Coastal Cutthroat Trout (Oncorhynchus clarkii clarkii) exhibit high variability in spawn timing across their range and also locally. However, little is known about the factors driving annual variation in spawn timing within a river system. In this study, we examined how specific abiotic factors influence spawn timing of Coastal Cutthroat Trout from Skookum, Totten, and Eld inlets in South Puget Sound, Washington. Weekly counts of Coastal Cutthroat Trout redds in Skookum Creek across nine years (2009–2017) and Kennedy and McLane creeks in 2015 were compared to measurements of streamflow, water temperature, air temperature, atmospheric pressure, precipitation, tidal coefficient, and photoperiod to identify determinants of spawn timing. Spawn timing (i.e., redd counts) varied across study years with streamflow representing the most important predictor of redd counts. Across study years the date of peak redd counts varied by more than two months (15 February 2013 versus 21 April 2011). Annual redd abundance was negatively correlated with mean streamflow during the spawning period and the highest redd abundances occurred during periods when streamflow was between 0.25 and 0.50 m3·s-1. These results are cause for concern when considering the expectation for increased storm intensity, flooding, and periods of low summertime streamflow conditions associated with climate change. This work builds a foundation for future projects aimed at understanding how climate change may influence the spawn timing and abundance of Coastal Cutthroat Trout.
Introduction
The spawn timing of Pacific salmonids is influenced by environmental factors that affect growth and maturity throughout their early life-history and by sympatric species that overlap temporally and spatially (Quinn 2018). Variation that occurs between and within natal tributaries is largely driven by adaptations to abiotic conditions like streamflow and temperature, which have been selected for over long evolutionary periods (Quinn 2018, Tillotson and Quinn 2018). The dependence of stream biota on variable environmental conditions has been described in detail (Fisher 1983, Schlosser 1990, Poff et al. 1997), and for fishes, exogenous factors such as water flow and temperature are often the leading determinant of spawning phenology (Heins 2020), including egg incubation duration, mate choice, and spawn timing (Quinn 2018). For many large-bodied Pacific salmonids that spawn in the fall, spawn timing is predictable and occurs at the same time each year. For instance, the date of peak live counts for fall run Chum Salmon (Oncorhynchus keta) in Skookum and Kennedy creeks (South Puget Sound, Washington) varied by less than one week (standard deviation [SD] of 6.07 and 5.08 days, respectively) across the span of a three decade-long dataset (Agha et al. 2021), aligning closely with long term averages in peak flow.
For Coastal Cutthroat Trout (O. clarkii clarkii; CCT), spawning ground surveys in Skookum Creek, where redds are enumerated on a weekly basis, have shown a more protracted spawn timing than expected based on early descriptions (Johnson et al. 1999), but also a high degree of interannual variability over the span of six years (Losee et al. 2016). This flexible spawn timing was also demonstrated in the Elwha River system above a human-made impassable dam (McMillan et al. 2014). There, the unusual fall spawn timing of the fluvial CCT population was hypothesized to result from limited fall flood scour and the dam-related exclusion of sympatrically spawning Coho Salmon (O. kisutch) from the spawning grounds. In Irely Creek, a stream in Washington's Olympic National Park, the number and timing of adfluvial CCT redds were variable and strongly related to the degree of lake dewatering that occurred downstream of the monitored spawning habitat the summer/fall prior to spawning (Vadas et al. 2016). This negative relationship with stream dewatering suggests that CCT spawn timing and redd abundance may be affected by instream flow conditions, similar to other salmonids (Quinn et al. 1997, Bennett et al. 2014, Jonsson et al. 2018). For CCT in South Puget Sound and elsewhere, the absence of sympatric adult salmonids during the entirety of the observed spawning period and the high degree of environmental variability (flow, temperature, etc.) suggest that abiotic variables, rather than interspecies competition, likely drives variability of spawn timing. It is known that streamflow and temperature are important for salmon migration and spawn timing (Vadas et al. 2016, Quinn 2018, Losee et al. 2024), but this has not been well tested for anadromous trout outside of Washington State (Losee et al. 2024).
In addition to variability in spawn timing, results of studies investigating anadromous CCT migration phenology have shown considerable variation in the period between entry into freshwater and spawning. For instance, Trotter (1989) reported a fall entry timing and freshwater overwintering prior to spawning. This tendency to enter freshwater in the fall, feed on salmon eggs, and remain in freshwater for months prior to spawning has been reported in other scientific studies and anecdotal reports (Quinn 2018) where they are commonly referred to as “harvest trout”. In contrast, Losee et al. (2018) documented high site fidelity of genetically distinct CCT populations at nearshore marine locations in South Puget Sound with less than a one-month lag between peak catches in marine waters and peak redd counts in natal streams, which suggests staging in marine water prior to spawning. Similarly, results from otolith microchemistry analysis (Claiborne et al. 2020) and other studies (e.g., Quinn and Myers 2004) have revealed extended marine overwintering and a short period (days to weeks) of freshwater residence during the spawning period for many anadromous CCT. Together, this information highlights the complex migration phenology of anadromous CCT and the diversity of factors that may influence freshwater entry and spawn timing.
In this study, we compared a suite of environmental variables to weekly redd counts of CCT across nine years and three streams in South Puget Sound. The specific objectives of this study were to: 1) identify abiotic factors influencing CCT spawn timing; 2) examine whether relationships between spawn timing and certain environmental variables were stable over the study period and across neighboring populations; and 3) compare total redd abundance to those environmental variables related to spawn timing.
Methods
Study Overview
This study is one component of a multidisciplinary project aimed at improving understanding of the marine and freshwater ecology, movements, and abundance of CCT (Losee et al. 2017a). A two-part approach was used to maximize the use of available redd count data and test the stability of relationships between environmental variables and spawn timing across years and between neighboring systems. To achieve this, a suite of seven environmental variables (described below) was used to investigate conditions at the time of redd construction compared to conditions across the entire time period in three streams (Skookum, Kennedy, and McLane creeks) where extensive CCT redd monitoring occurred in 2015 (Figure 1). The most significant environmental predictors were used in conjunction with an expanded set of years in Skookum Creek where nine years of survey data were available (2009–2017) to understand the stability of relationships between environmental variables and spawn timing.
Figure 1.
Study area in South Puget Sound. Colors indicate streams surveyed for Coastal Cutthroat Trout Oncorhynchus clarkii clarkii redds and monitored for temperature and streamflow. Black dots indicate upper and lower boundaries of spawning ground surveys; the yellow dot indicates location of tide station; the green dot indicates location of Shelton Airport weather station; and blue dots indicate locations of streamflow and temperature recording.
Study Area
Skookum, Kennedy, and McLane creeks are all fourth order (Hughes et al. 2011), independent streams that drain directly into three separate inlets of Puget Sound: Little Skookum Inlet, Totten Inlet, and Eld Inlet, respectively (Figure 1). Previous work revealed that each of the three study streams contained CCT populations that were genetically distinct from one another and a small proportion (< 5%) of each population had resident life histories relative to the anadromous form (Losee et al. 2017b, 2018). All three study streams support naturally spawning populations of fall spawning Coho Salmon and fall run Chum Salmon. Historically, spring spawning steelhead (O. mykiss) existed in each stream, but neither adults nor juveniles have been documented in over three decades (Robert Leland, Washington Department of Fish and Wildlife [WDFW], personal communication). These streams were chosen based on the ability to safely view and survey redds under the majority of flow conditions, and a Physical Habitat Simulation Model (PHABSIM) evaluation that suggested the index (survey) areas encompassed over 80% of the suitable spawning habitat available to salmonids (Boessow et al. 2021, Beecher et al. 2024). CCT redd density at the reach scale has been correlated with the quantity and quality of available spawning habitat (Magee et al. 1996); therefore, we are confident that our study area accounted for a large proportion of the total number of redds in each study stream.
Skookum Creek is a 14.5 km-long independent stream that drains into the head of Little Skookum Inlet in the southernmost reach of Puget Sound. Skookum Creek originates at an elevation of 195 m in the low hills approximately 12 km southwest of Shelton, Washington. The stream drains through a steep, forested valley until river kilometer (RKM) 9.7, where it enters a wide, flat valley characterized by moderate agricultural development and poor riparian cover. Kennedy Creek is 15.4 km long, and while slightly larger than Skookum Creek, is only accessible to anadromous fish up to RKM 3.7, where a 10-m tall three-tiered falls acts as a complete barrier to anadromy. Kennedy Creek's lower reaches, however, are significantly less developed than Skookum Creek's lower watershed. McLane Creek, at 8.0 km in length, is the smallest of the three study streams and is accessible to anadromous fish throughout its length. Like Skookum Creek, the lower watershed of McLane Creek is low gradient and comprised of moderate agricultural development and poor riparian cover, whereas the upper watershed is forested with low to moderate urbanization.
Survey Methods
Redd count surveys were conducted on Skookum Creek between RKM 9.6 and 11.7 from 2009 to 2017 and on Kennedy and McLane creeks in 2015 (RKM 2.4–3.7 and 5.6–7.7, respectively). Surveys were conducted once every seven days, as flows allowed, across the entire survey area by the same experienced staff and volunteers for the duration of the study. Survey areas encompassed greater than 75% of the available spawning area in each of the three systems and are referred to as ‘index areas’ in this paper. Redds were identified and marked with survey flagging in the pit and labeled with a flag in a nearby tree so new and old redds could be identified and new redds could be summed at the end of the season to calculate total redd abundance in each survey area. Observed live and dead fish were also enumerated and recorded by species. Chum and Coho salmon redds that were still visible in January were identified and marked at the beginning of each CCT survey season. The redds of Brook Lamprey (Lampetra richardsoni), which spawn sympatrically with CCT, were identified and excluded from the study based on criteria described in Stone (2006), and no other spring spawning species were present during the study period. For additional details related to CCT spawning ground survey methodology, see Losee et al. (2016).
Environmental Variables
Streamflow—Changes in streamflow influence migration patterns of anadromous salmon and trout (Banks 1969, Bjornn 1971, Mantua et al. 2010, Vadas et al. 2016, Lazzaro et al. 2017). Upstream spawning migrations often occur in association with freshets that increase streamflow. However, relationships between spawning migrations, active spawning, and patterns in streamflow are rare in the literature for salmonids and non-existent for CCT (Losee et al. 2024).
In the fall of 2014, a staff gage was installed near the mid-point of the spawning index area on Skookum Creek. Gage heights were coupled with instream flow measurements taken throughout the data collection period to create a hydrograph for the season. Specifically, WDFW staff and volunteers collected three gage height readings per day from 15 October 2014 to 11 June 2015. Instream flow measurements were taken on four dates across this time series with a Hach FH950 electromagnetic flow meter, following standardized United States Geological Survey methodology. The relationship between the natural logarithm of the streamflow measurements and the natural logarithm of the known gage heights at these stream flows (y = 3.41x + 2.202, R2 = 0.99) was used to create the hydrograph for the season, consistent with Boessow et al. (2021), Beecher et al. (2010), and Vadas et al. (2016). This method was also used on Kennedy (y = 4.866x + 0.337, R2 = 0.85) and McLane (y = 1.980x + 2.207, R2 = 0.99) creeks in 2015. Additionally, continuous streamflow measurements were collected at a site located approximately 9 km downstream of the bottom of the spawning index area in Skookum Creek. The relationship between these streamflow data (measured as daily average streamflow) and the upstream hydrograph (y = 0.709x + 0.062, R2 = 0.90) was used to estimate streamflow on the spawning grounds for days in 2015 when staff gage readings were not taken and for all other years in the study period.
Water Temperature—Temperature is an important determinant of spawn timing for Cutthroat Trout (Wyatt 1959, Johnson et al. 1999, Bennett et al. 2014) but has not been directly assessed for the anadromous CCT subspecies. To generate estimates of temperature on the spawning grounds of Skookum Creek, measurement methods varied in some years but were primarily based on measurements collected every five minutes using an Onset HOBO U22-001 Pro v2 logger approximately 9 km downstream of the spawning index area for the years 2011–2014 and 2016–2017 (data courtesy of Squaxin Island Tribe, Erica Marbet). Additionally, in 2015, HOBO Pendant temperature loggers were installed in the spawning index areas to collect direct estimates of temperature on the spawning grounds of Skookum, Kennedy, and McLane creeks. Temperature data collected on the spawning grounds were compared to temperature data collected in lower Skookum Creek using correlations. We used the relationship between daily average temperature in the lower creek and the upper creek in 2015 to estimate water temperature on the spawning grounds (y = 0.699x + 0.603, R2 = 0.96, P < 0.01) for the years 2011–2014 and 2016–2017. For the years 2009 and 2010, where lower creek temperature data were not collected, ten years of available air temperature data from the Shelton Airport (Sanderson Field, located 16.1 km NNE of our study area) were compared to the Squaxin Tribe's water temperature data (y = 0.527x + 1.417, R2 = 0.84, P < 0.01) to estimate water temperature on the spawning grounds for those two missing years, as was shown in Smith (1972) and Stefan and Preud′homme (1993).
Air Temperature—Air temperature, while strongly related to proximate water temperature in this and other studies (Stefan and Preud′homme 1993), was used for analysis as a separate environmental variable. In addition to partially driving changes in water temperature, periods of prolonged warm or cold air temperatures can result in inter-annual shifts of environmental responses, such as the onset of leaf growth in riparian plants (Ritchie and Nesmith 1991). The presence of such overhanging vegetative cover may play a role in redd site selectivity by spawning salmonids (Mull and Wilzbach 2007). We measured air temperature in Skookum, Kennedy, and McLane creeks with HOBO Pendant temperature loggers in 2015.
Atmospheric Pressure—Changes in atmospheric pressure are associated with observable biological responses in species across the animal kingdom, including fishes (Daskalov 1999, Pellegrino et al. 2013, Cooper et al. 2023). In salmonid populations, movement patterns are often correlated with shifting atmospheric pressure. Hatchery-reared Rainbow Trout (O. mykiss) in Wyoming expressed increased activity during periods of dropping pressure (Peterson 1972), and Rainbow Trout in New Zealand showed increased upstream spawning migration movements at the ends of both rises and falls in pressure (Dedual and Jowett 1999). Atmospheric pressure data for this study were obtained from the Shelton Airport and compared to timing of CCT spawning in Skookum, Kennedy, and McLane creeks in 2015. Specifically, values of daily mean sea level pressure and of total daily pressure change were used for analysis.
Precipitation—All three streams in this study are predominately rain-driven and as such, streamflow values are highly correlated with precipitation data. However, rain events can also drive water temperature fluctuation in both directions (Linsley 1967, Kope and Botsford 1990). Precipitation was therefore included as a separate environmental variable for analysis. Daily total precipitation measurements were obtained from the Shelton Airport and used for comparison with each of three study streams' spawn timing in 2015. Values of daily mean precipitation were used for analysis.
Photoperiod (Day Length)—The role of photoperiod for CCT spawning is unknown. However, day length is an important factor in the reproductive life of other salmonid species, including the closest relative to Cutthroat Trout, Rainbow Trout (Duston and Bromage 1986, Duston and Saunders 1990, Dietze et al. 2020). Photoperiod, defined as the total minutes between sunrise and sunset for each day at the study site's latitude, was obtained from an online day length calculator ( https://www.timeanddate.com/sun/) and compared to spawn timing in each study stream in 2015.
Tide and Moon—The relationship between upstream spawning movements of anadromous salmonids and tidal influence have been well documented (Olson and Quinn 1993). For instance, adult Atlantic salmon (Salmo salar) in Scotland showed upstream movements through estuarine reaches with ebb tides (Groot et al. 1975, Malcolm et al. 2010). However, later movements further upstream into the river were not tidally influenced. Similarly, the phase of the moon was an important predictor of Chinook Salmon (O. tshawytscha) spawning movements in California (Peterson et al. 2017). However, the phase of the moon is directly related to the tides, with full and new moons corresponding with the largest tides, and mid-wax and mid-wane corresponding with the smallest tides. As such, lunar phase is captured in tidal data and only tides estimated from Arcadia Point in Totten Inlet were used for analysis in this study. Specifically, daily tidal coefficient values were used ( www.tides4fishing.com).
Data Analyses
Variability in Spawn Timing—A Kolmogorov–Smirnov (K–S) two-sample test was used to investigate whether the proportional distribution of redd counts by week differed significantly between study streams (i.e., Skookum, Kennedy, and McLane creeks) in 2015 and between years (2009–2017) in Skookum Creek. We applied a Bonferroni correction to α (i.e., α = 0.05/6 = 0.008 and α = 0.05/36 = 0.0014, respectively) as a realized experiment-wise error rate across multiple comparisons. All statistical analyses were conducted using R statistical program version 1.2.5033 (R Development Core Team 2018).
Influence of Environmental Variables on Spawn Timing—Environmental conditions at the time CCT were spawning were compared to the environmental conditions during the entire spawning season (1 February through 15 June) to test if there were significant differences. Environmental conditions can vary considerably on hourly, daily, and weekly scales. However, comparisons between environmental variability and spawn timing were limited by the survey frequency of seven days (survey methodology described above). Redd life, the duration for which a redd can be accurately observed by a surveyor, for anadromous CCT in the study area averages 14 days (Losee et al. 2016), therefore we used a 14-day average of environmental variables when the duration between surveys exceeded redd life. This was achieved by assigning an estimated value of each environmental variable to individual redds. Environmental variables at the time of spawning were estimated by calculating a weekly mean for each variable for the week prior to the date each redd was initially observed. In this way, environmental variables were binned by week and each redd was assigned a value for each environmental variable representative of the conditions during spawning.
Environmental data binned by the week of spawning during the spawning season were not normally distributed (Shapiro–Wilk test, P < 0.05). Therefore, a Mann–Whitney U test was used to compare the estimated median streamflow (m3·s-1), air temperature, water temperature, atmospheric pressure, precipitation, photoperiod, and daily tidal coefficient during spawning with the median conditions across the entire spawning season. This analysis was performed for all environmental variables in all three streams in 2015 and in Skookum Creek in all years (2009–2017).
Effects of Spawn Timing and Environmental Variables on Redd Abundance—In Skookum Creek, a longer time period (2009–2017) of flow and redd count data was available. Here, a Mann–Whitney U test was used to compare the proportional contribution of redds constructed within 0.25 m3·s-1 streamflow bins to the proportional contribution of daily streamflow bins during the season. To investigate the relationship between flow and spawn timing across the entire time series, we used a Pearson correlation coefficient to compare annual mean streamflow and the annual count of redds.
Results
Variability in Spawn Timing
In 2015, a total of 233 redds, 89 live CCT, and zero dead fish were observed across the three creeks (Skookum, Kennedy, and McLane) in South Puget Sound. In 2015, total redd observations were highest in Skookum Creek (n = 146) and lower in Kennedy (n = 38) and McLane creeks (n = 49). In Skookum Creek, between 2009 and 2017, a total of 925 CCT redds were observed. The annual mean redd count was 103 ± 32 (mean ± SD), with 2015 representing the highest total (n = 146) and 2012 the lowest (n = 45).
CCT redds were observed in the index area as early as 1 February and as late as 10 June (Figure 2). Pairwise comparisons of spawning distributions revealed that the timing of CCT redd observations was not significantly different between streams within a year but was different within a stream (Skookum Creek) between years. Specifically, spawn timing of CCT between McLane and Kennedy creeks (K–S test: D = 0.13, P > 0.008), McLane and Skookum creeks (K–S test: D = 0.25, P > 0.008), and Kennedy and Skookum creeks (K–S test: D = 0.18, P > 0.008) in 2015 were not different (Figure 2a). However, significant differences were observed for each pair of study years in Skookum Creek (K–S test: P < 0.0014) except 2009 versus 2012 (K–S test: D = 0.28, P > 0.0014) and 2010 versus 2012 (K–S test: D = 0.24, P > 0.0014; Figure 2b). Across all years in Skookum Creek, the overall mean date by which 50% of redds were observed was 24 March, but annual values varied by more than two months (15 February 2013 versus 21 April 2011; Figure 2b). The observed CCT spawning period ranged from a minimum of 71 days in 2010 to a maximum of 123 days in 2017 (averaging 96.4 ± 19.3 days from 2009 to 2017).
Environmental Variables
Comparisons of environmental variables at the time of spawning to the environmental conditions during the entire spawning season (1 February through 15 June) in 2015 revealed that streamflow conditions at the time of spawning were significantly different than average streamflow conditions across the spawning season in Skookum, Kennedy, and McLane creeks (Mann–Whitney U-test: U = 6,643, 1,571, and 1,921, respectively, P < 0.05). In contrast, weekly means of water temperature, air temperature, atmospheric pressure, precipitation, photoperiod, and tidal coefficient during the time of spawning were not significantly different from the average of these environmental variables throughout the spawning season for any study stream (Mann–Whitney U test: P > 0.05; Figure 3).
Redd Abundance
In Skookum Creek, instantaneous streamflow at the time of spawning was significantly different from general streamflow during the spawning season in all years except 2012 (Mann–Whitney U test: U = 2,663–6,300, P < 0.05; Figure 4). When grouped into 0.25 m3·s-1 bins by streamflow in the days leading up to spawning, CCT redds occurred in conditions that were disproportionately represented during the season (Mann–Whitney U test: U = 131, P < 0.05). Specifically, CCT showed a preference toward 0.25–1.00 m3·s-1 streamflows in Skookum Creek (Figure 5). These streamflows represented spawning conditions for 84.8% (n = 717/846; Figure 5) of redds observed, however these conditions were present only 66.6% of the days during the spawning season (1 February to 15 June) because average streamflows were higher. Annual mean streamflow during the spawning season in Skookum Creek was negatively correlated to the total number of redds counted (Pearson correlation, r = 0.67, P < 0.05; Figure 6). Consistent with results in 2015 across the three study streams, water temperatures were relatively stable in Skookum Creek during the spawning period in all years. Specifically, mean weekly water temperatures when spawning activity was observed ranged from 6.8 to 11.7 °C and varied by less than 1.0 °C between consecutive weeks.
Discussion
Work presented here provides novel insight into environmental factors driving variability in spawn timing for CCT. Previously, it was shown that anadromous CCT have a more protracted and variable spawn timing relative to other Pacific salmon in South Puget Sound tributaries (less than two months; Agha et al. 2021) and the fluvial form of CCT in other parts of Washington State (McMillan et al. 2014, Vadas et al. 2016), spawning over a six-month time period with peak spawning ranging from early March to late April (Losee et al. 2016). However, factors driving this broad and variable spawn timing were previously unknown. Results from our study build on previous findings and reveal that fluctuations in streamflow are a primary determinant of spawn timing, leading to a high degree of interannual variability and coherence within a year for streams in close geographic proximity to one another.
Cutthroat Trout are typically spring spawners across their range (Trotter 1989, Gresswell 1995, Saiget et al. 2007, Trotter 2008, Muhlfeld et al. 2009, Bennett et al. 2014) but also exhibit a high degree of variability within the spawning period. For instance, CCT have shown unique flexibility with observations of spawning from December to June (Trotter 1989, Saiget et al. 2007, Losee et al. 2018) and even in fall months in unique instances for landlocked CCT (McMillan et al. 2014). The current study showed that this plasticity may be driven, to a large extent, by variability in streamflow. The relationship between streamflow and redd construction could be a result of limitations associated with CCT body size that determine swimming capabilities and redd site selection (e.g., depth, sediment size), as was revealed for similarly sized Brown Trout (Salmo trutta) and Brook Trout (Salvelinus fontinalis) (Witzel and Maccrimmon 1983). The maximum body length observed for anadromous CCT of approximately 50 cm (Losee et al. 2024) allows them to occupy a unique ecological niche in marine and freshwater ecosystems that likely reduces competition, predation, and hybridization (Buehrens et al. 2013, Quinn 2018, Nevoux et al. 2019). However, this small body size may restrict successful spawning to a relatively narrow window of suitable streamflows. While conclusions regarding the selective pressures driving spawn timing are speculative and need further investigation, the majority of CCT redds observed were constructed at 0.25–1.0 m3·s-1 streamflows. This preference is supported by the average streamflow of 0.60 m3·s-1 measured at the site of redds in Skookum Creek (Losee et al. 2016).
Figure 3.
Number of anadromous Coastal Cutthroat Trout redds compared to environmental variables in Skookum, Kennedy, and McLane creeks in 2015. See Methods section for in-depth explanation of environmental variables.
In contrast to previous studies, streamflow was the only variable that consistently represented an important predictor of spawn timing for CCT in South Puget Sound. Previous work has demonstrated a suite of environmental variables affecting spawn timing of salmonids including temperature, streamflow, and photoperiod (Trotter 1989, Quinn 2018, Nevoux et al. 2019). For other subspecies of Cutthroat Trout, where water temperatures are highly variable, temperature was the leading determinant of spawn timing (De Staso and Rahel 1994, Muhlfeld et al. 2009, Budy et al. 2012). For instance, Bennet et al. (2014) showed that mean daily maximum water temperature during the early weeks of the year was the best predictor of temporal variability in spawn timing in a Bonneville Cutthroat Trout (O. clarkii utah) population spawning in a tributary of the Logan River, Utah. Bennet et al. (2014) also observed Bonneville Cutthroat Trout populations beginning and ending spawning earlier in years of warmer water temperatures with optimal spawning occurring between 10–16 °C.
Figure 4.
Number of anadromous Coastal Cutthroat Trout redds compared to flow (m3·s-1) and water temperature (°C) in Skookum Creek in 2009–2017. Lines: red = Coastal Cutthroat Trout redds, blue = flow, and green = water temperature.
In our study, CCT were observed spawning when air temperatures ranged from 3.6 to 13.8 °C but variability in water temperature was low between consecutive weeks (< 1.0 °C) and remained within the optimal range of spawning throughout the season in all years (Quinn 2018). As a result, average weekly water temperatures across the season were not different than average water temperatures at the time of spawning. Water temperatures outside the preferred range were observed before and after the spawning season and could trigger the onset and end of the spawning period. Therefore, water temperature may be an important factor to consider for CCT and could shape broader temporal patterns of spawning in a way that was not detected in the current study. For instance, Vadas et al. (2016) documented a significant interaction between water availability and air/water temperature. A complete understanding of the complex interactions between water temperature and spawning and other environmental variables in marine and freshwater habitats is beyond the scope of this study, but should be investigated further given the life history diversity, complex migration phenology, and broad geographical range of CCT. Future work on CCT in areas with more extreme temperature profiles and featuring smaller intervals between redd surveys (i.e., daily) or controlled lab studies could detect brief periods of spawning activity in association with fine-scale environmental shifts, such as sudden changes in air and water temperature, and allow for the application of more complex analytical tools that account for temporal dependency.
The relationship between streamflow and total number of redds as reported here has implications for the long-term persistence of CCT, given the expectation for a continued changing climate. Recent work forecasting climate change in the Pacific Northwest suggests that areas occupied by anadromous CCT will experience more intense storms, increased flooding, and extended periods of low summertime streamflows (IPCC 2007, Mantua et al. 2010, Williams et al. 2015). Anticipated climate changes will likely negatively affect anadromous CCT based on our results demonstrating the negative relationship between streamflow and redd counts (Figure 6) as well as the role temperature may play in determining the duration of the spawning season. Together, this new information about the influences of temperature on spawning and streamflow on redd counts provides habitat improvement practitioners insight into the potential for increased habitat complexity to enhance adult abundance. Notably, our results do not consider the complex secondary effects that environmental variability may have on salmonids outside the spawning period such as scour, redd dewatering, and changes in conductivity, salinity, and turbidity (Ames and Beecher 2001, Quinn 2018).
Figure 5.
Proportional contribution of weekly mean streamflow discharge (0.25 m3·s-1 bins) to Coastal Cutthroat Trout spawning activity (red bars) and during duration of the spawning season (grey bars) on Skookum Creek, South Puget Sound, Washington in years 2009–2017.
Figure 6.
Relationship between annual mean streamflow (discharge) in Skookum Creek, South Puget Sound, Washington and total number of redds counted annually (2009–2017, r = 0.67). Dashed lines indicate 95% confidence intervals.
CCT have shown resilience in a changing environment(McMillanetal.2014);itistherefore likely that CCT can respond to future changes in environmental conditions with changes in behavior, phenology, and status such as spawning behavior, distribution, and abundance. Indeed, a landlocked population of CCT in Olympic National Park delayed the onset of spawning in recent years due to limited water supply (Vadas et al. 2024). Results from the current study highlight the ecological significance of dynamic flow regimes in fluvial environments which have been described previously (Poff et al. 1997, Vadas et al. 2016) and suggest that extensive periods of high streamflow and flooding during winter could result in the delayed onset of spawning, fewer redds, and associated lower abundance of anadromous CCT. When coupled with early summer drought, high temperatures, and associated secondary effects, an overall contraction of the period of suitable spawning and reduced abundance of adult anadromous CCT should be expected. With a need for additional work on this topic and increasing evidence that a changing climate will likely disproportionally affect organisms that specialize in a unique niche (Johnson et al. 1999, Williams et al. 2015), anadromous CCT should be managed in a way that maximizes adult abundance and life history diversity and protects CCT during a dynamic and protracted spawning and migration period. Tools available to achieve these management goals include conservative fisheries regulations like catch and release at all points in the life cycle, closure of fisheries during times when cutthroat are most vulnerable, habitat restoration actions that increase stream complexity, and other innovative approaches that need to be explored.
Acknowledgements
This work would not have been possible without the support of the Squaxin Island Indian Tribe, leadership and insight of the late Bill Young, technical support and dedication in the field by Erica Marbet, Joe Puhn, Hal Beecher, Steve Boessow, John Rohr, Bill Young, Jack Havens, Greg Shimek, and Brad Caldwell, and logistical support from Hans and Julie Cooke. We also appreciate Andrew Claiborne, Jessica Helsley, Dr. Robert Vadas Jr., and three anonymous reviewers for reviewing an earlier version of this manuscript. Funding was provided by the Washington Department of Fish and Wildlife, the Coastal Cutthroat Coalition, and The Wild Salmon Center.
© 2025 The Authors. This open access article is licensed under a Creative Commons Attribution CC-BY-NC-ND 4.0 International License ( https://creativecommons.org/licenses/by-nc-nd/4.0/).
Statement of Contributions
JPL conceived and designed the study, assisted in data acquisition, analysis, and interpretation, and contributed to drafting and revising the manuscript. RF conceived and designed the study, assisted in data acquisition, analysis, and interpretation, and contributed to drafting the manuscript. GM assisted in data acquisition, analysis, and interpretation, and contributed to drafting the manuscript.
English (US) ·