Coastal cutthroat trout (Oncorhynchus clarkii clarkii) are found in many streams in coastal North America from Southeast Alaska to Northern California, and nearly all Puget Sound streams with access to anadromous fish (Trotter, 1989). Cutthroat trout often favor small streams, including those rarely occupied by other salmonids (Rosenfeld et al., 2000; Matzen and Berge, 2008), and they have several advantages that allow them to thrive in smaller streams than other salmonids. First, their smaller mature body size (ca. 20 cm fork length) compared to other anadromous salmonids, allows for holding, feeding, and spawning in shallow streams. Second, cutthroat trout also have a diverse and flexible life history that may include a mix of adfluvial, fluvial, and anadromous fish in a single watershed (Trotter, 1989; Zydlewski et al., 2009). Finally, cutthroat trout have a diverse diet, ranging from instream and terrestrial insects to piscivory, that allow them to adjust to seasonal changes in environmental conditions (Nowak et al., 2004; Romero et al., 2005). Following spring spawning, juveniles emerge in the late spring and early summer, and may spend 1–5 years in freshwater prior to smolting, while others may remain resident in freshwater (Trotter, 1989). Those that smolt tend to reside in estuarine and nearshore habitats while becoming piscivorous. Coastal cutthroat trout are iteroparous, with mature and immature individuals returning to freshwater annually to spawn, find overwinter refuge, or forage.

Although cutthroat trout can often be found in smaller ephemeral or intermittent streams, they are also viewed as an indicator of stream health, as they may be sensitive to water quality or other habitat issues (Scott et al., 1986), including the 6PPD-quinone that can be prevalent in urban runoff and cause high mortality in some salmonids (Shankar et al., 2025). Despite their potential sensitivity, some urban streams have extensive cutthroat trout populations (DeGasperi et al., 2009; Silver et al., 2018; Ostberg et al., 2024). Although there is substantial documentation of cutthroat trout occurrence in small urban streams, without tagging and tracking studies or serial mark and recapture, the temporal extent of their use of urban streams is largely unknown. Adult cutthroat trout may regularly spawn in streams that function as poor-quality habitats or demographic sinks, where little or intermittent juvenile survival results in a net population loss.

There are several common metrics for assessing fish performance in stream habitats. From single capture events, fish density and size at age (length or mass compared to age from scales or otoliths) or condition factor (e.g., Fulton’s K) are often used to assess how rapidly fish are growing or recent feeding success (Bolger and Connolly, 1989). Others have used biochemical indicators of recent growth like the insulin-like growth factor-1 (Beckman, 2011; Bond et al., 2014). Serial sampling of marked individuals allows for the calculation of individual growth rates in length or mass, often expressed as specific growth rate (Lugert et al., 2016). However, any growth or condition indices require comparative context to provide meaningful information for researchers or managers. That is, comparison of performance indices among the species from different populations or watersheds can provide valuable insights into the health of any single system among the spectrum of habitats available. Growth limitations in any watershed may stem not only from poor habitat or water quality, but from fundamental bioenergetics constraints. Factors such as the thermal environment, prey quality, intraspecific competition, and foraging trade-offs (e.g., predator avoidance) all converge to determine the net energy available for fish growth.

Here, we compare several performance metrics of juvenile cutthroat trout growth (e.g., condition, size at age, year-over-year growth rate) of individuals captured and tagged in a very urbanized watershed in Puget Sound, Pipers Creek, with similar indices from relatively undeveloped Deep, East Twin, and West Twin Creeks in nearby Strait of Juan de Fuca. Although there are no true control streams for this type of comparative assessment, streams that are geographically and climatically similar can still provide a useful context for the benefits and condition of the highly altered urban landscape.

We also took a bioenergetic accounting approach to comparing growth between Pipers Creek and Strait of Juan de Fuca streams (Urabe et al., 2010; Deslauriers et al., 2017). In bioenergetic terms, fish growth is the remaining resources from the consumption of prey minus the energetic costs of respiration and waste. Respiration in fish is driven by three primary components: temperature, body size, and activity level (Kitchell et al., 1977). Because cutthroat trout are poikilothermic, their internal body temperature and resulting metabolic rate are largely driven by temperature of the water they inhabit. However, there are allometric effects of body size on metabolic rate such that the metabolic rate for each gram of tissue decreases with increasing body size. Finally, movement increases metabolic rate, thus occupancy in high velocity habitats and migration require more energy that residence in lentic habitats. For each species and life stage there is a maximum consumption rate that individuals can achieve under conditions of essentially unlimited food availability that depends on prey handling time, digestion and gut evacuation rate. The physiologically maximum consumption rate for a fish of a specific size and thermal experience is called Cmax. Therefore, with laboratory derived species and size-specific metabolic parameters and temperature experienced by an individual, we can estimate the average proportion of Cmax (P_Cmax) that must have been achieved to result in a known change in body size between two measurement events. In this sense, P_Cmax represents how well fish were acquiring resources relative to maximum possible for a given thermal environment, and allows us to make comparisons of growth while controlling somewhat for differences in water temperature among habitats.

Differences in growth among streams could be due to several factors like temperature affecting bioenergetics (Elliott et al., 1995), as well as food availability and threat of predation (Billman et al., 2011), both of which may be widely mediated by the density of conspecifics and other species alike (Sabo and Pauley, 1997). Therefore, density of competitors may be a key driver in limiting growth potential (Jenkins Jr. et al., 1999), yet density is difficult to estimate for stream salmonids. While electrofishing is a standard tool for estimating stream fish abundance, methodological differences in both field practice (e.g., single vs. multiple passes) and statistical analyses (e.g., total catch vs. depletion modeling) complicates comparisons among studies. Consequently, even when field designs appear similar, researchers must exercise caution when comparing density data across studies. In contrast, mark-and-recapture datasets offer more consistency, some information is inherently comparable, such as year-over-year recaptures and the resulting growth information.

Streams heavily impacted by the effects of urbanization may have reduced species richness and invertebrate diversity (Jones and Clark, 1987), poor water quality from run-off, or be susceptible to warm summer temperatures as a result of degraded riparian areas. We hypothesize that these factors may result in reduced growth rates and size at age cutthroat trout rearing in urbanized streams, after controlling for temperature and body size, when compared to nearby streams with a more natural landscape. Here, we will use comparable factors from both stream types to assess fish growth as a key performance metric.

In Pipers Creek and its tributaries, 1239 cutthroat trout were sampled over the four sampling years (2006-2009) at six sampling reaches. Although we classified all trout sampled as cutthroat trout in Pipers Creek, though only 3% (38 fish) of those sampled were< 60 mm fork length, the threshold that excluded smaller unknown trout in Strait streams. In Pipers Creek cutthroat trout comprised nearly all the salmonid abundance (95.7%), with coho salmon and chum fry comprising 2.5% (33 fish) and 1.7% (23 fish) of total abundance, respectively. In stark contrast to Pipers Creek, cutthroat trout (> 60mm fork length) in the Strait of Juan de Fuca streams comprise only 1% (1,949 fish) of all salmonids sampled (174,928), with the remainder composed of coho salmon (49.7%, 87,010 fish), rainbow trout (21.3%, 37,359 fish), and trout fry (27.7%, 48,610 fish).

Although the difference in sampling methodology makes statistical comparison of fish densities between streams impractical, a descriptive comparison between the two system reveals that among all salmonids, the densities are broadly similar among Strait of Juan de Fuca (0.701 salmonids·m^-², 0.32 SD) and Pipers Creek reaches (0.733 salmonids·m^-², 0.12 SD). That said, data from single-pass electrofishing in the Strait streams should be viewed as more of a minimum density since it is unlikely all fish were captured in one pass.

Fish length varied significantly by river (Kruskal-Wallis χ² = 385.9, df = 3, p< 0.001). Post-hoc pairwise comparisons using Dunn’s test revealed that cutthroat trout in the Pipers Creek were significantly smaller than those in all other locations (all comparisons p< 0.001). In contrast, no significant differences in length distributions were detected among the three Strait streams: Deep Creek, East Twin, and West Twin (all p > 0.43; Figure 4). Although there is considerable overlap in the size distribution, Pipers fish also had a smaller maximum size indicating that cutthroat trout may not remain resident in Pipers Creek for as long as the Strait streams and may initiate anadromous behavior once they reach ca. 200 mm fork length. In contrast, cutthroat trout in the Strait streams exhibited both a larger median and maximum size with some individuals > 300 mm fork length.

Although our analysis of relative condition factor (Kn) found statistically significant differences among cutthroat trout from nearly all streams (Kruskal-Wallis χ² = 186.71, df = 3, p< 0.001; Figure 5), the magnitude of these differences was small. Median Kn values ranged from 0.973 to 1.03 (Deep Creek = 0.983, East Twin = 0.995, West Twin = 0.973, Pipers Creek = 1.03). Pairwise comparisons confirmed significant differences between almost all locations (Dunn’s test, p< 0.01), with the exception of Deep Creek and West Twin (p = 0.29). All told, cutthroat trout from Pipers Creek were typically 3% heavier for their length than expected from the fitted relationship of all sampled fish.

We calculated annual specific growth rate for all PIT tagged fish that were recaptured after approximately one year at liberty. In Pipers Creek, there were 39 recapture events from 35 individuals indicating at least some likely extended residence time and juvenile survival despite the small creek size. Four individuals were captured in two sequential years following tagging. In the Strait, there were 21 recapture events, though no individual was recaptured more than once. In Strait streams, 20 of 21 cutthroat trout were recaptured in the same habitat unit where they were initially tagged. However, despite the site fidelity of a few individuals, 20 same-site recaptures of 1,750 tagged cutthroat trout, a 1.2% recapture rate may indicate that there is considerable movement, and fish that stay in the same reach year-over-year are relatively rare. As expected, specific growth rate (percent change per day) declined markedly with increasing initial body size (Figure 6), regardless of tagging stream. To compare growth among systems while accounting for large variation in body size, we leveraged the large time (ca. one year) between capture events to estimate annual growth and von Bertalanffy growth curves for each population. We estimated the asymptotic maximum fork length (L_inf) to be 245 mm (95% CI: 206, 406) in Pipers Creek and 316 mm (95% CI: 205, 1,020) in Strait streams, not surprising given the larger fish captured in Strait streams. In the von Bertalanffy comparison of growth, the important parameter is K, which determines how quickly fish reach their asymptotic size, and thus, the growth rate. The value of K for Pipers cutthroat trout was 0.313 (95% CI: 0.141, 0.468) tended to be larger than that of Strait at 0.196 (95% CI: 0.009, 0.446) (Figure 7). That said, our analysis of K and L_inf from 1,000 bootstrapped samples indicated that the 95% confidence interval in the difference between both parameters from the two regions spanned zero, indicating no discernable difference in size at age. Similarly, median difference in expected growth of a 60 mm fork length was 5.9 mm longer in Pipers Creek than the Strait regions. However, the 95% confidence interval in growth also encompassed zero. So, although Pipers fish may grow slightly faster, we could not find a statistically significant difference between the two populations.

Finally, for Pipers Creek fish, we used our bioenergetic analysis to compare realized annual growth to the growth that would be expected if fish were feeding at their size-specific maximum ration, C_max. We estimated the proportion of C_max, as P_Cmax using the temperature in each stream during days at liberty between captures. We found that, on average, Pipers fish were feeding at a P_Cmax of 0.41 (SD = 0.035), while Strait fish were feeding at an average P_Cmax of 0.46 (SD = 0.045). For Pipers Creek fish, a larger number of recaptures were initially sampled, and recaptured on the same date in successive years. For this subset of fish that all experienced the same temperature profile between measurement events, we also estimated the growth potential at a wide variety of P_Cmax to determine how much growth might have been realized if feeding rates were higher or lower. In general, we found that with elevated P_Cmax fish could have achieved much higher growth (Figure 8) with the same temperature and prey energy density.

Our expectation of cutthroat trout performance was low for Pipers Creek; a small urbanized watershed with potential exposure to contaminated runoff and concerns about temperatures, dissolved oxygen, eutrophication, sedimentation, and riparian disconnection. In addition, Pipers Creek has been classified as having a poor benthic index of biological integrity (Morley et al., 2010). It is now recognized that cutthroat trout are very susceptible to some of the same chemicals that have been well documented to cause mortality events in coho salmon in similarly urbanized environments, such as Seattle’s Thornton Creek (McIntyre et al., 2021; Tian et al., 2021; Shankar et al., 2025). Cutthroat trout are also often viewed as sensitive species, with their presence serving as an indicator of high ecological integrity and elevated water quality (Silver et al., 2018). Paradoxically, studies in Puget Sound (though not in Pipers Creek or the Strait) have found that the proportion of cutthroat trout may increase as streams become more urbanized, potentially indicating a more robust tolerance to the aforementioned stream impairments commonly associated with increased urbanization (Matzen and Berge, 2008). Cutthroat trout are also highly mobile and facultatively anadromous; individuals may be able to avoid poor stream conditions by moving out of creeks and into marine habitats (Northcote, 1997; Zydlewski et al., 2009; Losee et al., 2018). Despite these issues, cutthroat trout were by far the most common fish found in Pipers Creek, and the small size of many individuals indicates many were likely spawned locally. Additionally, although recapture rates were low, year-over-year recaptures indicated that at least some fish were likely remaining in the stream year-round.

Perhaps the most striking biological difference between the two regions is that Pipers Creek is a cutthroat trout dominated system, while Strait streams are largely coho and rainbow/steelhead trout systems (Hall et al., 2016). The caveat is that smaller (< 60 mm fork length) trout in Strait streams may be either rainbow trout or cutthroat trout and cannot be identified in the field (Kennedy et al., 2009). Unfortunately, no comparable genetic analysis discerning the two species has been conducted on the Strait streams. However, over a hundred steelhead females may spawn in each of the Strait streams each year, producing ≥ 5,000 eggs each, while the small body size of cutthroat trout may limit them to an order of magnitude fewer (ca. 200-500, Scott and Crossman (1973)). Other salmonids may avoid Pipers Creek due to its small size and distinct lack of habitat complexity; coho salmon require deep pools and off-channel habitats for overwinter survival (Peterson, 1982; Bisson et al., 1988; Quinn and Peterson, 1996). The lack of instream complexity in Pipers Creek likely also limits steelhead that, as larger bodied sea-run adults, may benefit from larger deeper pool habitats for spawning, and rearing in freshwater for up to three years before seaward migration (Quinn, 2018). Similarly, the Strait streams may have a more fluvial population of cutthroat trout in addition to anadromous individuals, while Pipers Creek may be comprised of anadromous individuals that have extensive growth and maturation outside of the creek environment. Sampling during the peak spawning months of January-March, may help determine the portion of the populations that are comprised of larger bodied individuals that may enter freshwater to spawn and return rapidly to Puget Sound while remaining undetected in summer sampling. Overall, the reduced species diversity in Pipers Creek is not surprising given its much smaller watershed, which has also been truncated by urban development and the disconnection of much of its watershed through stormwater routing away from the creek (Terui et al., 2021).

Despite the limitations in statistically comparing fish densities among the streams, the overall numerical density of all salmonids in mid-summer was broadly similar among the two streams. Therefore, the competition for resources may stem more from interspecific competition in the Strait streams, while intraspecific competition dominates in Pipers Creek. The result is that, although we observed a conserved length-weight relationship among the regions, Pipers Creek fish tended to be slightly heavier for their length. This may result from the lack of interspecific competition in Pipers Creek, and despite the intermediate benthic invertebrate indices, fish are readily obtaining prey resources, perhaps more so than Strait cutthroat trout.

Although Pipers Creek cutthroat trout were slightly heavier for their size, two different methods of growth rate estimation between the regions indicate no discernible difference between the creeks. Part of this may be due to the small sample of year-over-year recaptures. A larger sample size may have been able to separate more subtle differences between the regions that our bootstrapping did not reveal. However, despite the small sample size, the growth experienced by cutthroat trout in both regions is at least broadly similar. In an earlier study of cutthroat trout growth in urban and rural streams, Scott et al. (1986) found that fish grew comparatively faster in the more urbanized watershed. Our size at age analysis also indicated that although Strait cutthroat trout reach a larger maximum size, they achieve that size by remaining in the creeks longer rather than growing faster than Pipers creek fish. The average annual P_Cmax of ca. 0.4 in both streams indicates that growth in both systems is likely limited by availability of prey, although whether production or competition limits that access, or whether the mechanism is similar among the streams, remains unknown (Urabe et al., 2010; Roon et al., 2025). What we can say is that maximum temperatures do not currently limit growth in either system, though extended periods of cold temperatures likely limit both fish growth and prey production in winter months. Overall, cutthroat trout appear to be well suited to urban streams compared to other salmonids. In particular, their ability to spawn in small streams during high instream flow months (ca. Jan-March) when streams are warming, while avoiding stream entry during the very low flow periods and the difficulties that larger bodied species (e.g., coho salmon) may encounter spawning in very small streams, while also avoiding the worst of winter floods that may disturb redds. In addition, many urban streams have been simplified to remove wood and off channel habitats that may threaten urban infrastructure. While coho salmon have suffered under such conditions, cutthroat trout may be better suited to deal with simpler channel forms.

In this study, we employed multiple analytical approaches to rigorously compare the growth of coastal cutthroat trout between a small, urbanized stream and several larger, intact forested watersheds that are part of the long-standing Intensively Monitored Watershed (IMW) program. Despite the stark differences in the landscape context, our analyses did not discern a notable difference in growth rates between the populations. Ultimately, this finding may reveal as much about the remarkable plasticity and resilience of cutthroat trout to exploit marginal habitats as it does about the quality of the urban habitat itself. The ability of cutthroat trout to maintain comparable growth in a highly modified system underscores its adaptability and highlights the challenge of assessing habitat quality for a generalist species. Studies of this kind, which rely on parameter estimation and the detection of subtle population responses, could not be completed without the foundational datasets provided by these IMW efforts. Continued investment in long-term monitoring is essential for generating the sustained, high-quality data necessary to accurately model resilient species.