A Study of Variation in Eastern Timber Rattlesnakes. 

 Crotalus horridus Linnae (Serpentes: Viperidae) 



Christopher W. Brown and Carl H. Ernst 



Department of Biology, 



George Mason University, 



Fairfax, Virginia 22030 



ABSTRACT. — Variation was examined in specimens of Crotalus 

 horridus from the eastern United States in an attempt to investigate 

 the status of its two described subspecies, C. h. horridus and C. h. 

 atricaudatus, as defined by Gloyd. A particular effort was made to 

 duplicate the results of a study by Pisani, Collins, and Edwards, who 

 concluded that the subspecies were invalid. Maximum likelihood fac- 

 tor analysis and step-wise discriminant analysis on the same morpho- 

 logical characters, plus several others relating to adult size and pattern, 

 produced evidence that the two subspecies of C. horridus are valid in 

 the eastern portion of its range. However, standard morphological 

 characters alone are not sufficient to discriminate between the two 

 forms. Rather, adult size and pattern differences, in conjunction with 

 the number of dorsal scale rows and ventral scales, best differentiate C. 

 h. horridus from C. h. atricaudatus. 



Two subspecies of the rattlesnake Crotalus horridus are thought to 

 occur in the eastern United States (Conant 1975): C. h. horridus, the 

 timber rattlesnake, and C. h. atricaudatus, the canebrake rattlesnake. 

 Gloyd (1940) defined the former as having 23 dorsal scale rows, a lower 

 number of ventral and caudal scales, an absent or faint postocular 

 stripe, and less brilliant contrast between the ground color and pattern. 

 He defined the latter as having 25 dorsal scale rows, a higher number 

 of ventral and caudal scales, larger size, and more brilliant markings. 

 The geographic range of the two races is shown in Figure 1. 



Crotalus horridus shows considerable variation in the western por- 

 tion of its range; C. h. atricaudatus is not known to occur in Oklahoma, 

 yet specimens of C. h. horridus from southeastern Oklahoma resemble 

 C. h. atricaudatus in color and pattern (Webb 1970). Anderson (1965) 

 found that populations of C. h. horridus from western Missouri pos- 

 sessed a reddish-brown middorsal stripe like that of C. h. atricaudatus 

 from southeastern Missouri. Gloyd (1940:186) also reported that "the 

 middorsal stripe of reddish brown, although very conspicuous in typical 

 (C. h. atricaudatus) specimens, is not a good definitive character because 

 of its common occurrence in specimens of C. h. horridus from western 

 localities." Smith (1961) regarded Illinois specimens from Jackson 



Brimleyana No. 12:57-74, September 1986 57 



