MULTIVARIATE ANALYSIS OF PLEISTOCENE 



AND RECENT COYOTES (CANIS LATRANS) 



FROM CALIFORNIA 



BY 



EUGENE GILES 

 (A contribution from the University of California Museum of Paleontology) 



ABSTKACT 

 Multivariate analysis of cranial measurements of Raneho La Brea and McKittrick (California 

 Pleistocene) coyotes demonstrates that they form a well-delineated group that should receive at 

 least the subspecifiic distinction from Recent California coyotes that is usually accorded them 

 (Canis latrans orcutti). C. lepophagus (Texas Blancan) seems to rate only subspecifie distinction 

 within C. latrans. Discriminant analysis indicates that C. petrolei from Raneho La Brea is 

 dubiously distinguished from C. latrans orcutti. The separate species status of C. andersoni, also 

 from Raneho La Brea, is questioned. 



INTRODUCTION 



In paleontological classification the paleontologist is required to make a ju- 

 dicious assessment of the relative values of numerous morphological features of the 

 specimen or specimens he is classifying. As an aid to this evaluation he uses taxo- 

 nomic principles based on studies of existing forms of life, either his own or those 

 of others. But in applying these principles the paleontologist lacks, to varying de- 

 grees, evidence that is available to the taxonomist of living animals and plants. The 

 technique used in this study provides a means of analyzing morphological differ- 

 ences and assessing their significance objectively; at the same time, it can be recon- 

 ciled with variations in individual growth and, to a lesser degree, with the genetic 

 concepts underlying classification. 



If the focal point of zoological classification is considered to be the definition of 

 the species with reference to discrete, potentially interbreeding animal popula- 

 tions, the paleontologist must reconcile his morphological groupings with this con- 

 cept of general zoological systematics. Wright (1950) has observed that diagnosis 

 by morphology does not at all indicate abandonment of a genetic basis for the 

 classificatory units. The morphological similarities and differences intimately 

 associated with the underlying genetic system inevitably provide in practice the 

 basis for paleontological classification, with whatever assistance relevant strati- 

 graphy may offer. The difficulty lies in determining the significance of morpho- 

 logical differences. 



The degree to which morphological variables may be influenced by genetic, eco- 

 logical, or other factors is one critical problem. Since such factors do not operate 

 independently or remain the same through time according to any theory of natural 

 selection, the problem is complex. Even if the groups under consideration were 

 geographically stable and not subject to infusions from without, evolutionary con- 

 vergence and parallelism might interfere with the tracing of lineages by means of 

 the material paleontologists ordinarily use. Differing rates of evolution, possible 

 polyphyly, and confused stratigraphy pose additional difficulties. 



[ 369 ] 



