How urban impervious surface shapes bird foraging behavior in an arid city

Abstract

Wildlife behavior and interactions in urban ecosystems can vary across landscape types and species, with some birds dominating human-derived resources. This study investigates the relationship between urbanization, measured as impervious surface cover, and bird foraging behavior in the Phoenix metropolitan area in Arizona, USA. We conducted 126 feeding trials across 13 sites along a gradient of urbanization and bird interactions with anthropogenic food sources present. Trials were conducted using popcorn placed at random distances and orientations from a trash can. We recorded bird species visiting feeding stations and time to first visit during 20-minute trials and then quantified relationships among visitation timing (latency), species richness, and impervious surface coverage. Time to first visit was negatively correlated with the amount of impervious surface, with the highly urban areas having birds arrive at the feeding station the soonest. Great-tailed Grackles (Quiscalus mexicanus) comprised the most common visitor across all impervious surface levels. Non-native doves like Rock Pigeons (Columbia livia) and Eurasian Collared-doves (Streptopelia decaocto) were quicker to visit feeding stations compared to native doves like Mourning Doves (Zenaida macroura), Inca doves (Columbina inca), and White-winged Doves (Z. asiatica). Small urban-adapted generalists, like House Sparrows (Passer domesticus) and House Finches (Haemorhous mexicanus), tended to be more frequent visitors at highly urbanized sites compared to larger birds. These emphasize how species-specific foraging behaviors can differentiate resource use by birds in urban areas.

Keywords: Avian behavior; Coupled human and natural systems; Food waste; Foraging behavior; Human-wildlife interaction; Scavengers; Urban ecosystems.

Figure 1. Map of feeding trial locations across range of impervious surfaces.

Locations of 13 sites with feeding trials within the Phoenix metropolitan area, Arizona, USA. Values of impervious surface are depicted here based on the 2021 National Land Cover Database (United States Geological Survey, 2024). We characterized impervious surface in our analyses using the i-Tree Canopy tool (i-Tree, 2024). Study site locations ranged in longitude from −112.306 to −111.583 and in latitude from 33.221 to 33.608.

Figure 2. Five most common species/groups seen during feeding trials at stations with visitors across 13 sites in the Phoenix metropolitan area, Arizona, USA.

Mean in seconds (and SE) to first visitor to the feeding station (popcorn). Non-native doves include Eurasian Collared-dove (Streptopelia decaocto) and Rock Pigeon (Columbia livia). Native doves include Mourning Doves (Zenaida macroura), Inca Dove (Columbina inca), and White-winged Dove (Z. asiatica). The most common visitor was Great-tailed Grackles (Quiscalus mexicanus; photo).

Figure 3. Relationships of impervious surface area with time of first bird to feeding station.

Relationship of impervious surface area with time of first bird feeding station visits. Out of 126 feeding stations, a total of 62 were visited by any bird species. Across all species, time of first visits (in seconds) was negatively associated with sites with higher percentage of impervious surface. Point colors indicate the species identity of the first visitor to each station (in order of visitation frequency), with some species grouped for visualization purposes: native doves (Mourning Dove, White-winged Dove, Inca Dove), non-native doves (Eurasian Collared-dove, Rock Pigeon), and other species (Abert’s Towhee, Rock Wren, Cactus Wren, Northern Mockingbird, Greater Roadrunner, Curve-billed Thrasher, and European Starling). Shaded area represents the 95% confidence interval of the trendline, as approximated by the ‘geom_smooth’ function in the R package ‘ggplot2’.