Development of a Functional-Structural Plant Model to Simulate the Distribution of Transpiration within a Greenhouse-Grown Tomato Plant.

MSc-thesis abstract (submitted 7 March 2016):
Transpiration is an important physiological process in plants. It is functional for the plant in terms of its cooling effect as well as its effect on element distribution. In addition, it has implications for the energy balance of the greenhouse. Transpiration is affected by a large number of crop characteristics, environmental factors and management-related aspects. An effective and non-destructive way to quantify the transpiration rate of greenhouse-grown plants under a wide range of conditions, is to make use of models.

Most transpiration models are one dimensional. These models are a much simplified version of reality, since they generally assume that the whole canopy is subjected to the same environmental conditions.
A more useful type of model would therefore be a functional-structural plant model (FSPM), which is a three dimensional model that performs physiological calculations on organ level and takes into account the spatial distribution of the individual plant organs. This allows for example to determine the distribution of transpiration rate over the individual leaves of a plant, which can be useful in determining the distribution of elements (calcium) and hormones (cytokinins).
Currently, most FSPMs have a main focus related to light interception and photosynthesis, whereas calculations regarding transpiration are much simplified. Therefore, FSPMs require improvement with regard to their transpiration module.

Aim: The aim of this research was to improve the transpiration module of an existing static FSPM, in order to be able to simulate the vertical distribution of transpiration rate within a greenhouse-grown tomato plant.

Approach: The transpiration module of the existing FSPM was analyzed and improved using information and strengths of other models that focus on transpiration, such that the model gives a more accurate prediction of transpiration rate (mmol m-2 s-1) of each leaf. Main focus was on improvement of the energy balance of individual leaves, in particular in terms of both shortwave and longwave radiation fluxes. Distinction was made between PAR and NIR absorption by the leaves and TIR exchange with various objects in the greenhouse. After development of the model, simulations were carried out to determine the vertical distribution of transpiration rate through a tomato plant. In addition to that, the effect of row orientation was examined, and a sensitivity analysis was performed to examine the extent to which various environmental and plant-specific factors affected the model outcome.

Results: The simulation study showed that, according to expectations, the transpiration rate was not homogeneously distributed over the individual leaves of a tomato plant. Instead, an exponential distribution in relation to plant height was found, a pattern that could be explained by the underlying factors absorbed radiation and leaf temperature. In addition to that, a higher transpiration rate was found in an east-west oriented crop than a north-south oriented crop, especially in the leaves positioned low in the canopy. The sensitivity analysis showed that light intensity and relative humidity were the most important environmental factors determining transpiration rate. Especially relative humidity also changed the distribution of transpiration over the individual leaves, as the effect of a change in humidity was increasingly strong towards the bottom of the plant. Also canopy temperature had substantial effect on the model outcome, particularly with regard to the lower leaf layers. The ratio between diffuse and direct light, wind speed and object temperatures only had marginal effect on the model outcome. With regard to plant-specific parameters, especially leaf area was important for the transpiration rate. This factor changed the distribution of transpiration as well, as the effect was less strong in the top leaves of the plants. Also vapor pressure conductance had considerable effect on transpiration rate, whereas boundary layer conductance and leaf angle barely affected the model outcome.

Conclusions and recommendations: This project resulted in an FSPM that allows a more accurate calculation of transpiration rates of individual leaves in a greenhouse-grown tomato plant. Although some interesting results were found with the simulation study, the model still needs further improvement. The main limitation of the current model is that it does not respond correctly to direct light. As a result of shadowing by the greenhouse walls, the light interception and transpiration rate of part of the plants is underestimated, resulting in high variations between plants. To increase the accuracy of the model, it is recommended to improve the model with regard to the response to direct light. In addition to that, the model needs to be validated with experimentally acquired data.