Skip to Main Content

Time for Review

By Ashley Peele

An image of a white crowned sparrow in the grass; this bird is grey with brown wings and a black and white cap

White-crowned Sparrow (no breeding code needed, despite all their singing in VA) – CO Bob Schamerhorn

As we continue to wind down our third breeding season, we encourage Atlas volunteers to take the time for a little self-evaluation of this season’s data.  A quick check for common errors can help save the Atlas review team a lot of time in the off-season AND help you improve your Atlas knowledge for next season.

As mentioned in a previous article on eBird Virginia’s data guys, we have implemented a breeding data review process to identify common errors in the database, which is described in more detail at the end of this article.  First, here are the ten most common breeding code errors identified during the data review process…

1. Presence does not equate to breeding

Be sure that a species is likely to be a breeder in your part of Virginia, prior to tagging it with an H (In Suitable Habitat) or S (Singing Bird) code.  These should only be applied to species that are potential breeders. Boreal forest or arctic breeders like most of our wintering and migratory species (Northern Pintails, Blackpoll and Cape May Warbler, White-throated Sparrow, etc.) should not get an H or S code.

This can be confusing, as many of our wintering birds are still present in Virginia when resident species begin to breed.  See our Breeding Timelines charts for information on what species are likely to breed in your region of Virginia and when their breeding windows begin.

2. Not all vocalizations relate to breeding

The S code should only be applied to breeding-specific vocalizations.  For songbirds, this is usually a straightforward distinction. The songs of warblers, vireos, sparrows, finches, tanagers, etc. are distinctive and well-known, whether given by the male or female bird.  However, some species, e.g. White-breasted Nuthatch, Cedar Waxwing, or Blue Jay, give the same calls all year round.  Because these calls are not breeding-specific, they should not be coded as S.

For more detailed information on the singing codes, check out the first article in our series on Tackling Tricky Breeding Codes.

3. Flip-side – Don’t forget to code all those possible breeders

Frequently, we review Atlas blocks that have many un-coded species, which are almost certainly possible breeders where they are being reported.  For example, folks often forget to begin coding Cedar Waxwings, Chimney Swifts, or Ruby-throated Hummingbirds as H (In suitable habitat), once their individual breeding season arrive.

We cannot upgrade such species for you, so spare your reviewer the angst and add codes to all the possible breeders in your blocks.

*Pro-tip: When looking at a block level data summary in the Atlas ebird portal, you can click on the ‘Breeding Evidence’ column to arrange the species according to their status.  This allows you to quickly see who is or is not coded as a possible breeder.

4. P only applies to dimorphic species

Remember to only use the Pair in Suitable Habitat (P) code, if you can distinguish between male and female plumage characteristics.  Species with identical sexes should never be coded as P.

If you observe other breeding behavior which suggests that a pair of birds is in fact a breeding pair, then a different and possibly higher-level code should apply.  See Tackling Tricky Codes (Pt 2): the Probables for more detail.

5. Distraction Display is a fairly rare code

From comments provided by you (the observers), we have noticed that many observations coded as DD are actually more appropriately categorized as agitated behavior (A) or territorial defense (T).

Distraction displays specifically refer to instances when a bird is feigning an injury or physically attacking you.  Since this code confirms that a species is breeding, please use caution when applying it.

6. Care with counter-singing

Observations of counter-singing birds should only be coded as T when the birds are obviously exchanging songs (what often sounds like duetting) along territory boundaries.  Simply observing multiple males singing in the vicinity of one another is not sufficient.

Note! While the M code still appears on the drop-down list of breeding codes via the eBird phone app, it is not a VABBA2 code and should be avoided.

7. Be wary of the wandering birds

Many species of colonial waterbirds (shorebirds, herons, gulls, terns, pelicans, etc.) are highly precocial when borne.  Fledglings will often move quickly away from natal areas, thus observations like FL (Fledged Young) and FY (Feeding Young) should only be used if the nest site is known and within the same block.  In fact, care should be used when applying any breeding codes to colonial nesting species.  Be sure that you are confident that nesting colonies are within a given block before adding breeding codes to such species.

Similarly, birds of prey like vultures and eagles forage miles away from their nest sites, even during the peak breeding months.  Use care when applying breeding codes to these species.

8. Be careful when applying CF to large bird species

Species such as hawks, eagles, osprey, terns, gulls, corvids, kingfishers, and herons will often carry large food items some distance, prior to eating them.  This is often because these birds have favorite perches from which they like to eat.  For these larger species, repeated carrying of food in the same direction is more likely to be for the purpose of delivering food to a mate or young.

**Tip from an Atlas Reviewer: Please provide notes on the observation when applying nuanced codes, such as DD, CF, or C, or when applying an unusual code to any species.  Reviewers read ALL notes for records that are flagged during the review process. However, if your atypical observation is not accompanied by any explanation, they are forced to assume it was an error.

9. Don’t forget about the S7 or C codes

The S7 (Singing 7+ days) code is an easy way to upgrade species from the possible to probable breeding category.  If you observe a species singing in the same location seven days apart AND within the safe breeding window (Breeding Timeline Charts), then you can tag it with an S7 code.

Similarly, the C (Copulation or Courtship Display) is another underappreciated tool for getting a species into the probable category.  Many species including Ruffed Grouse, Wild Turkey, American Woodcock, etc. use behavioral displays to advertise their territory.

10. eBird and the VABBA2 eBird portal are not the same thing

Don’t forget that breeding codes only go to the VABBA2 project, if you submit your checklists through our portal (ebird.org/atlasva).  The rule of thumb is that if a checklist has ANY breeding codes, it should be submitted to us.  If you have questions about changing portals, check out this article on getting your data to the right place.


Update on Breeding Data Review Process

In a recent Atlas article on the data guys behind Virginia’s eBird review process, we mentioned that a second tier of data review occurs outside of the eBird system.  This process looks for common errors related to breeding codes or breeding dates for records reported to the VABBA2 eBird portal.  If records are flagged by these filters, then reviewers assess each one to determine whether the error should be validated or reinterpreted.  If a record needs to be reinterpreted, we do not actually change your observation.  Instead!  We edit the breeding category assigned to a given code.

Below, you will see an example of a flagged record and the reason that it was reinterpreted.  The updated category will now be displayed on the species map, block summary pages, etc., but the original code will remain in your records.  If you’d like to learn more about this process, please read this article provided by eBird Central.

An example of a breeding code describing the signs on must look for in a breeding pair of birds

Example of Breeding Code Review & Reinterpretation

Note that an initial review has only been completed for the 2016-2017 datasets.  If you decide that you have old records needing adjustment, then please feel free to edit those.  Additionally, let us know if you think that a record was adjusted in error by emailing us at vabba2_at_vt.edu.

This current phase of Atlas data review is only a stepping stone towards a more developed review process. We continue to work with Cornell to develop these tools, including a communication system that will eventually allow us to communicate with you directly about records that have been reinterpreted.

For now, please remember not to feel bad about having some flagged breeding records!  We guarantee that all Atlas volunteers experience a dramatic learning curve when participating in this project.  Everyone from coordinators to data reviewers to volunteers have had records flagged. We are working to develop a system that not only improves the data quality, but also helps you learn more as the project progresses.

Thanks again for all your continued efforts throughout the first three years of this project.  We’re excited to dig into the 2018 dataset and learn more about the breeding information that you collected this year!

~ Dr. Ashley Peele, VABBA2 Coordinator

  • August 15, 2018