Increase the Scientific Value of eBird Lists

Cornell Lab has several great articles with best practices. Here are a few of the key points:

Do “complete” checklists wherever possible

• Why: Complete checklists help to indicate both presence and absence of species. Only complete checklists go into research and conservation projects that use eBird data, as well as eBird’s own Status and Trends pages.

• How: Where birding was your primary purpose (not just something spotted while driving or doing yard work), include all wild bird species that you are able to identify and select “Yes” when prompted if you are submitting a complete list.
  • It’s okay if there were species you were not able to identify! Marking “yes” just means that you have included all species you were able to ID by sight or by sound to the best of your ability.
  • Include all wild birds, even non-natives that have established populations (like European Starlings and feral pigeons), even the ones you may consider “uninteresting.” (If we only reported “interesting” species, that biases the data.)
  • It’s okay to leave off domestics like escaped Muscovy ducks, but if you do include them, make sure to use the appropriate species label that explicitly includes “(domestic type)”
  • Per eBird guidelines, do NOT include captive birds or free-roaming pets (like your neighbor’s Guinea hens, or a pinioned Mute Swan at the local clubhouse), or dead birds or eggs.

Use accurate date/time stamps

• Why: Date and time of day are important to any scientific data sets. eBird also has special flags for “nocturnal” observations which can be automatically triggered based on recorded start time.
• How: Select an accurate date and actual time (or best estimate) that you actually started the observation. Be sure that the duration (exact or best estimate) reflects the time you were actually birding
  • If submitting a list after the fact, do NOT use the current date/time! Otherwise, it may look like you were observing birds later than when they were actually seen, which can become problematic
  • When manually selecting a time, be mindful of AM versus PM!
  • On the mobile app, adjust the date/time as needed before tapping “Start Checklist”
  • On the website, select location, select start date, select protocol (usually traveling or stationary), and then select start time.
• Note: Stationary or Traveling are the options to pick if you can honestly record start time and effort. “Historical” protocol is more appropriate in some scenarios. See more about eBird Protocols

Report species to the best of your ability

• Why: Accuracy (to the best of one’s ability) is the foundation of eBird data’s usefulness.
• How: Report species you were able to positively identify by sight or by sound. (Please, please, PLEASE don’t rely solely on Merlin or similar apps for ID!) If uncertain, leverage the group or other resources for help (even a “bad” photo may contain diagnostic clues!). You can also use a “slash” (e.g. “Greater Scaup/Lesser Scaup”) or “sp.” option (e.g. swallow sp.), or leave it off altogether.
  • It’s okay to be uncertain! Even the most experienced birders can’t identify every single bird they come across. And reaching out for guidance can be a great learning experience, for you and for the other folks in the group.
  • Seriously, take Merlin ID, BirdNet, Sound Sleuth, Google Lens, and similar identification apps with a hearty dose of salt. They can sometimes be accurate, but sound ID especially is easily fooled by mimics (Northern Mockingbird, Brown Thrasher, etc), and technology is not 100%. It’s always a good idea to do a “sanity check” by comparing your photo or recording to other known recordings of the suggested species.
  • It’s also okay to leave a bird off the list if you truly couldn’t identify it; you can still consider your list “complete” if it includes everything you were able to ID to the best of your ability.
  • You may think a species is one thing in the field, then later review pictures or audio later (or someone else may review these) and realize it’s something else. There’s no shame — we have all done that! It is okay to change your list later upon this realization.

Acknowledge “Rare” flags and document accordingly

• Why: “Rare” flags help with data integrity where a species would be unusual, either absolutely or during a certain time of year. This layer of review prevents “accidental” reports. If a rare report is not accepted by a local reviewer, it will not be included in eBird data.
• How: Rule out other similar species who are more expected (please, please, please don’t rely solely on Merlin app for ID). If you are certain, document the sighting with descriptive comments and/or media.
  • If something gets flagged as “rare,” this is an opportunity to pause and think about the ID.
    • If you saw the bird, can you positively rule out other similar-looking species that are more expected?
    • If you heard a call, can you positively rule out a bird with similar calls, or a mimic like a Northern Mockingbird, European Starling, or a Brown Thrasher?
  • When you do report the rarity, add descriptive comments to the list documenting what you actually saw or heard and how you ruled out other, more expected species. (“Seen well” or “I used to see them all the time when I lived X” is usually not helpful.) If something is early or late, acknowledge that in the comments
  • If you have diagnostic photos or a recording, make sure to upload the media to eBird as soon as you can. Even a photo you may not consider “great” can still be diagnostic if the distinguishing field marks are visible!
  • “Continuing bird” is often acceptable if you are following up on a rarity that someone else previously found and is already well-documented. But if it’s something more than just an early or late bird, it’s a good idea to still try to get photos or a recording and put other comments if possible.

Include actual or estimated counts of birds, no “X”

• Why: Data with an X is excluded from many projects. X indicates presence, but that could mean anything from 1 to 1 million – which is not helpful for measuring population trends.
• How: See eBird’s tips for how to count. If you can’t get an exact count, your best estimate (counting by 5s or by 10s or more if needed) is more than fine. Even saying 40 if there were actually 50 is a lot more precise than a simple X!

Strive for accurate (but conservative) counts

• Why: Data helps to measure population trends, so accuracy is important.
• How: See eBird’s tips for how to count. A few things are common sense — if you double back on a trail, don’t want count the same birds twice — but they have a lot of great tips and best practices!

Use appropriate locations

• Why: Locations help scientists track more accurately where birds are found over time and in certain habitats (and as a bonus, also help other birders using eBird)
• How: Consider doing multiple checklists for multiple stops, or if you switch hotspot locations or habitats.
  • See more from eBird about hotspots versus personal locations and how to choose the right hotspot.
  • Keep traveling lists to less than 5 miles (one-way distance) where possible

Example A: Birding in Cades Cove

• Instead of: Doing one big eBird list covering 11+ miles of walking and driving at various locations.
• Try this: Start a new list for each actual stop using appropriate specific hotspots (e.g. Sparks Lane, Hyatt Lane, Gum Swamp). If needed, do separate “driving” checklists for in between.
  • eBird prefers that you do separate checklists for walking versus driving at any point of a “traveling” protocol.
  • However, if you are driving something like Hyatt Lane for a bit before stopping and birding some more, at minimum this can still be done as a separate Hyatt Lane checklist versus the other “hotspots” or driving the rest of the loop.

Example B: Birding on the greenway in Maryville near Bicentennial Park

• Instead of: Creating a “personal location” in the same territory that’s covered by the hotspot
• Try this: Use the existing “Maryville-Alcoa Greenway – Maryville Section” hotspot

Example C: Birding on your own property/neighborhood/friend’s house, or where there is not an appropriate hotspot nearby (e.g. various locations along the Little River between Walland and Townsend that are definitely not “Laurel Lake” or within the GSMNP)

• Instead of: Picking a random hotspot that is several miles away and not indicative of where you were actually birding
• Try this: Create a “personal location” the first time birding that area.
  • Bonus: When you bird there again, you can select that same personal location next time.
  • If a true public hotspot is later created for that area, you can merge your “personal location” with that hotspot to consolidate things later!