Fascinating Science

On the other side of the scale, People often ask me why I'm studying Dragonflies - what could be so interesting about them?

As it turns out, Dragonflies have IMHO gotten a pretty weird wrap on pop-culture. Yes, they are beautiful fluttery things, but they are also one of the worlds most successful predators, arguably the most successful pursuit predator. When people think of successful predators, they often think of lions, sharks, or falcons, but Lions only have about a 30% capture success rate, and falcons only 60%. Dragonflies are 90+. But it gets more impressive.

Most animals have to deal with groups of prey when they hunt, because most prey species have figured out that they can reduce their individual chance of being killed by associating with many other members of the species. This is called the 'Dilution effect', and is part of why fish school, birds flock, and many prey mammals herd. There are a few reasons this works - the first is basic statistics. If I'm the only fish it's pretty likely to shark is going to eat me, but if I am one of a billion fish, the chance is now roughly 1-in-a-billion.

But actually, the chances are *lower* than that, because dense prey items don't just dilute the probability of being eaten, they overwhelm the predator's visual system, causing them to make more mistakes. Studies have shown that when the density of prey rises, the success rate of predators falls (even if the number of successes rises, just due to their being more). Think of it this way - with highly dense prey, the number of attempts increases and so does the number of catches, but the likelihood of a success for each attempt plummets. Fast. The most revealing experiment, for me, was one done in a fish called a stickleback, which predates on aquatic insect larvae. When presented with just a few mosquito larvae, it had a capture success up in the 90's as well, with fantastic targeting abilities. But in a swarm? The fish often accidentally made a predatory strike on the space in between two targets, suggesting it's visual system was representing the two targets as one target, at the average position. The brain was being overwhelmed.

Enter Dragonflies, who are pretty unique among animals for being immune to this 'confusion effect.' It doesn't matter if a dragonfly is hunting 1, 10, 100, or 1000 mosquitoes, it's capture:sucess rate remains stable around 90%. How could this be?

In order to understand what makes the dragonfly attentional system so robust to these conditions, i record intracellularly from individual neurons. Using a glass probe with a tip diameter smaller than the smallest visible wavelength of light, i pierce an individual neuron in the dragonfly brain to 'eavesdrop' on it's electrical signalling. I then show it visual stimuli on a computer monitor and record how the neuron responds, in order to try understand what and how it is representing.

Now to really appreciate what the dragonfly is doing, I gotta' tell you about some other animals. I'm not the only one doing this kind of science, people have recorded (using different techniques that give similar/comparable readouts from other primates, birds, and even us -- all animals that fall to the confusion effect. What these research find is that the brains of primates, birds, and even our brains represent attention choices in a 'weighted' manner. Even when we think we are paying '100%' attention to something, the neural activity is still being drawn by the distractions. In our brains, attention 'weights' the neuronal response - so say we have two targets, A, and B, and are attending A. Well, we are only giving about 80% of the response to A, and the remaining 20% is going to B. So we make errors. Things get trickier if we want to attend to something that's a bit hard to see, while there are more 'attention-grabbing' stimuli around. In order to do this, our brains use what is called 'gain enhancement' - basically boosting the signal of the target to be attended. Because the brain is 'hard-wired' to attend to the strongest stimulus at any given moment, if we want to attend to something a little weaker with distractions around, our brain has to artificially boost it so it becomes the strongest, in neural terms. However there is a big problem with this - boosting the signal also ends up boosting a lot of the noise.

The best human example I can give for this is a carnival game, but I don't know if it happens in the US or not. Here, there is a game where you can stand in a small room filled with fans and pieces of coloured paper -- and a few notes. You get 60 seconds to grab as much as you can, and any money you grab you keep. From a neuroscience perspective, your visual system has to identify and track the money so you can grab it, with hundreds of visually similar distractions about. Its hard, almost impossible to win (that's why they let you do it), but this is what dragonflies do every day when they are feeding.

So what does the Dragonfly brain do differently? Well it turns out we come across a neuron in the dragonfly brain that doesn't do any of that 'weighted' encoding, it does true 'absolute' encoding - when presented with multiple targets, the dragonfly brain selects one and responds to it as if it where literally the only one. No weighting, no boosting. The dragonfly is able to maintain a very high-fidelity representation of the target it has selected, without interference from distraction. We think this is how the Dragonfly is able to show such remarkable predatory success, so consistently under very difficult conditions.

There are still lots of questions to ask about how this attentional system works, but i guess that is how science works. A little bit at a time.

You'll probably never find discussion of this on CNN, but I think CNN is missing out. For me, understanding how a brain (a bunch of cells and bio-goop) produces behaviour, consciousness, perception, cognition - is the most interesting question in the world. Of course, since I'm doing the work, I'm very biased.

Incidentally, if I piqued your interest,. I recently published an article about Dragonflies that you can read here (https://www.sciencedirect.com/science/article/pii/S2214574520300882?dgcid=author) -- that link should bring you to a free version, but if it doesn't work, feel free to PM me or e-mail me (benjamin.lancer@adelaide.edu.au) The weird thing about scientific publications is that while I am not allowed to share anything publicly, I can send anyone who asks privately a copy.

Yep, they have six legs and they can indeed walk - but they rarely do it, and it is quite awkward when they do. The legs are more specialised for grasping, than movement. They have small 'spines' on them that allow the dragonfly to hold on to something. For example, in conditions too windy to fly they will try to land somewhere (on a tree, for example) and walk up the branch trying to find shelter. Then, the spines help the dragonfly keep hold on the branch in the wind.

The front two legs are used for walking and grasping, but they also function in prey capture. They are a bit longer than the mid- and hind-legs, and have longer spines. When catching prey, they fly towards it and extend their front legs down and across each-other so the spines form a 'basket', which they use to catch their prey.

So its true they have 6 legs, but they can walk - it's just a little awkward, and they don't really use it to get around, more just to find shelter after they land in the wind or rain.