Strange behaviour after taxon swaps/splits

Prompted by this thread, I just had a look at my maverick IDs, and found some strange situations, which I haven’t touched yet for others to look at.

In one case, there are four IDs on the observation following a taxon split. One is of the now-inactive taxon, and the other three are all denoted as mavericks! How can the 3/4 majority of IDs (and all valid IDs) be maverick?

In a second case, my genus-level ID (Nephila) agreed with subsequent species IDs (Nephila clavipes). Then those species-level IDs were updated with a taxon swap (to Trichonephila clavipes), but my genus ID was not. In this second case, shouldn’t my genus have automatically rolled back to family, rather than being interpreted as conflicting? I have my settings set to automatically update with taxon changes, and while I can see that Nephila was probably split into multiple genera, I would prefer an ID rolled back to the common ancestor than left as the invalid taxon.

1 Like

This definitely seems like buggy behavior, so I moved it to the Bug Reports category.

1 Like

There are no taxon changes associated with the genus Nephila, so that observation appears to be expected behavior. When the species was transferred to a different genus, the genus taxon should have been split.

2 Likes

Thanks! So the species was transferred to a different genus, but the genus was not split, in that case? That makes sense of that case, at least. Is it worth flagging the genus to be split, as I doubt I’m the only person to have fallen foul of this?

Yep I’d think so

Just a couple of comments:

  • anything marine is a problem for doing splits because atlases can only cover land, there is no way to atlas a marine range

speaking as the one who did the spider change

  • I’m not sure this was a great candidate to handle with a split. This was a change where there were 2 existing genera and the species within them were remixed between those 2
  • because of introductions there is overlap in the geography of the 2 genera even after the changes
  • I’m not sure if it is a matter of semantics, but the curator guide specifies that splits should be used when a species is determined to actually contain multiple species. It actually uses the word species.
  • if the above is not valid, then I guess other than spelling mistakes what basis is there for not requiring every change to be done as a split?
1 Like

Sounds like the right thing would be to bump up the genus IDs to the common ancestor, which would require a split.

1 Like

I understand there is a potential problem with genus level id’s, but I still don’t understand how this would be in line with the curator guide explanation of when a split should be used, which is to turn 1 species into more than 1 species, not to switch one thing into a different one thing.

I’m not seeing anything that says splits must be only for species. Here are a few genus splits by Scott:
https://www.inaturalist.org/taxon_changes/45966
https://www.inaturalist.org/taxon_changes/45964
https://www.inaturalist.org/taxon_changes/45962

and one I did recently because North American Cicadetta genus IDs needed to be bumped back: https://www.inaturalist.org/taxon_changes/80453

Curator Guide:

Splits a single “input” taxon into several “output” taxa, e.g. when a species has been revised and determined to contain several distinct, named species.

For Taxon Splits, there is one input and multiple output taxa, so the input goes in the “Single taxon” box on the left and the outputs go in “Multiple taxa” on the right.

For splits, help people out by explaining how to differentiate between the output taxa so people know how to update their content if it couldn’t be updated automatically.

Taxon Change description of split

Old taxon split into multiple new taxa

The “new”/output taxa would be the new taxonomic concepts of those genus names (in this case, same names, new meanings).

1 Like

Basically what this is saying then is every single taxa change that is not a spelling correction needs to be done as a split because of the possibility of genus level ID’s.

And / or that there needs to be some way to find identifications of a specific taxon id but not its children to determine if said ID’s exist.

Is there even a way to do that search? Searching https://www.inaturalist.org/identifications does not work this way, it finds all id’s plus id’s of children. So for example https://www.inaturalist.org/identifications?taxon_id=6922 finds all id’s of taxa 6922 (Mallards, Pintails and Allies), and any children of it.

The API takes rank as a parameter for identifications search, but https://www.inaturalist.org/identifications does not. You can use pisum’s tool to see IDs only for 6922: https://jumear.github.io/stirfry/iNatAPIv1_identifications.html?taxon_id=6922&rank=genus

1 Like