Flying senses: How insects find objects in a big, big world

In 1974, 39-year-old Ronald Prokopy was living in a small trailer with his wife and new-born son. Barely a year earlier, the couple had moved to Bailey's Harbor in Wisconsin, a tiny town with a population of about 600. The move was precipitated by two of Prokopy’s chief passions. His newly established orchard, ambitiously named ‘Prokopy Bio-Experimental Farm’ was one. His lifelong obsession with the apple maggot fly was the other.

Apple maggot flies, or appleflies, are noted pests in apple orchards. Apples form the core of their existence – they mate on its surface, lay eggs beneath its skin, and newly-hatched larvae feed on its juicy flesh.

One of the most respected entomologists of his generation, Prokopy spent over 30 years studying appleflies in their natural habitat. Some of his most ingenious experiments involved devising fly-traps out of cardboard and sticky glue, which he then hung on apple trees. Surprisingly, the little flies proved really adept at distinguishing shapes and colours, consistently preferring sticky red spheres to sticky yellow rectangles. This simple experiment hinted at something big.
Our survival depends on our ability to identify objects, including food, mates and predators. In a chaotic world, mistaking a predator for a harmless piece of scenery may spell death, and for a fly with a brain the size of a pinhead, identifying objects becomes a remarkable feat. Yet, as Prokopy observed, it is a task the flies excel in. And at the time, no one knew how they achieved this.

Four decades later, a group of scientists are trying to answer the same question, in a place halfway across the world from Bailey’s Harbor – at the National Centre for Biological Sciences (NCBS) in Bangalore, India. Driven by a desire to understand how insects recognize objects, these scientists travel from woodlands dominated by wild elephants to the high mountain passes of the Himalayas. They have at their disposal, tools that Prokopy lacked – a 3D printer, an insect brain imager, and a home-grown virtual reality machine.

Their experiments took them to Sikkim’s Lachen Valley, a popular tourist destination that lies more than 2500 meters above sea level. Yet, the group of researchers who gathered there had little time for sightseeing. For nearly three weeks, they took turns watching a patch of flowers, patiently awaiting an elusive visitor. The unusual conditions of the Himalayan heights took their toll on some of the scientists.
"The UV is extremely, extremely intense, which is horrible for me with my Irish skin. I just turn beet-red immediately", says Shannon Olsson, team leader.

Their quarry was a little fly called the hoverfly. Although Olsson had previously encountered the hoverfly in Uppsala, Sweden, where her close friend and collaborator, Karin Nordstrom, was working on its neurobiology, her fascination truly began when a student captured one 4800 meters up in the mountains of Sikkim. “And that just blew my mind” says Olsson, “That something could actually exist at such different climates and manage to make its way in the world.”

In Sikkim, Olsson and her students were trying to understand how the hoverfly identifies the flowers that it chooses to feed on. The team first collected data on the hoverfly’s floral preferences, and passed the information through what Olsson calls a “statistical soup”, a form of data analysis. This helped them identify the colours and odours that the hoverfly seems to prefers most – a blueprint to design the ideal flower for a hoverfly, if you so prefer.

The team then painstakingly brought this design to life, using of bits of paper and plastic. These ‘superflowers’ may resemble nothing in nature, but were expected to set the hoverfly’s heart fluttering. The  group went to Sikkim to test their hypothesis. In Lachen valley, the researchers placed the superflowers unobtrusively within a patch of real flowers that the hoverflies were known to frequent. And they waited.

Finally, their efforts were rewarded. "When we saw the first hoverfly land on our flower, we were just screaming, we were so happy," says Olsson.

The scientists’ results have shown that hoverflies can identify flowers which have specific visual and odour patterns. As predicted by their data analyses, the hoverflies are attracted to some patterns and avoid patterns expected to be repulsive to them. From paper and plastic, the team is now moving on to flowers that are 3D printed in the laboratory according to custom-made designs, which they plan to test soon in the field.

Olsson’s group constitutes the Naturalist-Inspired Chemical Ecology (NICE) lab at the NCBS. Taking inspiration from naturalists like Ronald Prokopy and E.O. Wilson, she and her students look to nature for answers on how insects make sense of their world. One needs to ‘think like a fly’, says Olsson, a motto she picked up from Thomas Eisner, her PhD advisor. “And that requires you to understand how that organism exists in its natural environment, whatever that might be – in the water, on the ground or underground,” she says.
Olsson’s research may turn out to have practical applications. India is the sixth biggest coffee producer in the world, exporting nearly 300,000 metric tons of it each year. Yet, coffee yields suffer losses of up to 40% annually, all thanks to a single organism – the coffee white stem borer.

The coffee white stem borer is a beetle that lays its eggs almost exclusively in the stems of coffee trees. As the larvae grow, they eat the tree alive from the inside. The only way to be rid of the pest is to cut down the tree and burn it before the larvae can pupate and more adult insects can emerge to lay eggs on other trees. Since coffee trees can live for over 50 years, this is a drastic measure that coffee farmers grieve to go through.

On being approached by one such desperate coffee plantation owner, Olsson’s NICE team took a hike down to Coorg, deep in the heart of India’s coffee county. The coffee plantations here stretch for miles, and the little field station that the team put up attracted the attention of wild elephants, who took exception to some of their equipment.

In spite of such minor setbacks, the team began observations, led by junior research fellows Sriraksha Bhagavan and Santosh Rajus. In a move that would probably have sent the coffee growers’ hearts pounding, the scientists purposefully released stem borer beetles into the air, albeit within the safety nets of the enclosed field station. The researchers were driven by their fundamental question, that of object recognition. “How does this coffee white stem borer know what coffee is?” says Olsson. To answer this question, the team started testing different scenarios with the released beetles.

Bhagvan and Rajus have found that even though the beetle larvae make their home within coffee stems, the adult beetles rarely landed on stems. Instead, the adult beetles preferred to land on leaves instead. A defoliated plant, it turned out, attracted very few beetles. Equally important was the strong aroma of the coffee plant, as the team found out when they covered plants with clear plastic, which preserved visual information, but masked the smells. The team have now carried these ‘smells’ back to the lab, adsorbed on pieces of charcoal, where they are currently testing how the beetle’s antennae respond to the different chemicals that make up the plant’s odour.

However, not all of the NICE lab’s studies begin in the field. Inside a small dark room within the lab, next to a table littered with a tangle of wires and electronic parts and equipment, Pavan Kaushik, a PhD student in Olsson’s lab, has built a virtual reality world that seems to be straight out of science fiction. Three computer monitors are positioned to completely enclose a triangular area. Inside this enclosure is a virtual world displayed by the monitors – a strange land with vast expanses of green grass, blue skies, and giant orbs coloured red or green strewn on the grass. The virtual world moves around, and one of the green orbs zooms into view.

A closer look tells you that human hands are not controlling this motion. A tiny applefly, tethered to a needle mounted inside the enclosure, is navigating this virtual world, flying as it were through an alien landscape. The movement of the virtual world is synched to its wing beats, allowing the fly to use its wings like a joystick to navigate its way through the virtual world. Using a small nozzle located near the fly’s antenna, Kaushik can control the odours the fly smells as it flies.

In science, a black box is a system whose internal workings are hidden from us, but we know its input and outputs, and this is how Kaushik refers to this problem. “I have a black box, and I'm trying to figure out what this black box does,” he says. The inputs to his black box are odour and vision. The output is the fly’s behaviour: we know the flies go towards apple trees. But how does the fly take than input and turn it into behaviour, the output?

Kaushik  is essentially trying to answer the same questions that fascinated Ronald Prokopy 40 years ago – how do appleflies manage to find apples in a world filled with not-apples? With this setup, Kaushik is hoping to understand how an applefly builds up the image of an apple in its head with visual and odour cues, which it then uses to make decisions.

Ronald Prokopy passed away in 2004, working with insects till the last day of his life. Perhaps if he were to see the virtual world his appleflies now inhabit in the NICE lab, he would have been happy to see scientists building up on the work he was so passionate about. Or else, having been notorious for avoiding technology, he might have urged these researchers to return to fields and orchards outside to become a new generation of starry-eyed naturalists. 

This article was written as part of an exercise on Feature Writing conducted at the Sixth Annual Science Journalism Workshop, and has been edited by Anil Ananthaswamy.