SciELO - Scientific Electronic Library Online

vol.36 número1Effect of rainfall regimes on seed production and quality of Avena barbataEarly individual growth of Eucryphia cordifolia and Laurelia sempervirens planted under different competition conditions in south-central Chile índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Servicios Personalizados




Links relacionados


Ciencia e investigación agraria

versión On-line ISSN 0718-1620

Cienc. Inv. Agr. v.36 n.1 Santiago abr. 2009 

Cien. Inv. Agr. 36(1): 77-84. 2009



Genetic divergence of parents and F2 segregation in grain Amaranths


Ram Milan Pandey

Department of Cytogenetics, National Botanical Research Institute, Lucknow-226001, India. Corresponding author: 


Twenty six accessions of grain Amaranths (Amaranthus hypochondriacus), including both indigenous and exotic introductions, were evaluated in the rabi season (winter crop, November-March). Based on D2 analysis, the accessions were grouped into eleven clusters. Clusters I, II, and III had seven, four, and three accessions, respectively; clusters VII, VIII, IX and X had only one accession in each case. The accession in cluster V had the greatest divergence, closely followed by those of clusters IV and I. The maximum and minimum divergences were revealed between clusters VIII and XI and between II and VIL respectively. The pattern of clustering did not show any relationship with geographic origin. In a study of pattern of F2 segregation in 25 crosses involving 10 parents distributed in six clusters, eight crosses showed high breeding potential. The parents involved in these eight potential crosses showed modérate genetic divergence.

Key words: Amaranthus hypochondriacus, D2 statistics, genetic diversity, grain amaranths.


Se evaluaron 26 líneas de amaranto para grano (Amaranthus hypochondriacus), incluyendo líneas silvestres y exóticas, durante la estación del rabí (cultivo de invierno en noviembre-marzo). Según los análisis D2, las líneas se diferenciaron once grupos. Los Grupos I, II y III tuvieron siete, cuatro y tres líneas, respectivamente; los Grupos VII, VIII, IX y X solo tuvieron una línea cada uno. Las líneas en el Grupo V tuvo la mayor divergencia, seguido muy cerca por aquellas de los Grupos IV y I. La divergencia maxima y mínima se obtuvo entre los Grupos VIII y XI, y entre los Grupos II y VII, respectivamente. El padrón de agrupamiento no se relacionó con el origen geográfico de las líneas. Al estudiar el padrón de segragación F2, ocho cruzamientos mostraron un alto potencial de diversidad genética, los que derivaron de 25 cruzamientos que comprometieron 10 padres distribuidos en seis Grupos.

Keywords: Amaranto grano, Amaranthus hypochondriacus, estadística D2, diversidad genética.


Grainamaranths of the gemisAmaranthus (Family Amaranthaceae), which are under-utilized crops, are native of South and Central América (Sauer, 1967). The major producing countries are in South and Central América and in some countries of Asia and África (e.g., China, India, Ethiopia, Kenya and Nepal). The crops are cultivated in India from Kashmir to Arunachal Pradesh. They are protein-rich pseudocereals that hold cultural significance in remote and tribal areas in many parts of India, particularly the Himalayan tract. Grain amaranths in general are high in nutrients, easily digestible, and have sacred ethnic significance. Commercial amaranth production for human consumption is found in México and South American countries (Berghofer and Sochoenlechner, 2002).

Amaranth crops show wide adaptability and good performance under normal conditions and under stress. Grain Amaranths have a high yieldpotential and could become very useful cereal crops given sufficient consumer demand. Genetic divergence analysis using D2 statistics (Mahalanobis, 1936) can be a powerful tool for quantifying the degree of genetic divergence. Minimal information is available regarding the amount of genetic divergence in amaranth grains. Thus, the  present investigation was carried out to measure genetic diversity among grain Amaranth, aiming to identify suitable parents for crop improvement by way of hybridization.

Two field experiments were conducted with the following objectives: (i) Experiment I was conducted to evaluate the performance of the 26   accessions of Amaranthus hypochondriacus L. with respect to yield and other related traits, and to assess genetic divergence among them. (ii) Experiment II aimed to study the pattern of F2 segregation in 25 crosses involving 10 parents, and to determine the respective relationships with genetic diversity of the parents.

Materials and methods

Experiment I had 26 accessions (Table 1) consisting of indigenous and exotic collections. Experiment II had 25 crosses (F2s), involving ten accessions, used along with their parents (Table 2). The experiments were conducted in randomized block designs with three replications during the rabi season (Winter crop, November-March) 2005-2006. In both experiments, each plot consisted of four rows, 3.0 m long, with a spacing of 40 x 15 cm.

Observations in Experiment I were recorded from ten randomly chosen plants in the two middle rows for seven quantitative characteristics (Table 1). Sample means were used for multivariate analysis (Mahalanobis, 1936). Group constellations were made following Tocher's method (Rao, 1952), and canonical analysis was done following Johnson and Wichern (1988).

In Experiment II, ten randomly chosen competitive plants from the middle rows of the parents and 20 plants from crosses were sampled for recording observations on five traits, i.e., plant grain yield, panicles per plant, inflorescence length, test weight of 1000 grains, and branches per plant (Table 3). The variations of individual F2s for different characteristics were analyzed in terms of range, variance, and phenotypic and genotypic coefficients of variation. The segregation pattern of the crosses was analyzed in terms of the frequency of positive transgressive segregates (FPTS) and the magnitude of transgression. The average positive transgression of a cross was measured as the difference of the mean of the PTS and the better parent mean. The relationship of the parameters of variation for yield in F2 with parental diversity was measured by D2 statistics and examined by computing correlation coefficients among all parameters of variability. Furthermore, the performances of the 10 top yielding F2s were examined in relation to parental diversity, classified as high, medium or low, based on their performance for each of the five traits, including yield.

Results and discussion

The accessions differed significantly with re-spect to the seven traits (days to flowering, plant height, inflorescence length, panicles and branches per plant, test weight grain, and yield per plant), indicating substantial variation in the material (Table 1). The low level of grain yield per plant ranging from 109.2 to 237.5 g in the AG-23 and AG-67 accessions, respec-tively, is possibly due to the thermo- and photo-period sensitivities of the genotypes under the short day conditions that characterize the rabi season. Considering grain yield and its components, such as number of panicles per plant, inflorescence length and test weight, five accessions (AG-67, AG-101, AG-114, AG-21 and EC-338810) were found to be superior with respect to yield. Among these, three accessions, AG-21, AG-67 and AG-114, are grown in India, partic-ularly in Uttar Pradesh state, while AG-101 is cultivated in México (Table 1).

On the basis of the general means () and standard errors (SE) of each of the seven characteristics, the genotypes were grouped into three categories: high (H) for values more than (+SE), medium (M) for values (±SE), and low (L) for values less than (-SE). This classification also indicated considerable genetic variability. The D2 values ranged from 0.00 to 876.45 (Table 4). Following Tocher's method, the 26 accessions were grouped into eleven clusters (I - XI), six of which were multi-accession clusters (Table 1). These multi-accession clusters included accessions from different countries and states of India, irrespective of geographical distances. On the other hand, the accessions from the same country and state were included in different clusters. In the present study, the three standard accessions AG-21 (a field selection from Barabanki, U.P), AG-67 (a selection from the cross AG-24 x AG-21) and AG-114 (a selection from the interspecific cross of A. hybridus xA. hypochondriacus (AG-20) were included in three different clusters (IV, I and II). Thus, the clustering pattern was not related to geographical origin. Rather, the clustering pattern was related to selection pressure, which played a large role in determining the genetic closeness/divergence among the accessions. Similar conclusions were made by Murty et al. (1965), Murty and Arunachalm (1966), Bhatt (1970), Gaur et al. (1978), Kumar (1977), Narayan and Macfield (1976) and Yadav et al. (2007), who suggested that selection may cause greater diversity than geographical distance.

In multivariate analysis following the D2 technique, the major factors contributing to genetic divergence were grain yield per plant (48.0%), followed by panicles per plant (23.08%) and inflorescence length (11.69%). According to canonical analysis in Root 1 (Z1), the important characteristics responsible for genetic divergence in the major axis of differentiation were grain yield per plant (0.7865) followed by panicles per plant (0.4645); and in Root 2 (Z2) were panicles per plant (0.6184) followed by inflorescence length (0.3001). These results are in complete agreement with the characteristic contributions assessed by D2 analysis.

Based on the seven characteristics, canonical analysis confirmed the clustering pattern obtained by D2 statistics. The compositions of the clusters and their respective dispositions remained almost the same (Figure 1).

The average inter-cluster values ranged from 6.51 to 29.60, while intra-cluster distances var-ed from 0.000 to 6.50. The average D2 values within and between clusters revealed that intra-cluster divergence showed a maximum in cluster V (D2= 42.26), which included two accessions. The inter-cluster distance had a maximum between clusters IV and VII (876.45, Table 4), indicating a greater genetic divergence between the accessions belonging to these categories. The minimum inter-cluster difference was observed between clusters II and IX (42.40, Table 5), indicating that accessions of these clusters had a maximum number of common gene complexes.

Significant differences among the means and variances for individual crosses (F2s) were found with respect to grain yield and four of its components, thus offering ampie scope for selection for yield improvement. On the basis of mean, variance, heritability and genetic advance for yield, five crosses (AG-16xAG-21, AG-21xAG-24, AG-21xAG-30 AG-21xAG-45 and AG-21xAG-198) showed high breeding potential for advancement through pedigree breeding. Furthermore, analysis of the nature of the magnitude of transgressive segregation and mean performance of the top 10% yielding plants, revealed high breeding potential of some crosses, suggesting that selection for yield in F2 and later generations would result in substantial improvement (Table 2). Average positive transgressions in these crosses were high, ranging from 0.37 in AG-21xAG-24 to 0.88 in AG-21-xAG-30.

Significant positive correlations among all parameters of variability and transgressive segregation were observed, except between the frequency of positive transgressive segregates and the mean of positive transgressive segregates (Table 5). On the other hand, negative correlations of parental diversity were not significant with all parameters, except with the mean of positive transgressive segregates. However, these negative associations with F2 mean, frequency of positive transgressive segregates and mean of the top 10% plants were of modérate magnitude, suggesting that crosses among parents with wide diversity may not generate potential hybrids. Furthermore, the D2 values between the parents of the five potential crosses ranged from 905 in AG-21xAG-24 to 12067 in AG-21xAG-198, and were consideredtobe modérate. Thus, it can be concluded that a cross has a higher potential when the parents are moderately diverse, rather than less or highly diverse (Chauhan and Singh, 1982, and Arunachalam and Bandyopadhyay, 1984).

The breeding potential of crosses in relation to parental performance was assessed by classifying the 26 accessions in Experiment I into high, medium and low on the basis of performance for yield and four of its components. The crosses were classified as high x high, high x medium, high x low, medium x medium, medium x low and low x low combinations. The performances of the 10 top yielding F2s were examined in relation to combination of parental performance for each of the five traits. It was observed that seven out of the 10 top yielding crosses (F2s) showed high x low combinations for yield; the remaining crosses were either high x medium or medium x low combinations (Table 4). However, high x high, medium x medium and low x low combinations did not give crosses with high yield potential. When the parental combinations of these crosses were considered for all five traits, it was found that 22 out of 50 crosses were either high x medium or medium x low, 19 high x low, and the remaining nine were of similar parental performance. Thus, the results indicated that high x medium and medium x low crosses are better than high x high, medium x medium, low x low and even high x low crosses for yield improvement. This supports the results obtained for parental divergence as measured by D2 statistics.

From the results of the present study, it may be inferred that accessions with moderate genetic divergence, as measured by D2 values or by per se performance with respect to yield and its components, are likely to produce potential crosses. Thus, while choosing parents in a hybridization program for yield improvement, assessment of yield performance and important yield components of the accessions may be considered as a simpler and more convenient method than complicated statistical analysis for measuring genetic divergence by D2 statistics.


The authors would like to thank R. Tuli, Director, National Botanical Research Institute, Lucknow, for supporting these experiments.



Arunachalam, V, and A. Bandyopadhyay. 1984. Limits of genetic divergence for occurrence of heterosis-experimental evidence from crop plants. Indian. J. Genet. 44:548-554.

Bhatt, G.M. 1970. Multivariate analysis approach to selection of parents for hybridization aiming at yield improvement in self-pollinating crops. Aust. I Agric. Res. 21:1-7.

Berghofer, E., and R. Schoenlechner. 2002. Grain amaranth. Pages 219-260. In: PS. Belton, and J.R.N, Taylor (eds.). Pseudocereals and Less Common Cereals. Springer-Verlag, Berlin, Germany.

Chauhan, V.S., and B.B. and Singh. 1982. Heterosis and genetic variability in relation to genetic divergence in Soybean. Indian, J. Genet. 42:324-328.

Gaur, PC, PK. Gupta, and H. Kishore. 1978. Studies on genetic divergence in potato. Euphytica 27:361-368.

Johnson, R., and D.W. Wichern. 1988. Applied Multivariate Statistical Analysis. Second ed.. Prentice Hall, NY, USA. 607 pp.

Kumar, N. 1997. Genetic diversity among chickpea accessions. Indian J. Genet., 57:87-90.

Mahalanobis, PC. 1936. On the generalized distance in statistics. Proc. Nati. Inst. Sci. India 2:49-55.

Mraty, B.R., and V. Arunachalam. 1966. The nature of genetic divergence in relation to breeding system in crop plants. Indian J. Genet. 26:188-198.

Mraty, B.R., J.B.L. Mathur, and V. Arunachalam. 1965. Self incompatibility and genetic divergence in Brassica campestris var. Brown Sarson. Sankhya B. 27:272-278.

Narayan, P.K.J., and AJ. Macfield. 1976. Adapted response and genetic divergence in a world germplasm of chickpea. Theor. Appl. Genet. 47:104-110.

Rao, C.R. 1952. Advanced Statistical Methods in Biometrical Research. John Wiley and Sons, NY, USA. 390 pp

Sauer, J.D. 1967. The grain amaranths and their relatives. A revised taxonomic and geographic survey. Ann. Missouri. Bot. Gar. 54:103-137.

Yadav, H.K., S. Shukla, and S.P Singh. 2007. Genetic divergence in parental genotypes and its relation with heterosis, F1 performance and general combining ability (GCA) in opium poppy (Papaver somniferum L.). Euphytica 157:123-130.


Received 04 June 2008. Accepted 11 November 2008. 

Creative Commons License Todo el contenido de esta revista, excepto dónde está identificado, está bajo una Licencia Creative Commons