Improving winter cereals for moisture-limiting environments: an international seminar

Abstract

Most of the agricultural productivity problems in the southern part of the Mediterranean region, like North Africa, Turkey, and the Middle East, are related to climate. The large variability of rainfall and thermal patterns of the region needed a climatic classification in order to define the most relevant groups of the area. According to this goal, a large FA0 dataset of meteorological stations included in a rectangular region ranging from 20°North/300West to 40°~orth/500~ast has been used for classification and ordination. A multivariate matrix n x m (where n = 227 is the number of the stations and m = 48 is the number of variables) was used. Variables were 12 values of monthly average rainfall, 12 values of monthly average air temperature, 12 values of monthly average minimum temperature, and 12 values of monthly estimated potential evapotranspiration. Hierarchical cluster analysis indicated a good partition with nine main classes. For each generated class a complete average climatic pattern, the related variance, and geographical distribution are given. An ordination method (Principal Component Analysis) finally allowed to better interpret the class meaning and the main relationships among stations. The most relevant climatic aspects involved in agricultural production are outlined and discussed.