Effects of Hatchery Rearing on Brain Structures of Rainbow Trout, Oncorhynchus mykiss

In this study, we contrast brain morphology from hatchery and wild reared stocks to examine the hypothesis that in salmonid fishes, captive rearing produces changes in brain development. Using rainbow trout, Oncorhynchus mykiss, as a model, we measured eight regions of the salmonid brain to examine differences between wild and hatchery reared fish. We find using multiple analysis of covariance (MANCOVA), analysis of covariance (ANCOVA) and discriminant function analysis (DFA) that the brains of hatchery reared fish are relatively smaller in several critical measures than their wild counterparts. Our work may suggest a mechanistic basis for the observed vulnerability of hatchery fish to predation and their general low survival upon release into the wild. Our results are the first to highlight the effects of hatchery rearing on changes in brain development inbreak fishes.     

Identifying footprints of selection in stocked brown trout populations: a spatio‐temporal approach

Studies of interactions between farmed and wild salmonid fishes have suggested reduced fitness of farmed strains in the wild, but evidence for selection at the genic level is lacking. We studied three brown trout populations in Denmark which have been significantly admixed with stocked hatchery trout (19–64%), along with two hatchery strains used for stocking. The wild populations were represented by contemporary samples (2000–2006) and two of them by historical samples (1943–1956). We analysed 61 microsatellite loci, nine of which showed putative functional relationships [expressed sequence tag (EST)‐linked or quantitative trait loci]. F_ST‐based outlier tests provided support for diversifying selection at chromosome regions marked by three loci, two anonymous and one EST‐linked. Patterns of differentiation suggested that the loci were candidates for being under diversifying hitch‐hiking selection in hatchery vs. wild environments. Analysis of hatchery strain admixture proportions showed that in one wild population, two of the loci showed significantly lower admixture proportions than the putatively neutral loci, implying contemporary selection against alleles introduced by hatchery strain trout. In the most strongly admixed population, however, there was no evidence for selection, possibly because of immigration by stocked trout overcoming selection against hatchery‐derived alleles or supportive breeding practices allowing hatchery strain trout to escape natural selection. To our knowledge, this is the first study demonstrating footprints of selection in wild salmonid populations subject to spawning intrusion by farmed fish.     

Comparison of brook trout reproductive success and recruitment in an acidic adirondack lake following whole lake liming and watershed liming

Limestone applications to the catchment of one tributary to Woods Lake were highly effective in reducing stream acidity and stabilizing seasonal fluctuations in pH. The resulting improvement in stream water quality also led to a dramatic shift in reproductive strategy of the Woods Lake brook trout population. Prior to catchment liming, brook trout in Woods Lake were restricted to spawning on poor quality near shore substrate with limited ground water seepage. Reproductive success was limited by high mortality of eggs and larvae and recruitment from in lake spawning was not successful. Spawning brook trout did not utilize the tributary for spawning prior to watershed liming. Mitigation of acidity in the tributary, by catchment liming, effectively extended the spawning habitat available to the Woods Lake brook trout population and one year following treatment brook trout spawned successfully in the tributary for the first time in 6 years of observation. Significant recruitment of young trout into the lake population occurred from 1991 through 1993, although the absolute number of fish captured was relatively small. In the fall of 1993, four year classes of naturally spawned brook trout were present in the lake. Although reproductive success was enhanced by improving tributary spawning habitat in the Woods Lake basin, self maintenance of the population may be limited by low recruitment rates of young trout, due to high levels of summer mortality resulting from predation. Mitigation of this constraint would require substantially higher levels of fry production than were observed in Woods Lake and/or enhanced refugia for young trout. The results of this experiment suggest that re-establishment of tributary spawning populations of brook trout may be possible, with future reductions in acidic deposition, in acidic Adirondack lakes with limited in-lake spawning habitat.     

Variation in reproductive success and effective number of breeders in a hatchery population of steelhead trout (Oncorhynchus mykiss): examination by microsatellite-based parentage analysis

Conservation programs that release captive-bred individuals into the wild to mix with naturally produced individuals are an increasingly common method of supporting or enhancing weak or reduced populations that otherwise may not be self-sustaining. Captive and supportive breeding can be important conservation tools for species with small or declining populations; however, in the case of hatcheries producing salmonid fishes, detailed evaluation of spawning programs is rare. We examined variation in reproductive success, measured by adult offspring production, from three parental generations of hatchery-bred steelhead trout (Oncorhynchus mykiss) using an exclusion-based method of genetic parentage assignment. Reproductive success varied greatly among individuals (especially males) and was correlated with fecundity and maternal spawning date. Estimates of egg to smolt survival for the population as a whole among years ranged from 64% to 95%, marine survival ranged from 0.32% to 2.30%, and the number of adults produced per female ranged from 0 to 18 and the number of adults produced per male ranged from 0 to 32. The effective number of breeders ranged from 11% to 31% of the census population size for that brood year. These ratios fell within estimates from estimates of Ne/N in chinook (O. tshawytscha) and rainbow trout (O. mykiss) hatchery populations.     

Conservation of population structure and genetic diversity under captive breeding of remnant coaster brook trout (<Emphasis Type="Italic">Salvelinus fontinalis</Emphasis>) populations

Rebuilding wild populations often involves captive broodstocks derived from small, remnant populations. We measured a hatchery program’s ability to conserve genetic diversity when founding captive broodstocks from such populations. Migratory coaster brook trout were extirpated from most of their historic range in US waters of Lake Superior and were proposed for listing under the Endangered Species Act. Two captive broodstocks, one with 19 founders and another with 99 founders, were established to rebuild US populations. We used microsatellite markers to examine genetic variation in source populations and early hatchery generations. Broodstocks retained the strong differentiation found between source populations; however, one founder, with a low probability of belonging to either source population, sired 5.7% of F₁ progeny. We found small changes in within-population genetic variation across successive wild and hatchery generations of broodstocks. Evaluation of stage-specific survivorship indicated that equalizing family sizes of embryos produced modest gains in the effective number of breeders, and that survival in the hatchery was nearly random across families. Our study demonstrates the value of genetic monitoring during initial stages of hatchery programs for small and declining populations.     

Disentangling stocking introgression and natural migration in brown trout: survival success and recruitment failure in populations with semi‐supportive breeding

1. Introgression into natural salmonid populations from stocked conspecifics has been widely studied. Outcomes vary from no effect even after decades of stocking, to population replacement after only a couple of generations. Potential introgression caused by semi‐supportive breeding (i.e. using a mixture of local strains as brood stock) is, however, less well studied. 2. We investigated population structure of brown trout (Salmo trutta) in a regulated alpine lake with three natural, environmentally contrasting tributaries used as spawning and rearing habitat. Massive semi‐supportive breeding of admixed local strains has been implemented for decades. Stocked trout represented c. 17% of the total lake population, and a substantial post‐release survival reflects a considerable potential for introgression. However, the mark‐recapture studies indicate no spawning runs of stocked fish. 3. Using 13 polymorphic microsatellite loci, we found natural straying and non‐native reproduction, especially among wild populations inhabiting environmentally unstable habitat. Retained genetic structure across tributaries indicated low reproductive success of wild‐born non‐natives. Moreover, the genetic structure among tributaries has probably not been influenced by semi‐supportive breeding, because of recruitment failure of stocked trout.     

A genetically distinct wild redband trout (<Emphasis Type="Italic">Oncorhynchus mykiss gairdneri</Emphasis>) population in Crane Prairie Reservoir,Oregon, persists despite extensive stocking of hatchery rainbow trout (<Emphasis Type="Italic">O. m. irideus</Emphasis>)

Crane Prairie Reservoir in the upper Deschutes River Basin has historically supported a wild population of migratory Deschutes River redband trout. Owing to its status as a premier destination for recreational angling in Oregon, the reservoir has been stocked with domesticated hatchery rainbow trout since 1955. In recent years the wild redband trout population has experienced a substantial decline. Effects on productivity related to genetic interaction with naturally spawning hatchery-origin fish (fitness risks) have not been determined. The species Oncorhynchus mykiss has been characterized with substantial genetic diversity throughout the Deschutes River Basin that further heightens the challenge of identifying specific conservation needs of wild populations. A conservation plan for Crane Prairie wild redband trout requires a better understanding of the natural reproductive success of out-of-basin hatchery trout in the reservoir tributaries, and the similarity between Crane Prairie redband trout with other extant redband trout populations in the basin. Using a suite of 17 microsatellite nuclear DNA markers, we evaluated the genetic structure among Crane Prairie Reservoir redband trout, hatchery rainbow trout, and two adjacent populations of redband trout from within the Upper Deschutes River Basin. We observed significant heterogeneity between the hatchery and wild Crane Prairie populations that may reflect differences in life histories, differential productivity and assortative mating. The genetic distinctions observed among the three redband trout populations suggest restricted gene flow and genetic drift within the upper basin. Temporally stratified sampling and larger numbers of samples will be necessary to confirm these conclusions.     

Phylogeography,genetic structure,and conservation of the endangered Caspian brown trout, <Emphasis Type="Italic">Salmo trutta caspius</Emphasis> (Kessler, 1877), from Iran

The Caspian Sea, the largest inland closed water body in the world, has numerous endemic species. The Caspian brown trout (Salmo trutta caspius) is considered as endangered according to IUCN criteria. Information on phylogeography and genetic structure is crucial for appropriate management of genetic resources. In spite of the huge number of studies carried out in the Salmo trutta species complex across its distribution range, very few data are available on these issues for S. trutta within the Caspian Sea. Mitochondrial (mtDNA control region) and nuclear (major ribosomal DNA internal transcribed spacer 1, ITS-1, and ten microsatellite loci) molecular markers were used to study the phylogeography, genetic structure, and current captive breeding strategies for reinforcement of Caspian trout in North Iranian rivers. Our results confirmed the presence of Salmo trutta caspius in this region. Phylogenetic analysis demonstrated its membership to the brown trout Danubian (DA) lineage. Genetic diversity of Caspian brown trout in Iranian Rivers is comparable to the levels usually observed in sustainable anadromous European brown trout populations. Microsatellite data suggested two main clusters connected by gene flow among river basins likely by anadromous fish. No genetic differences were detected between the hatchery sample and the remaining wild populations. While the current hatchery program has not produced detectable genetic changes in the wild populations, conservation strategies prioritizing habitat improvement and recovering natural spawning areas for enhancing wild populations are emphasized.     

A phylogeographic survey of brook charr (Salvelinus fontinalis) in Algonquin Park, Ontario based upon mitochondrial DNA variation

Forty-nine populations of brook charr (Salvelinus fontinalis) from Algonquin Park lakes and rivers were analysed for mitochondrial DNA variation. Haplotypic distributions of wild fish in the Algonquin Park region of Ontario, Canada, predominantly reflect postglacial dispersal patterns into the region in spite of substantial hatchery plantings. Two major refugial groupings colonized this region. Northern and eastern watersheds (Amable du Fond, Bonnechere, and northern Petawawa), were colonized primarily by haplotype 1 fish (B1 phylogenetic assemblage), while Oxtongue River, southern Petawawa, and York River populations were colonized predominately by fish from the B2 and A mtDNA phylogenetic assemblages. Fish with haplotypes in the A and B2 phylogenetic assemblages are common in the Lake Huron drainage. All watersheds in the Park drain into the Ottawa River, except the Oxtongue drainage (part of the Lake Huron watershed). This suggests that early glacial outflows south of the Algonquin Park region (Kirkfield-Trent) may have been colonized by fish that initially invaded ‘Ontario island’ (south-western Ontario), while fish which invaded northern Algonquin Park were derived from a different refugial grouping(s) which may have involved colonization both up the Ottawa River drainage, and/or from a more westerly (Mississippian) refugial grouping. A majority of the populations in Algonquin Park have been planted with hatchery reared brook charr since the 1940s. The Hills Lake or ‘Domestic’ strain was used almost exclusively for these plantings. Comparisons of mtDNA haplotypic distributions in hatchery and wild fish suggests that hatchery females had minimal spawning success and/or their progeny had poor survivorship in the wild.     

Analysis of gene origin in the first adult returns to the Cultus sockeye salmon captive breeding program

Rigorous evaluation of the utility of captive breeding for the restoration of depleted wild salmonid fish populations has not been undertaken. In particular, little is known about the reproductive success of captively-bred individuals that are released back into an extant population and their capacity to assist in long-term population persistence. For the endangered Cultus Lake sockeye salmon population, we examined the potential genetic contribution of the first juvenile fish released from a captive breeding program upon their maturity in the natural Cultus Lake environment. Genetic analysis of 792 Cultus sockeye salmon that were spawned in captivity in 2004 and their adult progeny of 2007 and 2008 revealed a genetic bottleneck originating from 20 wild sockeye salmon hatchery-spawned at Cultus Lake in the previous generation. Pedigree analysis revealed that six of the 20 sockeye salmon spawned in 2001 (grandparents) gave rise to a majority of the hatchery spawners in 2004 (parents) and provided more than 30% of the genes in the progeny that survived to maturity in the wild. Allele frequencies and genetic diversity of the age three progeny that returned to Cultus Lake from their marine migration in 2007 reflected the bottleneck, but its genetic signature was faint among the more genetically diverse age four fish that returned in 2008. Two-generation analysis of gene origin among fish resulting from 2004 hatchery production indicated that they contained the genetic diversity expected from 36 effective ancestors.     

Habitat use and fish activity of landlocked Atlantic salmon and brook charr in a newly developed habitat compensation facility

Degradation and destruction of valuable spawning and rearing habitat due to anthropogenic changes (e.g., flow modification and channelisation) is known to have dramatic impacts on fish populations. To compensate for habitat losses due to hydropower development, an artificial fluvial habitat channel (‘Compensation Creek’) was constructed in south-central Newfoundland, Canada. The creek was designed to include appropriate habitat features for the two dominant salmonid fish species, landlocked Atlantic salmon (Salmo salar L.) and brook charr (Salvenius fontinalis Mitchell). The study examines the habitat use of landlocked Atlantic salmon and brook charr in the Compensation Creek using electromyogram (EMG) radio telemetry. Ten landlocked Atlantic salmon and eight brook charr were captured and tagged with EMG transmitters. In laboratory swimming experiments, the EMG values were calibrated against swimming speed. Fish were then released in the Compensation Creek and tracked on a daily basis. The results show that (1) during residence in the creek, both species used preferentially the habitat features designed to match their rearing habitat preferences, and (2) swimming speed did not vary among habitat types for either species.     

Assessing the impact of stocking northern-origin hatchery brook trout on the genetics of wild populations in North Carolina

The release of hatchery-origin fish into streams with endemics can degrade the genetics of wild populations if interbreeding occurs. Starting in the 1800s, brook trout descendent from wild populations in the northeastern United States were stocked from hatcheries into streams across broad areas of North America to create and enhance fishery resources. Across the southeastern United States, many millions of hatchery-origin brook trout have been released into hundreds of streams, but the extent of introgression with native populations is not well resolved despite large phylogeographic distances between these groups. We used three assessment approaches based on 12 microsatellite loci to examine the extent of hatchery introgression in 406 wild brook trout populations in North Carolina. We found high levels of differentiation among most collections (mean F′_ST = 0.718), and among most wild collections and hatchery strains (mean F′_ST = 0.732). Our assessment of hatchery introgression was consistent across the three metrics, and indicated that most wild populations have not been strongly influenced by supplemental stocking. However, a small proportion of wild populations in North Carolina appear to have been strongly influenced by stocked conspecifics, or in some cases, may have been founded entirely by hatchery lineages. In addition, we found significant differences in the apparent extent of hatchery introgression among major watersheds, with the Savannah River being the most strongly impacted. Conversely, populations in the Pee Dee River watershed showed little to no evidence of hatchery introgression. Our study represents the first large-scale effort to quantify the extent of hatchery introgression across brook trout populations in the southern Appalachians using highly polymorphic microsatellite markers.     

Differential Reproductive Success of Sympatric, Naturally Spawning Hatchery and Wild Steelhead, Oncorhynchus mykiss

Hatchery propagation of salmonids has been practiced in western North America for over a century. However, recent declines in wild salmon abundance and efforts to mitigate these declines through hatcheries have greatly increased the relative abundance of fish produced in hatcheries. The over-harvest of wild salmon by fishing mixed hatchery and wild stocks has been of concern for many years but genetic interactions between populations, such as hybridization, introgression and outbreeding depression, may also compromise the sustainability of wild populations. Our goal was to examine whether a newly established hatchery population of steelhead trout successfully reproduced in the wild and to compare their rate of reproductive success to that of sympatrically spawning native steelhead. We used eight microsatellite loci to create allele frequency profiles for baseline hatchery and wild populations and assigned the smolt (age 2) offspring of this parental generation to a population of origin. Adults originating from a generalized hatchery stock artificially selected for early return and spawning date were successful at reproducing in Forks Creek, Washington. Although hatchery females (N = 90 and 73 in the two consecutive years of the study) produced offspring that survived to emigrate as smolts, they produced only 4.4–7.0% the number produced per wild female (N = 11 and 10). This deficit in reproductive success implies that the proportion of hatchery genes in the mixed population may diminish since deliberate releases into the river have ceased. This hypothesis is being tested in a long-term study at Forks Creek.     

Temporal constraints on the potential role of fry odors as cues of past reproductive success for spawning lake trout

Deciding where to reproduce is a major challenge for most animals. Many select habitats based upon cues of successful reproduction by conspecifics, such as the presence of offspring from past reproductive events. For example, some fishes select spawning habitat following odors released by juveniles whose rearing habitat overlaps with spawning habitat. However, juveniles may emigrate before adults begin to search for spawning habitat; hence, the efficacy of juvenile cues could be constrained by degradation or dissipation rates. In lake trout (Salvelinus namaycush), odors deposited by the previous year's offspring have been hypothesized to guide adults to spawning reefs. However, in most extant populations, lake trout fry emigrate from spawning reefs during the spring and adults spawn during the fall. Therefore, we postulated that the role of fry odors in guiding habitat selection might be constrained by the time between fry emigration and adult spawning. Time course chemical, physiological, and behavioral assays indicated that the odors deposited by fry likely degrade or dissipate before adults select spawning habitats. Furthermore, fry feces did not attract wild lake trout to constructed spawning reefs in Lake Huron. Taken together, our results indicate fry odors are unlikely to act as cues for lake trout searching for spawning reefs in populations whose juveniles emigrate before the spawning season, and underscore the importance of environmental constraints on social cues.     

Selection and genetic drift in captive versus wild populations: an assessment of neutral and adaptive (MHC-linked) genetic variation in wild and hatchery brown trout (<Emphasis Type="Italic">Salmo trutta</Emphasis>) populations

Captive bred individuals are often released into natural environments to supplement resident populations. Captive bred salmonid fishes often exhibit lower survival rates than their wild brethren and stocking measures may have a negative influence on the overall fitness of natural populations. Stocked fish often stem from a different evolutionary lineage than the resident population and thus may be maladapted for life in the wild, but this phenomenon has also been linked to genetic changes that occur in captivity. In addition to overall loss of genetic diversity via captive breeding, adaptation to captivity has become a major concern. Altered selection pressure in captivity may favour alleles at adaptive loci like the Major Histocompatibility Complex (MHC) that are maladaptive in natural environments. We investigated neutral and MHC-linked genetic variation in three autochthonous and three hatchery populations of Austrian brown trout (Salmo trutta). We confirm a positive selection pressure acting on the MHC II β locus, whereby the signal for positive selection was stronger in hatchery versus wild populations. Additionally, diversity at the MHC II β locus was higher, and more uniform among hatchery samples compared to wild populations, despite equal levels of diversity at neutral loci. We postulate that this stems from a combination of stronger genetic drift and a weakening of positive selection at this locus in wild populations that already have well adapted alleles for their specific environments.     

Non-native species limit stream restoration benefits for brook trout

Success of stream restoration can be difficult to define because many interacting abiotic and biotic factors across spatio-temporal scales can have measurable effects. Consequently, failure in habitat restoration to achieve targeted biological goals may reflect interactions of habitat restoration with unaccounted risks that have yet to be addressed on the landscape. This is particularly true within invaded landscapes, where habitat restoration can benefit non-native competitors as much as the native fishes for which restoration is designed. We tested for interacting effects of a reach scale habitat restoration effort and non-native trout competition on habitat use by a brook trout (Salvelinus fontinalis) metapopulation within a productive main stem corridor of the Shavers Fork watershed, West Virginia. We used a joint species occupancy model within a BACI sampling design to show that brook trout occupancy of main stem habitat was highest post-restoration within restored sampling reaches, but this benefit to native brook trout was conditional on brown trout (Salmo trutta) not being present within the main stem habitat. Collectively these results indicate that habitat restoration was only beneficial for native brook trout when non-native trout were absent from the restored sampling area. Proactive approaches to restoration will be integral for supporting resilient ecosystems in response to future anthropogenic threats (e.g. climate change), and we have shown that such actions will only be successful if non-native competitors do not also benefit from the restoration actions.     

Salmonid biomass and habitat relationships for small lakes

Lake morphometry and water chemistry were analyzed as predictors of brook trout and total salmonid biomass (brook trout, Atlantic salmon and Arctic char) in water bodies of Newfoundland. Lake morphometric variables included surface area, depth, perimeter and catchment area while water chemistry variables included conductivity/TDS and total Phosphorus. The broadly used Morpho-Edaphic Index was also assessed. Fish biomass/ha varied by an order of magnitude in the study lakes, despite similar water chemistry. Interactive stepwise multiple regression analysis suggested that Basin Permanence Index best explains brook trout biomass while surface area best explains the variance of total salmonid biomass in the relatively small, unproductive water bodies of Insular Newfoundland. In contrast, the Morphoedaphic Index performed poorly. The results suggest that biomass relationships with easily measured habitat variables such as surface area provide useful information, especially in management of water bodies under similar climate and nutrient regimes.     

Admixture analysis of stocked brown trout populations using mapped microsatellite DNA markers: indigenous trout persist in introgressed populations

Admixture between wild and captive populations is an increasing concern in conservation biology. Understanding the extent of admixture and the processes involved requires identification of admixed and non-admixed individuals. This can be achieved by statistical methods employing Bayesian clustering, but resolution is low if genetic differentiation is weak. Here, we analyse stocked brown trout populations represented by historical (1943–1956) and contemporary (2000s) samples, where genetic differentiation between wild populations and stocked trout is weak (pairwise F_ST of 0.047 and 0.053). By analysing a high number of microsatellite DNA markers (50) and making use of linkage map information, we achieve clear identification of admixed and non-admixed trout. Moreover, despite strong population-level admixture by hatchery strain trout in one of the populations (70.8%), non-admixed individuals nevertheless persist (7 out of 53 individuals). These remnants of the indigenous population are characterized by later spawning time than the majority of the admixed individuals. We hypothesize that isolation by time mediated by spawning time differences between wild and hatchery strain trout is a major factor rescuing a part of the indigenous population from introgression.     

Comparing stream-specific to generalized temperature models to guide salmonid management in a changing climate

Global climate change is predicted to increase air and stream temperatures and alter thermal habitat suitability for growth and survival of coldwater fishes, including brook charr (Salvelinus fontinalis), brown trout (Salmo trutta), and rainbow trout (Oncorhynchus mykiss). In a changing climate, accurate stream temperature modeling is increasingly important for sustainable salmonid management throughout the world. However, finite resource availability (e.g. funding, personnel) drives a tradeoff between thermal model accuracy and efficiency (i.e. cost-effective applicability at management-relevant spatial extents). Using different projected climate change scenarios, we compared the accuracy and efficiency of stream-specific and generalized (i.e. region-specific) temperature models for coldwater salmonids within and outside the State of Michigan, USA, a region with long-term stream temperature data and productive coldwater fisheries. Projected stream temperature warming between 2016 and 2056 ranged from 0.1 to 3.8 °C in groundwater-dominated streams and 0.2–6.8 °C in surface-runoff dominated systems in the State of Michigan. Despite their generally lower accuracy in predicting exact stream temperatures, generalized models accurately projected salmonid thermal habitat suitability in 82% of groundwater-dominated streams, including those with brook charr (80% accuracy), brown trout (89% accuracy), and rainbow trout (75% accuracy). In contrast, generalized models predicted thermal habitat suitability in runoff-dominated streams with much lower accuracy (54%). These results suggest that, amidst climate change and constraints in resource availability, generalized models are appropriate to forecast thermal conditions in groundwater-dominated streams within and outside Michigan and inform regional-level salmonid management strategies that are practical for coldwater fisheries managers, policy makers, and the public. We recommend fisheries professionals reserve resource-intensive stream-specific models for runoff-dominated systems containing high-priority fisheries resources (e.g. trophy individuals, endangered species) that will be directly impacted by projected stream warming.     

Comparative diets and foraging strategies of subyearling Atlantic salmon,brown trout,and rainbow trout during winter

Over‐winter survival of salmonids in streams is thought to be an important population regulation mechanism. Yet because of the difficulty of conducting field studies due to adverse weather or ice conditions, compared to other seasons, salmonid ecology during winter is least understood. Consequently, we sought to examine interspecific feeding associations of an important salmonid stream assemblage in the Lake Ontario watershed during winter. The diets of Atlantic salmon (Salmo salar) parr, brown trout (S. trutta) parr, and rainbow trout (Oncorhynchus mykiss) parr were significantly different in February but not in March. Salmonid diets differed from the benthos and the drift during both months. Dipterans (chironomids, simuliids, and tipulids) and ephemerellids were the major prey taxa consumed. All three species fed more heavily on prey items from the benthos than from the drift. The diet of Atlantic salmon had the highest similarity to the benthos whereas the diet of brown trout had the lowest similarity to the drift. All three salmonid species generally selected ephemerellids, limnephilids, and chironomids and avoided elmids. These winter feeding observations are the first reported for this specific salmonid assemblage and will help managers better understand interspecific associations during this critical period.