15.3.1 Water use efficiency of crops

Printer-friendly version

Water use efficiency is defined here as yield of plant product (tonnes of wheat grain, Y) per unit of crop water use (megalitres of water lost by evapotranspiration, ET), and is important in all areas of plant production.

Water use efficiency (Y/ET) is the outcome of an entire suite of plant and environmental processes operating over the life of a crop to determine both Y and ET. Consequently, biomass production per unit ET, has been used extensively as an interim measure of water use efficiency.

ET comprises non-productive evaporation (E) of water from the soil surface and productive transpiration (T) of soil-stored water by the plant. Evaporation of free water from leaf surfaces adds to non-productive evaporation (interception evaporation). The basic equation describing ET distinguishes productive and non-productive evaporation:

In contrast to ratios of water use efficiency based on ET, the ratios W/T and Y/T (interim and ultimate transpirational water use efficiency, respectively) do not involve E (Equation 15.2) and so serve to focus attention on physiological aspects of water use efficiency. Note, however, that non-productive evaporation from leaf and soil surfaces in a wheat crop ac-counted for 49% of ET at Wagga Wagga (Leuning et al. 1994). Even higher values (60%) have been reported for a sparse crop of barley growing in Syria.

According to these definitions, maximum water use efficiency is achieved by maximising both T as a proportion of ET (because water lost as evaporation from soil is non-productive; Leuning et al. 1994) and transpirational water use (TWUE) (because maximal TWUE requires maximal yield per unit of water transpired; Tanner and Sinclair 1983; Richards 1991).

Interrelations between growth, yield and transpiration were studied extensively by Bierhuizen and Slatyer (1965). They concluded from assimilation and transpiration characteristics of cotton leaves that plant growth is directly proportional to transpirational water use, but inversely dependent on atmospheric vapour pressure deficit. Their work was developed by Tanner (1981) and Tanner and Sinclair (1983), who described daily biomass production (W) in terms of daily transpiration (T) and mean daytime vapour pressure deficit (D) where:


Figure 15.10 Transpiration efficiency of wheat at Wagga Wagga, New South Wales, declines curvilinearly as evaporative condition intensify. Daily evaporation from a Class A pan increases from winter, through spring to summer (indicated by month of observation). (Based on Richards 1991, plus unpublished data of Warren and Lill, cited by Fischer 1979).

k accounts for several factors, including the biochemical pathway for photosynthesis, which varies between species (e.g. C3 v. C4), and yield of hexose per unit of carbon assimilated. Production value, PV, accounts for differences in biomass production per unit of hexose, and also depends on the species. A consequence of the inverse relationship between growth rate and D, incorporated in Equation 15.3, is seen in Figure 15.10 where data of Warren and Lill (cited by Fischer 1979) demonstrate that transpiration efficiency (W/T) decreases sharply with increasing pan evaporation (where vapour pressure deficit would feature prominently in evaporative conditions).

Wilson and Jamieson (1985) successfully applied Equation 15.3 to wheat grown in three seasons in New Zealand. The resultant linear relationship between biomass (W) and D-corrected transpiration (T/D) is shown in Figure 15.11. HI was ignored as a source of variation in Y. HI has attracted considerable attention as a means of crop improvement (Section 6.3). In determinate crops, HI increases in ac-cordance with water use (T) after flowering. Post-flowering water stress will increase use of pre-anthesis assimilate in seed yield (Hall et al. 1989; Sadras and Connor 1991; Section 15.3.3).

A simple rearrangement of Equation 15.3 shows that PV-adjusted growth, W/PV, is directly proportional to D-adjusted transpiration, T/D. That is:


Sadras and Connor (1991) related PV-adjusted HI of wheat and sunflower to the D-adjusted ratio, q, post-flowering T:(post-flowering T)/(pre- + post-flowering T). They demonstrated that variability in HI of sunflower is satisfactorily explained on the basis of the large decrease in PV after flowering (due to the large cost of synthesising high-energy, oil-rich seed), and a simple, linear decline in the use of pre-anthesis assimilate in yield, dependent on q. Sadras and Connor (1991) related PV-adjusted HI (HIPV) to q by the relationship




Figure 15.11 Above-ground biomass produced by wheat in these situations in New Zealand increases linearly as transpiration per unit vapour pressure deficit increases (Based on Wilson and Jamieson 1985)

on the basis that water-limited PV-adjusted yield comprises all post-flowering biomass production, supplemented by a quantity, D, of remobilised pre-flowering biomass, where D/Y = a - bq. Parameter a quantifies the potential contribution of pre-anthesis assimilates to grain biomass, and 1/(1 – a + b) the potential HI. Both are genotype dependent, with a varying between 0.81 and 0.92, and b varying between 1.19 and 1.86. The resultant term 1/(1 – a + b) varied between 0.52 and 0.72 for three different cultivars of sunflower. The ultimate balance between a and bq therefore indicates the contribution of pre-anthesis assimilate to yield, Y.

Application of Equation 15.5 (Figure 15.12) demonstrates the ability of the Sadras–Connor approach to define the role of pre-anthesis assimilate in HI, yield and water use efficiency. An apparent drawback to its wider application is the need to estimate coefficients a and b by statistical curve-fitting procedures. Definition of pre- and post-flowering periods is an added problem in indeterminate species.

Relationships between PV-corrected biomass and D-cor-rected transpiration therefore provide essential information for estimation of yield. In terms of the framework adopted here (Monteith 1977), biomass production depends on total D-adjusted T (Equations 15.3, 15.4) and Figure 15.12 demonstrates that yield, at least in determinate crops, depends on the proportion of D-adjusted T realised in the pre- and post-flowering periods, but modified by contributions of pre-flowering assimilate.


Figure 15.12 Relationship between nexose-based harvest index (HI) and the proportion of post-flowering D-adjusted transpiration in (a) wheat and (b) three different cultivars of sunflower. Curves were generated by application of Equation 15.4. (Based on Sadras and Connor 1991)

Physiological and climatological features of water use efficiency outlined by Bierhuizen and Slatyer (1965) remain a cornerstone for present understanding of the TWUE component of water use efficiency. Many developments need sophisticated techniques that allow precise and direct measures of processes and rates in both field and laboratory. Estimates of canopy conductance via aerodynamic methods and direct measures of transpiration via sap flux measurements (heat-pulse techniques) are thus finding widespread application in field analysis of water use efficiency.