Wood turtle (Glyptemys insculpta) COSEWIC assessment and status report: chapter 8

Limiting Factors and Threats

There are several threats, of varying degrees of imminence and potential severity, to the Wood Turtle throughout its range. Imminent threats with population-scale impacts include: increased mortality rates, particularly of adults, by traffic on roads (Anon., 2002; Seburn and Seburn, 2004; R. Brooks, pers. comm. 2004); modern agricultural machinery (Saumure, 1997; Anon., 2002; Saumure, 2004 and pers. comm. 2006, 2005); collection of all age classes, but especially of adults for personal and commercial use as pets and food (Lazell, 1976; Harding and Bloomer, 1979; Garber and Burger, 1995; Anon., 1996; Litzgus and Brooks, 1996; Seburn, 1997; Galois and Bonin, 1999; Cameron and Brooks, 2002; R. Brooks, pers. comm. 2004; J. Harding, pers. comm. 2004; R. Saumure, pers. comm. 2004); increased predation of nests and females by larger populations of edge predators (increase in edge predators due to increased fragmentation of habitat by forestry, agriculture and urban expansion) (Oldham, 1998; NatureServe, 2004; Bourgeois et al., 2004; R. Saumure, pers. comm. 2004) and destruction of nests by vehicles such as ATVs (J. Trottier, pers. comm. 2004; R. Knudsen, pers. comm. 2004; P. Brewer, pers. comm. 2004, 2005). On land, Wood Turtles are exceptionally vulnerable to collection, especially in spring when they bask and feed along watercourses before vegetation leafs out and makes the turtles much harder to find. Thus, the turtles are easily found and readily captured as they have no capacity to escape when on land. Construction of new forest access roads may increase the potential for collection to occur, as previously inaccessible areas become readily accessible, and often heavily travelled by outdoors people. As well, these new roads and water crossings may provide new nesting locations, which are less than ideal due to the vehicular traffic (trucks and ATVs) associated with them (Trute et al., 2004), and therefore, act as population sinks.

Like most turtles, Wood Turtles are long-lived and vulnerable to chronic increases in rates of mortality of adults or older juveniles. Recent modelling studies suggest that a chronic annual increase in adult mortality rate as low as 1% can lead to extirpation of a population of Wood Turtles (Compton, 1999). In other turtle species with similar age of maturity and reproductive output, increases of 5% in rates of adult mortality lead to population declines, whereas an annual mortality of 70% of eggs can be tolerated, all else being equal (Congdon et al., 1993: based on models from long-term study of a population of Blandings Turtle (Emydoidea blandingii)). Wood Turtle females reproduce only once per year at most and lack the ability for compensatory recruitment if population sizes decrease (Brooks et al., 1991, 1992).  Effects on population structure of mortality associated with roads and vehicles are demonstrated in a recent study of sex ratios in turtle populations associated with different densities of roads nearby. In Chrysemys picta and Chelydra serpentina, in 18 wetlands surrounded by “low” road density (>1.5 km of roads/km² of landscape) and 17 wetlands with “high” road density (> 1.5 km of roads/km² of landscape) in New York, significantly male-biased sex ratios were associated with high road densities, but not with low road densities (Steen and Gibbs, 2002). The authors concluded that this difference was a consequence of biased mortality of females on the roads. A recent meta-analysis compared sex ratios of 38,166 turtles from 157 studies and found a consistently larger proportion of females in populations sampled along roads (61%) than those sampled off roads (41%) (Steen et al. 2006). The authors concluded that the observation that freshwater turtle populations are becoming more male-biased was because females are more likely to go on roads and be killed than are males (Steen et al. 2006). For Wood Turtles, the sex ratios were 0.51 and 0.68 female:male for off and on road samples respectively (Steen et al. 2006). Also, telemetry studies of Wood Turtles usually show that males remain closer to water compared to females (Foscarini 1994, Walde 1998, Smith 2002) and that females often nest in agricultural fields (Saumure and Bider, 1998; Saumure, 2004) or on roads. Thus, female turtles are more likely to be killed, and even when they are not, their hatchlings in roadside nests are often killed (Ashley and Robinson, 1996). Nesting success and survival of hatchlings over their first year is extremely low (usually between 0-30%). On average, adult females are likely to nest successfully (i.e. “clutch” survives to maturity) only once in their, often long, reproductive lifetime. One population in Ontario is predicted to become extirpated within 50 years, because over a few days, collectors removed about 60% of the adult population (Cameron and Brooks, 2002). Two populations in Connecticut were extirpated within 10 years after allowing human access to a reserve used for picnics and family outings (Garber and Burger, 1995). These authors speculated that the decline and extirpation of the population occurred because of occasional removal of turtles by hikers and picnickers. A population in Quebec has declined because female adults are killed by agricultural equipment (Saumure, 2004; Daigle and Jutras, 2005), and such losses are occurring in other parts of the Wood Turtle’s range in agricultural areas (R. Saumure, pers. comm. 2004).

Therefore, humans contribute to Wood Turtle mortality in several ways, including road kills (Brooks et al., 1992; Seburn, 1996); injury or killing of Wood Turtles with farm machinery (Saumure and Bider, 1998; Saumure, 2004; M. Pulsifer, pers. comm. 2005); destroying nests with ATVs, dirt bikes, or 4WDs (R. Knudsen, pers. comm. 2004); and intentional shooting of turtles (Harding and Bloomer, 1979; Litzgus and Brooks, 1996). Collection for the pet trade does not directly cause mortality, but removes adults from the population, which, given the species’ long-lived life history with its consequent low annual reproductive output, greatly reduces recruitment, and collection has most likely been the cause of some drastic declines in the past (Lazell, 1976; Harding and Bloomer, 1979; Garber and Burger, 1995; Anon., 1996; Litzgus and Brooks, 1996; Galois and Bonin, 1999; Cameron and Brooks, 2002; R. Brooks, pers. comm. 2005; R. Saumure, pers. comm. 2004).

The crux of the problem for the Wood Turtle is that even compared to other late maturing, long-lived turtles, this species is exceptionally vulnerable to increased loss of adults because of its terrestrial habits and extreme “tameness” (i.e. it can’t evade capture when on land). Thus, even casual collection for pets (Garber and Burger, 1995), when added to the “usual” commercial collection, road mortality, and mortality from farm machinery, increased predation by racoons and coyotes and mortality from off-road vehicles, leads inevitably to population declines and extirpation if these losses are not mitigated more or less completely.

Perhaps a growing threat is collection for the exotic food trade (a US turtle researcher in Wisconsin tracked a transmittered turtle to a dumpster and found remains of over 60 other Wood Turtles that had been killed for food by one individual) (R. Saumure, pers. comm. 2004). Recently, different anonymous sources have reported directly to the A and R cochair examples of extensive private “harvesting” of turtles (and other taxa) for food (e.g., students on the University of Guelph campus capturing and keeping wildlife in their apartments for food or sale) (Anon. pers. comm. 2006) and the existence of a network to provide native turtles and other wildlife to restaurants in southern Ontario (Anon. pers. comm. 2007). Another threat is from the loss of nesting habitat and hibernacula due to stream and riverbank alteration and stabilization (Galois and Bonin, 1999; D. Coulson, pers. comm. 2004; Wesley 2006). Lesser threats include being buried alive during bank stabilization (Saumure, 2004); and perhaps pollution of waterways inhabited by Wood Turtles (Ernst, 2001; Trute et al., 2004).

Some attempts at mitigating the effects of collecting, agriculture and forestry are being made, including: a website listing captive breeding and a registry thereof to undercut the black market pet trade (R. Saumure, pers. comm. 2004); suggestions for agricultural methods that decrease the likelihood of injuring or killing turtles, though these are entirely unlegislated and voluntary (R. Saumure, pers. comm. 2004); forest management planning guidelines for the protection of Wood Turtle habitat (Anon., 2002; Trute et al., 2004; Wesley 2006); and headstarting of hatchlings (M. Malhiot, pers. comm. 2004). These are all quite recent initiatives, however, and their effectiveness has not been tested.

One of the interesting properties of the Wood Turtle’s current distribution is that it occurs discontinuously across its range (Ernst et al., 1994). Many researchers have noted that there seem to be many suitable rivers and creeks in areas occupied by Wood Turtles that are not used by the turtles (but see Mitchell et al. 1997; Boyd and Brooks, 1998; Wesley 2006). These absences could be explained by poor dispersal capabilities or key habitat requirements that are missing (Wesley 2006), or by random extinctions characteristic of small, isolated populations where isolation is accentuated in some areas by anthropogenic activity. M. Elderkin of Nova Scotia Wildlife has proposed that these patterns of presence and absence may reflect other historical events that extirpated turtles from some areas (M. Elderkin, pers. comm. 2005). Basically, Elderkin suggests that logging drives extirpated Wood Turtles, and perhaps other species. In Nova Scotia, the drives occurred in May and early June when the turtles would still be in the rivers, and occurred through about 250 years ending around 1950 (M. Elderkin, pers. comm. 2005). Logging drives could kill turtles directly and scarify and destroy nesting beaches, and oxbows in which the turtles spend much of their time (R. Knudsen, pers. comm. 2004; Wesley 2006). Although, logging drives no longer are a significant threat, they may have destroyed much Wood Turtle habitat and populations and may still have an effect on current abundance and distribution of the species. In a similar vein, some populations may have been extirpated by harvesting for food by Aboriginals as Wood Turtle remains do occur in native middens, or for pets by the much larger post-European settlement population.

Page details

Date modified: