The standardization of STS aimed at facilitating the combination, but there was a large gap between scores. However, as the STS was ordinal data recommended by Egdane et al,it had a greater priority than the growth character for tolerance selection. Therefore, the standardization was considered still relevant to determine genotype tolerance. SaTI results showed a normal distribution pattern in the population; and thus, it was comprehensive and effective. Moreover, the SaTI had an additive pattern. Formation of normal distribution or standardization was also carried out by Peternelli et al  through discriminant analysis in the formation of their selection index. The effectiveness of SaTI  was also shown in a more selective selection than that only based on STS values. Anshori,who used the same genetic materials and methods, found 45 good tolerance genotypes to salinity stress using only STS. The genotype adaptability index for tolerance to salinity stress based on a combination of GAI and SaTI was an indirect selection approach for a targeted environment. The indirect approach aimed at optimizing the potential of two different characters. Simultaneous selection with two or more characters would be tighter if they were not correlated. Our results indicated that GAI and SaTI were not correlated, and therefore the IASI model became important to obtain appropriate genotypes in saline areas.

Based on the correlation of the Sukra yield, IASI had the highest significance correlation to the yield. That is, the combination of GAI and SaTI was more selective and effective in predicting the actual yield in saline environments than the selection solely based on tolerance. Besides,mobile vertical farm the number of selected DH lines  was considered quite appropriate for preliminary selections. This indicated that IASI was a good method for preliminary selections of DH lines under salinity or other stresses. The validation in Sukra showed that the combination of both experiments was more effective than a focus on the tolerance selection in hydroponic screening because the yield is not correlated to the tolerance trait in early vegetative stages. Therefore, the selection of sustainable genotype under stress should combine agronomic and tolerant characters. IASI was an appropriate indirect approach in selecting the adaptive genotype under a saline environment. Its correlation value to the Sukra yield reached 0.69. The low repeatability showed that the environmental factors affected the yield variance and influenced the bias in correlation and regression analyses. Although the regression had a low variance value in predicting the Sukra yield,the genotype distribution was still within prediction interval,except for one genotype. In addition, based on direct adaptability index  in Truntum as a saline prone area  in the dry season,IASI also has a significant correlation to DAI with the value of 0.59. This indicated that IASI is still relevant in predicting the response of adaptive genotypes in saline areas. DH lines have high homozygous levels; however, they do not ensure stability in stressful areas. Stressful areas can be strongly influenced by the environment and G × E interaction. Hidayatullah et al  showed that DH rice lines have significant G × E interactions in all agronomic characters. In addition, the heritability of DH wheat lines can decrease for all characters in a saline environment. 

IASI may increase the effectiveness and efficiency in DH line selection, especially in the initial stage of breeding during the development of DH lines adaptive to salinity stress. In conclusion, multivariate analysis in the selection process can improve the effectiveness of DH rice line selection in preliminary selection stages. We found that GAI formula was 0.465 yield + 0.433 NPT + 0.31 NFG. In all, 24 DH rice lines were considered to have good agronomic characters. SaTI was developed by an average of SSI based on discriminant analysis and standardized salinity tolerance score. There were 34 DH rice lines with good tolerance to salinity stress in hydroponic culture. SaTI, in selecting tolerant genotypes, was more effective than that based solely on STS. IASI was effective based on the correlation in Sukra validation and DAI, and 28 DH rice lines were selected as adaptive to salinity stress.The bio-fortification of staple foods has been accepted as a practical and cost-efficient way to alleviate malnutrition by increasing the concentrations of essential nutrients in crops and ultimately in humans. Paddy rice,one of the most widely planted staple crops and the source of 80% of the daily caloric and micro-nutrients for over half of the world population, is the most suitable candidate for crop bio fortification strategies. Compared with medical supplementation and dietary diversification, it is easier to directly benefit the malnourished conditions of people living in poor rural regions via agronomic or genetically bio-fortified rice. Rice is adapted to diverse local edaphic and climatic conditions, resulting in the development of thousands of varieties and genotypes by selective breeding,however, yield and environmental adaptation are likely the most important factors for varietal breeding to enrich micro-nutrients content. For example, superfluous soil iron  is one of the most significant conditions decreasing rice production in Southeast Asia and South America. 

Farmers likely prefer rice with a lower Fe uptake to ensure optimal yield. Therefore, it is important to screen for the elemental concentration status of rice varieties. Moreover, plants take up and translocate nonessential elements through sharing the same pathways as essential elements due to similar chemical properties. For instance, zinc  and cadmium  share the same zinc-iron transport protein  influx transporter into plant roots ; and plant roots can absorb arsenic  and selenium  via the phosphate transporter.These nonessential elements can be enriched through the food chain from crops to humans, and the consumption of contaminated crops can pose a significant health risk. Rice is easy to accumulate toxic metal under flooded conditions due to high affinity. Many previous studies have reported the public concern and health risks of Cd and As,and the radioactive contaminant elements cesium  and strontium  in crop grains harvested from contaminated areas, such as Fukushima in Japan and Chernobyl in Ukraine. Hence, increasing the essential nutrient uptake and reducing nonessential elemental contamination in rice is a pressing issue for human health. Elements with similar chemical properties sharing the same transportation pathways can result in competitive uptake between them. Ionomics, the study of the elemental profiles of all essential and nonessential elements in organisms, tissues and cells, provides a rapid screening strategy for elemental interactions and varietal identification of plant species using high-throughput element analytic methods, such as inductively coupled plasma-mass spectrometry  and ICP-atomic emission spectrometry. Numerous ionomic studies for the rapid identification of elemental interactions among plant varieties have been conducted. Chen et al  performed ICP-MS to quickly characterize Lotus japonicus mutants with altered ionomic profiles. Chu et al  screened extreme Cs accumulation among Amaranthus species to find the ionomic basis for low Cs varietal breeding.

Watanabe et al  used ionomic methods to screen vegetable varieties to determine the characteristics of mineral accumulation and nutritional values. Therefore, it is feasible to screen rice varieties for bio-fortification and safety by ionomic-based research methods. Consequently, we cultivated 120 rice varieties at the early seedling stage in identical hydroponic systems to the external environment. The concentrations of nutrients, trace elements and anions of the rice species were analyzed by ICP-MS and capillary electrophoresis  to screen for rice varieties with high concentrations of nutrients or low concentrations of toxic minerals or both. The ionomic interactions among genotypes and the ionomic variations corresponding to phylogenetic relationships were also characterized.Meanwhile, a principal component analysis  using correlation coefficients conducted to compare the correlations among subspecies, indicated a 34.0% of the variance was partitioned in the first PCA axis, and root and shoot subgroup samples were separated on that axis. Root samples also showed a greater degree of separation on the 2nd PCA axis, indicating that the correlations among subspecies showed a differentiation between the shoots and roots. In addition, the PCA correlation scores in the roots among the subspecies were located in the same first quadrant, while the location in the shoots, showing a highly significant difference among subspecies, was in a different quadrant. Consistent to PCA, the elemental correlation patterns of roots were similar among subspecies, while that of shoots were largely different which the strongest correlation pattern was in japonica, and the less was in aus. K and Mg correlated with only a few elements in aus and indica, and P and Al only correlated with more elements in the shoots of japonica and roots of aus, respectively. In contrast, Ca, Cu, Zn and Sr were detected significantly related with many elements in all plots. Mg in the roots of japonica correlated significantly negatively with As, but positive relations showed in the shoots of japonica and indica. The same situations were also detected in Mn and Co, Mn and Cs, Zn and Cs. Although different rice subspecies and organs showed different elemental interactions, many still remained the same.

Obviously, correlations between S and SO4 were always significantly positive. Ca interacted significantly and positively with Ba and Sr, and Ba also significantly and positively correlated to Sr in all rice organs.To identify the ionomic differences between organs of the subspecies and to observe the geographic factors, we compared the element concentrations in all samples,shoots  and roots  using PCA, and displayed the loading plots. As shown in Fig. 4-A, there was a significant separation between shoots and roots in all the rice genotypes. Meanwhile, all the macro-nutrients along with Mn and B loaded on the negative x-axis to explain the shoot ionomes, vertical farming racks while the root ionomes were mainly determined by the micro-elements and anions. The PCA results in Fig. 4-B and -C showed that the clusters were largely determined by the different origins of the varieties, whereas the rice subspecies from the same origins were not significantly separated in shoots or roots. For example, the japonica shoots from Japan were separated from japonica shoots from South Asia and Southeast Asia, but not from the indica shoots from Japan in the PCA results. According to the loading plots, the differences in shoots and roots from Japan were mainly explained by most nonessential and toxic elements such as As and Cd, whereas K, Na, Li and anions mainly contributed to differences in rice varieties from South Asia.There were significant differences in the concentrations of all elements except Ba among japonica, indica and aus, but the magnitudes of element concentrations generally displayed similar, indicating that the ionomic variations were within the pre-framework of the phylogenetic factors of the genus Oryza, although the genomic differences among subspecies have been. As shown in Figs. 1 and 2, all rice subspecies follow the rule that trace elements and anions were concentrated higher in the roots than in the shoots, and it was confirmed by the separation between shoots and roots in all the rice varieties, as well as the loading of micro-elements on roots in PCA. Rice root is the main barrier to limit translocations of heavy metal and toxic element to shoots by chelation and compartmentalization. However, micro-nutrients for the goal of bio-fortification, such as Fe and Zn, are also indiscriminately fixed in roots. Interestingly, opposite to S, the SO4 2- concentration in the roots was higher than that in the shoots, primarily due to most of the inorganic S being fixed in root vacuoles or converted to organic S  in leaf for protein synthesis. Thus, it would be a meaningful challenge to identify absorption and translocation mechanisms for specific micro-elements in roots. As a response to the element concentrations, dry biomass weights among subspecies were showed in boxplot. The concentrations of essential elements in the shoots or roots of japonica were markedly the lowest, while that of harmful elements in japonica showed the highest. Correspondingly, the biomass weights of both shoots and roots in japonica were the lowest among subspecies. The results showed that improving nutrients and reducing toxic elements also showed a crucial correlation on improvement of the biomass and yield of rice. Significant differences among diverse rice genotypes detected in most elements in both shoots and roots showed that the phylogenetically-changed elements were more marked in the shoots than in the roots, and further indicated that the variations in elements among the rice varieties were mainly attributed to their translocation from roots to shoots, consistent with previous studies.