Fluorescence as a biological feedback signal for energy optimising spectra for greenhouse LED-illumination

Abstract: The introduction of LED-lighting for greenhouses has enabled new ways of controlling the light, both spectrum and intensity, to save energy and increase crop production. We have investigated a new method of automatically tuning the spectrum based on biological feedback.  More specifically, we have performed experiments in order to evaluate if steady-state chlorophyll a fluorescence, measured at canopy level, can be a candidate signal to be used as feedback for spectrum optimisation. The experimental results show a strong correlation between fluorescence and photosynthetic rate, both in the case when one LED colour is used at a time, and under various background light. This indicates that fluorescence can indeed be used as an indirect measure of growth. Furthermore, we have investigated different methods of evaluating how much the different diodes in the armature enhance the fluorescence and short term photosynthetic rate. At different background light, one LED colour at a time was changed, and the consequent change in fluorescence was measured. We refer to the fluorescence gain, for the specific LED colour, as the change in fluorescence divided by the change in photon flux, caused by the specific LED colour. When comparing the fluorescence gains for the different LED colours, the mutual relation was constant for a given plant species, independent of the background light spectra. Hence, if implementing the suggested controller for a predefined intensity, aiming to find the optimal spectra for maximised plant growth, the controller will maximise the intensity input to the most efficient LED colour and minimise the intensity to all other LEDs. When evaluating the performance of the different LED colours, as a function of different intensity levels, changes were noticed in the mutual relations of the fluorescence gains. For cucumber, we found that when passing the light level where the photosynthesis saturate, no further changes in the mutual relation of the fluorescence gains were noticed. This opens up for using our method, i.e. comparing the fluorescence gains caused by excitation of light of different colours, to determine the light level where light saturation occurs.