over the last years I observed that spider ID (at least in Europe) is only done by a comparably small group of people in iNat. Many of them aren’t specialists in the strict sense, i.e. people with a scientific background/publishing history. This holds true for many other arthropod groups as well, for example true bugs, crabs or even moths. Please don’t get me wrong. Most of the ID’s within the community are correct and I think the arachnological (and others) community here work astonishingly well given the many pitfalls of spider (and other arthropod groups) determination. Many of the naturalists/identifiers know a lot about the topic and the limitations of photo ID, and I myself learned a lot from them. However, because we are so few in comparison with the sheer amount of observations uploaded, I think several important topics don’t get the attention they deserve:
Pushing back wrong RG ID’s to the genus level. I could name at least 5 genera where this needs to be done and for which iNat contains several hundred (or even thousands) of wrong or doubtful ID’s
Identifications to the species group. For many amateurs, this is enough and keeps them motivated, while no ID’s/interactions demotivate them
Identifications with microscopic images. Several observers (specialists, but also photographers) upload records with microscopic images. Because we are so few with a strict taxonomic background, most of these remain at the “needs ID”-level. The “spider genitals”-project contains nearly 5000 (!!) such observations, most of them could probably be identified by specialists and often represent the first record of the species for the database. For a species I don’t see frequently or which is from a difficult species group, I need at least 10 min for identifications via images and literature to be fully sure. As my time is limited (maybe 30 mins a day for iNat), you can easily calculate why I only have 5000 ID’s, as I often focus on these records.
So, does anybody has an idea how to make iNat more attractive for spider/arthropod specialists? I thought about this for a long time. I’m probably one of only a handful people who still say something like this, but I didn’t came up with a good idea or strong opinion. I like iNat how it is structured, I like how it looks and how communication here works, and I don’t fully understand why many of my fellow colleagues (there are much more arachnologists than you may think) don’t look into it at least casually and help out with ID’s.
Do name them. Or write a journal post – which you can link to from this forum post, and from the obs where you have corrected the ID to Genus. A good and thoughtful explanation, written once, is worth the effort. You have a captive audience of potential interested identifiers here … and also the observers and other identifiers on those obs.
I have seen IDs where the identifier emailed a taxon specialist, then brought their answer to iNat. Not everyone is willing to do social media, it is not part of the ivory tower community.
These days I don’t think this is the main bottleneck, and therefore I don’t think that this should be the immediate goal. I think we mostly just need more laypeople to step up and share the burden of learning how to ID the common species and then doing so for others.
If someone does something professionally, getting paid to do so and in a structured environment where things seem to “count” – e.g. people think that physical specimen IDs “count”, peer-reviewed publications “count” – I think iNat can seem like it is too unstructured and therefore has low ROI.
One of the biggest frustrations that specialists deal with when coming to iNat is the quantity of low-quality identifications that they have to overturn. I’ve heard of multiple specialists expressing that iNat is a low quality dataset, or has inadequate taxonomy, or some related issue that makes it not worth bothering with. And this is difficult to argue with because it is true that if you start out on iNat as a specialists you will have disagreements with the current taxonomy and you will have to re-identify hundreds or thousands of observations and all of that will probably involve some frustrating conversations with observers, other identifiers, and curators who may well know less about the subject than you do. So there is some initial frustration but with a bit of effort those issues can be resolved to some extent and I think there can be a lot of mutual benefits once you’re over that hump.
Especially if you are a busy academic, as specialists often are, it can be difficult to make the case that any number of positive aspects can overcome that time investment. My strategy has been to try to dive into a difficult taxon myself, and if I get stuck or find any particularly exciting observations, I’ll email the specialist(s) and ask their thoughts, linking to the iNat obs. After a year or two of doing this every so often they start to recognize that there are cool data showing up on the platform that can’t be found any other way.
A common suggestion to make things easier for specialists is to modify the identification system in some way to give more authority to recognized specialists. There’s a long thread here on the subject, but generally people are pretty happy with the current egalitarian structure of the ID system. (And another long old wide-ranging thread here on how to attract more identifiers, for more ideas - the AI summary feature might help.)
I believe it takes a special kind of patience and pragmatism to put up with the nonsense that goes on in iNat. Either that, or you have to be faced with a choice between participating, and a much worse alternative (difficult as that may be to imagine).
For many people, the actual question would be “Why on earth would I want to subject myself to that?”
There ARE cool data showing up on iNaturalist, but there’s so much low-hanging fruit of common species with no obvious ROI for a specialist that I think it’s still hard to justify not waiting for someone else to deal with it sometimes? But that’s sort of what everyone does… waiting for others to make the IDs … it’s a tragedy of the commons, because most people just aren’t going to be interested in becoming identifiers. I can honestly relate to that quite a bit.
Like, as a taxon expert, I’m eager to go thru the observations that are already IDed as Mydidae in southeast Asia, but I’m not really willing to sort thru the unsorted-to-family observations for below 0.1% of interest.
I do make a lot of IDs personally, but I don’t think it’s justified by actual ROI for research, etc. I think it’s just fun, and as an observer myself, I know that it’s nice when things go from needs ID to RG.
Yes, 100%! There are a lot of taxa where IDing some of/most of the common species is pretty straight forward. And for the more ambitious, many more difficult taxa aren’t out of reach either given the right literature (which usually either exists for free on the internet somewhere, or someone with access is happy to provide).
I also think that if you work with a taxon for 40-ish hours a week, a lot of those people would rather spend their free time with something that is not that taxon. (There’s obviously still a lot of people who still ID despite that, and I respect that a lot, but I can understand anyone who doesn’t want to do that).
So, I think if the common species within that taxon would all be dealt with by laypeople, and only the “more interesting” stuff remains, that may also be a motivation to become more active.
The specialist has to “suck it up” and condescend to do some thankless grunt work
The specialist has a number of reliable “minions” who do the grunt work, and refer the more difficult IDs to their Lord and Master
Some time ago, I suggested that perhaps it would be useful if there was some way of flagging observations that require expert curation - perhaps another status (beyond Casual, Needs_ID, and Research_Grade), or a data quality flag.
This could have 2 benefits - it would make it easy for specialists to focus their limited resources where they are most needed, and also, it would deflect non-experts who might naively apply an incorrect ID (which would require even more effort to reverse/outvote).
I think we should also remember that someone can be a specialist on a particular organism (or group of organisms), and not necessarily be very good at identifying them in the iNat context. Even if they are good at identifying some narrow group, maybe that group inhabits some distant part of the world where there are few iNat users, and they’re handicapped when it comes to identifying local species.
Maybe they are experts on some narrow aspect of the group’s biology, and haven’t spent a lot of time sorting through field photos. A talented amateur might be able to run circles around the professional, precisely because they’ve spent countless hours honing their skills on iNat.
I’m guessing one of them is Dysdera. There are two species in Britain, distinguishable by the position of small spines on a leg. The UK’s National Biodiversity Network (a coalescence of biological records databases, some with more verification than others) has 676 records of D. crocata and 274 of D. erythrina. So the latter is less common but not orders of magnitude so. Whereas iNaturalist has over 1000 UK observations identified as D. crocata and 2 (yes, just two and neither of them RG) identified as D. erythrina. I could go through punting 100s of crocatas back to genus but spiders are not my speciality and its possible that someone who knows them well can tell them apart without seeing the leg spines.
Afterthought: One reason that D, crocata records are more numerous on the NBN could be because iNaturalist records are finding their way on to the NBN.
Theoretically the tagging system already works this way, except that it’s difficult to coordinate the associated ettiquette and so many tags to specialists are likely to be from impatient observers who want their common species confirmed to Research Grade within a few hours of upload. So far this kind of behaviour hasn’t been considered serious enough to warrant any more than annoyed comments from the tagged identifiers, but maybe as iNat scales this needs to be thought about more seriously…
But an observation field or traditional style project would also work for this function, the specialist(s) just need to be taught how to use them to filter observations. This system is already used by leafmine and gall identifiers fairly effectively.
Sure, but it relies on the folks who are looking at the observation knowing who the expert is, and it assumes that the expert can respond in quasi real time (if they don’t, the notification will quickly get buried under other notifications, which is a separate issue with iNat). Also, it targets/annoys a specific expert, and can’t be revoked. If the observation has a status/flag, it can be searched for and addressed by any expert who feels qualified, when they have time to do so. The onus is on the expert to look up observations of their preferred taxon/region that have been flagged as needing expert input - there’s no harassment. And if set inappropriately, the status/flag can be downvoted/revoked.
In theory yes, but the disadvantage is that those options don’t ward off the naive do-gooders, who are just trying to tidy up observations that they think are part of the needs_id backlog. If there was something like an “awaiting_expert_input” status, someone searching “needs_id” wouldn’t see them, and if they did see them, they might hesitate to provide their non-expert input. Yeah, yeah, I know, it smacks of elitism, but this thread is specifically about how to encourage experts. I think two things that discourage experts are:
not having an easy way of finding observations that require their expertise
having to struggle with “undoing” the well intentioned errors made by non-expert identifiers.
Note that in no way does this preclude the non-expert from participating. It’s entirely self-policing. The status/flag just indicates that somebody thinks “this is a tricky ID - you might want to be careful with this one unless you know the taxon reasonably well”. It should be easy to implement given that it’s just an extension of things iNat already has (observation status and data quality flags).
Another option is to somehow give expert IDs more weight (so that they can easily outvote non-expert IDs), but that’s WAY more elitist and WAY more difficult to implement. But it’s not unheard of on citizen science platforms - eButterfly does this. This doesn’t help with making difficult IDs easier for the expert to find, and I don’t advocate this approach.
Thank you for the many responses. I read them all, but I can’t cite all of them directly time-wise, so please excuse that my answer is a simple list.
Everybody who describes iNat as a “low-quality” database does simply either has a very normative (or maybe simplistic) view of biodiversity data in general or worked only casually with larger datasets. You just need to know how to use and categorize such biodiversity data for your research. The possibilities are endless. I can say, as somebody working with large datasets, that this is nonsense.
Academia: My knowledge of iNat etc is a true advantage for me career-wise. Maybe this needs to be communicated more. I always use iNat in my research, and I wrote papers based on online-available data. As an editor of Check List (Pensoft), I don’t accept manuscripts in which the authors did not search for additional records online.
I think there are some misconceptions here what most species experts really do at work. Only alpha-taxonomists or some freelancers in conservation really work all day on their group or talk about it. I’m an ecologist, and I spend my fair share of work identifying spiders (often tens of thousands at once), but I also have months in which I only do statistics, work with data from other groups, do database work or simply write manuscripts, reports etc. This holds true for most experts, especially in well-known regions like Central Europe. I actually come here because I can’t look at spiders enough at work and to search for rare records for the record schemes.
Frustration/discussions: That’s maybe the point that has the biggest impact over the long term. I received some very bigheaded answers on iNat from people who never identified a spider under a microscope. I sometimes wish for a stricter moderation of such answers, especially when already rebutted with literature. In some cases, members really think they know more than experts who worked on the group for decades and who explicitly state in papers they can’t be somatically determined. However, there are ways to deal with something like this. If somebody annoys with me with an uninformed opinion after I explained the ID in detail based on literature and morphology, I simply tag all my spider friends here and submerge the thread in genus identifications.
Tagging system. This is something that would help me personally a lot. I would like to help more with difficult ID’s especially with microscopic images (e.g. in Thomisidae or Philodromidae), but I often have a hard time finding them and not all of them are to be found in the spider genitals-project.
Special people: Yes, iNat can be a special place. But I worked at a Natural History Museum for 10 years. I would say iNat is a fairly normal community compared to what I experienced there.
@tobiasbauer I’m glad to hear there’s already a project relevant to your identifications; that’s exactly what I’m suggesting here. I’m guessing there aren’t very many observers who post microscopic images and that given that effort these observers are especially motivated to assist identifiers in identifying their specimens. So communicating with these observers (something like “Hey! I noticed you’ve posted some great spider observations; please consider adding your observations to this project to make it easier for me to find them in the future.”) would probably have good results.
@rcavasin yeah a new system would be necessary if it were to affects the Needs ID status (actually I think iNat used to have something like this many years ago before I joined, before the current Needs ID system was established?). In the meantime you could get something similar with an observation field if it’s with a taxon that doesn’t get many identifiers; then you can just tell all the relevant identifiers about the field and how it’s used and they can act accordingly. Something like “Specialist review:” and a dropdown menu with “Requested / In process / Completed” which the observer and identifiers can all modify as appropriate (they could also simultaneously use the DQA not identifiable further checkbox to take it out of Needs ID if that workflow works for everyone involved).
To make inaturalist more pleasant for identifiers (especially those who ID athropods), a new optional feature could be introduced to sort the observations in the ID module according to picture size. High resolution pictures can be identified to species level with greater probability, making the ID process less frustrating if you don’t need to look through dozens of low quality pictures
I’d trust the IDs of the experts and specialists a hell lot more than I do some random guy who’s trying to up his ID count (or who just has no clue but wants to try his luck), so not sure why there isn’t some way to distinguish the two and weight differently.
As to whether that’s “elitist” - I’d rather have a trained pilot at the helm of a plane than Joe the Guy at seat #23A
I would like to emphasize that the project (spider genitals) does not help bring in more experts. Indeed, the project shows that we lack them, as many observations remain unidentified. Personally, I think, with the answers here in mind, some things could bring in more experts.
A tagging system that doesn’t give expert ID’s more weight, but maybe puts certain record on hold before they become RG. This should only be used/accessible by curators for a small number of cases or certain groups (Like Araniella…). However, on the other hand this would mean that some people would need to be qualified as experts in an own category. Like I said before, I don’t have a strong opinion on this, as I hate this expert/non-expert categories. Many museum taxonomists no nothing about somatic determination. When implented, this specialist category needs to be a iNat-internal qualification, earned by activity, kind explanations and (but not always) external publications pertinent to the group, and can be rewoked. For example, a typical expert I ask for molluscs is Mathijs Zonneveld. It’s easy to see from his profile why this user could easily become an expert. Such people need to be rewarded in some way.
I often have to argue with people about certain groups like Trochosa or Araniella. Even when specimens look like a certain species, it does not mean it is 100%. This is something that annoys a lot of experts and keeps them away. A (by me) highly-valued expert from Brazil for example wrote this in his profile: "I published the taxonomic review and the phylogeny of Sidymella, Stephanopis and Epicadus based on morphological traits. I also revised Epicadinus, Coenypha and Paratobias, and proposed the new genera Spinaarachne and Kryptochroma. So, please give me some credit when I’ll try to ID the spiders you photographed." This tells you a little bit how many experts feel here when they try to help out.
Wrong RG ID’s. That’s something that annoys me personally, as I do a lot of modelling and the GBIF datasets are, in some cases, riddled with mistakes from all kind of plattform sources. Like I said, you need to handle these data and that does not make the iNat-data inherently bad (as this only holds true for certain very common species), however, when RG-grade data of certain species would have an additional column (like expert-input yes/no), I could easily re-download them here and sort them accordingly, hence, use them in macroecological and bieogeographical research.
(this is possibly not contributing to the main topic of the thread but) - with Dysdera crocata and erythrina, they live in different habitats. So say I have one under my sofa (which I regularly do) - it’s pretty much certain to be crocata, as erythrina lives in undisturbed natural habitats. If you punt my crocata back to genus I don’t think that’s truly warranted. And most of the observations on here will similarly be from amateur naturalists taking photos in ordinary places (which can be checked on the map). So maybe that’s why there are a lot of crocatas - they’re just more common and they’re in the places where most people observe. I don’t doubt there will be some misidentified. But if we insist on having to check their legs then there will be no records at all for people to use in their research. Is there not a happy medium, taking into account all the clues? (I’ve also read their colouring can lend some certainty too). Most identifiers probably use clues like range and habitat to discount rare but similar-looking species, for all sorts of species?
Anyway my feeling is yes, it would be great to have more people who have specialist knowledge to contribute. But they can only see as much as anyone in a photograph - they can’t stick the animal under a microscope and run it through a key. So that implies anyone who is suitably determined can seek out information and come to the same conclusion. I.e. they can develop expertise. If you know an identification is wrong, you can leave a kindly encouraging note to the effect that there are other species that it may easily be? This shares that information and that person can pass it on to the next (I’ve seen this happen). Eventually the ‘suggestion box’ can pick up differences if more correct ids are made. I just feel like the next generation of spider or whatever enthusiasts are out there and they’re on inaturalist. They need encouraging, then they’ll want to look into difficult groups more deeply themselves. Not be put off by being told no only specialists are allowed to identify things. I guess that’s not really what this thread is saying. But I think it’s important.
If there was some way of indicating that a species needs careful identification and better photos than the run-of-the-mill snap, then that would indeed be useful.
This is why adding notes to the observation is so important. Things like “observed among dust under a chair indoors” or “seen feeding on basswood leaves”, or even some zoomed-out shots showing the organism in situ can be useful for an ID. If the only picture is a closeup of the organism with an ambiguous background, no commentary, and an accuracy bubble that includes both the inside of a house and the outdoors, then it’s hard to warrant an ID based on context. I use the observer’s notes to assist in IDs every day. If you’re of the opinion that “If you punt my crocata back to genus I don’t think that’s truly warranted”, then make sure to explain why on the observation. I do find it frustrating when I kick something back to genus, then the observer comes back at me with a bunch of further information about context to support their original ID… Why wasn’t that information presented up front??