Remote identification of specific behaviors by three-dimensional acceleration in tegu lizards S. merianae

• Grant number: 20/07520-3
• Support type: Scholarships in Brazil - Scientific Initiation
• Effective date (Start): August 01, 2020
• Effective date (End): January 31, 2022
• Field of knowledge: Biological Sciences - Physiology - Compared Physiology
• Principal researcher: Kênia Cardoso Bícego
• Grantee: Ane Guadalupe Silva
• Home Institution: Faculdade de Ciências Agrárias e Veterinárias (FCAV). Universidade Estadual Paulista (UNESP). Campus de Jaboticabal. Jaboticabal , SP, Brazil

Abstract

With the advent of contemporary technologies for remote behavior monitoring, it is possible to determine daily and seasonal activity patterns and organism-environment interactions in free-living animals without experimenter interference. Among these devices, accelerometers are used to quantify animal activity by means of changes in acceleration in three orthogonal axes, X Y and Z, corresponding to impulse, balance and elevation acceleration, respectively. The validation of patterns of behavior to be detected by accelerometers can be used for remote measurements in the field. In the present study, we will use three-dimensional acceleration sensors to describe behavioral patterns of the tegu lizard Salvator merianae. This species exhibits a seasonal cycle of hibernation during winter and high activity in spring and summer, with reproduction occurring during spring. Thus, the objective of this work is to validate the use of three-dimensional acceleration to distinguish specific behaviors of lizards S. merianae in captivity, by comparing visual observation of specific behaviors with three-dimensional acceleration data. Our hypothesis is that it will be possible to detect specific behaviors of the tegu from the three-dimensional acceleration data without direct observation of the animals. For this, 5 females and 5 males will be used, all equipped with a vest where an accelerometer will be anchored in the dorsal region, between the shoulder blades. The video recordings of the behaviors will be made by means of three cameras positioned on the walls of the animals' enclosure, thus avoiding the presence of the experimenter on the spot. From the videos, patterns of behavior will be identified and classified and the acceleration data will then be extracted using the SAS 9.4 software. for the calculation of variables related to the activity of tegu. Later, these variables will be used in the mathematical statistical analysis of the Random Forest type, which will validate the behavior models. At the end, the acceleration data precision will be presented to identify each specific behavior and also the best variable (s) for this identification. Our preliminary results, with some female ticks during the reproduction phase, indicate the possibility of identifying some specific behaviors from three-dimensional acceleration data. (AU)