Wood turtle (Glyptemys insculpta) COSEWIC assessment and status report: chapter 7

• Previous page
• Table of contents
• Next page

Population Sizes and Trends

Search effort

Prior to 1990, there were very few studies on any aspect of Wood Turtle biology in Canada, but since the species was listed as Vulnerable (Special Concern) by COSEWIC in 1996, many studies have been initiated, especially in Quebec and Ontario, that have filled gaps in our knowledge of the species’ numbers, demography, habitat requirements and distribution. A number of known populations were studied using radio telemetry (e.g., Brooks and Brown, 1991; Quinn and Tate, 1991; Foscarini, 1994; Walde 1998; Compton, 1999; Arvisais et al., 2002, 2004; Cameron et al., 2002; Compton et al. 2002; Dubois, 2006; Smith 2002; Peiman and Brooks, 2003; Saumure, 2004; Wesley et al., 2004; Wesley, 2006). Surveys of new areas usually were completed by walking (one to four people) alongside a river thought to provide good habitat, sometimes with one person walking or canoeing the river or stream. In Ontario, systematic surveys of this type were carried out across the province from the extreme southwest to western Algoma (Mitchell et al., 1997; Boyd and Brooks, 1998). Geographic Information Systems were used in some cases to focus search effort (Smith, 2002). One survey (of New Brunswick) used an Internet newsgroup to solicit reports of sightings (McAlpine and Gerreits, 1991), while a second (Nova Scotia) interviewed local residents (Adams, 2002). When population estimates were given, they were usually based on mark-recapture rates. Population estimates are not available for all of the populations, and search effort was quantified only on rare occasions (Daigle, 1996).

Abundance

The total number of Wood Turtles in Canada is not known, but given that there are several recently completed studies on Wood Turtles, and several more ongoing across the species’ Canadian range,, it is possible to derive a rough estimate of the number of adults in Canada. To obtain this estimate, the COSEWIC Amphibians and Reptiles SSC co-chair (R. Brooks) used information from provincial Recovery Teams and other resources and contacted by email in June-August 2005, all researchers and government biologists known to have an interest in Wood Turtles and asked them to estimate numbers of adult Wood Turtles in their research areas, jurisdictions or whatever area they felt competent to cover. The following information was requested; location, number of adults marked, estimated number of adults, and method of estimation. A summary of responses from Ontario and Quebec is in Table 1. From some regions, individuals were also asked to report the number of rivers searched and the number with or without Wood Turtles, and the number of rivers that may have Wood Turtles, but have not yet been searched. This endeavour allowed calculation of reasonable estimates of adult abundance in the Carolinian, Great Lakes/St. Lawrence and Canadian Shield faunal provinces in Ontario and Quebec, but estimates for the Appalachian/Atlantic Coast region were considerably more uncertain, because this region had fewer studies and surveys.

In providing estimates, some people who responded to the request for numbers assumed that the density of turtles per km of river in their study area could be extrapolated to the entire length of the river. This assumption likely gives an overestimate in that most study areas only include a limited section of a watercourse and are selected for study because they are already known to have a population of Wood Turtles. However, experience indicates that even in rivers with apparently good habitat throughout their length, the turtles are usually patchily distributed with most of the river unoccupied except by transients (R. Brooks, pers. comm. 2005; Wesley, 2006). For example, on a major river in central Ontario, virtually all sightings of Wood Turtles along a 20-km stretch of the river occurred at 2 sites, one 1.2km and the other 0.4km. in length. Extrapolation from these sites over the 20-km surveyed would have given an estimate over 2000 adults, when the real population of adults is likely fewer than 150 (R. Brooks, pers. comm. 2005). Three other examples that indicate the species’ patchy distribution follow, “In my primary study site (oldest) and other sites in Michigan, the turtles are definitely clumped in distribution, and essentially absent (or just transient) on much of the river's length. Attached is an aerial view of my study site, showing the prime habitat. Portions of the river that lack the grassy-edged sandbars are MUCH less frequently used. The sloughs are not used at all” (J. Harding, pers. comm; 2006). ”I also think that taking localized “sub-population” estimates and extrapolating them to the entire stream will result in a serious over-estimate of metapopulation size”. (P. Wesley pers.comm 2006). “As for the River X, there is a meta-population, so to speak. We have one in the agricultural village of XXX proper and my site upstream from XY Lake. Nothing in between. You can calculate rough distance on a map. So, as far as we know, there are likely 3 small populations along the entire X river that probably don't intermingle that much (my longest move for a male was just over 5km). So, does that constitute three populations or a metapopulation? As to guessing on how many turtles there are along an unsurveyed river, estimates generalized to an entire river or watershed will grossly over-estimate metapopulation size” (R. Saumure pers. comm., 2006)

In addition, it appears that there is a significant positive correlation between Wood Turtle density and number of frost-free days (Walde, 1998; Smith, 2002). Therefore, Canadian populations, especially those near the northern limits of the species’ distribution have much lower densities (up to two orders of magnitude) than those in the central part of the species’ distribution in the USA.

Populations of the Carolinian Faunal Province

There has never been much information on Wood Turtles in this faunal province although the species evidently did occur there in small scattered “populations” in historic times (NHIC 2004a,b, Seburn and Seburn 2004). Most Element Occurrences are Extinct or Historic and it is unlikely that any viable populations still occur in this region (see Figure 4). Recent surveys that focused on Wood Turtles failed to find any evidence of the species in the Carolinian Region (Mitchell et al. 1997, Boyd and Brooks 1998).

Populations of the Great Lakes/St. Lawrence Faunal Province

All extant populations of Wood Turtles in Ontario and those in the upper St. Lawrence around Montréal are in Great Lakes/St. Lawrence habitat. The only population remaining in “southern” Ontario (i.e. south of Algonquin Park) was estimated at 412-420 individuals in 1993 (Foscarini, 1994), but a 65%-75% decrease occurred in 1994-95 (Mitchell et al., 1997), and a recent Population Viability Analysis indicated that the population is expected to be extirpated within 50 years if there is not active intervention (Cameron and Brooks, 2002). Headstarting of hatchlings is being attempted with this population in an effort to restore it (M. Malhiot, pers. comm. 2004). A more recent estimate suggests that the total population is <50 adults (K. Beriault, pers. comm. August 2007). In 1991 and 1992 respectively, 144 and 157 adults were captured in this area compared to <25 in 2007 using similar effort (Foscarini, 1994, K Beriault, pers. comm.,Sept. 2007). There may be another small and possibly viable population south of Georgian Bay, but neither its size nor viability are known (M. Oldham pers. comm. 2007).

A “population” occurs in low density over a wide area in and near eastern Algonquin Provincial Park (Quinn and Tate 1991; Brooks and Brown, 1992; Brooks et al. 1992; Mitchell et al. 1997; Boyd and Brooks 1998; Smith 2002) and has been studied over approximately the past 14 years with population estimates of 48 (Quinn and Tate, 1991); 108 over three parts of the area (Brooks et al., 1992); and 121 turtles at 5 different sites within this area (Smith, 2002). Sightings of Wood Turtles have been recorded at this location from 1972 to 2005 (B. Steinberg, pers. comm. 2005). The most intensively studied area within this population’s distribution is in Madawaska River Provincial Park, and this population has shown a slow decline over the past 15 years, possibly related to increasing ATV and 4WD truck use (R. Brooks, pers. comm. 2005). Another segment of this population occurs along the Opeongo river in Opeongo Provincial Park and there have been 15 adults captured there (L. Trute, pers. comm. 2005). Overall, there may be over 200 adults in this whole region. In the wild river parks, the numbers are likely declining with increasing human activity.

A “population” occurs in scattered locations across Renfrew County where there have been 19 adults marked over the past few years and a rough estimate of 100 adults is based on capture data from 5 rivers (L. Trute, pers. comm. 2005).

In 2003, a population was discovered on a river near Sudbury (Keable and Kearns, 2004), and to date 34 adults have been marked, but there has been no estimate of population size (J. Litzgus, pers. comm. 2005, 2007).

Further west, there are a number of streams in Algoma district that have populations of Wood Turtles. In Western Algoma District, 6 streams/rivers have been found to harbour Wood Turtles, 22 streams/rivers appear suitable but have no Wood Turtles, and 12 streams/rivers look suitable for Wood Turtles but have not been searched (P. Wesley, pers. comm. 2006). In the 6 rivers in which Wood Turtles have been observed, a total of 158 adults were captured and marked over the past 6 years (Knudsen, 2004; Wesley et al. 2004; P. Wesley, pers. comm. 2006).

In Eastern Algoma district, Wood Turtles have been found in 5 rivers with 114 adults marked over the past several years and an estimated population (assuming similar densities along the whole length of the rivers, of ~600 adults (Cameron et al. 2002; Peiman and Brooks, 2003; Trottier, 2004; J. Trottier, pers. comm. 2005) (see Table 1).

There have been numerous recent surveys and studies of the Wood Turtle in Quebec. Several biologists collaborated to produce estimates from 16 rivers from across the species’ range in the province. (J. Jutras, pers. comm. 2005: see Table 1). Of these rivers, ~10 are in or partly in, the Great Lakes/St. Lawrence FP (Table 1). Of these 10, 6 were thought to contain declining populations and 4 were “unknown” with respect to population trends and one was “stable” (Table 1).

From these data, a total estimate for the Great Lakes/St. Lawrence FP was calculated as ~ 1600 adults (Table1). This value may be slightly inflated because some of the numbers from Ontario were extrapolated from small study sites to entire lengths of rivers (see text above).

Populations of the Canadian ShieldFaunal Province

In Quebec, ~6 rivers are in or partially in the Canadian Shield FP. Of these, two are thought to be stable, two are declining and four have population status “unknown” (Table 1). The total numbers of adult Wood Turtles is estimated to be 1320 (Table 1). As noted earlier, the high estimate from simple extrapolation is likely too high.

Populations of the Appalachian/Atlantic Coast Faunal Province

In Quebec, ~ 7 streams are in, or partially in, this Faunal Province (Table 1). Of these, four are thought to be declining and status is “unknown” in the other three. Estimated number of adults is ~500. In New Brunswick, there have been sightings recorded from several areas (hence the 79 Element Occurrences for the province), but there is no abundance information for these sightings (McAlpine and Gerreits, 1999; E. Tremblay, pers. comm. 2004). A private study by a landowner over the past 8 years has found and marked nearly 100 Wood Turtles (66 adults) to date, with observations of nesting and juveniles (P. Brewer, pers. comm. 2005). A study begun on federal land in the province captured 111 turtles in the first field season, including hatchlings, juveniles and 57 adults (V. Roy, pers. comm. 2005). Overall, turtles have been reported from ~13 watersheds (streams/rivers (McAlpine and Gerrits, 1999).Total numbers for the province are difficult to estimate, but based on numbers from similar regions in Quebec and Ontario may be in the order of 1,000 adults.

In Nova Scotia, there have been few studies on Wood Turtles and the estimates of numbers are generally much less certain than in Ontario and Quebec. Estimates from J. Gilhen cover 12 watersheds which are known to have populations of Wood Turtles. These estimates suggest there is a total of about 2200 adults in these rivers (J. Gilhen, pers. comm. 2005). These estimates are based on Gilhen’s long experience and contacts (see Gilhen, 1990). A set of much higher estimates has been documented by M. Elderkin and M. Pulsifer (pers. comm. 2005). On 8 km of one tributary of River A, 55 person hours of searching yielded 21 adults (0.38 turtles per hour) and a “density” of 2.65 adults per km. Given about 500 km of streams and creeks in this river system, they extrapolated to an estimate of 1300 adult Wood Turtles for this single river (J. Gilhen estimated 240 adults for this stream, pers. comm. 2005). For the River B watershed, M. Pulsifer estimated about 1000-1500 Wood Turtle adults (M. Elderkin, pers. comm. 2005). For this same watershed, Gilhen estimated 800 adults (J. Gilhen, pers. comm. 2005). Both sets of estimates are much higher than was reported in the 1996 COSEWIC report, which stated that there were thought to be 9 “populations” in Nova Scotia, most with fewer than 100 turtles (Litzgus and Brooks, 1996). However, it does appear that Rivers A and B are outliers, and that most or all other populations in Nova Scotia have fewer than 100 adults and many of these populations are thought to be declining and in “big trouble” (T. Herman, pers. comm. Minutes of A&R SSC Meeting Akwesasne, Sept. 21, 2007).

Contents of the table are based on interpretation of information forwarded to T. Kraus and R.J. Brooks in 2005 by Ontario: J. Trottier (OMNR and Co-chair of Wood Turtle Recovery Team in Ontario) and L. Trute (OMNR, former co-chair of the Wood Turtle Recovery Team in Ontario; and Quebec: J. Jutras (MRnFQ) and P. Galois (Coordonateur et rédacteur du plan multi-tortues). Trends for Quebec populations were updated in Nov. 2007 based on new information provided by D. Banville.

Estimates of total numbers for Nova Scotia will have a high degree of uncertainty given the disparity in estimates and the limitations of extrapolation. It seems a range of 2000-7000 would cover most guesses. Therefore, the totals for the Appalachian/Atlantic Coast FP range from ~3000-9000 adults.

Canada

If we combine the above estimates from the four provinces, the values range from a low of ~6,000 to a high of 12,000 adult Wood Turtles in Canada.  However, it should be remembered that some of the estimates are based, especially high values, upon simple linear extrapolations from small study areas.  These extrapolations and their high estimates have a greater uncertainty for reasons given earlier. The turtles are not found along entire lengths of watercourses, but occur in discontinuous patches. Second, given the visibility of this turtle at nesting areas and along creeks and rivers in early spring, it is unlikely that such high numbers could have remained undetected for so long.  Third, the experience from long-term studies (Foscarini, 1994; Cameron and Brooks, 2002; Saumure, 2004; Wesley et al., 2004, Wesley and Brooks, 2005) also indicates a high number of new captures in the first 2-3 years, followed by a rapid decline thereafter and few new turtles over subsequent years.  The initial burst of new turtles is, of course, expected but the rapid decline is unusual when compared to other species, even other turtles (R. Brooks, pers. comm. 2005). It appears that compared to other species a larger proportion of Wood Turtle populations are captured fairly quickly, and this situation is presumably a reflection of the ease of capturing them on land. Therefore, population estimates using capture-mark-recapture from only 1- to 2-year studies are likely to be higher and to have higher uncertainty than estimates based on longer studies.

Fluctuations and trends

The general decline of Wood Turtles indicated by anecdotal evidence has spurred the initiation of studies throughout the range of the Wood Turtle. The majority of these studies have been underway for less than 4 years and so only a few studies can provide more “long-term” quantitative data from Canadian populations. So far no population has been judged to be increasing (see Table 1 and above text).

In Quebec, population trends from demographic studies or knowledge of regional biologists are available from ~16 rivers (Table 1). Of these, ~9 are considered declining and two are thought to be stable in Wood Turtle abundance (Equipe de rétablissement des tortues du Quebec, 2005). More specifically, one well-studied population was believed to be stable (Walde et al., 2003), but recent increases in predators have brought about a rapid decline in the population (Bourgeois et al., 2004). Two estimates made seven years apart show a decline of close to 50% (Daigle and Jutras, 2005). Additionally, this population could decline further due to increased mortality from agricultural machinery (Saumure, 2004). The situation is likely to be similar in other rivers located in the agricultural region of Quebec, which contains about half the provincial population of Wood Turtles (C. Daigle, pers. comm. 2005). Wood Turtle populations in more forested landscapes are likely more secure and declines, if any, are expected to be smaller (Walde et al., 2003; C. Daigle, pers. comm. 2005), as long as road access is not extensive.

In Nova Scotia, there is no published information on population trends, but there is anecdotal evidence that many populations are declining (Litzgus and Brooks, 1996; T. Herman, pers. comm. Sept. 2007).

In New Brunswick, there has been no formal long-term monitoring. However, anecdotal information indicates a population decline and lack of protection of at least one site in the southern part of the province, where illegal ATV activity has occurred in nesting areas (P. Brewer, pers. comm. 2005). This informal study, begun in 1998, has reported that the population has declined in the past few years, and suggested that the decline is from increasing ATV traffic and its effects. i.e., turtles being run over, nests destroyed and turtles picked up or deliberately killed (P. Brewer, pers. comm. 2005). Five of 6 nesting areas are driven on daily by several ATVs and as water levels decline over summer, ATV crossings increase to about 15 crossing points per km of river.

In Ontario, the lone population in southern Ontario was modelled after it underwent a sharp decline in 1994-5, presumably because of collection for the pet trade (Table 1). The model predicted that the population would be extirpated in 50 years without significant intervention (Cameron and Brooks, 2002). Currently, this population is being bolstered by headstarted juveniles, and it remains to be seen if this strategy will help restore its numbers (M. Malhiot, pers. comm. 2004); however, it appears that the population is continuing to decrease (K. Beriault, pers. comm., Sept. 2007). A second population near Algonquin Provincial Park has been monitored since 1987 (Quinn and Tate, 1991; Brooks et al., 1992; Brooks and Boyd, 1998; Smith, 2002), and it appears also to be declining (R. Brooks, pers. comm. 2005) (Table 1). This population is in a Provincial Park, but as with the population in New Brunswick (see above paragraph), the area is open to ATVs and increasing recreational use (R. Brooks, pers. comm. 2005). As reported in a 20-year study in Connecticut, even opening an area to hikers and picnicking can lead to the complete extirpation of a Wood Turtle population through collection (Garber and Burger, 1995). Other populations are reported as “healthy”, although road mortality and high levels of foot and vehicular traffic are seen on nesting sites and along some roads near the sites (J. Trottier, pers. comm. 2004; R. Knudsen, pers. comm. 2004; J. Litzgus, pers. comm., 2007). None of these populations has been investigated long enough that trends in abundance could be detected or inferred at this point. No other Ontario populations have been studied long enough to detect any potential trends, but all are subject to increasing contact with people and their vehicles. Expanding human impacts have placed virtually all Ontario Wood Turtle populations under threat.

The overall trend in both Canada and the USA (J. Harding, pers. comm. 2004) is one of decline, and although some populations are described as stable, the exceptionally high vulnerability of Wood Turtle populations to anthropogenic sources of mortality means that any population to which humans have access (all current known populations, to some degree) are susceptible to decline.

Rescue effect

In the United States, most Wood Turtle populations are declining (NatureServe, 2004), and there is no exchange of individuals known to occur between these populations and those in Canada. It is highly unlikely that any Canadian population would be bolstered or renewed by another population (from Canada or from the United States) were it to become extirpated (see “Distribution” and “Dispersal/Migration”).