Bi-environmental evaluation of oat (Avena sativa L.) genotypes for yield and nutritional traits under cold stress conditions using multivariate analysis

Abstract

Oat being a rabi/winter crop in Kashmir, experiences extremely low temperatures which has detrimental effects on its growth and development. Therefore, this study was designed to evaluate a set of 130 oat genotypes in multi-location trials across temperate conditions of Kashmir valley from 2018 to 2022. From the preliminary data of 56 genotypes, including five checks, were selected and evaluated for nutritional and yield attributing traits under cold stress conditions at two locations. The results demonstrated significant genetic variation and high heritability for majority of traits, except for days to 50% flowering, days to maturity and dry fodder. Positive correlations were observed between green fodder yield and other traits, indicating their potential for enhancing yield. Principal component analysis identified four principal components that accounted for 69.87% of the total variation. Cluster analysis categorized the genotypes into two main clusters and six sub-clusters. Frost damage assessment was conducted at tillering stage after the snow melted in late January 2021 and 2022 using cold tolerance rating scale and subsequently tested for chilling injury through an electrolyte leakage test. From field and lab data analysis, five most promising cold tolerant, nutritious and high-yielding genotypes were identified. These genotypes have significant potential for utilization in future breeding programmes to improve cold tolerance in cultivated oats within the Kashmir valley thus promoting agricultural productivity and sustainability. The outcomes also provide valuable insights into the genetic variation, heritability, genotype-by-environment interactions, correlations and cold tolerance of oat genotypes in Kashmir.

Introduction

Oats rank sixth in the world cereal production with an annual production of 22.7 million metric tonnes (USDA, 2021) following wheat, maize, rice, barley and sorghum (Ahmad et al., 2014). Fodder oat is grown in 0.9 lakh ha in Jammu and Kashmir, with an overall production of 33.6 lakh tonnes (Anonymous, 2014). The Poaceae family includes the genus Avena. Oat is a multipurpose crop used for grain and forage. It is regarded as one of the best dual-purpose cereal crops, fitting well into both human and cattle diets. However, to qualify as a dual-purpose crop, a variety must have ability to produce high green fodder and grain yield but the same crop can only be used as dry fodder if grains are to be harvested. Oat consumption has increased due to its nutritional benefits resulting from antioxidants and high soluble fibre content (Rasane et al., 2015). Oats contain a high amount of lysine and a low amount of prolamins (Rodehutscord et al., 2016) which is considered superior to other cereals that have reduced levels of lysine and elevated levels of prolamins. Several diseases such as Protein Energy Malnutrition (PEM), kwashiorker and marasmus are known to be caused by deficiency of amino acid lysine in diet. Protein content in whole grain oat kernels ranges from 9 to 15% in dehulled kernels. Oats contain a high concentration of glucans, cellulose and arabinoxylan, as well as lipids, vitamins, minerals, antioxidants, phenolic compounds, flavonoids, tocopherol and sterols (Premkumar et al., 2017). Recent research has examined the effects of oat consumption on health, and the benefits go beyond lowering cardiovascular risk factors such as diabetes, controlling blood pressure, lowering blood cholesterol concentrations, controlling and maintaining weight and improving gastro-intestinal health (Clemens and Klinken, 2014).

Morphological evaluation of a germplasm collection is used to describe its genetic diversity and helps in identification of agronomically significant variation. This type of evaluation entails characterizing variations for various morphological traits. Before beginning a breeding programme, it is critical to evaluate germplasm collection for key agronomic traits such as seed quality and defensive traits, flowering, maturity, plant height, protein content, oil content, primary branches, number of capsules, pest and disease resistance, drought and cold tolerance and other desirable traits (Krull and Borlaug, 1970). During the early vegetative or seedling stage, cold stress injury has been evaluated in prior research using a rating scale ranging from 0 to 9 (Rizza et al., 1994). Cold stress can also disrupt membrane integrity, resulting in ice formation in plant tissues and solute leakage. The extent of damage caused by stress in plant tissues can be determined through electrolyte leakage testing, which is a commonly used method (Demidchik et al., 2014). The aim of this study was to compare the performance and quality of oat genotypes bred in other parts of the world to advance different oat genotypes for the Himalayan region, as well as to identify ‘exotic’ sources of traits with the potential to broaden genetic variation within oat breeding programmes. In addition, the study aimed to evaluate cold/frost tolerance by using both field screening and electrolyte leakage testing (a) to evaluate genetic variability within the diverse germplasm for cold tolerance and (b) to identify cold-tolerant genotypes for use in subsequent breeding programmes. No such comprehensive study of cold tolerance in oats involving 4 years of selection using multivariate analysis is reported from temperate regions of the world.

Materials and methods

Plant material

During rabi 2018–2019, 130 oat accessions from Nordic gene banks in Sweden were sown in non-replicated field plots at Faculty of Agriculture, SKUAST-K, Wadura for seed multiplication and in order to obtain preliminary observation data on the adaptation of the genotypes to Himalayan growing conditions (data not included). Based on yield, earliness and plant height 70 genotypes were selected which were further evaluated in rabi 2019–2020 in a similar fashion. Fifty-one promising genotypes were further evaluated using augmented block design at multi-location trials in rabi 2020–2021 (November–April) and rabi 2021–2022 (November–April) against five checks (SFO-1, SFO-2, SFO-3, SFO-4, SFO-6) to identify genotypes with higher green fodder yield and grain yield. These 56 (51 + 5) genotypes were also extensively screened for cold stress in the same experiment under field conditions and laboratory conditions for electrolyte leakage and cold tolerance rating (CTR) (Rizza et al., 1994). The list of genotypes evaluated each year is presented in online Supplementary Table S6.

Field trials

Four field trials were conducted during consecutive years of 2020–2021 and 2021–2022 with one trial each year at Research Farm of the Division of GPB FoA, Wadura, Kashmir (34°17′ N and 74°33′ E at an altitude of 1594 m AMSL) and Dryland Agriculture Research Station (DARS), Rangreth, Kashmir (33.99′ N and 74.80′ E at 1599 m AMSL) given a plant geometry of 20 × 10 cm². The experimental design used was augmented block design. Crop management was done according to local practice for oat production.

Climate and weather conditions

The climate of Kashmir valley is humid temperate characterized by hot summers and severe winters. The average annual precipitation is 944.6 mm ranging from 676 to 1193 mm. More than 80% of the precipitation is received from December to April owing to western disturbances. The minimum and maximum temperature range between 8.0 and 33°C exhibits a considerable fluctuation both in summer and winter. The cold weather conditions that prevailed in the Kashmir valley during the winter season were found to be ideal for screening 56 accessions for cold stress tolerance. During the winter of 2020–2021 (November–February), average day temperatures ranged from 12 to 14°C, while average night temperatures ranged from 2 to 4°C. During this time, the average amount of snow accumulation ranged from 12.6 cm (December) to 14.9 cm (January). Plants were covered in snow for 5 d on average. The average day temperatures in the autumn of 2021–2022 (November–December) ranged from 15 to 17°C, while the average night temperatures ranged from 2 to 5°C. During this time, the average snowfall ranged from 30.4 cm in January to 11.4 cm in February. The crop was snow-covered for a maximum of 25 d (Fig. 1(a)–(c)).

Figure 1. (a, b) Month-wise variation in temperature during oat cropping season (2020–2021 and 2021–2022) and (c) monthly snowfall in cm during 2020–2021 and 2021–2022, at SKUAST-K, Wadura, Jammu and Kashmir, India.

Agronomic observations

Agronomic traits were observed and recorded using five competitive representative plants, randomly selected from each experimental plot within every block, along with checks and tagged for biometrical observations. Observations days to 50% flowering, days to maturity, plant height, total count of tillers at the flowering stage, green fodder yield per row (q/ha), dry fodder yield per row (q/ha), panicle length, 1000 seed weight (g) and number of seeds per panicle were taken at the appropriate phases of plant growth. The seed yield per plant was measured in rows and then converted to quintals/ha.

Quality traits

The biochemical traits of selected plants were estimated in the laboratory of the Division of Genetics and Plant Breeding FoA, Wadura. Seed crude protein (%) was estimated from total nitrogen measured using the micro-kjeldahl nitrogen evaluation method. Total soluble sugar (TSS) estimation was done using the anthrone method. The contents of iron (Fe), zinc (Zn) and other micronutrients in the extract were estimated using an atomic absorption spectrophotometer (Model Perkin-Elmer 2380).

Screening for cold stress

Injury by cold was assessed using a 0–9 scale (Rizza et al., 1994) (Table 1). Data were recorded for cold stress tolerance in the month of January after the plant's exposure to cold temperature in the autumn. These accessions were further assessed for the chilling injury in the laboratory with electrolyte leakage testing from damaged leaf tissues. Electrolyte leakage measurements were recorded (Steel and Torrie, 1980) and electrolyte leakage index (ELI) was calculated as given by Hepburn et al.'s (1986) formula:

$${\rm ELI}\;( {\rm \% } ) {\rm} = \displaystyle{{( {L_{\rm t}-L_ 0} ) } \over {( {L_{\rm b}-L_ 0} ) }}{\rm \times 100}$$

where L _t is the sample's electric conductivity after cold exposure, L ₀ is the sample's electric conductivity at plant growth temperature and L _b is the same sample's electric conductivity after boiling. The ELI represents electrolyte leakage from damaged tissues as a percentage of leakage from destroyed tissues (100%).

Table 1. Evaluation of 56 oat genotypes for cold tolerance

Statistical analysis

For calculation of all statistical analysis including G × E interaction, homogeneity of error variance was tested using Bartlett's test of homogeneity of variances. In Randomised block design (RBD), the combined plot means from two environments were subjected to a pooled analysis of variance. For the estimation of multivariate variability analysis, diversity analysis and selection of candidate genotypes for high yield and fodder quality, the data obtained were analysed using R software to compute cluster analysis, principal component analysis (PCA) and Pearson's correlation analysis. The phenotypic data were analysed separately for CTR and ELI and various statistical parameters, including mean, range, standard deviation and Pearson's correlation coefficient (r), were calculated using R software.

Results

Genetic variability analysis across environments

The analysis of variance for all traits revealed that the mean sum of squares due to environments and genotypes were highly significant, while the mean sum of squares due to G × E interaction was also significant for all but days to 50% flowering, days to maturity and dry fodder weight as shown in online Supplementary Table S1. The descriptive statistics of different yield and yield attributing traits and quality traits recorded in the current study are presented in online Supplementary Table S2. Significant variations were observed in all the studied traits and the coefficient of variation (CV) are ranged from 1.64 to 45.39% with the highest CV recorded for dry fodder yield and lowest CV for seed crude protein. CV also varied with respect to the environment. The magnitude of the phenotypic coefficient of variance was greater than the genotypic coefficient of variance for all of the characters studied, indicating that the environment has a significant impact on character expression (online Supplementary Table S3).

Correlation analysis and PCA studies

Seeds per panicle showed a significant positive correlation with seed yield and main culm diameter. Plant height showed a significant positive correlation with green fodder weight. Panicle length showed a significant positive correlation with seeds per panicle, main culm diameter and seed yield. Number of tillers/m² showed a significant positive correlation with plant height, green fodder weight and days to maturity. Green fodder yield and dry fodder yield showed a significant positive correlation with plant height, days to maturity, number of tillers/m², dry fodder weight and days to 50% flowering. Seed yield showed a significant positive correlation with seeds per panicle. The quality parameters estimated in 56 oats genotypes also showed a diverse pattern of results when they were compared with the green fodder yield and other parameters of the respective genotypes. The results are presented in Fig. 2. The principal components formed were equal to that of total characters studied i.e. 11. The eigenvalues for different principal components are shown in online Supplementary Table S4 and Fig. 3. Only four principal components having eigenvalues greater than one (>1) contributing 69.87% of total cumulative percentage were selected with component − 1 having the maximum percentage of 31.78. For principal component − 1, the plant height contributed the highest percentage of 19.41. For component − 2 highest percentage was contributed by seeds per panicle (26.94), while as in component − 3 and component − 4, the traits iron and seed crude protein contributed highest with percentage of 28.36 and 41.50, respectively. From online Supplementary Table S4, it can be predicted that component − 1 and component − 2 contributed highly towards yield and yield attributing traits while as component − 3 and component − 4 contributed mostly for quality traits.

Figure 2. Correlation between yield and yield attributing and quality parameters.

Figure 3. (a) 3D PCA plot, (b) distribution of characters, (c) distribution of genotypes and (d) biplot of genotypes versus characters.

Estimation of genetic divergence

A Ward D ² method was used to group the genotypes into different clusters based on Euclidean distances. The distribution pattern of genotypes into various clusters is given in online Supplementary Table S5 and Fig. 4. Fifty-six genotypes including five local check varieties were grouped into two main clusters which were further cut to make a total of six sub-clusters with sub-cluster-II (16) having maximum number of genotypes followed by sub-cluster III (12) and sub-cluster-IV (12).

Figure 4. Cluster dendrogram.

Screening and evaluation of oat germplasm for cold tolerance

The level of cold tolerance in all oat genotypes was revealed by comparing mean CTRs (Table 1) and Fig. 5. The CTR scale consisted of scoring from 0 to 9: 0: no damage, 1: slightly yellowed leaf tips; 2: half yellowed basal leaves; 3: fully yellowed basal leaves; 4: whole plants slightly yellowed; 5: whole plants yellowed and some plants withered; 6: whole plants yellowed and 10% plant mortality; 7: whole plants yellowed and 20% plant mortality; 8: whole plants yellowed and 50% plant mortality; 9: all plants killed. The highest CTR score (8) was recorded in BELIDA and VIRMA whereas minimum score of 2 was recorded in NGB 12221.2, BALLET I, MAGNE and LINDA. The ELI (%) was used as a physiological index in the current study to identify the magnitude of injury among studied genotypes after cold stress. Under natural conditions, the ELI test analysis revealed significant differences between different genotypes in the materials under study. In NGB 12221.2, the overall ELI score was estimated to be a minimum of 23.6%, followed by BALLET-I (26.51%), BLENDA (36.54%), LINDA (37.22%) and DIANA (39.17%). BELIDA (81.77%), VIRMA (79.27%), KYOTO I (78.32%) and GALLOP (77.69%) had the highest leakage percentages (Table 1). The average ELI score for the genotypes studied was found to be 53.57%.

Figure 5. (a) Correlation between CTR and ELI (%) in oat (A. sativa L.) germplasm. (b) Histograms showing mean frequency distribution in oat germplasm for CTR (x^ctr, trait under study, y^frq, number of plants).

Discussion

This study was carried out to identify agronomically desirable oat genotypes with cold tolerance traits. Kashmir valley receives ample snowfall and night temperatures during rabi season may well go below zero degree for several hours and for several days during chillaikalan (40 d harsh winter). Therefore, it is essential to identify oat varieties/genotypes suitable for Kashmir valley as oat is the most important rabi crop in Kashmir. These attributes make this study the only first of its kind in the Kashmir valley amidst booming white (milk) revolution-driven demand for oats as a primary fodder crop and a novel cereal for health-conscious people. However, cold tolerance cannot be the only criteria for selection. Agronomic desirability and yield are equally important features. Here we have employed multivariate analysis to first identify genotypes on the basis of fodder yield, food grain yield and nutritional quality. On the basis of our findings, the most promising genotypes are presented in Table 1.

The pooled analysis of variance based on two environments and 2 years of data from both locations revealed that genotype-related mean squares were highly significant for all traits under study. The results show that there are significant impacts of traits on overall variability and we can select the genotypes from the present set on the basis of these contributing traits. The cluster means for yield and yield attributing traits and quality traits indicate that sub-clusters III and IV are the best fit for yield and yield attributing traits and thus genotypes from these two clusters can be used in future breeding programmes in oats to develop high-yielding short duration varieties.

Based on correlation analysis and PCA studies we demonstrate that genotypes can be selected based on seeds per panicle, plant height, panicle length, seeds per panicle and number of tillers/m². The yield contributing traits are known to be selection yardsticks for section in crop breeding (Bibi et al., 2012). Buerstmayr et al. (2007), Moradi et al. (2005) and Choubey et al. (2001) reported that green fodder yield exhibited a significant positive association with tiller number, plant height, leaf width and dry fodder yield. Bityutskii et al. (2017) reported that a significant positive correlation was found between Fe and Zn. Pankaj (2019) reported that TSSs showed a positive and significant correlation with plant height and green fodder weight. However, based on PCA a significant contribution in variance due to crude protein and iron content was recorded which means that there is a scope of selection for better nutritional quality. The cluster means for yield and yield attributing traits and quality traits indicate that sub-clusters III and IV are the best fit for yield and yield attributing traits and thus genotypes from these two clusters can be used in future breeding programmes in oats to develop high-yielding short duration varieties (online Supplementary Table S5).

Cold tolerance screening in the field (winter sown; 2020–2021 and 2021–2022) as well as for cell membrane stability (ELI %) leads to identification of promising cold tolerance oat genotypes (Table 1). On a scale of 0–9 (Rizza et al., 1994), the genotypes varied significantly ranging from 2 to 8 with an average CTR of 5.52. Chawade et al. (2012) also used field cold tolerance screening to select promising genotypes in Sweden. Murariu et al. (2017) also used the Rizza et al.'s (1994) scale to check cold tolerance in 104 European oat genotypes and found great phenotypic variability in genotypes studied.

The ELI was used as a physiological index in the current study to identify the magnitude of injury among the studied genotypes after cold stress. The results of the ELI test revealed significant differences between genotypes (in Fig. 2(c)). The electrolyte leakage analysis concurred with the observed data for CTR, indicating more electrolyte leakage in the cold susceptible genotypes. A highly significant correlation was discovered between ELI score and CTR (Fig. 2(b)). Mir et al. (2021) also reported highly significant positive correlation between CTR and ELI (%) during their study on chickpea accessions.

Out of 56 accessions evaluated for cold tolerance on a scale of 0–9, the genotypes ranged from 2 to 8 with an average CTR of 5.52 (Rizza et al., 1994). Three promising genotypes were identified for grain yield (BALLET-I, NGB1222221.2 and MAGNE), three for fodder yield (KUNGES, ADLER and JADDER) and five for Dual-purpose (DIANA, GULDREGNI I, LINDA, BIRI and ALDEN). Genotype GALLOP was found with the highest TSS (5.61%), VIRMA with the highest seed crude protein (15.2%), ORN with the highest iron content (101.2 mg/kg) and ASLAK with the highest zinc content (68.52 mg/kg). The promising genotypes with high yield potential and cold tolerance for the Kashmir region were identified as: NGB-12221.2, BALLET-I, MAGNE, BLENDA, LINDA, DIANA, ALDEN, GULDREGNI-I, ADLER, KUNGES and JADDER (online Supplementary Table S7). The pooled 2 year data of all genotypes are presented in online Supplementary Tables S1 and S2. Chawade et al. (2012), Murariu et al. (2017) and Hasanfard et al. (2021) also used ELI to score the cold tolerance and the results obtained during the present study were in agreement with earlier published reports. However the current study is the first of its kind in the Kashmir valley.

The promising genotypes identified in this study could be tested further for use in varietal development, crossing programmes to converge desirable genes for yield correlated traits, cold tolerance and for further use in genomic selection and the development of superior oats varieties suitable for the Kashmir valley. However some might be suitable for other states in the northwestern Himalayan agroclimatic zone.