This is the eighteenth entry in an ongoing monthly (or almost monthly!) series profiling the amazing identifiers of iNaturalist.

Gall Week September 2023 runs from September 2nd to September 10th (gall observations from anywhere in the world are accepted), so I thought I’d feature a few of the folks involved with the creation of Gallformers.org, a website that is “The place to identify and learn about galls on plants in the US and Canada,” and was started by people who met through iNat and contribute a lot of gall identifications and information to gall observations made in that region! 

Plant galls are, according to Gallformers:

Abnormal growths of plant tissues, similar to tumors or warts in animals, that have an external cause--such as an insect, mite, nematode, virus, fungus, bacterium, or even another plant species…Plant galls are often complex structures that allow the insect or mite that caused the gall to be identified even if that insect or mite is not visible.

These structures are bizarre, beautiful, and sometimes barely noticeable, but they’re always fun to find and post to iNat - at least I think so. I encourage everyone to look for them the next time they’re out iNatting.

Below is an “oral history” from five people who were part of creating Gallformers. It’s cool to hear how they met on iNat and collaborated to create an excellent resource that’s its own entity and also using iNat data. And while Gallformers’ focus is on the United States and Canada, I think that it, like the Fly Guide, is a great model for how identification resources can be made for any taxa and region.

Finally, I kept this piece to five people for simplicity’s sake, but there are many, many people who’ve helped grow the gall community on iNaturalist, so I want to emphasize that. 

The people you’ll hear from below are:

Jeff Clark (@jeffdc): Gallformers co-founder and coder, based in the Washington, D.C area

Timothy Frey (@calconey): a Postdoctoral researcher studying microbial communities associated with crop rotation, cover crops, and soil health characteristics at the Ohio State University, based in the Ag Campus in Wooster, Ohio

Noriko Ito (@norikonbu): amateur naturalist and Gallformers volunteer, based in the San Francisco Bay Area

Adam Kranz (@megachile): the director of Gallformers and a private school math and science teacher based in Austin, Texas

Kimberlie Sasan (@kimberlietx): an Herbarium & Research Assistant at the Botanical Research Institute of Texas, FWBG|BRIT 

(Some quotes have been lightly edited and I’ve tried to arrange them in a way that flows well, narratively.)

Adam: When I first became interested in galls in 2018, I was living in Detroit, Michigan and working as an in-home private tutor. My work took me all over the Detroit suburbs and I often had awkward periods of time between sessions, so I spent a lot of time visiting little parks. At first, I made it my goal to learn as many local taxa as I could and identify them on iNat. This was a dream of mine for many years, and I had a big backlog of photos I'd never been able to identify for lack of expert guidance. iNat’s computer vision and social network suddenly made all of that possible.

Kimberlie: I've been a big fan of iNat for many years now and I often find myself going from one rabbit hole to another, as my journal posts will show.  In one such post in December 2019 I documented the various wooly galls of Texas because I couldn't find anyone who knew anything about them. I had also created a project for DFW Galls to help me see their diversity and begin to recognize differences in morphology. Over time, I met some other iNat folks interested in galls and we started to talk more and more about how to ID them. Once I tapped into the Biodiversity Heritage Library and found the original gall descriptions, the flood gates opened wide. 

Timothy: I got interested in galls during my undergrad when I took a class called Field and Woodland Fungi, which randomly had a lecture on galls. The biology was briefly explained in the class (hijacking plant cells for their own benefit) and I was fascinated. The class also included lots of pictures and I was struck by the incredible phenotypic diversity. This fascination with galls influenced my decision of which lab I chose for grad school and I chose to work in the lab of Chris Taylor on a gall-forming nematode, Meloidogyne incognita (Root-knot Nematode). My Ph.D. work focused on how it uses host chemical cues to locate a host root.  

I found out about iNat in 2019 as I was finishing up grad school, an intern of mine that summer suggested I check out the app. I was initially skeptical as other plant ID apps that I had tried had been quite lackluster, but I found the iNat computer vision actually worked pretty well, and I loved the distribution maps that the app makes. 

Noriko: Born and raised in Tokyo, I moved to the San Francisco Bay Area in the mid 90’s and have spent much of my adult life here. I always liked nature and taking random photos but was not good at organizing them. I joined iNat in early 2020 when the pandemic started, so it wasn’t too long ago. Like many people during that time, I was desperately longing for something to do outside. I never liked any form of social media so I was a bit skeptical about sharing anything online in the beginning, but now I’m glad my positive experiences have been confirmed when the New York Times’ “The Nicest Place Online?” article came out. I found iNat was a good way to catalog what/where I’ve seen, otherwise I’d forget. I started out with spring wildflowers which had been my primary interest, then got into mushrooms when not much else was to be observed in the California winter, but didn’t discover galls until the summer of 2021.

Adam: The thing that kept me hooked was that there were so many easily filled gaps. If I could just upload a few dozen observations of a common local species, its ID would be automated forever. Because the site was growing rapidly at that point, the sooner we got those training observations collected, the more the computer vision model could help us stay on top of the flood of new ones. 

Of course, for that to be worth doing it had to be done right, which meant focusing on groups with accessible visual ID traits backed up by reliable scientific sources. I gravitated to things that were smaller and weirder, and started collecting resources for any groups that seemed accessible for me and set myself the task of identifying those observations in Michigan and nearby states. One of the most exciting resources I found in that process was @ceiseman's book Tracks & Signs of Insects and Other Invertebrates. That book had a brief chapter that just dipped into galls, so for the first year or so I worked on galls alongside leafminers and eggs and rust fungi and aphids and slugs and all sorts of other things. Whenever I found something I couldn't find in the book, I would tag Charley on iNat and he would try to answer from the resources he'd collected. Eventually he emailed me a copy of Weld's 1959 catalog Cynipid galls of the Eastern United States, and I started a little spreadsheet to keep track of everything I found there and in other sources. From there, I expanded my focus to all gall inducing taxa in North America.

Jeff: Not long after I found iNaturalist (Spring 2020) I was walking around in some local woods when I found a couple of oak galls. I knew what they were but as it turns out I knew hardly anything and there was so much to learn. I took photos of these galls and uploaded them to iNat. They were identified to species and my interest was piqued. 

Noriko: It was a bit of a shock when I realized that I didn’t know anything about galls because by that time, I thought I knew most things in nature pretty well. So I started looking them up and came across Ron Russo’s article called “Confession of a Gall Hunter” [PDF]. He explained beautifully how he got into it, and I was mesmerized. Soon after I started posting my gall observations on iNat, I found a project called Galls of California. “Is there a such thing?” I thought. “I wasn’t the only weirdo fascinated by this???” 

I joined the project in August 2021, bought Russo’s newly published book Plant Galls of the Western United States and read it cover to cover. Through the Galls of California project, I’ve learned so much from experienced gall observers/identifiers such as @nancyasquith, @garth_harwood, @merav, @leslie_flint, @graysquirrel (this list goes on)… and of course, Adam Kranz.

Kimberlie: By May 2020 Adam had created his first iNat guide on Fuzzy Oak Galls.  We were furiously messaging back and forth about all the species we could find information about. I was also learning more about how to photograph the galls to get a better identification.  Many of the taxon images for Celticecis sp came about from my experiments with cross sections and light boxes.

Timothy: I had been somewhat casually tracking gall incidence in my area for a few years prior and decided to upload a few galls to the app (and a few that I could not find an ID for on my own), this is where I first ran into Adam, who was the biggest gall identifier at the time. I quickly figured out that this was a place where my interest, curiosity, and passion could be put to use by helping to ID these interactions/organisms. I devoured most of the readily available gall literature (Weld’s cynipid books and Gagne’s midge book and papers), so that I could become more adept at IDing galls on the platform.

Jeff: Most of the gall literature was either PDF scans of old journal articles or books that were difficult or very expensive to obtain. This got me thinking about ways to improve the situation via technology. Almost at this exact same time Adam reached out on the iNat forums asking if anyone was interested in helping him build an ID tool and database for galls. Next thing I knew, I was building out Gallformers.org. I wrote and maintain the code and infrastructure for the site.

Adam: It quickly became clear that gall ID traits lent themselves extremely well to a non-dichotomous key format. Once the vision for that tool had planted itself in my mind, it became a minor obsession to make it a reality. I started posting about it on the iNat forum, first asking if the site's developers could help me expand the existing Guides feature, then asking for other existing platforms it could be built on and collaborators to do that building, and finally for collaborators to build it from scratch. I eventually got tremendously lucky and received a reply from Jeff Clark, a web developer who was then on sabbatical between jobs, had exactly the skill set we needed, and was one of a few users who had become excited about galls because I was interacting eagerly with their gall observations. He built the site in late 2020 and we've now been operating since 2021. 

I put a big push in to finish the oak gall wasp portion of our database over the summer of 2021, and in the process became personally obsessed with their taxonomy and biology. Fortunately, once the site existed, it attracted a lot of new and existing gall enthusiasts who took up responsibility for adding data on many other gall-inducing taxa.

Kimberlie: While Adam and Jeff were connecting and brainstorming about Gallformers.org I was collecting, rearing, and photographing as many oak galls as I could find.  It gave us some great photos for the website, and I learned a lot about rearing galls first-hand which I could share with others who were also new to learning about galls. Gallformers.org has been a great community and I'm thrilled to watch it take off like it has!

Noriko: About a year later, in the fall of 2022, Adam invited me to join Gallformers.org, probably because I was pestering him and other Gallformers admins too much about updating their website. I felt flattered but was going to decline his invitation because I didn’t think I could contribute much. I was (and still am) just an amateur naturalist without any academic/scientific background in this field. But he wrote back, “There are no commitments involved.” So I decided to give it a try.

After joining Gallformers, I started helping Adam with some tedious data entry for the phenology tool that he was developing (I’ve just written a little bit about it here). In the early months of this year, we started planning the first Spring Gall Week with @merav, and @naturesarchive produced a wonderful podcast interview with Adam about it. This was the first event where I actively started making IDs on other people’s gall observations.

Adam: I've shifted my focus from building the website to collaborating with scientists to study oak gall wasps. A lot of my engagement with iNat now is to encourage people to collect and rear undescribed and understudied wasps for description and sequencing projects. To calibrate the timings for those collections, one of my big projects has been to accumulate a database of gall phenology data and share it using a web interface. I recently built a web extension that dramatically reduces the friction in adding observation field values to observations, which is necessary for collecting that data from iNat.

Jeff: As for the identification of galls, I tend to stick to the Cynipids, but these days I end up identifying tons of Eastern North American Oaks as well. I use Gallformers to help, much better than my faulty memory! After knowing Adam only virtually for 2+ years we finally had the chance to meet this past July when I was in Austin on a business trip (see the photo). Besides meeting Adam and a bunch of other great people via iNat, I also met the love of my life (@izafarr) via iNat!

Kimberlie: In October 2020, the year of the Gallformers boom, local botanist Bob O'Kennon (@bob777) invited me to see a gall he had found that I didn't recognize from the literature. It turned out to be an undescribed species, and with the help of Miles Zhang (@mileszhang), we published it in July 2022 as a new species, Druon laceyi. 

Adam: All of that work paid off this summer when I had the chance to go to the 8th International Plant Gall Symposium in Chico, California. I brought something like 60 undescribed wasp collections to send back to taxonomists in Australia, Hungary, and Spain. But the coolest part of that conference was that almost every presentation about North American gall inducers included photos by Gallformers contributors on iNat and acknowledgements for us at the end. It was really satisfying to see that our work was paying off in connecting the amateur and scientific communities in such a mutually rewarding way.

Noriko: After Adam returned from the 8th International Plant Gall Symposium, he is now full of new ideas/projects and busier than ever, so I hope to continue helping anyway I can, and am excited to see where Gallformers is headed.

I consider Adam and Nancy as my mentors because they really took me under their wings and taught me with patience. Without their support, I don’t think I’m here today. And thanks to iNat and Gallformers, I’m meeting many interesting people: other iNatters who invite me to go hiking together because they also think what I see is “cool”. Oak experts who comment on my hybrid oak observations and share their knowledge and passionate discussions. Top researchers who ask me to send specimens and let me know what they found about them. Who’d have thought I could find a community that I feel proud to be a part of, through these anomalies called… “galls”!

Gall tips and resources:

Noriko recommends @nancyasquith’s resources here, which focus on North American galls but also includes resources/info for galls in general. 

From Timothy

Galls can be found anywhere, this is one of the things that I find appealing about them – rest areas, edges of parking lots, cemeteries and other forgotten spaces usually contain galls, you will be surprised by how many you can find in your backyard.

1. Consider that galls have evolved with their host plants – this means that you are much more likely to find galls (population # and diversity) on native plants rather than exotic plants (although they often contain some gall species that have traveled with them as well). A site with a large diversity of native plants is likely to have the largest diversity of galls.
2. Consider a site where trees have branches that are accessible to you without having to climb – parks, cemeteries, forests along roads or other edges are easier to find galls in compared to an open forest which may have few accessible leaves.
3. Consider a site with a grassland component – in addition to having trees with more accessible limbs, many grassland plants also have quite a lot of gall diversity (ex: goldenrods and other Asteraceae here in the Midwest US)
4. Take the time to turn over leaves, many species only occur on the bottom surface of leaves and may not be noticeable before looking at the underside.
5. Many gall-forming organisms are carried at least short distances with prevailing winds and I have found that edges of woods that have an area that is exposed to prevailing winds often have a high gall density (usually western or southern edges here in NE Ohio.).
6. Galls are seasonal – you may find different species on the same plants at different times of the year (particularly true of the Oak Gall Wasps). There are often different species present in spring than are present during summer and fall. In general late summer (August/September here in NE Ohio) seems to have the greatest density of species.

The most important thing to ID galls is to “Know thy host” – Can you ID the host? If so, ID as narrowly as possible, galls are often fairly host-specific.

1. I recommend the following shots of a gall for it to get identified on iNat: 
2. Upper surface 
3. Lower surface (even if there appears to be nothing there) 
4. A wider shot with context (leaf, stem etc that the gall is on, ideally with the gall itself still visible), this shot is missed frequently but is extremely helpful in confusing cases. 
5. Further pictures of the host and gall if the ID of the host is uncertain. 
6. A photo of the cross-section.

Photos, from top to bottom:Spiny Leaf Gall Wasp (Diplolepis polita) by @mileszhang, Adam Kranz (left) and Jeff Clark (right), Timothy Frey, Noriko Ito, Kimberlie Sasan, Adam Kranz (center) with other researchers at the 8th International Plant Gall Symposium.

Our Observation of the Day is this Petrolisthes violaceus porcelain crab, seen in Chile by @matias_saa!

“I believe that the people of my country…are not very familiar with the animals or other kinds of life that coexist in the same areas where they live,” says Matias Sáa, a recent veterinary medicine graduate.

Conservation depends on the motivation of the people who live by nature, and what is not known is not protected - for this reason I believe that photography is a very good tool to bring this unknown world closer to people, to show how fascinating nature is.

Matias lives in the city of Viña del Mar, and likes to visit the coast at low tide. The crustacean you see above was found when Matias and a friend were out looking for creatures that Matias’s friend could put in nature guide they’re working on. 

This large and impressively colored specimen caught our attention and was the perfect candidate to include in this work. These animals are quite common in the area and I have always been struck by how little known they are to the inhabitants of my country, whose coast and sea are famous for having only “gray and not very colorful fish.” For this very reason I like to share my photos and make these colorful animals known, thus showing the true natural wealth of the Chilean coasts.

While they look like crabs, “porcelain crabs” (Porcellanidae) are more closely related to squat lobsters, forming the super family Galatheoidea. Porcelain crabs are often quite flat, which allows them to hide in crevices and under rocks. Their large claws, which are used for territorial disputes, detach easily as a form of defense, and is the basis of their common English name. To feed, they use special limbs by their mouths to trap plankton. Some are dull-colored, but many others are vibrant. 

“I had a special interest in nature from an early age thanks to the fact that my family really likes to spend time outdoors in natural spaces,” says Matias (above). 

My dad took me on excursions to hike through different hills in my country where little by little I learned about all kinds of creatures and other forms of life…In the future I plan to dedicate myself to the area of ​​exotic animal and wildlife medicine to contribute to nature conservation.

I use iNaturalist mainly as a means to identify species, but also to keep informed about the “rare observations” that appear near me, so I can try to find and photograph them. It’s also very useful in gathering information for my research.

(Some quotes have been lightly edited for clarity. Google Translate was used to translate text from the original Spanish.)

- here’s some footage of a porcelain crab feeding.

Our Observation of the Week is this Banded Caecilian (Scolecomorphus vittatus), seen in Tanzania by @pius_mollel!

Living and growing up in Tanzania immersed me in some of the most beautiful land,” says Pius Mollel, who’s currently pursing a master’s degree in Wildlife Ecology at the University of Dar es Salaam. 

Parks, mountains, forests, and rivers are all within a few miles from my home; these places have been literally in my backyard for most of my life. All these made me interested in the outdoors from a young age. The birds, insects, frogs, and other animals always caught my attention.

I am in love with everything in nature from land to sea but specifically am interested in herpetology (the study of reptiles and amphibians) and using my knowledge and skills I wish to unlock what resides in biodiversity hotspots in Tanzania - specifically forests in the Eastern Arc Mountains that are well-known for their exceptional biological diversity and high number of endemic species.

Pius led a research team in Tanzania’s Mamiwa Kisara North Forest Reserve this past June, and the group was hit by rain one afternoon. While at the camp site, Pius spotted something moving on the ground.

At first, I thought it was a snake and when I moved closer, I realized it was a caecilian because of its shining skin, I then noticed its coloration (dark brown and pink on the ventral side) and I confirmed it was Scolecomorphus vittatus. They are a hard species to spot because they spend most of their time hiding in soil, but when it rains they normally come out and that is how I got lucky to observe this one. They are also fast-moving and don’t settle when disturbed. This makes them hard to photograph so I was lucky to time that awesome shot.

Caecilians are an order of legless, mostly ground-dwelling amphibians that occur in tropical areas of the world (although not in Australia or the islands north of it). Not much is known about caecilians, but we believe that adults, at least, are carnivorous. Skin secretions of the banded caecilian are apparently toxic to humans. 

Pius (above) discovered iNat in 2021 when looking for plant identification apps in the Google Play Store, but didn’t post anything until last year. 

I use iNaturalist as a platform to share the fascinating flora and fauna found in Tanzania and I find it the right place to communicate my observations because it is a network of scientists and nature enthusiasts. I use iNaturalist as a place where I can connect with other researchers, share findings, and work together on projects related to specific species, regions, or topics of interest.

I see plenty of iNaturalist observations of wildlife, plants, and fungi, which makes me realize everything in nature is important and there is a need to take care of them so that future generations can enjoy these beautiful and fascinating organisms.

(Some quotes have been lightly edited for clarity.)

- young caecilians have been documented eating extra skin grown by their mother.

- some caecilians eschew land for the water, like these aquatic caecilians at the California Academy of Sciences.

- check out this Observation of the Week from 2017 that features a different legless underground animal seen by @hydaticus.

Expanding our capacity to address the biodiversity crisis in a new phase of growth and impact to 2030

We’re excited to announce a generous $10 million grant from the Gordon and Betty Moore Foundation. On the heels of our spin-off as an independent nonprofit in July, this grant will enable iNaturalist to embark on a new phase of growth and impact through 2030.

Donations Matched 2x or 3x

As part of the grant, we’re eager to share that donations through December 31, 2023 will be doubled and recurring donations will be tripled. This turns a $10 one-time donation into $20 and $10 per month into $30 per month.

Challenge and Opportunity

This grant—the largest in iNaturalist’s history—is a testament to everything the iNaturalist community has accomplished to date and an investment in iNaturalist’s capacity to address the biodiversity crisis by scaling our conservation impact.

The biodiversity crisis is becoming more acute every day. Human-induced climate change and land use alterations are now causing species extinctions at a rate a thousand times faster than any time in history. iNaturalist already has millions of participants engaged in documenting biodiversity with the click of a camera, but we need millions more people involved to address the growing threats to biodiversity.

These are big challenges for iNaturalist: expanding our community, increasing engagement, and exploring new ways to translate biodiversity data into conservation impact. The generous grant from the Moore Foundation is a major first step to address those challenges. We’ll need a lot more support to reach these ambitious goals, but every person can make a positive difference for iNaturalist.

Many, Many Thanks

We can't say thank you enough to everyone who generously contributes their time, energy, knowledge, and resources to support iNaturalist. This grant and matching donations will go toward ensuring your contributions will have a lasting impact. Thank you for everything that you have already done, and everything that you will do, in support of iNaturalist.

Together, through iNaturalist’s unique global community working to understand, document, and protect nature, we can make a profound difference in safeguarding our planet's precious biodiversity.

DOUBLE or TRIPLE MY GIFT NOW

FAQs

Will you be hiring?
Yes, we anticipate advertising for new positions in the next few months.

How can I learn more about the Moore Foundation?
The Gordon and Betty Moore Foundation has been a supporter of iNaturalist since 2017. Gordon and Betty Moore established the foundation to create positive outcomes for future generations. Visit Moore.org or follow @MooreFound on Twitter to learn more.

Why do you still need donations?
As a nonprofit, donations of all sizes are essential to a long and sustainable future for iNaturalist. We truly appreciate every donor and gifts of any size, and we can't maximize the Moore grant without matchable donations from the community.

Do I need to do anything special to get my donations matched by the Moore Foundation?
Nope!

I donated last month. Was my donation matched?
Yes! All donations we receive from July 1 through December 31, 2023 are matched.

Will you match workplace giving?
Yes! Employer matches, payroll donations, and volunteer hours will also be matched. This is the best time to take advantage of workplace giving programs. Your initial gift could ultimately be matched 3:1, 4:1, or even 5:1! You can search to see if your employer has a program.

Will you match donations from stock, IRAs, or Donor Advised Funds?
Yes! You can find more details on our other ways to give page about how to direct them appropriately.

What counts as a "recurring donation"?
Monthly, Quarterly, or Annual donations are considered "recurring".

Our Observation of the Week is this Murici Lancehead snake (Bothrops muriciensis, Jararacuçu de Murici in Portuguese), seen in Brazil by @josi_goncalves.

“I have always been very curious to know about animals,” says Josi Gonçalves.

At school I loved biology. I always fancied myself a microbiologist, but in college I fell in love with zoology and discovered that I was born to be an ornithologist. I'm currently developing research in my master's degree related to the interaction of insectivorous birds and army ants here at the Centro de Endemismo Pernambuco.

Last year, while collecting data at the Estação Ecológica de Murici,

my teammates and I found this Jararacuçu de Murici. In fact, I almost stepped on it as I didn't see it, it was hidden between two fallen logs. It was so small that it probably wouldn't survive if someone stepped on it.

What Josi and her teammates nearly stepped on was a rarely seen species of viper that was only described in 2001. In their 2012 paper, Graboski, Verissimo, et al say that not much is known about this species, but conclude that like many other members of Bothrops, juvenile murici lanceheads use their white tail tip to lure prey (likely frogs and toads), then switch to being generalist predators as they mature. The species is listed as Critically Endangered by the IUCN, threatened by habitat loss from agriculture.

Josi (above) only joined iNat a few weeks ago and has since been posting mostly older photos (including some really cool stuff like velvet worms).

I discovered iNaturalist through friends and started using it out of curiosity and also so that my records could contribute in some way to science. For me, it has helped me a lot to get to know new species of so many different types and to realize that there are always new things to be discovered and that there is never too much knowledge!

(Some quotes have been lightly edited for clarity.)

- here’s some footage a murici lancehead slithering about in the leaf litter, which seems to be its preferred habitat

- check out the diverse beauty of the genus Bothrops.

We released a new computer vision model today. It has 79,797 taxa up from 78,387. This new model (v2.7) was trained on data exported last month on August 20th.

Just a short post this time compared to past posts as we're focused on announcing the iNaturalist Geomodel with this v2.7 release.

Here's a graph of the models release schedule since early 2022 (segments extend from data export date to model release date) and how the number of species included in each model has increased over time.

Here is a sample of new species added to v2.7:

• Mammalia
• Protozoa
• Plantae
• Fungi
• Other Animalia
• Mollusca
• Arachnida
• Actinopterygii
• Reptilia
• Insecta
• Amphibia
• Aves

Today, we’ve changed the Seen Nearby label on suggestions to Expected Nearby. The label comes from predictions made by the iNaturalist Geomodel that we’re introducing for the first time.

What is the Geomodel?

Most of you are familiar with the iNaturalist Computer Vision Model which takes an image as input and returns the most likely species based on visual similarity as an output. We train that model on a set of about 80 thousand species with enough data and update it monthly (we released version 2.7 today).

The iNaturalist Geomodel takes a location as input and returns the most likely species at that location as output. Like the Computer Vision Model, it is a Deep Learning model trained on the same set of taxa and updated on the same monthly schedule. We developed and published the Geomodel in collaboration with the same Visipedia team that assisted with the iNaturalist Computer Vision Model. The map below shows Geomodel predictions for American Pika. The Geomodel is trained only on iNaturalist observations and an elevation map.

From Gridded Observations to Geomodel Predictions

iNaturalist has been using the Geomodel to weight computer vision suggestions since June of the year. We started using the Geomodel to apply the Expected Nearby label today.

Previously, we used a gridded version of the raw observations to weight Computer Vision suggestions and apply the Nearby label. We counted the relative number of observations for each species onto a 1-degree grid. If there were any observations of the taxon in the surrounding 9 grid cells, we applied the Seen Nearby label to suggestions. We used the relative number of observations in the grid cells to weight the Computer Vision suggestions. Note the grid cell for Mexican Treehopper in southern Brazil likely due to a misidentified observation.

We’re now using the Expected Nearby Map predicted by the Geomodel to apply the Nearby label. You can think of the Expected Nearby Map as an estimate of whether the species is present near the location.

The change in name from Seen Nearby to Expected Nearby is intended to make it clear that the label comes from a model prediction rather than a grid of observations. Note that these predictions aren’t perfect. For example, Mexican Treehopper probably doesn’t occur in the Galapagos or Cuba despite the predictions. For some species the Geomodel performs remarkably well, while for others predictions have very high error. Work to better understand these and experiment with improvements is ongoing. But as we show below, on average the Geomodel improves upon the 1-degree grid approach it replaces and we expect continued improvements with future Geomodel versions.

We use an Unthresholded version of the Expected Nearby Map to weight Computer Vision suggestions. You can think of the Unthresholded Map as the relative probability that a species occurs at a location.

You can explore the Expected Nearby Maps and Unthresholded Maps we use to weight Computer vision suggestions on new Geomodel prediction pages we’ve linked from the taxon pages of all approximately included 80,000 species.

Why the Geomodel and Next Steps

We transitioned from the 1-degree gridded data to the Geomodel for four main reasons:

1. Improvements to Computer Vision suggestions

As detailed in the Evaluating the Geomodel section below, the Geomodel improves the accuracy of Computer Vision suggestions compared to the 1-degree grid approach. Version 2.7, released today, is about a 4% improvement over the 1-degree grid approach for Top 1 suggestion accuracy and we anticipate more accuracy gains with future Geomodel versions as we refine the modeling approach and more observations are uploaded.

2. Future direction: Fast/offline geospatial information

The number of Geomodel parameters is less than 2% the size of the 1-degree grid cell data. This means the Geomodel is small and fast enough to run on the mobile device like the Computer Vision Model does on Seek. This opens up the potential for including geospatial information in features such as the Seek in-camera suggestions and displaying taxon maps on mobile devices offline. We haven’t built these features yet, but the Geomodel will make them possible.

3. Future direction: Surfacing unusual observations

As iNaturalist grows, the community needs better tools to surface unusual observations that may represent misidentifications or important discoveries such as a range extension or the early detection of an invasive species.

The figure below shows 2.1 million dragonfly observations ranked by their geographic unusualness as predicted by the Geomodel. The right side of the histogram shows the most unusual 0.01% of observations. We sent these 223 unusual observations to dragonfly expert @dennispaulson to vet. 197 observations (88%) were misidentified observations (red bars) such as this Rainpool Spreadwing misidentified as a Slender Spreadwing. The remaining 26 represented some legitimately unusual records (white bars) such as this Slaty Skimmer range extension from Colorado.

Some observations in the white bars were unusual to our model but not to @dennispaulson, such as this Highland Meadowhawk from Haiti that the Geomodel thinks is unusual. With more observations and identifications from poorly sampled regions, the accuracy of the Geomodel will improve over time.

Fly expert @zdanko helped with a similar experiment with 500,000 hoverfly observations. Similar to dragonflies, of the 365 most unusual observations, 267 observations (73%) were misidentifications.

We’re excited about the potential to build tools around the Geomodel to help more quickly surface these unusual observations for more attention from experts so that misidentifications can be fixed and important discoveries like species range extensions aren’t missed.

4. Future direction: Context about range size

One of the most important characteristics of a species from a conservation perspective is its geographic range size. All other things being equal, small-range species tend to be at much greater risk of extinction than species that are widely distributed. In order to prioritize scarce conservation resources and attention, land managers need tools to determine which species are small-ranged local endemics (species that occur nowhere else in the world) from other more widely distributed species.

As described in the Evaluating the Geomodel section below, Geomodel predictions of range area are well correlated with the areas of range maps such as the Taxon Ranges that appear on some taxon pages that come from external sources.

The figure below shows Geomodel predictions of range area for 10 small-ranged birds from around the world. We hope to build tools around the Geomodel to make it easier to determine which observations belong to small ranged endemic species in order to help the land management community prioritize these conservation targets.

The Expected Nearby Maps are being rendered on the Geomodel prediction pages at a coarse 1.8 thousand square-kilometer resolution and therefore are not publicly revealing precise information about sensitive species. We continue to improve iNaturalist channels that securely mobilize sensitive species data and precise predictions for conservation purposes.

Evaluating the Geomodel

We have evaluated the Geomodel by measuring:

1. Improvements to suggestion accuracy
2. Retaining the correct suggestion in the Expected Nearby subset
3. Overlap between Expected Nearby maps and Taxon Ranges

1. Improvements to suggestion accuracy

On average, Top 1 suggestion accuracy improved from 75% to 83% (+8%) by weighting the raw Computer Vision scores with the 1-degree grid. Weighting with the Geomodel instead improved Top 1 suggestion accuracy to 87% (+12%). We repeated this analysis within geographic and taxonomic groupings and in all cases the Geomodel outperformed the 1-degree data.

2. Retaining the correct suggestion in the Expected Nearby subset

By default, we only show the subset of Nearby suggestions. This has the advantage of removing suggestions that are unlikely based on location, but there’s also a risk of removing the correct suggestion. We calculated Recall statistics measuring how often the correct suggestion was retained in the Nearby subsets derived from the Geomodel and the 1-degree grid. On average, both approaches yielded the same Recall of 0.94 meaning for every 100 observations the correct result was included in the Nearby subset 94 times.

3. Overlap between Expected Nearby maps and Taxon Ranges

To measure how well the Expected Nearby maps compared to the Taxon Ranges displayed on the iNaturalist taxon pages, we compared them and calculated Precision and Recall statistics. The Taxon Ranges aren’t perfectly accurate either so for evaluation purposes we used the subset of around 5,000 Taxon Ranges that contained at least 90% of the observations for the taxon.

We repeated this analysis comparing the 1-degree grids and Geomodel to the Taxon Ranges. The Geomodel predictions improved the average of Precision and Recall. The F1 statistic (the harmonic mean of Precision and Recall) improved by 9% for the Geomodel compared to the 1-degree grid.

The Geomodel also does a better job of matching Taxon Range area than the 1-degree grids as measured by Mean Logarithmic Squared Error (MLSE).

Thank you

We want to extend special thanks to our research collaborators, including Oisin Mac Aodha (University of Edinburgh), Elijah Cole (Caltech), Grant Van Horn (UMass Amherst), Christian Lange (University of Edinburgh), Pietro Perona (Caltech), and @tbrooks (IUCN), as well as the generous support from a Climate Change AI 2021-2022 Innovation Grant that helped make this work possible.

We’re excited about the gains in suggestion accuracy the Geomodel is making possible today and the potential for future directions that it opens up for us to pursue in the coming months. Thank you to the entire iNaturalist community for generating all of the observations and identifications that make training powerful models like the Geomodel possible!

Our Observation of the Week is this Prevost's Squirrel (Callosciurus prevostii), seen in Malaysia by @dixonlau!

Last year, Dixon Lau, who was born in Sarawak, Malaysian Borneo, took a road trip that included a stop at Similajau National Park. It was at this park that he saw an adult male Black Hornbill (Anthracoceros malayanus) chased from its perch by the squirrel you see above. 

“The squirrel photographs had been resting on my hard drive for over a year until recently,” says Dixon, until he shared some of his squirrel photos with his friend Hobart WQH (@hobostew), who suggested to Dixon that he post the photo to iNat. 

Prevost’s squirrels range through the Thai-Malay Peninsula and on various islands in southeast Asia, including Borneo and Sumatra. They’re pretty big, at about 20-27 cm (8-11 in) in body length (not including the tail) and mostly stick to trees, where they feed on plants and, occasionally, insects. 

“I've always been interested in photography since I was a child, but I've never taken it seriously,” says Dixon (above). 

Today, I am an avid wildlife photographer and observer, with a special interest in birds. In my spare time, I like to be outside and observe nature. At the same time, I strive to film wildlife behavior in their natural habitats wherever feasible.

Unawareness of a squirrel may pique the interest of some people, and it may also aid in scientific community studies of this animal. Since [Hobart’s suggestion], I've become more eager to submit my discoveries of various wildlife species to iNaturalist. I'm hoping someone finds my post useful in some manner.

- you can see Dixon's wildlife videos on YouTube.

- Prevost's squirrels are noted for having a variety of color patterns. Check out the different patterns on iNat!

This is the nineteenth entry in an ongoing monthly (or almost monthly!) series profiling the amazing identifiers of iNaturalist.

The other week, I chose @eijimyorin’s Pagurus japonicus hermit crab as the Observation of the Day. When I checked out his profile, I saw that he’s a hermit crab expert who’s making his way through all hermit crab observations on iNat (currently about 60,000 total) and giving them a reappraisal. I wanted to learn more about this so I reached out to him about being featured in the iNat blog.

“I have a particular respect for hermit crabs,” says Eiji Myorin, a researcher at the Kuroshio Biological Research Institute in Japan. 

They are very intelligent and I consider them to be representatives of “tool-using creatures.” The tool is the shell that is their home. They also know that when they find a new shell, they can move the sand out of it by rolling the right-facing shell to the left. Some hermit crabs also attach anemones to their shells to protect themselves from outside enemies.

Eiji’s other area of research, which he calls his day job, is coral photochemistry. 

I make optical measurements of corals to determine what fluorescent proteins they have, make optical measurements in the ocean to determine the light requirements of corals, and design LED lighting to reproduce them. 

He likes identifying hermit crabs on iNaturalist for two main reasons:

First, I like to communicate with the observer and provide he or she more information about the observation. This will increase the accuracy of the person's future identifications and allow me to advise other novices.

I also correct the identification and provide information to make the records useful to others in the future.

As for identification resources, he relies mostly on internet image searches and scientific publications, but is building his own Hermit Crab Visual Dictionary for Japanese hermit crabs, a continuing endeavor which he calls “my life's work at this point.” 

Since I don’t like killing creatures, I am trying to avoid the method of identifying specimens by killing hermit crabs, and am working hard to establish identification information to be able to judge them based on ecological photographs as much as possible, and to promote this method. My hope is that the general public will then be able to identify hermit crabs without the need to kill them.

[In the process of creating the website,] I sometimes find new species, in which case I ask a specialist to write a paper on them. I am mainly engaged in observing their ecology and securing specimens of new species.

There are still many undescribed species of hermit crabs registered in iNat, and all the data on rare hermit crabs, including such records, are my favorite. For example, I have observed Clibanarius ransoni, which is native to Indonesia and sometimes washes ashore in Japan, but I first learned about an undescribed species with blue eyes similar to Clibanarius ransoni in India, which I first learned about through iNat. Such discoveries are one of the enjoyments of iNat.

(Some quotes have been lightly edited for clarity.)

Found a hermit crab? Here are some photography tips for making an identifiable iNat observation:

Carefully observe the color and pattern of the eyes, antennae, pincers, and legs. Each part of the body often has a distinctive color and pattern, so you need to be careful not to miss them.

Also, many iNat users hold hermit crabs in their hands and photograph them in the air, which makes it difficult to identify them because the colors and shapes are different from their original features due to the reflection of water droplets and hairs sticking to the body. Therefore, I recommend photographing hermit crabs underwater, if possible, and using those photos as a basis for identification.

- check out Eiji's Instagram feed, as well as his YouTube videos, he's got some excellent hermit crab footage.

Here are our favorite stories from September 2023 highlighting the iNaturalist community and impact. Thank you for making all this possible!

iNaturalist’s Science Impact

Yesterday, @thebeachcomber appeared in this Sky News clip and @jonodashper was quoted in this article. Both pieces addressed a new Australian study revealing the outsized contribution of iNaturalist and other citizen-science projects towards scaling biodiversity data. Here are some of our favorite September examples of how iNaturalist data are being used for science.

Invasive Species Science

This PBS News Hour segment on the $423 billion annual costs from invasive species endorsed iNaturalist as a way to engage the public in the early detection of invasive species. Minnesota Public Radio’s wonderful story about @krmenzel’s discovery of an invasive beetle is a great example of this.

In New Zealand, this study used iNaturalist to assess the spread of an invasive epiphyte. In Florida, this story described using iNaturalist to find invasive chameleons, and this Scientific Reports paper predicted chameleon invasions with iNaturalist and other data. We love seeing more and more bioblitzes like this one in Canada that incorporate monitoring and removal of invasives like Purple Loosestrife.

Climate Change Science

This Scientific Reports article used iNaturalist data to identify future refugia for endangered ghost orchids and their pollinators from the effects of climate change. Also published last week, this story about a Nature Climate Change paper by researcher @moore-evo-eco made use of observations from iNaturalist to measure the challenges flying insect face moving to higher elevations as climate changes.

Science Discoveries

We enjoyed last week’s story on @mathewlbrust‘s adventures with iNaturalist to confirm a new species of beetle in Wyoming and this paper describing a new praying mantis species in Brazil with help from iNaturalist.

@nicolasjouault made the first discovery of a leech in Britain on the back of a ray. Meanwhile, @kirbymorejohn shared observations documenting the first records of scorpions on Palmerston Island. Other range extension research papers that made use of iNaturalist included a paper by @johnascher and colleagues on a bee in California, the first record of a bug in Europe, and the arrival of a webspinner in Cyprus.

Conservation and Monitoring

In Canada, this study used iNaturalist data to measure motorboat damage to turtles, and this story described the use of iNaturalist to monitor Hudson Bay wolves. In Africa, stories detailed @brandonkemp and colleagues’ wildlife monitoring efforts in Tanzania and @lova_marline and colleagues’ bryophyte surveys in Benin. This study, leveraged iNaturalist to help assess fish populations remotely from a war zone.

iNaturalist Impact on AI Research

iNaturalist data was used in computer vision applications ranging from this robot designed to kill invasive Spotted Lanternflies to this study revealing color variation in salamanders. We’re thrilled that the new NSF funded AI Biodiversity Center will partner with iNaturalist to advance these types of applications at the intersection of AI and biodiversity.

iNaturalist for Land Stewardship

We loved this story about a iNaturalist Biodiversity Survey Sign erected at Quick Point Nature Preserve in Florida by @robcybulski and other Florida Master Naturalists. Other great land stewardship stories included this story about @lizcarr and fellow parishioners using iNaturalist to create EcoChurches, this story about Green Acres Farm, and many stories about local bioblitzes such as this one about the ongoing Baltimore Wood’s bioblitz.

Explore this innovative application of iNaturalist observations to rank US beaches by species diversity and this interactive feature showcasing local Humboldt plants and animals through iNaturalist. In Esquire magazine, Jeff VanderMeer, the author of Annihilation, shared his excitement at discovering a firefly with only 33 observations on iNaturalist while re-wilding his yard. iNaturalist observations beautifully capture the unique flora and fauna that make a piece of land truly exceptional.

September iNaturalist Events

We loved this video promoting the 2023 MarbleBowl and this video promoting EuroBioBlitz 2023. There was also great coverage of India Big Butterfly Month 2023 found here, California Biodiversity Day found here, and Parks for Pollinators found here.

iNaturalist’s Human Health Impact

In this study from the current issue of Ecology and Society, researchers suggest that engaging with iNaturalist and other citizen-science apps enabled individuals to experience the mental health advantages of connecting with nature during the COVID-19 lockdown.

Don't miss @deborahcostine's insightful article highlighting using iNaturalist for stress relief. Other stories promoting healthy lifestyles describe incorporating iNaturalist into activities ranging from walking to school and work to nurturing family bonds on Grandparent’s Day.

iNaturalist’s Education Impact

This study published last week from researchers at the University of Ioannina in Greece found that using iNaturalist significantly increased a group of future Kindergarten teacher’s connection to nature.

This story features @mossy_megan while describing an campus bioblitz in Indiana. Other stories highlight September student bioblitzes in Virginia and Ohio. Thank you to everyone holding iNaturalist trainings in your communities including @rogerritt in Maine, @akilee in Massachusetts and an online training by @naturalist2073.

iNatters in the News

Our favorite September piece profiling an iNatter was this story about @sallyslak’s journey from iNatting bugs during the pandemic to co-authoring a scientific paper on tortoise beetle poop architecture (not kidding).

Other September stories include this excellent profile on Florida iNatters @billpranty and @dmfraser, this profile on New Zealand iNatter @carey-knox-southern-scales, as well as stories about bot flies on chipmunks featuring @equalrights4parasites, summer cicadas featuring @pufferchung, New Jersey Mushroom hunting featuring @gabrielamushroom, giant carabid beetles featuring @jtweed, and stories about bee research by @kjhung and research at the San Diego Natural history museum by @sclerobunus. We also enjoyed @emilymstone’s excellent natural history article on observing a mite on a beetle larvae and @greatauntcarolyn’s article about photographing butterflies around Ontario.

Lastly, there were several pieces on iNaturalist's recent generous grant from the Gordon and Betty Moore Foundation including in Science Magazine, on KQED Forum, and in BayNature. This grant is matching donations through December. If you’re inspired to donate, you can do so via the link below.

Donate to iNaturalist

Thank you to everyone who participated in iNaturalist this September!