Energy budget for tomato plants grown in a greenhouse in northern China

Highlights

• Energy budget and physical and physiological factors governing λET in greenhouse was first investigated.

• Mixed convection dominated most of the study period, followed by pure forced convection.

• High α and Ω for most of days indicated that λET was primary controlled by R_n in greenhouse.

• G_c was an important physiological influence on λET, especially during middle growth period.

Abstract

Research is ongoing to increase our understanding on the mechanisms responsible for the variation in energy fluxes in greenhouses. In this study, a four-year experiment (2016, 2017, 2019, and 2020) was carried out to investigate the energy budget for drip-irrigated tomato plants in a greenhouse, where the latent heat flux (λET) was measured by two weighing lysimeters. Factors that determine energy budget and λET were also investigated. The results indicated that λET was the principal component of net radiation (R_n), accounting for 66.4–71.7%, followed by sensible heat flux (H) and ground soil heat flux (G). The low values (0.25–0.32) of the seasonal mean midday Bowen ratio (β=H/λET) also indicated that seasonal λET was greater than H for well-watered tomato plants. Leaf area index (LAI) strongly influenced the energy budget. The ratio λET/R_n increased linearly as LAI increased, whereas H/R_n decreased linearly and G/R_n and β decreased logarithmically. Seasonal mean λET was 76.8±4.7 W m⁻², which was less than the reported values for field crops. This difference was attributed to the semi-closed microclimate in the greenhouse. The high values of the Priestley–Taylor coefficient (α=1.03±0.05) and the decoupling factor (0.69±0.05) showed that λET was principally determined by R_n. These values support the conclusion that λET was energy limited rather than water limited in the greenhouse. Canopy conductance (G_c) also influenced λET as indicated by the high correlation between α and G_c, especially during the middle growth stage. These findings are of great importance in creating an energy-driven model and will lead to improved water management in greenhouse agriculture.

Introduction

Greenhouse agriculture occupies a land area of ~405,000 ha worldwide (Katsoulas and Stanghellini, 2019). Solar greenhouses (Fig. 1) supply a major proportion of vegetable crops in China; they provide optimal crop production environments and maximize grower profits. Water and energy transfer between soil surface and the greenhouse atmosphere governs the physiological behavior of crops, drives water circulation and energy storage, and transforms vegetation and soil (Baldocchi, 2001, Gu et al., 2005). Energy from net radiation (R_n) that is converted into sensible heat flux (H) and latent heat flux (λET) is significantly affected by irrigation, vegetation growth, and the greenhouse microclimate (Gong et al., 2017a, Gong et al., 2017b, Jiao et al., 2018, Liu et al., 2019). A good understanding of the behavior of energy components is important for improving water management and optimizing irrigation scheduling (Baldocchi, 1994; Chen et al., 2019; Ding et al., 2013; Youssef and Giuseppe, 2005). There have been many studies of water and heat energy transfer within open field ecosystems (Ai and Yang, 2016; Gong et al., 2017a, Gong et al., 2017b; Liu et al., 2019; Tian et al., 2017; Wang et al., 2020; Yan et al., 2017; Yu et al., 2017), in which energy flux and crop evapotranspiration (ET) vary under different climate conditions and soils. The microclimate of a solar greenhouse differs considerably from that of an open field. The greenhouse receives less energy due to the polyethylene sheeting covering the structure; internal wind speed is low due to the greenhouse being partially closed to external meteorological activity; and internal temperature and relative humidity are higher than outside (Gong et al., 2021). These factors all affect water and energy fluxes and energy partitioning in a greenhouse. Energy partitioning is primarily affected by irrigation methods, soil water constraints, crop management practices, and greenhouse whitening, misting and ventilation (Baille et al., 2001, Katsoulas et al., 2001, Kittas et al., 2001, Qiu et al., 2011). Thus energy fluxes and partitioning in a solar greenhouse differ from those in an open field. To our knowledge, there have been few studies that investigated the energy budget in a solar greenhouse, especially on an hourly scale.

The main component of energy flux is λET, which needs to be accurately measured when studying an energy budget. There are many techniques that are used to directly or indirectly measure λET in an open field, such as water balance, lysimeters, eddy covariance, Bowen ratio energy balance, scintillometry, sap flow plus microlysimeters, remote sensing energy balance, and satellite-based methods (Allen et al., 2011, Chen et al., 2021, Qiu et al., 2021). The water balance method has been used extensively to measure λET in a greenhouse (e.g. Qiu et al., 2015), with a recommended calculation period of seven days or longer to ensure accuracy (Allen et al., 2011). Li et al. (2020) attempted to measure hourly λET in a greenhouse using the Bowen ratio energy balance, but this approach has limited applicability due to the uneven ground and small surface area of a greenhouse (Papadakis et al., 1994, Yan et al., 2019). The weighing lysimeter has in the past been used as a precise method of measuring crop λET (Libardi et al., 2018), and have been extensively used in open fields to provide accurate hourly λET (Anapalli et al., 2016, Benli et al., 2006, Ding et al., 2010, Marek et al., 2016, Puppo et al., 2019, Xu and Chen, 2005). However, the use of weighing lysimeters to measure λET for greenhouse crops such as tomatoes is now rare, although the use of this technology has been reported for prairie grass and sugarcane (Libardi et al., 2019). Long-term continuous and accurate determination of λET is necessary to accurately quantify exchange of water and energy between atmosphere and soil surface.

Seasonal variation in λET is influenced by various physical and physiological factors. Physical factors can be quantified by the Priestley–Taylor coefficient (α), the Bowen ratio (β), and the decoupling factor (Ω); physiological factors are represented by canopy conductance (G_c) (Ding et al., 2015, Jarvis and McNaughton, 1986, Jiao et al., 2018). The Priestley–Taylor coefficient α is defined as the ratio of λET to equilibrium evapotranspiration. Its use eliminates the influence of weather and it can be used to analyze the factors that control λET. The Bowen ratio β is the ratio of sensible heat flux to latent heat flux; it is an important indicator of subsurface moisture and also indicates and reflects the energy distribution in the subsurface. The canopy conductance G_c represents the overall response of the crop to the environmental factors; it is a key parameter in the calculation of λET models. Jarvis and McNaughton (1986) investigated the response of λET to stomatal behavior. They rearranged the Penman-Monteith equation and decomposed it into two linear equations for boundary conditions. When the aerodynamic conductance was very large or tended to infinity, the corresponding hydrothermal transport was very efficient, leading to a leaf temperature that was close to the air temperature, with little effect from the input radiation; the leaf surface was well coupled to the environment when the aerodynamic conductance was very small or tending to zero, there was little hydrothermal transport between the surface and the atmosphere; the leaf surface was poorly coupled to the environment. These phenomena can be quantified and analyzed using Ω. Many studies of field crops have investigated the influence of these factors on λET, and results for different crops are inconsistent (Ding et al., 2010, Ding et al., 2015, Suyker and Verma, 2008, Zhang et al., 2016). Liu et al. (2019) found that λET was determined mainly by R_n and that daily λET was significantly influenced by G_c, shown by a good correlation between α and G_c. In contrast, G_c had a stronger influence than R_n on λET for plastic film-mulched cotton (Tian et al., 2017). Jiao et al. (2018) compared G_c, Ω, α, and β between maize and grapevine canopies; they found that λET for maize was primarily determined by R_n and λET for grapevines was primarily determined by G_c. However, there have been few solar greenhouse studies that characterize the physical and physiological drivers of λET, possibly because it is difficult to accurately measure hourly λET. The calculation of aerodynamic conductance G_a, a key parameter for determining G_c in a solar greenhouse, differs from the calculation for an open field because wind velocity in a solar greenhouse is generally low (Qiu et al., 2013, Zhang and Lemeur, 1992); thus the heat transfer coefficient has generally been used to determine G_a in a greenhouse. The equations for calculating G_a in a greenhouse vary for different modes of convection (free, forced or mixed convection), so identification of the convection mode is critical for accurate calculation of G_a. The accuracy of G_a will in turn affect the accuracy of G_c (Bailey et al., 1993, Montero et al., 2001, Qiu et al., 2013, Yan et al., 2018, Zhang and Lemeur, 1992).

Hence, in this study, to address the current scientific problems such as unclear mechanisms of energy budget and physical and physiological factors that affect λET in greenhouse grown tomato with drip irrigation, we characterize the modes of convection in the solar greenhouse, quantify λET and the dynamics of the energy budget, and identify the physical and physiological factors that govern λET.