How many Arthropods are on iNaturalist as of now?

[insectilluminatigetshrekt]

https://forum.inaturalist.org/t/how-complete-is-arthropoda-on-inaturalist/57838

In November 2024, I made a post detailing the percent of species in each arthropod order that was observed on inaturalist. Since then, I have made several updates expanding on the data, such as comparing high vs low diversity orders, small vs large body size, and aquatic vs terrestrial. My last update, on April 29th 2025, was comparing arthropoda to other phyla.

But, I now want to investigate something different. How much has changed? Which taxa have gained the most additional species since then? Which have remain, and have some actually lost species? Lets find out. (If any mods see this, I prefer to keep this as a separate thread from the original post since this a review of how things have changed rather than additional random insights in the same time frame).

In total, the number of arthropod species observed on inaturalist increased from 237,817 on 04/29/2025 to 256,012 as of 01/27/2026. This is a gain of 18,195 species, or an increase of around 7.65%.

Subphylum	Observed Species (Circa April 29, 2025)	Observed Species (Circa January 27, 2026)	Number of Species Gained	Percent Increase
Chelicerata	16632	18351	1719	10.33549784
Crustacea	8463	9245	782	9.240222143
Hexapoda	210923	226416	15493	7.345334553
Myriapoda	1799	2000	201	11.17287382

As expected, the number of new species was the most in subphyla that already had the most species. But Myriapoda had the highest proportional increase, with its number of species on inaturalist increasing by 11%, and Hexapoda the lowest by 7%. This is simply because additional species cause a disproportionate impact if there are fewer species in a group to begin with.

86 orders of arthropods have gained at least one species since my last post on this subject, 38 have remain unchanged, and 2 orders, Ricinulei and Scalpellomorpha, have actually lost 1 and 2 species respectively. This is most likely due to synonymizations brought about by taxonomic changes.

59 orders experienced a percent increase in the number of observed species higher than that of arthropoda as a whole, and 26 orders experienced a percent increase lower than arthropoda as a whole. Orders showed a similar trend to subphyla, but even more extreme. The orders that experienced a huge percent increase in the number of species a taxa with very few species to begin with, such that even a small number of new additions can make make a disproportionate impact. The clearest example is the millipede order Stemmiulida, which had one species and now has two, experiencing a 200% increase.

Who are the biggest winners? Well, it depends on what you mean by winner. Of course, the orders which already had the most species have had the most species added, but it does not radically increase their percentage. Species poor orders increase a lot in proportion but still remain tiny in absolute terms. Well, I would argue there is one order that is the biggest winner.

In October of 2025, the user Efrain Chavez uploaded the first ever observations of the rare crustacean taxon Thermosbaenacea to Inaturalist, specifically of the species Tulumella unidens. This order previously had 0 observations on inaturalist, and now has two. I would say Thermosbaenacea is the biggest winner, graduating from absent to present on this site. Welcome to the club, thermal shrimp!

https://www.inaturalist.org/observations/323511879

Keep in mind though, this does not account for species which have been observed but not identified, which may well be in the tens of thousands. As always, these numbers are subject to change.

As a bonus, here is a map of observed arthropod species on inaturalist by country. As always, tropical under sampling rears its head.

[keeana]

i love ur posts sm

[evanwahm]

Fantastic breakdown!

Happy to have contributed the first iNaturalist observation of Procambarus epicyrtus, a crayfish endemic to Georgia, US this past year. https://www.inaturalist.org/observations/308315219

[DianaStuder]

From your map - there are over a million arthropods waiting in Africa for an ID. (Almost half in South Africa where we have identifiers but) Africa needs more arthropod observers.

[sebastiandoak]

Its nice that the rare find link on https://glauberramos.github.io/inat/achievements allows you to see all the species you are the solo observer of. Since the 29th of April last year (Which seems to be your period) I only have three

https://www.inaturalist.org/taxa/1530880-Prostygnus-vestitus
https://www.inaturalist.org/taxa/541343-Cynortoplus-albimaculatus
https://www.inaturalist.org/taxa/385178-Aka-westlandica

Though, I have plenty of harvestmen, which maybe one day might be figured to species? Though several are probably yet to be described. https://www.inaturalist.org/observations?lrank=genus&taxon_id=47367&user_id=sebastiandoak&verifiable=any

Can you tell how many newly described species have been added in your period, and how many of those have observations already.

Like the salticid genus Ourea was accepted last year, and has obs of 12/12 species.

https://www.inaturalist.org/observations?taxon_id=1643886&view=species

[jtkindt]

This was a pleasant diversion. I have no solo observations, but 3 or 4 firsts (one is, I think, ambiguous.)
I’m just as proud of my only first ID, which was confirmed by a leading expert in the field - if I hadn’t spent a half-hour doing research, maybe nobody would know that this bee had been recorded on iNat.

[edanko]

It’s not just tropical under sampling – even the observations that ARE made in tropical areas are simply harder to identify because there tend to be more cryptic species

[sebastiandoak]

Thats cool, the research side can certainly be challenging. Many of the descriptions of the harvestmen I am trying to ID are in german scanned documents from the early 1900s. Which isnt something I am so good as reading, and google lens translations are still dubious on biological terms.

[spiphany]

@sebastiandoak

german scanned documents from the early 1900s

As long as you’re not talking about huge amounts (and the texts aren’t printed in Fraktur), feel free to send me a message on iNat and I can take a look.

[insectilluminatigetshrekt]

I covered species, so now lets cover observations. (NOTE: Endeostigmata was lumped with Trombidiiformes in previous analysis, but I have now separated them, so if the numbers for Trombidiiformes look off that is why).

First off, several orders did not gain any observations. Ascothoracida, Bathynellacea, Cephalocarida, Mormonilloida, Mystacoacrida, and Tantulocarida have the same number of observations then that they do now, and Gelyelloida, Mictacea, Misophrioida, Palaeocopida, Platycopida, Platycoptioida, and Siphoniulida have no observations then and continue to have none now.

Percentage wise, the biggest winners are similar to last time, observation poor taxa where a small number of observations can add a large percentage to their existing pool. Though, unlike with species, in terms of observations nearly every order got a substantial increase. 69 orders got a proportional increase higher than for Arthropoda as a whole, and 50 orders got lower.

In terms of numbers, the orders with the most number of observations also got the most numbers added, no surprise. Though, if you noticed there is some weird things going on.

Pentastomida and Spelaeogriphacea each lost an observation, despite the former actually gaining one species. Most likely this is due to observations that were misidentified observations being corrected. This is especially likely for pentastomids, which may be confused with other parasitic worm like animals such as rhizocephalans or tapeworms. This is the opposite of what happened to Ricinulei and Iblomorpha, which each lost species (likely to taxonomic revisions) despite gaining observations.

23 orders gained new observations without gaining any new species. These are, in declining orders of number of new observations, Pollicipedomorpha, Xiphosura, Callipodida, Notostraca, Siphonophorida, Onychopoda, Calanticomorpha, Zoraptera, Verrucomorpha, Cyclestherida, Opilioacarida, Laevicaudata, Anaspidacea, Haplopoda, Holothyrida, Monstrilloida, Siphonocryptida, Canuelloida, Iblomorpha, Acrothoracica, Remipedia, Craterostigmormopha, and Facetotecta. In a few cases like Xiphosura and Craterostigmormopha, this is because we have genuienly observed every species in the group that exists, but in most cases this is almost certainly because many of these new observations have not been properly identified. Zygentoma is technically on the list too but it just gained a species today.

The over all ratio of both species and observations between orders has remained largely similar. There are no huge upsets like scorpions suddenly becoming the second most observed order or anything like that.

Subphylum	Observations (Circa April 29, 2025)	Observations (Circa January 29, 2026)	Observations Gained	Percent Increase
Chelicerata	7083118	8872102	1788984	25.25701252
Crustacea	1508698	1892749	384051	25.45579036
Hexapoda	62125492	78153510	16028018	25.79942224
Myriapoda	673848	837291	163443	24.25517327
ARTHROPODA	71391156	89755652	18364496	25.72376892

On the subphylum level, this can be seen even more strongly. Each one increased in observations by essentially the same amount.

Something that is clearly obvious from this whole thing is that for most arthropod orders and every subphylum, and thus for arthropods as a whole, the number of new observations is far outpacing the number of new species. Over 18 million new observations resulted in just over 18 thousand newly observed species. While many of those are of course observations of existing species, this alone cannot account for the massive discrepancy. Among these many millions of observations are surely at least tens of thousands of species yet to be officially even described, let alone added to inaturalist’s database. There is currently 256,015 species of arthropods on inaturalist, and there is 1,351976 extant species of arthropods according to Catalog of Life. Around 18.9% of described extant arthropods have been observed on Inaturalist, which is largely unchanged from last time and is actually a tiny bit lower than the previous number because more species have been described since.The ultimate take away from this is that we still have a very long way to go to observing even just all the known arthropod species, let alone the potentially millions yet to be described. Still, it’s amazing we got even this far! Thanks to all the wonderful observers and identifiers on inat who have contributed to these numbers as of yet. As always, the solution is to go out and find more bugs. Happy trails!

[sebastiandoak]

Cheers, yeah its quite a bit really, this sort of thing https://mndi.museunacional.ufrj.br//aracnologia/pdfliteratura/Roewer/Roewer%201928a%20WW%20II.pdf

But I appreciate the offer.

[spiphany]

Oh that’s splendid, complete with a liberal sprinkling of snarky comments about the mistakes and lack of expertise of colleagues and the poor state of Opiliones research.

[sebastiandoak]

Interesting to hear, to be fair, as hard as people are working on it, I feel Opiliones research is still in a poor state, even if moreso due to the lack of specialists, than thier expertise. At least in certain countries. In other countries is probably a lot further.

[sjl197]

@sebastiandoak

Interesting to hear, to be fair, as hard as people are working on it, I feel Opiliones research is still in a poor state,

Well, i’m interested in this general discussion for ‘state of the art’ for arthropods and the reflection on iNat. There’s lots of great information in this core post, but i’m not sure why there’s sidelines going on for issues about harvestmen and their literature.

@sebastiandoak I suggest you make an entirely new forum post just about the issues and concerns you have. I could’t help but smile when you said “Opiliones research is still in a poor state” as in last years i’ve tried to become mindful of ‘state of the art’ for diverse arthropod groups and actually i’ve become dismayed about how research on many other arthropod lineages seem absolutely dismal compared to Opiliones and such others!

You linked to a pdf in German from 1927 where you struggle with content?
(1) The entire literature on every taxonomic act on Opiliones is accessible online, the majority as open access through that website hosting that pdf. Whilst imperfect, that and others are available and usually searchable with OCR, etc.
(2) The entire history of taxonomic actions for Opiliones is cataloged and reflected online via WCO-lite and reflected on Catalog of Life. That pulls in other databases which adopt taxonomy, and any discrepancies have been essentially cleared by dialog. If the CoL ever manage to finalise their long promised data exchange with GBIF, then that can beomce inline also.
(3) A companion document was published with a complete bibliography, overview of approaches to resolve historic inconsistencies etc - see ReserchGate “WCO-Lite: online world catalogue of harvestmen (Arachnida, Opiliones). Version 1.0”
(4) A summary checklist of all taxa published as a static reference point by Kury et al. 2025 “Checklists of extant harvestman (Arachnida: Opiliones) species for all the countries of the world” with discussion of works in relation to biogeography.
(5) [then generally] Multiple modern studies recently published in peer reviewed journals by many different authors, increasingly combining morphology (plus more detailed morphometrics) with genomics, biogeography, behavioral ecology etc.

Let me say, is there still much to do and improve, yes! But in this context if you need me to highlight the many other arthropod lineages where the active researchers can’t even be bothered to do what seem like the basics of making a consenus list of the described species in their focal family, tribe, etc, i can assure you my list will be long! You want access to the literature? You want to know all the synonyms to understand the literature trail to get to a name combination? You want data on biogeography etc?! Ha. Haaaaaa. >> Sigh!<<

Anyway, sorry - none of what i said directly addressed the concern. Absolutely right that even with access to detailed literatire then it’s often a whole other game to extract something meaningful, such as for making identifications. Here then, i’ll direct to cross-linkage onto wikispecies, where myself and others have gradually building the data to reflect the other authoritative sources. In general papers before 2025 are available as open access via BHL, and sometimes the quality is better than on the link where you found the other. It can be better for searching, translation etc. See e.g. https://species.wikimedia.org/wiki/Carl_Friedrich_Roewer
Secondly, i see amongst your observations you gathered some valuable records some South American ones. Over there, cosmetids are a diverse group and much of that above linked work by Roewer 1927/1928 is on cosmetids. I’d put a smattering of Roewer translations onto a “Opiliones Wiki” (fandom page) which was [to my view at least] something of a testing ground how to revise info on harvestmen via a wiki based framework [https://opiliones.fandom.com/wiki/Cynorta_coxalis example page]. Much info there is outdated now after a few years, other initiatives took over, but moving from there, in other context then i think there’s really still scope to get many of the historic descriptions translated and updated in modern web-searchable format. In that, i’m saying i’ve likely already tried what you’re struggling with for translations - and i’ll say in this context if you think his 1927/28 works are challenging, prior to this in 1923 he wrote a massive work where he went rampant with abbreviation, then requires knowledge of the attributes but also what his abbreviations mean, mostly undefined. “To mit 1 Dö-Paar u. davor mit 1 Zch-Paar ,..” << Google lens probably doesn’t have a translation setting for “Roewer circa 1923!”

[sjl197]

@insectilluminatigetshrekt

There is currently 256,015 species of arthropods on inaturalist, and there is 1,351976 extant species of arthropods according to Catalog of Life. Around 18.9% of described extant arthropods have been observed on Inaturalist, which is largely unchanged from last time

To now go into general issues for the overview, citing some statistics. Those are interesting numbers, i knew for CoL, but i had not found a way to total what’s already on iNat, so that’s interesting. When you say “256,015” then sounds like that’s available names at least one suggested observation? There will of course be a good few that are nonsense (someone once tagged a name!) but it’s good to see fraction. However when the numbers are gauged against CoL, then i’m sorry - then to my anecdotal view, the number of taxa on CoL is still a rather terrible reflection of the state of knowledge for many arthropod lineages. In particular, vast swathes of described beetle especially are missing. Many other lineages are relying on data from decades ago. I hate to say this way, as expect may sound like my view negates their massive efforts. Actually, i’d say what they’ve achieved is remarkably impressive, the closest approximation to a comprehensive listing of species that i’ve seen in my last 30 years of interest. Any who built into that are to be commended.

I do ask caution and realism though. 19% of a subset of the unknown is not 19%. Let’s not be blind to the vastness of invertebrate biodiversity, especially arthropods. Here also to be clear, i’m not aiming to be speculative about the actual unknown arthropod biodiversity, e.g. “this one tropical tree contains more arthropod species than all the known species of mammals and birds combined”. What i really wish to highlight instead is that CoL or any yardstick still seems to fall short of being a comprehensive catalog of what’s ALREADY described so far for arthropods! Maybe in another 10 years we might know what everyone said in the last 300 or so years of Linnean based taxonomy? Maybe by then, much of both the unknown and known biodiversity will be lost anyway via habitat destruction and such, shouldn’t be a problem!

[insectilluminatigetshrekt]

@sjl197

To now go into general issues for the overview, citing some statistics. Those are interesting numbers, i knew for CoL, but i had not found a way to total what’s already on iNat, so that’s interesting. When you say “256,015” then sounds like that’s available names at least one suggested observation?

These are the number of species with at least one verifiable observation on inaturalist. The numbers have increased in the days since making this post, as expected.

@sjl197

However when the numbers are gauged against CoL, then i’m sorry - then to my anecdotal view, the number of taxa on CoL is still a rather terrible reflection of the state of knowledge for many arthropod lineages.

It’s the closest thing there is to an at least somewhat comprehensive list of known species. No doubt many species are missing, so the numbers on CoL should be taken as a conservative under estimate. There is also GBIF but you can’t filter to sort out extinct species on there I don’t think.

@sjl197

In particular, vast swathes of described beetle especially are missing

Funny you say that, I find mites to be the worst sampled group on CoL. It seems around half of known extant species of mites are on CoL.

@sjl197

I do ask caution and realism though. 19% of a subset of the unknown is not 19%. Let’s not be blind to the vastness of invertebrate biodiversity, especially arthropods.

I made it very clear at the end of my post that there is of course vast swaths of undiscovered arthropods. Again, the 19% number should be taken as a percent of a conservative estimate of known species.

@sjl197

“this one tropical tree contains more arthropod species than all the known species of mammals and birds combined”.

Tbh I’ve always mildly disliked this approach to estimating arthropod biodiversity since it treats everything that isn’t a terrestrial herbivore as irrelevant. Extrapolating from tropical canopy beetles says nothing about mites, copepods, etc so you cannot estimate the biodiversity of “arthropods” from this.

[sebastiandoak]

Prehaps you read my post a little more literally than the intention.

Rower in that paper mentions many frustrations, and no doubt you would be aware that many of the frustrations still exist 100 years later, even with the wealth of tech we have, and the number of specific experts. That was essentially my point.

Of course there has been a lot of amazing work since, and I am massively appreciate of the many resources I have used from the omnipaper site, likewise of the skills of a small number of Iders who have provided Ids with links of species I am currently the only Inat observer of. Tsssss has really helped me there.

There is a massive wealth of information, across a multitude of sources, and I appreciate the several specialists who I have been able to talk with on some different groups. (Kury, Giribet, Derkarabatian). Here in NZ after Forster, there was limited work, as is often the case, but really appreciate the work of Powell / Taylor.

Still I hope kury doesnt mind if I quote him as saying “The taxonomy of Peruvian harvestmen is in a state of chaos. […] The problem is that the literature on Opiliones is horrible—very difficult to access and understand—and perhaps 70–80% of the species remain undescribed “

Based on my knowledge of his work, I am prone to trust him, because the omnipaper project is one of my go to resources, especially when many other resources I try and use arent so open access. When he mentions several of my pics hes looked at are probly undescribed, I know thats not a guarentee, but I trust his level of knowledge of South American Harvestmen.

If you have been working on translations, thats awesome. I have tried machine translations, and they definitly seem to struggle, and when they are good, I am not sure enough of certain things, to know they are good.

Still I wouldnt say it needs a new post, as my point was never about harvestmen, as much as the same point you make, in relation to this threads topic. That we simply dont have enough dedicted experts to know all species of everything yet.

In the past couple years, I have helped with base sorting of pitfall collections. With certain groups, of organisms being sent to specialists who are available to accept specimens and do the work on them. Like many projects its limited by budget, person hours, and expertise. Though we have had several new species due to having the likes of a amphipod specialist. I am definitly not up enough on the state of amphipods to speculate on the state of their research. Still as much as I love these critters I am more of a passioned hobbiest, trying to help where I can than any specific taxonomist. Sure I engage in some level of paid taxonomic work, and as such would tell people I do some taxonomy. But to any actual taxonomist, I am but a plebeian. In my explorations of South America, I try to photograph as many harvestmen species as possible. I would love to ID them all, but at my level, much of it is still just pointing out what I see.

I do love that behavioural work is starting to come more into play. That was what I specialised in when I was at uni.

Also if you are doing the work on the fandom, much appreciated, I have certainly referenced it for certain species.

[spiphany]

@insectilluminatigetshrekt

for most arthropod orders and every subphylum, and thus for arthropods as a whole, the number of new observations is far outpacing the number of new species. While many of those are of course observations of existing species, this alone cannot account for the massive discrepancy.

Actually I think it does pretty much explain the discrepancy. It is much the same as the fact that the growth of one’s personal species count or the species count at a site is not going to resemble a line as much as an asymptote – the more species one has already observed, the more difficult it is to find new ones. It is always going to be easy to make observations of common, readily identifiable species. People are not going to stop observing these species just because they are already represented on iNat.

The remaining as-yet undocumented species are going to fall into one of several categories: a) rare; b) present only in regions that are remote/inaccessible, or with extremely limited ranges; c) small/difficult to find because of habitat preferences; d) difficult to identify from field photographs. This means that the numbers of such observations are always going to be fairly low. In addition, it is likely that there are some species already represented by iNat observations that merely have not been identified as such, either because the evidence is insufficient to distinguish them from other species or because there are no experts on iNat with the skills to identify that taxon. Common wisdom in Germany is that only about 30-50% of bee species can be identified in field by an expert; this is probably roughly similar to the number that can be identified from the average iNat photo. To get the remainder represented on iNat, you therefore need people who are posting photos that are better than the average iNat photo (e.g. high-quality specimen photos) and someone who has the skills to identify them. I’ve also watched several taxa get a major “bump” in the number of species represented when someone joined iNat who was knowledgeable about the taxon or decided to dedicate themselves to learning it. But this requires an influx not of observers but of knowledge, which is a bit more difficult to predict or promote.

[insectilluminatigetshrekt]

@spiphany

In addition, it is likely that there are some species already represented by iNat observations that merely have not been identified as such

This is exactly what I was referring to though. Of those over 18 million new observations, it is plausible there is at least tens of thousands of unidentified inat firsts among them.

[spiphany]

I don’t think it would be accurate to call those “firsts” though – for difficult to ID but common species, there are probably already dozens of observations that simply lack sufficient detail to be identified. The identifiability from average photos is also not going to change substantially over time (except perhaps if IDers develop new ways to distinguish some species that don’t require specimens). So this is going to be a constant factor that won’t be reflected in growth of total numbers of species on iNat.