I’m fine with the wikipedia section and I think it’s nice to have a small summary of info about a species. I think the taxonomy section is really useful too.
With that being said though, I wish there was another section describing characteristics of a species to help with identification. Sometimes the wikipedia section does this, but only occasionally. this hypothetical section could include field marks to help differentiate between similar species. For example, it could note whether a specific plant species has a fuzzy stem or not, or whether a bird species has stripes on its undertail. It could also note other basic identifying traits like size, shape, color, etc.
This would be really useful because it can be hard to find identification info online for some species, especially plants. I know there’s field guides, but I don’t really want to pull out a whole field guide to add an ID to someone’s inat observation. I also don’t own a field guide for every single species in the world. I love inat, but it’s not as useful of a learning tool as it could be. You really just have to rely on your own knowledge or the AI tool to make identifications, and the AI tool doesn’t tell you its reasoning for suggesting a certain species.
What are your thoughts on this? Is it something you’d like to see? Is there something else you would change about inat’s species descriptions? Are you fine with the descriptions the way they are currently?
edit: I’m a newer user to the forum, so it’s not letting me add any more replies to this for 24 hours to prevent spam. I really appreciate all the conversation going on here and I hope it continues. I am reading all comments, just not able to reply yet!
I would like to see this, although I could see it being difficult to implement because of the number of species to write descriptions for. I wonder if it could be implemented so the community could create descriptions, so you could submit a comment and then have others check it.
I usually do get my field guides out when IDing to species, because I filter it to my geographic region so I can practice my field ID skills, but that is a fairly unusual use case.
Based on the latest staff updates, it seems like the developers are looking to just use AI to collect the best ID comments and display those rather than generating new AI text (although they may still be considering using genAI for photo-taking tips). The wiki idea seems like it might be a no-go with staff for reasons that are unclear, perhaps because of the extent of moderation that would be required.
yeah, there’s definitely an insane amount of species, so I would be fine with a slower rollout where they maybe start with more common/popular species and work from there. community descriptions could work as well, or maybe enlisting trusted sources like university professors, researchers etc to do the descriptions, but that would take money of course.
using a field guide to identify iNat observations is a good idea. I’ve been looking for ways to learn plant and insect ID during the colder months when they’re not around as much (besides studying a field guide for hours) and that seems like a good solution!
Yes, this would be helpful. I also wish that there was some kind of guide to list the time of year you need to observe species to be able to identify them. For example I know that some Hawthorn species can only be identified when they are in flower (April thru June-ish in Midwest).
I fear this would be extremely difficult to implement on a global scale, due to the extremely large number of species as well as the huge amount of variability that plants are capable of; a reason why most (if not all) identification guides and keying materials are created with a country or smaller region in mind – and even so they manage to miss a few things. Even the ‘AI’ (‘CV’, image recognition), which has no duty to provide a “reasoning” whatsoever, can fail tremendously whenever geographical input is missing.
Compiling “ID tips” and ordering them in reliable, workable ways (assuming it is possible) could end up as a mere rehashing/rephrasing of existing guides, with the addition of endless debates between taxonomists, lab or field experts, different “national traditions” etc. Is it really worth it?
I’ve said this before, but I think a taxonomic-based forum would be great. Like, imagine there is a page for every species and people can leave comments discussing how to ID certain taxa, curators could star the best comments, etc. I’ve had to pick up many specific tips for certain species just based on comments on years-old observations I’ve stumbled upon. I think it would make the learning curve for new ID’ers much easier and do a great deal with helping to increase the number and accuracy of research grade observations.
that’s unfortunate, the wiki thing could be a really good solution to this. I mean technically a wiki could still be created without connecting it to iNat in any way though. just call it “species wiki” or something like that, it doesn’t necessarily have to be the “iNat wiki” or anything
BugGuide managed to do this with a staff of 1, so I’m sure iNaturalist with it’s army of volunteers could manage. I’m disappointed iNat staff has thus far refused to engage with this idea, even to shoot it down. Maybe they just have too many other projects to deal with at the moment.
I agree. The more information identifiers and observers have access to, the better. I don’t think that iNaturalist should duplicate Wikipedia, however. That said, the iNaturalist journal is a good way to record identification tips.
I disagree with this. I use the algorithm and what little I know to make identifications but that is far from the end of it. I have three hardbound flora on the table next to me plus access to numerous reliable online sources. I can’t imagine being an identifier (at scale) without having access to this expert knowledge. We (as identifiers) should strive to transfer the recorded knowledge of experts into iNaturalist. Every time we make an identification based on some reliable source (online or offline), we do just that.
Well, there is already a wiki for each species : that’s called Wikipedia. There you can write how to distinguish one species from the other (for the plant taxa it’s a suggested section in the guideline for plant articles).
There already are tons of website, more or less specialized, with or without wiki sections. I doubt make a new wiki tool will change the game. It’ll only dissolve efforts.
And the global scale of iNat won’t make it possible in a foreseeable future to hope for a data base to be build from scratch. What you describe is something every naturalist would dream of, but it simply to big of a workload to begin with. That’s the reason Wikipedia articles are often too superficial. But I still think it’s the best tool with have now to freely share knowledge about species.
So go add your input in the wikipedia articles of species you’re confident with !
Today I just released an article about Ranunculus sect. Batrachium (in french).
Yeah that is an option. There’s a discussion about that here but no one has committed yet to try and actually create something: Where do we share how to ID species?
It’s very common outside of the world of birds and mammals. I do a lot of insects, and wiki pages and info is sparse. I’m working on creating Wikipedia pages for insects, but it’s slow going and way more are sans info than have pages.
As it turns out, your example illustrates why what you’re asking for may be unrealistic (at least somewhat). In Vermont, the red maple is notoriously difficult to ID from photos. It is very similar to the sugar maple (which is the Vermont State Tree). To make matters worse, the two species have a named hybrid.
For the red maples, I usually add an ID at the section level and move on. It doesn’t matter how many resources I consult, the red maple remains difficult to ID. There’s not much that can be done about that.
There are many species that are similarly difficult to ID, too many to list here. Moreover, some species are difficult to ID in one locale and simple to ID in another locale. A good example is the plant genus Tiarella. If you happen to live in New England, there’s just one species, and therefore a species-level ID is fairly straightforward. In the southeastern United States, however, there are multiple look-alike species. It’s almost impossible to distinguish them from photos.