Simulation of shoot dry matter partitioning and yield prediction of processing tomato

Abstract: At present, Xinjiang produces about 90% of total processing tomato production in China, and which has become the most important and the largest producer of processing tomatoes in China. Light, heat, water, and soil are the vital components of special ecological factors, ensuring high quality, high yield, and high efficiency of processing tomatoes in Xinjiang. Compared with traditional furrow irrigation, one of the new irrigation strategies of crop production is drip irrigation. The introduction of drip irrigation in Xinjiang has provided the means to increase crop production and water use efficiency. When drip irrigation was used to grow processing tomatoes in Xinjiang, its yield and quality far exceeded the average level in China. However, no previous studies have examined the development and dry matter partitioning of Xinjiang's drip irrigated processing tomatoes. Simulation models of crop growth and production provide a widely accepted tool for assessing agricultural production opportunities in different agro-ecological zones in response to weather and management. Thus, the aim of the present study was to develop model for the growth and production of drip irrigated processing tomatoes in Xinjiang.Field experiments were conducted in three subsequent years in Shihezi, Xinjiang, China. The relationships between the partitioning indexes of organ dry matter and physiological development time (PDT) were systematically studied with the experiment of different sowing dates and varieties. And simulation models for shoot dry matter partitioning and yield in drip irrigated processing tomato were developed based on a partitioning index (PI) and a harvest index (HI) in which the PI of leaves and the HI were the functions of PDT, which were also altered by relative thermal effectiveness (RTE), relative photoperiod effectiveness (RPE), and intrinsic development factor (IDF). Model validation with three years of weather and independent crop growth data showed that the growth and yield of processing tomatoes are simulated satisfactorily. R2, root mean square error (RMSE) and relative estimation error (RE) of simulated and observed dry matter under four different growing stages (emergence to flowering, flowering to fruit-setting, fruit-setting to maturing, and maturing to ending date), total dry weight of whole growth period, stem dry weight, leaf dry weight, and fruit dry weight were 0.9754, 0.029t/hm2, 11.43%; 0.9936, 0.074t/hm2, 5.09%; 0.9840, 0.250t/hm2, 6.83%; 0.9713, 0.102t/hm2, 5.71%; 0.9940, 0.504t/hm2, 8.06%;0.9629, 0.332t/hm2, 14.62%; 0.9828, 0.200t/hm2, 10.84%; 0.9585, 0.549t/hm2, and 18.30%. The R2、RMSE, and RE between the predicted and the measured yield based on the 1:1 line were 0.9658, 5.806t/hm2, and 8.07%, respectively, which indicated that the model could predict well the dynamic accumulation of dry matter in different organs under diverse conditions of a drip irrigated processing tomato. We concluded that this model provide a tool to assess development, growth and production of processing tomatoes in various ecological zones in response to temperature and incoming radiation.