Variation in temperate cereals in rainfed environments III. Water use and water-use efficiency

Abstract

In rainfed environments in southern Australia the yield of both grain and biomass is greater in barley than it is for bread wheat, durum wheat, triticale and oats. Barley also has a faster leaf area development and biomass accumulation than the other species. To establish reasons for the greater biomass of barley in environments where soil water is limited, variation in the following were evaluated in field experiments: (i) water use or evapotranspiration (ET); (ii) water-use efficiency (WUE), i.e. the ratio of of aboveground biomass (AGBM) to (ET); (iii) transpiration efficiency (TE), of either biomass production, i.e. the ratio of AGBM to transpiration ( T), or of leaf gas exchange, estimated here as the proportional variation in intercellular to external CO 2 concentration using carbon isotope discrimination (Δ) as an integrated measure; and (iv) evaporation of water from the soil surface ( E s) relative to T. Total ET in barley was not greater than in the other species. Variation in ET was largely accounted for by variation in anthesis date; long-duration cultivars used more water than short-duration cultivars. ET averaged 283 mm over 3 sites and for each day delay in anthesis water use increased by about 1 mm. However, this was offset by a decline in WUE as anthesis became later. WUE decreased by 0.67 kg ha −1 mm −1 for each day delay in anthesis whereas TE declined by 0.92 kg ha −1 mm −1. The average WUE at two sites where root mass was measured was 32 kg ha −1 mm −1 whereas TE averaged 48 kg ha −1 mm −1. When roots were included values increased to 42 kg ha −1 mm −1 for WUE and 62 kg ha −1 mm −1 for TE. Barley, cv. O'Connor, had the greatest WUE and TE at all sites. Evaporation of water from the soil surface ( E s) was also smaller in barley. The separation of ET into T and E s was about 20 mm smaller in barley than in the other species at all sites. Variation in leaf gas exchange among cultivars, measured as variation in Δ, was positively related to measured values of TE, not negatively as might be expected. Including root mass in the calculation of TE reduced the slope of the relationship somewhat but it remained positive and significant. Standardizing TE to correct for differences in vapour pressure of the air over the life of the crop also reduced the slope of the relationship between Δ and TE, but it remained positive. Variation in TE was also positively related to an index composed of height and anthesis date suggesting that TE was greatest in early-flowering cultivars that elongated first and this may be related to the maintenance of a cooler canopy and a greater boundary layer conductance. It is concluded that the high yield achieved by barley in this environment was due to faster leaf area growth and earlier flowering. These factors ensured that water loss from the soil surface was minimized and that growth was completed before the rapid rise in temperature and VPD.