Projected impacts of climate change on stream salmonids with implications for resilience‐based management

The sustainability of freshwater fisheries is increasingly affected by climate warming, habitat alteration, invasive species and other drivers of global change. The State of Michigan, USA, contains ecologically, socioeconomically valuable coldwater stream salmonid fisheries that are highly susceptible to these ecological alterations. Thus, there is a need for future management approaches that promote resilient stream ecosystems that absorb change amidst disturbances. Fisheries professionals in Michigan are responding to this need by designing a comprehensive management plan for stream brook charr (Salvelinus fontinalis), brown trout (Salmo trutta) and rainbow trout (Oncorhynchus mykiss) populations. To assist in developing such a plan, we used stream‐specific regression models to forecast thermal habitat suitability in streams throughout Michigan from 2006 to 2056 under different predicted climate change scenarios. As baseflow index (i.e., relative groundwater input) increased, stream thermal sensitivity (i.e., relative susceptibility to temperature change) decreased. Thus, the magnitude of temperature warming and frequency of thermal habitat degradation were lowest in streams with the highest baseflow indices. Thermal habitats were most suitable in rainbow trout streams as this species has a wider temperature range for growth (12.0–22.5 °C) compared to brook charr (11.0–20.5 °C) and brown trout (12.0–20.0 °C). Our study promotes resilience‐based salmonid management by providing a methodology for stream temperature and thermal habitat suitability prediction. Fisheries professionals can use this approach to protect coldwater habitats and drivers of stream cooling and ultimately conserve resilient salmonid populations amidst global change.     

Efficacy of a polyvalent injectable vaccine against Flavobacterium psychrophilum administered to rainbow trout (Oncorhynchus mykiss L.)

Flavobacterium psychrophilum is one of the most important pathogens affecting cultured rainbow trout (Oncorhynchus mykiss). Recent information from UK salmonid farms showed country‐wide distribution of genetically and serologically divergent clones, which has hampered the development of a vaccine for rainbow trout fry syndrome. The current study assessed the efficacy of an injectable polyvalent vaccine containing formalin‐inactivated F. psychrophilum in rainbow trout. The vaccine was formulated with an oil adjuvant (Montanide ISA 760VG) or formalin‐killed cells alone. Duplicate groups of trout (60 ± 13 g) were given phosphate‐buffered saline or vaccine formulated with Montanide by intra‐peritoneal (i.p.) injection and challenged by intra‐muscular (i.m.) injection with a homologous and a heterologous isolate of F. psychrophilum at 525 degree days post‐vaccination (dd pv). Significant protection was achieved in vaccinated fish (p = 0.0001, RPS 76% homologous, 88% heterologous). Efficacy of the adjuvanted vaccine was also demonstrated by heterologous challenge at 1155 dd pv resulting in 100% protection, whereas survival in the un‐adjuvanted group was not significantly different from control fish. Levels of specific antibody at 1155 dd pv, as measured by ELISA, were significantly higher in the fish vaccinated with adjuvant when compared with unvaccinated fish.     

Are brown trout Salmo trutta fario and rainbow trout Oncorhynchus mykiss two of a kind? A comparative study of salmonids to temperature‐influenced Tetracapsuloides bryosalmonae infection

Proliferative kidney disease (PKD) of salmonids caused by Tetracapsuloides bryosalmonae causes high mortalities of wild brown trout (Salmo trutta fario) and farmed rainbow trout (Oncorhynchus mykiss) at elevated water temperatures. Here the aim was to compare the temperature‐dependent modulation of T. bryosalmonae in the two salmonid host species, which display different temperature optima. We used a novel experimental set‐up in which we exposed brown trout and rainbow trout to an identical quantified low concentration of T. bryosalmonae for a short time period (1 hr). We followed the development of the parasite in the fish hosts for 70 days. PKD prevalence and parasite kinetics were assessed using qPCR. Exposures were performed at temperatures (12°C and 15°C) that reflect an environmental scenario that may occur in the natural habitat of salmonids. T. bryosalmonae infection was confirmed earliest in brown trout kept at 15°C (day 7 post‐exposure) while, in all other groups, T. bryosalmonae was not confirmed until day 15 post‐exposure. Moreover, significantly greater infection prevalence and a faster increase of parasite intensity were observed in brown trout kept at 15°C than in all other groups. These results indicate that PKD is differentially modulated by water temperature in related host species.     

Distribution and prevalence of T. bryosalmonae in Austria: A first survey of trout from rivers with a shrinking population

The first evidence of proliferative kidney disease (PKD) in an Austrian river (the River Kamp) was documented in 2016, and no information on the PKD infection status of trout in other rivers was available. Since then, brown trout (Salmo trutta fario) and rainbow trout (Oncorhynchus mykiss) have been collected from rivers in Upper and Lower Austria for different diagnostic purposes. In this study, we summarize the recent findings of a first survey concerning the distribution of Tetracapsuloides bryosalmonae, the causative agent of proliferative kidney disease (PKD), from these samples. Between September 2015 and October 2017, a total of 280 brown trout and 39 rainbow trout were collected from 21 rivers in the provinces of Upper and Lower Austria. T. bryosalmonae was detected by PCR of kidney tissue in 17 of 21 sampled rivers and in 138 of 280 brown trout as well as in 11 of 39 rainbow trout. Pathological signs of PKD (e.g., hypertrophy of the kidney) were observed in 33 analysed brown trout and six rainbow trout samples. No correlations between fish infected by T. bryosalmonae and the parameters size and age class, condition factor, geological origin of the streams and distribution within the river course were found, while positively tested fish are significantly increased at sampling sites exceeding water temperatures of 15°C for median periods of 115 days. The prevalence within the affected streams or stream sections is highly variable, and in single rivers, infection rates of up to 90% are confirmed.     

Effects of reduced summer flows on the brook trout population and temperatures of a groundwater‐influenced stream

Withdrawal of water from streams and groundwater is increasing in Midwestern North America and is a potential threat to coldwater fishes. We examined the effects of summer water withdrawals on brook trout Salvelinus fontinalis populations and water warming rates by diverting 50–90% of summer baseflow from a 602‐m treatment zone (TZ) in a groundwater‐influenced Michigan stream during 1991–1998. We compared density of brook trout in fall, and spring‐to‐fall growth and survival of brook trout, between the TZ and an adjacent reference zone (RZ) whose flows were not altered. Flow reductions had no significant effects on the density of brook trout in fall or spring‐to‐fall survival of brook trout. However, spring‐to‐fall growth of brook trout in the TZ declined significantly when 75% flow reductions occurred. Cold upstream temperatures and the relatively short study reach kept thermal habitat conditions excellent for brook trout in the TZ throughout the dewatering experiments. These findings suggest that brook trout can tolerate some seasonal loss of physical habitat if temperature conditions remain suitable. In summer 1999, we experimentally assessed the influence of flow reduction on the warming rate through the TZ by diverting from 0% to 90% of flow around the TZ in 3‐ or 4‐day trials on a randomised schedule. Average daily temperature increased exponentially as stream flows declined from normal summer levels. Our findings suggest the risk of trout habitat loss from dewatering is potentially large and proportional to the magnitude of withdrawal, especially as thermal conditions approach critical levels for trout.     

First detection and isolation of infectious haematopoietic necrosis virus from farmed rainbow trout in Nyeri County,Kenya

Infectious haematopoietic necrosis virus (IHNV) is the causative agent of infectious haematopoietic necrosis, a disease of salmonid responsible for great economic losses. The disease occurs in most parts of the world where rainbow trout is reared but has not been previously reported in Kenya. In this study, rainbow trout fry and growers from two farms in Nyeri County were screened for IHNV. Whole fry (n = 4 from each farm) and kidney samples from growers (n = 15 and n = 6 from the two farms, respectively) were collected and preserved for cell culture examination or PCR analysis. Screening of samples was done by PCR followed by sequencing of the glycoprotein gene of the virus. Demonstration of the virus was done by propagation in EPC cells followed by the indirect fluorescence antibody test (IFAT). The results revealed the presence of IHNV at low prevalence of 0.1 and 0.4 for the two farms. The virus was confirmed both by IFAT and by partial sequencing of the G gene. Phylogenetic analysis revealed that the Kenyan isolates were identical to those of the J genogroup found mostly in Asia. The findings have implications for biosecurity measures and import regulations for the Kenyan rainbow trout industry.     

Experimental evaluation of rainbow trout Oncorhynchus mykiss predation on longnose dace Rhinichthys cataractae

Laboratory and in‐stream enclosure experiments were used to determine whether rainbow trout Oncorhynchus mykiss influence survival of longnose dace Rhinichthys cataractae. In the laboratory, adult rainbow trout preyed on longnose dace in 42% of trials and juvenile rainbow trout did not prey on longnose dace during the first 6 h after rainbow trout introduction. Survival of longnose dace did not differ in the presence of adult rainbow trout previously exposed to active prey and those not previously exposed to active prey ( = 0.28, P = 0.60). In field enclosures, the number of longnose dace decreased at a faster rate in the presence of rainbow trout relative to controls within the first 72 h, but did not differ between moderate and high densities of rainbow trout (F_2,258.9 = 3.73, P = 0.03). Additionally, longnose dace were found in 7% of rainbow trout stomachs after 72 h in enclosures. Rainbow trout acclimated to the stream for longer periods had a greater initial influence on the number of longnose dace remaining in enclosures relative to those acclimated for shorter periods regardless of rainbow trout density treatment (F_4,148.5 = 2.50, P = 0.04). More research is needed to determine how predation rates will change in natural environments, under differing amounts of habitat and food resources and in the context of whole assemblages. However, if rainbow trout are introduced into the habitat of longnose dace, some predation on longnose dace is expected, even when rainbow trout have no previous experience with active prey.     

The effects of swimming exercise and dissolved oxygen on growth performance,fin condition and survival of rainbow trout Oncorhynchus mykiss

Swimming exercise and dissolved oxygen (DO) are important parameters to consider when operating intensive salmonid aquaculture facilities. While previous research has focused on each of these two variables in rainbow trout Oncorhynchus mykiss, studies examining both variables in combination, and their potential interaction, are absent from the scientific literature. Both swimming exercise (usually measured in body lengths per second, or BL/s) and DO can be readily controlled in modern aquaculture systems; therefore, we sought to evaluate the effects of these variables, separately and combined, on several outcomes in rainbow trout including growth performance, fin health and survival. Rainbow trout fry (18 g) were stocked into 12 circular 0.5 m³ tanks, provided with either high (1.5–2 BL/s) or low (approximately 0.5 BL/s) swimming exercise and high (100% saturation) or low (70% saturation) DO, and grown to approximately 1 kg. By the conclusion of the study, higher DO was independently associated with significantly (p < .05) increased growth performance. Significant differences were not noted in other outcomes, namely feed conversion, condition factor and mortality, although caudal and right pectoral fin damage was associated with low oxygen and low swimming exercise treatments respectively. Cardiosomatic index was significantly higher among exercised fish. These results suggest that swimming exercise and DO at saturation during the culture of rainbow trout can be beneficial to producers through improved growth performance and cardiac health.     

Evaluation of brook trout production in a coldwater recycle aquaculture system

Brook trout (Salvelinus fontinalis) are a commercially important coldwater species reared in Wisconsin and the Midwestern United States. Brook trout are raised by private, tribal, state, and federal fish hatchery facilities in Wisconsin. Approximately 10% of private coldwater aquaculture operations are presently raising brook trout of various strains for stocking uses and a limited amount for food markets. Growing brook trout to a larger size, if they can be reared in a shorter time span, may present a potential new sector for the aquaculture market in the Midwestern US. The present study reports hatchery production attributes, i.e., growth, survival, fin condition, feed efficiency, water chemistry requirements and general husbandry of Lake Superior strain (Nipigon) brook trout reared in a recirculating aquaculture system (RAS), operated at an average temperature of 13 °C. The recycle system at NADF reared 1379 kg of brook trout over a 10-month period from fingerling (9 g) to market size (340–454 g). The trout grew faster (0.84 g/day and 0.64 mm/day) in the RAS than fish cultured in traditional flow-through tank culture utilizing ground water at 7.6 °C (0.14 g/day and 0.35 mm/day). Final average weight of RAS fish was 260 g, while the flow-through fish averaged 65 g. Final tank densities for the RAS averaged 40.4 kg/m³ while flow-through tanks averaged 31.2 kg/m³. Throughout the project, feed conversions in the RAS ranged from 0.9 to 1.3. Water quality variables such as TAN, nitrite, DO, temperature, TSS, CO₂, ph, etc. were within safe limits for brook trout and will be discussed. It does appear from this initial research project that market size brook trout can be raised successfully in a recycle system within a similar time frame as a rainbow trout produced in a Wisconsin typical flow-through facility.     

Isolation of bacterial probiotic candidates from the gastrointestinal tract of rainbow trout,Oncorhynchus mykiss (Walbaum), and screening for inhibitory activity against Flavobacterium psychrophilum

In this study, 318 bacterial strains were isolated from the gastrointestinal (GI) tracts of 29 rainbow trout, Oncorhynchus mykiss (Walbaum). These bacteria were screened in vitro for their ability to inhibit growth of Flavobacterium psychrophilum, the causative agent of coldwater disease. Bacteria observed to inhibit F. psychrophilum growth were further screened against rainbow trout bile, as an indicator of their ability to survive in the GI tract. This screening resulted in narrowing the pool to 24 bacterial isolates. Those 24 isolates were then tested for pathogenicity in rainbow trout by intraperitoneal injection. Following a 28‐day challenge, eight isolates were shown to cause direct mortality and were eliminated from further study. As a result, 16 bacterial isolates were identified as probiotic candidates with the potential to control or reduce disease caused by F. psychrophilum.     

Response of resident brown trout, Salmo trutta L., and rainbow trout, Oncorhynchus mykiss (Walbaum), to the stocking of hatchery-reared brown trout

Two strains of hatchery-reared adult brown trout, Salmo trutta L., [208–334 mm total length (TL); n =  591] were individually marked and released into a limestone stream. The estimated survival after one month (86%; n =  508) was comparable to that for resident brown trout and rainbow trout, Oncorhynchus mykiss (Walbaum), (89%; n =  771), but declined to 14% ( n =  83) after 8 months compared with 52% ( n =  451) for resident trout. The movement of resident trout out of stocked stretches was higher (14%) than from control sites (5%), but the population size in both individual sites and the overall study area were unaffected. The growth of resident brown trout was unaffected by stocking, but rainbow trout showed lower growth rates in stocked versus unstocked stretches both one and 8 months after stocking ( P <  0.002).     

Competitive interactions among multiple non‐native salmonids and two populations of Atlantic salmon

Competitive interactions with non‐native species can have negative impacts on the conservation of native species, resulting in chronic stress and reduced survival. Here, juvenile Atlantic salmon (Salmo salar) from two allopatric populations (Sebago and LaHave) that are being used for reintroduction into Lake Ontario were placed into semi‐natural stream tanks with four non‐native salmonid competitors that are established in Ontario streams: brown trout (S. trutta), rainbow trout (Oncorhynchus mykiss), Chinook salmon (O. tshawytscha) and coho salmon (O. kisutch). Brown trout and rainbow trout reduced the survival and fitness‐related traits of Atlantic salmon, whereas Chinook salmon and coho salmon had no impact on these traits. These data support theories on ecological niche overlap and link differences in observed aggression levels with competitive outcomes. Measurements of circulating hormones indicated that the Atlantic salmon were not chronically stressed nor had a change in social status at the 10‐month time point in the semi‐natural stream tanks. Additionally, the Sebago population was better able to coexist with the non‐native salmonids than the LaHave population. Certain populations of Atlantic salmon may thus be more suitable for some environments of the juvenile stream phase for the reintroduction into Lake Ontario.     

Seasonal movement,growth and survival of brook trout in sympatry with brown trout in Midwestern US streams

Seasonal patterns in growth, survival and movement of brook trout Salvelinus fontinalis were monitored in two southeastern Minnesota streams divided into study reaches based on brown trout Salmo trutta abundance. We estimated survival and movement while testing for effects of stream reach and time using a multistrata Cormack–Jolly–Seber model in Program MARK. Multistrata models were analysed for three age groups (age‐0, age‐1 and age‐2+) to estimate apparent survival, capture probability and movement. Survival varied by time period, but not brown trout abundance and was lower during flood events. Age‐0 brook trout emigrated from reaches with low brown trout abundance, whereas adult brook trout emigrated from downstream brown trout‐dominated reaches. Growth was highest in spring and summer and did not differ across streams or reaches for the youngest age classes. For age‐2+ brook trout, however, growth was lower in reaches where brown trout were abundant. Interspecific interactions can be age or size dependent; our results show evidence for adult interactions, but not for age‐0. Our results suggest that brook trout can be limited by both environmental and brown trout interactions that can vary by season and life stage.     

Growth,osmoregulation and energy budget of rainbow and steelhead trout under different salinity acclimation methods and the best transition size of steelhead trout

In order to meet the demand of salmon market, Chinese scientists and entrepreneurs are working on salmon mariculture far offshore in the Yellow Sea, China. Rainbow, steelhead trout and Atlantic salmon were selected as the main culture species. The aims of the present study were as follows: (a) investigate the effect of the salinity acclimation method on the growth, osmoregulation and energy budget in two forms of Oncorhynchus mykiss, rainbow and steelhead trout and (b) explore the optimal size of steelhead trout for the seawater entry. In trial I, rainbow (mean = 99.44 g) and steelhead trout (mean = 99.01 g) were reared for 40 days after undergoing salinity acclimation at three rates: an abrupt increase in salinity from 0 to 30 g/L (T30); an abrupt increase in salinity to 14 g/L, followed by a daily increase of 2 g/L (T2) or 6 g/L (T6) until reaching 30 g/L; and no salinity exposure (control treatment) (T0). In trial II, steelhead trout with body weights of approximately 100 and 400 g were cultured for 60 days with two treatments, T0 and T2, and the specific growth rate (SGR) was calculated every 10 days. In trial I, in both kinds of fish, the optimal growth performance, survival rate, osmoregulation and energy budget were observed in the T0 treatment, followed by the T2 treatment. These results indicate that O. mykiss with a body weight of approximately 100 g can adapt to sea water with a gradual transition (T2), but they are still not suitable for the seawater entry because of low growth. Based on the recorded SGRs in trial II, our formulated regression formula revealed that approximately 200 g is the optimal size of steelhead trout for the transition to sea water.     

Heavy loads of parasitic freshwater pearl mussel (Margaritifera margaritifera L.) larvae impair foraging,activity and dominance performance in juvenile brown trout (Salmo trutta L.)

The life cycle of the endangered freshwater pearl mussel (Margaritifera margaritifera) includes a parasitic larval phase (glochidia) on the gills of a salmonid host. Glochidia encystment has been shown to affect both swimming ability and prey capture success of brown trout (Salmo trutta), which suggests possible fitness consequences for host fish. To further investigate the relationship between glochidia encystment and behavioural parameters in brown trout, pairs (n = 14) of wild‐caught trout (infested vs. uninfested) were allowed to drift feed in large stream aquaria and foraging success, activity, agonistic behaviour and fish coloration were observed. No differences were found between infested and uninfested fish except for in coloration, where infested fish were significantly darker than uninfested fish. Glochidia load per fish varied from one to several hundred glochidia, however, and high loads had significant effects on foraging, activity and behaviour. Trout with high glochidia loads captured less prey, were less active and showed more subordinate behaviour than did fish with lower loads. Heavy glochidia loads therefore may negatively influence host fitness due to reduced competitive ability. These findings have implications not only for management of mussel populations in the streams, but also for captive breeding programmes which perhaps should avoid high infestation rates. Thus, low levels of infestation on host fish which do not affect trout behaviour but maintains mussel populations may be optimal in these cases.     

The effect of competition among three salmonids on dominance and growth during the juvenile life stage

Although non‐native species can sometimes threaten the value of ecosystem services, their presence can contribute to the benefits derived from the environment. In the Great Lakes, non‐native brown trout (Salmo trutta) and rainbow trout (Oncorhynchus mykiss) support substantial recreational fisheries. With current efforts underway to restore once‐native Atlantic salmon (Salmo salar) to Lake Ontario, there is some concern that Atlantic salmon will impede non‐native contributions to the recreational fishery because Atlantic salmon exhibit niche overlap with brown trout and rainbow trout, particularly during the juvenile life stage. We therefore examined competition and growth of juvenile Atlantic salmon, brown trout and rainbow trout in semi‐natural streams. We found that brown trout were the most dominant and had the greatest growth rate regardless of what other species were present. Rainbow trout were more dominant than Atlantic salmon and consumed the most food of the three species. However, in the presence of brown trout, rainbow trout fed less frequently and exhibited negative growth as compared to when the rainbow trout were present with only Atlantic salmon. These data suggest that, outside of density‐dependent effects, Atlantic salmon will not impact stream production of brown trout and rainbow trout.     

Competitive interactions between native and exotic salmonids: a combined field and laboratory demonstration

Abstract –  We studied the impact of two exotic salmonid species (brook trout, Salvelinus fontinalis and rainbow trout, Oncorhynchus mykiss ) on native brown trout ( Salmo trutta fario ) habitat, growth and survival. Habitat selection and vertical distribution between young-of-the-year of the three species were examined in a stream aquarium under different sympatric and allopatric combinations. In addition, similar species combinations were introduced in a Pyrenean mountain stream (southwest France) in order to extend laboratory results to growth and apparent survival. Both laboratory and field results indicated that rainbow trout significantly affected native brown trout habitat selection and apparent survival. On the contrary, brown trout habitat, growth and apparent survival were hardly affected by brook trout. These results support the idea that rainbow trout negatively influence native brown trout, and that competition could influence the outcome of fish biological invasions in freshwater ecosystems.     

Changes in native bull trout and non‐native brook trout distributions in the upper Powder River basin after 20 years,relationships to water temperature and implications of climate change

Many bull trout populations have declined from non‐native brook trout introductions, habitat changes (e.g. water temperature) and other factors. We systematically sampled the distribution of bull trout and brook trout in the upper Powder River basin in Oregon in the 1990s and resampled it in 2013–2015, examined temperature differences in the habitats of the two species and analysed trends in temperatures in the light of possible increases associated with climate change. The species’ distributions are currently similar to those in the 1990s, except in one stream where bull trout declined. However, bull trout consisting of resident forms remain restricted to a few kilometres of habitat at the upper end of fish distribution. In streams where both species occur, the typical pattern was an intermediate zone of mixed bull trout, brook trout, and hybrids downstream of allopatric bull trout and allopatric brook trout extending farther downstream. Temperature differences between where bull trout and most brook trout occurred were small (0.5–1.0°C August mean). There were no statistical increases in water temperatures in nearby streams since the 1990s and no warming trends in air temperatures for the past 25–60 years. However, peak summer water temperatures are occurring about 3 weeks earlier than 25 years ago. Future effects of climate change, including possible increases in temperature, changes in timing and other factors (e.g. snowpack, flow and extreme events) remain a concern for the persistence of these populations. However, it is difficult to precisely predict where those changes will occur and what they will be.     

Myxobolus cerebralis (Hofer) infection risk in native cutthroat trout Oncorhynchus clarkii (Richardson) and its relationships to tributary environments in the Yellowstone Lake Basin

Conservation of native species is challenged by the introduction of non‐native pathogens and diseases into aquatic and terrestrial environments worldwide. In the Yellowstone Lake basin, Yellowstone National Park, the invasive parasite causing salmonid whirling disease Myxobolus cerebralis (Hofer) has been identified as one factor contributing to population declines of native Yellowstone cutthroat trout Oncorhynchus clarkii bouvieri (Jordan & Gilbert). In 2002 and 2003, we examined relationships between the stream environment and severity of M. cerebralis infection in native trout. Coefficients of variation of environmental features were calculated to examine variability. Ten years later, we reassessed infection levels at 22 tributaries broadly across the system. Results of principal component analysis (PCA) of physical features (2003) were negatively correlated with infection severity, mostly in lower jaw cartilage of cutthroat trout, and PCA of chemical features (and temperature) correlated with infection severity in cranial cartilage. Pelican Creek, where M. cerebralis prevalence and severity was high 2002–2003, remained high in 2012. We did not find evidence that the parasite had dispersed further within the system. Variable environmental features (physiological stress) across short spatiotemporal scales within a stream or season may possibly predispose salmonids to infection in the wild and facilitate parasite establishment.     

Relationship between stream temperature, thermal refugia and rainbow trout Oncorhynchus mykiss abundance in arid-land streams in the northwestern United States

Abstract – Warm stream temperatures may effectively limit the distribution and abundance of Pacific salmon Oncorhynchus spp. in streams. The role of cold thermal refugia created by upwelling groundwater in mediating this effect has been hypothesized but not quantitatively described. Between June 21 and September 15, 1994, rainbow trout O. mykiss abundance within 12 northeast Oregon (USA) stream reaches was inversely correlated with mean ambient maximum stream temperatures ( r =−0.7, P <0.05). Some rainbow trout used thermal refugia (1–10 m² surface area) that were on average 3–8°C colder than ambient stream temperatures. Within the warmest reaches, high ambient stream temperatures (>22°C) persisted from mid-June through August, and on average 10–40% of rainbow trout were observed within thermal refugia during periods of midday maximum stream temperatures. Frequency of cold-water patches within reaches was not significantly associated with rainbow trout density after accounting for the influence of ambient stream temperature ( P =0.06; extra sum of squares F -test). Given prolonged high ambient stream temperatures in some reaches, the thermal refugia available in the streams we examined may be too small and too infrequent to sustain high densities of rainbow trout. However, these refugia could allow some rainbow trout to persist, although at low densities, in warm stream reaches.