To date, agriculture is officially and commercially worldwide, releasing numerous new cultivars every year, with breaking records of yield quantity and quality (https://mvd.iaea.org/#!Search). Nevertheless, these production rates are too slow to meet the forecasted demand for the world population food demands [23–25]. In order to meet the global food needs, the productivity of staple cereal crops has to increase by more than double (60% to 110%) than the current yield in the coming 32 years, according to the United Nations Food and Agriculture Organization prediction [23,26,27]. Conventional plant breeding, which is based on hybridization of parents and phenotypic selection of offspring has been going on for millennia [28] and is still commonly used today with a great contribution to the annual cultivar release. Even though these hybrid seeds have had a tremendous impact on agricultural productivity, the breeding process is still relatively slow, where an average breeding program can take up to 10-12 years for annual crops [27,28] depending on the heritability of the trait which has been selected. Maintaining the current yield production improvement rate by itself is facing another major challenge as the climate change imperiling agriculture worldwide, and productivity is expected to drop by up to half in many parts of the world by 2080 [27] due to this climate changes. In fact, many of the new cultivar breeds under non-stress conditions and most of the new cultivar show dramatic yield loss under abiotic stress [29–32]. Breeding under abiotic stress conditions is a much greater challenge [33,34] and produces much less new abiotic stress tolerant cultivars every year. This challenge is combined with several factors, which include the complex nature of the abiotic tolerance, the instability, and uncertainty of the environmental conditions and the lack of clear definition for the stress scenarios, tolerance markers, and anticipated trait phenotypes [35–37]. These circumstances may explain the low efficiency in stress breeding, namely, the fact that much less new cultivars are released during much longer breeding periods[38].
In recent decades, the development of technologies associated with molecular markers and Genomic selection (GS) has provided new molecular tools that enhanced some of the classical breeding selections and make the breeding process more efficient for simple and complex traits [39–42]. In fact, to date, genotyping does not serve as a limiting factor in the prediction accuracy of GS [40]. The technical challenge of implementing the GS in crop plants is the reliability of phenotypic data that creates a Genotype-Phenotype Gap (GP gap). The anticipated improvement of genomic tools as well as other “omics” technologies (i.e. metabolomics, proteomics, and transcriptomics) will yield a flood of information. Breeding process is expected to upsurge [28,43,44][13], yet, it is expected to increase the GP gap, as the pace of determining DNA sequence is currently much faster than that of determining function of the organism or in exploiting the genes in crop improvement [45]. Thus, harnessing the true benefits of new technologies will have to be combined with high-throughput phenotyping for achieving the valuable genetic gain from complex traits [38]. It is important to emphasize that the GP gap will become more complex when it involves the interactions with the environment in general and the water-limited environment in particular.
2. Current phenotyping technologies and plant functional phenotyping
Phenotyping is the comprehensive assessment of complex plant traits such as growth, development, tolerance, resistance, architecture, physiology, yield, and the basic measurement of individual quantitative parameters that form the basis for more complex traits under certain environment [46]. The terms plant phenotype and phenotyping are interpreted in diverse ways [47], and the full set of phenotypic features of an individual, their use (and abuse) need a cautious not conclude them to ‘high-throughput measurements’ since ‘high-throughput measurements’ are instrumental means but are not a goal per se [48].
Phenotyping technologies are facing a challenge, which is the capacity to collect both precision and high throughput corresponding phenotypic data. This challenge has moved advanced global public and private plant research institutes to build phenomic facilities [28], which has raised the number of phenotyping facilities from 5 before 2009 to 44 by 2015 [49]. Most of these facilities are capturing information throughout the plant life cycle in controlled environments using robotics and automatic image acquisition and analysis [13,47,50–52]. The most common phenotyping approaches are using different wavelength-range sensors and cameras to capture signals from the plants by scanning plant population using versatile platforms (Moving plants on tracks, moving the sensor on cables, drone etc.). Nevertheless, these technologies have difficulties in providing meaningful information regarding the dynamic plant x environment interactions and its dynamic water stress response[50,53–55].
[Table 2 about here]
Plants are very sensitive to many signals coming from their close environment (e.g. light intensity, relative humidity, CO2 concentration, soil moisture etc.) most of these ambient environmental conditions are not stable and rapidly changing even under controlled conditions. For example, plants that are exposed to drought will rapidly develop different (from each other) soil water conditions, pending their transpiration rate. Moreover, even under similar ambient conditions (e.g. light in a greenhouse) measuring different plants at different hours will be conducted, de facto, under different light conditions due to the natural changes in the light intensity along the day.Thus, measuring single (or few) trait at a single date on a numerous genotype does not necessarily insights into the plant functioning or into the genetic control of this trait [48] in general, and under stress in particular.
2.1 Targeting the appropriate trait and well-defining breeding goals
The definition and terminology of stress-related beneficial traits are complex and must be well defined in breeding programs [37]. Creating a drought tolerant plant does not necessarily means increased productivity. This fact is shown by the work of Yang et al, [56] in which altering the expression of regulators of drought responses has often succeeded in enhancing drought tolerance, at least in laboratory conditions, this typically comes at the expense of development inhibition, and a significant yield penalty. Likewise, breeding for enhanced water use can lead to a yield penalty [57] because as an improved yield potential can translate into better performance under stress, it also places a greater demand on water and other resources [58]. In fact, modern crops use an immense amount of water due to high transpiration (see fig.1 table 1.), yet, their large stomatal conductance hazards their survival ability under water stress [59]. Thus, in order to improve crop drought tolerance, it is critical to strike the appropriate balance between defining clear breeding objectives and ensuring flexibility within the breeding strategy is important[60] [61]. One of the common misjudges in crop breeding is relating survivability traits as tolerant or resistance traits.
o