The Experiment I revealed that sugarcane and energy canes presented similar photosynthetic capacity in the top leaves of the canopy (Fig. 1a), those exposed directly to sunlight. However, such similarity between the top leaves did not explain the differential biomass production when comparing IACSP95-5000 to Vertix 2 and Vertix 3 (Fig. 4a). Aiming a comprehensive evaluation of canopy photosynthesis, we performed the Experiment II to investigate the photosynthesis in all leaves of the main tiller under natural conditions, focusing our analysis on how it varied along the canopy. In principle, such variation among genotypes could be caused by leaf aging and/or self-shading. This latter can be ruled out as potted plants exhibit lower shading due to neighbors when compared with plants growing in field condition, consequently light availability along the canopy profile was similar among cultivars (Fig. S2). Briefly, our findings revealed that the variation in photosynthetic traits along the canopy profile resulted in differential CO2 uptake among genotypes – corroborated by the regression analysis shown in Fig. S3 – and caused higher biomass production in Vertix 2, the energy cane type II. Such photosynthetic variation caused by leaf aging and its effects on biomass production in IACSP95-5000, Vertix 2 and Vertix 3 are discussed below.
Canopy photosynthesis in sugarcane and energy canes
Leaf photosynthetic capacity reduces with aging due to degradation of chlorophyll, Rubisco and other proteins – endogenous sources of nitrogen to other organs (Yin et al. 2009). However, decreases in Aa and Am observed along the canopy profile of IACSP95-5000 and Vertix 3 were not associated with the chlorophyll index as no variation occurred with leaf aging (Figs. 5a, b, h and S3a, c). The concomitant variations of Rubisco abundance, LNC and CO2 assimilation revealed to be more complex than one proposed by Yin et al. (2009), as Vertix 2 showed decreases in Rubisco and LNC without changing Aa or Am (Figs. 5a, b, i, 6b and S3b). According to Bindraban (1999) and Tominaga et al. (2015), PARinc within the canopy, rather than LNC variation with leaf aging, has a predominant effect on photosynthesis of wheat and sorghum plants. This would explain why leaf CO2 assimilation did not change along the canopy profile of Vertix 2 even with this genotype presenting significant reduction of LNC (Figs. 5a, b, i and S3b). As shown in Fig. S2, there was a gradual decrease in PARinc along the canopy profile of all genotypes, but the light availability in each canopy layer was similar among genotypes. Even with PARinc values being similar among the genotypes, IACSP95-5000 exhibited the highest decrease in photosynthesis along the canopy profile (Figs. 5a, b, S2 and S3). Such finding reveals a capacity of Vertix 2 and Vertix 3 canopy to counterbalance the leaf aging with a better light utilization when compared to IACSP95-5000.
The main difference among the genotypes was how the photosynthetic traits varied along the canopy profile. IACSP95-5000 showed the highest decrease in Aa, Am, qP and φPSII, a sharp reduction of Rubisco in leaf + 4 and the lowest relative PEPC abundance when compared to Vertix 2 and Vertix 3 (Figs. 5a-g and 6). Vertix 3 – the type I energy cane – showed a lower variation of the photosynthetic traits when compared to IACSP95-5000, a higher relative PEPC abundance and maintained the amount of Rubisco unchanged along the canopy profile (Figs. 5a-g and 6). Then, the reduction of CO2 assimilation due to leaf aging in Vertix 3 was lower than one observed for IACSP95-5000 (Figs. 5a-b and S3). Surprisingly, Vertix 2 – the type II energy cane – exhibited non-significant reduction in leaf CO2 assimilation and photochemical activity along the canopy profile, with similar relative PEPC abundance in leaves + 1, +4 and + 7 and a gradual and slight decrease of Rubisco with leaf aging (Figs. 5a-g and 6). Taken together, such characteristics of Vertix 2 maintained CO2 assimilation along the canopy profile and caused the highest total CO2 uptake by the main tiller (Figs. 5a, b, 7a and S3). Our findings revealed that the leaves of Vertix 2 and Vertix 3 presented an important difference between the chronological and physiological ages, which resulted in older leaves exhibiting high photosynthetic activity like the newer leaves.
Even with the highest PEPC abundance being found in Vertix 3 and the highest Rubisco abundance found in IACSP95-5000, these genotypes did not present the highest leaf CO2 assimilation (Figs. 1–3, 5a, b and 6). Such results could be explained by differences in the activation state of PEPC and Rubisco, in leakiness and in PEP and RuBP regeneration rates. In fact, reduced activation state of PEPC in Vertix 3 and Rubisco in IACSP95-5000 would explain similar Vpmax, Vcmax, Aa and Am in leaves + 1 (Figs. 2a-b, 3b, d, 5a, b and 6). Salesse-Smith et al. (2018) showed that decreases in the activation state of Rubisco balanced its overexpression in maize and there is no proportional increase in CO2 assimilation. On the other hand, leakiness represents the portion of fixed CO2 that leaks back to the mesophyll cells from the bundle sheath cells, being dependent on [CO2] gradient between both cell types and on the ratio of PEPC and Rubisco activity (Yin et al. 2011). As the C4 cycle provides the CO2 for the C3 cycle, the balance between PEPC and Rubisco activity defines the final CO2 assimilation rate. Herein, Vpmax:Vcmax ratios were similar among genotypes and then leakiness would be likely similar for leaves + 1 (Fig. 3b, d). In addition, one would argue that high PEPC abundance could be a physiological strategy to counterbalance low Rubisco abundance in Vertix 3, a hypothesis to be further explored by addressing the role of the bundle-sheath conductance and anatomical traits on photosynthesis (Tofanello et al. 2021).
Finally, the PEP and RuBP regeneration rates are known as limiting factors for CO2 assimilation and depend on ATP produced by the photochemical reactions along the chloroplast electron transport chain (Yin et al. 2011; Zhou et al. 2019; von Caemmerer 2021). Herein, there were no differences among genotypes when considering φPSII of leaves + 1 under optimum conditions – Experiment I (Figs. 1b and 2f). When the top leaves were analyzed under natural conditions in Experiment II, Vertix 3 exhibited lower qP and φPSII and higher NPQ than Vertix 2 (Fig. 5d, e, g). Such low photochemical efficiency would affect the production of ATP, and then limit the regeneration of PEP and RuBP of top leaves in Vertix 3. However, we believe such slight differences in φPSII are not enough to induce any significant change in PEP and RuBP availabilities, as all genotypes presented similar photosynthesis in leaves + 1 (Figs. 2a, b and 5a, b). Overall, Vertix 2 and Vertix 3 maintained the photochemical efficiency along the canopy profile, which may explain why leaf aging did not reduce leaf CO2 assimilation as in IACSP95-5000 (Figs. 5a, b, g, and S3).
Photosynthesis, respiration and biomass production
Besides all genotypes have presented similar photosynthetic rates in leaves + 1 (top canopy), seven-month-old plants showed similar tillering, number of leaves in the main tiller (6.7 ± 0.2, BF10 < 0.5) and top biomass (Figs. 1a, 2j and 4c). Facing these similarities, one could expect the same biomass production among genotypes, but surprisingly, Vertix 2 produced 25% and Vertix 3 17% more biomass than IACSP95-5000 (Fig. 4a). This latter exhibited the highest total leaf area per pot in Experiment I (Fig. 2l), which could support high canopy photosynthesis and then high biomass production, but the opposite is true. Vertix 2 and Vertix 3 have narrow leaves than IACSP95-5000 (Alexander 1985). As compared to the energy canes – mainly Vertix 3 – IACSP95-5000 exhibited the highest leaf area for a given leaf length (Fig. S4), which is a likely consequence of higher leaf width.. However, Vertix 2 and Vertix 3 presented more well-developed tillers – given by higher stalk biomass – than IACSP95-5000, which resulted in more leaves per soil surface area (Figs. 2k and 4b). Such canopy traits of Vertix 2 and Vertix 3 revealed to be more important to the overall photosynthesis than a high photosynthetic rate per leaf area unit, as it resulted in higher biomass production by Vertix 2 and Vertix 3 as compared with IACSP95-5000 (Fig. 4a). Moreover, we evaluated only the maximum photosynthetic rates of top leaves in Experiment I, and then the actual photosynthesis and the role of older leaves within canopy were not addressed.
When evaluating the canopy profile of photosynthesis, we found that – in spite of the similarities observed in Experiment I – IACSP95-5000 was more responsive to leaf aging than Vertix 2 and Vertix 3. Decreases in photochemical activity and Rubisco abundance along the canopy profile of IACSP95-5000 caused a large variation of Aa and Am when comparing the new (upper position) and old (bottom position) leaves (Figs. 5a, b, g and 6b). Such larger variation is evident when evaluating the slope of the linear regression of Aa against leaf age, which is two times higher for IACSP95-5000 as compared to Vertix 3 (Fig. S3). Although At revealed non-significant differences among the genotypes, the main tiller of Vertix 2 and Vertix 3 had more biomass than one of IACSP95-5000 (Figs. 7a and 8f). However, one must consider that these energy canes have more tillers than the sugarcane (Fig. 7b) and it is reasonable to assume similar profiles of photosynthesis among tillers from a clump. In fact, the biomass production is an integrative process that takes into account not only all tillers but also the interaction among them in a given surface area along the growth period. Interestingly, we found differences between Vertix 2 and Vertix 3 for photosynthesis and biomass production. Vertix 3 – the type I energy cane, bred for sugar and fiber production – exhibited higher variation of Aa and Am and lower At, top, stalk and total biomass of the main tiller than Vertix 2 – the type II energy cane, bred primarily to fiber production (Figs. 5a, b, 7a, 8f-h and S3b, c). Similar to our findings in Experiment I, Vertix 2 and Vertix 3 exhibited more well-developed tillers than IACSP95-5000 – given by its higher tillering, number of leaves per pot and stalk biomass – and the total biomass production by Vertix 2 and Vertix 3 was 85% and 61% higher than by IACSP95-5000, respectively (Figs. 7b, c and 8a, b).
Biomass accumulation is the net result of photosynthesis – which produces photoassimilates – and respiration – which consumes such photoassimilates to generate energy and carbon skeletons (Singels et al. 2005; Zhu et al. 2010). The overall respiration is composed by the maintenance respiration and growth respiration. While the maintenance component occurs in all living cells and form highly energetic and reducing compounds to maintain the existing biomass (Amthor 2000), the growth respiration is associated with the cost of producing new structures in cells undergoing division during the process of biomass production (Singels et al. 2005). Herein, Vertix 2 exhibited an increasing trend in respiration with leaf aging (Fig. 5c) and an average respiration of all leaves of the main tiller (0.90 ± 0.03 µmol m− 2 s− 1) significantly higher (BF10 > 3.7) than IACSP95-5000 (0.69 ± 0.06 µmol m− 2 s− 1) and Vertix 3 (0.47 ± 0.04 µmol m− 2 s− 1). As increased maintenance respiration reduces the photoassimilate conversion rate and then the biomass production (Zhu et al. 2010), our data suggest that such higher respiration of Vertix 2 is associated with the growth respiration as this genotype exhibited the highest accumulation of biomass in the top of the main tiller (Fig. 8h).