Using leaf gas exchange to quantify drought in cotton irrigated based on canopy temperature measurements

Plant gas exchange provides a highly sensitive measure of the degree of drought stress to which a crop is exposed. However, equipment costs and time requirements for gas exchange measurements are major obstacles to the use of gas exchange measurements in real-time irrigation scheduling systems. Canopy temperature (T(c)) provides a much easier to acquire indication of crop water deficit that has been used in irrigation scheduling systems, but interpretation of this measurement has proven difficult. Our goal was to test the ability of T(c) to quantify the degree of crop water deficit by comparing T(c) with simultaneous measurements of leaf-level gas exchange parameters, which were viewed as alternative indicators of water deficit. To provide a wide range of plant water deficit conditions for the comparison of T(c) with leaf-level gas exchange parameters, cotton (Gossypium hirsutum L.) was subsurface drip irrigated using T(c) according to the stress-time index method of irrigation scheduling during two growing seasons. Comparisons between T(c) and leaf-level gas exchange were accomplished by measuring T. diurnally with hand-held infrared thermometers and controlling cuvette leaf temperature (T(L)) equal to T(c) and then measuring leaf level net assimilation (A) and stomatal conductance (g) at a photosynthetically active radiation (PAR) level of 1500 mu mol (photons) m(-2) s(-1). In general, as plant water deficit became more severe, leaf level gas exchange tended to decline with rising T(L). However, we found that A and g could vary by more than twofold at a given T(L), indicating that T(c) was not a particularly robust indicator of the degree of drought stress. Furthermore, we found the leaf minus air temperature differential (T(L) - T(a)) and vapor pressure deficit calculated based on leaf temperature (VPD(L)) were better predictors of the degree of drought stress, as indicated by gas exchange parameters, than T(L) alone. Regression of A and g against (T(L) - T(a)) and VPDL indicated that the combination of these two variables accounted for >79% of the variability in A and g. We conclude that the term (T(c) - T(a)) either alone or in combination with VPD should provide a better predictor of the degree of drought stress in cotton than T(c) alone.