An indica rice genotype showed a similar yield enhancement to that of hybrid rice under free air carbon dioxide enrichment

Although the rice growth response to FACE (free-air CO2 enrichment) has been widely studied and is considered important within the scientific community, few studies have attempted to examine the effects of FACE on the yield of indica rice, which is typically the parent of indica hybrids in China. The effects of FACE on the yield, yield components, biomass, N uptake and leaf photosynthesis of Yangdao 6 Hao (an indica rice) in China were examined over 2 years. The grain yield increased over 30%, the panicle number increased 12.4% on average, and the spikelet number per panicle also showed an average increase of 8.2% at elevated CO2. FACE caused a significant enhancement in both the filled spikelet percentage (+5.9%) and the individual grain weight (+3.0%). Compared with three prior FACE studies on rice, a similar enhancement of yield in hybrid indica was shown under FACE, with much a higher value than for the japonica rice cultivar (approximately + 13%) because of indica’s stronger sink generation and N uptake capacity, which help coordinate the C/N balance to avoid photosynthetic acclimation. The high enhancement of the indica rice yield under FACE holds promise for improved cultivar selection for future food security.

Yangdao 6 Hao has large panicles, a high yield potential, resistance to disease and pathogens, and an anti-lodging ability, which suggest that this cultivar may be an important gene resource for rice breeding 14 . Hence, we chose this cultivar as the study subject. The aim of this study was to investigate whether the indica rice under consideration has a similarly strong response to elevated [CO 2 ] as hybrid rice using FACE (free air carbon dioxide enrichment) treatment.

Results
Effects of CO 2 on grain yield. FACE significantly increased the grain yield of the rice (P < 0.01) ( Table 1). The enhancement of the grain yield was 36.2% and 29.6% in 2012 and 2014, respectively. There was a strong effect of year on the yield response (P < 0.05) ( Table 1), but interactions between CO 2 and year were not detected.
Effects of CO 2 on yield components. As shown in Table 1, the panicle number per m 2 was increased to a similar extent (13.5% for 2012 and 11.2% for 2014, P < 0.05) under elevated [CO 2 ], and there was no significant interaction between CO 2 and year. The number of spikelets per panicle increased by 9.6% for 2012 and by 6.8% for 2014, and there was a strong year effect (P < 0.01). There was no interactive effect of CO 2 × year on spikelets per panicle (Table 1).
For the filled spikelet percentage, FACE rice showed a similar increase in the two years (5.8% for 2012 and 5.9% for 2014). The individual grain weight increased by 3.5% and 2.3% for FACE vs. ambient plants in 2012 and 2014, respectively ( Table 1). The interaction between CO 2 and year was not detected for the two yield components.
Effects of CO 2 on phenology, shoot and tiller biomass, and plant height. There was no change in phenology upon reaching 50% panicle emergence and grain maturity ( Table 2). In contrast, FACE significantly increased the shoot and tiller biomass and plant height at maturity. When averaged across years, the shoot and tiller biomass and plant height were increased by 29.0%, 14.3% and 4.5%, respectively. There was no interactive effect of CO 2 × year on phenology, shoot and tiller biomass or plant height ( Fig. 1 and Table 2).
Effects of CO 2 on gas exchange. There was a significant stimulation of flag leaf photosynthesis with elevated [CO 2 ] (relative to ambient) observed at the mid-filling stage for plants grown under either ambient or FACE conditions ( Fig. 2A). There was no difference in the net photosynthetic carbon assimilation rate of plants grown under ambient and FACE conditions when measured at the same [CO 2 ] (590 μ mol mol −1 ), indicating that the photosynthetic efficiency was not reduced by elevated [CO 2 ], even at the mid-filling stage ( Fig. 2A). The leaf temperatures were not different during measurement. The stomatal conductance and transpiration were not significantly different under ambient CO 2 and FACE conditions, although there was a declining trend due to the elevated [CO 2 ] (Fig. 2B,C). Considering the meaningless in stomatal conductance and transpiration with elevated [CO 2 ] for plants grown under ambient conditions, we did not show these parameters. N uptake. Averaged across two years, the N uptake in the FACE condition was increased by 15.9% during the vegetative stages (from transplanting to heading) and 15.7% during the reproductive stages (from heading to maturity) (Fig. 3). There was a clear variation between the two growth periods (Fig. 3), but the CO 2 effect and stage interactions were not significant.

Discussion
The 2-year FACE study showed that elevated [CO 2 ] increased the yield of Yangdao 6 Hao by over 30% (Table 1), which was a similar enhancement to that of hybrid rice in previous reports of China FACE 9,11 . This value is far higher than the range reported in inbred japonica rice FACE studies 5,8,10,12 . Undoubtedly, this study reveals the potential for taking full advantage of higher [CO 2 ] levels for inbred rice varieties. The FACE condition significantly enhanced the panicle density by 12.4% in Yangdao 6 Hao (Table 2), 10.3% in Shanyou 63 and 7.8% in Liangyoupeijiu 9,11 , all of which are smaller than the 18.8% increase in Wuxiangjing 14 in China Wuxi FACE 8 . The different responses to FACE among these varieties may be associated with differences in the dimensions of the leaf laminae 9 . Compared to rice cultivars with small and erect leaf blades (e.g., Wuxiangjing 14), Yangdao 6 Hao, Shanyou 63 and Liangyoupeijiu, with large and drooping leaves, would suffer more from mutual shading during crop development, thus presumably resulting in a weak stimulation of CO 2 induction for tillering and the resulting panicle number 9 .
In the present study, the number of spikelets per panicle in Yangdao 6 Hao increased by 8.2%, which was slightly lower than in hybrid indica but higher than in japonica under FACE conditions ( Table 2). Shimono et al. reported that the N uptake before the heading stage was closely correlated with the spikelet density rather than the [CO 2 ] and the cultivar type 10 . N uptake by Yangdao 6 Hao increased by 15.9% before the heading stage under elevated [CO 2 ]. The enhanced N uptake before heading was beneficial to increasing the spikelet number for Yangdao 6 Hao under FACE (Tables 1 and 2). In addition, the substantial enhancement in panicle size in this study was supported by the corresponding responses of plant height (Fig. 1A) and shoot biomass (Fig. 1B, Table 2) to elevated CO 2 , which were consistent with the findings that the height and tiller biomass were correlated positively and significantly with panicle size 15 . The spikelet number per panicle is the result of the difference in the number of differentiated and degenerated spikelets 8 . It is well accepted that cytokinins are mainly produced in the plant root and distributed in the shoot by the transpiration stream 16,17 , which impacts rice spikelet formation and development. The decreased spikelet number of japonica rice varieties with a low response to elevated CO 2 in China FACE (Table 2) may be attributed to the decrease in root activity that reduces cytokinin synthesis 18,19 .  6 Hasegawa et al. 12 Yang et al. 8 Liu et al. 9 Yang et al. 11 Table 2 and Fig. 1 in this paper This condition is unlikely to be the case for Yangdao 6 Hao, in which the spikelet number per panicle and the N uptake of the vegetative and reproductive stages were significantly increased by the elevated CO 2 . These findings suggest that Yangdao 6 Hao can maintain root activity for cytokinin synthesis under elevated CO 2 . However, the potential physiological and molecular mechanisms underlying the different responses to elevated CO 2 require further study. Elevated [CO 2 ] increased the actual grain sink per panicle (filled spikelet ratio × spikelet number × per panicle grain weight) by 18.0% in this study (Table 2), which is slightly lower than in hybrid rice but higher than the values obtained in previous FACE studies 6,8,9 . Obviously, grain-filling abilities are related to photosynthate assimilation after heading. At the mid-filling stage, Yangdao 6 Hao maintained a strong increase in the net photosynthetic carbon assimilation rate and avoided photosynthetic acclimation under elevated [CO 2 ] (Fig. 2A). This response is similar to that exhibited by the hybrid rice Shanyou 63, for which elevated [CO 2 ] resulted in an increased spikelet number and grain weight, increased sink:source ratio, and continued stimulation of photosynthesis up to grain maturity 20 . Overall, these results suggest that the greater response of this rice line to elevated [CO 2 ] may be associated with enhanced panicle sinks relative to sources and the ability to maintain photosynthetic capacity during grain development. Under FACE conditions, these rice lines can avoid the photosynthetic acclimation that is common in C 3 cereals 5,12,20,21 . As a result of the balance between carbon and nitrogen metabolism within the leaf under elevated [CO 2 ], the greater sink and the significant enhancement of N uptake during the filling stage (Fig. 2), these plants avoided the suppression of photosynthetic system genes and the resulting decrease in photosynthetic capacity 20,21 . Maintaining photosynthetic efficiency during the grain-filling stage ensures a strong yield enhancement under FACE conditions. This is the first study to confirm that an inbred indica genotype exhibits a yield enhancement similar to that of hybrid rice under elevated [CO 2 ]. To ensure food security in the future, additional indica genotypes with potentially strong responses to elevated [CO 2 ] should be evaluated to take full advantage of the predicted increases in [CO 2 ].

Materials and Methods
Research site. The experiment was conducted at the FACE facility located in Zhongcun Village (119°42'0"E, 32°35'5"N), Yangzhou City, Jiangsu Province, a typical Chinese rice-growing region 22 . The soil was classified as Shajiang-Aquic Cambiosol with a sandy loam texture. The soil properties at 0-15 cm relevant to this experiment are as follows: bulk density 1.16 g cm −3 , soil organic carbon 18.4 g kg −1 , total nitrogen 1.45 g kg −1 , available phosphorous 10.1 mg kg −1 , available potassium 70.5 mg kg −1 , and pH 6.8 23 . The operation and control systems for the FACE facilities were the same as those used at the Japan FACE site 24  . N was applied as a basal dressing (40% of the total) 1 day prior to transplanting and as a top dressing at early tillering (30% of the total) and at the panicle initiation (PI) stage (30% of the total) at 22.5 g N m −2 . Phosphorous (P) and potassium (K) were applied as a compound fertilizer at 9 g P 2 O 5 m −2 and 9 g K 2 O m −2 ; both P and K were applied as a basal dressing 1 day before transplanting.
Photosynthetic gas exchange measurement. Before the measurements, we measured the Chl content in 5-6 flag leaves per treatment plot non-destructively using a Chl meter (SPAD-502, Konica Minolta Optics, Inc., Japan). We then used two leaves with representative Chl content for the gas exchange measurements with a portable photosynthesis system with blue and red LED light sources (LI-6400, LI-COR Bioscience, USA). This measurement was conducted between 09:30 and 14:30 h on September 21, 2014 (mid-filling stage); the block temperature in the cuvette was fixed at 28 °C, and the photosynthetic photon flux density was fixed at 1,800 μ mol m −2 s −1 , with a flow rate of 500 μ mol s −1 . A 6400-01 CO 2 injector attached to the main system was used to control the [CO 2 ] in the cuvette. P N at Rice sampling and biomass measurements. The rice plants were sampled at the heading stage and at grain maturity. Six hills per plot were randomly selected and destructively sampled. The samples were separated into green and senescent leaves, stems (including leaf sheaths), and panicles.
All the plant parts were oven-dried at 80 °C to constant weight before being weighed. The N content in plant tissue was measured using an elemental analyzer (PE 2400, Series II CHNS/O, US). The N uptake was determined by multiplying the N content by the biomass of the panicles, stems and leaves.
The grain yield and yield components were measured according to Zhu et al. 20 . At the maturity stage, the grain yield component characteristics (i.e., panicles per m 2 , spikelet number per panicle, filled spikelet percentage and individual grain weight) were tested using six hills of rice. In addition, a 1.5 m 2 area of rice was harvested at ground level and separated into straw and grain components. The collected grains (seeds) were soaked in 1.00 specific gravity tap water, and the number of sunken and floated grains were counted to determine the filled spikelet percentage. The dry weight of the ripened (sunken) grains was measured after they were oven dried at 80 °C for 72 h. The weight per grain and grain yield were expressed by incorporating a 14% moisture content basis 8 .
Statistical analysis. The experimental design was a split plot arranged within a randomized complete block with 3 replications (three rectangular paddy fields). Using the software Statistical Package for the Social Sciences 19.0 (SPSS Inc., Chicago, USA), we first performed an analysis of variance for the main factors of [CO 2 ] and year on the yield and its components, including biomass and plant height, as shown in Table 1 and Fig. 1. We also performed an analysis of variance for the main factors of [CO 2 ] and stage on N uptake, illustrated in Fig. 3. In addition, post hoc comparisons were performed to detect the effects of [CO 2 ] on the gas-exchange parameters, as shown in Fig. 2.