PerspectiveEvolution

Wingless Insects and Plucked Chickens

See allHide authors and affiliations

Science  21 Mar 2003:
Vol. 299, Issue 5614, pp. 1854-1855
DOI: 10.1126/science.1083465

Plato, strolling through the grove of his Academy and holding forth to his students, is said to have defined man as a featherless biped, whereupon his rival, that wag Diogenes, tossed a plucked chicken over the wall to land at his feet. Those of us interested in arthropod relationships can understand how Plato felt.

Arthropods rule the world, at least among multicellular animals, and we'd like to think that we understand the basic outlines of their evolutionary relationships. Unfortunately, we don't. The last time experts on all branches of arthropods came together to discuss their phylogeny, back in 1996 (1), consensus was conspicuously absent. Since then, it is safe to say, virtually everyone accepts that the arthropods are a monophyletic group—that is, their last common ancestor was something we would recognize as an arthropod. Beyond that rather modest point, argument still rages about many of the evolutionary relationships within and between the four major extant arthropod groups: the chelicerates (spiders, mites, and scorpions), myriapods (centipedes, millipedes, and other, more obscure groups), crustaceans (lobsters, crabs, and barnacles), and hexapods (insects and the six-legged wingless groups, the collembolans and proturans, considered basal to the insects). One of the few points of general agreement has been that the hexapods are monophyletic, that is, they arose just once from a single common six-legged ancestor. Now, into this small grove of consensus, wander Nardi and colleagues (2) on page 1887 of this issue, who essentially lob a naked fowl into the midst of this accord. Their mitochondrial genome sequence data strongly support the separation of collembolans from the rest of the hexapods by several intervening species of crustacean (see the figure). This in turn suggests that today's terrestrial hexapods are products of at least two independent invasions of land and that some of the features shared by all hexapods have arisen convergently.

Six-legged relationships.

Mitochondrial genome sequence data suggest that six-legged terrestrial arthropods evolved more than once (1). Extant terrestrial hexapods (insects and the wingless collembolans and proturans) seem to be products of at least two independent land invasions and, thus, certain shared features of all hexapods have arisen convergently.

CREDIT: PRESTON MORRIGHAN/SCIENCE

Collembolans are a very common group of primarily soil-dwelling arthropods. They are often called springtails because of the unique appendage on their abdomen that allows them to jump many times their body length, making them the envy, proportionally, of any Olympic pole vaulter. As a group, the Collembola have a very long history—they were wandering around Scotland in the Lower Devonian, some 395 million years ago (3)—and now it appears they kept company with a marine hexapod in Germany at the same time (4).

Nardi et al. have sequenced the mitochondrial genomes of two collembolan species and a thysanuran (or silverfish). The silverfish are members of the Zygentoma, which are widely held to be the closest relatives of the winged insects. The authors took each of the 13 protein-coding genes typically found in animal mitochondrial genomes and aligned them with those of other arthropods whose mitochondrial genomes have been completely sequenced. To root their phylogenetic trees, they also included mitochondrial genome sequence data for several out-groups, making a total of 35 taxa subjected to phylogenetic analysis.

Their initial phylogenetic analysis is interesting because it reveals the effects of systematic biases in sequence data on the recovery of a believable tree of relationships. The most spectacularly unbelievable result of this intentionally naïve analysis is a strongly supported grouping of two insects (honeybee and louse) with two ticks (chelicerates). Inspection of the data reveals at least part of the reason for this obvious anomaly—a convergent high A + T base composition in the mitochondrial genome sequences. Another potential source of systematic bias is the lineage-specific differences in rates of evolution—rapidly evolving lines may group artifactually, the so-called “long branches attract” phenomenon (5).

Nardi and colleagues mitigate the effects of these biases by winnowing their data to remove taxa that are not compatible in base composition and relative rates of evolution with the taxa (collembola and silverfish) that they deem crucial for testing the hypothesis of hexapod monophyly. They lose more than half their taxa in this exercise, leaving 15, but they remove obvious bias from the remaining data. Now they find strong evidence linking hexapods with crustaceans. This in itself is another point of major contention among those studying arthropod phylogenetics. For many years, Myriapods have been considered the closest relatives of hexapods. But, more recently, new data from molecular phylogenetics and developmental biology support a close relationship between hexapods and crustaceans. This grouping has had its proponents in the past: A hundred years ago a former director of my institution—then known as the British Museum (Natural History)—placed the hexapods and crustaceans together (6). The findings of Nardi et al. and other recent work force a careful look at data from all sources (7).

The final analyses of Nardi and co-workers appear to be very conservative and strongly support the separation of the collembolans from the insects by the two remaining species of crustaceans (see the figure). Many arthropod experts will not be entirely convinced by these data. Systematics is a very contentious field, so we can count on criticisms about the small number of species, the single data type, and the method of analysis. But at the very least, these data will spur both the collection of more sequence data from more taxa and also the extension and reevaluation of morphological work. Whatever the outcome, we will have a more solid understanding of how six-legged animals colonized and then took over the terrestrial world.

References

View Abstract

Stay Connected to Science

Navigate This Article