​Effect of Genotype and Sowing Period on Chickpea Quality, Bioactive and Antioxidant Traits

Chickpea (Cicer arietinum L.) is considered to be one of the most important pulse for human food, ranking second in growing area and third in production around the world. Despite the potential nutritional and health-promoting value of antinutritional factors, their presence in chickpea limits its biological value and usage as food, by interfering with digestion of proteins. However, antinutritional factors can be reduced or eliminated by soaking and cooking (Alajaji and El-Adawy, 2006). 
       
Cooking time (CT) is a significant aspect of cooking quality of legumes that differs widely among genotypes (Kaur et al., 2005). There are also some equally important aspects of cooking quality such as seed size, hydration index (HI), hydration capacity (HC), seed coat percentage (SCP) and protein content (Williams et al., 1983). There is a significant correlation between seed size, CT and HC (Williams et al.,1983). However, Singh et al., (1988) stated that the correlation between seed size and CT neutralized when the seeds were previously soaked. This likely happened due to the negative correlation between seed size and SCP. According to Cobos et al., (2016) there is non significant correlation between HI and CT, whereas, there is a moderate positive correlation between HI and seed size.
       
Chickpea seed composition differs in different genotype, environment and their interaction. Genotype × environment interaction was significant for protein concentration and 1000 seed weight (seed size) (Frimpong et al., 2009). According to Ruggeri et al., (2017) environmental conditions and genetic factors affect grain yield, nutrient and anti-nutrient compositions of chickpea seeds. Smaller pod-filling period, caused by delayed sowing, improved protein content (Dehal et al., 2016). Although several studies have shown the genotype and environment effects on nutritional and quality traits of chickpea, few studies are available related to their influence on bioactive traits. Phenolic compounds are higher in Ascochyta blight resistant genotypes compared to susceptible genotypes (Kumar et al., 2013). Total tannins are strongly affected by the genotype × environment interaction and are negatively correlated with seed size (Nikolopoulou et al., 2006). Primi et al., (2019) showed that the concentration of some bioactive compounds is influenced by genotype and climatic conditions. Unpredictable climate change is the major constraint for chickpea as it increases the frequency of drought and temperature extremes. Environmental stresses during seed development have negative effect on chickpea’s grain quality (Kaur et al., 2008). Sowing chickpeas off-season could be considered a simulation of the expected conditions caused by the climate change, as the plants will be exposed in higher temperature and less soil water content in the crucial periods of anthesis and pod-filling.

Therefore, the objective of the present study was to investigate the effects of normal and late sowing period as well as genotypic effect on quality and bioactive traits, as well as antioxidiant activity.

Plant material and growing conditions
 
The genetic material used in the present study is presented in Table 1. Field trials were carried out at Institute of Industrial and Forage Crops (IIFC) (longitude 22°25' E, latitude 39° 36' N). All the five genotypes were planted at two different sowing dates, one during the normal sowing (NS) period according to the Greek climatic conditions (February 28, 2019) and one off-season (late sowing: LS, April 1, 2019) in order to achieve dry-heat conditions during the chickpea’s critical stages. During the experimental period, no irrigation was applied. Chickpea seeds were sown in randomized complete block design with three replications. Each plot consisted of three rows, each of 2 m length and 25 cm distance between rows. Climatic parameters of the cultivated study site are given in Fig 1.




 
Physicochemical and cooking characteristics
 
1000 seed weight (1000 SW, g)
 
Average weight of three random samples of 100 seeds.
 
Seed coat percentage (SC, %)
 
Seed hulls of each variety were removed manually after soaking 10 chickpea seeds of each variety for 8 h in distilled water at room temperature. The dehulled seeds were then dried and weighted.
 
Hydration capacity (HC, g water/seed)
 
10 seeds from each variety were soaked with 250 ml of distilled water overnight at room temperature. Afterwards, the seeds were reweighed. The HC was evaluated based on the equation:                                                                
 
Hydration index (HI)
 
 It is the ratio between HC and original weight based on the equation:
Cooking time (CT, min)
 
The method followed was a modification of the method described by Iliadis (2003) for lentils. The samples’ cooking time was tested starting from 20 minutes, with 5 minutes intervals, until the penetrometer’s depth of penetration was ≥ 9 mm.
 
Protein content (%)
 
Total nitrogen in the samples was determined by kjeldahl method (Horwitz and Latimer, 2005) and the crude protein was calculated using a factor of 6.25.
 
Total phenolic content (TPC)
 
The total phenolic content (TPC) of extracts was determined by the Folin-Ciocalteu reagent method (Singleton et al., 1999). TPC was expressed as mg of gallic acid equivalents per 100 g sample (mg GAE/100 g).
 
Total tannins content (TTC)
 
The difference in the TPC values before and after the polyvinylpolypyrrolidone treatment (unbound) represented the tannin levels as gallic acid equivalent (Makkar et al., 1993).
 
Antioxidant activity assay
 
ABTS assay was employed to determine the antioxidant activities of the chickpea seed extracts (Re et al., 1999). Trolox was used as the standard compound for calibration curves and the results were expressed in mg of trolox equivalents (TE) per 100 g of chickpea seed (mg TE/100 g).
 
Statistical analysis
 
Two-way analysis of variance (ANOVA) was performed to evaluate the effect of genotype (G), sowing period (S) and their interaction (G×S). Cluster analysis was performed by using Ward’s method in order to explore relationships between genotypes. Principal component analysis (PCA) with varimax rotation was performed to explore relationships between traits. All statistical analyses were performed using SPSS statistical software v.20.

The ANOVA results (Table 2) revealed that genotypic effect was significant (p≤0.05) for all studied traits (physical, quality and bioactive traits), however, both sowing period and the GxS interaction significantly (p≤0.05) affected most of the studied traits, except for protein content and HI. Genotypic effect was stronger than sowing effect, accounting >85% for the physical traits (Wood et al., 2008). The genotypic effect was stronger than the effect of sowing time for some quality traits (Protein, HI, HC) (>64%) (Cobos et al., 2016), whereas CT was largely affected by genotype, sowing time and their interaction. Bioactive traits and antioxidant activity were mostly affected by the genotype, but the effect of sowing period and the interaction GxS, were also significant. Brankovic et al., (2015) working with bread wheat, also, mentioned the same hierarchy (G>S>G×S) of importance in sources of variation of both traits.



The values of each characteristic measured per variety in NS and LS, are presented in Table 3. Higher temperatures and redistribution of rainfall are expected, so the results in late sowing could be used as an indication of the future cultivation of chickpeas under stress environments. No differences were detected in physical traits between the means of the two sowing periods, although 1000SW decreased in LS for all the varieties (Yücel, 2018). Among the five varieties, Macarena had the highest values of 1000SW and SC in both NS and LS, as well as the highest mean, across sowing periods. With regards to protein content, no differences were detected between mean values of NS and LS. However, in LS the protein content of Amorgos, Line9/14 and Can-01 decreased (Dehal et al., 2016), whereas it increased for Gavdos and Macarena. Amorgos and Gavdos had the highest protein content in NS and LS, correspondingly, whereas Gavdos showed the highest mean across sowing periods. Furthermore, significant differences were not detected between the means of NS and LS for HI and HC. Macarena indicated the highest values for HI and HC in both sowing periods, as well as the highest mean values, across sowing periods. Concerning CT, a significant decrease was observed in LS. Macarena, Line 9/14 and Can-01 had the lowest CT within and across sowing periods. Worth noticing is that Amorgos’ CT showed a vast decrease of 22 minutes in LS. Off-season sowing resulted in significant increase of TPC and TTC values (Patel et al., 2013). Especially, Amorgos, which is an Aschochyta blight resistant genotype with low 1000SW, showed the highest values of TPC and TTC within and across sowing periods. Kumar et al., (2013) mentioned the positive correlation between Aschochyta blight resistance and high TPC values, whereas Nikolopoulou et al., (2006) stated the negative correlation between 1000 SW and TTC. Antioxidant activity was mostly affected in different genotype, thus the mean values of the two sowing dates did not differ. Amorgos and Line9/14 had the highest antioxidant activity in NS, while in LS Amorgos had by far the highest value. Moreover, Amorgos presented the highest mean antioxidant activity across sowing dates.



According to the principal component analysis the first three principal components selected explained 77.56% of the total variance (Table 4). PC1 was strongly represented by physical traits, HI and HC with a positive relationship between them (Khattak et al., 2006). TTC and ABTS showed moderate negative loadings (-0.52 and -0.51 respectively) on PC1. The negative loading of TTC and ABTS indicated the negative relationship of these traits with the rest of the traits in PC1. PC1 explained 41.9% of the total variation. PC2 explained a high percentage of the total variance in CT and proteins. PC2 explained 22.41% of the total variation. The variation in bioactive traits was represented in PC3, with high loadings of TPC (0.77) and TTC (0.68) along with a moderate negative loading of 1000 SW (-0.56), indicating the negative relationship between these traits (Nikolopoulou et al., 2006). PC3 explained 12.61% of the total variation.



PCA biplot and cluster analysis, based on the studied traits of the five genotypes under NS and LS, revealed six clusters divided in three categories (Fig 2). Clusters with the same genotypes from different sowing periods (clusters 1, 4, 6) indicating that LS did not affect the performance of these genotypes and underlining their suitability for both NS and LS, clusters 2 and 3 that consisted of single genotypes that were significantly affected by sowing period and weren’t grouped with others, or cluster 4 that was consisted of two different genotypes sown the same period (NS).


       
Cluster 4 grouped the genotypes with the highest values for CT, proteins and a relatively high value for ABTS (PC2). Cluster 6 had the highest values for 1000 SW, HI, HC and SC (PC1). Contrariwise, Cluster 6 had a relatively low value for ABTS and the lowest value for TTC (Fig 2). The negative relationship between ABTS, TTC and the rest of this component’s traits was also confirmed by the correlation analysis (data not shown). Cluster 3 (Amorgos in LS) had the highest values for antioxidant activity, bioactive traits and the lowest value for 1000 SW (PC3). Amorgos’ increased bioactive and antioxidant activity traits, in combination with the decreased CT and good protein content in drought and heat stress conditions indicated the suitability for cultivation in the expected unfavourable environmental conditions. Moreover, Amorgos in LS could be a valuable genetic material for crossing with Macarena, in order to end up in a genotype of great dietary value with high 1000 SW and low CT, that will be able to retain these high quality standards under water scarcity and heat stress.

The study was a part of the Ph.D thesis “Breeding Chickpea under xerothermic conditions for seed yield and quality traits” supervised by the Laboratory of Genetics and Plant Breeding of University of Thessaly and supported by the Institute of Industrial and Forage Crops of Hellenic Agricultural Organization.