Yield and redesigned photosynthesis

On the basis of current evidence, it would appear that achieving a 40% to 50% increase in yield is probably beyond the capacity of rice with its current photosynthetic pathway. Such a large increase in potential yield will certainly require an RCF of 3.3 g DM MJ1 (the value for maize). As a consequence, achieving yields beyond 12.5 t ha 1 in the next 50 years could require the introduction of the C4 photosynthetic pathway to rice (Ku et al 1999). Increases in atmospheric CO, (Schimel et al 1996) could produce a similar yield increase by about 2050. However, several potential obstacles could prevent crops with very high photosynthetic rates from using the extra photosynthate to increase grain yield without reducing grain quality.

The sink size in rice is proportional to the number of spikelets. At harvest, a large fraction of the spikelets is filled and is known as grain: a smaller fraction remains unfilled or partially filled. Unlike in wheat or barley, grain weight in a rice cultivar is almost constant: the desired range is 23-27 mg. Consequently, yield improvements must result from an increase in spikelet number and that depends on the nitrogen status of the crop (Fig. 3). However, the critical nitrogen content of C4 crops is less than that of C, crops growing under current CO, conditions.

Several questions arise if it is assumed that a C4 rice were to have a critical nitrogen concentration typical of other C4 crops. Would the number of spikelets per unit ground area be smaller in a C4 rice crop than in a C, rice crop? Would this reduction lead to a limitation of yield potential? Would the C4 syndrome reduce the protein content of the grain? To answer these questions, it is necessary to make some calculations. A grain yield of 15 t ha"1 (equivalent to 13.3 t ha"1 dry weight) would require 57.826 grains m each weighing 23 mg. If the filled percentage were 80%, a total of 72,282 spikelets m : would be needed to produce such a yield. The maximum number of juvenile spikelets (glumous flowers at the late differentiation stage of the panicle) observed per square meter in the new plant type in the experiment described above was 116,325. If approximately 65% of these remained after floret abortion, 76% of them would have to be filled to produce the yield required. Therefore, the sink size for a crop with a conventional photosynthetic pathway would be adequate. However, one cannot assume that a rice crop having a C4 pathway would have the same number of juvenile spikelets because that number depends on plant N status. In an experiment here at IRRI, in which only a moderate amount of N was supplied (120 kg ha1), the number of juvenile spikelets per square meter was only 54,000.

Horie et al (1997) derived a relationship between spikelet number and the concentration of N at flowering (Fig. 3). The number of spikelets a C4 rice crop might produce is calculated by assuming that the shoot weight of a C4 crop at flowering would be 12.4 t ha"1 (a value typical for high-yielding C, crops). The Greenwood model suggests that the N content would be approximately 14.5 g N m:. The number of spikelets for a C4 rice crop was calculated to be 60,000 m2. To obtain a 151 ha"' yield, 96% of the spikelets would have to be filled, a remarkably high filling percentage.

The relatively lower intrinsic N content of a C4 crop may mean that the nitrogen reservoir in its shoots would be inadequate to support the nitrogen demand of the developing grains. To investigate this possibility, the simple model of Greenwood et al (1990) linking critical nitrogen concentrations in the aboveground biomass (%N) and crop dry weight (W) was used. For both C,and C4 crops in the vegetative stage, the model equation is

Table 3. Crop N contents for two typical grain N content scenarios: (1) high-yielding rice (1.4%) and (2) low-yielding rice (0.9%); straw N is assumed to be constant (0.8%) as is the total crop yield (151 ha1).




(kg ha"1


186 (1.4%


106 (0.8%

, N)


120 (0.9%


106 (0.8%

i N)


Greenwood found that for both C, and C4 crops the model accounted for 86% of the variance when b = 0.5, a = 5.7 (C,), and a - 4.1 (C4). Using this model, assuming a grain yield of 15 t ha-1, the total nitrogen content in the shoots of a C, and a C4 crop was calculated to be 292 kg ha~' for the C, crop and 211 kg ha-1 for the C4 crop. Next, the N contents were calculated for two grain N content scenarios: (1) typical of high-yielding rice (1.4%) and (2) typical of low-yielding rice (0.9%) (Table 3). By comparing the N contents in Table 3 with that predicted for a C4 rice containing the critical nitrogen concentration, it can be seen that such a C4 rice would contain insufficient N to support a grain yield of 15 t ha-1 at a grain N content as low as 0.9% N. The grain N concentration of a 15 t ha-1 crop would have to be approximately 0.8%, a very low value, assuming that straw N concentration could not be reduced.

A C, crop growing under elevated C02 conditions with suppressed photorespiration may resemble a C4 crop in terms of its N content (Nakano et al 1997) and yield. Several experiments have been conducted to investigate such effects. Many such experiments have been carried out in pots or enclosures of limited size and so the results are somewhat difficult to extrapolate to normal field conditions, especially for ceiling-yield scenarios. Ziska et al (1997), using open-top field enclosures, observed increases in grain yield with elevated C02, but the control yields were almost 50% lower than the 10 t ha-1 obtained outside the enclosures. Nonetheless, they reported that protein contents in the grain declined significantly with increasing CO:. Yoshida (1973) elevated C02 to approximately 900 ppmv in small open-top field enclosures (0.36 nr) during different growth stages. He reported a yield of 13.3 t ha"1 and a filling percentage of 87.6% when CO: was elevated for 30 d before flowering. Baker et al (1992a, 1994). using outdoor sunlit environment chambers, reported no effect of elevated C02 on grain yield and in other experiments a significant increase in yield (Baker et al 1990, 1992b, 1997).

If a C4 rice plant were to be developed and atmospheric CO, concentrations continued to rise, so that C, and C4 rice photosynthesized at the same rate, the C4 rice could use significantly less water in transpiration than the C, rice. For example, if CO: levels increased to 510 ppm, the stomatal conductance of C4 rice could fall by more than 50% while maintaining the same rate of photosynthesis per unit leaf area as would occur at current concentrations. The sunlit leaves of a C4 rice would be unable to shed as much heat as a C, rice under such conditions and it can be estimated that close to midday their temperature would rise above that of the C, by about 2 °C (Sage, this volume) and their rate of transpiration would be about 50% lower (Sheehy et al 1998b). In a future world of elevated CO:, the greatest advantage of a C4 rice may well be its drought tolerance in upland and rainfed ecosystems. The adverse effects on yield of temperature increases that may result from global climate change are documented elsewhere (Matthews et al 1995;.

Was this article helpful?

0 0

Post a comment