Vulgichneumon brevicinctor is a North American species. It hasn’t been recorded from UK but there are 45 UK observations (which actually belong to a variety of species).
iNat shouldn’t offer that taxon in the UK!
Actually the list of Ichneumonidae taxa is pretty poor. Lots missing and lots I don’t recognise.
We have a digital taxon dictionary for eukaryotic UK taxa (UKSI, maintained by Chris Raper). You should use it.
if this species doesn’t occur in the UK, it means these records are misidentified. The way to resolve this is to correct/bump back those IDs with an explanation
If a species name exists in iNat, but is yet to have any observations, it’s impossible for it to appear as a thumbnail in Explore; those thumbnails show species with at least one observation.
iNat is a global database, so our taxonomy aims to (where possible) follow global authorities
Maybe, but how many other extralimitals are there that I don’t know about without searching for every known species? Or going thru all 21,464 ichnemuonidae observations?
I was referring to the list of names it offers when you type in a suggested id. This should be restricted by the location of the record (or original poster).
Chris Raper is at the London Natural History Museum and it’s their list he maintains. That’s as global as it gets.
This is in part due to the new way in which the geomodel is produced. There are no Research Grade observations of this species in the UK, and only two in Europe (which I have just now knocked back to genus) but the geomodel predicts that the species could occur across Europe. This is a known problem with the geomodels and affects other species, too. See other discussions about the geomodel on the Forum (e.g.). How the geomodel is calculated is still being refined by @alex and others on the iNaturalist team.
There will be species missing, because nobody has yet observed them on the platform and had them confirmed. There are also likely to be some errors, because, as you know, some of these species are difficult or impossible to identify from photos. If you see any errors, and have expertise to correct them, please do! That’s how the platform works. iNaturalist is very effective for taxonomic groups where records are actively reviewed and corrected by taxon specialists, and not so reliable for groups lacking active, expert identifiers.
Anyone with expertise in a taxon can make an important contribution to improving and maintaining data quality on the site. In a group such as Ichneumonidae, where there are many errors in identification, I’d suggest to prioritise correcting Research Grade observations, as there will be fewer errors to correct and you’ll have a direct effect on the data exported to GBIF.
You could, for example, scan through the 136 species with Research Grade observations in the UK (from the link above), starting from the bottom (species with a single record) and check for errors there. If you find any, please feel free to suggest a different ID on the observation and to ask how it was identified.
Global for the hwole globe but with about 5% of our spp missing. But can pick them up from external lists, but a beginner user is just going to assume he’s wrong.
It doesn’t say they resemble them, just that they have been misidentified as such (or vice versa). Quite dissimilar species can end up in this list if a wrong ID is applied, whether because of a CV suggestion (possibly of a poor photo) or someone just not being familiar with the species. The tab can be a useful place to look for species that are easy to confuse with this species, but may also collect some fairly random other species. Clicking on the number will take you to the record(s) thus confused, which makes it easy to review potential problem observations you might be able to help ID.
The tab showing “commonly misidentified as this species” merely shows taxa for which someone at some point chose that ID on an observation that has since been corrected. Many users do not have the knowledge to evaluate the computer suggestions or they do not know that they can add an ID of their own that is not on the list, so they will use one of the suggestions even if it seems to have little resemblance. Depending on the photo quality (e.g., if the wings are folded over the body and it is difficult to see the color of the abdomen) they may think that their individual resembles the thumbnail of the suggested species. I find that hymenopterans seem to be really really unintuitive to a lot of people compared to other taxa (say, butterflies or ladybeetles) because we tend to focus on features that are only of limited use (e.g. color patterns) and miss traits that are important (subtle differences in body shape).
As I’m sure I don’t need to tell you, ichneumonids are a notoriously difficult group. This is particularly the case on a platform like iNat where ID is based on photos of variable quality rather than specimens. There are currently only a couple of active IDers in Europe who are willing to try to tackle this challenge.
As a result, a very large percentage of observations are never looked at by anyone with expertise and many people rely on the computer suggestions, which are very incomplete. The image recognition algorithm is only trained on taxa that have a certain number of observations, because a single photo would not provide it with enough data to realistically distinguish it from other similar species. So not only are many ichneumonid species not represented on iNat because there are no observations of them (or no observations that anyone has ID’d as such), the computer vision only knows (i.e., can suggest) a small fraction of those species that are represented.
In addition, the computer vision tends to struggle in general with arthropods where ID often depends on tiny details. This problem is not unique to ichneumonids, but I suspect the computer suggestions are worse for this group than many others because there are few specialists correcting the wrong IDs, so there is a negative feedback loop (it suggests a few species because they are the only ones it knows, observers use these suggestions and therefore it keeps suggesting them).
One thing you can do to help change this situation is adding/correcting IDs yourself. It doesn’t need to be “all 21,464 ichneumonidae observations”; even if you clean up one or two species this can make a real difference. Ideally you would be able to get support from other users (either people already on iNat or experts who would be willing to join iNat to help you).
One challenge here is that there are few good resources that are any use at all for photo ID and I imagine comparatively fewer experts than in many other insect taxa. I have made several attempts to make sense of ichneumonids for my own observations and so far have had no success even narrowing down IDs beyond family except for a few highly distinctive species. So if there are resources that you can point people to, I imagine I would not be the only person who would be grateful for guidance.
Oh, yes, and for cases where you know microscopy/male genitalia are needed for a species ID, you can go to the data quality assessment (bottom of observation page/rightmost tab in the “Identify” module) and check the box “ID cannot be improved”. This will allow the observation to become “Research Grade” at genus level and it can be helpful for indicating which observations cannot be refined further vs. ones that have merely not been looked at by someone who can evaluate them.
What do you by “our”? What do you mean by “for the whole globe”? It is a UK species list, it is even in the name. The title page says “flora and fauna in the UK”. 70 000 species is way too little for the whole world.
