How do I observe the absence of something?

[phma]

I went to Earth Fare as usual today and saw and heard no sparrows. They live in treehouses in dogwood trees. I’ve observed them even in winter. How can I observe that they weren’t there? Should I make an observation with no photo or sound, identify it as a sparrow, and explain in the notes that I’m observing their absence?

[myrmepropagandist]

I don’t know if this is a literal question or a deeper philosophical one. Perhaps it’s both.

But being practical to note the absence you could mention, if you see them again, or if you see others near by how long it has been since you saw this population.

You could observe what other life is in their place as well.

[yongestation]

Technically the trips feature allows you to do this, though it seems somewhat neglected.

[pfau_tarleton]

Don’t make an observation, even if you explain that it’s a non-observation because the assumption is that every observation is an observation of an organism that is present at that time and location. It’s not easy to filter your non-observation out from real observations even with the explanation.

[AdamWargon]

I was reading about trips, and I found a comment from @loarie from 2019. If you want to learn more about trips, or if you want to refresh your memory, here is the comment:

@loarie

Here’s an attempt to provide a little background on Trips.

The basic back end of trips was added to iNat back in 2014 to support this app was built by a third party developer on the iNat back end with a very small grant from BHL. The basic idea was for the app to show you nearby historic GBIF points, for you to try to relocate them and record presence with a new iNat obs, and to store absences in this trips structure. The app never really worked either because it didn’t have the funding to properly build/support it and/or it was just poorly conceived. The app is not maintained and no longer works but I think all of the older trips cmcheatle mentioned aside from a few us on the team made for testing are from this app. The website also had the basic web UI for trips stubbed out since 2014 but because the tool was only there to support the biocaching app, we didn’t put too much thought into it being user facing.

This year, our Citizen Science colleagues at the California Academy of Sciences has a ‘UBIF’ collaboration with land managers that are interested in developing this work a bit further. Their interest is in the presence/absence of surrogate species to better understand how their managing habitat - e.g. whether this habitat in the San Francisco Presidio has California Man-root which could be a surrogate for the ecosystem they’re trying to manage for https://www.inaturalist.org/trips/24446-ubif-lobos-creek-oak-woodland-plants-1-23-2019

Internally, we’re also interested in better understanding/modeling species distributions on the site to improve iNaturalist as both a species identification tool and conservation tool (understanding species distributions is important for both of this). I don’t think we know exactly where we’re going with this yet but we’ve been doing a little bit of work modeling distributions from iNat data and validating it with eBird presence/absence data and its possible that having access to presence/absence data for non-bird data may help validate and improve these models. Here’s a paper I like from a colleague at Berkeley
https://besjournals.onlinelibrary.wiley.com/doi/10.1111/2041-210X.12242
and some of the cool work folks are doing at eBird
https://ebird.org/science/status-and-trends/
if you want to get a sense for the general distribution modeling arena we’re exploring

Back to the trips feature - ( a ) the UBIF collaboration, ( b ) our interest in understanding/modeling/validating distributions, and ( c ) vague but persistent interest in this area fro the iNat community were behind the recent updates to the trips feature - which mainly entailed revamping what was there in 2014, creating the infrastructure for rendering out a presence/absence table from trips and accompanying observations https://www.inaturalist.org/trips/tabulate?taxon_id]=36204, better documentation, and some rounds of user testing with the UBIF collaborators. We didn’t publicize the updates because (a) its been a busy time of year with Spring, CNC, Seek updates, and other things we’ve been involved in and (b) we’re still trying to get a handle on use-cases for this kind tool and how it does or does not fit into the iNat universe.

An important thing to note about trips which is relevant to some of the forum threads, is that we’re generating presence absence data from plain ol’ observations and the context that trips provide around them rather than collecting a whole new parallel type of presence/absence data. This is appealing because it means we can explore the utility of this kind of data and data use without building and having to maintain a parallel system more or less unrelated to the core of iNat (obs and identifications) - think Guides which is a good example of what we don’t want to repeat. Its also cool because from a presence absence standpoint you can exactly mimic eBird checklists with trips - its not the most pleasant user experience to do so, but its possible to store and render out the exact same data formats. Notice I said presence/absence not abundance as a major caveat is that eBird checklists accomodate counts and trips accomodate presence/absence only.

As for whether this feature will be supported, I think we’re still in the exploratory phase trying to understand the use case more than the exact user interface. For example, at the moment one thing that I think is cool about trips is that you can put a clade like ‘Reptiles’ on your target list and then even if you only saw zero, one, or two species etc. the trip still can be used to generate absence data for every other reptile species. Contrast this with having to put Reptile A, Reptile B, Reptile C etc explicitly on your target list. But after looking at the trips people are posting, it seems like there’s a lot of potential for people to create ‘false-absences’ from this tool. For example:

1. alot of people seem to be putting things like ‘Plants’ on their target list but ignoring things like mosses which are covered by Plants (thus creating data that could be misinterpreted as absences for alot of moss species)
2. if you have Ferns on your list (e.g. https://www.inaturalist.org/trips/24758-city-inaturalist-challenge-class-field-trip) but post an obs ID’d to genus (https://www.inaturalist.org/observations/23714394) the trip is recording this as an absence for every species in the genus (because there were no obs of any species in that genus) but clearly at least one species was present.

One option might be to make it so that only species can go on the target list, to trade off ‘power’ for correctness of the data. You could also imagine a ‘without taxon’ filter added to the target list (e.g. Ferns without Horsetails), etc.

There also remains a lot of unclarity about how this kind of tool would work well with other issues that have been discussed a lot on the forum like obscured observations or captive/cultivated observations etc.

I’m definitely interested in bug reports but not so much UI optimization at this point. One thing I’d love is to get a little pile of data that we can play with in validating distribution modeling analogous to how we’ve been using eBird data for a non-bird group such as amphibians and reptiles. I’ll try in the next week or two to put together a little ‘call to action’ that describes the structure of a trip that would be most useful for this exercise (e.g lets each spend ~30m making a trip surveying for reptiles and amphibians) along with a little demo of how we’ve been using eBird presence/absence data to validate bird distributions and then assuming we get enough of the right kind of data we can do the same exercise with the trip data we collect and get a sense for whether this data is even that useful for distribution modeling and if not what need to change to make it more useful. Sound ok?

@aaravmishra Note the comment about guides

I guess this is an attempt to replicate the functionality of eBird? I barely know anything about eBird, so I’d love to hear any thoughts from birders!

[jnstuart]

Absence of evidence is not evidence of absence.

[myrmepropagandist]

How can I explain how thick the summer nights were with fireflies years ago, how the air hummed with the songs of the cicadas? How can I explain all of the insects that have vanished?

Someone told me that windshields stay clean, even down south, even in the deep summer because of “aerodynamics” I don’t believe them.

But more seriously, I do think it would be perfectly possible, scientific and proper to record a great ant colony now vanished by documenting their empty wooden halls or their silent mounds.

[fishkeeper]

Well, an empty ant mound is evidence of the ants having been there. A photo of where there used to be an ant mound before someone swept it away, with no signs of the ants, wouldn’t really be a viable basis for a non-casual observation.

Recording the absence of something is vitally important to conservation, but it doesn’t really fit into how INaturalist works. We kinda have to read between the lines for that, not rely on a specific feature.
(It would probably also not be a very good feature to try to add- I suspect a lot of observations-of-absence would be from people who just aren’t noticing the organism in question somehow. Which can’t be helped the way incorrect IDs can, because uploading a photo of a tree without sparrows in it isn’t exactly proof that there aren’t sparrows in the next tree over.)

[ncb1221]

The real answer, at least for birds, is to use eBird. iNaturalist just isn’t built for this kind of data. eBird specifically is. So I suppose you could use the trips feature if you’re intent on staying on iNat, but I don’t see how that would be particularly beneficial, even to yourself.

[aaravmishra]

I have not added any replies here, what do you mean?

[phma]

I added a note to the observation of when I was there and didn’t observe them.

I’m not trying to replicate a feature of eBird, since I haven’t used eBird.

[misumeta]

I usually make an observation of some random other species and write „negative report for yxz“ in the notes section.

[ruth-palsson]

Another possibility is to go back to your most recent visit when you saw them and add a note that on date XXX, none were sighted.

I do that with some mistletoes that disappear.

[samantha_knight]

I think it’s kind of tough if it’s a place you don’t venture often. I looooove shiny blue insects so much, so at my house I take pictures of my Zimmerman’s mud daubers pretty much every day when they’re here. For me, the first picture of the year is the first day they appeared that season and the last picture is the last day they were at my house.

I can know for certain that there aren’t any this year because I don’t have a single observation at my house. (Also because my pest control guy destroyed their nests on my house by “mistake” last year, which is a whole other story…) I also track dogbane leaf beetles at my local nature preserve pretty much daily and can tell you the same thing. I couldn’t imagine trying to say they were absent from somewhere else just because I didn’t see any the day I went. What if it was bad weather? What if there were too many people in the area? I just don’t think iNat can give accurate accounts of what is missing unless someone is tracking those organisms very regularly.