Measurement of Cotton Transpiration

Abstract

There are a few field methods available to directly measure water evapotranspiration ( ET ) along with its two components, evaporation from the soil ( E ) and from the crop ( T ). One such technique that measures T , uses sensors to calculate the sap flow ( F ) of water through the plant stem and is based on the conservation of mass and energy, i.e., the stem heat balance method. This instrument consists of a flexible heater that is wrapped around the plant stem with temperature sensors to measure the difference in temperature of F below and above the heater. This is a null method, where all inputs and outputs are known and the calculated F is a direct measure of T . This method has been used to measure <i>T</i> in a variety of crops, including cotton, grapes, olive trees, soybean, ornamental and horticultural crops. A new version of the EXO-Skin TM is the Stem Gauge Dual Channel Design (SGDC TM ), which was commercially introduced and had a radically new design resulting in a different energy balance, compared to the original design, which needed experimental verification. An initial evaluation was done with potted cotton ( Gossypium hirsutum , L.) plants in a greenhouse experiment showing that values of cotton- T measured with the new sensor were accurate; however, this comparison was limited to daily T < 2 mm/d. Thus, our objective was to expand the initial evaluation of the new sensor under field conditions and for daily values of cotton- T in the 2 - 7 mm/d range, representative of the semiarid Texas High Plains. For this purpose, cotton was planted on 12 June 2017 on a 1000 m2 plot in a soil classified in the Amarillo series at the facilities of the USDA-ARS, Lubbock, TX. For a period of 15 days, 2 to 16 Sep 2017, we measured hourly cotton- T with the new sensors and with portable growth chambers (0.75 m × 1 m cross-section, and 1 m height) where water vapor flux was measured at a 10 Hz frequency using an infrared gas analyzer. We used three chambers and, in each chamber, the new sensors were installed on four cotton plants. We used linear regression analysis to compare hourly and daily values of cotton- T measured with the sap flow gauges against T measured by the chambers. Using a t-test ( p < 0.001) we tested if the slope of the line was significantly different than 1 and if the intercept was significantly different than 0. Pooling all data yielded an almost 1:1 relation between values for a daily transpiration range from 2 to 7 mm/d. We concluded that the new sensor provides a robust and direct measure of hourly and daily cotton- T for a wide range of environmental conditions.