Phylogenetic Systematics of I guanine Lizards 1 5 



taxa of this analysis. To ignore these characters would be to throw away information; for 

 although they contain ambiguities, they also suggest relationships among certain taxa. One 

 very important reason for retaining characters that vary within basic taxa is that these are the 

 only characters that can reveal the paraphyletic status of a basic taxon. I have thus retained 

 certain variable characters in my analysis. However, in cases where the variation within 

 basic taxa was so great that it obscured any pattern of variation among them, the character 

 was eliminated. Because characters form a continuum from those that vary within all basic 

 taxa to those that vary only among them, the decision as to which characters were too 

 variable to be useful at this level of analysis was necessarily arbitrary. 



Variation within basic taxa is not itself a problem. For example, variation within basic 

 taxa involving different characters from those that vary among them is simply irrelevant to 

 an analysis of the relationships among these taxa. Even when variation within and among 

 basic taxa involves the same characters , the situation is not necessarily problematic. If the 

 variation within basic taxa characterizes all recognizable subgroups that are units of 

 phylogenetic ancestry and descent (e.g., evolutionary species), or if the basic taxa 

 themselves are such units, then the variable presence of a derived character in two or more 

 basic taxa may be attributable to a polymorphic ancestral population. In contrast, when 

 variation within basic taxa occurs among one or more of their monophyletic subgroups, 

 this variation represents homoplasy. Because homoplastic variation within a basic taxon 

 makes an assessment of its ancestral condition ambiguous, such variation is problematic. 

 The conclusion that variation within a taxon represents homoplastic similarity with a 

 condition occurring in another taxon rests on the assumption that the first taxon is 

 monophyletic. This is a critical point. The variable presence of a derived character in a 

 paraphyletic taxon does not necessarily require homoplasy. In fact, the very characters that 

 give evidence of a taxon's paraphyletic status necessarily vary within that taxon. For this 

 reason, the monophyletic status of basic taxa should be continually reevaluated in light of 

 variation within them. 



Variation within basic taxa that is assumed to represent homoplasy can be handled in 

 various ways, none of which is without drawbacks. The most appropriate method for a 

 particular case will depend on the amount and nature of the variation, as well as on the 

 assumptions that the investigator is reasonably able to make. I discuss below four methods 

 for handling variation within basic taxa. 



1. One means of handling variation within basic taxa is to abandon these taxa in favor 

 of less inclusive basic taxa.: for example, one might use species instead of genera. 

 Unfortunately, this practice carries no guarantee that variation will not exist within the new 

 basic taxa. A practical disadvantage of this method is that it necessarily increases the 

 number of basic taxa. Furthermore, the number of specimens may be distributed in such a 

 way that choosing the more inclusive groups as basic taxa gives a more accurate picture of 

 the range of variation. A lack of variation within less inclusive taxa may result simply from 

 small samples. 



