4. Discussion
Diurnal variation of rice Fv/Fm in this study was consistent with what reported by Wu et al . (2007) and Panda et al . (2011) (Fig. 2). The decrease of Fv/Fmat midday was thought to be photoprotection under high-light intensity and high-temperature rather than photodamage to PSII (Roden & Ball. 1996; Huang et al ., 2006), resulted from reversible inactivation of PSII reaction centers, which was responsible for the midday depression in photosynthesis (Roden & Ball, 1996; Panda et al ., 2011). The negative correlation between Fv/Fm and PPFD and temperature supported photo-protection theory (Fig. 3).
Elevated [CO2] enhanced PSII thermotolerance showed by the greater Fv/Fm under high temperature (Taub et al ., 2000; Pan et al ., 2018). Stepwise increase of [CO2] (SI) further enhanced high-light and heat resistance than short-term abrupt increase of [CO2] (AI160), demonstrated by alleviated midday depression in Fv/Fm across three growth stages (Table 2; Fig. 2). However, constant increase of [CO2] (CI) did not show this advantage than short-term abrupt increase of [CO2] (AI200). Different effects of the two treatments of elevated [CO2] suggested that simulations closer to the actual pattern of [CO2] increment may be more effective in reflecting crop responses to elevated [CO2]. Consistent with conjecture of Taub et al . (2000), increased PSII thermotolerance under elevated [CO2] may be associated with lower initial (Fo) at midday, while Pan et al . (2018) reported that elevated [CO2] protected PSII by reducing the heat stress-induced reactive oxygen species accumulation. The different results remind us that more mechanism researches should be carried out.
Regardless of different treatment of [CO2] increase, the midday decrease in Fv/Fm of rice were 8%, 11% and 7% at jointing, heading and grain-filling stages, respectively (Table 2), lower than 37% at booting stage reported by Wu et al . (2007). In addition to the differences in growth periods, it may be related to the different cultivar of rice used in our study (japonica) and the experiment of Wuet al . (2007) (indica). When it came to the tolerance to high light, as reported by Li et al . (2002) that the decrease of Fv/Fm in high-light resistant japonica rice cv. 9516 was the least, while the depression in strong-light-sensitive indica rice cv. shanyou 63 was the most. Even for two photosensitive indica rice, the Fv/Fm of rice cv. IR42 reduced more than cv. FR13A at midday (Panda et al ., 2001). Planting high-light and high-temperature tolerated varieties of rice may help to maintain higher PSII efficiency at noon, thus reducing the midday downregulation of photosynthesis.
Short-term elevation of [CO2] (one growing season) enhanced Fv/Fm in japonica rice cv. Fujiyama-5 (Ziska & Teramura, 1992). Multi-year [CO2] elevation (four generations) caused higher Fv/Fm than short-term increase of [CO2], whether it was a constant increase or a stepwise increase of [CO2] (Fig. 2), indicating that long-term increase of [CO2] enhanced beneficial changes to improve photosynthesis in rice. As reported by X. Li et al . (2019), multi-generational exposure to elevated [CO2] could reinforce the response occurred in short-term exposure, so that long-term response of crop to increasing [CO2] would not be completely predicted by short-term response to elevated [CO2].
The SI treatment, consistent with actual increases pattern of atmospheric [CO2], was more conducive to strengthen the advantage mentioned above. It can be indicated by the result that the higher Fv/Fm occurred under SI than AI160 across all three growth stages, while Fv/Fm in CI was only higher than AI200 at jointing stage (Table 2; Fig. 2). Roden & Ball (1996) showed that the depression in Fv/Fm was associated with the reduced reaction center as reflected by an increase in the minimum fluorescence (Fo) and the increased levels of nonstructural carbohydrates. Since the corresponding nonstructural carbohydrates were not determined, we could not derive the relationship between them, but the reduced Fo corresponded to the increased Fv/Fm under SI and CI (Fig. S1), suggesting that long-term exposure to higher [CO2] increased rice reaction center, which helps PSII capture more light quantum for electron transfer.
Significant interactions between [CO2] treatments and observation time points on Fv/Fm (p = 0.003 at heading and p = 0.087 at jointing), PIABS ( p = 0.046) and φEo (p = 0.066) suggested that attention should be paid to the observation time points in studying the effects of increased [CO2] on PSII efficiency (Table 2). Measurements that span a long time in a day may amplify or mask treatment effects due to diurnal variation. Predawn observations of plants that have been dark-adapted overnight may be more effectively to reflect the effects of treatments (Kalaji et al ., 2014). Values observed at sunset were almost completely restored to those in the state of predawn, which can be used for analysis. In addition to the treatment effects of elevated [CO2], the observations at noon and afternoon may also involve the response of plants to high-light and high-temperature, and the ability to recover.
Contrary to the results of Zhu et al . (2018) that short-term elevation of [CO2] decreased ψo of wheat, in this study, the marked increase of ψo under SI than AI160 at grain filling stage reflected that SI increased plastoquinone pool, thus reduced the accumulation of QA, along with the unhindered donor side of PSII and electron transfer chain (Fig. 4; Fig. S2). This attribution was supported by previous studies revealing that the accumulated QA- and reduced PQ pool caused higher fluorescence of OJ-phase (Kalaji et al ., 2011; Tsimilli-michael, 2019). While, CI did not affect ψo, speculating that long-term increment of [CO2] mitigated adverse effects of short-term elevation of [CO2], and the stepwise increase would be more beneficial to PSII. On the other hand, this might the difference between wheat and rice in response to increased [CO2].
The φEo, which was the product of ψo and Fv/Fm, showed stronger correlation with ψo than Fv/Fm(Table 4), indicating that elevated [CO2]-induced changes in electron transport had more to do with the changes in ψo than in Fv/Fm. The process of transferring electrons by captured photons in the reaction center may be the rate-limiting step of the initial photochemistry. The dominant role of ψo was also reported by Jiang et al . (2008). The correlation coefficient between PIABS and Fv/Fmwas the highest, suggesting that Fv/Fmcontributed the biggest fraction to the increase in PIABS due to elevated [CO2] (Table 4).
The significant decrease in ψo, φEo and PIABS along with the growth periods indicated that rice regulate the PSII efficiency according to growth needs, which may lead to a reduction in photosynthesis (Table 5). The lower attenuation under CI and SI indicated that long-term increase of [CO2] slowed down the reduction of PSII efficiency from jointing to grain filling stage, which may be a strategy to maintain higher photosynthesis.