Suggestion: Allow hybrids that specify a genus only

Curators
#1

Previous related discussions here:

https://forum.inaturalist.org/t/does-inat-do-hybrids-that-only-have-a-genus/9701
https://forum.inaturalist.org/t/treating-cultivars-different-than-wild-taxa-to-have-the-option-to-exclude-them-from-searches/8158

I would like to be able to specify that an organism is a hybrid within a single genus without specifying the exact parental taxa. @bouteloua and I think this would be feasible within the current system (i.e. no changes to the code needed), but there should be an agreed-upon policy before we start doing this.

The example I gave previously is genus Callithrix. There are locations where you could reasonably expect to find C. jacchus, C. penicillata and C. aurita, plus hybrids of any combination of those. Right now, giving a genus level ID means a) I don’t know which species it is, b) the photo quality is too poor to identify which species it is, or c) it is a hybrid. A genus level hybrid taxon would rule out both a and b. (For those interested, this is a clear photo of an individual who is a hybrid – possibly of all three species, maybe just two of them?)

From a data-entry point of view, I imagine the taxon would look something like this:

1035×505 41 KB

I don’t see a lot of downsides to this, other than the fact that it complicates the taxonomic tree and may be challenging for some to use correctly.

The best alternative within iNat’s current functionality is to give a genus level ID and use an observation field to indicate hybrid status. This isn’t terrible, but some users disable the ability to add observation fields.

I don’t deal much outside mammals, so I would particularly appreciate input on how this might be good or bad for other parts of the tree.

5 Likes
Does iNat do hybrids that only have a genus?
How to specify idetifiable hybrids to differentiate from specimens not identifiable to species level
Treating cultivars different than wild taxa to have the option to exclude them from searches
#2

This would be helpful with Manzanitas, there are lots of one-off apparent hybrids between many different species out in the wild. Leaving them at just genus is suboptimal for the same reasons you specify and having to add a new taxon page for each one doesn’t seem like a great solution.

#3

I just checked and there is now a genushybrid category. This should hopefully address this problem.

#4

The Genushybrid category is for hybrid taxa of two different genera, not the same genus.

1 Like
#5

I see. Nevermind then.

#6

I’m not seeing much downside either, and maybe it would reduce demand/temptation for lots of individual hybrid taxon pages in genera where hybrids are rampant but whose exact parentage is often speculative at best.

Theoretically each one would come with a “not external” Taxon Framework Relationship deviation, which might or might not complicate @loarie’s life more.

Since any genus could potentially need such a hybrid bucket, I wonder if it would make sense to have a “hybrid” checkbox as a built-in property of any genus-level ID, rather than having to create an indefinite number of separate taxon records as possible hybrids show up (presumably in response to an equal number of taxon curation flags). But then that would definitely involve new coding instead of fitting in the existing system as you are suggesting.

6 Likes
#7

Just to clarify, the main upside is for searching purposes, right? To be able to filter by only these unknown/unnamed hybrids and exclude hybrids from genus searches?

If the hybrid genus taxon is “below” the rank of genus, it’ll still appear in typical searches for observations in that genus. And from a “Needs ID pool” management perspective it would still require people to know to mark the Community Taxon as not needing improvement in the DQA (or are you proposing it be treated the same as species rank in that regard?)

In either proposed system – separate taxa for hybrid bin genus taxa vs. a hybrid checkbox, would a “normal” genus ID ever count as a disagreement with a “hybrid” genus ID?

1 Like
#8

Good question. Ideally, if something is currently identified as a “hybrid bin” taxon, then

  • ID as another specific (“known”) hybrid should “refine” the hybrid bin ID (but I suppose simple disagreement is OK there too)
  • IDs as genus (non-hybrid) should have the same soft or hard disagreement options as any other higher-level IDs - it is this genus and could be a hybrid, or it is this genus but not a hybrid.

So thinking that through, a “hybrid bin” should either be given the existing hybrid rank, or should be a special rank right above the current hybrid level. If it was inserted as a separate rank just above the current hybrid rank, like:

…
complex
species
unknown hybrid (= our “hybrid bin” idea)
known hybrid (= the current “hybrid” rank)
subspecies
…

Then

  • Unknown hybrid IDs could have the same soft or hard disagree options relative to known hybrids as any other “higher-level” ID
  • Genus IDs (but also anything else species and above, hmmm…) would have hard or soft disagree options relative to unknown (and known) hybrids).

Now I’ll have to think more about whether hybrid rank(s) are more appropriately placed just below, or just above, species rank for disagreement purposes… :thinking:

1 Like
#9

Hm, I’m not sure I agree. It’s also nice to have a taxon page for these hybrids where you can get the photos all together, and potentially even build a wikipedia page to link to. It also helps to prevent other people from making erroneous species level IDs – if they see a hybrid ID already on the observation, I think they’re more likely to consider before adding a species ID.

I think I could be talked into going either way on the needs improvement checkbox. I kind of suspect those who are aware enough to use a genus hybrid ID would also be aware enough to use the checkbox, but I also think it would probably be ok to automatically go to RG.

In theory I think a genus (non-hybrid) ID should give the option to disagree with a hybrid bin. In practice, I wonder if it would actually be useful? To go back to my Callithrix example, if someone gives a Callithrix-hybrid ID, I can’t really imagine a scenario where I want to add genus Callithrix and disagree with the hybrid ID. The only way I know it’s not a hybrid is to be familiar with all the phenotypes of the species in the genus Callithrix, and if I’m familiar with what all the Callithrix species look like, I can just ID to the correct species.

2 Likes
#10

yeah I wasn’t advocating for the creation of these, just looking for more info on why someone would want them. thanks!

3 Likes
#12

This would also be valuable for plant people: named man-made hybrids/selections within a single genus are common e.g. https://www.inaturalist.org/observations/47624160. See also https://forum.inaturalist.org/t/feature-request-new-option-hybrid/12947/6. How do I vote in favour?

1 Like
#13

That links to a private message, which other users cannot see.

This topic is not a Feature Request, it is in the Curators category here on the forum. Thus, it cannot be voted on as only the Feature Request category allows voting.

#14

To me, a big upside would be to avoid incorrect species-level IDs via the AI Vision suggestions. As an example, there are lots of hybrid Coreopsis (=plant) cultivars that have a distinct look that I think Vision could master. But since there is no way to specify them below genus, Vision keeps naming them Coreopsis tinctoria or Coreopsis basalis, which are the closest-looking wild types. If there were a hybrid category, Vision could learn what it looks like and suggest the hybrid first, then presumeably mark is as cultivated. As it is, I’m just going through correcting the IDs back to genus all the time. So sorry if someone already brought up this point - I didn’t think it was here exactly.

3 Likes
#15

I tried to do this with cultivated hybrid Penstemons a while ago and got completely shot down. It would be nice to have hybrids “bin” for the cultivated things with untraceable parentage that is separate from the genus level wild organisms that can’t be identified to species.

If I’m identifying Penstemons and see an obvious garden specimen that comes from multiple parent species, I want to assign it to something. I don’t want to mark it as “Penstemon” and “No, it can’t be confirmed or improved” and leave a comment that it’s a hybrid. I don’t want to just leave it as “Penstemon” and mark it as reviewed because I won’t see it again but other identifiers will see it when doing the same search. As an identifier, I would like to weed out the cultivated things so other identifiers don’t have to see them too. We have enough to do without duplicating our efforts. Observers also seem to like having a more precise label on their observations.

My earlier attempt was reverted because “Misc Hybrids” wasn’t an official taxon supported by the taxon framework but I see these as really being outside of the taxon framework because they are artificial.

3 Likes
#16

It seems better to have a design improvement–for it not to even suggest particular species when it’s not very confident to a specific lower taxon.

1 Like
#17

That would be fine with me as well. And your suggestion would help with several other recurring problems, too.

#18

Although, since the hybrid flowers I’m talking about are readily identified as hybrids, it would be nice to have a place to put them instead of having them clogging up the “needs identification” channels forever.

2 Likes
#19

That’s confusing because it is listed below Genus in the “rank” listing. I would think a genushybrid that was the combination of more than one genus would be above “Genus”.

2 Likes
#20

In my head I think intergeneric hybrids are equal in rank to a genus, almost like they are one. I guess the use of mashed-up binomials gives me that idea.
https://www.inaturalist.org/taxa/349781---chitalpa-tashkentensis
https://www.inaturalist.org/taxa/516232--chiranthofremontia-lenzii

2 Likes
#21

If we think phylogenetically, the ancestors for genus hybrids would appear more recently than the ancestors of the component genera in a tree without the genus hybrids, so I guess that’s why they appear below genus.

1 Like