Response of wheat canopy CO2 and water gas-exchange to soil water content under ambient and elevated CO2

The nature of the interaction between drought and elevated CO2 partial pressure (pC(a)) is critically important for the effects of global change on crops. Some crop models assume that the relative responses of transpiration and photosynthesis to soil water deficit are unaltered by elevated pC(a), while others predict decreased sensitivity to drought at elevated pCa. These assumptions were tested by measuring canopy photosynthesis and transpiration in spring wheat (cv. Minaret) stands grown in boxes with 100 L rooting volume. Plants were grown under controlled environments with constant light (300 mu mol m(-2) s(-1)) at ambient (36 Pa) or elevated (68 Pa) PCa and were well watered throughout growth or had a controlled decline in soil water starting at ear emergence. Drought decreased final aboveground biomass (-15%) and grain yield (-19%) while elevated pC(a) increased biomass (+24%) and grain yield (+29%) and there was no significant interaction. Elevated PCa increased canopy photosynthesis by 15% on average for both water regimes and increased dark respiration per unit ground area in well-watered plants, but not drought-grown ones. Canopy transpiration and photosynthesis were decreased in drought-grown plants relative to well-watered plants after about 20-25 days from the start of the drought. Elevated pC(a) decreased transpiration only slightly during drought, but canopy photosynthesis continued to be stimulated so that net growth per unit water transpired increased by 21%. The effect of drought on canopy photosynthesis was not the consequence of a loss of photosynthetic capacity initially, as photosynthesis continued to be stimulated proportionately by a fixed increase in irradiance. Drought began to decrease canopy transpiration below a relative plant-available soil water content of 0.6 and canopy photosynthesis and growth below 0.4. The shape of these responses were unaffected by pC(a), supporting the simple assumption used in some models that they are independent of pC(a).