Historical escapement and run size of fall chum salmon Oncorhynchus keta was reconstructed from incomplete sonar, weir, counting tower, mark-recapture, aerial survey, and foot survey data of varying precision from 1974 to 2007. The resulting estimates of drainage-wide escapement were fitted to an age-structured Ricker spawner-recruit model. Bayesian statistical methods were employed, which allowed for realistic assessment of uncertainty in the presence of measurement error, serial correlation, and missing data.

The objectives of this report are to: 1) describe the productivity and capacity of the Kuskokwim River Chinook salmon stock, as quantified from stock-recruit analyses, 2) recommend a drainage-wide escapement goal based on this information, and 3) revise escapement goals for selected tributaries.

An age-structured state-space spawner–recruit model was fit to estimates of relative and absolute abundance, harvest, and age composition for Copper River Chinook salmon (Oncorhynchus tshawytscha) from 1980 to 2018. Bayesian statistical methods were employed to assess uncertainty in the presence of measurement error, serial correlation, and missing data. Ricker spawner–recruit parameters and management reference points were estimated, including the escapement that provides for maximum sustained yield (SMSY). An additional analysis was performed on a subset of data from 1999 to 2018 that used only mark–recapture estimates of escapement and excluded abundance indices used in the full data set. The full data set produced an SMSY of 22,844 spawners (95% credibility interval of 12,920–84,942 spawners) and the 1999–2018 data set produced an SMSY of 26,951 spawners (95% credibility interval of 15,371–98,262 spawners). Different results may be partially explained by differences in data structure and may also indicate reduced productivity in more recent years. It is important to note that many escapements observed since 1999 were greater than those observed prior to 1999 and the increased contrast in observed escapements increased information and our understanding of true underlying stock production. After examining both analyses, it is recommended that a sustainable escapement goal range of 21,000 to 31,000 fish be adopted for Copper River Chinook salmon. Escapement is evaluated by subtracting estimates of inriver harvest from estimates of inriver abundance. Escapements within this range have a high probability of producing sustainable yields.

Historical abundance of Andreafsky River, Anvik River, and aggregated other upriver stocks of summer chum salmon Oncorhynchus keta was reconstructed from mixed stock harvest data and incomplete sonar, weir, counting tower, and aerial survey data of varying precision from 1972 to 2007. The resulting estimates of escapement and harvest of East Fork Andreafsky River chum salmon were fitted to an age-structured Ricker spawner-recruit model. Bayesian statistical methods were employed, which allowed for realistic assessment of uncertainty in the presence of measurement error, serial correlation, and missing data.

After decades of using the DFO mark-recapture border program as the sole U.S./Canada border passage estimator oil the Yukon River, the cooperative ADF&G-NOAA drainage-wide radio telemetry project and ADF&G-DFO sonar projects provided independent estimates of border passage for 2002-2004 and 2005- 2007, respectively. Marked and variable differences between the annual DFO mark-recapture estimates and estimates derived from radio telemetry and sonar projects strongly suggests that the DFO mark- recapture estimate was biased low and a new sonar-based escapement goal was necessary. An exploratory analysis was initiated to reconstruct the historic drainage escapement database and make it compatible with new sonar-based escapement estimates. Analyses were conducted that ranged from simple scaling of historical data to more robust regression models. Seventeen escapement databases were created through these analyses. Ultimately, four escapement databases, reconstructed from drainage escapements derived from radio telemetry and sonar projects (2002-2007) and regressed against escapement indices, were considered viable candidates. A secondary goal of these analyses was to investigate potential escapement goal ranges based oil the viable escapement databases. Interestingly, all Sustainable Escapement Goal (SEG) ranges and Marcov Table analysis indicated similar escapement goals for all four reconstructed escapement database scenarios. Spawner-recruit statistics were also very similar. Therefore, two escapement goal recommendations were suggested: 1) Set a conservative escapement goal range that reflected all escapement database scenarios, 40,000 to 60,000 Chinook salmon, and 2) Do not select a reconstructed escapement database for the period 1982-2001, but build on the current escapement database derived from radio-telemetry estimates (2002-2004) and sonar counts (2005-present).