Via Mongabay, an article on an initiative in Sri Lanka to identify individual marine turtles in the island’s waters which uses photos taken by recreational divers to build up a database based on their unique facial patterns:
The Polhena reef and the surrounding shallow seas in southern Sri Lanka are home to a number of marine turtles that stay there year round. Randunu Dimeshan, a managing partner at the Polhena Diving Center who frequently swims in the area with his diving clients, has frequently encountered these turtles and been able to identify a few individuals from notable physical features such as a scar on a flipper or a damaged carapace.
Meanwhile, Chathurika Munasinghe, a marine biologist at the Ocean Conservation and Education Alliance (OCEA), had recently returned from a research project in the Maldives armed with a special set of knowledge and skills: identifying sea turtles based on photo identification.
Dimeshan met Munasinghe during a discussion of underwater cleanups, and this gave rise to the idea of setting up a similar citizen-science initiative in Sri Lanka. After preparations, the pair launched the Sri Lanka Turtle ID project in August 2019, several months before the COVID-19 pandemic struck.
“Turtles have specific facial scale patterns that are unique to each individual,” Munasinghe told Mongabay. “Just like fingerprints [on humans], the facial scale patterns can be used to identify turtles.”
To build up this database of turtle mug shots, the researchers needed, well, mug shots: clear photographs of both sides of the face and, optionally, an image of the shell. They worked with dive centers to get the latters’ clients to take photos during dives. These photos can be uploaded to the Turtle ID project website, where special software is able to pick up facial patterns and compare them with patterns from already identified individuals stored in the database. If the facial pattern is new, the photo contributor is given the chance to name the new individual.
Facial identification
The Turtle ID project has so far identified 18 hawksbill turtles (Eretmochelys imbricata) and three green turtles (Chelonia mydas), all of them female. Of the seven marine turtle species found around the world, five can be observed in Sri Lankan waters. The olive ridley sea turtle (Lepidochelys olivacea) is the most common, but these are mostly found farther out in the open ocean.
“The hawksbill turtles stay closer to the coral reefs as they prefer to feed on sponges that are mostly found in association with coral reefs,” Munasinghe said. “These are the areas where most of the researchers and recreational tourists dive, so the chances of encountering hawksbill is higher.”
Dimeshan said they’ve put a name on all of the identified turtle faces. Tammy was the first to be identified and was named after the nickname of a mutual friend of the project founders. Some of the others are Alice, Avondster, Shelah, Polly, Keyara, Olya, and Chuta.
The project allows the observer who uploads photos of a previously unidentified turtle to name it. Image courtesy of Terney Pradeep Kumara.
Creating a database of sea turtles to assess the population size of each species is the main aim of the initiative. Some of the turtles use Sri Lankan waters as feeding ground, especially where reefs abound, so learning how they use certain reefs for feeding and breeding is another project aim. The team also expects the database to shed some light on turtle migratory patterns in the long run.
Munasinghe said the idea was inspired by her firsthand experience in the Maldives initiative. There, a similar project begun in 2011 has to date compiled individual records for more than 1,270 turtles, according to Marine Savers, the group running the initiative. The results indicate that the Maldivian hawksbills remain in their home reefs throughout the year, traveling only between reefs less than 2 kilometers (1.2 miles) apart, while green turtles tend to use multiple reefs for feeding.
“The Sri Lankan data show that we mainly observe female and juvenile turtles on the reefs, with few males of either species being spotted by our researchers,” Munasinghe said.
The technique of using facial pattern-based identification was introduced by French scientist Claire Jean of the Kélonia aquarium in the Indian Ocean island of Réunion, an observatory specializing in marine turtles.
Before then, most studies on marine turtle populations relied on capturing animals and tagging them with a marker such as a flipper tag or transmitter, which can be costly. Tags are also dif?cult to apply to turtles, as they remain in the water unless they reach beaches for nesting. So almost all physical tags are generally applied to nesting females. The photo identification method is both more cost effective and avoids putting the animals under any stress, Jean wrote in a 2010 paper.
Turtles get unique IDS
The Sri Lanka Turtle ID project uses open-source software called I3S pattern at its back end, employing machine identification of the turtles. The process is simple: reference points are first taken at the tip of the nose, the inner edge of the eye, and the furthest scale. The software then outlines the other identification zones and automatically selects 35 points within the zones as identification marks. Once this is completed, the program shows which turtles have the closest match based on the identification marks, and a user can either mark it as a previously identified turtle or a new individual.
The same technique can be used to identify individuals from other species, including whale sharks based on spot patterns.The technology is also thought to have the potential to identify large rays.
Many of Sri Lanka’s reefs have resident turtles, and some local groups have started feeding them in order to habituate them for tourism purposes. Marine biologists say they disapprove of this practice. Image by Malaka Rodrigo.
Munasinghe and Dimeshan have led several introductory sessions for researchers and recreational divers, their main focus being to get dive centers to support the photo ID project.
But the Turtle ID project got off to a rocky start, with COVID-19 hitting just months after its launch. In Sri Lanka, the government imposed a nationwide lockdown in early 2020, forcing the closure of dive centers and scuppering the Turtle ID project. Marine research in general has come to a halt in the year and a half since then.
In recent months, however, there’s been a greater sense of urgency around turtle conservation in Sri Lanka, following the sinking of the MV X-Press Pearl cargo ship off the western coast of the island in early June. The ship was carrying a cargo of nitric acid and plastic pellets, among other items, and was also loaded with 378 metric tons of bunker fuel. In the months since its sinking, more than 200 marine turtles have washed up dead on the beaches and in the waters in the vicinity.
Munasinghe said she’s hopeful of being able to dive soon and to update the database, with a view to contributing to turtle conservation efforts.
,
Read More » Posted in Open Source | Comments Off on In Sri Lanka, Biologists and Divers Build a Facebook for Sea Turtles
Via Mongabay, an article on how – armed with data and smartphones – Amazon communities are fighting deforestation:
>Equipping Indigenous communities in the Amazon with remote-monitoring technology can reduce illegal deforestation, a new study has found.
> Between 2018 and 2019, researchers implemented technology-based forest-monitoring programs in 36 communities within the Peruvian Amazon.
Compared with other communities where the program wasn’t implemented, those under the program saw 52% and 21% less deforestation in 2018 and 2019 respectively.
> The gains were concentrated in communities at highest risk of deforestation due to threats like logging and illegal mining.
Teaching Indigenous communities in the Amazon to tap on remote-monitoring technologies during forest patrols can reduce illegal deforestation, a new study has found.
Researchers, whose work was published July 12 in the Proceedings of the National Academy of Sciences (PNAS), implemented technology-based forest-monitoring programs in 36 communities in Loreto, the northernmost department of Peru, between 2018 and 2019. They trained and paid three representatives from each community to patrol their forests monthly and verify reports of suspected deforestation using a smartphone application and satellite data.
Compared with 37 other communities in Loreto where the program wasn’t implemented, those under the program saw 52% and 21% less deforestation in 2018 and 2019 respectively. The gains were concentrated in communities at highest risk of deforestation due to threats like illegal mining, logging, and the planting of illicit crops such as coca to manufacture cocaine, the researchers found.
The collaboration between Rainforest Foundation US (RFUS), the World Resources Institute (WRI), Indigenous leaders and independent researchers is the latest in a growing body of research that says recognizing and protecting Indigenous rights is the most effective way to preserve natural rainforests. In Latin America, studies have shown Indigenous people to be by far the best guardians of forests in the region, with deforestation rates up to 50% lower in their territories than elsewhere.
One-third of the Amazon Rainforest falls within formally acknowledged Indigenous peoples’ territories. Community-based forest monitoring programs coupled with enforcement support from local officials could save one-fifth of the 2.7 million hectares (6.7 million acres) of rainforest in Brazilian and Peruvian Indigenous territories likely to be lost over the next decade, RFUS estimated.
Deforestation alerts from satellite data have long been publicly available. WRI’s Global Forest Watch (GFW) tool relies on an algorithm developed by university researchers to detect changes in forest cover through satellite imagery. In Peru, the national Geobosques platform uses GFW data to issue early alerts of suspected deforestation.
However, these alerts rarely filter down to remote rainforest groups lacking reliable internet access, resulting in communities often detecting illegal deforestation activities only when they are well underway and difficult to halt.
“The whole point is to put the deforestation information into the hands of those most affected by its consequences and who can take action to stop it,” Tom Bewick, who is the Peru country director for RFUS and who was involved in the study, said in a statement.
During the two-year study, researchers hired couriers to traverse the Amazon River and its tributaries every month to deliver USB drives containing Geobosques reports of suspected deforestation to remote communities.
Trained representatives, or monitors, would then upload this information into a specialized smartphone application, which they used to navigate to the locations of forest disturbances during their monthly patrols. Where they identified cases of unauthorized deforestation, monitors would take photos as evidence and flag them to the community, which could then decide to report it to local authorities.
Monitors use the smartphone app Locus Map to identify GPS coordinates of deforestation for their regular patrols. Photo credit: Cameron Ellis
“We are helping them set up this system by which they can collect the evidence but our hope is that then we walk away,” Suzanne Pelletier, executive director of RFUS, said in a video. “They can then train others and be the model for thousands of other communities across the Amazon.”
Over the two-year period, communities under the program saved 456 hectares (1,127 acres) of rainforest, preventing the release of more than 234,000 metric tons of carbon emissions at a cost of $5 a ton. This makes it slightly more expensive than the $4.30 a ton average price of nature-based, forest management carbon credits in 2019, according to data from Ecosystem Marketplace.
But while nature-based credits have traditionally been plagued by the problem of leakage — where ecosystem conservation projects, even if successful in one area, often shift deforestation to another location — the researchers observed no such displacement of deforestation for the communities in their study.
They theorized this could be due to the inaccessibility of the forests in Loreto. “In the region that we study, in the general absence of roads, most transportation occurs by boat. As a result, the areas most vulnerable to deforestation are located close to navigable rivers,” they wrote in their report. Since Indigenous communities in Loreto also tend to live along the river, community-based forest-monitoring programs increase the cost of resource extraction, they said.
A Kichwa monitor fills out a report confirming an occurrence of illegal deforestation after returning from a forest patrol. Photo credit: Melvin Shipa Sihuango
“The study provides evidence that supporting our communities with the latest technology and training can help reduce deforestation in our territories,” Jorge Perez Rubio, president of the Indigenous group Regional Organization of the People of the Eastern Amazon (ORPIO), said in a statement. ORPIO worked with RFUS and WRI to implement the forest-monitoring programs in the study.
“Our network is ready to partner with Rainforest Foundation US to apply this technology-enabled model to our community forest protection initiatives basin-wide,” Gregorio Mirabal, general coordinator of the Coordinator of Indigenous Organizations of the Amazon Basin (COICA), said in a statement. COICA, which was not involved in the study, is an umbrella association for Indigenous organizations in the Amazon lowlands, of which ORPIO is a part.
Via The Wall Street Journal, an article on how conservationists hope newly launched open-source tools will lend more insight into endangered species and the effects of climate change:
Cameroon’s Dja Faunal Reserve is one of the largest undisturbed areas in Africa. Elephants trample through the trees, monkeys call to one another, and insects hum, all with barely any human interference. The lack of roads and dense forest have protected the biodiversity there, but the same remoteness has left the impact of climate change on the area little understood. Instead of trying to set up camp for a few weeks at a time, researchers are turning to tools they hope will gather data delicately and indefinitely.
“So often the science stops when you leave these places,” says Shah Selbe, a former rocket scientist and the co-founder of FieldKit, a Los Angeles nonprofit building open-source tools to gather conservation data. “If we can create low-cost tools that make monitoring easier, we can start to get more data.”
Big Data has transformed industries from finance to drug discovery. Conservationists, however, haven’t had the same access to deep data sets because of the difficulty and cost involved in gathering data in the wilderness.
To help, startups are developing open-source technology built to monitor the environment, from equipment that records the sounds of the forest to devices that collect data on weather conditions. The hope is that open-source tech will make it cheap enough to gather data. In turn, the data could lend greater insights into where to focus efforts to save endangered species and tackle the effects of climate change.
The open-source movement advocates sharing design information so that anyone can inspect and improve upon the tools that are built. Often, the groups behind the tech are nonprofits, typically resulting in hardware and software that are cheaper than commercial counterparts. Researchers can adapt the tools without worrying about breaking a user agreement or warranty. With more adaptable tools, projects could range from learning about a single species to ecosystems as large as the polar regions.
FieldKit makes a water-resistant device about the size of a coffee-table book that allows users to gather data on water temperature, weather conditions, pH levels and more. Launched this month, the $150 device includes a computer chip that can accommodate a variety of environmental sensors and add-ons, which run $50 to $205 each. FieldKit also offers two prebuilt models, one for water quality and another for weather conditions. A prebuilt model for air-quality is coming soon. Prices include a small margin to support Conservify, FieldKit’s parent nonprofit, which currently relies on grants to fund its operations.
The units were built to withstand environments from rainforests to freezing tundra, and to collect data for weeks or even months at a time, says Mr. Selbe. He expects that many users will be full-time researchers affiliated with nonprofits or universities, but he says it was important to sell prebuilt kits to expand the potential audience to students and individuals without technical backgrounds.
“You can take the FieldKit out of the box, download the app, and be monitoring in five minutes,” he says.
FieldKit’s website hosts the data collected on the devices. Mr. Selbe says that he hopes that most researchers will put their data on the platform, enabling even researchers and enthusiasts without FieldKits to use it. There is an option to disguise location information to protect endangered species or sensitive locations, Mr. Selbe says.
Much of the conservation technology builds on open-source tools made for more general uses, such as Raspberry Pi’s simple, single-board computer chips and Tensorflow, Google’s machine-learning platform. These advances have been essential for the cash-strapped environmental research community, says Alasdair Davies, co-founder of the Arribada Initiative, a nonprofit that builds open-source technology with partners who require technical expertise.
“We are riding the wave going on in the commercial space and repurposing it for conservation,” Mr. Davies says.
More devices in the field mean more data, but analyzing all of it can prove impossible for solo researchers and small teams. Citizen scientists have volunteered to label, transcribe or otherwise organize data on scores of different projects, including labeling wildlife caught on camera in a New York forest and transcribing old weather logs from the U.S. Navy. Some research even requires volunteers, such as Arribada’s Carnivore Bytes project, which aims to learn more about wild dogs in part by gathering data on people’s pets.
Carnivore Bytes in January sent recording devices that clip onto collars to hundreds of dog owners. With an app, volunteers note when their dogs are doing things like panting, playing or eating. Eventually, the data is meant to help train an algorithm to go through wild-dog noises for insights into how climate change is impacting their lives. For example, wild dogs making fewer eating noises could suggest that prey is scarce and hunger is becoming an issue.
Open Acoustic Devices, based in the U.K., sells acoustic recording devices called AudioMoths for $60, compared with hundreds or thousands of dollars for a commercial version. Since launching in 2018, AudioMoth has sold about 20,000 units, according to co-founder Andy Hill. The sales, along with occasional work building custom software to go along with AudioMoth projects, allow Mr. Hill and another founder to work on the technology full-time, he says.
AudioMoths have been used to record thousands of hours of sound. In one project, researchers are recording marine mammal vocalizations in an effort to decode the different sounds that dolphins, manatees and whales use to communicate. Mr. Hill estimates that projects have, on average, used 10 to 100 AudioMoths, which has greatly improved the quality and quantity of data.
“Without open-source technology, you are going to be limited to one-species studies with one device,” he says. “That hasn’t really gotten us anywhere when it comes to conservation.”
Open-source technology is also allowing more people to engage in environmental research, says Lydia Gibson, an ecologist and National Geographic explorer who uses it in her fieldwork.
Still, conservation isn’t just about technology, she says. It is also about the people who use the tools and the local communities that participate, actively or not. Conservation has gotten better about including more voices and respecting local expertise, and those strides can’t be lost as shiny new tools are added to research tool kits, Ms. Gibson says.
“A focus on technology needs to include a holistic view of how it’s used,” she says. “Technology on its own is not the solution.”
,
Read More » Posted in Open Source, Sensors | Comments Off on Better Data, Cheaper Tech Promise to Unlock Nature’s Secrets
Courtesy of Outside, an interesting article on how – over the past two decades – eBird has become the go-to online platform for scientists and hobbyists alike to upload and share bird observations:
In July 1992, two Danish birders visiting Patagonia, Arizona reported the first-ever, mega-rare cinnamon hummingbird in the United States. Back then, reporting rare birds required phoning in observations to a “rare bird phone tree,” usually via the nearest pay phone—and hoping that word got out. Sometimes it did, sometimes it didn’t. In this case, a couple of other out-of-towners—a birder from Mississippi and another from Nebraska—saw the species. The Nebraskan photographed the hummingbird, flew home, developed the slide film, and snail-mailed photos to the Arizona committee in charge of validating unusual sightings. Only then did word spread, but it was too late: the hummingbird was gone, and Arizona birders missed it.
This kind of tragedy would never befall Arizona birders today. Now, within minutes of seeing a rarity, birders can text friends, alert listservs, post sightings to Facebook rare-birds groups, and—the choice of many—submit observations to eBird, a global online database.
At its most basic level, eBird documents bird sightings. A team at the Cornell Lab of Ornithology created the platform in 2002, and it became widely used by birders within a few years. As of 2020, it has collected more than 860 million global bird observations from over 597,000 registered eBirders. By sheer numbers alone, eBird is one of the world’s largest citizen-science projects. It is now used to understand species distributions, population trends, migration pathways, and even habitat use.
“If used properly, it should be a tool to understand bird populations at scale in ways we never have before, and to apply that to conservation actions,” Marshall Iliff, an eBird project leader, told me. Scientists use eBird’s open-access data to study evolution and movement of invasive species and to highlight the importance of public lands in conservation. The eBird team has also created conservation-oriented status and trend maps for hundreds of species, and eBird data are used to create live bird-migration forecasts.
At least 120 million observations are submitted per year, many through the handy eBird app, a kind of Strava-Yelp-Pokémon Go hybrid for birders. The app doesn’t ID birds for you—Cornell offers another app called Merlin for that—but instead provides an easy way to record and upload the birds you spot. To log sightings, you start a checklist (similar to the way you’d start a run on a smartwatch) and the app automatically pulls your location via GPS. You can choose hot spots near you, which generate lists of species you’re likely to see created from data submitted by users in those areas. The app tracks time and distance traveled while you “tick” species and numbers of birds seen and heard. It even lets you keep an offline checklist, so you aren’t inconvenienced without cell service. On the web platform, users can upload photos and audio recordings to beef up checklist documentation. Once submitted, the observations join thousands of others being made on the platform at any given time.
The scope and accessibility of eBird make it a resource for birders and scientists alike. The majority of eBirders use the platform as a handy bird-logging tool. I study birds for my Ph.D. at the University of New Mexico, and I also watch them recreationally. I use eBird almost daily for everything from tracking how far I walk while scanning treetops, to planning vacation birding spots, to scouting remote Andean field sites for my doctoral work. Anyone can review lists of species in hot spots like the Ramble in Central Park stretching back decades, study maps of where birds are seen, analyze how frequently certain birds appear at different times of the year, and peruse photos and audio recordings from all over the globe.
Undeniably, eBird has changed birding culture, a quirky world already full of strong opinions. It brings birders together and allows for rapid information sharing. It’s also created new—and sometimes contentious—etiquette and social dynamics.
The closest thing to an official guide for birding etiquette is the American Birding Association’s Code of Birding Ethics, which emphasizes a few basic tenets that can be summarized as: Respect fellow birders, their diverse interests, and skill levels. Welcome and encourage beginners. Respect birds and other wildlife. Don’t trespass on private or restricted property. Be mindful of space and privacy when birding in groups. (Since the pandemic, a new set of guidelines on birding and social distancing has been added to the code, titled “Keep your eyes on the sky and your butt close to home!”)
Birders are typically friendly, both in person and online, with email exchanges often ending in well-wishes of “Good birding!” But a code of ethics is necessary because, as with any activity that can become competitive, birding has a dark side. Rivalry, animosity, and ego have long been hallmarks of the bird world. Even the famous naturalist John James Audubon plagiarized and invented species to convince members of the English nobility to promote his work. Birders sometimes go to semi-desperate lengths to track down birds, and online platforms like eBird that rank birders and sightings, akin to athletes on leaderboards, can amplify competition.
Although eBird is primarily an observation tool and a scientific database, the site still allows users to size each other up: anyone can view rankings of the top eBirders in different hot spots, counties, states, and entire countries. You can even peruse a list of the top 100 eBirders in the world. These types of competitive lists have birthed trends like endless Big Years, in which birders constantly compete to see who can spot the most species in a year. In turn, such fads have spurred counterinitiatives, like the five-mile-radius challenge, which encourages birders to enjoy birds in local areas rather than seeking them out in far-flung places. Local birding has soared during the COVID-19 pandemic, as many work from home and explore their own backyards.
Screenshots from the author’s eBird app showing checklists and hot spots in her home state of New Mexico (Photo: eBird)
If an eBird user makes their profile public, other eBirders can view their recently submitted checklists and photos. Essentially, this means that birders can keep tabs on one another. Last October, when a friend was in Belize, I lived vicariously through his trip by peeking at his eBird checklists each day, and they brightened my mood amid dropping fall temperatures in New Mexico. I’ve also received unexpected text or Facebook messages from birders with quips like, “Looks like you had an awesome day!” after they saw checklists I submitted. These interactions shouldn’t be surprising, given the public nature of eBird data, but it occasionally strikes me as odd that people I don’t know well can see exactly where I walked for eight hours and exactly how many Wilson’s warblers I counted while I did so.
The eBird database also maintains a frequently updated alerts bulletin called the Rare Bird Alert. The RBA, as many lovingly call it, pushes notifications to users so they can quickly find out who’s seeing what and where. The excitement of rare, sought-after species fuels cultures of chasing and listing that emphasize the prestige of finding rare birds and seeing more species than others.
Chasing rarities is certainly exciting—like an ephemeral, high-stakes treasure hunt where the pot of gold has wings—but the hobby can also turn into an obsession. I’ve heard stories of constant “twitching,” or compulsive bird chasing, nearly ending serious romantic relationships. For some, a reputation for finding rare birds becomes a noteworthy part of their identity. Last fall I met a birder at a popular migrant trap, a small patch of trees with a trickle of water, on the windswept plains of eastern New Mexico. He introduced himself to me by name, followed by, “You might recognize me from the Rare Bird Alert.”
A thirst for finding rarities can also encourage behavior that goes against common courtesy. One eBird app reviewer, “Notta Realname,” noted that after spotting an unusual bird for their locality, birders rang their doorbell, asking to sit in their backyard so they could see the bird. Notta Realname reported being “flummoxed” but welcomed the birders into their backyard anyway and then became frustrated when the unexpected guests displayed “questionable” ethics. Notta Realname turned away subsequent birders and then changed their privacy settings. All birders I know would agree: showing up at a stranger’s doorstep unannounced is bad form.
Because eBird is not a social-media site—there is no way to follow friends or comment on sightings—these types of interactions filter onto other platforms. Last fall an 11-year-old friend and beginning birder ticked the wrong species of quail on her checklist, which made it look like a bird from Africa and the Arabian Peninsula had been sighted in central New Mexico. Rather than wait for eBirders to flag the mistake respectfully, someone made fun of her in a Facebook birding group.
Occasionally, eBird itself is the site of bad behavior. Recently, a respected birder misidentified a common lazuli bunting for a more unusual species: a dickcissel, or “DICK” in four-letter shorthand speak, a sparrowlike bird of open grasslands easily recognized by its “flatulent buzz” calls. Several experienced birders tried correcting his mistake, but he stubbornly refused to change his ID, insisting the bird was simply “odd-looking.”
As with any activity that can become competitive, birding has a dark side.
I had my own run-in with bad behavior on eBird last November. I’d gotten wind via the eBird Rare Bird Alert that a vagrant woodcock had been spotted along the Rio Grande near Albuquerque, New Mexico, just 15 minutes from my house. American woodcocks are iconic little solitary shorebirds that live in forests and constantly bob while they walk, and they’re rarely seen out west. Naturally, I had to chase this bird. At 7 A.M. on a Sunday, I found myself walking along the river, kicking up piles of dead leaves in an attempt to flush the woodcock.
After a few hours, I’d had no luck. As I headed back to my car, I passed a group of birders also searching for the woodcock. We chatted for a bit before a well-known birder—the one who misidentified the DICK—recognized me. With a facetious smile, he asked, “How’s your goose ID going?”
The other birders stared blankly while I brimmed with silent shock and anger. He was publicly mocking me—a week before, he’d emailed me about a misidentified Ross’s goose I posted on eBird. Embarrassed, I quickly updated my observation. Our interaction should have ended there, but instead he was now calling me out for my mistake—gleefully—in front of others.
“Fine,” I said curtly, before walking back to my car.
When I got home, I ranted to my significant other, who is used to hearing too much about birds. He thought I sounded more wound up than usual—eBird can sometimes do that to you.
In recent years, eBird has grown tremendously. Between 2019 and 2020 alone, observations submitted to the database increased by 24 percent. Some believe that the rise in new eBird users is associated with a dangerous level of data imprecision. Can the data be trustworthy if they come from millions of observers who might not be able to correctly identify common backyard birds? As one well-known California birder has been known to say, “The average birder is below average.”
This is where data-vetting steps, like eBird’s review process, come in. Each eBird reviewer is a volunteer selected for their knowledge or experience in a state, region, or country. Reviewers act as quality filters and check observations for accuracy, detail, and validity. They may contact observers to request specific details about unusual sightings, point out misidentifications, or ask for justification about higher-than-expected numbers reported for a particular species. Some reviewers even go out of their way to coach users unfamiliar with eBird on how to use the database and app to enhance the quality of the information. This verification effort, in turn, makes eBird data more valuable to birders, citizen scientists, and professional scientists.
“My goal when reviewing is to make sure that an observation is documented well enough so that, in 100 years, someone who doesn’t know who the observer is can say, ‘This is reasonable,’” says Lauren Harter, an eBird reviewer of more than nine years for the Colorado River area.
Given their status, reviewers are privy to moments of birding vulnerability, such as when birders make identification mistakes. Errors are expected—even the world’s best can confuse extremely similar-looking immature gulls or drab flycatchers. If an eBird reviewer catches an ID mistake, usually from a photo, they reach out to the eBird user, typically with a polite template email that starts with, “Thank you for being a part of eBird. To help make sure that eBird can be used for scientific research and conservation, volunteers like me follow up on unusual sightings as a part of the eBird data quality process.” They’ll then explain why the species is listed incorrectly and request that the user change the ID to the correct species.
Given their status, reviewers are privy to moments of birding vulnerability.
This mutual respect between reviewers and birders tracks with offline birding etiquette, but sometimes interactions can turn to rudeness. The birder who made fun of me for my goose mistake, for example, was a New Mexico eBird reviewer, and one birder friend, after hearing the story, called the reviewer’s comment to me “way out of line.”
The relationship between reviewers and observers can be tricky to navigate. Reviewers sometimes screen as many as several hundred sightings per month, and they certainly deal with their fair share of user mistakes. I became increasingly respectful of the work they do as I spoke to more reviewers for this story. But some believe that reviewers exercise their power unfairly—for example, by accepting rare sightings by birders with good reputations, even with scarce documentation—and impose personal rules about how birding should be done in “their” territory.
Last year a friend birded at a popular eBird hot spot outside Raleigh, North Carolina, during a work trip. After submitting his checklist, he was contacted by a local reviewer who, in typical birder fashion, sent him overly detailed instructions about how to walk around the lake. My friend, a birder of 34 years, felt like his freedom to explore had been violated. There was a right and wrong way to walk around a lake now? “I was bemused that someone would want to exert control over how others experience a place,” he told me. “The idea that a hot spot has to be birded in a certain way and recorded in a certain way really takes the enjoyment out of visiting new places.”
Despite the fact that eBird has become an almost unstoppable force, some birders have resisted the eBird tide. They see the platform—and the “Cornell mafia,” as one birder put it—as supplanting traditional methods of birding that many still prefer. Observers who don’t use eBird still rely heavily on listservs or Facebook birding groups, but this can limit access to information.
“It makes you almost have to be an eBirder to keep track of this stuff anymore,” says Gary Rosenberg, a professional bird-watching guide of more than 35 years. “I call it eBorg,” he says, referencing the Star Trek character who transforms people into drones through assimilation. “If you’re not on eBird, you’re currently just sort of left out in the cold.”
On the flip side, eBird has encouraged people who may not have birded previously to contribute sightings in a popular forum. This citizen-science participation aspect of the platform, coupled with movements like #BlackBirdersWeek, are important for creating a diverse and equitable outdoors community. Increased representation and environmental awareness are sorely needed, given the estimated 2.9 billion birds lost since 1970. “Anything we can do to supercharge an interest in nature is a worthwhile goal unto itself,” says eBird’s Iliff. “I don’t think we’re going to have people who are willing to vote for climate change or preservation of public lands or endangered species, or really care about the world around us, without a level of public engagement.”
For all its unexpected dynamics, eBird has succeeded in connecting birders and scientists in ways that weren’t possible before. Last fall, while browsing through images of the species I study for my Ph.D., the giant hummingbird, I came across a photo of a bird that appeared to be wearing one of the tracking devices I use to research their migration. The eBirder who posted the photo listed his email address publicly, so I reached out to see if he had others. He was friendly, and he happily sent more my way. I flipped through them that night, amazed that a stranger’s photos might have unintended value for my research, and I wondered what other gems remained to be discovered on eBird.
Via CXO Today, a look at how the Internet of Things may be used to support bird conservation:
India is a biodiversity hotspot and home to over 1300 species of birds. While, these beautiful creatures are vital in maintaining the overall ecological balance, their natural habitats are increasingly under threat. A recent report, State of the World’s Birds 2018, found that one in eight birds is in danger of extinction, and 40% of the world’s 10,000+ species are declining thanks to climate change, logging, industrial farming and various other factors. Hence there is a pressing need to build an impactful and scalable conservation model, so as to engage more people, organizations and communities working in this space.
Realizing the urgency to identify and preserve birds as well as spread awareness in this area, IT consultancy firm Accenture has developed an Artificial Intelligence (AI)-based Birds platform – a first of its kind in India – that identifies bird species found in the country, in association with the Bombay Natural History Society (BNHS), a non-governmental wildlife research organization.
In an engaging conversation with CXOToday, Sanjay Podder, Managing Director at Accenture Labs (Asia Pacific) explains the technology and the rationale behind this initiative.
“We realized that most people find it difficult to identify rare birds unless they are experts or have read books on ornithology. While our country is home to a spectacularly diverse set of birds, it lacked a large centralized repository to help amateur bird lovers and conservationalists accurately identify them. This is when Accenture Labs started working with BNHS to design and develop the Internet of Birds platform,” he says.
Based in Mumbai, BNHS has been working on nature conservation and research projects in the sub-continent for the last 136 years. Launched initially as a web portal, Accenture has recently launched a mobile application for BNHS, which extends the functionality of the platform to mobile devices. The Internet of Birds platform initially identified nearly 100 species of birds in India, and can now identify nearly 700.
On the technology behind the concept, Podder says, “Internet of Birds is a cloud-based, image recognition and deep learning platform which uses AI, including machine learning and computer vision. The platform has been trained on birds found in the Indian subcontinent using a convolutional neural network, which is a deep learning algorithm that can take in an image and assign importance to various aspects in the image, with the ability to differentiate between images.”
He adds that it also uses a unique citizen crowd sourcing approach under which bird watchers can contribute to the platform by uploading information on rare birds that they come across. Data from BNHS helps confirm the identity of the birds and trains the application’s deep learning model. Each time a picture is contributed to the system, it teaches itself, increasing the platform’s accuracy in the recognition of bird species.
“In addition, its AI at the edge functionality enables mobile access to the information in remote locations such as deep jungles with poor or no internet connectivity,” says Podder.
Accenture Labs in Bangalore has provided pro-bono services to BNHS to design and build the Internet of Birds platform as part of its broader corporate focus on using technology for good. The IT firm’s Tech4Good program leverages emerging digital technologies and works with social innovators, academia, startups and government to address sustainable development goals across education, health, environment, inclusion and diversity.
Asked if the Internet of Birds use case can be replicated for other conservation solutions that leverage image identification, Podder says, “The AI at the edge functionality of the app could potentially be used in scenarios where connecting to the cloud introduces a lag or is simply not possible due to lack of connectivity.”
“The model can be applied to explore locations that are either inaccessible or hazardous for human beings such as in mines, underwater, or in space, and can be used in a variety of mission-critical industry use cases that enable safety, equipment diagnostics and troubleshooting.”
Needless to say, advanced technology is enabling new ways of mining data in the wildlife preservation activities around the world. In India, the concept is relatively nascent. There’s still a long way to go in terms of new product development and research.
While more research is going on in areas, such as IoT, big data, and advanced analytics for species protection and the public good in a scalable and sustainable manner, the Internet of Birds is a powerful example of how technology such as AI and deep learning can help drive innovation that benefits our communities and our planet.
,
Read More » Posted in Open Source, Sensors | Comments Off on Internet of Birds: Harnessing AI for Bird Tracking, Conservation
An announcement of an important new initiative that will help catalyze data-driven wildlife conservation by harnessing the power of technology and science to unite millions of photos from camera trap projects around the world and reveal how wildlife is faring, in near real-time:
Wildlife Insights is combining field and sensor expertise, cutting edge technology and advanced analytics to enable people everywhere to share wildlife data and better manage wildlife populations. Anyone can upload their images to the Wildlife Insights platform so that species can be automatically identified using artificial intelligence. This will save thousands of hours, freeing up more time to analyze and apply insights to conservation.
By aggregating images from around the world, Wildlife Insights is providing access to the timely data we need to effectively monitor wildlife. We are creating a community where anyone can explore images from projects around the world and leverage data at scale to influence policy.
Wildlife Insights provides the tools and technology to connect wildlife “big data” to decision makers. This full circle solution can help advance data-driven conservation action to reach our ultimate goal: recovering global wildlife populations.
,
Read More » Posted in Cloud, Google, Open Source, Sensors | Comments Off on Wildlife Insights: Data-Driven Conservation Uniting Camera Trap Projects With Artificial Intelligence
New technical innovations such as location-tracking devices, GPS and satellite communications, remote sensors, laser-imaging technologies, light detection and ranging” (LIDAR) sensing, high-resolution satellite imagery, digital mapping, advanced statistical analytical software and even biotechnology and synthetic biology are revolutionizing conservation in two key ways: first, by revealing the state of our world in unprecedented detail; and, second, by making available more data to more people in more places. The mission of this blog is to track these technical innovations that may give conservation the chance – for the first time – to keep up with, and even get ahead of, the planet’s most intractable environmental challenges. It will also examine the unintended consequences and moral hazards that the use of these new tools may cause.Read More
