Tuesday, July 26, 2016

Artistry of the Insect Kind

I've been posting a lot about mimicry lately, so I figured I would post about a few interesting insects I've come across that appear to take mimicry to another level. They mimic things in the way an artist would - by painting them on a canvas. Of course, the insects aren't really painting anything. Instead, they're growing the images using some complicated developmental processes.

Goniurellia tridens
Goniurellia tridens is a species of small fly which appears to have ants/spiders/somethings on its wings. The group of flies to which this species belongs have many examples with marked/patterned wings, but usually the patterns are simpler. Generally the flies use their patterned wings in courtship displays, dancing with their wings held to the side (like in this photo) while they shift and flap one wing after the other.

It is easy to imagine them using the patterns on their wings to discourage ants/spiders/etc. from attacking them, but it isn't clear if they do this behavior. It would seem to be far less risky to simply take to the air and fly away when faced with a predator. There are quite a few jumping spiders that would be able to leap and grab them out of the air. (As well, they have an impressive visual system that would let them see the images as ants/spiders/etc. in a way similar to how we do.) If such a spider were to jump at one of the high-contrast  "prey" images instead of the fly itself, the fly might gain time to escape. A jumping spider might also see the patterns as a pair of predatory ants and choose to be elsewhere. I would love to see some video footage of how this fly responds to predators as this would help clarify if any of the above hypotheticals have any basis in reality.

Siamusotima aranea
Or next mimic is the moth, Siamusotima aranea, that appears to have a bunch of spidery legs painted on its wings. This image, again, would be most useful against a predator with acute eyesight like a jumping spider. The photo at right provides some additional evidence. Moths typically rest flat to the surface, with their wings held close. In this image the moth is holding its abdomen raised from the surface and extending its wings, effectively displaying its image in a way not associated with typical moth behavior.

Other moths (like Callimorpha dominula) will actively flash their bright hind-wings to discourage predators (or photographers) who are expressing too much interest, so it is reasonable to interpret the behavior of S. aranea in this photo as being an active display to discourage predation.

Another moth, Macrocilix maia, appears to be a very good mimic of something different - bird poop and flies. The moth even has behavior to back this hypothesis up, often choosing to rest adjacent to real bird poop. Since it rests flat to a surface, like most moths, the image isn't one presented to an predator coming along the surface. Instead the image is presented continuously to the air above, where birds are using their acute color vision to hunt for insects.

It can take some experimentation to determine if an animal is really using its patterns as mimicry. We can make a hypothesis about what their wing patterns are used for and then test the hypothesis using the "natural experiment" of the moths' behavior. If their behavior is consistent with the mimicry hypothesis, then the experiment has a positive result and we can more strongly state their wing patterns are a form of mimicry. In the moth examples above, each species definitely has behavioral traits that align with the mimicry hypothesis.

In the fly example, I haven't been able to find enough information about their behavior to be able to interpret it as a natural experiment. It could be that female flies like male flies with wing-men, so to speak. (This would be consistent with the the use of wing markings in courtship displays seen generally in the group.) It could be that we humans are seeing a pattern in the image that is essentially random with respect to predation. I definitely like the anti-predation model, but without some form of experiment, we really don't have enough information to determine which hypothesis is better aligned with reality.