s CM XI B HARY OF THE UNIVERSITY OF ILLINOIS 550-5 v.7-9 GEOLOGY UNIVERSITY OF ILLINOIS LIBRARY AT URBANA-CHAMPAJGN GEOLOGY Te st amped below. I ao , '3 I HE LIBRARY Of THE SEP 2 1944 UNIVERSITY OF ILLINOIS GEOLOGICAL SERIES OF FIELD MUSEUM OF NATURAL HISTORY Volume 8 Chicago, August 14, 1944 No. 13 THE AELURODON SAEVUS GROUP By Paul O. McGrew Assistant Curator, Paleontology The holotypes of the species of Aelurodon consist almost exclu- sively of jaws, maxillaries, or fragments thereof. Some (A. ferox, A. meandrinus, A. brachygnathns, A. wheelerianus) are based upon inadequate specimens and their specific status cannot be determined. The characters that have been considered as diagnostic for the vari- ous species are those of general size, relative size and massiveness of teeth, and general proportion. While differences in these characters do exist between the specimens described as representative of dis- tinct species the assumption that the differences are of specific significance has been based upon purely subjective evidence. The known species of Aelurodon fall clearly into two quite dis- tinct groups. One of these is represented by the types of saevus, haydeni, inflatus, and mortifer and the other by the type of taxoides and a number of specimens referred to wheelerianus. The taxoides group is readily distinguished from the saevus group by the following characters: (1) Horizontal ramus not tapering anteriorly, symphysis deep, (2) premolars little reduced, (3) heel of M T short, (4) entoconid of M T much reduced, (5) paraconid, metaconid and entoconid of M^ reduced. Actually the differences here are so great that there seems to be sufficient justification for the separation of the taxoides group as a distinct genus. In the collections of Field Museum are several jaws that are referable to the saevus group of Aelurodon. These specimens are from two fossil quarries, one at Big Spring Canyon, South Dakota, and the other near Ainsworth, Nebraska. As nearly as can be deter- mined from associated mammals the two deposits are of equivalent age. The morphologic range among the specimens seems superficially No. 563 79 VAT. HIST. 80 Field Museum of Natural History— Geology, Vol. 8 to be extensive. In size and general proportions they range from that of the holotype of A. saevus, one of the smallest and earliest of species, to that of the type of A. mortifer, one of the larger species. Two specimens from Big Spring Canyon — specimens that in all proba- bility represent the same local population — differ considerably more in some respects than do types of supposedly distinct species. These facts naturally create some doubt concerning the validity of several of the described species. In dealing with an extinct phylum such as Aelurodon, in which few specimens are known, and those from scattered localities, there is no direct way of determining the range of morphologic variation within a species. If, however, the range of variability is known in a related group of the same size, the variations that might be expected can be determined. At least this kind of comparison offers objective evidence that is considerably more than guess work. Thus, the use of the wolf, 1 Canis lupus, as a standard in assessing the taxo- nomic significance of characters in species of Aelurodon is justified because the wolf is closer in genetic relationships and size than any other species whose morphologic range can be reasonably well estimated. Fortunately nearly every character that has been used to dis- tinguish the species of Aelurodon is measurable, and quantitative comparison is therefore possible. Measurements of a series of characters of various species of the saevus group whose types are lower jaws are here compared with homologous measurements of specimens of the single species Canis lupus. The variants compared include nearly all of those that have been used for diagnosis of the various species. The specimens used are as follows: Type A. saevus; Niobrara River Type A. mortifer; Snake Creek, Nebraska Type A. inflatus; Big Spring Canyon, South Dakota F.M. No. 26307; Ainsworth, Nebraska F.M. No. 26306; Ainsworth, Nebraska F.M. No. 15800; Big Spring Canyon, South Dakota F.M. No. 15801; Big Spring Canyon, South Dakota If it is assumed that all of these fossil specimens represent one species, values for the coefficient of variation (V) and standard 1 Although the wolves have at times been placed in several species, the most authoritative recent work (Pocock, 1935; Goldman, 1937) indicates that, with the exception of Canis rufus of the lower Mississippi basin and the probable exception of Canis lycaon of eastern Canada, all belong to a single species, C. lupus. V.? U The Aelurodon Saevus Group 81 deviation (S.D.) should be close to those of the living form. If the values are significantly greater and no reason therefor may be found it may be assumed that more than one species is represented in the sample. The statistics for the specimens studied are as follows : Aelurodon Variant Observed range S. D. V* Length M T 27.9-33.3=5.4 1.79±.43 6. Oil. 6 Breadth M T 10.9-12.6=1.7 0.63±.17 5.3±1.4 Length P* 15.6-19.1 = 3.5 1.50±.43 8.8±2.3 Breadth P* 9.3-10.8=1.5 0.56±.16 5.6±1.5 Length P 7 11.1-12.9=1.8 0.60±.17 5. Oil. 3 Depth of jaw 26.9-36.3=9.4 3.33±.96 10.6±2.8 Cards lupus Variant Observed range S. D. V* Length M T 25.2-30.4=5.2 1.57±.30 5.65±1.0 Breadth M T 10.0-12.5=2.5 0.74±.14 6.6 ±1.2 LengthP* 14.5-17.0=2.5 0.69±.13 4.4 ±0.8 Breadth P* 6.8-8.7=1.9 0.49±.09 6.2 ±1.1 Length P 3 11.7-15.1 = 3.4 0.87±.16 6.2 ±1.2 Depth of jaw 22.8-31.2 = 8.4 2.27±.48 8.7 ±1.6 *V (coefficient of variation) is calculated from the standard deviation. The agreement here is surprisingly close in all variants except the length of Pt. The differences in the coefficient of variation and the standard deviation for the depth of the jaw are not great, consid- ering the variable nature of the character, which is strongly affected by age, sex, and physical development of the individuals. Ordinarily the differences in the values of the coefficient of vari- ation and the standard deviation for the length of Pt would indi- cate that the unknown sample includes more than one taxonomic unit of the same magnitude as that of the sample with which com- parison is made. In this particular case, however, there is good reason to believe that the normal range in this character in a single species of Aelurodon was much greater than in Canis lupus, and that more than one species is not necessarily indicated. The three specimens from Big Spring Canyon in all probability represent a single local population. The observed range in these specimens is 2.9 mm. (15.8-18.7). This is within 0.6 mm. of the observed range for the entire sample and is 0.4 mm. greater than the observed range for the fourteen specimens of Canis lupus. Such high variability in this character should be expected in Aelurodon, because, being closely related, or ancestral, to such genera as Osteoborus and Boro- phagus, with highly modified Pt's, this character was in a state of U. OF ILL LIB. 82 Field Museum of Natural History — Geology, Vol. 8 rapid change. Contrary is the condition in Canis, which is a long- established, conservative form with unmodified Pt. This analysis suggests the possibility that all species of Aelurodon (except A. kaydeni) whose types are lower jaws, should be referred to A. saevus. It does not prove that all belong with one species but does prove that on the characters so far used more than one species cannot be demonstrated. The specimens here included in Aelurodon saevus range in age from Barstovian to late Clarendonian. The earlier specimens appear on the whole to be smaller than the later. It will probably be shown that there was a progressive increase in size. A popula- tion sample from the Barstovian will probably prove significantly different from a similar sample from the Clarendonian, but the gap theoretically will be filled by intergrading series from the inter- mediate horizons. The entire morphologic range is of about the same magnitude as is found in a comparable living species. Thus, if the extinct group is to be broken into smaller taxonomic units, the smaller ones will logically be regarded as subspecies. The entire series, then, may represent a chronocline, 1 as defined by Simpson (1943). Aelurodon haydeni, judging from the original description and the preservation of the type, was found in the same deposits as A. saevus. Aelurodon haydeni is surely a valid, distinct species. Comparing the type of the larger form with the group of specimens here regarded as A. saevus we find that the breadth of My and length of P^ are sig- nificantly greater, with P values 2 of less than 0.01; the length of Mt and the breadth of Pt are probably significantly greater, with P values of 0.02 or less; the length of Pt and the depth of the jaw are not necessarily significantly greater. The principal differences between A. haydeni and A. saevus are the large size and slightly different proportions of the former. These differences correspond rather closely to those between the coyotes and wolves of today. An interesting possibility is suggested by the fact that Aelurodon haydeni, as a contemporary of the smallest member of the saevus 1 Prior to Simpson's first published use of the term "chronocline" (1943, p. 174) Alfred E. Emerson used the term "chronological cline" (Amer. Nat., 77, p. 102, 1943) with the same meaning. Because of advantages in using the shorter, perfectly clear single word, and with Dr. Emerson's approval, I prefer to use and help establish the paleontologically significant "chronocline." 2 P represents the probability that "the single deviation could have been equaled or exceeded by one drawn at random from the population represented by the sample." (Simpson and Roe, 1939, p. 189.) Canis lupus Length Mt Breadth Mr Length Pr Breadth Pt Length Pu Depth of jaw under Mt Length Mt Breadth Mt Length Pr Breadth Pr Length Pr Depth of jaw under Mt Aelurodon saevus (incl) .06 OH .02 i O 02 .OH .06 Fig. 26. Logarithmic plot showing relative range of morphologic variation in Canis lupus and Aelurodon saevus (inch). For Aelurodon those specimens listed on page 80 were used, and for Canis lupus fourteen specimens in the col- lections of Field Museum. Horizontal lines show the range of variation and vertical lines indicate the mean of each character. 83 84 Field Museum of Natural History— Geology, Vol. 8 group, is distinguished primarily by its larger size. If A. haydeni had been found in deposits of later age it might be construed as a terminal member of the A. saevus chronocline. If a contemporary morphologic intergrading series were found connecting A. saevus and A. haydeni we would have a chorocline more or less duplicating a chronocline. Such conditions might well be found to exist in some groups when sufficient data are available. I am indebted to Dr. Anne Roe Simpson for checking on my statistical procedure. Any errors that may exist, however, are my own. REFERENCES Goldman, E. A. 1937. The Wolves of North America. Jour. Mamm., 18, pp. 37-45. Pocock, R. I. 1935. The Races of Canis lupus. Proc. Zool. Soc. Lond., 1935, pp. 647-686, pis. 1, 2. Simpson, G. G. 1943. Criteria for Genera, Species and Subspecies in Zoology and Paleozoology. Ann. N. Y. Acad. Sci., 44, pp. 145-178. and Roe, Anne 1939. Quantitative Zoology, xviii + 414 pp., 52 figs. McGraw-Hill, New York and London.