We passed 300,000 species observed on iNaturalist!!

It’s interesting to compare this number with the total numbers of species that we think are out there. Most agree that there are around 2 million species with names and that many more species exist that haven’t yet been named. We use IUCN’s numbers from 2010 which tally 1,740,330 species of plants, animals and fungi. Using this denominator, iNaturalist has now censused about 17% of all named species.

The figure below shows this split out by taxonomic category. Each square represents 1,000 species. There are 1,740 total squares and 300 green squares representing the subset observed on iNaturalist.

Most of the species on iNaturalist represent plants and insects but these are also the most speciose groups. In fact, even though the same number of plant and insect species have been observed on iNaturalist (roughly 100,000 each), this represents a much smaller fraction of the total insect diversity (11%) versus the total plant diversity (33%).

The group with the smallest percentage of species observed was Arachnids (8%) and the four terrestrial vertebrate groups (birds, reptiles, amphibians, and mammals) were the only groups with greater than 50% observed. The denominators are probably underestimated as new species have been described since these IUCN numbers were released in 2010, but these percentages are likely not too far off.

In addition to thinking about how many species we’ve observed on iNaturalist, it’s interesting to consider how quickly we’re accumulating new species. The graphs below show the numbers of species observed on iNaturalist over time. The graph on the left is for North American Birds and the graph on the right is for South American Fish. For North American Birds, we do seem to be reaching a plateau (i.e. we've running out of missing species). In fact, over the last 12 months we’ve averaged 19,696 North American Bird observations for each new species tallied. In contrast, for South American Fish the number of species does not seem to be plateauing at all. Over the last 12 months we’ve tallied a new species for every 14 observations of South American Fish added on average.

Here’s the species groups and continents (e.g. South American Fish) where we’ve averaged fewer than 100 observations to tally a new species sorted by this obs/sp stat. Underrepresented groups like fishes and mollusks are over represented in this subset as are continents like South America and Africa.

These are the species groups and continents where we’ve averaged more than 100 and fewer than 1,000 observations to tally a new species.

Lastly, these are the species groups and continents where we’re averaging over 1,000 observations to log a new species. Here, North America and Europe dominate alongside over represented groups like birds and other terrestrial vertebrates.

To tally a new species on iNaturalist, several things have to happen: we need observations from that taxonomic group and location to be posted, the images must reveal enough detail for the species to be identified, and someone with the skills, time, and interest needs to provide an identification. There are steps you can do at each point in this process to increase the overall rate.

To generate more observations, spend some time observing taxa in places you don’t normally visit. If you focus mainly on terrestrial vertebrates, try observing some fish and invertebrates. Invest in some gear to attract or capture creatures you don’t normally encounter like a moth light or aquatic sampling gear or a macro lens for your phone. Try to get others observing by organizing a bioblitz.

To increase the probability your observations will be identified, do your best to take identifiable photos. If you are an identifier specializing in a certain group, write a journal post to encourage people to pay attention to the characters you know are important. Also, engage with observers. Often they can relocate the critter and take photos of the characters you want as in this exchange between @agapakisnikos and @naufalurfi:

Lastly, to help grow the expertise needed to identify species, consider investing in becoming a more active identifier yourself. Pick a location or taxonomic group (or both) to specialize in. Dig into the scientific literature and learn the jargon for the museum characters the literature tends to focus on. Learn how to follow a dichotomous key and interpret a taxonomic description for your group. Sites like research gate and sci-hub provide access to an increasing number of papers. Don’t be afraid to learn from the community. Ask identifiers how they were able to identify particular observations you want to learn from. If they provide information and references try to understand them and see if you come to the same conclusion. If you can’t, explain where you’re stuck and ask for help. Encourage people you know with expertise to join iNaturalist and share their knowledge.

Of the 54,000 datasets tracked by the Global Biodiversity Information Facility, the most speciose is the Smithsonian Natural History Museum dataset with about 500,000 species represented. Based on the rates of growth in new species on iNaturalist and the relatively large percentages of species we’re still missing for most locations and groups, I expect we will also pass this 500,000 milestone in the next two years or so. But as more and more of the common species are observed, maintaining the growth in new species will rely on adventurous observers and identifiers working together reveal the identity of increasingly rarely encountered and poorly known groups of organisms. If you have other ideas for how to invigorate this process, we’d love to hear them!

Posted by loarie loarie, October 13, 2020 18:04

Comments

Thumb

Wow, what an amazing milestone! I recently passed 3000, which is just 1% of that.

Posted by alexis_orion 7 months ago (Flag)
Thumb

I have only 2000, which is so little of all the species! Hoping to get so many more! Thanks so much for this milestone!

Posted by yayemaster 7 months ago (Flag)
Thumb

i have less then 600 observer but In my list species there is 3 new species, proud of it

Posted by salimeh 7 months ago (Flag)
Thumb

Would love to see more from Saint Helena, South Atlantic Ocean. They are replanting the forest, and it will be interesting to see if plants return. if insects emerge from tiny surviving populations. But their internet is painfully slow.

Posted by dianastuder 7 months ago (Flag)
Thumb

The first graphic came just when I needed it. Thanks for sharing!

Posted by mich_croc 7 months ago (Flag)
Thumb

I'm curious what the 2% of remaining bird species are.

Posted by joe_fish 7 months ago (Flag)
Thumb

you can see a subset of them by going to the bird taxon page in the "Wanted" section of the "Trends" tab (also its more like 8% rather than 2% if we use the Clements denominator rather than the IUCN denominator - but I decided to simplify things by just using IUCN stats consistently in the post)

Posted by loarie 7 months ago (Flag)
Thumb

Do we have breakdown by country?
So southern Africa has contributed 31,400 species (leaf "species", or 29.300 "taxon" species) or 10% to the total species (which I believe is the same as our proportion contribution to the global species richness, at least in plants).

I am curious how well the other megadiverse regions are contributing to the total? Targeting "expeditions" to these regions is likely to get us observations of many new species, the problem is finding specialists to identify them (and taxonomists to typify and name those still undescribed).

Posted by tonyrebelo 7 months ago (Flag)
Thumb

I’d love to learn better methods of observing invertebrates.

Posted by jaquiring 7 months ago (Flag)
Thumb

Most impressive !

Posted by jtch 7 months ago (Flag)
Thumb

Hey @loarie, that's a neat summary of iNat's current taxonomic coverage, thanks a bunch! Shout out over the Atlantic, hope you're fine!

Posted by jakob 7 months ago (Flag)
Thumb

thanks jakob - and thanks for being one of the early experts on iNat - you are pretty much singlehandedly credited with identifications from the continent of Africa for the first 5 years or so!

tonyrebelo - I love the idea of expeditions - but how to coordinate? I agree that finding ways to engage specialists is key and tricky

Posted by loarie 7 months ago (Flag)
Thumb

@loarie @tonyrebelo an iNaturalist scholarship to visit remote places?? I'm willing to be the guinea pig :D

Posted by thebeachcomber 7 months ago (Flag)
Thumb

The graph for North American arachnids is mislabelled as "S. America". Fixed!

Posted by jeremyhussell 7 months ago (Flag)
Thumb

Was thinking something more along the lines of the CNC or GSB, but in remoter areas. Sponsors - either accompanying or remote. A meeting venue for those interested at a specific date (anyone welcome, but pay your way there). A 7-10 (-14) day period. Using local ecotourist guides to visit hotspots or target iconic species, over half a dozen themed routes, to be planned by local biologists/guides. Bioblitz everything that lives. Either provide roving wifi hotspots, or support to upload while asleep. Get local communities to help meet checklisted targets. Mobilizing specialists, taxonomists, student to help with IDs, at both international institutions, but local institutions as well: and also to annotate or comment on observations or special local interest or biology. The iNat community following postings real time, and helping out remotely with any snags, and discussing observations, and providing debrief and evaluation.

Spinoffs: good species and distributional data, global awareness of local tourism potential, local exposure to specialized ecotourists, local community with exposure to iNaturalist, a fun live adventure for the entire iNat community.

Posted by tonyrebelo 7 months ago (Flag)
Thumb

I love this. Nice job Scott. I encourage everybody that, with some effort, you can do this for your own areas/places too. For example, in Alaska, one new taxon was recorded about every 56 iNaturalist observations in 2019. 99% of resident and regularly occurring Alaska birds have been observed, and <12% of Alaskan insect and arachnid species have been observed and identified on iNaturalist. Combining all the state checklists I could find, iNaturalist observers have documented around 3,324 unique taxa, or approximately 25% of the combined checklist species total of 13,205 species.

Posted by muir 7 months ago (Flag)
Thumb
Posted by loarie 7 months ago (Flag)
Thumb

very cool muir - I hadn't thought of the place specific emphasis - your Alaska analyses definitely a great example of that

Posted by loarie 7 months ago (Flag)
Thumb

Thank you for featuring my observation, @loarie!
I do agree we need more identifiers of underrepresented taxa. The majority of my observations from Indonesia of these taxa goes unidentified but when they do, they often are species new to iNat. Got rather plenty of these first on iNat observations.
Also if youre an expert, please do identify observations from Indonesia. We desperately need identifiers here!

Posted by naufalurfi 7 months ago (Flag)
Thumb

@loarie @tonyrebelo that idea sounds great! Even less ambitious expeditions for those of us in well-observed areas would be neat, for example there are a lot of gaping holes without any iNat observations in the mountains of West Virginia.

I've seen an exponential growth in the expert resources I can access on iNaturalist for Diptera recently. There are more experts that have become active identifiers recently than I can keep track of, not to mention other experts that come onto iNat and make identifications only when prompted by email. And it seems like @phycus , among others, knows who to ask off-line for almost anything. I'm not sure it would be wise to create a publically available listing of experts to ID iNat observations, but maybe this is something a few of us should start compiling for individual groups like the Diptera, with individual experts' consent to be included?

Posted by edanko 7 months ago (Flag)
Thumb

Is there a way to see, for a given project, area, time period, or user, species that have never otherwise been observed on iNaturalist? This would be interesting to examine for a project like https://www.inaturalist.org/projects/2017-inat-athon-southeast-arizona and I know that local CNC organizers would use this a lot as well.

Posted by edanko 7 months ago (Flag)
Thumb

@edanko - it seems to me that the only way you can determine what has not been seen on iNaturalist in an area, time, etc. is to have an independent checklist. And then see what is not on the list. For instance, these Monocots have not been recorded on iNaturalist yet in Cape Town:
https://www.inaturalist.org/check_lists/1634623-Master-Plant-List-for-CoCT?observed=f&page=5&q=&rank=species&taxon=47163&view=photo
There are also tools for comparing lists, where lists can be projects or any filter you desire:
So species Cape Town dipped out on for the Great Southern Bioblitz versus the City Nature Challenge can be seen here: Vertebrates CoCT dipped
The problem with these lists is that many of these species are peripheral and only just get into the area, so they may be very well recorded just outside of the area (for any time, user, etc).
There must be a way to see what is in the iNaturalist dictionary that has not been observed (i.e. observations = 0), but I have not yet discovered how to do so as a url I can send to other observers. The problem of course is that one needs a checklist or atlas for the area to compare it too: iNat somehow needs to know what occurs in Arizona or Cape Town to know what has not been seen there for a user or time or place or project.

But a summary of species not yet observed on iNaturalist (globally as wild) from a place checklist would be a really good way to incentivize locals.

Posted by tonyrebelo 7 months ago (Flag)
Thumb

@tonyrebelo @loarie I like the idea very much as well, and I'd be happy to see it happening. As it helps with biodiversity inventories, some international societies such as the SCB (and it's many chapters) could also be interested in helping things out. Regading having experts able to make time to join, facilitating with research permits so that it can result in scientific publication could help.

Posted by amarzee 7 months ago (Flag)
Thumb

I just love these sorts of posts! They make me wish I lived in a less well-studied area than Massachusetts in the US or, at least, was able to travel right now. But one of these years, I'll get to the mountains of West Virginia - that's easy enough - and, with luck, I'll get to visit my sister in Australia someday, too. In the meantime, I could go iNaturalize the blank spot in Massachusetts (the town of Granville, essentially) or figure out which plants are still needed for Massachusetts or really dig into lichens, the way I've been meaning to for a while now, or start harassing, er, observing the inverts in my local ponds, and so on. There's no end to what can be done. As far as I know, I've only added one new species to iNaturalist in close to 10,000 observations overall. Surely, there are more to be observed, even here!

Posted by lynnharper 7 months ago (Flag)
Thumb

Does anybody know what general % of arthropods are identifiable from photos? So, if the global iNat community has observed and IDed ~10% of insects, spiders, custaceans, etc., what can reasonably be expected to be identified in the remaining 90% from the information available in most iNat observations?

Posted by muir 7 months ago (Flag)
Thumb

There are a lot of arthropod groups that can be IDed from photos but only by a serious expert, because the literature is based on bristles and other things that are easier to see in preserved specimens. Untangling these from the species that actually can't be separated is going to be difficult.

Posted by edanko 7 months ago (Flag)
Thumb

Great! Is there any data on the coverage at different taxonomic levels? (genera, families, orders etc.)

Posted by splendiddesolation 7 months ago (Flag)
Thumb

For branches covered by complete taxon frameworks down to species (which includes all vertebrates, some insect branches like Odonata, spiders, and some marine invert and plant groups), then if you go to any taxon page on that branch (family genus etc. - e.g. butterflyfishes) you'll see the total species observed ratio (e.g. 117 of 132) and in the 'trends' tab you'll see recent 'discoveries' (when new species were added such as Wrought Iron Butterflyfish added 4 months ago) and a subset of 'wanted' species (ones we're still missing such as Pete Basabe's Butterflyfish) - explained in a bit more detail here.

Posted by loarie 7 months ago (Flag)
Thumb

Is there a url or widget for the "Wanted" species and does it only work at generic level?

Posted by tonyrebelo 7 months ago (Flag)
Thumb

there's an undocumented api endpoint behind that 'wanted' carosel that takes any taxon and returns species level descendants that haven't been observed

Posted by loarie 7 months ago (Flag)
Thumb

Way to go 300000 species!

Posted by rangermyles 7 months ago (Flag)
Thumb

I love graphs.

Posted by predomalpha 7 months ago (Flag)
Thumb

Mollusks 13% -- lots of work to do!

Most of the smallest ones are missing.

Posted by susanhewitt 7 months ago (Flag)
Thumb

@susanhewitt In New Zealand, only 832 of the 3000 or so described species have been recorded. Thats still pretty good, nearly 1/3.

Posted by predomalpha 7 months ago (Flag)
Thumb

@predomalpha. --that's really impressive! Well done Kiwis!

Posted by susanhewitt 7 months ago (Flag)
Thumb

@tonyrebelo @loarie Really loving this idea, especially doing something along the lines of the CNC. Instead of having a single trip, there could be a dozen or so locations set around the world to maximize participation and findings. They don't all have to be at the same time like the CNC but could occur during the time of year with the most biodiversity in each respective location. This way, hotspots of biodiversity can be targeted in an efficient manner while maximizing the fun.

Posted by rynaturalist 7 months ago (Flag)
Thumb

Love this post!

Posted by calloftheloon 7 months ago (Flag)
Thumb

Hi, I'm a biologist in Rio de Janeiro and I joined the inaturalist 1 year ago. Every day I am impressed with the quality of the information on the site. It is sensational. I'm happy to interact with people who like to share images and knowledge about biodiversity from all over the world. I think the idea of @loarie @tonyrebelo to promote trips to isolated places and little sampled in the inaturalist is a genius idea. Super would love to participate, I have a lot of experience in the field, for being a biologist and passionate about registering biodiversity. Here in Brazil for example, we have many places with few records, either in the Amazon, Atlantic Forest, Cerrado and so on. I have been committed to register the fauna and flora of Rio de Janeiro and the Inaturalist has been an excellent place to identify the species. And excuse my bad English, hehehe

Posted by diogoluiz 7 months ago (Flag)
Thumb

Sorry, @loarie I was not clear. I meant: are there data on how many families/genera/orders etc. have been observed? Did we observe all the families of insects, for example?

Posted by splendiddesolation 7 months ago (Flag)
Thumb

I wish you wouldn't encourage amateurs to go out and "capture". Please don't.

Posted by ellen5 7 months ago (Flag)
Thumb

Impressive numbers !
To further increase it might also be worthwhile to think about how we can help experts to find "interesting" observations, it can be frustrating to scan through hundreds and thousand of observations of the same most common and easily identifiable taxa uploaded by the same users again and again (there might be valid reasons to do so) just to spot the few interesting findings... it might be a checkbox similar to wild/captive, it would mark observations

that have a good picture quality allowing for an identification
that are not common and possibly needs some expertise to identify
Potentially just a crazy idea...

However, as South American fish are explicitly mentioned as underrecorded, I'm even more motivated to check in my older picture archives if I can find some more... I definitely have quite a few undescribed ones to add and here we definitely do not have the problem of too many observations hiding the interesting ones :-)

Posted by karsten_s 7 months ago (Flag)
Thumb

"how we can help experts to find "interesting" observations" - I think this is a matter of networking. I tag experts I think would find an observation interesting, and others tag me.

This also requires volunteers to make IDs other than the experts themselves, for groups with more than a handful of observations. I don't think it's fair to expert the average expert to spend much time on iNaturalist, many of them have extensive responsibilities.

Posted by edanko 7 months ago (Flag)
Thumb

I'm doing basically the same, for observations with a reasonable picture quality (for which I think they could be identifiable) I'm tagging the experts that I know, if I know them. However, there are many cases where I don't know the expert but I see high quality pictures of some non-common organism, the experts are not responding (perhaps they are tagged to often)...
And quite often also I'm tagged and I'm not the right one (which is not really a problem) and sometimes I also tag the wrong ones...
As said I'm not sure if this is the right approach, but the time of the experts is very valuable and this should not be wasted with "non-interesting" observations IMHO.

Posted by karsten_s 7 months ago (Flag)
Thumb

Except that there is the real problem that the field guides show the common species, and the AI is trained on the common species, so the rarer and more exciting species often get misidentified as the more common species. And the undescribed species also tend to end up there.
So, as an expert in some groups, I enjoy trawling through the dross looking for gems. The Identify tool (really a curation tool), is really superb in this regard! The real challenge is not finding experts, but getting them to learn the tools that will make their work easier, more efficient and great fun. Otherwise this site it totally overwhelming! So: 300k species in 50M observations? Where does one begin?

Posted by tonyrebelo 7 months ago (Flag)
Thumb

Thats an amazing number!

Posted by ajott 7 months ago (Flag)
Thumb

Anyone who has good close-up ability, a really good macro set-up, I recommend that when they go to a beach pretty much anywhere really, that they should to try to find an area of beach drift that has a lot of very small shells, tiny shells, in it. And to photograph those teeny tiny shells one by one.

We are sorely lacking in the minute mollusk species, and they greatly outnumber the regular-sized ones.

I am not necessarily talking about microscopic shells, just anything from half an inch down would be really welcome!

Same is true when looking in flood debris along rivers, and finding live tiny land snails under logs or stones etc.

Posted by susanhewitt 7 months ago (Flag)
Thumb

Will be at a beach next week and will pay attention to the tiny snails... anyways I seem to photograph more molluscs here in egypt then I would have expected in a desert country

Posted by ajott 7 months ago (Flag)
Thumb

OK, good luck with it! Tiny snails and tiny clams are good too.

Posted by susanhewitt 7 months ago (Flag)
Thumb

I wish I had access to the coast... any similar tiny snails in the Great Lakes, @susanhewitt?

Posted by mws 7 months ago (Flag)
Thumb

There will be tiny land snails living in the land areas surrounding the Great Lakes. And you may have Pomatopyrgus jenkinsii in some small freshwater habitats, as well as probably some very small planorbids and maybe some tiny freshwater limpets.

But I don't know about tiny snails in the Great Lakes themselves.

Posted by susanhewitt 7 months ago (Flag)
Thumb

Thanks for the info! Time to get that mollusc number up

Posted by mws 7 months ago (Flag)
Thumb

@mws There are many minute molluscs everywhere. Look for a spot of native bush and get a bucket of leaf litter from a damp, shaded spot and take it home and have a close look through it with a magnifying glass. This should yield many minute species, under 10mm.
I do not know the names of the ones from the USA though.
It might also take some practice to figure out the right spots and the right ways to sort through that. When you are finished, you can use a USB microscope, a really good macro camera or an actual microscope to take some good photos of it. Try to take photos as similar to @openlabnz 's photos. I expect many of the species will be extremely difficult to identify, so you could send some photos to experts to assist. Even so, many will only be brought to genus or family level. For instance, Punctidae has an unknown number of pretty much identical shells, with the only difference being genetic.
If you wish to get even better identification, you can write a label showing the location, date, habitat description, collection method and name of the collector(s) and place the snails in 94% pure ethanol. This will prevent any DNA from decaying and it can be stored properly for centuries to come. This should be sent to a museum or university, which both have the necessary tools to properly identify them. Chances are, you'll find a few unnamed species, and maybe even species new to science. I certainly have found a few in the past 2 years.

Posted by predomalpha 7 months ago (Flag)
Thumb

The fun of iNaturalist is the relative lack of restrictions.

The moment you start collecting, you need permits, written permission from the land owner, you have to register a project with the provincial conservation authority, and if you send it to a museum or university outside of the province you need to apply for an export permit. And very recently authorities have been insisting on ethical clearance for such activiites as lifting stones.
If you have any intention of collecting, I would strongly advice you to make contact with a local herbarium, museum or university and participate under one of their established projects if at all possible.

Posted by tonyrebelo 7 months ago (Flag)
Thumb

@predomalpha that sounds pretty doable ! Maybe I’ll try collecting some leaf litter samples in a few days, before the first frost hits

Posted by mws 7 months ago (Flag)
Thumb

I liked the framing of this post.
For me, I tried (and succeeded) to do the 1000 species in 1 sq km challenge this year.
Even in a UK town centre, with relatively low biodiversity, I did not struggle to find species which were new to me on a near-daily basis.
Finds included various rarely observed winged-insects, and an undescribed species of collembola.

The biggest barriers for me were in finding experts with the time, interest and ability to identify from photographic records.
The limitations of my equipment (and budget).
The time it took to process all my observations.

@loarie @tonyrebelo - as well as sending people to remote locations, one could also just send equipment - an iNaturalist sponsored field kit for example, could be cheaper and quicker to send out ...and might trigger more integrated localised interest as well.

The problem of attracting more expertise has been discussed a lot in the forum, but the discussions are naturally very skewed toward opinions of existing users. Perhaps there could also be more rigorous surveyance of experts from outside of iNaturalist about what deters them from taking part. From the ones I've spoken to in UK, they seemed more deterred by misconceptions about the platform than anything else.

Posted by sbushes 7 months ago (Flag)
Thumb

@predomalpha I wouldn't encourage specimen collection, especially in a thread referencing rare species (for obvious reasons)!

As for ideas for increasing observations of unobserved species:

1) The figures above point out one way: getting out of your comfort zone! 98% of birds is great, but how about some effort to document other non-bird species while birding?

Building off of effort-shifting, maybe there could be an integrated system where each week (or month) of the year is dedicated Arachnid or Crustacean month, where those observations would be given greater prominence on the front page to get those IDs moving along faster?

2) It'd also be interesting to use the geo-tagged observations that exist as a reverse map for areas which are relatively accessible, but which have not been documented by many people..I found one large area in NV this way, which turned out to be near a former nuclear testing site (https://www.inaturalist.org/observations/36306269)!

2a) building off of lesser-observed areas, would be lesser-observed times to capture crepuscular/noctural species (bugs, mammals, etc), as well as lesser-observed seasons (warmer months are better observed in my region of the Northern Hemisphere).

2b) Short tutorials on how to better observe hard-to-detect species?

3) Identifying biodiversity hotspots which already have lots of coverage, but not much ID activity and highlighting them in the 'Explore' tool for local experts to look into.

Maybe having an accompanying rare or 'most wanted (not yet documented on iNat)' species list for these areas as well? A lot of species may already be documented, but not yet recognized as such.

4) Thinking more systematically could help too, #1 here is shifting effort from well-documented to less well-documented species, #2 and #3 are basically the same for geographic locations.

So there's redirecting observation/detection effort, but also important to think about increasing correct IDs (cryptic, lesser-known species)--and decreasing false-negatives. How to actually do this is more difficult to imagine.

5) Maybe thinking about false-positives as well, how can we be sure that all of the 300,000 species are actually correct? If there's a 5% error rate (a conservative guess?) overall, there'd be 15,000 species that are mis-assigned!

Maybe there needs to be a systematic study into this, kind of how (crowd-sourced) Wikipedia was checked against Encyclopedia Brittanica (https://www.cnet.com/news/study-wikipedia-as-accurate-as-britannica/) to validate accuracy.

6) Finally off of the above, it started making me think about the automated IDs, and how to increase confidence in IDs, maybe we could get some sort of rudimentary badge/flair system to identify experts in various fields (or even people whose IDs tend to end up correct more often than not) to give more weight to their IDs?

I know this would be controversial and possibly problematic in many ways, but could improve IDs and reduce the amount of observations that are stuck in limbo with not enough experts to overturn a community ID of a more common or AI-determined species.

6b) Easier idea based on above, just silly flair for IDing or documenting, for example, 5, 10, 50, 100 insects, mammals, etc. This could be an easy and fun way for people to track their own contributions, since it's probable that many people don't realize how little they're actually documenting across the tree of life, much less are looking at their life lists to see what they've documented from a more zoomed-out level.

Just a few ideas..

Posted by yerbasanta 7 months ago (Flag)
Thumb

@yerbasanta
Some of these points, especially no.6 touched on and discussed in forum posts like these (amongst others):
https://forum.inaturalist.org/t/why-not-empower-recognised-experts/14419
https://forum.inaturalist.org/t/gamify-accuracy-award-value-to-quality-not-just-quantity/14428/4

I´d also wondered about 2b - would be great to hear some tips from some of the invertebrate pros.

Posted by sbushes 7 months ago (Flag)
Thumb

@tonyrebelo @loarie Maybe a partnership with Earthwatch? (https://earthwatch.org/expeditions/browse) They've been organizing expeditions for almost 50 years.

Posted by michaelpirrello 7 months ago (Flag)
Thumb

@loarie I love these blogs and I feel like I'm always hungry for more charts and visualizations. For the 300k milestone, can we get a bar chart of # of unique taxa that have singleton observations, # of unique taxa with 2 observations, 3 observations, etc? I remember a presentation you gave that had a similar figure for biodiversity data, with the point that for so many species, we only know them from a single record. Curious if iNat has exactly the same curve as other gbif sources, or if our tail is longer, or our singleton peak higher, etc.

Posted by muir 7 months ago (Flag)
Thumb

@muir here's some quick an dirty stats (e.g. 93,728 taxa with 1 obs)
1: 93728, 2: 45372, 3: 29082, 4: 20902, 5: 16274, 6: 10756, 7: 9107, 8: 8113, 9: 6940, 10: 6152
its not super accurate because (1) it includes nodes other than species, (2) it includes casual/needs_id obs, (3) its a cached value that sometimes gets out of sync. But its easy to extract from the db and should be within the ballpark of what you're looking for.

Posted by loarie 7 months ago (Flag)
Thumb

Cool, thanks @loarie! So, 93,728/300,000 = about 31% of the 300,000 species that have been observed on iNaturalist are known from a single record (i.e, a singleton). That's not so different from past estimates of singleton representation in other, more traditional biodiversity data sources. This type of sampling is way outside my area of expertise, but see this article on sampling and rarity:

"Singletons—species only known from a single specimen—and uniques—species that have only been collected once—are very common in biodiversity samples. Recent reviews suggest that in tropical arthropod samples, 30% of all species are represented by only one specimen (Bickel 1999, Novotny and Basset 2000, Coddington et al. 2009), with additional sampling helping little with eliminating rarity."

I think a possible convergence around 30% by both iNaturalist and traditional sampling is interesting!

Posted by muir 7 months ago (Flag)

Add a Comment

Sign In or Sign Up to add comments