*3.1. Chlorophyll Fluorescence*

The maximum quantum efficiency of PSII photochemistry (Fv/Fm) indicates how effectively PSII uses absorbed light energy to reduce the primary quinone acceptor of PSII (QA) [37]. In practice, this measure can be used to assess stress in plants, as a value of ~ 0.83 is very consistent across species in non-stressed leaves [38]. Values below 0.83 indicate stress and a reduced maximum photosynthetic capacity; however, photosynthesis may not be reduced under ambient conditions as the quantum yield of PSII (ΦPSII) is generally considerably lower than Fv/Fm, especially under high light intensity. A low Fv/F<sup>m</sup> is one of the symptoms of red light syndrome [11–13].

Fv/F<sup>m</sup> was qualitatively lower in G, R, and RG than all other treatments, indicating a reduced maximum photosynthetic efficiency with values suggesting mild stress (Table 4). These qualitative differences are supported by previous findings that a comparatively higher level of blue light in LED treatments increased Fv/F<sup>m</sup> relative to high pressure sodium treatments [39]. Others have also concluded that blue light enhances PSII photochemistry relative to red light [11–13].

PSII operating efficiency decreases with increasing light intensity, primarily due to a reduced ability to oxidize Q<sup>A</sup> rather than an increase in non-photochemical quenching (NPQ) [37,40]. Therefore, it is not surprising that under saturating light the PSII operating efficiencies observed were much lower than Fv/Fm. PSII operating efficiency was significantly lower in G, R, and RG than all other treatments (with ΦPSII in R significantly lower than RG, which was significantly lower than ΦPSII in G) (Table 2). It is not possible to estimate electron transport rate or the quantum yield of CO2, since we cannot account for alternative electron sinks to PSII because these measurements were taken under atmospheric O<sup>2</sup> concentrations. Nevertheless, ΦPSII gives an estimate on the upper limit of possible photosynthetic carbon assimilation under a given condition, and the trend observed is very similar to the trend in net photosynthesis observed.

Despite the lower ΦPSII values, ΦNPQ, the quantum yield of light-induced quenching, and ΦNO, non-light-induced quenching, are both significantly higher in the G, R, and RG treatments than all other treatments (Table 2). Together, these data suggest that electron acceptors downstream of PSII are insufficient in the G, R, and RG treatments compared to the other treatments, and that G, R, and RG treatments are compensating by increasing nonphotochemical quenching to reduce photo-induced damage. Since the spectral quality and intensity used to excite the photosystems were identical across treatments during light-saturated measurements, one would expect differences in net photosynthesis and chlorophyll fluorescence to be related to adaptive differences between light treatments.


**Table 4.** Color breakdown for light treatment spectra as a percentage of total photosynthetic photon flux density (PPFD). Wavelength ranges for the traditional method indicate the typically defined range for each color, while the wavelength range for the bar method indicates the range in which >99% of the light is emitted from a given bar color.

> Plants have a variety of mechanisms to respond to changes in light quality. In the short term, light-harvesting complex II (LHC-II) can be transferred from PSII to PSI to help balance excitation energy between the two systems to improve electron transport efficiency [41]. In the long term, algae, cyanobacteria, and higher plants adjust the stoichiometry of photosystem I (PSI) and photosystem II (PSII) in response to light quality to improve photosynthetic efficiency [42–45] as well as their pigment composition [42,46] to more efficiently absorb ambient light.

> PSII is primarily excited by wavelengths at ~450–640 nm while PSI uses light above 680 nm much more efficiently than PSII [47]. Since the blue LEDs are the only source of photons at 450 nm in our light treatments, treatments lacking these wavelengths (G, R, and RG) likely have adjusted stoichiometry to decrease the number of PSI complexes relative to PSII to improve electron transport efficiency due to less efficient excitation of PSII relative to PSI. As the green LEDs supply light within the range that PSII can use effectively, this stoichiometric adjustment would be expected to be most pronounced in the R treatment, and less so in the G and RG treatments.

> When the plants were exposed to the novel light treatment (90% red, 10% blue) during A vs. C<sup>c</sup> measurements they could use transient LHC-II to improve the balance of excitation between PSI and PSII, but the capacity to balance in the G, R, and RG treatments may have been limited by extreme stoichiometric differences not seen in the other treatments. This would explain the much poorer performance of these three treatments relative to the other treatments and the poorer performance of R relative to G and RG during A vs. C<sup>c</sup> measurements, while the same long-term adaptations may have allowed for the trends seen in Figure 1 during measurement under ambient lighting.

> In any case, neither the photosynthesis measurements under saturating 90% red, 10% blue light at 1000 µmol photons m−<sup>2</sup> s <sup>−</sup><sup>1</sup> nor ambient light at 170 µmol photons m−<sup>2</sup> s −1 correlate well with shoot dry weight. There are many potential reasons for the lack of correlation between shoot biomass and photosynthesis measurements. First, the dry weight data include only shoot biomass, not root biomass. It is possible that with root biomass, the whole-plant biomass values would correlate well with the net photosynthesis measurements observed. The photosynthesis values presented are on a per-area basis. Leaf area and specific leaf weight (leaf area divided by leaf dry weight) vary between treatments.

It is therefore possible that plants with equivalent net photosynthesis rates could have very different whole plant growth rates due to differences in leaf area [48].
