The high requirements imposed by modern pasta processing enterprises on the grain quality of both spring and winter Triticum durum dictate the necessity to develop new methods, approaches for estimating breeding material and selecting high-quality genotypes. Correlation analysis, which makes it possible to identify practically positive and negative factors, minimize the most unfavorable ones, accelerate, and increase the efficiency of the breeding process, is one of them.The purpose of the current work was to study the correlation between quality traits of winter durum wheat and to determine the most accessible and informative ones as selection criteria in the breeding process.The field and laboratory (determination of grain quality parameters, rheological properties of dough and pasta) study was carried out on the basis of the FSBSI “ARC “Donskoy”, from 2014 to 2020. The material for the study were the varieties and lines of Competitive Variety Testing (n = 35).There have been presented the analysis results of a correlation between quality traits of kernels, hard semolina, pasta of winter durum wheat. There have been identified the most accessible, informative and significant traits, such as protein percentage, gluten quality (GDI), alpha-amylase (FN) activity, which have a decisive effect on the strength, cooking properties of pasta, rheological properties of the dough, which can be used as selection criteria in the breeding process, including at the early stages. Thus, protein percentage has positively correlated with pasta strength (r = 0.73), dough durability and elasticity (r = 0.86 and r = 0.61), valorimetric estimation (r = 0.55); GDI and FN have correlated with pasta strength (r = 0.69 and r = 0.57), durability (r = 0.92 and r = 0.57), elasticity (r = 0.75 and r = 0.50). The cor­relation between cooking properties of pasta and the above-mentioned traits was negative (positive correlation): digestibility by weight and volume with protein percentage (r = –0.60 and r = –0.71), GDI (r = –0.49 and r = 0.47), FN (r = –0.48 and r = –0.56), dry matter losses, respectively, r = –0.87, r =–0.85, r = –0.78. The content of carotenoids and the color of pasta are characterized with a mean positive correlation (r = 0.46). There has been found that informatively significant quality traits (protein, GDI, FN, carotenoids) are positively correlated with each other from moderate (protein with carotenoids r = 0.36) to strong (protein with FN r = 0.94) links, but negatively with such important parameters for winter durum wheat as 1000-kernel weight and grain unit. Therefore, when selecting breeding forms and lines with increased values of the main traits, it is necessary to control 1000-kernel weight and grain unit, or at least one of them.

The main task of agricultural producers has always been and still remains to obtain high and stable yields, which can be provided by new developed varieties, which are the main reserve for increasing grain production. The current paper has presented the study results of the work in 2018–2021, namely, the development of promising varieties and the main features of seed production. The purpose of the study was to estimate promising winter durum wheat varieties according to the main economic and biological traits and properties, quality indicators of grain and pasta, and seed production features. The studied varieties were able to give consistently high yields (from 5.56 to 12.67 t/ha), combining the main economic and biological traits (drought resistance, frost resistance, resistance to major diseases and lodging) and quality indicators of grain and pasta. The study of the physicochemical properties of grain showed that the winter durum wheat varieties met the requirements of GOST 9353-2016 and belonged to the 1–2 quality classes, which was quite enough to produce good pasta and cereal. The pasta color of the varieties ‘Yakhont’, ‘Yantarina’, ‘Uslada’ was yellow (4.3–4.6 points); it was creamy in the variety ‘Yubilyarka’ (3.5 points); the content of carotenoids varied from 507 µg/% (the variety ‘Yubilyarka’) to 610 µg/% (the variety ‘Uslada’). In order to preserve and maintain the main features and properties characteristic for each variety, their seed production is being currently carried out with a comprehensive estimation of both phenotypic (approbation) traits and genotypic ones using protein markers. The varieties ‘Yakhont’, ‘Yubilyarka’, ‘Yantarina’ and ‘Uslada’ have clearly distinguishable spectra of gliadins 3.2.3T2, 13.2T+6T.3.T1, 5.2.4T1 and 13X.4T.3.2, which makes it possible to control their varietal purity at all stages of seed production.

The current paper has presented the results of environmental testing of winter durum wheat varieties in the conditions of the southern part of the Rostov region. The purpose of the study was to estimate productivity and grain quality of winter durum wheat varieties of various maturity groups. The objects of the study were 53 winter durum wheat varieties of various environmental origins of breeding, namely FSBSI “ARC “Donskoy”, FSBSI RCG named after P. P. Lukiyanenko, FSBSI FRAC and other institutions. The variety ‘Kristella’ was used as a standard variety. According to the results of the study there was found that 22.6 % (12 pcs.) of winter durum wheat samples belonged to the middle-early maturity group with a heading stage starting on May 20–22; 49.1 % (26 pcs.) of samples belonged to the middle-maturing group (May 23–25); 28.3 % (15 pcs.) of samples belonged to middle-late maturing group (May 26–29). The maximum productivity for the study period (2020–2022) was identified among the samples of middle maturity group (8.40 t/ha). High quality indicators were formed by winter durum wheat varieties of the middle-early maturity group with 15.24 % of protein, 27.0 % of gluten, 38 ml of SDS-sedimentation, 567 Mkg/% of carotenoids. On average for three years, the samples of the middle maturity group (78 %; 777 g/l) have formed high kernel hardness and grain unit in the trial. During the correlation analysis, there have been obtained significant positive correlations in middle-early genotypes (with protein content (r = 0.95±0.04), gluten content (r = 0.94±0.04), kernel hardness (r = 0.48±0,12), SDS-sedimentation (r = 0.92±0.04), carotenoid content (r = 0.96±0.04)) and in middle-late genotypes (with kernel hardness (r = 0.41±0.13)). Among middle maturing samples there have been found no significant correlations.

Introduction and methodology. The value of the initial material is primarily determined by the tasks of modern breeding process, collection material playing an important role in it. The purpose of the current study was to estimate the initial material of winter durum wheat according to their yield structure elements in the south of the Rostov region. The objects of research were 159 winter durum wheat samples among which there were 69 of domestic breeding and 90 of foreign one.Results. The study has shown that in 2019–2020 grain productivity of the collection samples ranged from 227.9 g/m2 to 735.9 g/m2. All studied samples were divided into three groups according to the trait ‘productive stand’ (number of productive stems per 1 m2). There were 17.6% of samples with a small value of the trait (201–400 pcs./m2); with a mean value of the trait (401–600 pcs./m2) there were 72.3% of samples and with a large value of the trait (601–800 pcs./m2) there were 10.1% of samples. According to the trait ‘plant height’ more than a half of the samples (57.0%) belonged to the ‘semi-dwarf’ group (61–85 cm). Large part (70%) of winter durum wheat samples formed ‘length of head’ in the range of 6.5–7.5 cm. In 2019–2020, number of spikelets per head ranged from 17 pcs. up to 24 pcs. The trait ‘kernel weight per head’ of the studied collection samples varied from 0.77 g to 2.08 g, all the samples had different kernel sizes. The trait ‘number of kernels per head’ varied from 22 pcs. (the variety ‘Novinka 4’) up to 51 pcs. (the sample ‘1015/16’). The trait ‘1000-kernel weight’ in the trial ranged from 23.9 g to 49.0 g, the standard variety had 34.1 g. The productivity of the collection samples was formed due to 1000-kernel weight and kernel weight per head, the correlation coefficients for these productivity elements being r = 0.45 ± 0.07; r = 0.35 ± 0.07.