The effects of light on plant growth can exhibit substantial variation across plant species, cultivars, developmental stages, and environmental factors, including plant nutrition, water availability, light intensity, and spectra. This variance stems from the diverse photosynthetic capacities inherent to different species, as well as the intricate interplay of multiple environmental factors (Boyer, 1970; Penning de Vries et al., 1989; Rosati and Dejong, 2003; Zhao et al., 2003; Weaver and Van Iersel, 2020). For instance, cultivating crops under identical DLI values but with lower light levels and extended photoperiods can lead to equivalent or even heightened daily growth. This is attributed to the enhanced efficiency of photosynthesis at both the leaf level (Weaver and Van Iersel, 2020) and the crop level (Hui et al., 2001). Furthermore, light wavelengths outside the Photosynthetically Active Radiation (PAR) range, 400 – 700 nm, can significantly influence growth and development. Far-red light (700 to 800 nm), for instance, promotes leaf expansion and photosynthesis in the 700-750 nm range (Zhen et al., 2021). In light of these findings, an extended PAR range (ePAR) from 400 to 750 nm has been proposed to account for photosynthesis enhancements (Zhen et al., 2021). Therefore, due to the complex interplay of these factors, comparing and adopting light levels across different situations is challenging and should ideally be confined to the specific conditions under evaluation. Nevertheless, for general reference, we provide a comparison with other cannabis studies, crops, and indoor systems employing the conventional DLI range (DLI 400-700nm) while offering the extended range values as a reference ( Table 1 ).

Within the realm of cannabis cultivation, this study, as well as prior investigations employing exclusive LED lighting sources (Moher et al., 2021; Rodriguez-Morrison et al., 2021), provides compelling evidence to classify cannabis among the most elevated lighting requiring crops. For comparison, optimal lettuce production is associated with a maximum of 17 mol m^-2 d^-1 for marketable leaf growth, with higher light levels potentially leading to increased biomass but also an elevated risk of leaf tip burn (Both, 2002). High-light greenhouse crops like tomatoes and cucumbers require at least 30 mol m^-2 d^-1 for optimal yields and fruit quality (Dorais et al., 2017), although significant research into higher light levels may not have been pursued due to economic feasibility constraints for production (van Iersel and Gianino, 2017). In corn breeding, supplemental greenhouse lighting is recommended to reach up to 750 µmol m^-2 s^-1 or ≥38 mol m^-2 d^-1, depending on the photoperiod, while corn research in chamber conditions employs as much as 1000 µmol m^-2 s^-1 or 58 mol m^-2 d^-1 (Eddy and Hahn, 2010). Remarkably, indoor farming studies have shown that cannabis crops can utilize 82 mol m^-2 d^-1 for shoot growth (Moher et al., 2021) and 78 mol m^-2 d^-1 for flower production (Rodriguez-Morrison et al., 2021).

Despite growth increments in the vegetative stage of cannabis under high light, the light-to-biomass conversions in Moher et al. (2021) were lower than in our study. For instance, the biomass yield in the Moher et al. (2021) study was 246 g m^-2 with 52 mol m^-2 d^-1, and it increased to 271 g m^-2 with 82 mol m^-2 d^-1. In contrast, our study yielded an average of 285 g m^-2 with 52 mol m^-2 d^-1 (eDLI: 53; Table 1 ) and showed no indications of light saturation ( Figure 2A ); importantly, both studies shared a consistent 21-day timeframe, facilitating a valid DLI-based comparison. It is plausible that the improved greenhouse study result in our research could be attributed to differences in plant genetics (Vanhove et al., 2012; Suchoff et al., 2021), light spectra (Lalge et al., 2017; Zhen et al., 2021), temperature conditions (Chandra et al., 2008, 2011), and/or the effects of water and fertilizer supply. Nevertheless, it is evident that further research into cannabis cultivation and production conditions is necessary to gain a deeper understanding and achieve consistent maximum yields and product quality.

In this investigation, plants were uniformly spaced at intervals of 30.5 cm, resulting in a plant density of 10.8 plants per square meter. In crop production, manipulating plant densities is a strategy to optimize yields and cost-efficiency by capitalizing on resource utilization and enhancing the efficacy of lighting, cooling, and heating per unit area. For instance, elevating plant density leads to an early expanded leaf canopy and improved light interception capacity. Moreover, an optimal leaf area is key to increasing crop photosynthesis and water use efficiency (Hui et al., 2001). However, an excessive plant density does not necessarily translate to increased yields, as seen in cannabis cultivation (Vanhove et al., 2012). The Leaf Area Index (LAI) assesses light interception and quantifies the total leaf area relative to the ground area (m² m^-2). Previous research across multiple plant species has consistently reported that an LAI in the range of 3 to 4 intercepts approximately 95% of the incident light, whereas an LAI of 2 captures roughly 80% (Shibles and Weber, 1965; De Visser et al., 2007; Brodrick et al., 2013; Tabarzad et al., 2016). Given that an LAI of 3 to 4 represents an optimal threshold for light capture, and values exceeding four may not confer any significant advantages in terms of light interception (De Visser et al., 2007; Brodrick et al., 2013; Tabarzad et al., 2016) and growth (Shibles and Weber, 1965), growers are likely to enhance production efficiency by adjusting plant spacing when the LAI surpasses four. Consequently, for a duration exceeding 21 days and high-light conditions like those presented in this study (leading to an LAI of 4.4), plant density could be reduced or branches trimmed while still maintaining a critical light interception of around 95%. Conversely, for crops subjected to lower light levels ( Figures 1 , 2B ), pruning for harvesting cuttings or conducting formative pruning for flower production may have a negative impact on the LAI and plant growth.

Responses of plant height in conjunction with internode extension are key to understanding whether the plant stretches mainly because of growth or to compete for more light, triggering shade avoidance syndrome or SAS (Kump, 2020).

In this study, the combination of light quantity and quality ( Tables 1 , 2 ), did not lead to significant changes in plant height across various light treatments ( Figures 1 , 3 ). Still, the light level and spectrum affect SAS responses (Kump, 2020), and perhaps they limited the observed differences in plant height in this study. For example, the lower light condition treatments may have contributed to some stem and internode extensions and vice versa. Furthermore, lower red to far-red ratios have increased stem length in several plant species (Demotes-Mainard et al., 2016). For example, in cucumber and lettuce, stem elongation was only significant when far-red light exceeded 9% at a DLI of 11.5 mol m^-2 d^-1, while the effects were marginal under 29 mol m^-2 d^-1 (Kusuma and Bugbee, 2023). In the present cannabis study, the daily average far-red ranged from 8% to 3% across DLIs of 18 to 52 mol m^-2 d^-1. Therefore, significant light spectrum effects would not be expected.

The overall findings and data suggest that the specific light conditions implemented (as outlined in Tables 1 and 2 ) had only a minor influence on plant height and internode length. These characteristics would be more substantially influenced by plant genetics and developmental factors than light intensity for cannabis.

Cannabis branches are fundamental in enhancing cutting production and supporting flowers. To assess the impact of light quantity on branch development, the total number of branches was measured and categorized into two groups: primary branches, which develop directly on the main stem, and secondary branches, which sprout from primary branches. The formation, dormancy, and outgrowth of buds that will give place to new stems, leaves, flowers, and branches are regulated by external and internal factors such as genetics, apical dominance, photoperiod, temperature, nutrition, and the amount and quality of the light, among others (Lortie and Aarssen, 1997; Cooke et al., 2012; Leduc et al., 2014). Among those factors, increasing the quantity of light plays a crucial role in reducing SAS responses like internode extension and axillary bud dormancy, while promoting bud outgrowth and overall growth in various species (Gommers et al., 2013; Leduc et al., 2014). Furthermore, the striking similarities in the main stem heights ( Figure 3 ) and differences in primary and secondary branches ( Figure 4 ) suggest that photochemical resources were strongly prioritized and influenced by the main apical meristem. The impact of this effect can be so pronounced that, in certain species or under specific growth conditions, the strong control of branching by the shoot apex may only cease upon removal (Lortie and Aarssen, 1997). Notably, in our study, cannabis branching proliferation was effectively achieved from ~18 to 52 mol m^-2 d^-1 ( Figure 4 ) while maintaining a water and nutrient supply upon plant demand.

The effects of increased supplemental lighting on water use and water use efficiency (WUE) were investigated to understand how water is used at a whole plant and production unit (cutting) in a 21-day production cycle. The primary increases in crop WUE observed in the present study are likely attributed to the pronounced benefits of enhanced photosynthesis at both leaf and crop levels, outweighing the potential drawbacks of increased transpiration under increasing lighting conditions. The increment in WUE at the crop level correlated with higher WUE at the leaf level, which is discussed later in this study. The findings of this study are corroborated by prior research conducted on greenhouse-grown sunflowers, wherein continuous measurements of crop photosynthesis and water use were undertaken under greenhouse solar conditions. In this previous study, it was observed that crop WUE throughout the day aligned closely with solar radiation levels, with peak values coinciding with the highest levels of solar radiation and photosynthesis (Hui et al., 2001).

Despite the correlations observed between branches and WUE ( Figure 5C ), it is imperative to note that the improvement in water efficiency is likely primarily attributable to impacts on photosynthesis and transpiration rather than the specific type of branches.

Currently, there exists no standardized protocol for growers and researchers to systematically select the most optimal branch for propagating cuttings. Instead, cutting selection predominantly relies on individual preferences. In order to address the various potential scenarios, Figure 5D elucidates the relationship between branch selection and water usage under different light conditions. Based on these findings, on average, a nursery crop operating under conditions similar to those in this study is anticipated to require between 20 and 33 liters of water for every 100 cuttings, assuming that primary branches are mainly used to produce cuttings from secondary branches. Moreover, it’s noteworthy that the water demand per cutting is projected to be higher under lower greenhouse lighting levels (<18 mol m^-2 d^-1).

For flower producers, the shoot WUE results imply that to grow the same shoot biomass in 21 days, growers need 0.37 L g^-1 under 18 mol m^-2 d^-1 of light and only 0.24 L g^-1 under 52 mol m^-2 d^-1, a 35% water reduction in evapotranspiration. Similar effects were found in a greenhouse hydroponic lettuce crop that transpired about 0.5 L g^-1 under 8 mol m^-2 d^-1 and 0.27 L g^-1 under 22 mol m^-2 d^-1, a 46% water reduction in transpiration (Both, 2002).

This study’s results align with similar findings reported for cannabis cultivation under sole-source LED lighting, which exhibited stronger photosynthetic responses and DLI variations (Rodriguez-Morrison et al., 2021). Furthermore, higher leaf maximum-A capacities, as seen in this study and Rodriguez-Morrison et al. (2021), are expected under light augmentation due to increments in leaf weight per area (Penning de Vries et al., 1989). Thicker or denser leaves can produce more RuBisCO (Penning de Vries et al., 1989) and chlorophyll in the same area ( Table 3 ), potentially increasing photosynthesis (Emerson, 1930). On the other hand, higher respiration rates, driven by the elevated biomass and metabolite levels per area ( Table 3 ), may provide an explanation for the lower net photosynthetic rate observed in thicker leaves in cannabis research. In summary, our leaf photosynthetic rates results support the remarkable crop yield increments even at maximum greenhouse radiances of ~2000 µmol m^-2 s^-1 ( Supplementary Figure S3B ).

In this study, the observed rise in E with increased light intensity was attributed to the increased stomatal conductance (g_sw). Typically, a higher g_sw is linked to either a reduction in leaf internal CO₂ concentration (Ci) or an enhancement in photosynthesis (A) through light supplementation (McAusland et al., 2016; Driesen et al., 2020). However, despite the linear increase in g_sw with light intensity, the responses of A and Ci did not parallel the stomatal conductance trend observed in our study ( Figures 6C, A ; Supplementary S4A ).

The linear response in g_sw compared to the plateau response of A could be explained by additional mechanisms affected by light other than C_i reduction or photosynthesis enhancement. For instance, blue and red radiation is mentioned as direct influencers on guard-cell water status (Driesen et al., 2020), while shortwave radiation, encompassing most or all solar wavelengths, demonstrated an impact on leaf water status (Lawson et al., 2010; Pieruschka et al., 2010). Pieruschka et al. (2010) proposed a regulatory mechanism where the balance between radiation-driven water vapor production and transpiration rate governs stomatal conductance. Moreover, their observations found no specific wavelength effects, establishing only a linear correlation between stomatal conductance and absorbed radiation energy. Another factor affected by radiation is leaf temperature, a pivotal influence in stomatal conductance. Feller (2007) and Urban et al. (2017) demonstrated significant increments in stomatal aperture across temperatures ranging from ~20°C to over 40°C. Moreover, these increments persisted despite reductions in A and water potential and an increase in C_i (Urban et al., 2017). In the present cannabis study, results showed marginal leaf temperature differences (26.1 ± 0.17 to 26.9 ± 0.35°C) when elevating the PPFD from 150 to 2000 µmol m^-2 s^-1 ( Supplementary Figure S4B ). In conclusion, light can affect photosynthesis and transpiration differently and may act on the stomatal aperture via various mechanisms.

Our study found considerably higher E and g_sw values per photon of light at lower PPFDs than at higher light intensities, as previously shown in the result section. Moreover, leaf transpiration at 150 µmol m^-2 s^-1 constituted a noteworthy 57% of the total water loss found under 2000 µmol m^-2 s^-1 ( Figure 6B ). This can be explained by reports across diverse species (Hui et al., 2001; de Dios et al., 2015; Dayer and Gambetta, 2021; Dayer et al., 2021) that suggest rapid and excessive increments in stomatal opening at low light or in the absence of light, which can increase water losses. Furthermore, data from different species, spanning ferns, gymnosperms, and angiosperms, also revealed that at very low light levels (50 µmol m^-2 s^-1), stomatal conductance constituted between ~16% to 64% of the aperture at 1000 µmol m^-2 s^-1 (Deans et al., 2019). These results also align with increments above typical nighttime transpiration rates of 5% to 15% (Caird et al., 2007). Moreover, the observed rapid and predawn stomatal aperture increments have been associated with specific wavelengths (Lawson et al., 2010) and circadian regulation (de Dios et al., 2015; Dayer et al., 2021), while stomatal conductance did not correlate with photosynthesis at low light (Deans et al., 2019).

In addition to increasing photosynthesis, enhancing supplemental lighting strongly reduced the water use per photon of light. These findings align with the broader context of light effects on stomatal conductance observed across various species (Deans et al., 2019), as previously discussed. Additionally, similar trends in crop transpiration have been documented in beans and cotton subjected to maximum PPFDs of 500 and 1500 µmol m^-2 s^-1. Notably, nocturnal water losses constituted 23% and 12% of daytime requirements under lower and higher PPFDs, respectively (de Dios et al., 2015). These results suggest that optimizing lighting conditions can significantly mitigate the impact of both night-time and day-time water losses.

Stomatal conductance and transpiration exhibited linear increases by increasing PPFD, as detailed earlier. However, the absence of statistical evidence for DLI effects between 18 and 52 mol m^-2 d^-1 was unexpected. It is plausible that the lack of statistical differences is linked to the substantial variability in transpiration and stomatal conductance due to the interaction of our procedure’s sampling speed and stomatal acclimation periods (Lawson and Blatt, 2014; McAusland et al., 2016). Nevertheless, the information and results reviewed emphasize the complex interplay between direct and indirect light effects on guard cells, transpiration, and physiological dynamics, contributing to notable increases in gas exchange within the context of our research. Additionally, it is noteworthy that cannabis leaf temperatures remained similar to or below an air temperature of 27°C. This observation suggests that transpiration rates were sufficient to maintain optimal metabolic functions without the risk of overheating (Feller, 2007; Nelson and Bugbee, 2015). Our investigation revealed no physiological constraints on stomatal conductance and transpiration, even under the more demanding conditions of ~2000 µmol m^-2 s^-1 of actual greenhouse LED lighting and solar radiation (unshown data). This and previous research stress the robustness of optimal cannabis growth under high-light environments.

The leaf WUE of upper cannabis leaves remarkably increased as a consequence of the curvilinear and linear responses observed for photosynthesis and transpiration ( Figures 6A, B , and 7 ). Consistent with our findings, Deans et al. (2019) reported a substantial increase in intrinsic WUE (A/g_sw) under elevated lighting conditions across species, which was driven by a more pronounced increase in photosynthesis than stomatal conductance. In our study, there was weaker evidence indicating that the leaf WUE of the 52-DLI treatment might surpass that of the 18-DLI WUE under light levels of 1000 µmol m^-2 s^-1 or higher (p-values >0.05 to 0.09); however, at α 0.05, the statistical support was not robust enough to substantiate our hypothesis that leaves developed under the 52-DLI treatment exhibit higher WUE than leaves under the 18-DLI treatment.

In conclusion, the observed increase in crop WUE with higher lighting, consistent with other species, lighting conditions (Hui et al., 2001; Both, 2002; de Dios et al., 2015), and supported by increments in leaf WUE across various cannabis cultivars (unpublished data) and fifteen different species (Deans et al., 2019), suggests a potential for elevated WUE and yield in unstressed crops exposed to increased lighting conditions.