Misty Lake sticklebacks COSEWIC asessment and status report: chapter 3

Species Information

Name and classification

Class:
Actinopterygii (ray-finned fishes)
Order:
Gasterosteiformes
Family:
Gasterosteidae
Genus:
Gasterosteus
Lentic (Lake) form:
Gasterosteussp.
Lotic (Stream) form:
Gasterosteussp.

Common name

English:
Misty Lake Lentic Stickleback, Lake Form
Misty Lake Lotic Stickleback, Stream Form
French:
Épinoche lentique du lac Misty
Épinoche lotique du lac Misty

The marine form of the threespine stickleback, sometimes called trachurus in the literature, migrates to freshwater to spawn (sea-run) while the freshwater type, referred to as leiurus, remains in fresh water for its entire life (Scott and Crossman 1973). The taxonomy of Gasterosteusaculeatus has been known to be problematic for some time (see e.g., Hagen and McPhail 1970; Scott and Crossman 1973). Classical approaches to systematics are based on morphometrics and assume that morphological traits rarely evolve in parallel. However, parallel evolution is widespread in freshwater sticklebacks and their taxonomy does not reflect the vast diversity that exists (e.g., Hagen and McPhail 1970; Hagen and Gilbertson 1972; Lavin and McPhail 1985; Reimchen et al. 1985; Schluter and McPhail 1992; Bell and Foster 1994; McPhail 1994). McPhail and Lindsey (1970), and Bell and Foster (1994) consider G. aculeatus to be a species complex.

Thompson et al. (1997) noted that the radiation of forms within G. aculeatus found in the eastern North Pacific is characterized partly by three sets of divergences involving sympatric or parapatric biological species (reviewed by McPhail 1994): parapatric sea-run and freshwater sticklebacks (e.g., Hagen 1967); sympatric lacustrine limnetic and benthic forms (Larson 1976; McPhail 1984, 1992); and parapatric lake dwelling (lentic) and stream dwelling (lotic) sticklebacks (Moodie 1972, Reimchen et al. 1985; Lavin and McPhail 1993).  Highly divergent parapatric lake-stream pairs of sticklebacks are known to occur in three drainages: Mayer and Drizzle lakes, on northeast Graham Island (Queen Charlotte Islands), and Misty Lake on northeastern Vancouver Island (Fig. 1). These forms are endemic to British Columbia (BC). In particular, the magnitude of the ecological and genetic divergence seen in the Misty Lake parapatric lake-stream pair is similar to that found in benthic-limnetic species pairs (e.g., Paxton Lake and the Vananda Creek drainage on Texada Island, BC) (Taylor, pers. comm. 2005). Although several other lake-stream population pairs of sticklebacks have since been found in BC (Hendry and Taylor 2004), to date, none are known to combine the genetic, morphological and ecological differences observed in the Misty Lake pair (Taylor, pers. comm. 2005).

Figure 1. MistyLakesticklebacks – the upper fish is typical of the lake form and the lower fish is an example of the inlet stream form (photo courtesy of Dr. Eric Taylor, UBC).

Figure 1. MistyLakesticklebacks – the upper fish is typical of the lake form and the lower fish is an example of the inlet stream form (photo courtesy of Dr. Eric Taylor, UBC)

Hendry et al. (2002) conducted a detailed study of the sticklebacks from different sites in the Misty system. The authors concluded that morphological differences were greatest between the lake sticklebacks and the inlet sticklebacks (both in specimens collected in the wild and in crosses raised in common garden experiments conducted to determine if the differences seen in the wild had a genetic basis). These differences were thought to have evolved in response to their different foraging environments (Lavin and McPhail 1986; McPhail 1994). Genetic divergence (mitochondrial DNA and microsatellites) was also greatest between the sticklebacks from the lake and those in the inlet and least between those from the lake and fish from the outlet stream (Hendry et al. 2002). In their study on adaptive morphology, Moore and Hendry (2005) concluded that gene flow did not constrain adaptation in the inlet stream but strongly constrained adaptation in the outlet stream; unlike inlet fish (where a sharp shift in morphology occurred between lake and inlet) the outlet fish were characterized by a gradual increase in divergence from the lake fish as distance from the lake increased.

Morphological description

Generally, the threespine stickleback is a small fish (average about 51 mm in standard length); the body is compressed laterally and elongate, tapering to a slender, depressed caudal peduncle; there are three isolated, stout, serrated dorsal spines; the very short last spine, precedes, but is not attached to the soft rays; pelvics are thoracic with one strong spine and 1 soft ray; all fin spines can be locked in an erect position (Scott and Crossman 1973).

In the systems where highly divergent parapatric lake-stream pairs of sticklebacks have been observed the morphological divergence is almost identical (Lavin and McPhail 1993). The inlet stream form is mottled brown and robust bodied while the lake form is melanistic, slender-bodied and has longer spines and more gill rakers than its corresponding parapatric stream form; the same ecotypes from different drainages are morphologically more similar than different ecotypes from the same drainage (Lavin and McPhail 1993).

In their study, Hendry et al. (2002) found that in the Misty Lake system, lake and outlet fish were usually longer than inlet fish (Table 1). Other data showed Misty Lake inlet sticklebacks usually had deeper bodies than either Misty Lake or Misty Lake outlet fish; lake fish had a higher number of gill rakers than stream fish; Misty Lake inlet fish had shorter pelvic spines than either Misty Lake or Misty Lake outlet fish; and Misty inlet fish had wider pelvic girdles than either lake or outlet fish. In morphology, lake and inlet fish were at the extremes while outlet fish were found to be intermediate, but more similar to lake fish (Moore and Hendry 2005).

Morphological differences between Misty Lake and inlet sticklebacks are inherited: crosses between stream forms produced only stream-form phenotypes; crosses between lake forms produced only lake-form phenotypes (Lavin and McPhail 1993; Hendry et al. 2002). This indicates that the two forms comprise separate gene pools, not a complex polymorphism (Lavin and McPhail 1993).

Table 1. Average morphological measurements of threespine stickleback from different collection sites in the MistyLakesystem (adapted from Table 1, Hendry et al. 2002). Body depth, pelvic spine length and pelvic girdle width were standardized to a common body length of 55.4 mm. U indicates upper site (1.8 km upstream) and L the lower inlet site (0.9 km upstream). Homogenous subsets of collections based on Tukey tests are indicated with letter superscripts.
 
Upper Inlet 1997 Upper Inlet 1999 Lower Inlet 1997 Lower Inlet 1998 Lake 1997 Lake 1998 Outlet 1998
N
7 30 21 30 30 30 29
Body length
51.6b,c 49.1b 36.2a 52.3b,c 62.8d 61.6d 58.1c,d
Gill raker number
15.9a 16.7a,b 16.8a,b 16.1a,b 19.5d 18.8d 16.8a,b
Body depth
14.5f 13.6d 14.2e,f 13.8d,e 12.4c 11.6b 12.7c
Pelvic spine length
8.1a 8.7a,b 8.2a 9.0b,c 9.3b,c 9.2b,c 9.4c
Pelvic girdle width
4.6e 4.1d 4.5d,e 4.3d,e 3.1a,b 2.9a 3.2a,b,c

Genetic description

Different scales of evolutionary divergence occur in Gasterosteus of the North Pacific, an ancient divergence of mitochondrial DNA clades as well as a more recent, postglacial, divergence of ecotypes within the major clades (Thompson et al. 1997). Although it was initially only found around the Queen Charlotte Islands, Thompson et al. (1997) extended the known distribution of the Japanese (Argonaut Plain) clade to the Misty Lake drainage system and referred to it as the Trans-North Pacific clade (TNPC). Both the TNPC and the Eastern North Pacific lineage of the Euro-North American clade (ENAC) are present in the Misty Lake system (Fig. 2). The TNPC is much less common in British Columbia than the ENAC, although Johnson and Taylor (2004) have found it to be more common and distributed more widely than previously thought. It is now known to occur in 12 of the 45 new locations surveyed in their study: four of nine Cook Inlet (Alaska) lake populations; three of four Quadra Island lake populations; one of 15 lake populations on northern Vancouver Island; one of four adjacent mainland populations; and four of 12 anadromous or marine populations. Johnson and Taylor (2004) reported a significant association between the presence of the TNPC and lake elevation; it was more common in lakes below 42 m than in lakes at higher elevation. This factor indicates that post-glacial lake accessibility was important in determining the distribution of the two clades in the eastern Pacific (Johnson and Taylor 2004).

Figure 2.

  1. (a) Relationships among mtDNA haplotypes found in stickleback populations sampled from northern Vancouver Island and Graham Island, Queen Charlotte Islands, including Misty Lake inlet (predominantly haplo 2, plus haplo 3), and Misty Lake and outlet (predominantly haplo 3, plus 2 and 4). “A” and “B” refer to two major clades of mtDNA, the “Trans-North Pacific Clade (TNPC)” and the “Euro-North American Clade”. Numbers on branch points represent % support from 1000 bootstrap replicates.
  2. (b) Neighbour-joining tree of estimated relationships among populations of lake and stream sticklebacks from Vancouver and Graham islands.Fish “cartoons” represent general shape and shading differences (both figures adapted from Thompson et al. 1997). 
Figure 2. (a) Relationships among mtDNA haplotypes found in stickleback populations sampled from northern Vancouver Island and Graham Island, Queen Charlotte Islands, including Misty Lake inlet (predominantly haplo 2, plus haplo 3), and Misty Lake and outlet (predominantly haplo 3, plus 2 and 4).

Hendry et al. (2002) were able to show that genetic divergence (mitochondrial DNA and microsatellites) was greatest between Misty Lake and the upper inlet (1.8 km upstream of the lake), intermediate between the lake and the lower inlet (0.9 km upstream) and least between the lake and the outlet stream (1.2 km downstream of the lake). The TNPC was nearly fixed in upper inlet fish (95.5%), dominant in the lower inlet (67.5%), less common in outlet fish (20.5%) and rare in Misty Lake (4.4%); the TNPC was predominant in the inlet stream while the ENAC was dominant in lake and outlet stream fish (Thompson et al. 1997; Hendry et al. 2002). This juxtaposition of two populations each with a preponderance of a different clade appears to be unusual in these parapatric pairs.

Hendry et al. (2002) showed that within the Misty Lake system all pairs of collections (site- and year-specific) differed significantly in allelic frequencies at 5 nuclear DNA microsatellite loci (P < 0.010) and pairwise FST values were all significantly greater than zero (P < 0.001). Differentiation between years and within sites was small (FST = 0.023 – 0.030). Differentiation among sites ranged from small (lake vs. outlet, FST = 0.005 – 0.046) to moderate (lower inlet vs. lake, FST = 0.129 – 0.157) to large (upper inlet vs. lake, FST = 0.289 – 0.345), demonstrating concordance between nuclear and mitochondrial DNA differentiation. Misty Lake and outlet collections clustered together and were distinct from the lower inlet and upper inlet collections (98% bootstrap support). Lower inlet and upper inlet were also quite distinct from each other (99% bootstrap support), indicating the stream form may comprise more than one population (Fig. 3).

Figure 3. A dendrogram of Nei’s (1978) unbiased genetic distance showing general relationships of Misty Lake stickleback from different habitats to each other and to other stream-lake pairs. Labels at the end of terminal branches refer to specific collections, with a location abbreviation followed by the collection year. O, outlet; I, inlet; and L, lake. Additional abbreviations appear in the Misty watershed (UI, upper inlet), the Boot/Merrill watershed (ML, Merrill Lake; BL, Boot Lake; O/I, the stream connecting these two lakes), and the Amor de Cosmos watershed (LS, the southern McCreight Lake collection, which was adjacent to the inlet; LN, the northern McCreight Lake collection, which was adjacent to the outlet). Bootstrap values are shown for all nodes and genetic distance is based on variation at five microsatellite loci. See Hendry and Taylor (2004) for details.

Figure 3.  A dendrogram of Nei’s (1978) unbiased genetic distance showing general relationships of MistyLakestickleback from different habitats to each other and to other stream-lake pairs. Labels at the end of terminal branches refer to specific collections, with a location abbreviation followed by the collection year. O, outlet; I, inlet; and L, lake.

The estimated effective number of migrants between sites (total exchanged) was lowest between Misty Lake and upper inlet at 0.54 migrants per generation, slightly higher between Misty Lake and lower inlet at 1.59 migrants per generation, and much higher between Misty Lake and outlet at 18.23 migrants per generation (based on Wright’s FST calculations, Hendry et al. 2002).

Designatable units

All of the stickleback populations in the Misty Lake watershed are important components of the evolutionary processes that are occurring in this system. The fact that these populations have diverged and co-exist as parapatric pairs is unusual and of conservation value. The morphological, genetic and ecological information indicate that the Misty Lake inlet stream and lake populations are divergent and reproductively isolated, behaving as two biological species. The two populations should be considered distinct designatable units much like benthic-limnetic stickleback species pairs. Currently, the outlet fish do not appear to meet the criteria for a separate designatable unit (DU) due to the reported similarities between the lake and outlet populations. Gene flow from the lake population to the outlet population is high enough to constrain adaptation in the outlet population, although some degree of ecological divergence is evident (Moore and Hendry 2005). For now, the outlet population could be considered a component of the lake DU.  The outlet population contributes to the overall Misty Lake lake-stream complex and may have population dynamics separate from either the inlet stream or lake populations.

Thompson et al. (1997) argue that their molecular data strongly suggest the lake-stream pairs on Graham Island and northern Vancouver Island evolved separately through parallel evolution: neither the Misty Lake form nor its corresponding inlet stream form clustered with similar forms from the Drizzle or Mayer lakes’ systems; the Misty Lake lake-stream pair did not share any mtDNA haplotypes with either of the pairs from Graham Island; and the haplotypes were phylogenetically independent from the Graham Island haplotypes. Microsatellite data (Hendry and Taylor 2004) support the conclusion that all of the pairs from different watersheds have a long history of evolutionary independence and represent independent events of post-glacial divergence.

Data from Thompson et al. (1997) also support an allopatric divergence for mtDNA clades seen in the Misty Lake pair, because mtDNA nucleotide divergence between the lake and inlet stream populations was estimated at 1.75% indicating that the predominant clades in the Misty Lake and Misty Lake inlet populations diverged about 875,000 years ago (assuming a whole molecule substitution rate of 2% per million years), whereas Misty Lake and its surrounding area were glaciated only 12,000 years ago. The two divergent mtDNA clades observed in the Misty Lake pair are believed to have originated from historical isolation in the two major glacial refugia proposed for the North Pacific, namely the Beringia and Cascadia (Thompson et al. 1997).

Lavin and McPhail (1993) noted that the lake form was predominantly found in the lake, but small numbers of lake fish were collected in the stream as far as 300 m upstream. In contrast, none of the inlet stream fish were collected in the lake, but both forms were taken in the swamp, a transition zone, at the mouth of the inlet stream. The distribution of the two forms did overlap during breeding season. But the authors found that hybrids were rare in the wild; only one of the 274 fish collected from the Misty Lake system was found to be intermediate in morphology between the two forms, implying that hybridization is rare or hybrid survival is poor. The authors argue that this indicates that lake and inlet forms are distinct ecological and genetic entities and not an environmentally induced dimorphism within a single population.

Page details

Date modified: