Our Observation of the Week is this Pseudobiceros bedfordi flatworm, seen in Hong Kong by @josylai!

A scuba instructor living in Hong Kong, Josy Lai tells me that early in her career she focused on teaching skills and general marine knowledge, 

but then, there had always been times that we saw mysterious marine organisms and I didn’t quite know what they were. Having no background in biology or ecology, I have spent time reading marine guides and browsing the internet. The more I study, the more I am amazed by nature; the more I learn, the more I realize that I don’t understand much and would love to know more.

One of the amazing creatures she recently encountered was the flatworm documented in the observation of the week, which Josy saw when she was exploring a remote dive site off of Hong Kong island.

In general, underwater visibility in Hong Kong varies from 0.5 meters to 5 meters. During that dive, visibility was surprisingly good and visibility hit up to 10 meters. Remotely, I saw something “floating” in mid-water. At one point I thought that was a broken piece of plastic bag, and wanted to grab it. When I swam closer, I noticed immediately that it was a beautiful Pseudobiceros bedfordi “dancing in the air”! I spent five minutes watching its beautiful movement and almost forgot to photograph it. I was lucky enough to take a few pictures of this gorgeous marine flatworm until it finally landed on a rock.

Known colloquially as the Persian carpet flatworm or Bedford’s flatworm, Pseudobiceros bedfordi’s dorsal side is intricately patterned with both large stripes and many, many tiny spots. Ranging from the Maldives to Fiji, these 8-10cm long worms feed on tunicates and crustaceans and, yes, undulate beautifully when they swim. 

When it comes to reproduction, these hermaphroditic invertebrates engage in behavior known as “penis fencing,” as do many other flatworms. Generally, each individual tries to use one of its two penises to stab and inseminate the other without in turn getting inseminated. The inseminated individual usually takes on a greater physiological burden, which we believe is why they try to avoid being inseminated. However, in some species both individuals often inseminate each other. 

“As I spend most of my time scuba diving now,” says Josy (above, with a sea fan), “my current nature interest is marine animals. I also conduct marine conservation workshops in Hong Kong, and organize underwater and beach clean-ups on a regular basis.”

She joined iNat just over a year ago and tells me it’s

an important platform for me to learn more about marine life. Biologists and citizen scientists have helped me a lot in suggesting IDs of the animals I see. From iNaturalist, I learn about what other divers in the Indo-Pacific region have observed; I learn about the characteristics and sometimes habitats of the animals. When I dive now, I have a clearer idea of what to look for in a certain environment and I look more into details. I enjoy sharing my observations on the platform, and just can’t wait to receive responses or ID suggestions from other users after uploading a new observation. 

I like the way that iNaturalist has created an active community for professionals and citizen scientists to exchange ideas and knowledge. I am so glad to be part of your community.

(Some quotes have been lightly edited for clarity.)

- You can follow Josy on Facebook and Instagram.

- And you can check out her diving videos and watch her sing on her YouTube channel!

- David Attenborough and the BBC show us these worms in action.

- See the diversity and beauty of iNat’s flatworm observations!

Our Observation of the Week is this Marble Cone Snail (Conus marmoreus) eating a Chocolate Spotted Auger Snail (Terebra subulata) in the Marshall Islands! Seen by @uwkwaj.

The uwkwaj iNat account belongs to Scott and Jeanette Johnson, and Scott wrote back to me for this week’s blog post. “I guess we have been interested in nature as far back as both Jeanette and I can remember,” says Scott. Growing up in Los Angeles, Jeanette would seek out insects in her backyard and Scott sought out snakes, turtles, and frogs in a nearby pond in New England. “But what focused each of us on marine life was moving to Kwajalein Atoll in the Marshall Islands,” he explains.

In my case, my parents dragged me out there when I was 8 years old, and I immediately became fascinated by the water and reefs surrounding these small islands. Some years later, Jeanette and I returned to Kwajalein and spent the next 30 years diving and documenting [the atoll’s] marine life. Permanently living in an area gave us a more complete overview of the reefs and animal populations that come and go than sporadic expeditions ever could. While my main focus was mollusks, particularly nudibranchs, we are interested in and photograph anything that gets in front of our cameras. Unfortunately, we were also witness to the decline in health of the reefs. In all my years out there, I did not see any sign of coral bleaching until 2009, when Kwajalein had its first recorded episode. It happened again in 2013, 2014, 2016 and 2018, with more coral failing to recover each time. I fear that our documentation of the marine life of the Marshall Islands might turn out to be just a record of what used to be there.

Scott has mostly shot underwater video since 1994, and the image you see at the top of the page is a still from footage he took in 2009. 

For years we mostly dove the coral rich lagoon pinnacle reefs or the steep seaward drop-offs, but after a while discovered that lagoon sand flats bearing dense fields of Halimeda and other algae were surprisingly diverse, and the Halimeda patches became some of our favorite sites. Sand dwelling mollusk species such as cones and augers were abundant. I came across this particular scene of a Conus marmoreus sucking the animal out of a Terebra subulata in the sand around one such algae patch.

Famously venomous, cone snails are all predators that immobilize their prey by injecting toxins via their harpoon-like radular tooth, then consuming them. Marble cones seem to specialize in eating other mollusks. Chocoloate spotted auger snails are also predatory and venomous, using a similar venom delivery mechanism to that of cone snails. However, they predate marine worms.

Scott and Jeanette (Scott middle, Jeanette below) joined iNaturalist just over three years ago and have since added nearly eight thousand observations. The oldest photos are Scott’s, taken from his time as a graduate student in Hawaii and as a research staff member at the Mid-Pacific Research Lab at Enewetak. Most of the photos since the mid-90’s are Jeanette’s.

“Although I knew it existed and had looked at a number of species there,”  says Scott, 

I became more aware of iNat when a researcher asked me to post my images of a Marshall Islands endemic anemonefish there so they could be referenced in a paper in preparation. That made me realize that iNat would likely be a great way to make our observations more widely available and would almost certainly have a greater longevity than our own website on Marshall Islands marine life. I find it very useful as a means to compare variation within a species across its range; I’ve always been interested in biogeography and how and where species are distributed.

- Scott and Jeanette’s website can be found here.

- PBS’s Deep Look has a great video breaking down cone snail predation.

- Check out the remarkable observations Scott and Jeanette have posted from their archives!

- This paper goes over the potential benefits and drawbacks of using conotoxins in pharmacology.

- Speaking of Hawaii, check out this old Observation of the Week, documenting a crab that’s found a curious to place to call home.

We’ve released a new computer vision model for iNaturalist. This is our first model update since July 2021. The iNaturalist website, mobile apps, and API are all now using this new model. Here’s what’s new and different with this change:

• It includes 55,000 taxa (up from 38,000)
• Hybrid taxa are excluded
• It’s more likely to suggest the correct taxon

To see if a particular species is included in this model, you can look at the “About” section of the taxon page.

Here’s more details about what changed and why:

It’s a lot bigger

Our previous model included 38,000 taxa and 21 million training photos. When we announced that we were kicking off a new vision training job in October, we planned to train a new model on 47,000 taxa and 25 million training images. However, in November, after training had started, we realized that we could make improvements to how we were picking our training data that would improve the quality of the vision models. After making those changes, we re-started training with over 55,000 taxa and over 27 million training images.

We changed a few things about how we generate training data

As I mentioned above, we made some changes to how we generate our data export for the training data. Patrick’s new code for making the data export has a few improvements worth mentioning:

• It includes more taxa right on the borderline of inclusion (taxa with at least 100 photos but fewer than 100 observations will now make it into the export, but didn’t previously),
• It limits the number of photos used per observation (max 5 photos from each observation), and
• It is more clever about picking distinct photos for each taxon, preferring to choose photos from as many different observations as possible. This increases the visual diversity of the dataset, which in turn helps the computer vision model learn.

Our testing suggests this new approach produced a much better dataset. Overall accuracy numbers went up with this model, even though the model’s job has gotten considerably harder (choosing between 55,000 options instead of between 38,000 options).

We excluded hybrids this time

We also chose to exclude hybrid taxa for this training run. The previous production model, released in July 2021, was the first to have significant amounts of training data for many hybrid taxa. Including those hybrid species in the model made it much less likely that the first suggestion would be correct for clades like Genus Anas which includes Mallard Ducks, the most-observed species on iNaturalist.

Our CV models are trained to recognize discrete, mutually exclusive, distinct taxa. Given a photo, there should be one right answer as to what discrete taxon it belongs to. Hybrid taxa, while being potentially useful taxonomic entities, make it hard for our CV models to visually distinguish hybrid taxa from their hybridized origins, and to confidently recommend any of these taxa in any scenario given their visual overlap. So we decided to remove hybrid taxa thinking it would make the classifier’s job easier and thus improve accuracy, and our testing showed this to be the case. We believe it’s better to accurately identify distinct species than inaccurately identify hybrids and their origins. This is a reminder that taxonomy is an abstraction trying to put hard edges on what is often a continuum. Hybrid taxa are good examples of where this abstraction is an oversimplification but our CV doesn’t do well with some of these edge cases like hybrids and we’ve found the benefits from simplifying outweigh the loss in accuracy from trying to accomodate hybrids.

Future work

In addition to training and deploying this new model, we’re working on a few areas to generally improve the quality of suggestions.

First, we are continuing to work on new approaches to improve suggestions by combining visual similarity and geographic nearness. We can’t share anything concrete right now, but we’re excited to be working with some Visipedia researchers on this.

Second, we are running more experiments to improve training speed. Updating our vision models once or twice a year creates a lot of pressure on each model. There are techniques for speeding up training (such as transfer learning and finetuning) that we haven’t been completely taking advantage of, that may someday allow us to release new models every month or two.

Third, we will continue to look at new approaches to including or excluding nodes in the taxonomy. For example, we may someday find a technique to include hybrid species without causing significant reductions to suggestion accuracy for neighboring species. Or we may be able to identify other ranks or nodes in the taxonomy that should not be represented directly in the vision model.

Fourth, we’re still working to compress these larger models for on-device use. The in-camera suggestions in Seek continue to use the older model from March 2020.

We couldn't do it without you

Thank you to everyone in the iNaturalist community! Sometimes the computer vision suggestions feel like magic, but it’s all people. None of this would work without the millions (!) of people who have shared their observations and the knowledgeable experts who have added identifications.

Beyond adding observations and identifications, here are other ways you can help:

• Share your Machine Learning knowledge: iNaturalist’s computer vision features wouldn’t be possible without learning from many colleagues in the machine learning community. If you have machine learning expertise, these are two great ways to help:
• Participate in the annual iNaturalist challenges: Our collaborators Grant Van Horn and Oisin Mac Aodha continue to run machine learning challenges with iNaturalist data as part of the annual Computer Vision and Pattern Recognition conference. By participating you can help us all learn new techniques for improving these models.
• Start building your own model with the iNaturalist data now: If you can’t wait for the next CVPR conference, thanks to the Amazon Open Data Program you can start downloading iNaturalist data to train your own models now. Please share with us what you’ve learned by contributing to iNaturalist on Github.
• Donate to iNaturalist: For the rest of us, you can help by donating! Your donations help offset the substantial staff and infrastructure costs associated with training, evaluating, and deploying model updates. Thank you for your support!

Our Observation of the Week is this bird’s nest fungus (potentially Woolly Bird's Nest Fungus (Nidula niveotomentosa)), seen in Chile by @eduardomunozfotografia!

Last July, Eduardo Muñoz traveled to Chile’s Nahuelbuta National Park with friends Vicente Valdés (@vicentevaldesguzman) and Dalila Parraguez (@dalaila).

We walked the trails of the park in search of fungal life, where we found a great variety of beautiful species of fungi of different sizes, shapes, and colors, one of them was [this Nidula], which captivated me the first moment I saw it. Its shape and similarity to a bird's nest is impressive and I couldn't pass up a fungus with that different aspect.

I photograhed its two stages: immaturity and maturity. When it is mature, it lacks an epiphragm and reveals its interior, where the spores are contained.

Bird's Nest Fungi (Family Nidulariaceae) can be found in much of the world. They are saprobes, meaning they use decaying matter for their nutrients, and you can findt their tiny fruiting bodies on rotting logs and soil. As Eduardo noted, the fruiting bodies are covered by an epiphragm (or membrane), which comes off when they’re mature, exposing the egg-like periodoles, or spore-containing structures. When water drops hit the “nest” at the proper angle, the periodoles are forced out. Check out this fantastic paper (which includes slow motion video!) for more information about spore dispersal. Note that the species studied in that paper are in a different genus, Cyathus.

Now twenty-three years of age, Eduardo (above) credits his father for his interest in nature as he “took me to the hills to watch birds and look for the insects that live there.” He’s studied ecotourism and tells me

I dedicate myself to educating others about the environment with my photographs through social networks. My mission is to photograph all the possible biodiversity of Chile and thus generate awareness for the care of the environment and its species. As well as contributing to science, thanks to my records and observations.

Speaking of which, Eduardo joined iNat about three years ago and says 

I currently use INaturalist to recognize species that I observe and photograph, search for information, help identify species, and contribute with my observations in citizen science.

The use of iNaturalist has undoubtedly greatly improved my vision of the natural world, since the contribution and great content on the page is impressive, and it also helps me day by day to learn about new species that surround me.

- You can find more of Eduardo’s photos on Instagram.

- Why not take a look at the over 24k bird’s nest fungus observations on iNat?

In preparation for increased iNaturalist activity during the upcoming City Nature Challenge, iNaturalist will implement some temporary changes. From April 27 to May 9, we will temporarily restrict some types of content on iNaturalist that are more intensive. Most users will not notice these changes because they do not directly impact observations, identifications, comments, or projects. However, for anyone planning to use the features below, we want to give advance notice so you can plan and prepare accordingly.

Large places cannot be created or edited
Starting on April 27, any new or edited places must contain fewer than 10,000 observations and be smaller than roughly the size of West Virginia (~24,000 square miles or 62,361 square km). If you try to add or edit a place above these thresholds, it will give you a warning message.

All places added or edited during this time may experience extended times to reflect the edits or collect all of the observations. If you can delay adding or editing places, please do so.

“Search external providers” disabled
If you enter a taxon name that can’t be found in iNaturalist, normally you can “Search external providers”. This feature will be temporarily disabled to prevent the addition of new taxa that cannot be curated during this time period (see below).

Taxon changes & ancestry edits paused (applicable for curators only)
No taxon changes or edits to taxon ancestry (including grafting taxa) can be implemented starting April 26. If you try to do this, you’ll get a message that such changes are temporarily unavailable. You can still draft taxon changes and save them to be committed after the restriction.

These temporary limitations will be in place through May 9, which includes the observation period of the City Nature Challenge as well as the upload/identification period.

Other activities that are not restricted but should be deferred if possible:
-csv uploads: If you are uploading a csv of observations, expect considerable delays. Do not attempt the same upload more than once.
-csv data downloads: If you are trying to download a csv of observations, expect considerable delays. Do not attempt the same download more than once.

Thank you all for your patience and understanding as we prepare for this busy time of year.

Our Observation of the Day is this Usambara Eyelash Viper (Atheris ceratophora), seen in Tanzania by @jvl!

“I was born and raised at the foot of Mount Kilimanjaro, in Rombo district, Tanzania, near the Tanzania-Kenya border,” says John Lyakurwa, a researcher with the University of Dar es Salaam. Surrounded by world class protected areas, John says he often encountered wildlife as he was growing up. “Meeting chameleons, geckos and snakes while hand picking coffee was a very common experience at the age of 5,” he explains.

When I joined the university, I was deeply inspired by Prof. Kim Howell who made me realize how little is known about reptiles and amphibians, and how most snakes and some lizards (e.g. chameleons and agamas) are misunderstood and killed needlessly. Since then, I decided to photograph endemic amphibians and reptiles and use them to raise awareness, which partly involves traveling to remote places across the country, photographing unique and endemic species and share with people who have no access (majority) to these localities.

A few weeks ago, John and Ardgard Mwamgeni visited the East Usambara Mountains to photograph reptiles and amphibians. 

After two days and nights of hiking inside the bush and photographing a couple of species, it was on the last day that we found a medium sized, nicely colored subadult A. ceratophora coiled near the leaf node of an Aframomum plant (~1m from the ground). The snake was calm and was left undisturbed after the observation.  

Endemic to the Eastern Arc Mountains of Tanzania, this small arboreal viper “used to be common throughout its range, but has been difficult to find recently, mostly due to illegal collection for pet trade,” explains John. While he’s seen it several times in the Udzungwa Mountains, last March was his first encounter with this species in the Usambara Mountains (it’s type locality). 

A member of iNat since 2017, John (above) joined the community so he could share his nature photos and get ID help (especially for invertebrates, which he’s not as familiar with).

I thought that on iNaturalist my observations could be identified by people using the platform and other people could have access to my records (rather than just keeping them in my hard drive for myself). Later I learnt that through iNaturalist, one can document the biodiversity of an area and the project can be customized depending on specific needs/targets. INaturalist has enabled me to learn many things related to nature, link with specialists who are in the same field with me or are interested in observations I post, and has broadened my interest to even photograph arthropods.

(Photos of John by Ardgard Mwamgeni (middle) and John Lyakurwa (bottom). Some quotes have been lightly edited for clarity.)

- Check out John's publications on ResearchGate. He also co-authored a beautiful book, Amphibians of the Tanzanian forests, which is available to download for free here.

- African bush vipers are pretty stunning, check out these observations of them!

- Some other viper encounters have been observations of the week, like this bamboo viper by @prasannaparab and a Siamese peninsula pitviper by @rharris70!