16 University of California Publications in Zoology 



2. Kluge and Farris (1969) handled variation within basic taxa by subdividing each 

 taxon into two taxa, each coded invariable for one of the alternative states of the variable 

 character. This method makes no assumptions about phylogenetic relationships other than 

 those involved in the delimitation of the basic taxa, and it should be satisfactory as long as 

 there are few variable characters. As the number of variable characters increases for a 

 given taxon, however, the number of new "taxa" created by subdivision increases 

 geometrically. Therefore, this method will be impractical if any of the basic taxa exhibit 

 more than a few variable characters. 



3. Information about relationships within a basic taxon can be used to assess the 

 ancestral condition of a character that varies within the taxon. The ancestral condition can 

 then be assigned to the taxon because it follows that the alternative condition bears 

 homoplastic rather than synapomorphic resemblance to similar conditions in other basic 

 taxa. An obvious drawback of this method is that it requires information about 

 relationships within basic taxa, information that is not always available. An advantage of 

 this method is that it does not increase the number of basic taxa. 



4. A fourth means of dealing with variation within basic taxa is to arbitrarily choose 

 one of the alternative conditions as the ancestral condition for the variable taxon. Although 

 it might intuitively seem that the condition that appears to be ancestral on morphological 

 grounds (i.e., the condition that is found in outgroups) is the ancestral state, this 

 conclusion has a hidden bias. Given that the variation represents homoplasy, assigning the 

 morphologically "ancestral" state to the taxon amounts to asserting that the homoplasy is a 

 convergence; assigning the morphologically "derived" state to the taxon amounts to 

 asserting that the homoplasy is a reversal. Furthermore, the most reasonable interpretation 

 of character transformation after construction of a cladogram or phylogenetic tree may 

 conflict with the original choice of an ancestral condition for a variable basic taxon. 

 Therefore, the level at which characters that vary within basic taxa are considered to be 

 synapomorphies should always be reevaluated after an initial analysis that does not take the 

 variation into consideration. 



I have already presented reasons for not using less inclusive taxa than the iguanine 

 genera as my basic taxa. In this study, character variation within basic taxa was high 

 enough that the method of Kluge and Farris (1969) would have been impractical. The third 

 method for handling variation within basic taxa could not be used since not enough is 

 known about relationships within iguanine genera to use this information to assess the 

 ancestral condition of characters that vary within them. Therefore, I have been forced to 

 use the last, and perhaps the least satisfactory, means of dealing with variation within basic 

 taxa. 



CONSTRUCTION OF BRANCHING DIAGRAMS 



The branching diagrams presented in this study are all intended to be cladograms rather 

 than phylogenetic trees (Nelson, cited in Wiley, 1979); in other words, no attempt is made 



