American Scientist

Amniote Origins: Completing the Transition to Land. Stuart S. Sumida and Karen L. M. Martin, eds. 510 pp. Academic Press, 1997. $74.95.

Everett C. Olson was one of the most innovative and eclectic paleontologists of this century. His multifaceted approach has been an inspiration to generations of paleontologists, and it is therefore fitting that this new volume is dedicated to his memory. True to the Olsonian method, this book goes beyond skeletal anatomy and phylogeny to explore many other facets of the transition from amphibious to truly terrestrial existence.

Laurin and Reisz begin this volume with a detailed phylogenetic analysis of basal tetrapods. The large data matrix (38 taxa and 157 characters) is difficult to digest. Although their conclusions do not differ a great deal from other recent analyses, their placement of modern amphibians within the batrachomorph clade very near basal amniotes contrasts with the almost universally held opinion that modern amphibians are only very distantly related to amniotes. No doubt this conclusion will provoke a great deal of heated debate.

Since amniotes must have appeared by the Middle Carboniferous period, information on Upper Carboniferous and Lower Permian terrestrial faunas is essential for an understanding of their origin and early diversification. For many years, nearly all of this information came from North American deposits. Recent discoveries in Europe, reviewed by Berman, Sumida and Lombard, place these events in a more detailed biogeographical and paleoecological context, permitting a reexamination of earlier assumptions about the ecological factors associated with the origin and diversification of amniotes.

The critical event in the origin of amniotes, by definition, is the appearance of the cleidoic or amniotic egg, a complex structure considered essential for effective terrestrial reproduction. Unfortunately, eggs rarely fossilize, and the absence of data has hampered our ability to adequately define fossil amniotes. Lee and Spencer suggest a way out of this dilemma: Abandon the traditional apomorphy-based definition of amniotes (the first tetrapod with an amniotic egg and all of its descendants, extant and extinct) in favor of a crownclade definition, which includes only extant members and the first tetrapod to possess the amniotic egg. This permits a more precise definition of the amniote, but at the cost of excluding some probable basal amniotes.

Two papers examine amniote origins indirectly by reviewing the anatomy and physiology of the modern amniotic egg. Packard and Seymour argue that the essential event permitting terrestrial oviposition was the elimination of the thick jelly capsule of the amphibian egg and development of a fibrous shell membrane. This simultaneously improved mechanical support and increased the rate of gas exchange, permitting a larger, metabolically more active embryo to reach full development before hatching. Extraembryonic membranes and a calcified shell, though improvements on the design, were not essential and could have evolved later. If the development of a terrestrial egg was this easy, could it have happened more than once? Probably not, according to Stewart. Basic similarities in the development of the embryo and its membranes in a variety of amniotes support the homology of the amniotic egg throughout.

The adaptive advantage of the amniotic egg is indisputable, but this in itself does not explain why the egg evolved when it did. Graham, Aguilar, Dudley and Gans suggest that a gradual increase in the concentration of atmospheric oxygen throughout most of the Paleozoic may have created an opportunity for amphibious tetrapods with otherwise marginal abilities of aerial respiration to convert to the terrestrial environment. Once this occurred, there would be strong selective pressure to develop a terrestrial reproductive strategy. As oxygen levels fell through the Permian, these primitive amniotes, now irrevocably committed to aerial respiration, were forced to evolve more efficient respiratory and circulatory mechanisms.

Terrestrial adaptations of the vertebral column and appendicular skeleton thought to be correlated with the evolution of the amniotic egg are reviewed by Sumida, who points out that these adaptations are not restricted to amniotes, but characterize a much larger radiation of primitive tetrapods. The loss of dermal scales and replacement with keratinized epidermal scales, passed over quickly in most accounts, is usually assumed to be related in some way to the development of an impermeable skin covering. Frolich points out that keratin, by itself, cannot prevent desiccation. Instead, he argues that the transition to lighter, keratinized scales was driven by the need to reduce bony-scale mass around the periphery of the body. The integument of the first amniotes, as in living lower tetrapods, probably exhibited a wide range of resistance to the passage of both water and gases.

Martin and Nagy suggest that a key adaptation of the first amniotes, reduction of cutaneous water loss through reduction in skin permeability, also reduced cutaneous water uptake, respiration, utility of ammonia as a waste product and temperature regulation through evaporative cooling, all of which play important roles in the physiology of living amphibians. Increased body temperature, metabolic rate, activity, and growth rates associated with the need to improve metabolic homeostasis were all important factors in the success of amniotes.

Lauder and Gillis review experimental data on feeding mechanisms of fish and amphibians in an attempt to reconstruct the feeding mechanism of the immediate amniote ancestor. Hotton, Olson and Beerbower point out that among living amniotes, only those specializing in high-fiber herbivory exhibit clear skeletal evidence of diet. Omnivory and low-fiber herbivory, being primarily behavioral rather than structural adaptations, were probably much more common practices among the first amniotes and possibly their immediate predecessors than generally thought.

Garland, Martin and Diaz-Uriarte demonstrate how the method of squared-change parsimony can infer characteristics of hypothetical ancestral organisms (in this case the first amniote) using data from extant taxa. A major strength of this method, as compared to standard phylogenetic analyses, is that it can process continuously variable characters. It is clear that if generally accepted methods enabling the combination of continuous variable data with qualitative morphological characters into a single algorithm for generating hypotheses of relationships can be developed, systematists will possess a much more powerful tool with which to address phylogenetic problems.

Whenever we try to extract so much information from fossil bones and reason by analogy from living surrogates, the results are, by necessity, somewhat speculative. However, in contrast with many past efforts, the contributors have made a concerted attempt to generate testable hypotheses, and the editors have done an admirable job of tying this diverse range of topics into a coherent work. This volume will make a significant contribution to the construction of a research program for the future.—Robert Holmes, Paleobiology, Canadian Museum of Nature, Ottawa