There comes a point in the lives of young owl monkeys when they leave their parents and strike out on their own to find a mate.

In a new study of a wild population of Azara’s owl monkeys in northern Argentina, Yale researchers reveal that a combination of social and ecological factors influences when these tree-dwelling monkeys peel away from their parents and siblings.

According to their findings—which were based on 25 years of genetic and demographic data for several generations of owl monkeys, covering more than 330 individuals—none of the individuals, regardless of their sex, reproduced in the same group where they were born. In all cases, the researchers found, the animals either departed from their natal group or died before reproducing.

The researchers also found that particular factors—such as the arrival of a stepparent, or seasonal changes in climate and the availability of food—affected the age and timing when individual owl monkeys left the group.

“Owl monkeys are pair-living, serially monogamous, and both parents directly care for their offspring, which makes them very unusual among mammals,” said the lead author Margaret K. Corley, an associate research scientist in the Department of Ecology and Evolutionary Biology in Yale’s Faculty of Arts and Sciences (FAS). “Better understanding what influences owl monkeys’ decisions to leave their natal groups helps us to evaluate the evolution and functioning of their unique and interesting social structure.”

The study, published on Aug. 7 in the journal Royal Society Open Science, was coauthored by Eduardo Fernandez-Duque, professor of anthropology in FAS. In 1996, Fernandez-Duque founded the Owl Monkey Project, a multi-disciplinary research site in the Argentinian Chaco. Research conducted at the project’s field site—one of the few places where owl monkeys are studied in the wild—provided the life history and genetic data used in the analysis.

For the new study, the researchers evaluated two prominent hypotheses on the evolution of animal dispersal. One, known as the inbreeding avoidance hypothesis, posits that dispersal is driven by the risks of individuals mating with close kin if they remain in the natal group. The other, the competition avoidance hypothesis, suggests that owl monkeys leave their natal groups to avoid competing over potential mates and resources with other members of the group.

If avoiding inbreeding drives dispersal, then owl monkeys would consistently disperse before or near sexual maturity, the researchers said. However, the study found that while some individuals dispersed around the time they reached sexual maturity, others remained with their groups up to two years after they were sexually mature. The finding demonstrates that inbreeding avoidance does not fully explain the timing or age of dispersal, but it does help to explain why all the monkeys dispersed before reproducing, the researchers explained.

The study also revealed that the replacement of a parent by a genetically unrelated adult—a stepparent—affects the age at which individuals disperse. Offspring delayed dispersal if their stepparent is of the opposite sex, the researchers found.

“Older offspring will stick around longer when their stepparent is of the opposite sex and represents a potential mate. The situation puts them in competition with their genetic parent of the same sex,” Corley said. “If mating competition with kin were driving dispersal, we would expect to see offspring disperse sooner, rather than later, when their opposite-sex parent is replaced.”

The study found that owl monkey dispersals are highly seasonal; most occur in the spring when food is abundant and temperatures are relatively warm, indicating that individuals look to begin solitary ranging when conditions are mild and there is plenty to eat. The monkeys were less likely to disperse during the winter when food is less available, suggesting that resource competition within groups does not drive the timing of dispersal, the researchers concluded.

“Owl monkeys are unusual in a lot of ways, and we have a lot left to learn about their social structure and behavior,” Corley said.

Researchers working on large artificial intelligence models like ChatGPT have vast swaths of internet text, photos and videos to train systems. But roboticists training physical machines face barriers: Robot data is expensive, and because there aren’t fleets of robots roaming the world at large, there simply isn’t enough data easily available to make them perform well in dynamic environments, such as people’s homes.

Some researchers have turned to simulations to train robots. Yet even that process, which often involves a graphic designer or engineer, is laborious and costly.

Two new studies from University of Washington researchers introduce AI systems that use either video or photos to create simulations that can train robots to function in real settings. This could significantly lower the costs of training robots to function in complex settings.

In the first study, a user quickly scans a space with a smartphone to record its geometry. The system, called RialTo, can then create a “digital twin” simulation of the space, where the user can enter how different things function (opening a drawer, for instance).

A robot can then virtually repeat motions in the simulation with slight variations to learn to do them effectively. In the second study, the team built a system called URDFormer, which takes images of real environments from the internet and quickly creates physically realistic simulation environments where robots can train.

The teams presented their studies—the first on July 16 and the second on July 19—at the Robotics Science and Systems conference in Delft, Netherlands.

“We’re trying to enable systems that cheaply go from the real world to simulation,” said Abhishek Gupta, a UW assistant professor in the Paul G. Allen School of Computer Science & Engineering and co-senior author on both papers.

“The systems can then train robots in those simulation scenes, so the robot can function more effectively in a physical space. That’s useful for safety—you can’t have poorly trained robots breaking things and hurting people—and it potentially widens access. If you can get a robot to work in your house just by scanning it with your phone, that democratizes the technology.”

While many robots are currently well suited to working in environments like assembly lines, teaching them to interact with people and in less structured environments remains a challenge.

“In a factory, for example, there’s a ton of repetition,” said lead author of the URDFormer study Zoey Chen, a UW doctoral student in the Allen School. “The tasks might be hard to do, but once you program a robot, it can keep doing the task over and over and over. Whereas homes are unique and constantly changing. There’s a diversity of objects, of tasks, of floorplans and of people moving through them. This is where AI becomes really useful to roboticists.”

The two systems approach these challenges in different ways.

RialTo—which Gupta created with a team at the Massachusetts Institute of Technology—has someone pass through an environment and take video of its geometry and moving parts. For instance, in a kitchen, they’ll open cabinets and the toaster and the fridge.

The system then uses existing AI models—and a human does some quick work through a graphic user interface to show how things move—to create a simulated version of the kitchen shown in the video. A virtual robot trains itself through trial and error in the simulated environment by repeatedly attempting tasks such as opening that toaster oven—a method called reinforcement learning.

By going through this process in the simulation, the robot improves at that task and works around disturbances or changes in the environment, such as a mug placed beside the toaster. The robot can then transfer that learning to the physical environment, where it’s nearly as accurate as a robot trained in the real kitchen.

The other system, URDFormer, is focused less on relatively high accuracy in a single kitchen; instead, it quickly and cheaply conjures hundreds of generic kitchen simulations. URDFormer scans images from the internet and pairs them with existing models of how, for instance, those kitchen drawers and cabinets will likely move.

It then predicts a simulation from the initial real-world image, allowing researchers to quickly and inexpensively train robots in a huge range of environments. The trade-off is that these simulations are significantly less accurate than those that RialTo generates.

“The two approaches can complement each other,” Gupta said. “URDFormer is really useful for pre-training on hundreds of scenarios. RialTo is particularly useful if you’ve already pre-trained a robot, and now you want to deploy it in someone’s home and have it be maybe 95% successful.”

Moving forward, the RialTo team wants to deploy its system in peoples’ homes (it’s largely been tested in a lab), and Gupta said he wants to incorporate small amounts of real-world training data with the systems to improve their success rates.

“Hopefully, just a tiny amount of real-world data can fix the failures,” Gupta said. “But we still have to figure out how best to combine data collected directly in the real world, which is expensive, with data collected in simulations, which is cheap, but slightly wrong.”

On the URDFormer paper additional co-authors include the UW’s Aaron Walsman, Marius Memmel, Alex Fang—all doctoral students in the Allen School; Karthikeya Vemuri, an undergraduate in the Allen School; Alan Wu, a masters student in the Allen School; and Kaichun Mo, a research scientist at NVIDIA. Dieter Fox, a professor in the Allen School, was a co-senior author.

On the URDFormer paper additional co-authors include MIT’s Marcel Torne, Anthony Simeonov, Tao Chen—all doctoral students; Zechu Li, a research assistant; and April Chan, an undergraduate. Pulkit Agrawal, an assistant professor at MIT, was a co-senior author. The URDFormer research was partially funded by Amazon Science Hub.

After a 4.9-meter saltwater crocodile (Crocodylus porosus) killed a 40-year-old doctor in Far North Queensland this week, the illegal feeding of wild crocodiles has become a point of major concern.

The victim was not feeding crocodiles; he was reportedly just walking along a path when the river bank gave way, and he fell into the river. His wife had tried to save him but the doctor let go of her arm, with the woman quoted as saying:

He saved me—his last act was to not pull me in with him.

The doctor was reportedly taken by the crocodile within seconds.

Since the tragic attack, which occurred at the Annan River south of Cooktown, videos have surfaced appearing to show people feeding a large crocodile in that area.

This has prompted Queensland’s Department of Environment, Science and Innovation to post a media release stating, among other things, that the penalty for illegally feeding wild crocodiles is A$6,452.

I have been researching human–crocodile conflict for years. If it’s true crocodiles in this area had been fed in the past, that is extremely concerning.

Illegal feeding linked to human-crocodile conflict

There have been concerns in the past over the illegal feeding of crocodiles in Queensland. Media outlets reported on people feeding crocodiles in the Prosperine and Russell rivers in 2022.

Outside of Australia, illegal feeding has long been associated with increased human-crocodile conflict.

At a bridge over the Tarcoles River in Costa Rica, a group of large American crocodiles (Crocodylus acutus) have been illegally fed by people for years.

Despite being less aggressive and responsible for far fewer deaths (typically only between one and three annually) than the saltwater crocodile, this feeding appears to have resulted in changes to the behavior to these crocodiles. Normally wild crocodiles avoid humans but these crocodiles, who may have come to associate humans with food, appear to have grown bolder about approaching humans.

In 2013 a man was attacked and consumed by these crocodiles shortly after entering the waterway below the bridge.

The year prior, a photographer narrowly avoided being attacked while on shore.

Illegal feeding has also been implicated in conflict involving the American alligator (Alligator mississippiensis) in the United States.

What effect does feeding wild crocodiles have on risk?

When crocodiles are fed by people they tend to lose their typically timid behavior regarding humans. They may even begin to seek people out in anticipation of being fed.

If crocodiles are consistently fed in the same location, they are likely to remain at or near the same spot awaiting the next feeding.

In the town of La Manzanilla, Mexico, for example, media reports detail how another group of large, wild American crocodiles are fed from the mangrove boardwalk on a daily basis and rarely leave the spot.

As the Department of Environment, Science and Innovation put it in their statement released this week:

Feeding of crocodiles at riverbanks or boat ramps encourages them to hang around, waiting for their next meal. This can place future visitors to the area at a much greater risk of attack if they approach or enter the water.

Even in areas with extremely high numbers of saltwater crocodiles, people frequently do irresponsible things such as wading into water. Yet no attacks have occurred (so far) in this area.

This is likely due to a number of factors, including the abundance of natural prey. However, the fact these crocodiles aren’t fed by people (as far as we know) means they’re less likely to be waiting around seeking humans out.

How can we stop illegal feeding?

Harsher punishments, such as significantly increased and consistently enforced fines or jail time, might help.

After all, illegal feeding is linked to higher risk for both human and crocodile lives (a common refrain in my field is that a “fed croc is a dead croc”).

Targeting known trouble spots and consistently prosecuting offenders could also help reduce offending.

In this age of social media influencers, irresponsible and dangerous behavior around crocodiles is sadly all too common.

Authorities could increase efforts to monitor social media sites (particularly Instagram), so they know where and who to target for investigation and, ultimately, prosecution.

This article is republished from The Conversation under a Creative Commons license. Read the original article.