Studies on improvement of tomato productivity in a large-scale greenhouse: Prediction of tomato yield based on integrated solar radiation

As there are currently many large-scale production facilities that have contracts with the large retailing companies, accurate prediction of yield is necessary. The present study developed a method to predict tomato yield accurately using the data on the outside solar radiation. The present study was conducted in a Venlo-type greenhouse (29,568 square m) at Sera Farm Co., Ltd. in Sera-cho in Hiroshima prefecture. The cultivar used for this experiment was plum tomato. The sowing took place on July 18, the planting took place on August 30, and the harvesting started on October 9, 2002. The planting density was 2.5 plants msup(-2). As the results of the analysis of correlation between the weekly tomato yield and the integrated solar radiation for the period from October 7 to July 28 (43 weeks), the highest correlation (r=0.518) between the weekly tomato yield and the solar radiation integrated from seven to one weeks before the harvesting was observed. Further investigation by the same correlation analysis was conducted for the 25 weeks period from December 8 to May 26, during which time the effect of growing stages and air temperature were considered to be relatively small. The results showed the highest correlation (r=0.730) between the weekly tomato yield and the solar radiation integrated from eight to one weeks before the harvesting. The tomato yield occasionally needed to be adjusted at Sera Farm. Consequently, the correlation between the three-week moving average of tomato yield and the integrated solar radiation was calculated. The results showed the highest correlation was obtained for the period from eight to one weeks before the harvesting (r=0.860). This study therefore showed that it was possible to predict the tomato yield (y: kg.msup(-2).weeksup(-1)) using the following equation on the solar radiation integrated from eight to one weeks before the harvesting(x: MJ.msup(-2)): y=7.50 x 10 sup(-6)x + 0.148 (rsup(2)=0.740).