Predicting the weekly fluctuations in glasshouse tomato yields

There is considerable interest among tomato growers in systems to predict yields. However, while the relationships between aerial environment and cumulative crop yields are reasonably well understood, and a number of crop models have been developed, accurate predictions of the weekly patterns of crop yield are hard to achieve. A number of experiments have been conducted at HRI to investigate how light integral, temperature, and fruit load affect the pattern of yield so as to identify why yields fluctuate. The cyclical pattern of yield was largely caused by variation in the number of fruits picked per week rather than variation in the mean fruit size, which suggests that assimilate availability was not the primary cause. The variability in the number of fruits picked per week (even when all trusses had five fruits) indicates that the time from flowering to fruit maturation must have changed. A slight hastening of fruit maturation time tended to be associated with high yields, while low yields were often recorded when there was a delay in fruit maturation time. The changes in fruit development time were a result of fruits becoming more sensitive to temperature as they approached maturity. Therefore, a flush of ripe fruits followed a period of elevated temperature. However, rapid fruit ripening decreased the number of fruits nearing maturity on the plant, which resulted in depressed yield in subsequent weeks. This paper reports on progress in quantifying these relationships and in developing a model that can be used by growers, combined with better weather forecasting systems, for prediction of crop yields. Such a system should also provide information on how the aerial environment can be manipulated to modify the pattern of yield.