The future for lepidoptera larvae isn’t very bright when a tachinid fly decides to lay its eggs on the hapless larvaes’ head: this parasitic relationship ends in the larvae’s death. But first the flies have to hunt the larvae down. Yoshifumi Yamawaki wants to know how a tiny insect, with relatively poor vision, manages to solve complicated visual tasks, like host pursuit. It seems that for the tachid fly, the most important feature of it’s moving target is the leading edge, which seems to hold the strongest attraction for the fly.
The flies, which have relatively sturdy legs, pursue their prey on foot, rather than from the air. Once they’ve reached the target larva, they lay eggs on its head or thorax, to prevent the larva from dislodging them. But the larva doesn’t just sit there and take it! It often bites back, so the fly has to stay on its toes if it’s to reproduce successfully.
Surprisingly, flies don’t seem too choosy about what the host is, Yamawaki’s colleague, Chihura Tanaka, has seen the flies lay eggs on rubber tubes and slowly moving magnets; the only criterion that mattered to the flies was that they laid eggs on the front end of the moving decoy. Yamawaki took this a step further. He videoed more than half an hour of individual fly pursuits, ranging from as little as 5 seconds, to over 3 minutes to see if he could identify what attracted the fly about that wriggling larvae. But before he could extract the information that he needed, he had to design his own image analysis system to deconvolute the relative positions of the hunter and its victim.
As he analysed the films, Yamawaki saw that the flies didn’t run towards the larvae in a single rush, but they kept pausing as they approached the target. He charted the fly’s position at several points along its trajectory. Then he plotted the equivalent position of the larvae, so that he could correlate the visual stimulus from the larva with the fly’s response.
After a month of analysis, Yamawaki realised that there is a strong correlation between the larva’s position and the fly’s response. He noticed that if the larva was in the fly’s peripheral vision, the fly tended to travelled sideways (rather than forwards) and made large turns. The fly also tended to move forwards when the larva was far off. He also points out that the size that the larva appears to the fly has a much stronger affect on the fly’s movements than how central the larva is in the fly’s vision. ‘However’ says Yamawaki, ‘the interaction between these effects has not been examined… We just know the relationships between each stimulus and the responses.’
He also explains that there are several possible reasons why the fly doesn’t run at the larva in a single rush. One is to help the fly distinguish the slowly moving larva from the changing background as it runs towards the larva. From the few sensory cells that make up the fly’s retina, it is probably very difficult to resolve the larva from an apparently moving background. It probably also gives the fly more time to plan its approach, because it’s in little danger of being outrun by a wriggling larva. Yamawaki also explains that the fly’s ‘stop and go’ style of pursuit might arise from the flies sense of self preservation as well as it’s poor eyesight. The larvae aren’t passive victims of the fly’s life cycle; pauses in the fly’s approach may give it more opportunities to take flight if the need is urgent.
