Ecology and natural history of the wood turtle (Clemmys insculpta) in southern New Hampshire

Tuttle, Sheila E.
Meade Cadot
Department of Environmental Studies
An intensive one-year ecological study of the wood turtle (Clemmys insculpta) was conducted in southern New Hampshire from April 1993 to June 1994. A study population consisting of 82 turtles (29 adult females, 17 adult males, 36 juveniles) was captured, measured, and marked for individual identification. The male to female adult sex ratio was 1.0:1.8, and did not differ statistically from a 1:1 ratio. There was a highly significant intersexual difference in body size, with males having a significantly larger mean carapace length than females (males, 178 ± 17.4 mm; females, 171 ± 6.1 mm). In body mass, there was no statistically significant difference between males (= 750.3 ± 134.0 g) and females (= 711.4 ± 85.9 g). Ten adults (5 of each sex) and two juveniles were radio-tracked and their locations were mapped. Home range sizes were calculated using three methods; concave polygon, minimum convex polygon, and summed 20-meter quadrats of locations. Using the last method because of its comparability to a previous study, home ranges of wood turtles in New Hampshire were: 5.8 ± 3.3 ha (adult males), 3.9 ± 3.0 ha (adult females), and 6.0 ± 6.8 ha (juveniles). There was no significant difference in home range size between adult males and females, or between adults nad juveniles. There was no correlation between turtle body size or its estimated age and its home range size. Home ranges among and between sexes and age classes overlapped spatially and temporally throughout the study. During the active season, 33% of all captures of radio-tagged turtles (n = 757) were in a brook. Females were more terrestrial than were males throughout the active season, seldom being observed in a brook during the summer. Males were more frequently encountered in water. Fifty-seven percent (n = 429) of the total captures were in wetlands or in brooks, including turtles in brooks which ran through wetlands (deep shrub swamp, emergent wetland, open water wetland), and turtles in brooks which flowed through uplands. Additional observations of turtles in wetlands but not actually in water (n = 204) included 36% of all terrestrial observations (n = 574), in shrub swamp, wet meadow, or forested wetland habitat (or 27% of all radio-tagged turtle locations). Habitat was evaluated by assessing 28 habitat units which ere based on the areas most frequently used by the 12 radio-tagged turtles. In these habitat units, dominant (most frequent) cover types and dominant species within cover types were recorded. Fifteen (54%) of these habitat units were wetland in character, and 13 (46%) were upland. Cover types identified were: herbaceous, shrub, tree, sapling, and woody vine. Among the five cover types, shrub cover (42%) was the most frequently occurring cover type within a habitat unit. A total of 368 measurements of cloacal and environmental temperatures were recorded between April and December 1993. Activity categories identified were: basking, dormant in water, swimming, courtship, foraging, walking, and dormant in terrestrial forms. Cloacal temperatures of wood turtles during these activities were positively correlated with environmental temperatures. Cloacal temperature ranged from 3.8 C for a turtle found dormant in water to 32.0 C for one found dormant on land, and the mean cloacal temperature for all activities and months was 23.4 ± 5.7 C. Wood turtles showed peak activity periods during May and June, although this may partially be ascribed to an increased number of females observed nesting in June. Observations of nesting behavior were made from 5 to 13 June, after the first nest was discovered on 2 June. Nests (n = 9) were excavated by females in sandpits located a mean distance of 60.3 ±18.4 m from a brook. All nests were protected by close-fitting hardware cloth enclosures. The natural incubation period ranged from 66-82 days, with a mean of 76.0 ± 6.2 days (n = 8). (The date of egg deposition was unknown for one nest.) Mean clutch size for the nine nests documented was 7.8 ± 1.0 (6.0-9.0) eggs per clutch. Seventy eggs measured resulted in a mean egg width of 22.5 ± 1.0 mm, and a mean egg length of 35.3 ±2.5 mm. Overall hatching success (70 eggs) was 77%. Four additional nests not experimentally covered were all destroyed by mammalian predators, suggesting that many natural (uncovered) nests were predated. Post-emergent behavior of hatchlings during migration from nest to water was observed in August and September. Hatchling wood turtles were powdered with fluorescent pigments and tracked at night with long-wave ultraviolet light. Of 53 turtles marked, 12 individuals were successfully tracked to a nearby brook. Hatchlings were monitored daily form 0600 h until emergence from their terrestrial forms, or until 1300 h. Habitat selection was most frequently in herbaceous vegetative cover or in hayfield banks bordering sandpit nesting areas, and in dense woody and herbaceous ecotones along dirt roads. Mean distance traveled per movement was 26.2 ± 25.0 (0.2 - 109.0) m (n+134) the mean total distance traveled for the 12` hatchlings tracked to a brook was 131.7 ± 119.7 (27.8 ; 445.4) m; mean travel rate was 23.4 ± 9.5 (14.5 ; 40.0) m/day. The mean time taken to reach the brooks was 6.2 (1 ; 24) days. Post-hatchling observations of young wood turtles suggested conspecific scent-trailing. Based on field observations, olfaction, vision, positive geotaxis, and auditory cues of rushing water in brooks may be employed as orientation mechanisms in hatchling C. insculpta. The estimated population density of male, female, and juvenile wood turtles was 2.6 turtles per ha. Population size and density may be somewhat conservative, as other studies of C. insculpta report greater population densities. Home range sizes for adult male and female turtles were most similar to those reported in central Pennsylvania, using the same 20-m grid count method. Habitat in New Hampshire was diverse, as reported in other areas within the wood turtle's range. The New Hampshire site appeared to be similar to that in central Pennsylvania, southern Connecticut, northeastern New York, and Cape Breton Island, Nova Scotia. Wood turtles chose dense habitat most frequently, which was both wetland and upland in character. This field study provided basic life history data for understanding the wood turtle's ecological variation throughout its range.

Read Full Text