Should New Disruptive Technologies be Used for Classification in Ancient Linnaean Rankings

And the same would be true for the taxonomic observation fields. iNat, in the end, could agree that the prompting is needed for say… Taraxacum and not needed for say, “Poecile”… since there is risk of the feather thing that Lothlin mentioned. If the relationship is close with the taxonomic meta data supplier… it wouldn’t be a big deal. Not even a need to add a flag at each rank to flip off the prompting.

The feather thing is interesting tho. Because it might show how new disruptive technology shouldn’t always be used in ancient linnaean rankings… at least if we’re going to split Chickadees at species.

Surely if the taxonomists are going to split species they can define the meta data required to prompt data collection for the new species. I imagine you’d agree with this Charlie. No? I mean, it kinda has to be them.

1 Like

well, yeah.

Just to clarify, I was using the two chickadee species as a case example of an exception to a commonly assumed guideline. There isn’t a move to lump them into a single species as they are geographically isolated (except at the hybrid zone where contact is still relatively limited because of habitat/elevation factors as ddubois2 mentioned), and they have subtle but distinguishable field markers that don’t require microscopy, dissection, or sequencing.

2 Likes

Right. And so what prompting would be required for an iNat observer to gather data to justify their identification at species?

A clear photo from an angle that shows the wings and/or nape. Or clear audio (although audio isn’t foolproof where they overlap because their song isn’t innate and they do sometimes learn the “wrong” song).

Ok… so maybe we could rely on observation only for that one. And I imagine a taxonomist (supported by a bird biologist) would know this… and not add any meta data for those taxa for prompting. The risk is… lots of data with mis-identified chickadee species. Because the prompting might convince someone who flipped to gold to switch back to the default of green (realizing they don’t have clear audio or the right image).

The point is… it is the taxonomist (in consultation with experts on the taxa) who determine what traits signify rank. And so that data must be part of the taxonomic meta data. (at least where RG iNat data is concerned). But it only needs to be clearly defined via meta data if it isn’t obvious to a casual observer.

So like, it’s ok if no meta data is defined for Chickadees, but it might result in worse data.

In reality… many here know the worst ones (Taraxacum)… I’m not sure Chickadees justifies it. But the framework, once in place… would allow for others.

Or the innumerable number of recreational birders.

This isn’t exactly true. It’s kind of a chicken-egg situation. Generally there is a presumed difference first, which is then tested to verify that the presumed field markers are actually restricted to the respective species. Also, a difference in field markers alone doesn’t make a species.

Doing my best. I imagine Charlie could come up with wayyyy better examples than Chickadees.

Part of what we’d be trying to do with the prompting… is cause the observer to realize they aren’t qualified to identify that taxa to species. And it’d be more apt to go to RG green instead of gold.

I think what you are describing sounds basically like some combination of observation fields and projects and/or a third-party app built on top of the inaturalist platform/using the naturalist API. The platform itself is pretty mature technologically, supports a lot of use cases, and doesn’t really have the resources or inclination to implement taxon-specific ID solutions. There is some discussion of various third-party apps here: https://forum.inaturalist.org/t/3rd-party-inaturalist-tools/25971/13

The other thing you are talking about it encouraging users to take better pictures or relevant features. To an extent, that is actually baked into the computer vision apps like Seek; they are meant to encourage decent photos in part by providing more a useful ID when the photo is decent. That feature could certainly be improved further, and playing with iterations of that is possibly another good opportunity for third-party app development. And of course, communication with IDers can help users take better pictures.

The site also supports copy-pasting an entire DNA barcode into an observation field. That feature is mostly used in fungi right now. I don’t think the site would ever allow it to directly affect the RG status-tier of an observation because it is not something that can be checked with citizen science.

What external projects can/have done when they want a bunch of physical samples or DNA barcodes or whatever is to organize some kind of event in which the project organizers provide citizen scientists with a shipping address/label and sample collection/preparation instructions (or even a sample collection kit!) to send samples to in exchange for having their name on the eventual paper or whatever. If a project really wants those DNA analyses done they will have to pay for them themselves though; that isn’t a reasonable thing to expect citizen scientists to do. Obviously, care must also be taken to ensure that whatever sample collection and shipping you are asking citizen scientists to do is legal, ethical, and ecologically sound in their jurisdiction.

This kind of thing has actually already been done to an extent for taraxacum papers in North America: for example the data in Genome size variation among common dandelion accessions informs their mode of reproduction and suggests the absence of sexual diploids in North America was based on dandelion samples collected and sent in by volunteers in North America (though not organized through inat, as far as I know). I could imagine a larger or more detailed physical specimen collection project like that in combination with a set of associated observations posted to inat could be an interesting paper for someone to write. And I’m sure you could recruit a decent amount of volunteers even just from these forum threads.

4 Likes

yeah i posted some examples in the dandelion thread that got unreasonably locked because it was ‘off topic’ to talk about taxonomy in a taxonomy thread.

1 Like

Thank you for such a thoughtful response. What I can’t know is how many other taxa are like Taraxacum. If there are several, species will lose its “real world” meaning unless taxonomists give it meaning by sharing with iNat, the traits that define each species uniquely. These same traits could then be used to better prompt data collection… or not.

The other option is to negatively influence the proliferation of multiple species indistinguishable by the average iNat user.

As many have pointed out… there are consequences for conservation and preservation.

Either way, I feel like I had the opportunity to present what I was able to come up with. I’ll add it to my long list of “tried but failed”. It’s a complex thing. I won’t lose sleep. :o)

1 Like

I think the logic was that it wasn’t specifically dandelion taxonomy. But yes, this supports my contention that the “off topic” rationale is used very unevenly and is influenced by whether someone likes what is being said. Some posts are allowed to go further “off topic” than others.

And as conservationists/preservationists are increasingly pointing out, there is danger in what is termed “silo” thinking. If taxonomy and field ecology are separate silos, what are the consequences for ecosystems and species?

3 Likes

Right. The subject was discussing OTHER species where i feel there has been too much species-level splitting and it has been problematic, so that examples could be discussed. I suppose the initial post wasn’t about dandelion taxonomy exclusively, but the posts had been about taxonomy for literally days, and so it seemed pretty weird to me. But whatever. It’s not my forum and really they can do whatever they want with it, including silence me until i decide to leave (again), which may happen. But probably someone will deem this off topic too and close this post so…

It’s 2050. Children are playing outside again. They have their iNat glasses on. One of them focuses on a dandelion. Do we want the prompting Id to bounce around on 2,000 different Taraxacum species? Or do we want the glasses, recognizing that the child is at the green (default) level to present the child with “Taraxacum (dandelion)”. I mean, it seems plausible that some children might evolve via exploration.

Sometimes, a child’s first view is Linnaean. But some of these children, crawling on hands and knees, might look closer. Upon closer inspection, when that child blinks their right eye, would we want their view to become gold? Again, it seems plausible that some children might evolve via exploration.

I completely agree with this point. What I think you might be hinting at is that the more mature technology sometimes loses some mind share among my people… the developers. They’re constantly looking at the next thing. This is unfortunate because sometimes the mature technology can present exactly the use cases that the new technology should be focused upon.

You go on to mention, in no particular order… computer vision, seek, features that could be improved further, and playing with iterations of what is possibly another good opportunity for third-party app development. @DianaStuder mentioned computer vision up above somewhere and I had started looking through it a little… but wasn’t ready to go there at the time.

Computer vision must have access to similar taxonomic data as iNat. I don’t know where or how that taxonomic data is created, stored, and updated. But it is theoretically possible to add a “role” or at least a boolean at each level telling an app to bundle or un-bundle based on a taxonomists understanding of both the child’s view and the modern taxonomists.

Every named species is associated with a published description which does exactly that. Well, the authors don’t send it directly to iNat–but iNat curators base the iNat taxonomy directly (or indirectly) on those published descriptions. Nothing is being hidden or not provided. I’m still finding it very confusing as to what exactly your question or concern actually is.

Information on the iNat taxonomy framework can be found here: https://www.inaturalist.org/pages/taxon_frameworks

5 Likes

It’s a significantly more sophisticated system than just making a description to justify a name, all valid names (in botany not sure about elsewhere) also have a type specimen associated with them. That means if anyone is unsure of what so and so meant by their description they can reference back to a physical specimen of what that name represents. Maintaining type collections is one of the really powerful and sophisticated technologies (although generally hidden) in taxonomy. While there is always dispute over how much of a taxon a name describes we can always be certain that a name describes the specific type individual. That means new technologies can be applied to very old names is a standardized way.

It also gives names flexibility in handling splits, while maintaining a meaning. If a species is split the new species concept that includes the type keeps the old name and the new species gets a new type.

I’ve heard it described as a name fundamentally only applies with 100% certainty to the type, and it is taxonomists job to make and support hypotheses that that name also applies to populations and groups that are not that single individual.

Taxonomy is not static, and plenty of people have made plenty of bad hypotheses of how to apply names over the centuries. Without a physical concrete reference of what a name really refers to taxonomy would be complete anarchy, and there would be a lot of really pointless disputes and redundant names all over the place, much moreso than there already are.

3 Likes

Then they shut down the herbarium.
https://forum.inaturalist.org/t/duke-herbarium-closure/48997

3 Likes