A Functional Framework for Behavioral Network Theory in Urban Birds

Thesis Abstract

Urban birds inhabit rapidly changing landscapes where individual and collective behaviors emerge from dynamic interactions with anthropogenic environments. Yet a cohesive theoretical framework linking individual behavioral rules to network-level dynamics in urban avifauna remains underdeveloped, limiting predictive understanding of how urban structure, resource distribution, and human activity shape social coordination, information flow, and resilience. This study aims to develop a Functional Framework for Behavioral Network Theory in Urban Birds, articulating how micro-level behavioral rules modulate meso- and macro-level network properties under urban pressures, and to evaluate the framework’s explanatory and predictive power across species and urban contexts. Specific objectives are to (i) identify core behavioral modules—foraging, vigilance, communication, and social affiliation—and delineate their interaction rules; (ii) construct a functional network model that links individual actions to emergent network metrics (density, modularity, centrality) in two urban centers; (iii) quantify how landscape features (green cover, impervious surface, feeding hotspots) and disturbance regimes (traffic, noise) modulate network structure; (iv) test predictive relationships between network configuration and fitness proxies (reproductive success, survival) over a multi-year temporal window; and (v) compare cross-species applicability of the framework using dataset from urban populations of the Great Tit (Parus major) and the House Sparrow (Passer domesticus). A mixed-methods design integrates observational ethology with quantitative network analysis. The study samples 320 individually tagged birds across four urban neighborhoods in each city, sampled over three breeding seasons. Observational data are collected via focal follows and scan sampling to record perching duration, foraging bout structure, vigilance frequency, call sequences, and affiliative interactions, coded with established ethograms. For network construction, proximity and interaction data are translated into weighted adjacency matrices representing co-occurrence, communication exchanges, and cooperative foraging ties. Landscape context is quantified using high-resolution GIS metrics foliage density, impervious surface proportion, proximity to feeding sites, and noise index. Data analysis employs dynamic social network analysis (DSNA) to derive temporal networks, exponential random graph models (ERGMs) to test structural tendencies, and multi-level Bayesian hierarchical models to relate network metrics to fitness proxies such as fledgling success, juvenile recruitment, and survival rates. Structural equation modeling (SEM) will integrate behavioral modules as latent constructs to assess direct and indirect pathways to network properties. Model comparison will utilize information criteria (WAIC, LOOIC) and posterior predictive checks. Additionally, generalized linear mixed models (GLMMs) will test moderation effects of urban landscape variables on the strength of behavioral-to-network linkages. Theoretical grounding draws on Behavioral Ecology, Social Network Theory, and the Information Transfer Theory, with explicit references to centrality and modularity concepts, and integrates the niche-based perspective of habitat-driven behavioral stratification. Expected findings include (i) identification of robust behavioral modules whose interaction strengths predict network density and modularity; (ii) evidence that high-quality green corridors increase clustering and reduce fragmentation in interaction networks; (iii) demonstration that noise and disturbance reduce effective communication ties, altering centrality distributions and potentially elevating the role of certain “hub” individuals in information propagation; (iv) positive associations between favorable network configurations (moderate density, clear modular structure) and higher reproductive success and survival; and (v) partial cross-species generalizability with species-specific weighting of modules reflecting differing reliance on social information. The study contributes to knowledge by integrating micro-level behavioral rules with macro-level network dynamics in urban birds, offering a unified framework to predict how urbanization reshapes social information flow, coordination, and fitness. It has practical implications for urban biodiversity management, suggesting design principles that preserve functional social networks—such as maintaining habitat heterogeneity and reducing sudden sensory overload—thereby enhancing resilience of urban avifauna. The main conclusion anticipates that functional network structure mediates the adaptive value of social behaviors under urban pressure, and recommendations emphasize landscape planning that supports stable interaction networks and promotes continued social learning among urban bird populations.