January 2005] Reviews 367

Trivers's theory of parent-offspring conflict associated fauna and flora, biotic history of has shed relatively little empirical light on sib- Australia, possible feeding habits, and the like. licide in birds will undoubtedly provoke some The book's concept, organization, and visual raised eyebrows. But Mock's perspectives are so presentation are brilliant, but the execution has clearly articulated and thoughtfully explained some serious flaws. that even readers with dissenting views will be The first known species, Dromornis australis, unlikely to object strenuously. was described in 1874 by Richard Owen, and I highly recommend this book to anyone inter- for almost a century and a quarter the drom- ested in the evolutionary biology of family con- ornithids were associated with paleognathous flict. It will be especially useful to ornithologists ratites such as emus and cassowaries. The name working on such topics as hatching asynchrony "mihirung" was originally adopted for these siblicide, brood reduction, and parental care. birds by Rich (1979) from Aboriginal traditions And for anyone wanting to know how to write of giant emus (mihirung paringmal) believed pos- a scholarly biological book that will appeal to a sibly to apply to Genyornis. It was not until the general audience. More Than Kin and Less Than seminal paper of Murray and Megirian (1998), Kind should be essential reading.•RONALD L. based on newly collected Miocene skull mate- MUMME, Department of Biology, Allegheny College, rial, that the anseriform relationships of the 520 North Main Street, Meadville, Pennsylvania Dromornithidae were revealed. Six years later, 16335, USA. E-mail: [email protected] Murray and Vickers-Rich glibly and rather mis- leadingly refer to these birds as gigantic geese and imply that their nonratite nature should have been apparent earlier. E. C. Stirling and A. H. C. Zietz, who were director and assistant director, respectively, of the South Australian Museum, excavated and The Auk 122(1):367-371, 2005 published impressive monographs on extensive © The American Ornithologists' Union, 2005. fossil material of G. newtoni from 1896 to 1913. Printed in USA. According to Murray and Vickers-Rich,

Stirling and Zietz's comparisons and Magnificent Mihirungs. The Colossal Flight- discussion of Genyornis morphology and less Birds of the Australian Dreamtime.•Peter relationships are so intently focused on F. Murray and Patricia Vickers-Rich. 2004. Indiana ratite osteology that as each structural University Press, Bloomington, Indiana, vii + 410 incongruity is realized they dutifully note pp. ISBN 0253342821. Quarter cloth and boards, it, adjust their spectacles, and move on as $75.00.•This is an exhaustive, superbly illus- though there were no alternatives, (p. 59) trated treatment of the Dromornithidae, a family of large to very large flightless birds known from Furthermore, the skull "contains ample evi- fossils from Australia and Tasmania. The fossils dence to have placed the Genyornis [sic] among range in age from a rather equivocal early Eocene the Neognathae even at the time Stirling partial footprint to the late Pleistocene Genyornis described it" (p. 60). The penetrating clarity of newtoni, which died out about 50,000-30,000 Vickers-Rich's hindsight is a scientific marvel, years ago, presumably as a result of human pré- for in her own monographic treatment of the dation and habitat modification by fire. Dromornithidae (Rich 1979) they were ratites The book is divided into four major sections: from start to finish. "Discovery," a short chapter on the history of For the record, the first person in the history fossil finds and their discoverers; "Systematics of the Dromornithidae to insist that these birds and Morphology," the longest section, con- could not be ratites was unlovable old cladist- taining paleontology, descriptive osteology, baiting moi (Olson 1985). This seemingly signifi- systematics and phylogeny, and evolutionary cant fact, acknowledged by Murray and Megirian origins; "Paleobiology," consisting of functional (1998), was omitted by Murray and Vickers-Rich. morphology, biomechanics, weight estimates, Although the reference can be found in the bibli- and so forth; and "Paleoecology," which treats ography, it is not cited in the text. 368 Reviews [Auk, Vol. 122

Although I was pleased to learn of the evi- lacking the hindlimbs, from the London Clay dence that the dromornithids were derived and was referred to the Anseranatidae on the from Anseriformes, one might now fairly ask basis of highly distinctive derived characters of "Derived how many times?" Nowhere here the pectoral girdle. The bill morphology indi- or in Rich (1979) is there a clearly articulated cates very clearly that it was a filter-feeder. The argument or character analysis demonstrat- type species of Anatalavis from marine deposits ing that the Dromornithidae constitute a in New Jersey is either late Cretaceous or earli- monophyletic group. The skull and foot struc- est Paleocene. Thus, the earliest certain member ture of the clade that includes Genyornis are of the Anseriformes, which is also the earliest very different from the skull and foot, when member of the Anseranatidae, was a filter- known, of the other members of the family. If feeder. This strongly suggests that the macro- all these birds are large flightless derivatives feeding Magpie Goose is secondarily derived of Anseriformes, why might not the Genyornis from a filter-feeder. Screamers, too, may thus clade have evolved large size and flightless- be so derived. ness independently within the Anseriformes? Unfortunately, the two known early Eocene In fact, because details of the skull of Genyornis taxa of Anhimidae, from Wyoming and are poorly known (owing to the incomplete, England, have never been described. Those crushed nature of the available specimens), the were not filter-feeders and lacked many of the evidence that the relationships of this group lie autapomorphic characters of modern scream- with the Anseriformes is less satisfactory than ers, such as great skeletal pneumaticity and in older taxa for which better skulls are known. the double-spurred carpometacarpus. Murray For example, the large, blade-like retro-articular and Vickers-Rich make a weak case for drom- process shown for Genyornis, one of the most ornithids being screamer-like, on the basis of characteristic features of the Anseriformes, the supposed lack of uncinate processes and appears to be an almost entirely hypothetical the presence of a knob at the distal end of the construct (fig. 107, p. 127). Monophyly of the pectoral crest of the humérus. Eocene scream- Dromornithidae is, therefore, an issue that still ers lack that knob, however. Furthermore, the needs to be addressed. authors show a rib of D. stirtoni (fig. 60) with a Given that the dromornithids, or at least some large, very distinct facet for an uncinate process. of them, belong in the Anseriformes, where do Such a facet would seem to indicate a synovial their relationships lie within the order? This joint, and 1 doubt that such would form were question gets more consideration than that of there not a bony uncinate process to articulate monophyly, but its treatment is badly distorted with it. by prejudices and by another, more serious, In the end, no good case is made for the omission of pertinent literature. relationships of the Dromornithidae within The authors are dismissive, even derisive, the Anseriformes, even at the level of family. of the suggestion of Olson and Feduccia (1980) Although the authors favor a closer relationship that screamers (Anhimidae) might be second- with either the Anhimidae or Anseranatidae, it arily derived macrofeeders that evolved from would seem that even the Anatidae, through a filter-feeding ancestor. They conclude that a terrestrial goose-like form such as Cereopsis, macrofeeders such as screamers and the Magpie cannot be ruled out. Goose (Anseranas semipalmata) represent the The generic-level systematics used in this primitive condition in Anseriformes and that book is a complete mess. To begin with, the filter feeding is derived. They also state that the type species of Dromornis, Owen's D. australis, Anseranatidae have no fossil record. Did they is known only from a single femur and remains simply overlook the early Eocene Anatalavis practically the only Pliocene fossil of the fam- oxfordi (Olson 1999)? That species was published ily. The paucity of Pliocene material makes it in one of the quadrennial proceedings of the uncertain that the late Miocene D. stirtoni is cor- Society of Avian Paleontology and Evolution, rectly referred to Dromornis. The close relation- each of which has become a primary source ship between the middle Miocene Bullockornis in avian paleontology and ought to be famil- planei and the larger D. stirtoni that succeeds iar to everyone in the field. Anatalavis oxfordi it is emphasized repeatedly. At first implicitly was based on an excellent associated skeleton. (p. 273), and then more explicitly (p. 330), it is January 2005] Reviews 369 suggested that this is an ancestral-descendent farrago of indecision before such an important relationship. Why, then, are the two species book was published. Still, because of the supe- maintained in separate genera? rior nature of the illustrations, it is possible for Then we have the species that was origi- an intelligent reader to make some sense of part nally described by Rich (1979) as Ilbandornis'? of the evolutionary history of these birds, in lawsoni. The only change in status for this spe- spite of the disastrous nomenclature. cies in 25 years is that quotation marks were A lengthy chapter on body mass estimates added around the genus, so that everywhere goes into great detail to document three dif- it appears as "Ilbandornis?" lawsoni. Do the ferent methods of estimating mass and rather quotation marks make the identity even more diffidently concludes that D. stirtoni may have uncertain than before, or make us more certain been the heaviest bird that ever lived. Here, as of the uncertainty, or what? The illustrations elsewhere throughout the book, there is much and descriptions make it very clear that this additional information and speculation on late Miocene species has nothing to do with other large flightless "ground birds" such as Ilbandornis but shares distinctive characters Diatryma, moas, and phorusrhacids, so that the with Genyornis. It would have been far prefer- volume should become an essential reference able to have simply called it G. lawsoni rather for anyone studying those birds. than carry it through the entire book, befogged Sections on appearance and posture, locomo- in punctuation, in a genus to which it patently tion, and feeding apparatus are replete with does not belong. Another taxon is referred to beautiful anatomical reconstructions. As with throughout as "TBullockornis" sp. Does the fact any group of organisms known only from fos- that the question mark comes before the genus sils, one must rely to a greater or lesser extent here, rather than after, have any significance? on conjecture, and there will doubtless be The quotation-mark fetish reaches its apogee more than one interpretation of the structure with "Dromornis" australis (p. 304). Because aus- of dromornithids. But I have never known any tralis is the type-species of Dromornis, it belongs two functional anatomists to agree on anything, in Dromornis by definition, and the quotation even concerning living birds, so this should not marks add nothing but confusion. be regarded as a detraction. The following example epitomizes the addled Complementing the functional anatomical systematics used in this book: discussions are chapters on paleoecology and the fauna and flora associated with dromorni- The genera [sic = species] Ilbandornis thids through the Cenozoic. What emerges is woodburnei and "?Bullockornis" sp. have a most useful overview of the evolution of ter- slightly more derived morphological states restrial biotas on the Australian continent that and share no definite synapomorphic should be particularly useful for those outside states with Genyornis or "Ilbandornis'?" Australia who are unfamiliar with this history. lawsoni. They retain several plesiomorphic These sections are not without a good measure states but also share some derived states of advocacy, conjecture, and redundancy, but with Bullockornis planei and Dromornis the overall conclusions seem quite reasonable stirtoni. Despite the current retention of and believable to me. distinct generic names, they probably In a nutshell, the dromornithids are presented represent species of a single genus, among as large, browsing herbivores that evolved in which were structurally suitable ancestors open-canopy scleromorphic forests. They were for Bullockornis planei and ultimately capable of moving over considerable distances Dromornis stirtoni. "IBullockornis" sp. with reasonable celerity, and the taxa at any represents a species close to the ancestry given period are believed probably to have of Ilbandornis woodburnei. Cranial been widespread on the continent. The authors fragments are known for "7Bullockornis" repeatedly argue that most of Australia was sp. and Ilbandornis woodburnei, indicating not a wet, closed-canopy rainforest as appar- a close relationship, (p. 329) ently has often been asserted. From the middle IVfiocene onward, dromornithid diversity It is a great pity that a competent systematist declined, as did overall body size; in contrast, was not enlisted to sort out this horrendous diprotodontids and other large herbivorous 370 Reviews [Auk, Vol. 122 marsupials show overall increases in size over is defined only as "the space occupied by the the same interval. Those animals are believed to brain," which called to my mind the jar of alco- have been better adapted than dromornithids hol in which Einstein's brain now floats. Is that to the decreasing quality of browse as climatic an endocranial fossa? We can only imagine the conditions became drier and plants responded perplexity of the nonscientist who, fifteen lines accordingly. down, finds that "fossa" is "a slender cat-like Following the main text is an idiosyn- carnivorous mammal from Madagascar." cratic appendix entitled "Basic Avian Skeletal Minor errors of every description are Anatomy," which emphasizes comparisons disturbingly frequent: typographical, spell- between taxa thought relevant to understanding ing, grammatical, word-choice, factual, bib- the osteology of dromornithids, namely emu, liographic. These start on the first page of text magpie goose, and megapode. Terminology is (Acknowledgments: p. vii), where Brad Livezey supposed to be based on Howard (1929) rather appears as "Brad Linezen," and continue than adopting "the strictly formal Latin nomina throughout the book to the last page of the bib- anatómica" (p. 337), as though that were a medi- liography, where we find the following refer- eval incunabulum rather than a citable reference ence: "Zeitz, A. C. 1894. Nature 50:184-208." The (Baumel 1993). The reader is unlikely to be much name should be spelled "Zietz," not "Zeitz," enlightened by the tedious descriptive text, but and the article was not authored by Zietz any- the section is salvaged by the excellent illustra- way, but by Stirling. The title should not have tions. Discrepancies in anatomical nomencla- been omitted, and the actual pagination is 184- ture are evident: e.g. "furculum" (fig. Al) vs. 188, 206-211. Out of curiosity, I checked each of the correct "furcula" (fig. A12) or "pygidium" the other 10 references on that less than half a (fig. A12) vs. the correct "pygostyle" (fig. Al). I page and found 11 additional errors, including suspect that there are even more serious errors completely erroneous pagination for another to be found here, especially among the details reference. of the cranium. What is termed at one point the "Magnificent Because many of the legends are long Teratorn" (p. 255) is later called "Great Teratorn" and complex, one wishes that more labeling (p. 275) but is never identified by its scientific had been included directly on the figures. name (Argentavis magnificens). Then, Teratornis Nevertheless, the quality of the copious illus- is erroneously stated to have been "the largest trations is uniformly excellent. The stunning flying bird known" (p. 319), when that honor life-reconstruction of a pair of Bullockornis planei actually goes to Argentavis. And so on. by Peter Trusler that adorns the dust jacket is Don't get me wrong. This is still a highly one of the most arresting of its kind that I have meritorious and impressive book. Anyone with ever seen. Because many librarians discard dust an interest in morphology, paleontology, and jackets, it is fortunate that this painting is repro- evolution of birds; in the evolution of terrestrial duced in color on the half-title and title pages. ecosystems; or simply in exquisite scientific The two-page black-and-white reproduction illustration, will find much to learn and enjoy that appears later in the text is washed-out and in these pages. In concluding, the authors note completely ineffectual. some recent discoveries and remark that "after a The Glossary may have been an afterthought, decade of digging, preparing, comparing, mea- perhaps at the insistence of an editor, in which suring, and hypothesizing, our effort remains a a selection of words were randomly picked work in progress" (p. 335). My sincere hope is from the text and crude definitions supplied. that, in another decade, knowledge of dromor- This makes for entertaining reading, as there nithids will have advanced so far as to merit a are some real howlers. Definitions may be com- reissue of this work in which the new informa- pletely wrong ("Alpha taxonomy: A classifica- tion can be incorporated and all the flaws of tion of organisms based on overall similarity the present edition corrected. This might then of morphology") or misleading or unhelpful become one of the great classics in both orni- ("Dorsal: The top side"; "Lacrimal: A bone of thology and paleontology •STORRS L. OLSON, the skull"). Mandibles are defined as "lower Division of Birds, National Museum of Natural jaws," though the term "upper mandible" is History, Smithsonian Institution, Washington, D.C. used throughout the book. "Endocranial fossa" 20560, USA. E-mail: [email protected] January 2005] Reviews 371

LITERATURE CITED study•in this case, the reality of species. They unapologetically justify the biological species BAUMEL, J. J., ED. 1993. Handbook of avian anat- concept as an appropriate framework for study- omy: Nomina anatómica avium, 2nd ed. ing the origin of species. And they are explicit Publications of the Nuttall Ornithological about what they consider to be the central con- Club, no. 33. ceptual theme in speciation research: the origin HOWARD, H. 1929. The avifauna of the Emeryville and evolution of reproductive isolating barriers. shellmound. University of California Coyne and Orr do not dabble in semantics or Publications in Zoology 28:301-394. philosophy, and the book cuts quickly to the MURRAY, P. F., AND D. MEGIRIAN. 1998. The skull of process of species formation, relegating the dromornithid birds: Anatomical evidence for traditional debate (or quagmire) over species their relationship to Anseriformes. Records concepts to a carefully worded appendix. of the South Australian Museum 31:51-97. The remainder of the book is a tour and sta- OLSON, S. L. 1985. The fossil record of birds. tus review of the most significant facets of the Pages 79-238 in Avian Biology, vol. 8 (D. S. speciation process: the geography of speciation, Farner, J. R. King, and K. C. Parkes, Eds.). the nature of isolating barriers, the genetics of Academic Press, New York. reproductive isolation, speciation by reinforce- OLSON, S. L. 1999. The anseriform relationships of ment, polyploidy, speciation by hybridization, Anatalavis Olson and Parris (Anseranatidae), the relative importance of natural selection with a new species from the Lower Eocene and genetic drift, and macroevolutionary con- London Clay. Smithsonian Contributions to siderations. Each chapter follows a predictable Paleobiology 89:231-243. format: the authors examine theoretical and OLSON, S. L., AND A. FEDUCCIA. 1980. Presbyornis experimental evidence, as well as evidence and the origin of the Anseriformes from nature. They critically revisit the literature (Aves: Charadriomorphae). Smithsonian (including their own work) and provide their Contributions to Zoology, no. 323. own conclusions and synthesis. Through it all, RICH, P. V. 1979. The Dromornithidae, an extinct the authors demand testable hypotheses and family of large ground birds endemic to insist on examples from nature wherever pos- Australia. Bureau of National Resources, sible. Such hard-nosed empiricism, from two Geology and Geophysics Bulletin 184:1-196. scientists who clearly understand the theory, is very refreshing. Some readers may find Coyne and Orr to be overly critical•perhaps downright negative•in their assessment of previous research. They hold all studies to a hard standard, and it occasionally seems difficult to do anything properly in their The Auk 122(l):371-373, 2005 world. But they are at least consistent, and they © The American Ornithologists' Union, 2005. clearly indicate what needs to be done in future Printed in USA. studies. Their extensive discussion of sympatric speciation (Chapter 4) is a good example of this. To satisfy sympatric speciation, they hold a long Speciation.•Jerry A. Coyne and H. Allen list of potential cases to extreme scrutiny and a Orr. 2004. Sinauer Associates, Sunderland, standard of evidence that is difficult to obtain. Massachusetts, xiii + 545 pp. ISBN 0-87893- They conclude that, although several promis- 091-4. Cloth, $89.95. ISBN 0-87893-089-2. Paper, ing cases exist, sympatric speciation appears to $54.95.•This is an important book, perhaps the receive far more attention than it warrants. One most important work on the subject of speciation may disagree with their dismissal of some puta- in decades. The species problem•explaining tive cases of sympatric speciation, or with their the origin of discrete groups living together null hypothesis of allopatry•that speciation is in nature•is undoubtedly one of the greatest allopatric until proven sympatric. But in the end, questions in all of biology. 1 found the book's the lack of evidence for sympatric speciation logic extremely compelling. The authors first when it should be detected (e.g. among species establish the existence of the phenomena under on small oceanic islands or among host-specific