DRY-MATTER PARTITIONING IN TOMATO - VALIDATION OF A DYNAMIC SIMULATION-MODEL

A model for dynamic simulation of dry matter distribution between repr oductive and vegetative plant parts and the distribution among individ ual fruit trusses in glasshouse tomato, is validated. The model is par t of the crop growth model TOMSIM and is based on the hypothesis that dry matter distribution is regulated by the sink strengths of the plan t organs, quantified by their potential growth rates, i.e. the growth rates at non-limiting assimilate supply. Within the plant, individual fruit trusses are distinguished and sink strength of a truss is descri bed as a function of its development stage. Truss development rate is a function of temperature only. The same potential growth curve, propo rtional to the number of fruits per truss, is adopted for all trusses. In a simple version of the model, vegetative plant parts are lumped t ogether as one sink with a constant sink strength. In a more detailed version, vegetative sink strength is calculated as the sum of sink str engths of vegetative units (three leaves and stem internodes between t wo trusses). The model was validated for six glasshouse experiments, c overing effects of planting date, plant density, number of fruits per truss (pruning at anthesis), truss removal (every second truss removed at anthesis), single- and double-shoot plants and a temperature exper iment conducted in climate rooms at 17, 20 or 23 degrees C. Daily incr ease in aboveground dry weight, average daily temperatures and number of set fruits per truss were inputs to the model. Both the simple and the more detailed model showed good agreement between measured and sim ulated fraction of dry matter partitioned into the fruits over lime. F or the simple version of the model, the slope of the lines relating si mulated to measured fraction partitioned into the fruits (16 data sets ), varied between 0.92 and 1.11, on average it was 1.04, implying 4% o ver-estimation for this fraction. For the detailed model these numbers were slightly better: 0.89, 1.08 and 1.01, respectively. The temperat ure experiment revealed no important direct influence of temperature o n the ratio between generative and vegetative sink strength. Simulated truss growth curves showed reasonable agreement with the measurements , although both models over-estimated (17% on average) final dry weigh t of the lower trusses (truss 1-3) on a plant. Modelling dry matter pa rtitioning based on sink strengths of organs is promising, as it is a general, dynamic and flexible approach, showing good agreement between measurements and simulation for a range of conditions. Applicability of the model is, however, still limited as long as the number of fruit s per truss (flower and/or fruit abortion) is not simulated, as this i s a major feedback mechanism in plant growth. (C) 1996 Annals of Botan y Company