Tips for transcribing eBird checklists to iNaturalist

I use eBird a lot. I have tonnes of sightings pending transcription to iNaturalist. Any tips on doing so?

More of a meta comment, but it might be worth thinking about what you value about the transcription of the data/observations. Since ebird data are already “out there” science-wise, the observations you put into iNat won’t add much to that value.

However, they might add to your personal life lists on iNat, observations of rarer species that haven’t been included in the CV yet would be great, unusual observations that others might be interested in, great pics that aren’t in eBird data, etc. So considering some of those things might help you prioritize a list/save time from transcribing tonnes of sightings.

7 Likes

Thank you very much for the tip!

I agree,

To piggyback on this a bit it may be useful to understand the type of data submitted for each;

For eBird; Data is submitted for a list of species and individuals over a time period and distance (or stationary), with data mostly relying on the “honors system” as verification.

For iNat; Data is submitted for a single organism at a single point in time and single location. For the most part physical evidence needs to be provided for confirmation of the observation’s accuracy, although field sketches are accepted and you can always add casual observations (without media) if you want to capture something for your personal records.

If you’d like to keep a list of all of the species you’ve seen you could always submit an observation or two of each bird species you have listed on eBird so that it’s captured on iNat, but really it’s whatever you’d like to do. Me personally, if I’m out taking pictures I’ll pop any bird photos onto iNat, but for the most part I use eBird to keep track of most of my bird sightings.

4 Likes

(we/us pronouns) From our experience, ebird does not require evidence (photo or audio) of observation. Even for rarer species, a detailed written description can sometimes suffice for “confirmation,” especially for observers with high community credibility. Therefore, it would seem the amount of carryover from ebird to inat is dependent on what evidence the observer has collected, since our experience is that inat is built around submitting evidence.

Also, our experience has been that audio submitted to inat can take much longer for identification than photographs.

Our perception is also that ebird promotes quality of evidence submissions since observers are allowed/encouraged to rate photographs and sound recordings for Cornell’s Macaulay Library (highly rated evidence can become part of the featured description for the species). We personally tend to submit to ebird only higher quality images and recordings unless it of a rare species needed for confirmation.

Maybe it’s just us, but we are satisfied to submit to inat lesser quality evidence because we are aiming for identification, not a high star rating (five-star scale) like on ebird. These observations are not intended to disparage inat in any way.

2 Likes

I also use eBird a lot, and I am relatively new to iNaturalist. I would agree that observations on iNaturalist without photo or other evidence probably have little value outside of your own personal list. If you intend to do this mainly to contribute to science, and most of your bird sightings do not have photos, you may want to reconsider, as they will never reach the research grade level that scientist data users will require. Non research grade records aren’t even available for download in bulk anywhere I believe. If you are doing this to use iNaturalist as a personal record keeping system where all your sightings are in one place, that’s a valid consideration of course. I know that you can export all your sightings easily from eBird, but I don’t think a bulk import tool exists for iNaturalist.

I trust eBird observations, even without hard evidence most of the time, because the review process is quite thorough. A rarity flagging system is established for each bird species at a pretty fine scale. We know so much more on birds than most other taxa, that we can say that a Yellow Warbler seen on May 1st at a given site would be unusual, but may be totally fine 5 days later. Any observation that gets flagged then gets evaluated by a regional expert, who may then follow up if not enough details are provided. A truly rare bird would generally require hard evidence, or a very convincing description, and many of those will be eventually reported by other observers that can confirm the sighting. I don’t think this approach (not using hard evidence) could work for almost any other taxonomic groups.

Even on iNaturalist, I would only generally trust records of birds that came with evidence because they are not vetted through the same rigorous process as eBird records are. And even those with the research grade level, as a data user, I would be inclined to validate myself the photos for anything unusual, because it seems a bit too easy for incorrect records to get pass that initial filter. It works well most of the time, but since I focus on finding unusual records, I have also come across enough cases where the research grade was assigned on erroneous records. A downside of eBird is that common birds may get misidentified and fall through the review cracks, but this problem is somewhat drowned out by the quantity of data available for studies of population dynamics, which tends to be the main value of eBird.

Not all scientists require research grade (though it is required for export to GBIF).


You can indeed export casual/needsid observations from the export tool, e.g.



You can bulk import observations with the following fields: taxon name, date observed, description, place name, lat/lon, tags, geoprivacy. Additional info, like photos, annotations, etc. would need to be added manually.



Personally, I don’t particularly trust circularly verifying setups, i.e. “we expect it to be there, so we can safely assume when someone says it’s there, they’re right”.



Correct, they are vetted through a different process. Whether you consider something rigorous or not is rather subjective. I value the fact that iNat’s process is transparent and community-driven.

5 Likes

I agree with @cthawley and @natemarchessault - probably not worth importing eBird observations into iNat. eBird is great, there’s no need to cross post to iNat unless you really want to note some interesting records, especially if you have media evidence for them. I think you’re time is much better spent doing more birding in the field. ;-)

2 Likes

I use both eBird and iNaturalist. I can understand some redundancies, especially in well researched and surveyed areas like North America and Europe. But in areas with very few listings or where citizen science data is still only used by a few people (such as where I am, South Korea), using both is valuable for building a baseline. There’s still lots not known about birds in South Korea, so it’s worth punching in the same bulbuls and magpies when wandering through the area into eBird. Same goes for iNaturalist, even more so I think, cause the computer vision still recommends ridiculous suggestions for even common species. Adding the sightings improves the CV algorithm for future use, builds a baseline, and plus it’s a little fun entering all this stuff anyways (for me at least).

1 Like

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.