About 66 million years ago, when dinosaurs were at the peak of their diversity, a major cataclysm precipitated by a huge asteroid wiped them from the face of the planet -- but entirely spared the mammals, who by default inherited the earth's ecosystems. This is a popular explanation, having been publicized by scientific reviews, news articles, and television documentaries. It is also, in several key respects, most probably wrong. And the problem reflects more than the fact that dinosaurs are not technically extinct -- that as saurischian descendants, birds are classified biologically as a type of dinosaur, vernacular usage notwithstanding.

By the mid - 1990s, a more complex picture of the end of the Mesozoic had emerged. J. David Archibald's Dinosaur Extinction and the End of an Era offers a broad survey of paleontological and geological evidence in an attempt to determine just what happened at the Cretaceous -- Tertiary ("K-T") boundary. The results are engrossing, although any reader seeking easy answers and simple causes may be dissatisfied. All the better, because Archibald's book challenges the reader to grapple with the kinds of ambiguities and complexities that are inescapable in paleontological study.

Professor David Archibald, a member of the biology faculty at San Diego State University, brings to the issues an expertise in North American mammals of the late Cretaceous, the small beasts who lived, in Stephen Jay Gould's phrase, "in the interstices of the dinosaur's world." Patterns of extinction and survival across the K-T boundary in these and other species offer a powerful way to test hypotheses about the causes and duration of the late Mesozoic extinctions. The focus of Archibald's book is the Hell Creek Formation of eastern Montana, described as the only site worldwide that has yielded detailed evidence about the fate of individual species across the K-T boundary. Study of this site is important, because if some (non-avian) dinosaurs persisted into the early Cenozoic, as has been claimed from work in China, or if North American dinosaurs were experiencing a long-term decline in the late Mesozoic, then the extinctions may have occurred over a protracted time period under the influence of a variety of causal factors.

Archibald introduces the reader to the many elements that must be considered in paleontological interpretation. Definitive determination of when a taxon has gone extinct is not as easy as one might think, for disappearance of a species in a given site may reflect only a local extirpation, or some combination of diminution and sampling error, or even anagenetic transformation of the species into another. Furthermore, fossils can be "reworked" into overlying sediments under certain conditions. Additional problems may attend analyses by David Raup and others that suggest that mass extinctions have occurred periodically every 26 million years. An example cited by Archibald is that of the taxonomic artifact, for when non-monophyletic taxa or families are analyzed instead of species, spurious patterns can emerge. In view of such factors, the book's subtitle and recurrent refrain that we must listen to "what the fossils say" is curious; if anything Archibald's account demonstrates that sophisticated analysis of observations is crucial, and that more than a single interpretation is often possible.

Most of Dinosaur Extinction s devoted to evaluation of the evidence for various possible causes of the extinctions, including the effects of the giant asteroid that is thought to have struck the earth at the close of the Cretaceous. Over 51% of the 107 species at Hell Creek went extinct across the K-T boundary. About 75% of these extinctions occurred among the marsupials, lizards, sharks, and dinosaurs; less affected were such groups as turtles, ray-finned fishes, crocodilians, lissamphibians, and placental mammals. That the extinctions were neither near-total nor randomly distributed among the taxa offers a useful data base for testing predictions generated from the hypothetical causes. For example, Archibald reasons that massive world-wide conflagration following asteroid impact ought not have spared some lineages over others, whereas acid rain (indirectly resulting from the impact) would mainly have affected aquatic species. Alternatively, extensive marine regression should have especially affected marsupials (by allowing eutherian competitors to invade from Asia over the Bering land bridge) and the dinosaurs (which are mainly found in the coastal sites represented at Hell Creek).

From his analysis, Archibald concludes that no single factor can account for the patterns of extinction and survival of vertebrates at the Hell Creek site. Rather, Dinosaur Extinction opts for a "cacophony of causes" that occurred over a period of several million years, including massive volcanism and marine regression. Asteroid impact may have simply delivered a final, massive blow to groups of organisms that were already diminished before the end of the Cretaceous.

Archibald's analysis undoubtedly will stimulate debate, for his conclusions are sensitive to assumptions about the differential effects of the possible causes of extinction. For example, the facile assumption that massive global cooling would have affected ectotherms far more than endotherms is debatable. Given the high food requirements of endotherms, which expend 80-90% of their assimilated energy on thermoregulation, the associated effects on ecosystems could have had devastating impacts on endotherms. Likewise, to presume that extensive acid rain of pH below 1.0 would have affected amphibians and fishes more than terrestrial forms overlooks the likely effects of the burning rain on terrestrial ecosystems, including plants of the food chains.

A more serious concern has to do with the broad-brush approach taken in the taxonomic comparisons. The taxa are categorized into two groups according to their level of extinction; thus lizards, with only a 30% survival rate are classified as showing significant extinction, and multituberculates with a 50% survival rate are not. These two categories of taxa are then used to test the hypothetical causes, with each taxon being scored as a simple "yes" or "no"; thus, intermediate effects are not considered. Yet persistence of only one additional species would have placed lizards into the "survivor" category -- entirely reversing all the predicted outcomes for that group. Similarly, with loss of only two additional species, multituberculates would represent a "significant extinction" -- likewise reversing the associated predictions. The resulting conclusions therefore represent parsimonious interpretations, but are not necessarily robust. Listening to "what the fossils say" also requires assessing the durability of their message.

As another example, species are divided into two undefined categories by body size ("large" and "small") to allow examination of the effects of habitat fragmentation following marine regression. The rationale is that larger organisms are more subject to such fragmentation. But comparison of species belonging to distantly-related taxa is highly problematic. Can we assume, without considering species ecology, abundance, or geographic distribution, that any "large" turtle or crocodilian would have been much more vulnerable to habitat fragmentation than any smaller lizard or bony fish, especially when small barriers can be impossible for small organisms to surmount? Such concerns lend weight to criticisms in the geological literature that habitat fragmentation may be untestable in the context of the fossil record.

For a work of this taxonomic scope and disciplinary breadth, errors of fact and presentation are to be expected; fortunately, they are rare and minor in this work. Crocodilians do not have a connection between left and right atria (p. 24), the site of cardiac influx; the connection (the foramen of Panizza) actually lies between the systemic arches after they leave the ventricles. The homeothermy conferred upon large tetrapods by virtue of their surface area / volume ratios makes the archaic terms "cold-blooded" and "warm-blooded" meaningless; yet these terms are used. Elsewhere, the author uses the more preferable "endothermy" and "ectothermy", but the latter is not defined. Parental care in squamate reptiles is widespread and more elaborate than the book recognizes. Finally, one might well question whether the distinction between speciation by budding and by bifurcation (p. 63) is at all real, especially in the context of punctuated equilibrium. But these are minor points that do not diminish the value of this book.

Dinosaur Extinction and the End of an Era is a stimulating work that offers a significant challenge to those who maintain that the late Cretaceous extinctions can be attributed to a single causal explanation. Archibald makes no secret of his disagreement with alternative interpretations -- not only with the Alvarez asteroid hypothesis, but with those of other paleontologists. Far from the final word on the subject, this book has elicited further discussion and analysis. Study sites in China and in South America that post-date this work offer independent tests of the hypotheses Archibald defines. For its time, however, this book offered an excellent guide to the ongoing research, debate, and analysis of how life changed and evolved at the end of the Mesozoic.

Whether interested in dinosaurs or in larger theoretical questions about patterns of speciation and extinction, biologists, students, and laypersons will find this book an excellent addition to their libraries. Other books in the Perspectives in Paleobiology and Earth History series are also well-worth seeking out; they include analyses of extinctions at the end of the Devonian, the Permian, and the Eocene, as well as the early emergence of animals during the distant Cambrian. ( )

Another terrific book in the Critical Moments of Life and Earth History series. ( )