U.S. Fish & Wildlife Service

Abundance and Run Timing of Adult Pacific Salmon in the East Fork Andreafsky River, Yukon Delta National Wildlife Refuge, Alaska, 2018 Alaska Fisheries Data Series Number 2019-2

Fairbanks Fish and Wildlife Conservation Office Fairbanks, Alaska February 2019

The Alaska Region Fisheries Program of the U.S. Fish and Wildlife Service conducts fisheries monitoring and population assessment studies throughout many areas of Alaska. Dedicated professional staff located in Anchorage, Juneau, Fairbanks, and Kenai Fish and Wildlife Offices and the Anchorage Conservation Genetics Laboratory serve as the core of the Program’s fisheries management study efforts. Administrative and technical support is provided by staff in the Anchorage Regional Office. Our program works closely with the Alaska Department of Fish and Game and other partners to conserve and restore Alaska’s fish populations and aquatic habitats. Additional information about the Fisheries Program and work conducted by our field offices can be obtained at:

http://alaska.fws.gov/fisheries/index.htm

The Alaska Region Fisheries Program reports its study findings through the Alaska Fisheries Data Series (AFDS) or in recognized peer-reviewed journals. The AFDS was established to provide timely dissemination of data to fishery managers and other technically oriented professionals, for inclusion in agency databases, and to archive detailed study designs and results for the benefit of future investigations. Publication in the AFDS does not preclude further reporting of study results through recognized peer-reviewed journals.

Cover Photo: View upriver in the East Fork Andreafsky River valley, from a hill above the weir site, 2018. Photo courtesy of M. Larson, USFWS.

Disclaimer: The findings and conclusions in this article are those of the author(s) and do not necessarily represent the views of the U.S. Fish and Wildlife Service. The use of trade names of commercial products in this report does not constitute endorsement or recommendation for use by the federal government.

Alaska Fisheries Data Series Number 2019-2, February 2019 U.S. Fish and Wildlife Service

Chinook and Summer Chum Salmon Abundance, Run Timing, and Age, Sex, Length Composition in the East Fork Andreafsky River, Yukon Delta National Wildlife Refuge, Alaska, 2018 Jan M. Conitz

Abstract A weir was operated on the East Fork Andreafsky River from June 25 through July 29, 2018, to estimate salmon escapement and run timing. The primary target species were Chinook and summer Chum salmon, but all fish species passing the weir were enumerated. An underwater video system was used to record images of fish passing the weir 24 hours per day, every day of the operating period. Crew members counted fish, by species, from video recordings each day. For the days before and after the operation period, that fell within the historically known migration period, a Bayesian statistical procedure was used to estimate Chinook and summer Chum salmon passage. Chinook and summer Chum salmon were sampled to estimate age, sex, and length composition of the escapement. Chinook Salmon counted at the weir totaled 4,114 fish, and the total estimated escapement was 4,171 (95% Bayesian highest posterior density interval, HPDI, 4,116–4,358) Chinook salmon. Summer Chum Salmon counted at the weir totaled 36,330 fish, and the total estimated escapement was 38,250 (95% HPDI, 36,710–41,690). There were 4 age classes of Chinook Salmon present: age-1.1 (males only), -1.2, -1.3, and -1.4). The largest age-sex group of Chinook Salmon was age-1.3 males, representing about 51% of the run past the weir, followed by age-1.2 males (about 24%) and age-1.3 females (about 18%). There were also 4 age classes of summer Chum Salmon in 2018: age-0.2 (males only), -0.3, -0.4, and -0.5. The largest group of summer Chum salmon was age-0.3 males (about 31%), followed by age-0.3 females (about 29%) and age-0.4 males (about 22%). The same brood years, 2012–2015, were represented in both species sampled at the weir. Incidental counts of other species included 96,349 Pink Salmon O. gorbuscha, 1,198 Sockeye Salmon O. nerka, 12 Coho Salmon O. kisutch, and 2,270 whitefish (Coregoninae). This was the 25th consecutive year of weir operation in this lower Yukon River tributary.

Introduction

The Andreafsky River is a significant salmon producing tributary of the lower Yukon River, and supports some of the largest runs of Chinook Oncorhynchus tshawytscha and summer Chum O. keta Salmon in the Yukon River drainage. Located within the Yukon Delta National Wildlife Refuge, the Andreafsky River is designated a wild and scenic river, one of seven wild and scenic rivers in the National Wildlife Refuge system in Alaska.1 The Andreafsky River is also the lowest major tributary in the Yukon River drainage, entering the Yukon River main stem about 160 river kilometers (rkm) from the Bering Sea. The Yukon River extends over 3,190 rkm, from icefields in British Columbia, Canada, near the Gulf of Alaska through the Yukon Territory and Alaska to its mouth on the Bering Sea coast. The escapement monitoring project on the East Fork of the Andreafsky River is a key point in a drainage wide network of salmon monitoring and assessment projects, representing lower river stocks and providing information on salmon abundance, run timing, and stock composition. Escapement estimates from the East Fork

1 https://www.fws.gov/alaska/water/wildrivers.htm Author: Jan Conitz is a fisheries biologist with the U.S. Fish and Wildlife Service, and can be contacted at the Fairbanks Fish and Wildlife Conservation Office, 101 12th Ave., Room 222, Fairbanks, Alaska 99701; or [email protected]. Alaska Fisheries Data Series Number 2019-2, February 2019 U.S. Fish and Wildlife Service

Andreafsky River weir are combined with aerial survey estimates from both the East and West Forks and used by fisheries managers as in-season and post-season indices of lower river escapement.

Subsistence fisheries in the U.S. portion of the Yukon River drainage are jointly managed by the State of Alaska, under the Alaska Department of Fish and Game (ADF&G), and the U.S. Fish and Wildlife Service (USFWS) under the Federal Subsistence Management program. Commercial, sport, and personal use fisheries are managed by ADF&G. Within the National Wildlife Refuge, federal managers are responsible for conserving fish and habitats in their natural diversity, fulfilling international treaty obligations, supporting subsistence resource use by local residents, and protecting water quality.2 Both state and federal managers are responsible for ensuring priority of subsistence fishing use, protecting stock diversity, ensuring equitable distribution of harvest opportunity across the drainage, and meeting Canadian border passage objectives as specified by the Yukon River Salmon Agreement. In addition, ADF&G establishes escapement goals for various salmon populations in Yukon River tributaries. Among these are East Fork Andreafsky River Chinook and summer Chum salmon (Volk et al. 2009; Fleischman and Evenson 2010).

Salmon returning to the Andreafsky River contribute substantially to subsistence and commercial harvests in the lower Yukon River. Residents of some 20 communities, small villages, and fish camps located on the Yukon Delta and Yukon River below the Andreafsky River confluence harvest salmon from mixed stocks entering the Yukon River. Most residents harvest fish for subsistence use, but many also participate in small-scale commercial fisheries. Income from commercial fishing enables families to purchase the necessary gear and supplies to sustain their subsistence fishing activities. Chinook Salmon are the primary target for subsistence fisheries in the lower river, but summer and fall Chum Salmon and Coho Salmon (O. kisutch) are also harvested. Commercial fisheries harvest summer Chum Salmon during the early season, and shift to fall Chum and Coho salmon after mid-July, in the lower river.3

The Andreafsky River has been an important indicator for lower Yukon salmon runs for many years. Aerial surveys were conducted on both the East and West forks in most years since 1954.4 ADF&G experimented with sonar on the East Fork in the early 1980s (Buklis 1982) and operated a fish counting tower there from 1986 to 1988. The East Fork Andreafsky River weir in its current configuration has been operated annually by USFWS since 1994; and this year was the 25th consecutive year of operation. The escapement information from East Fork Andreafsky River is useful for in-season management because it generally lags run timing through the lower river fisheries by only a few days.

Most subsistence and commercial fishing in the lower Yukon River occurs in the main stem, harvesting from mixed stocks of salmon returning to spawning areas throughout the drainage. Although individual stocks cannot be specifically targeted or avoided, in-season genetic information and daily updates from escapement counting projects are monitored closely to ensure that certain stocks are not disproportionately harvested. Genetic mixed-stock analysis (MSA) compares genotypes of harvested fish with a baseline of genotypes from known

2 https://www.fws.gov/refuge/Yukon_Delta/what_we_do/resource_management.html 3 The commercial sale of Chinook salmon has been prohibited since 2012 (Estensen et al. 2017). 4 http://www.adfg.alaska.gov/CommFishR3/Website/AYKDBMSWebsite/Default.aspx

2 Alaska Fisheries Data Series Number 2019-2, February 2019 U.S. Fish and Wildlife Service

spawning populations. Genetic baselines have been developed for both Chinook5 and Chum salmon (Flannery et al. 2007) in the Yukon River. In both species, genetically distinct lower river stocks cover a broad geographic area, encompassing spawning populations in tributaries and the main stem from the Yukon River Delta up to the Gisasa River in the lower Koyukuk River drainage. All populations within the lower Yukon River Chum Salmon stock are genetically identifiable as summer Chum Salmon. Current genetic baselines cannot distinguish between individual Chinook or summer Chum salmon spawning populations within the lower Yukon River stock. Therefore, protection of these individual, geographically distinct populations still requires escapement monitoring projects such as the East Fork Andreafsky River weir. Furthermore, most MSA samples are obtained from test fisheries at the Pilot Station sonar project, with the primary purpose of helping managers ensure that border passage goals are met. Samples from this location, upstream of the Andreafsky River confluence, would exclude those portions of the lower river stock that entered spawning tributaries or were harvested downriver of the sonar site. Hence, the East Fork Andreafsky weir project remains a critical indicator of the effect of fisheries on lower Yukon River stocks.

Similar to previous years, daily migration counts of all fish species, and sex, length, and scale samples from Chinook and summer Chum salmon were collected at the East Fork Andreafsky River weir in 2018. The video system incorporated into the primary fish passage chute in 2014 continues to be used for counting fish passage through the weir. Environmental conditions were also monitored daily. Preparations were made at the end of the season to begin rebuilding the floating weir panels in the spring of 2019.

Objectives The following objectives were established for the 2018 season. 1. Estimate daily and seasonal escapement and run timing of adult Chinook and summer Chum salmon.

2. Estimate the age, sex, and length composition of the adult Chinook and summer Chum salmon escapements, with 95% confidence intervals for age-sex proportions no larger than ± 0.1.

3. Identify and count other fish species passing through the weir daily, and collect water level and temperature data twice daily.

Study Area The project was located on the East Fork of the Andreafsky River, about 40 rkm upriver from the village of St. Mary’s (Figure 1). St. Mary’s is on the main-stem Andreafsky River, near its confluence with the Yukon River. The weir and campsite are on Nerklikmute Native Corporation land and leased to the U.S. Fish and Wildlife Service for the project.

The Andreafsky River is in the lower reaches of the Yukon River drainage, flowing into the main-stem Yukon River about 160 rkm upriver from the Bering Sea. Its headwaters in the Nulato Hills give rise to two roughly parallel tributaries (East and West forks) which flow in a southwesterly direction for over 200 rkm before converging about 7 rkm upstream of the

5 http://www.adfg.alaska.gov/index.cfm?adfg=fishinggeneconservationlab.yukonchinook_baseline

3 Alaska Fisheries Data Series Number 2019-2, February 2019 U.S. Fish and Wildlife Service

confluence with the Yukon River. The total watershed area is approximately 5,450 km2. The Andreafsky River support runs of Chinook, summer Chum, Pink O. gorbuscha, Sockeye O. nerka, and Coho salmon, and is one of the largest salmon-producing tributaries in the lower Yukon River. Almost all salmon fishing on these stocks occurs below the confluence of the East and West forks and in the Yukon River. Other anadromous fish present in the Andreafsky River include Humpback Coregonus clupeaformis, Broad C. nasus, and Round Prosopium cylindraceum whitefish; Sheefish Stenodus leucichthys; Least C. Sardinella and Bering C. Laurettae cisco; and Dolly Varden Salvelinus malma. Resident freshwater species include Arctic Grayling Thymallus arcticus, Northern Pike Esox lucius, Burbot Lota lota, and Longnose Sucker Catostomus catostomus.

The East Fork Andreafsky River flows from its source for approximately 50 rkm through alpine tundra, and then for approximately 130 rkm through a forested river valley bordered by hills mostly less than 400 m in elevation, with willow, spruce, alder, and birch dominating the riparian zone and much of the hillsides. The streambed in this section is characterized by glides and riffles with gravel and rubble substrate. The river widens in its lowermost portion, flowing through wetlands, interspersed with forest and tundra, and bordered by hills that are typically less than 230 m elevation. In this low gradient section of the river, water levels can be affected by fluctuations on the Yukon River.

The regional climate is subarctic. Temperature records from St. Mary’s for the most recent 10- year period (2009–2018) show a maximum temperature of 27 ˚C (June 19, 2015) and a minimum temperature of -38˚C (January 3–4, 2012). The average winter temperature for this period was - 11.2 ˚C and the average summer temperature was 10.9 ˚C.6 Recent precipitation records from St. Mary’s are too sporadic to be useful. Normal annual precipitation records for 1981–2010 from Bethel, the nearest city on the Yukon-Kuskokwim Delta with long-term climate records, show annual rainfall was approximately 47 cm, and annual snowfall was approximately 137 cm.7 The month of heaviest normal rainfall was August and the month of heaviest normal snowfall was November for this period.

The median ice breakup date on the main-stem Yukon River at Mountain Village, 28 rkm downriver from the Andreafsky River confluence, was May 17, during the period 1980–2018. The earliest breakup date during this period was April 29, 2016; and the latest breakup date was June 1, 2013.8 Freeze-up dates, from sporadic records between 1949 and 2001, were typically mid-October, for first ice, through early November, for safe travel on foot.9

Methods Weir Design and Operation The weir structure is a modified resistance board design (Tobin 1994; Tobin and Harper 1995) with picket spacing 4.8 cm from inside edge to inside edge. This spacing allows smaller Pink Salmon and resident fish to pass through the weir undetected, but the slightly wider spacing

6 National Oceanic and Atmospheric Administration, National Centers for Environmental Information. Climate Data Online, https://www.ncdc.noaa.gov/cdo-web/datasets#LCD, accessed 020518. 730-year official Climate Normals period (1981-2010); NOAA, National Centers for Environmental Information, https://www1.ncdc.noaa.gov/pub/data/ccd-data/avgsnf15.dat. 8 National Weather Service, https://www.weather.gov/aprfc/breakupDB. 9 National Weather Service, https://www.weather.gov/aprfc/freezeUp.

4 Alaska Fisheries Data Series Number 2019-2, February 2019 U.S. Fish and Wildlife Service

allows better passage of water. The weir site coordinates are N 62o 07', W 162o 48.4', and the channel width is 105 m at that point.10 The fish passage chute was located at the deepest part of the channel, approximately 30 m from the left bank, and it connected to a sampling trap, which was closed as needed to conduct biological sampling.

A video system was used to capture images of all fish passing through the main passage chute, and recorded images were used for species identification and counting. A camera box attached to the side of the passage chute contained a CAM-AM070 Color Analog Video Camera (Applied MicroVideo) connected by a cable to a computer system in the main cabin. Motion capture (Security Spy) and automatic video recording software were used to preserve images of all fish passing through the chute. Installation of the weir and video system was completed on June 24, and normal operations began on June 25 and ended on July 29, 2018.

The weir and video system were operated 24 hours a day, 7 days a week. The weir was inspected each day and was cleared of debris at least once every 3–4 hours. Minor repairs were also made to the weir as needed to ensure that the weir remained structurally sound and provided a fish-tight barrier. The video system was set to record a motion capture video file every hour. These files were replayed during or after the end of each crew shift, and all fish that passed through the video chute were counted by species. Counts were entered from a computer keypad into a Visual Basic program, which automatically compiled a text file of fish counts for each hour and daily summary in Excel of fish counts by species. Water depth and temperature were manually measured and recorded twice daily at approximately 0800 hours and 2000 hours. A YSI 6920 Sonde (Yellow Springs, Ohio; http://www.ysi.com) was also set to record hourly average water temperatures through the season, and these data were uploaded to a computer file post-season. Each morning, a summary of the previous day’s data was reported by phone to the Fairbanks Fish and Wildlife Conservation Office, where the data were added into a daily update report distributed to managers, biologists, and stakeholders.

Weir operation typically begins each season on June 16 and ends July 30. However, water conditions on the river and weather can delay or otherwise affect the dates of operation. At the end of each season, guidelines are used to ensure that most target species have passed the weir before ending operation. The migrations of Chinook and summer Chum salmon are assumed to be nearly complete when daily passage counts are less than 1% of the season total passage for at least 3 consecutive days.

Estimates for Missed Weir Passage Days and Total Escapement

For days within the typical migration period for Chinook and summer Chum salmon, but outside the actual 2018 operational period for the weir, a statistical arrival model was used to estimate the total number of fish in the escapement (Sethi and Bradley 2016). This procedure would also be used to estimate fish passage on days when fish were able to escape through the weir uncounted due to high water or other problems. However, if the weir was functional and fish were counted through the passage chute for at least 6 hours in a given day, only the count and not the estimate was reported. The model could potentially include low probability events of fish passage on days far outside the typical migration period. To constrain the model to a reasonable

10 This is the same location and configuration used since 1995 (Tobin and Harper 1996). In 1994, the first year of operation, the East Fork Andreafsky weir was located approximately 2.4 rkm upstream of its 1995-2017 site.

5 Alaska Fisheries Data Series Number 2019-2, February 2019 U.S. Fish and Wildlife Service range of dates, the migration period was assumed to include the range of dates historically observed for Chinook and summer Chum salmon passage at the East Fork Andreafsky River weir (1994–2017). The range used in 2018 was June 15 through August 10.

The arrival model was fit to the weir data to estimate passage for days when no fish counts were made at the weir. To model arrival of fish at the weir, three possible distributions were considered: Normal, skew-Normal, and Student-t, with two degrees of freedom. Variability around this smooth arrival curve was modeled by Negative Binomial random variables, such that the magnitude of the variability scaled positively with the estimated daily passage counts. Three correlation structures were included for process variation in the Negative Binomial probability model: none (white noise), lag-1 autoregressive, and lag-1 moving average.

All combinations of run shape and process variation structure were fit to the data, resulting in a candidate set of nine potential models. Model fits were made in a Bayesian framework, allowing predicted estimates for each missing passage date and the associated variability to be easily obtained. Prior distributions were specified (Table 1 in Sethi and Bradley 2016) and models were run in WinBUGS (Lunn et al. 2000) using R2WinBUGS (Sturtz et al. 2005). Two chains were run with a burn-in period of 1,000 iterations and a thin rate of 10, resulting in 2,000 posterior draws stored. For each model, convergence was ensured by confirming that the Gelman-Brooks-Rubin statistic for the estimated total population abundance was less than 1.2. Deviance information criteria were used to select the best model from the candidate set of nine. Predicted passage from the initial and concluding tails of the run was summed separately and reported as derived parameters. For days within the weir operation period, only the observed counts were reported for days during which the weir was fully operational for at least 6 hours. If counting was conducted for less than 6 hours, estimated passage was reported. The total escapement estimate was the sum of fish counted during the weir operation period and arrival model estimates for the tails of the run and any days with missed passage estimates. A 95% highest posterior density interval (HPDI), which is roughly a Bayesian equivalent of a 95% confidence interval, was formed by adding 2.5 and 97.5 percentiles of the posterior distribution to the total counted fish.

Biological Sampling and Data Analysis Sex and length data and scale samples were collected from Chinook and summer Chum salmon in a trap integrated into the main passage chute. To avoid temperature stress, sampling and fish handling were avoided when water temperatures remained above 17°C for three consecutive days or any time temperatures reached or exceeded 20°C (von Biela et al. 2018). The sample size goal for each species was 220–240 fish for the season. This goal was based on a statistical calculation indicating a minimum desirable sample size of 180 fish with readable scales (Bromaghin 1993), plus an allowance of 25% additional samples for unreadable scales and other sampling errors. To ensure sampling was distributed throughout the run roughly in proportion to escapement, the season was divided into 4 periods having approximately equal fish passage numbers based on historical fish passage counts. The periods designated in 2018 were June 17– July 6, July 7–10, July 11–15, and July 16–31. The sample target was 55–60 fish of each species within each period. Daily targets were set to ensure consistent sampling effort while ensuring the target for each period was met. Salmon were sampled opportunistically during a day until the daily target was met. Fish length was measured from mid-eye to fork of tail, and sex was visually determined from external characteristics such as kype development or the presence of an

6 Alaska Fisheries Data Series Number 2019-2, February 2019 U.S. Fish and Wildlife Service

ovipositor. For age determination, three scales from each Chinook Salmon and one scale from each summer Chum Salmon were collected from the preferred sampling area (INPFC 1963). Scales were mounted on gum cards matched with length and sex data from each fish and sent to the Stock Biology group at ADF&G Division of Commercial Fisheries in Anchorage for age analysis (Eaton 2015).

During data analysis, the contribution of samples from each period was weighted by actual escapement during the same period to avoid bias due to over- or under-representation of samples in any period. Weighted estimates from all four periods were then pooled for seasonal estimates of age and sex composition.

Means and standard errors were calculated and reported for fish lengths within each age and sex class. Lengths within each age-sex class were assumed not to vary substantially over the seasonal progression of the run, so length calculations were made on pooled, unweighted samples within each age-sex class.

Age and sex composition was calculated as population proportions based on the age class and sex of the sampled fish. In each sampling period i, a number of fish ni were sampled, and of these a number nij were determined to be of class j (by sex, or combined sexes). The proportion pij of fish of class j in sampling period i was estimated as: = . 𝑛𝑛𝑖𝑖𝑖𝑖

𝑝𝑝�𝚤𝚤𝚤𝚤 𝑛𝑛𝑖𝑖 The variance of the proportion of class j in period i was: = (1 ). 𝑛𝑛𝑖𝑖

𝑉𝑉𝑉𝑉𝑉𝑉� �𝑝𝑝�𝚤𝚤𝚤𝚤� 𝑛𝑛𝑖𝑖−1 𝑝𝑝�𝚤𝚤𝚤𝚤 − 𝑝𝑝�𝚤𝚤𝚤𝚤 Sample proportions were weighted by the proportion Ni/N of the total population available for sampling within period i (i.e., proportion of the total seasonal escapement counted during period i). Weighted sample proportions were summed across all K periods to provide a pooled𝑁𝑁 seasonal estimate of the proportion of class j, as: = . 𝐾𝐾 𝑁𝑁𝑖𝑖 𝚥𝚥 𝑖𝑖= 1 𝚤𝚤𝚤𝚤 The variance of each seasonal age-sex class𝑝𝑝� proportion∑ 𝑁𝑁 𝑝𝑝� was estimated as: ( ) = 1 . 𝑛𝑛𝑖𝑖 𝑁𝑁𝑖𝑖 2 𝑝𝑝�𝚤𝚤𝚤𝚤 1− 𝑝𝑝�𝚤𝚤𝚤𝚤 𝐾𝐾 𝑉𝑉𝑉𝑉𝑉𝑉� �𝑝𝑝�𝚥𝚥� ∑𝑖𝑖=1 � − 𝑁𝑁𝑖𝑖� � 𝑁𝑁 � 𝑛𝑛𝑖𝑖−1 Using the above equation for variance and assuming asymptotic normality, 95% confidence intervals (CI) were constructed as: ± 1.96 .

𝑝𝑝�𝚥𝚥 �𝑉𝑉𝑉𝑉𝑉𝑉� �𝑝𝑝�𝚥𝚥� Other Species Counts and Environmental Data

Counts of other salmon species and non-salmon fish species were recorded in the video system. Daily passage counts of these other fish species were documented and reported in-season, but no corrections were made for any missed passage before, after, or during the weir operation period. Because the counts for non-target species may not have encompassed the entire migration period

7 Alaska Fisheries Data Series Number 2019-2, February 2019 U.S. Fish and Wildlife Service

of each species, they should not be considered reliable indicators of the species’ run timing, migration behavior, escapement, or stock status. However, they do provide information about presence, absence, and co-migration with the target species.

Water temperature and depth data were collected at the location of the fish passage chute upstream of the weir. A staff gauge was installed next to the sampling trap to measure daily water levels. The staff gauge was not calibrated to a fixed benchmark. Thus, it only provided a relative measure of water depth during the season. Water temperature was collected continuously using the YSI 6920 Sonde which was set to record hourly average temperatures. Data from the sonde were not uploaded on a daily basis, but water temperature was measured with a handheld analog thermometer twice daily at the same location to report with other in- season data. The sonde records were uploaded to a computer file at the end of the season, and summarized as daily average and maximum temperatures.

Results Weir Operation Placement of the weir in 2018 was completed on June 24, following a delay of about one week due to high water. Normal weir operations began just after midnight on June 25 and continued without interruption through the end date of July 29 (Table 1).

Water depth at the weir fluctuated slightly between 0.88 m and 1.02 m during the 2018 season except for one low reading of 0.74 m on July 1. Aside from the one lower reading, the lowest sustained water depths (0.88–0.89 m) occurred during the last 8 days of the season (July 22–29), and coincided with high water temperatures.

Daily average water temperatures at the weir ranged from 12.2ºC to 19.5ºC during weir operations (Table 2). Daily maximum temperatures were about 0.5ºC to 2.5ºC higher than daily average temperatures. Maximum and daily average temperatures approached guideline thresholds for safe fish handling (von Biela et al. 2018) in early July and exceeded these thresholds during part of the last week of weir operation in the 2018 season (Figure 2). The earlier warm water period began on June 30 and persisted until July 7. Maximum daily temperatures between 17ºC and 20ºC were recorded throughout this period, but the daily average temperature only exceeded the 17ºC threshold on 2 days, July 6–7. In late July, daily average water temperatures exceeded 17ºC for 6 consecutive days, July 24–29, and daily maximum temperatures reached or exceeded 20ºC on 3 days, July 25–27. During the earlier warm water period, air temperatures fluctuated diurnally and cooled in the early morning, but in late July, air temperatures remained high throughout each 24-hour day (Table 2).

Chinook and Summer Chum Salmon Escapement and Age-Sex-Length Composition

During 2018 weir operations 4,114 Chinook Salmon and 36,330 summer Chum Salmon were counted (Table 1). Both species were present in low numbers when weir operation began on June 25 and passage increased rapidly after June 27. The highest counts of Chinook and summer Chum salmon occurred on July 6 and 7; the two-day totals were 1,481 Chinook (about 36% of total run) and 6,412 summer Chum salmon (about 17% of total run).

8 Alaska Fisheries Data Series Number 2019-2, February 2019 U.S. Fish and Wildlife Service

Arrival model estimates for escapement before the beginning (June 15–24) and after the end of weir operation (July 30–August 10) added approximately 57 Chinook Salmon and 1,920 summer Chum Salmon to the passage counts from the weir (Table 1). Because the weir remained fish tight throughout weir operations in 2018, no missed passage estimates were needed. Combining counts and estimates, the total escapement estimates were 4,170 Chinook Salmon (95% HPDI11 4,116–4,358 fish) and 38,234 summer Chum Salmon (95% HPDI 36,710–41,690 fish; Table 1).

A total of 242 Chinook Salmon were sampled for age, sex, and length, close to the target range of 220–240 fish. Sampling errors including unreadable scales occurred in 13 of the fish sampled, reducing the effective sample size to 229 fish. Samples over-represented Chinook Salmon passage during the first and third sampling periods, and under-represented passage during the second and fourth periods (Table 3). The largest estimated age-sex components of the escapement were age-1.3 males (about 51%), age-1.2 males (about 24%), and age-1.3 females (about 18%). Only 3 age-1.4 fish (one male, two females) were sampled; these older fish represented only 1% of the 2018 estimated escapement. Confidence intervals for all estimated Chinook Salmon age and sex proportions were large (Table 4). Uneven sampling proportions across the 4 sampling periods, especially the small sample sizes in the second and fourth periods, likely contributed substantially to this uncertainty.

Most Chinook Salmon (98%) in the 2018 escapement were age-1.2 and -1.3 fish. Average length of the age-1.2 fish was 562 mm (SE ± 6 mm); males in this age class were more abundant but smaller (average 556 ± 6 mm) than females (average 602 ± 11 mm; Table 5). Average length of the age-1.3 fish was 674 mm (SE ± 5 mm); males in this age class were also more abundant but smaller (average 666 ± 5 mm) than females (average 693 ± 8 mm; Table 5).

For summer Chum Salmon age, sex, and length estimates, a total of 244 fish were sampled and 224 had readable scales, meeting the target range of 220–240 fish. Chum Salmon sampling was unevenly distributed over the 4 sampling periods, but was more evenly distributed than Chinook salmon sampling, with the exception of the third period (Table 6). The largest components of the summer Chum Salmon escapement were age-0.3 males (about 31%) and females (about 29%), and age 0.4 males (about 22%; Table 7). As with Chinook Salmon, confidence intervals for the summer Chum Salmon age and sex classes were wide, likely due in part to under-sampling during the third period.

Within the dominant age-0.3 and -0.4 classes, average lengths were 535 mm (SE ± 3 mm) for age-0.3 summer Chum Salmon and 567 mm (SE ± 5 mm) for age-0.4 Chum Salmon. Age-0.3 males were on average about 30 mm longer than age-0.3 females, and age-0.4 males were on average about 60 mm longer than age-0.4 females (Table 8).

Other Species Counts

The peak daily passage of Pink Salmon in 2018 occurred on July 21 when 11,697 fish were counted. High Pink Salmon passage overall occurred from July 17 through the last day of weir operations on July 29, when over 6,500 fish were counted. The total Pink Salmon count from June 25 through July 29 was 96,349 fish, but because of the high passage counts at the end of operations, this should be considered a partial escapement count.

11 Highest posterior density interval.

9 Alaska Fisheries Data Series Number 2019-2, February 2019 U.S. Fish and Wildlife Service

Steady numbers of Sockeye Salmon were counted through most of the season from June 28 through July 29; the season total passage was 1,198 fish. Only 12 Coho Salmon, which migrate later than Chinook and summer Chum salmon, were counted during 2018 weir operations. The total count of all whitefish (Coregoninae) was 2,270 fish, of which, most were Humpback Whitefish (2,219 fish), 49 were Broad Whitefish; and 2 were Round Whitefish. In addition, 2 Northern Pike were counted, but no Dolly Varden, Arctic Grayling, Burbot, or Longnose Sucker in 2018 (Table 1).

Discussion and Recommendations The 2018 Chinook Salmon escapement estimate of 4,171 fish is close to the historical median escapement of 4,157 (1994–2018; Table 9) and near the upper bound of the escapement goal range of 2,100–4,900. These 2018 escapement estimates are the median of the Bayesian posterior density interval, but to account for uncertainty in the estimates, they are evaluated on the basis of the 95% highest posterior density interval (HPDI). The HPDI for summer Chum Salmon escapement was 36,710–41,690 fish and the escapement goal (>40,000 fish) falls within this range. Nevertheless, the summer Chum Salmon escapement estimate of 38,250 fish is substantially below the historical median escapement of 57,959 (1994–2018; Table 9).

Achieving, but not exceeding, target sample sizes in each sampling period was challenging in 2018, in part because of run timing, high water temperatures, and various operational issues. For example, the 2-day spike in daily passage July 6–7 and unexpectedly strong passage of both species in late July likely contributed to failure to keep sample sizes within target ranges. Normal variation in run timing can easily confound attempts to sample in proportion to fish passage. Disproportionate sampling within some sampling periods, in comparison with fish passage numbers, likely contributed to failure of some age-sex proportion estimates to meet precision objectives. Reducing the number of sampling periods from 4 to 2 or 3 would help ensure sample sizes were large enough in each period to meet precision objectives while still addressing the need to distribute sampling effort in proportion to run timing.

A significant problem in biological sampling at the East Fork Andreafsky River weir site is the increasing frequency of high water temperatures. Sampling was suspended as mandated by current protocol when water temperatures exceeded specified thresholds (von Biela et al. 2018). The guidelines specify sampling should cease when water temperatures remain above 17°C for 3 consecutive days. However, the guidelines do not specify the intervals when temperature should be measured. The sampling crew was instructed to measure water temperature twice daily and assess temperature thresholds based on those measurements. The sonde measured temperature continuously and recorded measurements every hour, but the data were not available to the crew in-season. The hourly temperature data from the sonde, examined post-season, indicated temperatures reached or exceeded 17°C from several hours to 24 hours per day on more days than indicated by the twice-daily thermometer readings.12 However, daily average and maximum temperatures (average and maximum of the 24 hourly recorded temperatures per day; Table 2) confirmed that the crew’s assessment of temperature thresholds was approximately correct. Nevertheless, this comparison shows that the twice-daily hand measurements of water temperature are insufficient to confidently assess temperature thresholds. The sonde is

12 June 30–July 7 and July 22–31. Contact USFWS Subsistence Fisheries Branch staff, Fairbanks Fish and Wildlife Conservation Office, 101 12th Avenue, Room 222, Fairbanks AK, 99701 for detailed data.

10 Alaska Fisheries Data Series Number 2019-2, February 2019 U.S. Fish and Wildlife Service

cumbersome to deploy and has no direct data display or easy interface for uploading data. The other functions it provides are not needed for this project. Thus, it should be replaced with a smaller and simpler device that can measure and record temperature at specified intervals and from which data can be retrieved at least once per day. Post-season evaluation of temperature data also indicates that heat stress may also be better avoided by sampling during the morning hours, rather than in the late afternoon and evening when heat has been building in the river through the day.

Concerns about fish handling in warm water have been noted at the Andreafsky weir for the last several years, and are expected to continue or get worse due to climate warming and environmental change. The current temperature protocols for fish handling were based on research showing evidence, in protein markers, of stress effects for salmon migrating in moderate warm temperatures (18–20°C). Andreafsky River Chinook Salmon were sampled and showed evidence of stress via changes to proteins and gene transcription to a greater degree than salmon from other locations within the Yukon River drainage (von Biela et al. 2018). The value of the information provided by biological sampling needs to be weighed against actual and potential harm this activity can cause salmon in a warming environment. Furthermore, the quality of the information may be diminished when sampling schedules are constrained by the need to avoid fish handling during warm water conditions. As warm-water conditions become more frequent in the Andreafsky River, consideration should also be given to the added stress to migrating salmon caused by the weir itself. When large numbers of fish are present, they can build up downstream of the weir and their passage may be delayed. This is especially concerning when the water is low, which often coincides with the warmest days. In 2018, water temperatures were high enough to cause physiological stress in migrating salmon for 15 of 34 days, i.e., 44% of the days the weir was operated.

Acknowledgements

Funding for the East Fork Andreafsky River weir project in 2018 was provided by the Research and Management Fund under the Yukon River Salmon Agreement and by the National Fish and Wildlife Foundation. The USFWS Region 7 office provided additional funds for replacement of old weir materials, which will continue through the beginning of the 2019 season. Staff who contributed to this project includes subsistence branch chief and fishery manager Fred Bue, supervisory biologist Matt Keyse, biometrician Catherine Bradley, and crew members Matt Larson (crew leader), Claire Crawbuck, and Chris Horvath. Crew members Mykel Long and Trysten Aguchak were hired through the Yupiit of Andreafski Tribe, and we acknowledge administrative support for these employees by the Ilisagvik College Youth Engagement, Education, and Employment Program, in coordination with Neesha Stellrecht at FFWCO. Other USFWS staff, including Gerald Maschmann, Jeremy Carlson, and Sarah Conn provided additional assistance with weir installation, operation, and breakdown. Matt Larson and Amber Bohlman conducted post-season data quality review. Larry Dubois and Jim O’Rourke from ADF&G, Commercial Fisheries Division, provided training, salmon age analysis, and data quality control, archival, and publication within the ADF&G public-facing database system. Editorial review of this report was provided by Sarah Conn, Matt Keyse, Gerald Maschmann, and Mike Buntjer. The success of this project continues to be dependent on support from the people of St. Mary’s and we appreciate their continued involvement and interest in this project.

11 Alaska Fisheries Data Series Number 2019-2, February 2019 U.S. Fish and Wildlife Service References Bromaghin, J. F. 1993. Sample size determination for interval estimation of multinomial probabilities. The American Statistician 47(3):203-206. Buklis, L. 1982. Anvik, Andreafsky, and Tanana River salmon escapement studies. Alaska Department of Fish and Game, Anchorage. Eaton, S. M. 2015. Salmon age, sex, and length (ASL) sampling procedures for the Arctic- Yukon-Kuskokwim Region. Alaska Department of Fish and Game, Division of Commercial Fisheries, Regional Information Report 3A15-04, Anchorage. Estensen, J. L., S. N. Schmidt, S. Garcia, C. M. Gleason, B. M. Borba, D. M. Jallen, A. J. Padilla, and K. M. Hilton. 2017. Annual Management Report Yukon Area, 2015. Alaska Department of Fish and Game, Fishery Management Report No. 17-12, Anchorage. Flannery, B. G., T. D. Beacham, R. R. Holder, E. J. Kretschmer, and J. K. Wenburg. 2007. Stock structure and mixed-stock analysis of Yukon River Chum Salmon. Alaska Fisheries Technical Report Number 97, U.S. Fish and Wildlife Service, Anchorage. Fleischman, S. J. and D. Evenson. 2010. Run reconstruction, spawner-recruit analysis, and escapement goal recommendation for summer chum salmon in the East Fork of the Andreafsky River. Alaska Department of Fish and Game, Fishery Manuscript Series No. 10-04, Anchorage. INPFC (International North Pacific Fisheries Commission). 1963. Annual Report 1961. Vancouver, British Columbia. Lunn, D. J., A. Thomas, N. Best, and D. Spiegelhalter. 2000. WinBUGS —a Bayesian modelling framework: concepts, structure, and extensibility. Statistical Computing 10: 325– 337. Sethi, A. S. and C. Bradley. 2016. Statistical arrival models to estimate missed passage counts at fish weirs. Canadian Journal of Fisheries and Aquatic Sciences 73:1–10. Sturtz, S., U. Ligges, and A. Gelman. 2005. R2WinBUGS: a package for running WinBUGS from R. Journal of Statistical Software 12:1–16. Tobin, J. H. 1994. Construction and performance of a portable resistance board weir for counting migrating adult salmon in rivers. U.S. Fish and Wildlife Service, Kenai Fishery Resource Office, Alaska Fisheries Technical Report 22, Kenai, Alaska. Tobin, J. H., and K. C. Harper. 1995. Abundance and run timing of adult salmon in the East Fork Andreafsky River, Yukon Delta National Wildlife Refuge, Alaska, 1994. U.S. Fish and Wildlife Service, Kenai Fishery Resource Office, Alaska Fisheries Progress Report 95- 5, Kenai, Alaska. Tobin, J. H., and K. C. Harper. 1996. Abundance and run timing of adult salmon in the East Fork Andreafsky River, Yukon Delta National Wildlife Refuge, Alaska, 1995. U.S. Fish and Wildlife Service, Kenai Fishery Resource Office, Alaska Fisheries Progress Report 96- 1, Kenai, Alaska. Volk, E. C., M. J. Evenson, and R. H. Clark. 2009. Escapement goal recommendations for select Arctic-Yukon-Kuskokwim Region salmon stocks, 2010. Alaska Department of Fish and Game, Fishery Manuscript No. 09-07, Anchorage.

12

Alaska Fisheries Data Series Number 2019-2, February 2019 U.S. Fish and Wildlife Service von Biela, V. R., L. Bowen, S. D. McCormick, R. J. Brown, M. P. Carey, S. Larson, S. Zuray, S. Waters, A. M. Regish, and C. E. Zimmerman. 2018. Assessing heat stress in migrating Yukon River Chinook salmon. Abstract for presentation at the joint Alaska Chapter/Western Division American Fisheries Society meeting, May 21–25, 2018, Anchorage AK.

13 Alaska Fisheries Data Series Number 2019-2, February 2019 U.S. Fish and Wildlife Service

Table 1. — Daily and cumulative escapement estimates of Chinook, summer Chum, and Pink salmon, and daily estimates of other species, at the East Fork Andreafsky River weir, Alaska, 2018. Dolly Chinook Salmon Summer Chum Pink Salmon Sockeye Coho Whitefish Pike Grayling Varden Date Daily Cum Daily Cum Daily Cum Daily Daily Daily Daily Daily Daily

25-Jun 2 2 41 41 4 4 0 0 8 1 0 0 26-Jun 0 2 31 72 1 5 1 0 9 0 0 0 27-Jun 4 6 28 100 10 15 0 0 18 0 0 0 28-Jun 18 24 1,429 1,529 21 36 12 0 26 0 0 0 29-Jun 51 75 1,731 3,260 51 87 11 0 27 0 0 0 30-Jun 38 113 964 4,224 38 125 9 0 11 1 0 0 1-Jul 67 180 1,129 5,353 89 214 18 0 40 0 0 0 2-Jul 52 232 542 5,895 103 317 17 0 60 0 0 0 3-Jul 41 273 735 6,630 123 440 38 0 97 0 0 0 4-Jul 166 439 2,105 8,735 365 805 22 0 80 2 0 0 5-Jul 116 555 958 9,693 363 1,168 25 0 83 0 0 0 6-Jul 680 1,235 3,210 12,903 1,762 2,930 125 0 163 2 0 0 7-Jul 751 1,986 3,202 16,105 2,134 5,064 113 0 197 2 0 0 8-Jul 78 2,064 2,458 18,563 1,019 6,083 51 0 98 3 0 0 9-Jul 49 2,113 555 19,118 357 6,440 15 0 100 1 0 0 10-Jul 44 2,157 577 19,695 140 6,580 27 0 49 0 0 0 11-Jul 102 2,259 2,881 22,576 342 6,922 33 0 35 0 0 0 12-Jul 114 2,373 1,440 24,016 260 7,182 28 0 29 0 0 0 13-Jul 120 2,493 830 24,846 635 7,817 57 0 27 0 0 0 14-Jul 218 2,711 1,042 25,888 1,006 8,823 26 0 37 1 0 0 15-Jul 263 2,974 1,259 27,147 957 9,780 20 0 16 0 0 0 16-Jul 256 3,230 598 27,745 723 10,503 26 0 18 0 0 0 17-Jul 299 3,529 1,626 29,371 5,195 15,698 73 0 68 1 2 0 18-Jul 99 3,628 694 30,065 3,836 19,534 34 0 77 2 0 0 19-Jul 68 3,696 639 30,704 4,316 23,850 27 0 73 1 0 2 -continued-

12

Alaska Fisheries Data Series Number 2019-2, February 2019 U.S. Fish and Wildlife Service

Table 1. — Page 2 of 2. Dolly Chinook Salmon Summer Chum Pink Salmon Sockeye Coho Whitefish Pike Grayling Varden Date Daily Cum Daily Cum Daily Cum Daily Daily Daily Daily Daily Daily 20-Jul 48 3,744 686 31,390 4,225 28,075 22 0 68 2 0 0 21-Jul 156 3,900 1,002 32,392 11,697 39,772 33 0 112 2 0 0 22-Jul 70 3,970 998 33,390 8,485 48,257 33 0 112 2 0 0 23-Jul 49 4,019 501 33,891 4,765 53,022 26 0 87 3 0 0 24-Jul 19 4,038 578 34,469 6,423 59,445 25 1 83 0 0 1 25-Jul 22 4,060 600 35,069 9,092 68,537 57 0 96 1 0 0 26-Jul 15 4,075 338 35,407 7,005 75,542 38 0 68 2 0 0 27-Jul 11 4,086 236 35,643 8,600 84,142 45 1 59 1 0 2 28-Jul 8 4,094 304 35,947 5,622 89,764 53 4 67 0 0 0 29-Jul 20 4,114 383 36,330 6,585 96,349 58 6 72 3 0 0 Sum of daily 4,114 36,330 96,349 1,198 12 2,270 33 2 5 counts Est. 6/15-6/241 1 16 Est. 7/30-8/102 56 1,904 Total estimate 4,171 38,250 95% HPDI3 4,116 – 4,358 36,710 - 41,690 1Approximate numbers of Chinook and summer Chum salmon passing from starting date of 15 June, based on historical run timing, until first day of weir operation, 25 June. 2Approximate numbers of Chinook and summer Chum salmon passing after weir operations ended to end date of 10 August, based on historical run timing. 395% Highest posterior density estimate (Bayesian equivalent to confidence interval).

13 Alaska Fisheries Data Series Number 2019-2, February 2019 U.S. Fish and Wildlife Service

Table 2. — Daily water levels, turbidity, and water and air temperatures at the Andreafsky River weir, 2018.

Water temp (°C) Air temp (°C) Water level Turbidity Date (m)1 (visual) Daily average2 Maximum2 AM PM 26-Jun na na 15.7 17.0 8.9 15.3 27-Jun na clear 13.5 14.3 10.5 na 28-Jun 1.00 clear 14.2 15.6 10.2 18.4 29-Jun 1.00 clear 15.2 16.0 12.4 14.7 30-Jun 0.97 clear 15.7 17.1 15.0 18.8 1-Jul 0.74 clear 16.0 17.2 na 16.4 2-Jul 0.93 clear 15.4 16.7 10.0 15.4 3-Jul 0.93 clear 15.4 18.2 4.8 na 4-Jul 0.91 clear 17.0 18.0 13.3 15.2 5-Jul 0.91 clear 16.6 19.0 10.3 20.8 6-Jul 0.91 clear 17.9 20.0 11.1 24.5 7-Jul 0.91 clear 17.9 19.7 13.8 13.1 8-Jul 0.91 clear 14.8 16.7 9.1 10.2 9-Jul 0.95 semi turbid 13.3 14.2 9.5 11.6 10-Jul 1.02 semi turbid 13.2 14.3 9.3 14.8 11-Jul 0.99 semi turbid 14.1 15.2 9.9 na 12-Jul 0.99 semi turbid 14.0 14.7 8.6 na 13-Jul 0.92 semi turbid 13.4 14.2 9.9 13.7 14-Jul 0.91 clear 13.7 14.3 9.9 14.1 15-Jul 0.90 clear 13.4 14.1 9.7 na 16-Jul na na 12.2 12.7 na na 17-Jul na na 12.8 13.4 11.6 13.8 18-Jul 0.90 clear 12.6 13.3 na 11.9 19-Jul 0.90 clear 12.3 12.8 10.5 13.1 20-Jul 0.89 clear 13.3 15.8 6.1 19.8 21-Jul 0.90 clear 15.7 16.5 15.4 16.2 22-Jul 0.89 clear 16.4 17.8 13.7 15.5 23-Jul 0.89 clear 16.8 17.6 16.2 19.0 24-Jul 0.89 clear 17.1 18.4 19.0 23.6 25-Jul 0.89 clear 18.6 20.6 19.6 19.9 26-Jul 0.86 clear 19.5 20.5 16.1 18.8 27-Jul 0.88 clear 19.1 20.0 16.6 19.4 28-Jul 0.88 clear 18.4 19.6 16.7 16.6 29-Jul 0.88 clear 17.5 18.4 na 17.3 Minimum 0.7 12.2 12.7 4.8 10.2

Maximum 1.0 19.5 20.6 19.6 24.5

Average 0.9 15.4 16.6 11.9 16.5

1Measured from streambed to water surface at fish passage chute; not calibrated to shoreline elevation. 2From data logged hourly on sonde; bold font indicates temperatures above 17 and 20 °C stress thresholds for salmon.

14 Alaska Fisheries Data Series Number 2019-2, February 2019 U.S. Fish and Wildlife Service

Table 3. — Stratified sampling periods, and numbers of Chinook salmon sampled and counted at the East Fork Andreafsky River weir, 2018.

Number of fish: Proportion in each period: Sampling period Sampled1 Counted Sampled1 Counted June 25 - July 6 78 1,235 0.341 0.300 July 7 - 10 24 922 0.105 0.224 July 11 - 15 113 817 0.493 0.199 July 16 - 29 14 1,140 0.061 0.277 Totals 229 4,114 1Sample sizes are only those fish for which age could be determined.

Table 4. — Numbers sampled and estimated proportions of Chinook Salmon by age and sex in the East Fork Andreafsky River, 2018.

Estimated 95% Confidence interval bounds Number in proportion of Sex Age sample total passage1 Lower Upper Interval width By age and sex Male 1.1 2 0.011 0.002 0.020 0.018 Male 1.2 71 0.241 0.198 0.284 0.086 Male 1.3 98 0.507 0.437 0.577 0.140 Male 1.4 1 0.002 0 0.009 0.009

Female 1.2 12 0.050 0.034 0.065 0.031 Female 1.3 43 0.178 0.145 0.212 0.067 Female 1.4 2 0.011 0.002 0.020 0.018 By age

1.1 2 0.011 0.002 0.020 0.018 1.2 83 0.291 0.246 0.336 0.091

1.3 141 0.685 0.609 0.761 0.151

1.4 3 0.013 0.004 0.022 0.018

By sex

Male 172 0.761 0.683 0.840 0.157 Female 57 0.239 0.202 0.276 0.075 1Proportions estimated from samples grouped by four periods and weighted by fish passage during those periods.

15 Alaska Fisheries Data Series Number 2019-2, February 2019 U.S. Fish and Wildlife Service

Table 5. — Average lengths of Chinook Salmon by age and sex sampled at the East Fork Andreafsky River weir, 2018.

Age Sex Average length (mm) SE (avg length) Number of fish1 1.1 M 459 2 2 1.2 M 556 6 71 1.2 F 602 11 12 1.2 All 562 6 83 1.3 M 666 5 98 1.3 F 693 8 43 1.3 All 674 5 141 1.4 M 773 na 1 1.4 F 761 55 2 1.4 All 765 32 3 1Samples used for average length calculations include all fish that could be aged and were not weighted by run timing (based on the assumption that length within a given age-sex class does not change over the season).

Table 6. — Sample sizes and weighting factors used to estimate summer Chum Salmon age-sex composition in the East Fork Andreafsky River, 2018.

Number of fish: Proportion in each period: Sampling Period Sampled1 Counted Sampled1 Counted June 25 - July 6 67 12,903 0.299 0.355 July 7 - 10 53 6,792 0.237 0.187 July 11 - 15 9 7,452 0.040 0.205 July 16 - 29 95 9,183 0.424 0.253 Total 224 36,330 1Sample sizes are only those fish for which age could be determined.

16 Alaska Fisheries Data Series Number 2019-2, February 2019 U.S. Fish and Wildlife Service

Table 7. — Numbers sampled and estimated proportions of summer Chum Salmon by age and sex in the East Fork Andreafsky River, 2018.

Estimated 95% Confidence interval bounds Number in proportion of Sex Age sample total passage1 Lower Upper Interval width By age and sex Male 0.2 3 0.008 0.003 0.013 0.009 Male 0.3 61 0.308 0.256 0.359 0.103 Male 0.4 51 0.222 0.184 0.259 0.075 Male 0.5 3 0.011 0.005 0.016 0.010

Female 0.3 73 0.287 0.245 0.329 0.085 Female 0.4 31 0.159 0.121 0.196 0.075 Female 0.5 2 0.006 0.002 0.010 0.008 By age

0.2 3 0.008 0.003 0.013 0.009

0.3 134 0.595 0.531 0.659 0.128

0.4 82 0.380 0.329 0.432 0.103 0.5 5 0.017 0.010 0.023 0.013 By sex

Male 118 0.548 0.486 0.610 0.123 Female 106 0.452 0.397 0.507 0.110 1Proportions are estimates from samples grouped by the four sampling periods and weighted by fish passage during those periods.

17 Alaska Fisheries Data Series Number 2019-2, February 2019 U.S. Fish and Wildlife Service

Table 8. — Average lengths of summer Chum Salmon by age and sex sampled at the East Fork Andreafsky River weir, 2018.

Age Sex Average length (mm) SE (avg length) Number of fish1 0.2 M 515 9 3 0.3 M 552 4 61 0.3 F 521 3 73 0.3 All 535 3 134 0.4 M 590 4 51 0.4 F 528 6 31 0.4 All 567 5 82 0.5 M 569 2 3 0.5 F 510 17 2 0.5 All 545 15 5 1Samples used for average length calculations include all fish that could be aged, and were not weighted by run timing (based on the assumption that length within a given age-sex class does not change over the season).

18 Alaska Fisheries Data Series Number 2019-2, February 2019 U.S. Fish and Wildlife Service

Table 9. — Annual escapement estimates and confidence intervals for Chinook and summer Chum salmon in the East Fork Andreafsky River, 1994-2018.

Chinook Salmon Summer Chum Salmon 95% HPDI1 95% HPDI1 Year Estimate Lower Upper Estimate Lower Upper 1994 7,957 7,831 8,329 240,551 203,100 405,207 1995 5,842 5,841 5,850 172,427 172,200 172,900 1996 2,984 2,974 3,006 108,997 108,700 110,500 1997 3,188 3,186 3,200 51,162 51,150 51,180 1998 4,143 4,084 4,292 67,712 67,660 67,820 1999 3,459 3,413 3,567 32,740 32,660 32,790 2000 1,824 1,549 2,236 28,861 26,750 32,861 2001 NA2 - - NA2 - - 2002 4,124 4,123 4,131 45,621 44,830 47,340 2003 4,342 4,337 4,349 22,508 22,460 22,620 2004 8,377 8,195 8,693 70,154 66,370 76,421 2005 2,382 2,270 2,689 23,727 20,330 33,360 2006 7,813 6,541 12,632 140,623 103,100 278,310 2007 5,306 4,589 7,172 72,354 70,080 78,581 2008 4,300 4,242 4,613 58,788 57,510 63,870 2009 3,992 3,052 9,608 9,453 8,832 11,911 2010 3,239 2,508 5,117 74,915 73,490 77,871 2011 5,271 5,218 5,482 101,227 100,600 102,800 2012 4,359 2,614 15,146 83,901 59,970 139,412 2013 2,123 2,084 2,171 97,076 62,920 220,402 2014 5,983 5,951 6,099 39,253 37,950 44,730 2015 6,979 6,068 8,966 51,625 49,940 55,580 2016 2,759 2,686 2,990 52,030 51,050 53,730 2017 2,975 2,972 2,982 57,130 55,930 59,840 2018 4,171 4,116 4,538 38,250 36,710 41,690 1Highest posterior density interval. 2Weir operation was delayed until July 15 due to high water; and no estimates were reported.

19 Alaska Fisheries Data Series Number 2019-2, February 2019 U.S. Fish and Wildlife Service

Figure 1. — Project location on the East Fork Andreafsky River, Alaska, 1995-2016. Note: the 1994 site was 2.4 rkm upstream of the site shown on this map (Tobin and Harper 1996).

20 Alaska Fisheries Data Series Number 2019-2, February 2019 U.S. Fish and Wildlife Service

25.0

20.0

15.0

10.0

Avg Max 17°C 20°C

5.0

0.0 25-Jun 30-Jun 5-Jul 10-Jul 15-Jul 20-Jul 25-Jul 30-Jul

Figure 2. — Average and maximum daily water temperatures at the East Fork Andreafsky River weir, 2018.

21