Wood turtle (Glyptemys insculpta) COSEWIC assessment and status report: chapter 3

Species Information

Name and classification

Family:

Emydidae

Species

Glyptemys insculpta (Agassiz 1857) (No subspecies recognized)

The common name for this species is Wood Turtle, or, in French, tortue des bois. The scientific name was recently changed from Clemmys insculpta to better reflect the genetic relationships of species previously included in the genus Clemmys (Feldman and Parham, 2002; Holman and Fritz, 2001; NatureServe 2004). Emys orbicularis, Emydoidea blandingii and Clemmys marmorata were designated a monophyletic group, and Clemmys muhlenbergii and Clemmys insculpta were placed in a second monophyletic group (Feldman and Parham, 2001; Holman and Fritz, 2001). The type species for the genus Clemmys (Clemmys guttata) was retained as the only member of that genus (NatureServe 2004), Clemmys muhlenbergii and Clemmys insculpta were placed in the genus Glyptemys (Agassiz 1857), as recommended by Feldman and Parham (2002).

Morphological description

The Wood Turtle is a medium-sized turtle with adult carapace length ranging between 16cm and 25cm (Litzgus and Brooks, 1996, Smith, 2002). The grayish-brown to yellow carapace is broad and low, sometimes having dark lines and dots. Each scute has pyramidal concentric ridges (growth lines), giving the carapace a sculptured appearance. The carapace is strongly keeled and is serrated at the posterior margin (Babcock, 1971; Litzgus and Brooks, 1996). In older turtles, the ridges on the scutes may become worn smooth to some degree. The plastron lacks a hinge and is yellow with black splotches on the outer posterior corners of each scute in patterns that vary among individuals. The skin is generally brown, and the legs and neck often have yellow, orange or reddish colouring. The feet are slightly webbed with strong claws. The irises of the eyes are yellow or brown, and the upper jaw forms a beak, arched downwards over the slightly shorter lower jaw. Males are larger than females (Foscarini, 1994), and have a deeply notched plastron that generally contains grooves running cranio-caudally. The plastron is flat in juveniles and adult females, but it becomes strongly concave in males as they reach sexual maturity. In adult males, the cloacal vent is distal to the posterior margin of the carapace, whereas in females it is not. The male also has a longer thicker tail than does the female.

Figure 1. Wood Turtle, Glyptemys insculpta (by Rosemarie Schwab) (reproduced with the permission of the Ministère des Ressources naturelles et de la Faune, du Québec).

Figure 1. Wood Turtle, Glyptemys insculpta

Genetic description

A recent study of the phylogeography of Wood Turtles using 750 bp of the mitochondrial control region sampled 117 turtles from 29 localities across the species’ range (Amato et al. in press). Twenty-one haplotypes were identified and there was little genetic variation, which is typical of turtles in general and the genus Glyptemys in particular (Avise et al. 1992, Rosenbaum et al. 2007). Nested clade analysis indicated a main postglacial dispersal up the east coast from a southern refugium to Nova Scotia with subsequent westward dispersal (Amato et al. in press). A BEAST analysis using a Bayesian skyline plot, indicated the Wood Turtle population had been growing rapidly over the last 12,000 years. One clade is represented along the eastern USA and Canada and in the west into states south of the Great Lakes. A second clade occurs in Ontario and adjacent Quebec west of the St. Lawrence, although some presence of the first clade was also found in this region. In contrast to work on mitochondrial DNA,  genetic studies on six populations of Wood Turtles in Quebec found these populations were highly polymorphic and characterized each of the six using five microsatellite loci (Tessier and Lapointe, 2002; Tessier et al., 2005). There was high variability within all populations indicating that putative past declines have not yet led to significantly reduced variability, although the most genetically distinct populations had the lowest diversity (Tessier et al., 2005). Ultimately, it appears that there were three genetically different units, two on the north shore of the St. Lawrence, and a single homogeneous group (of four populations) south of the river (Tessier et al., 2005). Despite their small size, these populations showed high levels of heterozygosity and allelic diversity (H0 ranging from 0.561-0.886, an average of 10 alleles per locus per population, Tessier et al. 2005).These data suggest that these Wood Turtles have genetic variation expected of a relatively abundant species and that until “recently” (given the long generation times of the species) these populations existed in “long-term genetic neighbourhoods” comprising several thousand individuals  based on equilibrium considerations and the stepwise mutation model (Ohta and Kimura, 1973; P. Bentzen, email communication, Nov. 22, 2007). These conclusions further suggest that these populations have undergone rapid, large declines, recent enough that they still show little genetic evidence of inbreeding despite their small sizes and relative isolation.

In Canada, most Wood Turtle populations tend to be associated with watersheds that are often widely separated, and thus the populations become isolated (Foscarini, 1994; Arvisais et al., 2002, 2004; Smith, 2002; Seburn and Seburn, 2004; Wesley et al., 2004, Tessier et al., 2005, Wesley and Brooks, 2005), because turtles tend to move along streams and rarely move between streams even when they are only a few kilometres apart (e.g. Foscarini, 1994, Foscarini and Brooks, 1997).  The observed genetic distinctness among nearby populations is likely a reflection of low vagility.  For example, in a long-term study (>15 years) located on two Ontario creeks that were less than 5km apart and were tributaries of the same river, no turtle was ever recorded in both streams even though virtually every turtle was individually marked throughout the study (Cameron and Brooks, 2002).

Designatable units

Across its Canadian range, the Wood Turtle is exposed to different degrees of threat.  In southern Ontario, and in Quebec around Montréal, the species has been extirpated or has declined (Table 1), probably from increased mortality on roads, private and commercial collection as pets, and loss of key habitat features (Wesley, 2006) caused by alteration and pollution of streams and their associated riparian habitats.  The species has likely not declined much over the most northern parts of its range in these provinces, but there its density is low and numbers are small (Walde, 1998, Smith, 2002, Wesley, 2006) particularly in Ontario.  On the other hand, in Nova Scotia, limited data suggest that one or two watersheds have populations that are comparatively robust, and it is possible that the species has larger numbers in this province and that some populations are still relatively secure.

The different status of the Wood Turtle across its range in Canada is somewhat reflected by biogeographic and phylogeographic distinctions.  The species’ Ontario distribution is/was in the Canadian Shield, Great Lakes/St. Lawrence and Carolinian herpetofaunal provinces (COSEWIC Operations and Procedures Manual F5, p. 5).  The species is now effectively extirpated from the Carolinian province, and from almost all of the southern and eastern parts of the Great Lakes/St. Lawrence (Mitchell et al., 1997; Boyd and Brooks, 1998; Galois and Bonin 1999; Cameron and Brooks, 2002; NHIC, 2004ab; Seburn and Seburn, 2004; Equipe de retablisement des tortues du Quebec, 2005). Populations of the species occupying the Canadian Shield herpetofaunal province may be declining and threatened in the southern part of this distribution and comparatively secure in the north (Table 1). Shield populations tend to be small (<200 adults, Table 1). and isolated by topography (Wesley 2006). Presumably, Shield populations were stable until recently (Tessier et al. 2005, see comment by P. Bentzen above in section on Genetic Description), but are increasingly threatened by new road access and associated activities (Arvisais et al. 2002, 2004; Saumure 2004; Seburn and Seburn 2004; Smith 2002; Crowley 2006). Finally, many of the Appalachian/Atlantic Coast populations of Wood Turtles appear to be under less immediate threat than those further west, but they have been less studied (see Population trends/ Abundance). 

Despite these differences among the Faunal Provinces, there are no clear distinctions in genetic structure (Tessier et al., 2005, Amato et al. in press) in either microsatellites or mitochondrial genes, certainly none that can be associated with specific faunal provinces. Furthermore, there are no clear disjunctions among populations according to Faunal Province, indeed many populations seems to span the boundaries of these (see map in Figure 3). Finally, the putative conservation status also does not match well with the distribution in the Faunal Provinces. Therefore, potential separate Designatable Units based on the criteria for assigning such units (see COSEWIC O and P Manual, Appendix F5, Nov. 2007) do not appear defensible at this time.

Page details

Date modified: