A molecular analysis of hybridization between native westslope cutthroat trout and introduced rainbow trout in southeastern British Columbia, Canada

Journal of Fish Biology (2001) 59 (Supplement A), 42 54 doi:10.1006/jfbi.2001.1761, available online at http://www.idealibrary.com on A molecular analysis of hybridization between native westslope cutthroat trout and introduced rainbow trout in southeastern British Columbia, Canada E. RUBIDGE, P. CORBETT AND E. B. TAYLOR Department of Zoology and Native Fish Research Group, University of British Columbia, 6270 University Blvd, Vancouver, B.C., Canada V6T 1Z4 and Mirkwood Ecological Consultants Ltd, P.O. Box 138, Winlaw, B.C., Canada V0G 2J0 Restriction site variation in the Ikaros gene intron was used to assess the incidence of westslope cutthroat trout (Oncorhynchus clarki lewisi), rainbow trout (O. mykiss) and interspecific hybrids at 11 localities among eight streams tributary to the upper Kootenay River system in south-eastern British Columbia, Canada. Out of 356 fish assayed by this technique, hybrids (n=16) were found at seven of the 11 sites across five different streams. Rainbow trout (n=6) were found at two of the 11 sites. Analysis of hybrids with a second genetic marker (heat shock 71 intron) indicated that most represented either backcrosses to both westslope cutthroat and rainbow trout, or post F 1 hybrids. Mitochondrial DNA analysis indicated that hybrid matings occur between male rainbow trout and female westslope cutthroat trout and vice versa. Comparison of present hybridization in five tributaries relative to an allozyme-based analysis in the mid-1980s, that documented hybrids in only a single tributary of seven that were common to the two studies, suggests that hybridization and introgression has increased in upper Kootenay River tributaries. The present analysis is a conservative estimate of genetic interaction between the species because introgression was not tested in the majority of samples. Identification of genetically pure westslope cutthroat trout populations, and why they might be resistant to introgression from rainbow trout, are crucial conservation priorities for this unique subspecies of cutthroat trout. 2001 The Fisheries Society of the British Isles Key words: hybridization; introgression; extinction; westslope cutthroat trout; British Columbia; rainbow trout. INTRODUCTION Hybridization has been broadly defined as interbreeding between genetically distinct populations (Arnold, 1997) and hybrid zones are geographic areas where genetically distinct populations come into contact and hybridize. Hybrid zones have been described in a number of taxa (Arnold, 1997) and are often associated with areas of secondary contact between previously isolated taxa (Hewitt, 1989). The study of hybrid zones, and in particular the evolutionary fate of hybrid genotypes, has contributed considerably to knowledge of processes involved in speciation and divergence with gene flow (Barton & Hewitt, 1985; Hewitt, 1989; Jiggins & Mallet, 2000). Fishes have long been considered one of the animal groups where hybridization and introgression (the movement of alleles from one population into another) are relatively common (Hubbs, 1955; Verspoor & Hammer, 1991; Arnold, 1997). In addition to its relevance to fundamental studies of genetic divergence and speciation, hybridization and introgression may occur when naturally allopatric species are brought into artificial contact Author to whom correspondence should be addressed. Tel.: +1 604 822 9152; fax: +1 604 822 2416; email: etaylor@zoology.ubc.ca 42 0022 1112/01/59A042+13 $35.00/0 2001 The Fisheries Society of the British Isles

INTROGRESSIVE HYBRIDIZATION AND CONSERVATION IN TROUT 43 through anthropogenic activities such as habitat modification and intentional or accidental introductions. Although exotic species introductions have been clearly recognized as one of the major contributors to species extinctions, their effects have focussed mostly on ecological interactions such as predation, competition, or habitat destruction. Only relatively recently have the potential impacts of hybridization and introgression been highlighted as a possible cause of species extinction (Rhymer & Simberloff, 1996; Huxel, 1999). The cutthroat trout Oncorhynchus clarki (Richardson) is a salmonid fish native to western North America. The species has been subdivided into several well recognized sub-species each with largely parapatric ranges relative to one another (Allendorf & Leary, 1988). One of the sub-species, the westslope cutthroat trout O. clarki lewisi Suckley, is native to tributaries of the upper Columbia, Snake, South Saskatchewan and Missouri rivers in south-eastern British Columbia and adjacent areas of Alberta, and in Washington, Idaho, and Montana of the U.S.A. with the majority of the distribution in the latter three areas. In most of the U.S.A., populations of westslope cutthroat trout have experienced extensive hybridization and introgression with rainbow trout Oncorhynchus mykiss (Walbaum) which have been widely introduced throughout the western states (Allendorf & Leary, 1988). The upper Kootenay River system (a tributary of the upper Columbia River) is the heart of the geographic range of westslope cutthroat trout in Canada, the sub-species is found only in scattered localities east of this area in south-western Alberta or west in headwater tributaries of the South Thompson River (Fraser River) drainage (Haas, 1998). Given the limited distribution of westslope cutthroat trout in Canada and the impacts of introgressive hybridization with rainbow trout in the core of its range in the U.S.A., the westslope cutthroat trout is arguably one of the most endangered salmonid fishes in Canada. This status has been recognized recently by the British Columbia government with the designation of O. clarki lewisi as a blue-listed species (one of special concern owing to its vulnerability to extinction). The inherent risk of a limited geographic range in Canada is compounded by the fact that the native range of westslope cutthroat trout in British Columbia has also been subjected to stocking of non-native rainbow trout for at least the last 40 years. In fact, between 1996 and 1998, a minimum of c. 45 000 rainbow trout has been stocked in East Kootenay streams (British Columbia Ministry of Fisheries stocking records, unpubl. data). A stocking programme in Montana also introduced large numbers of exotic rainbow trout into the Koocanusa Reservoir, an impoundement of a portion of the Kootenay River downstream of the Canada-U.S.A. border (Fig. 1). Consequently, rainbow trout resulting from these stocking programmes probably enter the upper Kootenay River watershed in Canada. Given experience in other areas (Leary et al., 1984, 1987; Carmichael et al., 1993), hybridization between westslope cutthroat and introduced rainbow trout is a real possibility in the upper Kootenay River. Hybrids between species may show reduced fitness (Arnold & Hodges, 1995) and, therefore, represent a loss of reproductive potential for westslope cutthroat trout in their native habitats. Alternatively, if hybrid progeny show no reductions in fitness or perhaps even elevated fitness relative to the parental genotypes, introgression and the production of hybrid swarms can occur resulting in the loss

44 E. RUBIDGE ET AL. 1 Upper Kootenay River North Pacific Ocean Canada USA Upper Columbia River 2 10 N BC Kootenay Lake 6 4 3 5 11 9 AB 7 8 50 km WA Kootenay River ID Koocanusa Reservoir MT FIG. 1. Sample localities examined for the presence of westslope cutthroat trout, rainbow trout, and their hybrids. 1, Mainstem upper Kootenay River; 2, White River; 3, lower Skookumchuk Creek; 4, upper Skookumchuk Creek; 5, lower St. Mary s River; 6, upper St. Mary s River; 7, Gold Creek; 8, Wigwam River; 9, lower Elk River; 10, upper Elk River; 11, Bull River. Tributaries were sampled previously by Leary et al. (unpubl. data). Inset shows study area in western North America., Canada U.S.A. border,, provincial or state boundaries; BC, British Columbia; AB, Alberta; WA, Washington State; ID, Idaho; MT, Montana. of genotypes unique to westslope cutthroat (Rhymer & Simberloff, 1996). Introgression with introduced rainbow trout has been identified as the single greatest threat to the persistance of native westslope cutthroat trout in most of its native range in the U.S. (Leary et al., 1984; Deeds et al. 1999). R. F. Leary, F. W. Allendorf & K. L. Knudson (unpubl. data) investigated hybridization and introgression between westslope cutthroat trout and rainbow trout in the upper Kootenay River in British Colombia and reported hybrids in only the White River (Fig. 1), of nine sampled. Given the continued and perhaps more widespread stocking of exotic rainbow trout in the East Kootenays since the completion of Leary et al. s study in 1986 1987, and the potential deleterious influence of hybridization and introgression on native westslope cutthroat trout, it is important to reconsider the issue in the watershed and provide more current and continued monitoring of the genetic interactions between the species. The goal of this study was to revisit the localities sampled by Leary et al. (unpubl. data) to assess whether or not the incidence of hybridization among sites was stable, had increased, or perhaps decreased. Achieving this goal can contribute to conservation of native westslope cutthroat trout by identifying priority areas for protection (e.g. areas with no apparent hybridization), and

INTROGRESSIVE HYBRIDIZATION AND CONSERVATION IN TROUT 45 understanding the potential outcomes of hybridization (lack of introgression versus production of a hybrid swarm). MATERIALS AND METHODS SAMPLE WATERSHEDS Fin clippings were obtained from 356 specimens collected by angling, electroshocking, and minnow trapping, where possible 30 40 individuals per locality. Fish were collected from eight systems, three of which (Skookumchuk Creek, St. Mary s River, and Elk River) were sampled both above and below putative migration barriers with a total of 11 sites examined (Fig. 1), seven of which were sampled by Leary et al. (unpubl. data). The fin clips were stored in 95% ethanol and fork length (L F ) and tentative species identification was determined for each fish. Westslope cutthroat trout were distinguished from rainbow trout by an upper jaw extended past the posterior margin of the eye, a bright red-orange slash under the base of the lower jaw, and reduced black spotting on the anterior body below the lateral line. Fish with intermediate or ambiguous characters were tentatively classified as hybrids. Samples were collected in as non-biased a manner as possible, by collecting fin clips from fishes as they were encountered to achieve the target sample size without regard to presumed genotypic status. Genomic DNA was extracted from the fin clips using standard Proteinase K/salt extraction techniques. DNA ANALYSIS Sequence variation in intron regions has found considerable application in population and species identification, particularly in fishes (Moran et al. 1997). Several intron regions have been reported to have species-specific sequence variants in salmonids including cutthroat (coastal and interior subspecies) and rainbow trout (J. Baker, unpubl. data). Five such introns were amplified (Table I) from genomic DNA preparations of trout using the polymerase chain reaction (for PCR conditions see Table I). Speciesspecific sequence variants were assayed by incubating the amplified DNA fragments with restriction enzymes to produce diagnostic restriction fragment length polymorphisms (RFLPs). A portion of the mitochondrial DNA genome was also amplified (Table I) in all hybrids that were identified to determine the sex-species combinations involved in the production of hybrids. Because mtdna is maternally inherited, RFLP haplotypes diagnostic for each species can identify the female parent involved in each hybrid mating. Fragments were incubated with appropriate enzymes (New England Biolabs, Table I, J. Baker, pers. comm.) at 37 C and the resultant fragments electrophoresed in 2% agarose gels. Gels were visualized after staining with ethidium bromide or SybrGreen dyes which bind with the DNA and fluoresce under UV illumination. DATA ANALYSIS To conduct a synoptic survey for the presence or absence of individuals of mixed genetic ancestry rather than a detailed analysis of hybridization and genotypic frequency, individual fish were identified as pure westslope cutthroat (WSCT) or rainbow trout (RBTR) if they were homozygous for RFLP alleles diagnostic for the parental species at the Ikaros locus (Table I). Individuals were considered to be hybrids if they were heterozygous for the RFLP alleles at this locus. Individuals identified as hybrids were assayed at a second locus to provisionally identify them as F 1 or later generation hybrids or backcross genotypes. These classifications were intended to describe the genotype of individuals, but do not necessarily reflect their parentage. Because only limited nuclear analysis was used in this study the numbers of pure parental and F 1 individuals could have been overestimated and the numbers of backcross individuals underestimated. Scoring additional nuclear markers would have reduced this uncertainty (for example the percentage of first generation backcrosses heterozygous at all loci is 6 25, 3 12, 1 56, and

TABLE I. Primers, PCR conditions (annealing temperatures/cycle number), and species specific diagnostic restriction fragment allele sizes for molecular markers used in DNA analyses of westslope cutthroat trout, rainbow trout and their hybrids. All sequences are written 5 3 Primers Source Sequence 5-3 Annealing temperature ( C) and cycle no. MgCl 2 (mm) Enzyme and allele size (bp) HSC 71F ctg cgt atc atc aat gag cc Taq I WCT: 550 HSC 71R gat cag gac ggt cat gac 60,56/8,32 1 5 RBTR: 590 IK F ctt cga gtg caa cct ctg Hinf I WCT: 500 IK R att ttc ttt gcc acc gag g 48/45 1 5 RBTR: 750 p53-7-f1 cag gtg gga tca gag tg Alu I WCT: 350,320 p53-7-r2 gga ttg tct cct gct gct tc 60,54/10,30 1 5 RBTR: 350,320 RAG3-F1 gat atg ttt tag cag gag c Dde I WCT: 750 RAG3-R1 agt aat gga gcc tac tgc 56/40 1 5 RBTR: 750 SL-F13 aac agc ctg aca gac agc Xba I WCT: ND SL-3 -R gat aat cac aaa cgt act gtg cc 52/40 1 5 RBTR: ND GluDG tga ctt gaa gaa cca ccg ttg Ava II WCT: 1400, 800 12Sar ata gtg ggg tat cta atc cca gtt 56,54/5,30 4 0 RBTR: 1600, 600 HSC, Heat shock protein; IK, Ikaros gene; p53, protooncogene; RAG, recombination activation gene; SL, somatolactin gene. The primer combination GluDG/12Sar amplifies a 3 0 kilobase pair fragment of mitochondrial DNA spanning the cytochrome b gene, the control region, and a portion of the 12 rrna gene. ND, not determined.

INTROGRESSIVE HYBRIDIZATION AND CONSERVATION IN TROUT 47 0 78% for four, five, six, and seven nuclear loci), but the aim was to record the presence of hybridization and introgression rather than to accurately determine their frequencies at each locality. Results are expressed as the percentage of each genotype (rainbow trout, westslope cutthroat trout hybrids) within the total number of fish examined per site and tested for differences in occurrence of hybrids with χ 2 randomization tests using MONTE of the REAP software package (McElroy et al., 1992). A rank correlation between the percentage hybridization at each site and the geographic distance of that site from the point where the Kootenay River crosses the Canada U.S.A. border (a putative major source of exotic rainbow trout) was also calculated. RESULTS AMPLIFICATION OF INTRONS Five intron regions were successfully amplified in samples of rainbow and cutthroat trout (Table I). Upon restriction of these products with enzymes suggested by J. Baker (pers. comm.), two combinations (Ikaros intron restricted with Hinf I and the Heat Shock protein restricted with Taq I) produced differences in restriction fragment lengths that consistently distinguished rainbow trout from westslope and coastal cutthroat trout (and the latter from one another). In particular, the Ikaros/Hinf I combination produced a greater size separation among fragments and was chosen for all subsequent analyses (Table I). The Ikaros sequence lacks a restriction site for Hinf I in rainbow trout resulting in a single DNA fragment of c. 750 base pairs (bp) whereas westslope cutthroat trout have one restriction site that results in two smaller fragments (Table I). The diagnostic status of the Ikaros/Hinf I assay was confirmed by surveying representative samples of rainbow trout (n=100) and cutthroat trout from outside the East Kootenays (n=50). HYBRID DETECTION Sixteen hybrid individuals and six rainbow trout were observed across all samples (Table II). The highest percentages of hybrid trout were found in the upper Kootenay River and Gold Creek (21 and 11%, respectively). No rainbow trout were found at the mainstem upper Kootenay River site, but 18% of the fish from Gold Creek were tentatively identified as rainbow trout based on morphology and variation at the Ikaros locus. The next most impacted systems appeared to be the lower St. Mary s River and below the rapids on Skookumchuck Creek (9 7 and 6 7% hybrids, respectively, Table III). The lower site on Skookumchuk Creek was the only other locality in which rainbow trout were observed, but the White River, upper Skookumchuk Creek, and the lower site on the Elk River also contained hybrids (Table III). Considering only the Ikaros genotypes, the distribution of hybrid genotypes was significantly different among the 11 localities (χ 2, P=0 035). In Gold Creek, five out of five fish suspected to be rainbow trout from external appearance were confirmed as such by the molecular analyses. Similarly, two suspected hybrids on the basis of morphology were confirmed as hybrids. The molecular analysis, however, indicated that one fish originally identified as a westlope cutthroat trout by external appearance was, in fact, a hybrid. In addition, in two other rivers there was disparity between molecular- and

48 E. RUBIDGE ET AL. TABLE II. Incidence of westslope cutthroat trout rainbow trout hybrid individuals among eleven localities tributary to the upper Kootenay River system, British Columbia. Hybrid individuals were identified as those fish heterozygous for species specific alleles at the Ikaros locus. Putative genotype classifications based on morphology are given in parentheses Locality n WSC Hybrids RBTR Percentage hybrids UKR (1) 15 12 (15) 3 (0) 0 (0) 20 0 WR (2) 33 30 (0) 3 (31) 0 (2) 9 1 SK-1 (3) 39 38 (39) 1 (0) 0 (0) 2 6 SK-2 (4) 33 30 (33) 2 (0) 1 (0) 6 1 SM-1 (5) 31 28 (31) 3 (0) 0 (0) 9 7 SM-2 (6) 31 31 (31) 0 (0) 0 (0) 0 0 GC (7) 36 28 (29) 3 (2) 5 (5) 10 7 WW (8) 34 34 (34) 0 (0) 0 (0) 0 0 ER-1 (9) 38 37 (38) 1 (0) 0 (0) 2 6 ER-2 (10) 30 30 (30) 0 (0) 0 (0) 0 0 BR (11) 36 36 (36) 0 (0) 0 (0) 0 0 UKR, Mainstem Upper Kootenay River; WR, White River; SK-1, lower Skookumchuck Creek; SK-2, upper Skookumchuck Creek; SM-1, lower St Mary s River; SM-2, upper St Mary s River; GC, Gold Creek; WW, Wigwam River; ER-1, lower Elk River; ER-2, upper Elk River; BR, Bull River. Numbers in parentheses after stream name represent localities in Figure 1. WSC, Westslope cutthroat trout; RBTR, rainbow trout; n=sample size. morphological-based species identifications. A small fish from the White River that was identified as rainbow trout by morphology was identified as a westslope cutthroat trout by the Ikaros assay, and one from the lower Skookumchuck Creek site that was a westslope cutthroat trout in appearance was identified as a rainbow trout using the Ikaros assay. The 16 fish identified as hybrids by the Ikaros assay were also characterized with the heat shock 71 RFLP diagnostic alleles. One sample did not amplify, four fish were also identified as hybrids, and the remaining 11 fish were homozygous for either the rainbow trout RFLP allele (n=6) or the westslope cutthroat trout RFLP allele (n=5, Table II). The mitochondrial DNA analysis of each hybrid (Table II) indicated that both species RFLP haplotypes were present among the 16 hybrids; rainbow trout mtdna characterized four of the hybrids and westslope cutthroat mtdna was found in the remaining hybrids. Most of the hybrids (74%) were either fingerlings or juveniles, while the minority were found at the sub-adult or adult stages (26%). A comparison of the percentage of hybrids within samples and the rank distance of sites from the U.S.A. border on the Koocanusa Reservoir revealed no apparent relationship between these two variables (r s =0 005, P>0 2). To assess more precisely the possible role of migration distance to tributary streams of the Kootenay River from Koocanusa Reservoir on extent of hybridization, sample sites were eliminated from above impassible migration barriers (n=2) and from the mainstem upper Kootenay River. Under these conditions there was a negative, but non-significant correlation between percentage of hybrids and rank migration distance (r s = 0 54, P>0 1).

TABLE III. Genetic characterization of westslope cutthroat trout rainbow trout hybrids Individual Morphological phenotype Ikaros genotype Heat shock genotype mtdna haplotype Probable genotype SK 1-37 WSC Hybrid Hybrid WSC F 1 hybrid, male RBTR female WSC SK 2-17 WSC Hybrid WSC UD WSCBC or post F 1 hybrid SK 2-25 WSC Hybrid Hybrid WSC F 1 hybrid, male RBTR female WSC ER 2-23 WSC Hybrid RBTR WSC RBTRBC or post F 1 hybrid WR 1-1 RBTR WSC RBTR RBTR Post F 1 hybrid WR 1-25 Hybrid Hybrid WSC WSC WSCBC or post F 1 hybrid WR 1-26 Hybrid Hybrid Hybrid WSC F 1 hybrid, male RBTR female WSC UKR 1-1 WSC Hybrid WSC WSC WSCBC or post F 1 hybrid UKR 1-13 WSC Hybrid UD UD Hybrid UKR 1-15 WSC Hybrid Hybrid WSC F 1 hybrid, male RBTR female WSC SM 1-1 WSC Hybrid RBTR RBTR RBTRBC or post F 1 hybrid SM 1-3 WSC Hybrid WSC WSC WSCBC or post F 1 hybrid SM 1-5 WSC Hybrid WSC WSC WSCBC or post F 1 hybrid GC 1-6 WSC Hybrid RBTR RBTR RBTRBC or post F 1 hybrid GC 1-13 Hybrid Hybrid RBTR RBTR RBTRBC or post F 1 hybrid GC 1-20 Hybrid Hybrid RBTR WSC RBTRBC or post F 1 hybrid SK, Skookumchuck Creek; WR, White River; SM, St Mary s River; GC, Gold Creek; UKR, mainstem upper Kootenay River; WSC, westslope cutthroat trout; RBTR, rainbow trout; F 1, first generation hybrid; RBTRBC, backcross to rainbow trout; WSCBC, backcross to westslope cutthroat trout; UD, undetermined owing to repeated PCR failure.

50 E. RUBIDGE ET AL. DISCUSSION INCIDENCE OF HYBRIDIZATION Of the five intron regions utilized, two provided diagnostic tools for identifying rainbow and westslope cutthroat trout, and their interspecific hybrids. The presence of at least first generation hybrids between westslope cutthroat trout and rainbow trout at seven of the 11 sites strongly suggests that interspecific hybridization is not an isolated event in the upper Kootenay River watershed. In addition, the present survey must be considered as a conservative estimate of genetic interaction between the species because later generation hybrids or backcrosses were not identified in the majority of samples. Observing fish homozygous for one or the other species heat shock allele when those same fish were identified as hybrids at the Ikaros locus clearly establishes that these individuals represent post F 1 hybrids or backcross individuals and indicates that introgression between the species is occuring at some scale, particularly in the White and St. Mary s rivers, and in Gold Creek (Table II). These results are striking in the light of the previous study by Leary et al. (unpubl. data) who used six diagnostic allozyme loci to sample seven of the same tributaries (using 16 40 individuals) but detected introgression in only a single tributary (White River). Leary et al. (unpubl. data) estimated that they could detect as little as 1% rainbow trout genetic material in the westslope cutthroat trout samples with probabilities of between 0 91 and 0 99. Consequently, despite the higher resolution of the allozyme analysis, present data indicated hybridization in six additional tributaries in the upper Kootenay River. These comparative analyses, therefore, strongly suggest that hybridization and introgression have increased on a geographic basis (i.e. affecting more tributaries) in the upper Kootenay River since the mid-1980s. Because only a single diagnostic locus was employed for most analyses, putative parental genotypes may actually represent backcrosses or advanced generation hybrids, and the levels of introgression within tributaries may be underestimated. For instance, an F 1 hybrid westslope cutthroat trout mating would produce an Ikaros RFLP homozygous for the westslope cutthroat trout allele with a frequency of 0 5. Further, an F 1 F 1 mating would produce offspring homozygous for the westslope cutthroat trout allele at a frequency of 0 25. Consequently, the present analysis should be viewed as conservative because only hybrids were assayed with a second locus to assess whether or not introgression, in addition to hybridization, was occurring. Determining the extent of introgression between species requires the use of only a modest number (i.e.<10) of genetic markers and would be suitable for determining general levels of introgression (Leary et al., unpubl. data) or for distinguishing among parental, F 1 hybrids, and backcross individuals more precisely (Boecklen & Howard, 1997). Nevertheless, in the 16 samples assayed at a second locus, 11 exhibited genotypes that could be produced only by post-f 1 hybrids or backcrosses (i.e. homozygous at one locus and heterozygous at the second). These genotypes indicate that hybrids are viable and fertile, an inference that is consistent with other studies of hybridization between rainbow trout and various subspecies of O. clarki (Leary et al., 1984; Campton & Utter, 1985; Carmichael et al., 1993;

INTROGRESSIVE HYBRIDIZATION AND CONSERVATION IN TROUT 51 Henderson et al., 2000). Experimental crosses between coastal cutthroat trout O. clarki clarki and steelhead trout anadromous O. mykiss indicated no evidence of reduced viability of hybrids (Hawkins & Foote, 1998). This is consistent with data for westslope cutthroat trout experimental crosses, although there is some evidence of reduced growth rates in hybrids between these species (Leary et al., 1995). The mtdna variation in the hybrids indicated that hybridization and introgression were reciprocal; matings occurred with both species acting as the female parent. This observation is consistent with reciprocal hybridization between rainbow trout and Yellowstone cutthroat trout O. clarki bouvieri (Jordan & Gilbert) in south-eastern Idaho (Henderson et al., 2000). By contrast, Baxter et al. (1997) and Redenbach & Taylor (pers. comm.) reported that hybridization between two species of char, Dolly Varden Salvelinus malma (Walbaum) and bull trout Salvelinus confluentus Suckley was highly asymmetrical (Dolly Varden males bull trout females) which may result from sneaking of Dolly Varden [which are much smaller than adult bull trout in sympatry) on bull trout pairs (Wirtz, 1999). The present data, therefore, suggest that there may be generally a more similar size at maturity between westslope cutthroat trout and rainbow trout which may promote reciprocal hybridization. Nevertheless, all inferred F 1 hybrids in this study had westslope cutthroat females (Table III). This suggests that first generation hybrids may be biased towards having westslope cutthroat trout mothers which may reflect the relative rarity of rainbow trout in the studied tributaries (Wirtz, 1999). Although these analyses are preliminary, the observation that hybrids were found across four different life history stages indicates that genetic interactions between westslope cutthroat trout and rainbow trout are not isolated incidents, but have been occurring over a number a years. Futhermore, if hybrids are indeed more abundant at younger life history stages, samples consisting largely or exclusively of adults (e.g. upper Skookumchuck Creek, Wigwam River) may significantly underestimate the frequency of hybrids across all life stages at those sites. At the two localities with westslope cutthroat trout, hybrids, and rainbow trout tests were made for deviations from Hardy-Weinberg equilibrium at the Ikaros locus that might stem from selection against heterozygotes (i.e. hybrids). There were fewer heterozygotes than expected, assuming random mating at the Gold Creek site (P=0 017), but not at the lower Skookumchuck Creek site (P=0 13). Although these data suggest that selection may be operating against hybrids at the Gold Creek site, there is little direct evidence of reduced performance of hybrids between rainbow trout and subspecies of cutthroat trout (Leary et al., 1984; Hawkins & Quinn, 1996; Hawkins & Foote, 1998; Young et al., 2001). Alternatively, deviations from equilibrium expectations in Gold Creek may indicate that assortative mating is more common there owing to greater similarity in relative species abundance, or to better opportunities for reproductive habitat partitioning. For instance, Henderson et al. (2000) indicated that the degree of spatial overlap between rainbow trout and Yellowstone cutthroat trout varied among tributaries of the South Fork Snake River. Such variability in spatial overlap during reproduction could contribute to differences among upper Kootenay River tributaries in incidence of hybridization, and patterns of non-random mating between rainbow trout and westslope cutthroat

52 E. RUBIDGE ET AL. trout utilizing the same tributary (e.g. Gold Creek). More intensive sampling within the St. Mary s watershed is underway to assess better the factors that may influence spatial variation in hybridization and gene flow between rainbow trout and westslope cutthroat trout. The observation that the Gold Creek site, closest to the reservoir, contained five rainbow trout and had a high incidence of hybrids suggest that the reservoir may be a major source of exotic rainbow trout. Alternatively, three sites closest to the reservoir had few or no F 1 hybrids (e.g. two in the Elk River, Bull River), but these sites were located above migration barriers. Rainbow trout from the reservoir probably have no opportunities to enter these sites because of the migration barriers. Consequently, physical features of the tributaries themselves may be the important determinants of potential introgression between westslope cutthroat trout and exotic rainbow trout. The identification of one backcross or post F 1 hybrid in the Elk River (above the dam at Elko, B.C.), however, indicates that hybridization with rainbow trout cannot be attributable only to dispersal of rainbow trout into the upper Kootenay River from the Koocanusa Reservoir. Some exotic rainbow trout may originate from introductions into individual tributaries which have taken place in the recent past (Ministry of Fisheries stocking records, Victoria, British Columbia). CONSERVATION IMPLICATIONS Although further analysis is required to determine more precisely the extent of introgression in tributaries of the upper Kootenay River, current data strongly suggest that hybridization and introgression between westslope cutthroat and rainbow trout has increased since its original documentation (Leary et al., unpubl. data). The most likely reason for the apparent increase is the continued and expanded introductions of rainbow trout into the Koocanusa Reservoir and adjacent tributaries. Given the increase in hybridization documented here and the likelihood of introgression between species (Leary et al., 1984; Henderson et al., 2000), the most obvious step to minimize impacts on native westslope cutthroat trout would be to cease all further introductions of exotic rainbow trout into the geographic range of O. clarki lewisi. Even if introductions of rainbow trout were stopped (and considerable pressure exists to ensure that they continue), in the absence of selection against hybrid genotypes introgressed westslope cutthroat trout populations will persist indefinitely. Consequently, to help protect native westslope cutthroat trouts it will be important to identify populations that are genetically pure, and to understand why these systems appear to be experiencing less, or are at less risk from, introgression. Clearly, such populations should be afforded increased priority for conservation. In addition, it should be standard policy that all exotic rainbow trout are marked so that their geographic sources can be identified. Alternatively, representative samples from introduced populations could be subject to genetic analysis to facilitate monitoring. For instance, high resolution microsatellite DNA assays combined with statistical analysis employing individual assignment procedures have good potential for identification of exotic genotypes (Hansen et al., 2000). Identifying the source of exotic rainbow trout will be important for understanding the relative contributions of different stocking sources, past and present introductions, and the possible contributions of naturalized rainbow trout to

INTROGRESSIVE HYBRIDIZATION AND CONSERVATION IN TROUT 53 introgression. These activities are especially crucial for the upper Kootenay River because of the apparently lower degree of introgression relative to populations in the adjacent states of Idaho and Montana and the fact that tributaries of the upper Kootenay River represent the last stronghold of a unique lineage of salmonid fish endemic to north-western North America. We thank J. Baker (University of Washington) and P. Moran (National Marine Fisheries Service) for sharing their intron markers with us, M. Stamford for technical assistance, and S. Latham and G. Haas for providing some tissue samples. Funding for our research was provided by a Natural Sciences and Engineering Research Council of Canada operating grant awarded to EBT and by Global Forest Science (Vancouver, B.C., project number GF-18-1999-32). We appreciate the comments made on our research by P. Miller and anonymous reviewers. References Allendorf, F. W. & Leary, R. F. (1988). Conservation and distribution of genetic variation in a polytypic species, the cutthroat trout. Conservation Biology 2, 170 184. Arnold, M. L. (1997). Hybridization and Evolution. Oxford: Oxford University Press. Arnold, M. L. & Hodges, S. A. (1995). Are natural hybrids fit orunfit relative to their parents? Trends in Ecology and Evolution 10, 67 71. Barton, N. & Hewitt, G. M. (1985). Analysis of hybrid zones. Annual Reviews in Ecology and Systematics 16, 113 148. Baxter, J. S., Taylor, E. B., Devlin, R. H., Hagen, J. & McPhail, J. D. (1997). Evidence for natural hybridization between Dolly Varden (Salvelinus malma) and bull trout (S. confluentus) in a northcentral British Columbia watershed. Canadian Journal of Fisheries and Aquatic Sciences 54, 421 429. Boecklen, W. J. & Howard, D. J. (1997). Genetic analysis of hybrid zones: numbers of markers and power of resolution. Ecology 78, 2611 2616. Campton, D. E. & Utter, F. M. (1985). Natural hybridization between steelhead trout (Salmo gairdneri) and coastal cutthroat trout (S. clarki) in two Puget Sound streams. Canadian Journal of Fisheries and Aquatic Sciences 42, 110 119. Carmichael, G. J., Hanson, J. N., Schmidt, M. E. & Morizot, D. C. (1993). Introgression among Apache, cutthroat, and rainbow trout in Arizona. Transactions of the American Fisheries Society 122, 121 130. Deeds, S. A., Kaeding, L. R., Lohr, S. C. & Young, D. A. (1999). Status Review for the Westslope Cutthroat Trout in the United States. Portland, Oregon: United States Dept. of the Interior, U.S. Fish and Wildlife Service. Haas, G. R. (1998). Indigenous Fish Species Potentially at Risk in BC with Recommendations and Prioritizations for Conservation, Forestry/Resource Use, Inventory, and Research. Fisheries Management Report, No. 105. Vancouver: B.C. Ministry of Fisheries. Hansen, M. M., Ruzzante, D. E., Neilsen, E. E. & Mensberg, K-L. M. (2000). Microsatellite and mitochondrial DNA polymorphism reveals life-history dependent interbreeding between hatchery and wild brown trout (Salmo trutta). Molecular Ecology 9, 583 594. Hawkins, D. K. & Foote, C. J. (1998). Early survival and development of coastal cutthroat trout (Oncorhynchus clarki clarki), steelhead (Oncorhynchus mykiss), and their reciprocal hybrids. Canadian Journal of Fisheries and Aquatic Sciences 55, 2097 2104. Hawkins, D. K. & Quinn, T. P. (1996). Critical swimming velocity and associated morphology of juvenile coastal cutthroat trout (Oncorhynchus clarki clarki), steelhead trout (O. mykiss), and their hybrids. Canadian Journal of Fisheries and Aquatic Sciences 53, 1487 1496.

54 E. RUBIDGE ET AL. Hewitt, G. M. (1989). The subdivision of species in hybrid zones. In Speciation and its Consequences (Otte, D. & Endler, J., eds), pp. 85 110. Sunderland, Mass.: Sinauer Assoc. Henderson, R., Kershner, J. L. & Toline, C. A. (2000). Timing and location of spawning by nonnative wild rainbow trout and native cutthroat trout in the South Fork Snake River, Idaho, with implications for hybridization. North American Journal of Fisheries Management 20, 584 596. Hubbs, C. L. (1955). Hybridization between fish species in nature. Systematic Zoology 4, 1 20. Huxel, G. R. (1999). Rapid replacement of native species by invasive species: effects of hybridization. Biological Conservation 89, 143 155. Jiggins, C. D. & Mallet, J. (2000). Bimodal hybrid zones and speciation. Trends in Ecology and Evolution 15, 50 255. Leary, R. F., Allendorf, F. W. & Knudsen, K. L. (1984). Introgression between westslope cutthroat trout and rainbow trout in Clark Fork River drainage, Montana. Proceedings of the Montana Academy of Sciences 43, 1 18. Leary, R. F., Allendorf, F. W., Phelps, S. R. & Knudsen, K. L. (1987). Genetic divergence and identification of seven subspecies of cutthroat trout and rainbow trout. Transactions of the American Fisheries Society 116, 580 587. Leary, R. F., Allendorf, F. W. & Sage, G. K. (1995). Hybridization and introgression between introduced and native fish. American Fisheries Society Symposium 15, 91 101. McElroy, D., Moran, P., Bermingham, E. & Kornfield, I. (1992). REAP: An integrated environment for the manipulation and phylogenetic analysis of restriction data. Journal of Heredity 83, 157 158. Moran, P., Dightman, D. A., Waples, R. S. & Park, L. K. (1997). PCR-RFLP analysis reveals substantial population-level variation in the introns of Pacific salmon (Oncorhynchus spp). Molecular Marine Biology and Biotechnology 6, 315 327. Rhymer, J. M. & Simberloff, D. (1996). Extinction by hybridization and introgression. Annual Reviews in Ecology and Systematics 27, 83 109. Verspoor, E. & Hammer, J. (1991). Introgressive hybridization in fishes: the biochemical evidence. Journal of Fish Biology 39(Suppl. A), 309 334. Young, W. P., Ostberg, C. O., Keim, P. & Thorgaard, G. H. (2001). Genetic characterization of hybridization and introgression between anadromous rainbow trout (Oncorhynchus mykiss irideus) and coastal cutthroat trout (O. clarki clarki). Molecular Ecology 10, 921 930. Wirtz, P. (1999). Mother species-father species: unidirectional hybridization in animals with female choice. Animal Behaviour 58, 1 12. doi:10.1006/anbe.1999.1144.