Observations of absence vs presence

[pisum]

just blue skies here, but i was wondering if there has ever been any thought into possibly one day giving iNat the ability to capture observations of absence of an organism? it might be hard to do community confirmations on such observations, but it might be interesting data, especially if those observations of absence can be combined with observations of presence.

for example:

1. obs 1 (presence): May 1 2018: here’s a majestic oak in a field. such a beautiful tree!
2. obs 2 (presence): Nov 30 2018: here’s the same oak from obs 1. its leaves are starting to fall. and uh oh! there’s construction machinery right over there.
3. obs 3 (absence): Mar 15 2019: the oak is gone. there’s a parking lot here now.

or:

1. obs 1 (presence): April 1 2018: here’s a Chinese tallow in a section forest with lots of small Chinese tallows popping up. we’d better get to work!
2. obs 2 (absence): April 15 2018: we pulled out the Chinese tallow, and now it’s all gone.
3. obs 3 (absence): Nov 15 2018: still no new tallow in this section of the forest. (hooray!)
4. obs 4 (presence): April 1 2019: dang! the Chinese tallow is back!

or:

1. obs 1 (presence): Sep 30 2018: first Monarch of the season in my backyard
2. obs 2 (presence): Oct 1 2018: another Monarch passing through
3. obs 3 (presence): Oct 9 2018: went out of town for a few days. but I still saw this Monarch in the backyard when I got back.
4. obs 4 (absence): Oct 10 2018: no Monarch sightings today.
5. obs 5 (absence): Oct 15 2018: no Monarch sightings today, and none since obs 4. they must all have passed through for the year.

[charlie]

It’s been discussed, and would be neat if there were a way to do it, but i don’t think there are any plans by the devs to implement anything like this (obviously they could correct me if i am wrong here)

[kimberlietx]

Interesting thought! @pfau_tarleton and I are currently tracking absence/presence locations of Anemone caroliniana in DFW/Texas. (They look very similar and frequently grow side-by-side the more commonly observed A. berlandieri, thus getting “overlooked” in observations.) Another application for your blue skies :)

[pisum]

@kimberlietx – how are you capturing absence? would you add a couple of links to examples?

EDIT: nevermind. i thought maybe you meant you were tracking this in iNat, but that’s probably not the case.

[htrudell]

I would love absence data options - surveying for dead birds needs all the absence data we can get! Three birds in one month could be three birds in one survey, or three in three surveys: the building could be bad, or it could be awful.

[jdmore]

Many organizations struggle with the need for this, including one I am associated with, NatureServe. Such data are invaluable for training species habitat models, not to mention informing everyday conservation decisions.

One nitpicky issue to keep in mind – it’s hard to prove a negative. “Absence” is a convenient term for communication, but what we usually mean is something like “undetected.” The distinction becomes important when the perception of “absence” serves as a green light to go forward with some kind of destructive activity.

[andrewgillespie]

One way you could do it is to keep added photos to the same observation, but that does actually violate the official definition of an observation. A special observation type for tracking something over time would be a really neat feature.

[tony_wills]

@jon_sullivan has always wanted this too, and has tried a few approaches - hopefully he will see this @ ping.

[charlie]

you can add a common plant (please don’t do fake data though! There are plenty of plants!) and use a field to document absence of something else.

[pisum]

@charlie

you can add a common plant (please don’t do fake data though! There are plenty of plants!) and use a field to document absence of something else.

i think no picture and “evidence of organism” DQA fields marked might be better and would be less likely to cause issues for people wanting to do IDs. still, if that observation showed up on maps of unverifiable observations, i think that could still cause people to misinterpret data as presence vs absence if they didn’t look closely at the observation details.

EDIT: nevermind. i reread your post and it sounds like you want to add a special “absent species” observation field to a present observation. i guess that could work.

[pisum]

@jdmore

“Absence” is a convenient term for communication, but what we usually mean is something like “undetected.”

very true. the other thing is that it’s easy to say i detected this organism somewhere in this circle, but more problematic to define the area that you didn’t detect something in.

[charlie]

Maybe I’m just angling for more plant data

[tony_wills]

To quote Jon_sullivan from Observation field: Taxon sought but not found www.inaturalist.org/observation_fields/10102

For several years now I’ve used the Sought but not found field (Yes/No) in combination with my observation ID to specify what I was looking for when I didn’t find anything. For example, if I searched for butterflies but found none, I’d upload an observation with the ID Papilionoidea and set “Sought but not found” to “Yes”.

This hasn’t been ideal because every time I do this it puts a pin on the map for Papilionoidea.

I’ve just added this new field, “Taxon sought but not found”, so I can do this a better way. Now I can add a new observation with no ID so it doesn’t show on the iNat map. Then I can specify the taxon I was looking for with “Taxon sought but not found”.

I could also answer the Data Quality Assessment question “Based on the evidence, can the Community ID still be confirmed or improved?” to “No” so nobody needs to bother trying to ID it.

[charlie]

yeah, if you looked for a species but did not find it, you should NOT add a point for it, even without a photo, because it will add a point to the map and create false data. Please don’t do that.

[KitKestrel]

@pisum

“Absence” is a convenient term for communication, but what we usually mean is something like “undetected.”

Or, as many have said before,

Absence of evidence is not evidence of absence.

On the other hand, since most of us are not all-knowing, looking and not finding is about as close as we’ll get to establishing that something isn’t there.

[jdmore]

@charlie

if you looked for a species but did not find it, you should NOT add a point for it, even without a photo, because it will add a point to the map and create false data.

If this were done in combination with downvoting the “Evidence of Organism” DQA, would that help, since Casual observations can be turned off on maps?

[charlie]

i suppose that could work. It still doesn’t seem ideal, since it isn’t really saying absence, it’s saying ‘i saw it but i don’t have evidence for it and also i didn’t see an organism’ and probably neither of those are true. Though i personally don’t care much as long as it’s not on the map.

[jdmore]

@KitKestrel

looking and not finding is about as close as we’ll get to establishing that something isn’t there

Quite so! The debate becomes, how many times, and how often, do we look and not find, before deciding we are confident enough that something isn’t there? If I can see a new shopping center on aerial imagery, that is one thing, but otherwise…

[jdmore]

@charlie

It still doesn’t seem ideal

Yes, for sure. But for someone who insists on using iNaturalist to track negative sightings, maybe a minimally disruptive work-around. At least until iNat either decides to find a better accommodation, or to lay down the law and say “no absence data here.”

[bouteloua]

@jdmore

If this were done in combination with downvoting the “Evidence of Organism” DQA, would that help, since Casual observations can be turned off on maps?

No, please don’t do this, since an observation currently means that you saw that organism then and there.