“Complex” is indeed used in scientific literature, at least botanical – for species of uncertain affinity, or hard to distinguish, or insufficiently studied, or… – thus obviously acceptable when defined and used consistently, just like “aggr. aggregate”, “gr. group” etc. Not sure if evading the formal ICBN ranks makes such groupings “invalid” or “unofficial” - but they have their uses and merits nonetheless.
iNaturalist appears to be using “complex” liberally, as a generic equivalent to various types of groupings existing in literature. For such situations where no strict 1:1 equivalent exists at other databases, it makes sense that these iNat delicacies get translated to the next higher level available (i.e. genus at GBIF - or do they have [sub-]sections?). No big deal, provided all operations are explicitly shown and traceable.
Or, because seeing that there is a complex causes them to doubt the species-level ID that they were about to choose. Unless I already knew about the complex and the species in it, I end up choosing the complex to “be on the safe side.” The species may be seperable by range or even morphology, and I may be in a location where only the main species of the complex is found (invasives, for instance), but seeing the word “complex,” in itself, makes the observation “beyond my abilty to make a definitive ID.”
Because of their merits I thought it might be a good idea to report them to GBIF. But as GBIF doesn’t support complexes it doesn’t make much sense. Some liberality is of course part of the complex concept. Though they are in use in the scientific community there are no formal rules like for a taxon. They are by definition not part of the formal taxonomy.
Well, it means following iNat rules to use a broader ID if the observer or IDer is unsure. As far as I can see people who feel capable to ID to species level do so.
People who work with difficult taxa that are not fully represented in GBIF’s taxonomy are free to access the observations directly from iNat; they may find that there are advantages to doing so (such as the fact that observations with a community ID above species will not be sent to GBIF if they are not RG, or observations that may not become RG for a very long time because there is only one specialist active on iNat, etc.).
The way GBIF translates IDs for species complexes is not a good reason to knowingly ID something to species if you are aware that it is actually one of several similar species but you don’t know which of them it is.
One reason that some species complexes (like Draba verna or Portulaca oleracea) end up getting ID’d to species on iNat is because people have different species concepts – e.g. if they are using literature that treats all the members of the complex as a single species. (One of the resources that many people in Germany use for IDing plants differs from the taxonomy followed on iNat in various places; there isn’t a good solution for this except for IDers to cultivate awareness of how their literature translates to iNat taxonomy.)
A “sensu strictu” and “sensu latu” option would not solve the issue, because the IDs still need to be translated into something in GBIF’s taxonomy, so either you end up with a species ID on GBIF that may be wrong, or a broader ID (genus or subspecies or whatever). And “sensu latu” is vague; it does not tell you what species someone includes within that broad concept.
As far as I can tell, GBIF does not even allow “official” (as per ICBN) infrageneric ranks such as sections, maybe to make things a bit simpler to manage. :) On the other hand they support synonymy between various ranks, and also explicitly expose any “translation” performed to fit external data into their taxonomic scheme - which means that they at least include “research-grade iNat complex observations” even if not displaying them as such.
I posted a related topic a few weeks ago: https://forum.inaturalist.org/t/use-of-the-term-species-complex-in-inaturalist-taxonomy/58938
There seems to be little consistency in what qualifies as a “species complex”, mainly because “difficult to identify” is a relative distinction. A few counter-examples to the ones listed in the original post, taken from my other post:
-Datana major/drexelli complex and Halysidota tessellaris/harrisii complex- caterpillars are easy to identify, just adult moths are difficult. People often see “it’s a complex” and question their species ID, placing the caterpillars at “complex” level, despite them being easily identified further.
-Chionodes pereyra complex- the two species involved have virtually no range overlap, so are almost always identifiable based on location alone.
-Feltia subgothica/tricosa complex- you just need the antennae to be in focus and fairly hi-res to identify, which is the case in many of the photos
-Eulithis diversilineata/gracilineata complex- until a couple years ago were considered inseparable as adults, but consistent ID features have recently been found, so most at the “complex” level can be taken to species
I could keep listing examples, but the point is that there are tens of thousands of observations currently at “complex” level that absolutely can be taken to species. In fact, certain “complexes” can get frustrating as an identifier when every observation on a whole continent at “complex” level represents the same species- the only one that occurs there. (Manually adding the species name to all of them is much less efficient than just clicking “agree”.) If all these observations were taken out of the “Needs ID” pool and thrown to research grade, it would make identification extremely difficult. Half the RG “complex” observations would be truly unidentifiable to species, and half would be perfectly identifiable. The only way to distinguish between these is to do so manually, as is the case now.
Was the ID displayed on the observation on iNat for the complex or for the species?
Sometimes it happens that someone will add a non-disagreeing higher-level ID (i.e., first ID for species, second for complex) and then click “ID cannot be improved”. Although the community ID is for the complex, it will be displayed on iNat and shared with GBIF as the species.
No matter wether they have RG there is nothing wrong to through them to ID. You would also need to go through the others at RG species level because you can expect that some of them are wrong. BTW I regularly do such.
It’s the same with some subspecies if you have one native and one that is introduced.
They would be also free to distinguish between observations that had been previously reported as complex using the data fields of GBIF. Actually they need to because complexes due to DQA selection are already in GBIF
And most identifiers don’t go below complex level in most cases just because they don’t have the specific knowledge needed or it’s just impossible with the evidence provided.
And a great part of specimens stored in museums or are documented in scientific literature don’t do below complex because it’s either difficult or the micro species hadn’t been distinguished at the time.
I mean we are holding back valuable data because we don’t like to set filters to id difficult species.
It’s not always impossible and there are a lot of identifiers that do have specialized knowledge, I and many other identifiers I know do ID past complex level routinely
The concept of complexes is to use data eg for distribution maps that otherwise would get lost because there are only very limited data for the species. Delivering the data would make them available.
Remember the P oleracea example. Some 40000 observation are not qualified to be identified to species in contrast less than 100 really are.
Or take Achillea millefolium. There tons of observations held back as complex. Almost all observations that are identified to species don’t provide a better evidence. Instead those have the „luck“ of an observer and an identifier that both had not been aware of the problem of this difficult group.
As a specialist in one or more complexes you can not rely on RG if you think you would like to leave your suggestion in ech observation. And just to mention: I am one of those who does this (not only for complexes).