SAR and SAS are estimated and how time-series data are analyzed. Report PIT-tag detections for SAR components (downstream, ocean, upriver) where applicable. Maintain the integrity of long-term SAR and SAS datasets by comparing results of different marking and analytical methods, developing robust conversions where appropriate, and reporting CWT-based SAS estimates for representative stocks throughout the basin. Augment SAR reporting in publicly accessible databases to include SAS. Where appropriate for the application, adjust SAR and SAS estimates to a common age at maturity and provide the rationale and methods for adjustments. Use SAR and SAS metrics from surrogate populations with caution and explain how well the surrogate represents the population of interest. 
