In this study, animal-borne data loggers were deployed on a Reeves’ pond turtle for more than 3 h and on a red-eared slider for 26 days. In previous studies on freshwater turtles, methods for deployment of transmitters differed based on the experiment [24, 44]. Two deployment methods were used in this study: 1) deployment on the carapace using a quick setting epoxy and 2) fixing using the holes of the marginal scutes with cable ties. The data logger fell off after approximately 1 week at the longest employing the deployment method using epoxy because the scales covering the surface of its carapace peeled off. In contrast, in the method of fixing with cable ties using holes, the data logger was retained on the animal for nearly 1 month before being retrieved. The scutes covering the surface of the carapace in red-eared sliders peel off during growth, and this phenomenon occurs more frequently in juveniles [45].
Therefore, the deployment method to fix with cable ties using holes in marginal scutes is particularly effective for young individuals in which scales tend to peel off easily. Because holes of marginal scutes are used for individual identification in general, it is necessary to distinguish the holes for fixing data loggers from the holes for individual identification, for instance, using holes of different diameters. In addition, because there are no studies evaluating the impact of drilling on individual behavior and survival rate, the size and number of holes should be kept to a minimum [38].
The total weight of equipment in biologging studies should ideally not be more than 3–5% of the body weight [46–48]. Studies on radiotelemetry in freshwater turtles have shown that the weight of an animal-borne transmitter is less than 5% of the body weight [24, 44]. The impact of deployment on the animals in this study was considerably small because the percentage of the total weight of equipment, including the data logger, transmitter, cable ties, and basement, to the body weight of animals was 4% for the Reeves’ pond turtle and 2% for the red-eared slider.
Acceleration data obtained from the Reeves’ pond turtle showed that the turtle moved between 15:38 and 18:10. This is consistent with the time period of movement estimated from the GPS data. Moving speed of 3.3 m/min of the Reeves’ pond turtle in this study is the first record of the moving speed of freshwater turtles, as no studies on their moving speed have been conducted.
The turtle moved to the capture site within a few hours of being released, 200 m from the capture site. It is considered that the turtle stayed around the capture point because the activity of the animal decreased. It has been noted that several freshwater turtle species, such as the western pond turtle and Blanding’s turtle, navigate based on information about their surrounding environment [49, 50]. Although the turtle was kept in the laboratory for 5 days after capture, it moved back to the capture site after release, which suggests that the Reeves’ pond turtle may have homing ability.
Previous studies have shown that red-eared sliders bask primarily during the day, forage in the early morning and late afternoon [51], and rest at the bottom of the water at night [52]. Our results partly support those of previous studies in that the less activity of the red-eared slider from sunset to sunrise indicates resting at night.
The Japanese pond turtle, an endemic species, faces threats of habitat degradation [13–16], interspecific competition [14, 16, 18], and reproductive interference with non-native species [16, 18–20]. However, the relative impact of each of these multiple factors remains unclear, and conservation measures for this species have not advanced significantly. Unfortunately, owing to the low capture numbers of Japanese pond turtles in the surveyed area, we were unable to conduct research in this study. In the future, it will be necessary to understand the behavioral ecology of this species, evaluate the impact of each factor on the Japanese pond turtle, and employ the acquired knowledge in devising conservation strategies.
In this study, we established a method for attaching animal-borne data loggers to freshwater turtles. Prior research has shown that acceleration can be used to classify animal behaviors, such as walking and swimming, and to calculate the proportion of the classified behaviors (behavioral time budgets [53, 54]). It is expected that the behavioral ecology of freshwater turtles, such as behavioral time budgets, will be revealed by utilizing the data obtained from animal-borne data loggers.