SPATIAL VARIATION OF THE ISOTOPIC COMPOSI TION OF WATER (O-18, H-2) IN THE SOIL-PLANT-ATMOSPHERE SYSTEM .2. ASSESSMENT UNDER FIELD CONDITIONS

We have made intensive sampling for isotopic measurements (H-2/H-1 and O-18/O-16) conducted on water in the soil-plant-atmosphere system. Du ring two daily cycles, we have carried out the measurement of the isot opic composition of water in soil, collar, leaf and atmospheric vapor under different climatic conditions (sunny and cloudy days). We want t o identify the processes that alter the isotopic composition of the le af water in the water flow in a maize crop field (station of Bioclimat ology, INRA, Grignon, Yvelines, France). In all cases, the isotopic co mpositions of leaf water show a periodic evolution with a maximum in t he early afternoon and a minimum at the end of the night. This daily e volution is in response to the daily variations observed in relative h umidity. Experimental verifications of the isotopic model of the trans piration process enable one to establish a physical model which shows a satisfactory agreement between the measured isotopic compositions in the leaf water and the model calculations. We modelled the changes in the H-2/H-1 and O-18/O-16 ratios of leaf water in response to the cha nges in the climatic factors (temperature and hygrometric deficit of s aturation of the atmosphere), the H-2/H-1 and O-18/O-16 ratios of the atmospheric water vapor and of the source water (i.e. water from soil) , and the turnover of water in the leaf. At first, this simple physica l model consists of two compartments (a well=mixed leaf water reservoi r and air water vapor). A diffusional boundary layer in the leaf stoma ta separates these two compartments. We try to explain the discrepanci es between the measured and the calculated values by the effects of an incomplete mixing of water in the leaf. Taking into account the diffe rences between the isotopic compositions of the water pools in the lea f enables one to calculate the daily evolution of the volume of leaf w ater marked by the transpiration process. Vein water and apoplasmic wa ter are unfractionated concerning the soil water. We have found no lin ear covariance between the isotopic composition of the leaf water and the leaf water potential in this study. Taking into account the differ ent origins of water supply in the leaf (soil and internal reserves of the leaf) and the amounts of water lost then recovered by the leaf en ables one to explain the hysteresis loop observed between the isotopic composition of the leaf water and leaf water potential. With experime ntal data on the actual evapotranspiration flux, the isotopic model of transpiration allows a precise determination of the nature (transient or stationary) of the isotopic state of the water in the leaf with an increasing accuracy. During the day, the isotopic composition of the leaf water fitted a steady-state model because of the high values of t he turnover of the foliar water. In this case, the value of the kineti c enrichment factor was obtained to be 28.5 parts per thousand (O-18). These values agree well with theoretical predictions in pure diffusiv e conditions in the stomata chamber. Examining the relationship of the two isotopes tracers in the leaf water, following previous investigat ors, the intersection between the steady-state transpiration line and the meteoric line cannot be used to determine the isotopic composition of the source water in the plant in a deltaH-2-deltaO-18 diagram. The re is no fractionation during water uptake by the roots. Then the root s density in the soil and the isotopic profile of the soil water show considerable heterogeneities with depth. The effects of the root syste m and the mechanisms involved for the water transport from the soil la yers to the active roots increase the difficulties of sampling and ana lysis. The isotopic composition of the source water determined at the collar level shows a per-iodic evolution with a minimum in the afterno on and a maximum at dawn. These changes may reflect the use by the pla nt of available water from the most active sites of root absorption in deeper soil layers during transpiration under natural conditions.