Residential yards are a form of urban land use that cover a considerable amount of area in cities worldwide and provide important habitat for wildlife, especially when landscaped with native plants. Nevertheless, most native-plant landscaping and wildlife research in the northern temperate regions of the world has been conducted during the spring and summer breeding periods, leaving a gap in our understanding of the importance of residential yards and native plants as habitats for animals during the nonbreeding period. To fill this gap, we quantified the ecological role of native-landscaped yards to avifauna throughout Greater Los Angeles, California (L.A.) during the winter nonbreeding period, which is a time of year when the region hosts a high abundance and diversity of migratory and resident birds. We surveyed birds and habitat features from October-March of 2020 and 2021 at 22 pairs of native and nonnative-landscaped yards plus ten additional native-landscaped yards. We had three objectives for our study. First, we compared avifaunal communities, including feeding and nonfeeding behaviors, and habitat features between native and non-native-landscaped yards. Second, we quantified relationships between habitat features and bird richness, abundance, and feeding and nonfeeding behaviors — focusing on species affiliated with urban or natural terrestrial ecosystems of the region. Third, we documented feeding and nonfeeding behaviors by birds with native and non-native plants. Native-landscaped yards had a greater cover of native trees, shrubs, and herbaceous plants, and a higher cover of natural habitat elements, including leaf litter and bare-ground cover. Bird richness and abundance — especially bird species affiliated with tree and shrub-dominated ecosystems — were greater in native than nonnative-landscaped yards. Further, yards with a higher cover of native plants supported greater numbers of feeding birds, with individuals focusing their foraging behaviors on distinct native trees and shrubs, including Quercus spp. (oak), Heteromeles spp. (toyon), Arctostaphylos spp. (manzanita), and Salvia spp. (sage). Our results suggested that residential yards landscaped with native plants provide important habitats for birds during the nonbreeding period and are a viable approach for residents and cities if improving conditions for birds throughout the annual cycle is a goal.

Bird surveys and feeding and nonfeeding behavioral observations: We conducted three 20-min area-search bird surveys from October 2020 through March 2021 where we recorded the species and number of individuals using each yard (Watson 2003). We completed surveys between 06:30 - 11:00 hrs. and visited both yards of each pair on the same day to minimize weather and seasonal biases.

 Additionally, for each bird detected, we opportunistically recorded whether the individual was exhibiting feeding or nonfeeding behavior. Because most yards were small (Fig. 1), after a bird was detected, we observed the individual, while continuing to count additional species. We then documented whether a bird exhibited feeding or nonfeeding behaviors while in the yard during surveys. Feeding behaviors included gleaning for prey items on the surfaces of plants, foraging on the ground, feeding on seeds or fruit, obtaining nectar from flowering plants, and aerial maneuvers (e.g., sallying) (Wood et al. 2012).

Habitat features: We conducted vegetation and habitat surveys in each yard during the third round of surveys from January to March 2021. We used relevé methods to visually estimate the percent cover of vegetation at the ground, shrub, and tree canopy levels (Minnesota Department of Natural Resources 2007, Wood et al. 2010). 

Additionally, we used spatial analysis tools in ArcGIS to quantify habitat variables at the landscape scale in two ways (ESRI 2016). First, we used the 'measure' tool in ArcGIS to quantify the Euclidean distance from the centroid of each yard to the boundary of the nearest protected area (e.g., San Gabriel Mountains, Griffith Park, Bolsa Chica Ecological Reserve, Ballona Wetlands) and to the nearest urban greenspace (e.g., community gardens, urban parks, golf courses). 

Second, to quantify habitat features at the extent of the surrounding neighborhood, we used a light detection and ranging (LiDAR) derived data product that provided urban-form data for every parcel in Los Angeles County (Galvin et al. 2016). Using a spatial join to merge with our sample locations, we included seven variables from the parcel-level data: the average land value and the average percent cover of trees, grasses, bare soil, buildings, paved, and impervious surfaces. 

We used R for all data analyses. We have attached R Markdown files, with code, adjoining the datasets used for analyses related to our objectives.

R Core Team. 2017. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R-project.org/.