Hi everyone, As one of the people at the Lab of O who regularly works with data from eBird, I’ll give you my take on answering Deb’s question, from the perspective of someone who is interested in using the data from eBird for research, both for basic science, and applied conservation and management purposes. However, much of what I’ll write also applies to birders wanting to know when and where they can find a species of interest to them. For all of these purposes, it is important that we can learn both where a species of bird exists, as well as where a species does *not* exist. In order to understand where a species does not exist, eBird uses two types of information.
First, there is the answer to the question “is this a complete checklist of all bird species that you detected and identified?” If the answer to this question is “no, this is not a complete list”, then we have no clue whether any particular species not on that checklist was actually present. However, if the question is answered “yes, I am reporting all of the species that I saw and identified” then we at least know that the species in question was either: (1) really not present, or (2) present but undetected. The second type of information collected by eBird is needed in order to help distinguish between a species not being present, or that species just evading detection even though it was actually at the location. This second type of information is what we generally refer to as “effort information”, things like: the length of time spent birding, the distance traveled while birding, the time of day, and the number of people in the group that was birding. The longer someone spends looking for birds, the more likely it is that they’ll find and identify a species, when that species is actually present. The more pairs of eyes and ears looking and listening for birds, the more likely that any bird will be found…at least up to a point: we’re found that as the size of birding groups gets too large, the likelihood of finding some species will decline. The time of day is important, because some bird species have times of day (or night) when they’re easily found, but other times at which it’s essentially impossible to detect a species, for example because the species becomes silent and inactive. Simplifying things (a lot), it’s possible to figure out where a species is not found, by giving more weight to checklists on which a species was very likely to have been reported, *if* the species had actually been present (i.e. checklists from observations collected with sufficient effort and at a time of day when a species would likely to have been detected if it was present). Also, it’s important to have information from a large number of checklists, such that you’re more confident that a species is absent if many observers haven’t reported a species. It’s still impossible to be absolutely certain that a species is absent, especially for hard-to-see species like owls and rails, so in the end the best anyone can do is to conclude that it is highly likely that a species is not present. Some people might know that there are methods that have been created for analyzing “presence-only” data, but we’ve concluded that they should basically never be used with data from eBird. We’ve experimented with dumbing down eBird data --- removing all of the effort information --- and trying out these presence-only analysis methods on the resultant data. We’ve found that these presence-only methods do a worse job of describing where a species is and is not present. Even the creator of the most widely used method for analyzing presence-only data has told us that it makes no sense to use his analysis methods with information from eBird. These presence-only analysis methods are a sort of desperation option for use with information that comes from sources like museum specimens, for which there are just a bunch of presence “observations” sitting in boxes and drawers. So…what has this got to do with camera traps? The problem is that there’s no good way to report effort (would you report just the minute in which a bird triggered the camera, the entire hour before the picture was taken, the half-hour afterward?). The idea of a “complete” checklist is also stretched to the breaking point, because these cameras don’t identify birds by their vocalizations which is very different than many or most birders, so a camera trap may have close to zero chance to detecting most of the birds species in an area. Also, is someone likely to report every species photographed, or just the species that they think are particularly interesting? Admittedly, it would be possible to cobble together some sort of effort information, but cameras are so different from human beings that any information from camera traps in eBird would just introduce noise into any sort of analysis or interpretation that is based on assuming that all of the observations are made by people. It would be a completely different story if *all* of the observations are being made by camera traps, in which case observations from the cameras are comparable (Snapshot Serengeti is a good demonstration of this, I think), and the occasional human-made observation would just cause problems. Basically, eBird was not designed for, and shouldn’t be used for information from camera traps (or reports of dead birds, or reports of birds seen and heard from live-streaming feeder cams). That’s not the sort of information that eBird was designed to collect. Having written that, personally, I see nothing wrong with including a camera trap’s photograph as an accompaniment to an eBird checklist documenting observations that a person has made, even if the human observation was from a different (but nearby) period of time, like hearing an owl at dusk and then having the camera trap photograph it later in the evening. A photograph like that is just support for the auditory identification. I will hasten to add that I am just expressing my own personal opinion here, and I am absolutely, totally *not* stating or creating any official eBird policy. Finally, I’ll quickly note that eBird checklists of camera-trap “observations” are problematic from a birding perspective, because it’s not clear if a bird watcher has actually seen or heard a species themselves, and there’s a potentially slippery slope running from reporting that one has seen a species in a photo from their camera trap, through to having seen species on a feeder cam, and further into a gray zone of what it means to put a species on one’s life list. An arbitrary line has to be drawn somewhere, and eBird (and, I think, birders in general) draw that line at requiring an actual human to have detected a species in real time. For those of you who made it to the end of this uncharacteristically long CAYUGABIRDS-L post, apologies for my verbosity, but I just cannot think of a briefer way of answering Deb’s question. Wesley From: bounce-124627226-3494...@list.cornell.edu <bounce-124627226-3494...@list.cornell.edu> On Behalf Of Deborah G Lauper Sent: Tuesday, May 12, 2020 5:35 PM To: CAYUGABIRDS-L <cayugabird...@list.cornell.edu> Subject: [cayugabirds-l] ebird reporting question re: motion activated photos Maybe someone can clarify or refer me to the right place. I have lived and birded in the Cayugabirds region for 30+ years and use ebird. I also spend winters in Arizona, specifically Pima County (near Tucson, Madera Canyon etc). Owls have been visiting our bird baths nightly. Great Horned and Western Screech, got great pictures on motion activated cams. I recently learned that ebird does not want motion activated pictures and also, don't bother reporting it unless you saw it directly. My question is why? The ebird expert/monitor from this county (great guy by the way) advised me of this and said he doesn’t agree with it and doesn’t know the rationale either. Anyway, it’s amazing what we get on our motion cameras, mammals and birds, and it would be great to be able to include the birds on ebird. I want to be as accurate as possible and follow the rules so I deleted my owl pictures and sightings but I’m disappointed and hope I can get an explanation. Deb Lauper (Corning, NY and Sahuarita, AZ) -- Cayugabirds-L List Info: Welcome and Basics<http://www.northeastbirding.com/CayugabirdsWELCOME> Rules and Information<http://www.northeastbirding.com/CayugabirdsRULES> Subscribe, Configuration and Leave<http://www.northeastbirding.com/CayugabirdsSubscribeConfigurationLeave.htm> Archives: The Mail Archive<http://www.mail-archive.com/cayugabirds-l@cornell.edu/maillist.html> Surfbirds<http://www.surfbirds.com/birdingmail/Group/Cayugabirds> BirdingOnThe.Net<http://birdingonthe.net/mailinglists/CAYU.html> Please submit your observations to eBird<http://ebird.org/content/ebird/>! -- -- Cayugabirds-L List Info: Welcome and Basics<http://www.northeastbirding.com/CayugabirdsWELCOME> Rules and Information<http://www.northeastbirding.com/CayugabirdsRULES> Subscribe, Configuration and Leave<http://www.northeastbirding.com/CayugabirdsSubscribeConfigurationLeave.htm> Archives: The Mail Archive<http://www.mail-archive.com/cayugabirds-l@cornell.edu/maillist.html> Surfbirds<http://www.surfbirds.com/birdingmail/Group/Cayugabirds> BirdingOnThe.Net<http://birdingonthe.net/mailinglists/CAYU.html> Please submit your observations to eBird<http://ebird.org/content/ebird/>! -- -- Cayugabirds-L List Info: http://www.NortheastBirding.com/CayugabirdsWELCOME http://www.NortheastBirding.com/CayugabirdsRULES http://www.NortheastBirding.com/CayugabirdsSubscribeConfigurationLeave.htm ARCHIVES: 1) http://www.mail-archive.com/cayugabirds-l@cornell.edu/maillist.html 2) http://www.surfbirds.com/birdingmail/Group/Cayugabirds 3) http://birdingonthe.net/mailinglists/CAYU.html Please submit your observations to eBird: http://ebird.org/content/ebird/ --
