Reaching natural growth: Sources of variation in plant traits between indoor and outdoor experiments

One of the main problems of indoor plant production that especially plant researchers are confronted with, is a clear difference between plants grown under indoor versus outdoor conditions. This reduce the comparability between indoor and outdoor experiments as well as the portability of findings from indoor experiments to real world conditions (Matsubara, 2018). Poorter et al., (2016) suggested multiple reasons why this may occur, with major effects coming from lower light quantities, higher plant density and shorter experiment durations in indoor compared to outdoor experiments. Other sources of variation have been pointed out, including age of the plants, leaf temperature, soil temperature, soil microorganism, lack of UV light and the light quality in indoor experiments (e.g. Hogewoning et al., 2010 b). In general, the artificial conditions in indoor growth facilities often produce higher specific leaf area, leaf nitrogen content and relative growth rate, as well as lower maximum photosynthesis, plant height and shoot dry weight, compared with outdoor experiments (Poorter et al., 2016).
Light, as one of of the principal determinants of plant growth and development, is consider an important source of deviation between indoor and outdoor conditions. For example, the effect of either light quantity or quantity has been well described in plants from different species, by Arnott and Mitchell (1982). To compensate a growth limitation in plants due a possible lack of light in greenhouse or indoor growth facilities, additional lighting is well stablished in agriculture, especially in areas at higher latitudes with year-round lower levels of natural sunlight (e.g. Grammans et al., 2018). Poorter et al., (2016) suggest that an important difference between indoor and outdoor climates for plant growth is a significant lower daily light integral (DLI) radiation in indoor facilities compared with outdoor conditions. Especially in combination with a lack of light variation along the day may lead to plant growth in indoor conditions that deviates considerable from field grown plants. It was not until the development and mass production of light emitting diodes (LED) that dynamic and specifics wavelengths changes as well as fast fluctuations of light intensity became possible to be used in indoor plant growth facilities. Previous attempts in plant biological research to recreate sun-like lighting with conventional light sources used very complex and fault-prone setups (e.g. Thiel et al., 1995) which were thus never widely used or considered for commercial plant production.
With the technical improvements in controlled environment capabilities, the use of indoor cultivation systems has increased worldwide. In indoor experiments several authors have demonstrated the positive effects of incorporating closer-to-natural environmental conditions in indoor facilities (e.g. Arve et al., 2017, Kaiser et al., 2020,), what can help without adding higher levels of complexity to reach either closer to natural plant growth under indoor conditions and thereby increase the quality of food production to taste, smell and look more natural, attributes that are desired by consumers (Arve et al., 2017) Due to the high degree of absorption of blue (B) and red (R) light by chlorophyll, and the higher electric efficiency of LED in these spectral ranges (Overdieck, 1978), these two wavelength ranges tend to be dominating in commercial LED lamp systems (Fujiwara and Sawada, 2006). Many studies have investigated the responses of plants to different B to R ratios.
These studies revealed that independent of the light intensity, a required minimum percentage of B is need for plant growth (e.g. Miao et al., 2016), and suggestions to reproduce near to natural plant growth by correctly adjusting the B:R ratio in LED lamps has been done (Hogewoning et al, 2010 a), however without directly comparing indoor grown plants with an outdoor control. In the vast majority of studies related to light quality effects on plants, either low light levels (Macebo et al., 2011; Hogewoning et al., 2010 a; Hernandez and Kubota, 2016; Kim et al., 2004; Schuerger et al., 1997) or much higher than natural red to far ratios have been use (e.g. Bae and Cho, 2008; Hogewoning et al., 2010 a; Hernandez and Kubota, 2016; Hernandez et al., 2016; Kim et al., 2004; Shengxin 2016; Zhen and van Iersel, 2017). However, interactions between light quantity and quality have been reported previously (Furuyama et al., 2014), and modifications of the light spectra, especially in the red to far ratio, has shown to induce more natural like plant growth (Hogewoning et al., 2010 b). This highlights the requirement of finding light spectral combinations in LED lighting that results in the most natural like plant growth in indoor facilities. One challenge is that different species might react differently do changes in the applied light spectrum. Tests for the effect of a light spectrum on plant performance should thus be done across different plant species (as in this thesis) in order to reveal general patterns as well as species-specific responses.
In principal, lamps with multi-channel LEDs enable the application of lighting that can mimic close to natural light quality and intensities changes during plant cultivation in indoor growth facilities (Bula et al., 1991). However, although the newest generation of LED lighting systems are equipped with 4 or more individually controllable spectral channels, growth facilities generally do not apply dynamic and natural changes in the light spectra on a standard base. The knowledge about the changes in light quality related to the solar elevation angle, latitude, as well as the presence or absence of clouds (e.g. Smit, 1982; Goldberg et al., 1977) has been so far reported mainly from an atmosphere-physical point of view, and has not been transferred to actual lighting systems used for plant culture in greenhouses or growth chambers.
Additionally, it has been shown that light quality effects on plants can interact with other environmental factors, like temperature (e.g. Chiang et al., 2018). This highlights the importance of understanding the role of the light quality variation on plant development, especially in order to correctly predict the effect of climate crisis on plants from indoor experiments.
Although it is known that the fluctuation of environmental factors has an effect on plant phenology and development, it is common practice to apply static environmental conditions in indoor experiments. Fixed day and night time climates may be oversimplified reductions of natural conditions and may lead to plant growth significantly deviating from field grown plants (Poorter et al., 2016). Especially, it is well-known that random and daily fluctuations of temperature and light, can affect plant performance in both positive and negative ways (e.g. Myster and Moe,1995; Kaiser et al., 2015; Kaiser et al., 2018). Several studies have measured the effect of light or temperature variations on plant performance under semi-controlled and controlled conditions, but again, simultaneous comparisons with outdoor grown plants are rare in the literature and normally just Arabidopsis thaliana has been used (e.g. Vialet-Chabrand et al., 2017; Annunziata et al., 2017; Annunziata et al., 2018). Nevertheless, from these studies it could be derived that changes in light quantity along the day may induce lower biomass but also higher maxium photosynthesis, especially per unit of leaf mass (Vialet-Chabrand et al., 2017), even though fast fluctuations in light intensity have been shown to reduce photosynthesis and productivity in the long term (Kaiser et al., 2018). Additionally, these studies have shown more evidence of the difference of plants grown under totally fixed climatic conditions compared with semi-controlled environments (i.e. greenhouses), highlighting the necessity of a better knowledge for a minimum requirement of environmental fluctuations for natural like growth in indoor experiments. To investigate more closer the potential causes for the differences in plant performance between indoor and outdoor plant experiments, and to enable more natural-like plant growth in indoor facilities, a joint project had been stablish between the University of Basel (Basel, Switzerland) and Heliospectra A.B. (Gothenburg, Sweden) within the research consortia PlantHUB (European industrial doctoral programme (EID) funded by the H2020 PROGRAMME Marie Curie Actions- People), coordinated and managed by the Zurich-Basel plant science center. The project consisted of 18 months of basic research at the University of Basel, followed by 18 months of applied research, software development and documentation at Heliospectra A.B. As a result of this collaboration, the present thesis aims to identify how climatic conditions (especially, light quality and fluctuation of light intensity, temperature and air humidity) need to be adjusted in growth chambers in order to reach the most natural like plant growth under indoor conditions. To avoid documentation about only species-specific reactions, several species from different functional plant types were always used. The work on this thesis was divided in 5 main modules that aimed to:
1) Understand and quantify the natural light quality changes along the day and along a whole season, assess the effect of cloudiness on the natural light spectrum, and correlated these findings to previous studies on light quality effects in trees (Chapter 1)
2) Investigate which light spectral combination of LED-lights can induce the most
natural-like growth in plants grown in indoor chambers with constant climatic conditions (Chapter 2)
3) Identify the minimal degree of environmental fluctuations (of light, temperature and air humidity) necessary to reach natural-like growth in indoor grown plants (Chapter 3) 4) Understand the effect of asynchrony environmental fluctuations in indoor growth chambers, were potential interaction and/or synergies may occur depending of the degree of variability of each environmental variable (Chapter 4) 5) Test possible applications of light fluctuations to improve crop quality and develop software applications for optimized light control of multi-wavelengths LED assimilation lamps (Chapter 5 and Appendix)