Use of the Term "Species Complex" in iNaturalist Taxonomy

General
1

This is probably me being a pedant, but if there are any fellow pedants out there, feel free to chime it.

I’ve been increasingly confused by the use of the term “Complex” to refer to groups of species on iNat. My understanding of a species complex is that it refers to a group of closely related species for which the boundaries between the species are unclear. This usually means that some undescribed cryptic species are in there, or two species may in fact be synonyms, or there are two names but many individuals fail to fall neatly under one name or the other. Species complexes are the targets of taxonomic revisions, which hopefully resolve the complex and define the species involved in a clearer way.

Three examples of species complexes on iNat that I think fit this description well are “Complex Anacampsis conclusella”, “Complex Exoteleia pinifoliella”, and “Complex Enchenopa binotata” (sorry, my examples will all be bugs, because that’s what I know, but I’m sure this applies across all taxa). In these cases, there are only two or three named species listed under the complex, but everyone “in the know” is aware that there are other undescribed ones in there, and the groups are in need of revision to clarify how many species there are and how to separate them. So “Complex” level is usually the best an ID can be for these observations. “Complex Sphinx gordius” is another great example of one of these- current research is ongoing into what differentiates the species in this complex, so any species-level ID at this point in time is tentative at best.

But there are a lot of other cases I see where the term “Complex” does not apply. “Complex Chionodes pereyra” includes two species with ranges that don’t overlap, and, as far as I know, no known undescribed species. The two species are sister taxa, and they do look basically identical in photos, but there’s no “complexity” to them beyond that. They’re two known entities that can be differentiated based on known characteristics, albeit with some difficulty if you have a live photo with no location data. The same is true of “Complex Chionodes trichostola”, except that the species look even more different.

There’s a laundry list of other “Complexes” for which the species involved can be clearly differentiated, just not based on photos of live adults.
-“Complex Desmia funeralis” includes two species that you just need to look at the underside of the body to identify
-“Complex Feltia subgothica” you need to see the antennae up close
-“Complex Burnsius communis” you need to dissect the genitalia
The list could go on and on…

I don’t see how these species groups fit the definition of “Complex”. There’s no ambiguity as to how to differentiate the species and where the boundaries are; there’s just difficulty seeing the important features in a typical photo. But in these cases, at least the taxa included form actual taxonomic clades…

“Complex Datana major” includes two species that are not sister taxa, but happen to look the same in photos of adults. They have extremely different larvae, different host plants, phenology, ranges, genitalia, DNA barcodes, etc. There’s no question about species boundaries at all, and taxonomically, D. major is sister to D. robusta, not the other species in the iNat complex with it. The adult moths just happen to look the same in photos. Under no concept of the term “species complex” should it apply to this pair.

To be clear, I very much support the inclusion of all these taxonomic levels on iNat. They are very useful, as photos can often only be placed to these levels. In fact, I’d like to see more of them (Euchlaena muzaria and Euchlaena obtusaria, for example, are probably just one variable species, and which name is applied to which observations on iNat currently is pretty much random.) So I’m not suggesting that these “Complex” levels should be eliminated- just that the terminology is very confusing. In different places on iNat "complex can mean:
-Ambiguous species groups with unclear delineation between species
-Clearly defined species groups that just present ID challenges from photos
-Pairs that aren’t even in a species group together that present ID challenges from photos
It’s like the term “complex” is sometimes being used in the taxonomic sense, and other times as a placeholder for “cryptic species pair” or “species group”. It has real consequences too: to someone who doesn’t know a taxon very well, seeing that a species is “part of a complex” on iNat could mean anything from “don’t even bother with a species ID because even the taxonomists are confused” to “a closely related species to this one exists 2,000 miles away”.

Are there any fellow pedants out there who are also bothered by this? And if so, is there better terminology that could be used to keep the term “complex” reserved for actual taxonomically ambiguous situations?

4 Likes
2

I’m a pedant also, but my recollection is that these terms are sometimes used interchangeably even in the primary literature. I do agree though that complex suggests unresolved taxonomy whereas group suggests closely related but resolved taxa that are perhaps hard to tell apart. But that’s my interpretation.

3

Yes, it bothers me too. I recently flagged a taxon (an invalid name used for hybrids of several plant species in cultivation). In the end the resolution was to create a species complex containing those species, so that the existing observations wouldn’t be bumped all the way back to genus. I pointed out at the time that this use was against the curation guidelines but others in the discussion thought that it was more important to make a functional category than to follow the letter of the rules. And I understand that impulse - but it is frustrating when rules are applied inconsistently by different curators and in different contexts.

3 Likes
4

Well, to me, a complex is just another taxonomic rank, and it doesn’t necessarily depend on how often species are confused with one another. Using “similarity” as a criterion for defining taxonomic ranks could be somewhat arbitrary, especially given the fact that some completely unrelated species may be difficult to distinguish from one another

I do agree that the complexes added to iNat may feel a little arbitrary. Sometimes they have little use at all like in the examples that you gave, although maybe there were other reasons for them being added. For example, even when species are easily distinguishable, they might be part of a well-defined clade, and grouping them into a complex reflects their evolutionary relationships. If the complex needed to be raised to genus level, it would be easier to do so. Also, sometime complexes are added for consistency with how other taxa are organized. If other groups in a genus have complexes, creating a complex for closely related species—even if they’re easily distinguishable—keeps the taxonomy structured similarly across the board.

5

FWIW, I suspect species group vs complex is mostly zoology terminology–I don’t think botany uses either term too much or makes very clear distinctions.

I think the pressure to create them on iNat mostly comes from the existence of what you’re calling species groups. Presumably “complexes” get resolved into “groups” as systematics advances: in complexes, there’s a single name for the members because the components of the complex are still cryptic/undelineated/undescribed, and in groups, names for most or all members exist (but we find we need a taxon for the group because it’s rarely possible for us to apply those names based on the evidence available in observations).

It may be one of those things that’s just gotten baked into the system because iNat has gotten bigger and more complex than originally anticipated. i.e., the taxonomic hierarchy only supports one infraspecific level, which isn’t quite right nomenclaturally but is a sane decision for post-molecular taxonomy.

2 Likes
6

The flip side of this is that there are many groupings that do not exist on iNaturalist which are very useful for, say, birders in the field, but there doesn’t seem to be any accepted taxonomic category for them.

This is especially obvious in the case of IDs which are normally made based on sound. For example, most birders can only distinguish Alder and Willow Flycatchers by ear, and it’s debated whether it’s even possible to do it without sound. In birding this pair is routinely reported as “Traill’s Flycatcher”, a combination which has its own banding code. But there’s no equivalent on iNaturalist, so those observations are stuck at genus Empidonax, even though you have narrowed down the ID to 2 species out of 14.

This kind of grouping would make it a lot easier to monitor range expansions in species that have historically been separated by range, like Carolina and Black-capped Chickadees. The inability to say “this bird is one of these two species” can lead to false confidence about identifications based on range, which is a self-reinforcing cycle with the range maps. (…they hybridize and they learn to make the “wrong” sounds, also.)

But there doesn’t seem to be any accepted standard for this kind of grouping in birds, so we just don’t have it. :woman_shrugging:

2 Likes
7

I wish the complex ID would list the possible species below it and give a custom explanation for it’s existence personally.

2 Likes
8

I’m not bothered by the ambiguity. I see the complex as a way to allow us to apply a name to taxa that might otherwise by impossible to name based on the photos. I think the original idea was that each complex should be a clade of very similar species, but the term becomes even more useful if we’re not confining it to clades (groups that are each other’s closest relatives). Maybe we can set up a Bombus vosnesenskii clade. (Species almost identical to B. vos. aren’t its sister taxa.)

4 Likes
9

I think confusion about species complexes comes largely from a combination of a very broad interpretation of how to apply that rank:

As well as a preference from staff to co-opt other ranks (like species complex) rather than make a new taxon rank for species groups:

I think most taxonomy curators, as fellow pedants, don’t like to incorrectly use plant ranks for insects (I’m only aware of one instance, the genus Platycheirus, where this has actually been done), which means everyone is sticking to formal subgenera and a broad use of informal “complex” as the least worst compromise.

10

Cryptic monophyletic groups with uncertain taxonomic divides between species are a common situation that warrant a complex. However, I disagree that completely cryptic groups (i.e., species only discernable by DNA analysis) or species groups with uncertain taxonomic divides are strict qualifiers of the term “complex”, because the term is often used in a looser sense off iNat too. Complex Desmia funeralis and Complex Hyla versicolor are both used off iNat. The species within these complexes are difficult to decern (only near-cryptic as there are ways to decern species visually or auditorily) and there is no (or at least very little) disagreement about the taxonomic divides within these complexes.

2 Likes
11

It is worth mentioning that in mycology the term ‘species complex’ has a long tradition and is being used in the literature with increasing frequency. It is very useful. This isn’t a zoology-only issue. Sequence data often show that what we thought were single widely distributed species are actually multiple related cryptic species. Sometimes the related species are allopatric and sometimes sympatric. Sometime members of a complex can be macroscopically identical (and sometimes even microscopically identical), whilst others easily separated.

On iNat there are vast numbers of macro-fungal observations incorrectly assigned at species level where we know that complexes are involved. It is very hard to shift observer’s from their ‘pre-molecular’ concepts and names used in guide books. The species complex provides a painless way of injecting a bit more scientific rigor. Although correcting the identifications can be an uphill struggle against the CV.

In iNat the term Species Complex has a specific definition in the curator’s guide.
https://www.inaturalist.org/pages/curator+guide#complexes
Allopatry is not part of that definition.

4 Likes
12

Various threads have discussed the use of this term, and I appreciate that iNat has a relatively pragmatic approach to it. I don’t think the term ‘complex’ is used exclusively in the ways suggested even outside iNat. For example it often refers to groups of sister-species (or presumed sister-species) which are very hard to identify (e.g. genital determination is required), or where only one sex can be distinguished.

As mentioned above, in insects especially it is common for the literature to refer to ‘species groups’, which are often larger and less cryptic, but since iNat does not have this rank as an option, ‘section’ or ‘complex’ are often press-ganged into service in lieu of it. Platycheirus is an instructive example where both ‘section’ and ‘complex’ are used in this way because there are groups within groups! (all widely accepted ‘species groups’ outside of iNat, and used by recording schemes - but only one is a ‘true’ complex (males are visually identifiable, females completely unidentifiable). I find it good practice in these cases to advertise the fact that the taxon rank is ‘substituting’ for a ‘species group’ in the common name. For example making the common name ‘Platycheirus albimanus group’. That, after all, is how ordinary people (who are interested in hoverflies do refer to it…

The key thing is not to invent groups with no literature basis just because someone thinks they are confusing. Otherwise I don’t think the literature basis needs to be very strong. So long as the group can be defined from the literature and its members determined. Which exact rank it is placed at is less important.

3 Likes
13

At least the first part of your wish is already true. Just go to the taxon page an hit the taxonomy tap, where the included species should be listed

14

The curator guide is pretty clear:
https://www.inaturalist.org/pages/curator+guide#complexes

I had never thought complexes were only for unresolved groups. The ones I’ve created are only really used for groups that have cryptic species but they have been sorted out taxonomically. Although the literature probably tends to refer to these as species groups.

1 Like
15

“Species complex” in iNat is obviously used for any group of species which are difficult to identify from photo, to avoid having to drop back to the first ‘safe’ rank of identification, typically genus, which would make getting Research Grade more difficult. Put another way, the species concept iNat uses is “a group of organisms which can be identified from any random photo”, and groups for which this does not hold true become species complexes.

OP is right that “species complex” in science is used for groups of species in need of revision, and what iNat usually means by “species complex” is more a special case of what in science is referred to as “cryptic species”.

However, “species complex” and “cryptic species” are informal terms, not taxonomic ranks, and not governed by any rules.

My personal preference is that when an organism in a photo cannot be confidently identified to species, it should only be identified to the first ‘safe’ level, usually genus, but I realize there is too much birder DNA in iNat for this to happen and that the only realistic alternative to “species complex” is that all the difficult-to-separate species get lumped into a wastebasket species.

So while I mostly agree with OP, I feel that referring to groups of species which are difficult or impossible to separate in photos as a “species complex” is the least-worst realistic solution to identification problems resulting from iNat’s species concept.

2 Likes
16

In my opinion we should not be overly strict and pedantic over the application of complexes. The reason? The are two, I think. First is connected to the primary goal of iNaturalist – re-connect people with nature. And the users are so much more happier when they have the name in the form of a complex than just the name of a genus, especially in the case of large genera. I learned that people accept, for example, complex Rhizocarpon geographicum much more happily than genus Rhizocarpon (which is quite large and includes very different species, even though the complex does not neesarily include taxonomically closely related species. The second point is the secondary aim of iNaturalist – collecting data and its quality. Species-level identifications in the cases when IDs cannot be done from a photo is a big nuisance when it comes to cleaning up. Complexes go into certain wastebin and make less clutter.