Aim: Island biotas face an array of unique challenges under global change. Monitoring and research efforts, however, have been hindered by the large number of islands, their broad distribution and geographic isolation. Global citizen-science initiatives have the potential to address these deficiencies. Here, we determine how the eBird citizen-science program is currently sampling island bird assemblages annually and how these patterns are developing over time.
Methods: We compiled occurrence information of non-marine bird species across the world’s islands (n = 21,813) over an 18-year period (2002-2019) from eBird. We estimated annual survey completeness and species richness across islands, which we examined in relation to six geographical and four climatic features.
Results: eBird contained bird occurrence information for ca. 20% of the world’s islands (n = 4,205) with ca. 8% classified as well surveyed annually (n =1,644). eBird participants tended to survey larger islands that were more distant from the mainland. These islands had lower proximity to other islands and contained a broader range of elevations. Temperature, precipitation, and temperature seasonality were at intermediate levels. Precipitation seasonality was at low and intermediate levels. Islands located between 10-60° N latitude and 30-40° S latitude were overrepresented, and islands located between 60-130° W longitude were underrepresented. From 2002 to 2019, the number of islands surveyed annually increased by ca. 96.3 islands/year. During this period, island size decreased, distance from mainland did not change, proximity to other islands increased, and elevation range decreased.
Main conclusions: The eBird program tends to survey larger islands containing intermediate climates that are more isolated from the mainland and other islands. These findings provide a framework to support the rigorous application of eBird data in avian island biogeography. Our findings also emphasize citizen science as a resource to support ecological research, conservation, and monitoring efforts across remote regions of the globe.
The bird occurrence data in this dataset was acquired from eBird (ebird.org).
The zip file contains 6,573 RData files organized by month, day, and year. Each file contains eBird checklists of non-pelagic bird species that were observed on that date on the world's islands. A total of 4,205 islands are represented in the dataset. Each file contains the the scientific name of the bird species (SCI_NAME), the unique submission ID for the eBird checklist (SUB_ID), the longitude and latitude of the submission, and the ISLAND_ID where the observation occurred. ISLAND_ID corresponds to OBJECTID in the Sayre et al. (2018) global islands database (doi:10.1080/1755876X.2018.1529714).