Westslope cutthroat trout COSEWIC assessment and status report: chapter 3

Species Information

Name and classification

The cutthroat trout (Oncorhynchus clarkii) is a polytypic species of salmonid native to western North America. It is widespread in both coastal and interior drainages in a range of habitats, from lakes and headwater streams, to estuaries and large rivers. At least 14 subspecies are currently recognized but only two occur naturally in Canada: the coastal cutthroat trout (O. c. clarkii) and the westslope cutthroat trout (O. c. lewisi); the latter being the focus of this review.  A third type, described by Dymond (1931) from the Revelstoke area in British Columbia as O. c. alpestris, is now considered to be synonymous with O. c. lewisi (see below):

Phylum: 
Chordata
Class: 
Actinopterygii
Order: 
Salmoniformes
Family: 
Salmonidae, subfamily Salmoninae (salmon, trout, charr)
Genus:  
Oncorhynchus (formerly Salmo)
Species:
Oncorhynchus clarkii (formerly Salmo clarkii)
Subspecies:
Westslope cutthroat trout O. clarkii lewisi (Girard)    formerly Salmo clarkii lewisi; considered synonymous with S. clarkii alpestris (Dymond)
Common name:   
 
         English:   
Westslope cutthroat trout
         French:    
Truite Fardée
         Other:   
cutthroat, interior cutthroat, westslope cutthroat, mountain cutthroat, cutty, spotted trout, (Montana) black-spotted trout, black spots, red-throated trout, Lewis’ trout

 

The extensive phenotypic variation exhibited by this species (in terms of size, colouration, and life-history characteristics) has led to considerable confusion and disagreement among taxonomists in its description, particularly in the number of genuine types and of the proper taxonomic epithets used in describing them. At one time, up to 40 taxonomic designations existed for the species, and relationships within the group remain controversial. Most taxonomists currently recognize 14 subspecies of cutthroat: four major subspecies showing substantial divergence (Coastal, Westslope, Lahontan, and Yellowstone cutthroat), and ten minor subspecies of limited range (Allendorf and Leary 1988; Behnke 2002).  Many of the interior cutthroat trout subspecies appear to be of fairly recent origin (i.e., since the most recent glaciation) so that no one phenotypic or meristic character clearly differentiates them. Considerable overlap in morphological and meristic characters also exists between cutthroat trout and rainbow trout (Oncorhynchus mykiss).  Morphological (Behnke 1992), karyotypic (Thorgaard 1983), and genetic data (Gyllensten et al. 1985; Shedlock et al. 1992), however, confirm that while substantial overlap exists, all cutthroat trout subspecies are indeed more closely related to each other than any is to rainbow trout (Figure 1).

Figure 1.  Phylogenetic relationships between the various cutthroat trout subspecies and rainbow trout. Diploid chromosome number (2N) is shown for cutthroat trout subspecies. Modified from Behnke (1997).


Figure 1.  Phylogenetic relationships between the various cutthroat trout subspecies and rainbow trout. Diploid chromosome number (2N) is shown for cutthroat trout subspecies. Modified from Behnke (1997).


Morphological description

Westslope cutthroat trout (hereafter referred to as WCT) have the streamlined body typical of salmonids (terminal mouth, small cycloid scales, and presence of an adipose fin) and are generally trout-like in appearance, with dark spots on a lighter background (Figure 2). Spots are small and irregularly shaped, forming a characteristic arc from the anterior base of the anal fin forward to the pectoral fin (more numerous posteriorly and concentrated above the lateral line). Body colouration ranges from silver to yellowish-green with red on the front and sides of the head. A narrow pink band may be present along the sides, but to a much lesser degree than in the closely related rainbow trout (hereafter referred to as RBT). Spawning fish often develop a bright red colouration over the entire body. Westslope cutthroat trout do not tend to become very large; generally 15-23 cm (28-142 g) with larger ones rarely over 41-46 cm (<1.4 kg) (Behnke 2002).

Figure 2.  Westslope cutthroat trout from the WigwamRiver (Upper Kootenay drainage). Photo courtesy of Ernest Keeley, Idaho State University.

Figure 2.  Westslope cutthroat trout from the WigwamRiver (Upper Kootenay drainage). Photo courtesy of Ernest Keeley, Idaho State University.

The most conspicuous character distinguishing cutthroat trout throughout its range in Canada is the presence of orange-red slashes beneath the lower jaw. The slashes, however, may be faint or absent in juveniles, making the field identification of WCT and RBT difficult. While field guides and taxonomic keys are available (e.g., McPhail and Carveth 1993; Pollard et al. 1997; Joynt and Sullivan 2003), considerable phenotypic variation exists between individual populations in the size, colouration and degree of spotting. Cutthroat trout generally tend to have a larger mouth than RBT with a longer maxillary, which extends past the hind portion of the eye. As well, a series of small basibranchial teeth at the back of the throat are considered to be diagnostic of pure cutthroat trout throughout much of their range (Behnke 1992; Leary et al. 1996; Weigel et al. 2002). Hybridization with RBT leads to a host of alternate spotting patterns and to the appearance of spots on the top of the head and anterior portion of the body.  Hybrids may also lack the basibranchial teeth and the slash beneath the lower jaw, and have a larger head-tail length ratio (Behnke 1992; Weigel et al. 2002).

This meristic overlap between forms has undoubtedly been exacerbated by the indiscriminate stocking of non-native species and variously hybridized fishes in the past (See LIMITING FACTORS).  While diagnostic testing now exists to identify the genetic composition of introgressed populations (McKay et al. 1997; Baker et al. 2002; Ostberg and Rodriguez 2002), the ecological and taxonomic status of hybridized populations remain largely unresolved (e.g., US Federal Register 1996; Allendorf et al. 2004).

 

Genetic description

Relatively few studies have investigated population structure in the westslope cutthroat subspecies. Early genetic assays of WCT using allozymes suggested that population subdivision was substantial, with Fst values (a widely used measure of genetic subdivision) ranging from 0.08 to 0.45 (Loudenslager & Gall 1980; Leary et al. 1987; Allendorf & Leary 1988). Populations appeared well differentiated and were often characterized by unique alleles or those that, while locally abundant, were uncommon over a larger geographic area. More recently, Taylor et al. (2003) examined population structure in 32 WCT populations in southeastern British Columbia (including sites in the upper Kootenay, upper Columbia, and upper Fraser drainages). Consistent with previous studies, while the total number of alleles per microsatellite locus ranged from 5 – 20 across the study area, the average number of alleles per microsatellite locus in any one population was low, averaging ~ 3.9. Expected heterozygosities averaged 0.56 but varied widely among populations (from 0.05-0.61). Habitat heterogeneity, in terms of migration barriers, appeared to be a significant factor in structuring this variation. Populations isolated above impassable migration barriers consistently showed significantly reduced variation and increased differentiation compared to populations not similarly constrained {allelic richness (2.1 vs. 2.9), expected heterozygosity (0.303 vs. 0.463), Fst (0.45 vs. 0.18); p< 0.005 for all tests}.

Population subdivision appears extensive throughout this region (overall Fst value of 0.32) and a large proportion of the total genetic variation (32%) is partitioned among populations (i.e., certain populations have a high frequency of alleles that are uncommon over the larger region). Based on the distribution of this allelic variation, Taylor et al. (2003) suggested the existence of four main groups of WCT in southeastern BC, loosely corresponding to geographic proximity (Figure 3). Populations isolated above migration barriers had significantly lower levels of genetic variation and were generally more divergent from one another than was observed between below-barrier populations. However, the significant divergence among populations lacking any obvious migration barriers (e.g., Kootenay mainstem populations with Fst of 0.12) suggests significant reproductive isolation and a high degree of demographic independence among even mainstem populations. The authors suggest that each individual population likely acts as a distinct biological entity and that the conservation of genetic biodiversity in the region’s WCT population will require the maintenance of many such populations throughout the region.

In Alberta, Potvin et al. (2003) examined the levels and partitioning of genetic diversity in 24 lakes from Banff and Waterton Lakes national parks. Again using microsatellites, they found populations with low to moderate levels of genetic variation (average heterozygosity ranging from ~ 0.1 – 0.5).  The number of alleles per locus was

Figure 3.  Principal Components Analysis (PCA) of the genetic relationships between southeastern British Columbia WCT populations. Modified from Taylor et al. (2003).

Figure 3.  Principal Components Analysis (PCA) of the genetic relationships between southeastern British Columbia WCT populations. Modified from Taylor et al. (2003).

Figure 4.  Global/Canadian ranges of native coastal and westslope cutthroat trout. Modified from Behnke (2002), because of the scale the range is only coarsely delineated.

Figure 4.  Global/Canadian ranges of native coastal and westslope cutthroat trout. Modified from Behnke (2002), because of the scale the range is only coarsely delineated.

significantly lower in Banff National Park (BNP) than in Waterton Lakes National Park (WLNP; 2.5 vs. 3.5, respectively; p = 0.0039).  Factorial correspondence analysis found native populations to cluster closely with low levels of variation (He=0.17).  In contrast, populations stocked into previously fishless habitat are widely scattered in the plot and have the highest levels of genetic variation (He = 0.43).  Those containing both native and introduced stocks appear intermediate (He = 0.29).  The authors suggested that the high levels of variation in introduced populations could be due to the nature of past stocking in the region.  The majority of lakes stocked in WLNP were fishless prior to introduction so it may be that the lack of competition with sympatric species allowed more of the introductions to become established, resulting in a genetically heterogeneous mix. Importantly, while levels of variation were lower in native populations, the amount of genetic divergence between them was significant. Genetic subdivision in BNP was, in fact, greater than in WLNP (Fst 0.45 vs. 0.19, respectively).

A more recent study addressing rates of hybridization among WCT populations over a larger area in Alberta provides some data on the levels of genetic variation in that region.  Janowicz (2004) reported levels of genetic variation at six microsatellite loci that were consistent with studies in other portions of the range (e.g., Leary et al. 1987; Taylor et al. 2003). Variability was generally low in the study’s reference WCT populations (Job Lake, Picklejar Lakes #2 and #4, and Marvel Lake), averaging 3.3 alleles per locus and heterozygosities of 0.36. When a larger subset of WCT populations identified as “pure WCT” as part of the hybridization assay were included, a larger number of alleles per locus were found ranging from 4 – 21, with marginally higher heterozygosities.  The authors did not provide a discussion of population structure; however, barriers again appeared to be a significant factor influencing levels of genetic diversity and genetic divergence.

Two microsatellite loci (Omy77 and Ssa85) were shared between these three studies (Table 1) and allowed for some comparison between the two regions. The allelic size range is essentially the same across regions for Omy77, and slightly larger in Alberta for Sfo8.  However, for both loci, there are fewer alleles across the allelic size range in Alberta than in BC.

 

Table 1.  Microsatellite loci (Omy77 and Ssa85) allelic size range for Alberta and British Columbia populations (most common allele in parentheses).
Source Area Omy77 Ssa85
Taylor et al. 2003 SE BC 80 - 140 bp (110 bp) 100 - 164 bp (136 bp*)
Potvin et al. 2003 BNP, WLNP 85 - 141 bp (85 bp) 91 - 191 bp (137 bp*)
Janowicz 2004 AB 79 - 107 bp (81 bp) 137 - 155 bp (141 bp)
*likely same allele; different scoring systems

The reduced subset of alleles in Alberta is not unexpected considering that WCT likely recolonized Alberta through headwater transfers across low-lying mountain passes from BC (McPhail and Lindsey 1986). Serial founder events associated with recolonization of Alberta early during the deglaciation process could have led to such a pattern and have been observed in other species in the region (e.g., Costello et al. 2003). Although it was not possible to directly compare allele frequencies at the shared loci because of differences in allele scoring between studies, it is apparent that the most common allele at these two loci differed between the two regions. In BC, Omy77*110 and Ssa85*136 are the most common alleles while over a wide range in Alberta, Omy77*81 and Ssa85*141 predominate. This, and the lack of recent dispersal opportunities between the two regions, suggests that significant genetic differentiation likely exists between the two regions. The increased isolation of populations in headwater stream reaches further suggests that the majority of populations in Alberta may show an even greater degree of reproductive isolation and demographic independence than that observed in British Columbia.

 

Designatable units

In light of the disjunct distribution of populations across the Rocky Mountain divide and expected genetic differentiation between regions, it is appropriate that there be two Designatable Units (DUs) within the subspecies for WCT in Canada:

  1. AlbertaDU
  2. British ColumbiaDU

Recognition of the two DUs is supported by the biogeographic ecozones inhabited by the two groups: Alberta populations inhabit National Freshwater Ecological Area 4 (Saskatchewan-Nelson) while populations in British Columbia inhabit National Freshwater Ecological Area 11 (Pacific); while these ecozones are adjacent, they are separated by the Rocky Mountains.

 

Assessed populations

B.C. and Alberta populations of WCT have experienced a large degree of manipulation by humans. The results of these manipulations, particularly those involving stocking activities, raise questions as to which populations (or individuals within populations) are representative of the original range and diversity of the DUs and can be legitimately included for assessment purposes. It should be qualified that this discussion refers only to the portion of the COSEWIC assessment that counts existing populations.  All WCT and related hybrid populations should be considered in evaluating conservation threats.  The following section provides guidelines for determining which populations should be included in counts.

In general, only native, genetically-pure populations within the original WCT distribution should be included in the count of remaining WCT populations at this time.  However, the following situations involving ‘managed populations’ may also be included in the count:

  1. A population from a ‘pure’ source (from within the native range of the original DU) that is introduced to a new location (usually also in the native range of the original DU) as a sanctioned recovery or management activity designed to conserve the DU (e.g., genetic refugium);
  2. A population from within a DU that has been supplemented (i.e., conservation-based activities using hatchery or wild stock to increase natural production) by hatchery (or wild stock) additions with the source for the latter originating from a population within the same DU, as part of formally sanctioned recovery or management activities, where persistence of the population is not solely dependent on supplementation;
  3. A naturally reproducing population within a DU that has reportedly been stocked with WCT at least once (usually to augment fishery, not to increase natural production) but with no evidence to indicate that the receiving system was originally WCT-free or that the existing population has been genetically altered by introductions; and
  4. A population within a DU showing evidence of <1% introgression with RT or other CT subspecies. Below this level of introgression, the population is assumed to be non-hybridized since it is difficult, if not impossible, to distinguish between intra-specific polymorphism and a slight amount of introgression (see Allendorf et al. 2001, 2004, Allendorf et al. 2005).

In contrast, hybrid and backcrossed individuals are not WCT and do not contribute to the DU.  Hybridized populations of WCT may contain pure individuals that could be used for captive breeding purposes as a tool for recovery.  However, hybrid populations with introgression greater than 1% should not generally be included in population counts (Allendorf et al. 2001, 2004, 2005, also see note below*).  Other situations where managed populations of WCT should not be counted for assessment purposes include:

  1. A population from a ‘pure’ WCT source introduced to a new location outside of the native range of the original DU (except under rare cases when this might occur for sanctioned recovery and conservation activities where habitat within the original DU no longer remains);
  2. A population of WCT introduced into a system that did not originally contain native WCT (e.g., fishless lake), outside of sanctioned recovery or management activities designed to conserve the DU; and
  3. A population with > 1% introgression (but see note* below).

*Note: Hybridization is a complicated issue, and hybrid populations should be considered on a case-by-case basis for assessment, protection and recovery purposes as some populations may still have some conservation value (Allendorf et al. 2001).  Such populations may be recoverable through captive breeding efforts, or may have elevated value if very few ‘pure’ populations remain within a DU.  With respect to the survey work that has been conducted for hybrids in WCT populations in Canada, survey design including geographic scope should be considered.  An introgression estimate in a small tributary based on a representative sample is not equivalent to an introgression estimate based on limited sample size for a large system such as the upper Kootenay River.  Furthermore, small tributaries may contain physical or temperature barriers to upstream or downstream movement; thus limiting hybridization spatially.  Thus, finding hybridization in the downstream section of a system (e.g., lower Bull River) does not mean the entire tributary is affected (e.g., upper Bull River is still ‘pure’).

Page details

Date modified: