Agriophyllum squarrosum (L.) Moq. is a desert-adapted pseudocereal that has recently attracted attention as a climate-resilient crop and source of valuable phytochemicals and nutritionally relevant metabolites. Despite their ecological and nutritional importance, comprehensive studies combining lipid and phenolic profiles across natural populations remain limited. In the present study, five populations of A. squarrosum from ecologically contrasting regions of Kazakhstan were analyzed to evaluate biochemical diversity and potential for functional food applications. Total lipid content was determined using near-infrared spectroscopy, fatty acid composition was assessed by GC-MS, and phenolic compounds were quantified by HPLC. Multivariate approaches, including PCA, MANOVA, PLS analysis, correlation networks, and TOPSIS ranking, were applied to evaluate population differentiation and relationships between biochemical traits and environmental conditions. Total lipid content in seeds ranged from 7.71% to 15.40%, linoleic acid represented 50.20–57.67% of total fatty acids, and oleic acid ranged from 24.80% to 40.10%. Isorhamnetin was the dominant phenolic compound in leaves, with concentrations between 0.24 and 0.65 mg/g. Populations from Aktobe showed higher lipid and oleic acid contents, whereas Almaty populations accumulated greater flavonoid levels, including isorhamnetin, quercetin, and kaempferol. These findings reveal substantial metabolic differentiation among populations and suggest possible associations with ecological conditions. The observed accumulation of unsaturated fatty acids and phenolic compounds, including isorhamnetin, quercetin, and kaempferol, identifies promising germplasm resources for future studies on functional food development and biological activity evaluation. The results further support the potential utilization of A. squarrosum in sustainable agriculture in arid regions.
Agriophyllum squarrosum (L.) Moq. (sand rice) is a drought-tolerant psammophytic species with high potential as a climate-resilient food crop due to its nutritional value and adaptation to arid environments. This study evaluates morphometric and biochemical variation in seeds from five natural populations across the deserts of Kazakhstan to assess their breeding potential. Seed morphometric traits showed moderate variability (CVs of 4.71–17.98%), with strong positive correlations among seed length, width, and thousand-seed weight, indicating coordinated development. In contrast, biochemical traits, particularly amino acid composition, exhibited substantially higher variability (CV up to 174.9%), reflecting metabolic flexibility under different environmental conditions. Among the amino acids reliably quantified in this study, histidine was the most abundant, while cysteine, tyrosine, and alanine showed high variability. Total protein content remained relatively stable, reaching up to 34.96% in superior accessions. Multivariate analyses revealed significant population differentiation: Akt1 was the most distinct, whereas Alm1 exhibited superior seed size and mass. Weak correlations between morphometric and biochemical traits suggest their partial independence. Integrated multivariate evaluation identified Akt2 and Alm1 as the most promising populations for breeding. Overall, the observed variation highlights strong potential to select genotypes that combine improved seed size with favorable biochemical characteristics, based on the five amino acids quantified above the LOQ, thereby supporting breeding and domestication efforts.