Evaluation of seed yield determining factors of winter oilseed rape (Brassica napus L.) by means of crop growth modelling

A simple crop growth model of winter oilseed rape 'LINTUL-BRASNAP', is briefly described and evaluated. The model simulates crop growth and development under optimum growth conditions and calculates 1) light reflection and absorption by the green canopy and flower layer, 2) total dry matter production based on light absorption and light-use efficiency by the green canopy 3) seed density (seeds m(-2)), 4) partitioning of dry matter to the seeds based on whether seed growth is limited by source or sink, 5) accumulation and mobilization of reserve carbohydrates and 6) crop phenological development. Two hypotheses on the duration of a critical period (CP) for seed set in a crop, (1) the duration of the flowering period and (2) a period of 350 degrees Cd since onset of flowering, were tested by means of the model. Variation in seed density was better simulated with the second hypothesis. When the model was tested with data used for parameterization (cultivar Jet Neuf) and with other data (cultivar Victor), highly significant correlations were found between simulated and experimental data of various crop characteristics such as total dry matter production at various stages of crop development, seed density, pod density and seed yield. The differences between sowing dates were simulated better than the differences between years per sowing date. LINTUL-BRASNAP appears to be a useful tool for identifying the main crop characteristics responsible for variation in winter oilseed rape yield such as leaf area development, density of the flower layer and subsequent light absorption during the critical period of seed set. It can be used to explore options for improving the seed yield of winter oilseed rape. (C) 1997 Elsevier Science B.V.