Gold Creek Bull Trout Population

Overview

Gold Creek is one of several tributaries of Keechelus Reservoir, the uppermost storage reservoir in the Yakima Basin. Keechelus was a natural lake prior to the construction of Keechelus Dam on its outlet in 1917. The dam is the upstream terminus of the Yakima River, which continues 214 river miles downstream to the Columbia River. At full pool, the reservoir holds 158,000 acre-feet of water with a surface area of 2562 acres. The dam is a complete barrier to upstream migration isolating the only population of bull trout residing in Keechelus Reservoir, which spawns and rears in Gold Creek. Individuals from this population are regularly entrained through the unscreened outlet works of the dam; see the threat related to entrainment below.

Gold Creek originates in the Alpine Lakes Wilderness Area and flows into Keechelus Reservoir at its northern end. About 6.8 miles of Gold Creek is accessible to migratory fish up to a barrier waterfall. A bedrock cascade about a half-mile below this barrier may also impede upstream migration (Craig 1997). A little less than half of the stream’s length below the waterfall is in the wilderness. Once Gold Creek exits the wilderness, land use is a mix of National Forest, State and private. In this reach the channel dewaters annually between July and October, with the timing and extent largely dependent on snowpack and summer rainfall. The length of stream which dewaters varies, but the affected reach begins just above the outlet of Golf Creek Pond near RM 0.5 and has been observed to extend up to around RM 2.2 (Craig 1997, Abbe and Ericcson 2014). In recent drought years (2024-2025) dewatering has also been observed in the wilderness area upstream of the USFS trail 1314 creek crossing.

Population Information

Population Distribution and Life History

The Gold Creek population displays an adfluvial life history type. The spawning area extends from the barrier waterfall downstream to the lake. A major avalanche occurred sometime during the early spring of 2008 completely covering about a quarter-mile section of Gold Creek with large wood and rock debris, earth, snow, and ice. The affected reach did not turn out to be a passage barrier. Rearing juveniles are present throughout the length of the stream. Keechelus Reservoir provides FMO habitat for subadult and adult fish (Figure X - MAP). Some sub-adult bull trout have also been observed utilizing other tributaries for rearing, including Coal Creek, Resort Creek, Rocky Run Creek, and Townsend Creek (James 2025). These fish are assumed to be of Gold Creek origin. Timing of migration into Gold Creek is dependent on stream flows. Early studies on this population showed some adult fish begin to move into the stream in late July to mid-August; but in years when dewatering occurred, there was a subset of adults that migrated after the stream rewatered with fall rains (James 2002a). More recent PIT-tag studies indicate that adults and subadults enter lower Gold Creek starting in June (Beebe et al. 2025 a,b). Dewatering of river mile (RM) 0.5 - RM 2 is now an annual event and it appears to be disconnecting earlier in the season. While there are still some individuals that migrate to upstream reaches prior to dewatering, the majority of spawners migrate to upstream spawning grounds once fall rains reconnect the creek. Some individuals migrate into downstream reaches of the creek, encounter fish passage issues, and consequently spawn in the reach between the pond outlet channel and the reservoir.

Natural Barriers limiting distribution

Figure X. A picture of the lower bedrock cascade under a high flow scenario. A redd was found above this feature on this particular survey.

There is a partial fish passage barrier (bedrock cascade) around RM 6.5 (47.45074, -121.34149), and a full passage barrier at RM 6.8 (47.45409, -121.33688) (Figure X).

Population Genetics

Results of genetic analyses show the Gold Creek population is genetically distinct from all other populations in the Yakima Basin (Reiss 2003; Small et al. 2009). Initial genetic samples for the baseline were collected from juvenile bull trout during a snorkel survey conducted in 2001 (Reiss 2003). Additional samples were collected in 2010 by the USFWS. Upstream connectivity was eliminated by the construction of Keechelus Dam in 1917. Genetic exchange with other Yakima populations of bull trout may occur downstream of the dam due to entrainment, but it is undocumented.

Population Monitoring

The first official documentation of the presence of bull trout in Keechelus Lake comes from 1982 when WDFW captured five adults in gill nets (Mongillo 1982). As noted above, spawning surveys in Gold Creek were initiated two years later and consistent monitoring of the Gold Creek bull trout population began. Other than these spawning surveys, Gold Creek did not receive much attention until 1996 when CWU graduate student, Scott Craig, investigated habitat conditions affecting bull trout spawning areas in the creek (Craig 1997). To do so he used the redd count data collected during the annual spawning surveys.

In 1998, the USFS surveyed 725 meters for habitat and fish use (USFS 1998b). The short segments included one downstream of the pond outlet, one in the wilderness along the trail, and one upstream of the trail crossing. They found adult and juvenile trout, including bull trout, in all 3 segments.

CWU researcher Paul James studied the population status and life history characteristics of the Gold Creek population (James 2002a). To determine outmigration timing he attempted to trap post-spawn bull trout in the channel adjacent to Gold Creek Pond in 1999 without success. They did manage to trap three adults (and one juvenile) in 2000. That same year his crew conducted four daytime snorkel surveys to determine migration timing, observing 16 adults in the creek by the end of July. The stream became intermittently dewatered in late August, and he documented a few adults unable to migrate upstream of the Gold Creek Pond. Also in 2000, William Meyer conducted nine nighttime snorkel surveys in Gold Creek between 19 July and 9 November. He observed both adults and juvenile bull trout and reported juvenile densities lower than those observed in the upper Kachess River by 25-50% (Meyer 2002).

In 2001, snorkel surveys were conducted in Gold Creek by Yuki Reiss to capture juvenile bull trout and obtain genetic samples. Twenty samples were obtained, less than the number generally desired (30) to establish a genetic baseline Reiss (2003). The USFWS and WDFW returned to the creek in 2010 to collect additional genetic samples to supplement this baseline (See genetics baseline).

WDFW day and night snorkeled and electroshocked Gold Creek in 2003 as part of a project to develop a bull trout presence/absence sampling protocol (Hoffman et al. 2005). Larsen et al. (2003) examined these data in more detail with Peterson et al. (2005) providing final analysis.

In 2010, the USBR conducted an entrainment study directly below Keechelus dam. A screw trap was deployed in the river channel to capture fish entrained through the outlet works of the dam and passed to the river below. The sampling was done over a range of flow releases from mid-May through August. A total of 526 fish were captured during the course of the study, representing at least 11 species but no bull trout were collected (USBR 2010).

Ahead of the Gold Creek Valley Restoration Project (See Restoration Actions section below) the US Fish and Wildlife Service conducted nighttime snorkeling surveys at randomized locations within and outside of the proposed restoration site from Aug-October, 2024 to document changes in the fish community. During four surveys they observed 231 cutthroat trout, 84 rainbow trout, 64 sculpin, 60 juvenile bull trout,13 mountain whitefish, 2 YOY bull trout, and 2 brook trout. One of the brook trout was observed far upstream in the wilderness area, about 1 km downstream from the US Forest Service trail crossing (Craig Haskell, US Fish and Wildlife Service, unpublished data).

WDFW Fish Rescue

WDFW and its partners occasionally and opportunistically attempted daytime fish rescue from disconnected pools in Gold Creek through 2018.  In 2019 WDFW began leading frequent nighttime fish rescue efforts with the help of bull trout recovery partners including Mid-Columbia Fisheries Bull Trout Task Force and Yakama Nation biologists.  Rescue efforts typically occurred in the dewatering reach between RM 0.5 - 2.21, though in 2019 and 2025 one opportunistic rescue each year also occurred in the inundation zone.  Rescues were typically completed with an electrofisher and dipnets, though some night rescues were completed with dipnets only.  From 2019 to 2023 many westslope cutthroat, sculpin species, and bull trout were rescued and returned to perennially flowing water (usually the Gold Creek Pond outlet channel).  In 2024 and 2025, only 3 bull trout were rescued each year, despite similar rescue efforts. 470 young-of-year and 105 juvenile bull trout have been rescued over the 9 years.  Between 2019 and 2023, 75% to 100% of young-of-year rescued were transferred to the Yakama Nation for their bull trout rearing and release program.

During USBR funded Reservoir Fish Passage Monitoring, WDFW conducts fish rescue across the reservoir bed from isolated pools and side channels as reservoir levels decline during the summer. All fish species are rescued and returned to the nearest area of the creek that provides the highest chance of fish survival.  Most fish rescue is conducted during daytime and using dipnets only.  Electrofishing may be used as a last resort for fish rescue if a blind or isolated channel is at risk for complete dewatering.

Yakama Nation Bull Trout Population Restoration and Monitoring Project (Rescue and Rear Program)

In 2019, the Yakama Nation (YN) initiated their Bull Trout Population Restoration and Monitoring Project with the goal of maintaining and increasing population numbers and monitoring population trends. As part of this project, young of year (YOY) bull trout are rescued from the dewatering reach of Gold Creek and temporarily relocated to La Salle fish rearing facility, where they are fed a natural diet. Since 2019, YN has successfully reared and released a total of 356 YOY Bull Trout from Gold Creek. Using adaptive management, rearing survival has increased annually from 72% to 100%. The number of fish reared and released back into Keechelus Reservoir varies year to year and has ranged from 34 to 97. Total fish released from the LaSalle Rearing Facility are as follows: 2019: 78, 2020: 63, 2021: 84, 2023: 97, 2024: 34. For the past two years, only three individual bull trout were rescued each year, thus YOY were not available for the rearing program during those years.

U.S. Fish and Wildlife Service Trap and Haul

Starting in 2019, the US Fish and Wildlife Service began trap and haul work below Keechelus Dam as an interim fish passage measure for entrained bull trout. Since that time they have collected 54 Gold Creek-origin fish. In 2024 they also began encountering fish from the La Salle Rescue and Rear program. Most have been genetically identified as Gold Creek-origin fish, while some are assumed Gold Creek origin based on their collection in Gold Creek for the La Salle Rescue and Rear program. Four fish apparently entrained through Kachess Dam based on their genetic assignment to Kachess Reservoir populations have also been collected below Keechelus Dam. The fish are a mix of previously unencountered, recaptured fish indicating multiple entrainments, and fish from the La Salle Rescue and Rear program.  PIT-tag monitoring in Gold Creek has indicated that transported fish do enter Gold Creek and presumably spawn despite dewatering in the creek, though the detections of these fish are in successive years, indicating that they generally don’t reach the spawning grounds during the same year they are transported above Keechelus Dam (Beebe et al 2025 a,b).

Monitoring of PIT tagged fish in Gold Creek and below Keechelus Dam is ongoing. In lower Gold Creek there is an antenna near the I-90 bridge, an array of two flat plates about 2.5 miles upstream, and a single flat plate antenna in the Gold Creek Pond outlet channel. A five antenna array continues to operate just downstream of Keechelus Dam- this data is uploaded routinely to PTAGIS. While the antennas in lower Gold Creek, the Gold Creek Pond outlet channel, and the Keechelus Dam outlet channel operate year-round, the upstream antennas in Gold Creek operate seasonally (generally, June-November). Acoustic telemetry in Keechleus Reservoir is ongoing (see Entrainment section) but will likely end in December 2026 because the main source of fish for tagging, the La Salle Rescue and Rear program, has collected few fish from Gold Creek in recent years.

Redd Surveys

Figure X. Gold Creek redds over time

The historic spawning period for the Gold Creek population began in early September and extended through mid-October. However, since 2009, the majority of redds have been observed in October or November. Complete bull trout redd surveys have been conducted since 1984 and cover the entire spawning area from the FS Road 4832 bridge up to the barrier waterfall. Redd counts have been highly variable (Figure X - REDD GRAPH). This probably reflects, at least in part, the migration difficulties that spawners frequently encounter as a result of the annual dewatering of the channel. Up to five survey passes are completed on Gold Creek during spawning period to capture different migration timings. Most of these surveys are successful, but occasionally low flow conditions change to high flows in a very quick time period, which may hinder surveys or obscure new redds.

An analysis of WDFW redd data between 2009 - 2024 showed that in years with good passage, very few redds have been found downstream of the pond. Whereas in years with poor passage, most of the redds have been found downstream of the pond. Very little spawning occurs in the reach that dewaters between the pond and RM 2 (3 redds in 15 years) (Conley, 2025).

Other Distribution Data (eDNA, etc.)

Between 2009 and 2025, CWU researchers have done yearly sampling in short reaches of tributaries to Keechelus Reservoir (FMO Habitat) in the vicinity of the I-90 corridor. Over this 16 year period, 16 sub-adult (120 -265 mm) bull trout have been captured (James 2025). 11 of these bull trout were captured in Coal Creek, while the rest were captured in Resort Creek (2), Rocky Run Creek (2), and Townsend Creek (1). The researchers noted that these fish were all within about 200 meters of the reservoir and that sampling efforts were limited, suggesting that there are likely many more sub-adult bull trout from the Gold Creek population that are using these tributaries for foraging until they are large enough to feed on kokanee in the reservoir (James 2025).

In 2017 an eDNA sample from Townsend Creek returned positive for bull trout (Parrish 2017). The creek had a barrier culvert removed just prior to the observation. The one recorded bull trout in Townsend Creek, noted above, was captured around the same time the eDNA sample was taken.

Population Status and Trend

The USFWS (USFWS 1998) considers the Keechelus subpopulation (i.e., Gold Creek) to be depressed, decreasing, and at risk of stochastic extirpation. WDFW rates the status of the Keechelus Lake stock as critical (WDFW 2004). There has never been a clear trend in population abundance but redd counts over the past 6 years (2020-2025) have remained below the 30-year average of 15 redds.

Habitat

Habitat Overview

FMO habitat for the Gold Creek bull trout population is in Keechelus Lake. There is no shoreline development and water sports activities on the lake are not believed to influence habitat quality. While the active pool of the reservoir is significantly depleted over the course of the irrigation season (up to 94%), an inaccessible pool of over 156,000 acre-feet remains which cannot be accessed.

Elevations on Gold Creek range from 2,530 feet at its mouth to around 3,500 feet at the barrier waterfall. Approximately 3.2 miles of the stream is in the Alpine Lakes Wilderness; the remainder flows mostly through National Forest lands (formerly WSDOT, Plum Creek Timber Company, and North Pacific Railroad Company land) and private ownership in the lower 1.5 miles. Gold mining occurred in the upper part of the watershed in the late 1800s and early 1900s (prior to wilderness designation). It is unknown what impact these operations may have had on Gold Creek. From 1968 through the mid-1980s extensive commercial logging of old growth occurred adjacent to the creek in the lower mile. Some of the logged lands were subsequently sold by Plum Creek and subdivided; a large number of seasonal-use private cabins are now present on the east side of Gold Creek between RM 1.5 and 1.8. These cabins rely on a domestic water supply that does not meet the definition of municipal water rights and risk an impairment claim when the Yakama Nation’s time immemorial water right is not being met. There are also plans for new development in the Gold Creek Valley, with a total of 60 circular lots just southwest of the current cabin community. There has been grading, road, and utility work in preparation for the community. This community does not have a valid water right from Gold Creek and plans to rely on trucking in water that will be stored in cisterns. The lower mile of the creek sustained a major impact in the late-1970s when more than 750,000 cubic yards of gravel was mined from the floodplain for the construction of Interstate 90. The most notable gravel mine site has filled with water and is now called "Gold Creek Pond". The major current land use activity in the watershed is recreation. No campgrounds are present in the watershed, and dispersed campsites are not an issue.

The upper portion of Gold Creek from the barrier waterfall down to the wilderness boundary (RM 2.5 - 6.5) contains reaches with excellent habitat conditions for bull trout. The highest quality habitat, in terms of complexity, is located between the barrier waterfall and the 2008 avalanche zone (RM 4 - 6.5). Unfortunately, dewatering has been observed in this reach during the drought years of 2023-2025. Downstream of the avalanche zone (RM 2.5 - 4) to about the wilderness boundary has a relatively straight channel that lacks complexity and has few suitable spawning gravels, but maintains perennial flow and adequate temperature. Below this point, several of the cabins mentioned above are located close to the stream’s banks; mature riparian vegetation has been removed and the banks have become unstable. As a result property owners have made efforts to stabilize them to the detriment of instream habitat quality. Heavy equipment has operated in the channel, fine sediment has been mobilized, LWD has been removed, and the channel has been redirected in places. Adjacent to and downstream of the cabin development, the stream channel becomes progressively wider and more braided. Riffle habitat containing coarse substrate materials (cobbles and boulders) is predominant, and LWD is less prevalent; what is present is often isolated on cobble bars. The active base-flow channel is primarily located a significant distance from any riparian influence and intermittent dewatering occurs annually in this reach.

The most serious effects of past land use practices on Gold Creek are in the lower mile of the creek. As a result of past old-growth harvest, almost no key pieces of LWD (i.e., large and immobile) are present to stabilize the channel and stream banks (Haring 2001, Abbe and Trotter 2013). The creek has essentially “mined” those banks, increasing bed load and creating a channel that is extensively braided. An early study estimated stream widths frequently exceeding 200 feet (Meyer 2002), while a more recent habitat assessment saw widths closer to 350 ft. at bank full width (WDFW, unpublished data, 2024). The channel dewatering prevents migrating adult bull trout from reaching spawning habitat upstream, strands adult and juveniles present in the reach, and makes fish in isolated pools vulnerable to predators. Wissmar and Craig (1997) documented stranding mortality in Gold Creek in 1993 and 1994, estimating that 63% and 24%, respectively, of adult post-spawn bull trout died in these years. The loss of continuous surface flow is an annual event, which typically occurs starting in July (Thomas 2001b; Meyer 2002, Babik 2025). A dry streambed is usually first encountered just above RM 0.5 where the outlet of Gold Creek Pond enters the creek. The affected reach extends upstream to RM 2.21, a distance of 1.71 miles. It characteristically has very short sections of flowing water and isolated pools indicating the stream is flowing not far below the surface of its bed. Groundwater monitoring from June 3rd to October 13th, 2013, showed gaps in data where the water elevation dropped to below the groundwater monitoring wells, which were installed 10 feet below the surface. This indicates that the stream is deeper below the bed surface than previously thought (Abbe and Trotter 2013). The causal mechanism for this phenomenon has been investigated and determined that, in addition to legacy land use impacts on the channel, excavation of the 22-acre gravel pit (Gold Creek Pond) as well as other anthropomorphic hydrologic disturbances, such as the Starwater drain, are contributing factors (Abbe and Ericsson 2014).

A buried 8-inch drain system dubbed the “Starwater drain” parallels Snowshoe Lane and Gold Creek Road (Natural Systems Design 2025). The drain terminates at a partially buried manhole with the bottom broken, allowing the water to drain into a relict floodplain channel that flows via surface water into Gold Creek Pond. The depressed groundwater levels created by Gold Creek Pond provides much of the hydraulic head needed to make the drain effective. Without the pond, groundwater levels at the downstream end of the drain would be much higher, backwatering the lower end of the drain and decreasing the hydraulic gradient at the northern end of the drain. Thus, the drain magnifies the effect of the pond by accelerating flow through the eastern floodplain, and into the pond (Natural Systems Design 2025).

Figure X. A drone image captured in August, 2025 by Josh Rogala at WDFW shows Gold Creek flowing across the Keechelus Reservoir bed, meandering through many different channels. The creek's origin is the valley in the upper right corner of the photo.

Often overlooked because of the problems upstream are potential passage problems on the reservoir bed. By late summer Keechelus Lake is depleted, and Gold Creek flows for a considerable distance across the reservoir bed. The channel becomes extensively braided, and the main channel is poorly defined (Figure X - Drone Pic). As of 2017, WDFW fish passage biologists have monitored the alluvial fan of Gold Creek throughout the summer, opening passage or blocking of small braided channels to increase success of passage through one main channel (Rogala 2024). At minimum pool (6%), Gold Creek extends ~1.65 miles across the exposed reservoir, depending on the channel configuration (WDFW 2025). As soon as Keechelus Reservoir is drawn down, passage at Gold Creek could be affected. Other issues like recreational rock dams, and vehicles driving through the stream occurs after drawdown. WDFW tracks and fixes issues as they arise by using small scale fish passage features that are constructed and maintained by hand. Dozens of features may be removed, modified, or built each year to maintain and improve fish passage across the reservoir bed. If stream conditions deteriorate and water depths are not sufficient for fish passage, a temporary fish passage flume may be built to facilitate fish passage. One emergency fish passage flume was constructed in 2001.

According to the K-to-K 2016 DEIS No Action alternative: "When Keechelus Reservoir level falls below elevation 2,466, tributary access for bull trout would be adversely impacted for approximately 115 days in 81 percent of years. This would be a significant impact to fish passage."

Habitat Monitoring

Field Habitat Surveys

The Forest Service constructed a spawning channel between Gold Creek Pond and Gold Creek in 1972 and 1992 (Deichl et al. 2011). Goetz (1997) attempted to determine diel behavior of juvenile bull trout and its influence on sampling techniques. Gold Creek was one of the sites in the study in which the author determined that surveys of abundance, distribution, and habitat use of bull trout should include both day and night surveys. Mongillo and Faulconer (1980) discussed benthic macroinvertebrate densities and suggested fertilization of Keechelus Reservoir for improved fishing opportunities. USBR examined Keechelus Lake limnology from 1998-2001 to estimate the anadromous fish runs that could be sustained above Keechelus Dam if passage were restored (Ackerman et al. 2002). The data suggested that only sockeye salmon could be re-established.

Sediment samples were collected in Gold Creek in 1990. Fine sediment levels averaged 4.7-13.2% across the three reaches that were sampled (Mayo 1998). The Forest Service completed a stream survey of seven 100 meter sections of Gold Creek in 1998, using Timber, Fish and Wildlife protocol (USFS 1998b). Channel type, LWD, flow, temperature, bankfull width, and pool/riffle ratios were measured during this survey. The summary of results focused on "segment 1" where authors suggested it had the most problematic habitat. Segment 1 was the reach starting at the Frontage Bridge and going upstream to the Gold Creek Pond outlet channel (RM 0 - 0.5) . They noted a lack of canopy cover, minimal large woody debris, and a wide, unstable channel with active bank erosion. They also noted that the temperature data exceeded water quality standards set by Department of Ecology (16°C) (USFS 1998b). In 2004, a Federal Roads Analysis was completed for the Gold Creek watershed (USFS 2004). The authors suggested that Gold Creek has a high risk of road-related sediment problems and that the width of the Frontage Road bridge is insufficient.

Craig (1997) monitored habitat conditions that affect bull trout spawning in several Yakima Basin spawning tributaries including Gold Creek, noting that dewatering often prevents upstream and downstream migration of adult bull trout. His first observation of subsurface conditions was on August 24th, 1996. Thomas (Thomas 2001b) summarized dates, years, flows, and reservoir elevations when Gold Creek was not passable to upstream migrating adult bull trout. Similar data for Coal, Cold, and Meadow creeks, and Keechelus Reservoir tributaries was summarized. Monitoring of the dewatering section in Gold Creek has increased in recent years. Kittitas Conservation Trust (KCT) staff began weekly pedestrian surveys of the restoration reach (RM 0.5 - 3) from July through October in 2024 and 2025, which were both drought years. In 2025, dewatering began 10 days earlier, reached the full extent (1.71 miles) earlier, and lasted 7 days longer than in 2024.

In 2025, a more concerted effort was put into understanding the timing and extent of dewatering upstream of RM 4.5 in the wilderness.

In preparation for the in-stream restoration at Gold Creek (construction slated for 2026), a modified USFS level 2 habitat inventory was done by Mid-Columbia Fisheries' BTTF in 2024. The survey included RM 0 - 3, which covered the restoration reach, and an upstream reference reach (RM 3 - ~3.5). The BTTF also did a separate monthly "pool survey" in the dewatering zone in the summers of 2023 and 2024. The data from these habitat surveys was delivered to WDFW, USFWS, and KCT and has not been summarized or reported on as of writing this.

KCT installed groundwater and surface water monitoring wells in relevant areas to the restoration project in 2013 and 2014. USFWS installed their own groundwater monitoring wells in XXXX? CRAIG / KATY?. They intend to re-install these wells for post-project monitoring. There are also plans to monitor other habitat features post-project.

Stream Temperature Data

USFS (1998b) reports summer temperatures in 3 segments of Gold Creek. Willey (2007) collected temperature data in Gold Creek from 2005-2007. The USFS deployed temperature loggers during the summer low flow period at three sites in 2007 and two other sites in 2010 (USFS 2011b). Kittitas Conservation Trust, working with Natural Systems Designs, installed surface water monitoring stations that included temperature in 2013 and 2014, but most of those were dry during peak summer temperatures. USFWS has been monitoring temperature at their PIT antenna sites since 2019, and in the pond outlet channel since 2023. WDFW Water Science Team has monitored temperature at the USFS Frontage Road bridge since 2022, and at the USFWS UGC PIT antenna site since 2024, and in the Gold Creek outlet channel starting in 2025. Finally, the Bull Trout Task Force started monitoring temperature in the wilderness near the Gold Creek Trail crossing in 2024. There have been many short term monitoring sites in Gold Creek, but data compilation and long term analysis has not occurred.

A longitudinal analysis of temperature data was done for 2023-2025. It showed that maximum weekly maximum temperature (MWMT), a measure of the highest 7-day average of daily maximum stream temperatures, ranges from 13-18 °C from upstream to downstream sites. The mean August temperatures are lower, between 11 - 16 °C. It appears that Gold Creek upstream of the dewatering zone (RM 2.5 +) remains suitable for bull trout spawning and rearing, while downstream reaches during the warmest parts are less suitable for rearing but probably suitable for migration.

Restoration Actions

Former land owned by the railroad and private timber companies has been acquired and now the primary landowner in the Gold Creek watershed is the USFS.

Approximately 400-ft downstream of the USFS road 4832 (Frontage Road) bridge, I-90 crosses the Gold Creek Valley. In 2013, the Washington DOT (WSDOT) completed the construction of two 1,000-ft wide I-90 bridges over the Gold Creek floodplain and 120-ft wide wildlife undercrossing bridges. This was a critical part of the FHWA/WSDOT I-90 SPE project, which seeks to address wildlife connectivity needs while improving the capacity and safety of the interstate in this region

In 2019, the Yakama Nation (YN) initiated their Bull Trout Population Restoration and Monitoring Project with the goal of maintaining and increasing population numbers and monitoring population trends.

Planned restoration

The Gold Creek Valley Restoration project restores natural streamflow to Gold Creek, improving habitat for ESA listed Bull trout, reducing thermal barriers, and supporting the Yakama Nation’s time immemorial water right and YBIP’s goal of restoring healthy, harvestable fish populations. Phase 1, to be implemented in 2026, will restore 2.5 miles of instream habitat by installing 141 engineered log jams (ELJs), constructing 5 side channels, planting 20 riparian acres, restoring a small borrow pit, and filling 30% of the larger pit. Phase 2 is dependent on funding and will fill the 22-acre gravel borrow pit to restore groundwater gradients, increase summer baseflows, and eliminate the modeled siphon effect of streamflow loss. Phase 3, also dependent on funding, will restore groundwater aquifers by removing a levee, reconnecting 245 acres of floodplain, constructing 3 side channels, and establishing wetland and forested floodplain habitat.

USFS is also seeking construction funds to replace an existing 70' wide bridge across USFS road 4832 and over the Gold Creek floodplain and to construct a new 120-ft wildlife undercrossing and install wildlife exclusion fencing. USFS states that the bridge constrains GC to only 80% of its channel width.

Threats

Connectivity

Fish Passage Barriers

Threat Severity: Significant

About 1.7 miles of lower Gold Creek dewaters annually. The dewatering typically begins at around RM 1.6 and extends upstream and downstream from there.  The maximum downstream extent is just upstream of the confluence with the Gold Creek Pond outlet channel (RM 0.5), and the maximum upstream extent is just downstream of the upstream PIT tag array (RM 2.21), with flows starting to go fully subsurface between mid-July and mid-August. This is believed to occur due to past land use activities including mining and logging (Meyer 2002). These activities increased channel width, reduced LWD recruitment, and resulted in massive amounts of alluvial material settling in the valley bottom. The issue is exacerbated by the adjacent Gold Creek Pond, which increases the rate of dewatering due to groundwater flow to a slightly lower elevation. The river generally remains disconnected from its upstream reaches until fall rains reconnect it, typically sometime in October. See Habitat Overview section above.

In recent years (drought), dewatering has been observed in upper Gold Creek (RM 4.5+) as well. The timing and extent is not well documented, but weekly surveys were conducted in 2025 in an attempt to understand the conditions. A survey report from late August 2025 notes "0.15 miles were dewatered completely, and 0.23 miles appeared to be close to dewatering" (Acosta 2025). This survey was after a mid-August rainstorm that actually restored flow to about 0.8 miles of the downstream dewatering area (Babik 2025). A field assessment in 2024 showed similar results with 0.25 miles of dewatering (Scott Kline, WDFW, personal communication).

In addition to Gold Creek dewatering and causing adult and juvenile passage and mortality issues, the construction of Keechelus Dam in 1917 blocked upstream fish passage into the reservoir and Gold Creek. The only passage from below the dam is through USFWS' trap and haul program, and only Gold Creek origin fish that were entrained through the dam are transported back over.

Passage barriers (culverts and bridges) along the I-90 corridor tributaries (Rocky Run, Resort, and Townsend creeks) were replaced between 2013 and 2016.

Entrainment

Threat Severity: Significant

A 2010 study by USBR trapped fish below Keechelus Dam to understand entrainment, but no bull trout were captured (USBR 2010). However, since 2019, the USFWS has collected 54 entrained Gold Creek-origin fish below Keechelus Dam from September-November. A higher number (n=6) of these fish have been La Salle Rescue and Rear fish in recent years. A single Gold Creek-origin fish was also collected below Kachess Dam, apparently migrating downstream to Easton Reservoir and up the lower Kachess River. Overall, more entrained bull trout have been collected below Keechelus Dam in drought years suggesting a relation between pool elevation, discharge, and entrainment. PIT-tag detections downstream of the dam suggest most fish entrain when Keechelus Reservoir elevation drops below 2,460 ft (Beebe et al 2025 a,b).

From July 2024 through December 2025, USFWS tagged 63 bull trout and used acoustic telemetry from an acoustic array in Keechelus Reservoir and PIT tag detections from a PIT antenna below Keechelus Dam to evaluate entrainment at Keechelus Dam. Data collection and analysis is ongoing, but thus far, 10 of the fish entrained resulting in an entrainment rate of about 25% annually with most entrainment occurring during July and August. The preliminary analysis indicated that overall, seasonality (day-of-year) explained more deviance than pool elevation and discharge. Entrainment timing coincided with both peak irrigation releases and pre-spawn movements toward Gold Creek, making it difficult to separate the effects of Keechelus Dam discharge and reservoir pool elevation from seasonal behavioral movements (Connor Cunningham, USFWS, in prep).

Dewatering due to flow management

Threat Severity: _______________________

A 2000 study to evaluate the effect of Keechelus Reservoir drawdown on groundwater levels in the lower Gold Creek basin found that the water levels of the pond, and nearby groundwater levels did not drop in relation to drawdown of the reservoir (Didricksen 2001). Thus, the only dewatering issues due to flow management are probably within the inundation zone where sediment delivered by Gold Creek can rapidly drop out at the pool interface. As the pool is drafted down, the creek carves different pathways through the sediment each year resulting in shallow and heavily braided channels that would limit fish passage if there was not monitoring and intervention by WDFW fish passage team.

According to the K-to-K 2016 DEIS No Action alternative: "When Keechelus Reservoir level falls below elevation 2,466, tributary access for bull trout would be adversely impacted for approximately 115 days in 81 percent of years. This would be a significant impact to fish passage."

Land-use Issues

Forestry

Threat Severity: Moderate

Figure X. An aerial photo of the lower Gold Creek, before and after logging.

The Kachess Reservoir watershed experienced substantial timber harvest between the 1940s and 1980s (Deichl et al. 2011, Meyer 2002). Logging went along the river corridor all the way to the banks of the creek. Roads were built to haul materials and logs out of the site (Figure X).The effects of clear cut logging and associated roads include reduced recruitment and storage of large woody debris, bank instability, excessive cobble deposition in the valley bottom, increased stream temperature, and to some extent, dewatering. The second generation regrowth along the banks cannot withstand Gold Creek’s high velocities and erosion continues to degrade the riparian forest.

The current concern with forestry is the monoculture secondary regrowth that has not been managed for fire risk. There is a high risk of a severe wildfire in the watershed. (Suggestion to look at USFS land management maps and determine what is protected Owl habitat vs slated for harvest... anyone know where to access these maps?)

Agriculture and Grazing

Threat Severity: Insignificant

There is no agriculture or grazing in the Gold Creek watershed.

Recreation

Threat Severity: Unknown, likely insignificant

Gold Creek Pond is a popular recreation destination in both summer and winter. There is a large paved parking lot (in the historic Gold Creek floodplain) and paved walking path all the way around Gold Creek Pond. Few people realize that Gold Creek is adjacent to the pond because it is separated by a thick band of riparian vegetation. This limits recreation impacts to lower Gold Creek. The USFS maintains Gold Creek Trail, which begins at the end of USFS Road 146. The trail follows the creek for about 3.5 miles before crossing and going upslope toward Alaska Lake. The trail only has a few places where creek access is very easy. The road, USFS 146, is gated at the bottom near the Gold Creek Pond parking lot to reduce traffic and theft in the Gold Creek community. Walking to the trailhead from the pond adds an extra ~2 miles round-trip and thus limits use of the trail.

There is currently a 10-year seasonal (summer) closure in place at Gold Creek to protect public safety during the restoration project. The area will be open to winter recreation when safety can be maintained. The I-90 wildlife corridor is legally closed to recreation

For many years, the main threat related to recreation was vehicles driving on the Keechelus Reservoir bed and crossing back and forth through Gold Creek. This threat has been drastically reduced by the installation of a locked gate at the Keechelus Recreation area. The gate was installed in 2023 and is locked when the reservoir pool elevation is 2,480 ft and below. ORV can still access the reservoir bed from portions of the Palouse to Cascades Trail and from other border areas around the reservoir. There are still tire tracks found crossing Gold Creek almost weekly, but USFS has plans to install additional barriers to further reduce vehicle access to the reservoir bed while also making the reservoir bed closure permanent.

Roads and Development

Threat Severity: Moderate

A large number of seasonal-use private cabins are present on the east side (Ski Tur community) of Gold Creek between RM 1.5 and 1.8. There are still vacant lots that could be purchased and developed. These cabins rely on a domestic water supply that does not meet the definition of municipal water rights and risk an impairment claim when the Yakama Nation’s time immemorial water right is not being met. There are also plans for new development (Starwater) in the Gold Creek Valley, with a total of 60 circular lots just southwest of the current cabin community. There has been grading, road, and utility work in preparation for the development. This community does not have a valid water right from Gold Creek and plans to rely on trucking in water that will be stored in cisterns. Development of Starwater and other property will reduce the ability to deal with the issues related to Starwater drain (See Habitat Overview above for more details on the drain). Additionally, some cabin owners with property adjacent to Gold Creek are placing rock to armor the banks.

Private properties to the west of the creek have also been proposed for large developments, but that has been limited, again by the lack of domestic water supply. Expansion of water utilities may facilitate future development.

Monitoring of water quality has not occurred, but water quality issues are a concern.

Mining

Threat Severity: Insignificant (historical effects still significant).

While there are no current mining operations in the Gold Creek watershed, three major mining companies operated in the watershed in the early 1900s (Deichl et al. 2011). Gold, silver, and copper were the most common target minerals. Mine associated development included hundreds of feet of shafts and tunnels, construction of Chilean mills and stamp mills, and an unknown quantity of outbuildings (Deichl et al. 2011). There was one mining claim that remained active in the Gold Creek headwaters until it was abandoned or forfeited in 1992(https://thediggings.com/mines/ormc5682).

One of the most impactful mining operations was the mining of gravel from the Gold Creek floodplain to support the construction of I-90 in the 1970s. Pit site PS-S-256, now known as Gold Creek Pond, and other nearby quarries and pits were excavated to provide a haul of 55,255 cubic yards, a haul of nearly 100,000 cubic yards of stockpiled strippings, crushed surfacing top course estimated at 51,800 tons, and ballast totaling ~87,200 tons (Deichl et al. 2011). Even though congress passed NEPA in 1969, there was no mention of NEPA in the design plans for I-90 projects in the 1970s. The plans did include some environmental elements like the construction of a "pervious dam" from the pond to a spawning channel. They also proposed revegetation of the Gold Creek area (Deichl et al. 2011). This gravel pit is now the primary cause of dewatering of the adjacent Gold Creek, as groundwater flow draws water from the creek down to the lower-elevation pond.

Ecological Interactions

Brook Trout

Threat Severity: Unknown, likely significant

Brook Trout have been documented in Gold Creek since the first fish surveys occurred in the late 1990s (Craig 1997, USFS 1998b). The forest service noted that brook trout were only observed in lower Gold Creek in or near beaver-altered habitat. Meyer (2002) described the fish community in Gold Creek and observed 97 individual brook trout, or 16% of the total community observed. A small number of brook trout are captured in most years during fish rescue efforts (0 - 6) and are opportunistically culled. It is likely that the annual dewatering is helping to limit the upstream expansion of brook trout in the watershed. However, one brook trout was observed upstream of the dewatering near RM 3 during a USFWS snorkel survey in 2024. Warming stream temperatures, increased beaver activity, and future restoration of quality fish habitat might result in increased brook trout abundance. There is no documentation of hybridization with bull trout to date in either the Kachess or Keechelus watersheds based on ongoing assessments of fish collected through the trap and haul and La Salle Rescue and Rear programs (Beebe et al. 2025b).

Other Invasive Species

Threat Severity: Insignificant

There are no reports of other invasive species in Gold Creek.

Diminished Prey Base

Threat Severity: Unknown, likely significant

A 2017 study looked at the food web dynamics in Keechelus Reservoir (Hansen et al. 2017). The authors suggest that prey base available to bull trout in Keechelus Reservoir may be limited by food-web dynamics within the reservoir. Research indicated that the reservoir relies heavily on pelagic production, likely due in part to reservoir drawdown that reduces littoral habitat and associated productivity. Kokanee, an important prey item for bull trout, appear to experience limited growth and feeding opportunities because zooplankton densities, particularly Daphnia, are relatively low and often concentrated in the warm epilimnion during summer stratification. As surface waters warm, kokanee avoid these upper layers and remain deeper in the water column, reducing their access to key zooplankton prey. WDFW stocks an average of 250,000 kokanee annually in Keechelus, which supplements prey for bull trout. Other piscivorous species such as burbot and northern pikeminnow occur in the reservoir and may compete with bull trout for prey fish or consume juvenile fish that could otherwise contribute to the prey base (Hansen et al. 2017). Together, these factors may constrain forage availability for bull trout that migrate from Gold Creek into Keechelus Reservoir to feed.

The macroinvertebrate and forage fish base in Gold Creek itself is understudied (ARE THERE ANY MACRO STUDIES TO CITE?). However, data from annual fish rescue operations in Gold Creek show a relatively high density of sculpin and juvenile cutthroat trout when compared to other systems like Kachess River or Deep Creek (Scott Kline, WDFW, unpublished data). For example, the 2025 fish rescue season at Gold Creek resulted in the capture of 958 individual fish that were not bull trout.

Disease

Threat Severity: Insignificant

There have not been any observations of disease in Gold Creek.

Water Quantity and Quality

Flow issues/dewatering

Threat Severity: Significant

Seasonal dewatering of Gold Creek is probably the most significant threat to this population. The situation has been described above in Habitat Overview and the threat relating to Fish Passage Barriers. It is thought that filling and returning Gold Creek Pond to a more natural wetland condition, and reducing or eliminating the effects of Starwater drain will mediate this threat in areas adjacent to the pond. However, in a changing climate with increased drought, decreased snowpack, and earlier peak runoff, dewatering will likely continue to be a threat to this population, both in the headwaters and downstream.

Current and modeled future temperature conditions

Threat Severity: Moderate

Figure X. Gold Creek Mean August Temperatures at four sites (upstream to downstream) from 2023-2025

The NorWeST stream temperature model indicates that climate warming may increase summer stream temperatures throughout the Gold Creek watershed (Isaak et al. 2017). Under the 2040 climate scenario, headwater reaches are projected to remain relatively cool while mid- and lower-basin segments are expected to warm to approximately 14–16 °C during August. By the 2080 scenario, additional warming is projected across the watershed, with lower valley segments potentially exceeding 16 °C and headwater areas approaching 10–12 °C. However, recent temperature monitoring in Gold Creek shows a similar longitudinal pattern to 2040 projections, with mean August temperatures ranging from approximately 10–11 °C in upper reaches to roughly 13–16 °C in downstream valley segments between 2023 and 2025. Maximum weekly maximum temperature in lower reaches have reached approximately 16–18 °C during summer. These data include the pond outlet channel, which is known to deliver warmer water relative to the mainstem creek. These observations indicate that portions of lower Gold Creek are already experiencing temperatures comparable to those projected under mid-century climate scenarios, suggesting that suitable cold-water habitat for bull trout may become increasingly restricted to upper reaches of the creek. Something to note is that the NorWeST temperature model was built on data from several years with cooler temperatures, so it may be biased to lower "current" conditions than the long term average.

Thermal conditions in the watershed also interact with seasonal flow conditions. When lower Gold Creek dewaters during mid-summer, reduced surface flow and isolated pools can warm rapidly, further increasing temperature stress for fish attempting to migrate or rear in downstream portions of the watershed. These combined effects of warming and seasonal dewatering may compound habitat limitations for the Gold Creek bull trout population during critical summer periods.

Fisheries Impacts

Angling Regulations/Fisheries Use/Poaching (Recreational)

Threat Severity: Unknown

Little is known about the effects of Angling to the Gold Creek population of bull trout. Angling is unlikely to be a significant threat to this population for a couple of reasons. 1) Gold Creek is closed to fishing year-round and there are very few documented instances of angling in the closed area. 2) Keechelus Reservoir is the first to be drafted down for irrigation water lower in the basin. The rapid lowering of water surface elevation usually leaves the boat launch high and dry early in the summer, and it does not become accessible again until late spring the following year. Thus, fishing access is restricted to the banks. 3) As discussed above, public access to the reservoir is now limited during summer due to the gate that was installed to reduce vehicle damage to the creek, so access is on foot. 4) The bull trout population has low abundance and people probably aren't encountering them very often.

The Bull Trout Task Force visits the Keechelus Reservoir each summer and has a good relationship with residents of the area. Residents are frequently fishing from shore, but report very few bull trout interactions (1 instance across 6 years of outreach; a sub-adult bull trout was caught and released). Their main target is Kokanee salmon, but also report that their primary catch is pike minnow (Aimee Taylor, MCF, personal communication). Residents are aware of how to identify bull trout, and report that they help with outreach regarding bull trout to non-residents.

Management/Monitoring (Research)

Threat Severity: Unknown, likely insignificant

Monitoring actions for this population include annual redd surveys, fish rescue from the dewatering zone, captive rearing, and trap and haul of entrained individuals. Biologists will opportunistically measure, weigh, PIT or acoustic tag, and collect genetic tissue samples from Gold Creek bull trout. Management of this population is hopefully increasing survival, not posing a threat.

Other Threats

Low abundance

Climate Change

Climate projections indicate that watersheds in the Pacific Northwest will experience significant hydrologic changes as temperatures warm. Warmer winters are expected to increase the intensity of winter storm events and shift precipitation from snow to rain, resulting in higher winter peak flows and reduced snowpack storage. These changes will reduce spring and summer baseflows and lengthen low-flow periods (Mantua et al. 2009). Transition watersheds, which receive both rain and snowmelt runoff, are particularly vulnerable to these changes. Gold Creek is classified as a transition watershed and is already experiencing these climate-driven hydrologic shifts, which directly affect the quantity, timing, and temperature of water available for fish habitat. Regional streamflow projections indicate increasing peak flows and declining summer low flows across the Yakima Basin (Yoder et.al, 2022). In the Gold Creek watershed, peak flows are projected to increase by 23% while summer low flows may decline by as much as 38% between 2020 and 2049 compared to the 1980–2009 baseline (University of Washington Climate Impacts Group 2023, RCP 8.5 scenario). Increased winter flooding may scour redds and reduce egg-to-fry survival if adequate refugia are not available (Mantua et al. 2009). Climate change is also expected to increase stream temperatures in the watershed. Bull trout require cold water, with a critical temperature threshold of approximately 15°C. Rising air temperatures may push stream temperatures into the stressful range for salmonids (17–21°C) by 2040 and potentially into lethal ranges (21–24°C) by 2080. In Gold Creek, this risk is amplified by the existing gravel borrow pit, which warms groundwater before it returns to the stream. Without restoration, the combined effects of warming temperatures, reduced snowpack, and seasonal dewatering could significantly degrade habitat for multiple life stages of Bull trout and other aquatic species.

Summary of Primary Limiting Factors and Threats

One of the highest severity threats to this population is the frequent channel dewatering within lower Gold Creek that results in direct mortality and limits access to spawning habitat upstream. Other high severity threats include low population abundance and the passage barrier at Keechelus Dam. Other moderate or high severity threats include: development in the lower reaches of Gold Creek, entrainment at Keechelus Dam, potential expansion and introgression with brook trout, and future temperature conditions.

While forest management and recreation issues are present, they are not thought to be significant. Frontage road crosses Gold Creek at its mouth (when the reservoir is at full pool) significantly impacting floodplain function; however effects on bull trout are unknown. The threat of angling is unknown. Agriculture, altered flows, grazing, limited habitat, and mining are not present in this population area.

Recovery Strategy

Population-level Recovery Strategy

This population has been identified as a high priority “Action” population (see Prioritization of Actions). The highest priority action for this population is to complete the Gold Creek Valley restoration project, slated to begin in 2026 and continue for up to 10 years. is a hydrologic assessment and subsequent restoration project The project connect dewatered sections in the stream, which strand fish and prevent access to spawning grounds. The rescue and rear program should continue to support increased population abundance while habitat is being restored. Other high priority actions include passage at Keechelus Dam and an evaluation of further supplementation to address low abundance. Other actions to address threats include outreach, protection from future development, and carcass analog placement if a pilot study conducted elsewhere in the basin is successful. There have been documented hybrids in the system, and Introgression with brook trout should continue to be monitored. A study on the current distribution of brook trout in the watershed would help prioritize areas for targeted suppression or eradication. although no large-scale removal actions are recommended at this time.

---

Actions

Future link/ iframe to PowerApp will go in this section.

Completed Bull Trout Recovery Actions

• WSDOT purchased 550 acres on west side of lower Gold Creek around 2008 (ownership has since been transferred to the Forest Service).
• Forterra (formerly The Cascade Land Conservancy) purchased a total of 250 acres on the east side of lower Gold Creek in 2009.
• Length expansion of the Interstate-90 bridge over lower Gold Creek, a WSDOT project benefitting the lower Gold Creek floodplain was completed in 2013.
• Fishing regulations have been implemented to protect bull trout in Gold Creek.

Recommended Actions

Relevant Multiple Population Actions

Multiple Populations #1: Provide outreach on bull trout conservation issues (landowners, recreationists, anglers, school groups, and others).

Multiple Populations #2: Continue redd surveys within established index areas to monitor long-term trends in abundance.

Multiple Populations #3: Continue temperature monitoring.

Multiple Populations #4: Evaluate supplementation (see Appendix D).

Multiple Populations #5: Carcass analog placement if pilot studies demonstrate success.

Multiple Populations #7: Continue to screen all collected genetic samples for evidence of genetic introgression with brook trout.

Multiple Populations #9: Periodic entrainment studies at dams.

Gold Creek Actions

Gold Action #1: Provide Connectivity at Keechelus Dam

Gold Action #2: Instream and Floodplain Habitat Restoration

Gold Action #3: Restore Groundwater Hydrology

Gold Action #4: Long Term Habitat Protection in the Gold Creek Valley

Gold Action #5: Monitor and Improve Passage Conditions in Gold Creek

Gold Action #6: Replace USFS Road 4832 Bridge Over Gold Creek

Gold Action #7: Supplementation, Fish Salvage and/or Translocation

Gold Action #8: Implement Passage & Habitat Projects in Keechelus Lake Tribs.

Gold Action #9: Outreach Coordination & Strategic Planning

Actions in the Yakima Steelhead Recovery Plan that benefit this population

None

Update Notes

2012 BTAP text copied into Yakipedia and edited to match new format in March 2026 by Aimee Taylor. ADD LINK TO 2012 BTAP pdf placemark

Additional edits proposed by Alex Conley, Aimee Taylor and Gold Creek Population small group. Reviewed, and updated and approved by BTWG in XXXXXXXXXXXXX

Specific detail on out of cycle updates: