Riddles are everywhere (2015.10.05 [Mon.])

 

  I’ve been interested in the scorpionfly Panorpa japonica for some time now. This species exhibits complete metamorphosis but is grouped in the primitive lineage Mecoptera, which also includes the Bittacidae family of hangingflies. None of these are particularly well known, but they nonetheless make a nice addition to any collection (see Fig. 1). The name derives from the downward curled abdominal structure, seen only in males, but I have to say it always reminds me of Yoshiharu Tsuge’s manga collection “Hissatsu surume gatame.”

  So what is the fascination with this nondescript insect? Well, it turns out that they show the same pattern on both their fore- and hindwings. The patterns vary between species, but within the same species it is constant across both sets of wings, which indicates wing pattern diversification through evolutionary processes that somehow does not involve differential patterning in the fore- and hindwings. In nearly all winged insects, not just those that undergo complete metamorphosis, these two sets of wings have evolved into distinct structures, which suggests that these scorpionflies are not merely primitive, but belong to a lineage that is subject to some form of constraint. We may talk about ‘diversification,’ but in reality this involves a multitude of evolutionary processes about which, if pressed, we really know precious little.

  If you look at individual scorpionflies of the same species, you begin to notice subtle inter-individual differences. The fly pictured in Fig. 1, Panorpa japonica, is known for its high rate of mutations, resulting in distinct patterns (Fig. 2). In some cases, the fore and hindwings on one side of the animal will simultaneously exhibit a mutation, making it possible to observe wing differentiation phenomena that might be thought impossible from the perspective of developmental network structure in which slight mutations should manifest only in the fore or hind wings, revealing that a more complex mechanism is at work.

  I think what you can take away from such observations is that the story of insect evolution is in fact a story of the evolution of wings. Previous scholars have come to similar conclusions, noting that each insect order exhibits wing-associated traits. For example, in the differential evolution of fore- and hind wing morphologies, one can observe the emergence of hierarchical patterns. In some primitive moths, such as the ghost moth Endoclita excrescens and members of Micropterigoidea, distinct from the majority of lepidopteran taxa, the wing veins in fore and hind wings nearly always exhibit the same pattern. It looks like the time has come for us to revisit the notion that distinct fore and hind wing morphology was established by insects that undergo complete metamorphosis.

  Another interesting feature of the scorpionfly is the manner in which its larvae develop abdominal parapodia (Fig. 2). In other lepidopterans, abdominal segments that have shed their transiently appendages generate cell populations that do not express homeotic selector genes specifying abdominal fate, such as Ubx and abdA, and reestablish expression of genes such as Distal-less that are crucial for appendage specification (a process known as atavism; see Carroll et al., From DNA to Diversity, Wiley, 2004), but if we see parapodia in a species squarely in the main lineage that undergoes complete metamorphosis, it calls the conventional thesis on evolutionary origins into question. What’s more, we see the same parapdodial development in the larva of the sawfly (Hymenoptera). Are these parapodia across different lineages homologous, or did they evolve independently and converge on this phenotype?

  Speaking of parapodia, I have been wondering about a similar phenomenon in the geometer moths, whose larvae are the all too familiar inchworms, which once again shed the appendages reacquired by the lepidopteran ancestors, allowing them to wriggle down among the branches and twigs. So does this loss of abdominal appendages represent a second atavism, or a new system to control the parapodium developmental program that was innovated latterly? Intriguingly, the numbers and sites of secondarily lost parapdodia varies by lineage within the geometer moths, which seems reminiscent of the vertebral formula seen in mammals, showing orderly correlations between lineages, although I would need to study it more closely before saying anything definitive…

  In any case, an awareness of the quirkiness of morphological traits and mutant phenotypes, and how developmental programs come to be (through the influences of the epigenetic landscape, etc.) can help link our understanding of evolutionary changes to the actuality of animal form. In this regard, insects in all their diversity present an embarrassment of choices. Of course, one solution is to focus on just delving deeply into a single example or forte, which is what keeps most of us busy. My own specialty is vertebrates, and there are plenty of incongruities there as well. I can’t do them justice here, but just as there are questions that may be solved through diligent effort, or by poring over older reports for clues, there may be others on the verge of resolution in another lab, or which could be answered with just a pinch of the tweezers. Most of us in research recognize that these kinds of questions must abound. Unfortunately, we rarely encounter them in the wild, but I nonetheless hope to seduce a reader or two. The geneticist Hitoshi Kihara once published a book ‘Little Experiments’ (Chiisa na Jikken). Perhaps he did so out of the same concerns.