Hi All! New to iNat Forums so if this is in the wrong category, apologies in advance.
I had a question about bird population maps. As I’ve been getting more into birding I notice population maps on merlin/ebird vs inaturalist are significantly different. Some species Merlin classifies as “rare” in LA for example are tagged everywhere on inat and seem to in fact be fairly common (thinking: greylag geese or even neotropic cormorants). For another example I’ve recently been on the hunt for some white-breasted nuthatches in my area which are not uncommon based on inat, but every time I include it on my ebird checklists they’re flagged as unusual. Any reason for this?? Does the Cornell suite only track native populations? Is it just a more stringent database? Thanks in advance!
One thing to keep in mind is that iNat bird data will be biased towards folks posting their rare bird sightings. There might be many sightings of rare birds because each time they are sighted people will report them, which they would not necessarily do for every common bird that they see, so it is hard to judge relative rarity by iNat data. eBird data is a bit better suited to that because when people submit checklists they ideally report every bird that they see, common or not, as well as estimated numbers.
That’s a really good point I hadn’t thought of! Thanks for your input.
I’m not sure that eBird is necessarily more stringent, but the system works very differently. Data quality is assured through filters that flag birds as rare if they are unusual for the location and/or date. Volunteers then review the rare observations and make a judgement based on details you provide. This is very different from iNat, where there are no filters that flag unusual reports, and IDs are confirmed by the general public rather than by designated reviewers. Merlin can mark things as “rare” but this generally should be ignored, as it’s an automated thing that is strongly biased by the amount of people using Merlin/eBird in the area.
Regarding Graylag Goose, this is just a wild vs. domestic issue. The species is common in captivity in the form of domestic geese, but a genuine wild bird from Eurasia would be very rare in the US. The many iNat reports in the US are almost all of domestic birds. eBird can be used to report free-living domestic birds as well, but you will see the label “(domestic type)” associated with these reports, and they might be left at the “domestic goose” level rather than being ascribed to a species since most domestic geese have muddled hybrid ancestry.
As @elsemikkelsen noted, iNat will be biased towards rarities since observations are not part of checklists with other species. People will report something rare and ignore all the other birds in the area. You can do this on eBird, but many checklists are complete and contain all the species observed, rare or not.
Finally, consider habitat/precise location. White-breasted Nuthatch is not very regular at lower elevations of LA, but it is common in the mountains. This might be the issue you’re noticing.
It’s also worth noting LA is an extremely heavily iNatted place. There are over 1000 white breasted nuthatch observations which certainly looks like a ton when you zoom in on the map, but despite that it is still by my count only the 98th most common bird there. If by uncommon you just mean the half circle next to the name on ebird this is not an extremely selective criteria (it’s not saying this is extremely rare or anything, just not a bird you should expect to see all that regularly there), and a bird that just barely cracks the top 100 birds in a location seems fitting to earn it to me.
Are you referring to LA as Los Angeles or Louisiana?
The rare bird reports I get for my state highlight birds that are not commonly reported from a given area and/or for the time of year. I notice sometimes when someone uses a subspecies for a bird, that record also gets highlighted (included in the emailed rarity report) even if the species is not that unusual for that location/date. So not sure exactly how the algorithm works for determining what’s unusual.
One other consideration to add is that eBird can flag things that are out-of-season for birds whose abundance varies throughout the year. For example, a common spring migrant warbler might get flagged as rare if reported in the middle of winter. This can mean that the first early arrivals or latest stragglers get (rightly) flagged as rare when just a week later or earlier they would not.
One factor is surely the different demographics of users. If you are on EBird then you are already pretty serious about birding in some form. Most of the users on iNat are probably more into hiking in nature, looking for flowers, or whatever. Personally, I have comparatively little interest in birds (I’m far more into insects, spiders, reptiles and amphibians) yet I have over 2,800 bird obs logged on iNat nevertheless, some 17% my overall total. An awful lot of these are just mallards and pigeons and crows though – whatever I may happen across when I’m out and about.
I travel a lot for wildlife, but very rarely are birds my main objective. And I mainly take smartphone shots, which is great for frogs and flowers etc, but hopeless if I was after smaller or shier birds.
Any half serious birder on the other hand surely travels to different sorts of places, probably gets up before the sun much more often than I do, and always has their telephoto lens screwed onto on their camera. As a result, they’re sure to record a distinctly different profile of birds than I do.
One thing that both iNaturalist and eBird maps have a hard time portraying is how common a bird is in a location that it’s been reported. Species that are uncommon and only reported a few times a year can look to be common if you have many years of data, but those dots on the map represent single birds only seen for a few days, and then never again. eBird uses different shades to represent how often a species is reported from a location.
iNat users can also bias the data by reporting large numbers of individuals of certain species but not others, whereas in eBird users submit “complete checklists”, which means that you report all the species present to the best of your ability.
This is also probably something you are running into since “rare” changes throughout the season.
A friend of mine is a county reviewer for eBird and they said reviewers can mark certain species as Rare (or remove Rare from it) for a county. So my secondhand understanding is that the listing can be handled at the local level.
Looks like most of the OP’s observations are in southern California.
I’m a reviewer for eBird, and I can confirm that some things are flagged because of identification issues, even if they aren’t super rare. For instance, Hutton’s Vireos are flagged in Riverside County, California even though they are resident breeding species in some places, because about 20% of all photos of them before they were flagged as rare were misidentified.
Opposite that, some things that would be incredibly rare in a particular location aren’t flagged. For instance, a Bell’s Vireo in a backyard in downtown Los Angeles would be completely unexpected, but wouldn’t be flagged as rare because they are common in nearby preserves.
I think the thing to keep in mind is that iNatters photograph for many different reasons, a lot being opportunistic. iNatters may not post a complete list of what they see, perhaps because they are unable to get an image/sound of what they saw which is the requirement for an iNaturalist observation.
For example, I do not know birds, so I am not selective in what I photograph. If I capture an image or sound of a bird, then it is posted as an observation. I not sure that I’ve even posted a rare bird image - maybe I have, I don’t know.
Edit - just checked and found I do have a couple of rare ones from the Coorong, South Australia.
Yes, that is correct. I review for eBird as well and most of the rarity filters are set at the county level. Some places (including LA) have very fine-scale filters that go below the county level and focus on specific parts of a county.
I do inat, but not ebird. I will often find time to visit natural spaces whenever I travel. On my anniversary trip to Hawaii this spring, I spent some time seeking out endemics, some of which are highly endangered. I got pictures of some of the common stuff in the area because it was there and new to me, but if you looked at my observations from that trip, you would get a very incorrect impression about relative abundance because I SOUGHT the rare and endemic stuff.
Also the fact that eBird unlike iNat doesn’t rely on photo or audio evidence. You can of course include that evidence in eBird reports and it’s important for verifying rarities, but a huge part of the eBird database is checklists without supporting evidence.
I’ve seen a lot of arguments over which is stricter, but the answer is neither – they’re different with different goals in terms of data and different emphasis. They’re useful for different things to scientists looking for data as well.
As mentioned, with the Graylag specifically it’s mainly going to be a result of domestic birds (whether in captivity or dumped/escaped), which are not handled the same on eBird.
iNat only has the captive/cultivated or wild designations, and an escapee or released animal (i.e. dumped by the owners) Graylag living on its own is going to count as “wild” and therefore can obtain Research Grade on iNat.
On the other hand, eBird has three main designations for exotic species, which are naturalized, provisional, and escapee. Graylags would be “escapees” on eBird. Graylags do not have an established, self-sustaining breeding population in LA (whether you mean California or Louisiana, haha, I’m not sure). They are abandoned by owners or otherwise fugitives from captivity.
If someone does post a Graylag in the US on iNat, the vast majority of them are Anser anser var. domesticus, the domestic Graylag Goose. There are extremely rare actually-wild vagrants from Europe along the northeastern coast which need to be viewed very critically.
eBird’s page on exotics: https://support.ebird.org/en/support/solutions/articles/48001218430
iNat’s page on captive/cultivated v. wild: https://help.inaturalist.org/en/support/solutions/articles/151000169932-what-does-captive-cultivated-mean-
Escapee birds should be recorded in complete checklists on eBird, but people on eBird are not going out of their way to record a Graylag as an incidental if they happen to see one. On the other hand, even an iNat user who mostly looks at bugs might post that a domestic Graylag that was standing out at the edge of a pond.
iNat is significantly biased towards species you can record easily. Birds that are easy to photograph are therefore over-represented compared to even extremely common birds that are more difficult to get photos of (for example, birds that rarely come out from thick foliage, birds that spend more time really high up in the canopy, etc.). This is especially common for iNatters who are not hardcore birders. Cormorants and geese are both large birds that hang out in clear view and sit still for a long time, and relatively easy for even generalist naturalists to recognize and spot.
Many birds might be easier to ID and record as audio, and are rarely seen (including many night birds, but also a lot of birds like rails and so on). Audio recordings are far less commonly posted on iNat than photos, and are often under-IDed on iNat (with many IDers saying they just refuse to ID them at all) so even birds that are vocal but sneaky are underrepresented.
Birds that are quiet skulkers or high flyers that you can easily ID but not record just aren’t going to show up much on iNat, even if they’re super common. This means some birds that show as common on eBird are going to be seemingly uncommon on iNat, just like the reverse!
Also many of those “rare” designations on eBird are still going to have more sightings of a bird than on iNat, total, for most areas, just from the sheer number of eyes specifically looking at birds. There are 567 Graylag obs on iNat from Los Angeles County, total, both RG* and Needs ID, while there are 7712 on eBird. 355 iNat Neotropic Cormorant† versus 2818 on eBird. So rare is also relative!
* All of them that I saw at a quick glance should be Domestic Graylag. Some of these should not be RG, as they’re domestic Swan Geese or Graylag x domestic Swan Goose. I’m far from a waterfowl expert but I’ll go make IDs on the ones I’m confident about.
† A couple of the RG grade ones in the county are mis-IDed Double-crested Cormorants.
I find this interesting as, here in South Australia, I have been impressed with the willingness of birdos to listen to my recordings and help me figure our which is which. A good number of my bird observations are of birds I’ve had trouble seeing.
I post a fair number of audio observations. I don’t have the sort of camera (or the patience) required to get photos of a lot of birds. But I can do audio observations pretty easily. Kinda the same with frogs, but slightly different challenges photographing many of them.
I’ve noticed reasonable willingness to ID my audio observations. It’s usually the same folks, though.
As a birder who uses primarily eBird for posting birds, I 100% use eBird over iNat when searching for locations to find certain species, elucidate trends over seasons and years, pretty much anything. The data is imo much more complete, for reasons others have stated - eBirders tend to report ‘complete’ checklists with all present species, eBird is much more heavily used by birders to post sightings over iNat and by experienced birders moreover, and it’s not limited to just species that a birder has photographed or recorded. Every time I’ve searched a bird species on iNat the distribution map is a pale shadow of the completeness of eBird, making a lot of species seem rare and missing fine-scale detail about distribution and seasonality. Birders on eBird return to the same places throughout the year, creating a comprehensive picture of a site, a state, a species, etc. iNatters are more often ‘one and done’/opportunistic and target rarities and easy to photograph/identify species. eBird also has numerous visual tools to compile data, like the bar charts and many others. It’s much more well-suited to quickly and effectively summarizing vast amounts of birder data, submitted by, on average, more thorough and experienced birders than iNat.
This is not putting down iNat at all - it would be nice though if birders posted more on here to fill in some gaps, but I think that in itself is part of the discrepancy: seasoned birders tend to post on eBird and only carry over a handful of sightings to iNat if at all, so a much greater proportion of bird observations on iNat are from laypeople observing common city birds or rarities, but little in between.
Just one example, since it was something I was looking up not long ago. Vesper sparrow sightings in southern New Hampshire.
eBird:
iNat:
Data are also just laid out v differently on eBird, click on any of those points and it’s a location, not an observation. It will bring up a clickable list of every checklist the species has been submitted on at that location, with dates and observers, that bring you to even further details.
Until iNat catches up on most bird species, I’d use eBird when plotting out where to go looking for a target species.