POLYPHYLETIC OR MONOPHYLETIC ANCESTRY OF MAMMALS, OR: WHAT IS A CLASS?

Abstract

A recent issue of EVOLUTION carried two articles (Olson, 1959; Simpson, 1959) on the origin of mammals. Olson looked up, as it were, from the murky depths of the Pennsylvanian and saw the ongoing evolution of the Synapsida culminating in the late Triassic in several lines which he could call mammals. Simpson adopted the opposite point of view, and looked down the mammalian evolutionary sequence from the end of the Mesozoic. However, he arrived at the same time-level and the same conclusion as did Olson: mammals are polyphyletic, and the several lines involved (possibly nine, probably no fewer than four) independently crossed the reptilian-mammalian boundary in the late Triassic or near the division between the Triassic and Jurassic. Simpson wrote further, far as I know, all subsequent authors who have explicitly studied the Mesozoic mammals have agreed that they were polyphyletically derived from reptiles. Olson stated essentially the same (p. 348), although he did mention that a few students disagreed. However, the paleontologists, at least, are in almost total agreement. Simpson, with his usual prescience, proceeded to clarify the possible implications, in terms of formal taxonomy, of the meaning of the words 'Class Mammalia' if the present conclusions as to polyphyletic origins of the class are considered valid. Of various alternatives presented (p. 413) he personally preferred his third, The groups known to have reached mammalian grade (by whatever of the diagnoses one prefers) may be retained as a phylogenetically based Class By 'phylogenetically based' he means . arose wholly from one of lower categorical level, as Class Mammalia from Order Therapsida. As matters stand at present, then, several different groups within the Class Mammalia (as commonly accepted) are related to each other only by descent through several groups of therapsid reptiles. Such a concept of the meaning of 'Class' and of 'mammal' will be strange to most zoologists, trained as they have been in the tradition of monophyletic descent of all members of any taxon. Simpson does, indeed, suggest the possibility of a logical monophyletic beginning for all mammals by shifting the origin of the class backward in time to a common ancestor (his alternative 4, p. 413): Some or all of the animals now called mammallike reptiles may be included in the Mammalia. consequences of such a decision would obviously complicate the work of the paleontologists by changing the bases for differentiating between mammalian and reptilian fossils to different, even if more indefinite, characters than now used. Also, I suspect, such a decision would be distasteful to their present feeling of the proper order of things. Furthermore, continued Simpson, . such a radical change would make all current and earlier literature confusing and would greatly complicate teaching and popularization. Perhaps it is time that one who has not . explicitly studied the Mesozoic mammals , and thus one not hampered by the necessity of identifying specific fossils, had an overall look at the Droblem. UDon doing so, several