Statistical Problems in Karyotype Analysis

N a karyotype analysis it is often difficult to identify the single I chromosomes and their arm relations in a plate and consequently to reproduce the correct idiogram.The following idiogram of Lurix decidua (Mill.)chromosomes (Fig. l ) , will serve as a model for our further treatment of this problem (cf.SIMAK, 1962).For the sake of simplicity we shall assume that the material is haploid.Such material is sufficiently available for larch and other conifers in general in the young endosperm tissues.The usual method of karyotype analyses is that one measures the length of the chromosomes and their arms (in photographs or drawings etc.) in several plates and from the averages arranges an idiogram; for example, according to chromosome length.If secondary constrictions, etc., occur, these are naturally added to the chromosome concerned.A chromosome which possesses a satellite (e.g.chromosome VII) is usually easy to identify in a plate.If, however, the length of the chromosomes and centromere-index are the only criteria for identification, difficulties soon arise (e.g.chromosomes IX, X, XI and XII).The variation in length of the chromosomes caused by artifacts is often much greater than the genetic variation.This can lead to the resplt that two chromosomes of approximately the same length but with similar centromere-indexes may be impossible to distinguish from one another.For example, take the case of chromosome XII.It is questionable whether the shortest heterobrachial chromosome in every plate is really represented by this chromosome.It may happen that chromosome XI, through strong contraction, etc., appears in several cases as the "shortest" in the plate.A further demonstration of these erroneous measurements is provided by chromosomes I1 and 111.Both of these chromosomes can easily be differentiated from the other chromosomes