Variation of evapotranspiration in different spatial scales for solar greenhouse tomato and its controlling meteorological factors

Abstract: Knowledge of multi-scale evapotranspiration (ET) is the basis of increasing agricultural water use efficiency and realizing optimal water resources allocation. However, there is little information about the characteristics of multi-scale ET and the difference in the quantitative response of leaf and plant transpiration and farm-scale ET to the meteorological factors, especially for the solar greenhouse, which limits our understanding of agricultural water and energy cycles. In this study, theET at different scales was investigated in solar greenhouse tomato and their main influential factors were analyzed. Experiments were carried out in a typical solar greenhouse to investigate variation in leaf and plant transpiration and farmland ET of drip-irrigated tomato under different water conditions in 2 continuous seasons. Field trials were conducted during the growing season from March to June, 2015 and 2016 at Xinxiang Comprehensive Experimental Station, Chinese Academy of Agricultural Sciences (35°9′N, 113°47′E, and an altitude of 78.7 m), located in Xinxiang, Henan province. Tomato was used in the experiment on a plot of 8.8 m2 (8.0 m long by 1.1 m wide) with 2 rows and 50 plants. There are 4 replications for each treatment. Tomato seedlings were transplanted on March 8th, 2015 and March 9th, 2016 respectively and irrigated with a drip irrigation system. Irrigation amount was determined based on the accumulated evaporation in 20 cm pan, and 2 treatments were designed with full irrigation and deficit irrigation. Leaf transpiration was measured with a photosynthesis system, plant transpiration (expressed as sap flow rate) measured by a sap flow meter and farm-scale ET measured by a weighing lysimeter. Besides, solar radiation, net radiation, relative humidity and air temperature inside the solar greenhouse were constantly monitored with an automatic weather station at 2.0 m height above ground level. And then the difference of quantitative response of leaf transpiration, plant transpiration and farm-scale ET to the meteorological factors was performed using the path analysis method. Results showed that leaf transpiration rate and stomatal conductance gradually increased with the increase in solar radiation, and reached their maximum values at 10:00-14:00, and then rapidly decreased with solar radiation. Difference in leaf transpiration rate between full irrigation and deficit irrigation appeared at 54-58 days after transplanting. Daily average of leaf transpiration rate in the full irrigation was 10.8% and 14.7% higher than that in deficit irrigation, respectively. Tomato sap flow rate changed with weather conditions, and the difference in plant transpiration between the full irrigation and deficit irrigation was the largest in the sunny day, and the smallest in rainy days, and the plant transpiration was related to solar radiation but lagged behind 1 hour. Daily average of leaf transpiration rate in the full irrigation was 54.8% and 41.2% higher than that in deficit irrigation, respectively in the 2 years. The daily farm-scale ET in the full irrigation and deficit irrigation ranged between 0.32-6.65 mm/d and 0.15-5.91 mm/d, respectively, over the whole growth stage, and the farm-scale evapotranspiration reached the maximum in the full fruit period, accounting for 31.7%-34.7% of the total ET. The path analysis of evapotranspiration at different scales and its controlling meteorological factors indicated that leaf transpiration rate, plant transpiration and farm-scale evapotranspiration could be characterized by the net solar radiation, and the vapor pressure deficit also had significant influence on the scale of individual plant and farmland scale. The influence of wind speed should be considered in estimating plant transpiration rate and farm-scale evapotranspiration. Besides, adding leaf temperature could significantly improve the estimation accuracy of plant transpiration and farm-scale evapotranspiration under the condition of deficit irrigation. This paper provides valuable information for constructing conversion methods and theoretic model at various spatio-temporal scales.