Numerical taxonomy of Italian populations of maize

A classification of maize populations extracted from the Italian germplasm collection has been undertaken. A good taxonomy of populations showing high adaptability to different environments is of great importance not only for a descriptive and phylogenetic purpose, but also because local populations are a very important reserve of genetical variability, useful in breeding programs.

105 populations were extracted from the germplasm collection (including more than 500 local populations), according to morphological criteria and climatic-geographical considerations.

The populations were grown during the 1975 season in a randomized block design with plots of eight plants. The following traits were considered: dimensional traits of the ear (ear weight, length, base and apex diameter, row number, kernel number per row, 50-kernel weight, thickness, length and width of kernels); morphological-vegetative traits of the plant (plant height and number of leaves--when growth is completed--, number of ears and total yield); physiological-adaptive traits (leaves emission rate--time from 4 to 8 leaves and from 8 to 12 leaves--, tassel flowering time and time from tassel flowering to silking).

All traits show significant differences between populations. The values measured on individual plants were analyzed by multivariate statistical methods (Canonical Analysis). The first three canonical variates (linear transformation of the original ones) explain 71% of the differences between populations. Particularly: the first canonical variate is connected with the vegetative "size" of the plant (growth behavior, cycle length and yield); the second variate is connected with the shape and weight of kernels; the third with the ear size. Mean values of populations were plotted according to the canonical axes (first and second variate) and the points with common 90% significance area in the three dimensions were clustered (see figure).

Figure.

According to increasing values of the first variate the clusters change from early maize patterns (quarantini, cinquantini, etc.) to late patterns; according to the second variate, from types with small ears and numerous rows (multi-rowed) to types with few rows and large kernels (8-rowed, etc.). Many populations (about 30%) don't cluster: they were regarded as transition forms or as points of adaptation to peculiar environments.

On the basis of physiological, vegetative and adaptive traits it was possible to subdivide groups of populations with very similar morphological features: this shows a true variability that, according to morphological traits alone, can be underestimated.

The next phase will imply a design of crosses between highly differentiated populations to evaluate the genetical distance: it will be possible to obtain a taxonomy of phylogenetical significance.

A. Camussi, M. Zambelli and E. Ottaviano


Please Note: Notes submitted to the Maize Genetics Cooperation Newsletter may be cited only with consent of the authors.

Return to the MNL 51 On-Line Index
Return to the Maize Newsletter Index
Return to the Maize Genome Database Page