PS 3-28 - Combining data from citizen science sources and a bird banding research station to evaluate patterns of change in American Goldfinch (Spinus tristis) populations in the south Bay Area, CA

Abstract: Bird populations across North America are declining, and this is expected to accelerate as climate change and habitat loss increase. These are global phenomena but result from what happens in specific regions. At the local scale, how can we quantify the nature of ecological change in bird populations? Scientifically-grounded bird capture data is valuable but, because it is highly localized, may be limited in its scope to capture complex patterns within regions. Citizen science bird counts, such as summer and Christmas counts, and sightings (such as those on eBird) may be useful supplements but must be evaluated judiciously.



Coyote Creek Field Station (CCFS) in Milpitas, CA has seen significant decline in American Goldfinches (AMGO) since 1990. Once among their most common captures, the numbers have declined precipitously: from 2330 captures between 1990-1999 to 45 in 2010-2019. We wanted to supplement CCFS data with citizen science data to understand what is driving this decline.



Audubon-sponsored bird count data shows similar trends in AMGO numbers to those seen at CCFS. In the southern San Francisco Bay Area region, analysis of data from three Christmas Bird Count circles shows that American Goldfinch numbers sighted (total individuals or numbers normalized by observer hours) showed a steady decline from 1950 to 2022, with a significant (p< .001) negative correlation across time. A weaker but still significant decline was seen in summer bird counts, and the same pattern of decline held in a smaller analysis of the years 2012-2021 (a time frame chosen for comparisons with eBird data).



Our community-college-based team compiled eBird observational data from 30+ sites in the South Bay, selected within random sampling “territories”. Sites were included based on extent of data record, number of observers, and minimum number of AMGO observations. We compared sightings from 2013-2017 with 2018-2022. In this short time, there was, on average, a 13% decrease in American goldfinches observed, and the pattern of decrease was not spatially random. Sites closer to San Francisco Bay were more likely to show declines, compared to sites further from the shoreline. This is somewhat surprising, because these areas have the riparian zone type that local American Goldfinches prefer. However, microclimates closer to the Bay are more likely to be affected by the declines in fog that have been noted over the last 40 years, and research in Marin and Monterey counties has shown American Goldfinch populations being associated with coastal fog.