Last November I posted pictures of some bizarre treehoppers (membracids, a family in the order of true bugs, Hemiptera) that had strange structures branching from the top of their bodies. Because these insects were so incredibly weird, it proved to be the most linked-to post I’ve ever written. At any rate, go back for a minute and look at some of those things. And here are some more, taken from a new paper in Nature by Benjamin Prud’homme et al. (click to enlarge):
We’ve seen the one at the upper left before: it’s Bocydium globulare. The function of these “helmets” isn’t always clear: some of them may deter predators (e.g. left images, rows 2 and 3), others may serve as camouflage (last row, left and middle), and still others may mimic distasteful or dangerous insects like ants (lower right). Regardless of their function, the Prud’homme paper has made a significant contribution to understanding where these “helmets” come from.
They are highly modified homologs of the insect’s wings: an ancient winglike structure that was long repressed but appeared again to take on a variety of new functions as a helmet.
The “helmet” appears on the first segment of the insect’s thorax (“T1”), and is the only known dorsal appendage on that segment in insects (wings of modern insects are always on the last two segments, T2 and T3). Prud’homme et al. suggest, with good evidence, that the helmet is a pair of fused wing primordia: remnants of the early structures that gave rise to modern wings.
How do we know these are wing homologs? Prud’homme gives several lines of evidence:
- The helmet is not simply an outgrowth of the thorax, but is connected to it, as are the wings and other appendages, by a “complex articulation”: a flexible joint. Here’s a cross section of a dissected treehopper: the joint between thorax and helmet (top, red box) is similar to that between the wings and thorax (bottom, blue box):
The development of the helmet is similar to that of the wing: both “unfold” on emergence (many times I’ve watched the wings of newly eclosed flies unfurl from nubbins to full wings, a remarkable process that takes only a minute or two).
- There are other morphological similarities between wings and helmets: both consist of two layers of cells connected by columns, and both are suffused with a complex network of veins.
- The genes that are expressed in the developing helmet are the same as crucial genes expressed in the developing wing, including Nubbin, Distal-less, and homothorax.
If the helmet is a re-expression of a repressed wing structure, does that mean that early insects had more than two wings? Well, even some modern insects have more than two wings: dragonflies and bees, for example. But these wings are always, as I mentioned, on the second and third thoracic segment. We know that insect flight evolved about 350 million years ago (insects are the only flying invertebrates), but fossils are scanty. One theory, espoused by Jarmilla Kukalova-Peck and supported by more recent genetic work, suggests that wings evolved from gill plates in early insects (I’ve written before about the likelihood that the ancestors of modern insects were closely associated with water). In one of her papers, Kukalova-Peck gives a reconstruction of a Paleozoic mayfly nymph (an aquatic life stage), showing winglike structures—gill plates—on every segment of the thorax and abdomen:

From Kukalova-Peck, J., J. Morphology 156:53-125 (1978)
Another theory is that wings arose from branches of the ancestral insect limb. Both theories, though, posit that there were more than one or two ancestral structures that were winglike, and that the evolution of the two or four wings in modern insects involved genetic repression of these ancestral features. The gene Scr (sex combs reduced) seems to be involved in this repression: when it’s inactivated in some insects, extra wing primordia form on T1.
The authors thus posited that somehow, in the last 100 million years (the time when membracids arose), the Scr gene lost its ability to repress wing promordia in the membracid lineage, allowing the helmet to evolve. To test this, they actually inserted the membracid Scr gene into Drosophila (which has a normal, “repressive” Scr), expecting that perhaps wing primordia would then arise on the first segment of the fly thorax. They didn’t. They thus suggest that other genes—genes normally repressed by Scr in membracids—have lost their ability to be repressed, and these genes are involved in making the helmet.
Here are the authors’ conclusions from the paper. Even if you’re not a biologist you should be able to understand them, for if you haven’t, I have not written clearly enough!
Our results show that treehoppers have evolved a T1 dorsal appendage, thereby departing from the typical winged-insect body plan, by expressing a developmental potential that had beenmaintained under the repression of a Hox gene for 250Myr. This argues that the constraint preventing extra dorsal appendage formation in insects is not developmental but rather selective. We submit that morphological innovations can arise from the deployment of existing but silenced developmental potentials, therefore requiring not so much the evolution of new genetic material but instead the expression of these potentials.
The breadth of morphological diversity in helmets that has evolved in less than 40 Myr (ref. 27 and C. Dietrich, personal communication) is unusual for an appendage. The pace of appendage evolution is generally slow, probably because of the strong selective pressure associated with their role in locomotion. This is particularly true for the wings, and we speculate that, initially alleviated from functional requirements, the recently evolved helmet was free to explore the morphological space through changes in its developmental program.
As your reward for reading this far, here’s another really weird membracid (Heteronotus sp., from Ecuador), photographed by Alex Wild and taken from his wonderful website, Myrmecos. You can see both the helmet and the wings.

__________________
Prud’homme, B. et al. 2011. Body plan innovations in treehoppers through the evolution of an extra wing-like appendage. Nature 473: 83-86.