Table Of ContentsNext Page

Improving harvest index is an effective way to increase crop water use efficiency

Jianhua Zhang1 and Jianchang Yang2

1 Hong Kong Baptist University, Hong Kong, China www. Email
Yangzhou University, College of Agriculture, Yangzhou, Jiangsu, China Email


Water use efficiency (WUE), if defined as the biomass accumulation over water consumed, may be a highly inherited characteristics of a specific genotype. In practice, WUE can also be enhanced by less irrigation, particularly via stomatal regulation. However, such enhancement is largely a trade-off between lower biomass production and higher WUE. We have presented a case here that WUE may be enhanced through an improved harvest index. Harvest index has been shown as a variable factor in crop production, especially in cases where whole plant senescence of rice and wheat is unfavourably delayed. Such delayed senescence can delay the remobilisation of pre-stored carbon reserves in the straw and results in lower harvest index. A controlled soil drying, i.e. moderate drying such that overnight rehydration of plants is still possible, should enhance whole plant senescence and therefore improve the remobilisation of pre-stored carbon reserve. The gains from the improved harvest may outweigh any possible biomass loss due to shortened photosynthetic period in grain filling, such as the cases with high N nutrition, lodging-resistant cultivars that stay green for too long, and hybrid cultivars with excessive heterosis.

Media summary

Controlled soil drying at late grain filling can make better use of pre-stored carbon reserves and enhance harvest index and water use efficiency for yield.

Key Words

Water use efficiency, harvest index, grain filling, soil drying, wheat, rice.

Can WUE be increased?

Plants unavoidably lose a large quantity of water when they open their stomata for CO2 uptake in a less saturated air. Vapour diffuses out from the substomatal cavity into the air while CO2 goes in the opposite way. Mathematical modelling of these two opposing diffusion processes has shown that water use efficiency (WUE) is largely a function of the CO2 and vapour concentration gradients between the inside and outside of the leaf (Jones, 1992). These two opposing fluxes are regulated by stomata. Therefore the stomatal behaviour will determine the WUE of a particular species or cultivar.

Indeed WUE has been shown a conservative parameter with high inheritance. It is well known that C4 plants have higher WUE than C3 plants. Within C3 plants, many reports have shown that genotypes can be selected for higher WUE according to their carbon isotope discrimination, a function of the CO2 concentration gradient between the inside and outside of the leaf (e.g. Craufurd et al. 1991, Ehdaie et al. 1991).

Why is WUE a conservative characteristics of plants? Jones (1992) explained that stomata operate in such a way that a reduction in carbon assimilation during stomatal closing or partial closing is only responsible for a small proportion of the total assimilation. It is well known, however, that plants growing under water-limited conditions will have higher WUE. It has been predicted that plants generally have the capability to optimise their water use in the short term and maximize their chance of survival over a drought in a longer term. In a world where rainfall is unpredictable, long-term regulation means that plants must be able to 'detect' the soil drying and then 'respond' to it by regulating their stomata. Such a mechanism may be termed as a feed-forward mechanism, since the decreasing availability of water in the soil does not cause any significant water deficit (Davies and Zhang 1991). Jones (1992) concluded that such a responsive pattern of stomatal behaviour should be the best pattern for both plant survival and carbohydrate production.

Such optimised stomatal behaviour means that plants should have a higher WUE with less water supply. If we hold the conventional view that plant biomass production is linearly coupled with the amount of water used, it is not a surprise that higher WUE is a trade-off for lower biomass production. In agriculture, many ways of conserving water have been investigated and techniques such as partial irrigation, deficit irrigation or drip irrigation have shown that WUE can be enhanced. In general, these techniques are a trade-off: a lower yield for a higher WUE.

Is it possible to increase WUE without much reduction of yield? There are many examples where grain yield, a large proportion of the total biomass, shows a negative parabolic relationship with the amount of irrigation. This suggests that when water supply is sufficient, excessive vegetative growth may lead to less root activity, unhealthy canopy structure and a lower harvest index. That means that high biomass production, supported by high water supply, will not lead to high WUE if defined as the grain production per unit amount of water irrigated. Therefore the goal is to increase WUE of grain yield by limiting water supply to increase harvest index. Our recent research has shown that in some irrigated situations, grain yield can be improved while reducing the amount of water applied to the crop (Yang et al. 2000; 2001a; 2002), mainly via improved harvest index which has been shown as a key component to improve WUE of yield (e.g. Ehdaie and Waines 1993).

Remobilization of pre-stored carbon, the variable fraction in grain filling

Grain filling is the final stage of growth in cereals where fertilized ovaries develop into caryopses. At this stage, about 40 to 50 percent of total biomass is deposited into the grains. Monocarpic plants such as rice and wheat need the initiation of whole plant senescence so that stored carbohydrates in stems and leaf sheaths can be remobilised and transferred to their grains. Normally when these crops are grown in, high-input systems, pre-stored food contributes 1/4 - 1/3 to the final weight of a grain. Delayed whole plant senescence, leading to poorly filled grains and unused carbohydrate in straws, is a new problem increasingly recognized in rice and wheat production in recent years. Slow grain filling may often be associated with delayed whole plant senescence. Although farmers can choose cultivars of early-maturation, there are still some situations that have made delayed senescence a serious problem that needs attention:

A).Heavy use of nitrogen fertilizers is well known to lead to a delayed senescence and in worst cases canopy lodging. This is possibly related to the intensive nature of today’s irrigated agriculture and the need to use less arable land to feed increasingly more people.

B).Selection of lodging-resistant cultivars to cope with the lodging problem has led to another problem in some cases, i.e. stems are short and strong but stored carbohydrate is poorly used because the plants may stay “green” for too long, particularly in some cases with short-grain rice cultivars.

C).Introduction of hybrid rice has been a fantastic success in China. Utilization of heterosis, e.g. a hybrid between the two subspecies (or ecotypes) of rice, the Japonica and Indica rices, has however also met the problem of delayed senescence. Grains are poorly filled or un-filled (Yuan 1996). Such hybrid genotypes seem too vigorous in terms of keeping “young”.

Cases described above may be defined as unfavourably delayed senescence, which means that little or no gain is obtained from the extended grain-filling period. We should however, distinguish this situation from that in favourable conditions in which early senescence should be avoided because it reduces the photosynthesis during the grain-filling period and therefore reduces grain weight (e.g. Zhang et al. 1998).

A controlled soil drying promotes whole plant senescence at grain filling

Our recent experience with field-grown wheat has found that a soil drying during the grain-filling period can enhance early senescence (Zhang et al. 1998). We found that while the grain filling period was shortened by 10 days (from 41 to 31 days) in unwatered (during this period) plots, a faster rate of grain-filling and enhanced mobilization of stored carbohydrate minimized the effect on yield. It seems possible that a controlled soil drying during the later stages of grain-filling may promote whole plant senescence, leading to increased re-translocation from the stem and sheath of pre-stored carbon reserves.

A controlled soil-drying means that crops should not be soil-dried to a degree that over-night rehydration cannot be completed and photosynthesis is too severely inhibited. It should be emphasised that the soil drying should be at the later stage of grain filling because early development of embryos (at the rapid cell division stage), i.e. the “grain-setting” stage, is very susceptible to water deficit (Boyle et al. 1991).

With highly lodging-resistant rice cultivars, our results (Yang et al. 2001a) showed that if a water deficit during grain filling of rice is controlled properly so that plants can rehydrate overnight, photosynthesis should not be severely inhibited. A benefit from such a water deficit is that it can enhance plant senescence and lead to a fast and better remobilisation of pre-stored carbon from vegetative tissues to the grains (Table 1). The early senescence induced by water deficit does not necessarily reduce grain yield even when plants are grown under normal N conditions. Furthermore, in cases where plant senescence is unfavourably delayed such as by heavy use of nitrogen, the gain from the enhanced remobilisation and accelerated grain filling rate may outweigh the loss of photosynthesis and shortened grain-filling period and increase the grain yield and harvest index.

Table 1. Remobilisation of pre-stored assimilates from stems of rice subjected to various N and soil moisture treatments.


Water deficit treatment

Nitrogen applied

Remobilised C reserve%

Contribution to grain

NSC Residue mg g-1 DW

Total dry matter
g m-2

Harvest index

g m-2

Wuyujing 3



47.5 c

14.4 c

142.3 b

1707.2 a

0.47 c

802.4 b




24.5 d

7.6 d

218.5 a

1741.2 a

0.41 d

713.9 c




74.6 a

28.5 a

64.5 d

1537.9 b

0.52 a

799.7 b




61.2 b

21.5 b

103.5 c

1716.6 a

0.50 b

858.3 a


Yangdao 6



58.9 b

11.3 c

97.8 b

1788.0 a

0.51 c

911.9 ab




46.3 c

5.3 d

151.6 a

1743.6 a

0.47 d

819.5 c




82.7 a

27.8 a

41.2 d

1578.8 b

0.56 a

884.1 b




65.9 b

17.9 b

85.3 c

1768.3 a

0.53 b

937.2 a

The two cultivars used were highly lodging-resistant and stay green when grains are mature. NN and HN indicate normal and high levels of nitrogen application at heading time. WW and WS are well-watered and water-deficit treatments during grain filling. Values are means of 20 plants. Letters indicate statistical significance at P0.05 within the same cultivar. NSC stands for nonstructural carbohydrate in stems.

When senescence is unfavourably delayed, rice will show a prolonged and slow grain filling, e.g. under high N condition (Fig. 1). Controlled soil drying increases the grain-filling rate and shortens the grain-filling period. The increased rate and shortened period are especially remarkable at high N condition (Figs. 1C and 1D). Our results (Yang et al. 2001b) with lodging-resistant cultivars also showed that the final grain weight was not significantly different between well-watered and water-stressed treatments when normal amount of

Fig. 1 Grain filling process (A and B) and grain filling rate (C and D) of the japonica cultivar Wuyujing 3 (A and C) and indica cultivar Yangdao 6 (B and D) subjected to various nitrogen and soil moisture treatments. The treatments are: normal N (NN) +well watered (WW) (●), NN + water stressed (WS) (), high N (HN) + WW (■), and HN + WS (□). Grain filling rate was calculated according to Richards equation. Arrows in the figure indicate the start of withholding water. Vertical bars in the figure A and B represent SE of the mean (n=2) where these exceed the size of the symbol.

N was applied. However, final grain weight was significantly increased under water-stressed plus high N treatment, implying that the gain from accelerated grain-filling rate outweighed the possible loss of photosynthesis as a result of a shortened grain-filling period when subjected to water stress during grain filling.

Similar results were also obtained with hybrid rice which shows very strong heterosis but a slow grain filling as a result of delayed whole plant senescence (Yang et al. 2002). The stronger the heterosis, e.g. the hybrid between japonica and indica subspecies, the higher the harvest index can be improved by a controlled soil drying at the grain filling stage. The grain filling process and rate of the hybrid cultivars were substantially enhanced by the controlled soil drying. Grain yield was actually improved, rather than reduced, by moderate soil drying during the grain filling stage in cases where heterosis is very strong, e.g. the japonica and indica hybrids.

Heterosis of hybrid rice is usually a function of the genetic relations of the two parents. Hybrids from japonica and indica rice show stronger heterosis and also tend to stay green longer in later grain filling than the indica/indica rice. When the whole plant stays green, its non-structural carbon (NSC), mainly the starch, in the culm also stays high for long. We found that concentration of NSC in the culm and sheath during grain filling was very different between indica/indica and japonica/indica hybrid(s) under WW treatments (Yang et al. 2002). NSC in the culm and sheath of the two japonica/indica hybrids showed a “V” shape pattern, i.e. initially decreasing from 7 to 21 days after anthesis, but increasing thereafter. NSC concentrations at maturity for both japonica/indica hybrids were nearly the same as at anthesis. For the indica/indica hybrid, NSC in the culm and sheath decreased sharply from 7 to 32 days after anthesis and slowly thereafter. Water deficits substantially reduced NSC in the culm and sheath of all hybrids. The more severe the water deficit, the more the NSC was reduced. Under mild and severe deficit treatments the patterns of the NSC were similar for both indica/indica and japonica/indica hybrid(s).

Wheat grain yield is more vulnerable to shortened grain filling period than rice. Zhang et al. (1998) showed with wheat under reduced irrigation that grain size can be greatly reduced by less irrigation during this period. However, in cases where high N and very strong lodging-resistant cultivars are combined (Yang et al. 2000), moderate soil drying at later grain filling enhances the harvest index substantially with an improved WUE, if defined as the grain yield over the irrigated amount. Soil drying at grain filling stage can greatly shorten the filling period and reduce the gain size and yield. However, in cases where staying green is a problem as a result of heavy use of nitrogen, the moderate soil drying may not necessarily reduce the grain yield of wheat. An extra benefit from such practice is that less water may be required for wheat irrigation under this situation.


We have demonstrated that in three cases where whole plant senescence is unfavourably delayed and harvest index is low, a controlled soil drying at the later grain filling stage can greatly promote grain filling rate and lead to a much enhanced harvest index. In cases where senescence was severely delayed, the gain from a better utilization of pre-stored reserves actually outweighed the possible loss due to a shortened grain filling period.


Boyle, M.G., Boyer, J.S. and Morgan, P.W. (1991) Stem infusion of liquid culture medium prevents reproductive failure of maize at low water potentials. Crop Science, 31, 1246-1252.

Craufurd, P.Q., Austin, R.B., Acevedo, E. and Hall, M.A. (1991) Carbon isotope discrimination and grain-yield in barley. Field Crops Research 27, 301-314.

Davies, W.J. and Zhang, J. (1991) Root signals and the regulation of growth and development of plants in drying soil. Annual Review of Plant Physiology and Plant Molecular Biology 42, 55-76.

Ehdaie, B., Hall, A.E., Farquhar, G.D., Nguyen, H.T. and Waines, J.G. (1991) Water-use efficiency and carbon isotope discrimination in wheat. Crop Science 31, 1282-1288.

Ehdaie, B. and Waines, J.G. (1993) Variation in water-use efficiency and its components in wheat: I. Well-watered pot experiment. Crop Science 33, 294-299.

Jones, H.G. (1992) Plants and Microclimate: A Quantitative Approach to Environmental Plant Physiology, 2nd edn, Cambridge University Press, Cambridge.

Yang, J., Zhang, J., Huang, Z., Zhu, Q. and Wang, L. (2000) Remobilization of carbon reserves is improved by controlled soil-drying during grain filling of wheat. Crop Science 40, 1645-1655.

Yang, J., Zhang, J., Wang, Z., Zhu, Q. and Wang, W. (2001a) Remobilization of carbon reserves in response to water-deficit during grain filling of rice. Field Crops Research 71, 47-55.

Yang, J., Zhang, J., Wang, Z., Zhu, Q. and Wang, W. (2001b) Hormonal changes in the grains of rice subjected to water stress during grain filling. Plant Physiology 127, 315-323.

Yang, J., Zhang, J., Liu, L., Wang, Z. and Zhu, Q. (2002) Carbon remobilization and grain filling in japonica/indica hybrid rice subjected to post-anthesis water deficits. Agronomy Journal 94, 102-109.

Yuan, L. (1996) Strategies of selection for hybrid combination between japonica and indica rice. Hybrid Rice 2, 1-3.

Zhang, J., Sui, X., Li, B., Su, B., Li, J. and Zhou, D. (1998) An improved water-use efficiency for winter wheat grown under reduced irrigation. Field Crops Research 59, 91-98.

Top Of PageNext Page