A half a billion biodiversity records

By Team eBird August 22, 2014

Recently, eBird updated the data we share and publish through the Global Biodiversity Facility (GBIF), an international infrastructure that provides open access to biodiversity data. One result of this refresh is that data accessible through GBIF’s network now exceeds 500 million records—a true milestone for access to biodiversity information. This short article explains how data are made available and includes an interactive map showing where observations come from.

With the assistance of our collaborators at DataONE and VertNET, we have now developed the pipeline for providing continued open access to eBird data to a broad and diverse constituency. eBird developers at the Cornell Lab of Ornithology publish the eBird Observational Dataset (EOD) to DataONE. The EOD is published annually and contains the basic information to identify a location and date where an observation of a bird species was made—anywhere on Earth.

The VertNET team has developed a suite of sophisticated informatics software that formats and curates the version of the EOD incorporated into GBIF. As a result, anyone has access to eBird data, through eBird, DataONE (dataone.org), VertNET (vertnet.org), and GBIF (gbif.org).

When viewing the EOD through GBIF.org, users can also search, filter and download a selected subset of the records in place without having to download the entire dataset. One can filter the data according to various criteria including taxonomic classification, location and date. So, for example you can limit a search for eBird observations of Strigiformes (Owls) in Brazil or Argentina since 2010, returning just over 3,500 results. These data are also accessible through a thoroughly documented API (which likewise supports filtering). Here are some key points regarding eBird and data access:

  • eBird data in the EOD format are released annually. The time span of the current release of the EOD is from March 1810 to 31 December 2013.
  • eBird data have now been collected from all countries and have recorded the occurrence of more than 97% of the World’s species of birds. The map below provided by GBIF shows all the locations where eBird submissions have been made. You can view the map at the bottom of this page, or view it at GBIF. EOD Metadata are also available at GBIF.
  • eBird data contributions are growing exponentially. The graph below shows the growth of eBird data since its inception in 2002. It took eBird ten years to gather its first 100 million records (reached in August 2012). But by fall 2014, eBird will exceed 200 million records, making eBird one of the largest and fastest growing biodiversity data resources.

eBird Growth

  • More than 14.5 million hours have been volunteered by eBird contributors to collect bird observations!
  • To date, over 120 peer-reviewed publications have used eBird data. There have been over 6500 data requests in the last 18 months.
  • Remember that all eBird data—in its full richness—is available for download at the eBird Download page. This is our eBird Basic Dataset (EBD). We ask for a brief abstract to document how the data are being used.
  • A final eBird data product is the eBird Reference Dataset (ERD), which is a carefully curated subset of eBird data that meet ideal standards for science. These include complete checklists and effort-based protocols (i.e., not Incidental), and the data are made richer by association with landscape variables (e.g., MODIS, NLCD) and zero-filling for species not detected. These are very large files, but are what make the dynamic animated migration maps possible from our research team.

Your efforts are increasing our knowledge of the dynamics of species distributions, and having a direct impact on the conservation of birds and their habitats. In the upcoming months we will provide more information on how YOUR DATA are being used in the research and conservation of birds, biodiversity, and the environments they live.

Thank you!

View map fullscreen