Lepospondyli: Overview

Introduction

The lepospondyls consist of four to six groups of mostly small aquatic tetrapods. Generally, they include Nectridia, Aïstopoda, Adelogyrinidae, Microsauria, Lysorphia, and possibly a few other types. All are limited to the Carboniferous and/or Cisuralian (Early Permian). Lepospondyls are one of a very few large vertebrate taxa whose phylogenetic position and definition both remain vague. In fact, of the 50-odd groups in which we have arbitrarily divided the chordates, only the anapsids turtle lineage) are more loosely constrained.

Even with anapsids, it is easy to give a reasonable a phylogenetic definition: for example, turtles > turtle doves. That isn't possible with lepospondyls. It's quite possible that they are not a good clade. That is, the descendants of the last common ancestor of all lepospondyls may also include the Lissamphibia living amphibians), some or all reptiliomorphs, or even the entire amniote clade. Things have settled down a bit in the last few years, and the following table lists the main recent contenders.

In addition, there are variants of each possibility, depending on how the lepospondyls and amniotes are nested within the various reptiliomorph groups. The degree of apparent high-level agreement is somewhat illusory. For example, among lepospondyl groups, the Microsauria are probably paraphyletic. That is, all the other types probably derive from microsaurs.

Nevertheless, this is a considerable improvement over the situation in the early 1990's. At that time, there was some feeling that the lepospondyls might be a series of unrelated groups scattered along the line from Tetrapoda to Amniota. That may still be the case, but most recent papers have allowed that the Lepospondyli may be paraphyletic (including Lissamphibia and possibly Amniota) but are not polyphyletic. Whether this consensus survives remains to be seen. Westlothiana see image at the Dinosauromorpha site) has been very helpful as a sort of key transition critter between the reptiliomorph-amniote line and the lepospondyls. However, if it should turn out to be something different, the whole scheme could fall apart.

The reason we engage in this pessimistic and dire speculations is that this region of phylospace is warped. Like some peculiar non-Euclidian fold in the universe of physics, this is a region in which the fundamental laws of cladistics may not hold. To see why, we must consider the peculiar difficulties raised by the recent discovery that the stem tetrapods were probably aquatic, or only minimally terrestrial. See, e.g., Clack (2002). If true (and we are decidedly undecided), no one seems to have considered the next question: if all of the stem tetrapods were aquatic, then how on earth (so to speak) did the transition to terrestrial life take place?  If everyone from Acanthostega through the colosteids, baphetids, Greererpeton, etc. was aquatic, then we have no reason to suppose that the early reptiliomorphs and temnospondyls were any different. These animals are all designed along the same general plan. But, if this plan -- which included four legs, with feet and digits, a bona fide neck, moderately strong vertebral column, sacrum, lungs, and so forth -- was not sufficient to make a good terrestrial animal, how do we explain terrestrial life before the acquisition of the full amniote suite of terrestrial adaptations? 

The answer seems to lie in a developmental anomaly acquired by derived temnospondyls and derived reptiliomorphs: two groups which seem to have, independently, adopted a terrestrial component to their lifestyle. This shared characteristic consists of a prolonged juvenile stage followed by a change to a markedly different adult form.  That is, temnospondyls and many, or perhaps all, reptiliomorphs have markedly different juvenile and adult forms. See Steyer (2000) for a discussion of the temnospondyl case. Compare, for example, the images here showing the very different construction of the related seymouriamorphs Discosauriscus (juvenile) and Seymouria subadult).

It is particularly clear in the case of seymouriamorphs that the juvenile form is essentially aquatic, while the adult morph is almost certainly terrestrial. We will adroitly leap to the conclusion that this, in fact, is the way in which the transition to land was actually made. Now, if true, it follows that evolution is going to have an unusually free hand. The juvenile reptiliomorph is specialized for growth and development in water until some critical point is reached. Then, like a teen-ager getting a driver's license, it rather abruptly begins to concentrate on sex and terrestrial mobility, often accompanied by a radical change of diet, complexion, and degree of cooperation. Needless to say, these are very different lifestyles, and probably caused no end of worry for their Viséan reptiliomorph parents, too. More to the point, this functional separation means that quite different selective pressures will come to bear on the juveniles and adults. And, the more terrestrial the adult, the more diverges from the aquatic juvenile and, hence, the greater the differential selective pressure on pre-amniotes tending to separate the juvenile and adult morphs yet further.  As with teen-agers, it's a positive feedback system limited only by the fact that the same genome and life cycle have to contain both morphs.

Take it one step further. Metamorphosis offers significant adaptive advantages for juveniles and adults, but the process itself is awkward, metabolically expensive, and risky. There are excellent actuarial reasons why automobile insurance is expensive for young drivers; and most of these factors apply to tetrapods as well as teens: inexperience, inappropriate reflexes, and a surfeit of hormones -- to mention but a few. So, while some selective pressures tend to drive the juvenile and adult forms apart, equally compelling factors offer a rich evolutionary reward for species which can eliminate this awkward and hazardous transition phase by doing away with one or the other morph -- an option sadly unavailable to insurers. Thus, the inexorable logic of metamorphosis is a trichotomy of tetrapods: (1) amphibians with two dissimilar life stages, (2) generally aquatic pedomorphic forms, and (3) strongly terrestrial peramorphic species. While the details of development are not always easy to reconstruct in organisms extinct for 300 Ma, this is roughly what we see: a primary radiation of true amphibians (temnospondyls and early reptiliomorphs), followed closely by the evolution of peramorphic terrestrial animals (amniotes and their close relatives), and a group of strongly pedomorphic, generally aquatic species -- the lepospondyls.

But here's the problem.   How can we do good cladistics on animals which may have more than one morph? This is particularly true of the pedomorphic lepospondyls, because they have jettisoned the adult form. For all we know, the adult form of some microsaur might look exactly like an early amniote. Microsaurs don't have definitive adult forms, and amniotes generally don't have an aquatic juvenile form.   In short, the usual rules of cladistics don't necessarily apply. It's hard to say how seriously metamorphosis will throw things off. However, so far as we know, no one has even looked at the problem from the vantage point of cladistics, and it certainly suggests one reason for the inconsistent results in this area. The tendency of the pedomorphic lepospondyls to cluster based on juvenile characters may very well mask a very complicated relationship between lepospondyls (especially microsaurs) and the reptiliomorphs.

ATW061003