Distinguishing species

Abstract

Given two organisms, how can one distinguish whether they belong to the same species or not? This might be straightforward for two divergent organisms, but can be extremely difficult and laborious for closely related ones. A molecular marker giving a clear distinction would therefore be of immense benefit. The internal transcribed spacer 2 (ITS2) has been widely used for low-level phylogenetic analyses. Case studies revealed that a compensatory base change (CBC) in the helix II or helix III ITS2 secondary structure between two organisms correlated with sexual incompatibility. We analyzed more than 1300 closely related species to test whether this correlation is generally applicable. In 93%, where a CBC was found between organisms classified within the same genus, they belong to different species. Thus, a CBC in an ITS2 sequence-structure alignment is a sufficient condition to distinguish even closely related species.

FIGURE 1.

Correlation of the expected number of mutations (Jukes Cantor distance) to CBCs. This figure correlates the evolutionary distances of pairwise ITS2 sequences from genus alignments to CBC frequencies. The straight line shows the robust linear regression, while the curve shows the local robust regression (lowess). While the curve and the line are fairly similar, we claim there is a real linear relation between the evolutionary distance and the CBCs mean for genus alignments. The mean differences of the lowess function and the regression line is 0 (P <0.001). The residuals reflect that the lowess function is consistently below (0, 0.5) and above (0.5, 1) the regression line.