Wikimedia Commons

A new Yale study has discovered how the walking fruit fly interacts with complex smells. 

The research paper was published on Nov. 3 and was written by Yale’s Emonet Lab, which studies the biological functions that allow organisms to navigate their chemical environments. In the experiment, smoke was used to track the flies’ responses to complex odors in an effort to better understand how insects maneuver themselves in the natural world.

“We didn’t really understand yet how insects interact with odors,” said Thierry Emonet, professor of molecular, cellular and developmental biology and the head of the Emonet Lab. “The motivation was to try to discover the strategy insects use to navigate this kind of signal.”

According to Emonet, previous studies have investigated how insects interact with simple odors that are continuous and have a smooth gradient. In these cases, Emonet said, the flies move towards the more concentrated odor by following the concentration gradient. 

Most odors in nature, however, are not smooth gradients. Instead, wind and motion protrude the flow so the odor comes in individual odor packets that are disconnected from one another. It had been previously unknown as to how insects use these complex smells to guide themselves to their sources. 

“Because of the shifting winds, the odor direction is not constant,” Mahmut Demir, an associate research scientist in the lab and one of the lead authors on the study, said. “It doesn’t give the insect a direct route. Because of the turbulence, the odor is broken into little packets so there is no gradient the insect can follow.”

According to Demir, the goal of the study was to see how flies guide themselves to the odor signal when the odor is broken into these packets. These interactions were impossible to study because of the lack of visualization until Demir discovered an odor that was both attractive to the flies and easily trackable: smoke. 

According to Nirag Kadakia, a postdoctoral fellow in the Emonet Lab and the other lead author on the study, the flies were put in a dark two-dimensional walking column, and smoke was added by burning a wick. The smoke was then visualized using infrared light and the flies’ walking movements were analyzed, said Kadakia. 

“What we discovered is that, when the odors become more complicated, those flies, instead of deterministically turning whenever they encounter the odor, adopt a more probabilistic strategy to navigate,” Emonet said. 

Demir explained that, with simple smells, insects respond in a deterministic — or predictable — way, where they turn and walk in an upwind direction when encountering the odor signal. In complex environments, such as the environment set up in the study, the insects transition to stochastic — or randomly determined — behavior. Instead of turning upwind, the flies assign a probability of interacting with each odor packet, and use these probabilities to determine their actions, Demir said. 

In this behavior, the flies randomly turn at a constant rate, whether or not there is an odor. When the frequency of the odor packets is high, however, there is a higher probability that the fly will turn in the up wind direction, Emonet said.

“One of the most important findings in this paper,” Kadakia said, “is that the animals use the frequency of those hits to bias their behavioral actions in particular.” 

According to Kadakia, these findings could have future implications in terms of technology. The study provides an understanding for how animals use chemical signals to maneuver complex environments. He said this could then be used to design robots that track chemical signals in order to find a chemical source in environments that are unsafe for humans.

According to Demir, the study also has potential applications in information processing and understanding how neural networks extract relevant information from complicated situations. 

The next step in the research, according to Kadakia, is to transition the experiment — which only tracked the walking movement of the insects — to flight. 

“It would be interesting to see if a similar type of behavior was followed when they fly,”  Kadakia said, “because flies can fly much faster than they can walk.”

According to Kadakia, the flies choose when they stop very selectively based on the odor signal. They cannot stop when they are flying, however, because it is hard to stay put, he said. Emonet mentioned that Demir is planning to re-do the experiment in three dimensions to track the insects’ flying trajectories. 

Demir believes another future step in the research is to figure out what pathways in the flies’ brains account for the different behaviors exhibited when interacting with simple and complex odors. 

“We know that animals have different behaviors and different conditions,” Demir said, “but we don’t know the neural networks in the brain that causes this transition — which neurons are responsible for which behavior. The next step is to figure out which neurons are responsible for which behavior.”

Insects produce odor signals to interact with one another and are able to detect smells using antennae or other sense organs.

Kaitlin Flores | kaitlin.flores@yale.edu