Figure 6.

NJ tree of RISP haplotypes from four focal regions for T. californicus. Tree is unrooted. Branches that were also supported by consensus trees of partial parsimony analyses of these RISP haplotypes are indicated by an *. We were unable to complete a parsimony analysis for the entire dataset but could complete analyses for pruned datasets to examine relationships between regions. Secondarily, the relationships of haplotypes within each region were separately analyzed using a parsimony method and compared to the relationships obtained from the NJ analysis. Intragenic recombination and reticulation of sequences within regions could make parsimony analyses difficult by yielding numerous equally parsimonious trees.