Treating cultivars different than wild taxa to have the option to exclude them from searches

Some wild species have cultivars that go by the same name. Take for instance Portulaca umbraticola which is a native plant of North and South America and its cultivars like “Wildfire Mix” or Euphorbia graminea which is a wild plant that is spreading through cultivated plants and its cultivars like “Diamond Ice”. The cultivars are readily distinguishable from their wild counterparts, but iNaturalist is not great at distinguishing them. One could theoretically create a cultivar name for every cultivar that comes out, but this is limited by: (1) new cultivars coming out continuously and (2) requiring an identification to cultivar which can be difficult.

In the case of Portulaca umbraticola, in particular, many of the observations cannot be verified to be cultivated. This has distorted the map to such an extent that it is difficult to tell what the wild range of the organism actually is or even the observations of the wild organisms themselves.

Potential solutions:

  1. Adding new annotations or data quality fields to indicate the organism itself is of cultivated origin.
  2. Adding a general “cultivars” taxon always labeled “cultivars” (e.g., Portulaca umbraticola cultivars) as a bin to add all the cultivated taxa to that could have individual cultivars added under it if desireable.
  3. Decide to mark these as cultivated regardless of whether it spread without the help of people (I personally don’t like this option at all).

I strongly prefer number 2 as I think it is more consistent and useful to put it in as part of the taxonomic framework, though it would have to act or be treated differently from the way we usually use taxa. Instead of specifying what taxon it is, it would just be labeled “cultivars”. I suppose we could do the same thing now using the subspecies or hybrid taxa, but I don’t really like that option as I still don’t think it treats the cultivated organisms sufficiently differently and seems difficult to standardize.

Added 25 Nov 2019:
If a cultivar taxon were created for every species that possessed them, excluding them from searches would be as easy as typing in “&without_taxon_id=” into the URL. Perhaps the developers could even add an option to exclude escaped cultivars by checking a box.


@nathantaylor just a note: since there didn’t seem to be a specific actionable request here yet, I approved this for the General category to start with. If further discussion results in a clearer solution, then that could be submitted as a feature request with reference back to this discussion.


Having a cultivars “bin” taxon under the species seems reasonable to me, it’s similar to the “complex” used for some multi-species taxa. It would still take a little URL manipulation to find e.g. observations of Portulaca umbraticola that are not identified as cultivars, but not a big deal for someone who really wants to.

In these cases what I sometime do is mark as cultivated, but leave a stock message along the lines of:

Cultivated plants should be marked as captive/cultivated. I suspect this observation is planted, but am not totally sure. I’ve marked as cultivated, let me know if I should change.

Would work on it so it is nicer/politer, but you get the idea.


I like the cultivars “bin” idea too. Even if all the cultivated plants were correctly marked I think it would still be useful to be able to easily distinguish observations of the wild form growing wild from cultivars growing wild.

I like your wording. I think I prefer it to the wording on the frequently used response page.


It’s not quite the same situation. That text is more for where things are obviously cultivated, not where it is unclear.

Here’s a slightly nicer version of what I wrote above:

Please mark planted/cultivated plants as captive/cultivated. This option is available as a checkbox on the app, or in the “Data Quality Assesment” at the bottom of the observation page on the web.
I suspect this observation is planted as [explanation], but it isn’t clear from the photo. I’ve marked it as cultivated, but let me know if this was actually found growing in an area where it likely wasn’t planted, and I can change my mark to “wild”.


Another reason why I think it is important to use a unique taxon instead of the other two solutions I proposed is that, if 80% of the observations in a small place are marked as cultivated, iNaturalist automatically marks that species within that location as cultivated (at least, that’s what I’ve been hearing). As iNaturalist grows or as new trends come and go, I can think of instances where this could cause observations to be incorrectly labeled automatically. For example, the wild P. umbraticola is found in the Dallas/Fort Worth area. Though the cultivated form of P. umbraticola doesn’t seem to be commonly planted there now and not too commonly observed, it doesn’t take a lot for that to change. All you need is for the plant to start trending and a large influx of inexperienced users (the latter is fairly normal across iNaturalist). Although the scenario is somewhat unlikely (at least with this example), I’d really prefer to account for it so that it doesn’t become an issue in the future.


Yes, that’s correct. iNaturalist gets a vote just like a user. That means that species would default to cultivated unless I flag as not cultivated when I upload, so it’s not a big problem for any experienced user who knows how the system works.

It does also mean that the cultivar taxon would probably get flagged as cultivated by default in most areas.


The problem is simply that iNaturalist uses the “species concept”. Cultivars are not a different species and thus should not be treated any differently from any other variation within a species (which in some species can be considerable). We thus have a little problem.

In addition to cultivars being selected from within a species, many cultivars are hybrids. In this case, it is not valid to pretend that they are a species, when in fact they may be hybrids between two or more species. {{of course we can get around this by making the [cultivar or hybrid cultivar] a child of the genus an thus your “Wildfire Mix” will be found under Euphorbia “cultivar” rather than Euphorbia graminea “cultivar”, but this then bins lots of very different cultivars into one black box, and treats all cultivars as equivalent to a species. }}

You seem to be muddling two issues here. 1. The one is the use of cultivars, and how to do it.
2. The other is if the organisms are planted/captive or if they are running wild. In the majority of cases cultivars that run wild rapidly hybridize with other cultivars or escaped “species” or wild individuals and very rapidly lose their identity as cultivars. Even where it is easy to identify cultivars, once they start intebreeding it becomes a nightmare and the cultivar concept usually becomes meaningless. In the few cases where cultivars are becoming a “pure” problem, can this not simply be done in the description and notes? Do we need it to be done in the identification per se, given that it is the same species?
Might I suggest that were the Notes for the ID is not enough, we rather resourt to an Observation Field, and put the entry there?

Note that this is not similar to the complex “level” - a complex consists of several species that cannot easily be told apart in some situations. The children of this complex are the constituent species.
For cultivars, they are children of the species (or a hybrid between species), and iNat does not include the different cultivars because they are biologically identical, even though humans artificially breed them by preventing crossing with other members of the same species, and removing intermediates. Without this constant attention they would just melt away within their species (or if sterile, would just die out).

Why not just identify it as the species, and in the ID notes mention the cultivar? Alternatively, use an Observation Field.

We have a similar but different issue with “Frankenflora”: this is were cultivars or subspecies or species are moved around in gardens, urban areas or restoration areas, and then hybridize with related wild plants in the adjacent natural areas. They are clearly hybrids (or “hybrids” of subspecies which would never normally meet), and are a conservation issue as they spread at the expense of the true species, by potentially hybridizing it out of existence. But fortunately the hybrid option on iNaturalist has coped with all recorded instances of this so far (although it is true that some of these hybrids contain several horticultural cultivars, which we simply treat as a hybrid, listing the cultivar names in the ID notes where they can be determined, or where it is important).

There are garden sites which deal specifically with cultivars. People who want to document or display cultivars should rather use them?
Unlike animal and plant scientific names, cultivar names are under the control of dozens of different groups, and identifying cultivars with certainty is often a far more complex task than species identifications (which are difficult enough).

The 80% rule is easy enough to overrule, even though when it first catches one out, it creates quite some bewilderment …


If I understand your comment correctly, I actually don’t care much about “1.” at all except in the context of “2.” I don’t care about cultivars except when they make it difficult to interpret ranges of native populations. I am not muddling the two but am conflating them simply because I don’t care about the problem when they are separate. Reason: if they are not separate, they can simply be marked as cultivated and we don’t have to deal with them as if they are wild organisms.

Well, not quite. iNaturalist accepts varieties and subspecies which in function are similar to cultivars. In principle, these freely interbreed between populations of the same species and if translocated, can respond similarly to what you are describing in cultivars. I know some biologists have decided to not accept varieties nor subspecies, and I don’t know if you are one of them. If you are, please understand that I do accept varieties and subspecies and think they are useful. I am not the only one. Many iNaters also use them and they are allowed in the taxon framework.

  1. This assumes they will interbreed which is not always true and/or 2. This assumes it is not worthwhile to maintain distinctions when hybridization is possible. Using the logic of “2.”, it would be useless to have subspecies and varieties at all, an opinion that I am opposed to.

The answer to this is literally in my original post.

This cannot be resolved by just adding it in the description and adding observation fields does not fix the map.

Not exactly a true statement. Just because two morphotypes can hybridize, even to the point of complete intergradation, does not mean they are biologically identical. Cultivars are not species and do not adhere to the biological species concept, but this does not at all mean that they are biologically identical to wild organisms. Euphorbia polycarpa var. polycarpa is not biologically identical to Euphorbia polycarpa var. carmenensis, though they do theoretically have the ability to hybridize if they contact each other and appear to intergrade. Depending on your taxonomic philosophy, this is essentially what is meant by a variety.

Again, I don’t care about cultivars except when they make it hard to see or distinguish native forms. I don’t care about specifically what cultivar it is (at least, in the context of iNaturalist) just that it is a cultivar and not a wild form. This site is primarily designed for native organisms and the goal of what I’m proposing is simply to make it easier to see the native populations through the cultivars.

In my opinion, the “Frankenflora” dilema actually supports my idea that we should have separate categories for cultivars. This way, we can more efficiently monitor the truely wild forms and compare them to the cultivated or “blended”/hybrid forms by having a standard bin that we can use to compare and more efficiently monitor the threat that cultivars of native wild species pose.

That’s only if the area is well-curated. Some regions are not so lucky. Keep in mind that I am trying to think up a system that requires less curation and not more. The solutions you have included involve specialized fields or descriptions to work (and descriptions are much more difficult to use when it comes to curating data). Most members of the iNat community are not aware of the specialized fields but are quite familiar with how identifications work. This requires an active curator to be adding observation fields constantly while if this were part of the taxon framework, there is a greater sense that this is a better identification and identifications are the reason why (probably) most non-scientists post on iNaturalist.

I am well aware of the complexity of the issue. I would recommend that a cultivar bin could be created under genus level and under species level. This retains maximum flexibility to cope with circumstances that need it. For instance, Portulaca has many hybrid cultivars that would fit well under a genus-level cultivar bin. It also has the readily recognizable cultivars of P. umbraticola and to a lesser extent P. grandiflora. These would need a species-level bin in order to account for cultivars (especially escaped cultivars as often happens somewhere in the world) that are identifiable to species.

Minor nitpicking, but “Wildflower Mix” is a Portulaca umbraticola cultivar. “Diamond Frost” is the Euphorbia graminea cultivar I mentioned.


Very briefly (IMHO of course):

pros: it would provide an additional way to make users understand that what they have photographed and posted is something “artificial” that is also likely not wild.

cons: which standardized taxonomy for cultivars is to be used?

PS: in the case, it would be helpful also to add the cultivar for genera. E.G.: most cultivated roses actually are complex hybrids while unexperienced users continue to identify them as Rosa chinensis


I think, treating a general cultivars bin as a taxon, would create some problems: (1) We would put different unrelated organisms in a single category. (2) These cultivar-“taxa” would need to have their own hierarchies as well (Pelargonium zonale cultivars being a part of Pelargonium cultivars) in addition to them belonging to a species or genus. Otherwise one would not be able, e.g. to find all Pelargonium cultivars. (3) Where a cultivar name is known it may be more desirable to use this name, than the general cultivars bin, leading to more conflicting hierarchies as in the previous point. [But then we’d actually have to deal with further problems. Which cultivar names to accept? Only those in certain registries? - they only exist for few taxa and are usually not complete. And in which language? - as far as I remember the ICNCP, cultivar names are translatable…]
Wouldn’t it be sufficient to have cultivars documented in observation fields or something similar? Maybe this can be implemented in a way that they are still easily searchable?


So, I think I see what you’re getting at, but I believe they can be alleviated:

  1. Not necessarily if we have the cultivar bins under the respective genus.
  2. I think you may be correct here (it looks like I made a stupid reasoning mistake here in the previous comment). However, I think it would only have to go to genus or maybe family. I can’t imagine it being necessary to go any further. Alternatively, if the cultivar bins were placed directly below each taxon and the developers made an option where cultivated forms could be excluded from searches, I think that would be the best solution all around. Even just classifying them differently under each regular species/genus would allow the option to manually exclude cultivated taxa and be common enough that people would be more likely to incorporate the difference (i.e., if there is a taxon available for it, someone is much more likely to add an ID than an observation field).
  3. While I think it might be worthwhile to add specific cultivar names in some circumstances, I think that it is unnecessary and complicates the taxonomy more than it needs to be as I mentioned above:

Again, I don’t think this adequately addresses the problem simply because I’m not trying to search for them, but exclude them from searches.

Just so everyone’s clear, I don’t want to know about the cultivated organisms and I don’t think it is useful to add a taxonomy for the cultivated organisms. I am interested in the organisms that share the same name as the cultivated organisms that are wild and I think that creating a cultivar bin directly below species and/or genus level is the best way to facilitate this. If this isn’t clear enough from my original post, I’ll see what I can do to make it more clear.


I think that it would be helpful to have something like a bin taxon called “… unassigned hybrids and cultivars” with a “hybrid” rank; for example “Rosa unassigned hybrids and cultivars”, and also a cultivar bin inside specific species called “… … cultivars”, with a “variety” rank; like “Portulaca umbraticola cultivars”. So an extension to the option two:


I would very much like to have a misc cultivars bin to send them to instead of the species level. I identify a lot of penstemons and there are so many hybrids that come from multiple parents. I don’t want to spend any time trying to id which hybrid it is, but I also don’t really like leaving it at the genus level with a note that it’a an artificial hybrid.

I think the solution of a “unassigned hybrids and cultivars” hybrid rank taxon is a great idea.


This was posted recently:
With this in mind, would the use of “[genus] cultivars” at “hybrid” taxon be acceptable to curators? It’s still not perfect as many cultivars are not hybrids per se, but substantially modified forms of the original species.
@bouteloua do you think it would be useful to add a post for this discussion in the curators category as well?

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.