Roboticists have developed robots based on a wide variety of animals over the past few decades, including dogs, snakes, birds, spiders, bats, octopuses, and various insects. These robots were mostly made to solve real-world problems like getting into tight spaces, moving on uneven ground, or flying reliably over long distances.
A small robot that looks like a tadpole was recently created by Stanford University researchers to study the parenting behavior of poison frogs (R. imitator), which are poisonous and brightly colored frogs that are native to Central and South America. The frogs’ reactions to the robot and their interactions with their offspring were the subject of some very interesting observations in their paper, which was published on the arXiv pre-print server.
Billie C. Goolsby, one of the study’s authors, told Phys.org, “Animal models offer a treasure trove of how organisms respond and adapt to ecological pressures.” We are interested in poison frogs because of their wide range of behaviors, especially when it comes to providing parental care. We can opportunistically manipulate tadpole pools, or nurseries, with robotic tadpoles to observe how parents respond to changes in offspring signals because many frogs recognize space rather than kin directly (through facial recognition). This is impossible for many other species because parents directly recognize kin or other signals, such as sibling competition, must be taken into account.
In some previous research, robotic frogs were also put in the same environment as real frogs to observe how they socialize and make decisions. On the other hand, Goolsby and his colleagues specifically wanted to find out if a robot that looked like a tadpole could influence poison frogs’ parenting styles.
The commonly observed tendency of poison frog mothers to release unfertilized eggs and feed their offspring after the latter exhibit “begging” behavior (i.e., approaching the frog’s backside while vigorously wiggling their tail) was one behavior that was particularly interesting to them. They also wanted to learn about the behavior of father frogs, who typically check out their offspring and nurseries to see which ones need more food.
Goolsby elaborated, “Maternal provisioning evolved in the specific species we study, where mothers release unfertilized egg meals to begging offspring.” Paternal calling is used to coordinate maternal provisioning, and fathers also check on nurseries to see how their children are doing. As facets of parental effort in this species, we investigate parental visitations, paternal calling, and maternal provisioning.”
“The species under study evolved maternal provisioning, in which mothers release unfertilized egg meals to begging offspring. Paternal calling coordinates maternal provisioning, and fathers also watch nurseries to evaluate kids. We investigate parental visitations, paternal calling, and maternal provisioning in this species as aspects of parental effort.”
Billie C. Goolsby, one of the researchers who carried out the study,
The researchers started by creating a small robot that could mimic the wriggly body movements of tadpoles in order to carry out their experiments. They then placed this robot inside a large aquarium tank with poison frog parents, remotely controlling its movements to imitate the begging of the tadpoles.
Tony G. Chen, another researcher involved in the study, stated, “A motor outside of our robot rotates a crank, which pulls a string that’s attached to the robot inside.” As a result, the TadBot’s tail is pulled and pushed by this and the robot’s internal spring, resulting in the wiggling motion. The TadBot is able to be as small as a real tadpole thanks to this setup, which is its main advantage.
Goolsby, Chen, and their coworkers created a robot that is similar in size, appearance, and movement to tadpoles. Notably, its motor can be mounted outside of an aquarium tank to keep the frogs from being frightened or disturbed by the noise it makes.
The robot “passed the test,” according to the researchers, and deceived the parent frogs, particularly the father frogs, who appeared to treat it like one of their offspring. In many instances, the fathers called the mothers to the nurseries so that they could feed the robots unfertilized eggs when the tadpole-like robots started “begging” and vigorously waving their tails.
It is interesting to note that, despite their visits to the tadpole nursery and answering the call, the mothers did not release the unfertilized eggs, suggesting that they were aware that the robots were not their offspring. In their paper, the researchers suggest that the mothers may rely on additional signals, such as vibrations or other subtle movements, when attempting to decide whether or not to deposit an egg.
According to Goolsby, “We hope that our work can provide a toolkit that can be tractably used to manipulate offspring signals, revealing how vibrational cues may be represented in an amphibian and where parental decision-making occurs in the brain.” We can begin to comprehend the ancestry and evolution of the neural circuitry involved in parenting by combining these findings.”
The recent study by Goolsby, Chen, and their colleagues demonstrates how robots can be safely used to study animal social behavior. Bio-mimetic robots that mimic the behavior of different animal species and use them to interact with them could be developed from it in the future.
Chen continued, “Based on the data we have been gathering, we now plan to improve TadBot mechanically to be more like a real tadpole, as well as conduct further behavioral studies.”
More information: Tony G. Chen et al, Feed Me: Robotic Infiltration of Poison Frog Families, arXiv (2023). DOI: 10.48550/arxiv.2305.14570