MATERIAL AND METHODS
One hundred thirty-two specimens were collected from Brazil and Northern Argentina (Figure 1 and Table I). Tissue collection and DNA extraction are described in Dergam (1996). Molecular analyses were done at the Department of Microbiology, Colorado State University, Fort Collins, CO, USA.
We used RAPD-PCR (Williams et al., 1990), with primers OPA09, OPA05, and OPAC13 (Operon Technologies). All samples were run simultaneously for each primer reaction to minimize repeatability problems (Black, 1993). We used a binary matrix ("1" and "0" for presence and absence of bands, respectively) for each specimen to record bands for cluster analysis. A measure of distance was calculated for each pairwise comparison, yielding a matching matrix (where similarity was assessed by the shared presence or absence of bands) with the RAPDPLOT program (Kambhampati et al., 1991; Black, 1996). This program calculates the fraction of matches between all individuals in the matrix:
M = NAB/NT
where NAB = total number of matches (shared presence or absence of bands) between individuals A and B, and NT is the total number of loci of the binary matrix.
Molecular similarity was analyzed by unweighted pair-group mean average (UPGMA) cluster analysis computed with the NEIGHBOR option of PHYLIP 3.5C (Felsenstein, 1993). A phenogram was drawn with the DRAWGRAM program from the same computer package. Because this similarity analysis is exclusively qualitative and does not deal with allelic frequencies or variances at each locus, we assumed that uneven sample sizes did not affect our results.
Answers to all your Biology Questions
