Effects of growth conditions on external quality of cut chrysanthemum; analysis and simulation

For many years the emphasis in floricultural research laid with quantity rather than quality. Nowadays, since the prices are often determined on the basis of visual quality aspects, the so-called external quality, chrysanthemum growers aim to provide a high and constant product quality throughout the year. The external quality of cut chrysanthemum is usually evaluated in terms of stem and leaf morphology and flower characteristics. The priority within the external quality attributes depends on the particular market for the product.Chrysanthemum cultivation is one of the most controlled and intensive crop production systems in horticulture. This quantitative short-day plant can only be cultivated year-round in greenhouses by controlling several growth conditions. However, many combinations of these conditions are possible, according to the growth strategy being employed. To produce year-round high quality chrysanthemum is a constant challenging problem for the grower, as the seasonal variations in daily light integral will produce large seasonal fluctuations in yield and quality. Therefore, in order to choose the optimal strategy adjusted to the planting week, it is necessary to know how the growth conditions influence plant quality. Thus, the factors involved in chrysanthemum external quality need to be carefully analysed and effectively combined to achieve the production of flowers with the maximum ornamental value year-round, while maintaining a high yield and an acceptable low energy input. Considering the complexity of cut chrysanthemum production, with its many options for control and its range of product quality attributes, management of such a system would be expected to highly benefit from the use of simulation models. For instance, the explanatory models are a valuable tool to integrate knowledge and to assist in the decision support. The development of such models for product quality is still a weak feature in crop modelling research, since priority has been given to simulation of productivity. To develop an explanatory model for the external quality of cut chrysanthemum, detailed knowledge about its growth and morphological development is needed.The main aim of the present study was to quantify and understand the effects of the aboveground growth conditions on the external quality of cut chrysanthemum at harvest. Special attention has been paid to the integration of this knowledge and its incorporation into an explanatory model to predict the main external quality aspects of cut chrysanthemum. The focus was on the effects of the climate conditions (temperature, light intensity and CO 2 concentration) and cultivation practices (duration of the long-day period and plant density) on plant height (stem length), number of flowers and flower size.Chapter 2 presents an overview of the growth conditions involved in the different chrysanthemum external quality aspects, and identifies the gaps in literature. A synthesis of the available models that have been built to predict some external quality attributes of chrysanthemum is also given.The DIF concept states that internode length is dependent upon the DIFference between day (DT) and night (NT) temperature, and is independent of the mean 24 h temperature. This controversial proposition was investigated by means of an experiment described in Chapter 3.1. Chrysanthemum 'Reagan Improved' was grown in growth chambers at all 16 combinations of four DT and four NT (16, 20, 24 and 28 ºC) with a 12 h daylength. The length of internode 10, the number of internodes and the stem length were measured periodically. The experiment ended when internode 10 had reached its final length in all temperature combinations employed (27 days). A significant positive linear relationship between DIF and the length of the fully developed internodes was observed over the range of temperatures studied (16-28 ºC). It was also found that internode lengths recorded in early stages of development do not bear a close relationship to the final internode lengths, which explained contradictions in literature. In addition to being dependent on the developmental stage of the internodes, the effectiveness of DIF was related to the range of temperatures. It was shown that the DIF concept is valid only within a temperature range where the effects of DT and NT are equal in magnitude and opposite in sign (18-24 °C). Therefore, it was concluded that the response of internode length to temperature is strongly related to DIF, but this response is simply the result of independent and opposite effects of DT and NT. Internode appearance rate, as well as stem length formed during the experiment, showed an optimum response to DT.Chapter 3.2 is the description of an in depth study on the sensitivity of several flower characteristics to temperature, with the aim of obtaining a better understanding of the underlying physiological processes of flower initiation and development. An attempt has been made to analyse the effects of temperature, applied at different phases of the cultivation period, on each of the studied flower characteristic. Plants were grown in glasshouse compartments at two constant temperatures (17 and 21 °C), and in growth chambers at 32 temperature combinations (from 15 to 24 °C). In the growth chamber experiment the temperature treatments were based upon a division of the cultivation period into three consecutive phases: from planting until the end of the long-day (LD) period (phase I; 18 and 24 °C), from the start of the short-day (SD) period until the visible terminal flower bud (phase II; 15, 18, 21 and 24 °C), and finally from the visible terminal flower bud until harvest (phase III; 15, 18, 21 and 24 °C). Of the characteristics investigated only the flower position within the plant was independent of temperature. The number of flowers and flower buds per plant (NoF), individual flower size and colour (pink) were strongly affected by temperature. It is shown that the temperature effect was largely dependent on the cultivation phase and on the flower characteristic itself. In general, flower characteristics were less influenced by temperature applied during the LD period, compared to the SD period. A higher temperature increased NoF, mainly by increasing the number of flower buds. NoF was affected positively by temperature mainly during phase III, whereas individual flower size increased with temperature during phase II but decreased with temperature during phase III. Lower temperatures during phase III significantly enhanced flower colour intensity. It was concluded that it is not possible to ascribe to each phase of the cultivation a common optimum temperature for all the flower quality aspects. Hence, to define the most suitable temperature in each cultivation phase it is necessary to decide which quality attribute is to be maximised.The effects of the assimilate availability on the NoF, individual flower size and plant height is described in Chapter 4.1. Seven greenhouse experiments were conducted in different seasons using the cultivar 'Reagan Improved' (spray type). One extra experiment was carried out to extend this study to two other cultivars ('Goldy' and 'Lupo': 'santini' type), focusing on their response to plant density. Assimilate availability, measured as total plant dry mass (TDM, g plant -1), increased with higher light intensity, higher CO 2 concentration, lower plant density or longer duration of the LD period. In contrast, variation in the growth conditions produced hardly any effect on flower mass ratio (FMR), and only an increased duration of the LD period had a negative linear effect on the partitioning towards the flowers. The season also had an effect on chrysanthemum FMR: when planted in September (lowest light levels during the SD period), FMR was reduced compared to the other seasons. It is concluded that within a wide range of growth conditions chrysanthemum invests the additional assimilates, diverted to the generative organs, in increasing NoF rather than in increasing flower size. Individual flower size was only affected by assimilate availability when average daily incident photosynthetically active radiation during the SD period was lower than 7.5 mol m -2d -1, resulting in lighter and smaller flowers. When incident photosynthetically active radiation (PAR) during the SD period was higher than this threshold value, a constant flower size was observed for the fully open flowers (0.21 ± 0.10 g plant -1and 25 ± 2 cm 2plant -1). Excluding the positive linear effect of the duration of LD period, assimilate availability had no relevant influence on plant height (< 10 % increase). Irrespective of the growth conditions and season, a positive linear relationship between NoF and TDM was observed (NoF = 1.938TDM - 2.34; R 2= 0.90). The parameters of this relationship are cultivar-specific. The generic nature of these results is discussed is this chapter. The functional relationships developed for predicting NoF and flower size were incorporated as 'modules' in a photosynthesis-driven growth model for cut chrysanthemum (Chapter 5.2).The influence of assimilate availability on flower size can also be tested by manipulating sink-source ratio. This allows the estimation of the potential flower size, which is defined as the size reached under conditions of non-limiting assimilate availability. In Chapter 4.2 sink-source ratio was manipulated by flower bud removal (leaving one, two or four flowers and a control), by the presence or absence of axillary shoots, and by varying the light intensity. To investigate whether flower size is dependent on flower position within the stem, the apical terminal flower, the apical lateral flowers (from first order axillary shoots) and the first flower locate in a second order axillary shoot were compared. The results indicated that in treatments where a limit on number of flowers was imposed, individual flower dry mass and area increased significantly under conditions of lower competition for assimilates (for example, by decreasing sink-source ratio by either leaving fewer flowers per plant, removing axillary shoots or using supplementary assimilation light). The effect of flower position on flower size, in both the disbudded and control plants, was found to be only important when comparing flowers located on the first order axillary shoots with flowers on the second order axillary shoots, the latter being 40 % smaller than the former. Monoflower plants without side shoots represented the potential flower size, and their flower was up to 2.4 times as heavier and 76 % as larger in area as the control flower in 'Reagan Improved'. The 'santini' cultivars also produced their maximum flower size on the monoflower plants, but the increase in size relative to the control plants was cultivar specific. Higher leaf starch content and lower specific leaf area (thicker leaves) were observed in the monoflower treatments, reflecting an abundance of assimilates. Plant dry mass was only reduced at the lowest sink strength treatment (monoflower plants without axillary side shoots), whereas FMR showed a saturation response to the number of flowers per plant with a maximum value of 0.22.The data obtained in the previous chapters were further explored to model and validate some external quality attributes. In Chapter 5.1 a process-based model was developed to describe internode elongation in time as a function of temperature. This model was calibrated with the data from Chapter 3.1, and it was built based on three plausible physiological processes occurring in chrysanthemum elongation: (1) the accumulation of elongation requirements during the day, (2) elongation during the night using the accumulated elongation requirements, and (3) the limitation of the internode length due to low turgor pressure unable to counteract cell wall elasticity. Simulated and measured internode length showed a good agreement ( R 2= 0.91). The presented model may be extended to include variable light conditions and other plant species that show elongation control by DIF.In Chapter 5.2 a case study is presented on the interactive effects of duration of the LD period (2, 9 and 16 days) and plant density (48, 64 and 80 plants m -2) on several external quality aspects. An existing photosynthesis-driven crop growth model for cut chrysanthemum (Lee et al. , 2002) was validated and used to simulate total dry mass for the nine treatments. The possibility of a trade-off between the cultivation measures was analysed, while aiming to maintain a certain quality at harvest. It was concluded that a similar total plant fresh mass could be obtained using several combinations of plant density and number of LDs without affecting either NoF or individual flower size. This trade-off is, however, very dependent on the planting date of the crop, which emphasises the need for a crop simulation model as a decision support tool. Furthermore, special attention should be paid to plant height when choosing a combination of the cultivation measures studied, since this is strongly and positively influenced by the duration of the LD period. The modules developed in Chapter 4.1, for number of flowers and flower size were validated and the measured values were accurately predicted.The main achievements and limitations of this study are discussed in Chapter 6, and suggestions for future research are presented.