New Annotation: Evidence of Presence

We just introduced a new type of annotation called Evidence of Presence to iNaturalist, and originated from suggestions by members of the iNaturalist community.

To take a step back, remember that annotations are iNaturalist-controlled terms which are used to add metadata to observations and provide benefits to other parts of the site (e.g. taxon pages, collection projects).

An iNaturalist observation records an encounter with an organism or recent evidence of an organism, so the Evidence of Presence annotation provides a way for users to indicate what kind of evidence is depicted in an observation’s photos. We hope it will be especially useful for taxa where the organism itself is not often seen. We attempted to find existing, defined values for these types of data (even consulting the folks at GBIF) but were unable to do so, which is a) why it’s taken a long time to implement this and b) why we’re only starting with a few values.

The criteria we used to decide which types of evidence to include for the rollout:

  • Will most people understand what the term means?
  • Is it commonly observed and relevant to a broad group of taxa?
  • Will it noticeably improve parts of iNaturalist such as the taxon photo browser, searches, or collection projects?

Evidence of Presence annotations will only appear for observations within Kingdom Animalia (except for humans), and there are six possible values at launch, defined as follows:

Organism: Whole or partial organism.
Scat: Fecal matter (not owl pellets or other regurgitated matter).
Track: Impression in ground or snow made by an organism.
Feather (within Aves only): One or more feathers not attached to an organism.
Molt (within Arthropoda and Reptilia only): Discarded skin or exoskeleton.
Bone (within Vertebrata only): Predominantly endoskeletal remains. Partial bone exposure in an otherwise intact organism should be labeled “organism”.

I did a little annotating on our test server to make some examples of how this annotation can be used in the taxon page photo browser. For example here are photos from Great Horned Owl observations that have been annotated as Feather:

And photos from Sidewinder observations annotated as Track:

Finally, some FAQs:

Why only Animalia?

We have nothing against plants, I promise! Unlike most other organisms, animals leave behind traces of their passing, which are commonly the only evidence of these organisms found by naturalists. Plants do drop leaves and fruit, but those are far less commonly observed than the actual plants. This seemed most beneficial for Animalia, at least at this time.

What about “shells” for mollusks?

I actually tried this out on our test server but found that it was of little use. Most observations of shelled mollusks, whether dead or alive, already include photos of the shell so it didn’t do much for separating out the types of evidence one might come across. Probably best to use the Alive or Dead annotation for shell finds.

What about [insert animal-related value here]?

There are certainly other values we’re considering for addition in the future, but again we wanted to start this off with the most commonly seen, useful, and least ambiguous values we could. Some commonly seen and useful values like “domicile” or “construction” seemed too complex for now (do we make different values for a burrow and a bird’s nest and a bee hive, or just use a blanket term? How does this translate to non-English languages?) and others a bit too uncommon or difficult to define for this first phase (eg bear rubs). For more niche or specific data needs, observation fields would be a better choice.

Do any existing observation fields map to this annotation?

Yes, if you use the Animal Sign and Song observation field, some values like scat, tracks, and fur/feathers map to the corresponding Evidence of Presence values, so if you add them to an observation they’ll be annotated as well. In the background we’ll be retroactively annotating observations that, before this annotation was launched, already have the relevant observation field values. A few other observation fields that map to the Evidence of Presence annotation are Tracks, Scat/Excreta, Scat?, and Bone(s).

If there are commonly used non-English observation fields that would be relevant, please let us know.

What if none of the terms fit?

Just like existing annotations, choose the best value or, if none of the values are a good match, don’t annotate the observation.

71 Likes

Nice.

Here in Calgary, the most popular “Evidence of Presence” would be beavers. Most observations are of tree cuttings, lodges, and dams. The closest annotation with the new additions would likely be “track” but it still doesn’t quite fit. I look forward to seeing this progress over time.

Happy to see this new annotation! Thank you!

11 Likes

Galls, nests, burrows and similar structures are also evidence of presence, I think they should be added to the list.
It should also be possible to add multiple, for example in this observation https://www.inaturalist.org/observations/76912147 I photographed both the organism and its exuvia.

12 Likes

how do we proceed with eggs, in some elasmobranch. reptils and birds?

1 Like

This is perhaps true in general, although there are a decent number of live bivalve observations with no view of the shell. I agree that there probably isn’t an easy annotation option for that, and we do have a project so I guess that has it covered.
https://www.inaturalist.org/projects/living-bivalves

I would assume that counts as organism. Unless you’re thinking it’s evidence of the adult that laid it.
Screen Shot 2021-06-14 at 4.07.13 PM

2 Likes

I would like to batch edit my observations to add the new annotations. I could not figure out how to do that so far. Is it possible?

Leaf mine” would also be a useful annotation value for winged and once-winged insects.

13 Likes

Personally my take is to leave the “Alive or Dead” field blank, as I consider that to apply to what’s pictured, and the molt isn’t alive.

7 Likes

Does scat include frass? Or does it have to be…bigger poop?

1 Like

It’s a just a bit confusing to me about the first item, whole or partial organism.

A feather or bone are separate data values, but fur is not? I’m afraid, I miss the rational for that distinction?

1 Like

Can we get an “other” for ‘evidence of feeding’, ‘burrow’, ‘nest’, etc?

6 Likes

Will the ability to add multiple values be added? For example, an observation of an owl that has both pictures of the organism itself and a feather it dropped, which would be both ‘organism’ and ‘feather.’

1 Like

Tony described it as just “fecal matter” so I suppose it would include frass.

2 Likes

Thanks for all the feedback, everyone.

With the caveat being that I’m not an entomologist, I think molt would be the only necessary one here. The evidence the observation records is not of a dragonfly nymph, just the molt. And you don’t know whether or not it was alive when you found the molt, so I’d either leave the Alive or Dead annotation blank or put in Cannot be determined.

Yeah, like I said there are plenty of more specific types of evidence we left out on purpose for now, but I’d like to have ones for things like beaver chews, dear/bear rubs, and evidence of feeding, etc. Again, we wanted to start simple and broad.

As noted in the original post, trying to figure the best term for constructions and homes was a bit too complex for this at launch. As for galls, I’d be OK with it (would need a list of applicable taxa) but since most obserations of gall-inducing organisms are photos of the galls and not the organism itself, it seemed less pressing. But I think annotating observations that show photos of the gall-inducer as Organism would be helpful for sure, so people could find photos of the organism among the sea of gall photos.

You can’t really batch edit them, but you can use Identify to annotate your observations fairly quickly (although there are some performance issues which I want to look into later this week).

Ooh, good suggestion!

Any type of fecal matter.

I think it’s best to have specific values than a catch-all.

Yup, I forgot to make this annotation multi-valued. Should be working now.

13 Likes

Since we annotate observations and not individual photographs, though, a side-effect of this will be that both the organism and the feather images would show up when “feather” is selected in the photo browser. Not sure how much of a concern that is to folks, just something to be aware of.

10 Likes

You would just search for “adult” for insect gall-inducers. If I had a gall that I cut open and it showed the larva I would count that as “organism”.

3 Likes

These annotations are meant to be used when the organism is not available for you to observe

Bone: Predominantly endoskeletal remains; partial bone exposure in an otherwise
intact organism should be labelled organism

Feather: One or more feathers not attached to an organism

Fur would be a good one to add for mammals. You can find tuft of hair caught in vegetation or that has been pulled out by a predator (who took the rest of the corpse away)

6 Likes

Cool! Updated some of my annotations.

An excellent addition! A few pointers from me:

Moulting occurs in more than just reptiles and arthropods, and it would be good if it could be expanded to include those groups too (even if finding a moult from them is less likely than from reptiles and arthropods). Principally, everything in Ecdysozoa moults as far as I’m aware - on iNat this corresponds to the phyla Kinorhyncha, Nematoda, Nematomorpha, Loricifera, Onychophora, Priapulida, and Tardigrada - and several amphibians moult as well so it could be useful to expand it to them as well.

Would we be able to have somewhere to put a concise list of future additions to this annotation? I feel like many of the suggestions will get lost here and people will start repeating themselves.

I agree with not adding in an ‘other’ category, at least for now, but is it possible to search for images/sightings that have not been annotated? I get that this probably isn’t all that helpful because lots of people do not annotate their sightings at all, but it could be useful to at least remove photos that are definitely not what you’re looking for if you’re looking for something that isn’t an option yet.

6 Likes