Some relationships between plant population, yield components and grain yield of wheat in a Mediterranean environment

Abstract

Eight spring bread wheat cultivars (Triticum aestivum L.), differing widely in their nominal yield component characteristics, were tested under rain-fed conditions for three years at sowing densities ranging from 50 to 800 seeds m-2. The objectives of the experiments were to estimate the relationship between grain yield and particular yield components, the expression of plant type (yield components) in relation to plant density, and the plant population x cultivar interaction for grain yield over a range of seasons in a given environment. The 'optimum' plant population (at maximum grain yield) varied over 30-220 plants m-2, depending on season and cultivar. In general, variation in the 'optimum' population was greater between seasons for a given cultivar than between cultivars within seasons. The relationship between grain yield and yield components was examined at the 'optimum' population rather than at an arbitrary population at which grain yield may have been suboptimal for some cultivars or seasons. Grain yields at the optimum populations for the various cultivar x season combinations were positively related to culms m-2, spikes m-2 and seeds m-2. They were not clearly related to culm mortality (%). When averaged across seasons, cultivar grain yields were positively related to harvest index, but the general relationship was not so clear when seasons and cultivars were examined individually. Spike size (seeds spike-I or spike weight) and seed size were also not clearly related to grain yield at the 'optimum' population, and it was thus postulated that the production and survival of large numbers of culms, which in turn led to large numbers of seeds per unit area, were the source of large grain yields. Some interactions were found between yield components and plant population for some cultivars that could have implications for plant breeders selecting at low plant densities. The implications for crop ideotypes of the individual plant characters at the 'optimum' population are also discussed. Interactions between cultivars and plant populations implied that some cultivars required different populations to achieve maximum yields in some seasons. There was a tendency for larger yields to be achieved from cultivar x season combinations where the optimum population was larger, which suggested that commercial seed rates should be re-examined when changes to plant types or yield levels are made.