A Model for Predicting Spatial Movement Patterns in Urban Bats

Thesis Abstract

Urban environments pose both opportunities and challenges for chiropteran species, with their complex spatial dynamics influenced by infrastructural development, resource availability, and anthropogenic disturbances. Despite the ecological significance of bats in urban ecosystems, there remains a limited understanding of their spatial movement patterns within human-modified landscapes. This study aims to develop a predictive model that delineates urban bat movement, facilitating informed conservation and urban planning strategies. The specific objectives include identifying key environmental and anthropogenic factors influencing bat movements, quantifying movement patterns across different urban zones, and formulating a robust predictive framework based on empirical data. The research adopts a mixed-methods, explanatory sequential design grounded in a positivist paradigm. The study population comprises insectivorous and frugivorous bat species commonly found within metropolitan areas, specifically targeting populations within the metropolitan city of Silvertown. A stratified random sampling approach was employed to select 150 individual bats, with spatial locations tracked using radio telemetry and GPS loggers over a six-month period to capture seasonal variations. Data acquisition involved the deployment of lightweight tracking devices fitted to individual bats, complemented by environmental data collected through GIS mapping of urban features such as green spaces, water bodies, building densities, street lighting, and noise levels. Data analysis incorporates both descriptive and inferential statistical techniques. Spatial movement data will be processed using Geographic Information System (GIS) tools to generate movement trajectories and density heatmaps. Analytical frameworks include multiple regression analysis to identify significant environmental predictors of movement distances and directions, and spatial autocorrelation tests (Moran’s I) to assess clustering tendencies. To address the predictability aspect, the study will employ machine learning algorithms, specifically Random Forest models, to construct a data-driven predictive framework. This framework will be validated using cross-validation procedures to assess accuracy and robustness. The theoretical foundation integrates the Landscape of Fear theory and the Central Place foraging theory, providing conceptual lenses to interpret movement decisions within urban landscapes. Expected findings posit that elements such as proximity to water bodies, green corridors, reduced light pollution, and lower noise levels significantly enhance bat movement frequency and range. The developed model is anticipated to reliably predict movement patterns based on urban environmental variables, offering a tool for urban planners and conservationists. The anticipated contribution to knowledge includes establishing a quantitative, empirically validated model of urban bat movement that integrates various ecological and anthropogenic factors, filling existing gaps in spatial ecology and urban wildlife studies. The study concludes that understanding spatial movement is crucial for mitigating human-wildlife conflicts and promoting biodiversity conservation in cities. Recommendations include implementing green corridor networks, reducing light pollution in key bat habitats, and incorporating movement models into urban planning policies to enhance urban biodiversity resilience. Future research directions suggest longitudinal studies to examine long-term movement dynamics in relation to urban development trends, as well as expanding the model to include additional species and ecological interactions. This research advances the scientific understanding of urban bat ecology and provides practical tools for sustainable urban landscape design, contributing to the integration of ecology-based solutions into metropolitan development strategies.