Contacting lepidoptera identifiers in Africa

I’ve been asked by the LepSoc Africa council to find a way to make it easier for its members to get high quality identifications on iNaturalist. In iNat there are 6864 identifiers of Lepidoptera in Africa and 3557 in South Africa. Apart from a league table of number of identifications there’s no information on who does what or what their quality is. Just knowing who tends to specialise in what higher taxon such as superfamily would help. I only do Papilionoidea and I know a few who do general Lepidoptera. I also know that a lot of the specialists don’t use iNat for various reasons and that’s fine, but I’ve met a couple (purely by luck) who do. It would be great to have a database with these fields: [Identifier_ Username][No_of_identifications][taxon_most_often_identified] so we could publish it (on a forum) and advise members on how to tag people appropriately.

2 Likes

I tag (with his permission) @traianbertau for insects in South Africa.

Can you remind the members to annotate (the more the better if they know - this is an adult female, this is a caterpillar, these are the eggs.) They could also help to annotate the already uploaded obs.
Since the recent update - iNat’s CV will offer caterpillar pictures where appropriate. But that again means - someone competent needs to make the taxon pictures show the relevant field marks. No taxon picture of a caterpillar = no joy / luck.
https://www.inaturalist.org/blog/106449-better-image-matches-photo-similarity-update-to-computer-vision-suggestions

As a generalist I ID as Lepidoptera, and feel that the IDs move along.

Also the leaderboard gives no indication that ‘John Smith’ has explicitly demanded NOT to be @mentioned.

@mention one at a time. (Ask nicely, everyone else is busy too)
Having asked for help, follow your notifications, and respond - withdraw or agree, or discuss why it is your Species A not their B.

We need taxon specialists to pick over Pre-Maverick Lepidoptera for Southern Africa only 221 obs. Please ?
Update 1 March now 185 obs. We make progress!

Ultimately - the best way to get IDs, to know who to ask next time - is to use iNat. Both to upload your own obs, and to help with the backlog for others. Not the answer you were looking for.

PS https://www.inaturalist.org/observations/151593606 random example of one from my garden. I am awed by the response I got on iNat. More - I could not ask for! as Tony Rebelo said there - just requiring a critical eye

3 Likes

That’s great - I can publish that in LepSoc News Africa. I was able to nail a few myself and I’m not a ‘not-butterfly’ expert!

1 Like

You might want to edit the topic to be “Contacting lep identifiers in Africa” or something like that. I took a look at this because I was curious, but I was kind of expecting it to be another thread about the etiquette around contacting IDers, not a concrete request for identifier contact preferences in an area.

I think I’m probably on your list because I incidentally ID some caterpillars in Africa. But that’s mainly when I’m looking through the Micromoths project, since its filters include observations that have an order-level ID or conflict in the project. So I can often add a family/superfamily ID to African caterpillars but not anything more specific. I can ID hesperiid caterpillars in a bit more detail though.

6 Likes

If the members of this society have some training and knowledge themselves, it seems like recruiting knowledgeable people to join iNat and become IDers would provide a net benefit to everyone.

Another possibility might be to create an iNat project for the society. Project journal entries are a good way to share information and having a project to network around may attract skilled identifiers.

The phrasing “how to get high-quality IDs” feels like a rather transactional view of iNat – i.e., assuming that observers are entitled to IDs and IDers have some obligation to provide this service, rather than just everyone being users contributing however they can to a collective endeavor. Since members of the society are presumably people who are interested in learning about lepidopterans themselves, why not approach this instead as “how to help members evaluate the IDs they receive”?

I would not be particularly happy if someone were to compile and publish a list of “recommended IDers to contact” with my name on it without asking me if I was willing to be included. This essentially amounts to volunteering my time and energy on my behalf, even if I would prefer to decide myself how I want to direct my efforts. Others might feel differently.

(For context: I am a layperson who has acquired some specialist knowledge in my taxon of interest but would not consider myself an expert. This is stated clearly on my profile. I am one of the more active IDers in this taxon in the wider region and thus increasingly find myself getting tagged, often unnecessarily or for observations I cannot help with.)

6 Likes

Please consider asking people if they want to have their usernames published so lots of people can tag them. I like to help, but there are limits. (For example, over the past year one observer has tagged me for every single grass he’s posted and hasn’t provided tentative identifications himself. I have gone from happy to help to feeling exploited.) Don’t over use identifiers.

Also, remind people to provide a tentative ID if they can and, most of all, to learn! There are no doubt difficult species groups where every ID should be checked, but if the species is relatively easy to ID and the same observer tags the same identifier on it repeatedly, the observer can easily feel that the observer isn’t doing his part of the process.

9 Likes

I wonder how feasible it would be to introduce a “people who’ve identified observations like this” algorithm to replace the “top identifiers of X” currently shown on observations.
For example, on this observation of a Tineoid larva from Australia, my name shows up in the bottom right corner as “Top Identifier”: https://www.inaturalist.org/observations/263132733
But I couldn’t possibly be less qualified to help with the ID of this observation. The source of my Tineoid observation numbers is my identification of adults of one Western Hemisphere genus in the superfamily. Surely an algorithm that takes into consideration the geographic locations, life stage annotations, and specificity of my identifications would not name me as a “Top Identifier” of observations like this. We have the annotations, the leaderboards, and the Geo Model… I wonder if it would be possible to somehow combine these into a CV-esque “Identifier Recommendation” list.

4 Likes

These are great suggestions, thanks. And I’ll take what you recommend seriously. Ultimately it is about learning, but a lot of our members are unused to iNat and get frustrated when they upload something and see a load of conflicting IDs.I’d like a way to encourage them to identify something to at least superfamily level and have someone knowledgeable pick up on it and give an accurate identification. I specialise in Papilionoidea and have done tens of thousands of identifications, coaching users how to get closer next time. I know how to get a simple report on unknown Papilionoidea in Africa but I am inexpert in preparing more sophisticated reports. It looks like I’ll get the help I need here.

That is kind of what I was driving at. If we know who the top identifiers of a particular high level taxon are, it becomes possible to contact them and ask if they don’t mind being tagged. There were some good suggestions in this thread and I have some things I can try out now.

people should self-publish / self-identify. they can start a project with the taxa that they are willing to monitor, and others who want to help can join that project. they can use the journal to add helpful hints for identification or resources to share knowledge, and they can use the project description to post rules of engagement (ex. folks asking for identifications need to have made a tentative ID to family).

someone can collect each of these projects in an umbrella, and that can be thing that random folks can be directed to if they want to find someone to help with identification.

if anyone wants to stop identifying, they can remove themselves as a curator of their project or if they’re the only one, they can remove the project from the umbrella project or they can post a journal entry to say the project is no longer active (or temporarily inactive).

I find Tony Rebelo amazingly helpful. I asked him for some help on this as well and he’s responded. I’m going to follow up on some of the suggestions below as well, so we can improve the experience for our members and also improve the quality of data going into iNat. Ultimately the Society wishes to use reports from iNat to inform decisions on conservation and biodiversity improvement projects

2 Likes

Hi Steve, First a thanks for all you have done for butterfly id in southern Africa through your books, Facebook group, etc.

I think that outside groups often have an unrealistic idea of what iNat can do. It won’t answer all id questions but it certainly helps a lot, both through the CV and through individual identifiers. For Australian leps, the CV will id most butterflies and, for moths, it will get to species for the more distinctive ones or genus/family/superfamily for the others. (I no longer page through the plates of “The Moths of Australia” - I now see what the CV suggests and follow up those suggestions.) I guess that it would be similar for African leps.
It will only get better as more people submit observations and help with the ids. This is a synergy we need to nurture.

I suggest that you look at the project “Australasian Fishes”. This has really become an invaluable resource, showing the synergy I mention above as more and more people (fisherpeople, divers, naturalists, beachwalkers) submit their observations to growing body of identifiers. The identifiers are both very experienced amateurs but also a growing number of professionals in museums and unis who see the value of this growing database. Regular journal posts, highlighting recent discoveries and range extensions, help to keep this interest going. Contact @markmcg to see what tips he has for starting and running such a successful project.

Perhaps you need to set up a project which will follow this model but you need to encourage as many identifiers as possible to participate.
Good luck, Pete

3 Likes

Does this happen a lot? Actual people disagreeing about the ID for an observation rather than conflicting CV suggestions?

I ask because this hasn’t been my experience, but it is possible that the iNat community in Africa has very different dynamics than in the area where I am active in Europe.

What I normally see here is that the vast majority of the time IDers are in broad agreement. Occasionally over-enthusiastic users will suggest or confirm wrong IDs based on the CV, or there may be specific taxa that are regularly a cause for confusion (e.g., because of lack of consensus or differing ideas about how to distinguish them even beyond the iNat community). And of course everyone makes the occasional mistake. But my impression is that most people suggesting IDs for observations other than their own are doing so responsibly – i.e., their IDs reflect their familiarity with the taxon in question – regardless of whether they are specialists or not.

What is fairly common is for observations that are entered with a broad ID to go through several iterations of refinement: order to family to tribe or genus and only then to species. This is because many users who aren’t experts in a particular taxon nonetheless have enough general knowledge that they can help sort the observations into finer piles where the specialists will see them (e.g., Tineoidea experts). This is sometimes confusing for new users, who don’t know how to interpret the different IDs that they are getting and don’t understand why users are suggesting broad IDs. But these aren’t conflicting IDs.

Another thing that can happen is that if the observer uses a wildly wrong CV suggestion on their observation, they may get lots of IDs at different levels as users try to correct an ID that is often outside their area of expertise. So understanding the way the community ID is calculated and withdrawing one’s own ID if it proves to be wrong is also an important part of learning to use iNat.

And if users get conflicting IDs they don’t understand, it is always OK to ask the IDer(s) about it. It is fairly common for IDers to disagree without comment because our experience teaches us that a lot of users don’t care or may no longer be active on iNat. But I think most of us appreciate it when people take an interest in the taxa we care about and are happy to explain our IDs if asked.

4 Likes

It happens most of the time with the less familiar non-Papilionoidea species (the ones we call ‘moths’ but I prefer ‘not-butterflies’). Few records get to RG status and I suspect it’s because the real experts aren’t seeing them for some reason. Thanks to the new book ‘Southern African Moths and their Caterpillars’ by Staude, Griffiths and Picker, some users are able to get to the superfamily at least and if they are persistent, by using the ‘Afromoths’ site get to tribe or even genus. Sometimes an expert responds but often they don’t.

Hi nyoni-pete, I’ve created a project https://www.inaturalist.org/projects/african-lepidoptera that collects every one of the over half a million observations in Africa (more than Australasian fishes). Here it is https://www.inaturalist.org/projects/african-lepidoptera. What I don’t know how to do, is get it to open in that cool website format. Can you point me in the right direction?

just noting Steve, that since you specified ‘Lepidoptera’ as one of the taxonomic filters, there is no need to also list the superfamily children of that order as filters as well

1 Like

Yeah, I’ve worked that out. Umbrella projects with multiple collection projects for countries would be the answer there. Perhaps multiple umbrella projects for superfamilies, on top of collection projects at family or subfamily level.

Many of the butterfly and moth identifiers won’t read here.

Better you make a journal post on an iNat project or jour profile and tag the relevant identifiers