Do chiasmata disappear? An examination of whether closely spaced chiasmata are liable to reduction or loss.
The questions of whether closely spaced crossovers could be misidentified as single chiasmata or could cancel out each other and whether exchange could occur without chiasma formation were examined in Locusta migratoria. Monochiasmate bivalents that showed differential sister chromatid staining following bromodeoxyuridine incorporation were screened for their patterns of label distribution. Half of the chiasmata were associated with an exchange between dark and lightly stained chromatids, as expected if recombination involved any two non-sister chromatids chosen at random. Two variant types of label distribution were also seen: approximately 10% of all monochiasmate bivalents had an anomalous distribution of dark and light chromatids around the chiasma, and in three of the 1365 bivalents screened a second type of anomalous pattern was observed for the first time, in which all four chromatids had a label exchange at the chiasma. The observed incidence of the latter was considerably less than expected if they originated through a four-strand double crossover with closely positioned exchanges. Analysis of label distribution patterns in monochiasmate bivalents did not produce evidence of the other configurations expected if two closely spaced exchanges could be misidentified as single chiasmata or could cancel one another out so as not to form a chiasma. We conclude that analysis of chiasma frequencies and distributions offers an accurate means of assessing recombination in organisms with favourable cytology.