Warning banner on IDs for taxa that can't be IDed using photos

I was thinking that it may be beneficial for ID’s to include banners at the top (in the way that introduced/native species are listed) that stresses that a certain taxon cannot be identified using images alone. I don’t know how often users visit the actual taxa when making an identification vs. trusting the automod—but I feel that it may be effective as some deterrent that prevents people from making ID’s based on the images of other individuals that are there that may have been ID’d with dissection or other methods. For example, I work with leafhoppers (Cicadellidae) and there are numerous taxa that cannot be ID’d beyond a certain extent that causes some issues with frequent mis-ID’s based on a single observation that was ID’d using a separate method in tandem with the photograph (Gyponana, Scaphoideus, Osbornellus, etc.). Anyway, I thought I’d suggest this as it had been on my mind for a while.

In addition to this, as a much larger and more difficult change, I was wondering if a community wiki could be available for each taxa that could be edited to key-in users on how to ID their observation (maybe only accessible to editing by curators). I know wikipedia pages are linked to taxa, but I thought an iNat-specific community wiki would be more helpful. If something of this sort already exists, please direct me to it!

Welcome to the forum!

2 Likes

It was already discussed on the forum, probably not once.
From my point of view the term “photos” is complex, so any species that can be ided not only by DNA (or RNA) can be ided using photos.

4 Likes

I would encourage you to read and contribute to this forum topic:
https://forum.inaturalist.org/t/adding-useful-id-info-to-taxon-pages/657

and also read
https://forum.inaturalist.org/t/proper-way-to-store-and-organize-hints-about-identification-of-taxa/8962

How would you propose that we “teach” iNaturalist which taxa or groups should display that banner? And who gets to make those decisions?

3 Likes

How would that treat things that have been separately IDed after being photographed? Quite a few of the things I post are live photos of specimens I subsequently collected and then IDed with a microscope.

I agree with the above said, that the banner would not be a good idea in this case. In fungi, for example, some people add microphotos or photos of chemical reactions which supports and validates the ID of otherwise unindentifiable organism (from a usual photo). I will repeat myself (from the other post),that it would be much more useful, make the users happier and save time and patience of the IDers, if for some groups in larger taxonomical units, more complexes would be made. There are some large genera or families in which certain groups can be identified as a „group“,“aggregate“, „complex“, etc. which have no taxonomic circumscription or has lost it recently due to newest molecular findings, or are distinguishable only by chemistry or microanatomy, but which are distinct in macromorphology. As an example – users will be more happy getting an answer “Cladonia chlorophaea complex” than just Cladonia with an explanation that “one of Cladonia chlorophaea complex in which species can be identified by chemistry only”. Cladonia is a huge and morphologically very variable genus, so it is difficult to accept that your big and conspicuous lichen cannot have a name. And there are still lots of internet pictures and older references which indicate only C. chlorophaea without mentioning other species or treat them as C. chlorophaea s. lat. which is not possible to do on iNat. Probably there are similar examples in other organism groups, too.

7 Likes

As a UK recorder, this is one of the (few) downsides IMO in comparison with iRecord.

From an amateurs POV I would love to see this integrated, as it saves time and reminds me of things I often forget. As an amateur and pan-species lister its time-consuming to have to dig out the necessary information each time to delineate between species. Reminders such as on Naturespot in UK… where it tells you that it can only be reliably separated if you have a view of X Y or Z also help encourage you to capture that detail next time around.

Also when identifying others recently, I was often repetitively adding flags in comments that something was not identifiable to species and linking to reliable source… if automated surely this would save a lot of verifier time and questioning?

I think if its just a helpful flagging up also, then it can be easily ignored by those with the necessary microscopic photos ( if I am understanding the comments of @melodi_96 , @kmagnacca @jurga_li correctly which perhaps I am not :) )
…and it could also be switched off if people didnt wish to use it.

The complexes comment of @jurga_li sounds also really useful, though not sure I 100% understand why it couldn’t be in tandem with some sort of flagging / further info.

…In response to @jdmore I imagine iRecord have a process that must be replicatable. Presumably an offshoot of having verified experts and connected recording schemes though. But are there not other national lists like this? if not it would be an amazing resource to create and build/integrate a global one! Albeit time-consuming :) … could it not be crowdsourced like other things here?

3 Likes

I think the complexes would have to be another layer of taxonomy, essentially clusters of species that can be identified from a photo where getting it down to species level is next to impossible for most photos you can realistically expect to get. I think this would be useful but it’s easy to see it getting out of hand so it’s hard to see it being implemented here with the overly strict taxonomy control they use. I can’t even get them to add valid species that aren’t in a database, or to add databases to cover more insects.

I do not think it is so difficult to make complexes. It is just additional work for curators. One needs to collect all species names for the complex and some further mechanism I do not know. For example, you will see that there is such complex as Complex Rubus fruticosus. You will not even be able to ID it as species Rubus fruticosus.

Species complex taxa can be and have been created in iNaturalist. There are links and discussion ongoing on that topic here:
https://forum.inaturalist.org/t/large-genera-with-formal-species-groups/11073

This could be one limited way to implement @nomolosx request. If a species is included in an iNat complex, then that could be the trigger for a warning banner. And maybe Computer Vision could be restricted from suggesting anything finer than the Complex taxon, if any of its component species are among the suggestions.

7 Likes

thank you!

1 Like

Thank you for those links, they definitely have the same goal in mind. I think it would make sense if curators were able to activate this banner on a taxon. I’m sure this could present it’s own issues, such as if the curator’s information isn’t generally accepted or is wrong, so it would be good to be able to flag a banner like this so the community can have their own input.

I think the banner should be purely informative, so it wouldn’t prevent any observations from being associated with the taxon. It would act more so as a deterrent from users ID’ing a species based on “looks like.”

I agree with what you stetted about “group,” aggregate," and “complex” — this seems to be a much needed change and is one that many other websites utilise. My idea is that the banner wouldn’t prevent users from ID’ing to the taxon. It would instead act as a deterrent to users that would ID based purely on appearance and encourage them not to ID as specifically. One well-verified organism is attributed to its taxon and then other users see it/autoID suggests it and the user IDs to that taxon with insufficient information. Of course, the community can contest the observation, but I’ve seen some snowball effects that take hours to correct due to mis-ID’s.

Oh wow, I wasn’t aware they’d been implemented here! I like this idea a lot (having the banner triggered by the complex). I agree; I think limiting Computer Vision’s suggestions would also help achieve this goal.

You are very optimistic about banner as a deterrent :-) Some users are not deterred from IDing unindentifiable things even after showing where they are wrong. I just had an experience with one of them IDing green terrestrial microalgae to a species from a photo of a tree trunk covered with alge (shot from two meters or so).

2 Likes

Yeah, I think it’s that kind of issue specifically that is the problem. It may not be possible to solve it (I’ve seen so many threads of experts arguing with users on observations), but really anything to minimise it would be great. I think @jdmore suggested a more viable treatment than me (and if that’s the case, do I click the little “solution” button? I’m new here hahah).

2 Likes

I guess I’m curious how that would be regulated, given the resistance to using anything not found in a major database. Species groups and complexes are typically described in taxonomic monographs, and are not recorded in (for example) GBIF or ITIS because they don’t have a formal taxonomic category.

I will repeat myself again (from the other thread). It is based on purest human psychology. An OB can be named as a species complex by an expert IDer. Any user (and particularly less experienced ones) will be much happier having received more exact ID (a complex) than just genus name, especially in the cases of large and morphologically variable genera.

1 Like

The usual way on iNaturalist, when someone feels that we should deviate from the standard databases that guide iNat taxonomy, is to flag the appropriate taxon page for curation, provide supporting information in the flag comments (not in the flag reason, which should be brief), and @ tag other appropriate iNat users in the comments to initiate a discussion. Depending on how strong a consensus is achieved, deviations can be implemented, and fairly frequently are.

4 Likes