Population size.—The calculated population estimate and corresponding 95% Confidence Interval for sexually mature individuals in the population during 1997 was 231 (152 - 309). This corresponds to an estimated sexually mature female population of 114 (75 - 153) turtles. Fiftytwo individual females (46% of estimated female population) were observed in the nesting area. An additional 11 gravid female turtles were observed on the staging ground adjacent to the nesting area but not observed to nest. Thus, 63 female G. insculpta were observed in the vicinity of the nesting area, which represents 55% of the estimated female population. Movement patterns.—Seven of the 15 turtles followed by radio-telemetry were observed to nest in the nesting area and two others were suspected of nesting therein or close by. These latter turtles were included in analyses of movement to the nesting area because movement patterns put them in the general area of the gravel pit. The remaining six turtles were not observed nesting but some (n = 4) were gravid and telemetry data suggest that these turtles nested in other areas. Data on a single turtle are not included because of transmitter failure.
Movement to the nesting area were along waterways for all turtles observed (n = 9); therefore, distances reported are not straight-line distances. Two general movement patterns to the nesting area were observed. The first movement pattern was a relatively slow migration over more than 14 days (n = 3), which began shortly after emergence from hibernation. The second pattern was a quick migration (n = 6), whereby the turtles stayed relatively close to where they hibernated and on approximately 10 June (three days before nesting season began) moved directly to the nesting area over two to seven days. There was no significant difference between the slow and fast groups for mean distance traveled to the nesting area, 2240 ± 1041 m and 2113 ± 1360 m, respectively. The mean distance traveled by all turtles to the nesting site was 2156 ± 1196 m (n = 9, range = 840 - 3740 m). Two turtles had exceptionally long movements during one day: one was observed to move 2940 m and the other 2400 m, both of these against the current.
Staging and timing of nesting.—We observed 29 and 33 females digging nests in 1996 and 1997, respectively. The nesting periods lasted from 9 - 21 June, 1996 (13 days) and 13 - 28 June, 1997 (16 days) (Fig. 1). In 1996, a straggler nested on 1 July but was deemed an outlier because of abnormal behavior and movements for weeks prior to nesting. During 1996, one turtle was observed visiting the nesting area on 7 June and nine on 9 June, which was the day the first nest was completed. Seventysix percent (n = 93) of all visits to the nesting area occurred from 10 - 14 June 1996. During the same five day period, 66% of the known nests were constructed (Fig. 1A). On 13 June, a heavy midday rain stimulated nesting activity and 34 turtles were identified in the nesting area with 38% (n = 11) of all nests being constructed. The remaining 9 nests were constructed over the following week and fewer turtles visited the area during this period. The last nest, as well as visit, was on 21 June, the end of the nesting period for 1996. In 1997, the first turtle arrived at the nesting area on 10 June, and the following day another was seen. From 12 - 15 June visits were more numerous. The first nest was constructed on 13 June. From 16 to 21 June, there were 130 visits, which accounts for 58% of all the visits to the gravel pit during the nesting period. From 16 - 21 June, 24 out of the total 33 nests were constructed (73%) (Fig. 1B). The remaining eight nests were completed during the following week, from 22 to 28 June.
Many turtles were captured for the first time at the nesting area in 1996; therefore, the number of visits and activities may not be accurate for that year because some of the turtles were not identified or captured on their first visit to avoid disturbing other turtles. In 1997, 224 visits to the nesting area represented more than 375 hr of activity. From this, the mean time spent for all visits, for both nesters and non-nesters was 103 min. The mean number of days nesting turtles were observed at the nesting area from first sighting until nesting was completed was 3.3 days, with a range of 1 - 9 days. For turtles that were observed to nest in the 1997 season, the average time spent at the nesting area that did not result in a nest was 121 min (n = 71). During visits that resulted in nest completion, the mean time spent at the gravel pit was 232 min (n = 24). The mean nesting time (digging, laying, and covering the nest) of these turtles was 131 min and ranged between 86 and 202 min. On average, a female spent 101 min to choose the nest site location on the day she actually nested. Upon covering a nest, the turtle left the nesting site immediately, typically retreating to the river.
Excavations of nests were started at almost all hours during daylight, albeit with a predominantly bimodal distribution of nesting activity (Fig. 2). In the morning, 38.5% (n = 22) of nests were started between 05:00 and 09:00. Forty-four percent of nests (n = 25) were started in the evening between 16:00 and 21:00 with most, 37% (n = 21) from 18:00 to 21:00. The remaining 17.5% (n = 10) of nests were started at various times during midday. During the time that the nine radio-tracked turtles were in the vicinity of the nesting area, they were regularly observed in the adjacent staging ground. This area consisted of a 200 m length of river and flood plain plus a bog-like habitat centered on the nesting area. This staging area was utilized by 80% of the nesting females during both years of the study. During the study, 63 females (55% of population) were documented within the staging area and/or at the nest site. While staging, turtles tended to use natural habitats that were unaltered by human disturbances. Approximately 20 m of the waterfront within the staging area had been cleared of alder (Alnus rugosa) thickets (beside a cottage) and was devoid of turtles, while 28 females were observed in the adjacent alder thicket areas where the vegetation was undisturbed. Nest-site fidelity and reproductive output.—Of the 44 turtles that visited the nesting area in 1997, 30 had visited in 1996. Thus, a minimum of 68% were likely gravid in two consecutive years. Of the 33 that nested in 1997, 64% (n = 21) were known to have nested in the nesting area in 1996. One turtle that nested at this area in 1996 nested 200 meters upstream in 1997. Therefore, of the turtles that nested in two consecutive years, 95% (n = 21) returned to the same nest site. Of the turtles that returned to nest, 43% (n = 9) nested within 10 m of their nest from the previous year.
The mean carapace length (CL) and plastron length (PL) of nesting turtles in 1996 and 1997 were not significantly different from female turtles not observed to nest in the population (Table 1). Nesting turtles, however, had significantly more annuli than non-nesting turtles (Table 1). Turtles that nested in both years were significantly larger for mean CL and PL than those that nested in only one year, but the mean number of annuli was similar (Table 2). There was a significant positive correlation between number of eggs and female CL, PL, and mass (Table 3).
Mean clutch size was significantly different between 1996 (9.4 ± 2.4 eggs, range = 5 - 13, n = 27) and 1997 (10.6 ± 2.1, range = 5 - 15, n = 30; t = -2.045). In 1997, one nest containing 20 eggs was excluded from calculations as it was deemed an outlier. Nest success was 74% in 1996, and 65% for the 1997 season. Nests constructed during the first half of the nesting period had significantly greater nest success than those constructed during the latter half (Gadj = 7.984) with 31 of 40 nests (77.5%) constructed during the first half of the 1996 and 1997 nesting season being successful; whereas only 9 of 22 (40.9%) constructed during the second half were successful. The number of hatchling recruits in 1996 was 148 (total eggs = 253) and in 1997 was 175 (total eggs = 337). Thus, recruitment success for 1996 and 1997 were 59 and 52%, respectively. Twelve infertile eggs were observed; eight in 1996 and four in 1997.
Incubation period.—Hatching dates were 17 August - 7 October in 1996 and 19 August - 5 October in 1997 (Fig. 3). The average incubation period was 86.8 ± 12.4 days (range = 65 - 116, n = 19) in 1996 and 77.5 ± 9.4 days (range = 60 - 99, n = 20) in 1997. The incubation periods were significantly different between years (t = 2.37, df =37, P = 0.023).
Mortality.—Of the six nests left to overwinter in 1996, two had a total of 10 dead hatchlings in them in the spring. In 1997, 11 nests were excavated in mid-November. Only one nest had a living hatchling and it had a large yolk sac, this turtle was reburied in the nest and was found dead two weeks later. All other failed nests in both years had eggs that had not completed development, rotting eggs, and/or eggs associated with insects. No mammalian predation of nests at the nesting area occurred during the study. Raccoons (Procyon lotor), red fox (Vulpes fulva), skunks (Mephitis mephitis), as well as feral and domestic cats (Felis domestica), and dogs (Canis familiaris) were observed within the nesting area.