I’ve come across a lovely plant today, which iNaturalist recognises as Aceria echii. From the other pictures posted both on iNaturalist and on the web, I had no reason to doubt the algorithm given my utter lack of knowledge on plants.
When it came time to look for more information, scarce as it was, I noticed iNaturalist and most reliable sources (starting with French INPN) listed Aceria echii as a mite, which from the name taxonomically makes sense.
Is this specifically how Echium vulgare (which from everything I’ve seen seems to be the host plant) presents when infested with Aceria echii mites, or is this an example of citogenesis?
Is there a way to deal with these cases on iNaturalist, so that the algorithm doesn’t confuse laypeople like me?
Edit: Moved thread to General, seemed more appropriate thinking about it.
To be sure I understand the question, is the confusion over whether you can upload this as the mite?
I’m not familiar with this particular species of mites, but I am familiar from the family it’s from. These types of mites will form galls on plants. I’d say your photos would be probably be good for both the mite, and the echium, assuming that’s what’s causing it to grow funny, and could be two observations if you wanted. It doesn’t physically show the mite, but it is evidence of a gall former, and that’s fine. Similar to a deer track being acceptable evidence of an organism.
From a glance it does at least look like what other people are calling Aceria echii in their observations in Europe, but once again, I’ve never heard of it, no my comments here aren’t grace.
In general though as a way avoid confusion, perhaps a toggling feature could be developed in the future on the AI species suggestion feature to say whether you’re looking for a host ID or a pathogen ID, that way instead of a mixture of plants, fungi, and arthropods, it would just show you similar plants so the drop down lists aren’t all over the place. Or the other way around if you’re more after identifications for fungi or arthropods.
For now, the algorithm will probably keep suggesting them if it’s a common gall former. May become more of a common issue as more galls and rusts get photos uploaded and those species become part of the computer vision model. Many only have a handful of uploads and wouldn’t cause confusion for you now, but could someday in the future when there are 100+ uploads of it.
I know nothing about this viper’s bugloss gall, but it seems that iNat’s computer vision is operating as designed.
I think your photo likely does show Echium vulgare with an infestation of Aceria echii mites. I’m guessing the mites trigger that bushy growth. If so, then iNat is accurately identifying the parasite based on previous photos uploaded to iNat. Unsurprisingly, it’s not doing such a good job of identifying the host plant because most Echium vulgare photos are of plants without galls and look quite different.
As an iNat user, it’s best if you treat the computer vision taxa as just suggestions and still use your own judgment as to whether any of them is a good fit. In this case, seeing that iNat was suggesting a gall mite might have been a clue that you were looking at a deformed plant.
The computer vision doesn’t know what you’re trying to identify unless you tightly crop. It will readily identify creosote plants as creosote galls and vice versa, because the training sets are essentially identical.
I’d argue it’s both and a good opportunity to duplicate the observation and mark one as the plant and the other as the gall forming creature. For the gall, add a link or note the name of the host plant, it could be useful to gall identification.
is the confusion over whether you can upload this as the mite?
Most of my confusion stemmed from the UI, to be honest. Seeing a picture of plants among other plants (Artemisia, mostly, but not Echium vulgare) in the recommendations and finding very little info online, I was puzzled by later realising the ID was an arachnid. The only cases of gall mites I’ve seen (as far as I know, I guess!) where easily identifiable as parasites on an otherwise recognisable plant, not something quite as transformative.
As you and vreinkymov have suggested, I will duplicate the observation and take this as a lesson on due diligence!
Do you have your display setting set to show both common and scientific names? The common name for Aceria echii is Blueweed Gall Mite which could clue one in to the fact that it is a mite and not a plant.