Description
Medium to large (125-260 mm) turtle (Ernst and Barbour 1972) with a yellow chin and throat and a long neck. The elongated, domed, smooth, blue-black carapace is neither keeled nor serrated. Each pleural and vertebral scute has tan to yellowish spots or slightly radiating lines; the marginals are heavily spotted. The bright yellow plastron has large, dark, symmetrically arranged blotches on the posterior lateral third of each scute; the blotches may hide the yellow color on older adults. A well-developed hinge lies between the pectoral and abdominal scutes. The tail and limbs are blue-gray, black, or brown with some yellow or light brown spots. The head is large and flat, black or dark brown, possibly with scattered yellow spots. The eyes protrude. The upper jaw is notched terminally and may be marked with dark bars. Hind feet are weakly webbed.
Sexual dimorphism is not pronounced. In females the tail is more slender and the cloacal opening is located anterior to the rear carapacial margin. The males have plastrons more noticeably concave (Vogt 1981, Gibbons 1968). Kofron and Schreiber (1985) in Missouri found that concavity of the plastron does not appear until the turtle has a plastron length of 174-180 mm. They considered all individuals over 180 mm plastron length with a flat plastron to be females. Detailed observations of 74 turtles by Graham and Doyle (1979) revealed that only males showed plastron erosion and pitting.
The following measurements and ratios were gathered by Graham and Doyle (1979) from their study population in Massachusetts and may be useful to further separate the sexes. The ratio of plastron length to carapace length is significantly greater in females (ratio = 1.008, n = 33) than in males (ratio = 0.956, n = 41). The mean carapace length is greater in males (mean = 15.5 mm, range 182.4233.7) than in females (mean = 204.2 mm, range 178.9-217.5), but plastron lengths are not significantly different. Maximum shell height is significantly greater for females (mean = 84.5 mm, range 70.3-94.3) than for males (mean = 80.6 mm, range 69.8-90.6) as is the ratio of shell height to carapace length (females ratio = 0.415, males ratio = 0.375). Carapace width is significantly greater in males (mean = 146.5 mm, range = 125.6-161.3) than in females (mean = 139.0 mm, range = 121.7-148.6) and the ratio of carapace width to plastron width is also greater for males (ratio = 1.307) than for females (ratio = 1.240).
The carapace of a hatchling is rounded and keeled, 28-35 mm in length, and dark brown to black or gray, usually with some light spots or dashes on the carapace though some may be patternless. Ewert (1979) reported hatchlings to be 36.5 mm carapace length and 32.1 mm plastron length (1979). The top of the head, neck, and soft parts are black with light markings on the chin and rather prominent light lines on the tail. The plastron has a large, black central blotch or is predominantly black with light yellow on the edges. The tail is proportionately much longer than in the adult, being 65 to 70 % of the carapace length or even as long as the carapace. The tail and the deeply notched jaw like the adult distinguish them from the hatchlings of other species. The hinge may not always be apparent on young ones (Conant 1951). The yellow undersurface of the neck appears at 3 years of age (Vogt 1981).
Diagnostic Characteristics
Box turtles, Terrapene carolina and T. ornata, have a plastral hinge, but neither has a yellow throat and chin or a notched upper jaw; T. carolina has a keeled carapace. Clemmys guttata has a blue-black carapace with yellow spots, but it lacks the plastral hinge.
Habitat
Marshes, ponds, swamps, lake shallows, backwater sloughs, shallow slow-moving rivers, protected coves and inlets of large lakes, oxbows, and pools adjacent to rivers; waters with soft bottom and aquatic vegetation (Kofron and Schreiber 1985, Nyboer 1992, Ernst et al. 1994). Sometimes leaves water and walks overland.
Although E. blandingii has been said to be semi-aquatic, this idea evidently is based primarily on observations of nesting females and has been questioned by several researchers (Pope 1939, Lagler 1943, Gibbons 1968, Kofron and Schreiber 1985) who regard this species as primarily aquatic. Gibbons (1968) found turtles on land between aquatic areas only in April and September (in addition to females in June). Conant (1951) considered it to be unusual for turtles in Ohio to be more than 100 yards from the water. However, Rowe and Moll (1991) found that terrestrial excursions were a significant part of activity in Illinois.
Hibernation most often occurs within organic substrate of ponds and creeks. Two turtles overwintered within 10 m of each other for 31 days in Wisconsin (Ross and Anderson 1990). Somewhat communal hibernation may be due to scarcity of suitable hibernacula. Five of the six overwintering turtles in the Wisconsin study used one of their summer activity centers for overwintering. Most moved from marshes, shallow ponds, and ditches to deeper ponds after September 1. The deeper ponds probably provide stable water levels during the critical overwintering period and a longer period of warm water temperatures in early fall. Individuals in Missouri hibernated in shallow marsh areas under 15 cm mud below 9.5-21 cm of water. One turtle in Illinois was observed wintering under a brush pile that she had used as a refugium during August (Rowe and Moll 1991) and, in Ohio, turtles have been reported overwintering beneath leaves several feet from the water (Conant 1951).
Eggs are laid usually in sandy soil in upland areas, usually in warm sunny sites. In Michigan, nests were 2-1115 m (average 135 m) from the nearest water; most females returned to same general nesting area each year; nesting area usually was not immediately adjacent to female's marsh habitat (Congdon et al. 1983). In Wisconsin, nested usually in grassland, average of 168 m from water and average of 620 m from non-nesting activity center (Ross and Anderson 1990). Wisconsin turtles nested in large (>6 ha) contiguous grassland habitat (Ross and Anderson 1990). About 50% of the cover at the Wisconsin nest sites was grasses and Pennsylvania sedge (Carex pensylvanica) (Ross and Anderson 1990). Illinois females wandered overland for 5-17 days and up to 1670 m away before nesting 650-900 m from their home ponds (Rowe and Moll 1991). Turtles in Nova Scotia were nesting 5 miles across a lake from their probable activity centers (Bleakney 1963). In Maine, nests were 70-410 m (mean 242 m) from the nearest water (Joyal et al. 2000). In Nova Scotia, nests were predominantly on lakeshore cobble beaches (Standing et al. 1999).
Females in Ontario chose areas with little or no vegetation (Petokas 1986). However, nests were found in a clustered distribution, likely because of herbaceous cover along the perimeter of the chosen site where turtles could hide and survey the area before advancing into the open to seek a nest site (Petokas 1986).
Like other freshwater and terrestrial turtles, Blanding's turtle often chooses disturbed sites. Petokas (1986) suggested that the turtles probably nested in available clearings, on sand and gravel bars, and on muskrat lodges or beaver lodges and dams prior to the modification of the landscape by humans. However, all of the females in his study chose disturbed sreas such as tilled plots, cemeteries, a powerline right-of-way, and a road. He found no nests on the available beaver dams.
In Massachusetts, turtles were observed crossing an area of open sandy soil to nest almost exclusively in the middle and far reaches of the light sandy soil of a cornfield frequently disturbed by tilling or earth-moving; such areas afforded the highest elevation (about 10 m above the marsh) and most open land near the marsh (Linck et al. 1989). Similarly, one of the earliest natural history reports of the nesting Blanding's turtle stated that the female nested on a sand hill 10 feet above the level of open water (Brown 1927). Ross and Anderson (1990) found nests an average of 18.4 m from shrubs
There is some evidence that Blanding's turtles are faithful to nesting sites. Petokas (1986) found that some females who nested more than once during his five-year study returned to the same nest. Congdon et al. (1983) also observed nest fidelity. Eight of 11 females in Michigan returned to a specific nesting area. and 36.3 m from trees.
Further information on habitat in specific areas follows.
Prairie marsh or wet prairie, especially associated with sandy soils, is the preferred habitat in the western part of the range, (Kofron and Schreiber 1985, Nyboer 1992). In the Nebraska sandhills, Blanding's turtles, particularly juveniles, were significantly more numerous in marshes and small ponds than in lakes and open waters (Bury and Germano 2003).
In Wisconsin, ponds were used more often than the marshes that were available (Ross and Anderson 1990); the authors surmised that the use of ponds and ditches might be due to their use as travel routes between feeding or activity centers; use of ponds with sand substrate and no aquatic vegetation was minimal. Wetlands in which the cattails had been cleared in some areas were used by the turtles but not those with dense cattail mats, indicating that availability of open water affects wetland use. Marsh habitat use was highest in early summer. Higher water quality encourages increases in invertebrate prey populations, and those habitats in Wisconsin with higher dissolved oxygen (>5.0 ppm) had greater use.
In the Massachusetts Great Meadows National Refuge near Concord, Massachusetts, Graham and Doyle (1977) found more turtles in a pool with greater amounts of decaying vegetation and algal growth than in an otherwise comparable pool. Dominant submerged aquatics there were coontail (Ceratophyllum demersum), waterweed (Anacharis occidentalis) and pondweed (Potamogeton sp.) (1977). Kofron and Schreiber (1985) also found that this turtle seemed to be dependent on waters with aquatic vegetation. In Illinois, turtles moved from less vegetated ponds that they inhabited in May and June to more highly vegetated ponds and adjacent open marshes in July (Rowe and Moll 1991) perhaps due to increased competition during times of high feeding rates.
In Minnesota, the preferred habitat is calm, shallow water with rich aquatic vegetation. The turtles are found in marsh areas in large river floodplains in the state where there also occurs sandy upland areas for nesting (Coffin and Pfannmuller 1988). See J. Herpetol. 26:233-234 for information on habitat selection by juveniles in Minnesota. In Michigan, the turtles use shallow weedy bodies of water such as permanent ponds or open marshes (Harding 1992). In Ohio, the turtles have been reported uncommon in deeper or more exposed parts of lakes but frequently found in protected coves (Carr 1952). A population in Nova Scotia is located within the central marshy area of a peninsula projecting into a lake. At this site the turtles at least formerly were abundant (Bleakney 1963). In Nova Scotia, distribution parallels that of dark acidic waters and peaty soils; in summer, often near the outflow of streams into lakes (Power et al. 1994). Sphagnum appears to be the primary indicator of juvenile and subadult habitat (McMaster and Herman 2000). Hatchlings do not seek open water upon emergence from the nest (McNeil et al. 2000).
Turtles on the northern edge of Lake Erie in Ontario also were reported as very common in a similar peninsular habitat where they were rarely seen in the open water of Lake Erie but were common on the marshy banks of the ridges that line a 20-mile-long sandspit (Adams and Clarke 1958). On Grenadier Island of the Thousand Islands region of the upper St. Lawrence River in Ontario, Canada, the turtles inhabit a 70-ha, continually flooded sedge meadow in the center of the island (Petokas 1986). Although Cook (1984) stated the turtles are often found in boggy areas in Canada, they do not inhabit bogs in Wisconsin (Vogt 1981). Perhaps Cook was referring to marshy areas as "boggy" nor are there other specific references to a bog habitat.
Ecology
Age class structure of Emydoidea populations often appears to be highly skewed toward adults. It has been suggested that trapping techniques and locations may be missing the juveniles that do not share the same habitat as the adults. However, in Nova Scotia, juveniles, subadults, and adults shared the same macrohabitat (McNeil et al. 2000). Possibly juveniles are more secretive (Congdon et al. 1993) or truly scarce due to nest failure (Congdon et al. 1983). In Nebraska, Germano et al. (2000) trapped many juveniles but few small individuals less than 10 cm CL.
Sex ratios are usually considered 1:1. Gibbons (1968) reported an "apparent" abundance of females though admitted that all turtles over 90 mm not identified as male by plastron shape were considered females. Most authors now consider these turtles immature under 180-190 mm.
Populations incurred extensive nest predation by carnivores in Michigan and Wisconsin (Congdon et al. 1983, Ross and Anderson 1990). Raccoons were important egg predators in Nova Scotia (see Herman et al. 1995). Standing et al. (2000) documented predation on neonates by ants, short-tailed shrews, and possibly raccoons in Nova Scotia. In Michigan, characterized by low recruitment due to high level of nest loss from predation and low nesting frequency; egg survivorship to hatching was 0.18 (Congdon et al. 1983); subsequent study revealed variable annual nest survival rate (0-63%), with annual adult survivorship exceeding 93%; demographic analysis indicated that population stability was most sensitive to changes in adult or juvenile survival and less sensitive to changes in age of sexual maturity, nest survival, or fecundity (Congdon et al. 1993, 2000). In Nova Scotia, Herman et al. (1995) and Standing et al. (2000) also found low egg and hatchling survival and noted the importance of high survivorship of adults to population stability. Annual egg failure in Nova Scotia ranged from 26.5-94 percent; in the absence of predation (nests were protected in exclosures), low incubation temperatures and nest flooding appeared to be the major causes of egg failure (Standing et al. 1999). Flooding may be a significant cause of nest failure in some areas in some years (Herman et al. 1995).
Ross and Anderson (1990) defined activity centers as the area within a habitat used by turtles for at least five days. In their Wisconsin study, activity centers were well defined and separated by long distances. Size of activity centers for both males and females were about 0.7 ha (range 0.27-0.94 ha). Activity centers of females overlapped with those of other females (average overlap: 26%) and juveniles (7.4%) as well as males (12%). Male activity centers did not overlap with those of other males. In northeastern Illinois, activity centers, defined as the distribution of a plot of radiolocations throughout the season, were not significantly different between males and females. For all turtles, activity centers ranged from 0.1 to 1.2 ha. (mean = 0.6) (Rowe and Moll 1991).
Distance between an individual's activity centers did not significantly differ between males and females in Wisconsin (Ross and Anderson 1990). Length of such movement for two males was 260 m and 635 m, respectively. For six females the mean distance moved between activity centers was 489 m. Distance between activity centers was no higher than 100 m in Illinois, and sometimes activity centers overlapped (Rowe and Moll 1987). Home range/movement was less than 100 m for individuals in Michigan (Gibbons 1968). Home range size was 1.7-22 ha in Illinois (Rowe and Moll 1987). In Minnesota, home range size averaged 7.8 ha in adults, 5.9 ha in juveniles (Piepgras and Lang 2000). In Nova Scotia, total range and movements increased with age and correlated positively with the amount of suitable habitat in an area (McMaster and Herman 2000).
Daily movements however, differed significantly between females (mean: 95.1 m) and males (mean: 48.4 m) in Wisconsin, probably due to reproductive activities of the females (Ross and Anderson 1990). In Illinois, distance of daily movements by males (mean: 48.9 m) was in agreement with that found in Wisconsin, but females did not move as far (mean: 32.4 m) because Rowe and Moll did not include movements associated with reproductive activity. Females in Illinois moved farther in May than in August, probably due to greater resource needs during reproductive periods.
In Nova Scotia, sometimes moved 5-11.5 km overland to establish residency in a different drainage (Power et al. 1994, Herman et al. 1995
Other than movement by females to locate nesting sites, Blanding's turtles exhibit three other types of terrestrial movement, as noted by Rowe and Moll (1991). During reproductively active periods, males may move long distances overland to locate mates. Secondly, short overland excursions to other water bodies are common and probably indicate explorations for improved ecological conditions or in response to social interactions. Thirdly, turtles have been observed to "estivate" on land for several hours to several days in both Illinois and Wisconsin. Ross and Anderson (1990) recorded this behavior for Wisconsin turtles in July and August during periods of cool water temperatures. Rowe and Moll (1991) observed turtles with the anterior portion of the carapace nestled in leaf litter early in the season. They suggested the Illinois turtles were also avoiding cold water temperatures and practicing thermoregulation, but the behavior might better be referred to as basking, since the turtles were quite alert.
Reported density ranges from about 6/ha (Graham and Doyle 1977) to 55/ha (Kofron and Schreiber 1985).
Reproduction
Copulation occurs primarily March-May. Nesting occurs mid-June to mid-July in Nebraska, during 2-3 week period in June in Massachusetts, late May and June in northeastern Illinois, mid-May to early July in Michigan (Congdon et al. 2000), late May-June in Minnesota (Pappas et al. 2000), mid-June to early July in Wisconsin, mid-June in Maine (Joyal et al. 2000), mid-June to early July in Nova Scotia (Standing et al. 1999). Lays eggs in early evening (in Massachusetts, typically begins nesting when light, continues into darkness).
Nesting may vary annually by as much as 2 weeks in the same area. Nesting activity in Ontario does not commence until daily maximum temperatures reach at least 19.5 C and daily average temperature is 21.4 C (Petokas 1986). The nesting season lasts 2-3 weeks. Nesting generally occurs in the evening, beginning when it is still light but rarely completed until after dark; average time from first digging to leaving the nest by the female is 2.5 hours (Congdon et al. 1983, Linck et al. 1988). Turtles in southeastern Ontario averaged slightly less than 2 hours to complete nesting though began about the same time as those in Michigan (Petokas 1986).
Clutch size 3-22; average 15 in Nebraska, 10 in Minnesota (Pappas et al. 2000), 8-9 in Maine (Joyal et al. 2000), 8 in Ontario, 10-11 in Nova Scotia (Standing et al. 1999, 2000). DePari et al. (1987) found clutch sizes ranging from 9 to 16 eggs (mean = 12.9) for Massachusetts females whose size ranged from 200 to 220 mm. Mature females in southwestern Michigan have been measured at 160-162 mm (mean = 184.5). These turtles produced an average clutch size of 10 (Congdon, et al. 1983). Individual females lay one clutch per season in Michigan and Nova Scotia. Females may retain a single oviductal egg that is released post-nesting (Petokas 1986). Adult females may not nest in some years (68 percent nested less than annually in Nova Scotia, Standing et al. 1999).
Gutzke and Packard (1987) studied the clutches of 6 gravid females removed to the laboratory from the Valentine National Wildlife Refuge in Nebraska. The following information on the response of eggs to temperature is illuminating. At 31.0 C, 77.3% of the eggs hatched, required approximately 49 days to develop, and 100% of the hatchlings were females. At 26.5 C 95.2% of the eggs hatched, required 13-14 additional days of development, and all of the hatchlings were males.In contrast to the eggs of painted turtles of which a significant number will hatch at temperatures as low as 22 C, none of the eggs in this study hatched at 22 C. These incubation periods are in line with those reported by Ewert (1979): 30-32 C, 47.4 days; 29.5-30 C, 49.3 days; 27.4 C, 52.4 days; 25-25.5 C, 71.3 days; 24 C, 81.6 days. These relatively short incubation periods are a selective advantage for a species nesting on ephemeral or unstable substrates such as sandbars and beaches.
Hatchlings emerge in 10-17 weeks; mid-August to early October in Michigan, mid-August to late September in Minnesota (Pappas et al. 2000), late August to early October in Maine (Joyal et al. 2000). A mid- to late October emergence was observed in Minnesota (Moriarty and Linck 1995, Herpetological Review 26:99). Average time from egg laying to emergence in Nova Scotia was timed at 88 days (Bleakney 1963).
Hatchlings may emerge from the nest synchronously or asynchronously over several days. Nine of 14 nests in Michigan hatched over several days and three of these were preyed upon after emergence of the first hatchling (Congdon et al. 1983). Average time from egg laying to hatching was 84 days. There is some evidence that hatchlings may overwinter on rare occasions with emergence in the spring (Congdon et al. 1983). In Massachusetts, hatchling moved from nests to wetlands in from less than 12 hours to 9 days (Butler and Graham 1995).
Hatchlings emerge from the egg at less than 37 mm carapace length (Ewers 1979) and grow about 12 mm per year (Moriarty 1988). This rate is slowed significantly at maturity but there are few studies of growth rate in this species. Graham and Doyle (1977) found that growth rate dropped at age class 13-15 in Massachusetts turtles. This corresponded with the size of adults at sexual maturity (1977). Diet quality affects growth rate so that size at a given age is highly variable.
The age of sexual maturity of females is 14-20 years, though maturity seems to be more tied to size than to age; very long-lived, not uncommonly lives several decades (Congdon et al. 1983, Moriarty 1988, Congdon and van Loben Sels 1991; Herp. Rev. 20:53; Herman et al. 1995; Pappas et al. 2000). Kofron and Schreiber (1985) in Missouri found that males were sexually mature at 174-208 mm. Graham and Doyle (1979) believed that individuals with a carapace length less than 190 mm were immature. Male maturity may not be achieved until the 12th growing season when plastron lengths range from 181 to 190 mm in Massachusetts populations (Graham and Doyle 1977).
Recaptured turtles in Michigan have been aged at 42 years (Congdon et al. 1983) and in Massachusetts at 30-40 years (Graham and Doyle 1977). Gibbons reported in 1987 that the oldest known Blanding's turtle was 48 years old. However, Brecke and Moriarty (1989) reported a turtle in Chisago County, Minnesota, to be a minimum of 77 years old. The turtle was marked in 1926 presumably as an adult of at least 14 years. In spite of advanced age, the turtle broke no size records but was still large at 217 mm. The longevity of Blanding's turtles is a life history characteristic of the K-strategist. Combined with delayed maturity, single clutches, and a short annual reproductive period, this species is banking on many productive years by adults in a population. According to Congdon et al. (1983) 23-48% of the females in a population reproduce in a given year, and adults, barring death on the highway, can look forward to at least 15 years of reproductive activity. In this way, populations can be maintained through sufficient reproduction effort and an occasional good year in spite of long periods of low recruitment due to nest failure, predation, or hatchling mortality (Petokas 1986).
Emydoidea exhibits temperature-dependent sexual differentiation (TSD) that may actually favor males if nesting habitats become cooler due to shading.