Measuring Water Relations Traits
Breeding for drought tolerance has to consider various traits responsible for plant water status. Leaf water potential may be the most important parameter, but it is impossible to measure water potential nondestructively. Thus, various other methods have been tested that assess either leaf and canopy
(See color insert.) Estimation of the leaf angle distribution from a 3D reconstruction of a sugar beet canopy: (a) The stereo camera setup (see insert) has been mounted on a cherrypicker 3.5 m top of canopy; (b) shows a region of interest with ~6 plants in the original left camera RGB; (c) filtered depth (disparity) map of (b) with pixel colors indicating the object depth; (d) single leaf segmentation, which is used for further individual leaf surface modeling; (e) leaf surface models are used to derive different leaf traits like leaf area, canopy area, or leaf angle distributions; (f) the leaf angle distribution of the zenith angle of the previous reconstruction; the zenith angle ranges from 0° (flat leaf surface) to 90° (errected surface) and can derived locally or for the complete leaf.
water content or approximate transpiration rates (Farrar et al. 2011; Sampoux et al. 2011). In this context, measurements of near-infrared (NIR) and thermal imaging have been introduced as the most promising measurement approaches in the lab and in the field (Fiorani and Schurr 2013).
A qualitative approach to estimate plant water content is NIR measurements using the relative depth of the water absorption band in the NIR region (1370 and 1870 nm). This approach generally provides a good relative estimate of water content but to our knowledge there is no study available that describes the retrieval of plant water content as absolute physical values.
A second approach is exploiting thermal cameras, which are sensitive within the infrared region (9-13 pm spectral range) to evaluate plant transpiration and evapotranspiration (ET). The principle of passive thermography is that surfaces are cooled by ET, so surface temperatures are lower than ambient temperature, which is proportional to the rate of ET. However, leaf temperature (TL) does not only depend on ET rates but also depend on the leaf boundary layer, which is a thin layer of air at the leaf surface. The thickness and composition of the leaf boundary layer determine how fast heat can be dissipated, that is, increasing leaf boundary layer decreases the transfer of heat from the leaf to the atmosphere and vice versa (Leuning et al. 1989). Furthermore, leaf boundary layer and thus TL respond in a dynamic way to variable environmental conditions. Parameters such as solar irradiance and ambient air temperature are highly fluctuating and highly affect the leaf boundary layer.
To overcome these problems, Jackson et al. (1981) developed the crop water stress index (CWSI), which normalizes leaf temperature against the prevailing environmental conditions. This index is based on the comparison of leaf temperature to wet and dry reference surfaces. The CWSI and other indices have been shown to be sensitive to evapotranspiration and can be used to detect drought stress-induced stomatal conductance (Jackson et al. 1981; Jones 1999; Cohen et al. 2005; Grant et al. 2006; Moller et al. 2007; Alchanatis et al. 2009).
Passive thermography, and particularly the CWSI, has become a widely used tool for measuring plant evapotranspiration to analyze high numbers of plants in a short period of time. Passive thermography measures only one part of the overall plant-water relations, namely, ET. To understand plant- water relations in whole plants in response to changing environmental conditions (e.g., drought), it is essential to know how different water fluxes between roots, stems, and leaves are connected.
A promising approach is active thermography, where TL is actively manipulated by a short heat pulse. An additional heat pulse increases TL transiently. After a short time, TL will decrease again approaching the former steady state. The time constants (t) of heating or cooling can be measured. This time constant (t) depends on the leaf heat capacity per unit area (CA-f and the leaf heat transfer coefficient (hleaf). High leaf water content leads to a higher heat capacity and consequently higher t. The second parameter affecting т is hleaf, which describes how fast a leaf is able to dissipate heat. hleaf depends on the boundary layer and is therefore highly affected by environmental conditions. For instance, increasing wind decreases the boundary layer, which in turn increases hleaf and decreases t. Also, stomatal conductance affects hleaf. High stomatal conductance accelerates leaf heat dissipation and consequently hleaf increases and т decreases. Active thermography was successfully tested in laboratory at leaf scale and under greenhouse conditions at canopy scale. At leaf scale, a linear relationship between т and LWC and consequently CAle1af was found. This relationship changed when the leaf boundary layer and thus hleaf was manipulated by wind.
We would like to highlight the need to combine passive and active measurements to fully understand the overall plant-water relations in response to drought. Intuitively, the water content limits water loss by transpiration and thus controls transpiration rates. However, neither the CWSI alone, nor т alone are appropriate to reflect this relation. By combining т with CWSI, one may be able to track changes in LWC, boundary layer conductance, and transpiration. This will facilitate a better understanding of the dynamic responses of plants to optimize their water relations and help to better understand the strategies to cope with drought stress.