263 A Conceptual Dynamic Model for External Quality in Kalanchoe B. Eveleens-Clark 1 , S.M.P. Carvalho 1,2 and E. Heuvelink 2 1 Applied Plant Research 2 Wageningen University Glasshouse Horticulture Horticultural Production Chains Group Linneauslaan 2a Marijkeweg 22 1431 JV Aalsmeer 6709 PG Wageningen The Netherlands The Netherlands Keywords: Kalanchoe blossfeldiana, modelling, number of flower heads, plant density, plant height, temperature, timing of spacing, visual quality Abstract Modelling quality in pot plants is still a weak feature in crop modelling research. This work aims at building a conceptual dynamic model for external quality (plant height and number of flower heads) in Kalanchoe blossfeldiana. Four experiments were conducted to quantify the effects of temperature and spacing schedules at two different planting dates. Five constant temperatures (18, 20, 22, 24 and 26°C) and nine spacing schedules, resulting from combinations of spacing at a given leaf area index (LAI: 1.6, 2, 3, 4 and 4.4) into a final plant density (31.1, 34.6, 43.4, 52.0 and 55.6 plants m -2 ) were studied. Stem length was modelled as the number of internodes (same as number of leaf pairs) times the average internode length. Plant height results from the stem length plus the uppermost pedicel length. The spacing schedules did not affect plant height, but total aerial dry mass (TDM, g plant -1 ), leaf area and number of flower heads were increased at earlier and wider spacing. In contrast, temperature had a significant effect on the average internode length (internode elongation rate linearly increased with temperature) and on the generative length (optimum response to temperature). However, internode appearance rate was not affected by temperature. Number of flower heads showed a positive linear relationship with TDM. Therefore, by simulating TDM the number of flower heads can be predicted. TDM was simulated based on the intercepted photosynthetic active radiation and light use efficiency. The reaction time showed a quadratic response to temperature, with a maximum rate of progress to flower at 23°C (10 days delay at 18°C). These modules need to be validated to form the basis of a year-round decision support system for Kalanchoe. INTRODUCTION In The Netherlands Kalanchoe blossfeldiana is produced year-round and in 2002 it represented the first flowering pot plant in terms of area (44.5 ha) (CBS, 2002) and the second in terms of trade value in the Dutch auctions (37 million Euro), after Phalaenopsis (VBN, 2002). Since the production of pot plants is strongly influenced by market requirements a well-defined product must be available at a specific time. Therefore, to increase client satisfaction kalanchoe growers aim at increasing certainty of the delivery date and maintaining quality throughout the year (compact plants, high number of flower heads, and one to three open flowers at harvest). Although this crop is largely mechanised, at present the tools for an adequate planning and quality control are insufficient and need to be improved to face market demands. Image processing is a suitable tool for recording objective plant features, and this was successfully applied to monitor plant height in Ficus benjamina (Dijkshoorn-Dekker, 2002). The use of a growth and quality model, alongside the existing automatic grading and measuring systems for kalanchoe, would provide growers with a valuable tool for decision support. The model should be preferably dynamic, to give the possibility to predict at any moment of the cultivation whether the delivery date and quality targets can be achieved. Such a dynamic model would allow different batches to be subjected to different treatments to meet the final quality requirements. Proc. Int’l. WS on Models Plant Growth & Contr. Prod. Qual. in Hort. Prod. Eds. M. Fink and C. Feller Acta Hort. 654, ISHS 2004