Washington Streams Research Articles

Aquatic and riparian ecosystem recovery from debris flows in two western Washington streams, USA.

An exceptionally powerful storm struck southwestern Washington in December 2007 causing large debris flows in two adjacent streams. The two affected streams had been studied prior to the storm, providing a rare opportunity to examine ecosystem recovery. We monitored the streams and their riparian zones for six years after the disturbances to determine whether recovery rates of biota, physical habitat, and water temperature differed, and if so, what factors affected resilience. Along both streams, the debris flows removed wide swaths of soil, rock, and coniferous riparian forests, widening the active channel and increasing solar exposure and summer water temperatures. Initially depauperate of vegetation, after four years red alder trees dominated the riparian plant communities. The warmer water, greater solar radiation, and unstable substrates likely contributed to variable benthic insect and tailed frog tadpole densities over time, although benthic insect communities became more similar after three years. The debris flows also decreased channel slopes and removed channel step barriers such that cutthroat trout were able to rapidly occupy habitats far upstream, but sculpins were slower to recolonize and both fish species exhibited some differences in recovery between the two streams. Crayfish were severely impacted by the debris flows; this may be due to attributes of their life history and the timing of the flows. Overall, we found that recolonizing aquatic species exhibited varying levels of resilience and recovery after the disturbances being related to the influence of physical habitat conditions, species dispersal ability, and the presence of nearby source populations.

A field evaluation of the growth and survival of age-0 Oncorhynchus mykiss tagged with 8-mm passive integrated transponder (PIT) tags

BackgroundIn fish tagging studies, tag size limits the size of fish that can be tagged, the fraction of a population that can be represented, and ultimately inferences that can be made about the study population, particularly when juvenile fish are the subject of interest. Introduction of an 8-mm passive integrated transponder (PIT) reduced the minimum taggable size of fish, but it has not been evaluated in field trials. We evaluated the growth and survival of age-0 Oncorhynchus mykiss tagged with 8-mm PIT tags in four streams in southwest Washington, USA.ResultsA total of 351 PIT tagged fish and 340 control fish (marked with pelvic fin clips) were released, but recapture rates were low, particularly for control fish. Growth in length and mass did not differ between small (42–54 mm) and large (55–64 mm) PIT tagged fish. There was a slightly positive, but weak, relation between tag burden and growth in mass; however, there was considerable variability in this relation (R2 = 0.115). Summer to autumn joint probability of fish surviving and remaining in the study area estimated with a Bayesian mark-recapture model ranged from 0.228 to 0.478 in study streams. We found no significant relation between tag burden and survival, suggesting neither tag burden nor fish size at tagging affected survival.ConclusionsAlthough this study was limited in scope, it provided insight into how age-0 O. mykiss tagged with 8-mm PIT tags grew and survived under natural conditions. We showed that fish as small as 42 mm could be tagged without detrimental effects, which should allow researchers to represent a larger portion of study populations through PIT tagging.

Morphometric Analysis of Two Enigmatic Sculpin Species, Cottus gulosus and Cottus perplexus (Scorpaeniformes: Cottidae)

The freshwater sculpin species Cottus gulosus and Cottus perplexus present a challenge to biologists because of their overlapping sympatric distributions, morphological similarity, and historical inconsistency regarding species limits and ranges. We applied linear morphometric analyses in conjunction with meristic and discrete character counts to 1) assess the validity of previously cited diagnostic characters, 2) test multivariate combinations of characters to elucidate novel diagnostic characters useful in distinguishing the 2 species, and 3) apply these diagnostic characters to identify previously unidentified specimens obtained from Washington streams, where the 2 species occur in sympatry. Our analysis of 270 Cottus specimens encompassing the range of both species uncovered little support for the diagnostic utility of previously cited characters, although significant differences between mean measurements for each group were detected in the proportion of individuals with palatine teeth and a median chin pore, as well as in mean mouth-width to body-width ratio. Additionally, we were unable to detect linear transformations of continuous measurements that could be used to reliably identify individuals of these 2 species. These results prevented us from confidently identifying the unknown Washington specimens and illustrate that multi-locus phylogenetic studies are needed to assess species boundaries for these morphologically similar fishes.

Empirical and simulation evaluations of an abundance estimator using unmarked individuals of cryptic forest-dwelling taxa

Conservation and management programs use populations of sentinel taxa, such as stream-associated amphibians, as indicator species due to their perceived sensitivity to environmental change. Estimating population size with traditional tools such as mark–recapture estimators may be impractical for forest-dwelling stream-associated amphibians, many of which are cryptic and have low detection probabilities. In addition, sampling techniques can alter habitat conditions, particularly with repeat sampling, and compromise inferences about management impacts. We used simulated and empirical data and N-mixture models to estimate detection probabilities and abundances for two amphibian genera, giant Dicamptodon and torrent Rhyacotriton salamanders, sampled with a less-invasive approach than other methods. We surveyed forested headwater streams located in western Washington, USA, for salamanders 7 July–27 August 2008. We assessed model sensitivity to changes in animal abundance (5 and 15) and detection probability (0.05–0.5), as well as study design alternatives including number of sample plots (50–150) and number of sampling visits (2–4). We also evaluated the effects of stream temperature and stream order on detection probability using data collected from forested streams in Washington, USA. Precision of detection probability estimates improved as the number of plots and sampling occasions increased. Variability of estimated population sizes decreased with higher detection probability, although species abundance had little effect on precision of detection probability estimates. Detection probability estimated from empirical data ranged from 0.07 to 0.65 for giant salamanders and 0.06–0.67 for torrent salamanders. Giant salamander detection probability was positively associated with stream temperature regardless of stream order, and was higher in second- and third-order streams than first-order streams. Detection probability for torrent salamanders varied with stream temperature, order, and the interaction of those covariates, with detection increasing with temperature for second- and third-order streams but showing a flat or decreasing trend for first-order streams. N-mixture models, paired with careful consideration of study design alternatives, can produce robust estimates of abundance and obviate the traditional reliance on indices of relative abundance for many rare and sensitive taxa. Use of a less invasive technique for sampling stream-associated amphibians resulted in sufficient animal captures for suitable model performance. Our simulation results can inform study design and direct efficient allocation of field effort.

Elevated Nitrogen and Phosphorus Concentrations in Urbanizing Southwest Washington Streams

In southwest Washington, rapid population growth and associated land use change have resulted in elevated stream nutrient concentrations. To evaluate the extent and nature of human alterations to stream nutrient concentrations in this region, we compiled four water years of total phosphorus (TP) and dissolved inorganic nitrogen (DIN) data from two long-term monitoring programs. We also quantified watershed characteristics likely to affect aquatic nutrient loading, and tested for correlations between these characteristics and stream nutrient concentrations. Average nutrient concentrations in study streams were significantly elevated relative to EPA recommended nutrient criteria in all sites for DIN and in nine out of 14 sites for TP. Of the watershed characteristics investigated, percent “impervious” ( ) and percent “forested” (-) were the best predictors of TP concentration (R2 = 0.41 and 0.64, respectively, and — indicate the slope of the regression). Percent “developed” ( ) and percent “forest and woody wetland” (-) were the best predictors of DIN concentration (R2 = 0.75 and 0.73, respectively). In urban streams, the mean dry season DIN concentration was significantly higher than the mean wet season DIN concentration, but this pattern was reversed in less urban watersheds. Urban streams also had significantly higher DIN than non-urban streams. The strong relationship between DIN and “developed land” suggests that as southwest Washington's population continues to grow, targeted N management will become increasingly important. The strong negative relationship between “forest and woody wetland” and both TP and DIN concentration suggests that this land use type is particularly important in reducing stream nutrient loading.

Pyrethroid insecticides in urban salmon streams of the Pacific Northwest

Urban streams of the Pacific Northwest provide spawning and rearing habitat for a variety of salmon species, and food availability for developing salmon could be adversely affected by pesticide residues in these waterbodies. Sediments from Oregon and Washington streams were sampled to determine if current-use pyrethroid insecticides from residential neighborhoods were reaching aquatic habitats, and if they were at concentrations acutely toxic to sensitive invertebrates. Approximately one-third of the 35 sediment samples contained measurable pyrethroids. Bifenthrin was the pyrethroid of greatest concern with regards to aquatic life toxicity, consistent with prior studies elsewhere. Toxicity to Hyalella azteca and/or Chironomus dilutus was found in two sediment samples at standard testing temperature (23 °C), and in one additional sample at a more environmentally realistic temperature (13 °C). Given the temperature dependency of pyrethroid toxicity, low temperatures typical of northwest streams can increase the potential for toxicity above that indicated by standard testing protocols.

Trends in Pesticide Concentrations in Streamsof the Western United States, 1993-20051

Trends in pesticide concentrations for 15 streams in California, Oregon, Washington, and Idaho were determined for the organophosphate insecticides chlorpyrifos and diazinon and the herbicides atrazine, s-ethyl diproplythiocarbamate (EPTC), metolachlor, simazine, and trifluralin. A parametric regression model was used to account for flow, seasonality, and antecedent hydrologic conditions and thereby estimate trends in pesticide concentrations in streams arising from changes in use amount and application method in their associated catchments. Decreasing trends most often were observed for diazinon, and reflect the shift to alternative pesticides by farmers, commercial applicators, and homeowners because of use restrictions and product cancelation. Consistent trends were observed for several herbicides, including upward trends in simazine at urban-influenced sites from 2000 to 2005, and downward trends in atrazine and EPTC at agricultural sites from the mid-1990s to 2005. The model provided additional information about pesticide occurrence and transport in the modeled streams. Two examples are presented and briefly discussed: (1) timing of peak concentrations for individual compounds varied greatly across this geographic gradient because of different application periods and the effects of local rain patterns, irrigation, and soil drainage and (2) reconstructions of continuous diazinon concentrations at sites in California are used to evaluate compliance with total maximum daily load targets.

Fish Passage Effectiveness of Recently Constructed Road Crossing Culverts in the Puget Sound Region of Washington State

AbstractFish passage barriers at road–stream crossings are widely recognized as salmon Oncorhynchus spp. habitat restoration priorities in Washington State and throughout the Pacific Northwest of the USA. More than 3,500 fish passage barriers (mostly culverts) have been repaired in Washington streams since 1999, costing more than US$139 million. We evaluated fish passage at 77 randomly selected culverts (new and repaired) that were issued permits during three time periods (1998, 2003, and 2007). This sample represents approximately 85% of the fish passage culverts permitted in the Puget Sound region of Washington State during the last 10 years. All study culverts were permitted for fish passage under the Washington Department of Fish and Wildlife's (WDFW) hydraulic project approval (HPA) permitting process and evaluated using physical criteria in the WDFW's fish passage barrier standard. Our results indicate that 30% of culverts (23 of 77) permitted under the HPA process for fish passage were, in fact, barriers. Culverts permitted as no‐slope (one of the most common design types) or as an unknown design type were barriers in 45% of cases. Most culvert failures were due to noncompliance with permit provisions, particularly culvert slope, and a lack of critical evaluation of proposed plans in the context of site conditions by permitting biologists. We found no relationship between barrier status and permit date, experience of permitting biologists, quality of permit, or project sponsor type (private, public, or restoration entity). These results indicate the need for mechanisms to ensure better compliance from project sponsors and an improved process to critically evaluate the adequacy of proposed plans in the context of project site conditions.

Differential Incorporation of Natural Spawners vs. Artificially Planted Salmon Carcasses in a Stream Food Web: Evidence from δ15N of Juvenile Coho Salmon

AbstractPlacement of salmon carcasses is a common restoration technique in Oregon and Washington streams, with the goal of improving food resources and productivity of juvenile salmon. To explore the effectiveness of this restoration technique, we measured the Δ15N of juvenile coho salmon (Oncorhynchus kisutch) as an index of the incorporation of salmon‐derived nutrients (SDN) in streams with varying amounts of natural spawners and artificially placed carcasses. Although Δ15N of juvenile coho salmon increased with salmon carcass mass in the stream channel, this relationship was only significant for natural spawners. Artificially placed carcasses did not appear to increase SDN utilization by juvenile coho salmon. Possible explanatory factors include the restricted spatial and temporal distribution of carcasses, general absence of eggs, and lack of substrate bioturbation associated with artificially placed carcasses. Our study suggests that artificially placed carcasses at the levels added in the Oregon Coast Range may not directly mimic the role of natural spawners in stream foodwebs.

The occurrence of glyphosate, atrazine, and other pesticides in vernal pools and adjacent streams in Washington, DC, Maryland, Iowa, and Wyoming, 2005–2006

Vernal pools are sensitive environments that provide critical habitat for many species, including amphibians. These small water bodies are not always protected by pesticide label requirements for no-spray buffer zones, and the occurrence of pesticides in them is poorly documented. In this study, we investigated the occurrence of glyphosate, its primary degradation product aminomethylphosphonic acid, and additional pesticides in vernal pools and adjacent flowing waters. Most sampling sites were chosen to be in areas where glyphosate was being used either in production agriculture or for nonindigenous plant control. The four site locations were in otherwise protected areas (e.g., in a National Park). When possible, water samples were collected both before and after glyphosate application in 2005 and 2006. Twenty-eight pesticides or pesticide degradation products were detected in the study, and as many as 11 were identified in individual samples. Atrazine was detected most frequently and concentrations exceeded the freshwater aquatic life standard of 1.8 micrograms per liter (microg/l) in samples from Rands Ditch and Browns Ditch in DeSoto National Wildlife Refuge. Glyphosate was measured at the highest concentration (328 microg/l) in a sample from Riley Spring Pond in Rock Creek National Park. This concentration exceeded the freshwater aquatic life standard for glyphosate of 65 microg/l. Aminomethylphosphonic acid, triclopyr, and nicosulfuron also were detected at concentrations greater than 3.0 microg/l.

Comparison of Two Techniques for Surveying Headwater Stream Amphibians

Abstract: We compared rubble‐rousing versus light‐touch stream amphibian survey techniques in multiple 1‐m plots across 10 streams in southwest Washington, USA. Specifically, we wanted to determine if light‐touch surveys provide unbiased estimates of abundance (i.e., provide counts correlated with rubble‐rousing counts) and which method would provide more cost‐effective presence or absence information. Rubble‐rousing, a common technique for surveying stream‐associated amphibians in the Pacific Northwest, took 12 times as long as light‐touch to apply. Abundance estimates and standard errors for rubble‐rousing were consistently higher than those for light‐touch for all life stages for the coastal tailed frog (Ascaphus truei) and Columbia torrent salamander (Rhyacotriton kezeri). Except for eggs, light‐touch detected all life stages found during rubble‐rousing. For frogs, only some rubble‐rousing abundance estimates, mostly involving second‐year larvae, were highly correlated with their light‐touch counterparts, whereas for salamanders, similar comparisons generated high correlations across most life stages. Correlations between methods were consistently greater for salamanders than for frogs. However the smaller tailed frog sample sizes and the cryptozoic nature of some life stages may have contributed to this pattern. Depending on the degree to which researchers can tolerate false‐negative error rates, light‐touch may prove less costly than rubble‐rousing for detecting species presence. For the cost of obtaining one rubble‐rousing sample, many light‐touch samples can be used across a range of habitats for detecting species patchily distributed.

Change in the Upper Extent of Fish Distribution in Eastern Washington Streams between 2001 and 2002

AbstractThe upper extent of fish distribution in streams has received increasing attention in recent years because fish‐bearing streams in the Pacific Northwest receive greater protection from timber harvest than do non‐fish‐bearing waters. Significant amounts of time and funding have been spent surveying streams for the upper extent of fish distribution and, more recently, in developing models that predict where fish use ends in streams. The reliability of each approach rests in part on the assumption that the upper extent of fish distribution can be accurately determined by a single survey and that fish distribution boundaries (FDBs) do not shift appreciably through time. To examine changes in the upper extent of fish distribution, streams were surveyed throughout 10 forested watersheds in eastern Washington in the summers of 2001 and 2002. In 2002, 308 FDBs were resurveyed. Resurveys resulted in the establishment of 172 terminal boundaries (occurring mid‐channel or at the confluence of two non‐fish‐bearing streams) and 136 lateral boundaries (occurring where a non‐fish‐bearing channel laterally intersects a fish‐bearing channel). The number of terminal boundary changes was approximately evenly distributed among upstream shifts, no change, and downstream shifts. Excluding two relatively large downstream shifts, the mean distance between 2002 and 2001 terminal boundaries was −2.5 m, which is not statistically or biologically significant. Terminal boundaries most often occurred immediately below small impasses created by large woody debris. Of the 136 streams occurring above lateral boundaries in 2001, 134 were again found to support no fish. The upper extent of fish distribution was similar between these 2 years, perhaps because of similar streamflows during the two sampling sessions. The number of terminal‐to‐lateral boundary shifts equaled the estimated number of lateral‐to‐terminal boundary shifts, indicating that no net change in the use of small tributaries occurred between the 2 years.

Spatial Distribution and Habitat Use of Pacific Lamprey (Lampetra tridentata) Ammocoetes in a Western Washington Stream

ABSTRACT We investigated the spatial distribution and habitat associations of Pacific lamprey (Lampetra tridentata) ammocoetes in a southwest Washington stream. The ammocoetes exhibited a highly aggregated, non-random spatial distribution pattern. Conductivity, dissolved oxygen, and gradient influenced ammocoete distribution at the large scale (across 50 m reaches), whereas wetted width, percent fines, canopy density, and velocity influenced distribution at the small scale (1 m2 quadrats). We observed optimum water depths of 70 cm and current velocities of 0 to 10 cm/sec. In general, ammocoetes preferred fine substrates though larger ammocoetes were found at low densities in large gravel substrates. Our findings illustrate the importance of slow water environments for the rearing stage of Pacific lamprey. Restoration activities should be designed to restore watershed function and process in a way that naturally allows systems to retain these critical habitats. Additionally, when considering ammocoete habitat use and distribution, it is important to take into account the scale of the observations.

Use of Natural Temperature Patterns to Identify Achievable Stream Temperature Criteria for Forest Streams

Abstract Almost 90% of the streams listed on the EPA's nationwide database as water-quality impaired for temperature are in the Northwest. Historic records, monitoring of streams in federal wilderness areas in Oregon, and available data for least-impaired streams in Oregon, Washington, and Idaho show that many of these streams cannot achieve state temperature criteria. Forest management often is cited as a cause for increased stream temperature above state standards. The expectation that all forested streams should be below state targets has led to unnecessary listing of streams as impaired, wasting limited watershed protection resources. State water-quality programs should base water temperature criteria on natural patterns of stream temperature and on factors that have biological relevance to beneficial uses. West. J. Appl. For. 19(4):252–259.

Spatial Distribution and Habitat Use of the Western Pearlshell Mussel (Margaritifera falcata) in a Western Washington Stream

ABSTRACT We investigated the spatial distribution and habitat associations of western pearlshell mussels (Margaritifera falcata) in a southwest Washington stream. Variation in mussel occurrence differed with the scale of the observations, being lower among study reaches and higher within reaches. Additionally, mussels exhibited a highly aggregated, non-random spatial distribution pattern. The distribution of mussels at large scales (across reaches) was associated with dissolved oxygen and shear stress. Mussel distribution at small scales (with the 50 m reaches) was associated with wetted width, canopy, abundance of small gravel substrate, and distance from the stream bank. Mussels were found in locations having reduced shear stress, turbulence, and gradient and increased wetted width, abundance of small gravel, dissolved oxygen, and conductivity. Optimum water depth was 0.2–0.6 m and optimum current velocity was 0.23–0.30 m/sec. Mussels preferred substrates where boulders increased bed roughness, allowing small gravel and sand to create a stable, heterogeneous substrate.

Effects of sampling error on bioassessments of stream ecosystems: application to RIVPACS-type models

AbstractWe evaluated the influence of various sources of sampling error on the precision, accuracy, and sensitivity of bioassessments based on River Invertebrate Prediction and Classification System (RIVPACS)-type models. We used data from 98 minimally altered streams in western Oregon and Washington to generate 18 models representing 2 field collection techniques (fixed-area riffle and multiple-habitat collections) and 9 levels of subsampling (50, 100, …450 fixed counts). For each model, we generated 2 observed-to-expected taxa ratios (O/E). The 1st O/E was based on all predicted taxa and the 2nd excluded rare taxa (i.e., those taxa with predicted probabilities of occurrence <0.5). We then compared O/E values to determine the extent to which subsampling effort, field collection method, different field personnel, and individual site characteristics altered model performance. We also generated O/E values for 63 streams with varying amounts of watershed and channel alteration (test sites) to quantify the ex...

Ecological Effects of Hatchery‐Reared Juvenile Chinook and Coho Salmon on Wild Juvenile Salmonids in Two Washington Streams

AbstractWe observed wild salmonid fry at sites downstream (treatment) and upstream (control) from experimental releases of hatchery‐reared Chinook salmon Oncorhynchus tshawytscha smolts and coho salmon O. kisutch fry in two streams to estimate the effects of hatchery releases on the density, group size, microhabitat use, and size of wild fish. Most hatchery‐reared Chinook salmon smolts that were released left treatment sites within a few days; hatchery‐reared coho salmon fry remained at treatment sites for at least several weeks. We found limited evidence to suggest that the release of hatchery‐reared Chinook salmon smolts or coho salmon fry caused consistent, significant effects on the density, group size, microhabitat use, or size of wild coho salmon fry in the two streams. We observed significant decreases in the densities of wild coho salmon and trout fry (steelhead (anadromous rainbow trout) O. mykiss and cutthroat trout O. clarki combined) after hatchery‐reared Chinook salmon were released, but they did not occur when large numbers of hatchery fish were present at treatment sites. We observed an increase in the group size of wild coho salmon in one stream when Chinook salmon were released, but not in the other stream. We observed few changes in microhabitat use by wild coho salmon fry associated with the release of hatchery‐reared fish; most changes were observed in only a single stream or occurred after only some of the releases. The lengths and weights of wild coho salmon fry over the sampling season were not significantly different between treatment and control sites in one study stream. These results suggest that small‐scale releases of hatchery‐reared Chinook or coho salmon have few significant ecological effects on wild salmonid fry in small, coastal Washington streams, particularly when wild fry occur at low densities.

Predicting fish passage design flows at ungaged streams in eastern Washington

The Washington Department of Fish and Wildlife recently established statewide guidelines for incorporating adult fish migration into culvert design. One of the design options requires an estimate of fish passage design flow for ungaged culvert sites. A new model is presented for estimating fish passage design flows at ungaged sites in Eastern Washington. The model is founded on a unique definition of fish passage design flow and an area-based approach for estimating this flow at ungaged sites. The fish passage design flow was developed by combining the concepts of allowable fish delay, established by fisheries experts to be three days, with a consecutive day analysis. This design flow ensures that fish are not delayed for more than three consecutive days during a water year. A fish passage design flow per unit area is assigned to previously delineated subwatersheds in Eastern Washington using similarity relationships derived from basin parameters and US Geological Survey gaging stations. The percent standard error for this model is 36%, a significant improvement from the 75% standard error calculated for the model previously developed for fish passage design flows in Eastern Washington.

Evaluation of Depth and Velocity Preferences of Juvenile Coho Salmon in Washington Streams

We determined depth and velocity preference curves for juvenile coho salmon Oncorhynchus kisutch for input into the physical habitat simulation system of the instream flow incremental methodology. Preferences were determined by scaling the maximum value of the ratio of observed use to expected use to 1.0. We developed composite preference curves by compiling weighted preferences for depth and velocity from four separate studies. The values of depth and velocity preference determined in each of the four studies were weighted by the sample size of the study, summed for each depth or velocity interval, and then normalized such that the sums peaked at 1. Juvenile coho salmon showed maximum preference for depths of 76–100 cm and velocities of 3–6 cm/s. To test the composite preference curves, we used an independent data set, marking the locations of juvenile coho salmon and comparing the products of preferences for depth and velocity (Pref (dv) = Pref (d) · Pref (v), where Pref (d) is depth preference and Pref (v) is velocity preference) where coho salmon were observed with those at measurement points on a grid. In June and September 1996, most coho salmon were found most frequently at locations with high combined preferences.

Habitat Use by Fishes and Pacific Giant Salamanders in Small Western Oregon and Washington Streams

Abstract The habitat use patterns of juvenile salmonid fishes Oncorhynchus spp., Pacific giant salamanders Dicamptodon spp., torrent sculpins Cottus rhotheus, reticulate sculpins C. perplexus, and larval lampreys Entosphenus tridentatus and Lampetra spp. were examined in 30 small streams in western Oregon and Washington. Fish and salamander densities and sizes were compared between different habitat types (pools and riffles) in summer and winter. Species density, length, and diversity (richness and dominance) were also correlated with physical variables measured at the reach and watershed scales. In the summer, densities of coho salmon O. kisutch, cutthroat trout O. clarki, and larval lampreys were significantly higher in pools than in riffles (P < 0.01), whereas densities of age-0 torrent sculpins were higher in riffles than in pools (P < 0.01). In winter, densities of coho salmon, cutthroat trout, steelhead O. mykiss, and young-of-the-year trout fry were higher in pools than in riffles (P < 0.01). Cutth...