Contact: Dan Lagiovane (412) 622-3361

EMBARGOED BY NATURE
UNTIL 2 PM EDT, MARCH 25, 1999

Roots of the Mammalian Family Tree
Timothy Rowe

An exceptionally complete skeleton, dating back roughly 140-150 million years, offers our closest look yet at the last common ancestor of modern mammals.

Mesozoic mammals are among the most challenging and prized entries on a fossil hunter's life-list of discoveries. The essence of the challenge is size - the Mesozoic history of mammals was played out by tiny animals (fig. 1). Few specimens survived the destructive agencies of fossilization, and those that did are supremely difficult to find and collect. Most are fragmentary and most named species are based on isolated teeth and jaws. In nearly two centuries of searching, only a few precious complete specimens have been recovered and, without better fossils, long segments of mammalian history have remained in the dark. But [in their article], Ji, Luo, and Ji¹ describe one of the most complete and exquisitely preserved specimens ever found. It comes from the same Late Jurassic/Early Cretaceous deposit of Liaoning, China that recently yielded spectacular feathered dinosaurs² and a complete therian mammal skeleton³. The latest discovery helps to fill a wide gap in the fossil record, and brings new information to classic problems on the origin and interrelationships of early mammals.

The first Mesozoic mammals were discovered in 1812 by a mason in a tilestone quarry near Heddington, England. These fossils came from the Middle Jurassic (roughly 165 million year old) Stonesfield Slate, and consisted of two isolated lower jaws, each belonging to a different species. Today the Stonesfield jaws are still the oldest-known fossils of the 'crown clade' Mammalia⁴ - the lineage founded by the last common ancestor of living mammal species (Fig. 2).

The tiny jaws quickly found their way to the University of Oxford where, in 1818, Baron Georges Cuvier examined them on a sojourn to England. Renowned for his ability to judge the nature and affinities of an extinct animal from a part or even a single fragment of a skeleton, Cuvier pronounced the Stonesfield specimens to be mammalian⁵. He lived up to his reputation, and his identification was the first of many violations of what had been considered a very general rule - that mammals did not live during the Age of Reptiles.

A century-and-a-half later and hundreds more Mesozoic mammal fossils had been discovered, yet the two Stonesfield jaws remained among the most complete specimens known. Screen washing techniques pioneered by Claude Hibbard in the 1940s offered the first clues to an unsuspected diversity of Mesozoic mammalian species⁶. Over several decades and a huge investment of labor, many tons of Mesozoic sediments were sieved through a succession of screens graded to trap even the smallest fossils. But most recovered specimens consisted of only teeth and broken jaws, and the emerging view of early mammalian history became overly focused. In 1968 Alfred Sherwood Romer admonished, "So great has been this concentration on dentitions that I often accuse my 'mammalian' colleagues, not without some degree of justice, of conceiving of mammals as consisting solely of molar teeth and of considering that mammalian evolution consisted of parent molar teeth giving birth to filial molar teeth and so on down through the ages"⁷.

The first great advance toward a more complete knowledge of the structure and relationships of Mesozoic mammals came in the 1960s, when the Polish Academy of Sciences sent a series of expeditions into central Asia. Dozens of Late Cretaceous mammal skeletons representing several different lineages were collected⁶. Throughout the 1990s, Asian expeditions led by the American Museum have been collecting hundreds more Late Cretaceous specimens that document, in even greater detail, the initial diversification of therian mammals (fig. 2)^8-9. And discoveries elsewhere have added to this wealth of new data, initiating a renaissance in our knowledge of early mammalian history.

Computer-assisted cladistic analyses of data from these more complete specimens ^4,8-9 profoundly altered the picture of Late Cretaceous mammalian diversity that was painted in Romer's time¹⁰. For example, Romer's generation accepted that Mammalia arose in the Triassic (which immediately preceded the Jurassic), whereas the new analyses indicate that the species ancestral to living mammals probably lived in the Early or Middle Jurassic, and that Mammalia is 20 to 40 million years younger than once believed. Until very recently, however, the earliest details mammalian history were obscured owing to the lack of complete fossils.

The remarkable specimen described by Ji and colleagues¹ is, along with a primitive therian mammal announced last year from the same locality³, by far the most complete and informative ever discovered for a 20 Myr or longer segment of Jurassic and Early Cretaceous time. Ji et al. Present an analysis of evolutionary relationships, including dental evidence and data from throughout the skeleton, that places the new find very near the base of the mammalian crown clade (Fig. 2). Their analysis indicates that Triconodontidae - a group once believed to contain the direct ancestors of modern mammals - is not a natural group. Originally founded on the dental attributes that inspired its name, some triconodonts seem to be closer to mammals than to other so-called triconodonts.

But discovery of Ji, Luo and Ji also highlights the presence of homoplasy - the independent evolution of similar features- in skeletal characters, corroborating an earlier finding that no region of the skeleton is immune to homoplasy¹¹. Still, the data to be gleaned from the skeleton are strong enough to overthrow the apparent dental resemblance of the triconodonts to one another. And it is skeletal characters that largely support the sister group relationship of multituberculates (a long-lived and long-enigmatic lineage of extinct mammals) with therian mammals, once again contradicting hypotheses derived from dental evidence.

This beautiful specimen also offers new insight into what the ancestor of modern mammals was like. Working out when mammals first moved into the trees, and whether this happened more than once, has been problematic. Ji and colleagues' find indicates that the crown lineage Mammalia arose as a terrestrial species, and that only later did its therian descendants become arboreal.

Even with this spectacular new find, long gaps still punctuate our Mesozoic record of mammals and their extinct relatives. But this exciting Chinese locality has now produced so many exquisite tetrapod fossils that additional complete specimens of early mammals are likely to be unearthed. We can then expect rapid increases in the resolution of what was once the most fragmented segment of our early history.

This article first appeared in the March 25, 1999, issue of the journal Nature. Timothy Rowe is in the Department of Geological Sciences and the Vertebrate Paleontology Laboratory, The University of Texas at Austin, Austin, Texas 78712 USA

email: rowe@mail.utexas.edu

1. Ji, Q., Luo, Z., & Ji, S.-A. A Chinese triconodont mammal and mosaic evolution of the mammalian skeleton. Nature 398, 326-330 (1999).

2. Ji, Q., Currie, P. J., Norell, M. A., & Ji, S.-A. Two feathered dinosaurs from northeastern China. Nature 393, 753-761 (1998).

3. Hu, Y., Wang, Y., Luo, Z., & Li, C. A new symmetrodont mammal from China and its implications for mammalian evolution. Nature 390, 137-142.

4. Rowe, T. Definition, diagnosis and origin of Mammalia. J. Vert. Paleontol. 8, 241-264 (1988).

5. Owen, R., Observations on the fossils representing the Thylacotherium provostii, Valenciennes, with reference to the doubts of its mammalian and marsupial nature recently promulgated; and on the Phascolotherium bucklandi. Geological Transactions (2nd series) 6, 47-65 (1839).

6. Lillegraven, J. A., Kielan-Jaworowska, Z., & Clemens, W. A. Mesozoic Mammals - the first two-thirds of mammalian history. University of California Press, Berkeley (1979).

7. Romer, A. S. Notes and comments on Vertebrate Paleontology. University of Chicago Press, Chicago (1968)

8. Novacek, M. J., et al. Epipubic bones in eutherian mammals from the Late Cretaceous of Mongolia. Nature 389, 483-486 (1997).

9. Rougier, G. W., Wible, J. R., & Novacek, M. J. Implications of Deltatheridium specimens for early marsupial history. Nature 396. 459-463 (1998).

10. Dingus, L., & Rowe, T. The Mistaken Extinction - Dinosaur Evolution and the Origin of Birds. W. H. Freeman & Co., New York (1997).

11. Rowe, T. in Mammalian Phylogeny (eds Szalay, F. S., Novacek, M. J., & McKenna, M. C.) 129-145 Springer-Verlag, New York (1993).

• The Original Press Release
• Images to Accompany this Article