Variability , heritability and genetic advance in upland cotton ( Gossypium hirsutum L . )

The analysis of variance study indicated the presence of significant difference among all the traits in Gossypium hirsutum accessions. The highest phenotypic coefficient of variation (GCV) and genotypic coefficient of variation (GCV) were recorded by seed index, plant height, lint index and boll weight. Genotypic co-efficient of variation had a similar trend as PCV. High heritability along with high genetic advance was observed in traits viz., number of sympodia per plant, single plant yields, seed index and micronaire value. The combinations of high heritability with high genetic advance will provide a clear base on the reliability of that particular character in selection of variable entries. Based on per se performance, the accessions MCU5, TCH1715, TCH1716 and G cot 16 were identified as potential donors for single plant yield (g), number of bolls per plant, 2.5% span length (mm) and bundle strength (g/tex). So these accessions may be utilized for crossing programme to improve a particular character in crop improvement.


INTRODUCTION
Cotton is an important fibre crop of global importance which is grown in tropical and subtropical regions of more than 60 countries of the world.Despite threat from synthetic fibre or manmade fibre, cotton retains its reputation as "queen of the fibre plants"".For multiple uses of lint and byproducts, cotton is also referred to as "white gold".In any crop improvement programme, knowledge on nature of gene action and inheritance of traits is essential so as to choose a suitable breeding methodology in crop improvement (Vineela et al., 2013).Development of an effective breeding programme depends on the existence of genetic variability for various economic characters in the gene pool.Selection is effective only when there is enough magnitude of variability in the breeding population.An understanding of precise magnitude of variability present in a population is important in formulating the most appropriate breeding technique for improvement of various characters.The present investigation was carried out with 54 Gossypium hirsutum lines of diverse origin to estimate their per se performance, variability, heritability and genetic advance on the genetic architecture of 16 yield, yield components and fibre quality traits.

MATERIALS AND METHODS
The study was conducted in the Department of Cotton, Tamil Nadu Agricultural University, Coimbatore during winter 2010.Fifty four G. hirsutum cotton genotypes were planted in randomized block design with two replications.Uniform spacing of 90 x 45 cm and all the recommended field operations were carried out.In each replication five competitive plants were randomly selected and observations were recorded for 16 characters viz., days to 50% flowering, plant height (cm), internode length (cm), number of sympodia/plant, number of ovules/plant, number of bolls/plant, boll weight (g), number of seeds/plant, seed setting percentage, seed cotton yield/plant, lint index, seed index, ginning outturn (%), 2.5% span length (mm), bundle strength (g/tex) and fibre fineness.
Analysis of variance was carried out statistically utilizing the mean values (Panse and Sukhtame, 1995).The phenotypic and genotypic coefficient of variation was estimated using the formula suggested by Burton (1952) and expressed in percentage.The phenotypic and genotypic variances were calculated by utilizing the  (Lush, 1940).Heritability, expected genetic advance and genetic gain in the broad sense was calculated according to the formula suggested by Johanson et al. (1955).

RESULTS AND DISCUSSION
The per se performance of yield and fibre quality characters were recorded on 54 germplasm accessions and the range of variations observed in respect of all the 16 traits studied are presented in the Table 1.Based on per se performance, the accessions such as MCU 5, TCH1715, TCH1716 and G cot 16 in G. hirsutum were identified as potential donors which recorded highest mean values for single plant yield, number of bolls per plant, 2.5% span length (mm) and bundle strength (g/tex).These accessions may be exploited for further improvement of the above traits by breeding programmes.Regarding the identification of donor for specific trait, the highest performer of that particular trait can be considered.Among the accessions, TCH 1715 (5.5 g of boll weight), RB 488 (46.1% of GOT) were expressed the highest per se for that particular traits.The culture TCH 1710 (34 mm, 23.1g/tex) recorded the highest 2.5% span length.So these accessions may be utilized for crossing programme to improve that particular character in crop improvement.Anjali was found to be compact type which can be utilized in development of genotype suitable for high density planting.The analysis of variance showed highly significant differences among genotypes for all the characters studied (Table 2) and infers existence of considerable genetic diversity among genotypes.Phenotypic variance, genotypic variance, phenotypic coefficient of variation, genotypic coefficient of variation, heritability in broad sense and genetic advance as percentage of mean which were estimated for 16 characters are shown in Table 3.The knowledge of nature and magnitude of variability available in the genotypes for different characters is an important prerequisite for making simultaneous selection over more number of characters to bring remarkable improvement in cotton.The analysis of variance study indicated the presence of significant difference among all the traits in the accessions.The heritable (genotypic) variation is usually masked by non-heritable variation creating   difficulty in exercising selection.Hence it becomes necessary to partition overall variability into heritable and non-heritable components to enable the breeders to plan for proper breeding programme.The plant height recorded the highest value for phenotypic variance (342.6) and the single plant yield recorded highest genotypic variance (219.9).The micronaire value exhibited the lowest phenol-typic variance (0.2) and in case of genotypic variance, the traits viz., inter node length, boll weight and micronaire value had recorded the lowest value (0.1).The co-efficient of phenotypic and genotypic variance were calculated for all the characters under study.The PCV ranged from 4.5 (days to 50% flowering) to 25.9% (single plant yield).The highest PCV was followed by seed index (17.7%),plant height (17.0%), lint index (17.0%)and boll weight (15.4%).Genotypic co-efficient of variation had a similar trend as PCV.The range varied from 3.5 (Days to fifty percent flowering) to 25.8 (Single plant yield).The maximum genotypic coefficient of variation (GCV) was observed for single plant yield (25.8) followed by seed index (13.5),number of sympodia per plant (12.1), lint index (11.2) and number of bolls per plant (10.9), plant height (10.2).The lowest PCV (4.5%) and GCV (3.5) values was observed in days to 50% flowering.In the present study, there was a close corresponddence between phenotypic and genotypic variance for days to 50% flowering, inter node length, boll weight, single plant yield, lint index, micronaire value and 2.5% span length indicating less environmental influence.But plant height, number of sympodia per plant, number of bolls per plant, number of seeds per boll, seed setting percentage and ginning outturn showed higher variation indicating the influence of environment on these characters.Since the variations are influenced by the magnitude of the units of measurements of different traits, a measure of coefficient of variation which is independent of the unit of measurement is more useful in comparing between populations.In G. hirsutum accessions PCV was higher than the GCV for all characters.

Source of variation
From this, we can understand that the environment plays a major role on expression of all these traits leading to increase in the PCV more than GCV.
The highest PCV and GCV estimates were recorded for single plant yield indicating the presence of significant genetic variability in this character.
Selection pressure can be applied on this character to isolate promising genotypes.Similar observations in cotton was reported by Dheva and Potdukhe (2002) and Preetha and Raveendran (2007).
Moderate PCV and GCV estimates were noticed for the characters such as plant height, number of sympodia per plant, number of bolls per plant, lint index and seed index.Girase and Mehetne (2002) and Harshal (2010) also reported the moderate PCV and GCV in various traits with the suggestion that these characters can be improved though rigorous selection.The characters such as days to 50% flowering, 2.5% span length and micronaire value exhibited low PCV and GCV which indicated that the breeds should go for source of high variability for these characters to make improvement.Similar suggestion were given by Kowsalya and Raveendran (1996), Do Thi Ha An et al. ((2006) in their conclusion.In a population, the observed variability is a combined measure of genetic and environment causes, where as the genetic variability is the only estimate heritable from generation to generation.However, a measure of heritability alone does not give an idea about the expected gain in the next generation but it has to be considered in conjunction with genetic advance.The traits which expressed high heritability and high genetic advance as percentage of mean could be used as a powerful tool in selection process.According to Panes and Sukhatme (1995) such characters were found to be governed by additive genes and had minimum environment influence.The heritabi-lity estimates ranged from 23.5 (internode length) to 89.1% (single plant yield).
The high heritability estimates of 89.1% were recorded by single plant yield followed by 2.5% span length (96.2 %) and bundle strength (91.8%)where as the low heritability was observed in inter node length (23.5%).
The high heritability was registered in the traits viz., number of sympodia per plant, single plant yield, seed index, 2.5% span length and bundle strength.The inter node length exhibited low heritability in this investigation.For efficient selection, we cannot completely depend on heritability alone.The combinations of high heritability with high genetic advance will provide a clear base on the reliability of that particular character in selection of variable entries.The genotypic advance as percent of mean for 16 traits ranged from 7.0 to 27.3%.The higher genetic advance as percent of mean was recorded by single plant yield (27.3%) followed by seed index (27.1 %), number of sympodia per plant (26.2 %), number of bolls per plant (21.8 %).The lowest value of 7.0% was observed in the trait days to 50% flowering.High heritability along with high genetic advance was observed in traits viz., number of sympodia per plant, single plant yields, seed index and micronaire value in G. hirsutum (Do Thi Ha An et al., 2008).These traits are highly reliable during selection.High heritability combined with moderate genetic advance was found in the 2.5% span length.It was in accordance with of Muhammad et al. (2004).Among the study materials some of the accessions were identified as potential donors for the improvement of different characters (Table 4).The accessions with high mean performance are generally preferred for all the traits except days to 50% flowering, since earliness is the preferred attribute and early flowering was taken into consideration.From the results of the present study, it can be concluded that direct selection can be done for most of the yield attributing traits since it exhibited high genetic variability and high range of variation.A high PCV over GCV for the characters studied indicated that environment influences the expression of these characters under study.High genetic advance, genetic gain and heritability were recorded for number of sympodia per plant, single plant yield, seed index, micronaire value and 2.5% span length indicated that selection can be resorted for the improvement of these characters in the future crop improvement programmes.

Table 1 .
Mean and range performance for different characters among G. hirsutum accessions.

Table 2 .
Analysis of variance for the different characters among the germplasm accessions of G. hirsutum.

Table 3 .
Components of variance for yield and fibre quality characters of G. hirsutum accessions.