Data for Scent mark and social dynamics in free-ranging dogs
Description
This dataset investigates scent-marking behavior in free-ranging dogs in West Bengal, India. The data includes video recordings and behavioral observations of dog groups responding to controlled presentations of scent marks from different individuals (male and female) and groups (intragroup and intergroup) at various locations within their territories (near resources, resting sites, and boundaries). The study examined the dogs' reactions to these scents, including sniffing, overmarking, and other behaviors categorized using an established ethogram. The findings reveal that both male and female dogs display significant interest in scent marks, with males exhibiting stronger territorial responses. Overmarking was primarily observed in males, particularly in response to scents from other males and neighboring groups. Distinct behavioral clusters were identified in response to different scent mark types, highlighting the complexity of olfactory communication in free-ranging dogs and its role in territorial defense and intrasexual competition.
Files
Steps to reproduce
To reproduce the analyses, the dataset should first be prepared with detailed behavioral observations, categorized by experimental conditions, such as scent mark versus control, group identity, location, and individual responses. All statistical analyses are conducted in R Studio (version 4.2.0). The Wilcoxon rank-sum test is used to compare latencies between test and control groups and across genders, while the Kruskal-Wallis test evaluates differences across locations, followed by Dunn’s test for pairwise comparisons. Generalized Linear Models (GLMs) are performed using the `lme4` package to assess the effects of sample categories, group identity, and experimental positions on investigation durations and territorial response scores. Territorial responses are quantified using an ethogram-based scoring system, assigning numerical scores based on behavioral intensity. These scores are summed per individual to calculate final territorial response scores. Model assumptions, dispersion, and residual diagnostics are checked using the `performance` package to ensure analytical robustness. For behavioral clustering, a dissimilarity matrix is created based on the frequency of behaviors observed across responder categories, using the Bray-Curtis method. Hierarchical cluster analysis is conducted with the `vegan` package to construct dendrograms representing behavioral patterns, including all observed behaviors or focusing specifically on territorial responses. All statistical tests are conducted with an alpha level set at 0.05. The dataset must be accurately coded and aligned with the experimental framework to ensure reproducibility of results.
Institutions
Categories
Funding
Department of Biotechnology, Ministry of Science and Technology, India
BT/HRD/NBA-NWB/39/2020-21 (YC-1)