Evapotranspiration and crop-water relations in a peach orchard
Résumé
Transpiration (T) and several water stress indicators were followed in a Prunus persica orchard in central Portugal, with the aim of: (i) quantifying the water consumption of the orchard; and (ii) analysing the usefulness of the variables observed for irrigation scheduling purposes. The experiments were conducted 50 km east from Lisbon during June and July, 1994. A 4-ha orchard planted on a sandy soil was used. Three treatments were considered: A - (well irrigated control, 2.88 ha), B - (drought + normal ambient atmosphere, 0.5 ha), C - (drought + elevated air humidity, 0.02 ha). The large dimensions of plot A allowed micrometeorological measurements. Drought treatments (B, C) consisted of 2 successive periods of complete water shortage. Evapotranspiration (ET) was measured with the eddy covariance method in plot A and the results were compared with the sum of T and soil evaporation (Es) measured by means of a sap-flow method and micro-lysimeters, respectively. Since both methods agreed reasonably well, their results were assumed equivalent. Sensible heat (H) and latent heat (LE) fluxes represented 33% and 39% of net radiation (Rn) during daytime periods; the remaining 28% was heat flux to the soil (G). The relative average contributions of T and Es to total ET, for 24-h periods, were 82% and 18%, respectively. The crop coefficient (Kc) slightly decreased between 0.6 and 0.4 for the 2 months. The adequacy of water stress indicators for irrigation scheduling is discussed using the relationships between stomatal conductance (g s) and predawn leaf water potential (psi p) as well as between relative transpiration (RT, T on stressed pot divided by T on irrigated plot) and psi p or relative daily trunk shrinkage (RDTS, daily trunk shrinkage on stressed plot divided by the correspondent value on irrigated plot). A threshold value of 0.45 MPa for psi p corresponded to a 50% reduction in g s, a 35% reduction in T, and a RDTS of 2.