Study site.—The present study began in May of 1996 and continued until December of 1997 along a river in the Municipalité Régionale de Comté Le centre de la Mauricie, Québec, Canada. The exact location of the population is not revealed due to the potential threat of collection for the pet trade (e.g., Garber and Burger 1995; Niederberger and Seidel 1999). Although detailed descriptions of the research site have been published elsewhere (Arvisais et al. 2002, 2004; Walde et al. 2003), these papers do not adequately describe the nesting area. The nesting area was an old, yet operational, gravel pit. Nesting was confined to a portion of the gravel pit that has remained essentially undisturbed by mining activities for > 50 years. The substrate in the pit was comprised of a finemedium gravel and sand. Very little vegetation was present, except around the perimeter of the gravel pit itself. The nesting site was approximately 1.5 ha and located from 10 - 100 meters of the river. Three infrequently used access roads crossed the gravel pit.
Techniques.—During May of 1996 - 97, intensive searches of the entire study area were undertaken to locate as many G. insculpta as possible. Each subject was marked, measured, weighed, aged, and sexed as described in Walde et al. (2003). In addition, radio transmitters were attached to 16 adult female G. insculpta in May 1996 (Arvisais et al. 2002) to document nesting sites as well as movement patterns to nesting grounds. In 1996, turtles were not relocated frequently enough for movement patterns to be described accurately; therefore, telemetry data reported for pre- and post-nesting movements are for the 1997 field season only. Turtles were located daily from 26 May until the end of June. Locations were recorded with a Trimble GeoExplorer I (Trimble Navigation Ltd., Sunnyvale, California, USA) Global Positioning System (see Arvisais et al. 2002) and plotted on a topographic map to obtain information on female nesting migrations.
Nesting.—The nesting area was observed daily from 26 May until 1 July 1996 - 97 for signs of nesting activity and/or the presence of turtles. Observations at the nesting area were made using binoculars and/or spotting scopes. As female turtles in a pre-nesting state are known to be sensitive to visual or auditory disturbances (e.g., Kaufmann 1992; Horne et al. 2003), care was taken to remain silent and hidden behind a blind. Once females began to arrive at the nesting site, they were watched carefully for signs of nesting. A constant watch was kept from approximately 05:00 to 21:00. Typically, turtles that remained in the gravel pit after sunset were nesting; in which case, observations continued until the turtle retreated. The date, time and general location of gravel pit entry and exit points were recorded for each turtle. Each time a turtle was observed in the gravel pit it was termed a “visit”; a visit represented one bout of activity in the gravel pit. If a turtle exited and returned to the gravel pit within 60 minutes it was considered a single visit. The first day that a female was seen attempting to nest was recorded as the initiation of the nesting season and the last observed nest excavation marked the end of the season, as per Congdon et al. (1983). Once a subject was deemed to be making a true nest, constant watch was initiated to determine the time when the first egg was laid. As turtles were no longer visibly disturbed once egg-laying had begun, subjects were approached so the number of eggs could be counted, provided no other turtles were in the gravel pit or within sight of the laying female. If direct approach was not possible, observers tried to position themselves so that the number of eggs laid could be ascertained directly or indirectly, i.e., head retractions (Harding and Bloomer 1979). The time when the nest was covered completely and the female began to move away from the site was recorded as the completion of nesting. At this point, an observer approached the turtle to confirm its identification and notes were taken on behavior of the turtle and direction of retreat. The exact location of the nest in relation to three numbered stakes (triangulated) was measured. The nest itself was staked with a metal spike and numbered tag, placed 1m north of the nest. If the turtle had not been identified previously, it was marked and all aforementioned data were recorded. For each year, we divided the nesting season in half and nests constructed during the first half were categorized as early nests, while those in the latter period were deemed late nests. The data from the two years were pooled to assess whether early nesters had greater nest success than those nesting in the late category.
Nest monitoring and hatching.—Nests were visited two to three times a week throughout the summer to obtain information on predation, natural, and unnatural disturbances. In mid-August, nests were visited daily to determine when hatching began. Hatchling emergence was characterized by a hole in the substrate approximately 3 - 4 cm in diameter. Upon observing the first of such holes, 1 m2 covers were placed over all remaining nests to restrain the hatchlings when they emerged. These wooden frames measured 2.5 cm by 10 cm and 1 m per side and were covered with 1.27 cm hardware cloth. These covers were partially buried in the ground to prevent hatchling escape. Moreover, fern fronds were placed on top of the cover, in the northeast corner, to provide some shade for newly emerged hatchlings. The nests were then checked two to eight times per day for the emergence of hatchlings and the date(s) of emergence were recorded. The number of eggs and hatching success were calculated by adding together the number of young that emerged from the nest, plus, the number of unhatched and/or rotten eggs in the nest cavity. The number of eggs that hatched was compared with the number of eggshell fragments and the appearance of these shell membranes. Eggshell membranes from which hatchlings emerged were white, soft, and leathery compared to those of eggs that had rotted, which were brownish, hard, and brittle (pers. obs.). Undeveloped eggs were defined as eggs enclosing a dead embryonic turtle; whereas, infertile eggs did not contain an embryo. If no hatchlings emerged from a given nest, the number of eggs was determined by a direct count following excavation. If a single hatchling emerged, the nest was considered successful. When no hatchlings emerged, nest failure was evaluated based on characteristics of the destroyed nest, the appearance of the eggs, and/or shell fragments. In 1996, nests from which no hatchlings emerged were left undisturbed to be excavated in the spring to assess if hatchlings spent the winter in the nest cavity. In 1997, all nests were dug up during the second week of November to determine if hatchlings entered the winter months alive in the nest cavity.
Terminology.—For the purpose of clarity the following terminology is used. Nesting refers solely to the act of nesting, i.e., excavating a nest, laying eggs, and covering it up. Nest success is the relationship between the number of nests from which at least one hatchling emerged and the total number of nests constructed. Hatchling recruitment is the number of hatchlings that emerge from nests. Recruitment success is the percentage of hatchling recruits compared with the total number of eggs known to have been laid in the nest site.
Although nest-site philopatry was first defined as the mechanism by which “females come back to the same area where they hatched to lay their own eggs.” (Reinhold 1998), it was subsequently redefined as the “return of females to the same geographic location, irrespective of whether this nest site is their natal site” (Valenzuela and Janzen 2001). Due to the long and somewhat ambiguous history of the word “philopatry” in the literature (e.g., Mayr 1963; Greenwood 1980; Waser and Jones 1983), we follow Rowe et al. (2005) in using “nest-site fidelity” to represent what Valenzuela and Janzen (2001) termed nestsite philopatry.
Statistical analysis.—To obtain information on the number of females in the population, estimates of mature turtles (>180 mm) were calculated using the Mt model (Rivest and Lévesque 2001). Differences between means were tested for using the Student’s t-test. Contingency tables for analysis of nest success as a product of nesting time were analyzed with the adjusted G-tests of independence (Gadj) using William’s correction (Sokal and Rohlf 1995). Correlations among populations were analyzed using linear regressions in SigmaPlot 8.02 statistical software (SPSS Inc. 2002). Statistical significance was accepted at P
Answers to all your Biology Questions
