Introduction

Underutilized or neglected crops are ancient plant species that have been used by local communities for a long time and have been maintained by cultural preferences and traditional practices (^{Mayes et al., 2011}; ^{Bhargava and Srivastava, 2013}). These crops have played an important role in strengthening the food security, nutrition, and income generation of the rural poor in several countries worldwide. However, these crops have remained inadequately characterized and neglected in terms of research, extension services and conservation (^{Magbagbeola et al., 2010}; ^{Mayes et al., 2011}). A number of other terms, such as orphan, minor, abandoned, lost, underused, local, promising, traditional, forgotten and alternative crops, are often used for these less utilized species (^{Padulosi and Hoeschle-Zeledon, 2004}). The cultivation of these traditional crops has declined over time, yet many of these wild and underutilized plants have the potential for more widespread use and could contribute to agricultural diversification, income generation, food provision and nutritional security (^{Vietmeyer, 1986}; ^{Chivenge et al., 2015}; ^{Bhargava and Srivastava, 2013}). These crops are more important than other commercially cultivated crops since they can be cultivated with low inputs and therefore provide economic and environmental benefits. It can be safely said that the new century will see renewed interest and utilization of these underutilized crops. In the last few decades, chenopods have shown potential as food crops for agricultural diversification due to their climate adaptability and nutritional superiority (^{Bhargava et al., 2010}; ^{Bhargava and Ohri, 2015}).

The genus Chenopodium (family: Amaranthaceae), known as the goosefoot genus, comprises approximately 250 species (^{Giusti, 1970}) and includes mostly facultatively autogamous annuals distributed in temperate regions of the Americas, Asia and Europe (^{Wilson, 1990}; ^{Jellen et al., 2011}). Of the several species of the genus that have been used, only three, C. quinoa (2n=36), C. pallidicaule (2n=18) and C. berlandieri ssp. nuttalliae (2n=36), have been cultivated as food crops (^{Bhargava et al., 2010}; ^{Zurita-Silva et al., 2014}). However, the leaves and tender stems of several other species of the genus are used as food and fodder (^{Facciola, 1990}; ^{Moerman, 1998}; ^{Partap et al., 1998}; ^{Bhargava et al., 2007a}, ^2008a, ^2008b; ^{Fuentes and Paredes-Gónzalez, 2015}). Chenopods are a rich source of nutrients such as protein, vitamins and a wide range of minerals (^{Prakash et al., 1993}; ^{Bhargava et al., 2007a}, ^2007b; ^{Miranda et al., 2012}). Recently, chenopods have been on the forefront of research, especially as an alternative crop for agriculturally marginal lands and stressful environments in many parts of the world (^{Bhargava et al., 2003a}, ²⁰⁰⁶; ^{Fuentes et al., 2012}; ^{Bazile et al., 2016}).

Although the crop is nutritionally very important, information on the genetic improvement of vegetable chenopods is scarce (^{Bhargava et al., 2003b}, ^2003c, ^2007a, ²⁰¹⁹) and is restricted to a single species, C. album. Improvement of foliage requires knowledge of the extent of genetic variability in the available germplasm, the interrelationships between different traits, the extent of the environmental effects on these characters, trait heritability and the potential genetic gain of the material (^{Bhargava et al., 2007b}; ^{Litrico and Violle, 2007}; ^{Pistorale et al., 2008}). Surprisingly, this information is completely absent for numerous species of Chenopodium. To fill this gap, the present investigation was conducted to obtain in-depth knowledge of the different selection parameters for different quantitative and qualitative traits of vegetable Chenopodium.

Materials and Methods

The experiment was carried out in the experimental field of the CSIR-National Botanical Research Institute, Lucknow (26.5°N and 80.5°E), situated at an altitude of 120 meters above mean sea level. The region has a subtropical climate characterized by long, hot summers, cool winters and a marked rainy season between these two seasons. In the northern part of India, Chenopodium is generally grown as a winter crop, when the temperature ranges from 2-29 °C. The experimental material for the present study comprised 13 germplasm lines of Chenopodium spp. (Table 1), which included 7 lines of C. giganteum, 2 of C. murale and 1 each of C. bushianum, C. strictum, C. ugandae and C. opulifolium. These lines, obtained from different sources, have different ploidy levels and have been maintained for several years at the research institute.

The germplasm lines were sown in a randomized block design with 3 replications. The plot size for each treatment was 4 m² with 6 rows per plot. The plant-to-plant distance was maintained at 30 cm. Irrigation was provided as required, and hand weeding was periodically carried out. No chemical fertilizer or pesticide was used. A total of 4 cuttings were performed. The first cutting was carried out the 4^th week after sowing, and thereafter, three more cuttings were performed at intervals of 15 days. Foliage yield was recorded on a plot basis for different cuttings in kg/plot and then converted to t ha^-1. Observations were recorded for each cutting of 10 randomly selected plants in each replication for five morphological traits, namely, plant height (cm), leaves per plant, leaf size (cm²), branches per plant and stem diameter (cm). Fresh leaves of each cutting were analyzed separately for three qualitative traits, namely, moisture (%), carotenoid content (mg 100 g^-1) and protein content (%). Qualitative data were collected from randomly selected samples for all 4 cuttings separately. Moisture content was estimated as the ratio of fresh leaf weight to 100 °C dry weight. The extraction and estimation of leaf carotenoid and leaf protein contents were carried out per the methods suggested by ^{Jensen (1978)} and ^{Lowry et al. (1951)}, respectively.

The data were subjected to statistical analysis of variance according to ^{Panse and Sukhatme (1978)}. The phenotypic and genotypic coefficients of variation (PCV and GCV), heritability in broad sense (%) and genetic advance (%) were calculated according to the methods outlined by ^{Singh and Chaudhary (1985)} and ^{Johnson et al. (1955)}.

Results and Discussion

The analysis of variance (ANOVA) revealed significant differences in the germplasm for individual cuttings and traits, except for branches per plant and moisture content in the IV^th cutting (Table 2). The foliage yield was variable, ranging from 1.07-2.66 t ha^-1, with an average yield of 2.00±0.15 (Table 2). Plant height ranged from 12.89-22.40 cm with an average of 15.97±1.68, while branches per plant ranged from 7.76-12.65, with a mean of 9.39±0.60. The leaf size and number of leaves per plant ranged from 16.35-35.90 cm² and 10.20-28.94, with average values of 27.94±2.96 and 17.28±3.25, respectively. The stem diameter ranged from 0.37-0.51 cm with a mean of 0.46±0.03. Among the quality traits, carotenoid and protein content ranged from 12.12-16.24 mg 100 g^-1 and 3.26-3.98%, with average values of 13.47±0.56 and 3.70±0.09, respectively. Leaf moisture was in the range of 78.23-85.48%, with an average of 81.00+1.22.

The range and mean of individual cuttings for leaves per plant increased with increasing cuttings and were highest in the IV^th cutting (Table 2), while the plant height, leaf size, protein and foliage yield values increased with each cutting until the III^rd cutting. However, the values for these traits decreased after the III^rd cutting.

The present study clearly indicates that the species assessed can serve as potential vegetable crops due to their high biomasses and ample amounts of proteins and carotenoids. The protein content, though lower than that of the commonly used foliage crop C. album (^{Prakash et al., 1993}; Bhargava et al., 2007a), was higher than those reported for amaranth (^{Shukla et al., 2003}) and spinach (^{Kuti and Kuti, 1999}). Likewise, the material contained appreciable amounts of carotenoids, with levels higher than that of C. album. These species seem to have more regeneration capacity in comparison to C. album, since 4 cuttings were performed for the species in this study, whereas a maximum of 3 cuttings have been reported for C. album by ^{Bhargava et al. (2003a)}. Thus, there is a need to further explore the potential of these underutilized species of Chenopodium for use as fodder crops and for human consumption.

The estimates of variance, phenotypic coefficient of variation (PCV), genotypic coefficient of variation (GCV), heritability in broad sense and genetic advance are shown in Table 2. The PCV had higher estimates than the corresponding GCV for all the characteristics, and in most cases, the two values differed only slightly. The small difference between PCV and GCV for all the traits except stem diameter indicates that the variability was primarily due to genotypic differences. Foliage, followed by plant height and leaf size among the morphological traits, and carotenoid and protein content among the quality traits exhibited high PCV and GCV values, which indicates the potential for improvement of these traits through selection to enhance the potentiality of foliage yield. A high coefficient of variability for foliage yield has previously been reported in other foliage crops, such as vegetable amaranth (^{Shukla and Singh, 2000}) and C. album (^{Bhargava et al., 2003b}). However, the lower estimates for one or both of these parameters observed for the rest of the traits implies that the chances of obtaining substantial gain under selection are likely to be lower for these traits.

It is difficult to determine the amount of the variation that is heritable merely based on the extent of the coefficient of variation values. Knowledge of the heritability of a trait is important because it indicates the amount of advance to be expected through phenotypic selection (^{Robinson et al., 1949}; ^{Burton and DeVane, 1953}). Heritability, the proportion of the individual differences in a trait due to genetic difference, has been extensively utilized as an effective tool to assess the degree to which a trait is transmitted from parent to progeny (^{Piepho and Möhring, 2007}). This important parameter in quantitative genetics also indicates the relative importance of heredity and environment in the expression of these traits. Narrow-sense heritability (designated as h²) is the ratio of additive genetic variance to the total phenotypic variance and includes only the additive effects of variation (^{Nishio et al., 2014}). In the present investigation, the heritability estimates were high for all the traits in most cuttings individually as well as on a pooled basis, which suggests that all of the variability in traits comes from genetic differences. High heritability values have also been obtained for different foliage cuttings of lamb's quarters (^{Bhargava et al., 2003a}; ^{Basavaraj et al., 2018}) and Amaranthus tricolor (^{Shukla et al., 2004}).

High heritability alone does not guarantee a large improvement from selection unless sufficient genetic gain attributable to additive gene action is present (^{Mora and Saavedra, 2012}). Genetic advance in a trait is the product of the heritability and selection differential, expressed in units of standard deviation, and has an added advantage over heritability as a guiding factor in a selection program where improvement in particular traits is desired (^{Xu et al., 2017}; ^{Li et al., 2018}). The expected genetic advance, expressed as a percentage of the mean, was lowest for leaves per plant, moisture and stem diameter, which indicates the presence of non-additive gene effects; hence, these traits are unlikely to respond favorably to selection. In contrast, foliage yield, carotenoid content and protein content had high values for genetic advance (72.00, 44.17 and 43.78%, respectively). These traits also had high heritability and high coefficients of variability, indicating that the genotypic variance for these characters was due to additive gene effects. Hence, the selection based on the phenotypic performance for these characters would be beneficial for achieving the desired gain.

Conclusion

From the present study, it was concluded that many species of Chenopodium could be utilized as vegetables with high protein and carotenoid contents. Significant enhancements in foliage yield and foliage quality traits can be achieved through rigorous selection of plant types based on the component traits.