Changing of ID from specific species to higher taxa

One great things with iNaturlaist is ID help, especially if you make an error when entering or an actual ID mistake, but it is frustraing when reviewers change the ID from a specific species to the taxa - I fail to understand why - if you cannot make a positive ID from that of the original observer - then do not make it, but why change it to the upper level taxa if you have no idea what specific species is yourself? Just leave it - as it will not become research grade - I have had people do this, when if they knew more about the species distribution in that Country they would not be changing the original ID. All they are doing is basically messing up your own records by doing this, and is of little benefit - great if you want to give me a correct ID, but if you cannot just leave it as is - all I end up doing is withdrawing from the community ID,

It is rare for me to do so, but if I know for certain that a suggested species ID is incorrect, I will give a disagreeing ID at the lowest level I feel confident. I try to make sure to explain why.
It does matter, because I don’t want false positives. Even non RG observations are data points, that can be used by others. Also, if someone with less experience comes along, they may agree to the incorrect ID without knowing any better.

If I’m not sure either way I’ll usually just leave a comment.

12 Likes

If I’m confident that the plant is not species A, but I’m confident it belongs to the genus, I will change the identity to the genus, always. I think I should. I should say why and often do, but sometimes get lazy and don’t.

I could be wrong, of course! If you disagree, or if you wonder why I changed it, comment or ask! We can have a conversation in comments on the observation, and I can learn something.

17 Likes

Thats exactely the point, I have observed this phenomenon several times, here on the Caribbean islands, where i do most IDs. Once there is a observation that was wrongly identified (e.g. as a species that occurs on the American mainland only) than other identifiers, will see this name in the compare feature (and start using it for their identifications!). This means that a single wrong species-level-ID, in a certain area, is often a starting point for a chain reaction of wrong IDs …

If i cant make a better species-level-ID, i usually push these wrong species-level-IDs, up to a higher taxonomic level. As higher taxonomic levels are often not shown in the compare feature, they are less likely to cause a chain reaction.

7 Likes

Many organisms simply cannot be identified to species level from the kind of photos typically taken in the field. The Community ID should be based on the available evidence. If the evidence only supports identification to a higher taxonomic level, that is what the Community ID should eventually settle on.

7 Likes

I personally do this in order to keep incorrectly identified species off the distribution maps. I find these maps very helpful in my own learning process. Large numbers of e.g. New Zealand native plants in e.g. New Jersey stand in the way of my own learning, so I use higher level taxon IDs to remove them.

11 Likes

but I assume unless they are verified as research grade, they should not be appearing on distribution maps - look at from this angle observations with no photos get left and stay as such, if a photo, then if anyone agrees they become research grade, if some disagrees then they should be able to say what it is or at least leave a note why it is not what it is claimed to be explain why they do not know exactly what it is - but I have had people changeign things because they do not agree 100% what it is but cannot say what it is, yet that species is the only one that occurs there - I think my point is unless you know excalty what it is or are abolsutely 100% sure it is not and explain that, just going in and changing to genus is little help, just leave it and it wll stay non research grade as all observations with no photos do…the system is used by some for self recording purposes as well as wishing to contribute to the research grade data - in the UK on iRecord the recorder for that species will aslo verify records without photographs based on the observers recording history and the status of the species in the area, and will be shown a positive or probably or not proved etc… and then into the local biological records data bank - I have had items changed with photos verified by local recording experts on here to genus - which is not overly helpful and to keep the record for personal purposes I then have to elect out of communtiy ID because of this…where the real value of these sites is help where yyou are not sure of the ID, or in some cases I have put things in worng, either through ID error at the time or oddly slecting the wrong species on data entry -

For what it’s worth, if there is only one species of a genus present in a location, that’s a valid argument for a species level ID. Though with a grain of salt!

You can certainly argue about whether or not “Needs ID” observations should appear on distribution maps, but as of now they do. If they didn’t, I would have less incentive to move observations that I feel confident are incorrectly identified to coarser taxa.

I can’t say I understand the rest of what you wrote, except to say that you should keep in mind that iNaturalist is a global community, and what you do does affect everyone else. It’s perfectly fine to keep personal notes. It’s also fine for the community to speak up.

3 Likes

What exactly do you mean by “changed”? It would be helpful if you gave links to some iNat observations showing community IDs that you think have had a detrimental effect.

1 Like

while ‘explicit disagreement’ can be jarring and the criteria for doing so (whether one knows the ID is wrong or just isn’t sure) somewhat controversial, the way iNat works, explicit disagreement is definitely appropriate if you are confident the higher-level ID is wrong. Having ‘needs ID’ observations on the map is important among other things so that one can find and inspect these possible out of range examples. But, one can filter for research grade only in their queries if they want, and they do have a different looking pin when you are zoomed in.
Note that if you really don’t want people bumping your ID to a higher taxonomic level you can just opt out of community ID for that observation.

5 Likes

@clamsdell The problem here is that it can actually be pretty hard to get to “research grade”, especially with poorly known taxa. Let me give you some numbers. Within sect. Anisophyllum (the group I study), there are currently 22,995 observations IDed to species. Out of these, 21,081 are research grade. This seem like a good ratio on the surface, but what happens when you look at taxa at locations that are poorly represented (and in greatest need of representation)? Let’s take South America as our example. There are 336 observations IDed to species. Out of these, only 236 are research grade. Only about 2/3rds are research grade, but it gets worse. 116 of these are of E. hirta (a common weedy species) and 102 are research grade. If we exclude this single species, the numbers become 134 out of 220. I.e, only a little over half are research grade for the entire continent! I have inspected all of them and added an ID to most of them. If we look at some of the species, at least one (E. jamesonii) has no research grade observations.

While South America is a bit of an outlier in many respects, it illustrates the point well. To me, these kinds of observations are usually the most important and interesting. With the more well known groups, observations often don’t become research grade because they are of poor quality. In less well known groups, observations don’t become research grade because we don’t know enough. The latter matters a lot more to me than the former in any case.

3 Likes

This is a complicated issue as the data is used by others. What seems to be happening is that personal data management is interfering with community data management. I seem to remember some discussions similar to this when talking about users opting out of community ID, but I don’t think they went very far. @clamsdell One way you might try to get around this if you don’t want to opt out of community ID is by using tags or observations fields. I’m not sure if this will help or not.

Regardless, I wonder if a discussion about disabling community identifications for
some types of casual observations would be fruitful (and the ability to make observations casual and simultaneously disable identifications for observations that are being used only as data for an independent project). I have shared a few observations that have photos, but changed them to casual because I knew the evidence wouldn’t be enough for me to confirm as that plant much less anyone else. iNaturalist is useful for managing data (there are some problems, yes, but it’s still useful) and I can think of instances where this would be helpful and might alleviate the problem here. @charlie What do you think of this? Worth a discussion independent of the problem here? There are some problems I can foresee and it’s not exactly in line with what iNaturalist is here for, but might at least be worth the conversation.

Hmm. I agree that some things can’t reach species and unfortunately at the current time, you can’t tag something as not needing species ID to ‘request’ agreement with genus or higher ID. otherwise, with two genus level IDs and checking the ‘no, it’s good as it can be’ ID box at the bottom of the page you can reach research grade for genus. I do wish there were a way to tag something as not needing further ID without ‘breaking’ the observation by flagging data quality inappropriately (like marking my own observation as having the wrong date when it doesn’t). I too occasionally share things where it seems valuable to have a photo for whatever reason but the photo does not have diagnostic features. If the photo ends up totally useless (like too blurry to make anything out) I sometimes delete it and just leave the observation as casual.

I’m not sure if that is what you are getting at or not?

Sort of. I also wonder if it would be useful to be able to turn off identifications altogether (making the observation casual). That way, users could still use them for there lists and personal data, but not have it come into conflict with the other iNaturalist data.

Also, I’ve sometimes wanted to change one of my observations to casual without messing without doing it in a roundabout way without (for instance) marking yes or no here: “Based on the evidence, can the Community Taxon still be confirmed or improved?” or manipulating the data in some other way (as you mentioned above). It always just seemed too easy to me for someone to come along and say, “actually, there is enough evidence” when there isn’t.

Overall, this would give the observer a little more control over their data if they didn’t want to be a part of the community portion. Does that make sense?

Would a combination of making Geoprivacy “Private” and opting out of Community ID get close to what you are trying to do?

1 Like

I’ve thought before that it might be nice to be able to have private observations, just because the database and interface are so wonderful, and I do sometimes come across either questionable data or data i can’t share for other reasons (but not so secret I can’t have it online). However, as I understand it the vision and direction for iNat is to be a community of naturalists, so while obscuring and a few other ‘privacy’ or ‘solo’ features do exist, they aren’t likely to be developed much further. I personally just wish that a user checking that box would (still? again?) bump things to casual on their own observation for reasons you mention.

This would still be eligible for research grade, right?

1 Like

That’s close to what I’m thinking about. The difficulty of opting out of Community ID is that it applies to all your observations. This is actually something that, for somewhat unrelated reasons, I think would be useful to change too. I would really enjoy being able to opt out of community ID for Euphorbias and Plants of the Llano Estacado while not opting out for pollinators (of which I know essentially nothing about).

1 Like

I was thinking not, but just experimented with one of my own RG observations, and Private geoprivacy did not change its RG status.

Good to know. That’s a little sad to hear as I think it could alleviate some difficulties and/or tensions, but I suppose it really isn’t crucial or anything.